Optical CO 2 Gas Sensing Based on TiO 2 Thin Films of Diverse Thickness Decorated with Silver Nanoparticles

)e fabrication, characterization, and CO2 gas detection performance of single component-based and hetero-nanostructurebased optical gas sensors are reported in the present work. Single component-based structures include (i) TiO2 thin films with varied film thickness (37.45 nm, 51.92 nm, and 99.55 nm) fabricated via the RF sputtering system for different deposition times and (ii) silver nanoparticles (AgNPs) deposited on the glass substrate by the wet chemical method. Hetero-nanostructures were achieved by decorating the AgNPs on the predeposited TiO2 thin films. )e structural, morphological, and optical characteristics of prepared samples were studied by X-ray diffraction (XRD), scanning electron microscopy (SEM), and ellipsometry, respectively. XRD analysis of AgNPs confirmed the crystalline nature of prepared particles with average crystallite size of 21 nm, however, in the case of TiO2 films XRD results suggested amorphous structure of all as-deposited films. size 21 nm. )e SEM micrographs confirmed the deposition of AgNPs on the TiO2 thin films.With increasing sputtering time, TiO2 films were found to be denser and more compact, indicating a reduced porosity and higher film thickness. CO2 gas-sensing properties were investigated by measuring the optical transmission spectra in alone air and in CO2 gaseous atmosphere at room temperature. It was observed that neither TiO2 thin films even with higher thickness nor alone AgNPs could demonstrate any substantial gas-sensing activity. Nevertheless, TiO2/AgNP hetero-nanostructured substrates exhibited excellent CO2 gas-sensing performance as indicated by a huge change in the transmission spectra. )e enhanced sensing efficiency of TiO2/AgNP nanostructures owing to synergistic effects suggests a promising role of our manufactured sensors in practical applications.


Introduction
To meet the needs of rapidly increasing human population, industry production has drastically increased.During this process, a lot of industrial waste is causing environmental pollution due to improper disposal and management.e most common forms of industrial contaminants are pollutant gases such as CO, C 6 H 4 , NO x , NH 3 , CH 4 , SO 2 , and CO 2 .e exhaust fumes from automobiles are further intensifying the quantity of such pollutants in air making them more toxic and harmful for humans and for the natural ecosystem [1,2].e increasing public health issues demand strong efforts to reduce the impact of these noxious gases.
e development of efficient sensors for gas monitoring could be the first step to address this challenge.
For the quantitative detection of various gases in the environment, different types of traditional gas sensors are used including pellistors, semiconductor devices, and electrochemical gas sensors.Although these sensors are widely employed in several applications, they suffer from different drawbacks.For example, pellistors being robust in nature respond to combustion on a catalyst bead.ey are suitable for flammable gases detection around the lower explosive limit but face zero drift problems at parts per million (ppm) levels.Semiconductor gas sensors can work effectively at the low ppm levels; however, they also suffer from drift and humidity variations and cross-respond to other gases.Similarly, electrochemical gas devices can be sensitive at ppm or ppb levels and comparatively explicit to individual gases; nevertheless, they have issues such as limited lifetimes, known cross-response, and changing humidity levels [3,4].
Owing to aforementioned limitations and drawbacks of conventional gas-sensing devices, optical gas sensors are attracting more attention.ey are based on materials that can change their optical response when exposed to target gaseous atmosphere.
ey are fast in action (even below 1 sec) and have negligible drift and high gas specificity, with minimal cross-response.ey provide a wide range of operative parameters, and any change in intensity, frequency, polarization, and phase of reflected or transmitted light as compared to incident light can be used in the sensing mechanism.Measurements made by optical gas sensors are self-referenced and inherently reliable because the intensity of incident and transmitted/reflected light can be determined.Furthermore, contactless readout measurements and independency from electromagnetic noise make them suitable for optical fibers and integrated photonic devices for possibly fast and easy signal transport and in situ measurements with a compact, flexible system [1,5,6].
Carbon dioxide (CO 2 ), a colorless, generally inert in nature, and highly oxidizing gas, belongs to those greenhouse gases that occur both naturally (by decomposition of organic matter) and from human activities (including burning fossil fuels for power generation, oil refining, natural gas production, and transportation).Higher concentration of CO 2 in the atmosphere is not only dangerous for living beings but also has severe impact on global climate.For example, maximum limit of CO 2 required by safety regulations in a mixture of gases for mines, wells, and sewers operations is 5-20%.Beyond this limit, it becomes hazardous for humans health.Likewise, the anomalous increase of CO 2 can cause huge global climate changes such as greenhouse effect, global warming, possible expansion of subtropical deserts, and increase in sea levels [7][8][9].erefore, the qualitative and quantitative detection of CO 2 is of substantial interest in various fields such as biotechnology, food packaging and beverage, marine and environmental science, medicine and health, air quality, and industrial monitoring [10,11].
Titanium dioxide (TiO 2 ) owing to low cost, low oxidation potential, and nontoxic in nature along with high refractive index, high dielectric constant, and chemical stability is of much interest in various applications such as catalysis, optoelectronic devices, photocatalysis, and sensors [12][13][14].TiO 2 -based gas sensors are becoming popular for their potential use in medicine, biology, self-safety, transportation industries, and environmental protection purposes [15,16].To improve the sensing performance of these sensors, various techniques have been employed by the researchers.Manufacturing of TiO 2 -based heteronanostructures using the metallic nanoparticles is a promising approach in the gas-sensing mechanism.e combined effect of the functional TiO 2 matrix and large active surface areas of nanoparticles could boost the sensing efficiency as compared to single component-based sensors [15,17,18].e aim of this work was to explore the synergistic effect of TiO 2 /AgNP hetero-nanostructures for CO 2 gas sensing.TiO 2 thin films of varied thickness were fabricated via the RF magnetron sputtering system, and AgNPs were decorated on them by the wet chemical method.
e optical CO 2 gas detection experiments were carried out by measuring the transmission signals in a normal incident configuration of a variable-angle ellipsometer at room temperature.CO 2 gassensing performance of alone AgNP films, pristine TiO 2 thin films having different thickness, and AgNP-loaded TiO 2 films was investigated.
e effect of enhanced TiO 2 film thickness on the AgNP deposition and CO 2 gas-sensing efficacy was also studied.
e focus of this study was to work out the best possible composite morphology for most effective and reliable CO 2 gas sensors.

Preparation of Samples.
ree types of substrates were prepared: (i) alone TiO 2 thin films having thickness of 37.45 nm, 51.92 nm, and 99.55 nm, (ii) alone silver nanoparticles deposited on glass substrates, and (iii) TiO 2 /AgNP hetero-nanostructure substrates with varied abovementioned TiO 2 film thickness.Titanium dioxide (TiO 2 ) thin films were fabricated on soda lime glass substrates via the RF magnetron sputtering technique.e glass slides were cleaned ultrasonically for 20 min first in acetone and then for further 20 min in IPA and dried in nitrogen (N 2 ) flow, prior to film deposition.TiO 2 films were sputtered using the titanium dioxide target, in the argon (Ar) environment with a fixed flow rate of 50 sccm at room temperature and 10 −6 Pa of pressure.To achieve a varied film thickness, the deposition time was changed as 30 min, 60 min, and 90 min, keeping all other parameters unchanged.e TiO 2 thin film samples were used as-deposited without any annealing for further application.
In the next step, deposition of sliver nanoparticles (AgNPs) was carried out on bare glass substrates and on TiO 2 thin film samples.AgNPs were first synthesized following a chemical reduction route described in detail in our previous study [19].Briefly, 100 mL of 1 mM aqueous solution of silver nitrate (AgNO 3 ) was heated to boiling temperature under vigorous stirring and refluxing conditions.
en, 10 mL of aqueous solution of 1% trisodium citrate (Na 3 C 6 H 5 O 7 ) was added dropwise with the help of a dropper.
e solution changed color from transparent color to pale yellow to bright yellow and finally to greenish yellow, indicating the completion of the reaction as shown in Figure 1.Dissolving of silver precursors produced silver ions (Ag + ) in the aqueous solution.ese Ag + ions were reduced to free silver atoms (Ag 0 ) by the reducing agent (trisodium citrate) which resulted in the initial color change (pale yellow).Owing to nucleation and growth processes, produced free silver atoms (Ag 0 ) gathered into the oligomeric clusters and finally into silver nanoparticles [20].e color change from bright yellow to greenish yellow indicated the 2 Advances in Materials Science and Engineering di erent phases taking place in the reaction.Trisodium citrate served as a reducing agent and a stabilizing agent in this case.e colloidal solution was cooled to room temperature for further characterization and applications.Before decorating with AgNPs, all substrates (bare glass and TiO 2 thin lms) were functionalized with 3-aminopropyltriethoxysilane (APTES) to obtain a high surface a nity for the colloidal AgNPs adopting a solution-based approach [21].Brie y, bare cleaned glass substrates and TiO 2 lm samples were immersed into a 10% APTES methanolic solution for 60 min, followed by thoroughly rinsing with pure methanol and drying in N 2 ow.en, the samples were immersed in water for 30 min and dried in N 2 ow after rinsing with water.Finally, APTES-derivatized substrates were immersed in the silver colloidal suspension for the whole night, rinsed with water, and dried in N 2 ow.All of these procedures were performed at room temperature.

Characterization of Prepared Samples.
All three types of fabricated samples were characterized with di erent characterization techniques for their morphological, optical, and structural analyses.Structural investigations were carried out by using X-ray di raction (XRD) methods with an X-ray di ractometer (D-MaxIIA; Rigaku, Tokyo, Japan) operated at 40 kV and 25 mA using Cu Kα line radiation (λ 1.5406 Å).Few drops of the silver colloidal sample were dried on the clean glass substrate to form a thick lm for XRD measurements.A eld-emission scanning electron microscope (FEI Nova NanoSEM 450; Hillsboro, USA) was used to study the morphology of obtained structures at di erent magnications.A thin gold layer (10 nm) was also deposited, if needed, to avoid any possible charging on the sample.e thickness and the optical constants of TiO 2 thin lms fabricated at di erent deposition times were determined by spectroscopic ellipsometry.e measurements were carried out using a variable-angle ellipsometer (M-2000; J.A. Woollam, USA) in air (without chamber) at a xed incidence angle of 65 °with respect to the surface normal operating in the wavelength range of 300-900 nm at room temperature.

Optical Gas Sensing.
In order to examine the optical CO 2 gas-sensing performance of all three aforementioned categories of samples, optical transmission measurements were performed (in air and in CO 2 gas) within a custom-made transparent airtight gas-sensing chamber (8 cm × 6 cm × 12 cm) using a spectroscopic ellipsometer in the vertical con guration (as shown in Figure 2) in the wavelength (λ) range of 300 nm to 900 nm at room temperature.CO 2 gaseous environment was established by a constant ow rate of 50 sccm.Measurements were carried out by exposing each sample to air and to CO 2 gaseous atmosphere for 2 min.
e light beam was allowed to fall on the substrate at right angle (as schematically shown in Figure 2) which covered an area of about 2.8 mm 2 on the substrate.
e change in transmission due to CO 2 gas was recorded for each sample.

Results and Discussion
Figure 3 shows TiO 2 thin lm samples before and after AgNP deposition as an upper panel and a lower panel, respectively.It can be seen that the shade of the color changes from lighter to darker as the deposition times enhanced from 30 min to 90 min, which could be an indication of increased lm thickness.After decorating with silver NPs, a further change in the color was noticed that depicts the successful adsorption of AgNPs.

X-Ray Di raction Analysis.
e crystalline nature of asdeposited TiO 2 thin lms with sputtering times of 30 min, 60 min, and 90 min (without any further heat treatment) and synthesized AgNPs was determined by the XRD technique at room temperature.
e XRD indexed pattern of alone AgNPs deposited on the glass substrate by the drop-casting method is shown in Figure 4. Four major characteristic peaks at 2θ values of 38.45 °, 46.35 °, 64.75 °, and 78.05 °can clearly be observed.ese peaks can be discerned to the re ection planes (111), ( 200), (220), and (311) of the face-centered cubic (FCC) pure metallic silver according to JCPDS le number 04-0784 [22].us, XRD results con rmed the pure crystalline metallic nature of our synthesized AgNPs.e broadening of the major di raction peak (111) was utilized to calculate the crystallite size of prepared AgNPs using Debye-Scherrer's formula [23] as follows: where λ is the wavelength of X-ray used (1.54 Å), k is the shape factor (0.9), β hkl (taken in radians) is the full width at Advances in Materials Science and Engineering half maximum (FWHM) of the di raction peak, and θ is the di raction angle at that intensity.All measured vales are listed in Table 1.e average crystalline size of AgNPs was found to be 21 ± 1 nm as listed in Table 1.e XRD spectra obtained for TiO 2 thin lms for deposition times of 30 min, 60 min, and 90 min are shown in Figure 5(a).No sharp peak can be observed in all three spectra which can help to identify the crystalline nature or any phase of deposited TiO 2 lms.Absence of any strong re ection peak revealed apparently the amorphous nature of as-deposited TiO 2 thin lms in our case.is indicated that, without any further annealing or heat treatment, there was not enough crystalline structure developed which could be detected by the XRD.Perhaps, proper annealing could be helpful to develop su cient crystalline nature of the lms.However, a slight increase in the intensity of the spectra can be noticed by increasing the lm deposition time which might be due to increased lm thickness.
For the measurements of fabricated lm thickness, refractive index, and lm roughness, spectroscopic ellipsometry was employed, and calculations were made using the wellknown Tauc-Lorentz oscillator model [24,25].e obtained results are presented in Figure 5(b) by plotting the refractive indices as a function of wavelength and are listed in Table 2.It can be observed that, by increasing the deposition time, TiO 2 lm thickness, refractive index, and lm roughness increased which can be attributed to enhanced atomic packing density and decrease in the porosity of the deposited lm [25,26].

Scanning Electron Microscopy Analysis.
In order to determine the size and the shape of prepared AgNPs and the morphology of the TiO 2 thin lms with and without AgNP deposition, a eld-emission scanning electron microscope (FE-SEM) was used.
Figure 6 shows the SEM micrographs of AgNPs deposited on glass substrates (by drop-casting methods).It is evident from SEM images that shapes of the silver particles were not perfectly spherical, although some round-shaped particles can also be noticed in magni ed view (Figure 6(a)).e size of synthesized particles was found in the range of 20 nm to 60 nm.e observed agglomeration and clustering of particles could be the result of depositing and drying processes of silver colloids on glass by the drop-casting method.
SEM images of TiO 2 thin lms prepared at di erent deposition times (30,60, and 90 min) on glass substrates with and without AgNPs are presented in Figure 7.It can be For the TiO 2 samples with moderate lm thickness of 51 nm (Figure 7(c)), it appeared that TiO 2 lm became dense and compact, indicating the decreased porosity and enhanced lm thickness.e occurrence of TiO 2 clusters at many places can also be noticed increasing the lm roughness.e introduction of AgNPs (Figure 7(d)) over the 51 nm thick TiO 2 lms depicted that density of scattered

Advances in Materials Science and Engineering
AgNP clusters appeared to increase.Perhaps, the agglomeration of AgNPs occurred over the already developed TiO 2 clusters which increased by increasing film deposition time.
In the case of 90 min deposition time, the film thickness and roughness further increased, and more clusters and bump-like structures can be observed on the TiO 2 films (Figure 7(e)).e film morphology after depositing the AgNPs revealed that hill-and valley-like structures covered the TiO 2 films (Figure 7(f )).
e deposition of AgNPs in the form of clusters and bunches instead of uniform distribution of individual nanoparticles could be due to nonuniform adsorption of APTES layers on TiO 2 thin films by various chemisorption and physisorption mechanisms leading to scattered binding sites for AgNPs [27].Besides, the formation of TiO 2 clusters in the films might be another reason of AgNP agglomeration during the deposition process that enhanced the manufactured structure roughness.
It can be concluded from the SEM results that, by increasing the film deposition times from 30 to 90 min, the film morphology clearly changed by improving the thickness and roughness of the deposited film.e deposition of AgNPs not only developed the hetero-nanostructures but also enhanced the roughness by nonuniform decoration of silver clusters and hill-and valley-like assemblies on TiO 2 films.

CO 2 Gas-Sensing Analysis.
In order to describe the measured optical response of all the samples in the air and in CO 2 atmosphere, the substrates were divided into two categories: (a) Substrates without silver nanoparticles (AgNPs) (b) Substrates with silver nanoparticles (AgNPs).

Substrates without AgNPs.
In Figure 8, transmission spectra of all four samples, bare glass and TiO 2 thin films, having different film thickness (37.54 nm, 51.92 nm, and 6 Advances in Materials Science and Engineering 99.55 nm), in air and in the presence of CO 2 , are presented.It can be seen that, in case of air (Figure 8(a)), bare glass exhibited the typical spectrum with maximum transmission.Apparently, the transmission value of the glass substrate (∼60%) seems lower than the reported value (<90%) [28].All optical measurements were carried out inside a transparent box to generate a gaseous environment around the sample (Figure 2).e lower values of the bare glass substrate might be due to re ections o the windows of the gas cell apparatus.For the TiO 2 thin lm samples, transmission was found to decrease by increasing the lm deposition. is can be understood by considering the improvement in the lm thickness and density which absorbed more light causing a decrease in transmission.In the case of CO 2 gaseous atmosphere, transmission spectra (Figure 8(b)) of all four samples without AgNPs appeared similar to those as were in the case of alone air.us, no clear di erence could be noticed in the presence of CO 2 gas.

Substrates with AgNPs.
e optical transmission spectra of all samples decorated with AgNPs in air and in the presence of CO 2 gas are demonstrated in Figure 9.A slight Advances in Materials Science and Engineering change in the shape of all spectra can be noticed.As evident from SEM images (Figure 7), deposition of AgNPs formed di erent-sized Ag aggregates along with some scattered distinctive AgNPs.ese Ag aggregates could act as scattering centers causing a decrease in the transmission signal which are discussed in the next section.

Comparison of Gas-Sensing Performance.
A detailed comparison of optical characteristics in the presence and absence of CO 2 gaseous atmosphere for all samples is reported in Figure 10.
Figure 10(a) represents the transmission response of bare glass and alone AgNP-deposited glass substrate in air and in CO 2 gas ow.A huge decrease in the transmission spectra (in air) of the glass substrate decorated with AgNPs as compared to bare glass indicates the improved surface coverage of the glass substrate by the deposition of AgNPs.However, no signi cant change in the optical spectra due to the CO 2 gas atmosphere can be noticed except a slight decrease in case of AgNP-loaded samples.
is revealed substrates covered with AgNPs alone were not su cient enough for CO 2 gas detection.
In the case of TiO 2 thin lm (37 nm thickness) samples with and without AgNP deposition (Figure 10(b)), the overall decrease in the optical response was found to be larger as compared to samples without TiO 2 lms (Figure 10(a)).is depicts that morphology and thickness of the material on glass substrates had changed due to TiO 2 8 Advances in Materials Science and Engineering thin lms.However, absence of any prominent change in the transmission in the CO 2 gaseous environment proved that TiO 2 thin lms with insu cient thickness despite the presence of AgNPs were also incapable of noticeable CO 2 gas sensing.
For the TiO 2 samples with 51 nm lm thickness (Figure 10(c)), a signi cant decrease in the transmission spectra, starting from 600 nm to higher wavelength values, was noticed when TiO 2 thin lms decorated with AgNPs were exposed to CO 2 gas ow. is indicates that increased TiO 2 content in hetero-nanostructures (TiO 2 /AgNPs) had improved the CO 2 gas-sensing properties of the samples.is argument was further strengthened when TiO 2 thin lms of higher thickness (99 nm) loaded with AgNPs were analyzed in the CO 2 gaseous environment as shown in Figure 10(d).A huge decrease in the transmission of hetero-nanostructured samples exposed to CO 2 gas (green line) was noticed as compared to air alone (blue line).
ese optical results revealed that neither alone silver nanoparticles nor alone TiO 2 lms of various thicknesses were pro cient CO 2 gas sensors nevertheless; hetero-nanostructures (TiO 2 /AgNPs) with higher TiO 2 content provided the maximum CO 2sensing properties.

Dynamic Gas
Sensing. Figure 11 illustrates the dynamic CO 2 gas-sensing response of the heteronanostructured sample TiO 2 /AgNPs with maximum TiO 2 lm thickness (99 nm).
e transmission signal measurements were carried out at room temperature for a xed wavelength (λ 600 nm) by owing air and CO 2 gas alternatively.
e response time (air-to-CO 2 gas transient) is de ned as the time required to achieve 10% to 90% of maximum transmission signal, while the recovery time (CO 2 gas-to-air transient) is de ned as the time required to achieve 90% to 10% of maximum transmission.e response time was found to be about 10 sec, while recovery time was about 110 sec.e dynamic gas-sensing results (Figure 11) revealed the stability and reversibility of our prepared CO 2 gas sensors.

Transmission Change Ratio (TCR).
e sensitivity of various samples toward CO 2 gas can more clearly be observed in Figure 12, which reports the calculated transmission change ratio (TCR) values.TCR is de ned as the di erence between the transmissions measured during CO 2 gas exposure (T gas ) and in air (T air ), normalized to the transmissions level in air, and is given by the following formula [17]: e TCR plots of substrates without AgNPs and with AgNPs are presented in Figures 12(a) and 12(b), respectively.e almost same behavior of TCR around zero line for all the samples without AgNPs (Figure 12(a)) indicates that increased TiO 2 lm thickness alone was not enough to produce any sharp change in transmission as exposed to CO 2 gas.However, in the case of heteronanostructured morphology, a combination of TiO 2 thin lms of higher thickness with AgNPs, a substantial variation in TCR trend was noted.A perturbation in TCR (blue line) of 51 nm thickness (TiO 2 /AgNPs) sample starting from 400 nm showing peak value around 550 nm and ending at higher wavelength values can clearly be seen.Similarly, a negative trend in TCR for 99 nm thickness (TiO 2 /AgNPs) substrates indicates a large decrease in the transmission when exposed to CO 2 gas atmosphere.Again a perturbation starting from 460 nm and ending at higher wavelength values can be observed.From these TCR results, it can be inferred that enhanced TiO 2 lm thickness decorated with AgNPs con rmed e ectiveness of hetero-nanostructures for good CO 2 gas sensors.

Optical Sensing Mechanism for CO 2 Gas.
e interaction of gases with solid surfaces depends on various factors such as the type of the gas (reducing or oxidizing in nature), chemical nature of solid surface (n-type or p-type), surface morphology, and the surrounding atmosphere including temperature and humidity [29][30][31].Furthermore, type of sensing layers' con guration, whether a single component-based nanostructure or a multimaterial heteronanostructure, also plays a vital role to control surface reactions and thus gas-sensing mechanism [15].TiO 2 is an n-type semiconducting material having d 0 electronic con guration and belongs to the binary transition metal oxide category which is considered potential candidate for real gas-sensing applications.
e gas-sensing mechanism of TiO 2 -based sensors can be described by considering the adsorption of oxygen on the surface of sensors which could be due to physisorption and chemisorption processes.
ese processes are found to be surface temperature and activation energy dependent.When TiO 2 -based sensors are exposed to air at room temperature, oxygen molecules may adsorb on the surface of TiO 2 -based materials by physisorption process due to Van der Waals and dipole interactions.ese adsorbed oxygen molecules, then, generate chemisorbed oxygen species (O − 2 ) on the sensor surfaces.e reaction process can be described as follows [15,30]: Carbon dioxide (CO 2 ) is an oxidizing gas having a linear bonded atoms' stable structure with no lone pair of electrons.When TiO 2 -based surfaces are exposed to CO 2 gaseous atmosphere at room temperature, the moisture present on the surface of sensor come into action.e water molecules split into hydroxyl and hydrogen ions which subsequently react with gaseous CO 2 to form carbonate ions as follows [31]: TiO 2 -based gas sensor properties are also in uenced by the TiO 2 crystal phases, and rutile and anatase phases are found to be more promising for the sensing applications.
In our case, alone pristine TiO 2 thin lms even with increased lm thickness (37-99 nm) exposed to CO 2 gaseous ow could not exhibit any notable sensing performance (Figures 10(a)-10(c)).One of the probable reasons of this little or no sensing response of as-deposited TiO 2 lms could be the lack of TiO 2 crystalline phases (Figure 5(a)) or higher sensitivity of TiO 2 polymorph toward reducing gases (such as CO and H 2 ) than oxidizing gases such as CO 2 [17,29].Besides, negligible CO 2 gas-sensing properties of alone AgNP substrates (Figure 10(a)) suggested that neither alone as-deposited TiO 2 thin lms nor alone AgNPs were substantial materials for CO 2 gas-sensing mechanism.Instead, hetero-structure fabricated by decorating AgNPs on the predeposited TiO 2 thin lms demonstrated excellent CO 2 gas sensing (Figure 12(b)).Manera et al. [12] in the case of alcohol vapor sensors using TiO 2 /gold nanocomposite and Wang et al. [30] during their CO gas-sensing studies have already reported the signi cance of hetero-nanostructures in improving the gas-sensing properties.
is enhanced gas-sensing abilities of our TiO 2 /AgNPs hetero-structured sensors can be explained in the way that introduction of silver nanoentities on the TiO 2 lms generated high oxygen dissociation active sites which increased the probability of CO 2 adsorption and reaction on the sensor surfaces [15,32].
is perhaps provoked the prominent change in optical transmission when sensors exposed to in CO 2 gaseous environment.Furthermore, enhanced adsorption of CO 2 near the AgNPs could lead the changes in the refractive index of TiO 2 [33] which caused a change in optical response.Another reason of improved sensing e ect due to AgNPs on the TiO 2 lms could be the surface plasmon resonance (SPR) e ect which is in uenced by the variation in thickness and refractive index of the active sensing layer.Moreover, the size and density of AgNP islands on the TiO 2 lms could also a ect the SPR gas sensitivity [33,34].e increased CO 2 sensing properties of TiO 2 lms decorated with AgNPs with higher lm thickness could be due to the enhanced SPR e ect owing to increasing the thickness of TiO 2 lm along with growing size and density of AgNP islands.us ndings of this work proved that TiO 2 /AgNPs heteronanostructures were more e cient CO 2 gas optical sensors rather than alone AgNPs or as-deposited TiO 2 lms.e excellence performance of prepared hetero-nanostructured CO 2 gas sensors makes them promising candidates for practical gas-sensing applications.

Conclusion
In summary, CO 2 gas-sensing characteristics of different substrates including alone AgNPs, TiO 2 thin films with different thickness prepared by different sputtering times and TiO 2 /AgNPs hetero-nanostructures were studied.e surface morphology of all the samples was determined by SEM and TiO 2 film thickness and roughness was found enhancing by increasing film sputtering times.e deposition of AgNPs introduced the Ag islands on the TiO 2 films and thus further increased the surface roughness.XRD investigations confirmed the amorphous nature of TiO 2 thin films and FCC crystalline structure of AgNPs.By ellipsometry measurement using Tauc-Lorentz model, film thickness was found to be 37 nm, 51 nm, and 99 nm for sputtering times of 30 min, 60 min, and 90 min, respectively.Optical CO 2 gas detection studies, performed by ellipsometry in alone air and CO 2 gas flow, showed that only TiO 2 /AgNPs hetero-nanostructures exhibited substantial sensing performance as compared to single componentbased structures.
ese results revealed the vital role of hetero-nanostructures in improving the gas-sensing properties by virtue of synergistic effects.

Figure 1 :
Figure 1: Color change during the synthesis of AgNPs depicting di erent stages of the reaction.(a) Dissolving of silver precursors into water remained transparent.(b-d) Introduction of the reducing agent (trisodium citrate) caused the reduction of precursors as appeared in color change from transparent to greenish yellow.

Figure 5 :
Figure 5: XRD (a) and ellipsometry spectra (b) of TiO 2 thin lms for di erent deposition times.

Figure 8 :Figure 9 :
Figure 8: Transmission spectra of TiO 2 thin lms with thickness of 37 nm, 51 nm, and 99 nm in air (a) and CO 2 gas ow (b).

TiO 2 (TiO 2 (Figure 10 :
Figure 10: Comparison of optical responses of all samples in air and CO 2 gas ow.

Table 2 :
Calculated refractive indices for di erent deposition times of TiO 2 thin lms.