Introduction

In the past few decades, transition metal oxides have attracted tremendous attention in scientific and technological research due to their unique electrical, optical, and mechanical properties1,2,3. Among transition metal oxides, TiO2 plays a vital role because of its chemical stability, and inexpensive and non-toxic signatures. TiO2 maintains efficient charge transport capability even with a relatively wide band-gap (~3–3.35 eV), and high specific surface area in many of its nanoscale morphologies4,5,6. The anatase phase of TiO2 is largely preferred in photovoltaics, solar cell and gas sensing applications, because of its superior electron mobility and chemical stability compared with rutile and brookite phases7. Gopal et al., reported the Ti-O phase diagram and concluded that anatase TiO2 is more stable than the rutile phase at room-temperature (27 °C) and atmospheric pressure (760 mmHg)8. During the past several decades, control in the formation of TiO2 film with specific morphology for the purpose of improving solar cells and low-temperature gas sensor performance has proven the benefits of TiO2. Extremely great number of methods developed in literature for the preparation of TiO2 nanocrystalline films containing nanowires9, nanotubes10, nanorods11, and hollow microspheres12, etc., morphologies included spray pyrolysis13, sol-gel14, wet chemical15, electrodeposition16, magnetron sputtering17, and chemical vapor deposition18, etc. These methods are being undertaken to direct synthesize TiO2 nanocrystalline films on conducting/non-conducting substrates for both perovskite solar cell and gas sensor applications. The vast majority of reported methods involve either long deposition times or complicated processing steps, with lack of clarity in the roles of some of the synthetic steps in these growth processes. The valuable works motivated us to handle this challenge in the synthetic route and applications potential, such as the use of successive ionic layer adsorption and reaction (SILAR) technique. In spite of its simplicity, SILAR has a number of advantages: unlike physical technique, SILAR does not require high quality target and/or substrate nor it does require vacuum at any stage; the deposition rate and the thickness of the film can be easily controlled over a wide range by changing the deposition cycles; there are virtually no restrictions on substrate material, dimensions or its surface profile; moreover, it is convenient for large area direct deposition, which helps to enhance interconnection and adherence of film with superior properties19.

In this work, we unveil SILAR-mediated binder-free chemical synthesis of phase pure anatase TiO2 films directly grown on flourine-tin-oxide (FTO) and soda-lime glass (SLG) substrates for perovskite solar cell and chemical gas sensor applications. By tuning the SILAR cycles, anatase TiO2 nanocrystalline films grown at low-temperature (90 °C) on FTO were compatible with a sequentially deposited and full coverage perovskite layer20. As a result, the perovskite solar cells based on this TiO2 ETL demonstrate a PCE of ~10%, which is the one of the best figures for directly grown TiO2 ETL-based perovskite solar cells prepared at such low temperatures. Secondly, this anatase TiO2 nanocrystalline films on SLG  substrate were employed as a sensor material to investigate selectivity for various volatile oragnic compunds (VOCs) such as ammonia, petrol, formaldehyde, ethanol, and acetone. The transient response, repeatability and response time for ammonia sensing were also determined as function of temperature and gas concentration. Finally, the stability of TiO2 film sensors with potential commercial viability is demonstrated.

Results and Discussion

Surface morphology and elemental analysis studies. The FE-SEM image shown in Fig. 1a demonstrates the spherical surface morphology of 15 L, optimized (discussed later), SILAR TiO2. Inset of Fig. 1a shows the cross-sectional view of TiO2 ETL which is relatively uniform of ~149 nm in thickness. The elemental surface composition analysis carried out by EDX (Fig. 1b) confirms stoichiometric quantities of Ti and O in for TiO2. The lattice parameters were measured from the HR-TEM images and the positions of the main diffraction peaks in the SAED patterns (Fig. 1c,d). The, 0.36 nm distance between two adjacent lattice planes was in accordance with the (101) reflection plane of anatase TiO2 (JCPDS card no. 21-1272)21. Being of lower surface energy, the (101) crystal face of TiO2 was more stable than the other facets22, and our HR-TEM images also demonstrated the strongest ring pattern for (101) plane than others in SAED spectrum. Moreover, the diffraction bright and sharp ring pattern confirms nanocrystalline nature of SILAR-mediated TiO2 film. The XRD pattern of SILAR-mediated TiO2 film on SLG substrate, which avoids FTO reflection contributions, is shown in Fig. 2a where well-defined reflections confirmed tetragonal crystal structure of anatase TiO2 (JCPDS card no. 21-1272)23. No other impurity/phase peak was noticed, indicating high phase purity of the product. The various structural parameters such as grain-size, dislocation densities, and texture coefficient values, determined from XRD pattern of TiO2 have tabulated in Table S1† (Electronic Supporting Information, ESI). Furthermore, the dislocation density and texture coefficient values investigated from every diffraction peak are displayed in Table S1†. From Table S1†, the maximum texture coefficient of 2.011 was associated with (101) plane, which is also the preferred growth direction. The Williamson and Hall (W-H) plot (Fig. 2a inset) and the peak broadening method were utilized to estimate the average crystallite size as well as the micro-strain in the TiO2 particles. The positive values of slope and micro-strain (ε = 12.6 × 10−3) revealed an existence of tensile strain in TiO2 film. The average crystallite-size measured from the intercept value of W-H plot was 7 nm (±1 nm) which was in close agreement with the average value (7.2 nm) obtained from Scherrer’s formula and HR-TEM analysis i.e. 8 (±2) nm. The micro-strain was assigned to a relative change in interplanar distance which shifted diffraction peak positions, the difference in crystallite-size values (measured by W-H plot and Scherrer’s formula) and dislocation densities in TiO2 crystals24. The Raman spectrum of anatase TiO2 film (Fig. 2b) confirmed tetragonal space group D4th (I41/amd) having six Raman active modes (1A1g + 2B1g + 3Eg). The Raman peaks located at 143 (Eg), 199 (Eg), 396 (B1g), 514 (A1g), and 636 cm−1 (Eg) were assigned to the anatase TiO225. In addition, as-presented in Fig. 2c,d, XPS analysis was used to confirm Ti4+ at TiO2 film surface. The XPS spectrum of TiO2 (Fig. S1†) showed Ti, O and C core level photoemissions, and there was no trace of any other impurity peak. The carbon peak at 284.8 eV was assigned to residual carbon in the sample, and adventitious carbon. The core-level emissions at binding energies of 458.9 and 464.6 eV are of Ti 2p3/2 and Ti 2p1/2, respectively26. The peak with a binding energy of 530 eV is from O1s, confirming presence of oxygen anion in the lattice (Ti-O-Ti)27. The BET-derived values of 121 m2/g specific surface-area and 6.54 nm average pore-size were confirm the mesoporous nature of the TiO2 material. A H1-type hysteresis loop supported for mesoporosity in anatase TiO2 film (obtained using scratched film powder)28. The adsorption-desorption isotherm and pore-size distribution plots for TiO2 are shown in Fig. S2† and Table S2†. Since nanoparticles in TiO2 film were mesoporous in nature, with a high surface area and small pore-size, and uniformly distributed across the substrate surface, we investigated their potential for perovskite solar cell as well as efficient and selective gas sensor applications.

Figure 1
figure 1

(a) FE-SEM plan-view image, (b) EDX pattern, (c) HR-TEM image, and (d) SAED pattern of SILAR-grown TiO2 film.

Full size image
Figure 2
figure 2

(a) XRD pattern (inset shows the W-H plot), (b) Raman spectrum of the TiO2 film, (c) Ti2p and (d) O1s core-level photoemission spectra.

Full size image

Energy band positions

To verify the energy level alignment, UPS measurements were performed, and the results are presented in Fig. 3a–c. The high binding energy cut-off (Ecut-off) value of the TiO2 nanostructured film was determined to be 8.40 eV. The highest occupied molecular orbital (HOMO) energy onset value, the binding energy onset (Eonset), relative to the Fermi level of Au (set at 0 eV), was measured to be −5.23 eV. The valence band minimum (VBM) was calculated according to eq. 1:

$${\rm{VBM}}={\rm{h}}{\rm{\nu }}-(\mathrm{Ecut} \mbox{-} \mathrm{off}-{\rm{Eonset}})$$
(1)

where, h is Planck’s constant and ν is the frequency of vibration of light, corresponding to −7.58 eV for TiO2. The optical band gap energy (Eg) of the TiO2 film was estimated using eq. 2:

$$({\rm{\alpha }}{\rm{h}}{\rm{\nu }})2={\rm{A}}({\rm{h}}{\rm{\nu }}-{\rm{Eg}})$$
(2)

where, A is the proportionality constant and α is the absorption coefficient. Fig. S3† presents UV-vis spectrum of TiO2 film as an ETL where the Eg value was obtained by extrapolating the linear portion of the curve near the onset of the absorption edge to the energy axis (inset of Fig. S3†). The Eg value was determined to be 3.47 eV for the TiO2, which was close to the reported value29. Using the optical band gap, the conduction band minimum (CBM) energy level was also evaluated using eq. 3:

$${\rm{CBM}}={\rm{VBM}}+{\rm{Eg}}$$
(3)
Figure 3
figure 3

UPS spectra for different energy edge ranges: (a) cut-off region, (b) onset region of TiO2 film. (c) Schematic energy diagram of FTO/TiO2 ETL/perovskite/spiro-OMeTAD/gold solar cell device showing the energy level of each component calculated from UPS analysis.

Full size image

The calculated CBM level for TiO2 was −4.0 eV. To obtain a clear understanding of the electron transfer pathway through the ETL, an energy band diagram of the FTO/TiO2 ETL/perovskite/spiro-OMeTAD/gold solar cell was proposed based on their appropriate band levels.

Photovoltaic performance

The UV-vis absorption spectrum of the perovskite absorber layer deposited on TiO2 ETL was obtained (Fig. 4a). The perovskite layer on TiO2 ETL demonstrates a wide absorbance in the visible region (centered at ~750 nm). The nanocrystallinity, smaller particle-size, mesoporous signature and higher surface area of the prepared SILAR-based anatase TiO2 film increased the overall perovskite adsorption, which was corroborated from the enhanced UV-Vis absorption. The performance of the perovskite solar cell was tested under illumination of simulated AM1.5 G simulated solar light (100 mW cm−2), and electronic parameters are listed in Table 1. The TiO2 ETL exhibited perovskite solar cell performance with an open circuit voltage (VOC) of 0.81 V, short-circuit current density (JSC) of 18.0 mA cm−2, and a fill factor (ff ) of 66.3% corresponding to a PCE (ƞ%) of 9.7% (Fig. 4b). To optimize the anatase TiO2 ETL thickness for best perovskite solar cell performance, the TiO2 ETL thickness was increased by changing SILAR cycles as 10, 15 and 20 L during deposition process, and detailed solar cell performances are shown in Fig. S4† and Table S3†, respectively. For the 10 L SILAR deposition, the FTO/TiO2 ETL/perovskite/spiro-OMeTAD/gold device exhibited 8.5% PEC. Whereas, the device assembled with 20 L SILAR deposition had a PCE of 7.0%, and the best performance i.e. 9.7% was achieved for device fabricated for 15 L SILAR deposition at 8.7 mV s−1 scan rate in reverse scan direction. The comparative study of our TiO2 ETL with different synthesis methods for perovskite solar cell application was demonstrated in Table S4†30,31,32,33,34,35, signifying potentiality of proposed chemical method followed developed phase-pure TiO2 nanostructure in perovskite-based solar cells, which can be applied in developing other metal oxide nanostructured thin films too. To further verify the cell performance, EQE spectra of TiO2 ETL was obtained and presented in Fig. 4c. The TiO2 ETL device exhibited higher EQE value at 520 nm (~84.1%). Impedance spectroscopy in Fig. 4d (Nyquist plot) for the device conducted under 1 sun illumination at an applied bias voltage of 0.2 V, revealed two distinct semicircles (one in the high-frequency range and another in the low-frequency range) in the measured frequency range of 0.1 Hz to 100 kHz. The series resistance (RS) is related to the wiring and FTO substrate measured in the high-frequency region. R1 is the interface resistance of the counter electrode within the equivalent circuit representation, and CPE1 is its capacitance. In the Nyquist plot, the main arc is responsible for the TiO2 ETL/perovskite/spiro-OMeTAD interfaces as expressed by a combination of the interface recombination resistance (R2) and chemical capacitance (CPE2)36. The PL measurement confirms the photo-induced charge transfer and charge recombination loss in optolectrochemical devices37. Figure 5a presents the PL spectra of the glass/perovskite and glass/TiO2/perovskite layer configurations. The PL intensity of the glass/TiO2/perovskite film was less than that of the glass/perovskite film, indicating that the charge transfer effectively occurs prior to non-radiative carrier recombination at the interface, and this improved the electron extraction rate from the perovskite absorber layer. The TRPL results of the perovskite films deposited on glass/perovskite and glass/TiO2/perovskite are shown in Fig. 5b. The perovskite film deposited on glass/perovskite has an estimated charge carrier lifetime of 5.74 ns where, in contrast, the charge carrier lifetime of the perovskite film deposited on glass/TiO2/perovskite was ~37% times shorter (τ = 2.12 ns), evidencing a faster electron transfer process in the latter case. The decrease in the PL intensity supported the enhancement in PCE for the glass/TiO2/perovskite-based device counterpart. To verify the reproducibility of the perovskite solar cells, 10 devices at each TiO2 ETL were fabricated. Fig. S5† presents the average values of the photovoltaic device parameters. The perovskite solar cells assembled from SILAR-mediated TiO2 ETL devices exhibited good reproducibility of Jsc, Voc, ff, and PCE37. To confirm the hysteresis behavior of TiO2 ETL, different scan rates (i.e. as high as 520 mV s−1 and as slow as 8.7 mV s−1) and scan directions (reverse scan: black solid symbol and forward scan: black open symbol) were operated on the TiO2 ETL-based perovskite solar cell (Fig. S6a–c†) whose data is summarized in Table S7†. A very little hysteresis was obtained at slow scan rate i.e. 8.7 mV s−1 as there was a light difference between the PCE values obtained during the reverse (9.7%) and forward (9.1%) scans, suggesting chemical stability and mechanical robustness of developed ETL layer (Fig. S6c†).

Figure 4
figure 4

(a) UV-vis absorbance spectrum of deposited perovskite absorber layer FTO/TiO2 ETL/perovskite, (b) J-V curve data collected at 8.7 mV s−1 scan rate in reverse scan direction, (c) EQE spectrum for perovskite absorber layer, and (d) EIS spectra for FTO/TiO2 ETL/perovskite/spiro-OMeTAD/gold. All the data were collected using overnight aged cells.

Full size image
Table 1 Summary of photovoltaic performance of TiO2 ETL/perovskite/spiro-OMeTAD/gold device fabricated with 10 SILAR cycles.
Full size table
Figure 5
figure 5

(a) Steady-state PL, and (b) Time resolved PL spectra of glass/TiO2 ETL/perovskite (excited with 325 nm He-Cd laser).

Full size image

Gas sensor activities

It is well-known that the gas sensing mechanism is based on the alteration in the resistance of the sensing material, which is influenced by the adsorption-desorption process of target gas molecules via charge transfer processes. The resistance of the TiO2 film was stabilized before exposure to various VOCs viz. ammonia, petrol, formaldehyde, ethanol, and acetone to a maximum of 100 ppm. Sensitivity, operating temperature and transient gas response study of the anatase TiO2 nanocrystalline SLG  substrate supported film (25 L) was envisaged for sensitivity study for various gases as for gas sensor thick film is essential. Above this condition, films started peeling off from the substrate surface due to its weak mechanical adhesion. The sensitivity study of TiO2 sensor for 100 ppm gas concentration at room-temperature is shown in Fig. 6a. The TiO2 nanocrystalline film showed better sensitivity (80%) towards ammonia over other target gases. The ammonia gas response of the TiO2 sensor as function of operating temperature was studied (Fig. 6b). The response was recorded maximum at room-temperature (300 K) whereas, it decreased with at higher temperatures. The maximum ammonia response at room-temperature could be because of high number of ambient oxygens as well as target gas molecules adsorption on the TiO2 sensor surface; whereas decreased adsorption process for higher operating temperatures deteriorated the ammonia detection response38. The transient ammonia response of anatase TiO2 nanocrystalline film sensor is shown in Fig. 6c. As the ammonia exposed in testing chamber, the diffusing ammonia molecules adsorb on the TiO2 film sensor surface over time. After specific time, sensor reached to its equilibrium (saturation level), where ammonia response remained constant. Furthermore, the sensing system was opened to external atmosphere after saturation. The ammonia molecules were desorbed and hence, the ammonia response decreased. With time, TiO2 sensor achieved its original state of resistance. The response and recovery time values of TiO2 film sensor for ammonia sensing were 30 and 85 s, respectively, signifying the use of TiO2 material as potential room-temperature ammonia sensor. The dynamic repeatability study of TiO2 sensor was studied and is shown in Fig. 6d. The 100 ppm ammonia gas was tested for several times at room-temperature in order to observe sensor. It was observed that TiO2 sensor shows approximately same ammonia response over a few cycles, signifying an effective and repeatable contribution of sensor adsorption/desorption sites. The ammonia response of the TiO2 sensor was studied for various concentrations at the room-temperature (Fig. 6e). The lowest ammonia concentration detected response at 10 ppm was 2.5%. The ammonia response increased with concentration. The ammonia response of TiO2 film sensor increased until the sensor surface available for adsorption of target ammonia molecules. The highest ammonia response for 1000 ppm concentration was 98%. The comparison of reported sensor performances of different materials used for ammonia detection is provided in theTable S5† and also the sensor performances of previously reported TiO2 sensor materials for different gases are given in the Table S6†39,40,41,42,43,44,45,46,47,48. For higher ammonia concentration above 1000 ppm, the gas response gets was saturated, which we assume is due to maximal coverage for the TiO2 surface by ammonia molecules to the point where limited resistance changes are detected. The stability plot of TiO2 film for ammonia sensing is shown in Fig. 6f. The response stability of TiO2 sensor towards 100 ppm concentrations of ammonia was studied for 15 days and observed to be approximately constant.

Figure 6
figure 6

SILAR-mediated anatase TiO2 film sensor; (a) selectivity of towards various VOCs gases at room temperature (300 K), (b) operating temperature optimization, (c) transient response and recovery curves, (d) repeatability towards ammonia at 100 ppm concentration, (e) response for various ammonia concertation levels (10–1000 ppm), and (f) long-term stability test towards ammonia.

Full size image

Sensing mechanism

The gas-sensing mechanism of sensor involves the adsorption of gas molecules on the metal oxide film surface and charge redistribution between the surface and the adsorbed molecules, which eventually produces changes in the electronic structure and conductivity of sensor material49. The ammonia sensing mechanism on TiO2 film sensor followed two stages i.e. oxidation and reduction. In oxidation process, air oxygen species oxidizes the TiO2 sensor surface by capturing conduction electron offering depletion region underneath TiO2 sensor surface and consequently conductional channel shrinks50. The width of depletion region formed depends on number of air oxygen species adsorbed and number of conduction electrons available in TiO2 sensor at the measurement temperature51. These interconnected space-charge depleted TiO2 nanoparticles within the sensor film offer additional resistance to that of the grain boundary resistance, because of an increased Schottky barrier potential (qVb) in air52. In the second stage, ammonia gas absorption reduces the TiO2 sensor surface by donating electrons via chemisorbed anion species according to38:

$${{\rm{O}}}_{2({\rm{air}})}\to {{\rm{O}}}_{2({\rm{ads}})}$$
(4)
$${{\rm{O}}}_{2({\rm{ads}})}+{{\rm{e}}}^{-}\to {{\rm{O}}}_{2({\rm{ads}})}^{-}$$
(5)
$$4{{\rm{NH}}}_{3({\rm{gas}})}+5{{\rm{O}}}_{2({\rm{ads}})}^{-}\to 4{\rm{NO}}+6{{\rm{H}}}_{2}{\rm{O}}+5{{\rm{e}}}^{-}$$
(6)

In this reduction process, the overall depletion width of TiO2 reduces by gaining electrons from oxygen ion species and increasing overall conductivity and reducing the Schottky barrier height. Consequently, the inter-particle conduction pathway width is increased rendering the entire film more conductive.

Conclusion

In summary, mesoporous anatase TiO2 nanocrystalline films were directly growth on FTO and soda-lime glass substrates by using a simple SILAR method at low-temperature i.e. 90 °C. The morphological, structural, optical and electrical properties of TiO2 nanocrystalline films are consistent across the substrate allowing ETL layers to be grown for perovskite solar cells. The best performing perovskite solar cell was constructed using a 15-layer SILAR grown TiO2 ETL achieved a PCE of ~9.7%, with JSC of 18.05 mA/cm2, VOC of 0.81 V and ff of 66.3%. The considerable 84% room-temperature sensor response of anatase TiO2 film sensor towards ammonia among various volatile gases demonstrates the film’s capability for resistance-based sensing. A wide range of gas concentrations detected by TiO2 film sensor at room-temperature confirmed the 2.5% response as lowest ammonia detection limit at 10 ppm whereas, at 1000 ppm it was 98%. The fast response and recovery time values of present sensor for ammonia sensing were recorded as 30 s and 85 s, and nearly fifteen days stability, respectively. As-prepared anatase TiO2 nanocrystalline film-based ammonia sensor demonstrated good repeatability, high selectivity, stability, and fast transition times, signifying its potentiality as room-temperature VOCs based gas sensor. Furthermore, the use of this novel process may in future find great potential in large-scale fabrication of perovskite solar cell and gas sensor models in practical applications.

Experimental Section

All the chemicals used as reactants were of analytical grade, and used without any further purification. All the other materials were purchased from Sigma-Aldrich, including titanium (IV) chloride (TiCl4), potassium persulfate (K2S2O8, 99%), lead iodide (PbI2, 99.9%), methylammonium iodide (MAI, Dyesol), dimethyl sulfoxide (DMSO, 99.9%), N-N-dimethylformamide (DMF, 99.8%), isopropanol, and chlorobenzene (99.8%) were used in the as-received condition. In addition, 2, 2′, 7, 7′-tetrakis (N, N-di-p-methoxyphenyl-amine) 9, 9′-spirobifluorene (spiro-MeOTAD) was obtained from Borun Chemicals (98%, Ningbo, China). All VOCs were obtained SD FineChem Limited, Mumbai, for gas sensing purpose. The laser-patterned FTO (15 Ω sq−1) and soda-lime glass substrates (Borosil) were cleaned with diluted detergent solution, ultra-sonicated for 10 min with deionized water, acetone, and ethanol, and dried with clean dry nitrogen air. These substrates were ultrasonically cleaned for 30 min prior to TiO2 deposition.

Preparation of SILAR TiO2

Synthesis of anatase TiO2 nanocrystalline films on two substrates viz. FTO and glass was carried out by low-temperature SILAR chemical deposition. Briefly, 0.1 M TiCl4 was prepared in 50 ml de-ionized water as the Ti4+ source (eq. 1). In another beaker, 0.1 M (50 ml) K2S2O8 was prepared (eq. 2) and kept at 90 °C constant temperature and used as sulphate/oxide source (eq. 3). The substrate (either FTO or glass) was dipped in 0.1 M TiCl4 solution for 20 s and then the same substrate was dipped in de-ionized water for 10 s to remove remaining loosely bonded ions. A similar procedure was performed in 0.1 M K2S2O8 solution at 90 °C using the same substrate, representing one cycle (eq. 4). After about ten such cycles, a whitish film started to appear on both substrates. The dipping cycles were continued until twenty on FTO substrate and twenty-five on the glass substrate were coated, to form uniform and relatively thick whitish TiO2 films. The deposited films were rinsed with de-ionized water and air calcined at 450 °C temperature for 1 h to obtain anatase TiO2 nanocrystalline films (eq. 5), whose chemical reaction mechanism is proposed below which further were characterized and employed for perovskite solar cell and chemical gas sensing applications.

$${{\rm{TiCl}}}_{4}\to {{\rm{Ti}}}^{+4}+4{{\rm{Cl}}}^{-}$$
(7)
$${{\rm{K}}}_{2}{{\rm{S}}}_{2}{{\rm{O}}}_{8}\to 2{{\rm{K}}}^{+}+{{\rm{S}}}_{2}{{{\rm{O}}}_{8}}^{--}$$
(8)
$${{\rm{S}}}_{2}{{{\rm{O}}}_{8}}^{2-}\,\mathop{\longrightarrow }\limits^{{\rm{\Delta }}90\,{}^{\circ }{\rm{C}}}\,2{{{\rm{SO}}}_{4}}^{--}$$
(9)
$${{\rm{Ti}}}^{+4}+{{{\rm{SO}}}_{4}}^{2-}\,\mathop{\longrightarrow }\limits^{{\rm{\Delta }}90\,{}^{\circ }{\rm{C}}}\,{\rm{Ti}}{({{\rm{SO}}}_{4})}_{4}$$
(10)
$${\rm{Ti}}{({{\rm{SO}}}_{4})}_{4}\,\mathop{\longrightarrow }\limits^{{\rm{\Delta }}450\,{}^{\circ }{\rm{C}}}\,{{\rm{TiO}}}_{2}$$
(11)

TiO2 ETL layer-perovskite device fabrication

The PbI2 was dissolved in DMF at a concentration of 462 mg per ml under stirring at 70 °C. The solution was kept at 70 °C during the whole deposition procedure. The PbI2 precursor was spin-coated on SILAR-based anatase TiO2 nanocrystalline ETL at 4000 rpm for 30 s and dried at 70 °C for 10 min. After cooling to room-temperature, the films were then dipped in a solution of MAI in isopropanol (15 mg per ml) for 20 s, rinsed with isopropanol, and dried by nitrogen gas. A volume of 60 μl spiro-OMeTAD solution was spin-coated on the perovskite/TiO2 layer at 3000 rpm for 30 s. The solution was prepared by dissolving 72 mg spiro-OMeTAD in 1 ml of chlorobenzene, to which 28.8 μl of 4-tert-butyl pyridine and 14.4 μl of lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) solution (520 mg Li-TFSI in 1 ml of acetonitrile, Sigma-Aldrich, 99.8%) were added. Finally, 70 nm of gold was deposited at ~10−6 bar via thermal evaporation on the spiro-OMeTAD coated film for electrical contacts.

Gas sensing measurement details

For the gas response measurement, changes in electrical resistance of TiO2 film sensors during the interaction of target gases were recorded and the response (S) of sensor was calculated according to:

$${\rm{S}}=\frac{{{\rm{R}}}_{{\rm{a}}}-{{\rm{R}}}_{{\rm{g}}}}{{{\rm{R}}}_{{\rm{a}}}\,}\times 100$$
(12)

where, Ra and Rg are the resistances of TiO2 film sensor in presence of air and target gas, respectively. The desired concentration of target gas was obtained by the static liquid gas distribution method, which was calculated by the following formula27:

$$C\,({\rm{ppm}})=\frac{22.4{\rm{\rho }}\mathrm{TV}\text{'}}{273{\rm{MV}}}\times 1000$$
(13)

where, C (ppm) is the desired target gas concentration; ρ (58.08 g mol−1) is the density of the liquid, V′ is the volume of liquid acetone (µL), T (298 K) is testing temperature, M is the molecular weight of acetone (0.788 g cm−3), and V is the volume of testing chamber (0.250 L).

Characterization tools

Surface morphology and elemental composition of TiO2 films were analyzed using field-emission scanning electron microscopy (FE-SEM, JSM-7001F, JEOL Ltd., Tokyo, Japan) digital images and energy dispersive X-ray (EDX) pattern, respectively. The high resolution transmission electron microscopy (HR-TEM, Titan 80–300, FEI, Hillsboro, OR, USA) and selected area electron diffraction (SAED) images were recorded using a FEI TECNAI G2 20 S-TWIN equipped with a LaB6 cathode and a GATAN MS794 PCCD camera. These micrographs were obtained at an acceleration voltage of 200 kV. The powder of TiO2 film sample was suspended in ethanolic solution separately and dropped onto a Formvar/carbon, 200 mesh copper grids before HR-TEM measurements. The phase of TiO2 film was confirmed from X-ray diffraction (XRD) pattern (XRD-6000, Shimadzu, Japan) with Cu-Kα radiation (λ = 0.1542 nm). Diffraction pattern was recorded from 20° to 70° (2θ) with a 2° min−1 scan speed. The phase analysis was additionally performed using a Raman scattering spectrum scanned on Renishaw Invia Raman Microscope. The laser radiation (λ = 532 nm) was focused on the TiO2 film surface. The spectrum was measured from 100 to 1000 cm−1. The surface-sensitive quantitative spectroscopic technique of X-ray photoelectron spectroscopy (XPS) and ultraviolet photoelectron spectroscopy (UPS) spectra of the TiO2 nanostructures were acquired with PHI 5000 Versa Probe (Ulvac-PHI) under high vacuum conditions (6.8 × 10−8 pa) using a monochromatic Al Kα X-ray source (1486.6 eV). The data was collected from a spot size of 100 × 100 μm2. The carbon 1 s photoemission line (284.6 eV) was used for internal calibration as a reference. The optical properties of the TiO2 nanostructured films were measured at room-temperature using a UV-vis spectrophotometer (V-530, Jasco). The photovoltaic performances and current-voltage (JV) characteristics of perovskite solar cells, for various synthesis conditions, were determined using a solar simulator (Sol3A Class AAA, Oriel Instruments) and a Keithley 2400 source measurement unit. The AM1.5 G simulated solar light (100 mW cm−2) light intensity level was calibrated using a standard Si reference cell certified by the Newport Corporation. All the devices were measured in a light-tight sample holder, with an active area of 0.06 cm2 for each cell fixed using an aperture mask. The external quantum efficiency (EQE) was recorded using a quantum efficiency measurement system (QEX10, PV Measurements, Inc.) as a function of wavelength from 300 to 850 nm. The steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) spectroscopy spectra of perovskite-coated on both the glass and glass/TiO2 were also performed using a micro-photoluminescence measurement system (IK350IR-G, 325-nm He-Cd laser). The gas sensing properties of TiO2 films gown directly on soda-lime glass were measured in a stainless-steel cylindrical chamber with 250 ml volume capacity. Resistivity in air and target gases were measured using a two-electrode arrangement, with silver paint metal contacts of 10 mm × 10 mm area, using a Keithley 6514 electrometer as a function of time and gas concertation. In order to control temperature, a small heater equipped with a thermocouple was used.