Activated Biochar-Based Organomineral Fertilizer Delays Nitrogen Release and Reduces N₂O Emission

Abstract

Leaching and nitrous oxide (N₂O) emissions can represent substantial nitrogen (N) losses from chemical fertilizers, and slow-release fertilizers (SRFs) can mitigate these effects. Thus, biochar can be an alternative from an agronomic and environmental point of view to synthesize SRFs due to its physicochemical characteristics. We investigated the effect of nitrogenous organomineral fertilizers (OMF-N) formulated based on activated biochar on N losses by leaching and N₂O emissions. The OMF-N were developed from a dry mechanical pelleting process with different biochar and urea proportions (2:1; 1:2, and 1:4). Three experiments were conducted using four fertilizer sources (urea, OMF-N 2:1, OMF-N 1:2, and OMF-N 1:4): i. to analyze the kinetics of N release from OMF-N at times: 5, 15, 30, 60, 90, and 120 min; ii. for N₂O emission analysis determined at 3, 6, 10, 14, 24, 44, 54, 64, 74, 84, 104, and 118 days after application to the soil; and iii. for a double factorial design that was adopted to analyze N leaching, consisting of the combination of applying 160 kg N ha⁻¹ of fertilizers in PVC columns at different depths (20, 40, 60, and 80 cm) and analyzed at five times (1, 7, 14, 21, and 28 days). FTIR spectroscopic analysis, specific surface area, porosity, and surface morphology showed physicochemical interactions of N of the OMF with biochar; the N from the OMF interacts physically and chemically binds to the functional surfaces of biochar, delaying the dissolution flow. The OMF-N proved capable of retaining 48% to 60% more NH₄⁺ and reduced the release of N_total from urea from 27% to 60%, as well as reduced N₂O emissions from 47% to 66%. Although absolute CO₂ emissions intensified with the application of OMF-N, its use provides C sequestration in the soil to due to the recalcitrant C of the biochar, which results in a positive input-output balance in the system. The NO₃⁻ concentration profiles revealed that the OMF-N application was able to reduce leaching in the soil to a depth of 80 cm. These studies enabled better understanding of the processes involved in the biochar:urea interaction and revealed that biochar can be used as an organic matrix in the synthesis of SRF.

1. Introduction

With the widespread use of chemical fertilizers, negative impacts on the environment are receiving increasing attention worldwide, as the estimated loss of fertilizers to the environment is around 40–70% N, 80–90% P, and 50–70% K [1,2]. The indiscriminate use of nitrogen fertilizers has allowed successive gains in agricultural productivity; however, the NO₃⁻ leaching, NH₃ volatilization, and N₂O emission processes cause high N losses, resulting in low N utilization efficiency, as well as high economic costs and potential for environmental pollution [3].

Nutrient losses from chemical fertilizers can be minimized by using slow or controlled-release fertilizers (SRFs). These technologies have gained more and more global attention as they provide a slow release of nutrients without affecting the synchrony with the demand required by the crop, thus ensuring availability for plants and reducing losses in the environment [1].

SRFs are generally produced by coating the conventional fertilizer granule with polymers derived from fossil fuels (i.e., petroleum), such as polyethylene, polystyrene, and resins [4]. Although these materials have been shown to be useful to increase the efficiency of nutrient use and decrease losses, they can be toxic and non-biodegradable, in addition to having a high production cost [5].

In this context, biochar with its porous carbonaceous structure, high surface area, and abundant functional groups has been intensively studied as an adsorbent, and more recently as a promising organic matrix to produce SRFs [6,7]. The characteristics of biochar allow chemical elements which are essential to plants to be retained on its surface by physical and chemical adsorption. In addition, it is possible to increase the number of pores and functional groups on the biochar surface through the activation process to improve the interaction abilities with chemical elements of interest [8].

Biochar can be obtained from different raw materials, with emphasis on agro-industrial waste, and thus can present different physicochemical characteristics with objectives ranging from use as soil conditioner to pesticide adsorption, removal of pollutants such as toxic metals, greenhouse gas reduction, to fertilizer synthesis. [9,10,11,12,13,14]. In the case of fertilizers, depending on the type of biochar, the composite formed with the inorganic matrix of mineral fertilizers can result in interactions which have not yet been fully elucidated, and these interactions govern the response magnitude to the N release efficiency from these fertilizers. Thus, central questions need answers, such as: Would biochar from wood industry residues (environmental liabilities) constitute an organic matrix capable of controlling the N release flow from organomineral nitrogen fertilizers? If so, could controlling the N release flow from biochar-based organomineral nitrogen fertilizers reduce GHG emissions and soil N leaching?

In order to elucidate these questions, the study was guided by an agronomic and environmental perspective supported by the following basic hypotheses: (i) activated biochar reduces the flow of mineral N dissolution when used as an organic matrix; (ii) controlling the N release flow in an organomineral nitrogen fertilizer based on activated biochar reduces N₂O emissions; (iii) controlling the N release flow in an organomineral nitrogen fertilizer based on activated biochar reduces the leaching of NO₃⁻ in the soil; and (iv) application of an organomineral nitrogen fertilizer based on activated biochar has the potential to sequester C in the soil and thus fit into a regenerative agriculture model.

Therefore, the objective of this study was to investigate the effect of organomineral fertilizers (OMF-N) formulated with activated biochar on N losses in NO₃⁻ and NH₄⁺ leaching and on N₂O emissions.

2. Materials and Methods

2.1. Study Site

The experiments were carried out on the premises of Embrapa Agrossilvipastoril and in a greenhouse at the Federal University of Mato Grosso (UFMT) in Sinop, Mato Grosso, Brazil, from November 2020 to October 2021.

2.2. Activated Biochar Production

The biochar came from the slow pyrolysis of sawmill residues from native species obtained in the Sinop, MT region, produced in the pyrolysis shed of Embrapa Agrossilvipastoril, Sinop, Mato Grosso, Brazil. The batch reactor with the capacity to process about 9 kg of biomass per batch was adjusted to pyrolyze the sawdust at 450 °C with a retention time of 40 min and activation with the injection of water vapor (3 L of water) in the last 15 min. The physicochemical composition of activated biochar is shown in

Table 1. Physicochemical composition of activated biochar used in the experiment.

Elements	Unit	Concentration
Total Nitrogen (N)	g kg−1	3.30
Phosphorus (P2O5 citric acid)		0.44
Phosphorus (P2O5 total)		-
K2O		0.60
CaO		3.78
MgO		1.32
Sulfur (S)		0.14
Copper (Cu)	mg kg−1	10.0
Zinc (Zn)		9.0
Molybdenum (Mo)		1.35
Cobalt (Co)		1.93
Boron (B)		12.0
Total carbon (C)	g kg−1	760.0
Moisture		50.0
C:N ratio		230
Specific surface area (BET)	m2 g−1	537
Actual density	g cm−3	1.27
Apparent density	g cm−3	0.31
Porosity	%	75
Average pore diameter	Nm	2.62

.

The C, N, and S contents were determined by elemental analysis via the CNHS Vario TOC Cube analyzer. Nutrients were determined according to fertilizer analysis methodology [15]. The analysis of specific surface area and pore size was performed according to the physical adsorption method in a Quantachrome NOVA 4200e surface area analyzer.

2.3. Preparation of Organomineral Nitrogen Fertilizers (OMF-N) Based on Activated Biochar

The OMF-N were developed from the mixture of biochar and urea proportions (m:m) (2:1; 1:2, and 1:4) which resulted in the following N contents for the OMF-N: 16%, 32%, and 36%, respectively. The urea was initially placed in a rotor mill in the defined proportions and then the biochar and urea were uniformly mixed without the use of binders in a pelletizer (Biobrax PBX-100/05), which resulted in pellets with an average of 5.30 mm in length and 3.87 mm in diameter. The pellets were evaluated for hardness using a Texture Analyzer TA HD PLUS device (Stable Micro Systems) [16]. The results obtained corresponded to 17.6, 12, and 1.82 kgf, respectively, for the OMF-N 2:1, 1:2, and 1:4 formulations.

2.4. Fourier Transform Infrared (FTIR) Spectroscopic Analysis

OMF-N samples were submitted to FTIR spectroscopy in order to verify the physicochemical interaction of activated biochar with urea. To do so, an attenuated total reflectance (ATR) mid-IR spectrometer (Agilent Technologies Cary 630, Dansbury, CT, USA) was used. The Cary 630 laboratory spectrometer is interfaced with a custom three-reflection diamond ATR sampling accessory with a 1 mm diameter sampling surface and an active area of 200 μm. IR spectra were collected using the Agilent MicroLab PC software program (Agilent Technologies, Dansbury, CT, USA) in the spectral region between 4000 and 650 cm⁻¹. The resolution used was 8 cm⁻¹. Each spectrum was composed of the average of 64 samples (64 scans). The spectrometer crystal was cleaned with isopropanol (PA) at each sample application and a new background spectrum (I₀) background was performed before each sample application.

2.5. Specific Surface Area (SSA) and Pore Size and Distribution

The biochar and OMF-N samples were subjected to degassing and removal of other surface impurities through vacuum pre-treatment at 105 °C for 12 h. SSA was determined by nitrogen adsorption isotherms using the Brunauer–Emmett–Teller (BET) method [17]. The pore size and distribution were characterized using the theory of progressive pore emptying with decreasing pressure [18].

2.6. Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray Spectroscopy (EDS)

The surface morphology of the biochar and the OMF-N were characterized by scanning electron microscopy (SEM-JSM-6610-Jeol) equipped with an X-ray energy dispersive spectroscopic device (EDS-Thermo Scientific NSS) with an acceleration voltage of 10–15 kV and 1000 magnification.

2.7. OMF-N Solubilization Kinetics

The nitrogen solubilization kinetics in OMF-N was performed in triplicate for the following times: 5, 15, 30, 60, 90, and 120 min, stirring 1 g of the samples in 50 mL of water at 96 cycles per minute. After filtering the samples through a 45 μm filter, the total nitrogen was measured by the Total Organic Carbon (TOC) Analyzer in liquid mode (vario TOC cube model, Elementar Analysensysteme GmbH). The injection volume was adjusted to 0.2 mL of the sample diluted (1:1) with distilled water. The combustion column was maintained at 690 °C throughout the analysis and compounds measured by a near infrared detector (NDIR). A calibration curve (20 to 100 ppm of C) was performed with potassium biphthalate.

2.8. OMF-N N₂O and CO₂ Emission Kinetics

The N₂O and CO₂ emission kinetics were performed by incubating urea and OMF-N at 25 °C under field capacity, which is the optimal condition for most microbial processes to maximize microbial activity. Thus, 500 mg of urea and OMF-N were incubated in 1 × 1 mm open mesh screens and covered in 30 g of soil. The calculations were performed with due compensation for the N content of the fertilizer sources. CO₂ and N₂O were measured at 3, 6, 10, 14, 24, 44, 54, 64, 74, 84, 104, and 118 days in gas chromatography equipped with a flame ionization detector (FID) for CO₂ and an electron capture detector for N₂O, both with column and injector. Ultrapure nitrogen was used as carrier gas at an inlet pressure of 300 kPa (40 psi) and as detector make-up gas at a flow rate of 25 mL min⁻¹.

2.9. NH₄⁺ and NO₃⁻ Leaching in Soil Fertilized with Urea and OMF-N

Soil was collected on a rural property in the municipality of Sinop, Mato Grosso, at a depth of 0 to 20 cm. The soil was classified as Dystrophic Red Yellow Latosol, medium texture (Clay: 258 g kg⁻¹; Silt: 50 g kg⁻¹ and Sand: 692 g kg⁻¹), with the following chemical characteristics: pH (H₂O): 5.8; P: 64.4 mg/dm³ (Mehlich method); K⁺: 70.60 mg/dm³; Ca²⁺: 3.22 cmolc/dm³; Mg²⁺: 1.10 cmolc/dm³; Al³⁺: 0.00 cmolc/dm³; H + Al: 3.40 cmolc/dm³; V%: 57; CEC: 7.9 cmol/dm³; O.M.: 3.02 g/dm³; Fe: 114 mg/dm³; Mn: 8.8 mg/dm³; Zn: 3.7 mg/dm³; and Cu: 0.9 mg/dm³. The deformed soil samples were taken to a greenhouse, air dried, sieved, and homogenized. The columns to simulate soil depths (20, 40, 60, and 80 cm) were composed by the superposition of four PVC rings, 15 cm in diameter and 20 cm in height, joined by adhesive tape. A hole was made in the lower portion of each ring to insert the hoses, with removable covers, to collect the leaching water. Each soil column was filled with 14 kg of soil, and they were irrigated after 24 h to reach the maximum water holding capacity. Fertilization was performed 24 h after irrigation, making a superficial ring in the soil, applying the fertilizers evenly, and then covered by a thin layer of soil. Leaching water collections took place on days 1, 7, 14, 21, and 28 days after fertilization. Next, 2 L of artesian well water was added to the surface of the column on the day of each collection, while 1 L of water was added every 3 days in the days between collections. The leached water was collected from each depth of the columns in properly labeled sterile pots with approximately 40 mL. The volumes were filtered on quantitative black belt filter paper, then 3 drops of chloroform were added and stored in a cooler with ice. The pots were placed in a refrigerator after each daily collection for further analysis. The samples were submitted to centrifugation in the laboratory in Falcon tubes with 14 mL aliquots for 15 min at 4000 rpm. The spectrophotometry technique was used for the NO₃⁻ readings in the wavelengths of 220 and 275 nm, without adding reagents and with a 1:4 mL dilution (sample: distilled water) of the aliquots. The Berthelot reaction was performed with sodium salicylate to determine NH₄⁺, also by spectrophotometry at a wavelength of 647 nm. The NO₃⁻ and NH₄⁺ readings were taken using a Global Analyzer UV-VISIBLE spectrophotometer, according to the spectrophotometric method [19].

2.10. Statistical Analyses

The descriptive analysis of the data estimated the mean, median, minimum and maximum values, coefficient of variation, kurtosis, and data asymmetry. Data normality was verified by the Shapiro–Wilk test [20]. There was a need to remove some outliers from the data set, however, respecting the limit of 5% of the total observations. Values of three times the interquartile range in the box plot chart were used to define outliers [21]. Data analysis of variance was performed, followed by the F-test, with means compared by Tukey’s test at 5% error probability for significant qualitative data, and quantitative data were analyzed by regression by the Student’s t-test when significant. Finally, the standard error of the solubilization and emission coefficients of the respective models generated by the properly fitted curves was used to differentiate the regression analyses in the solubilization and emission kinetics studies.

3. Results and Discussion

3.1. Physicochemical Interaction of OMF-N

The FTIR spectra of OMF-N showed similar characteristics to those of urea (Figure 1), with peaks associated with stretching and deformation of N-H at 3430 cm⁻¹ and 1148 cm⁻¹, respectively, and N-H out of plane at 786 cm⁻¹ [22,23]. The stretching frequency of C=O appears at 1677 cm⁻¹ and 1589 cm⁻¹ and C-N stretching at 1459 cm⁻¹ and 1002 cm⁻¹ [22,24].

The spectra associated with urea and OMF-N clearly reveal that urea overlapped the biochar’s organic matrix peaks without major changes in the OMF-N spectra, suggesting that the main interaction of urea with the OMF-N biochar occurs physically. However, it is possible to verify that bands inherent to the N-H stretches in the region of 3430 cm⁻¹ and 1148 cm⁻¹ were weakened, as well as C=O in the region of 1587 cm⁻¹, suggesting that there is interaction, even if weak, via hydrogen bonds between the amino (NH₂) and carbonyl (C=O) groups of urea with the carboxylic (-COOH) or hydroxyl (-OH) groups of biochar. These observations are confirmed by the suppression of bands inherent to the carboxylic and hydroxyl groups of biochar (dashed lines) after interaction with urea. These results corroborate the interactions for urea-functionalized biochar and for a physical mixture of urea, mineral, and biochar [14,25].

In view of this, it seems evident to us that the mechanisms which govern the interaction between urea and biochar in the dry pelletization process occur physically (with greater intensity) and chemically (less intensity). The physical interaction would mainly occur through urea deposition in hydrophobic partition zones, but especially in the innermost pores of the organic matrix where the partial oxidation of the polycondensed aromatic structures of the biochar is lower [26]. On the other hand, the chemical interaction would occur at the outer edges of the peripheral polycondensed aromatic structures of the biochar, a region that is subject to greater partial oxidation and consequently greater presence of reactive functional groups [27]. This interaction seems to be favored by the high temperature observed during the pelleting process which reached values between 92 and 95 °C.

3.2. Physical Properties of the OMF-N

The specific surface area and porosity of biochar and OMF-N 2:1, 1:2, and 1:4 investigated through N₂ adsorption-desorption isotherms reveal the presence of mesoporous structures, characteristic of isotherms of the type IV (IUPAC). The presence of hysteresis at relatively low pressures is noteworthy (Figure 2a), indicating spontaneous filling of the mesopores due to capillary density [12]. The hystereses verified are similar to the H3 type, characterized by parallel loops typical of mesoporous materials consisting of non-uniform pores regarding their shape and size.

The pore size distribution curves of biochar and OMF-N reveal that although micropores are present, most are characterized by mesopores with a diameter ranging from 2 to 22 nm (Figure 2b). The results also reveal that because there is a reduction in the biochar-urea proportion, there is a drastic reduction in the pore volume, suggesting that the biochar mesopores were filled with urea. Additionally, the ASE of OMF-N 2:1 (17.6 m² g⁻¹), 1:2 (13.5 m² g⁻¹), and 1:4 (10.2 m² g⁻¹) were reduced by 96.7%, 97.4%, and 98.1% compared to biochar ASE (537 m² g⁻¹), thus confirming that urea was inserted into the mesopores and exposed external cavities of biochar, corroborating results that showed a 98% ASE reduction in fertilizer prepared from the hydrothermal synthesis of biochar-bentonite-urea [23].

3.3. Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray Spectroscopy (EDS)

The morphological characterization of the biochar surface revealed a porous structure with a planar texture (Figure 3a,b), while the OMF-N surfaces were characterized by filling exposed pores and cavities of the biochar by urea, presenting a heterogeneous, irregular, and rough structure (Figure 3d,e,g,h,j,k). This type of structure can control water permeation and hence nitrogen release.

The irregular heterogeneous textural aspect contributed to form external cavities to the pores, which can result in forming partition zones (mainly hydrophobic) due to the molecular nature of the biochar. Urea was inserted in such external cavities during the OMF-N synthesis process, which in turn would explain the physical interactions observed in FTIR spectroscopic analyses. Although hygroscopic in nature, the urea contained in the hydrophobic biochar zones may have its dissolution flow retarded, making the N release a gradual process.

In addition, a semi-quantitative analysis of the atomic composition of C, N, O, H, and S was simultaneously performed with the morphological characterization analysis of the biochar surface and the OMF-N (Figure 3c,f,i,l). Surface atomic mapping demonstrated N concentrations below the total elemental concentrations determined via the CHNS elemental analyzer. This is due to the N concentrations filled in pore spaces not determined in the surface analysis via EDS. These results clearly show that urea is retained on both the external and internal surfaces of the biochar, reinforcing the presence of physical interactions between the biochar and urea.

3.4. NH₄⁺ and N_total Release

The NH₄⁺ release from fertilizers is characterized by kinetics which show higher levels from urea (7755 mg kg⁻¹), while the lowest release came from OMF-N 2:1 (3045 mg kg⁻¹) (Figure 4a). It is known that the aqueous solution adsorption in the biochar follows a combination of physical and chemical processes, decreasing the nutrient release rate [28]. The quantitative delay observed in the release of NH₄⁺ from OMF-N was 60.7%, 54.0%, and 48.3% for the 2:1, 1:2, and 1:4 formulations, respectively, compared to conventional urea. The N based on biochar promotes lower losses by volatilization of NH₃⁺ and N₂O.

The adsorption of NH₄⁺ on the acidic functional groups of biochar (-COO-; -OH) would largely explain the delay in its release, as verified in other studies [14,29,30]. Furthermore, interactions can be weak and NH₄–N can easily exchange with other cationic species contained in biochar ash [31]. These values are expressive, as NH₄⁺ is one of the first products of the urea reaction in aqueous medium from ((NH₄)₂CO₃), which, given its chemical instability, is a primary source for the synthesis and consequent emission of NH₃⁺ and N₂O [32,33,34].

Thus, it is expected that the lower NH₄⁺ release from biochar based OMFs will promote lower losses by NH₃⁺ volatilization and N₂O emission. In this sense, we propose a mechanism for controlled N release based on the diffusive flow delay as a function of the chemical and physical adsorption of NH₄⁺ in the OMF-N pellets (Figure 5), which in theory would also delay the subsequent processes, reducing potential NH₃⁺ and N₂O losses.

The Weibull kinetic model best expressed the N_total release profiles (Figure 4b), thus corroborating the studies for urea-based fertilizers [35]. Although it is considered an empirical model, the Weibull model is currently suitable to apply to dissolution and controlled release processes [36]. In principle, due to the non-physical nature of its parameters, the Weibull model had its use restricted in kinetic applications; however, the constant b of the Weibull equation can also be considered an indication of the transport mechanism in a matrix [36,37].

The “b” values of the Weibull equation seem to demonstrate that the N_total release from the OMF-N is partly dependent on diffusion processes (Fickian model), and that it increases with the biochar concentrations in the OMF-N (reduction of b-values). However, one should not consider the constant diffusivity as in the Higuchi model (parabolic diffusion), because biochar’s organic matrix does not seem to provide some events which, in theory, are considered in the parabolic diffusion model, e.g., the diffusion in a single direction, negligible edge effect, perfect dissipation, and electrochemical interaction [38,39].

That said, the “a” parameter of the Weibull model for the kinetic profiles reveals higher and lower N_total release from urea and OMF-N 2:1, respectively. The N_total release after 120 min was 41% for urea and 30%, 29%, and 16% for the OMF-N 1:2, 1:4, and 2:1 formulations, respectively. This means that the OMF-N of 1:2, 1:4, and 2:1 formulation reduced the N_total release by ~27%, 29%, and 60%, respectively, in the first 120 min. It seems evident to us that the N dissolution flow in the biochar organic matrix is anchored in two main effects: i. in the physicochemical interaction between urea and the biochar organic matrix as already reported; and ii. on the physical characteristic of the organic matrix after the formulation process.

The physicochemical interaction of urea with the biochar organic matrix would justify the delay in the N_total release from the OMF-N due to the forces of hydrogen bonds and N-C=O electrostatic interactions and functional groups containing oxygen which would delay the urea release into the solution (adsorption) and by the physical protection that the pores (2.6 nm in diameter, located more internally in the biochar organic matrix) exert on the urea. On the other hand, the delay in the N_total release from the OMF-N due to the physical characteristic of the biochar organic matrix after the synthesis process is justified by the increase in compaction of the formulations since the reduction was proportional to the increase in the pellet hardness (compaction). There is the formation of empty space in the pores after urea solubility, which in theory could result in cracks in the organic matrix [35]. However, the OMF-N did not undergo a crack propagation process as they are composed of non-expandable organic matrices, which is a common process in expandable organic matrices [40]. The increased compaction of the OMF-N formulations resulted in a smaller contact surface with water, reducing the urea solubility and the nutrient mass transport to the external solvent phase.

3.5. N₂O and CO₂ Emissions

The N₂O emissions reveal similar temporal variation patterns in all treatments throughout the experiment (Figure 6). The N₂O flows were fitted to the sigmoidal model with two distinct phases which reveal an increase in flows: one after the application of fertilizers, and another at 58 days. The highest flows in the first days after application were verified for the urea treatment, while the highest flows after 58 days were verified in the OMF-N 2:1 treatment. Cumulative N₂O emissions from OMF-N 1:4 (0.51 mg kg⁻¹) and 1:2 (0.33 mg kg⁻¹) were significantly lower than treatment with urea (0.98 mg kg⁻¹). This means reductions of 47% and 66% in N₂O emissions from these fertilizers, respectively, corroborating other studies in laboratory and field conditions with the use of biochar and nitrogen fertilizer with a biochar base [41,42].

The higher N₂O emissions in the first days are related to the higher solubility of urea present in the peripheral spaces of the biochar. This is due to faster penetration of the liquid into these spaces, providing greater displacement of the nutrient mass via diffusion to the solid-liquid interface. On the other hand, the lower N₂O emissions from the OMF-N in the first days are possibly related to the lower urea solubility inside the pellets provided by the physicochemical interactions. As already reported, these interactions would result in lower permeability of the aqueous matrix to the interior of the porous structures of the OMF-N biochar, reducing urea dissolution and delaying diffusion of the nutrient mass to the solution caused by the chemical bonds of urea with the biochar contained in the pellets.

The increase in emissions from OMF-N 2:1 from 58 days onwards is possibly due to biological processes influenced by the C:N ratio of the granules. The greater number of pores available in OMF-N with a greater proportion of biochar results in a long-term condition which is favorable to solvent accumulation inside, since the pores are not completely filled with urea. Anaerobic sites intensify N₂O emissions under these conditions, as the transformation of urea into N₂O is dependent on denitrification processes [39]. The data show that there is an optimal biochar concentration (~26%) in the OMF-N, for which there is reduction, but above this concentration an increase in N₂O emissions is observed (Figure 7), whose effects are supported by physicochemical interactions.

Similar to N₂O, CO₂ emissions reveal similar temporal variation patterns in all treatments throughout the experiment (Figure 8). The CO₂ flows fitted the exponential model, revealing that 80% to 89% of emissions occurred up to 58 days. The highest (p < 0.05) accumulated CO₂ emissions throughout the study were observed in OMF-N 1:2 (305 mg kg⁻¹), 1:4 (292 mg kg⁻¹), and 2:1 (278 mg kg⁻¹) treatments, while the lowest emissions were observed for urea (220 mg kg⁻¹). These values are ten times lower than those verified in other studies [43].

The higher CO₂ emissions from OMF-N are understandable, since these formulations proportionally contained more carbon to urea, and this increase is probably due to the partial mineralization of labile C contained in the biochar. Considering the biochar:urea proportions in the OMF-N, the 2:1, 1:2, and 1:4 formulations contained 57.2%, 38.6%, and 31.2% of C, respectively, while the urea only contained 20%. However, the increase in CO₂ emissions did not linearly follow the same proportion of the total C contents of OMF-N, showing that part of the C contained in these fertilizers has high molecular stability. This means that even with an increase in CO₂ emissions from OMF-N compared to urea, its use provides C sequestration in the soil, as the final net balance of C is favorable, corroborating other studies [44,45]. Considering an environmental approach, this means that the potential for C sequestration from biochar-based OMF-N would be more than enough to neutralize CO₂ emissions from urea, thus meeting the assumptions of a regenerative agriculture.

3.6. NH₄⁺ and NO₃⁻ Leaching

The NH₄⁺ concentration profile in the soil solution for 28 days revealed higher levels up to 40 cm depth with the application of OMF-N 2:1 and there was no significant difference for the other fertilizers (Figure 9). The highest NH₄⁺ levels from 21 days after fertilizer application at a depth of 60 cm were verified with the OMF-N application, while there was no difference between the fertilizers at a depth of 80 cm. Higher NH₄⁺ concentrations were also verified when analyzing the leached solution 30 days after applying a fertilizer synthesized from a physical mixture of urea, mineral, and biochar [14].

Higher concentrations considering the NH₄⁺ release kinetics (Figure 4a) were to be expected with the application of urea, which was not verified. A possible explanation for the lower NH₄⁺ levels in the soil profile with the urea application may be theoretically related to the greater N loss by volatilization of NH₃⁺. Although we did not determine NH₃⁺ flows in our studies, there are reports in the literature that report losses of 38% to 44% by NH₃⁺ volatilization from urea [46,47]. On the other hand, there are several studies that report a reduction in NH₃⁺ volatilization and higher NH₄⁺ contents in the presence of biochar, whether incorporated into the soil or in the form of a nitrogen fertilizer based on biochar [48,49]. That said, it seems to us that the greater N loss by volatilization of NH₃⁺ and the greater ability of biochar to retain and gradually make NH₄⁺ available may in fact justify the lower NH₄⁺ concentrations in the solution along the soil profile with the urea application.

The urea and OMF-N 1:4 applications provided the greatest leaching of NO₃⁻ in the entire period evaluated, whose greatest differences were verified at 20 cm depth (Figure 10). Only OMF-N 2:1 provided a reduction in NO₃⁻ leaching below a depth of 60 cm from 21 days after fertilizer application. It seems evident that the increase in the biochar concentration in the OMF-N permeates in the sense of linearly reducing the leaching of NO₃⁻ at 28 days after fertilization, especially in deeper layers. Reduction in nitrate leaching after applying biochar-based nitrogen fertilizers was also observed in other studies [50,51].

The lower NO₃⁻ leaching with the OMF-N application is related to higher NH₄⁺ retention in the biochar. Nitrification is a process whose base substrate is NH₄⁺, which converts it into nitrite and nitrate via the catalytic activity of enzymatic complexes (nitrosomona and nitrobacter, respectively). Thus, the reduction in available NH₄⁺ consequently results in a reduced NO₃⁻ content. Then, we can logically say that the delay in the NH₄⁺ release is a beneficial process as long as it is low intensity (which does not extrapolate the synchrony of release with crop demand) since in addition to potentially reducing NH₃⁺ and N₂O emissions, also reduces the synthesis of NO₃⁻, which is the main leachable form of N in soil.

4. Conclusions

Our studies reveal that biochar can be an alternative for the synthesis of controlled-release nitrogen fertilizers from agronomic and environmental points of view. This is possible due to physicochemical interactions between urea and biochar even in a dry mechanical melting pelletizing process. OMF-N were able to significantly retain more ammonium and reduce N_total release from urea, as well as reduce nitrous oxide emissions. Although absolute CO₂ emissions intensified with OMF-N application, their use provides C sequestration in the soil given the recalcitrance of pyrogenic carbon which results in a positive net C balance. Nitrate concentration profiles revealed that the OMF-N application was able to reduce soil leaching to a depth of 80 cm.

Finally, our studies reveal that biochar from wood industry residues (environmental liabilities) can be used as an organic matrix for N anchoring aiming at the synthesis of controlled-release fertilizers. Moreover, we have presented better understanding of the processes involved in the biochar:urea interaction, and how they govern the dynamics of N release to the system. However, it is not yet known how the N release in these fertilizers affects plant growth and what is the effect on N₂O emission in the field, so field trials are imperative to elucidate these issues.