Voltammetric determination of leucovorin in pharmaceutical preparations using a boron-doped diamond electrode
- 111 Downloads
Method for voltammetric determination of leucovorin, a drug frequently applied to decrease some unfavorable effects of anticancer drugs such as methotrexate or to increase the therapeutic effect of 5-fluorouracil, has been developed employing a bare boron-doped diamond electrode. It is the first method for leucovorin determination based on its electrochemical oxidation. Although at least three anodic and three cathodic voltammetric peaks could be recorded under the used conditions, only the anodic response situated at about + 900 mV (vs. saturated Ag|AgCl electrode) was suitable, namely due to its shape and position, for analytical purposes. Using differential pulse voltammetry with optimized parameters and supporting electrolyte of pH 3.0, the linear dynamic range of leucovorin determination was recorded from 0.15 to 25 μmol dm−3. Under such conditions, low limit of quantification of 0.050 μmol dm−3 and limit of detection of 0.015 μmol dm−3 as well was reached. Relative standard deviation calculated from 11 repeated measurements amounted to 0.7% and calculated from five repeated determinations amounting less than 3.0%. Applicability of the developed method was verified by repeated analysis of the pharmaceutical preparation with excellent results (recovery 98.7–102.8%, relative standard deviation 1.81%).
KeywordsBoron-doped diamond electrode Determination Leucovorin Pharmaceutical samples Voltammetry
As it is evident from the above-mentioned information, it is highly important to determine LV in pharmaceutical products and in body fluids. Various analytical methods have been used for these purposes up to now. From non-electrochemical methods application of mass spectrometry for these purposes can be mentioned . As in other cases, different separation methods have been used most frequently, e.g., high performance liquid chromatography (HPLC) with UV detection , with fluorescence detection , or with gradient elution with following dual UV–fluorescence detection [9, 10]. Pre-separation of an analyzed sample using solid-phase extraction has been described in literature as well [6, 7, 11]. LV levels in urine or serum samples without pre-separation steps have been also analyzed using spectrophotometric techniques [12, 13]. Some authors have reported the application of capillary zone electrophoresis [14, 15] or of kinetic fluorimetry  for LV determination.
On the other hand, only a little attention has been recently paid to the application of electroanalytical methods for LV determination. Using these methods (mainly voltammetry and polarography) FA and its derivatives and metabolites can be easily determined. Because these compounds are electrochemically reducible and oxidizable, respectively, voltammetric techniques could be employed for their analysis. Utilization of the different electrodes have been described in literature sources, e.g., modified carbon electrodes [17, 18], multi-walled carbon nanotube-modified gold electrodes , single-walled carbon nanotube–ionic liquid paste electrode [18, 20], mercury electrodes , as well as amalgam electrodes [22, 23]. MTX can be analyzed non-electrochemically (e.g., using HPLC ) as well as electrochemically using different electrodes (e.g., amalgam or boron-doped diamond electrodes) [24, 25] too. Electrochemical behavior and determination of LV was described in details many years ago using dropping mercury electrode (DME)  and all of the following works deal also with the utilization of mercury  or silver solid amalgam electrodes (AgSAE) .
LV reaction mechanisms were described in detail in, e.g., [26, 27, 28]. Three oxidation signals were recorded on DME . Heyrovský et al.  observed a peak pair (anodic peak at about − 800 mV, cathodic peak at about − 950 mV) on hanging mercury drop electrode (HMDE). Two voltammetric signals corresponding to the oxidation of tetrahydropteridine ring were registered at potentials of − 150 and of 0 mV. The oxidation products, which are adsorbable at the electrode surface, can be reduced at about − 400 mV. This signal was successfully used for LV determination on HMDE  as well as on AgSAE .
All the above-mentioned voltammetric methods of LV determination are based on its reduction. In the present paper, electrochemical oxidation of LV was studied and the procedure of LV determination on boron-doped diamond electrode (BDDE) was developed. Electrodes based on boron-doped diamond film have been so far successfully applied in the voltammetric analysis of various biologically active compounds, e.g., [29, 30, 31, 32]. In the past, there was published the determination of FA  and MTX  on a bare BDDE using differential pulse voltammetry (DPV). Therefore, this paper focuses on development and verification of an electrochemical method of LV determination on this electrode too. Optimum conditions for DPV determination of LV were found and this sensitive method was tested by analysis of LV in a commercially available pharmaceutical preparation.
Results and discussion
Voltammetric behavior of leucovorin in dependence on pH
It is obvious that the cathodic signals were much smaller than the anodic ones (Fig. 2) independently on tested pH of the supporting electrolyte. Therefore, the anodic signals seemed to be more suitable for analytical purposes. Moreover, peaks 1 and 2 were much higher than signal 3; therefore, we paid attention to them in all subsequent studies. From the evaluation of the dependences of anodic peak heights (Ip) on pH of the supporting electrolytes (Fig. 2, inset), it could be concluded that the highest signal was recorded in BRB of pH 2.0 for both peaks (BRB was used as the supporting electrolyte in the pH range from 2.0 to 12.0 and the H2SO4 solution as the supporting electrolyte of pH 1.0). On the other hand, the repeatability of the anodic peak located at about + 900 mV was not sufficient in this medium (relative standard deviation (RSD) of Ip values evaluated from 11 repeated measurements of 50 µmol dm−3 LV achieved 23%). Therefore, BRB with pH of 5.0 was used for the following experiments focused on the voltammetric behavior of LV in dependence on scan rate. Furthermore, the attention has been paid to the finding a suitable supporting electrolyte pH during the optimization of DPV again.
The influence of scan rate on voltammetric behavior of leucovorin
Determination of leucovorin in model solutions
Repeatability of DPV measurement of 10 µmol dm−3 LV in dependence on pH
161.4 ± 8.8
132.8 ± 1.7
107.09 ± 0.62
64.53 ± 0.83
Statistical parameters of LV concentration dependences registered under conditions given in the legend for Fig. 7
Slope/nA dm3 µmol−1
16.389 ± 0.062
0.61 ± 0.42
17.19 ± 0.25
0.55 ± 0.42
17.201 ± 0.084
0.089 ± 0.086
16.392 ± 0.03
5.4976 ± 4.1
Results of five repeated LV determinations in model BRB solutions
10.10 ± 0.17
3.030 ± 0.035
0.3000 ± 0.0029
Determination of leucovorin in pharmaceutical preparation
It was confirmed that a bare BDDE, as a working electrode, could be used for voltammetric detection and determination of LV based on its electrochemical oxidation. BRB, particularly of pH 3, proved to be suitable supporting electrolyte. Using either CV or DPV, two anodic and two cathodic significant and well developed voltammetric LV peaks (at about + 850 and + 1450 mV) and one pair of small and hardly evaluable peaks (at about + 150 mV) could be recorded. Finally, the DPV anodic peak located at about + 850 mV was found to be suitable for analytical purposes. Its height was the most sensitive to LV concentration changes, it was the best developed and reproducible under optimized conditions. The highest and simultaneously the most reproducible peak was recorded in BRB of pH 3.0, which was chosen for all other analysis. The DPV method was applied for determination of LV in deionized water (linear dynamic range from 0.15 to 25 μmol dm−3, LOQ 0.050 μmol dm−3, and LOD 0.015 μmol dm−3). Similarly, determination of LV in the commercial pharmaceutical preparation “LEUCOVORIN CA LACHEMA 10” was found to be successful considering the achieved results, which were consistent with the declared LV content (recovery 98.7–102.8%).
It could be concluded, that our proposed method represents simple but very precise and sensitive tool for determination of the important bioactive compound LV in the pharmaceutical samples. It is the first voltammetric method for LV determination based on its oxidation and simultaneously the first described method using non-mercury working electrode.
The 1 mmol dm−3 solution of LV was prepared by dissolving of the appropriate amount of calcium folinate, European Pharmacopoeia (EP) Reference Standard (Sigma-Aldrich, Czech Republic) in distilled water and stored in the dark at + 4 °C. The analyzed solutions were prepared daily fresh by dilution of the BRB stock solution.
All chemicals used to prepare stock solutions and basic electrolytes were of p.a. purity. BRB of pH values from 2.0 to 12.0 were prepared from an alkaline component of 0.2 mol dm−3 NaOH and an acidic component consisting of 0.04 mol dm−3 H3PO4, 0.04 mol dm−3 H3BO3, and 0.04 mol dm−3 CH3COOH (all these chemicals Lachema, Czech Republic). Solutions of H2SO4 were prepared by dilution of concentrated 96% H2SO4, p.a. (Ing. Petr Švec-PENTA, Czech Republic) by deionized water. Deionized water (conductivity < 0.05 µS cm−1) produced by Milli-Q-Gradient, Millipore, Prague, Czech Republic, was used for all described measurements.
The pharmaceutical preparation in powder form for injection solution preparation “LEUCOVORIN CA LACHEMA 10” was purchased from Pliva-Lachema, Brno. Declared content of calcium folinate pentahydrate was 12.7 mg (corresponding to 10 mg of LV in 1 cm3 of prepared injection solution). Moreover, this preparation contained sodium chloride (10 mg) and sodium hydroxide (8 mg).
The Eco-Tribo Polarograph (Polaro-Sensors, Czech Republic) controlled by POLAR.PRO software (version 5.1, Polaro-Sensors, Czech Republic) and by Multielchem software (version 3.1, J. Heyrovský Institute of Physical Chemistry of the Czech Academy of Sciences, Czech Republic) was used for voltammetric measurements. They were carried out in a three-electrode arrangement where commercially available BDDE (Windsor Scientific, UK, active surface area of 7.07 mm2, inner diameter of 3 mm, resistivity of 0.075 Ω cm with a B/C ratio during deposition 1000 ppm) was used as a working electrode. A saturated argent chloride electrode (Ag|AgCl(KCl), sat.) served as a reference electrode and a platinum wire (diameter 1 mm) (both Monokrystaly, Czech Republic) served as an auxiliary electrode.
Accumet pH-meter AB150 (Fisher Scientific, Czech Republic) was used for the pH measurements. All realized experiments were performed at laboratory temperature (23 ± 2 °C).
At the beginning of every series of measurements, BDDE was activated in 0.5 mol dm−3 H2SO4 solution by insertion of − 1000 mV for 60 s and of + 2000 mV for 60 s. Then, the electrode surface was rinsed with deionized water. Subsequently, 20 cyclic voltammograms were realized in the potential range from − 1000 to + 2000 mV. A positive regeneration potential (Ereg) of + 2000 mV for a regeneration time (treg) of 5 s was inserted on the used BDDE before the start of each measurement. This step provided the O-terminated surface of the BDDE for the realized measurement and, at the same time, ensured oxidation of the most of the impurities trapped on the electrode surface.
Elucidations of the supporting electrolyte effect (pH) (v = 100 mV s−1) and of the scan rate effect were realized using CV from Ein = 0 mV to Efin = + 2000 mV and reversely. Supporting electrolyte was represented either by the solution of H2SO4 (pH 1.0) or by BRB (pH 2.0–12.0). The dependence of cyclic voltammograms of LV (cLV = 5 × 10−5 mol dm−3) on the scan rate was investigated from 25 to 500 mV s−1 in BRB (pH 5.0).
DPV was applied with the following parameters (if not stated otherwise): Ein = 0 mV, Efin = + 1200 mV, v = 40 mV s−1, pulse height = + 50 mV, pulse width = 20 ms (where the current values were registered in next 20 ms), BRB of pH 3.0, which was chosen based on the study, where supporting electrolyte of pH from 1.0 to 7.0 was employed.
The values of LOD and of LOQ were calculated as three times and ten times, respectively, a standard deviation of the blank solution divided by the calculated slope of the calibration curve . The parameters of the calibration curves (i.e., slope, intercept, correlation coefficients) were calculated and all of the graphical dependences were constructed using MS Excel 365 software (Microsoft, USA). All confidence intervals were calculated at the level of significance α = 0.05.
Pharmaceutical sample analysis
A commercially available pharmaceutical preparation “LEUCOVORIN CA LACHEMA 10” (in the powder form), representing a real sample of LV, was after dissolving analyzed using DPV. The declared content was 10 mg of LV per vial. The sample was prepared for analysis according to the manufacturer’s instructions. i.e., by dissolving of the vial content in 1 cm3 of distilled water and further diluted ten times. 10 mm3 of sample solution thus prepared was added to 10 cm3 of BRB (pH 3.0). All quantitative analyses were performed by the standard addition method (1 addition = 20 mm3 of the standard solution of 1 mmol dm−3 LV). The LV determination was repeated 5 times.
This work was supported by the grant project of the Czech Science Foundation (project no. 17-03868S) and by The University of Pardubice (projects nos. SGSFChT_2018_003 and SD373001/82/30350(2016)).
- 1.Murray RK, Bender D, Botham KM, Kennelly PJ, Rodwell VW, Weil PA (2012) Harper’s illustrated biochemistry. McGraw-Hill Lange, ColumbusGoogle Scholar
- 27.Stejskal D, Heyrovsky M (1994) Research report no. 1544-2. General Teaching Hospital, PragueGoogle Scholar
- 36.Meloun M, Militky J, Forina M (1992) Chemometrics for analytical chemistry, volume 1: PC-aided statistical data analysis, volume 2: PC-aided regression and related methods. Ellis Horwood, ChichesterGoogle Scholar