Formulation, characterization, and optimization of aripiprazole-loaded lyotropic liquid crystalline nanoparticle for sustained release and better encapsulation efficiency against psychosis disorder

Lyotropic liquid crystalline nanoparticles (LLCNPs) have recently received much attention in the application of drug delivery systems, due to their ordered and versatile internal nanostructures that are considered as a key factor in improving loading efficiency of various poorly soluble therapeutic agents. To take advantage on their unique well-defined and flexible internal nanostructures, aripiprazole-loaded LLCNPs consisted of a binary mixture of soy phosphatidylcholine (SPC) and citric acid ester of monoglyceride (citrem) were developed in this study. Despite exhibiting low aqueous solubility which lead to difficulties in formulation, aripiprazole, a class of psychotropic drug called atypical anti-psychotics has been used in the treatment of schizophrenia and bipolar disorder with few side effects. The utmost interest in this study is to explore the potential of LLCNPs in improving the percentage of encapsulation efficiency (EE%) of aripiprazole, their effect on the internal nanostructure of LLCNPs mesophases as well as the drug release performance from LLCNPs. The particle size of drug-loaded LLCNPs produced was in the range of 161–186 nm, with polydispersity index (PDI) between 0.11–0.16, and negative zeta potential of -21.5 to -23.8 mV. Small-angle X-ray scattering (SAXS) measurements indicated that the internal nanostructures of LLCNPs are of inverse hexagonal (H 2 ) with a negligible difference in the lattice parameter before and after drug loading. Transmission electron microscopy (TEM) was used to observe the morphology and overall size distribution of drug-free and drug-loaded nanodispersions, which supported both SAXS and particles size findings. Differential scanning calorimetry (DSC) and Fourier transform infrared (FTIR) spectroscopy demonstrated that aripiprazole interacted physically with binary mixture of citrem/SPC within the nanodispersions. Moreover, the results showed that aripiprazole was successfully encapsulated into LLCNPs nanoparticles, where the EE% was all above 92%. These LLCNPs were not only have a high EE% value, but also exhibited a sustained release performance of aripiprazole with the release capacity of around 97% up to 96 h. From the current study, the potential use of LLCNPs as a promising nanocarrier for aripiprazole delivery is anticipated to improve the pharmacokinetics of this drug whilst enduring the internal nanostructural stability of the LLCNPs upon exposure to physiological environment.


ABSTRACT
Lyotropic liquid crystalline nanoparticles (LLCNPs) have recently received much attention in the application of drug delivery systems, due to their ordered and versatile internal nanostructures that are considered as a key factor in improving loading efficiency of various poorly soluble therapeutic agents. To take advantage on their unique well-defined and flexible internal nanostructures, aripiprazole-loaded LLCNPs consisted of a binary mixture of soy phosphatidylcholine (SPC) and citric acid ester of monoglyceride (citrem) were developed in this study. Despite exhibiting low aqueous solubility which lead to difficulties in formulation, aripiprazole, a class of psychotropic drug called atypical anti-psychotics has been used in the treatment of schizophrenia and bipolar disorder with few side effects. The utmost interest in this study is to explore the potential of LLCNPs in improving the percentage of encapsulation efficiency (EE%) of aripiprazole, their effect on the internal nanostructure of LLCNPs mesophases as well as the drug release performance from LLCNPs. The particle size of drug-loaded LLCNPs produced was in the range of 161-186 nm, with polydispersity index (PDI) between 0.11-0.16, and negative zeta potential of -21.5 to -23.8 mV. Small-angle X-ray scattering (SAXS) measurements indicated that the internal nanostructures of LLCNPs are of inverse hexagonal (H2) with a negligible difference in the lattice parameter before and after drug loading. Transmission electron microscopy (TEM) was used to observe the morphology and overall size distribution of drug-free and drug-loaded nanodispersions, which supported both SAXS and particles size findings. Differential scanning calorimetry (DSC) and Fourier transform infrared (FTIR) spectroscopy demonstrated that aripiprazole interacted physically with binary mixture of citrem/SPC within the nanodispersions. Moreover, the results showed that aripiprazole was successfully encapsulated into LLCNPs nanoparticles, where the EE% was all above 92%. These LLCNPs were not only have a high EE% value, but also exhibited a sustained release performance of aripiprazole with the release capacity of around 97% up to 96 h. From the current study, the potential use of LLCNPs as a promising nanocarrier for aripiprazole delivery is anticipated to improve the pharmacokinetics of this drug whilst enduring the internal nanostructural stability of the LLCNPs upon exposure to physiological environment.

INTRODUCTION
The agriculture sector in Malaysia contributed RM96.0 billion to the Gross Domestic Product (GDP) in 2017. Oil palm was a major contributor to the GDP of the agriculture sector at 46.6% followed by other agriculture sectors (18.6%), livestock (11.4%), fishing (10.5%), rubber (7.3%) and forestry products (5.6%) (Department of Statistics Malaysia, 2018). There has been an increased usage of chemical fertilizer as it helps in growing crops on the economic scale. Chemical fertilizers (inorganic) are widely used in modern agriculture to amend nutrient deficiencies; to provide high levels of nutrition, which aid plants in withstanding stress conditions; to maintain optimum soil fertility and improved crop quality.
The drug solubility and its desired concentration in systemic circulation are some of the important criteria that need to be achieved in order to have a good therapeutic outcome, which would improve the success rate in the pharmaceutical development (Savjani et al., 2012). It turns out that most of the drug candidates are having disadvantage of poor water solubility which renders a low potential for their bioavailability (William et al., 2013;Kalepu and Nekkanti, 2015). Aripiprazole, an anti-psychotic drug that is being used to treat schizophrenia and bipolar disorder, also showed poor aqueous solubility which limits its clinical use .
Many attempts have been made to improve the solubility of aripiprazole including via nanoprecipitation or nanosuspension of the drug, formation of complexes with cyclodextrin as well as solubilization in different type of oils and lipids (Xu et al., 2012;Badr-Eldin et al., 2013). The previous findings of a dramatically higher percentage of cases reported with the depot formulation suggests that the risk of developing impulse control disorders (ICDs) while taking aripiprazole may be dosage dependent (Lertxundi et al., 2018). Another technique is by incorporating aripiprazole into lyotropic liquid crystalline nanoparticles (LLCNPs) which were described to have good tolerability, minimal cytotoxic effects, and could provide an alternative solution for poorly soluble drug therapeutics without altering the molecular structure (Akbarzadeh et al., 2013;Cerpnjak et al., 2013;Kalepu and Nekkanti, 2015).
LLCNPs are a promising drug delivery mechanism with the ability of loading pharmaceutical molecules of any polarity including proteins, peptides as well as nucleic acids (Krishna Sailaja et al., 2011;Asha Spandana et al., 2020). Because of their biocompatible and biodegradable features, stability, nontoxicity, potential to improve safety, effectiveness, and ability to penetrate across physiological barriers, LLCNPs have received a lot of interest in drug delivery research (Chountoulesi et al., 2018;Dutta et al., 2018;Tan et al., 2018;Chountoulesi et al., 2020;Maslizan et al., 2021;Yaghmur and Mu, 2021). Nanostructured liquid crystals, like liposomes (internal lamellar phase), cubosomes (internal bicontinuous cubic phase), and hexosomes (internal hexagonal phase), can be formulated by dispersing these structures into the aqueous medium with the presence of suitable stabilizers, e.g. the citric acid esters of monoglycerides (citrem) (Spicer, 2004;Bhosale et al., 2013).
In this work, we present the possibility of using LLCNPs consisted of a binary lipid system citrem and phospholipid (soy phosphatidylcholine) at different lipid ratios, to improve the encapsulation efficiency of aripiprazole. Alongside with the important role of nanoformulation to tackle the solubility issue of aripiprazole, we have also addressed the potential of LLCNPs to offer a sustain release profile (long-acting therapy), which is important for nonadherence patients of anti-psychotropic medication. By understanding the physicochemical characteristic of the drug-free and drug-loaded LLCNPs as well as the drug release profile, the development of drug delivery system for anti-psychotic treatment will have higher chance to thrive as several studies on LLCNPs which have reported their ability to improve the aqueous solubility of hydrophobic medicines, high encapsulation efficiency, stability in biological fluids, drug retention capability, and tendency to transport drug across the blood brain barrier (Elezaby et al., 2017;Piazzini et al., 2020).

Materials
Aripiprazole with purity 99.0% was purchased from Sigma Aldrich (Darmstadt, Germany). Soy phosphatidylcholine (SPC) with purity of 99.1% was purchased from Lipoid AG (Steinhausen, Switzerland). Grinsted® citrem LR10, which is a citric acid ester of monoglycerides made from sunflower oil, was received as a gift from Danisco A/S (Copenhagen, Denmark). Dulbecco's phosphate-buffered solution (PBS) was purchased from Thermo Scientific (Hampshire, UK).

Preparation of Drug-Free LLCNPs
Preparation of aqueous nano-dispersions involves a procedure that requires high-energy input (ultrasonication) (Azmi et al., 2016). Binary mixture of citrem/SPC was prepared at 1:1 ratio where the total lipid concentration was kept constant at 5% w/v. PBS was then added to the lipid mixtures to give 100% of total weight dispersion. The emulsification process was performed by means of ultrasonication (QSonica sonicator Q125, Newtown, USA) for 5 min in pulse mode (5 s pulses interrupted by 2 s breaks) at 20 and 50% of its maximum power until well-dispersed and stable milky solutions were obtained.

Preparation of Drug-Loaded LLCNPs
Aripiprazole-loaded citrem/SPC formulations were prepared via the ultrasonication method as well. Initially, aripiprazole at four different concentrations of 1.0, 2.0, 2.5, and 5.0 mg/mL with corresponding weights of 10, 20, 25, and 50 mg were added to already-prepared citrem/SPC nanodispersion (prepared as described in section 2.2.1) until achieving well-homogenized milky formulations. Subsequently, each formulation was dissolved in 10 mL of PBS/methanol (90:10 v/v) mixture with the aid of an ultrasonic bath on 37°C at 80 rpm to homogenize the drug formulation until well-dispersed and stable milky solutions were obtained (Piazzini et al., et al., 2020).

Encapsulation Efficiency of Drug-Loaded LLCNPs
UV-Vis spectrophotometer (UV-1601, Shimadzu, Japan) was used to study the percentage of encapsulation efficiency (EE%) of aripiprazole-loaded LLCNPs. A series of standard aripiprazole solution concentrations in a range of 0.001 to 0.050 mg/mL was prepared and a calibration curve was plotted in order to determine the EE%. All measurements were based on x mg of free aripiprazole in 1 mL of solution taken from the sample solution, x mg/mL, to calculate back the total amount of free aripiprazole. Evenly distributed free aripiprazole in the solution was ensured by swirling the solution well before performing the measurement. A series of measurements was performed to ensure the stability of the encapsulated drug into the LLCNPs for five consecutive days. The EE% was calculated using the given formula in 2.3.2 Particle Size, Zeta Potential, and Polydispersity Index The particle size, zeta potential, and polydispersity index (PDI) of the aripiprazole-loaded LLCNPs were determined by Zetasizer Nano ZS90 (Malvern Instruments Ltd, Malvern, UK) adopting the dynamic light scattering mechanism. This instrument was equipped with a 633 nm laser and 173 detection optics. To prevent multiple scattering negative impact, the aripiprazole-loaded LLCNPs (1 mL) have been diluted with PBS (9 mL) and the instrument was calibrated prior to the analysis at 25°C. Collection of data and analysis were conducted using Malvern DTS version 6.34 software. All the experiments were performed in triplicates and the average values were taken. The value of PBS solvent was used for the viscosity and refractive index.

Structural Characterization
Small-angle X-ray scattering (SAXS) technique was used to evaluate the liquid crystal phase and determine the important parameters of the structures obtained (Azmi et al., 2016;Saari et al., 2018). An analytical SAXS from SAXSpace (Anton Paar, Austria) was employed to characterize the dispersion scattering of the aripiprazole drug-loaded LLCNPs. The instrument was equipped with an X-ray tube (DX-Cu 12x0.45, SERFERT) emitting Cu-Kα of wavelength, λ = 0.1542 nm at 50 mA and 40 kV. The Goebel-mirror focused and Kratky-slit collimated X-ray beam was line shaped (17 mm horizontal dimension at the sample) and scattered radiation from the samples (measured in the transmission mode) was recorded with a one dimensional MYTHEN-1k microstrip solid-state detector (Dectris Switzerland) within a q-range of 0.1 to 6.0/nm at a sample-to-detector distance of 317 mm, where q is the magnitude of the scattering vector applying the conversion q [1/nm] = 4π(sinθ)/λ, with 2θ being the scattering angle with respect to the incident beam and λ the wavelength of the X-rays in nm. Silver behenate which has a periodicity of 5.84 nm (where d [nm] = 2π/q1 is the lamellar spacing obtained from the position of the first order reflection q1) was used as standard. The sample (~60 μL) was transferred onto a 1 mm quartz capillary cell holder and placed into the X-ray machine. The measurement was set to room temperature (25.0 ± 0.1°C) using a Peltier system (TCStage 150). An acquisition time of 15 min was applied for all measurements. The collected data were calibrated by normalizing the primary beam using SAXStreat software. The liquid crystal phases and their corresponding lattice parameters were analyzed using the SGI software (Space Group Indexing, V.03.2012).

Morphological Characterization
Transmission electron microscopy (TEM, JEOL JEM-2100, Japan) was used as a visualizing aid for the nonlamellar LLCNPs. Two samples were selected as representative i.e. drug-free and aripiprazole-loaded nanodispersions. Each nanodispersion was diluted 10 times with distilled water. A droplet of each diluted samples was then negatively stained with 1% aqueous solution of uranyl acetate. The stained samples were placed on the 400-mesh carbon-coated copper grids and allowed to stand for 5 min at room temperature. Subsequently, the observation of the two samples were made with a transmission electron microscope at a voltage of 200 kV. The images were viewed with magnification between 20,000× to 50,000×.

Fourier Transform Infrared Spectroscopy (FTIR)
Fourier transform infrared spectra of aripiprazole, drugfree, and aripiprazole-loaded LLCNPs were evaluated to investigate the viable chemical interaction between aripiprazole and LLCNPs. The measurements were recorded on a Shimadzu IR Tracer-100 spectrometer (Japan) using the attenuated total reflectance (ATR) method. All data were obtained in the 4000-400 cm range, with a spectral resolution of 4/cm and collected after 4 scans.

Differential Scanning Calorimetry (DSC)
Differential scanning calorimetry measurements were performed to investigate the change in physical behavior and phase transition temperature of aripiprazole, SPC, drugfree, and aripiprazole-loaded LLCNPs and also to study the presence of interaction between the aripiprazole with the LLCNPs. The measurements were carried out on a Mettler Toledo DSC 822e calorimeter (Columbus, OH, USA). Indium was used as the standard sample to calibrate DSC runs. Samples of 5-15 mg were sealed in aluminum pans. The samples were heated at scanning rate of 10°C/min from 30 to 300°C under a nitrogen atmosphere. Heat flowtemperature curve, melting point and specific heat energy were analyzed using STARe Excellence Thermal Analysis software.

In Vitro Drug Release Profile of Aripiprazole-Loaded LLCNPs
In vitro drug release study was performed by a dialysis technique (Nethravani et al., 2014) shown in Fig. 1. A ratio of 1:1 (citrem: SPC) with aripiprazole concentration of 2.5 mg/mL was selected to be incorporated in the nanodispersion. Aripiprazole-loaded LLCNPs (5 mL) was placed in a dialysis bag (pore size ~12,000 Da) that was previously soaked for 2 h in the PBS (pH 7.4), cleaned and sealed at both ends. The whole dialysis bag containing 5 mL aripiprazole-loaded LLCNPs was placed into 250 mL PBS (pH 7.4) and shaken (80 rpm) at 37.0 ± 0.5°C. The water bath sonicator was covered to prevent the evaporation of release medium. At time intervals of 2, 20, 24, 48, 50, and 96 h, 4 mL aliquots were withdrawn from the medium followed replacement with an equivalent volume of fresh PBS solution in order to maintain constant volume throughout the experiment. The amount of the drug diffused was estimated from the samples at 254 nm using UV-vis spectrophotometer (UV-1601, Shimadzu, Japan). All the experiments were performed in triplicates and the average values were taken. The formula for determination of cumulative percentage release of aripiprazole is given in Equations 2 and 3 as follows: In vitro drug release test method using dialysis technique. The appearance of drug-free concentration in the release medium is the result of diffusion from the drugloaded nanoparticle followed by diffusion across the dialysis membrane was measured using UV-vis spectrophotometer (Modi and Anderson, 2013).

Statistical Analysis
All data were expressed as average ± SD (standard deviation) unless particularly outlined. The data were compared using one sample t-test for independent sample and one-way analysis of variance (ANOVA) for multiple group comparison procedure. All statistical comparisons were calculated using SPSS software version 16.0 (SPSS Inc., Chicago, IL). A p-value < 0.05 was considered statistically significant.

Encapsulation Efficiency of Drug-Loaded LLCNPs
The encapsulation efficiency (EE%) of aripiprazoleloaded LLCNPs at a constant citrem/SPC ratio of 1:1 (CS1:1) is shown in Fig. 2. Four different aripiprazole concentrations of 1.0, 2.0, 2.5, and 5.0 mg/mL have been selected for this study. The results demonstrated high EE% value of ~92% for all understudied drug-loaded nanodispersions and was relatively independent of aripiprazole concentrations. This finding evidences the possibility of LLCNPs to improve the encapsulation of poorly water-soluble aripiprazole as compared to other reported techniques such as using poly (lactic-co-glycolic acid) (PLGA) nanoparticles which the EE% values ranged from 20-40% only . Statistical analysis revealed that the mean score of citrem/SPC-loaded aripiprazole at the concentrations of 2.0, 2.5, and 5.0 mg/mL was significantly different (p > 0.005) compared to that of formulation with 1.0 mg/mL aripiprazole. Thus, the high EE% of aripiprazole into the LLCNPs indicates maximum amount of the drugs that have been successfully loaded into the LLCNPs. were analyzed by one sample t-test. The means marked with * and ** are significantly different with p value less than 0.05 and 0.01, respectively, as compared among the tested formulations. Fig. 3 shows the stability study for different concentrations of aripiprazole-loaded LLCNPs (1.0, 2.0, 2.5, and 5.0 mg/mL) through 5 days of storage at 4°C. Stability test of nanodispersions for several days were done in the lab to endure the stability of drugs in the body system while modulating their physicochemical and biological properties (Silki and Ranjan, 2018). The stability of encapsulated drug was attempted to analyze the presence of any metastable formulations in the batches so that we can obtain a stable formulation to be used for a longer duration (Khumbhar et al., 2021). From Fig. 3, formulation with 1.0 mg/mL aripiprazole showed a decreased value from 92.5 to 88.8% with total reduction of 3.7% within 5 days of storage. On the other hand, formulation with 2.0 mg/mL aripiprazole displayed better stability with a reduction of 1.6% from 92.1% to 90.5%. Similar total reduction was observed for the stability of formulation with 2.5 mg/mL aripiprazole (Fig. 3). The highest concentration of 5.0 mg/mL aripiprazole-loaded LLCNPs exhibited the least drug lost with total reduction of only 1.0% from 92.0 to 91.0%. In overall, the 1.0 mg/mL of aripiprazole-loaded LLCNPs was relatively unstable compared to the other three nanodispersions that showed minimal drug lost. It can be observed that as the concentration of aripiprazole increases, the stability of encapsulated drug increases as well. Statistical analysis revealed that the mean score for EE% stability within 5 days of storage among the four aripiprazole-loaded LLCNPs showed no significant different (p < 0.005).

Stability Study for Encapsulation Efficiency of Drug-Loaded LLCNPs
Hydrophobic interaction between phospholipid bilayer of SPC increases with increasing amount of aripiprazole . The high hydrophobic interaction will cause the drug to diffuse at a slower rate from the bilayer to the outer aqueous phosphate buffered saline medium . LLCNPs have a hydrophobic core that can encapsulate aripiprazole and a hydrophilic part that acts as an outer corona on lipid bilayers (Piazzini et al., 2020). This type of drug delivery system may extend the duration of the drug nanocarrier travels in the blood circulation (Piazzini et al., 2020) and thus, offers larger drug loading capacity across the blood brain barrier (Cagel et al., 2017;Piazzini et al., 2020).

Particle Size, Zeta Potential, And Polydispersity Index Characterization for Drug-Free LLCNPs
The particle size, zeta potential, and polydispersity index (PDI) of the nanodispersions were assessed by dynamic light scattering. Fig. 4 shows the sonication time, particle size, zeta potential, and PDI of the drug-free LLCNPs prepared at two different sonication powers of 20 and 50%.
After 30 min of 20% sonication power, the particle size obtained was 475 nm with high PDI of 0.50 which indicates non-homogeneous particles size ( Fig. 4(a) and 4(c)). Increasing the time to 60 min successfully reduced the particle size to 149 nm with low PDI of 0.11 ( Fig. 4(a) and 4(c)) denoting a homogenous particle size was attained.
Danaei and co-workers (2018) stated that a PDI value of 0.30 or less was desirable, implying a relatively homogeneous formulation (Danaei et al., 2018). Herein, the zeta potential values of all the prepared nanodispersions were negatively-charged ( Fig. 4(b)). Such a tendency could be attributed to the presence of the negatively-charged head group of the citrem (Azmi et al., 2016). In principle, zeta potential values above +30 mV or below -30 mV provide a good stability in the dispersion medium (Gabr et al., 2017). As shown in Fig. 4(b), all zeta potentials after sonication were recorded with a magnitude lower than 30 mV implying electrostatic force of attraction between nanoparticles exceeds the repulsion force. Specifically, increasing the sonication time to 60 min further reduced the negative zeta potential value, thus signifying lack of colloidal stability compared to that of 30 min sonication time (Fig. 4(b)). Nonetheless, both mean particle size and PDI of the drugfree LLCNPs at 60 min still demonstrates relatively good homogenous formulation compared to the 30 min sonication period. However, 20% sonication power at 60 min was consider as time-consuming process for the preparation of stable and well-dispersed milky formulation. Therefore, the method was optimized to higher sonication power (50%) but with shorter time interval (5, 7, 10, and 13 min).
Using 50% sonication power at different time intervals of 5, 7, 10, and 13 min, all drug-free LLCNPs produced particle sizes of less than 167 nm with small PDIs of less than 0.30 implying the homogeneous particle size distribution were formed. Different sonication times ranging from 5 to 13 min were studied since they can impact the dispersion behaviors and aggregation size of nanoparticles (Mahbubul et al., 2015). The zeta potential for drug-free LLCNPs prepared through sonication for 5 min was -26.3 mV (Fig. 4(b)). At longer sonication time of 7, 10, and 13 min, the negative zeta potential showed slightly different values from -26.1 to -23.1 mV (Fig. 4(b)). Smaller particle size and homogenous formulations were achieved upon longer sonication time. The particle size for 5 min sonication time was 166 nm with PDI of 0.27 (Fig. 4(a) and 4(c)), whereas for 7 min sonication time, the particle size was 152 nm with PDI of 0.21 (Fig. 4(a) and 4(c)). Sonication time of 10 min formed particle size of 154 nm with PDI of 0.17 (Fig. 4(a) and 4(c)). The optimized condition was obtained at 13 min sonication time which gave particle size of 150 nm with PDI of 0.13 (Fig. 4(a) and 4(c)). Statistical analysis revealed that the mean score of particle size, PDI, and zeta potential for citrem/SPC-loaded aripiprazole at the concentration of 2.0 and 2.5 mg/mL including the drug-free nanodispersion showed no significant different (p < 0.005) in comparison between group.
Based on the findings, the particle size, polydispersity index, and zeta potential of drug-free LLCNPs decreased as the time for sonication increased indicate less stable nanodispersions. The decrease in particle size and PDI was due to sufficient energy to break the particle into smaller parts and evenly distributed in the aqueous medium (Olson, n.d.;Uner, 2006;Larsson et al., 2012). The decreased negative zeta potential implies less stabilization of the nanodispersions due to the lack of electrostatic force of repulsion between particles (Gabr et al., 2017). Thus, 10 min of 50% power of sonication was selected as the optimum condition to prepare the drug-free and drug-loaded LLCNPs because the differences in particles size, polydispersity index and zeta potential between 10 and 13 min of sonication time shows no differences in the stability of nanoparticles.

Particle Size, Zeta Potential, and PDI for Drug-Loaded LLCNPs Upon Storage Conditions
Using the optimized method obtained (from subsection 3.2.1), the particle size, zeta potential, and PDI values of drug-free and drug-loaded LLCNPs are summarized in Table 1. Since the value of EE% (Fig. 2) was not improved for aripiprazole concentration higher than 2.5 mg/mL, thus 5.0 mg/mL aripiprazole-loaded LLCNPs was not considered for additional stability study upon incubation. The particle size of drug-free LLCNPs obtained was 154 nm with PDI of 0.17. Loading 1.0 mg/mL of aripiprazole into LLCNPs gave an increment of particle size to 161 nm with PDI of 0.11. Meanwhile, LLCNPs loaded with 2.0 mg/mL and 2.5 mg/mL of aripiprazole resulted in larger particle size of 186 nm (PDI of 0.16) and 183 nm (PDI of 0.10), respectively. On the other hand, the zeta potentials of drugfree LLCNPs, 1.0, 2.0, and 2.5 mg/mL of aripiprazoleloaded LLCNPs were recorded within a range of -21.5 to -23.8 mV. In general, the LLCNPs prepared for both drugfree and drug-loaded were below 186 nm, with PDI less than 0.30. The value of negative zeta potentials of the drug-free and drug-loaded LLCNPs are acceptable, thus implying small particles size with homogenous particle size distribution of LLCNPs were formed.
Upon storage, the changes in particle size, zeta potential, and PDI of drug-free and drug-loaded LLCNPs were illustrated in Fig. 5. For drug-free LLCNPs, the particle size increased from 154 to 404 nm with PDI increased from 0.17 to 0.51 after 17 days (Fig. 5(a) and (c)). The negative zeta potential value also increased in magnitude from -23.3 to -27.8 mV indicating an improved electrostatic stabilization of the nanoparticles (Figure 5(b)). Significant changes in particle size and zeta potential were also obtained with aripiprazole-loaded LLCNPs. The 1.0 mg/mL of aripiprazole-loaded LLCNPs showed an increment in the particle size (from 161 to 1414 nm) as well as the PDI (from 0.11 to 0.91) as the storage time increases (Fig. 5(a) and (c)).  5. Stability analysis of (a) particle size, (b) zeta potential, and (c) PDI using dynamic light scattering analysis. Data are expressed as mean ± SD of triplicates (one independent experiment). Results were analyzed by one-way ANOVA.
The corresponding negative zeta potential was increased from -21.5 to -29.1 mV (Fig. 5(b)). As for 2.0 mg/mL of aripiprazole-loaded LLCNPs, the particle size was elevated from 186 to 784 nm with PDI raised up from 0.16 to 0.62 upon 9 days of storage. However, after 17 days of storage, we observed a decrease in both particle size and PDI values from 784 to 630 nm and 0.62 to 0.56, respectively (Fig. 5(a) and (c)). On the other hand, the corresponding negative zeta potential value was increased from -23.8 to -28.3 mV as the storage time increases up to 17 days (Fig. 5(b)). The 2.5 mg/mL of aripiprazole-loaded LLCNPs exhibited similar trend as the 2.0 mg/mL aripiprazole concentration for the particle size (Fig. 5(a)), zeta potential ( Fig. 5(b)), and PDI (Fig. 5(c)) values. Such observation for high drug concentrations could be explained by physical and chemical instability of cubosome nanoparticles (Fig. 6), which often synthesized in solution as colloidal systems and cannot be kept for lengthy periods of time (Malheiros et al., 2021). As a result, it is critical to investigate alternative storage settings for cubosomes in order to extend their shelf life (e.g. freeze drying method) (Malheiros et al., 2021). Some processes are employed in the pharmaceutical industry to improve the stability of a given formulation (Malheiros et al., 2021). Statistical analysis revealed that the mean score for stability study of particle size, PDI, and zeta potential over 17 days among drug-free and drug-loaded LLCNPs at the concentration of 1.0, 2.0, and 2.5 mg/mL showed no significant different (p < 0.005) in comparison between group.
The 2-week storage stability of aripiprazole-loaded LLCNPs after optimization is crucial for future applications as it is related to the electrostatic charge of nanoparticles (Luangtana-Anan et al., 2010;Cai et al., 2017). The increase in both particle size and PDI of the drug-free and 1.0 mg/mL of aripiprazole loaded LLCNPs was probably due to the occurrence of coalescent between LLCNPs. This causes them to aggregate and become bigger in particle sizes whilst their steric stabilization was sufficient to keep them together.

Structural Characterization of Drug-Loaded LLCNPs
Binary citrem/SPC (CS1:1) mixtures of drug-free and loaded-aripiprazole of different drug concentrations listed in Table 2 were characterized by small-angle X-ray scattering (SAXS). An internal inverse hexagonal (H2) phase (or known as hexosomes) with reciprocal spacing ratios of √1, √3, and √4 was observed for all formulations in Fig. 6. The internal nanostructure of these nanodispersions was retained with almost similar lattice parameter as the aripiprazole concentration increases (Table 2). It is suggested that the asymmetric insertion of citrem occurs through preferential charged-dipole interactions between the positively charged choline group of the zwitterionic headgroups of SPC molecules and the anionic moieties of citrem resulted in the stabilization of this curved state of asymmetric bilayers (Azmi et al., 2016).

Morphological Characterization of Drug-Free and Drug-Loaded LLCNPs
Transmission electron microscopy (TEM) was used to observe the morphological and the distribution of LLCNPs. The drug-free and aripiprazole-loaded LLCNPs at low (Figs. 7(a) and 8(a)) and high (Figs. 7(b) and 8(b)) magnifications were obtained. The morphology results for both formulations were in a good agreement with SAXS analysis of inversed type LLCNPs. The occurrence of internal nanostructure of H2 phase was observed at higher magnification. Interestingly, aripiprazole-loaded LLCNPs illustrated more closely packed formation of lipid bilayers, thus indicated significant interaction between aripiprazole and lipid bilayers formed by SPC and citrem. Since aripiprazole is a poorly water-soluble drug with limited solubility, it is relatively hydrophobic and showed good affinity towards the lipid bilayer formed by SPC. This induced an interaction through van der Waals forces (Chong et al., 2015). The interaction is therefore, maintained the internal nanostructure of H2 phase with slightly smaller aqueous nanochannels (Fig. 6). The FTIR spectra of aripiprazole, drug-free, and drugloaded LLCNPs are presented in Fig. 9. The FTIR spectrum of aripiprazole ( Fig. 9(a)) displayed an absorption band at 3472/cm indicated the N-H stretching of the secondary amide group of the lactam ring. Aromatic stretching = C-H vibration showed two absorption bands at 3103 and 3063/cm (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). The absorption band at 1375/cm corresponded to the C-N stretching vibration of the aromatic amine (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). Aromatic = C-Cl out of plane bending vibration showed a strong absorption band at 78/cm. This peak also indicated the orthosubstituted chloride in the benzene ring (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). The peak at 675/cm is due to the out of plane vibration of N-H of the secondary amide group (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). While absorption bands in the region of 407-591/cm are attributed to C-C-C out of plane bending vibration of the alkyl chain (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014).

Fourier Transform Infrared Spectroscopy Analysis
In the FTIR spectrum of drug-free LLCNPs ( Fig. 9(b)), the absorption band at 3221/cm belongs to O-H stretching of carboxylic acid, citric acid and glycerol (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). Typical peaks around 2959-2868/cm corresponded to the stretching vibration of alkyl groups (CH2) of phospholipids. A strong absorption at the 1158/cm indicated the presence of C-O stretching vibration of carboxylic acid O = C-OH. It also may be due to the P = O stretching vibration of the phosphate group of soy phosphatidylcholine (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). The peak at 1059/cm corresponded the symmetric stretching vibration of the C-O-C of ester bond (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). The absorption band observed at 943/cm belongs to the asymmetric stretching vibration of P-O-C, while that at 854/cm can be attributed to symmetric stretching vibration of P-O-C of the bond formed between phosphate group and glycerol (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). Absorption band of 739/cm was from C-C-OH out of plane bending of alcohol like in citrate or glycerol (Stuart, 2004;Ardiana et al., 2013;Begam et al., 2014). For drug-loaded LLCNPs (Fig. 9(c)), the peak located at 2922 and 2853/cm (CH2) was similar to that of drug-free LLCNPs. The O-H stretching band at 3221/cm was shifted to 3395/cm which suggested some interaction by intermolecular hydrogen bonds. The band for C-N stretching vibration of aromatic amine in aripiprazole (1375/cm) became reduced in intenisty in the drug-loaded LLCNPs. This suggested that aripiprazole was entrapped within the LLCNPs, possibly by hydrophobic interactions. Fig. 10 illustrates the DSC thermograms of aripiprazole, SPC, drug-free, and drug-loaded LLCNPs. Three endothermic peaks of aripiprazole were detected at 89, 140, and 149°C corresponding to its melting points for the different crystal forms (Fig. 10(a)). In general, the DSC curve of aripiprazole could be divided into two regions: (i) below 130°C is the dehydration/melting of hydrate form and (ii) above 130°C is the melting of anhydrous form (Ayala et al., 2010). The peak at 140°C is typically associated with the melting of the anhydrous form, while the observation of another melting point at 149°C evidenced the phase transition to another crystal form during the heating process. SPC thermogram disclosed three endothermic peaks at 97, 190, and 263°C which are correlated with its liquid crystalline nature (Fig. 10(b)). The first phase transition occurred at 97°C corresponding to the change from ordered gel phase to less ordered crystalline phase (planar gel). The second phase transition was observed at 190°C indicating transition to a new crystalline phase (rippled phase), while the third phase transition at 263°C signifies the transition to lamellar liquid crystalline phase Popova and Hincha, 2011). As expected, the DSC thermograms of both drug-free ( Fig. 10(c)) and drug-loaded ( Fig. 10(d)) LLCNPs experienced different endothermic event from those of aripiprazole and SPC which implies the formation of a new structure with different thermal characteristics. For drugfree LLCNPs, a double peak was observed at 120 and 125°C, while for drug-loaded LLCNPs, a double peak was detected at 120 and 127°C. Their peaks appeared to be broader than the pure aripiprazole and SPC due to the particles size distribution of the nanoparticles (Longo et al., 2011). A slight shift towards higher temperature value in case of drug-loaded LLCNPs than drug-free LLCNPs suggests an interaction between aripiprazole and the binary mixture of citrem/SPC. Moreover, the absence of aripiprazole endothermic peaks in the thermogram of drug-loaded LLCNPs implies that aripiprazole was well encapsulated within the nanodispersion.

In Vitro Drug Release Profile of Drug-Loaded LLCNPs
The drug release study for the 2.5 mg/mL of aripiprazoleloaded LLCNPs in Fig. 11 shows a sustained release profile up to 96 h. More than 50 and 70% of aripiprazole were released from LLCNPs after 24 and 48 h, respectively. After 96 h, around 97% of aripiprazole was released from the nanodispersion. The sustained release rate of LLCNPs is possibly due to the interaction of the drug with the lipidic bilayer (Kumar et al., 2007). Previously, Babu and coworkers (2014) reported the aripiprazole release study from PLGA nanoparticles up to 35 days with 80% of the drug was released out but with limited encapsulation efficiency . Additionally, the in vitro release data was in a good agreement with the stability result for 2.5 mg/mL of aripiprazole-loaded LLCNPs in which this inversed typed LLCNPs offer an exceptional long-term stability of particle size, zeta potential, and PDI with excellent drug encapsulation efficiency Table 1 and Fig. 2).

CONCLUSION
An inversed type LLCNPs was successfully prepared using a binary mixture of SPC and citrem-loaded with aripiprazole. The EE% of aripiprazole-loaded LLCNPs was relatively high (92%) with all formulations at different drug concentrations (1.0, 2.0, 2.5, and 5.0 mg/mL). The optimized method of preparation using high-energy ultrasonication method at 50% power of sonication, was significantly efficient to form homogenous solution in 10 min. The particles size for the drug-free and drug-loaded LLCNPs (1.0 to 2.5 mg/mL) was in the range of 154 to 186 nm, PDI from 0.10 to 0.17 and negative zeta potential around -21.5 to -23.8 mV. The stability studies up to 17 days of incubation period, resulted in coalescent between LLCNPs causing them to aggregate and become larger particles. However, the steric stabilization from zeta potential of LLCNPs was improved. Based on the SAXS and TEM analysis, an inverse hexagonal (H2) phase LLCNPs was observed. An interaction between aripiprazole with SPC and citrem was also observed from the FTIR and DSC results. The drug release from 2.5 mg/mL of aripiprazole-loaded LLCNPs revealed sustained release performance up to 96 h. In conclusion, this study proposed the promising usage of inversed type LLCNPs as a potential nanocarrier for aripiprazole delivery, which is projected to enhance the drug encapsulation and sustained release of this drug-loaded LLCNPs upon administration to a living organism. However, biological studies such as an in vitro toxicity study for aripiprazole-loaded LLCNPs should be further investigated to ensure its biocompatibility and efficacy with the related cells.