Alternating Current Electroluminescent Sensor for Visual Detection of Trace Water in Oil

Abstract

The trace water content in industrial oil critically affects the operational stability and service life of industrial equipment and serves as a key indicator for evaluating oil quality. Therefore, the rapid, sensitive, and visual detection of trace water in oil is of great engineering significance for equipment condition monitoring and early fault warning. Existing detection methods predominantly rely on precision instruments; although they enable quantitative analysis, their operational procedures are complicated and time-consuming, which are unsuitable for on-site real-time monitoring. Consequently, there is an urgent need for a novel trace water detection sensor that offers high sensitivity, visualization, and adaptability to oil-phase environments. Herein, a coplanar electrode alternating current electroluminescent (ACEL) sensor is developed for the visual detection of trace water in oil. The ACEL sensor features a multilayer structure comprising a substrate layer, a coplanar electrode layer, and a humidity-sensitive luminescent layer. The humidity-sensitive luminescent layer consists of humidity-sensitive hydrogel and ZnS: Cu electroluminescent powder, forming a loose and porous film that enables high-sensitivity humidity sensing and simultaneously electroluminescent visual signal output. The sensing mechanism study reveals that variations in trace water content modulate the dielectric properties of the humidity-sensitive layer, which further affect the electroluminescent intensity of the ACEL sensor. In addition, the ACEL sensor enables the rapid, naked-eye recognition of humidity changes under trace water conditions without the need for precision instruments, achieving a rapid response time of 3 s and a detection limit as low as 60 ppm, all making it applicable for different types of industrial oils. Thus, this ACEL sensor features a novel detection mechanism, excellent universality, fast response, and ease of operation, offering a new visual sensing strategy for trace water detection in industrial oil and holding broad prospects for practical applications.

1. Introduction

Industrial oils, including lubricating oils, insulating oils, and hydraulic oils, serve as core media for equipment lubrication, insulation, and power transmission. Their trace water content is a key indicator for evaluating oil quality [1,2,3]. The presence of trace water readily leads to problems such as oil emulsification, the corrosion of metal components, and the degradation of insulation performance, which directly affect the operational stability and service life of industrial equipment. Therefore, the development of rapid, sensitive, and visual detection methods for trace water in oil is of great engineering significance for equipment condition monitoring and early fault warning [4,5,6]. Currently, the mainstream methods for detecting trace water in oil include Karl Fischer titration [7,8], capacitance-based methods [9,10], and spectroscopy [11,12,13]. Although these methods enable quantitative detection, most rely on precision instruments and involve cumbersome operational procedures, making on-site real-time monitoring difficult [14]. Portable detection devices suffer from high detection limits and lack intuitive visual signal feedback [15]. Moreover, conventional contact-type sensors are susceptible to oil contamination, leading to performance degradation over time [16]. Consequently, there is an urgent need to develop a novel sensor that combines high sensitivity, visual readout, fast response, and adaptability to oil-phase environments.

To address these challenges, we report a coplanar electrode alternating current electroluminescent (ACEL) sensor for the visual detection of trace water in oil. The proposed sensor adopts a multilayer structure comprising a substrate, a coplanar electrode layer, and a humidity-sensitive luminescent layer. The humidity-sensitive luminescent layer is composed of humidity-sensitive hydrogel and ZnS: Cu electroluminescent phosphor. The hydrogel enables the rapid response to humidity changes [17,18,19,20,21], while the ACEL effect produces visible light signals [22,23,24,25,26,27]. Specifically, the high hydrophilicity of the hydrogel induces dielectric property changes in the humidity-sensitive luminescent layer, which in turn modulates the electroluminescence intensity, thereby enabling visual sensing of trace water in oil. The detection limit, response speed, and applicability for different oil types are systematically investigated, and its utilization for the naked-eye visual detection of trace water in oil is demonstrated. This study presents an ACEL sensor for on-site industrial oil analysis, providing a theoretical and practical basis for visual, rapid trace water detection.

2. Structure and Characterization of the ACEL Sensor

2.1. Structure and Characterization of the ACEL Sensor

A schematic diagram of the ACEL sensor is illustrated in Figure 1a, which consists of a substrate layer, a coplanar electrode layer, and a humidity-sensitive luminescent layer. The bottom substrate of the sensor is an FR4 printed circuit board (PCB), which provides robust mechanical support for the entire device. The electrode layer consists of coplanar copper films with a thickness of 35 μm. The humidity-sensitive luminescent layer is formed by hydrogel and zinc sulfide (ZnS: Cu) electroluminescent phosphor in a specific ratio to form a loose and porous film, which can simultaneously respond to water content and emit visible signals. An AC power supply is connected to both ends of the coplanar electrodes, and visible light emits from the gap of the coplanar electrodes. Upon dropping oil-phase liquid onto the surface of the device, the dielectric constant of the humidity-sensitive luminescent layer would be changed, resulting in a variation in luminescence brightness. A photograph of the as-prepared ACEL sensor is presented in Figure 1b (the detailed fabrication process is provided in the experimental section (Appendix A)). When an AC power supply with a voltage of 400 V and a frequency of 1000 Hz is applied, green visible light is generated in the gap region of the humidity-sensitive luminescent layer, as shown in Figure 1c. Figure 1d shows the scanning electron microscopy (SEM) characterization of the humidity-sensitive luminescent layer, which clearly presents the microscopic morphology of the humidity-sensitive layer. The ZnS: Cu electroluminescent particles size is approximately 20 μm. The hydrogel closely adheres to the surface of ZnS: Cu particles and binds the luminescent particles due to dehydration, forming a loose and porous humidity-sensitive luminescent layer. This structure facilitates the penetration of water molecules and greatly improves the sensitivity of the sensor. The infrared absorption spectrum of the hydrogel is displayed in Figure 1e. Characteristic absorption peaks of the amide group (-CONH-) can be clearly observed: the C=O stretching vibration peak of the Amide I band at about 1650 cm⁻¹ and the N-H bending vibration peak of the Amide II band at about 1550 cm⁻¹. This confirms the existence of amide groups in the hydrogel matrix, which is the chemical structural basis for the efficient humidity response performance of the hydrogel. Meanwhile, an obvious O-H stretching vibration peak is observed at about 3300 cm⁻¹, indicating the presence of water molecules.

2.2. Optimization of Electrode Parameters and Theoretical Simulation of the ACEL Sensor

To obtain an ACEL sensor with excellent visualization performance, each component of the sensor should be optimized. As we all know, the spacing of the coplanar electrode has an important impact on the electric field [28], which would in turn affect the electroluminescent intensity. Therefore, we designed sensors with different electrode spacings to achieve high-intensity electroluminescence (see the experimental section for detailed preparation methods). Figure 2a shows the photographs of the prepared ACEL sensors with different spacing distances, and the electrode spacing of the sensors was gradually set to six distances: 0.2 mm, 0.4 mm, 0.6 mm, 0.8 mm, 1.0 mm, and 1.2 mm. Figure 2b is a luminescence image of the device driven by an AC power supply. It can be observed that effective electroluminescence is achieved in all spacing regions of the sensor, and a smaller electrode spacing would introduce stronger luminescence intensity. The luminescent intensity under each electrode spacing was measured with a fiber optic spectrometer, as shown in Figure 2c. The inset in Figure 2c is the spectra peak of each electrode spacing. The spectra and peak curve indicate that as the electrode spacing gradually increases from 0.2 mm to 1.2 mm, the spectra peaks are 44,217, 13,139, 6637, 2324, 1565, and 683 counts, showing a gradual downward trend. Figure 2d is the spectra peak monitoring of each spacing, which is obtained by moving the sensor in Figure 2b at a constant speed, and the optical fiber probe of the spectrometer is fixed to continuously collect the spectra peak of each electrode spacing region from 1.2 mm to 0.2 mm. It can be seen that when the electrode spacing decreases from 1.2 mm to 0.2 mm, the spectra peak follows a stepwise increment pattern, and the luminescence intensity of the smallest spacing (0.2 mm) is about 65 times that of the largest spacing (1.2 mm).

To figure out the intrinsic physical mechanism of the electrode spacing regulating the electroluminescent performance of the sensor, ACEL devices with different electrode spacings were prepared. The capacitance of each device with different electrode spacings is measured by an LCR meter as in Figure 2e. It shows that as the electrode spacing of the device increases, the capacitance value gradually decreases. Finite element simulation for electric field distribution of the ACEL sensor with different electrode spacings is performed in Figure 2f. It can be observed that when the electrode spacing is smaller (0.4 mm), the electric field among the electrode spacing region is much higher. The electric field intensity curves of different electrode spacings are also simulated in Figure 2g, and the highest electric field (3.6 MV/m) corresponding to the smallest electrode spacing (0.2 mm) is about five times the peak value (0.7 MV/m) corresponding to the largest spacing (1.2 mm), which further confirms that the electric field increases significantly with the decrease in electrode spacing. Although the 0.2 mm spacing shows the highest luminescence intensity (44,217 counts), the spacing size is too small and thus the luminescent area is tiny, which brings inconveniences to the data collection operation in the actual test process; thus, the optimal electrode spacing of the ACEL visual humidity sensor was finally determined to be 0.4 mm. Therefore, the sensor would not only possess high luminescent intensity but also is suitable for visual observation.

Except for the electrode spacing, the thickness of the humidity-sensitive luminescent layer also has a significant influence on the luminescent intensity [29,30]. Figure 3a–c illustrate the optimization of the humidity-sensitive luminescent layer thickness for the ACEL sensor, achieved by systematically investigating the influence of layer thickness on both the electroluminescent performance and capacitance of the device. First, sensor devices with humidity-sensitive luminescent layer thicknesses of 50 μm, 100 μm, 150 μm, 200 μm, and 250 μm were fabricated individually (see the experimental section for detailed preparation methods). Figure 3a shows the spectra and the spectral peaks with the thicknesses (inset) of sensors upon dropping pure oil. It indicates that the luminescent intensity of the sensor first increases and then decreases with increasing thickness, reaching a peak at approximately 200 μm. Furthermore, upon adding a mixed liquid (oil–water ratio of 400:1), the electroluminescent intensity of the sensor is shown in Figure 3b, which shows a consistent variation trend of luminescent intensity as that in Figure 3a, also reaching a peak value at 200 μm. The capacitances of the sensors when dropping two kinds of liquids are tested in Figure 3c, which indicate that the device capacitance increases continuously with the increase in layer thickness despite the amplitude of capacitance variation being quite small. Therefore, although the increase in the thickness of the humidity-sensitive luminescent layer promotes the capacitance of the dielectric layer and enhances the electric field between the electrode gaps, which is beneficial for electroluminescent intensity, on the contrary, an excessive thickness will hinder the emission of light. Therefore, 200 μm is the optimal thickness for the sensitive luminescent layer.

In addition, the mass ratio of electroluminescent powder to hydrogel in the humidity-sensitive luminescent layer significantly influences both the luminescence performance and the mechanical stability. To investigate a proper ratio, ACEL sensors were fabricated with powder-to-binder mass ratios of 1:3, 2:3, 3:3, 4:3, and 5:3 (see the experimental section for detailed preparation procedures). Figure 3d presents the electroluminescent spectra of the sensors with different powder-to-binder ratios; the inset shows the corresponding variation in spectra peak. The results indicate that the electroluminescent intensity increases progressively with an increasing powder-to-binder ratio, reaching a maximum at a ratio of 5:3. However, further increasing the powder content leads to a marked decline in the film-forming ability of the humidity-sensitive layer, potentially resulting in incomplete film formation. This, in turn, significantly compromises the mechanical integrity of the device and reduces its operational lifetime. Therefore, to achieve both high luminescence intensity and good mechanical stability, an optimized powder-to-binder ratio of 5:3 is determined.

For ACEL devices with coplanar electrodes, the driving voltage and frequency are key electrical parameters that influence the luminescent intensity [31,32]. Figure 3e illustrates the relationship between the luminescent intensity and the driving voltage. As the voltage increases from 150 V to 400 V At a fixed frequency of 1000 Hz, the luminescent intensity rises from 878 counts to 43,528 counts, presenting an approximately a 50-fold enhancement. The expression describing the relationship between the layered ACEL luminescent intensity and applied voltage is as follows [33]:

$$L=L_0\exp \left(-{\displaystyle \frac{\beta }{\sqrt{V}}}\right)$$

(1)

where L is the luminescent intensity, V is the applied voltage, and L₀ and β are parameters determined by the device materials and structure. By fitting the experimentally measured data points, the formula for ACEL luminescent intensity is obtained as follows:

$$L=1.245\times {10}^8\exp \left(-{\displaystyle \frac{31632}{\sqrt{V}}}\right)$$

(2)

It can be seen in Figure 3e that the measured points are in good agreement with the fitted curve, and the goodness of fit is 0.998.

Similarly, at a fixed voltage of 400 V, the luminescent intensity of the ACEL sensor with varying AC frequency is shown in Figure 3f. The results indicate that the luminescent intensity of the ACEL device increases gradually to 2.45 times when the frequency rises from 200 Hz to 1000 Hz. The expression for the relationship between the luminescent intensity and the frequency of the ACEL sensor is as follows [34]:

$$L=C\times f^m$$

(3)

The formula between luminescent intensity and frequency is obtained by fitting the experimentally measured data points as follows:

$$L=1263f^{0.515}$$

(4)

As shown in Figure 3f, the measured points are in good agreement with the fitted curve, with a goodness of fit of 0.986 and minimal residuals.

2.3. Stability Analysis of the Device in Simulated Industrial Oil Environments

After completing the structural design, parameter optimization, and simulation verification of the visual humidity sensor, we further conducted a systematic analysis of the device’s operational stability under simulated industrial oil conditions. Figure 3g shows the variations in the device’s capacitance and luminescence intensity over the temperature range of 10–60 °C, exploring the influence of temperature on device performance. The test results show that as the temperature increased from 10 °C to 60 °C, the device’s capacitance decreased slightly from 2.702 pF to 2.687 pF, and the luminescence intensity decreased marginally from 19,456 counts to 18,633 counts, with both indicators showing extremely low levels of variation. This result indicates that temperature changes have an extremely small impact on device performance, and the device exhibits good operational stability under common temperature fluctuations in industrial oils.

To further explore the long-term operational stability of the device in oil environments, the device was continuously immersed in oil for 72 h, and its capacitance and luminescence intensity were measured every 6 h. The corresponding test results are presented in Figure 3h. In the initial 24 h of immersion, the capacitance of the device increased slightly from 2.645 pF to 2.754 pF, and the luminescence intensity rose from 12,889 counts to 13,902 counts. During 24–48 h, the variation trend of device performance slowed down significantly, with the capacitance increasing marginally from 2.754 pF to 2.764 pF and the luminescence intensity rising slowly from 13,902 counts to 14,611 counts. After 48 h of immersion, the device performance basically stabilized, with the capacitance maintained at approximately 2.766 pF and the luminescence intensity stabilized at around 14,678 counts.

The above test results indicate that trace polar impurities, degradation products, functional additives, acids, alcohols and various contaminants in oil can exert a slight effect on the dielectric properties of the device during long-term oil immersion, yet the overall interference level is extremely low. In terms of dielectric characteristics, pure water has a dielectric constant of approximately 80, while organic acids, esters, aldehydes and ketones in oil have dielectric constants of only 3–6, alcohols of 2.5–30, and various polar additives of 2.5–6. The dielectric constant of water is far higher than that of all impurity components in oil. According to the structural and response characteristics of hydrogel, water molecules feature strong polarity and small molecular size and can construct rapid transmission channels in the hydrogel polymer network through intense hydrogen bonding interactions to realize efficient response. In contrast, polar impurity molecules in oil have larger molecular sizes and complex spatial structures, requiring overcoming the network steric hindrance of hydrogel during diffusion and penetration, and their hydrogen bonding capacity is much weaker than that of water molecules. Meanwhile, nonpolar oil molecules are repelled by hydrophobic effects and have extremely poor affinity with the hydrogel skeleton, making it difficult to invade the hydrophilic network structure. Most metal impurities in oil exist in the form of worn metal particles or macromolecular complexes, whose overall size is far larger than the pore size of the hydrogel network, so they cannot penetrate into the sensitive layer of the device. Furthermore, the single test duration of the device is only 1–2 min, and the short-term detection mode can further avoid the penetration and pollution interference of various impurities. Combined with the overall test trends, the hydrogel structure of the device’s humidity-sensitive luminescent layer shows no deactivation, degradation or damage after long-term oil immersion. It is fully demonstrated that the coplanar electrode ACEL device possesses excellent long-term operational stability in simulated oil environments.

3. Experimental Results and Discussion

3.1. Detection of Trace Water in Oil by the ACEL Sensor

To realize the detection of the trace water content in oil by the ACEL sensor, Figure 4 systematically investigates the luminescent performance and capacitance variation of the sensor under different trace water contents in mineral lubricating oil and reveals the internal working mechanism of the sensor for detecting trace water in oil through finite element simulation. Figure 4a shows the comparison of luminescent spectra of the ACEL sensor under the conditions of no liquid dropping, pure oil dropping, and mixed liquid dropping (oil–water ratio of 2000:1). The spectra show that the luminescent peak with pure oil dropping is lower than that without any liquid, while the luminescent intensity of the sensor is significantly enhanced with the oil–water mixture dropping, indicating that the presence of water can remarkably increase the luminescent intensity. The luminescence photographs of the ACEL sensor upon dropping different oil–water ratios are presented in Figure 4b. With the decrease in the oil–water ratio (increase in trace water content), the brightness of the luminescent region of the ACEL sensor shows a significant increasing trend. Figure 4c calibrates the water content (ppm) of different oil–water ratios using a coulometric titrator, where the measured water content is 60.3 ppm at an oil–water ratio of 2000:1, 66.9 ppm at 1600:1, 91 ppm at 1200:1, 104.7 ppm at 800:1, and 110.5 ppm at 400:1. The relationship between the luminescent intensity of the ACEL sensor and the oil–water ratio is shown in Figure 4d, and the luminescent intensity is positively correlated with water content. The capacitance of the ACEL sensor when dropping liquids with different water contents is measured in Figure 4e, which indicates that an increase in trace water content (or a decrease in the oil–water ratio) leads to a corresponding increase in the capacitance of the ACEL sensor, which is consistent with the spectral curve in Figure 4d. Therefore, the ACEL sensor could realize visual detection and a quantitative response of trace water at a level as low as 60 ppm for mineral lubricating oil.

In addition to the commonly used mineral lubricating oil, we also evaluated the sensor’s performance in detecting trace water in other typical industrial oils, including insulating oil and hydraulic oil. The electroluminescent spectra measured upon dropping three types of oils is presented in Figure 4f, which show that under different oil–water ratios (400:1, 1200:1, and 2000:1), the luminescent intensity of the ACEL sensor deceases monotonically with an increasing oil–water ratio. It indicates that the ACEL sensor achieves a stable visual luminescence response to trace water across different oil types, demonstrating its broad applicability. The capacitances of the ACEL sensors upon dropping oils with varying water contents are also measured in Figure 4g. For all three oils, the device capacitance decreases with a rising oil–water ratio, a trend consistent with the luminescent intensity variation in Figure 4f. Therefore, owing to compositional differences among different oils, their intrinsic dielectric properties result in different baseline capacitances even at the same water content, which in turn leads to the differences in luminescent intensity.

To figure out the sensing working mechanism of the ACEL sensor, a finite element simulation model was established to analyze the variation in electric field upon dropping oil–water mixtures with different water contents [35]. A comparison of the simulated electric field distributions under high and low water content conditions is presented in Figure 4h. Obviously, under a high water content, the electric field intensity in the humidity-sensitive layer region is significantly higher than that under low water content, accompanied by a more concentrated and stronger electric field distribution. This observation explains the enhancing effect of water content on the electric field. Figure 4i further simulated the electric field under different water content conditions. The electric field distribution curves reveal that as the water content gradually increases, the peak electric field intensity exhibits an increasing trend, which in turn promotes the electroluminescent intensity of the ACEL sensor.

3.2. Application Demonstration of Visual Detection for Trace Water in Oil

Based on the device optimization and performance characterization of the ACEL sensor, an application demonstration for the visual detection of trace water in oil was designed. As shown in Figure 5a, a comparative ACEL sensor was fabricated, featuring two luminescent regions: a reference area and an experimental area. The experimental area was immersed in lubricating oil with an oil–water ratio of 2000:1 (corresponding to a trace water content of approximately 60 ppm) and continuously observed for 6 min. As shown in the luminescent photograph in Figure 5b, the luminescent intensity of the experimental area is significantly enhanced after immersion. Interestingly, the reference area, which is not in direct contact with the oil sample, also exhibits a slight increase in brightness. This phenomenon can be explained by the dynamic change in device capacitance shown in Figure 5c. During the 6 min immersion process, the device’s total capacitance increases continuously over time, indicating an overall rise in device capacitance and a corresponding enhancement of the electric field intensity, which in turn leads to the increased luminescent intensity observed in the reference area. Nevertheless, the brightness increase in the reference area is substantially lower than that of the test region.

To further verify the humidity sensitivity of the ACEL sensor to trace water content, the experimental area was immersed in lubricating oil of three different conditions: pure oil, low water content (2000:1, 60 ppm), and high water content (1200:1). The spectra peak of the device was continuously monitored, as shown in Figure 5d. The results indicate that even under the low trace water content condition of 60 ppm, the device brightness increases from 5250 counts to 5500 counts, still generating a detectable light intensity response. This fully demonstrates the excellent humidity sensitivity of the ACEL sensor. Moreover, a higher water content would lead to a greater peak intensity change.

Furthermore, the humidity response speed of the ACEL sensor was evaluated, as shown in Figure 5e. An oil sample with an oil–water ratio of 400:1 was dropped onto the test region of the device. The spectra peak monitoring results reveal that within 3 s, the brightness of the experimental area rapidly increases from 5250 counts to 8250 counts, exhibiting a sharp enhancement, and the change in light intensity is distinguishable by the naked eye. Therefore, the ACEL sensor enables the naked-eye visual detection of trace water without relying on precision measuring instruments.

4. Conclusions

In summary, a coplanar electrode ACEL sensor for the visual detection of trace water in oil was proposed, featuring a loose and porous humidity-sensitive luminescent layer composed of hydrogel and ZnS: Cu electroluminescent powder. First, the sensing mechanism of the ACEL sensor was systematically elucidated. Increasing the trace water content can elevate the capacitance of the humidity-sensitive layer, which strengthens the internal electric field and further positively regulates the luminescence intensity. This design realizes dual functions of humidity perception and electroluminescent signal output. Moreover, practical application tests verified that the proposed sensor can produce distinct luminescent responses toward trace water with a detection limit as low as 60 ppm. In addition, it delivers a rapid response time of around 3 s and supports visual detection without relying on sophisticated testing equipment. Consequently, the developed ACEL sensor possesses a rational working mechanism, favorable practicability and fast response performance. It provides an innovative visual sensing route for trace water monitoring in industrial oils and holds great promise for practical industrial deployment.