Crystal Defect Dependent Gas Sensing Mechanism of the Single ZnO

Nov 2, 2018 - Crystal Defect Dependent Gas Sensing Mechanism of the Single ZnO Nanowire Sensors ... into electrical signal (i.e. resistance change) is...
0 downloads 0 Views 1MB Size
Download PDF
Subscriber access provided by University of Virginia Libraries & VIVA (Virtual Library of Virginia)

Article

Crystal Defect Dependent Gas Sensing Mechanism of the Single ZnO Nanowire Sensors Xinyuan Zhou, Anqi Wang, Ying Wang, Luozhen Bian, Zai-xing Yang, Yuzhi Bian, Yan Gong, Xiaofeng Wu, Ning Han, and Yunfa Chen ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.8b00792 • Publication Date (Web): 02 Nov 2018 Downloaded from http://pubs.acs.org on November 2, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Crystal Defect Dependent Gas Sensing Mechanism of the Single ZnO Nanowire Sensors Xinyuan Zhou,†,‡ ,§ Anqi Wang,†,‡ Ying Wang,† Luozhen Bian,ǁ Zaixing Yang,*,ǁ Yuzhi Bian,†,‡ Yan Gong†,‡, Xiaofeng Wu†, Ning Han,*,†,§ and Yunfa Chen*,†,§

†

State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China ‡ University of Chinese Academy of Sciences, No. 19A Yuquan Road, Beijing 100049, China § Center for Excellence in Regional Atmospheric Environment, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China ǁ Center of Nanoelectronics and School of Microelectronics, Shandong University, Jinan 250100, China

KEYWORDS: ZnO nanowires; micro-photoluminescence; diameter; space charge layer; crystal defects; gas sensor ABSTRACT: Though the chemical origin of a metal oxide gas sensor is widely accepted to be the surface reaction of detectants with ionsorbed oxygen, how the sensing material transduces the chemical reaction into electrical signal (i.e. resistance change) is still not well recognized. Herein, the single ZnO NW is used as a model to investigate the relationship between the microstructure and sensing performance. It is found that the acetone responses arrive at the maximum at the NW diameter (D) of ~110 nm at the D range of 80 nm to 400 nm, which is even the temperature independent at the temperature region of 200 ℃-375 ℃. Then the electrical properties of the single NW field effect transistors illustrate that the electron mobility decreases but electron concentration increases with the D ranging from ~60 nm to ~150 nm, inferring that the good crystal quality of thinner ZnO NWs and the abundant crystal defects in thicker NWs. Subsequently, the surface charge layer (L) is calculated to be a constant of 43.6 ± 3.7 nm at this D range, which can not be explained by the conventional D-L model that gas sensing maximum appears when D approximates 2L. Furthermore, the crystal defects in the single ZnO NW are probed by employing the micro-photoluminescence technique. The mechanism is proposed to be the compromise of the two kinds of crystal defects in ZnO, i.e. more donors and less acceptors favor the gas sensing performance, which is again verified by the gas sensors based on the NW contacts.

D

ue to the advantages of high sensitivity, low cost and easy fabrication 1-3, metal oxide (MOX) gas sensors have attracted enormous research interests. As for the sensing mechanism, Yamazoe group built one milestone of the D-L model 4, which compared the nanoparticle size (D) and depth of surface charge layer (L): the gas sensitivity is high if D is comparable to or less than 2L. Afterwards, many researchers prepared the hierarchical or hollow MOX materials in order to decrease the D to enhance their gas sensitivity 5-7. Nevertheless, some limitations still exist in this D-L model. For example, it is difficult to obtain the precise L value by using current instruments. Besides, some researches 8-10 have reported their results contradictory to this D-L model where the gas sensitivities are independent or even reversely dependent on the D. On the other hand, some researchers change their mind to study the crystal defects of MOX materials by employing the photoluminescence (PL) and try to build the relationship between gas sensing performance and crystal defects of the materials. For example, Hong et al. 11 reported that the sensitivity to

NO2 was linearly proportional to the PL intensity of oxygen vacancy (VO) in ZnO nanowires (NWs). Han et al. 10, 12 proposed that ZnO nanoparticles which had more VO and/or zinc interstitial (Zni) showed higher response to the formaldehyde. The ZnO nanodashes fabricated by Xue et al. 13 exhibited the excellent response to ethanol due to the rich crystal defects Zn i and VO. However, these PL spectra were superposition of crystal defects coming from many individual nanostructures (i.e. NWs, nanospheres or nanodishes) and thus neglected the crystal defect variations of individual nanostructures, resulting from the morphology, size, crystalline 14-15, exposed facet 16 as well as growth orientation 17. Nowadays, the crystal defects of the individual nanostructures such as NWs, nanobelts etc. can be investigated by utilizing the micro-(μ)PL 18-19. And the single NW based gas sensors fabricated by using the micro-electromechanical systems have become one of the research focuses due to their simple operation, low cost and great miniaturization potential 20. Importantly, the single NW gas sensors eliminate the average effect of different NWs 21, in favor of choosing the appropriate

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

size to improve the sensing property. In this paper, ZnO NWs with diameters ranging from 50 nm to 425 nm have been prepared successfully by the chemical vapor deposition (CVD). Then the structural properties of the products are characterized by using X-ray diffraction (XRD), scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM). Next, their gas sensing and electrical properties are tested on the basis of the individual NW by single NW gas sensors and single NW field effect transistors (FETs), followed by testing the PL properties of single NWs using the μPL technique. The acceptor level (AL)- and donor level (DL)-related luminescences are discriminated by Gaussian deconvolution, which are found to be beneficial to the gas sensing property enhancement by more DL and less AL, promising for highly sensitive material synthesis.

EXPERIMENTAL SECTION Synthesis and Characterization of ZnO NWs. The ZnO NWs were grown by CVD method in Figure S1 in the supporting information. An Al2O3 boat containing ~0.3 g ZnO and ~0.1 g graphite powders was placed in the middle of upstream zone at the temperature of 950 ℃ and the SiO2/Si substrate covered with ~50 nm Au nanoparticles was put in the middle of downstream zone at the temperature of 750 ℃. After evacuation to ~10-2 Torr by using a rotary pump, 50 standard cubic centimeter per minute (sccm) argon and 2 sccm oxygen were introduced in the quartz tube and the inner pressure was maintained at ~3 Torr. The dual-zone horizontal tube furnace was then heated at the rate of 20 ℃/min and the growth time was set as 0.5 h. Finally, the white sample was obtained. The crystal structure of the ZnO sample was studied by XRD on a Panalytical X’Pert PRO system with Cu Kα radiation (1.5406 Å). The morphology and size of the samples were characterized under SEM (JEOL JSM-6700F, Japan) operated at 15 kV. Different diameters of ZnO NWs were investigated by using HRTEM (JEOL JEM-2100F, Japan) operated at 200 kV. The μPL spectra were obtained at room temperature on a spectrometer (inVia Reflex, Renishaw) with a 325 nm laser. Fabrication of the Single ZnO NW FETs. As shown in Figure S2 in the supporting information, ZnO NWs were dispersed by sonication in ethanol from the substrate and then drop-casted onto the p-type silicon wafer (2×2 cm2) with a 50 nm SiO2 layer. Standard photolithography was employed, followed by Ni metal deposition and lift-off process, and ZnO NW FETs were ultimately fabricated. Electrical characteristics of ZnO NW FETs were measured by using the Keithley 4200 semiconductor analyzer on a standard probe station (Lakeshore). Acetone Sensing Performance Measurement. The ultrasonic aluminum wire bonder was utilized to fix the two Ni electrodes of ZnO NW onto the Al2O3 chip (1×2 cm2). Then two Pt wires were connected to this Al2O3 chip. As depicted in Figure S3 in the supporting information, the gas sensor was put in the testing chamber, heated by a horizontal tube furnace. In all cases, the acetone concentration was controlled by introducing the synthetic air (20.9 vol.% O2, 79.1 vol.% N2) and standard acetone gas (49.8 ppm for acetone in synthetic air) into the mixing chamber using the digital mass flow controllers 22. Prior to the measurements, initial stabilization was achieved by flowing the synthetic air until the electrical current became stable at a constant voltage of 5.0 V. The electrical current as a function of time was recorded by using the Keithley 2602B during the measurements. The sensing per-

Figure 1. The morphology, crystalline structure, crystal orientation and diameter distribution of the ZnO NWs prepared by using the 50 nm Au nanoparticles as the catalyst. SEM image (a), XRD pattern (b) and TEM image (c) of ZnO NWs. The inset in (c) is the corresponding HRTEM image. (d) Diameter distribution histogram of the grown NWs. oriented ZnO NWs have the length of ~8 μm, with the diameter varying from 50 nm to 425 nm.

formance was measured at the temperature from 200 ℃ to 425 ℃. The response is defined as Response = Ra/Rg, where Ra and Rg are the sensor resistances in air and in acetone gas, respectively.

RESULTS AND DISCUSSION Morphology and Structure of ZnO NWs. The morphology of the ZnO sample is characterized by SEM (Figure 1a). Figure 1a shows that ZnO NWs with the length of ~8 μm grow almost vertically on the substrate. The crystalline structure of the sample is characterized by XRD, as shown in Figure 1b, where there is a sharp diffraction peak at 34.4°, corresponding to the facets (002) of the wurtzite phase ZnO. However, the diffraction peaks of other facets are extremely weak, which indicates a preferential orientation along c-axis. The growth of the ZnO NWs can be resulted from the lowest surface free energy of the (002) plane in ZnO. It is known that the three lowest densities of the surface free energy in ZnO are 9.9, 12.3, and 20.9 eV/nm 2 for (002), (110) and (100) planes in the equilibrium state respectively 23. Therefore, the ZnO NWs usually exhibit (002) texture with the c-axis perpendicular to the substrate surface due to its low surface free energy 24-25. The microstructure of ZnO NWs is further investigated detailedly by TEM and HRTEM in Figure 1c, with more details shown in Figure S4. It is found that the d-spacing of 0.26 nm of the lattice fringes corresponds to the (002) plane of the hexagonal structure of ZnO according to the HRTEM images. At the same time, the HRTEM images in Figure S4 illustrate that different diameters of ZnO NWs grow along crystal orientation , which is consistent with the XRD results. The same crystal orientation would eliminate its influence on the gas sensing performance 17 of ZnO NWs below. In addition, based on ~100 NWs observed by TEM, the diameter of ZnO NW varies from 50 nm to 425 nm with 60-70% of NWs in the region of 100-200 nm (Figure 1d). Gas Sensing Properties of Gas Sensors Based on the Single ZnO NW. Then gas sensing performances of the single ZnO NW sensors are tested. Figure 2a-c shows the curves of

ACS Paragon Plus Environment

Page 2 of 9

Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 2. Responses to 5 ppm acetone versus operating temperature of the single ZnO NW with different diameters of (a) ~80 nm, (b) ~110 nm and (c) ~210 nm. The insets in (a)-(c) are the corresponding SEM images of the gas sensors respectively (all scale bars are 2 μm). (d) Response to 5 ppm acetone as a function of the ZnO NW diameter. Each error bar is estimated by statistics of 3-4 gas sensors.

responses to acetone at different operating temperatures from 250 ℃ to 425 ℃ (Each error bar is estimated by statistics of 34 gas sensors). Taking the single NW of ~110 nm gas sensors for example, the acetone responses initially increase with the temperature increasing, which can be due to the fact that the high temperature can activate gas molecules to overcome energy barrier and thus facilitate the surface reaction. After reaching the maximum, the responses decrease with further increasing the temperature. The reason is that the higher temperature might weaken the gas adsorption, thus hinder the surface reaction 26-27. In all, the optimal temperature is 350 ℃ and the maximum response is 42 ± 3 to 5 ppm acetone. The gas sensors prepared by other diameters of ZnO NWs share the similar feature. Figure 2d shows the gas response to 5 ppm acetone versus diameter of ZnO NW in order to obtain the optimal diameter of NW. It should be noted that the peak response is obtained for NW with ~110 nm and the response displays a dramatic slump when the diameter decreases from ~110 nm to ~80 nm at the temperature of 350 ℃ and 375 ℃. The operating temperature is further reduced to verify this counterintuitive result. As depicted in Figure S5a in the supporting information, the ~110 nm ZnO NW shows the highest response to 200 ppm acetone and the dramatic slump reappears with the diameter decreasing from ~110 nm to ~80 nm with the temperature down to 200 ℃ and 225 ℃, which is consistent with the data in Figure 2d. Electrical Properties of the Single ZnO NW FETs. In order to explore the counterintuitive gas sensing performance, the electrical characteristics of ZnO NWs should be assessed because they are intimately associated with the microstructure of nanomaterials. The electrical properties at the temperature ranging from 25 ℃ to 375 ℃ are tested by using the fabricated single NW FETs with back-gate configuration, from which the intrinsic transport property of the single NW such as the carrier mobility as well as carrier concentration can be deduced. The typical SEM image and schematic diagram are demonstrated in Figure 3a inset. Figure 3a shows that the transfer curves of the representative single ZnO NW of ~110 nm have

Figure 3. Electrical properties of the single ZnO NW FET. (a) Transfer curves of the ~110 nm ZnO NW configured in the back gated field-effect transistor at the temperature from 25 ℃ to 375 ℃. The inset (panel a) gives the corresponding SEM image and schematic diagram of the device fabricated with Ni source/drain contacts. (b) The scatter diagram of the peak field-effect electron mobility and electron concentration versus operating temperature. The mobility remains 38.3±3.0 cm2/Vs over this temperature range. (c) Peak field-effect electron mobility and electron concentration at room temperature as a function of NW diameter, with the diameter ranging from 60 nm to 150 nm. (d) Calculation of space charge layer at room temperature according to the formula L = (2ε0εZnOφ/ene)1/2 using the data in (c). The L is a constant of 43.6 ± 3.7 nm in this diameter range.

similar slope but shift negatively gradually when the temperature varies from 25 ℃ to 375 ℃. Meanwhile, these transfer curves (Figure 3a and Figure S6a) exhibit their depletion mode 28 with n-type semiconducting behavior. In addition, output curves in Figure S6b show that the contacts are ohmic between the single ZnO NW and Ni source/drain electrodes because the work function of Ni metal (5.1 eV) 29-30 is lower than that of ZnO NWs (5.3 eV) 31. Furthermore, the field-effect electron mobility can be calculated based on cylinder-on-plate model 3233 . The low-bias (i.e. VDS = 0.1 V) transconductance (gm) can be extracted from the transfer characteristics, gm = dIDS/dVGS. The electron mobility is expressed as 34-36:

μn = gml2/(VDSCox) (1) Cox = (2πε0εl)/cosh-1(1+d/r) (2) where Cox is the gate capacitance, l is the NW length, r is the NW radius, d is the thickness of silicon oxide layer (50 nm), ε0 is the absolute dielectric constant, and ε is the dielectric constant of the silicon oxide layer (3.9). The electron mobility of an individual NW will be calculated accurately once the NW channel length and radius are given. The curve of electron mobility (μn) versus VGS at 375 ℃ for a typical single ZnO NW FET (D is 110 nm and l is 6.3 μm) is shown in Figure S6c, where the maximum mobility of the NW is found to be ~36 cm2 V-1 s-1. And the electron concentration ne is estimated to be 2.2×1018 cm-3 using the equation ne = σ/(eμn) 37, where e is electronic charge and σ is the electrical conductivity (12.5 Ω-1cm-1) at the point of the maximum mobility. As shown in Figure 3b, the electron mobility maintains stable value of 38.3 ± 3.0 cm2/Vs while the electron concentration tends to increase from 5 to 221017 cm-3 with the temperature increasing from 25 ℃ to 375 ℃. Notably,

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the temperature independent electron mobility would be attributed to the fact that the increased temperature enhances the phonon scattering, thus reduces the electron mobility while the defect scattering is opposite 38-40. Figure 3c summarizes the maximum μn and ne at room temperature for all of the measured different diameters of NW FETs. On one hand, it illustrates a decrease of electron mobility as the NW grows thicker; however, it is interesting to find that there is an inflection point at the NW diameter of ~100 nm. That is, the reduction is steep from ~60 to ~100 nm, after which the change is slower, similar to those in the literatures 40-41. This is possibly because that defect scattering dominates over the phonon scattering 40. On the other hand, the electron concentration is also found to increase from ~6 to ~101017 cm-3 when the D grows thicker to 150 nm, which infers the good crystal quality of thinner ZnO NWs and the high concentration of crystal defects in thicker NWs because the electrons come from the crystal defects in the undoped ZnO materials. Based on this, the surface charge layer (L) in Figure 3d is calculated by the formula (3) 42-43:

L = (2ε0εZnOφ/ene)1/2 (3) where εZnO is the dielectric constant of ZnO (8.66) 44-45, φ is the surface barrier potential (1.53 eV as measured by ultraviolet photoelectron spectroscopy) 42. Figure 3d illustrates that the L is a constant of 43.6 ± 3.7 nm for various diameters of NWs, which is in good agreement with the literature 42. Nevertheless, the responses to acetone should be higher when the D of NW approaches 2L (i.e. D ~87 nm) according to the D-L theory 21, 46-47 , which is not consistent with our experimental results in this context. Furthermore, at working temperature of 375 ℃, the electron concentration is ~4 times and thus the L is half of that at room temperature according to equation (3). This means 2L at optimal working temperature should be ~43.6 nm, which can also not explain our result by the D-L mode which would otherwise predict that small diameter favors high response at diameter > 43.6 nm. Crystal Defects in the Single ZnO NW. Meanwhile, the crystal defects of ZnO NWs are found to play a vital role in the electron mobility and concentration as mentioned above and need the deeper investigation to reveal the gas sensing mechanism. The μPL spectroscopy is employed to probe the crystal defects because it originates from either the photo-induced electron-hole bandgap recombination or in the intrinsic defects 48. In our experiment, μPL spectra of the single ZnO NW are shown in Figure 4. Obviously, μPL spectra of ZnO NWs at room temperature consist of two emission bands in the ultraviolet (UV) and visible regions. The UV emission band is the near-band edge emission which originates from the recombination of free excitons through an exciton-exciton collision process. The visible emission band is named as deep level emission, which comes from recombination in the crystal defects. It is clear that the UV peak is stronger than the visible peak for the single ZnO NW with ~80 nm while it is reverse for the thicker ZnO NW. The relative intensity normalized at ~378 nm is plotted logarithmically in Figure 4 inset, from which the visible-toUV luminescence ratios (IVL/IUV) are found to be ~0.711, ~2.12, 9.38, 330, 12.2 with the diameter of NW increasing from ~80 nm to ~400 nm. The relative low ratio of the thin ZnO NWs infers the low concentration of crystal defects and thus high crystal quality, which is in good agreement with the measured high electron mobility. However, though the single NW of ~330 nm possesses the maximum IVL/IUV, it exhibits a very low gas response, reverse to the pre-

Figure 4. μPL spectra of the single ZnO NW with different diameters and relative intensity normalized at 378 nm (inset). The IVL/IUV are found to be ~0.711, ~2.12, 9.38, 330 and 12.2 with the NW diameter ranging from ~80 nm to ~400 nm.

vious reports 49-50. Therefore, the gas sensing properties of the single NWs with different diameters cannot be interpreted by the IVL/IUV, which actually depends on the sample type and the experimental conditions 33. It is obvious that there exist clear differences among these five visible peak positions in Figure 4. So it is necessary to further divide the μPL spectra into the donor (DL) and acceptor (AL) related subpeaks. Normally, the origin of deep level emission is intrinsic defects (VO, Oi, Zni, VZn, ZnO and OZn) 51 and extrinsic impurities (e.g. Mn) 52-53. As for pure ZnO NWs, the deep level emission mainly results from the intrinsic defects which can be divided into two parts: one is DL, such as VO and Zni, and the other is AL, including VZn, Oi, and OZn 48, 54-55. Therefore, the whole μPL spectra are Gaussian deconvoluted into eight sub-peaks as shown in Figure 5 a-d and Figure S7 in the supporting information. These subpeaks are assigned to different defects. The UV peak at 378 nm is attributed to the conduction band to valence band (CB-VB) combination 56; the peak at 387 nm is indexed to a shallow donor (might be Zni related complex defect) 57-58 ; the peak at 460 nm is attributed to VZn 59; the peak at 490 nm is assigned to VO 60-61; the peak at 520 nm is related to OZn 62 and the yellow and orange luminescence peaks (> 540 nm) are ascribed to Oi 62-65. The resultant content of these crystal defects are seen in tables in Figure 5 and Figure S7. As for ~110 nm ZnO NW, the DL content is calculated to be 10.7 by summing up the content of VO (9.58) and the content of Zni (1.14), as shown in table in Figure 5b. Then increasing sequence of DL subpeaks content in ZnO NWs (~400 nm< ~80 nm< ~330 nm< ~210 nm< ~110 nm) is in good accordance with the order of the acetone response (~400 nm< ~80 nm< ~330 nm< ~210 nm< ~110 nm), revealing that the more the DL content is, the higher the responses are 13, 16. In addition, although ~80 nm NW and ~110 nm NW have the similarly low AL percentages, the former DL percentage is half of the latter, which leads to the fact that ~80 nm NW shows a sharply decreasing response to the acetone compared to ~110 nm NW. In order to demonstrate clearly the relationship between acetone responses and the contents of DL & AL subpeaks, the tendency curves are plotted in Figure 6 and Figure S5b in the supporting information. To minimize errors resulting from the Gaussian deconvolution of μPL spectra, 3-4 ZnO NW samples

ACS Paragon Plus Environment

Page 4 of 9

Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 5. Gaussian deconvolution of μPL spectra of the single ZnO NW with different diameters of (a) ~80 nm, (b) ~110 nm, (c) ~210 nm, (d) ~400 nm. The requirements in Gaussian deconvolution are: (1) the peak position is relatively fixed and each peak has its corresponding physical meaning. (2) Strong peaks (obvious peaks in μPL spectra) first and then weak ones (residue of μPL spectra after decomposing obvious peaks). (3) Relatively high correlation coefficient (r2> 0.999).

with similar diameters are measured. Taking ~80 nm ZnO NWs for example, the NWs with diameter of 62, 86, 78 and 81 nm are measured by the μPL, which are further divided into the donor and acceptor related subpeaks with relatively high correlation coefficient (r2>0.999). Then the results of Gaussian deconvolution are shown by using the error lines in Figure 6 and Figure S5b in the supporting information. In addition, the PL spectra originating from the defects in ZnO at the temperature of 200 ℃ do not shift clearly to the lower energy and are nearly parallel with those at room temperature 66. Consequently, it seems to be reasonable to interpret the gas response at 200 ℃ and 375 ℃ by using μPL spectra at room temperature. From Figure 6 and Figure S5b, it is found that where the DL percentage is high and the AL percentage is low, the response to acetone is high for thicker ZnO NWs (D>110 nm), implying that DL can positively improve the gas sensitivity whereas AL plays a negative role 10. As for thinner ZnO NWs (D≤110 nm), AL content is similarly low, but the large increase of the DL content contributes to the great enhancement of gas response. In all, ZnO NW of ~110 nm has the most DL and the least AL. More donors can release more electrons for the redox reactions, leading to the best response. Gas Sensing Properties of the ZnO NW Contact Based Gas Sensors. In traditional gas sensing films, the sensor resistance is made of not only the semiconductor materials but also the material contacts. To further compare the effect of crystal defects and material contact, ZnO NW contact based devices have been fabricated. The SEM image and schematic of a typical ZnO NW contact based device are shown in Figure 7a. The device channel length and diameter are 8.6 μm and 110 nm respectively with a junction produced between two NWs. The typical output curves are plotted in Figure 7b, where this device is an ohmic-like contact. Then these devices are subject to the gas sensing performance measurement as depicted in Figure 7ce (Each error bar is estimated by statistics of 4-6 gas sensors). For instance, the best operating temperature of NW (~110 nm) contact based gas sensors is 350 ℃,

Figure 6. Relationship between acetone response of the ZnO NWs at 375 ℃ (△) and the DL (○) & AL (□) subpeaks contents. Each error bar in the black and red curves is estimated by statistics of 34 NW samples. More donors and less acceptors favor the gas sensing performance.

with the response of 56 ± 3 to 5 ppm acetone in Figure 7d. It should be noted that the NW of ~110 nm has the maximum response to 5 ppm acetone compared to those of ~80 nm and ~160 nm according to Figure 7f, which reveals that the junctions between the two NWs have not changed the tendency of the responses versus the NW diameters. Moreover, the two types of gas sensors to 5 ppm acetone is tested once a day for 15 days at 350 ℃ as shown in Figure S8, displaying their good stability. For comparison, the sensor responses of the ZnO NW thin film with the same geometry are investigated in Figure S9. The ZnO NW thin film sensors have much lower responses than those of the single ZnO NWs and NW contacts 67, showing again the significance of the individual NW investigation to eliminate the averaging effect of the NW films. In all, the responses to acetone are more dependent on the crystal defects (DL and AL) than the junction of NWs. Sensing Mechanism of Gas Sensors. From the above discussions, the superior gas response of ZnO NW of ~110 nm can be ascribed to the more DL and the less AL, and vice versa. Thus, the defect species are introduced into our proposed gas sensing mechanism. When the ZnO NW is exposed to air, oxygen molecules are adsorbed and become negatively charged via capturing electrons from the conduction band of ZnO, such as O2-, O-, and O2-. It has been reported that 300 ℃-400 ℃ oxygen adsorbs on ZnO predominantly as O- and about 200 ℃ as the mixture of O2- and O- 68-71, which can be described as equations (4)-(5) 72:

O2 (ads) + 2e- → 2O- (ads) (4) O2 (ads) + e → O2 (ads) (5) Although oxygen species are temperature dependent and Ois deemed to favor the gas sensing performance (measured by using the temperature changeable XPS) 73, ~110 nm ZnO NW still shows maximum response to acetone and is temperature independent at the region of 200 ℃-375 ℃, as seen in Figure 2d and Figure S5a in the supporting information. For convenience, the case at high temperature of 375 ℃ is further elaborated because O- is the dominant species. According to equation (4), the more electron donors and the less electron acceptors are, the more oxygen species O- because the donors in

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 9

which is responsible for the sharp decrease of sensing performance. Briefly, three ideal models of sensing mechanism are schematized in Figure S10. These three models are the thinner single crystalline ZnO NW, the medium ZnO NW with lots of donorrelated defects and the thicker ZnO NW with equal donor- and acceptor-related defects. It is found that the second model has many free electrons produced by the donor level whereas other two models have less free electrons. That is because the intrinsic excitation in the single crystalline ZnO merely generates a few electrons or many electrons in the thicker ZnO NWs are consumed by the acceptor level. As is mentioned above, the more free electrons likely to adsorb more oxygen molecules, which contribute to the redox reactions with the target gas to improve sensing performance. In all, the gas mechanism is thought to be that the more the DL is and the less AL is, the higher gas response is.

CONCLUSION

Figure 7. (a) (Top) SEM image and (bottom) schematic of a backgated NW (~110 nm) contact based device. (b) Output curves of this device. Responses to 5 ppm acetone versus operating temperature of gas sensors based on ZnO NW contact with different diameters of (c) ~80 nm, (d) ~110 nm and (e) ~160 nm (Insets: the corresponding SEM image of the device in (c)-(e). (all scale bars are 2 μm). (f) Responses to 5 ppm acetone as a function of the diameter of the contacting ZnO NWs. Each error bar is estimated by statistics of 4-6 gas sensors.

crease the free electrons because the donors increase the free electrons while the acceptors consume free electrons in the ZnO NW, as expressed in equations (6)-(9) 74, where VO and Oi are the examples of the donors and acceptors respectively.

VOx → VO· + e- (electrons donating) (6) VO· → VO·· + e- (electrons donating) (7) Oi + e → Oi′ (electrons consuming) (8) Oi′ + e → Oi″ (electrons consuming) (9) Therefore, exposed to air, ZnO NW of ~110 nm with maximum DL and minimum AL, is able to adsorb the highest content of oxygen molecules on its surface. Of course, the number of the free electrons is a result of comprehensive effect of the donors and acceptors. Anyhow, free electrons are enough to ensure the pure ZnO is an n-type semiconductor. In contrast, once the acetone gas is introduced, it can react with the adsorbed oxygen ions. This process returns the electrons back to the conduction band of the ZnO NW, further decreasing the electrical resistance of the NW, as can be expressed in equation (10) 53: CH3COCH3(ads) + 8O- → 3CO2(g) + 3H2O + 8e- (10) Thus, it can be observed that due to its higher content of the adsorbed O- species, ZnO NW of ~110 nm can tend to react with the more acetone molecules and hence exhibits the higher gas response. Notably, despite ~80 nm NW has the almost same contents of AL as ~110 nm NW, the former has much less DL than the latter, leading to much less adsorbed oxygen molecules,

In summary, the roles of donor and acceptor crystal defects in sensing performance of ZnO NWs with various diameters have been studied by utilizing the μPL technique. The results show that ~110 nm ZnO NW has the maximum donors and minimum acceptors and thus displays the best gas response. More importantly, the tendency of the gas response versus NW diameter is independent of the temperature at the region of 200 ℃-375 ℃. At the same time, this tendency is little affected by the junction in the NW contact based device. Therefore, the donor and acceptor defects play a predominant role in the gas sensing performance. Finally, a gas sensing mechanism model is proposed: The higher gas response of the single ZnO NW results from more electron donors and less electron acceptors releasing more electrons for adsorbing more oxygen species, which tend to react with more target gas, causing the higher response.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Figure S1: Schematic representation of the dual-zone horizontal tube furnace for the ZnO NW growth; Figure S2: The flow chart for the fabrication of the single ZnO NW fieldeffect transistor; Figure S3: The schematic representation of the experimental system for gas sensing; Figure S4: TEM images and the corresponding HRTEM images of ZnO NWs with various diameters of (a) ~80 nm, (b) ~110 nm, (c) ~210 nm and (d) ~400 nm; Figure S5: (a) Response to 200 ppm acetone as a function of the NW diameter. (b) Relationship between acetone response of the ZnO NWs at 200 ℃ (△) and the DL (○) & AL (□) subpeaks contents; Figure S6: Electrical properties of the single ZnO NW FET at 375 ℃; Figure S7: Gaussian deconvolution of μPL spectra of the single ZnO NW of ~ 330 nm. Figure S8: The stability of the single (red) and two (blue) ZnO NWs with ~110 nm to 5 ppm acetone at 350 ℃; Figure S9: Responses to 5 ppm acetone versus operating temperature of the ZnO NW thin film; Figure S10: Schematic view of gas sensing process of ZnO NW with three different diameters considering the crystal defects. (PDF).

ACS Paragon Plus Environment

Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. *E-mail: [email protected]. *E-mail: [email protected].

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This research was financially supported by the National Key R&D Program of China (2016YFC0207100), the National Natural Science Foundation of China (61504151, 51602314 and 51272253), Shandong Provincial Natural Science Foundation of China (ZR2017MF037), “Qilu young scholar” program of Shandong University, Guangdong Innovative and Entrepreneurial Research Team Program (No. 2014ZT05C146), the CAS-CSIRO project of the Bureau of International Co-operation of Chinese Academy of Sciences (122111KYSB20150064), and the State Key Laboratory of Multiphase Complex Systems (MPCS-2014-C-01).

REFERENCES (1) Yamazoe, N., Toward innovations of gas sensor technology. Sens. Actuators B: Chem. 2005, 108 (1-2), 2-14. (2) Guntner, A. T.; Koren, V.; Chikkadi, K.; Righettoni, M.; Pratsinis, S. E., E-nose sensing of low-ppb formaldehyde in gas mixtures at high relative humidity for breath screening of lung cancer? ACS sens. 2016, 1 (5), 528-535. (3) Wang, C.; Cui, X. B.; Liu, J. Y.; Zhou, X.; Cheng, X. Y.; Sun, P.; Hu, X. L.; Li, X. W.; Zheng, J.; Lu, G. Y., Design of superior ethanol gas sensor based on Al-doped NiO nanorod-flowers. ACS sens. 2016, 1 (2), 131-136. (4) Xu, C.; Tamaki, J.; Miura, N.; Yamazoe, N., Grain size effects on gas sensitivity of porous SnO2-based elements. Sens. Actuators B: Chem. 1991, 3, 147-155. (5) Lee, J. H., Gas sensors using hierarchical and hollow oxide nanostructures: Overview. Sens. Actuators B: Chem. 2009, 140 (1), 319-336. (6) Lai, X. Y.; Li, J.; Korgel, B. A.; Dong, Z. H.; Li, Z. M.; Su, F. B.; Du, J. A.; Wang, D., General synthesis and gas-sensing properties of multiple-shell metal oxide hollow microspheres. Angew. Chem. Int. Ed. 2011, 50 (12), 2738-2741. (7) 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 (1), 102-110. (8) Wang, C. H.; Chu, X. F.; Wu, M. W., Detection of H2S down to ppb levels at room temperature using sensors based on ZnO nanorods. Sens. Actuators B: Chem. 2006, 113 (1), 320-323. (9) Zhang, L.; Yin, Y., Large-scale synthesis of flower-like ZnO nanorods via a wet-chemical route and the defect-enhanced ethanolsensing properties. Sens. Actuators B: Chem. 2013, 183, 110-116. (10) Han, N.; Wu, X. F.; Chai, L. Y.; Liu, H. D.; Chen, Y. F., Counterintuitive sensing mechanism of ZnO nanoparticle based gas sensors. Sens. Actuators B: Chem. 2010, 150 (1), 230-238. (11) Ahn, M. W.; Park, K. S.; Heo, J. H.; Park, J. G.; Kim, D. W.; Choi, K. J.; Lee, J. H.; Hong, S. H., Gas sensing properties of defectcontrolled ZnO-nanowire gas sensor. Appl. Phys. Lett. 2008, 93 (26), 263103. (12) Han, N.; Hu, P.; Zuo, A. H.; Zhang, D. W.; Tian, Y. J.; Chen, Y. F., Photoluminescence investigation on the gas sensing property of ZnO nanorods prepared by plasma-enhanced CVD method. Sens. Actuators B: Chem. 2010, 145, 114-119. (13) Xue, Z. G.; Cheng, Z. X.; Xu, J.; Xiang, Q.; Wang, X. H.; Xu, J. Q., Controllable evolution of dual defect Zni and VO associate-rich ZnO nanodishes with (0001) exposed facet and its multiple sensitization effect for ethanol detection. ACS Appl. Mater. Interfaces 2017, 9 (47), 41559-41567.

(14) Li, Y.; Meng, G. W.; Zhang, L. D.; Phillipp, F., Ordered semiconductor ZnO nanowire arrays and their photoluminescence properties. Appl. Phys. Lett. 2000, 76 (15), 2011-2013. (15) Fan, S.-W.; Srivastava, A. K.; Dravid, V. P., UV-activated roomtemperature gas sensing mechanism of polycrystalline ZnO. Appl. Phys. Lett. 2009, 95 (14), 142106. (16) Wang, Z.; Xue, J.; Han, D.; Gu, F., Controllable defect redistribution of ZnO nanopyramids with exposed {10-11} facets for enhanced gas sensing performance. ACS Appl. Mater. Interfaces 2015, 7 (1), 308-317. (17) Roso, S.; Güell, F.; Martínez-Alanis, P. R.; Urakawa, A.; Llobet, E., Synthesis of ZnO nanowires and impacts of their orientation and defects on their gas sensing properties. Sens. Actuators B: Chem. 2016, 230, 109-114. (18) Kim, J. C.; Rho, H.; Smith, L. M.; Jackson, H. E.; Lee, S.; Dobrowolska, M.; Furdyna, J. K., Temperature-dependent microphotoluminescence of individual CdSe self-assembled quantum dots. Appl. Phys. Lett. 1999, 75 (2), 214-216. (19) Madel, M.; Jakob, J.; Huber, F.; Neuschl, B.; Bauer, S.; Xie, Y.; Tischer, I.; Thonke, K., Optical gas sensing by microphotoluminescence on multiple and single ZnO nanowires. Phys. Status Solidi A 2015, 212 (8), 1810-1816. (20) Das, S. N.; Kar, J. P.; Choi, J.-H.; Lee, T. I.; Moon, K.-J.; Myoung, J.-M., Fabrication and characterization of ZnO single nanowire-based hydrogen sensor. J. Phys. Chem. C 2010, 114 (3), 1689-1693. (21) Tonezzer, M.; Hieu, N. V., Size-dependent response of singlenanowire gas sensors. Sens. Actuators B: Chem. 2012, 163 (1), 146152. (22) Li, W. H.; Wu, X. F.; Chen, J. Y.; Gong, Y.; Han, N.; Chen, Y. F., Abnormal n-p-n type conductivity transition of hollow ZnO/ZnFe2O4 nanostructures during gas sensing process: The role of ZnO-ZnFe2O4 hetero-interface. Sens. Actuators B: Chem. 2017, 253, 144-155. (23) Fujimura, N.; Nishihara, T.; Goto, S.; Xu, J.; lto, T., Control of preferred orientation for ZnOx films: control of self-texture. J. Cryst. Growth 1993, 130, 269-279. (24) Lu, J.; Ye, Z.; Huang, J.; Wang, L.; Zhao, B., Synthesis and properties of ZnO films with (100) orientation by SS-CVD. Appl. Surf. Sci. 2003, 207 (1-4), 295-299. (25) Kim, S. S.; Lee, B. T., Effects of oxygen pressure on the growth of pulsed laser deposited ZnO films on Si(001). Thin Solid Films 2004, 446 (2), 307-312. (26) Chang, J. F.; Kuo, H. H.; Leu, I. C.; Hon, M. H., The effects of thickness and operation temperature on ZnO:Al thin film CO gas sensor. Sens. Actuators B: Chem. 2002, 84 (2-3), 258-264. (27) Jinkawa, T.; Sakai, G.; Tamaki, J.; Miura, N.; Yamazoe, N., Relationship between ethanol gas sensitivity and surface catalytic property of tin oxide sensors modified with acidic or basic oxides. J. Mol. Catal. A: Chem. 2000, 155 (1-2), 193-200. (28) Kim, K.; Debnath, P. C.; Park, D.-H.; Kim, S.; Lee, S. Y., Controllability of threshold voltage in Ag-doped ZnO nanowire field effect transistors by adjusting the diameter of active channel nanowire. Appl. Phys. Lett. 2010, 96 (8), 083103. (29) Zhong, K.; Xu, G.; Zhang, J.; Liao, R.; Huang, Z., Work function change of Ni, HfO2 films and Ni/HfO2 interfaces as a function of external electric field. Int. J. Mod Phys B 2015, 29 (23), 1550168. (30) Mittendorfer, F.; Eichler, A.; Hafner, J., Structural, electronic and magnetic properties of nickel surfaces. Surf. Sci. 1999, 423, 1-11. (31) Thu, C.; Ehrenreich, P.; Wong, K. K.; Zimmermann, E.; Dorman, J.; Wang, W.; Fakharuddin, A.; Putnik, M.; Drivas, C.; Koutsoubelitis, A.; Vasilopoulou, M.; Palilis, L. C.; Kennou, S.; Kalb, J.; Pfadler, T.; Schmidt-Mende, L., Role of the metal-oxide work function on photocurrent generation in hybrid solar cells. Sci. Rep. 2018, 8 (1), 3559. (32) Song, S.; Hong, W.-K.; Kwon, S.-S.; Lee, T., Passivation effects on ZnO nanowire field effect transistors under oxygen, ambient, and vacuum environments. Appl. Phys. Lett. 2008, 92 (26), 263109. (33) Shi, W. S.; Cheng, B.; Zhang, L.; Samulski, E. T., Influence of excitation density on photoluminescence of zinc oxide with different morphologies and dimensions. J. Appl. Phys. 2005, 98 (8), 083502. (34) Choe, M.; Jo, G.; Maeng, J.; Hong, W.-K.; Jo, M.; Wang, G.; Park, W.; Lee, B. H.; Hwang, H.; Lee, T., Electrical properties of ZnO

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

nanowire field effect transistors with varying high-k Al2O3 dielectric thickness. J. Appl. Phys. 2010, 107 (3), 034504. (35) Jeong, S.; Choe, M.; Kang, J. W.; Kim, M. W.; Jung, W. G.; Leem, Y. C.; Chun, J.; Kim, B. J.; Park, S. J., High-performance photoconductivity and electrical transport of ZnO/ZnS core/shell nanowires for multifunctional nanodevice applications. ACS Appl. Mater. Interfaces 2014, 6 (9), 6170-6176. (36) Opoku, C.; Dahiya, A. S.; Oshman, C.; Daumont, C.; Cayrel, F.; Poulin-Vittrant, G.; Alquier, D.; Camara, N., Fabrication of high performance field-effect transistors and practical Schottky contacts using hydrothermal ZnO nanowires. Nanotechnology 2015, 26 (35), 355704. (37) Wang, F.; Du, R.; Ren, Q.; Wei, C.; Zhao, Y.; Zhang, X., Improving crystallization and electron mobility of indium tin oxide by carbon dioxide and hydrogen dual-step plasma treatment. Appl. Surf. Sci. 2017, 426, 856-863. (38) Lin, W.; Ding, K.; Lin, Z.; Zhang, J.; Huang, J.; Huang, F., The growth and investigation on Ga-doped ZnO single crystals with high thermal stability and high carrier mobility. CrystEngComm 2011, 13 (10), 3338-3341. (39) Yang, X.; Xu, C.; Giles, N. C., Intrinsic electron mobilities in CdSe, CdS, ZnO, and ZnS and their use in analysis of temperaturedependent Hall measurements. J. Appl. Phys. 2008, 104 (7), 073727. (40) Mohammad, S. N., Quantum-confined nanowires as vehicles for enhanced electrical transport. Nanotechnology 2012, 23 (28), 285707. (41) Koley, G.; Cai, Z.; Quddus, E. B.; Liu, J.; Qazi, M.; Webb, R. A., Growth direction modulation and diameter-dependent mobility in InN nanowires. Nanotechnology 2011, 22 (29), 295701. (42) Chen, C. Y.; Retamal, J. R. D.; Wu, I. W.; Lien, D. H.; Chen, M. W.; Ding, Y.; Chueh, Y. L.; Wu, C. I.; He, J. H., Probing surface band bending of surface-engineered metal oxide nanowires. Acs Nano 2012, 6 (11), 9366-9372. (43) Hong, W. K.; Sohn, J. I.; Hwang, D. K.; Kwon, S. S.; Jo, G.; Song, S.; Kim, S. M.; Ko, H. J.; Park, S. J.; Welland, M. E.; Lee, T., Tunable electronic transport characteristics of surface-architecture-controlled ZnO nanowire field effect transistors. Nano Lett. 2008, 8 (3), 950-956. (44) Fan, Z. Y.; Lu, J. G., Zinc oxide nanostructures: synthesis and properties. J. Nanosci. Nanotechnol. 2005, 5 (10), 1561-1573. (45) Pearton, S.; Norton, D.; Ip, K.; Heo, Y.; Steiner, T., Recent progress in processing and properties of ZnO. Prog. Mater Sci. 2005, 50 (3), 293-340. (46) Lupan, O.; Ursaki, V. V.; Chai, G.; Chow, L.; Emelchenko, G. A.; Tiginyanu, I. M.; Gruzintsev, A. N.; Redkin, A. N., Selective hydrogen gas nanosensor using individual ZnO nanowire with fast response at room temperature. Sens. Actuators B: Chem. 2010, 144 (1), 56-66. (47) Kolmakov, A.; Zhang, Y. X.; Cheng, G. S.; Moskovits, M., Detection of CO and O2 using tin oxide nanowire sensors. Adv. Mater. 2003, 15 (12), 997-1000. (48) McCluskey, M. D.; Jokela, S. J., Defects in ZnO. J. Appl. Phys. 2009, 106 (7), 071101. (49) Li, W. H.; Wu, X. F.; Han, N.; Chen, J. Y.; Qian, X. H.; Deng, Y. Z.; Tang, W. X.; Chen, Y. F., MOF-derived hierarchical hollow ZnO nanocages with enhanced low-concentration VOCs gas-sensing performance. Sens. Actuators B: Chem. 2016, 225, 158-166. (50) Zhang, Y.; Xu, J. Q.; Xiang, Q.; Li, H.; Pan, Q. Y.; Xu, P. C., Brush-like hierarchical ZnO nanostructures: synthesis, photoluminescence and gas sensor properties. J. Phys. Chem. C 2009, 113 (9), 3430-3435. (51) Chen, M.; Wang, Z.; Han, D.; Gu, F.; Guo, G., High-sensitivity NO2 gas sensors based on flower-like and tube-like ZnO nanomaterials. Sens. Actuators B: Chem. 2011, 157 (2), 565-574. (52) Hu, P.; Han, N.; Zhang, D. W.; Ho, J. C.; Chen, Y. F., Highly formaldehyde-sensitive, transition-metal doped ZnO nanorods prepared by plasma-enhanced chemical vapor deposition. Sens. Actuators B: Chem. 2012, 169, 74-80. (53) Wang, J. X.; Yang, J.; Han, N.; Zhou, X. Y.; Gong, S. Y.; Yang, J. F.; Hu, P.; Chen, Y. F., Highly sensitive and selective ethanol and acetone gas sensors based on modified ZnO nanomaterials. Mater. Des. 2017, 121, 69-76.

(54) Janotti, A.; Van de Walle, C. G., Native point defects in ZnO. Phys. Rev. B 2007, 76 (16), 165202. (55) Xu, P.; Sun, Y.; Shi, C.; Xu, F.; Pan, H., The electronic structure and spectral properties of ZnO and its defects. Nuclear Instruments & Methods in Physics Research Section B-Beam Interactions with Materials and Atoms 2003, 199, 286-290. (56) Lee, M. K.; Tu, H. F., Ultraviolet emission blueshift of ZnO related to Zn. J. Appl. Phys. 2007, 101 (12), 126103. (57) Srikant, V.; Clarke, D. R., On the optical band gap of zinc oxide. J. Appl. Phys. 1998, 83 (10), 5447-5451. (58) Look, D.; Hemsky, J.; Sizelove, J., Residual Native Shallow Donor in ZnO. Phys. Rev. Lett. 1999, 82 (12), 2552-2555. (59) Roro, K. T.; Dangbegnon, J. K.; Sivaraya, S.; Leitch, A. W. R.; Botha, J. R., Influence of metal organic chemical vapor deposition growth parameters on the luminescent properties of ZnO thin films deposited on glass substrates. J. Appl. Phys. 2008, 103 (5), 053516. (60) Cheng, W. D.; Wu, P.; Zou, X. Q.; Xiao, T., Study on synthesis and blue emission mechanism of ZnO tetrapodlike nanostructures. J. Appl. Phys. 2006, 100 (5), 054311. (61) Børseth, T. M.; Svensson, B. G.; Kuznetsov, A. Y.; Klason, P.; Zhao, Q. X.; Willander, M., Identification of oxygen and zinc vacancy optical signals in ZnO. Appl. Phys. Lett. 2006, 89 (26), 262112. (62) Tsai, C.-H.; Wang, W.-C.; Jenq, F.-L.; Liu, C.-C.; Hung, C.-I.; Houng, M.-P., Surface modification of ZnO film by hydrogen peroxide solution. J. Appl. Phys. 2008, 104 (5), 053521. (63) Studenikin, S. A.; Golego, N.; Cocivera, M., Fabrication of green and orange photoluminescent, undoped ZnO films using spray pyrolysis. J. Appl. Phys. 1998, 84 (4), 2287-2294. (64) Greene, L. E.; Law, M.; Goldberger, J.; Kim, F.; Johnson, J. C.; Zhang, Y. F.; Saykally, R. J.; Yang, P. D., Low-temperature waferscale production of ZnO nanowire arrays. Angew. Chem. Int. Ed. 2003, 42 (26), 3031-3034. (65) Wei, X. Q.; Zhang, Z.; Yu, Y. X.; Man, B. Y., Comparative study on structural and optical properties of ZnO thin films prepared by PLD using ZnO powder target and ceramic target. Opt. Laser Technol. 2009, 41 (5), 530-534. (66) Margueron, S.; Clarke, D. R., The high temperature photoluminescence and optical absorption of undoped ZnO single crystals and thin films. J. Appl. Phys. 2014, 116 (19), 193101. (67) Ra, H. W.; Khan, R.; Kim, J. T.; Kang, B. R.; Im, Y. H., The effect of grain boundaries inside the individual ZnO nanowires in gas sensing. Nanotechnology 2010, 21 (8), 85502. (68) Chon, H.; Pajares, J., Hall effect studies of oxygen chemisorption on zinc oxide. J. Catal. 1969, 14, 257-260. (69) Chang, S.-C., Oxygen chemisorption on tin oxide: Correlation between electrical conductivity and EPR measurements. Journal of Vacuum Science and Technology 1980, 17 (1), 366-369. (70) Yamazoe, N.; Fuchigami, J.; Kishikawa, M.; Seiyama, T., Interactions of tin oxide surface with O2, H2O and H2. Surf. Sci. 1979, 86, 335-344. (71) Lenaerts, S.; Roggen, J.; Maes, G., FT-IR characterization of tin dioxide gas sensor materials under working conditions. Spectrochimica Acta Part A-Molecular and Biomolecular Spectroscopy 1995, 51 (5), 883-894. (72) Barsan, N.; Weimar, U., Conduction Model of Metal Oxide Gas Sensors. J. Electroceram. 2001, 7 (3), 143-167. (73) Gao, L.; Ren, F.; Cheng, Z.; Zhang, Y.; Xiang, Q.; Xu, J., Porous corundum-type In2O3 nanoflowers: controllable synthesis, enhanced ethanol-sensing properties and response mechanism. CrystEngComm 2015, 17 (17), 3268-3276. (74) Gu, F.; You, D.; Wang, Z.; Han, D.; Guo, G., Improvement of gassensing property by defect engineering in microwave-assisted synthesized 3D ZnO nanostructures. Sens. Actuators B: Chem. 2014, 204, 342-350.

ACS Paragon Plus Environment

Page 8 of 9

Page 9 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors The gas sensors based on the single ZnO nanowire (NW) devices are utilized to explore the gas sensing mechanism. The results show that ~110 nm ZnO NW possesses the highest response to 5 ppm acetone, especially compared to ~80 nm ZnO NW. However, the surface charge layer (L) keeps constant of 43.6 ± 3.7 nm with D from 80 to 150 nm, calculated by electrical property measurement of the single NW field effect transistors. This is controversial to the conventional model that gas sensing maximum appears when D approximates 2L. The gas mechanism is attributed to the fact that the more the donors are, the higher the response of the sensing materials is.

ACS Paragon Plus Environment