Preparation and characterisation of alkali-activated blast furnace slag and Na-jarosite catalysts for catalytic wet peroxide oxidation of bisphenol A

In this study, cost-e ﬀ ective alkali-activated materials made from industrial side streams (blast furnace slag and Na-jarosite) were developed for catalytic applications. The catalytic activity of the prepared materials was examined in catalytic wet peroxide oxidation reactions of a bisphenol A in an aqueous solution. All materials prepared revealed porous structure and characterisation expressed the incorporation of iron to the material via ion exchange in the preparation step. Furthermore, the materials prepared exhibited high speci ﬁ c surface areas (over 200 m 2 /g) and were mainly mesoporous. Moderate bisphenol A removal percentages (35% – 37%) were achieved with the prepared materials during 3 h of oxidation at pH 7 – 8 and 50°C. Moreover, the activity of catalysts remained after four consecutive cycles (between the cycles the catalysts were regenerated) and the speci ﬁ c surface areas decreased only slightly and no changes in the phase structures were observed. Thus, the prepared blast furnace slag and Na-jarosite-based catalysts exhibited high mechanical stability and showed good potential in the removal of bisphenol A from wastewater through catalytic wet peroxide oxidation.


Introduction
Alkali-activated materials (AAMs) are zeolite-like, amorphous aluminosilicate compounds in which the calcium concentration plays an important role in the chemical structure [1].AAMs with low calcium concentration, known as geopolymers, have a three-dimensional aluminosilicate framework [2], while in the case of high-calcium AAMs, the calcium-sodium aluminosilicate hydrate (C-(N-)A-S-H) gel formed contains more cross-linked and non-cross-linked tobermorite structures [1,3].AAMs are produced through a reaction between a solid aluminosilicate precursor, and a concentrated alkaline activator, which is usually an aqueous solution of alkali hydroxide, silicate, carbonate or sulphate [4].The compounds most widely used as alkali activator solutions are MOH and M 2 O•xSiO 2 , where M is either a sodium or a potassium ion [5], while the most common raw materials for AAMs are industrial side streams, such as calcium-rich blast furnace slag (BFS) and other metallurgical slags, lowcalcium fly ash and red mud [1,6,7].In the past few decades, AAMs have attracted much interest worldwide in construction industry due to their inexpensive raw materials, low energy cost, great strength and good durability when compared to, e.g.Portland cement [8][9][10].In addition to their applications in the building industry, AAMs' uses in the field of water and wastewater treatment have also been investigated.AAMs, especially geopolymers, have been examined for, e.g. the removal of ammonium [11], heavy metals [12] and methylene blue [13] from aquatic environments.Furthermore, an alkali-activated BFS has been used as an adsorbent to remove nickel from aqueous solutions [14].Thus, AAMs show high potential in water treatment applications.
Bisphenol A (BPA) is a synthetic chemical that can be described as an environmental oestrogen, and it is used in the production of polycarbonate plastic and epoxy resin.These materials are then utilised in the production of, e.g.plastic bottles, food packaging, inner linings and coatings in metal food cans and beverages, and electronic equipment [15].BPA is classified as an endocrine-disrupting chemical and meaning that it can cause damage to the reproductive and neurological systems of animals and humans [16].Thus, several traditional separation methods, such as coagulation [17] and flocculation [18], as well as biological processes, e.g.activated sludge [19] and biofiltration [20], have been used to remove BPA from water, but with limited removal efficiency.More progressive separation methods, such as adsorption and ion exchange, have shown promising results in BPA removal [21], as have advanced oxidation processes (AOPs), such as catalytic wet air oxidation [22], ozonation [23] and ultravioletassisted oxidation reactions [24,25].One highly effective AOP is catalytic wet peroxide oxidation (CWPO), which shows great potential for removing organic pollutants from wastewater.In CWPO, organic compounds from wastewater are decomposed using hydrogen peroxide as a strong oxidising agent.Most catalysts used in CWPO processes are based on iron salts; however, other transition metals, such as Cu, Mn and Co, can also be used for this purpose [26].The active metals must be dispersed on a suitable support to maximize the number of active sites available to the reactants, therefore increasing the catalytic activity.Since the activity of heterogeneous catalysts is mainly promoted by atoms present at the accessible surface of the material [27], it is important to choose a support that has a high specific surface area (SSA).Numerous carbon-based materials, such as activated carbon [28], carbon nanotubes [29] and biomass-based carbons [30], have been used as supports in CWPO.Different clay materials [31,32] and zeolites [33] have also been utilised as catalyst supports, which in turn suggests zeolite-like AAMs as another interesting support alternative.
The novelty of this study is to use BFS and Na-jarosite (NaFe 3 (SO 4 ) 2 (OH) 6 ), a hydrous iron sulphate mineral often found in mine tailings and produced as a byproduct of sulphide oxidation and zinc refinement, as raw materials to produce cost-effective catalysts for the oxidation of BPA.The use of AAMs in the catalytic water treatment is rather limited and therefore such applications are examined in this research.

Materials and methods
In this section, the preparation process and different characterisation methods of studied AAMs are described.Furthermore, reaction conditions of the CWPO of an aqueous solution of BPA are given.

Preparation of alkali-activated materials and Fe catalysts
All AAMs were prepared by using industrial side streams from industry as raw materials.Two different powdered blast furnace slag samples (referred to as the first and second lot) were obtained from a European steel manufacturing company and Na-jarosite was obtained from a Finnish nickel processing industry.The compositions of the BFS and Na-jarosite, as determined by X-ray fluorescence (XRF) analysis are given in Table 1.
A total of 27 samples were prepared by mixing BFS and Na-jarosite (denoted by J) at different ratios (3:1 or 6.5:1 wt%), with 8.7 g of alkali activation solution (AASol) for 10 min at 500 revolutions per minute.The 27 samples were prepared in three batches: the first and second batch of samples were prepared with the first lot of BFS, while the third batch of samples had the second lot of BFS as a raw material (Table 1).The AASol was either 10 M of sodium hydroxide (NaOH The cured, mechanically stable samples were crushed and sieved to a particle size of <250 μm, washed and dried overnight at 105°C.The AAMs were washed with three different methods to study the effect of material pH on the removal of BPA.The materials in the first batch were washed with Milli-Q water (denoted by AQ) while the materials in the second and third batches (denoted by HAc and 5Hac) were washed with 1 M acetic acid once and five times, respectively, to investigate the effect of acid wash on the material's catalytic activity in the CWPO.Fe was chosen as the active metal for the catalysts, and it was impregnated on the AAMs prepared without Na-jarosite (denoted by Fe) through ion exchange.The ion exchange procedure was carried out by mixing the AAM and a 0.01M aqueous solution of ammonium iron (II) sulphate ((NH 4 ) 2 Fe(SO 4 ) 2 • 6 H 2 O, 99.0%-101.5%,Merck, Darmstadt, Germany) in an 80°C sand bath overnight.The amount of (NH 4 ) 2 Fe(SO 4 ) 2 used was 100 mL per 1 g of AAM.After filtration and washing with Milli-Q water, the samples were dried at 105°C overnight and then calcined at 500°C, raising the temperature from room temperature to the final value at a rate of 1°C/min.

Chemical stability of AAMs
The chemical stability of the prepared AAMs was examined by measuring the conductivity of the aqueous solution of the material.A sample of 1 g of the studied AAM was mixed with 25 mL of Milli-Q water; the conductivity was measured by using conductivity probe (Hach Lange HQ40D portable meter) immediately after mixing, and then after 6 and 24 h (results not shown).The most stable materials, i.e. the samples whose conductivity varied the least after 24 h, were chosen as catalysts or catalyst supports for the CWPO of BPA (Table 2).The samples were named according to their raw materials, curing temperature and whether the moulds were covered or not.In addition, the activator type was used as a part of the sample name, e.g.SH for sodium hydroxide, and SS for sodium silicate.

Characterisation of AAMs
The compositions of the BFS and Na-jarosite were analysed using an XRF spectrometer (PANanlytical Axios mAX XRF, Almelo, the Netherlands).The measurements were performed in an He atmosphere using loose powders through a transparent Mylar film.Leco CS (USA) elemental analyser was used to measure the sulphur contents.The phase compositions of the AAMs were determined using powder XRD with a PANalytical X'Pert Pro X-ray diffractometer (Malvern PANalytical, Amelo, The Netherlands).XRD analysis was performed by scanning 2θ values between 5 and 90°with monochromatic Cu Kα1 (λ = 1,5406 Å) at 45 kV and 40 mA at a scan speed of 0.015°/s.Crystalline phases were identified using HighScore Plus software with the Powder Diffraction File standards from the International Centre for Diffraction DATA (ICDD).DRIFTS was used to analyse the chemical and structural information of the materials' surfaces.DRIFT spectra were documented on a Bruker PMA 50 Vertex 80 V (Bruker, Billerica, MA, USA) equipped with a Harrick Praying Mantis diffuse reflection accessory and a high-temperature reaction chamber through baseline measurement using potassium bromide.The analysis was performed at room temperature in ambient atmosphere in the range of 400-4000 cm −1 , with 500 scans per minute and a resolution of 4 cm −1 .The SSA and porosity were obtained from nitrogen adsorption-desorption isotherms at the temperature of liquid nitrogen (−196°C) using the BET method on a Micromeritics ASAP 2020 system (Micromeritics Instrument Corporation, Norcross, GA, USA).Before analysis 0.1-0.2g of each sample was degassed 2 µm Hg pressure and 140°C for 2 h to clean the surface.The pore size distribution was calculated using density functional theory.In addition, the main elements of the prepared materials (Al, Ca, Mg, and Si) were analysed using ICP-OES analysis (Agilent 5110 SVDV, Agilent Technologies Inc., Santa Clara, California USA) and the active metal Fe using atomic absorption spectroscopy (AAS) analysis (Varian AA240FS, Agilent Technologies Inc., Santa Clara, California, USA).

CWPO experiments
Oxidation experiments were conducted with the AAMs listed in Table 2.They were performed in a four-necked flask reactor equipped with a reflux condenser, and the reactor was loaded with 160 mL of BPA aqueous solution (60 mg/L).The catalyst loading was 4 g/L, and the H 2 O 2 concentration was 0.15 wt%.The experiments were conducted at a reaction temperature of 50°C for the effective decomposition of H 2 O 2 to form active •OH radicals.The oxidation reaction was initiated with the addition of H 2 O 2 when the reaction temperature was reached.H 2 O 2 was added in a continuous flow with a flow rate of 0.33 mL/min during the 180 min test.Water samples were taken every 30 min and filtered using 0.45 μm cellulose nitrate filter paper.During the experiment, the pH and dissolved oxygen (DO) of the water samples were measured with an HQ40D IntelliCAL® LDO101-electrode (Hach).To assess the stability and reusability of the prepared materials, the AAMs AQ-Fe/BFS/SH + SS-60C, AQ-6.5 BFS + J/SH + SS-60C, 5HAc-Fe/BFS/SH + SS-60C and 5HAc-6.5 BFS + J/SH + SS-60C were examined in four consecutive runs.Furthermore, the regeneration of used catalysts was performed by washing the catalysts with 100 mL of Milli-Q water and drying them for 24 h at 105°C.

Analysis of oxidised water samples
The BPA concentration of the water samples was determined using high-pressure liquid chromatography with a Waters 996 photodiode array detector (Waters Corp., Milford, MA, USA) at a wavelength of 226 nm.A mixture of 0.1% trifluoracetic acid (TFA) in methanol and 0.1% TFA in water at a flow rate of 0.4 mL/min was used as the eluent mixture to separate compounds on a SunFireTM C18 5-m 2.1 × 100 mm column (Waters Corp., Milford, MA, USA) operated at 30°C.The total organic carbon (TOC) concentration of the water samples was determined on a Skalar FormacsHT Total Organic Carbon/total nitrogen analyser (Breda, The Netherlands).After oxidation, the final water samples were examined using ICP-OES analysis (Agilent 5110 SVDV, Agilent Technologies Inc., Santa Clara, California USA), to study possible leaching of Al, Ca, Mg and Si.

Results and discussion
In this section, the stability, SSAs, phase compositions and DRIFT spectra of the prepared materials are discussed.Furthermore, the performance of the AAMs in the CWPO of BPA is evaluated.

Chemical stability of AAMs
The conductivities of the aqueous solutions were measured at room temperature with an AAM concentration of 40 g/L (1 g/25 mL).According to the results (not shown), the AAMs prepared solely with BFS were more stable (i.e.their variability in conductivity after 24 h was lower) than those prepared with BFS and Najarosite.The conductivity varied less in samples prepared with sodium hydroxide and silicate than in those prepared solely with sodium hydroxide.The results indicate that the addition of sodium silicate stabilised the AAMs.Based on these stability tests, a total of 14 samples whose conductivity varied the least were characterised using XRD, ICP-OES and surface area techniques.In addition, these AAMs were chosen as catalysts for the CWPO of the BPA aqueous solution (Table 2).

Characterization of AAMs
Table 3 lists the results of the BET surface areas, pore diameters and pore volumes of the prepared AAMs.
All stable AAMs exhibited mainly mesoporous structures (80%-90%) in which, the pore diameters was 2-50 nm.Only a small percentage of micro-and macropores (defined as possessing pore diameters of <2 nm and >50 nm, respectively) were observed.HAcs and 5HAs exhibited greater surface areas as well as a higher percentage of mesopores than AQs but the acid wash considerably reduced the number of macropores in the AAMs' structures.Similar results were reported by [34] where an acetic acid treatment was performed on gasification slag and an increase in the number of mesopores was observed in the material's structure.Interestingly, the acid wash increased the number of micropores in SHs, but in SH + SSs, the percentage of micropores was reduced.In addition, the samples prepared with sodium hydroxide and silicate exhibited higher surface areas and pore volumes than those prepared solely with sodium hydroxide.When comparing materials that differed only by the composition of the AASol (i.e. by the denotation SH vs SH + SS), the SSAs doubled in size.Moreover, samples cured at 60°C for 24 h exhibited slightly higher surface areas and pore volumes than those cured at room temperature.This is because the higher curing temperature removed excess water from the materials' structures, which led to higher porosity.High curing temperatures (>50°C) have been reported to increase the number of mesopores in materials [35].In addition, the surface areas of samples cured at 60°C for 24 h and impregnated with Fe exhibited greater surface areas than those prepared with Na-jarosite.This was because of the calcination process performed on the Fe catalysts.Calcination of the catalysts resulted in the evaporation of water and carbon dioxide from the AAM structures, which gave rise to the increased surface area and pore volume [36].
Figure 1 shows the X-ray diffractograms of AAMs prepared in the first, second and third batch, as well as the raw materials (BFS and Na-jarosite) used in the preparation of the third batch.In the X-ray diffractogram of the BFS, no crystalline peaks were observed.Instead, the broad hump at the 2θ values of 23-37°revealed an amorphous structure.On the other hand, several crystalline peaks were observed in the X-ray diffractogram of Na-jarosite.The highest-intensity peaks at the 2θ values of 15.0, 28.After the alkali activation of the BFS and the mixture of the BFS and Na-jarosite, peaks were observed at the 2θ values of 11.3, 22.8, 34.7, 39.1 and 60,6°(denoted by ○), which corresponded to meixnerite (Mg 6 Al 2 (OH) 18 • 4 H 2 O, ICDD file 00-035-0965).However, these peaks were not observed in the X-ray diffractogram of 5HAc-Fe/BFS/SH + SS-60C.According to this result, the calcination at 500°C of the Fe catalyst led to the decomposition of the meixnerite [36].Furthermore, all materials prepared with Na-jarosite exhibited peaks that corresponded to Na-jarosite (denoted by ◇), while only HAc-6.5 BFS + J/SH-60C and HAc-6.5 BFS + J/SH + SS-60C showed peaks corresponding to iron (III) oxide monohydrate (denoted by ◆).The AAMs washed five times with acetic acid in the third batch (denoted by 5HAc) exhibited fully amorphous structures, indicated by wide halos at the 2θ values of 17-36°, while all the AAMs in the first and second batches exhibited only partially amorphous structures.A high-intensity crystalline peak at the 2θ value of 29.4°in addition to those at 35.9, 43.1, 47.4, 48.4, and 57.4°(denoted by ▪) corresponded to CaCO 3 (ICDD file 04-023-8700).In the X-ray diffractograms of 6.5 BFS + J/SH + SS-60C and BFS/SH + SS-60C a high-intensity peak slightly overlapping with the 29.4°peak and appearing as a right-handed shoulder at 30.0°(denoted by •) corresponded to wollastonite (CaSiO 3 (ICDD file 04-010-0710)).After the ion exchange of iron, low-intensity peaks were observed at 24 45 , ICDD file 04-017-1504); however, these peaks were not visible in the X-ray diffractograms of AQ-BFS/SH + SS-RTO, HAc-6.5 BFS + J/SH + SS-60C, HAc-BFS/SH + SS-RTO, HAc-BFS/SH + SS-RTC or HAc-BFS/SH + SS-60C.According to these results, the formation of katoite which was identified in the sodium hydroxide-activated materials, appears to have been hindered by the addition of sodium silicate to the AASol.Furthermore, the crystalline peaks at 17.5, and 36.9°(denoted by ^) and the peaks at 29.3, 59.7 and 71.5°(denoted by +) corresponded to Na 4 FeO 3 (ICDD file 00-034-0891) and SiO 2 (ICDD file 04-015-7166), respectively, and they were visible only in the X-ray diffractogram of HAc-6.5 BFS + J/SH-60C.In the X-ray diffractograms of the materials washed with water, peaks at 34.5°and 48.6°(denoted by C) revealed the presence of a calcium aluminosilicate hydrate (C-A-S-H) compound (ICDD file 00-018-0276), while the peaks at 32.1°and 50.1°(denoted by *) corresponded to calcium silicate hydrate (C-S-H) compound (ICDD file 00-033-0306), which is primarily a strengthening compound found in cement-based materials [37].These peaks were observed in the X-ray diffractograms of materials washed with water, except AQ-BFS/SH + SS-RTO, which indicated that this material did not achieve the same strength when cured in an open vessel at room temperature as those cured at 60°C in closed vessels.It was also reported by [38] that a higher curing temperature leads to greater strength in AAMs.Instead, the peaks observed at 43.2 and 48.5°(denoted Figure 2 shows the DRIFT spectra of BFS, Na-jarosite and the AAMs BFS/SH + SS-60C, 5HAc-BFS/SH + SS-60C, 5HAc-Fe/BFS/SH + SS-60C, 6.5 BFS + J/SH + SS-60C and 5HAc-6.5 BFS + J/SH + SS-60C from the third batch.Only a few peaks were observed in the DRIFT spectrum of BFS.The broad peak at 1490 cm −1 was assigned to the stretching vibration of the O-C-O bonds of CO 2− 3 groups while the strong peak at 1110 cm −1 corresponded to pure silica [39,40].In the DRIFT spectrum of Na-jarosite, a sharp peak at 1053 cm −1 was observed, which is typical for compounds containing SO 2− 4 [41].The alkali-activated materials washed with acetic acid exhibited multiple peaks in the regions analysed.A weak signal was observed at 3741 cm −1 , corresponding to silanol groups [42].After the acetic acid washes and the impregnation of iron, the intensity of the peak increased and sharpened, as can be seen in the DRIFT spectrum of 5HAc-Fe/BFS/SH + SS-60C.A shouldering peak at 1565 cm −1 on the right-hand side of the peak at 1639 cm −1 can be observed only in the materials washed with acetic acid.The peak corresponded to the O-C-O asymmetric stretching vibrations of carboxylate groups [43].Furthermore, in the DRIFT spectra of the materials washed with acetic acid, a slight shift to higher wavenumbers was observed for the peaks at 1170 and 1110 cm −1 .
All of the AAMs exhibited a broad band at 3600-3200 cm −1 following the absorption bands at 1639, 1170, and 686 cm −1 , which indicate hydroxyl groupsspecifically, strongly acidic bridged hydroxyl groups [44,45].In addition, the broad band at 3500-3400 cm −1 corresponded to silanol groups interacting with other atoms, e.g. in silanol nests [42,46].Furthermore, the band at around 3700-2800 cm −1 and the peak at 1639 cm −1 observed in all of the AAMs corresponded to the stretching vibrations of O-H groups and H-O-H bending vibrations of adsorbed water, respectively [46,47].The asymmetric stretch peaks at around 1490 and 1429 cm −1 split slightly into two parts in the DRIFT spectra of BFS/SH + SS-60C and 6.5 BFS + J/SH + SS-60C; this peak is characteristic of amorphous calcium carbonate [48].The broad peak at around 1490-1430 cm −1 disappeared after the introduction of iron to the sample.Finally, the peak at about 600 cm −1 corresponded to the Si-O-Si and Al-O-Si symmetric stretching vibrations [49].The iron species Fe 2 O 3 and Fe 3 O 4 should have shown absorbance peaks at 550 and 780 cm −1 and 571 and 590 cm −1 , respectively [50,51], but because of the overlap of the Si and Al vibrations in this wavenumber region, peaks corresponding to iron could not be observed in the DRIFT spectra of the prepared AAMs.
Table 4 lists the metal concentrations of Al, Ca, Mg, Si (determined by ICP-OES) and Fe (determined by AAS) in all AAMs.
Differences in the concentrations of, e.g.Ca, Mg, Si and Fe can be observed when comparing the AAMs from the first and second batch with the AAMs from the third batch.This is due to the different elemental composition of the two blast furnace slag batches used in this study (Table 1).According to the results obtained from the ICP-OES and AAS analyses, the Ca concentrations of the acid-washed materials were much lower than those of materials washed with water, while the Al and Si concentrations remained practically unchanged.A slight change in the Mg concentration was observed,  as the concentration was lower in the acid-washed materials than those washed with water.

CWPO experiments with AAMs
The stable AAMs, as listed in Table 2, were used as catalysts in the CWPO of a BPA aqueous solution and the results of these experiments are shown in Figure 3. Oxidation reactions were performed at 50°C at an initial pH of BPA (∼7.5), a catalyst concentration of 4 g/L and a H 2 O 2 concentration of 0.15 wt%.The oxidant was added in a continuous flow at a flow rate of 0.33 mL/min during the 180 min test.In the tests performed without a catalyst, only ∼10% BPA removal was reported [52], while with the AAMs, BPA removal of 32%-50% was achieved.No significant difference in the removal percentage was detected between the water-and acid-washed AAMs, and there was no clear correlation between catalytic activity and iron loading.Furthermore, measuring the TOC revealed that the removal of organics was 20%-22%.Several studies have demonstrated that the surface acidity is an important factor in the catalytic oxidation of organic compounds.For example, the strong surface acidity of hydroxyl-Fe-pillared bentonite catalyst enhanced the catalytic degradation of Orange II [53] while acid activated carbon catalyst improved the removal of phenol [54].In addition, in the study of Juhola et al. [30] the acidic character of iron impregnated carbon materials was more favourable in the degradation of bisphenol A than materials with basic surface.Although prepared materials expressed suitable SSA and porosity for catalytic oxidation (Table 3) probably the basic character of prepared materials was the main reason for moderate activity of catalysts.The lack of acidic surface sites limited the catalytic reaction on studied catalysts.Moreover, the DO concentration of the water samples changed from around 9.9 mg O 2 /L to 7.6 mg O 2 /L.This indicates that oxygen was still present after 180 min of oxidation, probably because the low reaction temperature (50°C) was not enough to the decompose H 2 O 2 to active •OH radicals.It has been reported that a higher reaction temperature promotes the degradation of H 2 O 2 and thereby that improves pollutant removal [55][56][57].

Stability of the used catalysts
Because no significant variation in the removal of BPA and TOC was noticed between any of the prepared AAMs, the samples with the highest surface areas were further examined in four consecutive cycles of CWPO of an aqueous solution of BPA.Furthermore, to study the effect of Na-jarosite on the performance of the catalyst, one sample prepared with only BFS and another sample with BFS and Na-jarosite were chosen, as well as the corresponding samples washed five times with 1 M acetic acid (i.e.samples AQ-Fe/BFS/SH + SS-60C, AQ-6.5 BFS + J/SH + SS-60C, 5HAc-Fe/BFS/SH + SS-60C, and 5HAc-6.5 BFS + J/SH + SS-60C).To have enough material for consecutive tests and regeneration, 12-21 tests were performed in total for each material, and the catalysts used in these experiments were collected and combined.Between consecutive tests, the catalyst used was filtered from the aqueous solution and dried at 105°C for the subsequent runs.
A BPA removal of 35%-37% was observed for all four catalysts after the first oxidation reaction; this percentage decreased by 2% on average after the second oxidation reaction using the same catalysts.Furthermore, the BPA removal after four consecutive runs decreased by around 1%-4% (Figure 4), except in the catalyst AQ-Fe/ BFS/SH + SS-60C, whose catalytic activity increased slightly.The BPA removal with this catalyst after four runs was ∼1.5% higher than after the first oxidation reaction.This indicates that the catalyst remained stable and active after multiple cycles in the CWPO of BPA.
Furthermore, the possible leaching of metals from the prepared AAMs was examined using ICP-OES between the consecutive tests.The water samples were filtered after 180 min of oxidation using a 0.45μm cellulose nitrate filter to remove the solid catalysts.The Al, Ca, Fe, Mg, and Si concentrations of the oxidised water samples were measured.According to the results, leaching of Ca and Si was observed in every catalyst.In all the oxidised water samples, the Si concentration after oxidation was around 12-32 mg/L, while the Ca concentration was 7-30 mg/L after one oxidation experiment and slightly higher in the following cycles, around 17-40 mg/L.The Ca concentration was slightly lower in the water samples catalysed by AAMs without impregnated iron.Similarly, the Mg concentration varied noticeably between water samples catalysed by AAMs treated with high temperature, and samples catalysed without heat-treated AAMs, from 1.3-13-0.1-5mg/L, respectively.Therefore, it is evident that the heat treatment of the catalysts led to the increased dissolution of Ca and Mg in the water phase during the oxidation experiment.The Al concentration of the oxidised water samples was quite negligible (at a maximum of 0.2 mg/L), as was the Fe concentration (at a maximum of 0.16 mg/L).
In iron catalysed wet peroxide oxidation reaction hydrogen peroxide oxidates Fe(II) to Fe(III) and at the same time hydroxyl radicals are generated (Equation 1).Fe(III) is further reduced to Fe(II) (Equation 2) with the generation of more hydroxyl radicals (Equations 3-5) [58,59].
Despite the leaching of Ca, Si and Mg, the DO concentration of water samples was min.6.5 mg O 2 /L after consecutive experiments which indicates that hydroxyl radical generation was not disturbed on the dissolved elements.In addition, the BPA removal decreased only slightly after four consecutive experiments.

Conclusions
Cost-effective AAMs made from BFS and Na-jarosite were prepared and their catalytic activity in the CWPO of an aqueous solution of BPA was evaluated.AAMs cured at 60°C were selected as catalysts for CWPO reactions, as they proved to be more stable in aqueous solution than those cured at room temperature.The effect of pH on the removal of BPA was studied by washing the prepared materials with either water or acetic acid, but no significant difference was observed between the different materials in the percentage of BPA removal.The catalytic activities of the materials were examined in a 180 min oxidation reaction at a reaction temperature of 50°C and at pH 7-8.The catalysts AQ-Fe/BFS/SH + SS-60C, AQ-6.5 BFS + J/SH + SS-60C, 5HAc-Fe/BFS/SH + SS-60C and 5HAc-6.5 BFS + J/SH + SS-60C exhibited moderate BPA removal (35%-37%), even after four consecutive experiments.Although prepared AAMs thus prove to be interesting alternatives for catalytic water treatment applications, leaching of Ca and Si, as well as of Mg, was observed from the AAMs during oxidation.Therefore, in future studies, additional attention should be paid to the stability of AAMs.

Table 1 .
Composition of raw materialsblast furnace slag and jarositeas determined by XRF analysis.Results are reported with wt% > 0.01.

Table 2 .
Specific preparation methods (curing temperature and method), raw materials and number of acetic acid washes (HAc) performed to produce mechanically and chemically stable materials.

Table 4 .
Metal concentrations of prepared materials from all batches, as determined by ICP-OES (Al, Ca, Mg and Si) and AAS (Fe) analysis.