1 Introduction

Soil is the largest terrestrial carbon pool, containing approximately 1.5 × 1012 kg organic carbon (Wang et al. 2020e), which stores over twofold higher than the atmospheric and biotic pools (Lehmann and Kleber 2015). Organic carbon and organic matter normally represent a similar series of organic compounds in soil, whereas the concept of organic carbon is typically used to emphasize the carbon element in organic matter compared with organic nitrogen and phosphate in the soil. The concentration of soil organic carbon (SOC) ranges widely (Lamichhane et al. 2019). For instance, the SOC concentration in China could change from 16.01 to 42.62 Mg C ha− 1 due to the different soil properties, and it increased by about 140 kg C ha− 1 year− 1 (Zhao et al. 2018b). In addition to the changeable concentration, SOC has a complicated and variable composition, including plant and animal residues with different degrees of decomposition, substances synthesized through various natural and anthropogenic reactions in soil, and the biomass of living soil organisms and their associated metabolic products. The varying composition and concentration of organic carbon greatly affect the biogeochemical processes in different soils.

Humus (humic acid, fulvic acid, and humin) and some non-humic dissolved organic matter (e.g., organic acids) are the most widely studied organic carbon components to obtain a deep understanding of their crucial roles in different soil reactions (Ondrasek et al. 2019; Paul 2016; Polak et al. 2019; Wang et al. 2020e; Yang et al. 2020). Recently, pyrogenic carbon (e.g., biochar) from biomass burning in fields, which is rich in Amazonian dark earth, has garnered increasing attention owing to its high stability for long-term carbon sequestration (Lehmann 2007; Lehmann et al. 2021). In addition, the porous structure, rich surface functionality, and condensed graphitic carbon of biochar offer outstanding capacity for the regulation of other soil properties, including the immobilization of pollutants (Gong et al. 2022; Wang and Wang 2019; Xu et al. 2021c) and fertility (Baki and Abedi-Koupai 2018; Marcińczyk and Oleszczuk 2022). In addition to naturally generated organic carbon, human activities might result in the introduction of some organic carbon in the soil, such as carbon-based materials for soil application, leading to further variations in soil properties (Marcińczyk and Oleszczuk 2022; Xie et al. 2015). Living organisms, including bacteria, fungi, actinomycetes, soil animals, and archaea, also comprise a non-negligible portion of organic carbon in the soil; however, their content and composition are highly site-specific and time-dependent.

The impact of different organic carbon on fertility, soil remediation, and carbon sequestration is systematically studied (Hoffland et al. 2020; Lehmann and Kleber 2015; Maestrini et al. 2015; Ondrasek et al. 2019; Smreczak and Ukalska-Jaruga 2021; Verbeeck et al. 2020; Woolet and Whitman 2020), and their potential role in the abiotic redox-induced transformation of pollutants in soil has been highlighted recently (Aftabtalab et al. 2022; LaCroix et al. 2021; Xu et al. 2020c, 2022b). The considerable redox capacity of organic carbon could directly affect the redox reaction of potentially toxic elements (PTEs), which are the trace elements with the potential toxicity to humans. The mobility and toxicity of PTEs will be greatly altered during the redox-induced transformation process (Ali et al. 2020), thus affecting their risks to humans. For instance, the oxidation of arsenic (As(III) to As(V)) and thallium (Tl(I) to Tl(III)) by the oxidative moieties of SOC decreases their mobility (Chen et al. 2022; Cuong et al. 2021; Xu et al. 2021b). In addition, the reduction of chromium (Cr(VI) to Cr(III)) by organic carbon reverses the charge of ions from negative to positive, leading to different combination processes in the soil (Xu et al. 2022a, b). The indirect impact of organic carbon on redox-induced transformation is also critical, such as coupling the redox reaction with iron minerals or electron mediation (Qiao et al. 2021). The chelation effect of dissolved organic carbon, especially for that with rich O- or N-moieties, might affect the transformation of PTEs because it can increase the availability of redox-active moieties (e.g., Fe ions) in the soil, thereby enhancing the redox-induced transformation of PTEs. The transformation of PTEs is vital because it not only affects their mobility and stability but also their toxicity. For instance, As(V) and Cr(III) are usually less toxic than their other valence states (As(III) and Cr(VI)).

To the best of our knowledge, no critical review has comprehensively studied the abiotic redox-induced transformation of typical PTEs in soil with a wide range of organic carbon content. Various kinds of non-living organic carbon, including humus, biochar, non-humic dissolved organic matter, and anthropogenic carbon, are included in this review because they represent the majority of SOC (> 80%) and are relatively stable compared to living organic carbon. Notably, although the concept of humus has been refuted recently since the contemporary techniques failed to find convincing evidence and the definition was not precise enough (Lehmann and Kleber 2015; Weng et al. 2021), we still use these concepts to describe the humic substances as a composition of SOC in this review since they are widely used in related papers. This review critically summarizes the following: (1) typical SOC with different redox properties; (2) direct redox-induced transformation of PTEs with SOC; (3) indirect transformation of PTEs with SOC coupled with active soil moieties; and (4) remediation strategies and potential risks based on the PTEs transformation process through SOC. This review can advance our understanding about the fates and risks of PTEs with organic carbon in the soil and guide sustainable remediation of PTEs.

2 Organic carbon in the soil and its redox reactivity

2.1 Humus

Humus (humic substances) is a traditional concept with different definitions: strict operation according to what can be extracted with an alkaline solution, with further subcategories of humic acids (HA), fulvic acids (FA), and unextractable humin; as an existing substance that is not merely an operational construct; or as a combination of the two (Lehmann and Kleber 2015). As stated before, we still use these concepts to describe the humic substances as a composition of SOC since they have been extensively used in published papers, although the concept of humus has been refuted recently (Lehmann and Kleber 2015; Weng et al. 2021). Different small assimilable biopolymers and monomers are generated through the decomposition process of dead plants and animal matter under different strategies, and humus comprises these molecules with supramolecular structures (Fig. 1) (Lehmann and Kleber 2015; Paul 2016). Different theories, such as humification, have been developed for describing the formation of humus, and they all agree that the detailed composition of humus is highly related to the feedstock precursors and formation strategies, leading to the highly variable properties of humus (Lehmann and Kleber 2015).

Fig. 1
figure 1

Existence and transformation of organic carbon in the soil with the following processes: (1) progressive decomposition of biomass residues to CO2 by microbes; (2) humification of different biopolymers and monomers by microbes and minerals; (3) pyrolysis or incomplete combustion of biomass with wildfire; (4) release of a small fraction of organic carbon through physical, chemical, and biological degradation; and (5) deployment of carbon-based materials for different applications, wherein artificial carbon could be partially degraded over a long duration

Full size image

Humin is an insoluble component even during strong-base extraction because of its highly condensed structure and combination with minerals, and the inherent quinone and sulfur-containing moieties are the redox-active centers (Pham et al. 2021, 2022). HA and FA are soluble in a strong-base solution, but only HA precipitates when the strong-base extract is acidified, owing to its higher molecular weight and less surface functionality of the HA (Tang et al. 2014). Both HA and FA contain various redox-active functional groups, such as phenol, hydroxyl, and quinone, thereby enabling them to participate in the redox reaction directly (Eqs. 1–3 and Fig. 2) (Bai et al. 2020; LaCroix et al. 2021). For instance, the quinone can accept electrons with the formation of hydroquinone, thus causing oxidation (Eq. 1). In addition, the transition among hydroxyl, carbonyl, and carboxyl accompanied by the electron exchange also leads to the redox reaction. The quinone-hydroquinone/phenol moieties (E0 = 0.70 V, Table 1), instead of the aliphatic structure, are the primary electron-donating or withdrawing moieties (He et al. 2019; Klüpfel et al. 2014b; Walpen et al. 2018), offering a reversible redox reactivity (on 822 μmol e− g of HA) (Aeschbacher et al. 2010). In addition, the persistent free radicals (PFRs), hydroxyl, and N/S-containing functionalities might be responsible for up to 50% of the redox reactivity of the humic substances (He et al. 2019; Shi et al. 2021). The higher molecular weight fraction of humus may contain a higher amount of carbon-centered radicals, leading to a higher reducing capacity than that of the lower molecular weight fraction with oxygen-centered radicals (Shi et al. 2021).

(1)
$$-\mathrm{C}\mathrm{HO}+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to -\mathrm{C}{\mathrm{H}}_2\mathrm{OH}$$
(2)
$$-\mathrm{COOH}+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to -\mathrm{CHO}+{\mathrm{H}}_2\mathrm{O}$$
(3)
Fig. 2
figure 2

Major electron-donating and electron-accepting moieties on the organic carbon (left), and electron transfer with the organic carbon via different mechanisms (right)

Full size image
Table 1 Standard electrode potentials related to the redox reactions of PTEs in the soil environment
Full size table

In addition to the direct redox reaction with PTEs, humus can serve as the redox buffer and mediator to facilitate electron transfer (Fig. 2) (Bai et al. 2019; Zou et al. 2021). The potential capacity of humus for radical generation has been evaluated (Chen et al. 2018; Wu et al. 2021), which significantly affects the transformation of PTEs. The PFRs or reductive functionality of humus can reduce O2 with the formation of H2O2 and free radicals (Eqs. 4–7) (Kim et al. 2022; Page et al. 2012). Although several studies have also reported that humic substances are radical scavengers (Cheng et al. 2022; Donham et al. 2014), the radical activation process by humus can be facilitated after complexation with minerals (Liao et al. 2019a; Wang et al. 2021b; Yu et al. 2021; Zeng et al. 2020), leading to a higher probability of radical activation in the soil. The redox-active minerals could donate parts of electrons to complete the formation of radicals with HA (Yu et al. 2021), and the homogeneous H2O2 decomposition by HA-chelated metal complexes would facilitate ·OH formation (Liao et al. 2019a; Zeng et al. 2020). Solar irradiation could also enhance the generation of radicals with humus owing to the excited states of humus (Wan et al. 2021a; Wang et al. 2021b; Wu et al. 2021), and this phenomenon might mainly occur in the surface layer of soil.

$${\mathrm{O}}_2+\mathrm{SOC}\ \left({\mathrm{e}}^{-}\right)\to {{\mathrm{O}}_2}^{\cdotp -}+\mathrm{SOC}$$
(4)
$${{\mathrm{O}}_2}^{\cdotp -}+{\mathrm{H}}^{+}\to \mathrm{H}{{\mathrm{O}}_2}^{\cdotp }$$
(5)
$$2{\mathrm{H}}^{+}+{{\mathrm{O}}_2}^{\cdotp -}+\mathrm{SOC}\ \left({\mathrm{e}}^{-}\right)\to {\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{SOC}$$
(6)
$${\mathrm{H}}^{+}+{\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{SOC}\ \left({\mathrm{e}}^{-}\right)\to {\mathrm{H}}_2\mathrm{O}+\mathrm{O}{\mathrm{H}}^{\cdotp }+\mathrm{SOC}$$
(7)

2.2 Biochar

Biochar is a relatively newer concept than humus, but it has a similarly long history in the soil. Biochar is mainly produced from biomass combustion in the complete or partial absence of oxygen (Fig. 1), leading to the formation of carbon-rich residues with higher stability (Pignatello et al. 2017). Previously, it had some older names with similar concepts, such as black carbon, carbonized biomass, pyrogenic carbon, and charcoal, but the expression of biochar was first defined in 2006 by Lehmann et al. (2006), highlighting its carbon sequestration nature. Although current studies mainly focused on the engineered preparation of biochar for different soil applications, its natural source from the on-site combustion of plant-based biomass is still critical.

The redox properties of biochar are highly related to the production temperature because of the transition in surface functionality (Klüpfel et al. 2014a; Xin et al. 2021; Zhang et al. 2018b, 2019b). Similar to humus, the quinone and phenol moieties are vital for the redox performance of biochar (Fig. 2). Moreover, phenol would transform to quinone with the increase of temperature, leading to the transition of reductive biochar to oxidative biochar (Klüpfel et al. 2014a; Xin et al. 2021). In addition, other oxygen-containing functionalities (e.g., carboxyl, hydroxyl, and carbonyl) may be responsible for the electron transfer capacity of biochar (Xu et al. 2020c; Zhang et al. 2019b; Zhong et al. 2019). In addition to the reversible electron transfer through surface functionality (geo-battery), the conductivity of biochar facilitates a fast electron transfer routine, also denoted as the geo-conductor mechanism (Fig. 2) (Sun et al. 2017, 2018; Yang et al. 2021). The formation of a graphitic-like structure with condensed aromatic rings during high-temperature pyrolysis (> 700 °C) is vital to the fast electron transfer process, and this might be the unique electron transfer routine for biochar as opposed to humus. Moreover, the highly ordered aromatic ring might stabilize the PFRs on the biochar, leading to higher PFRs content than humus (Ruan et al. 2019). Biochar-derived dissolved organic matter also contains considerable activity for the redox reaction (Xu et al. 2021a; Zhang et al. 2019a).

Engineered biochar produced by the pyrolysis of selected waste under designated conditions is increasingly studied as a soil amendment. Compared with the naturally occurring biochar or other pyrogenic carbon in the soil, biochar manufactured purposefully through a wide range of feedstocks and precise temperatures with different modification or activation processes could accomplish high performance for the desired applications (Li et al. 2020; Lü et al. 2022; Wen et al. 2017; Xu et al. 2022a). The engineered biochar can demonstrate a higher redox reactivity with rich surface functionality and a higher surface area than natural biochar for the redox-related remediation of PTEs (Alkurdi et al. 2019; Li et al. 2020; Lü et al. 2022). However, if the biochar is designed for soil conditioning and fertility enhancement, its impact on the redox-induced transformation of PTEs needs further consideration based on the biochar properties such as surface functionality, carbon structure, particle size, and surface area. In addition, the newly-added biochar may have a different redox reactivity than the naturally occurring biochar, because the latter generally undergoes a long-term aging process with increasing O-functionalities and decreasing electron-donating capacities (Wang et al. 2020c; Xu et al. 2021c).

Owing to the rich mineral content in the soil, biochar is known to extensively combine with soil minerals (Yang et al. 2016; Zhao and Zhou 2019), resulting in a variable electron transfer capacity and route (Xu et al. 2021c, 2022b). The formation of an organo-mineral coating on the biochar surface can protect the carbon phase of biochar with a decrease in the redox reactivity (Jing et al. 2022; Yang et al. 2016). In addition, the electrons from biochar might be “captured” by minerals with the formation of reductive species (e.g., Fe(II)), causing the decrease in electron reactivity (Xu et al. 2020b, 2022b). In contrast, some redox-active metal ions might facilitate the electron transfer through valence transformation (Xu et al. 2021c), and cationic ions can overcome the charge repulsion between biochar and anions, enhancing the electron transfer probability (Xu et al. 2021c; Yang et al. 2018).

2.3 Non-humic dissolved organic matter

Non-humic dissolved organic matter refers to a series of free, small organic carbon constituents without supramolecular humus structures (Polak et al. 2019), as the widely used concept of dissolved organic matter (DOM) or dissolved organic carbon (DOC) might have some overlapping characteristics with humus. The main components are organic acids, poly-/mono-saccharides, phenol-aromatics, and lipids (Yang et al. 2020). Although the content of DOC is lower than that of humus and its composition is highly variable, its impact is still critical for the redox processes of PTEs, especially for organic acids. Recent studies have been conducted on the impact of organic acids and saccharides on the redox-induced transformation of PTEs in soil (Deng et al. 2020; Liu et al. 2019; Xu et al. 2019c).

Organic acids are a significant part of the non-humic dissolved organic matter in the soil, especially in the rhizosphere (Adeleke et al. 2017; Hubova et al. 2017). Among the different organic acids, oxalic, citric, malic, acetic, formic, and tartaric acids have been identified to play significant roles in soil because of their relatively higher contents (Adeleke et al. 2017; Xu et al. 2019c). The α-hydroxyl on organic acids has a higher reducing capacity and might participate in the electron-donating process, while the carboxyl group typically has a limited impact on the direct electron donation (Liu et al. 2019; Xu et al. 2019c). The strong chelation effect of hydroxyl and carboxyl groups on organic acids might also affect the redox-induced transformation of PTEs, because of their different redox potentials and solubilities after chelation (Liu et al. 2019; Saad et al. 2017). For instance, organic ligands could lead to the formation of Cr-ligands through chelation, facilitating the Cr(VI) reduction process (Liu et al. 2019). In addition, chelation with redox-active metals (e.g., Fe) contributes to further electron transfer with PTEs (Saad et al. 2017; Xu et al. 2020b). The N−/S-containing functional groups (e.g., −NH2 and −SH) in amino acids also play a non-negligible role in the redox cycle of the elements in soil owing to their strong redox reactivity and chelation effect (Bhattacharyya et al. 2019).

2.4 Anthropogenic carbon sources

In addition to naturally occurring organic carbon, different carbon-based materials have been added to soil for various applications, including carbon sequestration, soil fertility enhancement, and environmental remediation (Mazarji et al. 2021; Pignatello et al. 2017; Rashid et al. 2021). A wide range of carbon-containing materials have been added to the soil, and they might have different reactivities for the redox-induced transformation of PTEs in the environment.

In addition to biochar-based materials, as mentioned before, there has been a significant interest in using natural organic materials for soil amendment, such as biopolymers (starch and lignin) (Pang et al. 2019; Rashid et al. 2021; Yaashikaa et al. 2022). These materials might also have a considerable redox potential, which could be enhanced after modification (He et al. 2019). Composting is a common method for organic modification and depolymerization (Farid et al. 2022), and it degrades the stable carbon structure with the formation of a humus-like structure containing rich quinone and phenol moieties (Duran et al. 2022; He et al. 2019), thus leading to high redox reactivity and performance. The addition of inorganic catalysts, such as minerals, during the composting process can accelerate depolymerization and humic substance formation (Jiang et al. 2021; Wu et al. 2018; Zheng et al. 2021).

Other carbon-based materials, such as activated carbon, graphene, and carbon nanotubes, can also be added to the soil for remediation (Li et al. 2018; Mazarji et al. 2021; Park et al. 2016). The primary redox-reactive moieties on these carbon materials should be similar to pyrogenic carbon (e.g., phenolic and quinone moieties); however, they usually exhibit higher reactivity, a large surface area, rich active sites, small particle size, and well-formed graphitic structure, thereby significantly influencing the transformation of PTEs (Wang et al. 2020a). The aging process also has a critical impact on the redox performance of the anthropogenic carbon materials, and it is a combined process involving oxidation, aggregation, and fragmentation (Ding et al. 2022a; Park et al. 2016), leading to different performances in the redox-induced transformation of PTEs.

3 Direct redox reaction between PTEs and organic carbon

3.1 Reduction

The SOC had a higher reducing potential, leading to the direct reduction of several PTEs. Chromium (Cr) is the most widely studied PTEs because the reduction of Cr(VI) to Cr(III) decreases the toxicity of Cr. The electrical properties of Cr also change from negative (CrO42−/HCrO4−) to positive (Cr(OH)X(3-X)+), leading to a different fate of Cr in the soil. Surface functionalities such as phenolic functionality, PFRs, and α-hydroxyl on organic carbon are the main reducing moieties for Cr(VI) reduction (Xu et al. 2020b, c; Zhao et al. 2018a). The structural defects on the graphitic carbon structure, which widely occur on the pyrogenic carbon and graphene, also provide considerable reducing activity for Cr(VI) reduction (Xu et al. 2020c). The reduction of Cr(VI) by organic carbon is facilitated under the following conditions: (i) an acidic environment (Xu et al. 2019b); (ii) accelerated electron transfer with redox-active metal ions (i.e., iron (Fe) and manganese (Mn)) (Xu et al. 2020b, 2021c); (iii) condensed graphitic structure on pyrogenic carbon (e.g., biochar) for fast electron transfer (Sun et al. 2017; Xu et al. 2020c); (iv) cations for overcoming the charge repulsion (Xu et al. 2021c); and (v) accumulation of Cr(VI) and organic carbon on the solid surface.

In addition to Cr(VI) reduction, SOC can reduce other PTEs with similar critical reducing moieties. The formation of zero-valent mercury (Hg(0)) through Hg(II) reduction by the reduced quinones from organic carbon was found, which was two to six times faster than that from the photochemical reduction of Hg(II) (Zheng et al. 2012). The reduction of copper (Cu(II) to Cu(I) or even Cu(0)) could also be achieved by organic carbon with a high reducing capacity and the reaction was sensitive to the O2 concentration, because the oxic environment causes the re-oxidation of Cu(I) to Cu(II) (Fulda et al. 2013; Xing et al. 2020). In addition, the reductive formation of nano-sized silver particles might occur with the phenol-like groups of organic matter, which contribute to 11–31% of the Ag(0) particles in soil (Huang et al. 2019). The reduction of Tl(III) to Tl(I), Sb(V) to Sb(III) (antimony), As(V) to As(III), and Sn(IV) to Sn(II) (tin) by the reductive surface functionalities of organic carbon has also been observed in soil under reducing conditions with lower Eh, which increases the solubility and toxicity of these elements with significant potential risks (Aftabtalab et al. 2022; Rinklebe et al. 2020; Zhong et al. 2022).

3.2 Oxidation

The oxidizing potential of organic carbon is normally lower than its reducing potential due to the nature of organic carbon. Quinone and semi-quinine-types PFRs are the primary oxidizing moieties on organic carbon, which can directly oxidize the As(III) to As(V) under an alkaline environment (Zhong et al. 2019). A rich surface oxidative functionality with a good aromatic structure usually offers a higher capacity for the oxidation of As(III). Hg(0) oxidation by the surface functionality of organic carbon was also evaluated, and the low-molecular-weight thiol compounds, glutathione, and mercaptoacetic acid could be potential oxidizing moieties (Zheng et al. 2012, 2019). However, the organic carbon might fail to directly oxidize some PTEs because of the higher demand for the oxidizing potential, while the indirect impact of the redox cycle on other elements might be the main reaction route.

4 Redox-induced transformation of PTEs by soil moieties with organic carbon

4.1 Redox-induced transformation of PTEs by the redox cycle of O2 with organic carbon

The redox cycle of organic carbon with O2 usually results in a higher redox capacity and can even be the determining process for the redox-induced transformation of PTEs that can hardly be affected by organic carbon directly. The interaction between O2 and surface functionalities, such as the phenolic functionality, is believed to affect both the reduction and oxidation of PTEs with the formation of reactive oxygen species (ROS).

4.1.1 Formation of ROS with O2

Reactive oxygen species can be considered as a series of oxidants obtained by the incomplete reduction of O2 compared with the complete reduction to H2O (Yu and Kuzyakov 2021). The main ROS in soil include the superoxide anion (O2·-), hydroperoxyl radical (HO2·), hydrogen peroxide (H2O2), and hydroxyl radicals (OH·) (Fig. 3 and Table 1) (Yu and Kuzyakov 2021). The Fenton-like reaction with iron, photocatalytic reaction, and input from rainwater are believed to be the primary input of ROS in soil, while the reduction of O2 by organic carbon might also potentially form ROS (Eqs. 4–7) (Trusiak et al. 2018). Although several studies have found that ROS have a higher possibility of reacting with organic carbon and generating H2O and CO2 (Patel et al. 2021; Yu and Kuzyakov 2021), the incomplete reduction by organic carbon could still contribute to the formation of a considerable amount of ROS, thereby altering the fate of PTEs (Zhong et al. 2019).

Fig. 3
figure 3

Redox reactions between organic carbon and O2, Fe-minerals, and/or Mn-minerals in the soil environment

Full size image

Reduced humic acids with rich reductive surface functionalities could support the reduction of O2 with the formation of H2O2 and OH· under dark conditions, and the yield could reach 42–160 mmol per mole of electrons donated by the reduced HA (Page et al. 2012). Similarly, the reduction of O2 with the formation of ROS occurred with pyrogenic carbon in the absence of light (Fang et al. 2015; Zhong et al. 2019, 2020). In addition, DOM has a high potential to activate O2, especially for those with lower molecular weights (Du et al. 2021). The potential reactive moieties on the organic carbon for the formation of ROS could be the oxygen functionalities, nitrogen functionalities, and PFRs (Chen et al. 2017; Huang et al. 2021; Navalon et al. 2017). Furthermore, defects and heteroatom doping are believed to have a higher reactivity for ROS generation (Navalon et al. 2017; Song et al. 2019; Wan et al. 2021b), which mainly occurs on the anthropogenic carbon after modification. The acidic environment could facilitate the electron transfer process with more ROS formation, and the interaction with redox-active ions would further enhance this process (Du et al. 2021); this will be discussed in a subsequent section. Recent studies have also found that the photocatalytic process could drive ROS formation with organic carbon, including humus (Zhang et al. 2020), biochar (Chen et al. 2017; Rangarajan et al. 2022), and DOM (Fu et al. 2016).

4.1.2 Reduction of PTEs with the ROS

The generated ROS, especially O2·- with a high reducing potential (E0 = − 0.33 V, Table 1), could reduce different PTEs based on their high reactivities. During this electron transfer process, the ROS serves as the reducing moiety, which then changes back to O2 or oxidative species. For instance, phenolics and PFRs on humin could trigger the generation of O2·- with O2, and the generated O2·- could reduce Cr(VI) to Cr(III) (Xu et al. 2020a). This electron transfer process contributed to almost 30% of the total Cr(VI) reduction by humin, while the remaining reduction was achieved by direct electron transfer (Xu et al. 2020a). In addition, the ROS generated by biochar might cause the reduction of Cr(VI), although their contribution could be limited compared to the direct reduction by reductive surface functionality and PFRs (Zhu et al. 2020). Several studies have reported that the generated ROS can directly react with PFRs on pyrogenic carbon (Zhao et al. 2018a; Zhu et al. 2020), resulting in the lower contribution of ROS to Cr(VI) reduction by pyrogenic carbon. A lower pH should be the premise for reducing Cr(VI) by ROS, while a higher pH might cause the re-oxidation of Cr(III) by the generated ROS (Liang et al. 2021). In addition to Cr(VI) reduction, other PTEs such as Ag(I) might also be reduced by O2·- generated by the organic carbon (Huang et al. 2019). Notably, the reduction of Cu(II) to Cu(I) by O2·- might have a lower possibility because the oxidation of Cu(I) by O2·- or O2 is dominant under oxic conditions, especially with organic carbon (Xing et al. 2020).

4.1.3 Oxidation of PTEs with the ROS

Compared to the reduction of PTEs in soil, ROS usually has a higher oxidation potential for PTEs. The oxidation of As(III) is a widely studied redox process in the soil; and different ROS generated from organic carbon, including hydrogen peroxide (H2O2) and hydroxyl radical (OH·), are responsible for this process (Xu et al. 2022a; Zhong et al. 2019, 2020). The formation of ROS from organic carbon through electron donation is considered to be the rate-limiting step for the entire electron transfer process; thus, an acidic environment and rich reductive moieties on the organic carbon could facilitate this redox process (Zhong et al. 2019). The Cr(III) oxidation by ROS has been reported, and this process could be enhanced by an alkaline environment and the formation of a Cr(III)-ligand complex (Liang et al. 2021; Luo and Chatterjee 2010; Oze et al. 2016). The oxidation of Cu(I) also occurs with the formation of ROS, such as O2·-, leading to the limited Cu(I) content in the oxic environment (Xing et al. 2020). Although Tl(I) oxidation can be achieved with species such as hydroxyl radicals (OH·) (Li et al. 2019; Xu et al. 2019a), the oxidation of Tl(I) with organic carbon and O2 has seldom been reported owing to the relatively lower concentration of Tl in soil (Belzile and Chen 2017) and the higher oxidation demand (E0 = 0.77 V, Table 1).

4.2 Redox-induced transformation of PTEs with Fe and organic carbon

4.2.1 Transformation of Fe minerals with organic carbon

Iron (Fe), the most abundant redox-active metal element in the soil, plays a vital role in the dynamics of toxic elements in the environment (Han et al. 2019; Xu et al. 2022b), and its speciation is greatly affected by organic carbon in the soil (Chen et al. 2021; Frankl et al. 2022). The direct redox reaction is widely found in soil because of electron transfer from the reductive moiety to the ferric mineral, resulting in the formation of Fe(II)-containing minerals, sorbed Fe(II), or Fe2+ ions (Fig. 3) (Cismasu et al. 2016; Kim et al. 2019; Klüpfel et al. 2014b; Li et al. 2021a; Xu et al. 2022b). This process also transforms the Fe mineral (Fig. 3) because the formed Fe(II) can serve as a reactive moiety to accelerate the transformation process with organic carbon (Boland et al. 2014; Chen et al. 2015; Sheng et al. 2020; ThomasArrigo et al. 2018). Relevant studies have suggested that organic carbon, such as dissolved organic ligands, can modify the relative rates of olation and oxolation reactions that assemble the labile Fe(III) into various product minerals. This leads to the formation of different iron minerals, including goethite, magnetite, and lepidocrocite, based on the DOM concentration (Sheng et al. 2020). In addition, the presence of organic carbon inhibits the oxidation of Fe(II) to Fe(III) as well as the crystallization of iron minerals, and more iron minerals might remain in the amorphous phase or low-crystallinity phase (i.e., ferrihydrite) with highly-disordered structure (Chen and Thompson 2021; ThomasArrigo et al. 2019). Co-precipitation of the iron mineral with organic carbon would also occur during the interaction; this causes a highly dynamic phase in the iron mineral, namely, the mixing of solid Fe with the aqueous Fe(II), even without secondary mineral formation (Zhou et al. 2018). The chelation effect of organic ligands causes the dissolution of iron minerals, resulting in a higher amount of soluble iron with potent redox reactivity (Saad et al. 2017; Xu et al. 2020b; Zeng et al. 2020). All these organic carbon-related transformations of iron minerals affect the redox-induced transformation of PTEs in the soil.

4.2.2 Reduction of PTEs by Fe mineral with organic carbon

The Fe(II) formed by organic carbon can serve as a direct reducing agent for PTEs, such as Cr(VI) (Liao et al. 2019b; Xu et al. 2020b). During this process, Fe served as the mediator to transfer the electrons from organic carbon to PTEs; namely, soluble Fe(III) could obtain electrons from biochar with the formation of Fe(II), which quickly reacts with Cr(VI) with the re-formation of Fe(III) (Xu et al. 2020b). In addition to the soluble Fe(II), the Fe(II) combined with clay mineral could serve as the reducing moiety for the Cr(VI) with organic carbon (Liu et al. 2019). However, some studies have reported that Fe(II) complexed with organic carbon might have lower reactivity for Cr(VI) reduction with a higher probability of reacting with O2 (Pan et al. 2017). This electron transfer process may also occur in the solid phase of iron minerals. For this process, organic carbon can combine with the iron mineral and reduce Fe(III) to Fe(II). The electron on Fe(II) can be transferred through the iron mineral by Fe(II)/Fe(III) transformation, which reduces the sorbed Cr(VI) afterward with the re-formation of Fe(III) (Liao et al. 2019b; Xu et al. 2022b). This process should be more vital in soil because of the prevalent combination of iron mineral-organic carbon-PTEs. Moreover, the low-crystallinity iron minerals might have a higher possibility than high-crystallinity iron minerals owing to their lower stability and higher reactivity (Xu et al. 2022b).

Compared to those on Cr(VI) reduction, studies on the reduction of other PTEs are relatively insufficient, which might be related to the low possibility of redox transfer with Fe(II). For instance, the abiotic reduction of As(V) by the Fe(II) with organic carbon can hardly be achieved in soil under normal temperature (Johnston et al. 2018), and it often requires support from organisms (Qiao et al. 2019).

4.2.3 Oxidation of PTEs through Fe mineral and organic carbon

The generated Fe(II) could serve as a reducing agent for the activation of O2 with ROS formation (i.e., Fenton-like reaction, Eqs. 8–10) (Merino et al. 2021; Usman et al. 2022), although the chelation of organic carbon with Fe might buffer this electron transfer process (Daugherty et al. 2017; Jones et al. 2020; Yu et al. 2021). This Fenton-like reaction could also happen with Fe sorbed on the clay mineral with organic carbon (Hong et al. 2019). Some studies have suggested that organic carbon could act as a mediator to expand the area of ROS production with Fe(II)-containing minerals, leading to the higher and faster production of ROS (Yu et al. 2021). Further comparison between the ROS generation efficiency of organic carbon with and without Fe minerals is required to evaluate the potential synergistic impact. The generated ROS can be the determining oxidative species for different PTEs, such as As(III), Cu(I), and Pu(IV) (Gubler and ThomasArrigo 2021; Pan et al. 2021; Xu et al. 2019a, 2022a), with a mechanism as previously mentioned. Transformation and crystallization of iron minerals with organic carbon also alter the direct electron-accepting capacity of iron (Aeppli et al. 2019a, b), and the direct oxidation by Fe(III) (e.g., As(III) oxidation by Fe(III)) could be achieved with the support of rich quinone functionality on organic carbon (An et al. 2022).

$${\mathrm{O}}_2+\mathrm{Fe}\left(\mathrm{II}\right)\to {{\mathrm{O}}_2}^{\cdotp -}+\mathrm{Fe}\left(\mathrm{II}\mathrm{I}\right)$$
(8)
$$2{\mathrm{H}}^{+}+{{\mathrm{O}}_2}^{\cdotp -}+\mathrm{Fe}\left(\mathrm{II}\right)\to {\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{Fe}\left(\mathrm{II}\mathrm{I}\right)$$
(9)
$${\mathrm{H}}^{+}+{\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{Fe}\left(\mathrm{II}\right)\to {\mathrm{H}}_2\mathrm{O}+\mathrm{O}{\mathrm{H}}^{\cdotp }+\mathrm{Fe}\left(\mathrm{II}\mathrm{I}\right)$$
(10)

4.3 Redox-induced transformation of PTEs by the redox cycle of Mn with organic carbon

4.3.1 Transformation of Mn minerals with organic carbon

Manganese (Mn) minerals are ubiquitous oxidants in the soil systems, and the redox reactions between organic carbon and Mn oxide occur extensively (Fig. 3) (Ma et al. 2020; Zhang et al. 2021). Manganese could be the dominant redox moiety in soil, especially in the surface layer, even when compared with iron minerals (Jones et al. 2020). Redox reactions between organic carbon and birnessite (MnO2) cause the reduction of Mn and the oxidation of organic matter (Ding et al. 2022b; Trainer et al. 2021). The DOM was more aliphatic and had a lower molecular weight after reacting with Mn, and the highly aromatic DOM underwent greater changes (Trainer et al. 2021). This redox process is generally accompanied by the dissolution of Mn oxide and the release of Mn(III)-ligand complex, which subsequently transforms into Mn(II) and oxidized organic compounds (Wang and Stone 2006). In addition to the reduction process, the oxidation of Mn(II) to Mn(III) occurs during the interaction with organic carbon (Wang et al. 2020d; Xu et al. 2021c). Organic carbon in the soil can serve as the oxidizing moiety, activating agent for ROS, and ligands to support Mn(III) formation, and the generated Mn(III) contains a solid oxidizing potential (Gao et al. 2018; Oldham et al. 2017). This Mn(III)-ligand complex has a higher possibility for internal electron transfer to transform back into aqueous Mn(II) and oxidized degradation products, especially in an acidic environment (Li et al. 2021b); this process might also contribute to ROS generation (Jones et al. 2020).

The multiple redox processes of Mn minerals with organic carbon are strongly related to the basic properties of organic carbon in the soil. Different surface functionalities might lead to the contrasting transformations of Mn minerals owing to the different redox reactivity. For instance, −OH and − NH2 functionalities could promote the electron transfer to MnO2, whereas some electron-withdrawing functionalities, such as −NO2 and − COOH groups, might inhibit the electron transfer (Zhao et al. 2020). Environmental conditions, such as the pH and O2 content, may affect the redox-induced transformation of Mn with organic carbon. For example, co-existing O2 may oxidize Mn(II) newly produced through reductive dissolution, especially with support from the mineral surface (Li et al. 2021b; Ma et al. 2020).

4.3.2 Redox-induced transformation of PTEs with Mn and organic carbon

Mn(IV) can effectively oxidize the reduced PTE species (e.g., As(III), Cr(III), Tl(I), and Sb(III)) in soil (Ao et al. 2022; Cruz-Hernández et al. 2019; Fu et al. 2018; Liang et al. 2021; Wick et al. 2020; Zhang et al. 2018a), while the Mn(III) and Mn(II) formed from the reduction of Mn(IV) usually contain lower oxidizing potential (Lafferty et al. 2010; Zhang et al. 2018a). In particular, reducing Mn(IV) together with organic carbon might alleviate the oxidative transformation of some PTEs. In contrast, some studies found that labile Mn(III) could be more active than Mn(IV) with the formation of a Mn(III)-organic ligand complex (Luther et al. 2018; Wang et al. 2020d). This could form ROS through the Mn(II)/Mn(III) cycle for the direct oxidation of PTEs. In addition, the reductive formation of Mn(III) is generally accompanied by the formation of vacancy-rich Mn oxides, which could be effective for the oxidative sorption of PTEs such as Tl(I) (Wick et al. 2019; Xu et al. 2019a). Mn(II) could also serve as an electron-mediating agent to facilitate the electron transfer between PTEs and organic carbon, such as the reduction of Cr(VI) by organic acids and biochar (Sarkar et al. 2013; Xu et al. 2021c). In addition, Fenton-like reaction caused by the activation of O2 by Mn(II) has been reported, which might also affect the fate of PTEs (e.g., Sb(III) oxidation) (Wang et al. 2022; Wu et al. 2022).

5 Implications to the soil PTEs remediation

Based on the above information, we conclude that organic carbon can significantly affect the speciation of PTEs through redox reactions, and this process is highly related to the redox moieties in soil, including O2, Fe, and Mn. Therefore, it is critical to consider these redox processes during the remediation of soil PTEs (Fig. 4).

Fig. 4
figure 4

Positive and negative roles of organic carbon during the PTEs remediation in soil

Full size image

5.1 Strategies for the utilization of organic carbon for PTEs immobilization

The addition of carbon-based materials has been widely studied to alter the fate of PTEs for lower toxicity and mobility. Introducing biochar into the soil is a common method to improve the carbon content in the soil, and it is believed to have great electron-donating capacities (Xu et al. 2022a, b). This supports the reduction of highly toxic Cr(VI) to less toxic Cr(III), which has a higher potential for immobilization in soil. The addition of catalysts to facilitate the electron transfer between organic carbon and PTEs can be a potential method to effectively utilize the redox reactivity of organic carbon (Fei et al. 2022). For instance, the graphitic structure can act as an electron conductor to facilitate the electron transfer from organic acids to Cr(VI), thus achieving the Cr(VI) reduction (Xu et al. 2019c). Redox-active minerals and surface functionalities can also participate in this electron transfer process through valence transformation (Xu et al. 2021c). Notably, this process might cause the oxidation of organic carbon, posing the potential risks to soil health and carbon emissions.

Although the inherent electron-accepting capacity of organic carbon is relatively low, it can support the ROS formation as a rich electron pool for the oxidation of several PTEs with the addition of oxidizing agent. For instance, oxidants, such as persulfate or O2, could be activated by electrons from organic carbon, leading to the oxidation of pollutants (Wu et al. 2020; Zhou et al. 2021). For example, the oxidation of As(III) to As(V) leads to lower toxicity and mobility, which are critical for As immobilization in soil. In addition, the oxidative transformation of Tl(I) to Tl(III) could result in rapid precipitation and sorption, leading to a higher immobilization potential (Biagioni et al. 2017; Chen et al. 2022). However, organic carbon might also inhibit ROS generation with the oxidant, which needs more consideration during practical applications.

Moreover, organic carbon also acts as an activating agent for mineral-based materials through the redox process. For example, the surface passivation of zero-valent-iron with the formation of Fe(III) is the rate-limiting step for redox-related immobilization; however, a suitable amount of organic carbon could lead to the reductive dissolution of Fe(III) to Fe(II), resulting in a higher removal rate of Cr(VI) (Lv et al. 2013; Xie et al. 2022). In addition, the reduction of Mn-based minerals may be critical for the oxidation-determined immobilization process (Wick et al. 2019). For instance, vacancy-containing birnessite (MnO2) with rich Mn(III) has higher oxidation and immobilization potentials for Tl(I) (Marafatto et al. 2021; Wick et al. 2019); therefore, organic carbon could potentially support the Mn(III) formation to enhance the Tl(I) immobilization. Reductive dissolution by organic carbon might also support the de-passivation of MnO2, leading to a higher oxidizing potential for PTEs (Lafferty et al. 2010). Notably, organic carbon has a dual impact on mineral materials for the immobilization of PTEs based on its concentration and properties (Di Palma et al. 2018; Xie et al. 2022), and this negative impact will be discussed in the following section.

5.2 Potential risks related to SOC during the PTEs immobilization

In addition to the direct redox reaction with PTEs, increasing the organic carbon content has been widely applied in other applications. According to its redox reactivity, organic carbon might cause unwanted transformation or mobilization of PTEs. For instance, adding organic carbon to the soil would increase the mobility of As (Gong et al. 2022), and its reducing potential might cause the reduction of As(V) to As(III), which is highly toxic (Alkurdi et al. 2019; Qin et al. 2020). The reductive formation of Tl(I), Sb(III), and Sn(II) with higher solubility was also reported owing to the broader Eh caused by the addition of biochar, increasing their mobility and potential toxicity (Rinklebe et al. 2019, 2020). In addition, rich organic carbon causes the reductive dissolution of Fe and Mn minerals with the formation of Fe(II) and Mn(III). As mentioned before, the formation of Fe(II) and Mn(III) is highly related to mineral transformation and the generation of ROS, leading to the potential oxidative formation of Cr(VI) with high toxicity.

Organic carbon in the soil requires more consideration when applying metal-based materials (e.g., zero-valent-iron and MnO2) to the soil for PTEs remediation. Owing to the redox-related transformations between organic carbon and Fe/Mn minerals, the organic carbon content significantly affects the effectiveness of the metal-based materials. The interactions between organic carbon and iron-based materials might cause the dissolution of Fe(II) into the soil and thus facilitate the redox-induced transformation of PTEs, which can be a potential risk. For instance, adding zero-valent-iron materials could effectively reduce and immobilize Cr(VI) through reduction and co-precipitation, whereas a high organic carbon concentration might cause the re-oxidation of Cr(III) by Fe (Di Palma et al. 2018, Xie et al. 2022), possibly owing to the formation of Fe(II) and ROS (Liang et al. 2021; Usman et al. 2022). This process could cause the potential remobilization of Cr(VI), and its contribution has been largely ignored compared with the competitive impact of organic carbon (Xie et al. 2022). Similar interactions might also occur with Mn minerals, and several studies have reported the decrease in the oxidation potential with the reductive dissolution of Mn minerals (Lafferty et al. 2010, Zhang et al. 2018a).

6 Future perspectives

In this review, we scrutinized the composition of SOC, primary redox-active moieties on organic carbon, and their impact on the redox-induced transformation of PTEs. The phenolic and quinone functionalities, PFRs, and graphitic structure could be the critical redox-active moieties of organic carbon in the soil, and they could directly affect the redox-induced transformation of Cr, As, Cu, Hg, etc., in the environment. In addition, the interactions between organic carbon and O2, Fe, or Mn minerals lead to the formation of ROS, the reductive formation of Fe(II) and Fe(II)-organic ligands with the catalytic transformation of iron minerals, and the transformation of Mn(IV)/Mn(III)/Mn(II) with Mn-containing minerals. These processes significantly affect the fate of PTEs through either oxidation or reduction reactions. It is critical to consider both the positive and negative impacts of organic carbon on the transformation of PTEs during soil remediation. Utilizing organic carbon as the redox reagent or activator can be a promising method for the transformation and immobilization of PTEs, but the potential risk caused by organic carbon also needs further evaluation. Such fundamental knowledge of PTEs transformation with organic carbon can help us better understand the bio-geochemistry of PTEs and guide the remediation of PTEs in soil.

Several issues regarding organic carbon and its impact on the redox-induced transformation of PTEs need to be addressed in future studies. First, organic carbon contains both reducing and oxidizing moieties, and their capacities are strongly related to the chemical structure and surface properties (Li et al. 2020). Although several studies have revealed the structure-function relationship of the redox reactivity of specific organic carbon, the overall concern of all types of SOC and their synergistic impact on redox-related performance is still required. Moreover, a more detailed consideration of both micro (i.e., molecular and atomic structures) and macro aspects (i.e., different properties related to organic carbon evolution and peculiarity) is critical for evaluating the redox potential of organic carbon and its impact on the redox-induced transformation of PTEs. A theory-driven Pourbaix Diagram (Eh/pH diagram) of PTEs with varying SOC containing different moieties is also suggested in future studies.

Second, the addition of carbon into the soil has been widely conducted as a potential method for fertility enhancement, environmental remediation, and carbon sequestration, leading to the increased organic carbon content in the soil. Notably, this newly added carbon could have different properties compared to the indigenous organic carbon owing to its formation process and aging impact (LaCroix et al. 2021; Wang et al. 2020c). Thus, it is vital to consider the unintended consequence and potential risk of adding carbon to the soil in future studies, including both the direct mobilization of PTEs and the indirect transformation with the soil moieties.

Third, current studies have provided detailed explanation for the potential redox interactions between organic carbon and several primary soil moieties, respectively, including O2, Fe-minerals, and Mn-minerals. An integrated consideration of these soil moieties with organic carbon and their impact on the redox-induced transformation of PTEs under temporally and spatially variable geochemical conditions could be a promising research area in future studies, which is much more relevant to natural soil conditions. In addition, potential contribution of other redox moieties (e.g., sulphur (Wang et al. 2021a, Zhao et al. 2019) and chlorine (Hou et al. 2018; Wang et al. 2020b)) and electron transfer between different PTEs (e.g., As and Cr (Dong et al. 2014)) need further consideration.

Fourth, this review focused on the abiotic transformation processes of organic carbon with soil moieties and PTEs. In field conditions, biotic processes involving organic carbon could also be critical for determining the fate of PTEs, such as the redox-related transformation of As and Tl (An et al. 2022; Wei et al. 2020, 2021). The consideration of both abiotic/biotic processes and their synergistic interplay could be an important research area in future studies.