Double Coating as a Novel Technology for Controlling Urea Dissolution in Soil: A Step toward Improving the Sustainability of Nitrogen Fertilization Approaches

This research introduces a novel technology for reducing ordinary urea (OU) dissolution by developing double-coated urea (DCU) using phosphate rock (PR) as an outer layer to reduce its hydrolysis and sodium thiosulfate (STS) as an inner layer to inhibit the urease enzyme and nitrification process. Due to the double coating, the nitrogen content of DCU was lower than that of the OU (36.7% vs. 46.5%). The ultramorphological analysis using scanning electron microscopy (SEM) indicated that the controlled coating of urea, resulting from the outer layer of PR, was due to the adhesive effect of urea formaldehyde (UF), which was used as a glue. In addition, the transmission electron microscopy (TEM) analysis of the DCU revealed its high degree of agglomeration. The mechanical hardness of DCU was higher compared to that of OU (1.38 vs. 1.08 kgf). The seven-day dissolution rate test showed that OU reached 100% dissolution on the fifth day. The rate of DCU, however, was significantly lower (32% dissolution in the seventh day). Cumulative NO3− and NH4+ losses from a clay soil sample reached 68.3% and 7.6%, respectively, with OU measuring 40.5% compared to 4.9% for DCU 70 days after application. Field experiments showed a significant improvement in the marketable yield and agronomic nitrogen efficiency (ANE) of maize grains and zucchini fruits fertilized with DCU. Furthermore, the macro and micronutrient concentrations in maize grains and zucchini fruits showed an increase in the plants fertilized with DCU. In summary, double coating can be introduced as a novel technique to control urea dissolution in soil.


Introduction
Intensive agricultural systems have profoundly changed the global pedosphere, hydrosphere, and atmosphere [1][2][3]. In this regard, the increasing global mineral fertilizer consumption (108.744 million tonnes in 2020) has caused several deteriorative impacts on the ecosphere [4]. Due to its higher nitrogen content (46%), ordinary urea (OU) is the most commonly used nitrogen fertilizer worldwide [5]. The low nitrogen use efficiency of OU (20-35%) is associated with various potential losses: (i) leaching through high rainfall or irrigation, (ii) volatilization as ammonia and nitrous oxide emissions [6], (iii) denitrification dissolution rate as compared with OU; (3) evaluating cumulative ammonium (NH 4 + ) and nitrate (NO 3 − ) losses from soil following DCU and OU application; and (4) evaluating the effect of OU and DCU fertilizers on marketable yield, nitrogen efficiency (ANE) and nutrient concentrations in maize grains and zucchini fruits in field investigations.

Materials
Ordinary urea (46.5% N), urea formaldehyde (UF), and PR were obtained from El-Delta Fertilizers Plant, Mansoura, Egypt. Sodium thiosulfate (ReagentPlus ® , 99%) and all chemicals used of analytical grade were purchased from Sigma Aldrich Ltd. Alluvial soil (Vertic Torrifluvents) was obtained from the Agricultural Research Farm of Mansoura University, Egypt, in conjunction with starting the first field experiment.

DCU Preparation
OU (10 kg) was mixed with UF (2.5% w:w) in a fluidized bed granulator (Glatt GmbH, Binzen, Germany) before adding the first coating layer ( Figure 1). The full description of the equipment is illustrated in Figure 1. The fluidized bed granulator was made from stainless steel for granulation, drying, and pellet coating. The delivery of the coating material was carried out via a dual-fluid spray nozzle inserted from the side of the coating/granulating chamber. A particle-free zone was made to prevent the over-wetting of particles using low pressure, which was supplied through a tube that links the coating chamber with the inlet air plenum.
the physicochemical properties of double-coated urea (DCU); (2) assessing its me hardness and dissolution rate as compared with OU; (3) evaluating cumulative nium (NH4 + ) and nitrate (NO3 − ) losses from soil following DCU and OU applicat (4) evaluating the effect of OU and DCU fertilizers on marketable yield, nitrogen e (ANE) and nutrient concentrations in maize grains and zucchini fruits in field invest

Materials
Ordinary urea (46.5% N), urea formaldehyde (UF), and PR were obtained Delta Fertilizers Plant, Mansoura, Egypt. Sodium thiosulfate (ReagentPlus ® , 99% chemicals used of analytical grade were purchased from Sigma Aldrich Ltd. Allu (Vertic Torrifluvents) was obtained from the Agricultural Research Farm of Manso versity, Egypt, in conjunction with starting the first field experiment.

DCU Preparation
OU (10 kg) was mixed with UF (2.5% w:w) in a fluidized bed granulator (Glatt Binzen, Germany) before adding the first coating layer (Figure 1). The full descr the equipment is illustrated in Figure 1. The fluidized bed granulator was ma stainless steel for granulation, drying, and pellet coating. The delivery of the coa terial was carried out via a dual-fluid spray nozzle inserted from the side of t ing/granulating chamber. A particle-free zone was made to prevent the over-w particles using low pressure, which was supplied through a tube that links the chamber with the inlet air plenum. The first layer (STS at the rate of 2% w:w) was added using the nozzle of the pump. The coated urea was air-dried outside the fluidized bed granulator to obta ter homogeneity. Thereafter, the air-dried product was reinserted inside the fluidi granulator and mixed with UF (2.5% w:w) before adding the second coating layer. PR (<0.002 mm diameter) was added at a rate of 17.5% (w:w) inside the fluidiz granulator. The final product was then air-dried before packaging ( Figure 2). The first layer (STS at the rate of 2% w:w) was added using the nozzle of the sprayer pump. The coated urea was air-dried outside the fluidized bed granulator to obtain a better homogeneity. Thereafter, the air-dried product was reinserted inside the fluidized bed granulator and mixed with UF (2.5% w:w) before adding the second coating layer. Ground PR (<0.002 mm diameter) was added at a rate of 17.5% (w:w) inside the fluidized bed granulator. The final product was then air-dried before packaging ( Figure 2).

DCU Characterization
The nitrogen concentration in the DCU was determined using a CN analyze Thermo Scientific, Flash 2000, Milan, Italy). The K-factor calibration curve was pe using 90-100 mg of urea standard (46.65% N). Then, subsamples (90-100 mg) DCU were analyzed 10 times to optimize the repeatability. To investigate the surf acteristics of DCU and OU, the samples were sputter-coated with gold before surf acterization using a SEM (JEOL JSM 6510 lv). The surface elemental composition and OU was observed simultaneously using an energy dispersive X-ray spectrosc ford X-Max 20). Microstructure investigation was performed using a TEM (JEO 2100) after dispersion in warm DI water for 20 min. To study the crystallographic ties of the DCU and OU, digital selected-area electron diffraction (SAED) was pe with a scattering electron beam source equipped with a TEM.
Compression testing was performed to investigate the mechanical hardness as compared with OU. As the engineered urea granules were not of a uniform or shape, 24 granules were subjected to force using Instron equipment to obtain the crushing strength (kgf) vs. granule diameter. Granule diameter was obtained micrometer device. Granules were placed on a flat metal plate and the downwa sure of the Instron equipment was increased gradually until the granule fract evaluate the efficacy of the double coating technique on urea release, a seven-d lution rate test was performed by adding 50 g of OU or DCU to 250 mL of distill without shaking. The temperature was adjusted to 25 °C using a water bath an fractive index was measured daily in the fertilizer solution using a digital refra at a wavelength of 589 nm, as compared with distilled water [33]. The amoun released in the water was calculated using the standard calibration curve betwee fractive index and standard urea concentrations.

Cumulative Nitrate and Ammonium Losses
To assess the protective effect of the double coating technology against nit ammonium losses from the soil matrix, PVC columns (7.0 cm diameter and 40.0 cm were uniformly packed with 2.0 kg of air-dried clay soil (Vertic Torrifluvents). Som ical and chemical analyses of the tested soils are presented in Table 1. An acid sand layer (0.2-2.0 mm average diameter and 5.0 cm height) was inserted at the e PVC columns to ensure a uniform flow of leachates. A filter paper (Whatman ®

DCU Characterization
The nitrogen concentration in the DCU was determined using a CN analyzer (model Thermo Scientific, Flash 2000, Milan, Italy). The K-factor calibration curve was performed using 90-100 mg of urea standard (46.65% N). Then, subsamples (90-100 mg) from the DCU were analyzed 10 times to optimize the repeatability. To investigate the surface characteristics of DCU and OU, the samples were sputter-coated with gold before surface characterization using a SEM (JEOL JSM 6510 lv). The surface elemental composition of DCU and OU was observed simultaneously using an energy dispersive X-ray spectroscopy (Oxford X-Max 20). Microstructure investigation was performed using a TEM (JEOL JEM-2100) after dispersion in warm DI water for 20 min. To study the crystallographic properties of the DCU and OU, digital selected-area electron diffraction (SAED) was performed with a scattering electron beam source equipped with a TEM.
Compression testing was performed to investigate the mechanical hardness of DCU as compared with OU. As the engineered urea granules were not of a uniform diameter or shape, 24 granules were subjected to force using Instron equipment to obtain data for the crushing strength (kgf) vs. granule diameter. Granule diameter was obtained using a micrometer device. Granules were placed on a flat metal plate and the downward pressure of the Instron equipment was increased gradually until the granule fractured. To evaluate the efficacy of the double coating technique on urea release, a seven-day dissolution rate test was performed by adding 50 g of OU or DCU to 250 mL of distilled water without shaking. The temperature was adjusted to 25 • C using a water bath and the refractive index was measured daily in the fertilizer solution using a digital refractometer at a wavelength of 589 nm, as compared with distilled water [33]. The amount of urea released in the water was calculated using the standard calibration curve between the refractive index and standard urea concentrations.

Cumulative Nitrate and Ammonium Losses
To assess the protective effect of the double coating technology against nitrate and ammonium losses from the soil matrix, PVC columns (7.0 cm diameter and 40.0 cm depth) were uniformly packed with 2.0 kg of air-dried clay soil (Vertic Torrifluvents). Some physical and chemical analyses of the tested soils are presented in Table 1. An acid-washed sand layer (0.2-2.0 mm average diameter and 5.0 cm height) was inserted at the end of the PVC columns to ensure a uniform flow of leachates. A filter paper (Whatman ® Ashless, Grade 42) was inserted between the soil and sand layers, and the bottom of each column was tightly sealed with silicone adhesive. A glass tube (5.0 mm internal diameter) was inserted inside the bottom of the columns to collect leachates inside closed-glass bottles using rubber stoppers with a hole for inserting into the other side of the glass tube. Water was applied to the soil columns at a 70% saturation percentage. The average values of temperature and relative humidity during the experiment were 25.0 ± 0.5 • C and 65.0 ± 0.5%, respectively. Columns were weighed every two days, and the loss of water was compensated by volume, assuming that water specific gravity was 1.0 g cm −3 . Nitrogen was applied to the soil columns in the form of DCU or OU at the rate of 2.0 g column −1 . Control treatments (without nitrogen addition) were performed to distinguish between the NO 3 − and NH 4 + derived from the soil and those derived from fertilizers. Cumulative nitrate (CNP) and ammonium (CAP) percentages in leachates were calculated according to the following equations:

Field Experiments
Two field experiments were carried out in the Agricultural Research Farm of Mansoura University in a clay soil (Vertic Torrifluvents). Representative surface soil samples were collected from the experimental field before starting the field experiments. Collected samples were air-dried, crushed, and passed through a 2 mm sieve (ISO 11464). Maize seeds (Zea Maize L. cv Single Hybrid 10) were cultivated in the summer season to present a cereal crop. Thereafter, zucchini (Cucurbita Pepo L. cv Eskandrani) was cultivated as a successive planting to investigate the residual effect of the developed fertilizer. In a completely randomized block design, three fertilization treatments, OU, DCU, and a control (without nitrogen application), were arranged in three replicates. Nitrogen was applied at the rates of 285 and 225 kg ha −1 for maize and zucchini, respectively. Plants were irrigated following 50% depletion of the available soil water (the difference between field capacity and the permanent wilting point). Soil water content was gravimetrically measured according to the difference in mass of soil samples before and after drying at 105 • C ± 5 (ISO 11465). All agricultural practices were carried out according to the recommendations for maize and zucchini crops. Plants were harvested on 30 August and 30 October for maize and zucchini, respectively. Maize grains and zucchini fruits yield was measured, and the agronomic nitrogen efficiency (ANE) was calculated as additional economic yield per applied fertilizer unit [34].
where Yield (F) and Yield (C) are the obtained yields of fertilized and unfertilized (control) treatments, respectively. Representative soil samples were collected following maize harvesting to identify the effect of DCU on residual nitrogen retained in the soil for the zucchini crop.

Soil, Leachate and Plant Analyses
Soil analyses were carried out using the standard methods: particle size distribution by the pipette method [35], soil reaction (pH) in soil paste and soil electrical conductivity in soil paste extract [36], total carbonate as the gasometric determination of released CO 2 [37], NH 4 + and NO 3 − concentrations in soil column leachates, and the soil availability of nitrogen using the Kjeldahl method [38]. Grain and fruit samples were oven-dried at 70 • C ± until reaching a constant weight. The total nitrogen in grain and fruit yields was determined using the automatic Kjeldahl method (model Raypa, DNP-2000) after digestion using sulfuric (H 2 SO 4 ) and perchloric (HClO 4 ) acids (1:1). Total phosphorus was determined colorimetrically at a wavelength of 680 nm using a spectrophotometer (Spekol), and total potassium was determined using a Gallen Kamp flame photometer and micronutrients (Fe, Zn and Mn) with an atomic absorption spectrophotometer (PerkinElmer model 5000).

Statistical Analysis
At a significance level of (0.05), data were statistically analyzed using CoStat (Version 6.303, CoHort, Berkeley, CA, USA, 1998-2004) based on a one-way ANOVA [39]. Standard deviation values (represented in error bars) were found using the Microsoft Excel Software (version 2010, Microsoft Corporation, Redmond, WA, USA). Data was further presented using the OriginPro 9.1. software.

Fertilizer Characterization
Nitrogen concentration in the DCU decreased by about 21% as compared to the OU (36.7% vs. 46.5%, Figure 1). The incorporation of inner and outer coating layers in the modern urea fertilizer led to a reduction in the final concentration of nitrogen as compared with the initial concentration in the ordinary form. The ultramorphological analysis of the DCU indicated the presence of a spherical geometry outer layer with rough surfaces (Figure 3c), as compared with the soft, dense, compact, and rigid surface structures of OU (Figure 3a). The EDS spectra showed a lowering in the N peak in the DCU compared with the OU, which confirmed the data obtained by the CN analyzer. Moreover, a high peak of Ca was observed in the DCU with the disappearance of other peaks observed in the OU (e.g., Zn and Cu), which indicated the controlled coating of urea by the outer layer of rock phosphate ( Figure 3d).
As observed in the TEM micrographs, DCU also presented a spherical morphology with dark spots and a high degree of agglomeration, given the presence of an outer coating layer (Figure 4c). Added to this, the colloidal nature of urea formaldehyde with its adhesive properties might increase the agglomeration of DCU [40]. In contrast, OU appeared separated with a smaller particle size and non-uniform shape as compared with the DCU (Figure 4a). Although urea is one of the few organic compounds that possesses a highly crystallographic structure, its high solubility decreased its crystalline appearance in the SAED micrographs (Figure 4b). In contrast, the SAED pattern of DCU (Figure 4d) showed spots and ring patterns in a random orientation, illustrating the crystalline nature of the existing compounds: phosphate-bearing minerals with various degrees of crystallinity [41], the original urea granules [42] UF [43], STS [44], and the potential formation of crystalline thiourea following the reaction between urea and sodium thiosulfate [45].
A significant increase in the particle diameter was observed following the double coating (1.79 vs. 2.40 mm average diameter for OU and DCU, respectively). The mechanical hardness test showed that the mechanical crush strength increased with an increasing particle diameter of fertilizer granules (1.08 vs. 1.38 kgf average strength for OU and DCU, respectively), and this relation is expressed as a linear relationship between granule diameter (mm) and crushing strength (kgf) in Figure 5. Furthermore, various physical properties (e.g., hardness) could be enhanced after coating with PR. This increase in mechanical hardness could improve the ability of DCU to withstand fracture. ure 3c), as compared with the soft, dense, compact, and rigid surface structures of OU (Figure 3a). The EDS spectra showed a lowering in the N peak in the DCU compared with the OU, which confirmed the data obtained by the CN analyzer. Moreover, a high peak of Ca was observed in the DCU with the disappearance of other peaks observed in the OU (e.g., Zn and Cu), which indicated the controlled coating of urea by the outer layer of rock phosphate (Figure 3d). As observed in the TEM micrographs, DCU also presented a spherical morphology with dark spots and a high degree of agglomeration, given the presence of an outer coating layer (Figure 4c). Added to this, the colloidal nature of urea formaldehyde with its adhesive properties might increase the agglomeration of DCU [40]. In contrast, OU appeared separated with a smaller particle size and non-uniform shape as compared with the DCU (Figure 4a). Although urea is one of the few organic compounds that possesses a highly crystallographic structure, its high solubility decreased its crystalline appearance in the SAED micrographs (Figure 4b). In contrast, the SAED pattern of DCU (Figure 4d) showed spots and ring patterns in a random orientation, illustrating the crystalline nature of the existing compounds: phosphate-bearing minerals with various degrees of crystallinity [41], the original urea granules [42] UF [43], STS [44], and the potential formation of (e) The seven-day dissolution rate test showed that the dissolution rate of OU was 22% in the first day compared with only 3% for the DCU (Figure 6). This dissolution rate increased gradually as OU reached 100% in the fifth day. This rate, however, was relatively slow compared with the DCU, which reached the final day test with only 32% dissolution. The low dissolution rate of DCU was mainly attributed to the low solubility of PR, which served as the outer coating layer [46]. This dissolution rate is expected to be higher under wet soil conditions, given the supporting bio-leaching from soil microorganisms [47]. a highly crystallographic structure, its high solubility decreased its crystalline appearance in the SAED micrographs (Figure 4b). In contrast, the SAED pattern of DCU (Figure 4d) showed spots and ring patterns in a random orientation, illustrating the crystalline nature of the existing compounds: phosphate-bearing minerals with various degrees of crystallinity [41], the original urea granules [42] UF [43], STS [44], and the potential formation of crystalline thiourea following the reaction between urea and sodium thiosulfate [45].  A significant increase in the particle diameter was observed following the double coating (1.79 vs. 2.40 mm average diameter for OU and DCU, respectively). The mechanical hardness test showed that the mechanical crush strength increased with an increasing particle diameter of fertilizer granules (1.08 vs. 1.38 kgf average strength for OU and DCU, respectively), and this relation is expressed as a linear relationship between granule diameter (mm) and crushing strength (kgf) in Figure 5. Furthermore, various physical properties (e.g., hardness) could be enhanced after coating with PR. This increase in mechanical hardness could improve the ability of DCU to withstand fracture. The seven-day dissolution rate test showed that the dissolution rate of OU was 22% in the first day compared with only 3% for the DCU (Figure 6). This dissolution rate increased gradually as OU reached 100% in the fifth day. This rate, however, was relatively slow compared with the DCU, which reached the final day test with only 32% dissolution. The low dissolution rate of DCU was mainly attributed to the low solubility of PR, which

Cumulative Nitrate and Ammonium Losses
This experiment simulated the potential losses of available nitrogen forms from the soil matrix to groundwater following urea dissolution. In general, NO3 − losses in leachates were more pronounced than those of NH4 + losses. This is mainly attributed to the fast transformation of NH4 + to NO3 − following the nitrification process, and to the binding force of NH4 + to the negatively charged soil surface compared with leaching from the soil matrix, given the positive charge of ammonium ions, whereas nitrate is negatively charged. Cumulative nitrate and ammonium losses decreased significantly with the DCU as compared with the OU (Figure 7). In the first week, approximately 4.4% and 0.8% were lost from the total OU in forms of NO3 − and NH4 + , respectively. Meanwhile, the NO3 − and NH4 + losses were significantly lower following the addition of DCU, with losses of only 0.1% and 0.05%, respectively. After reaching the fifth week, approximately 50% and 6.2% of the total applied OU was lost in the form of NO3 − and NH4 + , respectively, but only 20.7% and 2.3% was lost from the DCU. At the end of the experiment (70 days after fertilizer addition), about 68.3% and 7.6% were lost from the OU vs. 40.5% and 4.9% from the DCU, in the forms of NO3 − and NH4 + , respectively.
As previously discussed, in the seven-day dissolution rate test, the low solubility of PR (the outer coating layer) reduced the dissolution of DCU as compared with a relatively fast dissolution of OU. On the other hand, STS (the inner layer) reduced the NH4 + transformation to NO3 − due to its inhibitory effect on the urease enzyme and nitrification process. Few studies have discussed the potential role of STS as a urease inhibitor. Thiosulfate does not have a direct effect on urease enzyme. However, there are several mechanisms proposed for its inhibition of the urease enzyme: (i) the abiotic and rapid reaction with soil, forming tetrathionate and liberating Fe 2+ and Mn 2+ ions, which are able to inactivate urease enzyme by binding to sulfhydryl groups [29]; (ii) the formation of tetrathionate following oxidation, which has an inhibitory effect on the urease enzyme [48]; and (iii) the potential formation of thiourea, which is considered to be a urease inhibitor following the reaction between urea and STS [49]. Besides its role as a urease inhibitor, STS has a remarkable effect as a nitrification inhibitor, with a 64% inhibition of nitrification and a 41% increase in apparent nitrogen recovery as compared with a sole urea application [50]. The mechanisms responsible for nitrification inhibition by STS are not fully understood;. However, a few reports have suggested the following inhibition mechanisms: (i) liberating Fe 2+ and Mn 2+ ions following the conversion of thiosulfate to tetrathionate can retard the nitrification process [51,52]; (ii) tetrathionate itself can inhibit the microbial or enzymatic activities responsible for nitrification [48,53]; (iii) the formation of carbonyl sulfide, which has the ability to inhibit nitrification [54]; and (iv) the direct inhibitory effect of thiosulfate on Nitrobacter sp. and other NO2 − oxidizing microorganisms [55].

Cumulative Nitrate and Ammonium Losses
This experiment simulated the potential losses of available nitrogen forms from the soil matrix to groundwater following urea dissolution. In general, NO 3 − losses in leachates were more pronounced than those of NH 4 + losses. This is mainly attributed to the fast transformation of NH 4 + to NO 3 − following the nitrification process, and to the binding force of NH 4 + to the negatively charged soil surface compared with leaching from the soil matrix, given the positive charge of ammonium ions, whereas nitrate is negatively charged. Cumulative nitrate and ammonium losses decreased significantly with the DCU as compared with the OU (Figure 7). In the first week, approximately 4.4% and 0.8% were lost from the total OU in forms of NO 3 − and NH 4 + , respectively. Meanwhile, the NO 3 − and NH 4 + losses were significantly lower following the addition of DCU, with losses of only 0.1% and 0.05%, respectively. After reaching the fifth week, approximately 50% and 6.2% of the total applied OU was lost in the form of NO 3 − and NH 4 + , respectively, but only 20.7% and 2.3% was lost from the DCU. At the end of the experiment (70 days after fertilizer addition), about 68.3% and 7.6% were lost from the OU vs. 40.5% and 4.9% from the DCU, in the forms of NO 3 − and NH 4 + , respectively.

Field Experimentation
Significant increases in the grain and fruit yields of maize and zucchini were recorded with DCU fertilization as compared with the OU and control treatments ( Table 2). The yield increase achieved by DCU application was 19.7% and 35.4% for maize and zucchini, respectively. This obtained yield increase resulted in maximizing the ANE of DCU as compared with OU (19.8 vs. 11.9, and 8.7 vs. 21.5 kg kg −1 for maize and zucchini, respectively). It is also clear that the increase in ANE with zucchini was more pronounced than with maize. This is mainly due to the residual effect of N retained in the soil after maize harvesting (Table 1), which might improve zucchini growth in its first growth stage. Several mechanisms could be responsible for improving the yield productivity of plants: (i) As previously discussed, in the seven-day dissolution rate test, the low solubility of PR (the outer coating layer) reduced the dissolution of DCU as compared with a relatively fast dissolution of OU. On the other hand, STS (the inner layer) reduced the NH 4 + transformation to NO 3 − due to its inhibitory effect on the urease enzyme and nitrification process. Few studies have discussed the potential role of STS as a urease inhibitor. Thiosulfate does not have a direct effect on urease enzyme. However, there are several mechanisms proposed for its inhibition of the urease enzyme: (i) the abiotic and rapid reaction with soil, forming tetrathionate and liberating Fe 2+ and Mn 2+ ions, which are able to inactivate urease enzyme by binding to sulfhydryl groups [29]; (ii) the formation of tetrathionate following oxidation, which has an inhibitory effect on the urease enzyme [48]; and (iii) the potential formation of thiourea, which is considered to be a urease inhibitor following the reaction between urea and STS [49]. Besides its role as a urease inhibitor, STS has a remarkable effect as a nitrification inhibitor, with a 64% inhibition of nitrification and a 41% increase in apparent nitrogen recovery as compared with a sole urea application [50]. The mechanisms responsible for nitrification inhibition by STS are not fully understood;. However, a few reports have suggested the following inhibition mechanisms: (i) liberating Fe 2+ and Mn 2+ ions following the conversion of thiosulfate to tetrathionate can retard the nitrification process [51,52]; (ii) tetrathionate itself can inhibit the microbial or enzymatic activities responsible for nitrification [48,53]; (iii) the formation of carbonyl sulfide, which has the ability to inhibit nitrification [54]; and (iv) the direct inhibitory effect of thiosulfate on Nitrobacter sp. and other NO 2 − oxidizing microorganisms [55].

Field Experimentation
Significant increases in the grain and fruit yields of maize and zucchini were recorded with DCU fertilization as compared with the OU and control treatments ( Table 2). The yield increase achieved by DCU application was 19.7% and 35.4% for maize and zucchini, respectively. This obtained yield increase resulted in maximizing the ANE of DCU as compared with OU (19.8 vs. 11.9, and 8.7 vs. 21.5 kg kg −1 for maize and zucchini, respectively). It is also clear that the increase in ANE with zucchini was more pronounced than with maize. This is mainly due to the residual effect of N retained in the soil after maize harvesting (Table 1), which might improve zucchini growth in its first growth stage. Several mechanisms could be responsible for improving the yield productivity of plants: (i) the outer coating layer (PR) led to the release of nitrogen in a controlled manner in synchrony with the sequential needs of the plants until the final growth stage [19,56]; (ii) the partial contribution of PR in phosphorus nutrition and improving soil quality properties [57,58]; (iii) the contribution of the inner coating layer (STS) to sulfur nutrition, and its inhibitory effect on the urease enzyme [48]; (iv) the role of STS as a nitrification inhibitor through retarding the biological oxidation of NH 4 + -N to NO 3 − -N [50,59]; and (v) the beneficial effect of urea formaldehyde as a slow-release N source and biological soil activator [60]. Values are the average mean of three replications ± standard error.
Due to the beneficial effect of the double coating technology on retarding urea dissolution/hydrolysis and inhibiting the nitrification process, the final concentration of nitrogen in maize grains and zucchini fruits was higher with DCU (1.88% and 4.69%, respectively) as compared with the OU (1.65% and 2.45%, respectively) and the control (1.53% and 2.02%, respectively) treatments. Phosphorus concentration was significantly higher in plants fertilized with DUC due to the slow-release supply of phosphorus derived from phosphate rock, which was reflected in an increasing available phosphorus concentration by about 40.1% following the DCU application. The increase in phosphorus concentration in maize grains and zucchini fruits was, respectively, 1.40 and 1.14-fold greater than with OU. Likewise, the available potassium concentration in the soil showed a significant increase following DCU application (227 mg vs. 335 mg kg −1 ), which led to increased potassium concentrations in the grains and fruits. Additionally, micronutrient concentrations in the maize grains and zucchini fruits also increased with DCU application. The slow-release supply of nitrogen might have activated the enzymes involved in N metabolism, thereby forcing the plants to absorb higher amounts of micronutrients (Ai et al., 2017).

Conclusions
Controlling urea dissolution is urgently needed to meet the nitrogen requirements of plants in all growth stages. The double coating technology created a controlled outer layer of PR with a high degree of agglomeration given the sticky effect of UF, and an inner layer of STS protected from fast degradation in soil. Research results concluded that the double coating technology improved urea resistance against fast dissolution, maximized the mechanical hardness of fertilizer granules, reduced the cumulative NO 3 − and NH 4 + leaching from the soil matrix, maximized the ANE of maize and zucchini crops, and increased N concentration in grains and fruits. Furthermore, DCU application modulated the concentrations of macro and micronutrients in maize grains and zucchini fruits. The findings of this study highlight the applicability of double coating as a novel technology to improve urea efficacy in terrestrial ecosystems. Further investigations should be undertaken using different application rates to improve the economic revenues of DCU fertilizers.