Talanta 51 (2000) 889 894

www.elsevier.com/locate/talanta

Flame atomic absorption spectrometric determination of

silver after preconcentration on Amberlite XAD-16 resin
from thiocyanate solution
Adalet Tunceli, A. Rehber Turker *
Gazi U8 ni6ersitesi, Fen-Edebiyat Fakultesi, TR-06500 Ankara, Turkey

Received 8 July 1999; received in revised form 23 November 1999; accepted 25 November 1999

Abstract

A method of silver preconcentration by using a column containing Amberlite XAD-16 resin and this future
determination by a flame AAS after elution is proposed. The effect of the factors such as pH, the nature of
complexing agent, sample volume, flow rate, the type and concentration of elution solution on the preconcentration
efficiency have been investigated. The influence of some matrix elements on the recovery of silver were also examined.
It was found, that the quantitative recovery of thiocyanate complex of silver was obtained from nitric acid solution
(pH 2) as 99.20 90.07% at the 95% confidence level. A preconcentration factor up to 75 could be obtained. The
detection limit of silver was 0.047 mg l 1. The adsorption of silver onto Amberlite XAD-16 can be formally described
by a Langmuir equation with maximum adsorption capacity 4.66 mg g 1 (0.043 mmol g 1). The proposed method
was applied to determination of silver in standard alloy with relative error 6.25%. 2000 Elsevier Science B.V. All
rights reserved.

Keywords: Silver; Preconcentration; Amberlite XAD-16; Flame atomic absorption spectrometry

1. Introduction creasing in the use of highly-pure materials [13],

Silver is a metal of commercial importance for
use in high-strength and corrosion resistance al-
loys and jewellery. For example, because of the
marked antibacterial properties of silver its com-
pounds and alloys have been widely used in dental
and pharmaceutical preparations and in im-
planted prosthesis. In addition, because the in-
* Corresponding author. Fax: +90-312-2131235.
E-mail address: rturker@quark.fef.gazi.edu.tr (A.R. Turker)
it is one of the most important components in
electronic devices. Therefore, the development of
high sensitive and high precise methods of trace
amounts of silver determination is very important.
Although, trace amount of silver can be deter-
mined in different media by expensive analytical
methods such as inductively coupled plasma mass
spectrometry [3,4], inductively coupled plasma
atomic emission spectrometry [5,6] and elec-
trothermal atomic absorption spectrometry [7],
the flame atomic absorption spectrometry (FAAS)

0039-9140/00/$ - see front matter 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 3 9 - 9 1 4 0 ( 9 9 ) 0 0 3 4 8 - 3
890 A. Tunceli, A.R. Turker / Talanta 51 (2000) 889894
is not sufficient sensitivity for directly determina-
tion of silver traces. But it is an accepted method
for an easy determination of silver, as, the inter-
ferences of many elements are unimportant and
there is only a small increase of the silver signal
when the concentration of these elements (Fe, Ni,
Mg, Pb, Ca, Cu, Na) is at high level [8].
FAAS determination of silver, needs their pre-
vious separation and preconcentration by using
extraction [1,2,9], ion-exchange [4,5,10] or adsorp-
tion [6,11,12]. The adsorption on various matrixes
is known to be simple and reliable method of
elements preconcentration. For this purpose,
XAD resins [6,13], chitin [12] and silica gel [11]
have been used. Because of purity and good ad-
sorption properties, the commercially available
Amberlite XAD copolymers (XAD-2, -4, -7 and
-8) have been applied to recovery and determina-
tion of various elements [14 16]. In our previous
paper Amberlite XAD-16 has been proposed as
an adsorbent for the determination of gold in
geological samples and anode slime [17].
In this study, Amberlite XAD-16 was used for
the separation and preconcentration of silver from
acidic medium. The optimum conditions for ad-
sorption and determination of silver in alloys are
also discussed.
2. Experimental
2.1. Apparatus
A Philips PU 9285 atomic absorption spectrom-
eter equipped with an air-acetylene burner and
silver hollow cathode lamp operated at 11.2 mA
was used for the determination of silver without
background correction. The operating conditions
were as follows: wavelength: 328.1 nm, bandpass:
0.5 nm and fuel flow rate: 1.1 l min 1. All pH
measurements were made with a Consort digital
pH meter and a combined glass electrode.
2.2. Column preparation
In this study, a glass column (30 mm length and
10 mm id) containing 0.2 g of Amberlite XAD-16
resin with a glass-wool plug over its stopcock was
used. Another plug of glass-wool was placed on
top to avoid the distribution of resin during sam-
ple passage. The column diameter and amount of
resin were chosen taking into account the infor-
mation obtained from our previous studies [18].
Hydrochloric acid (1 mol l 1) was passed through
the column for cleaning and preconditioning after
each using.
2.3. Reagents and standard solutions
All chemicals were of analytical reagent grade.
Deionized and distilled water was used in all
experiments. A stock solution of silver (1000 mg
l 1) was prepared by dissolving the appropriate
amount of silver nitrate (AgNO3) in water. Work-
ing standard solutions (100 mg l 1) were pre-
pared by dilution.
Amberlite XAD-16 (surface area 800 m2 g 1,
wet mesh size 2060) from Rohm and Haas was
used after washing with methanol, 1 mol l 1
hydrochloric acid and deionized and distilled wa-
ter and dried at 60C.
2.4. Optimized method for preconcentration
Silver spiked solution (100 ml, 0.25 mg l 1) was
prepared by adding 1 ml of 0.5 mol l 1 potassium
thiocyanate to 0.25 ml of standard silver solution
(100 mg l 1) and making up to 100 ml with
water. The pH of the solution was adjusted to pH
2 with nitric acid solution. The resulting solution
was passed through the column at a flow rate
about 2.5 ml min 1. The retained metal ions on
the column were eluated into a 10 ml calibrated
flask by using 10 ml of 1 mol l 1 potassium
cyanide solution. Silver content in the eluate was
determined by FAAS. The resin was used repeat-
edly after washing with 10 ml of 1 mol l 1
hydrochloric acid and 10 ml of water. The recov-
ery of silver was calculated from the ratio of
concentration found by FAAS to the concentra-
tion calculated theoretically.
2.5. Sample preparation
The described method was applied to the stan-
dard reference material (NBS 54d). The tin based
 A. Tunceli, A.R. Turker / Talanta 51 (2000) 889894
alloy was dried at 110C for 2 h and  0.5 g of
sample was weighed accurately into a beaker and
heated with 2 ml of concentrated nitric acid up to
reddishbrown NO2 fumes was finished. Then, 10
ml of 50 (v/v)% nitric acid and 2 g of urea (to
reduce NO 2 forming to nitrogen) were added and
continued to heat.
Then, the solution obtained was boiled with 10
ml of water during 10 min. The tin and antimony
oxides formed were filtered through a filter
paper and the insoluble part was washed with
hot water three times. The filtrate and washings
were collected in a 100 ml calibrated flask.
One milliliter of 1 mol l 1 potassium thiocyanate
and 0.02 ml of 0.1 mol l 1 sodium fluoride
have been added to the solution in order to
formation of thiocyanate complex of silver and to
prevent the interference of iron, respectively.
The total volume was adjusted to 100 ml
with water. Silver was preconcentrated from this
acidic solution (pH  1) with the column
procedure described above and determined by
FAAS.
3. Results and discussion
3.1. Effect of pH of solution of sil6er on its
reco6ery
The retention of silver on the resin was studied
as a function of hydronium ion. The pH of the
silver spiked solution (0.25 mg l 1) was adjusted
to desired values with HNO3 and the general
procedure was applied by using potassium
cyanide as an eluent. The results obtained are
shown in Fig. 1. As can be seen, maximum recov-
ery (\99%) was obtained at pH 1 2 and pH 2
was chosen as the optimum pH for subsequent
experiments. However, it was found that 10 2 1
mol l 1 nitric acid solutions have also been ap-
propriate as a medium for preconcentration of
silver. So it can be said that silver could be
preconcentrated from the acidic solution
(pHB2). This is a desired and useful feature,
because the sample solutions are generally acidic
and at higher pH values some precipitation may
occur.
891
Fig. 1. Effect of pH of silver solution on the recovery of silver
(CAg, 0.25 mg l 1; V, 100 ml; complexing agent, KSCN;
eluent, 10 ml of 1 mol l 1 KCN; flow rate, 2.5 ml min 1).
3.2. Effect of nature of eluent
In our pre-experiments, the recovery of silver
ions from acidic medium as its chloride complex
has been studied. For this purpose, pH
of the spiked solution (0.25 mg l 1 Ag) was
adjusted to 2 with HCl and the solution was
passed through the columns at a flow rate of
about 1 ml min 1. The retained silver ions as its
chloride complex were eluated from the columns
by using various eluents. Ammonia, sodium
thiosulfate, potassium iodide, potassium
cyanide, potassium thiocyanate solutions and
some of their mixtures have been tested as an
eluent (Table 1).
As can be seen from the Table 1, maximum
recovery ( 95%) has been obtained by using 1
mol l 1 potassium cyanide solution and its mix-
ture with 1 mol l 1 potassium iodide. This recov-
ery is satisfactory and acceptable for many
purposes. It can be concluded that silver can be
preconcentrated as its chloride complex from
acidic medium.
In order to further increase the recovery of
silver, in the following experiments, the effect of
complexing agent, potassium thiocyanate, has
been investigated by using the same eluent (KCN)
used for chloride complex of silver. This eluent
was also found as a suitable elution solution for
892 A. Tunceli, A.R. Turker / Talanta 51 (2000) 889894

Table 1
The effect of nature and composition of eluation solutions on the recovery of silver

Solution R (%)

No. Component I (C, mol l1) V (ml) Component II (C, mol l1) V (ml)

1 NH3 (6) 10.0 43.2

2 NH3 (conc.) 10.0 61.3
3 Na2S2O3 (1) 10.0 65.2
4 KI (1) 10.0 79.8
5 Na2S2O3 (1) 5.0 NH3 (6) 5.0 73.4
6 KI ( 1) 5.0 NH3 (conc.) 5.0 81.4
7 KI in methanol (0.3) 5.0 Methanol 5.0 82.1
8 KSCN (0.5) 5.0 NH3 (6) 5.0 83.2
9 KSCN (1) 5.0 NH3 (conc.) 5.0 85.3
10 KCN (1) 10.0 95.0
11 KI (1) 5.0 KCN (1) 5.0 95.4

eluting the thiocyanate complex of silver from
acidic medium. Therefore other eluents have not
been examined again.
3.3. Effect of complexing agent
In order to further increase of silver recovery,
potassium thiocyanate has been examined as a
complexing agent. For this purpose, 1 ml of 0.5
mol l 1 potassium thiocyanate was added to 0.25
ml of 100 mg l 1 silver solution and pH was
adjusted to 2 10 with nitric acid (instead of HCl)
and potassium hydroxide. After passing through
the columns retained ions were eluated by using
10 ml of 1 mol l 1 potassium cyanide. The recov-
ery was 99.5% at pH 2. Therefore, potassium
thiocyanate was used as a complexing agent for
subsequent experiments.
3.4. Effect of sample 6olume and analyte
concentration
When dealing with real samples containing very
low concentrations of element ions, the maximum
applicable sample volume must be determined.
For these purposes solutions (100 1000 ml) con-
taining 0.025 mg silver were passed through the
columns by applying general procedure men-
tioned above. The recovery of silver was approxi-
mately quantitative (95%) up to sample volume
of 750 ml. So maximum concentration factor was
75 (sample volume was 750 ml, eluate volume was
10 ml). This result indicates that silver could be
determined with the concentration of  0.033 mg
l 1 by applying the preconcentration procedure
mentioned above. Normally, silver can not be
determined by FAAS at levels below 1 mg l 1
accurately and precisely. Analytical range for sil-
ver is 1.020 mg l 1 [19].
3.5. Effect of flow rate of sample solution on
sorption
The effect of the flow rate of the sample solu-
tion was examined under the optimum conditions
(pH, eluent type, etc.). The silver recovery was 99,
98 and 88% at flow rates 1, 2.5 and 5 ml min 1,
respectively. The flow rate 2.5 ml min 1 was
chosen for subsequent experiments. The flow rate
of eluent was adopted as 1 ml min 1 in all
experiments. In these conditions the preconcentra-
tion procedure takes 50 min for 100 ml of sample
solution (40 min for filtration and 10 min for
elution).
3.6. Adsorption isotherm and adsorption capacity
The adsorption (a, mg g 1) of silver onto resin
as a function of their concentration in solution
was investigated. The results obtained are shown
in Fig. 2. Isotherm of silver adsorption can be
 A. Tunceli, A.R. Turker / Talanta 51 (2000) 889894 893

formally described by a Langmuir equation [20] 3.8. Effects of matrix elements

and linearized in the coordinates [C]/a [C].
Equation y = (6.4 9 0.2) + (0.2148 9 0.0009)x. According to literature [8], low amount of Fe,
The maximum chelating capacity of resin was 4.66 Cu, Ni, Na, K, Ca and Mg have not interfered the
mg g 1 (0.043 mmol g 1) and conditional con- silver determination by FAAS, the signal of silver
stant was found as 3500 l mol 1. was increased at their high concentration. So the
effects of these elements on silver determination at
3.7. Precision of method 1000-times their quantity were investigated. A
fixed amount of silver was taken with various
The precision of the method was examined at amount of matrix elements and the recommended
the optimum conditions (pH, eluate, flow rate, procedure was applied.
etc.) by using 100 ml of solutions containing 0.25 Copper and iron interfered the general precon-
mg l 1 Ag. The mean recovery was 99.209 0.07% centration procedure at the step of preparation of
(n= 8, P= 0.95). RSD was 3.5%. sample solution.
As-well known [21], iron forms with thio-
cyanate a red complex and prevents the formation
of soluble thiocyanate complex of silver. In order
to eliminate this interference, 0.1 mol l 1 NaF
solution was added to the solution drop by drop
until the red colour of solution was dissappeared.
Copper forms with thiocyanate solution black
coloured precipitate of Cu(SCN)2 [22]. However,
it was found experimentally that no precipitate
formed at pH 2. Therefore, pH of the solution
containing silver and copper was adjusted to 2
with nitric acid and then, 1 ml of 0.5 mol l 1
potassium thiocyanate solution was added to this
solution. This solution was passed through the
column and the general preconcentration proce-
dure was applied. The results were given in Table
2.
As can be seen, the increasing of silver signal in
Fig. 2. The isotherm of silver adsorption onto XAD-16 (V, 100
the presence of diverse ions at their C: 250 mg l 1
ml; pH 2).
was 5 8%. These results agree with the results
obtained by literature [8].
Table 2
Effect of matrix elements at a concentration of 250 mg l1 on 3.9. Application
the recovery of silver (0.25 mg l1)
The method proposed was applied to the deter-
Element Recovery of silver (%) mination of silver in a tin base alloy (NBS 54d)
None 99.2
containing (%m/m) Ag: 0.0032, Pb: 0.62, Sb: 7.04,
Cu 106.8 Sn: 88.57, Ni: 0.0027, Bi: 0.044, Cu: 3.62, As:
Ni 101.3 0.088, Fe: 0.027. It was found Ag: 0.00309 0.0001
Fe 103.8 (n=6, at 95% confidence level). The detection
Na 101.4 limit was calculated on the basis of three times of
K 102.5
Mg 102.5
the standard deviation of the absorbance using a
Ca 100.4 series of measurements of blank solutions pre-
pared according to the procedure given in sample
894 A. Tunceli, A.R. Turker / Talanta 51 (2000) 889894
preparation without taking the sample. As can be
seen, the preconcentration, separation and deter-
mination of trace amount of silver could be per-
formed successfully. The relative error was 6.25%.
4. Conclusion
Silver can be determined successfully by using
simple column preconcentration method with
Amberlite XAD-16 in the presence of high con-
centrations of foreign cations such as Fe, Cu, Ni,
Na, K, Ca and Mg from acidic medium. The
precision and accuracy were satisfactory. The pro-
posed method can be used for determination of
silver in alloy.
References
[1] I. Kojima, A. Takayanagi, J. Anal. At. Spectrom. 11
(1996) 607.
[2] E. Vassileva, S. Arpadjan, Analyst 115 (1990) 399.
[3] K. Chiba, I. Inamoto, M. Saeki, J. Anal. At. Spectrom. 7
(1992) 115.
[4] M. Rehkampor, A.N. Halliday, Talanta 44 (1997) 663.
[5] R.P. Singh, E.R. Pambid, Analyst 115 (1990) 301.
[6] Y.S. Chung, R.M. Barnes, J. Anal. At. Spectrom. 3
(1988) 1079.
[7] P. Anderson, C.M. Davidson, D. Littlejohn, M.A. Ure,
C.A. Shand, M.V. Cheshire, Anal. Chim. Acta 327 (1996)
53.
[8] K.C. Thomson, R.J. Reynolds, Atomic Absorption, Fluo-
rescence and Flame Emission Spectroscopy, 2nd ed.,
Charles Griffin and Company, England, UK, 1978.
[9] C.S. Mendoza, S. Kamata, K. Sodeyama, Anal. Sci. 12
(1996) 969.
[10] S. Tao, Y. Shijo, L. Wu, L. Lin, Analyst 119 (1994) 1455.
[11] R.L. Bruening, B.J. Tarbet, K.E. Karakowiak, M.L.
Bruening, R. Izatt, J.S. Bradshaw, Anal. Chem. 63 (1991)
1014.
[12] A. Hase, T. Kawabata, K. Terada, Anal. Sci. 6 (1990)
747.
[13] K. Brajter, E. Dabek-Zlotorzynska, Analyst 113 (1998)
1571.
[14] X.G. Yang, E. Jackwerth, Fresenius J. Anal. Chem. 327
(1987) 179.
[15] X.G. Yang, E. Jackwerth, Fresenius J. Anal. Chem. 331
(1988) 588.
[16] E. Jackwerth, X.G. Yang, C. Xu, Fresenius J. Anal.
Chem. 334 (1989) 514.
[17] A. Tunceli, A.R. Turker, Analyst 122 (1997) 239.
[18] A.R. Turker, A. Tunceli, Fresenius J. Anal. Chem. 345
(1993) 755.
[19] J.W. Robinson, Atomic Spectroscopy, Marcel Dekker,
New York, 1990, p. 134.
[20] A.W. Adamson, Physical Chemistry of Surfaces, 4th edn.,
Wiley Interscience, New York, 1990, p. 521.
[21] D.A. Skoog, D.M. West, Fundamentals of Analytical
Chemistry, 5th edn., Sounders College Publishing, New
York, 1988.
[22] T. Gunduz, Yari-mikro kalitatif analiz, Ankara
U8 niversitesi Fen fakultesi yayinlari, Ankara, 1985