Synthesis of Alpha Ferrous Oxalate Dihydrate from Ferrotitaniferous Mineral Sands via Hot Pressurized Aqueous Oxalic Acid: Kinetics and Characterization

Abstract

Ferrous oxalate dihydrate is a versatile organic mineral with applications across fields. However, little is known about the feasibility of its synthesis directly from iron-bearing minerals using binary subcritical water (sCW) systems and its associated kinetics. In this study, the sCW+oxalic acid system at either 115 °C or 135 °C was investigated as a reaction medium for ferrous oxalate dihydrate (α-FeC₂O₄∙2H₂O) synthesis, starting from ferrotitaniferous sands. The kinetics of the synthesis reaction were studied, and the physicochemical characterization of the as-synthetized ferrous oxalates was performed. Overall, the sCW synthesis was temperature-dependent, following second-order reaction kinetics according to the proposed precipitation pathway. A high reaction rate constant, significantly high yields (up to 89%), and reduced reaction times (2–8 h) were evident at 135 °C. The as-synthetized product corresponded to the monoclinic α-FeC₂O₄∙2H₂O, showed relatively high specific surface areas (from 31.9 to 33.7 m²∙g⁻¹), and exhibited band gap energies within the visible light range (~2.77 eV). These results suggest that α-FeC₂O₄∙2H₂O can be synthesized using an organic dicarboxylic acid and iron-rich, widely available, low-cost mineral precursors. In addition, the as-prepared α-FeC₂O₄∙2H₂O could be further optimized and tested for catalytic and visible light photocatalytic applications.

1. Introduction

Ferrous oxalate dihydrate (FeC₂O₄∙2H₂O), also referred to as Humboldtine, is a relatively rare yellowish ionic organic mineral, naturally found in geological environments [1,2]. In addition, synthetic analogs of natural ferrous oxalate have been obtained starting from reagent-grade ferrous salts and oxalates [1,3] as well as from iron-bearing mineral rocks and oxalic acid-producing fungi by biomineralization [4]. In addition, several patents have been reported aiming at the conversion of industrial upstream waste into ferrous oxalate [5,6,7,8]. Among them, patent 101717331A deals with a multi-stage process for the conversion of ferrous sulfate, a by-product from TiO₂ production via the sulfate method, into ferrous oxalate.

In general, ferrous oxalate is commercialized as a photographic developer agent, pigment for optical glass, plastics, and paints, and as a precursor of lithium iron phosphate (LiFePO₄) cathode material for lithium-ion batteries [9,10,11,12,13,14]. Interestingly, the iron oxalate market is expected to grow further if used in anti-anemic drug preparations [10].

Beyond current commercial applications, ferrous oxalate dihydrate has been studied as a reagent for the synthesis of valuable materials such as siderite [15], maghemite and hematite nanorods adsorbents [16], and iron carbide catalysts [17]. In addition, the catalytic and photocatalytic activities of ferrous oxalate dihydrate have been explored for the decontamination of wastewater from various organic pollutants such as methylene blue, caffeine, and rhodamine B [18,19,20,21], with promising results because of its combined photocatalytic and photo-initiated Fenton activities [20,21]. Moreover, ferrous oxalate has been incorporated into different complex photocatalytic systems, including aramid nanofibers, polyphenylene sulfide microfibers, graphene, and iron oxides for the efficient photodegradation of methylene blue, bisphenol A, and the antibiotic norfloxacin, respectively [22,23,24,25,26].

Although applications of ferrous oxalate dihydrate have been investigated across various fields [27,28], studies focusing on the kinetics of its chemical synthesis are scarce [29]. Similarly, little is known about the feasibility of using emerging technologies such as subcritical water (sCW) extraction for the production of ferrous oxalate dihydrate. In our previous study, the synthesis of ferrous oxalate dihydrate together with titanium dioxide polymorphs from ferrotitaniferous mineral sands (low-cost iron-bearing mineral ore) using aqueous oxalic acid (0.1, 0.5 to 1.0 M) at sCW conditions (155 °C/12 h/50 bar) was confirmed and reported as a proof-of-concept [30].

Based on the current knowledge, the main objective of this study was to investigate further whether or not α-FeC₂O₄∙2H₂O can be synthesized from ferrotitaniferous mineral sands within reduced reaction times, using aqueous oxalic acid of higher molarity but at mild temperatures that fall within the low-temperature range (100 °C < T < 155 °C) of the sCW region (100 °C < T < 374 °C), which have not yet been investigated for α-FeC₂O₄∙2H₂O synthesis from such iron-bearing mineral. In addition, another goal was to gain more insight into the synthesis kinetics, aiming for large-scale production, as well as the physicochemical properties of the obtained product.

Therefore, the sCW synthesis was carried out with pressurized 1.5 M aqueous oxalic acid at two different temperatures 20 °C apart (115 °C and 135 °C, respectively) and reaction times of 2, 4, 8, and 12 h. The corresponding ferrous oxalate percent yield and reaction rates were evaluated. As a result, a possible mechanistic pathway for ferrous oxalate synthesis with aqueous oxalic acid under sCW conditions has been proposed, and the kinetic parameters determined. In addition, the physicochemical characterization of the as-synthetized ferrous oxalates, including the phase composition, particle morphology, specific surface area, and band gap energy, was performed.

2. Materials and Methods

2.1. Materials

Ferrotitaniferous sands (0.6 FeTiO₃ ∙ 0.4 Fe₂O₃), with particle size smaller than 38 µm [31], and oxalic acid dihydrate (>99%) from DQI S.A (Medellín, Colombia) were the precursors used for ferrous oxalate synthesis. Distilled water from a Milli-Q system (18.2 MΩ·cm, Millipore, Billerica, MA, USA) was used as the solvent, and N₂ gas (99.9% purity) from Linde (Quito, Ecuador) as the purge and pressurized gas.

2.2. Synthesis of Ferrous Oxalate by Pressurized Aqueous Oxalic Acid

Ferrous oxalate synthesis was carried out in a Berghof high-pressure batch stirred reactor system with a movable head design (BR-500, Berghof Products Instruments, Eningen unter Achalm, Germany) equipped with a 500 mL PTFE-lined vessel, an electrical heating jacket, a thermocouple probe, and a pressure transducer. The temperature was measured inside the reactor (fluid temperature) and controlled using a temperature controller (BTC-3000, Berghof Products Instruments, Eningen unter Achalm, Germany).

A 1.5 M oxalic acid solution was used as the reaction medium, and the synthesis variables were temperature (115 and 135 °C) and reaction time (2, 4, 8, and 12 h). Briefly, for the sCW synthesis (Figure 1a), the reactor vessel was loaded with ferrotitaniferous sand (1.75 g) and 1.5 M oxalic acid solution (300 g), and the resulting mixture was purged with N₂ for 12 min. Next, the stirred reactor system (700 rpm) was semi-pressurized with N₂ up to 17 bar to keep the reaction medium in the liquid phase. Then, it was heated to either 115 °C or 135 °C (Figure 1). The recorded temperature and pressure profiles of the synthesis reaction are shown in Figure 1c. According to Figure 1c, in both cases, the synthesis reaction was isothermal, while pressure was raised as a function of reaction time. Upon completing the set reaction time, the heating system was turned off, and the reactor was allowed to cool naturally. The reaction product, a solid–liquid mixture, was poured into a beaker to settle for easy separation of the solid precipitate from the supernatant.

Finally, the yellowish precipitate (ferrous oxalate dihydrate; vide infra, Section 3) was recovered and washed until neutral pH prior to air drying at 80 °C for 6 h (Figure 1a). The ferrous oxalate dihydrate precipitate in powder form was stored in a sealed container for further characterization.

The supernatant was drained and collected in a separate glass container for quantification of its iron and titanium contents. In addition, the titanium-rich supernatant was further processed at 155 °C/12 h (Figure 1b) for TiO₂ precipitation.

The synthesis experiments were carried out in duplicates for each condition, and the percent yield of ferrous oxalate precipitate was quantified according to Equation (1).

$$\mathrm{Ferrous} \mathrm{oxalate} \mathrm{percent} \mathrm{yield} (\%)={\displaystyle \frac{\mathrm{Actual} \mathrm{yield} \left(\mathrm{g}\right)}{\mathrm{Theoretical} \mathrm{yield} \left(\mathrm{g}\right)}}100\%$$

(1)

where the ‘theoretical yield’ is the amount of ferrous oxalate dihydrate (FeC₂O₄∙2H₂O) obtained from stoichiometric reaction considering the complete conversion of the iron present in 1.75 g of ferrotitaniferous sand (0.6 FeTiO₃∙0.4 Fe₂O₃) loaded into the reactor, that equals to 2.84 g of FeC₂O₄∙2H₂O. The ‘actual yield’ is the amount of FeC₂O₄∙2H₂O obtained from the synthesis reaction at given processing conditions. To evaluate whether temperature and reaction time influenced the ferrous oxalate percent yield, a two-way analysis of variance (ANOVA) and Tukey’s multiple range test (p-value < 0.05) were performed using Minitab software v.17 (Minitab Inc., State College, PA, USA).

2.3. Kinetics of Ferrous Oxalate Synthesis

The rate of dissolution of solid ferrotitaniferous sand with the subsequent formation of ferrous oxalate dihydrate within a pressurized 1.5 M oxalic acid medium at either 115 or 135 °C was measured under dynamic conditions. Therefore, the change in ferrous ion concentration [Fe⁺²_aq] over 2, 4, 8, and 12 h in the solution was quantified by atomic absorption spectroscopy (AAS) as described in Section 2.4.6.

Based on ferrous ion concentration [Fe⁺²_aq] data, the reaction pathway and the order of reaction were determined. In addition, the rate constant (k), activation energy (E_a), and frequency factor (A) values were calculated according to the rate law for a second-order reaction and the Arrhenius model, respectively. Furthermore, the rate constant at 155 °C (k_155°C) was calculated, and ferrous oxalate dihydrate percent yield was predicted. For comparison purposes, an additional synthesis was carried out at 155 °C/2 h to compare the experimental and predicted FeC₂O₄∙2H₂O percent yield.

Finally, to determine whether the 1.5 M oxalic acid reaction medium degraded to CO₂ under the moderate sCW conditions used herein, experimental runs loaded with 300 g of 1.5 M oxalic acid solution were conducted at 135 °C for 4 and 12 h. Then, the generated headspace gas was collected from the reactor by using a gas sampling double-ended cylinder for further CO₂ determination by gas chromatography (ASTM D1945). A Clarus 500 chromatograph (Perkin Elmer, Oak Brook, IL, USA) equipped with a TCD detector, with helium as the carrier gas, was used.

2.4. Characterization of As-Synthesized Ferrous Oxalate and Remnant Supernatant

2.4.1. Phase Composition and Crystallographic Structure

Ferrous oxalate powders synthesized at either 115 or 135 °C and reaction times of 2, 4, 8, and 12 h were analyzed by X-ray diffraction (XRD) to study the effect of reaction conditions on phase composition and crystallographic structure.

The XRD measurements were carried out for pooled samples on a D2 Phaser Benchtop diffractometer (Bruker, Billerica, MA, USA) equipped with a LYNXEYE XE-T detector, which has a superb energy resolution < 880 eV (Cu Kα) that enables the filtering of unwanted scattering, from K_β to fluorescence radiations, and delivers data with superior signal-to-noise. The measurements were performed using Cu K_α (λ = 1.5406 Å) as the radiation source operating at 30 kV and 10 mA. The samples were scanned within a 2θ angle range of 10 to 70°, with 0.01° step size and 0.250 s per step.

For phase quantification, powder XRD data were refined via the Rietveld refinement program BGMN using Profex 5.1.1 graphical interface [32]. As such, the Rietveld refinement strategy considered the scale factor and preferred orientation modeled by a six-order spherical harmonic function, anisotropic crystallite size broadening, and micro-strain broadening. The peak profile fitting was calculated based on the Fundamental Parameters convolution Approach (FPA), which considers the contribution of the Cu Kα emission profile, optical elements in the beam path, and the physical variables of the specimen [32].

2.4.2. Raman Spectroscopy

Raman spectroscopy was used to obtain information about the polymorphism in the Ti-based precipitate (Figure 1b). The Raman spectra were collected by a LabRAM HR Evolution confocal microscopic Raman spectrometer (HORIBA Scientific, Montpellier, France) with a 633 nm laser irradiation and power of 12.5 mW, following the methodology previously described by Valdivieso et al. [30].

2.4.3. Morphology

Scanning electron microscopy (SEM) was used to observe morphological variations in the as-synthesized ferrous oxalate powders as a function of reaction conditions. The as-obtained ferrous oxalate powders were mounted on specimen stubs using an electrically conductive double-sided carbon tape prior to the analysis. The images were acquired using a field emission SEM, Tescan Mira3 (Brno, Czech Republic), with a secondary electron detector at 20 kV electron beam energy.

2.4.4. Specific Surface Area

The specific surface area of ferrous oxalate powders was determined by nitrogen adsorption on a NOVAtouch gas sorption analyzer (Quantachrome Instruments, Boynton Beach, FL, USA), and the results were evaluated using the Brunauer–Emmett–Teller (BET) theory and Barrett-Joyner-Halenda (BJH) pore size distribution analysis. The powders were vacuum degassed at 150 °C for 180 min prior to the sorption analysis using nitrogen at cryogenic conditions as the adsorbate.

2.4.5. Band Gap Energy

The optical band gap of synthesized ferrous oxalate powders was estimated using UV-Vis diffuse reflectance spectroscopy (DRS) according to the methodology described by Makuła et al. [33]. A UV/VIS Lambda 365 spectrophotometer equipped with an integrating sphere (Perkin Elmer, Oak Brook, IL, USA) was used to measure the diffuse reflectance (% R) of the samples. The collected spectra were converted to pseudo-absorption spectra by means of the Kubelka–Munk function (F(R_∞)), and the band gap energy was calculated from the Tauc plot method (Equation (2)).

$${\left(\mathrm{F}\left({\mathrm{R}}_{\infty }\right)·hv\right)}^{1/\gamma }=\mathrm{B}\left(hv-{\mathrm{E}}_{\mathrm{g}}\right)$$

(2)

where F(R_∞) is the Kubelka–Munk function, h is the Plank constant (1.24 × 10⁻⁶ eV), ʋ is the frequency of the photon, B is a constant, E_g is the band gap energy (eV), and γ is a factor that equals to 1/2 or 2 for direct and indirect transition band gaps, respectively.

2.4.6. Iron and Titanium Contents in the Supernatant

The supernatants from the reaction synthesis at either 115 or 135 °C and reaction times from 2 to 12 h were analyzed AAS to determine the iron and titanium contents as a function of time. The analysis was performed using the AAnalyst 300 flame atomic absorption spectrophotometer (Perkin Elmer, Inc., Waltham, MA, USA).

3. Results and Discussion

3.1. Synthesis of Ferrous Oxalate by Pressurized Aqueous Oxalic Acid

Ferrous oxalate dihydrate was obtained from ferrotitaniferous Ecuadorian sand as a yellowish powder using pressurized 1.5 M oxalic acid as reaction medium at either 115 or 135 °C. The effects of temperature and reaction time on ferrous oxalate percent yield, in accordance with Equation (1), are shown in Figure 2. Overall, both temperature and reaction time influenced the ferrous oxalate percent yield, as it significantly increased (p < 0.05) at high temperatures and longer reaction times. The yield at 135 °C, for example, was significantly higher than that at 115 °C at any given reaction time. As such, ferrous oxalate percent yield at 135 °C/12 h was 2.3-fold higher than that at 115 °C/12 h. A similar trend was observed when the reaction time increased from 2 to 8 h at either 115 or 135 °C since ferrous oxalate percent yield increased by 17.4% and 27.5%, respectively. Reaction times longer than 8 h at 115 °C or 135 °C did not significantly influence the percent yield. These results indicate that the formation and crystallization of ferrous oxalate dihydrate in aqueous oxalic acid medium were predominant within 8 h of reaction regardless of the temperature used, with faster reaction rates during the first 4 h since steeper-sloped curves were apparent during this period (Figure 2).

In addition to the classical nucleation theory, the critical supersaturation (S_c) phenomenon and the nucleation rate at the S_c have been more recently reported [34,35]. The S_c refers to the level of supersaturation above which spontaneous nucleation occurs instantly. He et al. [35] reported that S_c was lower at higher temperatures; therefore, the nucleation rate was higher [34,36]. Thus, steep-sloped curves in Figure 2 may indicate that (i) a 20 °C temperature rise lowered the S_c, thus increasing the nucleation rate, resulting in a more rapid crystallization and higher yields of FeC₂O₄∙2H₂O, (ii) due to the high supersaturation levels, a primary nucleation mechanism prevailed within the first 4 h of FeC₂O₄∙2H₂O crystallization [37], and (iii) upon a decrease in the supersaturation level with time, secondary nucleation mechanisms and ferrous oxalate crystal growth could occur [38].

Therefore, it can be inferred that the synthesis of ferrous oxalate dihydrate with hot pressurized 1.5 M oxalic acid was primarily driven by temperature. The higher the temperature, the faster the primary nucleation occurs within the first 4 h of the reaction. Thus, longer reaction times at low temperatures did not necessarily ensure a maximum yield.

In addition, the effect of temperature on pressure was also evident (Figure 1b) since pressure increased considerably at 135 °C. Interestingly, this variation in pressure increased with time and was more prevalent between 8 and 12 h of reaction when the ferrous oxalate percent yields were not significantly different. This increase in pressure can be associated with thermal degradation of the reaction medium since decarboxylation of aqueous dicarboxylic acids such as oxalic acid has been observed at temperatures below 140 °C with subsequent emission of CO₂ and H₂O [39]. The gas chromatography results revealed that degradation of 1.5 M oxalic acid solution to CO₂ occurred indeed at 135 °C over time. As such, the mole percentage for CO₂ in the headspace gas obtained after 4 h and 12 h of reaction was 20.97% ± 0.30% and 31.59% ± 0.59%, respectively. Thus, since temperature accelerates reactions, faster degradation of 1.5M oxalic acid solution occurred during the synthesis of FeC₂O₄∙2H₂O at 135 °C than at 115 °C, as indicated by the pressure profiles in Figure 1c.

Although high pressure can also accelerate some organic reactions in solution [40,41], such reactions are particularly associated with negative activation volume (∆V^≠) [41,42,43,44,45]. For ferrous oxalate synthesis, neither the autogenous pressure generated at 135 °C falls within the high-pressure range (>1 kbar) nor a negative activation volume has been reported [42]. Therefore, ferrous oxalate precipitation was not expected to be sped up by autogenous pressure but through an increase in temperature.

3.2. Kinetics of Ferrous Oxalate Formation

Since ferrotitaniferous sand is an ilmenite–hematite solid solution (0.6FeTiO₃∙0.4Fe₂O₃) [31,46], its dissolution prior to FeC₂O₄∙2H₂O precipitation can be envisioned according to Equation (3), which to some extent, resembles the reported dissolution mechanism of pure hematite in oxalic acid [47]. Upon Equation (3), the required time for complete dissolution of ferrotitaniferous sand in pressurized 1.5 M oxalic acid was experimentally determined by performing short-time reactions (from 30 to 90 min) at either 115 °C or 135 °C until a single-phase system was visually observed, which corresponded to 90 and 54 min, respectively.

In addition, a net ionic equation was proposed for the combination reaction and FeC₂O₄∙2H₂O precipitation as outlined in Equation (4), where the total amount of iron present in the ferrotitaniferous sand and the amount of oxalic acid present in the 1.5 M solution were the limiting and the excess reagents, respectively.

$${\mathrm{FeTiO}}_3·{{\mathrm{Fe}}_2{\mathrm{O}}_3}_{\left(s\right)}+5{\mathrm{H}}_3{\mathrm{O}}^{+}+5{\left({\mathrm{HC}}_2{\mathrm{O}}_4\right)}^{-1}\to 3{\mathrm{Fe}}^{+2}+{\mathrm{TiO}}^{+2}+2{\mathrm{CO}}_2+4{\left({\mathrm{C}}_2{\mathrm{O}}_4\right)}^{-2}+10{\mathrm{H}}_2\mathrm{O}$$

(3)

$${\mathrm{Fe}}^{+2}+{\left({\mathrm{C}}_2{\mathrm{O}}_4\right)}^{-2}+{\mathrm{H}}_2\mathrm{O}\stackrel{\mathrm{k}}{\to }{\mathrm{FeC}}_2{\mathrm{O}}_4·2{\mathrm{H}}_2\mathrm{O}$$

(4)

Therefore, the rate of dissolution of solid ferrotitaniferous sand with subsequent formation of ferrous oxalate dihydrate within hot pressurized 1.5 M oxalic acid was measured according to Equation (4).

Based on experimental data (Section 3.1), the dissolution of the iron-bearing sand (Fe⁺²_aq) in the hot pressurized oxalic acid medium followed second-order kinetics and was used to fit the curve of (1/Fe⁺²_aq) versus time. The corresponding differential rate law is detailed in Equation (5):

$${\displaystyle \frac{d \left[\mathrm{Fe}\right]}{dt}}=-k{\left[\mathrm{Fe}\right]}^2$$

(5)

where the rate constant k (L∙mg⁻¹·s⁻¹) was estimated from the slope of the linear plot of the inverse concentration of iron ($1/\left[\mathrm{Fe}\right]$) in supernatant versus time, d[Fe]/dt is the reaction rate or the change in the concentration of the limiting reagent with time (mg of iron∙L⁻¹ of solution·s⁻¹), [Fe] is the concentration of iron at a given time (mg of iron∙L⁻¹ of solution), as determined by AAS, and the concentration of iron initially present in 1.75 g of the iron-bearing sand loaded into the reactor equals to 2945.83 mg∙L⁻¹.

Figure 3 showcases the decay trend of Fe⁺²_aq ions from dissolved ferrotitaniferous sand because of the precipitation of ferrous oxalate dihydrate at either 115 °C or 135 °C. It is clear, however, that at 135 °C, both Fe⁺²_aq decay and FeC₂O₄∙2H₂O formation through precipitation occurred faster (Figure 3B). Furthermore, the color change in the resulting supernatants (inset in Figure 3B(b–e)) also indicated the increasing trend of iron precipitation from solution at higher temperatures. In general, more yellow the supernatant, the higher the Fe⁺²_aq concentration. At 135 °C, the supernatant turned from yellow to transparent within 8 h, indicating a rapid iron precipitation from the solution. Conversely, at 115 °C, a yellowish supernatant was still visible even after 12 h, indicating that the reaction was temperature-dependent.

To assess the sensitivity of the reaction to temperature variation, a second-order reaction equation (Equation (4)) and the Arrhenius model (Equation (6)) were used to estimate the rate constant (k) and activation energy (E_a) values, respectively.

$$\ln \left(k\right)=\left(-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)\times {\displaystyle \frac{1}{\mathrm{T}}}+\ln \left(\mathrm{A}\right)$$

(6)

where E_a is the activation energy (kJ∙mol⁻¹), R is the ideal gas constant (8.314 × 10⁻³ kJ∙mol⁻¹∙K⁻¹), T is the temperature (K), and A is the frequency factor (L·mg⁻¹∙s⁻¹).

The estimated kinetic parameters are summarized in

Table 1. Reaction rate constants (k) and Arrhenius activation energy (E_a) for α-FeC₂O₄∙2H₂O synthesis with pressurized 1.5 M oxalic acid at 115 and 135 °C.

T	k		A	Ea
(°C)	(L∙mg−1·s−1)	R2	(L∙mg−1·s−1)	(kJ∙mol−1)
115	8 × 10−9	0.94	1.84 × 1012	151.52
135	8 × 10−8	0.99

T: temperature, R²: coefficient of determination of the regression line derived from the plot (1/[Fe]) vs. time. A: pre-exponential factor of Arrhenius equation.

. According to Table 1, the reaction rate constant value at 135 °C was one order of magnitude higher than the calculated for 115 °C, indicating that a 20 °C temperature rise considerably increased the kinetic energy of the system, and ions were more likely to collide, react and yield ferrous oxalate. The temperature effect on the reaction was also reflected in the Arrhenius model by the high value of the frequency factor (A), which indicated that, a high rate of molecular collisions per second occurred as the temperature increased.

Based on the Arrhenius model, the higher the activation energy, the more difficult it is for the reaction to occur and, thus, the higher the temperature sensitivity of the reaction rate. Based upon the E_a value estimated for ferrous oxalate synthesis of 151.52 kJ∙mol⁻¹, a considerable activation barrier must be overcome, showing that the rate of the reaction is temperature dependent. In contrast, physical events that are less dependent on temperature, such as physisorption, exhibit E_a lower than 40 kJ∙mol⁻¹ [48]. Although the E_a values for ferrous oxalate synthesis with pressurized hot oxalic acid have not been yet reported, the E_a value estimated herein was in agreement with the range of E_a values found for hydrolysis (142 kJ∙mol⁻¹) [49] and degradation (203 kJ∙mol⁻¹) [50] reactions of some other organic compounds, such as polygalacturonic acid and hemicellulose, in hot pressurized citric and malic acids, respectively.

According to the Arrhenius model, the temperature dependence of reaction rate for the synthesis of ferrous oxalate dihydrate by hot pressurized 1.5 M oxalic acid can be expressed as follows (Equation (7)):

$$k_{\left(\mathrm{T}\right)}=1.84\times {10}^{12}·e^{\left(-{\displaystyle \frac{151.52 }{\mathrm{R} \mathrm{T}}}\right)}$$

(7)

Then, based on Equations (5) and (7), the concentration of ferrous oxalate dihydrate can be estimated for a higher temperature range (>135 °C), which is of great interest for optimizing operating conditions and maximizing yield. For such purpose, an additional synthesis reaction was carried out at 155 °C for 2 h, to compare both the experimental and predicted ferrous oxalate dihydrate percent yield. Based on Equations (5) and (7), the estimated rate constant k_(155°C) and the predicted ferrous oxalate dihydrate percent yield were $6.75\times {10}^{-7}$ L·mg⁻¹∙s⁻¹ and 93.7%, respectively, whereas the experimental ferrous oxalate dihydrate percent yield obtained at 155 °C/2 h was 94.9% (vide infra, Section 3.3)). This confirms that the experimental yield was successfully aligned with the predicted one.

Dissimilar to the decay trend of iron Fe⁺²_aq in the supernatant, displayed in Figure 3, the titanium concentration did not change over time regardless of the temperature and time. Further research is needed to determine whether titanium is still in its ionic tetravalent form (Ti⁺⁴) or as a coordinated compound in the oxalic acid reaction medium. In some previous studies, water-soluble titanium (IV) oxalate complexes such as TiO(C₂O₄)₂⁻² have been reported in an acid medium [51,52].

In addition, water-soluble titanium complexes have also been synthesized from hydroxo-peroxo titanium complexes and hydroxycarboxylic acids such as citric, glycolic, and lactic acids to be used as precursors for the synthesis of TiO₂ polymorphs [53]. Interestingly, hydrothermal decomposition of such aqueous titanium-peroxo-citrato and titanium-peroxo-glycolato complexes at 200 °C/24 h and pH = 6 has resulted in anatase and rutile structures, respectively [54].

Therefore, the iron Fe⁺²_aq depleted supernatant obtained at 135 °C/12 h (Figure 3B(point e)) was investigated as a precursor for TiO₂ synthesis. Accordingly, this supernatant was processed further in the reactor at 155 °C for 12 h [30]. As a result, a fine greyish precipitate was obtained and analyzed by Raman spectroscopy (Figure 4).

Figure 4 shows a Raman spectrum with peaks at wavenumbers of 246, 442, and 610 cm⁻¹, which may correspond to the rutile crystalline phase [55,56]. However, an intense peak at 150 cm⁻¹, which is distinctive to the anatase phase, was also observed [56,57]. Thus, Raman spectrum in Figure 4 indicate features from both TiO₂ polymorphs, rutile, and anatase, with anatase having a weaker contribution as its peaks at 197, 395, 512, and 635 cm⁻¹ cannot be distinguished clearly [56,58]. The Raman spectrum suggested that TiO₂ in the rutile phase was indeed synthesized together with the anatase phase from the acidic iron-depleted supernatant resulting from the reaction at 135 °C/12 h when further heated at 155 °C/17 bar/12 h. Therefore, the liquid co-product of ferrous oxalate synthesis from ferrotitaniferous mineral sands might be used as a promising precursor for obtention of TiO₂ polymorphs without the shortcomings associated with the use of strong mineral acids and energy-intensive processes [54,59].

3.3. Characterization of the As-Synthesized Ferrous Oxalate

3.3.1. Phase Composition and Crystallographic Structure

According to the XRD patterns presented in Figure 5A, the as-synthesized products are monophasic and exhibited strong well-defined peaks at 2θ angles of 18.6, 18.97, 23, 24.7, 24.9, 30, 33.9, 34.2, 34.5, 40.2, 45, 48.3, and 50.44°, which corresponded to α-FeC₂O₄∙2H₂O crystalline phase (indexing PDF # 72-1305).

Interestingly, it was observed that ferrous oxalate dihydrate was also rendered even under the mildest condition of 115 °C/2 h, suggesting that the reaction occurring in a hot pressurized 1.5 M oxalic acid medium is highly feasible. In addition, peaks observed in Figure 5A(a–d) became sharper and more intense as the reaction time approached 12 h, regardless of the temperature condition used. At 115 °C/12 h, for example, the intensity of the representative peak at 2θ angle of 18.6° was 3.6-fold higher than that obtained at 115 °C/2 h. Similarly, at 135 °C/12 h, the intensity of the peak at the same 2θ angle was 1.9-fold higher than that obtained at 135 °C/4 h. Therefore, the high intensity and sharp peaks in the XRD patterns obtained at long reaction times can be attributed to crystal growth. Based on the XRD, the estimated kinetic parameters, and percent yield data, it can be concluded that pressurized 1.5 M oxalic acid at 135 °C/4 h may be a suitable reaction medium for fast production of pure α-FeC₂O₄∙2H₂O from ferrotitaniferous sand without sacrificing the yield calculated as 77% for this reaction condition.

Therefore, the XRD pattern of FeC₂O₄∙2H₂O powder synthesized with 1.5 M oxalic acid at 135 °C/4 h was further subjected to Rietveld refinement to verify its crystal structure. The Rietveld refinement profile of the XRD pattern of FeC₂O₄∙2H₂O synthesized at 135 °C/4 h is depicted in Figure 5B. Rietveld results confirmed that, at 135 °C/4 h, the α-FeC₂O₄∙2H₂O phase was formed as the only crystal phase, with a crystal structure corresponding to the monoclinic crystal system belonging to the C2/c space group.

The lattice parameters werecalculated as a = 12.0166 Å; b = 5.5563 Å; c = 9.9227 Å; ß = 128.536° with a unit-cell volume of 518.235 ų, and a 1.1 µm average crystallite size. The reliability parameters of the refinement were R_wp = 3.73%, R_exp = 3.24%, and goodness-of-fit (GoF) = 1.15. The unit-cell values were in agreement with those reported in the powder diffraction file PDF # 72-1305, corresponding to iron oxalate hydrate (monoclinic, C2/c, a = 12.060 Å, b = 5.550 Å, c = 9.804 Å, ß = 127.97°) as well as to the cell-unit parameters previously reported by Ogasawara and Koga [60] for synthetic α-FeC₂O₄∙2H₂O.

3.3.2. Morphology

Figure 6 depicts an overall trend of α-FeC₂O₄∙2H₂O crystallizing as overlapped octahedra of different sizes. The effect of temperature on the rate of crystal growth was apparent. As a result, crystal growth was faster at 135 °C, resulting in larger and better-defined octahedral crystals regardless of the reaction time used (Figure 6a,c). In addition, at 135 °C, as the reaction time approached 8 h, ferrous oxalate crystal seeds grew new octahedron crystal patterns (Figure 6e). Interestingly, irregular cavity shapes were observed in many crystals (Figure 6b–e). It was evident, however, that after a 12 h reaction at 135 °C, oxalate crystals became more compact (Figure 6f). This suggests that the crystallinity increased with longer reaction times, as indicated by the high-intensity peaks in the XRD pattern obtained at 135 °C/12 h (Figure 5A(d)). Furthermore, at long reaction times, ferrous oxalate crystal habit (Figure 6e,f) resembled the (010) lattice plane (white lines in Figure 6e) of the monoclinic unit cell, probably due to its morphological stability (Bravais–Friedel–Donnay–Harker law) in the reaction medium [61,62].

3.3.3. Specific Surface Area and Optical Band Gap Energy

To determine whether the severity of thermal treatment influenced the specific surface area and band gap energy of synthesized ferrous oxalate dihydrate, both products from mild (115 °C/2 h) and moderate (135 °C/12 h) reaction conditions were evaluated.

According to Figure 7, nitrogen gas adsorption curves obtained for ferrous oxalate dihydrate resembled a type IV isotherm characteristic of mesopores that involved capillary condensation (hysteresis loop) [63]. Then, in the limit of low relative pressures (P/P₀) corresponding to <0.1 (see at Figure 7), ferrous oxalate exhibited a Langmuir-like behavior being covered by an adsorbate monolayer. However, the low gas uptake at such low relative pressures and the lack of a clear inflection point indicated weak interactions between adsorbate and adsorbent. The calculated surface areas in the low relative pressure region were 33.7 and 31.9 m²∙g⁻¹ for ferrous oxalate synthesized at 115 °C/2 h and 135 °C/12 h, respectively. Such variation could be associated with particle size growth as the size of ferrous oxalate synthesized at 135 °C/12 h was larger (>10 µm, Figure 6f) than that synthesized at 115 °C/2 h (<5 µm, Figure 6a), resulting in a reduction in the specific surface area.

In addition, the specific surface area of the as-synthesized ferrous oxalates was larger than the previously reported for ferrous oxalate dihydrate nanorods as 27.5 m²∙g⁻¹ [16]. Furthermore, at medium pressures, multilayer formation was apparent on the ferrous oxalate surface as adsorption continued to increase. In high relative pressures, capillary condensation also occurs regardless of the synthesis conditions. For ferrous oxalate obtained at 115 °C/2 h, capillary condensation occurred at a relative pressure of 0.94 for 13.3 nm-diameter pores, whereas for ferrous oxalate obtained at 135 °C/12 h, it occurred between 0.94 and 0.99 for 1.7 nm-diameter pores. In addition, a type B hysteresis loop corresponding to slit-shaped open pores (Figure 6f) [63,64,65] was observed in both isotherms. Interestingly, the ferrous oxalate obtained at 135 °C/12 h exhibited the smallest hysteresis loop, which correlated with its smaller pore size and more compacted structure (fewer hollow cores, Figure 6f). Conversely, a low relative pressure hysteresis (<0.5) was observed for ferrous oxalate obtained at 115 °C/2 h, suggesting swelling of the adsorbent or irreversible chemical adsorbate–adsorbent interaction [63].

Therefore, the conditions of the thermal treatment influenced the number and size of pores as well as the specific surface area of ferrous oxalate. Consequently, it can be expected that high temperatures and long reaction times can lead to the compaction of ferrous oxalate particles into bigger aggregates with reduced specific surface area but increased chemical stability.

Figure 8b shows band gap energies corresponding to a direct electronic transition as indicated by linear fit in (F(R_∞)∙hʋ)² versus hʋ plot. For ferrous oxalate dihydrate, the reported literature is not conclusive on whether its band gap is direct or indirect. Nonetheless, polymorphism (α or β), crystal system (monoclinic or orthorhombic), and particle size of synthesized FeC₂O₄∙2H₂O as well as the selected technique for crystal growth and purity of used iron precursors are some factors that can be associated with the reported type of band gap. For example, direct band gap values of 2.17 eV [20] and 2.10 eV [21] have been reported for α-FeC₂O₄∙2H₂O with the morphology of cuboid-like microrods synthesized from FeSO₄∙7H₂O (purity > 99%) and (NH₄)₂C₂O₄∙H₂O (purity > 99.5%) at 80 °C/2 h and 95 °C/6 h, respectively. Also, a direct band gap of 2.52 eV has been measured for the rectangular prism shaped β-FeC₂O₄∙2H₂O [66] synthesized from FeSO₄(NH₄)₂SO₄∙6H₂O and H₂C₂O₄ at 80 °C/2 h.

In addition, a smaller direct band gap of 1.9 eV has been determined for commercial β-FeC₂O₄∙2H₂O [67]. Conversely, an indirect band gap of 1.91 eV has also been reported for irregularly shaped α-FeC₂O₄∙2H₂O nanoparticles obtained from reagent grade FeSO₄∙7H₂O, and (NH₄)₂C₂O₄∙H₂O at 80 °C/2 h [68], and an even smaller indirect band gap value of 1.36 eV has been calculated using DFT method based on a correlation-consistent basis set model (PBE/ucc-pVDZ) [12]. Moreover, band gap values of 2.36 eV and 2.24 eV of unspecified transition type have been reported for β-FeC₂O₄∙2H₂O nanosheets obtained by a mechanochemical method [69] and β-FeC₂O₄∙2H₂O nanorods [16], respectively.

As depicted in Figure 8b, the conditions of the thermal treatment did not drastically influence the optical band gap energy of as-synthesized FeC₂O₄∙2H₂O. As such, the band gap energy of α-FeC₂O₄∙2H₂O obtained under moderate conditions (135 °C/12 h) was 2.77 eV (447.6 nm), whereas at mild conditions (115 °C/2 h) it was 2.74 eV (452.5 nm). In both cases, however, the band gap values corresponded to the photon energy of the blue color (2.75 eV, λ = 450 nm), meaning that synthesized α-FeC₂O₄∙2H₂O only absorbs light with λ ≤ 450 nm, which leads to its characteristic yellow color (2.14 eV, λ = 580 nm).

Based on the estimated optical Eg values of as-synthesized ferrous oxalates and the previously reported photocatalytic and photo-initiated Fenton activities of reagent grade α-FeC₂O₄∙2H₂O [18,20,21], it can be inferred that as-synthesized ferrous oxalates can behave such as small band-gap semiconductors (1.5 eV < Eg < 3.0 eV) [70]. Therefore, their use as efficient metal-organic photocatalysts [20] via the electron transfer mechanism is promising [71].

4. Conclusions

This study reports a novel approach to synthesizing high-purity alpha-ferrous oxalate dihydrate directly from ferrotitaniferous mineral sands using hot pressurized 1.5 M oxalic acid in subcritical water conditions.

The evaluated temperatures, 115 and 135 °C, were in the low-temperature range of the subcritical water region and yielded between 38.7% and 89.9% of α-FeC₂O₄∙2H₂O within 12 h. For the synthesis, a precipitation reaction was proposed as the mechanistic pathway, in which the rate of reaction followed second-order kinetics.

Additionally, for the first time, as the authors are aware, the reaction rate constants, Arrhenius activation energy, and frequency factor for α-FeC₂O₄∙2H₂O synthesis from ferrotitaniferous sands using hot pressurized aqueous oxalic acid are reported. According to the kinetic parameters, temperature drives the synthesis reaction. Nonetheless, temperature had little effect on crystal morphology, specific surface area, or band gap energy. Interestingly, all as-synthesized α-FeC₂O₄∙2H₂O powders had optical band gap energies in the visible-light range.

These findings provide fundamental knowledge about the feasibility of using a hot pressurized oxalic acid for the conversion of iron-bearing mineral precursors into α-FeC₂O₄∙2H₂O with potential solar-driven photocatalytic activity.