Sensitive MnFe2O4–Ag hybrid nanoparticles with photothermal and magnetothermal properties for hyperthermia applications

In this study, we present an experiment showing that designing multifunctional MnFe2O4–Ag nanoparticles to act as a dual hyperthermia agent is an efficient route for enhancing their heating ability. Interestingly, the specific absorption rate of the heteromeric MnFe2O4–Ag nanoparticles increased 2.7 times under simultaneous irradiation of a 100 Oe magnetic field and 0.14 W cm−2 laser compared to the action by the magnetic field alone, and more interestingly, is 30% higher than the sum of the two individual actions. The synergistic benefit of the magneto- and photo-thermal properties of the heteromeric structure can reduce the strengths of the magnetic field and laser intensities as well as their irradiation time to levels lower than those required in their hyperthermia applications individually. In vitro cytotoxicity analysis performed on HepG2 liver cancer and Hela cervical cancer cell lines showed that IC50 values were 83 ± 5.6 μg mL−1 (for HepG2) and 122.6 ± 19.8 μg mL−1 (for Hela cells) after 48 h of incubation, therefore, the nanoparticles are moderately cytotoxic and nontoxic to HepG2 and Hela cells, respectively; which offers the potential of safe therapy.


Introduction
Recent attempts to increase specic absorption rate (SAR) in magnetic nanomaterials used in magnetic uid hyperthermia (MFH) have led to the discovery of multifunctional hybrid nanoparticles. Hybrid nanostructures with two or more distinctive functional components oen have unique advantages that cannot be gained by individual components. 1,2 To date, several magnetic hybrid carbon nanotubes have been developed for magnetic resonance (MR) imaging and MRguided photothermal treatment of tumors. 3,4 Hybrid NPs with core/shell or composite nanostructures made from plasmonic metals (Au, Ag, or Cu) and magnetic nanoparticles (Fe 3 O 4 , g-Fe 2 O 3 or MnFe 2 O 4 ) are used in imaging diagnostics and hyperthermia therapeutics. In such applications, optical imaging and photothermal are based on the strong localized plasmon resonant optical response of the metallic components of these multifunctional hybrid nanostructures, whereas contrast enhancement for MRI, targeted delivery, and thermal heating can be provided by magnetic components. [5][6][7][8] Regarding the heating properties, using such multifunctional hybrid nanoparticles can benet synergistically the individual magnetothermal and photothermal efficiencies to improve the safety in tumor treatments via reducing strengths of stimulus electromagnetic elds and/or the required amount of the nanoparticles agents. Therefore, this topic has recently attracted a large attention of the scientic community, especially aer the discovery of amplied heating efficiency in magnetite nanocubes irradiated simultaneously by an alternating magnetic eld (AMF) and a laser. 9 In the past decade, among plasmonic nanostructures gold nanoparticles (NPs) had most been used as therapeutic agents because of their adjustable plasmon resonance in the near infrared (NIR). 10 This feature is useful in the treatment of cancer through in vitro 11,12 and in vivo10,13 photothermal tumor ablation. Hence, if gold nanoparticles and magnetic nanoparticles are used simultaneously, the properties of these nanomaterials can act as therapeutic and diagnostic agents. For example, Balasubramanian et al. 14 demonstrated the application of a magnetic hybrid Au nanocomposite for combined chemo-, photo-, and magnetic hyperthermia therapies against pancreatic and breast cancer. However, the fabrications of such multifunctional gold nanoparticles require multistep and expensive processes. The limited scalability and reproducibility of the process hinder the commercialization of the nanoparticles. Therefore, developing alternative materials with increased reproducibility and reduced costs is signicant.
In addition to Au NPs, silver (Ag) NPs are classied as good plasmonic heating agent 11 along with being as excellent anticancer or antibacterial biological actions, and created new opportunities for the extensive study of using Ag NPs in biomedical elds. [15][16][17] Ag NPs can enhance the efficiency of hyperthermia. [18][19][20] For instance, Wang et al. 21 demonstrated that Ag NPs improve the thermosensitivity of C6 cells. Ding et al. 22 showed that multifunctional magnetite-silver hybrid NPs with core-shell (Fe 3 O 4 @Ag) or heteromeric (Fe 3 O 4 -Ag) structures exhibit higher biocompatibility with SMMC-7721 and L02 cells than individual Ag NPs. They also studied the inuence of a combination of Fe 3 O 4 with Ag NPs on magnetic hyperthermia and found that Ag NPs enhance cancer cell death both in vitro and in vivo, although no enhancement in SAR by the hybridization was found. Interestingly, the SAR value of the heteromeric structure is about 10% larger than that of the core shell structure. The above results suggested that fabrication of a nanocomposite-like structure with magnetic NPs separated from Ag NPs might be sufficient to gain a hyperthermia agent for efficient hybrid magneto-and photothermal treatments. As to the best of our knowledge, there has been no report studying on the dual heating modality of an Ag containing hybrid nanostructure.
Manganese ferrite (MnFe 2 O 4 ) shows a higher magnetic saturation than several other magnetic materials. Moreover, manganese ferrite is an inexpensive, air-stable, biocompatible, and magnetically recoverable material with a high saturation eld 23 and has thus emerged as a potential candidate material for various biomedical applications, such as MRI contrast agent, 24 hyperthermia treatment, 25 and drug delivery. 26 Multifunctional nanomaterials designed to compose of MnFe 2 O 4 and inorganic compounds, such as silver or graphene oxide have various purposes, such as increasing antibacterial efficiency, enhancing electrochemical performances, etc.27,28 However, studies on hyperthermia therapy using MnFe 2 O 4 silver hybrid NPs are still scarce.
Aiming at improving the heating properties of nanoparticles, we attempted to synthesize heteromeric MnFe 2 O 4 -Ag structure with different mass fractions of Ag NPs for combining magnetic hyperthermia and photothermal techniques. The nanoparticles were prepared with a facile co-precipitation technique, and the morphology, structure, magnetic and optical properties as well as heating capability of the resultant hybrids were systematically investigated for potential applications in cancer hyperthermia treatments. We measured and compared the heating power of MnFe 2 O 4 -Ag NPs in three protocols, i.e. (i) magnetic inductive heating (MIH), (ii) photothermal heating (PTH) and (iii) both MIH and PTH modes were applied simultaneously (DUAL). It is showed that the SLP can be maximized at equal mass ratio of MnFe 2 O 4 and Ag NPs. The in vitro biocompatibility of the naked MnFe 2 O 4 -Ag hybrids was evaluated by analyzing the IC 50 data of the agents against HepG2 liver cancer and Hela cervical cancer cell lines in order to deduce their toxicity level acceptable for the biomedical application. Overall, this study aimed to provide a novel and effective approach to improve hyperthermia treatment efficiency.

MnFe 2 O 4 NPs synthesis
Nano-sized MnFe 2 O 4 powder was synthesized using a coprecipitation technique. The details of the synthesis procedure are the same as that described in the work of Pereira et al. 29 with slight adjustments. 4 mmol of manganese chloride (MnCl 2 $4H 2 O) and 8 mmol of ferric chloride (FeCl 3 $6H 2 O) were dissolved in deionized water and thoroughly mixed. The liquid precipitate was warmed up to 90 C and aerward 4 mmol NaOH as a precipitating agent was added to beyond solution drop by drop with constant stirring, till pH maintained at 11-12. The mixture kept in that condition for 60 min. Aer the reaction, the product was cooled naturally to room temperature. Finally, the black precipitate was magnetically separated and washed several times with deionized water and dried under vacuum. The produced crystalline MnFe 2 O 4 nanoparticles were collectively labeled as S0.
synthesis steps is illustrated in Fig. 1. First, 35 mL of deionized water and then 20 mg of MnFe 2 O 4 (obtained in Section 2.2) were added to a beaker and the mixture was stirred well. Then, 2 mL of C 6 H 5 Na 3 O 7 $2H 2 O (0.54 mmol) was added and stirred vigorously at 500 rpm for 10 minutes. The mixture was heated to 95 C while stirring for 15 minutes. Then, AgNO 3 was added directly to the solution with 4 various MnFe 2 O 4 : Ag mass ratio of 2 : 1, 1 : 1, 1 : 2 and 1 : 2.5 for S1, S2, S3 and S4 sample, respectively. The used concentrations of AgNO 3 and the Ag mass in all the synthesized MnFe 2 O 4 -Ag NPs samples were gathered in Table 1. The mixture was allowed to react for 1 h at a constant temperature of 95 C while stirring at 500 rpm. Aer cooling to room temperature, the resulting black precipitate was separated with a magnet, washed for several times with deionized water and dried under vacuum. The nal product was used for conducting the characterization. 2.4.2 Raman spectroscopy. All spectra were collected at room temperature using a Jobin-Yvon Lab RAM HR800 Raman Spectrometer with a laser of 830 nm wavelength and power of 17 mW to not affect the sample. The laser beam with its radius of 1.0 mm was focused on the surface of the sample with a 50Â objective. The Raman system was calibrated with a silicon semiconductor using the Raman peak at 520 cm À1 . All spectra were taken in the wavenumber region from 300 to 800 cm À1 , with a resolution of 0.6 cm À1 .
2.4.3 Scanning electron microscopy. The morphology of the samples was analyzed using a eld emission scanning electron microscopy (FESEM) of Hitachi S-4800. Powder samples were deposited on a Si wafer, dried and inserted in the instrument without further coating. The measurement was performed at energies between 5 and 10 keV. Energy-dispersive X-ray mapping of MnFe 2 O 4 -Ag NPs was obtained using the Hitachi S-4800 FESEM equipped with an energy-dispersive X-ray spectroscopy (EDXS) attachment.
2.4.4 Transmission electron microscopy (TEM). The morphology and dimensions of the materials were also analyzed in a JEOL JEM-1010 system operating at 120 kV. Particles were measured using ImageJ soware and the resulting histograms were constructed by the OriginPro 8.5 (Origin-Lab, Northampton, MA, USA) soware package.
2.4.5 UV-Vis-NIR spectroscopy. Optical absorbance of nanoparticle powders was measured using a Cary 5000 UV-Vis-NIR double beam spectrophotometer (Agilent Technologies, Santa Clara, USA) equipped with an integrating sphere attachment using BaSO 4 as background, over the 300-800 nm range of wavelength.  2.4.6 Fourier transform infrared spectrometer. Fourier transform infrared (FT-IR) spectra were collected within a wavenumber range of 4000-400 cm À1 at a resolution of 1.29 cm À1 using a Nicolet 6700 FT-IR spectrometer for the analysis of the bonds of the heteromeric MnFe 2 O 4 -Ag NPs. Specimens were prepared by dilution of MnFe 2 O 4 -Ag powder (2 wt%) in KBr and compressing the mixture into a pellet.
The size distribution of the hydrodynamic diameter of the heteromeric MnFe 2 O 4 -Ag NPs and the stability of the suspension were examined using Zetasizer (Zetasizer Nano, Malvern Instruments, UK). The Zetaseizer system uses the laser diffraction technique to detect the scattered laser light at room temperature. The particles were diluted in water at 10 mg L À1 and ultrasonicated (Sonics Vibra Cell, 8 kJ, power 70%, pulse on/off 1 s/1 s). Each sample was measured 5 times to improve the statistics.
The DC magnetic properties of the samples at room temperature were investigated in dry samples using a homemade vibrating sample magnetometer (VSM) under the magnetic eld up to 11 kOe.
2.4.7 Magneto-and photo-thermal effect measurements. The induction heating rates of the samples were measured using a commercially available generator UHF-20A combined with laser irradiation (Verdi G SLM-Series). The magneto-optical heating performance of 2 mg mL À1 of heteromeric MnFe 2 O 4 -Ag uids was assessed under simultaneous stimulation by 532 nm continuous laser light irradiation with power density of 0.14 and 0.25 W cm À2 and an alternating magnetic eld (AMF) at constant frequency of 340 kHz and amplitude range from 100 to 300 Oe for 14 minutes. Temperature of the sample was monitored in situ along with the irradiation by using a platinum resistance thermometer (PT 100 sensor).
Efficiency of conversion from the absorbed energy of stimulating eld (AMF and/or light) into heat is characterized by specic absorption rate, SAR, which can be experimentally determined using the following equation: 31 where C is the specic heat capacity of water (4.185 J g À1 K À1 ), m is the concentration (mg mL À1 MnFe 2 O 4 -Ag) of the magnetic material in the solution, and dT/dt is the slope of the measured temperature-time curve. In the current experiments, the temperature slope was calculated via analyzing the temperature versus time curves for the whole time range, i.e. rst to t experimental curves by the following equation: 32 to gain: DTtemperature difference between the initial and steady state, and t mthe time constant of heating; and then dT/ dt is determined as equal to DT/t m .
2.5. Cytotoxicity assays 2.5.1 Preparation of stock solutions of drugs. Doxorubicin hydrochloride (Adriamycin®) was provided from Sigma-Aldrich (USA). Stock solutions of 2.5 mg mL À1 doxorubicin hydrochloride and further dilutions (0.039, 0.078, 0.156, 0.312, 0.625 and 1.25 mg mL À1 ) were prepared in RPMI1640. The stock solutions were sterilized using 0.22 m microlters under laminar ow hood and stored frozen.
2.5.2 Cell line and culture conditions. HepG2 liver cancer and Hela cervical cancer cell lines were assessed, which were procured from the American Type Culture Collection. The cell lines were maintained and propagated in 90% Dulbecco's Modied Eagle's Medium containing 10% fetal bovine serum and 1% penicillin/streptomycin. All media, serum, and antibiotics were provided by Gibco, USA. The cells were cultured as adherent monolayers (i.e., cultured at approximately 80% conuence) and controlled at 37 C in a humidied 5% CO 2 atmosphere.
2.5.3 MTS viability assay. In vitro cytotoxicity to HepG2 liver cancer and Hela cervical cancer cell were evaluated using MTS assay and were compared with the untreated controls. A total of 180 mL of cells growing in the log phase, were seeded into four 96-well at-bottom culture plate at a density of 5 Â 10 3 cells per well and subsequently incubated for 24 h. Then, in plate 1 and 2, serial dilutions of doxorubicin (20 mL; nal concentration: 0.039-2.5 mg mL À1 ) were added to a nal volume of 200 mL and incubated for another 72 h. In plates 3 and 4, serial dilutions of MnFe 2 O 4 -Ag (20 mL) were added. Aer an incubation period of 48 h, the medium was aspirated and the cells were washed in PBS. Doxorubicin was used as positive controls (40 mL in each well), and the cells treated only with medium were considered as negative controls.
To evaluate cell survival, 20 mL of MTS [3-(4,5dimethylthiazol-2-yl)-5-(3 carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] was added to each well (100 mg mL À1 in medium) and incubated for 3 h at 37 C. Colored formazan formed inside the cells was dissolved by adding dimethylsulfoxide to each well, and the optical density of the suspension was measured at 540 nm in a microplate reader (Elisa Spectra MAX Plus 384, USA). The following formula was used to calculate the viability of cell growth: cell viability (%) ¼ (mean of IA value of treated group/mean of IA value of control) Â 100.

Characteristics of the heteromeric MnFe 2 O 4 -Ag NPs MNPs
The room temperature powder XRD patterns of the Ag-free and various Ag-containing MnFe 2 O 4 samples are shown in Fig. 2(a). A single cubic phase for pure parent MnFe 2 O 4 (S0 sample) was observed. The composites showed two different sets of XRD peaks, all of which were successfully indexed with cubic structured MnFe 2 O 4 and fcc Ag phases. The reections indexed as (220), (311), (400), (511), and (440) planes (marked with asterisks) conrmed the formation of well-dened single-phase cubic spinel structure characteristics with space group Fd3m, and the reections recorded as (111), (200), (231), (220), and (311) planes (marked with hashtags) are attributed to the fcc Ag. No additional peak of any other phase was detected. With increasing Ag content, the intensities of the Ag peaks gradually increased, whereas the intensities of the MnFe 2 O 4 peaks clearly decreased. Hence, the composites were homogeneous systems composed of cubic MnFe 2 O 4 and fcc Ag. However, the occurrence of both phases in heteromeric NPs did not affect the crystal structure of each phase, indicating the absence of additional chemical interactions between manganese ferrite and Ag metal phases and the conservation of their symmetry in the composites. To further conrm the coexistence of two phases in the heteromeric MnFe 2 O 4 -Ag NPs, the Rietveld renements of the S2 sample is exhibited in Fig. 2(b) as an example. The crystallite sizes were calculated using Scherer's equation with diffraction angle using prole of the (311) and (111) peaks for MnFe 2 O 4 and Ag nanoparticles, respectively. The obtained crystallite size values were listed in Table 2. As can be seen from Table 2, the crystallite size of Ag NPs increases with the increase of Ag concentration, which is opposites with that of MnFe 2 O 4 nanoparticles. Fig. 3(a)-(e) present the SEM micrographs of the heteromeric MnFe 2 O 4 -Ag NPs at the same magnication. Obviously, except in sample S0, no signicant difference in shape was found among particles with different Ag contents. The energy dispersive X-ray (EDS) spectrum analysis results for the red-dotted box region in sample S2 (Fig. 3(f)) is shown in Fig. 3(g). Notably, the peaks in the EDS spectra conrmed the elemental composition (Mn, Fe, Ag, and O), and no other element was detected, verifying that the heteromeric nanoparticles were free of impurity. In combination with the EDX spectra in Fig. 3(g), the corresponding elemental mappings of Mn, Fe, Ag, and O elements are shown in the inset of Fig. 3(g). The elements' map indicated that Mn, Fe, Ag, and O elements were evenly distributed in the heteromeric MnFe 2 O 4 -Ag NPs.  weight rations of 2 : 1 (S1), 1 : 1 (S2), 1 : 2 (S3), and 1 : 2.5 (S4). The average particle sizes determined from TEM analysis were 49.8 for the all samples.
For characterization of the heteromeric MnFe 2 O 4 -Ag NPs, the FTIR spectrum was recorded in the 400-4000 cm À1 range. The results are depicted in Fig. 5. All the samples exhibited strong absorption peaks at 574, 1640, and 3446 cm À1 . The two broad peaks at 1640 and 3446 cm À1 corresponded to the bending and stretching modes of the hydroxyl groups of a small amount of water. 33 The spectra peak observed at 574 cm À1 was the characteristic peak of the intrinsic stretching vibration of iron and oxygen (Fe-O) bond of MnFe 2 O 4 at the tetrahedral site. 34 The peak observed at 1395 cm À1 represented the COOasymmetric stretching vibration of COOH in sodium citrate tribasic dehydrate. This observation was in good accordance with the XRD data conrming the phase and typical surface bonding state changes in the samples before and aer coating with silver metal.
The localized surface plasmon resonance (LSPR) peak of Ag NPs was identied using the optical property of MnFe 2 O 4 -Ag hybrid NPs at 300-800 nm and UV-Vis absorption spectra. As shown in Fig. 6, except the S0 sample, the samples with various   Ag concentrations exhibited the characteristic LSPR peaks of Ag NPs at approximately 432-444 nm. As Ag concentration increased, the LSPR peaks showed slight blue shis from 444 nm for S1 to 432 nm for S4, because of the local dielectric effect according to the classical Mie theory. 35 Similar phenomena were reported in literature. 22,36,37 The inuence of Ag on the Raman scattering of MnFe 2 O 4 nanoparticles was studied at room temperature through Raman spectroscopy. The spinel ferrite crystal structure MnFe 2 O 4 has ve Raman active modes, namely, A 1g + E g + 3T 2g . 38 Fig. 7 describes the Raman spectra of MnFe 2 O 4 -Ag hybrid NPs at 200-800 cm À1 . Fig. 7 shows that the Raman spectra of all the samples exhibited three Raman modes, T 2g , E g , and A 1g , at 220-240, 280-320, and 60-620 cm À1 , respectively, which closely matched the modes in previously reported data. 39 In addition, in the heteromeric MnFe 2 O 4 -Ag, the Raman intensities for all bands increased relative to those of pure MnFe 2 O 4 NPs, and this increase may be attributed to the improvement in electric eld induced by the LSPR of the Ag NPs. 40 The peak intensity signicantly increased with Ag content in the MnFe 2 O 4 NPs.
The magnetic properties of the heteromeric MnFe 2 O 4 -Ag NPs were studied according to their magnetic hysteresis loop at 300 K under an applied magnetic eld (H ¼ 10 kOe), as shown in Fig. 8. Maximum magnetization decreased from 66.5 emu g À1 (for S0 sample) to 16.1 emu g À1 (for S4 sample) with increasing Ag concentration because of the increase in the amount of nonmagnetic silver. Initially, coercivity (H c ) did not change initially in the S0, S1, and S2 samples but subsequently decreased rapidly with further increase in Ag concentration. The highest values of coercivity (H c ) were 113.5 Oe for S0, S1, and S2 samples and 6.8 Oe for the S4 sample. In general, the coercivity of the magnetic nanoparticles depends on magneto-crystalline anisotropy, microstrain, interparticle interaction, grain size, and shape. 41 In our study, the H c of the heteromeric MnFe 2 O 4 -Ag NPs decreased with increasing Ag content in the S3 and S4 samples. This property can be attributed to the low microstrain of Ag.

Heat generation of the heteromeric MnFe 2 O 4 -Ag NPs
The magneto-and photo-thermal efficiency of the heteromeric MnFe 2 O 4 -Ag NPs was analyzed in aqueous dispersions, and three modes were used. First, the magnetic uids were subjected to an alternating magnetic eld at 340 kHz and 100 Oe, and magnetic inductive heating (MIH) was evaluated. Then, the hybrid NPs uids were irradiated with a laser at 532 nm and 0.14 W cm À2 for the assessment of photothermal heating (PTH). Finally, both modes of thermal stimulation were applied simultaneously (DUAL, MIH + PTH). Fig. 9(a)-(e) show the timedependent temperature increase in the MnFe 2 O 4 -Ag uids with a xed concentration of 2 mg mL À1 . In the pure MnFe 2 O 4 uid, the temperature increase (DT) only reached under MIH and DUAL protocols, as shown in Fig. 9(a). In the other samples, temperature increased in the three experimental modes ( Fig. 9(b)-(e)). Moreover, the rate of temperature increase was higher when dual-mode stimulation was used compared to the cases of single individual ones. Interestingly, the therapeutic temperature window (42-45 C) was reached within very short irradiation time, namely less than 2 min for the S2-based uid. SAR values for the uids were then calculated based on the DT and t m parameters gained by tting eqn (2) to the experimental temperature versus time curves. The results are presented in Fig. 9(f). The maximum SAR value was obtained in the S2 sample at three measuring conditions. The heating effect is improved obviously, which indicates that a high localized heating can be rapidly achieved by combining a moderate laser irradiation and a low magnetic eld, simultaneously. As shown in Fig. 9(f), SAR value of S2 sample under DUAL mode (219.5 W g À1 ) is 2.7 times larger than the one by MIH (81.5 W g À1 ) alone, and more interestingly is about 30% larger than the sum of individual MIH and PTH (87.8 W g À1 ) ones. Another noteworthy nding is that the time, t c , to reach cancer therapeutic temperature range (40-45 C) is much shorter for the DUAL mode than for the single modality. As can be seen in Fig. 9(c), the temperature of 42.5 C is reached aer about t c ¼ 100 s and 800 s by the synergistic irradiation of 0.14 W cm À2 laser light combined with magnetic eld of 100 Oe/340 kHz, and only the    Therefore, the photothermal and magnetothermal properties of MnFe 2 O 4 -Ag nanoparticles show great potential for localized cancer treatments, when considering the advantages in reduction of both strengths of stimulating elds and their irradiation time. In addition, in both PTH and DUAL modes, two regimes were observed, namely, a rst linear increase in SAR is observed at low concentrations of Ag, and which tends to have a constant value with increasing Ag concentration, as recently reported. [43][44][45] According to the Beer-Lambert law, 46 the heating transfer of nanoparticles produced by PTH is strongly dependent on nanoparticle concentration, and the light absorption of laser intensity increases with concentration. However, heating capacity reaches a maximum value at a xed high concentration. This effect can be explained by the following light-to-heat energy transfer equation: where I 0 (W) is the incident laser power, S (cm 2 ) is the illuminated area, A is the absorbance of the sample at the irradiation wavelength, h is the photothermal conversion efficiency from irradiation laser energy to thermal energy, m sample (g) is the weight of sample, C (J g À1 K À1 ) is the specic heat of water, and dT/dt ( C s À1 ) is the initial slope of the heating curves. In eqn (2), 10 ÀA becomes negligible at a high concentration. Therefore, the SAR remained nearly constant with increasing Ag concentration. This result was also obtained using the PTH and DUAL protocols. Conversely, heating generation efficiency for MIH was lower than the values obtained using the PTH and DUAL protocols. Initially, SAR decreased from sample S0 (without Ag) to sample S1 and then reached its maximum value in sample S2. Finally, it gradually decreased in sample S4. Apparently, nonmonotonically changes in SAR with decreasing Ag concentration may be caused by another relaxation mechanism apart from the hysteresis loss and Neél-Brownian relaxation, for instance, dipole-dipole interactions, which can be similarly tuned by decreasing magnetic particle volume fraction. A similar property was observed by Martinez-Boubeta et al. 47 Obviously, SAR obtained from the S2 sample in all three conditions reached the highest value. This result implied that an equal mass ratio between MnFe 2 O 4 and Ag would achieve the  optimum experimental SAR value. Therefore, the dependence of SAR on the intensity of the magnetic eld and the power of the laser and the evaluation of in vitro cytotoxicity should be explored for the purpose of application in cancer therapy. Fig. 10(a)-(c) show the time-dependent temperature increase of S2-based uid at 2 mg mL À1 concentration. First, the magnetic uid was subjected to an alternating magnetic eld with a xed frequency of 340 kHz and amplitudes of 100, 150, 200, 250, and 300 Oe for the assessment of MIH alone ( Fig. 10(a)). Then, at each of the above magnetic eld the temperature increases of the hybrid uid were conducted under simultaneous irradiation by the laser of power of 0.14 W cm À2 (Fig. 10(b)) and 0.25 W cm À2 (Fig. 10(c)). As shown in Fig. 10(a)-(c), the heat generated increased with magnetic strength and laser power. The SAR value markedly increased when laser and magnetic stimulation were combined, and this method increased the SAR from 75-250 W g À1 with MIH alone to 625 W g À1 at the maximum laser power (0.25 W cm 2 ; Fig. 10(d)).
Magnetic and photo-induced heating can thus be tuned by varying the magnetic eld and laser parameters. Espinosa et al. 42 obtained the highest SAR value of 150 W g À1 using Fe (16) @Au nanoparticles with MIH/PTH protocols, but the laser power in their study was twofold (0.5 W cm À2 ) that in the current study (0.25 W cm À2 ). Owing to the low concentration (2 mg mL À1 ) of MnFe 2 O 4 -Ag hybrid NPs in our experiment, magnetic hyperthermia combined with photothermal treatment for MnFe 2 O 4 -Ag hybrid NPs optimized the therapeutic condition, as indicated by the SAR. In addition, laser power density (0.25 W cm À2 ) used in this experiment was in the physiologically threshold. A low MnFe 2 O 4 -Ag dose is one of the important advantages in biomedical applications.

Colloidal stability of the heteromeric MnFe 2 O 4 -Ag NPs
The hydrophilic nature of materials is an important parameter for biomedical applications. Fig. 11(a) represents the mean  hydrodynamic size (D H ) of an S2-based uid. The average hydrodynamic size (D H ) obtained by DLS technique was $48 nm, whereas the average core size estimated from TEM was 37.2 nm. The D H was larger than the core size obtained from TEM because the D H included the shell of the surfactant and dissolved molecules around the NPs. The D H in this study was signicantly less than 250 nm, so the uid is quite suitable for biomedical applications. The MnFe 2 O 4 -Ag hybrid NPs were well-dispersed and colloidally stable in aqueous solutions and showed no sign of physical agglomeration or aggregate formation for at least 3 weeks (inset in Fig. 11(a)). The highly hydrophilic property of the MnFe 2 O 4 -Ag hybrid NPs was mainly attributed to the existing of OH-group in the sample. Moreover, zeta potential was measured and used to determine the longterm colloidal stability of MNPs in aqueous solutions. The result is shown in Fig. 11(b). The high negative charge (À32 mV) of the surface implied increase in electrical charge on the surface of the NPs, causing electrostatic repulsion between particles. This result indicated that the long hydrophilic chains of the OH-group played an important role in dispersing the NPs. Moreover, the suspended MnFe 2 O 4 -Ag hybrid NPs were easily collected by a magnet, as shown in the inset in Fig. 11(a), suggesting that they can be controlled by an external magnetic eld.

In vitro cytotoxicity assay of the heteromeric MnFe 2 O 4 -Ag NPs
In general, assessing the biocompatibility of heteromeric MnFe 2 O 4 -Ag NPs using cytotoxicity assay is important. The cytotoxicity of the S2 sample with various MnFe 2 O 4 -Ag concentrations was evaluated against HepG2 liver cancer and Hela cervical cancer cell lines with a standard MTS assay. The MTS assay of the HepG2 cells and Hela cells aer 48 h of contact with 23.45, 46.9, 93.75, 187.5, 375, 750, and 1500 mg mL À1 of  MnFe 2 O 4 -Ag NPs. The results obtained from these studies are depicted in Fig. 12. Cell viability was found to be concentration dependent, and with the increasing concentration of the S2 sample, cell viability decreased. Aer the concentration of the MnFe 2 O 4 -Ag NPs increased from 187.5 mg mL À1 to 1500 mg mL À1 , cell viability percentage was reduced. Aer 48 h of incubation, control treatments showed the highest cell viability percentage (>100%), and the lowest (<19%) was observed at 1500 mg mL À1 nanoparticle concentration. IC 50 values of 83 AE 5.6 mg mL À1 (R 2 ¼ 0.98) and 122.6 AE 19.8 mg mL À1 (R 2 ¼ 0.97) were observed for HepG2 and Hela cells aer 48 h of incubation, respectively. In contrast, doxorubicin (positive control) in the tested concentrations of 2.5 mg mL À1 killed most of the cells. IC 50 values were 0.23 AE 0.07 mg mL À1 (R 2 ¼ 0.98) and 0.25 AE 0.05 mg mL À1 (R 2 ¼ 0.97) for HepG2 and Hela cells, respectively, evidencing toxic effect on these cells of doxorubicin. Similarly, cytotoxic activity against N27 dopaminergic neuronal cells of Mn nanoparticles was induced in time-and dose-dependent manners was reported in the literature. 48 To illustrate the progress of the experiment on the cytotoxic activity, we presented optical micrographs in Fig. 13-16. Two control samples had nearly the same viability percentage in the HepG2 cells and Hela cells. Both types of cells survived at least not shorter than 48 h. When the cells were exposed to higher concentrations of MnFe 2 O 4 -Ag NPs, they showed progressive increases in apoptotic and necrotic activity. At 750 mg mL À1 , the cells were difficult to observe (see Fig. 13(g), (h) and 14(g), (h)) aer 48 h of incubation. The high concentrations of MnFe 2 O 4 -Ag NPs increased the apoptotic and necrotic activities in the HepG2 cells to a higher degree than in the Hela cells. The mechanism of toxicity of magnetic nanoparticles was discussed in detail in literatures. The studies showed that magnetic nanoparticles of high concentrations induce cell necrosis. 49,50 Most authors proposed that the mechanism of necrosis inducing is arisen from the generation of reactive oxygen species (ROS). 51 High ROS levels can damage cells membranes by lipid peroxidation, interrupting DNA, modulating gene transcription, modifying proteins and consequence in decline in physiological function and cell death. However, the high IC 50 indicate that MnFe 2 O 4 -Ag based-uid is totally nontoxic to the Hela cells and approximately nontoxic to the HepG2 cells. With such toxicity behavior of the naked MnFe 2 O 4 -Ag nanostructure, we believe the fabricated nanohybrid is quite safe in cancer hyperthermia application.

Conclusions
In this work, the magnetic/plasmonic nanoparticles of heteromeric MnFe 2 O 4 -Ag structure were fabricated and thoroughly characterized, and their biocompatibility with HepG2 and Hela cells was demonstrated. Through XRD study, it was conrmed that formation of hybrided Ag NPs did not change the crystal structure of MnFe 2 O 4 nanoparticles, and the enhancement of Raman spectrum was claried to be due to the localized surface plasmon resonance. The magnetization at room temperature decreases with the silver content, with the maximum of 65.5 emu g À1 obtained. The synthesized MnFe 2 O 4 -Ag displayed excellent chemical stability, colloid stability, photostability and biocompatibility. Furthermore, multifunctional MnFe 2 O 4 -Ag is able produce an extra increment of the heating effect in the simultaneous irradiation of magnetic eld and laser, which makes the synergistic effect of magnetically induced heating and photothermal heating of 30% stronger than the sum of the two individual ones. The highest SAR was achieved to be 625 W g À1 for MnFe 2 O 4 /Ag with equal mass ratio at combined magnetic eld of 340 kHz and 300 Oe and laser at a power density of 0.25 W cm À2 . In particular, the role of combining photothermia with magnetic hyperthermia was addressed. Our results imply that the combination of magnetic and plasmonic properties in the nanohybrids directly and strongly inuence the heat generation and, therefore, lead to large SAR in MnFe 2 O 4 /Ag nanouids. Moreover, the therapeutic temperature can rapidly reach with interval during lower than for classical magnetic hyperthermia treatment. The features of much higher SAR and lower t c are particularly important for cancer treatment applications as providing possibility to lower the NPs concentration, and/or lower the strengths as well as shorten the irradiation time of the magnetic and optical stimulating sources. Additionally, bare nanouid is non-toxic to cells thus it is able to apply them in hyperthermia cancer treatment.

Conflicts of interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to inuence the work reported in this paper.