Systematic Evaluation of Chromatographic Parameters for Isoquinoline Alkaloids on XB-C18 Core-Shell Column Using Different Mobile Phase Compositions

Chelidonium majus L. is a rich source of isoquinoline alkaloids with confirmed anti-inflammatory, choleretic, spasmolytic, antitumor, and antimicrobial activities. However, their chromatographic analysis is difficult because they may exist both in charged and uncharged forms and may result in the irregular peak shape and the decrease in chromatographic system efficacy. In the present work, the separation of main C. majus alkaloids was optimized using a new-generation XB-C18 endcapped core-shell column dedicated for analysis of alkaline compounds. The influence of organic modifier concentration, addition of salts, and pH of eluents on chromatographic parameters such as retention, resolution, chromatographic plate numbers, and peak asymmetry was investigated. The results were applied to elaborate the optimal chromatographic system for simultaneous quantification of seven alkaloids from the root, herb, and fruit of C. majus.

Due to the durability, silica-based RP-18 stationary phases are widely used for HPLC separation of plant extracts [18,19]; however, RP chromatography of alkaloids is rather di cult because they may exist both as free bases and charged forms. Cationic forms strongly interact with residual silanol groups of the RP-type stationary phase and cause the occurrence of the dual retention mechanism (RP and ion-exchange retention mechanism) and result in the peak tailing, irreproducible retention, and poor system efciency [20]. Due to the basic character of isoquinoline alkaloids, it would be preferable to conduct chromatographic separation at alkaline pH to avoid their ionization; however, silica-based adsorbents are unstable at this condition [21].
Di erent approaches may be used to eliminate these problems. Alkaline additives to mobile phases, mostly organic amines such as diethylamine, triethylamine, or dimethyloctylamine, are applied to suppress the ionization of the analyte and as silanol blockers [14,15]. Addition of anionic ion-pairing reagents, for example, sodium dodecyl sulphate or salts (e.g., ammonium acetate, ammonium formate, and sodium phosphate) is also used to improve the chromatographic separation [5,13,17,22]. On the other hand, the silanol-masking e ect may be achieved by additional modi cation of the sorbent surface, for example, endcapping [23,24].
XB-C18 sorbent is relatively new column lling with trimethylsilane endcapping and additional isobutyl chains. In the present work, an XB reversed-phase column was used to separate the isoquinoline alkaloids typically found in the C. majus extract. e in uence of organic modi er concentration, addition of salts, and pH of eluents on chromatographic parameters such as retention, resolution, chromatographic plate numbers, and peak asymmetry was investigated. e results were applied to elaborate the optimal chromatographic system for simultaneous quantication of alkaloids from the root, herb, and fruit of C. majus.
Chromatograms were recorded in the range of wavelength from 220 to 400 nm. e identity of compounds in plant extracts was con rmed by comparison of retention times and spectra with corresponding standards. Peak homogeneity was established comparing the spectrum recorded at the three peak sections upslope, apex, and downslope with the reference spectrum. Additionally, the chromatographic fractions eluted at the retention time characteristic for the investigated alkaloids were collected using a Foxy R1 fraction collector (Teledyne Isco, Lincoln, USA), and their identity was con rmed by direct injection mass spectrometry (micrOTOF-Q II, Bruker Daltonics, Bremen, Germany) using Compass DataAnalysis software version 4.1.

Sample Preparation.
e extraction conditions were based on literature [13]. e root, leaf, and fruit of C. majus (1 g) were extracted in ultrasonic bath (3 × 15 min) with 10 mL of methanol acidi ed with 0.05 M HCl. Subsequently, the extracts were combined, evaporated to dryness, and dissolved in 20 mL of methanol.

Results and Discussion
Su cient resolution between neighbouring peaks, symmetric peaks, and narrow peaks are the most important for the optimal chromatographic system. A stationary phase and a mobile phase have a crucial impact on these parameters. In our work, di erent variants of eluent compositions were tested for their suitability in HPLC of isoquinoline alkaloids on a new-generation XB-C18 endcapped core-shell column. e in uence of the three variables: concentration of the organic modi er, salt, and pH on resolution (R S ), peak asymmetry (A S ), and system e cacy (N-theoretical plate numbers) for methanol/water and acetonitrile/water solvent systems was investigated.

Optimization of Chromatographic Condition.
e chromatographic parameters were established in the range   Taking into consideration the resolution, the total separation of investigated compounds was obtained for concentration of 20% ACN in the whole tested pH and salt concentration range. At 25% of ACN, the compound All/Che (at pH ≥ 4) or Che/Cop (at pH ≤ 4) partially coeluted. e exemplary R s values are given in Table 1.
In higher concentrations, the majority of compounds were eluted below 10 min (k values for protopine, allocryptopine, and chelidonine were lower than 1), and the resolution was poor (Table S1). e concentration of ammonium acetate and pH also a ected the alkaloid retention; at lower pH values and at higher salt amounts, retention times were shortened. Moreover, e ciency of the system (N), symmetry of peaks (A s ), and resolution (R s ) strongly depended on these variables. N, A s , and R s values versus pH and salt concentration are presented in Figure 1 and Figure S1.
As can be seen, at pH 5 and at concentration of salt 5 mM, theoretical plate number decreased and peak asymmetry increased signi cantly, and it resulted in the peak broadening and decreased peak resolution. In contrast to salt concentration, pH had a major impact on R s . Resolution between Che/Cop, Cop/Sang, and Sang/Berb increased at lower pH; in turn, for All/Che, the opposite e ect occurred. Moreover, the change of elution order was observed for chelidonine and coptisine at pH 5.
Methanol showed lower elution strength, and the concentration in the range of 30-35% was required to obtain the elution of alkaloids at a reasonable time. Moreover, the order of elution strongly depended on pH and amount of organic modi er and salt ( Table 2 and Table S2).
As can be seen, no composition of mobile phases provided su cient separation within compounds with weaker retention such as protopine, allocryptopine, chelidonine, and coptisine, and all tested chromatographic parameters were worse for methanol/water than for acetonitrile/water eluents.
In further experiments, ammonium acetate/acetic acid in ACN/water eluents was replaced by ammonium formate/formic acid; however, it had a minor impact on chromatographic parameters. e R s values and retention times did not di er signi cantly, and only a slight increase in system e cacy (narrower peaks) was observed. Exemplary chromatograms of the standard mixture obtained at various mobile phase compositions are shown in Figure 2.   (1) protopine, (2) allocryptopine, (3) chelidonine, (4) coptisine, (5) sanguinarine, (6) berberine, and (7) chelerythrine. e comparison of N and A s for various mobile phase compositions is presented in Figure 3.
Based on the obtained results, acetonitrile at concentration of 20% and water at pH 3-4 with addition of 10-20 mM ammonium acetate or ammonium formate were considered as optimal for isoquinoline alkaloid separation on an XB-C18 core-shell column.
A lot of chromatographic systems for RP separation of isoquinoline alkaloids in C. majus were described in literature; however, most of them were more complicated [14,15,22] or did not provide su cient separation for quantitative analysis [25]. Due to additional modi cation of endcapped octadecyl silica by isobutyl chains in the XB-C18 stationary phase, the interaction of the basic analyte with residual silanol decreased signi cantly. It allowed to conduct chromatographic separation using milder pH and lower amount of salt addition compared to eluents proposed in literature [13], and it is bene cial for the HPLC system.

Chromatographic Analysis of C. majus Extracts.
Chromatography of the C. majus extract was carried out using the mobile phase consisting of ACN (solvent A) and 10 mM water solution of ammonium acetate adjusted to pH 4 with acetic acid (solvent B) (20 : 80, v/v). High R s values between coptisine and sanguinarine allowed to use the simple  348.123 gradient program to shorten the total time of analysis. After 20 min, the elution strength of the mobile phase was increased to accelerate elution of strongly retained sanguinarine, berberine, and chelerythrine. e gradient program was as follows: A 20% and B 80% during 0-20 min, A 25% and B 75% during 20-27 min, and A 30% and B 70% during 27-40 min. e obtained chromatograms are presented in Figure 4. e data used for identi cation of the investigated compounds are given in Table 3.
e results of quantitative determination of isoquinoline alkaloids in the root, leaf, and fruit of C. majus are given in Table 4, and validation parameters are summarized in Table S3.

Conflicts of Interest
e authors declare that there are no con icts of interest regarding the publication of this paper. Table S1: the comparison of retention times and peak resolutions of investigated alkaloids in 30% of acetonitrile in water at di erent pH and ammonium acetate concentration. Table  S2: the comparison of retention times and peak resolutions of investigated alkaloids in 30% of methanol in water at di erent pH and ammonium acetate concentration. Table S3: calibration data for quanti cation of investigated alkaloids. Figure S1: the relationship between theoretical plate numbers (N), peak asymmetry (As), resolution (Rs) and pH/ammonium acetate concentration. (Supplementary Materials)