1 Introduction

Dynamic changes of redox potential have been widely observed during underground water recharge of unconfined river aquifers, water table fluctuations of lake sediments, and flooding-drainage cycles of cultivated lands (Du et al. 2021; Honma et al. 2016; Zhang et al. 2020a). The frequent redox fluctuations in these water-soil environments lead to intense interactions between reduced substances and oxygen (O2), which consequently induce the production of reactive oxygen species (ROS) (Chen et al. 2021b; Page et al. 2013; Yuan et al. 2017). Ubiquitous ROS production in redox oscillation events has arisen increasing attention and the potential roles of ROS in multiple biogeochemical processes have been proposed (Yu and Kuzyakov 2021). As one of the most reactive ROS, hydroxyl radical (·OH, E0 = 2.8 V vs. NHE) has critical impacts on the transformation of redox-active metals (Kong et al. 2015; Liu et al. 2022b), degradation of organic pollutants (Xie et al. 2021), and geochemical cycling of nutrients (Zhao et al. 2021).

In terrestrial systems, ferrous iron (Fe(II)) species are the main active components to induce ·OH production, especially the aqueous Fe(II) (Fe2+), ligand-complexed Fe(II), and Fe(II)-bearing mineral species (e.g., green rust, siderite, and pyrite), with different morphology and redox sensitivity that determine their ·OH production capacity (Fang et al. 2021; Tan et al. 2022; Zhang et al. 2016). Formation of Fe(II) species is usually regulated by microbial Fe reduction processes that Fe(III)-reducing microorganisms utilize iron minerals as the terminal electron acceptors for anaerobic respiration (Weber et al. 2006). In paddy fields, agricultural strategies such as flooding-drainage management and application of remediation agents can greatly influence microbial community structure and soil redox conditions (Huang et al. 2021a). To maintain agricultural production, amendments such as pyrogenic carbon (PyC) have been widely applied to paddy fields with several tons per hectare per year (Liu et al. 2022c). Pyrogenic carbon is the porous carbonaceous material obtained from feedstock pyrolysis (300–900 °C) under O2-limited conditions, which can greatly affect soil microbial communities and physicochemical properties, including pH, water-holding capacity, and nutrient availability (Chen et al. 2022; Kappler et al. 2014; Liu et al. 2021; Yu et al. 2021b). Owing to its redox-active oxygen-containing functional groups (e.g., quinone/hydroquinone groups) and aromatic conjugated carbon structure, PyC can facilitate the electron transfer processes in multiple biotic and abiotic reactions, especially in the reduction of Fe(III) minerals, while pyrolysis temperature is a key factor in determining its electron exchange capacity (Kappler et al. 2014; Sun et al. 2021; Wu et al. 2021; Yu et al. 2022). Although the facilitation of microbial Fe reduction by PyC has been frequently reported (Kappler et al. 2014; Xu et al. 2016; Zhou et al. 2016), it is unclear whether and to what extent PyC affects the formation of active Fe(II) species (biotically or abiotically) and the consequent ·OH production.

Redox-active moieties (e.g., phenolic hydroxyl) of PyC can also serve as electron shuttles to facilitate the production of active intermediates (e.g., ·OH and Fe(IV)) via promoting the Fe(III)/Fe(II) cycling and oxidant decomposition (e.g., H2O2 and ferrate) for the control of organic pollutants in aquatic environments (Feng et al. 2021; Seo et al. 2015; Tian et al. 2020). The electrochemical properties may enable PyC to promote electron transfer from reducing agents to O2 to produce ROS (Wang et al. 2021). Even though PyC has been widely applied for soil amelioration and contaminant remediation, its influences on ROS production and associated  contaminant degradation have been extensively overlooked. Given the importance of PyC in participating in various biogeochemical processes, the integrated effects of PyC on Fe cycling and ·OH production need to be systematically investigated.

Herein, we investigated the Fe(II)/Fe(III) cycling and ·OH production in paddy soils with amendments of PyC prepared under various pyrolysis temperatures (300–700 °C). The objectives of this study were to explore (i) the effects of PyC on the Fe cycling and associated ·OH formation during redox fluctuations of paddy soils, and (ii) the underlying mechanisms of PyC participating in redox reactions. This study deepens the understanding of PyC-mediated redox processes and broadens the applicability of PyC for remediation strategies in agricultural fields that are contaminated with organic pollutants.

2 Materials and methods

2.1 Chemicals and materials

Pyrogenic carbons were prepared by pyrolyzing rice straw biomass at 300, 500, and 700 °C (denoted as PyC300, PyC500, and PyC700) under the oxygen-limited condition in a muffle furnace (Wang et al. 2021). A dwell time of 4 h for all pyrolysis temperatures was applied to stabilize the carbon structure. The synthesized PyC was ground and passed through a 100-mesh sieve. Reduced pyrogenic carbons (rPyCs) were prepared to mimic those pyrogenic carbons reduced by microorganisms or other reducing species during anoxic incubation of soils (Wang et al. 2021; Zhu et al. 2022). The rPyCs were prepared by further thermal annealing of PyC300, PyC500, and PyC700 at the same temperature and duration time as that of PyC under a nitrogen atmosphere (1 L min−1). Accordingly, these prepared rPyCs were named as rPyC300, rPyC500, and rPyC700, respectively. Detailed information on physicochemical properties is provided in Additional file 1: Table S1. Sources of chemicals are  included in Additional file 1: Text S1.

2.2 Paddy soil sampling and anoxic incubation

Two paddy soils were collected from paddy fields (0–20 cm topsoil) in Changsha city (CS) in Hunan province and Yingtan city (YT) in Jiangxi province, China. The CS and YT soils were classified as Entisol and Inceptisol based on US soil taxonomy, respectively. The total Fe and soil organic carbon (SOC) contents were 39.3 g kg−1 and 22.8 g kg−1 for CS, and 33.3 g kg−1 and 14.2 g kg−1 for YT. Details of soil sampling and their properties were described in our previous study (Chen et al. 2021b).

Batch incubations were conducted at a soil to water ratio of 1:2.5 under anoxic conditions. Briefly, 20 g dry soil, 50 mL deoxygenated ultrapure water (99.9% N2, 1 h), and 0.2 or 0.6 g (1% or 3%, w/w) PyC or rPyC were well mixed in 100 mL serum bottles in an anaerobic glovebox (Braun Co., Germany). Serum bottles were sealed with butyl rubber stoppers and aluminum caps and then placed on an orbital shaker (150 rpm, 25 °C, dark) for anoxic incubation. Control experiments without PyC were also performed under the same conditions. All treatments were conducted in triplicate. When Eh and pH values of soils became metastable after 20-day incubation (Additional file 1: Fig. S1), serum bottles were sacrificed for oxygenation experiments. Similar procedures were widely applied in previous studies to investigate ROS production (Zhang et al. 2020b), contaminant transformation (Chen et al. 2021a), and element cycling in soils (Zhao et al. 2020). To further evaluate the effects of PyC500 amendments during redox fluctuations, three more paddy soils with low SOC content (7.9–17.0 g kg−1) were collected from Chongqing (CQ), Hebei (HB), and Sichuan (SC) province of China for incubations. The main properties of soils are presented in Additional file 1: Table S2.

To distinguish the contribution of biotic/abiotic Fe reduction by PyC or microorganisms, sterilization was performed by adding 1% (m/v) HgCl2 to eliminate soil microorganisms (Tong et al. 2016). After 20-day anoxic incubation, the number of soil microorganisms was evaluated using the spread plate method and the results showed that no colony was observed in the plates (Additional file 1: Fig. S2), indicating that the addition of 1% (m/v) HgCl2 effectively inhibited the growth of microorganisms.

2.3 Oxygenation experiments

To quantify ·OH production, 5 mL homogenized soil suspension was withdrawn into 40 mL brown vials in the glovebox, and then moved out with the addition of 15 mL benzoic acid (BA, 10 mM). These vials were exposed to air for 1 min via opening caps and then shaken at 150 rpm under dark (25 °C). Anoxic control groups (99.9% N2) were also performed under the same conditions. At the predetermined time intervals, 0.5 mL suspension was withdrawn and mixed with 0.5 mL methanol. The mixture was filtered through a 0.22 μm membrane for the analysis of p-hydroxybenzoic acid (p-HBA) (Additional file 1: Text S2). The ·OH production was estimated based on the concentration of p-HBA with a conversion factor of 5.87 (Mopper and Zhou 1990). The ·OH production was also confirmed by electron paramagnetic resonance (EPR) spectrum (Additional file 1: Text S3). The dynamic changes of H2O2 were monitored with a modified acridinium ester chemiluminescence method (Additional file 1: Text S2) (Jiang et al. 2018; Zhang et al. 2020a). The capacity of PyC per se to produce ·OH was evaluated as described in Additional file 1: Text S4. To explore the mechanisms of H2O2 decomposition for ·OH production, the peroxidase-like activity of paddy slurries was determined according to the protocol of Jiang et al. and Du et al. (Additional file 1: Text S5) (Du et al. 2021; Jiang et al. 2018). To investigate the effects of biotic processes on ·OH production during oxygenation, anoxic paddy slurries were sterilized using γ irradiation (50 kGy, Co source, China) and then oxygenated as above. To investigate the changes of Fe species during oxygenation, anoxic paddy slurries were oxygenated without BA under the same conditions. At the predetermined time intervals, 1 mL mixture was sampled for sequential extraction of Fe (described in Sect. 2.5.1). All experiments were performed in triplicate with standard deviations reported in the results.

2.4 Microbial inactivation and contaminant degradation induced by ·OH

To explore the environmental implications of ·OH, cell viability of soil microorganisms and imidacloprid (IMI) degradation were investigated during oxygenation of paddy slurries. Given the important roles of microorganisms in soil redox fluctuations, the effects of ·OH on microbial activity were also explored. The number of bacteria before and after oxygenation was evaluated based on the spread plate method. Live and dead bacteria were also observed on a fluorescent microscope (IX71, Olympus, Japan) using the LIVE/DEAD BacLight™ Bacterial Viability Kit. Imidacloprid (IMI) is a commercial neonicotinoid pesticide with increasing world production (Simon-Delso et al. 2015), and thus was chosen as the model contaminant. IMI degradation was examined after 4-h oxygenation of anoxic slurries. Anoxic and quenching experiments were conducted under the same conditions to verify the critical role of ·OH. The detailed information on these experiments is described in Additional file 1: Text S6.

2.5 Analytical methods

2.5.1 Sequential extraction of Fe species

Sequential extraction was applied to examine the changes of Fe species in slurries during anoxic incubation and oxygenation experiments according to our previous work (Chen et al. 2021b). In brief, 1 mL homogenized soil suspension was sequentially extracted using 1 M CaCl2, 0.5 M HCl, 5 M HCl, and 1.3 M HF with 1.8 M H2SO4 to obtain four Fe phases including (i) exchangeable Fe, (ii) surface-bound/complexed Fe and Fe in low-crystalline minerals (e.g., ferrihydrite, siderite, and green rust-like phases), (iii) high-crystalline Fe-bearing minerals (e.g., hematite), and (iv) Fe in silicates and clays, respectively (Additional file 1: Text S7).

2.5.2 Soil bacterial 16S rRNA gene amplification, Illumina sequencing and data analysis

To investigate the effects of PyC on soil microbial communities, 1 mL homogeneous anoxic paddy slurry was centrifuged at 14,000×g for 1 min (4 °C), and the total DNA of microbial communities was extracted using FastDNA™ SPIN Kit for Soil (MP Biomedicals, Irvine, California USA). The quality and concentrations of total DNA were identified by NanoDrop2000 spectrophotometry (Thermo Scientific, USA). The V3–V4 region of the 16S rRNA genes was amplified using universal primer 341F (forward primer, 5′-CCTAYGGGRBGCASCAG-3′) and 806R (reverse primer, 5′-GGACTACNNGGGTATCTAAT-3′). Purified PCR products were subjected to Illumina MiSeq platform (Illumina Inc., San Diego, USA) for high-throughput sequencing. After filtering out the poor-quality sequences and those did not match the reference database, sequences for per sample ranged an average from 50,251 to 59,345. The raw sequence data are deposited in the National Genomics Data Center (NGDC) (https://ngdc.cncb.ac.cn/gsa) (GSA: CRA010021). The α/β diversity estimation and taxonomic classification were performed on Quantitative Insights into Microbial Ecology (QIIME2) software (Additional file 1: Text S8).

2.5.3 Other analyses

57Fe Mössbauer spectra were applied to identify the Fe species before and after oxygenation (Additional file 1: Text S9). Elemental composition of PyC was analyzed by Elementar vario EL cube and the contents of surface functional groups were characterized by X-ray photoelectron spectroscopy (XPS). Attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR, VERTEX 70v, Bruker Corporation, Germany) analyses were conducted to further examine the surface functional groups of PyC (Additional file 1: Text S10). Persistent free radicals (PFRs) in PyC were monitored by EPR analysis (Additional file 1: Text S3). Electron accepting capacity (EAC) and electron donating capacity (EDC) of PyC were analyzed using Ti(III) citrate and ferricyanide as described by Xin et al. (Additional file 1: Text S11) (Xin et al. 2019). Scanning electron microscope equipped with an energy dispersive spectrometer (SEM–EDS, Quanta FEG-250) was used to determine the morphology of minerals attached or deposited on the surface or pore structure of PyC.

3 Results and discussion

3.1 ·OH production during oxygenation of soils after anoxic incubation

Extensive ·OH was produced during oxygenation of CS and YT slurries (Fig. 1), while negligible ·OH was trapped under anoxic conditions (Additional file 1: Fig. S3a). ·OH production was further confirmed by the EPR spectrum with four-equidistant lines and the peak intensities ratio of 1:2:2:1 (Additional file 1: Fig. S3b). After 8-h oxygenation, ·OH production was 150.6 ± 2.8 and 132.0 ± 3.5 μM in the CS- and YT-control, respectively. The addition of 1% PyC negligibly affected ·OH production in CS slurries (Additional file 1: Fig. S4a). The presence of 3% PyC500 prominently increased ·OH production by 12.6% (169.5 ± 6.9 μM, p < 0.05) in CS slurry, whereas the addition of 3% PyC300 or 3% PyC700 rarely influenced ·OH production (Fig. 1a). Compared to CS slurries, all PyC amendments increased ·OH production in YT slurries by 7.6–7.9%, 15.6–24.8%, and 15.9–21.0% (p < 0.05) for 1–3% PyC300, PyC500, and PyC700 treatments, respectively (Fig. 1b and Additional file 1: Fig. S4b). The highest ·OH production (164.8–169.5 μM) in the 3% PyC500 treatments indicated the strongest capacity of PyC500 in promoting ·OH production during oxygenation. Within the initial time points (0, 10, and 30 min), ·OH production was well fitted with pseudo-zero-order kinetics (R2 > 0.86, Additional file 1: Table S3). Interestingly, the observed pseudo-zero-order rate constants (kobs,·OH) increased with higher amendments of PyC. With the amendment of 3% PyC500 and PyC700, kobs,·OH significantly increased to 164.9 ± 4.7 and 161.7 ± 18.9 μM h−1 in CS slurries, and 143.5 ± 0.8 and 151.9 ± 2.1 μM h−1 in YT slurries, much higher than that of CS-control (112.9 ± 2.5 μM h−1) and YT-control (101.8 ± 3.9 μM h−1). These results indicated that the addition of PyC accelerated ·OH production during the oxygenation of anoxic paddy slurries. Besides, compared to CS soil, a more significant promotion of ·OH production by PyC500 was observed in YT soil with lower SOC content. To further evaluate the critical role of PyC500 on ·OH production, oxygenation experiments were conducted for three more paddy soils with low SOC content. Results showed that the most significant promotion by PyC500 amendment was observed in HB soil with the lowest SOC content (7.90 g kg−1) (Fig. 1d). In the presence of 3% PyC500, ·OH production increased from 63.5 ± 2.0 to 92.2 ± 2.9 μM (~ 1.5-fold) in CQ soil, 5.1 ± 1.6 to 40.0 ± 6.1 μM (~ 7.8-fold) in SC soil, and 11.3 ± 8.7 to 241.6 ± 1.2 μM (~ 21.3-fold) in HB soil, respectively.

Fig. 1
figure 1

·OH production in a Changsha (CS) and b Yingtan (YT) slurries without (control) and with 3% (w/w) pyrogenic carbon (PyC300, PyC500, and PyC700); c ·OH production in CS slurries with 3% (w/w) reduced pyrogenic carbon; d ·OH production in Chongqing (CQ), Hebei (HB), and Sichuan (SC) slurries with 3% (w/w) PyC500 during 8-h oxygenation

Full size image

The relative contribution of the aqueous phase to ·OH production was also examined (details in Additional file 1: Text S12). Within the initial 1 min, ·OH quickly formed in the aqueous phase of all CS treatments, corresponding to the quick Fe2+ oxidation (Additional file 1: Fig. S5a, b). Linear correlation was observed between the ·OH production and initial Fe2+ concentration (R2 = 0.98, p < 0.05, Additional file 1: Fig. S5c), rather than dissolved organic matter (DOM) (R2 = 0.80, Additional file 1: Fig. S5d). To investigate the role of DOM in ·OH production, ferrozine was used as Fe(II) stabilizing ligand (Hudson et al. 2022) to rule out the contribution of Fe2+. Results showed that ·OH production decreased by 87.5–91.2% with the addition of 2 mM ferrozine (Additional file 1: Fig. S6), indicating that Fe2+ was the main contributor to ·OH production in the aqueous phase. In all treatments, the contribution of the aqueous phase to total ·OH production was < 1.5%, indicating that the solid phase was mainly responsible for ·OH production.

3.2 Changes of Fe(II) species during anoxic incubation

To investigate the effects of PyC on Fe reduction, the formation of Fe(II) species was examined during anoxic incubation. In CS-control, the total content of Fe(II) was 7.78 ± 0.10 g kg–1 after 20-day incubation. The contents of CaCl2-, 0.5 M HCl-, 5 M HCl-, and HF-Fe(II) were 0.66 ± 0.10, 4.56 ± 0.01, 1.45 ± 0.07, and 1.11 ± 0.04 g kg–1, respectively (Fig. 2). Specifically, 0.5 M HCl-Fe(II) was the most abundant Fe(II) species, accounting for 58.5% of total Fe(II).

Fig. 2
figure 2

Dynamic changes of a CaCl2-Fe(II), b 0.5 M HCl-Fe(II), c 5 M HCl-Fe(II), and d HF-Fe(II) of Changsha (CS) paddy soil without (control) and with 3% (w/w) pyrogenic carbon (PyC300, PyC500 and PyC700) during 20-day anoxic incubation

Full size image

In CS slurries, the addition of 1% PyC showed negligible influence on Fe reduction whereas 3% PyC significantly affected Fe(II) formation during incubation (Additional file 1: Table S4). Compared with the control, the addition of 3% PyC300 increased the contents of CaCl2-, 0.5 M HCl-, 5 M HCl-, and HF-Fe(II) by 25.8%, 4.4%, 37.1%, and 12.5%, respectively (Fig. 2). Addition of 3% PyC500 increased 11.7% of 0.5 M HCl-Fe(II), 52.7% of 5 M HCl-Fe(II), and 14.1% of HF-Fe(II), whereas these Fe(II) species decreased by 0.4%, 32.8%, and 45.1% in the 3% PyC700 amended slurries. In general, total Fe(II) content increased by 13.5% and 17.4% with the addition of 3% PyC300 and 3% PyC500 but decreased by 15.3% with the addition of 3% PyC700 (Additional file 1: Fig. S7a).

Compared to CS slurries, the amendment of PyC greatly increased Fe reduction in YT slurries even with 1% amendment dosage. The addition of 1% or 3% PyC300 and PyC500 increased the total Fe(II) content (by 6.8–16.7% with PyC300 and 9.5–27.7% with PyC500). Specifically, the addition of 3% PyC300 and 3% PyC500 increased the content of 0.5 M HCl-Fe(II) by 25.6% and 46.3%, 5 M HCl-Fe(II) by 62.8% and 97.4%, and HF-Fe(II) by 11.8% and 19.0%, respectively (Additional file 1: Fig. S8). The increased contents of 5 M HCl-Fe(II) and HF-Fe(II) in both CS and YT soils suggested that PyC300 and PyC500 promoted the crystallization of Fe(II) phases. On the contrary, the addition of 1% or 3% PyC700 decreased 2.8–6.3%, 0.6–19.4%, and 9.3–42.6% of CaCl2-Fe(II), 5 M HCl-Fe(II), and HF-Fe(II), respectively, but increased 23.8–45.2% of 0.5 M HCl-Fe(II) in YT soil. The formation of 0.5 M HCl-Fe(II) and decrease of 5 M HCl-Fe(II) and HF-Fe(II) in PyC700 amended slurries (especially YT-3%PyC700) suggested that PyC700 may prevent the conversion of low-crystalline Fe(II) to high-crystalline Fe(II) phases or promote the transformation of high-crystalline Fe(II) to low-crystalline Fe(II). SEM spectra showed that fine Fe mineral particles existed on the surface or inside the inner pores of PyC after anoxic incubation (Additional file 1: Fig. S9). The close aggregation of PyC and Fe minerals may favor the electron transfer from cells to PyC and minerals, thus promoting Fe reduction (Bonneville et al. 2009; Gao et al. 2022). The similar distribution of C and Fe further indicated that newly formed Fe minerals associated with PyC to form stable organometallic C–O–Fe complexes (Yang et al. 2016), and thus probably prevented the transformation of Fe species (Xu et al. 2022).

3.3 Oxidation and transformation of Fe(II) species during oxygenation

3.3.1 The critical role of Fe(II) in ·OH production

To validate the contribution of Fe(II) to ·OH production, 2,2′-bipyridine (BPY) was added to chelate Fe2+ and surfaced-exposed Fe(II) (Katsoyiannis et al. 2008). In the presence of 5 mM BPY, ·OH production was suppressed by 74.5–80.1% in CS slurries (Additional file 1: Fig. S10), indicating that Fe(II) species, rather than redox-active SOC or microorganisms, were mainly responsible for ·OH production. Application of γ irradiation before oxygenation rarely affected ·OH production (Additional file 1: Fig. S11), which further confirmed that microorganisms negligibly impacted ·OH production. The amount of ·OH produced by PyC or rPyC was <2.2 μM (Additional file 1: Fig. S12a), which accounted for <1%  of total ·OH production in the CS and YT slurries. Each gram of microbially-reduced PyC300, PyC500, and PyC700 was oxidized to produce 13.2, 23.4, and 33.5 nmol ·OH, respectively (Additional file 1: Fig. S12b and Table S5). Thus, ·OH generated by microbially-reduced PyC at 3% (w/w) amendment dosage contributed to <0.5% of total ·OH produced in the CS and YT soil. Correlation analysis showed ·OH production positively correlated with total Fe(II) in CS slurries (p < 0.05, Additional file 1: Fig. S13a). However, no significant linear correlation between ·OH production and total Fe(II) was observed in YT slurries (p > 0.05, Additional file 1: Fig. S13b), suggesting that different Fe(II) species were responsible for ·OH production probably due to the presence of PyC. To further investigate the effects of PyC on ·OH production, sequential extraction was applied to examine the changes of Fe species during oxygenation.

3.3.2 Changes of different Fe(II) species during oxygenation

For CS soils, the addition of 1% PyC rarely promoted ·OH production (Additional file 1: Fig. S4a) and the oxidation of Fe(II) species showed no significant difference (Additional file 1: Fig. S14). Therefore, we mainly discussed Fe(II) oxidation in the 3% PyC treatments. For CS slurries, CaCl2-Fe(II) was completely oxidized within 0.5 h in all treatments (Fig. 3a). For YT-control, about 80.2% of CaCl2-Fe(II) was oxidized after 8-h oxygenation (Fig. 3e). However, complete oxidation of CaCl2-Fe(II) was observed in PyC treatments, and the oxidation rates increased from 1.36 ± 0.06 g kg–1 h–1 (YT-control) to 1.65 ± 0.13 g kg–1 h–1 (YT-3%PyC300), 1.54 ± 0.14 g kg–1 h–1 (YT-3%PyC500), and 1.51 ± 0.04 g kg–1 h–1 (YT-3%PyC700), respectively (Additional file 1: Table S6). During oxygenation, the pH of YT-control slurries gradually decreased from 7.1 to 5.3, but that of PyC treatments was higher than 6.1 after oxygenation (Additional file 1: Fig. S15). Higher pH was probably responsible for the higher oxidation rates of CaCl2-Fe(II) in PyC treatments (Garg et al. 2018).

Fig. 3
figure 3

Changes of extractable Fe(II) species (CaCl2-Fe(II), 0.5 M HCl-Fe(II), 5 M HCl-Fe(II), and HF-Fe(II) in a–d Changsha (CS) and e–h Yingtan (YT) slurries without (control) and  with 3% (w/w) pyrogenic carbon (PyC300, PyC500, and PyC700) during 8-h oxygenation

Full size image

During the initial 1 h oxygenation, oxidation rates of 0.5 M HCl-Fe(II) in CS- and YT-PyC treatments (2.65–2.74 g kg–1 h–1 and 1.67–1.85 g kg–1 h–1, Additional file 1: Table S6) were higher than those in CS-control (2.54 ± 0.07 g kg–1 h–1) and YT-control (1.29 ± 0.18 g kg–1 h–1). As for 5 M HCl-Fe(II), the oxidation rates were 0.23 ± 0.18, 0.34 ± 0.07, 0.63 ± 0.06, and 0.21 ± 0.10 g kg–1 h–1 in the CS-control, CS-3%PyC300, CS-3%PyC500, and CS-3%PyC700 slurries, respectively. Similar results were observed in YT slurries that oxidation rates of 5 M HCl-Fe(II) with the addition of PyC300 and PyC500 were higher than that of the YT-control. Negligible oxidation of HF-Fe(II) was observed in the CS- and YT-control treatments (Fig. 3d, h). Compared with other active Fe(II) species, HF-Fe(II) was resistant to be oxidized. Even so, approximated 15.3% and 10.1% of HF-Fe(II) were oxidized in the CS-3%PyC300 and CS-3%PyC500 slurries, whereas 15.6%, 48.7%, and 63.3% of HF-Fe(II) were oxidized in the YT-3%PyC300, YT-3%PyC500, and YT-3%PyC700 slurries, respectively. These results suggested that oxidation of HF-Fe(II) was enhanced in the PyC - amended soils, especially in YT slurries, indicating that HF-Fe(II) also participated in ·OH production.

3.3.3 Transformation of Fe species during oxygenation

Results of sequential extraction of Fe species showed that contents of CaCl2-Fetotal decreased during oxygenation of CS and YT slurries, indicating the transformation of exchangeable Fe to other Fe phases (Additional file 1: Fig. S16a). In CS slurries, 0.5 M HCl-Fetotal and 5 M HCl-Fetotal decreased while more HF-Fetotal was produced, which indicated that these Fe species gradually transformed into high-crystalline Fe species (Additional file 1: Fig. S16b–d) (Chen et al. 2021b; Winkler et al. 2018). In YT slurries, the quick decrease of CaCl2-Fe(II)total was accompanied by the rapid formation of 0.5 M HCl-Fetotal during the initial period, suggesting that these aqueous/exchangeable Fe species transformed into low-crystalline Fe species (Additional file 1: Fig. S16e, f). Generally, rapid oxidation of Fe2+ usually generates low-crystalline Fe oxides (e.g., ferrihydrite) (Cismasu et al. 2016; Park and Dempsey 2005). However, the increase of 0.5 M HCl-Fetotal was less than the consumption of CaCl2-Fetotal, suggesting that 0.5 M HCl-Fetotal further transformed into higher-crystalline Fe phases simultaneously, consistent with the decrease of 5 M HCl-Fetotal and formation of HF-Fetotal (Additional file 1: Fig. S16c, d, g, h).

Transformation of Fe phases in the CS-control and CS-3%PyC500 slurries was characterized using Mössbauer spectroscopy. Five major components were distinguished in both slurries, including QSD-FeII, QSD-FeIII, HFD-Hae, HFD-Gt, and HFD-Fe (Fig. 4). For CS-control, the fraction of QSD-FeII declined from  10.3% to 5.9% after 4 h and slightly decreased to 5.4% after 24 h (Fig. 4a, c). For CS-3%PyC500, more QSD-FeII (9.3%) was oxidized, indicating that the presence of PyC facilitated Fe(II) oxidation. The decline of HFD-Fe (i.e., amorphous Fe oxides) after 24-h oxygenation in both slurries suggested the transformation of low-crystalline Fe to higher-crystalline Fe phases. The increased proportion of QSD-FeIII, HFD-Hae, and HFD-Gt indicated that more crystalline Fe phases were produced during oxygenation processes (Liptzin and Silver 2009; Thompson et al. 2006), which was consistent with the formation of 5 M HCl-Fetotal and HF-Fetotal as mentioned before (Additional file 1: Fig. S16). Within the initial 4 h, QSD-FeIII and HFD-Gt species increased by 1.0–2.0% and 2.2–4.6% in the CS-control and CS-3%PyC500 soil, respectively. However, the fractions of QSD-FeIII and HFD-Fe decreased by 1.7% and 4.1% in CS-control thereafter, respectively, which was accompanied by the increase of HFD-Gt (7.2%). These results indicated that QSD-FeIII and HFD-Fe species slightly transformed into HFD-Gt species. As aforementioned components of QSD-FeIII and HFD-Fe species, including ligand/surface-complexed Fe and Fe in the most disordered phases, might undergo a slight transformation into high-crystalline phases due to the surface electron transfer from electron donors (Huang et al. 2021b; Pedersen et al. 2005). Different from CS-control, the QSD-FeIII species increased by 4.5% from 4 to 24 h in the CS-3%PyC500, which was probably ascribed to the preservation and stabilization of newly formed QSD-FeIII species by PyC due to its porous structures (Additional file 1: Fig. S9), thus preventing its transformation to high-crystalline phases during oxygenation (Chen et al. 2015; Jones et al. 2009).

Fig. 4
figure 4

Fitted Mössbauer spectra (13 K) of Changsha slurries (CS-control and CS-3%PyC500) before (0 h) and after oxygenation (4 and 24 h). In each spectrum, grey line is the total calculated fit, through discrete data points. The resolved spectral components and assignments are: (1) QSD-FeII (red), Fe(II) in primary minerals, clays or adsorbed; (2) QSD-FeIII (yellow), Fe(III) in silicates, clays, and organic complexes (+paramagnetic FeIII-(oxyhydr)oxides); (3) HFD-Hae (dark blue), Fe(III) in hematite; (4) HFD-Gt (light blue), Fe(III) in goethite; (5) HFD-Fe (lilac), FeIII-(oxyhydr)oxides near their blocking temperature (TN < collection temp). The detailed fitting parameters are presented in Additional file 1: Table S7

Full size image

3.4 Critical role of PyC in Fe cycling and ROS production

3.4.1 Mechanisms of increased Fe(II) reduction with PyC during anoxic incubation

The aforementioned results showed that PyC300 and PyC500 significantly increased the formation of Fe(II) species in both CS and YT slurries (Sect. 3.2). Pyrogenic carbon can promote electron transfer between microorganisms and Fe minerals by acting as electron shuttles (Lu et al. 2021). To evaluate the electron exchange capacity of PyC, the electron donating/accepting capacity of three PyC was examined (Additional file 1: Fig. S17a). The EDC values of PyC300, PyC500, and PyC700 were 0.22 ± 0.04, 0.28 ± 0.01, and 0.08 ± 0.02 mmol e– g–1, respectively. The EAC values were the lowest for PyC300 (0.04 ± 0.08 mmol e– g–1) but similar for PyC500 and PyC700 (1.57 ± 0.04 and 1.57 ± 0.12 mmol e– g–1). The electron exchange capacity (EEC, sum of EAC and EDC) of PyC500 (1.84 mmol e– g–1) was the highest among the three PyC. FTIR spectra showed the enhanced intensity of oxygen-containing functional groups (i.e., C=O and C–O) at 1621 and 1091 cm–1 in PyC300 and PyC500 (Additional file 1: Fig. S17b). Elemental analysis showed that C content increased from 62.4% to 81.4% with increasing pyrolysis temperature, coupled with a decrease in H and O contents (Additional file 1: Table S1). These results demonstrated that the fraction of oxygen-containing functional groups significantly declined with increasing pyrolysis temperature, and more phenolic moieties were presented in PyC300, whereas quinone moieties predominated in PyC500 and PyC700 (Klupfel et al. 2014). Raman spectra showed that the ratio of D band (∼ 1350 cm−1) to G band (∼ 1580 cm−1) peak intensity (ID/IG) increased from 0.854 to 0.956 with increasing temperature (Additional file 1: Fig. S17c), indicating more graphitic structure presented in PyC700 (Ferrari and Robertson 2000). The electrical conductivity of PyC300 and PyC500 was negligible, whereas that of PyC700 was 0.42 S·cm−1 (Additional file 1: Fig. S17d). The highest conductivity of PyC700 was attributed to the graphitic structure, which can facilitate electron transfer processes (Si et al. 2022; Sun et al. 2017). EPR  spectrum analysis revealed that PyC500 contained abundant carbon-centered persistent free radicals (1.09 × 106 spin g−1, g-factor of 2.0027, Additional file 1: Fig. S17e), which are also involved in electron transfer processes (Xu et al. 2016).

To distinguish the contribution of PyC (abiotic) and microorganisms (biotic) to Fe reduction, soils were sterilized by 1% (m/v) HgCl2 to eliminate microorganisms. Interestingly, despite soil microorganisms being eliminated, Fe reduction was observed in both CS and YT soils with the addition of PyC300 (1.44 and 1.17 g kg−1) and PyC500 (1.16 and 0.76 g kg−1) (Additional file 1: Fig. S18). These results indicated that PyC300 and PyC500 reduced Fe minerals abiotically, which was consistent with their high EDC (Additional file 1: Fig. S17a). Abiotic reduction of Fe minerals by PyC300, PyC500, and PyC700 contributed to 22.2%, 17.1%, and 4.0% of total Fe reduction in CS soil, and 19.7%, 11.4%, and 4.6% in YT soil, respectively (Additional file 1: Table S8). Abiotic reduction by different PyC resulted in the transformation of different Fe(II) species in soils (Additional file 1: Fig. S19). Specifically, PyC300 significantly promoted the formation of 5 M HCl-Fe(II), and PyC500 promoted the formation of 0.5 M HCl-Fe(II) and 5 M HCl-Fe(II). The formation of high-crystalline Fe(II) in the presence of PyC300 and PyC500 may be due to the reversible redox reactions between quinone and hydroquinone groups, which can transform low-crystalline Fe minerals to high-crystalline minerals through a dissolution-reprecipitation mechanism (Lian et al. 2022). Biotic reduction of high-crystalline Fe minerals (e.g., goethite and magnetite) was slower than that of low-crystalline Fe minerals (e.g., ferrihydrite and lepidocrocite) (Bonneville et al. 2009; Dong et al. 2020). Hence, based on the results of unsterilized and sterilized experiments, we suggest that 0.5 M HCl-Fe(II) in PyC300- and PyC500-amended slurries was mainly produced by Fe(III)-reducing microorganisms, while the formation of 5 M HCl-Fe(II) was ascribed to the abiotic reduction by PyC300 and PyC500. For sterilized CS-3%PyC700 and YT-3%PyC700, total Fe(II) contents were similar to those of the sterilized control groups (Additional file 1: Fig. S18). However, the content of 0.5 M-Fe(II) significantly increased, accompanied by the decrease of HF-Fe(II) (Additional file 1: Fig. S19), suggesting that conductive PyC700 may participate in Fe cycling and electron transfer processes through motivating electrons from Fe-bearing silicate minerals to low-crystalline Fe species and thus accelerated the electron cycling. Similar results were observed in the rPyC treatments (Additional file 1: Figs. S20 and S21), and interestingly, the abiotic Fe reduction by rPyC700 was more significant than that of PyC700 and contributed to more than 10% of total Fe reduction, probably owing to its higher EDC than PyC700 (0.57 ± 0.01 vs. 0.08 ± 0.02 mmol e– g–1). Both the redox-active functional groups and graphitic structure of PyC contributed to electron transfer processes in abiotic Fe reduction in soils (Additional file 1: Fig. S22). The redox-active functional groups in PyC300 and PyC500 were mainly responsible for the electron exchange process, whereas direct electron transfer of graphitic structure dominated for PyC700, corresponding to the results from Sun et al. (2018).

Compared to humic substances that also act as electron shuttles, PyC abiotically reduced Fe minerals and induced the formation of different Fe(II) species, while soil humic substances showed limited Fe reduction capacity in the absence of microorganisms (Roden et al. 2010; Stern et al. 2018). During anoxic incubation, significant promotion of Fe reduction by PyC was observed in YT soil with lower SOC content versus CS soil (14.2 g kg−1 vs. 22.8 g kg−1). The enhanced Fe reduction by PyC500 in YT soil suggested that PyC500 acted as efficient electron shuttles to accelerate redox processes, especially in soils with lower SOC content. Therefore, anoxic incubation was conducted with/without the addition of 3% (w/w) PyC500 for three more paddy soils (SC, CQ, and HB) with low SOC content (7.9 − 17.0 g kg−1). Results showed that PyC500 significantly increased total Fe(II) content after 20-day anoxic incubation and therefore promoted ·OH production during oxygenation (Additional file 1: Fig. S23 and Fig. 1d). The changes of Fe reduction (especially redox-active Fe species, Additional file 1: Fig. S24) ultimately regulated ROS production, and a more significant enhancement of ROS production induced by PyC was observed in soils with lower SOC content (i.e., HB soil with 7.90 g kg−1 SOC). These results showed that PyC500 exhibited a strong capacity to promote Fe cycling and ROS production during redox fluctuation of different paddy soils.

3.4.2 Microbial community analyses of anoxic soils

To explore the mechanisms of microbially mediated Fe reduction, the influence of PyC on soil microbial community composition was also investigated via 16S rRNA gene analyses. Amplicon sequence variants (ASVs) were identified using DADA2. The α-diversity indexes at a sequencing depth of 50,000 are listed in Additional file 1: Table S9. Sequencing results showed that the dominant bacteria at the phylum level in CS treatments were Firmicutes, Bacteroidetes, Proteobacteria, Acidobacteria, and Actinobacteria, accounting for more than 80% of total reads (Fig. 5a). For YT slurries, the dominant bacteria were Firmicutes, Bacteroidetes, Proteobacteria, and Acidobacteria (Additional file 1: Fig. S25a). Principal component analysis (PCoA) showed that the microbial community composition of PyC-amended slurries were significantly different from that of the control groups (PERMANOVA’s p < 0.01, Fig. 5b and Additional file 1: Fig. S25b), indicating that PyC changed the microbial community.

Fig. 5
figure 5

a Microbial compositions of Changsha (CS) slurries without (control) and with 3% (w/w) pyrogenic carbon (PyC300, PyC500, and PyC700) at phylum level; b principal component analysis (PCoA) of microbial community; c relative abundance of the dominant Fe(III)-reducing microorganisms at genus level (more than 0.5%)

Full size image

Previous studies have reported that amendment of PyC promoted the microbial Fe(III)-reduction in paddy soils by elevating the relative abundance of Fe(III)-reducing microorganisms including Geobacter (Zhou et al. 2017), Methannosarcina (Liu et al. 2011), Anaeromyxobacter (Wang et al. 2020), Clostridium (Li et al. 2017), and Bacillus (Wang et al. 2017). The dominant Fe(III)-reducing microorganisms in CS-control was Anaeromyxobacter (8.8%) (Fig. 5c), whereas other species, like Methanosarcina and Geobacter, were the major bacteria in the PyC-amended treatments. Compared to the control, the relative abundance of these Fe(III)-reducing microorganisms varied with PyC addition. For example, the relative abundance of Methannosarcina significantly increased from  0.9% to 5.8% in 3% PyC300 treatments. The amendment of 3% PyC500 and 3% PyC700 increased 4.3% and 3.5% (p < 0.05) of the relative abundance of Geobacter. The total relative abundance of the aforementioned Fe(III)-reducing microorganisms in CS-3%PyC700 treatment (9.0%) was less than that of CS-control (14.0%), which could account for the less Fe(II) formation during anoxic incubation. Similar results were also observed in YT slurries (Additional file 1: Fig. S25c). Although the abiotic Fe reduction by PyC500 was lower than that of PyC300 (17.1% vs. 22.2% in CS soil and 11.4% vs. 19.7% in YT soil), amendment of PyC500 significantly elevated the relative abundance of Geobacter, resulting in higher Fe(II) formation after anoxic incubation. Meanwhile, the higher EEC of PyC500 than that of PyC300 also accelerated the electron transfer from microorganisms to Fe minerals, thus facilitating microbial Fe reduction. The respective abundance and Fe-reducing capacity of these bacteria, together with abiotic reduction by PyC, were mainly responsible for the Fe(II) formation in PyC-amended treatments (Kugler et al. 2019). Nonetheless, unraveling the contribution of biotic or abiotic processes to Fe reduction is extremely challenging, especially in complex soil matrices (Kappler et al. 2021). The promotion of Fe reduction in the presence of PyC cannot be separately ascribed to the electron exchange capacity of PyC or the elevated abundance of Fe(II)-reducing microorganisms, but largely to their synergistic effects (Yang et al. 2021).

3.4.3 Relations between Fe(II) oxidation and ·OH production

During oxygenation, ·OH production slowed down after 2 h, probably because of the quick consumption of active Fe(II) species, followed by the declined electron transfer from less-reactive interior Fe(II) species to O2 (Yu et al. 2021a). Correlation analyses showed that the content of 5 M HCl-Fe(II) was positively correlated to ·OH production in the CS treatments (p < 0.05, Additional file 1: Fig. S26b), whereas a linear relationship was observed between 0.5 M HCl-Fe(II) and ·OH production in the YT treatments (p < 0.05, Additional file 1: Fig. S26e). For CS slurries, the proportion of CaCl2-Fe(II) in CS slurries was < 10%, while 0.5 M HCl-Fe(II) and 5 M HCl-Fe(II) accounted for 76.4–86.7% of total Fe(II) The sum of kobs,0.5 M HCl-Fe(II) and kobs,5 M HCl-Fe(II) was positively correlated to kobs,·OH (p < 0.05, Additional file 1: Fig. S26d). These results indicated that 0.5 M HCl-Fe(II) and 5 M HCl-Fe(II) mainly contributed to ·OH production in CS slurries. In YT slurries, CaCl2-Fe(II) accounted for more than 30% of the total extracted Fe(II), and the higher oxidation rates of CaCl2-Fe(II) in the PyC-amended slurries were consistent with their higher ·OH production rates, suggesting that CaCl2-Fe(II) was also the active Fe(II) species contributing to ·OH production. Besides, CaCl2-Fe(II) and 0.5 M HCl-Fe(II) accounted for 65.0–80.8% of total Fe(II), while the proportion of 5 M HCl-Fe(II) was in the range of 12.2–18.8% in YT slurries. Therefore, the rapid oxidation of CaCl2-Fe(II) and 0.5 M HCl-Fe(II) contributed more to ROS production than that of 5 M HCl-Fe(II) in YT slurries. Together, active Fe(II) species including CaCl2-Fe(II), 0.5 M HCl-Fe(II), and 5 M HCl-Fe(II) were mainly responsible for ·OH production during oxygenation.

3.4.4 Mechanisms of ·OH production in the presence of PyC

To explore the electron transfer processes, ·OH production was examined in the presence of NBT (O2.– scavenger) and catalase (H2O2 scavenger) during oxygenation. Results showed that ·OH production was suppressed by 39.1–55.8% in the presence of 1 mM NBT and by 46.2–56.0% with the addition of 4000 U L–1 catalase in all treatments (Additional file 1: Fig. S10). These results indicated that O2.– and H2O2 were the necessary intermediates for ·OH production via the Haber–Weiss mechanism (Voinov et al. 2011), and both one- and two-electron transfer mechanisms contributed to ·OH production.

Instantaneous H2O2 concentration was also monitored to further confirm the H2O2 production during oxygenation (Additional file 1: Fig. S27). The instantaneous H2O2 concentrations decreased rapidly within 2 h, which was well consistent with the rapid ·OH production during initial periods (Additional file 1: Table S3). To further explore the H2O2 decomposition, the peroxidase-like activity of these slurries was examined. Before anoxic incubation, both CS and YT slurries (with or without PyC) rarely exhibited peroxidase-like activity (Additional file 1: Fig. S28a, c). After 20-day anoxic incubation, the peroxidase-like activity of CS-control was 0.37 ± 0.03 U mg–1, which increased to 0.51 ± 0.02, 0.55 ± 0.02, and 0.52 ± 0.06 U mg–1 for the CS-3%PyC300, CS-PyC500, and CS-PyC700, respectively (Additional file 1: Fig. S28b). Similar results were observed in YT slurries, which showed that peroxidase-like activity increased from 0.24 ± 0.04 U mg–1 (YT-control) to 0.33 ± 0.01, 0.36 ± 0.02, and 0.34 ± 0.01 U mg–1 for the YT-3%PyC300, YT-3%PyC500, and YT-3%PyC700 (Additional file 1: Fig. S28d). Peroxidase-like activities of PyC300, PyC500, and PyC700 alone (0, 0.003 ± 0.001, and 0.021 ± 0.002 U mg–1) were much lower than those of anoxic slurries. Thus, the higher peroxidase-like activity of PyC-amended slurries might be attributed to the formed Fe(II) species instead of PyC per se. These results demonstrated that PyC amendment significantly enhanced the peroxidase-like activity of paddy slurries and thus promoted H2O2 decomposition, which was consistent with their higher ·OH production rates (Additional file 1: Table S3).

Although PyC itself produced negligible ·OH (Additional file 1: Fig. S12 and Table S5), they can participate in electron transfer processes to facilitate Fe(II) oxidation and ROS production during oxygenation of anoxic soils. After normalization by the content of oxidized Fe(II), the normalized yields of ·OH (μmol ·OH per mM Fe(II)) occurred in the following order: CS-3%PyC500 (6.86 ± 0.23) ≈ CS-3%PyC700 (6.82 ± 0.25) > CS-control (5.57 ± 0.47) ≈ CS-3%PyC300 (5.49 ± 0.14). Although PyC300 amendment increased 13.5% Fe(II) formation, its promotion on ·OH production was negligible (Fig. 1), probably because of the higher ·OH consumption by DOM (122.4 ± 1.8 mg L–1 vs. 60.5 ± 0.8 mg L–1 in the control, Additional file 1: Fig. S29). Higher Fe(II) utilization in the PyC500 and PyC700 treatments was also consistent with their higher kobs,·OH (Additional file 1: Table S3). Although the amount of total Fe(II) of CS-3%PyC700 was the lowest in all treatments, it exhibited a fast ·OH production rate and high ·OH production efficiency. Our previous study found that PyC700 efficiently transferred electrons from sulfides to O2, resulting in higher ·OH yields than those of PyC300 and PyC500 (Wang et al. 2021). Thus, besides the increase of soil pH, the critical role of PyC700 in promoting ·OH production might be also ascribed to its strong capacity to accelerate electron transfer from reducing agents (e.g., Fe(II) and S2–) to O2. As Additional file 1: Fig. S30 shows, PyC with higher EAC (e.g., PyC500 and PyC700) quickly accepted electrons from Fe2+ and then delivered electrons to O2 to produce ·OH, acting as effective electron shuttles.

Compared to PyC, higher contents of phenolic C–OH groups were observed in rPyC (Additional file 1: Figs. S31b, S32), and thus rPyC exhibited stronger EDC and EEC (Additional file 1: Fig. S31a). Furthermore, to better distinguish the main parameters of different PyCs in regulating Fe cycling and ·OH production, CS and YT soils added with 3% (w/w) rPyC300, rPyC500, or rPyC700 were anoxically incubated following the same procedure of PyC treatments. Similar to the results of PyC, the addition of rPyC500 also promoted Fe reduction in CS and YT soils after 20-day anoxic incubation (Additional file 1: Fig. S33). Oxygenation experiments showed that rPyC500 prominently increased ·OH production by 36.2% (207.8 ± 8.4 μM, p < 0.05) in CS soil and by 26.3% (192.7 ± 12.6 μM, p < 0.05) in YT soil after 8-h oxygenation (Fig. 1c and Additional file 1: Fig. S34), and the associated kobs,·OH also significantly increased (Additional file 1: Table S10). Besides, accelerated oxidation of active Fe(II) species was observed in the presence of rPyC (Additional file 1: Fig. S35 and Table S11). The normalized yields of ·OH also increased with the addition of rPyC in CS and YT slurries (Additional file 1: Table S12), suggesting that rPyC increased the utilization efficiency of electrons to generate ROS during Fe(II) oxidation. The results of rPyC were similar to those of PyC, indicating that PyC (especially PyC500 and rPyC500) greatly increased the Fe(II) oxidation and ·OH production during the oxygenation process. Pyrolysis temperature affected the physicochemical properties of PyC and rPyC, and authentically impacted various biochemical processes in soil, such as Fe cycling and ROS production. Based on the results of PyC and rPyC, correlations among physicochemical properties of pyrogenic carbons and ·OH production rates and yields were characterized by a Pearson correlation analysis (Additional file 1: Fig. S36). The kobs,·OH was highly correlated to EEC and EDC (R2 = 0.82, p < 0.05), demonstrating the critical role of the electron shuttle capacities of PyC in regulating ROS production during oxygenation processes.

3.5 Environmental significance of ·OH

3.5.1 Microbial inactivation

To investigate the impacts of generated ·OH on soil microbial activity, the numbers of bacteria before and after oxygenation were examined. Compared to anoxic groups, approximately 22.5% and 23.1% of total bacteria were inactivated in CS-control and CS-3%PyC500 slurries after oxygenation (Additional file 1: Fig. S37a). With the addition of 100 mM ethanol to quench ·OH, no significant difference was observed under anoxic or oxic conditions. Fluorescent microscopy images further showed that more cells were inactivated after oxygenation without ·OH scavenger (Additional file 1: Fig. S38b, e). These combined results demonstrated that the produced ·OH was mainly responsible for microbial inactivation during oxygenation.

3.5.2 Pollutant degradation

Under the anoxic condition, less than 3% of IMI was eliminated after a 4-h reaction in all slurries (Additional file 1: Fig. S39). After 4-h oxygenation, only 7.8% of IMI was lost in CS-3%PyC500 slurry, probably due to ·OH quench induced by the high content of soil organic matter. In comparison, more IMI degradation (18.3%) was observed in YT slurries in the presence of 3%PyC500. With the addition of 1 M methanol, only 0.1% and 6.5% of IMI were diminished in YT-control and YT-3%PyC500, suggesting the critical role of ·OH in IMI degradation. Despite the fact that recent studies have reported that O2·– and H2O2 participated in oxidation of pollutants (Hong et al. 2023; Liu et al. 2022a; Si et al. 2022), higher redox potential of ·OH (E0 = 2.8 V) may favor the degradation of organic pollutants in soils (Yu and Kuzyakov 2021). Degradation of IMI in the YT and CS slurries was enhanced with the addition of PyC500, which was ascribed to the higher ·OH production as aforementioned (Fig. 1). Except for exogenous anthropogenic organic pollutants, the oxidation of soil organic carbon by reactive ·OH has also been proposed (Chen et al. 2021c). Assuming that 1 mol ·OH can effectively react with humic substances to produce 0.3 mol CO2 (Goldstone et al. 2002; Page et al. 2013), the enhanced ·OH production in CS- and YT-3%PyC500 slurries would result in 5.6–9.8 μmol L–1 more CO2 production after 8-h oxygenation.

4 Conclusion

Given the importance of PyC in participating in various biogeochemical processes, an investigation into the integrated effects of PyC on Fe cycling and ROS production in soil systems is required. Under the anoxic condition, PyC (especially PyC500 with high EEC) promoted the formation of active Fe (II) species mainly through abiotically reducing Fe minerals and enhancing the relative abundance of Fe(III)-reducing microorganisms, and therefore increased ·OH production during oxygenation. With PyC amendment, accelerated Fe(II) oxidation was consistent with the higher ·OH production rate, indicating that enhanced Fe(II)/Fe(III) cycling and ·OH production occurred. The produced ·OH was capable of inducing microbial inactivation and organic pollutant degradation during oxygenation processes. Overall, this study helped to better understand the critical roles of PyC in driving Fe(II)/Fe(III) cycling and ROS production under redox conditions. Given the significance of Fe(II)/Fe(III) cycling and ROS production, these processes consequently influence the mobility of toxic metals (e.g., arsenic), degradation of organic pollutants, and nutrient cycling (e.g., C, N, and P). As a frequently used functional material, the utility of PyC to increase the content of soil organic carbon should be examined during redox fluctuations in long-time scales  due to the enhanced ·OH production and probable organic carbon decomposition.