See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233461762

Atomic Absorption Spectroscopy: A Tutorial Review*

Article in Applied Spectroscopy Reviews · February 2007

DOI: 10.1081/ASR-100100844

CITATIONS READS
44 14,561

1 author:

Anthony Lagalante
Villanova University
79 PUBLICATIONS 1,600 CITATIONS

SEE PROFILE

All content following this page was uploaded by Anthony Lagalante on 03 June 2021.

The user has requested enhancement of the downloaded file.

APPLIED SPECTROSCOPY REVIEWS, 34(3), 173–189 (1999)

Atomic Absorption Spectroscopy:

A Tutorial Review*
ANTHONY F. LAGALANTE
Chemical Science and Technology Laboratory
Physical and Chemical Properties Division
National Institute of Standards and Technology†
Boulder, CO 80303

I. INTRODUCTION.......................................................................... 174
II. GENERAL DESCRIPTION.......................................................... 174
III. ATOMIZATION ............................................................................ 175
A. Flame Atomization................................................................. 175
B. Electrothermal Atomization ................................................... 179
IV. EXCITATION SOURCES ............................................................ 180
A. Hollow-Cathode Lamps ......................................................... 180
B. Electrodeless Discharge Lamps ............................................. 182
C. Continuum Sources ................................................................ 182
V. BACKGROUND CORRECTION................................................. 183
A. Continuum Source ................................................................. 183

*Contribution of the U.S. Government. Not subject to copyright in the United

States.
†Certain commercial instruments, materials, or equipment are identified in this
tutorial in order to specify adequately the experimental procedure or discussion. Such
identification does not imply recommendation or endorsement by the National Insti-
tute of Standards and Technology, nor does it imply that the materials or equipment
identified are necessarily the best available for this purpose.

173

Published by Marcel Dekker, Inc. www.dekker.com

174 LAGALANTE

B. Pulsed Hollow-Cathode Lamp............................................... 184

C. Zeeman Effect ........................................................................ 185
VI. INTERFERENCES ........................................................................ 185
A. Chemical Interferences .......................................................... 185
B. Ionization Interferences.......................................................... 186
C. Matrix Interferences ............................................................... 186
D. Spectral Interferences............................................................. 188
VII. CONTINUING RESEARCH ........................................................ 188
BIBLIOGRAPHY .......................................................................... 189

I. INTRODUCTION

Atomic absorption spectroscopy is one of the important instrumental

techniques for both quantitative and qualitative analysis of metallic and non-
metallic elements in inorganic or organic materials. In this tutorial review,
the principal components of atomic absorption instrumentation will be dis-
cussed. In order to understand the instrumental variables associated with
these particular instruments, a brief discussion of the theory behind atomic
absorption is presented, as well as some rudimentary aspects of the spectral
information that is obtainable.

II. GENERAL DESCRIPTION

The phenomenon of atomic absorption was first observed in 1802 with

Wollaston’s observation of dark bands in the emission spectrum of the sun.
In 1859, Kirchoff and Bunsen correctly explained Wollaston’s observation
by showing that the dark bands were due to the absorption of emission radia-
tion by ground-state gas-phase atoms in the sun. However, the process of
absorption of radiation by atomic vapors would not be used as a quantitative
analytical tool until nearly a century later when Alan Walsh fabricated the
first analytical atomic absorption spectrophotometer in 1953.
Since its invention, atomic absorption spectroscopy (AAS) has gained
acceptance as a standard method for the analysis of both metallic and nonme-
tallic elements. AAS is widely accepted because the technique is element-
selective and it provides analytical sensitivities at the parts-per-million level
and less (where 1 part-per-million is 0.0001%). Example application areas
that use AAS as a routine method of metal analysis include various forms of
industrial manufacturing, geology, medicine, and agriculture. The technique
of AAS is divided into two steps: the conversion of an analyte molecule into
ATOMIC ABSORPTION SPECTROSCOPY 175

its constituent gas-phase atoms (atomization), and the subsequent absorption

of radiation by these free atoms.
Quantitation in AAS is accomplished by measuring the amount of ab-
sorbing species produced at a given analytical wavelength. The principles of
quantitation in AAS rely on the adherence to the Beer-Lambert law, namely
that the increase in absorbance (measured as a decrease in transmittance) has
a linear relationship to the concentration of gas-phase atoms. The instrumen-
tation required for AAS (see Figure 1) is similar to that required by other
high-resolution spectroscopic techniques. AAS differs from other spectro-
scopic techniques predominantly in the nature of the radiation source that is
used and the use of heat to produce the absorbing species. The atomic lines
produced by the radiation source must be of narrow enough width to over-
come limitations due to the finite bandpass of the instrument. Usually, this
requirement is met by using a line source, such as the hollow cathode lamp.

III. ATOMIZATION

The atomization step can be successfully executed by either flame or

electrothermal methods. In either of the two methods, thermal energy is used
to vaporize the analyte material and to break the chemical bonds within the
component molecules. The percentage of the atoms of interest in the mole-
cule that are converted to the gas-phase during the atomization process is
called the atomization efficiency. The atomization efficiency is an important
limiting factor in the analytical sensitivity of AAS.

A. Flame Atomization

Heating the sample in a flame is the most frequently used method for
producing free atoms in AAS. In flame AAS, the sample to be analyzed must
first be dissolved in a suitable solvent. The analyte-containing solution is
typically converted to an aerosol and passed into the flame with a pneumatic
nebulizer. Alternative techniques, such as ultrasonic nebulization have also
been used. Pneumatic nebulizers utilize the flame oxidant gas as a nebuliza-
tion gas. The differential pressure of the oxidant gas, between the solution to
be analyzed and the burner head, draws solution through a sample capillary

FIG. 1. The essential components of an atomic absorption spectrophotometer.

176 LAGALANTE

into the oxidant gas stream (see Figure 2 for a typical burner head configura-
tion). The stream is expanded through a small orifice to produce droplets
with a broad size distribution. Viscosity, surface tension, and density of the
analyte-containing solution are important factors influencing the droplet size
obtained. Removal of large droplets (>20 µm) by a downstream impinger
results in less than 10% of the actual sample being converted to an aerosol
and reaching the burner head. The removal of the large droplets greatly re-
duces the flame noise, but fewer molecules of the analyte are available for
atomization. Therefore, fewer free atoms are produced, and analytical sensi-

FIG. 2. The nebulizer and burner assembly for flame AAS. (Reproduction
courtesy of Perkin-Elmer Corporation)
ATOMIC ABSORPTION SPECTROSCOPY 177

tivity is sacrificed. The aerosol portion of the sample is diluted in approxi-

mately 100 L min−1 of expanded flame gases. Although a high signal-to-noise
ratio can be obtained by using flame atomization, the nebulization procedure
and subsequent dilution by the flame gases greatly reduces the sensitivity of
flame AAS.
Typical flame gases used in AAS include air-acetylene or nitrous ox-
ide-acetylene, providing maximum operating temperatures of 2400°C and
2800°C, respectively. The air-acetylene flame typically uses a burner path-
length of 10 cm, whereas the nitrous-oxide acetylene flame requires a shorter
pathlength (typically 5 cm). For safety reasons, care must be taken to avoid
flashback occurring when using the nitrous oxide-acetylene flame, although
the risk has been greatly reduced through computer control of the flame
gases. The actual temperature of the flame, determined by the oxidant/fuel
ratio, dictates the number of atoms in the excited state, Nex, relative to the
number of atoms in the ground state, Ngr . This ratio is defined by the Max-
well-Boltzmann equation:

Nex gex −E/kT

= e (1)
Ngr ggr

where E is the excited-state transition energy, k is the Boltzmann constant, T

is the system temperature (expressed in Kelvin), and ggr and gex are the statis-
tical weights of the ground and excited states, respectively. According to the
Maxwell-Boltzmann expression, the number of atoms in the excited state is
negligible compared to the number of atoms in the ground state at typical
flame temperatures. This distribution is fortuitous because the signal mea-
sured in AAS is due only to atoms that undergo electronic transitions from
the ground-state to the excited-state. Any processes that remove atoms from
their ground state are deemed interferences, and these are discussed later.
Within the flame, the sample in aerosol form undergoes various processes:
evaporation of the solvent, volatilization of the metal analyte, dissociation of
the analyte to the gaseous metallic element, and excitation of the gaseous
metallic element to the excited state. These are by no means trivial processes,
and the flame chemistry must be tuned by adjusting the oxidant/fuel ratio to
optimize the flame temperature for gaseous metal atom formation. Flames
that operate in fuel-rich conditions are reducing flames, sometimes called
“yellow” flames, whereas flames operating under fuel-lean conditions are
oxidizing flames, sometimes called “blue” flames. The hotter flame tempera-
ture obtained by using the nitrous oxide-acetylene flame is necessary to atom-
ize elements with higher heats of vaporization such as the rare earth elements.
Of the 68 elements listed in Table 1, half are best analyzed using the nitrous
oxide-acetylene flame.
178 LAGALANTE

Table 1.
Elemental Atomic Lines and Recommended Flame Gases
for Analysis

˚ Flame ˚ Flame
Element λ (A) Gases Element λ (A) Gases

Ag 3281 A-Ac Nb 3344 N-Ac

Al 3093 N-Ac Nd 4634 N-Ac
As 1937 A-Ac Ni 2320 A-Ac
Au 2428 A-Ac Os 2909 N-Ac
B 2497 N-Ac p 2136 N-Ac
Ba 5536 N-Ac Pb 2833 A-Ac
Be 2349 N-Ac Pd 2476 A-Ac
Bi 2231 A-Ac Pr 4951 N-Ac
Ca 4227 A-Ac Pt 2659 A-Ac
Cd 2288 A-Ac Rb 7800 A-Ac
Co 2407 A-Ac Re 3460 N-Ac
Cr 3579 A-Ac Rh 3435 A-Ac
Cs 8521 A-Ac Ru 3499 A-Ac
Cu 3247 A-Ac Sb 2176 A-Ac
Dy 4212 N-Ac Sc 3912 N-Ac
Er 4008 N-Ac Se 1960 A-Ac
Eu 4594 N-Ac Si 2516 N-Ac
Fe 2483 A-Ac Sm 4297 N-Ac
Ga 2874 N-Ac Sn 2863 A-Ac
Gd 4079 N-Ac Sr 4607 A-Ac
Ge 2651 N-Ac Ta 2715 N-Ac
Hf 2866 N-Ac Tb 4326 N-Ac
Hg 2536 A-Ac Tc 2615 A-Ac
Ho 4104 N-Ac Te 2143 A-Ac
In 3039 A-Ac Ti 3653 N-Ac
Ir 2640 A-Ac Tl 2768 A-Ac
K 7665 A-Ac Tm 3718 N-Ac
La 5501 N-Ac U 3585 N-Ac
Li 6708 A-Ac V 3184 N-Ac
Lu 3360 N-Ac W 2551 N-Ac
Mg 2852 A-Ac Y 4102 N-Ac
Mn 2795 A-Ac Yb 3988 N-Ac
Mo 3133 A-Ac Zn 2139 A-Ac
Na 5890 A-Ac Zr 3601 N-Ac
A-Ac = Air-Acetylene
N-Ac = Nitrous Oxide-Acetylene
ATOMIC ABSORPTION SPECTROSCOPY 179

The atomization of an analyte is not homogeneous throughout the

flame. The maximum production of free atoms in the flame is dictated by
the variation in temperature in the different zones of the flame and by the
rate of diffusion of the flame gases that serve to dilute the free-atom popula-
tion. To achieve maximum sensitivity, the radiation from the line source must
pass through the area of the flame that contains the greatest number of free
atoms. This condition is met by optimizing the height and alignment of the
burner head relative to the radiation beam as it passes through the flame
parallel to the major axis of the burner.

B. Electrothermal Atomization

To overcome the sensitivity problems associated with flame AAS, the

nebulizer and burner assembly of a flame AAS instrument can be replaced
with an electrothermal atomizer. Electrothermal atomizers include the graph-
ite furnace (Figure 3), carbon rod, and tungsten ribbon atomizers. In these
atomizers, a 5 to 50 µL aliquot of the sample is placed in the resistively
heated element and then heated stepwise to the temperatures conducive to
gaseous atom formation. The heating element design commonly used in the
graphite furnace atomizer is a modified L’vov platform. A typical heating
program involves a solvent evaporation step (110°C), an ashing step for the
removal of any volatile organic materials (450–900°C) and a 5 to 10 second
atomization step (2000–3000°C). The entire heating cycle can be accom-
plished in less than one minute. In contrast to flame AAS, electrothermal

FIG. 3. A graphite furnace electrothermal atomizer. (Reproduction courtesy

of Perkin-Elmer Corporation)
180 LAGALANTE

AAS samples are not diluted by flame gases, nor do they depend on nebuliza-
tion efficiencies. The entire sample to be analyzed is vaporized within a small
internal volume (,2 mL) by using electrothermal atomizers, and this results
in an increase in sensitivity of three orders of magnitude over flame AAS.
Typically, analytes present in picogram quantities can be determined by elec-
trothermal AAS. Furthermore, electrothermal atomizers can be used to di-
rectly analyze solid materials without any form of modification. A major
limitation of electrothermal atomizers, however, is the reduced precision of
the results obtained, compared to flame AAS. The difference is that the atom-
ization occurs as a single event in time compared to the continuous process
of flame atomization. In electrothermal atomizers, the lifetime of a free atom
in the optical path is typically on the order of 0.01 seconds, in comparison
to flame AAS, in which a sample can be continuously nebulized in a flame
and where the signal can be averaged over long periods. Replicate results are
difficult to obtain with electrothermal atomizers because of the reproducibil-
ity associated with the furnace operation, although computer control of the
furnace has reduced the impact of many of these difficulties.

IV. EXCITATION SOURCES

The full width at half-height of the absorption line produced by either

flame or electrothermal atomization as measured by emission spectroscopy
is approximately 0.002 nm. Factors that modify (broaden) this line width
include natural, pressure, resonance, and Doppler broadening. To provide
both sensitivity and selectivity, the line sources used for excitation must be
stable and narrower than the full width at half-height of the absorption of the
analyte. The narrow full-width-at-half-height criterion ensures that all of
the radiation produced by such a line source is available for absorption by
the gas-phase atoms. Line sources typically used include the hollow cathode
lamp and the electrodeless discharge lamp. With proper instrumental design,
even continuum sources can be used.

A. Hollow-Cathode Lamps

The line source of radiation most frequently employed in AAS is the

hollow-cathode lamp (HCL)—see Figure 4 for a schematic rendering. Com-
mercially, hollow cathode lamps are available as both a single element and
multi-element sources. A potential of 350V to 500V (producing a current of
1 to 50 mA) is placed across the cathode and anode terminals. The element
selectivity of the HCL is achieved by fabricating the cathode from, or coating
the interior of the cathode with, the element of interest. The anode and cath-
ATOMIC ABSORPTION SPECTROSCOPY 181

FIG. 4. Schematic of a hollow-cathode lamp.

ode are sealed in a Pyrex tube containing a quartz window at the opposite
end for transmission capability throughout the ultraviolet spectral region. The
interior of the tube is evacuated and filled with an inert gas, such as neon.
Under the applied potential, the neon atoms become ionized at the anode and
are accelerated toward the cathode. The kinetic energy of the accelerated
neon ions is imparted to the material of the cathode upon impact, causing
metal atoms to be sputtered off the cathode surface into the gas phase. The
sputtered metal atoms undergo collisions with other neon ions, producing
excited state metal atoms. Upon relaxation to the ground state, the metal
atoms emit radiation at their characteristic wavelengths. The full width at
half-height of the atomic line produced is approximately 0.0002 nm, and
therefore all of the radiation produced is available for absorption by the atom-
ized sample. The intensity of the atomic line radiation produced depends on
the number of sputtered metal atoms, which depends on the kinetic energy of
the neon ions, which, in turn, is dictated by the lamp current. Consequently, a
higher lamp current produces a greater spectral line intensity from the source,
which in turn provides a greater analytical sensitivity. However, at high lamp
currents, an excess of metal atoms is sputtered from the cathode surface. The
resulting dense cloud of metal atoms absorbs radiation emitted from nearby
excited-state metal atoms. When self-absorption of this type occurs within
the HCL, the level of radiation available for absorption by the free atoms
from the sample, produced via atomization, is reduced and sensitivity is lost.
This process of self-absorption is called “self-reversal” and dictates the maxi-
mum operating current of the HCL. Following emission of radiation, the
gaseous metal atom will ideally redeposit on the cathode surface, making it
available for removal by other accelerated neon ions. Alternatively, the metal
atoms may migrate and be deposited on the Pyrex wall of the HCL, a process
that reduces the lifetime of the lamp. Restrictive shields placed around the
cathode, and optimization of the wall geometry of the HCL can minimize
migration of the gas-phase atoms to the HCL walls.
182 LAGALANTE

The multi-element HCL operates under the same principles governing

the singleelement HCL, except that the cathode is fabricated from pressed
alloys or powders of several elements, instead from just a single element.
For a cost comparable to that of the single-element HCL, the multi-element
HCL provides simultaneous detection of several metals. Despite the eco-
nomic advantages, the multi-element HCL typically has a shorter lifetime
than the single-element HCL, primarily because the metal atoms of the cath-
ode have different degrees of volatility. In time, the element with the highest
volatility will redeposit preferentially over the surface of the cathode, causing
the other metals of the cathode to be rendered inaccessible to the impacting
neon ions. The sensitivities of the inaccessible metal atoms will decrease
over time and ultimately shorten the effective lifetime of the HCL.

B. Electrodeless Discharge Lamps

The electrodeless discharge lamp (EDL) is another atomic line source

used in AAS. The EDL is also available in a multi-element format. The EDL
is fabricated by sealing a small amount of the metal or metal iodide of inter-
est in a quartz tube filled with an inert gas, usually argon, at reduced pres-
sures. The EDL is placed in a microwave, radio-frequency field to ionize the
inert gas. The metal or metal iodide species collides with the free electrons
from the ionized inert gas, producing excited-state metal atoms. In a process
analogous to that described for the HCL, these excited-state metal atoms emit
the characteristic atomic lines of the metal as they relax to the ground state.
In comparison to the HCL, the EDL produces an atomic line source of greater
intensity, resulting in enhanced analytical sensitivities. However, the EDL re-
quires a separate power supply that increases the instrumental cost of the AAS.

C. Continuum Sources

For purely economic reasons, the use of a continuum source in AAS is

attractive because it eliminates the need to purchase multiple HCL sources.
However, in practice the bandpass of a typical monochromator (0.1 to 0.2
nm) is much greater than the atomic line-width of the absorbing atoms (0.002
nm). Consequently, a large fraction of the radiation produced by the source
falling on the detector is unabsorbed. The small amount of radiation that is
actually absorbed results in a low analytical sensitivity. Continuum sources
utilizing scanning Echelle grating monochromators have largely overcome
the bandpass difficulty, but overall they still provide lower sensitivity in com-
parison to the narrow lines produced by the HCL. When using a continuum
source, the high-intensity xenon arc lamp is preferred over the deuterium
continuum source because of the higher analytical sensitivity obtainable.
ATOMIC ABSORPTION SPECTROSCOPY 183

V. BACKGROUND CORRECTION

In AAS, a background correction is necessary to compensate for non-

atomic absorption of the radiation produced by the source and for emission
from interfering species in the flame. When non-atomic absorption occurs, a
falsely high analyte concentration is detected because the signal measured
according to the Beer-Lambert law is due to the concentration of atomic and
non-atomic species at a given wavelength. In the case of non-analyte emis-
sion, a falsely low analyte concentration is detected because, at a given wave-
length, both the non-absorbed radiation from the source and the background
emission are detected. Background emission is an excess of radiation that is
detected which is caused by emitting species in the flame. This excess radia-
tion includes (1) the band spectra produced by various molecular fragments
and radical species, such as OH, O2, CH, CO, CN, CH, and NH, and (2) the
continuum spectra produced by hot solids and chemiluminescent reactions.
The three primary methods for background correction are (1) the addi-
tion of a broadband or continuum source, (2) pulsing the HCL source, and
(3) application of the Zeeman effect to the source or sample.

A. Continuum Source

The continuum radiation sources previously described find more appli-

cability in AAS as an additional source used for background correction. In-
corporation of a continuum source (hydrogen or deuterium lamp) in combi-
nation with a line source permits correction for non-atomic absorption.
Non-atomic absorption is broad in comparison to the narrow absorption line
of atoms produced by atomization. At the wavelength produced by the HCL,
the absorption is due to both atomic and non-atomic species distributed in
the flame. Over the bandpass of the monochromator, the absorption of the
continuum source radiation is attributable to primarily non-atomic species.
The difference between the absorption due to the HCL and the absorption
due to the continuum source represents the absorption due to atomic species
in the flame. By modulating the signal received at the detector between the
line source and the continuum source (i.e., chopping), a background-cor-
rected absorbance signal is obtained. For background correction at longer
wavelengths, where the hydrogen or deuterium sources are less intense, a
tungsten or quartz halogen lamp can be used. When using continuum sources
for background correction, care must be taken to align the line source and
the continuum source so that the irradiation passes through the same spatial
volume of the flame (i.e., the sources must be co-axially aligned). This align-
ment will ensure measurement of the same species in the flame, this allows
184 LAGALANTE

FIG. 5. A schematic of a dual-beam AAS instrument operating in the back-

ground mode. (Reproduction courtesy of Perkin-Elmer Corporation)

for the fact that the flame chemistry will differ between the various tempera-
ture zones within the flame.

B. Pulsed Hollow-Cathode Lamp

A pulsed HCL can also be used as a method of background correction.

Often called “the poor man’s continuum source,” it relies on the previously
mentioned process of self-reversal in the HCL. By pulsing the HCL between
low (5 to 20 mA) and high (100 to 500 mA) currents, a background-corrected
spectrum can be obtained. At low currents, absorption by species in the flame
again accounts for both atomic and non-atomic absorption. At high currents,
the lamp emission spectra are broadened owing to Doppler effects and, over
the bandpass of the instrument, it mimics a continuum source. Therefore, the
absorbance in the high-current mode is due to non-atomic absorbance, and
the two signals can be subtracted to give a corrected absorbance signal. Al-
though the pulsed design requires no special lamp alignment, as a continuum
source does, the HCL must be stable at the high currents used. In contrast to
the continuum-background correction method, the high current of the HCL
does not produce a constant value over all wavelengths passing to the detec-
tor. Therefore, the atomic absorption can constitute a small but significant
amount of the high current signal. The non-negligible amount of atomic ab-
ATOMIC ABSORPTION SPECTROSCOPY 185

sorption during the high current part of the cycle reduces the sensitivity of
the calibration curve.

C. Zeeman Effect

Zeeman background correction is yet another method of providing a

background-corrected signal for non-atomic absorption during atomization.
Two configurations of Zeeman splitting are often used, the analyte-shifted
Zeeman correction and the source-shifted Zeeman correction. Zeeman correc-
tion is based on the principle that the atomic lines of the analyte or the source
can be split into degenerate spectral lines of differing polarization in an ap-
plied magnetic field. For example, in analyte-shifted Zeeman splitting, an
atomic line is split into σ (perpendicularly polarized) and π (parallel polar-
ized) components. The π component remains at the same wavelength as the
original atomic line, but the wavelength of the σ component is shifted away
from the original wavelength. Absorption measured from the source radiation
at the π wavelength can be attributed to both the atomic and background
absorbance, whereas absorption measured at the σ wavelength can be attrib-
uted solely to the background absorbance. By placing a polarizer either be-
fore (dc technique) or after (ac technique) the atomizer, the σ and π compo-
nents can be measured independently and subtracted to give the background-
corrected absorbance.
The analyte shifted Zeeman correction is easier to implement in elec-
trothermal AAS than in flame AAS because the atomization chamber in an
electrothermal atomizer is smaller, requiring a less powerful magnet. In
source-shifted Zeeman corrections, an EDL is often used because the HCL
is not a magnetically stable lamp. In this case, the source atomic line is
split into its π and σ components and is measured independently. Zeeman
background correction is attractive because it is essentially a double-beam
technique contained in a single-beam instrument. However, the technique
sacrifices sensitivity because the full intensity of the analytical atomic line is
compromised when it is separated into its lower-intensity degenerate spectral
lines.

VI. INTERFERENCES
A. Chemical Interferences

Of all the interferences, chemical interferences are predominantly re-

sponsible for removing metal atoms from their ground state. Chemical in-
terferences arise during the vaporization processes in atomization. Radical
species are produced that can react with the metal compound undergoing
186 LAGALANTE

vaporization to produce metal hydroxides, metal hydrides, or metal oxide

species. In flame AAS, these species cannot be dissociated at cooler flame
temperatures. This fact effectively reduces the number of free metal atoms
produced in the flame, resulting in a loss in sensitivity. One method to reduce
the effects of chemical interferences is to adjust the flame-gas stoichiometry.
A fuel-rich flame produces a reducing environment, with correspondingly
fewer of the reactive oxide species. An alternate approach is to use a hotter
flame, such as the nitrous oxide-acetylene flame, to vaporize and dissociate
any refractory metal oxides. A third commonly employed technique is to use
a releasing agent that can compete with the metal-oxide formation equilib-
rium. An example often cited is the addition of lanthanum when analyzing
for calcium in the presence of high concentrations of phosphate ions. In this
example, the lanthanum competes with the calcium for complex formation
with the phosphate ions in the flame. The lanthanum phosphate complex is
more stable than the calcium phosphate complex and predominates, resulting
in a greater number of free calcium atoms in the flame providing a higher
sensitivity for the calcium. Another useful scavenging agent is ethylenedi-
aminetetraacetic acid (EDTA). EDTA forms stable complexes with most
metals that can be vaporized in a flame. Furthermore, the carbon atoms of
EDTA surrounding the metal effectively produce a reducing environment
during the dissociation step.

B. Ionization Interferences

To some extent, all metallic atoms produced at the temperatures of the

flame undergo ionization. This can certainly be minimized by using cooler
flames such as the acetylene-air flame and by adjusting the flame-gas stoichi-
ometry. Another method is to chemically alter the ionization equilibrium
within the flame. The ionization of a metal written as an equilibrium expres-
sion is:

M0 → Mn+ + ne− (2)

Therefore, elements that provide electrons to the flame will shift this equilib-
rium to the left. Such a species is called a “radiation buffer” or an “ionization
buffer”. Radiation buffers are species that are easily ionized, such as cesium.
To ensure similar atomization efficiencies for quantitation, the buffer is typi-
cally added in excess to both standards and unknowns.

C. Matrix Interferences

A matrix interference involves the bulk physical properties of the sam-

ple to be analyzed. In practice, first a blank solvent is atomized and the
ATOMIC ABSORPTION SPECTROSCOPY 187

instrument zeroed so that the measured voltage at the photo-multiplier tube

(PMT) corresponds to 100% transmittance. Solutions containing the analyte
of interest are subsequently atomized, and their atomic absorbances are mea-
sured. The Beer-Lambert law is used to relate the unknown analyte concen-
tration to the concentration of standard solutions. To ensure that the Beer-
Lambert law is followed, atomization efficiencies must be matched between
the calibration curve and the unknown sample. In flame AAS, matrix interfer-
ences will affect nebulization efficiencies during the atomization stage, pri-
marily because nebulization efficiencies are affected by the physical proper-
ties of viscosity, surface tension, and density. For quantitative results using
the Beer-Lambert law, these properties must be matched in both standard and
unknown solutions. Instead of taking the time to matrix-match standards and
unknown solutions, if the unknown is sufficiently concentrated, it can be
diluted to approximately match the physical properties of the standard solu-
tions, which are typically dilute.
The most common way to overcome a matrix interference is to use the
method of standard additions (Figure 6). This method effectively creates a
calibration curve by using incremental additions of a standard solution to an
unknown. The incremental additions are of a minuscule volume so as not to
alter the bulk physical properties of the unknown, effectively providing a
matched matrix. The plot obtained is extrapolated to the x-axis intercept. The
absolute value of the x-axis intercept gives the analyte concentration in the

FIG. 6. An example of the method of standard additions for the analysis of

copper concentration in an unknown.
188 LAGALANTE

unknown sample. When using the method of standard additions, care must
be taken to ensure that one is operating in the linear region of the calibration
curve. In AAS, the Beer-Lambert typically holds at low concentrations, and
the calibration plot often exhibits nonlinearity at higher analyte concentra-
tions. This is caused by stray source radiation that reaches the detector and
to flame-diffusion effects, both of which can severely reduce the accuracy of
the measurement and the results.

D. Spectral Interferences

Spectral interferences in AAS, although rare, do occasionally occur and

can be corrected by manipulation of instrumental variables. Spectral interfer-
ences arise when radiation of a wavelength different from that of the wave-
length of the element of interest falls on the detector. An example of direct
overlap of spectra produced by atomic species in the unknown occurs when
sodium and magnesium are both present. When analyzing for dilute amounts
of magnesium in a sample with a high concentration of sodium, the 285.28
nm sodium line will overlap the 285.21 nm magnesium line. Therefore, in
the flame, sodium atoms will absorb radiation from the magnesium HCL,
resulting in a falsely high magnesium concentration detected at the PMT. As
most monochromators do not have a resolution sufficient to separate the two
closely spaced lines, an alternative wavelength for magnesium quantitation
must be selected in such cases. Another similar spectral interference occurs
when using a multi-element HCL. In this instance, lines from other elements
in the HCL occur near, or overlap, the line of the analytical element of inter-
est. This is the case when using the nickel 232.0 nm line in a multi-element
HCL containing iron, chromium, and manganese. To overcome the problem,
either the slit can be narrowed to permit transmission of only the nickel
wavelength, or a single-element HCL can be used. Other spectral interfer-
ences include emission interferences and non-atomic absorption. These have
been addressed in the section on background correction.

VII. CONTINUING RESEARCH

Although the major innovations in AAS appear to have been accom-

plished, research still continues in many laboratories to improve atomization
efficiencies. Despite the fact that flame AAS instruments are the most com-
monly used type of instrument, new electrothermal atomizers are being de-
veloped to increase sensitivity levels of AAS. Modern computer control has
reduced much of the operator error in use of both flame and electrothermal
atomizers, and if the precision of electrothermal AAS can be increased to
ATOMIC ABSORPTION SPECTROSCOPY 189

rival flame AAS, the technique could become more widely accepted. AAS is
finding increasing use as an element-selective detector in liquid chromatog-
raphy, gas chromatography, and even supercritical fluid chromatography. The
element selectivity of AAS has secured the place of AAS in diverse analyti-
cal laboratories, just as the inefficiency of the atomization process will surely
stimulate novel sensitivity breakthroughs in the future.

BIBLIOGRAPHY

Brown, S. D. “Zeeman Effect-Based Background Correction in Atomic Absorption

Spectrometry,” Anal. Chem. 1977, 49, 1269A.
Christian, G. D., Feldman, F. J. Atomic Absorption Spectroscopy, John Wiley and
Sons, New York, 1970.
Harnley, J. M. “Multielement Atomic Absorption with a Continuum Source,” Anal.
Chem. 1986, 58, 933A.
Ingle, J. D., Crouch, S. R. Spectrochemical Analysis, Prentice Hall, New Jersey, 1988.
Robinson, J. W. Atomic Spectroscopy, Marcel Dekker, New York, 1996.
Slavin, W. “Atomic Absorption Spectroscopy,” Anal. Chem. 1982, 54, 685A.
Slavin, W. “Flames, Furnaces, Plasmas,” Anal. Chem. 1986, 58, 589A.
Zander, A. T., “Atomic Emission Sources for Solution Spectrochemistry,” Anal.
Chem. 1986, 58, 1139A.
Sneddon, J., “Atomic Absorption Spectrometry,” Handbook of Instrumental Tech-
niques for Analytical Chemistry, ed. Frank Settle, pp. 373–393, Prentice Hall
PTR, NJ, 1997.
Dean, J. A., “Atomic Spectroscopy,” Analytical Chemistry Handbook, McGraw Hill,
NY, 1995.
Kellner, R. A., Mermet, J.-M., Otto, M., and Widmer, H. M. Eds., “Elemental Analy-
sis: Atomic Absorption Spectrometry,” Analytical Chemistry, pp. 453–463, Wi-
ley-VCH Verlag GmbH, Weinheim, 1998.
Parsons, M. L., “Atomic Absorption and Flame Emission Spectrometry,” Analytical
Instrumentation Handbook, 2nd Edition, Galen Wood Ewing Ed., pp. 257–325,
Marcel Dekker, NY, 1997.
View publication stats