A low-cost portable electrical sensor for hydroxyl ions based on amorphous InGaZnO4 thin film at room temperature

Sensors and Actuators B 239 (2017) 679–687

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb

A low-cost portable electrical sensor for hydroxyl ions based on amorphous InGaZnO4 thin film at room temperature Dali Sun a,∗,1 , Hiroaki Matsui a,b , Hiroyasu Yamahara b , Chang Liu c , Lei Wu d , Hitoshi Tabata a,b a

Department of Bioengineering, The University of Tokyo, 1-3-7 Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan Department of Electrical Engineering and Information Systems, The University of Tokyo, 1-3-7 Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan c Department of Nanomedicine, Houston Methodist Research Institute, 6670 Bertner Avenue, Houston, TX 77030, United States d Department of Epidemiology, Institute of Geriatrics, Chinese PLA General Hospital, China b

a r t i c l e

i n f o

Article history: Received 27 May 2016 Received in revised form 1 August 2016 Accepted 10 August 2016 Available online 12 August 2016 Keywords: Amorphous InGaZnO4 OH− sensor Impedance sensing Resistivity

a b s t r a c t The measurement and control of hydroxide ion (OH− ) concentration in solution are essential in industrial processes. However, no portable sensing method directly targeting OH− ion with low-cost has been reported till date. Herein, we demonstrate an electrical detection method for OH− concentration in solution based on impedance spectroscopy of hydroxyl ions (OH− ) attached to amorphous InGaZnO4 (aIGZO) film surfaces. The systematic examination of impedance response reveals that the resistance component of impedance is sensitive to the OH− ions interaction with the film surface. Results of X-ray photoemission spectroscopy confirm that the change of the impedance property is directly attributed to the amount of hydroxyl radical on the film surface originated from OH− ions in the solution. The impedance behavior of the film upon interaction with OH− was reasonably described by the theoretical analysis of optical measurements based on a vacancy-dependent model. Developed by applying this mechanism as a reference application, an easy-to-use aIGZO thin film based resistance OH− sensor at room temperature shows superior sensitivity, reproducibility, and linearity in the alkali range. This study extends the understanding and usage of aIGZO thin film regarding surface-sensing for the detection of surface interaction and process involving chemical ions and species. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Quantitative assessment and control of hydroxide ion (OH− ) concentration in solution are crucial in most chemical processes and reactions. To date, the detection of OH− ions has not received much attention since most applications have focused on hydrogen (H+ ) ions as the counterpart. The concentration of H+ ions in solution can be evaluated by pH value [1,2], and has been applied to gross evaluations [2–5] and electrical pH sensing [1,6,7]. Nevertheless, these methods involving optical and electrical pH sensing suffer from the alkaline error, which limits their applications for OH− sensing [8]. Some reported optical pH sensing for high alkaline solution, but they were by no mean of portable [9,10]. Meanwhile, electrical pH meters commercially available for concentrated alka-

Abbreviations: aIGZO, amorphous indium gallium zinc oxide; XPS, X-ray photoelectron spectroscopy. ∗ Corresponding author. E-mail address: [email protected] (D. Sun). 1 Present address: Department of Nanomedicine, Houston Methodist Research Institute, 6670 Bertner Avenue, Houston, TX 77030, United States. http://dx.doi.org/10.1016/j.snb.2016.08.060 0925-4005/© 2016 Elsevier B.V. All rights reserved.

line solution requires costly electrode. A low-cost portable sensing scheme targeting at OH− become desirable specifically for alkalinity measurement at high concentration. Although several gas-phase sensing methods of hydroxyl ions have been reported [11,12], surface-sensing methods targeting at hydroxyl ions in solution have not been advanced. Transparent oxide semiconductors (TOSs) have been applied for electrochemical and nano-optical surface-sensing [13,14]. Among amorphous transparent oxide semiconductors, amorphous InGaZnO (aIGZO), a typical inorganic metal oxide, has attracted considerable attention due to its superior electrical properties and its environmental/thermal stability [15,16]. In particular, aIGZO can be widely tunable in terms of controlling electron carriers following the incorporation of Ga ions to suppress excessive carrier generation via oxygen vacancies [17,18]. We reported in our previous studies, that aIGZO thin film has not only a superior but also tunable affinity for hydroxyl species on the surface [19,20]. K. Takechi et al. and other groups has employed aIGZO for pH sensing [21–24], nevertheless, the focus was drawn on the low pH range (≤8). Inspired by these studies and aforementioned physical and chemical prop-

680

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

erties, in this work, aIGZO films were chosen as a host material for the electrical detection of hydroxyl ions. Intensively employed in sensing research [25–28], impedance spectroscopy, a high sensitivity tool for surface reaction, was employed as an indicator of the electrical activity of aIGZO films in this work. The electrical activity on aIGZO films was examined before and after OH− treatments. The hydroxyl originated from the solution attached to the aIGZO film surface caused impedance changes. Above all, we demonstrate electrical detections of OH− ions based on the impedance and resistance behavior of OH− ions adsorbed onto aIGZO film surfaces. As an application reference, a ready-to-use electrical sensor for OH− in solution at room temperature was disclosed showing superior sensitivity, reproducibility, and linearity in the alkali range.

centrations (0, 50, 100, 120, 140, 160, 180, and 200 mM) for various durations (10, 20,30,40,50, and 70 min), and then air dried for measurement after 3 times DI water washing. Impedance measurements were conducted before and after NaOH treatment. All impedance equal circuit models were fitted by ZView software (Scribner Associates Inc., USA), with a limited range from 104 Hz to 106 Hz. Resistance measurements were conducted using a digitalelectrical multi-meter (Sanwa DA32). Gold electrode patterned film samples were treated with NaOH solutions at 100, 120, 140, 160,180, and 200 mM. The measurements were also conducted before and after NaOH treatment, and sensor responses were then calculated according to the equation defined. 3. Result and discussion

2. Experimental section 2.1. Film fabrication and surface-chemical pretreatment The aIGZO films with a thickness of 70 nm were deposited on flat glass substrate by direct-current (DC) magnetron sputtering at room temperature (RT) [29]. A stoichiometric InGaZnO4 ceramic was chosen as the target material. Film growth was performed at room temperature (RT) with an Ar gas flow of 200 cc. Interdigitated array gold electrodes were deposited by the dc sputtering method for 5 min (45 nm). The films on glass substrates were sliced into small pieces (16 × 22 mm), and sonicated in deionized (DI) water for 15 min before drying by an air blower to clean the surface. Film samples were then treated chemically using NaOH solutions at different concentrations and varied treatment time. After rinsing three times with deionized (DI) water and air-dry, electrical measurements were conducted both before and after the OH− treatment. The oxygen-related species attached to the film surfaces was examined using O (1s) core-level spectra derived from X-ray photoelectron spectroscopy (XPS) on an instrument equipped with a monochromatized Al K␣ source in an ultra-high vacuum chamber above 10−10 Torr (JEOL JPS-9010MC). XPS spectrum was calibrated by utilizing the Au 4f5/2 peak at a photon energy of 87.5 eV. Samples treated with 50, 100, 150, and 200 mM NaOH solution for 30 min were mounted in line on the XPS stage with a standard gold film slice in the end for calibration. For each sample, high-resolution spectra were collected 10 times for averaging. XPS scan range was set to from 522 to 534 eV to target at the OH− (∼529 eV) peak. XPS peak curve fitting (Multi-peaks Gaussian fitting) and area under peak calculations were carried out using Origin Lab 8.6 software. Optical responses of the film samples were examined to investigate optical structures in the vicinity of band edges using a UV–vis spectrometer (JASCO V-670). The absorption spectra were recorded over a range from 200 to 1000 nm. The clean glass substrate was scanned first as the baseline for all samples. Film samples treated with 0, 100, 150, 200 and 250 mM NaOH solution for 30 min were then characterized sequentially for their absorption spectra. Sodium hydroxide (NaOH, +96%), and other chemicals were obtained from Wako Pure Chemical (Tokyo, Japan). Nanopure water (>18.0M) purified by the Millipore Milli-Q was used in all experiments. 2.2. Electrical measurement The impedance spectra in the frequency domains from 104 to Hz were measured using an impedance analyzer under a zero bias with voltage oscillation amplitude of 50 mV (Solartron SI 1260) at room temperature. The electrodes comprised two-terminal configurations with interdigitated arrays. NaOH treatment: the samples were then treated with NaOH solutions at different con106

3.1. Impedance OH− sensor Impedance signal shift was evaluated in this work to indicate the hydroxyl group attached to the surface, which is eventually related to the OH− concentration in the solution. The overall procedure of the sensing detection is schematically outlined in Fig. 1. Patterned with thin film gold electrode by dc sputtering, the aIGZO thin films were measured by impedance spectroscopy for the first time. The films were then dipped into the target OH− solutions for various durations. After the OH− interaction with the surface, impedance was measured again. The impedance variance before and after OH− interaction are closely related to the quantity of hydroxyl species on the film surface, and indicates OH− concentration in the solution. Impedance obeys Ohm’s law, which can be described as follows: Z = Z  + jZ"e;

(1)

where Z’ and Z” indicate real and imaginary parts, corresponding to resistance and inductance, respectively. Both terms are referred in defining the magnitude of impedance: |Z| =

 2

Z  + Z  2

2

(2)

In an effort to elucidate the correlation between impedance shift and OH− interaction with the surface, we defined the change in impedance as follows:











Z = Z OH − Z 0 = Z OH − Z 0 + j Z "e;e; − Z "e;e; OH 0



(3)

where Z0 and ZOH represent the impedance before and after OH− interaction with the surface, respectively. Fig. 2 exemplifies the processing of impedance response. The sample was treated by immersed into 0.2 M NaOH solution for 30 min with impedance measured before and after OH− treatment. Distinctive impedance shift upon OH− interaction can be observed in Fig. 2(a) and (b) in the form of bode plot showing that both Z’ and |Z| components were subject to a clear shift from 70 to 150  in the Z’ component upon OH− treatment. A Nyquist plot is also depicted in Fig. 2(c) to clarify the origin of the calculated Z result shown in Fig. 2(d) according to Eq. (3). Following the same procedure, systematic examinations on the impedance changes of all samples were conducted under different OH− treatment conditions. Fig. 3 summarizes the impedance shifts (Z) at various OH− concentrations. The dipping time for all films was fixed for 30 min in order to evaluate the correlation between the impedance change and NaOH concentration of the solution. To illustrate frequency response of the impedance shift, Fig. 3(a) and (b) are presented in bode plot. Noticeable variance in the imaginary parts (Z”) upon OH treatment is only observed in high frequencies from 105 to 106 Hz while there is negligible or no change at lower frequencies. On the other hand, changes in the real parts (Z’) following the treatments is observed in all frequency ranges. Changes in Z’

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

681

Fig. 1. Schema of the impedance based OH− sensor. The inset on the left top is the image of the sensor chip with aIGZO film and gold electrode. The sample aIGZO films were treated sequentially: (1) electrode deposition, (2) impedance measurement before OH− treatment, (3) dipping into NaOH solution, (4) impedance measurement after OH− treatment. The impedance was compared between the first and second measurements with the variance indicating OH− concentration in the target solution.

are more significant compared to Z”. Treatment of film surface with OH− ions strongly influences the real part of the impedance which can be also endorsed by the impedance magnitudes |Z| shown in Fig. 3(c). Moreover, Fig. 3(d) demonstrates the OH− concentration dependency of the real parts (Z’) more explicit in Nyquist plot. Low concentration treatments (below 100 mM) exhibit minute shifts shown in Fig. 3(a)–(d) indicating the limitation of the impedance analyzer. Nevertheless, above the limitation (OH− concentration more than 100 mM), the imaginary, and real part is systematically detectable, and the change in impedance magnitude is well correlated with OH− concentration (Fig. 3(e)). This result becomes the origin for using impedance response on aIGZO thin film to evaluate OH− concentration in solution. It is important to notice that the impedance shift of the imaginary parts with increasing OH− concentration is positive, which suggests the impedance response may be decomposed in term of an R − L model (R: resistance and L: inductance). This model has been applied to impedance-based surface sensing in thin films based on oxide semiconductors [30–35], in which the impedance comprises bulk and surface components. Before OH− interaction with the surface, bulk impedance dominates. After OH− interaction, surface impedance contributes most to the impedance shift. Hence, the impedance shift (Z) can be raised by surface impedance changes upon OH− interaction following the equation [31]: Z = RS + j2␲fLS

(4)

where Rs, and Ls represent the resistance and inductance, respectively, and f indicates the frequency. This model could describe the remarkable changes in impedance response with increasing OH− ion concentration. As mentioned, it is clear that surface modifications of film samples were reflected in the changes of the real rather than imaginary part of the impedance. According to Eqs. (2) and (4), Z ought to increase with rising frequency. But the frequency dependence of |Z| was insignificant as shown in Fig. 3(c) and (e), suggesting the second term related to the Ls component in Eq. (4) is small. Theoretical fittings to Eq. (4) shows that the component of Rs is much larger than that of LS as shown in Fig. 3(f) (see Supporting information). It suggests that the second term of 2 fL is almost constant and independent of the OH− ion concentration. In other words, the value of Z is mainly determined by the Rs component rather than the Ls component. The real part related to Rs plays an important role in determining the impedance response, implies that a simple resistor model is efficient for analysis, which is employed for analysis later in this work.

3.2. Dipping time The impedance responses accompanying OH− interaction with the aIGZO film are related to the interface interactions between the film surfaces and hydroxyl ions. As a key factor for interaction dynamics, the reaction time was also evaluated. Systematic

682

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

Fig. 2. Impedance data processing example: (a) Impedance spectra of Z’ and (b) |Z| in the form of bode plot before(Z0 ) and after OH− interaction with the surface (ZOH ) treated with 200 mM NaOH solution for 30 min. (c) A Nyquist plot (Z’ towards Z”) showing the impedance shift of the film sample upon OH− interaction with the surface. (d) Impedance variance (Z) calculated from (c) according to Eq. (3).

impedance measurements at various OH− concentrations were conducted with different dipping duration to examine the dipping time impact on impedance shift. The impedance shift presented as Nyquist plot in Fig. 4, demonstrates significant dependency on dipping time as well as OH− concentration. At the lower concentration (100 mM), the impedance shifts in Z’ and Z” is trivial under different dipping durations suggesting less OH attaching on the surface, which may also because of the detection limitation mentioned in the previous section. However, from 120 mM to 180 mM, the impedance shift shows apparent relevancy to dipping durations. It is also noticeable that the relevancy is not as simple as a linear correlation. The impedance shift increment upon dipping duration grows exponentially along with the OH− concentration rise in solution. This behavior is associated with the surface coverage extent of OH on the film surface, and is further validated in the following sections. 3.3. Surface properties o the film surface From the aforementioned results, the impedance changes dramatically on the aIGZO thin film related to OH− interaction. Its dependency to both OH− concentration and dipping duration implies the occurrence of surface reaction at the solid-liquid interface. The impedance shift is related to the amount of OH− ions as

a surface reactant. To use the impedance as an indicator for OH− concentration in solution, it is indispensable to elucidate the relationship between the surface hydroxyl radical (OH• ) quantity on the film and the OH− concentration in solution for treatment. In order to estimate the hydroxyl radical amount on the film surface before and after the surface OH− treatment, we employed XPS area analysis of OH- and O-related signals in O(1s) core-level spectra, as estimated by theoretical-fitting to the Gaussian function [36–38]. The relation between OH• on the film surface ([OH• ]surface ) and the bulk solution can be expressed with the help of Boltzmann factor [27]: •

[OH ]surface



= OH −

 solution

 exp

−e D kB T

 (5)

where [OH− ]solution is the OH− concentration in the solution. D and T are the surface potential and temperature, respectively; e is the elementary charge (1.60218 × 10−19 C); kB is the Boltzmann constant (1.38066 × 10−23 J/K). This relation is based on the mechanism that ion interaction with solid surface is related to the electric potential at liquid-solid interfaces, which obeys the Boltzmann statistics. Fig. 5 shows the correlation between the OH-related XPS peak area (A) and NaOH concentration in solution. The experimental data can be fitted with a linear relation to ∼exp(-e D /kB T) (a dot

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

683

Fig. 3. Impedance shifts (Z) under various OH− concentrations. (a) Bode plot of the imaginary part (Z”) towards frequency. (b) Bode plot of the real part (Z’) towards frequency. (c) Bode plot of the impedance magnitude |Z| towards frequency. (d) Nyquist plot of the impedance shift (Z) (e) Impedance magnitude shift (|Z|) at specified frequencies (2, 4, 6, 8, and10 × 105 Hz) for films treated with different NaOH concentrations. (f) Parameter-fitting results to an R-L series equivalent circuit.

line in Fig. 5), which shows that the surface OH• quantity is linearly related to the NaOH solution concentration and can be used as an indicator for OH− concentration in the solution. Discussed in the previous section, the real part of the impedance has shown its deterministic to the impedance shift upon OH− interaction. The impedance shifts can be simplified to the resistance variance (R) using the following resistor model: R =

R OH R0

(6)

where R0 and ROH indicate the resistance before and after OH− interaction with the surface, respectively. With constant treatment duration (30 min), the resistance variance correlates with OH− concentrations as plotted in Fig. 6(a). Besides the concentration, Fig. 6(b) reveals significant dependency of the resistance variance on dipping duration. Regarding oxide semiconductor, several models have been proposed for resistance-based sensors [32,39–42], and have been applied to semiconductor gas sensors [31,43,44]. We propose that

684

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

a vacancy-dependent model is suitable in accounting for explaining the electrical detection of OH− ions attached to the aIGZO film surface. The first reason is that it has been reported that aIGZO films deposited at room temperature have large amounts of defects in the host contributing to the generation of electron carriers [45], which may be depressed by thermal annealing at high temperatures [46]. Furthermore, our previous study revealed that the thermal annealing remarkably affected hydroxyl species affinity on the aIGZO thin film surface [20] since vacancies related to cations/anions in the host leads to an increase in electron carriers in the film [19]. Accordingly, a vacancy-dependent model is suitable accounting for the relationship between OH− interaction with the surface and changes in resistance of the film surface: k[OH − ]

v ..o → Ox

Fig. 4. Dipping duration effect on impedance variation at various OH− concentrations in solution (100, 120, 140, 160, and 180 mM) in Nyquist plot. The dipping durations are 10, 15, 20, 30, 35, 40, 50, 60, and 70 min.

(7)

where v..o denotes the surface oxygen vacancy density, k the rate constant, Ox the natural lattice, and the vacancy is formulated as a first order process: v ..o = v 0 e−ktc

(8)

OH-related XPS Peak Area (a.u.)

OH−

concentrawhere v0 the initial vacancy density, c the surface tion. The change in vacancy following a surface reaction with OH− elicits modulation of the carrier content. Assuming oxygen vacancies are primary contributors in the film, the carrier content (n) can be denoted as follows:

Expt. Fit

2000

n ∝ 2v ..o 1800

(9)

The resistance R is related to the carrier concentration as: u u R= (10) ∝ n 2v ..o where the proportionality constant (u) contains both the mobility and geometrical factors. The resistance variance (R) upon OH− interaction with the surface, referring carrier concentration change, can be expressed by:

1600

0.05

0.10

0.15

0.20

[OH-] (M) Fig. 5. Correlation between OH-related XPS peak area (A) on the film surfaces and the NaOH concentrations in the solutions. The OH-related XPS peak area (A) was fitted with a linear relation to ∼ exp(-e D /kB T) defined in Eq. (5).

R =

R OH n0 = ∝ ektc R0 n

(11)

The above deduction suggests that the resistance variance is exponentially related to OH− concentration in solution and dipping time. With time t assigned to a constant value of 30 min, best parameterfit of the resistance variance R can be found to Eq. (11) in Fig. 6(a)

Fig. 6. Resistance variance’s (R) correlation to OH− concentration and dipping time. Dash line is the model fitting to Eq. (11). (a) Correlation to NaOH concentrations (R2 = 0.98198) with a constant dipping duration (30 min). (b) Correlation to dipping duration and NaOH concentrations (R2 = 0.9956, 0.98823, 0.96347, and 0.97855 for 180 mM, 160 mM, 140 mM, and 120 mM, respectively).

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

685

a

c

b

Fig. 7. (a) Optical transmission spectra of the aIGZO films with OH− treatment at 0, 100,150,200,250 mM. (b) Tauc plot for transmission spectra of the aIGZO films with OH− treatments. (c) The optical band gap (Eg) and carrier concentration (n) as a function of NaOH calculated by Eq. (13) and Eq. (12), respectively.

and (b) plots the fitting results at various reaction duration with different NaOH concentrations. The experimental data agrees well with the theoretical evaluations which signify that the vacancy model is efficient to explain the resistance change upon OH− interaction with the surface. To characterize changes in electron concentration of the films, optical transmission measurements were conducted. According to Burstein–Moss (BM) model [47,48], the band gap widening (EgBM ) for an n-type semiconductor with parabolic band is related to carrier concentration: E BM = E g − E 0g = g

h2  2 2/3 3 n 2m∗

(12)

where Eg0 , Eg , m∗ , h are the intrinsic band gap, the optical band gap, effective electron mass and plank constant, respectively. Accordingly, by evaluation optical band gap, it is possible to estimate carrier reduction introduced by OH− treatment. We extracted the optical energy band gap (Eg , also called Tauc gap) by applying the Tauc method [49]: 2



(␣hv) = B hv − Eg





(13)



where ˛ = −ln T/d , T the measured transmission (%), d the film thickness (m), hv the photon energy and B a constant slope parameter. Fig. 7(a) is the optical transmission spectrum of the aIGZO thin films treated with OH− solutions at various concentrations. By parameter-fitting to Eq. (13), Tauc plot as in Fig. 7(b) highlights the optical band gap change upon OH− treatment. At low concentration of OH− (from 0 to 100 mM), the band gap changes little suggesting the concentration limitation of effective OH− treatment. Whereas, at high concentrations (from 200 to 250 mM), the decreasing band gap reduction implies possible saturation effect for OH− treatment on aIGZO thin film. Carrier concentrations of the aIGZO films under different OH− concentrations were estimated according to Eq. (12) with the following parameters: E 0g , 3.4 eV [15]; m∗ , 0.32me (me is the mass of the free electron) [50]. The carrier concentration trend plotted in Fig. 7(c) demonstrates that elevating OH− concentration of the OH− treatment results in the reduction of the carrier concentration of the aIGZO film. This result is consistent with the oxygen vacancy-dependent model we proposed. 3.4. Resistance response of OH− interaction As mentioned above, we conducted impedance measurements of hydroxyl species attached to aIGZO film surfaces, which were evaluated using a theoretical approach. We found that changes in resistance component of the equivalent circuit model played an

Fig. 8. Resistance electrical response (ER) of a common multi-tester (blue dot) on aIGZO thin films treated by NaOH under various OH− concentrations (100,120,140,160,180,200 mM). The average standard error of the resistance electrical response (ER) is about 0.00366, and is too small to be displayed. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

important role in estimating the concentration of hydroxyl species attached to the film surfaces. Thereupon, measurement of the resistance response using an electrical multi-meter was performed to detect the presence of hydroxyl species on the film surfaces as an easy application of the OH− sensor. To simply the sensor response, regarding Eq. (11), the output electrical response (ER) is defined using the constant A0 as ER = log (R) = A0 ktc

(14) OH−

ER is linearly related to the concentration at a constant treatment duration t. Fig. 8 shows the dependency of ER on the ion OH− concentration. Herein, we defined the detection sensitivity (S) of ER using the following equation: S=

dlog (R)



d OH −



(15)

The detection sensitivity of ER is S = 0.58 per mM, which was correlated in a high linear manner with a mean-square error of 0.942 in the range of 100–200 mM. The minimum detectable concentration variance for the resistance sensor is 1.5 mM. There is no other portable pH or OH− sensor can ever reach this range to date with comparable cost. Electrical pH meters commercially available for concentrated alkaline solution requires costly electrode. The absorption of other species to the electrode like sodium ions (sodium error) may interfere the pH sensor output because the whole measurement was performed in solution [51]. After OH−

686

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687

treatment, the OH− sensor proposed in this work, was washed intensively with DI water to remove interfering ions, and then airdried for measurement. Therefore, only the target reacted hydroxyl species remaining on the film surface caused the output response variance without sodium error. In addition, the stability of the sensor under various humidity environment at 21 ◦ C was also tested. The response was slightly affected by humidity, but very limited to 0.0002 per humidity increment (see Supporting information). It can be considered very stable in humidity environment. However, it is still recommended to keep the sensor chip in a dark vacuum container (<10−4 Pa) at room temperature for long time storage. The reproducibility was evaluated by calculate coefficient of variation (CV) with 5 repeats (different electrodes) for each OH− concentration. The maximum CV was 0.93%, and the average CV was 0.37%, which can be consider as highly reproducible. These results indicate that hydroxyl ions attached to aIGZO film surfaces are detected well by electrical resistance, and confirm the impedance analysis mentioned ahead. The results also soundly nominate an easy-to-use and portable sensor to evaluate OH− concentration in solution with high sensitivity, linearity and stability while extending the usage of aIGZO thin film. Because the OH− interaction with the film surface is irreversible, it is difficult to restore the film surface for reusability. However, since the sensor response depends on relative increment, it is still possible to reuse the electrode chip again after measuring lower OH− concentration for higher OH− concentration. Besides, the cost can be compensated by minimizing the sensor chip size. 4. Conclusions This paper demonstrated the utilization of aIGZO films for the surface sensing of hydroxyl species using electrical detection. Fabricated as a conducting film, it was found in a systematic examination of different dipping times and OH− concentration in the solutions that the impedance of the aIGZO thin film increased accordance with the OH− concentration while treated by the alkali solution. The observed behavior could be reasonably accounted for by utilizing a vacancy-dependent model, which was also supported by optical measurements. It was clarified that the change in impedance response of the film was mainly determined by the surface resistance, and was confirmed using simple electrical measurements. Correspondingly, a ready-to-use, stable resistance based OH− sensor using aIGZO thin film at room temperature was also presented showing superior sensitivity and linearity in the alkali region. These results point out that the hydroxyl species attached to the aIGZO film surfaces can be detected well by measuring electrical resistance, which provides another insight toward the development of surface sensing devices on aIGZO films. Acknowledgments Thanks Dr. Ryosho Nakane, Dr. Akira Isogai, Dr. Tuguyuki Saito, for their invaluable advice throughout the project. This research was supported in part by a grant-in-aid from the JSPS Core-to-Core Program. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.snb.2016.08.060. References [1] N.F. Sheppard Jr., M.J. Lesho, P. McNally, A. Shaun Francomacaro, Microfabricated conductimetric pH sensor, Sens. Actuators B Chem. 28 (1995) 95–102, http://dx.doi.org/10.1016/0925-4005(94)01542-P.

[2] Z. Zhujun, W.R. Seitz, A fluorescence sensor for quantifying pH in the range from 6.5 to 8.5, Anal. Chim. Acta 160 (1984) 47–55, http://dx.doi.org/10.1016/ S0003-2670(00)84507-9. [3] H.R. Kermis, Y. Kostov, G. Rao, Rapid method for the preparation of a robust optical pH sensor, Analyst 128 (2003) 1181–1186, http://dx.doi.org/10.1039/ b304902g. [4] S.H. Lee, J. Kumar, S.K. Tripathy, Thin film optical sensors employing polyelectrolyte assembly, Langmuir 16 (2000) 10482–10489, http://dx.doi. org/10.1021/la0011836. [5] E. Pringsheim, E. Terpetschnig, O.S. Wolfbeis, Optical sensing of pH using thin films of substituted polyanilines, Anal. Chim. Acta 357 (1997) 247–252, http://dx.doi.org/10.1016/S0003-2670(97)00563-1. [6] K.G. Kreider, M.J. Tarlov, J.P. Cline, Sputtered thin-film pH electrodes of platinum, palladium, ruthenium, and iridium oxides, Sens. Actuators B. Chem. 28 (1995) 167–172, http://dx.doi.org/10.1016/0925-4005(95)01655-4. [7] X. Zhang, B. Ogorevc, J. Wang, Solid-state pH nanoelectrode based on polyaniline thin film electrodeposited onto ion-beam etched carbon fiber, Anal. Chim. Acta 452 (2002) 1–10, http://dx.doi.org/10.1016/S00032670(01)01435-0. [8] D.J. Graham, B. Jaselskis, C.E. Moore, Development of the glass electrode and the pH response, J. Chem. Educ. 90 (2013) 345–351, http://dx.doi.org/10. 1021/ed300246x. [9] A. Safavi, M. Bagheri, Novel optical pH sensor for high and low pH values, Sens. Actuators B Chem. 90 (2003) 143–150, http://dx.doi.org/10.1016/S09254005(03)00039-X. [10] A. Safavi, H. Abdollahi, Optical sensor for high pH values, Anal. Chim. Acta 367 (1998) 167–173, http://dx.doi.org/10.1016/S0003-2670(98)00079-8. [11] Y. Wang, Mesostructured tin oxide as sensitive material for C2H5OH sensor, Talanta 57 (2002) 875–882, http://dx.doi.org/10.1016/S0039-9140(02)001248. [12] K.-M. Kim, H.-R. Kim, K.-I. Choi, H.-J. Kim, J.-H. Lee, ZnO hierarchical nanostructures grown at room temperature and their C2H5OH sensor applications, Sens. Actuators B Chem. 155 (2011) 745–751, http://dx.doi.org/ 10.1016/j.snb.2011.01.040. [13] U. Yogeswaran, S. Chen, A review on the electrochemical sensors and biosensors composed of nanowires as sensing material, Sensors 8 (2008) 290–313, http://dx.doi.org/10.3390/s8010290. [14] H. Matsui, W. Badalawa, A. Ikehata, H. Tabata, Oxide surface plasmon resonance for a new sensing platform in the near-Infrared range, Adv. Opt. Mater. 1 (2013) 397–403, http://dx.doi.org/10.1002/adom.201200075. [15] K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, H. Hosono, Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors, Nature 432 (2004) 488–492, http:// dx.doi.org/10.1038/nature03090. [16] T. Kamiya, K. Nomura, H. Hosono, Origins of high mobility and low operation voltage of amorphous oxide TFTs: electronic structure, electron transport, defects and doping, J. Disp. Technol. 5 (2009) 468–483, http://dx.doi.org/10. 1109/JDT.2009.2034559. [17] M. Ito, M. Kon, C. Miyazaki, N. Ikeda, M. Ishizaki, R. Matsubara, et al., Amorphous oxide TFT and their applications in electrophoretic displays, Phys. Status Solidi 205 (2008) 1885–1894, http://dx.doi.org/10.1002/pssa. 200778910. [18] M. Ito, C. Miyazaki, M. Ishizaki, M. Kon, N. Ikeda, T. Okubo, et al., Application of amorphous oxide TFT to electrophoretic display, J. Non Cryst. Solids 354 (2008) 2777–2782, http://dx.doi.org/10.1016/j.jnoncrysol.2007.10.083. [19] D. Sun, H. Matsui, C.-N. Wu, H. Tabata, Surface treatment on amorphous InGaZnO4 thin film for single-stranded DNA biosensing, Appl. Surf. Sci. 324 (2015) 310–318, http://dx.doi.org/10.1016/j.apsusc.2014.10.115. [20] D. Sun, H. Yamahara, R. Nakane, H. Matsui, H. Tabata, Hydroxyl radical and thermal annealing on amorphous InGaZnO4 films for DNA immobilizations, Colloids Surf. B. Biointerfaces 130 (2015) 119–125, http://dx.doi.org/10.1016/ j.colsurfb.2015.04.024. [21] K. Takechi, S. Iwamatsu, T. Yahagi, Y. Abe, S. Kobayashi, H. Tanabe, Bottom-gate amorphous InGaZnO4 thin-film transistor pH sensors utilizing top-gate effects, Jpn. J. Appl. Phys. 53 (2014), http://dx.doi.org/10.7567/JJAP. 53.076702. [22] J.-C. Lin, B.-R. Huang, Y.-K. Yang, {IGZO} nanoparticle-modified silicon nanowires as extended-gate field-effect transistor pH sensors, Sens. Actuators B Chem. 184 (2013) 27–32, http://dx.doi.org/10.1016/j.snb.2013.04.060. [23] K. Takechi, S. Iwamatsu, S. Konno, T. Yahagi, Y. Abe, M. Katoh, et al. Gate-to-source voltage response in high-sensitivity amorphous InGaZnO4 thin-film transistor pH sensors, 078004 (n.d.) 7–10. [24] K. Takechi, S. Iwamatsu, T. Yahagi, Y. Abe, S. Kobayashi, H. Tanabe, Sensitivity evaluation on amorphous InGaZnO4 thin-film transistor pH sensors having various capacitances of ion-sensitive and bottom-gate insulators, ECS J. Solid State Sci. Technol. 3 (2014) Q3076–Q3080, http://dx.doi.org/10.1149/2. 014409jss. [25] J. a Lee, S. Hwang, J. Kwak, S. Il Park, S.S. Lee, K.-C. Lee, An electrochemical impedance biosensor with aptamer-modified pyrolyzed carbon electrode for label-free protein detection, Sens. Actuators B Chem. 129 (2008) 372–379, http://dx.doi.org/10.1016/j.snb.2007.08.034. [26] M.W. Shinwari, D. Zhitomirsky, I. a. Deen, P.R. Selvaganapathy, M.J. Deen, D. Landheer, Microfabricated reference electrodes and their biosensing applications, Sensors 10 (2010) 1679–1715, http://dx.doi.org/10.3390/ s100301679.

D. Sun et al. / Sensors and Actuators B 239 (2017) 679–687 [27] T.C. Hang, A. Guiseppi-elie, T. Christopher, A. Guiseppi-elie, T.C. Hang, A. Guiseppi-elie, Frequency dependent and surface characterization of DNA immobilization and hybridization, Biosens. Bioelectron. 19 (2004) 1537–1548, http://dx.doi.org/10.1016/j.bios.2003.12.014. [28] C. Tronstad, Developments in Biomedical Sensors Based on Electrical Impedance Dr. Thesis., 2012 https://www.duo.uio.no/handle/10852/34384. [29] J. Kang, K.-J. Moon, T. Il Lee, W. Lee, J.-M. Myoung, Migration of indium ions in amorphous indium–gallium–zinc-oxide thin film transistors, Appl. Surf. Sci. 258 (2012) 3509–3512, http://dx.doi.org/10.1016/j.apsusc.2011.11.104. [30] S. Agarwal, G.L. Sharma, Humidity sensing properties of (Ba, Sr) TiO3 thin films grown by hydrothermal–electrochemical method, Sens. Actuators B Chem. 85 (2002) 205–211, http://dx.doi.org/10.1016/S0925-4005(02)00109-0. [31] A. Labidi, C. Jacolin, M. Bendahan, A. Abdelghani, J. Guérin, K. Aguir, et al., Impedance spectroscopy on WO3 gas sensor, Sens. Actuators B Chem. 106 (2005) 713–718, http://dx.doi.org/10.1016/j.snb.2004.09.022. [32] A. Labidi, E. Gillet, R. Delamare, M. Maaref, K. Aguir, Ethanol and ozone sensing characteristics of WO3 based sensors activated by Au and Pd, Sens. Actuators B Chem. 120 (2006) 338–345, http://dx.doi.org/10.1016/j.snb.2006.02.015. [33] J. Bisquert, Theory of the impedance of electron diffusion and recombination in a thin layer, J. Phys. Chem. B 106 (2002) 325–333, http://dx.doi.org/10. 1021/jp011941g. [34] S. Ghosh, C. Roychaudhuri, R. Bhattacharya, H. Saha, N. Mukherjee, Palladium-silver-activated ZnO surface: highly selective methane sensor at reasonably low operating temperature, ACS Appl. Mater. Interfaces 6 (2014) 3879–3887, http://dx.doi.org/10.1021/am404883x. [35] N. Barsan, D. Koziej, U. Weimar, Metal oxide-based gas sensor research: how to? Sens. Actuators B Chem. 121 (2007) 18–35, http://dx.doi.org/10.1016/j. snb.2006.09.047. [36] B. Erdem, R. Hunsicker, G. Simmons, XPS and FTIR surface characterization of TiO2 particles used in polymer encapsulation, Langmuir 17 (2001) 2664–2669, http://pubs.acs.org/doi/abs/10.1021/la0015213 (accessed 11.12.13). [37] I. Kojima, B. Li, T. Fujimoto, XRR and XPS studies of SiO 2 thin films formed by r. f. magnetron sputtering, Surf. Interface Anal. 270 (1999) 267–270. [38] K. Shimizu, C. Phanopoulos, R. Loenders, M.-L. Abel, J.F. Watts, The characterization of the interfacial interaction between polymeric methylene diphenyl diisocyanate and aluminum: a ToF-SIMS and XPS study, Surf. Interface Anal. 42 (2010) 1432–1444, http://dx.doi.org/10.1002/sia.3586. [39] R. Martins, E. Fortunato, P. Nunes, I. Ferreira, A. Marques, M. Bender, et al., Zinc oxide as an ozone sensor, J. Appl. Phys. 96 (2004) 1398, http://dx.doi.org/ 10.1063/1.1765864. [40] S. Kim, Oxygen surface exchange in mixed ionic electronic conductor membranes, Solid State Ionics 121 (1999) 31–36, http://dx.doi.org/10.1016/ S0167-2738(98)00389-0. [41] S.R. Aliwell, J.F. Halsall, K.F.E. Pratt, J.O. Sullivan, R.L. Jones, R.A. Cox, et al., Ozone sensors based on WO3: a model for sensor drift and a measurement correction method, Meas. Sci. Technol. 12 (2001) 684–690, http://dx.doi.org/ 10.1088/0957-0233/12/6/305. [42] J. Guérin, K. Aguir, M. Bendahan, Modeling of the conduction in a WO3 thin film as ozone sensor, Sens. Actuators B Chem. 119 (2006) 327–334, http://dx. doi.org/10.1016/j.snb.2005.12.005. [43] D.E. Williams, Semiconducting oxides as gas-sensitive resistors, Sens. Actuators B Chem. 57 (1999) 1–16, http://dx.doi.org/10.1016/S09254005(99)00133-1. [44] D.E. Williams, K.F.E. Pratt, Resolving combustible gas mixtures using gas sensitive resistors with arrays of electrodes, J. Chem. Soc. Faraday Trans. 92 (1996) 4497, http://dx.doi.org/10.1039/ft9969204497. [45] H. Yabuta, M. Sano, K. Abe, T. Aiba, T. Den, H. Kumomi, et al., High-mobility thin-film transistor with amorphous InGaZnO4 channel fabricated by room temperature rf-magnetron sputtering, Appl. Phys. Lett. 89 (2006) 112123, http://dx.doi.org/10.1063/1.2353811.

687

[46] K. Ide, K. Nomura, H. Hiramatsu, T. Kamiya, H. Hosono, Structural relaxation in amorphous oxide semiconductor, a-In-Ga-Zn-O, J. Appl. Phys. 111 (2012) 073513, http://dx.doi.org/10.1063/1.3699372. [47] J.G. Lu, S. Fujita, T. Kawaharamura, H. Nishinaka, Y. Kamada, T. Ohshima, et al., Carrier concentration dependence of band gap shift in n-type ZnO:Al films, J. Appl. Phys. 101 (2007) 083705, http://dx.doi.org/10.1063/1.2721374. [48] Y. Sato, T. Ashida, N. Oka, Y. Shigesato, Carrier density dependence of optical band gap and work function in Sn-Doped In2 O3 films, Appl. Phys. Express 3 (2010) 061101, http://dx.doi.org/10.1143/APEX.3.061101. [49] T. Kamiya, K. Nomura, M. Hirano, H. Hosono, Electronic structure of oxygen deficient amorphous oxide semiconductor a-InGaZnO 4- x: optical analyses and first-principle calculations, Phys. Status Solidi 5 (2008) 3098–3100, http://dx.doi.org/10.1002/pssc.200779300. [50] A. Takagi, K. Nomura, H. Ohta, H. Yanagi, T. Kamiya, M. Hirano, et al., Carrier transport and electronic structure in amorphous oxide semiconductor, a-InGaZnO4, Thin Solid Films 486 (2005) 38–41, http://dx.doi.org/10.1016/j. tsf.2004.11.223. [51] S. Licht, pH measurement in concentrated alkaline solutions, Anal. Chem. 57 (1985) 514–519, http://dx.doi.org/10.1021/ac50001a045.

Biographies Dali Sun completed his Ph.D. at the bioengineering department, The University of Tokyo in 2015. The subject of his research was to develop biosensor based on oxide materials. He joined the nanomedicine department of Houston Methodist Research Institute after graduation and is researching on plasmonic enhanced biosensor for cancer detection. Hiroaki Matsui, assistant professor at the department of bioengineering, The University of Tokyo. His research focus on using light source in frequency region from near-infrared to THz having high bio-transparency and small bio-damage, meanwhile contributing to development of medical optics using space and time spectroscopes for biomaterials by plasmonic functionalities. Hiroyasu Yamahara received his Ph.D. in bioengineering from The University of Tokyo in 2014 where he focused on oxide material and its application in bioresearch. Currently, as a research associate in the same institute, he continues his research on spin-glass behaviors of the oxide semiconductor materials. Chang Liu received his Ph.D. in biomedical engineering from Florida International University. Since he joined Houston Methodist Research Institute as a research associate in the department of nanomedicine 2014, he has been working on the development and engineering of nanoporous-silica materials, and their clinical application on the detection of proteomic and genomic biomarkers of human. Lei Wu is technologist-in-charge in the department of epidemiology at The Institute of Geriatrics, People’s Liberation Army General Hospital, China. She received her master’s degree from the Chinese People’s Liberation Army Medical School in 2014. Her main research interests focus on the biological engineering, epidemiology and geriatrics. Hitoshi Tabata graduated from Osaka University, Ph.D., is a professor at The University of Tokyo, Graduate School of Engineering. Through his research on expanding the potential of new-age memory devices through researches of ferroelectrics and magnetics, he is particularly interested in the quintessential, natural form of memory storage: DNA.