Gd 3+ Doped CoFe 2 O 4 Nanoparticles for Targeted Drug Delivery and Magnetic Resonance Imaging

: Nanoparticles of CoGd x Fe 2 − x O 4 ( x = 0%, 25%, 50%) synthesized via sol–gel auto combustion technique and encapsulated within a polymer (Eudragit E100) shell containing curcumin by single emulsion solvent evaporation technique were formulated in this study. Testing of synthesized nanoparticles was carried out by using different characterization techniques, to investigate composition, crystallinity, size, morphology, surface charge, functional groups and magnetic properties of the samples. The increased hydrophilicity resulted in sustained drug release of 90.6% and 95% for E1(CoGd 0.25 Fe 1.75 O 4 ) and E2(CoGd 0.50 Fe 1.5 O 4 ), respectively, over a time span of 24 h. The relaxivities of the best-chosen samples were measured by using a 3T magnetic resonance imaging (MRI) machine, and a high r 2 /r 1 ratio of 43.64 and 23.34 for composition E1(CoGd 0.25 Fe 1.75 O 4 ) and E2(CoGd 0.50 Fe 1.5 O 4 ) suggests their ability to work as a better T 2 contrast agent. Thus, these novel synthesized nanostructures cannot only enable MRI diagnosis but also targeted drug delivery. reasons for higher r 2 and relaxivity ratio can be the hydrophilicity factor. It was observed, during the synthesis, that the solubility of CoGd x Fe (2 − x ) O 4 in the solvent enhanced with the increment in the Gd 3+ content in the sample. Moreover, E2 composition has higher value of positive charge and bigger size, as compared to E1, leading to low r 2 and relaxivity ratio. Hydrophilicity plays an important role because this can affect the relaxivity to allow proximity of more water molecules to induce short spin–lattice relaxation time. The degree of hydration is strongly inﬂuenced by the nature of hydrophilic or hydrophobic coatings on MNPs and, thus, their capability of MRI [59].


Introduction
Inorganic nanoparticles (NPs) are widely examined for targeted drug delivery (TDD), early stage treatment and diagnosis of major diseases, such as cancer, for instance, for the past 20 years [1,2]. For time efficient and effective treatments, theranostic nanomaterials have been extensively used for clinical purposes. Such materials combine the functionalities of both diagnosis imaging and therapy into one single nanoscale entity [3,4]. Among these nano materials, the utilization of protein-based and polymer-coated magnetic nanoparticles (MNPs) [5,6] in biomedical applications, such as TDD, magnetic hyperthermia [7], magnetic separation [8], targeted imaging in vivo [9,10], magneto-control of subcellular signaling pathways [11], tracking of stem cells [12], magneto-responsive therapy [13] and magnetic resonance imaging (MRI) contrast agents (CAs), have attained serious consideration [14]. The long-term aim of diagnosing diseases is the advancement of imaging techniques having adequate sensitivity and specificity. Therefore, theranostic nanomaterials providing therapy and generating imaging signals when subjected under external stimuli are particularly of great demand [15]. Among all the imaging techniques used in biomedical field for diagnosis and molecular imaging purposes, MRI is the most powerful and versatile diagnostic tool because of its fundamental advantages, such as high spatial resolution, non-invasiveness, excellent soft tissue contrast and safety [16]. In most of the clinical applications, MRI CAs are used for enhancing contrast which in result boosts sensitivity and image quality for more accurate diagnosis.
MRI CAs are generally classified into two categories: paramagnetic metal ions based T 1 CAs and superparamagnetic iron oxide (SPIONS) based T 2 CAs. High spin paramagnetic metal ions or their complexes containing gadolinium (Gd 3+ ) ions having unpaired electrons in their outer orbits give rise to magnetic dipoles when subjected under external magnetic field, due to a (700 times) larger magnetic moment of electron, as compared to that of a proton. These paramagnetic metal ions generate large fluctuating magnetic fields, resulting in the enhancement of proton relaxation if this fluctuation frequency has components close to the Larmor frequency. Such CAs are responsible of providing positive contrast in the case of T 1 weighted MR image, while SPIONS based MR agents shorten the transverse (T 2 ) relaxation time and show negative contrast on T 2 weighted images [17]. The Food and Drug Administration (FDA) permitted the utilization of Gd 3+ based contrast agents (GBCAs) for clinical purposes for about 30 years. However, later it was found that GBCAs contain harmful side effects, including impaired kidney function, fibrosis of skin, joints and internal organs caused due to leached free Gd 3+ ions [18]. Moreover, GBCAs get excreted through the body via urine, thus hindering their use in high-resolution imaging application, which demands a long scan time [19]. Moreover, it has been confirmed that repeated exposure of GBCAs under MRI results in the accumulation of Gd 3+ ions in human brain [20] and bones [21] and stay in the organs for years. Moreover, the capability for Gd 3+ dechelation elevated the concerns over the safety of GBCAs. Consequently, in February 2018, the UK government suspended the licenses of Omniscan and Magnevist, two main commercial GBCAs, until further investigation [22]. Thus, there is a considerable need for new clinically applicable MRI CAs meeting the standards of traditional GBCAs. In

X-ray Diffraction (XRD)
The XRD patterns of synthesized CoGd x Fe (2−x) O 4 (x = 0%, 25%, 50%) MNPs calcined at 850 • C for 5 h are presented in the Figure 1. For CoGd x Fe (2−x) O 4 (x = 0%), the patterns can be indexed to single-phase cubic spinel crystal structure (Fd3m) of CoFe 2 O 4 MNPs (JCPDS card 22-1086) with (220), (311), (400), (422), (511) and (440) diffraction peaks. The lack of any additional or impurity peak in XRD is proof of pure CoFe 2 O 4 MNPs [27]. Moreover, in XRD pattern of CoGd x Fe (2−x) O 4 (x = 25%, 50%), the presence of secondary phase is visible, which is identified as peaks of GdFeO 3 , after comparing with JCPDS card 74-1476 data and the literature [28,29]. The crystallite sizes of CoGd x Fe (2−x) O 4 (x = 0%, 25%, 50%) calculated by using the Scherrer formula were 36.89, 36.90 and 36.92 nm, respectively. The increased peak intensity of secondary phase along with reduced crystallinity is observed with the increment in dopant concentration. The increase in crystallite size with increment in Gd 3+ content agrees with the reported data and is due to the ionic radii difference of Gd 3+ (0.094 nm) and Fe 2+ (0.067 nm) [30]. The bigger ionic radii of Gd 3+ ions as that of Fe 3+ and Co 2+ ions result in expansion of the crystal structure, lattice strains and disordered lattice structure. Such improvements impede the growth of grain and therefore raise the lattice parameter [31], whereas the reduction in lattice parameter "a" is observed with the doping of large sized Gd 3+ ions [32]. This is because rare-earth (RE) ions prefer to occupy octahedral sites [33]. The decrease of "a" can also be associated with the existence of Fe vacancies in samples [34]. visible, which is identified as peaks of GdFeO3, after comparing with JCPDS card 74-1476 data and the literature [28,29]. The crystallite sizes of CoGdxFe(2 − x)O4 ( = 0%, 25%, 50%) calculated by using the Scherrer formula were 36.89, 36.90 and 36.92 nm, respectively. The increased peak intensity of secondary phase along with reduced crystallinity is observed with the increment in dopant concentration. The increase in crystallite size with increment in Gd 3+ content agrees with the reported data and is due to the ionic radii difference of Gd 3+ (0.094 nm) and Fe 2+ (0.067 nm) [30]. The bigger ionic radii of Gd 3+ ions as that of Fe 3+ and Co 2+ ions result in expansion of the crystal structure, lattice strains and disordered lattice structure. Such improvements impede the growth of grain and therefore raise the lattice parameter [31], whereas the reduction in lattice parameter "a" is observed with the doping of large sized Gd 3+ ions [32]. This is because rare-earth (RE) ions prefer to occupy octahedral sites [33]. The decrease of "a" can also be associated with the existence of Fe vacancies in samples [34].  Figure 2a shows the FTIR spectra of CoGdxFe(2 − x)O4 ( = 0%, 25%, 50%) MNPs. The higher-frequency peak (ʋ1) represents the metal-oxygen vibration at the tetrahedral sublattice, and the lower-frequency peak (ʋ2) represents the intrinsic vibrations of metal-oxygen bond at the octahedral sites [35]. For CoGdxFe(2 − x)O4 ( = 0%), the spectra shows absorption bands position at around 384 and 590 cm −1 for octahedral and tetrahedral sites, respectively. With the doping of Gd 3+ ions in CoFe2O4, the positions of ʋ1 and ʋ2 bands shift towards higher frequencies, due to the lattice distortion with the addition of RE ions [27] and increased bond length at octahedral site [36]. Figure 2b shows the FTIR spectra of the best-chosen emulsions E1 and E2. The IR peaks are mostly shifted from their original position due to the specific interactions (most commonly due to hydrogen bonding) between the chemical entities of different components. Changes in the strength of these interactions are expected to cause the change of FTIR peak position and shape of the functional groups [37]. Figure 2b confirms the presence of functional groups of all the components used in  Figure 2a shows the FTIR spectra of CoGd x Fe (2−x) O 4 (x = 0%, 25%, 50%) MNPs. The higher-frequency peak (υ 1 ) represents the metal-oxygen vibration at the tetrahedral sub-lattice, and the lower-frequency peak (υ 2 ) represents the intrinsic vibrations of metaloxygen bond at the octahedral sites [35]. For CoGd x Fe (2−x) O 4 (x = 0%), the spectra shows absorption bands position at around 384 and 590 cm −1 for octahedral and tetrahedral sites, respectively. With the doping of Gd 3+ ions in CoFe 2 O 4 , the positions of υ 1 and υ 2 bands shift towards higher frequencies, due to the lattice distortion with the addition of RE ions [27] and increased bond length at octahedral site [36]. Figure 2b shows the FTIR spectra of the best-chosen emulsions E1 and E2. The IR peaks are mostly shifted from their original position due to the specific interactions (most commonly due to hydrogen bonding) between the chemical entities of different components. Changes in the strength of these interactions are expected to cause the change of FTIR peak position and shape of the functional groups [37]. Figure 2b confirms the presence of functional groups of all the components used in emulsion formulation. For example, the peak at 556 cm −1 confirms the presence of CoGd x Fe (2−x) O 4 MNPs [38,39]. Similarly, the peak around 1110.5 cm −1 represents the functional groups of ether present in the drug, i.e., curcumin, the peaks at 1455.79 and 1733.13 cm −1 represent the CH bending and carbonyl group of polymer Eudragit E100, respectively [37,40]. The peak at 3432.16 cm −1 represents hydroxyl group stretching of curcumin [37], solvent (ethanol) or water used in the formation of organic phase. Hence, the FTIR spectra confirmed the compatibility of all the components used in emulsion formulation and its stability.

Functional Group Analysis by FTIR
Magnetochemistry 2021, 7, x FOR PEER REVIEW 4 of 16 emulsion formulation. For example, the peak at 556 cm −1 confirms the presence of CoG-dxFe(2 − x)O4 MNPs [38,39]. Similarly, the peak around 1110.5 cm −1 represents the functional groups of ether present in the drug, i.e., curcumin, the peaks at 1455.79 and 1733.13 cm −1 represent the CH bending and carbonyl group of polymer Eudragit E100, respectively [37,40]. The peak at 3432.16 cm −1 represents hydroxyl group stretching of curcumin [37], solvent (ethanol) or water used in the formation of organic phase. Hence, the FTIR spectra confirmed the compatibility of all the components used in emulsion formulation and its stability.

Particle Size Determination and Morphological Analysis by SEM
SEM images represent the morphology and grain size of the prepared samples. Figure 3a shows the SEM image of diluted CoGdxFe(2 − x)O4 ( = 0%) and powdered CoGdxFe(2 − x)O4 ( = 25%, 50%) samples. The particle size was found ranging from 22 to 46 nm of the synthesized samples. These SEM images also verify particle size increment with the increase in Gd 3+ content which is in accordance to the reported data [30]. This can be explained based on ionic radius of the dopant used. Gd 3+ has ionic size of 0.094 nm, which is greater, as compared to that of Fe 2+ ions, resulting in increment of particle size [41]. Moreover, the magnetic forces present between the MNPs and high calcination temperature of 850 °C cause their agglomeration [30,42]. Figure 4a,b represents the micrographs of E1 and E2. The average particle size calculated for emulsions E1 and E2 was around 103.5 and 124 nm, respectively. The elemental composition of the synthesized samples was analyzed via EDS. Figure 3b confirms the presence of elements O, Fe, Co and Gd in the prepared samples, while the absence of any traceable impurities confirms the purity of the synthesized MNPs.

Particle Size Determination and Morphological Analysis by SEM
SEM images represent the morphology and grain size of the prepared samples. Figure 3a shows the SEM image of diluted CoGd x Fe (2−x) O 4 (x = 0%) and powdered CoGd x Fe (2−x) O 4 (x = 25%, 50%) samples. The particle size was found ranging from 22 to 46 nm of the synthesized samples. These SEM images also verify particle size increment with the increase in Gd 3+ content which is in accordance to the reported data [30]. This can be explained based on ionic radius of the dopant used. Gd 3+ has ionic size of 0.094 nm, which is greater, as compared to that of Fe 2+ ions, resulting in increment of particle size [41]. Moreover, the magnetic forces present between the MNPs and high calcination temperature of 850 • C cause their agglomeration [30,42]. Figure 4a,b represents the micrographs of E1 and E2. The average particle size calculated for emulsions E1 and E2 was around 103.5 and 124 nm, respectively. The elemental composition of the synthesized samples was analyzed via EDS. Figure 3b confirms the presence of elements O, Fe, Co and Gd in the prepared samples, while the absence of any traceable impurities confirms the purity of the synthesized MNPs.

Magnetic Hysteresis Evaluation by Using VSM
Magnetic hysteresis loops of samples CoGd x Fe (2−x) O 4 (x = 0%, 25%, 50%) calcinated at 850 • C for 5 h are shown in Figure 5. The size, composition and morphology of the synthesized material have a great influence on its magnetic properties [43]. The reduction in saturation magnetization (Ms) and remanent magnetization (Mr) values with the increment in Gd 3+ content supports the reported data. The reduction in Ms values can also be assigned to the increasing crystallite size of prepared samples with increasing Gd 3+ content. It has been reported that the surface of ultrafine MNPs and spins are canted on their surfaces due to lack of balanced exchange interactions [44]. Hence, the surface of synthesized MNPs seems to have nonmagnetic layer. Moreover, the magnetic moments of RE ions can be seen only at very low temperature (less than 40 K) [45]. The arrival of RE ions thus appears to be replacing magnetic Fe 3+ ions by non-magnetic RE ions at octahedral locations [46], resulting in a decrease of Ms values. The values of Ms, Mr and coercivity (Hc) are listed in Table 1.

Hydrodynamic Size and Zeta Potential
The hydrodynamic size and zeta potential of the synthesized samples CoGdxFe(2 − x)O4 (x = 0%, 25%, 50%) analyzed by Malvern Zetasizer were 89.4, 97.2 and 105 nm and +15, −18.8 and −12 mV, respectively. The sharp peaks shown in the Figure 6 confirm the presence of NPs of almost same size range. The average size of 103.9 nm and 125.9 nm and zeta potential of +19.6 ± 7.83 mV and +31.8 ± 6.36 mV was calculated for E1 and E2, respectively. The increased size confirms the successful drug loading and covering of polymer shell over the synthesized MNPs. This result supports the grain size obtained by SEM analysis of these emulsions. The positive charge on the emulsions is attributed to the cationic nature of the polymer used, i.e., Eudragit E100. The suspensions of formulations containing zeta potential values between +30 and −30 mV are believed to be stable. Zeta potential either positive or negative in surface loading than +30 and −30 mV is more prone to inhibit agglomeration in electrostatic stabilization emulsion [47]. The polydispersity index (PDI) of 0.397 and 0.456 for E1 and E2 confirms the stability of the emulsion. The PDI values less than 0.5 correspond to the polydispersity of uniformly dispersed samples. Meanwhile, PDI values greater than 0.7 indicate the presence of NPs having broad size distribution and are not suitable to be analyzed by using the dynamic light-scattering method [48][49][50].

Hydrodynamic Size and Zeta Potential
The hydrodynamic size and zeta potential of the synthesized samples CoGd x Fe (2−x) O 4 (x = 0%, 25%, 50%) analyzed by Malvern Zetasizer were 89.4, 97.2 and 105 nm and +15, −18.8 and −12 mV, respectively. The sharp peaks shown in the Figure 6 confirm the presence of NPs of almost same size range. The average size of 103.9 nm and 125.9 nm and zeta potential of +19.6 ± 7.83 mV and +31.8 ± 6.36 mV was calculated for E1 and E2, respectively. The increased size confirms the successful drug loading and covering of polymer shell over the synthesized MNPs. This result supports the grain size obtained by SEM analysis of these emulsions. The positive charge on the emulsions is attributed to the cationic nature of the polymer used, i.e., Eudragit E100. The suspensions of formulations containing zeta potential values between +30 and −30 mV are believed to be stable. Zeta potential either positive or negative in surface loading than +30 and −30 mV is more prone to inhibit agglomeration in electrostatic stabilization emulsion [47]. The polydispersity index (PDI) of 0.397 and 0.456 for E1 and E2 confirms the stability of the emulsion. The PDI values less than 0.5 correspond to the polydispersity of uniformly dispersed samples. Meanwhile, PDI values greater than 0.7 indicate the presence of NPs having broad size distribution and are not suitable to be analyzed by using the dynamic light-scattering method [48][49][50].

Drug Release and Encapsulation Efficiency
A graphical representation of the drug-release study of the two best chosen emulsions is displayed in Figure 7. An instant and persistent release of 88.44% and 91.32% drug was observed in the first 12 h, which was then followed by a decelerate release of leftover drug over the next 12 h, with the total of 90.6% and 95.0% drug release over 24 h by E1 and E2, respectively. The rapid release of bursts can be attributed to the release of the drug at or nearby the surface of MNPs, due to its hydrophilic nature [51]. Diffusion of the drug inside the MNPs core in the dissolution media could lead to exponential delayed release [52]. The reason observed for the increased drug release in the case of E2 is its hydrophilicity. It was observed during this research that the hydrophilicity of E2 increased with the increased Gd +3 content, which, as a result proved to release more drug as compared to E1, having less Gd +3 content. The encapsulation-efficiency percentages of 87.7% and 74.52%, along with the loading-capacity percentages of 5.84% and 4.62% measured using Equations (1) and (2) for emulsion E1 and E2, respectively, are in accordance with the reported literature [53].  tions (1) and (2) for emulsion E1 and E2, respectively, are in accordance with the reported literature [53].

Korsmeyer-Peppas Model
After confirmation from Higuchi model that drug release follows diffusive mechanism, Korsmeyer-Peppas model was applied for finding out the type of dissolution. It is achieved by calculating its exponent value "n" [54]. This value should be less than 0.5 and, in our case, it came out to be 0.186 and 0.263 for E1 and E2, respectively. The R 2 values for this model are 0.9955 and 0.991 for E1 and E2, respectively, which are very close to 1. The predicted data are very close to those obtained experimentally which confirmed that the drug diffusion of the prepared formulations followed Korsemeyer-Peppas model.

Magnetic Resonance Imaging (MRI)
The inverse relaxation times (1/T1 = r1, 1/T2 = r2) of the compositions E1 and E2 are shown in Figures 8 and 9 respectively. For E1 the values of r1 and r2 are found to be 1.64 and 71.57 s −1 , with a relaxivity ratio of 43.64. Whereas for E2 values of r1 and r2 are found to be 1.85 and 43.18 s −1 , with a relativity ratio of 23.34. For the estimation of the utility of CAs in relaxation r1 or r2, the soothing ratio is a significant parameter: The more calming the agent is, the more efficient the agent can contrast T2 [25]. The relaxivity ratio higher than 10 generally suggests a good T2 imaging ability of CAs, while a relaxivity ratio less than 5 indicates the capability for T1 imaging [26,55,56]. The MRI contrast enhancement relies upon the particle size, magnetic properties, surface composition and charge of the material. Ms value is one of the most important factor that affect the relaxation of T2. [57].

Application of Kinetic Models Korsmeyer-Peppas Model
After confirmation from Higuchi model that drug release follows diffusive mechanism, Korsmeyer-Peppas model was applied for finding out the type of dissolution. It is achieved by calculating its exponent value "n" [54]. This value should be less than 0.5 and, in our case, it came out to be 0.186 and 0.263 for E1 and E2, respectively. The R 2 values for this model are 0.9955 and 0.991 for E1 and E2, respectively, which are very close to 1. The predicted data are very close to those obtained experimentally which confirmed that the drug diffusion of the prepared formulations followed Korsemeyer-Peppas model.

Magnetic Resonance Imaging (MRI)
The inverse relaxation times (1/T 1 = r 1 , 1/T 2 = r 2 ) of the compositions E1 and E2 are shown in Figures 8 and 9 respectively. For E1 the values of r 1 and r 2 are found to be 1.64 and 71.57 s −1 , with a relaxivity ratio of 43.64. Whereas for E2 values of r 1 and r 2 are found to be 1.85 and 43.18 s −1 , with a relativity ratio of 23.34. For the estimation of the utility of CAs in relaxation r 1 or r 2 , the soothing ratio is a significant parameter: The more calming the agent is, the more efficient the agent can contrast T 2 [25]. The relaxivity ratio higher than 10 generally suggests a good T 2 imaging ability of CAs, while a relaxivity ratio less than 5 indicates the capability for T 1 imaging [26,55,56]. The MRI contrast enhancement relies upon the particle size, magnetic properties, surface composition and charge of the material. Ms value is one of the most important factor that affect the relaxation of T 2 . [57]. Jun et al. indicated that Ms values depend upon the size and composition of MNPs [58]. Metal NPs, including Fe, Ni and Co, have higher Ms values than those for oxide NP. In accordance with the reasons explained above, both compositions confirmed their suitability to be used as T 2 contrast agents. However, E1 displayed an r 2 /r 1 ratio twice that of E2, which is because E1 has a high Ms value of 47.81 emu/g, as compared to 29.719 emu/g of E2. One of the reasons for higher r 2 and relaxivity ratio can be the hydrophilicity factor. It was observed, during the synthesis, that the solubility of CoGd x Fe (2−x) O 4 in the solvent enhanced with the increment in the Gd 3+ content in the sample. Moreover, E2 composition has higher value of positive charge and bigger size, as compared to E1, leading to low r 2 and relaxivity ratio. Hydrophilicity plays an important role because this can affect the relaxivity to allow proximity of more water molecules to induce short spin-lattice relaxation time. The degree of hydration is strongly influenced by the nature of hydrophilic or hydrophobic coatings on MNPs and, thus, their capability of MRI [59]. accordance with the reasons explained above, both compositions confirmed their su ity to be used as T2 contrast agents. However, E1 displayed an r2/r1 ratio twice that which is because E1 has a high Ms value of 47.81 emu/g, as compared to 29.719 em E2. One of the reasons for higher r2 and relaxivity ratio can be the hydrophilicity fa was observed, during the synthesis, that the solubility of CoGdxFe(2 − x)O4 in the s enhanced with the increment in the Gd 3+ content in the sample. Moreover, E2 comp has higher value of positive charge and bigger size, as compared to E1, leading to and relaxivity ratio. Hydrophilicity plays an important role because this can aff relaxivity to allow proximity of more water molecules to induce short spin-lattice tion time. The degree of hydration is strongly influenced by the nature of hydroph hydrophobic coatings on MNPs and, thus, their capability of MRI [59].
After stirring magnetically for 1 h, the solution was neutralized by dropwise addition of aqueous ammonia and heated at 100 °C until the formation of gel. The gel then converted automatically into fluffy powder form upon self-ignition. This prepared powder was dried further, in oven, at 100 °C, for a few hours, for complete removal of moisture.  O and citric acid, were prepared, separately, by dissolving their stoichiometric quantities in deionized water. Then a 1:1.5 molar ratio of metal nitrates to citric acid was taken, and the solutions were then mixed.

Materials
After stirring magnetically for 1 h, the solution was neutralized by dropwise addition of aqueous ammonia and heated at 100 • C until the formation of gel. The gel then converted automatically into fluffy powder form upon self-ignition. This prepared powder was dried further, in oven, at 100 • C, for a few hours, for complete removal of moisture. Which was then ground by using mortar and pestle to get fine powder and calcined at 850 • C in a muffle furnace for 5 h.

Emulsion Formulation
Solvent evaporation technique was used to formulate O/W emulsions. Firstly, the organic phase was prepared by taking fixed amount of polymer, drug and MNPs in specific quantity of solvent in separate glass vials and sonicated in ultrasonication water bath for 10 min. These three solutions were then combined in one vial and again sonicated for 10 min. For aqueous phase, fixed amount of Tween ® 80 was dissolved in deionized water, by stirring magnetically for about 30 min. Then, the organic phase was poured, dropwise, into aqueous phase in ultrasonication water bath, for 30 min, to form a uniform solution. At the end, the prepared emulsion was heated at 30 • C, for 15 min, in Rota vapor, to remove solvent from it. The emulsion obtained at the end was stored and further used for different characterizations.
For getting a uniform and stable emulsion, different parameters, such as concentration of polymer, drug, MNPs and volume of aqueous phase, were altered. For this purpose, Design Expert software was used, which generated a list of different formulations. From that list, the one formulation showing the best stability for over three months for both compositions, i.e., CoGd 0. 25

Characterizations
The structural analysis of all the samples were performed, using STOE Powder X-Ray diffractometer θ-θ between 2θ values ranging from 20 • to 80 • , at room temperature, using Cu Kα (λ = 1.5406 Å) radiation. The patterns were evaluated by X'pert Highscore software and plotted, using Origin software. For the investigation and confirmation of required functional groups in the synthesized samples, FTIR transmission spectra were taken, using PerkinElmer, SpectrumTM100 in a range from 300 to 1000 cm −1 for NPs and from 400 to 4000 cm −1 for emulsions. For morphological and elemental analysis, Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS) were performed, using Vega3 Tescan microscope (Czech Republic) connected with EDS apparatus of Oxford instrument. The diluted and powdered samples were firstly deposited on the glass substrate and dried at room temperature and then coated with gold, under vacuum, by cathodic sputtering. SEM images of the samples were obtained under an accelerating voltage of 15 kV. For investigation of magnetic behavior of MNPS, vibrating sample magnetometry (VSM) was performed, using Lake Shore 7407 calibrated with pure nickel as a calibration sample at room temperature with maximum applied field of 10kOe. The charge and average hydrodynamic size (D h ) of MNPs and emulsions was measured, using Zeta Sizer Nano ZS (Malvern Instruments, UK) by taking highly diluted dispersion of MNPs, by using distilled water as a dispersion medium, at 25 • C.

In Vitro Drug Release Kinetics
To study in vitro drug release kinetics of the emulsions, 5 mL from each prepared emulsion was taken and placed inside a dialysis membrane which was then dipped in conical flask containing 50 mL phosphate buffer solution (pH 7.4). This conical flask was then placed in a mechanical shaker for 24 h, at 37 ± 0.5 • C temperature, and 1 mL of sample was taken out after fixed time intervals from the buffer solution and replaced with fresh buffer solution, to keep the total volume constant. Ten samples were taken during a time interval of 24 h and analyzed by UV-Vis spectrometer model HALO DB-20 series. Similarly, for the measurement of drug-encapsulation efficiency and loading capacity, emulsions were centrifuged at 4500 rpm for 1.5 h. The supernatant was discarded, and the obtained pellet was centrifuged again for 1.5 h, at 4500 rpm, by adding ethanol. The supernatant was then used for UV-Vis analysis. The drug encapsulation efficiency and loading capacity was calculated by using Equations (1) and (2) [60], respectively, as given below.
Diffusion and dissolution of drug encapsulated within the polymer matrix was monitored and different kinetic models, i.e., Zero Order kinetics, First Order kinetics, Higuchi kinetics and Korsemeyer-Peppas kinetics models, were applied, using a software named "DD Solver 1.0" to analyze drug release [60]. The best fitted model was then selected depending upon the correlation coefficient (R 2 ), which is a statistical measure of closeness of predicted data to experimental ones under the same conditions.

Magnetic Resonance Imaging (MRI) Analysis
The performance of the synthesized samples as MRI agents was evaluated by using clinical MRI system GE Optima MR450w 3.0T. For this purpose, the samples were prepared at different concentrations by using different TE and TR sequence. For T 2 contrast, TR sequence of 5000 ms with varying TE of 14.27, 42.8, 76, 108, 131, 164 ms and for T 1 , fixed TE value of 10.64 ms with TR of 101 and 685 ms was used. Two glass vials, one containing water and one containing commercial Gd sample, respectively, were used as references for comparing the imaging ability of the formulated emulsions with that of water and Gd. After taking MRI images, a software named "Kpacs" was used for calculating the mean intensities of the samples by using mean Region of Interest (ROI) values of the respective samples and the relativity values of the samples were calculated, as well [61,62].

Conclusions
In this research, magnetic polymeric emulsion of Gd 3+ doped CoFe 2 O 4 MNPs, using Eudragit E100 (copolymer) and curcumin (anti-cancer drug), was formulated. The Gd 3+ doped CoFe 2 O 4 MNPs having single phase cubic structure uniform spherical morphology and enhanced magnetic properties were synthesized by sol-gel auto-combustion method. These MNPs were then encapsulated within a copolymer shell along with the drug curcumin by forming oil-in-water (O/W) emulsion, using solvent evaporation technique. The formulated emulsion showed spherical morphology with a zeta potential of +19.6 ± 7.8 mV and +31.8 ± 6.3 mV, indicating its stability. FTIR analysis confirmed the presence of all the major functional groups of the materials used for their formation. In vitro drug release study performed at pH of 7.4 pursued Korsemeyer-Peppas kinetic model, along with sustained and continuous drug release of 90.3% and 95.0% over a time span of 24 h and encapsulation efficiency of 88% and 76% for E1 and E2, respectively. The contrast enhancement ability of these nanostructures was confirmed by MRI analysis. Moreover, a r 2 /r 1 ratio of 43.64 and 23.34 for E1 and E2, respectively, confirmed its ability to work as a better T 2 contrast agent. Such synthesized nanostructures provide a successful route for bio-compatible and selective drug delivery and enhanced MRI, thereby promoting further research into the field. Funding: This research has not received any external funding.

Data Availability Statement:
The experimental data used to support the findings of this study are available for the readers.