Development and validation of GLC/FID- and GLC/MS- procedures of secnidazole determination by the methods of additions

Gas-liquid chromatography is widely used in the process of forensic toxicological examinations, but data about application of GLC with flame-ionization and mass-spectrometry detection for secnidazole determination in analytical toxicology are absent. Aim. To develop GLC/FIDand GLC/MS-procedures for the quantitative determination of secnidazole and carry out step-by-step validation of the procedures developed in the variant of the method of additions. Results and discussion. The chromatographic conditions has been chosen for secnidazole determination by the method of GLC in two variants of performance with flame-ionization and mass-spectrometry detection with the temperature program changing during the analysis from 70 °C to 250 °C or 320 °C. Retention times for secnidazole are 8.97 min and 11.74 min. To prove the possibility of application of the procedures proposed in further analysis their validation has been carried out in the variant of the method of additions. Such validation parameters as in-process stability, linearity, accuracy and precision have been estimated by model solutions. Experimental part. The GLC/FID-analysis: HP 6890 Hewlett Packard; НР-1 ø0.32 mm × 30 m, 0.25 μm; thermostat – 70 oС (3 min), 40 oС/min to 180 oС (2 min), 40 oС/min to 250 oС (3 min); injector – 280 oС; detector – 280 oС; volume rate of a carrier gas (helium) – 1.5 ml/min; split mode – 1 : 2. The GLC/MS-analysis: Agilent 6890N/5973N/7683; НР-5MS ø0.25 mm × 30 m, 0.25 μm; DB-17MS ø0.25 mm × 30 m, 0.15 μm; columns are connected sequentially through Deans switch; thermostat – 70 oС (2 min), 45 oС/min. to 210 oС, 6 oС/min to 320 oС (12.56 min); transfer line – 280 oС; ion source – 230 oС; quadrupole – 150 oС; electron impact – 70eV; 40-750 m/z; injector – 250 oС; splitless mode; inlet carrier gas (helium) pressure: 1st column – 26.06 psi, 2nd column – 19.30 psi. Conclusions. New procedures for the quantitative determination of secnidazole by the method of GLC/FID and GLC/MS have been developed. Their validation has been carried out, and acceptability for application has been shown.

Ключевые слова: секнидазол; газо-жидкостная хроматография; валидация Gas-liquid chromatography (GLC) with different types of detection is widely used in the process of forensic toxicological examinations for screening and confirming investigations -with the purpose of detection, identification and determination of analytes [1,2]. The method is applied in the analysis of 5-nitroimidazole derivatives [3][4][5], but data about application of GLC with flame-ionization (FID) and mass-spectrometry (MS) detection for secnidazole determination in analytical toxicology are absent.
The aim of our paper was to develop GLC/FIDand GLC/MS-procedures for the quantitative determination of secnidazole and carry out step-by-step validation of the procedures developed in the variant of the method of additions [10,11] in order to confirm their acceptability for further application in analytical toxicology.

Results and discussion
Since secnidazole was readily soluble and stable in aqueous solutions, distilled water was proposed by us for preparation of the reference and model solutions in developing GLC/FID-and GLC/MS-procedures for the quantitative determination of secnidazole.
Previously, the chromatographic conditions were chosen for secnidazole determination by the method of gas-liquid chromatography in two variants of performance with flame-ionization and mass-spectrometry detection with the temperature program changing during the analysis from 70 °C to 250 °C or 320 °C, respectively. The typical chromatograms of secnidazole are presented in Fig. 2 and 3.
The mass-spectrum of secnidazole obtained under the GLC/MS-conditions proposed is presented in Fig. 4.
To prove the possibility of application of the procedures proposed in further analysis their validation was in the variant of the method of additions [10,11].
Such validation parameters as in-process stability, linearity/calibration model, accuracy and precision (repeatability) were estimated by model solutions.
The validation provides application of the normalized coordinates: were: i. e. transition from the equation 2 , it allows to calculate the validation characteristics, which do not depend on the analyte and features of the method of analysis [12,13]. The secnidazole concentration in the model solution for the point of 100 % in the normalized coordinates was chosen as the concentration provided the "signal/noise" ratio at the level of 40 [10].
For normalization of the experimental data obtained the reference solution with the analyte concentration of was used. The analytical range D of the method application was 25-175 %; the number of concentration levels g equaled 7 in constant increments of 25 %.
Acceptability criteria for validation parameters were formed on the basis of systematic application of the "insignificance concept" [12,13] and proceeding from the value of the extreme uncertainty Δ As , which equaled 20 % for the method in analytical toxicology [1,14].
Acceptability criteria for validation parameters were calculated proceeding from the assumption that uncertainty of the analyte quantification in model solutions was insignificant as compared with the total uncertainty Δ As : Validation results. In-process stability of secnidazole in the model solution was verified by chromatographing the reference solution immediately and in 1, 12, 24 and 48 hours after its preparation, and the systematic error σ model stability was calculated and assessed (Tab. 1). In-process stability of secnidazole in model solutions satisfied the acceptability criteria for all periods of time.
To determine linearity/calibration model the model solutions 1-7 were analyzed within 1 run, the correlation coefficient , the rest standard deviation , as well as the absolute term a model were calculated and assessed (Tab. 2).
To estimate precision and accuracy the model so-   Table 1 The results of in-process stability verification for secnidazole in model solutions The values "found/given" were calculated and used to determine the confidence interval and the systematic error σ model MA , respectively: The values of the confidence interval and the systematic error were compared with the corresponding acceptability criteria.
The results obtained within one analytical run are presented in Tab. 3-4. The total results of validation allow making the conclusion about acceptable linearity, accuracy and pre-cision of GLC/FID-and GLC/MS-procedures for the quantitative determination of secnidazole in the variant of the method of additions. It gives us the possibility to recommend this procedure for further application in analytical toxicology with the purpose of development of methods of biological liquids analysis for the quantitative determination of secnidazole.

Experimental part
Secnidazole was of pharmacopoeial purity. Instrumentation and chromatographic conditions. The GLC/FID-analysis conditions were as follows: • device -HP 6890 Hewlett Packard; • column -НР-1 ø0.32 mm × 30 m, 0.25 μm, 100 % dimethylpolysiloxane; (4)  Table 3 The results of accuracy and precision verification of GLC/FID-procedure for secnidazole determination in the variant of the method of additions Weighing was carried out using a digital analytical balance АN100 (AXIS, Ukraine) with d = 0.0001 g.
The glassware satisfied ISO 648:2008 "Laboratory glassware -Single-volume pipettes", ISO 1042:1998 "Laboratory glassware -One-mark volumetric flasks", ISO 4788:2005 "Laboratory glassware -Graduated measuring cylinders", ISO 385:2005 "Laboratory glas- and "Guidelines for calibration in analytical chemistry" [15] was used throughout this study. Reference and model solutions (Scheme 1 and 2). The stock solutions 1, 2 and 3 (100 μg/mL) were prepared by dissolving 50.0 mg of secnidazole in distilled water, and the solutions were diluted to 500.0 mL with the same solvent. The reference solution (8 μg/mL) was prepared by diluting 4.00 mL of the stock solution 1 to 50.0 mL with distilled water. The stock solution 2 was diluted with distilled water to prepare the model solutions 1-7 with the concentrations of 2; 4; 6; 8; 10; 12 and 14 μg/mL, respectively.
The addition solution 1 (300 μg/mL) was prepa red by dissolving 60.0 mg of secnidazole in distilled water, and the solution was diluted to 200.0 mL with the same solvent. The stock solution 3 was diluted with distilled water to prepare the model solutions 8 -13 with the concentrations of 10; 10; 20; 30; 40; 40 μg/ml, respectively. The model solutions 8.1-13.1 were prepared by diluting 10.00 mL of the model solution 8 -13 to 50.0 mL with distilled water. To prepare the model solutions 8.2-13.2 10.00 mL of the model solutions 8-13 were mixed with 1.00 mL of the addition solution 1 and diluted to 50.0 mL with the solvent.
When carrying out experiments each solution (except the in-process stability study) was chromatogra-phed 3 times or more, as required, following the requirements to repeatability of peak areas S for replicate injections offered by us [10] -the relative standard deviation of the mean RSD nom calculated towards the nominal value of the peak area S nom should not exceed: where: S nom -is the mean peak area obtained when analyzing the model solution 1. The mean values were used in further calculations.

Conclusions
New procedures for the quantitative determination of secnidazole by the method of GLC/FID and GLC/MS have been developed. Their validation by such parameters as stability, linearity, accuracy and precision in the variants of the method of additions has been carried out, and acceptability for application has been shown.
Conflict of interests: authors have no conflict of interests to declare.