Visible-Light-Active Iodide-Doped BiOBr Coatings for Sustainable Infrastructure

The search for efficient materials for sustainable infrastructure is an urgent challenge toward potential negative emission technologies and the global environmental crisis. Pleasant, efficient sunlight-activated coatings for applications in self-cleaning windows are sought in the glass industry, particularly those produced from scalable technologies. The current work presents visible-light-active iodide-doped BiOBr thin films fabricated using aerosol-assisted chemical vapor deposition. The impact of dopant concentration on the structural, morphological, and optical properties was studied systematically. The photocatalytic properties of the parent materials and as-deposited doped films were evaluated using the smart ink test. An optimized material was identified as containing 2.7 atom % iodide dopant. Insight into the photocatalytic behavior of these coatings was gathered from photoluminescence and photoelectrochemical studies. The optimum photocatalytic performance could be explained from a balance between photon absorption, charge generation, carrier separation, and charge transport properties under 450 nm irradiation. This optimized iodide-doped BiOBr coating is an excellent candidate for the photodegradation of volatile organic pollutants, with potential applications in self-cleaning windows and other surfaces.


INTRODUCTION
Bismuth oxyhalides, with formula BiOX (X = Cl, Br, and I), have received broad interest in areas such as water splitting, pollutant degradation, and photoelectrochemical sensing, 1,2 with particular focus as emerging materials in photocatalytic applications. 3,4The tetragonal crystal structure of BiOX materials consists of [Bi 2 O 2 ] 2+ layers sandwiched between double X − slabs. 5Due to the induced polarization, this layered arrangement promotes the formation of an internal static electric field perpendicular to the layers.The existence of this electric field is thought to favor the separation of photogenerated charge carriers, contributing to improved photocatalytic efficiency. 6In addition, since BiOX materials are indirect semiconductors, their band structures also facilitate the separation of photogenerated charge carriers, thus further reducing recombination loss. 7−10 The band gap energy of BiOBr (E bg = 2.7 eV) allows for the harvesting of photons in the high-energy end of the visible range (up to 460 nm), which is particularly relevant to solar applications. 10BiOI has the smallest band gap energy (E bg = 1.8 eV) among bismuth oxyhalides, with photon absorption across the entire visible range; however, its narrow band gap accompanies poor redox abilities and also promotes rapid charge recombination. 11,12On the other hand, BiOCl (E bg > 3.2 eV) can barely absorb photons in the visible range, and thus, it is less suitable for solar applications.
−22 Conveniently, BiOX materials form similar crystal structures and compositions, with Cl − and Br − or Br − and I − showing similar ionic radii (Cl − = 1.67 Å, Br − = 1.82 Å, and I − = 2.06 Å), 23 allowing for the continuous adjustment of band structures upon doping with minimum formation of detrimental crystal defects. 24hloride-and iodide-doped BiOBr materials have been reported as promising photocatalysts for applications involving organic pollutant decomposition, 25 photoelectrochemical biosensors, 26 or carbon dioxide (CO 2 ) reduction. 27Most of these materials, however, have been produced in powder form from hydro/solvothermal methods, 20,28,29 which poses a challenge for scale-up and commercialization.Powder photocatalysts can pose additional challenges to their end use, as they may require embedding within a host material or, when used in solution, may require extraction and recovery steps.Therefore, some work has been reported for the thin-film fabrication of doped BiOX by successive ionic layer adsorption and reaction (SILAR), sol-assisted, and drop-casting methods. 25,30,31However, in sol-assisted and drop-casting processes, BiOX powders have to be prepared first, and then extra calcination or coating steps are necessary for film deposition.In addition, the SILAR method can only be utilized to form samples of small sizes on the lab-scale due to its complicated steps and high cost.In this work, iodide-doped BiOBr thin films were deposited on substrates directly using aerosolassisted chemical vapor deposition (AACVD), which is a feasible, scalable, and cheap method for thin-film deposition.The photocatalytic properties of the iodide-doped BiOBr films were investigated using the standard smart ink method based on the reduction of resazurin (Rz) dye. 32The influence of iodide concentration on the crystal structure, compositions, morphologies, and optical properties of the deposited films was fully characterized, as well as the impact of iodide loading on their photocatalytic performance under UV (365 nm) and visible light (450 and 627 nm).A significant improvement in the ink degradation rate was observed after doping iodide in BiOBr under visible-light irradiation, with the best performance seen for 2.7 atom % dopant concentration.Photoluminescence spectroscopy and photoelectrochemical measurements were carried out to gain an insight into the potential underlying mechanism for the observed photocatalytic behavior of these materials.

EXPERIMENTAL SECTION
2.1.Film Fabrication.Stoichiometric amounts of BiBr 3 and BiI 3 (total 1 mmol) were dissolved in 25 mL of anhydrous dimethylformamide (DMF) and ultrasonicated for 10 min to ensure the mixing of precursors.An AACVD setup was used for the film deposition, where a flask containing the precursor solution with a humidifier was attached to the CVD reactor and the generated mist was carried into the reactor under compressed air flow (1.0 l min −1 ).The AACVD process was carried out until the precursor solution was exhausted.Iodide-doped BiOBr x I 1−x films were grown on FTO glass (NSG TEC 15) substrates (15 cm × 4.5 cm) at 300 °C for 60 min.Substrates were left to cool to room temperature under air flow.
For comparison, parent materials, BiOBr and BiOI, were deposited under the same conditions from BiBr 3 (1 mmol) and BiI 3 (1 mmol), respectively, in anhydrous DMF (25 mL), following a literature procedure. 9.2.Physical Characterization.Grazing-incidence X-ray diffraction (GIXRD) patterns were obtained between 5 and 60°(2θ) (0.05°steps, 1.5 s/step) using a Panalytical Empyrean diffractometer, with Cu Kα radiation (λ = 1.5406Å) at 40 kV and 40 mA emission current.The angle of the incident beam was 1°.Rietveld refinement of XRD data was carried out using MDI Jade 6.The film morphology was studied using a JEOL JSM-7600 field emission scanning electron microscope (SEM) and energy-dispersive spectroscopy (EDS) system.UV−vis−NIR transmission spectra were measured within 300−1100 nm using a Shimadzu UV-3600i Plus spectrometer.X-ray photoelectron spectroscopy (XPS) was carried out using a Thermo Scientific K-Alpha spectrometer with monochromated Al Kα 1 radiation (λ = 8.3418 Å).A dual beam system was employed for charge compensation.Survey scans (0−1200 eV) were obtained at 50 eV pass energy.Depth profiling was performed using an Ar + ion beam for surface etching.All peak positions were calibrated to adventitious carbon (284.8 eV) by using the software CasaXPS.Photoluminescence (PL) spectra were obtained at room temperature using a Renishaw RM1000 spectrometer with an excitation wavelength of 325 nm.Atomic force microscopy (AFM) was performed by using a Keysight 5500 scanning probe microscope.Images over a projected area (5 μm × 5 μm) were recorded in tapping mode using a Si cantilever (NuNano SCOUT-70) with a resonant frequency of ∼70 kHz and ∼2 N m −1 spring constant.The roughness factor was obtained by dividing the measured surface area by the projected area.
2.3.Functional Test.The photocatalytic properties of the films were assessed using the smart ink test upon evaluation of kinetics in the transformation of resazurin (Rz) to resorufin (Rf) under illumination. 32The ink consists of an aqueous hydroxyethyl-cellulose solution containing the redox dye Rz and a sacrificial electron donor (typically glycerol).Positive holes generated upon photoexcitation of the film react with glycerol, while the blue Rz dye is reduced irreversibly to pink Rf by photogenerated electrons.In this work, the Rz ink was prepared following a reported recipe. 330.006 mmol of hydroxyethyl-cellulose was dissolved in 98.5 mL of distilled water under stirring conditions for 8 h.0.53 mmol of resazurin sodium salt and 144.78 mmol of glycerol were added to the aqueous polymer and then stirred overnight.The prepared ink was stored in the refrigerator and fully mixed before use.After preirradiation of the films (1.5 cm × 1.5 cm) under UVC light for 10 min, 0.5 mL of the ink was spincoated on the surface at 6000 rpm for 10 s.
The photocatalytic testing was carried out using a board containing a set of LED lights ranging from 365 to 627 nm (Figure S1). 34In this work, top-down irradiation was carried out at 365 nm (2.87 mW cm −2 ), 450 nm (2.01 mW cm −2 ), and 627 nm (1.32 mW cm −2 ).Average RGB values across the film area were recorded as a function of irradiation time from digital images using ImageJ software, where the normalized red component (R t = RGB R,t /(RGB R,t + RGB G,t + RGB B,t )) was used to monitor the color change in the Rz ink. 35.4.Photoelectrochemical Measurements.Transient photocurrent was collected in a three-electrode configuration with an applied voltage of 1.0 V RHE at room temperature and under intermittent irradiation.A 220 W Xe lamp with a 420 nm cutoff filter was used as the light source.The electrolyte was an aqueous 0.5 M Na 2 SO 4 solution at pH 6.6.The deposited iodide-doped BiOBr film was used as the working electrode with Ag/AgCl/3M-KCl and Pt mesh acting as reference and counter electrodes, respectively.An electrochemical workstation (IVIUMSTAT, Netherland) was implemented to apply voltages and measure currents.In the same system, current density−voltage (J−V) curves were collected by sweeping the voltage from 0.1 to 1.1 V RHE at a rate of 10 mV s −1 .The applied voltage in our current density−voltage curves was reported against the reversible hydrogen electrode (V RHE ), upon conversion using the Nernst equation , where V Ag/AgCl is the applied potential versus the Ag/AgCl reference electrode and V Ag/AgCl ⌀ is the standard reference potential (0.197 V NHE at 25 °C).

Synthesis of Iodide-Doped BiOBr Films and Dopant (Iodide) Distribution.
Iodide-doped BiOBr films were grown via AACVD from mixtures of BiBr 3 and BiI 3 with predetermined molar ratios of 9:1, 8:2, 7:3, 6:4, 5:5, 4:6, and 3:7.Henceforth, the obtained samples will be denoted as I1-BB, I2-BB, I3-BB, I4-BB, I5-BB, I6-BB, and I7-BB, respectively, where the number represents the amount of iodine added to the precursor solution per 10 halide anions.During the formation of BiOX solid solutions, iodide was more difficult to be doped into the lattice structure compared with other halides due to its larger radius. 21In addition, in the AACVD process, it is relatively hard to form a mist from DMF solutions containing BiI 3 , leading to loss of the iodide source.Therefore, the doping efficiency, as determined from XPS analysis of the halide concentrations in the as-deposited films, was roughly a factor of 10 lower for the BiI 3 source as compared with the BiBr 3 source.Interestingly, once the ratio of BiBr 3 and BiI 3 in precursor solutions reached above 3:7, such as 2:8 and 1:9, the amount of iodide increased dramatically and finally dominated in the deposited films with poor catalytic performance.Therefore, only pure BiOI and doped films fabricated from precursors with BiI 3 less than 70% were investigated further.
Due to heavy peak overlap between the Sn Lα from the substrate and the I Lα from the films, EDS was not possible to obtain accurate atomic ratios in films.Therefore, surface composition of the films and elemental valence states were studied using high-resolution XPS.For parent BiOBr and BiOI materials, besides Bi and O elements in both, Br and I elements were observed separately, with peak positions comparable to previous reports observed (Figures S2 and  S3). 9,36In all doped samples, Bi, Br, O, and I elements coexisted.The Bi 4f spectra were best fitted with a doublet of doublets.The primary Bi 4f 7/2 was centered at 159.6 (±0.1) eV and assigned to Bi 3+ , while the secondary Bi 4f 7/2 peak at 158.5 (±0.3) eV closely matched metallic bismuth that often presents itself in XPS due to the photoreduction of BiOX under incident X-rays (Figure 1a). 9,37The Br 3d signals were best fitted using a doublet with the 3d 5/2 centered at 68.8 (±0.1) eV, matching Br − (Figure 1b).As for the highresolution XPS spectra of I 3d (Figure 1c), the signal produced I 3d 5/2 peaks centered at 619.2 (±0.3) eV, indicating the successful doping of I − into the BiOBr films. 9,38n enhancement in the I 3d peak intensity was clearly observed due to the increasing iodide content in the films (Figure 1d).Atomic ratios of Br and I in I-BB films were calculated from the high-resolution XPS Br 3d and I 3d signals, as summarized in Table 1.
XPS depth profiling studies showed that the peak intensity from I 3d decreased dramatically with increasing etch time, thus demonstrating that iodide was surface segregated (Figure 1e), which is advantageous for heterogeneous photocatalysis as this is a surface-based process.

Structural Characterization of Iodide-Doped BiOBr Films. The structural properties of the iodide-doped
BiOBr films were first determined by GIXRD (Figure 2a).The patterns of the parent materials, BiOBr and BiOI, corresponded to pure tetragonal phases (PDF 78-0348 and PDF 10-0445, respectively).The incorporation of iodide in the films resulted in a gradual shift of the (102) peak toward smaller angles (Figure 2b), due to the larger ion radius of I − (2.02 Å) compared to that of Br − (1.82 Å). 23 This shift was indicative of the formation of a solid solution rather than a simple mixture of the parent phases.Within the dopant concentration range explored in our work, all samples exhibited the tetragonal phase of BiOBr, which further confirmed the doping of iodide in the I-BB films.Unit cell parameters and average crystallite sizes of the deposited films were determined from XRD data (Table S1).It can be seen that the introduction of iodide in the structure has significant impact on the c-lattice axis, which is consistent with the literature. 39The increasing trend in c values showed a slight deviation from Vegard's law (Figure S4), which could be attributed to the intrinsic difference in the halide anionic radii and weak van der Waals forces between adjacent layers.The preferential surface segregation of iodine over bromine, as determined by XPS depth profile analysis, may also be responsible for this deviation.
Top-down SEM was used to study the morphology of the film surface (Figures 3 and S5).All deposited films featured compact and uniform plate-shaped grains.These square nanoplatelets were commonly observed in the investigations of BiOX materials. 9,40,41Considering the smaller calculated average crystallite size (Table S1), the particle size observed in top-down SEM images indicated that nanograins in the asfabricated films were constituted of multiple crystallites.Significant morphological differences were observed among the BiOBr and I-BB samples.The grain edge became sharper with more iodide doped.In addition, the grain size of the films decreased from ∼1.3 to ∼0.6 μm.A similar trend was also reported in other iodide-doped BiOX. 21All films showed a similar thickness of around 400 nm.
As determined by AFM, the roughness factors of the films were ∼2.03, ∼ 1.67, and ∼1.76, respectively for BiOBr, I4-BB, and BiOI films deposited on FTO (Figure 3).The AFM images of other doped samples are also shown in Figure S6.Calculated roughness factors of film surfaces ranged from 2.03 to 1.48, as summarized in Table S2.It was observed that the surface roughness decreased slightly with the increased dopant amount.This likely resulted from the more compact film growth induced by smaller grains, which was shown in the SEM images as well.Greater surface roughness and nanostructuring of films are possible to enhance the photocatalytic activity.However, the roughest BiOBr film did not exhibit the best performance in visible-light photocatalysis, and the change trend of photocatalytic activity, which is discussed later, is not in accordance with that of the roughness factors, indicating that the dramatic difference in visible-light photocatalytic activity could not be attributed to the marginal difference in the samples' surface areas.

Optical Properties of Iodide-Doped BiOBr Films.
The color of the as-deposited I-BB films changed from white to yellow with increasing iodide content, evidenced by a gradual red shift of the absorption edge of the films from 400 to 475 nm (Figure 4a).The sharp edge of pure BiOBr at 400 nm was attributed to an intrinsic transition from the valence band to the conduction band. 42Upon doping, however, the absorption curves of the films showed long tails, as it corresponds to the formation of impurity levels in the forbidden band of doped semiconductors. 43,44The red shift of the absorption onset involved a significant decrease in band gap energy from 2.72 to 2.34 eV upon incorporation of iodide, as estimated from Tauc plot approximations (Figure 4b). 20he regular decreasing trend determined that the band gap could be tailored through adjusting the doping content in the films.
Computational analysis of BiOX materials has shown that the valence band maximum (VBM) is largely composed of X np states, and the conduction band minimum (CBM) is mainly due to the Bi 6p state with minor contributions from X np states. 7,39,45Therefore, while uplift of the VBM and downward shift of the CBM would be induced simultaneously during the doping process, the VBM is affected more significantly by the introduction of iodide.As a result, even a trace amount of I − dopant can drastically impact the VBM position leading to a reduced band gap. 46.4.Photocatalysis Test.The photocatalytic behavior of the as-prepared I-BB films was investigated following the smart Rz ink test under UV (365 nm) and visible light (450 and 627 nm).The Rz-to-Rf transformation was monitored through a digital photographic method from digital images of the Rz ink coated on films and recorded as a function of irradiation time.
The photocatalytic activity of the samples under UV light was monitored from digital images of the Rz ink up to 720 s at an initial interval of 10 s (up to 240 s), followed by an interval of 30 s.No transformation of the Rz dye was observed in the absence of a photocatalyst (Figure 5a).A rapid color change over the initial irradiation period (150 s) was observed in pure BiOBr, followed by a gradual decrease in kinetics upon I − loading (Figure 5b).From the results of calculated rates of Rz conversion (dR t /dt), it was found that this variation trend was nearly linear with the exception of sample I4-BB, which showed an unusual peak in dR t /dt (Figure 6a).The test under 450 nm irradiation showed a very different scenario.Digital images were obtained for up to 990 s with intervals of 30 s. Figure 5c shows selected images of a representative I4-BB sample over an initial irradiation period of 450 s.It was observed that the ink-coated I4-BB film followed a smooth color change from blue to pink, while no color change was observed for the ink in the absence of a photocatalyst (Figure 5c).The blank experiment on an FTO substrate showed that the ink was stable, and there was no sign of photolysis under these irradiation conditions.Photosensitization of the film was also ruled out given the peak of the absorption spectrum of the ink at ca. 600 nm (Figure S7).Thus, the color change of the ink was only attributed to the photocatalytic behavior of the films.This assumption was confirmed from photocatalytic tests under 627 nm, where the ink slowly bleached in the absence and presence of bismuth oxyhalides, indicating that the self-degradation of the ink occurred through photolysis under those irradiation conditions (Figure S8). 47erived from recorded images, a series of R t values vs irradiation time t could be obtained (Figure 5d).From inspection of the figure, sample I4-BB (2.7 atom % I − doping) showed the fastest color change, reaching a plateau at around 300 s under 450 nm illumination.The trend of kinetics reaches a maximum with this sample, gradually increasing from pure BiOBr to I4-BB and decreasing upon further I − loading.Pure BiOI showed no activity under these conditions, which was attributed to an ultrarapid recombination of charge carriers in this material.The remarkable photocatalytic activity of I4-BB exemplified from such a rapid photodegradation of the Rz dye was significantly superior to that of commercial photocatalysts, such as TiO 2 , WO 3 , and C 3 N 4 , and comparable to CdS and BiOCl/Plaster of Paris composites, even under the light source with a lower power density. 47,48n order to compare the ink degradation rates of all samples intuitively and quantitatively, dR t /dt was calculated (Figure 6b) and is summarized in Table 2.With the increased I − amount, the values of dR t /dt improved and then declined, where I4-BB was able to degrade the Rz dye at a rate of 4.96 × 10 −4 s −1 , showing the optimal performance among all measured films.The pure BiOBr was even one order of magnitude lower (5.22 × 10 −5 s −1 ) in the Rz conversion rate than I4-BB.
These observations can be explained by consideration of band shifting upon incorporation of the iodide dopant.Under UV irradiation (3.40 eV), electrons can be excited into the conduction bands (CB) of all of the samples, with BiOBr having the most positive valence band (VB) and thus the highest oxidation ability within the family.The sudden relative increase observed for sample I4-BB under UV irradiation suggests there may be other factors influencing the photocatalytic activity of these samples, outside the question of VB edge potential.With a relatively stable CBM position, the uplift of VB edge position upon I − loading will narrow the band gap of the materials, favoring the absorption of low-energy photons but also weakening the oxidation ability of holes in the VB. 49,50hese two competitive factors led to an optimized performance in I4-BB (2.7 atom % I − ) in the case of the tests under 450 nm.
3.5.Photoluminescence and Photoelectrochemical Measurements.Further insight into the photocatalytic properties of the I-BB films was gathered from photoluminescence (PL) and photoelectrochemical (PEC) analysis.PL emission results from the recombination of photogenerated electrons and holes after excitation back in the ground state.Thus, the weaker PL emission intensity is widely attributed to less recombination and better separation efficiency of photogenerated electron−hole pairs. 51,52Figure 7a shows the contrast in PL emission between the I-BB samples and the parent materials, BiOBr and BiOI, under UV laser irradiation (λ = 325 nm).As observed, pure BiOBr and BiOI showed no emission peaks, indicating that there was no radiative  recombination of photogenerated charge carriers in these materials due to their indirect band features.−57 On the other hand, the impurity states generated by I − doping led to strong PL emission spectra around 600−625 nm in the I-BB samples.These spectra were consistent with characteristic emissions from defect states in iodide-doped BiOBr. 15,54n addition, the PL band maxima gradually red-shifted upon increasing dopant concentration, as a result of band gap narrowing and increasing photon adsorption (Figure 7b).It is interesting to note that the band intensities decreased upon I − dopant incorporation up to 2.7 atom % (I4-BB), where the intensity reached its lowest value (Figure 7a).This was consistent with our previous results during the photocatalytic tests and suggested that I4-BB had the lowest carrier recombination rate within the family of I-BB materials in this study.Furthermore, the separation efficiency of photogenerated charge carriers was investigated by PEC studies.As shown in Figure 8, all of the I-BB films and parent materials displayed a stable, reversible and positive transient photocurrent response during three on/off intermittent visible-light irradiation cycles, indicating an effective transfer of charge carriers and successful electron collection for these samples in the PEC cell. 29The photocurrent levels remained similar for each sample, demonstrating an excellent photoresponse stability.The figure shows intense photocurrents in the I-BB samples compared to the parent materials, following a trend in accordance with that of the photocatalytic behavior of these samples in the smart Rz ink test.The strongest transient photocurrent intensity was recorded for I4-BB.−60 In addition, compared with pure BiOBr, the enhanced photon absorption of the doped samples, shown in the transmittance spectra (Figure 4a), was also one of the reasons for the larger photocurrent response.
Current density−voltage (J−V) measurements were carried out with voltage swept from 0.13 V RHE to 1.10 V RHE and J−V curves were obtained under on/off illumination, where the anodic photocurrent increased steadily with the increased voltage (Figure S9).Pure BiOBr and BiOI showed small photoanode currents with onset potentials of ∼0.33 and ∼0.81 V RHE , respectively, which could result from the poor absorption of low-energy photons and high charge recombination rates.Nevertheless, the onset potential in I4-BB shifted negatively to ∼0.25 V RHE and the current density of this sample (46.5 nA cm −2 at 1.0 V RHE ) increased by almost 10 times compared to pure BiOBr, which was attributed to its fast charge separation and excellent charge transfer ability.It has been reported as well that through the introduction of iodide in BiOBr, the charge transfer resistance of materials could be reduced, leading to higher transfer efficiency of photogenerated charge carriers. 41,61,62It is worth noting that the BiOI sample showed photocathodic behavior under cathodic electrode polarization (<0.8 V RHE ) and photoanodic behavior under anodic electrode polarization (>0.8 V RHE ) (Figure S9), which is consistent with the observed transient anodic photocurrent response of BiOI with the applied voltage at 1.0 V RHE (Figure 8).−65 In bulk semiconductors, photogenerated charge carriers are separated by the electric field of the space charge region, and their transport direction can be determined.The potential drop in the direction perpendicular to the plane of BiOI platelets is considered too small to influence charge transfer due to the small platelet thickness.Therefore, the transfer of photogenerated charge carriers is determined by the electrode−electrolyte interface rather than by the built-in electric field of BiOI.Under cathodic polarization, the conduction band edge of BiOI takes a more negative potential than the redox potential of the electrolyte, which makes the electron transfer from the BiOI electrode to the electrolyte possible.The opposite trend takes place under anodic polarization. 65,66ased on the above analysis of PL and PEC measurements, an improved separation and transport behavior of photogenerated charge carriers in the I-BB samples, especially I4-BB, have been demonstrated compared to those in parent materials.Our results confirmed previous reports in the literature 41,67 suggesting that hybridization of Br 4p and I 5p orbitals could significantly reduce the mobility of photoninduced holes, whereas the mobility of electrons was rarely influenced, leading to reduced charge recombination.In addition, impurity defects induced by doped iodide may contribute to the trapping of charge carriers in the I-BB materials, facilitating charge carrier separation to a certain degree. 68,69An excess of dopant sites (beyond 2.7 atom % dopant concentration, I4-BB in our case), however, can serve as recombination centers, leading to the detriment of the photocatalytic activity. 43The electronic structures of materials are also able to be reconstructed with the increased dopant amount, resulting in the narrowed band gaps while sacrificing the redox ability. 70In addition, the potential internal electric field within the intrinsic layered structure of BiOX materials can play an important role in the photocatalytic performance of these materials. 71The polarization of the internal electric field is possible to occur due to anionic doping, which is beneficial to the charge transport directed by the electric field force. 22,72However, our GIXRD results showed that c values increased with I − doping (Table S1), corresponding to an enlarged interlayer space, and thus a longer distance for the charge transport, against the efficient separation of photogenerated charge carriers. 73,74As a result, synergistically influenced by the above factors, the charge separation and transport efficiency of iodide-doped BiOBr was improved and maximized when the I − concentration reached 2.7 atom %, making for the optimized visible-light photocatalytic performance.

CONCLUSIONS
Iodide-doped BiOBr (I-BB) thin films were successfully deposited in a single step on FTO glass substrates by aerosol-assisted chemical vapor deposition (AACVD).Dopant levels ranged from 0.7 to 6.3 atom %, as confirmed by XPS.Structural analysis was carried out by GIXRD, SEM, and AFM.The photocatalytic properties of the I-BB films were explored using the smart ink test, based on the photodegradation of resazurin dye under UV and visible light (LED source).An optimum photocatalytic performance in the visible range (450 nm) was observed for the sample containing 2.7 atom % dopant concentration.This optimized behavior was attributed to a balance between an enhanced photon absorption from band gap narrowing and improved charge carrier separation and transfer ability, based on photoluminescence and photoelectrochemical measurements.The methodology employed in the synthesis of these visible-light-active coatings is a step forward in the engineering of efficient materials for sustainable infrastructure.The pleasant yellow color of these I-BB coatings is attractive for their use on self-cleaning windows, for instance, acting on volatile organic pollutants under sunlight.

■ ASSOCIATED CONTENT
* sı Supporting Information

Figure 1 .
Figure 1.High-resolution XPS spectra of (a) Bi 4f, (b) Br 3d, and (c) I 3d from the I4-BB film grown via AACVD.(d) High-resolution XPS spectra of I 3d from all I-BB samples.(e) XPS depth profile analysis of I 3d from the I4-BB film on FTO.

Figure 2 .
Figure 2. (a) GIXRD analysis of I-BB films with increasing concentration of iodide dopant (from bottom to top, molar ratios of 0.7, 1.0, 1.9, 2.7, 3.3, 5.3, and 6.3).The patterns of parent BiOBr and BiOI materials were included for reference.The symbols (*) highlight peaks due to the FTO substrate.(b) Selected region highlighting the shift of the (102) peak upon increasing dopant content.

Figure 4 .
Figure 4. (a) Transmittance spectra of BiOBr, BiOI, and I-BB films deposited on FTO.(b) Indirect allowed optical band gap transitions were calculated through the Tauc plot method.

Figure 5 .
Figure 5. (a) Images of the ink coating on BiOBr (upper) and FTO (lower) under 365 nm irradiation.(b) R t vs t plots of the smart Rz ink on BiOBr, BiOI, and I-BB films deposited on FTO under 365 nm irradiation.(c) Images of the ink coating on I4-BB (upper) and FTO (lower) under 450 nm irradiation.(d) R t vs t plots of the smart Rz ink on BiOBr, BiOI and I-BB films deposited on FTO under 450 nm irradiation.

Figure 6 .
Figure 6.Rates of Rz conversion (dR t /dt) of BiOBr and I-BB films deposited on FTO under the irradiation of (a) 365 and (b) 450 nm.

Figure 7 .
Figure 7. (a) Photoluminescence spectra of BiOBr, BiOI, and I-BB films deposited on FTO excited at 325 nm and (b) normalized spectra of I-BB films.

Figure 8 .
Figure 8. Transient photocurrent response of BiOBr, BiOI, and I-BB films deposited on FTO, with the voltage kept at 1.0 V RHE .A 220 W Xe lamp with a 420 nm cutoff filter was used as the light source (1 sun illumination).The electrolyte was an aqueous 0.5 M Na 2 SO 4 solution at pH 6.6.

Table 1 .
Halide Atomic Concentrations in As-Fabricated I-BB Films on FTO

Table 2 .
dR t /dt of BiOBr and I-BB Films Deposited on FTO Under 450 nm Irradiation