- Email: [email protected]
Potentiometric determination of chlorides with an “Air-Gap” cyanide sensor
0039-9140/93$6.00+ 0.00 Copyright 0 1993Pergamon Press Ltd
Talanra,Vol. 40, No. 9, PP. 1465-1471,1993 Printed in Great Britain. All rights reserved
POTENTIOMETRIC DETERMINATION OF CHLORIDES WITH AN “AIR-GAP” CYANIDE SENSOR RYSZARDBARANOWSKI and TOMASZKUBIK Department of Analytical and General Chemistry, Silesian Technical University, PL 44-101 Gliwice, Poland (Received 4 October
1991. Revised
19 November
1991. Accepted
9 March 1993)
Snmmary-A new electroanalytical method for determining chloride ions with an Air-Gap cyanide sensor system is described. The method is based on the reaction of chloride with mercury(I1) cyanide in dilute sulphuric acid. This reaction leads to hydrogen cyanide which can be determined with an Air-Gap cyanide sensor. Optimum concentrations of mercury(I1) cyanide and sulphuric acid were established and an analytical curve was prepared for 1 x 10-‘-l x 10e5M Cl-. The slope of the calibration curve was equal to 62.8 mV/log C. The correlation coefficient (R ) was equal to 0.9992. The method can determine chloride with good results in high saline solutions and in the presence of surfactants, which is in contrast to direct potentiometry with a chloride electrode. The method was applied for chloride determination in fuses used for initiating explosions. The chlorides were determined both in the raw materials used to prepare the fuse braids and in the other fuse components. Chloride was also determined in drinking water and river water. In dependence of source, chloride amount analyzed in drinking water was in the range 2.18-182.6 mg/l. and 25.8 mg/l. in river water. A comparative analysis was carried out. In the first case, chloride was determined by a turbidimetric method, whereas in the second one by potentiometric titration against a chloride-ISE.
The chloride anion occurs widely in nature. Small quantities of chloride present in natural water and soils are necessary for plants and animals to function correctly. Chlorides participate in oxygen release in photosynthesis, increase hydration of colloids, and influence water balance in plants. They act as a stimulant on plants by influencing the swelling of plasma colloids. The plants absorb chloride from a soil solution through their root system. Chloride anion content in plants is variable and depends on the plant species and organs. Mechanical processing of plant raw materials such as flax and hemp used in the textile industry removes only adsorbed chlorides. Chlorides as components in cellular plasma and combined in organic compounds are not removed during processing. Flax and cotton fibres are contained in fuses used to initiate explosion, when construction elements in high pressure industrial boilers for the nuclear power industry are connected. Due to the corrosive action on construction materials, the fuse components contain as little chloride as possible. For a number of plants, high concentrations of chlorides in soil are toxic. Potassic fertilizers, used for soil manuring, usually contain some of chlorides ballast. These excessive chloride
concentrations can penetrate natural watercourses and reservoirs, considerably increasing their salination. The natural environment, rivers and water underflows are permanently contaminated with some surfactants, coming from municipal and food industry wastes, which unfortunately are not completely biodegradable. The possibility of ion determinations in the presence of surface active agents has been considered in other works.‘-’ The concentrations of the contaminants in the present work was selected based on actual amounts in environmental samples from Upper Silesia. Two potentiometric sensors, for chloride determination, were applied (an Ag, S/AgCl cell and an Air-Gap cyanide sensor). Many industrial electrochemical processes, e.g., copper plating, nickel plating, metals electropurification, aluminum electropassivation, require constant chloride concentration. An excess of chloride leads to irregular build-ups on the electroplating surface (fir-tree crystal formation) so the quality of coppering surfaces gets worse. The presence of chloride causes the anodic dissolution velocity to increase. This way an increasing cathodic current density is possible while electroplating. The oxygen evolution
1465
RYSZARDBARANOWSKI and Tomsz
1466
from the electrolyte solution and the passivation process of the anode is observed while the concentration of chlorides is too small. One method of aluminum electropassivation requires the Cl- level to be less than 20 mg/16 If Fe3+ is present, then Cl- amounts as low as 4 mg/l. lead to formation of pits on the electropassivated aluminum surface.6 Chloride determination in these baths was made with two ion-selective electrodes: Ag, S/AgCl cell and the Air-Gap cyanide sensor. The method described in this paper consists of potentiometric determination of chlorides by applying an Air-Gap cyanide sensor system. Our studies were based on the reaction of mercury cyanide with chloride anions in sulphuric acid, described in the literature’ Hg(CN), -t 2H + + 2X - = HgX, + 2HCN where X- =Cl-,
Br-, I-, SCN-.
Hydrogen cyanide is formed in this reaction. As a weak acid it diffuses from acid solution to the gaseous phase. We proposed to determine the evolving hydrogen cyanide by means of the Air-Gap cyanide sensor system. Therefore, studies have been carried out to determine the analytical parameters such as the concentrations of reagents taking part in the above reaction and the response time of the cyanide sensor. Mercury(I1) cyanide has a high stability constant 1034.7,8which is exploited in this method. In the literature, it is stated that this complex undergoes decomposition in sulphuric acid solution whose concentration exceeds 0.56M.9 The studies performed by other authors prove that a mercury(I1) cyanide does not undergo decomposition in diluted acid solution even at an elevated temperature. lo Besides sulphuric acid, other acids like tartaric, nitric, phosphoric were examined. It was found that they do not cause mercury(I1) cyanide to decompose, but that the complex decomposes in diluted hydrochloric acid solution. Adding chloride, bromide, iodide or thiocyanate to Hg(CN)* solution in diluted sulphuric acid also caused decomposition of mercury(I1) cyanide with hydrogen cyanide evolution.” The design and principle of operation of the Air-Gap cyanide sensor system has been described in the literature.” The cyanide sensor has previously been used to determine cyanides at low concentrations and mercury.9
KUBIK EXPERIMENTAL
Reagents Analytical grade reagents and reagents and redistilled water were used in the studies. Mercury(II) cyanide was synthesized by a literature method.12 From weighed samples of mercury cyanide, solutions with concentrations from 0.1 to 2.0% were prepared. Solutions of 0.1 and 0.3M sulphuric acid, O.lM nitric acid, O.OlM silver nitrate and O.lM sodium chloride were prepared. Sodium chloride solutions in concentrations from 0.01 to lo-‘M were obtained by successive dilutions from 0. 1M stock sodium chloride solution. The electrolyte, O.OlM potassium dicyanoargentate in 1% methylcellulose, was buffered by addition of borax to pH 9.3. The following reagents were used for investigating the influence of surfactants on chloride analysis: -an anionic SULFAPOL (sodium alkylbenzosulphonate) -a weak anionic ROKSOL (block copolymer made from ethylene oxide, propylene and polyoxyethylated alcohols phosphoric acid ester) -a nonionic ROKANOL K-20 (made from addition of ethylene oxide to unsaturated fat alcohols) -a cationic KAMINOX L-l 1 (alkylpolyoxyethylenemethylammonium sulphate). Apparatus The Air-Gap OCN-01 METRON cyanide sensor system and type CT1 ELFA-JOT magnetic stirrer with CP-311 ELMETRON microcomputer pH/mV meter were used. For turbidimetric determination of chlorides, a Spekol spectrophotometer with a TK Carl Zeiss Jena attachment was used. Chlorides were also determined by a potentiometric method using a CRYTUR chloride ISE or RADELKIS chloride ISE or Polish private manufacture chloride ISE and a Ag/AgCl reference electrode with electrolytic nitrate bridge. Mineralization of the examined samples The raw materials used to manufacture fuse braids and its components were mineralized in an oxygen-bomb calorimeter. Approximately 0.5 g samples were placed in a quartz crucible containing wire coil resistors. Wire ends were attached to electrodes. Redistilled water (5 ml), in which combustion products were absorbed, was placed in the bomb calorimeter. The bomb
1467
Potentiometric determination of chlorides
calorimeter was filled with oxygen up to 2.5 MPa. After cornbusting the sample and cooling the bomb, the solution containing the mineralization products was quantitatively transferred to a 25-ml measuring flask. Procedures About 5 ~1 of electrolyte (potassium dicyanoargentate, see Reagents) was placed between an Ag/AgCl reference electrode and a silver electrode of the Air-Gap cyanide sensor system. Next, 1 ml of mercury(I1) cyanide solution, 1 ml of standard sodium chloride solution or sample and 1 ml of sulphuric acid solution were pipetted into a 5-ml sample cell. The sample cell was then closed with a ground-in stopper. A magnetic stirrer was started. After 2 min, the stopper was replaced with the cyanide sensor connected to a millivoltmeter. The response time of the sensor was recorded, i.e., the time elapsed from sensor insertion to the moment when the potential was constant for 1 min. The potential was read with a precision of 1 mV. Each measurement was repeated five times and the average value was calculated. After each measurement, the sensor was taken out of the sample cell, the electrode was rinsed with redistilled water, dried with filter paper and a new electrolyte portion was poured in. The sensor was kept in a vertical position in the neck of a lOO-ml measuring flask containing about 50 ml of distilled water. After each measurement, the test solution was poured out of the sample cell. The cell was flushed with redistilled water, and then rinsed with methanol and thoroughly dried.
constant insignificant increase of the added titrant.
a
Preparation of investigation baths for analysis The samples of copperizing and aluminum electropassivation baths were neutralized with 1M sodium hydroxide and then diluted IO-fold with distilled water. The sample of nickel electroplating baths was also diluted IO-fold, then an additional lOO-fold with distilled water. RESULTS AND DISCUSSION
The effect of mercury(I1) cyanide concentration at constant concentration of sulphuric acid and chloride ions on the potential and response time of the Air-Gap cyanide sensor system was studied. Mercury(I1) cyanide solutions with concentrations from 0,l to 2.0%, 0.3M sulphuric acid and standard 0.1-10-5M sodium chloride solutions were used in the studies. Figure 1 presents the relationship of sensor potential vs. mercury(I1) cyanide solution concentration at a constant sulphuric acid concentration in standard chloride solutions from pC1 = l-5. For pC1 = 1 to 3, the cyanide sensor potential decreases with increasing mercury(I1) cyanide concentration from 0.1 to 0.5%. However, for mercury(I1) cyanide solutions from 0.5 to 2.0% concentration, the cyanide sensor potential is constant. Hence, there is no need to use a mercury(I1) cyanide solution, whose concentration exceeds 0.5%. It can be, therefore, expected that mercury(I1) cyanide reacts quantitatively with chloride ions. When the effect of
-llO-
Turbidimetric determination of chlorides Turbidimetry was applied to determine chlorides in raw materials for manufacturing the fuse braids and other fuse components after mineralization of the samples in a bomb calorimeter. The solutions were filtered by hard filters free of chlorides, and then chlorides were determined as AgCl using the Spekol spectrophotometer with TK attachment.13
-160
-210 -
\
/*
l~C*_*_e_*-*-*
-*
5 9 cq
Potentiometric titration of chlorides Potentiometric titration was applied to determine chlorides in drinking and river water. The interfering species were removed from water samples,i4 and then the chlorides were titrated with O.OlM silver nitrate solution until the largest difference in potential was reached at
Fig. 1. The effect of mercury(I1) cyanide concentration on the potential of the Air-Gap cyanide sensor.
RYSZARD BARANOWSKI and
1468
Hg(CN)2 concentration on a cyanide sensor response for chloride solutions of pC1 = 4 and 5 was examined, a stable potential was observed in the more narrow concentration range of O&0.8%. Due to these facts, a 0.5% Hg(CN)* solution was used in further studies. The effect of Hg(CN)2 concentrations on the response time of the cyanide sensor was examined. Also in this case, 0.L2.0% Hg(CN), solutions, 0.3M sulphuric acid solution and chloride solutions from pC1 = 1 to 5 were used. This relationship is illustrated in Fig. 2. For chlorides of pC1 = 1 and 2, the increasing Hg(CN), concentration did not cause any change in response time of the cyanide sensor. In each case, the response time was 2 min. The effect of Hg(CN)2 solution concentration on the response time of the cyanide sensor for chloride solutions from pC1 = 3 to 5 is more important. The shortest response times were recorded for Hg(CN), solutions from 0.3 to 0.5% concentration. For higher concentrations, the response time was doubled. This study proved that the 0.5% Hg(CN), solution is suitable for determination of chlorides. The effect of sulphuric acid solution concentration on the replacement of cyanide ion by chloride ion with simultaneous evolution of hydrogen cyanide from a liquid phase to a gaseous one was investigated. The cyanide sensor potential was recorded us. mercury(I1) cyanide concentration at pC1 = 3, for 0.3 and O.lM sulphuric acid. The relationship is illustrated in Fig. 3. In case of O.lM sulphuric acid, the measured cyanide sensor potentials dropped
Xl-
16-
%(CW% Fig.
2. The effect of mercury(I1) cyanide concentration the time response of the Air-Gap cyanide sensor.
on
Towsz
KUBIK
CIllKmr
%
Fig. 3. The effect of mercury(I1) cyanide concentration on the potential of the Air-Gap cyanide sensor in various sulphuric acid concentrations.
sharply with increasing mercury(I1) cyanide concentrations. This potential became stable at a concentration above 1.6%, but with a prolonged response time. When a 0.3M H,S04 solution was used, the potentials underwent slight variations within a wide range of concentrations from 0.4 to 2.0%. As in this case, the application of >0.8% mercury(I1) cyanide causes a considerable extension of sensor response time. The 0.5% solution is considered to be optimum. The analytical curve for determining the chlorides within a concentration range of O.l-lo-‘M was plotted for a 0.5% mercury(I1) cyanide solution. Its equation is E = 62.8 . pCl467.5. A correlation coefficient of 0.9992 enables us to determine chlorides within a specified concentration range with high precision. Studies were performed to define the effect of other ions on the determination of chloride. The results are collected in Table 1. Each ion was introduced to 35.5 pug/ml chloride solution at molar ratios of 1: 1, 1: 10 and 1: 100. The anions recorded in the table do not significantly affect the determination of chloride even at loo-fold excess. However, interferences are observed in the presence of halides and pseudohalides. The suitability of the Air-Gap cyanide sensor system for determining chlorides is evaluated in Table 2 by comparing the results with a standard method. When chlorides are determined in raw materials for fuse braids and other
Potentiometric
determination of chlorides
Table 1. Interferences from various anions [A-] in determination (35.5 &ml) Anion
sq-
b-W-1
Cl - determined (&ml )
l-100 36.3-39.6
NO,
BO;-
I-100 34.742.2
l-100 34.2-29.3
components, the best conformity of the test results was achieved for chlorides in braid 2. This braid was cleaned before determination many times by repeated boiling in distilled water to remove chlorides adsorbed on the surface. Similarly, satisfactory agreement of the test results was achieved, when chlorides were determined in braid 3 by the two methods. In the remaining case, slight differences in determinations were found to be due to a distinct inhomogeneity of the tested materials. Chloride was also determined in river and drinking water by the Air-Gap cyanide sensor and by potentiometric titration. The results in Table 3 show good agreement. Small differences lie within experimental error. A low chloride amount in water samples taken from a deep well, required, that titration with an ISE be carried out in a mixed solvent system whereas
of chloride anion HW-
Hzmi I-100 36.8-37.2
l-100 35.8-31.8
the newly developed method determine chlorides in aqueous medium with good accuracy, without additions. The possibility of applying the Air-Gap cyanide sensor to chloride analysis in the presence of surfactants was studied. The analysis of chloride with widely available ISE (Ag,S/AgCl) was made to compare the results obtained. Four surfactants were used for these studies: anionic, weak anionic, cationic and nonionic. The 0.1, 0.01 and 0.005% solutions of these agents of 35.5 mg/l. each, were investigated. Results are compiled in Table 4. The surfactants added to the analysed solutions considerably disturbs the results of chloride determination when the ISE was used, because of the influence of the surfactant on the membrane material. Only the nonionic surface active agent did not cause a negative interference. Chloride analysis with an
Table 2. A comparison of chloride determinations in materials used for fuse preparation performed by the described and comparative methods Found Cl-, % . l@ Sample
n
Air-Gap
s
Raw-material no. 1
5
5.06 *0.41*
0.33
5
5.45 +0.35
0.28
5
3.09*0.15
0.12
5
3.15kO.12
0.10
5
4.86 k 0.10
0.08
5
3.90+0.18
0.14
5
7.50 kO.69
0.55
5
6.95 kO.35
0.28
5
4.88 f 0.32
2.63
5
3.07 + 0.08
0.06
5
1.48kO.32
0.09
5
1.41 * 0.09
0.07
5
2.30 kO.13
0.10
5
2.96 f 0.11
0.09
Raw-material no. 2
Raw-material no. 3 Fuse Braid no. 1 Fuse Braid no. 2 Fuse Braid no. 2 (purified) Fuse Braid no. 3
n -
Turbidvmetric -
s
-
-
5
-
4.22 + 0.53
-
0.42
-
5
2.75 + 0.23
0.19
5
1.48 f0.14
0.11
5
2.43 f 0.22
0.18
n = number of measurements. * = confidence interval, probability, P = 0.95. s = standard deviation.
RYSWRD BARANOWSKI and
1470
TOUA.QZ KUB~K
Table 3. A comparison of chloride determination in selected water samples performed by the described and comparative methods Found Cl-, mgll. Water sample
n
Well water from Myszkow Deep-well water from Myszkow Water from Czama Struga River Drinking water from Gliwice
Air-Gap
s
n 3
5
22.4* 1.20’
1.0
5
2.18 kO.15
5
25.8 f 1.00
5 182.6 f 6.4
Potentiometric titration with HE
s
21.6kO.6
0.2
0.12 3
2.05 f 0.29
0.12
0.08 3
20.8 f 0.4
0.1
190.6 f 4.3
1.7
5.1
3
n = number of measurements. * = confidence interval, probability, P = 0.95. s = standard deviation.
Air-Gap cyanide sensor in solutions containing surfactants was not subject to significant error. Good results are obtained because the Air-Gap electrode is not in contact with the solution. The possibility of applying the Air-Gap cyanide sensor to chloride determination in copperizing, nickel electroplating and aluminum electropassivation was also studied. The results of chloride analysis with the Air-Gap cyanide sensor and an ISE in three of the electrolytic baths are shown in Table 5. These baths are equivalent to those applied in industry. Those baths contain high metal concentration, e.g., copperizing baths contain 170-200 g of CuSO, . 5Hz0 in 1 1. and the nickel plating bath contained 250 g of NiS04. 7H,O and 30 g of NiCl, .6H,O in 1 1. In spite of IO-fold dilution, or lOO-fold dilution in the case of the nickel bath, the concentration of anions and cations
was so large that chloride analysis with ISE was impossible. A very large error was observed because of interaction between dissolved ions and the ion selective electrode membrane material. In the case of the Air-Gap cyanide sensor application, the chlorides were determined in solutions 1-3, with a relative error not higher than 5.6%. The other anions and cations present in the solution did not interfere because the sensor was not directly connected to the solution. In an acidic (pH 1) solution of mercury(I1) cyanide, when hydrogen cyanide was generated as a result of the reaction with chloride, the cyanide complexes of other metals cannot be formed, so measurements by the Air-Gap cyanide sensor were correct. The above results show the advantage of the developed chloride method in high saline
Table 4. Comparison of results of chloride analysis made with Air-Gap cyanide sensor and ISE in solution of surface active agents Taken 35.5 mg Cl-/i.
Found Cl- mg/l.
C, % (w/w)
ISE (Polish)
RE %
Anionic SULFAPOL
0.005 0.01 0.1
37.2 38.2 25.5
4.8 7.6 28.2
Weak Anionic ROKSOL
0.005 0.01 0.1
40.3 45.0 55.0
Nonionic ROKANOL K-20
0.005 0.01 0.1
Kationic KAMINOX L-11
0.005 0.01 0.01
Surfactant
RE = relative error.
ISE (RADELKIS)
RE %
Air-Gap
RE %
32.7 30.9 21.2
7.9 13.0 40.3
36.4 37.1 37.6
2.5 4.5 5.5
13.5 26.8 54.9
36.9 38.2 42.8
3.9 7.6 20.6
35.9 36.9 37.2
1.1 3.9 4.8
36.4 36.4 37.0
2.5 2.5 4.2
33.9 32.7 32.7
4.5 7.9 7.9
34.5 36.9 37.2
2.8 3.9 4.8
56.2 65.3 72.8
58.3 83.9 105.1
41.6 57.1 61.2
17.2 60.8 72.4
35.9 35.9 37.3
1.1 1.1 5.1
Potentiometric
determination of chlorides
1471
Table 5. Comparison of results of chloride analysis made with Air-Gap cyanide sensor and ISE in selected electrolytical baths Found, ClElectrolytic bath
Taken, Cl-
ISE (Polish)
RE %
ISE (RADELKIS)
RE %
Copperizing bath
70.9 mgli.
27.6 mgll.
61.1
61.9 mgll.
12.7
74.8 mgll.
5.5
Nickel plating bath
8.90 m&d
10.9 m~lm~
22.5
IO.1 mg~ml
13.5
8.4 m&m/
5.6
Aluminium electropassivation
20.0 mgll.
50.4 mgll.
152.0
33.5 mgll.
67.5
19.8 mgll.
1.1
Air-Gap
RE %
RE = relative error.
solution as compared to direct potentiomet~. Our investigations show that an Air-Gap cyanide sensor could be applied to chloride ion determination in such systems where the classical ion selective electrode action is disturbed by some active agents, e.g., surfactants or highly concentrated electrolytes.
strated in comparison to the direct method with ISE. The advantages of this method make it more competitive than other well-known methods. Furthermore, it extends the applicability of the Air-Gap cyanide sensor system used until now solely in the analysis of cyanides and mercury, to trace chlorides.
CONCLUSIONS
REFERENCES
A new indirect electroanalytical method for determining the chlorides using the Air-Gap cyanide sensor system has been developed. It has a number of advantages. The method is quick, not requiring the expensive analytical equipment. Owing to simple construction, the cyanide sensor can be made at low expense, even in laboratories that are not equipped for regular man~acture of such eq~pment. The small dimensions of the sensor, measuring cell and magnetic stirrer permit analysis of trace chlorides in small volume solutions. The use of a milli-voltmeter and battery supplied magnetic stirrer enables the method to be performed in the field. This new method makes it possible to determine chlorides over a wide concentrations range i.e., 0.1-10-5M. The suitability of the Air-Gap cyanide sensor for chloride analysis in high saline solutions and su~actant-containing solutions was demon-
I. P. Pakalns and Y. J. Farrar, Water Res., 1976,10, 1087. 2. A. Hulanicki, M. Trojanowicz and E. Pobozy, Analyst, 1982, lW, 1356. 3. A. Craggs, G. J. Moody, J. D. R. Thomas and B. J. Birch, ibid., 1980, 105, 426. 4. A. J. Frend, G. J. Moody, J. D. R. Thomas and B. J. Birch, ibid., 1983, 108,1072. 5. H. Hara and S. Okazaki, ibid., 1985, 110, 11. 6. Multi-author work. Poradnik Gufwurzotechnikn, WNT, Warszawa 1985. 7. M. K. Bathy and P. C. Uden, Tahta, 1971, 8, 799. 8. J. Ciba et al. Poradnik Chemiku Analityka, Vol. I, p. 123. WNT, Warszawa, 1989. 9. P. Czicholi, J. Fligier and Z. Gregorowicz, Anal. Chim. Actu, 1981, 126, 221. 10. F. Fe@ and F. L. Chan, Mkrochim. Acta, 1%7,2,339. Il. J. Fligier, P. Czichori and Z. Gregorowicz, Anal. Chim. Acta, 1980, 118, 145. 12. J. Supniewski, Preparatyku Nieorganiczna, 1st Ed., p. 675. PWN, Warszawa, 1958. 13. Z. Marczenko, Spectrophotometric Determination of Elements, 2nd Ed., Ellis Horwood, New York, 1976. 14. W. Hermanowicz, Ftzyczno-chemiczne Badanie Wody I Seiekriw, 1st Ed., p. 504. Arkady, Warszawa, 1976.
Talanra,Vol. 40, No. 9, PP. 1465-1471,1993 Printed in Great Britain. All rights reserved
POTENTIOMETRIC DETERMINATION OF CHLORIDES WITH AN “AIR-GAP” CYANIDE SENSOR RYSZARDBARANOWSKI and TOMASZKUBIK Department of Analytical and General Chemistry, Silesian Technical University, PL 44-101 Gliwice, Poland (Received 4 October
1991. Revised
19 November
1991. Accepted
9 March 1993)
Snmmary-A new electroanalytical method for determining chloride ions with an Air-Gap cyanide sensor system is described. The method is based on the reaction of chloride with mercury(I1) cyanide in dilute sulphuric acid. This reaction leads to hydrogen cyanide which can be determined with an Air-Gap cyanide sensor. Optimum concentrations of mercury(I1) cyanide and sulphuric acid were established and an analytical curve was prepared for 1 x 10-‘-l x 10e5M Cl-. The slope of the calibration curve was equal to 62.8 mV/log C. The correlation coefficient (R ) was equal to 0.9992. The method can determine chloride with good results in high saline solutions and in the presence of surfactants, which is in contrast to direct potentiometry with a chloride electrode. The method was applied for chloride determination in fuses used for initiating explosions. The chlorides were determined both in the raw materials used to prepare the fuse braids and in the other fuse components. Chloride was also determined in drinking water and river water. In dependence of source, chloride amount analyzed in drinking water was in the range 2.18-182.6 mg/l. and 25.8 mg/l. in river water. A comparative analysis was carried out. In the first case, chloride was determined by a turbidimetric method, whereas in the second one by potentiometric titration against a chloride-ISE.
The chloride anion occurs widely in nature. Small quantities of chloride present in natural water and soils are necessary for plants and animals to function correctly. Chlorides participate in oxygen release in photosynthesis, increase hydration of colloids, and influence water balance in plants. They act as a stimulant on plants by influencing the swelling of plasma colloids. The plants absorb chloride from a soil solution through their root system. Chloride anion content in plants is variable and depends on the plant species and organs. Mechanical processing of plant raw materials such as flax and hemp used in the textile industry removes only adsorbed chlorides. Chlorides as components in cellular plasma and combined in organic compounds are not removed during processing. Flax and cotton fibres are contained in fuses used to initiate explosion, when construction elements in high pressure industrial boilers for the nuclear power industry are connected. Due to the corrosive action on construction materials, the fuse components contain as little chloride as possible. For a number of plants, high concentrations of chlorides in soil are toxic. Potassic fertilizers, used for soil manuring, usually contain some of chlorides ballast. These excessive chloride
concentrations can penetrate natural watercourses and reservoirs, considerably increasing their salination. The natural environment, rivers and water underflows are permanently contaminated with some surfactants, coming from municipal and food industry wastes, which unfortunately are not completely biodegradable. The possibility of ion determinations in the presence of surface active agents has been considered in other works.‘-’ The concentrations of the contaminants in the present work was selected based on actual amounts in environmental samples from Upper Silesia. Two potentiometric sensors, for chloride determination, were applied (an Ag, S/AgCl cell and an Air-Gap cyanide sensor). Many industrial electrochemical processes, e.g., copper plating, nickel plating, metals electropurification, aluminum electropassivation, require constant chloride concentration. An excess of chloride leads to irregular build-ups on the electroplating surface (fir-tree crystal formation) so the quality of coppering surfaces gets worse. The presence of chloride causes the anodic dissolution velocity to increase. This way an increasing cathodic current density is possible while electroplating. The oxygen evolution
1465
RYSZARDBARANOWSKI and Tomsz
1466
from the electrolyte solution and the passivation process of the anode is observed while the concentration of chlorides is too small. One method of aluminum electropassivation requires the Cl- level to be less than 20 mg/16 If Fe3+ is present, then Cl- amounts as low as 4 mg/l. lead to formation of pits on the electropassivated aluminum surface.6 Chloride determination in these baths was made with two ion-selective electrodes: Ag, S/AgCl cell and the Air-Gap cyanide sensor. The method described in this paper consists of potentiometric determination of chlorides by applying an Air-Gap cyanide sensor system. Our studies were based on the reaction of mercury cyanide with chloride anions in sulphuric acid, described in the literature’ Hg(CN), -t 2H + + 2X - = HgX, + 2HCN where X- =Cl-,
Br-, I-, SCN-.
Hydrogen cyanide is formed in this reaction. As a weak acid it diffuses from acid solution to the gaseous phase. We proposed to determine the evolving hydrogen cyanide by means of the Air-Gap cyanide sensor system. Therefore, studies have been carried out to determine the analytical parameters such as the concentrations of reagents taking part in the above reaction and the response time of the cyanide sensor. Mercury(I1) cyanide has a high stability constant 1034.7,8which is exploited in this method. In the literature, it is stated that this complex undergoes decomposition in sulphuric acid solution whose concentration exceeds 0.56M.9 The studies performed by other authors prove that a mercury(I1) cyanide does not undergo decomposition in diluted acid solution even at an elevated temperature. lo Besides sulphuric acid, other acids like tartaric, nitric, phosphoric were examined. It was found that they do not cause mercury(I1) cyanide to decompose, but that the complex decomposes in diluted hydrochloric acid solution. Adding chloride, bromide, iodide or thiocyanate to Hg(CN)* solution in diluted sulphuric acid also caused decomposition of mercury(I1) cyanide with hydrogen cyanide evolution.” The design and principle of operation of the Air-Gap cyanide sensor system has been described in the literature.” The cyanide sensor has previously been used to determine cyanides at low concentrations and mercury.9
KUBIK EXPERIMENTAL
Reagents Analytical grade reagents and reagents and redistilled water were used in the studies. Mercury(II) cyanide was synthesized by a literature method.12 From weighed samples of mercury cyanide, solutions with concentrations from 0.1 to 2.0% were prepared. Solutions of 0.1 and 0.3M sulphuric acid, O.lM nitric acid, O.OlM silver nitrate and O.lM sodium chloride were prepared. Sodium chloride solutions in concentrations from 0.01 to lo-‘M were obtained by successive dilutions from 0. 1M stock sodium chloride solution. The electrolyte, O.OlM potassium dicyanoargentate in 1% methylcellulose, was buffered by addition of borax to pH 9.3. The following reagents were used for investigating the influence of surfactants on chloride analysis: -an anionic SULFAPOL (sodium alkylbenzosulphonate) -a weak anionic ROKSOL (block copolymer made from ethylene oxide, propylene and polyoxyethylated alcohols phosphoric acid ester) -a nonionic ROKANOL K-20 (made from addition of ethylene oxide to unsaturated fat alcohols) -a cationic KAMINOX L-l 1 (alkylpolyoxyethylenemethylammonium sulphate). Apparatus The Air-Gap OCN-01 METRON cyanide sensor system and type CT1 ELFA-JOT magnetic stirrer with CP-311 ELMETRON microcomputer pH/mV meter were used. For turbidimetric determination of chlorides, a Spekol spectrophotometer with a TK Carl Zeiss Jena attachment was used. Chlorides were also determined by a potentiometric method using a CRYTUR chloride ISE or RADELKIS chloride ISE or Polish private manufacture chloride ISE and a Ag/AgCl reference electrode with electrolytic nitrate bridge. Mineralization of the examined samples The raw materials used to manufacture fuse braids and its components were mineralized in an oxygen-bomb calorimeter. Approximately 0.5 g samples were placed in a quartz crucible containing wire coil resistors. Wire ends were attached to electrodes. Redistilled water (5 ml), in which combustion products were absorbed, was placed in the bomb calorimeter. The bomb
1467
Potentiometric determination of chlorides
calorimeter was filled with oxygen up to 2.5 MPa. After cornbusting the sample and cooling the bomb, the solution containing the mineralization products was quantitatively transferred to a 25-ml measuring flask. Procedures About 5 ~1 of electrolyte (potassium dicyanoargentate, see Reagents) was placed between an Ag/AgCl reference electrode and a silver electrode of the Air-Gap cyanide sensor system. Next, 1 ml of mercury(I1) cyanide solution, 1 ml of standard sodium chloride solution or sample and 1 ml of sulphuric acid solution were pipetted into a 5-ml sample cell. The sample cell was then closed with a ground-in stopper. A magnetic stirrer was started. After 2 min, the stopper was replaced with the cyanide sensor connected to a millivoltmeter. The response time of the sensor was recorded, i.e., the time elapsed from sensor insertion to the moment when the potential was constant for 1 min. The potential was read with a precision of 1 mV. Each measurement was repeated five times and the average value was calculated. After each measurement, the sensor was taken out of the sample cell, the electrode was rinsed with redistilled water, dried with filter paper and a new electrolyte portion was poured in. The sensor was kept in a vertical position in the neck of a lOO-ml measuring flask containing about 50 ml of distilled water. After each measurement, the test solution was poured out of the sample cell. The cell was flushed with redistilled water, and then rinsed with methanol and thoroughly dried.
constant insignificant increase of the added titrant.
a
Preparation of investigation baths for analysis The samples of copperizing and aluminum electropassivation baths were neutralized with 1M sodium hydroxide and then diluted IO-fold with distilled water. The sample of nickel electroplating baths was also diluted IO-fold, then an additional lOO-fold with distilled water. RESULTS AND DISCUSSION
The effect of mercury(I1) cyanide concentration at constant concentration of sulphuric acid and chloride ions on the potential and response time of the Air-Gap cyanide sensor system was studied. Mercury(I1) cyanide solutions with concentrations from 0,l to 2.0%, 0.3M sulphuric acid and standard 0.1-10-5M sodium chloride solutions were used in the studies. Figure 1 presents the relationship of sensor potential vs. mercury(I1) cyanide solution concentration at a constant sulphuric acid concentration in standard chloride solutions from pC1 = l-5. For pC1 = 1 to 3, the cyanide sensor potential decreases with increasing mercury(I1) cyanide concentration from 0.1 to 0.5%. However, for mercury(I1) cyanide solutions from 0.5 to 2.0% concentration, the cyanide sensor potential is constant. Hence, there is no need to use a mercury(I1) cyanide solution, whose concentration exceeds 0.5%. It can be, therefore, expected that mercury(I1) cyanide reacts quantitatively with chloride ions. When the effect of
-llO-
Turbidimetric determination of chlorides Turbidimetry was applied to determine chlorides in raw materials for manufacturing the fuse braids and other fuse components after mineralization of the samples in a bomb calorimeter. The solutions were filtered by hard filters free of chlorides, and then chlorides were determined as AgCl using the Spekol spectrophotometer with TK attachment.13
-160
-210 -
\
/*
l~C*_*_e_*-*-*
-*
5 9 cq
Potentiometric titration of chlorides Potentiometric titration was applied to determine chlorides in drinking and river water. The interfering species were removed from water samples,i4 and then the chlorides were titrated with O.OlM silver nitrate solution until the largest difference in potential was reached at
Fig. 1. The effect of mercury(I1) cyanide concentration on the potential of the Air-Gap cyanide sensor.
RYSZARD BARANOWSKI and
1468
Hg(CN)2 concentration on a cyanide sensor response for chloride solutions of pC1 = 4 and 5 was examined, a stable potential was observed in the more narrow concentration range of O&0.8%. Due to these facts, a 0.5% Hg(CN)* solution was used in further studies. The effect of Hg(CN)2 concentrations on the response time of the cyanide sensor was examined. Also in this case, 0.L2.0% Hg(CN), solutions, 0.3M sulphuric acid solution and chloride solutions from pC1 = 1 to 5 were used. This relationship is illustrated in Fig. 2. For chlorides of pC1 = 1 and 2, the increasing Hg(CN), concentration did not cause any change in response time of the cyanide sensor. In each case, the response time was 2 min. The effect of Hg(CN)2 solution concentration on the response time of the cyanide sensor for chloride solutions from pC1 = 3 to 5 is more important. The shortest response times were recorded for Hg(CN), solutions from 0.3 to 0.5% concentration. For higher concentrations, the response time was doubled. This study proved that the 0.5% Hg(CN), solution is suitable for determination of chlorides. The effect of sulphuric acid solution concentration on the replacement of cyanide ion by chloride ion with simultaneous evolution of hydrogen cyanide from a liquid phase to a gaseous one was investigated. The cyanide sensor potential was recorded us. mercury(I1) cyanide concentration at pC1 = 3, for 0.3 and O.lM sulphuric acid. The relationship is illustrated in Fig. 3. In case of O.lM sulphuric acid, the measured cyanide sensor potentials dropped
Xl-
16-
%(CW% Fig.
2. The effect of mercury(I1) cyanide concentration the time response of the Air-Gap cyanide sensor.
on
Towsz
KUBIK
CIllKmr
%
Fig. 3. The effect of mercury(I1) cyanide concentration on the potential of the Air-Gap cyanide sensor in various sulphuric acid concentrations.
sharply with increasing mercury(I1) cyanide concentrations. This potential became stable at a concentration above 1.6%, but with a prolonged response time. When a 0.3M H,S04 solution was used, the potentials underwent slight variations within a wide range of concentrations from 0.4 to 2.0%. As in this case, the application of >0.8% mercury(I1) cyanide causes a considerable extension of sensor response time. The 0.5% solution is considered to be optimum. The analytical curve for determining the chlorides within a concentration range of O.l-lo-‘M was plotted for a 0.5% mercury(I1) cyanide solution. Its equation is E = 62.8 . pCl467.5. A correlation coefficient of 0.9992 enables us to determine chlorides within a specified concentration range with high precision. Studies were performed to define the effect of other ions on the determination of chloride. The results are collected in Table 1. Each ion was introduced to 35.5 pug/ml chloride solution at molar ratios of 1: 1, 1: 10 and 1: 100. The anions recorded in the table do not significantly affect the determination of chloride even at loo-fold excess. However, interferences are observed in the presence of halides and pseudohalides. The suitability of the Air-Gap cyanide sensor system for determining chlorides is evaluated in Table 2 by comparing the results with a standard method. When chlorides are determined in raw materials for fuse braids and other
Potentiometric
determination of chlorides
Table 1. Interferences from various anions [A-] in determination (35.5 &ml) Anion
sq-
b-W-1
Cl - determined (&ml )
l-100 36.3-39.6
NO,
BO;-
I-100 34.742.2
l-100 34.2-29.3
components, the best conformity of the test results was achieved for chlorides in braid 2. This braid was cleaned before determination many times by repeated boiling in distilled water to remove chlorides adsorbed on the surface. Similarly, satisfactory agreement of the test results was achieved, when chlorides were determined in braid 3 by the two methods. In the remaining case, slight differences in determinations were found to be due to a distinct inhomogeneity of the tested materials. Chloride was also determined in river and drinking water by the Air-Gap cyanide sensor and by potentiometric titration. The results in Table 3 show good agreement. Small differences lie within experimental error. A low chloride amount in water samples taken from a deep well, required, that titration with an ISE be carried out in a mixed solvent system whereas
of chloride anion HW-
Hzmi I-100 36.8-37.2
l-100 35.8-31.8
the newly developed method determine chlorides in aqueous medium with good accuracy, without additions. The possibility of applying the Air-Gap cyanide sensor to chloride analysis in the presence of surfactants was studied. The analysis of chloride with widely available ISE (Ag,S/AgCl) was made to compare the results obtained. Four surfactants were used for these studies: anionic, weak anionic, cationic and nonionic. The 0.1, 0.01 and 0.005% solutions of these agents of 35.5 mg/l. each, were investigated. Results are compiled in Table 4. The surfactants added to the analysed solutions considerably disturbs the results of chloride determination when the ISE was used, because of the influence of the surfactant on the membrane material. Only the nonionic surface active agent did not cause a negative interference. Chloride analysis with an
Table 2. A comparison of chloride determinations in materials used for fuse preparation performed by the described and comparative methods Found Cl-, % . l@ Sample
n
Air-Gap
s
Raw-material no. 1
5
5.06 *0.41*
0.33
5
5.45 +0.35
0.28
5
3.09*0.15
0.12
5
3.15kO.12
0.10
5
4.86 k 0.10
0.08
5
3.90+0.18
0.14
5
7.50 kO.69
0.55
5
6.95 kO.35
0.28
5
4.88 f 0.32
2.63
5
3.07 + 0.08
0.06
5
1.48kO.32
0.09
5
1.41 * 0.09
0.07
5
2.30 kO.13
0.10
5
2.96 f 0.11
0.09
Raw-material no. 2
Raw-material no. 3 Fuse Braid no. 1 Fuse Braid no. 2 Fuse Braid no. 2 (purified) Fuse Braid no. 3
n -
Turbidvmetric -
s
-
-
5
-
4.22 + 0.53
-
0.42
-
5
2.75 + 0.23
0.19
5
1.48 f0.14
0.11
5
2.43 f 0.22
0.18
n = number of measurements. * = confidence interval, probability, P = 0.95. s = standard deviation.
RYSWRD BARANOWSKI and
1470
TOUA.QZ KUB~K
Table 3. A comparison of chloride determination in selected water samples performed by the described and comparative methods Found Cl-, mgll. Water sample
n
Well water from Myszkow Deep-well water from Myszkow Water from Czama Struga River Drinking water from Gliwice
Air-Gap
s
n 3
5
22.4* 1.20’
1.0
5
2.18 kO.15
5
25.8 f 1.00
5 182.6 f 6.4
Potentiometric titration with HE
s
21.6kO.6
0.2
0.12 3
2.05 f 0.29
0.12
0.08 3
20.8 f 0.4
0.1
190.6 f 4.3
1.7
5.1
3
n = number of measurements. * = confidence interval, probability, P = 0.95. s = standard deviation.
Air-Gap cyanide sensor in solutions containing surfactants was not subject to significant error. Good results are obtained because the Air-Gap electrode is not in contact with the solution. The possibility of applying the Air-Gap cyanide sensor to chloride determination in copperizing, nickel electroplating and aluminum electropassivation was also studied. The results of chloride analysis with the Air-Gap cyanide sensor and an ISE in three of the electrolytic baths are shown in Table 5. These baths are equivalent to those applied in industry. Those baths contain high metal concentration, e.g., copperizing baths contain 170-200 g of CuSO, . 5Hz0 in 1 1. and the nickel plating bath contained 250 g of NiS04. 7H,O and 30 g of NiCl, .6H,O in 1 1. In spite of IO-fold dilution, or lOO-fold dilution in the case of the nickel bath, the concentration of anions and cations
was so large that chloride analysis with ISE was impossible. A very large error was observed because of interaction between dissolved ions and the ion selective electrode membrane material. In the case of the Air-Gap cyanide sensor application, the chlorides were determined in solutions 1-3, with a relative error not higher than 5.6%. The other anions and cations present in the solution did not interfere because the sensor was not directly connected to the solution. In an acidic (pH 1) solution of mercury(I1) cyanide, when hydrogen cyanide was generated as a result of the reaction with chloride, the cyanide complexes of other metals cannot be formed, so measurements by the Air-Gap cyanide sensor were correct. The above results show the advantage of the developed chloride method in high saline
Table 4. Comparison of results of chloride analysis made with Air-Gap cyanide sensor and ISE in solution of surface active agents Taken 35.5 mg Cl-/i.
Found Cl- mg/l.
C, % (w/w)
ISE (Polish)
RE %
Anionic SULFAPOL
0.005 0.01 0.1
37.2 38.2 25.5
4.8 7.6 28.2
Weak Anionic ROKSOL
0.005 0.01 0.1
40.3 45.0 55.0
Nonionic ROKANOL K-20
0.005 0.01 0.1
Kationic KAMINOX L-11
0.005 0.01 0.01
Surfactant
RE = relative error.
ISE (RADELKIS)
RE %
Air-Gap
RE %
32.7 30.9 21.2
7.9 13.0 40.3
36.4 37.1 37.6
2.5 4.5 5.5
13.5 26.8 54.9
36.9 38.2 42.8
3.9 7.6 20.6
35.9 36.9 37.2
1.1 3.9 4.8
36.4 36.4 37.0
2.5 2.5 4.2
33.9 32.7 32.7
4.5 7.9 7.9
34.5 36.9 37.2
2.8 3.9 4.8
56.2 65.3 72.8
58.3 83.9 105.1
41.6 57.1 61.2
17.2 60.8 72.4
35.9 35.9 37.3
1.1 1.1 5.1
Potentiometric
determination of chlorides
1471
Table 5. Comparison of results of chloride analysis made with Air-Gap cyanide sensor and ISE in selected electrolytical baths Found, ClElectrolytic bath
Taken, Cl-
ISE (Polish)
RE %
ISE (RADELKIS)
RE %
Copperizing bath
70.9 mgli.
27.6 mgll.
61.1
61.9 mgll.
12.7
74.8 mgll.
5.5
Nickel plating bath
8.90 m&d
10.9 m~lm~
22.5
IO.1 mg~ml
13.5
8.4 m&m/
5.6
Aluminium electropassivation
20.0 mgll.
50.4 mgll.
152.0
33.5 mgll.
67.5
19.8 mgll.
1.1
Air-Gap
RE %
RE = relative error.
solution as compared to direct potentiomet~. Our investigations show that an Air-Gap cyanide sensor could be applied to chloride ion determination in such systems where the classical ion selective electrode action is disturbed by some active agents, e.g., surfactants or highly concentrated electrolytes.
strated in comparison to the direct method with ISE. The advantages of this method make it more competitive than other well-known methods. Furthermore, it extends the applicability of the Air-Gap cyanide sensor system used until now solely in the analysis of cyanides and mercury, to trace chlorides.
CONCLUSIONS
REFERENCES
A new indirect electroanalytical method for determining the chlorides using the Air-Gap cyanide sensor system has been developed. It has a number of advantages. The method is quick, not requiring the expensive analytical equipment. Owing to simple construction, the cyanide sensor can be made at low expense, even in laboratories that are not equipped for regular man~acture of such eq~pment. The small dimensions of the sensor, measuring cell and magnetic stirrer permit analysis of trace chlorides in small volume solutions. The use of a milli-voltmeter and battery supplied magnetic stirrer enables the method to be performed in the field. This new method makes it possible to determine chlorides over a wide concentrations range i.e., 0.1-10-5M. The suitability of the Air-Gap cyanide sensor for chloride analysis in high saline solutions and su~actant-containing solutions was demon-
I. P. Pakalns and Y. J. Farrar, Water Res., 1976,10, 1087. 2. A. Hulanicki, M. Trojanowicz and E. Pobozy, Analyst, 1982, lW, 1356. 3. A. Craggs, G. J. Moody, J. D. R. Thomas and B. J. Birch, ibid., 1980, 105, 426. 4. A. J. Frend, G. J. Moody, J. D. R. Thomas and B. J. Birch, ibid., 1983, 108,1072. 5. H. Hara and S. Okazaki, ibid., 1985, 110, 11. 6. Multi-author work. Poradnik Gufwurzotechnikn, WNT, Warszawa 1985. 7. M. K. Bathy and P. C. Uden, Tahta, 1971, 8, 799. 8. J. Ciba et al. Poradnik Chemiku Analityka, Vol. I, p. 123. WNT, Warszawa, 1989. 9. P. Czicholi, J. Fligier and Z. Gregorowicz, Anal. Chim. Actu, 1981, 126, 221. 10. F. Fe@ and F. L. Chan, Mkrochim. Acta, 1%7,2,339. Il. J. Fligier, P. Czichori and Z. Gregorowicz, Anal. Chim. Acta, 1980, 118, 145. 12. J. Supniewski, Preparatyku Nieorganiczna, 1st Ed., p. 675. PWN, Warszawa, 1958. 13. Z. Marczenko, Spectrophotometric Determination of Elements, 2nd Ed., Ellis Horwood, New York, 1976. 14. W. Hermanowicz, Ftzyczno-chemiczne Badanie Wody I Seiekriw, 1st Ed., p. 504. Arkady, Warszawa, 1976.