Removal Mechanisms of Cr(VI) By Redox-Active Moieties on HNO3 Modified Biochar Under Different pH and O2 Conditions


 Nitric acid (HNO3) modified biochar (NBC) has been demonstrated to be a promising sorbent. However, the roles of their redox-active moieties (RAMs, i.e., environmentally persistent free radicals (EPFRs) and oxygen-containing function groups) in Cr(VI) removal under varying pH and O2 conditions remain poorly understood. In this study, HNO3 oxidation caused an obvious increase in specific surface area, porous volume, RAMs content, and surface potential of the biochar, leading to the more effective removal of Cr(VI) (with the removal rate reached 100% at pH 2.0) than that of the untreated biochar. Kinetics experiments revealed that O2 and pH are of great importance for the reduction efficiency and rate of Cr(VI). RAMs on NBC can either directly reduce Cr(VI)(predominant pathway) or activate O2 to produce •O2− for indirect Cr(VI) reduction. In addition, we examined the changes in the compositions of RAMs during the reaction by tuning the RAMs compositions using methanol and hydrogen peroxide. The results of electron paramagnetic resonance and X-ray photoelectron spectroscopy analysis demonstrated that the main electron donors on NBC were different at different pH values: oxygen-containing groups, e.g., –OH and C–O–C, played a dominant role in reducing Cr(VI) under acidic conditions while the neutral condition was beneficial to EPFRs-dominated reduction. This study investigated the roles of the EPFRs and oxygen-containing function groups on HNO3 modified biochar, which may provide new insights into the promoted reduction of Cr(VI) by applications of biochar.


Introduction
Due to industry and economic development, heavy metal contamination of soil and water has been a global challenge. In particular, chromium(Cr) poses great risks to the environment and human health because of its carcinogenicity, teratogenicity, and mutagenicity (J. Zhong et al., 2017). Chromium is released from electroplating, textile dyeing, tanneries, and other industrial processes (Arslan et al., 2010).
Chromium has more complicated chemical behaviors in the environment because it exists in diverse oxidation states (i.e., trivalent Cr(III) and hexavalent Cr(VI)) ( Mohan and Pittman, 2006). Cr(VI) is much more soluble, mobile, and toxic than Cr(III), and it is regarded as a priority pollutant by the U.S. EPA (Guo et al., 2012). Thus, an effective strategy to decrease the mobility and toxicity of Cr is through the adsorption-reduction method in view of its high e ciency and cost-effectiveness (Jiang et al., 2014;Wang et al., 2015).
As a carbon-rich byproduct of biomass pyrolysis, biochar is a cost-effective and environmental-friendly adsorbent with a highly porous structure and a variety of oxygen-containing functional surface groups . Among these modi cation/activation methods, chemical oxidation is a simple and effective way to add more reactive sites (e.g., oxygen-containing functional groups) to the surface of biochar (J Jin et al., 2018). For instance, HNO 3 oxidized biochar rich in carboxyl functional groups (i.e., -COOH) showed remarkably greater stabilization ability for zinc, copper, and lead (Uchimiya et al., 2012). Furthermore, HNO 3 modi ed biochar exhibited enhanced uranium (VI) adsorption ability than the virgin biochar due to the increase in -COOH content and negative surface charge (J Jin et al., 2018). However, the overall roles of HNO 3 Zhu et al., 2020). In addition to the well-known oxygen-containing functional groups (e.g., −OH, -C-O-C, and -C=O moieties) (X. Xu

Characterization
A series of analytical techniques were used to characterize BC and NBC. The morphology and elemental compositions of the biochar were investigated by Scanning electron microscopy energy-dispersive X-ray spectroscopy (SEM-EDX, Zeiss Sigma 300). Fourier transform infrared spectroscopy (FTIR) (Thermo Scienti c Nicolet IS) and the modi ed Boehm titration analysis according to the method of Zhong et al.
(D. Zhong et al., 2019) were conducted to analyze the functional groups of the biochar. The BET-surface area, pore size, and pore volume were determined using a Micromeritics ASAP 2020 Plus1.03 surface area and pore size analyzer. X-ray photoelectron spectroscopy (XPS, Thermo Scienti c K-Alpha) was employed to identify the speciation of C, O, and Cr on the biochar. The surface potential of the biochar was measured by using Malvern Zetasizer Nano-ZS90 at 25°C temperature. The EPFRs in biochar samples were measured by an electron paramagnetic resonance spectrometer (EPR, Bruker MS-5000). The ltrates were then analyzed for total Cr and Cr(VI) concentrations using atomic absorption spectroscopy (AAS, PinAAcle900F) and the 1,5-diphenylcarbazide method at 540 nm (UV-2550, Shimadzu), respectively. The aqueous Cr(III) concentration was calculated based on the difference between the total Cr and Cr(VI) in solutions. The resulting solid residues were dried in a vacuum oven at 60°C for 24 h and then subjected to the characterization mentioned above.

Adsorption experiments
To explore whether NBC can transfer electrons to O 2 to generate •O 2 − , we conducted the anoxic experiments to examine the effect of active oxygen. All solutions were purged with N 2 for at least 1 h before the anoxic reactions to ensure no dissolved O 2 existed. Once the NBC and Cr(VI) solutions were mixed in every 100 mL headspace vial sealed with caps, the bottle was immediately vacuumed and lled with N 2 . The adsorption experiment followed the same procedures as above, except for the absence of

Data analysis
All experiments were run in triplicate with well reproducible results, and the data was reported as the mean with standard deviations (SD). Error bars on the graphs represent the SD of the averages. All the results were evaluated with one-way analysis of variance (ANOVA) at a signi cance level of P ≤ 0.05 using Prism 8.

Biochar Characterization
The difference in morphology and surface structure between BC and NBC was directly displayed in the SEM-EDX images ( Fig. 1a and b). BC had an intact structure with many impurities on its surface, lacking tunnels and visible pores. In contrast, HNO 3 treated BC led to the removal of ash and alkali-metal salts (D. Zhong et al., 2018), and the corrosion made the NBC show tunnel-like structures with inner pores. Furthermore, the SEM-EDX result showed the slightly increased oxygen and nitrogen contents of NBC, which also indicated the success of HNO 3 modi cation.
The N 2 adsorption and desorption isotherm of BC constituted typical type IV curves with hysteresis loop of H3 , while that of NBC was more in line with the hysteresis loop of H4, indicating the NBC had the characteristics of microporous and mesoporous (Fig. 1e) . As expected, the NBC had a larger BET surface area (27.066 m 2 g −1 ) and pore volume (2.266×10 -2 cm 3 g −1 ) than those of BC (with 4.864 m 2 g −1 BET surface area and 9.750×10 -3 cm 3 g −1 pore volume). However, the average pore size of NBC (4.383 nm) was smaller than that of BC (21.435 nm) (Table S1). Density Functional Theory (DFT) model was used for further understanding the pore size distribution of BC and NBC. The results showed that the pore size of BC extended to the range of 11 nm -100 nm (Fig. 1c), while the NBC concentrated within the limit of 11 nm (Fig. 1d), again con rming the presence of micropores and mesopores in NBC. All these differences between BC and NBC were caused by the oxidative corrosion of HNO 3 , which results in the conversion of BC into the NBC of smaller pore size but the larger surface area and pore volume (Zhao et al., 2017).
The zeta potentials of BC and NBC were shown in Fig. 1f. BC was negatively charged across the pH range from 1 to 11. In comparison, the pH pzc of NBC was found to be at pH 4.02. Thus, the increased surface potential of NBC may favor a weaker electrostatic attraction between the biochar and negatively charged Cr(VI) species.
To further verify the changes of functional groups on biochar before and after HNO 3  . It was observable that these polar functional group peaks mentioned above had enhanced intensities on the NBC surface, indicating the oxidation reaction caused by HNO 3 treatment (Fig. S2a). Meanwhile, the dramatic decline of transmittance at 1380 cm -1 (-CH 3 ) for NBC suggested decreased nonpolar aliphatic components. The titration result was in agreement with the FTIR results, which showed that the content of the Ph-OH group in the NBC increased after HNO 3 treatment (Fig. S2b). Therefore, it can be reasonably deduced that the different removal capacity between BC and NBC is mainly attributed to the changes of surface structure, surface potential, and functional groups from the HNO 3 treatment. The species of Cr adsorbed on NBC after 24 h were examined by XPS analysis (Fig. 2d -e). In the Cr 2p spectrum, the binding energy values at 579.5 and 588.2 eV were assigned to Cr(VI), and 576.9 and 586. where t (h) is the time of adsorption; q e (mg g −1 ) and q t (mg g −1 ) are the removal capacity at equilibrium and time t, respectively; and k 1 (h −1 ) and k 2 (g mg −1 h −1 ) are the pseudo-rst-order and pseudo-secondorder rate constants, respectively.
The kinetic parameters calculated from the two models are presented in Table S2, and the corresponding plots are shown in Fig. S3. It can be seen that pseudo-second-order model was superior to the pseudorst-order model in data tting with high correlation coe cients (R 2 = 0.922, 0.945, 0.989 for pH 2.0, 5.0, 7.0, respectively). Therefore, the chemisorption of Cr(VI) was the determining step of the adsorption process. The rate-controlling step might be a chemical interaction involving surface chelation reaction or ion exchange between Cr anions and the polar functional groups on NBC ( The different Cr(VI) removal e ciencies by NBC at different initial pH values were primarily connected with Cr(VI) speciation in the aqueous solution and the surface charge of biochar. At solution pH 1.0 -6.8, Cr(VI) ions are present as Cr 2 O 7 2− and HCrO 4 − , which have more negative adsorption free energy than that of CrO 4 2− (predominant form at pH > 6.8) (Huang et al., 2016). Under the solution pH < pH pzc (4.02), the hydrated surface of NBC was protonated and positively charged, which resulted in a remarkable electrostatic attraction between the NBC and the anionic Cr(VI). Besides, a lower pH can produce a higher redox potential of Cr(VI)/Cr(III), which is conducive to the reduction of Cr(VI) by NBC (Mohan et al., 2006). Therefore, the removal e ciency of NBC under oxic condition at pH 2.0 could reach 97.54% within 24 h (Fig. S4). Nevertheless, when the solution pH > pH pzc , the hydrated surface of NBC was deprotonated and negatively charged with the increasing pH values, which inhibited the Cr(VI) adsorption owing to the electrostatic repulsion between the NBC and Cr(VI) anions. Meanwhile, the OHcan compete with Cr(VI) ions for the available adsorption sites on the surface of NBC at a higher pH (Ahmadi et al., 2016). The removal e ciencies of NBC at pH 5.0 and 7.0 reduced to 50.00% and 23.83% (Fig. S4). For the pristine BC, there was almost nil Cr(VI) removal at pH 5.0 and 7.0 due to its negatively charged surface (Fig. S4).

Direct and indirect reduction of Cr(VI) by NBC
In the Cr(VI) reduction system, the reducing agent can either directly transfer electrons to Cr(VI)(direct The removal e ciency and rate of NBC under anoxic conditions were lower than that under oxic conditions. The overall removal rate of Cr(VI) by NBC was reduced by 5.74% (pH 5.0) and 8.97% (pH 7.0) in the absence of O 2 , respectively, compared to the oxic condition (Fig. 3a). Besides, Cr(VI) removal rate constants under oxic conditions were 1.12-fold (pH 2.0), 1.22-fold (pH 5.0), and 4.46-fold (pH 7.0) higher than those at the corresponding pH values under anoxic conditions (Fig. 3b). These results con rmed that the contribution of indirect reduction to Cr(VI) removal increased with the increasing initial pH values. Besides, competitive adsorption of O 2 and Cr(VI) may happen on the NBC surface. The increasing pH converted NBC's surface charge from positive to negative, which not only weakened the electrostatic adsorption between the NBC and the anionic Cr(VI) but also inhibited the direct electron transfer from the NBC to Cr(VI (pH 5.0) and 28.43% (pH 7.0) (Fig. 3a), indicating that the direct reduction dominated the Cr(VI) removal process of NBC under varying pHs.

Roles of EPFRs
We speculated that the Cr(VI) reduction capacity of NBC might be related to EPFRs, and therefore used the EPR technique to investigate the changes in EPFRs on biochars. Both BC and NBC had the broad singlet EPR signals, which con rmed the presence of free radicals in both biochars. Besides, the EPFRs content of NBC was about twice that of BC (Fig. 4a). It was reported that EPFRs with g-factors below 2.0030 are attributed to carbon-centered EPFRs (e.g., cyclopentadienyls), those of 2.0030 -2.0040 are for carbon-centered radicals with an adjacent oxygen atom (e.g., phenoxy-derived species), and > 2.0040 for oxygen-centered radicals (e.g., semiquinone-type radicals) (Odinga et al., 2020). The g-factors of BC and NBC were 2.0057 and 2.0054, respectively. Hence, both were characteristic of oxygen-centered EPFRs, namely, semiquinone-type EPFRs. Semiquinone-type EPFRs are reported to participate in redox transformation of contaminants, such as As(III) oxidation (D. Zhong et al., 2019) and Cr(VI) reduction . Therefore, the EPFRs on NBC may directly and/or indirectly donate electrons to reduce Cr(VI) to Cr(III), accompanied by forming quinone groups on the biochar.
To investigate the effect of EPFRs in the biochar/Cr(VI) systems in depth, EPFRs were recorded before and after the Cr(VI) treatment under varying pHs (i.e., 2.0, 5.0, and 7.0) and O 2 (i.e., anoxic and oxic) conditions. As illustrated in Fig. 4b, the intensity of EPR signals under oxic and anoxic conditions decreased with decreasing solution pH, consistent with the residual Cr(VI) concentration in the solution (Fig. 2a -c). This is in accordance with the previous result (D. Zhong et al., 2018) that EPFRs could directly transfer electrons to reduce Cr(VI), and more Cr(VI) could be reduced by EPFRs at a lower pH. At the same time, these EPFRs of NBC under anoxic and oxic conditions were consumed to different extents. EPFRs were consumed to a lesser extent under anoxic conditions than under oxic conditions. Yet, the removal e ciency was the opposite, further validating that the semiquinone-type EPFRs could participate in electron transfer to O 2 for the indirect Cr(VI) reduction. These results demonstrated that semiquinonetype EPFRs may be directly and indirectly responsible for the Cr(VI) reduction, and there was high dependence of Cr(VI) reduction by EPFRs on the solution pH and O 2 .
To further investigate the role of EPFRs on Cr(VI) transformation, methanol was used to scavenge surface EPFRs before treating Cr(VI). As seen in Fig. 4a, the EPR measurement of the NBC after the methanol treatment revealed a pronounced decrease in EPFRs concentration. Meanwhile, the removal capacity of NBC-CH 3 OH was reduced by 5.83% (pH 2.0), 13.06% (pH 5.0), and 43.26% (pH 7.0) in the presence of O 2 when compared to the NBC, respectively (Fig. 5). Therefore, the decrease of EPFRs on NBC-CH 3 OH led to an apparent suppression of the removal e ciency of Cr(VI), and the contribution of EPFRs to Cr(VI) removal increased with the increasing initial pH values. This observation agreed with the previous nding that EPFRs probably act as the reductants to reduce Cr(VI) in the neutral condition instead of hydroxyl or catechol groups on the biochar (Zhao et al., 2018). Likewise, the removal capacity of NBC-CH 3 OH under oxic condition increased by approximately 4.80% (pH 2.0), 14.16% (pH 5.0), and 82.70% (pH 7.0) when compared with that under anoxic condition, respectively (Fig. 5). It further demonstrated that EPFRs could transfer electrons to O 2 for the indirect Cr(VI) reduction, and the contribution of indirect reduction increased with the increasing initial pH values.

Roles of oxygen-containing functional groups
According to the above analysis of the role of EPFRs on Cr(VI) reduction, we speculated that oxygencontaining functional groups on NBC may also play similar roles in Cr(VI) reduction by direct and/or indirect pathways. As shown in Fig. 6a, signi cant changes in FTIR spectra of the NBC before and after the reaction with Cr(VI) were observed. For example, FTIR spectra showed decreases of the peak assigned to -OH (3407 -3430 cm -1 ) and C-O-C (1030 -1160 cm -1 ) groups after reacting with Cr(VI), suggesting that they may be the key regulators for Cr(VI) reduction. The peaks representing aromatic C-H bending vibrations (800 ± 10 cm −1 ) also became inconspicuous after the reaction. Besides, the slight red shift of C=C/C=O (1596 -1612 cm -1 ) occurred mainly due to the complexation of the -COOH group with Cr (J. Xu et al., 2020). Although the involvement of oxygen-containing functional groups in the Cr(VI) reduction can not be determined by using FTIR alone (e.g., complexation also affects the FTIR peak and positions), these results still indicated that the -OH, C-O-C, and C=O were involved in the Cr(VI) reduction to some extent.
The content of acidic functional groups (Ph−OH and -COOH) on NBC was determined by a modi ed Boehm titration method before and after adsorbing Cr(VI) at different conditions. Before reacting with Cr(VI), it is worth noting that the methanol used in this study had little impact on the content of Ph−OH and -COOH groups of the NBC-CH 3 OH (Fig. 6b), again proving that the decrease of EPFRs on NBC-CH 3 OH was responsible for the reduced removal e ciency of Cr(VI) (Fig. 5). However, after Cr(VI) treatment, the Ph−OH content of NBC sharply decreased (Fig. 6c). In contrast, the -COOH content increased correspondingly, which suggested that the Ph−OH may be oxidized to the -COOH group by transferring electrons to Cr(VI) (Fig. 6d). removal e ciency deciphered the greater contribution of oxygen-containing groups in the lower pH. It is also apparent that the removal capacity of NBC-H 2 O 2 was much lower than that of NBC-CH 3 OH at pH 2.0 and pH 5.0 (Fig. 5). Noteworthily the EPFRs content of NBC-H 2 O 2 was higher, but the content of the Ph−OH group was lower than that of NBC-CH 3 OH (Fig. 4a & 6b). Therefore, the Ph−OH group plays a dominant role in reducing Cr(VI) under acidic conditions. On the contrary, the Cr(VI) removal capacity of the NBC-H 2 O 2 was getting closer to that of NBC-CH 3 OH at pH 7.0, again proving that a neutral condition was more favorable for EPFRs-dominated reduction of Cr(VI).
The transition of oxygen-containing functional groups of biochar at varying pH values for Cr(VI) reduction was also investigated by XPS. From the O 1s spectrum (Fig. 7a), the peaks around 531 eV, 532 eV, 532.9 eV, and 533. This result can be further consolidated by the XPS C1s spectra (Fig. 7b). The binding energies at 284.2, 285.6, 287.9, and 289 eV corresponded to the C-C, C-O (phenol and alcohol), C=O, and -COOH groups, respectively (Liu et al., 2020). A relative decrease in the C-O content coupled with the increases in the -COOH and C=O content were observed after reaction with Cr(VI). Therefore, these characterization results suggested that C-O in the form of phenol, alcohol, or ether probably donated electrons to Cr(VI) coupled with the formation of the -COOH and C=O groups, which was consistent with FTIR and titration analysis (Fig. 6). Correspondingly, the Cr(III) species in the NBC increased with decreasing pH (i.e., 72.4% at pH 7.0, 73.7% at pH 5.0, and 89.12% at pH 2.0), suggesting the formed -COOH and C=O groups would complex with Cr(III). Besides, the increasing initial pH led to the decreasing contents of -COOH & C=O groups and the absorbed Cr(III) on the NBC (Fig. 2, 6d, and 7), which again supported that less Cr(VI) reduction will occur in the neutral pH because of the reduced contribution of oxygen-containing groups (

Conclusion
In this work, corn straw biochar was modi ed by HNO 3 to prepare NBC with semiquinone-type EPFRs and oxygen-containing functional groups. We systematically investigated the key roles of these two redoxactive moieties (RAM) in the Cr(VI) removal at varying initial pH and O 2 levels from water. Furthermore, oxygen-containing functional groups, especially -OH and C-O-C groups, played a dominant role in reducing Cr(VI) under acidic conditions, and the formed -COOH and C=O groups could be involved in complexation with Cr(III). This study provides a better understanding of the removal of Cr(VI) by RAMs in the biochar.

Declarations
Ethics approval and consent to participate Consent for publication Not applicable Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests