Study of the Catalytic Activity and Surface Properties of Manganese-Zinc Ferrite Prepared from Used Batteries

Department of Inorganic Chemistry, Faculty of Engineering and Economics, Wrocław University of Economics, Komandorska 118/120, PL53345 Wrocław, Poland Institute of Low Temperature and Structure Research, Polish Academy of Sciences, Okólna 2, PL50422 Wrocław, Poland Department of Advanced Material Technologies, Faculty of Chemistry, Wrocław University of Science and Technology, Wybrzeże Wyspiańskiego 27, PL50370 Wrocław, Poland Section of Waste Technology and Land Remediation, Faculty of Environmental Engineering, Wrocław University of Science and Technology, Wybrzeże Wyspiańskiego 27, PL50370 Wrocław, Poland


Introduction
One of the significant aspects of waste management is the recycling of worn out batteries and accumulators.e main aim of the battery industry is to reduce the negative impact of the used batteries, accumulators, and other portable cells on the environment by reducing the amount of hazardous substances in batteries and accumulators and enabling the proper collection and recycling of battery waste.Due to the existing regulations and the increasing amount of battery and accumulator waste, new technologies for the recycling of used cells, or modifications of existing methods, are being sought.
e processing methods are mainly based on pyrometallurgical or hydrothermal processes [1][2][3].Recently, there have been many scientific studies focused on improving the efficiency of the hydrometallurgical recovery of zinc and manganese from battery waste by using additional reducing agents (e.g., oxalic acid or urea) in the sulphuric acid (VI) leaching process [4,5], the selective leaching of zinc with sodium hydroxide [6], or leaching with a mixture of sulphuric acid or ammonia with the addition of SO 2 [7].Besides the studies on the recovery of Zn, MnO 2 , or ferromanganese alloy from Zn-Mn and Zn-C batteries [5][6][7][8][9], there are also studies on the processing of battery waste by chemical methods leading to prepared Mn-Zn ferrites, soft magnetic materials commonly used in microelectronics.In this case, used batteries were subjected to mechanical crushing and then leached with sulphuric acid [10,11], hydrochloric acid [12], or nitric acid [13,14].Ferrite can be prepared from such solutions by co-precipitation [15][16][17], when sodium hydroxide or ammonium oxalate is used as a precipitating agent.Attempts to synthesise ferrite powders by combustion methods in the presence of citric acid as a fuel and nitrate ions (derived from nitric acid leaching of crushed battery waste) as an oxidant have also been described [13,14,18].e studies [13,14,18] were focused on determining the degree of leaching of Mn, Zn, and Fe, the main components of ferrite (the content of the other elements in the solution after leaching was not studied), after which the microstructure and phase composition of the obtained products were characterised.Nan et al. [10] and Kim et al. [16] also investigated the magnetic properties of the ferrites prepared from waste batteries and found that the saturation magnetization of the powders obtained by coprecipitation from the solution after leaching battery scrap is similar to that of Mn-Zn ferrites synthesised by other chemical methods.Cheng-hong et al. [11] described the Mn-Zn ferrite prepared from battery waste with magnetic properties that were not much worse than those of typical ferrites produced on a large scale by a ceramic method.e effect of the precipitating agent on magnetic and thermal properties was also studied in [19].
Simultaneously, the recent literature points to new potential application areas of ferrite synthesised by chemical methods from pure reagents, such as adsorbents [20], catalysts [21][22][23], gas sensors [24], drug delivery [25], and in thermotherapy of cancers [26][27][28].Ferrite powders of a metastable character can be obtained by low-temperature synthesis and can be used in catalysis due to their high reactivity/activity [29,30].Based on our previous studies conducted on model systems, we found that the prepared Mn-Zn ferrites are catalytically active in the test of butan-1ol conversion [23,31].However, there are no data on the catalytic activity of such materials derived from secondary sources.
In this study, Mn-Zn ferrites were prepared by coupled co-precipitation and combustion methods from solutions after the acid leaching of battery waste.e method proposed in this paper, Mn-Zn ferrite synthesis, is an attempt to broaden the technology of processing battery waste by hydrometallurgical methods.A careful analysis of the microstructure and surface condition with regard to acid-base properties and redox centres allowed the differences in the catalytic activity of the prepared materials to be discovered and explained.

Mn-Zn Ferrite Preparation.
e procedure of leaching Zn-C and Zn-Mn battery scraps was described in detail in [4,32].e quantity and quality of metal ions in the solution after leaching (determined by ICP-AES) are shown in Table 1.
e sulphate solution after leaching was a source of metal ions in the subsequent stages of preparation of the ferrite.Precursors of metal ions (Fe, Mn, and Zn) from the sulphate solution were obtained by co-precipitation.ere were three precipitating agents applied: ammonium oxalate, sodium carbonate, and sodium hydroxide to precursor preparation.
e conditions of co-precipitation for the precipitating agents used are given in Table 2.
e ions deposited (precursors) were subjected to combustion processes.Table 2 gives the ratios of citric acid to nitrate ions in the experiments performed, depending on the ion precursor used.e pH � 7 was adjusted by adding aqueous ammonia (25 vol.%).e resulting sols were heated to evaporate the solvent and form a gel.en, the gel was placed in a dryer for 10 days at 70 °C.Subsequently, the gels were heated, which led to spontaneous ignition, giving a free-flowing powder of ferrite.

Mn-Zn Ferrite Characterisation.
e inductively coupled plasma-atomic emission spectrometry was conducted to determine the chemical composition of solution after leaching and co-precipitation.Analysis was performed on ICP-AES iCAP 7400 spectrometer ( ermo Scientific).e identification and phase composition of the prepared ferrite powders was determined by X-ray diffraction analysis at room temperature on a Siemens D5000 device with a scintillation counter (using CuKα radiation).Measurements were done at an angle range of 10-70 °and 30-40 °.  e average crystallite size (d XRD ) was estimated from XRD line broadening for the (311) peak using the Scherrer equation: d XRD � 0.9λ/B cos θ, where d XRD is the mean crystallite size in the direction perpendicular to the (hkl) plane of reflexes in nm and K � 0.9 is a Scherrer constant (λ � 0.154 nm is the X-ray wavelength used in the measurement;

􏽱
, where β FWHM and β 0 is the full-width at half-maximum (FWHM) of diffraction peak at angle θ and the corrected instrumental broadening (in radian), respectively).Scanning electron microscopy (FEI Quanta TM 250) and transmission electron microscopy (Philips CM20 SuperTwin) allowed the morphology and agglomerate size/grain size to be determined.All of the samples for scanning electron microscopic observation were sputtercoated with a thin (∼10 nm) layer of carbon.e samples were deposited on a Cu grid with a carbon film.
XPS surface analysis was performed using a SPECS PHOIBOS 100 spectrometer and Mg anode as X-ray source (1253.6 eV), operating at 250 W (high-resolution spectra).Surface cleaning during measurements was carried out by Ar + sputtering with the beam energy of 2 keV and a beam current density of 4.1 µA•cm −2 .Spectra were processed and fitted by the SPECLAB software using Gaussian-Lorentzian curve profile and Shirley baseline.e accuracy of the reported binding energies was ±0.1 eV.
e C 1s peak at 284.8 eV, as a contamination carbon, was used as a reference for all spectra.
e texture of the samples (BET surface area and porosity) was measured using an Autosorb 1, Quantachrome Autosorb Automated gas sorption system (Quantachrome Instruments).
e reducibility was determined based on temperature-programmed reduction using hydrogen as a component of the reducing mixture (TPR-H 2 ).Measurements were performed using chemisorptiometry with Pulse Chemisorb 2705.e TPR-H 2 measurements consisted of passing through a ferrite catalyst bed reducing mixture (5% H 2 in Ar, 30 cm 3 •min −1 ) with a controlled rise in temperature.e 50 mg of sample was placed in a U-shaped quartz reactor and heated from room temperature to 1000 °C with a slope of 5 °C•min −1 .A TCD detector was used for the quantitative determination of hydrogen consumption.Acidbase properties of the studied spinels were determined by the test reaction of cyclohexanol (CHOL) conversion.Experiments were carried out in a continuous-flow fixed bed reactor.e catalyst (0.5 g) was diluted with SiC (grain size from 0.5 to 0.6 mm) to the volume of 1 cm 3 and placed in the glass reactor with an internal diameter of 8.0 mm. e reactor was put in a furnace chamber heated up to 300 °C.e mixture of dry N 2 (99.999%) with cyclohexanol (1.3 mmol•dm −3 ) was passed through a reactor with a total gas flow of 20 dm 3 •h −1 .
e products of CHOL decomposition were analysed using a GC-FID detector.e total acidity and acid strength distribution were evaluated by the temperature-programmed desorption of ammonia (TPD-NH 3 ) method.
e sample (1 g) was placed in the reactor and heated to 750 °C under argon for 1 h in the argon stream to remove the adsorbed pollutants.After cooling to 180 °C, gaseous ammonia was passed through the sample for 30 minutes, and then, the sample was purged with argon at 180 °C for 1 h in order to remove physically adsorbed NH 3 .Finally, TPD-NH 3 measurements were started in argon with a heating rate of 10 °C•min −1 up to 550 °C.e amount of desorbing ammonia was determined using the TCD detector.
e catalytic test reaction of ferrite samples was examined as for the butan-1-ol conversion reaction.e catalytic experiments were carried out in a vertically heated flow quartz reactor under air pressure.Typically, 3 cm 3 (d � 0.6-1.2mm) of catalyst was placed in the reactor.Butan-1-ol was fed continuously from 200 °C using a Medipan 610-2 dosing pump with a liquid flow of 3 cm 3 h −1 without any carrier gas.Reaction products collected in the temperature range from 350 to 470 °C were analysed with INCO N-505 gas chromatograph.

Results and Discussion
e co-precipitation yield for each precipitating agent was determined by ICP-AES on the basis of the balance of the content of elements in the leach solution and the filtrate.e precipitation yield exceeded 95% for all metals, while it was almost 100% for the main ferrite components (Fe, Zn, and Mn) when sodium hydroxide and sodium carbonate were used as precipitating agents.In turn, when ammonium oxalate was applied as the precipitating agent, the values of the yield of zinc, manganese, and iron co-precipitation were slightly lower.In the case of Zn, this effect may be associated with the formation of a soluble ammine complex Zn(NH 3 ) 4

2+
. e powder X-ray analysis (Figure 1) confirmed the nanocrystalline nature of the materials prepared by the combustion method.All diffraction peaks are characteristic for a typical spinel ferrite (confirmed by JCPDS-ICDD file No. 01-074-2401).In the XRD pattern, there were no additional peaks indicating the presence of other phases, for example, bixbyite or hematite.e average crystallite size was 18.6, 33.6, and 27.5 nm for P1-C2O4, P2-CO3, and P3-OH sample, respectively.e particles' size observed in TEM pictures is consistent with the results of the XRD analyses (Figure 2).
On the basis of the recorded nitrogen adsorption and desorption isotherms (Figure 2), both porosity and the specific surface area of the prepared ferrites were characterised.e recorded isotherms are intermediate between the type I and IV isotherms according to the IUPAC classification [33].Such a type of isotherm may indicate the presence of both mesopores and micropores.At higher relative pressure values, a hysteresis loop appears, which is the narrowest for the P2-CO3 sample.
e shape of the hysteresis loop with no limitation of adsorption at high p s /p 0 values corresponds to H3-type hysteresis loop [33]. is hysteresis is characteristic for a material with plate-like particles connected in such a way that they form slitshaped pores.SEM and TEM microscope images confirm the presence of flakes in shape agglomerates (Figure 2).e specific surface area of the obtained powders has a similar value for the P1-C2O4 and P3-OH samples and amounts to 4.61 and 4.98 m 2 •g −1 , respectively, while the P2-CO3 sample has half of that value (1.96 m 2 •g −1 ).However, the grain size calculated on the basis of the specific surface area is 10 times larger than crystallite size.An explanation of this phenomenon (differences of these values of surface area) may be the particles agglomeration.It is known that when the powders are strongly agglomerated, the results from BET in fact do not reflect the size of particles.
e low value of specific surface area in relation to the expected value for nanocrystalline powders is related to the formation of agglomerates that limit the adsorption of nitrogen.is phenomenon was also described in the literature for α-Al 2 O 3 synthesised by combustion methods [34].SEM micrographs of the powders obtained from the hydroxide (P3-OH) and oxalate (P1-C2O4) precursors show a sponge-like structure with numerous pores.us, the P3-OH sample total pore volume is high and amounts to 4.13 10 −3 cm 3 •g −1 .e   e powder P2-CO3 is more compact, so the bed is less available.
is sample has the lowest value of the pore volume, which equals to 2.22 10 −3 cm 3 •g −1 .e pore size distribution of P1-C2O4 and P2-CO3 samples is wide and is also included in both the meso-and micropore ranges (inset Figures 2(a) and 2(b)), whereas the P3-OH sample has a sharp maximum at approx. 5 nm in the graph of pore size distribution, which means that there is a significant contribution of mesopores in this material (inset Figure 2(c)).
Depending on the metal ion precursor and synthesis conditions, the morphology of powders produced by the combustion method is varied (Figure 2).e grains of the ferrite powders are combined into agglomerates observed in the micrographs as flakes of different sizes.Typically, in the combustion synthesis, the adsorbed water contributes to the formation of hard agglomerates, although additional drying could slightly reduce this negative trend.Ferrites prepared from the hydroxide precursor (which requires a more concentrated nitric acid (V) to be dissolved) are more finely comminuted and have a "sponge" structure, with many open pores (Figure 2).Materials obtained from the oxalate and carbonate precursor do not outwardly differ.However, the P1-C2O4 sample is more porous, while the P2-CO3 sample is formed by numerous flakes of different sizes.Transmission electron microscopy allows the precise analysis of the microstructure.As shown in Figure 2, the aggregated grains are combined into larger agglomerates.Magnetic materials of nanometre grain size are related to the existence of a permanent magnetic moment proportional to the volume [35] and have a tendency to form agglomerates. e P1-C2O4 samples have much smaller irregularly shaped particles of irregular shape compared to those of the P2-CO 3 sample.
e material obtained from the hydroxide ion precursor is characterised by aggregated particles with the size in the 20-30 nm range.
In the next step, the state of the surface was determined because it participates in catalysis.e surface composition of the ferrite powders obtained by the combustion method was defined on the basis of the data obtained by XPS.Table 3 shows the surface elemental composition (at.%) of C, O, Mn, Zn, and Fe. e reported differences in chemical composition can significantly affect the catalytic properties of these powders.It can be seen that the surfaces of the analysed samples in the "as received" form contain significant amounts of carbon within the range of 20-30 at.%, which does not raise any objections.Moreover, it considered the components of carbon bonds separated from the C 1s spectra C-C (C-H), C-O, C�O, and O-C�O/CO 3 as being typical, with the assigned binding energies of 284.8, 286.3, 287.6, and 288.8 eV, respectively (Figure 3).e structure of the C 1s bonds was also analysed on the surface of the powders after removing most of the carbon contamination by Ar + beam, which resulted in a drop in the surface total carbon content to 12-13 at.% for all samples.Due to the complexity of the combustion process, it is impossible to precisely correlate the carbon bonds with oxygen and the applied precursors.However, it was noted that the proportion of C-O groups after Ar + sputtering was similar and amounted to approx.14% in relation to the total carbon amount, while the proportion of CO 3 /O-C�O groups in the P2-CO3 sample was 2 and 2.5 times higher than that in the P3-OH and P1-C2O4 samples, respectively; this probably resulted from the incomplete oxidation of carbonate precursors.e characteristic peak of a carboxyl group (Figure 3) is the most intensive for the P1-C2O4 sample, and it could be connected with the remaining citric acid (the fuel of the combustion process).On the analysed "as received" surface, there is also a quite intensive peak with a maximum at 287.6 eV, which is characteristic for the C�O bond.It can be related to both the remaining citric acid and oxalate ions that, as a metal oxalate precipitate, were the precursor ions to the combustion synthesis.After slight etching with Ar + ions, this component was no longer observed.For the P3-OH samples, obtained from the hydroxide precursor, the intensities of the peaks characteristic for C�O and COO/CO 3 groups were lower than those for the P1-C2O4 sample.
In the O 1s spectrum (Figure 3), an intensive peak, which is characteristic for lattice oxygen, is observed at the energy of ∼530 eV for all of the samples.Another peak at the energy of approx.531 eV for the P1-C2O4 and P2-CO3 samples is characteristic for the carbon and oxygen bond in carbonates, whereas the peak at 533 eV can be related to surface water, as well as the products of combustion an organic carbon compound (citrates and oxalates).e O 1s spectrum of the P3-OH sample differs from the previously described ones only in the fact that the peak at ∼531 eV is slightly shifted towards higher bond energies.
is suggests that oxygen joins not only the carbonate carbon, but it also indicates a higher proportion of oxygen originating from OH − groups.
Zn 2p spectra for the P1-C2O4 and P2-CO3 samples are similar (Figure 4).After deconvolution of the Zn 2p 3/2 spectra, the main peak at ca. 1021.3 eV was assigned to ZnO as lattice oxide.Another less intensive peak at the energy of ∼1023 eV was interpreted as the Zn bond with the oxygen in the carbonate group.
is component is present on the surface of all three samples of the Mn-Zn ferrites.Its proportion was estimated to be 7.0, 21, and 25% in the analysed "as received" P1-C2O4, P2-CO3, and P3-OH samples, respectively.After etching the surface with Ar + beam, the proportions of CO 3 2− groups were similar and ranged between 10 and 15% of the total amount of Zn.On the P3-OH sample surface, after the deconvolution of the Zn 2p envelope, the third component with a max.at approx.1022 eV was distinguished and was assigned to Zn(OH) 2 .Its proportion was estimated to be 18% of the total content of surface Zn in the ferrite (P3-OH; (A) in Figure 4).e presence of surface hydroxyl groups bound in Zn(OH) 2 may be important for the catalytic activity of the ferrite, since they catalyse many reactions [36].e formation of the surface Zn in the form of hydroxide as a result of the degradation of ZnMn 2 O 4 ferrite was also described previously [37].
Mn 2p 3/2 spectra for all three samples are asymmetrical at the side of higher energy values.Initially, the possible presence of Mn in the +2, +3, or +4 oxidation state was assumed, but the lack of a satellite peak characteristic for Mn 2+ , located approx.5 eV above Mn 2p 3/2 , means that Journal of Chemistry Mn 2+ did not appear on the surface of the studied samples.Moreover, deconvolution of the Mn 2p 3/2 spectrum for the P1-C2O4 sample made on the basis of the Handbooks of Monochromatic XPS Spectra [37] (i.e., preserving the same relative positions of the peaks and the FWHM values, (A) in Figure 5) showed a very good agreement with the observed spectrum of pure MnO 2 .e increase in the FWHM value from 3.0 eV for P1-C2O4 to 3.8 and 3.9 eV for P2-CO3 and P3-OH, respectively, may indicate the presence of a greater amount of Mn 2 O 3. e multiplet of Mn 2 O 3 (4 peaks) [38] includes a dominant peak with a maximum at approx.642 eV. is component widens the top of the peak for the observed P2-CO3 and P3-OH spectra ((B) and (C) in Figure 5).Unfortunately, overlapping of the MnO 2 and Mn 2 O 3 envelopes excludes a quantitative distinction between these phases.
Analysis of the Fe 3p spectra of the studied materials confirms the presence of Fe 3+ on the surface (Figure 6).Both  6 Journal of Chemistry Mn 3+ and Fe 3+ are present in the form of connections with the spinel lattice oxygen occupying octahedral sites.Jacobs et al. [39] pointed out that only octahedral sites in the spinel lattice are strongly surface-exposed and, therefore, participate in the catalytic reaction.Table 3 presented the results of the quantitative XPS analysis.It should be noted that there are clear exceptions of each sample surface composition to the stoichiometry of Mn 0.6 Zn 0.4 Fe 2 O 4 ferrite, whose presence was confirmed by XRD analysis.is apparent contradiction was interpreted as the presence of the ferrite crystal phase constituting the core of grains coated with Mn, Zn, and Fe compounds.Such compounds located on the surface may be non-stoichiometric and/or provide a much thinner layer than the Mn 0.6 Zn 0.4 Fe 2 O 4 crystallite size.ese conclusions are based on the specificity of the XPS technique, which was very surface-sensitive, recording only the photoelectrons emitted from the layers of a thickness of at most a few nm.Moreover, the compositions of objects with sizes >10 nm will always be underestimated in relation to those in the nanometre size range.ese relationships can be clearly seen in the obtained results.us, the calculated ratios of Mn/Fe and Zn/Fe for each sample are clearly higher than the nominal; in the ferrite, they are equal to 0.3 and 0.2, respectively.is means that Fe ions are arranged inside the analysed grains and are probably Fe(III) (present in the crystal lattice of the ferrite).Enrichment of the surface of the ferrite powders with Mn(IV)/Mn(III) and Zn(II) compounds resulting from Mn and Zn segregation also implies the disordering of ferrite volume.is involves the formation of a cation defect spinel structure with oxygen vacancies balancing the overall positive and negative charge in the crystal lattice.
e iron content in the analysed sample surfaces varied and was the lowest in the P2-CO3 sample (4.8 at.%) and the highest in the P3-OH one (7.3at.%). ese differences may also indicate variations in the size of the ferrite grains, which are visible in TEM images, and the diversity of the catalytically active surface.
To understand how the differences in the surface revealed in the XPS analysis affect the surface properties, i.e., the ability to reduce and the amount and quality of the acid-base centres, the prepared powders were subjected to a typical catalytic test.Figure 7 shows the profiles of susceptibility to a reduction of the obtained ferrites.e studied samples are reduced at elevated temperatures, but all of the recorded peaks are multimodal.Such a course of the TPR-H2 curves indicates that more than one reduction reaction takes place at a similar temperature, which can be an effect of the complex nature of the Mn-Zn ferrite.Both manganese and iron may occur in different oxidation states.During the reduction, a ferrite poor in oxygen (Mn x Zn 1−x Fe 2 O 4−δ ) and then mixtures of MnO, ZnO, and Fe 3 O 4 or FeO oxides (the product of the reaction depends on the reduction mechanism) are formed in the solid phase.At high temperatures, the reduction of ZnO to metallic zinc and its sublimation, as well as a reduction of Fe 3 O 4 to FeO or metallic Fe, can take place.e high value of the Mn 2+ /Mn reduction potential precludes the formation of metallic Mn under the TPR-H2 measurement conditions [40].e P1-C2O4 sample starts reducing at a temperature of about 100 °C lower than the P2-CO3 and P3-OH samples.e described effect may be related to a reduction of organic groups (residue of citric acid) remaining after the combustion of the gel.e presence of such surface organic groups was confirmed by the XPS analysis (characteristic peak of carboxyl groups seen in Figure 3 is the most intensive for the P1-C2O4 sample).For the P2-CO3 sample, which, as indicated by analysis of the XPS spectra, is characterised by a high proportion of  CO 3 /O-C�O groups, there is an additional small peak at approx.350 °C on the TPR-H2 curve (I at Figure 7).A slight inflection occurring above 410 °C may be associated with a reduction in the surface Mn(IV) and Zn(II) compounds (II at Figure 7).Quite a broad and intense peak with its maximum at approx.600 °C is observed for all samples (III at Figure 7). is peak is asymmetrical towards lower temperatures.At higher temperatures (IV at Figure 7), as a result of the reduction reaction, an oxygen-deficient ferrite may form and the ZnO to "volumetric" Zn and Fe 2 O 3 to Fe 3 O 4 reduction reactions may subsequently take place [41].e peak occurring at above 900 °C (V at Figure 7) can be attributed to the reactions: FeO ⟶ Fe or Fe 3 O 4 ⟶ Fe, if the reduction of magnetite occurs according to the mechanisms given by Wimmers et al., in which an intermediate product (FeO) is not formed [42].
e characteristics of the acid-base properties of the ferrite catalyst surface were determined in the test reaction of cyclohexanol conversion.Defining the nature of the acidbase active centres on the surface of the catalyst is based on the determination of the selectivity of cyclohexanol conversion to its dehydration (cyclohexene CHEN-S CHEN ) or dehydrogenation (cyclohexanone CHON-S CHON ) product.
e dehydration of a secondary alcohol, which leads to the formation of an unsaturated C�C bond, occurs at acid centres, whereas the simultaneous interaction of acid and base centres is needed for the dehydrogenation of cyclohexanol [43].e surface of the P2-CO3 sample obtained from a carbonate precursor is clearly distinct from those of the P1-C2O4 and P3-OH samples in terms of the acid-base nature.e selectivity of the dehydration reaction reaches nearly 100% for this sample, which indicates the sole participation of acid centres in the cyclohexanol decomposition reaction.is sample is characterised by the highest value of total acidity, which is shown in Figure 8. e P1-C2O4 and P3-OH samples contain a similar proportion of acid and base sites, but acid centres are predominant in the materials (S CHEN � 62.7 and 67.0% for P1-C2O4 and P3-OH, respectively).It was also noted that the quantitative Mn/Fe and Zn/Fe ratio on the catalyst surface (determined by XPS analysis) is comparable for these samples (Table 3).As shown by previous studies [23,31], such a nature of the surface, where both acid and base centres are present, improves the ability of the material to selective ketone (heptanone-4) formation from butan-1-ol.e proportion of the centres versus their strength is shown in Figure 8. Weak acid sites dominate in the P2-CO3 sample, and their proportion is almost twice as high as in the P1-C2O4 and P3-OH samples.In contrast, the proportions of the medium acid centres are comparable for all the samples.e studies of catalytic activity for ferrite prepared by combustion methods from a variety of ion precursors confirm both the quantitative (surface area and pore volume) and qualitative (differences in the elemental composition of the catalyst surface) diversity of the samples.For the P1-C2O4 and P3-OH samples, selectivity towards ketone (heptanone-4) increases with temperature, as can be seen in Figure 9.However, the P3-OH sample has the highest value of selectivity of the dehydrogenation reaction, with consecutive condensation to heptanone-4.
e apparent selectivity towards aldehyde formation decreases.Butyraldehyde, as an intermediate product of the reaction, did not undergo consecutive transformations to ketone and/or ester.
e P2-CO3 sample, at the same catalyst loading (1 h −1 ), exhibits the selectivity of dehydrogenation only to aldehyde, and the selectivity to by-product formation is higher than that of the P1-C2O4 and P3-OH catalysts.
e surface studies showed that the P2-CO3 catalyst surface differs from that of the other samples.e reason for the low catalytic activity of the P2-CO3 ferrite (butan-1-ol conversion is only 40%), compared to the P1-C2O4 and P3-OH samples, may be the high proportion of low-strength acid centres and the presence of only, or the vast majority of, acid centres on the surface (S CHEN ≈ 100%).e interaction of strong Lewis acid sites and those of a base nature is required to form ketone as a result of bimolecular condensation [44].
Only an increase in the contact time allows the occurrence of consecutive reactions and the formation of butyl butyrate and heptanone-4 on the P2-CO3 catalyst (Figure 10).e significant increase of contact time (a butan-1-ol flow rate of 1 cm 3 •h −1 , LSHV � 0.33 h −1 ) caused creation of similar products, as in the case of the P1-C2O4 and P3-OH samples.For the P2-CO3 sample, the increase in contact time caused an almost double increase in conversion degree and allowed the conversion of butan-1-ol to butyl butyrate.is could indicate that for a less catalytically active ferrite, the occurrence of bimolecular condensation reactions is only possible in the presence of a significant proportion of newly appearing aldehyde particles.In contrast, for the P1-C2O4 and P3-OH powders, the ester formation yield decreased.ere are two probable reasons for this phenomenon: (1) If the ester is an intermediate product in primary alcohol transformation [45,46], which occurs in the order: n-alcohol ⟶ aldehyde ⟶ ester ⟶ ketone, then too short a contact time does not allow the full conversion of ester to ketone (2) If the ester is a product which is formed parallel with ketone, it may lead to the secondary thermal decomposition of butyl butyrate to aldehyde according to the scheme: n-alcohol ⟶ aldehyde ⟶ ketone and n-alcohol ⟶ aldehyde ⟶ ester ⟶ aldehyde ⟶ ketone.
In both cases, the proportion of ketone appearing as the result of consecutive reactions increases.
For materials with a structure as complex as that of ferrites, there are many factors affecting their catalytic activity: both the presence of cations in the crystal lattice, which can change the oxidation state, the existence of defects in the anion network, and the formation of a coreshell structure.e metal oxidation degree of cations (Fe and Mn) present in the studied ferrite materials can vary according to the reaction: 2Fe(Mn) 3+ + O lattice 2− ⟷ 2Fe(Mn) 2+ + ½ O 2 [47].
e Fe 3+ and Mn 4+ ions, as coordinatively unsaturated metal centres, may be Lewis acid sites, and the oxygen anions and hydroxyl groups, as proton acceptors, are Brønsted base sites [42,48].e  [49] underlined that the combination of two or more transition metal oxides displays higher catalytic activity than single metal oxides because of the synergistic effects.Also, the high lattice-defect density and surface coverage of OH groups or oxygen ions endow spinel oxides strong catalytic activity [49].Qiu et al. denoted that the kind of metal constituting the bivalent cation states of the spinels' structure has an influence on the activity and selectivity of the oxidative dehydrogenation reaction [50].e formation of the core-shell-type structure, in which the ferrite is the cation-deficient core and the oxygen vacancies compensate for the total positive and negative charge in the crystal lattice, forming active sites of oxygen adsorption at elevated temperatures, should therefore enhance the catalyst activity in the dehydrogenation reaction of the primary alcohol.Based on the above data and the experimental literature, it can be concluded that the increased catalytic activity of the P3-OH sample can also be related to the physical state of the catalyst: the material was less dusty, and therefore, the accessibility of the bed during catalysis was better.e sum of the yields of all dehydrogenation products (aldehyde + ester + ketone) was highest for the P3-OH sample (up to 470 °C).e highest catalytic activity of this ferrite may also be related to the surface condition.e iron content in the analysed samples (from XPS) varied: it was lowest in the P2-CO3 sample (4.8 at.%) and highest in the P3-OH ferrite (7.3 at.%).It can be concluded that the highest total Mn and Fe content (acid Lewis centre) and presence originating from OH − groups contributed to enhance the activity of this catalyst compared to P1-C2O4 and P2-CO3 samples.

Conclusions
Catalytically active Mn-Zn ferrites were obtained from the solution after acid leaching of used battery waste in a twostep process: co-precipitation and combustion.Only the spinel phase, characteristic for ferrite, is present in the powders prepared after combustion.TEM observations and the XPS analysis of ferrite surfaces revealed the formation of a core-shell-type structure.
e crystalline spinel phase forms the core, and the compounds Mn, Fe, and Zn are located on the surface.e type of precursor used changed the ferrites obtained in terms of particle size, elemental composition, development, and characteristics of the surface.e ferrites obtained from the hydroxide and oxalate precursors were similar in terms of the surface area (S BET ) and the number of acid-base sites (predominance of acid sites).However, materials obtained from the hydroxide precursor proved to be far more efficient catalysts.It is less dusty which makes better accessibility to the catalyst bed and faster catalysis reaction.Moreover, the observed presence of both surface hydroxide groups (not observed for the other studied samples) and the largest proportion of iron for this ferrite could contribute to the increased catalytic activity through the formation of oxygen vacancies, which became the location of oxygen anion adsorption.erefore, for the ferrite obtained from the hydroxide precursor, the reaction of both aldehyde formation at lower temperatures and ketone formation at elevated temperatures was observed, with the butan-1-ol conversion being almost 100%.

Figure 2 :
Figure 2: Adsorption and desorption isotherm of nitrogen for the samples: P1-C2O4 (a), P2-CO3 (b), and P3-OH (c).Inset: pore size distribution.SEM and TEM micrographs of the ferrite powders are given below the adsorption and desorption isotherms.

Figure 8 :Figure 9
Figure 8: e proportion of the sites of acid character and dependence of butan-1-ol conversion on temperature for the ferrite powders obtained by combustion.

Table 1 :
e content of elements in the solution after acid leaching of battery waste.

Table 3 :
XPS surface composition of Mn-Zn ferrites.Nominal value for Zn 0.4 Mn 0.6 Fe 2 O 4 .* * Calculated on the basis of Zn 3p spectral line. ).