Isopropyl myristate (99.99%) and 2-propanol (99.99%) were purchased from Sigma-Aldrich, while Tween-80^® (99.99%), Tween-60^® (99.99%) and Tween-20^® (99.99%) were purchased from VWR Chemicals. Schazoo Zaka Lab (PVT), Lahore, Pakistan, generously provided Brinzolamide (working standards). Distilled and deionized water was used for dilution.

μEs of three different formulations were prepared using the titration method by mixing IPM as oil, 2-propanol as cosurfactant, Tween-80, Tween-20 and Tween-60 as surfactants. Surfactant/co-surfactant (S_mix) were used in a 1:1 ratio. Water was then added to form the optimized μE formulation. Formulation A consists of IPM, Tween-80, 2-propanol and water. Formulation B, on the other hand, comprised of IPM, Tween-20, 2-propanol and water. While, Formulation C consisted of IPM, Tween-60, 2-propanol and water. Ternary phase diagrams were constructed using random ratios of oil, water and S_mix. A dilution line AB as shown in Figure 1, was created at a constant S_mix ratio to investigate the structural changes of the μE from w/o to o/w phase. The optimal μEs were selected from the dilution line.

BZD was dissolved in each optimized μE formulation under continuous stirring at pH 7.2, 7.4 and 7.8 respectively at room temperature (30°C). Optimized μEs were consist of 38% Tween-80, 38% 2-propanol, 8% IPM and 16% water for formulation A, 40% Tween-20, 40% 2-propanol, 9% IPM and 11% water for formulation B and 40% Tween-60, 40% 2-propanol, 8% IPM and 12% water for formulation C μE A, B and C dissolved 2.0wt%, 2.0wt%, and 1.0wt% of BZD, respectively. All drug-free and drug-loaded optimal formulations were stable and transparent over 9 months of storage.

The developed μE formulations were centrifuged for 15 min at 5,500 rpm using Hermle Z200 centrifuge equipment to check the stability of each drug-free and drug-loaded μE. Meanwhile, a biological microscope (LABOMED FLR Lx 400; Jenoptic, Germany) with 4x/10x/40x/×100 magnification was used to reveal the structural transitions in all drug free μE formulations. Viscosity was measured at 25°C ± 0.5°C with a calibrated Brookfield viscometer (LVDV-2T) at 150 rpm by washing and cleaning the viscometer at each measurement. Electrical conductivity measurement has been used to study and validate different types of phase inversion and continuous phase of μEs using a conductivity meter (ADWA AD-3000, Hungry).

Using a Zetasizer (Malvern, Nano ZSP), the average droplet size of BZD-free and BZD-loaded formulation and the polydispersity index (PDI) were determined at room temperature without sample filtration. The equipment had a laser with a wavelength of 635 nm and backscattering technology, also known as NIBS (Non-Invasive Back-Scatter), was used to measure light scattering at an angle of 173°. For accuracy, each individual measurement was performed three times.

The IR spectra of drug-free and drug-loaded μEs in the range of 500–4,000 cm^-1 were recorded with a resolution of 2 cm^-1 using a Bruker FTIR (Alpha series), while the steady-state fluorescence was measured using a spectrofluorophotometer (manufactured by Shimadzu RF-6000). The fluorescence spectra of BZD were recorded in the range of 300–800 nm. These spectra were recorded in aqueous phase, single oil phase, S_mix (1:1), and in each of the optimal μE systems.

The in-vitro release of all formulations was evaluated using simulated tear fluid (STF) at 50 rpm. The 2 mL freshly prepared STF medium used consisted of several salts, i.e.,; sodium chloride 0.65 g, sodium bicarbonate 0.25 g, calcium chloride dehydrate 0.009 g with 100 mL of water at a simulating eye temperature of 34ºC ± 0.5°C. The fixed amount of all formulations, which was 1 g, was added into the dialysis membrane and sealed. The medium was sampled at the prescribed intervals (0, 5, 1, 2, 4, 8, 10, 12 and 24 h). An equal amount of STF medium was removed from it and added to maintain the volume, and the sinking condition was achieved. The BZD concentration in the retained samples was determined using the High performance liquid chromatography (HPLC) method.

The released BZD content was examined using HPLC. Briefly, 25 μL of solvent system was injected into the column before samples were diluted with the same volume. The column used was a C18 column (Zorbax SB, 4.6 mm250 cm, 5 m packing-L1, Agilent). Triethylamine phosphate buffer/acetonitrile/methanol was used as the mobile phase in a ratio of 70/20/10 v/v with a flow rate of 1.5 mL/min. The detector wavelength was 254 nm and the sample run time was approximately 10 min.

The phase behavior of μE formulations consisting of oil, water and surfactant was studied using the ternary phase diagram as shown in Figure 1. Mapping the pseudoternary phase diagram is important in formulating μEs (Rahman et al., 2017; Saleem et al., 2018). It enables the phase compatibility of the ingredients and the average area of ​​μE formation to be determined. Different ratios of oil, S_mix and water were used (Lawrence and Rees, 2012; Subramanian et al., 2005; Rahdar et al., 2018). The behavior of each phase in the μE system was investigated through water dilution as this offers the advantages of fast, accurate and cost-effective operation at room temperature (Takamura et al., 1979). The present study comprised of oil (∼8%), water (∼16%), Tween-80 (∼38%) and 2-propanol (∼38%) as Formulation-A, oil (∼9%), water (∼11%), Tween-20 (∼40%) and 2-propanol (∼40%) as Formulation-B and oil (∼8%), Water (∼12%), Tween-60 (∼40%) and 2-propanol (∼40%) as Formulation-C. Figure 1 represents μE region (shaded area), while AB represent the dilution line and highlighted mark represents the optimal μE selected for further studies on physical stability based on the visual appearance for all formulations. The final formulation contained an excessive proportion of water compared to oil phase. This demonstrates the development of an o/w μE, as supported by previous studies (Saleem et al., 2019; Siddique et al., 2021). Table 1 lists some measured physical parameters of the optimal BZD-free μE and BZD-loaded μE of all μE sets.

Electrical conductivity is convenient tool that determines the type of µE formed and the phase transitions of µE formation from water-in-oil (w/o) to oil-in-water (o/w) through the bicontinuous network channel (Yadav et al., 2018). µEs exhibit a sudden change in electrical conductivity when the composition is varied at a fixed temperature (Kahlweit et al., 1993). Figure 2 represents the changes in electrical conductivity for formulation A, B and C with increasing water content as a function of the weight fraction of the aqueous component (Φw) along the dilution line AB of the oil/S_mix. The bicontinuous channel of formulation A starts at ∼16% Φ_w, which is called the percolation threshold (Φ_p), below this point the w/o µE at Φ_w < 16%. At a Φ_w value of 28%, the abrupt change occurs, indicating the formation of o/w µE with increasing water content and triggering the Φ_b phase transition (Pal et al., 2017; Nazar et al., 2018). As the Φ_w value increases, the first derivative (dσ/dφ) becomes even more helpful in determining the µE domain phase transition (Zavgorodnya et al., 2017). At Φ_w ∼25%, the maximum value of the first derivative was observed, indicating the existence of a stable bicontinuous microstructure in this particular region.

Formulation-B showed the corresponding values of Φw for all suggested transitions in the microstructure are the percolation threshold (Φp ∼ 21%), phase inversion (Φb ∼ 36%), and bicontinuous channel (dσ/dΦ ∼ 30%). Similarly the Formulation-C showed values of Φw for are the percolation threshold (Φp ∼ 7%), phase inversion (Φb ∼ 22%), and bicontinuous channel (dσ/dΦ ∼ 15%).

An optical microscope was used to visualize the variations in microstructure of μE from w/o to o/w via a bicontinuous network channel. Microstructural transitions in µE were studied under a biological microscope. The expected phase changes present in µE containing w/o, o/w, and bicontinuous network are shown along with the proposed microstructure, also sketched in Figure 3 (Nazar et al., 2020). Dispersed spherical oil and water droplets in their respective continuous phases constructing w/o and o/w μE are shown in Figures 3A, C respectively, while Figure 3B depicts a bicontinuous channel formed by a network of spherical droplets. The results agree well with previously reported observations by researchers (Nazar et al., 2018). The oil-rich composition showed the presence of μE with water droplets dispersed in the continuous oil phase (w/o), as shown in Figure 3A. Conversely, water-rich compositions exhibited μE with oil droplets dispersed in water (o/w), shown in Figure 3C. Between these two extremes, a bicontinuous μE emerged, revealing an aqueous phase-dependent microdomains, as shown in Figure 3B. The distinctive microstructures are influenced by the hydrophilic-lipophilic balance (HLB) value of surfactant and co-surfactant (Paria and Khilar, 2000; Rahdar et al., 2018; Qadri et al., 2022b). When the HLB leans towards the hydrophilic side, a two-layer o/w system is formed in which oil is dispersed in the water phase and vice versa (Khan et al., 2016). For μEs in which neither water nor oil droplets form, the term bicontinuous emerged, signifying the percolation behavior (Dasilva-Carvalhal et al., 2003; Rahman et al., 2016).

Size distribution analysis of BZD-free and BZD-loaded μE of all formulations was performed to check the stability of the dispersed partciles using dynamic light scattering (DLS) (Khan et al., 2016). As shown in Figure 4, the particle size of BZD-free μE droplet was 26.3 ± 1.4, 45.3 ± 2.8 and 36.5 ± 1.8 nm for formulation A, B and C, respectively each represented monostructured droplets by a single peak (Soomro et al., 2016). When BZD was loaded into the μE, the average partcile size increased to 36.2 ± 2.5, 53.5 ± 2.5 and 58.2 ± 2.3 nm for formulation A, B and C respectively, indicating the accumulation and encapsulation of BZD inside the interfacial layers of microstructres (Richard et al., 2017; Xiang et al., 2023). The low polydispersity index (PDI <0.3) indicates uniform homogeneity of the droplet distribution of the μE formulation (Khan et al., 2016). Furthermore, the negative ζ-potential values indicated the high colloidal stability, as the highly stable μE has a ζ-potential value of either >30 mV or < −30 mV due to the electrostatic repulsion between the droplets (Nazar et al., 2017; Pavoni et al., 2020). Other researchers have found and reported similar results (Nazar et al., 2018; Siddique et al., 2024). The values of diffusion coefficients D) are listed in Table 1, which were calculated using the Stokes−Einstein relationship ( D = k B T 6 π η r ) at a temperature of 298 K (where r is the hydrodynamic radius of droplets). The higher D value for μE indicated greater diffusion across the membrane barriers, while lower D values show longer retention and are therefore considered responsible for sustained permeation (Erdal et al., 2016; Nakamura et al., 2016; Nazar et al., 2018).

FTIR analysis was performed to check the compatibility of BZD with μE excipients. The FTIR spectra of pure drug, BZD-free and BZD-loaded μE of all formulation are shown in Figure 5. In the FTIR spectrum of pure BZD drug, a signal in the range of 3,110–3,350 cm⁻¹ was observed, probably due to N–H stretching vibrations, while a weak peak in the range of 3,100–3,000 cm⁻¹ is attributed to the C–H stretching mode. The peak observed at 1,644 cm⁻¹ was due to the asymmetric and symmetric C–C stretching vibrations, while a strong band at 1,450 cm^-1 corresponds to the bending of the C–H bond. The FTIR spectrum of the BZD-loaded μE is completely different from that of the BZD-pure powder; however the spectrum of BZD-free μE and BZD-loaded μE showed no significant change in all formulations, indicating that no interaction with μE excipients could be observed, demonstrating the stability of BZD within the microstructures (Schneider et al., 2011; Dinache et al., 2020).

The FTIR spectrum of BZD-loaded μE demonstrates that the BZD is completely dissolved in μE and the peak-to-peak correlation showed that there are no larger aggregates due to the absence of any additional peak (Nazar et al., 2009). However, some slight changes in intensities are observed, which are most likely due to the different molecular environments experienced by BZD in different μE formulations due to the weak physical interactions such as hydrogen bonding or intermolecular forces between the functional groups of the BZD and excipients (Saleem et al., 2019). However, these connections may facilitate sustained release of the BZD.

The steady-state spectrophotometry is widley used to measure the partioning of the drug in microdomains of the μE as it depends on the polarity of medium provided (Lissi et al., 2000; Pal et al., 2011). The emission spectra of BZD in water, IPM, S_mix of all formulations are depicted in Figure 6. The maximum emission (λ_em) of BZD in water was observed at 345 nm, while in IPM it was at 360 nm. The λ_em of BZD in S_mix A, B and C were at 422 nm, 456 nm and 465 nm, while the λ_em of optimal μE-A, μE-B and μE-C was at 408 nm, 406 nm and 404 nm. A red shift was observed in S_mix and optimal μE formulations as compared to water. These results suggests that BZD molecules are firmly contained to the non-polar portion of the interface and are shielded by any pure bulk domain, such as water or oil (Pal et al., 2011). All of these results indicate that micro-aggregate-assembled probes may be found close to the μE interface, where excitation propagation was inhibited and additive rotation is further limited (Nazar et al., 2020; Saleem et al., 2020).

New formulations are created and their efficacy is examined in term of in-vitro release also an effort to reduce the use of animal models (Bachhav and Patravale, 2009; Yasir Siddique et al., 2021). The in-vitro release of BZD from μE was assessed by STF as a release medium that simulated the ocular environment after several interval of time (Wu et al., 2013). The release of BZD was found >99.9% of formulation A in interval of 10 h, while the interval for the formulation B and C was 4 and 7 h demonstrating faster release than formulation A owing various factors, such as viscosity, surfactant, oil phase and drug solubility which can affect the release rate (Siafaka et al., 2015). 2-Propanol tends to lower the surface tension of the surfactant ultimately enhances the drug release as it makes the surface of droplet smooth and dynamic (Figueiredo et al., 2016). In Figure 7, the release rate of BZD from all three optimal μE has shown different BZD release at different time due to change in surfactant. In start, all μE showed a slow release of BZD until 5 h. After that a sudden increase in release rate has been observed (Siafaka et al., 2021).

Formulation A showed ∼35% BZD release at 4 h and rest of the 60% get released round about 10 h. In formulation B, 50% BZD was released at 4 h and rest of the BZD was released at 8 h, similar trend was followed by formulations C. Additionally, formulation A showed a sustained release in 10 h among all 3 μE formulation indicating lower dosage quantity because of sustained release.