Figure 1 shows the nitrogen sorption results of different montmorillonite samples. All isotherms correspond to H3 hysteresis adsorption associated with capillary condensation in nanopore structures (Sing 1985). The type H3 is normally corresponding to plate-like particles which cause slit-shaped pores (Maged et al. 2020). After acid treatment of montmorillonite, an increase in the surface area of montmorillonite samples was observed (from 28 up to 92 cm2/g, Table 1), indicating an increase in the porous structures on the montmorillonite layers. This is due to the destruction of the octahedral structure of the montmorillonite (Bieseki et al. 2013, Franco et al. 2016). Besides, the pores volume increased hugely after acid treatment, indicating the formation of nanopores (Table 1). The Fe saturation hardly affected the nitrogen sorption results of the montmorillonite or the acid-treated one. The Fe is enriched on the surfaces of the montmorillonite, however, it is proposed that the Fe is saturated in the removed octahedral sheets of the nanoporous montmorillonite. After acid treatment, the porosity of the montmorillonite (such as specific surface area and pore volume) was increased. Besides, according to XRF analysis (Table 2), most cations (Ca2+ and Na+) were leached out from the structure. In addition, the amount of Mg and Fe was slightly reduced, indicating the effect of acid treatment on the montmorillonite.
Table 1
Results of nitrogen sorption analysis
Sample
|
BET surface area (m2 g−1)
|
Pore size (nm)
|
Total pore volume (cm3 g−1)
|
Mt
|
28
|
|
0.12
|
Mt-Fe
|
26
|
|
0.11
|
NP-Mt
|
92
|
|
0.40
|
NP-Mt-Fe
|
89
|
|
0.46
|
Table 2
XRF analysis of montmorillonite samples
Sample
|
Mass (%)
|
SiO2
|
Al2O3
|
CaO
|
Fe2O3
|
K2O
|
MgO
|
TiO2
|
Na2O
|
LOI
|
Mt
|
54.5
|
16.4
|
1.9
|
4.1
|
0.4
|
3.7
|
0.4
|
0.6
|
18.0
|
Mt-Fe
|
51.2
|
13.5
|
1.9
|
11.1
|
0.3
|
3.5
|
0.4
|
0.5
|
17.6
|
NP-Mt
|
60.8
|
13.3
|
0.2
|
2.1
|
0.5
|
2.1
|
0.4
|
0.1
|
20.5
|
NP-Mt-Fe
|
56.7
|
10.6
|
0.2
|
11.8
|
0.5
|
2.2
|
0.2
|
0.1
|
17.7
|
The SEM images of samples confirm the layered morphology of montmorillonite and the acid-treated ones (Figure 2). Cracked aggregates, segmented and flaky pieces, and reduced particle size are some of the noteworthy signs of changes in the montmorillonite structure via acid treatment. On both Mt and NP-Mt samples, no significant changes in the structure of montmorillonite occurred after Fe saturation.
FTIR analysis was performed to investigate the changes made to montmorillonite structure through acid treatment (Figure 3). After acid treatment (to provide nanoporous montmorillonite), the intensity of bands at 835, 875, and 914 cm−1 decreased, which are corresponding to the bending vibration of AlMgOH, AlFeOH, and Al2OH, respectively (Franco et al. 2016). This could be due to the leaching of the cations of octahedral sheets by order of Al, Mg, and Fe. In addition to cation removal, and consequently their hydroxyl groups (3620 cm−1), some changes were detected in the intensity of the bands at 1113, 1035, 792, 610, 522, and 462 cm−1 (Figure 3). These decreases showed the acid has influenced tetrahedral and octahedral sheets (Si-O-Si, Al-O-Si), and Si-O bands were decomposed partially (Madejová 2003, Maged et al. 2020). Comparing the spectrum of montmorillonite with Fe saturated samples, a new peak emerged at 700 cm−1, which represents the FeOOH bending vibration. This peak suggested that Fe ions are hydrolyzed into the interlayer spaces of the montmorillonite by pillaring (Yuan et al. 2008).
The kinetics of adsorption and desorption of TC onto different montmorillonite samples were plotted in Figure 4 as well as the results of two kinetic models (pseudo-second-order and intraparticle diffusion models) fitted to the data (details of kinetics models are presented in Table 3). For all samples, the equilibrium of adsorption reached after 60 minutes and the Mt and Mt-Fe samples adsorbed more TC than the nanoporous montmorillonite samples to some extent at all pH which is due to the partial destruction of the nanoporous montmorillonite after acid treatment and lower cation exchanges.
Table 3
Details of kinetics models fitted to the adsorption data
Sample
|
Adsorption pH
|
Pseudo second order
|
|
Intraparticle diffusion
|
k
|
R2
|
|
k1
|
R12
|
k2
|
R22
|
k3
|
R32
|
Mt
|
3
|
0.0121
|
1
|
|
32.5
|
0.98
|
8.4
|
0.91
|
0.2
|
0.98
|
Mt-Fe
|
0.0125
|
1
|
|
34.5
|
0.98
|
8.4
|
0.91
|
0.1
|
0.97
|
NP-Mt
|
0.0137
|
1
|
|
27.9
|
1
|
7.8
|
0.92
|
0.2
|
0.99
|
NP-Mt-Fe
|
0.0133
|
1
|
|
30.1
|
1
|
7.6
|
0.91
|
0.2
|
0.98
|
Mt
|
5
|
0.0115
|
1
|
|
33.3
|
0.98
|
12.3
|
1
|
0.5
|
0.98
|
Mt-Fe
|
0.0118
|
1
|
|
31.7
|
0.99
|
9.1
|
0.96
|
0.6
|
0.97
|
NP-Mt
|
0.0134
|
1
|
|
27.8
|
1
|
7.9
|
0.98
|
0.2
|
0.95
|
NP-Mt-Fe
|
0.0138
|
1
|
|
28.9
|
1
|
7.1
|
0.96
|
0.3
|
0.95
|
Mt
|
8
|
0.0113
|
1
|
|
33.6
|
0.98
|
12.4
|
0.97
|
0.4
|
1
|
Mt-Fe
|
0.0124
|
1
|
|
31.2
|
0.98
|
11.2
|
0.99
|
0.3
|
0.97
|
NP-Mt
|
0.0171
|
1
|
|
22.6
|
1
|
5.8
|
0.99
|
0.2
|
0.99
|
NP-Mt-Fe
|
0.0184
|
1
|
|
21.8
|
1
|
5.1
|
0.98
|
0.3
|
1
|
The pseudo-second-order model fitted to all data (Figure 4 and Table 3) indicated a similar trend and the adsorption capacity was reduced by increasing the pH. A higher amount of TC adsorbed at lower pH (5 and slightly at 7) which is due to the higher positive charge of TC at this condition (Figueroa et al. 2004) and resulted in a facile cation exchange mechanism (Akbari Alavijeh et al. 2017, Ramazani Afarani et al. 2018). By increasing the pH to 9 higher negative charges of TC resulted in the lower tendency in the adsorption onto the montmorillonite. In addition, lower interlayer adsorption was detected at this condition.
The intraparticle diffusion model results indicated that the adsorption of TC on the external surfaces of all montmorillonite particles was almost the same at all pH of the solution (Figure 4 and Table 3 the first linear region of the plots). At acidic and neutral pH the intraparticle diffusion was another important mechanism of adsorption for all samples which implies the adsorption of TC in the interlayer spaces which will be discussed in the XRD results as well. However, by increasing the pH of adsorption, the intraparticle diffusion reduced for nanoporous samples (acid treated ones) indicating a lower tendency of TC to be diffused in the interlayer spaces due to the higher negative charges of the TC and lower chance of cation exchange mechanism with montmorillonite interlayer cations (Chahardahmasoumi et al. 2019).
The changes in the layered structure of montmorillonite after treatments and adsorption of TC were analyzed using XRD analysis (Figure 5). After acid treatment the basal spacing of the montmorillonite was increased from 11.6 to 13 Å which could be due to the destruction of the octahedral structure of the montmorillonite upon acid treatment (Temuujin et al. 2004). The saturation of both montmorillonite and nanoporous montmorillonite samples with Fe ion did not affect the basal spacing very much which is in accordance with what was reported previously (Hong et al. 2019, Zhao et al. 2020). The XRD results indicated that at two acidic and neutral pH the TC was adsorbed mainly on the interlayer spaces of all montmorillonite samples (the basal spacing was increased for 6 to 7 Å, Figure 5). However, at pH 9 for two samples that were not acid-treated no changes in the interlayer spaces was detected showing no adsorption of TC between interlayer spaces. while for the acid-treated (nanoporous montmorillonite samples) at pH of 9, a tween peak was detected showing a mutual adsorption of TC in the interlayer spaces and on the external surfaces of the samples (this trend is in accordance with the intraparticle diffusion models, Figure 4 and Table 3). The reason for adsorption of TC in the interlayer spaces of the nanoporous montmorillonite at pH 9 is the increase in the cation charges (H+) in the interlayer spaces created during the acid treatment which allows TC to diffuse in the interlayer spaces and increase the basal spacing (TC is negatively charged at this pH) (Chahardahmasoumi et al. 2019). Enrichment of the montmorillonite with Fe ion did not change the trend of TC adsorption on the external surface or into the basal spacing of different samples (Figure 5).
Surprisingly, more than 30% of adsorbed TC was desorbed from all samples with a linear trend at all pH conditions (Figure 4), showing that the montmorillonite (natural or nanoporous one) could not fully immobilize TC in its structure and release it back to the environment in high amount. Higher amounts of TC were desorbed from the montmorillonite samples at pH of 9 (Figure 4) indicating more repulsion forces between the montmorillonite and the TC. The observed desorption of TC from montmorillonite samples was somehow expected as reported previously, however, compared to the desorption rate from agricultural soil (Conde-Cid et al. 2020), such a high amount of desorption was contrary to expectation (Figure 4). This phenomenon is partly due to the different nature of the adsorbents and partly due to the extremely low concentration of TC in the solution (in the range of 0.003 to 0.2 mg/ml) for adsorption onto the agricultural soil (Conde-Cid et al. 2019). However, by increasing the TC concentration in the solution the amount of TC adsorbed was increased (Figure 4) and the desorption rate onto the montmorillonite samples increased as well, which is a realistic prediction of the empirical situation. Hence, as the montmorillonite might reach its maximum adsorption capacity it finally releases the adsorbed TC to the environment. This could be a big challenge in the use of such environmentally friendly and cheap adsorbents for adsorption/separation of antibiotics such as TC. One possible remediation for such a problem is to apply a second capability on the montmorillonite to degrade the TC in the adsorption-desorption cycle. Hence the desorbed pollutants do not harm the environment upon desorption. A possible remediation process is to apply a photo-Fenton enhanced surface for photodegradation of TC. The Fe ions were loaded on the montmorillonite for this purpose and the properties of montmorillonite on the degradation of TC during the desorption process were analyzed.
To check the photocatalytic properties of montmorillonite samples, after adsorption of the TC, the montmorillonite was separated and put in a solution of the same pH and the light irradiation was applied under shaking. Then the antibacterial activity test on the solution was done in different duration and the results were compared with the results of the antibacterial activity of the same sample in which the adsorption-desorption cycle were carried out in dark. Antibacterial activity results and the inhibition zone summary results of released TC from montmorillonite samples were plotted in Figures 6 and 7, respectively.
For the Mt and NP-Mt sample, almost no changes in the antimicrobial activity of the TC were detected after irradiation under light for up to 180 minutes at all pH compared to samples in dark (the inhibition zone in Figure 6a1, 6b1, 6c1, 6a3, 6b3, 6c3, and 7). By icreasing the desorption duration the inhibition zone increased (Figure 7) up to 120 minutes due to the increasing amount of TC desorbed and after that almost remained constant. Comparing the samples under light and in dark (comparing disk 1 with 4, 2 with 5, and 3 with 6 in Figure 6) the inhibition zone did not vary very much when placed under light for these two montmorillonite samples. Hence, the main conclusion is that the TC is desorbed from the montmorillonite and reasonably strong antimicrobial activity was detected even after being under light for 180 min. Considering the application of such adsorbent in the environment, montmorillonite does not perfectly separate the TC from the soil and release it back to the environment with almost no changes.
On the other hand, for the Fe saturated samples, the antibacterial activity of TC was reduced significantly even after 60 minutes of light irradiation (Figure 6a2, 6b2, 6c2, 6a4, 6b4, 6c4, and 7). The reduction in the antimicrobial activity of TC was higher for the nanoporous Fe saturated montmorillonite than the normal Fe saturated montmorillonite. The main reason could be the higher accessible site for the adsorption and possible photodegradation process. Indeed, the higher transparent nanoporous structure provides a higher chance for interlayer adsorbed TC to absorb light and be degraded. By increasing the light irradiation time from 60 to 180 the antimicrobial activity reduced steadily for Fe saturated samples compared to the same samples in dark. Therefore, during the desorption, the degradation of the TC occurs for Fe saturated samples and this process continues in an equilibrium adsorption-desorption-photodegradation cycle. Meanwhile, the degradation of TC was facilitated at acidic conditions due to the higher amount of TC desorbed from the montmorillonite samples (Figure 4) and a lower chance of degradation.
It is proposed that the photodegradation happens through a photo-Fenton process in which the Fe cations seemingly played a key role in the photo-Fenton deactivation of TC (Wang et al. 2020b). Fe cations degraded the TC by the production of active species via Fenton-type reaction. The production of ˙OH radicals is the key element in the degradation of the TC via creating complexation (Brillas &Garcia-Segura 2020, Fareed et al. 2021, Liu et al. 2021b). Although the mechanism of the H2O2 decomposition in the photo-Fenton reaction is not well understood, the following reactions have been suggested corresponding the reduction of Fe3+ with the production of ˙HO2 radicals, followed by regeneration of Fe3+ which resulted in the generation of the hydroxyl radicals (Ramirez et al. 2007):
X-Fe3+ +H2O2 → X-Fe2+ + ˙HO2 + H+ (1)
X-Fe2+ +H2O2 → X-Fe3+ + OH− + ˙HO (2)
TC + ˙OH → degraded TC (3)
in which the X represents the surface of the montmorillonite surface. The concentration of ˙OH radicals was evaluated using the coumarin method in which the produced ˙OH radicals would react with the coumarin and produce 7-hydroxycoumarin that is highly fluorescent (Louit et al. 2005) and can be detected with a spectrophotometer (Figure 8). The proposed reactions take place on the montmorillonite surface when enriched with the Fe cations (Scheme 1). Based on the proposed mechanism, the Fe saturation of the non-porous montmorillonite mainly occurs on the external surfaces (Scheme 1). The TC adsorbed on the surfaces of the montmorillonite would be in a part of the Fenton reaction in which the created ˙OH radicals (Figure 7) helps to degrade TC. However, for the nanoporous montmorillonite, the nanopores (Scheme 1), giving rise to a protonated (H+) amorphous hydrous silica, provide high accessible surface area for adsorption of Fe cations (Wei et al. 2017). In this structure, the absorption of light would be much higher than the non-porous montmorillonite. Hence a great opportunity for the production of ˙OH radicals was proved by the fluorescence spectra (Figure 8) and hence the photo-Fenton reactions would occur.