You are on page 1of 5

Determination of traces of As, B, Bi, Ga, Ge, P, Pb, Sb, Se, Si and Te

in high-purity nickel using inductively coupled plasma-optical


emission spectrometry (ICP-OES)
S. Thangavel, K. Dash, S.M. Dhavile, A.C. Sahayam
n
National Centre for Compositional Characterization of Materials, Bhabha Atomic Research Centre, ECIL-Post, Hyderabad-500062, India
a r t i c l e i n f o
Article history:
Received 22 May 2014
Received in revised form
7 August 2014
Accepted 8 August 2014
Available online 19 August 2014
Keywords:
Nickel
Impurities
Nickel perchlorate
Nickel alloys
a b s t r a c t
A method has been developed for the determination of traces of arsenic, boron, bismuth, gallium,
germanium, phosphorus, lead, antimony, selenium, silicon and tellurium in nickel matrix. The sample
was dissolved in HClO
4
(150 1C) and nickel was settled as crystalline nickelperchlorate [Ni(ClO
4
)
2
] on
cooling. The mixture was ultrasonicated and after the separation of Ni(ClO
4
)
2
, analytes of interest were
determined in the supernatant using ICP-OES. Similarly, it was also found that, after the dissolution of
nickel in perchloric acid, when the solution temperature was maintained at 100 1C, long needle like
crystals of nickel perchlorate were formed. The crystals were separated from the mixture and trace
elements in the supernatant were determined using ICP-OES. In both methods the matrix removal was
499% and the recoveries of analytes were in the range 9297%. The limits of detection for As, B, Bi, Ga,
Ge, P, Pb, Sb, Se, Si and Te were found to be 0.18, 0.21, 0.07, 0.06, 0.25, 0.11, 0.09, 0.10, 0.17, 0.20 and
0.07 g g
1
respectively. The procedure was applied for the analysis of a standard reference material
nickel oxide (SRM 761, Nickel Oxide No.1, NBS, USA) and the values obtained are in close agreement with
the certied values.
& 2014 Elsevier B.V. All rights reserved.
1. Introduction
Nickel base super alloys, by virtue of their excellent high tempera-
ture properties coupled with corrosion resistance, have been widely
used in jet engine, gas turbine engine components where high
strength at high temperature is needed [1,2]. The quality of nickel-
base alloys is highly dependent on the amounts of trace elements
present as the presence of trace elements has serious effects on the
mechanical and physical properties of these high temperature alloys
[3,4]. Among the impurities in nickel base alloys, lowmelting elements
such as As, Se, Te, Pb etc. are detrimental to the strength of high
temperature alloys even when present at very low concentrations [5].
Therefore it is essential to determine traces of these low melting
elements (As, Se, Te and Pb) in nickel matrix.
Hydride generation atomic absorption spectrometry (HG-FAAS)
and Hydride generation Inductively Coupled Plasma Optical Emis-
sion Spectrometry (HG-ICP-OES) are among the most sensitive
techniques for the determination of traces of As, Se and Te [6,7].
However, due to the strong interference of nickel in hydride
generation process, it is difcult to analyze nickel matrices for
low melting elements mentioned above by hydride generation
[8,9]. In these techniques, the interference caused by nickel is
minimized by masking the interferents [6,10] or by rapid hydride
generation followed by fast separation of hydride from the liquid
phase [11]. Inductively Coupled Plasma Mass Spectrometry (ICP-
MS) is a powerful technique for the trace multi-element analysis.
The determination of traces of Te in alloys by ICP-MS suffers from
severe non-spectroscopic interference when conventional pneu-
matic nebulizer was used for sample introduction. This problem
was overcome by applying the vapor generation of Te and its
introduction to ICP-MS [12]. Graphite Furnace Atomic Absorption
Spectrometry (GFAAS) methods were developed for the determi-
nation of selenium, tellurium in nickel after the matrix separation
by reductive co-precipitation with palladium using ascorbic acid
[13,14]. As per our knowledge, no single method has been reported
to date for trace level determination of these impurities (As, Sb, Bi,
B, Ga, Ge, Pb, P, Se, Si and Te) in nickel matrix.
Herein, a new method has been described for the determina-
tion of traces of As, Sb, Bi, B, Ga, Ge, Pb, P, Se, Si and Te in nickel
matrix. The sample was dissolved in con.HClO
4
at hot condition
and after cooling, the matrix (nickel) was separated from the
sample solution either as crystalline precipitate or needle like
crystals of Ni(ClO
4
)
2
. The crystalline precipitate was ultrasonicated
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/talanta
Talanta
http://dx.doi.org/10.1016/j.talanta.2014.08.026
0039-9140/& 2014 Elsevier B.V. All rights reserved.
n
Corresponding author. Tel.: 91 40 27121365.
E-mail address: sahayamac@gmail.com (A.C. Sahayam).
Talanta 131 (2015) 505509
to leach out the impurities from crystals and the concentration of
trace impurities in the supernatant was measured using ICP-OES.
In the second procedure, needle like crystalline Ni(ClO
4
)
2
was
prepared by maintaining the sample solution at 100 1C and the
supernatant was analyzed for trace impurities by ICP-OES.
2. Experimental
2.1. Apparatus
A Horiba scientic model of ULTIMA-2 Inductively Coupled
Plasma-Optical Emission Spectrometer (ICP-OES) was used. The
instrumental specications are given in Table 1. A Bandelin model
of SONOREX DIGITEC 35 kHz ultrasonic bath with heater (20
80 1C) was used.
2.2. Reagent and standard solutions
Suprapur grade HClO
4
(70%,w/v) from Merck, Germany was
used. Standard stock solutions of all the elements (1000 mg L
1
)
were obtained from Merck, Germany. Working standard solutions
were freshly prepared from these stock solutions as required.
2.3. Procedure
2.3.1. Matrix separation after precipitation of Ni(ClO
4
)
2
followed
by ultrasonication
Around 0.5 g of nickel sample was taken in a at bottom quartz
test tube of 25 mL capacity, 10 mL suprapur grade HClO
4
(70%,w/v)
was added, the tube was covered with watch glass and heated
over a hot plate at 150 1C (solution temperature) for 30 min.
After the complete dissolution of the sample, the sample solution
was allowed to cool to 60 1C. Then the sample solution was
sonicated at 50 1C (35 kHz) for 40 min. After ultrasonication, the
sample solution was taken out from the ultrasonic bath and cooled
to room temperature and was centrifuged between 1000 and
1500 rpm and the supernatant was separated using a micropip-
ette. Then the supernatant (3 mL) was diluted to 25 mL and
impurities present were measured using ICP-OES. If the concen-
tration of impurities is not detectable, pre-concentration of impu-
rities can be carried out by evaporating the supernatant on a hot
plate (surface temperature of 130 1C) in a fume hood till no more
dense fumes came out from the sample solution. Then after
cooling it was diluted to 10 mL using mill-Q water and the analytes
were measured using ICP-OES.
2.3.2. Matrix separation after the formation of needle like crystals
of Ni(ClO
4
)
2
Around 0.5 g of nickel sample was dissolved in perchloric acid as
described above. After complete dissolution of sample, the test tube
was removed from the hot plate and temperature of the solution
was maintained at 100 1C using Infra Red (IR) lamp (IR lamp was
kept by the side with respect to sample solution). After the
appearance of needle like crystals, the sample solution was kept
at this temperature for around 30 min. Then the IR lamp was
switched off and allowed the solution to cool to room temperature
(25 1C). The sample solution was carefully transferred into a clean
25 mL quartz beaker using a micropipette. Then the separated
sample solution (3 mL) was diluted to 25 mL and measured for
the impurities using ICP-OES. To reduce the dilution factor, the
separated sample solution could be diluted to 10 mL after the
evaporation of excess HClO
4
as described in the above procedure.
3. Result and discussion
3.1. Role of HClO
4
on dissolution and separation of nickel
Nickel/nickel oxide reacts with con. HClO
4
(70%,w/v) at
150 1C (solution temperature) forming intense green color Ni
(ClO
4
)
2
solution. When this solution is allowed to cool 499% of
nickel is settled as crystalline form of Ni(ClO
4
)
2
. This unique
property of Ni(ClO
4
)
2
is utilized to separate the matrix for its trace
element analysis. Due to high solubility of Ni(ClO
4
)
2
in water, after
the crystallization, any dilution of HClO
4
with water leads to
redissolution of the matrix and decrease in matrix separation.
Therefore, sample dissolution and matrix separation were carried
out using concentrated HClO
4
. The effective matrix separation was
not affected when the concentration of HClO
4
is maintained 495%
(v/v). Any increased dilution of HClO
4
drastically increases the
dissolution of Ni(ClO
4
)
2
and leads to complete dissolution when
the concentration of HClO
4
is o50% (v/v). However, the matrix
separation (99%) was independent of the amount of HClO
4
(70%,
w/v) used for the sample dissolution.
3.2. Ultrasonic extraction of analytes
Preliminary experiments were carried out for the separation of
Ni as Ni(ClO
4
)
2
precipitate and the recovery of analytes in the
supernatant was found to be only 2030%. Lower recovery of
analytes could be due to the occlusion of the analytes during the
formation of bulky molecule of Ni(ClO
4
)
2
(5 times the weight of
nickel) precipitate. Therefore, after the precipitation of crystalline
form of Ni(ClO
4
)
2
, ultrasonication was carried out to leach out the
analytes from Ni(ClO
4
)
2
precipitate. Even after the ultrasonication,
the recovery of analytes was in the range of 80%. Increase in the
temperature of the sample solution and ultrasonication bath lead
to further improvement in recovery of analytes. However, it also
leads to increase in the dissolution of Ni(ClO
4
)
2
which resulted in
poor matrix separation. Therefore these two parameters are
required to be optimized. After the separation of nickel, a high
concentration of HClO
4
remains in the sample solution requiring
large dilution prior to the determination of impurities using ICP-
OES which lead to poor limit of detection (LOD). Hence, to obviate
large dilution, excess HClO
4
from sample solution needed to be
evaporated. Due to hazardous nature of HClO4, it is very essential
to optimize its amount to as minimum as possible.
3.3. Optimization of ultrasonic extraction of analytes
Based on the recovery of arsenic, ultrasonic extraction was
optimized for each parameter, i.e., amount of HClO
4
, temperature of
the sample solution/ sonication bath and time duration of ultrasonica-
tion. To study the effect of amount of HClO
4
on matrix removal and
analyte recovery, after the dissolution of sample, ultrasonication was
carried out at a solution temperature of 60 1C and ultrasonic bath
temperature of 50 1C for 30 min. The amount of HClO
4
was
increased from 5 mL onwards (minimum 5 mL of HClO
4
is required
for the dissolution of 0.5 g of nickel) in steps of 2 mL and when it was
Table 1
Operating conditions of ICP-OES.
Power 1.1 kW
Wavelengths (nm) Se (196.090), As (197.198), Sb (206.833),
Ge (209.426), P (213.618), Te (214.281),
Pb (220.353), Bi (223.061), B (249.772),
Si (251.611) and Ga (294.364)
Plasma gas ow rate 12.1 l min
1
Nebulizer gas ow rate 0.83 l min
1
, concentrique nebulizer
Auxiliary gas ow rate 0.52 l min
1
Solution uptake rate 2.0 mL min
1
S. Thangavel et al. / Talanta 131 (2015) 505509 506
10 mL or higher the recovery of analyte (As) was more than 92%
(Fig. 1). From this gure, it can be seen that no signicant increase in
the separation of matrix was observed by further increasing the
amount of acid. Hence, 10 mL of acid was used for further studies.
The effect of temperature of the sample solution and the temperature
of the ultrasonication bath on matrix separation and recovery of
analytes are given in Figs. 2 and 3 respectively. For these studies, after
the dissolution (0.5 g of sample in 10 mL of HClO
4
) the sample solution
was cooled to various temperatures (10025 1C) and kept inside the
ultrasonication bath maintained at a temperature between 25 and
80 1C. When the temperature of ultrasonic bath and sample solution
were between 25 and 60 1C matrix separation was 99%. However,
when these temperatures were between 50 and 60 1C, the recovery of
analytes was found to be more than 93%. Therefore optimized
temperature of between 50 and 60 1C was employed in subsequent
studies. At optimized temperatures, the required time duration of
ultrasonic bath was also studied and the results are shown in Fig. 4. It
can be seen from this gure that optimized time duration was
40 min to achieve near quantitative recovery of all the analytes.
3.4. Preparation of needle like crystals of Ni(ClO
4
)
2
In our preliminary experiments after the dissolution of nickel,
when the sample solution was allowed to cool, during transition of
temperature formation of needle like crystals was observed
occasionally. When these crystals were formed, the recovery of
analytes was in the range of 9296%, thus obviating the need for
ultrasonic extraction. However, it was very difcult to achieve
reproducible preparation of long needle like crystals. From
repeated experiments, it has been observed that the following
parameters namely heating apparatus, the crystallization tempera-
ture and dimension of the container (where crystallization is
carried out) were crucial for the formation of long needle like
crystals. As in the case of ultrasonic extraction, even in this matrix
separation method, no signicant variation in the separation of
matrix (499%) was observed with the amount of HClO
4
used for
the dissolution sample.
3.5. Optimization of crystallization
Preliminary experiments were carried out for the formation of
needle like crystals by maintaining the temperature of the sample
solution using a hot plate. On hot plate it is very difcult to
maintain the temperature of solution and hence it was found to be
very intricate to produce needle like crystals. After the dissolution
of sample at 150 1C, when the temperature of the hot plate was
decreased nickel perchlorate was precipitated in crystalline form
and is completely dissolved when the temperature was increased.
Therefore, crystallization was attempted by maintaining the tem-
perature of the sample solution using IR lamp. In this case also, it
was stringent to develop needle like crystals when IR lamp was
kept above or below the sample solution. When the IR lamp was
kept in horizontal position (by the side) with respect to sample
solution, long needle like crystals were formed that were not
settled down as precipitate (Fig. 5). The distance between IR lamp
and test tube was also optimized. In our laboratory atmosphere
the optimized distance was found to be 10 cm. When the
Fig. 1. Effect of amount of con.HClO
4
on 0.5 g of sample on matrix (A)/analyte
(B) separation (No. of replicates (n)3).
Fig. 2. Effect of temperature of ultrasonication bath/sample solution on separation
of matrix (n3).
Fig. 3. Effect of temperature of ultrasonication bath/sample solution on recovery of
analyte (n3).
Fig. 4. Effect of time duration of ultrasonic bath (50 1C) on matrix/analyte
separation. A-matrix separation; B-analyte recovery (n3).
S. Thangavel et al. / Talanta 131 (2015) 505509 507
temperature of the sample solution was 4120 1C, Ni(ClO
4
)
2
remained in the solution form. As soon as the temperature was
decreased below 115 1C, needle like crystals started to form and
did not settle down as precipitate till the temperature of the
sample solution remained above 90 1C. As described earlier,
settling down of crystals as precipitate leads to poor recovery of
analytes. Therefore, to keep the Ni(ClO
4
)
2
as needle like crystals,
optimized temperature of the solution was found to be
100710 1C. If this temperature was maintained for 30 min,
499% of 0.5 g of nickel was formed as long needle crystals of Ni
(ClO
4
)
2
. The IR lamp was switched off to bring the solution to room
temperature. After reaching room temperature, the sample solu-
tion needs to be separated from the crystals within one hour.
Otherwise, the needle like crystals starts to settle as precipitate,
which leads to lower recovery of analytes.
It was observed that the dimensions of container also play an
important role in the formation of needle like crystals. Since the
volume of HClO
4
optimized is 10 mL for the dissolution of 0.5 g of
sample, diameter of the tube should be as low as possible.
Therefore at bottom quartz test tubes with different diameters
were taken for crystal formation. If the vessel is very narrow the
crystals started to form from the wall of the vessel in horizontal
direction and hence complete crystal growth could not taken
place. The crystals formed on the wall of the tube fell down and
settled as precipitate. The optimized diameter of the tube was
found to be between 2 and 3 cm.
3.6. Recovery of other elements
In both matrix separation methods (crystallization and ultra-
sonication), since most of the transition elements such as Fe, Al,
Cu, Mn etc. form perchlorates, these elements will co-precipitate
with Ni(ClO
4
)
2
resulting in poor recovery (o30%). However, after
the matrix separation in the form of needle like crystals, poorer
recovery of B, Ga, Ge, Pb and Si was observed (Table 2). It could be
due to the adsorption of perchlorates of B, Ga, Ge and Pb on Ni
(ClO
4
)
2
resulting in poorer recovery for these analytes. Silicon is
prone to hydrolysis on treatment with concentrated acids (except
HF). In this method since crystallization was carried out in
concentrated HClO
4
medium, silicon might be hydrolyzed and
Fig. 5. Ni(ClO
4
)
2
in solution form (A) and needle like crystalline form (B).
Table 2
Recovery study and limit of detection.
Element After evaporation of HClO
4
Without evaporation of HClO
4
Limit of Detection (g g
1
)
After crystallization After ultrasonication After crystallization After ultrasonication
As 95 94 94 96 0.18
B 46 28 42 93 0.21
Bi 94 96 97 95 0.07
Ga 45 95 42 99 0.06
Ge 45 22 46 92 0.25
P 95 97 96 99 0.11
Pb 42 94 48 97 0.09
Se 94 95 93 94 0.17
Sb 38 36 95 96 0.10
Si 32 40 32 96 0.20
Te 94 94 97 96 0.07
S. Thangavel et al. / Talanta 131 (2015) 505509 508
adsorbed on Ni(ClO
4
)
2
. However, matrix separation followed by
ultrasonication, improved recoveries of these analytes to 492%. It
could be due to the fact that ultrasonication was carried out at
50 1C and at this condition these analytes are quantitatively
leached out from Ni(ClO
4
)
2
precipitate. After the separation of
matrix, when excess HClO
4
is evaporated on hot plate, poor
recovery was found out for B, Ge and Sb. It could be due to the
loss of these analytes on evaporation of HClO
4
.
3.7. Sample analysis, limit of detection (LOD) and recovery
To validate the developed method, this procedure was applied
to a standard reference material of Nickel oxide (Standard Sample
671, Nickel Oxide, No.1, National Bureau of Standards, Washington)
and the results are in close agreement with certied values
(Table 3). The limit of detection (LOD), given as three times the
variation on repeated measurements of blank (n6) for all the
analytes in solid sample is given in Table 2 and it was calculated
based on the dilution factor of 0.5 g of sample made upto 10 mL
whereas for B, Ge and Sb it was calculated based on the dilution
factor of 0.5 g made upto 25 mL as loss of these analytes was
observed on evaporation of HClO
4
.
4. Conclusion
This new sample digestion and matrix separation approach is
simple and all the critical elements in nickel matrix could be
determined. HClO
4
is the only reagent used for the dissolution of
sample as well as for the separation of matrix and is available in
very pure grade (suprapur grade), which results in very less
process blank. The proposed method can also be used for crystal
growth of Ni(ClO
4
)
2
for the preparation of high-pure Ni(ClO
4
)
2
.
References
[1] D.R. Wood, R.M. Cook, Metallurgia 67 (1963) 109117.
[2] B. Geddes, H. Leon, X. Huang, Super Alloys: Alloying and Performance, ASM
International, Ohio (2010) 15.
[3] D.G. Andrews, J.B. Headridge, Analyst 102 (1977) 436445.
[4] S. Backman, R.W. Karlsson, Analyst 104 (1979) 10171029.
[5] R.J. Henricks, M.L. Gel, Trace Element Effects in Cast Nickel-Base Superalloys,
III, ASM Materials Engineering Congress, Chicago, 1973 (October 3).
[6] I. Wickstrm, W. Lund, R. Bye, Analyst 120 (1995) 26952698.
[7] H. Wiltsche, I.B. Brenner, K. Prattes, G. Knapp, J. Anal. At. Spectrom. 23 (2008)
12531262.
[8] C. Rigby, I.D. Brindle, J. Anal. At. Spectrom. 14 (1999) 253258.
[9] B. Weltz, M. Melcher, Spectrochim. Acta B 36 (1981) 439462.
[10] I. Wickstrm, W. Lund, R. Bye, J. Anal. At. Spectrom. 10 (1995) 803808.
[11] W.W. Ding, R.E. Sturgeon, Anal. Chem. 69 (1997) 527531.
[12] Y.L. Chen, S.J. Jiang, J. Anal. At. Spectrom. 15 (2000) 15781582.
[13] T. Ashino, K. Takada, Anal. Chim. Acta 312 (1995) 157163.
[14] Hui-Ming Liu, Shi-Yang Chen, Pin-Hsuan Chang, Suh-Jen Jane Tsai, Anal. Chim.
Acta 459 (2002) 161168.
Table 3
Analysis of nickel oxide (SRM 761, Nickel Oxide No.1, NBS, USA).
Element SRM value (g g
1
) Present method value (g g
1
7 ) n4
Pb 1672 1572
Se 2.070.3 1.970.4
As 59
a
, 45
b
5674
Te o0.2 o0.2
a
Average of atomic absorption methods.
b
Average of extraction-photometric and distillation-photometric procedures.
S. Thangavel et al. / Talanta 131 (2015) 505509 509

You might also like