Multifunctional Core-Shell NiFe2O4 Shield with TiO2/rGO Nanostructures for Biomedical and Environmental Applications

Multifunctional core@shell nanoparticles have been synthesized in this paper through 3 stages: NiFe2O4 nanoparticles by microwave irradiation using Pedalium murex leaf extract as a fuel, core@shell NiFe2O4@TiO2 nanoparticles by sol-gel, and NiFe2O4@TiO2@rGO by sol-gel using preprepared reduced graphene oxide obtained by modified Hummer's method. XRD analysis confirmed the presence of both cubic NiFe2O4 spinel and tetragonal TiO2 rutile phases, while Raman spectroscopy analysis displays both D and G bands (ID/IG = 1.04) associated with rGO. Morphological observations by HRTEM reveal a core-shell nanostructure formed by NiFe2O4 core as confirmed by SAED with subsequent thin layers of TiO2 and rGO. Magnetic measurements show a ferromagnetic behavior, where the saturation magnetization drops drastically from 45 emu/g for NiFe2O4 to 15 emu/g after TiO2 and rGO nonmagnetic bilayers coating. The as-fabricated multifunctional core@shell nanostructures demonstrate tunable self-heating characteristics: rise of temperature and specific absorption rate in the range of ΔT = 3–10°C and SAR = 3–58 W/g, respectively. This effectiveness is much close to the threshold temperature of hyperthermia (45°C), and the zones of inhibition show the better effective antibacterial activity of NTG against various Gram-positive and Gram-negative bacterial strains besides simultaneous good efficient, stable, and removable sonophotocatalyst toward the TC degradation.


Introduction
In the modern research world, multifunctional nanoparticles (NPs) have attracted much attention with numerous fascinating properties and potential diverse applications in optics, magnetic, electronics, and catalysis. Single functional NPs must overcome the requirements for numerous applications such as water purification [1], biosensing [2], and cancer treatment [3,4]. In this context, the proposed solution consists of a combination of two or more different functional nanomaterials to form multifunctional nanoparticles [5,6]. Nowadays, the whole world is facing human health issues and environmental pollution. Generally, we are focusing on human health issues more than environmental problems because pharmaceutical companies synthesize more and more medicinal products; this medicinal side products and wastes are also one of the causes of environmental pollution. Hence, the product should have human therapeutic nature as well as environmental pollutiontreated nature. It is very essential to this world. Specifically, core@shell NPs have attracted great attention due to the distinctive properties originating from a well-engineered and good functional choice of materials; thereby, the fabricated nanostructures demonstrate promising novel applications [7,8].
Hyperthermia treatment has been identified as one of the potential cancer treatments, which depends on raising the temperature of the tumor cells up to around 42-47°C, which is enough to destroy the cancer cells if applied for at least 10 minutes [9]. Magnetic NPs are good functional materials for hyperthermia cancer treatment due to their biocompatibility and nontoxicity and can be remotely heated by applying an external magnetic field [10].
is heating ability has also been used in different biomedical applications, such as magnetic separation and target drug delivery [11,12]. e induction heating ability of NPs is usually estimated by the specific absorption rate (SAR) within the medium.
Normally, the huge global antibiotic usage per year is near 200,000 tons of medical compounds. Tetracycline is one of the most required antibiotics. e discharge of antibiotics from hospitals' effluents, municipal wastewaters, and drugs' industrial drainage have been identified as very polluting to the natural environment and ecosystems. Antibiotics possess a low biodegradability, high toxicity, and high-water solubility in nature and hence can be easily absorbed by humans as well as animals [13]. Even at low concentrations of antibiotics, their discharge into the environment leads to teratogenic problems in pregnant women, function disorders of endocrine glands, and chronic toxicity [14,15]. Hence, it can produce antibiotic resistance genes and bacteria. A variety of methods have been adopted for the antibiotics' removal from water bodies. In the photocatalytic process, renewable sources such as solar energy utilize visible light photocatalysis, and ultrasonic irradiation is the most important and efficient method. However, the sonophotocatalytic approach by coupling the two processes boosted the antibiotics' degradation [16][17][18]. e magnetic energy of magnetic NPs, which is transformed into thermal energy, shows promising application behavior in hyperthermia treatment [19]. Magnetic transition metal ferrites with a general formula MFe 2 O 4 (M � Ni, Co, etc.) can be written based on ions distribution A 3+ [B 2+ , B 3+ ]O 4 2− , where tetrahedral cation sites are denoted by "A" and octahedral cation sites are denoted by "B." e Fe 3+ cations occupy half of the tetrahedral A-sites, and the rest of the Fe 3+ cations and divalent M 2+ cations occupy the octahedral site in the spinel structure, resulting in improved conductivity. Among the different metal ferrites, nickel ferrite (NiFe 2 O 4 ) exhibits soft magnetic semiconducting properties with ferromagnetic behavior, better chemical stability, low conductivity, and catalytic performance [20,21]. ey are used in many fields such as catalyst carrier, rectangular hysteresis material, microwave absorbing, and supercapacitor. Magnetic NPs are aggregating in nature because of their high surface energy and strong van der Waals interactions. e formation of a shell belonging to another functional material on a spinel ferrite surface has been proposed as an alternative approach for better control of NPs aggregation and novel functional abilities toward a broad range of applications [22]. TiO 2 is one of the remarkable functional materials used in photocatalysis, photovoltaic cells, photodegradation, electrochromic devices, white pigment in paint, cosmetics, and food coloring due to its strong optical absorption, high chemical stability, nontoxicity, and low cost [23][24][25]. Among all applications, TiO 2 has been undoubtedly recognized as a promising functional material for biomedical applications. Hence, it may play an important role in the development of healthcare, mainly for cancer treatment. TiO 2 has excellent photocatalytic behavior because of its instance photoexcitation which determines the capability to kill cancer cells. e photoexcited activity of TiO 2 has determined the capability to destroy cancer cells effectively. TiO 2 possesses a large bandgap energy (∼3 eV), and because of that, it becomes active only under ultraviolet irradiation [26,27]. Doping with another element has been reported to be one of the most effective methods for turning over new electronic structures and heteroatomic surface structures, consequently resulting in the enhancement of its photocatalysis efficiency in solar light irradiation. TiO 2 also acts as a functional material for photodegradation as well as photodynamic cancer therapy [28][29][30].
Reduced graphene oxide (rGO) is an exciting material with excellent thermal stability and electrical conductivity as well as high surface area and thereby has been used in numerous applications, including batteries, photodetectors, and sensors [31][32][33]. Metal and ferrite-based rGO materials have increased the ability of rGO material toward other fields such as energy harvesting, drug delivery, wastewater contaminants removal, antibacterial efficiency, and cancer therapy [34,35].
Several techniques have been adopted for the synthesis of metal ferrite NPs such as electrochemical, chemical, photochemical, and biological methods [36,37], wherein some are still carrying on use (with help) of hazardous chemicals and complex reaction conditions. Chemical synthesis methods are the most common to synthesize nanoparticles explicitly due to the short reaction time [38][39][40], where hazardous chemicals are used as reducing and capping agents that produce environmental pollution. Modern research focuses on the synthesis of metal ferrite NPs using green synthesis because of its eco-friendly nature, simplicity, and low cost. Green synthesis draws the great inspiration from nature plants and microorganisms of yeast, bacteria, and fungi [41][42][43]. e different parts of plant extracts help to synthesize metal ferrites via different chemical methods as a result reducing the risk of hazardous chemicals. e plant extracts can produce a variety of metabolites with amino acids, phenols, vitamins, polysaccharides, and carbohydrates, which can react as reducing and capping agents in addition to serving as fuel. Nowadays, many researches report on the green synthesis of NPs using plant extracts as stabilizing and reducing agents [44,45], even though this method takes a longer reaction time than chemical synthesis. Microwave irradiation method helps to overcome this problem and offers several advantages, including short reaction time under controlled environment, high purity product, uniform heating, and high yield [46,47]. e rapid uniform heat production and the energy can interfere the sample with the molecular motion. e generated enormous heat originates through molecular collisions, while its uniform distribution is transferred to the interior of the material resulting in an exothermic reaction with the evolution of gases to produce nanoparticles [48].
Pedalium murex is of the family of Pedaliaceae generally known as Gokshru presented throughout the world especially in India, Pakistan, and Sri Lanka. is plant was used as medicine in India, including analgesic, kidney stone treatment, stomachache, urinary retention, and intestinal infections. e leaves and roots extract of this plant are used as an aphrodisiac, antimicrobial, antioxidant, antibacterial, and neuroprotection [49][50][51][52][53].
Numerous methods have been adopted to prepare core@ shell NPs such as electrostatic layer-by-layer, hydrolysis, and hydrothermal methods. Nevertheless, the sol-gel method is the most frequently used because of its simplicity and low cost in preparing different and stable core@shell NPs [54,55]. In this paper, NiFe 2 O 4 @TiO 2 @rGO magnetic core@ shell NPs have been successfully prepared by the sol-gel method, considering NiFe 2 O 4 as a core and the fact that its outer layer is functionalized simultaneously by TiO 2 and rGO. Structural, morphological, optical, and magnetic properties have been investigated. e synthesized samples have been studied in hyperthermia and antibacterial treatment as well as environmental pollution of TC degradation in water bodies.

Preparation of Graphene Oxide (GO).
Using the modified Hummer methods, GO was synthesized. Under vigorous stirring in an ice bath, 1 g of natural graphite powder was added to 30 ml conc. H 2 SO 4 , and the temperature was maintained at 20°C. en, 6 g of KMnO 4 was added slowly with regular intervals of time for 16 h, and the as-obtained mixture was stirred at room temperature. About 500 ml of ice-cold water was slowly added to the mixture for the endorsement of the complete oxidation, where the color changed from murky brown to dark brown. For the change of color from dark brown to yellowish-brown, 5 ml of 30% H 2 O 2 was added. With 200 ml of 1 : 10 HCl aqueous solution, the graphite oxide solution was washed numerous times with DIW to remove any residual metal ions. Finally, the graphite oxide solution in DIW was subjected to sonication, centrifugation at 6000 rpm, and ultrasonication for 30 min in order to remove the multilayered graphite oxide.

Core@Shell NiFe2O4@TiO2@rGO
NPs. In the previously prepared NiFe 2 O 4 @TiO 2 homogenous solution, the prepared 0.1 mol liquid GO was added slowly dropwise under continuous stirring. For 2 days at room temperature, the obtained colloidal solution was dried and then calcined at 600°C for 3 h in a muffle furnace. us, the final powder obtained was NiFe 2 O 4 @TiO 2 @rGO.

Characterizations.
e structural analysis by X-ray diffraction (XRD) was performed using a high-resolution Rigaku Ultima IV diffractometer equipped with a CuKα radiation source (λ �1.5418Å). Rietveld refinement was analyzed by the PDXL program. FTIR was studied using a 4000-400 cm −1 range of Perkin Elmer infrared spectrophotometer. HR-SEM analysis was carried out by VEGA 3 TESCAN, and EDX was fiend out by Bruker Nano. Morphological observations were checked by high-resolution transmission electron microscopy (TEM) using JEOL 2000 EX operating at 200 kV. Raman spectra were recorded under 633 nm laser excitation using a Raman spectrometer. Magnetic measurements were performed at room temperature by using a vibrating sample magnetometer (VSM) PMC Micro Mag 3900 equipped with 1 Tesla magnet and a resolution of 0.5 μemu.
2.5. Self-Heating Characteristics. Self-heating capacity of the as-prepared NPs was tested using Magnetherm from Nanotherics Ltd. e temperature versus time curves for all samples was recorded under an alternating current (AC) magnetic field of 32.5 mT and a frequency of 622 kHz. For each sample, three different concentrations were applied (5, 10, and 20 mg/mL). All powders were dissolved in distilled water by using pulsed ultrasound vibrations for 10 min. e experiment was conducted starting from the room temperature for all samples (room temperature).
To further evaluate the self-heating characteristics of the as-prepared nanoparticles, the specific absorption rate (SAR) was calculated by using the following equation [50]: where C is the specific heat capacity of water (4.180 (J/ (g.C))); ∆T/∆t represents the initial slope of the temperature change within the first minute; m s and m w are the mass of the sample and water, respectively.

Sonocatalytic, Photocatalytic, and Sonophotocatalytic
Activities. e sonocatalytic, visible photocatalytic, and sonophotocatalytic degradation of TC antibiotics was investigated by using NiFe 2 O 4 (NiF), NiFe 2 O 4 @TiO 2 (NiT), and NiFe 2 O 4 @TiO 2 @rGO (NiTG) core@shell nanostructures as catalysts. e 30 mg of the catalyst was dissolved separately in the 100 ml of 20 mg/L TC. Prior to the degradation experiments, the mixture was stirred in a dark condition for 30 min to attain the adsorption and desorption equilibrium of TC on the photocatalyst.
In the sonocatalytic degradation, the reaction mixture was taken in the beaker, and then the 2/3rd portion was immersed in the ultrasonic bath and sonicated with 40 kHz frequency. During the reaction, the ultrasonic bath temperature was maintained at 25 ± 5°C by adding ice cubes and exchanging water frequently. e sample was collected every 15 min for further UV-visible characterization.
In the photocatalytic process, a 150 W Halogen lamp was used as visible light in the degradation of TC. e reaction mixture was introduced into the quartz tube placed at 8 cm from the Halogen lamp. During the degradation, the temperature was controlled with a continuous water flow. During the irradiation reaction, 2 ml of solution was collected and then characterized by using UV-visible spectroscopy to estimate the rate of TC degradation.
In the sonophotocatalytic method, the degradation of TC was carried out with the help of both ultrasonic and visible light irradiation. e reaction mixture was placed in the ultrasonic bath with visible light irradiation, and the exchange of ice water controlled the reaction temperature. Every 15 min, a portion of the solution was collected for UV-visible characterization. e following equation was used to determine the degradation efficiency: where C 0 is the initial concentration of the antibiotic and C t is its concentration after a certain time.

Antibacterial Studies.
e antibacterial activity of the asprepared NiF, NiFT, and NiFTG nanostructures was tested against both Gram-positive and Gram-negative bacterial strains using the agar diffusion method. e bacterial cells were cultured on solidified nutrient agar in Petri plates, five wells were created by sterile well-borer for each plate, and then the sample was loaded to the corresponding well. After 24 h of incubation at 37°C, the antibacterial activity of the nanoparticles was analyzed by measuring the diameter of the clear area of the inhibition zone around the wells. Figure 1 displays powder X-ray diffraction patterns of the prepared NiFe 2 O 4 , NiFe 2 O 4 @TiO 2, and NiFe 2 O 4 @TiO 2 @rGO core-shell NPs. From Figure 1 [56]. Figure 1 shows that the main peaks of the NiFe 2 O 4 did not change when coated with TiO 2 and rGO. is confirms the stability of the spinel phase of the NiFe 2 O 4 structure after annealing.

X-Ray Diffraction Analysis.
In Figure 1(c), the diffraction peaks of rGO cannot be observed in the recorded XRD pattern. is can be associated with several concurrent effects, including the following: (i) the position rGO (002) plane is very close to the anatase (101) plane at 25°; thereby, overlapping occurs, resulting in lowering the anatase (101) plane intensity [57]; (ii) in X-ray diffraction, peaks' intensity is dependent on the atomic number of the constituents; in this case, rGO (carbon-6) is much lower compared to NiFe 2 O 4 (nickel-28 and iron-26) as well as TiO 2 (titanium-22) in addition to (iii) the small amount and amorphous structure of rGO compared to that of NiFe 2 O 4 (dominant) and TiO 2 [58]. It is important to highlight that no additional peaks belonging to impurities can be detected within the resolution limit of the X-ray diffractometer, thereby signifying that the as-prepared core@shell samples have a single phase with high purity. e crystallite size (D) of the particles has been determined by the Scherrer formula: where λ is the wavelength of the X-ray radiation source (1.5418Å), β is the full width at half maximum of the most intense diffraction peak (in radians), θ is its corresponding diffraction angle, and 0.9 is a constant for spherical shaped particles. e calculated value of the crystallite size of NiFe 2 O 4 , NiFe 2 O 4 @TiO 2 , and NiFe 2 O 4 @TiO 2 @rGO coreshell NPs is found to be 35, 26, and 23 nm [59]. From Figure 1, it can be observed that the diffraction peaks of NiFe 2 O 4 @TiO 2 and NiFe 2 O 4 @TiO 2 @rGO become broad compared to that of NiFe 2 O 4 , an indication of crystallite reduction, that is, from 35 to 23 nm [60]. is reduction in crystallite size can be explained as follows: the formation of the TiO 2 layer and subsequently TiO 2 @rGO double layers onto NiFe 2 O 4 nanoparticles during the synthesis process inhibited grain growth. It is well known that nanoparticles' formation occurs mainly in 3 stages: nucleation, coalescence, and growth. During the coalescence of NiFe 2 O 4 nanoparticles already formed, TiO 2 and then rGO start clustering on their surface, thereby hindering their growth. Figure 2 illustrates the Rietveld refined XRD patterns of NiF, NiT, and NiTG core@shell photonanocatalysts. e upper field data depicts the observed diffraction and calculated patterns, whereas the lower field shows the difference between the observed and calculated diffraction patterns. In e (101) peak intensity in Figure 2(c) is smaller than that in Figure 2(b); it may be the overlapping between the rGO (002) and TiO 2 (101) planes since both are very close to 25°. No additional peaks can be observed, indicating that the as-prepared photonanocatalysts are of high purity. e Rietveld refinements are performed to determine the crystallite size, microstrain, and lattice parameters. Table 1 displays the fitting parameters. Both S and χ 2 values are statistical findings that indicate how close the experimental pattern is to the calculated one. e S value and χ 2 are very close to 1, indicating a very good fit. e goodness of fit value is around 1.179, 1.166, and 1.175 for NiF, NiT, and NiTG, respectively, which shows the good agreement between the observed and calculated XRD patterns [61]. e crystallite size as obtained from the Rietveld refinements is found to be in the same range, that is, 57, 52, and 47 nm for NiF, NiT, and NiTG, respectively. Similarly, the value of the microstrain varies slightly, that is, 0.057, 0.035, and 0.084% for NiF, NiT, and NiTG, respectively. e lattice parameters of TiO 2 are found to be 3.787 and 3.790Å. Figure 3 shows the vibration of the absorption bands of the as-prepared coreshell NPs in the lattice within the range 400-4000 cm −1 .    Consistently, two strong high-frequency bands located in the ranges of 540-600 cm −1 and 400-450 cm −1 correspond to tetrahedral and octahedral vibration sites of Ni-O and Fe-O complexes, respectively. e strong highfrequency peaks located at 558 cm −1 and 428 cm −1 can be assigned to the tetrahedral Ni-O stretching vibration (Ni 2+ ) and Fe-O stretching vibration (Fe 3+ ), respectively [62], thereby confirming the formation of the spinel NiFe 2 O 4 phase. e broad peak observed at 900-570 cm −1 corresponds to the Ti-O-Ti bending of the anatase phase [63]. e vibration of bands centered at 1650-1550 cm −1 and 3600-3300 cm −1 refers to the characteristic band of the -OH group, indicating the absorbed hydroxyl groups at the surface of nanoparticles [64]. e common dissolved CO 2 antisymmetric vibrational band is observed at 2330-2350 cm −1 .

Fourier Transformed Red Analysis.
e functional group associated with rGO is located at 1632 cm −1 , 1020 cm −1 , and 1116 cm −1 , ascribed to the vibration modes of C � O, C-O-C, and C-OH, respectively [65]. Figure 3(c) indicated that the oxygen-containing functional group peak intensity is very small, hence demonstrating the reduction of GO to rGO [66]. Finally, the FTIR spectra of the as-prepared core-shell nanoparticles exhibit typical characteristic bands of NiFe 2 O 4 , TiO 2 , and rGO. Figure 4 illustrates the optical properties of NiF, NiT, and NiTG core@ shell NPs through UV-visible DRS. e optical bandgap energy value has been determined by means of the Kubelka-Munk function. e graph represents [F(R)hv] 2 versus hv plots, where F(R) is denoted as Kubelka function F(R) � (1-R) 2 /2R and R is denoted as reflectance in UVvisible spectra. Figure 4 shows that the bandgap values are 1.7, 2.4, and 2.9 eV for NiF, NiT, and NiTG, respectively. As reported by the formula λ � 1240/Eg, the respective core@ shell nanophotocatalysis wavelength absorption is found to be 729, 516, and 427 nm for NiF, NiT, and NiTG [67]. Normally, TiO 2 has a bandgap energy of around 3 eV, whereas NiT shows an energy bandgap of 2.9 eV (Figure 4(b)).

Diffuse Reflectance Spectroscopy Analysis.
is reduction in the energy band gap can be attributed to the formation of the TiO 2 layer onto the surface of NiFe 2 O 4 nanoparticles.
is shell formation induces a redshift, as a consequence narrowing the electronic structure level between valence and conduction bands and enhancing the absorption of visible light. e formation of the rGO shell onto the lattice of NiT showed 2.9 eV and gap energy, which will also have the visible light absorption ability. e DRS study indicates that NiTG core@shell NPs possess a good photocatalytic activity and then pure individual compounds.

SEM Observations.
Scanning electron microscopy analysis was carried out to characterize the microstructure of the as-prepared nanoparticles; see Figure 5. SEM image of pure NiFe 2 O 4 powder ( Figure 5(a)) reveals the formation of irregular spherical-like particles at the nanoscale with the tendency to agglomeration due to the interaction between the magnetic dipoles of NiFe 2 O4 NPs. e average diameter of pure NiFe 2 O 4 NPs is in the range of 70-80 nm. Figure 5   TiO 2 and then TiO 2 and rGO onto NiFe 2 O 4 will facilitate transfer of electrons charge separation [68]. SEM observations indicate that the average size of the nanoparticles is larger when compared with the crystallite size obtained by XRD analysis. is is caused due to the high tendency of nanoparticles for agglomeration in the form of µm sized particles [69].

Particle Size Distribution.
e average particle size distribution of core@shell nanoparticles has been confirmed by dynamic light scattering (DLS) analysis. Figures 5(d)   106.7. e mean particle size values obtained by DLS results are also different from that of HR-SEM, which may be due to the salvation properties step carried out in the DLS studies. e DLS determines the particle size distribution of the sample only in the solvated state, whereby there would be solvent molecules associated with the core@shell nanoparticles [70]. e dry state of the powdered samples is only analyzed by HR-SEM.

Energy Dispersive X-Ray Analysis.
e chemical composition of the as-prepared NiF, NiT, and NiTG core@shell NPs has been determined by energy dispersive X-ray (EDX) analysis; see Figure 6. e observed peaks in Figure 6(a) for the NiF sample correspond to nickel, iron, and oxygen elements, which confirm once again the formation of the nickel ferrite phase. Figure 6(b) shows similar elements but with the appearance of titanium peaks, which indicates the formation of the TiO 2 layer onto NiFe 2 O 4 core NPs. For the NiTG sample ( Figure 6(c)), carbon peaks appear in addition to Ni, Fe, O, and Ti elements, which manifests the formation of NiTG core@shell NPs. e EDX elemental analysis indicates that only Ni, Fe, O, Ti, and C are present, hence confirming the purity and formation of the core@shell NPs.

HRTEM Analysis.
e morphology of NiFe 2 O 4 @TiO 2 @ rGO core@shell NPs has been further investigated by HRTEM analysis. Figure 7 is SAED pattern confirms that the NiFe 2 O 4 @TiO 2 @rGO core@shell nanostructures are polycrystalline in nature, which corroborates X-ray diffraction analysis. Figure 8 illustrates the magnetization-field (M-H) curves of the as-prepared NPs recorded at room temperature. It can be observed that all samples exhibit a ferromagnetic behavior with a relatively low hysteresis loop, typical of the soft magnet. However, the magnetic properties (saturation magnetization, remanence, and coercivity) of the core NiFe 2 O 4 are found to be strongly affected by the coating shells (TiO 2 , rGO); see Table 2. It can be noticed that the saturation magnetization value of NiFe 2 O 4 @TiO 2 (16 memu/g) and NiFe 2 O 4 @TiO 2 @rGO (15 memu/g) core@shell NPs decreases by almost three times compared to that of NiFe 2 O 4 (45 memu/g) NPs [62].

Magnetic Measurements.
Usually, the saturation magnetization value of the sample mainly depends on the amount of magnetic component M s � φm s , where φ and m s represent the magnetic particles volume fraction and a single magnetic particle saturation magnetization, respectively. e magnetic component amount present in NiFe 2 O 4 @TiO 2 and NiFe 2 O 4 @TiO 2 @ rGO is less than that of NiFe 2 O 4 NPs [63]. e observed decrease in the saturation magnetization of the spinel ferrite NiFe 2 O 4 phases is explicitly related to the intrinsic nonmagnetic nature of TiO 2 and rGO shells. Asha D Patil et al. studied the magnetic property of TiO 2 -doped nanocrystalline Ni-Cu-Zn ferrites, which also exhibited a decrease in saturation magnetization value with the increase in the amount of TiO 2 in Ni-Cu-Zn ferrites. Similarly, to the saturation magnetization, the remanence of both NiFe 2 O 4 @ TiO 2 and NiFe 2 O 4 @TiO 2 @rGO NPs decreases due as well to the effect nonmagnetic (TiO 2 and rGO) layers formed onto the surface of magnetic NiFe 2 O 4 NPs. However, the coercivity is found to increase gradually from 65 to 74 and 80 Oe for NiFe 2 O 4 , NiFe 2 O 4 @TiO 2 @rGO, and NiFe 2 O 4 @TiO 2 @ rGO NPs, respectively. e TiO 2 and rGO shell formation induces an increase in the magnetic particle space interval, which results in an increase in coercivity [71]. Also, the calcination at the relatively high temperature promotes significantly the disorder in the arrangement of the magnetic moment of local atoms within the crystal lattice, consequently resulting in the linear increase of coercivity and the saturation magnetization, in addition to the inverse dependence of particle size and coercivity as reported in the literature [65,72].

Raman Spectroscopy.
Raman spectroscopy is mainly used to identify the structural defects and disorders present in carbon and carbon-based materials. Figure 8 shows the Raman spectrum of TiO 2 , rGO, NiFe 2 O 4 , and NiFe 2 O 4 @ TiO 2 @rGO. As shown in Figure 8(a), anatase TiO 2 Raman peaks appear at 143, 394, 516, and 632 cm −1 corresponding to Eg (1) , B 1g , A 1 g + B 1 g (2) , and Eg (2) modes. e pure rGO Raman peak of the D-band and G-band appear at 1321 and 1593 cm −1 in Figure 9(b). e Raman spectrum of pure NiFe 2 O 4 located at 259, 310, 463, 546 cm −1 , and 999 cm −1 correlates to T 2g (1) , Eg, T 2 g (2) , T 2 g (3) , and A 1 g. In Figure 8(c), the four high intense Raman peaks appearing at 143, 394, 516 cm −1 , and 632 cm −1 can be ascribed to the anatase TiO 2 phase Eg (1) , B 1 g, A 1 g+B 1 g (2) , and Eg (2) modes, respectively [68]. In Figure 8(c), the rGO D-band and G-band of NiFe 2 O 4 @TiO 2 @rGO are observed at 447 and 487 cm −1 , respectively. e sp 2 vibration-bonded carbon atoms are denoted as G-band and the carbon materials of carbon atoms vibration with dangling bonds which is indicated as D-band. ese G-and D-bands help to know the inadequacy and structural disordered in carbon and carbon-based compounds. e quantity of defects in the graphitic compound is characterized by I D /I G intensity ratio. e lower value (below 1) represents a degree of GO higher due to the high sp 2 hybrid carbon atoms. e higher value of I D /I G (above 1) manifests a higher structural disorder upon the reduction of GO [60]. In Figure S1 (in Supplementary Materials), the intensity I D / I G ratio value of NiFe 2 O 4 @TiO 2 @rGO core@shell NPs is found to be 1.04, confirming the successful reduction of  GO during the core@shell thermal formation. e Raman peak of NiFe 2 O 4 is unable to find out in Figure 8(d) due to the high-intensity peaks of TiO 2 overlapping the NiFe 2 O 4 Raman peaks.

Hyperthermia Application.
e generation of heat from the as-synthesized magnetic core@shell nanoparticles in an AC magnetic field for the application of hyperthermia can be examined by the variation of temperature in a function of measured time. e curves of temperature varied with time for pure NiFe 2 O 4 , NiFe 2 O 4 @TiO 2 , and NiFe 2 O 4 @TiO 2 @ rGO core@shall nanoparticles with the help of AC magnetic field of 32.5 mT at a 622 kHz frequency. Hence, in the presence of an AC magnetic field, there are four possible mechanisms to explain the temperature increase (Neel relaxation, Eddy current, hysteresis magnetic field, and Brownian relaxation). In the case of our nanoparticles, the temperature rise is due to the particles spinning due to the AC magnetic field. e energy deposition in this case is related to Neel and Brownian relaxations, which are the most dominant in nanoscale [69]. Overall, all the tested samples manifested a considerable heat emission when AC magnetic field was applied.   mL. All heating curves demonstrate the same behavior of temperature increase. In the first minute, a rapid temperature increase occurred and then continued to increase gradually and in an exponential manner. e rapid increase in temperature is due to the rearrangement of particles when the AC magnetic field was applied, and then the next gradual temperature increase could be associated with the hysteresis loss due to the ferromagnetic nature of the as-prepared samples. After certain times, the temperature' curves start to flatten. is might be explained by the low dispersion of the NPs in the solution as particles tend to gradually accumulate in the sample holder base lowering heat dissipation ability. Furthermore, to evaluate the self-heating efficiency, the specific absorption rates (SAR) were calculated for the samples. e SAR explains the heating effect created per unit mass of the sample. It depends on the initial slope (the 1st minute) of the temperature-time curve as well as the concentration of the sample (equation (1)) [73]. In Table 3, the SAR values are shown along with the XRD crystallite size and magnetization. e temperature rise in degrees is also recorded for 20 minutes. It is essential to consider the rise of temperature degrees as the living cell (at 37°C) can be destroyed at a temperature range of 42 to 47°C for a period of time (10 to 20 mins) depending on the type of treated tissue [74,75]. Table 3 shows the variation of SAR values of the samples as well as the temperature change depending on the concentration of the sample and also on the coatings applied. NiFe 2 O 4 experienced the highest SAR value of 58.2 W/g at the lowest concentration of 5 mg/mL. Higher concentrations of NiFe 2 O 4 (10 and 20 mg/mL) lowered the calculated SAR value; however, at a concentration of 10 mg/mL, the rise in the temperature reached 10°C in only 20 mins. At 10 mg/mL concentration, NiFe 2 O 4 was dispersed more in the solution giving higher heat dissipation ability. More concentration (20 mg/mL) may lead to agglomeration of NPs and hence reduce their self-heating effect. NiFe 2 O 4 @TiO 2 NPs also showed considerable SAR values at a maximum of 18 W/g at 10 mg/mL. Lower heating ability is expected for the coated NiFe 2 O 4 @TiO 2 samples as the saturation magnetization reduces from 45 to 15 emu/g, which leads to less magnetic induced self-heating ability. When rGO was introduced to the samples, a considerable decrease was noticed in the slope and the increased temperature, especially when increasing the concentrations. At a concentration of 5 mg/mL, NiFe 2 O 4 @ TiO 2 @rGo samples have shown an initial temperature change that is higher than that of the NiFe 2 O 4 @TiO 2 samples; this also led to a higher SAR value of 40.5 W/g in comparison with 31 W/g for the NiFe 2 O 4 @TiO 2 . is can be explained by the formation of powerful anisotropy and reduced aggregation of the TiO 2 and rGO shells formed on the NiFe 2 O 4 nanoparticles, which helps the NiFe 2 O 4 @TiO 2 @ rGO heating effect, which is better than NiFe 2 O 4 and NiFe 2 O 4 @TiO 2 [76,77]. However, at higher concentrations (10 and 20 mg/mL), NiFe 2 O 4 @TiO 2 @rGO samples tend to be less dispersive in distilled water solution and to accommodate quickly at the bottom of the sample holder, which may explain the low self-heating parameters as shown in Table 3. Table 4 shows the comparative values of the as-prepared core@shell nanoparticles. Overall, our prepared samples have shown significant SAR values when compared with other values calculated by a previously published work in the literature, taking into consideration the properties of the nanoparticles as well as the self-heating experiment parameters (magnetic field strength and frequency). Table 4 displays a comparison of the SAR values obtained in this study with some recent similar works. Manoher and Hong have prepared NiFe 2 O 4 through the solvothermal reflux method, which resulted in very small nanoparticles (9 nm) with a high saturation magnetization of 46.86 emu/g [78]. A very high SAR value (96.86 w/g) was obtained for these NiFe 2 O 4 samples at a high magnetic field of 294 mT and a frequency of 316 kHz, whereas the combusted NiFe 2 O 4 samples obtained in our work exhibited a SAR value of 58.2 W/g with only 32.5 mT magnetic field strength. Other ferrites nanoparticles have also shown high SAR values at higher magnetic field strengths such as ZnFe 2 O 4 and CoFe 2 O 4 [79,80]. Other studies that conducted the selfheating experiment at lower magnetic field strength experienced lower SAR values such as autoclave MgFe 2 O 4 samples, only 20 W/g at 17 mT [81]. For hyperthermia application, it is very important to consider the clinical magnetic field, which is reported to be around 20 mT (Zhang et al., 2019), as well as the temperature rise [9]. High magnetic fields and high-rate temperature rise could cause harmful side effects to the surrounding healthy tissues [82]. It is also essential to mention that the core@shell design gives the particles the potential to be applied as an efficient drug carrier [83]. Our prepared NiFe 2 O 4 @TiO 2 @rGO core-shell NPs have shown considerable self-heating efficiency in terms of temperature change as most of the samples showed a temperature rise of 6 to 10 degrees in 20 min, which is within the hyperthermia therapeutic range [84]. e obtained SAR values at relatively low magnetic field strength have also shown a great potential for further research on hyperthermia and drug delivery.

Sonophotocatalytic Degradation.
e TC degradation efficiency of NiF, NiT, and NiTG core@shell photonanoparticles has been assessed by sonocatalysis, photocatalysis, and sonophotocatalytic methods. e TC degradation with respect to time was evaluated by UV-Vis spectroscopy.
e UV-Vis spectrum exhibits two main absorption bands at 258-298 nm and 326-378 nm wavelengths corresponding to the presence of tricarbonyl amide and phenolic diketone group resonating, respectively. Initially, the reaction mixture adsorption-desorption equilibrium of catalysis and TC solution was carried out for 30 min under dark condition. No significant antibiotic degradation rate on the catalytic surface of all three photocatalytic NPs has been observed. Figure 11(a) illustrates the UV-Vis absorbance results of TC after the sonocatalytic degradation using NiF, NiT, and NiTG nanophotocatalysts. In the sonocatalysts process, the ultrasonic irradiation cleans the surface of the catalyst NPs rapidly to enhance the material mass transfer and organic compound accumulation reduction and avoid unwanted byproducts. For a contact time of 75 min, the degradation efficiency of TC increases gradually and significantly in the following order: 20% for NiF, 40% for NiT, and 60% for NiTG. Figure 11(b) depicts the TC degradation by a visible light photocatalytic method using NiF, NiT, and NiTG nanophotocatalysts. e visible light energy enhances the photocatalytic electron transfer from the valence band to the conduction band, thereby accelerating the degradation of TC. e degradation rate of TC is enhanced significantly from only 40% for NiF to 60% for NiT, and then reaching 80% for NiTG was for a contact time of 75 min. Figure 11(c) displays the TC degradation by the combined sonophotocatalytic process using NiF, NiT, and NiTG nanophotocatalysts. A similar trend is also observed; the efficiency of TC degradation for a contact time of 75 min is enhanced significantly, achieving 50%, 70%, and 90% for NiF, NiT, and NiTG, respectively.
For the three tested TC degradation methods, it is clearly noticed that the ternary NiTG core@shell NPs demonstrate the highest degradation efficiency compared to the binary NiF and NiT core@shell NPs (Figures 12(a)-12(c)). At the same time, comparing the three different photocatalytic methods, the combined sonophotocatalytic process stands as the most efficient compared to sono-and UV-visible photocatalytic methods taken individually (Figure 12(d)).

Tetracycline Degradation Mechanism.
e degradation pathway of TC is analyzed using the LC-MS technique to determine the intermediate products obtained from visible light degradation. e prominent peak of TC is at m/  Z � 445. In general, functional groups such as double bonds, amine, and phenolic groups are degraded due to their relatively high electron density, out of which double bonds have higher reactivity than others and therefore readily undergo hydroxylation to produce primary product 2 (m/Z � 467) [88]. Due to the weak binding energy of N-C, product 3 (m/Z � 381) and product 4 (m/ Z � 359) are obtained by N-demethylation and deamidation. Product 4 is demethylated with the removal of CH 2 C(OH) to produce product 6 (m/Z � 304). Upon further reaction, product 7 (m/Z � 274) and product 8 (m/ Z � 242) are obtained. Upon oxidation and dealkylation processes, product 3 is converted to product 5 (m/ Z � 318). Upon further reaction progress of dihydroxylation, dealkylation, and ring opening, product 9 (m/Z � 261) and product 10 (m/Z � 218) are attained [89]. e determination of the degradation pathway of TC under visible light is shown in Figure 13, and the LC-MS spectra of the derived products are elaborated in more detail in Figure S2 (in Supplementary Materials) [90]. e obtained results of the degradation pathways indicate that, upon further degradation reactions, the TC molecules would oxidize and progressively degrade to small molecules. e TC degradation by NiFe 2 O 4 @TiO 2 @rGO using visible light achieved 90% in 75 min. e degradation mechanism under visible light is examined to determine the involved pathways. First, TC molecules are adsorbed on the NTG surface by electrostatic attraction and π − π stacking. When NTG is irradiated under visible light, the molecules absorb photon energy which surpasses their bandgap energy to produce electron-hole pairs. e electrons (e -) on the surface of NTG are transferred from the valence band (VB) to the conduction band (CB), and holes (h + ) will be produced in the VB. e CB and VB for NiFe 2 O 4 are found to be 0.62 and 2.91 eV, respectively. e observed CB value is higher than the redox potential of O 2 /O 2 − and hence cannot participate in the reduction process as it is a strong oxidant. Likewise, the holes are at larger oxidation potentials, which favor the oxidation of TC molecules. is is because of the greater value of VB than the redox potential of OH/OH − . e holes readily form OHwhich undergoes oxidation of

Antibacterial Activity Study.
e antibacterial activity of the as-prepared NiFe 2 O 4 , NiFe 2 O 4 @TiO 2 , and NiFe 2 O 4 @ TiO 2 @rGO core-shell NPs has been investigated by using both Gram-positive and Gram-negative bacterial strains ( Figure 14). All tested NPs are smaller in size, so they have a high surface area; thereby, it is expected they can easily reach the bacteria cell to damage the respiration of bacteria and may cause bacterial cell death. Table 5 indicates that the asprepared NiTG shows higher antibacterial activity than NiF and NiT. However, all the three samples show higher activity toward Gram-negative bacteria than Gram-positive bacteria.
e Gram-negative bacteria have a thin layer wall of peptidoglycan polymer (∼7v-v8 nm), whereas Gram-positive bacteria cells have a thicker wall of multilayer of peptidoglycan (<8 nm). For this reason, the as-prepared core@shell NPs can easily reach the Gram-negative bacteria cell membrane and then Gram-positive bacteria. Besides, NiTG core@shell NPs exhibit good magnetic properties, so they can be easily removed from the solution with an external magnetic field, hence avoiding the environmental pollution. From the antibacterial activity results, the as-prepared NiTG core@shell NPs demonstrate good activity and are environmentally friendly.

Conclusion
In this paper, multifunctional core@shell of NiFe 2 O 4 @ TiO 2 @rGO nanostructures have been successfully synthesized through the green synthesis method and characterized. XRD analysis confirms the formation of single nanocrystalline phases. e TiO 2 and rGO shell formation onto NiFe 2 O 4 nanoparticles' surface has been confirmed with interlattice spacing, that is, 0.25 nm for NiFe 2 O 4 (311) plain and 0.35 nm for TiO 2 (101) plane by high-resolution transmission electron microscopy. SAED analysis indicates the polycrystalline nature of NiFe 2 O 4 @TiO 2 @rGO core@ shell nanoparticles. e reduction of GO present in the prepared NiFe 2 O 4 @TiO 2 @rGO core@shell nanoparticles is evidenced by Raman spectroscopy analysis. VSM analysis establishes the ferromagnetic nature with a noticeable decrease in the saturation magnetization upon the deposition of nonmagnetic layers of TiO 2 and rGO. A SAR value of 40.5 W/g is obtained for NiFe 2 O 4 @TiO 2 @rGO core@shell nanoparticles at the lowest concentration of 5 mg/mL under 32.5 mT applied AC magnetic field. e obtained SAR values at relatively low magnetic field strength have also shown a great potential for further research in hyperthermia and drug delivery. e TC degradation performed by means of the three methods reveals that NiTG NPs exhibit a higher degradation efficiency than binary NiF and NiT. e measured zone of inhibition indicates that NiTG core@shell NPs have an effective antibacterial activity in Gram-positive and Gram-negative bacterial pathogens. is promotes their potential use for many medical and environmental applications.

Data Availability
No data were used to support this study.    Figure S1 shows the Raman spectrum of intensity I D /I G ratio value of GO (I D /I G � 0.94) and rGO (I D /I G � 1.04). is different value confirms the successful reduction of GO. Figure S2 shows the LC-MS spectra for the derived product TC degradation pathway under sonovisible light with the help of NiFe2O4@TiO2@rGO nanoparticles. Table S1 shows the Raman spectrum data values for the reference. (Supplementary Materials)