Synthesis and application of g-C3N4/Fe3O4/Ag nanocomposite for the efficient photocatalytic inactivation of Escherichia coli and Bacillus subtilis bacteria in aqueous solutions.

Contamination of water with bacteria is one of the main causes of waterborne diseases. The photocatalytic method on the basis of bacterial inactivation seems to be a suitable disinfectant due to the lack of by-products formation. Herein, g-C3N4/Fe3O4/Ag nanocomposite combined with UV-light irradiation was applied for the inactivation two well-known bacteria namely, E. coli and B. subtilis. The nanocomposite was prepared by a hydrothermal method, and subsequently it was characterized by XRD, FT-IR, SEM, EDX and PL analyses. The optimum conditions established for the inactivation of both bacteria were as follows: nanocomposite dosage 3 g/L and bacterial density of 103 CFU/mL. In the meantime, the efficient inactivation of E. coli and B. subtilis took 30 and 150 min, respectively. The results also revealed that inactivation rate dropped with an increase in the bacterial density. It is also pointed out that OH˚ was found out to be the main radical species involved in the inactivation process. Finally, the kinetic results indicated that the inactivation of E. coli and B. subtilis followed the Weibull model. It is concluded that C3N4/Fe3O4/Ag nanocomposite along with UV-light irradiation is highly effective in inactivating E. coli and B. subtilis bacteria in the aqueous solutions.

Introduction

Given on-going growing population and climate change phenomena, provision of high quality and clean water from reused aqueous sources remains a great challenge (Widi et al. 2018). This highlights the importance of water purification with regard to chemical and microbial contamination (Rojviroon and Sirivithayapakorn, 2018; Feilizadeh et al. 2015). It is crystal clear that the presence of pathogens in water has become a big concern worldwide (Fang et al. 2013). Bacteria, viruses and fungi are widely found in water resources and pose significant health risks to humans and animals (Xia et al. 2017).

Recently, the development of new technologies for the treatment of pathogens in aquatic environments has dramatically increased (Widi et al. 2018). Until now, various methods such as chlorination, UV and ozone have been used to disinfect and remove pathogens (Ouyang et al. 2016). The major drawback of the chlorination process is the reaction of chlorine with the natural organic matter (NOM) present in water which leads to the formation of disinfectionby-products (DBPs) namely, Trihalomethanes (THMs) and Haloacetic acids (HAAs) (Li et al. 2015; Zazouli et al. 2017). These compounds have high carcinogenic effects even at low concentration levels (Zazouli et al. 2017). UV method is effectively used to inactivate microbial agents, but it is expensive and requires high level of energy (Zhang et al. 2014). Moreover, many pathogens are resistant to UV and chlorine (Xia et al. 2017). In recent years, advanced oxidation processes have widely been applied for water and wastewater treatment (Abeledo-Lameiro et al. 2016).

Advanced oxidation processes (AOPs) are founded on the basis of the production of reactive oxygen species (ROSs) (Zhang et al. 2017). The ROSs produced during the photocatalytic process can damage biologically vital macromolecules including DNA, proteins and lipids and alter cell permeability (Wang et al. 2015b).

Heterogeneous catalysts are widely implemented to decompose organic pollutants and inactivate microbial agents and pathogens (Armon et al. 2004). In these catalysts, the electrons lying within the valence band are stimulated and pushed through the conduction band leaving behind a hole in the valence band (Zhang et al. 2017). The main feature of these catalysts are non-toxicity and high stability (Di Palma et al. 2019).

Lately, Carbon nitride graphite (g-C3N4)-based nanocomposites have extensively been used for the photocatalytic degradation of various pollutants (Wang et al. 2015a). g-C3N4 is a polymeric organic semiconductor and has properties such as environmental compatibility, high chemical stability, low cost. It is of two-dimensional structure and low-energy band width (2.7 eV) (Mousavi and Habibi-Yangjeh, 2017). However, the rapid re-coupling of electron–hole pairs remains the main problem associated with the use of g-C3N4 is, which results in the reduction of photocatalytic activity (Mousavi and Habibi-Yangjeh, 2016). As a remedy, metallic/ nonmetallic doping and combination with different semiconductors have been introduced (Pant et al. 2017). In addition, the separation of the catalysts used in the photocatalytic processes is another setback to be overcome (Mousavi and Habibi-Yangjeh, 2017). To fix the above-mentioned problem, the combination of Fe3O4 nanoparticles with g-C3N4 sheets has been proposed. This facilitates the quick separation of catalysts from the refined solutions using an external magnet (Akhundi and Habibi-Yangjeh, 2017). A number of studies have shown that g-C3N4/Fe3O4 can improve the performance of photocatalytic processes (Ding et al. 2018). It should also be noted that Fe3O4 nanoparticles can act as intermediates for the rapid transfer of producing electrons due to their high conductivity (Li et al. 2019). Thus, it seems that the introduction of highly conductive elements within Fe3O4 nanoparticles could certainly improves the separation efficiency of charge carriers (Ghodsi et al. 2020). On the other hand, loading semiconductor surfaces with metals such as Pt, Au and Ag can enhance the photocatalytic activity under light irradiation (Mousavi and Habibi-Yangjeh, 2015).

Amongst the mentioned conductive metals, the disinfection properties of Ag have well been understood for a long time. The advances in nanotechnology have also improved the efficiency of its disinfection. On the other hand, Ag is not associated with the production of any by-products, nor the creation of odor, taste, color, etc. As well as being highly effective in disinfection. Ag is non-toxic, non-irritating, non-allergic, hydrophilic, tolerant to various conditions (ie., very stable), environment friendly, heat resistant and does not escalate the resistance and adaptability of microorganisms (Ma et al. 2016; Tran and Le, 2013; Sondi and Salopek-Sondi, 2004).

Briefly, in this study, g-C3N4/Fe3O4/Ag nanocomposite was initially synthesized and characterized by respective hydrothermal and SEM, EDX, XRD, FT-IR and PL methods. The applied nanocomposite was used to inactivate the target bacteria. The Gram-positive bacterium (Bacillus subtilis ATCC 6636) and the Gram-negative bacterium (Escherichia coli ATCC 25922) were used as the target models throughout. Following on, the Weibull, Log-Linear, and Biphasic models were also used to describe the kinetic behavior of the bacterial inactivation.

Experimental

Materials

Melamine (C3H6N6, 99%), silver nitrate (AgNO3, 99.8%), iron chloride tetra hydrate (FeCl2.4H2O, 98%),, Polyvinylpyrrolidone (PVP, 99%), tert-butanol (C4H10O), ammonium oxalate (C2H8N2O4), benzoquinone (C2H4O3), sodium hydroxide (NaOH), hydrochloric acid (HCl, 99%), ethanol (C2H5OH, 98%), methanol (CH3OH, 98%), EMB agar culture medium, BHI culture medium, Loria Bertani (LB) culture medium, glycerol (C3H8O3, 10%), blood agar (BA) culture medium, sulfuric acid (H2SO4, 98%), Barium chloride (BaCl2), ammonia (NH3, 28%) were obtained from the Merck Company.

Synthesis of nanocomposite

Yellow powder g-C3N4 was obtained by heating melamine in the furnace at 550 ºC for 4 h. To prepare g-C3N4/Fe3O4/Ag nanocomposite, a hydrothermal method was implemented. Briefly, 50 g g-C3N4 was added to 30 mL distilled water and dissolved with aid of ultrasonic waves. After that, 0.5 g of FeCl2.4H2O, 0.025 g of AgNO3 and 0.1 g PVP were added to the solution while stirring for 3 h. Then 2.5 mL NH3 was added to the above solution and it was vigorously agitated for 10 min. The obtained suspension was transferred to the Teflon cell and autoclaved at 140 °C for 3 h. After that, at room temperature, the suspension was filtered and washed with water and ethanol and subsequently dried at 80 °C for 12 h (Pant et al. 2017).

Characterization

Following the preparation of g-C3N4/Fe3O4/Ag, the physico-chemical properties of the nanocomposite was determined by the identification of the crystalline phase by the XRD (X-ray diffraction) experiments within the range of 2θ = 20–80°. The presence of Fe and Ag elements in the structure of the nanocomposite was confirmed by the energy dispersive X-ray (EDX) analysis. Scanning electron microscopy (SEM) was used to determine the morphology of the synthesized catalyst. Finally, to identify the functional groups of the nanocomposite the Fourier-transform infrared spectroscopy (FT-IR) and photoluminescence (PL) techniques were implemented.

Preparation of bacterial samples

The bacterial strains used in this study included the Gram-negative bacterium (Escherichia coli ATCC 25922) and the Gram-positive bacterium (Bacillus subtilis ATCC 6636). The bacteria were lyophilized from the collection center of the Iranian industrial microorganisms. To remove the lyophilized bacteria, each bacterium was inoculated with 1 ml of the BHI liquid medium and incubated for 24 h. Then, the BHI medium containing 10% glycerol was prepared for the long-term storage of the bacteria. Following on, 10 µl of the desired bacteria were individually transferred to the containing-glycerin BHI media and stored at −18 °C. For daily inactivation tests, the standard bacterial strain samples were placed in the incubator to freeze. After that, 0.1 ml of each sample was heated on a shaker at a speed of 180 rpm in the LB culture medium at 37 °C for 18 h. The bacteria were finally separated by centrifugation (5000 rpm, 15 min) and washed with the normal saline water (0.9% w/w) to remove the residual culture medium (Ruales-Lonfat et al. 2016).

Figure 1 shows the stages of bacterial extraction from lyophilization phase, in which the bacterium was cultured in TSA and BHI Culture medium after leaving the lyophilization state and finally cultured on the final culture medium to multiply the bacteria.

Fig. 1

Removal of the bacteria from the lyophilized phase followed by the plate transfer

Photocatalytic experiments

To perform the photocatalytic disinfection, a number of fixed levels of nanocomposite (0.5, 1.5, 3, 5 g/L) were added in 100 ml sterile salt (0.9% w/w) while stirring by ultrasonic waves (35 kHz) for 1 min (Blatchley et al. 2005). Following that, the specific bacterial densities (Escherichia coli ATCC 25,922 and Bacillus subtilis ATCC 6636) were prepared by an optical density (OD) method and added to the solution (pH 7) and subsequently exposed to the UV lamp (3.3 mw/cm2) placed 10 cm above the reactor. At each sampling stage, 100 µL of the diluted sample was added to the culture medium. In a further development, the samples containing the Escherichia coli were homogenized on the EMB agar medium and incubated at 37° C for 24 h. The Bacillus subtilis sampling was similar to that of Escherichia coli, except for the fact that the sample was incubated in the BA culture medium at 30 °C for 24 h. The number of colonies was then counted using the counter colony according to the following Equation (Spuhler et al. 2010; Matin et al. 2018).C: CFU / mL, n: Number of colonies on a plate, d: Growth Factor, V: Size of transitional sample for culture on plate.

Kinetic models

Logarithmic model (log-linear model)

The linear logarithm or the Chick-Watson model is the modified form of the Chick model. The model assumes that there is a stoichiometric relationship between the disinfectant molecules and the number of inactivated microorganisms. The Chick-Watson equation is described as follows:where N0 and Nt are the microbial density before and after the inactivation process. C, K and t are the concentration of disinfectant, the first-rate inactivation and the inactivation time, respectively (Sun et al. 2007).

Weibull model

Mafart et al. developed a deactivation kinetic model based on the Weibull distribution (Mafart et al. 2002). Unlike the first-rate model, which assumes the bacterial population is homogenous, Mafart et al. hypothesized that the microbial population would be so heterogeneous that each cell death in the face of external stresses would require different contact times depending on their level of resistance. The cells follow the Weibull distribution model which is expressed as presented below:δ represents the time required for the first part of the reduction, the duration at which the first logarithmic decline occurs in the bacterial population. β values are varied with the shape of the equation curve. At β > 1, the curve has a downward concave shape whilst at β < 1 an analogous upward shape is observed. At β = 1, the first-order linear logarithm model is seen (Albert and Mafart 2005).

Biphasic model

This model, which is based on two fractions, was proposed by Cerf in 1977. It is assumed that there are two subpopulation groups with different susceptibility to disinfection.where P represents a fraction of living microorganisms related to the group 1 subpopulation and (1-P) is indicative of a fraction of living microorganisms related to the group 2 subpopulation. K1 is the kinetic constant for the sensitive population and K2 is the kinetic constant for the higher resistance population (Cerf 1977).

Results

Characterization of nanocomposites

SEM analysis was used to identify the morphology and the composition of the compounds within their surface layers. Figure 2a(I) shows the structure of pure g-C3N4 with overlapping irregular plate structure. Figure 2a (II) exhibits the structure of g-C3N4/Fe3O4/Ag nanocomposite in which the distribution of Fe3O4/Ag onto g-C3N4 is marked by the arrow.

Fig. 2

a SEM image of pristine g-C3N4 sheets (I), SEM image of g-C3N4/Fe3O4/Ag (II), b EDX spectra of g-C3N4/Fe3O4/Ag nanocomposite, c Magnetism property of C3N4/Fe3O4/Ag nanocomposite, d XRD images of g-C3N4 and g-C3N4/Fe3O4/Ag, e FT-IR images of g-C3N4 and g-C3N4/Fe3O4/Ag, f PL Spectra of g-C3N4 and g-C3N4/Fe3O4/Ag

EDX analysis was performed to identify the elements present in the nanocomposite structure. The results confirmed the presence of Ag, Fe, O, N and C elements in the nanocomposite structure as well as the fractional weight for each element (Fig. 2b).

The magnetic properties of g-C3N4/Fe3O4/Ag nanocomposites is illustrated in Fig. 2c. As can clearly be seen, the nanocomposite was completely separated off the solution phase following the inactivation process underlying a remarkable magnetic propery of g-C3N4/Fe3O4/Ag nanocomposite.

The X-ray diffraction (XRD) pattern was used to identify the phase type as well as the crystalline properties of g-C3N4 and the nanocomposite. The XRD analyses were performed within the 2θ range of 20–80° for pure g-C3N4 and g-C3N4/Fe3O4/Ag nanocomposite. For pure g-C3N4, a strong peak at 27.6 was observed, which is consistent with the planes (002), referring to an aromatic compound in Fig. 5. In the case of the nanocomposite, the observed peaks were located at 30.2, 35.8, 43.5, 53.7, 57.3 and 62.7°, which are in agreement with the planes (220), (311), (400), (422), (511), (440), respectively. They are all related to Fe3O4 structure and consistent with the results obtained from the previous studies (Akhundi and Habibi-Yangjeh, 2016; Zhu et al. 2017). Also in the XRD pattern of the nanocomposite, a small drop in the peak intensity compared to that of pure g-C3N4 could be attributed to the accumulation of Fe3O4 in the nanocomposite structure. The respective peaks indexed at 38, 44.2, 64.4 and 77.4° refer to the planes (111), (200), (220) and (311) implying the presence of Ag in the nanocomposite structure(Zhu et al. 2016).

Fig. 5

a Effect of irradiation on inactivation of E. coli inactivation under the optimal conditions, b Effect of irradiation on inactivation of B. subtilis inactivation under the optimal conditions

In a further related confirmatory analysis, FT-IR was implemented to identify the organic functional groups within the structure of the compounds. Figure 6 represents the FT-IR spectra of pure g-C3N4 and g-C3N4/Fe3O4/Ag at the range of 400–3900 cm−1. Within the spectra shown, pure g-C3N4 is highlighted with a broad absorption band from 3000 to 3400 confirming the tensile (–NH) and (–NH2) modes. The peaks located at 1251 cm−1, 1325 cm−1, 1419 cm−1, 1463 cm−1, 1575 cm−1 and 1639 cm−1 are related to (C–N) and (C=N) bonds. In addition, the peak shown at 808 cm−1 is related to the s-triazine units and the broad band shown in the far right (430–650 cm−1) is attributed to Fe–O.

Fig. 6

Effect of the presence of scavenger on inactivation of E. coli and B. subtilis under the optimal

It is assumed that transfer and recombination process involving the electron–hole pairs plays an important role in a photocatalytic activity. Moreover, efficient separation of charge carriers can improve the photocatalytic activity. For this reason, photoluminescence (PL) technique was applied to investigate the coupling (i.e., recombination) of the cavities and electrons. The PL spectrum is depicted in Fig. 2f. As can be seen, the strong emission peak is observed for g-C3N4 compared to that of g-C3N4/Fe3O4/Ag. The lower PL intensity is attributed with the favorable electrical conductivity expected for Fe3O4 and Ag. This brings about an efficient electron transfer from the g-C3N4 conduction band to the Fe3O4 and Ag preventing the recombination of the charge carriers followed by an improvement in the photocatalytic activity. It is noted that our PL findings are in agreement with the results published elsewhere (Zhu et al. 2016; Pant et al. 2014).

The effect of nanocomposite dosage

The effect of g-C3N4/Fe3O4/Ag nanocomposite dosage on the photocatalytic disinfection of the target bacteria was investigated at a specified pH 7. Since the target bacteria (E. coli and B. subtilis) are sensitive to the change in the environmental conditions, the neutral pH (ca. pH 7) was temporarily chosen. The results exhibiting effect of nanocomposite dosage on the inactivation of the bacteria are illustrated in Fig. 3a, b. It was revealed that the inactivation rate for the target bacteria escalated with the rise in the nanocomposite dosage from 0.5 to 3 g/L. The inactivation rate for E. coli increased from 73.6 to 100% at 45 min while the corresponding value jumped for B. subtilis from 43.2 to 100% at 180 min.

Fig. 3

a Effect of catalyst dosage on E. coli inactivation rate, b Effect of catalyst dosage on B. subtilis inactivation rate, c bacterial inactivation efficiency for 3–5 repeated experiments using recycled sample

The effect of bacterial density on inactivation rate

The effect of bacterial density on the inactivation rate of E.Coli and B. subtilis is depicted in Fig. 4a, b. The results demonstrated that with increasing the bacterial density from 103 to 107 CFU/mL the inactivation rates decreased for both target bacteria over time.

Fig. 4

a Effect of initial bacterial density on E. coli inactivation, b Effect of initial bacterial density on B. subtilis inactivation

Effect of irradiation on inactivation of bacteria

In a yet further development, at the presence of the nanocomposite (g-C3N4/Fe3O4/Ag) the impact of irradiation on the inactivation rate of the bacteria of interest were thoroughly examined. For this reason, a number of experiments were performed under the dark and UV/Vis light conditions as illustrated in Fig. 5a, b. The experiments were done at the optimal conditions established earlier (pH 7, density 103 CFU/mL and 3 g/L nanocomposite). Applying the couple of UV/g-C3N4/Fe3O4/Ag led to a complete inactivation rate for E.Coli and B. subtilis, whilst the corresponding values for the joint Vis/g-C3N4/Fe3O4/Ag were 68.66 and 60.53%, after 30 and 150 min, respectively (Fig. 5b).

Effect of scavengers on inactivation of E. coli and B. subtilis

Various types of reactive oxygen species (ROS) are produced during a photocatalytic process. Generally, ROS damage the cellular components such as peptidoglycan layer, electron transfer chain, bacterial genome (DNA, RNA), protein and ribosome. It also alters cell the permeability and invades the cell membrane causing it to rupture and release the cytoplasmic content (Erdem et al. 2015). In this study, the effect of three common scavengers, namely ammonium oxalate (AO), tert-butanol (TB) and benzquinone (BQ) on the inactivation of the bacteria of interest under the optimum conditions were investigated. AQ, TB and BQ were used to control the holes (h+), hydroxyl radicals (OH˚) and superoxide radicals (O2˚−), respectively. As depicted in Fig. 6, the rate of inactivation for both E. coli and B. subtilis was reduced in the order of AO < BQ < TB. This indicates that OH˚ is the main active species in the current photocatalytic process and the roles of h+ and O2˚− are relatively negligible in this regard. Also the concentration of each scavenger was 0.1 mol.

Kinetic models

Three kinetic models namely, the Log-Linear, Weibull and Biphasic were employed to describe the kinetics of the inactivation process at three bacterial densities of 103, 105 and 107 CFU/mL (see Table 1). The GInaFiT software was used for the modeling, which was developed by Geeraerd et al. (2005). The statistical parameters including the coefficient of determination (R2) and the root mean sum of squared error (RMSE) were also used to determine the appropriate model describing the kinetics involved. Finally, a model with the maximum R2 and the minimum RMSE was selected as the appropriate kinetic model (Kashiri et al. 2018).

Table 1

Kinetic models used to describe the inactivation process

Model	Bacteria population	RMSE	R-Square	R-Square adjusted
E. coli
Log-linear	103	0.4043	0.9763	0.9723
	105	0.4480	0.8928	0.8571
	107	0.7712	0.8887	0.8510
Weibull	103	0.0374	0.9998	0.9990
	105	0.1368	0.9982	0.9968
	107	0.1392	0.9977	0.9943
Biphasic	103	0.2066	0.9948	0.9928
	105	0.2624	0.9755	0.9510
	107	0.2195	0.9915	0.9858
B. subtilis
Log-Linear	103	0.2729	0.9305	0.9450
	105	0.3636	0.9547	0.9417
	107	0.3624	0.9518	0.9242
Weibull	103	0.1053	0.9965	0.9955
	105	0.1153	0.9919	0.9878
	107	0.1705	0.9905	0.9865
Biphasic	103	0.1361	0.9849	0.9811
	105	0.1527	0.9839	0.9806
	107	0.2196	0.9825	0.9798

RMSE Root Mean Sum of Squared Error

DRS analysis

The photocatalytic activity of catalysts is closely related to their ability to absorb light. The UV–vis DRS is shown in Fig. 7 for g-C3N4/Fe3O4/Ag showed very strong absorption in the range of 400 to 700 nm. The adsorption peak is between 480 and 500 cm-1, which may be due to the intensification of surface plasmon resonance (SPR) of Ag species. The wider light absorption region g-C3N4/Fe3O4/Ag is able to maximize the use of light and produce load carriers by producing more efficient light, resulting in higher photocatalytic activity (Fig. 8).

Fig. 7

A schematic diagram representing the photocatalytic mechanism proposed for the inactivation of E. coli and B. subtilis

Fig. 8

The UV–vis DRS for g-C3N4/Fe3O4/Ag

Discussion

Increasing the nanocomposite dosage leads to a shape increase in the number of photons absorbed followed by the generation of further active radicals. However, with an increase in the nanocomposite dosage from 3 to 5 g/L, the respective inactivation rate fell down to 91.8% and 87.4% for E. coli and B. subtilis. It can be justified by the reason that with an excessive rise in the dosage of nanocomposite the turbidity increases. As a result, the UV photons are prevented from reaching the active species leading to a decline in the rate of inactivation (Helali et al. 2014; Benabbou et al. 2007). B. subtilis shows a high level of resistance due to the presence of a thick layer of peptidoglycan around it. The main difference between gram-positive and gram-negative bacteria arise from the cell wall and the amount of peptidoglycan membrane constituents (Al-Kobaisi, 2007). On the other hand, the different responses shown to the similar dosages of nanocomposite in the gram-negative and gram-positive bacteria might be attributed to the physiological differences, intra-bacterial metabolism and selective membrane permeability, all of which are dependent on the presence/absence of light (Felczak et al. 2012). It is worth noting that the results obtained are consistent with the findings in a further related report (Alikhani et al. 2013) inactivation rates decreased can be explained by the fact that the rise in the bacterial density prevents light from penetrating into the surface of the nanocomposite followed by a sharp decline in the production of active radicals (Widi et al. 2018). On the other hand, under the constant dosage of nanocomposite, increasing the bacterial density led to a drop in the number of radicals produced resulting in reduction of the inactivation rate (Zhan et al. 2014). Our findings are in agreement with other related reports in the literature (Wang et al. 2015a).

Compared to the Vis light, the application of UV light yeilded higher inactivation rates for both bacteria the UV light activates the radicals, and destroys the cellular and enzymatic structure. Furthermore, In the dark condition, the inactivation rates for the E. coli were determined to be 12.6 and 21.73%, whilst the corresponding values for the B. subtilis, were 13.63 and 17.66%, respectively. The application of the current advanced oxidation process (AOP) using UV/g-C3N4/Fe3O4/Ag, which is accompanied with the production of OH° and further reactive agents, results in a higher bacterial inactivity than any single AOP alone. It should be mentioned that in a photocatalytic process, the UV-excited catalyst is responsible for the production of the most active radicals (Mansoury et al. 2015).

The results demonstrated that the Weibull model fitted best to the inactivation process. Compared with the other models, this model produced the lower RMSE and higher R2 values (Table 1). According to the Weibull model, the resistance of the individual member of population of bacteria to the inactivation process is not the same. In other words, the microbial population is of high diversity and each cell needs a specific contact time to deactivate(C Mecha et al. 2019). In a further related study, Mecha et al. reported that the Weibull model described the inactivation of the bacteria much better than the first-order kinetic model (Mecha et al. 2016). In the Weibull model, for β < 1 values an upward concave trend is observed whilst for β > 1 a downward concave trend is seen. On the other hand, the first-order kinetic is accompanied with β = 1. The results depicted in Fig. 9 reveals that the survival curve is of upward concave shape (b < 1). This implies that over the time the bacterial cells have weakened and the damage to the bacterial cells has increased (van Boekel, 2002).

Fig. 9

Experimentally Survival values (Measured) and the model-fit values (Identified) derived from the Weibull model for E. coli and B. subtilis

Mechanism of photocatalytic process

When g-C3N4 is exposed to light, the electrons move from the valence bond to the conduction bond, leaving a series of holes (h+) behind. Because Fe3O4 has a high electrical conductivity, it rapidly transfers electrons to Ag with sufficient storage capacity. It also improves the charge separation process throughout the photocatalytic system. Following that, the electrons produced react with O2 to produce O2˚−. The generated O2˚− may react with h+ to produce OH˚ radicals. The radicals generated during the photocatalytic process (including OH) can inactivate the bacteria and subsequently damage their cells via various routs such as cell membrane destruction, inactivation of enzymes and essential proteins, and damaging DNA (Hamblin and Hasan, 2004).