Removal of Cr(VI) from aqueous solution using ball mill modified biochar: multivariate modeling, optimization and experimental study

Chromium (Cr(VI)) pollution has attracted wide attention due to its high toxicity and carcinogenicity. Modified biochar has been widely used in the removal of Cr(VI) in water as an efficient and green adsorbent. However, the existing biochar prepared by chemical modification is usually complicated in process, high in cost, and has secondary pollution, which limits its application. It is urgent to explore modified biochar with simple process, low cost and environmental friendliness. Therefore, ball milling wheat straw biochar (BM-WB) was prepared by ball milling technology in this paper. The adsorption characteristics and mechanism of Cr(VI) removal by BM-WB were analyzed by functional group characterization, adsorption model and response surface method. The results showed that ball milling effectively reduced the particle size of biochar, increased the specific surface area, and more importantly, enhanced the content of oxygen-containing functional groups on the surface of biochar. After ball milling, the adsorption capacity of Cr(VI) increased by 3.5–9.1 times, and the adsorption capacity reached 52.21 mg/g. The adsorption behavior of Cr(VI) follows the pseudo-second-order kinetics and Langmuir isotherm adsorption model rate. Moreover, the Cr(VI) adsorption process of BM-WB is endothermic and spontaneous. Under the optimized conditions of pH 2, temperature 45 °C, and adsorbent dosage 0.1 g, the removal rate of Cr(VI) in the solution can reach 100%. The mechanism of Cr(VI) adsorption by BM-WB is mainly based on electrostatic attraction, redox and complexation. Therefore, ball milled biochar is a cheap, simple and efficient Cr(VI) removal material, which has a good application prospect in the field of remediation of Cr(VI) pollution in water.


Material preparation
Preparation of WB: The wheat straw used in this study was collected from Nanyang, Henan Province, and ground by a grinder.After passing through a 40-mesh sieve, it was placed in a vacuum tube furnace (OTF-1200X, Shenzhen Kejing Zhida Technology Co., Ltd., China); before the start of pyrolysis, high-purity nitrogen was introduced at a flow rate of 100 mL/min for 30 min to exhaust the residual air in the pyrolysis system.After the start of pyrolysis, the flow rate of nitrogen remains unchanged, the heating rate is 10 °C/min, and the furnace body is heated from room temperature to the target pyrolysis temperature of 300 °C for 2 h.After the pyrolysis, it was cooled to room temperature in a muffle furnace, and then the residue was recovered and the biochar was washed three times with ultrapure water and dried at 60 °C for 12 h.Sealed in a brown glass bottle for subsequent analysis.N 2 (purity of 99.999%) was used in the pyrolysis process.The prepared sample was labeled as WB.
Preparation of BM-WB: 10 g WB and 100 g agate balls (8, 10, 15 mm in diameter) were mixed into agate bottles, and then the agate bottles were placed in a planetary ball mill (MITR-YXQM-2L, Changsha Miqi Instruments and Equipment Co., Ltd., China).Rotating at 1500 rpm for 60 min, passing 100-mesh sieve, sealed and stored in brown glass bottles for subsequent analysis.The prepared sample was labeled as BM-WB.

Characterization of biochar
In this study, the specific surface area and pore properties of BM-WB materials were determined by nitrogen adsorption-desorption method (ASAP 2460, Micromeritics, USA), and the surface morphology and structure of the materials before and after adsorption were observed by using scanning electron microscope (SEM) and energy dispersive X-ray spectrum (EDS) (ZEISS Gemini 300, OXFORD Xplore, Germany).By Fourier transform infrared spectrometer (FTIR) (Nicolet iS50, Semirfei, USA), 1 mg sample was mixed with 100 mg KBr, pressed into tablets, and scanned in the wavelength range of 450-4000 cm −1 , which was used to analyze the changes of surface functional groups before and after adsorption.In order to further study the surface morphology and elemental composition of the material, X-ray diffractometer (XRD) (Ultma IV, Rigaku, Japan) was used to analyze the form of the material and determine the crystal structure.The scanning rate was 5 °C/min, and the 2θ range was 10-80°.X-ray photoelectron spectroscopy (XPS) (K-Alpha, Thermo Scientific, USA) was used to study the presence of Cr(VI) on the surface of the material before and after adsorption and the oxidation state of chromium.According to China's environmental protection standard (GB 7467-87), the concentration of Cr(VI) was determined at 540 nm wavelength based on ultraviolet-visible spectrophotometer (DR1900, Beijing General Instrument, China), and the detection limit was 0.004 mg/L.

Batch adsorption experiments
All batch tests were carried out in a 100 mL conical flask with a volume of 50 mL, which was sealed by a silicone plug.In addition to study the effect of temperature on this experiment, all other experiments were carried out at room temperature 25 °C using a constant temperature water bath oscillator (SHZ-82B, Jinnan Instruments, China) at 150 rpm.Different parameters were optimized, such as adsorbent dose (1-10 g/L), initial pH (2-10), oscillation time (0-5 h), Cr(VI) concentration (10-110 mg/L) and temperature (25-45 °C).The pH of the solution was adjusted by adding 1 mol/L HCl or NaOH using a pH meter (PHS-3C, Lei, China).At the end of each experiment, the mixture was filtered through a 0.45 μm filter membrane and then the Cr(VI) concentration was quantified.All experiments were conducted in triplicate.The data were expressed as mean ± standard deviation.The test results were recorded and analyzed.The removal rate R(%) and adsorption capacity q e (mg/g) of Cr(VI) by WB were calculated according to Eqs. (1)- (2).In addition, detailed information on adsorption kinetics, adsorption isotherms, and adsorption thermodynamic models were provided in the chapters of support material information Sect. 1, Sect. 2 and Sect.3, respectively.where C 0 is the initial concentration of Cr(VI), mg/L; C e is the concentration of Cr(VI) in the solution at adsorp- tion equilibrium, mg/L; V is the volume of Cr(VI) solution, L; m is the dosage of biochar, g.

Experimental model optimization design
As the most commonly used method in RSM, Box-Behnken design (BBD) is the most widely used surface response method because it can accurately describe the linear interaction and quadratic effect of the secondorder polynomial model with the least number of experiments, and avoid extreme conditions.In addition, the response surface optimization method also considers the random error of the test.Through the regression fitting process, the response surface and contour are drawn, and the response values corresponding to each factor level can be found.Therefore, we applied it to study and verify the influence of the parameters we selected on the Cr(VI) removal efficiency of WB.
There were five laboratory factors evaluated in this study.Each factor was selected as an independent variable at three levels: initial solution pH (× 1), initial Cr(VI) concentration (× 2), biochar dosage (× 3), adsorption time (× 4), and temperature (× 5).Each independent variable was set at three levels (low, medium, and high, coded as − 1, 0, and + 1, respectively).The level and range of the selected independent variables were given in Table 1.This design illustrated the interaction between independent and dependent variables.The quadratic polynomial equation was used to perform multiple regression on the data from BBD to obtain the parameter estimation of the model (Eq.(3)): (1) where x i and x j are coded values; β 0 is constant; β i is linear coefficient; β ii is quadratic term coefficient; β ij is interaction term coefficient; ε is random error; k is 5 (independent variable value).

Reuse test
In order to evaluate the reusability of BM-WB as an adsorbent, the adsorbed BM-WB was added to 50 mL of 0.1 mol/L NaOH solution, and then regenerated by continuous magnetic stirring at 25 °C for 60 min.The regenerated BM-WB was washed three times with ultrapure water to effectively remove the desorbed Cr(VI).The regenerated BM-WB was repeated for 5 adsorption cycles to determine its effectiveness and stability.

Characterization
By comparing the physicochemical properties of WB and BM-WB (SI Table S1), it was found that the ball milling technique could effectively increase the specific surface area and pore volume by 2.87 and 127.77times, respectively, of WB.In addition, the particle size of biochar could be reduced by ball milling (SI Fig. S1).Based on the analysis of Cr(VI) adsorption by WB and BM-WB (SI Fig. S1), the adsorption amount of Cr(VI) by BM-WB was 3.5-9.1 times higher than that by WB, and BM-WB could provide more binding sites, which could facilitate the exposure of functional groups on the surface of biochar and thus contribute to the effective removal of Cr(VI) from aqueous media 31,32 .
The morphology and porous structure of BM-WB were characterized by SEM.The structure of BM-WB before adsorption was irregular, with an uneven and rough surface containing macropores and micropores of different shapes, which could provide effective adsorption sites and spaces for the adsorption process (Fig. 1a).The overall surface characteristics of BM-WB tended to be flat after adsorption, and its individual spots showed agglomerated morphology, indicating that the pores were covered by Cr(VI) (Fig. 1b).The elemental composition (Fig. 2 and SI Fig. S2) and content percentages (Fig. 1c and d) of BM-WB were analyzed using EDX, and only after adsorption a Cr elemental content of 6.01% was detected, favorably confirming that Cr(VI) was effectively adsorbed by BM-WB.In addition, significant changes in the elemental contents of Mg, K and Ca occurred before and after adsorption, which could be attributed to the ion-exchange reactions of these metal elements with Cr ions 33 .The changes of C, N and O elements may be attributed to oxidation reactions, complexation reactions and co-precipitation 34 .
The main functional groups of the different materials were examined by FTIR (Fig. 3a), and for the WB materials, diffraction peaks were observed at 3373 (-OH), 2929 (-CH 3 , -CH 2 ), 1609 (C=C), 1437 (-CH 3 ), 1376 (C-H), 1091 (C-O-C), 797 (Si-O-Si) and 466 cm −135-37 .The diffraction peaks of BM-WB were stretched and the intensity increased significantly, indicating that ball milling can change the content of functional groups 38 .The www.nature.com/scientificreports/diffraction peaks of BM-WB were significantly altered before and after adsorption, and the wave number spectral bands and depths became smaller after adsorption, which was attributed to the stretching vibration of the bonds inside the functional groups and the chattering generated during the adsorption of Cr(VI) 39 .Among them, the diffraction peaks at 3373 cm −1 , 2929 cm −1 , 1091 cm −1 and 466 cm −1 were significantly weaker, which indicated that functional groups such as -OH, -CH 3 , C-O-C and Si-O-Si were involved in the Cr(VI) removal reaction.
The chemical oxidation state and elemental composition of BM-WB were analyzed by XPS method.The changes of the main characteristic peaks on the surface of BM-WB before and after adsorption were O 1 s (+ 2.63%), C 1 s (− 2.41%), F 1s (− 1.1%), N 1s (+ 0.15%), Si 2p (− 0.38%), Al 2p (+ 0.18%), and Cr 2p (+ 0.62%)  (Fig. 3c).Before adsorption, no characteristic peak of Cr was detected on the surface of BM-WB, while the presence of the characteristic peak of Cr was clearly detected after adsorption (Fig. 3f), indicating that Cr was successfully adsorbed.The XPS O1s signal of BM-WB was decomposed into three peaks (Fig. 3d), with the binding energy concentrated at 531.0 eV (C-O), 532.1 eV (-OH) and 533.0 eV (C=O).After adsorption, the -OH and O-C=O ratios in BM-WB decreased by 4.77% and 2.69%, respectively, while the C-O increased by 7.23%.In addition, the C1s signal was also decomposed into three peaks (Fig. 3e), with binding energies concentrated at 284.8 eV (C-C), 286.3 eV (C-O-C) and 288.2 eV (C=O).The C-O and -OH characteristic peaks changed significantly after adsorption.It indicated that these functional groups were involved in the adsorption process and their contribution might be related to the electrostatic attraction and complexation reactions occurring in Cr(VI) 41,42 .Further analysis of the different valence states of Cr (Fig. 3f) showed that the Cr 2p peaks at 576.67 and 586.26 eV, 578.48 and 588.05 eV corresponded to Cr(III) and Cr(VI), respectively.The surface molar contents of Cr(III) and Cr(VI) were 50.55% and 49.45%, respectively, confirming that the BM-WB surface adsorption of Cr(VI) might be reduced to Cr(III) by surface functional groups (-COOH, -OH and -NH 2 , etc.) 43 .The XPS analysis results were consistent with the FTIR spectroscopy results (Figs. 3a and 4b).

Study of batch tests
The highest removal rate was achieved at pH 2.0 and decreased sharply at pH > 2 (Fig. 4a), a result that proved that acidic conditions favored the removal of Cr(VI).Cr(VI) in aqueous solution was mainly present in the form of HCrO 4 − (pH = 1.0-5.0)and CrO 4 2− (pH > 8.0) (Fig. 5a).The zero point charge (pH ZPC ) is strongly influenced by pH 44 .SI Fig. S4 shows the pH of the zero charge point of BM-WB, pH ZPC = 5.62.At pH < pH ZPC , the functional groups (-COOH, -OH and -NH 2 , etc.) on the surface of BM-WB are positively charged by protonation, which facilitates the adsorption of HCrO 4 − on the surface of BM-WB by electrostatic attraction 45 .When pH > pH ZPC , the functional groups on the surface of BM-WB underwent deprotonation and became negatively charged, and HCrO 4 − might be converted to CrO 4 2− and Cr 2 O 7 2− when using higher pH solutions, due to the negative surface charge and competition with the OH − competition, making BM-WB less attractive to Cr(VI) 46,47 .
The equilibrium adsorption amount of Cr(VI) by BM-WB increased from 48.83 mg/g to 54.06 mg/g when the test temperature was increased from 25 to 45 °C (Fig. 4b).This was because the increase in temperature accelerates the movement of ions in solution, which was favorable for the diffusion of Cr(VI) into the pore channel overcoming the resistance of the BM-WB surface 48,49 .Also, the activation energy to be overcome by the adsorption process was lower at higher temperatures, making the adsorption reaction easier to proceed 50 .Considering the energy-saving perspective, an operating temperature of 25 °C was used for the next experiments.www.nature.com/scientificreports/ The equilibrium adsorption of Cr(VI) by BM-WB increased from 5.00 to 48.83 mg/g when the test concentration was increased from 10 mg/L to 110 mg/L (Fig. 4c).The reactive sites on the surface of BM-WB were fully bound to Cr(VI) at high concentrations, while some sites failed to participate in Cr(VI) reduction and adsorption at low concentrations 51 .The adsorption amount of Cr(VI) showed a rapid increase with adsorption time and then stabilized (Fig. 4d), probably because there were more adsorption sites on the BM-WB at the initial stage of the adsorption reaction and the concentration difference between the surface and the solution Cr(VI) was larger, resulting in a larger mass transfer kinetic, and also there was a strong electrostatic attraction between the BM-WB surface and the solution Cr(VI) at this time 52 .The electrostatic attraction between the BM-WB surface and the solution Cr(VI) was also strong at this time 53 .However, as the adsorption proceeds, the adsorption sites were occupied by Cr(VI) and the concentration difference between the surface and solution Cr(VI) decreases gradually, and the electrostatic interaction between the BM-WB surface and Cr(VI) changes from gravitational force to repulsive force, which eventually caused the adsorption rate to stabilize 54 .
The removal of Cr(VI) was positively correlated with the amount of BM-WB dosing, and the removal of Cr(VI) increased from 76.86% to 100% (Fig. 4e).It was shown that the specific surface area of BM-WB exposed to Cr(VI) solution increased with the increase of the dosage, and the number of reaction sites involved in the reduction and adsorption of Cr(VI) increased 55 .To further determine the optimal dosage of the material, the adsorption of Cr(VI) was investigated at different BM-WB dosages (Fig. 4d).The Cr(VI) adsorption amounts were 38.43, 48.97, 25.00, 16.67, 16.67, 12.50, 12.50, 10.00 and 10.00 mg/g at the dosage of 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45 and 0.5 g.Among them, the adsorption efficiency was highest when the dosage was 0.1 g, so this optimal dose was used in the next experiments to ensure higher Cr(VI) removal.www.nature.com/scientificreports/

Adsorption kinetics
The kinetic properties of Cr(VI) adsorption by BM-WB were investigated by fitting first-order, second-order, Elovich and intraparticle models according to the kinetic equations SI Eqs. S (1)-( 4), and the model fitting results were shown in Fig. 5b, c, d and e, and the kinetic parameters were shown in SI Tables S2, S3 and S4.The fitted second order (R 2 = 0.999) and Elovich model (R 2 = 0.993) results were better.Among them, the fitted second-order results indicated that the overall rate between BM-WB and Cr(VI) was not only affected by physical diffusion, but the more dominant rate-limiting step was the chemisorption process 56 .Since the large shear force generated during ball milling could improve the surface properties of biochar and increase the number of functional groups, ball milling could enhance the removal of Cr(VI) 57,58 , which was consistent with the phenomenon reported by Zhang et al. 24 .The Elovich model further suggested that the removal of Cr(VI) by BM-WB might be on a homogeneous surface of chemisorption, where the initial concentration had a large effect on the adsorption rate α 59 .The adsorption rate α increased proportionally with increasing initial concentration, which led to an increase in chemisorption rate 60 .
To further elucidate the diffusion mechanism, an intra-particle diffusion model was used (Fig. 5e).q t and t 0.5 kinetic curves of BM-WB were both straight lines not passing through the origin, indicating that the adsorption process was controlled by multiple steps 61 .The adsorption process could be divided into three stages: liquid film diffusion, intraparticle diffusion and adsorption equilibrium, and k 1 > k 2 > k 3 (SI Table S5), so liquid film diffusion and intraparticle diffusion were the main diffusion rate determining steps 62 .

Adsorption isotherm
The adsorption studies of Cr(VI) solutions (10, 30, 50, 70, 90 and 110 mg/L) with different concentrations by BM-WB were carried out using the Langmuir and Freundlich isotherm models (Fig. 5f and g, and the relevant parameters are shown in SI Table S6).The results showed that the data of Cr(VI) adsorption by BM-WB fitted better to the Freundlich model (R 2 = 1), indicating that the adsorption of Cr(VI) by BM-WB was not monolayer adsorption, but the formation of multilayer adsorbent on the non-uniform surface of BM-WB 63 .This might be attributed to the irregular structure of the BM-WB surface and the properties of O-rich functional groups, which allowed Cr(VI) to form multilayer adsorption on the BM-WB surface and led to a non-uniform distribution of thermal energy 64 .

Thermodynamic studies
With the change of solution temperature, the removal rate of Cr(VI) changed obviously.Using the Van't Hoff law plot (Fig. 5h), H • and S • could be determined from the slope and intercept, respectively, and the value of G • could be further calculated.The calculation of various thermodynamic parameters was shown in Table 2.
When the temperature increased from 25 to 45 °C, G • was negative, and G • decreased from − 5.67 kJ/mol to − 16.47 kJ/mol.This indicated that the removal of Cr(VI) by BM-WB was spontaneous and the adsorption process was thermodynamically more favorable at higher temperatures 65,66 .The value of S • was 538.12 kJ/mol/K, indicating that BM-WB had a high affinity for Cr(VI) 40 .The value of H • was 155.02 kJ/mol, which confirmed that the adsorption process was an endothermic process.With the increase of temperature, q e (mg/g) increased, which was mainly driven by physical and chemical processes such as electrostatic interaction and ion exchange, rather than a single physical or chemical process 67,68 .
The maximum adsorption capacity of BM-WB for Cr(VI) was q m 52.21 mg/g.In comparison with other reported materials for Cr(VI) adsorption, such as Zhou et al. (10.6 mg/g) 69 , Nethaji et al. (57 mg/g) 70 and so on, the advantages and efficiency of this work were confirmed (SI Table S7).In addition, the adsorption-desorption cycle results of BM-WB are shown in Fig. 5i.The regenerated adsorbent can be reused for at least 5 adsorption-desorption cycles without changing its effectiveness.However, the observed decrease in adsorption efficiency after the third cycle can be attributed to the fact that the surface of BM-WB material was occupied by Cr(VI) during the regeneration process.

Statistical optimization of BBD and RSM in BM-WB adsorption process
According to the above batch test results, the adsorption conditions were optimized: BM-WB dosage (0.1 g), initial concentration of Cr(VI) (50 mg/L), reaction time (3 h), pH (2.5) and temperature (35 °C) were used as the central points, and BBD was selected for 45 times of optimization experiments.SI Table S8 provided the actual and predicted results of Cr(VI) removal rate.A quadratic polynomial model (Eq.( 4)) was established to predict the removal rate of Cr(VI) (Y%) : www.nature.com/scientificreports/ The adequacy of the model was determined by different experiments.The naming order model, the sum of squares, and the model summary statistics represented the removal effect of BM-WB on Cr(VI).The results in Tables 3 and 4 showed that the quadratic model fits the experimental data best, with the lowest standard deviation, the highest adjusted R 2 and predicted R 2 values, and the lowest P value.Therefore, the quadratic model was used as the best fit for further analysis.Because there were not enough points to estimate the coefficient pairs of the model, the cubic model was not selected.
In order to evaluate the reliability of the quadratic model, the F value in the S9 analysis of variance (ANOVA) in the SI table was used for analysis.The F value of the model was 39, and the p value was < 0.001, indicating that the model had a significant effect on the response 71 .Table 5 showed that the R 2 value of the model was 0.9701, indicating that 97.01% of the model was consistent with the experimental data, and only 2.99% of the data did not match the model.The adjusted R 2 value was 0.9453, and the R 2 and the adjusted R 2 values were close to 1.This confirmed that the model was highly significant under the support of extremely high R 2 and adjusted R 2 values 40 .Further analysis of the determination coefficient of the quadratic model in Table 4 showed that the signal-tonoise ratio was measured by 'sufficient accuracy' .If the value exceeded 4, the model was an ideal model.In this study, the Adeq was 22.7774, which indicated that the signal was sufficient, and this model could be used in the navigation design space 72 .Among them, the CV value was 3.92%, which was far from 10%, further indicating that the precision and reliability of the detection were high.The removal rate obtained by the experiment was compared with the predicted value 68,73 .In Fig. 6g, it could be detected that the value had a high correlation along the uniform distribution of the straight line 29 .It was proved that the model had good applicability to data fitting.
Figure 6 showed the results of various factors and their interactions.When only a single factor was considered (Fig. 6h), the removal rate was positively correlated with BM-WB dosage (B), reaction time (C) and temperature (D), and negatively correlated with initial concentration (A) and pH value, and showed a maximum at the center point.According to the F value in ANOVA, the influence degree of each parameter on adsorption was x 5 > x 2 > x 3 > x 1 > x 4 , that is, pH had the greatest influence on the removal of Cr(VI).Figure 6a-f were the results of the interaction of various factors.The initial concentration of Cr(VI) (30 mg/L), BM-WB dosage (0.15 g), reaction time (5 h) and temperature (45 °C) were fixed respectively.When the pH was reduced from 3 to 2, the   , respectively.At a lower pH value, the removal effect of Cr(VI) was better.This was consistent with the above pH batch test results.When the pH value was fixed at 2, the initial concentration of Cr(VI), the dosage of BM-WB, the reaction time and the temperature were changed respectively.It could be found that only by changing the dosage of BM-WB (0.05-0.15 g), the removal efficiency of Cr(VI) was greatly affected (from 89.34% increased to 100%).Maybe with the increase of BM-WB, there were more reaction sites involved in Cr(VI) adsorption, which made it fully combined with Cr(VI) to increase the removal efficiency of Cr(VI).

Adsorption mechanism
The reaction mechanism of Cr(VI) adsorption by BM-WB was analyzed based on the above experimental studies, FTIR and XPS results, and the fitting results of kinetic, isothermal and thermodynamic models (Fig. 7).The adsorption process mainly included the following mechanisms: (1) Analysis by SEM image showed that the original BM-WB surface structure was irregular and the surface was uneven, containing macropores and micropores of various shapes, providing effective sites and spaces for the adsorption of Cr(VI), and the pores on the surface of BM-WB were covered by Cr(VI) after adsorption filling.Physisorption was the most common and the first reaction to occur, but it was not the main adsorption 2+ in aqueous solution (Fig. 5a).With the increase of solution pH, the surface charge characteristics of biochar usually change from positive to negative 76 .Many scholars have studied the effect of pH on the adsorption of biochar during the adsorption process.Sun et al. 77 also determined the contribution of proton exchange by measuring the pH change during the adsorption process, indicating that the decrease of pH value during the adsorption process was accompanied by the release of biochar protons.Xu et al. 78 also confirmed that it is more obvious at strong acidity (pH = 2) than at weak acidity (pH = 4).When pH > pH zpc , the adsorption sites of functional groups (-OH, -COOH and -NH 3 , etc.) on the surface of BM-WB were deprotonated to make the overall surface of biochar negatively charged and electrostatically repelled by HCrO4-ions in the solution, resulting in a decrease in Cr(VI) adsorption and a weak effect on Cr(VI) removal efficiency.When pH < pH zpc , the surface functional groups of BM-WB are protonated, and HCrO 4 − ions are attached to the surface of biochar through electrostatic attraction, which is beneficial to the removal of Cr(VI) 24,40 .The response surface experiment also showed that the removal of Cr(VI) by BM-WB was closely related to the pH value; (3) The exchange between cations and heavy metals on the surface of biochar is also one of the important mechanisms of adsorption 79 .XRD analysis showed that the content of original ions (Na + , K + , Ca 2+ , etc.) on the surface of BM-WB changed significantly before and after adsorption of Cr(VI), which was attributed to the ion exchange of different forms of Cr on BM-WB during adsorption, so that Cr was attached to functional groups rather than original ions 80,81 .The study of Abbas et al. 82 also showed that C-C, C-O and COO-H of biochar were involved in the exchange of heavy metals.Sun et al. 77 also confirmed that ion exchange, especially Na+ biochar on the surface of biochar, promoted the removal of heavy metal pollutants by XPS analysis; (4) In addition, the surface of BM-WB is rich in functional groups (-OH, -COOH and -NH 3 , etc.), their H, N, and O, etc., could be used as coordination atoms for coordination complexation with heavy metal ions 83,84 .Qu et al. 85 also analyzed the changes of C-C, C-O and COO-H after adsorption of Cr(VI) by XPS, indicating that these groups play a leading role in the complex adsorption weight 75 ; (5) Since the surface of BM-WB also contained functional groups such as -C-O and -C=O in the form of phenols, ketones and amino groups, these functional groups could be used as electron donors in the process of oxidation to quinone groups and ammonium salts to reduce Cr(VI) to Cr(III) 75 , see Eqs. (5-6).Xu et al. 78 proved by electron shuttle test that -C-O and -C = O were electron donors for Cr(VI) reduction in biochar when removing Cr(VI) in solution.About 83.5% of Cr adsorbed on biochar was reduced to Cr(III), and the remaining 16.5% was Cr(VI).At the same time, the FTIR spectrum of this study was significantly weaker at the diffraction peak of -OH at 3373 cm −1 , which further proved that -OH may also be the electron donor part of Cr(VI) reduction; (6) In addition, the surface functional groups can further precipitate with Cr(III) on BM-WB, and can also indirectly affect and promote the coprecipitation behavior through other adsorption mechanisms 86 , see Eqs. (7-8).

Figure 3 .
Figure 3. Characterization analysis: (a) FTIR; (b) XRD; (c) XPS full spectrum before and after adsorption; (d) Spectrum of O 1 s before and after adsorption; (e) Spectrum of C1s before and after adsorption; (f) Spectrum of Cr(VI) 2p before and after adsorption.

Figure 6 .
Figure 6.Dual effects of input variables on Cr(VI) removal efficiency (3D plot): (a) pH and initial concentration effect; (b) pH and dosing amount effect; (c) pH and adsorption time effect; (d) pH and temperature effect; (e) initial concentration and dosing amount effect; (f) dosing amount and adsorption time effect; (g) predicted and actual values of model on Cr(VI) removal efficiency; (h) independent effects of five variables A, B, C, D and E on Cr(VI) removal.

Table 1 .
Factors and levels of the Box-Behnken design.

Table 3 .
Sequential model sum of squares.

Table 4 .
Model summary statistics.