Introduction

Micronutrient availability in calcareous soils—particularly iron (Fe), manganese (Mn), zinc (Zn), and copper (Cu)—is a critical determinant of agricultural productivity in arid and semi-arid regions. These soils, characterized by alkaline pH (7.5–8.5), high calcium carbonate content, and low organic matter (< 2%), inherently limit micronutrient solubility through adsorption, precipitation, and redox immobilization1. Conventional fertilization strategies often fail to address these constraints, necessitating sustainable amendments to enhance nutrient bioavailability. Humic substances (HS), the chemically complex fraction of soil organic matter, have emerged as promising candidates due to their ability to chelate metals and stabilize micronutrients in plant-available forms2,3. However, the efficacy of HS in calcareous systems remains inconsistently documented, with a poor understanding of how their molecular architecture and application parameters govern micronutrient dynamics.

HS are broadly classified into humic acids (HA) and fulvic acids (FA), differentiated by solubility and structural properties. FA, with lower molecular weight (< 2 kDa), higher oxygen-rich functional groups (e.g., carboxylic, phenolic), and solubility across a wide pH range (2–12), forms mobile metal complexes that resist precipitation4. In contrast, HA’s larger molecular size (> 10 kDa) and aromatic backbone promote adsorption to soil colloids, potentially reducing micronutrient mobility4,5. While both fractions interact with metals via ligand exchange and redox reactions, their functional group density and hydrophobicity—shaped by source material (e.g., lignite and compost) and extraction methods—introduce variability in reactivity6,7,8.

Despite commercial promotion of HS as ‘universal’ soil conditioners, their performance in calcareous environments is poorly predicted, with fragmented insights into dose–response relationships and soil-specific interactions9. This inconsistency stems primarily from three knowledge gaps: (1) insufficient characterization of molecular structural variability in commercial HS products, particularly regarding hydrophobic domains and lignin-derived motifs; (2) neglect of how soil mineralogy (e.g., clay type, and CaCO3 surface properties) modulates HS-metal interactions; and (3) overdependence on functional group abundance as a predictor of efficacy, overlooking architectural determinants of complex stability7,8. A critical knowledge gap lies in the mechanistic link between HS molecular properties and micronutrient mobilization in calcareous soils. While carboxylic and phenolic groups are recognized as key chelation sites10, their abundance alone does not fully explain HS efficacy, as structural features like hydrophobicity and lignin-derived moieties may regulate metal binding stability11. Furthermore, commercial HS products are rarely characterized at a molecular level, leading to inconsistent field results. To address these limitations, this study introduced an innovative approach by integrating: (i) advanced molecular fingerprinting (CPMAS 13C NMR and DRIFT) to quantify architectural features like alkyl/O-alkyl ratios and condensed aromaticity; and (ii) systematic evaluation of six calcareous soils with contrasting mineralogies (kaolinite/illite dominance and CaCO3 gradients). This dual-axis design—the first applied to calcareous systems—explicitly resolves how soil mineralogy filters HS structural functionality. Accordingly, we hypothesized that: FA would outperform HA in enhancing micronutrient availability due to superior solubility; HS efficacy would correlate more strongly with hydrophobic indices (A/OA) than acidic group density; and elevated HS rates would overcome soil constraints through architecture-mineralogy synergies. To test these hypotheses, this study investigated: (i) comparative impacts of HA/FA on DTPA-extractable micronutrients; (ii) quantitative links between NMR/FTIR-derived structural indices and nutrient mobilization; (iii) dose–response relationships; and (iv) soil mineralogical modulation via clay-CaCO3 interactions.

Materials and methods

Soil sampling and characterization

Six calcareous soil samples were collected from the top 20 cm of agricultural fields across diverse regions in Iran. After air-drying, soils were homogenized and sieved (< 2 mm) for analysis according to routine procedures12,13. Key physicochemical properties, including pH and electrical conductivity (EC), were measured in a saturated paste extract using standard methods. Organic carbon (OC) content was determined via the Walkley–Black wet oxidation method, while calcium carbonate equivalent (CCE) was quantified through acid neutralization. Soil texture was analyzed using the hydrometer method. Plant-available Fe, Mn, Zn, and Cu were extracted with 0.005 M DTPA-TEA-CaCl2 (pH 7.3) and quantified by flame atomic absorption spectrometry (PerkinElmer 1100)).

Selection and purification of HAs and FAs

Ten commercial samples each of humic acids (HAs) and fulvic acids (FAs) were sourced from diverse suppliers for analysis. The purification of HAs followed established protocols14, beginning with alkaline extraction in 0.1 M NaOH under a nitrogen atmosphere to minimize oxidation. The soluble fraction was isolated via centrifugation, acidified to pH 1.0 with 6 M HCl to precipitate HA, and centrifuged again. The precipitate was redissolved in a minimal volume of 0.1 M KOH under nitrogen and reprecipitated at pH 1.0 with HCl, a cycle repeated to enhance purity. Residual ash was removed by agitating the HA in 0.1 M HCl + 0.3 M HF for 36 h, a step repeated thrice. The purified HA was dialyzed against distilled water for 48 h, treated with Amberlite IR-120 resin (H⁺ form) to eliminate cations, and freeze-dried. For FAs, the supernatant from the initial NaOH extraction was acidified to pH 1.0 with HCl, and the soluble FA fraction was isolated via DAX-8 resin chromatography, where non-humic impurities were eluted, and adsorbed FA was desorbed using 0.1 M NaOH. The FA solution was de-ashed with 0.3 M HF, further purified through Amberlite IR-120 resin (H⁺ form) to remove residual ions, and freeze-dried14,15. All steps were conducted under controlled conditions to ensure reproducibility and minimize organic degradation.

Characterization of purified HA and FA samples

Elemental analysis

The carbon (C), hydrogen (H), nitrogen (N), and sulfur (S) contents of the purified HA and FA samples were determined using an Elemental Analyzer EA 1108 (Fisons Instruments). Samples were combusted at high temperatures in an oxygen-rich environment, converting the elements into gaseous compounds (CO2, H2O, NOx, and SO2), which were quantified via thermal conductivity detection. Oxygen (O) content was calculated by mass difference:

$${\text{O}}\% = {1}00\% - \left( {{\text{C}}\% + {\text{H}}\% + {\text{N}}\% + {\text{S}}\% + {\text{ash}}\% } \right)$$

Ash content, determined separately by heating samples at 750 °C to constant weight, was subtracted to express all elemental compositions on an ash-free basis. Atomic ratios of hydrogen-to-carbon (H/C) and oxygen-to-carbon (O/C) were calculated using the equations:

$${\text{H/C}} = \frac{{{\text{wt}}\% {\text{H/}}1.008}}{{{\text{wt}}\% {\text{C/}}12.01}}$$
(1)
$${\text{O/C}} = \frac{{{\text{wt}}\% {\text{O/}}16.00}}{{{\text{wt}}\% {\text{C/}}12.01}}$$
(2)

where 1.008, 12.01, and 16.00 are the atomic weights of hydrogen, carbon, and oxygen, respectively.

Diffuse reflectance infrared fourier transform spectroscopy (DRIFT)

Diffuse Reflectance Infrared Fourier Transform (DRIFT) spectra of the purified HA and FA samples were acquired using a PerkinElmer 1720-X FT-IR spectrometer (Waltham, MA, USA) equipped with a diffuse reflectance accessory. Prior to analysis, samples were homogenized with anhydrous potassium bromide (KBr) at a 1:100 (w/w) ratio and finely ground to achieve a particle size of < 2 µm. Spectra were recorded over the mid-infrared range (4000–400 cm⁻1) by averaging 32 scans at a resolution of 4 cm⁻1. A background spectrum of pure KBr was collected and automatically subtracted from each sample measurement to eliminate environmental interference. Data processing included baseline correction using a concave rubberband algorithm and normalization to the most intense absorption band to ensure comparability between spectra16.

Acidic functional groups

The total acidity, phenolic hydroxyl (─OH), and carboxylic acid (─COOH) group contents of the purified commercial HA and FA samples were quantified using established methods outlined by Swift14. Total acidity was determined via the barium hydroxide titration method, while carboxylic acid groups were measured using the calcium acetate method. Phenolic hydroxyl group content was derived by subtracting the carboxylic acid content from the total acidity, reflecting the contribution of non-carboxyl acidic groups. Following functional group characterization, five HA and five FA samples were selected to represent a spectrum of acidic functional group contents (low, medium, and high) for subsequent soil application experiments. This selection ensured variability in reactive group availability, enabling assessment of their influence on soil processes.

Solid-state CP-MAS 13C NMR analysis

Solid-state cross-polarization magic-angle spinning (CP-MAS) 13C NMR spectra of purified HA and FA samples were acquired using a Bruker AVANCE 300 spectrometer (Karlsruhe, Germany) equipped with a CP-MAS probe operating at a 13C frequency of 75.47 MHz. Samples were packed into 4-mm zirconia rotors sealed with Kel-F caps and spun at 13,000 Hz. Spectra were collected under the following conditions: 2 s recycle time, 30 ms acquisition time, 4000 scans, and 1 ms contact time, with a 1H ramp sequence applied to mitigate Hartmann-Hahn condition inhomogeneity. Data processing involved Fourier transformation using 4 k data points and exponential apodization with 100 Hz line broadening. The 13C chemical shift range (0–200 ppm) was segmented into six regions based on established assignments: alkyl C (0–45 ppm), methoxy/N-alkyl C (45–60 ppm), O-alkyl C (60–110 ppm), unsubstituted and alkyl-substituted aromatic C (110–145 ppm), O-substituted aromatic C (145–160 ppm), and carboxyl/carbonyl C (160–200 ppm)16,17. Spectral integration and quantification of regional contributions were performed using MestReNova 6.2.0 software (Mestrelab Research, 2010). Structural indices, including hydrophobicity (HB/HI: [Aryl-C + Phenol-C + Alkyl-C]/[Carboxyl-C + O-Alkyl-C]), alkyl-C/O-alkyl-C (A/OA), aromaticity (ARM: aromatic C [110–160 ppm]/total C), and lignin ratio (LigR: methoxy/N-alkyl C [45–60 ppm]/O-alkyl C [60–110 ppm]), were calculated to evaluate organic matter composition and degradation patterns18. These indices provided insights into the structural complexity and functional behavior of HA and FA in soil systems.

Selection of extractant

DTPA (diethylenetriaminepentaacetic acid) was selected as the extractant for soils treated with humic substances (HS), as it is widely employed to assess plant-available micronutrient concentrations. However, concerns arose regarding the potential limitations of FA and HA as complexing agents, as their binding capacity for micronutrients may be weaker than that of DTPA. To evaluate this hypothesis, preliminary experiments compared HS-treated soils extracted with DTPA to those treated with weaker extractants, such as sodium acetate (data not shown). Results revealed no significant complexing effect of HS on micronutrient availability when weaker extractants were used, suggesting that HA and FA may not enhance micronutrient solubility under low-binding-strength conditions. Consequently, DTPA was retained as the primary extractant to align with standardized protocols for evaluating plant-available micronutrient pools.

Incubation experiment

To enhance reproducibility, six soil samples were sieved through a 1-mm stainless steel sieve. Aliquots of 10.0 g of air-dried soil were transferred to 50-mL centrifuge tubes and amended with 1 mL of 0.01% or 1.0% purified HA or FA solutions, with untreated soils serving as controls. Soil moisture was adjusted to 40% of the saturation percentage (SP) by adding distilled water. Tubes were sealed with perforated parafilm to maintain aerobic conditions and incubated at 25 ± 0.5 °C for 15 days. Moisture content was monitored every 3 days by weight and replenished as needed. Post-incubation, plant-available Fe, Mn, Zn, and Cu were extracted using a 0.005 M DTPA, 0.1 M TEA, and 0.01 M CaCl2 solution buffered at pH 7.319. Each tube received 20 mL of the extracting solution, followed by shaking at 180 rpm for 2 h at 25 °C. Samples were centrifuged at 3000 rpm for 15 min, and supernatants were filtered through Whatman No. 42 filter paper. Micronutrient concentrations were quantified via flame atomic absorption spectrometry (PerkinElmer 1100).

Statistical analysis

All experiments were conducted in duplicate, with means derived from three independent replicates. Statistical analyses were performed using SAS software (SAS Institute Inc., Cary, NC, USA). Treatment means were compared using Tukey’s multiple range test, a post hoc analysis applied following ANOVA to assess significant differences at p < 0.05. Pearson correlation analysis was also conducted to assess linear relationships between structural indices and Mn availability, with statistical significance set at p < 0.05 (n = 6 soils).

Results

Soil characteristics

The physicochemical properties of the six soil samples are summarized in Table 1. Soils exhibited textures ranging from clay to loamy sand, with five classified as non-saline (EC < 4 dS/m) and one (Soil 5) displaying moderate salinity (EC = 5.81 dS/m). All soils were calcareous, with calcium carbonate equivalents (CCE) spanning 10.3–37.2%, and organic carbon (OC) content varying widely from 0.03% (Soil 2) to 2.46% (Soil 5). Soil pH was alkaline (7.7–8.1), consistent with calcareous substrates. Saturation percentage (SP), reflecting water-holding capacity, ranged from 28.7% (Soil 2) to 60.5% (Soil 5). Available micronutrient concentrations varied significantly across soils, with Fe (0.97–8.01 mg/kg) and Zn (0.09–4.30 mg/kg) showing the most pronounced variability.

Table 1 Some physical and chemical properties of six soil samples selected for the research.
Full size table

Characteristics of the purified commercial HA and FA samples

To focus on molecular properties most relevant to micronutrient dynamics, elemental ratios (H/C and O/C) and NMR-derived structural indices were consolidated (Table 2). Alkyl/O-alkyl ratio (A/OA) and aromaticity (ARM) were prioritized as they emerged as primary predictors of FA and HA efficacy, respectively (Section "Effects of FAs and HAs on Mn availability in soils"). Hydrophobicity (HB/HI) and O/C ratios provide complementary insights into metal binding behavior. Detailed carbon distribution data are available in Supplementary Table S1.

Table 2 Key molecular properties of purified humic (HA) and fulvic (FA) acids.
Full size table

Elemental composition

Elemental composition and atomic ratios (H/C and O/C) of purified HA and FA samples are summarized in Table 2 and Table S1. Fulvic acids exhibited lower average nitrogen (N: 0.84% vs. 1.60%) and oxygen (O: 29.0% vs. 36.7%) content compared to humic acids, but higher carbon (C: 58.7% vs. 56.8%), hydrogen (H: 5.0% vs. 3.2%), and sulfur (S: 6.3% vs. 1.7%). The H/C atomic ratio, a proxy for aromaticity, revealed that all HA and FA samples—except FA4 (H/C = 1.12)—displayed values ≤ 1.0, indicating aromatic-dominated structures. HA samples exhibited lower average H/C ratios (0.67 vs. 1.01 for FA), suggesting a more condensed aromatic framework in HA. Conversely, HA showed higher O/C ratios (0.49 vs. 0.37 for FA), reflecting greater oxygen-containing functional groups (e.g., carboxyl and hydroxyl) in their molecular architecture.

FTIR spectra analysis

The FTIR spectra of purified FAs and HAs are presented in Fig. 1. All FA samples exhibited nearly identical absorption bands (Fig. 1A). A broad peak at ~ 3400 cm⁻1 corresponded to O–H stretching vibrations from carboxylic acids, phenols, and alcohols. Bands at ~ 2940 cm⁻1 and 2845 cm⁻1 were assigned to asymmetric and symmetric aliphatic C–H stretching in CH2 and CH3 groups. A medium-intensity absorption at 1725–1716 cm⁻1 indicated C = O stretching of carboxyl groups (COOH), while the peak at ~ 1600 cm⁻1 arose from aromatic C = C stretching. A distinct signal at 1516 cm⁻1 was attributed to overlapping N–H bending (amide II) and aromatic C = C vibrations. Weak bands at 1460–1456 cm⁻1 and 1420 cm⁻1 corresponded to aliphatic C–H deformation and C–N stretching of primary amides (amide III), respectively. Additional absorptions in the 1260–1200 cm⁻1 region were linked to C–O stretching and OH deformation of COOH, as well as aryl ethers and phenols. A prominent band at 1039 cm⁻1 suggested C–O stretching in polysaccharide-like structures, while weaker peaks between 930–660 cm⁻1 reflected aromatic C–H out-of-plane bending.

Fig. 1
figure 1

FTIR spectra of the purified commercial FA (A) and HA (B) samples.

Full size image

All HA samples displayed highly similar FTIR profiles (Fig. 1B), with minor variations in relative band intensities. Key absorptions included: a broad O–H stretch at ~ 3400 cm⁻1, aliphatic C–H stretching at 2933–2926 cm⁻1, C = O stretching (COOH/ketones) at 1728–1710 cm⁻1, and overlapping aromatic C = C/H-bonded C = O vibrations at 1620–1600 cm⁻1. Aliphatic C–H deformation appeared at 1444–1440 cm⁻1. Bands at 1253–1236 cm⁻1 were associated with C–O stretching and OH deformation of COOH. A distinct C–O polysaccharide-like stretch at 1035 cm⁻1 was observed exclusively in HA2 and HA4, while a weak peak at 911 cm⁻1 (aromatic C–H out-of-plane bending) was unique to HA1.

Acidic functional groups

The acidic functional group contents of purified FA and HA samples are summarized in Table 3. Significant differences in functional group quantities were observed between FA and HA fractions. HA samples exhibited higher average carboxyl group (COOH) content (5.27 meq/g) compared to FAs (3.49 meq/g), while phenolic OH group levels were comparable across both fractions. Consequently, HA samples demonstrated greater total acidity than FAs. The COOH/phenolic OH ratio revealed phenolic OH dominance over COOH groups in all samples except HA1 (ratio = 1.35). FA samples showed particularly low ratios (0.27–0.82), with FA4 exhibiting the smallest proportion of COOH relative to phenolic groups (0.27). In contrast, HA samples displayed moderately higher ratios (0.51–1.35), though still reflecting phenolic OH prevalence in most cases.

Table 3 Amount of functional groups of the purified FA and HA samples.
Full size table

13C NMR spectra

Solid-state 13C CPMAS-NMR spectra of purified FA and HA revealed distinct structural features (Fig. 2A,B and Table 2). FA samples exhibited broadly similar profiles, though FA1 and FA5 displayed reduced aromatic content (145–111 ppm) with weak lignin-derived guaiacyl/syringyl signals (113–131 ppm) and prominent methoxyl (56 ppm) and carbohydrate-associated peaks (66–107 ppm). All FAs lacked detectable carboxyl/carbonyl carbons (160–199 ppm), while aliphatic resonances (0–45 ppm) suggested minor alkyl contributions. In contrast, HA spectra were homogeneous except for HA1, which showed dominant aromaticity (88.2% in 145–111 ppm) and negligible alkyl content. Structural indices highlighted differences: FA1 and FA5 exhibited elevated alkyl/O-alkyl ratios (A/OA: 1.14–1.16) and lignin ratios (LR: 2.11–2.72), indicating hydrophobic stabilization and lignin-derived phenolics, whereas FA2–FA3 (LR: 1.49–1.60) suggested non-lignin phenolic sources. HA1 displayed exceptional aromaticity (ARM: 7.45 vs. 1.31–1.57 in other HAs), while HAs universally exhibited higher hydrophobicity (HB/HI: 7.45–18.31) than FAs (1.64–1.95), underscoring their divergent molecular architectures.

Fig. 2
figure 2

Solid-state 13C CPMAS-NMR spectra of the purified FA (A) and HA (B) samples.

Full size image

Effects of FAs and HAs on Mn availability in soils

Notably, FA1 and FA5 exhibited distinct molecular architectures predictive of superior metal mobilization. Both samples displayed elevated alkyl/O-alkyl ratios (A/OA: 1.14–1.16), reduced aromaticity (ARM: 0.71–0.80), and high lignin ratios (LR: 2.11–2.72) (Table 2), indicating hydrophobic, lignin-derived domains rather than acidic functional group abundance (Table 3) as the primary drivers of efficacy. As demonstrated below, this structural profile correlated strongly with enhanced micronutrient release. The influence of purified FA and HA on Mn availability is summarized in Fig. 3. At the lower application rate (12.5 mg/kg), FAs did not significantly enhance Mn release, except for FA1 in Soil 1. However, at 1250 mg/kg, all FA treatments increased DTPA-extractable Mn by 26–3061% across soils (p < 0.05), with FA1 and FA5 exhibiting the highest efficacy and FA2 the lowest. Conversely, Mn release showed no correlation with acidic functional group content (Table 3). For instance, despite FA4 having nearly triple the total acidity of FA3 (15.43 vs. 5.78 meq/g), both elicited similar Mn availability. Notably, FA5, with double the COOH groups of FA1 (6.03 vs. 3.27 meq/g), released less Mn in most soils, underscoring the limited role of carboxyl groups in Mn chelation. HA treatments at 12.5 mg/kg had no significant effect, while 1250 mg/kg enhanced Mn availability in four soils (p < 0.05). In Soil 1, HA2 and HA5 increased Mn by 48% and 38%, respectively; in Soil 2, HA3 and HA4 raised levels by 108% and 117%; and in Soil 5, HA1 and HA5 boosted Mn by ~ 33%. Despite these increases, FAs universally outperformed HAs, particularly in Mn-deficient soils (Soils 3 and 6), where FA applications elevated Mn availability by 786–3061% compared to controls. These results highlight the superior chelating capacity of FAs, likely linked to their hydrophobic aliphatic and lignin-derived structures, rather than functional group density or acidity.

Fig. 3
figure 3

Impact of the purified commercial FAs and HAs on the availability of Mn in the six studied soils. Values with different superscript letters differ significantly (p < 0.05, Tukey’s multiple range test).

Full size image

To identify molecular drivers of FA efficacy, Pearson correlations were computed between FA characteristics and DTPA-extractable Mn across soils at the high application rate (1250 mg/kg) (Table 4). Key patterns emerged: Alkyl/O-alkyl ratio (A/OA) showed strong positive correlations with Mn availability in 5 of 6 soils (r = 0.60–0.72; p < 0.05 in Soils 1, 3, 6), confirming hydrophobic domains promote Mn mobilization. Aromaticity (ARM) exhibited significant negative correlations (r =  − 0.46 to − 0.68; p < 0.05 in Soils 1, 3, 6), indicating reduced efficacy in highly aromatic FAs. Acidic functional groups (COOH, phenolic OH) showed no consistent relationships (|r|≤ 0.34; p > 0.05), underscoring their limited predictive value. Soil-specific exceptions were observed. In Soil 5 (high OC and salinity), inverse trends emerged, with C content (r = 0.58) and O/C ratio (r = 0.54) emerging as dominant factors. These correlations validate that FA efficacy for Mn mobilization is primarily governed by hydrophobic aliphatic structures (A/OA) rather than acidic group abundance, with some soil-specific variations.

Table 4 Correlation between humic substances and available Mn in the soils treated with fulvic acid at a rate of 1250 mg/kg.
Full size table

Effects of FAs and HAs on Zn availability in soils

Consistent with trends observed for Mn (Section "Effects of FAs and HAs on Mn availability in soils"), Zn availability responded selectively to FA samples with hydrophobic lignin-derived architectures (FA1 and FA5). At the lower application rate (12.5 mg/kg), neither FAs nor HAs significantly altered DTPA-extractable Zn across all soils. However, at 1250 mg/kg, FA treatments significantly increased Zn availability in Mn-deficient soils 3 and 6 (p < 0.05; Fig. 4), with FA1 and FA5 demonstrating the greatest efficacy (FA5: + 8% in Soil 6; FA1: + 5% in Soil 3). HA applications improved Zn availability exclusively in these same soils for HA5 (p < 0.05). These results confirm that Zn release is enhanced by high doses of hydrophobic FA fractions in soils with inherently low micronutrient bioavailability, while limited responses elsewhere suggest saturation of Zn-binding sites or competitive soil matrix interactions.

Fig. 4
figure 4

Impact of the purified commercial FAs and HAs on the availability of Zn in six studied soils. Values with different superscript letters differ significantly (p < 0.05, Tukey’s multiple range test).

Full size image

Effects of FAs and HAs on Fe availability in soils

The influence of purified FA and HA on iron (Fe) availability in calcareous soils is detailed in Fig. 5. At the lower application rate (12.5 mg/kg), only FA1 and FA5 significantly increased DTPA-extractable Fe in Soil 2 (+ 32%), while HA3 and HA5 enhanced Fe availability in Soil 3 (p < 0.05). Conversely, all FA treatments reduced Fe availability in Soil 4 compared to controls. At the higher rate (1250 mg/kg), FAs significantly improved Fe availability in all soils except Soil 5, with FA1 emerging as the most effective: increases of 43% (Soil 1), 45% (Soil 2), 17% (Soil 3), and 9% (Soil 6) were observed. HAs at 1250 mg/kg elevated Fe availability in Soils 1, 2, and 5, with HA4 showing the highest efficacy (+ 17% in Soil 1, + 40% in Soil 2, and + 16% in Soil 5). Despite these enhancements, absolute increases in Fe availability remained modest across treatments, underscoring the limited solubilization capacity of both FAs and HAs in calcareous systems. These results highlight the dose- and sample-dependent effects of organic amendments on Fe mobilization, with hydrophobic FA fractions (e.g., FA1) exhibiting broader efficacy than HAs under high-application conditions.

Fig. 5
figure 5

Impact of the purified commercial FAs and HAs on the availability of Fe in six studied soils. Values with different superscript letters differ significantly (p < 0.05, Tukey’s multiple range test).

Full size image

Effects of FAs and HAs on Cu availability in soils

The impact of FA and HA on copper (Cu) availability in calcareous soils is summarized in Fig. 6. FA applications at both 12.5 and 1250 mg/kg had no significant effect on DTPA-extractable Cu in all soils except Soil 4. In contrast, HA treatments at 1250 mg/kg significantly reduced Cu availability (p < 0.05) in five soils, with suppression ranging from 3.6% to 35.5%. Soil-specific reductions included: 11–15% in Soil 1, 15–20% in Soil 2, 3.6–4.3% in Soil 3, 17.7–35.5% in Soil 4, and 6.7–9.6% in Soil 6. Soil 5 remained unaffected by HA treatments. Notably, HA1–HA5 exhibited consistent Cu-suppressive effects, with HA3 and HA4 causing the largest reductions in Soil 4 (32.3–35.5%). These results suggest that high-dose HA applications may immobilize Cu through ligand competition or adsorption processes, particularly in soils with elevated native Cu bioavailability. The lack of FA-induced Cu mobilization contrasts with their efficacy for Mn and Fe, underscoring metal-specific interactions in humic substance-metal dynamics.

Fig. 6
figure 6

Impact of the purified commercial FAs and HAs on the availability of Cu in six studied soils. Values with different superscript letters differ significantly (p < 0.05, Tukey’s multiple range test).

Full size image

FA/HA effects on micronutrient availability across six soils

As shown in Fig. 7, FA and HA treatments significantly affected only Mn availability, exhibiting a strong dose-dependent response. In contrast, Cu, Fe, and Zn availability remained largely unchanged compared to controls, regardless of treatment type or concentration. Mn availability increased significantly only at higher treatment concentrations; lower concentrations showed no effect. Among the FA treatments, only FA1 to FA5 significantly enhanced Mn availability, with FA1 demonstrating the strongest effect.

Fig. 7
figure 7

Impact of purified fulvic acid (FA) and humic acid (HA) on Mn, Fe, Cu, and Zn availability in six soils. Values with differing superscript letters are significantly different (p < 0.05). Error bars show ± 1 standard deviation (n = 12 replicates: 6 soils × 2 reps). Large error bars indicate high variability in micronutrient levels among the soils.

Full size image

Discussion

The efficacy of commercial humic substances (HS) in enhancing micronutrient availability in calcareous soils is intricately tied to their molecular architecture, application rate, and soil-specific mineralogy, as demonstrated by this study’s integration of structural characterization and incubation experiments. While fulvic acids (FAs) such as FA1 and FA5 significantly increased manganese (Mn) availability at high application rates (1,250 mg/kg), with improvements ranging from 87% in Soil 4 to 3,061% in Mn-deficient Soil 6, these effects were absent at agronomically relevant doses (12.5 mg/kg), underscoring the dose dependency noted by Gollenbeek and van der Weide20. The Mn mobilization mechanism likely involves competitive displacement by DTPA, as FA1 and FA5 formed labile Mn complexes that readily released Mn2⁺ upon DTPA addition, a phenomenon consistent with Essington21 observations on the thermodynamic superiority of DTPA-metal complexes. These complexes also arise from the structural properties of FA samples revealed by FTIR and CPMAS NMR: depolymerized lignocellulose fragments enriched in phenolic groups (FTIR: 1516 cm⁻1; NMR: 145–160 ppm) and carboxylic functions (FTIR: 1725–1716 cm⁻1). These molecular components facilitate reductive dissolution of Mn oxides via phenolic-mediated electron transfer and complexation of Mn2⁺ through carboxylate ligands, enabling dynamic interactions across variables like cation:OM ratio, pH, and soil mineralogy3,22. Crucially, FA1 and FA5’s efficacy correlated not with carboxyl or phenolic group density—contrary to assumptions by Ampong, et al.8—but with hydrophobic stabilization, evidenced by their elevated alkyl/O-alkyl ratios (A/OA: 1.14–1.16) and reduced aromaticity (ARM: 0.71–0.80), structural traits linked to biochemical stability by Monda, et al.11. This dominance of hydrophobic/lignin domains over acidic groups can be attributed to the micelle-like behavior of dissolved humic substances, where alkyl chains and lignin fragments mediate conformational rigidity of hydrophilic moieties (e.g., polysaccharides), enhancing pliability and surfactant properties of FA aggregates22,23,24,25. In dissolved humic materials, lignin functional groups play a critical role in metal interactions, as evidenced by NMR spectra showing FA1/FA5 dominance in lignin fragments and polysaccharide derivatives (Fig. 2, Table 2). The high lignin ratios (LR: 2.11–2.72) confirm that phenolic groups primarily originate from lignin units, while carboxylic groups partly derive from oxidized propyl-side chains of depolymerized lignin. Thus, lignin—not isolated acidic groups—is the central driver of binding capacity, enabling dual Mn-activation mechanisms: (1) hydrophobic stabilization of labile Mn complexes via micellar encapsulation, and (2) phenolic-mediated reductive dissolution of Mn oxides. This aligns with the reductive dissolution of Mn(III/IV) oxides proposed by Sparks et al. (2022), where phenolic groups in FAs facilitated Mn3⁺/4⁺ reduction to soluble Mn2⁺, particularly in low-organic carbon, CaCO3-rich soils (Soils 1–3). However, Soil 5’s high organic-bound Mn resisted solubilization, highlighting speciation-dependent responses, while Soil 4’s kaolinite content suppressed Mn release via HS-clay adsorption, as noted by Singh et al.26. Iron (Fe) availability showed modest, soil-specific improvements (e.g., 45% in Soil 2 for FA1), constrained by calcareous conditions that promoted Fe2⁺ re-precipitation as Fe(OH)3 or FeCO3, a limitation echoed in Jones et al.27. HA4’s Fe mobilization in Soils 1–2 and 5 (16–40%) was attributed to phenolic-mediated Fe3⁺ reduction28, yet HA’s high aromaticity (ARM: 1.31–7.45) fostered insoluble Fe-HA complexes, consistent with Stevenson29 findings on HA-metal precipitation. Zinc (Zn) remained unresponsive to HS, even at high doses, due to its 3d1⁰ electronic configuration, which impedes stable complexation30, while clay and CaCO3 competition for Zn2⁺ adsorption (Soil 2: 86% sand) further limited bioavailability, contrasting with Halim et al.31, who reported Zn mobilization in artificially enriched soils. Copper (Cu) availability paradoxically decreased with HA applications (11–35% in Soil 4), attributable to strong Cu-HA-clay ternary complexes that resisted DTPA extraction. This immobilization stems from HA’s molecular architecture: condensed aromatic moieties (NMR: 110–160 ppm; ARM: 1.31–7.45) create an extended network of shared π-electrons (FTIR: 1600 cm⁻1), which provide a large active surface for covalent Cu2⁺ binding. Despite lower acidic functional group density, these hydrophobic domains enable stable complexation via charge transfer and cation-π interactions, overriding solubilization effects22,29. As Eshwar et al.10 demonstrated, Cu exhibits strong covalent bonding with HA aromatic rings and amine groups, stabilizing Cu-HA complexes. This stabilization was further amplified in Soil 4’s kaolinite-rich matrix, where ligand exchange—highlighted by32 as a dominant mechanism in clay-rich soils—enhanced Cu adsorption onto HA. These findings align with Stevenson29 observations on synergistic interactions between humic substances (HS) and clay minerals. However, they contrast starkly with Bianchi et al.33, who documented Cu mobilization in marine sediments, underscoring how environmental conditions (e.g., soil mineralogy vs. aqueous sediment dynamics) critically influence Cu speciation and mobility. Structural indices—notably A/OA (alkyl/O-alkyl) and ARM (aromaticity)—emerged as superior predictors of HS efficacy over functional group metrics. While FTIR/NMR confirmed acidic group abundance (e.g., COOH at 1725 cm⁻1; phenolic OH at 3400 cm⁻1), their role in nutrient mobilization is governed by synergistic interactions with molecular architecture: alkyl domains in FA enhance Mn/Fe chelation through hydrophobic stabilization and conformational flexibility, whereas HA’s aromatic cores sequester Cu via π-electron networks22,23. This mechanistic distinction explains why FA4’s high total acidity (15.43 meq/g) failed to mobilize Mn, while FA1’s lignin-derived alkyl chains (NMR: 0–45 ppm; A/OA: 1.16) achieved 3,061% mobilization. Similarly, HA’s hydrophobicity (HB/HI: 7.45–18.31) and aromaticity drove Cu/Fe sequestration. These findings fundamentally challenge the carboxyl-centric model of HS chelation, advocating instead for prioritization of structural motifs (alkyl domains for mobilization; aromatic networks for immobilization) in HS design. Practical applications include: (1) deploying hydrophobic FAs (e.g., FA1/FA5) in sandy, Mn-deficient soils, and (2) developing soluble Fe-HS complexes for clay-rich systems. Future research should employ isotopic tracers to resolve DTPA extraction artifacts and optimize nanoformulations for field viability. Collectively, this study redefines HS efficacy as a function of molecular architecture–soil context interactions, urging a paradigm shift from functional group quantification to structural engineering for sustainable micronutrient management.

Conclusions

This study challenges the conventional paradigm that functional group density governs humic substance (HS) efficacy in calcareous soils, demonstrating instead that molecular architecture—specifically hydrophobicity and lignin-derived alkyl domains—dictates micronutrient mobilization. Our hypotheses were partially validated: while fulvic acids (FA) outperformed humic acids (HA) in enhancing manganese (Mn) availability at elevated rates (1,250 mg/kg), this superiority stemmed not from carboxyl/phenolic group abundance but from hydrophobic stabilization and structural flexibility. Conversely, HA reduced copper (Cu) availability via stable Cu-HA-clay complexes, underscoring their role as metal immobilizers. Zinc (Zn) and iron (Fe) dynamics were largely unresponsive to HS, constrained by Zn’s electronic inertia (3d1⁰ configuration) and Fe’s redox re-precipitation in alkaline matrices. Critically, soil mineralogy overrode HS effects; Mn mobilization succeeded in sandy, CaCO3-rich soils but failed in clay-rich or organic-bound systems. These findings refute the assumption that HS efficacy can be predicted by functional group metrics alone, advocating instead for molecular design prioritizing alkyl-aromatic balance.

The selective efficacy of hydrophobic FA in Mn-deficient calcareous soils positions these fractions as precision tools for regions grappling with Mn scarcity, such as arid zones with alkaline soils. Conversely, HA’s Cu-suppressive capacity could mitigate Cu toxicity in contaminated soils, aligning with sustainable heavy metal management strategies. However, the impracticality of high HS doses (1,250 mg/kg) necessitates nanoformulations or foliar delivery to enhance field viability. Policymakers and agribusinesses must shift from marketing HS as “universal amendments” to certifying products based on structural indices (e.g., alkyl/O-alkyl ratios), ensuring transparency and efficacy.