A comparison study on hydrogen sensing performance of Pt/MoO3 nanoplatelets coated with a thin layer of Ta2O5 or La2O3

Transcription

1 Title Author(s) Citation A comparison study on hydrogen sensing performance of Pt/MoO3 nanoplatelets coated with a thin layer of Ta2O5 or La2O3 Yu, J; Liu, Y; Cai, FX; Shafiei, M; Chen, G; Motta, N; Wlodarski, W; Kalantar-zadeh, K; Lai, PT The 8th Annual IEEE International Conference on Nano/Micro Engineered and Molecular Systems (IEEE NEMS 23), Suzhou; China, 7- April 23. In Conference Proceedings, 23, p Issued Date 23 URL Rights IEEE International Conference on Nano/Micro Engineered and Molecular Systems (NEMS) Proceedings. Copyright IEEE.

2 A comparison study on hydrogen sensing performance of Pt/ nanoplatelets coated with a thin layer of or J. Yu *, Y. Liu, F.X. Cai, M. Shafiei, G. Chen, N. Motta, W. Wlodarski, K. Kalantar-zadeh, P.T. Lai Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong SAR 2 School of Chemistry, Physics and Mechanical Engineering, Queensland University of Technology, Australia 3 School of Electrical and Computer Engineering, RMIT University, Australia *Corresponding Author: Jerry Yu, jcwyu@hku.hk Abstract In this work, we investigate how hydrogen sensing performance of thermally evaporated nanoplatelets can be further improved by RF sputtering a thin layer of tantalum oxide ( ) or lanthanum oxide ( ). We show that dissociated hydrogen atoms cause the thin film layer to be polarised, inducing a measurable potential difference greater than that as reported previously. We attribute these observations to the presence of numerous traps in the thin layer; their states allow a stronger trapping of charge at the Pt-thin film oxide interface as compared to the sensors without the coating. Under exposure to H 2 ( ppm), the maximum change in dielectric constant is 45.6 (at 26 C) for the / nanoplatelets and 3.6 (at 22 C) for the / nanoplatelets. Subsequently, the maximum sensitivity for the / and / based sensors is 6.8 and 7.5, respectively. Keywords- hydrogen; sensor; metal oxide; heterostructure I. INTRODUCTION With the rise of the capability to control and restructure matter at the nanoscale, research and development of cheap, low-power, miniaturized hydrogen sensors with enhanced performance have become a priority for renewable energy related devices. Such devices which convert hydrogen gas to energy, such as fuel cells, internal-combustion engines and turbines all require high-performance sensors, due to the highly volatile nature of hydrogen gas [4]. Sensors based on semiconducting metal oxides have arisen as one of the most favorable choices for industrial applications due to their simplicity, lightweight and portability and have been demonstrated down-scaling and compatibility with existing technologies [5]. Sensing performance of nanostructured metal oxide materials can excel far beyond those based on bulk materials, due to the unique properties that are exceptional only at the nanoscale [2, 3]. Herein, we examine two high-κ metal oxides ( and ) as they are deposited on nanoplatelets for hydrogen sensing. We use a quantum heterostructure design [6] in this work, as our motive is to implement the quantum confinement behaviour to trap charge and induce a strong Pt 3.3 ev 4.4 ev.35 ev Fig.. The proposed energy band structure of the metal oxides, and [2, 3] with Pt, as implemented in this work. polarisation at the interface between the Pt and the high- layers. The proposed energy band structure of the sensors is as illustrated in Fig.. II. 3. ev 5.65 ev METHODOLOGY A. Fabrication of nanoplatelets The nanoplatelet films were grown on n-type 6H-SiC substrates (Tankeblue Co.) using thermal evaporation deposition in a tube furnace. The details outlining the fabrication process can be referenced to our previous work [, 7, 8]. B. Fabrication of and high- thin film layers RF Sputtering deposition of the high- thin film layer was performed in a Denton vacuum discovery sputtering system as previously reported [9, ]. The thickness of the and layer were calibrated to 4 nm prior to sputtering using a Spectroscopic Ellipsometer (VASE VB4). C. Electrical and gas sensing testing procedure The developed sensors with the configuration of Pt/high-κ thin film/ nanoplatelets/sic were placed in a multi- Pt 2.4 ev 6. ev 2.3 ev 2.3 ev 3. ev

3 channel gas testing system for the electrical and sensing tests. Experimental measurements were conducted by using a needle probe similar to procedures as presented in our previous work [, 7, 9-2]. III. RESULTS A. Surface morphology and crystallogrphic structure Analysis of the surface morphology of the as-deposited nanoplatelets conducted by SEM is shown in Fig. 2. The insets of the figure also show the TEM of the nanoplatelets subsequent to the RF sputtering of or (Fig. 2) indicating the thickness of the deposited layer. The crystallographic structure of the nanoplatelets, with and as measured by XRD is shown in the Fig cts. 4 nm m nm 4 nm nm SEM Fig. 2. SEM of the nanoplatelets deposited onto SiC by thermal evaporation (insets shows TEM of the 4 nm RF sputtered and thin film). nanoplatelets [] XRD o Fig. 3. XRD diffractograms of the nanoplatelets [] deposited onto SiC by thermal evaporation with the presence of (as indicated by hollow squares) and (as indicated by the filled circles). The characterisation results indicate that the RF sputtered and layers can be applied onto nanoplatelets as a thin coating without significantly altering the surface morphology. The results from the XRD also show that the original crystal structure is preserved [, 8] with the exception of additional peaks of for the RF sputtered layer. Similar observations can be made for nanoplatelets with the RF sputtered layer. B. Current density vs Voltage (J-V) charcteristics and ideality factor The electrical properties of a metal/metal oxide can be characterised in terms of its current density vs voltage (J-V) characteristics as governed by the Schottky diode model. Based on the thermionic emission (TE) current transport mechanism, the forward J-V equation is given by [3]: ** 2 q B q VF J F A T exp exp () kt kt where A ** is the effective Richardson constant; T is the absolute temperature; q is the electron charge; B is the effective barrier height; k is the Boltzmann constant. According to the Poole-Frenkel (PF) current transport mechanism, the dielectric constant can be estimated from the slope of the forward bias semi-log J-V /2 characteristics in (2) and its equation is given by) [4]: log J F 2.3k T 2.3k T l 2 logb V where 3 q 2 t ox (2) B B The J-V characteristics of the diodes based on nanoplatelets without and with and coating measured at 25C are shown in Fig. 4. The inset of Fig. 4 shows the ideality factor, estimated from the J-V data selected between the region. V to.3 V as compared with the region of.5 V to V as selected in previous work [, 7, 9]. This region was chosen as the TE model is suitable for this range of data. J (A/cm 2 ) / 5 4 La 3 2 / 2 MoO Pt/ Pt/ / V (V) Fig. 4. J-V characteristics of the nanoplatelets, without and with and coating (inset shows the estimated ideality factor as extracted from the region between. V to.3v). The J-V characteristics of the nanoplatelets with the and coating shows strong non-ideal behavior and therefore, indicate that PF emission can be a dominant conduction mechanism. In this work, we have selected a range between.5 V to 2 V to calculate the -value, respectively as Pt/ /

4 the PF model is suited for this region of data. We have also neglected the overlapping of TE to PF to simplify the analysis. Of the J-V characteristics measured at 25 C (Fig. 4), we can use the ideality factor as an indicator to identify the number of traps in the material to compare the nanoplatelets with and without the or layers. According to the inset of Fig. 4, the ideality factor of the nanoplatelets with the and thin film layer were 5.92 and 5.23 at 25 C, respectively as opposed to an ideality factor of.77 for the nanoplatelets without these layers. We expect these values to indicate that the effect of RF sputtering or layer generating numerous additional interface traps between the Pt metal and their surface as identified in literature [2, 3]. C. Effective barrier height and dielectric constant In this section, we calculate the effective barrier height and dielectric constant using equation () and (2) and discuss their significance with respect to temperature. Fig. 5 shows the plot of the effective barrier height for the nanoplatelets. B (mev) / / TE PF Dielectric Constant.7 / Effective Barrier Height / For the nanoplatelets coated with (or ) we can interpret from the data that at temperatures less than 8 C, the electrical properties follow the TE mechanism and at temperatures greater than 8 C, the properties follow the PF mechanism. According to the TE model, electrons are barred from flowing through the interface unless they acquire sufficient thermal energy to flow over the Schottky barrier [3]. However, in the PF mechanism, when there is sufficient thermal energy, numerous interface traps (from the Ta 5+ or La 3+ atoms) are activated allowing electrons an alternative pathway through the flow barrier [3]. Therefore, we calculate the effective barrier height from the data between 25 C to Fig. 5. Effective barrier height B (as denoted by the hollow circles) and dielectric constant (as denoted by the filled squares) for the nanoplatelets, with and without and coating at different temperatures. The thermionic emission (TE) model is found to be dominant allowing for an accurate calculation of the effective barrier height, from 25 C to 35 C for the uncoated and from 25 C to 4C for the with the coatings. Thus, from 8 C to 3 C, the Poole Frenkel (PF) mechanism is dominant, allowing for the calculation of the dielectric constant for the nanoplatelets with and. C using equation (), and also calculate the dielectric constant from the data at temperatures above 8 C by equation (2). D. Static hydrogen sensing performance The plot of the change in effective barrier height B and change in dielectric constant (under exposure to hydrogen with, ppm concentration) at different temperatures is shown in Fig. 6. B (mev) / / / / With the presence of a (or ) layer, the maximum change in dielectric constant under exposure of hydrogen gas (, ppm) was calculated as 45.6 (at 26C) for the nanoplatelets with and 3.6 (at 22C) for the nanoplatelets with the. These values represent a significant change in the polarisation intensity at the Pt/metaloxide interface caused by the presence of hydrogen gas. These observations can be explained in terms of the hydrogen dissociation and adsorption mechanism [4, 5, ]. As hydrogen molecules adsorb on the Pt surface, they can undergo catalytic dissociation and diffuse through into the Pt metal. Subsequently, hydrogen atoms accumulate at the interface between Pt and the metal oxide as adsorbates. Here, a net positive charge layer is accumulated on the metal side of the interface, inducing a net negative charge layer on the metal-oxide side. This dipolar charges layer, polarises the interface [5] which can be measured by the change in the value (in terms of the PF model) and also the change in potential difference (from across this thin layer). Simultaneously, the dipolar charge causes the bending of the energy bands and thus, lowers the barrier height (in terms of the TE model). We have previously discussed the confinement effects of the 4 nm thin film layer and how potential difference is consumed to drive the PF current for coated nanoplatelets [7]. In this work, we have further discussed the Fig. 6. Plot of the effective change barrier height B (as denoted by the hollow circles) and effective change in dielectric constant (as denoted by the filled squares) for nanoplatelets, and and coated nanoplatelets.

5 degree of its effects on the polarisation at the interface in terms of the change in the -value. E. Sensitivity The plot of the hydrogen sensitivity (defined as S = I H2 I air / I air ) and the calculated maximum values for the coated and uncoated nanoplatelets are shown in Fig. 7. The results indicate that the coating can achieve a significantly higher hydrogen sensitivity than the coating, this is in good agreement with the as shown in Fig. 7. Sensitivity exhibited by the nanoplatelets, with and without and coating in the presence of hydrogen gas (, ppm) at different temperatures. Fig. 6. F. Dynamic hydrogen sensing performance The dynamic response towards hydrogen for and coated nanoplatelets at 22 C are shown in Fig. 8. We show that the dissociated hydrogen atoms cause the thin film layer to be polarised and also induce a measurable potential difference, greater than that as reported previously [, 7]. The results indicate that the coated nanoplatelets can exhibit a stable response and saturation to steady state at the presence of, ppm hydrogen. V (V) S S = 7.52 S =.33 S = 6.87 / ppm 25 ppm 2 5 ppm / / At A (22 o C) Time (min) Fig. 8. Dynamic hydrogen sensing performance of the and coated nanoplatelets as exposed to different concentrations of H 2 gas for 5 min at 22 C, under A bias current. 5 ppm ppm / IV. CONCLUSIONS In this work, we have presented how and as a thin film coating can substantially improve the hydrogen sensing performance of nanoplatelets. Under exposure to H 2 (, ppm), the maximum change in dielectric constant is 45.6 (at 26 C) for the / nanoplatelets and 3.6 (at 22 C) for the / nanoplatelets. The maximum sensitivity is 6.8 (at 26 C) for the / based sensor and 7.5 (at 3 C) for the / based sensor. We show that these materials can achieve steady state under exposure to different hydrogen concentrations of H 2, which indicate that these materials can be applied as potential candidates for future hydrogen sensing applications. ACKNOWLEDGMENT The authors of this work would like to acknowledge the University Development Fund (Nanotechnology Research Institute, 69) of The University of Hong Kong. REFERENCES [] J. Yu, et al., "Reverse biased Pt/nanostructured /SiC Schottky diode based hydrogen gas sensors," Applied Physics Letters, vol. 94, Jan 29. [2] J. Robertson, "High dielectric constant oxides," European Physical Journal-Applied Physics, vol. 28, pp , Dec 24. [3] J. Robertson and P. W. Peacock, "Bonding and structure of some high-k oxide: Si interfaces," Physica Status Solidi B-Basic Research, vol. 24, pp , Aug 24. [4] T. Hubert, et al., "Hydrogen sensors - A review," Sensors and Actuators B-Chemical, vol. 57, pp , Oct 2. [5] G. Korotcenkov, "Metal oxides for solid-state gas sensors: What determines our choice?," Materials Science and Engineering B-Solid State Materials for Advanced Technology, vol. 39, pp. -23, Apr 27. [6] J. Schalwig, et al., "Gas sensitive GaN/AlGaN-heterostructures," Sensors and Actuators B-Chemical, vol. 87, pp , Dec 22. [7] M. Shafiei, et al., "Improving the hydrogen gas sensing performance of Pt/ nanoplatelets using a nano thick layer of," Sensors and Actuators B: Chemical, DOI:.6/j.snb (in press). [8] J. Yu, et al., "Enhancement of electric field properties of Pt/nanoplatelet /SiC Schottky diode," Journal of Physics D-Applied Physics, vol. 43, Jan 2. [9] J. Yu, et al., "Hydrogen gas sensing properties of Pt/ Schottky diodes based on Si and SiC substrates," Sensors and Actuators, A: Physical, vol. 72, pp. 9-4, 2. [] G. Chen, et al., "A study on MIS Schottky diode based hydrogen sensor using as gate insulator," Microelectronics Reliability, vol. 52, pp , 22. [] M. Shafiei, et al., "Reversed bias Pt/nanostructured ZnO Schottky diode with enhanced electric field for hydrogen sensing," Sensors and Actuators, B: Chemical, vol. 46, pp , 2. [2] J. Yu, et al., "Nanorod based Schottky contact gas sensors in reversed bias condition," Nanotechnology, vol. 2, 2. [3] S. M. Sze and K. K. Ng, Physics of semiconductor devices: Wileyinterscience, 26. [4] W. Mönch, "Metal-semiconductor contacts: electronic properties," Surface Science, vol , pp , 994. [5] X. F. Chen, et al., "Microstructure, dielectric properties and hydrogen gas sensitivity of sputtered amorphous Ba.67Sr.33Ti thin films," Materials Science and Engineering B-Solid State Materials for Advanced Technology, vol. 77, pp , Aug 2.