- Email: [email protected]
Continuous monitoring for cyanide in waste water with a galvanic hydrogen cyanide sensor using a purge system
Talanta 48 (1999) 997 – 1004
Continuous monitoring for cyanide in waste water with a galvanic hydrogen cyanide sensor using a purge system Hiromitsu Hachiya a, Satoshi Ito a, Yoshito Fushinuki b, Takashi Masadome c, Yasukazu Asano c,*, Toshihiko Imato d a R & D di6ision, DKK Corporation, 4 -13 -14 Kichijoji, Kitamachi, Musashino-shi, Tokyo 180 -0001, Japan Scientific In6estigation Research Laboratory, Kagoshima Pref. Police HQ, 10 -1 Kamoikeshinmachi, Kagoshima-shi, Kagoshima 890 -0064, Japan c Ariake National College of Technology, 150 Higashihagiomachi, Omuta-shi, Fukuoka 836 -8585, Japan d Faculty of Engineering, Kyushu Uni6ersity, 6 -10 -1 Hakozaki, Higashi-ku, Fukuoka-shi, Fukuoka 812 -0053, Japan
b
Received 17 August 1998; received in revised form 2 September 1998; accepted 2 September 1998
Abstract A continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor and an aeration pump for purging was developed. Hydrogen cyanide evolved from cyanide solution using a purging pump was measured with the hydrogen cyanide sensor. The system showed good performance in terms of stability and selectivity. A linear calibration curve was obtained in the concentrating range from 0 to 15 mg dm3 of cyanide ion with a slope of −0.24 mA mg − 1 dm − 3. The lower detection limit was 0.1 mg dm − 3. The 90% response time of the sensor system was within 3.5 min for a 0.5 mg dm − 3 cyanide solution, when the flow rate of the purging air was 1 dm3 min − 1. The system maintained the initial performance for 6 months in the field test. The developed galvanic sensor system was not subject to interference from sulfide and residual chlorine, compared with a potentiometric sensor system developed previously. The analytical results obtained by the present system were in good agreement with those obtained by the pyridine pyrazolone method. The correlation factor and regression line between both methods were 0.979 and Y = 2.30 × 10 − 4 +1.12X, respectively. This system was successfully applied for a continuous monitoring of cyanide ion in waste water. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Galvanic cell; Chemical sensor; Cyanide; Continuous monitoring; Gas phase; Industrial waste water
1. Introduction
* Corresponding author. Tel.: +81-944-53-8876; e-mail: [email protected].
Cyanide is one of the most toxic compounds but it has been used widely in industrial fields such as hydrometalogy and metal plating due to its excellent chemical properties. Cyanide concen-
0039-9140/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 9 8 ) 0 0 3 1 4 - 2
998
H. Hachiya et al. / Talanta 48 (1999) 997–1004
tration in industrial waste water had been regulated strictly to be 1 mg dm − 3 in Japan [1,2] because the leakage of cyanide causes serious damage to the ecosystem. A potentiometric cyanide ion selective electrode has been widely used as a continuous cyanide-monitor to control cyanide in industrial waste water. However, the method needs several procedures such as sampling, adjustments of pH and ionic strength of the samples prior to determine. Therefore a simple and more inexpensive monitoring system for cyanide is needed to replace the system with the conventional ion selective electrode [3 –5]. Absorption spectrometry based on the Kno¨ig reaction such as the pyridine pyrazolone method has also been used to determine cyanide. However this method needs several procedures and is timeconsuming [6 – 8]. Recently, rapid and simple analytical methods for cyanide using ion chromatography [9,10] and flow injection analysis [11,12] have been reported. However, both methods may have some difficulty in application as a continuous, maintenance-free monitor because a sampling pump, a separation column and pretreatment of the samples are needed and the running costs seem to be high for continuous monitoring. Leakage of cyanide into the environment is accidental, therefore a continuous cyanide-monitoring system with the characteristics of simplicity, selectivity, reliability, low cost and long-stability in the field is desirable presently. We have developed a system for monitoring cyanide by combining a potentiometric gas sensor for hydrogen cyanide with an air-bubbling flow system [13,14]. Though the developed system is simple and of low-cost and is of commercial use, it is subject to interference from concentrated residual chlorine and traces of sulfide which coexist in the sample solution. These interferences may be due to the fact that the sensing element of the system consists of Ag2S which shows a response to residual chlorine and sulfide [14]. In this work we have developed a galvanic hydrogen cyanide sensor system in order to improve the interference from residual chlorine and sulfide. The monitoring system consists of an air pump, a purge tube, a galvanic hydrogen cyanide
sensor and an ammeter. The performance of the sensor system and its application to waste water monitoring are described.
2. Experimental
2.1. Apparatus The output-current of the galvanic hydrogen cyanide sensor was measured by an electrometer (model 612, Keithley Instruments). The signals from the electrometer were fed to a recorder (model LR4110, Yokogawa Denki). An aeration pump (model NS-SUN, Nissei) was used for airpurging. A silver ion-selective electrode (model 7084L, DKK) and a double junction type reference electrode (model 4083, DKK) were used to standardize the cyanide stock solution by argentometry.
2.2. Reagents All reagents used were of analytical grade. Deionized water (Milli-Q water) was used throughout the experiment. A stock solution of 1000 mg dm − 3 cyanide solution was prepared by dissolving 1.252 g potassium cyanide in 0.5 dm3 of a 0.1 mg dm − 3 NaOH solution. This solution was standardized by argentometry before use. A 0.25 mol dm − 3 phosphate buffer solution was prepared by dissolving 34.0 g KH2PO4 and 35.5 g Na2HPO4 in 1 dm deionized water.
2.3. Gal6anic hydrogen cyanide sensor The structure of the galvanic hydrogen cyanide sensor is shown in Fig. 1. An inner electrode, which consists of a silver working electrode and a silver counter electrode, was fabricated. The silver working electrode (2mm diameter, plate) was fixed at the end of the inner body and a silver wire (0.4 mm diameter, 350 mm long) was reeled around the inner body. The inner electrode was inserted into a sensor body (26 mm outer diameter, 65 mm long). The sensor body has a compartment for an internal filling solution. A gas-permeable polytetrafuluoroethylene (PTFE)
H. Hachiya et al. / Talanta 48 (1999) 997–1004
membrane was fixed at the end of the sensor body by an O-ring. The pore size and porosity of the membrane were 0.22 mm and 65%, respectively. An internal filling solution (4 cm3), which was a mixture of 4.6 mg dm − 3 silver nitrate and 0.02 mg dm − 3 nitric acid in 70% ethylene glycol to prevent evaporation of the inner solution, was filled into the inner compartment of the sensor body. The thickness of the internal filling solution layer between the working electrode and the PTFE membrane was less than 0.1 mm. Because the sensor was a galvanic cell, any potential between the working and the counter electrode was not applied.
999
2.4. Principle of determination of cyanide concentration by the hydrogen cyanide sensor In an aqueous sample solution containing cyanide, the following equilibrium is established between the hydrogen ion, cyanide ion and hydrogen cyanide; H + + CN − = HCN
(1)
The formation constant of hydrogen cyanide is expressed as follows: [HCN]/[H +] [CN − ]= 1/Ka
(2)
where Ka is the acid dissociation constant and is reported as 10 − 9.24. From Eq. (2), using the pH of the solution, the concentration ratio of HCN and CN − is expressed as follows: [HCN]/[CN − ]= 109.24 − pH
(3)
In the neutral pH region (pH 6.86), most of cyanide exists in the form of hydrogen cyanide judging from Eq. (3). When air is bubbled into the sample solution at the flow rate of more than 1 dm3 min − 1, hydrogen cyanide is evolved into a gaseous phase immediately. When the gas phase is transported to the hydrogen cyanide sensor, the hydrogen cyanide permeates through the PTFE membrane of the sensor and is dissolved in the internal filling solution. The hydrogen cyanide is oxidized according to Eq. (4) at the surface of the working electrode spontaneously, while the silver ion internal filling is reduced to silver according to the Eq. (5) [15] at the counter electrode. Thus the current generated is measured by the ammeter.
Fig. 1. Structure of the galvanic hydrogen cyanide sensor. 1, Terminal; 2, inner solution (4.6 ×10 − 1 M Ag + /0.2 M HNO3 in70% ethylene glycol); 3, inner body; 4, lead wire from counter electrode; 5, membrane for pressure balance; 6, counter electrode (f 4 Ag wire 350 mm); 7, O-ring; 8, gas permeable membrane; 9, membrane support; 10, working electrode (f 2 Ag plate); 11, lead wire from working electrode; 12, sensor body.
Ag+ 2HCNAg(CN)2 − + 2H + + e −
(4)
Ag + + e − Ag
(5)
Since the current changes are proportional to the hydrogen cyanide concentration in aqueous phase, and the hydrogen concentration is related to the cyanide in aqueous solution by Henry’s law, the cyanide concentration in the sample solution can be determined by measuring the current from the hydrogen cyanide sensor.
1000
H. Hachiya et al. / Talanta 48 (1999) 997–1004
cyanide-monitor is shown in Fig. 2. Air was bubbled through the Teflon tube (3 mm i.d.) at 70 mm from the surface of the sample solution by the aeration pump. The hydrogen cyanide evolved from the solution was transported to the sensor by air through the polypropylene tube (27 mm i.d., 1.0 m long) immersed in the sample solution at depth of 15 cm. The evolved hydrogen cyanide was detected by the sensor located at the top of the polypropylene tube. Therefore, the sensor signals was reproducible since the surface of the sensor was always kept clean. The current from the sensor was measured by the ammeter and its signals were fed to a recorder. The hydrogen cyanide and air were evacuated through a hole which was bored at 80 mm from the top of the polypropylene tube. 3. Results and discussion Fig. 2. Continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor.
2.5. E6aluation of the performance of the gal6anic hydrogen cyanide sensor The performance of the sensor in the gaseous phase was evaluated in a 1 dm3 brown colored closed vessel containing 0.1 dm3 cyanide solutions of different concentrations at pH 6.86 adjusted by phosphate buffer. The sensor was suspended in the aqueous phase of the vessel containing the cyanide solution (pH 6.86) and then the sensor was inserted into the gas phase of the closed sample solution. The temperature of the solution was kept constant. The current from the sensor was measured after gas – liquid equilibrium was attained. The accuracy of the results from the sensor was evaluated by measuring the hydrogen cyanide gas in the gas phase after absorbing hydrogen cyanide gas in a 0.1 mol dm − 3 NaOH solution and by argentometric titration of the absorbed solution.
3.1. Performance of the hydrogen cyanide sensor 3.1.1. Calibration cur6e The calibration curve of the hydrogen cyanide sensor, which was obtained for the cyanide solutions (pH 6.86) at 25°C in the closed vessel is shown in Fig. 3. The linear calibration curve was obtained in the range 0–15 mg dm − 3 cyanide concentration. The slope of the calibration curve was − 0.24 mA mg − 1 dm − 3. The lower detection limit of the sensor was 0.1 mg dm − 3 cyanide.
2.6. Construction of the continuous monitoring system for cyanide The schematic diagram of the continuous
Fig. 3. Typical calibration curve for cyanide.
H. Hachiya et al. / Talanta 48 (1999) 997–1004
Fig. 4. Response time of the galvanic hydrogen cyanide sensor (0.1 mg dm − 3 cyanide, pH 6.68).
3.1.2. Response time Fig. 4 shows the typical response curve of the 0.1 mg dm − 3 cyanide solution. The 90% response times for 0.1, 0.5, 3 and 10 mg dm − 3 cyanide solutions were 282, 208, 91 and 47 s, respectively. The 50% response times for 0.1, 0.5, 3 and 10 mg dm − 3 cyanide solutions were 80, 98, 24 and 13 s, respectively. The fast response time is characteristic of the present sensor even for a sample solution of low cyanide concentration.
1001
usually contains residual chlorine. Therefore the cyanide monitor should have good selectivity for cyanide against residual chlorine. Although the present hydrogen cyanide sensor was immersed in the brown colored closed vessel containing 100 dm3 of 500 mg dm − 3 hypochlorite solution for 30 min to examine the interference from residual chlorine, the current from the sensor was not changed. Even after immersing the sensor for 30 min. in the hypochlorite solution, the sensitivity to cyanide remained at the same level as the initial sensitivity of − 0.24 mA mg − 1 dm − 3. In addition, when the sensor was exposed to 1 ppm Cl2 gas for 5 min it did not show any response to the Cl2 gas. However, the potentiometric hydrogen cyanide sensor showed 65% errors for measurement of 1 and 10 mg dm − 3 cyanide when exposed to the same Cl2 gas. From these experiments the present galvanic sensor was confirmed to be applicable to the aqueous phase containing residual chlorine and the gaseous phase containing chlorine. On the other hand, the sensor was subjected to interference from sulfide as shown in Fig. 5. The current from the sensor for 0.1, 1 and 10 mg dm3 sulfide solutions corresponded to the one for 1.41, 15.8 and 65.6 mg dm − 3 cyanide solution, respectively. Sulfide exists as dissolved hydrogen sulfide
3.1.3. Reproducibility and stability The reproducibility and the long-term stability of the sensor are important for continuous monitoring in the field. The reproducibilities for 0.1, 1, 3 and 10 mg dm − 3 cyanide solutions were 1.0, 4.3, 8.4 and 7.8%, respectively. The variation of the current from the sensor for the 0.5 mg dm − 3 cyanide solution over a period for 6 months was within 920% reproducibility. These results indicate that the good reproducibility and long-term stability of the present sensor meet the requirement continuous monitoring. 3.1.4. Effect of interferences Because the waste water containing cyanide is generally treated with hypochlorite for decomposition of cyanide, the waste water after treatment
Fig. 5. Effect of hydrosulfide: , cyanide; , hydrosulfide.
1002
H. Hachiya et al. / Talanta 48 (1999) 997–1004
tiometric hydrogen cyanide sensor reported previously.
3.1.5. Effect of temperature The effect of temperature on the sensor response is shown in Fig. 7. The sensitivity and the slope of the calibration curve increased with increasing temperature. The increasing sensitivity of the sensor with temperature may be due to the fact that the concentration of hydrogen cyanide in the gas phase increases with increasing temperature of the aqueous phase. This indicates that the sensor response is subject to temperature variation of the sample solution. In order to reduce the effect of temperature a temperature compensation circuit was combined in the system.. 3.2. Performance of the continuous monitoring system for cyanide 3.2.1. Response time The performance of the continuous monitoring system for cyanide as shown in Fig. 2 was evaluated with respect to response time by aeration through the cyanide standard solutions instead of the sample solutions. The 90% response times of the system to 0.5 mg dm − 3 was 3.5 min. The longer response time of the system compared to the hydrogen sensor itself may be due to the fact Fig. 6. Comparison of the effect of sulfide between the galvanic and the potentiometric sensors.
in the solution at near neutral pH. The sensor signals to a 1 mg dm − 3 cyanide solution and a 1 mg dm − 3 sulfide solution correspond to 2 ppm HCN and 8 ppm H2S, respectively. However, the effect of sulfide on the signal of the present sensor was not so large compared with that of the potentiometric hydrogen cyanide sensor, as shown in Fig. 6 which indicates that sulfide is insoluble in the inner solution because the inner solution of the present sensor is acidic while that of the potentiometric sensor is alkaline. Accordingly, the reactivity of the present sensor to sulfide is very low and, as a result, the present galvanic hydrogen cyanide sensor is more selective for sulfide compared to the poten-
Fig. 7. Effect of temperature on the sensor signals.
H. Hachiya et al. / Talanta 48 (1999) 997–1004
1003
such as waste water. 3.2.3. Stability of the sensor system in the field The long term stability of the present system was tested by setting the system in the waste water from a plating factory and the results are shown in Fig. 9. The system was periodically checked with a standard solution of 0.5 mg dm − 3 cyanide during the period of the test. The variation for 6 months was within 920%. The initial performance of the cyanide monitor remained constant for 6 months.
4. Conclusion
Fig. 8. Correlation between the present method and the conventional method. The samples are waste water from a plating factory.
that there is a time-lag to attain the phase equilibrium and transportation of the evolved hydrogen cyanide to the sensor by air. Shortening the length of the gas purge tube and more powerful aeration may reduced the response time.
3.2.2. Correlation between the present amperometric method and a con6entional pyridine pyrazolone method The present continuous monitoring system for cyanide was applied to the determination of cyanide in waste water. Several waste water samples containing cyanide from a plating factory were used. The analytical results obtained by the present sensor system method were compared with those from the conventional pyridine pyrazolone method. The correlation between two methods was fairly good as shown in Fig. 8. From these data the regression line expressed by Y=2.30×10 − 4 +1.12X and a correlation factor of 0.979 were obtained. This good correlation indicates that the present sensor system can be applied for the continuous monitoring of cyanide directly in nearly neutral pH solutions
We have developed a galvanic hydrogen cyanide sensor, in which selectivity to cyanide against sulfide and residual chlorine was improved and used it in a continuous monitoring system with a purge unit for the detection of cyanide. A linear response curve was obtained in the range from 0 to 15 mg dm − 3 cyanide concentration. The lower detection limit was 0.1 mg dm − 3, and the 90% response time of the system was shorter than 3.5 min for 0.5 mg dm − 3 cyanide solution. The system maintained a good
Fig. 9. Long-term stability test of the continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor. Field, a plating factory. The system was checked with a standard solution of 0.5 mg dm − 3 cyanide periodically.
1004
H. Hachiya et al. / Talanta 48 (1999) 997–1004
performance for 6 months. This long-term stability was due to the gas phase measurement in which the surface of the sensor was kept clean. Residual chlorine and chlorine gas did not interfere with the present system. Although there was an error for the measurement of a cyanide sample in the presence of sulfide greater than 0.1 mg dm − 3, selectivity of the sensor system against sulfide was improved greatly compared to that of the potentiometric hydrogen cyanide sensor system. This improved selectivity against sulfide and residual chlorine was due to the low reactivity with the sensing element and the low solubility of those gases in the inner solution. A linear regression line with a correlation factor of 0.979 was obtained between the present amperometric method and the conventional pyridine pyrazolone method. It can be concluded that the present system is applicable as a continuous monitor for cyanide in industrial waste water.
.
References [1] Regulations by The Prime Minister’s Office on Waste Water Standards: Basic Six Environmental Laws, Chuohoki Publisher, Japan, 1995. [2] Testing Methods for Industrial Waste Water, Japanese Industrial Standard JIS K 0102-1991, pp. 122 – 123. [3] Y. Asano, Chem. Eng. 17 (1972) 90 – 97. [4] M.S. Frant, J.W. Ross Jr., J.H. Riseman, Anal. Chem. 44 (1972) 2227 – 2232. [5] R.J. O’Herron, EPA-670/4-75-005, April 1975, p. 7. [6] J. Estein, Anal. Chem. 19 (1947) 272. [7] Testing Methods for Industrial Waste Water, Japanese Industrial Standard JIS K 0102-1991, pp. 119 – 121. [8] E. Tanaka, S. Ohara, S. Adachi, M. Nunoura, J. Jpn. Water Works Assoc. 61 (1992) 21. [9] J.E. Girad, Anal. Chem. 51 (1979) 836 – 839. [10] Y. Liu, R.D. Rocklin, R.J. Joyce, M. Doyle, Anal. Chem. 62 (1990) 766. [11] O. Elshoz, W. Frenzel, C.-Y. Liu, J. Moller, Fresenius’ Z. Anal. Chem. 338 (1990) 159. [12] V. Kuban, Anal. Chem. 64 (1992) 1106. [13] Y. Asano and S. Ito, Japanese Patent, 1416967 (1987). [14] Y. Asano, S. Ito, Bunseki Kagaku 39 (1990) 785 – 787. [15] W.M. Latimer, Oxidation Potentials, 2nd edn, PrenticeHall, Englewood Cliffs, 1952, p. 191.
Continuous monitoring for cyanide in waste water with a galvanic hydrogen cyanide sensor using a purge system Hiromitsu Hachiya a, Satoshi Ito a, Yoshito Fushinuki b, Takashi Masadome c, Yasukazu Asano c,*, Toshihiko Imato d a R & D di6ision, DKK Corporation, 4 -13 -14 Kichijoji, Kitamachi, Musashino-shi, Tokyo 180 -0001, Japan Scientific In6estigation Research Laboratory, Kagoshima Pref. Police HQ, 10 -1 Kamoikeshinmachi, Kagoshima-shi, Kagoshima 890 -0064, Japan c Ariake National College of Technology, 150 Higashihagiomachi, Omuta-shi, Fukuoka 836 -8585, Japan d Faculty of Engineering, Kyushu Uni6ersity, 6 -10 -1 Hakozaki, Higashi-ku, Fukuoka-shi, Fukuoka 812 -0053, Japan
b
Received 17 August 1998; received in revised form 2 September 1998; accepted 2 September 1998
Abstract A continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor and an aeration pump for purging was developed. Hydrogen cyanide evolved from cyanide solution using a purging pump was measured with the hydrogen cyanide sensor. The system showed good performance in terms of stability and selectivity. A linear calibration curve was obtained in the concentrating range from 0 to 15 mg dm3 of cyanide ion with a slope of −0.24 mA mg − 1 dm − 3. The lower detection limit was 0.1 mg dm − 3. The 90% response time of the sensor system was within 3.5 min for a 0.5 mg dm − 3 cyanide solution, when the flow rate of the purging air was 1 dm3 min − 1. The system maintained the initial performance for 6 months in the field test. The developed galvanic sensor system was not subject to interference from sulfide and residual chlorine, compared with a potentiometric sensor system developed previously. The analytical results obtained by the present system were in good agreement with those obtained by the pyridine pyrazolone method. The correlation factor and regression line between both methods were 0.979 and Y = 2.30 × 10 − 4 +1.12X, respectively. This system was successfully applied for a continuous monitoring of cyanide ion in waste water. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Galvanic cell; Chemical sensor; Cyanide; Continuous monitoring; Gas phase; Industrial waste water
1. Introduction
* Corresponding author. Tel.: +81-944-53-8876; e-mail: [email protected].
Cyanide is one of the most toxic compounds but it has been used widely in industrial fields such as hydrometalogy and metal plating due to its excellent chemical properties. Cyanide concen-
0039-9140/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 9 8 ) 0 0 3 1 4 - 2
998
H. Hachiya et al. / Talanta 48 (1999) 997–1004
tration in industrial waste water had been regulated strictly to be 1 mg dm − 3 in Japan [1,2] because the leakage of cyanide causes serious damage to the ecosystem. A potentiometric cyanide ion selective electrode has been widely used as a continuous cyanide-monitor to control cyanide in industrial waste water. However, the method needs several procedures such as sampling, adjustments of pH and ionic strength of the samples prior to determine. Therefore a simple and more inexpensive monitoring system for cyanide is needed to replace the system with the conventional ion selective electrode [3 –5]. Absorption spectrometry based on the Kno¨ig reaction such as the pyridine pyrazolone method has also been used to determine cyanide. However this method needs several procedures and is timeconsuming [6 – 8]. Recently, rapid and simple analytical methods for cyanide using ion chromatography [9,10] and flow injection analysis [11,12] have been reported. However, both methods may have some difficulty in application as a continuous, maintenance-free monitor because a sampling pump, a separation column and pretreatment of the samples are needed and the running costs seem to be high for continuous monitoring. Leakage of cyanide into the environment is accidental, therefore a continuous cyanide-monitoring system with the characteristics of simplicity, selectivity, reliability, low cost and long-stability in the field is desirable presently. We have developed a system for monitoring cyanide by combining a potentiometric gas sensor for hydrogen cyanide with an air-bubbling flow system [13,14]. Though the developed system is simple and of low-cost and is of commercial use, it is subject to interference from concentrated residual chlorine and traces of sulfide which coexist in the sample solution. These interferences may be due to the fact that the sensing element of the system consists of Ag2S which shows a response to residual chlorine and sulfide [14]. In this work we have developed a galvanic hydrogen cyanide sensor system in order to improve the interference from residual chlorine and sulfide. The monitoring system consists of an air pump, a purge tube, a galvanic hydrogen cyanide
sensor and an ammeter. The performance of the sensor system and its application to waste water monitoring are described.
2. Experimental
2.1. Apparatus The output-current of the galvanic hydrogen cyanide sensor was measured by an electrometer (model 612, Keithley Instruments). The signals from the electrometer were fed to a recorder (model LR4110, Yokogawa Denki). An aeration pump (model NS-SUN, Nissei) was used for airpurging. A silver ion-selective electrode (model 7084L, DKK) and a double junction type reference electrode (model 4083, DKK) were used to standardize the cyanide stock solution by argentometry.
2.2. Reagents All reagents used were of analytical grade. Deionized water (Milli-Q water) was used throughout the experiment. A stock solution of 1000 mg dm − 3 cyanide solution was prepared by dissolving 1.252 g potassium cyanide in 0.5 dm3 of a 0.1 mg dm − 3 NaOH solution. This solution was standardized by argentometry before use. A 0.25 mol dm − 3 phosphate buffer solution was prepared by dissolving 34.0 g KH2PO4 and 35.5 g Na2HPO4 in 1 dm deionized water.
2.3. Gal6anic hydrogen cyanide sensor The structure of the galvanic hydrogen cyanide sensor is shown in Fig. 1. An inner electrode, which consists of a silver working electrode and a silver counter electrode, was fabricated. The silver working electrode (2mm diameter, plate) was fixed at the end of the inner body and a silver wire (0.4 mm diameter, 350 mm long) was reeled around the inner body. The inner electrode was inserted into a sensor body (26 mm outer diameter, 65 mm long). The sensor body has a compartment for an internal filling solution. A gas-permeable polytetrafuluoroethylene (PTFE)
H. Hachiya et al. / Talanta 48 (1999) 997–1004
membrane was fixed at the end of the sensor body by an O-ring. The pore size and porosity of the membrane were 0.22 mm and 65%, respectively. An internal filling solution (4 cm3), which was a mixture of 4.6 mg dm − 3 silver nitrate and 0.02 mg dm − 3 nitric acid in 70% ethylene glycol to prevent evaporation of the inner solution, was filled into the inner compartment of the sensor body. The thickness of the internal filling solution layer between the working electrode and the PTFE membrane was less than 0.1 mm. Because the sensor was a galvanic cell, any potential between the working and the counter electrode was not applied.
999
2.4. Principle of determination of cyanide concentration by the hydrogen cyanide sensor In an aqueous sample solution containing cyanide, the following equilibrium is established between the hydrogen ion, cyanide ion and hydrogen cyanide; H + + CN − = HCN
(1)
The formation constant of hydrogen cyanide is expressed as follows: [HCN]/[H +] [CN − ]= 1/Ka
(2)
where Ka is the acid dissociation constant and is reported as 10 − 9.24. From Eq. (2), using the pH of the solution, the concentration ratio of HCN and CN − is expressed as follows: [HCN]/[CN − ]= 109.24 − pH
(3)
In the neutral pH region (pH 6.86), most of cyanide exists in the form of hydrogen cyanide judging from Eq. (3). When air is bubbled into the sample solution at the flow rate of more than 1 dm3 min − 1, hydrogen cyanide is evolved into a gaseous phase immediately. When the gas phase is transported to the hydrogen cyanide sensor, the hydrogen cyanide permeates through the PTFE membrane of the sensor and is dissolved in the internal filling solution. The hydrogen cyanide is oxidized according to Eq. (4) at the surface of the working electrode spontaneously, while the silver ion internal filling is reduced to silver according to the Eq. (5) [15] at the counter electrode. Thus the current generated is measured by the ammeter.
Fig. 1. Structure of the galvanic hydrogen cyanide sensor. 1, Terminal; 2, inner solution (4.6 ×10 − 1 M Ag + /0.2 M HNO3 in70% ethylene glycol); 3, inner body; 4, lead wire from counter electrode; 5, membrane for pressure balance; 6, counter electrode (f 4 Ag wire 350 mm); 7, O-ring; 8, gas permeable membrane; 9, membrane support; 10, working electrode (f 2 Ag plate); 11, lead wire from working electrode; 12, sensor body.
Ag+ 2HCNAg(CN)2 − + 2H + + e −
(4)
Ag + + e − Ag
(5)
Since the current changes are proportional to the hydrogen cyanide concentration in aqueous phase, and the hydrogen concentration is related to the cyanide in aqueous solution by Henry’s law, the cyanide concentration in the sample solution can be determined by measuring the current from the hydrogen cyanide sensor.
1000
H. Hachiya et al. / Talanta 48 (1999) 997–1004
cyanide-monitor is shown in Fig. 2. Air was bubbled through the Teflon tube (3 mm i.d.) at 70 mm from the surface of the sample solution by the aeration pump. The hydrogen cyanide evolved from the solution was transported to the sensor by air through the polypropylene tube (27 mm i.d., 1.0 m long) immersed in the sample solution at depth of 15 cm. The evolved hydrogen cyanide was detected by the sensor located at the top of the polypropylene tube. Therefore, the sensor signals was reproducible since the surface of the sensor was always kept clean. The current from the sensor was measured by the ammeter and its signals were fed to a recorder. The hydrogen cyanide and air were evacuated through a hole which was bored at 80 mm from the top of the polypropylene tube. 3. Results and discussion Fig. 2. Continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor.
2.5. E6aluation of the performance of the gal6anic hydrogen cyanide sensor The performance of the sensor in the gaseous phase was evaluated in a 1 dm3 brown colored closed vessel containing 0.1 dm3 cyanide solutions of different concentrations at pH 6.86 adjusted by phosphate buffer. The sensor was suspended in the aqueous phase of the vessel containing the cyanide solution (pH 6.86) and then the sensor was inserted into the gas phase of the closed sample solution. The temperature of the solution was kept constant. The current from the sensor was measured after gas – liquid equilibrium was attained. The accuracy of the results from the sensor was evaluated by measuring the hydrogen cyanide gas in the gas phase after absorbing hydrogen cyanide gas in a 0.1 mol dm − 3 NaOH solution and by argentometric titration of the absorbed solution.
3.1. Performance of the hydrogen cyanide sensor 3.1.1. Calibration cur6e The calibration curve of the hydrogen cyanide sensor, which was obtained for the cyanide solutions (pH 6.86) at 25°C in the closed vessel is shown in Fig. 3. The linear calibration curve was obtained in the range 0–15 mg dm − 3 cyanide concentration. The slope of the calibration curve was − 0.24 mA mg − 1 dm − 3. The lower detection limit of the sensor was 0.1 mg dm − 3 cyanide.
2.6. Construction of the continuous monitoring system for cyanide The schematic diagram of the continuous
Fig. 3. Typical calibration curve for cyanide.
H. Hachiya et al. / Talanta 48 (1999) 997–1004
Fig. 4. Response time of the galvanic hydrogen cyanide sensor (0.1 mg dm − 3 cyanide, pH 6.68).
3.1.2. Response time Fig. 4 shows the typical response curve of the 0.1 mg dm − 3 cyanide solution. The 90% response times for 0.1, 0.5, 3 and 10 mg dm − 3 cyanide solutions were 282, 208, 91 and 47 s, respectively. The 50% response times for 0.1, 0.5, 3 and 10 mg dm − 3 cyanide solutions were 80, 98, 24 and 13 s, respectively. The fast response time is characteristic of the present sensor even for a sample solution of low cyanide concentration.
1001
usually contains residual chlorine. Therefore the cyanide monitor should have good selectivity for cyanide against residual chlorine. Although the present hydrogen cyanide sensor was immersed in the brown colored closed vessel containing 100 dm3 of 500 mg dm − 3 hypochlorite solution for 30 min to examine the interference from residual chlorine, the current from the sensor was not changed. Even after immersing the sensor for 30 min. in the hypochlorite solution, the sensitivity to cyanide remained at the same level as the initial sensitivity of − 0.24 mA mg − 1 dm − 3. In addition, when the sensor was exposed to 1 ppm Cl2 gas for 5 min it did not show any response to the Cl2 gas. However, the potentiometric hydrogen cyanide sensor showed 65% errors for measurement of 1 and 10 mg dm − 3 cyanide when exposed to the same Cl2 gas. From these experiments the present galvanic sensor was confirmed to be applicable to the aqueous phase containing residual chlorine and the gaseous phase containing chlorine. On the other hand, the sensor was subjected to interference from sulfide as shown in Fig. 5. The current from the sensor for 0.1, 1 and 10 mg dm3 sulfide solutions corresponded to the one for 1.41, 15.8 and 65.6 mg dm − 3 cyanide solution, respectively. Sulfide exists as dissolved hydrogen sulfide
3.1.3. Reproducibility and stability The reproducibility and the long-term stability of the sensor are important for continuous monitoring in the field. The reproducibilities for 0.1, 1, 3 and 10 mg dm − 3 cyanide solutions were 1.0, 4.3, 8.4 and 7.8%, respectively. The variation of the current from the sensor for the 0.5 mg dm − 3 cyanide solution over a period for 6 months was within 920% reproducibility. These results indicate that the good reproducibility and long-term stability of the present sensor meet the requirement continuous monitoring. 3.1.4. Effect of interferences Because the waste water containing cyanide is generally treated with hypochlorite for decomposition of cyanide, the waste water after treatment
Fig. 5. Effect of hydrosulfide: , cyanide; , hydrosulfide.
1002
H. Hachiya et al. / Talanta 48 (1999) 997–1004
tiometric hydrogen cyanide sensor reported previously.
3.1.5. Effect of temperature The effect of temperature on the sensor response is shown in Fig. 7. The sensitivity and the slope of the calibration curve increased with increasing temperature. The increasing sensitivity of the sensor with temperature may be due to the fact that the concentration of hydrogen cyanide in the gas phase increases with increasing temperature of the aqueous phase. This indicates that the sensor response is subject to temperature variation of the sample solution. In order to reduce the effect of temperature a temperature compensation circuit was combined in the system.. 3.2. Performance of the continuous monitoring system for cyanide 3.2.1. Response time The performance of the continuous monitoring system for cyanide as shown in Fig. 2 was evaluated with respect to response time by aeration through the cyanide standard solutions instead of the sample solutions. The 90% response times of the system to 0.5 mg dm − 3 was 3.5 min. The longer response time of the system compared to the hydrogen sensor itself may be due to the fact Fig. 6. Comparison of the effect of sulfide between the galvanic and the potentiometric sensors.
in the solution at near neutral pH. The sensor signals to a 1 mg dm − 3 cyanide solution and a 1 mg dm − 3 sulfide solution correspond to 2 ppm HCN and 8 ppm H2S, respectively. However, the effect of sulfide on the signal of the present sensor was not so large compared with that of the potentiometric hydrogen cyanide sensor, as shown in Fig. 6 which indicates that sulfide is insoluble in the inner solution because the inner solution of the present sensor is acidic while that of the potentiometric sensor is alkaline. Accordingly, the reactivity of the present sensor to sulfide is very low and, as a result, the present galvanic hydrogen cyanide sensor is more selective for sulfide compared to the poten-
Fig. 7. Effect of temperature on the sensor signals.
H. Hachiya et al. / Talanta 48 (1999) 997–1004
1003
such as waste water. 3.2.3. Stability of the sensor system in the field The long term stability of the present system was tested by setting the system in the waste water from a plating factory and the results are shown in Fig. 9. The system was periodically checked with a standard solution of 0.5 mg dm − 3 cyanide during the period of the test. The variation for 6 months was within 920%. The initial performance of the cyanide monitor remained constant for 6 months.
4. Conclusion
Fig. 8. Correlation between the present method and the conventional method. The samples are waste water from a plating factory.
that there is a time-lag to attain the phase equilibrium and transportation of the evolved hydrogen cyanide to the sensor by air. Shortening the length of the gas purge tube and more powerful aeration may reduced the response time.
3.2.2. Correlation between the present amperometric method and a con6entional pyridine pyrazolone method The present continuous monitoring system for cyanide was applied to the determination of cyanide in waste water. Several waste water samples containing cyanide from a plating factory were used. The analytical results obtained by the present sensor system method were compared with those from the conventional pyridine pyrazolone method. The correlation between two methods was fairly good as shown in Fig. 8. From these data the regression line expressed by Y=2.30×10 − 4 +1.12X and a correlation factor of 0.979 were obtained. This good correlation indicates that the present sensor system can be applied for the continuous monitoring of cyanide directly in nearly neutral pH solutions
We have developed a galvanic hydrogen cyanide sensor, in which selectivity to cyanide against sulfide and residual chlorine was improved and used it in a continuous monitoring system with a purge unit for the detection of cyanide. A linear response curve was obtained in the range from 0 to 15 mg dm − 3 cyanide concentration. The lower detection limit was 0.1 mg dm − 3, and the 90% response time of the system was shorter than 3.5 min for 0.5 mg dm − 3 cyanide solution. The system maintained a good
Fig. 9. Long-term stability test of the continuous monitoring system for cyanide with a galvanic hydrogen cyanide sensor. Field, a plating factory. The system was checked with a standard solution of 0.5 mg dm − 3 cyanide periodically.
1004
H. Hachiya et al. / Talanta 48 (1999) 997–1004
performance for 6 months. This long-term stability was due to the gas phase measurement in which the surface of the sensor was kept clean. Residual chlorine and chlorine gas did not interfere with the present system. Although there was an error for the measurement of a cyanide sample in the presence of sulfide greater than 0.1 mg dm − 3, selectivity of the sensor system against sulfide was improved greatly compared to that of the potentiometric hydrogen cyanide sensor system. This improved selectivity against sulfide and residual chlorine was due to the low reactivity with the sensing element and the low solubility of those gases in the inner solution. A linear regression line with a correlation factor of 0.979 was obtained between the present amperometric method and the conventional pyridine pyrazolone method. It can be concluded that the present system is applicable as a continuous monitor for cyanide in industrial waste water.
.
References [1] Regulations by The Prime Minister’s Office on Waste Water Standards: Basic Six Environmental Laws, Chuohoki Publisher, Japan, 1995. [2] Testing Methods for Industrial Waste Water, Japanese Industrial Standard JIS K 0102-1991, pp. 122 – 123. [3] Y. Asano, Chem. Eng. 17 (1972) 90 – 97. [4] M.S. Frant, J.W. Ross Jr., J.H. Riseman, Anal. Chem. 44 (1972) 2227 – 2232. [5] R.J. O’Herron, EPA-670/4-75-005, April 1975, p. 7. [6] J. Estein, Anal. Chem. 19 (1947) 272. [7] Testing Methods for Industrial Waste Water, Japanese Industrial Standard JIS K 0102-1991, pp. 119 – 121. [8] E. Tanaka, S. Ohara, S. Adachi, M. Nunoura, J. Jpn. Water Works Assoc. 61 (1992) 21. [9] J.E. Girad, Anal. Chem. 51 (1979) 836 – 839. [10] Y. Liu, R.D. Rocklin, R.J. Joyce, M. Doyle, Anal. Chem. 62 (1990) 766. [11] O. Elshoz, W. Frenzel, C.-Y. Liu, J. Moller, Fresenius’ Z. Anal. Chem. 338 (1990) 159. [12] V. Kuban, Anal. Chem. 64 (1992) 1106. [13] Y. Asano and S. Ito, Japanese Patent, 1416967 (1987). [14] Y. Asano, S. Ito, Bunseki Kagaku 39 (1990) 785 – 787. [15] W.M. Latimer, Oxidation Potentials, 2nd edn, PrenticeHall, Englewood Cliffs, 1952, p. 191.