Retention characteristics of serotonin receptor ligands
The retention behavior of serotonin receptor ligands was defined on two types of stationary phases that differ in polarities. The C8-modified silica gel stationary phase provides the hydrophobic interactions within the chromatographic systems, while the PFP (pentafluoro-phenyl propyl) phase enables the additional presence of dipole and ion exchange interactions. The influence of temperature (20–30 ℃), the volume fraction of acetonitrile (40–50%), and the concentration of ammonium acetate (15–25 mM) on the retention factor (k) were defined according to Eq. (1). The most important statistical characteristics of the obtained MLR models are given in Table S1 of the Supplementary material.
The volume fraction of acetonitrile (a) showed the most important influence on the retention behavior, and as the fraction of acetonitrile increases, the interaction between the compound and the stationary phases decreases. The temperature (t) influenced the separation mechanism within the PFP phase, while the effect of ammonium acetate (b) was more emphasized within the C8 phase, Table S1.
The differences in retention behavior between PFP and C8 stationary phases were graphically presented in Fig. 2. For olanzapine, mianserin, ziprasidone, and quetiapine there is a visible change between their interactions with the PFP and C8 phases. In these cases, the acidity of the mobile phase (concentrations of ammonium acetate (B)) showed a significant effect on the retention behavior, which can be expected considering the more emphasized basic characteristics of the tested compounds. In contrast, no significant differences were found for aripiprazole retention between PFP and C8 columns (Fig. 2).
Selectivity performances
Knowing the chromatographic behavior under different conditions can be useful for optimizing the simultaneous separation of structurally related pairs such as ziprasidone/risperidone (Z/R), aripiprazole/ziprasidone (A/Z), aripiprazole/risperidone (A/R), olanzapine/mianserin (O/M), olanzapine/quetiapine (O/Q) and mianserin/quetiapine (M/Q). Their separation was examined in relation to the obtained selectivity parameter values α (α = k1/k2) shown in Table 2.
Table 2
Obtained α values of structurally related chemical pairs in the tested conditions.
A | T | B | C8 | PFP |
(v/v) | (℃) | (mM) | Z/R | A/Z | A/R | M/O | Q/O | M/Q | R/Z | A/Z | A/R | O/M | O/Q | M/Q |
40 | 20 | 15 | 1.65 | 1.95 | 3.20 | 2.11 | 1.47 | 1.43 | 1.29 | 1.71 | 1.32 | 1.31 | 1.50 | 1.15 |
50 | 20 | 15 | 1.33 | 1.63 | 2.18 | 1.60 | 1.07 | 1.50 | 1.48 | 1.46 | 0.98 | 1.54 | 1.86 | 1.21 |
40 | 30 | 15 | 1.59 | 1.92 | 3.06 | 2.00 | 1.44 | 1.40 | 1.34 | 1.66 | 1.24 | 1.35 | 1.47 | 1.09 |
50 | 30 | 15 | 1.23 | 1.71 | 2.10 | 1.57 | 1.13 | 1.40 | 1.54 | 1.41 | 0.92 | 1.48 | 1.79 | 1.21 |
40 | 20 | 25 | 1.43 | 1.94 | 3.26 | 2.28 | 1.67 | 1.37 | 1.29 | 1.30 | 1.69 | 1.22 | 1.64 | 1.35 |
50 | 20 | 25 | 1.33 | 1.67 | 2.23 | 1.66 | 1.00 | 1.66 | 1.51 | 1.44 | 0.95 | 1.48 | 1.93 | 1.30 |
40 | 30 | 25 | 1.63 | 1.90 | 3.11 | 2.08 | 1.43 | 1.45 | 1.24 | 1.59 | 1.28 | 1.14 | 1.50 | 1.31 |
50 | 30 | 25 | 1.34 | 1.64 | 2.19 | 1.67 | 1.01 | 1.66 | 1.57 | 1.47 | 0.93 | 1.21 | 1.86 | 1.54 |
40 | 25 | 20 | 1.36 | 2.00 | 3.15 | 2.02 | 1.43 | 1.41 | 1.36 | 1.72 | 1.26 | 1.29 | 1.53 | 1.18 |
50 | 25 | 20 | 1.34 | 1.63 | 2.18 | 1.72 | 1.05 | 1.64 | 1.62 | 1.48 | 0.91 | 1.45 | 1.85 | 1.27 |
45 | 20 | 20 | 1.48 | 1.80 | 2.66 | 1.90 | 1.22 | 1.56 | 1.44 | 1.56 | 1.08 | 1.40 | 1.74 | 1.24 |
45 | 30 | 20 | 1.45 | 1.76 | 2.55 | 1.84 | 1.20 | 1.53 | 1.47 | 1.54 | 1.05 | 1.34 | 1.68 | 1.25 |
45 | 25 | 15 | 1.48 | 1.84 | 2.74 | 1.82 | 1.29 | 1.41 | 1.34 | 1.53 | 1.14 | 1.39 | 1.67 | 1.21 |
45 | 25 | 25 | 1.53 | 1.76 | 2.69 | 1.97 | 1.21 | 1.63 | 1.33 | 1.49 | 1.13 | 1.28 | 1.63 | 1.27 |
45 | 25 | 20 | 1.49 | 1.72 | 2.57 | 1.89 | 1.23 | 1.54 | 1.44 | 1.45 | 1.00 | 1.35 | 1.67 | 1.24 |
45 | 25 | 20 | 1.49 | 1.73 | 2.58 | 1.89 | 1.23 | 1.53 | 1.45 | 1.45 | 1.00 | 1.35 | 1.67 | 1.24 |
45 | 25 | 20 | 1.48 | 1.73 | 2.57 | 1.88 | 1.23 | 1.53 | 1.45 | 1.45 | 1.00 | 1.36 | 1.68 | 1.24 |
A: volume fraction of acetonitrile in the mobile phase (v/v %); T: temperature (℃); B: concentration of ammonium acetate (mM); Inversion in elution order: marked in yellow. |
From the data summarized in Table 2, it can be concluded that the change in elution order for ziprasidone-risperidone, mianserin-olanzapine, and quetiapine-olanzapine correlates with the observed changes in retention mechanisms of these compounds (see Fig. 2). The C8 phase provided better separation compared to the PFP phase, especially for 4-arylpiperazine derivatives (ziprasidone, aripiprazole, risperidone). It can be seen that with a 40% acetonitrile fraction in the mobile phase, the separation of olanzapine, risperidone and quetiapine showed improved characteristics compared to the results obtained in a similar study [25]. According to obtained α values, the C8 system with the 40% fraction of acetonitrile can be used for the successful separation of the investigated compounds.
Separation of structurally related compounds
Reactivity of the alpha position of the indol moiety in the molecule of ziprasidone contributes to its instability and to the formation of impurities (oxy-ziprasiodne). Aripiprazole N-oxide is formed from aripiprazole by the cytochrome P450 (CYP) isoforms CYP2D6 and CYP3A4, but can also be found in commercial preparations as an degradation impurity. The monographs of ziprasidone and aripiprazole in the European Pharmacopoeia (EP) [35] and the United States Pharmacopoeia (USP) [36] as well as the manufacturers require testing of their oxidative products. Therefore, C8 separation mechanism was investigated as a basis for further analysis of structurally related, oxidative impurities. Figure 3 shows the influence of mobile phase composition (acetonitrile fraction (A), ammonium acetate concentration (B)) on the separation of structurally related arylpiperazines, ziprasidone and aripiprazole.
Better separation characteristics can be observed with lower fractions of acetonitrile and lower concentrations of ammonium acetate in the mobile phase (Fig. 3). The highest selectivity between ziprasidone and aripiprazole (α = 2.00) was achieved at 25℃ using the lower content limit of acetonitrile (40%, v/v) and 20 mM ammonium acetate as mobile phase modifier (Table 2). It has been found that under selected conditions, arylpiperazine-related compounds including aripiprazole, ziprasidone and their oxidative degradants (aripiprazole N-oxide, oxy-ziprasidone) can be successfully separated, Fig. 4.
Although the reported HPLC methods for estimation of ziprasidone, aripiprazole, and their impurities in pharmaceutical formulations present adequate validation criteria, they show a number of limitations including long chromatography time and selectivity characteristics between the active ingredient and its impurities. The obtained selectivity between ziprasidone and oxy-ziprasidone was improved by the developed method (α = 1.85) compared to UHPLC-DAD analysis (α = 1.12) in ref. [12]. In the case of the separation of aripiprazole and aripiprazole N-oxide, progress was also made compared to the HPLC analysis in ref. [19], and ref. [20].
Validation of a method for determination of arylpiperazines in pharmaceutical dosage forms
The suitability of the developed method for the determination of aripiprazole was confirmed by testing sensitivity, selectivity, linearity, accuracy, and precision, i.e. method performance qualification was performed. The calculated regression parameters and recovery value fulfilled the acceptance criteria for linearity and accuracy, respectively. The linearity of the analytical method was estimated within the concentration range of 0.8–1.2 mg mL− 1 of aripiprazole. Statistical analysis of the obtained data was performed by linear regression analysis. The results are presented in Table 3. As the obtained correlation coefficient (r) was 0.99901, it can be concluded that the calibration diagram was within the acceptance criteria of linearity.
Table 3
The calibration and validation data.
Concentraction (%, mg mL− 1 ) | 120 | 110 | 100 | 90 | 80 |
1.2 | 1.1 | 1.0 | 0.9 | 0.8 |
Peak area (mAUs) | 1825 | 1552 | 1325.2 | 1105 | 822.9 |
Equation | 0.4601 + 0.0004 Peak Area |
r = 0.99901 |
The accuracy of aripiprazole was investigated at the three concentration levels (80%, 100%, 120%) each injected in triplicate, and RSD was 1.12%. Additionally, the obtained results showed no interference of the excipients with the peak of interest, thus the proposed method is suitable for the quantitative determination of aripiprazole in the tablets. The repeatability of the devised method was checked by the replicate sample injections (n = 6) of the six individual preparations of aripiprazole hydrochloride (1.00 mg mL− 1). The obtained RSD = 1.11% for aripiprazole hydrochloride and aripiprazole N-oxide fulfilled the acceptance criteria (RSD < 2% for active ingredients) [37]. Robustness was tested for the deliberate variations of the mobile phase composition and the column temperature (Table 2). From the data summarized in Table 2, it can be concluded that the proposed analytical method remains basically unaffected by the aforementioned deliberate changes in analytical conditions. The results demonstrated that the content of aripiprazole in analyzed tablets Bipodis 5mg was 99.45% (RSD = 0.5%) which was found within the acceptance criteria 98.0–102.0% for active ingredients.
The sensitivity of the method was checked by determination of LOD and LOQ, respectively. In order to examine the possibility of applying the method as a basis for the simultaneous examination of structurally related arylpiperazines and their oxide derivatives, LOD and LOQ values were calculated for ziprasidone, aripiprazole, oxy-ziprazidone, and aripiprazole N-oxide (Table 4). The LOD and LOQ were defined as the minimum concentration at which the analyte can readily be detected and quantified, respectively. The S/N ratios of 3:1 and 10:1 are generally considered acceptable for the estimation of the LOD and LOQ values.
Table 4
The LOQ and LOD values for aripiprazole and its related compounds (ziprasidone, oxy-ziprasidone, aripiprazole N-oxide).
Compound | LOD (mg mL− 1) | LOQ (mg mL− 1) |
Aripiprazole | 0.000800 | 0.00160 |
Aripiprazole N-oxide | 0.000400 | 0.00070 |
Ziprasidone | 0.000100 | 0.00030 |
Oxy-ziprasidone | 0.000014 | 0.00005 |
So far, there are no data on chromatographic methods that can be used for the simultaneous analysis of aripiprazole, ziprasidone and their oxide impurities. The sensitivity for the determination of aripiprazole and aripiprazole N-oxide in the developed method showed better characteristics compared to the HPLC analysis in ref. [19]. The LOQ of ziprasidone in the present method was found to be 0.00010 mg mL− 1 (< 0.001 mg mL− 1), as well as LOD of ziprasidone was 0.00030 mg mL− 1 (< 0.00046 mg mL− 1) [38]. The validation parameters of the developed HPLC method in ref. [11] indicate a better sensitivity of the proposed chromatographic conditions in determining the oxy-ziprasidone impurity (LOD < 0.00002 mg mL− 1; LOQ < 0.00006 mg mL− 1).