Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2026 Aug 19;11(34):52079–52087. doi: 10.1021/acsomega.6c07443

Understanding the Anomalous Gas-Concentration-Driven Response Polarity Transition Behavior of a Room-Temperature-Operated TiO2‑Based Alcohol Sensor

Yizheng Liu †, Xuelan Cheng †, Hongli Zhu †, Shuai Chen †, Ziliang Huang †, Yan Li †, Shuang Xiao †, Yonghui Deng ‡, Fang Xu †,*
PMCID: PMC13625166  PMID: 42819873

Abstract

Emerging p-type gas response of n-type oxide-based resistive gas sensors challenges classical theory, which posits that gas response polarity aligns with the metal oxide’s conductivity type. This work investigates the anomalous gas-concentration-driven gas response polarity transition to alcohols with various chain lengths of a conventional TiO2-based gas sensor at room temperature. A particular focus is on the relationship between the critical concentration and the gas molecular structures. The sensor shows anomalous p-type responses to four alcohols (methanol, ethanol, propanol, and butanol) and transitions to n-type as the alcohol concentration increases to approximately 9–20 K ppm, depending on the alcohol species. Interestingly, this critical concentration decreases generally as the alcohol chain length increases. This concentration-driven response polarity transition is attributed to a shift in the reaction sites governed by the gas concentration, specifically from the adsorbed water layer to the O2 – on the TiO2 surface. The p-type response is attributed to the disruption of proton migration in the adsorbed water layer on the TiO2 surface by alcohol molecules. This process becomes saturated as the alcohol concentration increases and the reaction with the O2 – on the TiO2 surface progressively regains dominance, restoring the n-type response. Longer-chain alcohols show stronger hydrophobicity, reaching water saturation at lower concentrations, thereby showing a lower critical concentration. This work links the anomalous concentration-driven gas response polarity transition to alcohol chain length, providing critical theoretical insights into anomalous gas-sensing behaviors.

graphic file with name ao6c07443_0008.webp

graphic file with name ao6c07443_0006.webp

Introduction

Gas sensors demonstrate broad applications in various fields such as environmental monitoring, , healthcare, and smart home systems. Metal oxide semiconductor (MOS) based gas sensors are distinguished by their low cost, ease of integration, and high response. − Reducing gases like ethanol are typically oxidized, releasing electrons to the n-type MOS and thereby decreasing the sensor’s resistance. , Conversely, oxidizing gases such as NO2 tend to extract electrons from the n-type oxide and increase its resistance. , This type of response is defined as an n-type response. In contrast, opposite response behaviors occur in p-type MOS, , termed a p-type response. Thus, within the classical theory framework, the gas response type generally aligns with the semiconductor type of sensing materials.

However, emerging p-type gas responses in n-type MOS reveal the limitations of the classical sensing mechanism. , Bagheri et al. found that ZnO-based sensor showed a p-type response to CO2 at room temperature. Liu et al. observed an anomalous p-type response of TiO2 to alcohols between room temperature and 120 °C. Nguyet et al. reported that V2O5 exhibited a p-type and an n-type response to NH3 at working temperatures below and above 50 °C, respectively. In our previous work, we also documented room-temperature anomalous p-type responses of n-type TiO2 to several gases. , These reports indicate that anomalous p-type responses usually appear at low temperatures, particularly room temperature. However, the mechanisms for the anomalous p-type gas responses among the above works are different. Bagheri attributed the p-type response to CO2 at room temperature to the p-type conductivity of ZnO due to the presence of Zn defects. Liu attributed the p-type response to ethanol of TiO2 to proton conduction within the water layer adsorbed on the TiO2 surface. Ethanol molecules hinder proton transport, leading to an increase in resistance. Nguyet attributed the p-type response to NH3 of V2O5 at room temperature to the adsorption of oxygen species that capture electrons from the semiconductor surface, thereby forming a p-type inversion layer. Therefore, the mechanism for the anomalous p-type response of an n-type MOS-based gas sensor remains unclear. Besides, the reported gas response polarity switching phenomena in these studies are mainly driven by varying operating temperature. The influence of gas-related parameters, such as gas concentration and molecular structure, remains insufficiently understood.

Room-temperature-operated gas sensors attract great attention due to their advantages of low power consumption, smaller device size for higher integration, improved safety, and compatibility with flexible substrates. , Therefore, further exploration of the anomalous p-type gas response phenomenon of MOS-based sensors working at room temperature is of great significance. However, the gas response polarity depends on the properties of target gases. For example, VO2 materials were found to exhibit an anomalous p-type response to NH3, while showing n-type responses to H2S, H2, NO, toluene, and ethanol. We also observed that the n-type TiO2 exhibits anomalous p-type responses to reducing alcohols, ketones, aldehydes, and oxidizing NO2, yet an n-type response to NH3. This indicates that the property of the target gas plays an important role in determining the gas response type of MOS-based sensors. Furthermore, for gas species that induce anomalous p-type responses, the increase in concentration can shift the p-type response toward n-type behavior. ,

Therefore, the gas response type is influenced by both the gas concentration and the gas species. However, the relationship between the gas-concentration-driven response polarity transition and the molecular structure of gas molecules, particularly the influence of molecular chain length, has not been systematically investigated. Thus, this study selects alcohols as the target gases and TiO2 as the sensing medium to investigate the variation patterns of response types to alcohols at room temperature as a function of concentration. Alcohols were selected as the target gases because they are widely used in industrial and daily applications. Additionally, alcohols feature diverse and tunable molecular structures, with variations in chain length, providing an ideal model for studying the effect of molecular structure on sensing performance. , TiO2 was selected because it is a representative material in MOS-based gas sensors. ,

In this study, anatase burr-like TiO2 nanorods were synthesized for the fabrication of room-temperature-operated gas sensors. The gas-concentration-driven p-n transition phenomenon in its response to the alcohols is investigated with an aim to elucidate the relationship between the molecular structure of alcohols and the transition critical concentration. The variation pattern of the critical concentration with the molecular structure of alcohols is reported, as is the mechanism behind it.

Experimental Section

The synthesis procedure for anatase burr-like TiO2 nanorods is shown in Figure . 0.1 mol/L (NH4)2TiF6 was mixed with 0.3 mol/L H3BO3 at room temperature. ZnO nanowires were added to the mixture, which was homogenized using magnetic stirring to form a reaction solution with well-dispersed ZnO nanowires. After standing for 60 h in a laboratory fume hood at room temperature (around 25 °C, the temperature in the lab is monitored by both the screen of the fume hood and a temperature monitor in the lab), the white product at the bottom of the solution was separated via centrifugation, washed with deionized (DI) water, and collected as Burr-TiO2 nanorods. The fabrication and characterization details could refer to our previous work.

1.

1

Open in a new tab

Schematic of the synthesis process for the Burr-TiO2 nanorods and the fabrication of a gas sensor.

For sensor fabrication, 10.0 mg of the as-synthesized TiO2 was dispersed in 5.0 mL of deionized water to form a slurry. 4.0 μL of this slurry was drop-casted onto a 2 mm × 2 mm ceramic electrode substrate (Beijing Ailite Technology Co., Ltd.), which has a pair of interdigitated Pt electrodes with a minimum spacing of 100 μm. The substrate was then dried on a hot plate at 70 °C for 5 min, followed by annealing at 400 °C in air for 3 h. The sensing medium is denoted as B-TiO2-NRs. The annealed ceramic chip was subsequently mounted onto a six-pin platform, and the sensor device was obtained.

The field emission scanning electron microscope (FESEM, GeminiSEM 300, Carl Zeiss Microscopy Ltd., England) and transmission electron microscope (TEM, Talos F200X G2, Thermo Scientific, Czech) were utilized to obtain scanning electron microscope (SEM) images, transmission electron microscope (TEM) images, high-resolution transmission electron microscope (HRTEM) images, and energy-dispersive X-ray spectroscopy (EDS) images for characterizing the morphology and elemental composition of the B-TiO2-NRs. The crystalline phase of B-TiO2-NRs was analyzed using X-ray diffraction (XRD, SmartLab XRD, Rigaku Corporation, Japan) with Cu Kα radiation (λ = 1.54056 nm). Absorption spectra were measured by an Ultraviolet–visible (UV–vis) spectrophotometer (Lambda1050+, PerkinElmer, USA) to calculate the band gap of B-TiO2-NRs. The chemical composition of B-TiO2-NRs and the energy gap between the valence band and the Fermi level were determined through X-ray photoelectron spectroscopy (XPS, Escalab Xi+, Thermo Fisher Scientific, USA) using Al Kα as the X-ray source at a power of 25 W. The surface states of B-TiO2-NRs were analyzed using a Fourier Transform infrared (FTIR) spectrometer (Nicolet iN10, Thermo Scientific, USA). The complex impedance spectroscopy (CIS) measurements were acquired using the electrochemical workstation (Jiangsu Donghua Testing Technology Co., Ltd. DH7000) to characterize the impedance of the sensor in air and alcohol atmospheres.

Gas-sensing measurements were conducted using a commercial gas-sensing measurement system, WS-30B (Zhengzhou Weisheng Electronics Technology Co., Ltd.), by monitoring the resistance variations of the sensor under various atmospheres. A static gas distribution method was employed, where specific gas concentrations were generated by vaporizing alcohol reagents

C=(22.4×d×ρ×Q)(M×V)×109 1

where C represents the concentration of the alcohol gas (ppm), d denotes the volume fraction of the alcohol reagent which is provided by the manufacturer of these alcohol reagents, ρ stands for the density of the alcohol liquid (g/mL), Q refers to the volume of the alcohol liquid (mL), M indicates the molar mass of the alcohol (g/mol), and V represents the volume of the chamber (mL).

The response of the sensor is defined as follows: when R a < R g, response=RgRa ; when R a > R g, response=RaRg ; (R a and R g represent the sensor’s resistance before and after gas exposure, respectively).

Results and Discussion

The XRD pattern of B-TiO2–NRs, as presented in Figure (a), exhibits diffraction peaks consistent with anatase TiO2 (JCPDS No. 21–1272). SEM and TEM images (Figure (b,c)) reveal a burr-like nanorod morphology, with diameters and lengths around 200 and 1000 nm, respectively. The HRTEM image (Figure (d)) displays lattice fringes with a spacing of 0.35 nm, corresponding to the (101) plane of anatase-phase TiO2. , The EDS mapping results (Figure (e,f)) show that Ti and O elements are uniformly distributed. The elemental composition and chemical states of B-TiO2-NRs were analyzed by XPS. The Ti 2p spectrum was deconvoluted as presented in Figure (g). Peaks at 458.89 and 464.55 eV are assigned to Ti4+ 2p3/2 and Ti4+ 2p1/2, respectively. Peaks at 458.21 and 461.35 eV correspond to Ti3+ 2p3/2 and Ti3+ 2p1/2, respectively. The presence of Ti3+ suggests the existence of electron-rich defects associated with oxygen vacancies. These defect states facilitate oxygen adsorption and electron transfer, contributing to the formation of reactive oxygen species involved in the n-type response. Figure (h) shows the deconvoluted O 1s XPS spectrum. The peak centered at 530.01 eV is attributed to lattice oxygen (Ti–O). The peak centered at 530.55 eV is assigned to oxygen vacancies (OV), which are due to nonstoichiometric defects. The peak centered at 532.21 eV is ascribed to surface-adsorbed hydroxyl groups (−OH). − The OV can serve as active sites for the adsorption of reactive oxygen species (O2 –), and −OH indicates the surface hydrated layer.

2.

2

Open in a new tab

(a) XRD pattern, (b) SEM image, (c) TEM image, (d) HRTEM image of the B-TiO2-NRs; EDS elemental mapping of Ti (e) and O (f); XPS spectra of Ti 2p (g) and O 1s (h).

The film thickness is around 11.5 μm, which is characterized by the cross-sectional SEM image as shown in Figure S1. To verify the fabrication consistency and reproducibility of the as-fabricated devices, 5 independent gas sensors were prepared under identical experimental conditions. The resistance baseline and sensing response toward 500 ppm n-butanol were characterized at room temperature, and the corresponding results are presented in Figure S2. The 5 sensors exhibit consistent baseline resistance and sensing performance, demonstrating the device-to-device reproducibility. Accordingly, sensors with comparable resistance baselines and sensing response characteristics were selected for the subsequent sensing measurements. The dynamic response curves to the four alcohols with increasing concentrations are shown in Figure (a–d), and their corresponding response magnitudes and polarities are summarized in Figure (e–h). When the sensor is subject to analysis, its sensing behavior at room-temperature in methanol atmospheres (Figure a), compared to its resistance baseline in air (∼750 MΩ), the sensor’s resistance increases with the methanol concentration in the range of 500–4000 ppm, indicating a p-type response. This p-type response enhances as the methanol concentration increases, marked with a red background (Stage I). As the methanol concentration reaches about 10 K ppm, a decline appears in the resistance after the initial increase (green lines) and the final resistance is ∼1030 MΩ, remaining above the baseline resistance, maintaining a p-type response. This response characteristic maintains until the methanol concentration reaches 20 K ppm, with a remarkable decline as the gas concentration increases, indicating a weakened p-type response, marked with a green background (Stage II). After 20 K ppm, a similar response remains with a decline in the resistance after the initial increase (blue lines). However, the final resistance falls below the baseline, ∼600 MΩ and ∼580 MΩ for 25 K ppm and 30 K ppm methanol, respectively. This represents increased n-type responses with the increased gas concentration marked with a blue background (Stage III).

3.

3

Open in a new tab

Dynamic response curves to various concentrations of (a) methanol, (b) ethanol, (c) n-propanol, and (d) n-butanol. Response–concentration relationships for (e) methanol, (f) ethanol, (g) n-propanol, and (h) n-butanol.

The dynamic response curves of the sensor toward ethanol (Figure b), n-propanol (Figure c), and n-butanol (Figure d) display similar characteristics to those of methanol. They all experience three stages as the gas concentration increases. At stage I, the resistance of the sensor increases first, which represents typical p-type responses. This p-type response increases as the gas concentration increases. At stage II, resistance increases first, followed by a decrease, weakening the p-type response, which declines further with increasing concentration. This stage starts from 10 K ppm for methanol, 4 K ppm for ethanol, 4 K ppm for n-propanol, and 2 K ppm for n-butanol, respectively, showing a decreased threshold concentration as the alcohol chain length increases. At stage III, the decline in their resistance strengthen, which causes the resistance to be lower than the baseline, indicating that the responses shifted back to n-type. The gas concentration at which the response type switches to n-type is defined as the critical concentration in this work, that is, the start gas concentration for stage III in Figure . They are approximately 20 K ppm, 20 K ppm, 15 K ppm, and 9 K ppm, respectively, showing a general decreasing trend as the alcohol chain length increases as well. Specifically, there is a plateau between methanol and ethanol before decreasing for longer-chain alcohols. A clearer demonstration of the aforementioned three-stage variation trend can be seen in Figure (e–h), including increased p-type responses (stage I), decreased p-type responses (stage II), and increased n-type responses (stage III). It is particularly worth noting that because the saturated vapor pressure of n-butanol is 9 K ppm, the maximum achievable gas-phase concentration of n-butanol should theoretically be 9 K ppm. Therefore, the maximum gas concentration of n-butanol is located around 9 K ppm in Figure d,h. 9 K ppm+ indicates a concentration slightly higher than 9 K ppm. Given that the p-type response of the sensor to n-butanol around 9 K ppm has already dropped to 1, it can be concluded that the critical concentration for n-butanol is 9 K ppm. To further demonstrate this, we conducted a measurement by injecting more n-butanol after 9 K ppm, namely, 9 K ppm+. Although theoretically, the gas-phase concentration is also 9 K ppm, in terms of butanol molecules interacting with the TiO2 surface, the 9 K ppm+ condition provides slightly more n-butanol molecules than the 9 K ppm condition does, which explains why the slight p-type response at 9 K ppm transitions to a slight n-type response at 9 K ppm+.

Based on the empirical gas-sensing mechanism, TiO2 is usually n-type, exhibiting n-type gas responses. Therefore, to clarify this unusual p-type gas response in this work, it is first necessary to characterize the conductivity type of the B-TiO2-NRs. UV–vis and valence band X-ray photoelectron spectroscopy (VB-XPS) measurements were performed on B-TiO2–NRs to determine the Fermi level, conduction band minimum (CBM), and valence band maximum (VBM). Figure presents the optical absorption and electronic structure characterization results of the B-TiO2-NRs. As shown in Figure (a), the UV–vis absorption spectrum of B-TiO2–NRs reveals an absorption edge at approximately 388 nm. The band gap is 3.13 eV, which is calculated from the Tauc plot model as shown in Figure (b). This aligns with the general band gap of TiO2. Combined with the VB-XPS (Figure (c)), the VBM of B-TiO2-NRs is located 2.28 eV below the Fermi level. The corresponding energy band diagram is illustrated in Figure (d), indicating that the surface of the B-TiO2–NRs is n-type.

4.

4

Open in a new tab

(a) UV–vis absorption spectrum, (b) Tauc plot showing the extracted band gap, (c) VB-XPS spectrum, and (d) energy band diagram of B-TiO2-NRs.

Therefore, the observed p-type responses to the alcohols and the concentration-driven p-n response type transition could not be explained by the classical gas-sensing theory. Considering the sensor is operated at room temperature at a humidity of 55%, it may be related to the water layer adsorbed on the TiO2 surface. To verify this, Fourier Transform infrared spectroscopy (FTIR) was used to characterize the B-TiO2-NRs surface (Figure (a)). A broad and strong absorption peak appears around 3400 cm–1, typically associated with O–H stretching vibrations. Additionally, a weaker absorption peak around 1630 cm–1 corresponds to H–O–H bending vibrations, confirming the presence of physically adsorbed water molecules on the TiO2 surface. , Therefore, the anomalous p-type response to alcohols in stage I in this work might be explained as follows: Initially, proton conduction occurs within the adsorbed water layer on the TiO2 surface (as shown in eq ). When alcohol molecules contact the TiO2 surface, due to their higher proton affinity compared to H2O molecules, the alcohols extract protons from water to form alcohol ions (as shown in eq ). Since CnH2n+1OH2 + (H2O)2 possesses larger ionic radii with lower diffusion coefficients than H3O+, the ionic conductivity decreases and the sensor’s resistance increases, resulting in the anomalous p-type responses in stage I. This proposed proton-exchange mechanism was further supported by CIS measurements. The CIS of the sensor in air and 500 ppm n-butanol are shown in Figure S3. The red line corresponds to the Nyquist curve in air. It exhibits a straight line in the low-frequency region, which is characteristic of Warburg impedance (Z w), indicating the presence of ionic transport behavior. Combined with the FTIR results mentioned above, which demonstrated water molecule adsorption on the TiO2 surface, this ionic conduction behavior can be attributed to the dissociation of adsorbed surface water, generating H3O+, which undergoes interfacial hopping and continuous diffusive transport via the Grotthuss mechanism. , The blue line corresponds to the Nyquist impedance spectrum of the sensor when exposed to 500 ppm n-butanol. The semicircle radius of the curve increases, and Zw appears at lower frequencies. This provides direct evidence that the interaction between alcohols and water molecules reduces the ionic conductivity of the water layer.

H2O+H3O+→H3O++H2O 2
H3O+(H2O)2+CnH2n+1OH→CnH2n+1OH2+(H2O)2+H2O 3

Based on this mechanism, the concentration-induced p-n response transition could be explained as follows: as the alcohol concentration increases, the reaction between the alcohol and the water layer becomes saturated, and excessive alcohol begins to react with the O2 – adsorbed on the TiO2 surface. This reaction releases electrons back to the TiO2, thereby reducing their electrical resistance as shown in eq . The O2 – redox reactions and the proton disruption process coexist, and the competition between the two reactions appears. At the beginning, the reaction between the alcohols and the surface water layer still dominates. Therefore, the sensor remains p-type response to alcohols while the p-type response magnitudes decrease as more alcohols participate in the redox reactions by the O2 – adsorbed on the TiO2 surface. This corresponds to stage II in Figure . As the alcohol concentration increases further, the reaction between the alcohols and the surface O2 – is strengthened. Finally, the O2 – redox reaction overcomes the proton disruption process and regains the dominance. This corresponds to stage III, where the response polarity transitions to n-type.

CnH2n+1OH+O2−→TiO2CO2+H2O+e− 4

In conclusion, the concentration-driven response polarity transition of B-TiO2-NRs-based gas sensor toward alcohols is attributed to the competition between proton disruption and O2 – redox process as summarized in Figure (b), using n-butanol as the example. The alcohol molecules preferentially interact with water molecules adsorbed on the TiO2 surface. This increases the sensor’s resistance by reducing the ion diffusion coefficient within the water layer, thereby exhibiting a p-type response in stage I. When this reaction approaches saturation as the alcohol concentration increases, the redox reaction between alcohols and the surface-adsorbed oxygen species on TiO2 begins (Stage II). This will release electrons to TiO2, thereby reducing its electrical resistance. As the alcohol concentration increases, the redox reaction between alcohols and the surface-adsorbed oxygen species strengthens, causing the decreased p-type response in stage II and progressively becomes dominant, causing the n-type response in stage III.

5.

5

Open in a new tab

(a) FTIR spectrum of B-TiO2-NRs. (b) Mechanism for the concentration-driven p-n response transition of B-TiO2-NRs to alcohols. (c) The hydrophilicity of the four alcohols. (d) Resistance of the sensor at various humidity levels, the inset is the magnification for humidity in the range of 55%∼85%. (e) Response to 4 K ppm n-butanol at various humidity levels. (f) Comparison of the response amplitudes toward four different alcohol vapors with concentrations of 500 ppm.

Therefore, the critical alcohol concentration for the p-n response polarity transition is related to the saturation concentration of alcohol molecules that react with the water molecules on the surface of TiO2. Alcohol molecules consist of a hydrophilic hydroxyl group and a hydrophobic alkyl chain. Consequently, as the carbon chain length increases, the hydrophilicity of the alcohol diminishes, i.e., hydrophilicity follows the order: methanol > ethanol > n-propanol > n-butanol, as shown in Figure (c). Accordingly, the saturation concentration for the reaction with the water layer follows the order: methanol > ethanol > n-propanol > n-butanol. Hence, the critical concentration for the p-n response transition decreases progressively with increasing alcohol carbon chain length, which is consistent with the experimental results.

To further validate this proton conduction mechanism, the resistance baseline and response characteristics to 4 K ppm n-butanol were investigated under various humidity conditions (Figure d,e). According to the proton conduction theory, as the humidity decreases, the conductivity of the sensor should decrease, its room-temperature response to alcohols should likewise decrease, and transition to an n-type response. As shown in Figure (d), the resistance of the sensor continuously decreases with increasing relative humidity (10%–85%), consistent with the proposed mechanism. As shown in Figure (e), at relative humidity below 40%, the sensor exhibits a normal n-type response to n-butanol and the response magnitude decreases as humidity increases. After 40% relative humidity, the response switches to anomalous p-type, indicating a humidity-induced n-p response transition, with the response magnitude increasing as humidity decreases. The results are in good agreement with the theoretical predictions, thereby validating the mechanism proposed in this work. Similar humidity-induced n-p response transition behavior was also observed for n-propanol, further confirming this mechanism (Figure S4).

The order of proton affinity is n-butanol (789.2 kJ/mol) > n-propanol (786.5 kJ/mol) > ethanol (776.4 kJ/mol) > methanol (754.3 kJ/mol) > H2O (691 kJ/mol), suggesting that the ability to capture protons decreases accordingly. Thus, theoretically, the room-temperature anomalous p-type response to the four alcohols should follow the same trend: n-butanol > n-propanol > ethanol > methanol. Moreover, as shown in Figure (c), the radii of the alcohol ions decrease in the order n-butanol > n-propanol > ethanol > methanol. A larger ion radius leads to greater hindrance of proton flow, resulting in a larger resistance change, and thus a higher response. This theoretically also leads to the same order of room-temperature anomalous p-type response: n-butanol > n-propanol > ethanol > methanol. To confirm this, the room-temperature responses of the sensor to the four alcohols at concentrations of 500 ppm (Figure f) and 1000 ppm (Figure S5) were compared. These two concentrations are selected because they are the common concentrations among the 4 alcohols in stage I. The observed orders are response (n-butanol) > response (n-propanol) > response (ethanol) > response (methanol), further validating the proposed gas-sensing mechanism. Nevertheless, the relatively small difference in response magnitude among the four alcohols suggests that the present sensing platform still has limitations in achieving high selectivity among similar alcohols.

Conclusions

In summary, a room-temperature-operated alcohol sensor showing an unusual p-type response was fabricated using anatase burr-like TiO2 nanorods as the sensitive layer. The anomalous gas-concentration-driven p-n gas-response-polarity-transition behavior was observed experimentally toward alcohol vapors, including methanol, ethanol, n-propanol, and n-butanol, in this work. The critical concentration decreases generally as the alcohol chain length increases, with a plateau between methanol and ethanol before it decreases for longer-chain alcohols. A proton–electron competing conductivity mechanism is proposed. In detail, alcohol molecules first react with the adsorbed water layer on TiO2, and this reduces the proton conductivity, causing the p-type gas response; as the alcohol concentration rises beyond water saturation, reactions with adsorbed O2 – on the TiO2 surface strengthen and dominate progressively, increasing electron concentration and inducing the decreased p-type gas response and the final increased n-type gas response. In this way, longer-chain alcohols containing stronger hydrophobicity reach water saturation at lower concentrations, causing a decreased critical concentration. This work establishes a new correlation between the gas-concentration-driven anomalous response polarity transition and alcohol chain length in resistive gas sensors using a common anatase TiO2, providing key insights into gas-sensing mechanisms of MOS resistive gas sensors.

Supplementary Material

ao6c07443_si_001.pdf (279KB, pdf)

Acknowledgments

This study was supported by the Guangdong Science and Technology Department [Grant No. MS202600196], the National Natural Science Foundation of China [Grant No. 62001308], the Shenzhen Science and Technology Program [Grant No. RCBS20210609103736094, SGDX20230116093205009], the Industry-University Cooperation Collaborative Education Program [Grant No. 231000532114622], and the Shenzhen Technology Education Reform Program [Grant No. 20251056010029].

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c07443.

  • Cross-sectional SEM image of B-TiO2–NRs (Figure S1); comparison of responses to 500 ppm n-butanol and resistance baseline of 5 independent B-TiO2–NRs sensors (Figure S2); CIS of B-TiO2–NRs in air and n-butanol (Figure S3); response to n-propanol at various humidity levels (Figure S4); and comparison of the response amplitudes toward four different alcohol vapors (Figure S5) (PDF)

F.X.: Conceptualization, methodology, writingreview and editing, supervision, project administration, and funding acquisition. Y.L.: Data curation, writingoriginal draft preparation, and writingreview and editing. X.C.: Methodology and formal analysis. H.Z.: Data curation and writingreview and editing. Z.H.: Data curation and writingreview and editing. S.C.: Investigation and data curation. Y.L.: Investigation and formal analysis. S.X.: Supervision and project administration. Y.D.: Supervision, project administration, and writingreview and editing. All authors discussed the results, analyzed the data, and reviewed the manuscript.

The authors declare no competing financial interest.

References

  1. Zong B., Wu S., Yang Y., Li Q., Tao T., Mao S.. Smart Gas Sensors: Recent Developments and Future Prospective. Nano-Micro Lett. 2025;17(1):54. doi: 10.1007/s40820-024-01543-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Li P., Yu Y., Li P., Wei M., Feng Y., Deng Y., Wei J.. Atomically Dispersed Noble-Metal-Sensitized Semiconductor Metal Oxides for Chemiresistive Gas Sensing. Micro. Nano Sci. 2026;1:3. doi: 10.20517/mns.2025.07. [DOI] [Google Scholar]
  3. Xue L., Li Y., Deng Y.. Multicomponent Porous Metal Oxide Semiconductors for Advanced Gas Sensing. Micro. Nano Sci. 2025;1:2. doi: 10.20517/mns.2025.03. [DOI] [Google Scholar]
  4. Fan C., Yang J., Ni W., Wu J., Liu X., Li Z., Zhang Y., Quan W., Zeng M., Hu N., Fang H., Wang T., Yang Z.. Real-Time and Wireless Transmission of a Nitrogen-Doped Ti3C2Tx Wearable Gas Sensor for Efficient Detection of Food Spoilage and Ammonia Leakage. ACS Sens. 2024;9(9):4870–4878. doi: 10.1021/acssensors.4c01394. [DOI] [PubMed] [Google Scholar]
  5. Yang X., Ju S., Xie K., Deng Y., Zhang Y., Du M.. Understanding and Quantifying the Contribution of Oxygen Species in Hydrogen Sensing by Pd/In2O3 Nanosheets. Angew. Chem., Int. Ed. 2025:e18988. doi: 10.1002/anie.202518988. [DOI] [PubMed] [Google Scholar]
  6. Yang Z., Rong Q., Zhang W., Fan X., Shao M., Li J., Wang C., Guo Y.. A ZnO Packaged MEMS Hydrogen Sensor for Reliable SIBO Breath Analysis with Anti-H2S Interference. Chem. Eng. J. 2025;522:167331. doi: 10.1016/j.cej.2025.167331. [DOI] [Google Scholar]
  7. Zhao N., Chang X., Liu X., Li J., Zheng W., Zhang J.. Probing the Structural Dynamics of In2O3 Using in Situ Raman Spectroscopy: Bridging Material Dynamics and Sensor Functionality. Angew. Chem., Int. Ed. 2025;64(45):e202512808. doi: 10.1002/anie.202512808. [DOI] [PubMed] [Google Scholar]
  8. Wang S., Tan C., Zong Q., Li S., Gao C., Yang H., Huang Q., French P., Ye H.. Selective Reduction Laser Sintering: A New Strategy for NO2 Gas Detection Based on In2 O3 Nanoparticles. Adv. Funct. Mater. 2025;35(25):2419057. doi: 10.1002/adfm.202419057. [DOI] [Google Scholar]
  9. Zhao M., Nitta R., Izawa S., Yamaura J., Majima Y.. Nano-Patterned CuO Nanowire Nanogap Hydrogen Gas Sensor with Voids. Adv. Funct. Mater. 2025;35(12):2415971. doi: 10.1002/adfm.202415971. [DOI] [Google Scholar]
  10. Ren W., Luan J., Yin L., Chen H., Wang C., Zhang P., Cui G., Lv L.. Ultrasensitive Room-Temperature NO2 Gas Sensor Based on MXene–Cu2O Composites. ACS Sens. 2025;10(5):3579–3588. doi: 10.1021/acssensors.5c00215. [DOI] [PubMed] [Google Scholar]
  11. Yang X., Liu X., Zhao N., Chang X., Zheng W., Zhang J.. Beyond Conventional Sensing Paradigms: Multiscale Analysis of Conductivity-Type Switching in Semiconductor Gas Sensors. Coord. Chem. Rev. 2026;549:217286. doi: 10.1016/j.ccr.2025.217286. [DOI] [Google Scholar]
  12. Miao L., Song P., Xue Y., Hou Z., Hasegawa T., Okawa A., Goto T., Seo Y., Maezono R., Sekino T., Yin S.. Novel Selectivity: Target of Gas Sensing Defined by Behavior. Adv. Mater. 2025;37:2413023. doi: 10.1002/adma.202413023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bagheri F., Haratizadeh H., Ahmadi M.. Improving CO2 Sensing and P-n Conductivity Transition under UV Light by Chemo-Resistive Sensor Based on ZnO Nanoparticles. Ceram. Int. 2024;50(1):1497–1504. doi: 10.1016/j.ceramint.2023.10.240. [DOI] [Google Scholar]
  14. Liu C., Lu H., Zhang J., Yang Z., Zhu G., Yin F., Gao J., Chen C., Xin X.. Abnormal P-Type Sensing Response of TiO2 Nanosheets with Exposed {001} Facets. J. Alloys Comp. 2017;705:112–117. doi: 10.1016/j.jallcom.2017.02.118. [DOI] [Google Scholar]
  15. Thi Nguyet T., Van Duy L., Nam N. C., Dat D. Q., Nguyen H., Hung C. M., Van Duy N., Hoa N. D.. Transition from P-Type to n-Type Semiconductor in V2O5 Nanowire-Based Gas Sensors: Synthesis and Understanding of the Sensing Mechanism. Sens. Actuators, B. 2025;424:136841. doi: 10.1016/j.snb.2024.136841. [DOI] [Google Scholar]
  16. Li Y., Cheng X., Zhang S., Huang J., Liu Y., Zhou H., Yang X., Shi J., Zhong A., Zhu Z., Xu F., Zhou C.. Alcohol Selectivity of a Room Temperature-Operated Gas Sensor Based on TiO2 Burr-like Nanorods. Chem. Eng. J. 2025;508:160923. doi: 10.1016/j.cej.2025.160923. [DOI] [Google Scholar]
  17. Cheng X., Liu Y., Zhong W., Li S., Li Y., Zhao Z., Zhang C., Shi J., Liu H., Zhu Z., Xu F.. An Ultra-Selective and Humidity-Resistant Room-Temperature-Operated NO2 Sensor Based on Black TiO2 . Adv. Sci. 2025;12(41):e09293. doi: 10.1002/advs.202509293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kumar S., Mirzaei A., Kumar A., Hoon Lee M., Ghahremani Z., Kim T.-U., Kim J.-Y., Kwoka M., Kumar M., Kim S. S., Kim H. W.. Nanoparticles Anchored Strategy to Develop 2D MoS2 and MoSe2 Based Room Temperature Chemiresistive Gas Sensors. Coord. Chem. Rev. 2024;503:215657. doi: 10.1016/j.ccr.2024.215657. [DOI] [Google Scholar]
  19. Guo C., Yang B., Xin Y., Meng H., Cheng X., Xiao X., Xu Y.. MIL-125­(Ti)-Derived Rutile Phase TiO2 Achieves Highly Sensitive Detection and Mechanism Study of n-Amylamine. Chem. Eng. J. 2025;505:159616. doi: 10.1016/j.cej.2025.159616. [DOI] [Google Scholar]
  20. Huang H., Gong H., Chow C. L., Guo J., White T. J., Tse M. S., Tan O. K.. Low-Temperature Growth of SnO2 Nanorod Arrays and Tunable n–p–n Sensing Response of a ZnO/SnO2 Heterojunction for Exclusive Hydrogen Sensors. Adv. Funct. Mater. 2011;21(14):2680–2686. doi: 10.1002/adfm.201002115. [DOI] [Google Scholar]
  21. Zhou J., Xue K., Liu Y., Liang T., Zhang P., Zhang X., Zhang W., Dai Z.. Highly Sensitive NO2 Response and Abnormal P-N Sensing Transition with Ultrathin Mo-Doped SnS2 Nanosheets. Chem. Eng. J. 2021;420:127572. doi: 10.1016/j.cej.2020.127572. [DOI] [Google Scholar]
  22. Zhao Y., Björk E. M., Yan Y., Schaaf P., Wang D.. Recent Progress in Transition Metal Based Catalysts and Mechanism Analysis for Alcohol Electrooxidation Reactions. Green. Chem. 2024;26(9):4987–5003. doi: 10.1039/D4GC00227J. [DOI] [Google Scholar]
  23. Anderson, G. ; Fahey, C. ; Richards, A. ; Sauer, S. S. ; Wegman, D. ; Booth, J. . Organic and Biochemistry Supplement to Enhanced Introductory College Chemistry; eCampusOntario Pressbooks: Toronto, 2024. [Google Scholar]
  24. Meng F., Xu J., Ji H., Yuan K., Yuan Z.. Dynamic Temperature Modulation Method of Ag-Modified SnO2 Sensor for the Identification of Different Alcohol Homologues. Sens. Actuators, B. 2026;449:139088. doi: 10.1016/j.snb.2025.139088. [DOI] [Google Scholar]
  25. Liu Y., He Y., Liu Z., Cheng X., Zhou C., Shi J., Li Y., Xu F.. One-Step Synthesis of Various Burr-like TiO2 Nanostructures: Nanotubes, Nanorods, Nanotetrapods, Nano-Urchins. Chem. Phys. Lett. 2023;832:140900. doi: 10.1016/j.cplett.2023.140900. [DOI] [Google Scholar]
  26. Qin S., Denisov N., Osuagwu B., Kolařík J., Sarma B. B., Badura Z., Schmuki P.. Minuscule Amounts of Pt Single Atoms Selectively Loaded on Minor (101) Facet of Anatase Crystallites Enables Outstanding Utilization Efficiency for Photocatalytic H2 Production. Adv. Funct. Mater. 2025;35(24):2423088. doi: 10.1002/adfm.202423088. [DOI] [Google Scholar]
  27. Umair M., Ruiz-Aguirre A., Berruti I., Rodriguez S. M., Palmisano L., Loddo V., Bellardita M.. Biomass Derivatives Photoreforming in Pilot Plant Scale to Obtain H2 under Green Conditions by Using Ball Milling Cu2O-TiO2 P25 Photocatalysts. Chem. Eng. J. 2025;504:158585. doi: 10.1016/j.cej.2024.158585. [DOI] [Google Scholar]
  28. Jung S., Sung M., Ju B., Yoon H., Park Y., Kim D.. Phase-Guided Assembly of RuO2 Epitaxial Films on TiO2 Nanofibers: Interfacial Coherence Outperforming Defect-Driven Mismatch for Active Site Generation in Li–O2 Batteries. Adv. Funct. Mater. 2025:e24403. doi: 10.1002/adfm.202524403. [DOI] [Google Scholar]
  29. Mekasuwandumrong O., Chaitaworn S., Panpranot J., Praserthdam P.. Photocatalytic Liquid-Phase Selective Hydrogenation of 3-Nitrostyrene to 3-Vinylaniline of Various Treated-TiO2 Without Use of Reducing Gas. Catalysts. 2019;9(4):329. doi: 10.3390/catal9040329. [DOI] [Google Scholar]
  30. Zhang Y., Xu Z., Li G., Huang X., Hao W., Bi Y.. Direct Observation of Oxygen Vacancy Self-Healing on TiO2 Photocatalysts for Solar Water Splitting. Angew. Chem., Int. Ed. 2019;58(40):14229–14233. doi: 10.1002/anie.201907954. [DOI] [PubMed] [Google Scholar]
  31. He J., Qi X., Liu J., Zhang H., Wang Z., Duo S.. Coaxial Electrospinning In2O3/TiO2 Core-Shell Nanofibers for Enhanced Triethylamine Sensing. Sens. Actuators, B. 2025;443:138313. doi: 10.1016/j.snb.2025.138313. [DOI] [Google Scholar]
  32. Wei T., Yin J., Zhang H., Luo M., Zhao J., Liu J., Huang B., Wu Z., Zhou H., Liu Y., He C., Lai B.. Layered Mesoporous TiO2 Loaded CQDs Activation of Periodate to Remove Sulfamethoxazole: Key Role of CQDs. Chem. Eng. J. 2025;519:165217. doi: 10.1016/j.cej.2025.165217. [DOI] [Google Scholar]
  33. Wu Y., Ge C., Zhao Y., Liu S., Zhang X., Liu G., Wang M.. Phosphorus-Doped TiO2 Nanoflowers for Ppb-Level and Highly Selective Detection of Oxygen Gas. Sens. Actuators, B. 2025;444:138431. doi: 10.1016/j.snb.2025.138431. [DOI] [Google Scholar]
  34. Srimurugan V., Suryakumar C., Jha S., Raj C. C., Prasanth R.. Synergistic Effects of MoS2/TiO2 Nanotubes p-n Heterojunction Photoelectrode for Hydrogen Evolution. Int. J. Hydrog. Energy. 2025;103:609–623. doi: 10.1016/j.ijhydene.2025.01.230. [DOI] [Google Scholar]
  35. Makuła P., Pacia M., Macyk W.. How To Correctly Determine the Band Gap Energy of Modified Semiconductor Photocatalysts Based on UV–Vis Spectra. J. Phys. Chem. Lett. 2018;9(23):6814–6817. doi: 10.1021/acs.jpclett.8b02892. [DOI] [PubMed] [Google Scholar]
  36. Mino L., Negri C., Santalucia R., Cerrato G., Spoto G., Martra G.. Morphology, Surface Structure and Water Adsorption Properties of TiO2 Nanoparticles: A Comparison of Different Commercial Samples. Molecules. 2020;25(20):4605. doi: 10.3390/molecules25204605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Belhadj H., Hakki A., Robertson P. K. J., Bahnemann D. W.. In Situ ATR-FTIR Study of H2O and D2O Adsorption on TiO2 under UV Irradiation. Phys. Chem. Chem. Phys. 2015;17(35):22940–22946. doi: 10.1039/C5CP03947A. [DOI] [PubMed] [Google Scholar]
  38. Hunter E. P. L., Lias S. G.. Evaluated Gas Phase Basicities and Proton Affinities of Molecules: An Update. J. Phys. Chem. Ref. Data. 1998;27(3):413–656. doi: 10.1063/1.556018. [DOI] [Google Scholar]
  39. Linstorm, P. NIST Chemistry WebBook, NIST Standard Reference Database 69. In National Institute of Standards and Technology; National Institute of Standards and Technology, 1997. [Google Scholar]
  40. Dai L., Yu H., Wang Y., Zhang X., Xu F., Pan X., Luo J.-T., Zhong A.. Ti3C2T x MXene Composite with Much Improved Stability for Superior Humidity Sensors. Langmuir. 2025;41(2):1368–1376. doi: 10.1021/acs.langmuir.4c04107. [DOI] [PubMed] [Google Scholar]
  41. Prakasha B. S., Kumar A., Mishra J., Tan P., Wang C., Rodríguez-Viejo J., Kumar M., Sledzinska M.. Neural-Network-Assisted Impedance Analysis for Humidity and Ammonia Detection Using MXene and PtSnO2 Sensors with Cross-Interference Suppression. ACS Sens. 2026;11(2):1741–1751. doi: 10.1021/acssensors.5c02376. [DOI] [PubMed] [Google Scholar]
  42. Fu C., Zhang Y., Sun X., Fang F., Huang W.. Role of Water in Methanol Photochemistry on TiO2 Nanocrystals: An In Situ DRIFTS Study. J. Phys. Chem. C. 2022;126(20):8615–8626. doi: 10.1021/acs.jpcc.2c00568. [DOI] [Google Scholar]
  43. Yaws C. L., Hopper J. R., Sheth S. D., Han M., Pike R. W.. Solubility and Henry’s Law Constant for Alcohols in Water. Waste Manag. 1998;17(8):541–547. doi: 10.1016/S0956-053X(97)10057-5. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao6c07443_si_001.pdf (279KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES