Open Access Article
Iqra Ramzan
,
Cesar III D. L. Reyes,
Ivan P. Parkin
and
Claire J. Carmalt
*
Materials Chemistry Centre, Department of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK. E-mail: c.j.carmalt@ucl.ac.uk
First published on 24th June 2026
Antimony trioxide (Sb2O3) thin films were deposited using aerosol-assisted chemical vapor deposition (AACVD). Varying the deposition temperature and precursor concentration increased the film thickness and resulted in a shift from phase-pure senarmontite Sb2O3 to mixed Sb2O3/Sb0 films due to a temperature-driven reductive transformation. These changes altered crystallinity, grain structure, defect chemistry, and light-scattering behaviour, which in turn controlled the electrical, optical, and surface properties of the coatings. Films deposited at higher temperatures produced thicker films containing metallic Sb crystallites that enhanced charge transport but reduced optical clarity. X-ray diffraction (XRD) was used to evaluate the structural properties of the films, while scanning electron microscopy (SEM) revealed distinct changes in surface morphology with varying deposition conditions. X-ray photoelectron spectroscopy (XPS) analysis revealed the oxidation state of Sb in the films, with temperature-dependent variations in residual carbon content. Despite its wide technological relevance, Sb2O3 remains largely unexplored using AACVD. This work represents a detailed study of antimony-based thin films prepared via AACVD, demonstrating tunable film properties and potential for use in optoelectronic and sensing applications.
Like many other metal-oxide semiconductors, Sb2O3 has been employed in optoelectronic devices such as solar cells and UV LEDs.21–24 It is also used in heat-reflective coatings (heat mirrors), where it enhances the infrared reflectivity of the surface and improves thermal efficiency.25,26 Antimony trioxide exists in two polymorphic phases: the low-temperature senarmontite phase, exhibiting a cubic shape, and the high-temperature valentinite phase, which displays an orthorhombic shape.10,27,28
Optoelectronic and structural properties of Sb2O3 thin films have widely been investigated using vacuum deposition methods.8,25,26,29,30 Tigau et al. reported that substrate temperature strongly influenced the crystallinity, optical transmittance and optical band gap of polycrystalline Sb2O3 thin films.8 Tigau et al. also demonstrated that post-deposition heat treatment significantly altered the electrical conductivity of Sb2O3 films due to thermally induced structural and microstructural changes.25 Divya et al. reported transition-metal-doped Sb2O3 thin films exhibited high near-IR transmittance together with anti-reflective and UV-blocking behaviour, highlighting the multifunctional optical properties achievable in Sb2O3-based coatings.26 Tigau further investigated the preparation and characterization of Sb2O3 thin films, reporting their optical transparency, band gap behaviour and surface morphology.29 In a separate study, Tigau et al. demonstrated that film thickness strongly influenced the crystallinity, optical transmittance and electrical resistivity of Sb2O3 thin films.30 Binions et al. deposited insulating Sb2O3 films via atmospheric pressure chemical vapor deposition (APCVD) and investigated their gas sensing properties.31
While most reported studies have used vacuum-based techniques, these methods often involve high-cost equipment and limited scalability for large-area coatings.32 In contrast, solution-based CVD routes such as AACVD offer a simpler, more controllable and cost-effective alternative for producing high-quality oxide films under atmospheric conditions. In addition, AACVD does not require highly volatile precursors, allowing greater flexibility in precursor and solvent selection compared to many vacuum deposition techniques. This technique has been successfully applied to deposit transparent conducting oxides such as ZnO33–35 and SnO2.36,37 Although film growth in AACVD can be influenced by aerosol transport and local gas flow dynamics, reproducibility is generally less challenging for undoped oxide systems than for multicomponent or heavily doped films, where precise dopant incorporation becomes increasingly important. In the present study, all depositions were repeated three times under identical conditions, with negligible variation observed in the optical and electrical properties of the films. However, to the best of our knowledge, no detailed study has yet reported the deposition and optimisation of Sb2O3 thin films via AACVD.
In this work, Sb2O3 thin films were deposited at varying substrate temperatures and precursor concentrations. The influence of these parameters on the structural, morphological, optical, and electrical properties of the films was systematically investigated. The study particularly focuses on the relationship between film composition and the resulting film conductivity, as well as the onset of partial reduction to metallic Sb at higher temperatures. This work provides new insights into the controlled synthesis of Sb2O3 thin films via AACVD, expanding their potential use in optoelectronic, sensing, and dielectric applications.
A horizontal bed AACVD reactor was used, with the glass substrate placed horizontally on a smooth graphite heating block. A top plate, suspended above the substrate, ensured laminar flow of the aerosol for uniform deposition. This entire setup was housed inside a quartz tube. The substrates were heated at different temperatures (480–580 °C). The precursor solution was nebulized using a piezo ultrasonic atomizer from Johnson Matthey, operating at a frequency of 1.6 MHz to generate a mode droplet size of 3 µm. The precursor solution and nebulizer were maintained at ambient laboratory temperature during deposition and no external heating or active temperature stabilisation was applied.
The generated aerosol mist of the precursor solution was transported by a carrier N2 gas into the deposition chamber. The entire setup was placed in a fume cupboard. The precursor solution was completely misted in 30 ± 5 minutes to deposit the film, after which the substrate was left to cool in reactor under continuous flow of N2 gas and removed when the temperature of the reactor reached below 50 °C.
X-ray photoelectron spectroscopy (XPS) analyses of the films were conducted using a Thermo Scientific spectrophotometer equipped with a monochromatic Al-Kα radiation source. The chemical compositions were determined through high-resolution surface scans of the Sb 3d, O 1s, and C 1s regions. These scans were carried out with a pass energy of 50 eV and a spot size of 400 µm. Identification and quantification of peaks were performed using Avantage Thermo Fisher scientific software and plotted using OriginPro. The XPS spectra were quantified to determine the O/Sb ratios in Sb2O3 films. Binding energies of the peaks were calibrated relative to the adventitious carbon peak at 284.5 eV to correct for charging effects.
UV-Vis transmittance and reflectance spectra were recorded using Shimadzu 3600i plus spectrometer over a wavelength range of 250–2500 nm. The band gap of films was calculated from acquired spectra in OriginPro using Tauc plot method.
Scanning electron microscopy (SEM) was employed to examine the surface morphology of the films in top-down configuration and to estimate the film thickness using cross-sectional configuration. JEOL IT-700HR was used to analyse the samples. All measurements were performed at an accelerating voltage of 5–15 kV. Prior to imaging, the films were vacuum sputtered with a thin layer of carbon to minimize surface charging.
Electrical properties were measured by Hall effect measurements using a four-point probe in van der Pauw geometry on an Ecopia HMS-3000 setup. Recordings were performed on 1 cm2 samples with an applied current of 1 nA–1 µA under a calibrated magnetic field of 0.58 T. It provided the values of resistivity, mobility, and carrier concentration.
Finally, static water contact angles were measured using a Krüss DSAE droplet shape analyser employing the sessile drop technique. Measurements were carried out at ambient conditions using 5 µL droplets of ultrapure water. The contact angle values were obtained by fitting the droplet profiles from the recorded images using the Krüss DSAE analysis software.
The films were deposited on float glass substrates positioned on the graphite block in the horizontal-bed AACVD reactor. All films were deposited on the bottom substrate plates. Deposition parameters including substrate temperature (480–580 °C) and precursor molarity (0.4–0.6 M) were systematically varied to study their influence on the structural, morphological, optical, and electrical properties of the resulting films. The specific deposition conditions for each sample are summarized in Table 1.
| Film | Deposition temp. (°C) | Antimony(III) ethoxide (mL) | Anhydrous toluene (mL) | Precursor molarity (M) |
|---|---|---|---|---|
| S1 | 480 | 3.1 | 24 | 0.4 |
| S2 | 500 | 3.1 | 24 | 0.4 |
| S3 | 540 | 4.7 | 24 | 0.6 |
| S4 | 580 | 4.7 | 24 | 0.6 |
A wider range of deposition temperatures was initially explored during preliminary optimisation experiments. Films deposited below 450 °C were found to be very thin and powdery, making them unsuitable for further analysis. The upper deposition temperature was also limited by the AACVD setup and glass substrate stability, as temperatures above ∼590 °C were not suitable for the substrates used in this study. The selected temperature range therefore enabled systematic investigation of the transition from phase-pure Sb2O3 to mixed Sb2O3/Sb films at elevated deposition temperatures.
To confirm reproducibility, all films were deposited three times under identical conditions. The reproduced samples showed negligible variation in optical transparency and electrical conductivity. The Sb2O3 films (S1–S2) adhered strongly to the substrate, passing the Scotch tape test. No film removal or edge peeling was observed after 20 test cycles, confirming the mechanical stability of the coatings. However, sample S3 and S4 deposited at 540 °C and 580 °C showed the formation of metallic Sb crystallites due to partial thermal reduction. The bright, faceted Sb crystallites were scratchable with a metal spatula, indicating poorer adhesion of the metallic phase compared to the oxide matrix. This reduced stability likely resulted from differences in thermal expansion and weaker interfacial bonding of metallic Sb to the underlying oxide and glass substrate.
All the films were stored in ambient air for 14 months, during which no measurable change in their electrical or optical properties was observed. This long-term stability indicates that the films properties were stable under normal atmospheric conditions.
The normalized powder X-ray diffraction (XRD) patterns of the as-deposited films (S1–S4) are shown in Fig. 1. The reference diffraction patterns of senarmontite (cubic, ICSD No. 1011201) and Sb (metal, ICSD No. 2310880) phases were obtained from the Inorganic Crystal Structure Database (ICSD) and plotted using the VESTA software package38 (Fig. 1). Films S1 and S2 were found to be phase-pure and crystallized in the senarmontite phase, which is the most reported polymorph for Sb2O3 when deposited below 600 °C.10 In contrast, S3 and S4 showed additional diffraction peaks corresponding to metallic Sb, indicating partial reduction of Sb3+ species at higher deposition temperatures (540–580 °C). This confirmed that partial thermal reduction of Sb2O3 occurred at the highest deposition temperature, leading to the formation of mixed Sb2O3–Sb phases. Such oxide-to-metal transitions at high deposition temperatures have previously been reported when antimony(III) ethoxide was used as a precursor.39,40
![]() | ||
| Fig. 1 Collated XRD diffractograms of Sb2O3 thin films (S1–S4) deposited by AACVD under the conditions listed in Table 1. Reference patterns of cubic senarmontite Sb2O3 and metallic Sb are included for comparison. | ||
While Sb2O3 is widely reported to exhibit excellent thermal stability as an oxide, the partial formation of metallic Sb observed at higher deposition temperatures in this study arises from the specific AACVD environment and precursor chemistry. At high deposition temperatures, the decomposition of antimony(III) ethoxide in an organic solvent can generate locally reducing conditions, promoting partial reduction of Sb3+ to Sb0. The use of toluene solvent together with N2 carrier gas may further contribute to a relatively low oxygen partial pressure within the AACVD reactor, favouring the formation of metallic Sb during deposition.
The crystallite sizes (D) were calculated using the Scherrer formula,41 as shown in eqn (1):
![]() | (1) |
The crystallite size was found to grow from 31.3 nm to 45.0 nm as the substrate temperature increased from 480 °C (S1) to 540 °C (S3). This trend can be attributed to the enhanced surface and grain boundary mobility at elevated temperatures, which can facilitate the migration and merging of adjacent grains.42 As a result, smaller crystallites can merge to form larger ones, leading to an increase in crystallite size. However, for S4 (580 °C, 0.6 M), the crystallite size decreased to 29.0 nm. This reduction is likely due to excessive nucleation and partial reduction of Sb2O3 to metallic Sb at elevated temperature and higher precursor concentration. The coexistence of oxide and metallic phases can introduce lattice strain and grain boundary pinning, which can disrupt uniform grain growth and lead to smaller crystallites.
The film thickness values were found to be between 440–1020 nm, as shown in Table 3. It is evident that the film was thinner at lower deposition temperature which may be attributed to the incomplete decomposition of the precursor. At higher temperature the decomposition became more complete, and the films grew thicker. The film thickness was found to increase linearly with substrate temperature. Increasing the precursor concentration also yielded thicker film which can be due to the greater flux of reactive species reaching the substrate.
The increase in film thickness with deposition temperature and precursor concentration significantly influenced the resulting film properties. Thicker films deposited at higher temperatures showed lower optical transmittance and improved electrical conductivity. In contrast, thinner films deposited at lower temperatures remained more transparent but showed higher resistivity. These results demonstrate that film thickness played an important role in controlling the balance between optical transparency and electrical transport in Sb2O3-based films.
X-ray photoelectron spectroscopy (XPS) was used to investigate the chemical state and composition of the antimony oxide films. XPS survey spectrum of S2, shown in Fig. S1, indicated the presence of only Sb, O, and C peaks. The Sb 3d XPS spectra of S1–S4 are shown in Fig. 2a. Within the energy range of 520–540 eV, spectra exhibited complex features due to the overlapping of O1s and Sb 3d5/2 photoemission lines at around 530 eV.43 As a result, in most of the reported literature, the complete deconvolution of O1s and Sb 3d5/2 peaks have not been conducted.44,45
![]() | ||
| Fig. 2 High-resolution XPS spectra of Sb2O3 thin films S1–S4 showing (a) Sb 3d and (b) C 1s regions. | ||
In this study, spectral deconvolution was carried out while considering spin–orbit coupling, setting the Sb 3d3/2/Sb 3d5/2 area ratio to be 2
:
3, and fixing the splitting energy of the doublet at 9.39 eV.43 A deconvoluted Sb 3d spectrum is shown in Fig. S2. This method allowed for the independent quantification of the contributions from the Sb 3d and O 1s peaks.46 The Sb 3d3/2 peak, which is free from O 1s overlap, was used to determine the antimony oxidation state. Sb(III) reportedly appears at 539.3–539.7 eV and Sb(V) at 540.3–540.8 eV.44,47,48 In all the studied samples (Fig. 2a), the Sb 3d3/2 peak was located at approximately 539.5 eV, which confirmed the presence of Sb(III) in all the films. Moreover, no contribution from Sb(V) was found in any of the films, which further confirmed the deposition of Sb2O3.
The XPS spectrum of sample S3 and S4 revealed the coexistence of Sb2O3 and metallic Sb species. The Sb 3d5/2 peak of metallic antimony was observed at approximately 528 eV, consistent with previously reported values for Sb0 in the literature.49 The simultaneous presence of these features confirmed partial thermal reduction of Sb2O3 at 580 °C, consistent with the XRD observations. This mixed chemical state suggested that metallic Sb crystallites formed within the Sb2O3 matrix during high-temperature depositions, resulting in a composite oxide–metal film.
The O 1s peak consisted of two components, corresponding to distinct oxygen bonding states present in the samples. The deconvolution of O 1s peaks in Sb 3d spectrum is shown in Fig. S2. The first peak centred at approximately 530.3 eV was assigned to lattice oxygen corresponding to Sb–O bonds. Similar O 1s signals at approximately same binding energies have been previously reported and linked to the metal–oxygen bond.45,50–54 However, the higher binding energy peak at around 532.2 eV was attributed to surface hydroxyl group (–OH).55,56
The C 1s spectra of all the films are shown in Fig. 2b. C 1s was deconvoluted into three components. The peak at 284.8 eV corresponds to C–C from adventitious hydrocarbons. The component at 286.5 eV is associated with C–O species (alcohol/ether or alkoxy groups). The high-binding-energy peak at 288.9 eV is due to O–C
O groups.57
XPS spectra were quantified to determine the O/Sb atomic ratio and the carbon content (at%), as shown in Table 3. The films contained 4–26 at% carbon, depending on the deposition parameters. The brown colouration and reduced transparency are consistent with elevated carbon content. Use of toluene as the solvent likely increased carbon incorporation into the films. Hassan et al. reported a similar brown tinge when toluene was used and attributed it to carbon contamination (>5 at% by EDS).58 In a broader solvent study on AZO, Potter et al. reported higher carbon contamination when toluene was used instead of methanol.59 In the present study, methanol was avoided due to precursor instability, and toluene was selected for its stable solubility of Sb(OC2H5)3. Brown colouration of the films can also due to appearance of metallic Sb0 in the films.
Although residual carbon was detected by XPS, no direct correlation between carbon content and enhanced conductivity was observed. In fact, higher carbon concentration is reported to introduce localized trap states at grain boundaries, which can limit carrier mobility.60,61 The significant reduction in resistivity observed for S3 and S4 is therefore primarily attributed to the formation of metallic Sb0 inclusions rather than carbon-related conduction pathways.
The XPS depth profiling (Fig. 3) spectra of sample S4 are summarized in Table 2. At the surface, the C 1s spectrum was dominated by the C–C/C–H component (79.1 at%), accompanied by minor oxidized carbon species, C–O–C (12.9 at%) and O–C
O (7.9 at%). The carbonate-related O–C
O contribution decreased steadily with sputtering time and became less than 6.0 at% beyond 100 s, confirming that these species are confined to the near-surface region and arise from air exposure or CO2 adsorption rather than bulk incorporation. In contrast, the C–C intensity remained nearly constant throughout the etching process, consistent with hydrocarbon contamination. XPS depth profiling of S1 (Fig. S3) also showed a systematic decrease in the C–C, C–O–C, and O–C
O contributions with depth profiling time, indicating that the carbon-related species were primarily confined to the near-surface region.
![]() | ||
| Fig. 3 XPS depth profiling spectra of sample S4 showing the evolution of (a) Sb 3d and (b) C 1s regions with increasing sputtering time. | ||
O), oxygen (O–Sb, Sb–OH), and antimony (Sb0, Sb3+) species obtained from XPS depth profiling of S4. Samples (S1–S4) correspond to the deposition conditions listed in Table 1
| Etching time (s) | C–C (rel. at%) | C–O–C (rel. at%) | O–C O (rel. at%) |
O–Sb (rel. at%) | Sb–OH (rel. at%) | Sb0 (rel. at%) | Sb3+ (rel. at%) |
|---|---|---|---|---|---|---|---|
| 0 (Surface) | 79.1 | 12.9 | 7.9 | 56.9 | 43.1 | 36.9 | 63.1 |
| 50 | 81.9 | 12.2 | 5.9 | 56.9 | 43.1 | 29.7 | 70.3 |
| 100 | 81.6 | 12.8 | 5.6 | 56.1 | 43.9 | 27.2 | 72.8 |
| 150 | 82.7 | 12.4 | 4.9 | 55.1 | 44.9 | 27.5 | 72.5 |
| 200 | 81.8 | 12.7 | 5.5 | 55.1 | 44.9 | 27.9 | 72.1 |
| Film | Film thickness (±50 nm) | Resistivity (×10−1 Ω cm) | Mobility (cm2 V−1 s−1) | Carrier conc. (cm−3) | Sheet resistance (Ω □−1) | F.O.M. (×10−6 Ω−1) | Band gap (eV) | Tλ550 (%) | Tλ380–750 (%) | Tλ750–1400 (%) | Tλ1400–2500 (%) | C (at% from XPS data) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| S1 | 440 | 2100 ± 200 | 8 | 3.7 × 1015 | 4 800 000 |
0.02 | 3.63 ± 0.03 | 79.2 | 78.6 | 82.8 | 86.7 | 4.0 |
| S2 | 490 | 590 ± 50 | 10 | 1.0 × 1016 | 1 200 000 |
0.07 | 3.72 ± 0.03 | 77.6 | 77.2 | 82.9 | 85.6 | 7.0 |
| S3 | 750 | 8.2 ± 0.8 | 14 | 5.5 × 1017 | 11 000 |
1.98 | 3.59 ± 0.04 | 68.2 | 67.7 | 75.4 | 78.9 | 26.0 |
| S4 | 1020 | 1.1 ± 0.1 | 11 | 5.2 × 1018 | 1100 | 0.03 | 3.47 ± 0.06 | 36.1 | 36.2 | 54.2 | 63.7 | 21.0 |
The lattice oxygen (O–Sb) and surface hydroxyl oxygen (Sb–OH) maintained a nearly constant ratios throughout the sputtering time. However, the relative fraction of Sb0 decreased from 36.9 at% to 27.9 at%, while Sb3+ increased from 63.0 at% to 72.1 at% with sputtering time. These results indicated that surface carbonates were confined to the outermost layer, while the OH-related oxygen species persisted with depth. The sustained presence of the Sb–OH component suggested that hydroxyl-related oxygen remained stable even after surface carbon removal.
The SEM images (Fig. 4) show a systematic evolution in surface morphology of the Sb2O3 films with changes in deposition temperature and precursor concentration. At lower deposition temperature (S1, 480 °C) the film showed small spherical grains. The film coverage improved at 500 °C (S2), producing smoother and compact surface composed of spherical grains of around 100 nm. This morphology reflects limited surface diffusion and slow grain growth, typical of kinetically controlled oxide deposition at lower temperatures.33,62,63 The high resistivity and low mobility of these samples can be attributed to their less dense surface morphology.
![]() | ||
| Fig. 4 Top-view SEM images of Sb2O3 thin films deposited under the conditions listed in Table 1: (a) S1 (b) S2 (c) S3 (d) S4, showing the evolution of surface morphology with increasing deposition temperature and precursor concentration. | ||
With increasing precursor concentration (S3, 540 °C, 0.6 M), the grains became larger, and the surface roughness increased due to accelerated particle agglomeration and presence of Sb0 as suggested by XRD and XPS. The improved grain connectivity and densification in this sample along with the presence of metallic antimony was consistent with the observed reduction in resistivity and higher carrier concentrations. While the increased surface roughness and presence of Sb0 contributed to lower optical transmittance.
At the highest deposition temperature (S4, 580 °C, 0.6 M), the surface morphology changed markedly. The film exhibited large, faceted grains with bright metallic contrast, indicating partial reduction of Sb2O3 to elemental Sb. These metallic crystallites were dispersed within the oxide matrix but did not form a continuous conducting network, resulting in a mixed-phase structure. The presence of these metallic domains explained the sharp decrease in resistivity compared to the oxide samples. However, the film still retained semiconducting characteristics due to the non-percolating nature of the Sb inclusions.
Large, faceted crystallites with well-defined polygonal geometry dominated the surface, consistent with the metallic phase observed in XRD. These micrometre-sized grains exhibited sharp edges and smooth faces, typical of the trigonal crystal structure of elemental antimony. This confirmed the thermal reduction of Sb2O3 to metallic Sb at higher deposition temperature. The loosely packed crystallites suggested localized metallic nucleation and growth within the antimony(III) oxide matrix. This observation is fully consistent with the XRD and XPS results that showed both Sb0 and Sb3+ states.
UV-Vis-NIR spectroscopy was used to evaluate the optical properties of the films. Fig. 5 shows the transmittance and reflectance spectra recorded in the wavelength range of 250–2600 nm. The average transmittance progressively decreased with increasing deposition temperature and precursor concentration.
![]() | ||
| Fig. 5 Optical transmittance (T) and reflectance (R) spectra together with photographs of Sb2O3 films S1–S4. | ||
Samples S1 (480 °C, 0.4 M) and S2 (500 °C, 0.4 M) showed the highest transparency, with visible-range average transmittances (Tλ380–750) of approximately 78.6% and 77.2%, respectively. Their high transparency is consistent with their phase-pure Sb2O3 composition, smaller grain sizes, and the absence of metallic antimony. The relatively low film thicknesses (440–490 nm) can also contribute to reduced scattering and absorption losses. The obtained transmittance values are in good agreement with the reported literature of Sb2O3 films.31,64,65 However a transmittance of 47% was reported for Sb2O3 film deposited by spray pyrolysis.66 The decline in transparency with increasing temperature and concentration can be correlated with higher film thickness, increased sub-bandgap absorption due to defect states, and enhanced light scattering from larger grain structures at elevated temperatures.31,64,65
However, the transmittance decreased significantly for samples S3 (540 °C, 0.6 M) and S4 (580 °C, 0.6 M). These films exhibited noticeably reduced visible transmittance values of 50% (S3) and 37% (S4). This loss of transparency might be linked to the combination of structural and compositional factors. The substantially greater film thicknesses can contribute towards the optical scattering. This can also be linked with the larger grain sizes and higher surface roughness as observed in SEM images. Additionally, the formation of metallic Sb inclusions (confirmed by both XRD and XPS) can introduce free-carrier absorption and plasmonic losses that further suppress transmittance.
The optical band gap (Eg) values were calculated using the Tauc method,67 and shown in Fig. 6. In this method, the linear region of a plot of (αhν)2 versus energy Eg (eV) near the high-energy absorption edge was extrapolated to the x-axis to provide an estimate of the band gap energy. In the Tauc equation (eqn (2)),67 α represents the linear absorption coefficient, A is a constant, and hν denotes the photon energy, where h is Planck's constant and ν is the frequency of the incident light.
| (αhν)2 = A(hν − Eg) | (2) |
The band gap was found to vary systematically with deposition temperature and precursor concentration, decreasing from 3.72 eV to 3.47 eV across the series. Samples S1 (3.63 eV) and S2 (3.72 eV), both deposited at lower temperatures (480–500 °C) and at lower precursor concentration, exhibited the highest band gap values. Their wider band gaps are consistent with phase-pure senarmontite Sb2O3, relatively low defect densities, and the absence of metallic Sb.
A notable decrease in the band gap was observed for S3 (3.59 eV) and S4 (3.47 eV). These films were deposited at higher temperature and higher precursor concentration and both XRD and XPS confirmed the presence of Sb0 species in these samples.
This reduction in the apparent optical band gap with increasing film thickness can be associated with enhanced structural disorder and partial incorporation of metallic Sb at higher deposition temperatures. Thicker films also exhibited larger grain structures and increased optical scattering, which can influence the absorption edge.
The appearance of metallic Sb inclusions in S3 and more prominently in S4 could led to enhanced free carrier absorption and localized plasmonic effects that can lower the band gap values. The progressive reduction from 3.72 eV (S2) to 3.47 eV (S4) therefore reflects the combined influence of increased defect density and metallic antimony incorporation at elevated deposition temperatures. A decrease in band gap from 3.98 eV to 3.71 eV when the film thickness was increased from 0.20 µm to 1.10 µm was previously reported for Sb2O3 films deposited by thermal evaporation method.30 Although metallic Sb inclusion was observed in S3 and S4, the oxide matrix remained predominantly Sb2O3. Therefore, the fundamental band gap of these samples was found to be close to the reported Sb2O3 films.65
The haze values in the visible region (380–750 nm), shown in Fig. 6, increased progressively from S1 (7.8%) to S4 (11.9%), indicating a rise in light-scattering behaviour with higher deposition temperature. The low haze of S1 and S2 reflects their smoother, compact microstructures, which minimise scattering. In contrast, S3 shows slightly higher haze (9.07%) due to increased grain growth and surface roughening at 540 °C. The highest haze in S4 corresponds to the formation of large, faceted grains and metallic Sb domains, which strongly scatter incident light. Overall, haze evolution confirms that microstructural coarsening and partial oxide reduction at elevated temperatures degrade optical clarity.
The Sb2O3 thin films deposited via AACVD exhibited n-type semiconducting behaviour across all deposition conditions. A clear trend was observed in the electrical transport properties as a function of deposition temperature and precursor concentration. The resistivity decreased systematically with increasing temperature and precursor molarity.
At lower deposition temperatures (S1: 480 °C and S2: 500 °C), the films displayed high resistivity and low carrier concentrations. This behaviour is consistent with their compact but fine-grained microstructure. XPS and XRD showed that those films were purely Sb2O3 with no inclusion of Sb0. The carbon content of 4–7% originated from incomplete combustion of toluene, leading to carbonaceous residues at grain boundaries that act as trap centres and scattering sites, thereby limiting carrier mobility.61 The properties of previously reported Sb2O3 films deposited by various techniques are summarized in Table 4. Resistivity values lower than those obtained in this work have been reported for Sb2O3 films prepared by thermal evaporation30 and spray pyrolysis.64
| Film | Resistivity (Ω cm) | Thickness (nm) | Transmittance (%) | Band gap (eV) | Precursor | Solvent | Deposition method | Ref. |
|---|---|---|---|---|---|---|---|---|
| Sb2O3 | 1.0 × 10−2 | 298 | 74 | 3.6 | SbCl3 | Ethanol | Spray pyrolysis | 64 |
| Sb2O3 | — | — | 70 | — | SbCl5 | Ethyl acetate | APCVD | 31 |
| Sb2O3 | — | — | — | — | Sb(BuO)3 | — | APCVD | 39 |
| Sb2O3 | 6.5 × 10−2 | 300 | 85 | — | Sb2O3 powder | — | Thermal evaporation | 30 |
| Sb2O3 | — | — | 47 | 4.0 | SbCl3 | Water | Spray pyrolysis | 66 |
| Sb2O3 | 26 × 102 | 189 | 75 | 3.43 | SbCl3 | Acetic acid | Spray pyrolysis | 65 |
A noticeable reduction in resistivity was observed for films deposited at 540 °C (S3). A decrease in resistivity values with the increase in temperature has also been widely reported.65,68 While excess carbon content can introduce additional trap states, the conductivity of S3 improved by several orders of magnitude compared to S1 and S2. This increase can be attributed to the first appearance of metallic Sb0 features (detected in both XRD and XPS), which suggests partial thermal reduction of Sb2O3. Although metallic Sb might not have formed a continuous conductive pathway at this stage, its presence increased local carrier density that led to enhanced electrical transport.
The film deposited at the highest temperature (S4: 580 °C) exhibited the lowest resistivity among the series. This can be linked to its greater film thickness, larger and faceted grain morphology, and extensive formation of metallic Sb inclusions. XPS depth profiling confirmed the coexistence of Sb3+ and Sb0 throughout the film, with the Sb0 fraction decreasing with etching depth. These metallic domains acted as electron-rich regions within the oxide matrix and significantly increased carrier concentration. Despite the large fraction of Sb0, the conductivity did not reach metallic behaviour, indicating that the Sb0 crystallites remained non-percolating. The mobility of S4 remained modest (∼11 cm2 V−1 s−1), likely due to enhanced scattering at interfaces between Sb2O3 and metallic Sb, as well as increased microstructural roughness, consistent with SEM observations.
Superior transport properties such as high carrier concentration are often achieved at the cost of reduced optical transmittance. To assess the optoelectronic performance of TCO materials the figure of merit (FOM) proposed by Haacke in 197669 was employed in this study. The FOM (Φ) is defined in eqn (3).
![]() | (3) |
The low-temperature films S1 and S2 exhibited very low FOM values (0.02 × 10−6 and 0.07 × 10−6 Ω−1, respectively), reflecting their high sheet resistances despite high visible transmittance. A pronounced improvement was observed for S3 (540 °C, 0.6 M), which achieved the highest FOM of 1.98 × 10−6 Ω−1 due to its significantly reduced sheet resistance (11
000 Ω □−1) and moderate transparency. In contrast, S4 exhibited a markedly reduced FOM (0.03 × 10−6 Ω−1), as its substantial loss in transmittance (caused by large metallic Sb crystallites formed at 580 °C) outweighed the benefits of its low sheet resistance. These trends demonstrate that although higher deposition temperature improved conductivity, excessive thermal reduction diminished transparency. Such oxide-to-metal transitions at high deposition temperatures have previously been reported when antimony(III) ethoxide was used as a precursor.39,40
Additional data are available from the corresponding author upon reasonable request.
| This journal is © The Royal Society of Chemistry 2026 |