## 2.1. Chemicals and reagents

Hydrochloric acid (HCl) 37% (Fluka, Germany), phosphoric acid (H3PO4), chitosan (CS-deacetylation degree: 92.0 wt-%; 1.0 × 106 Da molar mass), acetic acid (AAc) (99% purity), poly (vinyl alcohol) (PVA, hydrolysis degree of 99.0–99.8%) (Sigma Aldrich), gluteraldehyde (pentane-1, 5-dial) (India), graphite powder (purity ≥ 99.9%) (Changsha, China), sodium nitrate (NaNO3) (98%, Nice chemicals),potassium permanganate (KMnO4)(99%, RFCL),potassium dichromate (K2Cr2O7) (Carelabmed, India), sulfuric acid (H2SO4) ( 95–97%, (Fluka, Germany), sodium hydroxide (NaOH) assay > 97% (RPE, ACS-ISO for analysis), ammonium persulfate (NH4)2S2O8) assay = 99% ( BDH, England), 1,5- diphenylcarbazide (C13H14N4O assay > 99% (Sigma Aldrich, India) and hydrogen peroxide (H2O2) (30% ) were used.

## 2.2. Materials syntheses

**Graphene oxide (GO)** was made by Hummers method [22]. Briefly, 1.0 g of graphite was added into 9:1 (H2SO4 : H3PO4) concentrated solution in a flask, then 6 g of KMnO4 was added. The resultant mixed solution was heated to 50 ℃ and allowed to react for 6h under mechanical stirring and ultrasonication. Upon completion of the reaction, the flask was put into an ice bath containing H2O2 and enough ice water for 1h. Then the filtrate was calm and centrifuged at 5000 rev/min and washed with distilled water until pH 7.0, followed by coagulation with ethanol and vacuum drying.

As the procedure reported in the literature, water hyacinth leaf protein concentrate (WHLPC) was synthesized [23]. Water hyacinth leaves were collected from Lake Tana at Bahir Dar city, Ethiopia. First, the collected leaf material was destalked and washed in running tap water until it was clean. The water hyacinth leaves were then soaked in water at a 2:1 ratio for 30 min then macerated using a blender. Sodium hydroxide (NaOH, 0.1 M) was added to the slurry until reaching pH 9.0 allowing the solubilization of leaf proteins. The tissue slurry was then filtered through cheesecloth and the filtrate was collected. The protein from the filtrate was coagulated by adding 0.1 M HCl until pH 2.0. The coagulum was subjected to 80°C for five minutes to form large protein clumps that could be separated by filtration and oven-dried at 60°C.

**WHLPC/GO/CS/PVA** composites were synthesized following the procedure reported in the literature[24] where 250 mg of the prepared GO powder was mixed in 125 mL distilled water and followed by sonication for 30 min. 500 mg Chitosan was mixed homogeneously in 12 ml (2% acetic acid) until a clear solution was observed. 500 mg PVA was dissolved in 4.5 ml distilled water and heated at 90 oC for 30 min. Then, 4g WHLPC, 50 mL of the prepared GO suspension, 4.5 ml PVA solution, 12 mL CS solution, and 300 µL GA solution (25%) was mixed under ultrasonic dispersion for 1 h at 25°C. The resulting mixture was dropped into a gently stirred 5% NaOH for 48 h to form a stable cross-linked WHLPC/GO/CS/PVA composite. Finally, the composite was washed with ultrapure water several times until residual reagents were removed and oven-dried at 60 oC for 12 h.

## 2.3. Characterization of the WHLPC/GO/CS/PVA

X-ray diffraction is the most widely used technique for general crystalline material characterization. It is used to measure the average spacing between layers or rows of atoms and determine the orientation of a single crystal or grain. The XRD graph obtained for CS, synthesized GO, PVA, WHLPC, and WHLPC/GO/CS/PVA are shown in Fig. 1. The d-spacing of the most intense peaks was calculated by using Bragg's relationship.

\(n\lambda =2d\sin \theta ................(14)\)

Where λ is X-ray wavelength, n is an integer and θ is the angle between the incident and reflected rays.

The chemical composition of CS, synthesized GO, PVA, WHLPC, and WHLPC/GO/CS/PVA was analyzed using Spectrum 65 FT-IR (Perkin Elmer) in the range of 4000 − 400 cm− 1 (resolution: 4 cm− 1, number of scans: 4) using KBr pellets.

## 2.4. Adsorption experiments

The adsorption experiments were conducted in an Erlenmeyer flask with a capacity of 25 mL containing 10 mL of aqueous Cr(VI) solutions of different concentrations and adsorbent dosage of 30 mg of WHLPC/GO/CS/PVA. The solution pH was optimized and maintained at pH 1.0using 0.1 M HCl. The sample flasks were placed in an orbital shaker for agitation at 200 rpm in the time range of 5 to 420 min. Subsequently, the adsorbent was separated by filtering the respective solutions using 0.22 mm Whatman filter paper. Thereafter, the Cr(VI) content in the supernatant liquid was evaluated by using a UV–visible spectrophotometer. The maximum absorbance measured at 540 nm corresponds to the formation of the red-violet chromophore complex due to the reaction between Cr(VI) and 1,5-diphenyl carbazide in an acidic medium, and the maximum intensity of absorbance (350 nm) due to yellow color of high concentration Cr(VI)[25]. Accordingly, the adsorption capacity of Cr(VI) was calculated as follows:

\({q_t}=\frac{{({C_o} - {C_t})V}}{m}.................(1)\)

Where Co and Ct indicate the initial concentration of Cr(VI) and the concentration at any given time (t), respectively, m is the mass of the adsorbent in gram, and V (L) is the volume of the Cr (VI) solution. The qt represents the adsorption capacity at any time. The % removal of Cr(VI) was also calculated as follows:

\(\% \,\,Removal=\frac{{({C_o} - {C_t})}}{{{C_0}}} \times 100..........(2)\)

The reproducibility of the results was checked by triplication of every adsorption experiment and was found to be within acceptable limits. The effects of pH, contact time, adsorbent dose, temperature, the adsorption isotherms, and kinetics were explored for the removal of Cr(VI).

## 2.5. Kinetic studies

For adsorption kinetic tests, 30 mg WHLPC/GO/CS/PVA was added to 10 mL Cr(VI) solutions (100 mg/L) at a pH of 1.0 for 5–420 min. The kinetics behavior was studied by fitting into the pseudo-first-order and pseudo-second-order equation equations given by Eqs. 3–4[26]:

\(\log \,({q_e} - {q_t})=\,\log \,\,\,{q_e} - \frac{{{k_f}}}{{2.303}} \times t\,..............(3)\)

where qt (mg/g)is the amount of adsorbate adsorbed at time t(min), qe (mg/g) is the adsorption capacity in equilibrium, and kf (min− 1) is the rate constant.

\(\frac{t}{{{q_t}}}=\frac{1}{{{k_s}q_{e}^{2}}}+\frac{1}{{{q_e}}} \times t............(4)\)

where kS (g/mg min) is the rate constant of pseudo-second-order adsorption, and the initial adsorption rate h can be regarded as the initial adsorption rate as qt/t→0, hence: h = kSqe2.

## 2.6. Adsorption isotherms

Adsorption isotherm experiments were carried out by shaking Cr (VI) solutions of different concentrations (100–2000 mg/L) with 3.0 g/L WHLPC/GO/CS/PVA per 330 min at pH 1.0. The experimental data were then analyzed by fitting into the different adsorption isotherms described below:

The Langmuir isotherm is represented by the following equations [27]:

\(\frac{1}{{{q_e}}}=\frac{1}{{{K_L}{q_m}}}x\frac{1}{{{C_e}}}+\frac{1}{{{q_m}}}...............(5)\)

Where qe (mg/g) is the equilibrium adsorption capacity, Ce (mg/L) is the concentration in the liquid phase at equilibrium, KL (L/mg) is the Langmuir adsorption constant related to the energy of adsorption and qm (mg/g) signifies adsorption capacity.

\({R_L}=\frac{1}{{1+{K_L}{C_0}}}...................(6)\)

Where C0 (mg/L) is initial concentration. Adsorption is considered favorable if 0 < RL< 1, unfavorable if RL> 1, linear if RL = 1 and irreversible if RL< 0.

**The Freundlich isotherm** is represented by the following equation as follows[28]:

\(\ln \,\,{q_e}=\,\frac{1}{n}\ln \,{C_e}+\,\ln \,{K_F}......................\,(7)\)

Where KF (L/mg) is the Freundlich constant and 1/n is the heterogeneity factor indicating the adsorption intensity of the adsorbent.

The thermodynamic parameters for the adsorption process were calculated from the following equation[26]:

\(\Delta G= - RT\,\ln \,Kc\,\,..............\,(8)\)

Where R is the gas constant of 8.314 J/mol k, Kc is the equilibrium constant and T (K) is the temperature. The Kc value is calculated from the following equation:

\(Kc=\frac{{{C_A}}}{{{C_S}}}\,\,..................(9)\)

CA and CS (mg/L) are the equilibrium concentrations of Cr (VI) on the adsorbent and in the solution, respectively. Standard enthalpy (∆H) and entropy (∆S) of adsorption can be estimated from van’t Hoff equation given as:

\(\ln \,Kc\,=\frac{{ - \Delta H}}{{RT}}+\frac{{\Delta S}}{R}...................\,(10)\)

The slope and intercept of the van't Hoff plot which are equal to −∆H/R and ∆S/R, respectively, enable estimation of the enthalpy and entropy changes of the adsorption process.

## 2.7. Desorption and regeneration experiments

The desorption study was done by transferring 30 mg WHLPC/GO/CS/PVA mixed with 100 mg/L Cr (VI) to flasks containing 10 mL of 0.1 M NaOH. Afterward, the solution was shaken for 330 min, and the equilibrium concentration after desorption was measured. The WHLPC/GO/CS/PVA was thoroughly dried in an oven at 363K for 4 h and could be reused as an adsorbent for five consecutive cycles [26].

## 2.8. Error analysis

In this work, three error functions, the non-linear chi-square test (\({\chi ^2}\)), the coefficient of determination (R2), and the standard error of estimate (SEE) were used for analyzing the adsorption system. The advantage of using the Chi-square test was comparing all isotherms on the same abscissa and ordinate. The equivalent mathematical statement was [29]:

\({\chi ^2}=\sum {\frac{{{{({q_e} - {q_{e,m}})}^2}}}{{{q_{e,m}}}}} .......................(11)\)

Where qe,m equilibrium capacity was obtained by calculation from the model (mg/g) and qe was the equilibrium capacity (mg/g) from the experimental data. The coefficient of determination (R2) was calculated as follows [30]:

\({R^2}=\frac{{\sum {{{({q_m} - {q_e}^{ - })}^2}} }}{{\sum {{{({q_m} - {q_e}^{ - })}^2}+\sum {{{({q_m} - {q_e})}^2}} } }}......................(12)\)

Where 𝑞m is the equilibrium capacity obtained from the isotherm model, qe is the equilibrium capacity obtained from the experiment, and qe− is the mean of qe. The standard error of estimate (SEE) was calculated as follows [31].

\(SEE=\sqrt {\frac{{\sum {{{({q_e} - {q_m})}^2}} }}{{df}}} ....................(13)\)

Where qe is the experimental equilibrium adsorption capacity, qm is the equilibrium capacity obtained from the isotherm model ad df is the degree of freedom, which is the difference between several data points and the number of parameters in the isotherm model function. These error analysis parameters were calculated using Origin 16 software.