Brookite has an orthorhombic crystalline structure with a unit cell described by the space group Pbca [1]. The structure is composed of octahedra, each with a titanium atom at its center and oxygen atoms at its corners (Figure 2). The octahedra share edges and corners with each other to such an extent as to give the crystal the correct chemical composition. The octahedra are distorted and present the oxygen atoms in two different positions [2]. The bond lengths between the titanium and oxygen atoms are all different.

Oxygen atoms in the octahedron are distributed on the two faces nearly perpendicular to the (100) direction [2]. One face has oxygen atoms of type O1; the other one, oxygen atoms of type O2. Every octahedron shares three edges: one of them determines the crystal distribution along the (100) direction, and the other two determine it along the (001) direction (see Figure 3).

Octahedra arrangement produces a crystalline structure with tunnels along the c-axis, in which small cations like hydrogen or lithium can be incorporated.

X-ray diffraction (XRD) analysis is usually used to prove the presence of brookite in a sample. As shown in Figure 4, the existence of brookite in the XRD patterns is clearly evidenced from the presence of the (121) peak at 2θ = 30.81°. Anyway, for the interpretation of the diffractograms it is necessary to take into account that the main (101) diffraction peak of anatase at 2θ = 25.28° overlaps with the (120) and (111) peaks of brookite at 2θ = 25.34° and 25.69°, respectively, so that apparently pure brookite samples can be a mixture of anatase and brookite.

Ideal brookite has a I ⁽¹²¹⁾ _brookite /I ⁽¹²⁰⁾ _brookite ratio of ~0.9 (JCPDS No. 29-1360) but many of the claimed pure brookite powders reported in literature show I ⁽¹²¹⁾ _brookite /I ⁽¹²⁰⁾ _brookite ratios significantly lower than this value [3]. This does not imply necessarily the presence of anatase since samples synthesized at low temperatures are often scarcely crystalline and the corrected ratio could be easily reached by calcination at relatively high temperatures. On the other hand, an unusual single (121) peak was found for a film containing neither anatase nor rutile obtained by a modified sol–gel method [4]. Hu et al. [5] used one characteristic peak of anatase at 2θ = 62.57°, which does not overlap with any diffraction peak of brookite, to ascertain the existence of anatase.

According to Bokhimi and Pedraza [6] the only way to prove that the samples are single brookite phase is to run their XRD patterns with a good statistic to refine the crystalline structure of brookite in order to demonstrate that the modeling of only this phase is enough to reproduce the experimental diffraction pattern. The Rietveld refinement method has been often applied considering the whole pattern and not only the single peaks [7].

Raman spectroscopy is a very sensitive tool to confirm the existence of only brookite in a sample since brookite is characterized by a relatively complex vibrational spectrum when compared with the other TiO₂ polymorphs [8]. As shown in Figure 5, the overall spectral profile of brookite is characterized by a very strong band at 153 cm⁻¹ and a grouping of weaker, higher wavenumber bands. In comparison, the vibrational spectrum of anatase is relatively simple in accord with its greater crystal symmetry. The absence of the characteristic peak at 516 cm⁻¹ allows for exclusion of the presence of anatase in the samples.

By calcination at high temperatures, brookite is transformed to rutile. The transformation occurs directly or via anatase depending on several factors including crystallite size, size distribution and contact area of the crystallites in the powder. Zhang and Banfield [9] found a correlation among the surface enthalpies of the three polymorphs and their particle size. The energies of anatase, brookite and rutile are sufficiently close that they can be reversed by small differences in surface energies. Brookite is more stable than anatase for crystal sizes greater than 11 nm while rutile is the most stable phase at sizes greater than 35 nm. Zhu et al. [10] developed an empirical expression on a critical grain size of brookite, D_c, which dominates the transition sequence between anatase and brookite. When the size of brookite D_b is larger than D_c, brookite directly transforms to rutile whereas, when D_b is smaller than D_c, brookite transforms to anatase and then anatase to rutile.

The knowledge of the electronic band structure of the TiO₂ polymorphs is useful to understand the photocatalytic behaviour of the pure phases and of their mixtures. Theoretical and experimental works report band gap values for brookite both smaller and larger than that of anatase. In 1985, Grätzel and Rotzinger [11] were the first to estimate the band gap value, E_g, of brookite as 3.14 eV by extended Hückel molecular orbital calculations. The E_g, was intermediate between those of anatase (3.23 eV) and rutile (3.02 eV). Mo and Ching [12] used the self-consistent orthogonalized linear combination of atomic orbitals method to study the electronic structure and the optical properties of anatase, brookite and rutile. According to their calculations, brookite had a direct band gap of 2.20 eV, which is larger than that of anatase and rutile (1.78 eV for rutile and 2.04 eV for anatase, respectively). These results, as reported by the same authors, were an underestimation of the effective band gap values. Park et al. [13] calculated an E_g value of 2.1 eV for brookite using the Kohn-Sham method, which underestimates systematically the band gap often by more than 50%. Recently, standard density functional theory calculations [14] showed that brookite and rutile had a direct band gaps of 1.86 eV and 1.88 eV, respectively, whereas an indirect band gap of 1.94 eV was observed for anatase. Using a more modern approximation, the band gap underestimation was corrected and the E_g values of brookite, rutile and anatase were increased to 3.30 eV, 3.39 eV and 3.60 eV [14].

Experimental band gap energies ranging from 3.1 to 3.4 eV have been reported for brookite, but there is disagreement on whether the optical response is attributable to direct or indirect transitions. The band gap was usually determined by diffuse reflectance measurements, from the tangent lines to the plots of the modified Kubelka-Munk function, [F(R’_∞)hν]^1/2, versus the energy of the exciting light [15] considering brookite as an indirect semiconductor [16,17,18,19,20,21,22,23,24]. Direct band gap values have been also obtained [3,5,16,17,25].

The precise value of E_g is currently unknown since the experimental results often refer to samples not well crystallized. Zallen and Moret [26] reported that natural crystals of brookite had an indirect band gap of about 1.9 eV. This low value is ascribable to the presence of impurities in the mineral that was pale brown in color. On the other hand, different band gap values were obtained by varying the duration of the hydrothermal treatment [18] or the kind of nanostructure (nanosheets, nanospindles, nanoflowers) [27]. Koelsch et al. [21] determined the band gap of brookite nanoparticles as 3.4 eV, measuring the optical transmission of brookite dispersions at different concentrations and by UV fluorescence measurements. Shibata et al. [28] photoelectrochemically determined a band gap value of 3.26 eV for films consisting mostly of brookite phase with some anatase. A value of 3.29 eV was obtained by crystallization at 450 °C of a sample of brookite prepared by thermohydrolysis of TiCl₄ at 100 °C [24]. This value compares well with the value of 3.31 eV found for brookite nanorods [23].

The flatband potential, (E_FB), of a semiconductor is a fundamental property for the thermodynamics of the interfacial electron transfer steps. For an n-type semiconductor as TiO₂ it can be assumed that the positions of the flatband potential and the quasi-Fermi level (*E_f) are the same and very close to the lower edge of the conduction band [29].

Grätzel and Rotzinger [11] firstly reported a rough estimate of E_FB = −0.03 V vs. NHE for brookite based on a theoretical band gap value of 3.14 eV. Di Paola et al. [30] determined the flat band potentials of anatase, brookite, and rutile by a slurry method [31] measuring the photovoltage of the corresponding suspensions in the presence of methylviologen dichloride as a function of pH. As shown in Figure 6, the values obtained at pH 7 were −0.45 V for anatase, −0.46 V for brookite, and −0.37 V for rutile. It is worth noting that the experimental E_FB value found for brookite is practically coincident with that calculated theoretically (−0.44 V at pH 7 [11]).

Kandiel et al. [23] calculated the flat-band potentials of anatase and brookite by impedance spectroscopy using electrodes obtained by spreading the corresponding suspensions on conductive fluorine-tin oxide glasses. The quasi-Fermi levels (*E_f) were, furthermore, obtained under UV illumination by the slurry method. The E_FB values estimated at pH 7 versus NHE were −0.35 V for anatase and –0.54 V for brookite whereas the corresponding *E_f values were −0.68 V and −0.77 V. The differences between these values were attributed to the independent methods performed under dark and illuminated conditions. The conduction band edges of anatase and brookite were calculated as the mean value of E_FB and *E_f, i.e., −0.51 V for anatase and −0.65 V for brookite.

Recently, Truong et al. [32] reported values of −0.91 V and −0.96 V vs. NHE at pH 7 for the conduction band edges of anatase and brookite, respectively, without citing the methods used to determine them.

The few experimental values of flat band potential of brookite reported in literature and the uncertainty on its band gap do not allow to exactly locate the position of the conduction and valence band edges of the three oxides of titanium. Anyway, the most probable hypothesis is that the flat band of brookite is shifted negatively with respect to that of anatase.

The first works on the synthesis of brookite date back to the late 1950s. Classically, brookite was synthesized from titanium(IV) compounds in aqueous or organic media by hydrothermal methods at high temperatures and pressures. Mixtures of anatase and brookite were prepared by thermal treatment of the amorphous TiO₂ powders obtained by hydrolysis of titanium(IV) tetraethoxide [33,34,35]. Yamaguchi found that a film of brookite was obtained on the surface of a Ti foil by anodization in a solution of H₂SO₄ (50% by weight) [36]. Complete transformation of anatase to brookite was achieved by grinding for 30 h samples previously heated for several hours to 600 °C [37] or anatase-brookite mixtures obtained by hydrolysis of titanium alcolate [38].

Keesmann [39] was the first to synthesize pure brookite by hydrothermal treatment of amorphous TiO₂ obtained by hydrolysis of titanium tetraisopropoxide or alkaline titanates, in the presence of a solution of NaOH containing 3–25 atoms % of Na with respect to (Ti + Na). Na⁺ ions were considered to stabilize the lattice of brookite but Schwarzmann and Ognibeni [40] obtained well crystallized brookite by hydrothermal treatment of TiO₂·xH₂O between 100 and 300 °C in the absence of Na⁺ ions.

Basic environments are not essential for the formation of brookite and this phase was also prepared by oxidation of an aqueous acidic TiCl₃ solution containing CH₃COONa under a flux of air at 95 °C [41,42]. Moreover, brookite was synthesized by hydrothermal oxidation of Ti metal in a NaF solution at 200–500 °C [43]. It was also formed from systems containing TiH₂ or TiO under similar conditions [44]. The nucleation of brookite was related to the brookite-like structure stabilized by NaF.

These first reports indicate that the formation of pure brookite strongly depends on many factors, which are difficult to predict. For example, Mitsuhashi and Watanabe [45] synthesized brookite from fresh coprecipitates obtained by neutralization with NaOH of an aqueous solution of TiCl₄ and CaCl₂ at 220–560 °C for 0.5–220 h. Their conclusion was that also the Ca²⁺ ions, with an ionic radium similar to that of the Na⁺ ions, have a stabilization effect for the formation of brookite. Arnal et al. [46] observed the singular formation of brookite by the reaction of tert-butyl alcohol with TiCl₄ at 110 °C. The crystallization of brookite was related to the formation of a unique hydroxyl group during the alcoholysis of TiCl₄.

Brookite was hydrothermically synthesized at 110–200 °C within 24 h using amorphous TiO₂ and sodium salts as starting materials [47]. Almost single phase brookite was obtained at 200 °C and a TiO₂/NaOH molar ratio ~1. It was suggested that in alkaline solution the crystallization of brookite from sodium titanate occurred with release of Na⁺ and H⁺ ions accompanied by the oxidation of Ti and Ti-OH bond that changed to Ti-O-Ti bond in the structure.

To avoid any contamination with the other polymorphic TiO₂ phases, Kominami et al. [48] synthesized microcrystalline brookite by solvothermal treatment of oxobis(2,4-pentanedionato-O,O')titanium (TiO(acac)₂) in ethylene glycol, in the presence of sodium laurate and a small amount of water at 300 °C. As confirmed by Raman spectroscopy, pure brookite was obtained when Na/Ti = 2. Ethylene glycol showed a specific solvent effect for the crystallization of brookite.

Zheng et al. [49,50] synthesized brookite nanocystallites by using Ti(SO₄)₂ or TiCl₄ as the precursors. The formation of brookite was obtained under hydrothermal treatments by controlling the pH values and the reaction temperature. Different conditions were used depending on the titanium precursor. High basicity favoured the formation of brookite in the presence of TiCl₄ solutions.

In the last decade, an always increasing number of papers have concerned the preparation of powders and films of brookite. Inorganic and organic precursors such as TiCl₄ [21,24,49,50,51,52,53,54,55,56,57,58], Ti(SO₄)₂ [18,59,60,61,62], TiOSO₄ [63,64], TiCl₃ [3,19,65,66,67,68], TiOCl₂ [16,69,70], titanate nanotubes [69,71,72,73], titanium tetraisopropoxide [25,74,75,76], titanium butoxide [77,78,79,80,81] and titanium tetraethoxide [82] have been employed.

Pottier et al. [51] synthesized nanoparticles of brookite and rutile by thermolysis of TiCl₄ in concentrated HCl solutions at 100 °C. The Ti:Cl molar ratio was the key factor to determine the crystalline phases and their relative proportions. Pure brookite was easily separated by the mixtures of brookite and rutile by drying the stable brookite sols obtained by peptization of the solid phase with a HNO₃ solution [51] or by repeated washings with water [24]. Similarly, platelets of brookite were obtained by selective peptization of the mixtures of brookite and rutile resulting from the thermolysis of TiCl₄ (0.15 mol L⁻¹) in a HNO₃ solutions at 95 °C for 7 days [56]. Lee and Yang [55] synthesized crystalline TiO₂ nanoparticles with pure brookite structure by hydrolysis of TiCl₄ in HNO₃ solutions at 70 °C within 40 h. The aqueous TiCl₄ solutions contained not more than 0.8 mol L⁻¹ of Ti⁴⁺ and the concentration of HNO₃ was kept in the range of about 3.9–4.3 mol L⁻¹. Nanocrystalline particles of pure brookite were prepared by an ambient condition sol process hydrolyzing TiCl₄ in an acidic alcohol solution under refluxing [53,54]. The refluxing time and temperature had a fundamental role in controlling the particle size and the phase of the particles obtained [54]. These results highlight the strong influence of the precipitation conditions (acidity, nature of the anions, titanium concentration, etc.) on the formation of brookite.

Yang et al. [59] prepared pure brookite by using Ti(SO₄)₂ as the precursor. The phase formation was achieved by hydrothermal treatment at 200 °C for 24 h. Ti(SO₄)₂ was hydrolyzed using 2 M NaOH and a NaOH/ Ti(SO₄)₂ molar ratio of 4:1. Xie et al. [18] observed that the crystallinity of brookite increased with increasing the hydrolysis time. Ultrafine brookite powders were prepared by Luo et al. [62] by hydrolyzing Ti(SO₄)₂ in alkali at 199 °C under high pressure.

Pure brookite was synthesized by thermolysis of an aqueous solution of TiOSO₄ at 90 °C for 4 h [63,64]. In the presence of lithium oxalate with a molar ratio [C₂O₄²⁻]/[Ti⁴⁺] = 1, a titanium oxalate-based compound Ti₂O₃(H₂O)₂(C₂O₄)·3H₂O formed. Brookite was obtained by thermal decomposition of the oxalate hydrate at a temperature as low as 300 °C [63]. Samples with high specific surface area lying between 150 and 400 m² g⁻¹ were obtained by thermal decomposition of the oxalate hydrate at a temperature between 300 and 400 °C for 4 h under an air atmosphere [64]. As shown in Figure 7, the eggshell morphology of the Ti₂O₃(H₂O)₂(C₂O₄)·3H₂O compound was retained despite the creation of porosities arising from the removal of water and oxalate species.

Many years after the first pioneering work of Kiyama et al. [41], a few reports have concerned the synthesis of brookite by thermolysis of TiCl₃ in aqueous media. Li et al. [65,83], prepared phase-pure brookite of high crystallinity under ambient pressure at 70°C via reacting a mixed solution of urea and titanium(III) chloride. The resultant particles were monodispersed spheres (~154 nm) composed of brookite nanocrystals (~25 nm) that were stable in terms of phase purity and morphology up to ~ 500 °C. Particles with lower size distribution were obtained by adding polyethylene glycol to the solution of TiCl₃ and urea [66]. Štengl et al. [67] prepared transparent particles of brookite by hydrolysis of an aqueous solution of TiCl₃ in the presence of a solution of polyethylene glycol in ethanol at 70 °C. Selected area electron diffraction (SAED) analysis showed that the brookite particles grew in the presence of polyethylene glycol with a molar mass higher than 10,000. The crystallinity of the samples increased with the time of ageing in the aqueous solutions up to 60 days. Recently, pure brookite was obtained by hydrothermal treatment of TiCl₃ with NH₄OH at 200 °C for 20 h in the presence of 4.2 M NaCl [68].

The mechanism of the direct formation of titanium(IV) oxide from a titanium(III) solution is still unclear. Li et al. [3] selectively synthesized brookite nanocrystals via a redox route under mild hydrothermal conditions (180 °C, 3 h), employing TiCl₃ as the titanium source and H₂O₂ as the oxidant. Pure brookite was attained by carefully controlling the reactant concentration (0.0625–0.075 M) and the solution pH (1.1–1.32). Spindle-like particles of pure brookite were obtained at pH = 10 by using NaNO₃ as the oxidizing agent [19].

Small brookite particles were separated by centrifugation from a mixture of brookite and rutile obtained by self-hydrolysis of a solution of TiOCl₂ with HCl in autoclave for 24 h [69]. Liu et al. [70] synthesized nanocristalline brookite by thermohydrolysis of a titanium oxychloride hydrate obtained by slow hydrolysis of a commercial TiOCl₂ solution in a solution containing tetramethylammonium hydroxide (TMAOH) and H₂O. Brookite was obtained for 0.47 ≤ Ti/TMAOH ≤ 0.6.

Nanocrystals with brookite phase were prepared by Murakami et al. [73] by hydrothermal treatment of the solution obtained by dispersion of a commercial powder of titanate nanotubes in HClO₄. Single-phase brookite was extracted from the suspension by centrifugal separation.

Pure brookite nanoparticles were synthesized by hydrothermal treatment of amorphous titania prepared by reaction of titanium(IV) isopropoxide with water in ethanol [74]. The amorphous titania was successively treated with NaOH and water at 200 °C for 48 h. Titanium isopropoxide vapour, argon and oxygen mixtures were used to obtain pure brookite crystals with large surface area by plasma enhanced chemical vapour deposition (PECVD) at room temperature, under an applied d.c. bias voltage of −250 V [75]. Complex porous particles of brookite were prepared using hydrated pollen grains exposed to titanium isopropoxide solutions, followed by calcination at 500 °C for 3h [76].

Recently, pure brookite has been successfully synthesized by hydrothermal treatment of water-soluble titanium complexes [84] such as titanium-peroxo-glycolate [16,85,86,87,88] and titanium bis(ammonium lactate) dihydroxide [23,89].

A selective synthesis of brookite was realized by Tomita et al. [85] by using an aqueous solution containing the titanium-peroxo-glycolate complex (NH₄)₆[Ti₄(C₂H₂O₃)₄(C₂H₃O₃)₂(O₂)₄O₂]·4H₂O prepared from Ti powder dissolved in an ice-cooled solution of 30% H₂O₂, 28% NH₃ and glycolic acid. Single-phase brookite nanorods were obtained at 200 °C for 24 h, by adjusting the pH of the solution to 10 by addition of NH₃ or ethylenediamine [86]. The results demonstrated that a weakly basic environment was essential for the formation of brookite while ammonium ions did not play an important role. Pure brookite was also obtained by Štengl and Králová [16] by hydrothermal treatment of a solution of titanium peroxo-glycolate complex at pH 9.5 and 220 °C for 24 h. The formation of brookite was attributed to the structure of the complex anion that contains four titanium atoms coordinated by glycolic acid and peroxo groups in a dimer-like structure that is connected to another dimer through bridging oxo groups. As shown in Figure 8, this structure closely resembles the architecture of brookite [84].

Similarly, brookite particles with typical nanorod shapes were synthesized at pH 10 by hydrothermal treatment at 200 °C for 24 h of the titanium picolinato complex obtained by adding picolinic acid rather than glycolic acid to the peroxo-titanic acid solution [90].

A microwave-assisted hydrothermal treatment of the titanium-peroxo-glycolate complex at 200 °C for 0–60 min allowed to synthesize brookite particles with a size smaller than that obtained by the conventional hydrothermal method [87]. Lately, nanocrystalline single phase brookite was also prepared by decomposition at 200 °C for 4 h of a water-soluble titanium complex with ethylenediaminetetraacetic acid (EDTA) [91]. When the titanium concentration in the starting solution was higher than 0.25 M, brookite could be prepared in a wide range of pH including both acidic and basic media.

Bahnemann et al. [23,89], synthesized brookite nanorods by a simple hydrothermal method at 160 °C for 24 h using aqueous solutions of commercially available titanium bis(ammonium lactate) dihydroxide (TALH) in the presence of urea as an in situ OH⁻ source. High quality brookite was obtained at high concentrations of urea (≥ 6.0 M).

A drawback of the environmentally benign syntheses of single-phase brookite using water-soluble titanium compounds and water as solvent is the possibility to obtain anatase, rutile or mixed TiO₂ phases if the experimental conditions, e.g., water to precursor ratio, solution pH, reaction temperature, are not strictly controlled.

Brookite was also prepared from the lepidocrocite-type K_0.8Ti_1.73Li_0.27O₄ by alkali metal ion extraction with polytetrafluoroethylene (PTFE) and simultaneous heat treatment at 400 °C in flowing Ar [92]. Fully oxidized brookite formation and complete decomposition of the PTFE derivatives were achieved by further heating in flowing air, and coproduced alkali-metal salts were removed by washing with water.

Additionally, brookite was obtained by milling anatase-phase powders. Wakamatsu et al. [93] synthesized almost single phase brookite after 54 ks and 72 ks ball milling of anatase under the condition of medium intensity.

It is worth noting that the final products of several syntheses reported in the literature are not really single phase brookite since they contain traces of anatase or rutile, as clearly evidenced by the XRD patterns. Reyes-Coronado et al. [17] prepared brookite nanoparticles by hydrothermal treatment of amorphous titania obtained by titanium(IV) isopropoxide, water and 2-propanol. The white paste of amorphous titania was diluted with 3 M HCl and heated at 175 °C for 7 h. Under these conditions the phase purity of brookite was over 95% due to the presence of a small amount of rutile.

Likewise, TiO₂ nanoparticles with brookite as the majority phase (> 95%) were synthesized by co-thermohydrolysis at 60 °C of an equimolar solution of TiCl₃ and TiCl₄ [94]. The pH was adjusted to 4.5 with a NaOH solution and the obtained suspension was aged for one week at 60 °C without stirring. The particles were flocculated by increasing the pH up to 6 with NaOH.

Nanomaterials with special structures and morphologies, such as nanotubes, nanorods, nanospheres or nanoflowers have attracted growing interest because of their unique structures and properties. A summary of methods used to prepare brookite nanomaterials is reported in Table 1.

Nanometric particles of pure brookite TiO₂ were synthesized by modified thermolysis of reactant solutions containing titania powder, HCl, urea and PEG 10000 [66]. Unique flower-like brookite agglomerates with an average diameter of ~400–450 nm composed of single brookite nanocrystals of ~ 4–5 nm were obtained at 105 °C.

Deng et al. synthesized pure brookite nanotubes with a ~5 nm inner diameter and a ~0.78 nm interlayer space [71] or nanorods with a bipyramidal shape [72] using as precursors titanate nanotubes or nanosheets prepared by dispersing anatase TiO₂ powder in a 10 M NaOH solution at 120 °C or 90 °C, respectively. The corresponding brookite nanostructures were obtained by hydrothermal treatment at 200 °C.

Buonsanti et al. [57] developed a surfactant-assisted non aqueous strategy to prepare anisotropic brookite nanocrystals with tunable geometric features. The syntheses relayed on the high temperature aminolysis of titanium carboxylate complexes obtained by heating mixtures of TiCl₄, oleic acid and oleyl amine in 1-octadecene at 290 °C for 30 min under air-free conditions.

Well-facetted brookite crystals with the growth direction along the (001) axis were sinthesized by a low-temperature process using [Ti₈O₁₂(H₂O)₂₄]Cl₈·HCl 7H₂O prepared by hydrolysis of a commercial TiOCl₂ solution and tetramethylammonium hydroxide with a molar ratio R = Ti/TMAOH = 0.6 [70].

Lin et al. [27] synthesized brookite nanosheets with a regularly truncated tetragonal shape through a low-basicity hydrothermal process utilizing TiCl₄ as titanium source, urea as in situ OH⁻ source, and sodium lactate as complexant and surfactant. After 12 h of reaction at 200 °C, the [Ti(C₃H₄O₃)₃]²⁻ complex transformed into single-crystalline nanosheets with specific facets exposed.

High quality brookite flowers were obtained by Hu et al. [5] by hydrothermal treatment at 220 °C for 48 h of a suspension containing TiOSO₄ and NaOH at pH 12.5. Time-resolved experiments revealed that layered titanate transformed into spindle-like brookite nanoparticles that aggregated together to yield the flower morphology. The presence of Na⁺ ions was a key factor for the formation of brookite since the replacement of NaOH with LiOH or KOH favoured the formation of mixtures of anatase and rutile.

Brookite nanoflowers consisting of single crystalline nanorods were prepared by hydrothermal treatment of a solution containing titanium butoxide, NaCl and aqueous ammonia [78,79,80]. Pure brookite was obtained with a NaCl concentration of about 0.25 M at 180 °C and 0.50–0.75 M at 200 °C [79]. At the initial stage of the reaction, layered titanate formed by condensation of TiO₆ octahedra. After several hours of heating, the layered structure collapsed resulting in the formation of brookite.

Nguyen-Phan et al. [25] prepared hierarchical brookite with a humming-top-like morphology by heating a solution containing 0.1 M NaOH, titanium isopropoxide and 2-propanol at 200 °C for 72 h. The gradual transformation from layered titanate to brookite phase was well consistent with the hypothesized formation mechanism of the hierarchical superstructure.

Macroporous spherical brookite particles with sub-micrometer diameters were prepared by spray drying and evaporation of an aqueous suspension obtained by mixing a commercial suspension of brookite nanoparticles and spherical monodispersed polystyrene latex particles, which were used for colloidal templating [96]. Katagiri et al. [97] prepared hollow brookite spheres by alternate deposition of titanium-peroxo-glycolate complex and a positively charged polyelectrolyte on spherical polystyrene templates, by utilizing the electrostatic interactions for the shell formation. The hollow capsules obtained by removing the polystyrene cores by tetrahydrofuran were hydrothermally treated at pH 10 and 200 °C for 24 h to induce the crystallization of brookite.

Pseudo-cube shaped brookite nanocrystals with the four corners truncated were synthesized by hydrothermal treatment of a solution of titanium-peroxo-glycolate complex in the presence of sodium oleate at 200 °C for 6 h [95]. The particular morphology of the particles (Figure 9) was attained by selective absorption of oleate molecules onto specific crystal faces of brookite that reduced their specific surface energies and induced the formation of the cubic shape.

Brookite is very difficult to prepare in thin-film form since most of the usually employed techniques lead to amorphous or crystalline anatase. Takahashi et al. [98] were the first to prepare almost pure brookite films by vapor phase decomposition of tert-butyl titanate. Evidence for the formation of a brookite layer as an intermediate of the anatase/rutile transition was also observed by in situ Raman spectroscopybafter controlled oxidation of a Ti electrode in KOH electrolyte [99,100].

Moret et al. [101] prepared thin films of TiO₂ by pulsed laser deposition at 750 °C on silicon and other substrates. The laser-ablation target was a pellet prepared from a 99.95% TiO₂ powder. Brookite was the main phase present in the films, along with anatase and a small amount of rutile. Djaoued et al. [102] obtained brookite-rich films by using titanium tetraisopropoxide, diethanolamine and polyethylene glycol. The brookite phase was found preponderant only in films heated slowly to 600 °C. The weight fraction of brookite in the crystalline phase was around 70%.

Shibata et al. [28] deposited TiO₂ films on SiO₂-coated glass plates using a commercial brookite type sol, followed by calcination at 500 °C for 30 min. The films contained anatase and the content of brookite was about 75%. The same commercial sol was employed by Jiang et al. [103] to prepare mesoporous films that were tested as electrodes for dye-sensitized solar cells.

63.4% of brookite and 36.6% of anatase were present in a TiO₂ thin films deposited on a soda-glass substrate using a sol suspension obtained by hydrolysis of titanium tetraisopropoxide [104]. Only pure anatase was observed if the substrate was silicon. The formation of the brookite phase was attributed to the presence of Na dissolved from the soda-glass substrate during the calcination at 500 °C for 2 h.

Mild oxidation of Ti foils in air at 500 °C for 6 h resulted in brookite-rich films characterized by a strong absorption band in the visible spectral range [105].

Kuznetsova et al. [4] obtained pure brookite thin films by using TiCl₄, cellulose, ethylene glycol and oxalic acid as complexing agent. The XRD results revealed an unusual (121) orientation of the film due probably to a suitable selection of the chemical processing parameters. The presence of 5% of Na⁺ ions detected by XPS analysis confirms that the thermo-diffusion process of Na from the glass substrate induces the stabilization of brookite [106]. Single (121) oriented pure brookite films were also deposited on glass substrates using a solution of titanium butoxide and acetic acid [81]. The particle size was controlled by the water/acetic acid ratio.

Transparent TiO₂ thin films were deposited on soda lime glass at different substrate temperatures by a spray pyrolysis technique from an aqueous peroxo-polytitanic acid solution [107]. A low polycrystalline brookite structure was obtained when the amorphous as-deposited films were annealed in air at 500 °C for 3 h.

Novotna et al. [108,109] prepared pure brookite thin films by a very simple sol–gel method. Soda lime glasses were dipped into a solution of titanium isopropoxide, HCl, 2-propanol, and acetylacetone at 22 °C. Finally, the coated substrates were calcined at 500 °C for 2 h. The formation of the brookite phase was caused by the diffusion of sodium ions from the soda lime substrate.

An alternative possibility for the formation of brookite films is the synthesis of brookite particles and their following deposition on a substrate. Thin films were so obtained by coating stable sols of pure brookite separated from mixtures of brookite and rutile by peptization with HNO₃ [21,58], or by repeated washings with water [110,111,112]. Magne et al. [22,113] prepared pure brookite porous films using a viscous paste created by mixing ethyl cellulose powder dissolved in ethanol and brookite particles obtained from mixtures of brookite and rutile or by cohydrolysis of TiCl₃ and TiCl₄.

Kim et al. [114] prepared brookite colloidal sols for thin film coating by hydrothermal treatment at 120 °C for 10 h of the solution obtained dissolving Ti hydroxide in H₂O₂ in the presence of 5 M NaCl. Recently, nanostructured brookite films were deposited by the matrix-assisted pulsed laser evaporation technique [115,116]. The target was a frozen toluene suspension of brookite nanorods covered with an oleate/oleyl ammine capping layer synthesized by surfactant-assisted aminolysis of titanium oleate complexes at 280 °C under air-free conditions [57].

The difficulty in preparing brookite having both high purity and large surface area is probably one of the reasons for the limited works on its photocatalytic properties. Anyway, density functional theory calculations have shown that the commonly exposed brookite (210) surface is more reactive than the ubiquitous anatase (101) surface and might be useful in catalytic and photocatalytic applications [117].

The first study on the photocatalytic behaviour of brookite appeared many years after the publication of the pioneering works on the preparation of this TiO₂ polymorph [39,40,41,43,44,45]. In 1985, Ohtani et al. [42] found that extra-fine crystallites of brookite, prepared by air oxidation of TiCl₃in aqueous HCl solution, possessed good photocatalytic capacities for both the reactions of H₂ formation from 2-propanol and O₂ evolution from an Ag₂SO₄ solution. In 2003, Kominami et al. [118,119] studied the correlation between physical properties and photocatalytic activities of nanocrystalline brookite samples prepared by a solvothermal method. The powders were calcined at various temperatures and then tested for the photocatalytic mineralization of CH₃COOH in aqueous solutions and dehydrogenation of 2-propanol. Some samples exhibited photoactivities comparable to that of the representative active commercial photocatalyst, Degussa P25.

Nanocrystalline brookite particles obtained by reaction of TiCl3and urea at 70 °C revealed a good efficiency for the photodegradation of acetaldehyde [65] and 4-chlorophenol [66] but the activity of these samples was rather lower than that of Degussa P25.

Brookite was found efficient for the degradation of rhodamine B [18] but less active than P25 [19]. Macroporous brookite particles prepared by a spray drying process with a precursor solution of brookite nanoparticles and polystyrene latex particles were more active for the degradation of rhodamine B than dense particles as a result of their increased surface area [96]. The macroporous particles retained their structure in the liquid phase and were easily collected and reused.

Kobayashi et al. [88] prepared brookite nanoparticles by hydrothermal treatment of a titanate-glycolate complex. The samples were more efficient than Degussa P25 for the photooxidation of NO. Under illumination by only visible light with λ > 510 nm the photoactivity was four times higher than that of P25 due probably to the presence of doping nitrogen atoms. The photocatalytic activity of samples annealed at different temperatures was examined by using the degradation of Orange II dye [16]. With increasing temperature, the crystallinity increased and the photoactivity was higher. An enhancement of activity was achieved by employing brookite particles with a smaller size obtained by microwave heating of the titanate-glycolate complex [87]. Further studies evidenced that brookite synthesized from a water soluble Ti-EDTA complex had a considerably higher activity for the decomposition of NO than that of brookite prepared from the titanate-glycolate complex [91].

Transparent particles of brookite prepared by hydrolysis of an aqueous solution of TiCl₃ in the presence of polyethyleneglycol degraded Orange II more efficiently than P25 [67]. Li et al. [3] selectively synthesized anatase, rutile and brookite nanocrystals via a redox route under mild hydrothermal conditions employing TiCl₃ as titanium source. The brookite nanoplates gave the best performance for the bleaching of methyl orange under UV irradiation.

Brookite nanoparticles prepared by hydrothermal treatment of a solution containing tetrabutyl titanate and triethanolamine at pH 12 were more active than anatase nanorods prepared at pH 11 and P25 [77]. The higher activity was attributed to the larger surface of the brookite nanoparticles.

Di Paola et al. [24] synthesized mixtures of brookite and rutile by thermohydrolysis of TiCl₄ in HCl solutions. The brookite particles separated by selective peptization revealed good catalytic properties for the photodegradation of 4-nitrophenol. The powder calcined at 450 °C exhibited a higher reactivity, mainly because of its increased crystallinity.

Similarly, the photocatalytic activity of small brookite nanoparticles separated by centrifugation from a mixture of brookite and rutile prepared by self-hydrolysis of TiOCl₂ at 100 °C was enhanced by a further hydrothermal treatment at 100 or 200 °C for 24 h [69].

Pure brookite nanorods synthesized by thermal hydrolysis of titanium bis(ammonium lactate) dihydroxide were less active than anatase for the reaction of photodegradation of dichloroacetic acid due probably to its lower surface area [23,89]. At variance, brookite was more efficient than anatase and P25 for the production of H₂ from an aqueous methanol solution. This behaviour was attributed to the position of the flatband potential of brookite that is more cathodic than that of anatase so that it has a higher driving force for the proton reduction.

Nanomorphologies generally show better photocatalytic properties because of their larger specific areas that provide more active sites. Brookite nanocrystals prepared by hydrothermal treatment of titanate nanotubes exhibited photocatalytic activity for the degradation of acetaldehyde similar to that of P25 [73]. Lin et al. [27] prepared high-quality brookite nanosheets surrounded with four (210), two (101), and two (201) facets. The photoactivity of these nanosheets for the degradation of methyl orange was superior to that of P25 in terms of unit specific surface area. Other morphologies such as irregularly faceted nanoflowers and nanospindles were inactive indicating that the photocatalytic activity of brookite is highly morphology-dependent.

Brookite nanoflowers degraded methyl orange much more efficiently than anatase nanorods with similar specific area [78]. The photocatalytic activity of the nanoflowers was high also for the degradation of phenol and salicylic acid [79].

Zhang et al. [82] modified the surface morphology of brookite nanorods by chemical etching with a solution containing NH₃ and H₂O₂ or with concentrated H₂SO₄. Before etching, the nanorods were more active than P25 for the decomposition of toluene. After etching with the NH₃-H₂O₂ solution the photocatalytic activity of the samples increased, whereas the treatment with H₂SO₄ caused an appreciable reduction of activity. The enhancement of photoactivity was ascribed to the formation of highly active faces working as oxidation sites.

Addamo et al. [110] prepared thin brookite films that exhibited a high photocatalytic activity for 2-propanol and a promising long term stability after repeated usage. Low crystallized brookite films deposited by spray pyrolysis from a peroxo-polytitanic acid were active for the photodegradation of methylene blue [106].

Mattson and Österlund [20] presented a comparative study of the adsorption and photoinduced degradation of propanone and acetic acid on thin films of anatase, brookite, and rutile. The quantum yield for the photodegradation of propanone was larger for brookite than for anatase and much larger than for rutile. In contrast, the quantum yield for the degradation of acetate was lower for brookite than for anatase.

Anatase films deposited on soda lime glass precoated with a SiO₂ layer exhibited better activity for the photodecomposition of acid orange 7 and 4-chlorophenol than brookite films obtained on soda lime glass [108]. Similar results were reported by Krýsa et al. [109]. The low photocatalytic activity was attributed not to the presence of brookite but to the high content of Na⁺ ions diffused in the film from the glass substrate.

Anatase and brookite-rich films tested for the photocatalytic oxidation of methylene blue and cis-9-octadecenoic acid exhibited almost the same activity [28]. Anyway, the brookite-rich film was more hydrophilic than the anatase film as revealed by the lower contact angle under weak UV light irradiation. Zhou et al. [120] prepared anatase/brookite films on modified poly(ethylene terephthalate) substrates that were less active than anatase films for the photodegradation of rhodamine B but showed an improved hydrophilicity when brookite was present in the films.

Superhydrophilic properties, even in dark conditions, were exhibited by transparent brookite thin films prepared by using titanium tetraisopropoxide as the precursor [108]. Similar results were obtained by Bellardita et al. [121] who studied the photocatalytic activity and the hydrophilicity of anatase, brookite and rutile films deposited by sols obtained by thermolysis of TiCl₄ in aqueous or diluted HCl solutions. As shown in Figure 10, the anatase and rutile films were less hydrophilic than the brookite film that exhibited a contact angle of 10° in the dark.

In addition, N-doped brookite films showed a photoinduced hydrophilic ability highest than that of N-doped anatase and rutile films [122].

Brookite has been successfully tested for the photocatalytic production of hydrogen. Ohtani et al. [42] firstly reported that pure brookite powder, modified by Pt nanoparticles deposition, had a marked photocatalytic activity for the dehydrogenation of 2-propanol. A brookite sample with improved crystallinity and sufficient surface area obtained by calcination at 873 K exhibited the hydrogen evolution rate almost equal to that of P25 [118,119]. Chiarello et al. [123] proved that Pt-deposited brookite is a very good photocatalyst for obtaining hydrogen from methanol steam reforming under UV-vis irradiation, also in consideration of its high selectivity to CO₂ formation and low CO production. Kandiel et al. [23] reported that the photocatalytic hydrogen evolution of anatase/brookite mixtures and pure brookite was higher than that of pure anatase. Anatase-brookite-rutile mixtures with brookite as the major phase were superior to P25 and anatase in generating hydrogen from an aqueous EDTA solution under visible light [124].

Anji Reddy et al. [125] demonstrated the facile insertion of lithium into nanocrystalline brookite due to the presence of atom-free channels along the (001) direction that enable the lithium mobility into the structure (see Figure 2). Electrochemical and ex situ XRD studies showed that brookite is stable for lithium intercalation and deintercalation. The high reversible capacity and low polarization make brookite an attractive alternative negative electrode material for Li-ion batteries. Further studies revealed that the reversibility of Li insertion decreased with an increase in crystallite size of brookite [126]. Hybrid multi-walled carbon nanotube/brookite electrodes improved the specific capacity and capacity retention on cycling [127,128]. Dambournet et al. [63] prepared micrometer-size mesoporous brookite electrodes with a high specific area that provided higher volumetric energy density in a lithium ion battery than anatase or rutile electrodes.

Brookite-rich thin films obtained by oxidation of Ti foils in air in a temperature range from 500 to 700 °C were tested as electrodes for water splitting, under visible light [129]. The photocurrent densities were higher than those observed with anatase or rutile produced by oxidation of Ti foils. Mesoporous brookite films containing a certain amount of anatase were used as electrodes for dye-sensitized solar cells [103]. Magne et al. [22] prepared efficient cells employing porous layers of pure brookite. The comparison between anatase and brookite based-cell revealed that the latter phase is very interesting for this kind of applications [22].

Photocatalysis by TiO₂ has been mainly applied to degrade organic and inorganic pollutants both in vapor and in liquid phases. Until a few years ago, the photocatalytic reaction processes were considered highly unselective, particularly in water, but recent researches have demonstrated that it is possible to perform organic selective reactions for synthetic purposes. Very few examples of such reactions in aqueous solutions have been reported [129,130].

Brookite was also tested for the selective oxidation of 4-methoxybenzyl alcohol to 4-methoxybenzaldehyde (p-anisaldehyde) in water [131,132,133]. A good conversion and selectivity (52.5%) was obtained after reasonable irradiation times. The selectivity of the brookite sample was more than 3 and 6 times higher than that obtained with a commercial anatase or P25, respectively. The highest oxidizing rate of the commercial samples corresponded to a lower selectivity for the partial oxidation of the aromatic alcohol.

Defective brookite obtained by helium pretreatment of brookite prepared by thermal hydrolysis of titanium bis(ammonium lactate) dihydroxide was more active than the untreated sample for the photocatalytic reduction of CO₂ [134]. The production of CO and CH₄ was remarkably higher than that observed with defective and defect-free anatase and rutile. The enhancement was attributed to a facilitated formation of oxygen vacancies on brookite that could promote CO₂ activation both in the dark and under illumination.

The band gaps of anatase, brookite and rutile allow only the absorption of the ultraviolet part of the solar irradiation so that many researches have recently been carried out to obtain materials doped with either anions or cations or co-doped with several ions to shift the absorption edge to longer wavelengths. N-doping was found to be particularly effective in enhancing the photocatalytic efficiencies of TiO₂ in the visible-light range [135].

Yin et al. [122,136,137,138] prepared N-doped brookite nanoparticles by a solvothermal treatment at 190 °C of aqueous or alcoholic solutions of TiCl₃ and hexamethylenetetramine. The samples showed higher photocatalytic activity than that of P25 for the oxidative destruction of nitrogen monoxide under irradiation with visible light (λ > 510 nm) and UV-light (λ > 290 nm).

Au-loaded brookite samples prepared by photodeposition from a HAuCl₄·4H₂O solution were more active than pure brookite for the degradation of rhodamine B under visible light irradiation [139]. It is worth noting that highly stable catalytic systems for CO oxidation were obtained by depositing gold on the surface of brookite [140]. The interaction between gold nanoparticles and brookite plays the important role of stabilizing the catalyst against sintering in high temperature environments.

Lanthanide-doped TiO₂ samples have received an increasing attention for their enhanced photocatalytic properties. The increase in activity was attributed to a higher adsorption of the organic pollutants due to the formation of Lewis acid-base complexes between the f-orbitals of the lanthanides and the substrates [141,142]. Bellardita et al. [121,143], firstly, synthesized Sm-loaded brookite powders and films that were tested for the photodegradation of 4-nitrophenol and 2-propanol, respectively. Loading with Sm resulted in a significant improvement of the photoreactivity of brookite and the beneficial effect was ascribed to an increased separation efficiency of the photogenerated electron–hole pairs. Lanthanum-doping was also positive to improve the photocatalytic efficiency of brookite [94] and lanthanum-doped brookite films were studied for dye-sensitized solar cells [113].

The surface hydrophilicity of TiO₂ thin films can be increased by doping [144]. Eshaghi and Eshaghi [145] studied the effect of Cu loading on the surface properties of brookite thin films. The results indicated that the Cu-loaded brookite film was more hydrophilic than the undoped film both under irradiation and in the dark.

The addition of samarium to anatase, brookite and rutile films caused a reduction of the contact angle values [121]. In particular, the Sm-loaded brookite film revealed a nearly superhydrophilic property exhibiting a 3° water contact angle in the dark. The enhanced hydrophilicity was attributed to an increased surface hydroxyl concentration due to the presence of the rare earth metal f orbitals.

Molybdenum/carbon co-doped brookite samples prepared by hydrothermal treatment of Ti(SO₄)₂ in the presence of glucose and ammonium molybdate were active for the photodecoloration of rhodamine B under simulated solar light irradiation [146].

A not adequate attention has been paid to the photocatalytic activity of binary and ternary mixtures of brookite with anatase and rutile. Really, the photocatalytic activity of mixed TiO₂ phases is often higher than that exhibited by the pure single polymorphs. Usually, the contemporaneous presence of different phases of the same semiconductor is beneficial to enhance the photoactivity and in particular, the very efficient commercial TiO₂ Degussa P25 is composed of anatase and rutile. As already demonstrated for other systems [147,148,149], the coupling of different semiconductors allows the vectorial displacement of electrons from one semiconductor to another, leading to more efficient electron/hole separation and greater catalytic reactivity (see Figure 11).

Anyway, Ovenstone [150] observed that biphasic particles containing both anatase and brookite had lower photocatalytic activity than anatase for the decomposition of acetic acid. Similar behaviour was also found for anatase-brookite crystals that were tested for the gas phase oxidation of benzene [151,152]. It is worth noting that the amount of brookite present in these samples was very small.

Yu et al. [153,154,155,156] reported that anatase-brookite composite nanocrystals synthesized by a sonochemical sol-gel method were more active than anatase [153,154] and P25 [153,155,156] for the oxidation of propanone in air. Ozawa et al. [157] found that the photoactivity of an anatase-brookite sample for the oxidation of gaseous CH₃CHO was 5.4 times greater than that of a single-phase anatase sample with comparable crystallite size and surface area. The high efficiency was attributed to the junction between anatase and brookite.

Anatase-brookite composites were more efficient than P25 for the photodegradation of methyl orange [78,79], methylene blue [158], rhodamine B [159,160,161,162], phenol [79,159], salicylic acid [79], acetaldehyde [163] and propanone in air [164,165].

The photoactivity exhibited by a mixture of anatase and brookite prepared from a titanium-lactate complex was slightly higher than that of pure brookite obtained from a titanium-glycolate complex [88]. Both samples were more active than P25 for the oxidation of NO whereas the anatase-brookite composite prepared from a peroxotitanium complex was less active than anatase for the degradation of methylene blue [166].

Qin et al. [167] synthesized nanosized TiO₂ particles using different alcohols as solvents, TiCl₃ and variable amounts of hexamethylene tetramine. Mixed crystals of anatase and brookite photodegraded methyl orange more efficiently than mixtures of anatase and rutile or pure anatase.

Ardizzone et al. [168] prepared anatase-brookite nanocrystals that showed an enhanced activity both for the degradation of NO_x in gas phase and the oxidation of 2-chlorophenol in liquid phase. The photoactivity of the anatase–brookite composites was comparable to that of P25 [169] for the photoreduction of Cr(VI) but much higher for the degradation of toluene in gas phase [170].

Mesoporous anatase-brookite powders showed a high photocatalytic activity for the degradation of rhodamine B that was attributed to the porous structure, large specific surface area, bicrystallinity and small crystalline size [171]. A high photodegradation rate of rhodamine B was also exhibited by hollow submicrospheres with a rough surface synthesized via the combination of hydrothermal treatment and calcination of submicrospheres consisting of a polystyrene core and an amorphous TiO₂ shell [172].

Mahdjoub et al. [173] prepared anatase-brookite powders rich in brookite by hydrolysis of titanium tetraisopropoxide at room temperature. All the mixtures showed high photoactivity for the photodegradation of methyl orange and the best value was observed for the sample that contained the highest percentage of brookite.

The efficiency of the mixed samples depends on the percentages of the single TiO₂ polymorphs so that a great challenge is the possibility to tailor the composition of the biphasial mixtures. Mixed TiO₂ powders with brookite as the predominant phase were obtained in ambient conditions by hydrolysis of TiCl₄ in an acidic water-isopropanol medium [54,174]. The phase composition was tuned by varying the water-isopropanol ratio. The samples were active for the degradation of methyl orange under UV irradiation and a following treatment with N-methylpyrrolidone enhanced the photocatalytic activity under visible light [123,175,176,177].

Bahnemann et al. [23,89] synthesized anatase-brookite mixtures by thermal hydrolysis of a solution of titanium bis(ammonium lactate) dihydroxide in the presence of urea. The ratios between anatase and brookite were easily controlled by the concentration of urea. The anatase-brookite mixtures exhibited higher photocatalytic activity than anatase nanoparticles and P25 for the photocatalytic H₂ evolution from an aqueous methanol solution. On the contrary, the binary mixtures were less active than anatase for the degradation of dichloroacetic acid. For both reactions, the best sample contained 72% anatase and 28% brookite [89].

Shen et al. [68] synthesized TiO₂ polymorphs with a tunable anatase/brookite ratio by an alkalescent hydrothermal treatment of TiCl₃ in the presence of NaCl. The amounts of anatase and brookite were regulated by adjusting the NaCl concentration and the NH₄OH/H₂O volume ratio. The photocatalytic activity of the mixed phases was higher than that of pure anatase and pure brookite for the degradation of rhodamine B and the best sample contained 46.6% anatase and 53.4% brookite. TiO₂ nanoparticles with controllable phases of anatase and brookite were also prepared by heating at 200 °C the suspension obtained by adding tartaric acid and NaOH to an aqueous solution of TiCl₃ and NaNO₃ [19]. The contents of anatase and brookite were tuned by varying the additional molar ratio of C₄H₆O₆ to Ti. The sample containing 78.7% anatase and 21.3% brookite exhibited the highest photocatalytic rate per surface area superior to that of brookite, anatase and P25.

Boppella et al. [162] prepared brookite/anatase and brookite/rutile nanocrystals by thermal hydrolysis of TiCl₄ in a water/ethanol solution. The tuning of the phase compositions was achieved by changing the water/ethanol ratio. Binary mixtures were more active than P25 and pure anatase for the degradation of rhodamine B under visible light.

Recent reports have confirmed the enhanced photocatalytic activity of binary mixtures containing brookite. Jiao et al. [80] prepared a series of anatase-loaded brookite nanoflower hybrids by hydrothermal treatment of a suspension of flower-like brookite, NH₃·H₂O and tetrabutyl titanate. The powder containing 60% of brookite and 40% of anatase showed photoactivity superior to that of anatase and brookite for the degradation of methyl orange and 2,4-dichlorophenol.

Truong et al. [32] synthesized anatase-brookite composites from a titanium oxalate complex. The mixed samples were efficient for the reduction of CO₂ to CH₃OH under both UV–vis and visible light irradiation, mainly due to the presence of doped carbon and nitrogen.

Anatase-brookite mesoporous nanoparticles showed a removal efficiency of Reactive Red 195 similar to that of commercial P25 under optimal conditions [178]. The percentages of anatase (76.8%) and brookite (23.2%) were very similar to the percentages of anatase and rutile in P25.

Nguyen-Phan et al. [25] synthesized a hierarchical rose bridal bouquet nanostructure containing a mixture of layered titanate and brookite that revealed an excellent performance for the degradation of methylene blue. The contemporaneous presence of two phases with different band gap energy facilitated the interfacial electron transfer, reducing the electron/hole recombination and promoting the degradation efficiency (see Figure 12).

Binary and ternary mixed TiO₂ phases supported on kaolinite were synthesized by introducing polymeric Ti cations into the layered aluminosilicate by hydrolysis of TiCl₄ in HCl [179]. The as-prepared samples exhibited strong photocatalytic activity for the removal of acid red G and 4-nitrophenol due to the synergetic effects of the adsorbibility of kaolinite and the catalytic ability of TiO₂. The best results were obtained with the anatase-brookite mixture.

The photocatalytic activity of the anatase-brookite composites depends on the relative percentages of the two phases (see Figure 13). A small or a large amount of brookite is generally unfavourable to improve the activity exhibited by pure anatase or brookite. As general trend, the photoactivity increases with increasing the amount of brookite, reaches a maximum and then decreases [68,80,89,162].

Several authors have reported an optimum ratio between anatase and brookite corresponding to the highest photoactivity for a given reaction. These ratios are often very different and depend on the particular method of synthesis. Besides, the range of examined compositions is often small. Only a few works have recently studied the behaviour of mixtures with brookite percentages ranging from 9.5 to 95% [23,68,89]. A rough comparison among the results of the previous investigations seems to suggest that the effective optimum amount of brookite ranges between 20 and 40%. Anyway, the behaviour of the various anatase-brookite composites also depends on many other factors such as specific surface area, crystallinity degree, crystallite sizes of the different phases, surface hydroxylation and preparation method, that can exert contemporaneously their influence on the photocatalytic process [180,181].

Only few papers have concerned the photocatalytic behaviour of brookite-rutile mixtures. Mixed brookite and rutile particles obtained by hydrothermal post-treatment of H-titanate fibers in an acidic solution were active for the degradation of rhodamine B [182]. Xu et al. [183] synthesized mixed brookite-rutile nanocrystals using TiCl₄ as titanium source and triethylamine as the “adjusting reagent” to tune the brookite/rutile ratio. All the binary mixtures degraded rhodamine B under artificial solar light and the photoactivity of the sample containing 38% brookite and 62% rutile was 6 times higher than that of P25. Similarly, a mixed sample with 83% brookite and 17% rutile prepared by thermal hydrolysis of TiCl₄ exhibited a photoactivity superior to that of P25 and pure anatase for the degradation of rhodamine B under visible light [162]. The improved performance of the brookite-rutile mixtures was attributed to the increased charge separation efficiency resulting from the interfacial electron transfer from brookite to rutile.

The performances of ternary mixtures of anatase, brookite and rutile have been also considered. Lopez et al. [184] firstly reported that an anatase-brookite-rutile mixture was more efficient than a mixture of anatase and brookite, anatase and P25 for the decomposition of 2,4-dinitroaniline. The highest activity was attributed to the lowest E_g of the sample and to the coexistence of three different crystalline phases.

Anatase-brookite-rutile composites were active for the degradation of methyl orange [173], methylene blue [185], acetaldehyde [163], 4-nitrophenol [24] and 4-chlorophenol [66,183]. Samples with a mixed crystal lattice of anatase, brookite and rutile prepared by hydrolysis of titanium butoxide degraded NO_x under UV and visible light illumination [186].

Ternary mixtures of the three TiO₂ polymorphs were obtained by thermolysis of solutions containing Ti powder, HCl, urea and polyethyleneglycol [66]. The photoactivity of the best sample, containing 3.2% brookite, 42.9% anatase and 53.9% rutile, was higher than that of brookite but lower than that of P25.

Di Paola et al. [30] synthesized highly active TiO₂ photocatalysts by thermohydrolysis of TiCl₄ in water at 100 °C. Binary mixtures of anatase and brookite and ternary mixtures of anatase, brookite and rutile were obtained depending on the TiCl₄/H₂O ratio. The photoactivity of some samples was higher than that of Degussa P25 for the degradation of 4-nitrophenol. The most efficient sample contained 28% brookite, 65% anatase and 7% rutile.

Luís et al. [187] synthesized TiO₂ powders with different contents of anatase, brookite and rutile by hydrolysis of titanium tetraethoxide and post thermal treatment. The best results for the degradation of methylene blue were obtained with a binary mixture of anatase and brookite and a ternary mixture of the three TiO₂ polymorphs.

Recently, Liao et al. [188] synthesized anatase, brookite and rutile nanocomposites using toluene as the nonpolar solvent and toluene soluble tetrabutyltitanate as the precursor. The composition was easily tailored by varying the reagents volume ratio. The ternary mixtures were more efficient than P25 for the degradation of methyl orange. The photoactivity of the best sample, containing 29.9% anatase, 27.9% brookite and 42.2% rutile was almost twice as high as that of P25.

N-doped mixtures of brookite and rutile obtained by using 1-propanol or 1-butanol as solvents were less active than N-doped brookite [137]. Instead, N-doped (anatase-brookite) samples obtained in methanol were more active than N-doped anatase or N-doped (anatase-rutile) mixtures for the degradation of methyl orange under visible light [189].

Shao et al. [190] synthesized meso-macroporous N-doped TiO₂ materials with a bicrystalline (anatase and brookite) framework by thermal treatment of a mixture of hierarchical mesoporous-macroporous TiO₂ powder with an urea solution, followed by calcination at 350–550 °C. The samples showed a good photocatalytic activity for the photodegradation of methyl orange and rhodamine B under UV and visible-light irradiation. The photoactivity was significantly improved with increasing the nitrogen doping concentration.

Li and Liu [191] prepared N-doped (anatase-brookite) nanocatalysts by solvothermal treatment of titanium butoxide with dimethylformamide in the presence of acetylacetone at 180 °C. The binary mixtures were efficient for the degradation of methylene blue under both visible and UV light and the photoactivity of the most active sample containing 24.6% brookite and 75.4% anatase was comparable to that of P25. Visible light N-doped anatase-brookite mixtures were also obtained by hydrolysis of TiCl₄ in CH₃COOH/HNO₃ media followed by rapid calcination [192]. The samples were more active than a commercial N-doped anatase for the photodegradation of phenol and the visible light activity increased with increasing the brookite particle size.

Fe³⁺-doped TiO₂ aerogels containing both anatase and brookite were more efficient than P25 for the salicylic acid degradation uder UV irradiation [193]. Hao and Zhang [194] prepared Fe³⁺ and nitrogen co-doped mesoporous anatase-brookite mixtures by a modified sol-gel method from titanium butoxide using dodecylamine and Fe(NO₃)·9H₂O as nitrogen and Fe dopants, respectively. The photoactivity of the co-doped powders for the degradation of 2,4-dichlorophenol under visible light irradiation was higher than that of the N-doped sample and P25.

Yu et al. [195] synthesized F⁻-doped (anatase-brookite) powders by hydrolysis of titanium tetraisopropoxide in an NH₄F aqueous solution. The photocatalytic activity of the F⁻-doped samples for the oxidation of propanone under UV irradiation exceeded that of P25 when the NH₄F/H₂O molar ratio was in the 0.5–3 range.

I-doped mixtures of anatase and brookite tested for the photocatalytic reduction of CO₂ in the presence of H₂O vapor exhibited a significant enhancement of activity under both visible and UV–vis illumination [196].

Ag-loaded (anatase/brookite) powders synthesized via an alkalescent hydrothermal process were more efficient than the Ag-free composite and P25 for the degradation of methyl orange [197]. Yu et al. [198] prepared Ag–TiO₂ multiphase thin films on quartz substrates by liquid phase deposition from an aqueous solution of (NH₄)₂TiF₆, AgNO₃ and H₃BO₃, followed by calcination at 500 °C. The photocatalytic activity of the films consisting of anatase, brookite, rutile and silver nanoparticles exceeded that of an anatase film. High activity for the photodegradation of gaseous toluene was also exhibited by Ag–TiO₂ three-phasic powders synthesized from titanium butoxide and AgNO₃ using a microwave-assisted method [199]. The enhanced efficiency of the Ag–TiO₂ multiphase samples was attributed to the presence of many heterojunctions among the different TiO₂ phases and Ag.

Nassoko et al. [200] prepared Nd-doped (anatase-brookite) powders that showed better activity than the undoped TiO₂ mixture and P25 for the photodegradation of rhodamine B under visible light.