The use of polyoxyethylene (20) cetyl ether in assessing the hydrophobicity of compounds of biomedical importance and in the process of drug release from microemulsions

2.1 Materials

Surfactants cholic acid (ChA), polyoxyethylene (4) lauryl ether (Brij-30), polyoxyethylene (23) lauryl ether (Brij-35), polyoxyethylene (20) sorbitan monooleate (Tween-80), Tyloxapol, poly(propylene glycol) block-poly(ethylene glycol)-block-poly(propylene glycol) (Pluronic-31R1) and promethazine hydrochloride (PMT) were purchased from Sigma-Aldrich, USA. Polyoxyethylene (20) cetyl ether (Brij 58) was obtained from SERVA FEINBIOCHEMICA GmbH & Co. KG, Heidelberg, Germany. Polyethylene glycol tert-octylphenyl ether (Triton-X-100) was purchased from Ferak, Berlin, Germany.

2.2 Conditions of chromatographic experiments

Tryptophan (TP), phenylalanine (PA), caffeine (Caff), thiamine hydrochloride (B₁), pyridoxine hydrochloride (B₆), riboflavin (B₂), acetylsalicylic acid (AS), saccharin (SA), barbital (BR), phenobarbital (PB), benzobamyl (BM), hexamydin (HD), 4-aminobutyric acid (GABA) and 3-(3,4-dihydroxyphenyl)-DL-alanine (DOPA) were selected as model compounds.

The work was performed on a high-performance liquid chromatograph “Millichrom-4” (Nauchpribor, Oryiol, Russia), equipped with an ultraviolet detector (wavelength range 190 nm–360 nm) and an automated control system. A chromatographic column (62 × 2 mm) was packed with Kromasil-C₁₈ stationary phase with particle size 5 µm. The detection wavelength was set at 220 nm and 260 nm. The mobile phase was a micellar solution of a micelle-forming non-ionic surfactant, polyoxyethylene (20) cetyl ether or Brij-58, in phosphate buffer. The mobile phase was prepared on the basis of 0.05 M Na₂HPO4, the pH of which was adjusted to pH 7 with phosphoric acid. Brij-58 in the concentration range of 1–5 % was added to the above buffer solutions to elute the samples. Cholic acid powder was added to the Brij-58 micellar solution of a certain concentration to obtain its solutions of (1, 5, 44 and 100) mM. The solutions obtained were stirred until complete dissolution of the cholic acid. The mobile phase flow rate was 50 µl/min. The dead volume of the column was calculated from the first deviation from the baseline after injection of the sample by introducing a dilute sodium chloride solution into the column.

2.3 Calculating the hydrophobicity index

To determine the hydrophobicity index of substances, Equation (1) is used [39]:

where k is the sample retention factor, [M] is the concentration of the micellar solution, k _w is the sample retention factor in pure water at [M] = 0, K _SM is the sample-micelle binding constant.

The concentration of micellar solution is calculated by Equation (2):

where [C] is the surfactant concentration, CMC is the critical micelle concentration of surfactant, and N is the micelle aggregation number.

Graphs are drawn in the coordinates “1/k − [M]”. By extrapolating the obtained graphs, the value of the hydrophobicity index (k _w) is found.

2.4 Conditions for the release of drugs from microemulsions

Membrane microemulsions were prepared by dissolving a certain amount of Brij 58, Tyloxapol, Pluronic-31P1, Brij-35, Brij 30, Tween-80 in water. In the case of mixed microemulsions, certain amounts of both surfactants were dissolved in water. The drug-loaded microemulsion was placed in a cellulose dialysis bag (64 mm × 10 mm), which was then suspended in 50 ml of phosphate buffer (0.075 M Na₂HPO₄ and 0.025 M NaH₂PO₄, pH 7.4) at T = 30 °C for the drug release experiment. The system was stirred continuously using a magnetic stirrer (INTLLAB). Samples of 1 ml were taken after (15, 30, 45, 60, 75, 90, 120, 150, 180, 210 and 240) min. An equal volume of phosphate buffer was added to the receptor solution after sample collection to maintain the dilution volume. The samples were treated with 0.75 ml of 0.5 % sodium persulfate (Na₂S₂O₈) to oxidize them and then left for 30 min to stabilize the intensity of the reddish-pink color [40]. The concentration of the drug released, promethazine hydrochloride, was measured on a photocolorimeter at a wavelength of 490 nm.

2.5 Calculation of kinetic parameters of drug release

In order to estimate the kinetics of drug release, we processed the obtained results on the basis of the Korsmeyer-Peppas model with Equation (3) [22, 24]:

where M _t/M is the fraction of drug released in time t; M is the amount of drug included in the dialysis bag; M _t represents amount of drug released in time t; k is kinetic constant; n is the diffusion exponent.

The fraction of drug released in time t, M _t/M, is converted to %. This is done by multiplying the right side of the equation M t M = k t n by 100 and then then taking the logarithm to obtain the linear expression (Equation (4)):

The dependence of ln ( M t M , % ) on ln t is constructed; n and 100 k are calculated from the slope of the line in the x-axis and the point of intersection in the y-axis, respectively.

3.1 Elution with Brij-58 micellar mobile phases: evaluation of hydrophobicity of solutes

The chosen non-ionic surfactant Brij-58 is compatible with the chromatographic system, i.e. it does not contain chromophoric groups and therefore it has a low absorption in the UV region and can be used for both isocratic and gradient elution. In addition, Brij-58 has a high cloud point (>100 °C). It will not precipitate on the column and will not interfere with the passage of the mobile phase through the column. In addition, Brij-58 is non-toxic and has a similar structure to that of Brij-35, in fact they have the same HLB value (16.9). It can be assumed that the stationary phase modified with Brij-58 monomers, as in the case of Brij-35, can be considered as an artificial model of the cell membrane [16, 17].

The chromatographic behavior of biologically and pharmacologically important model compounds was studied using mobile phases containing increasing concentrations of Brij-58 to assess their hydrophobicity. It should be noted here that the mobile phase containing the lowest concentration of Brij-58 that was used for elution was a 1 % Brij-58 solution. This means that the molar concentration of such a solution is 8.9 · 10⁻³ mol/l, which is well above the critical micelle concentration of Brij-58 (CMC = 0.08 · 10⁻³ mol/l). The samples are then eluted through micellar mobile phases.

The micellar mobile phase contains both micelles and a pure liquid or solvent (in our case, water). Therefore, in micellar chromatography, the sample molecules are not only distributed between the mobile and stationary phases, but also between the liquid and the micelles in the mobile phase, i.e. the distribution of substances in micellar chromatography follows the generally accepted three-phase model [41, 42]. The chromatographic behavior of samples in biopartitioning micellar chromatography is determined by the dispersion forces of London-van der Waals attraction between the non-ionic micelle and the sample molecules. There is no electrostatic interaction between the sample and the micelle due to the use of a non-ionic surfactant [16].

The study of the effect of the concentration of Brij-58 on the retention factor of solutes shows that an increase in the concentration of Brij-58 in the mobile phase reduces the retention factor k of the samples. However k decreases relatively little for hydrophilic samples, whereas for hydrophobic solutes (HD, BM, PB) the decrease in k is more pronounced with an increase in Brij-58 concentration within the range of (1.5–5.0) % (Figure 1).

Figure 1:

Dependence of retention factor of the model compounds versus concentration of Brij-58 in buffer: BM (●), PB (∆), HD (■), BR (X), Caff (♦), AS (○).

It is generally accepted that in the micellar chromatography the surfactant acts as an organic modifier in reversed phase HPLC, i.e. an increase in the surfactant concentration is similar to an increase in the content of an organic solvent in a hydroorganic mobile phase [41].

In addition, Brij-58 is adsorbed on the stationary phase, interacts with the C₁₈ groups of the stationary phase with its hydrophobic chain, and is likely to further modify the stationary phase after CMC, contributing to a decrease in the retention factor of the samples [43].

In addition, the polyoxyethylene groups are oriented towards the water and increase the polarity of the stationary phase, which further reduces the retention of hydrophobic solutes. The retention factors of the model compounds in pure water, i.e. hydrophobicity indices, are shown in Table 1.

3.2 Elution with Brij-58 micellar mobile phases, modified with cholic acid

Cholic acid has been added to the Brij-58 micellar mobile phase in order to approximate the structure of the natural cell membrane. It should be noted that cholic acid is insoluble in water, although it dissolves well in the micellar solution of Brij-58, i.e. it is well solubilized in the hydrophobic core of micelles.

An increase in the concentration of Brij-58 in the presence of 1 mM cholic acid in the mobile phase causes a decrease in the retention factor of the model compounds (Figure 2), i. e. in this case the same trend is observed as without the addition of cholic acid.

Figure 2:

Dependence of the retention factor of the model compounds versus Brij-58 concentration in the presence of 1 mM cholic acid in buffer: BM (●), PB (∆), HD (■), BR (X), Caff (♦), AS (○).

However, the retention-concentration curves for solutes eluted with Brij-58 micellar mobile phases are located higher than those in the presence of cholic acid in the mobile phase and they are symbiotic. However, the distance between these curves is greater at low concentrations of Brij-58 (Figure 3). This can be explained by an increase in the surface polarity of the Brij-58 micelles due to the hydrophilic groups of cholic acid as a result of surface adsorption of its molecules on the surface of the Brij-58 micelles (Scheme 1), as well as an increase in the polarity of the stationary phase surface modified with Brij-58 monomers and few molecules of cholic acid (Scheme 2).

Figure 3:

Dependence of the retention factor of the model compounds versus Brij-58 concentration in buffer without cholic acid: PB (∆), HD (○), ES (□) and in the presence of 1 mM cholic acid in buffer: PB (▲), HD (●), ES (■).

Scheme 1:

Scheme of Brij-58 micelle modified with molecules of cholic acid.

Scheme 2:

Scheme of C₁₈-stationary phase surface, modified with Brij-58 monomers and with few molecules of cholic acid.

The calculated hydrophobicity indices for the model compounds in 1 mM cholic acid solution are given in Table 1. It should be noted that the order of the hydrophobicity values (k _w) is reversed for some pairs of samples after the addition of a fixed amount of cholic acid to the mobile phase, e.g. membrane permeation increases in the pairs caffeine (3.5) → saccharin (3.7), thiamine (1.4) → pyridoxine (1.7), gamma-aminoerboic acid (1.0) → phenylalanine (1.2) when eluted with the micellar mobile phase of Brij-58 (Table 1). This order is reversed when eluting with mobile phases with increasing concentrations of Brij-58 in the presence of a fixed concentration of cholic acid in the mobile phases: saccharin (1.6) → caffeine (2.9), pyridoxine (2.1) → thiamine (2.2), phenylalanine (0.9) → gamma-aminoerboic acid (1.2) (Table 1).

The results obtained indicate that the presence of cholesterol in the cell membrane affects the permeation of substances into the membrane, i.e. it is necessary to take the cholesterol factor into account when modelling biomembranes.

A decrease in the retention factors for acidic solutes is observed with an increase in the concentration of cholic acid at a fixed concentration of Brij-58 in the mobile phase (Figure 4). This can be explained by the fact that cholic acid is solubilized in Brij-58 direct micelles so, that the hydrophilic part (carboxyl and hydroxyl groups) is directed towards the dispersion medium, and the hydrophobic part is directed towards the hydrophobic core of the spherical micelles (Scheme 1). As a result, the polarity of the micelle surface increases, as it is influenced by the hydrophilic groups of cholic acid in addition to the oxyethylene groups of Brij-58. It should be noted here that a slight increase in the retention factor of BR is observed in the presence of 100 mM cholic acid in the mobile phase.

Figure 4:

Effect of cholic acid concentration on the retention factor of model substances, eluting with 3 % polyoxyethylene (20) cetyl ether mobile phase: 1. BR, 2. HD, 3. BM, 4. PB. (■) 1 mM ChA, () 5 mM ChA, () 44 mM ChA, (□) 100 mM ChA.

No clear dependence is observed for other model compounds, i.e. the retention of some substances (phenylalanine, dopamine, caffeine, ethosuximide, aminoerbic acid) either almost does not change or changes slightly (Figure 5), whereas the change in retention for the other samples (tryptophan, saccharin, vitamin B2) has an alternating character (Figure 6).

Figure 5:

Effect of cholic acid concentration on retention factor of model substances, eluting with 3 % polyoxyethylene (20) cetyl ether mobile phase: 1. Phenyl-alanine, 2. Caff, 3. DOPA, 4. GABA, 5. Ethosuximide, (■) 1 mM ChA () 5 mM ChA, () 44 mM ChA, (□) 100 mM ChA.

Figure 6:

Effect of cholic acid concentration on retention factor of model substances, eluting with 3 % polyoxyethylene (20) cetyl ether mobile phase: 1. Tryptophan, 2. Saccharin, 3. Riboflavin, (■) 1 mM ChA, () 5 mM ChA, () 44 mM ChA, (□) m100 mM ChA.

3.3 Study of the release profile of promethazine hydrochloride from microemulsion systems

The time dependency curves of the drug released from microemulsions (in wt%) prepared with different surfactants at the same concentration are represented in Figure 7. Different values of the weight % of promethazine hydrochloride are observed. However, this difference is greater for drug release from oil-water microemulsions prepared with polymeric surfactants (Figure 7, curves 4, 5) than from microemulsions prepared with monomeric surfactants (Figure 7, curves 1, 2, 3).

Figure 7:

Weight % of PMT release-time curves from microemulsions: 1 (20 mM Triton X-100); 2 (20 mM Brij-35), 3 (20 mM Brij-58), 4 (20 mM Tyloxapol), 5 (20 mM Pluronic). Concentration of PMT in all microemulsions is the same 19.5 mM. n = 3, SD = 2.04, RSD 6.0 %.

As for the monomeric surfactants, the drug release value (in wt%) from the microemulsion prepared on the basis of Brij-58 is the highest (Figure 7, curve 3), and from the microemulsion of Triton X-100 the lowest (Figure 7, curve 1). The release value (wt%) from the Brij-35 microemulsion (Figure 7, curve 2) is intermediate. This may be due to its hydrophilic-lipophilic balance: Brij-58 (HLB = 15.7) and Brij-35 (HLB = 16.9) have strong polar head groups, while Triton X-100 (HLB = 13.5) has a relatively low hydrophilic-lipophilic balance. Therefore, promethazine is more likely to be transported from the surface of micelles with an extended oxyethylene chain, namely Brij-58 and Brij-35 micelles, than from TritonX-100 micelles, which accounts for the comparatively high value of weight % of PMT release in the case of Brij-58 and Brij-35.

The result is the opposite for microemulsions prepared from polymeric surfactants, in particular, the HLB of Pluronic-31R1 is 1, whereas Tyloxapol has an HLB of 12.4. However, a higher drug release (wt%) is observed for Pluronic (Figure 7, curve 5) than for Tyloxapol (Figure 7, curve 4).

The drug release (wt%) decreases with increasing concentration of Brij-58 (Figure 8). This can be explained by the fact that the number of micelles increases with increasing surfactant concentration. As a result, the drug is dissolved in more micelles, thus each micelle is less loaded with drug and therefore the degree of drug release decreases.

Figure 8:

Weight % of PMT release-time curves from Brij-58 microemulsions: 3 (20 mM Brij-58); 6 (10 mM Brij-58); 7 (5 mM Brij-58); 8 (7.5 mM Brij-58); 9 (2 mM Brij-58). Concentration of PMT in all microemulsions is the same 19.5 mM. n = 3, SD = 2.04, RSD 6.0 %.

Brij-30 was chosen to study the drug release from a mixed microemulsion system with Brij-58 (C₁₆E₂₀). Brij-30 was selected for several reasons. Brij-30 has a relatively low HLB value (HLB = 9.7). Triton X-100 which has also a relatively low HLB, gave a low PMT wt% release as seen in previous experiments (Figure 7, curve 1). Brij-30 (C₁₂E₄) was preferred to Triton X-100 to minimize the effect of surfactant structural changes not due to HLB, as it belongs to the Brij-es group or the polyoxyethylene glycol monoether group.

A phase diagram was constructed by preparing a mixture of surfactants with oil and then titrating with water under continuous stirring to study drug release from the Brij-58/Brij-30 mixed microemulsion system (Scheme 3). The shaded area in the ternary phase diagram indicates the region where the microemulsion is formed.

Scheme 3:

Pseudo-ternary phase diagram of surfactant-water-oil system at 1:2 wt ratios of oil to Smix. Smix represents 1:1 wt ratios of Brij-58 to Brij-30. Oil phase is 8:1 (weight ratio of isopropyl myristate to butanol). The dark shaded area represents transparent and easily flowable nanoemulsion region.

The percentage by weight of promethazine released from the mixed system (Brij-58+Brij-30) is lower (Figure 9, curve 10) than from the Brij-58 microemulsion (Figure 9, curve 8). It appears that the addition of Brij-30, which has a low hydrophilic-lipophilic balance (HLB = 9.7), with a more hydrophilic surfactant Brij-58 (HLB = 15.7) reduces the weight % of PMT release from this mixture.

Figure 9:

Weight % of PMT release-time curves from microemulsions: 10 (7 mM Brij 58 + 22 mM Brij 30); 8 (7.5 mM Brij-58).

The numerical values of the diffusion exponent (n) and the kinetic constant (k), the main characteristics of the release process of Promethazine hydrochloride from microemulsions of different compositions are given in Table 2. It is known that the value of the diffusion exponent (n) depends on the mechanism of drug release [24, 26] (Table 3). This indicates that the release of promethazine hydrochloride obeys Fick’s law of diffusion only from the aqueous medium (#6) and from the microemulsion (#5) (Table 2).

In the other cases, the release of promethazine occurred by non-Fickian transport (n = 0.5603 ÷ 0.8387) except for Tyloxapol microemulsion (#9). Only in one case, the release from Tyloxapol microemulsion (#9) occurred by supercase II transport (n = 1.123) (Table 2). It can be seen in Table 2, that the value of the diffusion exponent has a relatively high value at a high surfactant concentration, in particular in the case of 2 mM Brij-58 (#1) and 10 mM Brij-58 microemulsion (#4) it is equal to 0.5603 and 0.8336 respectively. This means that the drug release mechanism changes with increasing in surfactant concentration. In addition, the kinetic constant has a relatively high value at low surfactant concentration, namely 5.20 and 0.97 in the case of 2 mM Brij-58 (#1) and 10 mM Brij-58 (#4) microemulsions, respectively. This pattern is not observed in the case of Tyloxapol (#9), where n = 1.123 and at the same time the kinetic constant has a high value of 100 k = 5.9.