A Capacitive Liquid-Phase Sensor and Its Sensing Mechanism Using Nanoporous Anodic Aluminum Oxide

Abstract

With the evolution of micro/nanotechnology, anodic aluminum oxide (AAO) has received attention for sensor applications due to its regular and high-aspect-ratio nanopore structure with an excellent sensing performance, especially for electrical and optical sensors. Here, we propose the application of these capacitance and porous properties in a facile nanoporous AAO liquid sensor and study an efficient and economical method for preparing AAO substrates for liquid-phase substance sensing. By applying hybrid pulse anodization (HPA), a growth rate of approximately 5.9 μm/h was achieved in AAO fabrication. Compared to traditional low-temperature (0–10 °C) and two-step anodization with a growth rate of 1–3 μm/h, this process is significantly improved. The effect of pore widening on the performance of electrical sensors is also investigated and discussed. After pore widening, the capacitance values of AAO for air as a reference and various liquids, namely deionized water, alcohol, and acetone, are measured as 3.8 nF, 295.3 nF, 243.5 nF, and 210.1 nF, respectively. These results align with the trend in the dielectric constants and demonstrate the ability to clearly distinguish between different substances. The mechanism of AAO capacitive liquid-phase sensors can mainly be explained from two perspectives. First, since an AAO capacitive sensor is a parallel capacitor structure, the dielectric constant of the substance directly influences the capacitance value. In addition, pore widening increases the proportion of liquid filling the structure, enabling the sensor to clearly differentiate between substances. The other is the affinity between the substance and the AAO sensor, which can be determined using a contact angle test. The contact angles are measured as values of 93.2° and 67.7° before and after pore widening, respectively. The better the substance can fully fill the pores, the higher the capacitance value it yields.

1. Introduction

As environmental and food safety issues gain more attention, the detection of toxic substances has become increasingly important. Various methods have been extensively discussed for this purpose, such as mass spectrometry, Surface-Enhanced Raman Scattering (SERS), micro-electro-mechanical systems (MEMSs), and electrical sensors or devices [1,2,3,4,5,6,7]. Anodic aluminum oxide (AAO) is a porous alumina film formed under specific electrochemical conditions [8,9,10,11,12] with good thermal stability and biocompatibility, which make it highly valued. In traditional process, high-purity aluminum (>99.99%) is used in the preparation of AAO templates through a two-step process or at low temperatures (0–10 °C) [13,14,15,16,17,18,19]. However, conducting electrochemical reactions at low temperatures or the removal required in the two-step process reduces the efficiency of AAO fabrication. From an economic point of view, its disadvantages include high costs and long fabrication times [9]. At present, AAO has maturely been used to improve the hardness, corrosion resistance, and coloration of aluminum metal surfaces [20,21,22] in industry. In applications within the sensor field, AAO is a famous nanomaterial for electrical and optical sensors, especially humidity sensors [23,24,25,26,27,28,29,30,31,32] and SERS templates [33,34,35,36,37]. However, nanoporous AAO sensors still have many more potential possibilities not yet discovered in substance sensing.

For AAO-based electrical sensors, humidity measurements are the most extensively developed applications. Humidity sensors are widely utilized in agriculture, industrial manufacturing, and material storage to monitor environmental conditions. Compared to other materials, the advantages of ceramic materials as humidity sensors include their corrosion resistance, good thermal stability, high mechanical strength, high response, and low response–recovery times. As a ceramic material with porous characteristics, AAO offers a high-aspect-ratio structure that enhances its water vapor adsorption, making it a well-studied candidate for capacitive humidity sensors. Consequently, both the sensing mechanisms and practical applications of AAO-based humidity sensors have been widely explored by various research groups [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38]. In 2010, Zheng et al. [24] investigated the effect of anion concentrations and annealing on AAO humidity sensors to improve their performance. It was reported that impurities could promote the adsorption of water vapor by the AAO pore walls, leading to an increase in capacitance values. To understand the effect of the AAO nanostructure, Kashi et al. [30] studied the relationship between sensitivity and AAO geometry. Anodization in phosphoric acid at high voltages of 165–185 V and the pore widening process have been applied to enhance sensors’ performance. In 2021, low-voltage anodization of AAO at 20 V increased the number of nanopores, resulting in a higher total circumference of the nanoporous structure, which was demonstrated as a method for enhancing the sensor’s response. Through theoretical derivation and experimental data validation, it was found that pore density is significantly correlated with the performance of capacitive AAO electrical sensors [31]. As the discussion of the humidity sensing and mechanisms in AAO-based capacitive sensors becomes more comprehensive, developments are gradually shifting toward practical applications and organic substance detection. For example, in 2022, Lee et al. proposed an AAO-based capacitive sensor for the continuous and real-time measurement of plant water, focusing on agricultural applications [38]. However, most of the current AAO-based electrical sensors primarily focus on humidity sensing, with limited research exploring applications to detecting other substances. Traditionally, analyses of liquid-phase substances have often relied on liquid chromatography techniques, which are complex and time-consuming. Therefore, developing a fast and simple detection method would be highly beneficial, and AAO-based electrical sensors present a promising method for making significant contributions. In 2020, Petukhov et al. investigated the capability of AAO for gas-phase sensing of various organic substances [39]. In 2022, Zhang et al. [40] proposed using AAO to measure liquid-phase methanol concentrations, enhancing the detection sensitivity for HCHO by incorporating BDC-NH₂. More recently, in 2023, different research groups published studies on using AAO to detect ethanol gas [41,42], highlighting the growing importance of nanoporous AAO sensors for detecting various organic solvents.

In this study, we propose using the HPA method for one-step anodization at a room temperature of 25 °C to fabricate AAO, which can significantly reduce the cost and preparation time compared to those of the traditional processes. We also explore the effects of pore widening and the resulting nanostructures on the performance of the electrical sensors. Nanoporous AAO sensors were utilized to detect four different substances, including air, deionized (DI) water, ethanol, and acetone. This demonstrated the ability to effectively distinguish between their signals based on capacitance. Furthermore, we established a theoretical mechanism for AAO liquid-phase sensors, providing an in-depth discussion on the dielectric constant of the substances and the sensor’s surface affinity. A scanning electron microscope (SEM), an LCR meter, and contact angle testing were employed to analyze the nanostructures of the sensors, the electrical measurements, and the surface properties. These techniques enable the rapid identification of different substances and hold potential for future applications in environmental safety and liquid substance detection.

2. Materials and Methods

The experimental method for sensor preparation is shown in Figure 1 below. First, 1050 aluminum alloy (AA1050) was cut into an appropriate size (2.5 cm × 2.5 cm) for our self-made holder. Then, 2-step electrochemical polishing was performed at 20 V in a mixture of HClO₄ and C₂H₅OH solution with a volume ratio of 1:1 for 1 min, followed by using a ratio of 1:4 for 5 min to reduce the surface roughness or scratches. Subsequently, anodization was performed using a 2-electrode system (JIEHAN, Taichung, Taiwan, PPS-2150) with Pt as the counter electrode and AA1050 as the working electrode. The distance between AA1050 and Pt was fixed at 10 cm. We achieved high-efficiency AAO preparation through hybrid pulse anodization (HPA) in 0.3 M oxalic acid with a voltage of 40/−2 V and a duty ratio of 5 s/5 s at 25 °C for 2 h [31]. The temperature was controlled at 25 ± 1 °C using external cooling equipment (DENGYNG, Kaohsiung, Taiwan, D610). After anodization, the AAO was rinsed with DI water for 1 min and dried with N₂ gas to ensure surface cleanliness. The AAO was then immersed in 5 wt% phosphoric acid at 35 °C for pore widening for 20 min. The subsequent analyses compared the nanostructure, electrical properties, and surface characteristics of AAO with and without pore widening. In the manufacturing of AAO capacitors, a metal deposition coater (JEOL, Peabody, MA, USA, JEC-3000FC) is used to sputter Pt using a current of 20 mA on top of the AAO to form the metal–ceramic–metal (MDM) structure for 180 s. This forms a parallel capacitor structure between the aluminum oxide and air, facilitating the entry of various liquid substances into the air-filled spaces, thereby enabling the effective detection of different substances. The capacitance data measurements were performed using an impedance analyzer (HIOKI, Dallas, TX, USA, IM3533) with a computer for the data analysis (Z-ware software ver. 1.3). The nanostructure of AAO was observed using a high-resolution field scanning electron microscope (HRFESEM, HITACHI, SU-5000, Tokyo, Japan) and analyzed using the ImageJ software (ImageJ ver. 1.53t). The contact angle test was conducted using a self-assembled measurement platform to capture images, and the angle analysis was performed using the contact angle function in ImageJ. Different test substances, including DI water, alcohol, acetone, and the pore widening AAO structures, were analyzed to discuss the surface hydrophilicity and hydrophobicity.

3. Results and Discussion

The AAO capacitive sensor was fabricated using the one-step HPA method. The nanostructure of the AAO, the sensor performance, and the sensing mechanism are discussed in detail in this section. After the anodization process, the AAO nanostructure was observed using the SEM, and the results are shown in Figure 2. Figure 2a shows the top view of a SEM image of AA1050 anodized at room temperature using HPA at 40/−2 V for 2 h. Through the ImageJ analysis, the average pore diameter was determined to be 31.4 ± 4.7 nm. Figure 2b illustrates the surface morphology shown in Figure 2a after the AAO underwent pore widening in 5 wt% phosphoric acid at 35 °C for 20 min, resulting in an average pore diameter of 70.5 ± 4.2 nm. This analysis revealed a pore widening rate of 2 nm/min, with the larger pore diameter expected to be more advantageous for facilitating the entry of substances into the pores. In addition, the AAO pores remain complete without burning, demonstrating that the HPA method successfully enhances the growth of AAO at room temperature. The HPA method involves applying a square-wave anodization voltage by coupling a normal positive potential with a smaller negative potential, allowing the current to remain at zero during the negative potential period. This effectively reduces Joule heating in the AAO growth process, thereby improving the reaction temperature, which increases to 25 °C [9,31]. It is noted that replacing the zero potential with a small negative potential during HPA is intended to prevent the reverse discharge caused by the capacitive properties of AAO. Figure 2c,d show cross-section images of AAO corresponding to the experimental parameters of Figure 2a,b, respectively. Based on the ImageJ analysis, the AAO’s thicknesses were measured to be 11.7 μm and 11.8 μm, respectively. The similar thicknesses of the two samples are attributed to the identical anodization parameters, with pore widening primarily affecting the pore diameter rather than the thickness of the AAO structure. This high-efficiency AAO fabrication achieved a growth rate of approximately 5.9 μm/h, significantly faster than the traditional rate of 1–3 μm/h, effectively reducing the overall processing time.

In the capacitance measurements of different substances, we dropped the test liquids onto the area using a sputtered Pt electrode. The porous property of AAO made it easy for the test substances to enter the pores and promoted the differences in the electrical properties. The measurement results for AAO prepared through HPA at 40/−2 V for 2 h are shown in Figure 3, with each substance measured three times and averaged. The measured capacitance values for air, DI water, alcohol, and acetone were 6.6 nF, 125.4 nF, 167.9 nF, and 162.2 nF, respectively. Since the AAO liquid-phase sensor was an MDM structure and could be regarded as a parallel capacitor, the measurements followed Equation (1) below:

$$C={\displaystyle \frac{\epsilon A}{d}}$$

(1)

where ε represents the dielectric constant, A is the electrode area, and d is the thickness of the AAO. When different substances enter the AAO’s pores, they cause changes in the capacitance value. This is due to the parallel contribution of the AAO and the test substance, which can be expressed by the following, Equation (2):

$$C={\displaystyle \frac{{\epsilon }_{AAO}A\left(1-\alpha \right)}{d}}+{\displaystyle \frac{{\epsilon }_{substance}A\left(1-\alpha \right)}{d}}$$

(2)

where ε_AAO represents the dielectric constant of AAO, approximately 9.3 [9], and ε_substance is the dielectric constant of the test substance. When no liquid is added, the dielectric constant of air (as the test substance) is 1. According to studies by Mohsen-Nia et al. [43] and Alejandre et al. [44], the dielectric constants of alcohol and acetone are 25.0 [43] and 20.8 [44], respectively. In addition, α represents the porosity of AAO, which was analyzed to be approximately 10% based on Figure 2a. Since AAO grows with vertically aligned pores, this ratio corresponds to the proportion of liquid that occupies the capacitive AAO liquid-phase sensor.

On the other hand, (1−α) represents the portion occupied by the AAO, which also corresponds to the proportion of the capacitance contributed by the AAO. However, in Figure 3, while the measurement results for air, alcohol, and acetone align with our expectations, DI water, which has the highest dielectric constant of 80, yields a lower measured capacitance of 125.4 nF. This discrepancy arises due to the surface characteristics, as illustrated in Figure 4. When DI water is dropped onto the nanoporous AAO sensor, it does not completely occupy the whole space in pores. This is due to the surface tension of water and the surface properties of AAO. In contrast, alcohol and acetone fully occupy the pores, contributing to their complete capacitance values. As a result, although DI water has a higher dielectric constant, its signal as measured by the sensor is lower.

To address this inconsistency with the expected trend, we measured AAO after the pore widening treatment in 5 wt% phosphoric acid at 35 °C for 20 min for comparison. The results are shown in Figure 5. After the pore widening process, the porosity of AAO significantly increased to approximately 30%, allowing a higher proportion of the test substance to occupy the capacitive AAO sensor. The measured capacitance values for air, DI water, alcohol, and acetone increased to 3.8 nF, 295.3 nF, 243.5 nF, and 210.1 nF, respectively, which aligned well with the trend in the dielectric constants. Comparing the capacitance measurements of air before and after pore widening, the value decreased from 6.6 nF to 3.84 nF. This is because the proportion of AAO decreased, reducing its contribution to the overall capacitance. Among the tested substances, DI water showed the largest change. This is attributed to the increased porosity of the AAO, which enabled the DI water to fill the AAO’s pores more effectively, thereby contributing more significantly to the capacitance value. For alcohol and acetone, the measured capacitance difference increased from 5.7 nF before pore widening to approximately 33.4 nF after the pore widening process. Thus, it is observed that as pore widening increases the contribution of the test substances to the capacitance, the resolution for distinguishing between different substances also improves. To understand the ability of AAO nanostructures to distinguish between substances, the dielectric constants of two substances, defined as ε₁ and ε₂, can be expressed using the following, Equation (3):

$$\begin{gathered} \Delta C=\left[{\displaystyle \frac{{\epsilon }_{AAO}A\left(1-\alpha \right)}{d}}+{\displaystyle \frac{{\epsilon }_{1}A\alpha }{d}}\right]-\left[{\displaystyle \frac{{\epsilon }_{AAO}A\left(1-\alpha \right)}{d}}+{\displaystyle \frac{{\epsilon }_{2}A\alpha }{d}}\right] \\ ={\displaystyle \frac{{\epsilon }_{1}A\alpha }{d}}-{\displaystyle \frac{{\epsilon }_{2}A\alpha }{d}}={\displaystyle \frac{({\epsilon }_1-{\epsilon }_2)A\alpha }{d}} \end{gathered}$$

(3)

From Equation (3), it is concluded that increasing the anodization area of AAO or reducing its thickness can achieve a higher resolution. Additionally, the pore widening process shown in Figure 5 primarily increases the porosity of AAO, which leads to greater capacitance differences when measuring different substances.

Figure 6 illustrates the contact angle measurements of AAO before and after pore widening, as well as with different test substances. In Figure 6a, the AAO sample was anodized at 40/−2 V for 2 h and coated with 10 nm of platinum. In Figure 6b, the AAO underwent pore widening treatment in 5 wt% phosphoric acid at 35 °C for 20 min before being coated with 10 nm of platinum. Through the ImageJ software analysis, the contact angles were measured three times and averaged, resulting in values of 93.2° and 67.7° for Figure 6a,b, respectively. This experiment demonstrates that after pore widening, the contact angle of the AAO nanostructure decreased, indicating an increase in hydrophilicity. The smaller contact angle also explains why more DI water was able to occupy the pores, leading to a greater contribution to the capacitance value. Consequently, DI water, with the highest dielectric constant, conforms to this trend by showing the highest measured capacitance value. Figure 6c,d show the contact angle measurement results for the non-pore-widened AAO liquid-phase sensor with alcohol and acetone. No liquid images can be observed in the figures because alcohol and acetone fully enter the AAO pores, resulting in a contact angle close to 0°. This phenomenon was also reported by Redón et al. [45], indicating that the liquid wetted the nanostructured layer, and substances such as ethyl acetate and hexane exhibited similar behavior. The same result was observed in the pore-widened AAO samples, indicating that while pore widening modified the surface properties to allow DI water to enter the pores more effectively, it did not affect the behavior of alcohol and acetone. The effect of nanostructures on surface hydrophilicity has been investigated by multiple research teams [46,47,48]. This phenomenon can be explained according to the surface tension of liquid and divided into three components: anti-wetting pressure (P_a), effective water hammer pressure (P_ewh), and wetting pressure (P_w) [48]. Among these, P_a is generated by capillary forces and prevents liquids from occupying the nanopore structures. P_a is proportional to the surface tension, meaning that the P_a value of DI water is much higher than that of alcohol and acetone, which is why a noticeable contact angle is observed in Figure 6a. P_ewh and P_w are the pressures caused by the impact of liquid droplets on the surface and the kinetic energy of the droplets, respectively. Since the conditions for measuring the contact angle are consistent, P_ewh and P_w are only affected by the density of the test liquid. On the other hand, according to research from Yong et al. [48], an increase in porosity effectively reduces P_a, thereby enhancing the surface hydrophilicity, as shown in Figure 6b. If the anti-wetting pressure is larger than P_ewh and P_w, an anti-wetting state is observed. Conversely, if P_a is smaller than P_ewh and P_w, a contact angle is difficult to observe. As a result, for liquids like alcohol and acetone with less pronounced surface tension, no significant contact angle can be observed on the surface (Figure 6c,d).

The reproducibility and the response–recovery time are also important issues for sensor performance. Figure 7 shows the re-measurement results for the AAO liquid-phase sensor after pore widening following two months of storage. The black line (a), the red line (b), and the green line (c) represent the response–recovery time diagrams for the measurements with DI water, alcohol, and acetone, respectively. The capacitance measured in air decreased from 3.8 nF to approximately 3.6 nF, mainly due to humidity variations on the measurement day. Meanwhile, the stable capacitance values for DI water, alcohol, and acetone were measured at 298.0 nF, 245.2 nF, and 213.0 nF, respectively, which were very close to the values measured two months earlier. This stability is attributed to the stability of AAO and platinum. For the DI water measurements, the response time was approximately 4 s, while the recovery time required was 10–11 s. On the other hand, in the alcohol and acetone measurements, the response time was only 2–3 s, and the recovery time was about 5–6 s. This difference is mainly due to the higher anti-wetting pressure of DI water, which causes a longer stabilization time for droplets making contact with the sensor to maintain a steady state. In contrast, ethanol and acetone diffuse into the pores quickly, resulting in shorter response times. Furthermore, even when the liquids were blown off the sensor’s surface with air, the recovery times for all three substances were generally longer than their response times. This is because evaporation requires time for the liquids to leave the surface. Among these, ethanol and acetone, due to their good volatility, exhibited shorter recovery times. The above results enable effective discrimination of the electrical signals for different liquid substances, facilitating the rapid identification of organic solvent types.

4. Conclusions

We fabricated complete, non-destroyed AAO nanoporous structures at a room temperature of 25 °C using the HPA method with a one-step anodization process in 0.3 M oxalic acid at 40/−2 V for 2 h. This process achieved a thickness of 11.7–11.8 µm and a growth rate of 5.9 um/h. Compared to the traditional two-step anodization process at low temperatures with high-purity Al, the growth rate was approximately 2–5 times faster, improving a more efficient and cost-effective AAO fabrication process. We have developed an effective liquid-phase sensor that can be used to detect various organic solvents. The effects of different nanostructures on the sensor performance is discussed. Pore widening was performed using 5 wt% phosphoric acid at 35 °C, achieving an etching rate of 2 nm/min. The pore widening process was found to significantly enhance the signal intensity and resolution of the sensor, aligning the measured capacitance more closely with the trends in the dielectric constants of liquid substances. After pore widening, the capacitance values of AAO for air, DI water, alcohol, and acetone were measured as 3.8 nF, 295.3 nF, 243.5 nF, and 210.1 nF, respectively. The sensing mechanism of the nanoporous AAO liquid-phase sensor can be explained by two main factors: the dielectric constant of the test substance and its hydrophilicity towards the AAO surface. Since the MDM structure of the nanoporous AAO sensor resembles a parallel capacitor, the dielectric constant of the test substance directly impacts the measured capacitance. Meanwhile, the hydrophilicity of the test substance affects whether it fully occupy the pores and contributes to the capacitance. Before and after pore widening, the contact angle of DI water decreased from 93.2° to 67.7°, indicating improved hydrophilicity in the widened AAO structure. Organic solvents such as alcohol and acetone exhibited near-zero contact angles, demonstrating excellent compatibility with the sensor. This study on AAO as a liquid-phase sensor provides new insights into the use of nanoporous AAO in electrical sensors and contributes to potential applications in the detection of solvents that are toxic to the human body and environmental safety.