A Flexible Wearable Sensor for In Situ Non-Destructive Detection of Plant Leaf Transpiration Information

Abstract

This paper investigates an in situ, non-destructive detection sensor based on flexible wearable technology that can reflect the intensity of plant transpiration. The sensor integrates four components: a flexible substrate, a humidity-sensing element, a temperature-sensing element, and a self-adhesive film. It is capable of accurately and continuously measuring the temperature, humidity, and vapor pressure deficit (VPD) on the leaf surface, thus providing information on plant transpiration. We combined the humidity-sensitive material graphene oxide (GO) with a PDMS-GO-SDS flexible substrate as the humidity-sensing element of the sensor. This element exhibits high sensitivity, fast response, and excellent biocompatibility with plant interfaces. The humidity monitoring sensitivity of the sensor reaches 4456 pF/% RH, while the temperature sensing element has a sensitivity of approximately 3.93 Ω/°C. Additionally, tracking tests were conducted on tomato plants in a natural environment, and the experimental results were consistent with related research findings. This sensor can be used to monitor plant growth during agricultural production and facilitate precise crop management, helping to advance smart agriculture in the Internet of Things (IoT) for plants.

1. Introduction

Transpiration is the process by which plants absorb water from the soil through their root systems, transport it upward via the stem, and release it into the atmosphere through leaf transpiration [1]. Traditional plant transpiration monitoring techniques, such as the steady-state porometer method [2], rapid weighing method [3], and isotope tracing method [4], have made certain advancements. However, these techniques have limitations in terms of cost-effectiveness, are prone to interference from low-noise signals and harsh weather conditions, and the accuracy and timeliness of the measurements are restricted.

Plant VPD (vapor pressure deficit) detection is a method used to reflect the intensity of plant transpiration. VPD refers to the difference between the actual water vapor pressure and the saturated water vapor pressure in the air. It is an important indicator of the dryness of the air and serves as a driving force for plant transpiration. Within an appropriate range, a higher VPD value indicates a greater level of air dryness, leading to more intense evaporation and transpiration. The environmental VPD (VPD_A) monitoring method can partially reflect the transpiration state of plants. However, because the environmental temperature and relative humidity data collected using this method are measured at a certain distance from the plant, the results are easily influenced by environmental factors between the sensor and the plant. This makes it difficult to accurately reflect the true vapor pressure deficit of the microenvironment surrounding the plant leaves [5]. In contrast, the leaf-scale VPD (VPD_L) detection method calculates the vapor pressure deficit at the leaf scale by simultaneously measuring the relative humidity on the leaf surface, the leaf surface temperature, and the ambient temperature. Since VPDL is measured closer to the leaf’s original position, it is less affected by environmental factors and can more accurately reflect the intensity of plant transpiration [6].

However, to detect leaf-scale VPD (VPD_L), it is necessary to obtain in situ temperature and humidity information from the leaf surface. Afzal et al. developed a clamp-type sensor that attaches to the leaf surface to detect real-time changes in leaf thickness and water content. However, the rigid structure of the sensor can damage growing plants [7]. With significant advancements in flexible electronics technology, new solutions have recently emerged to address these issues. Plant bioelectronic sensors based on flexible electronics have been developed, which can effectively convert plant life signals into electrical signals [8,9]. The integration of flexible electronics with plant science offers a new approach to the digital and precise monitoring of plant transpiration information [10,11,12].

The key performance indicators of humidity sensors, such as linearity and sensitivity, are primarily influenced by the sensing materials used. The abundant oxygen-containing functional groups in graphene oxide readily interact with water molecules in the environment, making it an ideal candidate for humidity sensing [13,14,15,16]. Yin et al. integrated a graphene-based humidity-sensing element and a gold-based thin-film thermistor on a polyimide (PI) sheet. Additionally, a 1 mm thick waterproof double-sided adhesive was applied to the edges of the PI sheet, allowing it to adhere to the leaf surface for measuring plant transpiration-related information [17]. Although current research has yielded some successful examples in monitoring external features such as temperature and humidity on leaf surfaces and the growth of plant organs [18,19,20], developing a highly sensitivity, fast-response, low-detection-limit, non-invasive wearable physiological sensor for crops remains a significant challenge.

This paper addresses the need for leaf temperature, leaf surface relative humidity, and ambient temperature data in leaf VPD_L detection. It integrates flexible wearable electronic technology and smart sensing technology to design and fabricate a novel flexible wearable in situ non-destructive detection sensor. This sensor utilizes graphene oxide as the sensitive material, combined with PDMS-GO-SDS flexible materials to construct the humidity-sensing element, while a platinum thin film serves as the temperature sensing element. Additionally, we developed a self-adhesive porous breathable film with excellent biocompatibility to connect the flexible wearable sensor to the plant leaves. The sensor was subjected to interference resistance testing and signal response performance evaluation and calibration. Finally, the flexible sensor was applied to the surface of tomato plant leaves to explore the biocompatibility between the sensor and the plant, and the sensor’s effectiveness in detecting leaf-scale VPD (VPD_L) was tested in a real environment.

2. Experimental Section

2.1. Chemical Reagents and Materials

PDMS (polydimethylsiloxane) sol was purchased from Dow Corning in the United States (model number: Sylgard 184). Graphene oxide (GO) aqueous solution (concentration 2 mg/mL) was obtained from Nanjing Xianfeng Nanomaterials Technology Co., Ltd., China. Monolayer graphene oxide powder was sourced from Shenzhen Suiheng Technology Co., Ltd., China. Sodium dodecyl sulfate (SDS) powder was purchased from Beijing Solarbio Science and Technology Co., Ltd., China. model number: S8010. Conductive ink was obtained from Tokyo Jujyo Ink Co., Ltd., Tokyo, Japan (model number: JELCON CH-8). Platinum thin film was sourced from Heraeus GmbH in Germany (model number: PT1000). Deionized ultrapure water used during the experiments was sourced from the Milli-Q ultrapure water system. Ethanol (C₂H₅OH) and acetone (C₃H₆O) were purchased from Shanghai China National Pharmaceutical Group Chemical Reagents Co., Ltd., Shanghai, China.

2.2. Equipment

Water bath ultrasonic cleaner: S800 Branson; magnetic stirrer: LC-MSB-S; precision electronic balance: JT3003D; laser engraving machine: Nano Pro III; constant temperature and humidity chamber: GDJ(S); LCR tester: VICTOR-4091B; scanning electron microscope: Gemini 300 Zeiss; UV-Vis spectrophotometer: Agilent Cary 60 Keysight.

2.3. Preparation of Flexible Interdigitated Electrodes

As shown in Figure 1a, a certain amount of monolayer graphene oxide powder is added to a beaker containing ultrapure water and subjected to ultrasonic treatment for 10 min to prepare a GO aqueous solution with a concentration of $4 \mathrm{m}\mathrm{g}/\mathrm{m}\mathrm{L}$. An appropriate amount of SDS powder is weighed using a precision electronic balance and added to the above GO aqueous solution, followed by an additional ultrasonic treatment for 10 min to prepare a 0.2 M SDS aqueous solution. The PDMS sol and curing agent are mixed in a volume ratio of 10:1 in the beaker and thoroughly stirred using a magnetic stirrer. Then, 1/5 of the volume of the PDMS solution of the GO-SDS mixed aqueous solution is added to the beaker, and stirring is continued until a uniform grayish-white mixture is obtained. We place the beaker containing the grayish-white mixture of PDMS, GO, and SDS into a vacuum chamber for a specific period to remove internal air bubbles. Then, we remove the beaker and drop the mixture into the flexible substrate mold, using a scraper to level the surface. The mold is then placed in a constant temperature and humidity chamber at 80 °C for 2 h to fully cure and form a PDMS-GO-SDS polymer flexible substrate.

As shown in Figure 1b,c, the interdigitated electrode structure is designed using Auto CAD 2018 software and imported into the laser engraving machine. The cured PDMS-GO-SDS polymer film is placed on a UV processing platform, where local exposure is performed on the surface of the film. The glass slide with the PDMS-GO-SDS polymer film is adhered and fixed to a flat tabletop, and the hollow copper mold is removed. Place the screen printing template with the interdigitated electrode directly above the polymer film. Pour in the conductive ink and use a dedicated screen printing squeegee to print the interdigitated electrode onto the PDMS-GO-SDS film. Simultaneously, print the same interdigitated electrode onto a PI film and allow it to air dry naturally. After the flexible substrate has dried, use a blade to cut the flexible interdigitated electrodes with varying spacings and peel them off the optical glass slide.

2.4. Modification and Integration of Temperature and Humidity-Sensing Elements

We place the flexible interdigitated electrode flat on a glass slide and use 100 μL of a $2 \mathrm{m}\mathrm{g}/\mathrm{m}\mathrm{L}$ GO aqueous solution to drop two times in the interdigitated area (Figure 2). After allowing it to air dry, the modification of the humidity-sensing element is completed. A small amount of PDMS and curing agent is mixed in a beaker at a ratio of 10:1 and thoroughly stirred using a magnetic stirrer. The bottom surface of the platinum thin film is then coated with the PDMS solution, and it is attached to a predefined position on the lower part of the flexible interdigitated electrode substrate. The assembly is placed in a constant temperature and humidity chamber at 50 °C for 2 h to cure, completing the modification of the temperature sensing element, as shown in Figure 2. The humidity and temperature sensing elements are integrated into the flexible substrate, with the front of the flexible wearable sensor containing the leaf temperature and humidity-sensing elements and a localized self-adhesive film, while the back features the environmental temperature sensing element.

2.5. Performance Testing of Temperature Sensing Elements

The sensor is placed in a constant temperature and humidity chamber, where the relative humidity is adjusted from 20% RH to 90% RH in increments of 10% RH at fixed temperatures of 10 °C, 25 °C, and 40 °C. Light intensity is varied from 0 Lux to 10,000 Lux in increments of 2000 Lux. An LCR tester is used to record the resistance changes of the temperature-sensing element, enabling the analysis of its resistance to interference.

The sensitivity S_T of the temperature sensing element is defined as

(1)$S_T={\displaystyle \frac{R-R_0}{T-T_0}}.$

In this procedure, T represents the final measured temperature, $T_0$ denotes the initial measured temperature, R is the resistance value at the final temperature, and $R_0$ is the resistance value at the initial temperature. The temperature in the constant temperature and humidity chamber is gradually adjusted from 10 °C to 40 °C in 5 °C increments, and the resistance values of the sensing element at different temperatures are recorded.

The temperature-sensing element is quickly transferred from a constant temperature and humidity chamber at 10 °C to another chamber set at 40 °C (with a relative humidity of 30% RH), and the time taken for the resistance value of the temperature-sensing element to stabilize is recorded as its response time. To test the stability of the temperature-sensing element over a period of 30 days, resistance values are measured at different temperatures (10 °C, 20 °C, 30 °C, and 40 °C) at intervals of 5 days.

2.6. Performance Testing of Humidity-Sensing Elements

Under conventional conditions of 25 °C and 40% RH in a constant temperature and humidity chamber, humidity response tests are performed on interdigitated electrodes with a gap of 200 μm using different detection frequencies. The changes in capacitance values over time are observed to determine the appropriate detection frequency.

At fixed humidity levels of 30% RH, 60% RH, and 90% RH, the temperature in the constant temperature and humidity chamber is adjusted from 10 °C to 40 °C in increments of 5 °C, while the light intensity varies from 0 Lux to 10,000 Lux in increments of 2000 Lux. An LCR tester records the capacitance changes of the humidity-sensing element. For bending interference analysis, as illustrated in Figure 3, the anti-interference performance of the humidity-sensing element is evaluated by altering its bending degree and measuring the corresponding capacitance values.

The sensitivity $S_{RH}$ used to characterize the humidity-sensing element is defined as

(2)$S_{RH}={\displaystyle \frac{C-C_0}{RH-R{H}_0}},$

where RH represents the measured endpoint relative humidity, ${RH}_0$ is the measured starting point relative humidity, C is the capacitance value corresponding to the endpoint relative humidity, and $C_0$ is the capacitance value corresponding to the starting point relative humidity. The relative humidity in the constant temperature and humidity chamber is gradually adjusted from 20% RH to 90% RH at intervals of 10% RH, and the capacitance value information of the humidity-sensing element under different relative humidity levels is recorded.

The humidity-sensing element is quickly transferred from a constant temperature and humidity chamber at 20% RH to another chamber at 80% RH (at a temperature of 25 °C), and the process is monitored until the capacitance value of the element stabilizes. The time taken for this process is recorded as the adsorption response time of the humidity-sensing element. Similarly, the element is transferred from a constant temperature and humidity chamber at 80% RH to another chamber at 20% RH (at a temperature of 25 °C), and the time required for the capacitance value to stabilize is recorded as the desorption time of the humidity-sensing element. Over a period of 30 days, the capacitance values of the humidity-sensing element are measured at different relative humidity levels (20% RH, 40% RH, 60% RH, and 80% RH) at 5-day intervals to assess the stability of the humidity-sensing element.

2.7. Detection of Plant VPD by Wearable Sensors

There are various formulae for calculating saturated vapor pressure, each applicable to different temperature intervals, and at room temperature, the Magnus formula is generally used [21]. Combined with the blade VPD_L calculation method [22], the blade VPD_L value at the time of detection is calculated as

(3)$VP{D}_L=V{P}_{sat}-V{P}_{air},$

(4)$V{P}_{sat}=0.61078\times e^{{\displaystyle \frac{17.2694\times T_L}{243.04+T_L}}},$

(5)$V{P}_{air}=0.61078\times e^{{\displaystyle \frac{17.2694\times T_a}{243.04+T_a}}}\times RH,$

where ${VP }_{sat}$ is the saturated vapor pressure inside the blade, and where $T_L$ is the blade temperature (unit of °C). ${VP}_{air}$ is the actual vapor pressure of the air on the blade surface, where T_a is the air temperature (unit of °C), and RH is the relative humidity on the blade surface. The flexible sensor is attached to the lower surface of the green leaf (avoiding the main leaf veins), and the electrode is connected to the LCR tester through a flexible wire to monitor the capacitance signal of the humidity-sensing element and the resistance signal of the temperature sensing element of the green leaf.

3. Results and Discussion

3.1. Characterization of Flexible Wearable Sensors

The interfacial polarization in the PDMS-GO composite material and the principle of micro-capacitors generated due to its unique structure can lead to an increase in the dielectric constant. As the content of GO increases, the hydrogen bonding between GO and PDMS results in a larger interfacial area, providing numerous sites for the enhanced Maxwell–Wagner–Sillars effect. This leads to the formation of many micro-capacitors within the composite material, resulting in an increased dielectric constant and improved electrical properties of the material [23].

As shows in Figure 4a, the XRD (X-Ray diffraction) characterization of the composite material reveals a broad peak appearing at around 10° to 15°, corresponding to the amorphous phase of PDMS [24]. The peak expected for GO, located between 11° and 12°, is not clearly visible, possibly due to the very low weight percentage of GO, which causes this peak to be dominated by the amorphous peak of PDMS.

As shows in Figure 4b, the FTIR (Fourier transform infrared spectrometer) characterization of the composite material indicates that the absorption peaks in the FTIR spectrum change upon the addition of GO to PDMS. The absorption peak corresponding to the epoxy $(\mathrm{C}-\mathrm{O}-\mathrm{C})$ group at $910 {\mathrm{c}\mathrm{m}}^{-1}$ shifts down to $905 {\mathrm{c}\mathrm{m}}^{-1}$ in the GO-PDMS composite. Additionally, the peak intensities of the alkoxy $(\mathrm{C}-\mathrm{O},1255 {\mathrm{c}\mathrm{m}}^{-1})$ and carbonyl $(\mathrm{C}=\mathrm{O},1639 {\mathrm{c}\mathrm{m}}^{-1})$ groups gradually weaken with increasing GO content. The shift of the epoxy group and the decrease in the intensities of the alkoxy and carbonyl peaks are generally considered evidence of hydrogen bonding [25].

As shows in Figure 4c, the XPS (X-Ray photoelectron spectroscopy) characterization of the material reveals the C 1s XPS spectrum of PDMS, which displays four signals corresponding to C, O, and Si at $\mathrm{C}-\mathrm{S}\mathrm{i}\left(284.2 \mathrm{e}\mathrm{V}\right),\mathrm{C}-\mathrm{C}\left(284.8 \mathrm{e}\mathrm{V}\right),\mathrm{C}-\mathrm{O}-\mathrm{C}\left(286.7 \mathrm{e}\mathrm{V}\right), \mathrm{C}=\mathrm{O} (287.6 \mathrm{e}\mathrm{V})$, and $287.6 \mathrm{e}\mathrm{V}$ $(\mathrm{C}=\mathrm{O})$. In the literature, the high-resolution C 1s XPS spectrum of GO shows the $\mathrm{C}-\mathrm{O}-\mathrm{C}$ and $\mathrm{C}=\mathrm{O}$ peaks at $287.1 \mathrm{e}\mathrm{V}$ and $287.8 \mathrm{e}\mathrm{V}$, respectively, with no $\mathrm{C}-\mathrm{S}\mathrm{i}$ peak present [26]. In the composite material, the $\mathrm{C}-\mathrm{O}-\mathrm{C}$ and $\mathrm{C}=\mathrm{O}$ peaks shift to $286.8 \mathrm{e}\mathrm{V}$ and $287.7 \mathrm{e}\mathrm{V}$, respectively, placing them between the peak positions of GO and PDMS. This indicates the presence of non-covalent interactions (hydrogen bonding) between the GO sheets and the PDMS copolymer. Furthermore, the C 1s XPS spectrum of the composite material also shows a $C-Si$ signal, indicating that PDMS has been successfully incorporated into the composite.

Figure 5a shows the SEM (scanning electron microscope) micrograph of the cross-section of the composite material. Figure 5b–d present the silicon mapping, oxygen, and carbon mapping images of the composite material, respectively, indicating that PDMS is present between the GO layers and is evenly dispersed within them.

The film WVRT value is commonly used in the laboratory to define the air permeability of a material [27], which is calculated as

(6)$WVTR(\mathrm{m}\mathrm{g}·\mathrm{c}{\mathrm{m}}^{-2}·{\mathrm{h}}^{-1})={\displaystyle \frac{W}{S\times T}},$

where S is the area, T is the time, and W is the weight change. As shown in Figure 6a, the slope indicates the water vapor permeability of the bottle, revealing that the water vapor permeability of the PDMS-SDS porous polymer film is approximately $19 \mathrm{m}\mathrm{g}·{\mathrm{c}\mathrm{m}}^{-2}·{\mathrm{h}}^{-1}$. To further demonstrate the gas permeability of the fabricated porous polymer film, tests are also conducted on an equivalent thickness of pure PDMS film, which showed that the water vapor permeability of the pure PDMS film without SDS is nearly zero. The water vapor permeability of the empty bottle without any film is approximately $43 \mathrm{m}\mathrm{g}·{\mathrm{c}\mathrm{m}}^{-2}·{\mathrm{h}}^{-1}$. From the above information, it can be concluded that the PDMS-SDS porous polymer film prepared in this experiment exhibits good gas permeability.

For light wavelengths, the range of 400 to 800 nm corresponds to visible light, while the range of 290 to 400 nm corresponds to ultraviolet (UV) light. Figure 6b shows that within the UV range of 290 to 400 nm, the average transmittance of the flexible substrate is approximately 63.48%, and within the visible light range of 400 to 800 nm, the average transmittance is approximately 72.47%. This indicates that the flexible substrate has good transmittance in both the visible and UV regions, demonstrating its excellent transparency under natural conditions. Using a 10 × 10 mm copper detection plate to hold the 0.5 mm thick PDMS-GO-SDS flexible substrate, the average capacitance measured is $10.27\pm 0.03 \mathrm{p}\mathrm{F}$, while the capacitance without any materials held is $1.99\pm 0.02 \mathrm{p}\mathrm{F}$. This yields a dielectric constant of approximately 5.16 for the PDMS-GO-SDS polymer material, indicating that this polymer material possesses good dielectric properties.

3.2. Performance Testing and Model Establishment of Temperature and Humidity-Sensing Elements

The experiment analyzed the changes in the resistance signal of the platinum temperature sensing element under different relative humidity and light intensity conditions. As shown in Figure 7a–c, at temperatures of 10 °C, 25 °C, and 40 °C, the average resistance values of the temperature sensing element were $1038.18$, $1077.79 \Omega$, and $1155.11 \Omega$, respectively, as the relative humidity increased from 20% RH to 90% RH in increments of 10% RH. The resistance change rates were all below 0.07%. This result indicates that the resistance information of the platinum sensing element is basically unaffected by humidity. Regarding light intensity, as shown in Figure 7d–f, at temperatures of 10 °C, 25 °C, and 40 °C, the average resistance values were 1037.96 Ω, 1078.09 Ω, and 1155.19 Ω, respectively, as the light intensity increased from 0 Lux to 10,000 Lux in increments of 2000 Lux. The resistance change rates were all below 0.060%, indicating that the resistance information of the platinum sensing element is largely unaffected by light intensity.

The response performance reflects the ability of the temperature sensing element to detect target information (temperature) during operation. As shown in Figure 8a, the sensitivity of the temperature sensing element is approximately 3.93 Ω/°C. Figure 8b indicates that when the temperature rapidly changes from 10°C to 40 °C, the response time t_r (the time required for the value to rise from 10% to 90% of the steady-state value) is approximately 9.5 s. Figure 8c illustrates the change in resistance values of the element at different temperatures at five-day intervals. Over the 30-day testing period, the resistance variations at 10 °C, 20 °C, 30 °C, and 40 °C are all below 0.036%, indicating that the temperature sensing element exhibits good stability. Figure 8d shows the curve of resistance values of the temperature sensing element as a function of temperature, establishing a linear model between the resistance value of the temperature-sensing element and the temperature. The resulting equation is

T = 0.2558 R − 255.61,(7)

where T represents the temperature, and R is the resistance value of the temperature sensing element. The goodness of fit is approximately 0.9985, indicating a strong correlation between the resistance value of the temperature sensing element and the temperature.

The capacitance values of the elements vary at different detection frequencies. As shown in Figure 9, the lower the detection frequency, the higher the measured capacitance value, resulting in better sensitivity. This is because, at lower frequencies, capacitors exhibit higher resistance, and the internal resistance of the capacitor affects its capacitance, leading to higher measured capacitance values. However, the detection frequency should not be too low, as a very low frequency can prolong the stabilization time of the capacitance. The stabilization times at frequencies of 400 Hz, 700 Hz, 1000 Hz, and 2000 Hz are approximately 5 s, 3 s, 1 s, and 0.9 s, respectively. This phenomenon is due to the polarization effect of the capacitor [28]. Polarization refers to the occurrence of a polarized electric field between the capacitor plates when a voltage is applied, causing charge to move from one electrode plate to the other. Over time, the charges reach a balanced state, and the capacitance stabilizes. Therefore, the polarization effect of the dielectric in the electric field requires some reaction time, known as relaxation time or polarization time. If the polarization time is too long, it can cause delays and instability in capacitance measurements. After careful consideration, a frequency of 1000 Hz will be chosen for subsequent testing.

After determining the detection frequency, the experiment analyzed the effects of temperature, light intensity, and bending degree on the capacitance response of the $200 \mathsf{\mu }\mathrm{m}$ spaced interdigital electrode humidity-sensing element at different humidity levels. Regarding temperature, as shown in Figure 10a–c, when the relative humidity was set at 30% RH, 60% RH, and 90% RH, the capacitance values of the humidity-sensing element increased from $20.15 \mathrm{n}\mathrm{F}$ to $366.77 \mathrm{n}\mathrm{F}$ as the temperature rose from 10 °C to 40 °C, with a capacitance change rate not exceeding 0.85%. This indicates that the capacitance signal of the interdigital electrode humidity-sensing element is essentially unresponsive to temperature. Figure 10d–f display the capacitance variations of the humidity-sensing element under different light intensities. The change rates in capacitance remained below 1.18%, indicating that the capacitance signal of the humidity-sensing element of the interdigital electrode has basically no response to the light intensity.

Figure 11 shows the curves of the capacitance values of humidity-sensing elements with different substrates as humidity changes. Under the same humidity conditions, the humidity-sensing elements made from two types of flexible substrates exhibited different sensitivities. The humidity-sensing element with a 200 μm electrode spacing on the PDMS-GO-SDS flexible substrate showed the highest sensitivity, i.e., approximately 4456 pF/% RH, while the sensitivity of the humidity-sensing element on the PDMS flexible substrate with the same electrode spacing is about 1436 pF/% RH. This difference is attributed to the microstructure of GO, which has a larger specific surface area. The addition of GO improved the electrical properties of PDMS and increased the number of binding sites, thereby enhancing water adsorption [29,30].

Additionally, at high relative humidity, the increase in capacitance is more pronounced compared to low relative humidity. This is because, at low relative humidity, water molecules primarily adsorb to the available active sites on the surface of graphene oxide (GO) through double hydrogen bonding. In this region, water molecules are constrained by double hydrogen bonds and cannot move freely, resulting in minimal capacitance variation. As relative humidity increases, water molecules are physically absorbed through single hydrogen bonds on the hydroxyl groups, allowing them to become mobile and gradually similar to water molecules in the liquid phase. The physically adsorbed water undergoes ionization under the influence of an electric field, generating a large number of hydrated hydrogen ions as charge carriers. At high relative humidity, the physically adsorbed water penetrates into the intermediate layers of the graphene oxide film, facilitating the hydrolysis of functional groups on the graphene oxide sheets [31,32]. Therefore, in high humidity conditions, the capacitance increases sharply.

As shown in Figure 12a, the response curve illustrates the variation in capacitance signal over time during rapid transitions in relative humidity from 20% RH to 80% RH. The results indicate that the response time t_r (the time required for the steady-state value to change from 10% to 90%) is approximately 12.5 s. Figure 12b displays the variation rates of capacitance values measured under 20% RH, 40% RH, 60% RH, and 80% RH over a 30-day testing period, all of which do not exceed 1.3%. This indicates that the humidity-sensing element possesses excellent long-term stability for use. During the bending process, where the bending degree changes from 0% to 30%, the variation rate of capacitance for the humidity-sensing element does not exceed 0.73%, suggesting that the capacitance information of the interdigital electrode humidity-sensing element is essentially unaffected by changes in bending degree. In summary, we selected the humidity-sensing element with a flexible PDMS-GO-SDS substrate and an electrode spacing of 200 μm. Under a detection frequency of 1000 Hz, we established a model for the relationship between the capacitance value of the humidity-sensing element and relative humidity. The model equation is given by

(8)$RH=-80.822·e^{\left({\displaystyle \frac{-C}{79.465}}\right)}+89.196,$

where RH represents the relative humidity, and C denotes the capacitance value of the humidity-sensing element. The goodness of fit for the model is approximately 0.9952, indicating a strong correlation between the capacitance value of the humidity-sensing element and relative humidity.

3.3. Monitoring of Plant Leaf Transpiration Using Flexible Wearable Sensors

Commercial sensors (Plant Nutrition Tester, model TYS-4N) and wearable sensors (with humidity- and temperature-sensing elements on the front) are used to monitor the temperature and humidity of tomato leaves. The commercial sensor performs measurements by clamping the leaf between two test clips, while the wearable sensor measures directly by attaching to the back of the leaf. Figure 13a,b show the monitoring results of leaf temperature and humidity from both sensors. The average error rate for leaf temperature monitoring is approximately 1.13%, while the average error rate for leaf humidity monitoring is 2.08%. This indicates that the wearable sensor can accurately monitor the temperature and humidity information on the leaf surface. Figure 13c displays the monitoring results for air temperature in tomatoes, using a commercial sensor (gas thermometer, model 0–50) and a wearable sensor (featuring an environmental temperature sensing element on the back). The average error rate for air temperature measurements between the wearable sensor and the commercial sensor is approximately 1.55%, indicating that the wearable sensor can accurately detect the air temperature information in a small environment.

Twenty tomato plants were tested in a controlled greenhouse, eliminating mechanical disturbances caused by wind and rain. The flexible wearable sensor was attached to the back of the tomato leaves (Figure 13d) to monitor the leaf VPD_L values from 6:00 a.m. to 11:00 p.m., as shown in Figure 14. During the daytime, as time progressed, the air temperature and solar radiation gradually increased, enhancing the plant’s transpiration and causing the VPD_L values to rise. Notably, in the initial phase of the day, the air temperature gradually increased to about 21 °C, and the average temperature of the leaves first rose to a level close to the air temperature before dropping to about 20.2 °C. The cooling of the leaves may be related to heat dissipation through transpiration. After 4:00 p.m., as light intensity decreased, photosynthesis and transpiration gradually stopped, leading to changes in the leaf temperature and surface relative humidity, resulting in a decrease in VPD_L. This monitoring result is consistent with existing studies [17], indicating that the flexible wearable sensor developed in this study for in situ non-destructive detection of leaf-scale VPD_L has good practicality.

4. Conclusions

In this work, a flexible substrate with excellent biocompatibility was fabricated based on PDMS-GO-SDS flexible materials through improved processing techniques. Additionally, a humidity-sensing element with high sensitivity and strong interference resistance was developed by integrating graphene oxide. According to relevant research conclusions, the leaf-scale VPD_L can more accurately reflect the true transpiration state of plants compared to the environmental scale VPD_A, which highlights the need for detecting information on leaf temperature, surface relative humidity, and microenvironment temperature when acquiring leaf-scale VPD_L. To address the issues of existing plant information sensors being functionally limited, not lightweight, and prone to damage, a flexible wearable sensor for in situ non-destructive detection of plant transpiration information was designed. This sensor features multifunctionality, compactness, flexibility, and high sensitivity. This study provides a new technical approach for the quantitative monitoring of physiological parameters throughout the crop lifecycle, contributing to the establishment of intelligent planting systems.

Author Contributions

H.M. provided project management and financial support; Z.L. (Zhikang Li) designed experiments and wrote papers; L.L. provided technical guidance; Y.W., Y.Y., M.Z. and Z.L. (Ze Liu) provided assistance during the experiment. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

The data presented in this study are available in the article.

Conflicts of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures

Figure 1. (a) Schematic diagram of the flexible substrate preparation process. (b) Schematic diagram of the flexible substrate after curing. (c) Printing process of the interdigitated electrode.

Figure 2. Schematic diagram of the flexible wearable sensor.

Figure 3. Schematic of deformation of flexible substrate under different degrees of bending.

Figure 4. (a) XRD characterization image of the PDMS-GO composite material. (b) FTIR characterization image of the composite material. (c) XPS characterization image of pure PDMS and the composite material.

Figure 5. (a) SEM image of the cross-section of the composite material. (b) Silicon mapping image of the composite material. (c) Oxygen mapping image of the composite material. (d) Carbon mapping image of the composite material.

Figure 5. (a) SEM image of the cross-section of the composite material. (b) Silicon mapping image of the composite material. (c) Oxygen mapping image of the composite material. (d) Carbon mapping image of the composite material.

Figure 6. (a) Permeability testing of the adhesive film. (b) UV transmittance of the PDMS-GO-SDS flexible substrate.

Figure 7. Influence of humidity on the resistance information of the temperature sensing element at (a) 10 °C, (b) 25 °C, and (c) 40 °C; influence of light intensity on the resistance information of the temperature sensing element at (d) 10 °C, (e) 25 °C, and (f) 40 °C.

Figure 8. (a) Relationship between the resistance value of the temperature sensing element and temperature. (b) Response curve of the temperature sensing element during rapid temperature changes from 10 °C to 40 °C. (c) Temperature response graph of the temperature sensing element over a long period. (d) Fitting of the resistance value R of the temperature sensing element with temperature T.

Figure 9. Capacitance response of the humidity-sensing element at 40% RH (normal relative humidity) under different frequencies.

Figure 10. Capacitive response of the humidity-sensing element under the influence of temperature at (a) 30% RH, (b) 60% RH, and (c) 90% RH; capacitive response of the humidity-sensing element under the influence of light intensity at (d) 30% RH, (e) 60% RH, and (f) 90% RH.

Figure 10. Capacitive response of the humidity-sensing element under the influence of temperature at (a) 30% RH, (b) 60% RH, and (c) 90% RH; capacitive response of the humidity-sensing element under the influence of light intensity at (d) 30% RH, (e) 60% RH, and (f) 90% RH.

Figure 11. The capacitance values of humidity-sensing elements using PDMS-GO-SDS and pure PDMS as flexible substrates vary with relative humidity.

Figure 12. (a) Response curve of the humidity-sensing element during rapid changes from 20% RH to 80% RH. (b) Detection stability of the humidity-sensing element under long-term storage conditions. (c) Fitting of capacitance value C of the humidity-sensing element with relative humidity RH.

Figure 13. (a) Leaf temperature curves measured by the wearable sensor and the commercial sensor. (b) Leaf surface relative humidity curves measured by the wearable sensor and the commercial sensor. (c) Leaf small environmental air temperature curves measured by the wearable sensor and the commercial sensor. (d) The wearable sensor detecting tomatoes in a real scene.

Figure 13. (a) Leaf temperature curves measured by the wearable sensor and the commercial sensor. (b) Leaf surface relative humidity curves measured by the wearable sensor and the commercial sensor. (c) Leaf small environmental air temperature curves measured by the wearable sensor and the commercial sensor. (d) The wearable sensor detecting tomatoes in a real scene.

Figure 14. (a) Detection of leaf and surrounding microenvironment temperature by the wearable sensor. (b) Detection of surface relative humidity of the leaf by the wearable sensor. (c) Detection of leaf VPDL.

References

1. Goo, H.; Roh, Y.; Lee, J.; Park, K.S. Analysis of bell pepper (Capsicum annuum L.) leaf spectral properties and photosynthesis according to growth period. Horticulturae; 2024; 10, 646. [DOI: https://dx.doi.org/10.3390/horticulturae10060646]

2. Yousefi, M.; Dehghanpour, H. A model and measurement technique for liquid permeability oftight porous media based on the steady-state method. Energy Fuels; 2022; 36, pp. 6860-6867. [DOI: https://dx.doi.org/10.1021/acs.energyfuels.2c01160]

3. Dixon, L.; Bellinger, B.; Carter, A.H. A gravimetric method to monitor transpiration under water stress conditions in wheat. Plant Phenome J.; 2023; 6, e20078. [DOI: https://dx.doi.org/10.1002/ppj2.20078]

4. Kulmatiski, A.; Forero, L.E. Bagging: A cheaper, faster, non-destructive transpiration water sampling method for tracer studies. Plant Soil; 2021; 462, pp. 603-611. [DOI: https://dx.doi.org/10.1007/s11104-021-04844-w]

5. Yan, H.; Li, M.; Zhang, C.; Zhang, J.; Wang, G.; Yu, J.; Ma, J.; Zhao, S. Comparison of evapotranspiration upscaling methods from instantaneous to daytime scale for tea and wheat in southeast china. Agric. Water Manag.; 2022; 264, 107464. [DOI: https://dx.doi.org/10.1016/j.agwat.2022.107464]

6. Nietupski, M.; Ludwiczak, E.; Olszewski, J.; Gabryś, B.; Kordan, B. Effect of aphid foraging on the intensity of photosynthesis and transpiration of selected crop plants in its early stages of growing. Agronomy; 2022; 12, 2370. [DOI: https://dx.doi.org/10.3390/agronomy12102370]

7. Afzal, A.; Duiker, S.W.; Watson, J.E.; Luthe, D. Leaf thickness and electrical capacitance as measures of plant water status. Trans. ASABE; 2017; 60, pp. 1063-1074. [DOI: https://dx.doi.org/10.13031/trans.12083]

8. Xing, D.; Wang, W.; Wu, Y.; Qin, X.; Li, M.; Chen, X.; Yu, R. Translocation and utilization mechanisms of leaf intracellular water in karst plants Orychophragmus violaceus (L.) o. E. Schulz and Brassica napus L. Horticulturae; 2022; 8, 1082. [DOI: https://dx.doi.org/10.3390/horticulturae8111082]

9. Yu, R.; Wu, Y.; Xing, D. The differential response of intracellular water metabolism derived from intrinsic electrophysiological information in Morus alba L. and Broussonetia papyrifera (L.) vent. Subjected to water shortage. Horticulturae.; 2022; 8, 182. [DOI: https://dx.doi.org/10.3390/horticulturae8020182]

10. Song, P.; Tao, J.; He, X.; Sun, Y.; Shen, X.; Zhai, L.; Yuan, A.; Zhang, D.; Ji, Z.; Li, B. Silk-inspired stretchable fiber-shaped supercapacitors with ultrahigh volumetric capacitance and energy density for wearable electronics. Chem. Eng. J.; 2020; 386, 124024. [DOI: https://dx.doi.org/10.1016/j.cej.2020.124024]

11. Ge, D.; Babangida, A.A.; Hu, Z.; Zhang, L.; Wang, M. Flexible pressure sensor based on a thermally induced wrinkled graphene sandwich structure. IEEE Sens. J.; 2022; 22, pp. 3040-3051. [DOI: https://dx.doi.org/10.1109/JSEN.2021.3130445]

12. Di Tocco, J.; Lo Presti, D.; Massaroni, C.; Cinti, S.; Cimini, S.; De Gara, L.; Schena, E. Plant-wear: A multi-sensor plant wearable platform for growth and microclimate monitoring. Sensors; 2023; 23, 549. [DOI: https://dx.doi.org/10.3390/s23010549] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36617147]

13. Lu, Y.; Yang, G.; Shen, Y.; Yang, H.; Xu, K. Multifunctional flexible humidity sensor systems towards noncontact wearable electronics. Nano-Micro Lett.; 2022; 14, 150. [DOI: https://dx.doi.org/10.1007/s40820-022-00895-5] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35869398]

14. Li, D.; Li, G.; Li, J.; Xu, S. Wearable crop sensor based on nano-graphene oxide for noninvasive real-time monitoring of plant water. Membranes; 2022; 12, 358. [DOI: https://dx.doi.org/10.3390/membranes12040358] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35448328]

15. Tao, B.; Yin, J.; Miao, F.; Zang, Y. High-performance humidity sensor based on Go/ZnO/plant cellulose film for respiratory monitoring. Ionics; 2022; 28, pp. 2413-2421. [DOI: https://dx.doi.org/10.1007/s11581-022-04478-7]

16. Li, S.; Wan, T.; Wei, H.; Wang, S.; Wang, B.; Cheng, B. Flexible highly-sensitive humidity sensor based on CGO/SMPLAF for wearable human skin humidity detection. Sens. Actuators B Chem.; 2022; 362, 131806. [DOI: https://dx.doi.org/10.1016/j.snb.2022.131806]

17. Yin, S.; Ibrahim, H.; Schnable, P.S.; Castellano, M.J.; Dong, L. A field-deployable, wearable leaf sensor for continuous monitoring of vapor-pressure deficit. Adv. Mater. Technol.; 2021; 6, 2001246. [DOI: https://dx.doi.org/10.1002/admt.202001246]

18. Wright, A.J.; Francia, R.M. Plant traits, microclimate temperature and humidity: A research agenda for advancing nature-based solutions to a warming and drying climate. J. Ecol.; 2024; 112, pp. 2462-2470. [DOI: https://dx.doi.org/10.1111/1365-2745.14313]

19. Li, R.; Zeng, F.; Zhao, Y.; Wu, Y.; Niu, J.; Wang, L.; Gao, N.; Zhou, H.; Shi, X.; Huang, Z. Analyzing the impact of various factors on leaf surface temperature based on a new tree-scale canopy energy balance model. Sustain. Cities Soc.; 2023; 99, 104994. [DOI: https://dx.doi.org/10.1016/j.scs.2023.104994]

20. Lo Presti, D.; Di Tocco, J.; Massaroni, C.; Cimini, S.; De Gara, L.; Singh, S.; Raucci, A.; Manganiello, G.; Woo, S.L.; Schena, E. et al. Current understanding, challenges and perspective on portable systems applied to plant monitoring and precision agriculture. Biosens Bioelectron; 2023; 222, 115005. [DOI: https://dx.doi.org/10.1016/j.bios.2022.115005]

21. Pamuta, H. Vapor Pressure of Water Calculator. Available online: https://www.omnicalculator.com/chemistry/vapour-pressure-of-water (accessed on 11 October 2024).

22. Grossiord, C.; Buckley, T.N.; Cernusak, L.A.; Novick, K.A.; Poulter, B.; Siegwolf, R.T.W.; Sperry, J.S.; McDowell, N.G. Plant responses to rising vapor pressure deficit. New Phytol.; 2020; 226, pp. 1550-1566. [DOI: https://dx.doi.org/10.1111/nph.16485] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32064613]

23. Borah, B.; Rajitha, G.; Dash, R.K. Correlation between the thickness and properties of the ethanol treated GO–PDMS based composite materials. J. Mater. Sci. Mater. Electron.; 2018; 29, pp. 20216-20224. [DOI: https://dx.doi.org/10.1007/s10854-018-0154-2]

24. Zhang, Y.; Zhu, Y.; Lin, G.; Ruoff, R.S.; Hu, N.; Schaefer, D.W.; Mark, J.E. What factors control the mechanical properties of poly (dimethylsiloxane) reinforced with nanosheets of 3-aminopropyltriethoxysilane modified graphene oxide?. Polymer; 2013; 54, pp. 3605-3611. [DOI: https://dx.doi.org/10.1016/j.polymer.2013.04.057]

25. Chen, K.; Shi, B.; Yue, Y.; Qi, J.; Guo, L. Binary synergy strengthening and toughening of bio-inspired nacre-like graphene oxide/sodium alginate composite paper. ACS Nano; 2015; 9, pp. 8165-8175. [DOI: https://dx.doi.org/10.1021/acsnano.5b02333]

26. Song, S.; Zhai, Y.; Zhang, Y. Bioinspired graphene oxide/polymer nanocomposite paper with high strength, toughness, and dielectric constant. ACS Appl. Mater. Interfaces; 2016; 8, pp. 31264-31272. [DOI: https://dx.doi.org/10.1021/acsami.6b08606]

27. Yin, X.; Zhang, J.; Xu, J.; Tian, M.; Li, L.; Tan, L.; Li, Z. Fast-acting and highly rechargeable antibacterial composite nanofibrous membrane for protective applications. Compos. Sci. Technol.; 2021; 202, 108574. [DOI: https://dx.doi.org/10.1016/j.compscitech.2020.108574]

28. Mosgaard, L.D.; Zecchi, K.A.; Heimburg, T. Mechano-capacitive properties of polarized membranes. Soft Matter; 2015; 11, pp. 7899-7910. [DOI: https://dx.doi.org/10.1039/C5SM01519G]

29. Lv, C.; Hu, C.; Luo, J.; Liu, S.; Qiao, Y.; Zhang, Z.; Song, J.; Shi, Y.; Cai, J.; Watanabe, A. Recent advances in graphene-based humidity sensors. Nanomaterials; 2019; 9, 422. [DOI: https://dx.doi.org/10.3390/nano9030422]

30. Bi, H.; Yin, K.; Xie, X.; Ji, J.; Wan, S.; Sun, L.; Terrones, M.; Dresselhaus, M.S. Ultrahigh humidity sensitivity of graphene oxide. Sci. Rep.; 2013; 3, 2714.

31. Wang, Z.; Shi, L.; Wu, F.; Yuan, S.; Zhao, Y.; Zhang, M. The sol-gel template synthesis of porous TiO₂ for a high performance humidity sensor. Nanotechnology; 2011; 22, 275502. [DOI: https://dx.doi.org/10.1088/0957-4484/22/27/275502] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21597157]

32. Fatima, Q.; Haidry, A.A.; Yao, Z.; He, Y.; Li, Z.; Sun, L.; Xie, L. The critical role of hydroxyl groups in water vapor sensing of graphene oxide. Nanoscale Adv.; 2019; 1, pp. 1319-1330. [DOI: https://dx.doi.org/10.1039/C8NA00135A] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36132612]