1 Introduction

Dissolved organic matter (DOM) is one of the key variables that drive soil and aquatic ecologies. The change of DOM in quantity and composition can directly affect carbon cycle, pollutant behavior, and microbial activity (Lapierre et al. 2013; Lou et al. 2018; Stegen et al. 2018). Recently, the change of DOM due to the frequent wildfires in forest watersheds was reported to pose a threat to water quality and affect drinking water treatment (Hohner et al. 2016, 2019). The changes of DOM in quantity and composition could affect the water treatment process and increase the challenge of coagulation and filtration. The further formed disinfection byproducts that are regulated in drinking water may pose threat to human health. Pyrogenic carbon (e.g., biochar), pyrolyzed from biomass during the wildfires and production, has accumulated or applied in the soil and surrounding ecosystem, which gained broad interest due to its significant effect on carbon sequential and soil amendment (Rudolf et al. 2013; Bistarelli et al. 2021; Zhang et al. 2021). The pyrogenic carbon in the soil and aquatic system may affect the DOM pool. Since biochar is easy to produce and apply, biochar is often used to represent pyrogenic carbon in laboratory and field research. Biochar may change the characteristics of soil DOM. For example, in a batch experiment, pyrogenic carbon increased the release of DOM from the soil due to the increased pH, and the composition of the released DOM correlated with the size exclusion effect of pyrogenic carbon pore structure (Smebye et al. 2016). Similar results were confirmed in the agricultural field experiment (Zhang et al. 2017; Liu et al. 2019), and the main DOM components released from the soil were fulvic acid-like and humic acid-like DOM. Conversely, it was reported that biochar decreased DOM leaching in a column experiment by stabilizing the complex compounds already present in the soil (Eykelbosh et al. 2015). Another 5-year field experiment showed that biochar had little effect on the soil DOM pool (Dong et al. 2019). The differences in these studies reveal that the effect of biochar on soil DOM is complex, which may be related to biochar and soil characteristics.

Biochar in soils and aquatic systems can be broken into small particles with micro- and nano-size (Spokas et al. 2014; Sigmund et al. 2018). Biochar colloids show strong mobility and are prone to vertically transport into groundwater under irrigation or precipitation (Georg et al. 2008; Yang et al. 2019). Moreover, these biochar colloids have a large surface area and abundant functional groups and show strong interaction with DOM components. It was reported that humic acid, proteins, and low molecular weight organic acids could complex with biochar (Xu et al. 2019; Yang et al. 2019). The presence of various DOM poses different effects on colloid transport through competing retention sites, masking charge heterogeneity, inducing stronger/weaker electrostatic repulsion, or enhancing surface roughness (Wang et al. 2015; Wu et al. 2018; Yang et al. 2019). Humic acid was reported to increase the transport of biochar colloids due to the enhanced electrostatic repulsion and weaker aggregation between biochar colloids. And the positively charged protein (cytochrome c) decreased the transport of biochar colloids and negatively charged protein (bull serum albumin) increased biochar colloid transport (Yang et al. 2019). Additionally, during the vertical transport, biochar colloids may act as vehicle for DOM transport and further cause an unknown impact on nearby aqueous systems. With more biochar accumulating in soils, it is anticipated that the transport of biochar colloids could significantly affect the release of soil DOM, but little attention has been paid to the release and composition change of soil DOM during the vertical transport of biochar colloids. In addition, biochar colloids suffer environmental aging processes (e.g., oxidation) continuously in the environment (Ward et al. 2014) and the surface properties of biochar colloids are modified. The changed properties (e.g., functional groups and porous structures) of aged biochar colloids may affect their interaction with DOM and further influence the cotransport of biochar colloids with soil DOM. To the best of our knowledge, little research focuses on the effect of biochar colloid transport on soil DOM leaching through combined optical EEM and molecular FT-ICR-MS characterizations.

In this study, softwood biomass was pyrolyzed as pristine biochar because wood-biochar commonly exists in the environment. The chemical aging process was chosen as a typical biochar aging process because it commonly exists in the natural environment. As one of the most widely used methods for simulating natural biochar chemical aging processes (Li et al. 2019; Wang et al. 2020), hydrogen peroxide (H2O2) was used to treat pristine biochar to obtain aged biochar. Three typical soil types were selected: black soil (BS), fluvisol (FS), and paddy soil (PS). The effect of biochar on the composition and concentration of DOM released from three soils was analyzed by both adsorption and transport experiments. The composition and molecular properties of leached DOM induced by pristine and aged biochar colloid transport were analyzed by EEM-PARAFAC and FT-ICR-MS. The objectives of this study were to (1) identify the role of biochar colloid transport in the soil DOM leaching and (2) evaluate the impact and mechanism of pristine and aged biochar colloids on the composition and properties of DOM leached from different soils.

2 Materials and methods

2.1 Biochar colloids and soils

Pristine biochar was prepared by pyrolyzing at 550 °C with a softwood pellet. To simulate the aging process of biochar, aged biochar was obtained by the chemical aging method with hydrogen peroxide (H2O2). Briefly, 5% H2O2 was added to pristine biochar and heated at 80 °C for 2 days until the solution dried out (Leng et al. 2019). The biochar sample was grounded into powder form by a ball mill (MM400, Verder Shanghai Instrument, Shanghai, China). The pristine and aged biochar powder was added to distilled (DI) water and sonicated for 30 min in a water bath. The suspension was settled for 24 h and the upper suspension was collected as pristine and aged biochar colloidal suspension (Zhao et al. 2021). The zeta potential and hydrodynamic diameter of biochar colloids were measured using a Zetasizer (Nano ZS90, Malvern, UK). The specific surface area of biochar was characterized by the Brunauer-Emmett-Teller (BET) method (Micromeritics ASAP 2460, USA). Static contact angles of biochar were measured by goniometer (Dataphysics OCA20, Germany) at room temperature. The characterizations of pristine and aged biochar colloids are in Table S1.

Three soils, a black soil (BS, loam soil) from Jilin, China (43°17′ N, 124°35′ E), a fluvisol (FS, sandy clay loam soil) from Beijing, China (40°06′ N, 116°13′ E), and a paddy soil (PS, clay loam soil) from Hunan, China (28°13′ N, 113°32′ E), were used in this study (Table S2). The soils were obtained from every field with a depth of 0–20 cm and were air-dried, sieved through a 2 mm mesh, and mixed thoroughly. The dissolved organic carbon (DOC) contents of three soils were 1.80% for BS, 1.44% for FS, and 0.96% for PS.

2.2 Column experiments

Soil column experiments were performed in an acrylic column (length 15 cm, inner diameter 5 cm) (Fig. S1). Glass beads of 1 cm thickness were placed at the bottom of the column and 2 layers of steel and nylon membranes were covered on glass bead to support the upper soil. The column was packed with three different soils (BS, FS, and PS) in an increment of 1 cm thick layer and 2 layers of steel and nylon membranes were covered on soil. After packing, the pore volume (PV) was calculated by subtracting the volume of soil from the volume of the column. Solution or biochar colloids suspension was injected into the column at the inlet using a peristaltic pump at a velocity of 0.3 mL min−1 in a down-flow direction. The soil columns were flushed with deionized (DI) water, then flushed with the background solution (1 mM NaCl) for more than 10 PVs to equilibrate the system. For the biochar colloid transport experiment, 3 PVs of 100 mg L−1 biochar colloids were injected into the soil column, followed by flushing with more than 3 PVs of background solution. Effluents were collected continuously with a fraction collector (CBS-A, Shanghai Huxi, China) and the concentration of biochar colloids was measured using a UV-vis spectrophotometer at 790 nm (TU-1900, Persee, China).

To investigate the effect of biochar colloid transport on HA release, sand column experiments were further conducted. Quartz sand was washed to obtain clean sand and iron oxides were coated on the clean sand to prepare iron oxide-coated sand (Fe sand) (Zhao et al. 2023). The clean or Fe sands were packed in the column and the water flow was maintained at 0.3 mL min−1. The column was firstly flushed with DI water and 1 mM NaCl solution, and then 10 mg L−1 humic acid (HA) solution was injected in the sand column, followed by 100 mg L−1 biochar colloids suspension. DOC content in the effluent was measured and the sand in the column was excavated after the transport experiment for SEM characterization.

2.3 Adsorption experiment

An adsorption experiment was performed to examine the effect of biochar colloids on the release of DOM from different soils. The pristine or aged biochar colloids suspension (sonicated for 30 min before use) was mixed with black soil, fluvisol, or paddy soil in a 15 mL vial. The vial was shaken on a shaker (150 rpm) for 24 h, and the mixture was centrifuged at 4000 rpm for 10 min and filtered through a 0.45 μm membrane to separate the dissolved fraction from the mixture. The filtrate was collected for fluorescence characterization (see below).

2.4 DOM characterization

Three-dimension excitation/emission matrix (EEM) fluorescence spectra of DOM were measured using a fluorescence spectrophotometer (LS-55, PerkinElmer). The excitation and emission wavelength ranges were 200–450 nm and 270–550 nm at 1 nm increments. A blank EEM spectrum of DI water was used to remove background noise, and Rayleigh and Raman peaks were removed from the EEM spectra. The detailed characterized fluorescence peaks identification and EEM-PARAFAC modeling process were in SI. The DOM in the leachate from the soil column was analyzed by Bruker Solarix FT-ICR-MS system with a 15.0 T superconducting magnet and an ESI ion source. The raw spectra of each sample were converted to m/z value and the molecular formulas with 0.2 ≤ H/C ≤ 2.3 and 0 ≤ O/C ≤ 1.2 were retained. The van Krevelen diagram was plotted using oxygen/carbon (O/C) and hydrogen/carbon (H/C) element ratios. The detection methods and DOM identifications were in SI.

The fluorescence intensities of each component in the EEM spectra were calculated by normalizing the regional fluorescence intensity of C1, C2, or C3 to the total fluorescence intensity. The relative intensity of FT-ICR-MS molecule was defined as the relative abundance of each formula. The component fluorescence intensities in the PARAFAC and the relative intensities of FT-ICR-MS were converted to ranks within a sample (Stubbins et al. 2014; Martinez-Perez et al. 2017; Hu et al. 2023). Since the relative intensities of the FT-ICR-MS peaks in several samples were not normally distributed, Person’s correlation could not be applied. The Spearman’s correlation was performed between PARAFAC and FT-ICR-MS data (in Rstudio). For an n of 9 samples, an average absolute-value Spearman’s rank correlation coefficient (r) was calculated to be significant at the 95% confidence limit by a Student’s t-test.

3 Results and discussion

3.1 Characterization of biochar colloids

The hydrodynamic diameters of pristine and aged biochar colloids were 453.9 ± 10.4 nm and 544.8 ± 20.1 nm, respectively (Table S1). Aged biochar colloids showed a more negative zeta potential (-40.6 ± 0.8 mV) compared with pristine biochar colloids (-35.7 ± 1.6 mV). The FTIR analysis of pristine and aged biochar colloids in our previous study (Zhao et al. 2023) showed that aging increased the oxygen-contained functional groups on the surface of aged biochar colloids, making their surface more negative. The specific surface area (SSA) of biochar colloids was increased from 38.86 to 314.11 m2 g−1 after aging. The EEM spectra of DOM in pristine and aged biochar colloids are shown in Fig. S2. DOM was hardly detected for pristine biochar colloids as there were no peaks in EEM spectra. After aging, tryptophan protein-like (peak II) and fulvic acid-like (peak III) substances appeared in the EEM, due to the oxidation and fragmentation of biochar during the aging process.

3.2 Transport of biochar colloids

Figure 1 shows the breakthrough curves and mass recovery rates in the effluent of pristine and aged biochar colloids in black soil, fluvisol, and paddy soil at pH 5, 7, and 9. In black soil and fluvisol, the transport of pristine and aged biochar colloids was enhanced with the increase in pH. In black soil, the transport of aged biochar colloids increased from 8.8% to 12.4% and 17.2% as increasing pH from 5 to 7 and 9 (Table S3). The increased transport of pristine and aged biochar colloids at high pH was attributed to enhanced electrostatic repulsion between biochar colloids and soils. The transport of aged biochar colloids in black soil and fluvisol was stronger than that of pristine biochar colloids at all pH levels. The transport of aged (8.8%, 11.1%, and 17.0% at pH 5, 7, and 9) biochar colloids in BS soil were higher than pristine (7.0%, 9.3%, and 10.2% at pH 5, 7, and 9) biochar colloids. This increase was attributed to the enhanced negative charge on aged biochar colloids resulting from an increase in oxygen-containing functional groups (Zhao and Shang 2023).

Fig. 1
figure 1

Breakthrough curves (a-c) and mass recovery rates (d) of pristine and aged biochar colloids at pH 5, 7, and 9 in black soil (BS), fluvisol (FS), and paddy soil (PS). C/C0 represents the relative concentration of biochar colloids; Meff represents the mass recovery rates of pristine and aged biochar colloids in the effluent

Full size image

Under all pH conditions, the transport of both pristine and aged biochar colloids was strongest in black soil and weakest in paddy soil. In the black soil, abundant DOM may increase the transport of biochar colloids. Yang et al. (2019) reported that DOM (e.g., humic acid and protein) could enhance the transport of biochar colloids through increased electrostatic repulsion and steric force. In paddy soil, the transport of both pristine and aged biochar colloids was weak (below 7.0%) and exhibited little variation as pH increased from 5 to 9 (Table S3). The high content of iron oxides in paddy soil increases soil grain roughness and decreases negative charge, potentially leading to a reduction in the transport of biochar colloids (Wang et al. 2013).

3.3 Release of DOM during biochar colloid transport

3.3.1 Enhanced release of DOM

The EEM spectra of DOM in the supernatant resulting from adsorption experiment, involving the mixing of soils with biochar colloids at pH 7, are shown in Fig. S3. The fluorescence peaks of all soils were mainly distributed at peak II (tryptophan protein-like substance), peak III (fulvic acid-like substance), and peak V (humic acid-like substance) without pristine and aged biochar colloids adsorption. The relative intensities of the three peaks constituted more than 80% of the total fluorescent intensity, representing commonly existing substances in soil DOM (Liu et al. 2009; Sun et al. 2022). In adsorption experiments for pristine and aged biochar colloids, the primary fluorescence peaks of the different soils were still tryptophan protein-like, fulvic acid-like, and humic acid-like substances. Furthermore, the relative intensity of the different peaks in the EEM spectra of the three soils showed little change, indicating that both pristine and aged biochar colloids had little effect on the release of DOM.

The impact of biochar colloid transport on DOM release from soils was assessed through soil column experiment at pH 7. The EEM spectra of DOM in the leachate are shown in Fig. 2. In transport experiments of pristine and aged biochar colloids, the fluorescence peaks of all soils were mainly distributed at peak II (tryptophan protein-like substance), peak III (fulvic acid-like substance), and peak V (humic acid-like substance). The relative intensity of humic acid-like substance in the black soil and fluvisol increased from 0.244 and 0.183 to 0.353 and 0.214, respectively, indicating that the transport of biochar colloids increased the release of humic acid-like substance. However, for paddy soil, the transport of pristine and aged biochar colloids had no effect on the fluorescence properties of the leachate.

Fig. 2
figure 2

The EEM spectra and relative intensity of fluorescent components in the effluent after pristine or aged biochar colloid (BC) transport at pH 7 in black soil (BS), fluvisol (FS), and paddy soil (PS). I-V: tyrosine protein-like, tryptophan protein-like, fulvic acid-like, soluble microbial by-product-like, humic acid-like substances

Full size image

The enhanced leaching of the humic acid-like substances in the column experiment was inconsistent with the results of the adsorption experiment, which showed little effect on soil DOM release. In pristine and aged biochar colloids themselves, almost no humic acid-like substance was detected in the EEM spectra (Fig. S2), revealing that the increased leaching of humic acid-like substance was from soils rather than biochar colloids. The discrepancy between the adsorption and transport experiments may be due to the difference in hydrodynamic conditions, surface roughness, and pore structure (Chrysikopoulos and Aravantinou 2012; Treumann et al. 2014). In the adsorption experiment, the solid phases were well mixed due to the continuous shaking, which might have facilitated the colloid collision with soil media. In the column experiment, the water flow is single direction. The colloid immobilization and rolling were determined by the balance of hydrodynamic torque (TH) and resisting adhesive torque (TA) (Bradford et al. 2011). The soil surface had a high degree of nano- to micro-roughness, which increased the lever arm for TA (Shen et al. 2010; Bradford et al. 2013) and enhanced the retention of biochar colloids. In addition, the pore structure (e.g., grain-grain contacts) increased the retention of biochar colloids in the column experiment through a straining mechanism (Bradford et al. 2013), which did not exist in the adsorption experiment due to the destruction of soil pore structures. Compared with the adsorption experiment, the biochar colloids in the soil column experiment might have increased their contact with DOM through the surface roughness and hydrodynamic forces, enhancing the release of DOM from soils. The discrepancy between the adsorption and transport experiments indicates that the prediction of biochar impact on soil DOM release based on adsorption results and lab incubation experiment may be inadequate under field conditions since biochar colloids always transport with precipitation and irrigation.

3.3.2 Mechanisms of enhanced DOM release

The release of soil DOM during biochar colloid transport at pH 5 and 9 was also conducted (Figs. S4 and S5). Similar to the result at pH 7, the relative intensity of the humic acid-like substance (peak V) in the black soil and fluvisol increased, revealing the increased release of the humic acid-like substance. During pristine and aged biochar colloid transport, the relative intensity of the humic acid-like substance in paddy soil showed little change. The results showed that the transport of biochar colloids enhanced the release of soil DOM in black soil and fluvisol under all pH conditions. The possible change in pH due to the injection of biochar colloids into the soil column was not the determining factor influencing the enhanced release of soil DOM.

The relatively stronger mobility of pristine and aged biochar colloids in the black soil and fluvisol compared to the paddy soil was consistent with the higher release of DOM in the black soil and fluvisol compared to the paddy soil. We speculate that an exchange between biochar colloids and soil DOM occurred during the downward transport of biochar colloids in the column experiment. According to a soil DOM cycling downward model, the surface-near soil DOM could reach deep soil through temporal immobilization (by adsorption or co-precipitation), followed by microbial processing, and re-release (by desorption or dissolution) into soil water of altered compounds (Kaiser and Kalbitz 2012). The concept model suggests that the surface-near DOM compounds are exchanged for older DOM compounds transported into deeper soil layers. The exchange is related to the physicochemical displacement of previously sorbed DOM by new DOM. In this study, the duration of column experiment was short, and the velocity was relatively high. Thus, microbial activity may have little effect on the degradation and release of DOM in such a short time (Frberg et al. 2007; Tipping et al. 2011). For the black soil and fluvisol, the active surface of biochar colloids might have exchanged the humic acid-like DOM through physicochemical processes, after which the exchanged DOM was leached out from the soils.

Sand column experiments were conducted to verify the mechanisms of enhanced release of humic acid-like substance during biochar colloid transport. Compared with the sand column with only HA transport, the DOC of the sand column with HA transport followed by pristine biochars injection was increased from 3.71 mg L−1 to 5.31 mg L−1 (Fig. 3a). The result indicated that more HA was released during biochar colloid transport. In addition, the SEM images also showed that after biochar colloid transport (Fig. 3c), less HA was retained on the sand surface than that without biochar colloids (Fig. 3b) transport, revealing the enhanced release of HA. The paddy soil had a high content of Fe element (Table S2), with a high isoelectric point (Wang et al. 2012), making the paddy soil surface less negative than the black soil and fluvisol. Due to the relatively weaker electrostatic repulsion between biochar colloids and paddy soil, the biochar colloids showed weak transport in the paddy soil column and exhibited little effect on DOM release. In addition, the large amount of Fe oxides in the paddy soil has a strong precipitation and immobilization ability for soil DOM, which might have also reduced the release of DOM. The sand column experiment also showed that Fe oxides decreased the release of HA. The DOC content in the Fe sand column was lower than that in the clean sand column (5.31 mg L−1 < 2.35 mg L−1) (Fig. 3a), which indicated that iron oxides decreased the release of HA. The SEM image also showed that much HA was retained on the sand surface in the column with Fe oxides (Fig. 3d). The less negative and rougher sand surface with Fe oxides than that without iron oxides decreased the release of HA.

Fig. 3
figure 3

The dissolved organic carbon (DOC) in the effluent (a) and the SEM images of quartz sand (b-d) in sand column experiments at pH 7. A: 10 mg L−1 humic acid (HA) in clean sand; B: 100 mg L−1 biochar colloids in clean sand; C: 10 mg L−1 HA followed by 100 mg L−1 pristine biochar colloids in clean sand; D: 10 mg L−1 HA followed by 100 mg L−1 pristine biochar colloids in iron oxides coated sand (Fe sand)

Full size image

3.3.3 Molecular characterization of released DOM

The FT-ICR-MS can provide detailed structural information on every molecule and the dynamics of the DOM released from soils (Sun et al. 2020; Liu et al. 2021; Hu et al. 2023). In biochar colloid transport experiments, relatively higher O/C, DBE, AImod, and NOSC values, and lower H/C values were observed in all soil leachate (Table 1), indicating that the transport of biochar colloids increased the unsaturation, aromaticity, and oxidation of the leached DOM. All formulas of the molecules mainly contained four subcategories based on their elemental composition: CHO, CHON, CHOS, and CHONS. CHO (relative abundance of 53%–64%) and CHON (relative abundance of 22%–38%) were the dominant formulas in the soil leachate (Fig. 4a and Table S4).

Table 1 Element (C, H, O, N, S) mass fraction, H/C, O/C, AImod, DBE, NOSC derived from assigned molecule of FT-ICR-MS from soil column leachate
Full size table
Fig. 4
figure 4

Relative abundance of formulas element classes (a) and compound classes (b) derived from van Krevelen diagram of DOM from soil column leachate. BS, FS, and PS represent black soil, fluvisol, and paddy soil, respectively

Full size image

The van-Krevelen diagram was used to present a graphic plot of compound distribution (Fig. S6), which revealed the changes in DOM molecules caused by the transport of biochar colloids. Lipid-like, protein-like, carbohydrate-like, lignin-like, tannin-like, condensed aromatic-like, and unsaturated hydrocarbon-like components were detected in the van-Krevelen diagram. The lignin-like component had the highest relative abundance in all soil leachate probably because of the degradation of terrestrial-plants sources (Fig. 4b; Ohno et al. 2010; Liu et al. 2021). The transport of pristine and aged biochar colloids increased the relative abundance of condensed aromatic-like and tannin-like compounds in the leachate from the black soil and fluvisol. The relative abundance of condensed aromatic-like compounds in the BS soil leachate increased from 9.44% to 15.24% and 15.69% with the transport of pristine and aged biochar colloids, respectively (Fig. 4b and Table S5). In the paddy soil, the relative abundance of the seven compounds showed minimal change, consistent with the slight change shown by the EEM spectra. Nevertheless, the relative abundance of condensed aromatic-like and tannin-like compounds in the paddy soil leachate slightly increased from 3.31% to 4.43% and 5.71%, and from 5.51% to 6.04% and 8.48% for pristine and aged biochar colloids, respectively. These results revealed that the transport of both pristine and aged biochar colloids enhanced the leaching of condensed aromatic-like and tannin-like compounds in all soils.

The condensed aromatic-like and tannin-like compounds could originate from biochar colloids and soils. To elucidate the sources of the increased DOM compounds in the leachate, we compared the number and types of molecules before and after biochar colloid transport in the three soils. Compared with the column experiments without biochar colloid transport, the molecule number of DOM in the leachate from pristine and aged biochar colloid transport experiments showed 1898, 1602, 2561, 1315, 1888, and 1768 newly emerged molecules, 1929, 1705, 2583, 2031, 1811, and 1472 vanished molecules, and 6724, 5923, 6336, 6026, 6369, and 7105 resistant molecules (Fig. S7). These molecules plotted in van Krevelen diagrams are shown in Fig. 5a-f. Since the molecules in pristine and aged biochar colloids were removed during the selection processes, these “New/Increased” and “Vanished/Decreased” molecules were from soils, not biochar colloids.

Fig. 5
figure 5

The van Krevelen diagrams (a-f), the number of “Vanished/Decreased” (g) and “New/Increased” (h) molecule in soil column leachate after biochar colloid transport. “New”, “Increased”, “Decreased”, and “Vanished” represent the newly emerged, intensity increased, intensity decreased, and vanished molecule after biochar colloid transport; BS, FS, and PS represent black soil, fluvisol, and paddy soil, respectively

Full size image

The “New/Increased” molecules with high O/C and low H/C ratios and “Vanished/Decreased” molecules with low O/C and high H/C ratios revealed the increased release of molecules with large molecular weight. The molecules released from soils with low H/C ratio suggested their low lability (Liu et al. 2009; Gu et al. 2019), which could change the soil organic carbon pool and alter the microbial communities since carbon lability affects microbial metabolic process. In addition, the results revealed that the transport of biochar colloids increased the release of soil DOM molecules with large molecular weight. This was different from the results in the previous studies, which reported that molecules with low molecular weights were easily leached out during precipitation and irrigation (Sun et al. 2022; Hu et al. 2023). This might be because the biochar colloids change the soil DOM toward molecules with larger sizes by sorbing smaller aliphatic species (Smebye et al. 2016). The elemental composition and compound distribution showed 191 “Vanished/Decreased” molecules (Fig. 5g) and 520 “New/Increased” molecules (Fig. 5h) in all soils after biochar colloid transport. CHO and CHON were the main formulas in “New/Increased” and “Vanished/Decreased” molecules. The lignin-like compound was the most abundant in “New/Increased” and “Vanished/Decreased” molecules. Condensed aromatic-like and tannin-like compounds had more “New/Increased” molecules with few “Vanished/Decreased” molecules, indicating that they were the most released DOM compounds during biochar colloid transport.

The optical characteristics of DOM in the EEM spectra were related to its molecular composition in the FT-ICR-MS. The EEM-PARAFAC model decomposed the EEM spectra into three individual characterized fluorescence components of DOM: C1 (at Ex/Em of 250 nm/ < 450 nm), C2 (at Ex/Em of < 250 nm/ < 450 nm), and C3 (at Ex/Em of < 270 nm/480 nm). C1 and C2 were microbial humic acid-like species with anthropogenic origin usually found in agriculturally dominated watersheds, while C3 was a typical terrestrial humic acid-like component primarily found in soil-derived DOM (Murphy et al. 2014; Peleato et al. 2016). The correlations between the three components (C1, C2, and C3) and the relative abundance of “New & Increased” and “Vanished & Decreased” formulas are presented in Fig. S10. There were 491, 281, and 421 “New & Increased” formulas significantly correlated to C1, C2, and C3 components, of which 64.6%, 6.1%, and 92.9% were positively correlated to C1, C2, and C3 components, respectively. This indicated that these formulas were mainly positively correlated with C1 and C3 components. For the C1 and C3 components, the positively correlated formulas were mainly distributed in the regions with high O/C and low H/C ratios and consisted of lignin-like (57.0% and 60.1%), condensed aromatic-like (22.8% and 19.4%), and tannin-like (18.4% and 18.9%) compounds. There were 297, 251, and 191 “Vanished & Decreased” formulas significantly correlated to C1, C2, and C3 components, of which 44.1%, 96.0%, and 2.1% were positively correlated to C1, C2, and C3 components, respectively. This showed that the formulas were mainly positively correlated with C1 and C2 components. For C1 and C2 components, the positively correlated formulas were mainly distributed in the regions with low O/C and high H/C ratios and contained lignin-like compounds (79.4% and 77.2%). The strong correlation between the optical characteristics and the molecular composition of DOM indicated the increased leaching of humic acid-like substances that mainly consisted of lignin-like, condensed aromatic-like, and tannin-like compounds. The decreased leaching of humic acid-like substance showed that the main component was a lignin-like compound.

3.4 Environmental implications

The increased release and downward movement of DOM may affect the soil carbon pool, microbial communities, and water quality, which is tightly related to composition and structure of DOM (Kellerman et al. 2015; Roth et al. 2019). Generally, the DOM enriched in components with a high H/C ratio (e.g., proteins and sugars) is considered more labile, whereas DOM composed of high O/C ratio structures (e.g., aromatic-like) is less labile (Sun et al. 2020). The most labile fractions of DOM are usually consumed firstly by the microbes, followed by less labile fractions. Our study showed that the DOM components with higher aromaticity were released during the transport of biochar colloids (Fig. 6), and these released DOM could not be metabolized by microbe or further change the microbial community in short term. However, the long-term effect of DOM with higher aromaticity should not be neglected. For example, an incubation experiment showed that after the initial faster consumption of labile DOM, the aromatic DOM tends to be more selective for the microbial community (Zhou et al. 2021). In addition, the degradation processes of DOM in lakes preferentially remove aromatic compounds rather than reduced and aliphatic compounds (Kellerman et al. 2015), and the light penetrates water preferentially degrading aromatic compounds (Stubbins et al. 2010). Thus, the different degradation process in aquatic systems compared with soils indicates that more attention should be paid to the watershed and water quality in regions with biochar application and frequent wildfires. Additionally, the high aromaticity and strong electron transfer ability of released DOM could change the transport and degradation of organic pollutants in aquatic systems (e.g., antibiotics) (Lou et al. 2018), and the preferential adsorption of aromatic compounds on mineral surfaces (Coward et al. 2019) may further affect pollutant behaviors and carbon sequestration in soils.

Fig. 6
figure 6

Schematics of the enhanced soil DOM release by biochar colloid transport

Full size image

4 Conclusion

This study showed the optical and molecular evidence highlighting the vital role of biochar colloid transport in soil DOM release. Aging process alters the properties of biochar colloids and slightly enhances their transport in black soil and fluvisol. The significant DOM release in three soils differed slightly in response to pristine and aged biochar colloid transport, revealing the critical role of both pristine and aged biochar colloids in soil DOM release. The minor effect of chemical aging on soil DOM release enhanced by biochar colloids revealed that chemical aging was not a detrimental factor contributing to the increased soil DOM release. Various aging processes may coexist in the natural environment, and further research is needed to investigate their effect on soil DOM release. Our study could provide a reference for evaluating the environmental aging process on the soil DOM release enhanced by biochar colloids during rainfall and irrigation. The different results between the adsorption and transport experiments indicated that the prediction of DOM release based on adsorption experiments might be inadequate in real fields where precipitation and irrigation frequently occur. The effect of biochar application on soil DOM release may be underestimated in areas with high rainfall and irrigation since soil DOM release could be enhanced during biochar colloids downward movement. The increased release of humic acid-like substance had large molecular weights and contained lignin-like, condensed aromatic-like, and tannin-like compounds, which had important implications for the soil organic carbon pool, microbial communities, and groundwater systems. Thus, more attention should be focused on the increasing accumulation of biochar colloids in the soil and aquatic environments due to frequent wildfires and artificial biochar application. The long-term effect of biochar field application on DOM release remains largely unknown and deserves further attention.