POTENTIOMETRIC STUDY FOR RAPID CONTINUOUS MONITORING OF TRACE LEVEL THIOCYANATE USING SOLID AND CONVENTIONAL TYPES PVC MEMBRANE SENSORS

A comparative study using two different thiocyanate electrode designs was conducted; a solid type electrode (type A) and conventional liquid inner contact electrode (type B). The fabrication of these electrodes was based on Al (III) [4-(2-Pyridylazo) resorcinol] (Al/PAR) and Mg (II) phthalocyanine (MgPC) complexes as a charged carrier, aliquate 336S and TDMAC as ion exchangers in plasticized poly (vinyl chloride) membrane. Electrodes type (A) revealed significantly enhanced response towards SCN ions with displayed near-Nernstian slope of -53.7 -55.8 mV decade over the concentration range 5.0x10-1.0x10 mol L and a detection limit of 0.4-3.7 μg mL. Type (B) sensors revealed near-Nernstian potential response to SCN with a slope of -45.9 -62.4 mV decade over a linear range of 5.0x101.0x10 mol L and a detection limit of 0.12-0.3 μg mL. Membrane sensors based on (Al/PAR) and (MgPC) using the so called "fixed interference method" (FIM) exhibited a good selectivity over different anions which differ significantly from the classical Hofmeister series. All sensors were integrated in a flow system for continuous monitoring of thiocyanate under hydrodynamic mode of operation. The sensors revealed a frequency of ~ 54 samples h. Application of the proposed sensors for SCN determination in biological fluid samples such as saliva collected from some non-smoker and smoker donors. Determination of cyanide content in electroplating wastewater samples after its conversion into thiocyanate was also applied. The results obtained from the proposed sensors were agreed with that obtained using the standard methods of thiocyanate and cyanide analysis.


INTRODUCTION
2][3][4][5] Ease of use, simplicity, ease of manipulation and no sample pre-treatment have replaced for long other wet analytical methods, because they offer high precision and rapidity, low cost, high selectivity and enhanced sensitivity over a wide linear range of concentrations.From all of the above, potentiometric sensors can be used as appropriate devices for process control.
Good analytical performances were noticed for conventional potentiometric sensors such as potential stability, short response time and high selectivity.On the other hand, low detection limit and lack of miniaturization were also noticed.7][8][9] This goal has been achieved in potentiometric sensors having a liquid internal contact by reducing zero-current transmembrane ion flux effects.For that, the primary ion activity in the internal filling solution is kept at a low level by using ion buffers 10,11 or ion-exchange resins. 12A carefully designed internal solution reduces trans-membrane concentration gradients through the partial exchange of primary ions by interfering ones on the inner membrane side. 13However, a significant drawbacks of these sensors are miniaturization and difficulty in use due to the presence of this internal filling solution.This requires avoiding leakage via alignment of the sensor in a vertical position, refilling the solution; moreover, there are problems with shape tailoring.
A new trial regarding solution-free sensors was proposed, based on inclusion of additional layer between the membrane and support, called solid contact ion selective electrodes (SC-ISEs). 14The first reported SC-ISEs were commonly referred to as coated-wire electrodes (CWEs). 15n which, a sensing polymeric membrane is directly coated onto an electronic conductor.These sensors have a high potential drift due to the "blocked" interface between the sensing membrane and the electronic conductor. 16These initial detections soonheless led to dense research to introduce materials that facilitate the charge transfer at the interface between the ion conducting membrane and the electronic conductor "ion-to-electron transduction".The most promising "ion-to-electron transducers" introduced to date include conducting polymers [17][18][19][20] and different carbon structures such as carbon nanotubes, 21,22 graphene, [23][24][25] and colloid-imprinted mesoporous carbon. 26r biomedical analysis, relevant contributors to the potential at the membrane of sample interface are thiocyanate itself and interfering ions.For most applications, a solvent polymeric membrane is used.Perm-selectivity is a necessary condition for thiocyanate and is achieved by incorporating cationic sites in the membrane, addition of an ion complexing agent or ionophore to the membrane.8][29][30][31][32] Although these sensors exhibit an anti-Hoffmeister selectivity order, they suffer from strong interference by I -, CN -, NO 2 -, IO 4 -, ClO 4 -, Br -, and N 3 -.In addition, some of these sensors suffer from the disadvantages of limited range of the linear response, and a long response time.
In this work, a comparative study concerning slopes, detection limits, concentration ranges and selectivity for a potentiometric determination of SCN -using ion selective membrane was made between two types of transducers: high double layer capacitance (Solid contact) e.g graphite/graphene sensors type (A) and classical inner filling solution (Liquid contact) sensors type (B).
Herein, we present a novel potentiometric thiocyanate membrane sensors based on the Al(III) [4-(2-Pyridylazo)  resorcinol] (Al/PAR) and Mg (II) phthalocyanine (MgPC) complexes as a charged carrier, aliquate 336S and TDMAC as ion exchangers in plasticized poly (vinyl chloride) membrane are described and developed.These sensors were used for thiocyanate determination in saliva and urine fluids, and for the determination of cyanide in waste-water and electroplating samples after conversion into thiocyanate.

EXPERIMENTAL
An Orion (Cambridge, MA, USA) Model 720/SA pH /mV meter was used for all potential measurements.Thiocyanate membrane sensor in conjunction with an Orion doublejunction Ag /AgCl reference electrode (Model 90-20) filled with 10% (w/v) KNO 3 were used for cell construction.For pH measurements, Ross glass electrode (Orion 81-02) was used.
Continuous monitoring of SCN -using flow system was carried out using a two-channel peristaltic pump (Ismatech Ms-REGLO) integrated with an Omnifit injection valve (Omnifit, Cambridge, UK) with a sample loop of 500 µL volume.The potential readings were contineously recorded through the interface ADC 16 (Pico Tech, UK) and Pico Log software (version 5.07).

Preparation of AL/PAR complex
Al/PAR ionophore (Figure 1) was prepared and structurally characterized using elemental and spectral analysis.To a solution containing 4-(2-pyridylazo) resorcinol monosodium salt monohydrate (PAR), (3.5 mmol) in 20 mL ethyl alcohol, K[Al(SO 4 ) 2 ].12H 2 O (1.5 mmol) was added.A solid precipitate was immediately produced, stirred at room temperature for 30 min, collected by filtration and washed with de-ionized water.The crude precipitate was re-crystallized from ethanol to give fine dark red crystals.

Sensor preparation and emf measurements
Liquid contact sensors were prepared as described previously. 34,35  2.0 mg portion of the ionophore was mixed with 126 mg of the plasticizer, 66 mg of PVC, 1.0 mg TDMAC and dissolved in 3 mL THF in a glass ring (2.2 cm diameter).The cocktail left to stand overnight for solvent evaporation at room temperature.The master membrane was sectioned with a cork borer (10 mm diameter) and glued using THF to PVC tubing (~3 cm length, 8 mm; internal diameter).The PVC tube was attached to a 1 mL micropipet tip.The blue tip was internally filled with 10 -2 mol L -1 thiocyanate solution.Homemade Internal reference electrode Ag/AgCl (1.0 mm diameter) was immersed in the internal reference solution.The conditioning solution was 10 -2 mol L -1 SCN -solution for 12 h and was stored in the same solution when not in use.
Solid contact sensors were fabricated as described previously 36 and as shown in Figure 2. A solid electrical contact is made from graphene which is characterized with high double layer capacitance.The epoxy metal mixture, used as supporting conductor, was obtained by pressing out an equal amount of both components onto the enclosed mixing tray.These two equal parts were mixed well with a synthetic spatula until a paste is obtained with a homogeneous color.This resulting paste was then mixed with graphite powder 1:1 mass ratio, introduced to the electrode body which is made from polypropylene and a copper disk with diameter 5.5 mm welded to a copper wire and immersed inside the paste.The electrodes were left to dry in air for 24 h.A volume of 10 µL of graphene solution (dissolved after sonication in THF) was applied onto the conductive layer.After dry, a volume of 20 µL of the membrane cocktail is applied at once by drop casting above the graphene layer.The membranes were left to dry at ambient temperature for 2 h before use.These solid contact electrodes were conditioned in a 1.0x10 -2 mol L -1 aqueous SCN -solution for 2 h before using and were stored dry when not in use.5.0x 10 -5 -1.0x 10 -2 5.0x10 -5 - 5.0x10 -6 -1.0x10 -2 Detection limit, mol L - 1 1.45x10 -6 2.68x10 -6 7.56x10 -6 3.88x10 -6 5.0x10 -6 2.9 x10 -6 5.3x10  The sensors were calibrated under a static mode of operation by transferring (0.5-1.0) mL aliquots of a 1.0 x 10 - 1 -1.0 x 10 -4 mol L -1 aqueous solution of SCN -to a 25 mL beaker containing 10.0 mL of 0.1mol L -1 acetate buffer of pH 5.0.The potential readings were recorded after stabilization to ±1 mV.

Determination of SCN -content in saliva
Different saliva samples were collected from cigarette smokers and non-smokers into polyethylene tubes.The samples were immediately centrifuged and stored at 4 o C. A 1 mL aliquot of the sample was transferred into 10 mL measuring flask and diluted with 0.1 mol L -1 acetate buffer (pH= 5.0).The proposed sensor in conjunction with the reference electrode were immersed in the solution, and the potential readings were recorded after stabilization.

Cyanide determination
An aliquot of 10 mL of standard 1.0 mol L -1 CN -test solution was boiled for 30 min using a hot plate with 1.0 g potassium polysulfide, pH adjusted at 10 by ammonia solution to prevent formation of volatile HCN.The solution was cooled, transferred into a 100 mL measuring flask and completed to the mark with 0.1 mol L -1 acetate buffer of pH 5.0.Further dilutions (1.0x10 -2 -1.0x10 -4 ) were made by the same buffer solution.The solutions were shaken well and (0.5-1.0) mL aliquots of a 1.0 x 10 -1 -1.0 x 10 -4 mol L -1 of converted CN -solution were transferred into a 25 mL beaker.The potential changes after each addition was recorded, and a calibration curve was constructed by plotting the potential reading against the concentration of cyanide ions.S1.The IR spectrum of the free ligand displayed strong bands at 3419 cm -1 , 1491.8 cm -1 , 1325.5 cm -1 and 1213.4cm - due to O-H, C=C, C-N and C-O.In the IR spectrum of Al/PAR, these bands are shifted to 3437.96 cm -1 , 1477.65 cm -1 , 1285.83cm -1 and 1192.18cm -1 , respectively.This confirms the complexation between Al 3+ ion and the PAR ligand.Mass spectroscopy analysis with molecular weight 536.0 confirmed the formation of 2:1 complex as shown in Figure S2.

Performance characteristics of thiocyanate sensors
Development of new potentiometric ion-sensors with better designs is a prospering area of research.An improvement in detection limit is more likely by introduction of solid inner contact instead of the traditional internal filling solution of the membrane.This may ideally give rise to a more reliable and robust sensors.
We have designed and characterized these new sensors, a conventional and solid contact types for thiocyanate monitoring and then compared their properties in light of these considerations.The optimum membrane composition was 33.0 wt % PVC, 1.0 wt % sensing material, 66.0 wt % of the plasticizer, and 0.5 wt % TDMAC as cation excluder in case of Al (III)-PAR complex and MgPC, whereas 33.0 wt % PVC, 1.0 wt % sensing material, 66.0 wt % of the plasticizer in case of aliquat and TDMAC ion exchangers.The response characteristics of thiocyanate sensors based on these ionophores were given in Table 1.Sensors based on Al(III)-PAR/TDMAC complex (sensor I) revealed a potentiometric response towards SCN -ions with near-Nernstian slope of -53.7±0.6 and -46.4±1.1 mV decade -1 (r 2 = 0.9998 and 0.9988) and a detection limit of 1.45x10 -6 and 5.0x10 -6 mol L -1 for type (A) and (B) sensors, respectively.For sensors based on MgPC/TDMAC (sensor II) exhibited a Nernstian potentiometric slope of -55.3±0.9 and -56.1±1.7 mV decade -1 (r 2 = 0.9999 and 0.9998) with a detection limit of 2.7x10 -6 and 2.9x10 -6 mol L -1 for type (A) and (B) sensors, respectively.For sensor III based on aliquat exhibited a potentiometric response with near-Nernstian slope of -54.7±1.5 and -55.1±0.8 mV decade -1 (r 2 = 0.9998 and 0.9996) and a detection limit of 7.56x10 -6 and 5.3x10 -6 mol L -1 for type (A) and (B) sensors, respectively.Sensors based on TDMAC only (sensor IV) exhibited a potentiometric response towards SCN -ions with near-Nernstian slope of -55.8±0.9 and -62.4±0.5 mV decade -1 (r 2 = 0.9999 and 0.9995) and a detection limit of 3.9x10 -6 and 2.0x10 -6 mol L -1 for type (A) and (B) respectively.Potentiometric responses of these sensors are shown in Figure 3.
The pH effect on the potentiometric response of the sensors was studied over the pH range of 2 to 11 at fixed concentrations of thiocyanate (1.0 x 10 -3 and 1.0 x 10 -2 mol L -1 ).The pH of these solutions was adjusted by adding NaOH and/or HCl.The solid contact sensors exhibited wider pH range 3.5-10, 4.5-10, 3.5-10 and 4-10 for sensor I, II, III, and IV, respectively.
For conventional sensors, they were pH independent in the pH range 4.0-7.0,2.5-8.0,4.0-10, 5.0-9.0 and 4.0-7.0 for sensor I, II, III, IV and V, respectively.The potentials of all sensors considerably declined with negative drift at higher pH values probably due to an increase of the OH - competition with the SCN -anion.The sensors were stored and conditioned in 10 -2 mol L -1 SCN -solution of pH 5.0.With all sensors examined, the detection limits, response times, linear ranges and calibration slopes were reproducible within ±3 % of their original values over a period of at least 12 weeks for both sensors types.One of the important features of the electrochemical sensor is the response time which can be evaluated by measuring the time required to achieve a steady-state potential with in ±0.2 mV of the final equilibrium value after successive immersion in a series of SCN -solution, each having a 10fold difference from low to high concentrations.The time required to achieve a steady potential response (±3 mV) using the proposed sensors in 10 -6 to 10 -4 mol L -1 SCN - solutions with a rapid 10-fold increase in concentration was < 20 s and < 10 s for sensors type (B) and type (A), respectively.Replicate calibrations for each sensor indicated low potential drift, long-term stability and negligible change in the response of the sensors.

Effect of foreign ions on thiocyanate sensors
Selectivity of the sensors was assessed by means of potentiometric selectivity coefficients (K POT SCN,J ), calculated by the fixed interference method (FIM). 37The potential measurements determined by using a solution contain a fixed activity of the interfering ion (10 -3 mol L -1 ) on acetate buffer pH 5.0 and varying the activity of the primary ion.The logarithmic values of log K POT SCN,J were indicated in Table 2.
The selectivity order for type (A) sensors were illustrated as follow: For sensor I, the selectivity order as follow sensor

Hydrodynamic monitoring of thiocyanate
Flow-through analysis using potentiometric detectors showed high advantages such as low cost, simple instrumentation and automation.In addition, the transient nature of the signal in flow injection analysis is help to overcome the effects of interfering ions if the electrode's response to the target analyte is faster than those interfering ions. 38A tubular-type detector incorporating Al (III)-PAR/TDMAC (sensor I), MgPC/TDMAC (sensor II) and Aliquat sensor (III) based membrane sensors were prepared and used under the hydrodynamic mode of operation for continuous SCN -quantification.A triplicate transient peaks were obtained from the flow injection analysis system under optimal experimental conditions and are shown in Figures 4A, 4B and 4C.A linear relationship between the SCN -concentrations and FIA signals was obtained over a concentration range of 1.0 x 10 -5 to 1.0 x 10 -1 mol L -1 using a 0.1 mol L -1 acetate buffer, pH 5.0.The general analytical features recorded under optimum flow conditions are presented in Table 3.The slope of the calibration plot was near-Nernstian (-46.5, -52.0 and -53.3 mV decade -1 ), there lower limit of detection was 1.0 x 10 -5 mol L -1 , and the sampling frequencies were about 44, 54 and 44 samples per hour respectively.

Thiocyanate determination in saliva
In the saliva of cigarette smokers, rhodanaze enzyme catalyzes the detoxification reaction of hydrogen cyanide forming thiocyanate.The possibility of using the proposed sensors to measure SCN -concentration in human saliva samples was examined.Table 4 indicates that the concentrations of saliva-SCN -in the groups under study fairly agreed with data obtained by the standard spectrophotometric method. 39e F-test showed no significant difference at the 95 % confidence level between the mean and the variance of the potentiometrric and spectrophotometric set of results.The calculated F-values (n = 5) were found to be in the range 1.1-5.2compared with the tabulate value (6.39) at the 95 % confidence limit. 1

Cyanide determination in electroplating baths
Determination of cyanide ion after its conversion to thiocyanate ion using the potassium polysulfide treatment was investigated.Such a conversion is well-documented, 40 and occurs according to the following reaction: K2Sx + CN -→SCN -+ K2Sx-1 A linear calibration curve for CN -ions (1.0 x 10 -5 -1.0x10 -2 mol L -1 ) after conversion into SCN -by a treatment with potassium polysulfide, followed by monitoring using the proposed sensors were obtained.The method was used for the potentiometric determination of free cyanide ions in industrial waste-water and electroplating samples obtained from an electroplating factory.These sensors are used under the hydrodynamic mode of operation for continuous monitoring of CN − ions (1.0 x 10 -4 -1.0 x 10 -1 mol L -1 ) after conversion into SCN − using a 0.1mol L -1 acetate buffer, pH 5.0 as a flow carrier.The obtained results agreed with data obtained using the solid state cyanide ion-selective electrode Table 5. 189 CONCLUSION Two Types of potentiometric ISE (Solid type (A) and conventional type (B)) were studied to give a fast and simple method for determination of thiocyanate under static (manual) mode of operation and hydrodynamic (FIA) mode of operation for type (B).A new Al (III)-PAR complex was prepared and developed by using TDMAC as anionic additive.ISE type (A) improved the response of sensor I towards SCN -1 ion with lower detection limit 1.45x10 -6 mol L -1 and slope of -53.7 mV decade -1 while that of type (B) exhibit detection limit 5.0x10 -6 mol L -1 and a Nernstian slope of -46.4 ± 1.1 mV decade -1 .So ISE type (A) can be preferred to use as a tool for low SCN -concentration determination in biological fluids (saliva).ISE type (B) used for determination of CN -ion in waste and electroplating.

Figure 2 .
Figure 2. Fabrication of a SC-ISE high double layer capacitance based on epoxy-graphite composite.

Figure 3 .
Figure 3. Calibration plot of thiocyanate sensors type (A) and type (B) I, II, III and IV; in 0. 1 mol L -1 acetate buffer pH 5.0

Table 2 .
Effect of foreign ions on SCN − membrane sensors type (A) and (B) by using FIM.

Table 4 .
Determination of thiocyanate in saliva samples using type (A) sensors I, II, III and IV.

Table 3 .
Potentiometric response characteristics of SCN - membrane type (B) sensors I, II and III using FI operation.

Table 5 .
Potentiometric determination of cyanide in electroplating baths after conversion into thiocyanate using a thiocyanate PVC membrane type (B) sensor I, II and III.