Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Aug 11;5(33):21104–21112. doi: 10.1021/acsomega.0c02750

Ethanol Gas Sensing by a Zn-Terminated ZnO(0001) Bulk Single-Crystalline Substrate

Taku T Suzuki †,*, Takeshi Ohgaki †, Yutaka Adachi †, Isao Sakaguchi †, Minoru Nakamura ‡, Hideyuki Ohashi ‡, Akihisa Aimi ‡, Kenjiro Fujimoto ‡
PMCID: PMC7450640  PMID: 32875247

Abstract

graphic file with name ao0c02750_0007.jpg

Metal oxide semiconductor gas sensors have been widely studied for the selective detection of various gases with trace concentrations. The identification of the reaction scheme governing the gas sensing response is crucial for further development; however, the mechanism of ethanol (EtOH) gas sensing by ZnO is still controversial despite being one of the most intensively studied target gas and sensing material combinations. In this work, for the first time, the detailed mechanism of EtOH sensing by ZnO is studied by using a bulk single-crystalline substrate, which has a well-defined stoichiometry and atomic arrangement, as the sensing material. The sensing response is substantial on the ZnO substrate even with a millimeter-size thickness, and it becomes larger with resistance of the substrate. The large sensing response is described in terms of the adsorption/desorption of the oxygen species on the substrate surface, namely, oxygen ionosorption. The valence state of the ionosorbed oxygen involved in EtOH sensing is identified to be O2– regardless of the temperature. The increase in the sensing response with the temperature is attributed to the enhanced oxidation rate of the EtOH molecule on the surface as analyzed by pulsed-jet temperature-programmed desorption mass spectrometry, which has been newly developed for analyzing surface reactions in simulated working conditions.

Introduction

Metal oxide-based gas sensors have been widely studied for various practical applications, such as mobile gas detectors. The fundamental gas sensing mechanism of ZnO, which is a typical sensing material, has been discussed for many years.1,2 An understanding of the detailed gas sensing mechanism is needed to develop a versatile gas sensor that enables selective gas detection with high sensitivity in the future.

Briefly, a metal oxide-based gas sensor detects the change of electrical resistance with a reduction/oxidation reaction that is induced by the target gas on the sensing material. The typical operating temperature is approximately 300 °C. It is generally accepted that the molecular oxygen in the air dissociatively adsorbs on the surface at elevated temperatures (above 100–200 °C), thus resulting in the adsorption of negatively charged atomic oxygen on the surface, which is referred to as oxygen ionosorption.3 The resistance of the sensing material changes according to the oxygen ionosorption as seen in the following equations2−4

graphic file with name ao0c02750_m001.jpg 1

and

graphic file with name ao0c02750_m002.jpg 2

where (gas) and (ads) denote gas and adsorbate molecules, respectively. Equations 1 and 2 indicate that conduction band electrons are trapped at the ionosorbed oxygen, causing the sensing material to have a large electrical resistance. In contrast, the resistance decreases with the elimination of the ionosorbed oxygen as a result of a reaction with a reductive target gas because the released electrons return to the conduction band. The ethanol (EtOH) sensing by ZnO is considered as a surface reductive reaction, and the partial reaction pathway is respectively written for O–(ads) and O2–(ads) as4,5

graphic file with name ao0c02750_m003.jpg 3
graphic file with name ao0c02750_m004.jpg 4

The reaction pathways including the above equations have been discussed by observing the reaction product using, for example, gas chromatography-related techniques5 and the surface-sensitive temperature-programmed desorption (TPD).6−9 Although EtOH sensing by ZnO is one of the most intensively studied target gas—sensing material combinations from both experimental and theoretical approaches,4,5,10−23 the central working mechanism of the gas sensing remains controversial.

The controversy concerning the sensing mechanism of EtOH by ZnO originates from the following two open questions: (i) what is the electrical conducting mechanism of ZnO involved in the EtOH sensing? and (ii) what is the reaction pathway of the adsorbed EtOH molecule on the ZnO surface?

Concerning the first question, ZnO is a naive n-type semiconductor because of oxygen vacancies and zinc interstitials. The unintentional impurities have also been considered as a source of the conductivity.24 The superior transport properties of ZnO, such as its high electron mobility, wide and direct band gap, high thermal conductivity, and large exciton binding energy, have attracted attention in many practical applications. Although a number of studies have attempted to reveal and control the electrical conductivity, even the origin of the unintentional n-type conductivity remains unclear. This is because it is particularly difficult to distinguish between the native point defects, such as oxygen vacancies and zinc interstitials, and unintentional impurities in the electrical conductivity of ZnO.24 However, gas sensing is due to changes in electrical conductivity; hence, to understand the sensing mechanism, it is crucial to reveal the origin of the conductivity related to gas sensing.

The second question is deeply related to the chemical states of the ionosorbed oxygen species (O2–, O22–, O–, and O2–), which are generally considered to be dependent on the temperature.3 A number of studies have indicated that molecular species are dominant below from 100 to 200 °C, while they dissociate into atomic species at high temperatures on sensing materials, such as ZnO and SnO2.1,4,25,26 The identification of these oxidation states is essential for revealing the reaction pathway, as seen in eqs 1–4. However, the mechanism is still unclear because there are no direct observations of ionosorbed oxygen species by a surface-sensitive analytical technique.27 Because of the so-called “pressure gap” problem, it is usually difficult to apply surface-sensitive analytical techniques that require a high-vacuum environment to a working gas sensor surface under atmospheric pressure.28,29

The gas sensing properties have typically been evaluated with powder and film samples because gas sensors in these forms have been widely used in practical applications. The microscopic structures of these samples consist of grains, grain boundaries, pores, and surfaces, and each of these structures has different electrical conducting properties. Indeed, the gas sensing response depends on the grain size.30 Moreover, the temperature-dependent microstructure affects the pathway of the electron carriers. Thus, the complicated structure of the sensing material hinders a straightforward analysis on the fundamental gas sensing mechanism. On the other hand, a clear-cut analysis is expected using a single-crystalline bulk substrate with a well-defined stoichiometry and atomic arrangement. On the basis of this motivation, theoretical studies concerning the gas sensing on a single-crystalline metal oxide surface have been recently reported.31 On the other hand, to the best of our knowledge, no experimental study has been reported for the gas sensing on a ZnO single-crystalline bulk substrate, although a high-quality ZnO single crystal without grain boundaries grown by the hydrothermal method is rather easily available in recent years. This probably comes from the fact that no substantial gas sensing response is anticipated on a thick single-crystalline substrate because the contribution of the surface is intuitively considered to be very small in the electrical conduction as below.

For evaluating gas sensing properties, the sensing response Rair/Rgas has been conventionally measured by the voltage drop at the reference resistance in the dc two-terminal method, where Rair and Rgas denote the resistance of the sample in air and in the target gas, respectively. The sensitivity S is defined by the saturated Rair (Rairsaturate) and saturated REtOH (REtOH) as

graphic file with name ao0c02750_m005.jpg 5

The relationship between the sensitivity and thickness of a sensing material has been explained from the model in terms of the electron-depleted space charge layer formed upon oxygen ionosorption. In this model, the sensitivity becomes large if the thickness of a sensing material is comparable to or less than double of the space charge layer thickness.32 This is because the electrical conductivity of the fully depleted sensing material is extremely sensitive to the change of the charge caused by the target gas adsorption.25 The space charge layer model has successfully explained the large sensitivity of a sensing material with a size of nanometers to a few tens of nanometers, which is the typical thickness of the space charge layer.25,33

The thickness W of the space charge layer is described as

graphic file with name ao0c02750_m006.jpg 6

where εr is the relative permittivity of the sensing material, ε0 is the dielectric constant of vacuum, φ is the surface barrier potential, e is the electron charge, and ne is the charge carrier density.34,35 For ZnO, it is reported to be 43 nm for a nanowire35 and 400 nm for a single-crystalline substrate at 300 °C.36 The large W of the single crystal is due to the low density of the defect-originated charge carrier. We are not aware of the value for a ZnO single-crystalline thin film in the literature, but it should be smaller than 400 nm of a single-crystalline substrate because the defect density is lower for the bulk single crystal than for the crystalline film grown on a substrate. Because the sensitivity to EtOH of an epitaxially grown ZnO single-crystalline film decreases with the thickness above 78 nm (Figure S1), the value for a single-crystalline thin film is considered to be several tens of nanometers.

The EtOH (50 ppmv) sensing sensitivity by a crystalline ZnO thin film with a thickness comparable to the space charge layer thickness (several tens of nanometers) has been reported to be in the range of 10–20 (Figure 1 and refs (22) and (37)). By taking this value into account with a space charge layer thickness of 400 nm for the single-crystalline substrate,36 the EtOH sensing sensitivity is estimated to be smaller than 10–2 on a ZnO bulk crystalline substrate with a thickness of 1 mm. Thus, the EtOH sensing response by a ZnO single-crystalline bulk substrate should be negligibly small. However, it is actually not the case as described in this paper.

Figure 1.

Figure 1

Open in a new tab

(a) Resistance R of the ZnO substrate as a function of the annealing time in air at 400 °C; (b) normalized resistance change as a function of the annealing time in high vacuum, where ΔR denotes the change from the initial resistance; (c) normalized resistance change as a function of the irradiation time of pulsed air or N2 gas jets in high vacuum at 300 °C; and (d) He+ LEIS spectra after annealing at 100 and 600 °C in UHV.

In this paper, we report the EtOH gas sensing by a ZnO bulk single-crystalline substrate with a thickness of 0.5 mm. The substrate surface was parallel to Zn-terminated (0001), which has been proposed to be an appropriate crystallographic plane for sensitive gas sensing.12 It was found that the EtOH sensing response was substantially large even on a ZnO bulk substrate with a millimeter-scale thickness and terminal distance. The sensitivity becomes larger with the larger resistance of the ZnO substrate. The relationship between the sensitivity and the resistance of the ZnO substrate, the EtOH concentration, and the temperature is discussed in the framework of the oxygen ionosorption model. It is indicated that the ionosorbed oxygen with a valence state of O2– is responsible for the EtOH sensing regardless of the temperature. Moreover, newly developed pulsed-jet TPD-mass spectrometry (TPD-MS) was applied to analyze the EtOH reaction with the ZnO surface, which simulates the realistic working environment on the sensing surface. In the pulsed-jet TPD-MS, the pressure periodically reaches the quasi-atmospheric pressure in the vicinity of the sample surface, while the background pressure is maintained low enough to allow the operation of TPD-MS. The oxidation product of EtOH on the ZnO surface, such as acetaldehyde, was successfully detected on the simulated sensing ZnO surface. It was found that the temperature dependence of S is consistent with that of the production rate of acetaldehyde on the ZnO surface. The novelty of the present work is the first demonstration of a gas sensing experiment on a ZnO single-crystalline bulk substrate in addition to the application of the newly developed TPD-MS combined with the pulsed-jet technique.

Results and Discussion

Figure 1a shows the resistance change of the ZnO substrate during annealing in air at 400 °C. The resistance increases by more than 4 orders of magnitude after annealing for approximately 20 h. The resistance was different among the substrates sliced from a different ZnO crystal, but the increasing behavior of the resistance by air annealing was consistent in our experiment. A similar increase in resistance of ZnO with air annealing has been repeatedly reported, which has been attributed to the elimination of point defects, such as oxygen vacancies, which act as donor defects in n-type ZnO.39−42 The initial steep increase may be partly due to the desorption of water-related species from the surface.

The elimination of the oxygen vacancy during air annealing is due to dissociative O2 adsorption, followed by uptake of oxygen interstitials into the subsurface. The isotopic oxygen gas–solid exchange study has reported substantial uptake of oxygen on the ZnO(0001) surface at several hundred degrees of temperature in this mechanism.43 The following diffusion of the oxygen interstitial leads to spontaneous annihilation between the oxygen interstitial and the oxygen vacancy in the bulk43,44 and hence enhances the resistivity of the ZnO substrate.

The uptake of oxygen into the ZnO substrate and the subsequent elimination of the oxygen vacancy were supported by the color change of the ZnO substrate with air annealing. Our as-received ZnO substrate exhibited a yellowish tinge, which has been attributed to oxygen vacancies.45 On the other hand, the substrate became more transparent with longer air annealing at 400 °C. Thus, it is indicated that the concentration of the oxygen vacancy decreases with the air annealing in the whole ZnO substrate with a thickness of 0.5 mm. A similar color change of a ZnO bulk single-crystalline substrate during air annealing has been reported, and it has also been attributed to the elimination of the oxygen vacancy.43

We observed that the reduced electrical conductivity of the ZnO substrate after air annealing can be recovered by annealing in high vacuum, which is a reductive atmosphere (Figure 1b). The recovery can be enhanced with a higher annealing temperature, as shown in the relative change of resistance normalized by the initial resistance, which was 738, 641, 596, and 482 Ω for 100, 200, 300, and 400 °C, respectively.

The effect of annealing ZnO in vacuum on the electrical conductivity has been investigated by a number of studies focused on powders and thin films.41,42,46,47 In these studies, increases in conductivity after vacuum annealing have been consistently reported and have been attributed to the elimination of negatively charged oxygen from the surface, grain boundaries, and pores. This is also thought to be the case in the present study. Because the sample is a single-crystalline substrate with an atomically flat surface in the present study, the contribution of the surface to the change in resistance is dominant. Thus, the temperature dependence of the resistance change in Figure 1b is due to the accelerated desorption rate of oxygen from the surface at high temperature.

The relationship between the adsorption/desorption of oxygen and the resistance was further investigated by pulsed air or N2 gas irradiation on the ZnO sample placed in high vacuum (Figure 1c) and elemental composition analysis of the topmost surface by He+ low-energy ion scattering (LEIS) (Figure 1d). Pulsed air or N2 gas irradiation was performed using the pulsed gas injection system of TPD-MS (see the Experimental Section and the inset of Figure 5). Thus, the ZnO sample surface was irradiated by periodic pulsed air or N2 jets, where the periodicity was 5 s. The exposure amount of the gas onto the surface was 250 Pa s for the single pulse.26 In Figure 1c, the resistance increases with air irradiation, while the resistance variation is absent for N2 irradiation. This clearly demonstrates that the resistance change of the ZnO substrate is induced by the adsorption of oxygen on the surface. The surface-sensitive He+ LEIS, which detects only a few atomic layers of an outermost surface, confirms the decrease of the surface oxygen concentration after vacuum annealing, as shown in Figure 1d. It is known that the effect of the preferential sputtering on the elemental composition is negligibly small on the Zn-terminated ZnO(0001) surface for 2 keV Ar+ ion sputtering,48 which was used for surface cleaning in the present study. Thus, the elemental composition ratio of O/Zn is unity after the sputtering, and the decrease of the He+ LEIS relative intensity ratio of O/Zn manifests the reduced oxygen concentration at the outermost surface after vacuum annealing.

Figure 5.

Figure 5

Open in a new tab

Pulsed EtOH (50 ppmv) jet TPD-MS spectrum for acetaldehyde measured at m/z = 43. The inset illustrates the schematic of the measurement configuration.

The series of our experiment shown in Figure 1 indicates that the adsorption and desorption of the ionosorbed oxygen are reversible on the surface, while the adsorbed oxygen is partly incorporated into the underlying subsurface, resulting into elimination of the donor defect. Thus, the resistance of the ZnO substrate depends on the annealing history. This is trivial but crucial for reproducing the gas sensing characteristics of a ZnO bulk substrate, as discussed below.

Figure 2a shows a typical EtOH sensing response curve of the ZnO bulk single-crystalline substrate. The temperature and EtOH concentration were 350 °C and 50 ppmv, respectively. The response and recovery times are both substantially slow compared with those of the powder and thin-film samples. For this reason, the EtOH sensing response is not saturated in the limited time of our experiment, whereas the saturated resistance is needed for evaluating the sensing sensitivity S. Thus, Rairsaturate and REtOH are evaluated from the exponential fitting, as shown in Figure 2a. Figure 2b displays S together with Rairsaturate and REtOH for four samples with different resistances. As discussed above, the resistance of the ZnO substrate varied from the kiloohm-to-gigaohm range depending on the annealing history in our experiment. The relationship between S and the resistance (Rairsaturate) is plotted in Figure 2c. It is clearly observed that S increases with Rair. It is worthwhile to note that the ZnO sample exhibits a large S of 21 with an Rairsaturate of about 6 GΩ. It is a surprisingly large sensitivity by considering that the ZnO single-crystalline thin film with a thickness of several tens of nanometers has an S of 10–20 (Figure S1 and refs (22) and (37)). The relationship between S and Rair in Figure 2c indicates that S may further increase with the increase of Rairsaturate. As discussed in the Introduction section, this large sensitivity is not expected from the space charge layer model assuming the reported space charge layer thickness for a bulk single crystal.36 Alternatively, the large sensitivity observed on the ZnO thick bulk substrate is explained by the resistance dependence of S in terms of the oxygen ionosorption, which is based on the theory by Hongsith et al. and is described below.4

Figure 2.

Figure 2

Open in a new tab

(a) EtOH (50 ppmv) sensing response curve at 350 °C and the exponential fitting for evaluating Rairsaturate and REtOH, (b) change in resistance after EtOH sensing at 350 °C for four samples with different annealing histories, and (c) sensing sensitivity S as a function of the saturated resistance in air Rairsaturate and the fitting curve that assumes that the resistance change is due to oxygen ionosorption.

According to eqs 1–4, for the reaction of EtOH with the oxygen-ionosorbed ZnO surface, the rate equation of electron density n is expressed as

graphic file with name ao0c02750_m007.jpg 7

and

graphic file with name ao0c02750_m008.jpg 8

Equations 7 and 8 are generalized as

graphic file with name ao0c02750_m009.jpg 9

where A is the coefficient, EA is the activation energy, kB is the Boltzmann constant, and T is the absolute temperature. By considering that electrical conductivity σ is proportional to n, integrating eq 9 gives the following relationship

graphic file with name ao0c02750_m010.jpg 10

where σ0 is the conductivity in air. Thus, S at equilibrium is written as

graphic file with name ao0c02750_m011.jpg 11

where c and c′ are coefficients. Equation 11 indicates that S is linearly dependent on Rairsaturate. This relationship is consistent with our experiment, as shown in Figure 2c. It is also easily derived that S at equilibrium is related to the EtOH concentration as

graphic file with name ao0c02750_m012.jpg 12

Thus, b, which reflects the charge of ionosorbed oxygen, is obtained from the slope of the log(S – 1) versus log[EtOH] plot, where b values of 1 and 0.5 correspond to charge numbers of −1 and −2, respectively.

On the basis of the discussion above on the charge parameter b, the EtOH sensing response was measured by systematically varying the EtOH concentration as well as the temperature to evaluate the valence state of the ionosorbed oxygen. More specifically, the EtOH concentration was varied in the order of 50, 25, 10, 5, 2, 1, and 0.5 ppmv, and the temperature was changed in the order of 450, 400, 350, 300, 260, 250, 220, and 180 °C. The representative response curve measured at 400 °C with changing EtOH concentration is displayed in Figure 3a as an example, and the relationship between S – 1 and the EtOH concentration obtained from these data is plotted in Figure 3b. The linear fitting for S – 1 versus the EtOH concentration demonstrated in Figure 3b gives b and thus the valence number of the ionosorbed oxygen. Figure 3c summarizes b as a function of the temperature, which was obtained from the same procedure.

Figure 3.

Figure 3

Open in a new tab

(a) EtOH sensing response at 400 °C for various concentrations (50–0.5 ppmv), (b) sensing sensitivity S – 1 as a function of the EtOH concentration derived from the response curve in (a) and the linear fitting, and (c) summary of the charge parameter b as a function of the temperature.

In Figure 3c, it is observed that the ionosorbed oxygen with a valence state of −2 is responsible for EtOH gas sensing regardless of the temperature. Thus, the oxygen species involved in the EtOH sensing is either O2– or O22–. The latter possibility is less likely by taking into account the thermally induced dissociation of oxygen molecules on a ZnO surface above 200 °C, which is widely accepted.1,2 This interpretation is supported by our pulsed-jet TPD-MS measurement, which will be discussed later.

It is observed that b drops from 0.5 below 220 °C in Figure 3c. This indicates that the EtOH sensing mechanism deviates from the reaction mechanism discussed above. The most reasonable explanation is that the concentration of the ionosorbed oxygen is dependent on that of the adsorbed EtOH at low temperature. It is assumed that EtOH adsorption is independent of oxygen ionosorption for deriving eqs 7–9. This assumption may not be valid at low temperature because low temperatures prolong the residence time of EtOH molecules on the surface, which finally leave the surface via thermally induced desorption, decomposition, or reaction. Thus, the concentration of ionosorbed oxygen may decrease because of EtOH adsorption at low temperature. Actually, such competitive adsorption has been proposed to be involved in the EtOH sensing mechanism from an ab initio study.18 The competitive adsorption between oxygen and EtOH reduces the slope of the log(S – 1) versus log[EtOH] plot because S becomes insensitive to the change in [EtOH]; thus, b decreases from 0.5.

Figure 4 shows the EtOH sensing response (a) and the sensitivity (b) as a function of temperature for three samples that have different annealing histories and hence different resistances from each other. For all three samples, the resistance increases with temperature (Figure 4a). This is due to the semiconducting nature of ZnO. It is also observed that the sensitivity S increases monotonically with temperature (Figure 4b). A similar temperature dependence of the sensitivity has been reported for EtOH sensing by ZnO with powder and film samples in a number of publications.5,11,13,14,16,20 The temperature-dependent behavior of the sensitivity in these studies has been attributed to many factors related to not only the EtOH molecule—O/ZnO surface interaction but also the complicated microscopic structure of the ZnO sensing material itself because the structure generally becomes sensitive to the temperature change with nanostructuring. On the other hand, it is known that no structure reconstruction of the atomically flat ZnO surface occurs in the temperature range of the present experiment (below 500 °C), and therefore, the temperature-dependent EtOH sensing is attributed only to the EtOH molecule—O/ZnO surface interaction in the present study.

Figure 4.

Figure 4

Open in a new tab

(a) Change of the resistance during EtOH (50 ppmv) sensing for three samples having different annealing histories and (b) sensing sensitivity S as a function of the temperature derived from (a).

To further analyze the origin of the temperature dependence of EtOH sensing, we applied newly developed pulsed-jet TPD-MS to evaluate the oxidation rate of the EtOH molecule on the ZnO surface. In pulsed-jet TPD-MS measurement, the ZnO surface was irradiated by a periodic pulsed EtOH (50 ppmv)–air jet to simulate a realistic sensing surface while keeping the background pressure low enough to operate TPD-MS, which requires high vacuum. The pulse width and periodicity were 10 ms and 5 s, respectively. In this irradiation condition, the EtOH–air gas pressure periodically reached above 2 kPa in the vicinity of the ZnO surface.38 The detailed relationship of EtOH sensing by ZnO between the pulsed-jet irradiation and the static atmospheric pressure has been discussed in our previous report.38 Briefly, the ZnO sample exhibits a similar EtOH sensing response with the above-mentioned irradiation condition to that under static atmospheric pressure. Therefore, pulsed-jet TPD-MS analyzes essentially the same surface as that in a realistic working environment, which is typically about 300 °C and atmospheric pressure, from the view point of EtOH sensing.

Figure 5 shows pulsed-jet TPD-MS spectra of acetaldehyde (m/z = 43), which is the reaction product of EtOH oxidation on the ZnO surface as shown in eqs 3 and 4. The quadrupole mass spectrometry (QMS) signal during the pulse-off time is plotted in Figure 5 to lower the effect of the background pressure. It is observed that the QMS intensity of acetaldehyde increases with the temperature above 150 °C, similar to the temperature dependence of the sensitivity in Figure 4b. This indicates that the temperature-dependent behavior of the sensitivity is caused by the oxidation rate of the EtOH molecule on the ZnO surface. It is noted that the increase of the QMS intensity with the temperature is rather monotonic, suggesting a consistent reaction scheme regardless of the temperature. Therefore, by considering the result of the ionosorption charge analysis shown in Figure 3, the ionosorbed oxygen with a valence state of O2– should continuously determine the EtOH sensing response in the whole temperature range below 500 °C.

Yang et al. recently discussed the oxidation mechanism of EtOH on an oxygen-adsorbed ZnO surface.20 They pointed out that the oxidation of EtOH to acetaldehyde is possible only on the ZnO surface with thermally dissociated atomic oxygen, where the thermal dissociation takes place above 170 °C. Thus, the EtOH reaction rate is dependent on the concentration of ionosorbed oxygen. In the present study, it is also likely that the behavior of surface oxygen (adsorption, desorption, decomposition, interdiffusion, ionization, and deionization) governs the EtOH sensing properties of ZnO.

Conclusions

Although metal oxide semiconductor gas sensors are widely considered to be promising in future devices for selectively sensing a target gas at an ultratrace concentration, the central sensing mechanism is still under debate. This is also the case for ZnO which is one of the most intensively studied sensor materials. The main open question is related to the electrical conduction mechanism as well as the surface chemical reaction pathway in sensing. In the present study, these points were analyzed by using a bulk single-crystalline substrate as the sensor material for sensing EtOH gas, which is one of the most intensively studied target gases in gas sensor research. The well-defined stoichiometry and atomic arrangement of the ZnO single-crystalline substrate enabled a detailed analysis of the EtOH sensing mechanism. The ZnO substrate tended to exhibit high electrical resistance with extended annealing in air. This was due to the adsorption/desorption of surface oxygen species. The EtOH sensing sensitivity was surprisingly larger with higher resistance of the ZnO substrate. This is understood in terms of the relationship between the sensitivity and the electron carrier concentration governed by oxygen ionosorption. The valence state of ionosorbed oxygen participating in the sensing mechanism was identified to be O2– regardless of the temperature. The newly developed pulsed-jet TPD-MS, which analyzes the target gas reaction on a surface by simulating a realistic sensing surface, detected a monotonic increase of the oxidation rate of EtOH on the surface. This explains the temperature-dependent behavior of EtOH sensing. The present study demonstrated that a ZnO bulk single crystal is adequate for studying the detailed gas sensing mechanism. The well-defined stoichiometry and atomic arrangement of the ZnO surface revealed that the sensing sensitivity becomes larger with larger resistivity of a sensing material. This may provide an insight into the material design for new semiconductor gas sensors.

Experimental Section

Materials

An atomically flat mirror-polished Zn-terminated ZnO(0001) substrate with a size of 10 × 10 × 0.5 mm3 and a surface roughness of <3.5 Å (K & R Creation, Japan) was degreased by successive ultrasonic baths of acetone and ethanol, and it was then rinsed with pure water. The ZnO crystal was grown by the hydrothermal method. Synthesized air (G1 grade) and diluted EtOH (50 ppmv) in air were purchased from Suzuki Shokan, Japan.

Gas Sensing Measurements

A custom gas sensing measurement apparatus was used to evaluate the gas sensing properties of the ZnO single-crystalline bulk substrate. The details of the apparatus have been described elsewhere.37 Briefly, the apparatus has essentially the same design as that for evaluating gas sensing properties by the conventional two-terminal method.

The surfaces of the ZnO substrate were covered with a 300 nm thick SiO2 insulating film by sputtering deposition with the exception of the (0001) plane (Zn plane). This was made to limit the contact of the gas to the (0001) plane. Two 100 nm thick Au electrodes were deposited on the (0001) surface by magnetron sputtering to evaluate the electrical resistance by measuring the voltage drop Vr at a reference resistance Rr using an electrometer (Keithley DAQ6510). The Au electrode has a square shape (3.5 × 10 mm2) with a distance between the two Au electrodes of 3 mm. The sample and the reference resistance were connected in series to a dc voltage supply of 7 V. Thus, the resistance of the sample is expressed as Rr × (7 – Vr)/Vr. The ZnO substrate was fixed on an alumina sample holder by alumina sample fixing plates. A Pt wire (a diameter of 0.5 mm) was used for the electrical connection between the ZnO sample and the electrometer, which was inserted between the Au electrode and the sample holder. Either air or diluted EtOH (50 ppmv) in air was introduced into the gas sensing measurement apparatus with a flow rate of 200 sccm.

He+ LEIS Spectroscopy Measurements

Experiments were performed in an ultrahigh vacuum (UHV) chamber (a base pressure of 7 × 10–9 Pa) equipped with a 2 keV Ar+ ion gun for sputtering, a 180° rotatable electrostatic hemispherical sector analyzer (Omicron SHA50), and a beamline of LEIS using a He+ ion as the projectile. In the He+ LEIS measurement, the scattering angle was 150°. The incident and exit angles were 0 and 30°, respectively, and these angles were measured from the surface normal direction. The incident energy was 1.7 keV. The scattering plane was parallel to (112̅0). The temperature of the ZnO sample was controlled by the electron bombardment above 500 °C, while it was controlled by radiation heating from a tungsten filament placed behind the sample below 500 °C. The ZnO sample surface was cleaned by Ar+ ion beam sputtering for approximately 20 min before the He+ LEIS measurement. The ion beam current density of He+ and Ar+ was about 0.1 and 10 nA/mm2, respectively.

Pulsed-Jet TPD-MS Measurements

Experiments were performed in a high-vacuum chamber (a base pressure of 5 × 10–7 Pa) equipped with a quadrupole mass spectrometer (AMETEK Dycor) and a UHV-compatible solenoid pulse valve (Parker 009-0582-900) for irradiating pulsed EtOH (50 ppmv)–air gas onto the ZnO substrate. The details of the pulsed-jet gas irradiation have been described elsewhere.38 Briefly, the ZnO sample surface was irradiated by periodic pulsed EtOH–air gas to simulate a realistic EtOH sensing surface in a vacuum environment. The pressure in the vicinity of the ZnO surface periodically reached the quasi-atmospheric pressure (above 2 kPa), while the background pressure was maintained low enough for operating TPD-MS. The pulse width and periodicity were 10 ms and 5 s, respectively. The distance between the ZnO substrate surface and the nozzle of the pulse valve was approximately 1 mm (inset of Figure 5). The ZnO sample temperature was controlled by a UHV-compatible infrared heating system (Thermo Riko GVJ298). The portion of the high-vacuum chamber for placing the ZnO sample and the quadrupole mass spectrometer was separated by an aperture with a diameter of 6 mm. The aperture, quadrupole mass spectrometer, and differential pumping system were integrated together and were retractable, and the setup was positioned very close to the ZnO sample during the TPD-MS measurement. The ramp rate was 20 °C/min.

Acknowledgments

This work was partly supported by the JSPS KAKENHI grant no. 19K12633 and the Innovative Science and Technology Initiative for Security, ATLA, Japan, grant number JPJ004596.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c02750.

  • Sensing sensitivity to EtOH contained in the air by a Zn–ZnO(0001) single-crystalline thin film as a function of the film thickness (PDF)

The authors declare no competing financial interest.

Supplementary Material

ao0c02750_si_001.pdf (401.3KB, pdf)

References

  1. Dey A. Semiconductor metal oxide gas sensors: A review. Mater. Sci. Eng., B 2018, 229, 206–217. 10.1016/j.mseb.2017.12.036. [DOI] [Google Scholar]
  2. Ji H.; Zeng W.; Li Y. Gas sensing mechanisms of metal oxide semiconductors: a focus review. Nanoscale 2019, 11, 22664–22684. 10.1039/c9nr07699a. [DOI] [PubMed] [Google Scholar]
  3. Barsan N.; Weimar U. Conduction model of metal oxide gas sensors. J. Electroceram. 2001, 7, 143–167. 10.1023/a:1014405811371. [DOI] [Google Scholar]
  4. Hongsith N.; Wongrat E.; Kerdcharoen T.; Choopun S. Sensor response formula for sensor based on ZnO nanostructures. Sens. Actuators, B 2010, 144, 67–72. 10.1016/j.snb.2009.10.037. [DOI] [Google Scholar]
  5. Xu J.; Han J.; Zhang Y.; Sun Y. a.; Xie B. Studies on alcohol sensing mechanism of ZnO based gas sensors. Sens. Actuators, B 2008, 132, 334–339. 10.1016/j.snb.2008.01.062. [DOI] [Google Scholar]
  6. Motin M. A.; Sadiq M.; Kim C. M. Reaction of CH3CH2OH and H on a ZnO Single Crystal Surface. Bull. Korean Chem. Soc. 2019, 40, 1226–1228. 10.1002/bkcs.11890. [DOI] [Google Scholar]
  7. Martono E.; Hyman M. P.; Vohs J. M. Reaction pathways for ethanol on model Co/ZnO(0001) catalysts. Phys. Chem. Chem. Phys. 2011, 13, 9880–9886. 10.1039/c1cp20132h. [DOI] [PubMed] [Google Scholar]
  8. Kwak G.; Yong K. Adsorption and Reaction of Ethanol on ZnO Nanowires. J. Phys. Chem. C 2008, 112, 3036–3041. 10.1021/jp7103819. [DOI] [Google Scholar]
  9. Vohs J. M. Site Requirements for the Adsorption and Reaction of Oxygenates on Metal Oxide Surfaces. Chem. Rev. 2013, 113, 4136–4163. 10.1021/cr300328u. [DOI] [PubMed] [Google Scholar]
  10. Zhou X.; Wang A.; Wang Y.; Bian L.; Yang Z.; Bian Y.; Gong Y.; Wu X.; Han N.; Chen Y. Crystal-Defect-Dependent Gas-Sensing Mechanism of the Single ZnO Nanowire Sensors. ACS Sens. 2018, 3, 2385–2393. 10.1021/acssensors.8b00792. [DOI] [PubMed] [Google Scholar]
  11. Wan Q.; Li Q. H.; Chen Y. J.; Wang T. H.; He X. L.; Li J. P.; Lin C. L. Fabrication and ethanol sensing characteristics of ZnO nanowire gas sensors. Appl. Phys. Lett. 2004, 84, 3654–3656. 10.1063/1.1738932. [DOI] [Google Scholar]
  12. Han X.-G.; He H.-Z.; Kuang Q.; Zhou X.; Zhang X.-H.; Xu T.; Xie Z.-X.; Zheng L.-S. Controlling Morphologies and Tuning the Related Properties of Nano/Microstructured ZnO Crystallites. J. Phys. Chem. C 2009, 113, 584–589. 10.1021/jp808233e. [DOI] [Google Scholar]
  13. Al-Hardan N. H.; Abdullah M. J.; Abdul Aziz A.; Ahmad H.; Low L. Y. ZnO thin films for VOC sensing applications. Vacuum 2010, 85, 101–106. 10.1016/j.vacuum.2010.04.009. [DOI] [Google Scholar]
  14. Prajapati C. S.; Sahay P. P. Alcohol-sensing characteristics of spray deposited ZnO nano-particle thin films. Sens. Actuators, B 2011, 160, 1043–1049. 10.1016/j.snb.2011.09.023. [DOI] [Google Scholar]
  15. Tian S.; Yang F.; Zeng D.; Xie C. Solution-Processed Gas Sensors Based on ZnO Nanorods Array with an Exposed (0001) Facet for Enhanced Gas-Sensing Properties. J. Phys. Chem. C 2012, 116, 10586–10591. 10.1021/jp2123778. [DOI] [Google Scholar]
  16. Wang L.; Kang Y.; Liu X.; Zhang S.; Huang W.; Wang S. ZnO nanorod gas sensor for ethanol detection. Sens. Actuators, B 2012, 162, 237–243. 10.1016/j.snb.2011.12.073. [DOI] [Google Scholar]
  17. Xiao Y.; Lu L.; Zhang A.; Zhang Y.; Sun L.; Huo L.; Li F. Highly Enhanced Acetone Sensing Performances of Porous and Single Crystalline ZnO Nanosheets: High Percentage of Exposed (100) Facets Working Together with Surface Modification with Pd Nanoparticles. ACS Appl. Mater. Interfaces 2012, 4, 3797–3804. 10.1021/am3010303. [DOI] [PubMed] [Google Scholar]
  18. Korir K. K.; Catellani A.; Cicero G. Ethanol Gas Sensing Mechanism in ZnO Nanowires: An ab Initio Study. J. Phys. Chem. C 2014, 118, 24533–24537. 10.1021/jp507478s. [DOI] [Google Scholar]
  19. Wei S.; Wang S.; Zhang Y.; Zhou M. Different morphologies of ZnO and their ethanol sensing property. Sens. Actuators, B 2014, 192, 480–487. 10.1016/j.snb.2013.11.034. [DOI] [Google Scholar]
  20. Yang T.; Liu Y.; Jin W.; Han Y.; Yang S.; Chen W. Investigation on the Transformation of Absorbed Oxygen at ZnO {10-10} Surface Based on a Novel Thermal Pulse Method and Density Functional Theory Simulation. ACS Sens. 2017, 2, 1051–1059. 10.1021/acssensors.7b00363. [DOI] [PubMed] [Google Scholar]
  21. Saito N.; Watanabe K.; Haneda H.; Sakaguchi I.; Shimanoe K. Highly Sensitive Ethanol Gas Sensor Using Pyramid-Shaped ZnO Particles with (0001) Basal Plane. J. Phys. Chem. C 2018, 122, 7353–7360. 10.1021/acs.jpcc.8b01936. [DOI] [Google Scholar]
  22. Adachi Y.; Saito N.; Sakaguchi I.; Suzuki T. T. Polarity dependent gas sensing properties of ZnO thin films. Thin Solid Films 2019, 685, 238–244. 10.1016/j.tsf.2019.06.023. [DOI] [Google Scholar]
  23. Spencer M. J. S. Gas sensing applications of 1D-nanostructured zinc oxide: Insights from density functional calculations. Prog. Mater. Sci. 2012, 57, 437–486. 10.1016/j.pmatsci.2011.06.001. [DOI] [Google Scholar]
  24. Janotti A.; Van de Walle C. G. Fundamentals of zinc oxide as a semiconductor. Rep. Prog. Phys. 2009, 72, 126501. 10.1088/0034-4885/72/12/126501. [DOI] [Google Scholar]
  25. Sun Y.-F.; Liu S.-B.; Meng F.-L.; Liu J.-Y.; Jin Z.; Kong L.-T.; Liu J.-H. Metal oxide nanostructures and their gas sensing properties: A review. Sensors 2012, 12, 2610–2631. 10.3390/s120302610. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Tricoli A.; Righettoni M.; Teleki A. Semiconductor gas sensors: Dry synthesis and application. Angew. Chem., Int. Ed. 2010, 49, 7632–7659. 10.1002/anie.200903801. [DOI] [PubMed] [Google Scholar]
  27. Gurlo A. Interplay between O2 and SnO2: Oxygen Ionosorption and Spectroscopic Evidence for Adsorbed Oxygen. ChemPhysChem 2006, 7, 2041–2052. 10.1002/cphc.200600292. [DOI] [PubMed] [Google Scholar]
  28. Degler D.; Barz N.; Dettinger U.; Peisert H.; Chassé T.; Weimar U.; Barsan N. Extending the toolbox for gas sensor research: Operando UV/vis diffuse reflectance spectroscopy on SnO2-based gas sensors. Sens. Actuators, B 2016, 224, 256–259. 10.1016/j.snb.2015.10.040. [DOI] [Google Scholar]
  29. Sänze S.; Hess C. Ethanol Gas Sensing by Indium Oxide: An Operando Spectroscopic Raman-FTIR Study. J. Phys. Chem. C 2014, 118, 25603–25613. 10.1021/jp509068s. [DOI] [Google Scholar]
  30. Rothschild A.; Komem Y. The effect of grain size on the sensitivity of nanocrystalline metal-oxide gas sensors. J. Appl. Phys. 2004, 95, 6374–6380. 10.1063/1.1728314. [DOI] [Google Scholar]
  31. Simon C. E.; Schipani F.; Papadogianni A.; Stanoiu A.; Budde M.; Oprea A.; Weimar U.; Bierwagen O.; Barsan N. Conductance Model for Single-Crystalline/Compact Metal Oxide Gas-Sensing Layers in the Nondegenerate Limit: Example of Epitaxial SnO2 (101). ACS Sens. 2019, 4, 2420–2428. 10.1021/acssensors.9b01018. [DOI] [PubMed] [Google Scholar]
  32. Yamazoe N.; Sakai G.; Shimanoe K. Oxide semiconductor gas sensors. Catal. Surv. Asia 2003, 7, 63–75. 10.1023/A:1023436725457. [DOI] [Google Scholar]
  33. Li Y.-X.; Guo Z.; Su Y.; Jin X.-B.; Tang X.-H.; Huang J.-R.; Huang X.-J.; Li M.-Q.; Liu J.-H. Hierarchical morphology-dependent gas-sensing performances of three-dimensional SnO2 nanostructures. ACS Sens. 2017, 2, 102–110. 10.1021/acssensors.6b00597. [DOI] [PubMed] [Google Scholar]
  34. Ponzoni A.; Baratto C.; Cattabiani N.; Falasconi M.; Galstyan V.; Nunez-Carmona E.; Rigoni F.; Sberveglieri V.; Zambotti G.; Zappa D. Metal oxide gas sensors, a survey of selectivity issues addressed at the sensor lab, Brescia (Italy). Sensors 2017, 17, 714. 10.3390/s17040714. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Zhou X.; Wang A.; Wang Y.; Bian L.; Yang Z.; Bian Y.; Gong Y.; Wu X.; Han N.; Chen Y. Crystal-defect-dependent gas-sensing mechanism of the single ZnO nanowire sensors. ACS Sens. 2018, 3, 2385–2393. 10.1021/acssensors.8b00792. [DOI] [PubMed] [Google Scholar]
  36. Han C. S.; Jun J.; Kim H. The depth of depletion layer and the height of energy barrier on ZnO under hydrogen. Appl. Surf. Sci. 2001, 175–176, 567. 10.1016/s0169-4332(01)00118-0. [DOI] [Google Scholar]
  37. Adachi Y.; Watanabe K.; Saito N.; Sakaguchi I.; Suzuki T. T. Gas sensing properties of c-axis-oriented Al-incorporated ZnO films epitaxially grown on (11-20) sapphire substrates using pulsed laser deposition. J. Ceram. Soc. Jpn. 2016, 124, 668–672. 10.2109/jcersj2.15319. [DOI] [Google Scholar]
  38. Suzuki T. T.; Adachi Y.; Ohgaki T.; Sakaguchi I.; Nakamura M.; Ohashi H.; Aimi A.; Fujimoto K. Electrical resistance response of a ZnO single-crystalline substrate to trace ethanol under pulsed air jet irradiation. Vacuum 2020, 179, 109526. 10.1016/j.vacuum.2020.109526. [DOI] [Google Scholar]
  39. Li G. R.; Hu T.; Pan G. L.; Yan T. Y.; Gao X. P.; Zhu H. Y. Morphology-Function Relationship of ZnO: Polar Planes, Oxygen Vacancies, and Activity. J. Phys. Chem. C 2008, 112, 11859–11864. 10.1021/jp8038626. [DOI] [Google Scholar]
  40. Gorai P.; Seebauer E. G.; Ertekin E. Mechanism and energetics of O and O2 adsorption on polar and non-polar ZnO surfaces. J. Chem. Phys. 2016, 144, 184708. 10.1063/1.4948939. [DOI] [PubMed] [Google Scholar]
  41. Gritsenko L. V.; Abdullin K. A.; Gabdullin M. T.; Kalkozova Z. K.; Kumekov S. E.; Mukash Z. O.; Sazonov A. Y.; Terukov E. I. Effect of thermal annealing on properties of polycrystalline ZnO thin films. J. Cryst. Growth 2017, 457, 164–170. 10.1016/j.jcrysgro.2016.07.026. [DOI] [Google Scholar]
  42. Fujitsu S.; Koumoto K.; Yanagida H.; Watanabe Y.; Kawazoe H. Change in the Oxidation State of the Adsorbed Oxygen Equilibrated at 25 °C on ZnO Surface during Room Temperature Annealing after Rapid Quenching. Jpn. J. Appl. Phys. 1999, 38, 1534–1538. 10.1143/jjap.38.1534. [DOI] [Google Scholar]
  43. Gorai P.; Ertekin E.; Seebauer E. G. Surface-assisted defect engineering of point defects in ZnO. Appl. Phys. Lett. 2016, 108, 241603. 10.1063/1.4953878. [DOI] [Google Scholar]
  44. Huang G.-Y.; Wang C.-Y.; Wang J.-T. First-principles study of diffusion of oxygen vacancies and interstitials in ZnO. J. Phys.: Condens. Matter 2009, 21, 195403. 10.1088/0953-8984/21/19/195403. [DOI] [PubMed] [Google Scholar]
  45. Simpson J. C.; Cordaro J. F. Characterization of deep levels in zinc oxide. J. Appl. Phys. 1988, 63, 1781–1783. 10.1063/1.339919. [DOI] [Google Scholar]
  46. Barnett C. J.; Mourgelas V.; McGettrick J. D.; Maffeis T. G. G.; Barron A. R.; Cobley R. J. The effects of vacuum annealing on the conduction characteristics of ZnO nanorods. Mater. Lett. 2019, 243, 144–147. 10.1016/j.matlet.2019.02.005. [DOI] [Google Scholar]
  47. Lalanne M.; Soon J. M.; Barnabé A.; Presmanes L.; Pasquet I.; Tailhades P. Preparation and characterization of the defect-conductivity relationship of Ga-doped ZnO thin films deposited by nonreactive radio-frequency-magnetron sputtering. J. Mater. Res. 2010, 25, 2407–2414. 10.1557/jmr.2010.0300. [DOI] [Google Scholar]
  48. Sulyok A.; Menyhard M.; Malherbe J. B. Stability of ZnO{0001} against low energy ion bombardment. Surf. Sci. 2007, 601, 1857–1861. 10.1016/j.susc.2007.02.011. [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

ao0c02750_si_001.pdf (401.3KB, pdf)

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

RESOURCES