Unraveling the Immobilization Mechanisms of Biochar and Humic Acid on Heavy Metals: DOM Insights from EEMs-PARAFAC and 2D-COS Analysis

Abstract

The structural complexity and variability of dissolved organic matter (DOM) significantly affect its binding capacity with heavy metals (HMs). This study evaluated the remediation efficacy of biochar (BC) and humic acid (HA) on Mn- and Cu-contaminated soils using four maize pot treatments: 3% BC (YB3), 6% BC (YB6), 3% BC + 1% HA (YB3H), and 6% BC + 1% HA (YB6H). The results showed that compared to the control (Y), Results showed Mn and Cu concentrations in rhizosphere soil decreased by 11.08–17.76%, while DOM content increased by 44.2–103.83%. BC enhanced DOM aromaticity and humification, further intensified by HA, leading to a more complex and stable DOM structure. PARAFAC identified four DOM components in BC (BC-DOM): C1 (fulvic-like), C2/C3 (humic-like), and C4 (protein-like), and in BC + HA (BC + H-DOM), an enhanced structural complexity with additional aromatic C–H groups was observed. 2D-COS analysis revealed that in BC-DOM, polysaccharides primarily interacted with Mn and Cu, followed by carboxylic acids and phenolic hydroxyl groups, but in BC + H-DOM, aromatic C–H groups preferentially bound Cu before polysaccharides, showing weaker affinity for Mn. These results elucidate the DOM-mediated immobilization mechanisms of BC and HA for HMs, offering insights for soil remediation and carbon sequestration strategies.

1. Introduction

The acceleration of modernization will trigger serious soil pollution problems [1,2,3,4]. Taking the Xiangtan manganese mine as an example, the manganese (Mn) content in the neighboring farmland has seriously exceeded the standard due to the mining activities and improper treatment of tailings and is accompanied by the pollution of other heavy metals [5,6,7]. A study showed that the Mn content in the farmland around the Mn mine exceeded the standard by 16.3 times, and, at the same time, the content of lead (Pb) and copper (Cu) also exceeded the standard by 15.4 times and 15.0 times, respectively [8]. In addition, the heavy metal (HM) content at the soil–water interface in the Mn mining area was more affected by Cu pollution than Mn pollution [9]. This complex and multifaceted pollution seriously threatens soil fertility and food security production. Therefore, there is an urgent need to remediate Mn–Cu contamination in agricultural lands around Mn mining areas and restore soil health to ensure food security.

Biochar is an organic material produced by the pyrolysis process under high temperature and oxygen-limited conditions [10]. Its surface is rich in a variety of functional groups, which can effectively immobilize HM in the soil and thus inhibit the migration of HM to crops to a certain extent [11,12,13]. A study reported that the application of tobacco biochar in a lettuce field (5 t hm⁻¹) resulted in a 23% reduction in the bioactive state of Cu in the soil and a 26.4% reduction in Cu concentration in the aboveground portion of lettuce [14]. Similarly, Mn concentration in spinach was significantly reduced by 78.0% after the application of broadleaf wood biochar to Mn-complex contaminated soil [15]. Furthermore, biochar can release dissolved organic matter (DOM) due to its porous structure and large specific surface area [16].

DOM, as the most active form of organic carbon in the soil, has a complex composition containing carboxyl, hydroxyl, aromatic ring, double bond, and other rich functional groups [17,18]. These functional groups can mediate the transport of HMs in soil. It was shown that there are differences in the functional groups of DOMs from different sources in their response to HMs. For instance, DOM derived from corn stalk undergoes static quenching in Pb solution, where the amide group binds to Pb first, followed by phenolic hydroxyl groups and C-O-O telescopic deformation to produce a quenching response [19]. Li et al. found that under the background of Cd and Cu contamination, the binding responses of phenolic hydroxyl and carboxyl groups in compost-DOM were the strongest, while the amide and phenolic hydroxyl groups in rice straw-DOM responded the fastest [16]. This indicates that the DOM and HM responses are different mainly related to functional groups. Moreover, the diversity of functional groups depends on the DOM chemical composition. Therefore, DOM substance composition directly affects its binding effect with HM.

The results of a large number of studies have shown that DOM is mainly composed of humic-like substances (e.g., humic acid or xanthic acid) and protein-like substances (e.g., tyrosine and tryptophan) [20,21]. Among them, humic-like substances, especially humic acid (HA), have large molecular weights and are rich in functional groups that can form stable complexes with HM ions [22,23]. Therefore, their binding capacity with HM far exceeds that of protein-like substances. In summary, the higher the proportion of HA in DOM, the more efficiently it binds to HM, and the complex formed is more stable, with a corresponding decrease in mobility. It is worth noting that HA, as a kind of macromolecular organic matter, not only enhances soil fertility but also regulates the adsorption of biochar on soil DOM [24,25]. Therefore, the following questions are worth exploring in depth: how will the strategy of combining HA with biochar affect the compositional composition of DOM? How does this strategy affect the binding effect with HM by changing the functional groups? And does this combination enhance the ability to remediate HM-contaminated soil?

With the advancement of spectroscopy, two-dimensional correlation mapping (2D-COS) has emerged, which is capable of displaying more informative two-dimensional information, such as the reaction sequences of functional groups, than the one-dimensional state of functional group categories [26,27]. Meanwhile, three-dimensional fluorescence scanning (3D-EEMs) is a fluorescence analysis technique that can be used to determine the composition of DOM in samples [28]. When combined with parallel factor analysis (PARAFAC), it can separate specific DOM components [29]. This combined EEM-PARAFAC approach provides an in-depth means for qualitative and quantitative studies of DOM. Based on this, this paper first designed a potting experiment with the addition of different amounts of BC and BC + HA to explore their remediation effects on HMs. Subsequently, DOM derived from BC (BC-DOM) and DOM derived from BC and HA (BC + H-DOM) were extracted to conduct quenching titration experiments and combined with 2D-COS and 3D-EEMs techniques to investigate (1) the differences in soil HM remediation capacity between BC and BC + HA treatments; (2) component composition and differences between BC-DOM and BC + H-DOM; and (3) major binding modes and response order of BC-DOM and BC + H-DOM to HM functional groups.

2. Materials and Methods

2.1. Material Preparation

The corn straw biochar for the experiment was purchased from ‘Henan Lize Environmental Protection Science and Technology Company Limited’ (Zhengzhou, China). The biochar was obtained by pyrolysis at 500 °C, in which C: 42.21%, N: 8.34%, P: 2.31%, ash content: 7.23%, pH: 9.46. Humus was purchased from ‘Sheng hua Fertiliser’, with organic matter ≥ 80.0%, and humic acid ≥ 40.0%, pH: 4.0–6.0. Maize variety ‘Xixing bainuo NO.2’ was purchased from ‘Shandong Denghai Seed Industry Company Limited’ (Yantai, China). Fertilizer purchased from ‘Jinfeng Seeds’ shop in Xiangtan, Hunan Province. The soil for potting was taken from the farmland around the Mn mine in Xiangtan City, Hunan Province (27°99′ N, 112°78′ E). After collecting the soil, stones and weeds were removed, and the soil was sun-dried. The basic physical and chemical properties of the soil were as follows: pH: 5.7, Mn: 841.82 mg kg⁻¹, and Cu: 50.97 mg kg⁻¹.

2.2. Pot Experiment and Characterization

The pot experiment was conducted at Hunan University of Science and Technology (HUST) from 10 May 2024 to 15 July 2024, where the prepared soil was incubated in the dark for 10 days before the start of the experiment. Potting experiments were conducted in 3-gallon pots and 4 kg of soil was placed in each treatment. The experiment consisted of five treatments supplemented with biochar/humic acid at different percentages of soil mass, i.e., 3% biochar (w:w); 6% biochar (w:w); 3% biochar (w:w) + 1% humic acid (w:w); 6% biochar (w:w) + 1% humic acid (w:w), and a control group with no remediation material added. The treatments were recorded as YB3 (3% biochar), YB6 (6% biochar), YB3H (3% biochar + 1%HA), YB6H (6% biochar + 1%HA), and Y(CK); each treatment was replicated four times with regular management during the experiment. To ensure that the restorative materials are fully mixed with the soil, the pre-weighed amount of biochar or humic acid required for each treatment should be mixed separately with its corresponding soil. At the same time, 1.28 g of nitrogen fertilizer (urea), 1.12 g of phosphorus fertilizer (P₂O₅), and 0.57 g of potassium fertilizer (KCl) were applied to each treatment.

The experiment was harvested on 15 July 2024 with one crop per pot and the soil around the root system of the harvested crop was shaken off into self-sealing bags and rhizospheric soil was obtained. The rhizosphere soil was sieved using 20 mesh and 100 mesh sieves and then stored at 4 °C for heavy metal (HM) analysis. The harvested crops were cleaned and dried, then ground using a grinder to separate the aboveground and belowground parts for HM analysis. Fresh soil was dissolved in deionized water (W:V = 1:10) and shaken on a shaker (180 rpm) for 24 h, then filtered through a 0.45 μm PTFE membrane. The solution was stored in the dark in a refrigerator at 4 °C and the DOM content was measured by a total organic carbon analyzer (TOC) and determined by spectroscopy (UV-Vis and 3D-EEM). For heavy metal measurements, 0.2 g of soil and plant samples were placed on a graphite hot plate and digested with 5 mL H₂SO₄ and 3 mL HClO₄ at 240 °C. After digestion, the samples were analyzed for heavy metals using ICP-MS (Agilent 7900, Agilent Technologies, Santa Clara, CA, USA), The variability of heavy metal content between treatments was obtained by ANOVA.

2.3. DOM Extraction and Quenching Titration

Before extraction of DOM, the biochar was sieved through a 60 mesh sieve and HA was sieved through a 200 mesh sieve and then dried in an oven to avoid the effect of moisture on DOM extraction. A total of 45 g of biochar was dissolved in 450 mL of deionized water (w:v = 1:10), and 45 g of biochar + 15 g of HA was dissolved in 600 mL of deionized water (w:v = 1:10) and shaken on a shaker for 24 h. After shaking, the solution was filtered through a 0.45 μm membrane by vacuum filtration, and the DOM was determined by filtration three times and diluted to 10 mg C L⁻¹ to reduce the effect of infiltration [30].

The DOM extracted from biochar was denoted as BC-DOM, and the DOM extracted from biochar and HA was denoted as BC + H-DOM. Mn and Cu were selected for the experimental titration in this experiment, with MnCl₂ and CuCl₂ at contents of 0.1 mg L⁻¹ and 0.02 mg L⁻¹, respectively. A certain amount of the above solutions was aspirated and titrated to a solution containing a range of contents of HM (Mn: 0, 100, 150, 200, 250, 300, 400 µmol L⁻¹; Cu: 0, 10, 20, 40, 60, 80, 100 µmol L⁻¹) were titrated into a 50 mL brown volumetric flask containing 30 mL of DOM solution. The volume aspirated was less than 0.25 mL for each HM content and the pH of the mixture was maintained at about 8.25 with NaOH and HCl during the titration to minimize the effect of changes in solution and acidity. After the titration, the mixed solution was placed on a shaker for 24 h to ensure complexation equilibrium [28]. After shaking, the solution was stored at 4 °C, and one part was used for spectroscopic (UV-Vis and 3DEEM) determination and the other part was freeze-dried at −80 °C for infrared analysis (Nicolet iS 10, Thermo Scientific, Waltham, MA, USA).

2.4. Spectroscopic Determination

After diluting the extracted DOM samples to 10 mg C L⁻¹, a portion was used to measure the absorbance of DOM at 254 nm and 260 nm (UV254 and UV260) using a UV-Vis spectrophotometer (Shimadzu UV360, Shimadzu, Kyoto, Japan). The absorbance coefficients at 254 nm and 260 nm (a254, a260) were calculated according to Equation (1), and SUVA254 and SUVA260 (L mg⁻¹ m⁻¹) were calculated according to Equation (2) [31].

$${\mathrm{a}}_{254/260}=U{V}_{254}/U{V}_{260}\times 2.303\times 100$$

(1)

$$SUV{A}_{254/260}={\displaystyle \frac{a_{254}/a_{260}}{DOM}}$$

(2)

Another portion was subjected to 3D fluorescence scanning (Hitachi F-7000, Kyoto, Japan), with excitation wavelength scanning ranges of 200–500 nm and emission wavelength scanning ranges of 200–550 nm, both at intervals of 5 nm. The scanning speed was 2400 nm/min, and the excitation/emission slit widths were 5 nm each. During sample scanning, deionized water was also scanned to subtract its spectral values during analysis, thus eliminating Raman and Rayleigh scattering effects [32,33].

After removing the Raman and Rayleigh scattering from the scanned spectra, the fluorescence indices were calculated by integrating the areas of DOM spectra at E(x) = 254 nm, E(m) = 435–480 nm, E(x) = 254 nm, E(m) = 300–345 nm, and the humification index (HIX) was calculated using Equation (3). The fluorescence index (FI) was calculated using the fluorescence intensity at E(x) = 370 nm, E(m) = 470/520 nm (Equation (4)); and the biological index (BIX) was determined using the fluorescence intensity at E(x) = 310 nm, E(m) = 380/430 nm (Equation (5)) [34].

$$HIX={\displaystyle \frac{{\displaystyle \sum E_{(m)}435-480\mathrm{nm}}}{{\displaystyle \sum E_{(m)}300-345\mathrm{nm}}}},E(x)=254\mathrm{nm}$$

(3)

$$BIX={\displaystyle \frac{E(m)=380\mathrm{nm}}{E(m)=430\mathrm{nm}}},E(x)=310\mathrm{nm}$$

(4)

$$FI={\displaystyle \frac{E(m)=470\mathrm{nm}}{E(m)=520\mathrm{nm}}},E(x)=370\mathrm{nm}$$

(5)

2.5. Parallel Factor Analysis and Complexation Modeling

The identification of DOM components using EEM-PARAFAC analysis was carried out with the DOM Fluor toolbox in MATLAB R2021a. After removing Rayleigh and Raman scattering, a three-dimensional factor matrix was constructed, and the matrix was estimated using the Alternating Least Squares (ALS) method. After the factorization, the generalization ability of the model was evaluated by cross-validation. Finally, the fluorescence intensities corresponding to different excitation and emission wavelengths, as well as the maximum fluorescence intensity (Fmax) of each component were obtained.

The quenched solutions were subjected to synchronous fluorescence measurement (Edinburgh FLS1000, Edinburgh, UK), with E(x) = 200–300 nm and E(m) = 250–600 nm, constant offset Δλ = 60 nm. The fluorescence intensities at the wavelengths corresponding to the excitation and emission wavelengths obtained from the PARAFAC analysis of each component were identified in the synchronous fluorescence spectra. The Stern–Volmer linear equation model (Equation (6)) was used for fitting [26,34]. This model assumes that in the fluorescence quenching reaction, the HM ions form a 1:1 complex with the fluorescent probe, which is more suitable for describing the actual complexation reaction.

$${\displaystyle \frac{F_0}{F_0-F}}={\displaystyle \frac{1}{f\cdot K_M\cdot C_M}}+{\displaystyle \frac{1}{f}}$$

(6)

In this model, F₀ and F represent the fluorescence intensities of each component before and after HM titration, C_M is the content of each HM ion, f is the fraction of the fluorescent component bound to the HM ion, and K_M is the complexation constant, reflecting the binding capacity of each component with the HM. K_M and f can be obtained from the linear fitting between F₀/(F₀ − F) and 1/C_M.

2.6. Two-Dimensional Correlation Spectral Analysis

The solutions after quenching titration were freeze-dried using a −80 °C freezer and a vacuum dryer, followed by infrared spectroscopy scanning (400–4000 cm⁻¹). After baseline correction and smoothing of the spectra, the spectral range of 800–1800 cm⁻¹ was selected for two-dimensional spectral analysis. The 2D-COS analysis was performed using the “2D Shige” 1.0.0.0 software released by Kansai University in Japan [18,35], and synchronous and asynchronous 2D correlation spectra were plotted using the 2D Correlation plugin in Origin 2024b.

3. Results and Discussion

3.1. Effects of Biochar and Humic Acid on Heavy Metal Content in Crops and Soil

3.1.1. Heavy Metal Content in Crops

In the analysis of HM contents in the crops, it was found that different treatments significantly altered the distribution pattern of HM within the crops (Figure 1). Notably, the Mn content in the aboveground parts of maize under the YB3, YB6, YB3H, and YB6H treatments was significantly lower than that in the Y treatment, with reductions of 10.81%, 11.05%, 26.03%, and 47.02%, respectively (p < 0.05). Furthermore, in the YB3H and YB6H treatments, the Mn content in the underground parts of maize decreased significantly by 7.07% and 11.55%, respectively, to the Y treatment (p < 0.05). Conversely, the Mn content in the underground parts of maize under YB3 and YB6 treatments did not differ significantly from that in the Y treatment (p < 0.05). This suggests that biochar alone was effective in reducing Mn uptake in the aboveground maize, and this deterrent effect was further enhanced when biochar was used in conjunction with HA [36]. On the contrary, Cu content in aboveground parts of maize under YB3, YB6, YB3H, and YB6H treatments was significantly higher than that under Y treatment, and it increased synchronously with the increase of biochar and HA application amount (YB3-YB6-YB3H-YB6H). The underground Cu content in YB6, YB3H, and YB6H treatments was significantly lower than in Y treatment by 23.85%, 23.83%, and 40.40% (p < 0.05), respectively. Overall, all of the high remediation material treatments (YB6, YB3H, and YB6H) increased aboveground Cu content but significantly decreased underground Cu content.

3.1.2. Heavy Metal Content and Morphology of Soils

Biochar and HA, when applied at different dosages and ratios, exhibited varying remediation effects on the immobilization of different HMs in the soil. The results in Figure 2A showed that the Mn content of maize rhizosphere soil was significantly reduced by 12.02% and 11.08% in treatments YB3 and YB6, respectively, compared with Y treatment (p < 0.05); the solidification and remediation efficiencies of Mn in the soil were further enhanced by the YB3H and YB6H treatments, which significantly reduced by 17.76% and 14.53%, respectively (p < 0.05) compared with Y treatment. This indicates that an excessive application of remediation materials may have a negative effect on the immobilization of soil HM, thereby counteracting the original remediation benefits. This is mainly attributed to the fact that excessive remediation materials may disturb the balance of the soil and its microenvironment, and change the competition of HM ions for the active adsorption sites in the soil, resulting in easier desorption and release of HM ions from the soil, which in turn increases their mobility and bioavailability [37,38,39].

Cu content of the maize rhizosphere soil in the YB3, YB6, YB3H, and YB6H treatments was significantly reduced by 16.15%, 15.29%, 16.42%, and 16.61% (p < 0.05), respectively, compared with that in treatment Y, but there was no significant difference between the treatments of YB3, YB6, YB3H, and YB6H. This indicates that the different materials and ratios of treatments did not significantly affect the rhizosphere soil Cu. Meanwhile, the curing repair effects of biochar and HA on different types of HMs (Mn and Cu) are significantly different, which were mainly attributed to the fact that the chemical properties of Mn and Cu, especially the degree of filling of their d-electronic layer and the electron arrangement, determine the oxidation state and coordination chemistry of the HMs, which largely affect their interactions with biochar and HA [40].

Biochar and HA not only affect the content of HMs in soil but also significantly affect their morphology. The results of the morphology of HMs in the rhizosphere soil of maize showed (Figure 2B) that in the treatments YB3, YB6, YB3H, and YB6H, the stable states (Residual and Oxidizable) of Mn increased by 16.12%, 24.8%, 18.68%, and 20.65%, respectively, in comparison to the Y treatment. Conversely, the stable state of Cu showed a slight decrease compared to the Y treatment. These findings indicated that the application of biochar and HA not only effectively decreased the total Mn and Cu content in soil but also successfully immobilized a certain quantity of Mn, which was verified in most of the studies [41,42]. Biochar and HA directly participate in HM immobilization through complexation and redox reactions in a soil micro-interaction environment. On the other hand, DOM released from biochar is involved in this process [43,44] which collectively resulted in alterations in the quantity or morphology of HM within the maize rhizosphere soil.

3.2. Effect of Biochar and Humic Acid on the Content, Characterization and Component of DOM in Soil

3.2.1. Soil DOM Content

The effects of different additions of biochar and HA on DOM content in soil varied, with the results in Figure 3 indicated significant increases in DOM content by 44.2%, 88.45%, 69.22%, and 103.83% (p < 0.05), respectively, in treatments YB3, YB6, YB3H, and YB6H compared to treatment Y. Specifically, the DOM contents in YB6 and YB6H treatments were as high as 8.56 mg L⁻¹ and 9.25 mg L⁻¹, respectively. The results indicated that the DOM content in the soil increased with both the addition of biochar and HA. However, biochar had a more significant effect on enhancing the soil DOM content than HA. Some studies have shown that biochar can release DOM in soil, which is the main reason for increasing soil DOM content [45,46].

3.2.2. Soil DOM UV-Vis Absorption and Fluorescence Index

In this study, the properties of soil DOM were evaluated using UV parameters and 3D fluorescence spectroscopy parameters. The results of the parameters are shown in

Table 1. Soil DOM UV-visible and spectral parameters. Different letters indicate significant differences between treatments.

	SUVA254	SUVA260	FI	BIX	HIX
Y	9.96 ± 1.05 c	8.80 ± 0.67 c	2.31	0.81	3.64
YB3	11.49 ± 0.46 c	10.65 ± 0.44 c	2.46	0.80	4.50
YB6	12.28 ± 0.55 c	11.35 ± 0.5 c	2.31	0.76	4.75
YB3H	86.11 ± 2.51 a	82.3 ± 2.5 a	2.21	0.67	10.25
YB6H	28.71 ± 3.88 b	28.48 ± 1.97 b	2.06	0.68	11.25

. Among them, SUVA254 denotes the ratio of absorbance to DOM at 254 nm, a larger value indicates a higher degree of aromatization in the organic matter [47]. The results demonstrated that the aromatization level of soil DOM in treatments YB3 and YB6 did not exhibit a significant increase compared to the Y treatment. Nevertheless, in treatments YB3H and YB6H, the aromatization level of soil DOM increased significantly by 760% and 188.25%, respectively (p < 0.05). This indicates that HA, serving as the primary raw material, incorporates a substantial amount of functional groups including benzene rings into DOM, thereby markedly enhancing its stability [48]. Furthermore, the alignment in trends between SUVA260 and SUVA254 further suggests that the incorporation of HA elevates the proportion of highly hydrophobic compounds within the DOM [47].

On the other hand, the fluorescence index (FI) presented in Table 1 provides insights into the origin of soil DOM. Specifically, an FI value below 1.4 suggests that DOM primarily originates from terrestrial inputs, whereas an FI value above 1.9 indicates that DOM is predominantly of endogenous origin, resulting from the decomposition of soil organic matter by microorganisms [49]. The results showed that the FI values of all treatments exceeded 1.9, indicating that DOM in maize rhizosphere soil was mainly derived from the decomposition and transformation process of added biochar and HA by soil microorganisms, rather than directly from land input.

Meanwhile, the data in Table 1 show that the humification index (HIX) of DOM in soil increased significantly with YB3-YB6-YB3H-YB6H treatments. In particular, the HIX values of YB3H and YB6H treatments increased by 181.59% and 209.07%, respectively, compared with the Y treatment. In terms of biodegradation index (BIX) of soil DOM, the BIX values of treatments with only biochar (YB3 and YB6) were not significantly different from those of Y treatment, but the BIX values of treatments with biochar and HA (YB3H and YB6H) were significantly lower than those of Y treatment, with decreases of 17.28% and 16.05%, respectively. This suggests that the addition of HA further deepened the humification of DOM and reduced the ability of DOM to be decomposed and utilized by microorganisms, thus enhancing the potential of DOM chelating HM when biochar is used alone [50].

3.2.3. PARAFAC Analysis of Soil DOM

The composition information of biochar-derived DOM (BC-DOM) and biochar + HA-derived DOM (BC + H-DOM) was obtained by three-dimensional fluorescence scanning. Subsequently, these fluorescence spectra were decomposed using parallel factor analysis (PARAFAC), with the specific component distributions illustrated in Figure 4A,B. From BC-DOM, four fluorescence components were identified based on their excitation (Ex) and emission (Em) wavelengths: C1 (Ex/Em = 240–260/405 nm), C2 (Ex/Em = 350/465 nm), C3 (Ex/Em = 260/460 nm), and C4 (Ex/Em = 270/320 nm). Likewise, four fluorescent components were isolated from BC + H-DOM as C1 (Ex/Em = 250/405 nm), C2 (Ex/Em = 270/450 nm), C3 (Ex/Em = 200/400 nm), and C4 (Ex/Em = 240/340 nm). Notably, C1 was identified as a fulvic acid-like substance, which appears in the conventional A-peak region (230–260/380–420 nm). Fulvic acids are low-molecular-weight compounds that are water-soluble, making them easily transported and absorbed within both environmental and biological systems [17]. C2 and C3 are humic acid-like substances or their compounds (Ex/Em = <250–350/400–465 nm). These substances correspond to the traditional humic acid-like A and C peaks, whose absorption peaks are typically associated with aromatic compounds and unsaturated structures [32,51]. C4 was the only protein-like component, with peaks located in the region of the protein-like T-peak (270–280/320–340), and was considered to be a tryptophan-like substance [52]. This component is closely related to microbial activities and organic matter degradation. The Ex and Em of the above components reveal that C3 exhibited a redshift relative to C2. This observation suggested that C2 possesses a higher molecular weight, a greater degree of humification, and typically resides in a more intricate environment [32]. In addition, C2 and C3 also showed a redshift in BC + H-DOM compared with their counterparts in BC-DOM, which indicated that the addition of HA improved the complexity and compactness of DOM molecules.

The F_max values for each component of DOM are depicted in Figure 5. In BC-DOM, the F_max values for C1 (fulvic acid-like), C2, C3 (humic acid-like), and C4 are 3.94, 2.60, 3.00, and 1.82, respectively. This indicates that C1 has the highest content, followed by C3 and C2, collectively accounting for 83.99% of the total components. In BC + H-DOM, the F_max values for C1, C2, C3, and C4 are 4.63, 2.85, 1.28, and 1.09, respectively, with C1, C2, and C3 together comprising 88.97% of the total components. The above results suggest that the combined use of biochar and HA can increase the humus-like composition of DOM by nearly 5% compared with the addition of biochar alone, which may be the main reason for the more complex structure of BC + HA-DOM [53,54].

3.3. DOM and HM Bonding Capability and Its Characteristics

3.3.1. Quenching Curve of PARAFAC Component

The excitation wavelengths (Ex) and emission wavelengths (Em) of the DOM components (C1, C2, C3, C4) under various treatments were determined through parallel factor analysis. Subsequently, the fluorescence intensities at specific response wavelengths were derived from synchronized fluorescence data, and the corresponding curves depicting the relationship between fluorescence intensity and HM content were plotted (Figure 6). In BC-DOM, the fluorescence intensity of the C1, C2, and C3 components progressively diminished as Mn content increased, suggesting that the humic acid-like substances (C1, C2, and C3) bound to Mn, resulting in effective fluorescence quenching. Similarly, the fluorescence intensities of C1, C2, and C3 components decreased with the increase of Cu content, but the quenching effect diminished after Cu contents exceeded 80 µmol L⁻¹, 60 µmol L⁻¹, and 60 µmol L⁻¹, respectively. This quenching response at varying HM contents resembles the ‘failure’ phenomenon in other related articles [55,56]. This may be attributed to the fact that at the beginning of the titration, a certain amount of Cu binds to the functional groups in HA, resulting in an effective quenching. However, as the Cu content further increases, its interaction with specific sites in C1, C2, and C3 can cause Aggregation-Induced emission changes (AIE), restricting internal molecular motion (RIE), which in turn increases the photoluminescence quantum yield (PLQY) [57,58]. Furthermore, the elevated Cu content may also induce the fluorescence resonance energy transfer (FRET) phenomenon, where energy is transferred among excited state molecules, serving as another factor influencing the fluorescence quenching effect [59].

In BC + H-DOM, the humic-like substances C1, C2, and C3 exhibited similar quenching effects on Mn as in BC-DOM. When Cu was used as the quenching agent, these substances did not show obvious ‘failure’ at high concentrations. However, the fluorescence intensity of the humic-like substances in BC + H-DOM was significantly lower than that in BC-DOM. This reduction may be attributed to intermolecular π-π stacking and ligand interactions, which may restrict molecular mobility [60]. Regarding the protein-like component C4, the proteinaceous substances in BC-DOM and BC + H-DOM exhibited different changes in fluorescence intensity during the quenching process with Mn and Cu. Specifically, the protein-like substances in BC + H-DOM demonstrated higher stability when complexed with Mn and Cu compared to those in BC-DOM. It has been demonstrated that the addition of HA may introduce additional carboxyl groups to DOM, allowing the protein-like substances to form complexes with HM through these groups, thereby enhancing the overall stability of DOM [22].

3.3.2. Binding Capability of Humic Acid-like Substances to HM

The improved Stern–Volmer model was used to analyze the fluorescence intensity, enabling a deeper investigation into the binding capabilities of various DOM components with different HM. The relevant fitting parameters were outlined in

Table 2. Binding parameters between Mn, Cu and BC-DOM and BC + H-DOM components calculated by Stern–Volmer model.

		Log KM	R2	f
Mn	BC-C1	2.01	0.88	0.61
	BC-C2	2.56	0.93	1.31
	BC-C3	3.21	0.89	2.79
	BC + H-C1	3.36	0.92	8.23
	BC + H-C2	3.68	0.99	5.42
	BC + H-C3	3.56	0.98	1.07
Cu	BC-C1	1.25	0.94	0.39
	BC-C2	0.87	0.84	0.48
	BC-C3	1.30	0.97	0.54
	BC + H-C1	0.71	0.82	0.69
	BC + H-C2	0.73	0.82	0.54
	BC + H-C3	0.86	0.83	0.67

. Based on the fitting parameters of DOM and BC + H-DOM with Mn, the Log K_M values range from 2.01 to 3.68. Among them, BC-C3 in BC-DOM exhibited the strongest binding affinity for Mn, with the corresponding f value being the largest. In BC + H-DOM, BC + H-C2 showed the strongest binding ability to Mn, followed by BC + H-C3. Notably, BC + H-C1 had the highest f value, indicating that although the binding affinities of these three components are comparable, their binding sites and processes differ. BC + H-C1 may bind to weaker binding sites but maintains a higher Mn-complexing effect, while BC + H-C3 showed the opposite trend. For Cu, the Log K_M of BC-DOM/BC + H-DOM binding Cu ranged from 0.71 to 1.30, and the f ranged from 0.39 to 0.69, which was in agreement with the results of a previous study [18], with the Log K_M value of BC + H-DOM being lower than that of BC-DOM, suggesting that compared to BC + H-C1, C2, and C3, BC-C1, C2, and C3 exhibit stronger complexing ability with Cu.

3.3.3. DOM-HM Binding Order

Through two-dimensional correlation analysis of the FTIR spectra of different DOM derived from biochar/biochar + HA at different HM contents demonstrated the vibrational modes of the strongly correlated characteristic peaks and the dynamic order of structural adjustments, as shown in Figure 7A,B. The synchrotron plots for BC-DOM and BC + H-DOM during Mn and Cu titrations exhibited three and four autocorrelation peaks, respectively. The peaks common to both were situated at 1582/1586 cm⁻¹, 1386 cm⁻¹, and 1106 cm⁻¹, and these peaks are related to the asymmetric stretching vibration of COO⁻, the bending (or deformation) vibration mode of -OH, and the stretching vibrations of the C–O bond in polysaccharides [61,62,63]. Notably, BC + H-DOM showed an additional peak in the range of 800–874 cm⁻¹, suggesting that the addition of HA affected the deformation of the aromatic C–H bond and thus increased the activity of DOM [47].

In the asynchronous plots of Mn, two negative correlation peaks (1582/1106 cm⁻¹ and 1386/1106 cm⁻¹) and one positive correlation peak (1582/1386 cm⁻¹) were identified. According to Noda’s rule [64], the sequence of functional group binding to Mn in BC-DOM is 1106 cm⁻¹, 1582 cm⁻¹, and 1386 cm⁻¹. However, after the addition of HA, an additional peak emerged at 801 cm⁻¹, altering the peak sequence of the functional group of Mn bound to BC + H-DOM to 1582 cm⁻¹, 1386 cm⁻¹, 1106 cm⁻¹, and 801 cm⁻¹. Therefore, the binding order of BC-DOM with Mn is polysaccharides’ C–O bonds–carboxyl groups–phenolic hydroxyl groups. After the addition of HA, the C–H stretching vibration in the aromatic compound altered the binding sequence of Mn to BC + H-DOM. Consequently, the functional groups of Mn bound to BC + H-DOM are, in order, carboxylate, phenolic hydroxyl, polysaccharide C–O bond, and the C–H bond in the aromatic compound.

In the asynchronous plots of Cu, BC-DOM exhibited three negatively correlated peaks at 1586/1386 cm⁻¹, 1586/1106 cm⁻¹, and 1386/1106 cm⁻¹. According to the principles of spectral interpretation, the sequence of peaks representing the functional groups of Cu bound to BC-DOM is 1106 cm⁻¹, 1386 cm⁻¹, and 1586 cm⁻¹. However, after the addition of HA, the original three negative correlation peaks transformed into positive correlation peaks, and new negative correlation peaks emerged at 1586/874 cm⁻¹, 1386/874 cm⁻¹, and 1106/874 cm⁻¹. Consequently, the sequence of peaks associated with the functional groups bound to Cu in BC + H-DOM shifted to 874 cm⁻¹, followed by 1586 cm⁻¹, 1386 cm⁻¹, and finally 1106 cm⁻¹. Specifically, in BC-DOM, the functional groups bound to Cu followed the order of C–O bond, phenolic hydroxyl group, and carboxyl group within polysaccharides. In contrast, for BC + H-DOM, the order of functional groups bound to Cu was C–H bond, carboxyl group, phenolic hydroxyl group, and lastly, the C–O bond in polysaccharides.

The results indicated that carboxyl, hydroxyl, and polysaccharide–CO were the primary functional groups present in the humic-like substances within BC-DOM. Notably, polysaccharide–CO exhibited the strongest binding affinity for Mn and Cu, a finding that aligns with previous research by He et al. [65]. Furthermore, additional studies have demonstrated that in the presence of polysaccharide–CO, carboxyl groups, and phenols also possess a high binding capacity for Cu. Specifically, phenols coordinate with Cu through oxygen atoms, forming a stable ring structure [66]. In BC + H-DOM, in addition to the above three functional groups, the C–H bond in aromatic compounds is also involved in the reaction, and its presence changes the response of other functional groups to Mn. Notably, aromatic compounds respond better to Cu than phenols, polysaccharides, and carboxylic acids. This may be due to the influence of the C–H bond in the aromatic ring on the electron density of other functional groups, or the spatial steric hindrance effect that alters the binding capacity of other functional groups to HM In BC + H-DOM, in addition to the above three functional groups of carboxyl, hydroxyl and polysaccharide-CO, the C–H bonds in the aromatic compounds were also involved in the binding to HM, simultaneously altering the response of other functional groups to HM. This may be attributed to the influence of C–H bonds in aromatic rings on the electron density of other functional groups, or to steric hindrance effects that change the binding ability of other functional groups with HM [64,67].

3.4. Reflections on HM Remediation with Biochar/Biochar + HA

The study confirms that the combined use of biochar and biochelator exhibits significant remediation effects on Mn, Cd, and Cu contamination in soil. Using 3D-EEM and PARAFAC techniques, it was discovered that the concurrent application of HA with biochar enhanced the stability of DOM and facilitated its humification process. This combination also increases the content of macromolecular structures, such as aromatic rings, enriches the functional group varieties within DOM, and notably improves the binding capacity between DOM and HM.

However, molecular aggregation was found to occur during the quench titration when DOM was exposed to high concentrations of Cu contamination, which limited the ability of DOM to bind Cu. In addition, the results of 2D-COS showed that the addition of HA changes the reaction order of functional groups. Therefore, how to avoid the ‘failure’ of HM repair with a high concentration of HM and how to utilize the ‘interference’ effect of HA to promote the binding of HM become a question we need to think deeply. Perhaps we can try to modify the biochar with HA before extracting DOM to enhance the binding of HA to DOM more effectively through the formation of chemical bonds or surface chemisorption, thus optimizing the repair effect of HM.

Furthermore, HA and biochar and their derivatives consist mainly of organic carbon. Applying them to soil is a strategy for carbon sequestration. It is of crucial importance to explore in depth how they affect the mineralization rate of soil organic carbon, whether they can effectively enhance soil fertility, reduce CO₂ emissions, and whether they can simultaneously achieve the dual goals of soil remediation and carbon sequestration.

4. Conclusions

This study verified the effectiveness of biochar/biochar + humic acid (HA) on the remediation of Mn and Cu soil pollution and revealed the effective components and functional groups of/binding to Mn and Cu from the perspective of dissolved organic matter (DOM). Analyses of DOM extracted from biochar/biochar + HA showed that humus-like components played a more important role in binding to HM. The incorporation of HA enhanced the molecular structure and density of DOM and provided more functional groups for binding to HM. The 2D-COS results indicate that the stretching vibrations of C–O bonds in polysaccharides and the protonation stretching vibrations of carboxylic acids, as well as the bending vibrations of O–H bonds in phenolic compounds, play a significant role in binding with HM. However, the introduction of HA also leads to the presence of aromatic ring C–H bonds in DOM, which introduces disturbances in the binding process with HM. In summary, the addition of HA enriches the functional groups of DOM and enhances its binding capacity with HM. This finding provides a theoretical basis and practical guidance for the development of efficient and sustainable soil remediation technologies.