Improvement of the CO2 Sensitivity: HPTS-Based Sensors along with Zn@SnO2 and Sn@ZnO Additives

Fluorescent pH-sensitive indicator dye, 8-hydroxypyrene-1,3,6-trisulfonic acid (HPTS), has become known as a preferred alternative for continuous and accurate monitoring of dissolved and/or gaseous CO2 in chemistry, medical, and biochemical research. The objective of this work is to enhance the HPTS dye’s CO2 sensitivity in the presence of Zn@SnO2 and Sn@ZnO additive particles. Sol–gel synthesized metal oxide semiconductors (MOSs) were characterized using XRD, XPS, and SEM. The fluorophore dye and the MOS additives were embedded in the ethyl cellulose (EC) polymeric matrix to prepare the sensing thin films. The steady-state and decay kinetic measurements of the HPTS-based composites were obtained by PL spectroscopy for the concentration ranges of 0–100% p[CO2]. As expected, the addition of MOSs improves the sensor characteristics, specifically its CO2 sensing ability, linear response range, and relative signal change compared to the free form of HPTS. The CO2 sensitivities of the HPTS-based thin films were found at 17.6, 23.2, and 40.9 for the undoped, Zn@SnO2,-doped, and Sn@ZnO-doped forms of the HPTS, respectively. Additionally, the response and recovery times of the HPTS-based sensor agent with Sn@ZnO were measured as 10 and 460 s, respectively. The obtained results demonstrate that materials composed of HPTS with MOSs are potential candidates for CO2 sensors.


INTRODUCTION
In the fields of chemistry, medicine, and biochemistry, accurate measurement and monitoring of carbon dioxide (CO 2 ) levels are important for understanding biological processes, studying enzymatic reactions, assessing metabolic activity, and evaluating the effectiveness of drugs or therapeutic interventions. Due to this, it is crucial that CO 2 gas can be precisely and constantly measured over a broad area. Techniques based on infrared absorptiometry, electrochemistry, and luminescence can be used to primarily identify the presence of CO 2 gas. Optical sensors have gained popularity in recent years as a result of their advantages, such as electrical isolation, minimal noise interference, lower cost, fast response, adaptability for miniaturization, and simplicity of production and use. 1−6 Currently, electrochemical and optical devices based on spectroscopic changes are used to measure CO 2 , but optical chemical measurement methods have advantages, such as shorter response times, greater sensitivity, stability, and lower prices. The most common pH-sensitive fluorescent indicator dye, 8-hydroxypyrene-1,3,6-trisulfonic acid (HPTS), is commonly used for the accurate and continuous monitoring of dissolved CO 2 gas in various research fields. It exhibits strong emission and excitation bands when excited at 468 nm. It is simpler to immobilize HPTS on appropriate polymeric support materials when it is used as a trisodium salt or an ion pair. 7−11 In photoluminescence-based CO 2 sensors, the studied pHsensitive dye is physically trapped in polymeric matrices such as silicone, ethyl cellulose (EC), polystyrene (PS), and poly(methyl methacrylate) (PMMA), which provide a stable and protective environment for the dye molecules while allowing for interaction with the surrounding environment. Despite the widespread use of HPTS dye in optical applications and its superior response to CO 2 , it exhibits some disadvantages, such as stability, low relative signal variation, and reproducibility. As such, HPTS dye has been used with various additives such as ionic liquids (ILs), 12 metal oxide nanoparticles (NPs), 1,2 coordination polymers (CPs), 13 and/or metal oxide semiconductors (MOSs) 14,15 to eliminate such disadvantages. By incorporating MOSs into the sensor design, it is possible to increase the sensitivity and selectivity toward gas molecules. The porous structures of MOSs allow for a larger surface area available for gas adsorption, leading to enhanced sensing capabilities. 14,15 Metal oxide semiconductors (such as ZnO, CuO, SnO 2 , TiO 2 , Co 3 O 4 , Fe 2 O 3 , and WO 3 ), which have n-type and p-type forms, perform numerous gas sensing applications due to their chemical and physical properties and distinctive structure. 1 26−29 In core@shell systems, heteroatom doping with different metals can successfully increase the gas sensitivity of the MOSs in addition to altering the material with noble metals. The inclusion of these metal heteroatoms modifies the size, porosity, and specific surface area of MOSs, which, in turn, influences the adsorption sites of gas molecules and diffusion pathways. The presence of these heteroatoms can lead to changes in the crystal structure, resulting in a modified lattice size or unit cell dimensions. Changes in size, porosity, and surface area can affect the accessibility and connectivity of these diffusion pathways, altering the diffusion rates and pathways of gas molecules within the material. 22 For applications involving gas sensing, non-noble metals have also been reported to be used as cores in metal core@metal oxide shells. 30 In metal core@metal oxide shell nanoparticles, the basic metal atoms of an MOS are replaced by heteroatoms, resulting in a smaller grain size. By increasing the gas sensing ability of MOSs, gas sensors can be produced that outperform those composed of metal oxides. 22 In recent years, metal core@metal oxide shell structures with SnO 2 or ZnO as the shell for gas sensing applications have received much research. Rai and co-workers presented the Au@SnO 2 core@shell structure for CO gas sensing applications and showed a greater response compared to pristine SnO 2 . 31 In their study on the Au@SnO 2 core@shell structure, Yu et al. claimed target gases like CO catalytically react with oxygen adsorbents. According to the electrical mechanism, the modulation of the Schottky barrier caused by the creation of depletion zones around the particles improves the sensing of CO. 31 Wu et al. reported that SnO 2 , Ag/SnO 2 , and Ag@SnO 2 structures were used to detect ethanol vapor. When exposed to 200 ppm ethanol gas, the response time (t 90 ) and recovery time (t R90 ) of the SnO 2 material were measured as 54 and 85 s, while those of the Ag@SnO 2 material were 34 and 68 s, respectively. 32 Chung et al. explained the production of HCHO gas-sensitive Au@SnO 2 core@shell structures by the sol−gel method. While the SnO 2 material did not react against HCHO gas, the 1 wt % Au-doped SnO 2 material increased the sensor response by 2.4, while the Au@SnO 2 material increased it by 2.9. It has also been reported that Au@SnO 2 material reduced the response times. 33 Li et al. created Au@ZnO NPs and compared their CH 2 O-sensing abilities to those of ZnO and 1 wt % Au-doped ZnO NPs. With a faster reaction and recovery time, Au@ZnO exhibited an improved response signal. 34 Yang and co-workers reported that hydrothermal synthesized CdS/ZnO core@shell nanowires (NWs) demonstrated excellent visible-light-activated gas sensing performance toward ppb-level NO 2 at room temperature, with responses ranging from 6.7% to 337% when exposed to NO 2 concentrations from 5 to 1000 ppb. 35 36 According to the literature, many gas sensor studies containing metal oxides are based on electrical measurement. The advantage of the luminescence-based measurement is that it is used as an alternative in the absence of electrical contact between the resulting impurities and nanostructures. In addition, real-time information about the variation of certain additives in the photoluminescence spectra can be observed. However, the effect of adsorbed gases on the photoluminescence of metal oxide powders has not yet been adequately studied. This situation motivated us to study photoluminescent gas sensing materials based on Zn@SnO 2 and Sn@ZnO additives for the determination of CO 2 gas levels. In the current study, the CO 2 sensitivity of HPTS when used along with Zn@SnO 2 and Sn@ ZnO additives exhibited a considerable increase compared with the additive-free form. There has been an enhancement in the sensitivity and linear CO 2 response of HPTS when used with Zn@SnO 2 and Sn@ZnO additives. This improvement should be attributed to the gas adsorption on the surface of HPTS structures. However, when HPTS is used in the presence of Sn@ZnO NPs, the observed hypersensitivity to CO 2 should be attributed to the increased surface defects and the high surface-to-volume ratio of the nanoparticles.

Materials.
Zinc acetate dihydrate (Zn(CH 3 COO) 2 · 2H 2 O; 99%), tin chloride dihydrate (SnCl 2 ·2H 2 O; 98%), and potassium hydroxide (KOH; 85%) used in the synthesis of Sn@ZnO were of analytical grade and used without any additional purification. Chemicals used in the production of Zn@SnO 2 consist of an ammonia solution (NH 3 ), tin(IV) chloride pentahydrate (SnCl 4 ·5H 2 O), hydrochloric acid (HCl), and zinc chloride hexahydrate (ZnCl 2 ·6H 2 O; 99%). Tetrahydrofuran (THF), tetrabutylammonium hydroxide (TBAOH), dioctylphatalate (DOP) as a plasticizer, the ionic liquid (IL) 1-butyl,3-methyl imidazolium tetrafluoroborate [BMIM] [BF 4 ] that is used to increase stability, and ethyl cellulose (EC) that is used as a polymeric membrane were obtained from Sigma-Aldrich. N 2 and CO 2 gas cylinders were 99.9% pure and were obtained from Tinsa Gas, Izmir, Turkey. Tetraoctylammonium bromide (TOABr), dichloromethane (CH 2 Cl 2 ), and 8-hydroxypyrene-1,3,6-trisulfonic acid trisodium salt (HPTS) were purchased from Sigma-Aldrich and were employed as ion pairs that had been prepared in earlier studies. In a solution of Na 2 CO 3 (1 wt %) and CH 2 Cl 2 , the trisodium salts of HPTS and TOABr were combined in a 1:4 ratio. The production of the ion pair required the use of a separatory funnel, and the organic solvent was evaporated to obtain the ion pair. 12 2.2. Instrumentation. The structural analyses of the produced nanopowder samples were performed with a Panalytical/Empyrean X-ray diffractometer (XRD) using CuKα radiation at a scanning rate of 0.01°/min. a Thermo Scientific Kα X-ray photoelectron spectroscope (XPS) with a beam size of 400 nm diameter and a monochromatic Al Kα Xray source was used to analyze the surface chemistry and elemental content of the materials. Microstructure images were taken at different magnifications with a Zeiss Sigma 300 VP scanning electron microscope (SEM) to perform the morphological characterization. The steady-state photoluminescence (PL) measurements were performed using an FLSP920 Fluorescent Spectrometer. The measurement of the decay time values was carried out utilizing the time-related single photon counting mode (TCSPC) of the FLSP920. For detection CO 2 measurements, CO 2 and N 2 gases were mixed with the Sonimix 7000A gas mixing system in a concentration range of 0−100%.

Synthesis of Zn@SnO 2 and Sn@ZnO Core@Shell Structures.
To synthesize Zn@SnO 2 powder, 4 g of SnCl 4 · 5H 2 O was dissolved in 200 mL of distilled water (DW) under vigorous stirring for 20 min. Aqueous NH 3 was added dropwise to the solution to adjust the pH to 12. To the pHbalanced solution was added 1 g of ZnCl 2 ·6H 2 O, and the mixture was stirred for 30 min. The precipitate was collected and washed with DW. After drying in the oven at 100°C for 24 h, Zn@SnO 2 crystals were obtained. In order to improve crystallinity, crystals were annealed for 2 h at 500°C in an ambient atmosphere. 37 For the preparation of Sn@ZnO powder, 0.5 M Zn-(CH 3 COO) 2 ·2H 2 O was dissolved in 50 mL of DW. SnCl 2 · 2H 2 O prepared in a concentration of 0.1 M with 20 mL of water was added dropwise to the solution, and the reaction mixture was mixed for 30 min. Then, 2 M KOH in 50 mL of DW was added to the final mixture, and the mixture was stirred magnetically at room temperature. The resulting dispersions were purified several times with DW and ethanol to remove impurities. The final product was dried at 100°C for 6 h and annealed at 300°C for 3 h. 38 2.4. Thin Film Preparation. The polymer-based sensing thin films were prepared by mixing 100 mg of EC as a polymeric matrix, 96 mg of DOP as a plasticizer, 24 mg of [BMIM][BF 4 ] as an ionic liquid, 2.5 mL of THF as a solvent, and 0.10 mg of HPTS dye as a fluorophore dye. To enhance the CO 2 sensitivity of the HPTS dye, 0.10 mg of Sn@ZnO and 0.10 mg of Zn@SnO 2 powders were added separately to the HPTS-based polymeric cocktails. The compositions of HPTSbased cocktails in the absence and presence of Zn@SnO 2 and Sn@ZnO additives are shown in Table 1. Herein, the prepared composites were spread on a polyester (Mylar TM type) support using the knife coating method, and the thickness measurements of thin films were performed with a Tencor Alpha Step 500 Profilometer. The average of the thickness measurement results was found to be 6.53 ± 0.04 mm (n = 5). Optical CO 2 measurements were performed after the prepared HTPS-based thin-film sensor agents were placed in a quartz cuvette with a septum. A gas mixing system capable of producing precise CO 2 concentrations was used. The different CO 2 concentrations were given to the system by immersing a diffuser needle in the sensing agents under working ambient conditions. Before the HTPS-based thin films were exposed to CO 2 gas, the gas was passed through thermostated wash bottles containing water heated to 25°C and maintained at a constant relative humidity level of 100% to humidify it. Additionally, neither impurity peaks with elemental zinc nor those for other zinc compounds were seen, demonstrating the effective integration of zinc atoms into the SnO 2 crystal lattice. 40 The XPS spectra are given in Figure 2a and b, and Table 2 shows the results of both synthesized powders. Binding energies were calibrated using the C 1s hydrocarbon peak at ∼284 eV. Zn 2p was completely consistent with the divalent oxidation state of Zn 2+ , since the 2p 3/2 and 2p 1/2 peaks were separated at ∼1022 and ∼1043 eV, respectively. Additionally, the O 1s spectrum results highlight the two distinct oxygen atoms that were discovered at ∼530 and ∼531 eV. The binding energy of Sn 3d can vary depending on the specific oxidation state and chemical environment of the tin species being analyzed. The binding energy of Sn 3d is typically around 486−489 eV. The results of the Sn 3d spectrum at ∼486 and ∼494 eV were shown to correspond to Sn 3d 5/2 and Sn 3d 3/2 , respectively (see Table 2). 41,42 Figure 3 shows the FTIR spectra of the synthesized MOS powders. While the strong and broad band centered at 627 cm −1 corresponds to the Sn−O−Sn stretch, the peak of 628 cm −1 was due to the Zn−O−Zn vibration. In the literature, the characteristic stretching vibrations of ZnO bonds were observed in the absorption peak of wavenumbers between 500 and 650 cm −1 . 43,44 The broad band between 750 and 500 cm −1 is from the vibrations of Sn−O; −OH stretching vibrations in the wavenumber range of 3600 cm −1 are also due to the presence of H 2 O in the ZnO structure. It has been determined that there is a peak around 2340 cm −1 due to acetate and CO 2 molecules in the air. There are symmetrical and asymmetrical stress modes belonging to the acetate group  37 The surface morphologies of Sn@ZnO and Zn@SnO 2 particles were characterized by SEM at different magnitudes. The hexagonal plate-like structures of Sn@ZnO nanoparticles range in size from 552 to 812 nm. While the structure consisted of Zn@SnO 2 microparticles, some agglomerations were observed on the surface. The homogeneous distribution of these agglomerations provided good surface morphology (Figure 4).

Characterization of Sn@ZnO and Zn@SnO 2 .
3.2. CO 2 Sensitivity of HPTS-Based Sensing Agents along with Sn@ZnO and Zn@SnO 2 . In this study, the increase in the CO 2 sensitivity of HPTS with the use of Zn@ SnO 2 and Sn@ZnO powders as additives in the EC matrix was investigated. Measurements based on the signal change in the emission spectra of the prepared thin films were obtained after exposure to concentrations of 0−100% p[CO 2 ] with the humidification of the gas. The humidification process of the CO 2 gas is necessary for the formation of carbonic acid that reaches the active regions of the HPTS. With the increase in the CO 2 concentration in the atmosphere, the dye changes into a less fluorescent form from green to light yellow and signal drops in the emission band at 515 nm are observed. The Stern−Volmer equation is widely used to describe the assumption of a dynamic quenching process where the quenching species interacts with the excited state of the fluorophore, reducing its emission intensity (eq 1).
where I 0 is the intensity of the fluorescence or phosphorescence in the absence of the quencher, I is the intensity of the fluorescence or phosphorescence in the presence of the quencher at concentration [CO 2 ], and K SV is the Stern− Volmer constant. 10 The CO 2 -induced spectrum behavior of additive-free HPTS in the 0−100% p[CO 2 ] concentration range is shown in Figure  5. Based on the excitation of thin films in the EC matrix at 467 nm, signal decreases were observed in emission intensities at 515 nm.    The calibration graphs of all analyzed thin films in the concentration range of 0−100% p[CO 2 ] are given in Figure 8 comparatively. Table 3 shows the calibration equation, the Stern−Volmer (K SV ) constant, the regression coefficient, and the sensor sensitivity (I 0 /I 100 ) values of all HPTS-based composites obtained due to the calibration graphics. The correct equation for the indicated CO 2 concentration range of the additive-free HPTS sensor slide was calculated as y = 0.0217x + 1, and the correlation coefficient was 0.9128. However, for the concentration range of 0−40% p[CO 2 ], the correct equations of HPTS_Sn@ZnO and HPTS_Zn@SnO 2 composites were found to be y = 0.1558x + 1 and y = 0.1089x + 1, and the correlation coefficients were 0.9815 and 0.9505, respectively. When the CO 2 -induced variations of the composites were examined, it was observed that the HPTS_Sn@ZnO sensor material had a considerably higher slope and superior linear response in the 0−100% p[CO 2 ] range compared to other composites.
In this study, the use of Sn@ZnO and Zn@SnO 2 crystals with HPTS in the EC matrix resulted in improved I 0 /I 100 values compared with previous studies in the literature. It was seen that the enhancement of the CO 2 sensitivity was a result of the addition of MOSs to the dye as an additive. While the I 0 /I 100 value indicating the sensor sensitivity for HPTS without  additives was 17.6, these values were found to be 23.2 and 40.9 with Zn@SnO 2 and Sn@ZnO additives, respectively (Table  3). It was observed that HPTS_Sn@ZnO thin films increased the sensitivity to carbon dioxide gas ∼10× compared to the additive-free form.
In comparison to the free form of HPTS-based composites, the MOS-doped HPTS-based sensing slides have a quick response time and better sensitivity. In many instances, heteroatom doping with different metals can also successfully increase the gas sensitivity of MOSs. The inclusion of these metal heteroatoms modifies the size, porosity, and specific surface area of the MOS, which in turn alters the adsorption sites and diffusion pathways of gas molecules. It has been known that the surface shape of the MOS structures changes when metal additives are added as core@shell structures. As can be seen from the results obtained by SEM analysis, the Sn@ZnO additive showed higher sensitivity, since it has a smaller grain size than the Zn@SnO 2 additive. Since nanoparticles have a large surface area and hence more chemisorbed oxygen ions and a higher barrier height, the gas sensing capability can be improved. 24

Interactions between HPTS and MOSs.
To better understand the interaction causing the increase in the CO 2 sensitivity, we separately recorded the excitation and emission spectra of the HPTS, Sn@ZnO, and Zn@SnO 2 particles when excited at 370 and 468 nm (Figure 9). One of the probable reasons for the observed enhancement may be energy transfer between the MOSs and the HPTS, since both the Sn@ZnO and Zn@SnO 2 additives are substantially absorbed between 270 and 430 nm and releas at a broad wavelength range between 420 and 700 nm, covering the HPTS excitation band. These ranges overlap with the excitation band of HPTS, suggesting the possibility of energy transfer between the metal oxide particles and HPTS. Energy transfer, also referred to as Forster resonance energy transfer (FRET), is the term used to describe the phenomena. FRET occurs when two fluorophores (light-absorbing molecules) are in close proximity, and the excitation energy from one fluorophore is transferred to the other through nonradiative dipole−dipole interactions. In this case, the MOSs and HPTS act as potential energy donors and acceptors, respectively. The higher-energy electron in the lowest unoccupied molecular orbital (LUMO) of the excited fluorophore can undergo a transition to the conduction band (CB) of metal oxide semiconductors, leading to electron transfer between the two systems. Regarding the increase in fluorescence quenching of HPTS in the presence of a CO 2 atmosphere, the transfer of electrons from the excited singlet state to the CB of metal oxide semiconductors may contribute to this phenomenon. The presence of CO 2 can lead to energy dissipation pathways, such as electron transfer to gas molecules or other reactive species, resulting in the quenching of fluorescence. 48 Figure 10 shows

ACS Omega
http://pubs.acs.org/journal/acsodf Article improves the surface adsorption and desorption of CO 2 gas to the sensor surface. 49,50 Both HPTS and metal oxide particles competed with each other for absorbance when immobilized in the proximity of the polymeric matrix and simultaneously excited by the light source. For HPTS, metal oxide particles likely act as lightcollecting centers, creating new opportunities for additional excitation. However, together with FRET energy transfer, emission from metal and oxygen vacancies, interstitial metal ions, oxygen anticides, and electronic transitions between interstitial metal ions during relaxation improves the emission performance of HPTS. This is due to the efficient transfer of electrons from the excited state of HPTS to the conduction band of the semiconductor nanoparticle ( Figure 10). 51 Table 4 shows all of the recorded decay time values. Measuring the decay kinetics of HPTS-based thin films gives us important information about the interaction mechanism between the quencher and fluorophore (see Figure 11). While the fluorescence decay time values of HPTS_Sn@ZnO and HPTS_Zn@SnO 2 were recorded as 4.26 and 3.46 ns in the N 2 atmosphere, all of the phosphorescence decay time values were reduced to 3.71 and 3.12 ns, respectively, when fully exposed to CO 2 (see Table 4). When combined with Sn@ZnO and Zn@SnO 2 additives, the multiple exponential decay time values of HPTS exhibited a greater decrease than those of the additive-free forms. The obtained results showed that factors such as surface defects, electrical conductivity, and charge transfer cause a reduction in decay time kinetics. The adsorbed or diffused gas in the MOSs decreases the carrier density, resulting in potential barriers between oxide particles that cause reduced electrical conductivity. This charge mobility also causes a decrease in the luminescence intensity and decay time kinetics.  Figure 11. Decay curves of (a) HPTS_Sn@ZnO and (b) HPTS_Zn@SnO 2 in the fully N 2 and CO 2 gas atmospheres.

ACS Omega
http://pubs.acs.org/journal/acsodf Article 3.5. Reproducibility of the Sensor Slides. The response and recovery time for gas sensing refers to the time it takes for a CO 2 gas sensor to detect the presence of CO 2 and then return to its baseline reading after the CO 2 concentration has changed. The MOS additives assist in the desorption of CO 2 molecules from the sensing material, improving the recovery time of the HPTS and its ability to repeat the sensing cycle. The regeneration measurement is a crucial parameter for CO 2 sensor applications, as it determines the sensor's ability to recover its sensing properties after exposure to CO 2 . The response and reversibility measurements of the EC thin filmembedded HPTS dye in the presence of MOSs were interpreted according to time and varying quencher concentrations in fully N 2 and CO 2 gas atmospheres. The results obtained were response time and reversibility for CO 2 sensing (see Figure 12). The signal changes of HPTS-based thin films with Sn@ZnO and Zn@SnO 2 were reversible during measurements after the third cycle, and the standard deviations of the upper and lower signal intensities were found to be less than 5.0%. The response and the recovery times of HPTS_Sn@ ZnO and HPTS_Zn@SnO 2 were determined as 10, 20, 460, and 520 s in fully CO 2 and N 2 atmospheres (see Figure 12).
The long-term stability of fluorescent dye is one of the important factors in optical sensor design. Since the long-term instability of HPTS-based optical sensors is an important factor to be improved, in the study, we used an ionic liquid during the preparation of thin films to improve the stability of HPTS. Even after 10 months, the HPTS-based sensing slides stored in the dark under laboratory conditions showed nearly the same emission-based intensity. Considering the results obtained in this study, we confirmed that all sensing membranes gave reproducible and stable responses for CO 2 measurements.

CONCLUSION
As a result, herein, the pH-sensitive HPTS dye was used in the presence of additives of Sn@ZnO and Zn@SnO 2 for the first time, and thus HPTS dye has been shown to increase its sensitivity to CO 2 gas. Although the I 0 /I 100 value with the additive Zn@SnO 2 was 23.2, this value increased 40.9-fold in the case of the Sn@ZnO additive. The results show that doping metal oxides with non-noble metals generates free electrons that facilitate the adsorption of CO 2 molecules on the surface and defect and morphology changes. Analysis based on measurement of the emission intensity with the addition of additives showed that HPTS showed a higher sensitivity to CO 2 gas, a higher K sv constant, and a better relative signal change. In addition, data were obtained with a better linear calibration graph compared to HPTS without additives in the range of 0−100% p[CO 2 ]. The charge transfer in the adsorption process is the basis of the gas sensing mechanism. As a result of the metal-doped semiconductor metal oxide addition, the results showed that these particles are among the most promising materials for CO 2 sensor design. The obtained results can be attributed to defects in the semiconductor metal oxide system. Surface defects play a crucial role in the adsorption of gas molecules on oxide surfaces, inducing charge transfer between adsorbents and substrates even in the presence of a small number of defects. This shows that the material has important effects on its electronic, structural, and optical properties.