3.1. ITIES studies
Initially, the comprehensive electroanalytical analysis of DANO was examined at the ITIES using ITV technique. The pH of the aqueous phase was adjusted to 2 being significantly lower than the DANO pKa1 and pKa2 values (6.07 and 8.50, respectively)32. Consequently, according to the analysis of the antibiotic’s structure and the concentration fraction diagram plotted for DANO (see Fig. 2A) all ionizable functional groups present within the studied analyte structure are protonated and hence the molecule is fully charged (exists in the aqueous phase as only cationic fraction). Figure 2B displays a graph depicting the DANO ion partition diagrams (dependency of the formal Galvani potential difference - \({\varDelta }_{org}^{aq}\varPhi\) - of the DANO ion transfer plotted in function of the aqueous phase pH). \({\varDelta }_{org}^{aq}\varPhi\) is taken from ITVs recorded in a broad pH range (2–12) of BRBs used as the aqueous phase, as illustrated in Fig. S1 (see electronic supporting information). In the structure of DANO we can distinguish a carboxylic group with pKa1 of approximately 6 and peripheral nitrogen atoms, which are part of the piperazine ring with pKa2 value around 8.5. With this in mind, at pH 2, the carboxylic acid groups are not dissociated, and the piperazine units are protonated, leading to a fully positively charged antibiotic molecule. To transfer the positively charged (cationic) molecules from the aqueous phase to the organic phase, the LLI was polarized from less positive to more positive potentials during the forward scan 6. When the pH of the aqueous phase is significantly lower than the DANO pKa value (pH 5), the positively charged analyte undergoes a direct ion transfer reaction from the aqueous to the organic phase upon application of a Galvani potential difference exceeding + 0.123 V. Since the DANO possess two functionalities that can be either positively or negatively charged, in the pH range from 5 to 10 a fraction of zwitterions exists in the aqueous phase with a peak concentration found at around 7. With an increase in the pH of the aqueous phase, neutral/zwitterionic DANO molecules distribute into the organic phase (denoted by the vertical black arrow pointing towards x axis in Fig. 2B). The presence of neutral DANO in the organic phase can facilitate the transfer of protons from the aqueous to the organic phase, necessitating Galvani potential difference values exceeding + 0.123 V. The expected behavior of DANO is depicted by the dashed red line calculated according to Eq. 1 which is in line with the experimental findings marked on the Fig. 2B with black data points .
\({\varDelta }_{org}^{aq}{\varphi }_{0.5}={\varDelta }_{org}^{aq}{\varphi }^{0}+\frac{RT}{nF}\text{ln}\left(\frac{{10}^{-pH}+{K}_{a}{K}_{D}+{K}_{a}}{{10}^{-pH}}\right)\) (Eq. 1)
In Eq. 1 the acid dissociation constant is represented as Ka (pKa value of 8.50). KD is the distribution constant, delineating the ratio between the concentration of the non-protonated form of DANO present in the aqueous [DANO]aq and the organic phase [DANO]org:
\({K}_{D}=\frac{{\left[DANO\right]}_{aq}}{{\left[DANO\right]}_{org}}\) (Eq. 2)
The experimental results demonstrated that the optimal correlation was obtained for a KD value of around 600, suggesting the inherently hydrophobic characteristics of DANO molecules in the neutral form. In more accessible terms, for every 600 molecules of DANO in the organic phase, 1 molecule will be present in the aqueous phase when the pH approaches the pKa2 value.
Figure 2C shows a series of ITVs recorded for fixed concentration of DANO (333.3 µM) and varied potential scan rate value. The observed process demonstrated reversibility, with the forward and backward currents intensity ratio being close to unity, while the peak-to-peak separation (ΔEp) was found to be ~ 75 mV (measured for ITV recorded at 25 mV s− 1). This value is in proximity to the anticipated theoretical value of 59 mV z− 1 (z = 1), indicating the mono-charged nature of the DANO cation undergoing ion transfer reaction. Deviation from the expected theoretical value of 59 mV z− 1, where z represents the molecular charge of the analyte, is commonly observed for polarized LLIs and arises from the resistive properties of the organic phase 6,33. By analyzing the linear fit equation of the relationship between the current signal and the square root of the scan rate (v1/2) (Fig. 2D) and the Randles – Ševčík equation we have calculated the aqueous and the organic DANO diffusion coefficients (D). Obtained values were equal to Daq→org = 1.13×10− 6 cm2·s− 1 and Dorg→aq = 0.14×10− 6 cm2·s− 1. Another parameter derived from ITVs is the formal Galvani potential of DANO (in cationic form) ion transfer (\({\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\)). This parameter is closely related to the hydrophobicity/-philicity of the analyte under study. For cationic species, a higher value of \({\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\) signifies greater hydrophilicity 34. The \({\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\) serves as a valuable parameter, which, in conjunction with Eq. 3, enables the calculation of the formal water | 1,2-DCE partition coefficient (\({logP}_{water/DCE}^{{\prime }}\)). This coefficient quantitatively characterizes the partitioning behavior of the charged molecule between the aqueous and the organic (1,2-DCE) phases 35.
\({logP}_{water/DCE}^{{\prime }}=-\frac{{\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}{z}_{i}F}{2.303RT}\) (Eq. 3)
where: \({\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\) is the formal Galvani potential of the ion transfer reaction (V); zi – charge of the investigated analyte; F – the Faraday constant (96485 C·mol− 1); R – the gas constant (8.314 J mol− 1·K− 1) and T – the temperature (298 K). For DANO, the calculated \({logP}_{water/DCE}^{{\prime }}\) is
-2.08, indicating its relatively high hydrophilicity (given that it is built from the aromatic rings and has fluorine substituent). This value is also in line with \({logP}_{water/octanol}^{ }\)= − 1.37 (pH 3) reported by G.M. Cardenas-Youngs and J.L. Beltrán 36 which also suggests that DANO is a hydrophilic compound. Ultimately, the \({\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\)was employed to calculate the formal Gibbs free Energy of the interfacial ion transfer reaction (\(\varDelta {G}_{ }^{{\prime }, aq\to org})\) according to Eq. 4:
\(\varDelta {G}_{ }^{{\prime }, aq\to org}={z}_{i}F{\varDelta }_{org}^{aq}{\varPhi }_{ }^{{\prime }}\) (Eq. 4)
All physicochemical parameters determined for DANO are presented in Table S1 in electronic supporting information.
Finally, we harvested the fact that the DANO is electrochemically active at the ITIES to developed the procedure for the electroanalytical determination of the concerned analyte. For this purpose, the ITVs were recorded for increasing DANO concentrations, as illustrated in Fig. 2E. Subsequently, the dependencies of forward and backward peak current intensities vs. CDANO were plotted, as shown in Fig. 2F. Notably, in both cases, DANO transfer from the aqueous to the organic phase and vice versa, the coefficients of determination (R2) approached unity. The linear relationship between Ip values (positive or negative currents) and CDANO within the LDR of 7.13–333.3 µM is evident from Fig. 3F. Based on these findings, crucial electroanalytical parameters including linearity, sensitivity, LODs, LOQs, were determined and are compiled in Table 1.
3.2. Electroanalytical study of DANO at GCE
The next step of this work involved utilizing GCE combined with SWV and CV techniques for the electrochemical investigation and determination of DANO. The electrochemical response of DANO was first investigated across wide pH spectrum provided by BRB (pH 2–10, Fig. S2A). The SWV studies were carried out within the potential range from + 0.4 V to + 1.5 V. Preliminary analysis revealed that DANO exhibits two oxidation signals, one at approximately + 0.25V and the other at around + 1.2V versus Ag/AgCl (3M KCl). The analytical signal with a better-defined shape and higher current intensity was observed at the potential ~ + 0.25V. Therefore, for the purpose of this study, this signal was subjected to further investigations and electroanalytical quantification of DANO. The most pronounced signals of DANO were detected under acidic conditions (pH 2.0), hence the BRB with pH = 2 was chosen as the supporting electrolyte for subsequent studies (Fig. S2B from electronic supporting information). Additionally, we have noticed that as the pH increased, the oxidation peak of DANO shifted towards more cathodic potentials, indicating the involvement (as expected) of protons in the electrochemical process 37. Furthermore, the relationships between the peak potential (Ep) and pH is linear only in pH range (3–7) (Fig. S2C), suggesting that the electrochemical reaction is more complex, and on factors other than hydrogen or hydroxyl ions concentration in the solution.
To scientifically assess the optimum conditions for determining DANO using SWV in conjunction with GCE, the influence of potential modulation parameters, such as frequency (f), amplitude (ESW) and step potential (∆E) were examined. The obtained results, indicated that the highest oxidation peak (at approximately + 0.25V) and the best shape of the DANO signal were observed with the following parameters: a frequency of 60 Hz, an amplitude of 90 mV and a step potential of 12 mV. Finally, the developed SWV procedure was employed for DANO determination in a model sample. SWV technique was used under the optimal experimental conditions. The usefulness of the SWV for the assay of DANO was estimated as a function of the peak current (Ip) of increasing DANO concentrations (CDANO) in three runs (n = 3). The developed procedure for SWV determination of DANO was also validated. The significant validation parameters, such as linearity, LOD, LOQ and precision were evaluated (see Table 1). The SWVs and the corresponding calibration graph are depicted in Fig. 3. As can be noticed from this figure, the oxidation peak current increased linearly in LDR of 13.71 to 373.6 µM.
The CV technique facilitates the extraction of valuable insights regarding the electrode process, including kinetic parameters, reversibility or its nature (diffusion- or adsorption controlled charge transfer processes) 38. In this paper CV technique was employed to elucidate the electrochemical behaviour of DANO. CV analyses of DANO were carried out within a potential range from − 0.4 to 1.40 V, at scan rates in the range of 10–500 mV s− 1. The recorded CVs, conducted in the presence of [DANO] = 146 µM in BRB solution at pH 2, revealed an electrochemical process spanning nearly over entire potential window (from − 0.2V to 0.9V) and only one analytical, anodic peak at approximately + 1.30 V, which was analysable (Fig. 3C). The latter most probably originated from the oxidation of the peripheric nitrogen atom from the piperazine ring. To assess the nature of the electrochemical process happening at the GCE during DANO oxidation, the dependence of the Ip on the v was analyzed (Fig. 3D). The relationship Ip vs. v shows a linear correlation, which indicates an adsorption-controlled process. To confirm the obtained results, a plot correlating the logarithms of the peak current (log Ip) and the scan rate (log v) was generated, yielding a slope of 0.4049 (R2 = 0.9798) (Fig. S3), closely aligning with the theoretically anticipated value of 0.5 for an diffusion-controlled process 38,39. Hence, in this case the character of DANO oxidation process at the GCE is not unequivocal and indicated a mixed adsorption-diffusion process 38,40,41.
Table 1
Electroanalytical parameters of DANO obtained at the ITIES and GCE.
Employed configuration Parameter | ITIES | GCE |
Number of repetitions | 3 |
Linear concentration range [µM] | 7.13–333.3 | 13.71–373.6 |
Slope (a) (A M− 1) | 0.1022 aq→org 0.1168 org→aq | 0.0300 |
Standard error of slope (SEa) [a] | 0.0004 aq→org 0.0004 org→aq | 0.0005 |
Intercept (b) (µA) | 0.1022 aq→org -0.1168 org→aq | 0.0300 |
Standard error of intercept (SEb) [a] | 0.0725 aq→org 0.0787 org→aq | 0.1037 |
Coefficient of determination (R2) | 0.9999 aq→org 0.9999 org→aq | 0.9986 |
LOD (µM) [b] | 2.13 aq→org 7.09 org→aq | 10.37 |
LOQ (µM) [c] | 2.02 aq→org 6.74 org→aq | 13.71 |
[a] SE = SD/n1/2; [b] LOD = 3SDb / a; [c] LOQ = 10SDb / a; a – slope and b – intercept; aq→org – corresponds to parameter calculated for the positive signals; org→aq – corresponds to parameter calculated for the negative signals.
3.3. Real samples analysis
The next stage of this research involved the application of both developed procedures (ITV and SWV at ITIES and GCE, respectively) for the DANO determination in samples of cow’s milk. The samples analyzed comprised ultra-high temperature (UHT) milk with a fat content of 1.5% obtained from a nearby supermarket.
3.3.1. DANO detection at ITIES
In ITIES-based experiments, the aqueous phase was substituted with a 3.5 mL of the milk sample, which did not necessitate prior purification (or treatment) to eliminate fats, proteins, saccharides, or other chemical constituents. Subsequently, suitable volumes of DANO standard solution were introduced into the test sample, and the ITVs were recorded in three runs (n = 3) as the DANO concentrations in the milk sample increased. Figure 4A depicts the ITVs obtained during the addition of specific volumes of DANO stock solution into the milk samples, after subtracting blank reading (recorded in the absence of DANO). The analytical signals, manifested as negative currents, displayed a linear correlation (Fig. 4B) with the increasing DANO concentration within the LDR of 14.24 to 104.9 µM. The LODs were derived from the calibration curves. The calculated LOD value for the determination of DANO in milk samples was determined to be 7.32 µM for DANO based on the signals attributed to the analyte transfer from the organic to the aqueous phase. The LOQ is defined as the lowest concentration contained within the calibration curve (14.24 µM).
3.3.2. DANO detection at GCE
To determine DANO, we have started by recoding the SWV at GCE immersed into 10 mL of milk sample (used instead of supporting electrolyte). Electroanalytical parameters applied at this stage were taken from the optimization study described in Section 3.2. Following that, the standard addition method was employed. Consecutive volumes of DANO stock solutions were introduced into the voltammetric cell utilizing a micropipette. The SWVs together with corresponding calibration graph are depicted in Fig. 4 (C and D). As can be noticed from Fig. 4C the oxidation peak current of DANO increased linearly within the LDR of 19.92 to 49.50 µM. The calculated LOD value for the determination of DANO in milk samples was determined to be 4.00 µM. The LOQ is defined as the lowest concentration contained within the calibration curve (19.92 µM).
3.4. Interference studies
One prevalent challenge in chemical analysis involves the impact of interfering agents (IA) on the recorded analytical signals. We can only claim that the method is selective when the impact of the interfering species do not exceed ± 10% 42. Despite the fact that both developed procedures exhibited high applicability, as they allowed for the determination of DANO in such a complex matrix as milk samples without difficulty, the impact of potential interfering species on the recorded DANO signals was also investigated. Consequently, the influence of potential interferents, including milk contaminants, such as: citric acid, galactose, lactose, glucose, calcium cations, potassium cations, magnesium cations, iron(III) cations, sodium lactate and orthophosphate (V) anions was evaluated by means of both elaborated procedures.
3.4.1. DANO detection at ITIES
Initially, ITVs were recorded for [DANO] = 70.40 µM, which served as a reference. Subsequently, the appropriate amount of interferent standard solution was added to the aqueous phase placed in the ITIES cell to achieve concentrations 14.3, 140.6, 277.4 and 1248.0 µM, respectively. After each aliquot of interfering agent (IA) addition the ITVs were recorded. Based on the results, it was observed that only in the case of iron (III) ions and citric acid, their significant influence on the recorded DANO signals was observed for each studied concentration. For the remaining IA, only at the highest IA concentration (1248 µM), their effect on the recorded DANO signals was observed. For the other IA concentration in the range from 14.3 to 277.4 µM, their influence on the analyte signals did not exceed 5.4%. Detailed information regarding the impact of potential IA is presented in Table S2 from electronic supporting information.
3.4.2. DANO detection at GCE
The interference study started with voltametric analysis of 10 mL of the supporting electrolyte placed in the voltammetric cell (BRB, pH = 2 ), followed by the addition of 143 µL of the DANO stock solution (CDANO in cell = 62.55 µM). Subsequently, the SWV of DANO was recorded. Then, specific volumes of IA stock solutions were added to the cell containing DANO at concentrations: 6.09, 56.20 and 572.4 µM, and voltammograms were recorded after each addition of IA. Unfortunately, the results of the experiments conducted at the GCE indicate low selectivity of the developed method. For all investigated IA, their significant influence on the recorded DANO signals was observed. Only in the case of potassium cations, magnesium cations, citric acid, sodium lactate and orthophosphate (V) anions, at their lowest concentration (6.09 µM), the influence of IA on the DANO signals did not exceed 10% (see Table S2 from electronic supporting information).