Both batch PBR and continuous DMPBR system were made of plexiglass with dimensions of 32 cm in height, 32 cm in length, and 30 cm in width, and a working volume of 24 L. It is illustrated schematically in Figure 1. A sparger connected to an air pump with an aeration rate of 20 L min^-1 is installed at the bottom of the batch PBR and DMPBR. In this study, a Kubota flat sheet membrane (Osaka, Japan) was immersed in the DMBPR. Table 2 presents the characteristics of the membrane.

The DMPBR operated continuously at room temperature during which synthetic wastewater was treated through the dynamic membrane. The membrane had a filtration area of 0.11 m², and the system was operated at a flux of 18 L m^-2. h^-1, resulting in a permeate flow rate of 1.98 L h^-1. The hydraulic retention time (HRT) was approximately 12 h.

The microalgae used in this experiment was Chlorella vulgaris, a kind of round-shaped green algae well known for its dye-removal capability from dye-contaminated wastewater (Chin et al., 2020). The microalgae bank of the Science and Research Branch of Islamic Azad University in Tehran, Iran, provided C. vulgaris. The culture medium used for its growth was based on BG-11 (Emamshoushtari et al., 2022).

The C. vulgaris inoculation into 30-L batch PBR, and DMPBR was 5% v. v⁻¹ with a dry weight (DW) of about 0.5 g. L^-1. The microalgae were exposed to a light-dark regime 24:0 under white LED lamps with an intensity of 3000 lux. When DW reached 2.5 g. L^-1 in both batch PBR and DMPBR systems, the experiments regarding the Maxilon red dye removal were initiated.

DM was formed on the surface of the membrane in the DMPBR. As the microalgae-dynamic membrane forms, the transmembrane pressure (TMP) steadily increases; this would indicate the formation of microalgae as a cake layer on the membrane’s surface. It was determined that if the TMP reaches a threshold of 200 bar, the formation of microalgae DM was assumed to be complete.

Maxilon Red dye, also known as Astrazon Red FBL; C.I. Basic Red 46; Cationic Red GRL, its chemical structure is 1,2-dimethyl-3-((4-(methyl (phenylmethyl)amino)phenyl)azo)-1,2,4 thiazolium bromide (Farouq, 2022). The dye concentration range in textile wastewater was selected between 10 and 50 mg L^-1 based on values reported in the literature (Yaseen and Scholz, 2019).

A microscopic analysis was conducted on the morphology of C. vulgaris before and after the removal of dye. Microscopic images were captured using an acceleration voltage of 25 kV and magnifications of 500. These images were then analyzed using a SEM.

Fourier-transform infrared (FTIR) was adopted to determine the functional groups that contributed to the adsorption of Maxilon Red dye by C. vulgaris, using a Thermo Scientific Nicolet NEXUS 870 FT-IR spectrometer (United States). The samples of microalgal biomass were collected before and after dye adsorption, washed with deionized water to clear off any residual dye or culture medium, freeze-dried, and ground into powder. An amount equal to 1 mg of the sample was mixed in with 100 mg of spectroscopic-grade potassium bromide (KBr) pressed onto a clear pellet in a hydraulic press. FTIR spectra spanning a 400–4000 cm^-1 range were recorded at a resolution of 4 cm^-1, averaging 32 scans per sample.

The dye removal efficiency was determined according to the literature (da Rosa et al., 2018). Briefly, a sample is taken from the system and filtered using filter paper to separate microalgae from the solution. Hence, cotton filters Whatman 1442–125, whose pore size is 2.5 µm, have been used. A spectrophotometer (Hack DR6000 UV VIS) was used to measure the optical absorption of the filtered sample at a 554 nm wavelength. The dye removal efficiency was obtained according to Equation 1: R = C 0 − C i C i × 100 (1) where R is the color removal efficiency (%), and C ₀ and C _i are the initial color concentration and the color concentration after filtration (mg. L^-1), respectively.

Moreover, microalgae DW was measured according to the literature (Shirazi et al., 2024). 50 mL of each sample were centrifuged at 10,000 g and 5°C for 20 min to separate the culture medium from the biomass. It was dried at 80°C for 24 h and then weighed.

The mechanism of biosorption can be clarified through kinetic studies. Once the kinetic parameters are determined, they provide essential information for reactor design, particularly concerning reactor volume and residence time. Therefore, identifying the kinetic model that best describes the system is critically important.

The equilibrium adsorption capacity of Cr (III) for each sample was calculated using the following equation: q e = C 0 − C e V m (2) where q e ​ shows the biosorption capacity at equilibrium (mg.g^-1), V is the volume of the solution ( L ), C ₀ is the initial concentration of the dye (mg.L^-1), C e ​ stands for the equilibrium concentration of the dye (mg.L^-1), and m is the mass of the adsorbent (g).

The biosorption capacity of Cr (III) at various time intervals was quantified using Equation 3. q t = C 0 − C t V m (3) where, q t is the biosorption capacity (mg. g^-1), and C _t is the concentration of dye in each time step interval.

The statistical analysis of the results was conducted using analysis of variance (ANOVA). Differences between means were considered significant when the p-value was less than or equal to 0.05. All the experiments were repeated three times at room temperature (25°C ± 2°C).

The biosorption process by C. vulgaris was estimated using different Maxilon red concentrations using the Langmuir and Freundlich isotherm models. The red dye initial concentration (C₀) of 5, 10, 15, 20, 25, and 30 mg. L^-1 reached the Ce of 1.32, 2.1, 2.2, 5.4, 9.5 and 14.1 mg. L^-1, respectively.

The Langmuir adsorption isotherm describes the monolayer biosorption of various metals onto bio sorbents. The Langmuir model is the most commonly used model for characterizing heterogeneous metal biosorption. The Langmuir adsorption isotherm is given in the following equation (Equation 8) (Pradhan et al., 2019). C e q e = 1 k l q max + C e q max (8) where, k l represents the Langmuir binding constant (L. mg^-1), q max is the maximum amount of Cr(III) adsorbed per unit weight of adsorbent (mg. g^-1). Consequently, a linear graph of C e q e vs C e using experimental data illustrates the Langmuir isotherm parameters (slope = 1 q max , and intercept = 1 k l q max ). The Langmuir isotherm was then analyzed through the interpretation of the Langmuir equilibrium parameter (R _L ), as given in Equation 9 (Maleki et al., 2015). R L = 1 1 + k l C 0 (9)

The value of R _L defines the nature of the biosorption process, it can be classified as below: unfavorable (R _L > 1), linear (R _L = 1), favorable (0 < R _L < 1), or irreversible (R _L = 0) (Sathvika et al., 2016).

The Freundlich model is understood to assume multilayer adsorption on heterogeneous surfaces. The model assumes that the adsorption sites are occupied in multilayers by the binding capacities. The model is expressed by the following Equation 10 (Pradhan et al., 2019). log q e = log ⁡ k f + 1 / n ⁡ log ⁡ C e (10) where, k f is the constant of Freundlich isotherm (mg.g^-1) which is related to adsorption capacity, and 1 n is another constant related to the intensity of biosorption. 1 / n is a value that characterizes the practicability of the isotherm, e.g., irreversible ( 1 / n = 0); favorable (0< 1 / n <1); unfavorable ( 1 / n > 1). By plotting log   q e against log   C e using experimental data, 1 / n and log ⁡ k f will be realized as the line slope and intercept, respectively. Furthermore, the calculated parameters of the Langmuir and the Freundlich isotherm models are shown in Table 4.

The poor value of R ² of the Freundlich model (0.653) in comparison with the R ² value for the Langmuir model (0.908) shows that the Langmuir model generated a more accurate fit to the experimental data. Meaning that adsorption occurs in a single layer on a homogeneous surface with a finite number of specific adsorption sites. Once the aforementioned sites are filled, no further adsorption can happen, which suggests that the adsorbent has a maximum capacity, represented as q max . The model is predicated on the assumption that there is no interaction between adsorbed molecules and that all adsorption sites have the same capacity for the adsorbate. This makes it suitable for uniform surface adsorption processes (Pradhan et al., 2019). The resulting values for R_L were 0.18 for the C ₀ of 15 mg. L^-1. The results of the calculations suggest that the Maxilon red dye biosorption process onto the microalgae biomass was a favorable one.

The optimal concentration for maximum dye removal in the PBR system as batch mode of operation was determined to be 15 mg. L^-1. This approach uses the same ideal dye loading conditions to evaluate the performance of batch PBR and continuous DMPBR. Figure 4 represents the removal efficiency of Maxilon Red when DMPBR is used, which is the continuous mode. The removal efficiency of dye removal is almost constant during the 10 h of operation and almost full dye removal can be reached 98% for initial dye concentrations of 15 mg. L^-1 Utilizing a DM offers a distinct advantage, as it ensures the maintenance of physical separation even after achieving equilibrium in the batch PBR after 2.5 h. This characteristic enhances the overall efficiency of the biosorption system, surpassing the efficiency of a single biosorption system. This suggests a potential for synergistic use of suspended biosorbent and microalgal DM in a DMPBR. These findings suggest that the DMPBR system’s synergy with suspended biosorbents and a dynamic membrane optimizes overall performance despite fouling’s limitations. Statistical analysis showed a significant effect on the dye removal efficiency by incorporating a DM into the system (p-value˂0.05).

When comparing these results with the ones from the literature (Berkessa et al., 2020), they had reached almost 98% of Remazol Brilliant Blue R (∼50% pure) removal in a reactor with working volume of 10 L. The dynamic membrane, with total filtration area of 0.01 m², was immersed in the cylindrical side stream reactor with a volume of 4 L (Berkessa et al., 2020).

Although the present results show that DMPBR cannot remove the dye much more than anaerobic dynamic membrane reactor and living membrane bioreactor (almost 98% for all cases). However, it can be argued that the potential of the current research is its green nature, meaning that it is less harmful to the environment and produces no dangerous chemicals.

The spectral analysis was conducted within the range of 400 to 4,000 cm^-1. It was determined that the cell walls of microalgae contain a variety of chemical groups, including hydroxyl, carbonyl, and sulfhydryl, and it is the responsibility of these proteins to determine the adsorption ability of the microalgal cell (Pradhan et al., 2019). As illustrated in Figure 5, the FTIR spectra demonstrates Chlorella vulgaris microalgae before and after Maxilon Red dye adsorption.

The broad O-H stretching band at 3425 cm^-1 shifts to 3480 cm^-1 with reduced intensity after dye adsorption, indicating the interaction of hydroxyl groups with the dye molecules, likely through hydrogen bonding (Chin et al., 2020). A shift of the C=O stretching peak from 1635 cm^-1 to 1665 cm^-1 suggests that carbonyl groups, possibly from ketones, aldehydes, carboxylic acids, primary amides, and esters, which are involved in the dye binding (Peng et al., 2015). The peak at 2100 cm^-1 wavenumber generally corresponds to the stretching of a carbon-carbon triple bond (C≡C), typical of alkyne groups, although it can also be associated with nitrile groups (C≡N) in some cases. Upon adsorption of Maxilon red dye, this peak disappeared, suggesting a significant interaction between the dye and the functional groups responsible for this absorption. The 710 cm^-1 peak is the most significant in the fingerprint region and correlates with aliphatic chloro compounds, C-Cl stretch, and has disappeared due to dye adsorption. This phenomenon can result in a modification of the vibrational frequencies of the functional groups present on the surface of the biomass. The binding of sorbate to functional groups, such as carboxyl (-COO-), hydroxyl (-OH), amino (-NH₄), or phosphate (-PO₄ ⁻³), has the potential to alter the associated peaks in the fingerprint region resulting in a decrease in wave number (Tattibayeva et al., 2022). The peak at 1040 cm^-1, associated with C-O stretching vibrations in polysaccharides or C-N stretching in proteins (Croué et al., 2003), showed a potential increase in intensity after dye adsorption. This could be due to the surface adsorption of dye molecules, hydrogen bonding interactions, or the formation of weak new C-O bonds resulting from the interaction of dye molecules with the polysaccharides and proteins in the microalgae cell wall (Pradhan et al., 2019). These results underscore the complexity of the adsorption process and reinforce the potential of Chlorella vulgaris as a promising biosorbent for dye removal.

Figures 6a,b shows SEM images of C. vulgaris before dye removal, and with adsorbed dye on the microalgal dynamic membrane, respectively, with magnification of 1000X.

SEM micrographs (Figure 5) display the morphological changes. Raw Chlorella vulgaris cells have smooth, spherical surfaces before dye adsorption (Figure 5a). On the contrary, the microalgal DM after dye adsorption shows a heterogeneous, porous, biofilm-like layer packed with irregular aggregates after dye treatment (Figure 5b). This texture and multilayered structure result from the deposition of microalgal cells, extracellular polymeric substances (EPS), and adsorbed dye particles. The participation of the hydroxyl and carbonyl groups in the FTIR shows that the EPS matrix forms the cake layer that forms the DM’s physical barrier in addition to capturing dye molecules (Liao et al., 2018). By overcoming equilibrium constraints and continuously renewing binding sites within the EPS matrix while maintaining a structured cake layer, the DMPBR enhances contact between biomass and dye, which explains why it outperforms the batch PBR (Diaz-Uribe et al., 2021).

DMPBR offers a promising method for a circular economy by considering wastewater not as a pollutant but as a resource (Goh et al., 2022). Textile wastewater contains effluents such as nitrate and phosphate, which would benefit microalgae growth, enabling simultaneous wastewater treatment, Maxilon red dye removal, and biomass generation. The dynamic membrane formed on the membrane’s surface improves wastewater removal and offers a low-energy alternative to conventional PBR systems (Liao et al., 2024). This study shows that DMPBRs are highly effective for Maxilon red dye removal. Moreover, the photosynthetic-based nature of microalgae would allow the occurrence of CO₂ capturing at the same time as microalgae cells require CO₂ for growth (Liao et al., 2024).

Furthermore, by optimizing resource use and reducing waste, the biomass produced in the DMPBR may be sustainably recycled into high-value products, supporting the zero-waste biorefinery concept. One important use is creating thin-film composite (TFC) membranes, which provide an economical and sustainable substitute for conventional materials. These membranes, which are made from algae biomass cultivated in wastewater, exhibit durability and robustness for a range of industrial applications, including organic solvent nanofiltration (OSN) systems, thus promoting the ideas of green chemistry. Furthermore, the biomass can be utilized to create carbonaceous anodes for lithium-ion batteries and optically pure (R)-γ-valerolactone ((R)-GVL) for biopharmaceuticals and bioplastics, which will help create more sophisticated energy storage systems and lower CO₂ emissions (Yang et al., 2023; Cha et al., 2024).

However, the feasibility of dye recovery remains a challenge. The reason is that most dyes (such as Maxilon red) undergo structural transformation or degradation during the treatment period, making them unusable for reuse; nevertheless, although partial recovery is theoretically possible, whether this recovered dye is efficient or economically logical needs further research and discussion (Sarioglu and Aşkal, 2018; Pimentel et al., 2023). Therefore, it is recommended that DMPBRs be used for resource recovery, pollutant neutralization, and wastewater treatment rather than dye recovery (Goh et al., 2022; Liao et al., 2024).