Introduction

Crystalline ferrites having general formula MFe2O4 (M = divalent metal ion, e.g., Ni, Co, Cu, etc.) are one of the most attracting materials due to their various applications such as transformer core, antenna rod, recording head, loading coil, memory and microwave devices [1]. Practical applications of ferrites include household, electronics, and computers. Apart from their technological importance in the electronic and magnetic industries, ferrites exhibit properties such as various redox states, electrochemical stability, pseudocapacitive behavior, etc., and nowadays, they can be used in supercapacitors and in Li batteries [2, 3]. It is established that various binary and ternary spinel ferrites are effective catalysts for a number of industrial processes such as oxidative dehydrogenation of hydrocarbons [4], decomposition of alcohols and hydrogen peroxide [5], treatment of automobile-exhaust gases [6], oxidation of various compounds such as CO [7], phenol hydroxylation [8], alkylation reaction [9], etc.

Nickel ferrite (NiFe2O4) with an inverse spinel structure shows ferrimagnetism, which originates from magnetic moment of anti-parallel spins between Fe3+ ions at tetrahedral sites and Ni2+ ions at octahedral sites [10]. As more and more attentions have been devoted to the nano-sized magnetic materials for their unique and amazing properties compared to their bulk counterparts, the scientific interests were focused on nano-sized nickel ferrite [11]. NiFe2O4 is also employed in a wide variety of catalytic and photocatalytic processes such as total oxidation of propane [12] and toluene [13], photocatalytic water remediation [14], decomposition of N2O [15], oxidative dehydrogenation of 1-butene to 1,3-butadiene [16], and methane combustion [17]. Moreover, this catalyst showed high degradation efficiency of di-n-butyl phthalate in ozonation catalytic system with convenient separation [18]. Nanoferrites are usually prepared by various physical methods [19, 20], as well as chemical methods like mechanical milling, inert gas condensation, hydrothermal reaction, oxidative precipitation, and sol–gel synthesis techniques [21, 22]. The sol–gel auto-combustion route, as a unique combination of the combustion and the chemical gelation process, is preferred in general because of its main advantages of inexpensive precursors, short preparation time, modest heating, and relatively simple manipulations [23–25].

Synthesis procedures of spinel ferrite nanoparticles were intensively studied in recent years. Large-scale applications of ferrites with nanoparticles and tailoring of specific properties have prompted the development of several widely used chemical methods [26] like sol–gel [27], host template [28], co-precipitation [29], and hydrothermal treatment [30]. Addition of cationic surfactants with hydrophilic head and hydrophobic tail into the solution results in the formation of reverse micelles [31]. Placing the aqueous ions inside these micelles can affect the controlling of the nucleation and growth of the particles. Due to the existence of surfactant, the surface tension of solution is also reduced, which lowers the energy of the formation of the new phase [31].

In the present work, a new synthesis method of NiFe2O4 was carried out by adding different amounts of polyvinyl alcohol (PVA) to citric acid (CA) via employing the sol–gel auto-combustion technique, which resulted in lowering the ignition temperature of formation of the pure crystalline NiFe2O4 phase. The synthesized catalysts were investigated by Fourier transform infrared (FT-IR) spectroscopy, Thermogravimetric and differential thermal analyses (TG/DTA), X-ray diffraction (XRD), and scanning electron micrograph (SEM) measurements. The catalytic hydroxylation of benzene to phenol in the presence of hydrogen peroxide was chosen as a test reaction and the effect of reaction parameters was investigated.

Experimental

Materials

The used chemical reagents are ferric chloride (Merck), nickel chloride (Merck), ammonia solution, CA, PVA (Fluka) of molecular mass (125,000), benzene, hydrogen peroxide, 4-aminoantipyrine, and potassium ferrocyanate (Fluka).

Synthesis of nickel ferrite (NiFe2O4)

NiFe2O4 was prepared by the sol–gel auto-combustion method. 8.5 mol of iron chloride and 3.6 mol nickel chloride powders in a molar ratio of Fe/Ni = 2.63 were mixed together with 8.5 mol of CA, wherein CA helps the homogenous distribution of the metal ions throughout the solution by forming a complex with Fe(III). After adding this solution into 200 mL of deionized water, different amounts of PVA (0.1–0.3 mol%) were added, followed by carefully dipping little drops of ammonia solution to the solution to attain the pH 7. The mixed solution was heated at 353 K for 3 h with continuous stirring to allow a sol–gel reaction to occur through condensation reaction between the metal ions and the PVA to yield a composite network in colloidal state known as sol. The resultant sol was obtained and then dried in an oven at 393 K for 8 h. The dried samples were heated in an oven air for 6 h at 673, 873, and 1173 K, respectively. The samples were referred as PVA-free NiFe2O4 for PVA-free sample and 0.1PVA/NiFe2O4, 0.2PVA/NiFe2O4, and 0.3PVA/NiFe2O4 where the onset of abbreviates indicate mol% of the loaded PVA.

Physicochemical characterization

XRD data of the samples were measured at room temperature using a Philips diffractometer (type PW 3710). The patterns were run with Ni-filtered copper radiation (λ = 1.5404 Å) at 30 kV and 10 mA with a scanning speed of 2θ = 2.5° min−1. The mean particle size (D) was calculated using the Debye–Scherrer Eq. (1) [32], in which K is a constant equal to 0.9, λ is the wavelength of the Cu Kα radiation, β is the half peak width of the diffraction peak in radiant, and theta is the Bragg scattering angle.

$$ D = K\lambda \beta \cos \theta . $$
(1)

The FT-IR spectra were recorded on a Perkin Elmer Spectrum (RXI FT-IR) system, single-beam spectrometer with a resolution of 2 cm−1. The samples were ground with KBr (1:100) as a tablet and mounted to the sample holder in the cavity of the spectrometer. The measurements were recorded at room temperature in the region 4,000–400 cm−1.

Morphology and size of the prepared materials were determined using SEMs for products obtained on a JEOL scanning microscope model JSM-T 330A at an accelerating voltage of 30 kV.

TG/DTA were carried out using Shimadzu-50 thermal analyzer units. The sensitivity of TG and DTA measurements was 0.01 mg and 25 μv, respectively. In each run about 10 mg of uncalcined sample was heated from 303 to 1,273 K at a heating rate of 10 K min−1 under a stream of flowing nitrogen at a rate of 30 mL min−1.

Thermal kinetic analysis

The thermal data under investigation were analyzed according to Kissinger equation [33]:

$$ \beta \frac{{{\text{d}}\alpha }}{{{\text{d}}t}} = A{\text{e}}^{{ - E_{{\text{a} /\text{RT}}} }}\, f(\alpha ), $$
(2)

where α is the fraction reacted (alternatively called conversion degree), T is the absolute temperature, β is the heating rate (β = dT/dt, where t is the time), A is the pre-exponential factor, E a is the activation energy, R is the gas constant, and f(α) is the differential form of the reaction model, i.e., a function of α which well fits the experimental data.

The kinetic analysis was performed by means of software kinetic program [34], which includes the majority of reaction models, ranging from 1st to nth order, nucleation-growth model, as well as the activation energy distribution models, and uses a non-linear regression method to determine the parameters of Eq. (2) to fit the experimental data. To check the presumed invariance of E a on α, the software kinetic program performs the isoconversional “model-free” analysis by means of expanded Horowitz–Metzger (HM) and modified Coats–Redfern (CR) methods [35]. Multi-heating rate application of the CR equation may be expressed by the equation:

$$ \log \left[ {\frac{{\log \{ f|(W_{\text{f}} - W)\} }}{{T^{2} }}} \right] - \log \left[ {\frac{AR}{{\theta E^{*} }} \left( {\frac{1 - 2RT}{{E^{*} }}} \right)} \right] - \frac{{E^{*} }}{2.303RT}, $$
(3)

where W f is the mass loss at the completion of the reaction, W is the mass loss at Kelvin temperature T, R is the gas constant, E* is the activation energy in kJ mol−1, θ is the heating rate, and (1 − 2RT/E*) ≈ 1. A plot of the left-hand side of Eq. (3) against 1/T gives a slope from which E* was calculated and A was determined from the intercept.

Alternatively, we applied the HM equation [36, 37] to the thermal data for the materials under study which may be written in the form:

$$ \log \left( {\frac{{W_{\infty } }}{{W_{\text{r}} }}} \right) = \frac{{\theta E^{*} }}{{2.302RT_{\text{s}}^{2} }} - \log 2.303, $$
(4)

where T s is the DTG peak temperature and θ = T − T s. A plot of log [logW ∞/W r] versus θ will give a straight line with a slope from which E* can be calculated. The pre-exponential factor A was calculated from the following equation:

$$ C = \left( {\frac{{\theta E^{*} }}{{RT_{\text{s}}^{2} }}} \right) \exp \left( {\frac{{E^{*} }}{{RT_{\text{s}} }}} \right). $$
(5)

The activation entropy (ΔS*), activation enthalpy (ΔH*), and the activation free energy (ΔG*) were calculated using the following equations [38]:

$$ \Delta S^{*} = 2.303\left[ {\log \left( \frac{Ah}{kT} \right)} \right], $$
(6)
$$ \Delta H^{*} = E^{*} - nRT, $$
(7)
$$ \Delta G^{*} = \Delta H^{*} - T\Delta S^{*} , $$
(8)

where k and h are the Boltzmann and Planck constants, respectively. The calculated values of E*, A, ΔS*, ΔH*, and ΔG* for the different decomposition steps are given in Table 2.

Catalytic tests

Liquid-phase hydroxylation of benzene to phenol with hydrogen peroxide over the prepared catalysts was conducted in a 250 mL batch glass reactor equipped with a condenser under atmospheric pressure. The catalyst was dried at 383 K in an oven to remove water before weighing into the reactor. 20 mL additional water, apart from that obtained by the H2O2 solution, was normally used, unless stated otherwise, for better dispersion of the catalyst. Magnetic stirring bar and additional water were added before placing the reactor into the oil bath and connecting with the condenser. After heating up this system to reaction temperature, liquid benzene was injected, thereby avoiding excess benzene evaporation during the heating up process. H2O2 was slowly added dropwise to the reactor with the help of an appropriate syringe during the first 10 min of the reaction. In a typical reaction, 0.011 mol of H2O2 and x mol of benzene (x = 0.011, 0.022, 0.033, 0.044, or 0.055 mol) were used. The reaction was carried out for 3 h at 343 K under vigorous stirring. After the reaction has stopped, the concentration of phenol being obtained solely was determined at λ max = 510 nm using spectrophotometery by the 4-aminoantipyrine method [39, 40].

Results and discussion

XRD analysis

The XRD examination of the thermally treated 0.2PVA/NiFe2O4 at 673 and 873 K showed strong diffractions at d-spacings 3.68, 2.71, and 1.69 Å related to α-Fe2O3, the others at 2.95 and 4.82 Å are due to NiFe2O4 (Fig. 1). Raising the calcination temperature to 873 K results in the increase of the peak intensity at d = 2.95 Å characteristic of NiFe2O4 phase with a subsequent decrease in the peak intensity of free α-Fe2O3 phase at d = 2.51 Å. A further increase in temperature to 1,173 K simulates only the existence of the nickel ferrite phase through vanishing reflections of α-Fe2O3 phase. The average size of the NiFe2O4 crystals varied relatively from 47 nm at 673 K to 145 nm at 1,173 K.

Fig. 1
figure 1

XRD patterns of CA/PVA/NiFe2O4 (molar ratio PVA:CA = 5:1) calcined at (a) 673 K, (b) 873 K, and (c) 1,173 K

Full size image

The effect of PVA loadings on the crystalline phase composition of NiFe2O4 in comparison with PVA-free NiFe2O4 is shown in Fig. 2. The PVA-modified NiFe2O4 samples generally exhibit reflections of various planes indexed as (220), (311), (222), (400), (422), (511), and (440) owing to the presence of the cubic spinel structure. This result imparts a clear evidence for the NiFe2O4 formation according to the standard JCPDS (Card No. 10-325) [41]. Phase analysis of the 0.1PVA/NiFe2O4 pattern revealed the presence of α-Fe2O3 at d spacings 2.949, 2.698, and 2.203 Å, along with key lines of NiFe2O4. When the PVA concentration increased to 0.2–0.3 mol%, a single phase comprising merely of NiFe2O4 was depicted; emphasizing the beneficial effect of PVA in eliminating impurities that deteriorate the formation of a distinct phase NiFe2O4.

Fig. 2
figure 2

XRD patterns of (a) PVA-free NiFe2O4, (b) 0.1PVA/NiFe2O4, (c) 0.2PVA/NiFe2O4, and (d) 0.3PVA/NiFe2O4

Full size image

The most intense XRD peak at 2θ = 36.16° and its respective plane at d = 2.483 Å corresponding to the (311) plane of NiFe2O4 are indicative of the presence of inverse spinel structure [42]. As reported earlier, the intensities of the (220), (440), and (511) planes are more sensitive to the cations in tetrahedral (A) and octahedral (B) coordinated sites, surrounded by oxygen atoms in the composite oxide [43, 44]. Fe3+ ions have tendency to occupy A sites, while Ni2+ ions appear to compete with Fe3+ ones to occupy B sites in the mixed ferrites. It is likely that the high concentrations of PVA significantly stabilize the Fe3+ ions in A sites in which the composition exhibit single-phase cubic spinel structure. Table 1 shows the observed intensities of the above three planes. The intensity of the (440) peak was observed to increase as the concentration of PVA increased; which infers the Ni2+ ions occupying the B site, i.e., the octahedral sites on the (440) plane. However, the intensity of the (220) plane increases by the continuous PVA addition indicating the preferential occupation of A sites by Fe3+ ions since NiFe2O4 is established to be an inverse spinel structure. Therefore, any distortion in the structure is caused by the migration of more number of Ni2+ ions to the tetrahedral site during sintering. The crystallite sizes of the synthesized powders are listed in Table 1 using the X-ray broadening of the (311) diffraction peak, via the well-known Scherrer’s equation.

Table 1 The effects of addition of PVA to citric acid on the intensity values of hkl planes of NiFe2O4 phase
Full size table

It must be noted that the citrate method may occur statistically at some regions of the nanoparticles causing a minor distortion. Thus, the steps taken for the formation of a single-phase ferrite by maintaining the citrate method are unsuccessful. Inversely, it seems that combining CA with PVA will stimulate the improved preparation of single-phase NiFe2O4. As the carbon chains in CA and PVA decompose during combustion, the adjacent atoms which are homogeneously distributed throughout the matrix can easily come into contact and form crystal lattice at a considerable rate.

IR analysis

The synthesis of the ferrite spinel NiFe2O4 by treatment primarily with CA solution in the absence and presence of PVA polymeric intermediates was probed by FT-IR spectroscopy, as shown in Fig. 3. All the spectra characterized by two bands at 597 and 410 cm−1, which are assigned to stretching vibration of overlap integrals between Fe3+ and O2− ions, involve tetrahedral and octahedral coordinations in typical inverse spinel ferrites, respectively [19, 45]. The O–H stretching vibration associated with CA/PVA is observed at 3,440 cm−1 for all the samples. However, the greater the concentration of the PVA, the more intense and broader the absorption of this band, which can be a result of the superposition of multiple hydrogen bonds associated with the PVA crystalline phase [30]. The observed bands at 2918, 2852, and 1457 cm−1 are assigned to the anti-symmetric, symmetric, and bending of CH2 vibrations, respectively, of the carbon chains in the PVA structure [46].

Fig. 3
figure 3

FT-IR absorbance spectra of (a) PVA-free NiFe2O4, (b) 0.1PVA/NiFe2O4, (c) 0.2PVA/NiFe2O4, and (d) 0.3PVA/NiFe2O4

Full size image

The ν4 vibration doublet at approximately 1,644 and 1,609 cm−1 for the protonated complex ≡Fe–O–COOH [47] was observed for 0.2PVA/NiFe2O4 and 0.3PVA/NiFe2O4 samples (Fig. 3c, d). It is noteworthy that this species grew largely as a function of PVA concentration, though the bands of this species were absent for PVA-free NiFe2O4 (Fig. 3a), which can verify the interaction between the Fe(III) ions and the CA/PVA matrix. This result was further confirmed by the presence of two bands at 1,549 and 1,401 cm−1 in the spectrum of the latter sample which are assigned to asymmetric and symmetric vibrations of the carboxylate –COO− groups, respectively. When compared with the frequencies of free carboxylate, these bands are shifted downward by small values to make firmer the presence of metal–carboxylate adducts through Fe(III) ions-chelating CA/PVA [48, 49]. The peak at 1,740 cm−1 for all samples prepared by adding PVA was assigned to the carbonyl stretch C=O group. The bands at 1,136 and 1,024 cm−1 in Fig. 3a, c, d are attributed to the organic network or COH groups [37]. These bands did not appear in the spectrum of 0.1PVA/NiFe2O4 sample, indicating the formation of pure NiFe2O4 compound at this concentration of incorporated PVA, i.e., 0.1 mol%. Meanwhile, the observed intense IR peaks of this organic species in the other samples indicate the incomplete combustion of the organic intermediates.

Thermal analysis

Thermal degradation of nickel ferrite–CA samples with respect to PVA concentration was studied by TG analysis (TG/DTG) (Fig. 4). The TG curve of PVA-free NiFe2O4 exhibits mass loss which occurs in two stages of temperatures in the range 303–1,273 K. The first mass loss (62.11 %) took place between 404 and 718 K due to the evaporation of absorbed water and decomposition of CA. The second stage indicates mass loss (2.33 %) at 1,022–1,268 K due to simultaneous condensation of the lattice oxygen and formulation of carbon metals/metal oxides. The addition of PVA (0.1–0.3 mol%) to form NiFe2O4 led to three mass loss stages in the temperature range 303–1,273 K. The first loss that occurred at 303–573 K is due to the decomposition of CA and evaporation of water molecules through dehydration reaction of PVA in the local segments of its molecular chain in which the main PVA could be turned into polyacetylene (–CH=CH–) n . The second stage at 579–885 K showed increased mass losses comprising 28.275 % for 0.1PVA/NiFe2O4, 21.996 % for 0.2PVA/NiFe2O4, and 52.835 % for 0.3PVA/NiFe2O4 due to mainly the elimination reactions that happened between residual acetate groups and partially can be products of hydrolyzed PVA. The third stage that experienced a mass loss of 2.958 % at 804–1,243 K in the profile of 0.3PVA/NiFe2O4 is indicative of further degradation process to yield hydrocarbons and metallic residues [50].

Fig. 4
figure 4

TG and DTG profiles of (a) PVA-free NiFe2O4, (b) 0.1PVA/NiFe2O4, (c) 0.2PVA/NiFe2O4, and (d) 0.3PVA/NiFe2O4

Full size image

DTA curves (Fig. 5) obtained for PVA-free NiFe2O4, and 0.1PVA/, 0.2PVA/, and 0.3PVA/NiFe2O4 showed exothermic peaks at 1150, 1099, 1105, and 1073 K, respectively, which are attributed to different degradation processes [51]. Upon addition of PVA, a random change in glass-transition temperature can be achieved as compared with PVA-free NiFe2O4 that presented a greater decomposition temperature comparatively. This suggests that the segmental mobility of amorphous PVA becomes more rigid [52]. The concurrent lowering of decomposition temperature from 1,150 to 1,073 K upon increasing the PVA concentration may be due to reduction in molecular mass of PVA and/or breaking of inter- and intra-molecular PVA hydrogen bonding [53].

Fig. 5
figure 5

DTA profiles of (a) PVA-free NiFe2O4, (b) 0.1PVA/NiFe2O4, (c) 0.2PVA/NiFe2O4, and (d) 0.3PVA/NiFe2O4

Full size image

The kinetic parameters are listed in Table 2 and the following remarks can be pointed out: (i) the correlation coefficient, r, which was computed using the least square method was higher than 0.99, indicating a good linearity for the experimental data; (ii) for most of samples, the negative ΔS* values indicate an ordered activated intermediate in spite of tending to show a disorder behavior as the PVA concentration increases [54]. However, the trend that directs to the disorder by increasing the concentration of PVA may be caused in consequence of water loss. The positive values of ΔH* mean that the decomposition processes are endothermic ones; (iii) as shown in Table 2, the E a values at the regarded temperatures revealed that the water and anion molecules are easily eliminated by increasing the PVA amounts, as shown by the following sequence:

$$ 0.3{\text{PVA}}/{\text{NiFe}}_{2} {\text{O}}_{4} > 0.2{\text{PVA}}/{\text{NiFe}}_{2} {\text{O}}_{4} > 0.1{\text{PVA}}/{\text{NiFe}}_{2} {\text{O}}_{4} > {\text{PVA free}}\;{\text{NiFe}}_{2} {\text{O}}_{4} . $$

The values of E a for all the samples somewhat refer to a weak bonding between NiFe2O4 and PVA molecules. Besides, the E a values of the first step are lower than the other steps as a result of its high rate of decomposition [55]; (iv) the values of the activation free energy ΔG* which increase significantly for the subsequent decomposition stages for nickel ferrites indicate the non-spontaneous behavior. This is due to increasing the values of TΔS* significantly from one step to another and thus overriding the values of ΔH*. Increasing the values of ΔG* for the subsequent steps reflects increasing either CA or PVA removal rate which will be lower than that of the precedent citrate with PVA. This may be attributed to the structural rigidity of the remaining nickel ferrite after the expulsion of one species and/or more (citrate or PVA), as compared with the precedent sample, which requires more energy, TΔS*, for its rearrangement before undergoing any compositional change; and (v) there is an apparent gap in the values of the heat of activation E* and the enthalpy of activation ΔH* of volatilized of species between 0.3PVA/NiFe2O4 (ΔH* = 29.62 kJ mol−1) and PVA-free sample (ΔH* = 46.84 kJ mol−1). This may be attributed to the packing structure of NiFe2O4 which may allow stronger interactions of CA molecules in the crystal.

Table 2 Decomposition temperature, order, and activation parameters of NiFe2O4 prepared by different concentrations of PVA
Full size table

Scanning electron microscopy (SEM)

SEM analysis was performed in order to depict the microstructure of the synthesized NiFe2O4 through treatment primarily with CA solution, in the absence and presence of different loadings of PVA polymeric intermediate; and calcination at 1,173 K (Fig. 6). For PVA-free NiFe2O4, the SEM micrographs exhibited foamy-like agglomerated particles with a wide distribution, along with the presence of large voids in the matrix (Fig. 6a). On the other hand, the image of 0.1PVA/NiFe2O4 was characterized by the presence of both large and small spherical particles that are distributed in a homogeneous manner in this sample (Fig. 6b). The former particles may correlate to α-Fe2O3, compatible with XRD. However, this image looks different from that of the main structure of ferrites. When PVA loading increased (i.e., 0.2PVA/NiFe2O4), the spheriform shape of the particles as well as slight agglomeration of the very fine particles emerged due to intrinsic nickel ferrite. This result strongly supports that the addition of PVA plays a structure-directing agent which improves the nucleation. The enhanced nucleation has thus led to the formation of smaller particle size.

Fig. 6
figure 6

SEM micrographs of a PVA-free NiFe2O4, b 0.1PVA/NiFe2O4, and c 0.2PVA/NiFe2O4

Full size image

Formation mechanism of NiFe2O4

Both types of chelating agents, namely PVA (hydroxyl chelating agent) and CA (carboxyl chelating agent) were used to coordinate Ni(II) and Fe(III) to form either a mono- or bi-dendate metal complexes [56], in which undesired spontaneous condensation reactions would be prohibited [57]. This can be visually verified through color changes observed when ferric and nickel precursors are added to water encompassing PVA with CA at ambient conditions. In the absence of PVA with CA, ferric chloride is oxidized by oxygen-enriched air and the solution turned turbid after a short period of time. However, no color change is observed when ferric chloride is added to the PVA solution, indicating that Fe3+ cations are sterically entrapped in the entangled network of the organic polymer [58].

The use of chelating agents was essential for the production of more nickel ferrite phases in the final product. Two possible mechanisms can explain this observation. The first mechanism is associated with the molecular structure of the hybrid organic–inorganic complex of both nickel and iron ions, since the chelating agent might prevent any initial condensation or any cross-linking between PVA and CA [59] during the preparation stage. The strongly charged ferric ions are neutralized by the PVA and CA anions, and would diffuse into their integrity until they become uniformly surrounded by the less charged nickel ions. This process of diffusion ensured that the energy of the system would approach a minimum value.

The second mechanism involved an enhanced T g of the solution while oxidizing PVA during heat treatment. The melting enthalpy (ΔH) decreased (from 1,150 to 1,073 K) as the PVA concentration increased, suggesting reduction in intermolecular forces in the presence of CA-crosslinked PVA networks. According to phase investigation of 0.2PVA by XRD and SEM, we found that this sample exhibited majority of the nickel ferrite phase. Plausibly, the chelating agents influence the reaction temperature of phase formation. It may also enhance the dynamics of the oxidation reaction from the ground state to the optimum condition to form molten nickel ferrite phase. Note that the PVA concentration must be sufficient to obtain a uniform distribution of cations, and hence isolate the particles by a polymer layer [60], as shown by SEM analysis (Fig. 6). The weak hydrogen bonding in the hydrated PVA also plays a role in promoting homogeneous physical entrapment between the hydroxyl groups and cations, which are solvated by water molecules [58], thereby hindering aggregation and resulting in a relatively uniform particle size distribution. Moreover, the chelate effect of mutual hydroxyl and carboxyl chelating agents of CA–PVA appears more effective than carboxyl chelating agent of CA merely in producing more completely molten splats and nickel ferrite phase.

Catalytic activity

Composite catalysts of NiFe2O4 prepared by adding varying amounts of PVA (0.1–0.3 mol%) have been used for the one-step direct hydroxylation of benzene toward phenol in the presence of hydrogen peroxide (30 % aq.) as a green oxidant at 353 K as water is the only by-product (Eq. (9)). The phenol concentrations produced from the hydroxylation reaction have been plotted against time (Fig. 7). The 0.1PVA/NiFe2O4 catalyst exhibited higher activity in benzene hydroxylation compared to other catalysts [18, 61, 62].

(9)
Fig. 7
figure 7

Hydroxylation of benzene as a function of type of catalyst; reaction conditions: pressure 1 atm, temperature 353 K, catalyst amount 0.10 g, concentration of H2O2 0.011 mol, concentration of benzene 0.022 mol, and reaction time 1 h

Full size image

Influence of catalyst amount

Figure 8 shows the relationship between conversion of substrate and the amount of 0.1PVA/NiFe2O4 catalyst in the hydroxylation of 50 mL pure benzene. It indicates that the catalytic activity increases to produce a higher yield of phenol (95 %) with increasing amount of catalyst till 0.1 g. Hence, the reaction rate of this reaction is dependent on the catalyst amount [63, 64].

Fig. 8
figure 8

Hydroxylation of benzene to phenol using different amounts of 0.1PVA/NiFe2O4 catalyst, reaction conditions: pressure 1 atm, temperature 353 K, catalyst amount 0.010, 0.015, 0.10, and 0.20 g, concentration of H2O2 0.011 mol, concentration of benzene 0.022 mol, and reaction time 1 h

Full size image

Effect of molar ratio of benzene/H2O2

As illustrated in Fig. 9, benzene conversion on 0.1 g of 0.1PVA/NiFe2O4 was found to increase with increasing benzene/H2O2 molar ratio up to 2:1, reaching 96 % conversion and level-off at this ratio even though the molar ratio was further increased to 4:1.

Fig. 9
figure 9

Conversion of benzene to phenol over 0.1PVA/NiFe2O4 as a function of molar ratio of benzene:H2O2, reaction conditions: pressure 1 atm, temperature 353 K, catalyst amount 0.1 g, concentration of H2O2 0.011 mol, concentration of benzene 0.011, 0.022, 0.033, 0.044 and 0.055 mol, 20 mL water, and reaction time 1 h

Full size image

Effect of added water to the reaction mixture

Figure 10 illustrates the effect of adding water to the reaction mixture composed of benzene/H2O2 and 0.1 g of 0.1PVA/NiFe2O4 on benzene conversion into phenol. Increasing the volume of added water to 30 mL causes a significant increase in the conversion percentage. However, blank experiments in the absence of H2O2 indicated that no direct reaction occurred under these conditions between either benzene and water or phenol and water. Increasing water volume to 40 mL decreases the conversion. This can be explained in the sense that water helps to strip phenol away from the catalytically active sites, thereby preventing further oxidation of phenol.

Fig. 10
figure 10

Conversion of benzene to phenol as a function of added water over 0.1PVA/NiFe2O4, reaction conditions: pressure 1 atm, temperature 353 K, catalyst amount 0.1 g, concentration of H2O2 0.011 mol, concentration of benzene 0.022 mol, and reaction time 1 h

Full size image

Conclusions

PVA was found to be an effective surfactant for the hydrothermal synthesis of NiFe2O4 at a molar ratio of Fe:Ni = 2.63 (w/w) in the presence of consistent concentration of CA. Distinctive preparations refer to the conditions of the calcination and period of aging. While calcination at 1,173 K exhibits only a single phase of NiFe2O4, lowering the temperature to 673–873 K presumed the formation of Fe2O3 and NiFe2O4 phases. As the period of aging was increased at 1,173 K, an enhancement generally occurred in crystallinity as revealed from the XRD analysis. The kinetic parameters established on both CR and HM “model free” methods indicated considerable variation of the kinetic parameters E* and A based on the diverse concentration of PVA-doped NiFe2O4. From these kinetic parameters, the elimination of water and organic CA/PVA species from nickel ferrites at the first temperature loss stage was facilitated according to the following order: 0.3PVA/NiFe2O4 > 0.2PVA/NiFe2O4 > 0.1PVA/NiFe2O4 > PVA-free NiFe2O4. The E* values for the first stage is lower than the analogous stages as a result of high decomposition rate of CA/PVA. The lower ΔH* value for 0.3PVA/NiFe2O4 (29.62 kJ mol−1) than that for PVA-free NiFe2O4 (46.84 kJ mol−1) may be attributed to the packing structure of NiFe2O4 which may allow stronger interaction of CA molecules with the PVA-free sample. The one-step direct hydroxylation of benzene with H2O2 has been extensively studied over the various preparations and demonstrated a range of catalytic activities following the order: 0.1PVA/NiFe2O4 > 0.2PVA/NiFe2O4 > 0.3PVA/NiFe2O4 > PVA-free NiFe2O4.