The industrial waste biomass, dried E. coli biomass (powder form) was supplied from Daesang Co. (Kunsan, Korea). The raw E. coli biomass was washed using 1M HCl solution for a day in order to remove remaining pollutants such as impurities and nutrient salts, and to protonate the biomass surface. The protonated E. coli was washed via distilled water for 3 times and freeze-dried for 24 h. The CS material (degree of deacetylation: 75~85%) was offered in powder form from SHOWA Co. (Tokyo, Japan) and was used without any modifications. The model anionic dye, RY2 (Cibacron Brilliant Yellow 3G-P, C₂₅H₁₉Cl₃N₉NaO₁₀S₃, MW: 831.02, dye content: 60–70%, and λ_max: 404 nm) and epichlorohydrin (ECH, 99%<) were obtained from Sigma-Aldrich Ltd. (Korea). HCl (35~37%) and NaOH (97%<) were purchased from Samchun Chemicals Co., Ltd (Korea) and Daejung Chemicals & Metals Co., Ltd (Korea), respectively.

CS powder (35 g) was mixed in 5% w/w acetic acid solution (475 g) and stirred at 300~400 rpm for approximately 12 h to obtain uniform CS solution. The prepared CS solution (7% w/w) was left to stand for some time in order to remove air bubbles. The E. coli was mixed with CS solution for biomass-CS suspensions with different biomass:CS ratio to fabricate biomass composite CS beads (BCCB). However, at a ratio exceeding 7:3 of biomass:CS, the viscosity of the mixture was too high to fabricate BCCB. In addition, as mentioned in introduction, by maximizing the biomass content of biomass-CS composite sorbent, the used amount of CS can be reduced, and price of the sorbent can be decreased accordingly. Therefore, the E. coli biomass content for BCCB was chosen 70% which the highest biomass ratio can make sorbent as a bead type in this study. To prepare BCCB containing biomass content 70%, the E. coli biomass (3 g) was mixed in 18 g of 7% CS solution for 10 h. The biomass-CS mixture was dropped into 500 mL of 1M NaOH solution to form beads by using a 300 μm inner-diameter needle tip. The BCCB were thoroughly washed with distilled water, then 10 g_wet of BCCB were cross-linked in 1M NaOH solution (100 mL) containing 2 mL of ECH. The cross-linking reaction conducted for 30 min at room temperature (25°C). Afterward, the cross-linked BCCB were washed for several times via distilled water and freeze dried. The pristine CS beads (CSB) were prepared by dropping 7% w/w CS solution in 1M NaOH solution. Thereafter, CSB were cross-linked by the same procedure as that of BCCB. The dried CS-based beads showed approximately 1.5~2 mm of bead sizes.

The RY2 sorption performances of E. coli biomass, CSB and BCCB were evaluated in a batch system. To determine the pH effect on sorption performance, pH edge experiments were performed. For this, 0.03 g of E. coli biomass, CSB and BCCB were each mixed in 30 mL of RY2 solution in conical tubes. The initial concentrations of RY2 solutions for pH edge tests using E. coli and CS-based beads were fixed at 335 and 1,313 mg/L, respectively. To evaluate the maximum sorption capacities (q_m) of sorbent, isotherm experiments were performed. For this, 0.03 g of sorbents were agitated in the 30 mL of RY2 solution which have different initial concentrations (0~1,313 mg/L). During the experiments, the solution pH of samples was controlled to the desired pH ranges by using 1M NaOH and 1M HCl solutions (pH effect: pH 2~11, isotherm: pH 3.95~4.05 (pH 4)). The samples were agitated for about 96 h in an auto-shaker at 140 rpm and 25°C. For kinetic experiment for RY2 adsorption, 0.1g of CSB and BCCB was continuously mixed into the 100 mL of RY2 solution (initial concentration: 2,021 mg/L, initial pH: pH 4) at 200 rpm by magnetic stirring. After then, the solution pH was continuously monitored and maintained as pH 4 using 1M NaOH and HCl solutions for approximately 92 hours. During kinetic experiment, RY2 solution samples were taken at different adsorption time. To measure remaining RY2 concentration of adsorption experimental samples, the samples were liquid-solid separated by centrifugation at 9,000 rpm. The RY2 concentration in the separated supernatants was measured using a UV spectrophotometer (UV-2550, Shimadzu, Kyoto, Japan) at 404 nm. The RY2 uptakes of applied sorbents were calculated using the following mass balance equation:

where q (mg/g) is RY2 uptakes of sorbents. V (L) and C (mg/L) are experimental volume and concentration of RY2, respectively. The subscripts i and f indicate the initial and final state, respectively. The final volume is calculated as sum of the initial volume and added volume of acid and/or base solution for pH control. M (g) is the mass of the adsorbents.

In addition, RY2 recovery rate could be calculated by the following equation.

To compare functional group properties of E. coli biomass, CSB, and BCCB, the infrared spectra properties of the sorbents were obtained from Fourier transform infrared spectrometer (FT/IR-4100, Jasco, Japan). The FT-IR analyses were conducted using KBr-sorbent pallets The FT-IR spectrum of each sample was measured in the wavelength range of 700–4,000 cm¹.

To compare morphological properties of CSB and BCCB, the SEM observation was conducted. Before SEM observation, the surface of CSB and BCCB were coated by platinum (Pt) in a vacuum chamber for 2 mins (8 nm coating). Then, the Pt-coated CSB and BCCB were analysis using an SEM instrument (Field emission scanning electron microscope, JEOL-7610F-Plus, Japan) at ×40 magnification.

The functional groups property of a sorbent is an important factor to confirm biosorption performance of biosorbents. Therefore, the FT-IR analysis of E. coli biomass, CSB, and BCCB were carried out to determine properties of their functional groups. The observed FT-IR spectra of sorbents are displayed in Fig. 1.

The FT-IR spectrum of the E. coli biomass in Fig. 1(a) showed various FT-IR peaks at 3,600~3,200, and around 2,925, 1,735, 1,645, 1,538, 1,453, 1,408, 1,233, and 1,075 cm⁻¹, which representing different kinds of functional groups on the surface of the biomass. Adsorption peaks in range of 3,600~3,200 cm⁻¹ indicate the O-H of hydroxyl group and N-H asymmetric stretching of the amines on the E. coli biomass [20, 21]. The peaks at around 2,925 cm⁻¹ are characteristic often seen due to presence of the -CH₂ and -CH₃ [22]. The peak at 1,735 cm⁻¹ indicates carbonyl stretching of un-ionized carboxylates [23, 24]. The strong peak at 1,645 cm⁻¹ are associated with the carboxyl groups on the surface of E. coli biomass [25]. The peak at 1,538 cm⁻¹ might attributed to the amide I and the carboxyl group (C=O chelate stretching) and amide II bonds [26, 27]. The peaks at 1,453 and 1,408 cm⁻¹ indicate the alkanes [28] and symmetrical stretching of carboxylic acids, respectively [29]. The peak at 1,233 cm⁻¹ indicate C-O stretching of the carboxyl groups [30]. In addition, the phosphonate group (P-OH stretching) of the E. coli biomass could be identified by the 1,075 cm⁻¹ peak [31]. In case of CS bead, shown in Fig. 1(b), the main spectrum peaks were obtained at around 3,462, 1,656, 1,589, 1,473, 1,380, 1,152, 1,071, and 1,032 cm⁻¹. The strong and broad band at around 3,462 cm⁻¹ corresponds to the N-H and O-H stretching vibrations of chitosan [32]. The peak at 1,656 cm⁻¹ is attributed to the secondary amide C=O bond that is contributed by the acetamide groups [33]. The peaks observed at 1,589 and 1,473 cm⁻¹ are attributed to -NH bending vibration in -NH₂ and -NH deformation vibration, respectively [34, 35]. The peak at 1,380 cm⁻¹ indicates aliphatic C-H stretching [36]. The peak at 1,152 cm⁻¹ is due to -CN stretching vibration [37]. Meanwhile, the peaks at 1,071 and 1,032 cm⁻¹ are due to the -CO stretching in COH [37, 38]. The FT-IR spectrum of BCCB showed the characteristic of E. coli biomass and CS, because BCCB is a composite of them (Fig. 1(c)). From FTIR spectrum of BCCB, the peak at 3,200 cm⁻¹, which is typical of CS, was enhanced after compositing CS with E. coli biomass. In addition, a joint peak emanating from the biomass (1,538 cm⁻¹) and CS (1,589 cm⁻¹) was observed at 1,560 cm⁻¹, and may indicate -NH bending vibration in -NH₂.

As a result of pH effect (Fig. 2), the RY2 sorption capacities of E. coli biomass, CSB and BCCB were highly affected by solution pH. As the pH increased, significant decreasing of RY2 sorption capacity of E. coli biomass was shown until around pH 6 and finally reached zero above pH 7. From the FT-IR analyses, possession of various anionic and cationic functional groups (carboxyl, phosphonate, and amine groups) on the E. coli biomass were confirmed. The amine groups on the biomass can be protonated and positively charged (-NH₃⁺) at pH less than their pK_a values (pK_a 8~11) [39]. Therefore, the negatively charged RY2 molecules in aqueous solution can be bound to amine groups at low pH by electrostatic attraction between positive amine groups and negative RY2 molecules. On the contrary, the pK_a values of carboxyl groups and phosphonate groups on the biomaterials are between 3.5~5.0 [40] and 6.1~6.8, respectively [39, 41]. Thus, the carboxyl and phosphonate groups are changed into negative charges by deprotonation of functional groups with increasing pH. The negatively charged functional groups can interrupt the adsorption of RY2 ions onto the biomass by electrostatic repulsion. Therefore, the RY2 sorption capacity of E. coli biomass decreased with increasing pH. The pH effect for RY2 sorption of CSB and BCCB also showed similar pH dependence like E. coli biomass. The RY2 uptake of CSB was higher than that of BCCB in the range of pH 2 to pH 6, however the RY2 uptake of CSB and BCCB were similar above pH 6. The matrix of CSB and BCCB consist of chitosan which has large number of positive amine groups as main functional groups. However, the pK_a value of amine group in the chitosan structure were reported as 6.4 [42], and the positively charged amine groups (-NH₃⁺) of CSB and BCCB may significantly lost their positive charges by deprotonation at pH around 6, similar to their pK_a values. Moreover, the performance of BCCB for RY2 was additionally influenced by electrostatic characteristics of the anionic functional groups in the composed E. coli biomass. Hence, the RY2 uptakes of CS-based beads, CSB and BCCB, decreased with increasing pH. Although the RY2 sorption capacity of sorbents can be maximized under acidic condition, the CS-based beads cannot be stable. Therefore, considering the stability of CS-based beads, the subsequent biosorption experiments were conducted at pH 4.

To evaluate the maximum RY2 uptakes of the E. coli biomass, CSB, and BCCB, isotherm experiments were carried out at pH 4. The CS-based beads were stable at solution pH 4 without CS matrix loss.

As shown in Fig. 3, the RY2 uptakes of sorbents increased with increasing initial concentration of RY2 and reached to the saturated states. To determine detail maximum sorption capacities of sorbents, the Langmuir isotherm equation [43] was applied to the experimental isotherm results of sorbents. The Langmuir isotherm model equation can be presented as follow:

where q is the experimental RY2 sorption capacity of sorbents (mg/g), C_e (mg/L) is the RY2 concentration at sorption equilibrium, q_m is the maximum sorption capacity (mg/g) of sorbent and b (L/mg) is a Langmuir constant [44]. The estimated parameters from Langmuir model equation are described in Table 1.

According to table 1, the maximum RY2 uptakes (q_m) of E. coli, CSB, and BCCB at pH 4 were 200.77±4.51, 934.71±50.97, and 679.13±23.76 mg/g, respectively. The Langmuir constants (b) were estimated to be 0.962±0.111, 1.346±0.262, and 1.274±0.219 L/mg for E. coli biomass, CSB, and BCCB, respectively. The RY2 uptake of CSB was about 1.4 times higher than BCCB. In addition, the RY2 uptake of BCCB which containing 70%_w/w of E. coli biomass in the bead matrix was 3.38 times higher RY2 uptake than raw E. coli biomass. The higher RY2 uptake of CS-based beads than E. coli biomass may be due to large amounts of amine groups (-NH₃⁺) located in the chitosan structure which can adsorb RY2 molecules by electrostatic interaction.

Considering the maximum RY2 uptake of E. coli biomass (200.77±4.51 mg/g), CSB (934.71±50.97 mg/g) and composite ratio of E. coli biomass (70%) and CS (30%) in the BCCB, the expectable maximum uptake of BCCB can be estimated to be 420.05 mg/g. However, the experimental maximum RY2 uptake of BCCB (679.13±23.76 mg/g) was higher than the expectable RY2 maximum uptake. For CS-based beads, it can assume that almost all the binding sites for RY2 sorption are possessed in the CS matrix. It indicates that the sorption efficiency of CS for BCCB was higher than that of CSB for RY2. Hence, comparison of the RY2 uptakes based on CS contents in CS-based beads was carried out. qkr

The practical RY2 adsorption capacity (q_p) indicated the RY2 uptake by CS in the beads. Therefore, the practical RY2 uptakes of CS-based beads were calculated from the isotherm experimental data of them by following simple mass balance equation. The CS ratios in the CSB and BCCB was 1 and 0.3, respectively.

The comparison of practical RY2 uptakes between the CSB and BCCB are shown in Fig. 4. The maximum q_p of the CSB (934.71 mg/g) was same as the result of isotherm for the CSB (Fig. 3). This is because the original composition of the CSB was purely CS. However, the maximum q_p value of BCCB was calculated as 2,276.71±107.09 mg/g based on the amount of CS in the bead (30%_w/w chitosan) by Langmuir equation, and this was 2.4 times higher than that of the CSB. It might be due to RY2 transfer interfering into the CSB by more dense structure of CSB compared to BCCB. It can be supported by SEM observation results of CSB and BCCB (Fig. 5). In the SEM observation results, it was found that the CSB shows dense and strict morphological property compared to that of BCCB. BCCB shows relatively rough surface. In addition, large pores can be found, which might be formed by mixed air bubbles during mixing of E. coli biomass with CS solution for bead fabrication. Although the CSB contained a lot of amine groups compared to BCCB, the rough surface and pores of the BCCB might contribute to easier and more efficient RY2 contact to the CS molecules in the BCCB. The RY2 sorption kinetic data also might support this phenomenon. The results of kinetic experiments using CSB and BCCB were displayed in Fig. 6. To detail compare of kinetics of CSB and BCCB, pseudo-2^nd model which can represented as below was applied on the experimental kinetic data.

where the q_t is experimentally determined RY2 sorption capacity of sorbent at sorption time (t, min). The terms of q_e2 and k₂ indicate the RY2 sorption capacities of sorbents (mg/g) at sorption equilibrium and rate constant (g/mg·min) calculated from pseudo-2^nd order model, respectively. From pseudo-2^nd order model (R² values for CSB and BCCB are 0.9789 and 0.9351, respectively), although the RY2 sorption capacities of CSB at equilibrium (q_e2) was calculated as higher value (1,214.28±52.24 mg/g) than BCCB (794.51± 64.16 mg/g), the rate constant (k₂) of BCCB ((1.511±0.431)×10⁻⁶ g/mg·min) was higher than that of CSB ((0.976±0.149)×10⁻⁶ g/mg·min). This higher rate constant of BCCB might be attributed by rough and porous morphological properties of BCCB.

In conclusion, it was found that chitosan can be used more effectively for RY2 sorption process when biomass was composited in the bead. In addition, the effective usage of CS can give the cost-benefit to CS-based sorbents. The main factor of cost for making the CS-based beads may be dependent on the amount of used CS, not E. coli biomass because the biomass is a waste material (zero-cost). Therefore, by blending waste E. coli biomass with CS, it is expected that the price of BCCB can be reduced to 30% compared to case of only CS use.

Based on the maximum sorption capacities of the sorbents, the adsorption process for RY2 treatment will require 1.4 times more BCCB than CSB when same amount of RY2 is to be treated. However, according to the amount of CS used for CS-based beads, the sorbent price of BCCB is cheaper than that of CSB. In other words, 58% of process cost for RY2 wastewater treatment can be reduced when BCCB is used as sorbent (Table 2), and as the main cost factor for RY2 treating process is price of sorbents, this is expected to create a huge economical boost in terms of cost reduction.