Poly(aquachlorobis(1,10–phenanthroline)copper(II)iodidemonohydrate)/GCE for simultaneous determination of caffeine and theophylline in human serum, tea, and tablet samples

2.2.1 Electrode modification

Poly(ACP₂CuIH)/GCE was synthesized following the procedure in our previous work (Kassa et al., 2021). Briefly: Glassy carbon electrode was mirror polished with 1.0, 0.3, and 0.05 µm alumina slurry, successively and then washed thoroughly with distilled water. Poly(ACP₂CuIH)/GCE was fabricated potentiodynamically by scanning the potential between −1.2 and + 1.8 V at 100 mV s⁻¹ for 15 cycles in pH 7.0 PBS containing 1.0 mM ACP₂CuIH. The modified electrode was then rinsed with distilled water and conditioned by sweeping it in monomer free 0.5 M H₂SO₄ between −0.8 to + 0.8 V at 100 mV s⁻¹ until a steady voltammogram was obtained.

2.2.2 Preparation of stock and working standard solutions

Stock solutions of CAF and TP, each with 10.0 mM concentration in 100 mL volumetric flask, were prepared by dissolving 194.2 mg of CAF and 180 mg of TP in the respective flask and fill to the mark with doubly distilled water. An equi-molar (1.0 mM) mixture of CAF and TP in pH 5.0 (0.1 M) acetate buffer solution (ABS) was prepared in a 50 mL flask by mixing measured volumes of each from the respective stock solution. Finally, the individual and mixed equi-molar standard TP and CAF working solutions of various concentrations (1.0, 5.0, 10.0, 20.0, 40.0, 80.0, 120.0, 160.0, and 200.0 μM) were prepared from the stock solutions by serial dilution using pH 4.5 ABS (Mekassa et al., 2017).

2.2.3 Preparation of real samples

3.2.1 Cyclic voltammetry

Cyclic voltammetry using [Fe(CN)₆]^3-/4- as a probe was used to characterize the poly(ACP₂CuIH)/GCE (Fig. 3). In contrast to the characteristic redox peaks with peak-to-peak separation (ΔE) of 0.424 V at the unmodified electrode, a pair of redox peaks with much enhanced current and ΔE of 0.096 V at the poly(ACP₂CuIH)/GCE for [Fe(CN)₆]^3-/4- verified surface modification by a material that improved reversibility of the probe. The observed current enhancement of the probe at the modified electrode might be ascribed to the reported increased effective surface area of the modified electrode (Kassa et al., 2021).

Close

Fig. 3 CVs of unmodified (a) and modified (b) glassy carbon electrodes in pH 7.0 PBS containing 10.0 mM [Fe(CN)₆]^3-/4- and 0.1 M KCl at a scan rate of 100 mV s⁻¹.

Export to PPT

3.2.2 Electrochemical impedance spectroscopy

As can be observed from the Nyquist plots of the unmodified and polymer modified electrodes in [Fe(CN) ₆]^3-/4- (Fig. 4), both electrodes showed semi-circles of different diameters and hence different charge transfer resistance (R_ct) at high frequency region, and a linear line at low frequency region representing the diffusion of the electroactive species from the bulk of solution towards the electrode surface.

Close

Fig. 4 Nyquist plots for (a) bare GCE and (b) poly(ACP₂CuIH)/GCE in pH 7.0 PBS containing 10.0 mM [Fe(CN)₆]^3-/4- and 0.1 M KCl at frequency range: 0.01–100,000 Hz, applied potential: +0.23 V, and amplitude: 0.01 V. Inset: proposed equivalent circuit where Z_d stands for the Warburg diffusion constant.

Export to PPT

The presumed RC-parameters (R_s, R_ct, and C_dl) for the studied electrodes as calculated using eq. (1) from the respective semi-circles, and electrode surface roughness ratio calculated using eq. (2) are summarized in Table 1.

where f_max, is the frequency (Hz) corresponding to the maximum value of –Z” at the semi-circle (316.2 Hz for GCE, and 40.7 Hz for modified GCE), R_ct is the charge transfer resistance given by the diameter of the semi-circle, R_s is solution resistance given by the x-axis value corresponding to the semicircle at maximum frequency, R_F is roughness factor, C_dl is the double layer capacitance of the modified electrode, and C_s is the capacitance of the smooth surface of the unmodified electrode.

While the higher double layer capacitance (C_dl 4.70 × 10^-6F) of the modified electrode is an indication of deposition of the polymer film on the surface of the electrode, the lower charge transfer resistance (R_ct 831 Ω) indicated surface modification by a material that facilitates charge transfer between the electrode and electroactive species. Lower ΔE (0.096 V) value for the probe at the modified electrode (curve b of Fig. 3) relative to the ΔE value (0.424 V) at the unmodified electrode could therefore be ascribed to the lower charge transfer resistance of the modified electrode. Moreover, the surface roughness (R_F) value of 52.4, calculated using eq. (2), indicated modification of the surface of the electrode by a material that increased the surface roughness and hence effective surface area as reported in our previous work Kassa et al., 2021).

3.3.1 Electrochemical behavior of TP and CAF at poly(ACP₂CuIH)/GCE

Simultaneous electrochemical behavior of TP and CAF in pH 5.0 ABS containing an equi-molar mixture of the two (1.0 mM) was investigated (Fig. 5). Both TP and CAF showed irreversible oxidation at both the unmodified and polymer modified glassy carbon electrodes although with different current intensity and peak potential. In contrast to the weakly resolved oxidative peaks at the unmodified electrode (curve A of inset), oxidative peaks with much enhanced peak current at a reduced potential at the modified electrode (curve B of inset) showed the catalytic property of the modifier towards irreversible oxidation of TP and CAF. Moreover, appearance of sufficiently separated (ΔE 220 mV) and resolved oxidative peaks for the two drugs at the polymer modified electrode indicated the potential applicability of the method for simultaneous detection of TP and CAF.

Close

Fig. 5 CVs of bare GCE (a, c) and poly(ACP₂CuIH)/GCE (b, d) in 0.1 M ABS pH 5.0 containing no (a, b) and 1.0 mM CAF and TP (c, d) at 100 mV s⁻¹. Inset: corrected for blank CVs of (A) bare GCE, and (B) poly(ACP₂CuIH)/GCE.

Export to PPT

While the observed peak current enhancement might be attributed to the improved surface roughness and effective surface area as reported elsewhere (Kassa et al., 2021), lowering of peak potential could be due to the improved charge transfer resistance of the modified electrode.

3.3.2 Effect of scan rate

The effect of scan rate on the simultaneous oxidation of TP and CAF at the poly (ACP₂CuIH)/GCE was investigated (Fig. 6). The observed potential shift with scan rate observed for both TP and CAF confirmed the irreversibility of the oxidation reaction they undergo (Fig. 6A). Better determination coefficient value (R²) for the dependence of oxidative peak current of both TP and CAF on the square root of scan rate (0.99890 & 0.99953, respectively) (Fig. 6C) than on the scan rate (0.99553 & 0.96040, respectively) (Fig. 6B) suggested predominantly diffusion-controlled irreversible oxidation of TP and CAF (Wang et al., 2016; Mekassa et al., 2017). Furthermore, slope values of 0.42 (TP), and 0.47 (CAF) for the plot of log (peak current) versus log (scan rate) (Fig. 6D), both values being very close to the ideal value for a diffusion controlled reaction (0.5), confirmed the diffusion mass transport kinetics of the oxidation of both TP and CAF at the polymer modified electrode (Jesny and Girish, 2017).

Close

Fig. 6 (A) CVs of poly(ACP₂CuIH)/GCE in pH 5.0 ABS containing equi-molar (1.0 mM) mixture of TP and CAF at various scan rates (a–l: 10, 20, 40, 60, 80, 100, 125, 150, 175, 200, 250 and 300 mV/s, respectively), plot of (B) Ip vs. ν, (C) Ip vs. ν^1/2, and (D) log (Ip) vs. log (ν).

Export to PPT

The number of electrons involved (n), and the electron transfer coefficient (α) for irreversible oxidation of CAF and TP at the poly(ACP₂CuIH)/GCE were calculated using eq. (3 & 4) (Kassa et al., 2021):

Taking E_p and E_1/2 for the cyclic voltammogram at a scan rate of 100 mV/s (Fig. 6A) to be (1435 & 1300 mV for CAF) and (1206 & 1120 mV for TP), the value of αn from eq. (3) was calculated 0.36 (CAF) and 0.56 (TP).

For an irreversible redox reaction, the dependence of E_p on ln(scan rate) is governed by the Nicholson equation(eq. (4)) (Jesny and Girish, 2017; Kassa et al., 2021).

Where E_p- peak potential, E^o - formal potential, K⁰- electrochemical rate constant (s⁻¹), T-experimental absolute temperature (298 °C), F- Faraday’s constant (96, 436C mol⁻¹), R- gas constant (8.314 J. mol⁻¹ K⁻¹), ν- scan rate, and D- diffusion coefficient (cm² s⁻¹).

From the plot of E_pa versus ln(scan rate) for both TP and CAF (Fig. 7), equations with slope $(slope=0.013/n-\alpha n)$ of 0.03 for TP and 0.04 for CAF were obtained at the experimental temperature of 25 °C. Using the αn from eq. (3) and n–αn from eq. (4), n and α values were calculated as 0.69 (∼1.0) & 0.52 for CAF,and 0.99 (∼1.0) & 0.57 for TP, respectively. The calculated α values, which re both close to the ideal 0.50 for the irreversible process, confirmed the irreversibility of the oxidation of CAF and TP at the modified electrode (Bard and Faulkner, 2001; Wang et al., 2018; Guan et al., 2021).

Close

Fig. 7 Plot of Ep vs. lnν/ V s⁻¹ of poly(ACP₂CuIH)/GCE in pH 5.0 ABS containing equi-molar (1.0 mM) mixture of TP and CAF at various scan rates (a–l: 10, 20, 40, 60, 80, 100, 125, 150, 175, 200, 250 and 300 mV/s, respectively).

Export to PPT

3.3.3 Effect of pH on Ip and Ep of TP and CAF

The dependence of the peak current and peak potential of TP and CAF on the pH of the 0.1 M acetate buffer solution (ABS) was investigated in the range from 3.0 to 6.0 (Fig. 8A). The peak current of both TP and CAF decreased slowly with increasing the pH of the supporting electrolyte (Fig. 8B), which might be ascribed to the increasing columbic repulsive force between the ligands in the complex (pka(phen) = 4.86) (Gans et al., 1976) and the analytes (TP and CAF pka1 = 3.5). Taking the peak current increment, associated background current, accompanied peak shape distortion, and peak potential separation into consideration, pH 4.5 was selected as the optimum pH for further analyses.

Close

Fig. 8 (A) Representative CVs of poly(ACP₂CuIH)/GCE in ABS of various pHs (a–g: 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, and 6.0, respectively) containing equi-molar (1.0 mM) mixture of TP and CAF, and plot of mean ( x ¯ ± % R S D ) (B) Ip vs. pH, and (C) Ep vs. pH in the entire pH range.

Export to PPT

The peak potential of TP and CAF shifted in the negative potential direction with increasing the pH (Fig. 8A) indicating the participation of protons during their oxidation at the modified electrode. As presented in Fig. 8C, plot of peak potentials vs. pH values in the range of 3.0 to 6.0 revealed regression equations of Ep (V) =  − 0.0486pH + 1.406 for TP, and Epa (V) =  − 0.017pH + 1.491 for CAF (Fig. 8C). While the slope of − 0.048 for TP, which is close to the theoretical Nernstian value of 0.059 V, suggested an oxidation process that involves electrons and protons in a 1:1 ratio (Amare and Admassie, 2012; Shu et al., 2017; Wang et al., 2018; Quan et al., 2021); the slope of − 0.017 for CAF is too lower than the ideal value of 0.059 V for a 1:1 ratio of proton and electron involvement. This might be an indication of the oxidation of CAF without the involvement of protons as is reported in literature (Chitravathi and Munichandraiah, 2016; Wang et al., 2016; Mekassa et al., 2017; Shu et al., 2017; Guan et al., 2021).

3.4.1 Optimization of square wave voltammetric parameters

The dependence of SWV peak current response of poly(ACP₂CuIH)/GCE for TP and CAF in pH 4.5 ABS on square wave parameters such as step potential, amplitude, and frequency was investigated. CAF and TP showed peak current increment with increasing each SWV parameter as theoretically expected at the modified electrode (SM-1 (A-C)). As a compromise between the current enhancement, and associated non-Faradaic current with each SWV parameter; the amplitude of 35 mV, step potential of 6 mV, and frequency of 20 Hz were taken as optimum values.

3.4.2 Calibration curve for simultaneous determination of TP and CAF

A calibration curve for simultaneous determination of CAF and TP was obtained after making a survey on the response of the fabricated modified electrode for each in the presence of constant amount of the other. Under the optimized method parameters, the current response for CAF at constant TP (Fig. 10A) showed linear dependence on the concentration of CAF in the range 1.0–200.0 µM while current of TP remaining constant irrespective of the CAF content. Current response for TP also varied with concentration in the same range of concentration with comparable sensitivity as does the CAF (Fig. 10B).

Close

Fig. 10 Background subtracted SWVs of poly(ACP₂CuIH)/GCE in pH 4.5 ABS containing different concentrations of (A) CAF (a–i: 1.0, 5.0, 10.0, 20.0, 40.0, 80.0, 120.0, 160.0, and 200.0 μM, respectively) and 60.0 µM TP, and (B) TP (a–i: 1.0, 5.0, 10.0, 20.0, 40.0, 80.0, 120.0, 160.0, and 200.0 μM, respectively) and 60.0 µM CAF. Insets: respective plot of peak current vs. concentration.

Export to PPT

Finally, the response of the electrode for simultaneously varying equi-molar concentrations of CAF and TP was evaluated in triplicate. As can be seen from Fig. 11, the method revealed two well-shaped and sufficiently separated peaks for CAF and TP whose current increased with the concentration in the entire studied range (1.0–200.0 µM). Identical sensitivity for each drug independent of the concentration of the other showed the suitability of the method for simultaneous determination of the two drugs irrespective of their concentrations.

Close

Fig. 11 (A) Corrected for background SWVs of varying equi-molar mixtures of TP and CAF in pH 4.5 ABS (a–i: 1.0, 5.0, 10.0, 20.0, 40.0, 80.0, 120.0, 160.0 and 200.0 μM, respectively) at poly(ACP₂CuIH)/GCE, (B) Plot of the mean oxidative peak current (%RSD as error bar) vs. concentration of (a) TP, and (b) CAF.

Export to PPT

The low associated %RSD values (≤3.62% for TP and ≤ 2.90% for CAF for n 3) showed the precision of the developed method based on the poly(ACP₂CuIH)/GCE for simultaneous determination of TP and CAF. While LoD (3 s/m for n = 6) was 8.92 × 10^-3, and 1.02 × 10^-2 µM, LoQ (10 s/m) was 3.4 × 10^-2, and 2.97 × 10^-2 µM for TP and CAF, respectively.

3.5.1 Tea samples

Extracts of the three tea samples, prepared in pH 4.5 ABS as described under the experimental part, were analyzed for CAF and TP content. Among the two oxidative peaks that appeared for the three tea brands (Fig. 12), while the peak at about 1.38 V was assigned CAF, the peak at about 0.88 V (peak x) was ascribed to theobromine (Nia et al., 2021) showing absence of detectable TP in all the analyzed tea brands. The detected CAF levels in the tea samples were Addis tea (25.98), Wush wush tea (29.08), and Black Lion tea (30.51) mg of CAF/g tea sample, which are all in good agreement with literature reported values (Yang et al., 2010; Guo et al., 2011; Tefera et al., 2016).

Close

Fig. 12 SWVs of poly(ACP₂CuIH)/GCE in pH 4.5 ABS containing (A) Addis tea, (B) Black Lion tea, and (C) Wush wush tea samples.

Export to PPT

3.5.2 Tablet samples

CAF and TP content of the two selected tablet brands; panadol extra (labeled 65 mg/tablet CAF), and theodrine (labeled 120 mg/tablet TP) was determined using the developed method. As can be seen from Fig. 13, whereas the voltammogram for panadol extra tablet showed only one peak (Ep 1.320 V) assigned for CAF (Fig. 13A), the voltammogram for theodrine also revealed one peak (Ep 1.133 V) assigned to TP (Fig. 13B). This showed absence of TP in the panadol extra tablet and CAF in the theodrine tablet sample.

Close

Fig. 13 Corrected for blank SWVs of poly(ACP₂CuIH)/GCE in pH 4.5 PBS containing A) panadol tablet, and B) theodrine tablet.

Export to PPT

Meanwhile, the detected CAF level in the panadol tablet is 100% of the company’s label, the TP content in theodrine tablet is 101.0% of the company’s label (Table 2). Agreement between the detected CAF and TP amounts with the nominal values showed the applicability of the method for determination of CAF and TP in tablet samples that are known to contain complex excipients.

3.5.3 Human blood serum sample

The proposed method was also applied for determination of both CAF and TP in a human blood serum sample. The SWV for the unspiked human blood serum sample (curve a of Fig. 14) showed no visible peak indicating absence of any oxidizable species including CAF and TP at a detectable level, which is in agreement with previous report (Wang et al., 2016).

Close

Fig. 14 Corrected for blank SWVs of human blood serum sample in pH 4.5 ABS spiked with equi-molar mixtures of TP and CAF (a–c: 0.0, 80.0, and 120.0, respectively) at poly(ACP₂CuIH)/GCE.

Export to PPT

3.6.1 Spike recovery

The accuracy of the developed method as one of the validation parameters was checked by the spike recovery analysis.

3.6.1.2

3.6.1.2 Spike recovery from tea samples

As discussed under the real sample analysis section, the three unspiked tea brands all showed a peak for CAF and a peak speculated for theobromine (TB), but no peak for TP. Fig. 15 presents overlaid SWVs for unspiked and spiked tea samples (Black Lion tea, Wush wush tea, and Addis tea). While the peak intensity for the TB remained constant, appearance of a new peak as a consequence of spiking the samples with TP confirmed initially absence of TP in the three tea brands. Moreover, similar current for same concentration of CAF, TP, and TB in the samples showed the reproducibility of the results.

Close

Fig. 15 Corrected for blank SWVs of A) Black Lion, B) Wush wush, and C) Addis tea samples spiked with TP and CAF (a–c: unspiked tea sample, a + 0.0 CAF + 100.0 µM TP, and a + mix of 100.0 µM TP & CAF, respectively) at poly(ACP₂CuIH)/GCE.

Export to PPT

As presented in Table 4, spike recovery results in the range 97.0–102.0% for TP, and 95.4–96.8% for CAF in the three tea brand samples showed the accuracy and hence validated the method for simultaneous determination of TP and CAF in food samples.

Table 4 Summary of recovery results of TP and CAF in Addis, Wush wush, and Black Lion tea brand samples spiked with 100.0 μM of TP, and 100.0 μM equi-molar mixture of TP & CAF.

Tea brand	Analyte	Detected before spike (μM)c	Spiked amount (μM)	Detected after spike (μM)d	Recovery (%)
Addis	TP	----	100.0	100.0 ± 0.025	100.0
			100.0	102.0 ± 0.022	102.0
	CAF	83.6 ± 1.312	---	83.6 ± 1.310	---
			100.0	180.0 ± 0.016	96.4
Wush wush	TP	----	100.0	97.0 ± 0.022	97.0
			100.0	97.0 ± 0.023	97.0
	CAF	93.6 ± 1.071	---	93.6 ± 1.070	---
			100.0	189.0 ± 0.024	95.4
Black Lion	TP	----	100.0	99.0 ± 0.032	99.0
			100.0	100.0 ± 0.032	100.0
	CAF	98.2 ± 1.170	---	98.2 ± 1.170	---
			100.0	195.0 ± 0.021	96.8

^{c & d} Detected mean ±%RSD for n = 3.

3.6.1.3

3.6.1.3 Spike recovery from tablet samples

As can be observed from Fig. 16, panadol and theobromine tablets spiked with both TP and CAF showed peak current increment in proportion to the amounts spiked. Excellent spike recovery results in the range 99.0–101.0% for TP, and 98.9–100.0% for CAF in both tablet samples (Table 5) further confirmed the accuracy of the developed method for its applicability for simultaneous determination of TP and CAF in pharmaceutical formulations.

Close

Fig. 16 Corrected for blank SWVs of poly(ACP₂CuIH)/GCE in pH 4.5 ABS containing (A) panadol tablet spiked with TP and CAF (a–c: unspiked panadol tablet sample, a + 80.0 µM TP, and a + 80.0 µM equi-molar mixture of TP and CAF, respectively), and (B) theodrine tablet brand spiked with TP and CAF (a–c: unspiked tablet sample, a + 80.0 µM CAF, and a + 80.0 µM equi-molar mixture of TP and CAF, respectively).

Export to PPT

Table 5 Summary of spike recovery results of TP and CAF in theodrine and panadol extra tablet samples.

Tablet sample	Analyte	Amount before spike (μM)	Spiked amount (μM)	Detected after spike (μM) e	Recovery (%)
panadol	TP	---	80.0	79.2 ± 0.027	99.0
			80.0	79.2 ± 0.026	99.0
	CAF	80.0	0	80.0 ± 1.500	100.0
			80.0	159.1 ± 0.021	99.4
theodrine	TP	80.0	0	80.8 ± 1.880	101.0
			80.0	160.0 ± 0.020	99.5
	CAF	---	80.0	79.1 ± 0.014	98.9
			80.0	79.1 ± 0.015	98.9

^eDetected mean±%RSD for n = 3.

3.6.2 Interference study

The selectivity of the method for detection of CAF and TP in the presence of several potentially interfering substances was examined.