Phytocannabinoids: Chromatographic Screening of Cannabinoids and Loading into Lipid Nanoparticles

2.2. Determination of Particle Size, Polydispersity Index, and Zeta Potential

Once the lipid nanoparticles (SLN and NLC) loaded with CBD were produced, the mean particle size (Z-Ave) and polydispersity index (PDI) of all developed formulations were determined directly after production and up to 30 days. Results on the day of production (day 0) were obtained for the empty-SLN and empty-NLC (both considered as standard samples) and reached about 201.5 ± 3.80 nm with PDI 0.281 ± 0.01 [n.u.] and 173.3 ± 4.40 nm with PDI 0.235 ± 0.02 [n.u.], respectively.

CBD-cSLN suspensions, containing 4% (w/w) of solid lipid (Compritol^® 888 ATO), 1.5% (w/w) of surfactant (Poloxamer^® 188) and 1% (w/w) of active compound (CBD extract), were produced using a probe sonication and stored at room temperature (25 °C) for 30 days. Then, Z-Ave, PDI, and zeta potential (ZP) were measured 1, 14, and 30 days after production. The results are presented in Figure 2. As a result, the Z-Ave for CBD-cSLN suspensions varied from 217.2 ± 6.50 nm (day 1) with a PDI of 0.273 ± 0.02 [n.u.]. Furthermore, after two weeks of storage at 25 °C, the particle size of cSLN suspensions loaded with CBD extract increased to 227.7 ± 11.10 nm (day 14) with a PDI of 0.301 ± 0.02 [n.u.]. It can indicate a slight agglomeration process. Finally, four weeks after the production, the CBD-cSLN suspensions gained 222 ± 5.00 nm (day 30) with a PDI of 0.320 ± 0.01 [n.u.].

Zeta potential (ZP) of empty cSLN was −18.09 ± 0.81 [mV], whereas of CBD-cSLN was −12.99 ± 0.90 [mV] (day 0) to −12.42 ± 2.14 [mV] (day 30). All of the results are presented in Table 2. It may indicate that the non-ionic nature of the surfactant has formed a spherically stabilizing adsorbed polymer layer in the SLN surface [34].

CBD-cNLC suspensions, consisting of 3% (w/w) of solid lipid (Compritol^® 888 ATO), 1% (w/w) of liquid lipid (Miglyol^® 812), 1.5% (w/w) of surfactant (Poloxamer^® 188) and 1% (w/w) of active compound (CBD extract), were also produced using a probe sonication and stored at room temperature for 30 days. The Z-Ave and PDI were measured 1, 14, and 30 days after production. Figure 3 shows that the Z-Ave for CBD-cSLN suspensions varied from 158.3 ± 6.60 nm (day 1), with a PDI of 0.325 ± 0.02 [n.u.]. Two weeks later, the particle size of cSLN suspensions loaded with CBD extract stored at 25 °C decreased to 128.6 ± 3.10 nm (day 14) with a PDI of 0.327 ± 0.01 [n.u.]. After one month after the production, the CBD-cSLN suspensions amounted to 146.2 ± 2.10 nm (day 30) with a PDI of 0.340 ± 0.01 [n.u.].

Zeta potential (ZP) of empty cNLC was recorded at −13.63 ± 0.89 [mV], whereas of CBD-cNLC was recorded at −5.51 ± 1.58 mV (day 0) to −12.06 ± 1.50 mV (day 30). Results are also presented in Table 3. Similar to empty-cSLN and CBD-cSLN, empty-cNLC and CBD-cNLC also showed that the presence of active compounds decreased the stability of the samples.

Zeta potential is a physical property exhibited by any particle in suspension, macromolecule, or material surface. A significant positive or negative value of the zeta potential of nanoparticles indicates good physical stability due to electrostatic repulsion. The dissociation of acidic groups on the surface of a particle will give rise to a negatively charged surface. Conversely, a primary surface will take on a positive charge. In both cases, the magnitude of the surface charge depends on the acidic or basic strengths of the surface groups and the pH of the solution. The surface charge can be reduced to zero by suppressing the surface ionization by decreasing the pH of negatively charged particles or increasing the pH of positively charged particles. The general dividing line between stable and unstable suspensions is generally taken at either +30 or −30 mV. Particles with zeta potentials more positive than +30 mV or more negative than −30 mV usually are considered stable. In our present case, our particles were slightly negative, showing that it is not the surface electrical charge that contributes to the physical stability of the particles, but rather the stereochemical hindrance promoted by the surfactants surrounding the particles.

The final size may depend on various factors, such as the chemical structure of the lipids and surfactants composing the systems and their chemical interaction. It should be pointed out that the largest particle size on day 14 corresponded to CBD-cSLN containing Compritol^® 888 ATO as solid lipid. NLC suspensions either consisted of one solid lipid or liquid lipid and showed significantly smaller particle size than cSLN. This could be attributed to the structural differences between lipid nanoparticles and the inclusion of Miglyol^® 812 in the cNLC formulations. In addition to that, PDI was similar in all CBD-loaded nanoparticles and varied in the optimal range from 0.2 to 0.3 [n.u.]. The stereochemical stability of lipid nanoparticles was monitored for 30 days. Additionally, previous research has revealed that the concentration of Miglyol^® 812 remarkably impacts the nanoemulsions’ stability, which is a crucial factor influencing particle size [35]. The slight reduction in particle size with the loading of CBD was attributed to the bonds between the CBD molecule and solid and liquid lipids, resulting in an imperfect crystal-type of NLC. It involves mixing spatially different lipids, consisting of many fatty acids, which introduce imperfections in the order of the crystals. The drug loading can further increase the molecular interactions, resulting in a smaller nanoparticle size.

2.3. Encapsulation Efficiency and Loading Capacity

The encapsulation efficiency (EE%) and loading capacity (LC%) were determined as qualitative and quantitative parameters of the production of process. EE and LC were calculated for CBD-cSLN and CBD-cNLC formulations to estimate how much CBD was inside each system. The results are shown in Table 3. The influence of varying lipid types (Compritol^® 888 ATO and Miglyol^® 812) on the more efficient loading CBD was analyzed. Consequently, CBD-cSLN achieved the EE of 74.23%, CBD-cNLC gained 70.44%, and the LC was 15.77% and 15.11%, respectively.

2.4. Rheology Study

The viscosity of the lipid nanoparticle-based formulations can vary from liquids to semi-solid systems. It also governs the selection of the ideal pharmaceutical formulation for the intended administration route. When nanodispersions are applied to the skin, they should show a thixotropic pattern with a high viscosity. Therefore, they can adhere and be retained on the skin site. On the other hand, in oral formulations, lipid nanoparticle-based dispersions can be used as granulating media in tablet production to convert the powder into a viscoelastic mass and then into dry-looking granules or load soft and hard gelatin capsules (either as liquids or after spray/freeze-dried).

To study the rheological behavior of the dispersions, an oscillation frequency sweep test was run for each formulation (CBD-cSLN, CBD-cNLC) at 22 ± 0.5 °C one day after production, over the frequency range of 0–10 Hz. For comparison purposes between CBD-cSLN (Figure 4, left) and CBD-cNLC (Figure 4, right), we recorded the storage modulus (G’), loss modulus (G”), and shear viscosity. The loss modulus was higher than the storage modulus over the applied frequency range for both lipid nanoparticle-based formulations. The storage modulus (viscous component) resembles how much energy the dispersion requires to be distorted, whereas the loss modulus (elastic component) translates the energy lost during the strain. At low frequencies, the shear rate is also low, which means that the capacity of the dispersions to maintain their original media strength is high. With the increase in the frequency range, the shear rate also increases, thus requiring more energy and consequently increasing the viscous component (G’, storage modulus).

The results showed that G” was always higher than G’ for both tested samples. It means that they were more elastic than viscous. When the stress was removed, the dispersions were restored to equilibrium but did not follow the same conformational path. Moreover, the dispersions did not maintain their shape, and this behavior is typical of a liquid-like character system.

In turn, the rise in the frequency increased the shear viscosity for CBD-cSLN (Figure 4, left), while for CBD-cNLC, the reverse was observed (Figure 4, right). The addition of an oily component (Miglyol^® 812) to the solid matrix of SLN (to obtain NLC) could result in more viscous dispersions, as seen with the higher shear viscosity values recorded for CBD-cNLC when compared to CBD-cSLN.

2.5. DSC

The thermodynamic stability of cSLN and cNLC is mainly governed by the lipid modifications encountered upon loading with the active ingredient. The selected solid lipid was Compritol (i.e., glyceryl behenate), a mixture of mono-, di- and tri-behenate of glycerol [36]. It is reported that Compritol has a typical crystalline structure of tryglycerols, containing tiny amounts of α-form [37]. The crystallinity index, which measures the amount of solid content inside the particles, shows the slightly solid character of SLN compared to NLC, as documented in the literature [38].

The production of SLN and NLC loading CBD resulted in the decrease in the melting peak down to 71.0 °C and 63.6 °C, respectively, with a significant depression of the crystallization index when we compare CBD-cSLN (90.41%) with CBD-cNLC (40.18%) (Table 4). These results are aligned with the loading capacity and encapsulation efficiency encountered for these particles.

Figure 5 shows bulk Compritol melted at 75.8 °C, with a recording melting enthalpy of 109.7 J/g. The bulk CBD analysis showed the active ingredient’s low melting peak at 37.5 °C (Figure 6). It has been loaded into lipid nanoparticles, contributing to the crystallinity index recorded for both types of particles.

3.2. Chromatography of Plant Extract

Approximately 150 mg of the sample was weighed into a centrifuge tube and flooded with precisely 10 mL of a MeOH: THF mixture. Then, it was shaken for 30 min, and the tubes were centrifuged at 5500 rpm. Samples were diluted 10-fold into an HPLC-type vial and placed on an HPLC autosampler. All analyzes were performed on a gradient of 0.1% H₃PO₄ in water with CAN, observing UV light absorbance at 230 nm on a chromatographic column with o stationary phase C18. Identification and calculation of concentration for cannabinoids were performed based on the chromatograms of the standards, from which the calibration curves were plotted.

3.3. Production of SLN and NLC

SLN and NLC were produced by hot high-pressure homogenization as described by Doktorovova et al. [39] with slight modification and based on preliminary studies as described by Souto et al. [40]. Briefly, the melted lipid phase (solid lipid in case of SLN; solid lipid and liquid lipid in case of NLC) was dispersed in the aqueous surfactant solution heated at the same temperature using a probe sonication (Sonics Vibracell, Newtown, CT, USA) for 15 min and 70% amplitude. A hot emulsion was obtained, then left to cool down at room temperature. For the loading of CBD, the extract was added to the inner oil phase before emulsification. The composition of the developed batches is shown in Table 5.

The authors are most grateful to René Beyer, Head of Strategic Sales Pharma of IOI Oleo GmbH Company in Germany, for gifting the solid lipids used in the study. Faezeh Fathi is grateful to Laboratório Associado para a Química Verde—Tecnologias e Processos Limpos—UIDB/50006/2020 that supports her grant REQUIMTE 2020-20. The authors are thankful to Salette Reis from the Chemical Sciences Department University of Porto (FFUP) for providing the equipment. Eliana B. Souto acknowledges the national funds from FCT—Fundação para a Ciência 414 e a Tecnologia, I.P., in the scope of the project UIDP/04378/2020 and UIDB/04378/2020 of the 415 Research Unit on Applied Molecular Biosciences—UCIBIO and the project LA/P/0140/2020 of the 416 Associate Laboratory Institute for Health and Bioeconomy—i4HB.