Adsorption of Phosphate and Ammonium on Waste Building Sludge

Two selected waste building sludges (WBS) were used in this study: (i) sludge from the production and processing of prestressed concrete pillars (B) and (ii) sludge from the production of technical stone (TS). The materials were used in their original and Fe-modified forms (BFe/TSFe) for the adsorption of NH4+ and PO43− from contaminated waters. The experiments were performed on a model solution simulating real wastewater with a concentration of 1.7 mmol·L−1 (NH4+) and 0.2 mmol·L−1 (PO43−). The adsorption of PO43− had a high efficiency (>99%) on B, BFe and TSFe, while for TS, the adsorption of PO43− was futile due to the high content of available P in the raw TS. The adsorption of NH4+ on all sorbents (B/BFe, TS/TSFe) had a lower efficiency (<60%), while TS proved to be the most effective. Leaching tests were performed according to the CSN EN 12457 standard for B/BFe and TS/TSFe before and after NH4+ and PO43− sorption when the contents of these ions in the leachates were affected by adsorption experiments in the cases of B and TS. For BFe and TSFe, the ion content in the leachates before and after the adsorption experiments was similar.


Introduction
In developed countries, the construction industry can create environmental issues, such as the depletion of natural resources and the production of several tons of construction waste [1,2]. Construction waste in various branches of the construction industry (the production of concrete, artificial stone, etc.) also includes dried powder building sludge (WBS), which is defined as a very fine material that is dispersed in water [1,2].
Concrete is one of the most widely used building materials, with an annual global consumption of 25 billion tons [3]. At present, various separation and recycling processes are used in its production, enabling the reuse of water and coarse aggregates [2,4,5]. The remaining concrete sludge (fine aggregates and cement particles) can be used in the production of ceramic materials [6], synthesis of geopolymers [7], or in the production of new concrete to reduce the required amount of cement [1,3,5].
All of these processes are relatively effective but insufficient for modern sustainable development. This is because the remaining concrete sludge (B) is landfilled without further use, along with several other wastes from the construction industry, such as powder waste from the production and treatment of technical stone (TS), which currently has no other application [4]. However, sewage sludges have a large specific surface area (S BET ), suitable structural properties and chemical composition (Si, Ca, Al and Fe content), which predetermine their possible applications in environmental technologies, for example, as adsorbents for removing toxic ions from contaminated waters [4,8,9].
Nitrogen and phosphorus in NH 4 + and PO 4 3− ionic forms are an integral part of living organisms and plants [3,8,10]. Both elements are important for good plant growth and development and are often applied in the form of fertilizers to satisfy the growing requirement for food, but high concentrations of NH 4 + and PO 4 3− in water result in excessive algae growth, which consumes dissolved oxygen and kill fishes and other organisms 2 of 11 living in the water (water eutrophication) [3,8,[10][11][12]. High concentrations of NH 4 + and PO 4 3− enter into natural streams from various sources, such as agricultural effluents, industrial wastewater and domestic wastewater [3,8,[10][11][12]. Addressing the issue of declining reserves of mineable phosphate ore requires new solutions for capturing and reusing phosphates from wastewater [3,[10][11][12]. Several types of absorbents (e.g., biochar, fly ashes, iron-enriched zeolites, etc.) have been developed for the regeneration of phosphates from wastewater [3,[10][11][12]. The adsorption of NH 4 + was studied, for example, on a polyurethane film prepared from ball-milled algal polyol particles to maintain low concentrations of this ion in fish and shrimp breeding tanks [13]. The coadsorption of NH 4 + and PO 4 3− in wastewater was not discussed in these studies.
As part of this study, selective, simultaneous and additional adsorptions of NH 4 + and PO 4 3− were monitored. The experimental data were fitted by the Langmuir and Freundlich adsorption isotherm to determine the sorption parameters (q max. -maximum equilibrium adsorption capacity, Q t -theoretical adsorption capacity, K L -Langmuir adsorption constant, R 2 -correlation factor, 1/n-heterogeneity factor, K F -Freundlich constant indicating adsorption capacity). The Langmuir adsorption isotherm is the simplified sorption model, which assumes the equivalence and even distribution of the active sites, to which only one series of non-interacting molecules can be bound [14][15][16]. The Freundlich adsorption isotherm is the first known model describing reversible multilayer adsorption with a different distribution of active sites [17]. Kinetic measurements were performed for NH 4 + and PO 4 3− adsorption, and the data for systems that could be described by the Langmuir model (NH 4 + -TS, PO 4 3− -B, PO 4 3− -B Fe and PO 4 3− -TS Fe ) were processed by pseudo-firstand pseudo-second-order formal kinetic models to find appropriate rate constants (k 1 for pseudo-first-order formal kinetic and k 2 for pseudo-second-order formal kinetic) [18].
The goal of this study was to find new possible applications of B and TS in their original and surface-modified forms (B Fe and TS Fe ) for the coadsorption of NH 4 + and PO 4 3− ions from wastewater and their subsequent use for improving the quality and nutritional values of agricultural soils.

Characterization of Used Building Waste Sludge
The WBS from the production of concrete (B) and artificial stone (TS) with a particle size of <0.1 mm was used. The B is formed during the production and abrasion of prestressed concrete columns, with a high cement content of~21%. The TS is created during the production and processing of technical stone from Technistone, Czech Republic. The mineralogical and elemental composition of both materials were determined using X-ray powder diffraction (XRD) and X-ray fluorescence analysis (XRF), and the results are discussed further in Section 3.1.
For the selective sorption of anions, the surfaces of B and TS were modified with Fe 2+ ions (B Fe , TS Fe ) according to the verified method [19][20][21][22]. The surface modification was performed with 0.6 M FeSO 4 solution for 24 h at the laboratory temperature (20 • C) upon stirring the mixture with a shaker. Then, the suspension was filtered, and the obtained modified sludge was washed with distilled water, dried (60 • C) and homogenized.

Model Solution
The ion concentrations in the model solutions were chosen according to the real values in the wastewater (pond from the contaminated area) in the Havlíčkův Brod vicinity (Czech Republic-Highlands).
Model solutions of selected ions and their mixture were prepared in the concentration of 1.7 mmol·L −1 NH 4 + and 0.2 mmol·L −1 PO 4 3− . The solutions were prepared from analytically pure inorganic salts NH 4 Cl, K 2 HPO 4 and distilled water at the original pH (~7.5).
Distilled water, tap water and 0.1 M KCl were used for leaching experiments.

Adsorption Experiments
The suspension of a defined amount of sorbent (5-40 g·L −1 ) and 50 mL of model solution was shaken in 100 mL sealed polyethylene containers for 24 h (chosen based on preliminary experiments) at laboratory temperature (20 • C), pH of the model solution (~7.5) and at a speed of 280 rpm. Subsequently, vacuum filtration was performed on 0.6 µm pore size filters. The residual NH 4 + and PO 4 3− concentrations in the obtained filtrates were analyzed. The experimental data were fitted by the Langmuir and Freundlich adsorption isotherm to determine the sorption parameters (q max. , Q t , K L , 1/n, K F , R 2 ). The accuracy of fitted data was supported by the triple measurement of the adsorption series.
The Langmuir isotherm is defined by Equation (1) [14][15][16]: and its linearized form by Equation (2) [14-16]: where q is an equilibrium concentration of an adsorbed ion in the solid phase [mmol·g −1 ], c is an equilibrium concentration of an adsorbed ion in the solution [mmol·L −1 ], Q t is the theoretical adsorption capacity [mmol·g −1 ], and K L is a Langmuir adsorption constant [L mmol −1 ]. The equilibrium ion concentration in the solid phase was calculated from the experimental data according to Equation (3) [14][15][16]: where V 0 is the volume of solution [L], c 0 is the initial concentration of adsorbate in solution [mmol·L −1 ], and m is the mass of the solid phase [g]. The Freundlich isotherm is defined by Equation (4) [17]: and its linearized form by Equation (5) [17]: where q is an equilibrium concentration of an adsorbed ion in the solid phase Kinetic data for the systems that could be described by the Langmuir model (NH 4 + -TS, PO 4 3− -B, PO 4 3− -B Fe and PO 4 3− -TS Fe ) were processed by the pseudo-first-and the pseudo-second-order formal kinetic models to find rate constants (k 1 and k 2 ) [18].
The pseudo-first-order kinetic model is described by Equation (6) [18]: Integrated Equation (4) and substituted the boundary conditions from t = 0 to t = t and q t = 0 to q t = q t , a linearized equation was obtained (Equation (7) [18]: ln q e − q t = ln q e − k 1 t (7) The pseudo-second-order kinetic model is described by Equation (8) [18]: Integrated Equation (6) and substituted the boundary conditions from t = 0 to t = t and q t = 0 to q t = q t , a linearized equation was obtained (Equation (9) [18]:

Leaching Tests
The leaching of both ions from the original and saturated WBS was performed according to the CSN EN 12,457 standard [23]. The defined amounts of B and TS before and after the sorption of NH 4 + and PO 4 3− were poured with the appropriate leaching solution (Section 2.2) at the solid-liquid ratio of 1:10.

Analytical Methods
X-ray powder diffraction (XRD) of solid samples was measured using a 2D Phaser (Bruker s.r.o., Billerica, MA, USA). A current of 10 mA, a voltage of 30 kV, a step size of 0.02 • and a range of angles (6-80 2θ) were used for the measurements.
The semi-quantitative chemical composition was determined by X-ray fluorescence analysis (XRF), which was performed using a NEX QC instrument (Rigaku Company, Tokio, Japan), where the powder sludge was measured at 50 kV using an SDD detector.
Zero-charge pH (pH zpc ) was measured using the Stabino ® , Version 2.0 (Particle Metrix GmbH, Inning am Ammersee, Germany). The stabilized suspensions of the solid sample and 0.1, 0.01 and 0.001 M KCl (solid: liquid ratio of 1:100) were dynamic with 0.1 M solution of NaOH or HCl to the isoelectric point (IEP). The resulting pHzpc value is the average of three pH values corresponding to the zero potential.
The Micromeritics ASAP 2020 (accelerated surface area and porosimetry) analyzer (Micromeritics ® , Norcross, GA, USA) was used to measure the specific surface area (S BET ) of the sludge used, which uses gas sorption (N 2 ) to study macropores and micropores using the Horvath-Kavazoe method (BJH method) bath at −195.8 • C. Prior to measurement, the samples were degassed at 313 K for 1000 min. NH 4 + and PO 4 3− concentrations were determined by UV/Vis spectrophotometry using an Evolution 220 instrument (Thermo Scientific ® , Waltham, MA, USA) at 425 nm for NH 4 + using potassium sodium tartrate and Nessler reagent [24], and at 820 nm for PO 4 3− using the molybdenum blue method [25]. properties were changed by the modification with Fe 2+ ions; the B Fe and TS Fe significantly differed in S BET , Fe and alkali content (Table 1), which affected PO 4 3− and NH 4 + adsorption. The chemical and surface properties of B, TS, B Fe and TS Fe are listed in Table 1.

Characterization of Original and Modified B/BFe and TS/TSFe
The XRD diffractograms for BFe and TSFe were identical to their original forms of B and TS (Figure 1) because Fe oxides were bound to the silicate skeleton of B or TS by chemisorption in an amorphous form during the modification of Fe 2+ ions when hydrated metal particles formed on the surface of the sorbents (BFe, TSFe) in reactive, ion-exchangeable positions and there were no changes in mineralogical composition [17,18]. The chemical and surface properties were changed by the modification with Fe 2+ ions; the BFe and TSFe significantly differed in SBET, Fe and alkali content (Table 1), which affected PO4 3− and NH4 + adsorption. The chemical and surface properties of B, TS, BFe and TSFe are listed in Table 1.

Adsorption of the Selected Ion (NH4 + or PO4 3− ) on Original and Modified B/BFe and TS/TSFe
All adsorption experiments were performed under the same conditions described in Section 2.3. Figure 2 shows the dependence of adsorption efficiencies ε (%) on the weight m (g·L −1 ) of B/BFe and TS/TSFe for NH4 + or PO4 3− . Table 2 shows the sorption parameters (theoretical adsorption capacities-Qt; adsorption constants-KL and KF; heterogeneity factor-1/n; root mean squared error-RMSE) calculated using the Langmuir and Freundlich model [14][15][16].  All adsorption experiments were performed under the same conditions described in Section 2.3. Figure 2 shows the dependence of adsorption efficiencies ε (%) on the weight m (g·L −1 ) of B/B Fe and TS/TS Fe for NH 4 + or PO 4 3− . Table 2 shows the sorption parameters (theoretical adsorption capacities-Q t ; adsorption constants-K L and K F ; heterogeneity factor-1/n; root mean squared error-RMSE) calculated using the Langmuir and Freundlich model [14][15][16]. PO 4 3− adsorption occurred with high efficiency (<99%) on modified forms B Fe and TS Fe (Figure 2b). Due to its high alkalinity, B did not primarily support the adsorption of anions. The high efficiency of PO 4 3adsorption on B can be explained by the precipitation of PO 4 3− into a poorly soluble amorphous form or as apatite (Ca 5 (PO 4 ) 3 (OH)). Modified forms of B Fe (Figure 2b, orange line) and TS Fe (Figure 2b, red line) achieved high sorption efficiencies with PO 4 3− because they were enriched with hydrated metal particles in reactive, ion-exchangeable surface positions (Section 3.1). These available Fe ions are sufficient for the adsorption of an oxyanion such as PO 4 3− onto Fe oxy(hydroxides). The TS released PO 4 3− into the solution, where the concentration of this ion increased by more than 50% at the highest dosage of sorbent (Figure 2b   The adsorption of PO4 3− on B, BFe and TSFe and NH4 + on TS corresponded to both the Freundlich and Langmuir models, but the worse correlation of experimental data for the Freundlich model (R 2 : 0.496-0.956 versus 0.897-0.999, Table 2) indicated the Langmuir isotherm more appropriate for investigated systems. The NH4 + adsorption on BFe a TSFe followed the Freundlich model but with very low correlation factors.  Kinetic experiments were performed under the same conditions described in Section 2.3. The dependence of concentration q t (mmol·g −1 ) of an adsorbed ion (NH 4 + or PO 4 3− ) in the solid phase on the contact time t (h) is shown in Figure 3. Kinetic experiments were performed under the same conditions described in Section 2.3. The dependence of concentration qt (mmol·g −1 ) of an adsorbed ion (NH4 + or PO4 3− ) in the solid phase on the contact time t (h) is shown in Figure 3. The PO4 3− and NH4 + adsorption equilibrium was reached around 19 h (Figure 3). The obtained rate constants (k1 and k2) and correlation factors (R 2 ) for the pseudofirst-and the pseudo-second-order formal kinetic models, which were used for the systems that could be fitted by the Langmuir model (Section 2.3), are reported in Table 3. The PO 4 3− and NH 4 + adsorption equilibrium was reached around 19 h (Figure 3). The obtained rate constants (k 1 and k 2 ) and correlation factors (R 2 ) for the pseudo-firstand the pseudo-second-order formal kinetic models, which were used for the systems that could be fitted by the Langmuir model (Section 2.3), are reported in Table 3. Table 3. Correlation factors (R 2 ) and velocity constants (k 1 and k 2 ) of the pseudo-first-order kinetics model and the pseudo-second-order kinetics model. Adsorption systems that could be fitted to the Langmuir model (PO 4 3− -B, PO 4 3− -B Fe , PO 4 3− -TS Fe and NH 4 + -TS) proceeded by chemisorption, according to the pseudo-secondorder kinetic model (Section 2.3). The other studied systems did not correlate sufficiently with any of the applied adsorption models, and prevailing physical adsorption could be assumed.

Additional Adsorption of NH 4 + and PO 4 3− on Original and Modified B/B Fe and TS/TS Fe
In order to determine the possible accumulation of NH 4 + or PO 4 3− and the effect of adsorbed NH 4 + or PO 4 3− on the possibility of further sorption, the most effective systems of selected adsorption (Section 3.2) were saturated with the oppositely charged ion. Figure 4 compares the sorption efficiencies of the ions adsorbed in the selective sorption (Sec.) and in the additional sorption (Add.) on the oppositely charged ion captured on the sorbent surface during the prior selective sorption (Section 3.2). Additionally, the PO 4 3− -B, PO 4 3− -B Fe and PO 4 3− -TS Fe systems were used for NH 4 + adsorption, while for the PO 4 3− adsorption, only the NH 4 + -TS system was used.  During the additional adsorption, the sorption efficiency increased from 6% for adsorption NH4 + on PO4 3− -TSFe system (Figure 4b) to 60% for adsorption of PO4 3− on the NH4 + -TS system (Figure 4a) because active sites formed on the surfaces of the formerly saturated sorbents with NH4 + or PO4 3− , causing the additional binding of oppositely charged ions, whereby the adsorption yield of additional adsorption increased. These active sites also supported the accumulation of nutrients in the sorbents for possible applications in agricultural soils.

Simultaneous Adsorption of NH4 + and PO4 3− on Original and Modified B/BFe and TS/TSFe
The tested ions can usually coexist in real water systems; therefore, their simultaneous sorptions (Sim.) on B, TS, BFe and TSFe were performed. Figure 5 shows the dependence of the adsorption efficiencies on the dosage of B/BFe and TS/TSFe for NH4 + (Figure 5a) and PO4 3− (Figure 5b) adsorption when the data obtained in this sorption experiment are During the additional adsorption, the sorption efficiency increased from 6% for adsorption NH 4 + on PO 4 3− -TS Fe system (Figure 4b) to 60% for adsorption of PO 4 3− on the NH 4 + -TS system (Figure 4a) because active sites formed on the surfaces of the formerly saturated sorbents with NH 4 + or PO 4 3− , causing the additional binding of oppositely charged ions, whereby the adsorption yield of additional adsorption increased. These active sites also supported the accumulation of nutrients in the sorbents for possible applications in agricultural soils.

Simultaneous Adsorption of NH 4 + and PO 4 3− on Original and Modified B/B Fe and TS/TS Fe
The tested ions can usually coexist in real water systems; therefore, their simultaneous sorptions (Sim.) on B, TS, B Fe and TS Fe were performed. Figure 5 shows the dependence of the adsorption efficiencies on the dosage of B/B Fe and TS/TS Fe for NH 4 + (Figure 5a) and PO 4 3− (Figure 5b) adsorption when the data obtained in this sorption experiment are compared with the sorption efficiencies of selective ion sorption (Sec.) mentioned in Section 3.2.
sorption NH4 + on PO4 3− -TSFe system (Figure 4b) to 60% for adsorption of PO4 3− on the NH4 + -TS system (Figure 4a) because active sites formed on the surfaces of the formerly saturated sorbents with NH4 + or PO4 3− , causing the additional binding of oppositely charged ions, whereby the adsorption yield of additional adsorption increased. These active sites also supported the accumulation of nutrients in the sorbents for possible applications in agricultural soils.

Simultaneous Adsorption of NH4 + and PO4 3− on Original and Modified B/BFe and TS/TSFe
The tested ions can usually coexist in real water systems; therefore, their simultaneous sorptions (Sim.) on B, TS, BFe and TSFe were performed. Figure 5 shows the dependence of the adsorption efficiencies on the dosage of B/BFe and TS/TSFe for NH4 + (Figure 5a) and PO4 3− (Figure 5b)

Leaching Experiments
The leaching experiments (described in Section 2.4) were performed to determine the possible use of both sludges (B and TS) as additives to agricultural soils to improve their quality. Figures 5 and 6 show the amounts of NH 4 + /PO 4 3− ions leached from the individual sludges (B/B Fe and TS/TS Fe ) before ( Figure 6) and after ( Figure 7) the adsorption of selected ions.
The leaching experiments revealed a relatively high release of PO 4 3− (Figures 6b and 7b) and NH 4 + (Figures 6a and 7a) from saturated and original sorbents, B and TS. For the B and TS, the leaching tests also showed that the leaching of PO 4 3− and NH 4 + was affected by the saturation of the PO 4 3− or NH 4 + on the sorbent surface (PO 4 3− and NH 4 + adsorption is discussed in Sections 3.2-3.4).
The B Fe and TS Fe were able to leach significantly lower contents than their original forms B and TS, and due to their affinity for oxyanions, PO 4 3− was almost not leached (Figures 6b and 7b, yellow and grey lines). The

Leaching Experiments
The leaching experiments (described in Section 2.4) were performed to determine the possible use of both sludges (B and TS) as additives to agricultural soils to improve their quality. Figures 5 and 6 show the amounts of NH4 + /PO4 3− ions leached from the individual sludges (B/BFe and TS/TSFe) before ( Figure 6) and after ( Figure 7) the adsorption of selected ions.
(a) (b) The leaching experiments revealed a relatively high release of PO4 3− (Figures 6b and  7b) and NH4 + (Figures 6a and 7a) from saturated and original sorbents, B and TS. For the B and TS, the leaching tests also showed that the leaching of PO4 3− and NH4 + was affected by the saturation of the PO4 3− or NH4 + on the sorbent surface (PO4 3− and NH4 + adsorption is discussed in Sections 3.

Leaching Experiments
The leaching experiments (described in Section 2.4) were performed to determine the possible use of both sludges (B and TS) as additives to agricultural soils to improve their quality. Figures 5 and 6 show the amounts of NH4 + /PO4 3− ions leached from the individual sludges (B/BFe and TS/TSFe) before ( Figure 6) and after ( Figure 7) the adsorption of selected ions.
(a) (b) The leaching experiments revealed a relatively high release of PO4 3− (Figures 6b and  7b) and NH4 + (Figures 6a and 7a) from saturated and original sorbents, B and TS. For the B and TS, the leaching tests also showed that the leaching of PO4 3− and NH4 + was affected by the saturation of the PO4 3− or NH4 + on the sorbent surface (PO4 3− and NH4 + adsorption is discussed in Sections 3.

Conclusions
B, B Fe and TS Fe proved to be promising sorbents for the sorption of PO 4 3− when such adsorptions were successfully fitted by the Freundlich and Langmuir adsorption models, with better parameters for the Langmuir fit. The TS spontaneously released PO 4 3− into the solution, and no adsorption occurred.
The adsorption of NH 4 + had a lower efficiency compared to the sorption of PO 4 3− , while the TS was found to be the most efficient sorbent. The adsorption of NH 4 + on the TS could be fitted by the Freundlich and Langmuir adsorption models when better correlation factors were achieved for the Langmuir fit. The NH 4 + adsorption on B Fe and TS Fe followed the Freundlich model but with very low correlation factors. Adsorption of NH 4 + proceeded with a lower sorption robustness compared to the PO 4 3− adsorption. The kinetic equilibrium for PO 4 3− and NH 4 + adsorption was reached around 19 h. For the selected adsorption systems that could be fitted by the Langmuir model (PO 4 3− adsorption on B, B Fe and TS Fe and NH 4 + adsorption on TS), the pseudo-second-order kinetic model was the most suitable, and these adsorption systems proceeded by chemisorption.
During the adsorption of oppositely charged ions on the sorbents formerly saturated with NH 4 + or PO 4 3− (i.e., the NH 4 + adsorption on B, B Fe and TS Fe saturated with PO 4 3− , and the PO 4 3− adsorption on TS saturated with NH 4 + ) the efficiency increased compared to the adsorption on the original sorbents due to the creation of new active sites on the sorbent surface. The simultaneous sorption of PO 4 3− and NH 4 + was more efficient when compared with the efficiency of selective ion adsorption.
The leaching experiments proved to have a relatively high release of PO 4 3− and NH 4 + from saturated sorbents, which made it possible to apply the saturated sorbents to agricultural soils, for example, to increase their nutritional values. The content of NH 4 + in the leachates decreased in the following order: TS sorption > B > B Fe sorption ∼ = B Fe > B sorption > TS > TS Fe sorption ∼ = TS Fe ; the content of PO 4 3− in the leachates decreased in the following order: TS > TS sorption > B sorption > B > TS Fe sorption ∼ = TS Fe > B Fe sorption ∼ = B Fe .
The waste concrete sludge B was found to be an effective PO 4 3− sorbent. It is unsuitable for NH 4 + sorption due to its high alkalinity, which can be considered a major disadvantage for its possible use as a soil additive. Waste from the production of artificial stone TS was found to be a relatively good sorbent for the NH 4 + , but a high dosage is necessary to achieve an acceptable sorption efficiency. For the sorption of PO 4 3− , the TS is completely unsuitable because of the spontaneous release of this ion into the solution. Due to significantly lower alkalinity, the TS represents a promising candidate for application to agricultural soils. The modified B Fe and TS Fe forms proved to be selective and efficient sorbents of PO 4 3− ions, while the adsorption of NH 4 + on B Fe and TS Fe was almost ineffective. The use of B Fe and TS Fe as soil additives was possible.