This paper examines potentiometric multisensory systems that consist of novel cross-sensitive PD-sensors (Potential Donnan-sensors). The analytical signal of PD-sensors is the Donnan potential at the ion-exchange polymer/electrolyte test solution interface. The use of novel sensors for the quantitative analysis of multicomponent aqueous solutions of amino acids, vitamins and medical substances is based on protolytic and ion-exchange reactions at the interfaces of ion-exchangers and test solutions. The potentiometric sensor arrays consist of PD-sensors and ion-selective electrodes. Such systems were developed for the multicomponent quantitative analysis of lysine monohydrochloride, thiamine chloride and novocaine hydrochloride solutions that contained salts of alkaline and alkaline-earth metals, as well as for mixed solutions of nicotinic acid and pyridoxine hydrochloride. Multivariate methods of analysis were used for sensor calibration and the analysis of the total response of sensor arrays. The errors of measurement of the electrolytes in aqueous solutions did not exceed 10%. The developed multisensory systems were used to determine the composition of a therapeutic “Mineral salt with low content of sodium chloride” and to determine concentrations of novocaine in sewage samples from a dental clinic.
UV spectrophotometry [
The advantages of potentiometric methods include the possibility of rapid,
Modern investigations in the field of potentiometric sensors have taken the following approaches: the search for novel materials useful for the construction of ISEs [
A multisensitive system includes an array of cross-sensitive sensors (i.e., sensors that are sensitive to several components in a given solution) and algorithms for processing the multidimensional data from a sensor array [
Recently, we described the development of a novel potentiometric sensor (PD sensor), which measures the Donnan potential at an ion-exchange polymer (IEP)/electrolyte test solution interface [
We have previously described the application of the PD sensor for the selective determination of lysine in the presence of neutral amino acids, ammonium ions [
The aim of this paper was to develop potentiometric multisensory systems with novel polymer-based ion-exchange PD sensors for the determination of inorganic ions and various ionic forms of organic electrolytes in multicomponent aqueous solutions. Specifically, the analytes of interest were amino acids, vitamins, and medical substances.
All chemicals were of analytical reagent grade. All solutions were prepared using distilled water with a resistance of 0.35 MΩ·cm. The following analytes of interest were dissolved in aqueous solutions: lysine monohydrochloride (LysHCl), thiamine chloride (ThiaminCl), pyridoxine hydrochloride (PyridoxinHCl), nicotinic acid (Niacin), novocaine chloride (NovHCl), and inorganic electrolytes (NaCl, KCl, CaCl2, and MgSO4). Concentrations of the various solution components ranged from
Systems containing sulphocation-exchange polymers with different structures (i.e., homogeneous perfluorinated sulphocation-exchange MF-4SK membranes and tubes, which are Russian analogues of Nafionc, and heterogeneous hydrocarbonic MC-40 membranes) and individual solutions of inorganic electrolytes (i.e., HCl, NaCl, and KCl) were previously researched for the selection of ion-exchange materials.
The structure of perfluorinated sulphocation-exchange polymers (PSPs) is formed from a system of nanopipes (10–17/5–10/3–5 nm) and pores (0.75–1.25 nm) with hydrophobic walls and hydrophilic sulphonate ionic groups within the channel volumes. The structural units of the hydrocarbonic polymers are represented by macroclusters of micropores with radii of 2-3 nm. The structural units of the hydrocarbonic polymers include ion-exchange groups and hydrophilic regions in a matrix that is divided by meso- and macropores with 5 ÷ 500 nm radii. The meso- and macropores are filled with test solution and include polymeric chains and an inert material [
PSPs are characterised by optimal selective properties as a result of fewer numbers of mesopores and the complete absence of macropores. Thus, the presence of hydrophobic (i.e., polytetrafluoroethylene chains) and hydrophilic (i.e., sulphonate ionic groups) regions in such polymers provides the matrix with labile structural components, which allows for the electrochemical properties of PSPs to be controlled by changing the ionic form of the polymers. Therefore, the hydrophobicity of the matrix from PSP and the absence of macropores together define a greater interface transition activation energy of hydrated ions compared to heterogeneous hydrocarbonic polymers. Hence, the use of PSPs in PD sensors provides increased signal, sensitivity, and accuracy in comparison with hydrocarbonic polymers. The comparison of the sensitivity of PSP-based PD sensors and MK-40 membranes for a number of inorganic ions is shown in Figure
Sensitivity of PD sensors based on PSP and MC-40 in corresponded ionic types in test solutions of HCl, NaCl, KCl, and CaCl2.
All solutions were analysed at
A scheme of an electrochemical cell for analysing multicomponent aqueous solutions of organic and inorganic electrolytes [
The scheme of the electrochemical cell for the determination of organic electrolytes in multicomponent test solutions:
The sensor array included PD sensors (
The PD sensor [
The electrochemical circuit (
The organisation of the PD sensor is as follows. The sum of all potential jumps in the EMF (
The stability, sensitivity, and selectivity of sensors were estimated in individual solutions of analytes. The determination of activity coefficients in the polyionic systems is a difficult scientific problem. Therefore, calibration of the sensors was performed in the
Multivariate calibration methods were used to deduce the calibration equations for the calculation of analyte concentrations in mixed test solutions. The values of factors of concentrations were changed with a constant step. Initial factors were encrypted so that the sum of the values of any two factors from all experiments was equal to zero, which satisfied the requirements of nondegeneracy and orthogonality for the experimental plan. The absence of systematic errors and insignificant distinction of variances was a necessary requirement for the responses of sensors. Statistical models that did not take into account (
The coefficients of multivariate calibration equations were determined by the method of least squares [
The PD sensors were organised so that quasi-equilibria, which are formed at PSP/test solution and PSP/reference solution interfaces, are stable as a function of time and are independent from each other. The ionic properties and concentrations in the solution phase volumes and the ion-exchanger were slightly changed [
In systems with inorganic electrolytes, ion-exchange reactions are potentially defining. In such reactions, a hydrated shell of ions partially breaks up in solution phase and then reorganises in the PSP phase. Therefore, the sensitivity of the PD sensor to solution phase inorganic ion concentrations increases with decreasing charge, crystallographic radius, and increase in degree of hydration (Figure
Various functional groups (–NH2, –COOH) are present in the structure of amino acids, vitamins, and medical substances. These functional groups are capable of participating in ion-exchange and proteolytic reactions in the solution phase, PSP, and at the interface. The quasi-equilibria at the PSP/multicomponent test solution interface for K+- and H+-type PSPs are shown in Figures
The quasiequilibria at the interface of PSP in K-type (a) and H+-type (b) with test and reference solutions in PD sensor: X+ is lysine, thiamin, and novocaine cations, Mez+ is cations of alkaline, and alkaline-earth metals.
The potential defining reactions of PSP-based K+-type PD sensors are ion-exchange reactions. The PSP-based K+-type PD sensor was characterised by high sensitivity to all organic (
Sensitivity of PSP-based K-type (a) and H+-type (b) PD sensors in test solutions of LysHCl, ThiaminCl, NovHCl, PyridoxinHCl, Niacin, NaCl, KCl, and MgSO4.
In PSP-based H+-type PD sensors, hydronium ions contributed to the Donnan potential. The hydronium ions competed with large organic cations during formation of the Donnan potential at the PSP/test solution interface because organic cations have hydrophilic and hydrophobic groups. Additionally, the singly charged organic cations LysH+, ThiaminH+, and NovH+ did not participate in ion-exchange reactions. These cations transformed into doubly-charged ions in the polymer phase as a result of a heterogeneous proteolytic reaction. Thus, adsorption at the interface of large doubly charged ions resulted in the decreased contribution of an organic component to the Donnan potential. As a result, the stability of the analytical signal of the PD sensor in such systems decreased. In some cases, the sensitivity to an organic component also decreased (e.g., by 1.2- and 1.5-times to LysH+ and NovH+, resp.). However, the sensitivity of PSP-based H+-type PD sensors to inorganic ions changed only slightly in comparison with PSP in salt forms (Figure
Thus, the use of PSPs in various ionic forms led to different PD sensor sensitivities toward the same organic component. However, the sensitivity of PD sensors toward inorganic ions changed insignificantly. The various influences of ionic forms of PSPs on the sensitivity to vitamins PyridoxinHCl and Niacin permitted the use of PD sensors for their joint determination in mixed aqueous solutions.
The potentiometric multisensory systems were developed for a multicomponent quantitative analysis of lysine monohydrochloride, thiamine chloride, and novocaine hydrochloride solutions that also contained chlorides of potassium and sodium. Additionally, solutions of nicotinic acid and pyridoxine hydrochloride were also analysed.
The sensor array for the analysis of solutions LysHCl + KCl + NaCl, ThiaminCl + KCl + NaCl, and NovHCl + KCl + NaCl included the PSP-based K+-type PD sensor (A1), K-SE (B1), Na-SE (B2), and silver chloride/silver reference electrode (C). The electrochemical circuits for the determination of the responses of the sensor array are described by
The sensor array for the analysis of solutions LysHCl + KCl + NaCl + MgSO4 included Mg(Ca)-SE (B3) and sensors A1, B1, B2, and C. The electrochemical circuits for the determination of the response of the sensor array are described by (
According to reference [
The factors for the estimation of the cross-sensitivity of sensors Ak, Bk.
Determining components | LysHCl, KCl, NaCl, MgSO4 | ThiaminCl, KCl, NaCl | NovHCl, KCl, NaCl | Niacin, PyridoxinHCl | ||||||||
Sensor | A1 | B1 | B2 | B3 | A1 | B1 | B2 | A1 | B1 | B2 | A1 | A2 |
40 | 27 | 29 | 13 | 41 | 28 | 44 | 46 | 34 | 42 | 29 | 33 | |
9 | 5 | 6 | 3 | 5 | 7.5 | 11 | 15 | 4 | 2 | 8 | 16 | |
1.4 | 0.08 | 0.05 | 0.09 | 2 | 0.1 | 0.2 | 2 | 0.1 | 0.2 | 0.3 | 17 |
Large sensitivity and nonselectivity factors for the PSP-based K+-type PD sensor in the groups of ions (Table
ISEs were not as highly selective in test solutions. For example, Mg(Ca)-SE had an average angle lower than 25 mV/pC (Table
The sensor array for the analysis of PyridoxinHCl + Niacin solutions included two PSP-based K+-(A1) and H+-type (A2) PD sensors and a reference electrode (C). The electrochemical circuit for the determination of PD sensor responses is described by (
Shown in Table
The mixed test solutions LysHCl + KCl + NaCl, ThiaminCl + KCl + NaCl, NovHCl + KCl + NaCl, and PyridoxinHCl + Niacin were studied for the multivariate calibration of the sensor array. All possible combinations of factors pC were examined for each analyte in the range of 2–4 with a constant step of pC = 1. The coefficient estimates from the multivariate calibration equations without taking into account any interference of components to the responses of sensors
The coefficient estimates from the multivariate calibration equations without taking into account any interference of components to the responses of sensors
Test solution | LysHCl (pC1) + KCl (pC2) + | ThiaminCl (pC1) + KCl | NovHCl (pC1) + KCl | Niacin (pC1) + | ||||||||
NaCl (pC3) + MgSO4 (pC4) | (pC2) + NaCl (pC3) | (pC2) + NaCl (pC3) | PyridoxinHCl (pC2) | |||||||||
Sensor | A1 | B1 | B2 | B3 | A1 | B1 | B2 | A1 | B1 | B2 | A1 | A2 |
— | — | — | ||||||||||
— | ||||||||||||
— | — | — | — | — | — | |||||||
— | — | — | — | — | — | — | — | — | — | |||
3 | 3 | 9 | 0.3 | 10 | 6 | 7 | 10 | 2 | 15 | 4 | 1.4 |
The coefficient estimates of the multivariate calibration equations, taking into account the interference of components on the responses of sensors
The coefficient estimates from the multivariate calibration equations, taking into account interference of components to the responses of sensors
Test solution | LysHCl (pC1) + KCl (pC2) + | ThiaminCl (pC1) + KCl | NovHCl (pC1) + KCl | Niacin (pC1) + | ||||||||
NaCl (pC3) + MgSO4 (pC4) | (pC2) + NaCl (pC3) | (pC2) + NaCl (pC3) | PyridoxinHCl (pC2) | |||||||||
Sensor | A1 | B1 | B2 | B3 | A1 | B1 | B2 | A1 | B1 | B2 | A1 | A2 |
— | — | — | ||||||||||
— | ||||||||||||
— | — | — | — | — | — | — | ||||||
— | — | — | — | — | — | — | — | — | — | |||
— | — | — | — | — | — | |||||||
— | — | — | — | — | — | — | ||||||
— | — | — | — | — | — | — | — | — | — | — | ||
— | — | – | — | — | — | |||||||
— | — | — | — | — | — | — | — | — | — | — | ||
— | — | — | — | — | — | — | — | — | — | — | ||
2 | 3 | 7 | 0.2 | 6 | 6 | 5 | 5 | 2 | 4 | 3 | 0.6 |
The equations were adequate at a confidence level of 0.05. The statistical models that took into account the interactions of factors reduced the errors of PD sensors and the errors of Na-SE by 1.3–2.3-times and 1.4–3.2-times, respectively, compared to statistical models that did not take into account the interactions of factors. The errors for K-SE and Mg(Ca)-SE did not change. Therefore, to calculate analyte concentrations, we used (
Shown in Tables
The actual and measured values of the analyte concentrations for some test solutions ThiaminCl + KCl + NaCl, NovHCl + KCl + NaCl.
Test solution | Added, M | Found, M | ||||
X+ | K+ | Na+ | X+ | K+ | Na+ | |
Thiamin+, ThiaminH2+(X) K+ Na+ Cl− | ||||||
NovH+ (X) K+ Na+ Cl− | ||||||
The actual and measured values of the analyte concentrations for some test solutions PyridoxinHCl + Niacin.
Test solution | Added, M | Found, M | ||
PyridoxinH+ | Niacin | PyridoxinH+ | Niacin | |
PyridoxinH+ | ||||
Niacin+ | ||||
Niacin± Cl− |
The number of replicate measurements was 6–8. The statistical data interpretation was made using a confidence coefficient of 0.95. The relative error of measurement was 2–10%.
The potentiometric multisensory system used to analyse LysHCl + KCl + NaCl + MgSO4 solutions was used for the analysis of therapeutic “mineral salt with low content of sodium chloride” samples. This product contained NaCl, KCl, MgSO4, and LysHCl in the following mass rations (%): 0.35–0.58; 0.31–0.40; 0.05–0.10 and 0.02–0.10 [
The potentiometric multisensory system used to analyse NovHCl + KCl + NaCl solutions was used to analyse sewage samples from a dental clinic. The responses of sensors A1, B1, and B2 were measured against reference electrode C in sewage samples. These samples were taken from a sewer knee before and after a patient’s reception. The disparity of NovH+ concentrations in the dental clinic sewage before and after the patient’s reception was (
Here, the development of potentiometric multisensory systems, in which novel potentiometric PD sensors were cross-sensitive was described. The analytical signal of PD sensors is the Donnan potential at the individual PSP/test electrolyte solution interface. The use of the membrane potential equilibrium component as an analytical signal, which is the Donnan potential at the IEP/test solution interface, resulted in the elimination of migration and diffusion problems inherent in potentiometric sensors. As a result, the accuracy and stability of the analysis subsequently increased. The use of novel sensors for the quantitative analysis of multicomponent aqueous solutions of amino acids, vitamins, and medical substances was based on proteolytic and ion-exchange reactions at the PSP/test solution interface. The potentiometric sensor arrays were developed for multicomponent quantitative analyses of lysine monohydrochloride, thiamine chloride, and novocaine hydrochloride solutions containing salts of alkaline and alkaline-earth metals as well as mixed solutions of nicotinic acid and pyridoxine hydrochloride. The sensor arrays consisted of cross-sensitive PD sensors and ISEs. The multivariate methods of the analysis were used for sensor calibrations in addition to the analysis of the total response of the sensor arrays. The relative errors of electrolyte measurements in aqueous solutions did not exceed 10%.
The developed multisensory systems were then used to determine the composition of therapeutic “mineral salt with low content of sodium chloride” and dental clinic sewage.
The PD sensor multisensory systems made it possible to perform quantitative analyses of multicomponent solutions of different electrolytes, which is in contrast to the majority of known potentiometric sensors arrays that only permit semiquantitative analyses.
The authors thank cand.ch.sc., the chief of the laboratory of Membrane Processes OSS “Plastpolymer” (St. Petersburg, Russia) Timofeev Sergey Vasilevich for giving samples of perfluorinated sulphonic cation-exchange membranes, tubes, and rods. This study was funded by the Russian Fund of Fundamental Research (09-03-97505 r_center_a).