Screening of Mono-, Di- and Trivalent Cationic Dopants for the Enhancement of Thermal Behavior, Kinetics, Structural, Morphological, Surface and Magnetic Properties of CoFe2O4-SiO2 Nanocomposites

CoFe2O4 is a promising functional material for various applications. The impact of doping with different cations (Ag+, Na+, Ca2+, Cd2+, and La3+) on the structural, thermal, kinetics, morphological, surface, and magnetic properties of CoFe2O4 nanoparticles synthesized via the sol-gel method and calcined at 400, 700 and 1000 °C is investigated. The thermal behavior of reactants during the synthesis process reveals the formation of metallic succinates up to 200 °C and their decomposition into metal oxides that further react and form the ferrites. The rate constant of succinates’ decomposition into ferrites calculated using the isotherms at 150, 200, 250, and 300 °C decrease with increasing temperature and depend on the doping cation. By calcination at low temperatures, single-phase ferrites with low crystallinity were observed, while at 1000 °C, the well-crystallized ferrites were accompanied by crystalline phases of the silica matrix (cristobalite and quartz). The atomic force microscopy images reveal spherical ferrite particles covered by an amorphous phase, the particle size, powder surface area, and coating thickness contingent on the doping ion and calcination temperature. The structural parameters estimated via X-ray diffraction (crystallite size, relative crystallinity, lattice parameter, unit cell volume, hopping length, density) and the magnetic parameters (saturation magnetization, remanent magnetization, magnetic moment per formula unit, coercivity, and anisotropy constant) depend on the doping ion and calcination temperature.

into the SiO 2 matrix gained considerable attention due to the possibility of controlling the particle size and minimizing the surface roughness and spin disorder, thus enhancing the magnetic properties of the obtained nanocomposites [22,23]. Additionally, the nonmagnetic SiO 2 matrix possesses a high surface area and does not affect the magnetic behavior or electric properties of the CoFe 2 O 4 nanoparticles due to its lower dielectric constant [29].
In the present work, we investigate the changes in the structural, morphological, surface, and magnetic properties of undoped and doped CoFe 2 O 4 with transition monovalent (Ag + , Ag 0.1 Co 0.95 Fe 2 O 4 ; Na + , Na 0.1 Co 0.95 Fe 2 O 4 ), divalent (Ca 2+ , Ca 0.1 Co 0.9 Fe 2 O 4 ; Cd, Cd 0.1 Co 0.9 Fe 2 O 4 ) and trivalent (La 3+ , La 0.1 CoFe 1.9 O 4 ) metal ions embedded in the SiO 2 matrix obtained through the sol-gel route, followed by calcination at various temperatures.
The novelties of this paper consist of the following: (i) A study of the influence of the SiO 2 matrix and dopant ion on forming metallic succinates and their decomposition into ferrites embedded in the SiO 2 matrix. (ii) A comparative study of the impact of monovalent, divalent, and trivalent ion doping of CoFe 2 O 4 embedded in the SiO 2 matrix on the morphological, magnetic and structural properties in order to find new strategies to increase their potential for existing and new possible applications. (iii) Filling the gap in the existing literature on the effect of Na + , Ag + , Cd 2+ , La 3+ , and Ca 2+ ion doping on the physicochemical properties of CoFe 2 O 4 embedded in the SiO 2 matrix; the embedding of CoFe 2 O 4 nanoparticles in the non-toxic, inert SiO 2 matrix via the sol-gel method allows the control of the particle growth, reduces the nanoparticle agglomeration and enhances the chemical stability and magnetic guidance [11]. (iv) The elucidation of the thermal behavior under isothermal and non-isothermal conditions and the formation kinetics of undoped and doped CoFe 2 O 4 . (v) Obtaining single crystalline ferrites at low calcination temperatures (400 and 700 • C) by doping CoFe 2 O 4 with mono-, di-and trivalent ions. Figure 1 shows the TG and DTA curves for the gels dried at 40 • C. The DTA curves ( Figure 1) of the gels dried at 40 • C show the following three processes: (i) the loss of residual water (physically adsorbed water and moisture) indicated by the endothermic peak with the maximum at 33.6-35.5 • C, with a mass loss of 5.3-7% between 25 and 110 • C; (ii) the formation of metal succinates indicated by two endothermic effects at 143.6-145 • C (formation of Co, Ag, Na, Ca, and Cd succinates) and 177-184 • C (formation of Fe and La succinates) with a mass loss of 14.4-16.3% and (iii) the decomposition of metal succinates to metal oxides and the formation of ferrites shown by two exothermic effects at 250-257 • C (mass loss of 18.8-20.9%), corresponding to the oxidative decomposition of the succinates to Ag, Na, Co, Ca, and Cd oxides and at 283-294 • C (mass loss of 10.1-13.2%), consistent with the formation of Fe and La oxides [21,22,29,30]. The two-stage (144-145 • C and 177-184 • C) evolution of the redox reaction between the metal nitrates and 1,4-butanediol (1,4-BD) is due to the free aqueous, stronger acid [Fe(H 2 O) 6 ] 3+ ion than [Co(H 2 O) 6 ] 2+ ion [21]. As a result, the Fe succinate is formed at higher temperatures [21]. The corresponding mass loss on the TG curve is due to the loss of crystallization water from the nitrates and volatile products (H 2 O, NO 2 ), resulting in the redox reaction [21]. An additional mass loss (0.1-1.6%) can also be observed on some TG curves between 900 and 1000 • C. This effect is the most visible in the undoped CoFe 2 O 4 . The highest total mass loss is observed for CoFe 2 O 4 (57.6%), while in the doped CoFe 2 O 4 , the total mass loss is slightly lower (53.4-54.6%). The SiO 2 matrix undergoes various transformations during the thermal process, which makes the processes' delimitation ascribed to the formation and decomposition of succinate precursors difficult [21][22][23]29,30]. various transformations during the thermal process, which makes the processes' delimitation ascribed to the formation and decomposition of succinate precursors difficult [21][22][23]29,30]. The decomposition of metal succinates under isothermal conditions (150, 200, 250, and 300 °C) is presented in Figure 2. In all cases, a rapid mass loss in the first 10 min, followed by a slow mass loss of up to 120 min, is observed. The mass loss is lower at 150 and 200 °C due to the incomplete formation of metal succinates and is comparable at 250 and 300 °C, confirming the formation of ferrites around 250 °C. Ag0.1Co0.95Fe2O4 presents the lowest mass loss on the 150 and 200 °C isotherms, while Na0.1Co0.95Fe2O4 presents the lowest mass loss on the 250 and 300 °C isotherms. The highest mass loss appears on the 150 °C isotherm for The decomposition of metal succinates under isothermal conditions (150, 200, 250, and 300 • C) is presented in Figure 2. In all cases, a rapid mass loss in the first 10 min, followed by a slow mass loss of up to 120 min, is observed. The mass loss is lower at 150 and 200 • C due to the incomplete formation of metal succinates and is comparable at 250 and 300 • C, confirming the formation of ferrites around 250 • C.

Kinetics of Doped and Undoped CoFe2O4 Formation
The rate constant (k) was calculated using the isotherms recorded at 150, 200, 250 and 300 °C according to the following first-order kinetic equation, Equation (1): where dx/dt is the reaction rate, xo is the initial mass (mg), x is the mass (mg) at time t, t is the time and k is the rate constant defined as k = A e (-E/RT) , where A is the pre-exponential factor, E is the activation energy, R is the ideal gas constant, and T is the temperature [31]. The integration of Equation (1) leads to Equation (2) as follows: Each isotherm's rate constant (k) value was computed using ten values in the 0-10 min range, where the highest mass loss occurs. A higher k indicates a faster reaction [31,32]. The k value increases with the increase in the calcination temperature. At 150 and 200 °C, the average k values of the doped ferrites are lower than that of the undoped CoFe2O4, except for Ca0.1Co0.9Fe2O4. At 250 and 300 °C, the average k value is higher for Ag0.1Co0.95Fe2O4, Ca0.1Co0.9Fe2O4, and La0.1CoFe1.9O4, and lower for Na0.1Co0.95Fe2O4 and Cd0.1Co0.9Fe2O4 compared with the average k value of the undoped CoFe2O4 (Table 1).

Kinetics of Doped and Undoped CoFe 2 O 4 Formation
The rate constant (k) was calculated using the isotherms recorded at 150, 200, 250 and 300 • C according to the following first-order kinetic equation, Equation (1): where dx/dt is the reaction rate, x o is the initial mass (mg), x is the mass (mg) at time t, t is the time and k is the rate constant defined as k = A e (−E/RT) , where A is the pre-exponential factor, E is the activation energy, R is the ideal gas constant, and T is the temperature [31]. The integration of Equation (1) leads to Equation (2) as follows: Each isotherm's rate constant (k) value was computed using ten values in the 0-10 min range, where the highest mass loss occurs. A higher k indicates a faster reaction [31,32]  The activation energy (E a ) and the pre-exponential factor (A) for the formation of doped and undoped CoFe 2 O 4 were calculated by plotting the logarithm of the rate constant (k) versus the inverse temperature, log k vs. 1/T ( Figure 3). The activation energy (E a ) was calculated according to the Arrhenius equation (Equation (3)) [32] as follows: where A is the pre-exponential factor, E a is the activation energy (J/mol), R is the ideal gas constant (8.314 J/mol·K) and T is the temperature (K). The E a value of the undoped CoFe 2 O 4 (1.217 kJ/mol) increases by doping with Ag + , Ca 2+ , Cd 2+ and La 3+ , and decreases by doping with Na + (Table 1). The activation energy (Ea) and the pre-exponential factor (A) for the formation of doped and undoped CoFe2O4 were calculated by plotting the logarithm of the rate constant (k) versus the inverse temperature, log k vs. 1/T ( Figure 3). The activation energy (Ea) was calculated according to the Arrhenius equation (Equation (3)) [32] as follows: where A is the pre-exponential factor, Ea is the activation energy (J/mol), R is the ideal gas constant (8.314 J/mol·K) and T is the temperature (K). The Ea value of the undoped CoFe2O4 (1.217 kJ/mol) increases by doping with Ag + , Ca 2+ , Cd 2+ and La 3+ , and decreases by doping with Na + (Table 1).

FT-IR Spectroscopy
The FT-IR spectra (Figure 4) of the gels heated at 40 °C show an intense band around 1380 cm −1 , which is characteristic to nitrate groups, two bands at 2953 and 2862 cm −1 , which

FT-IR Spectroscopy
The FT-IR spectra (Figure 4) of the gels heated at 40 • C show an intense band around 1380 cm −1 , which is characteristic to nitrate groups, two bands at 2953 and 2862 cm −1 , which are specific to C-H bond vibrations and a band at 3200 cm −1 , which is attributed to intermolecular hydrogen bonds in 1,4-BD [22,23]. These bands are not remarked in the spectra of the gels heated at 200 • C, confirming the decomposition of metal nitrates and the formation of the metal succinates, respectively. are specific to C-H bond vibrations and a band at 3200 cm −1 , which is attributed to intermolecular hydrogen bonds in 1,4-BD [22,23]. These bands are not remarked in the spectra of the gels heated at 200 °C, confirming the decomposition of metal nitrates and the formation of the metal succinates, respectively.  The vibrations of the OH groups in 1,4-BD and the adsorbed molecular water appear at 1642 cm −1 , while the band at 950-944 cm −1 is attributed to the deformation vibration of Si-OH that occurs during the hydrolysis of the -Si(OCH2CH3)4 groups in TEOS [21][22][23]. The band at 808 cm −1 is attributed to the stretching vibration of the Si-O chains in the SiO4 tetrahedron, the band at 1048 cm −1 is attributed to the stretching vibration of the Si-O-Si bonds, while the shoulder at 1176 cm −1 is attributed to the stretching vibration of the Si-O bonds in the SiO2 matrix [22,23]. As shown in the XRD data, in the samples calcined at low temperatures (400, 700 °C) the SiO2 is amorphous, while in the samples calcined at 1000 °C, the SiO2 (quartz and cristobalite) is crystalline. In general, macro-sized silica is crystalline, and nanosized silica is amorphous.
At 200 °C, the band around 1626 cm −1 is attributed to the vibration of the COOgroups, indicating the formation of a chelated complex through the coordination of succinates by metal ions. The disappearance of this bandin the FT-IR spectra of the samples calcined at 700 and 1000 °C suggests that the formed cobalt ferrite nanoparticles have no residual organic compounds. The band at 3408 cm −1 is attributed to O-H stretching and intermolecular hydrogen bonds in metal succinates [22,23]. The samples heated at 40 and 200 °C show that the absorption band around 560 cm −1 is assigned to the stretching vibrations of tetrahedral The vibrations of the OH groups in 1,4-BD and the adsorbed molecular water appear at 1642 cm −1 , while the band at 950-944 cm −1 is attributed to the deformation vibration of Si-OH that occurs during the hydrolysis of the -Si(OCH 2 CH 3 ) 4 groups in TEOS [21][22][23]. The band at 808 cm −1 is attributed to the stretching vibration of the Si-O chains in the SiO 4 tetrahedron, the band at 1048 cm −1 is attributed to the stretching vibration of the Si-O-Si bonds, while the shoulder at 1176 cm −1 is attributed to the stretching vibration of the Si-O bonds in the SiO 2 matrix [22,23]. As shown in the XRD data, in the samples calcined at low temperatures (400, 700 • C) the SiO 2 is amorphous, while in the samples calcined at 1000 • C, the SiO 2 (quartz and cristobalite) is crystalline. In general, macro-sized silica is crystalline, and nano-sized silica is amorphous.
At 200 • C, the band around 1626 cm −1 is attributed to the vibration of the COO − groups, indicating the formation of a chelated complex through the coordination of succinates by metal ions. The disappearance of this band in the FT-IR spectra of the samples calcined at 700 and 1000 • C suggests that the formed cobalt ferrite nanoparticles have no residual organic compounds. The band at 3408 cm −1 is attributed to O-H stretching and intermolecular hydrogen bonds in metal succinates [22,23]. The samples heated at 40

X-ray Diffraction of Undoped and Doped CoFe 2 O 4
As the oxidic phases at low temperatures are poorly crystalline or amorphous, the desired surface properties and crystallinity can be achieved by tailoring the calcination conditions. In addition, the reactivity of the amorphous SiO 2 allows its participation in various chemical transformations.
The XRD patterns of the gels calcined at 400, 700, and 1000 • C are presented in Figure 5. At 400 and 700 • C, the diffraction peaks matching with the reflection planes of (220), (311), (222), (400), (422), (511), and (440) confirm the presence of the pure, lowcrystallized CoFe 2 O 4 (JCPDS #00-022-1086) phase with a cubic spinel structure (space group Fd3m) [22][23][24]. At low temperatures (400 and 700 • C), doping ions did not produce any secondary impurity-associated reflections, and the spinel crystal structure of the produced gels was maintained. The absence of secondary phases points toward the successful insertion of doping ions. The broad peak at 2θ = 20-40 • reveals the low crystallization of gels calcined at 400 and 700 • C. At 1000 • C, the undoped ferrite displays the single, well-crystallized CoFe 2 O 4 phase, which is accompanied by cristobalite (JCPDS #89-3434) for the Na-and Ag-doped ferrites and by cristobalite and quartz (JCPDS #85-0457) for the Ca-, La-and Cd-doped ferrites. Calcination also led to a slight shift in the 2θ peak position, small changes in the peak width, and higher crystallite sizes [22,23]. The Cd 0.1 Co 0.9 Fe 2 O 4 gel displays the lowest intensity diffraction peaks, indicating the lowest crystallization compared to the other gels. The increase in the diffraction peaks' intensity with the calcination temperature indicates the increase in the crystallinity degree and crystallite size [22,23]. The presence of La 3+ ions did not generate any secondary impurity-related reflections, and the spinel crystal structure of the CoFe 2 O 4 was maintained [6]. Oppositely, Mansour et al. reported the appearance of LaFeO 3 as a secondary phase due to the diffusion of some La 3+ ions to the grain boundaries that react with Fe to form LaFeO 3 [8]. The XRD patterns are influenced not only by the calcination temperature and doping ions but also by the crystallite size, lattice strain, and defects [22]. 10   The DC is the ratio of the area of crystalline peaks over the total area of the diffractogram. A possible reason for the slight reduction in DXRD by Cd 2+ doping could be the local temperature increase to release the latent energy on the surface. This process leads to a strike in the crystal growth and lowers the concentration of ferrites in the vicinity [3]. The DXRD increases for the doped ferrites, except in the case of Cd 2+ doping ( Table 2). The observed expansion of the unit cell and the high structural distortion of the doped CoFe2O4 compared to the undoped CoFe2O4 are attributed to the difference in the ionic radii of the host and dopant ions, as well as to the change in the cation distribution that occurs due to the Ag + , Na + , Ca 2+ , Cd 2+ , and La 3+ doping in the spinel structure [1,4,6,27]. A possible explanation for the largest unit cell value of La0.1CoFe1.9O4 is the much larger ionic radius of La 3+ (1.216 Å) than of Fe 3+ (0.65 Å), with the unit cell expansion via doping with La 3+ taking place to compensate for the crystal deformation; accordingly, Shang et al. stated that the replacement of Fe 3+ ions by La 3+ ions result in a higher potential barrier for the formation of a spinel ferrite crystal structure [9]. The lattice parameter of the undoped and doped CoFe2O4 gels increases with the calcination temperature, which is also ascribed to the expansion of the unit cell [27]. A possible explanation for the difference between the theoretical and experimental values could be the assumption that the ions are rigid hard spheres [22]. The obtained results and the absence of supplementary phases in the XRD patterns indicate that the doping ions are incorporated into the CoFe2O4 structure [5,6]. The increase in the molecular weight is more significant than the increase in the V (Table 2); however, the molecular weight is more influenced by the increase in the V [22]. The decrease in the unit cell volume is expected with the The structural parameters, namely, the crystallite size (D XRD ), degree of crystallinity (DC), lattice parameter (a), unit cell volume (V), distance between the magnetic ions and the hopping length in the A (d A ) and B (d B ) sites, physical density (d p ), X-ray density (d XRD ) and porosity (P), of the gels calcined at 400, 700 and 1000 • C determined by XRD are presented in Table 2. The D XRD increases with the calcination temperature since at high temperatures (1000 • C), the crystallite agglomeration without subsequent recrystallization led to the formation of a single crystal rather than a polycrystal structure [22]. Mariosi [6].
The DC is the ratio of the area of crystalline peaks over the total area of the diffractogram. A possible reason for the slight reduction in D XRD by Cd 2+ doping could be the local temperature increase to release the latent energy on the surface. This process leads to a strike in the crystal growth and lowers the concentration of ferrites in the vicinity [3]. The D XRD increases for the doped ferrites, except in the case of Cd 2+ doping ( Table 2). The observed expansion of the unit cell and the high structural distortion of the doped CoFe 2 O 4 compared to the undoped CoFe 2 O 4 are attributed to the difference in the ionic radii of the host and dopant ions, as well as to the change in the cation distribution that occurs due to the Ag + , Na + , Ca 2+ , Cd 2+ , and La 3+ doping in the spinel structure [1,4,6,27]. A possible explanation for the largest unit cell value of La 0.1 CoFe 1.9 O 4 is the much larger ionic radius of La 3+ (1.216 Å) than of Fe 3+ (0.65 Å), with the unit cell expansion via doping with La 3+ taking place to compensate for the crystal deformation; accordingly, Shang et al. stated that the replacement of Fe 3+ ions by La 3+ ions result in a higher potential barrier for the formation of a spinel ferrite crystal structure [9]. The lattice parameter of the undoped and doped CoFe 2 O 4 gels increases with the calcination temperature, which is also ascribed to the expansion of the unit cell [27]. A possible explanation for the difference between the theoretical and experimental values could be the assumption that the ions are rigid hard spheres [22]. The obtained results and the absence of supplementary phases in the XRD patterns indicate that the doping ions are incorporated into the CoFe 2 O 4 structure [5,6]. The increase in the molecular weight is more significant than the increase in the V (Table 2); however, the molecular weight is more influenced by the increase in the V [22]. The decrease in the unit cell volume is expected with the introduction of smaller-sized monovalent (Ag + , Na + ) ions in the crystal lattice. The d A and d B of the gels calcined at 700 and 1000 • C are higher for the doped CoFe 2 O 4 than the undoped CoFe 2 O 4 and display a decreasing trend for the monovalent dopant (Ag + , Na + ) ion and an increasing trend for the trivalent dopant (La 3+ ) ion ( Table 2). The lower value of the d p (Table 2) of the undoped CoFe 2 O 4 compared to the doped CoFe 2 O 4 could be attributed to the pore formation through the synthesis processes [22]. The variation of the d p caused by small fluctuations in the lattice constant is probably due to the changes in the cation distribution among the A and B sites. The P ( Table 2) of the doped CoFe 2 O 4 is lower than that of the undoped CoFe 2 O 4 . Additionally, the rapid densification during the calcination and the growth of irregular shape grains decrease the porosity at higher calcination temperatures. The decrease in the P with the increase in the d p may result from the different grain sizes [22]. In conclusion, the increase in the D XRD , DC, a, V, d A , d B , and d p , and the decrease in the d XRD and P at higher calcination temperatures are observed.

Elemental Composition of Undoped and Doped CoFe 2 O 4
To investigate the elemental composition and verify the stoichiometric amount of each element in the undoped and doped CoFe 2 O 4 , the M/Co/Fe molar ratio was determined using inductively coupled plasma optical emission spectrometry (ICP-OES) after microwave digestion ( Table 2). The best fit between the experimental and theoretical data is observed for the gels calcined at 1000 • C.

Morphology and Surface Parameters of Undoped and Doped CoFe 2 O 4
The thermal treatment concomitantly enables the crystalline phase formation and grain growth with a relative coalescence between the particles. The coalescence can be attributed to the physical attraction forces between the small particles or the bonding bridges between the particles, especially at high calcination temperatures [34,35]. As the microscopic examination of gels failed due to the agglomeration of small particles into clusters, the powders were dispersed into deionized water under intense stirring to break up the powder clusters and release the free particles into dispersion. The water dispersion also prevents the particles' re-agglomeration and allows their transfer onto a solid substrate as a thin film through adsorption [36,37]. This method is usually used to prepare thin films of noble metal nanoparticles directly from the mother solution [38,39], but it was also successfully used for the Ni-and Mn-doped ferrites [29,30]. The obtained thin films were subjected to atomic force microscopy (AFM) (Figure 6). Mansour et al. suggested a coalescence phenomenon, resulting in large particle size of the obtained La-doped CoFe 2 O 4 nanoparticles [8].
The undoped CoFe 2 O 4 , after calcination at 400 • C, displays small spherical particles of around 25 nm (Figure 6a), which increase to 30 nm with the calcination temperature at 700 • C (Figure 6b), and to 40 nm at 1000 • C (Figure 6c). The polycrystalline particles formed at 400 • C consist mainly of low-crystalized CoFe 2 O 4 mixed with amorphous material. At 700 • C, the growth of CoFe 2 O 4 crystallites and the presence of amorphous matter are remarked. For the gels calcined at 1000 • C, the crystallite size estimated via XRD (37.2 nm) is comparable with the particle size observed using AFM (around 40 nm). The undoped CoFe 2 O 4 calcined at 1000 • C is mono-crystalline, as indicated by the spherical particle shape with slightly square corners (Figure 6c). These results are in good agreement with the data in the literature [40,41]. The Ag-doped cobalt ferrite calcined at 400 • C is low crystalline, with crystallites of 14 nm mixed with amorphous matter into small spherical particles of around 27 nm (Figure 6d). Increasing the calcination temperature to 700 • C leads to better-developed crystallites of 24.7 nm and increases the particle size to about 31 nm. The development of the crystalline phase and the reduction in the amorphous component determines a significant alteration of the particle shape, which becomes spheroidal (Figure 6e). The particle shape evolving tendency continues by calcination at 1000 • C, with an increase in size to 70 nm and a crystalline core of 65.1 nm (Figure 6f). The presence of secondary phases prevents the development of cubic shape features, which remains spherical, as reported by Prabagar et al. [42]. Mahajan [43].  The Na-doped cobalt ferrite shows no significant modification of the crystallite size for the gels calcined at 400 • C and 700 • C ( Table 2). Like the undoped CoFe 2 O 4 , the amorphous matter between the crystallites leads to slightly larger spherical particles of 24 nm at 400 • C, and 35 nm at 800 • C (Figure 6g,h). Surprisingly, after calcination at 1000 • C, well-structured particles with a 52 nm diameter and a ferrite core of 47.6 nm covered with some traces of cristobalite are remarked (Figure 6i). Due to the presence of the crystalline core, the particle shape becomes spheroidal. The observed size and shape are in good agreement with the data in the literature [44].
The Ca-doped cobalt ferrite, calcined at low temperatures, exhibits spherical particles of around 28 nm (Figure 6j), formed by a 14.6 nm ferrite crystallite core coated with amorphous material. Kumar and Kar also reported crystallites of 10 nm for this composition by calcination at 550 • C for 2 h [45]. Higher calcination temperature at 700 • C facilitates the crystal growth, leading to crystallites of 25.2 nm mixed with some amorphous matter, which generates particles of about 33 nm. Predominately spherical particles are accompanied by several right corners formed on the most representative particles (Figure 6k). Calcination at 1000 • C generates well-formed Ca 0.1 Co 0.9 Fe 2 O 4 crystallites of around 73.7 nm. Due to the traces of cristobalite and quartz crystalline phases, the crystallites become the core of the particles of about 75 nm (Figure 6l). The particle has cubic shapes with rounded edges, which is in good agreement with the data in the literature [46]. Only a few blunted cristobalite and quartz particles are observed around ferrites.
The Cd-doped cobalt ferrite gels calcined at 400 and 700 • C present small spherical particles of around 20 and 28 nm, respectively, containing small ferritic crystallites and amorphous material (Figure 6m,n). The heterogeneous nanostructure of Cd 0.1 CoFe 1.9 O 4 was also evidenced by Shakil et al. [47]. The calcination at 1000 • C leads to particles of around 40 nm (Figure 6o) having a ferrite core of 35 nm, which is in good agreement with the less developed XRD peaks. Particle sizes of 30-50 nm were previously reported for a similar composition [3].
The La-doped cobalt ferrite has a significant influence on the particle size and shape. The calcination at 400 • C results in spherical particles of around 30 nm containing crystallites of 15.5 nm mixed with amorphous material (Figure 6p). A possible explanation could be the higher value of the La atomic radius (2.50 Å) compared to that of Fe (1.26 Å) [48]. This effect is enhanced by calcination at 700 • C, resulting in particles of 38 nm with a crystalline core of 26.1 nm (Figure 6r), covered with amorphous material. Figure 6n shows several bigger particles that might indicate that the crystallization process is in progress, but most particles do not have enough time to reach a larger size. By calcination at 1000 • C, particles of around 90 nm, with a ferrite crystalline core of 81.8 nm covered by traces of cristobalite and quartz (Figure 6s), are formed. The spherical particle features a specific aspect derived from a cubic crystallite core, which is in good agreement with the data in the literature [49,50]. Some small particles of about 40-50 nm belong to the secondary phases. The particle adsorption onto the solid substrate develops thin films with specific topographic characteristics, which depend on the morphological aspects correlated with their density on the surface [51,52].
The tridimensional profiles of the undoped CoFe 2 O 4 and the Ca-, Na-and Cd-doped CoFe 2 O 4 reveal that the particles resulting after calcination at 400 • C build uniform films of well-individualized nanoparticles (Figure 7a,d,p). Interestingly, the doping with Ag and La leads to a slightly irregular film due to the occurrence of local heights (Figure 7g,m), probably due to the irregular adsorption generated by local influences related to the bigger crystallite sizes. The particles obtained by calcination at 700 • C generally display a complex topography of the thin film with randomly spotted bigger particles, which may contain a more developed crystalline core for all the doped CoFe 2 O 4 gels, except for Cd (Figure 7e,h,k,n). The undoped ferrite and Cd 0.1 Co 0.9 Fe 2 O 4 calcined at 700 • C form uniform and smooth films (Figure 7b,r).
The surface roughness (Rg) values are presented in Table 3. The particles formed by calcination at 1000 • C are well individualized and present crystalline topographic aspects and form relatively uniform thin films with no signs of coalescence tendency (Figure 7c,f,i,l,o,s). The observed topographic aspects may be helpful in the further processing of the obtained particles as the main ingredient for dedicated thin film preparation. The AFM investigation allows for the effective area of topographic features at a precise scanned area to be measured. Hence, the thin films obtained through adsorption from aqueous dispersion are uniform and compact (Figure 6), allowing the measurement of powder surface area. The particle number and diameter influence the variation of the obtained values (Table 3) since a large number of bigger particles leads to a large powder area, while a small number of particles with a small diameter leads to a small powder area [22,23]. Therefore, low calcination temperatures generate thin films with a small surface area by spreading the secondary phases among ferrite particles. The wellcrystallized ferrites obtained after calcination at 1000 • C form thin films with a significantly larger powder surface area.
Since the AFM topographic images reveal the exterior aspect of the particles, in order to obtain information on the internal structure of the particles, the transmission electron microscopy (TEM) images were recorded on the gels calcined at 1000 • C (Figure 8), considering that the ferrite crystallites are better developed at this temperature. The crystallites appear in dark grey shades surrounded by a lighter gray hollow, indicating a dense ferrite core covered by a thin layer of SiO 2 . Some particles are associated in clusters of about 90-100 nm, but these clusters are not observed in the AFM images due to their sedimentation in the aqueous dispersion before transferring the nanoparticles onto the glass slide. The surface roughness (Rg) values are presented in Table 3. The particles formed by calcination at 1000 °C are well individualized and present crystalline topographic aspects and form relatively uniform thin films with no signs of coalescence tendency ( Figure  7c,f,i,l,o,s). The observed topographic aspects may be helpful in the further processing of the obtained particles as the main ingredient for dedicated thin film preparation. to obtain information on the internal structure of the particles, the transmission electron microscopy (TEM) images were recorded on the gels calcined at 1000 °C (Figure 8), considering that the ferrite crystallites are better developed at this temperature. The crystallites appear in dark grey shades surrounded by a lighter gray hollow, indicating a dense ferrite core covered by a thin layer of SiO2. Some particles are associated in clusters of about 90-100 nm, but these clusters are not observed in the AFM images due to their sedimentation in the aqueous dispersion before transferring the nanoparticles onto the glass slide.  The undoped CoFe 2 O 4 ( Figure 8a) displays mainly fine particles of about 38 nm in diameter, which is in agreement with the AFM, but slightly bigger than the average value estimated by the XRD, probably due to the presence of the cristobalite and quartz exterior layer. A particle cluster of about 95 nm is observed on the central side of Figure 8a. Small and homogenously distributed particles result when the nucleation rate exceeds the growth rate, but the small particles tend to agglomerate into bigger structures. A possible explanation for the agglomeration tendency of small particles could be the interaction between the magnetic ions, van der Waals forces at the particle surface, and interfacial surface tensions [21,22,29,53,54]. The volume expansion and the internal energy produced during calcination may also lead to particle growth. The SiO 2 matrix reduces the number of particles that interact with each other and, thus, reduces particle agglomeration [10,11,18,42,53,54].
Ag 0.1 Co 0.95 Fe 2 O 4 presents spherical particles of about 67 nm (Figure 8b), which is very close to the value observed by the AFM. The dark core of the particles observed in the TEM images is very close to the average crystallite size estimated via XRD, while the light hollow surrounding the dark core suggests the ferrite coating by a thin layer of cristobalite and quartz, a fact also sustained by the spherical particle shape. The TEM images of Na 0.1 Co 0.95 Fe 2 O 4 ( Figure 8c) show a less dense distribution of spherical particles with a diameter of about 50 nm, which is in good accordance with the AFM. The intense gray shade indicates the presence of a 47.6 nm ferrite core inside the particles, which is in accordance with the crystallite diameter determined by the XRD data. Oppositely, the TEM image shows a compact structure of well-developed Ca 0.1 Co 0.9 Fe 2 O 4 particles with a diameter of about 76 nm (Figure 8d). The particle size is slightly bigger than that observed by the AFM. The dark core corresponds to the ferrite crystallite evidenced by XRD, while the lighter halo on the exterior is attributed to the cristobalite and quartz layer. Figure 8e reveals a compact and uniform package of fine Cd 0.1 Co 0.9 Fe 2 O 4 spherical particles of about 37 nm, which is in good agreement with the AFM. The core is darker due to the ferrite crystallite presence, and the lighter exterior shade corresponds to the crystalline SiO 2 layer. La 0.1 CoFe 1.9 O 4 presents big spherical particles of about 84 nm (Figure 8f), which is in good agreement with the AFM. The dark core corresponds to the ferrite, evidenced by XRD, and the outer halo to the cristobalite and quartz layer. The presence of some crystallite clusters indicates a local powder agglomeration.

Magnetic Properties of Undoped and Doped CoFe 2 O 4
The magnetic hysteresis loops, M(µ 0 H), and the magnetization first derivatives (dM/d(µ 0 H) of the gels calcined at 700 • C ( Figure 9) and 1000 • C ( Figure 10) indicate a typical ferromagnetic behavior. The derivative of the hysteresis loops (total susceptibility) is the local slope of the M-H curve. For the gels calcined at 700 • C, a single maximum in the dM/d(µ 0 H) vs. the µ 0 H curve, close to the coercive field, consistent with a single magnetic phase, is observed. These behaviors suggest crystalline samples with a single magnetic phase [21][22][23]. The magnetic hysteresis loops indicate moderate coercivity due to the coalescence of the particles accompanied by their magnetic coupling and improved magnetization. Although the magnetization first derivative dM/d(µ 0 H) of the undoped CoFe 2 O 4 calcined at 1000 • C shows two maxima (a more intense and better differentiated maximum next to a less intense one), one on each side of the coercivity, these two magnetic phases are magnetically coupled inside of the particle along their magnetic moments [21][22][23]. The doping effect of the monovalent (Ag + and Na + ) ions supports the formation of the two magnetic phases (an intense peak and one as a shoulder merged with the other for Ag 0.1 Co 0.95 Fe 2 O 4 , and a broader maximum peak suggesting the merging of the two maxima, characteristic of the two magnetic phases for Na 0.1 Co 0.95 Fe 2 O 4 ). Oppositely, the doping with the divalent (Ca 2+ and Cd 2+ ) and trivalent (La 3+ ) ions improves the magnetic properties, leading to the formation of a single magnetic phase characterized by a single maximum, which is very intense and sharp on the dM/d(µ 0 H) vs. the µ 0 H curve.
Due to the change in the magnetocrystalline anisotropy or the particle sizes by doping CoFe 2 O 4 with non-magnetic ions, the values of M S , remnant magnetization (M r ), H C , magnetic moment per formula unit (n B ), and K are higher than those of the undoped CoFe 2 O 4 , with few exceptions [1,4,26]. For the gels calcined at 1000 • C, the peak heights and their horizontal shifts are associated with the strength of the magnetic phases, with the broader peaks indicating a large particle size distribution accompanied by a large H C [21][22][23]. A significant increase in H C is observed on the hysteresis loops of the gels calcined at 1000 • C (they are much broader) compared to those at 700 • C. For the gels calcined at 1000 • C, the doping effect of the monovalent metals (Ag + , Na + ) increases the already large H C of the undoped CoFe 2 O 4 , while the doping effect of the divalent (Ca 2+ , Cd 2+ ) and trivalent (La 3+ ) ions leads to a decrease in H C , which is observable on the hysteresis loops of the gels calcined at 1000 • C.
The magnetic parameters M s , M R , H c , n B and K values determined using the hysteresis loops and M(H) curves are presented in Table 4. Generally, the M s for the spinel ferrites is dictated by the superexchange interactions between the A and B site cations. The M s decreases with the increase in the crystallite size due to the larger number of surface defects [55].  Due to the change in the magnetocrystalline anisotropy or the particle sizes by doping CoFe2O4 with non-magnetic ions, the values of MS, remnant magnetization (Mr), HC, magnetic moment per formula unit (nB), and K are higher than those of the undoped CoFe2O4, with few exceptions [1,4,26]. For the gels calcined at 1000 °C, the peak heights and their horizontal shifts are associated with the strength of the magnetic phases, with the broader Figure 9. Magnetic hysteresis loops of gels calcined at 700 • C. peaks indicating a large particle size distribution accompanied by a large HC [21][22][23]. A significant increase in HC is observed on the hysteresis loops of the gels calcined at 1000 °C (they are much broader) compared to those at 700 °C. For the gels calcined at 1000 °C, the doping effect of the monovalent metals (Ag + , Na + ) increases the already large HC of the undoped CoFe2O4, while the doping effect of the divalent (Ca 2+ , Cd 2+ ) and trivalent (La 3+ ) ions leads to a decrease in HC, which is observable on the hysteresis loops of the gels calcined at 1000 °C.  surface include forming of a dead layer, which contains broken chemical bonds; deviations from the bulk cation distribution; randomly oriented magnetic moments; lattice defects and non-saturation effects, resulting in depreciated magnetic properties [21].
Cd 0.1 Co 0.9 Fe 2 O 4 is a good candidate for various technological applications such as communication, data storage, and high-frequency inductors [1,4]. For both calcination temperatures, Ag 0.1 Co 0.95 Fe 2 O 4 exhibits the lowest M S value. Previous studies also reported that the doping of diamagnetic Ag + into the CoFe 2 O 4 spinel structure substantially decreases the M S of CoFe 2 O 4 ; a possible explanation for this is the high number of uncoordinated magnetic spins that are not able to align in the direction of the external magnetic field. Generally, the Ag doping enhances the nanoparticles' antibacterial activities, suggesting that Ag 0.1 Co 0.95 Fe 2 O 4 may be a potential candidate for antibacterial applications [4]. Moreover, considering the excellent electron conductivity of Ag, it is expected that Ag doping increases the catalytic activity of CoFe 2 O 4 [25].
As La 3+ is a non-magnetic ion, it does not participate in the exchange interactions with its nearest neighbor ion; thus, the superexchange interactions between the A and B sites' cations are depreciated [55]. Above the single-domain critical size, the competition between the magneto-static energy and the domain-wall energy favors forming domain walls and splitting the single-domain particle into multi-domain particles [22]. Mariosi et al. reported that the M S of the undoped CoFe 2 O 4 (44.6 emu/g) decreased to 29.0 emu/g for the first increase in the La 3+ concentration (sample CoLa 0.025 Fe 1.975 O 4 ); the possible mechanisms for the magnetic behavior of these nanoparticles are still widely discussed [6]. Moreover, a disorder in the crystal's surface results in a lack of collinearity of magnetic moments; this effect is generally attributed to a single magnetic domain configuration [6]. When non-magnetic La 3+ ions substitute Fe 3+ ions, the content of Fe 3+ ions at ferrite lattice sites is reduced, resulting in a decrease in the total magnetic moment and a weakening of the Fe 3+ -Fe 3+ interactions and, consequently, a lower M S value [9].
The remanent magnetization (M R ) for the undoped CoFe 2 O 4 calcined at 700 • C is 3.5 emu/g. The doping with monovalent cations (Ag + , Na + ) increases the M R to 7.7-7.9 emu/g, while doping with the divalent (Ca 2+ , Cd 2+ ) and trivalent (La 3+ ) cations decreases the M R to 1.8-3.0 emu/g. In the samples calcined at 1000 • C, the M R of the doped ferrites increases, compared to the M R of CoFe 2 O 4 (13.4 emu/g), except for the M R of Na 0.1 Co 0.95 Fe 2 O 4 , which slightly decreases (11.3 emu/g).
The slight decrease in the Hc values in the doped samples could result from the magnetocrystalline anisotropy, microstrain, size distribution and the decrease in the magnetic domain size [27]. For the gels calcined at 700 • C, the H C of the undoped CoFe 2 O 4 is 600 Oe, while that of the doped ferrites is lower, which is most probably due to changes in the crystallite size, anisotropy and formation of agglomerates that increase the average particle size above the critical single-domain, which results in a multi-domain structure and the reduction in pinning effects on the domain wall mobility at the grain boundary [22,55]. Oppositely, for the gels calcined at 1000 • C, the H c of the doped CoFe 2 O 4 with monovalent cations (Ag + si Na + ) are comparable, while for those doped with the di-and trivalent cations, the H c is lower than that of the undoped CoFe 2 O 4 (1750 Oe). The lower H c values of the obtained gels indicate a spin distortion on the surface, owing to the magnetocrystalline anisotropy [24]. The presence of SiO 2 generates stress on the surface of the ferrite particles, which hinders the rotation of the dead layer's magnetic moments and contributes to the reduction in the H c [21,22]. The increase in the surface potential barrier caused by crystalline lattice defects, such as the deviation of atoms from the normal positions in the surface layers, also determines the increase in the H C [29].
The n B of the gels calcined at 700 • C decreases from 0.935 (CoFe 2 O 4 ) to 0.857-0.815 by doping. Additionally, in the samples calcined at 1000 • C, except for doping with Cd 2+ and La 3+ , the n B of the doped samples is lower than that of the undoped CoFe 2 O 4 (0.977).
To calculate the magnetic anisotropy constant (K), we assumed that the spinel ferrite particles have a spherical shape. The value of K depends on the crystalline symmetry of the lattice, the crystalline anisotropy, and the particle size and shape [21]. The highest K was obtained for the undoped CoFe 2 O 4 (1.13·10 −3 erg/cm 3 in gels calcined at 700 • C, and 3.46·10 −3 erg/cm 3 in gels calcined at 1000 • C). The value of K increases with the increasing calcination temperature and decreases by doping. A possible explanation for this decrease could be the pinning of some surface spins in the magnetically disordered surface layer, which needs a higher magnetic field for magnetic saturation [22]. In addition, the magnetic disorder may originate in randomly oriented grains of different sizes and disordered vacancies [22]. The individual K of particles acts as an energy barrier and delays the switch of the magnetization direction to the easy axis [22]. Crystalline anisotropy is strongly affected by the volume strain in the crystal, which is determined by the substitution of Fe 3+ ions by the different sized (La 3+ ) ions [55].
To summarize, the embedding of the undoped and doped CoFe 2 O 4 in the nonmagnetic SiO 2 matrix promotes both the formation of single-phase spinel and minimization of the spin disorder and surface roughness, thus enhancing the magnetic properties of the ferrites. Combining the best magnetic properties and morphological configuration of the undoped and doped CoFe 2 O 4 can be of interest for several applications, such as high-density storage and biomedicine. Moreover, since SiO 2 is non-toxic, biologically inert, and widely accepted material by the living body, and even reduces the inflammatory risk, embedding the undoped and doped CoFe 2 O 4 could enhance their biocompatibility [21]. Although the properties of the obtained doped CoFe 2 O 4 could be further improved by optimizing the amount of dopant ions, the calcination temperature, or the SiO 2 -to-ferrite ratio, our study brings valuable baseline data on the properties of doped CoFe 2 O 4 -SiO 2 nanocomposites.

Characterization
The reaction progress was investigated via thermogravimetry (TG) and differential thermal analysis (DTA) in air, up to 1000 • C, at 10 • C·min −1 using alumina standards and a simultaneous SDT Q600 (TA Instruments, New Castle, DE, USA) thermal analyzer. The Fourier transform infrared (FT-IR) spectra of samples were recorded on KBr pellets containing 1% sample using a BX II FT-IR (Perkin Elmer, Waltham, MA, USA) spectrometer. The crystalline phases were investigated via X-ray diffraction using a D8 Advance (Bruker, Karlsruhe, Germany) diffractometer at ambient temperature with CuKα radiation (λ = 1.5418 Å), operating at 40 kV and 35 mA. The composition of gels calcined at 400, 7000 and 1000 was confirmed via Optima 5300 DV (Perkin Elmer, Norwalk, CT, USA) ICP-OES after microwave digestion using a Speedwave Xpert (Berghof, Germany) system. Atomic force microscopy (AFM) was performed using a JSPM 4210 (JEOL, Tokyo, Japan) microscope in tapping mode using an NSC 15 (Mikromasch, Sofia, Bulgaria) silicon cantilever with a nominal resonant frequency of 325 kHz and a nominal force constant of 40 N/m. Three different 1 µm x 1 µm areas of the thin films obtained by transferring nanoparticles onto glass slides via adsorption from aqueous suspension were scanned for each sample. Image processing and topography were performed using a WinSPM 2.0 software (JEOL, (Tokyo, Japan). Cantilever characteristics were considered in the particle size determination. The particles' morphology was visualized using an HD-2700 (Hitachi, Tokyo, Japan) transmission electron microscope (TEM). The magnetic measurements were performed using a 7400 vibrating sample magnetometer (VSM) (Lake Shore, Carson, CA, USA). The hysteresis loops were recorded at room temperature, up to an applied field of 2 T, while the magnetization (M) was measured in a high magnetic field of up to 5 T.

Conclusions
The influence of doping with monovalent (Ag + , Na + ), divalent (Ca 2+ , Cd 2+ ), and trivalent (La 3+ ) ions on the structural, morphological, surface, and magnetic properties of CoFe 2 O 4 was investigated. The kinetic formation of the doped and undoped CoFe 2 O 4 showed that the activation energy of CoFe 2 O 4 (1.236 kJ/mol) increased to 1.487-1.747 kJ/mol by Ag + , Ca 2+ , Cd 2+ , and La 3+ doping, and decreased to 1.102 kJ/mol by Na + doping, while the rate constant increased with the calcination temperature and depended on the doping ion. Poorly crystalline ferrites at 400 and 700 • C, and a well-crystallized single-phase ferrite in the undoped CoFe 2 O 4 at 1000 • C, were observed. By doping, besides the wellcrystallized ferrite, crystalline silica phases (cristobalite and quartz) were also formed. Although all the obtained gels have a cubic spinel structure, doping with different ions changed in the structural parameters determined via XRD. The AFM revealed that a low calcination temperature generated mainly spherical particles with a polycrystalline structure containing ferrite crystallites mixed with amorphous material. The increase in the calcination temperature led to a larger crystallite size, forming particles by a single-phase ferrite core covered by traces of secondary phases. The TEM measurements also indicate that thermal treatment is the main cause of the large size of the obtained nanoparticles; these results indicate a coalescence of nanoparticles, increasing the mean size. At 700 and 1000 • C, a single magnetic phase is generally observed, except in the case of doping with monovalent (Ag + , Na + ) ions at 1000 • C, when the formation of two magnetic phases is favored. Moreover, the magnetic parameters of the gels calcined at 1000 • C were higher than those at 700 • C. The doping with monovalent ions decreased the M S and increased the H C , while the doping with multivalent ions increased the M S and decreased the H C . The K value decreased with doping, with the undoped CoFe 2 O 4 displaying the highest anisotropy constant. The obtained results confirm that doping plays an important role in the tuning of the physical properties of promising CoFe 2 O 4 , which may be of great importance in the exploration of new applications in high-density information storage, drug delivery and tissue imaging.