Download Chemical (Elemental) Analysis - Fritz-Haber

Modern Methods in Heterogeneous Catalysis Research
29.10.2004
Chemical (Elemental) Analysis
Raimund Horn
Dep. of Inorganic Chemistry, Group Functional Characterization, Fritz-Haber-Institute of the Max-Planck-Society
Outline
1.
Introduction
1.1
Concentration Ranges
1.2
Accuracy and Precision
1.3
Decision Limit, Detection Limit, Determination Limit
2.
Methods for Quantitative Elemental Analysis
2.1
Chemical Methods
2.1.1 Volumetric Methods
2.1.2 Gravimetric Methods
2.2
Electroanalytical Methods
2.2.1 Potentiometry
2.2.2 Polarography
Outline
2.6
Spectroscopic Methods
2.6.1
Atomic Emission Spectroscopy
2.6.2
Atomic Absorption Spectroscopy
2.6.3
Inductively Coupled Plasma Mass Spectrometry
2.6.4
X-Ray Fluorescence Spectroscopy
2.6.5
Electron Probe X-Ray Microanalysis
3 Specification of an Analytical Result
4 Summary
5 Literature
Backup Slide
The Analytical Process
strategy
solution
problem
characterization
smaller
sample
measurement
sample for
measurement
evaluation
measurement
data
analysis
preparation
chemical
information
data
diminution
representative sampling
object of investigation
sample
analytical
result
Backup Slide
The Analytical Process
strategy
interpretation
problem
characterization
smaller
sample
measurement
sample for
measurement
evaluation
measurement
data
analysis
preparation
chemical
information
data
diminution
representative sampling
object of investigation
sample
analytical
result
Backup Slide
Analytical Problems in Science, Industry, Law…
Science
Industry
Law
example: Determination by Xray Absorption Spectroscopy of
the Fe-Fe Separation in the
Oxidized
Form
of
the
Hydroxylase
of
Methane
Monooxygenase Alone and in
the Presence of MMOD
example: manufacturer of iron Shall I by this iron ore or not ?
example: Is the accused guilty
or not guilty ?
Rudd D. J., Sazinsky M. H., Merkx
M., et al. Inorg. Chem. 2004, 43
(15), 4579-4589, 2004
Motivation
fundamental interest
future developments
Motivation
money
precision (e.g.)
53 ± 5wt% vs. 53 ± 1wt%
accuracy (e.g.)
53 ± 1wt% vs. 51 ± 1wt%
Motivation
human fate
murder conviction
genetic fingerprint fits – lifelong
prison
genetic fingerprint does not fit acquittal
1. Introduction
1.1 Concentration Ranges
concentration ranges in elemental analysis
side
main
components components
trace components
1 ppt
1 ppb
1 ppm
1‰
1%
ppt (w/w)
ppp (w/w)
ppm (w/w)
‰ (w/w)
% (w/w)
10-12 g/g
10-9 g/g
10-6 g/g
10-3 g/g
10-2 g/g
pg/g
ng/g
µg/g
mg/g
0.01 g/g
ng/kg
µg/kg
mg/kg
g/kg
10 mg/kg
100 %
1. Introduction
1.1 Concentration Ranges
mass ratio
10-2
10-3
1%
10-6
1‰
10-9
10-12
1 ppb
1 ppt
0.27 L
2.7 L
2 700 000 L
2 700 L 1 ppm
2 700 000 000 L
concentration ranges in trace analysis
2.7 g
1. Introduction
1.2 Accuracy and Precision
accuracy:
precision (uncertainty):
Ø can
be
defined
as
the
consistency between the mean
of an analytic result and the
value held as the true value
Ø describes in a positive (negative)
manner the influence of random
errors on an analytic result
Ø describing quantities:
Ø has to be verified before each
analysis by means of standards
s=
Ø describing quantities:
∑ (x
main and
side
component
analysis
wrong
results
Gaussian
distributed
errors!
i
errorx = ± x − x̂'
± x − xˆ '
relative errorx =
⋅ 100
xˆ '
− x)
2
i
v = s2 =
n−1
∑ (x
− x)
2
i
i
n−1
trace and
ultra trace
analysis
1. Introduction
1.3 Decision Limit, Detection Limit, Determination Limit
Ø the lower the concentration of
an analyte, the more difficult
becomes
its
qualitative
detection and its quantification
Ø stochastic errors become more
and more pronounced
Ø the decision limit and the
detection limit are lower limits
characterizing the performance
of a method to discriminate a
true value from a blank value
Ø the
determination
guaranties the quality
quantitative
result
specific)
yk = y B +
t f ,α ⋅ s B
n
limit
of a
(user
⇒ y k = y B + bx NG ⇒ x NG =
for α = β x EG = 2 ⋅ x NG
t f ,α ⋅ s B
b⋅ n
DIN 32645
2. Methods for Quantitative Elemental Analysis
2.1 Chemical Methods
2.2 Electroanalytical
Methods
2.3 Spectroscopic Methods
Ø atomic emission
Ø atomic absorption
Ø volumetric methods
Ø potentiometry
Ø gravimetric methods
Ø polarography
Ø inductively coupled
plasma mass
spectrometry
Ø X-ray fluorescence
Ø electron probe X-ray
microanalysis
2.1 Chemical Methods
2.1.1 Volumetric Methods - Principle
principle of volumetric methods
evaluation of the result
p = VB ⋅ k ⋅ N ⋅
100
in %
me
p = percentage of the analyte in the sample
VB = volume in ml dropped in from the burette
A+B→C+D
k = stoichiometric factor (mg analyte/ml)
N = correction factor for the theoretical value k
e = weighted sample (in mg)
2
Ø fast, complete and stoichiometric well
defined reaction
Ø the extend of the reaction can be
monitored (e.g. pH, electric potential)
Ø the point of equivalence can be
detected precisely (e.g. sudden color
change, steep pH or potential change)
2
2
 σ m e   σ VB   σ N 
σP
 + 
 + 
= 

p
m
V
N


 e   B 
inserting typical values
0.1% <
σP
< 1%
p
advantage: high precision!!!
2.1 Chemical Methods
2.1.1 Volumetric Methods – Acid Base Titration
acid – base reaction:
titration curve:
τ=
HnA + nBOH V An- + nB+ + nH2O (zA=n, zB=1)
z B ⋅ C B ⋅ VB
K
V +V  Kw

= − pH a
+ S
⋅  − pH − 10− pH  K w = [ H + ][OH − ]
z A ⋅ C A ⋅ V A 10
+ K a C S ⋅ VS  10

strong acid – strong base
weak acid – strong base
different Ka – strong base
- H+
+ H+
Ø choice of the indicator depends on Ka
pH = pK a , Ind ± 1
Ø very weak acids (bases) can be titrated in non-aqueous solvents as liquid
ammonia or 100% acetic acid
2.3 Chemical Methods
2.1.1 Volumetric Methods - Complexometry
complex formation:
nL + Mm+ V MLnm+
most used ligand - EDTA
H4-nYn- + Mm+ V [MY](m-4) + (4-n)H+ (zA=1, zB=1)
[M
m+
[ MY ( m − 4 ) ]
]=
K f ⋅ [ EDTA]total ⋅ α Y 4−
complex formation only with Y4⇒ buffer required
indication: HI2- + Mm+ V MI(m-3) + H+
e.g. Eriochromschwarz T
applicable for the determination of:
Mg2+, Ca2+, Sr2+, Ba2+ at pH 8-11
Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+, Al3+, Pb2+, VO2+ at pH 4-7
Bi3+, Co3+, Cr3+, Fe3+, Ga3+, In3+, Sc3+, Ti3+, V3+, Th4+ at pH 1-4
2.1 Chemical Methods
2.1.1 Volumetric Methods – Redox Titration
redox reaction:
nBOxB + nARedA V nBRedB + nAOxA (zA=nB, zB=nA)
MnO4- + 5Fe2+ + 8H+ V Mn2+ + 5Fe3+ + 4H2O
examples:
Ce4+ + Fe2+ V Ce3+ + Fe3+
τ=
z A ⋅ C A ⋅VA
z B ⋅ C B ⋅ VB
0 < τ < 1 U = U A = U AΘ +
RT
τ
⋅ ln
nB F
1 −τ
n AU BΘ + nBU AΘ
τ = 1 U eq =
n A + nB
τ > 1 U = U B = U BΘ +
indication:
RT
⋅ ln(τ − 1)
nA F
self indication (violet MnO4- → colorless Mn2+)
redox indicator (InRed F InOx + ne-)
potentiometric endpoint detection
titration curve
2.1 Chemical Methods
2.1.2 Gravimetric Methods
precipitation reaction:
ν + M z + ( aq ) + ν − A z − (aq ) ↔ Mν + Aν − ( s )
Ø precipitation of the ion to be determined by a suitable precipitating agent
Ø filtering and washing of the precipitate
Ø transformation into a compound of well defined stoichiometry (drying/calcination)
→ weighing
examples
Ba2+ + SO42- F BaSO4 (KL=[Ba2+]⋅[SO42-]=1⋅10-10mol2/l2) as BaSO4
Fe3+ + 3OH- F Fe(OH)3 (KL=[Fe3+]⋅[OH-]3=1⋅10-38mol4/l4) as Fe2O3
[ M z + ] = ν + +ν −
Ø for analytical purposes:
ν
K L  +
ν −
ν−



[ A z − ] = ν + +ν −
ν
K L  −
ν +
ν+



[ M z+ ]⋅V
α p = 1−
≥ 0.997
z+
[ M ]o ⋅ V0
Ø advantage
⇒ one of the most precise analytical methods
2
σm  σm 
σm
=  e  +  a 
m
 me   ma 
2
0.01% <
σm
< 0.1%
m
2.2 Electroanalytical Methods
electrode reaction, IFaraday=0 (potentiometry)
Ø ion-selective electrodes
Ø potentiometric titration
electrode reaction, DC,
IFaraday≠0 (voltametry)
Ø polarography
electroanalytical
methods
no electrode
reaction, AC, high
frequency
Ø conductometry
Ø coulometry
electrode reaction, AC,
IFaraday≠0
Ø AC polarography
2.2 Electroanalytical Methods
2.2.1 potentiometry, potentiometric titration
Ø principle ⇒ measurement of the cell potential (potential between two electrodes at
zero current)
Ø relation of the cell potential and the analyte concentration
typical reference electrode
calomel electrode
Hg2Cl2(s) + 2e- V 2Hg(l) + 2Cl-
E = E 0' −
0.05916V
log[Cl − ]2
2
Ø electrolyte = saturated KCl solution
(3.8mol/l at 25°C) ⇒ [Cl-] = constant
⇒ ref. pot. E = const.
2.2 Electroanalytical Methods
2.2.1 Potentiometry, Potentiometric Titration
pH measurement with the glass electrode
Cl- determination with a silver electrode
Ag+ + Cl- V AgCl (s)
KL=1.8⋅10-10mol2/l2
0
E Ag = E Ag
−
/ Ag +
2.303 RT
1
log
1⋅ F
a Ag +
0
E Ag = E Ag
+ 0.059 log K L − 0.059 log[Cl − ]
/ Ag +
Ag/AgCl/KCl sat/H+in/H+out/KClsat/AgCl/Ag
E = ∆ϕ As + ∆ϕ Diff +
RT
ln 10 ⋅ ( pH in − pH out )
F
2.2 Electroanalytical Methods
2.2.2 Polarography
Ø measurement of current – voltage curves (voltametry) gives information of the
chemical composition and the concentration of the analytes in the sample
(working range 1⋅10-6 < cA < 1⋅10-3 mol/l)
Ø polarography was invented in 1922 by Jaroslav Heyrovsky (Nobelprice 1959)
electrode arrangement for
polarographic measurements
Ø a polarographic cell contains:
a polarizable working electrode
(dropping mercury electrode)
a non-polarizable counter electrode
(e.g. Pt wire)
a reference electrode
(e.g. calomel electrode)
a supporting electrolyte
(e.g. Li+ Cl- in H2O)
2.2 Electroanalytical Methods
2.2.2 Polarography
example: Cd determination
Ø half step potential (E1/2) charcteristic for the
analyte
Ø hight of the step (id/2) proportional to the
concentration of the analyte
Ø working potential range -2.2V < Eappl < 0.3V
Ø applicable to reducible or oxidizable metal
ions or organic compounds
A: 5⋅10-4 mol Cd2+ in 1 mol/l HCl as
supporting electrolyte
B: pure supporting electrolyte
2.2 Electroanalytical Methods
2.2.2 Polarography
polarographic oligo-element determination
A: 1⋅10-4 mol/l Ag+, Tl+, Cd2+, Ni2+, Zn2+ in 1 mol/l NH3 and 1 mol/l NH4Cl supporting electrolyte
B: pure supporting electrolyte
2.3 Spectroscopic Methods
2.3.1 Atomic Emission Spectroscopy
Ø history: one of the oldest methods for elemental analysis
(flame emission - Bunsen, Kirchhoff 1860)
Ø principle: recording of line spectra emitted by excited atoms or ions during
radiative de-excitation (valence electrons)
optical transitions and emission spectrum of H
intensity of an emission line
 h ⋅ c ⋅ gm ⋅ A ⋅ N 
 − Em 
I = Φ
 ⋅ exp

4
π
⋅
λ
⋅
Z
kT




Ø intensity of a spectral line ~N, i.e. ~C
λ=
h⋅ c
E
Ø population of excited states requires high
temperatures (typical 2000 K-7000 K)
Ø relative method, calibration required
2.3 Spectroscopic Methods
2.3.1 Atomic Emission Spectroscopy
instrumentation
analytical performance
Ø multielement method
(with arc 60-70 elements)
Ø presiscion: ≤ 1% RSD for spark, flame
and plasma; arc 5-10% RSD
Ø decision limits: 0.1-1 ppm with arc, 110ppm with spark, 1ppb-10ppm with
flame, 10ppt-5ppb for ICP
application:
radiation sources (classification)
inductively coupled
plasma (ICP)
flame
liquid
samples
spark
arc
glow discharge
laser
solid
samples
Ø flame AES – low cost system for
alkali and earth alkaline metals
Ø ICP AES – suited for any sample
that can be brought into solution
Ø arc AES – with photographic plate
for qualitative overview analysis
Ø spark AES – unsurpassable for
metal analysis (steel, alloys)
Ø laser ablation – for direct analysis of
solids (transient signals)
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
Ø principle: measurement of the absorption of light by free atoms in their electronic
ground state
ground and excited states of Al and
optical transitions used for AAS
linear relation between absorbance and
concentration
Aabs
g m λ4
1
 I0 
= log  = 0.434 ⋅ A ⋅
⋅
⋅
⋅ n0 ⋅ l
I
g
8
π
c
∆
λ
 
0
eff
AAS instrumentation
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
radiation source
Ø atomic absorption lines are very
small (∆λ≤0.005nm) ⇒ narrow
excitation lines required
working principle of hollow cathode lamps
Ø hollow cathode lamp filled with Ar
(~5 Torr) ⇒ discharge in cathode cup
Ø cup-shaped cathode made from the
element to be determined ⇒ emission
of sharp de-excitation lines
source of free atoms
flame
Ø pneumatic nebulization of the
liquid sample into the flame
or hydrid generation
(e.g. for As, Bi, Ge)
Ø air/C2H2 for nonrefractory
elements (e.g. Ca, Cr, Fe)
Ø N2O/C2H2 for refractory
elements (e.g. Al, Si, Ta)
graphite furnace
Ø electrically heated graphite tube
(Tmax=3000°C, under Ar)
Ø T vs. t program ⇒ drying,
ashing, atomization, cleaning
Ø transient signals
Ø liquid and solid samples
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
interferences in atomic absorption spectroscopy
Ø flame and furnace AAS are sensitive to matrix interferences ⇒ i.e. matrix constituents
(anything but the analyte) influence the analyte signal
Ø there are two types of interferences ⇒ chemical and spectral interferences
chemical interferences
Ø limited temperature of the flame does
not ensure full dissociation and
atomization of thermally stable
compounds (e.g. Ca phosphates) ⇒
hotter N2O flame (La buffer)
Ø large amounts of easily ionized
elements (alkali metals) modifies the
equilibrium between atoms and ions
⇒ adding excess of Cs
Ø loss of volatile analyte species during
the ashing process in GF-AAS (e.g.
ZnCl2) ⇒ adding of matrix modifiers
NH4NO3 + NaCl → NH4Cl + NaNO3
spectral interferences
Ø two spectral lines overlap within the
band pass of the dispersive element
(e.g. ∆λd ≈ 0.1nm, Cd 228.802nm and
As 228.812nm) ⇒ select another line
Ø nonspecific absorptions from solid or
liquid particles in the atomizer (e.g.
light scatterring by liquid or solid
NaCl particles)
Ø emission of molecular bands by
molecules and radicals present in the
atomizer
(metal halides from 200 - 400nm)
Ø background correction necessary
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
deuterium background correction
Ø deuterium lamp emits a continuous spectrum in the range 200-380nm
Ø alternating measurement of deuterium and hollow cathode lamp absorption
setup sketch
Ø simple correction method
!!!
d
n
Ø requirements: smoothly rvarying
ou
g
k
c
background, λ < 360nm
a
gb
n
i
ry for strongly
Ø bad performance
a
v
:
mbackgrounds
varying
e
l
b
pro
working principle
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
Zeeman background correction
Ø Zeeman effect: splitting of atomic energy levels in a magnetic field
Ø applied in AAS for correction of strong varying backgrounds
setup sketch
normal Zeeman effect (S=0)
(e.g. Cd atoms)
working principle
2.3 Spectroscopic Methods
2.3.2 Atomic Absorption Spectroscopy
analytical performance
Ø AAS is most popular among the instrumental
methods for elemental analysis
flame-AAS:
C relatively good limits decision limits (ng/ml)
C working range mg/l (liner range)
C easy to use, stable, comparably cheap
C few and well known spectral interferences
D single element method (up to 6 in parallel)
D high sample consumption
graphite furnace – AAS:
C superior decision limits (pg/ml)
(trace and ultra-trace analysis)
C very low sample consumption
C analysis of solids and liquids
C analyte matrix separation possible
D demanding operation
2.3 Spectroscopic Methods
2.3.3 Inductively Coupled Plasma Mass Spectrometry
Ø a mass spectrometer separates a stream of gaseous ions into ions with different mass
to charge ratio m/z (mass range in inorganic mass spectrometry from 1 – 300 u)
Ø in combination with an ion source (ICP, spark, glow discharge, laser ablation) as
analytical method for elemental analysis
Ø most popular combination ⇒ Inductively Coupled Plasma + quadrupol Mass
Spectrometer = ICP - MS
setup scheme
sample introduction
Ø pneumatic nebulizer
(liquid samples)
Ø electro thermal vaporization
(graphite furnace)
liquid and solid samples
Ø laser ablation
solid samples
2.3 Spectroscopic Methods
2.3.3 Inductively Coupled Plasma Mass Spectrometry
ICP as ion source
Ø plasma generation by inductively heating of a gas (e.g. argon) using a high frequency
field
ICP assembly
Ø ICP works under atmospheric
pressure, MS under high vacuum ⇒
differentially pumped interface required
Ø typical RF frequencies 27 or 40 MHz
Ø power consumption 1 – 5 kW
starting an ICP
a) plasma off
b) RF voltage applied
c) ignition by HT spark
d) plasma formation
e) steady state
2.3 Spectroscopic Methods
2.3.3 Inductively Coupled Plasma Mass Spectrometry
Ø sample droplets + carrier gas (inner tube connected to
nebulizer) punch a channel through the toroidal plasma
Ø sample transformations in the plasma:
desolvation → vaporization → atomization → ionization
Ø ionization behavior depends on the temperature ⇒ Saha
equation (law of mass action for ionization equilibria)
α i (1 + α e )
= pK (pi +1) (T )
α i +1α e
K
( i +1)
p
g i λ3e
I 
(T ) =
exp i +1 
ge g i + 1 T
 T 
temperature profile
Ø example: Ar0 (1S0) → Ar+ (2P3/2)
Na0 (2S1/2) → Na+ (1S0)
I+=15.76 eV, 1 bar
I+=5.14 eV
2.3 Spectroscopic Methods
2.3.3 Inductively Coupled Plasma Mass Spectrometry
Ø determination of elements Z < 80 u by ICP - MS suffers from spectral interferences
caused by the plasma gas, solvent, sample matrix etc.
types of spectral interferences
Ø isobaric overlaps
Ø oxide-, hydroxide species
Ø doubly charged ions
interference
analyte
40 Ar+
40Ca+
40 Ar16O+
56Fe+
138Ba2+
69Ga+
Ø spectral interferences cause a high background ⇒ higher decision limit for disturbed
analytes (e.g. xd(Fe)≈10 µg/l, xd(Ag)≈0.03 µg/l)
2.3 Spectroscopic Methods
2.3.3 ETV – ICP – MS
Ø analyte – matrix separation by coupling of an ElectroThermal Vaporization to the
ICP-MS → ETV-ICP-MS
setup picture
scheme of the ETV unit
operation principle
Ø quenching of the sample vapor by mixing it with
cold carrier gas
d k = 4σ ⋅
Vm
k ⋅ T ⋅ ln S
S=
p
peq (T )
Ø sample transport as condensed particles if the
condensation nuclei exceed the critical diameter
2.3 Spectroscopic Methods
2.3.3 ETV – ICP – MS
Ø one touchstone in ICP-MS analytics is the determination of
light elements in saline solutions, e.g. seawater)
Ø example: Zn determination in sea water
relevant spectral interferences
ion
64 u
66 u
67 u
68 u
concentration
in seawater
Zn2+
48.6% n.a.
27.9% n.a.
4.1% n.a.
18.8% n.a.
~1 ng/ml
35Cl16O16O+
35Cl16O17O+
ca. 16 mg/l
Cl34S16O16O+
32S34S+
SO42-
32S16O16O+
33S33S+
32S32S+
32S17O17O+
32S16O18O+
33S16O17O+
Mg2+
24Mg40Ar+
26Mg38Ar+
26Mg40Ar+
36S16O16O+
33S34S+
32S36S+
32S17O18O+
34S17O17O+
33S16O18O+
34S16O18O+
33S17O17O+
34S34S+
34S16O17O+
33S17O18O+
ca. 2 mg/l
32S18O18O+
ca. 1 mg/ml
2.3 Spectroscopic Methods
2.3.3 ETV – ICP – MS
application to the Zn in seawater problem
13000
Ø analyte – matrix separation by ETV
12000
Ø sample matrix can be vaporized prior the
analyte or in reversed fashion
10000
9000
intensity in counts/s
Ø application of modifiers (ETV as
thermochemical reactor, cp. atomic
absorption spectroscopy)
11000
64amu, 48,6% n.H.
66amu, 27.9% n.H.
67amu, 4.1% n.H.
68amu, 18.8% n.H.
8000
7000
Zn signal, correct isotopic ratio
Zn vaporizes before the matrix
components
6000
5000
4000
3000
2000
1000
0
temperature in °C
step
T / °C
ramp time / s
hold time / s
function
1
150
10
60
drying
2
150
0
30
baseline
3
800
1
59
Zn vaporization
(prob. as ZnCl2
BP=732°C)
4
2500
5
5
cleaning
3000
2900
2800
2700
2600
2500
2400
2300
2200
2100
2000
1900
1800
1700
1600
1500
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
Ø resulting temperature – time program
2.3 Spectroscopic Methods
2.3.3 ETV – ICP – MS
Ø if matrix matched calibration samples are not available ⇒ standard addition
Ø principle of the standard addition method: - measurement of the analysis sample
- stepwise addition of analyte (x+=xA…4xA)
- evaluation of xA by extrapolation of the
calibration function to the intercept point
with the x-axis (y=f(x+)=0 ⇒ xA)
- prerequisite: no blank value!!!
result: 1 ng/ml < CZn < 3 ng/ml
certified value: 1.24 ng/ml
2.3 Spectroscopic Methods
2.3.3 Inductively Coupled Plasma Mass Spectrometry
analytical performance of ICP-MS
Ø multi-element method (more than 20 analytes in parallel)
Ø good performance in precision (a few percent),
accuracy, number of determinable elements, decision limit
(depending on the instrument 0.01-100 ng/l) and sample
throughput
Ø linear dynamic range ⇒ 3-5 orders of magnitude
Ø difficult for elements with m/z < 80 u (spectral interferences)
Ø mostly applied for liquid sample, ETV and laser vaporization enable direct analysis of
solid samples
Ø supplies information about the isotopic ratio of an element
- isotop dilution analysis for ultra-trace analysis (sub ppt range)
- age determination of biological and geological samples
Ø relatively young analysis method (1980) ⇒ high innovative potential
Backup Slide
Specifity and Selectivity of Analytical Methods
specificity…
Ø …describes the ability of a method to detect 1 particular analyte on 1 sensor
undisturbed from all other components present in the sample
Ø …refers to a single component analysis
Ø …is based upon the concept of partial sensitivities
vector of partial sensitivities
s A = ( S AA S AB ... S AZ )
spec( A / B,..., Z ) =
S AA ⋅ x A
Z
∑S
K=A
AK
⋅ xK
0 ≤ spec(A/B,…,Z) ≤ 1
for practical work spec(A/B,…,Z)>0.9 sufficient
S IJ =
∂y I
∂x J
Backup Slide
Specifity and Selectivity of Analytical Methods
selectivity…
Ø …describes the ability of a method to determine n analytes on n sensors undisturbed
and independent from each other and from other components present in the sample
Ø …refers to a multi-component analysis
Ø …is based upon the concept of partial sensitivities
S IJ =
matrix of partial sensitivities
 S AA

 S BA
S=
M

S
 NA
S AB L S AN S AN +1 L S AZ 

S BB L S BN S BN +1 L S BZ 
M
O
M
M L M 

S NB L S NN S NN +1 L S NZ 
N
sel ( A, B ,..., N / N + 1,..., Z ) =
∑S
I=A
N
Z
II
∑∑S
J=AI=A
⋅ xI
IJ
⋅ xI
0 ≤ sel(A,B,…,N/N+1,…,Z) ≤ 1
for practical work sel(A,B,…,N/N+1,…,Z) >0.9 sufficient
∂y I
∂x J
Backup Slide
Specifity and Selectivity of Analytical Methods
Ø application of the selectivity/specificity concept to compare PN-ICP-MS and ETV-ICPMS for the determination of Zn in seawater
 ∆I 64amu

 ∆C Zn 2 +
 ∆I 64amu

∆C −
S =  ∆I Cl
 64amu
 ∆C
SO 42 −

 ∆I 64amu
 ∆C
Mg 2 +

specificity matrix for the
sensors (masses) 64,66,67,68 u
test samples to
determine S
∆I 66amu
∆C Zn 2+
∆I 66amu
∆C Cl−
∆I 66amu
∆C SO 2 −
4
∆I 66amu
∆C Mg2 +
∆I 67amu
∆C Zn 2 +
∆I 67amu
∆C Cl−
∆I 67amu
∆C SO 2 −
4
∆I 67amu
∆C Mg 2 +
∆I 68amu
∆C Zn 2 +
∆I 68amu
∆C Cl−
∆I 68amu
∆C SO 2 −
4
∆I 68amu
∆C Mg2 +
sample
[Zn2+] in
ng/ml
[Cl-] in
mg/ml
[SO42-] in
mg/ml
[Mg2+] in
mg/ml
1
50
10
1
0.5
2
100
10
1
0.5
3
50
20
1
0.5
4
50
10
5
0.5
5
50
10
1
1.5












Backup Slide
Specifity and Selectivity of Analytical Methods
Ø experimental specificity matrices after t-test (comparison of two mean values)
SPN


2+
 Zn
=  Cl −

 SO42 −

2+
 Mg
64amu
66amu
67amu
616
1.22 ⋅ 10 − 3
378
− 3.48 ⋅ 10 − 4
63.9
0
18.6 ⋅ 10 − 3
0
0
28.5 ⋅ 10 − 3
14.3 ⋅ 10 − 3
2.52 ⋅ 10 − 3

68amu 
 2+

 Zn
265 
 S ETV =  Cl −
0
 2−

 SO 4
0

 Mg 2+

8.97 ⋅ 10 − 3 

64amu
66amu
67amu
5669
− 10.2 ⋅ 10−3
3557
− 5.67 ⋅ 10−3
524
− 1.01 ⋅ 10−3
0
0
− 23.0 ⋅ 10−3
− 40.4 ⋅ 10−3
0
0


2438 
− 4.13 ⋅ 10−3 

− 17.3 ⋅ 10−3 
− 33.4 ⋅ 10−3 
68amu
Ø ETV-ICP-MS about 10 times more sensitive than PN-ICP-MS
Ø positive partial sensitivities in SPN reflect spectral interferences
(e.g. interference of 24Mg40Ar+ and 26Mg38Ar+ on 64Zn+)
Ø negative partial sensitivities in SETV reflect non-spectral interferences
Ø influence of Cl- : Zn2+ + 2Cl- → ZnCl2(s) → ZnCl2(g)
(the less Cl- in solution the less Zn2+ vaporizes as ZnCl2) ⇒ non-spectral interference
due to sample transport mechanism
Ø influence of SO42- on the drying step: Zn2+ + SO42- → ZnSO4(s)
above 770°C ZnSO4(s) → ZnO (mp = 1974°C) + SO2+ ½ O2
⇒ as the vaporization step is done at 800°C SO42- ions cause a loss of Zn2+
Backup Slide
Specifity and Selectivity of Analytical Methods
Ø specificities for a typical seawater sample
ion
C
in ng/ml
Zn2+
75
Cl-
15⋅106
SO42-
2.5⋅106
Mg2+
1.0⋅106
Sprakt.
64amu
PN
Sprakt.
66amu
PN
Sprakt.
67amu
PN
Sprakt.
68amu
PN
Sprakt.
64amu
ETV
Sprakt.
66amu
ETV
Sprakt.
67amu
ETV
Sprakt.
68amu
ETV
0.33
0.59
0.66
0.69
0.74
0.59
0.72
0.57
1.0
spec PN , 64amu ( Zn 2+ / Cl − , SO42− , Mg 2 + ) =
2+
−
2−
4
2+
616 ⋅ x Zn2+
616 ⋅ x Zn2+ + 93300
spec ETV ,64amu ( Zn / Cl , SO , Mg ) =
5669 ⋅ x Zn2+
5669 ⋅ x Zn2+ + 153000
0.9
spec(Zn / Cl , SO4 , Mg)
no external calibration
(i.e. standards) possible
for ng/ml concentrations
0.8
ETV
0.7
0.6
PN
0.5
0.4
0.3
0.2
0.1
0.0
0
100 200 300 400 500 600 700 800 900 1000
xZn in ng/ml
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
Ø instrumental analytical technique for the elemental analysis of solids and liquids
Ø wide concentration ranges (ppm - %), minimal sample preparation
Ø principle: detection of element specific X – rays (characteristic X-rays) emitted from a
sample under X-ray excitation (X-ray fluorescence)
interaction of X-rays with matter
emission of a core level electron (here from
copper) by absorption of a X-ray photon
energetic relaxation by X-ray
fluorescence
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
Ø transitions are governed by quantum
mechanical selection rules
(∆n>1, ∆l=±1, ∆j=0,±1)
Ø K – lines originate from vacancies in
the K – shell, L – lines from
vacancies in the L – shell …
Ø denotation of lines:
i) IUPAC notation – Fe KL3
ii) Siegbahn notation – Fe α1
Ø line intensities depend on:
i) the energetic difference of the levels
(the higher the energetic difference the
lower the transition probability)
ii) the degeneracy of the levels
iii) the fluorescence yield
Ø K – lines best suited for analysis of
elements Z < 45 (Rh) (mostly K-L3,2)
Ø L – lines for analysis of elements Z > 45
(mostly L3-M5,4)
atomic energy levels and allowed transitions
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
instrumentation – X-ray source
Ø generation of X-rays by collision of high
energetic electrons with a pure metal
target (Rh, Mo, Cr, Ag, W)
Ø continuous X-rays by decelerating
collisions with the target atoms
(Bremsstrahlung)
Ø characteristic X-rays by refilling of core
level vacancies in the target atoms
created by the impinging electrons
Ø the short wavelength limit of the source
depends on the accelerating voltage, the
long wavelength tail depends on the Bewindow thickness
Ø only ~1% of the electric power is
converted to X-rays, rest is heat ⇒
effective water or air cooling required
Ø quantitative work ⇒ stable filament
heating and accelerating voltage
required
Rh
anode
45kV
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
instrumentation – wavelength dispersive instruments
geometric arrangement
Ø irradiated sample emits
polychromatic X-rays
detector
Ø crystal as dispersive element
Ø wavelength selection according to
Bragg’s law
n ⋅ λ = 2 ⋅ d ⋅ sin θ
Ø wavelength scan by varying θ ⇒
rotation of the crystal with respect to
the incoming beam
flow proportional
counter for low
energetic X-rays
scintillation
counter for high
energetic X-rays
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
instrumentation – energy dispersive instruments
Ø energy dispersion and X-ray photon
counting in one step
Ø detector ⇒ semiconductor crystal
(Si(Li), hyperpure Ge) at liquid nitrogen
temperature (-196°C)
example
Ø an Fe K-L3,2 X-ray photon (E=6.400keV)
hits a Si detector
Ø energy to create 1 electron – hole pair =
3.85eV ⇒ creation of 1662 ehp’s
Ø the preamplifier converts this current in
voltage pulse of e.g. 32mV
Ø incoming X-ray photons create
electron hole pairs ⇒ current ⇒ voltage
pulse
Ø the number of electron – hole pairs is
proportional to the energy
Ø best suited for energies > 2 keV
Ø a connected ADC translates the
amplified pulse height into a digital
number, e.g. 320 ⇒ channel 320 in a
memory is increased by 1
Ø common memories employ 1024 or 2048
channels (0-20 keV and 0 – 40 keV
respectively) ⇒ resolution ~20 eV
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
Ø application of XRF: qualitative and quantitative analysis of elements with Z ≥ 9 (F)
⇒ multi-element method
Ø relatively simple sample preparation
(bulk solids, pressed powder pellets, fused discs, liquids)
Ø information depth in the µm range ⇒ homogeneous and smoothed surface required
Ø qualitative analysis by recording the entire fluorescence spectrum ⇒ software
assisted element identification by assigning the element specific K,L,M – lines
qualitative analysis by XRF
WD-XRF
ED-XRF
2.3 Spectroscopic Methods
2.3.4 X-Ray Fluorescence Spectroscopy
Ø quantitative analysis requires relation between line intensity and element concentration
Ø 2 step process:
i) determination of the net peak intensity (background subtraction)
ii) conversion of the net peak intensity to concentration
quantitative analysis
complicated by
x-ray attenuation
x-ray enhancement
Ø excitation of the Fe - K lines by
the two Ni – K lines
I x = I 0 ⋅ exp(− µ ⋅ ρ ⋅ x )
µ sample = ∑ C i µ i
i
a) no matrix effect
b) attenuation
c) enhancement
fundamental parameters
approach (implemented in
software)
Ø observed Fe – K line intensity
= sum of primary and secondary
fluorescence
Ø Cr – K line intensity = sum of
primary, secondary and tertiary
fluorescence
2.3 Spectroscopic Methods
2.3.5 Electron Probe X-ray Microanalysis (EPXMA)
EPXMA at the Fritz-Haber-Institute
(Inorganic Chemistry Department, Building L)
2.3 Spectroscopic Methods
2.3.5 Electron Probe X-ray Microanalysis (EPXMA)
Ø 2D
elemental
analysis
by
combination of Scanning Electron
Microscopy
and
X-Ray
Fluorescence Analysis
SEM instrumentation
electron sample interaction
Ø topographic sample image by
synchronizing the TV – scanner
with the scanning coils
Ø X-ray detection by a solid – state
detector
(Si
single
crystal)
as in XRF (not shown above)
2.3 Spectroscopic Methods
2.3.5 Electron Probe X-ray Microanalysis (EPXMA)
information depths, spatial resolution
Ø in most cases, the SEM image
is created using secondary
electrons
Ø secondary electrons are of low
energy (<50eV), surface sensitive
and supply an topographic
image of the sample surface
Ø spatial resolution < 0.1 •
possible
secondary electron image
synthetic graphite (25keV, ×3000)
2.3 Spectroscopic Methods
2.3.5 Electron Probe X-ray Microanalysis (EPXMA)
X – ray generation
information depths, spatial resolution
Ø X-ray generation process similar to XRF
Ø X-ray spatial resolution much lower than by secondary electrons
2.3 Spectroscopic Methods
2.3.5 Electron Probe X-ray Microanalysis (EPXMA)
example
analytical performance
(pinhole for molecular beam setup)
Ø qualitative elemental analysis from
Z > 11 by identification of the
characteristic KLM peaks (software
assisted)
Ø detection limit ~1%
Ø minimal sample preparation and sample
damage
Pt
Ø quantitative analysis can be performed
from
element
concentrations
of
1 – 100% with a relative precision of 1-5%
Ø similar to XAF, relating net peak intensity
with element concentration requires
either sophisticated mathematical models
(e.g.
ZAF
correction)
or
suitable
calibration samples
Ni
3. Specification of an Analytical Result
Ø any quantitative analysis relies on a relationship between the measurement value y
and the analyte concentration x ⇒ y = f(x)
Ø if y = f(x) is known, x can be calculated by x = f-1(y)
Ø the mathematical form of f(x) and f-1(y) depends on the analytical method
analytical methods
relative methods
absolute methods
definite methods
reference methods
direct RM
indirect RM
B = yr/xr
Y = f(x)
Y = a+bx
y = b⋅x
b=F
b = k⋅F
gravimetry
m Aq B p =
M Aq B p
q⋅ MA
B
⋅ mA
ICP-MS, AES, AAS
photometry
XPS
B
I A FX S ( E k )σ ( E k )λ (E k )cos θ ⋅ n A n A
=
=
I R FX S ( E k )σ ( E k )λ (E k )cos θ ⋅ nR nR
Aλ = ε λ ⋅ d ⋅ c A
IA =
IR
⋅ nA
nR
B
3. Specification of an Analytical Result
Ø if the accuracy of the method is assured (e.g. by certified reference materials), the
analytical result can be specified as follows
x = x ± ∆x
x = mean of n p parallel determinations
∆x = prediction interval of x
Ø verbal: with a probability of P is the concentration x of the analyte A in the
concentration range x ± ∆x
Ø results from measurements with absolute, definite and direct reference methods can
be evaluated using np parallel determinations (entire sample preparation process) and
the uncertainties of all values in B using the rules of error propagation ⇒ x = B-1⋅y
Ø specification of results from measurements with an indirect reference method
(experimental calibration) requires to evaluate both the uncertainty of the analysis and
of the calibration
3. Specification of an Analytical Result
example 1: gravimetric determination of Fe2O3 in an iron ore
Ø results of np=5 parallel determinations 38,71% 38,99% 38,62% 38,74% 38,73%
mean : x =
1
np
∑x
i
= 38.76%
standard deviation : s =
i
degrees of freedom :
∑ (x
− x)
2
i
i
np − 1
= 0.1381%
confidence level : P = 95%
f = np − 1 = 4
t-table
result : x ± ∆x = x ±
s ⋅ t ( 0.95,4)
0.1381% ⋅ 2.78
= 38.76% ±
= 38.76% ± 0.17%
np
5
3. Specification of an Analytical Result
example 2: Zn determination by AAS
Ø calibration: m samples (xi, yi) ⇒ f = m-2
Ø xi taken as error free
Ø least squares fit
y = aˆ + bxˆ
bˆ =
∑ (x
i
∑ (x
i
aˆ =
∑y
s02 =
− x )( yi − y )
i
i
i
− x)
2
i
− b∑ x i
i
m
∑ (y
i
2
− yˆ i )
i
m−2
∆yˆ k = t ( P , f )


2
(
)
xk − x 
1
s02  +
m ∑ ( xi − x )2 


i


xi/ppm
2
4
6
8
10
yi (E=log(I0/I))
0.266
0.598
0.856
1.222
1.467
3. Specification of an Analytical Result
example 2: Zn determination by AAS
Ø analysis: np measurement samples (e.g. np=3 parallel determinations)
np
1
2
3
yi (E=log(I0/I))
1.064
0.930
0.942
Ø calculation of
Ø calculation of
x=
y = 0.979
y − aˆ 0.979 − (− 0.026)
=
= 6.656 ppm
0.151
bˆ
s
1
1
∆x = t ( P , f ) ⋅ 0 ⋅
+
+
m np
bˆ
(x − x )
∑ (x − x )
2
2
i
i
0.0314 ppm 1 1 (6.656 − 6) ppm 2
∆x = 3.18 ⋅
⋅
+ +
= 0.488 ppm
0.151
5 3
40 ppm 2
2
x ± ∆x = 6.656 ± 0.488 ppm
4. Summary
Ø a wide variety of methods are available to determine quantitatively the elemental
composition of inorganic samples
methods for
elemental analysis
chemical
methods
instrumental
methods
advantages
advantages
cheap, direct, easy to
understand, precise
cover concentrations from 100%
to sub ppt, solid and liquid
samples, multi-element
character
disadvantages
cumbersome, restricted to main
components, danger of
systematic errors, liquid
samples only
disadvantages
expensive, “black box”
character, operation requires
expertise
Ø the choice of the right method depends on numerous parameters ( liquid or solid
sample, available amount of sample, which and how many elements, expected
concentrations, accuracy and precision, decision limit…)
5. Literature
Ø General Analytical Chemistry
Kellner R.; Mermet J.-M.; Otto M.; Valcárcel M.; Widmer H. M.; Analytical Chemistry. WileyVCH, Weinheim, 2nd edition, 2004.
Ø Chemical Methods for Elemental Analysis
Ackermann G.; Jugelt W.; Möbius H.-H.; Suschke H. D.; Werner G.; Elektrolytgleichgewichte
und Elektrochemie. Lehrwerk Chemie, Vol. 5, VEB Deutscher Verlag für Grundstoffindustrie,
Leipzig, 5th edition, 1987
Ø Electroanalytical Methods
Scholz F.; Electroanalytical Methods. Springer Verlag, Berlin, 1st edition, 2002
Ø Analytical Atomic Spectrometry
Broekaert J. A.; Analytical Atomic Spectrometry with Plasmas and Flames. Wiley-VCH,
Weinheim, 1st edition, 2001
Evans E. H.; Fisher A.; An Introduction to Analytical Atomic Spectrometry. Jon Wiley and
Sons Ltd, 2nd edition, 2002
Welz B.; Sperling M.; Atomic Absorption Spectroscopy. Wiley-VCH, 3rd edition, 1998
Montaser A.; Inductively Coupled Plasma Mass Spectrometry, Wiley-VCH, 1998
5. Literature
Ø X-Ray Fluorescence Spectroscopy
Lachance G. R.; Claisse F.; Quantitative X-Ray Fluorescence Analysis. John Wiley & Sons
Ltd, 1995
Ø Energy Dispersive X-Ray Microanalysis
Goldstein J.; Scanning Electron Microscopy and X-Ray Microanalysis. Plenum Publishing
Corp., New York, 3rd edition, 2002
Ø Statistics and Chemometrics
Doerffel K.; Statistik in der Analytischen Chemie. Deutscher Verlag für Grundstoffindustrie,
Leipzig, 5th edition, 1990
Massart D. L.; Vandeginste B. G. M.; Buydens L. M. C.; De Jong S.; Lewi P. J.; SmeyersVerbeke J.; Handbook of Chemometrics and Qualimetrics Part A and B. Elsevier,
Amsterdam, 1997