Extraction and Puriﬁcation of Phosphorus from the Ashes of Incinerated Biological Sewage Sludge

: Phosphorus depletion represents a signiﬁcant problem. Ash of incinerated biological sewage sludge (BSS) contains P, but the presence of heavy metals (e.g., Fe and Al) is the main issue. Based on chemical characterization by SEM-EDS, ED-XRF and ICP-OES techniques, the characteristics and P content of bottom ash (BA) and ﬂy ash (FA) of incinerated BSS were very similar. On BA, P extraction carried out in counter- current with an S:L ratio of 1:10 and H 2 SO 4 0.5 M led to better extraction yields than those of a similar test with H 2 SO 4 1 M and an S:L ratio of 1:5 (93% vs. 86%). Comparing yields with H 2 SO 4 0.5 M (S:L ratio of 1:10), the counter-current method gave better results than those of the crossﬂow method (93% vs. 83.9%), also improving the performance obtained with HCl in crossﬂow (93% vs. 89.3%). The results suggest that the puriﬁcation of the acid extract from heavy metals with pH variation was impractical due to metal precipitation as phosphates. Extraction with H 2 SO 4 and subsequent treatment with isoamyl alcohol represented the best option to extract and purify P, leading to 81% extraction yields of P with low amounts of metals.


Introduction
On average, in 2020, more than 13 million tons of biological sewage sludge (BSS) were produced in EU27 [1]. This quantity is destined to increase in the coming years due to the expected increase in the number of wastewater treatment plants (WWTPs) due to the stricter limits for the discharge of treated effluents into the water body [2,3]. As for all waste, hierarchical criteria must also be followed for the management of BSS: (i) minimization of production, (ii) material recovery, (iii) energy recovery, and (iv) residual disposal [4].
To date, there are several techniques for minimizing BSS that make it possible to significantly reduce production within the WWTPs [5,6]. In any case, a residual share of BSS will still be produced by biological treatments. These must be valorized first through the recovery of matter and then through energy recovery. To date, one of the most applied techniques for material recovery is spreading in agriculture, which aims to supply the soil with nutrients (mainly C, N, and P) [7,8]. This is essential for the development of crops [9] and is extremely important, especially in territories such as the Mediterranean area (e.g., Spain, Portugal, and Italy), where soil organic matter and nutrient depletion has reached worrying levels [10,11]. Other forms of material recovery, currently a minority in Europe, concern reuse in the construction sector (e.g., bricks and cement production and road construction) [12][13][14] and as adsorbent materials [15].
After material recovery, the hierarchy criteria of the EU requires energy recovery to be adopted before final disposal of the residue [4]. To date, there are several techniques that can be used to recover energy from BSS, namely (i) combustion [16], (ii) pyrolysis [17], and (iii) gasification [18]. The main drawbacks of these technologies are represented by (i) the need for pre-dewatering treatment; (ii) the need for disposal of final residues in combustion, such as bottom ash (BA) and fly ash (FA); and (iii) limited experience of real scale plant management for pyrolysis and gasification [5,[19][20][21]. Recent studies [22,23] evaluated the presence of P within these residues, which could therefore be a potentially exploitable source. P depletion represents a significant problem [24][25][26]. Even if a complete depletion of P deposits is not expected in the short to medium term [27], the significant increase in its use has led to greater attention being paid to the possibility of its recovery and reuse. While BSS-based fertilizers proved to be more suitable alternatives of P sources than chemical fertilizers [28], P is also contained in ashes of incinerated sludge [29], but in this form, it is not available for crops [30]. Moreover, the presence of heavy metals (e.g., Fe and Al) represents a barrier [30]. The evaluation of which techniques can be adopted to separate and valorize P in the ashes resulting from the BSS incineration is fundamental.
Wet extraction of P from incinerated BSS is a widely studied technique [31][32][33][34]. It is based on the application of acids or other leaching agents to extract P from BSS ash and on the subsequent purification of the extract by different techniques, such as ion exchange, filtration, precipitation, and solvent extraction [32]. Xu et al. [35] extracted more than 95% of the total P content from BSS ash as struvite by applying 0.5 mol L −1 of HCl at a liquid/solid ratio of 50 mL g −1 . Moreover, they also found that extracted material had low heavy metal content and high P bioavailability [35]. Fang et al. [33] studied the minimization of P co-precipitation with heavy metals present in incinerated BSS ashes. They found that by using Ca(OH) 2 (at pH = 4) and adopting a two-step extraction method, the presence of heavy metals in the extract significantly decreased (50%) compared to a single-step method of extraction [33].
However, to determine the large-scale application feasibility of P recovery from incinerated BSS ashes, evaluation of how effectively this element can be extracted from the ashes and separated from the other metals contained therein is necessary to optimize the process. This work, therefore, aims to provide a basis on which to develop a future technical-economic analysis of the feasibility of recovering P from ashes of incinerated sludge. BA and FA samples of incinerated BSS were chemically characterized by nondestructive (SEM-EDS and ED-XRF) and destructive (ICP-OES) techniques to quantify the presence of P for potential recovery. On the BA sample, acid extractions with H 2 SO 4 and HCl were performed in crossflow and counter current modes to evaluate the best option. The acid extract was then subjected to two different purification processes (pH variation and addition of isoamyl alcohol) to evaluate which is more effective in separating the P from the metals present within the liquid matrix. The purification of phosphorus from phosphate rocks by organic solvents (specifically isoamyl alcohol) is a known and applied procedure [36,37]; however, the innovative aspect of this work is that the same approach has been applied to purify the phosphorus contained in the acid extract of BA.

Sample Preparation
Combustion residues were obtained from a fluidized bed incineration plant of BSS. BA (1 kg) was sampled from the bottom of the combustion chamber, while FA (1 kg) was sampled from cyclones placed on the fume's treatment line. Before the subsequent analyses, both samples were ground to reduce the particle size (≤200 µm) using a PM100 planetary ball mill (Retsch GmbH, Germany) in 500 mL zirconia jars with zirconia spheres with a diameter of 40 mm and a 20:1 (spheres mass:powder mass) ratio. The plate rotation speed was set to 400 rpm. These analyses were performed coupling an energy dispersion spectroscopy (EDS) X-max 50 mm 2 probe (Oxford Instruments, UK) with an EVO MA10 (Zeiss, Germany) scanning electron microscope (SEM). For the EDS analyses, 2 mm 2 of area was analyzed. In SEM, the samples were analyzed at 8.5 mm and with a voltage of 20 kV.
For the analysis of the matrices by inductively coupled plasma-optical emission spectrometry (ICP-OES), an iCAP7000 (Thermo Fisher Scientific, Waltham, MA, USA) was used following the method developed for the analysis of water and waste by the U.S. EPA [38]. The instrument allowed operation in both axial and radial modes. Based on the relative sensitivity of the spectral line, the measure modes of the instrument were chosen as a function of the expected signal (Table S1). The radiofrequency power was 1150 W. The nebulization gas flow, the cooling gas flow, and the auxiliary gas flow were 0.5 L min −1 , 12 L min −1 , and 0.5 L min −1 , respectively. The peristaltic pump speed and the frequency were maintained at 50 rpm and 500 Hz, respectively.

ED-XRF
A total of 1 g of Li 3 BO 3 was added to the samples (4 g) as a binder. The mixture was then placed in a homogenizer for 5 min. At the end of this phase, the powders were pressed (150 MPa) for 4 min, and 32 mm diameter tablets with a smooth surface were obtained. A Spectro XEPOS ED-XRF spectrometer (Spectro Analytical Instruments GmbH, Germany) was used.

Phosphorus Extraction
P was extracted from BA with crossflow and counter-current methods.

Extraction in Crossflow
A total of 80 g of the BA sample was placed in contact with 240 mL of H 2 SO 4 3 M (95-97%), purchased from Sigma-Aldrich (U.SA), using a Teflon propeller stirrer to keep the resulting suspension homogeneous and dispersed. The suspension was centrifuged after 240 min, and the solid was treated again with 80 mL of H 2 SO 4 3M under stirring for 240 min. This operation was then repeated, thus carrying out three consecutive extractions, maintaining the solid:liquid (S:L) ratios at 1:3, 1:1, and 1:1, respectively. The same tests were also conducted using HCl 3 M (30%) in place of H 2 SO 4 . These tests were performed in duplicate.
To study the optimal P extraction time defining the extraction curves, during the first step of extraction, small aliquots of suspension during and after 30 min, 60 min, 120 min and 240 min were sampled; filtered in a glass microfiber filter (GF/C); and analyzed by ICP-OES.

Extraction in Counter Current
Each portion of the solid sample that enters the process comes out after undergoing five consecutive extractions by an acid that is gradually enriching in the extracted elements. Two processes were set up in parallel, one using H 2 SO 4 1M in an S:L ratio of 1:5 (5 g of sample and 25 mL of H 2 SO 4 ), and one using H 2 SO 4 0.5M in an S:L ratio of 1:10 (3 g of sample and 30 mL of H 2 SO 4 ). Extractions were made in Falcon TM tubes (50 mL) purchased from Thermo Fisher Scientific (USA) moved by a rocking shaker. Each extraction step was kept under stirring for 4 h, and the resulting suspension was subsequently separated by centrifugation. After each extraction cycle, the pH and volume of the resulting liquid phases were measured.

Phosphorus Purification 2.4.1. Purification by pH Variation
A total of 15 g of bottom ash was placed in contact with 150 mL of H 2 SO 4 1 M for 4 h, and, subsequently, the sample was subjected to centrifugation. To purify the P in the form of H 3 PO 4 , the pH of the aqueous phase was varied by adding NaOH 0.25 M (Carlo Erba Reagents, Italy) and causing, at pH 3.7, the precipitation of the metals as hydroxides. The starting aqueous phase (pH < 1) and the aqueous phase with pH 3.7 were then analyzed in ICP-OES.

Purification by Organic Solvent Extraction
As suggested by Israel Mining Industries (IMI) [39,40] and subsequently taken up by others (e.g., [41][42][43]), the purification of phosphorus from phosphate rocks can be achieved with the extraction by organic solvent. In this work, this approach was tested on phosphorus contained in BA.
The procedure was tested on both the HCl (as reported in [39][40][41][42][43]) and H 2 SO 4 extracts. The extract with HCl derived from crossflow extraction kinetics tests (Section 2.3.1). As the H 2 SO 4 extract obtained during the kinetics tests was no longer available, it was obtained by treating BA in crossflow (ratio S:L equal to 1:10) for 4 h.
Acid solutions, after centrifugation, were treated for three consecutive extractions with isoamyl alcohol (purchased from Carlo Erba Reagents, Italy) in a 1:1 ratio. In this case, the three organic phases were kept separate and analyzed by ICP-OES.

FA and BA Characterization
According to the SEM-EDS analysis of both BA and FA, before grinding, BA exhibited a larger particle size than FA according to the literature [44] (Figure 1). Moreover, the content of P in BA and FA is equal to 6.80% and 6.88% (in weight percentage, wt%), respectively, with no substantial difference (Table 1).
Water 2021, 13, 1102 5 of 12 mineralization methods in relation to the recovery of P. From the statistical analysis (ANOVA) of the data related to the P element, the variances in the results cannot be considered the same for 95% of cases (p < 0.05). The results obtained with m1 differed significantly from those obtained with m2, m3, and m4, which on the contrary gave similar results (p > 0.05). Excluding the results of m1, the average value of P obtained with ICP-OES was equal to 5.1 ± 0.1% (P2O5 titer in BA equals 12%) ( Table S2).
The results obtained with IPC-OES were significantly different (p < 0.05) from the data obtained by the non-destructive techniques (SEM-EDS and ED-XRF), which provided an average value of 6.6 ± 0.3% of P content. P content in incinerated sewage sludge ashes was consistent with previous literature results. For instance, Kleeman et al. [45] found 7.2-7.5%, and Franz [29] stated that P content in sewage sludge ash varied from 4% to 9%. Moreover, Donatello et al. [46] studied seven different samples of incinerated BSS ash (precisely FA) and evaluated a P content of 5.7-7.8%.
This P content was decidedly lower than usual content in the minerals (e.g., phosphorites and apatite) used as source to extract P [47][48][49][50]. However, it was considerable and could make this waste a sustainable alternative source to produce phosphate fertilizers. The main barrier to the use of these materials in the conventional cycles of production of P compounds by wet processes is represented by the concentration of Fe (7-15.4% in BA and 15-17.5% in FA) and Al (1.9-4.7% in BA and 3.9-5.2% in FA). This limit was also highlighted by other authors (e.g., [29,34,51]) and, therefore, involves the need for purification post-treatments of the extracted P. These results are therefore particularly interesting for understanding the feasibility of the recovery of P from the ashes of BSS after incineration.   This corresponds to a P titer expressed in P 2 O 5 of approximately 16% for both BA and FA. The characterization using the ED-XRF technique highlighted comparable results detecting a content of 6.4-7.4% of P (equal to a P 2 O 5 titer of 15-16%). Therefore, concerning P composition, the results obtained with the two techniques were similar. The initial  The data obtained using ICP-OES after different mineralization were subjected to statistical analysis to verify the absence of outliers (Q test) and the differences between mineralization methods in relation to the recovery of P. From the statistical analysis (ANOVA) of the data related to the P element, the variances in the results cannot be considered the same for 95% of cases (p < 0.05). The results obtained with m1 differed significantly from those obtained with m2, m3, and m4, which on the contrary gave similar results (p > 0.05). Excluding the results of m1, the average value of P obtained with ICP-OES was equal to 5.1 ± 0.1% (P 2 O 5 titer in BA equals 12%) ( Table S2).
The results obtained with IPC-OES were significantly different (p < 0.05) from the data obtained by the non-destructive techniques (SEM-EDS and ED-XRF), which provided an average value of 6.6 ± 0.3% of P content. P content in incinerated sewage sludge ashes was consistent with previous literature results. For instance, Kleeman et al. [45] found 7.2-7.5%, and Franz [29] stated that P content in sewage sludge ash varied from 4% to 9%. Moreover, Donatello et al. [46] studied seven different samples of incinerated BSS ash (precisely FA) and evaluated a P content of 5.7-7.8%.
This P content was decidedly lower than usual content in the minerals (e.g., phosphorites and apatite) used as source to extract P [47][48][49][50]. However, it was considerable and could make this waste a sustainable alternative source to produce phosphate fertilizers. The main barrier to the use of these materials in the conventional cycles of production of P compounds by wet processes is represented by the concentration of Fe (7-15.4% in BA and 15-17.5% in FA) and Al (1.9-4.7% in BA and 3.9-5.2% in FA). This limit was also highlighted by other authors (e.g., [29,34,51]) and, therefore, involves the need for purification post-treatments of the extracted P. These results are therefore particularly interesting for understanding the feasibility of the recovery of P from the ashes of BSS after incineration.

Phosphorus Extraction
The use of acid attacks to extract P from BA allowed promising results. The results show that the complete equilibrium situation between the solid sample and the extracting acids (H 2 SO 4 and HCl) was reached between 2 h and 4 h (Figure 2a). This result agrees with the findings of Biswas et al. [52]. The result is also consistent with the findings of Donatello et al. [46]. Using H 2 SO 4 , they extracted about 85% of the P present in the incinerated BSS sample in the first 30 min and over 90% after 2h, while the total extraction was observed after approximately only 12 h [46]. However, it is not necessarily convenient to conduct extraction with extremely long acid attack times. For example, Ottosen et al. [30] found that the acid attack with H 2 SO 4 of incinerated sludge for 1 week led to a concentration of metals (i.e., Cu) in the extract greater than that obtained with an attack time of 2 h. incinerated BSS sample in the first 30 min and over 90% after 2h, while the total extraction was observed after approximately only 12 h [46]. However, it is not necessarily convenient to conduct extraction with extremely long acid attack times. For example, Ottosen et al. [30] found that the acid attack with H2SO4 of incinerated sludge for 1 week led to a concentration of metals (i.e., Cu) in the extract greater than that obtained with an attack time of 2 h.
In the present study, the extraction of P by acid attack in three-stage crossflow mode with S:L 1:3, 1:1 and 1:1 ratios granted recovery yields of 83.9% and 89.3% with H2SO4 and HCl, respectively (Figure 2b). Moreover, the results show that the highest share of P was extracted after the first attack with H2SO4 (83.5% of total P extracted) and HCl (88.2% of total P extracted). Taking into consideration a future industrial application, the extraction was conducted also as a five-stage counter current process. This type of extraction was carried out only with H2SO4, with two different S:L ratios (1:5 and 1:10) while maintaining the ratio of ash mass attacked and mole numbers of H2SO4. The results obtained show that the extraction carried out in conditions of a 1:10 S:L ratio and H2SO4 0.5 M led to better extraction yields compared to the similar test with H2SO4 1 M and an S:L ratio of 1:5 (93% vs. 86%) (Figure 3). Therefore, providing a higher volume of acid with the same moles number of H2SO4 allowed us to obtain better results in terms of P extraction. Comparing yields with H2SO4 0.5 M, the counter-current method gave better results than those of the crossflow method (93% vs.83.9%), also improving the performance obtained with HCl in crossflow (93% vs. 89.3%). In the present study, the extraction of P by acid attack in three-stage crossflow mode with S:L 1:3, 1:1 and 1:1 ratios granted recovery yields of 83.9% and 89.3% with H 2 SO 4 and HCl, respectively (Figure 2b). Moreover, the results show that the highest share of P was extracted after the first attack with H 2 SO 4 (83.5% of total P extracted) and HCl (88.2% of total P extracted).
Taking into consideration a future industrial application, the extraction was conducted also as a five-stage counter current process. This type of extraction was carried out only with H 2 SO 4 , with two different S:L ratios (1:5 and 1:10) while maintaining the ratio of ash mass attacked and mole numbers of H 2 SO 4 . The results obtained show that the extraction carried out in conditions of a 1:10 S:L ratio and H 2 SO 4 0.5 M led to better extraction yields compared to the similar test with H 2 SO 4 1 M and an S:L ratio of 1:5 (93% vs. 86%) ( Figure 3). Therefore, providing a higher volume of acid with the same moles number of H 2 SO 4 allowed us to obtain better results in terms of P extraction. Comparing yields with H 2 SO 4 0.5 M, the counter-current method gave better results than those of the crossflow method (93% vs.83.9%), also improving the performance obtained with HCl in crossflow (93% vs. 89.3%).

Phosphorus Purification
The acid extract still contained an important share of Fe and Al, which, overall, exceeded 2 wt% ( Table 2). The use of H2SO4 and the consequent presence of

Phosphorus Purification
The acid extract still contained an important share of Fe and Al, which, overall, exceeded 2 wt% ( Table 2). The use of H 2 SO 4 and the consequent presence of highconcentration SO 4 2− ions led to the precipitation of Ca 2+ as CaSO 4 , thus producing an extract with a much lower concentration of this element than that obtained by extraction with HCl. To obtain a separation between the phosphorus and metals, the acid extracts were subjected to a purification process. The need to proceed with a purification phase due to the co-presence of metals within the extract was also reported by several other studies [46,[52][53][54]. In the present study, two different treatments were tested: (i) separation by pH variation and (ii) separation with organic solvent extraction.

Purification by pH Variation
An onset of precipitation was observed at pH nearly 2. This phenomenon increased with the subsequent additions of NaOH up to a pH of 3.7. The contents of the liquid phase of P, Al, Fe, and Ca decreased drastically with increasing pH (Figure 4). The precipitate was, in fact, constituted by phosphates of Ca, Fe, and Al. It is reasonable to assume that at pH 3.7, the precipitation of Ca 3 (PO 4 ) 2 , FePO 4 , and AlPO 4 sequestered P from the solution, reducing its extraction yield. This result is consistent with the finding of He et al. [55]. Although at this pH range, P occurred in several forms (including H 2 PO 4 − and HPO 4 2− ), it precipitated in the form of complex phosphates [56]. Therefore, the extraction yield of P dropped from 90% to 25%, passing from pH lower than 1 to pH of 3.7. The purification of the acid extract from heavy metals with this technique was therefore impractical.

Purification by Organic Solvent Extraction
Since Fe and Al are elements to be removed in the acid extracts obtained, the purification procedure of isoamyl alcohol addition was tested. The purification tests were conducted on both H 2 SO 4 and HCl acid extracts.
The solution extracted in H 2 SO 4 in contact with isoamyl alcohol did not lead to the separation of two phases unless after adding tridistilled water in the first extraction (10 mL). The extraction of the acid solution from HCl in isoamyl alcohol instead immediately led to the formation of a clear separation surface between the two phases. In terms of P extraction, the addition of isoamyl alcohol on acid extract with H 2 SO 4 allowed us to obtain better results: 81% against 64% obtained from the solutions produced by P extraction with HCl ( Figure 5).

Purification by Organic Solvent Extraction
Since Fe and Al are elements to be removed in the acid extracts obtained, the purification procedure of isoamyl alcohol addition was tested. The purification tests were conducted on both H2SO4 and HCl acid extracts.
The solution extracted in H2SO4 in contact with isoamyl alcohol did not lead to the separation of two phases unless after adding tridistilled water in the first extraction (10 mL). The extraction of the acid solution from HCl in isoamyl alcohol instead immediately led to the formation of a clear separation surface between the two phases. In terms of P extraction, the addition of isoamyl alcohol on acid extract with H2SO4 allowed us to obtain better results: 81% against 64% obtained from the solutions produced by P extraction with HCl ( Figure 5).
The mass of Fe in organic phases after extraction with H2SO4 was clearly lower than that in the case of extraction with HCl, and the mass of P was mostly distributed into the organic phases ( Figure 5). The situation for the mass of P was also favorable with HCl, even if the recovery was lower than in the previous case. In the case of HCl extraction, the mass of Fe was very high (86%), especially in the first organic phase (84.2% of total Fe extracted). It is reasonable to assume that the high concentrations of chlorides led to the formation of undissociated FeCl3 species. For these reasons, extraction with H2SO4 and subsequent treatment with isoamyl alcohol represented the best option to extract and purify P from BA of incinerated BSS sludge.
It is established that the efficiency of extraction and purification of P significantly depends on several parameters, such as the initial matrix, the extraction mode (crossflow or counter current), the type of acid used during extraction and the S:L ratio, the type of purification, and the organic solvents used during L:L extraction. Thus, the present results can hardly be compared to results already present in the literature. However, similar results in different conditions were obtained by other authors. For instance, Hong et al. [54] separated 76% of P present in the acid extract (obtained with HCl extraction) by 1-butanol with an L:L ratio of 1.5:1.

Conclusions
In this work, the characteristics of BA and FA generated by the incineration of BSS were analyzed. According to the SEM-EDS analysis, the P content in BA and FA is equal to 6.80% and 6.88%, respectively. Comparable results were obtained with the characterization using the ED-XRF technique (6.4-7.4%). The results obtained with IPC-OES were significantly different (p < 0.05) from the data obtained by the non-destructive techniques, which provided an average value of 6.6 ± 0.3% of P content. The results show that the complete equilibrium situation between the solid sample and the extracting acids (H2SO4 and HCl) was reached between 2 h and 4 h. Moreover, the results show that in crossflow, the highest share of P was extracted after the first attack with H2SO4 (83.5% of total P extracted) and HCl (88.2% of total P extracted). Extraction carried out in counter current with an S:L ratio of 1:10 and H2SO4 0.5 M led to better extraction yields than those of the similar test with H2SO4 1 M and an S:L ratio of 1:5 (93% vs. 86%). Comparing yields with H2SO4 0.5 M (S:L ratio 1:10), the counter-current method gave better results than those of the crossflow method (93% vs. 83.9%), also improving the performance obtained with HCl in crossflow (93% vs. 89.3%). The results suggest that the purification of the acid extract from heavy metals with pH variation was impractical due to metal precipitation as phosphates. Extraction with H2SO4 and subsequent treatment with isoamyl alcohol represented the best option to extract and purify P, leading to 81% extraction yields of P with a low amount of metals. Other aspects that should be further investigated are (i) the op- Figure 5. Extraction yields of Fe, P and Al from acid phase obtained with crossflow extractions by HCl and H 2 SO 4 , after each of the three sequential extractions (E1, E2, and E3) with isoamyl alcohol. In all extractions, the ratio between organic phase and acid phase was 1:1. The analysis on samples was performed by ICP-OES.
The mass of Fe in organic phases after extraction with H 2 SO 4 was clearly lower than that in the case of extraction with HCl, and the mass of P was mostly distributed into the organic phases ( Figure 5). The situation for the mass of P was also favorable with HCl, even if the recovery was lower than in the previous case. In the case of HCl extraction, the mass of Fe was very high (86%), especially in the first organic phase (84.2% of total Fe extracted). It is reasonable to assume that the high concentrations of chlorides led to the formation of undissociated FeCl 3 species. For these reasons, extraction with H 2 SO 4 and subsequent treatment with isoamyl alcohol represented the best option to extract and purify P from BA of incinerated BSS sludge.
It is established that the efficiency of extraction and purification of P significantly depends on several parameters, such as the initial matrix, the extraction mode (crossflow or counter current), the type of acid used during extraction and the S:L ratio, the type of purification, and the organic solvents used during L:L extraction. Thus, the present results can hardly be compared to results already present in the literature. However, similar results in different conditions were obtained by other authors. For instance, Hong et al. [54] Water 2021, 13, 1102 9 of 11 separated 76% of P present in the acid extract (obtained with HCl extraction) by 1-butanol with an L:L ratio of 1.5:1.

Conclusions
In this work, the characteristics of BA and FA generated by the incineration of BSS were analyzed. According to the SEM-EDS analysis, the P content in BA and FA is equal to 6.80% and 6.88%, respectively. Comparable results were obtained with the characterization using the ED-XRF technique (6.4-7.4%). The results obtained with IPC-OES were significantly different (p < 0.05) from the data obtained by the non-destructive techniques, which provided an average value of 6.6 ± 0.3% of P content. The results show that the complete equilibrium situation between the solid sample and the extracting acids (H 2 SO 4 and HCl) was reached between 2 h and 4 h. Moreover, the results show that in crossflow, the highest share of P was extracted after the first attack with H 2 SO 4 (83.5% of total P extracted) and HCl (88.2% of total P extracted). Extraction carried out in counter current with an S:L ratio of 1:10 and H 2 SO 4 0.5 M led to better extraction yields than those of the similar test with H 2 SO 4 1 M and an S:L ratio of 1:5 (93% vs. 86%). Comparing yields with H 2 SO 4 0.5 M (S:L ratio 1:10), the counter-current method gave better results than those of the crossflow method (93% vs. 83.9%), also improving the performance obtained with HCl in crossflow (93% vs. 89.3%). The results suggest that the purification of the acid extract from heavy metals with pH variation was impractical due to metal precipitation as phosphates. Extraction with H 2 SO 4 and subsequent treatment with isoamyl alcohol represented the best option to extract and purify P, leading to 81% extraction yields of P with a low amount of metals. Other aspects that should be further investigated are (i) the optimization of the purification process; (ii) the research for alternative methods of purification, such as the use of diluted H 3 PO 4 as an extractant to produce H 3 PO 4 enriched in a percentage of P 2 O 5 ; (iii) the study of extracted P availability for crops, with leaching tests; and (iv) cost-benefit analysis of this type of P recovery.