Efficiency of Mechanochemical Ball Milling Technique in the Preparation of Fe/TiO₂ Photocatalysts

Abstract

Rapid population growth and widespread industrialization are the main contributing factors to the increasing contamination of the world’s diminishing freshwater resources. This work investigates Fe/TiO₂ as an efficient and sustainable photocatalyst for treating organic micropollutants in water. The photocatalysts prepared by these mechanochemical methods used a high-energy ball milling technique to manipulate Fe/TiO₂’s structural, optical, and catalytic properties for the photo-oxidation of 2,4-Dichlorophenol (2,4-DCP). Doping with iron effectively reduced the band gap of rutile TiO₂ from 3 to 2.22 eV. By reducing the ball/powder ratio from 34 to 7, the removal efficiency of 2,4-DCP increased from 65.2 to 84.7%. Measuring the TOC indicated 63.5 and 49.4% mineralization by Fe/TiO₂-7 and rutile TiO₂, respectively, after 24 h. The energy yields for the Fe/TiO₂ and rutile TiO₂ were 0.13 and 0.06 g 2,4-DCP/kW h, respectively.

1. Introduction

Global climate change, population growth, and urbanization will see more than four billion people suffer water stress for at least one month each year by 2050 [1,2]. Moreover, decades of unbridled agricultural and industrial expansion have heavily polluted lakes, rivers, and aquifers with heavy metals, chemicals, pesticides, herbicides, and pharmaceuticals that are detrimental to human health and the environment [3,4,5]. Among the emerging pollutants, chlorophenols pose the biggest challenge, being persistent and refractory toxicants that bioaccumulate in the food chain. In addition, their wide use in pesticides, herbicides, fungicides, dyes, and pharmaceuticals [6,7] means they can be found extensively in the environment [8].

The chemical compound 2,4-dichlorophenol (2,4-DCP) is widespread, and due to its toxicity, carcinogenicity, and mutagenicity, it is a priority pollutant of the US Environmental Protection Agency (EPA) [9]. In addition, conventional water and wastewater treatment processes are ineffectual [10,11,12] at treating it, as demonstrated by its increasing occurrence in drinking water at levels exceeding the permissible concentrations of 0.5 ppm [6]. Thus, providing the impetus to develop more efficient treatment technologies such as advanced oxidation processes (AOP) that include photocatalysis [13], UV/H₂O₂ [14], UV/O₃ [15], UV/H₂O₂/O₃ [16], and Fenton/sonolysis methods [17].

Photocatalysis is attractive as it can harness light, particularly sunlight, to generate reactive radical species to degrade pollutants. It provides a low cost, efficient, and sustainable remediation method for treating recalcitrant contaminants [18,19,20]. TiO₂ photocatalysts are popular due to their relatively low cost, excellent durability, superb chemical stability, and high electron transport [21]. A critical shortcoming of TiO₂ is its wide bandgap, which limits its light utilization to UV wavelengths [22]. Introducing foreign atoms such as Zn [23], Gd [24], Cu [25], Lu [26], and Al [27] can narrow its bandgap, and iron proved to be the most promising [28,29]. Iron’s ionic radii, comparable to Ti⁴⁺, are readily inserted into TiO₂’s crystal lattice. Iron is also inexpensive and non-toxic compared to the other dopants.

The mechanochemical catalyst preparation process affords a simple, inexpensive, and green preparation of the photocatalysts [30]. This technology can help to reduce particle sizes, which plays a vital role in improving photocatalytic performance and can produce nanoparticles of a large surface area that are enriched in oxygen-containing functional groups [31]. In mechanochemical synthesis, the type of ball mill (e.g., zirconium dioxide), nature of the milling material [32], and milling speed and time [33] are crucial variables. An appropriate ball-to-powder weight ratio (B/P) is another critical parameter in high-energy ball milling. A high B/P sees more ball-to-ball impact and inefficient milling [24], while a very low B/P from overfilling reduces grinding [34]. Prior studies have often neglected the effects of ball sizes when studying B/P, which will be considered in our study of Fe/TiO₂ photocatalyst preparation and its impact on Fe³⁺ insertion and the subsequent structure, morphology, optical, and photocatalytic properties of the photocatalysts.

2. Materials and Methods

2.1. Materials

Titanium (IV) oxide, rutile (99.5% trace metals basis, CAS number: 1317-80-2), and iron (III) nitrate nonahydrate (≥98%, CAS number: 7782-61-8) catalyst precursors and 2,4-dichlorophenol (99%, CAS number: 120-83-2) were purchased from Sigma-Aldrich and used without further purification. The 2,4-DCP solutions were in deionized distilled water.

2.2. Fe/TiO₂ Photocatalysts

The Fe/TiO₂ catalysts were synthesized in a dry planetary ball mill (PBM-0.4A) at room temperature in air. Three catalyst batches were milled at B/P = 34, 17, and 7 from a powder mixture of 5 wt.% Fe(III) and 95 wt.% TiO₂. Briefly, the Fe/TiO₂-34 was obtained by milling 2.5 g catalyst precursor mixture with zirconia beads of 15 mm (2), 12 mm (8), 8 mm (6), and 5 mm (42) in size; the Fe/TiO₂-17 was prepared by milling 5 g catalyst precursor mixture; and the Fe/TiO₂-7 was milled from 5 g catalyst precursor mixture with 8 mm (6) and 5 mm (42) beads. The milling speeds and durations were kept at 400 rpm and 10 h, respectively. The direction of the milling rotation reversed every 10 min. Before conducting photocatalytic reaction, the catalyst was dissolved in 15 mL DDI water and then centrifuged (5000 rpm, 3 min) to remove the possible iron ions on its surface. The catalyst was then dried at 65 °C.

2.3. Catalyst Characterization

The catalysts’ structures, morphologies, and surface textures were determined by X-ray diffraction (XRD), scanning electron microscopy (SEM), and nitrogen physisorption. The powder XRD was conducted in a PANalytical X-ray Diffractometer, Model X’pert Pro, using 2 kW Cu-Kα X-ray radiation with a graphite monochromator at a 1.5406 Å wavelength and a 7 degrees/min scan rate. The crystal size was determined using the Scherrer equation (Equation (1)) by taking the average of the dominant peaks in the diffraction data [35,36], as follows:

$$D_S =\frac{K\lambda }{{\beta }_{D}cos\theta }$$

(1)

where, D_s is the crystal size (nm), K is the shape factor, β_D is FWHM, λ is the wavelength of the CuKα radiation (1.5406 Å), and θ stands for the angle of the XRD. K is generally considered to be 1.0 for spherical particles.

The SEM imaging was performed using a JEOL JSM-7100F at 15 kV and 1.1 nm resolution on finely dispersed catalyst powder on a clean silicon wafer. Precautions were taken to minimize sample charging by using conductive carbon tape for sample-mounting the aluminum holder and sputtering the sample with a thin gold layer. The nitrogen physisorption isotherms of the catalysts were obtained using a Belsorp mini X (MicrotracBEL Inc., Haan, Germany). Herein, the powder catalyst was outgassed at 150 °C for 6 h before taking measurements, and the Barrett–Joyner–Halenda (BJH) model was used to determine the pore size distribution.

The catalysts’ surface compositions were analyzed with a Kratos X-ray Photoelectron Spectrometer with a monochromatic Al Ka X-ray source at 1486.6 eV and an X-ray power of 150 W. The binding energy (BE) was referenced to the C 1s signal at 284.8 eV. The catalysts were further analyzed with a UV-vis NIR spectrometer (Perkin Elmer, Lambda 950) for their optical properties and band gap energies. The Raman spectra of the catalysts were collected by a Renishaw InVia Raman spectrometer equipped with a notch filter up to 100 cm^–1 and a 100 mW Ar-laser (514 nm) as an excitation source at two scans and 30 s acquisition times under room temperature.

2.4. Photocatalytic Oxidation Reaction

Figure 1 displays the photoreactor for the batch photocatalytic oxidation of the 2,4-DCP. It consisted of a quartz tube reactor placed equidistant (10 mm) from three 6 W GE F6T5 35 lamps. The light from the lamps was filtered by a 3 mm thick polycarbonate UV-cutoff filter to remove wavelengths of <400 nm. The photoreactor setup was placed inside a stainless-steel isolation box to avoid ambient lighting. A pair of DC fan circulated air and maintained a temperature of 25 °C. A 2,4-DCP stock solution (16 mg/L) was diluted to 4.08 mg/L, and 100 mL was added to the reactor, along with 100 mg photocatalyst to create a 1 g/L catalyst loading. The initial pH of the solution was ~6.5 before starting the photocatalytic reaction. The reaction mixture was stirred in the dark for one hour to allow adsorption equilibrium to be reached before turning on the light. Then, at fixed time intervals, 2 mL of the reaction mixture was drawn and centrifuged at 6000 rpm (Labnet C-1200 Mini Centrifuge) for 4 min. The supernatant was analyzed in triplicate by a UPLC (WATERS Acquity H-class) equipped with an ACQUITY UPLC BEH C18 Column (2.1 mm × 100 mm, 1.7 µm). The concentration of 2,4-DCP was recorded by optical spectrum measurement at 284 nm. The mobile phase in the UPLC consisted of acetonitrile and 0.1% trifluoroacetic acid solution at a flow rate of 0.35 mL/min (25 °C). The total organic carbon (TOC) was determined with a Shimadzu TOC-LCPH/CPN. The reaction for each photocatalyst was repeated three times.

The 2,4-DCP conversion was calculated according to Equation (2):

$$\mathrm{conversion}(\%)=\frac{C_0-C_t}{C_0}\times 100$$

(2)

where C₀ and $C_t$ are the initial and final concentrations of the 2,4-DCP and t indicates the reaction time. Meanwhile, the mineralization and energy yield are presented in Equations (3) and (4):

$$\mathrm{mineralization}(\%)=\frac{TO{C}_0-TO{C}_t}{TO{C}_0}\times 100$$

(3)

where $TO{C}_0$ and $TO{C}_t$ are the initial and final concentrations of the TOC at time t, respectively and the energy yield (Y, g/kWh) represents the energy efficiency of the pollutant degradation and is strongly influenced by the reactor design and light source [37]:

$$\mathrm{Y}=\frac{[C_0]\times V\times \mathsf{ɳ}(\%)\times \left(1/100\right)}{P\times t}$$

(4)

where $C_0$ is for the initial concentration of the 2,4-DCP, $\mathsf{ɳ}$ is the degradation efficiency (%) at the reaction time t, and P is the nominal power of the light source (kW). G₅₀ is the energy yield when the degradation efficiency is 50% ($\mathsf{ɳ}$ = 50%) at a reaction time of t₅₀. The related cost is calculated according to the electricity cost in Hong Kong.

3. Results and Discussion

3.1. Fe/TiO₂ Photocatalysts

Figure 2 plots the X-ray diffraction of the rutile and ball-milled TiO₂ and the prepared Fe/TiO₂ samples, which display only diffraction peaks corresponding to the rutile TiO₂ [38,39]. Doping the samples with 5 wt.% Fe³⁺ showed a slight 0.033^° shift for the Fe/TiO₂-34 and Fe/TiO₂-17, and a 0.066^° shift for the Fe/TiO₂-7. In contrast, the ball-milled TiO₂ did not show any diffraction shift from that of the rutile TiO₂. This slight shrinkage in the lattice (i.e., 0.00003 and 0.006 Å, respectively) is consistent with the lattice substitution of Fe³⁺ for Ti⁴⁺. Indeed, the Fe³⁺ ion (0.64 Å) is slightly smaller than the Ti⁴⁺’s radii (0.68 Å), and i9t can readily be inserted into the TiO₂ lattice [40]. The crystal size was smaller following ball-milling, as indicated by the diffraction peaks broadening, and the ball-milled TiO₂ measured 48.3, 45.1, and 41 nm for the Fe/TiO₂-34, Fe/TiO₂-17, and Fe/TiO₂-7, respectively, compared to the rutile TiO₂ (73.3 nm).

The micro-Raman spectra in Figure 3 are consistent with the XRD data and display only the characteristic bands of the rutile TiO₂ at 144 cm⁻¹ (B_1g), 449 cm⁻¹ (E_g), and 611 cm⁻¹ (A_1g). The 237 and 1049 cm⁻¹ bands originate from multiple phonon-scattering processes and NO₃^- ions, respectively [41,42]. The absence of iron-related bands indicates the successful doping of iron into the TiO₂. In addition, there were no contaminants from the milling bar and jar detected in the samples. The ball-milled Fe/TiO₂ had a light-yellow color compared to the rutile and ball-milled TiO₂, as shown in Figure 4.

Figure 5 presents the scanning electron micrographs of the rutile TiO₂ (Figure 5a), Fe/TiO₂-34 (Figure 5b), Fe/TiO₂-17 (Figure 5c), and Fe/TiO₂-7 (Figure 5d). During ball-milling, the sample underwent collisions and attritions which subjected the particles to repetitive breaking–welding–breaking cycles that gradually abraded large particles into fine grains [43]. The ball-milled Fe/TiO₂ samples are less aggregated and better dispersed than the rutile TiO₂ (Figure 5a). Further, the Fe/TiO₂ powders prepared with balls of different sizes (i.e., 15, 12, 8, and 5 mm) have a broader particle size distribution (cf. Figure 5b,c) compared to the powder ball-milled with balls of similar sizes (i.e., 8 and 5 mm), as shown in Figure 5d. The Fe/TiO₂ milled at a higher B/P with less sample led to air-grinding and rounder particles.

Figure 6 plots the N₂ adsorption–desorption isotherms of the rutile TiO₂ and Fe/TiO₂-7. Both isotherms are type IV curves with hysteresis loops, which are associated with a mesoporous range (2–50 nm, according to IUPAC classification) [44]. Mesopores are known to enhance diffusion, and thus enhance photocatalytic reactions [45], and according to the BJH model, the mean pore diameters of rutile TiO₂ and Fe/TiO₂-7 are 7.39 and 14.83 nm, respectively.

Table 1. Surface properties of the rutile TiO₂ and produced Fe/TiO₂ catalysts.

Catalysts	BET Surface Area (m2 g−1)	Average Pore Diameter (nm)	Total Pore Volume (cm3 g−1)	Crystal Size (nm)
Rutile TiO2	2.69	7.39	0.013	73.3
Fe/TiO2-34	5.78	52.41	0.023	48.3
Fe/TiO2-17	6.46	16.20	0.028	45.1
Fe/TiO2-7	6.87	14.83	0.071	41

shows the B/P decreasing from 34 to 7, resulting in a smaller mean pore diameter (52.41 to 16.20 and 14.83 nm).

The UV-Vis diffuse reflectance spectra (DRS) and corresponding Tauc plots of the rutile TiO₂ and Fe/TiO₂-7 are presented in Figure 7. The TiO₂ has an absorption edge at the 368 nm wavelength (Figure 7a) and calculated bandgap energy (E_g) of 3.00 eV (at the 417 nm cut-off Landa), which is consistent with the literature [46,47]. Ball-milled Fe/TiO₂ shifts the absorption edge of the visible region (546 nm) and narrows the bandgap to 2.22 eV. It is believed the iron creates new e^-/h⁺ and traps energy levels (Fe⁴⁺/Fe³⁺ and Fe³⁺/Fe²⁺) between the conduction band and valence band of TiO₂ precursors [48,49]. Lattice defects from ball-milling can also play a role [50]. The earlier work by Carneiro et al. [50] prepared Fe/TiO₂ by ball-milling aeroxide TiO₂ P25 with 10 wt.% Fe to obtain photocatalysts that were 2.45 eV broader, in comparison with this study.

Figure 8 compares the electronic structure of the rutile TiO₂ and Fe/TiO₂-7. According to XPS measurements, the conduction band of Fe/TiO₂-7 is 0.67 eV compared to 0 eV for the rutile TiO₂ and the valence band is 2.89 eV against TiO₂’s 3.00 eV. Electrons and holes are generated by the photocatalyst as it absorbs photons with energy equal to or greater than its bandgap energy. The electrons and holes migrate to the catalyst surface and react with adsorbed O₂ or H₂O to generate reactive hydroxyl radicals (^•OH), superoxide radicals (O₂^•−), and holes (h⁺) that oxidize the pollutants to reactive intermediates. The highly reactive ^•OH (redox potential = +3.06 V) is generated by water decomposition or holes and an ^-OH reaction, while the superoxide anion (O₂^•−) can participate in the production of a hydroperoxyl radical (HO₂^•) and, subsequently, H₂O₂. The general mechanism of photocatalysis over TiO₂ is summarized in Equations (5)–(19) [51,52,53,54,55], as follows:

Light irradiation and photoexcitation: TiO₂ + hv → e⁻ + h⁺

(5)

e⁻ trapping: e⁻_CB → e^-_TR

(6)

h⁺ trapping: h⁺_VB → h⁺_TR

(7)

h⁺ + H₂O → ^•OH + H⁺

(8)

Hydroxyl oxidation: OH⁻ + h⁺ → ^•OH

(9)

h⁺ + 2H₂O → H₂O₂ + H⁺

(10)

e⁻ scavenging: (O₂)_ads + e⁻ → O₂^•−

(11)

Superoxide protonation: O₂^•− + H⁺ → HO₂^•

(12)

e⁻ co-scavenging: HO₂^• + e⁻ → HO₂⁻

(13)

H₂O₂ formation: e⁻ + O₂^•− + 2H⁺ → H₂O₂

(14)

e⁻ + H₂O₂ → ^•OH + OH⁻

(15)

O₂^•− + H₂O₂ → ^•OH + OH⁻ + O₂

(16)

2HO₂^• → O₂ + H₂O₂

(17)

Photocatalytic oxidation by ^•OH: R-H + ^•OH → R^• + H₂O

(18)

Photocatalytic oxidation by direct h⁺: R-H + h⁺ → ^•R⁺ → degradation by-products

(19)

The XPS analysis of the rutile TiO₂ and Fe/TiO₂-7 is shown in Figure 9. The O 1s peak at 530.1 eV (Figure 9a) belongs to the lattice oxygen [56], and the peak at 532.8 eV is from chemisorbed (O) [56]. The Ti 2p spectra in Figure 9b show two main peaks of Ti2p_3/2 and Ti2p_1/2 at 458.8 and 464.6 eV, respectively, with a signal separation of 5.8 eV, which corresponds to Ti⁴⁺ [57,58]. Compared to the rutile TiO₂, the Ti 2p_3/2 and Ti 2p_1/2 peaks shifted upward by 0.3 eV, which is attributed to lattice distortions following Fe³⁺ doping and the milling process. Figure 9c displays two peaks at 725 and 711.3 eV, assigned to Fe 2p_1/2 and Fe 2p_3/2, respectively. A satellite peak at 719.2 eV indicates the presence of Fe³⁺ [59]. All these peaks provide conclusive evidence for the presence of Fe³⁺ ions in the samples [56,60]. None of the binding energies of the Fe²⁺ (709 and 723 eV) and zero-valent iron (707 and 720 eV) were observed [61,62], which indicates the successful insertion of iron into the TiO₂ lattice.

3.2. Photocatalytic Oxidation of the 2,4-DCP

Figure 10 presents the photocatalytic oxidation of the 2,4-DCP over the rutile TiO₂ and ball-milled Fe/TiO₂ photocatalysts. After equilibrating under dark conditions, the 0.1 g TiO₂, Fe/TiO₂-34, Fe/TiO₂-17, and Fe/TiO₂-7 photocatalysts adsorbed 16.8, 18.8, 26.7, and 19.7%, respectively, of 4.08 mg/L 2,4-DCP (100 mL) and converted 59.4, 64.4, 71.8, and 81.4%, respectively, of 2,4-DCP after 24 h irradiation under visible light. A photolysis experiment exhibited a negligible reduction in the 2,4-DCP concentration (7.5%) after 24 h. The Fe/TiO₂ is a more efficient photocatalyst, as Fe³⁺ ions can transform into Fe²⁺ and Fe⁴⁺ ions trapping e^-/h⁺, as described in Equations (20) and (21) [63]:

Fe³⁺ + h⁺ → Fe⁴⁺

(20)

Fe³⁺ + e⁻ → Fe²⁺

(21)

The Fe²⁺ and Fe⁴⁺, being less stable than Fe³⁺ and their relative energy levels Fe³⁺/Fe⁴⁺ and Fe³⁺/Fe²⁺ (Figure 11), allow the trapping of photo-generated e^-/h⁺ and increase their lifetimes [64]. Prior work [65] has indicated that the energy level of the Fe³⁺/Fe²⁺ is below the TiO₂ conduction band, while that of Fe⁴⁺/Fe³⁺ is above the TiO₂ valence band [66]. Therefore, the transition of e^- from Fe 3d orbitals to CB of TiO₂ induces local states below the conduction band edge. This can significantly decrease the bandgap and increase the removal efficiency [59,66].

Figure 12 presents the photoreaction kinetics, assuming a Langmuir–Hinshelwood (LH)) model as described by Equation (22) [67]:

$$r=-\frac{dC}{dT}=\frac{k_{LH}\times K\times C_t}{1+K\times C_t}$$

(22)

where r, C_t, t, $k_{LH}$, and K refer to the reaction rate (mg/L·h), the concentration of 2,4-DCP at the reaction time t (mg/L), reaction time (h), photocatalytic reaction rate constant (h⁻¹), and equilibrium constant for adsorption (L/mg), respectively. In heterogeneous photocatalytic reactions and at low concentrations (K × C_t << 1), Equation (22) can be simplified to Equation (23) [68]:

$$r=-\frac{dC}{dT}=k_{app}\times C$$

(23)

where $k_{app}$ is the apparent rate constant and can be calculated by plotting the graph of ln(C_t) versus t, as shown in Figure 12.

Table 2. Kinetic parameters of the Fe/TiO₂ samples (C₀ = 0.025 mM, catalyst = 1 g/L).

Samples	Reaction Rate Constant (k, h−1)	Correlation Coefficient (R2)	Removal Efficiency (%)
Rutile TiO2	0.031	0.95	59.4
Fe/TiO2-34	0.032	0.91	64.4
Fe/TiO2-17	0.035	0.96	71.8
Fe/TiO2-7	0.057	0.98	81.4

summarizes the reaction rate constant (k) and the linear correlation coefficients (R²) for the rutile TiO₂ and Fe/TiO₂ photocatalysts.

Among the catalysts, Fe/TiO₂-7 indicated the best performance in 2,4-DCP conversion. Despite using low-power light in the present study, the result is favorably comparable to the photocatalytic yield of Fe/TiO₂ photocatalysts reported in other studies. The 10 wt.% Fe/TiO₂ ball-milled catalyst (B/P = 8, 250 rpm, 5 h) prepared by Carneiro et al. [50] degraded 50.7% rhodamine B (5 mg/L) after 2 h irradiation under a pair of 15 W UV lamps. The 5 wt.% Fe/TiO₂ (B/P = 5, 300 rpm, 5 h) prepared by Hadi et al. [69] converted 37% methylene blue (2 mg/L) after 4 h irradiation under 150 W visible light. Ramírez-Sánchez and Bandala [45] fabricated a hydrothermal sol-gel Fe/TiO₂ catalyst which converted 10.63% estriol (2.9 mg/L) within 8 h of irradiation under a pair of 15 W visible light lamps (without UV cut-off filters). Mancuso et al. [63] prepared 3.5 wt.% Fe/TiO₂ via a soft-templating method for treating 10 mg/L Acid Orange 7. They reported 32% conversion after a 3 h reaction time under 10 W white LED lighting.

Figure 13 plots the TOC with the reaction time for the rutile TiO₂ and Fe/TiO₂-7, showing that the latter has a greater mineralization of 2,4-DCP pollutants (i.e., 63.5 vs. 49.4%). In comparison, the 1 wt.% Fe/TiO₂ prepared by the NaBH₄ reduction method [70] reached 35% TOC removal for a 2,4-DCP solution after 3 h under an intense 100 W tungsten halogen lamp and vigorous aeration. Slow mineralization indicates the presence of refractory by-products that can pose more hazards than the main contaminant [71].

The energy consumed during the photocatalytic reaction is essential for calculating the operating cost of pollution treatment [72]. The G50s for the decomposition of 4.08 mg/L 2,4-DCP in the laboratory photoreactor setup using the rutile TiO₂ and Fe/TiO₂-7 were 0.13 and 0.06 g 2,4-DCP/kW·h. Thus, the Fe/TiO₂-7’s energy yield is approximately twice that of the rutile TiO₂. It costs $0.077 and $0.168 to degrade a gram of 2,4-DCP, according to Hong Kong electricity costs [73].

4. Conclusions

Ball-milling allows a simple and inexpensive method for inserting iron dopants into TiO₂ to produce a series of visible light, photoactive Fe/TiO₂ catalysts that include 5 wt.% iron. It is a green synthesis method that avoids the use of solvents. Applying different B/P ratios (34, 17, 7) and differently sized balls affect the particle morphology and size distribution. The mesoporous Fe/TiO₂-7 prepared under optimum milling conditions (B/p: 7, at 400 rpm and 10 h) displayed the best photocatalytic activity. Fe³⁺ was successfully inserted into the TiO₂ lattice, with no extraneous iron detected by XPS or micro-Raman. A concomitant change in lattice spacing was detected by XRD following the preparation of the Fe/TiO₂-7. It was two times more efficient at 2,4-DCP conversion and mineralization compared to the rutile TiO₂ and the other catalysts. The catalyst’s performance is commensurate to, or better than, similar Fe/TiO₂ photocatalysts and requires less energy to degrade pollutants ($0.077 with respect to Hong Kong electricity costs).