Experimental Study on the Purification of High-Hardness Arsenic-Containing Water Using Sintered-Clay Waste Materials Authors: Qinghao Xin Published: 2026 ABSTRACT Sintered-clay waste (arising from the crushing of red and blue bricks), as the main constituent of urban construction waste, has a porous structure and water purification components: the fixed adsorbents (SiO2, MgCa­(CO3)2) and the effective purification substances (sodium aluminum silicate, potassium aluminum silicate), which provide a new adsorbent for the purification of high-hardness arsenic-containing water. Analysis of the components of sintered clay and filtered sintered clay showed that adsorption, complexation, and chemical reactions were the main purification mechanisms of high-hardness arsenic-containing water. The pore-specific surface area of the large-particle sintered-clay filter media was smaller than that of their small-particle counterparts. The experimental results showed that the process of total hardness softening included three stages: the accelerated stage, the decelerated stage, and the stable stage; the arsenic removal process consisted of three stages: the decelerated stage, the accelerated stage, and the stable stage. Both the effective purification duration of the total hardness and arsenic was 40–60 > 20–40 ≈ 10–20 mesh; the removal rates were 40–60 > 20–40 > 10–20 mesh at each purification stage, and the optimal particle size of the sintered-clay filter media was 20–40 mesh. After filtration, the amounts of components SiO2 and MgCa­(CO3)2 increased in the sintered-clay media; SiAs2 and Pb28As12S46 also appeared. The precipitates and the strong-adsorbent flocculation substances Al­(OH)3 were produced during the reaction between Ca2+, Mg2+, and silicates and aluminates in the sintered clay. The ions Ca2+, Mg2+, and arsenic were further removed during Al­(OH)3 redissolution and surface group M–OH complexation with Ca2+ and Mg2+. Both the confidence ranges of the total hardness and arsenic were narrow, and the confidence range of total hardness softening showed a greater change in the early stage than in the later stage; the trend for arsenic ran contrary. The average correlation coefficients of the total hardness and arsenic adsorption models were, respectively, significant at 0.9410 and 0.9625, and both consistencies of experimental adsorption capacities with their theoretical counterparts were greater in the middle stages than in the end stages. The hardness removal rate was 69.36%, and that of arsenic was 53.45%, both higher than those achieved by existing adsorption filters for purifying high-hardness arsenic-containing water. Introduction The utilization of groundwater and industrial wastewater has come into human view with the gradual exploitation of high-quality surface water resources. However, it is difficult to directly utilize groundwater and industrial wastewater because of high hardness and the presence of small amounts of heavy metal ions. Therefore, the softening of high-hardness water and the removal of heavy metal ions are of great significance for improving water utilization efficiency and alleviating pressure on water resources . However, the large-scale use of existing adsorption materials is difficult due to limitations in treatment effectiveness and regeneration costs. It is of practical significance to explore adsorption materials that are low-cost, highly efficient , and capable of strong-sustainable purification to improve the efficiency of water softening and the removal of heavy metal ions. At present, the methods for removing Ca2+, Mg2+, and heavy metal ions include chemical precipitation, ion exchange, membrane separation, electrochemistry, and adsorption. Adsorption is a common method for water softening, and the main adsorption materials include lime, zeolite, alumina, bauxite, and fly ash. − In recent years, studies on the removal of Ca2+ and Mg2+ by exchange resins, polymer materials, weathered coal, bamboo charcoal, and nut shells have gradually emerged, but these methods were not widely utilized due to secondary pollution. The mixed zeolite A and X cocrystallized functionalized adsorption material promoted the synergistic effect of cation exchange and alkaline precipitation. The new sepiolite-polyacrylonitrile polymer composite resin could also effectively soften water. However, the above research could not meet the requirements of large-scale water treatment and was not easy to widely promote or apply . The permeable reaction membranes were relatively expensive, and the purification capacity decreased significantly with prolonged usage. − The application of electrochemical softening technology was severely restricted by the lack of environmentally friendly liquid–solid separation methods and numerous issues such as equipment, performance, and economy. − The methods for arsenic removal from water included commonly membrane technology, electrocoagulation, ion exchange, and adsorption. Membrane separation technology could not be used on a large scale due to high cost. The removal rate of electrocoagulation was significantly affected by factors such as metal electrodes, pH, current density, treatment time, and ion concentration. − Iron-containing substances had a good enrichment effect on arsenic; however, pure iron-containing substances were prone to agglomeration and hardening. − Adsorption was widely used due to advantages such as simplicity, high efficiency, and recyclability. At present, arsenic removal adsorption materials include carbon-based, inorganic, and waste-based ones. Arsenic adsorption by carbon-based and inorganic materials was greatly affected by pH, functional active groups, temperature, etc., and the leachate was unstable and even toxic. − The arsenic treatment using waste-based adsorption materials has been emerging, including electrolytic manganese slag materials, pyrolysis waste from discarded tires, and stainless steel slag. The adsorption materials of electrolytic manganese slag and pyrolysis waste had a relatively low arsenic removal rate. , Stainless steel slag permeable reactive barriers (PRB) had certain effects in remediating arsenic-contaminated groundwater, while the large-scale removal of arsenic-containing water was limited, and the sources were few. Obviously, research on the treatment of arsenic-containing water is a long and arduous task. The accumulation of demolished and abandoned construction waste occupies a large amount of land with the development of cities. Conventional covering and spraying have solved the problem of air dust, and filling and mine underground filling could reduce the land occupation of construction waste. However, existing methods do not consider the high-performance utilization of urban construction waste. The research on construction waste components is of great significance for enhancing its utilization value and conserving resources. Construction waste contains multiple components such as waste concrete, red and blue bricks, soil, etc. Red and blue bricks (sintered-clay waste) account for a large proportion in construction waste, which have well-developed micropore structures and good adsorption properties, and contain a small amount of arsenic-removing components such as aluminum and iron. Based on the component determination and analysis, the adsorption, complexation, and chemical reactions of sintered-clay waste were explored in high-hardness arsenic-containing water purification. The characteristics of the component migration and the mechanisms of the ion adsorption, complexation, and chemical reaction were clarified, and the optimal particle size was determined. Application tests verified the feasibility of scale treatment on hardness softening and arsenic removal of high-hardness arsenic-containing water by sintered-clay waste. This paper used sintered-clay waste for the first time to simultaneously achieve hardness softening and arsenic removal of high-hardness arsenic-containing wastewater, which saved treatment costs and alleviated environmental pressure. The research results achieved a win–win situation of efficient utilization of construction waste and low-cost, long-period-scale treatment of high-hardness arsenic-containing wastewater. Methods Clay Components and Adsorption Modification The sintered-clay waste was taken from an urban construction dump (Figure a), which was made from clay at high temperature. The main clay components (containing kaolinite) are hydrated aluminum silicate, calcium oxide, magnesium oxide, etc. The molecular formula of hydrated aluminum silicate is Al2O3·2SiO2·2H2O with a 1:1 type dihedral layered structure, and the crystalline water was present in the form of hydroxyl groups (Table ). The structural functional groups on the surface of kaolin included −Si­(Al)–OH, −Si–O–Al–, and −Si­(Al)–O (Figure b). Subjecting the clay to high-temperature sintering, calcination, dissolution, and crystallization, multidimensional pore materials similar to molecular sieves were formed, which possessed ion-exchange adsorption properties. Silicoaluminate (Na/KAlSiO4) produced in sintered clay could remove Ca2+ and Mg2+ through chemical reactions, and generate the flocculating constituent Al­(OH)3, which could adsorb Ca2+ and Mg2+ again in the treatment of high-hardness arsenic-containing water. During this period, some flocculating constituent Al­(OH)3 disintegrated to produce the ion Al3+, which reacted with H3AsO4 to form the precipitate of aluminum arsenate (AlAsO4). 1Characteristics and Structure of Kaolinite in ClayMolecular formulaStructure typeThe main structure of octahedronMain interlayer ionsCation exchange capacity (mmol+.g–1)Al2Si2O5(OH)4 One tetrahedron and one octahedronDioctahedronNo0.03–0.15 Characteristics and Structure of Kaolinite in Clay Clay structures and XRD patterns of sintered-clay waste. a) Covered urban construction dump; b) structures of main clay components: Al2O3·2SiO2.2H2O and Ca/MgO; c, d, e) 10–20, 20–40, 40–60 mesh XRD patterns of sintered-clay waste, respectively. Diffraction Analysis of Sintered-Clay Waste The sintered-clay waste was crushed and sieved into three particle sizes: 10–20, 20–40, and 40–60 mesh, and ground to less than 200 mesh for component analysis one by one. The components were determined by a D8 ADVANCE X-ray diffractometer with continuous scanning. The XRD patterns of the sintered-clay components are shown in Figure c–e, and the quantitative results are presented in Table . 2Component Quantitation of Sintered-Clay Waste by XRD (%)StateSintered-clay waste before filtrationParticle size (mesh)Group 1 (10–20)Group 2 (20–40)Group 3 (40–60)SiO2 42.351.645.4CaMg(CO3)2 8.77.822.1NaAlSi3O8 29.222.911.9KAlSi3O8 19.8//CaAl2Si2O8 /17.720.7 Component Quantitation of Sintered-Clay Waste by XRD (%) As illustrated in Figure c–e and Table , the components of the three sintered-clay wastes were mainly SiO2, CaMg­(CO3)2, calcium aluminosilicate, sodium aluminosilicate, and potassium aluminosilicate salts. SiO2 accounted for the largest proportion, ranging from 42.3% to 51.6%, CaMg­(CO3)2 from 7.8% to 22.1%, calcium aluminosilicate from 17.7% to 20.7%, sodium aluminosilicate from 11.9% to 29.2%, and potassium aluminosilicate at 19.8%. CaMg­(CO3)2 was present as a precipitate, which could not increase the ions Ca2+ and Mg2+ in the filtered water. Sodium aluminosilicate and potassium aluminosilicate, the main components for removing Ca2+ and Mg2+, could produce the flocculating constituent Al­(OH)3 enhancing arsenic removal during the removal of Ca2+ and Mg2+. Relationship between Pores and Particle Size of Sintered-Clay Waste Sintered-clay waste could be used as an excellent adsorption material due to its rich microporous structure. The pores of granular sintered-clay waste included internal micropores and particle voids (Figure a). The absorbed water was stored in the internal micropores of sintered-clay waste with poor water solubility. The particle gradation influenced the specific surface area of the filter media. The differences in the specific surface area of three particle-size sintered-clay filter media, namely 10–20, 20–40, and 40–60 mesh (Figure b–d), were mainly caused by particle voids, which were the key factors influencing the porosity, pore-specific surface area, and adsorption of the filter media. According to the Lauden method, the void-specific surface area of the granular sintered-clay waste was, 1 (FV)i′=6di×di+1 Pore characteristics and binarization of sintered-clay waste media. (a) Pore structures of sintered-clay waste; (b, c, d) porosity binarization process of 10–20, 20–40, and 40–60 mesh sintered-clay waste, respectively. Here, (F/V)′i represents the specific surface area, cm2/g; d i is the sieve diameter of grade i, mm. PCAS software was applied to binarize the images of three particle-size sintered-clay waste (10–20, 20–40, and 40–60 mesh), and the porosity was, respectively, found to be 40.78%, 39.15%, and 37.29% before filtration (Table ). The void-specific surface area was respectively calculated by eq . The smaller the particle size, the larger the adsorption and pore-specific surface area. 3Pore Specific Surface Area in the Granular Sintered-Clay Waste (%)Particle size (mesh)10–2020–4040–60Specific surface area of pores (cm2/g)4.2439.49421.202 Pore Specific Surface Area in the Granular Sintered-Clay Waste (%) Analysis of FE-SEM and EDS The microscopic pore structure characteristics of the sintered-clay waste were observed and imaged using an FEI Quanta FEG 250 field emission scanning electron microscope (Figure ). The abundant micropores in the sintered-clay waste were the main sites for Ca2+/Mg2+ adsorption, presented as linear or curved shapes with good connectivity, which was conducive to permeation. The mineral debris reduced the connectivity to some extent while being conducive to adsorption. The sintered-clay waste mainly consisted of hydrated aluminosilicate, calcium oxide, and magnesium oxide. Scanning images of the sintered-clay waste by FE-SEM. (a–f) Images magnified by 50, 100, 200, 300, 400, and 500 times, respectively. Based on XRD diffraction analysis, the field emission scanning electron microscope was used to obtain the ultraresolution EDS elemental cross-sections of Si, Al, Ca, K, Mg, and Na to further analyze the composition of the sintered-clay waste (Figure ). The distributions of Si, Al, Ca, K, Mg, and Na were relatively uniform, little agglomeration occurred, and the proportions were 68.7%, 13.8%, 11.3%, 3.2%, 1.7%, and 1.4%, respectively. Si and Al accounted for the largest proportion; Si was the main element in the pore structure of the sintered-clay waste, and Al played the greatest role as an effective purification substance. The analysis results were consistent with the XRD information. High-performance X-ray energy spectrum analysis (EDS) of sintered-clay waste. (a) ultraresolution EDS cross-section of elemental distribution, (b) ultraresolution EDS elemental cross-sections of Si, Al, Ca, K, Mg, and Na, respectively, (c) X-ray energy spectrum. Experimental Scheme Based on the flow velocity and the crushing cost, the sintered-clay waste was mechanically crushed and sieved into granules with particle sizes 10–20 mesh (2–1 mm), 20–40 mesh (1–0.4 mm), and 40–60 mesh (0.4–0.2 mm). Each filter medium was then respectively filled into a transparent column of φ45 mm × 300 mm in three layers. The water sample was taken from the Quaternary aquifer at a buried depth of 60 m in the south of Jiaozuo, with BOD5 ≤ 2 mg/L, CODMn ≤ 5 mg/L, 7.2 ≤ pH ≤ 8.2, and 600 ≤ EC ≤ 1800 μS/cm. The contents of arsenic, Ca2+, Mg2+, and total hardness were respectively 0.15, 98.17, 113.91, and 709.89 mg/L. The water sample was placed into a Bartensioner flask and fed into the column through a hose. 1000 mL water sample was taken once every 8 h at a flow rate of 0.0004m3/min, which ensured both the bed contact time and the replacement of the water sample in the column. The process was monitored using an inductively coupled plasma optical emission spectrometer until the contents of the ions Ca2+, Mg2+, and arsenic stabilized. Sampling was conducted over a period of 520 h. The experimental design is shown in Figure . Experimental design. (a) Sintered-clay waste, (b) crushing of sintered-clay waste, (c) screening of sintered-clay waste, (d) filling of granular sintered-clay waste for the infiltration test, (e) gradation of granular sintered-clay filters, and (f) infiltration test of granular sintered-clay filters. Results Analysis of Total Hardness Softening The curve of the Ca2+ content in the filtered water sample is shown in Figure a. In the initial filtration flushing stage, the Ca2+ content in the filtered water sample was higher than that in the raw water because the calcium components in the fine-grained filter media were dissolved. The washing times for 10–20, 20–40, and 40–60 mesh filter media were, respectively, 22, 22, and 30 h. Subsequently, the Ca2+ content rapidly dropped below that of the raw water. The accelerated purification stages for 10–20, 20–40, and 40–60 mesh filter media lasted for 144, 136, and 112 h, respectively. Over time, the adsorption effect of the sintered-clay filter media weakened; meanwhile, the strong-adsorption flocs Al­(OH)3 could be formed from the Ca2+ and Mg2+ with aluminate. Therefore, the decelerated purification time was prolonged, and the purification effect stabilized (representing the decelerated purification stage). The stable purification stage followed: the Ca2+ removal of sintered-clay filter media presented three stages after the initial filtration washing: the accelerated purification stage, the decelerated purification stage, and the stable purification stage. The curve of the Ca2+ removal rate is shown in Figure b. The smaller the particle size, the more Ca2+ dissolution occurred in the sintered-clay filter media during the filtration washing stage. In the accelerated purification stage, the smaller the particle size, the greater the Ca2+ adsorption and exchange, and the higher the removal rate. The maximum removal rates of the 10–20, 20–40, and 40–60 mesh filter media reached, respectively, 56.88%, 61.99%, and 67.27% at 144, 136, and 112 h, when the adsorption and ion exchange were sufficient. In the decelerated purification stage, the adsorption and exchange effects of small-particle filter media started to weaken, and the purification capacity of small-particle filter media rapidly decreased, as did the removal rate. The decelerated removal of Ca2+ continued, respectively, to 376, 376, and 400 h with removal rates of 54.07–56.88%, 57.80–61.99%, and 60.14–67.27% for 10–20, 20–40, and 40–60 mesh filter media. The stable purification stage then ensued. The Ca2+ removal rates of the three particle-size filter media were 40–60 > 20–40 > 10–20 mesh. Test results and analysis of hardness softening. (a) Curve of Ca2+ removal in the filtered water, (b) curve of Ca2+ removal rate, (c) curve of Mg2+ removal in the filtered water, (d) curve of Mg2+ removal rate, (e) curve of the total hardness softening in the filtered water, and (f) curve of the total hardness removal rate. The curve of the Mg2+ content in filtered water is shown in Figure c, and the removal rate is illustrated in Figure d. After the sintered-clay filter media were washed, the Mg2+ content in the filtered water was significantly lower than that in the raw water, and the Mg2+ adsorption effect was significant. The removal rate of Mg2+ was 40–60 > 20–40 > 10–20 mesh because the pore specific surface area of the filter media was 40–60 > 20–40 > 10–20 mesh. The accelerated purification stages of 10–20, 20–40, and 40–60 mesh filter media ranged from 8 to 32 h, during which the Mg2+ removal rates reached, respectively, 68.44%, 73.62%, and 82.25%, and Mg2+ contents were purified, respectively, to 35.95, 30.04, and 20.22 mg/L. This was followed by the decelerated purification stage (32–152 h), wherein Mg2+ adsorption was apparent in the early part thereof. The removal rate of 40–60 mesh filter media gradually approached that of the 20–40 mesh counterpart in the later stage due to the sufficient release of the ion Mg2+ in CaMg­(CO3)2. The Mg2+ removal rates of 10–20, 20–40, and 40–60 mesh filter media decreased to 52.17%, 55.69%, and 56.05%, with the maximum contents thereof being 54.48, 50.47, and 50.06 mg/L, respectively. Due to the complexation of the flocculating constituent Al­(OH)3 and the Al–OH functional group, the Mg2+ decelerated removal of 10–20, 20–40, and 40–60 mesh filter media was more stable and prolonged. The smaller the particle size of the filter media, the higher the Mg2+ removal rate was. The Mg2+ removal rate of the 40–60 and 20–40 mesh filter media was significantly better than that of their 10–20 mesh counterparts. In the later decelerated purification stage and the stable purification stage, the ion Mg2+ released from the CaMg­(CO3)2 significantly influenced the removal rate; the smaller the particle size, the greater the influence was. The Mg2+ removal rate of the 20–40 mesh filter media gradually closed in on that of the 40–60 mesh one. The curve of total hardness softening is shown in Figure e. The washing stage of the 10–20, 20–40, and 40–60 mesh filter media was respectively 16, 16, and 19 h, during which the total hardness was higher than that of the raw water due to the dissolution of cations in the sintered-clay media. The total hardness entered the accelerated purification stage after the washing stage. The accelerated purification stages of the 10–20, 20–40, and 40–60 mesh filter media were, respectively, 16–48, 16–56, and 19–56 h, with the minimum content 279.78, 255.83, and 217.48 mg/L. In the stable deceleration purification stage, the total hardness was, respectively, 327.03–301.69, 398.58–297.74, and 285.60–291.24 mg/L. The curve of the total hardness removal rate is shown in Figure f. In the accelerated purification stage, the total hardness removal rates of the 10–20, 20–40, and 40–60 mesh filter media increased rapidly in 16–48, 16–56, and 19–56 h, with the removal rates being 60.59%, 63.96%, and 69.36%, respectively, and gradually decreased in 48–152, 56–152, and 56–152 h, with the maximum removal rates 53.93%, 57.93%, and 59.82% in the deceleration purification stage. The stable purification then ensued and continued respectively to 376, 376, and 400 h. Obviously, the total hardness of the filtered water from the 10–20, 20–40, and 40–60 mesh filter media all decreased rapidly to less than that of the raw water, and entered the effective purification periods of 32–520 h, with the removal rates being 57.94–60.59%, 57.36–63.96%, and 57.74–69.36%, respectively, including the partial accelerated purification stage, the decelerated purification stage, and the stable purification stage. The total hardness removal rates of the three particle-size filter media were in the order: 40–60 mesh >20–40 mesh >10–20 mesh. Analysis of Arsenic Removal The curve of arsenic removal is shown in Figure . In the deceleration purification stage, arsenic adsorption by sintered-clay waste gradually increased. The arsenic content in the filtered water from the 10–20, 20–40, and 40–60 mesh sintered-clay media decreased, respectively, to 0.0930, 0.0943, and 0.0903 mg/L at 16 h, and gradually increased, respectively, to 0.1066, 0.1050, and 0.1034 mg/L within 48, 48, and 40 h. Subsequently, the long accelerated purification stage occurred until the arsenic content decreased to a stable level, at which point the complexation and adsorption capacity decreased significantly, and the stable purification stage ensued. The boundary points were, respectively, 184, 208, and 208 h between the accelerated purification stage and the stable stage for the 10–20, 20–40, and 40–60 mesh sintered-clay media. When the sintered-clay filter media were washed for 16 h, the arsenic removal rate increased due to the strong adsorption on the pore surface, but the adsorption capacity of the filter media was subsequently weakened. The arsenic removal capacity of the 10–20, 20–40, and 40–60 mesh filter media decreased, respectively, to 27.19, 28.29, and 29.34% at 48, 48, and 40 h, and the arsenic adsorption capacity gradually increased with decreasing particle size of the filter media. Subsequently, complexation, flocculation adsorption, and chemical reactions gradually intensified, and the accelerated purification stages for the 10–20, 20–40, and 40–60 mesh filter media lasted 48–184, 48–208, and 40–208 h, respectively, followed by a stable purification stage. The removal rates of arsenic during the entire purification stage for the 10–20, 20–40, and 40–60 mesh filter media were, respectively, 27.19–48.01, 28.29–50.63, and 29.34–53.45%. Test results and analysis of arsenic removal. a) Curve of arsenic removal in the filtered water; b) curve of arsenic removal rate. From the above, the purification stability and the removal rates of total hardness and arsenic from 20–40 and 40–60 mesh filter media were significantly better than those of 10–20 mesh ones. Considering the influence of particle size on flow rate, the reasonable particle size of filter media was selected to match the water consumption in applications. The confidence intervals of the average total hardness softening and the average arsenic removal for 10–20, 20–40, and 40–60 mesh sintered-clay filter media are shown in Figure . The confidence range was between the upper and lower limits of 95%. The total hardness softening started from a lower value than the raw hardness and continued to the end of each experiment, and arsenic removal was observed throughout the treatment period. The confidence interval of the total hardness was wide during the accelerated purification stage and decreased gradually during the decelerated purification stage and the stable stage, which showed the purification tended to stabilize over time. The confidence interval of arsenic removal was narrow, with small differences among the three particle-size media during the deceleration and acceleration stages, but it gradually expanded as differences among the three particle-size media increased in the later stable stage because of the rise in arsenic content. The change in the confidence range of total hardness softening was overall larger in the early stage than in the later stage, whereas the opposite trend was observed for arsenic removal. Confidence interval analysis. (a, b) Confidence intervals of the average total hardness softening and the average arsenic removal of 10–20, 20–40, and 40–60 mesh sintered-clay filter media, respectively. Mechanism Components of the Filtered Sintered-Clay Media To compare the component differences of the sintered-clay media before and after filtration, the 10–20, 20–40, and 40–60 mesh filtered sintered-clay media were respectively taken for monitoring. The monitoring method was the same as that used before filtration. The component quantification is presented in Table , and the XRD patterns of the filtered sintered-clay media are shown in Figure a–c. 4Component Quantification of the Filtered Sintered-Clay Waste by XRD (%)Particle size (mesh)10–2020–40–40 ∼ 60SiO2 66.156.755.0CaMg(CO3)2 13.515.321.5NaAlSi3O8 ///KAlSi3O8 ///CaAl2Si2O8 //19.2CaMgSi2O6 12.7//(Mg1.56Fe0.39Ca0.05)Si2O6 /21.1/Pb28As12S46 7.7/4.2SiAs2 /6.9/ Component Quantification of the Filtered Sintered-Clay Waste by XRD (%) XRD diffraction analysis and purification model of the filtered sintered-clay media. (a–c) XRD patterns of the 10–20, 20–40, and 40–60 mesh filtered sintered-clay waste, respectively; (d) Main component structure of the filtered sintered-clay filter media: CaMg­(CO3)2, SiO2, (e) structural model of aluminosilicate; f) Ca2+/Mg2+ complex structure; (g) arsenic complex structure, and h) three stages of arsenic removal. In the 10–20 mesh filtered sintered-clay media, SiO2 increased from 42.3% to 66.1%, MgCa­(CO3)2 increased from 8.7% to 13.5%, NaAlSi3O8 and KAlSi3O8 decreased respectively from 29.2% and 19.8% to undetectable levels, and 12.7% CaMgSi2O6 and 7.7% Pb28As12S46 were present in the filter media. In the 20–40 mesh filtered sintered-clay media, SiO2 increased from 51.6% to 56.7%, MgCa­(CO3)2 increased from 7.8% to 15.3%, CaAl2Si2O8 decreased from 17.7% to undetectable levels, and 21.1% (Mg1.56Fe0.39Ca0.05)­Si2O6 and 6.9% SiAs2 appeared. In the 40–60 mesh filtered sintered-clay media, SiO2 increased from 45.4% to 55.0%, MgCa­(CO3)2 decreased slightly from 22.1% to 21.5% with no significant change, and CaAl2Si2O8 (19.2%) and Pb28As12S46 (4.2%) appeared. The stable precipitates SiO2 and MgCa­(CO3)2 in all the filtered media showed a significant increasing trend, and their stable structures are shown in Figure d. Sodium aluminum silicate and potassium aluminum silicate salts were the main components responsible for Ca2+ and Mg2+ removal. The decreasing CaAl2Si2O8 trend indicated that some Al3+ dissolved and interacted with Ca2+ and Mg2+. The arsenic removal effect of the filter media was good, and the main products were Pb28As12S46 and SiAs2 due to the reaction between arsenic in the water sample and Pb and S in the sintered-clay media. The changes in media components before and after filtration indicated the complexity of the sintered-clay filter media in purifying high-hardness arsenic-containing water. Analysis of Filtration Purification According to the XRD results of sintered-clay waste, quartz (SiO2) was 42.3–51.6%, silicoaluminates (NaAlSi3O8, KAlSi3O8, CaAl2Si2O8) 32.6–49.0%, CaMg­(CO3)2 7.8–2.1%. Quartz (SiO2) was stable in sintered-clay waste. Silicoaluminate formed a framework structure with three-dimensional silicon–aluminum–oxygen tetrahedral units, which contained balancing electrical cations such as Na+ and K+ (Figure e). The surface of the silicoaluminate structure underwent hydration and dissolution during sintered-clay waste purifying high-hardness arsenic-containing water, as shown in eq . 2 2NaAlSi3O8⁡+⁡2CO2+11H2O→2Na++Al2Si2O5(OH)4+3H4SiO4+2HCO3− Meanwhile, the sintered-clay waste possessed significant ion exchange/adsorption capability because of the pore structure, which was the main way to remove Ca2+ and Mg2+, as shown in eq . 3 2KAlSi3O8⁡+⁡Ca2+→2K++CaAl2Si2O8+4SiO2 The sintered-clay waste contained a large amount of silicates and aluminates, which could react with Ca2+ and Mg2+ in water. The main reaction equations were, 4 Ca2+/Mg2++SiO32‐→Ca/MgSiO3↓ 5 Ca2+/Mg2++⁡2AlO2−⁡+⁡4H2O→Ca/Mg(OH)2⁡+⁡2Al(OH)3↓ 6 Ca2+/Mg2++⁡AlO2−⁡+⁡2H2O→Ca/Mg(OH)2⁡+⁡Al(OH)3↓ The strong adsorption flocculating constituent Al­(OH)3 is produced in the reaction of Ca2+ and Mg2+ with aluminate, which can physically adsorb Ca2+ and Mg2+. When the concentrations of Ca2+ and Mg2+ ions are sufficient , these ions react chemically with Al­(OH)3 to form the coprecipitates Mg2Al­(OH)7 and Ca2Al­(OH)7. The functional group OH– in Al­(OH)3 surface complexes with Ca2+ and Mg2+. The reaction equations are as shown in eqs and , and the complex structures are shown in Figure f. 7 Al‐OH+Ca2+/Mg2+→Al‐O‐Ca/Al‐O‐Mg 8 2Al‐OH+Ca2+/Mg2+→Al‐O‐Ca‐O‐Al/Al‐O‐Mg‐O‐Al CaMg­(CO3)2 in the sintered-clay waste could release Ca2+ and Mg2+ in sufficient CO2 water. The reaction formula was, 9 CaMg(CO3)2⁡+⁡CO2⁡+⁡H2O→Ca2++Mg2++⁡(OH)2+2HCO3− The forms of arsenic in water included H2AsO4 – (pH between 4 and 9), HAsO4 2– (pH between 4 and 9), H3AsO3 (pH between 4 and 10), and H2AsO3 – (pH between 4 and 10). Arsenic could be reduced by adsorption, complexation, and the chemical action of the sintered-clay waste. Arsenic in water usually exists in the form of AsO3 –, AsO4 –, and their salts. The ionization equilibriums of arsenic in water were, 10 3H++AsO33−=As(OH)3=As3++3OH− 11 5H++AsO43−=As5++⁡H2O+3OH− In acidic water, the aluminum oxides or hydroxides containing Al3+ partially dissolved, and the removal effect of As­(III) was poor in the case of aluminum-containing sintered materials. In neutral or alkaline water, As­(III) mainly existed in the form of neutral molecules (H3AsO3), some complex anions would appear, and physical adsorption (Coulomb force) would be generated between the surfaces of the sintered-clay media and the adsorbate, which improved the removal capacity of As­(III). pH had a small impact on As­(V) removal; the oxidized or dissolved Al3+ could react with H3AsO4 to form the precipitate AlAsO4 (reaction eq ). The main force for removing As­(V) was the chemical bonds between the surface functional groups of sintered-clay waste and As­(V). The reaction ratio of As­(V) to Al3+ was 1:1. 12 Al3++H3AsO4→AlAsO4↓+3H+ The M–OH groups on the surfaces of SiO2 and metal oxides in sintered-clay waste could complex with arsenic ions. The arsenic complex structure is shown in Figure g and the chemical reaction eqs and . The reaction ratio of arsenic ions to M–OH was 1:1–2:1. When aluminum oxides and silicates coexisted, the above complexation was more efficient, and the effect of removing arsenic was more significant. 13 M‐OH+As3+/As5+→M‐O‐As 14 2M‐OH+As3+/As5+→M‐O‐As‐O‐M Arsenic removal of sintered-clay waste was divided into three stages: the deceleration purification stage, the accelerated purification stage, and the stable purification stage (Figure h). As­(III) existed as H3AsO3 in raw water. The first stage, known as the deceleration purification stage (AC section), included the arsenic descending segment (AB) and the arsenic rising segment (BC). In the descending segment (AB), arsenic was removed by physical adsorption and complexation. Physical adsorption included pore adsorption of the sintered-clay waste and flocculation adsorption. The large pore-specific surface area of the the sintered-clay waste guaranteed a stable adsorption effect on As­(III), and the adsorption capacity of the flocculating constituent Al­(OH)3 for As­(III) was greater than that for As­(V). In the arsenic descending segment, less Al­(OH)3 was generated and participated in flocculation, so the KSP of Al3+ did not reach the reaction level with H3AsO4 to form the precipitate AlAsO4, and no chemical reaction occurred for arsenic removal. During the early deceleration purification stage, Al3+ and As­(V) did not meet the KSP conditions; thus, no precipitation occurred during pore adsorption and complexation reactions. In the later stage, the arsenic removal rate was lower than the arsenic release rate, causing the arsenic content in water (Figure a). The second stage, known as the accelerated purification stage (CD section), was the main stage for arsenic removal , during which both As­(III) and As­(V) underwent their maximum removal rates. Adsorption, complexation, and chemical reactions all occurred during this stage The third stage was the stable purification stage (DE section), in which arsenic removal capacity reached a relatively stable state and then entered the attenuation period (after E): the purification capacity of the sintered-clay waste weakened, and the arsenic content in the filtered water upturned. Transformation of arsenic removal in high-hardness arsenic-containing water. (a) Mechanism of Al­(OH)3 flocculation and dissolution for arsenic removal; (b, c) purification mechanisms of high-hardness arsenic-containing water in large-particle sintered-clay media and small-particle sintered-clay media, respectively. Influence of Particle Gradation on Purification Due to the differences of particle voids, porosity, and pore-specific surface area, the adsorption effect of filter media with different particle gradations was significantly different. The removal rate of Ca2+ and Mg2+ was related to the pore-specific surface area, silicate, and aluminate components in the filter media, while that of arsenic was closely related to the pore-specific surface area and the Al­(OH)3 generated during the removal of Ca2+ and Mg2+. The pore-specific surface area of the large-sized filter media was smaller than that of its smaller counterparts; therefore, the removal rates of Ca2+, Mg2+, and arsenic in the small-sized filter media were better than those in the larger ones. However, the small particle-size filter media had a greater effect on water flow than the large particle-size ones because the particle voids of the large-sized filter media were more voluminous than those of the smaller ones (Figure b and c). The smaller the particle size of the filter media, the larger the specific surface area, and the stronger the adsorption purification capacity. When water flowed in the pores, water molecules were attracted to the pore walls by molecular force. The smaller the particle size of the filter media, the smaller the pores, the larger the molecular force, and the greater the resistance to the flow of water; therefore, a moderately sized particle medium was preferred when trying to ensure the removal effect of Ca2+, Mg2+, and arsenic. Analysis of Adsorption Kinetics Model According to the experimental and sampling data, the pseudo-first-order kinetic model could represent the relationship between adsorption capacity and time for analyzing the kinetic process of adsorption capacities of total hardness and arsenic. The pseudo-first-order kinetic model was, 15 ln(qe−qt)=lnqe−k·t Here, q t represents the adsorption capacity at time t, mg/g; q e represents the adsorption capacity at the equilibrium state, mg/g; k is the pseudo-first-order adsorption rate constant, s–1. To establish the relationship between adsorption capacity and time, eq was simplified as, 16 qt=qe(1−e−k1t) The sintered-clay filter media with particle sizes 10–20, 20–40, and 40–60 mesh were filled to the same height in the column with masses of 430, 450, and 490 g, respectively. The adsorption kinetic fitting analysis was conducted in three stages of total hardness and arsenic removal. The termination times for the three stages of total hardness softening of sintered-clay filter media with particle sizes of 10–20, 20–40, and 40–60 were 376, 376, and 400 h, respectively; those for the three stages of arsenic removal were all 448 h. Figure shows the relationships between time and adsorption capacities for total hardness and arsenic. The total hardness softening of sintered-clay filter media with particle sizes of 10–20, 20–40, and 40–60 mesh all underwent three stages: accelerated purification, decelerated purification, and stable purification. The adsorption rate constants and correlation coefficients were 0.00587, 0.00585, and 0.00561 h–1 and 0.9421, 0.9371, and 0.9438 in the relationship between total hardness adsorption capacity and time, respectively. Arsenic removal also underwent three stages: decelerated purification, accelerated purification, and stable purification. The adsorption rate constants and correlation coefficients were 0.00467, 0.00466, and 0.00463 h–1 and 0.9617, 0.9634, and 0.9624 in the model, respectively. Both the adsorption rate constants of total hardness and arsenic changed little, and the total hardness adsorption rate constant was greater than that pertaining to arsenic. The average correlation coefficients of the total hardness and arsenic adsorption models were 0.9410 and 0.9625, respectively, suggesting significant overall correlations. The correlation of the total hardness adsorption model was lower than that for arsenic. Both consistencies of experimental adsorption capacities with the theoretical ones were greater in the middle stages than in the end stages with regard to total hardness and arsenic. In the three-stage purification, the optimal particle size of the filter media was 20–40 mesh for the optimal adsorption capacity of total hardness and arsenic. Relationships between time and adsorption capacities of total hardness and arsenic. (a–c) Relationships between time and total hardness adsorption capacities of the 20–40, 40–60, and 60–80 mesh sintered-clay media, respectively; (d–f) relationships between time and arsenic adsorption capacities, respectively. Application Process Design Based on the research in this paper, the sintered-clay filter media could simultaneously achieve hardness softening and arsenic treatment for high-hardness arsenic-containing water by physical adsorption and chemical processes. The application process was designed as shown in Figure a. The structure of the filtration system is shown in Figure b. The main filtration device included a packing port, an upper water distributor, a silo, and a lower water distributor. The bottom of the filtration device was covered with a layer of quartz sand, and the sintered-clay filter media were placed on the quartz sand, with quartz sand covering the top of the sintered-clay filter media. The water flowed in from the upper water distributor, passed through the filter material, and was collected by the lower water distributor before being discharged. This treatment of high-hardness arsenic-containing groundwater could meet the requirements for nonpotable water (such as that used for bathing, irrigation, etc.). Application process and structure of the filtration system: a) application process and b) structure of the filtration system. An application test was conducted to soften hardness and remove arsenic from high-hardness, arsenic-containing groundwater by sintered-clay waste media. The treatment comparison of the 20–40 and 40–60 mesh media is shown in Figure a and b. The morphology is respectively shown in Figure c and d, but there was little difference in the surface fouling of the particles. The total hardness softening duration meeting the Class II water body standard was 288 days, and the hardness removal rates of the 20–40 and 40–60 mesh sintered-clay filter media were initially high and then decreased, with rates of 46.40–58.52% and 44.72–68.32%, respectively. The accelerated removal of arsenic was completed after 224 days, at which point the arsenic content in the filtered water from the 20–40 and 40–60 mesh filter media was 0.0789 and 0.0777 mg/L, respectively. Subsequently, a long-term stable purification stage was observed. The effects of total hardness softening and arsenic removal are presented in Table . The application process was simple, and the treatment effect was significant. Purification effect on application test. (a,b) Comparisons of the 20–40 and 40–60 mesh filter media for total hardness and arsenic purification, respectively; (c,d) differences in surface fouling of the 20–40 and 40–60 mesh filter media, respectively. 5Quantification Effect of Total Hardness and Arsenic Purification The total hardness (mg/L) Arsenic conten (mg/L) Particle size (mesh)Working time (d)Purified water volume (m3)InflowOutflowTotal hardness capacity (mg/g)InflowOutflowArsenic capacity (mg/g)20–402882880419.37236.77591.680.13990.08270.185340–602882880419.37226.70624.310.13990.08070.1918 Quantification Effect of Total Hardness and Arsenic Purification Discussion It has become a research hotspot to develop adsorbents from waste materials for treating high-hardness arsenic-containing water due to their environmental and economic benefits. Common methods for treating calcium, magnesium, and arsenic mainly include adsorption, electrocoagulation, ion exchange, and membrane technology, among which adsorption is widely used due to its simplicity and low cost. Research on adsorption methods mainly focused on the modification of filter materials in recent times. At present, among filter materials made from waste, electrolytic manganese slag, pyrolyzed waste tires, stainless steel tailings, and fly ash are used for arsenic removal, while lime, alumina, zeolite, fly ash, and lime + alumina are used for water softening. The removal rate of each filter material is shown in Table . 6Comparisons of Arsenic and Total Hardness Removal Rates of the Current Filters with Those in the StudyArsenic removal rate (%)Hardness removal rate (%)Manganese slag<20reference Lime + alumina53.58reference Pyrolysis waste-tire46.4reference Zeolite73.5reference Fly ash85.4reference Fly ashno obviousreference Stainless steel tailings47reference Fly ash + lime64.33reference Sintered-clay waste 53.45 This study Sintered-clay waste 69.36 This study Comparisons of Arsenic and Total Hardness Removal Rates of the Current Filters with Those in the Study The arsenic removal rates of these filter media derived from waste materials are shown in Figure a. In ref , the removal rate of As­(III) was less than 20% in arsenic-containing water (marked as 16.5%) by manganese slag with 23–28% moisture content and 40–250 μm fine particles, which were obtained by leaching rhodochrosite (main component MnCO3) in sulfuric acid. In the reference , the As­(V) removal rate of pyrolysis waste-tire charcoal filter (unmodified) was 40.0% from 5 mg/L As­(V) wastewater at pH 2.3, with a reaction time of 24 h and water temperature of 25 °C, and was 46.40% from 2 mg/L As­(V) wastewater at pH 7, with a reaction time of 2 h and temperature of 25 °C, which closely resembled the actual situation. In reference , 0.5 g of stainless steel tailings (unmodified) was used to purify 0.2 mg/L arsenic-containing water (50 mL), achieving a removal rate of 47% and a removal capacity of 0.0094 mg/g. In ref , an arsenic adsorption experiment using fly ash (unmodified) was conducted in arsenic-containing wastewater, and the optimal removal rate was 85.4% without pH adjustment, using fly ash at 0.03g/mL, an adsorption time of 60 min, and a temperature of 25 °C. Comparisons of adsorbents for high-hardness water softening and arsenic removal. (a, b) Comparisons of this study to other adsorbents for arsenic removal and hardness softening, respectively. The comparison of the adsorbents for high-hardness water softening is shown in Figure b. In ref , the hardness treatment was optimized by lime coagulation (lime+aluminum chloride). The effluent quality was optimal, with the total hardness removal rate being 53.58% under the conditions of lime at 600 mg/L, pH = 8.5–9.5, polyaluminum chloride at 30 mg/L, and polyacrylamide at 0.5–1.5 mg/L. In reference , the softening methods of lime, fly ash, and fly ash + lime were adopted for wastewater hardness treatment. The hardness removal rate was 45.57%, with a minimum of 157 .95 mg/L under the optimal lime dosage of 261 mg/L and a reaction time of 55 minutes . When the raw water hardness was 375.14 mg/L, the optimal fly ash dosage was 270 mg/L, and the reaction time was 55 minutes. The hardness removal rate of fly ash was −0.68% at pH = 8, 2.45% at pH = 9, and 13.73% at pH = 11. Thus, the hardness removal effect of fly ash was not significant. The hardness removal rate of fly ash + lime was 64.33% under the conditions of optimal fly ash at 270 mg/L, optimal lime at 261 mg/L, raw water hardness at 375.14 mg/L, and a reaction time of 48 min. In ref , the total hardness removal rate of lime was 23.4% at pH = 10 when a 55% calcium oxide solution was added to 100 mL of brine. The total hardness removal rate of zeolite was 19.9% at pH = 10, with a magnetic stirring time of 10 minutes, when excess bauxite solid and an appropriate PAM solution were added to 100 mL of brine. The total hardness removal rate of the zeolite filter column was 73.5% at pH = 10, with a flow rate of 0.3–0.5 L/h and a brine water sample of 100 mL. The total hardness removal rate of the alumina adsorbent was 32.6% at pH = 10, with a flow rate of 0.3–0.5 L/h and a water sample of 100 mL. Among the various waste-utilizing adsorbents, fly ash had a higher arsenic removal rate than that observed in this study but could not be used to soften hardness; zeolite had a slightly higher hardness softening capacity than this study, while its arsenic removal capacity was not significant. In this study, the sintered-clay waste could simultaneously achieve water softening and arsenic removal of high-hardness arsenic-containing wastewater, with the removal rates of hardness and arsenic content reaching 69.36% and 53.45%, respectively. Obviously, the sintered-clay waste was a widely available, low-cost, environmentally friendly adsorbent that possessed significant advantages in both water softening and arsenic removal. For arsenic treatment, sintered-clay waste media (derived from the demolition of red and blue bricks from construction) was adopted for the treatment of wastewater, groundwater, or nonpotable use. Post-treatment or material modification would be required to meet stricter standards (e.g., the WHO guideline of 10 μg L–1). After being used, the sintered-clay filters could also be reused as filling materials. Conclusion The main clay components were hydrated aluminum silicate, CaO and MgO, and the surface structural functional groups formed the basis for high-temperature modification. The pores increased in both number and volume, and the pore-specific surface area expanded after the clay was sintered, which made the sintered-clay waste an ideal material for purifying high-hardness arsenic water. The main sintered-clay components were the fixed adsorbents (SiO2, MgCa­(CO3)2) and the effective purification substances (sodium aluminum silicate, potassium aluminum silicate salts, etc.). The total hardness softening of the filtered water from sintered-clay media showed three stages: accelerated, decelerated, and stable, while the arsenic removal process consisted of three stages: the decelerated stage, the accelerated stage, and the stable stage. During the purification periods, the total hardness removal rates of the 10–20, 20–40, and 40–60 mesh filter media were, respectively, 57.94–60.59%, 57.36–63.96%, and 57.74–69.36%; those for arsenic were, respectively, 27.19–48.01%, 28.29–50.63%, and 29.34–53.45%. The pore-specific surface area of the large-particle filter media was less than that of their smaller counterparts. The effective purification periods of the sintered-clay media for total hardness and arsenic were such that 40–60 > 20–40 ≈ 10–20 mesh. The purification rate at each purification stage was such that 40–60 > 20–40 > 10–20 mesh. Considering the purification cost and the purification rate for total hardness and arsenic, the optimal particle size of the sintered-clay media was 20 ∼ 40 mesh. The components SiO2 and MgCa­(CO3)2 increased and SiAs2 and Pb28As12S46 appeared in the filtered sintered-clay media, which showed that precipitates formed and strong adsorbent flocculation substances Al­(OH)3 were produced during the reaction between the ions Ca2+, Mg2+, silicates and aluminates in the sintered-clay media. Ca2+, Mg2+, and arsenic were further removed due to coprecipitation and complexation during Al­(OH)3 redissolution and the M–OH surface groups’ reaction with Ca2+, Mg2+, and arsenic. Both of the confidence ranges for total hardness and arsenic were narrow. The confidence range for total hardness softening changed more in the early stage than in the later stage, while the situation for arsenic was the opposite. Both the adsorption rate constants for total hardness and arsenic changed little, and the total hardness adsorption rate constant was greater than that pertaining to arsenic. The average correlation coefficients for the total hardness and arsenic adsorption models were, respectively, significant at 0.9410 and 0.9625. The use of sintered-clay waste was proposed for purifying high-hardness arsenic-containing water. In such an application, the total hardness removal rate and that for arsenic reached, respectively, 69.36% and 53.45%, both of which were superior to those of existing adsorbents, thereby realizing improved waste utilization. In the future, the research focus could usefully be on the modification technology for the sintered-clay filter media to further optimize the purification effect of high-hardness arsenic-containing water.