The iodine index indicates the degree of activation of activated carbon and is essentially used to measure the micropores (0–20 Å) present on the carbon. To gain an insight into the porosity structure of activated carbons, liquid-phase iodine adsorption has been used extensively in the characterization of activated carbon (Ndi et al., 2014; Sahira and Bhadra, 2014; Tchuifon et al., 2014; Ndifor-Angwafor et al., 2017; Ngakou et al., 2017; Kuete et al., 2018). The iodine index values of the prepared activated carbons are presented in Table 2.

Table 2 shows that, in the case of KOH, the iodine value increases with both concentration and precursor-to-activator solution ratios. For the H₃PO₄ activator, a maximum is reached at 1 M.

These results attest that the type of activating agents used in the preparation of activated carbons, the concentration, and synthesis conditions variously influence the porosity structure of the resulting carbonaceous material. These iodine index characterization tests show higher microporosity in activated carbons prepared from KOH with iodine indexes ranging from 489 to 591 mg/g. This microporosity is very low for carbons prepared from orthophosphoric acid (H₃PO₄) with iodine values between 328 and 395 mg/g. H₃PO₄ retains carbon (promoting the formation of cross-links through dehydration, cyclization, and condensation reactions) and redistributes bio-polymers, which facilitates their conversion to aromatic rings and forms larger pores, consequently reducing iodine adsorption efficiency (Anis et al., 2014). This promotes the formation of a mixture of micropores and large mesopores.

The high iodine values obtained during KOH activation indicate that the KOH oxygen eliminated cross-linking and stabilized the carbon atoms in the crystallites. During pyrolysis, the intercalation of K atoms, resulting from the reduction of potassium hydroxide, is the initial step in the degradation process of biomass activated with KOH. This intercalation of potassium metal forces the lamellae apart in the crystal. This spacing in the crystal promotes the development of the carbonaceous structure and the porous network (Sahira and Bhadra, 2014). In other words, the development of this porosity in KOH-activated carbons is associated with a gasification reaction as KOH is reduced to potassium metal during the carbonization process. The reaction of KOH and carbon is postulated to occur according to Eq. 8 (Li et al., 2004): 6 KOH + 2 C   →   2 K + 3 H 2 + 2 K 2 CO 3 . (8)

In general, higher iodine values denote greater adsorption capacity. Thus, the iodine values (Table 2) show that KOH is a better activating agent than H₃PO₄ for the preparation of carbons from peanut shells. However, the iodine values reported vary from the values reported in the literature for activated carbons activated with potassium hydroxide (Okieimen et al., 2007; Sahira and Bhadra, 2014; Kuete et al., 2018). This is due to differences in the composition of the precursor material and synthesis conditions.

In the remainder of this work, the activated carbon with the greatest iodine index value was used and coded as AC-PS, whereas the iron oxide-modified activated carbon was coded as AC-PS@FeII. The iodine index of PS@FeII was determined, and relative performance with pristine carbon is presented in Table 3.

Based on the iodine index values given in Table 3, the pristine activated carbon (AC-PS) is more porous than the functionalized activated carbon (AC-PS@FeII) consistent with the pore volumes reported in nitrogen adsorption–desorption studies. This is due to the occupation of the pores of the activated carbon by iron oxides during treatment.

FTIR spectroscopy was used to identify different surface functional groups, the pristine carbon, KOH-activated carbons, and ferromagnetic carbons (AC, AC-PS, and AC-PS@FeII), over a wavelength range of 500–4,000 cm⁻¹. The infrared spectra of the materials are shown in Figure 1.

The FTIR spectra of the three materials, namely, AC, AC-PS, and AC-PS@FeII, show that in the region between 2,500 and 2,000 cm⁻¹, characteristic peaks of C–C aliphatic stretching vibrations appear. The peaks appearing at 1,600 cm⁻¹ on these spectra are attributed to the stretching of the C=O carbonyl groups of the ketones, aldehydes, and/or lactones. The bands around 1,583 cm⁻¹ attributed to the elongation vibrations of C=O groups remained unchanged and have almost the same intensities on the dried peanut shells as on the activated, calcined, and magnetized peel. The slight decrease in the intensity of the peaks was observed following pyrolysis. The peaks centered at 1,100 cm⁻¹ for ferromagnetic activated carbon AC-PS@FeII and 1,050 cm⁻¹ are attributed to the stretching vibrations of ether C–O (Lall et al., 2020; Djouonkep et al., 2022). The appearance of two new bands on the AC-PS@FeII spectra around 750 and 1,000 cm⁻¹ is the characteristic of the OH in- and out-of-the-plane bending vibrations of the Fe–OH group in goethite, respectively (Khelifi et al., 2016). Additional differences between the three FTIR spectra were noted. For the spectrum of AC-PS@FeII, a band is observed at 560 cm⁻¹ corresponding to iron oxides (FeO). The band between 450 and 740 cm⁻¹ was attributed to the Fe–O vibrations of iron oxide nano AC-PS particles (Ngankam et al., 2020). This denotes the successful precipitation of iron oxide after the functionalization of AC-PS to obtain ferromagnetic activated carbon, AC-PS@FeII.

Energy dispersive X-ray spectroscopy (Figure 2) was used to inspect the elemental composition of the materials. The pie charts presented in each of these figures provide the percentages of the main chemical elements present in the structure of each of these three materials. The chemical composition includes the raw materials, i.e., carbon (72.2%), oxygen (24.9%), and potassium (2.9%), and activated carbon, i.e., carbon (82.3%) and oxygen (17.7%), while ferromagnetic activated carbon contains carbon (42%), oxygen (24%), sodium (8%), and iron (26%) (Figure 2C). EDX analysis revealed the presence of the elements Fe, O, and C on the surface of the ferromagnetic activated carbon, consistent with the formation of iron oxides observed from the FTIR spectrum. The disappearance of elements (K) could be due to dissolution into the aqueous phase and the appearance of the new elements (Fe and Na) arising from the elemental composition of the synthesis materials.

Figure 3A shows the SEM image of the precursor peanut shells. The biomass has a surface morphology in the form of heterogeneous cavities with irregularly agglomerated dispersed clumps of particles. The SEM image of the activated carbon (Figure 3B) shows a sponge-like structure with enhanced porosity as a result of the chemical activation process. The SEM image of the ferromagnetic activated carbon (Figure 3C) shows a surface covered with dispersed particles of iron oxides that clog the pores (Aurelien et al., 2022; Peng et al., 2022).

XRD was used to determine the crystallinity of the raw material compared to the activated carbon and mineral phases of iron in the ferromagnetic activated carbon. The diffractograms are given in Figure 4.

For the raw material, the peak observed at 2θ = 23° is attributed to different planar crystal structures of carbon (Dong et al., 2020). In terms of crystallinity, much of the structure of the activated carbon is amorphous based on the diffractogram pattern. The peaks around 2θ = 30° and 2θ = 44° were attributed to the planar crystal structure of the activated carbon (Sandrale et al., 2023). The diffractogram of AC-PS@FeII portrayed at least five well-pronounced peaks, indicating the existence of different crystalline phases. The peaks at 2θ values of 30°, 35°, 44°, 55°, and 65° of AC-PS@FeII are the characteristic of magnetite (Fe₃O₄), which indicates a cubic spinel structure of indexed peaks like (220), (311), (400), (422), and (400) (Dong et al., 2020). The presence of crystalline phases of iron confirms the insertion of iron ions into the carbon. These crystalline phases essentially consist of magnetite (Fe₃O₄) and maghemite (γ-Fe₂O₃) at 2θ = 35°. Only magnetite (Fe₃O₄) has a ferromagnetic character, and the other phase is paramagnetic.

The results of the nitrogen adsorption hysteresis at 77 K by different carbons are given in Figure 5A, and the pore-size distribution of the carbons is given in Figure 5B.

Figure 5A shows the N₂ adsorption–desorption isotherms and pore-size distribution obtained for the AC-PS and AC-PS@FeII samples. Although the isotherms present a combined type I/II characteristic (IUPAC, 1985), the type I isotherm characterized for microporous solids occupies a major part. The pore parameters given in Table 4 indicate that the formation of iron oxide nanoparticles inside the porous structure of the activated carbon causes a reduction in the BET surface area and pore volume (Zhang et al., 2007). The BET surface area and pore volume values are given in Table 2. The analysis of the pore size by the BJH method (Figure 5B) highlights the microporous nature of the activated carbon and AC-PS@FeII. Similar results were obtained by Kpinsoton (2019) on the development of catalysts based on activated carbon and laterites for the degradation of methylene blue by the heterogeneous Fenton process.

The catalyst thus prepared was used to degrade methyl orange dye by optimizing the percent degradation using the response surface methodology.

AC-PS@FeII was used in the optimization studies based on the central composite experimental design of the response surface methodology since it exhibited the highest catalytic activity. Table 5 displays the design matrix with 27 runs conducted in a single block, together with the experimental and predicted responses. For each of the 27 experimental runs, the residue values are much smaller than the experimental and theoretical values, indicating that the experimental and theoretical values are reasonably approximated. This is supported by the near-unity R ² (96.60%) and R² _adj (92.64%) values.

The quadratic mathematical model suitably represented the correlation between predicted values (Eq. 9) and the experimental values (Eq. 10).

ANOVA makes it possible to evaluate the influence, if any, of different factors and their interactions and verify the validity of the mathematical model. The ANOVA data of the model are given in Table 7:

For this analysis, the probability p was used to evaluate the significant effects of the factors and their interaction, and the indicator of the fineness of the second-order polynomial model obtained is the value of the coefficient of determination, R ². A factor or an interaction between factors is said to be significant when the probability p is less than 0.005 (p < 5%). In addition, the value of the coefficient of determination must be close to 1 (Kimouche, 2008). The results are given in Table 7. The degradation efficiency was significantly (p < 0.05) affected by the solution pH and the initial concentration of the dye. The solution pH alters the surface chemistry of the catalyst and, hence, the activity of the catalytically active sites. The initial concentration controls the mass gradient of the dye across the boundary layer to the catalyst surface and the radicals and accessibility to the limited catalyst. Beyond 50 mg/L, the dye concentration exceeds the available radicals, leading to decreased degradation efficiency. Uncharacteristically, the catalyst dosage had no statistically significant effect on the degradation efficiency. This suggests that the increase in the adsorbent dosage resulted in the degradation of the finite amount of H₂O₂ to O₂ and the concomitant quenching of OH* radicals by Fe²⁺ ions in the solution (Dat et al., 2023).

Figure 8 shows in 3D the effect of two variables on the degradation efficiency when the other variables are fixed at zero. A relative significant interaction was observed between the initial dye concentration and pH. A decrease in dye concentration and an increase in pH favored the degradation efficiency. In the literature, the removal efficiency of dyes using the heterogeneous Fenton process is dependent on pH. A pH range of 2–3 is often optimal for Fenton degradation because complexes and insoluble ferric hydroxide form in the solution at pH values above 3, and the quantity of OH^• radicals needed for the oxidation of the dye is significantly reduced, which explains the notable decrease in the percentage degradation obtained with higher pH values (Hassan and Hameed, 2011; Muhammad and Mashi, 2020; Ngankam et al., 2020). Fewer investigations, however, demonstrate that high yields of degradation at pH levels (in an acidic medium with pH 2 and 4) result from the impregnation of magnetic particles in the form of iron oxides (Figure 8). The optimum process conditions were identified from the 3D model plots. The optimal degradation conditions were as follows: 0.2 g/50 mL of AC-PS@FeII, the concentration of the pollutant at 50 mg/L, and pH = 2 at a contact time of 50 min for a degradation rate of 99.22%.

Determining the kinetics of organic pollutant degradation and H₂O₂ decomposition is a challenging task since numerous mechanisms can alter the apparent rate of the Fenton reaction (Bopda et al., 2022). To simplify, the Fenton process degradation reaction of an organic chemical can be characterized by a pseudo first-order kinetic rule (Eq. 10a) that assumes a quasi-stationary concentration of hydroxyl radicals created in the medium. d C d t = − k C → ⁡ ln C 0 C t = k t . (10a)

The second-order kinetic model (Eq. 11) was also used to describe the degradation of an organic dye by the Fenton process (Kenda et al., 2023): d C d t = − k C 2 → 1 C t − 1 C 0 = k t . (11)

The linear regression of the experimental points can be used to determine the kinetic rate constant by tracking the pollutant degradation over time. Only at the beginning of the reaction, before the substantial intermediates have formed, can the kinetic parameters be determined (Mafo et al., 2023):

The pseudo first- and pseudo second-order kinetics were studied, and the results are presented in Figures 9, 10, respectively. Table 8 presents the first- and second-order kinetic parameters. Figure 9 presents the pseudo first-order kinetic model applied to the degradation of methyl orange on ferromagnetic activated carbon.

As shown in Table 8, the pseudo first-order model best described the degradation kinetics of the methyl orange dye since the coefficient of determination was closest to unity (R² = 0.936).