All reagents were of analytical grade and were used without further purification. Milli-Q water (Milli-Q, Millipore, USA, 18 MΩ) was used for solution preparation.

The Britton-Robinson buffer solution (4.0 × 10^-2 mol L^-1; pH = 2) was prepared by dissolving 4.95 g boric acid (Kemika, Croatia), 4.79 g glacial acetic acid (VWR Chemicals, France), and 5.45 g orthophosphoric acid (Kemika, Croatia) in 2.0 L of Milli-Q water. The pH value was adjusted by adding a 2.0 mol L^-1 solution of sodium hydroxide (Kemika, Croatia) while monitoring the solution pH.

Acetate buffer solutions (pH = 3.5–4.5) were prepared by mixing appropriate volumes of 0.5 mol L^-1 sodium acetate (Kemika, Croatia) and 0.5 mol L^-1 acetic acid (Kemika, Croatia) solutions. The pH was adjusted by adding a 2.0 mol L^-1 sodium hydroxide solution.

A stock solution of 8.0×10^-3 mol L^-1 Cu(II) was prepared by dissolving 99.9 mg CuSO₄·5H₂O (Kemika, Croatia) in 50.0 mL Milli-Q water.

A 2.0×10^-3 mol L^-1 solution of bathocuproine disulfonic acid (BCS) was prepared by dissolving 56.5 mg of bathocuproine disodium salt (Alfa Aesar, Karlsruhe, Germany) in 50.0 mL of Milli-Q water.

A 1.0 × 10^-2 mol L^-1 stock solution of PEN was prepared by dissolving 0.1492 g PEN (Fluka Chemika, Buch, Switzerland) in 100 mL Britton-Robinson buffer solution (pH = 2) and stored at 4 °C in a dark bottle to ensure its stability for at least 30 days. Working standards were prepared daily by diluting the above stock solutions with Milli-Q water.

A perchloric acid solution was prepared and standardized for the titrimetric analysis of the pharmaceutical samples in accordance with the British Pharmacopeia procedure (British Pharmacopoeia 2018).

All pH measurements were conducted with a pH meter (SevenMulti, Mettler Toledo, Switzerland) equipped with a combined glass electrode (InLab®413, Mettler Toledo, Switzerland).

All kinetic measurements were performed in a closed-configuration unsegmented continuous flow analyzer for continuous sample introduction, shown in Fig. 2.

Figure 2. 

Schematic of the closed-configuration unsegmented continuous flow analyzer for studying the kinetics of the analytical reaction used in the determination of PEN.

Such analyzers are characterized by unsegmented flow and continuous introduction of samples as a flowing stream at a low flow rate. They are suitable for monitoring evolving systems such as the kinetics of the reaction between PEN and the [Cu(BCS)₂]^2- complex. The analyzer consisted of a peristaltic pump (IPC, Ismatec, Switzerland) for continuously propelling the reaction solution, which was located in a thermostated, jacketed reaction vessel and mechanically mixed using a magnetic stirrer. The components of the analyzer were connected by PTFE tubing (0.8 mm i.d.). The absorbance of the Cu(I)-BCS complex, produced in the reaction vessel, was monitored in time at 483 nm in the flow-through cuvette (160 μL internal volume and 10 mm optical pathlength, Hellma, USA) of a UV/Vis spectrophotometer (UV-1601, Shimadzu, Japan) connected to a laptop running Hyper UV-Vis software (Shimadzu, Japan). The recorded kinetic data, with a frequency of 1 s per reading, were then transferred to GraphPad Prism Ver. 4.03 for Windows (GraphPad Software, San Diego, CA) for curve-fitting, regression, and statistical analysis. A thermostat bath with a digital thermoregulator (Julabo, Germany) was used to circulate water at a preselected temperature through the jacketed compartment of the reaction vessel to maintain the reaction solution at the desired temperature.

Acetate buffer (10 mL, pH = 5.0), Cu(II) solution (0.45 mL, 8.0 × 10^-3 mol L^-1), BCS solution (1.80 mL, 2.0 × 10^-3 mol L^-1), and 1.75 mL of Milli-Q water were added to the thermostated jacketed reaction vessel under continuous stirring. The resultant reaction solution was recirculated through the flow-through cuvette of the analyzer for 1 min at a constant flow rate of 6 mL min^-1 before the analytical reaction was initiated by adding 1.0 mL of PEN standard solution, thus bringing the total reaction solution volume to 15.0 mL.

Stock solutions (0.2 mol L^-1) containing common inorganic ions (K⁺, Na⁺, NO₃-, and SO₄^2-) and excipients in pharmaceutical formulations were used in the interference study. They were prepared by dissolving KNO_3, Na₂SO₄, lactose, D-(+)-glucose, D-(-)-fructose, sodium citrate, citric acid, L-(+)-tartaric acid, and boric acid (Kemika, Croatia) in Milli-Q water. Appropriately buffered solutions containing 4.0×10^-5 mol L^-1 PEN and the individual interferents in interferent:PEN molar ratios of 5:1, 10:1, 50:1, 100:1, 250:1, and 500:1 were analyzed.

A commercially available pharmaceutical (Metalcaptase tablets, 150 mg PEN per tablet, Heyl, Chemisch-Pharmazeutische Fabrik, GmbH & Co. KG, Germany) was analyzed to validate the newly developed method. The analytical procedure involved the mixing of five tablets in powder form. An amount of powder corresponding to 150 mg PEN (declared amount of PEN per tablet) was dissolved in 500 mL of Milli-Q water, and after filtering (blue ribbon filter paper, S&S, Germany), the supernatant was analyzed by both the newly developed method and the British Pharmacopeia titrimetric method (British Pharmacopoeia 2018).

Recovery studies involved the analysis of tablet samples spiked with 50–200 mg PEN.

It was established that after the addition of PEN standards to the reaction vessel (Fig. 2), the absorbance of the [Cu(BCS)₂]^3- complex at 483 nm rapidly increased, reaching a plateau in 50 s (Fig. 3). Therefore, it was decided to study the effect of the main analytical reaction (Eq. (2)) parameters (i.e., solution pH and temperature, and Cu(II):BCS and [Cu(BCS)₂]^2-:PEN molar ratios) on the absorbance of the [Cu(BCS)₂]^3- complex at 60 s. Taking into account that 60 s were used for achieving a stable baseline signal prior to the standard addition, the actual absorbance value used in the abovementioned study was reached 120 s after the start of the measurement.

Figure 3. 

Transient absorbance curves at 3 different PEN concentrations. The reaction parameters have the initial values listed in Table 1.

The initial and optimal values of the analytical reaction parameters and the ranges within which they were studied are listed in Table 1.

When studying the effect of pH using Britton-Robinson buffer, it was found that the absorbance increased from pH 2 to pH 3 and then leveled off. This result allowed the use of acetate buffer at pH 5 in all subsequent experiments, thus avoiding the introduction of borate, a potential interferent, in the reaction mixture when using the Britton-Robinson buffer.

The temperature effect was negligible in the temperature range studied (Table 1), and all subsequent experiments were conducted at room temperature of 25 °C. Therefore, no thermostating was required, thus simplifying the analytical procedure.

It was observed that the absorbance increased when the [BCS]:[Cu²⁺] ratio was increased from 0.5 to 1.0. However, further increase of this ratio up to 3.0 did not produce higher absorbance, and therefore the ratio of 1.0 was selected as the optimal value of this reaction parameter.

The absorbance increased with increasing the reagent to PEN ratio (i.e., [Cu(BCS)₂^2-]:[PEN]) from 2.0 to 4.0 and then leveled off. Hence, 4.0 was selected as the optimal value of this reaction parameter.

Under optimal reaction conditions (Table 1), the method is characterized by a linear range of 3.0×10^-6 – 6.0×10^-4 mol L^-1 described by the following calibration equation: A = 8.65×10³ [PEN] (R² = 1.000). The limit of detection (LOD), calculated as 3 × standard deviation of the blank, was found to be 9.0×10^-7 mol L^-1.

Table 2 compares the analytical performance of the newly developed method with existing kinetic spectrophotometric methods for the determination of PEN. As can be seen, the newly developed method offers higher sensitivity than all other methods, except for the method involving the use of Hg²⁺ (Agarwal et al. 2016), which is a highly toxic reagent. The newly developed method and the method utilizing the Cu(II)-neocuproine complex (Kukoc-Modun et al. 2021) provide the widest linear ranges and the shortest measurement time of 2 min; however, the latter method has a higher LOD value. On the basis of the analytical figures of merit presented in Table 2, it can be concluded that the newly developed method offers superior analytical performance over the existing kinetic spectrophotometric methods for the determination of PEN.

The influence of possible interfering species that are commonly found in commercial pharmaceutical formulations (excipients) and inorganic salts often present in the reaction solutions was investigated in synthetic solutions containing 4.0×10^-5 mol L^-1 PEN and the species studied in large excess. The tolerance limit was defined as the highest interferent:PEN ratio that would still cause an error within ± 5%. The results of the interference study are shown in Table 3. Taking into account that the tested concentrations of the common excipients listed in Table 3 were much higher than those commonly found in commercial pharmaceuticals, it was concluded that the newly developed method offered the required degree of selectivity for the determination of PEN in common pharmaceutical formulations.

The analytical reaction kinetics was studied under optimal conditions (Table 1) in the closed-configuration unsegmented continuous flow analyzer for continuous sample introduction by measuring the initial reaction rate (V, Eq. (3)) with respect to the PEN concentration in the reaction vessel in the range 2.0×10^-7–4.0×10^-5 mol L^-1, where the reagent (i.e., [Cu(BCS)₂]^2-) was in sufficiently large excess. It should be noted that this concentration range corresponds to the linear calibration range of PEN.

$V=\frac{\Delta c}{\Delta t}=\frac{\Delta A}{b\Delta t}=k{c}^n$ (3)

where Δt = 15 s, b is the slope of the calibration curve (8.65×10³ L mol^-1), k is the kinetic rate constant, and n is the reaction order.

By plotting lgV versus lgc_PEN (Fig. 4) and fitting it by Eq. (4), n and lgk were determined as 1.041 and 5.345, respectively, thus indicating that under optimal conditions and in the PEN concentration range mentioned above, the analytical reaction was of pseudo-first order.

Figure 4. 

Linear plot of lgV vs. lgc_PEN obtained under optimal conditions (Table 1) where V and c_PEN are the initial reaction rate (Eq. (3)) and the molar concentration of PEN in the range 2.0×10^-7–4.0×10^-5 mol L^-1 respectively.

lgV = lgk + nlgc (4)

The amount of PEN in commercial PEN tablets containing 150 mg PEN per tablet was determined in triplicate by the newly developed method and the reference method (British Pharmacopoeia 2018), and the results were found to be statistically indistinguishable at the 95% confidence level (i.e., 150.8 ± 0.5 mg and 147.2 ± 1.8 mg).

The recovery experiments using the same PEN tablets produced recovery values in the range of 97.4–102.6% (Table 4), thus further confirming the validity of the newly developed method.