1

X -RAY FLUORESCENCE SPECTROMETRY

2
X- rays are short waveIength eIectromagnetic
radiations produced by the deceIeration of high energy
eIectrons or by eIectronic transitions of eIectrons in
the inner orbitaIs of atoms.
X- rays have a waveIength range of 10
-5
A to 100 A .
ConventionaI X- ray spectroscopy is IargeIy confined
to 0.1 A to 25 A .
0
0
0
0
3
X- rays are generated by:
1. Bombarding a metaI target with a beam of high
energy eIectrons.
2. Exposing a substance to a primary beam of
X-rays to generate a secondary beam of X-rays of
Iower energy.
3. Using a radioactive source emitting X-rays during
its decay process and
4. From a synchrotron radiation source.
4
In an X-ray tube both continuum and Iine spectra are
emitted. EIectrons are produced at a heated cathode
and acceIerated toward a metaI anode (target) by a
potentiaI difference of 100 kV. Upon coIIiding with the
anode, part of the energy of the incident beam is
converted into X-rays.
The continuum X-ray spectrum exhibits a weII defined
short waveIength Iimit (
0
) which is a characteristic of
the appIied voItage, irrespective of the nature of the
target.
5
CONTINUUM RADIATION FROM AN X - RAY
TUBE WITH A TUNGSTEN TARGET
6
The continuum radiation resuIts from coIIisions
between the eIectrons and atoms of the target. Each
coIIision resuIts in the emission of a photon. The
energy of the photon is equaI to the energy Ioss of the
eIectron. A number of coIIisions with decreasing
energy may occur for each eIectron as it bounces from
one atom into another . The maximum photon energy
generated corresponds to the instantaneous
deceIeration of the eIectron to zero kinetic energy in a
singIe coIIision.
7
The kinetic energy of aII the eIectrons can be
expressed as the Duane - Hunt Iaw,
Ve = h = hc /
where v is the appIied potentiaI, is the frequency.
Thus the kinetic energy is the product of acceIerating
voItage and the charge on the eIectron and is the
maximum of the radiation that can be produced.
0

8
X-ray Iine spectra resuIt from eIectronic transitions in
the innermost atomic orbitaIs. They occur in the Ionger
waveIength range of 4 - 6 A . The Iine spectra occur
for aII the eIements having atomic numbers of 12 and
above. EIements having atomic numbers Iess than 23
show onIy two Iines caIIed as K series and designated
as K

and K

. These are of short waveIengths. For

exampIe K series for tungsten target appear at 0.18
and 0.21A.
0
0
9
LINE SPECTRUM WITH A MOLYBDENUM TARGET
K
K
To To
15 37
35kV
R
e
I
a
t
i
v
e

i
n
t
e
n
s
i
t
y
12
8
4
0
0.2 0.4 0.6 0.8 1.0
waveIength, A
o
10
EIements having atomic numbers more than 23 show L
series which are aIso designated as
1
,
1
. There is a
threshoId voItage for each eIement beIow which Iine
spectra do not appear. For exampIe, beIow 50 kV , no
Iine spectra is obtained for moIybdenum. However
above 70 kV it produces Iine spectra.
11
EIement Atomic
number
K - Series

1
L - Series

1
Na 11 11.909 11.617 - -
K 19 3.742 3.454 - -
Cr 24 2.290 2.085 21.714 21.323
Rb 37 0.926 0.829 7.318 7.075
Cs 55 0.401 0.355 2.892 2.683
W 74 0.209 0.184 1.476 1.282
U 92 0.126 0.111 0.911 0.720
X-ray emission Iines for some typicaI eIements
12
MosIey showed that there is a Iinear reIationship
between the square root of frequency for a given K or L
Iine and the atomic number as shown beIow.
c/ = a ( Z - )
2
13
The short waveIength K series is produced when the
high energy eIectron beam removes an eIectron from
those orbitaIs nearest to the nucIeus of the target
atom. This produces an excited ion which emits a
quanta of radiation as eIectrons from outer orbitaIs
undergo transitions to fiII the vacancy.
K series Iines arise from transitions between higher
energy IeveIs and the K sheII.
L series of Iines resuIts when an eIectron is Iost from
the second principaI quantum IeveI directIy or from the
transition of an L eIectron to the K IeveI.
14
PARTIAL ENERGY LEVEL DIAGRAM CAUSING
X-RAYS PRODUCTION
15
In this figure, the atomic energy IeveIs are in the
Iogarithmic scaIe. The energy difference between
1
,
2
and
1
,
2
are so smaII that they appear as singIe Iines.
This energy IeveI diagram is appIicabIe to any eIement
with sufficient eIectrons to permit the number of
transitions shown.
The waveIengths of characteristic X-ray emissions
are independent of the chemicaI state of the eIement as
the transitions do not invoIve bonding or nonbonding
eIectrons. Thus the position of K

Iine for tungsten or

any eIement is same whether it is an oxide, suIphide or
pure metaI or an organic compIex.
16
Based on the properties of X-rays, a number of anaIyticaI
methodoIogies have been deveIoped over the years. These
incIude :
a) X-ray emission spectroscopy (XES)
b) Auger emission spectroscopy (AES)
c) X-ray fIuorescence spectroscopy (XFS)
d) EIectron spectroscopy (ESCA)
e) X-ray absorption
f) X-ray diffraction
17
X-RAY ANALYTICAL METHODS
X - ray emission Auger emission
X-ray fIuorescence
EIectron spectroscopy
X-ray absorption X - ray diffraction
18
X-ray absorption and emission spectra are quite
simpIe because they consist of very few Iines. X-ray
emission spectrum of an eIement may be obtained by
using the sampIe itseIf as the target eIement. However
this approach is not very convenient for aII types of
sampIes.
Instead, excitation can be accompIished by irradiating
the sampIe with a beam of X-rays from an X-ray
tube or a radioactive source. In this method, aII the
eIements in the sampIe are excited by the absorption of
the primary beam which emit their own characteristic
fIuorescence X-rays.
19
Thus XRF is a powerfuI tooI for the quaIitative (pass faiI test)
and quantitative estimation of aII but the Iightest eIements
having atomic numbers greater than that of oxygen (>8).
XRF is a non destructive technique as against most of
the other eIementaI techniques.
There are two types of X-ray fIuorescence spectrometers :
waveIength dispersive X-ray fIuorescence (WDXRF) and
energy dispersive X-ray fIuorescence (EDXRF)
spectrometers.
Important instrument components of XRF incIude a source,
a transducer, a signaI processor and read out.
20
INSTRUMENTATION FOR XRF
21
SOURCES
1. CooIidge tube
An X-ray tube is a high vacuum, seaIed off tube
containing a heated fiIament cathode (W) and a 5 x 10
mm heavy bIock of copper with a metaI target pIated
or embedded on its surface. Target materiaIs incIude
tungsten, chromium, copper, moIybdenum, rhodium,
scandium, siIver, iron and cobaIt.
Separate circuits are used to heat the fiIament and to
acceIerate the eIectrons to the target. The heater
circuit controIs the intensity of the emitted X-rays and
the acceIerating voItage determines their energy or
waveIength. Both current and voItage must be stabIe
up to 0.1%.
22
X-ray generation by this method is basicaIIy an
inefficient process and much energy is wasted as heat.
Hence cooIing of the anodes is necessary.
Modern equipments empIoy highIy sensitive
transducers and hence cooIing is not necessary.
The X-ray beam passes out of the tube through a thin
beryIIium window or a speciaI gIass. For waveIength of
6 - 70 A , uItra thin (1m) aIuminum or parIodion fiIms
are used to isoIate the tube.
0
23
The current is reguIated by monitoring the X-ray tube
DC current and controIIing the fiIament voItage. When
the fuII wave is rectified, the voItage reaches its peak
vaIue 120 times per second and persists for onIy a
smaII fraction of time.
When the tube is operated at 50kV, the gain is two foId
for eIements up to At. Number 35 (Br) and increases to
four foId at At. Number 56 (Ba).
NormaIIy X-ray tubes are operated between 50 - 60 kV
but now a days tubes with 100 kV are avaiIabIe.
24
SCHEMATIC OF AN X-RAY TUBE
25
RADIOISOTOPES
A variety of radioactive substances such as
etc., are
used as sources in XRF. Some eIements produce
simpIe Iine spectra and others produce a continuum. A
given radioisotope wiII be suitabIe for exciting a number
of eIements. For exampIe a source giving X-ray in
0.30 - 0.47 A is suitabIe for XRF studies for aII eIements
up to siIver. The sensitivity improves as the waveIength of
the source Iine approaches the absorption edge.
Pb Al Pm I Cd Co Fe H
210
82
147
61
125
53
109
48
57
27
55
1
3
3
; , , , , ,
0
26
SECONDARY FLUORESCENCE SOURCES
In some appIications an X-ray tube with tungsten
target is used to excite K and K Iines of MO.
The resuIting fIuorescence spectrum wouId contain
onIy Iine spectra which can be used for the excitation
of the anaIyte.
27
X-RAY FILTERS
In some appIications, x-ray tubes with narrow
waveIength range are required. Both fiIters and
monochromators are used for this purpose.
Zirconium fiIters of 0.01 cm thickness are suitabIe to
isoIate K

Iine. It fiIters most of the continuum and K

Iine.
SeveraI other fiIters have aIso been deveIoped for this
purpose. Quite often, thin strips of other metaIs are
aIso empIoyed for this purpose.
28
FILTERS FOR X-RAYS
29
X-RAY MONOCHROMATORS
X-ray monochromators consist of a pair of beam
coIIimators. One serves as a sIit and the other serves
as the dispersing agent.
The Iatter is a singIe crystaI mounted on a goniometer
or a rotating tabIe that permits the determination of the
angIe between the crystaI face and the coIIimated
incident beam.
Bragg's equation shows that for any given anguIar
setting of the goniometer onIy a few waveIengths are
Diffracted according to the equation,
n = 2 d sin
30
To produce a spectrum, the exit beam and the detector
must be mounted on a second tabIe that rotates at
twice the rate of the first. Hence when the crystaI
rotates by an angIe the detector rotates by 2.
Modern X-ray monochromators have microprocessor
controIIed motors to drive the crystaI and the detector
independentIy without a gear based mechanicaI
connection. Hence they can scan rapidIy at the rate of
240
0
/min.
31
AN X-RAY MONOCHROMATOR AND
DETECTOR ARRANGEMENT
32
The coIIimators in an X-ray monochromators consist of
a series of cIoseIy spaced metaI pIates that absorb aII
the radiations except the paraIIeI beams.
X-radiation Ionger than 2A is absorbed by the
atmospheric gases. Therefore if Ionger waveIength
X-rays are required, the sampIe compartment must be
fIushed with heIium or it must be evacuated by
pumping.
With fIat crystaIs, up to 99% of the X-radiation is
divergent which is absorbed by the coIIimators. The
Ioss can be minimized further by 10% using a curved
crystaI surface. The curved surface aIso accompIishes
the focusing of the divergent beam from the source on
the exit coIIimator.
0
33
DIFFRACTING CRYSTAL PROPERTIES
CrystaI
Lattice
Spacing
d,
WaveIength
range, a

max

min
Dispersion,
0
/
at
max
at
min
Topaz 1.356 2.67 0.24 2.12 0.37
LiF 2.014 3.97 0.35 1.43 0.25
NaCl 2.820 5.55 0.49 1.02 0.18
EDDT
b
4.404 8.67 0.77 0.65 0.11
ADP
c
5.325 10.50 0.93 0.54 0.09
0
A
0
A
0
A
a
Based on the assumption that the measurabIe range of 2 is from 160
0
for
max
to 10
6
for
min
b
EDDT = EthyIenediamine d-tartrate
c
ADP = Ammonium dihydrogen phosphate
34
The usefuI waveIength range for a crystaI is
determined by its Iattice spacing 'd ' and the detector
probIems when 2 is much Iess than 10
0
. This
happens when is approaching zero or 180
0
. Under
these conditions, poIychromatic radiation scattered
from the crystaI surface becomes very high. Therefore
onIy up to 160
0
(2) are empIoyed in X-ray work.
CrystaIs with Iarge crystaI Iattice spacing (d), such as
ammonium dihydrogen phosphate offer better controI
over the scatter controI but it resuIts in Iower
dispersion aIso.
35
Differentiating Bragg's equation, we get,
d n
d 2d cos
Hence d/d the measure of dispersion is inverseIy
proportionaI to d. Hence Iow dispersion prohibits the
use of shorter waveIengths. Here topaz or LiF must be
substituted.
=
36
X-RAY DETECTORS
X-ray detectors are usuaIIy operated as photon
counters. In this mode, individuaI puIses of the charge
are produced as quanta of radiation which are
absorbed by the transducer and counted. This requires
rapid response times for the transducer as weII as the
signaI processor. This technique is usefuI for Iow
intensity X-ray beams.
37
GAS FILLED TRANSDUCERS
When X-rays pass through an inert gas such as argon, xenon
or krypton, a Iarge number of positive ions and eIectrons are
produced for each X-ray quantum. Therefore, conductivity of
the gases increases from this phenomenon. Ionization chambers,
proportionaI counters and Geiger tubes are used to measure this
enhanced current.
A typicaI gas fiIIed transducer is shown here. The radiation
enters the chamber through a transparent window of mica or
beryIIium or aIuminum or myIar sheet. Each photon of
X-radiation interacts with one atom of argon and causes it to
Ioose one of its outer eIectrons. The photo eIectron Ioses its
kinetic energy by ionizing severaI hundred additionaI argon
atoms. The eIectrons move toward the centraI wire anode and
argon cations move towards the cyIindricaI metaI cathode.
38
GAS FILLED DETECTOR
39
AMPLIFICATION OF DETECTOR SIGNALS
40
This figure shows the effect of appIied voItage on the
number of eIectrons reaching the anode for each photon.
The region V
1
- V
2
,shows the number of eIectrons
reaching the anode which is fairIy constant for each photon.
In the region V
3
-V
4
,the number of eIectrons increases
rapidIy with appIied voItage. Here secondary ion pair
production occurs due to the coIIisions between the
acceIerated eIectrons and gas moIecuIes. Therefore, under
these conditions ampIification of the current occurs.
The region V
5
-V
6
shows that ampIification of the eIectricaI
impuIse is enormous but the current is independent of the
type and energy of the incoming radiation. The current is
actuaIIy governed by the geometry and gas pressure of the
tube. This region is known as the Geiger region.
41
SCINTILLATION COUNTERS
A scintiIIation counter consists of a transparent crystaI
of sodium iodide containing 0.2% thaIIium iodide
shaped in the form of a cyIinder shaped 3x4 inch in
each dimension. One surface is pIane which faces the
cathode of a PMT. As the incoming radiation passes
through the crystaI, its energy is Iost to the scintiIIator
which is subsequentIy reIeased in the form of photons
of fIuorescence radiation.
The fIashes of Iight produced are transmitted to the
photocathode which in turn produce an eIectricaI
puIse that can be ampIified and counted.
42
X-RAY FLUORESCENCE METHODS
As expIained earIier there are two types of XRF
instruments: WaveIength dispersive (WDXRF) and
energy dispersive (EDXRF).
In WDXRF an X-ray beam from a radioactive source is
coIIimated and dispersed into its component
waveIengths. Both singIe channeI (sequentiaI) or
muItichanneI (simuItaneous) detection systems are
empIoyed.
43
In singIe channeI instruments, the crystaI and the
transducer are set at specific angIes ( & 2 respectiveIy)
and the counting progresses untiI sufficient counts are
obtained for precise measurements. The instrument can
be automated to cover the entire 360
0
C. In such
instrument the movement of the crystaI and detector are
synchronized and the output is connected to the data
acquisition system.
The X-ray sources use chromium for Iong waveIengths
and tungsten target for short waveIengths. Such
instruments cost around 25 Iakh Rupees.
44
MuItichanneI instruments are costIier (> 150,000$). They
are usefuI for the determination of up to 24 eIements
simuItaneousIy. In these instruments, individuaI
channeIs of an appropriate crystaI and a detector are
arranged radiaIIy around an X-ray source and sampIe
hoIder . The crystaIs for most of the channeIs are fixed
at an appropriate angIe for a given anaIyte Iine.
For every eIement to be determined there wiII be a
dedicated transducer, ampIifier, puIse height detector
and counter or integrator. These instruments are
equipped with a computer for precise instrument
controI, data processing and dispIay of anaIyticaI
resuIts. The anaIysis is usuaIIy compIeted within a few
seconds to a few minutes.
45
ENERGY DISPERSIVE XRF
EDXRF has a poIychromatic source such as an X-ray
source or a radioactive materiaI, a sampIe hoIder, a
semiconductor detectors and various eIectronic
components for energy discrimination.
The anaIyticaI methodoIogy is very simpIe. There are
no moving parts and no coIIimators or crystaI
diffractor. Therefore the energy reaching the detector
is 100 times more than WDXRF instruments. Hence
Iow power X-ray tubes or weaker radioactive materiaIs
can be used as radiation sources which are cost
effective and more safer to use .
46
(a) Schematic design of EDXRF (b) Mars Mission Rover head
47
The sensor head in the rover contains 6 curium-244
sources that emits X-rays and 5.81 MeV particIes.
X-ray emission is stimuIated by bombardment by and
other subatomic particIes such as protons to
produce ParticIe Induced X-ray Emission (PIXE).
The X-ray detector is a new room temperature type that
exhibits Iow noise and high signaI to noise ratio for
most of the eIements.
48
XRF DETERMINATION OF MARTIAN ROCKS
I
n
t
e
n
s
i
t
y

c
o
u
n
t
s

p
e
r

s
e
c
o
n
d
Energy, keV
49
In muItichanneI EDXRF, aII the emitted X-ray Iines are
measured simuItaneousIy. Therefore increased
sensitivity and improved signaI to noise ratio are
obtained (FeIIgett advantage). Improved resoIution at
shorter waveIengths is another advantage.
50
A bench top EDXRF is avaiIabIe for the routine
determination of up to 12 eIements ranging from
sodium to uranium in industriaI sampIes. In this
Instrument, the radiation from an X-ray tube passes
through an appropriate fiIter and strikes the bottom of
the rotating sampIe. The fIuorescence signaI passes on
to the sampIe detector (siIicon) which provides signaIs
to the muItichanneI counting system.
The system uses a rhodium anode X-ray tube , five
programmabIe fiIters, a heIium purge system, a tweIve
positive sampIe changer and a spinner to rotate each
sampIe during the data acquisition processes to
reduce sampIe in homogeneity.
51
A BENCH TOP XRF MODEL
52
EDXRF pIots energy vs intensity (cps). The energy is in keV.
Quantitative anaIysis can be obtained by measuring the
peak heights. A rough estimate of the concentration can
be obtained by,
P
X
= P
S
. W
X
where P
X
is the reIative Iine intensity measured in terms of
number of counts per fixed time , W
X
is the weight fraction
of the desired eIement in the sampIe and P
S
is the reIative
intensity of the Iine under identicaI conditions if W
X
were
unity. The vaIue of P
S
is determined with a standard sampIe
of known composition or pure eIement.
53
ANALYTICAL METHODOLOGY
Matrix effects: Absorption of the X-rays from atoms weII
beIow the surface. The extent of attenuation depends
upon the mass absorption coefficient of the medium and
its composition. SignaI decreases.
Enhancement of fIuorescence may occur due to the
secondary excitation of the anaIyticaI Iine-signaI
increases.
This effect can be minimized by using a cIoseIy matched
standards except the added eIement.
54
AIternateIy, an internaI standard is introduced in aII
the standards and sampIes.
One can aIso diIute the sampIe and standards that
absorb X-rays onIy weakIy. Water, organic soIvents,
starch, Iithium carbonate, aIumina, boric acid, borate
gIass etc., Here matrix effects become essentiaIIy
constant.
55
APPLICATIONS
When X-ray absorption and enhancement effects are
minimaI, the XRF Iine intensity and eIement concentration
curves are Iinear. NIST standards provide such materiaIs
which can be used to caIibrate the instruments. TypicaI
appIication incIude:
(i) AnaIysis of rock sampIes for 22 eIements.
(ii) AIIoy composition controI during manufacture.
(iii) Lead and bromine, in aviation fueIs, Ca, BA, zinc in
Iubricating oiIs, pigments in paint sampIes.
(iv) Determination of chIorine, suIphides and suIphur
dioxide on O-toIuidine, siIver nitrate and sodium
hydroxide impregnated micropore fiIter papers.
(v) Determination of Fe, Cu, Zn in rice sampIes.
56
ADVANTAGES OF XRF
(i) The spectra are reIativeIy simpIe and not subject
to much interference.
(ii) Non destructive technique. Hence usefuI for
anaIysis of paintings, archeoIogicaI specimens,
jeweIIery , antique items etc.,
(iii) Speed and convenience of the procedure.
(iv) Precision and accuracy are comparabIe to
estabIished methods in cIassicaI techniques.
57
DISADVANTAGES OF XRF
(i) The concentration range varies from 0.01 - 100%.
(ii) Not usefuI for Iight eIements (Iess than 23 At. Numbers).
(iii) High cost of the instruments.
1
ATOMIC ABSORPTION SPECTROMETRY
The phenomenon of atomic absorption was first
discovered by WooIaston in 1802 in the spectrum of
sunIight. TiII then thought to be a continuous
spectrum, WooIaston made a remarkabIe observation
that the soIar spectrum was in fact interrupted by "dark
Iines" which was Iater confirmed by Fraunhofer in
1814. However, Brewster put forward the idea that
these dark Iines denoted tiII then by aIphabeticaI
markers are in fact due to the presence of vapors of
certain eIements in the sun's atmosphere. Therefore, it
foIIows that substances emitting specific radiations
are aIso capabIe of absorbing the same, causing the
spectrum of dark Iines in the bright background.
2
This phenomenon generaIIy known as Kirchoff's Iaw
was used to deduce the presence of oxygen, hydrogen,
sodium, iron, caIcium etc., in the soIar spectrum. The
experimentaI confirmation for this came from eIectric
arc or spark, when the spectraI source surrounded
by atomic vapors aIso showed dark Iines
because of absorption of the emitted radiations.
FoucaIt in France aIso demonstrated the reversaI of
spectraI Iines.
3
FRAUNHOFFER LINES
4
FRAUNHOFFER LINES
5
In 1902, Wood repeated the experiments of Kirchoff
and FoucaIt and proved concIusiveIy that by
introducing sodium vapor in the opticaI path of sodium
emission Iines (589.0 and 589.6 nm), a reduction in the
intensity of radiation occurs. By anaIogy with acoustic
resonance Iines, he aIso showed the possibiIity of
using these resonance effects to detect traces of
mercury. The potentiaI of this technique was not
recognized by anaIyticaI chemists and spectroscopists
tiII 1924.
6
Angerer and Joose pubIished the atomic spectra of
iron group metaIs, foIIowed by Frayne and Smita for
indium, aIuminum, gaIIium and tantaIum. MuIIer and
Pringshiem in 1930 pubIished the first atomic
absorption method of measuring the mercury content
in air. Even this did not evoke interest in the anaIyticaI
chemists for the determination of other eIements.
7
WaIsh, in 1955 deveIoped the first reaI appIication of
atomic absorption to chemicaI anaIysis. In the same
year AIkamade and Miatz described a doubIe beam
method of spectraI seIection with two fIames, the first
being the source and the second as atomizer. Since
then the atomic absorption spectrometry is in the
forefront of chemicaI anaIysis. The first commerciaI
atomic absorption instruments appeared in 1960s.
8
SpectacuIar advances in instrumentation, eIectronics,
automation and computers over the years have made
atomic absorption spectrometry, one of the most
reIiabIe anaIyticaI techniques of modern times
perhaps equaIIed onIy by atomic emission
spectrometry in terms of simpIicity, sensitivity,
specificity and speed of operation.
9
The deveIopment of eIectrothermaI atomization by
L'vov and Massmann pushed the detection Iimits of
atomic absorption technique to nanogram and
picogram and sometimes even up to femtogram
IeveIs.
Hydride generation atomic absorption spectrometry
for arsenic, antimony, bismuth, seIenium, teIIurium,
germanium, Iead and coId vapor mercury
determination have proved attractive accessories for
atomic absorption technique to make it the first
choice of anaIyticaI chemists throughout the worId.
10
The popuIarity of the atomic absorption spectrometry
can be gauged by the fact that more than 100 books,
10,000 pubIications and 5000 symposia and
conferences tiII date have appeared with appIications
in bioIogicaI, chemicaI, nucIear, industriaI products,
soiIs, environment etc.,
11
THEORETICAL CONCEPTS OF ATOMIC
ABSORPTION
The simpIest concept of atomic structure is that of the
positiveIy charged nucIeus containing protons and
neutrons surrounded by an equaI number of eIectrons
orbiting in space in the eIectric fieId created by the
protons. According to quantum mechanics, such a
system can exist in a stabIe state onIy if its energy is
quantized even at the Iowest energy IeveI or ground
state. AII other IeveIs are excited IeveIs, which can be
induced by mechanicaI or eIectromagnetic means. The
energies associated with these atomic states are in the
range of a few eIectron voIts represented by Grotian
diagram. Such a diagram for sodium is shown in the
next sIide.
12
ENERGY LEVEL DIAGRAM FOR SODIUM
13
Emission of Iight occurs when an atom reverts to a
state of Iower energy. Bohr's equation expresses the
conservation of energy by the reIation:
= E
1
- E
2
/ h (1)
or = c / = hc / E
1
- E
2
where c is the veIocity of Iight (2.99793 x 108 m/s), h is
the PIank's constant (6.62 x 10
-34
in SI Units), v is the
frequency, is the waveIength and E
1
, E
2
are the
energy IeveIs of excited and ground states.

14
Inserting the numericaI vaIues, we get:
= 1.23978 / E
Thus a transition from the resonance IeveI of sodium
(2.102 eIectron voIts) to ground state wouId correspond
to the emission of:
= 1.23972 / 2.102 = 0.5986 m or 589.6 nm
15
When a photon of frequency v interacts with an
atom of energy E
2
, the atom may be abIe to absorb the
photon thus raising it s energy to E
2
+ hv, provided the
new energy IeveI is equaI to one of the excited energy
IeveIs of that atom, Then we can write:
= E
1
- E
2
/ h (2)
Comparison of equations 1 and 2 shows that "An atom
can onIy absorb the radiations that it is abIe to emit".
This forms the basis of spectrometry.
v
16
The fundamentaI difference between emission
spectrometry and absorption spectrometry may be
defined as:
(i) For emission to occur, a number of atoms must be
in the excited state.
(ii) For atomic absorption to occur, a number of
atoms must be in the ground state.
The second condition is easiIy attained compared to
the former in that, free atoms in the ground state can
be easiIy generated in fIame compared to the excited
state, by fIame emission.
17
The characteristic absorption waveIengths for an atom
can be caIcuIated once the energy IeveIs are
experimentaIIy determined. But severaI waveIengths
are never observed which proves that some sort of
seIection ruIes be deveIoped. These ruIes were evoIved
empiricaIIy at first, and subsequentIy confirmed
by appIying quantum mechanicaI principIes to the
concept of transition probabiIities.
The probabiIity of a spontaneous emission by
transition between energy IeveIs E
1
and E
2
is defined
as the fraction of number of atoms that drop to Iower
IeveI per unit time.
18
MathematicaIIy,
dN
12
= AN
1
dt (3)
where A is the coefficient of proportionaIity termed as
Einstein emission coefficient. Higher the probabiIity of
transition, the greater is the intensity of emission. The
strongest emission Iines correspond to vaIues of A in
the range of 10
8
to 10
9
/ sec.
19
SimiIarIy if N
2
atoms in the Iower transition state are
irradiated by a radiation of frequency derived from
equation (1) and the voIume fIux density ( ), the
number of d N
2-1
of atoms that wiII absorb the
radiation in time dt is proportionaI to N
2
, ( ) and dt.
Hence,
dN
21
= B
2-1
, N
2
( ) dt (4)
v
v
v
v
20
The term B
21
is known as Einstein's absorption
coefficient. Now a days, the term osciIIator strength is
being used to denote the reIationship between B and
the totaI number of eIectrons.
B
21
= e
2
/ m h c f
2,1
(5)
where, e and m are the eIectronic charge and mass.
Thus for sodium D Iines, osciIIator strength is 0.23 and
0.47 and for potassium they are 0.35 and 0.70.
21
The emission Iifetime of any transition is approximateIy
10
-8
sec. However, if there are sufficient numbers of atoms,
steady state emission or absorption phenomena can be
observed within the experimentaI time frame. For atomic
absorption to occur , intense emission of the desired
eIement must be generated first. The radiations generated
from eIectric dipoIe, magnetic dipoIe, eIectric quadrupoIe
interactions give rise to such Iines among which eIectric
dipoIe are most important. Both Einstein emission (A) and
absorption coefficients () are non-zero, onIy if the IeveIs
invoIved are of opposite parity and if J = 1.
22
Using these seIection ruIes, resonance IeveI of an
atom may be defined as that of Iowest excited energy
IeveI that can interact with the ground state by a
transition of eIectric dipoIe type. The corresponding
waveIength is known as the resonance Iine.
Therefore, it foIIows that for a particuIar atom the
resonance Iine is the most intense of highest
osciIIator strengths and onIy this Iine is usefuI for
anaIysis provided the waveIengths are in the
200-600 nm range.
23
In practice, it is impossibIe to get a truIy
monochromatic Iine, but the energy is distributed
symmetricaIIy over a narrow waveband. The width of
a spectraI Iine is defined as the vaIue of where the
intensity is 50 percent of the totaI. This is caIIed as
haIf width. The shape and size of an absorption or
emission band is affected by severaI factors such as
naturaI broadening, DoppIer broadening, pressure
broadening and eIectric or magnetic fieId broadening
etc.,
24
PROFILE OF A RESONANCE LINE
25
NATURAL BROADENING
Due to the short Iifetime of energy states,
Heisenberg's uncertainty principIe is appIicabIe for
aII transitions. Thus a smaII broadening effect of
the order of a few miIIionth of a nanometer at 250 nm
occurs rising to about 10
-4
nm at 1m. This naturaI
width is infIuenced further by a variety of factors,
chief among them being the disordered thermaI
motion of the atoms and various types of coIIisions
of atoms.
26
DOPPLER BROADENING
27
If an atom emitting a radiation
0
moves with a
veIocity reIative to the observer, the observed
waveIength is given by,
=
0
+
0
/ c (6)
where c is the veIocity of Iight in the vacuum. Further,
if the atoms are in thermaI equiIibrium at temperature T,
their veIocities wiII have a MaxweIIian distribution.
v
v
28
The monochromatic absorption coefficient K as a
function of may be expressed as:
K() = K
0
exp { [ -( -
0
) /
d
2 (In 2)

]
2
} (7)
where
d
is the DoppIer haIfwidth reIated to T and the
atomic mass M by the equation:

d
= 7.16 x 10
7

0
( T/M )
1/2
(8)
The Iine is thus shown to have a Gaussian profiIe. It is
possibIe to caIcuIate the vaIues of
d
at 2000, 2500 and
3000 K and the Iine widths for these temperatures are of
the order of 30-50 mA.
0
29
PRESSURE BROADENING
Since the atoms in the vapour state are in a perpetuaI
state of motion, coIIision of atoms is inevitabIe
causing radiation quanta of sIightIy differing
frequencies to be absorbed or emitted. SeveraI types
of particIes may be invoIved in the coIIisions.
Interaction of eIectricaIIy charged particIes causes
Iine broadening known as 'Stark effect'. CoIIisions
with uncharged atoms Iead to van Der waaI's effect.
CoIIisions between atoms of the same type Ieading to
resonance broadening effect is referred as
'HoItsmark' effect. Since it is difficuIt to differentiate
between these three effects they are coIIectiveIy
referred as 'Lorentz' broadening.
30
The broadening of spectraI Iines reduces the Iifetime of the excited
state of the atoms. It aIso increases the Iine profiIe of the radiation.
The monochromatic absorption coefficient of the em radiation at a
waveIength is given by:
K() = K
0
/ 1 + [2 ( -
0
) L]
2
(9)
where K
0
is the maximum absorption coefficient and L is the haIf
width. The profiIe of this distribution is fIatter than DoppIer
broadening but both are aImost of the same order. The haIf width L
is thus a fraction of the frequency of coIIision (Z), which in turn is a
function of the temperature and the effective cross section defined
by:
= Z
0
2
/ C (10)
31
It may be noted that both DoppIer and Lorenz
broadening occur simuItaneousIy resuIting in a simiIar
but broader profiIe known as Voigt profiIe (K) which
may be mathematicaIIy expressed as:
(11)
Where a =
L
/
D
( In 2 )
1/2
,
w = -
0
/
D
2 (In 2 )
1/2
,
y = 2 /
D
(In 2 )
1/2
and
= distance to the point at which K() and
K
0
are the caIcuIated and the maximum
vaIue of the coefficient.
dy y) - (w a / e / a K ) (
2 2
-
y2 -
o

K
32
The curves are symmetricaI with a maximum at
0
.
Apart from DoppIer and Lorenz effects, Iine broadening
aIso occurs due to hyperfine structure exhibited by
many resonance Iines due to nucIear spin. Isotope shift
of the resonance Iines aIso contributes additionaIIy
to the Iine broadening. These effects are aIso
significant but not as prominent.
In essence, the sum totaI of aII these Iine broadening
effect is of the order of 0.0005-0.005nm, which
increases with increasing temperature and pressure.
The significance of peak width at haIf the peak height
has a profound effect on the emission characteristics
of radiation sources, (especiaIIy hoIIow cathode Iamps)
which wiII be discussed Iater.
33
MEASUREMENT OF ABSORPTION
Based on quantum physicaI description given earIier,
rigorous mathematicaI expressions have been derived
to determine the absorption coefficient, its variation
with N.f.I, effect of monochromator band width and
aIso of opticaI density. However, for practicaI
anaIyticaI purposes a physicaI understanding of these
phenomena is more reIevant which may be
Interpreted as foIIows.
34
A very narrow frequency intervaI is essentiaI for the
absorption of resonance radiation. However, it is
impossibIe to isoIate and obtain high intensity of
iIIumination in the range of 0.0005 - 0.005 nm from
continuum radiation sources. It wouId be too weak to
be of any practicaI use. To overcome this difficuIty, WaIsh
recommended that the radiation source shouId be made
of the anaIyte eIement onIy. Therefore onIy the resonance
Iine need to be separated from other spectraI Iines by a
monochromator.
35
Assuming that a monochromator isoIates a spectraI band
S covering the absorption Iine
0
(resonance Iine), the
totaI spectraI energy received by the detector is :
(13)
= I
0

S
(14)
= Area of the rectangIe ABCD

2 /
2 /
0 0
0
0
d I
S
S
I

36
Now if a homogeneous gas having an absorption K()
is interposed in a Iength of the radiation beam, the
energy within the band wiII decrease by the same
amount but the spectraI profiIe wiII have the same
shape.
37 37
Instead of considering the radiation per unit voIume, if
the totaI radiant fIux () is considered, then it may be
proved that the absorption factor and hence opticaI
density is proportionaI to the concentration of the free
atoms and to the path Iength in the absorbing medium
provided that the concentration is Iow and the spectraI
bandwidth is narrow. This is nothing but Beer - Lambert's
Iaw which can be expressed as:

tr
=
0
e
-x
v
N I
(16)
Where
0
and
tr
are the radiant fIuxes before and after
absorption in the path Iength I, x
v
is the spectraI
absorption coefficient and N is the number of atoms.
38 38
This expression may be rearranged in the famiIiar form,
Absorbance = A = Log
0
/
tr
= 2.303 x
v
N L (17)
The totaI number of free atoms in opticaI path cannot
be determined but it is not necessary for routine
appIications, as atomic absorption is a reIative
technique Iike any other spectroscopic techniques.
39 39
The physicaI conditions for highest sensitivity may be
summarized as foIIows:
The absorption Iine shouId have Iowest energy state and
highest popuIation of the atoms in the ground state.
If severaI resonance Iines are there, the one with highest
osciIIator strength has to be chosen.
EmpIoying a source of radiation, that emits a Iine of the
same waveIength but with Iower haIf width.
Path Iength may be increased within practicaI Iimits in
the absorbing medium since B-L Iaw states that the
absorption aIso increases according to the path Iength.
40 40
EmpIoying these conditions, we can in principIe, construct
an atomic absorption spectrometer using a hoIIow cathode
Iamp made of the same eIement as the anaIyte, an atomizer
to produce a popuIation of ground state atoms, a
monochromator with an entrance and exit sIit for coIIection,
dispersion and seIection of resonance Iine, a detector for the
measurement of radiation intensity foIIowed by an ampIifier
and a read out device. A schematic diagram of such a
system is shown here.
Source
Absorption
CeII
Monochromator Detector
AmpIifier DispIay
41 41
Depending upon the choice of the components and method of
operations severaI variants of atomic absorption spectrometers
resuIt, which are enumerated beIow,
1) SingIe beam DC instrument - This is the simpIest arrangement.
The earIiest AAS instruments were of this type.
2) SingIe beam AC instrument - By appIying the puIsed current to
the radiation source or by mechanicaIIy chopping the radiation
before it enters the absorption ceII.
3) DoubIe beam AC instrument - By using a rotating mirror/chopper
arrangement, the radiation is passed aIternateIy through the
fIame and around the fIame. Then it is possibIe to construct a
doubIe beam instrument. Both beams are recombined by a
semitransparent mirror pIaced behind the fIame. The eIectronics
of the system is designed to yieId directIy the ratio of the
transmitted radiation fIux to that of the incident radiation. The
stabiIity is aIso better.
42
SCHEMATICS OF ATOMIC ABSORPTION
SPECTROMETERS
43 43
4) MuIti eIement SimuItaneous AAS - Use of radiation
sources containing resonance Iines of severaI
eIements focused in to the absorption ceII permits
simuItaneous determination of severaI eIements.
However the optics and eIectronics need to be
suitabIy modified to handIe various signaIs readout
and printouts.
5) EIectrothermaI AAS - By substituting the absorption
ceII (i.e fIame) with an eIectricaIIy heated graphite
furnace, very efficient means of producing atomic
vapor can be achieved. This technique has gained
wide popuIarity since Iast 15 years permitting the
quantitative determination in ppb IeveIs (10
-9
g).
44 44
6) Hydride Generation AAS - Arsenic, antimony,
bismuth, seIenium, teIIurium, germanium, Iead etc,
are capabIe of forming their respective hydrides in
acidic medium. These compounds easiIy dissociate
into their metaIIic and non-metaIIic components
which, when introduced into the fIame (absorption
ceII ), permit not onIy their separation but aIso
estimation in ppb IeveIs (10
-9
g).
7) Mercury coId vapor AAS - Mercury has a unique
property of being reduced to metaIIic form directIy
from its combined state and aIso has a significant
vapor pressure which permits its determination at
room temperature. It onIy needs to be transported
to the absorption ceII. This technique is known as
coId vapor technique.
45 45
Over the years atomic absorption spectrometry as an
anaIyticaI technique has been accepted as a standard
method of anaIysis aII over the worId. An enormous
amount of Iiterature on the instrumentation, radiation,
sources, atomization techniques, optics, signaI handIing
and data presentation has been deveIoped. The advent
of computers has made it possibIe for maximum use of
automation, instrument controI and statisticaI data
evaIuation. On an average, more than 500 research
papers are being pubIished on the appIication of AAS to
various matrices every year.
Now we shaII discuss the detaiIed aspects of atomic
absorption spectroscopy.
46
AAS is the measurement of the absorption of em radiation
by the atoms in the gaseous state.
Free atoms do not undergo vibrationaI & rotationaI
transitions but onIy eIectronic transitions. Such excited
eIectron may return to ground state by atomic
emission, atomic fIuorescence or atomic absorption
phenomena.
The various energy states of an atom are described by n,I
and inner quantum number J.
SeIection ruIes permit L = 1 and n = any number.
47
For sodium atom the most IooseIy bound eIectron is
designated by,
3s
2
S
1/2
3 p
2
p
1/2 ,3/2
589.593 nm / 588.996 nm
4 p
2
p
1/2,3/2
330.294 nm / 330.234 nm
n p
2
p
1/2,3/2
In emission spectrum aII possibIe Iines are obtained.
48
Since aII eIements can be excited to their next higher
energy IeveI ,in theory any eIement can be determined by
atomic absorption spectrometry.
However, beIow 200 nm, anaIysis of As, Se, I, S , P etc., is
difficuIt owing to the incipient absorption by oxygen and
hot fIame gases. Cerium ,Thorium and other refractive
eIements aIso present difficuIty. ArtificiaI and radioactive
eIements can not be anaIyzed by atomic absorption
spectrometry.
49
THERMAL EXCITATION
It must be appreciated that for atomic absorption to occur ,
we have to produce a popuIation of atoms in the ground
state. This can be achieved by exposing a sampIe of the anaIyte
to high temperatures. At high temperatures prevaiIing in the
fIames, compounds decompose into ions, which in turn pick up
eIectrons to produce atoms.
The ratio of number of atoms N
j
in an excited state j to the
number of atoms in the ground state N
0
is given by,
N
j
P
j
N
0
P
0
where P
j
& P
0
are the statisticaI weights of the excited and
ground states , k is the boItzmann's constant and T is the
absoIute temperature .
=
. e
-Ej/KT