Experiment:
Different materials were used in the present study, including PES polymeric membrane as the primary polymer, polyvinylpyrrolidone (PVP) as an auxiliary polymer for porosity regulation, dimethylacetamide (DMA) as an organic solvent, SiO2 and TiO2 nanoparticles, and distilled water as anti-solvent (Table 1 presents the chemical structures of used materials). In addition, a UP400s-ultrasonic device (Sound Technology Development Ltd, Iran) was utilized in the coagulation technique for solution uniformization during the membrane synthesis. Furthermore, Shing Sing magnetic stirrer (Germany) with a variable round of 100–800 rpm was utilized for bubble removal from the solvent, besides a film applicator for filtrating the prepared solution.
Membrane preparation method
The weight ratios of 25% PES, 5% PVP, 1% SiO2, and 2% TiO2 were considered for membrane production. First, two separate containers were used to prepare the solutions, one containing additive and auxiliary polymer, and 20% of the required solvent was placed on a magnetic stirrer. The other containing nanoparticles and the remaining solvent were sonicated at the minimum temperature in the ultrasonic device. Stirring was continued for a long time following the preparation of a homogeneous solution and the addition of the contents of the first container, and then the bubbles were allowed to escape completely after the necessary period of time. Finally, a film applicator of a specific thickness was used to transfer the solution to the glass. Next, a coagulation bath was used for preparing the membrane that was then transferred to another distilled water container for solvent removal. In this phase, any quantity of PVP that can be dissolved in water is removed from the film by dissolving it in water. The scanning electron microscope (SEM) and atomic force microscope (AFM) (Figs. 1 and 2) imaging techniques were used to capture an image from the adsorbent surface structure for morphological analysis of the physical membrane surface. The number of adsorption-needed sites grows as the heterogeneous adsorbent surface is confirmed. The adsorption is also enhanced by the pores at the adsorbent surface. The synthetic membrane was characterized using the Fourier-transform infrared (FTIR) spectroscopy technique (Fig. 3).
The membrane hydrophilicity was determined by conducting a contact angle test (Table 2). The membrane surface-water contact angle decreases by adding nanoparticles to the polymer solution, indicating that hydrophilicity of the membrane containing nanoparticles increases in comparison with the plain polymeric membrane.
Table 2
The result of the contact angel test on the synthetic membrane.
|
membrane
|
Contact angel
|
|
PES
|
68
|
|
PES TiO2 – SiO2
|
57.4
|
Impact of various parameters on the cefotaxime adsorption
pH effect
Tests were conducted at the pH values of 5, 7, and typical ones for examining the impact of pH on the adsorption. The pH value was regulated using sodium hydroxide and 0.1 M hydrochloric acid. The first was to complete adsorption in three wastewater samples containing cefotaxime antibiotics. After that, samples were taken at various pH values, and the quantity of drugs left in the wastewater was recorded.
Impact of contact time and temperature on adsorption
The tests were repeated at three 60, 120, and 180 seconds contact times, three 21°C, 28°C, and 37°C temperatures, and the constant pH value of 5 to examine the impact of contact time and temperature on the adsorption rate by keeping the adsorbent constant from the prepared membrane containing nanoparticles. The solution containing the left drug was then collected, and a spectrophotometer was then used to read and record its quantity.
Impact of initial drug concentration in wastewater
Tests were conducted at 10, 25, and 50 ppm by keeping constant other parameters to examine the impact of the initial concentration of cefotaxime on the adsorption. The quantity of drugs left in the solution was then measured.
Determining the percent removal and adsorption capacity
The percent antibiotic elimination (R) and the adsorption capacity (qe) can be calculated as follows:
$$\text{R}=\frac{(\text{C}\text{o}-\text{C}\text{e})}{\text{C}\text{o}}\times 100$$
1
$$\text{q}\text{e}=\frac{(\text{C}\text{o}-\text{C}\text{e})}{\text{w}}\times \text{V}$$
2
where Co, Ce, V, w, R, qe denote the initial and final concentrations (mg/L) of antibiotic molecules in the solution, solution volume (L), solution mass (gr), removal efficiency (%), and adsorption capacity (mg/g), respectively. Finally, adsorption isotherms, corresponding associations, and the governing equations were evaluated. Then, for laboratory result confirmation, the Monte Carlo calculations were performed.
Data analysis
The pH effect on the cefotaxime adsorption using optimized synthetic membrane containing nanoparticles at the contact time of 120 min, the temperature of 28°C, and the membrane adsorption value of 2.5 mg is shown in Fig. 4. As shown in the figure, because the concentration of hydrogen ions is low at the pH value of 7, and hydrogen ions are absorbed rather than medicinal ions, the minimum adsorption rate is at this pH value. Adsorption efficiency reduces as the pH value increases, leading to achieving the minimum percent adsorption at the pH value of 7. The type and ionic state of the adsorbent causative agents significantly affect the pH value of antibiotic adsorption. Consequently, it can be concluded that the pH value affects the adsorption equilibrium; thus, the pH value of 5 was chosen as the best pH, with the maximum cefotaxime adsorption by a membrane containing nanoparticles. The quantity of hydrogen ions and the antibiotic ions-adsorbent bond reduces by increasing the pH value. This most is likely due to an increase in inadequate retention time for the synthesis of active hydroxyl radicals and adequate time for the hydroxyl radical reaction to the cefotaxime molecules.
The impact of contact time on the cefotaxime adsorption using optimized synthetic membrane containing nanoparticles at the pH value of 5, the temperature of 28°C, and the membrane adsorption value of 2.5 mg is shown in Fig. 5. As shown in the figure, the fast adsorption in the contact time of 120 min is due to the abundant surface sites for cefotaxime adsorption on synthetic membranes containing nanoparticles. When the external sites are saturated after 120 min, further time is required for the adsorption to occur at the internal active sites. Since practically, all internal and external sites are saturated after 180 min, and the adsorption process achieves a state of equilibrium, so the best time is 180 min for the maximum drug adsorption (100% adsorption).
The impact of temperature on the cefotaxime adsorption using optimized synthetic membrane containing nanoparticles at the pH value of 5, contact time of 180 min, and membrane adsorption value of 2.5 mg is shown in Fig. 6. As shown in the figure, the efficiency of cefotaxime adsorption increases at the temperature of 28°C. The cefotaxime adsorption was 95.1%, 96.2%, and 99% for all three actual hospital wastewater samples at the contact time of 180 min, respectively. Consequently, 28°C was found as the best temperature for cefotaxime removal from hospital wastewaters using synthetic membrane containing nanoparticles.
The initial concentration of study antibiotic is another key parameter in the adsorption system for the synthetic study of the adsorption and the effective and fast practical application of membranes. Figure 7 shows the impact of this parameter on the cefotaxime adsorption from the hospital wastewater using an optimized synthetic membrane containing nanoparticles at the pH value of 5, contact time of 180 min, temperature of 28°C, and membrane adsorption value of 2.5 mg. As shown in the figure, given the occupancy of sites in the membrane containing nanoparticles, the efficiency of adsorption rate reduces rapidly as the drug concentration in wastewater increases.
Adsorption isotherm
Adsorption isotherm representing the relationship between adsorbent concentration and adsorption capacity is the most significant parameter in the design of adsorption systems. The adsorption is described as a mass transfer defined by mathematical equations as a process of adsorption equilibrium and adsorption speed. Analysis of adsorption isotherm data is essential for developing the equations that reflect the achieved results used in the system design. Adsorption isotherms are useful quantitative expressions that tell how well an adsorbent can absorb a specific substance [21]. When both phases are in equilibrium, the equation of adsorption isotherm offers a relationship between the drug concentration in the solution and the quantity of drug absorbed at the solid phase [22]. This equilibrium analysis gives data, such as the final adsorbent capacity of a specific substance in a single-component system. Besides, the isotherm diagrams of the equilibrium data may be used to derive the isotherm constants required in the mathematical modeling of adsorption systems.
As reported in the related literature, although different adsorption isotherm models have been developed, only a few can be utilized to absorb drugs from hospital wastewaters. The adsorption mechanism can be used to describe the adsorption isotherm models [23]. The Langmuir, Freundlich, and Temkin isotherm models are the three most frequently models suitable for experimental data [24]. The linear equation of adsorption isotherm for the Langmuir isotherm model is given by (3) [25]:
$$\frac{1}{{\text{q}}_{\text{e}}}=\frac{1}{{\text{k}}_{\text{l}}\times {\text{q}}_{\text{m}\text{a}\text{x}}^{2}}\times \frac{1}{{\text{C}}_{\text{e}}}+\frac{1}{{\text{q}}_{\text{m}\text{a}\text{x}}}$$
3
where Ce, qe, qmax, and kL denote the ion equilibrium concentration (mg/L), the quantity of adsorbed ions in equilibrium per gram of absorbent, surface adsorption capacity (mg/g), and adsorption energy (L/g) of Langmuir constants, respectively. Also, RL, whose value denotes the mode and way of adsorption isotherm, is another significant and efficient parameter expressing the main features of the Langmuir equation. The adsorption is undesirable, irreversible, linear, and desirable for RL> 1, RL=0, RL=1, and 0 < RL<1, respectively [26]. The value of RL is given by:
$${\text{R}}_{\text{L}}=\frac{1}{1+{\text{K}}_{\text{L}}{\text{C}}_{\text{o}}}$$
4
where Co (mg/L) denotes the initial concentration of antibiotics in the aqueous solution.
The calculations of adsorption isotherm for cefotaxime were done at 21°C and 37°C because the temperature at which 100% adsorption occurred was 28°C. Based on the results, the maximum adsorption capacity at 0 < RL <1 and the temperatures of 21°C and 37°C is 901.091 mg/g and 1250 mg/g, respectively, indicating the linear and optimal adsorption of drug molecules using the adsorbent (Figs. 8 and 9).
The Freundlich isotherm model is another frequently utilized isotherm model known as an experimental model that may be used to describe how different adsorbents absorb organic and inorganic substances. The linear equation of adsorption isotherm for Freundlich isotherm model is given by:
$${\text{L}\text{n}\text{q}}_{\text{e}}=\text{L}\text{n} {\text{K}}_{\text{f}}+\frac{1}{\text{n}}\text{L}\text{n} {\text{C}}_{\text{e}}$$
5
where qe, Ce, Kf, and n denote the equilibrium adsorption capacity (mg/g), ion equilibrium concentration in solution (mg/L), and the Freundlich isotherm model's constants related to adsorption capacity and adsorption intensity, respectively.
The results from the cefotaxime adsorption by the adsorbent are represented in Figs. 8 and 9. In numerous investigations, the value of n has been reported to be between 1 (linear adsorption for all active adsorbent sites) and 10 (a higher degree of interaction between the adsorbent and study material ions) [27]. The n represents the distribution of adsorbent particles attached to the adsorbent surface. Also, 1/n with values ranging from zero to one represents surface heterogeneity. The surface heterogeneity increases as the value of n approaches zero. Further, 1/n = 0, 0 < 1/n < 1, 1/n > 1 represent the reversible, desirable, and undesirable adsorption, respectively.
The constants and parameters of the Freundlich isotherm model were also determined. The value of n in the cefotaxime adsorption from the actual hospital wastewater using the synthesized membrane containing nanoparticles at the temperatures of 21°C and 37°C was estimated to be 6.154 mg/g and 0.374 mg/g, respectively. The value of 1/n reflects the desirable and undesirable cefotaxime adsorption from hospital wastewater at 21°C and at 37°C, respectively. Based on the Freundlich isotherm model, the coefficient of determination (R2) was determined 0.8684 mg/g and 0.7522 mg/g at the temperatures of 21°C and 37°C for the cefotaxime adsorption from hospital wastewater using membrane containing nanoparticles, respectively, indicating the capability of Freundlich isotherm model in describing the isothermal adsorption of the drug (Figs. 10 and 11).
According to our findings, the cefotaxime adsorption from real hospital wastewaters using synthesized membranes containing nanoparticles is based on the Langmuir (but 21°C) and Freundlich isotherm models.
Simulation Studies
The Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) is an open-source molecular dynamics simulator that is utilized for computational investigations [28]. This simulator is one of the first MD codes introduced for particle modeling in various gas, liquid, and solid phases. Multiple force fields and potentials, as well as various boundary conditions, can be used to model different atomic, polymer, biological, metal systems and a combined system composed of these systems. The CFF91force field designed for usage in polymers and organic materials was used to perform calculations. The PES polymeric membrane was prepared using the polymer builder package utilized in the material studio program. To do so, the head and tail points of the monomer were determined, and two strains of PES polymeric membrane with a length of 15 monomers were established using the material studio program, as shown in Figs. 12 and 13.
Similarly, the material studio program was used in preparing the PVP polymer of 25 monomeric chains. A specific morphology was achieved by adding two SiO2 and TiO2 nanoparticles to the membrane. The desired percent composition was selected based on the mass percent composition utilized in the experiments. In the material studio program, all of the structures were given to the Amorphous Cell module, and the membrane structure was built using the default density to get the required membrane structure. Figure 14 shows a schematic representation of the achieved membrane.
In the LAMMPS molecular dynamics simulator, the equilibrium and associated polymer consistent force field (pcff) coefficients were determined for each atom type [29]. The cut-off radius for electrostatic was chosen as 12 Aº, and Van der Waals interaction was chosen as 12 Aº. The long-distance electrostatic interactions of the system were calculated using the k-space algorithm [30]. The temperature and pressure of the system were controlled using a Nose-Hoover thermostat and barostat during the membrane preparation [31]. Also, the Tdamp and Pdamp values were chosen as 100 and 1000, respectively. In all stages, the time step was set at 1 femtosecond.
Of designs established to compress and balance the membrane density, two cases are discussed as follows. In 2000, a 12-step preliminary compression design was introduced [28], followed by the 12-step annealing and compression design [32]. Later, in 2011, a general overview was developed based on the 12-step annealing and compression design [33]. Table 3presents the characteristics of this 21-step compression/equilibrium general design utilized for density balance. This 21-step design is known as a simulated annealing algorithm where the membrane is periodically cooled and heated to generate the final structure. During compression, the box changes dimensions and overcomes the energy barriers on the study system. As a result, the final density of the system will have a more precise equilibrium value besides its more reliable morphology.
Table 3
|
Stage
|
Slow decentralization
|
Time (ps)
|
|
1
|
NVT 600 K
|
50
|
|
2
|
NVT 300 K
|
50
|
|
3
|
NPT 0.02Pmax bar, 300 K
|
50
|
|
4 and 5
|
NVT 600 K, NVT 300 K
|
50, 100
|
|
6
|
NPT 0.6Pmax bar, 300 K
|
50
|
|
7 and 8
|
NVT 600 K, NVT 300 K
|
50, 100
|
|
9
|
NPT Pmax bar, 300 K
|
50
|
|
10 and 11
|
NVT 600 K, NVT 300 K
|
50, 100
|
|
12
|
NPT 0.5Pmax bar, 300 K
|
5
|
|
13 and 14
|
NVT 600 K, NVT 300 K
|
5, 10
|
|
15
|
NPT 0.1Pmax bar, 300 K
|
5
|
|
16 and 17
|
NVT 600 K, NVT 300 K
|
5, 10
|
|
18
|
NPT 0.01Pmax bar, 300 K
|
5
|
|
19 and 20
|
NVT 600 K, NVT 300 K
|
5, 10
|
|
21
|
NPT 1 bar, 300 K
|
100
|
| P max = 40 Gpa |
The system is fed into NVT and NPT ensembles regularly in this design. At each ensemble, the compaction is carried out by increasing the temperature and pressure of the system. Then, periodic boundary conditions were applied to all three dimensions following the Mont Carlo GCMC simulations using LAMMPS molecular dynamics simulator. The nature of the interactions between the adsorbent and adsorbed molecules was determined using a force field similar to that utilized for the membrane. An in-house program was used to convert the Ceftriaxone molecule's parameters into the files required for LAMMPS simulation. The adsorption coefficients were calculated at three temperatures of 21°C, 28°C, and 37°C and a radiation frequency of 12 Aº by applying the pressure ranging from 1 kPa to 10000 kPa. After balancing the system with 106 steps, additional 106 steps were completed to acquire and evaluate data.
Adsorption of cefotaxime molecules on membranes
Cefotaxime adsorption isotherms were determined on the membrane surface in the pressures ranging from 1kPa to 10000 kPa, at three different temperatures of 21°C, 28°C, and 37°C, as shown in Figs. 15–17.
It was revealed that the adsorption rate of cefotaxime on the membrane at 21°C and 28°C is approximately 2.6 times the adsorption rate at 37°C at a pressure of 1 kPa. Figure 16. shows the adsorption rate of cefotaxime at a temperature of 21°C. As shown in the figure, at extreme pressure, the membrane adsorption is about 15 molecules per membrane (by adding the percentages from the previous section). At this temperature and pressures near 3000 kPa, the cefotaxime adsorption is constant, indicating that membrane sites are saturated at this pressure and temperature. Figure 17. shows the adsorption rate of cefotaxime at 28°C. The adsorption curve shows the adsorption rate of 16 cefotaxime molecules at zero pressures so that the adsorption rate should be increased by increasing pressure. Although the membrane seems to exhibit evidence of saturation at a pressure of 9000 kPa, this figure does not show the degree of saturation for this membrane versus cefotaxime at 28°C.
When comparing the adsorption diagrams at 21°C and 28°C, it can be found that a similar pattern can be anticipated for cefotaxime adsorption at two temperatures, implying that the adsorption rate of cefotaxime molecules at both temperatures is comparable. Figure 17. shows the adsorption rate of cefotaxime at 37°C. The adsorption diagram does not follow the same pattern as the previous two temperatures, which is surprising. At 37°C, the adsorption rate of the membrane is much lower than at the previous two temperatures at pressures near zero. At high pressures, a similar pattern can be seen, which means that the membrane follows a similar pattern and is saturated at much lower pressures.
Figure 18 shows the morphological analysis of the membrane after maximum cefotaxime is absorbed at the three desired temperatures, in addition to the density of adsorbed cefotaxime as aqua fields or phases in the membrane (distribution of cefotaxime molecules on the membrane). In addition, the occupied membrane sites are shown at all three temperatures, representing the membrane cavities available to trap the cefotaxime molecules.