Impact of postdeposition annealing on the sensing and impedance characteristics of TbY_xO_y electrolyte–insulator–semiconductor pH sensors

Abstract

In this investigation, we explored the impact of postdeposition annealing (PDA) on the sensing and impedance characteristics of TbY_xO_y sensing films deposited on Si(100) substrates through reactive cosputtering for electrolyte–insulator–semiconductor (EIS) pH sensors. The TbY_xO_y EIS sensor annealed at 800 °C exhibited the best sensing characteristics (pH sensitivity, hysteresis voltage, and drift rate). In addition, the effect of PDA treatment on the impedance properties of TbY_xO_y EIS sensors was studied using the capacitance–voltage method. The resistance and capacitance of TbY_xO_y sensing films were determined using different frequency ranges in accumulation, depletion, and inversion regions. In our impedance spectroscopy analysis, the semicircle diameter of the TbY_xO_y EIS sensor became smaller, due to a gradual decrease in the bulk resistance of the device, as the PDA temperature was increased.

---

Introduction

An ion-selective field-effect transistor (ISFET) as a semiconductor device is currently constructed by replacing an ion-sensing membrane for the metal gate of a metal–oxide-semiconductor field-effect transistor (MOSFET).¹ The ISFET is able to respond to a change in the surface potential arising from the acid–base reactions occurring at the oxide–electrolyte interface. Semiconductor field-effect devices based on an electrolyte–insulator–semiconductor (EIS) system are generally the basic structural components of chemical and biological sensors. These devices have been established in versatile tools for determining pH and the concentrations of ions, as well as detecting enzymatic reactions, antigen–antibody recognition events, DNA hybridisation, etc.^2–4 The gate insulator material is an important element in an ISFET or EIS device, because the gate electrode is directly in contact with the solution. ISFET or EIS devices designed to determine pH, which is the most widely used application of these sensors, are fabricated with a large range of possible insulators (e.g., SiO₂, Si₃N₄, Al₂O₃, ZrO₂, and Ta₂O₅).^5–7 Although the Si₃N₄ film displays high strength, high thermal-shock resistance and good wear resistance, it is strongly reactive and can be readily oxidized upon exposure to air, thus leading to the degradation of sensing performance.^7,8 Apart from the Si₃N₄ film, Al₂O₃ and ZrO₂ films as sensing materials are poor as a result of their low sensitivity and narrow pH range.⁷ Compared to these materials, the Ta₂O₅ film is more sensitive and displays less drift,^7,9 but its deposition on the Si produces a tantalum silicide layer and it has a narrow band gap, which degrade its electrical properties.¹⁰ It has therefore not been easy to achieve a reliable pH sensor with high sensitivity and low drift.

The scaling of the gate dielectric has been accomplished by shrinking the physical dimensions of the complimentary metal–oxide semiconductor (CMOS). A number of fundamental problems, e.g., high gate leakage current and oxide reliability, arise when the thickness of the SiO₂ gate dielectric approaches the limit of ∼2 nm. To overcome this fundamental limit, new materials with high dielectric constant (κ) values have been studied.¹¹ Rare-earth (RE) oxide thin films, including La₂O₃, Nd₂O₃, Sm₂O₃, and Tb₂O₃,^12–14 have been widely investigated as alternative gate dielectrics for CMOS applications because they offer several advantages, including wide band gaps, large band offsets to Si, and improved thermal and thermodynamic stabilities, as well as high dielectric constants. Among these RE oxides, terbium oxide (Tb₂O₃) has recently attracted a lot of attention as a promising high-κ gate dielectric due to its high conduction band offset, large band gap, and good thermal stability.^12,15 However, most high-κ RE oxide films are hygroscopic under air, and hence a layer of hydroxide forms on their surfaces.¹⁶ As the moisture absorption of the RE oxide increases, so does its surface roughness, resulting in a deterioration of the electrical properties.¹⁷ This problem can be solved by incorporating Ti or Y into the RE oxide film, which results in less reactivity toward water.^16,18

In an EIS structure, the charge and mass transfer at the interface of the electrolyte and electrode have different time constants and can be modeled as a combination of resistors and capacitors. Generally, they consist of frequency-independent resistive and capacitive components that indicate the charge storage capacity and the relaxation times of an individual process.^19,20 Impedance measurements as a function of frequency allow one to monitor the effects of interfacial layers, by evaluation of the time constants related to the charge and mass transfer at the interface, and providing information on changes in capacitance/resistance occurring at conductive or semiconductive for sensing applications. Impedance spectroscopy is becoming an attractive electrochemical tool to characterize biomaterial films corresponding to electronic elements, and to thereby allow detection of biorecognition events at the surfaces of the films. So far, to the best of our knowledge, experimental data associated with the impact of postdeposition annealing (PDA) on the impedance properties of the TbY_xO_y sensing films have not yet been acquired. Here, we studied the sensing characteristics of the TbY_xO_y films deposited on an Si substrate by means of reactive co-sputtering and subsequent PDA at three temperatures. Furthermore, we investigated the impedance properties of the TbY_xO_y sensing films under accumulation, depletion and inversion regions, and at pH values of 4, 7, and 10.

Experimental

Fig. 1(a) shows the EIS sensor structure with a TbY_xO_y sensing film. The TbY_xO_y EIS device was fabricated on a 4-inch p-type (100) Si wafer. The Si substrates were cleaned by using a standard Radio Corporation America cleaning process and then by using diluted HF to remove the native oxide from the surface of the Si wafer. Thin (∼40 nm) TbY_xO_y films were deposited on the Si substrate through reactive co-sputtering from both Tb and Y targets in diluted O₂. Then, the PDA treatment of the films was carried out in a rapid thermal annealing (RTA) system with three annealing temperatures (700, 800 and 900 °C) under an O₂ ambient for 30 s. The backside contact (a 400 nm-thick Al film) of the Si wafer was deposited using a thermal coater. The sensing area of the deposited TbY_xO_y films was defined by an automatic robot dispenser with an adhesive silicone gel (S181) acting as an isolating layer. Finally, the EIS device was assembled on the copper lines of a custom-made printed circuit board by applying a silver gel, and another adhesive epoxy was used to prevent leakage of the electrolyte.

Fig. 1 Schematic illustration of (a) the structure of the TbY_xO_y EIS pH sensor and (b) the electrical measurement setup.

The pH-sensing performance of the EIS sensors was determined using standard buffer solutions in the pH range of 2 to 12. The capacitance–voltage (C–V) curves for the TbY_xO_y EIS sensors in the standard buffer solutions at different values of pH were then measured using a commercial Ag/AgCl reference electrode and using a Hewlett-Packard (HP) 4284A LCR meter. As shown in Fig. 1(b), a high potential (H_pot) and high current (H_cur) were applied to the Ag/AgCl reference electrode and a low potential (L_pot) and low current (L_cur) to the electrode of the EIS device in the 4284A LCR meter. The GPIB address of the 4284A LCR meter was extracted onto the interface program, and then the computer started communicating with the 4284A LCR meter. The impedance measurements of the EIS devices were taken using a combination of HP 4284A (100 Hz to 500 kHz) and 4285A (500 kHz to 10 MHz) precision LCR meters to achieve a total frequency range of 100 Hz to 10 MHz. We used three sensors to measure their electrical properties. For a given experiment, each condition was tested in triplicate. To reduce any interference, all experimental setups were maintained in the dark and performed at room temperature.

Results and discussion

A. Sensing characteristics of TbY_xO_y EIS devices

In an ISFET or EIS device, the flatband voltage (V_FB) is defined as²¹where E_Ref is the potential of the reference electrode, ϕ₀ is the pH-dependent surface potential, χ_sol is the surface dipole potential of the solution, Φ_S is the work function of electrons in the semiconductor, q is the elementary charge, Q_ox is the charge located in the oxide, Q_SS is the charge located on the surface and interface, and C_ox is the gate oxide capacitance. The flatband voltage was shifted by the surface potential, suggesting it to be mainly dependent on the charge concentration. The site-binding model was used to describe the proton transfer reactions at the oxide/solution interface. The oxide surface assumed the number of hydroxyl groups that can be found in the neutral MOH state, the positively charged (protonated) MOH₂⁺ state, or the negatively charged (deprotonated) MO⁻ state. The transduction of pH to a potential difference ϕ₀ between surface and bulk solution was performed through a pH-dependent proton or hydroxide density gradient. Under acid conditions, a Boltzmann–Poisson equation was used to relate the surface potential to the proton density difference. This equation is²²where [H⁺]_x is the proton concentration at the oxide surface or in the bulk of the solution, k_B is Boltzmann's constant, and T is the absolute temperature. At a high pH value (low proton concentration in the solution), the equilibrium is shifted toward a deprotonated surface, i.e., one that is negatively charged. And, indeed, the C–V curve was shifted to the right. The relationship between the surface potential and the pH of the solution can be interpreted using the combination of the site-binding model and Gouy–Chapman–Stern theory, which also characterize the properties of the interface between the buffer solution and the gate dielectric. The pH sensitivity of the gate dielectric surface can be expressed as²¹where α is a sensitivity parameter (dimensionless) that varies between 0 and 1 depending on the proton concentration of the buffer solution and the intrinsic buffer capacitance of the gate dielectric. The pH sensitivity value of the gate dielectric surface approaches a maximum (59.5 mV pH⁻¹, at room temperature) as the value of α approaches to unity.

In general, the pH sensitivity of an EIS device is determined by the change in the flatband voltage per unit change of pH of the buffer solution. Fig. 2(a)–(c) depict the C–V curves of the TbY_xO_y EIS sensors annealed at 700, 800 and 900 °C. In EIS pH sensing, a change in pH leads to a reference voltage (V_Ref) shift of the C–V curves as a result of ionization of the surface OH groups by H⁺ or OH⁻ ions,²² and thereby changes the surface potential by forming dipoles on the sensing film. The V_Ref values were here calculated from the C–V curves corresponding to at a normalized capacitance of 0.5. The capacitance was normalized by dividing it by the capacitance value at the substrate voltage of −1 V. The TbY_xO_y EIS pH sensor exhibited a linear response to increases in the pH from 2 to 12. To estimate the pH detection sensitivities of the TbTi_xO_y sensing films that had been subjected to various PDA temperatures, we recorded their C–V curves and then calculated their sensing performances. As shown in Fig. 2(d), the TbY_xO_y EIS sensor after PDA at 800 °C exhibited a pH sensitivity of 61.63 ± 3.5 mV pH⁻¹, greater than those at other temperatures (52.72 ± 3.4 mV pH⁻¹ for 700 °C and 54.57 ± 3.6 mV pH⁻¹ for 900 °C). This dependence on temperature could be attributed to PDA at 800 °C producing a rougher surface (see ESI†), thus causing a higher density of surface sites in the TbY_xO_y film.²³ Moreover, the pH sensitivity resulting from this annealing temperature was higher than the theoretical value of 59.5 mV pH⁻¹. The super-Nernstian pH response of a TbY_xO_y EIS sensor could be associated with the adsorption of specific molecules²⁴ because PDA at 800 °C under an O₂ atmosphere could fill in the oxygen vacancies and dangling bonds,²⁵ thus leading to increases in the dielectric constant and in the capacity of the compact layer to change the surface charge of the TbY_xO_y film. Furthermore, this behaviour can contribute to the mechanism involving one transferred electron per 1.5 H⁺ ions.²⁶ During high-temperature annealing, a thicker silicate layer can form at the TbY_xO_y–Si substrate interface.¹² The stability of the TbY_xO_y EIS sensor annealed at 800 °C was measured every three days. After 90 days, the sensitivity of this sensor was 60.48 ± 3.7 mV pH⁻¹. That is, the sensitivity of TbY_xO_y EIS sensor was stable.

Fig. 2 (a) C–V curve responses for TbY_xO_y EIS devices annealed at (a) 700, (b) 800, and (c) 900 °C, when inserted into solutions at different pH values. (d) Reference voltage plotted as a function of pH for TbY_xO_y EIS devices annealed at three temperatures and measured at room temperature.

In addition to the sensitivity detection, hysteresis and drift effects are two other important factors determining the success of an EIS sensor. Fig. 3(a) shows the hysteresis effect of the TbY_xO_y EIS sensor devices after PDA at three temperatures, with each device directly immersed in each pH standard solution for up to 1500 s in a set pH 7 → 4 → 7 → 10 → 7 cycle. Here, the hysteresis voltage was determined from the difference between the initial and terminal reference voltages measured in above cycle. The TbY_xO_y EIS sensor annealed at 800 °C displayed a hysteresis voltage of 1 ± 0.17 mV, lower than those at the other temperatures (14 ± 1.3 mV for 700 °C and 6 ± 0.6 mV for 900 °C), presumably suggesting a stoichiometric TbY_xO_y film in which the oxygen vacancies and chemical defects, and hence crystal defects, were suppressed. In contrast, the film annealed at 700 °C exhibited a higher hysteresis voltage than did the films subjected to other PDA temperatures, suggesting a higher density of crystal defects in this sensing film. We also estimated the drift effects of the EIS sensors by measuring the variation in the reference voltage after submerging them in a standard buffer solution at pH 7 for up to 12 h. The change in the reference voltage can be defined as ΔV_Ref = V_Ref(t) − V_Ref(0). Fig. 3(b) depicts the drift characteristics of the TbY_xO_y EIS sensors annealed at the three temperatures. In each case, the slope of the plot of the deviation from the reference voltage versus time was used as a measure of the stability of the tested EIS sensor. The TbY_xO_y EIS device after PDA at 800 °C exhibited better long-term stability (slope = 0.33 ± 0.02 mV h⁻¹) than did the other studied systems (0.71 ± 0.06 mV h⁻¹ for 700 °C and 0.41 ± 0.03 mV h⁻¹ for 900 °C), as shown in Table 1. This low drift rate may be attributed to the reduction in the oxygen vacancies and dangling bonds,²⁵ which would have led to lower ionic mobility. The sample annealed at 700 °C displayed the highest drift rate, possibly suggesting a relatively high density of chemical defects in this film.

Fig. 3 (a) Hysteresis characteristics as a function of time for TbY_xO_y EIS devices annealed at three temperatures, during the loop of pH 7 → 4 → 7 → 10 → 7. (b) Drift phenomena as a function of the amount of time that the TbY_xO_y EIS devices were annealed, for three temperatures, with each one tested at pH 7.

Table 1 pH response parameters, including pH sensitivity, hysteresis voltage, and drift rate, for all TbTi_xO_y samples

PDA temperature (°C)	pH sensitivity (mV pH^−1)	Hysteresis voltage (mV)	Drift rate (mV h^−1)
700	52.72 ± 3.4	14 ± 1.3	0.71 ± 0.06
800	61.63 ± 3.5	1 ± 0.17	0.33 ± 0.02
900	54.57 ± 3.6	6 ± 0.6	0.41 ± 0.03

---

The measured and extracted sensing parameters are summarized in Table 2, where the data from EIS or ISFET devices using Al₂O₃,²⁷ Ta₂O₅,²⁸ ZrO₂,²⁹ CeO₂,²⁵ Gd₂O₃,³⁰ Y₂O₃,³¹ and Tb₂O₃ (ref. 32) sensing films, are shown for comparison.

Table 2 Comparison of the sensing performances of EIS or ISFET sensors fabricated with Al₂O₃, Ta₂O₅, ZrO₂, CeO₂, Gd₂O₃, Y₂O₃, Tb₂O₃, and TbY_xO_y sensing films

Sensing film	pH sensitivity (mV pH^−1)	Hysteresis voltage (mV)	Drift rate (mV h^−1)
Al2O3 (ref. 27)	52–58	∼1	0.3
Ta2O5 (ref. 28)	55–58	∼1	<0.5
ZrO2 (ref. 29)	57.1	x	∼5.9
CeO2 (ref. 25)	58.76	5.97	0.96
Gd2O3 (ref. 30)	53.1	x	5.4
Y2O3 (ref. 31)	54.5	4.8	2
Tb2O3 (ref. 32)	56.97	12	0.446
TbYxOy	61.63	1	0.33

---

The TbY_xO_y sensing film displayed a better pH sensitivity than did the other sensing materials. The hysteresis voltage and the drift rate of the TbY_xO_y sensing film were also comparable to those of the Al₂O₃ and Ta₂O₅ films. We therefore suggest that the TbY_xO_y sensing film is suitable for applications in EIS devices.

B. Impedance characteristics of TbY_xO_y EIS devices

Impedance spectroscopy is an effective method for probing the features of surface-modified electrodes.³³ A perturbing small-amplitude sinusoidal voltage signal was applied to the electrochemical device, and the resulting current was measured. The oxide–solution interface can be represented by an equivalent circuit, as shown in Fig. 4(a), where Z_S denotes the Si substrate impedance (parallel R_S and C_S circuit), C_sc the space charge capacitance, R_it and C_it the interface trap resistance and capacitance, respectively, C_dl the double-layer capacitance, R_ct the charge (or electron) transfer resistance that exists if a redox probe is present in the electrolyte solution, Z_W the Warburg impedance arising from the diffusion of redox probe ions from the bulk electrolyte to the electrode interface, and Z_Ref the reference electrode impedance. Note that both Z_S and Z_W represent bulk properties and are not expected to be affected by an EIS structure on an electrode surface. On the other hand, C_dl and R_ct depend on the dielectric and insulating properties of the oxide–electrolyte solution interface.

Fig. 4 (a) Model of circuit elements present when the EIS device is immersed in a pH solution and (b) simplified equivalent circuit of an EIS device used in the study. Equivalent elements: Z_s, Si substrate impedance; C_sc, space charge capacitance; C_it, interface trap capacitance; R_it, interface trap resistance; C_ox, gate oxide capacitance; C_dl, double-layer capacitance; R_ct, charge transfer resistance; Z_W, Warburg impedance; Z_Ref, reference impedance; C_p, frequency-dependent capacitance; G_p, frequency-dependent conductance.

Fig. 5–7 show Nyquist plots obtained by plotting Z_Im vs. Z_Re for different solution pH values and different PDA temperatures. Several features can be identified in a Nyquist plot that relate to rate-determining processes at the oxide–solution interface. Each of the plots in these figures shows a semicircle with a centre located on the Z_Re axis at higher frequencies, followed by a straight line at lower frequencies. In general, the semicircle feature is related to the electron transfer-limited process, whereas the linear region is associated with the diffusion-limited process of the system. In addition, the semicircle is generally offset on the Z_Re axis (as ω → ∞) by a value corresponding to the magnitude of Z_S. For very rapid electron transfer processes, the Nyquist plot shows only the linear part of the impedance spectrum. In contrast, for an electron transfer process that is the rate-determining step, the impedance spectrum shows a large semicircle without a straight line. Note that the diameter of the semicircle equals R_ct and extrapolation of the semicircle to lower frequencies produces an intercept corresponding to (R_S + R_ct). For a MOSFET device under the depletion region, the conductance method is generally considered to be the most sensitive method to determine D_it in the depletion and weak portion of the band gap. It reflects the loss mechanism as a result of interface trap capture and emission of carriers.³⁴ Assuming small losses in capacitance from the oxide and depletion, the measured conductance under depletion stems from the contribution of the interfacial trap states. The simplified equivalent circuit of an EIS device appropriate for the conductance method is shown in Fig. 4(b).

Fig. 5 Nyquist diagrams (Z_im vs. Z_re) of the TbY_xO_y EIS devices annealed at three temperatures and tested in the (a) accumulation, (b) depletion and (c) inversion regions, when immersed in a pH 4 solution.

Fig. 6 Nyquist diagram (Z_im vs. Z_re) of the TbY_xO_y EIS devices annealed at three temperatures and tested in the (a) accumulation, (b) depletion and (c) inversion regions, when immersed in a pH 7 solution.

Fig. 7 Nyquist diagram (Z_im vs. Z_re) of the TbY_xO_y EIS devices annealed at three temperatures and tested in the (a) accumulation, (b) depletion and (c) inversion regions, when immersed in a pH 10 solution.

Based on the simplified equivalent circuit model in Fig. 4(b), two types of semicircles may be generally observed in impedance plots: one being related to the bulk membrane response at higher frequencies, and the other associated with the oxide–solution interface response at lower frequencies. As shown in Fig. 5–7, our TbY_xO_y EIS sensors yielded single semicircles, with frequencies ranging from 0.1 to 10 MHz.

As shown in Fig. 5(a)–(c), in which the films were tested in a pH 4 solution, that annealed at 700 °C yielded a semicircle radius larger than did the films produced at the other temperatures, indicating lower bulk resistances of the EIS devices annealed at higher temperatures. In addition, the radius of the semicircle in the inversion region was observed to be larger than those in the accumulation and depletion regions. Fig. 6(a)–(c) show the Nyquist plots of the impedance spectra of the TbY_xO_y EIS sensors annealed at the three temperatures and then tested in a pH 7 solution under the three regions. The diameters of the semicircles resulting from these EIS devices clearly decreased with increasing PDA temperature, and were greater in the depletion and inversion regions than in the accumulation region. Fig. 7(a)–(c) show the Nyquist plots of the impedance spectra of the TbY_xO_y EIS sensors annealed at the three temperatures and then tested in a pH 10 solution under the three regions. Here, the highest frequency led to the lowest real impedance value (small semicircle), while the lowest frequency caused the highest real impedance value (large semicircle).

The approximately semicircular plots in Fig. 5–7 are typically seen in simple electrochemical systems. The diameters of the semicircles resulting from testing the devices in the pH 10 (alkaline) solution were larger than those from the pH 4 (acid) solution, suggesting a gradual increase in the bulk resistance of the EIS devices with increasing pH. The rate of diffusion of H₃O⁺ ions was lower than that of the OH⁻ ions due to their larger size. Compared with both the accumulation and depletion regions, the TbY_xO_y EIS device under the inversion region exhibited a higher impedance. Furthermore, for each of the tested pH values, the TbY_xO_y EIS device after PDA at 700 °C exhibited a greater bulk resistance than did the devices from the other PDA temperatures.

Conclusions

In this study, we explored the impact of PDA on the sensing and impedance characteristics of TbY_xO_y sensing films deposited on Si(100) substrates through reactive cosputtering. The TbY_xO_y EIS sensor after PDA at 800 °C exhibited a high pH sensitivity of 61.63 mV pH⁻¹, a small hysteresis voltage of 1 mV and a low drift rate of 0.33 mV h⁻¹. In addition, the impact of PDA on the impedance characteristics of TbY_xO_y EIS sensors was investigated by means of the C–V method. The diameter of the Nyquist plot semicircle for TbY_xO_y EIS device measured at the inversion region was greater than those at the accumulation and depletion regions. Finally, the bulk resistance decreased upon increasing the PDA temperature, whereas it increased with increasing pH value. The TbY_xO_y EIS sensor is suitable for use in electrochemical sensors and biosensors.

Acknowledgements

The authors thank the Chang Gung Memorial Hospital and the National Science Council of the Republic of China, Taiwan, for financially supporting this research under contracts of CMRPD2C0151, CMRPD2C0152, CMRPD2C0153, and NSC 102-2221-E-182-072-MY3.

References

1. P. Bergveld, IEEE Trans. Biomed. Eng., 1970, 17, 70.
2. P. Bergveld, Sens. Actuators, B, 2003, 88, 1.
3. A. Poghossian and M. J. Schoning, Electroanalysis, 2004, 16, 1863.
4. M. J. Schoning and A. Poghossian, Analyst, 2002, 127, 1137.
5. L. Bousse, H. H. van der Vlekkert and N. F. de Rooij, Sens. Actuators, B, 1990, 2, 103.
6. D. Sobczynska and W. Torbicz, Sens. Actuators, 1984, 6, 93.
7. D. H. Kwon, B. W. Cho, C. S. Kimo and B. K. Sohn, Sens. Actuators, B, 1996, 34, 441.
8. S. M. Castanho, R. Moreno and J. L. G. Fierro, J. Mater. Sci., 1997, 32, 157.
9. P. Gimmel, B. Gompf and W. Gopel, Sens. Actuators, 1989, 17, 195.
10. G. D. Wilk, R. M. Wallace and J. M. Anthony, J. Appl. Phys., 2001, 89, 5243.
11. E. Gusev, Defects in High-k Gate Dielectrics Stacks, Springer, Dordrecht, 2006.
12. High-k Gate Dielectrics for CMOS Technology, ed. G. He and Z. Sun, Wiley-VCH Verlag GmbH & Co, Weinheim, 2012.
13. T. M. Pan, J. D. Lee and W. W. Yeh, J. Appl. Phys., 2007, 101, 024110.
14. T. M. Pan and C. C. Huang, Electrochem. Solid-State Lett., 2008, 11, G62.
15. S. Kitai, O. Maida, T. Kanashima and M. Okuyama, J. Appl. Phys., 2003, 42, 247.
16. Y. Zhao, K. Kita, K. Kyuno and A. Toriumi, Appl. Phys. Lett., 2006, 89, 252905.
17. Y. Zhao, M. Toyama, K. Kita, K. Kyuno and A. Toriumi, Appl. Phys. Lett., 2006, 88, 072904.
18. T. Schroeder, G. Lupina, J. Dabrowski, A. Mane, C. Wenger, G. Lippert and H. J. Mussig, Appl. Phys. Lett., 2005, 87, 022902.
19. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Wiley India Edition, 2006.
20. C. A. Betty, R. Lal and J. V. Yakhmi, Electrochim. Acta, 2009, 54, 3781.
21. L. Bousse, N. F. De Rooij and P. Bergveld, IEEE Trans. Electron Devices, 1983, 30, 1263.
22. M. W. Shinwari, M. J. Deen and D. Landheer, Microelectron. Reliab., 2007, 47, 2025.
23. T. M. Pan and C. W. Lin, J. Phys. Chem. C, 2010, 114, 17914.
24. A. S. Poghossian, Sens. Actuators, B, 1992, 7, 367.
25. C. H. Kao, H. Chen, M. L. Lee, C. C. Liu, H. Y. Ueng, Y. C. Chu, C. B. Chen and K. M. Chang, Sens. Actuators, B, 2014, 194, 503.
26. G. M. da Silva, S. G. Lemos, L. A. Pocrifka, P. D. Marreto, A. V. Rosario and E. C. Pereira, Anal. Chim. Acta, 2008, 616, 36.
27. T. Matsuo, M. Esashi and H. Abe, IEEE Trans. Electron Devices, 1979, 26, 1856.
28. M. J. Schoning, D. Brinkmann, D. Rolka, C. Demuth and A. Poghossian, Sens. Actuators, B, 2005, 111–112, 423.
29. K. M. Chang, C. T. Chang, K. Y. Chao and C. H. Lin, Sensors, 2010, 10, 4643.
30. L. B. Chang, H. H. Ko, M. J. Jeng, Y. L. Lee, C. S. Lai and C. Y. Wang, J. Electrochem. Soc., 2006, 153, G330.
31. T. M. Pan and K. M. Liao, Sens. Actuators, B, 2007, 128, 245.
32. T. M. Pan, C. W. Wang, Y. S. Huang, W. H. Weng and S. T. Pang, J. Electrochem. Soc., 2015, 162, B83.
33. E. Katz and I. Willner, Electroanalysis, 2003, 15, 913.
34. E. H. Nicollian and J. R. Brews, MOS (Metal Oxide Semiconductor) Physics and Technology, Wiley, Hoboken, 2003.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra11377j

---