Characterization of Single Superphosphate Powders – a study of Milling Effects on Solubilization Kinetics

Phosphorus (P) is an essential element for any living being that exists on Earth. It plays a primary role in the formation of vital biomolecules such as deoxyribonucleic acid (DNA) and nucleotide forms adenosine triphosphate and adenosine diphosphate (ADP and ATP), which store energy and are extremely important for cellular respiration and photosynthesis. Together with nitrogen (N) and potassium (K), P belongs to the class of the three essential macronutrients for development of any kind of plant.1-4 In the 19th century, phosphorus fertilization in crops was mainly accomplished with bone meal from slaughtered animals. However, the solubility of this material was not satisfactory enough to ensure the necessary supply of phosphorus to agricultural crops, and as the demand for foods increased, a search for new phosphorus sources became necessary. In 1840, Justus von Leibig found out that the solubilization of bone meal could be assisted by adding sulfuric acid, hence, increased release rates of phosphorus were obtained. Later the same scientist successfully applied the same chemical treatment on phosphate rocks. John Bennet Lawes developed this process at an industrial scale and renamed the product as superphosphate fertilizer, which was the first P-fertilizer commercially produced worldwide.5-7 Since its invention, the process for obtaining single superphosphate (SSP) only changed with respect to the phosphorus source. Currently the main source of phosphorus used by the industries to fabricate P-fertilizers is the phosphate rock, which is mainly composed of some kind of apatite, hydroxyapatite (Ca10(PO4)6(OH)2), fluorapatite (Ca10(PO4)6F2) or carbonateapatite (Ca10(PO4)6CO3). It is estimated that the current fertilizer industry consumes approximately 90% of all phosphate rock produced worldwide.8-9 The chemical reaction for SSP formation can be summarized in the equation below:


-Introduction
Phosphorus (P) is an essential element for any living being that exists on Earth.It plays a primary role in the formation of vital biomolecules such as deoxyribonucleic acid (DNA) and nucleotide forms adenosine triphosphate and adenosine diphosphate (ADP and ATP), which store energy and are extremely important for cellular respiration and photosynthesis.][3][4] In the 19th century, phosphorus fertilization in crops was mainly accomplished with bone meal from slaughtered animals.However, the solubility of this material was not satisfactory enough to ensure the necessary supply of phosphorus to agricultural crops, and as the demand for foods increased, a search for new phosphorus sources became necessary.In 1840, Justus von Leibig found out that the solubilization of bone meal could be assisted by adding sulfuric acid, hence, increased release rates of phosphorus were obtained.Later the same scientist successfully applied the same chemical treatment on phosphate rocks.6][7] Since its invention, the process for obtaining single superphosphate (SSP) only changed with respect to the phosphorus source.Currently the main source of phosphorus used by the industries to fabricate P-fertilizers is the phosphate rock, which is mainly composed of some kind of apatite, hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 ), fluorapatite (Ca 10 (PO 4 ) 6 F 2 ) or carbonateapatite (Ca 10 (PO 4 ) 6 CO 3 ).9] The chemical reaction for SSP formation can be summarized in the equation below: Essentially, the rock is reacted with dilute sulfuric acid (between 68-75%) for up to 4 hours.1][12][13] However, despite this apparent simplicity, only a few papers have described the fundamental aspects of SSP, i.e., phase composition, thermal degradation behavior and solubilization kinetics.Particularly, the importance of milling conditions and the resulting particle size was addressed only in empirical studies, where the focus on the agronomical aspects surpassed any evaluation of SSP in terms of microstructural characterization.The textural characterization of SSP is of scientific relevance since it could contribute to the understanding of how the properties of SSP could be tailored in order to obtain more efficient SSP fertilizers as well as to explain why SSP is much more active in soil than the pristine raw materials, despite the similar phase composition.
In this paper we report a detailed characterization of SSP powders milled under different conditions to obtain different particle sizes.Characterizations by thermogravimetry, infrared spectroscopy, total surface area measurements, and electron microscopy were correlated with solubilization tests in order to discuss the main factors determining the performance of SSP powders as a P-fertilizer.To the best of our knowledge, this is the first detailed study on relationships between textural property and solubilization kinetics of SSP fertilizers.

Milling of SSP fertilizer
The SSP powders were prepared from a commercial SSP fertilizer (original SSP) by using predetermined milling times of 2.5, 5, 10, 20, 40 and 80 minutes.The milling process was carried out using an orbital mill equipment comprised of a porcelain jar and alumina balls.
Chemical analyses was done by X-ray fluorescence (XRF) using the lithium tetraborate fusion technique.The 10 most common oxides found in ores were analyzed.
Fourier transform infrared spectroscopy (FT-IR) analyses were performed on a Paragon 1000 spectrometer (Perkin-Elmer  ) in the absorbance mode ranging the wavenumber from 4000 to 400 cm -1 .
The morphology of the milled SSP samples was observed by field emission gun scanning electron microscopy (FEG-SEM) using a JSM-6701F FEG microscope (JEOL, Japan).The imaging was carried out using an accelerating voltage of 2 kV and secondary electron detector.
Total surface area measurements were done by isothermal nitrogen adsorption on aMicrometritics ASAP 2020 equipment using the 5-points BET (Brunauer-Emmett-Teller) method. 14hermogravimetric analyses (TGA) were performed on a Q500 thermal analyzer (TA Instruments) using synthetic air atmosphere (80% N 2 and 20% O 2 ) with a flow of 60 mL min -1 .Samples were heated up to 1000 °C using a heating rate of 10 °C min -1 .Differential scanning calorimetry (DSC) analyses were conducted in a Q100 calorimeter (TA Instruments).Samples were placed in an aluminum crucible and heated at a heating rate of 10 o C min -1 from 25 to 400 o C under inert nitrogen atmosphere flowing at 40 mL min -1 .

Solubilization tests
The solubility of the SSP materials was directly tested in deionized water with a resistivity higher than 18.2 MΩ cm.The solubilization experiments were carried out for 7 days following the model proposed by Tomaszewska and Jarosiewicz 15 and modified by Pereira 16 .For each experiment, aliquots parts were taken periodically and used to determine the phosphorus concentration released by the SSP sample.In brief, aliquot of 5 mL was reacted with 2 mL of ascorbic acid solution (0.4 mol L -1 ), 0.2 mL of citric acid (0.03 mol L -1 ) and 2 mL of a mixed reactant consisting of 25 mL of a sulfuric acid solution (4.7 mol L -1 ), 5.5 mL of ammonium molybdate solution (0.08 mol L -1 ) and 0.6 mL of antimony and potassium tartrate solution (0.05 mol L -1 ).This mixture was reacted at 50 °C for 15 minutes to form a phosphoantimonylmolybdenum blue complex (PBC).The PBC concentration was determined by UV-Vis spectrometry at a wavelength of 880 nm [17][18] in a Lambda 25 UV-Vis spectrophotometer (Perkin Elmer).Linear function (r 2 = 0.998) was obtained using several dilutions of phosphorus standard solution in a concentration range from 0.5 to 5 mg L −1 .
1][22] The Ritger and Peppas model describes the classical diffusion (referred to as Fickian diffusion) into a thin film where up to 60% of the released material can be characterized by a constant k multiplied by time raised to a pre-exponential factor n, as described by the equation below: Where (M t ) is the amount released at time t, (M ∞ ) is the equilibrium amount, i.e. the release of 100% of the material, (k) is a kinetic constant, and (n) is the pre-exponential factor associated with the type of mechanism.If n = 5 the diffusion is assumed to be a Fickian diffusion process, and for different n values the diffusion process may be suffering interference of the diffusion matrix. 19,23,24he pre exponential factor (n) and kinetic constant (k) are determined by plotting the logarithm of the concentration versus the logarithm of time.Thus the equation takes a linear form where the pre factor (n) and the kinetic constant (k) are the slope coefficient and linear coefficient, respectively, as shown in the equation below: The other model used in this study was an equation that assumes that the release of nutrient is a second-order reaction where its rate law is given by: Where the velocity υ is proportional to a kinetic constant (k) multiplied by the concentrations of A (solvent) and B (solute).Considering that the concentration of solvent A (water) is infinitely greater than the concentration of the solute B (nutrient) its variation over time will be approximately zero or, in other words: The concentration of A can be encompassed into the constant k then a law of pseudo first order rate is obtained.
Where the new rate constant k' is k[A] 0 .By writing the equation above in a differential form and performing integration between the limits, one can obtain: For a typical exponential curve where t = 0 the amount of nutrient released is also equal to zero, and for an infinite time the amount of nutrient released is equal to 100%.
Table 1 displays the results of chemical analysis for 10 traces of surface oxides determined by X-ray fluorescence.It is known that the P 2 O 5 concentration in SSP fertilizers may vary due to the manufacturing process, phosphorus concentration in the pristine rock (used as a raw material) and other factors.According to previous reports 9,10,25-27 , the available P content values in SSP are in the range from 7% to 10% (16% to 23% P 2 O 5 ), thus the amount of phosphorus obtained in this study is in agreement with the literature and specifications for this fertilizer.It is noteworthy pointing out the high loss on ignition reported in Table 1, which is clearly related to the high amount of hydrated phases occurring in the sample.Other constituents listed in Table 1 were expected as they are contaminant minerals usually found in SSP fertilizers.
The micrographs in Figure 2 reveal that the increase of the milling time provoked a decrease in the SSP particle size.It is observed that the milling process was effective because particle deformation occurred, but one can also notice that small aggregates were formed with increasing milling time.Furthermore, sub-micron particles are visible, which is clearly seen even in the micrograph of the SSP_2.5-minsample.The nitrogen physisorption results presented in Figure 3 show the relative increase of surface area of the SSP powder with increasing milling time, which confirms the features drawn from the SEM micrographs.Earlier studies on grinding mills [28][29] showed that there is an optimal time for each assembled system (mill and grinding media), i.e., there is a specific time at which the smallest particle size, and consequent largest relative surface area, is reached.After this point, the particle deformation may lead to agglomeration, affecting the total surface area available for adsorption.From Figure 3 it is possible to suggest that the mounted system has not reached its maximum grinding efficiency, as the surface area values of SSP continued increasing with increasing milling time.
Figure 4 shows the FTIR spectrum of the original SSP fertilizer.The infrared spectroscopy was used to confirm the coexistence of two mineral phases, monocalcium phosphate and gypsum, in the majority of phosphate fertilizers.The corresponding vibration bands (OH -, SO 4 2-and PO 2 chemical groups) have been summarized in Table 2. [30][31][32][33][34][35] Thermogravimetric analysis for all samples showed the same profile as that illustrated in Figure 5.The first mass loss between 100 and 250 °C [36][37][38][39][40][41][42] can be ascribed to the release of water molecules of gypsum which occurrs in two steps as described by Equations 1 and 2: Calcium phosphate loses water molecules by dehydration reactions at the same temperature range 38,43,44 , also in two stages, as described by Equations 3 and 4: Figure 1.XRD pattern of original (not milled) SSP fertilizer.As the first gypsum dehydration step corresponds to a mass loss of 75%, one can assume that the DTG peak at 134 °C corresponds to this water loss of gypsum.On the other hand, the calcium phosphate dehydration is more difficult to be proven, since the phosphate phase is present at smaller quantities in the fertilizer than the gypsum phase.According to refs.[38-44] the complete dehydration for both phases occurs up to 300 o C.
Important thermal events were also observed in the range 550 °C -950 °C as shown in Figure 6 for the original SSP fertilizer and its corresponding milled samples.According to Pyldme et al. (1979) the dihydrogen phosphate undergoes a phase transition at temperatures of approximately 500 °C.In this case, the thermal events observed for the SSP powders cannot be associated with the phase transition of dyhydrogen phosphate since they occurred at higher temperatures (550 -950 °C) in the TG/DTG curves.Conversely, gypsum does not undergo any phase transition in the temperature range studied, remaining stable as an anhydrous phase up to 1300 °C. 38,40Thus, the thermal events observed at 550 °C -950 °C must be related to some other intermediary (non-stoichiometric) phases, taking into account that the pristine rocks used in the SSP fertilizer production are typically made up of numerous types of minerals.Nevertheless, it is interesting to note that the increase of milling time shifted the characteristic temperatures of these thermal events toward lower temperature values, from 722 ° C to 674 ° C, and from 789 ° C to 762 ° C for the first and second peaks, respectively.
Differential scanning calorimetry (DSC) revealed a similar thermal behavior for all milled SSP samples.A typical DSC trace is shown for SSP_20min in Figure 7.It is possible to note that in the range between 150 o C and 225 o C there are only endothermic events.These endothermic events occur within the same temperature range at which mass losses were observed by TGA.0][41] Table 3 summarizes the main thermal events observed by TGA and DSC, where these comparisons are clearly stated.In order to examine the milling effect on the kinetics of phosphorus release, representative SSP samples were tested in experiments of solubilization in deionized water.It is well known that this type of experiment is highly aggressive because the dihydrogen calcium phosphate is highly soluble in water.On the other hand, the phosphorus release to soil will involve different mechanisms mainly depending on environmental humidity conditions.The results presented in Figure 8 show a relation between milling time and release kinetics: the solubilization of 70% of the total available phosphate present in the granular SSP was attained after 3 days of immersion in water, whereas the same content of released phosphate was obtained for the SSP_80min sample after 2 days.It is noteworthy mentioning that the milling process was more effective at short milling time, since the SSP_2.5minsample exhibited a significant increase of solubilization and phosphorus release kinetics.
The calculated kinetic parameters are reported in Table 4.It can be seen that the rate constant (k) values calculated by the two models increased as the milling time was increased.But for milling times of 5, 10 and 20 minutes the same trend was not observed.According to the Peppas model, when the value of the factor (n) is greater than 1, diffusion is considered to be anomalous i.e. the diffusion process suffers some kind of delay.In this case it is possible to suggest that the plaster layers may be acting as a physical barrier, hindering the diffusion of SSP to the aqueous medium.This means that agglomeration of the plaster influences on the solubility of the SSP fertilizer, since the disassembling by milling was effective in increasing its solubilization.
This result should be carefully analyzed, since a rapid release in field may favor the immobilization of phosphate in soil by reacting with iron, aluminum or manganese species [8][9] .Nevertheless, this result demonstrates that it is possible to consider the orbital milling as a possible route to obtain more efficient P-fertilizers from other low-soluble phosphate fractions.

-Conclusions
The orbital milling process is an effective process to accelerate the solubilization of SSP by reducing the effective particle size of this fertilizer.Despite the regular increase of surface area with increasing milling time, the solubilization kinetics tends to a plateau probably due other effects, such as agglomeration of fertilizer particles.The thermal characterizations were noteworthy, as they revealed that the gypsum phase present in SSP is influenced by milling

Figure 3 .
Figure 3. Specific surface are of simple superphosphate (SSP) fertilizer milled at different times.Figure 4. FTIR spectrum of original (not milled) SSP fertilizer.

Figure 4 .
Figure 3. Specific surface are of simple superphosphate (SSP) fertilizer milled at different times.Figure 4. FTIR spectrum of original (not milled) SSP fertilizer.

Figure 6 .
Figure 6.DTG curves of SSP granular and milling samples.

Figure 7 .
Figure 7. DSC curve of the milled SSP_20 min fertilizer sample.

Figure 8 .
Figure 8. Kinetics of solubilization in aqueous medium for different samples of SSP fertilizers.

Table 1 .
Chemical analysis for the SSP fertilizer.

Table 2 .
IR band assignment for gypsum and monocalcium phosphate.

Table 3 .
Summary of TGA and DSC data of SSP fertilizer samples.

Table 4 .
Kinetic parameters of the phosphate release experiments a highest extent.The orbital milling process could be more effective on more brittle phosphate rocks, which are probably more susceptible to a regular particle size reduction without agglomeration. at