American Journal of Anal yt ical Chemistry, 2011, 2, 710-717
doi:10.4236/ajac.2011.26081 Published Online October 2011 (http://www.SciRP.org/journal/ajac)
Measuring Conditions for the Determination of Lead in
Iron-Matrix Samples Using Graphite Atomizers
with/without a Platform in Graphite Furnace Atomic
Absorption Spectrometry
Syun Morimoto, Tetsuya Ashino, Kazuaki Wagatsuma*
Institute for Materials Research, Tohoku University, Ka tahira, Sendai, Japan
Received June 27, 2011; revised August 2, 2011; accepted August 15, 2011
Abstract
In graphite furnace atomic absorption spectrometry (GF-AAS), the atomization process of lead occurring in
graphite atomizers with/without a platform plate was investigated when palladium was added to an
iron-matrix sample solution containing trace amounts of lead. Absorption profiles of a lead line were meas-
ured at various compositions of iron and palladium. Variations in the gas temperature were also estimated
with the progress of atomization, by using a two-line method under the assumption of a Boltzmann distribu-
tion. Each addition of iron and palladium increased the lead absorbance in both the atomizers, indicating that
iron or palladium became an effective matrix modifier for the determination of lead. Especially, palladium
played a significant role for controlling chemical species of lead at the charring stage in the platform-type
atomizer, to change several chemical species to a single species and eventually to yield a dominant peak of
the lead absorbance at the atomizing stage. Furthermore, the addition of palladium delayed the peak after the
gas atmosphere in the atomizer was heated to a higher temperature. These phenomena would be because the
temperature of the platform at the charring stage was elevated more slowly compared to that of the furnace
wall, and also because a thermally-stable compound, such as a palladium-lead solid solution, was produced
by their metallurgical reaction during heating of the charring stage. A platform-type atomizer with palladium
as the matrix modifier is recommended for the determination of lead in GF-AAS. The optimum condition for
this was obtained in a coexistence of 1.0 × 10–2 g/dm3 palladium, when the charring at 973 K and then the
atomizing at 3073 K were conducted.
Keywords: Graphite Furnace Atomic Absorption Spectrometry, Platform Atomizer, Gas Temperature,
Matrix Modifier, Lead, Iron, Palladium
1. Introduction
Several metallic elements, such as Cu, Pb, As and Bi,
exert a negative influence on the quality of manufactured
steel materials; for example, their segregation at grain
boundaries or their precipitation may cause cracking dur-
ing hot rolling in the steel-making process [1,2]. These
elements are called tramp elements, because they are
hardly removed through the conventional refining proc-
ess and thus remain as impur ities during recycling of the
steel materials. It is therefore required that trace amounts
of these elements are accurately determined to control
the mechanical properties of commercial steel products.
Graphite-furnace atomic absorption spectrometry (GF-
AAS) [3] is usually employed for the determination of
metallic elements having a concentration level of 10–6%
[4-7], and thus can be applied to the tramp elements in
steel materials [7].
It is known in GF-AAS that the absorption signal of an
analyte element is varied by co-existing matrix elements
in the sample solution, called a matrix effect, even if the
temperatures of GF are optimized. Therefore, the matrix
elements are usually pre-removed in the analytical ap-
plication of GF-AAS; however, it is in most cases fol-
lowed by a time-consuming chemical procedure. Koba-
yashi el al. reported on the direct determination of tramp

S. MORIMOTO ET AL.
711
elements in a co-existence of large amounts of iron in
order to analyze actual steel samples more rapidly, indi-
cating that an optimization of the operating condition s in
GF-AAS enabled these elements to be accurately deter-
mined, even when iron was not removed from the sample
solution [7]. Their result was interesting, because the
matrix effect would be reduced through the formation of
any compound with iron.
In practical GF-AAS, a matrix modifier comprising
various types of reagents is usually added when a sample
solution is injected into the furnace, for obtaining ana-
lytical values with higher sensitivity as well as better
precision [3]. The matrix modifiers in GF-AAS have
been investigated by many researchers, [8-13] and in the
case of a metallic modifier, the effect would be derived
from a thermally stable compound which is produced
through any metallurgical reaction between the analyte
element and the modifier element. Hirokawa et al. first
indicated that the role of a metallic matrix mo difier could
be considered using a phase diagram between the analyte
element and the modifier element [12]. The effectiveness
of a matrix modifier can be estimated from the thermal
properties of solid solutions and/or intermetallic com-
pounds appearing in the phase diagram [12]. We invest-
tigated absorption profiles of a cadmium line at the at-
omization stage, in order to discuss the matrix modifier
effect of palladium, iron, and a mixture of palladium and
iron for the determination of cadmium in GF-AAS [13].
It indicated that palladium or iron each became an effect-
tive additive for the cadmium determination probably
because their reactants having lower vapor pressure, ap-
pearing in the palladium-cadmium and the iron-cadmium
phase diagrams, could reduce a loss of cadmium at the
charring stage [13].
In order to introduce sample species rapidly into a hot
environment, a platform technique, which produces a
delay of heating between the graphite furnace and the gas
atmosphere, is extensively applied in GF-AAS, as was
first developed by L’Vov [14]. Use of a platform plate is
known to be effective for the determination of easily-
volatile elements in GF-AAS [15-17]. Several studies
have been conducted concerning the diffusion process on
a platform plate [18] or between a platform plate and a
furnace wall [19], indicating that a temperature gradient
between the platform and the furnace body could be a
determini ng factor for the funct i on of a platf orm plate.
This paper describes the fundamental characteristics of
GF-AAS in the determination of lead, which is a typical
tramp element in steel materials, under the condition
where no separatio n proced ures of iron ar e prepared. It is
expected in this system that the iron element itself acts as
a matrix modifier because large amounts of iron coexist
in the sample solution. Palladium was further added to
this sample solution as another matrix modifier. This
situation can be considered to be the determination of
lead using an iron-palladium binary matrix modifier,
which becomes a more practical example in discussion
of the matrix modifier. In this paper, vaporization/at-
omization of lead in the graphite furnace with or without
a platform plate was monitored by temporal change in
the absorbance of a lead atomic line, when added
amounts of iron and/or palladium were varied in the
sample solution. Furthermore, a temporal variation in the
gas temperature was in-situ observed during the atomiza-
tion stage by using the two-line method. The atomization
process of lead occurring in the graphite furnace is dis-
cussed, especially focusing on the effect of the platform
plate in it.
2. Experimental
2.1. Apparatus
A simultaneous multi-element atomic absorption spec-
trometer (Z-9000, Hitachi Corp., Japan) with an auto-
sampler system was employed, which could be equipped
with four individual hollow cathode lamps as the light
source. In this study, a hollow cathode lamp of lead
(Hamamatsu Photonics Corp., Japan) was installed to
measure the atomic resonance line at 283.3 nm, which
was assigned to the transition from the 6s26p2 3P0
(0.0000 eV) to the 6s26p7s 3P1 (4.3769 eV) levels, and
two iron hollow cathode lamps (Hamamatsu Photonics
Corp., Japan) were also installed to monitor two different
absorption lines of iron. Their discharge currents were
set to be 5.0 mA. A background correction was con-
ducted through the measurement program by using a
dc-polarized Zeeman effect. The graphite furnaces em-
ployed were a non-pyrocoated tube without a platform
(Part No. 180-7400, Hitachi Corp., Japan) and a pyro-
coated tube with a platform (Part No.190-6007 and
190-6008, Hitachi Corp., Japan). The method for con-
trolling the temperature of these furnaces was based on
the measure ment wi th an irrad iativ e ther mo meter in s ta l l e d
in the spectrometer, by monitoring the radiation from the
furnace; therefore, the temperature of the furnace wall
was directly estimated. The absorption values for each
spectral line were recorded on a personal computer
through an analogous-to-digital converter at an interval
of 0.02 s, and a temporal variation in the gas temperature
at the atomization stage was directly estimated on the
computer. Sampling of the absorbance data was con-
trolled by a trigger signal from the spectrometer.
2.2. Reagents and Procedure
A lead stock solution of 1.0 g/dm3, a palladium stock so-

S. MORIMOTO ET AL.
712
lution of 3.0 g/dm3, and an iron stock solution of 2.0
g/dm3 were prepared by dissolving each pure metal
(99.9%) with 7 M-nitric acid for lead and palladium or
with 7M-nitric acid containing a small amount of hydro-
chloric acid for iron. They were further diluted with
de-ionized water for a working solution containing lead
of 1.0 × 10–4 g/dm3 and appropriate amounts of palla-
dium and/or iron. The concentration of palladium or iron
in the working solution was each varied from 0, 1.0 ×
10–4, 1.0 × 10–3, 1.0 × 10–2, to 1.0 × 10–1 g/dm3. The pre-
pared sample solutions were injected into the furnace at a
volume of 2 × 10–5 dm3.
The temperature program during heating of the fur-
nace was optimized when a test solution containing lead
of 1.0 × 10–4 g/dm3 and iron of 1.0 × 10–3 g/dm3 was
measured at various charring and atomizing temperatures.
The drying stage was carr ied out at 353 - 423 K for 30 s
and the cleaning stage at 3273 K for 10 s. The charring
temperature was determined at 973 K for lead, because it
began to be vaporized at the charring stage at higher
temperatures. The charring stage was conducted by
holding the corresponding charring temperature for 30 s.
The atomizing stage was held for 8 s at a temperature of
3073 K. At the drying, charring, and cleaning stages,
argon gas was introduced into the furnace at a flow rate
of about 0.2 dm3/min to remove water and other volatile
species; however, no gas flow was required only at the
atomization stage to keep the analyte species in the fur-
nace and thus to obtain a better sensitivity.
2.3. Two-Line Method
The principle of the two-line method has already been
described elsewhere [20]. This method is based on the
difference in the number density between two energy
levels, which can be determined by a characteristic tem-
perature under thermodynamic equilibrium [21]. The
temperature can be estimated from the absorbance ratio
between two spectral lines of a probe element whose
lower energy levels are different. Variations in the gas
temperature directly depend upon the spatial and tempo-
ral distribution of the probe element in the furnace and
thus relate to its diffusion behavior [20]. The gas tem-
perature is usually different from the furnace wall tem-
perature, which is monitored in conven tional temperature
control for GF-AAS. We discussed the diffusion of ana-
lyte elements in the furnace with the progress of atomi-
zation, by comparing the gas temperatures which were
estimated between iron and nickel as the probe element
[22]. The gas temperature could yield dynamic informa-
tion on the gas atmosphere where the sample volume was
expanding in the graphite furnace.
In this study, a pair of iron atomic lines, Fe I 372.0 nm
and Fe I 373.7 nm, was employed as probe lines for the
temperature measurement. The former line is assigned to
the transition from the 3d64s2 5D4 (0.0000 eV) to the
3d64s4p 5F5 (3.3319 eV) levels having a gA value of
1.467 × 108 s–1, and the latter line to the transition from
the 3d64s2 5D3 (0.0516 eV) to the 3 d64s4p 5F4 (3.3682 eV)
levels having a gA value of 0.994 × 108 s
–1 [23]. The
energy difference between the 3d64s2 5D4 and the 3d64s2
5D3 levels is the most important factor in determining the
temperature [13].
2.4. Phase Diagram
In our previous paper, we discussed the matrix modifier
effect of iron and palladium on the determination of
cadmium by using the corresponding phase diagrams
[13]. It indicated that both the elements became effective
modifiers because their compounds could reduce a loss
of cadmium at the heating stages before the atomization
while cadmium alone was easily evaporated.
Lead is less volatile than cadmium. The temperature
representing a va por pressure of 1.3 Pa is rep orted at 573
K for cadmium and at 991 K for lead, although their
melting points are similar: 594 K for cadmium and 601
K for lead. On the other hand, palladium and iron are
much less evaporated compared to lead. The temperature
representing a vapor pressure of 1.3 Pa is reported at
1839 K for palladium and at 1720 K for iron. It can be
thus considered that the addition of palladium or iron is
generally effective in the determination of lead in
GF-AAS due to hindering of the evaporation loss, when
large amounts of them comprise the sample matrix in-
cluding a trace amount of lead.
Lead and cadmium have a similar metallurgical fea-
ture on the phase diagram relating to iron or palladium,
where similar metallic phases are formed at the corre-
sponding compositions and temperatures. In a lead-pal-
ladium binary alloy system, lead can be soluble in the
solid palladium up to the composition of ca 30 atomic %
Pb, where a stable solid solution between palladium and
lead is produced [24]. The formation of the solid solu tion
is expected to reduce the vaporization of lead, because
the melting temperature becomes much higher compared
to pure lead. On the other hand, in a lead-iron binary
alloy system, they are mutually insoluble not only in the
liquid state but in the solid phase all over the chemical
compositions, meaning that neither solid solution phases
nor intermetallic compounds are produced between lead
and iron [24]. In this case, a modifier effect by any
compound formation is not expected. However, because
a two-phase separation occurs at temperatures of more
than the melting poin t of lead due to a large difference in
specific gravity between lead and iron, lead clusters

S. MORIMOTO ET AL.
713
would be covered with excess amounts of iron matrix,
which eventually reduces the vaporization of lead physic-
cally.
3. Results and Discussion
3.1. Absorption Profile of Lead Line during the
Atomization Stage
Figure 1 shows temporal changes in the absorbance for
the lead and the iron lines during an atomization stage,
together with those of the wall and the gas temperatures,
when the iron element coexists or is not in the sample
solution. A graphite tube without a platform (a normal
tube cuvette) was employed as the atomization source.
After a charring temperature of 973 K had been kept for
30 s, it took 1.0 s to rise from the charring stage to an
atomizing stage of 3073 K and then the temperature was
held for 6 s, as shown using a straight line in Figure 1.
The gas temperature in the atomization process was es-
timated from the absorption of the iron sp ectral lines: Fe
I 372.0 nm and Fe I 373.7nm, after the charring stage.
The absorbance for lead of 1.0 × 10–4 g/dm3 was meas-
ured when iron of 1.0 × 10–3 g/dm3 was in the sample
solution or absent. The lead absorbance reaches a maxi-
mum at a duration time of 0.40 - 0.42 s when the wall
temperature is on the way of heating to the atomizing
temperature programmed at 3073 K, whereas an iron
peak appears at a duration time of 1.4 s after the wall
temperature rises up to 3073 K. As also shown in Figure
1, a maximum peak of the gas temperature appears to be
2550 K at a duration time of 1.6 s, whereas the peak of
the lead absorbance appears earlier at lower gas tem-
perature. This effect is because the absorption of lead
Figure 1. Temporal variations in the absorbance of Pb I
283.3 nm (a, b) and Fe I 372.0 nm (c), and in the wall (d)
and the gas temperatures (e) at an atomization stage after a
charring at 973 K for 30 s. The absorbance of the Pb I line
is measured with (a)/without (b) an iron matrix of 1.0 × 10–3
g/dm3. A normal tube cuvette is employed as the atomiza-
tion source.
begins to be detected when the furnace wall enables lead
atoms to be vaporized, regardless of the gas temperature
in the furnace, which indicates that the vaporization of
lead is mainly controlled by thermal conduct with the
graphite furn ace. It is also found in Figure 1 that the lead
absorbance in the iron-containing sample is 1.45-times
larger than that in the iron-free sample. Therefore, the
iron element becomes a matrix modifier to improve the
detection sensitivity of lead in GF-AAS.
Figure 2 shows temporal variations in the absorption
profile of lead and iron, in the same samples and under
the same heating conditions as in Figure 1 except that a
graphite tube with a platform (a platform cuvette) is em-
ployed as the atomizer. The lead absorbance of the
iron-free sample has a broad maximum peak at a dura-
tion time of 1.28 s, whereas that of the iron-containing
sample comprises the first peak at 1.28 s and the second
peak at 1.68 s having much larger intensity. It is likely to
see that the broad peak in the iron-free sample also com-
prises the first peak having larger intensity and the sec-
ond peak being a shoulder. Differing from the result of
the normal tube cuvette (see Figure 1), these absorp-
tion peaks appear after the furnace wall is heated up to
the atomizing temperature programmed at 3073 K, and
the absorption peak of iron also appears after a slight
delay at a duration time of 2.68 s. The reason for the
peak splitting of the lead absorbance would be that two
different chemical species of lead are left at the charring
stage, which is attributed to lower temperature of the
platform compared to the wall temperature. The tem-
perature of the platform is elevated more slowly due to
indirect irradiative heating ; as a result, the appearance of
the absorbance peaks becomes more delayed and broad-
ened, compared to those using the normal tube cuvette.
The splitting of the absorption peak is not favorable for
the determination of lead in analytical applications. The
Figure 2. Absorption profiles corresponding to Figure 1,
when a platform cuvette is employed instead of the normal
tube cuvette.

S. MORIMOTO ET AL.
714
addition of iron seems to improve this drawback. It
should be noted in Figure 2 that a maximum peak of the
gas temperature becomes also delayed to 2.7 s; however,
the maximum value of 2540 K is similar to the corre-
sponding gas temperature in the case of the normal tube
cuvette (2550 K). This implies that a similar gas atmos-
phere is eventually generated in both the atomizers under
the same heating program.
3.2. Iron Concentration in Iron-Lead Binary
System
Figure 3 shows absorbance profiles of the lead line at
atomization stages where the iron concentration is varied
from 0 to 1.0 × 10–1 g/dm3 in a sample solution cont a i n i n g
lead of 1.0 × 10–4 g/dm3. In this case, the normal tube
cuvette was employed as the atomizer. The temperature
was raised to the programmed atomizing temperature of
3073 K for 1 s, after the charring stage had been con-
ducted at a temperature of 973 K for 30 s. The lead ab-
sorbance is gradually elevated with an increase in the
iron concentratio n up to 1.0 × 10–2 g/dm3, indicating that
the iron element becomes a matrix modifier for the lead
determination. The maximum peaks of the lead absorb-
ance appear at duration times of 0.4 - 0.5 s and they are
not so shifted with the iron concentration. This observa-
tion implies that, although the addition of large amounts
of iron can reduce a vaporization loss of lead during the
charring stage, the iron matrix does not keep the lead
species at the atomization stage, where lead and iron
atoms would be atomized individually according to each
thermal property. It results from the fact that lead and
iron are insoluble with each other and never form ther-
mally-stable compounds, which is predicted from the
Figure 3. Absorption profiles of the absorbance of Pb I
283.3 nm at atomization stages of a lead solution of 1.0 ×
10–4 g/dm3 containing different amounts of iron: 0.0 (a), 1.0
× 10–4 (b), 1.0 × 10–3 (c), 1.0 × 10–2 (d), 1.0 × 10–1 g/dm3 (e),
when a normal tube cuvette is employed as the graphite
furnace.
corresponding phase diagram [24]. In iron-cadmium bi-
nary system, we reported a similar dependence of the
cadmium absorbance on the concentration of added iron
[13], which was also attributed to a metallurgical feature
of the iron-cadmium system being analogous to the
iron-lead system.
Figure 4 shows absorbance profiles of the lead line in
the iron-lead binary system when the platform cuvette is
employed as the atomizer. The appearance of the ab-
sorbance peak becomes more delayed and broadened,
compared to that using the normal tube cuvette as show n
in Figure 3, because the platform is heated more slowly
and indirectly by irradiative heating. In addition, the re-
sponse of the lead absorption profile is largely varied by
adding excess amounts of iron to the sample solution. A
dominant peak of the lead absorbance could be obtained
at an iron concentration of 1.0 × 10–1 g/dm3, whereas the
absorption peak was split in an absence of iron. This
phenomenon is probably because large amounts of iron
added contribute to the formation of an stable chemical
species of lead on the platform at the charring stage.
Therefore, the addition of iron is recommended as a ma-
trix modifier for the determination of lead in the case of a
platform cuvette. In our study, the measurement could be
optimized when the platform cuvette was employed with
iron additive of 1.0 × 10–1 g/dm3.
3.3. Palladium Concentration in Palladium-Lead
Binary System
Figure 5 shows absorbance profiles of the lead line at an
atomization stage in the normal tube cuvette, when the
palladium concentration is varied from 0 to 1.0 × 10–1
g/dm3 in a sample solution containing lead of 1.0 × 10–4
g/dm3. The temperature program was the same as in the
iron-lead system as shown in Figure 3. The lead absor-
Figure 4. Absorption profiles corresponding to Figure 3,
when a platform cuvette is employed instead of the normal
tube cuvette.

S. MORIMOTO ET AL.
715
Figure 5. Absorption profiles of the absorbance of Pb I
283.3 nm at atomization stages of a lead solution of 1.0 ×
10–4 g/dm3 containing different amounts of palladium: 0.0
(a), 1.0 × 10–4 (b), 1.0 × 10–3 (c), 1.0 × 10–2 (d), 1.0 × 10–1
g/dm3 (e), when a normal tube cuvette is employed as the
graphite furnace.
bance becomes elevated by the addition of palladium up
to 1.0 × 10–2 g/dm3, and t he ma ximu m p eak are observed
at longer duration time with increasing concentration of
palladium. The shift in the maximum peak can be at-
tributed to the formation of a thermally-stable compound
between palladium and lead; in this case, it may be a
palladium-matrix solid solution having higher melting
temperature and lower vapor pressure at the charring
temperature, which can be predicted from the palladium-
lead phase diagram [24]. This alloy phase reduces loss of
lead at the charring stage as well as contributes to a delay
of the vaporization at the atomizing stage.
Figure 6 shows absorbance profiles of the lead line in
the palladium-lead binary system when the platform
cuvette is employed as the atomizer. It is likely to say
that the variation in the ab sorption profiles is in a similar
manner to those observed i n t he i ron-l ead sy st em as shown
in Figure 4. The addition of palladium can eliminate the
splitting of the lead absorption peak. The first peak of the
lead absorbance drastically decreases when palladium is
added to the sample solution, and then the second peak
appears at longer duration time with an increase in the
concentration of palladium. Therefore, palladium is also
an effective matrix modifier in the determination of lead
when the platform cuvette is employed as the atomizer.
In our study, the optimum condition for this was
obtained in a coexistence of 1.0 × 10–2 g/dm3 Pd.
As shown in Figures 5 and 6, the lead absorbance is
largely reduced by adding palladium of 1.0 × 10–1 g/dm3.
This effect would be because the vaporization of lead is
suppressed by the excess addition of palladium at the
atomization stage; however, the exact reason for this
cannot be understood in our work. In the analytical ap-
plication using palladium, appropriate amounts of palla-
dium should be added within the concentration range
where the absorbance of lead does not de cr ease.
3.4. Addition of Palladium in Iron-Lead Binary
System
Figure 7 shows absorption pro files of the lead line when
palladium, at the concentration of 0 to 1.0 × 10–1 g/dm3,
is added to a sample solution containing lead of 1.0 ×
10–4 g/dm3 and iron of 1.0 × 10–3 g/dm3. In this case, a
platform cuvette was employed as the atomizer, by using
the same heating program as in Figures 4 and 6.
The effect of palladium on the absorption profile in the
palladium-iron binary additive is almost similar to that in
the case of palladium alone as shown in Figure 6. It in-
dicated that palladium could promote to form a dominant
chemical species at the charring stage, enabling a strong
Figure 6. Absorption profiles corresponding to Figure 5,
when a platform cuvette is employed instead of the normal
tube cuvette.
Figure 7. Absorption profiles of the absorbance of Pb I
283.3 nm at atomization stages where different amounts of
palladi um: 0. 0 (a), 1. 0 × 10–4 (b), 1.0 × 10–3 (c), 1.0 × 10–2 (d),
1.0 × 10–1 g/dm3 (e), are added to a mixture solution con-
taining lead of 1.0 × 10–4 g/dm3 and iron of 1.0 × 10–3 g/dm3,
when a platform-type cuvette is employed as the graphite
furnace.

S. MORIMOTO ET AL.
716
absorption peak to appear at the atomizing stage. Little
synergy effect of palladium and iron was able to be ob-
served in the palladium-iron-lead ternary system, imply-
ing that any metallurgical interaction with palladium
rather than iron would determine the absorption profile
of lead.
As shown in Figure 7, the gas temperature was also
estimated in this system. Temporal changes in the gas
temperature were almost the same at the atomization
stage, regardless of the palladium concentration in the
sample solution; therefore, the gas atmosphere became
unchanged when palladium was added to the iron sam-
ples. The maximum peak of the lead absorbance is more
delayed with an increase in the concentration of palla-
dium added; for instance, the maximum peak appears at
a duration time of 2.0 s in the sample containing palla-
dium of 1.0 × 10–2 g/dm3 and the lead atoms are intro-
duced into an argon atmosphere having higher gas tem-
peratures (1000 - 1500 K), which may reduce any nega-
tive reactions for atomic absorption such as clustering.
4. Conclusions
The data in this paper demonstrate the atomization proc-
ess of lead in two different types of atomizers: a normal
tube cuvette and a platform cuvette, with/without iron
and/or palladium added as a matrix modifier in GF-AAS.
Absorption profiles of the Pb I 283.3-nm line were in-
vestigated when the chemical composition of iron and
palladium was varied. Each addition of iron or palladium
to a lead sample solution increases the absorbance of the
lead line in the case of the normal tube cuvette, and fur-
thermore, it is much more effective in the platform cu-
vette. The addition of palladium contributes to producing
a dominant chemical species of lead at the charring stage
in the platform cuvette, and the reaction would occur at
relatively low temperatures to reduce a loss of lead at the
charring stage. The resulting chemical species of lead is
atomized on the platform which is heated more slowly
than the furnace wall. As a result, the lead species is in-
troduced into the gas atmosphere which is sufficiently
heated for the delay period. This effect can be a reason
for use of the platform cuvette for the actual analytical
application in GF-AAS. The optimum measuring con-
dition for the platform cuvette was obtained in a coexis-
tence of 1.0 × 10–2 g/dm3 palladium, when the charring at
973 K and then the atomizing at 3073 K were con-
ducted.
5. Acknowledgements
The authors gratefully acknowledge financial support by
a grant from Steel Industry Foundation for the Advance-
ment of Environment Protection Technology, Japan.
6. References
[1] L. Savov, E. Volkova and D. Janke, “Copper and Tin in
Steel Scrap Recycling,” Materials Geoenvironment, Vol.
50, No. 3, 2003, pp. 627-640.
[2] K. Kunishige and M. Hatano, “Surface Hot-Shortness of
Steels Induced by a Small Amount of Copper and Tin
from Scrap Steels and Its Suppression Methods,” Mate-
rial Science Forum, Vol. 539-543, 2007, pp. 4113-4118.
doi:10.4028/0-87849-428-6.4113
[3] J. A. Broekaert, “Analytical Atomic Spectrometry with
Flames and Plasmas,” Wiley-VCH Verlag, Weinheim,
2002, pp. 146-171.
[4] F. Betts and A. Yau, “Graphite-Furnace Atomic-Absorp-
tion Spectrometric Determination of Chromium, Nickel,
Cobalt, Molybdenum, and Manganese in Tissues Con-
taining Particles of a Cobalt Chrome Alloy,” Analytical
Chemistry, Vol. 61, No. 11, 1989, pp. 1235-1238.
doi:10.1021/ac00186a012
[5] T. Ashino, K. Takada and K. Hirokawa, “Determination
of Trace Amounts of Selenium and Tellurium in
High-Purity Iron by Electrothermal Atomic-Absorption
Spectrometry after Reductive Coprecipitation with Pal-
ladium Using Ascorbic-Acid,” Analytical Chimica Acta,
Vol. 297, No. 3, 1994, pp. 443-451.
doi:10.1016/0003-2670(94)00226-6
[6] T. Itagaki, T. Ashino and K. Takada, “Determination of
Trace Amounts of Gold and Silver in High-Purity Iron
and Steel by Electrothermal Atomic Absorption Spec-
trometry after Reductive Coprecipitation,” Fresenius
Journal of Analytical Chemistry, Vol. 368, No. 4, 2000,
pp. 344-349. doi:10.1007/s002160000455
[7] K. Kobayashi, S. Hasegawa, S. Itoh, K. Ide, H. Yamagu-
chi and Y. Yamada, “Analysis of Iron and Steels for
Tramp Elements by Graphite Furnace AAS,” Tetsu-To-
Hagané, Vol. 90, No. 2, 2004, pp. 86-91.
[8] T. W. May and W. G. Brumbaugh, “Matrix Modifier and
Lvov Platform for Elimination of Matrix Interferences in
the Analysis of Fish-Tissues for Lead by Graphite-Fur-
nace Atomic-Absorption Spectrometry,” Analytical Chemi-
stry, Vol. 54, No. 7, 1982, pp. 1032-1037.
doi:10.1021/ac00244a005
[9] M. Taga, H. Yoshida and O. Sakurada, “Determination of
Tin by Graphite-Furnace AAS with Matrix Modifier,”
Bunseki Kagaku, Vol. 36, No. 10, 1987, pp. 597-600.
doi:10.2116/bunsekikagaku.36.10_597
[10] K. Matsumoto, “Palladium as a Matrix Modifier in
Graphite-Furnace Atomic-Absorption Spectrometry of
Group IIIb-VIb Elements,” Analytical Sciences, Vol. 9,
No. 11, 1993, pp. 447-453. doi:10.2116/analsci.9.447
[11] Y. Morishige, K. Hirokawa and K. Yasuda, “The Role of
Metallic Matrix Modifiers in Graphite-Furnace Atomic-
absorption Spectrometry,” Fresenius Journal of Analyti-
cal Chemistry, Vol. 350, No. 6, 1994, pp. 410-412.
doi:10.1007/BF00325614

S. MORIMOTO ET AL.
717
[12] K. Hirokawa, K. Yasuda and K. Takada, “Graphite-Fur-
nace Atomic-Absorption Spectrometry and Phase-Dia-
grams of Alloys,” Analytical Sciences, Vol. 8, No. 3,
1992, pp. 411-417. doi:10.2116/analsci.8.411
[13] S. Morimoto, T. Ashino and K. Wagatsuma, “Role of a
Binary Metallic Modifier in the Determination of Cad-
mium in Graphite Furnace Atomic Absorption Spec-
trometry,” Analytical Sciences, Vol. 26, No. 7, 2010, pp.
809-813. doi:10.2116/analsci.26.809
[14] B. V. L’Vov, “Electrothermal Atomization—Way to-
ward Absolute Methods of Atomic-Absorption Analysis,”
Spectrochimica Acta, Vo l . 33B, N o . 5, 1 9 7 8 , pp. 153-193.
[15] M. L. Cervera, A. Nevarro, R. Montoro, M. Gerardia and
A. Salvador, “Platform in Furnace Zeeman-Effect Atomic-
absorption Spectrometric Determination of Arsenic in
Beer by Atomization of Slurries of Sample Ash,” Journal
of Analytical Atomic Spectrometry, Vol. 6, No. 6, 1991,
pp. 477-481. doi:10.1039/ja9910600477
[16] H. D. Narrce, C. M. Ohl and M. Stoppler, “Metal Analy-
sis in Difficult Materials with Platform Furnace Zeeman-
Atomic Absorption Spectrometry: I. Direct Determina-
tion of Cadmium in Crude Oil and Oil Products,” Inter-
national Journal of Environmental Analytical Chemistry,
Vol. 18, No. 4, 1984, pp. 267-279.
doi:10.1080/03067318408077008
[17] D. Ma, Y. Okamoto, T. Kumamaru and E. Iwamoto,
“Determination of Gallium by Graphite Furnace Atomic
Absorption Spectrometry with Combined Use of a Tung-
sten-Coated L’Vov Platform Tube and a Chemical Modi-
fication Technique,” Analytica Chimica Acta, Vol. 390,
No. 1-3, 1999, pp. 201-206.
doi:10.1016/S0003-2670(99)00144-0
[18] F. Fagioli, C. Locatelli, R. Vecchiett and G. Torsi, “Sur-
vey of the Diffusion Process in Atomic-Absorption Spec-
trometry with the Platform-in-Furnace Technique,” Jour-
nal of Analytical Atomic Spectrometry, Vol. 3, No. 1,
1988, pp. 159-162. doi:10.1039/ja9880300159
[19] T. M. Mahmood and K. W. Jackson, “Wall-to-Platform
Migration in Electrothermal Atomic Absorption Spec-
trometry .1. Investigation of the Mechanism of Chloride
Interference on Thallium,” Spectrochimica Acta, Vol.
51B, No. 9-10, 1996, pp. 1155-1162.
[20] H. Shimabukuro, T. Ashino and K. Wagatsuma, “Tem-
poral Variations in Gas Temperature in an Atomization
Stage of Cadmium and Tellurium Evaluated by Using the
Two-Line Method in Graphite Furnace Atomic Absor-
ption Spectrometry,” Analytical Sciences, Vol. 24, No. 9,
2008, pp. 1165-1170. doi:10.2116/analsci.24.1165
[21] P. W. J. M. Boumans, “Theory of Spectrochemical Exci-
tation,” Prenum Press, New York, 1966.
[22] T. Ashino, S. Morimoto and K. Wagatsuma, “Role of a
Binary Metallic Modifier in the Determination of Cad-
mium in Graphite Furnace Atomic Absorption Spectro-
metry,” Analytical Sciences, Vol. 26, No. 7, 2010, pp.
809-813. doi:10.2116/analsci.26.809
[23] J. R. Fuhr, G. A. Ma rtin and W. L. Wiese, “Atomi c Tran -
sition Probabilities, Iron through Nickel,” Journal Physi-
cal and Chemical Reference Data, Vol. 17, Supplement
No. 4, 1988, pp. 25-26.
[24] M. Hansen, “Constitution of Binary Alloys,” McGraw
Hill, New York, 1958, pp. 695, 1094-1095.