Optimization of total hardness removal efficiency of industrial wastewater using novel adsorbing materials

: One of the contaminants of industrial wastewater is hardness, which has harmful effects on our environment when discharge to surface water. Hence, various treatment methods have been studied to remove the hardness of wastewater. Herein, a series of oxide glasses with different compositions were prepared as novel adsorbents to dispose of hardness from industrial wastewater via adsorption treatment technique. It is found that the treatment with borosilicate glass (25% B 2 O 3 –20% SiO 2 – 5% CuO – 5% Bi 2 O 3 – 45% Na 2 O) exhibits the highest hardness removal efficiency. Based on such sample, the boundary conditions, including contact time, temperature, and adsorbent dosage are controlled to obtain the optimum removal efficiency. The optimum boundary conditions are found to be 150 min, and 15g L -1 for contact time, and adsorbent dosage, respectively. Fortunately, the efficiency is unaffected by the temperature of the adsorption process, which indicates no required expended energy and low cost water treatment. The effect of initial hardness concentration and pH of wastewater shows an initial increase in the removal efficiency to a maximum value at 150 mg L -1 and 6.5 – 7.2, respectively. The results from the contact time study were simulated by Langmuir and Freundlich isothermal adsorption models, and well fitted. Finally, the comparison with the previous studies shows that the experimental results exhibit excellent and acceptable hardness removal efficiency at the appropriate contact time and adsorbent dosage.

have an amorphous form; they are formed by the conventional melt quenching technique to the solid case without crystallization (Camilo et al., 2013;Jiménez et al., 2011). Yadav and Singh (2015) found that, forming of glass required a critical cooling rate. The aim of this study is reduce the concentration of the hardness, whatever water source, that to allow treated water to be used without exposure to the problems caused by hard water in terms of health problems when using it for drinking or problems when using it for other purposes. Where the World Health Organization stated that water containing a concentration of less than 60 mg L -1 of calcium carbonate is considered soft as reported by Takahashi and Imaizumi (1988), and this is what is expected to be reached. This study intends to investigate the optimum boundary conditions of the adsorption process by oxide glass, after reaching to an effective oxide glass formula. Finally, this research presents the influence of the wastewater sample properties on the adsorbent efficiency. The structural properties of the prepared oxide glasses were studied by XRD and FTIR spectra, as the structural properties play an important role in the efficiency of adsorbent.

Wastewater samples
In this study, wastewater samples were collected in the first experiment from the drainage of agricultural fertilizers manufactory and chemical industry, which was collected one day before the experimental work in June 2022. In the case of subsequent experiments, samples were synthesized. Synthesized wastewater samples have the same properties as original industrial wastewater samples in order to save the same conditions in all experiments. The wastewater samples have been collected or synthesized one day before the adsorption process, the samples were kept at 5 o C to experimental time, which matches with conditions mentioned by Torfs et al. (2016). Synthesized samples are made by adding the required quantity of CaCl2 and MgSO4.7 H2O into distilled water; this is to simulate the same Ca +2 and Mg +2 values in the initial sample of industrial wastewater and its pH value. The obtained samples have the initial concentration of total hardness 319 ± 7 mg L -1 and pH 8.9, concentrations of Ca +2 and Mg +2 are 104 -108 mg L -1 and 12.6 -13.6 mg L -1 respectively. Analysis was carried out according to Standard Methods (APHA, 2002;Federation, 1999).

Oxide glasses definition
There are two different types of solid materials: crystalline and amorphous. The atomic arrangement in solid crystalline is regular; amorphous solid has an irregular atomic arrangement such as glass, ceramic, gel, and polymers. The types of glasses are oxide glasses and non-oxide glasses, oxide glass is composed of former, intermediate, and modifier. The former is the main ingredient in oxide glass; boron oxide, phosphor oxide, silicon oxide, and germanium oxide are the known formers in oxide glasses synthesized. Not only single-network former glasses, but also double-network former glasses were investigated. Borosilicate glass is one of the double network former glasses. The interlinking of some borate and silicate networks is also possible with such glasses as reported by Axinte (2012). The modifier is an oxide of one of the transition metals, which is added to modifying properties of glass composition such as decreasing the melting point of the former. The intermediate is used to configure the links between the atomics of the glass composition.

Adsorbents (Oxide Glasses) synthesis
Conventional melt quenching is the technique used for oxide glasses synthesis. Three groups of oxide glasses were prepared: group I, group II, and group III. The chemical formula of group I is 45% GFO -5% CuO -5% Al2O3 -45% Na2O; where GFO (glass forming oxide) is B2O3 and SiO2, labeled A and B respectively. The chemical formula of group II is 45% GFO -5% Cu2O -5% Al2O3 -45% Na2O; GFO is B2O3 and SiO2, labeled C and D respectively, this group tested to compare CuO and Cu2O influence in adsorbent efficiency. Copper monoxide and dioxide were used as a modifier. The chemical formula of Hybrid borate and silicate (borosilicate) glasses (group III) is 25% B2O3-20% SiO2 -5% CuO -5% GIO -45% Na2O; GIO (glass intermediating oxide) is Bi2O3, SnO, Al2O3 and, labeled I, J, K respectively, in order to find a better adsorbent. Table 1 summarizes the chemical composition of each system. Sodium carbonate is used as a precursor of Na2O for reducing the melting point. The quantity of each chemical component in oxide glass composition was calculated based on its atomic weight by Eq. (1). All components of each oxide glass system were mixed and milled together in a porcelain mill, after determining the weight of each component for each oxide glass system. The porcelain crucible was used to contain the mixture, and then it was put in the furnace at 1000 o C for 60 min. Once the melt came out from the furnace to room temperature, the glass is formed as a piece with its properties. A grinding machine (Fritsch Mortar Automatic Grinder) was used to grind glass pieces to small particles size 125 -63 µm.

Adsorption process
The adsorption efficiency of hardness removal on adsorbent (oxide glass types) is tested in experimental work, and then investigates the influence of experimental boundary conditions such as contact time, temperature, and adsorbent dosage, and the influence of the wastewater sample properties such as initial hardness concentration and pH value. In the first step of experimental work, the adsorbent dosage was 7.5 g L -1 , and it was added to the wastewater sample. Then magnetic stirrer used to stir the mixture at 300 rpm for 30 min at 40 o C. Buchner funnel used to filter the mixture after the stirring step, and qualitative filter paper by size 0.45µm pore used in the filtration step. The filtrate was kept at 5 o C to analysis time after the filtration step. The spectrophotometer (HACH DR6000) used for hardness value determination. Figure 1 summarizes experimental setup steps. .

Fig. 1 Summarization of experimental setup steps
Oxide glass hardness removal efficiency (E) by percent and the adsorption capacity (Qt) were calculated by Eq.
(2) and (3) respectively: Where: Co is the initial concentration of hardness (mgH L -1 ), Ct is the concentration of hardness ion (mgH L -1 ) at time t, Cf is the final concentrations of hardness ion (mgH L -1 ), V is the volume of wastewater sample (L), m is the adsorbent (oxide glass) mass (g), and Qt is adsorption capacity of hardness ion per gram of oxide glass (mgH gG -1 ) at a time t.

3.Results and Discussions 3.1. Characterizations of oxide glass adsorbents 3.1.1 XRD analysis
Oxide glasses (adsorbents) were investigated by X-Ray Diffraction instrumental techniques (XRD). Bucker D8 Advance Diffractometer was used in the two theta range from 10 to 80 degrees. The XRD patterns of oxide glasses are illustrated in Figure 2. It is observed that no crystalline peaks appear in all XRD patterns, indicating the amorphous nature of all glasses.

FTIR analysis
Borate, silicate, and borosilicate glasses were investigated by Fourier-transform infrared spectroscopy (FTIR). Platinum -ATR (Bruker Alpha) was used in the wavenumber range of 450-4000 cm −1 . The IR of borate glasses are illustrated in Figure 3. The absorption peaks of main IR in B2O3 containing glasses reside at wavenumbers of 710, 1260, and 1420 cm −1 . Feller and Kasper (1987) reported that the absorption peak near 1420 cm −1 appeared according to the groups boroxol ring stretching, and the 1260 cm −1 vibration is according to the B-O-B bond formation the linkage of boroxol groups to neighboring groups. The absorption band of bending vibrations of BO3 triangles appears near 700 cm −1 , and stretching vibrations of BO3 units with nonbridging oxygens (NBOs) appears the absorption band near 1380 cm −1 (Ardelean and Toderaş, 2006;Hartman and Chan, 1993;Krogh-Moe, 1965). B-O stretching of trigonal BO3 units of vibrations 1200-1500 cm −1 , and B-O stretching of tetrahedral BO4 − units of vibrations 850-1200 cm −1 , and 600-800 cm −1 is due to bonding vibrations of B-O-B groups in different borate segments (Gautam et al., 2012;Jellison and Bray, 1978).   Figure 5a shows the IR spectra for borosilicate glass sample K. The bending vibration peak of Si-O-Si lies at 473.5 cm -1 , and the stretching vibration peak of O-Si-O near 800 cm -1 . The absorption peak near 1187.26 cm -1 is the bending vibration peak of Si-O-Si, which synchronizes with the stretching vibration peak of BO4 (antisymmetric). In addition, the absorption peaks near 725.35 cm -1 and 1450.35 cm -1 are bending vibration peak of BO3 and anti-symmetric stretching vibration peak of BO3, respectively (X. Li et al., 2018). Theoretically, when adding a small amount of Al2O3 into borosilicate glass, Al 3+ has priority to attract free oxygen to form [AlO4], accessing the structure of SiO2 to achieve the more stable structure (X. Li et al., 2018). Further, due to the oxygen deficiency, B 3+ essentially exists in the form of [BO3] instead of [BO4]. In Figure 5b and 5c, the same peaks mentioned are achieved in the case of the borosilicate glass with Bi2O3 and SnO respectively, but with slight shifts in the wavenumbers as a result of the use of another glass intermediating oxide.

BET Analysis
The surface area and pore size of oxide glasses (adsorbents) were investigated by Brunauer-Emmet-Teller (BET). Nova touch LX 4 Quantachrome instrument version 1.21 was used to measure the surface area and average pore radius of oxide glass samples in powder case.
The results are illustrated in Table 2. Determining the surface area and pore size of the adsorbents are extremely important because of their impact on the efficiency of the adsorption process. The surface area indicates the availability of active sites, which receiving atoms or molecules of adsorbate. The surface area can be controlled by controlling the size of the adsorbent particles, when the size of the adsorbent particles is minimized through the grinding process, the surface area increases for the same amount of adsorbent, that leads to improve the efficiency of the adsorption process (Ahn et al., 2008;Hwang et al., 2011;Mossa Hosseini et al., 2011). Therefore, the pore size of adsorbent and its arrangement, both have a significant impact on efficiency of adsorption process. Pore size controls if molecules of adsorbate can enter to them, where adsorbate molecules can be adsorbed in the pores of adsorbent molecules, which have size larger than molecules of adsorbate as reported by W. Li et al. (2019). As for the arrangement of the pores, Suresh Kumar et al. (2019) reported that it has an effect on the one hand if they present as long isolated cylinders or they are highly branched.

Influence of oxide glass composition
Influence of oxide glass chemical composition on hardness removal efficiency is shown in Figure  6. The calcium ions removal efficiency with adsorbents A, B, C, and D is around 36%, 37%, 31%, and 11% respectively, and the magnesium ions removal efficiency with adsorbents A, B, C, and D is around 35%, 28%, 33%, and 7% respectively. It's observed that the hardness removal efficiency of samples A and B is approximately similar; the hardness removal efficiency of sample C is close to A and B, as shown in Figure 6a. The efficiency of sample D in removing hardness is the lowest and with a large difference. Therefore, the modifier Cu2O was dispensed, which has a weak improvement effect on hardness removal efficiency in case of borate glass and significant decrease in case of silicate glass. Hence, the decision was made to prepare a glass system with merge boron oxide and silicone oxide in one chemical composition. : borosilicate glass samples This explains the reason for the desirable using CuO as a modifier in the oxide glass system at the next step. Based on the results, the formers used for the next stages were boron oxide and silicone oxide. Boron and silicon oxide together are the formers with modifier CuO to investigate intermediate types. Figure 6b shows the hardness removal efficiencies of samples I, J, and K; they have the same chemical composition except for the intermediate oxide. Oxide glass sample I gives the best hardness removal efficiency, which contains Bi2O3 as intermediate oxide, bearing in mind that the differences between the hardness removal efficiency are small, the calcium ions removal efficiency with adsorbents I, J and K is around 43%, 36%, and 38% respectively, and the magnesium ions removal efficiency with adsorbents I, J and K is around 43%, 39%, and 24% respectively. And by reviewing Table 2, which displays the specific surface area and pore radius of adsorbents, it was found that the increase in surface area of adsorbent is not necessary to improve the adsorption efficiency, also the pore diameter as long as the adsorbents are different in chemical composition as reported by Santoso et al. (2020). The efficiency of adsorption process depends on several factors, not only the surface area and pore size of adsorbent but also the surface charge of adsorbent and adsorbate has a significant effect, but in the case of using the same adsorbent, the specific surface area has an important role, as mentioned in previous studies, which presented the decreasing of adsorbent particles size leads to increase the surface area of adsorbent, which in turn enhances the quantity of active sites and thus increases the efficiency of adsorption process (Baral et al., 2007;Imchuen et al., 2016;Y. Wang et al., 2006). In this study the comparison is between more than one adsorbent, which have different characteristics and chemical composition. The surface area of an adsorbent may be larger than another, but the surface charge of adsorbent and adsorbate is an important factor to achieve the attraction and increase adsorption capacity (Garg et al., 2015;Liu et al., 2014), which is affected by pH value (Mahmoodi and Saffar-Dastgerdi, 2019;Olusegun et al., 2018). Adsorbent chemical composition, as the interactions between oxides to composite the oxide glass sample, it can create electrostatic interferences for instance the electrical surface charges on the oxide glass particles, which increase or decrease the attraction between the adsorbed in wastewater sample and the surfaces of the oxide glass particles (Malamis and Katsou, 2013). Moreover, the molar volume of adsorbate may be larger than the pore radius of adsorbent, which leads to difficulty in completing the adsorption process for this adsorbent, as the pores size and its arrangement has a significant effect on possibility of contaminants molecules accessing to the inner part of adsorbent (W. Li et al., 2019;Ouyang et al., 2020;Suresh Kumar et al., 2019). All this indicates that the adsorption process depends on the nature of adsorbent and adsorbate.

Influence of Boundary conditions
As the type of used adsorbent is important, boundary conditions of experimental work are no less important than it, to achieve the superiority of the adsorption process. Some boundary conditions were studied in this work and their influence on the adsorption process performance.

Contact time
Three types of oxide borosilicate glass I, J, and K were examined in five periods of contact time: (30, 60, 90, 120, and 150 min). Studies were done in same value of conditions as following: adsorbent dosage 7.5 g L -1 , pH 8.9, initial concentration of total hardness 319 ± 7 mg L -1 , and temperature 40 o C. Periods of contact time were measured and set by minutes. The performance of adsorbents in hardness removal at different contact times is illustrated in Figure 7. As observed, the efficiency of three adsorbents in the minimization of hardness was amended by increasing the contact time. Adsorbent I with intermediate Bi2O3 achieved the best removal efficiency of total hardness over the duration of the adsorption process. The removal efficiencies of Ca 2+ and Mg 2+ with adsorbent I at 30 min are around 43%, then they are increasing over time to be around 64% in 150 min. The removal efficiencies of Ca 2+ and Mg 2+ with adsorbent J at 30 min are around 36% and 39% respectively; these at 150 min are around 54% and 56% respectively. The removal efficiencies of Ca 2+ and Mg 2+ with adsorbent K at 30 min are around 38% and 24% respectively; these at 150 min are around 52% and 54% respectively. In general, with increasing the contact time, more active sites were occupied achieving its maximum capacity as mentioned by Ashour et al. (2017). There is high increase in removal efficiency was observed at the initial phases of contact time because of the availability of active sites for the adsorption process, and the adsorption process tends to equilibrium gradually as reported by Qin et al. (2010).
The results from the contact time study were analyzed by Langmuir and Freundlich isothermal adsorption models. Langmuir theory supports that the adsorption process is monolayer adsorption (Eq. (4)) (Langmuir, 1917), it is based on assumptions 1-Dynamic equilibrium (rate of desorption is the same of adsorption rate at equilibrium point) 2-The surface of adsorbent material is homogenous 3-Neglect the interaction between the adsorbate molecules and adsorbent surface

= 1 + . (4)
Freundlich theory supports that the adsorbent surface isn't homogenous (Eq. (5)) (Gawande et al., 2017). The theory works on the assumption that adsorption sites of the adsorbent surface have different adsorption energy.
The parameter Keq is Langmuir constant; the parameters KF, n are Freundlich constants; Ce is the adsorbate concentration at equilibrium adsorption (mg L -1 ); Qe is the adsorption capacity at equilibrium adsorption (mg g -1 ); Qmax is the maximum adsorption capacity (mg g -1 ). The adsorption isotherm data were reported in Table 3, and they were illustrated in Figures 8 and 9.
Log Q t (mg g -1 ) Fig. 9 Linear plot of Freundlich isotherm of Ca 2+ and Mg 2+ on adsorbents I, J, and K As shown in Figures 8 and 9, and Table 3, the conclusions from the Langmuir and Freundlich graphic relations model well fitted with both isothermal adsorption models. That means the adsorption process can be carried out chemically or physically for the three borosilicate glasses.

Temperature:
The influence of temperature in range of 30 -70 o C was investigated; its results were illustrated in Figure 10. The process was conducted under boundary condition as follow: contact time of 150 min and adsorbent dosage of 7.5 g L -1 , and the characteristics of the wastewater samples are, pH value of 8.9, initial concentration of total hardness 319 ± 7 mg L -1 . Figure 10 shows the semi constant behavior of hardness adsorption with increasing temperature in most cases, and even an increase in removal efficiency in some cases isn't noteworthy compared to the cost of temperature increase. That means the influence of temperature can be negligible in the hardness removal efficiency of tested adsorbents in the range studied, these results matched with what have been published in some previous studies (El-Nahas et al., 2020;Lodeiro et al., 2006;Rolence, 2014;Sanjuán et al., 2019;Sepehr et al., 2013a). Adsorbent I removal efficiency of Ca 2+ and Mg 2+ is in the range around 64 -65% and 63 -65% respectively, adsorbent J removal efficiency of Ca 2+ and Mg 2+ is in the range around 54 -55% and 56 -57% respectively, and adsorbent k removal efficiency of Ca 2+ and Mg 2+ is in the range around 51 -52% and 54 -57% respectively, at temperature degree from 30 to 70 o C. Then ambient temperature can consider as acceptable temperature for the removal of total hardness using this technique without affecting efficiency.

Adsorbent Dosage:
The influence of adsorbent dosage in range of 2.5 -15 g L -1 was examined in adsorption of total hardness; at contact time 150 min, pH 8.9, initial concentration of total hardness 319 ± 7 mg L -1 , and temperature 40 o C. The results were illustrated in Figure 11. As seen in Figure 11 removal efficiencies of Ca 2+ and Mg 2+ are improving with adsorbent dosage increases, that attributed to the availability of more active sites with increasing adsorption surface area by adsorbent dosage increase. In the case of adsorbent I, removal efficiencies of Ca 2+ and Mg 2+ are around 49% and 39% respectively with an adsorbent dosage of 2.5g L -1 . Removal efficiencies of Ca 2+ and Mg 2+ with adsorbent I increase to around 95% and 94% respectively with an adsorbent dosage of 15 g L -1 . The same behavior was observed with adsorbents J and K but with lower performance than adsorbent I. In the case of adsorbent J, removal efficiencies of Ca 2+ and Mg 2+ are around 37% and 32% respectively with an adsorbent dosage of 2.5g L -1 , then these increase to around 85% and 81% respectively with an adsorbent dosage of 15 g L -1 . In the case of adsorbent K, removal efficiencies of Ca 2+ and Mg 2+ are around 33% and 27% respectively with an adsorbent dosage of 2.5g L -1 , then these increase to around 84% and 80% respectively with an adsorbent dosage of 15 g L -1 . The previous notes mean the occupied sites of adsorbent I is greater than the occupied sites of adsorbents J and K, and the adsorption process improves with the adsorbent dosage increasing, as this allows the presence of active sites in a larger quantity to accommodate the hardness ions. Previous studies reported the same behavior with increasing removal efficiency with adsorbent dosage; they pointed that attributed to adsorption increasing following the increase dosage of adsorbent due to availability of adsorbent active sites (Das and Das, 2013;Esposito et al., 2001;Xie et al., 2015).

Influence of wastewater Properties 3.4.1 Initial hardness concentration
The influence of the initial hardness concentrations was studied, that in the case of initial concentration of total hardness from 50 : 318 mg L -1 and pH value 8.9, and boundary conditions: contact time 150 min, adsorbent dosage 7.5g L -1 , and temperature 40 o C. The results are illustrated in Figure 12. As it is observed, adsorption efficiency was improved by increasing the initial hardness concentration from 50: 150 mg L -1 , and then adsorption efficiency was decreased by increasing the initial hardness concentration from 150: 318 mg L -1 . As Figure 12a shows, in the case of the initial hardness concentration is 50 mg L -1 , the removal efficiency of Ca 2+ is around 65%, 47%, and 48% for adsorbents I, J, and K respectively. The removal efficiency of Ca 2+ is increasing with increasing the initial hardness concentration, which reaches around 79%, 60%, and 59% with a hardness concentration of 150 mg L -1 . The removal efficiency of Ca 2+ start to decrease with increasing the initial hardness concentration from 150 mg L -1 to 318 mg L -1 , as it is around 64%, 54%, and 52% for adsorbents I, J, and K respectively with initial hardness concentration of 318 mg L -1 . The same behavior was observed with Mg 2+ removal efficiency in Figure 12b. At the initial hardness concentration 50 mg L -1 , removal efficiency is around 64%, 46%, and 47% for adsorbents I, J, and K respectively, and then this reaches around 73%, 60%, and 58% with hardness concentration 150 mg L -1 . Then the removal efficiency of Mg 2+ decreases with initial hardness increasing, as it is around 64%, 56%, and 54% for adsorbents I, J, and K respectively with an initial hardness concentration 318 mg L -1 . In case of low hardness concentrations, the ions might be attracted to active sites of adsorbent. At the increasing of ions concentrations, more ions have the capability to occupy active sites. That refers to the differences in the mechanism of ion adsorption, ion adsorption is subjected to adsorbent pores, paths of the arranged lattice, and the need for interchangeable cations' displacement (Arancibia-Miranda et al., 2016;Es et al., 2019;Farrag et al., 2017). The results confirm that excessive initial hardness concentration leads to increase in the adsorbent capacity to adsorb but to a certain limit. If it is considered that surface adsorption is the ion removal driving force that means increasing the initial concentration of Ca 2+ and Mg 2+ leads to improve the driving force (Payus et al., 2019;Sepehr et al., 2013a). After an initial hardness concentration of 150 mg L -1 , the removal efficiency for Ca 2+ and Mg 2+ decreases with the initial hardness concentration increasing, which indicates the active sites for adsorbent, it is no longer sufficient to accommodate the same proportion from the initial hardness concentration. Thus, the adsorption efficiency of the adsorbent decreases because of active sites occupied with Ca 2+ and Mg 2+ cations (Kyzioł-Komosińska et al., 2015;Lee and Rees, 1987;Sanjuán et al., 2019).

pH value:
The pH of the sample has an important influence on the contaminants' adsorption processes as reported by Mahmoodi and Saffar-Dastgerdi (2019), as that effect contaminants' ionization and the adsorbent surface charge. The influence of pH on the removal efficiency of glasses was investigated from 4.1 to 11 at a contact time of 150 min, adsorbent dosage of 7.5 g L -1 , and temperature 40 o C, the initial hardness concentration for wastewater samples was 319±7 mg L -1 . The pH was controlled by adding 0.1M NaOH or 0.1M HCl solution and measuring by a pH meter (JENWAY 3510). As observed in Figure 13, the removal efficiency percentage of Ca 2+ and Mg 2+ with adsorbent I increases from around 49% to 93% and from around 47% to 91% with increasing pH from 4.1 to 6.5 respectively, and then these percentages decrease from around 92% to 49% and from around 88% to 47% with increasing pH from 7.2 to 11 respectively. The same behavior appeared with adsorbents J and K, the removal efficiency percentage of Ca 2+ increases with adsorbent J from around 44% to 85%, and with adsorbent K from around 42% to 84% with increasing pH from 4.1 to 6.5 respectively. Then the removal efficiency percentage of Ca 2+ decreases with adsorbent J from around 84% to 39% and with adsorbent K from around 82% to 36% with increasing pH from 7.2 to 11 respectively, and the same case is in the removal efficiency percentage of Mg 2+ . That is attributed to the high level of removed proton of the functional group on the hydrogen bond surface, which happens in case of high pH as mentioned by Aragaw and Ayalew (2019). The best removal efficiencies of oxide glasses I, J, and K were achieved at pH range 6.5 -7.2, which is close to the range reported by Barathi et al. (2014), as the adsorbent surface has a negative charge and the adsorbate cations have a positive charge, this is due to the effect of the pH value on a adsorbent surface charge as reported by Olusegun et al. (2018). The weak adsorption capacity at an acidic medium (from 4.1 -5.3) is due to ion exchange proton in functional groups or contest of H + with ions of metals to fill the active sites of the oxide glasses (Iqbal et al., 2009;F. T. Li et al., 2007;Saeed et al., 2005). As for the alkaline media (from pH 8.9 -11), Lodeiro et al. (2006) believe that the metal hydroxide formation mainly causes the metal adsorption decrease as it is low-soluble.   150 min 1000 mg L -1 5 g L -1 6.5 nano powder zeolite (El-Nahas et al., 2020) 90 % Ca 2+ , Mg 2+ 30 min >1000 mg L -1 10 g L -1 NA modified bentonite (Kadir et al., 2017) 6.67% Ca 2+ , Mg 2+ 75 min 120 mg L -1 NA NA pine nut cover (Kaur et al., 2021) 90.44% Ca 2+ 15 min 680 mg L -1 6 g L -1 6-7 Coffee ground activated charcoal (Khaerul and Syahputri, 2021) 97.8 % Ca 2+ 75 min 565.17 mg L -1 10 g L -1 7 Nano-zeolite-graphene oxide composite (Konale et al., 2020) 98% Ca 2+ 120 min 250 mg L -1 1 g L -1 7 Coconut shell activated carbons (Rolence, 2014) 60 % 10 h 655 mg L -1 240 g L -1 6.3 Natural zeolite (Hailu et al., 2019) 79.9% Ca 2+ , 85.6% Mg 2+ 30 min 658 mg L -1 (TH) 50 g L -1 6.9 Amberlite 748 (Yu et al., 2009) NA 8 h 100 mg L -1 120 mg L -1 9.5 Kaolinite smectile (Es et al., 2019) 98.24% Ca 2+ , 99.24% Mg 2+ 30 min 100 mg L -1 1 g L -1 8 Coconut shell powder (Tomar, 2018) 62.58 % 3 h 890 mg L -1 6 g L -1 7 Neem leaves (Tomar, 2018) 59.52 % 3 h 420 mg L -1 2 g L -1 7 Modified pumice (Sepehr et al., 2013b) 61% Ca 2+ , 51% Mg 2+ 240 min 150 mg L -1 6 g L -1 6 Borosilicate glass I (current study) 93% Ca 2+ , 91% Mg 2+ 150 min 319±7 mg L -1 7.5 g L -1 6.5  NA refers to no available information.

Conclusions
The results of this research referred to that borosilicate glass achieved noticeable success in adsorption process, especially sample (I). Hardness removal efficiency increases with increasing contact time from contact time 30 min to 150 min. The conclusions from the Langmuir and Freundlich graphic relations model well fitted with both isothermal adsorption models. That means the hardness adsorption process can be carried out chemically or physically for three borosilicate glasses. The improvement of hardness removal efficiency is slight and not noticeable with increasing temperature, that in range of tested temperature from 30 -70 o C. The adsorption process efficiency increases with increasing the adsorbent dosage, at the range of tested dosage from 2.5 g L -1 to 15 g L -1 . Adsorption process performance is variable with increasing the initial hardness concentration, hardness ions removal efficiency increases in case of increasing the initial hardness concentration from 50 to 150 mg L -1 , then it decreases with increasing the initial hardness concentration from 150 to 318 mg L -1 . Adsorption process performance achieves to its best case at pH value range (6.5 -7.2). As a conclusion, the optimum boundary conditions for enhancing removal efficiency of total hardness from industrial wastewater using borosilicate glass type I (B2O3 +SiO2 as former, and CuO as modifier, with intermediate Bi2O3) as adsorbent are; 150 min contact time, and adsorbent dosage 15 g L -1 , with the initial hardness concentration equal to 150 mg L -1 , and pH range 6.5 -7.2 for wastewater properties. It is recommend further investigation to identify the influence of use other oxides as a modifier or others as intermediate, in addition to study the influence of sample (I) on other contaminants in wastewater.

Declarations: Ethical approval
This article does not contain any studies with human participants or animals performed by any authors.

Competing interests
The authors declare they have no competing interests.