Novel prism shaped C3N4-doped Fe@Co3O4 nanocomposites and their dye degradation and bactericidal potential with molecular docking study

Novel prism shaped C3N4-doped Fe@Co3O4 nanocomposites were fabricated via a co-precipitation route for effective removal of organic pollutants from water and for bactericidal applications. Doping of C3N4 in the heterojunction significantly enhanced the photocatalytic and sonocatalytic activity against methylene blue ciprofloxacin (MBCF) dye. The main purpose of doping Fe atoms in the cobalt lattice was to generate crystal and surface defects. Moreover, the optimum doping amount of C3N4 for maximum degradation performance was evaluated. A detailed examination of the prepared nanocomposites was carried out systematically using various characterization tools for better understanding. HR-TEM images revealed the formation of novel prism shaped structures that exhibited outstanding degradation of the organic dye in water. Significant bactericidal potential was also observed for the synthesized nanocomposites against Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) bacteria. In silico, molecular docking studies against β-lactamase, DHFR and FabI enzymes served to elucidate the mechanism governing the bactericidal activity of the as-synthesized nanoparticles (NPs). Furthermore, a scavenging study by DPPH (2,2-diphenyl-1-picrylhydrazyl) assay and COD (chemical oxygen demand) analysis was performed in order to evaluate active species and the anti-oxidant potential of prepared composites.


Introduction
Water is the prime necessity for all life present on this planet. Water pollution, being a serious threat to life on Earth, poses a grave challenge for humanity to overcome. [1][2][3][4] About 1 billion people are affected every year by water pollution and about 1.8 million deaths were recorded in the year 2015 according to a report published in The Lancet. A major source of water pollution are the organic dyes emanating from industrial wastes which are extremely harmful due to their carcinogenic nature. 5,6 Various techniques developed by researchers in the past for water treatment are costly and energy-consuming. 7-12 A recent and cost-effective method that has emerged is the use of photocatalysts for water treatment and water splitting. [13][14][15][16][17][18][19] Apart from its numerous advantages over other techniques, a major drawback while using semiconductor photocatalysts is the corrosion and dissolution of catalysts in aqueous medium when irradiated with light. High recombination rate and agglomeration also inuence their performance as catalysts. Researchers are looking for catalysts that are more stable under light and have low recombination and agglomeration rates, which will serve to improve performance. 13 Heterojunction photocatalysts are produced to address the drawback of high recombination rates of semiconductors. [20][21][22][23][24] Heterojunctions, however, in bulk form are not effective and use of nanomaterials becomes necessary to boost their performance. Large surface area and high reactivity of nanocatalysts play an important role in photocatalysis for the removal of organic pollutants from water. 25 Among 2D nanomaterials, graphitic carbon nitride (g-C 3 N 3 ) has emerged as a strong candidate for photocatalysis, possessing mid-range band gap (2.7 eV) and tunable electronic properties. [26][27][28] However, g-C 3 N 3 suffers from high recombination rate, poor visible light absorption and small surface area. Various nanostructure designs have been synthesized to increase the performance of g-C 3 N 3 photocatalysts. [29][30][31] Production of g-C 3 N 3 based heterojunctions increases catalysts' charge transfer efficiency to improve the overall photocatalytic performance. 32,33 Moreover, addition of co-catalysts into the framework has shown to yield outstanding results by enhancing stability and activity of g-C 3 N 3 based catalysts. 34 Plenty of work has been reported recently on g-C 3 N 3 based photocatalysts embedded with various metal oxides such as ZnO, TiO 2 and WO 3 that are used as co-catalysts. [34][35][36] Among such co-catalysts, Co 3 O 4 emerged as a viable candidate for photocatalytic activity. Co 3 O 4 grasped the attention of researchers aer the work of Liao et al. on the degradation of water using Co 3 O 4 under visible light irradiation. 37 Vivid optical response under visible light and good charge transfer efficiency of Co 3 O 4 (B g ¼ 2.6 eV) makes it highly valuable. 28,38 One difficulty, however, is the severe aggregation of Co 3 O 4 nanoparticles, which causes massive reduction in photocatalytic activity. 39 To overcome this, suitable semiconductors can be doped in Co 3 O 4 or highly rened synthesis techniques could be adopted to fabricate dispersed Co 3 O 4 nanoparticles. 40 Appropriate solvents can signicantly affect the size, morphology and aggregation behavior of nanomaterials. In this regard, ethylene glycol (EG) has been reported as an effective solvent for suppressing severe aggregation of nanoparticles. Moreover, it can act as a surfactant also to regulate the uniformity of NPs diameter. 41 No work has been reported yet (to best of our knowledge) on the use of C 3 N 4 -doped Co 3 O 4 composites for dye degradation and anti-bacterial applications.
In the following work, a bottoms-up technique (coprecipitation) has been adopted to synthesize C 3 N 4 -doped Fe@Co 3 O 4 nanostructures for photocatalytic degradation of methylene blue ciprooxacin (MBCF) dye. A schematic representation of the synthetic route adopted is given in Fig. 1. Moreover, bactericidal activity of as-prepared samples was evaluated against Methicillin Resistant (MR) E. coli and S. aureus bacteria and possible mechanism for anti-microbial activity has been proposed using molecular docking studies. The docking predictions were performed against b-lactamase, DHFR and FabI enzymes belonging to cell wall, folate and fatty acid biosynthetic pathways. Structural and optical properties, chemical composition and surface morphology of various doped samples were studied. Moreover, comparison of pure Co 3 O 4 particles and doped Co 3 O 4 has also been provided for better understanding.  (Fig. 1).

Isolation and identication of S. aureus and E. coli
Sheep blood agar (5%) was utilized to culture mastitic milk samples (Bovine) collected from various veterinary (private and public) hospitals in Punjab, Pakistan. Aer incubating (for 24 Fig. 1 (a) Synthesis mechanism of C 3 N 4 . (b) Schematic representation of synthesis route adopted for fabrication of nanocomposites. hours) at 37 C, samples were further puried by streaking them in triplicate on Manitol Salt Agar (MSA) and MacConkey agar (MA) media to isolate S. aureus and E. coli effectively. Morphological study (Gram staining) and biochemical tests (Catalase and Coagulase) were employed for the identication of isolated colonies.

Antimicrobial activity
Agar well diffusion phenomenon was adopted to analyze the antimicrobial activity of as-obtained samples on isolated Gram positive and negative bacteria. Aer swabbing MSA-deposited Petri dishes with isolated S. aureus and E. coli cultures (1.5 Â 10 8 CFU mL À1 ), wells of diameter 6 mm were generated in the dishes via sterilized cork borer. Various constitutions $0.5 mg/ 50 mL and 1 mg/50 mL corresponding to low and high concentration respectively, of each sample were poured into the wells along with positive control (ciprooxacin 0.005 mg/50 mL) and negative control (DIW 50 mL). Further incubation of prepared dishes was conducted at 37 C for 24-48 hours to investigate the antibacterial activity of nanostructures. Inhibition zones formed aer 48 hours of incubation were measured using Vernier caliper (mm) and their relation with doping concentration was analyzed. One-way analysis of variance (ANOVA) was employed in order to calculate antibacterial efficacy of nanostructures in terms of inhibition zone diameter (mm).

Scavenging (DPPH assay)
The traditional DPPH scavenging assay was adopted with some modications in order to check the free radical active species and anti-oxidant behavior of the prepared nanostructures. Bare and doped Co 3 O 4 nanoparticles were utilized with various concentrations (0-500 mg mL À1 ) along with equal volume of 1 mM DPPH solution for evaluation. The reaction mixture was vortexed and incubated for 30 minutes in dark at ambient temperature. For reference sample, standard solution of ascorbic acid was used as strong anti-oxidant. The % scavenging activity of each sample was evaluated by measuring the degradation in maximum absorbance wavelength of DPPH solution (l ¼ 517 nm) as given in eqn (1).
where, A 0 is absorption of control (methanol + DPPH) and A 1 corresponds to the absorbance value in the presence of sample aer some time.

COD analysis
Chemical oxygen demand (COD) analysis is another important factor that decides the amount of oxygen required for complete oxidation of organic pollutant in wastewater. The conventional ferrous ammonium sulfate (FAS) assay was adopted in order to evaluate the COD of dye solution before and aer degradation under visible light irradiation. The prepared sample solutions were reuxed with standard solution of potassium dichromate (K 2 Cr 2 O 7 ) and sulfuric acid (H 2 SO 4 ) and excess of K 2 Cr 2 O 7 was titrated against FAS to check the COD values. Total amount of K 2 Cr 2 O 7 consumed in the procedure offers the extent of oxygen consumed by the solution for complete oxidation of organic pollutants. The following relation was used to calculate the COD values at various irradiation time (eqn (2)); where, A ¼ volume of FAS used for blank (mL), B ¼ volume of FAS used for sample (mL), M ¼ molarity of FAS, V ¼ volume of sample.

Molecular docking studies
The promising antibacterial potential of C 3 N 4 -doped Fe@Co 3 O 4 NPs revealed by in vitro studies prompted us to perform in silico molecular docking studies to identify possible interactions of NPs against selected enzyme targets. Here, we performed molecular docking predictions of NPs against key enzymes of cell wall biosynthetic pathway, folate biosynthesis and fatty acid biosynthetic pathway i.e. b-lactamase, dihydrofolate reductase (DHFR) and enoyl-[acyl-carrier-protein] reductase (FabI), respectively. The 3D-structural coordinates of b-lactamase, DHFR and FabI for S. aureus were retrieved from protein data bank having ID as 1MWU (resolution: 2.60Å), 42 2W9H (resolution: 1.48Å) 43 and 6TBC (resolution: 2.55Å), 44 respectively. The docking predictions were performed using molecular operating environment (MOE) soware 45 according to the method reported in our previous studies. 46,47 The main steps involved were removal of water and native ligand followed by addition of H-atoms and ultimately energy minimization of retrieved enzyme structures. Default parameters of energy minimization algorithm (MMFF94x force eld, gradient: 0.05) of MOE were utilized for energy minimization. Later, systematic conformational search was employed on these selected enzymes using default parameters of Site nder tool i.e. RMS gradient of 0.001 kcal mol À1 and binding pocket was specied within close vicinity (i.e. 10Å) of native ligand. Finally, 10 top scored docking conformations were generated in each case and the best-performing complexes were further evaluated for binding tendency prediction of NPs. Pymol soware was employed for analysis of interactions and 3D view representation of docked complexes. Ligand structures were prepared using builder tool of MOE soware.

Materials characterization
X-ray powder diffraction (XRD) measurements were carried out by collecting data from 5 to 80 (2q range) using a PANalytical-Xpert-PRO diffractometer with Cu-Ka radiation of l ¼ 1.5418Å to ascertain the crystallite size and phase constitution of nanomaterials. FTIR (Fourier Transform Infrared) spectra with PerkinElmer spectroscopy was employed to detect the presence of functional groups in undoped and co-doped CuO samples while Raman spectra were obtained through a Raman Ther-moscientic microscope equipped with 532 nm laser (6 mW). To observe the optical properties, UV-vis (ultraviolet-visible) and photoluminescence (PL) pictures of the samples were recorded via UV-vis (Genesys10S spectrophotometer) and PL analyzer (JASCO, FP-8300), respectively. Using INCA EDS soware, elemental composition was obtained through energy dispersive X-ray spectroscopy. Inter-planner d-spacing of the synthesized products were measured using HR-TEM equipment JEOL JEM 2100F.

Results and discussion
Specimens were characterized for structural and phase composition analysis with XRD ( Fig. 2a 48 This slight change in lattice parameters of Fe@Co 3 O 4 could be assigned to the difference in the atomic radii of Co (0.125 nm) and Fe (0.126 nm) or some intrinsic crystal defects. 49 However, this conrms successful incorporation of Fe into crystal lattice. 50,51 No signicant change in peaks position and intensity was observed upon C 3 N 4 doping into the composite. This indicates that the two-dimensional C 3 N 4 was not incorporated in the cobalt lattice and caused no signicant structural changes. It may have covered the Fe@Co 3 O 4 nanoparticles or served as host matrix for particle decoration/incorporation. Furthermore, no characteristic peak of C 3 N 4 appeared in any of the doped samples' spectra which is attributable to its relatively low content in the matrix. 52 Chemical composition and functional group identication of samples was investigated via non-destructive FTIR analysis (Fig. 2b). Spectra were observed at room temperature in the frequency range of 500-4000 cm À1 and two noteworthy peaks were observed at 546 cm À1 and 645 cm À1 for Co 3 O 4 nanostructures. The peaks attributed to 546 cm À1 and 645 cm À1 depict the stretching of Co-O bond and bridging vibration of O-Co-O bond, respectively. 53-55 Broad bands near 3412 cm À1 and 1420 cm À1 refer to the stretching and bending vibrations of -OH group respectively, which originate from the moist environment in which samples were prepared. Bands appearing near 1428 cm À1 are ascribed to CO 3 vibration in carbonate anions that might have originated due to the precursors used during synthesis (Co(CH 3 CO 3 O 4 ) 2 $4H 2 O). [56][57][58] Selected area electron diffraction (SAED) patterns for Co 3 O 4 and C 3 N 4 (50 mg, 100 mg) are shown in Fig. 2c-e. Images with concentric rings indexed with (111), (220), (222), (311) and (400) planes, are in agreement with XRD results thus conrming the FCC cubic crystal structure of Co 3 O 4 nanostructures. Interlayer spacing for Co 3 O 4 (d ¼ 0.247) was calculated from most intense peak (311) and veried by TEM observation using Digital Micrograph (Fig. 4).
Optical absorption spectra was investigated via UV-Vis spectrometer at room temperature using dispersed solutions of nanocatalysts in water (Fig. 3a). Two distinct absorption windows were observed for Co 3 O 4 around 430 nm and 750 nm that corresponded to charge transfer between metal and ligands. First absorption peak (l < 500 nm) is ascribed to O 2 À to Co 2+ charge transfer while second absorption window (l > 700 nm) refers to O 2 À to Co 3+ charge transfer. 59,60 Such charge transfer is ascribed to d (t 2g ) to d (t 2 ) electronic transition from    data has also been used to calculate band gap energies of specimens using the classical Tauc transformation for near edge absorption. Aer extrapolating linear portion of (ahn) 2 vs. hn plot on x-axis (Fig. 3b), allowed direct band gap for Co 3 O 4 was calculated to be 2.5 eV which is in agreement with reported literature. 62,63 Trivial change in band gap energies for Fe@Co 3 O 4 was observed for samples with increasing amount of C 3 N 4 dopant concentration $2.6 eV for C 3 N 4 (3%).
In order to study the change in electron transfer efficiency and recombination rate, PL spectra for all samples were examined (Fig. 3c). Co 3 O 4 nanostructures showed characteristic PL band from 410 to 425 nm (l exc ¼ 280 nm) with noteworthy decrease in peak intensity upon introduction of Fe. A substantial drop in peak intensity, which corresponds to low recombination rate and high electron transfer efficiency, was detected for low amount of doped-C 3 N 4 in the composite as well. Reduction in intensity might be due to the defects and trap sites introduced into the structure upon doping with C 3 N 4 . This can be understood by SRH (Defect Assisted) recombination process explained by Shockley, Read and Hall in 1952. 64 Generation of new energy levels in the forbidden band gap of Fe@Co 3 O 4 aer doping with C 3 N 4 act as trap sites for electrons and holes, hence reducing recombination rate. Further increase in doping concentration (C 3 N 4 2, 3%) caused an increase in PL intensity indicating reduction in electron transfer efficiency. This may be attributed to substantial agglomeration of particles or formation of 2d sheets of C 3 N 4 over the surface of nanostructures.
Room temperature Raman spectroscopy was utilized to further investigate phase composition and structural defects present in prepared nanostructures (Fig. 3d). Characteristic bands observed at 517 cm À1 and 680 cm À1 were attributed to F 2 2g and A 1g Raman-active modes of Co 3 O 4 nanoparticles. This conrmed the cubic spinel structure of Co 3 O 4 belonging to Fd 3m group with Co 3+ and Co 2+ atoms at tetrahedral and octahedral sites, respectively. 65 Two bands near 1365 cm À1 and 1580 cm À1 are respectively referred to as D-band (defective band) and G-band of carbon containing materials. The D-band (sp 2 Raman signature) which arises due to TO (transverse optical) phonons near k-point indicates the defects present in carbon hexagons breathing modes. On the other hand, G-band is the consequence of E 2g phonons at the center of Brillouin zone, indicating in-plane C-C stretching (double-degenerated) mode which is characteristic Sp 2 Raman signature of carbon based materials. 66,67 The ratio of G and D bands intensity (I D /I G ) can be used to evaluate the extent of disorder present in carbonaceous materials. Minimum I D /I G ratio (0.99) was calculated for C 3 N 3 (3%) conrming the least amount of defects.
Morphological and structural study of pure and doped samples was carried out using HR-TEM images up to 10 nm resolution (Fig. 4). As expected, pure Co 3 O 4 depicted aggregated structure with particles of size less than 15 nm (Fig. 4a). Prism shaped particles of size 25-35 nm were observed upon 1% doping with C 3 N 4 as shown in Fig. 4b. These prism-like structures exhibit improved catalytic performance by offering large surface area and increased number of active sites. For higher dopant concentration of C 3 N 4 , TEM images (Fig. 4c) revealed aggregated Fe@Co 3 O 4 nanoparticles uniformly decorated on C 3 N 4 sheets, hence con-rming the presence of heterojunction between C 3 N 4 and Fe@Co 3 O 4 . 28 Highly magnied images (10 nm) were used to show lattice fringes for identication of crystallographic planes to calculate d-spacing. The d-spacing calculated for Co 3 O 4 came out to be $0.243 nm that is consistent with theoretical d-spacing of (111) crystallographic plane of Co 3 O 4 , Fig. 4d-f. Electron dispersive X-ray spectroscopy (EDS) was carried out for elemental composition tracing of obtained nanostructures as depicted in Fig. 5a-f. Strong peaks of Co, Fe and O conrmed the presence of Fe@Co 3 O 4 composite while additional peaks of C and N assured the successful incorporation of C 3 N 4 into crystal lattice. Small amount of Na was detected which might have originated due to the use of NaOH to maintain pH of samples during synthesis.
The magnetic properties of bare and doped Co 3 O 4 materials have been previously reported in the literature. In a typical Co 3 O 4 spinel structure, the magnetic moment mainly arises due to Co 2+ species while Co 3+ ions do not exhibit any permanent magnetic moment. This is mainly owed to the fully lled t 2g orbitals arising from splitting of 3d orbitals in octahedral crystal eld. 68 Furthermore, the bulk state Co 3 O 4 structures exhibit antiferromagnetic behavior, but the nanostructures of cobalt have been reported to possess a slight ferromagnetic nature due to quantum size effects and uncompensated spins. 69 In addition to that, Co 3 O 4 nanorods demonstrate higher coercivity values in contrast to nanoparticles as well, which is owed to the shape anisotropy. 70 Studies have also revealed that doping of Fe also induces a slight ferromagnetism in Co 3 O 4 nanostructures. 71

Photocatalytic activity
Photocatalytic activity of as-obtained samples was measured under visible light irradiation (Fig. 6). Time dependency of absorbance was plotted for solutions containing 10 mg of samples in 40 mL of MBCF. Furthermore, the pH of the due solution was measured to be 6.1 and the evaluations were conducted in this slightly acidic pH. The solution was placed under visible light irradiation and 3 mL of dye was syringed out at regular intervals of time for UV-Vis analysis. Mercury lamp with 400 W intensity and wavelength ranging from 400-700 nm was employed as the irradiation source. Reduction in absorption maxima (l max ¼ 665 nm) was observed which evidenced the reduction of dye signicantly. Maximum degradation of dye was observed just within 40 minutes for all nanocatalysts. Co 3 O 4 showed 81% degradation of dye aer 40 minutes of light irradiation and upon Fe doping, degradation performance signicantly increased to 93% for the same duration. Upon 1% doping of C 3 N 4 in Fe@Co 3 O 4 , remarkable increase in photocatalytic activity was observed with maximum degradation of 99% noted within 40-45 minutes. Decrease in dye degradation performance was observed for higher doping concentration of C 3 N 4 (2 and 3%) implying that the optimum doping concentration of C 3 N 4 in Fe@Co 3 O 4 was 1% for best photocatalytic performance. Moreover, rate constants were calculated by applying pseudo rst order kinetics to linearly tted curves of ln(C o /C t ) vs. time plot from eqn (3) as follows: Increase in photocatalytic performance of Co 3 O 4 upon doping of Fe and C 3 N 4 can be attributed to reduced recombination rate of photo-induced charge carriers as depicted by PL spectroscopy (Fig. 3c). Apart from that, degradation performance majorly depends upon size, shape and surface area of nanocatalysts. Particles with large surface area provide greater number of active sites for atoms or molecules hence increasing the number of redox reactions, leading to degradation of MB to LMB. The photo-generated electrons and holes interact with surrounding oxygen (O 2 ) and water (H 2 O) molecules to form highly reactive cO 2 À , cOH À and H + radical species, respectively.
The excessive H + ions and electrons combine with dye molecules leading to the reduction of MB to LMB. 72 Addition of Fe atoms into Co 3 O 4 boasts the electron transfer efficiency of nanocatalysts by introducing defects in the lattice. This defect-assisted recombination phenomenon has been previously discussed in PL spectra analysis where defects introduced in the lattice act as trap sites for electrons and holes rendering their recombination. 64 Doping of C 3 N 4 into Fe@Co 3 O 4 also compliment the above-mentioned phenomenon but only to a certain extent of doping concentration. Further increase in doping amount reduces the catalytic activity of nanostructures which may be attributed to agglomeration of particles or the predominance of 2D sheet-like structure of C 3 N 4 enclosing the particles within it.
A possible mechanism for the photocatalytic degradation of MB over the prepared nanocatalysts is proposed in Fig. 7. The Fe@Co 3 O 4 exhibits a lower VB (valence band) and a higher CB (conduction band) values in contrast to C 3 N 4 sheets which signicantly result in better separation of photogenic charge carriers. As a consequence, electrons from CB of Fe@Co 3 O 4 composite jump to the CB of C 3 N 4 while the holes in the VB of C 3 N 4 get transferred to Fe@Co 3 O 4 atoms. This favors a type-II carrier transfer in the heterojunction resulting in efficient separation of photogenic excitons. The E CB of doped C 3 N 4 is more negative in contrast to E 0 of O 2 /cO 2 À (À0.33 eV). Hence, the electrons in CB of C 3 N 4 signicantly reduce the O 2 species into reactive radical cO 2 À species whereas the holes in the VB interact with surrounding H 2 O molecules and oxidize them to form radical cOH À species. These radical species interact with the surrounding dye molecules and complete degradation of organic pollutant occurs. The MB dye being cationic in nature get reduced by accepting electrons from these highly reactive species. Consequently, p-conjugation occurs aer addition of electrons which results in the formation of Leuco-MB which is colorless in nature.

Sonocatalytic activity
Sonocatalytic activity of nanocatalysts was also observed by sonicating the sample solution (10 mg of nanocatalyst in 40 mL of MBCF) in ultrasonic bath for at least 80 minutes in the absence of light (Fig. 8). Similarly, 3 mL of sample were taken for UV-Vis analysis at regular interval of time and degradation of dye was evaluated. Overall, sonocatalytic performance was found to be markedly better than photocatalytic activity for all samples. The maximum degradation for Co 3 O 4 were measured to be 95% while Fe@Co 3 O 4 showed a degradation of 98% aer 40-50 minutes of sonication. Maximum degradation of 99% was observed for C 3 N 4 (1%) whereas, decrease in degradation performance was detected for samples with higher concentration of C 3 N 4 (2, 3%). The mechanism involved in degradation of dye in the abovementioned procedure can be explained via hot spot theory which deals with the formation of microbubbles, their nucleation, growth and collapse to produce heat and light (sonoluminescence). 73 Ultrasonication helps in catalytic activity of nanocatalysts by following acoustic cavitation mechanism where  large number of hot spots are produced aer extensive energy accumulation. 74 The compressing and expanding ultrasonic waves cause change in bubble size (i.e., grow, compress) resulting in implosive collapse under extreme conditions. This implosion releases adequate amount of energy (heat and light) to generate electron hole pairs on the surface of nanocatalysts. These electrons and holes take part in redox reactions where electrons act as strong reducing agents while holes behave as strong oxidizing agent. Another consequence of this released energy is the breaking of surrounding water molecules to produce highly reactive radical H + , OH À and O À species. Addition of nanocatalyst in solution complements the formation of microbubbles while the number of active sites is signicantly increased. 75

Photo-sono catalytic activity
In order to check the combined effect of sonocatalysis and photocatalysis, samples were placed in sonicator under visible light irradiation for at least 80 minutes (Fig. 9). Surprisingly, reduction in catalytic activity was observed for almost all samples in photosono catalysis with Co 3 O 4 showing maximum degradation of 67% in 80 minutes. Decrease in activity was observed for Fe@ Co 3 O 4 as well with maximum degradation of 88% while C 3 N 4 (1%) gave a consistent degradation of 99% within 60 minutes. Samples with higher doping concentration, C 3 N 4 (2%) and C 3 N 4 (3%) showed degradation up to 58% and 47%, respectively.
In addition to above, COD analysis was performed to further evaluate the best-performing photocatalyst C 3 N 4 (1%) in dye solution under visible light irradiation in order to check the oxygen demand before and aer degradation of the dye (Table 1). For evaluation, 10 mg of photocatalyst in 1000 mL of dye solution was placed under visible light for 80 minutes and 1 mL was syringed out at regular intervals to check the COD values. The blank sample with no photocatalyst showed a high COD value of 1865 mg L À1 which gradually decreased aer light exposure in the presence of catalyst. Maximum COD degradation of up to 95% was achieved aer 60 minutes of dye degradation over C 3 N 4 (1%) catalyst. These results show that the prepared photocatalysts can act as potential candidates for pre-treatment of polluted water for effective oxidation of organic pollutants.

Scavenging (DPPH) assay
In order to evaluate the active radical species present in the photocatalyst and to measure their anti-oxidant activities, DPPH scavenging assay was employed. Antioxidant characteristics of compounds is tied to their electron or hydrogen atom donating capability to DPPH free radical, such that they create stable diamagnetic compounds. This DPPH free radical's reduction capability can be examined spectrophotometrically by evaluating the degradation in absorbance at 517 nm. A dose dependent behavior was observed for the anti-oxidant activities of all prepared samples (Fig. 10). The maximum scavenging performance of 85.56% was exhibited by pristine Co 3 O 4 aer 120 minutes at concentration of 100 mg mL À1 . Generation of highly reactive cOH and cO 2 À radical species, which have the potential to bond with the DPPH free radical can result in its degradation. Upon Fe addition, a signicant drop in antioxidant activity up to 61.82% was observed at the same concentration. Further decrease in scavenging performance was   (Fig. 11).
Bactericidal action of nanostructures is attributed to the formation of reactive oxygen species (ROS) and ionic radical species that interact with cell walls and DNA structure of bacteria. Reactive oxygen species (OH À and H 2 O 2 ) are formed in the presence of nano-catalysts under UV-Vis irradiation (electron-hole generation), which induce oxidative stress on the cell walls of bacteria. Another possibility involves the interaction of positively charged Co 2+ and Co 3+ with negatively charged cell membrane, resulting in cell death.
Role of nanoparticles as good antibacterial agents is welldocumented. Although various mechanisms such as disruption of cell wall synthesis and inhibition of key enzymes of metabolic pathways have been suggested as possible mechanisms governing bactericidal activity of NPs, however exact mechanism still needs to be explored. As key enzymes from cell wall biosynthetic pathway, 77,78 folate biosynthesis 79,80 and fatty acid biosynthetic pathway have been reported as attractive target for antibiotic discovery, here we performed molecular docking predictions against enzymes of these selected pathways. Therefore, the binding tendency of C 3 N 4 -doped Fe@Co 3 O 4 NPs was evaluated against b-lactamase, DHFR and FabI enzymes to analyze their role as inhibitor of these selected targets.
In the case of b-lactamase S. aureus , the best-docked complex obtained revealed H-bonding interaction of NPs with two amino acid residues i.e. Asn464 (2.2Å and 2.3Å) and Gln521 (2.4Å) alongside metal-contact interaction with Thr600 with the overall binding score of À7.833 kcal mol À1 as depicted in Fig. 12a and b.
On the other hand, the best-docked conformation of C 3 N 4doped Fe@Co 3 O 4 NPs against active site of DHFR S. aureus showed involvement of three amino acid residues of binding pocket in Hbond interaction like Ser49 (2.5Å), Asn18 (3.4Å) and Ile14 (2.5Å) with binding score of À9.418 kcal mol À1 . Similarly, metal-acceptor bond with Ser49 was also observed as shown in Fig. 12c and d. In addition, the docking predictions of C 3 N 4 -doped Fe@Co 3 O 4 NPs in the case of FabI S. aureus showed H-bonding interaction with Ser44 (2.5Å and 2.8Å) and Arg40 (2.1Å) as depicted in Fig. 13a and b with binding score of À5.893 kcal mol À1 .
In silico molecular docking predictions showed good agreement with in vitro antibacterial activity of C 3 N 4 -doped Fe@Co 3 O 4 NPs against S. aureus and suggested inhibition of these selected enzyme as a possible mechanism.

Conclusion
Cost-effective co-precipitation route was adopted to fabricate novel prism-shaped C 3 N 4 -doped Fe@Co 3 O 4 nanocomposites. Photocatalytic, sonocatalytic and photo-sonocatalytic potentials of synthesized catalysts were evaluated against MBCF dye. An outstanding degradation of >99% was observed for the novel heterojunction framework which followed the type-II transfer mechanism of photogenic excitons. The enhanced performance of the catalyst was attributed to the large surface area, enhanced  visible light harvesting capability of Co 3 O 4 atoms and effective separation of photogenic charge carriers due to the addition of well-matched energy levels of C 3 N 4 . Moreover, the photocatalytic and sonocatalytic activity of prepared catalysts was controlled by tuning the doping concentration of C 3 N 4 and an optimum doping amount was proposed. In addition to above, the prepared samples also depicted noteworthy bactericidal efficacy against S. aureus bacteria where inhibition of b-lactamase, DHFR and FabI enzyme of S. aureus was suggested as possible mechanism behind their bactericidal potential. Overall, this study offers new insights into the use of cobalt-based heterojunction photocatalysts for various photocatalytic, sonocatalytic and antibacterial applications.

Funding
Authors are thankful to higher education commission (HEC), Pakistan through project 21-1669/SRGP/R&D/HEC/2017. Support provided by the Core Research Facilities at King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia is highly appreciated.

Availability of data and materials
All data are fully available without restriction.

Author contributions
SOAA performed the whole experiments and wrote the manuscript. MI and M. Imran provided the novel idea to carry out the experiment. AH performed antimicrobial and participated in the data analysis of the results and discussion portion. S. Naz and J. Haider performed molecular docking study. A Shahzadi worked on schematic diagram and reviewed the manuscript. AUH carried out the FESEM and HRTEM analysis. All authors read and approved the nal manuscript.

Conflicts of interest
Authors conrm no conict of interest.