Analytica Chimica Acta 547 (2005) 14–17

Determination of copper, cadmium and lead in seawater and mineral water

by flame atomic absorption spectrometry after coprecipitation with
aluminum hydroxide
G. Doner ∗ , A. Ege
Istanbul Technical University, Faculty of Science and Letters, Chemistry Department, Maslak, 34469 Istanbul, Turkey

Received 9 November 2004; received in revised form 25 February 2005; accepted 28 February 2005
Available online 5 April 2005

Abstract

An aluminum hydroxide coprecipitation method for the determination of cadmium, copper and lead by flame atomic absorption spectrometry
in aqueous solutions, seawater and mineral water samples has been investigated. The coprecipitation conditions, such as the effect of the pH,
the amount of carrier element, the effect of possible matrix ions and the time were examined in detail for the studied elements. It was found
that cadmium, copper and lead are co-precipitated quantitatively (≥95%) with aluminum hydroxide at pH 7 with low R.S.D. values of around
2 to 3%. Detection limits (38) were 6 ng ml−1 for Cd, 3 ng ml−1 for Cu and 16 ng ml−1 for Pb. The method proposed was validated by the
analysis of HPS 312205 seawater standard reference material and spiked mineral water samples.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Coprecipitation; Aluminum hydroxide; Seawater; Mineral water; Cu; Cd; Pb; FAAS

1. Introduction ments. In addition, concentration factors by coprecipitation

The direct determination of the trace metals in sea water
and mineral waters by atomic absorption spectrometry (AAS)
is very difficult due to the low levels of metal ions and also
interferic influences of main components of the water sam-
ples. In many instances, it has been necessary to use a precon-
centration method prior to analysis. Several preconcentration
techniques are used for this purpose including solvent extrac-
tion, ion exchange, chelating resins, solid-phase extraction,
adsorption and coprecipitation. Among these techniques, co-
precipitation method is particularly proposed to preconcen-
trate trace elements from matrices that adversely influence
atomic absorption spectrometric detection. The combination
of coprecipitation with filtration and/or centrifugation offers
a simple and rapid preconcentration technique especially for
the recoveries of trace elements and determination of the suf-
ficient separation factors for alkali and alkaline earth ele-
∗ Corresponding author. Tel.: +90 212 285 6997.
E-mail address: doner@itu.edu.tr (G. Doner).
can reach 5000–40,000 by this technique. These concentra-
tion factors can only reach 10–500 when the other techniques
are used and they are time-consuming. Therefore, by using
a coprecipitation technique, the analytical sensitivity can be
considerably increased. The main requirement for this tech-
nique is that the collector should easily separate from the
matrix solution; this is done by filtering and/or centrifug-
ing and washing the precipitate. In addition, it is desirable
that the collector should be a pure and readily available sub-
stance. A number of metal hydroxides such as iron(III), in-
dium, gallium, cerium(IV), lanthanum, magnesium, hafnium,
scandium and zirconium have been widely used and success-
fully applied for the preconcentration of trace metal ions from
different media [1–9]. The method using hafnium hydroxide
is expensive. Although magnesium hydroxide is an excellent
carrier, magnesium, itself, causes serious high background
absorption in graphite furnace atomic absorption spectrome-
try (GFAAS). The method using iron(III) hydroxide cannot
avoid the coprecipitation of large amount of alkaline earth
metals, which may interfere with the cadmium determina-

0003-2670/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2005.02.073
 G. Doner, A. Ege / Analytica Chimica Acta 547 (2005) 14–17
tion. Although indium hydroxide is an excellent collector, in-
dium itself, causes serious background absorption. The back-
ground absorption can be eliminated by using the minimal
amount of indium and the volatilization of indium as bromide
during the pyrolysis stage can be proceeded in the presence
of thiourea. In addition, Hiraide et al. [10] used aluminum
as a carrier element to collect traces of copper(II) and cad-
mium(II) ions from an aqueous solution. However, aluminum
was not found effective carrier for the coprecipitation for cop-
per(II) and cadmium(II) due to matrix interferences and the
high background absorption and also loss of analyte at pyrol-
sis step in graphite furnace atomic absorption spectrometry.
The supernatant solution was analyzed by GFAAS to reduce
the background effect. Therefore, this procedure is not ac-
ceptable for the analysis of trace elements as a preconcentra-
tion technique. Indium hydroxide was also suggested by these
authors as a better collector for copper(II) and cadmium(II)
in aqueous solution. However, the suggested coprecipitation
method was not applied to any real sample.
To the best of our knowledge, there is no significant study
on the use of aluminum hydroxide for coprecipitation of trace
elements in seawater and mineral water samples.
Flame atomic absorption spectrometry (FAAS) is a much
more convenient technique and it has more advantages over
the other AAS method, graphite furnace atomic absorption
spectrometry. The matrices are not as effective in FAAS as
in GFAAS and also FAAS provides a larger working range
than GFAAS does. These are the most important advantages
of FAAS.
In this study, taking the advantages of FAAS and the copre-
cipitation technique, an aluminum hydroxide, which is read-
ily available and inexpensive, have been used as an carrier
for the separation/preconcentration of cadmium, copper and
lead in aqueous solutions, seawater and mineral water sam-
ples. The influence of the various analytical parameters on
the recovery for the examined elements, such as pH, amount
of carrier element and Na2 CO3 , matrix ions, the settling time
etc. have been discussed.
2. Materials and method
2.1. Apparatus and reagents
An Analytic Jena Vario 6 atomic absorption spectrometer
equipped with an air–acetylene burner was used for the deter-
mination of copper, cadmium and lead. Copper, cadmium and
lead hollow cathode lamps were used as the spectral radiation
sources. The wavelengths were 324.8, 228.8 and 217.0 nm
for Cu, Cd and Pb, respectively. The pH adjustment was con-
trolled using WTW pH 340-A/SET 2 pH meter.
All chemicals were of analytical-reagent grade (Merck).
Distilled-deionized water was used throughout. Stock solu-
tions (1000 mg 1−1 ) of Cu, Cd and Pb were prepared from
titrisol standards (Merck) and working standard solutions
were prepared fresh daily by appropriately diluting the stock
15
solutions. HPS 312205 seawater standard reference material
was used.
Seawater samples collected from Istanbul Bosphorous
were immediately filtered through cellulose nitrate (0.45 ␮m
pore size, 47 mm diameter – Whatman) and acidified to pH 3
with nitric acid and stored in precleaned polyethylene bottles.
Mineral water samples were purchased from a local market.
2.2. Coprecipitation procedure
The coprecipitation method was optimized with aqueous
solutions prior to its application to the real samples. First,
650 ␮l of 1 M Na2 CO3 solution and then 300 ␮l of 1 M
Al(NO3 )3 ·9H2 O solution (8.1 mg Al) were added to 50 ml
of a solution containing 5 ␮g of Cu and Cd and 10 ␮g of Pb
in a centrifuge tube. Sodium carbonate amount was found
enough to adjust the pH 7. The solution was centrifuged at
3500 rpm for 20 min. The supernatant was decantated and the
precipitate was dissolved with 500 ␮l of 4 M HNO3 . Cu, Cd
and Pb were determined in 2.5 ml of the final solutions by
FAAS.
3. Results and discussion
3.1. Effect of different bases on precipitation
The coagulation step is very important factor to get a
very fast coprecipitation and the easily separation of the co-
precipitated part from the supernatant with a simple centrifu-
gation procedure. Therefore, our initial method development
studies were performed with different bases such as NH3 ,
NaOH and Na2 CO3 for aluminum hydroxide precipitation.
The experiments were shown that coagulation of aluminum
hydroxide was much dependent on the type of the chemi-
cal. When sodium carbonate was used for this purpose, the
precipitated aluminum hydroxide was easily coagulated and
the physical nature of the precipitate was much convenient to
work on the precipitate. In addition, to control the pH gradi-
ent using sodium carbonate (rather than NaOH and NH3 ) was
found easy. Particularly, working with ammonia (NH3 ) so-
lution, the precipitate was gelatinous and suspended; hence,
separation of co-precipitate from supernatant by centrifuga-
tion was difficult and it takes longer time. In addition, recov-
eries for the studied elements were found very low. This may
be due to the lower surface area of the precipitate. It is known
that the higher surface area provides higher contact surface
with analytes during centrifugation step to collect the trace
elements from different media. As a result, Na2 CO3 was cho-
sen as a hydroxide ion source because of the above-mentioned
reasons.
3.2. Study of the optimum conditions for coprecipitation
The influence of the pH on the recovery of trace level of
Cu(II), Cd(II) and Pb(II) were examined. Aluminum hydrox-
16 G. Doner, A. Ege / Analytica Chimica Acta 547 (2005) 14–17

Table 1
Effect of aluminum amount on coprecipitation (amount of analytes: 20 ␮g
of Pb, 10 ␮g of Cu and Cd, n = 5)
Amount of Al (mg) Recovery (%)

Pb Cu Cd
4.0 90 ± 3 92 ± 3 87 ± 2
8.1 99 ± 2 98 ± 2 99 ± 3
16.2 100 ± 2 97 ± 3 100 ± 2

increased. The reason for that co-precipitate part has own

volume that is directly related with the amount of aluminum
Fig. 1. Effect of pH on the coprecipitation: (
) Cu, () Cd and (䊉) Pb.
hydroxide used. The results show that final volume is more
Amount of analytes: 20 ␮g of Pb, 10 ␮g of Cu and Cd, n = 5.
important to get higher concentration factor. To remain the
final volume constant to obtain higher concentration factor,
ide could be obtained between 4.2 and 9.0 pH range. Because higher concentration of HNO3 with small volume was used.
of the dissolution of the precipitate as an Al(OH)4 − in alka- The results show that co-precipitated part can easily digest
line solutions (above pH 9), working pH range was narrow. with 4 M HNO3 . Maximum concentration factor was found
The pH was adjusted to 4–9 by changing the amount of added to be 125 for real samples with 4 M HNO3 .
Na2 CO3 solution while the amount of aluminum in solution
was constant. As can be seen from Fig. 1, the recovery of all
3.3. Effect of the order of addition of reagents on
elements increased with an increase in pH and quantitative
precipitation
results were obtained for all studied elements at pH 7. For
further experiments, the pH 7 was chosen as an optimum pH.
The order of addition of reagents for each pH values stud-
Contact time, which is important for the quality of copre-
ied was found to be an important parameter for higher set-
cipitation, was optimized working with 5–30 min time pe-
tling rate and amount of precipitate. Experiments were carried
riod of the centrifugation. Centrifugation process not only
out between pH 4.0–9.0. Rapid settling was achieved when
provides the separation of the co-precipitated part from the
Al(III) solutions were added to Na2 CO3 solution near neutral
supernatant but also reduces the contact time between the ana-
pH which is close to the minimum solubility pH of Al(OH)3 .
lyte and co-precipitant aluminum hydroxide for the quantita-
For this pH value, higher amount of solid phase was obtained
tive recoveries. Another series of coprecipitation experiments
by the reverse order of reagent addition. These results were
were performed without applying any centrifugation. The co-
in good agreement with those reported in literature as well
precipitate part has been left for the settling by gravitation.
[11]. It is obvious that for the successful precipitation, alu-
Unfortunately, to get the clear part from the co-precipitate
minum(III) should be added to Na2 CO3 solution.
takes long time with lower recoveries. In this case, even 2
days settling time which is not practical and also open for the
3.4. Effect of matrix ions
contamination was not enough for the separation of solid part
by decantation from supernatant. Consequently, the required
To detect potential interferences on the coprecipitation of
centrifugation time for the successful coprecipitation with
studied elements various amount of Cl− (NaCl) and SO4 2−
quantitative recoveries of the analytes was determined to be
(Na2 SO4 ) ions, which are major components of mineral wa-
20 min. The necessary amount of aluminum for coprecipita-
tion was selected experimentally. The amount of aluminum
Table 2
hydroxide is very important for two reasons. First, enough Effect of matrix ions on coprecipitation (amount of analytes: 10 ␮g of Pb,
aluminum should be added to the sample solution to ensure 5 ␮g of Cu and Cd, n = 5)
that the precipitation is rapid, and that sufficient precipitate Ions Added as Concentration Recovery (%)
is formed for easy filtration or centrifugation. Second, at the (mg l−1 )
Pb Cu Cd
same time the amount of the carrier should be sufficiently
small for adsorption of interfering ions to be negligible. The Cl− NaCl 0 99 ± 3 97 ± 4 93 ± 6
3000 96 ± 2 97 ± 4 90 ± 6
latter can affect the preconcentration factor that is very im-
5000 95 ± 4 95 ± 5 91 ± 5
portant for the detection of analyte. To optimize these fac- 10,000 102 ± 5 94 ± 3 90 ± 7
tors, while analyte concentration was constant, coprecipita- 20,000 101 ± 3 90 ± 2 87 ± 5
tion was performed using different amount of aluminum. As 35,000 98 ± 4 105 ± 4 88 ± 6
it can be seen from Table 1, the recoveries for all studied SO4 2− Na2 SO4 0 99 ± 5 97 ± 2 93 ± 5
elements were quantitative after 8.1 mg of Al. 100 99 ± 4 102 ± 3 88 ± 4
When amount of aluminum hydroxide is increased, to di- 200 102 ± 4 99 ± 5 87 ± 5
gest the co-precipitate, necessary amount of nitric acid was 500 98 ± 5 105 ± 4 95 ± 7
1000 99 ± 6 106 ± 5 105 ± 7
increased. As a result, the final volume of the sample was
 G. Doner, A. Ege / Analytica Chimica Acta 547 (2005) 14–17 17

Table 3
Recovery of analyte spikes from seawater and mineral water (n = 5)
Sample Pb Cu Cd

Added (␮g) Founda (␮g) Recovery (%) Added (␮g) Founda (␮g) Recovery (%) Added (␮g) Founda (␮g) Recovery (%)
Seawaterb 0 n.d. – 0 n.d. – 0 n.d. –
6 5.8 ± 0.1 97 2 1.9 ± 0.1 95 2 1.8 ± 0.1 90
8 7.6 ± 0.1 95 4 3.9 ± 0.1 98 4 3.5 ± 0.1 88
10 9.5 ± 0.2 95 6 5.8 ± 0.1 97 6 5.2 ± 0.1 87
Mineral 0 n.d. – 0 n.d. – 0 n.d. –
Water 6 5.8 ± 0.1 97 2 1.9 ± 0.1 96 2 1.8 ± 0.1 90
8 7.7 ± 0.2 96 4 3.9 ± 0.1 97 4 3.6 ± 0.1 90
10 9.5 ± 0.3 95 6 5.8 ± 0.2 97 6 5.3 ± 0.2 88
Seawaterc 0 n.d – 0 2.4 ± 0.1 96 0 n.d –
5 4.8 ± 0.1 96 2.5 4.8 ± 0.1 96 2.5 2.2 ± 0.1 88
8 7.6 ± 0.2 95 4 6.2 ± 0.1 95 4 3.6 ± 0.1 90
Sample volume: 250 ml for sea water, 200 ml for mineral water.
a x ± ts/√N, uncertanity at 95% confidence limit.
b Seawater of Istanbul Bosphorous, n.d. not detected.
c Standard reference seawater (HPS 312205) (certified values: 0.004 mg kg−1 Pb, 0.01 mg kg−1 Cu, 0.0001 mg kg−1 Cd).

ter and seawater, were added to the solution containing trace
amount of analytes and the optimized procedure was ap-
plied. The results are given in Table 2. As can be seen from
Table 2, large amounts of chloride (<10,000 mg l−1 ) and
sulfate (<500 mg l−1 ) did not interfere with the determina-
tion of examined elements. Although in the presence of Cl−
(35,000 mg l−1 ), the tolerable level is high for the recoveries
of Pb and Cu. The recoveries of Cd were slightly reduced be-
low 90%. On the other hand, in the presence of SO4 2− ions
up to 1000 mg l−1 that is relatively high amount for mineral
waters, determination of cadmium was not affected. On their
labels of commercially available mineral waters, those ions
amounts are always given. The amount of Cl− and SO4 2−
ions depends on the source of mineral water. Although, in
some cases Cl− and/or SO4 2− are not present in mineral wa-
ters, typical concentrations for SO4 2− in mineral water are
between 29 and 86 mg l−1 . On the other hand, chloride level
(20–35 g kg−1 ) is relatively high in seawater.
the reagent blank, for FAAS measurements, were found to
be 6, 3 and 16 ng ml−1 for Cd, Cu and Pb, respectively.
4. Conclusion
Coprecipitation with aluminum hydroxide has been suc-
cessfully applied to the determination of trace amounts of
Cu, Cd and Pb in seawater and mineral water samples with
acceptable accuracy and precision. The proposed method is
inexpensive and fast. Using the optimized method, it is pos-
sible to increase the concentration factor as 250 by using less
volume of concentric HNO3 but in that case to run the small
volume of sample by FAAS, flow injection sample introduc-
tion system should be used.
References

3.5. Method validation and analytical performance
The utility of the present method was evaluated by exam-
ining the recoveries of studied elements from Bosphorus sea-
water, mineral water and standard reference seawater (issued
by HPS 312205) samples spiked with the studied analytes.
Five replicate analyses were performed for each sample. The
recoveries were above 95% for Pb and Cu and 87% for Cd.
The results are given in Table 3, indicating that the proposed
method is applicable to analyze the water samples contain-
ing trace amount of Cu, Cd and Pb after 125-fold precon-
centration. The detection limits, defined as the concentration
equivalent to three times the standard deviation (n = 10) of
[1] T. Duan, J. Kang, H. Chen, X. Zeng, Spectrochim. Acta Part B 58
(2003) 1679.
[2] O. Kujirai, K. Yamada, Fresenius J. Anal. Chem. 354 (1996) 428.
[3] K. Takeda, C. Akamatsu, Y. Ishikawa, Anal. Chim. Acta 298 (1994)
375.
[4] J. Nakajima, Y. Hirano, K. Oguma, Anal. Sci. 19 (2003) 585.
[5] U. Divrikli, L. Elci, Anal. Chim. Acta 452 (2002) 231.
[6] S. Nielsen, J.J. Sloth, E.H. Hansen, Analyst 121 (1996) 31.
[7] L. Elci, S. Saracoglu, Talanta 46 (1998) 1305.
[8] S. Saracoglu, M. Soylak, L. Elci, Talanta 59 (2003) 287.
[9] J. Ueda, N. Yamazaki, Bull. Chem. Soc. Jpn. 59 (1986) 1845.
[10] M. Hiraide, Z.S. Chen, H. Kawaguchi, Microchim. Acta 127 (1997)
119.
[11] B. Demirata, R. Apak, G. Gumus, H. Afsar, Water, Air, Soil Pollut.
133 (2002) 265.