Application of ion pair complexes to design novel potentiometric membrane sensors for direct determination of frusemide in pharmaceuticals

Fabrication of two membrane sensors using two acidic indicators among sulphonthalein dyes, namely bromophenol blue (BPB) and bromocresol green (BCG), and their use as indicative electrodes for the quantification of frusemide (FUR) is presented. The ion pair complexes of FUR with BPB or BCG are used to prepare the membranes in THF solvent, PVC matrix and dibutyl phthalate (DBP) as plasticizer and subsequently to fabricate FUR-BPB (Sensor I) and FUR-BCG (Sensor II) sensors. Sensors I and II are employable to determine 2.4 × 10-5–2.4 × 10-3 mol/L FUR at operative pH of 3.71. The calibration curve between the potentials against the concentration of FUR yielded the slopes of 58.73 ± 1 and 57.66 ± 1 mV/decade, respectively, using Sensors I and II, and this confirmed the Nernstian behaviour. Satisfactory correlation was obtained between the measured potentials and FUR concentration with the proposed sensors, and this was revealed by regression coefficient values of 0.9987 and 0.9980 for Sensors I and II, respectively. The LOD (limit of detection) values were calculated and reported for both the sensors. The experimental parameters were optimised to yield acceptable characteristics of both the sensors in the context of performance. The role of excipients of tablets and interferences were assessed by standard addition protocol. The obtained results confirmed the ineffective role of excipients of tablets and foreign species used as interferents. The designed sensors were validated to confirm the accurate, precise, robust and rugged functioning in determining FUR. The mean of recovered FUR, close to 100%, revealed the acceptable and effective functioning of the proposed sensors. Excellent results were obtained by FUR tablets’ analysis using both the sensors.

A couple of reports were found in the literature on potentiometric sensors used to quantify FUR. Santini et al. [34] described the construction and analytical application of a potentiometric sensor using platinum, mercury, mercuric nitrate, and graphite to estimate FUR. The article devoted by Dias et al. [35] describes the construction and use of polymeric electrode by incorporating the FUR-tricaprylmethylammonium chloride (FUR-aliquat 336S) ion pair complex into a membrane for assaying FUR. The reported potentiometric sensors are not selective to determine FUR in pharmaceuticals because of their incomplete validation and scanty competing analytical results. Therefore, in this paper, the preparation, validation, and application of two new membrane sensors are presented for FUR determination in pharmaceuticals.

Apparatus
A PICO model digital potentiometer (Chennai-32, India) was used in all potential measurements. An Elico (Mumbai, India) pH metre was used in pH recordings. The electrochemical cell utilised Ag/AgCl as a reference electrode and a copper wire as a conducting device.
High-purity distilled water and analytical grade chemicals were used. The reagents were prepared using analytical grade substances. Jubilant Life Sciences Ltd, Nanjangud, Mysore, India gifted the pure FUR (99.8%), and the sample was used as received. Lasix tablets (40 mg FUR/tablet) were obtained from the local market. Bromophenol blue dye (BPB), bromocresol green (BCG), tetrahydrofuran (THF), chloroform (CHCl 3 ), dibutyl phthalate (DBP), and polyvinyl chloride (PVC) were supplied from SD Fine Chem Ltd., Mumbai, India. Glacial acetic acid (HOAc) was purchased from Merck (Mumbai, India) and used to prepare a 2:3 (v/v) solution in water.
A stock of 3 mmol/L FUR solution was prepared in 2:3 HOAc using a calibration flask.

Designing of sensors
Accurately weighed 10.1 mg of FUR was dissolved in 25 mL 2:3 HOAc. To the same solution, 20.04 mg of BPB or 20.6 mg of BCG dye was added; the content was transferred into a separatory funnel and extracted thrice with 10 mL portions of chloroform. The organic layer was passed through anhydrous sodium sulphate and collected into a clean and dry beaker. After evaporating the solvent, the resulted residue in the form of ion pair was used. About 20.0 mg of FUR-BPB or FUR-BCG ion pair was dissolved in 10 mL THF. The solution took in a Petri dish (width 4 cm), 0.1 g each of DBP and PVC were added; it was dissolved, and the content was dried under laboratory temperature for 24 h. The obtained dry membrane was composed with 9% (w/w) of respective ion pair and 45.5% (w/w) each of DBP and PVC. The resultant sufficiently dried membrane was attached to one end of a plastic tube using THF and dried at room temperature for 24 h. A 3-5 mL internal solution of 0.005 mol/L FUR was filled, the lower terminal of copper wire (15-cm length and 2.0-mm diameter) coated with insulator (leaving 1.0 cm at the top end and 0.5 cm at the bottom not insulated) was immersed, and the upper terminal was taken into connection to a negative terminal of the potentiometer. At this moment, the resulted FUR-BPB (Sensor I) and FUR-BCG (Sensor II) sensors were soaked into FUR solution and were used for measuring potentials after 5 h. The calibration curves were constructed using the solutions' potentials data versus log [FUR] were prepared. The FUR concentrations in unknowns were calculated using calibration curves or regression data derived using potential versus log [FUR].

Study of interference
A 2-ml each of 3 mmol/L FUR and 0.1% interferent solutions were taken in an array of 10-mL calibrated flasks. The solutions were made to the mark with 2:3 HOAc and the potentials of each measured using FUR-BPB (Sensor I) or FUR-BCG (Sensor II) indicative sensors in reference to Ag/AgCl standard electrode.

Analysis of tablets
A group of twenty pre-weighed tablets were ground into a fine powder. The homogeneous powder corresponding to 50 mg FUR was measured and taken in a 50-ml calibrated flask, added 20 ml HOAc and shaken for 20 min and the volume raised to calibration mark with water. The content after mixing was filtered with Whattman No. 41 paper. A suitably measured volume of the resultant solution was drawn out, and potential was measured as given under the calibration curve construction. The FUR concentration in the taken extract was computed.

Optimisation of parameters Effect of dye content
The quantity of BPB or BCG was varied in the preparation of ionic aggregate with FUR. Sensors prepared with differential amounts of dyes were employed to measure the potential, and consequently, slopes of the linear curves were found out. Calibration lines' slopes were close to 59.16 mV/decade when the quantities of BPB and BCG added were 20.04 and 20.6 mg in Sensors I and II, respectively. At other quantities of dyes, varying slopes were observed. The resultant slopes from differential amounts of dyes are summarised in Table 1.

Response time and lifetime of the electrodes
The time required to achieve a steady-state potential (within ± 0.2 mV/min), for 2.4 × 10 -5 -2.4 × 10 -3 mol/L FUR solutions at pH 3.71. The potentials were reached steady state in 8 s for FUR solutions at concentrations above 1.0 × 10 -3 mol/L and about 20 s at concentrations near the limit of detection. The experimental results proved 60 days of lifetime with a total of 500 determinations using these FUR-BPB and FUR-BCG sensors. During this period, each sensor was in daily use and was stored in 0.010 mol/L FUR solution and pH adjusted at 3.71 when not in use. There were no significant changes in working concentration range, slope, and response time in this period.

Effect of plasticizer
The sensing membranes were prepared separately by adding different amounts of dibutyl sebacate (DBS), dibutyl phthalate (DBP), dioctyl phthalate (DOP), and onitrophenyl octylether (NPOE) as plasticizers. The membranes were prepared using 0.1 g of DBP for both; the sensors were found to behave in Nernstian response. The sensors were found to perform satisfactorily with respect to stable potential readings, ease of conditioning, and less response time. Therefore, DBP was used as plasticizer in fabricating the FUR-BPB and FUR-BCG membrane sensors.

pH effect
The pH effect on electrode potential using electrodes over the range of 2.0-8.0 for 2.4 × 10 -5 -2.4 × 10 -3 mol/ L FUR solutions was investigated. The resulting solutions' pHs were adjusted with either 2 M HNO 3 or dilute NaOAc solutions. As it can be seen in Fig. 2, the potential response was found maximum and acceptable at pH 3.71 for both the sensors. However, at pHs which differ from 3.71, significant change in either the solution's clarity or potentials was observed. Therefore, the pH of 3.71 was maintained as optimum in all potential measurements.

Evaluation of selectivity coefficients
The chemical interaction between the active sites of the membrane surface and the analyte is responsible for generation of membrane potential (E m ). The poor selectivity of membranes towards a single species is found due to the factor of chemical process on the signal. Practically, E mem is in proportion to the concentration ions capable of interacting at the membrane's active sites. The generalised Nernst equation to include the contribution of an interferent, I, can be written as follows [36].
where K is a constant and includes reference electrode's potential, liquid junction potential, and E m ; Z A and Z I are the charges on the analyte and interferent, respectively, with concentrations [A] and [I]. The parameter K A,I is known as selectivity coefficient and is defined as where [A] E and [I] E are the analyte's and interferent's concentrations yielding identical cell potentials, i.e. E. When the K A,I is equal to unity, the membrane responds for both analyte and interferent. A selective membrane must always possess K A,I value of less than unity [36].
Thus, as an important feature, the value of K A,I can be evaluated using the data of potentials of solutions   The analyte's and interferent's concentrations at the intersection of the two linear regions are used to calculate K A,I . Thus, K A,I values were calculated in the presence of different interferents and are given below in Table 2. Both the sensors were used to measure the potentials of FUR along with various inorganic and organic compounds, anions, and cations. A 0.1% solution of each interferent was spiked into pre-analyzed FUR solution, and analysis was done [37,38]. The summarised results indicated non-interference from the added species as showed by the calculated K A,I values in Table 2. Therefore, the developed sensors are the most suitable electrodes for assay of FUR in the presence of such studied interferents, too.

Validation of sensors
The developed FUR-BPB and FUR-BCG sensors were evaluated for linearity by following the IUPAC [37,38] recommendations. The sensors were validated for accuracy, precision, robustness, and ruggedness by following the ICH Guidelines [39]. The validation results for individual parameters are presented in the following sections.

Linearity and analytical and regression parameters
By following the IUPAC recommendations [37,38], the linearity of the proposed analytical methods using FUR-BPB and FUR-BCG ISEs was evaluated. The results reflected that the ISEs respond rapidly and linearly for FUR concentration ranges presented in Table 3. The calibration graphs were linear (Fig.  3), and the regression equations were y = 58.73x + 322.8 and y = 57.65x + 336.89 for FUR-BPB and FUR-BCG ISEs, and the corresponding Nernstian slopes were 58.73 and 57.66 mV/decade. The regression coefficients, slopes, intercepts of the calibration lines and other relevant performance characteristics of both analytical methods using proposed sensors were calculated as described in the literature [37,38,40] and are mentioned in Table 3. The detection limits (LOD) were also calculated [37,38] and reported in Table 3.  [41] and USP [42] recommendations and made their applicability for use as quantification sensors for FUR.

Robustness and ruggedness
The robust functioning of proposed ISEs was assessed by varying working pH and temperatures slightly. FUR solutions of 0.9, 1.5, and 2.1 mmol/L were used for analysis. The pH was varied by 0.02 [3.71 (± 0.02)] units for each sensor, and the calculated values of %RSD were from 1.78 to 3.33. Besides, the robustness was also been evaluated by varying the temperature by 1°C during the measurement of potentials of FUR solutions of different concentrations. The temperatures of the FUR solutions were brought to 23 ± 1 and 37 ± 1°C using both the sensors; potentials were measured; and %RSD values were reported. Range of %RSD from 2.68 to 3.52 indicated the robust functioning of proposed sensors. Analyses were done using three different potentiometers for 5 days by the same and different analysts to assess the ruggedness. The variations expressed in %RSD values for inter-potentiometric and inter-analysts were < 5%, reflecting sensors' rugged function. Summary of results of this study are presented in Table 5.

Application of ISEs to tablets analysis
Five replicates of each of 0.9, 1.5 and 2.1 mmol/L in FUR from tablet extracts were analyzed with proposed Sensor I and Sensor II by using the described procedure for tablet analysis. The mean of the measured potentials obtained and the percent recoveries of FUR were evaluated. The results of the United States Pharmacopeia (USP) assay method [2] were compared with the obtained results using the proposed sensors. The procedure in reference method describes titration of FUR (about 600 mg) in dimethylformamide against standardised 0.1 M NaOH using bromothymol blue indicator. The Student's t and and F values were intently obtained   Table 6. The t cal and F cal values are less than the values which are statistically tabulated. Therefore, the proposed sensors yielded accurate and precise results.

Recovery study
The accurate functioning of proposed ISEs was further assessed by recovery studies following the standard addition method. The aliquots of pre-analyzed tablet extracts were spiked with pure FUR solution in three different concentration levels. To a fixed amount of five replicates of FUR tablet extract, pure FUR in amounts corresponding to 50, 100, and 150% to that of the amount from tablets were spiked; pH was manipulated to optimum value, and content was diluted to 10 mL and potentials measured. Then, FUR concentrations were calculated using the regression data. The %FUR recovered was calculated for each level. The mean %FUR recovered was 98.33 with standard deviation of less than 3% revealed that good and acceptable recovery is possible with proposed sensors. The relevant results of recovery experiments are summarised below (Table 7).

Discussion
The development of ISEs using the ion pair complexes of therapeutic compounds and sulphonthalein dyes through their application to quantify FUR is described for the first time. The action of two sulphonthalein dyes viz. BPB and BCG is en-cashed to construct the membrane sensors or ISEs for FUR. Chemically, the nitrogenous group of FUR will possibly be utilised to react with anionic dye, as chromogenic reagent. In CHCl 3 , the resulting yellow product of FUR-BPB or FUR-BCG is due to the transformation of the lactoid ring structure to quinoid form [43] in the anionic dyes, BPB and BCG. The two tautomers (lactoid and quinoid) are supposed to be present in equilibrium. However, due to the acidic property of the sulphonic acid group, the latter (quinoid) form is assumed predominant. Finally, protonated FUR (FURH + ) forms ion-pair with the anionic dye (BPBor BCG -). These possible reaction mechanisms are presented in Scheme 1. Anionic dye such as BPB or BCG forms an ionic aggregate or ion pair complex with a positively charged drug, and the product formed between the drug and dye is an electrically neutral ion pair complex molecule. This formed the basis to develop novel membranes to fabricate potentiometric sensors of FUR with BPB and BCG in THF and PVC. The acidic solution of FUR reacted with either BPB or BCG and formed a stable aqueous insoluble product of ion pair of yellow-coloured appropriate grain size. The probable structures of ion pair complexes are given below (Scheme 1). The aqueous insoluble ion pair complex of FUR-BPB or FUR-BCG is useful as a recombinant material to fabricate the membrane sensor. The membrane is formed effectively with uniform thickness all over its area when PVC is used as matrix and DBP is used as plasticizer. The membranes constructed here should therefore selectively respond to FUR, as the artificial ion-selective membrane sensors. The potentiometry will enable their use to confirm the selective functioning by generating the potential difference due to the varied concentrations of FUR solutions at opposite sides of the membrane [44].
The obtained product was used successfully to construct two sensors, namely FUR-BPB (Sensor I) and FUR-BCG (Sensor II) using THF as solvent and PVC. These acidbase indicator-based Sensor I and Sensor II were utilised for the first time to quantify FUR in pure form and in tablets.  The membranes developed were used to fabricate Sensors I and II and to measure the potentials of the drug solutions. The systematic representation of the electrochemical cell is represented as

Conclusion
For the first time, the two membrane sensors for potentiometric determination of frusemide using two acid dyes, namely bromophenol blue and bromocresol green are presented. The sensors are applicable to quantify frusemide over a wider linear range of concentration with the Nernestian response. The sensors have been successfully used to reliably determine frusemide in pure state and tablets. Good agreement evolved out of the results while using these sensors to determine frusemide accurately and precisely. Potentiometry with these sensors is the most selective technique to determine frusemide in pharmaceutical laboratories. Since, potentiometry is neither highly sophisticated nor have stringent steps to be followed; these could be used as a conventional technique in place of other high-cost techniques such as HPLC, LC-MS, GC-MS, etc.