Application of Calcium Carbonate in the Pharmaceutical Removal Process

Abstract

One way to reduce the negative impact of human activity on the natural environment is to use natural, easily available and relatively cheap to produce compounds in industrial processes. One such compound is naturally occurring calcium carbonate (CaCO₃). This compound has adsorption properties so that it can be an alternative to commonly used adsorbents. The aim of this work is to determine the possibility of using CaCO₃ to remove pharmaceutical substances such as sulfadiazine and tetracycline from water. The CaCO₃ used in this work was synthesised using our own method, which allows the production of CaCO₃ particles with nanometric size. In the conducted research, calcium carbonate was used in the form of a suspension in purified solutions and as an inorganic filling of the developed membranes. The mass of pharmaceutical substances removed from their aqueous solutions was determined in the tests carried out. Based on the results obtained, it can be concluded that CaCO₃ has the ability to adsorb both tetracycline and sulfadiazine. In suspension tests, the mass of the removed substances per unit mass of adsorbent was 1.52 mg/g and 6.85 mg/g, respectively. In turn, in the case of the integrated process using the developed membranes, the mass of the removed substances per unit mass of adsorbent was 109 mg/g and 97 mg/g.

1. Introduction

Due to economic growth, increasing population, urbanisation, industrialisation, and agricultural production, some new toxic substances are detected in the water environment. A significant group of new chemical compounds is included in the group termed emerging contaminants (ECs). These substances are characterised by relatively low concentrations in water, typically in concentrations ranging from ng·L⁻¹ to μg·L⁻¹, and in many cases, their presence in the water environment is generally not monitored. However, they are recognised as significant and dangerous water pollutants due to their long-term stability in water and tendency to accumulate in the environment [1].

One of the types of EC are pharmacological substances belonging to a family of compounds called, in short, PPCPs (pharmaceuticals and personal care products). Most of these substances are toxic to organisms and increase the phenomenon of microorganisms’ drug resistance. Although pharmaceuticals are in low concentrations, their bioaccumulating capacity makes them unsafe for organisms [2,3]. Moreover, the studies show that the transformation products’ metabolites and derivatives are more toxic than the primary forms of pharmaceuticals [4].

Another aspect related to the presence of pharmaceuticals in the water environment is the phenomenon of antimicrobial resistance. It occurs when bacteria, fungi or other microorganisms no longer respond to medicines. Due to that, the medicines become ineffective and treating infections is harder and could lead to death at times [5,6,7]. One of the top 10 global public health threats facing humanity by WHO is antimicrobial resistance.

The source of pharmaceuticals in water might be domestic, urban, hospital, and industrial wastewater, as well as effluents from sewage treatment plants (STPs), aquaculture, and intensive livestock farming. However, it should also be noted that except for a few different sources of pharmaceuticals, the problem also refers to many different substances with other chemical properties detected in various types of water bodies [8,9,10,11]. For these reasons, using a single, efficient method to remove pharmaceutical substances from water is complicated, and it is necessary to develop new treatment methods or improve the efficiency of existing technologies.

Several treatment processes for removing PPCPs from water are known, such as membrane techniques, advanced oxidation processes, and biological methods [12,13,14]. Adsorption is one of the well-known and usually cheap methods for removing pollutants from water. During the process, substances bind to the solid’s surface from water or gas. The most significant advantages of this type of purification are the lack of harmful products and the independence of toxic substances contained in the cleaned stream [12,13,14,15]. Many materials are used as adsorbents, such as activated carbon, carbon nanotubes, and graphite. However, despite their high efficiency, the relatively high cost of sorbent and operation (incl. the need for regeneration or bed replacement after a number of cycles) is a serious disadvantage [12,16,17,18,19,20]. Due to that, natural materials have been evaluated into non-conventional adsorbents such as biochar, zeolites, chitosan, coconut shells, metal-organic frameworks (MOFs), and others [20,21,22]. They are developing not only because they are cheaper but also more environmentally friendly and available in large amounts from industries or agriculture.

One of the non-conventional adsorbents is the calcium carbonate (CaCO₃). This substance is a natural mineral which is known for its good thermal stability and low cost [23,24,25]. In the literature, many studies consider the molecular level of calcium carbonate precipitation, aggregation, and formation [26,27,28]. It is related to the increasing use of CaCO₃ nanoparticle aggregates in industry, e.g., as an adsorbent for cationic dye molecules [29], for pharmaceutical substances [30], or in biomedical applications [24,31,32].

This work aimed to determine the adsorption potential of CaCO₃ for PPCPs, which was produced using our own method. Additionally, the adsorption tests using the CaCO₃ as a sorbent were carried out in different systems, where CaCO₃ particles were suspended in pharmaceutical solutions or where CaCO₃ particles were an inorganic filler in the produced non-homogeneous polymeric membranes.

Additionally, the antibacterial properties of CaCO₃ particles and the membranes modified by CaCO₃ were determined. Antibacterial properties of a membrane, bacteriostatic or bactericidal, provide additional advantages in water filtration processes. Various microorganisms are ubiquitous in any water system, and their presence can cause severe problems to the process equipment, such as clogging catalytic or sorption beds, blocking filtration equipment, negatively affecting the operation of pumps, causing microbial corrosion, etc. The microorganisms spontaneously deposit on various surfaces, where they attach and proliferate. Secretion of extracellular polymeric substances (EPS) allows them to adhere and form a biofilm, which creates a serious problem for any filtration system. Usually, a drastic reduction of the system capacity is observed as a result of so-called “biofouling”. Polymeric surfaces, including membranes, are particularly susceptible to settling by various microorganisms and biofilm formation. This is why, preventing this detrimental effect is critical to the long-term reliable operation of membrane systems applied for water treatment. The incorporation of particles, which have proven antibacterial properties, is a reliable and cost-effective way to reduce the rate of biofouling, thus minimising the equipment downtime for membrane cleaning or replacement.

An innovation in our research is the proposal of an integrated adsorption-filtration system, which, using a single apparatus, will be able to address the growing problem of the presence of PPPCs in water. Another innovation is the utilisation of CaCO₃ as the adsorbent bound within the membrane structure.

2. Materials and Methods

2.1. Materials

To prepare pharmaceutical solutions for experiments, sulfadiazine (Sigma Aldrich, Darmstadt, Germany), tetracycline (Pol-Aura, Olsztyn, Poland) and ultrapure water (own reverse osmosis laboratory installation) were used. The solution concentration was 40 mg/dm³. Additionally, NaCl (Pol-Aura, Olsztyn, Poland) and SiO₂ (Sigma Aldrich, St. Louis, MO, USA) were used to prepare test solutions.

Calcium carbonate, CaCO₃, was produced by the original method, which is presented in the next section. Calcium hydroxide (Chempur, Piekary Slaskie, Poland), carbon dioxide (99.99%) (Multax, Stare Babice, Poland) and ultrapure water were used to produce calcium carbonate.

Polysulfone (PS) polymer (Sigma Aldrich, Finland Oy, Uusimaa, Finland) was used to produce the membranes. N-methyl pyrrolidone (NMP) (Sigma Aldrich, Poznan, Poland) (ACS reagent, ≥99.0%) was used as the solvent. The membranes were produced by the wet-phase inversion process [33]. All important parameters related to the membrane process production were determined during our own research.

Additionally, commercial flat sheet ultrafiltration polyvinylidene fluoride (PVDF) membranes (Microdyne Nadir, Wiesbaden, Germany) were used during experiments.

2.2. CaCO₃ Production Method

Calcium carbonate has been produced in the laboratory using the method previously developed by our team members [26,34]. The installation with a rotated disc is shown in Figure 1. In the first step, a calcium hydroxide solution, which was oversaturated and mixed for 48 h, was prepared. Then, the solution was filtered using a 0.1 μm filter. Both processes were performed under atmospheric pressure and at room temperature (25 °C). The temperature and pH of the calcium hydroxide solution were controlled during the production process. The initial pH of the solution was between 10.80–11.00. The production starts when the gas CO₂ flows and contacts with the solution. Gas flow was 2 dm³/min all the time. The process was stopped when the pH decreased to 7.00. It means that the solution became neutralised.

The solution of calcium carbonate from the reactor was also filtered on a vacuum set. The crystals were poured with ultrapure water, dried in an oven at 80 °C until the weight was stable, and finally stored in a desiccator.

2.3. Method of Membrane Preparation

The membranes were prepared by the wet-phase inversion method. The membrane-forming solution was prepared by dissolving PS in NMP with a concentration of 11% (w/w). Our own earlier research determined the concentration of the polymer [35]. Using a commercially available casting knife, a film of the appropriate thickness was formed. In turn, for the preparation of membranes with calcium carbonate particles as a filler, the CaCO₃ was added to the polymer solution in 10% or 20% concentration relative to the weight of the polymer. The suspension was stirred and placed in an ultrasonic bath before forming the layer with the casting knife. The temperature of the membrane-forming solution or suspension was 25 °C.

2.4. Testing of Materials Structure

To determine the surface area and pore size distribution of CaCO₃ particles, sorptometric analysis was used. The tests were conducted with the Sorptomat ASAP 2405 (Micrometrics Inc., Norcross, GA, USA). The parameters were determined using Brunauer–Emmet–Teller (BET) adsorption isotherms and the Barrett–Joyner–Halenda (BJH) method, respectively.

The cross-section of the membranes was examined by PhenomPro scanning electron microscopy (SEM) (PhenomWorld, Eindhoven, the Netherlands). Before tests, all the membrane samples were frozen in liquid nitrogen and broken.

Additionally, Fourier Transform Infrared Spectroscopy (FTIR) with the attenuated total reflectance (ATR) module was used to characterise the membrane and CaCO₃ particle surface. The tests were conducted with the Nicolet iS10 (Thermo Scientific, Waltham, MA, USA) spectrometer. The sample was scanned 32 times within the wave numbers 4000—400.

2.5. Testing of CaCO₃ Adsorption Properties

Testing of CaCO₃ adsorption properties was conducted in three different systems. In the first system—the adsorption stationary system (Figure 2), CaCO₃ particles acted as adsorbent. In this study, three distinct masses of calcium carbonate were prepared in samples and analysed in a stationary system. The pharmaceutical solutions were added to the adsorbents. Three concentrations of suspensions were considered—0.05%, 0.1% and 0.5% (weight of calcium carbonate/weight of suspension). After the adsorption process, the suspension had to be clarified before quantitative pharmaceutical analysis. In this system, centrifugation was used to separate CaCO₃ particles from the suspensions. The centrifugation time was included in the adsorption time because the particles remained in contact with the removed substance. High-speed centrifugation does not affect the adsorption process. The binding forces of pharmaceutical molecules to the adsorbent are mainly van der Waals forces [36], and their value is typically expressed in piconewtons [37]. In contrast, the inertial forces acting on an individual pharmaceutical molecule during sample centrifugation are five orders of magnitude smaller. Therefore, centrifugation cannot cause desorption.

In the second system—the adsorption flow system the CaCO₃ particles played the role of adsorption bed in the flow system (Figure 3). In this system, a suspension of CaCO₃ particles and pharmaceutical solution circulated during the test. The PVDF membrane was used to separate CaCO₃ particles from the solutions.

In the third system, two processes were combined into one integrated process. Therefore, the filtration-adsorption process was analysed (Figure 4). In this system, CaCO₃ particles were used as fillers in the produced membranes. Additionally, in this system, the impact of the presence of salt (NaCl) and solid particles (SiO₂ particles) on the filtration-adsorption process was tested.

In both flow systems, the standard membrane filtration process was performed. A laboratory filtration plant (Figure 5) was used to carry out the membrane filtration process. The research used a typical module with three ports to conduct the membrane filtration process in a crossflow system. The membrane filtration area was 0.014 m².

The effectiveness of pharmaceutical removal in all used systems was determined based on changes in the concentration of tested substances in water during the processes. In the case test with suspension, the mass of removed pharmaceutical was given per unit mass of CaCO₃ m_CaCO3 [g]. In turn, in the case of an adsorption-filtration system, the mass was given per unit surface of membranes A_mem [m²]. However, knowing the volume of the membrane and the filling particle content, it is possible to estimate the mass of removed pharmaceuticals per unit mass of CaCO₃.

The tested substance concentration was determined by using UV-vis spectroscopy on the Genesys 10S UV-Vis device (ThermoFisher Scientific, Waltham, MA, USA). The absorption peak for tetracycline and sulfadiazine was observed at wavelengths of 360 nm and 288 nm. Additionally, turbidity measurement was carried out to determine the efficiency of retaining calcium carbonate by the membrane in the case of a flow system and to determine the efficiency of suspension separation in a stationary system. An MI 415 turbidimeter (Milwaukee, Rocky Mount, NC, USA) was used.

These three systems were selected by the authors. The aim was to remove pharmaceuticals from water by a novel membrane with filling CaCO₃ during filtration (system three). It is an integrated adsorption process on calcium carbonate and membrane filtration. Due to that fact, both processes had to be analysed and considered. The first system, where CaCO₃ suspension was analysed, allowed us to define the capacity of those particles to remove tetracycline and sulfadiazine. The second system was a combination of both integrated processes. A filtration influence on removing substances from water was analysed by comparing results from the first two systems. Moreover, comparison results from all systems showed the effectiveness of the integration of processes.

2.6. Testing of CaCO₃ Antibacterial Properties

The evaluation of antibacterial properties of CaCO₃ particles and prepared membranes was carried out for two strains of bacteria commonly encountered in the aquatic environment: Escherichia coli (Gram-negative) and Staphylococcus aureus (Gram-positive). The procedure based on the ASTM E2149-20 “Standard Test Method for Determining the Antimicrobial Agents Under Dynamic Contact Conditions” was used in the tests [38]. Each test was repeated three times, and the results were averaged.

The percentage reduction in the number of bacteria RP was determined based on Equation (1)

$$RP=\frac{B-A}{B}·100\%$$

(1)

where A is the concentration of living cells after 1 h of contact with the sample [CFU/mL], and B is the concentration after 1 h in the control sample [CFU/mL].

The initial concentration of bacteria was 4.42 × 10⁴ CFU/mL for E.coli and 4.55 × 10⁵ CFU/mL for S. aureus.

3. Results and Discussion

3.1. Analysis of the Materials’ Structure

Based on the BET and BJH analysis, it can be concluded that the specific surface area of produced CaCO₃ particles is 13.48 m²/g, and the average pore diameter is 27.5 nm. The obtained value of surface area is within the range of values from literature data. According to the literature, depending on the substrate concentration, temperature or pH, the specific surface area of CaCO₃ particles ranges from about 3 m²/g to even 250 m²/g [39,40,41]. Additionally, based on the results of XRD analysis and particle size determination, which have been carried out by one of the authors, it can be concluded that the produced CaCO₃ in the disk reactor occurs only in the form of calcite, and the single crystallite size is lower than 50 nm [42].

Figure 6 shows the cross-sections through the membrane without and with the 20% filling addition. The calcium carbonate particles can be observed both on the outside surface of the membrane and in the pores. Moreover, it should be emphasized that adding the filling does not change the membrane structure visibly. Additionally, it should also be noticed that both the unmodified membrane and the membrane with CaCO₃ particles are characterized by a classic amorphous structure with finger-like pores linked by sponge walls.

The presence of CaCO₃ particles on the membrane surface can also be confirmed by FTIR-ATR analysis (Figure 7). Based on the result, for the membrane with CaCO₃ (PS+CaCO₃, PS+CaCO₃+tetr, PS+CaCO₃+sulf) can be observed a new peak compared to the unmodified membrane. This peak occurs with wave number 1411 cm⁻¹ corresponds to the asymmetric CO₃ band. In turn, the other characteristic peaks for the calcite form of CaCO₃, which occurs with wave number 870 cm⁻¹, corresponding to the asymmetric CO₃ band (out-of-plane vibration) and 717 cm⁻¹ (in-plane vibration) [43,44], overlap the peak typical of groups occurring in the PS structure. Additionally, for the membranes after the filtration process (PS+CaCO₃+tetr, PS+CaCO₃+sulf) no new peaks were noticed. It might be the result of a very small amount of pharmaceutical substance which has been collected on the membrane surface or the overlap of characteristic peaks for pharmaceutical substance with peaks that are typical of groups occurring in the PS structure [45].

3.2. Mechanisms of Pharmaceutical Substance Adsorption on CaCO₃ Particles and Membrane Materials

To better understand the obtained results, it is necessary to determine some mechanism responsible for the adsorption of pharmaceutical substances on CaCO₃ particles and membrane materials, so PS and PVDF.

Based on the literature data [46,47,48] and our own test [49], it may be concluded that the electrostatic interactions are responsible for the adsorption of pharmaceutical substances on CaCO₃ particles. For this reason, the surface charge of the pharmaceutical particles and adsorbent material is a basic parameter that might help to explain the obtained results. The parameter used to determine surface charge is the zeta potential. In neutral pH (used ultrapure water was 6.5), the zeta potential on the surface of calcium carbonate is positive [50]. In turn, tetracycline and sulfadiazine are amphoteric molecules with two dissociation constants, pK_a = 3.30; 7.68 [51] and pK_a = 2.20; 6.5 [52], respectively. The effect of pH on the intensity of the adsorption process of pharmaceutical substances on solid materials is widely described in the literature [53,54,55,56]. It should also be noted that the literature underlines the effect of ionic strength, physicochemical properties of adsorbent particles, and process parameters on the intensity of the adsorption pharmaceutical substance process.

Regarding pharmaceutical substance absorption on the membrane materials, it may be concluded that despite electrostatic interaction, the chemical interactions can be responsible for the adsorption process. This is due to the presence of different organic functional groups in the structure of pharmaceutical substances and polymeric materials. The presence of these groups enables the formation of п-п interaction, H-bonding or dipole-dipole H-bonding between substances [57,58]. As a result, the sorption intensity may increase.

3.3. Adsorption Stationary System

The first research step focused on the determination of CaCO₃ sorption properties in the stationary system. Experiments lasted 90 min, and samples were analysed after 90 min from the beginning of the experiments. The final results calculated as the mass of pharmaceuticals removed per unit mass of calcium carbonate q_I [mg/g] (Equation (2)) and process efficiency η [%] (Equation (3)) are presented in

Table 1. Turbidity of suspension after the centrifugation process.

The concentration of calcium carbonate in suspension [%]		0.05	0.1	0.5
Turbidity [NTU]	Initial	>1000	>1000	>1000
	After all stages	2.97	3.31	4.16

.

$$q_I=\frac{c_0-c_t}{m_{CaCO3}}\cdot V_I$$

(2)

where c₀ [mg/dm³] is the initial concentration of the pharmaceutical substance, c_t [mg/dm³] is the concentration of the pharmaceutical substance during the process, m_CaCO3 is the mass of CaCO₃ [g]; in turn, V_I [dm³] is the volume of tested solution.

$$\eta =\frac{c_0-c_t}{c_0}$$

(3)

The clear liquid is necessary for UV-Vis spectroscopy analysis, where particles could interfere with light and change the result. Due to that, centrifugation was used to separate adsorbent particles from suspension before analyses. The centrifugation parameters should allow for obtaining the solution with the lowest possible turbidity in a short time, indicating a high level of CaCO₃ separation from the pharmaceutical solution. The chosen centrifugation method has three stages, each lasting 3 min with 4000 rpm spin speed. The obtained values of solution turbidity are presented in Table 1.

The amount of CaCO₃ particles corresponding to the obtained solution does not negatively affect the measurements by UV-vis spectroscopy.

Based on the results (

Table 2. The process efficiency and the adsorbed mass of pharmaceuticals by using calcium carbonate for different suspension concentrations.

Substance	Suspension Concentration [%]	Efficiency, η [%]	Adsorbed Mass, qI [mg/g]
Tetracycline	0.05 0.10 0.50	1.07 ± 0.07 4.03 ± 0.03 10.01 ± 0.07	0.81 ± 0.01 1.52 ± 0.04 0.75 ± 0.04
Sulfadiazine	0.05 0.10 0.50	19.87 ± 0.21 17.66 ± 0.17 21.92 ± 0.19	15.40 ± 4.27 6.85 ± 1.96 1.69 ± 0.44

), it can be noticed that the adsorbed mass of sulfadiazine increases with a more significant amount of sorbent. In the case of tetracycline, removal efficiency at the end of the experiment for 0.05%, 0.1% and 0.5% are, respectively, 1.07%, 4.03% and 10.01%. In the case of sulfadiazine, this parameter reached 19.87%, 17.66%, and 21.91%, respectively. However, the process efficiency for sulfadiazine does not increase significantly. It can result from the particle agglomeration phenomenon. The particle agglomeration decreases mass transfer surfaces, and as a result, the intensity of the adsorption process decreases. The adsorbed mass of tetracycline varies without any trend. The results obtained for measurements for the suspension with the smallest amount of calcium carbonate are questionable. However, no reason for this result was found during testing.

Additionally, based on the results (Table 2), the difference in the adsorbed mass of tetracycline and sulfadiazine might be the result of the differences in the physicochemical properties of the tested substances. In the case of solid particles without chemical groups on their surface, such as CaCO₃, physical interactions are responsible for the adsorption of substances on their surface. The differences in the amount of adsorbed mass of tetracycline and sulfadiazine can be explained by electrostatic interactions between the tested substances and calcium carbonate. As mentioned, in neutral pH (used ultrapure water was 6.5), the zeta potential on the surface of calcium carbonate is positive. In turn, tetracycline and sulfadiazine are characterised by more than one dissociation constant, which depends on the value of pH. However, for pH = 6.5, the percentage of anionic species to neutral or cationic species is higher for sulfadiazine than tetracycline. As a result, the mass of adsorbed sulfadiazine is higher than tetracycline in the system used.

3.4. Adsorption Flow System

The second research step focused on the determination of CaCO₃ sorption properties in the adsorption flow system. During this process, the effectiveness of pharmaceuticals results from filtration on an ultrafiltration membrane and sorption onto calcium carbonate. Additionally, the ultrafiltration process is responsible for removing CaCO₃ from the feed solution.

In the stationary system, the highest adsorption concerning the mass of the material was achieved for 0.10% concentration of CaCO₃ in the suspension with tetracycline and for the lowest concentration of CaCO₃ in the case of sulfadiazine. A concentration of 0.05% was chosen to study the integrated filtration and adsorption process due to the results of sulfadiazine described above and the problem of membrane blocking at higher calcium carbonate concentrations in the suspension. The experiments lasted 90 min, and samples were analysed 90 min after the beginning of the experiments. The final results calculated as the mass of pharmaceuticals removed per unit area of membrane q_II [mg/m²] (Equation (4)) are presented in

Table 3. The adsorbed mass of pharmaceuticals in an adsorption flow system with CaCO₃ and without CaCO₃.

Substance	Suspension Concentration [%]	Adsorbed Mass qII [mg/m2]
Tetracycline	0.00	3477 ± 56
	0.05	3742 ± 76
Sulfadiazine	0.00	494 ± 61
	0.05	954 ± 85

.

$$q_{II}=\frac{c_0-c_t}{A_{mem}}\cdot V_I$$

(4)

where c₀ [mg/dm³] is the initial concentration of a pharmaceutical substance, c_t [mg/dm³] is the concentration of a pharmaceutical substance during the process, A_mem is the membrane area [m²]; in turn, V_I [dm³] is the volume of tested solution. For the comparison, in Table 3 also, the results for the tests carried out without CaCO₃ are presented, as a blind probe for membrane.

Based on the result, it can be concluded that the presence of CaCO₃ increases the adsorbed mass in the flow system compared to the pure membrane. However, there is also a significant difference in the results between the adsorbed mass of the tested substances. The adsorbed mass of the substance indicates a lower affinity of sulfadiazine for membranes made of PVDF. This is probably due to weaker interactions between PVDF chains and sulfadiazine molecules than between PVDF chains and tetracycline molecules.

Based on the results presented in Table 3, the adsorbed mass of pharmaceuticals re-moved per unit mass of calcium carbonate can be calculated according to Equation (2). The mass of tetracycline removed by calcium carbonate is 3.25 mg/g, and the mass of sulfadiazine is 7.35 mg/g. In this case, the obtained results are analogous to the results which were obtained in the stationary system.

Additionally, it should also be noted that the PVDF ultrafiltration membrane makes it possible to remove CaCO₃ particles from the feed stream, which is confirmed by the results of permeate turbidity. The initial turbidity of each suspension was higher than 1000 NTU. The average value of permeate turbidity for different tests is about 1 NTU. However, the presence of CaCO₃ particles in the feed solution is responsible for the fouling phenomena on the membrane surface. As a result, the permeate volume stream insignificantly decreases during the process.

3.5. Filtration-Adsorption System

The last part of the research focused on the use of calcium carbonate for membrane modification to improve the adsorption properties of membranes. In order to determine the effect of CaCO₃ particles on membrane adsorption properties, membranes with 0% (labelled as mem_0), 10% (mem_10) and 20% (mem_20) concentrations of CaCO₃ were prepared. The content of particles was calculated in relation to the weight of the polymer. Experiments lasted 90 min, and samples were analysed after 90 min from the beginning of the experiments. The final results calculated as the mass of pharmaceuticals removed per unit area of the membrane, q_III [mg/m²], are presented in Figure 8. The mass of pharmaceutical removed was calculated according to Equation (5).

$$q_{III}=\frac{c_0-c_t}{A_{mem}}\cdot V_I$$

(5)

where c₀ [mg/dm³] means the initial concentration of pharmaceutical substance c_t [mg/dm³] means the concentration of pharmaceutical substance during the process, A_mem [m²] means the area of the membrane, in turn, V_I [dm³] means the volume of tested solution. The concentrations refer to the number of pharmaceuticals in the feed stream.

Based on the results (Figure 8), the presence of CaCO₃ improved membrane adsorption properties. The higher the CaCO₃ concentration, the higher the amount of pharmaceutical substance removed mass from the feed solution. The removed mass of pharmaceutical is about four times higher for the process where the membrane with 20% concentration of CaCO₃ was used compared to the unmodified membrane. However, with over 20% CaCO₃ concentration, the membrane’s mechanical properties dramatically decrease. Therefore, the results for membranes with higher CaCO₃ are not presented. Additionally, a higher CaCO₃ concentration increases the intensity of agglomeration phenomena, which makes it impossible to prepare a homogenous membrane casting solution, and as a result, the membrane structure is not uniform.

According to the results shown in Figure 8, the mass of removed pharmaceuticals per unit mass of calcium carbonate can be calculated according to Equation (6):

$$m_{CaCO3}=x\cdot d\cdot A_{mem}\cdot \delta$$

(6)

where x [%] means the concentration of calcium carbonate in the membrane, d [g/m³] means the density of the membrane, which is 0.258 g/cm³, A_mem [m²] means the area of the membrane is 0.014 m², in turn, δ [m] means the thickness of the membrane, it is 98 ± 6 µm.

The mass of tetracycline and sulfadiazine removed by unit mass of calcium carbonate in the membranes with 10% of particles are 43 mg/g and 15 mg/g, respectively. For the membranes with 20% calcium carbonate, the mass of pharmaceuticals removed per unit mass of the filling is 109 mg/g in the case of tetracycline and 97 mg/g in the case of sulfadiazine. It shows that the removed mass does not increase proportionally with calcium carbonate content in the membrane.

The obtained results show that CaCO₃ suspension and the particles as membrane filling also have the ability to remove pharmaceuticals. Mechanisms of adsorption were elaborated in Section 3.2, where the pH influence is one of the main limitation parameters of physical adsorption, which is considered in that case. This is due to the amphoteric character of analysed substances. The driving force behind the adsorption is the concentration difference (between the pharmaceuticals’ concentration in the aquatic solution and on the adsorbent’s surface); it is connected with the diffusion. However, in the filtration process, the pressure difference is the driving force. Therefore, for the removal of trace tetracyclines and sulfadiazine in water, the pressure-related force is the same as in the studies discussed. Based on this, it is assumed that the use of an innovative membrane filled with CaCO₃ enables the filtration of both micropollutants and trace amounts of pharmaceutical substances because the solution-membrane contact is forced by pressure. Practical applications of these membranes are found in all wastewater treatment plants because PPCPs can be present in wastewater from households, hospitals and industrial plants.

For the integrated process system, an analysis of adsorption kinetics was presented for a membrane containing 20% CaCO₃ (mem_20). Figure 9 and Figure 10 depict the changes in concentration over time for tetracycline and sulfadiazine, respectively.

The adsorption kinetics was fitted to the pseudo-zero-order model (PZO), pseudo-first-order model (PFO), pseudo-second-order model (PSO) and pseudo-third-order model (PTO). Linear equations of these models are as follows [17,59,60].

• PZO

$$q_t=k_{0}t$$

(7)

• PFO

$$ln\left(q_e-q_t\right)-\ln \left(q_e\right)=k_{1}t$$

(8)

• PSO

$$1/\left(q_e-q_t\right)-1/\left(q_e\right)=k_{2}t$$

(9)

• PTO

$$1/{\left(q_t\right)}^2-1/{\left(q_e\right)}^2=k_{3}t$$

(10)

where k₀ [min⁻¹] is the rate constant of PZO adsorption, k₁ [min⁻¹] is the rate constant of PFO adsorption, k₂ [min⁻¹] is the rate constant of PSO adsorption, k₃ [min⁻¹] is the rate constant of PTO adsorption, q_e, q_t [mg/g] are the amount of analysed pharmacological substances that adsorbed at equilibrium and at time t [min].

Using the method of linear regression, data presented in Figure 10 and Figure 11 were analysed, respectively. For the kinetics of tetracycline adsorption, the coefficient of determination R² = 0.835 was obtained for the Pseudo-Zero-Order (PZO) model, R² = 0.905 for the Pseudo-First-Order (PFO) model, R² = 0.972 for the Pseudo-Second-Order (PSO) model, and R² = 0.694 for the Pseudo-Third-Order (PTO) model. This means that the kinetics of tetracycline adsorption in the integrated process can be well approximated by the pseudo-second-order model. On the other hand, for the kinetics of sulfadiazine, the coefficient of determination R² = 0.902 was obtained for the PZO model, R² = 0.779 for the PFO model, R² = 0.767 for the PSO model, and R² = 0.724 for the PTO model. Such results indicate that this kinetics cannot be well approximated by any of the proposed models, with the pseudo-zero-order model describing this kinetics best among the analysed models.

It should also be noted that the modified membranes have higher permeability than the unmodified membrane, which is confirmed by the results of the water permeability [dm³/(m²·bar·h)] (Figure 11). The increase in the permeability of the modified membranes might be a result of an increase in membrane wettability, pore size, and porosity. The increase in membrane wettability is related to the presence of hydrophilic CaCO₃ particles in the membrane structure [61]. Additionally, the presence of hydrophilic particles in the membrane casting solution affects the process of membrane formation by the wet-phase method. The presence of hydrophilic particles facilitates the penetration of non-solvent (water) into the formed membrane-forming film. In addition, the presence of hydrophilic particles reduces the affinity of the solvent (NMP) to the polymer chains. Both the facilitated penetration of water and faster diffusion of NMP into the non-solvent volume accelerate the extraction process, resulting in increased porosity and pore diameter [33,62].

In the last part of the research, the influence of present ions (NaCl) or solid particles (SiO₂) on the efficiency of removing pharmaceutical substances from water was investigated. The obtained results for different compositions of the feed solutions are presented in Figure 12.

Based on the results (Figure 12), the presence of NaCl in pharmaceutical solutions has a statistically insignificant effect on tetracycline or sulfadiazine adsorption. However, the impact of salt on pharmaceutical adsorption might be completely different than in the presented results because it depends on the type of salt, its concentration, and the physicochemical properties of pharmaceutical substances [52,54,63,64,65].

In the case of SiO₂ presence in the pharmaceutical solutions it can be noticed an increase in the removed mass of antibiotics compared to the solution without SiO₂ (Figure 8). The increase in the removed mass of antibiotics is a result of adsorption on the SiO₂ surface [66,67], but it might also be the effect of fouling phenomena. A newly created layer on the membrane surface or blocked pore with SiO₂ particles can improve the filtration effect of pharmaceutical substances from water. However, due to the fouling phenomena, the permeate volume stream increases during the process. Additionally, the insignificant differences in percentage change in the adsorbed mass of tetracycline and sulfadiazine, an increase of about 40% in ratio to the solution without SiO₂, indicated that the presence of SiO₂ similarly affects the adsorption of both substances.

An interesting issue in the application of membranes possessing adsorptive properties is the repeated utilization of these membranes. However, this issue was not thoroughly investigated in the present study. It should be noted that membranes containing adsorptive materials in their structure can be used until the adsorbent reaches its sorption capacity. However, it must be remembered that the saturation of the adsorbent within the membrane occurs relatively quickly. Of course, this process is slower the lower the concentration of pharmaceutical compounds in the water. Reusing such a membrane would only be possible after regenerating the adsorbent. However, this is often not feasible [68]. This constitutes a significant limitation of this method for removing pharmaceutical compounds from water. There are methods for regenerating adsorbents, such as washing with steam or alcohol. While effective for the adsorbent, these methods are simultaneously harmful to the membranes. Desorption of pharmaceuticals into ethanol seems to be a promising method for regenerating membranes used in this study. Detailed results of desorption studies are not the subject of this work. It should also be realized that the use of ethanol may not be safe for all polymeric membranes.

3.6. Antibacterial Properties

The results of bacteriostatic tests are presented in Figure 13. It can be concluded that CaCO₃ particles (CaCO₃ Ps) exhibit pronounced antibacterial properties against both Gram-negative and Gram-positive bacteria, for which the reduction in bacteria population clearly exceeds 80% and 90%, respectively. Accordingly, the presence of CaCO₃ particles on the membrane surface translates to their antibacterial properties. For both types of bacteria, the modified membranes (mem_10) have clearly better antibacterial properties than the unmodified membrane (mem_0).

It should be noted that CaCO₃ particles have the best antibacterial properties compared to analysed materials. The antimicrobial effects of calcium carbonate against gram-negative bacteria are also confirmed in the literature [69]. In the case of the membrane with CaCO₃ particles, the polymer plays an important role in the growth of bacteria [39,70].

3.7. Adsorbents Comparison

Table 4. Comparison of adsorbents.

Adsorbent	Adsorption Capacity [mg/g]		References
	Tetracycline	Sulfadiazine
PS membrane with CaCO3	109	97	This work
Graphene oxide	313	-	[56]
Activated carbon	208	30	[71,72]
Palygorskite	29	-	[73]
Rectorite	100	-	[74]
Silicon dioxide	0.86	-	[75]
Zeolite	-	1.06	[76]
Graphite	-	13.2	[53]
Nano-carbon	125	620	[77]

presents a comparison of the sorption properties of various adsorbents and compares them with the results of this study. It can be noticed that the proposed non-integrated sorption and filtration system in this study (see Section 3.4) allows for achieving results better than most of the adsorbents analyzed in the literature and only slightly worse than the best ones described in the literature.

4. Conclusions

To sum up, it can be noticed that the CaCO₃ particles, which were produced by our method, have the ability to adsorb tetracycline and sulfadiazine from water solutions, which was confirmed by the test carried out in three different systems. Additionally, the produced CaCO₃ can be used to modify the developed membranes to improve their adsorption properties. However, the amount of adsorbed mass of pharmaceuticals depends on their physicochemical properties as well as the properties of membrane materials. It is related to the possible mechanism of adsorption of tested substances, which might be electrostatic interactions between pharmaceutical molecules and materials. This phenomenon is confirmed by the obtained results. According to them, the best adsorption properties are observed for the integrated adsorption-filtration process. The mass of tetracycline or sulfadiazine removed by unit mass of particles is 109 mg/g and 97 mg/g, respectively. It should also be noticed that the produced CaCO₃ particles have antibacterial properties against both Gram-negative and Gram-positive bacteria.