Removal of calcium from water by zeolites with gravity-driven membrane filtration for water treatment without electricity

Abstract

Hard water creates issues due to scaling that impacts industrial and domestic applications. Zeolite adsorption is an effective method for calcium removal. However, treated water requires additional filtration to remove suspended solids, which uses energy under traditional methods. Considering applications at sites where electricity is unavailable, this study removed calcium from water using zeolite adsorption and gravity-driven membrane (GDM) filtration. Four zeolite types were assessed: powdered (P1 and P2), bead-shaped (B) and granular (G). The properties of the zeolites significantly affected calcium removal and GDM filtration performance. Zeolites with low Si/Al ratios (high Al content) exhibited high Ca removal because negatively charged lattice sites were created by replacing Si with Al. P1 and B with Si/Al ratios of 1.40 and 1.49, respectively, exhibited high Ca-removal efficiencies of 92.6% and 99.8%. In contrast, P2 and G with high Si/Al ratios from 4.71–5.35 showed lower removal efficiency of 29.8–43.7%. The average sizes of P1, P2, B, and G were 9.68, 5.73, 2134, and 3639 μm, respectively. At similar Si/Al ratios, smaller particles exhibited faster adsorption rates. However, as the zeolite size decreased, the GDM flux decreased. During GDM filtration after zeolite adsorption, the permeate flux of water treated with large B and G zeolites was higher than that of water treated with small P1 and P2 zeolites. Finally, the zeolites were evaluated based on three criteria: Ca removal, GDM water flux, and price, which provides useful guidance for selecting appropriate zeolites.

---

Water impact

In the areas where people do not have access to electricity, calcium removal from water can be accomplished by zeolite adsorption combined with gravity-driven membrane filtration. Zeolite properties, such as size and Si/Al ratio, significantly affect calcium removal and membrane filterability. This study provides useful guidance for selecting zeolites for calcium removal.

1. Introduction

Hard water refers to water with a high content of minerals, such as calcium and magnesium. A large amount of calcium ions adversely affects the use of industrial and domestic water via scaling in pipes and boilers, reducing the performance of water infrastructure and causing difficulty in using detergents and personal hygiene products.¹ Several studies have been conducted on the harmful effects of hard water on health, finding that hard water may increase the prevalence of atopic eczema and dermatitis.^2–4

Adsorption for calcium removal is an effective, inexpensive, easy-to-use, and easy-to-operate method from both environmental and epidemiological perspectives.^5–7 Zeolites, activated carbon, clay minerals, biomaterials, and industrial solid waste have been extensively used as adsorbents for water treatment.^5,8–11 Among these adsorbents, zeolites exhibit a high cation adsorption capacity for cations with high thermal and chemical stability.¹² Moreover, zeolites are abundant minerals globally and are characterized by high accessibility and cost-effectiveness.^8,13–15 Because of these attributes, zeolites find extensive use in industries for water softening including calcium removal.¹⁵

Zeolites are microporous crystalline aluminosilicates formed by AlO₄ and SiO₄ tetrahedra connected by the sharing of oxygen atoms.¹⁶ Zeolites are produced in nature via alkaline hydrothermal degeneration¹⁷ and can be synthesized artificially.^18–20 The negative charges of zeolites are generated by the homomorphic substitution of Al³⁺ for Si⁴⁺.^21,22 Interactions with additional cations can compensate for the negative charge of zeolites. Therefore, zeolites are effective in removing calcium ions from drinking water.^10,13,23

However, water from which contaminants are removed by zeolite adsorption cannot be used immediately. To use water treated by zeolites, a process capable of removing suspended solids, including zeolites, is required. Membrane filtration processes are suitable for this purpose. Among membrane processes, gravity-driven membrane (GDM) filtration has been considered considering its applications in developing countries and remote rural areas. GDM filtration can achieve a stable flux without using electrical energy, making it suitable for areas with low energy availability. Globally, approximately 40% of the population in developing countries does not have access to electricity.²⁴ GDM filtration has been successfully applied to surface water treatment for drinking water production in disaster-stricken areas, developing countries, and remote areas.^25–27

Previous studies have been conducted on water treatment combining zeolite adsorption and membrane filtration. Rohani et al. conducted a study on the removal of ammonia from surface water using flat membranes with added zeolite adsorbents.²⁸ Mulyati and Arahman studied the removal of iron ions from groundwater by combining ultrafiltration and zeolite adsorption processes.²⁹ Blöcher et al. combined zeolite adsorption and flotation processes with membrane filtration to increase the removal of heavy metals such as copper, zinc, and nickel.³⁰ Some authors have studied the effect of incorporating zeolite particles in membrane fabrication for the removal of heavy metals. Adam et al. conducted an experiment to remove Cr(VI) using a hybrid adsorptive hollow fiber ceramic membrane based on zeolite.³¹ Yurekli researched the combined process of adsorption and filtration of heavy metal-containing wastewater using a zeolite nanoparticle-impregnated membrane.³² Motsa et al. conducted a study on the preparation of polymer composite membranes with added zeolite to investigate their heavy metal adsorption properties.³³ Tijani et al. studied the removal of heavy metals using NaA zeolite membranes.³⁴ However, to utilize zeolite-based membranes, machinery to generate pressure and electrical energy is required. Moreover, the process utilizing this membrane-based adsorbent demonstrated lower adsorption performance than the conventional application of powdered adsorbents.³⁵ To the best of our knowledge, there has been no research that combines zeolite adsorption and GDM filtration without using electrical energy. In particular, research on the influence of zeolite properties on water quality and GDM water flux is scarce.

Therefore, this study developed a method to remove calcium from water via zeolite adsorption and GDM filtration. Four types of zeolites were tested, and their properties were analyzed in terms of particle size, elements, crystal structure, functional groups, and surface images. The analyzed properties were correlated with the calcium removal by the zeolites. The GDM filtration performance depending on the zeolite type was compared and analyzed based on the mass, size, and microscopy images of the zeolites deposited on the membrane surface. Finally, the zeolites were assessed for calcium removal, GDM water flux, and price.

2. Materials and methods

2.1. Chemicals and membranes

Four zeolite types were used and designated as follows: powder 1 (P1; Sigma-Aldrich, St. Louis, MO, USA), powder 2 (P2; Duksan Science, Seoul, Republic of Korea), beads (B; Wako Pure Chemical, Osaka, Japan) and granules (G; Duksan Science, Seoul, Republic of Korea). Flat-sheet membranes made of poly(vinylidene fluoride) electrospun fibers (Amogreentech, Gimpo, Republic of Korea) were used for GDM filtration. The nominal pore size of the membranes was 0.3 μm. CaCl₂ (>99.0%, Samchun Chemicals, Seoul, Republic of Korea) was used to prepare feed solutions.

2.2. Preparation of feed solutions

Feed solutions were prepared by adding CaCl₂ to water samples collected from a lake (Maeji Lake, Wonju, Republic of Korea). The water quality of the samples is listed in Table S1 in the ESI.† The calcium concentration in the feed solution was adjusted to approximately 200 mg L⁻¹ by adding CaCl₂.

2.3. Zeolite adsorption pretreatment for calcium removal

The removal of calcium from the feed solution using zeolites was performed as a pretreatment before GDM filtration. The zeolite concentration was adjusted to 10 g L⁻¹ by adding 2 g of each zeolite to 200 mL of feed solution. Adsorption was performed in a shaking incubator (ISS-3075R; Jeio Tech, Daejeon, Republic of Korea) at 25 °C and 250 rpm. In the preliminary experiments the calcium removal efficiency of each zeolite was measured over time. The zeolite adsorption time was set to 3 h.

2.4. GDM filtration

The GDM filtration experiments were conducted using a bench-scale GDM filtration system, as shown in Fig. 1. A flat-sheet membrane with an effective surface area of 12.7 cm² was placed at the bottom of the membrane module. The module had an overflow outlet 0.3 m above the bottom to maintain a constant water head by continuously circulating water from the feed tank, providing a constant filtration pressure of 2.9 kPa. There was no lid on the module; therefore, filtration was performed only by gravity, without electricity. The solutions obtained from the 3-hour zeolite adsorption pretreatment for Ca removal were allowed to stand for 10 s and then transferred to the feed tank. The mass of the permeate produced from the membrane module was measured to calculate membrane water flux. During filtration, the water temperature was kept constant at 25 ± 1 °C.

Fig. 1 Schematic diagram of the bench-scale GDM filtration system.

2.5. Analytical methods

The average particle sizes of the P1 and P2 zeolites were measured using a laser scattering particle size analyzer (Partica mini LA-350, Horiba Ltd., Kyoto, Japan). Zeolites B and G were too large to measure their sizes using a particle size analyzer. Therefore, their sizes were manually measured using vernier calipers. The small debris on the surfaces of B and G were measured using the particle size analyzer as follows: when immersed in water, the small debris of B and G was dislodged from the surfaces. Therefore, to obtain the small debris of B and G, each of B and G was separately put into deionized water, which was shaken well. After 10 s of settling the large zeolite particles, only small debris was collected along with the supernatant. The sizes of the small debris samples were measured using the particle size analyzer. Elemental analysis of the zeolites was performed using an X-ray fluorescence (XRF) spectrometer (S4 Pioneer, Bruker, Billerica, MA, USA) to determine zeolite Si/Al ratios. Scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDS, JSM-7001F, Jeol Ltd., Tokyo, Japan) was used for image analysis of the zeolites, including the zeolites deposited on the membrane during filtration. The zeolite samples deposited directly on the membrane surface were prepared by washing most of the zeolite cake layers with 50 mL of deionized water. X-ray diffraction (XRD) patterns of the zeolites were obtained using an X-ray diffractometer (Ultima IV, Rigaku Corp., Tokyo, Japan) at 40 kV and 40 mA. A Fourier transform infrared (FTIR) spectrophotometer (Vertex 70, Bruker, Billerica, MA, USA) was used to analyze the functional groups of the zeolites. For the FTIR analysis, sample pellets were prepared by grinding the zeolites with KBr.

Ca²⁺ concentrations were determined using inductively coupled plasma-optical emission spectrometry (ICP-OES, 720ES, Agilent Technologies Inc., Santa Clara, CA, USA). Turbidity was measured using a turbidimeter (2100N, Hach, Loveland, CO, USA). Ultraviolet absorbance at 254 nm (UV₂₅₄) and dissolved organic carbon (DOC) were measured using a spectrophotometer (Lambda 365 UV/Vis spectrophotometer, PerkinElmer Inc., Waltham, MA, USA) and a total organic carbon (TOC) analyzer (TOC-LCPH, Shimadzu Corp., Tokyo, Japan), respectively.

3. Results and discussion

3.1. Properties of zeolites

3.1.1. Physical properties of zeolites. Table 1 lists the zeolite sizes. The average diameters of P1 and P2 were 9.68 and 5.73 μm, respectively. For B and G, the average diameters were 2134 and 3639 μm, respectively.

Table 1 Average sizes of zeolites^a

	Zeolite type
	P1	P2	B	G
a The sizes of P1 and P2 were measured using a particle size analyzer (n = 3) whereas those of B and G were obtained by averaging the sizes of 100 grains measured using vernier calipers.
Diameter [μm]	9.68 (±0.17)	5.73 (±0.35)	2134 (±125)	3639 (±740)

---

SEM images of the zeolites are shown in Fig. S1.† P1 had a relatively regular cubic shape, but P2 was irregular in shape. B and G were much larger than P1 and P2 but had small debris on their surfaces (Fig. S1c and d†). The average diameters of debris for B and G were measured using the particle size analyzer at 1.91 and 7.30 μm, respectively (Table S2†). Particle size affects the adsorption rate of adsorbents.^36–38 In addition, the sizes and shapes of particles influence the water flux of membrane filtration.³⁹ In particular, small-sized particles can form a denser cake layer and block membrane pores, increasing hydraulic resistance and leading to a decrease in the water flux.

3.1.2. Elemental composition of zeolites analyzed by XRF. Table 2 presents the XRF analysis results for the zeolites. Si and Al accounted for the majority of the elements in all the samples. The Si/Al mole ratios of P1 and B were 1.40 and 1.49, respectively, indicating that the Si content was relatively low, corresponding to ‘low silica zeolite’.^40,41 In contrast, in the cases of P2 and G, the Si/Al ratio was close to 5.

Table 2 Elemental analysis data for zeolites using XRF with mole ratios of Si to Al

Element [% wt]	Zeolite type
	P1	P2	B	G
Si	48.25	65.19	49.71	62.73
Al	32.99	11.70	31.96	12.80
Na	17.60	2.56	13.60	2.69
K	0.31	8.65	0.63	7.79
Ca	0.23	4.18	0.56	4.45
Fe	0.04	5.03	1.74	6.35
Mg	0	0.62	0	0.70
Etc.	0.58	2.07	1.80	2.49
Si/Al (mole ratio)	1.40	5.35	1.49	4.71

---

A lower Si/Al ratio results in more negatively charged sites in the zeolites.⁴² The negatively charged lattice sites created by replacing Si with Al can be neutralized by other substitutable cations. The smaller the Si/Al ratio, the larger the ion-exchange capacity of zeolites.⁴³ In terms of molecular transport, zeolites with low Si content have large bonding and cage dimensions, which facilitate molecular transport. Low-silica zeolites prefer cations that are small and multivalent and thus have a high charge density.⁴⁰ Therefore, the elemental properties of zeolites influence their calcium adsorption efficiency. Moreover, as the size of zeolite particles decreases, the surface area per unit mass increases, resulting in an increased adsorption rate. Therefore, both physical and elemental properties may determine the calcium adsorption efficiency. This is further discussed below.

3.1.3. XRD patterns of zeolites. The XRD patterns of the zeolites were analyzed to determine the zeolite components (Fig. 2). The detailed XRD analysis plots are shown in Fig. S2.† P1 and B have XRD patterns consistent with that of zeolite A (Linde Type A; LTA) at several 2θ values, denoted as A in Fig. 2.⁴⁴ LTA is a family of aluminosilicate molecular sieves composed of sodalite cages connected by four-membered rings forming a three-dimensional network.⁴⁵ Owing to this structure, LTA is characterized by a markedly open zeolite framework.⁴⁶ This is in accordance with the large cage dimensions of zeolites with low Si content as mentioned above.

Fig. 2 X-ray diffraction patterns of zeolites.

The XRD pattern of P2 almost completely matched that of clinoptilolite (designated C as in Fig. 2). Clinoptilolite is a natural zeolite widely used for metal cation removal from water.^47,48

The peaks of G indicate that it is mainly composed of heulandite, marked as H in Fig. 2.⁴⁹ Both clinoptilolite and heulandite belong to the zeolite structure group HEU.⁵⁰

3.1.4. FTIR spectra of zeolites. FTIR spectral analysis provides information on the bonding composition and structure of zeolites.^51,52Fig. 3 shows the FTIR spectra of the four zeolites used in this study. As previously mentioned, P2 and G contained clinoptilolite and heulandite, respectively, with the HEU structure as the main component. The weak bands at 606–608 cm⁻¹ of P2 and G can be attributed to the 5-membered ring vibrations of the 4–4–1 complex of HEU.⁵² The bands at 790–793 cm⁻¹ can be assigned to the SiO₂ stretching vibration.⁵³ The Si/Al ratios of P2 and G were higher than those of P1 and B, implying that P2 and G had a higher SiO₂ content in the framework. Therefore, the bands at 790–793 cm⁻¹ were stronger for P2 and G than for P1 and B.

Fig. 3 FTIR spectra of zeolites.

The strongest peaks of the four zeolites appeared at 1004–1053 cm⁻¹. In general, all zeolite phases exhibit strong vibration bands in the range of 996–1080 cm⁻¹. This wavenumber range can be assigned to the asymmetric stretching of T–O bonds (T = Si or Al) in aluminosilicates that constitute zeolitic structures.^54,55 The bands in this spectrum depend on the Si/Al ratio of the zeolite framework, resulting in shifts to lower wavenumbers as the Si/Al ratio decreases. This is because the T–O–T angle becomes small when Si⁴⁺ is replaced with Al³⁺ and thus the corresponding Si/Al ratio is low.⁵³ As shown in Table 2, both P1 and B had low Si/Al ratios, and corresponding vibration bands were shown at 1004 cm⁻¹, which is lower than those of P2 and G (1040 and 1053 cm⁻¹, respectively).

In the FTIR spectrum of a zeolite, the broad band between 3000 and 3700 cm⁻¹ corresponds to the Si–OH–Al hydroxyl groups of its framework as well as water molecules absorbed by the zeolite.^56,57 The Si–OH–Al band appears more intense in the FTIR spectrum of the sample with a higher Al content. The FTIR patterns of P1 and B with high Al contents (low Si/Al ratios) showed stronger bands between 3000 and 3700 cm⁻¹ than those of P2 and G. The FTIR spectral analysis of these zeolites is consistent with the XRF analysis results.

3.2. Calcium removal by zeolites

Fig. 4 shows the calcium-ion removal rates over time for each zeolite. P1 exhibited the highest calcium removal rate of 99.8%, even after 15 min of zeolite addition. Calcium removal by B and P2 at 15 min was 28.3% and 27.9%, respectively, indicating similar initial Ca removal rates. However, over time, removal by B increased substantially, reaching 92.6% at 360 min, whereas that by P2 was only 43.7% at 360 min. The removal by G was initially 6.7% and remained at 29.8% even after 360 min, showing the lowest removal rate at all times compared with the other zeolites.

Fig. 4 Calcium removal efficiency using zeolites at a dose of 10 g L⁻¹ (n = 3).

Referring to the Si/Al ratios in Table 2, P1 and B, which had low Si/Al ratios, showed excellent calcium removal, whereas P2 and G, which had high Si/Al ratios, showed low calcium removal. As mentioned previously, the lower the Si/Al ratio of the zeolite, the better the cation adsorption capacity and molecular transport.

Although P1 and B had similar Si/Al ratios, P1 had a high calcium removal rate of 99.8%, whereas B had an initial removal rate of 28.3%. This can be explained by the size difference between P1 and B (Table 1). Generally, when the same amount of adsorbent is added, the adsorption rate increases as the size of the adsorbent decreases.^36–38 Therefore, when the same mass of zeolite is spiked, the adsorption rate of the smaller zeolite is faster.

3.3. GDM system operation

Fig. 5 shows the change in water flux over time during the GDM filtration of solutions from the zeolite adsorption pretreatment. In general, the permeate fluxes of the water treated with B and G were higher than those of the water treated with P1 and P2. G showed the highest flux and B showed the second highest flux.

Fig. 5 Water flux during GDM filtration with solutions from zeolite adsorption pretreatment.

Table S3† shows the mass of cake layers that accumulated on the membrane surface during GDM filtration. This cake layer generates resistance to water transport, causing a decline in water flux. The masses of the cake layers P1 and P2 were 9.928 and 12.367 g, respectively. In contrast, those of B and G were 0.006 and 0.228 g, respectively, which were much lower than those of P1 and P2, respectively. During the adsorption pretreatment, the mass of the zeolites was the same. However, most of the large zeolite particles of B and G were not transferred to the membrane module of the GDM system because the zeolites settled almost immediately after the adsorption treatment. This led to a mass difference between the small and large zeolites deposited on the membrane surface. In general, membrane resistance tends to increase as the particle size decreases and the mass of the cake layer increases, leading to a disadvantage in water flux.^58,59 The difference in water flux for cases B and G can be explained from this perspective. Overall, the smaller particles with larger masses, P1 and P2, exhibited lower fluxes than B and G (Fig. 5).

Initially, P1 exhibited a higher water flux than P2 (Fig. 5). This can be explained by the difference in the particle size distribution between P1 and P2 and the presence of small particles. The particle size of zeolites differed significantly, with P1 measuring 9.68 μm and P2 measuring 5.73 μm (Table 1). This would have had a greater impact on membrane resistance and flux decline in the early stages before thick fouling formation. In addition, P1 had a larger average diameter than P2 but a smaller standard deviation (Table 1). This indicates that P1 had a more uniform size and included fewer small pieces of debris than P2. P1 contained almost no small particles, whereas P2 contained a substantial number of small particles (Fig. S1a and b†). Fig. S3† shows the average size of the particles deposited on the membrane surface during GDM filtration, which were mostly zeolites. The value for P2 was lower than that for P1. Small powder particles may be caught between the electrospun fibers of the membrane, especially during the initial stage of filtration. These small particles can either block or be adsorbed onto the inner walls of the porous membrane, thereby completely or partially obstructing the fine pores. The resistance generated by this fine pore blockage is termed micro-pore blockage resistance, which adversely affects the flux.⁵⁸ The small particles was confirmed by the SEM images of the membranes after removing most of the cake layers using deionized water washing (Fig. 6). Compared to P1 (Fig. 6a′), many small particles were trapped between the fibers of the membrane for P2 (Fig. 6b′). The small particles between the fibers of the membrane were hardly removed, even after washing the surface with deionized water. This resulted in a difference in the initial fluxes between P1 and P2.

Fig. 6 SEM images of the zeolites (a) P1, (b) P2, (c) B, and (d) G deposited on the membrane surface. The images (a′), (b′), (c′), and (d′) correspond to the SEM images analyzed after removing cake layers using 50 mL of deionized water, which revealed zeolites were clogging the fibrous membranes.

As mentioned previously, B had a smaller accumulated cake mass than that of G. However, B exhibited a lower flux than G as shown in Fig. 5. This can be explained by the average size of the small debris. The average sizes of B and G debris were only 1.91 and 7.30 μm, respectively (Table S2†). The smaller zeolite debris of B caused more severe clogging between the membrane fibers, corresponding to membrane pore blockage (Fig. 6c'). This resulted in a lower flux for B than for G.

Table 3 shows the average water quality of the raw water, water treated by zeolite adsorption, and the GDM permeate. The pH and turbidity were measured without prefiltration. However, analyses of parameters such as Ca concentration, TOC, and UV₂₅₄ were conducted after pre-filtration using 0.45 μm filters to remove zeolite and impurities. Raw water represents the quality of the lake water used in the experiment. Zeolite adsorption for each zeolite indicated the water quality of the adsorption-treated water, including the zeolite, immediately after a 3-hour zeolite adsorption. The GDM permeate represents the water quality of the permeate passing through the GDM module membrane after zeolite adsorption. Because the average membrane pore size was 0.3 μm, corresponding to microfiltration, most of the zeolites were filtered out. Additionally, through this microfiltration process, effective removal of most suspended solids, turbidity, and some organic compounds was achieved.^60–63 As shown in Table 3, the turbidity increased greatly when zeolites were injected into raw water but was less than 0.5 NTU after the GDM filtration. However, calcium removal did not change significantly during GDM filtration. Calcium removal was achieved only through zeolite adsorption pre-treatment, which is one of the reasons why we combined the zeolite adsorption process with GDM filtration. Consequently, through the combined process, high-concentration calcium ion treatment and purification of the surface water were effectively carried out simultaneously. Through this combined process, we simplified the calcium ion removal and water purification processes, such as traditional electrochemical treatment, coagulation, precipitation, and disinfection. The combined process system can operate without electricity, thereby preventing unnecessary energy consumption and waste of resources. This combined system and evaluation data can be applied not only to the removal of calcium but also to the elimination of other cationic contaminants, such as heavy metals, offering various potential applications. The concentrations of organics in the raw water, in terms of TOC and UV₂₅₄, did not change significantly during zeolite adsorption and GDM filtration. The pH of the raw water was 7.53. For most zeolites, a similar pH of approximately 7 was maintained after adsorption. However, in the case of P1, the pH increased significantly to 10.79 after the adsorption pretreatment. This is because the P1 zeolite has a strong cation adsorption capacity, adsorbing protons (H⁺) in water and increasing the pH.⁶⁴

Table 3 Average water quality of raw water, water treated by zeolite adsorption and GDM permeate (n = 3)

	Raw water	Zeolite type
		P1		P2		B		G
		Zeolite adsorption	GDM permeate	Zeolite adsorption	GDM permeate	Zeolite adsorption	GDM permeate	Zeolite adsorption	GDM permeate
a The turbidity was too high to measure using the turbidimeter in this study (its measuring range is 0–4000 NTU).
Ca concentration [mg L^−1]	215	0	0	134	136	36	32	167	173
Turbidity [NTU]	1.82	—^a	0.31	—^a	0.15	39.75	0.15	152.50	0.13
TOC [mg L^−1]	2.64	2.96	2.94	2.74	2.78	2.70	2.72	2.41	2.52
UV254 [cm^−1]	0.027	0.027	0.028	0.020	0.021	0.025	0.025	0.023	0.024
pH	7.53	10.79	10.44	7.84	7.57	7.86	7.87	7.48	7.47

---

The cake layers were analyzed using SEM-EDS. Fig. S4† shows the surface elemental analysis results of the fresh zeolites before the experiments and Fig. S5† shows the SEM-EDS results for each zeolite cake layer that accumulated on the membrane surface after adsorption and GDM filtration. In all the zeolite cake layers, the elemental calcium content increased because of calcium adsorption. The calcium mass percentages of each zeolite before and after adsorption, as analyzed using SEM-EDS, are listed in Table S4.†

3.4. Comprehensive evaluation of zeolites in a zeolite-GDM system for calcium removal

Table 4 shows a comprehensive relative evaluation of zeolites for removing calcium from water together with GDM filtration. The zeolites were evaluated based on three criteria: Ca removal, GDM water flux and price. Grades were assigned as ‘Very Good’, ‘Good,’ ‘Fair,’ or ‘Poor’. The parameters and criteria for assigning grades are listed in Table S5.† First, the calcium removal was almost 100% for P1 and 92.6% for B at 360 min, indicating the excellent performance of Ca²⁺ using zeolites with low Si/Al ratios. Therefore, the grades of P1 and B were very good and good, respectively. In contrast, P2 and B exhibited low removal rates of 43.7% and 29.8%, respectively, after 360 min (Fig. 4). Regarding GDM water flux, B and G had higher flux values than P1 and P2 and had a good effect on permeate production. In terms of cost, P2 and G, which are mainly composed of natural zeolites, were cheap and, thus, highly valued. The price of each zeolite is listed in Table S6.† Because the prices of the synthetic (P1 and B) and natural (P2 and G) zeolites differ by up to 40 times, poor (expensive) and very good grades were assigned, respectively. Zeolites with large sizes and low Si/Al ratios, such as synthetic zeolite B, are suitable for Ca removal by adsorption combined with GDM filtration. However, for practical applications, the cost should be considered because synthetic zeolites are relatively expensive. For instance, if a moderate Ca removal efficiency is acceptable for an application, it is worth considering a cheap and large zeolite, such as zeolite G, as a Ca adsorbent before GDM filtration because of its low price and high GDM water flux. In this manner, an evaluation such as that presented in Table 4 will be useful for selecting the appropriate zeolites.

Table 4 Comprehensive relative evaluation of zeolites for Ca removal in water

Criteria	Zeolite type
	P1	P2	B	G
Remark: “−”, “±”, “+”, and “++” represent “Poor”, “Fair”, “Good”, and “Very Good”, respectively.
Ca removal	++	±	+	−
GDM water flux	±	−	+	++
Price	−	++	−	++

---

4. Conclusions

Calcium removal from water was performed via zeolite adsorption combined with GDM filtration using four types of zeolites. XRF analysis showed that P1 and B zeolites had high Al contents and low Si/Al ratios, whereas P2 and G zeolites had low Al contents and high Si/Al ratios, in accordance with the results of the XRD and FTIR analyses. The low Si/Al ratios of P1 and B led to higher Ca removal. Although P1 and B had similar Si/Al ratios, P1 had a higher calcium removal rate from the beginning, whereas B had a lower removal rate initially because P1 was much smaller than B. During GDM filtration after zeolite adsorption, the permeate flux of water treated with the large zeolites, B and G, was higher than that of water treated with the small zeolites, P1 and P2. When zeolites were evaluated in terms of calcium removal, GDM water flux, and price, they had different advantages and disadvantages depending on the criteria. This evaluation is useful for selecting zeolites suitable for Ca removal via zeolite adsorption combined with GDM filtration. This combined system and evaluation data can be applied not only to the removal of calcium but also to the elimination of other cationic contaminants including heavy metals, offering various potential applications. Because of insufficient research on the impact of zeolite characteristics on membrane fouling, our study is significant. From this perspective, our study provides new insights into membrane fouling and process flux related to zeolite adsorption adsorbents. The combined process system can operate without electricity, thereby preventing unnecessary energy consumption and resource waste, thus representing a sustainable technology. Consequently, this combined process has the potential to serve as a benchmark for enhancing water treatment efficiency.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study was supported by the Korea Ministry of Environment (MOE) and Korea Environmental Industry & Technology Institute (KEITI) as “Prospective green technology innovation project” [2020003160019] and by “Regional Innovation Strategy (RIS) project” through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (MOE) [2022RIS-005].

References

1. A. H. Mahvi, F. Shafiee and K. Naddafi, Int. J. Environ. Sci. Technol., 2005, 1, 301–304.
2. N. J. McNally, H. C. Williams, D. R. Phillips, M. Smallman-Raynor, S. Lewis, A. Venn and J. Britton, Lancet, 1998, 352, 527–531.
3. Y. Miyake, T. Yokoyama, A. Yura, M. Iki and T. Shimizu, Environ. Res., 2004, 94, 33–37.
4. S. G. Danby, K. Brown, A. M. Wigley, J. Chittock, P. K. Pyae, C. Flohr and M. J. Cork, J. Invest. Dermatol., 2018, 138, 68–77.
5. G. Gasco and A. Mendez, Desalination, 2005, 182, 333–338.
6. S. A. Critter and C. Airoldi, Geoderma, 2003, 111, 57–74.
7. D. Muraviev, J. Noguerol and M. Valiente, React. Funct. Polym., 1996, 28, 111–126.
8. S. Wang and Y. Peng, Chem. Eng. J., 2010, 156, 11–24.
9. M. N. Sepehr, M. Zarrabi, H. Kazemian, A. Amrane, K. Yaghmaian and H. R. Ghaffari, Appl. Surf. Sci., 2013, 274, 295–305.
10. Y. Hailu, E. Tilahun, A. Brhane, H. Resky and O. Sahu, Groundw. Sustain. Dev., 2019, 8, 457–467.
11. O. Karnitz, L. V. A. Gurgel and L. F. Gil, Carbohydr. Polym., 2010, 79, 184–191.
12. S. Narayanan, P. Tamizhdurai, V. Mangesh, C. Ragupathi and A. Ramesh, RSC Adv., 2021, 11, 250–267.
13. C. Qin, R. Wang and W. Ma, Desalination, 2010, 259, 156–160.
14. C. J. Rhodes, Annu. Rep. Prog. Chem., Sect. C: Phys. Chem., 2010, 106, 36–76.
15. F. M. Higgins, N. H. de Leeuw and S. C. Parker, J. Mater. Chem., 2002, 12, 124–131.
16. M. Trgo, J. Peric and N. V. Medvidovic, J. Environ. Manage., 2006, 79, 298–304.
17. E. Glover, A. Faanur and J. Fianko, West Afr. J. Appl. Ecol., 2010, 16, 95–105.
18. D. Breck, W. Eversole and R. Milton, J. Am. Chem. Soc., 1956, 78, 2338–2339.
19. T. Henmi, Soil Sci. Plant Nutr., 1987, 33, 517–521.
20. H. Ishimoto, T. Origuchi and M. Yasuda, J. Mater. Civ. Eng., 2000, 12, 310–313.
21. M. Al-Anber and Z. A. Al-Anber, Desalination, 2008, 225, 70–81.
22. V. Van Speybroeck, K. Hemelsoet, L. Joos, M. Waroquier, R. G. Bell and C. R. Catlow, Chem. Soc. Rev., 2015, 44, 7044–7111.
23. S. Babel and T. A. Kurniawan, J. Hazard. Mater., 2003, 97, 219–243.
24. E. Martinot, A. Chaurey, D. Lew, J. R. Moreira and N. Wamukonya, Annu. Rev. Energy Environ., 2002, 27, 309–348.
25. W. Schier, J. Romaker, H. Exler and F. B. Frechen, Water Supply, 2011, 11, 39–44.
26. M. Boulestreau, E. Hoa, M. Peter-Verbanets, W. Pronk, R. Rajagopaul and B. Lesjean, Desalin. Water Treat., 2012, 42, 125–130.
27. A. Ding, J. Wang, D. Lin, X. Cheng, H. Wang, L. Bai, N. Ren, G. Li and H. Liang, Environ. Sci.: Water Res. Technol., 2018, 4, 48–57.
28. R. Rohani, I. I. Yusoff, N. Khairul Zaman, A. Mahmood Ali, N. A. B. Rusli, R. Tajau and S. A. Basiron, Sep. Purif. Technol., 2021, 270, 118757.
29. S. Mulyati and N. Arahman, IOP Conf. Ser.: Mater. Sci. Eng., 2017, 180, 012128.
30. C. Blöcher, J. Dorda, V. Mavrov, H. Chmiel, N. Lazaridis and K. Matis, Water Res., 2003, 37, 4018–4026.
31. M. R. Adam, N. M. Salleh, M. H. D. Othman, T. Matsuura, M. H. Ali, M. H. Puteh, A. F. Ismail, M. A. Rahman and J. Jaafar, J. Environ. Manage., 2018, 224, 252–262.
32. Y. Yurekli, J. Hazard. Mater., 2016, 309, 53–64.
33. M. M. Motsa, T. A. M. Msagati, J. M. Thwala and B. B. Mamba, Desalin. Water Treat., 2015, 53, 2604–2612.
34. N. Tijani, H. Ahlafi, M. Smaihi and A. El Mansouri, Mediterr. J. Chem., 2013, 2, 484–492.
35. D. S. Dlamini, J. M. Tesha, G. D. Vilakati, B. B. Mamba, A. K. Mishra, J. M. Thwala and J. Li, J. Cleaner Prod., 2020, 277, 123497.
36. C. Jain and D. Ram, Water Res., 1997, 31, 154–162.
37. G. Annadurai, M. Chellapandian and M. Krishnan, Environ. Monit. Assess., 1999, 59, 111–119.
38. C. Jain, D. Singhal and M. Sharma, J. Hazard. Mater., 2004, 114, 231–239.
39. G. Pearce, Filtr. Sep., 2007, 44, 30–32.
40. E. Pérez-Botella, S. Valencia and F. Rey, Chem. Rev., 2022, 122, 17647–17695.
41. S. Kulprathipanja, Zeolites in Industrial Separation and Catalysis, John Wiley & Sons, 2010.
42. M. Akgül and A. Karabakan, Microporous Mesoporous Mater., 2011, 145, 157–164.
43. H.-R. Wenk and A. Bulakh, Minerals: Their Constitution and Origin, Cambridge University Press, 2016.
44. M. M. J. Treacy and J. B. Higgins, Collection of Simulated XRD Powder Patterns for Zeolites, Elsevier, 4th edn, 2001.
45. A. Julbe and M. Drobek, in Encyclopedia of Membranes, ed. E. Drioli and L. Giorno, Springer Berlin Heidelberg, Berlin, Heidelberg, 2016, pp. 2055–2056,  DOI:10.1007/978-3-662-44324-8_604.
46. P. Jacobs, E. M. Flanigen, J. Jansen and H. van Bekkum, Introduction to Zeolite Science and Practice, Elsevier, 2001.
47. G. Blanchard, M. Maunaye and G. Martin, Water Res., 1984, 18, 1501–1507.
48. A. Cincotti, N. Lai, R. Orrù and G. Cao, Chem. Eng. J., 2001, 84, 275–282.
49. M. M. J. Treacy and J. B. Higgins, Collection of Simulated XRD Powder Patterns for Zeolites, Elsevier, 5th edn, 2007.
50. P. Ambrozova, J. Kynicky, T. Urubek and V. D. Nguyen, Molecules, 2017, 22, 1107.
51. D. W. Breck and D. W. Breck, Zeolite Molecular Sieves: Structure, Chemistry, and Use, John Wiley & Sons, 1973.
52. W. Mozgawa, J. Mol. Struct., 2001, 596, 129–137.
53. K. Elaiopoulos, T. Perraki and E. Grigoropoulou, Microporous Mesoporous Mater., 2010, 134, 29–43.
54. E. M. Flanigen, H. Khatami and H. A. Szymanski, Infrared Structural Studies of Zeolite Frameworks, in Molecular Sieve Zeolites-I, ed. E. M. Flanigen and L. B. Sand, American Chemical Society, 1974.
55. W. D. Nesse, Introduction to Mineralogy, Oxford Univ. Press, 2012.
56. S. Rayalu, J. Udhoji, S. Meshram, R. Naidu and S. Devotta, Curr. Sci., 2005, 2147–2151.
57. G. Li, FT-IR Studies of Zeolite Materials: Characterization and Environmental Applications, The University of Iowa, 2005.
58. S. Shirazi, C.-J. Lin and D. Chen, Desalination, 2010, 250, 236–248.
59. A. McCarthy, P. Walsh and G. Foley, J. Membr. Sci., 2002, 201, 31–45.
60. S. Al Aani, T. N. Mustafa and N. Hilal, J. Water Process Eng., 2020, 35, 101241.
61. H. Guo, Y. Wyart, J. Perot, F. Nauleau and P. Moulin, Water Res., 2010, 44, 41–57.
62. S. Mallakpour, E. Azadi and C. M. Hussain, New J. Chem., 2023, 47, 17–40.
63. S. Hong, F. Miller and J. Taylor, Water Sci. Technol.: Water Supply, 2001, 1, 43–48.
64. C. Qin, R. Wang and W. Ma, Chem. Eng. J., 2010, 156, 540–545.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ew00438h

---