The process of struvite (MgNH4PO4.6H2O) crystallization for removing and recovering nutrients is technically viable. Since the high reagent cost in conventional methods for struvite precipitation comes from the magnesium (Mg2+) chemicals, a reduction in the overall struvite generation cost should be considered to be economically viable in wastewater treatments (LAHAV et al., 2013; CRUTCHIK et al., 2013). In this sense, some studies have been gaining prominence.
According to Huang et al.(2011), the use of renewable and low-cost sources of phosphate (PO43--P) and magnesium is an alternative to reduce the costs of the struvite precipitation process. The authors evaluated the use of decomposed magnesite and brucite as a low cost magnesium source, as well as an alkali reagent to remove ammonium (NH4+-N) as struvite from wastewater generated in the separation process for rare-earth elements. In an economic analysis, they reported that the operating costs of the struvite precipitation process can be reduced to approximately 34% by using decomposed magnesite (MgCO3) instead of pure magnesium chloride (MgCl2). Decomposed MgCO3 shows a higher reaction rate than MgCO3.
Magnesia (MgO) is mostly produced by the calcination of MgCO3. Depending on the thermal conditions, caustic MgO (for temperatures < 900ºC) or dead-burned MgO (for temperatures > 1200ºC) is obtained. Formed at relatively lower temperatures, caustic MgO is a material that is more porous and reactive. It has many industrial applications such as in agriculture, cattle feed, environmental control, manufacture of special cements and in several special uses. Dead-burned MgO produced at high temperatures is mainly utilized by the refractory industry (BIRCHAL et al., 2000).
Magnesium hydroxide (Mg(OH)2) can be found naturally as the mineral brucite, or can be produced from the MgO hydration reaction (VAN DER MERWE et al., 2004). The MgO hydration process has been studied since the 19th century and from these studies it was well established that the reaction mechanism comprises steps of magnesia dissolution with further Mg(OH)2 precipitation. The physical properties of the solid reagent were found to strongly influence this reaction (BIRCHAL et al., 2001).
According to Del Valle-Zermeño et al.(2012), low-grade MgO is eight to ten times cheaper than pure MgO and the revalorization of this by-product by a hydration process is very attractive for applications that require great quantities of Mg(OH)2. The authors have studied the hydration as a function of two parameters: hydrating agent and temperature. It was observed that an increase of temperature had a positive effect on the final degree of hydration as well the use of acetic acid as hydrating agent.
Aphane et al. (2009) evaluated the MgO hydration process at different time intervals in both water and magnesium acetate as hydrating agents. According to these authors, 85% of Mg(OH)2 was obtained from hydration in magnesium acetate after 500 min. For the hydration process in water, a conversion result was equal to 65% after 800 min. The process of MgO hydration should be carefully controlled in order to obtain a high quality product with the desired properties.
Münch & Barr (2001) used 60% Mg(OH)2 slurry to add the Mg2+required for the process of struvite crystallization and to obtain the alkaline pH value of 8.5 in a PO43--P removal study from anaerobic digester sidestreams. The struvite pilot-scale process achieved a PO43--P removal about 94% (from 61 to 4 mg P.L-1).
Recently, the possibilities for cost reduction in the process of struvite crystallization using alternative reagents sources is being widely discussed in literature (SAKTHIVEL et al., 2012; LIU et al., 2013; CRUTCHIK et al., 2013; SICILIANO & De ROSA, 2014). Struvite is a source of nutrients with minimal contamination. It is easy to handle, transport and store, characteristics that support its application as a potential fertilizer (CORNEL & SCHAUM, 2009). Foletto et al. (2013) have been evaluating an innovative application for struvite crystals recovered from swine wastewater treatments. In this study, struvite powder was used as adsorbent to remove an azo leather dye from aqueous solution. As result, the Langmuir and Freundlich isotherm models were fitted to the adsorption data and both satisfactorily represented the process with a maximum adsorption capacity equal to 38.14 mg.g−1.
The present study aims to evaluate the struvite crystallization process using Mg(OH)2 slurries obtained by the hydration of a high-grade industrial MgO compound. The hydration reactions were obtained using water, an organic and an inorganic salt as hydrating agents – magnesium acetate (Mg-Acet) and sodium chloride (NaCl), respectively. It is a feasible and is an alternative process to recover nutrients, generating an interesting product without potential health risks and with ecological criteria, mainly when integrated into a wastewater treatment line.
2. Materials and methods
First, the MgO hydration reactions were performed under different evaluated conditions in order to obtain Mg(OH)2 slurries that were properly characterized and used as Mg2+ and alkali sources in the struvite crystallization reactions.
A high-grade industrial MgO compound sample, Table 1, from Magnesita S.A. mine in Brumado, Bahia State, Brazil, was used in all assays to obtain the Mg(OH)2 slurries at different conditions. According to the producer, it was calcined at (700 ± 100)ºC and has 35% of particles as maximum residue in a 200 mesh sieve size with BET surface area equal to 21.91 m2.g-1.
|Parameters||MgO||Fe2O3||SiO2||MnO||CaO||Al2O3||Lost on Ignition|
|Mass content (%)||98.3||0.4||0.3||0.1||0.8||0.1||< 2.0|
The MgO hydration reactions were performed in triplicate by reacting 40% w w-1 of solids in a 500 mL of 0.1 mol L-1 hydrating solution (APHANE et al., 2009; MATABOLA et al., 2010). The reactions took place in a 1 L glass beaker immerse in a thermostatic water bath at 80 ± 2ºC and continuous mechanical stirring of 200 rpm.
The hydrating agents used were an organic and an inorganic salt, Mg-Acet and NaCl, respectively, besides a test containing only distilled water. Samples were taken at times of 0, 15, 30, 60, 120 and 360 min; immediately cooled and subjected to analysis (BIRCHAL et al., 2001).
The struvite crystallization was evaluated from a synthetic solution containing 500.0 mg N.L-1 at operational conditions established by an experimental design according to Castro et al. (2013) - pH equal to 8.5, Mg2+:NH4+:PO43- ratio equal to 1.5:1.0:1.25 and about 10 minutes reaction time.
The assays were performed in triplicate using analytical grade chemicals of NH4+-N and PO43--P, NH4Cl and Na2HPO4.7H2O, respectively, in addition to the Mg(OH)2 slurries obtained in the previous step.
The hydration reactions were conducted and all samples collected during the process were subjected to pH analysis (Digimed DM-2P pHmeter) and calculation of the MgO conversion rate to Mg(OH)2 (%). This last one was determined according to the variation of the weight of the solids according to Eq. 1, where mf and mi are the final and initial weights of the solids, respectively, and M is the molecular weight.
The Mg(OH)2 slurries were characterized by the citric acid method for reactivity determination (STRYDOM et al., 2005), together with density and total suspended solids analyses (TSS) in order to perform the proper stoichiometric dosage of Mg2+ required in the crystallization reactions (APHA/AWWA/WEF, 1998). With the characterized Mg(OH)2slurries, the struvite crystallization experiments were performed under conditions previously mentioned.
The solid samples obtained were filtered (AP40 Millipore filter), washed with distilled water, dried at 40ºC until constant weight and then characterized by X-Ray Diffraction (XRD) and Scanning Electronic Microscopy (SEM) coupled with Energy Dispersive X-Ray analysis (EDS) analysis.
The nutrient recovery and struvite crystallization was evaluated by nitrogen (NH4+-N) and phosphorus (PO43--P) analysis in the supernatant liquid, carried out according to Standard Methods, titulometric 4500-NH3 and colorimetric 4500-P methods, respectively (APHA/AWWA/WEF, 1998). The analytical methods were performed in triplicate.
3. Results and discussion
Some results of the MgO hydration reaction are shown in Figure 1, the kinetic analysis of MgO conversion to Mg(OH)2 and citric acid reactivity under differently evaluated conditions: Mg-Acet, H2O and NaCl solutions. In all of these situations, the conversion profile increased with time and reached values equal to 67.4 ± 1.6, 56.9 ± 1.8 and 58.5 ± 1.0 % respectively at 360 minutes reaction time, Figure 1a.
The optimum hydration condition was observed using a 0.1 mol L-1 Mg-Acet solution as hydrating agent. At the same reaction temperature, 80ºC, a similar result was obtained by Matabola et al. (2010) on an assessment of the influence of different hydrating agents in the MgO hydration reaction. According to these authors, the increased hydration degree from the reactions performed in Mg Acet can be attributed to the presence of acetate and additional Mg2+ ions which increases the MgO solubility by increasing the pH. It was also reported that the presence of Mg2+ ions accelerates the reaction rate (VAN DER MERWE et al., 2004).
A hypothesis test was used in a statistical analysis of MgO hydration results. The T Test was performed to compare the mean values obtained under each hydration condition with significance level (α) equal to 0.05. It was observed that the conversion rate obtained for the optimum hydration condition (in Mg-Acet solution) is significantly higher than the other two conditions evaluated. On the other hand, the observed difference between the mean values of hydration in H2O and NaCl solutions, 56.9 and 58.5% respectively, are not convincing enough to state that the conversion rate of MgO to Mg(OH)2 differs significantly in function of these hydrating agents.
In Figure 1a, the experimental data can be compared to a model proposed by Birchal et al. (2001), adapted to the reaction temperature of this study. Some adjustment is observed for the best condition of hydration. This kinetic model assumes no diffusive contribution and incorporates an additional resistance to the reactive flux, due to the transient variation of the porosity of the material during the MgO hydration reaction. According to this model, the porosity changes with time, and the obtained hydroxide particles deposit in the MgO pores; so this film of hydroxide imposes an increase in resistance to the hydration process (ROCHA et al., 2004).
Using water as a hydrating agent for the MgO conversion to Mg(OH)2, the values were similar to those obtained by Aphane et al. (2009) at the 360-minute reaction time with assays performed at a temperature of 80ºC. According to these authors, the same maximum hydration degree was reached for MgO samples with different reactivity degree. However, the time needed for maximum hydration differs for each agent, depending on the hydration agent. Also, according to these authors, the hydration rate is influenced by the calcination conditions and reactivity (or different surface areas). It indicates that the MgO hydration process proceeds with a physical step as the rate-determining mechanism.
The citric acid reactivity analysis corroborates the conversion results. Due to the higher MgO conversion and, consequently, higher content of Mg(OH)2, a shorter time was obtained for the slurry obtained in Mg-Acet solution, equal to 59.3 ± 0.5 seconds. According to Van der Merwe (2004), samples that exhibit reactivity in citric acid lower than 60 seconds are classified as highly reactive. An increase in reactivity with higher conversion rates is expected; in other words, the time required for the complete acid neutralization decreases.
In Table 2, the characterization results of the obtained slurries are presented. It was observed that with the different results obtained for the MgO conversion to Mg (OH)2 and TSS analysis, different amounts of each slurry should be used to achieve the same established stoichiometric conditions in the struvite crystallization reactions from the synthetic solution with 500 mg N.L-1.
|Parameter / Mg(OH)2 Slurry||Mg-Acet||H2O||NaCl|
|Density (kg L-1)||1.43||1.52||1.43|
|Conversion to Mg(OH)2 (%)||67.4||56.9||58.5|
|Mg(OH)2 (kg) / Slurry (kg)||0.341||0.338||0.297|
The struvite crystallization reactions have been carried out under the optimum conditions established by the experimental design mentioned previously: pH equal to 8.5 and Mg2+:NH4+:PO43-stoichiometric ratio equal to 1.5:1.0:1.25, from the synthetic solution (CASTRO et al., 2013)
The performance of the Mg(OH)2 slurries obtained in the hydration reaction as alkalinity and Mg2+ ions source on the struvite crystallization process was evaluated according to the calculation of nutrients removal. The results are given in Figure 2.
It was observed that the struvite crystallization reaction performed from the obtained slurry in Mg Acet solution allowed the best nutrients removal, about 62.3 ± 1.7 % and 67.2 ± 1.8 % of NH4+-N and PO43--P, respectively, in a single-stage process. Additional Mg2+ ions in the acetate solution probably contributed to these results. With the slurry obtained in NaCl the crystallization reaction promoted the lower removals, with similar values by using pure H2O. According to Kabdasli et al.(2006) the presence of Na+ ions contributes in increasing the induction time in the struvite crystallization process.
A calculation was performed to comparatively evaluate the Mg2+ reagent consumption. The slurry obtained in Mg-Acet solution, the optimum condition of nutrient removal, reached a ratio equal to 2.2 mol of consumed Mg2+ per mol of NH4+-N precipitated. By a stoichiometric analysis of the struvite, a value closer to 1.0 corresponds to an optimum condition of less Mg2+ reagent consumption. However, as the reaction occurs under supersaturation conditions, a stoichiometric excess ensures the struvite crystallization. For slurries obtained in H2O and NaCl solution, this ratio was higher and reached values equal to 2.6 and 2.7, respectively. It shows a higher Mg2+ consumption and therefore a higher cost in the process regarding this reagent.
In general, the results obtained by the crystallization reactions performed with the three different tested slurries of Mg (OH)2 were interesting in terms of the nutrient removal; however, better results can be achieved by optimizing the experimental conditions and parameters in an experimental design. As alkalinity sources, the slurries acted effectively, being capable of raising the pH to the predetermined value, 8.5, without the need for any further adjustment using another reagent.
In all of the evaluated conditions, the stoichiometry of the recovery ions suggests the formation of struvite with a molar ratio N:P around 1.0. Thus, the analysis of the generated solids from the crystallization reactions using the Mg(OH)2 slurries confirmed a majority presence of the struvite mineral. The results of SEM and XRD analysis are shown in Figure 3.
In Figure 3a, the images obtained by SEM analysis have shown, in all three cases, small crystalline formations in an orthorhombic system characteristic of struvite crystals. Small crystals were obtained due to the fast reaction time for the crystallization, 10 ± 3 minutes, which did not favor their growth, in addition to the shear forces generated by a rapid mechanical stirring - 150 rpm.
A semi-quantitative analysis confirmed in the three cases a purity degree above 60% for the obtained struvite crystals. This result was most pronounced in solids obtained by the slurry in Mg Acet solution, equal to 74%. In the light of the conversion rate obtained in the hydration reaction of MgO in Mg-Acet solution, a smaller amount of this slurry was required for the struvite crystallization, which was added to the reaction medium to achieve the predetermined stoichiometric ratio of Mg2+:NH4+:PO43-. This fact may have contributed to the purity degree of the crystals. In some samples, traces of impurities were identified by EDS analysis, such as Si, Al, Fe and Ca ions that may have originated from the MgO compound used.
In Figure 3b the formation of struvite crystals is qualitatively confirmed by the XRD analysis, according to Yilmazel & Demirer (2013) study. All the cases have characteristic peaks of the struvite pattern retrieved from powder diffraction (PDF) card 71-2089.
It can be concluded that the hydration reactions using Mg-Acet, H2O and NaCl as hydrating agents, promoted the MgO conversion to Mg(OH)2. It reached a maximum value equal to 67.4% in acetate solution at 360-minutes reaction time. The Mg(OH)2 slurries presented good results as alkalinity and Mg2+ions on the struvite crystallization reactions. An evaluation of this reaction in terms of the NH4+-N and PO43--P removals showed that the crystallization using the slurry obtained in Mg-Acet reached the highest removals of about 62.3± 1.7 % and 67.2 ± 1.8 % of NH4+-N and PO43--P, respectively. The formation of small struvite crystals was proved by the instrumental analysis. A considerable purity degree was observed, even more pronounced in solids obtained by using the slurry in Mg-Acet solution. Overall, struvite crystallization is a feasible process when using Mg(OH)2 slurries obtained through a hydration reaction of a low-cost industrial MgO compound. It can be used as an alternative source of reagent in recovering nutrients and obtains a solid with potential for use as fertilizer. It has been also noted that the optimization of parameters and operational conditions by using statistical tools may contribute towards improving the results and minimizing the process costs.