1 Introduction

As a consequence of wide application in industries such as metallurgy and leather tannery, Cr(VI) is frequently found in wastewater and groundwater (Kaur and Crimi 2014; Wang et al. 2018). The toxicity of chromium is highly dependent on its oxidation state. Cr(VI) is highly toxic and extremely mobile in environment, while Cr(III) is less toxic and easy to be removed from environment. Therefore, the reduction of toxic Cr(VI) to less toxic Cr(III) offers a feasible way for the treatment of Cr(VI) pollution (Yang et al. 2023; Zhuang et al. 2021). Currently, a variety of methods have been developed for the removal of chromium contamination from water media, including coagulation, biological treatment, membrane separation, ion exchange, adsorption, and electrochemical methods(Fang et al. 2021; Maamoun et al. 2023; Matsukevich et al. 2021). Among them, owing to the advantages including low cost and strong reducing capability, zero-valent iron (ZVI) has been extensively explored to remediate Cr(VI) contaminated surface waters and groundwater(Wang et al. 2021a). However, the wide application of ZVI for Cr(VI) removal faces a formidable challenge, i.e., easy formation of a passivation layer that inhibits electron and mass transfer (Prasad et al. 2011; Wu et al. 2020).

So far, a variety of enhancing strategies have been proposed to alleviate the passivation of mZVI such as sulfidation (Li et al. 2019; Lü et al. 2019), construction of composite with carbon materials (Li et al. 2016; Wang et al. 2022; Wang et al. 2020; Zhu et al. 2018), pre-magnetization (Sun et al. 2018; Zhang et al. 2020), bimetallic materials (Fu et al. 2015) and so on. Carbon materials take the advantages of low cost, excellent chemical stability, large surface area and abundant pores, leading the construction of composite with carbon materials being the most popular strategy. In particular, biochar is a fascinating carbon material with the features of low cost and eco-friendly, which make it a better choice compared with other carbon materials like carbon nanotubes and graphene (Zou et al. 2021). The presence of carbo nanomaterials can accelerate the corrosion of ZVI to retard the passivation (Li et al. 2016). Sulfidation can increase both the reactivity and selectivity of ZVI (Wang et al. 2018), due to the formation of FeSx species on the surface of ZVI particles. Nevertheless, the Cr(VI) removal capacity of ZVI-carbon composites and S-ZVI is still very low at circumneutral pH, which precludes their wide application in the remediation of Cr(VI) contaminated groundwater.

In this work, we propose a “capture-reduction” strategy for enhancing the Cr(VI) removal capacity of microscale ZVI through dual modification, i.e., simultaneous sulfidation coupled with sulfur-doped graphene-like biochar (S-mZVI/SGB). The Cr(VI) anion is captured by SGB and then reduced to Cr(III) by accepting electrons from ZVI. Density-function-theory calculation was employed to confirm the enhanced Cr(VI) adsorption by SGB. Tafel tests and Fe(II) scavenging tests were carried out to reveal the detailed Cr(VI) removal mechanism. Finally, the effect of natural organic matter and co-existing anions on Cr(VI) removal and the reusability of the S-mZVI/SGB were studied to evaluate its practical application prospect.

2 Experimental section

2.1 Materials and chemicals

All chemicals are of analytical grade and used as received without further purification. Sublimed sulfur (S), hydrochloric acid, sulfuric acid, phosphoric acid, acetone, potassium dichromate, sodium hydroxide, and diphenylcarbazide were purchased from Aladdin Biochemical Technology Co. Ltd. Micron-sized zero-valent iron (mZVI, 1–3 μm) was purchased from China Metallurgical Institute. Ultrapure water (18.25 MΩ.cm) was obtained through an Ulupure water purification system, and lake water was collected from Xinkai Lake in Nankai University.

2.2 Mechanochemical synthesis of S-mZVI/SGB

GB was synthesized using seaweeds as precursor similar to a method reported in a previous work (Liu et al. 2020). The dried and cleaned seaweed was pyrolysis at 800 °C in a tubing furnace for 2 h under a flow of nitrogen. The sample was named as graphene-like biochar and stored in a desiccator for further use.

The mechanochemical synthesis of S-mZVI/SGB was carried out on a PM4L planetary ball mill system (Zhuodi Instruments Co., Ltd., Shanghai, China). Briefly, mZVI (7.125 g), GB (1.5 g), and sublimed sulfur (0.375 g) were mixed together in a 1 L stainless steel jar with 1 kg agate balls (mass ratio Φ6mm: Φ10mm: Φ20mm = 2: 4: 4). After sealing, the stainless-steel jar was evacuated and back-filled with argon three times. The ball milling process was carried out at 275 rpm for 24 h.

For comparison, the sulfur doped GB (SGB), sulfidated mZVI (S-mZVI) and mZVI/GB were synthesized by omitting one component using a similar process. To investigate the effect of ball milling, mZVI was ball milled to get BM-mZVI.

2.3 Cr(VI) removal

Batch experiments were adopted to evaluate the performance of S-mZVI/SGB composites in removing Cr(VI) from aqueous solutions, and all experiments were performed in triplicate in deoxygenated 100 mL glass saline bottles. The pH was adjusted with 0.1 M HCl and NaOH using a pH meter. The initial concentration of Cr(VI) was 30 mg·L− 1, and the initial pH was 5.7 without adjustment. Unless otherwise specified, the above parameters were not adjusted in the experiment.

The material dose and the initial Cr(VI) concentration varied from 0.5–1.5 g·L− 1 and 0–200 mg·L− 1, respectively, to investigate their effect on Cr(VI) removal. The effect of pH on Cr(VI) removal was investigated by adjusting the initial pH to 4, 6, 7, 9. Finally, lake water was used to simulate the removal of Cr(VI) from natural aqueous solutions.

During reaction, the solution was sampled at pre-determined time intervals. After filtered with a 0.22 μm membrane, the concentration of Cr(VI) was measured at 540 nm using an UV-Visible spectrophotometer (Gao et al. 2022).

The removal efficiency (η), adsorption capacity at time t (Qt) and equilibrium (Qe) of the composites for Cr(VI) were calculated by eqs. (1), (2) and (3), respectively.

$$\upeta =\frac{\left({C}_0-{C}_t\right)}{C_0}\times 100\%$$
(1)
$${Q}_t=\frac{\left({C}_0-{C}_t\right)}{m}\times V$$
(2)
$${Q}_e=\frac{\left({C}_0-{C}_e\right)}{m}\times V$$
(3)

where, C0, Ct and Ce represent Cr(VI) concentration at time 0, t and equilibrium (mg·L− 1), respectively. m is the mass of the material (g), and V is the solution volume (L).

2.4 Characterization

Scanning electron microscope (SEM, TESCAN MIRA LMS, Czech Republic) equipped with energy-dispersive X-ray spectroscopy (EDS) was used to observe the morphology of S-mZVI/SGB before and after Cr(VI) removal. Vibrating sample magnetometer (VSM, LakeShore7404) was used to test the magnetic properties of the material. The phase and crystallography information of the material was analyzed by using a Rigaku Smartlab X-ray diffractometer. The functional groups on the surface of the composites were characterized using a Bruker Tensor 27 Fourier Transform Infrared Spectrometer (FTIR). Raman spectra of the composites were collected on Edinburgh RM5a Raman spectrometer. Surface wettability of the composites was characterized by measuring their contact angles. X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA) was used to examine the surface chemical composition and valence states of the material. Tafel curves and electrochemical impedance spectroscopy (EIS) tests were performed using an Correst CS150M electrochemical workstation (see Text S1 for details).

3 Results and discussion

3.1 Cr(VI) removal performance

As shown in Fig. 1a, commercial mZVI was ineffective for Cr(VI) removal. Only 2.16% of Cr(VI) could be removed by mZVI within 6 h. The poor performance was ascribed to the presence of an oxide layer on the surface of mZVI particles formed during storage and transportation. The ball milling treatment could increase the removal rate of Cr(VI) to 8.46%, which was attributed to the partial removal of the surface oxide layer. In sharp contrast, sulfidation processing could dramatically increase the Cr(VI) removal rate to 24.35%, which was consistent with previous studies (Garcia et al. 2021). The mZVI/GB composite had a remarkably high Cr(VI) removal rate of 63.76%, suggesting that the construction micro-electric field is a better strategy than sulfidation (Wang et al. 2020). But mZVI/GB composite exhibited a slow Cr(VI) removal kinetics compared to that of S-mZVI. The S-mZVI/SGB composite showed the highest Cr(VI) removal of 99.03%, and the corresponding Cr(VI) removal rate were between that of mZVI/GB and S-mZVI. The inset image in Fig. 1b shows that the yellow color of Cr(VI) solution almost disappeared. The Cr(VI) removal capacity at equilibrium of the S-mZVI/SGB composite reached 29.71 mg·g− 1 within 8 h, which was 1.55 and 4.08 times that of mZVI/GB and S-mZVI, respectively.

Fig. 1
figure 1

Cr(VI) removal performance of various materials. a Cr(VI) removal rate, and b equilibrium adsorption capacity of each material. The material dose was 1.0 g·L− 1, the initial concentration of Cr(VI) was 30 mg·L− 1, and the initial solution pH was 5.7 without adjustment

3.2 Characterization

The SEM images of mZVI, mZVI/GB, S-mZVI and S-mZVI/SGB are shown in Fig. S1. Under the action of mechanical forces, the mZVI particles were broken, deformed and welded (Cagnetta et al. 2017), and the composite material was deformed from the original spherical morphology to irregular shapes of flakes, blocks and granules. Compared with the original mZVI particles of 2.67 ± 1.42 μm, the size of the composite particles was significantly reduced (Fig. S2). The EDX spectrum of S-mZVI/SGB revealed the presence Fe, C, O and S elements (Fig. S3), and the elements were uniformly distributed in the composites (Fig. 2a). The results confirmed the successful preparation of S-mZVI/SGB composites.

Fig. 2
figure 2

SEM images of S-mZVI/SGB a before and b after Cr(VI) removal

After being used for Cr(VI) removal, there are obviously nanostructures formed on the surface of the composite particles to exhibit a petal-like morphology (Fig. 2b). The elemental mapping result indicated that areas with high concentration of Fe element had low concentration of Cr element (marked by blue circles), which was attributed to the easy formation of passivation layer on the Fe-rich area. The result clearly demonstrated that the significantly increased Cr(VI) removal capacity was attributed to the presence of sulfur species and SGB nanosheets. No apparent nanostructure growth was observed in mZVI/GB composite after Cr(VI) removal (Fig. S4). The growth of petal-like nanostructures may therefore be ascribed to the formation of sulfur-containing FexCryOz flakey nanostructures on GB nanosheets in the composite (Zhao et al. 2019). The result suggested that chromium ions were preferentially precipitated on GB nanosheets (Gao et al. 2022).

To investigate the interaction of sulfur, GB nanosheets and mZVI in the composite, FTIR was used to characterize the surface functional groups of various materials. As shown in Fig. 3a, absorption peaks positioned at 3435 cm− 1 and 1367 cm− 1 were ascribed to the stretching vibration of –OH and –C-O, respectively. The absorption band located at 1589–1610 cm− 1 was assigned to the stretching of C=C bond of the graphitic domain (Guo et al. 2009). Compared to GB, additional absorption peaks around 3000–2700, 1083 and 773 cm− 1 were observed in other samples, possibly due to the elimination of oxygen-containing functional groups (reduced by sublimated sulfur or mZVI) during preparation (Li et al. 2023). The absorption bands located around 3000–2700 cm− 1 were due to the asymmetric and symmetric CH2 stretching. The peak positioned at 1083 cm− 1 was attributed to C-O stretching vibrations of C-O-C. The peak position at 773 cm− 1 was assigned to the bending vibration of C-H. Besides the samples containing mZVI, the peak positioned at 483 cm− 1 (Fe-O) was also observed in sample SGB, which might be caused by the etching of ball milling jar by sulfur during the ball milling process. The peaks corresponding to C=C stretching in all other samples were blue shifted compared to GB, suggesting the migration of electrons from S and/or Fe atoms to neighboring C atoms.

Fig. 3
figure 3

a FTIR spectra, b XRD patterns, c water contact angles and d magnetization curves of various materials

The crystallinity and phases of GB, mZVI, mZVI/GB, S-mZVI and S-mZVI/SGB before and after Cr(VI) removal were characterized using powder X-ray diffraction (XRD). The diffraction pattern of mZVI showed two prominent diffraction peaks corresponding to the (110) and (201) planes of α-Fe0 (Fig. S5), which were present in all composites (Fig. 3b). The diffraction pattern of GB only exhibited a very broad and weak diffraction peak corresponding to graphene, demonstrating its graphene-like nature (Fig. S5). The intensity of the diffraction peak around 22.6o (attributed to graphene-like nanosheet structure) was stronger for S-mZVI/SGB than other samples (Yang et al. 2021), which indicated that the existence of aggregated or restacked graphene-like nanosheets. To characterize the morphology of SGB in the composite, the sample was treated with HCl solution and DMF in sequence to remove iron and sulfur species and then observed by using TEM. As shown in Fig. S6, SGB exhibits a sheet-like morphology with clear lattice fringes like graphene. Raman spectroscopy was employed to characterize the quality of SGB nanosheets formed in the composites. As shown in Fig. S7, the ratio of D band to G band intensity (ID/IG) was 2.01, indicating that the SGB in the composite contained considerable number of defects. The ID/IG ratio of the S-mZVI/SGB composite was 1.8, suggesting the GB nanosheets in the S-mZVI/SGB composite contained fewer defects than those in the mZVI/GB composite. The result indicated that strong interaction occurred between sulfur and GB nanosheets in the composite. After being used for Cr(VI) removal, the background noise of the XRD patterns increased significantly, suggesting the formation of FexOy. In addition, several weak peaks belonging to FeS appeared in the Cr-loaded S-mZVI and S-mZVI/SGB composite.

The reactivity of zerovalent iron was closely related with its hydrophilicity (Xu et al. 2020). So, water contact angle measurements were done to characterize the hydrophilicity of the materials. As shown in Fig. 3c, ball milling processing led to the decrease of the water contact angle, indicating that mZVI became more hydrophilic. The enhanced hydrophilicity might be due to the increase of specific surface area and the exposure of Fe0 caused by the destruction of the surface structure of mZVI (Wang et al. 2022), which is more likely to bind with hydroxyl group, resulting in enhanced hydrophilicity complexity of the surface structure of mZVI. The mZVI/GB composite had the lowest water contact angle among all samples, which was attributed to the presence of hydroxyl and carboxyl groups on the surface of GB nanosheets (Fig. 3a). Consistent with previous studies, sulfidation rendered mZVI more hydrophobic with water contact angle increased to 49.32o (Gu et al. 2019; Tian et al. 2022). The increase might be related to the formation of hydrophobic FeSx species on the surface of mZVI. Meanwhile, S-mZVI/SGB also showed a high hydrophobicity, corresponding to a contact angle of 46.4° (significantly higher than that of pure mZVI). This might be attributed to the hydrophobicity increase caused by sulfur doping. It was reported that increasing the hydrophobicity of mZVI could inhibit the water corrosion of mZVI (Li et al. 2022a), leading to the increase of the electron utilization efficiency of mZVI. Therefore, compared to mZVI/GB, the enhanced Cr(VI) removal performance of S-mZVI/SGB composite was related to the reduction of side reactions after sulfidation and the increase of reducing components (reducing properties of sulfur components).

mZVI is ferromagnetic and is well-known for its high magnetic susceptibility. The magnetic properties of S-mZVI, mZVI/GB and S-mZVI/SGB at room temperature were characterized. As shown in Fig. 3d, the saturation magnetization for S-mZVI, mZVI/GB and S-mZVI/SGB were determined to be 119.80, 96.75 and 113.61 emu·g− 1, respectively. The results indicated that sulfidation could enhance the magnetic susceptibility of mZVI material. Because the magnetic property of mZVI is directly determined by its magnetic domain structure, i.e., electronic structure, the result suggested that sulfidation could alter the electronic structure of mZVI (Wang and Yang 2022). Since the reactivity of ZVI material was highly related with its electronic structure (Liu et al. 2018), the enhanced Cr(VI) removal capacity of S-mZVI might also be related to the change of its electronic structure besides the rich chemistry of FeSx. Due to the high magnetic susceptibility, all three materials could be easily separated from solution using a permanent magnet after being used for Cr(VI) removal (Fig. S8). The magnetic separation allows the easy recollection and recycle of the material, which facilitates its environmental applications.

To reveal the Cr(VI) removal mechanism, XPS was employed to further characterize the surface chemical composition of the S-mZVI/SGB composite before and after Cr(VI) removal. The XPS survey spectrum of S-mZVI/SGB composite after Cr(VI) removal revealed a new peak at 580.0 eV, indicating the presence of Cr elements in the composite (Fig. 4a). The high resolution Fe 2p XPS peak of S-mZVI/SGB composite displayed two prominent peaks at 710.5 and 724.0 eV corresponding to Fe 2p3/2 and Fe 2p1/2, respectively (Fig. 4b) (Wang et al. 2020). The peaks at binding energies of 712.0 and 725.6 eV were characteristic of Fe(III), while the peaks at binding energies of 710.3 and 723.6 eV were ascribed to Fe(II) (Wang et al. 2021b). The absence of Fe0 peak (705.9 eV) and the characteristic satellite peak of Fe(III) (718.8 eV) were attributed to the formation of FeSx shell outside the mZVI particles (Gu et al. 2017). After Cr(VI) removal, the Fe(III) satellite peak at 718.8 eV enhanced significantly, suggesting the formation of iron oxide/hydroxide and the co-precipitation of Fe(III) and Cr(III): xCr3+ + (1 − x) Fe3+ + 2H2O → CrxFe1 − xOOH + 3H+(Xu et al. 2021; Yoon et al. 2022). The peak area corresponding to the Fe species varied significantly, indicating the significant change of the proportion of Fe species after the removal of Cr(VI). The Fe(III)/Fe(II) ratio in the S-mZVI/SGB composite was calculated to be 0.63, which increased to 0.82 after Cr(VI) removal. The C 1 s peak of S-mZVI/SGB (Fig. 4c) could be deconvoluted to three peaks positioned at 284.6 eV, 285.9 eV and 288.5 eV, corresponding to C-C, C-S/C-O-S and C-S-O, respectively (Xu et al. 2014). The existence of C-S and O-S chemical bonds was attributed to the reaction of sulfur with graphite oxide during ball milling. The C-S bonding also indicated the incorporation of S in the graphite lattice of GB (Jeon et al. 2013), demonstrating the formation of S-doped GB by mechanochemical synthesis. After Cr(VI) removal, the C 1 s XPS peaks shifted to higher binding energy, suggesting the oxidation of SGB in the composite. For the high resolution S 2p XPS spectra (Fig. 4d), peaks positioned at 163.5 eV, 166.2 eV and 168.3 eV were ascribed to Sn2−, SO32− and carbon-bonded SOx species (Wu et al. 2017). After the reaction, the content of Sn2−decreased while the content of C-SOx-C increased due to the oxidation. Moreover, the peak corresponding to SO32− disappeared, which was attributed to the coverage of the surface sulfur-containing groups by the formed ferrochrome oxide layer. To further verify the formation of SGB nanosheets, the S-mZVI/SGB composite was digested with 1.0 M HCl to remove the iron content and subsequently washed with DMF thoroughly to remove free sulfur species. The treated sample was then characterized using XPS. As shown in Fig. S9, the high-resolution C 1 s and S 2p spectra confirmed the existence of C-S and C-O-S bonds. The result unambiguously confirmed the formation of SGB. As shown in Fig. 4e, the characteristic peaks corresponding to Cr 2p3/2 and Cr 2p1/2 appeared at 576.9 and 586.6 eV, respectively, with a spin-orbit split interval of 9.7 eV indicating that Cr(III) was the overwhelming dominant chromium species (Wu et al. 2019). The absence of Cr(VI) species indicated that the contribution of flocculation removal was negligible. Therefore, the removal of Cr(VI) by the S-mZVI/SGB composite followed a reduction-precipitation mechanism.

Fig. 4
figure 4

XPS spectra of S-mZVI/SGB composite before and after Cr(VI) removal. a XPS survey spectra, b high-resolution Fe 2p spectra, c high-resolution C 1 s spectra, d high-resolution S 2p spectra, and e high-resolution Cr 2p spectrum

3.3 Influence of dose and initial pH on Cr(VI) removal

3.3.1 Effect of dose

Figure 5a shows the effect of S-mZVI/SGB dose on Cr(VI) removal. With the dose of the S-mZVI/SGB composite increased from 0.5 to 1.5 g·L− 1, the removal rate of Cr(VI) increased from 45.34% to 99.15%. At each dose, the removal of Cr(VI) by S-mZVI/SGB could reach equilibrium within 3 h. The low removal rate of Cr(VI) at the dose of 0.5 g·L− 1 was ascribed to the limited number of active sites (Gheju et al. 2017). With the dose of S-mZVI/SGB increased to 0.8 g·L− 1, the Cr(VI) removal rate increased significantly to 81.3%. When the dose increased to 1.0 g·L− 1, the Cr(VI) removal rate could reach more than 99%. The dose of 1.0 g·L− 1 was therefore selected for further experiment.

Fig. 5
figure 5

The effect of a dose and b initial pH on Cr(VI) removal by S-mZVI/SGB composite

3.3.2 Effect of solution pH

The pH of the water environment is an important geochemical parameter that regulates the reaction interface between the composite and pollutant (Sethy et al. 2021). Therefore, the removal of Cr(VI) by S-mZVI/SGB was evaluated over a pH range spanning acidic and alkaline (4.0–9.0) (Fig. 5b). Lower solution pH was conducive for Cr(VI) removal, with the removal rate reached 99.27% at pH of 4.0. The Cr(VI) removal rate declined slowly with the increase of the initial pH. Nevertheless, the removal rate was still higher than 91% at neutral pH, indicating the excellent Cr(VI) removal performance of the composite. When the initial pH increased to 9.0, the removal rate of Cr(VI) decreased to 62.03%. At the end of the reaction (Fig. S10), the final pH of the solution was higher than 9.0 due to the formation of the iron and chromium hydroxides. It is well-known that the iron and chromium hydroxides form passivation layer hindering the electron transfer from Fe0 to contaminants (Zhang et al. 2019). The decline in removal rate was mainly because Fe0 is more easily getting passivated as the pH of the solution increases.

3.4 Kinetics and isotherms

Adsorption kinetics and isotherms were used to further elucidate the mechanism of Cr(VI) removal by S-mZVI/SGB, and the relevant equations are given in Text S2. Regardless of the mechanism in the removal of Cr(VI) by mZVI-based composite, it always involves the adsorption of Cr(VI) anions on the surface of the composite (Fiúza et al. 2010). To characterize the dynamic adsorption process of Cr(VI) by S-mZVI/SGB, pseudo first-order, pseudo second-order and Elovich kinetic models were used to fit the Cr(VI) removal kinetics (Fig. 6a). The removal of Cr(VI) by S-mZVI/SGB increased rapidly in the first 60 min and basically reached equilibrium after 120 min reaction. As shown in Table 1, the pseudo-second-order (R2, 0.96) and Elovich (R2, 0.93) models outperformed the pseudo-first-order kinetic model (R2, 0.87) to describe the Cr(VI) removal kinetics. This verifies the importance of chemisorption in the removal of Cr(VI) by S-mZVI/SGB. The Qe  was calculated using the pseudo first-order and pseudo-second-order model and were 26.93 and 28.05 mg·g− 1, respectively. Obviously, the Qe determined by using pseudo-second-order model was closer to the measured value. Since pseudo-second-order was the best kinetic model to describe the Cr(VI) removal kinetics, it was also used to fit the Cr(VI) removal kinetics by S-mZVI and mZVI/GB composites (Fig. S11). As shown in Table S2, the S-mZVI composite had the highest second order rate constant, which was credited to its fast electron and mass transfer.

Fig. 6
figure 6

The Cr(VI) removal kinetics (a) and isotherm (b) by S-mZVI/SGB composite

Table 1 Fitting parameters of various kinetic and isotherm models for Cr(VI) removal by S-mZVI/SGB

To further explore the mechanism of Cr(VI) removal by S-mZVI/SGB, Langmuir and Freundlich adsorption isotherm models were used to fit the isotherm data (Fig. 6b). Both models can well describe the removal of Cr(VI) by the composite with the correlation coefficient R2 > 0.99 (Table 1). The result indicated that heterogeneous chemical adsorption process was involved in the removal of Cr(VI) by S-mZVI/SGB (Wang et al. 2020). The theoretical Cr(VI) removal capacity of S-mZVI/SGB was determined to be 70.20 mg·g− 1 according to Langmuir model.

3.5 Comparison with literature data

The comparison and summary of Cr(VI) removal rates by S-mZVI/SGB and reported materials are shown in Table S1. Here, the removal of Cr(VI) by S-ZVI/SGB is significantly higher than that by commodity mZVI (as shown both in literature and by experimental data), indicating that the method described in this work has a significant activation effect on mZVI. Although the modified ZVI materials prepared by wet chemical method showed high Cr(VI) removal capacity (56.8–79.12 mg·g− 1), the use of the toxic and expensive chemical NaBH4/KBH4 was the main limitation (Wu et al. 2022). Moreover, S-mZVI/SGB showed a higher Cr(VI) removal capacity than most mZVI materials modified by ball milling.

3.6 Tafel and EIS studies

Tafel and electrochemical impedance (EIS) tests were carried out to analyze the electron transport properties of various materials (Mishra and Farrell 2005). The Tafel curves are shown in Fig. 7a, and the fitted Tafel slopes are given in Table 2. In Cr(VI) solution, the corrosion potentials of mZVI, mZVI/GB, S-mZVI and S-mZVI/SGB were − 0.34, − 0.02, − 0.54 and − 0.56 V, respectively. The fitted data shown in Table 2 indicated that the Tafel slope followed the order: S-mZVI/SGB > S-mZVI > mZVI/GB. Obviously, S-mZVI/SGB had the most negative corrosion potential and the largest Tafel slope (8916.6 mV dec− 1), suggesting a fast corrosion rate. Compared to mZVI, the corrosion potential of mZVI/GB shifted to a higher value, which was consistent with previous study that used activated carbon as additive (Wang et al. 2020). The reason for the positive shift was still unclear, nevertheless, we surmise that it may be related to the specific electron migration pathway. As shown in Fig. S12 and Table S3, the corrosion potential of mZVI/GB was − 0.63 V in 3.5% NaCl solution, which was more negative than that of mZVI. Under given conditions, the corrosion potential of a material is only related to the potential difference between the surface reactive sites and the electrolytes. The disparate electrochemical behavior of mZVI/GB in Cr(VI) and NaCl solution should be attributed to the change of the reactive sites. Therefore, the shift of the corrosion potential of the composite compared to mZVI suggested the change of the reactive sites for Cr(VI) removal. Compared to S-mZVI and mZVI/GB, the more negative corrosion potential of S-mZVI/SGB composite indicated that additional reactive sites were presented. Considering the composition differences of the three samples, the additional reactive sites should be closely related to the SGB. The result of Tafel tests clearly manifested the importance of the type of reactive sites for Cr(VI) removal.

Fig. 7
figure 7

Electrochemical analysis. a Tafel curves and b electrochemical impedance spectroscopy curves of various materials. The measurement was carried out in 30 mg·L− 1 Cr(VI) solution in an anaerobic environment

Table 2 Tafel and EIS fitting results

To further demonstrate the electron transfer advantage between S-mZVI/SGB and Cr(VI), electrochemical impedance spectroscopy (EIS) was used (Fig. 7b) (Zhang et al. 2022). The composition of internal resistance involves ohmic resistance (R1), charge transfer resistance (R2) and diffusion resistance (R3), so the equivalent circuit as shown in the inset of Fig. 7b was designed (Cai et al. 2018). The results showed that the ohmic resistance (R1) between the pollutant-composite interface was small with large error, and the fitting results were ambiguous. Therefore, the differences between R2 and R3 of each material are compared (Table 2). It is interesting to see that the mZVI/GB composite had higher electron transfer resistance than that of mZVI, which was consistent with the slow Cr(VI) removal kinetics by mZVI/GB composite. The increase of the electron transfer resistance might be attributed to the presence of GB on the surface of the mZVI particles. The GB was negatively charged due to the presence of oxygen-containing functionalities (Fig. 3b), which would retard the adsorption of negatively charged Cr(VI) anion due to electrostatic repulsion (Bandara et al. 2019). S-mZVI/SGB composite had the smallest charge transfer resistance (854.20 Ω) and diffusion resistance (796.70 Ω), followed by mZVI, S-mZVI and mZVI/GB. The result demonstrated that the electrons of Fe0 core in the S-mZVI/SGB composite were easier to migrate to Cr(VI) than other samples, which was beneficial for Cr(VI) removal.

3.7 DFT calculation

The Tafel test clearly showed that Sulfur-doping had significant influence on the adsorption of Cr(VI) on GB nanosheets. The adsorption energies of Cr(VI) anion on GB and S-doped GB nanosheet were then calculated by using DFT simulations, and Cr2O72− and HCrO4 were selected as the model ions. The details about the DFT simulations are provided in Text S3. The optimized structure for Cr2O72− and HCrO4 adsorption on two types of GB nanosheets are shown in Fig. 8. The calculated adsorption energies of HCrO4 on GB and SGB nanosheets were determined to be − 2.05 eV and − 4.07 eV, respectively. For adsorption of Cr2O72−, the adsorption energies were calculated to be − 2.80 eV and − 3.69 eV for pristine and S-doped graphene nanosheet, respectively. All of the adsorption energies were negative, which means that the adsorption of Cr(VI) anion on both GB and SGB nanosheets occurred spontaneously. Both the adsorption energies of HCrO4 and Cr2O72− on SGB were more negative than those on GB nanosheet, suggesting that the S-doping could facilitate the adsorption of Cr(VI) anions (Geng et al. 2022).

Fig. 8
figure 8

The optimized structure of CrO42− and Cr2O72− adsorbed on a undoped and b Sulfur-doped GB nanosheets

3.8 Cr(VI) removal mechanism

The removal of Cr(VI) by mZVI-based materials usually follows a reduction-precipitation mechanism, which generally involves the Fe(II) as a reducing agent (Li et al. 2022b). Therefore, the role of Fe(II) in Cr(VI) removal was investigated by adding 1,10-phenanthroline as scavenger. As shown in Fig. 9, with 0.8 g·L− 1 of 1,10-phenanthroline, the Cr(VI) removal efficiency declined by 17.6%. The results showed that the reduction of Cr(VI) to Cr(III) by Fe(II) was not the predominant pathway. Based on the XPS result, most of the Cr(VI) anions have been reduced to Cr(III) in the used S-mZVI/SGB composite. Therefore, there must be other pathways accounting for the reduction of Cr(VI). FeS species formed on the surfaces of mZVI particles may be served as reducing agent for the reduction of Cr(VI). Moreover, FeS could also promote the electron transfer from mZVI to Cr(VI). However, the Cr(VI) removal capacity of S-mZVI was much lower than that of mZVI/GB due to the fact that the formation of precipitates on FeS species would block the mass transfer pathway. So, the SGB nanosheets played an paramount role for Cr(VI) removal in the S-mZVI/SGB composite, which served as cathode to reduce Cr(VI) using the electrons from the Fe0 core. Based on the DFT calculations and the Cr(VI) performance experiments, the plausible Cr(VI) removal mechanism was proposed: (i) Cr(VI) anions are captured and fixed by the SGB nanosheets, where they accept the electrons from Fe0 to be reduced to Cr(III); (ii) Fe0 lose two electrons to form Fe(II), which migrate to the SGB nanosheets to reduce adsorbed Cr(VI) anions to Cr(III); (iii) Cr(VI) anions directly react with FeS species to form precipitates. Based on the petal-like morphology of the used composite shown in Fig. 2a, the pathways (i) and (ii) dominated the removal of Cr(VI) with FeCr hydroxide preferentially formed on the SGB nanosheets, which could effectively avoid the passivation of mZVI particles.

Fig. 9
figure 9

The Cr(VI) removal by S-mZVI/SGB composite with/without the addition of 1,10-phenanthroline (Initial pH = 5.7, initial Cr(VI) concentration = 30 mg·L− 1)

3.9 Reusability and removal of Cr(VI) from natural water

The reusability of the composite for Cr(VI) was evaluated by doing four consecutive Cr(VI) removal experiments. The reusability of S-mZVI and mZVI/GB composite was also assessed for comparison. The Cr(VI) removal rate of S-mZVI decreased by 40% in the second reuse cycle, indicating poor reusability (Fig. S13). As shown in Fig. 10a, the Cr(VI) removal rate of mZVI/GB declined gradually with increased reuse cycles. While no obvious Cr(VI) removal rate decrease was observed for the S-mZVI/SGB composite in the first three consecutive cycles. The Cr(VI) removal declined slightly in the fourth cycle, but was still greater than 95%. The results demonstrated that the S-mZVI/SGB composite has an excellent anti-passivation capability.

Fig. 10
figure 10

a The reusability of S-mZVI/SGB composite with mZVI/GB as comparison. The dose of the material is 2.0 g·L− 1, the initial concentration of Cr(VI) in each cycle is 30 mg·L− 1, and the solution pH is not adjusted. b Cr(VI) removal by S-mZVI/SGB from lake water

To further evaluate the performance of S-mZVI/SGB composites for Cr(VI) removal in natural water, the Cr(VI) removal by the composite was carried out in lake water. As shown in Fig. 10b, the Cr(VI) removal rate in lake water was 63.97%, which was considerably lower than that in pure water. The dramatic decline was attributed to the organic matter and ions present in the lake water competed with Cr(VI) for the active sites of S-mZVI/SGB, leading to the decrease of Cr(VI) removal capacity (Fang et al. 2021; Sun et al. 2016). Nevertheless, the Cr(VI) removal rate of the S-mZVI/SGB composite was decent. It is therefore feasible to remove Cr(VI) from practical wastewater using the S-mZVI/SGB composite.

4 Conclusion

In summary, S-mZVI/SGB composite was successfully synthesized by a mechanochemical approach with mZVI, sulfur powder and biochar as precursors. During the synthesis process, the sulfidation of mZVI, and the sulfur doping of graphene-like biochar nanosheets occurred simultaneously. The Fe(II) scavenging test indicated that the reduction of Cr(VI) to Cr(III) by Fe(II) was not the dominant pathway. DFT calculations indicated that SGB nanosheets have a strong adsorption capability for Cr(VI) anions. Mechanism study suggested that the Cr(VI) anions were captured by the SGB, where Cr(VI) received electrons from Fe0 core to be reduced to Cr(III). Owing to the advantage of the “capture-reduction” mechanism, the S-mZVI/SGB composite had a Cr(VI) removal capacity of 70.2 mg·g− 1, which was 56 times higher than that of mZVI.