Separation of 9-Fluorenylmethyloxycarbonyl Amino Acid Derivatives in Micellar Systems of High-Performance Thin-Layer Chromatography and Pressurized Planar Electrochromatography

The problems with separation of amino acid mixtures in reversed-phase mode are the result of their hydrophilic nature. The derivatisation of the amino group of mentioned above solutes leads to their solution. For this purpose, 9-fluorenylmethoxycarbonyl chloroformate (f-moc-Cl) as the derivatisation reagent is often used. In our study, the separation of some f-moc- amino acid derivatives (alanine, phenylalanine, leucine, methionine, proline and tryptophan) with the use of micellar systems of reversed-phase high-performance thin-layer chromatography (HPTLC) and pressurized planar electrochromatography (PPEC) is investigated. The effect of surfactant concentration, its type (anionic, cationic and non-ionic) and mobile phase buffer pH on the discussed above solute migration distances are presented. Our work reveals that the increase of sodium dodecylsulphate concentration in the mobile phase has a different effect on solute retention in HPTLC and PPEC. Moreover, it also affects the order of solutes in both techniques. In PPEC, in contrast to the HPTLC technique, the mobile phase pH affects solute retention. The type of surfactant in the mobile phase also impacts solute retention and migration distances. A mobile phase containing SDS improves system efficiency in both techniques. Herein, such an effect is presented for the first time.


Results and Discussion
Effect of surfactant concentration. The effect of surfactant concentration in the mobile phase on the retention (HPTLC) or migration distance (PPEC) of f-moc amino acids is presented in Fig. 1a,b. In the chromatographic system, the presence of surfactant (Fig. 1a) affects several processes. The first is the interaction between solute, micelles and components of the eluent. The second is the activity between micelle and the stationary phase. This induces changes in the physicochemical character. With regard to the first process, the presence of surfactant in the mobile phase influences the formation of solute-micelle complexes within this part of the chromatographic system. Thus, depending on the surfactant concentration in the eluent, the solute is either dissolved in the interfacial part of the micelle at low surfactant level or it penetrates the micelle at high amphiphile content 17 . Hence, the interactions with the stationary phase affect the total separation process. Thus, the retention of solutes depends on the strength of their bonding in the stationary phase and their rate of dissolving in the mobile phase. It should be noted that in thin-layer chromatography, the micelle mobile phase can affect the demixing of eluent during plate development, resulting in two mobile phase fronts. In our experiment, the higher of these held a higher content of water-organic (aqueous-acetonitrile) solvent, whereas a higher concentration of surfactant was present in the lower phase front. What is more, in the lower part of the plate, the mobile phase was more viscous than in the higher. As the result of such behaviour, the eluotropic strength of the mobile phase differed below and above the first (lower) eluent fronts. This effect was described in 18 .
Regarding the HPTLC technique (Fig. 1a), adding SDS to the mobile phase reduces solute retention of f-moc amino acids. Such behaviour is the result of the surfactant, dissociated f-moc derivatives and silanol groups from the stationary phase holding identical charges. In the stationary phase system, the interactions between the non-polar part of the SDS molecules and the alkyl chain of C-18 sorbent are hydrophobic and result in the increasing polarity of the sorbent. Thus, the solute associated with the SDS micelle (solubilised in its interfacial part) and the change of sorbent character deteriorates retention. Herein, the smallest retention is observed at 100 mM of SDS in the mobile phase. Nevertheless, this fact is not associated with improvement of solute zones separations. Thus, the best zone separation was observed when 45 mM of SDS was present in the eluent. We also discovered that the surfactant concentration affected the compound retention order. Furthermore, this varied for some f-moc derivatives at 5 and at 15 mM of SDS. According to our results, the migration order of the f-moc derivatives for 45 mM SDS starting from the compound of the strongest retention are as follows: f-moc Met < f-moc Trp < f-moc Leu < f-moc Phe < f-moc Pro < f-moc Ala.
In the PPEC technique (Fig. 1b), contrary to HPTLC, we found that the increase of SDS concentration in the mobile phase slightly increased the migration distances. This effect results from the opposing direction of electrophoretic migration of the micelle-solute complex relative to the electroosmotic flow. Similarly to TLC, the best www.nature.com/scientificreports www.nature.com/scientificreports/ Effect of buffer pH. Since the interactions between investigated solutes, surfactant, stationary and mobile phases depend on ionization, we investigated the effect of mobile phase buffer pH on retention/migration distances. Two buffers of different pH were chosen for this purpose. The application of the first (pH = 2.5) resulted in the withdrawal of solute ionization. Thus, at this condition, interactions between neutral (hydrophobic) solute, surfactant and both stationary and mobile phases were made evident. The second buffer pH (7.0) is above the solute pK A (see Table 1). At such pH, the solutes are ionized, hence, their hydrophobicity decreased. The effect of the mobile phase buffer pH on the f-moc amino acid derivative migration distances in both techniques is presented in Table 2.
In TLC, the mobile phase buffer pH had insignificant influence on the retention of the majority of derivatives. Such behaviour is the result of the strong interactions between f-moc derivatives and the stationary phase being changed by surfactant addition. In comparison with the remaining solutes, f-moc Ala exhibited higher R F . This results from the effect of the weaker interactions with sorbent that comes about due to its aliphatic, more hydrophilic character and weaker dispersion forces (see log P, MW in Table 1). In our experiment, aromatic and more hydrophobic f-moc Phe and Trp showed the strongest retention (see Table 1). Application of the lowest pH buffer in the mobile phase, herein, led to poor resolution of f-moc Met and f-moc Pro zones, as well as f-moc Phe and f-moc Trp spots. In contrast, for the higher pH buffer, only two f-moc derivatives (f-moc Phe and f-moc Trp) did not separate. www.nature.com/scientificreports www.nature.com/scientificreports/ On using PPEC as the experiment technique, we noted that the mobile phase buffer pH had a significant effect on the migration distances of the investigated f-moc derivatives. Herein, the hydrophobic effect between neutral solute (at pH 2.5), the hydrophobic interior of SDS micelle and the stationary phase resulted in stronger interactions with the stationary phase. Similarly to TLC experiments, the physicochemical character of derivatives affects the length of their migration distances (f-moc Ala has the longest, f-moc Trp the shortest). We also saw poor f-moc-Phe and f-moc Leu zone separation. This came about due to the increase of buffer pH to 7.0 resulting in the ionization of f-moc derivatives and enhancement of their hydrophilicity. Such effect led to stronger interactions in the mobile phase and the elongation of migration distances. Electroosmotic flow (EOF) also contributed to the solute migration distance length in PPEC. Furthermore, we saw a stronger EOF at higher pH -as also noted by 19 . In addition, an increase the mobile phase buffer pH improved the separation of f-moc Phe and f-moc Leu, but worsened the separation of f-moc Met and f-moc Pro.

Comparison of surfactants.
We also investigated the effect of surfactant type (cationic, anionic, non-ionic) on solute retention, migration distances and selectivity. Tests were carried out using both techniques (TLC and PPEC). We established the composition of the mobile phase on the basis of previous experiments 20 . This was comprised of 45% acetonitrile in an aqueous buffer of pH 2.5. Such a buffer pH allows an investigation of the neutral form of f-moc derivatives (the solutes pK A are presented in Table 1). The following surfactants were applied during these experiments: Brij-35 (CMC in aqueous solution 0.1 mM), sodium cholate (CMC in aqueous solution 13-15 mM), hexadecyltrimethylammonium bromide (CTAB, CMC in aqueous solution 0.92 mM), sodium dodecyl sulphate (SDS, CMC in aqueous solution 8.1 mM) and tetramethylammonium chloride (CTMA, lack data of CMC in aqueous solution). Since both organic solvent (acetonitrile) and pH affects the surfactant CMC value, to be sure that the chosen micellar systems were applied, high concentrations of investigated surfactants were prepared. In this experiment, experiment time for HPTLC was 15 min, while for PPEC,the corresponding time was only 10 min (when the polarization voltage was 800 V). The results are presented in Table 3.
We saw that in both techniques, the surfactant type affects separation selectivity. This effect is due to interactions of all components of the system (tested substances, micelles and both the mobile and stationary phases). Regarding the HPTLC technique, the cationic surfactant of the longer hydrocarboneous chain (CTAB) enhances the solute retention, in comparison to the counterpart of this in the shorter chain (CTMA).This is the result of the CTAB system holding higher hydrophobicity in comparison to that with CTMA, due to the mentioned interaction (adsorption) of the polar head of cationic surfactant onto the negatively charged silanol group of the sorbent. Thus, mixed non-polar sorbent with C-18 and C-16 (from CTAB) hydrocarboneous chains is formed. Such mixed adsorbent brings about stronger retention of hydrophobic micelles containing f-moc amino acids. When CTMA is used, the above-mentioned interaction happens, but due to the C-1 group in CTMA, this affects only the free silanol group in the sorbent and does not protrude over on to the C-18 chain. Consequently, the hydrophobic interaction between the solute/CTMA micelle and the stationary phase is weaker, in comparison to the system containing CTAB. Nevertheless, the application of cationic surfactant produces little improvement in solute zone separation. Of note, the zones of Phe, Leu, Trp f-moc derivatives are not separated when the systems with the cationic surfactant are used.
Bearing in mind that the stationary phase interaction of anionic surfactants (sodium cholate and SDS) are only via hydrophobic interactions with the C-18 chain of sorbent, this result in changing the physicochemical character into that which is more hydrophilic. The bulky nonpolar cyclopenta[a]phenanthren group of sodium cholate additionally limits the specific interaction between sorbent and complex micelle-solute, and deteriorates the selectivity of solute separation (of note, the majority of f-moc derivatives are not, in fact, separated). In contrast, the flatter SDS molecule acts less obstructively on the above-mentioned specific interaction and engenders a better resolution of the solute zones. It also must be underlined that the solute retention in eluents containing SDS are more dissimilar.
Regarding the third surfactant type, non-ionic Brij-35, all possible interactions between all components of the system are naturally hydrophobic. Thus, this amfifile induces insignificant changes in the sorbent character. Consequently, a mobile phase inclusive of Brij-35 enhances retention of aliphatic and hydrophilic derivatives (f-moc Ala), while lessening the effect of more hydrophobic f-moc amino acids (Pro, Phe and Trp). Therefore, the zone separation is better than that for an eluent with sodium cholate, but poorer in comparison with that including SDS or cationic surfactants. www.nature.com/scientificreports www.nature.com/scientificreports/ In assessing solute migration distances, we noted that the order of an amphiphile inclusive system is as follows (longest to shortest): Brij-35 and CTMA > SDS > sodium cholate > CTAB. For the majority of tensides, this sequence is in accordance with increasing CMC values for aqueous surfactant solutions as presented at the beginning of this section. The exception is CTAB, while for CTMA, there is no such data.
As for PPEC, cationic surfactants (CTAB or CTMA) in the mobile phase shorten the solute migration distances, since in applied electromigrational techniques (i.e. capillary electrophoresis), they create sorbents more hydrophobic and/or according to literature, reverse the electroosmotic flow direction 21 . Furthermore, in PPEC, cationic surfactants reduce the flow of the mobile phase 22 . We also found that a system containing CTAB is unable to separate f-moc Met and f-moc Pro, and a mobile phase with CTMA cannot distinguish between f-moc Phe and f-moc Leu. Interestingly, the longest migration distance in the CTAB-doped eluent is that of f-moc Trp, while the shortest is that of f-moc Ala. Such behaviour is contrary to other surfactant included mobile phases.
We discovered that eluents containing anionic or non-ionic surfactant (SDS, sodium cholate and Brij-35) generate considerably longer solute migration distances (SDS > Brij-35 > sodium cholate > CTAB > CTMA). Regarding anionic surfactants (SDS), literature data states that these have positive outcomes on the PPEC mobile phase flow 23 . This leads to longer solute migration distances, and the SDS-doped eluents separate all investigated f-moc derivatives quite well. In contrast, sodium cholate produces a solute separation that is similar that in HPTLC (poor separation of f-moc Phe, f-moc Leu and f-moc Met), while Brij-35 inclusion results in similar migration distances as that of sodium cholate-doped systems, as it has insignificant influence on electroosmotic flow. However, a mobile phase containing Brij-35 has poor solute zone characteristics, and, consequently, f-moc Phe and f-moc Met zones are not separated.
Thus the PPEC technique may be applied for systems with surfactant different than SDS. It is additionally resulting in changing solute selectivity.

Statistical comparison of used techniques. Two f-moc amino acid derivatives(f-moc Ala and f-moc
Trp) were chosen for statistical comparison of TLC and PPEC. The results are presented in Table 4.
As indicated in our results, and in accordance with literature data, the migration distances of the investigated amino acid derivatives are longer in PPEC, than in TLC 19,[22][23][24] . However, after standard deviation and % RDS of results assessment, the above-mentioned technique is less reproducible. This phenomenon is characteristic of all electromigrational techniques.
Effect of surfactant presence on separation efficiency. The mixture separation process in liquid chromatography is the result of a solute partition between stationary and mobile phases. During it, the solute zones undergo a dispersion process observable on the chromatogram or electrochromatogram as peak broadening or tailing. Both effects have an effect on separation efficiency. The presence of surfactant in the mobile phase influences the latter process. In the mobile phase, the influence is the result of mass transfer kinetics during the separation process 25 . This effect in liquid chromatography systems has been investigated in 26,27 . In these works, the authors related peak asymmetry, measured as left and right peak half-widths, to the separation system efficiency. However, these papers cover only HPLC-micellar systems. Since the aforementioned parameters have influence on peak broadening (asymmetry factor, A s ) and peak tailing (peak tailing factor, T f ), thus to the derivation of data, we decided to compare the effect of surfactants on TLC and PPEC, using the model compounds: f-moc Ala and f-moc -Trp. The results are seen in Table 5.  www.nature.com/scientificreports www.nature.com/scientificreports/ For the system without surfactant, in TLC, both peak parameters (A S and T f ) for f-moc Ala and f-moc Trp significantly deviate from 1.0. Such behaviour indicates that mass transfer kinetics is disrupted. This effect affects peak broadening and/or tailing. Herein, solute interactions with the stationary phase have the essential effect that the solute of shorter migration distance (f-moc Trp) exhibits considerably higher A s and T f values when compared to the solute of longer migration distance (f-moc Ala). In PPEC without surfactant, A s and T f values are significantly less when compared to the TLC system results. Similarly, peak asymmetry and peak tailing factors are higher for the solute which migrates a shorter distance (f-moc Trp). On comparing separation efficiencies (H) for TLC and PPEC, the latter technique is better. Such behaviour is in accordance with literature data 19,22 .

Retardation factors (HPTLC technique)
The introduction of surfactant to the mobile phase significantly changes the property of the chromatographic system. In the TLC system and solute of shorter migration distance (f-moc Trp), A s and T f factors are lower when compared to the system without SDS. Moreover, for f-moc Trp, the deviation is close to 1.0. A similar effect is observed for f-moc Ala. In the PPEC technique, the inclusion of a surfactant in the eluent generated A s and T f values for the investigated f-moc derivatives that are close to ideal (1.0). Thus, the presence of SDS improves the shape of solute zones in both TLC and PPEC.
Hence, with regard to separation efficiencies (H obs values) for mobile phases containing surfactant, for both TLC and PPEC, the results are considerably better compared than eluent without SDS. However, the higher efficiency for f-moc Trp in a system without amphiphile and the effect of this on PPEC is worth further considering. As a general trend, the efficiencies for PPEC technique inclusive of eluent containing SDS are higher than that for TLC. The visual confirmation of above-mentioned thesis is presented on Fig. 2a-d. These Figs present densitograms (chromatograms and electrochromatograms) of the f-moc Trp detected at UV light of wavelength 262 nm. The zone shapes are better in PPEC technique when comparing to TLC one, especially for the mobile phase with surfactant (Fig. 2d) Table 5. Comparison of the surfactant effect on some chromatographic parameters in TLC and PPEC. Model solutes: f-moc Ala and f-moc Trp. The mobile phase: 40% acetonitrile, the aqueous universal buffer of pH 2.5. PPEC: polarisation voltage 1.5 kV, experiment time -15 min. TLC experiment time -17 min. *Peak asymmetry factor was calculated using the following equation: T f = b/a, where b is the distance from the peak midpoint (perpendicular from the peak highest point) to the trailing edge of the peak measured at 10% of peak height (left peak half-width) and a is the distance from the leading edge of the peak to the peak midpoint (perpendicular from the peak highest point) to the trailing edge of the peak measured at 10% of peak height (right peak half-width). **Peak tailing factor was calculated using the following equation: A s = (a + b)/2a, where b is the distance from the peak midpoint (perpendicular from the peak highest point) to the trailing edge of the peak measured at 10% of peak height (left peak half-width) and a is the distance from the leading edge of the peak to the peak midpoint (perpendicular from the peak highest point) to the trailing edge of the peak measured at 10% of peak height (right peak half-width). Both peak tailing and asymmetry factor formulas were found in 30 . ***Separation efficiency, the height of the theoretical plate, has been calculated using the following equation Where σ is the half of peak width at 0.607 height and Z x is solute zone migration distance 30 . www.nature.com/scientificreports www.nature.com/scientificreports/ experiments, 1.5 kV of polarization voltage was applied. The experiment time in both techniques lasted 15 min. Two test mixtures of f-moc amino acids were prepared. Mixture 1 contained f-moc derivatives of alanine, methionine, proline, tryptophan and citrulline (which was not investigated earlier; this amino acid derivative is more polar at log P = 1.95 (MervinSketch), in comparison with the other f-mocs). Mixture No 2 consisted of f-moc Ala, f-moc Leu, f-moc Phe, f-moc Pro and f-moc Cit. In the experiment, mixtures of the volume 4 μL were applied onto the sorbent with the use of ATS-4 applicator. The resulting chromatograms (TLC) and electrochromatograms (PPEC) are presented in Fig. 3a-d. The amino acid derivative zones were detected with the use of TLC scanner at 262 nm.
As indicated, the solute migration distances in PPEC are considerably longer in comparison with those from HPTLC. Additionally, the solute zones in HPTLC are quite well separated (compare Fig. 3a,c,b,d). What is more, calculating the peak resolution (R s ) value for both investigated techniques and two last solute zones, the higher values we obtained for PPEC technique. Considering TLC technique (Fig. 3a,b) and peaks 3 and 4 for both mixtures, the R s values are equal to 1.28 and 1.18, respectively. While for PPEC (Fig. 3c,d) peaks 4 and 5 for both above-mentioned mixtures the R s values are 1.49 and 1.42, accordingly The order of separated solutes is in accordance with their hydrophobicity decrease.

Conclusions
As a result of our experiments, we noted that the kind of surfactant applied and its concentration in the mobile phase has an influence on the f-moc amino acid migration distances in both HPTLC and PPEC. In HPTLC and PPEC experiments, eluents containing anionic surfactants (SDS or sodium cholate) are more convenient to use regarding the solute migration distances or retentions, in comparison with eluents containing cationic surfactants www.nature.com/scientificreports www.nature.com/scientificreports/ (CTAB, CTMA). This is because the latter have weaker interactions between the surfactant and the stationary phase. We also found that the concentration of surfactant in the eluent is a crucial factor affecting solute migration distances and the order of f-moc amino acids in both investigated techniques. In addition, depending on the technique used, the effect of the mobile phase buffer pH on retention, migration distance and zone separationof solutes vary. In HPTLC, the pH has insignificant influence on the solute retention of the majority solutes but affects the solute zone separations. In PPEC, pH affects both the solute migration distances and zone separations. Application of the PPEC technique also leads to significant improvement of the solute zone separations. We noted that the presence of surfactant in the mobile phase in the HPTLC technique results in two eluent migration fronts. Such a problem is eliminated in PPEC due to the pre-wetting process. Regarding a statistical comparison of these techniques, standard deviation and % RDS of the results indicate that PPEC is less reproducible. Still, introducing surfactant to the eluent results in a diminution of peak broadening in both techniques. What is more, the peak asymmetry and tailing factor values are lower for PPEC, in comparison to that of HPTLC. Of note, the efficiencies of separation of HPTLC are worse than that of PPEC. Therefore, due to its benefits, PPEC can be successfully applied, together with a micellar mobile phase, to separate mixtures.

Methods
Instrumentation. Prototype PPEC equipment was constructed in the Department of Physical Chemistry, Medical University of Lublin (Lublin, Poland). Its details are described in 24 . Horizontal DS-II-10 × 10 chambers for TLC were purchased from CHROMDES (Lublin, Poland). The automatic sampler ATS-4 and TLC scanner were provided by CAMAG (Muttenz, Switzerland). Preparation of chromatographic plates. Commercially available HPTLC RP-18W plates (10 × 20 cm) were cut using a chromatographic plate cutter to a size of 10 × 10 cm. For the PPEC experiments, the plate edges were impregnated with a solution of Sarsil W and hardener (ratio of 96:4, one layer) and, subsequently, with a solution of Sarsil H-50 and hardener (ratio of 96:4; three layers). The painted margins were about 5 mm of width. After applying all layers, the plates were dried in an oven at 100 °C for 60 min, and were then transferred to a desiccator for cooling. Next, they were washed with methanol in a horizontal DS-chamber. After the methanol evaporated, the plates were activated in the drying chamber at 100 °C for 15 minutes, following which they were placed again into the desiccator. The plates for the TLC experiments were cleaned with methanol, activated in the drying chamber at 100 °C for 15 minutes and placed into the desiccator.

Application of samples.
For the PPEC experiments, 2 µL of each f-moc amino acid derivative solution was applied 15 mm from the lower edge of the previously prepared electrochromatographic plate, at a zone with a length of 6 mm, utilizing the automatic sampler ATS4. The spot application began 15 mm from the right edge and finished 15 mm from the left edge of the plate. For the TLC experiments, the above-mentioned volume of each solution of f-moc amino acid derivative was applied 5 mm from the lower edge of the previously prepared chromatographic plates. The remaining preparation practices were similar as that in PPEC.

Development of chromatograms and electrochromatograms.
Chromatograms were developed at room temperature, using a Horizontal DS Chamber for TLC, model DS-II-10 × 10. The sorbent layer was first equilibrated by standing 15 min inside the mobile phase vapour before the plate development. The distance of the eluent was 45 mm from the application line. The development of electrochromatograms using PPEC equipment is described in 24 . Before development, the electrochromatographic plates were prewetted for 1 min with the mobile phase. The polarization voltage used varied from 800 to 1500 V depending on the experiment. Consequently, PPEC experiment times were also distinct.
Solutes detection. F-moc amino acids zones on the chromatographic and electrochromatographic plates were identified with the use of TLC scanner (CAMAG, Muttenz, Switzerland) at 262 nm.
Ethical approval. This article does not contain any studies with human participants or animals performed by any of the authors.