1. Introduction
“Wearable” plant sensors for monitoring plant health have been widely studied [
1]. They can be classified into plant physiology and microclimate sensors [
2], [
3], [
4], [
5], [
6], [
7], plant growth sensors [
8], [
9], [
10], and chemical sensors [
11], [
12], [
13]. Most of these sensors monitor physiological signals occurring in the later stages of stress response, so this monitoring is not timely and does not fully exploit the potential of wearable plant sensors. Timely plant health monitoring requires wearable sensors that can analysis biomarkers and indicators generated in earlier stages. Reactive oxygen species (ROS) and hormones are key signaling molecules generated in the early stages of plant stress response [
14]. ROS can pass through plant cells during system signal transduction, switching different cells and tissues from a normal growth state into a stress state [
1], [
15], [
16], [
17]. For example, when plants suffer from drought stress, ROS act as a second messenger to promote a change in the abscisic acid content in the plants, further leading to stomatal closure in the leaves [
18]. Hydrogen peroxide (H
2O
2) is a primary ROS that mediates rapid systemic signals in plants, due to its long half-life and its capability for long-distance and transmembrane transmission [
19], [
20]. However, the various molecular mechanisms related to H
2O
2 have not been fully revealed. Accurate and real-time monitoring of H
2O
2 in plants is crucial in studying plant stress response mechanisms and plant health status [
21].
At present, there are two main H
2O
2 detection methods in plants: ① tissue slice staining [
22], [
23] and ② spectrophotometry [
17], [
24]. It is easy to miss the best detection window using these methods, and the results are reliant on human operation; therefore, these methods cannot meet the requirements of high-time resolution and continuous monitoring. In addition, the wounds formed by sampling are easily infected by bacteria, causing plant diseases. In recent years, the development of implantable sensors has provided promising access to the sensing of real-time health information in plants. In particular, Wu et al. [
19] and Lew et al. [
20] have developed a series of optical nanosensors that can be injected into plants to monitor real-time molecular information, providing dynamic, spatial, and temporal biochemical measurements in plants. Although these studies were conducted in the laboratory, they provide novel analytical tools for studying plant physiology [
21]. H
2O
2 is a typical electroactive molecule, and the application of invasive electrochemical sensors in animal biomarker monitoring suggests a new approach for monitoring H
2O
2 in plants [
25], [
26], [
27]. Therefore, implantable electrochemical sensors are expected to be developed for the
in vivo analysis of H
2O
2 in plants. However, for electrochemical sensors implanted into plants, the power lines are not conducive to the stable fixation of the sensor, and the batteries require continuous charging or frequent battery replacement. These power supply issues limit the application of implantable wearable plant sensors [
28]. Therefore, it is necessary to find an alternative power supply method for electrochemical sensors implanted into plants.
Here, we develop an implantable and self-powered sensing system for the continuous
in vivo monitoring of dynamic H
2O
2 in tomato plants. The integrated system consists of a microsensor, a data-acquisition and -transmission module, and a photovoltaic (PV) module. The system can sustainably transmit the sensor’s signal over a long distance to a multi-channel monitoring interface for visual analysis (
Fig. 1(a)). The microsensor, which is based on three-dimensional (3D) porous laser-induced graphene (LIG), is the core of the sensing system; it can be inserted into a plant stem (a tomato plant in our case) to obtain real-time H
2O
2 information (
Fig. 1(b)). To solve the problem of a continuous power supply, a PV module was designed and integrated with the data-acquisition and -transmission module and collects sunlight or artificial light in the planting environment to continuously power the microsensor (
Fig. 1(c)). Using this self-powered system, the transmission process of the H
2O
2 signal was monitored
in vivo, and the time and concentration specificity of the H
2O
2 signal for abiotic stress were investigated.
2. Materials and methods
2.1. Materials and instruments
Polyimide (PI) film (80 μm thick) and PI tapes (50 μm thick) were purchased from DuPontTM (USA). Ag/AgCl ink (E2414) was purchased from Ercon, Inc. (USA). K2PtCl4 and potassium iodide (KI) were purchased from Sigma Aldrich (USA). Anhydrous sodium sulfate (Na2SO4) and polyethylene glycol (PEG)-6000 were purchased from Macklin Inc. (China). Trichloroacetic acid (TCA) was obtained from Fuchen Chemical Reagent Co., Ltd. (China). H2O2 (30%) was purchased from Sinopharm Chemical Reagent Co., Ltd. (China). Phosphate-buffered saline (PBS; 0.1 mol∙L−1) was purchased from Sangon Bioengineering Co., Ltd. (China), and 5% Nafion solution was purchased from DuPont Co., Ltd.
LIG was prepared by means of a computer-controlled laser scribing micromachining system with a 10.6 µm CO2 laser (50 W, PLS 6MW; Universal Laser Systems, USA). The morphology characterizations were obtained via scanning electron microscopy (SEM; SU8010; Hitachi, Japan), and the graphene structure was characterized by a Raman spectrometer (LabRAM HR; HORIBA Jobin Yvon, Japan). X-ray diffraction (XRD) was carried out using a D8 advance diffractometer (Bruker, Germany). Electrochemical characterizations were performed by a potentiostat (Sens 3; PalmSens, Netherlands). Ultraviolet (UV) spectroscopy was carried out by means of a microplate reader (BioTek, France). Deionized water was obtained using a Milli-Q purification system (Millipore, USA).
2.2. Preparation of LIG microelectrodes
The PI film was ultrasonically cleaned with ethanol and deionized water for 10 min each. The cleaned PI film was fixed onto the surface of a copper plate, and laser induced graphene sheet (LIGS) was induced on the PI film by the laser engraving machine. The implanted part of the microelectrode was 3 mm wide and 3 mm long. The diameter of the circular work area was 0.75 mm, and the widths of both the reference electrode and the counter electrode were 0.2 mm. The optimal laser power, velocity, and image density were 8%, 14%, and 750 pulses per inch (PPI), respectively. A mask layer (50 μm thick) made by laser cutting PI tape was laid on the surface of the PI substrate to protect the microelectrode from damage to the sensing layer when the microelectrode was pierced into a tomato stem. All laser engraving experiments were conducted in an air environment. A reference electrode was obtained by coating the reference area of the LIG with Ag/AgCl ink.
2.3. Preparation of microsensor and the electrochemical sensing method
An integrated three-electrode system based on LIGS was prepared according to the above steps. Before the modification of the sensing interface, the prepared electrode was electroactivated using the cyclic voltammetry (CV) method within five cycles in PBS electrolyte (with a voltage ranging from +0.3 to +1.4 V and a scan rate of 100 mV∙s
-1). Then, 5 μL of K
2PtCl
4 solution (2 mmol∙L
−1, dissolved in 0.1 mol∙L
−1 Na
2SO
4) was added to the three-electrode reaction region [
29]. Platinum nanoparticles (PtNPs) were electrodeposited onto the interface of the LIGS via the CV method within seven cycles in a voltage range from −0.4 to +0.5 V to obtain the LIGS/PtNPs microelectrode (with a scan rate of 50 mV∙s
-1). Next, 0.4 μL of 0.2% Nafion solution was added onto the working electrode and the reference electrode to obtain the microsensor (LIGS/PtNPs/Nafion).
An electrochemical sensing method for H2O2 was established using a potentiostat; the electrochemical behavior of H2O2 was characterized by CV, and the detection potential of H2O2 was optimized by means of chronoamperometry (CA).
2.4. In vitro simulation
The purpose of the
in vitro simulation was to simulate the intravital conditions of the tomato plant as much as possible and to investigate the feasibility of the microsensor for
in vivo sensing. A tomato stem was cut off at 5 cm from the root, and the wound was covered with absorbent cotton and sealed with a plastic bag to prevent the fresh bleeding sap from evaporation [
30]. The bleeding sap adsorbed by the cotton was squeezed out, collected in a centrifuge tube, and frozen at −20 °C for later use. Before use, the bleeding sap was filtered with a 0.22μm membrane filter to remove impurities [
31] and then heated at 90 °C for half an hour to inactivate enzymes. H
2O
2 solutions were prepared with the treated bleeding sap as the solvent and were analyzed using the prepared microsensor.
2.5. In vivo monitoring
The tomato plants used for the osmotic stress testing were obtained through soilless cultivation, while the other tomato plants were cultivated in nutrient soil. All plants were cultivated in an environmentally controlled plant factory with a temperature of 26 °C, relative humidity (RH) of about 70%, and light intensity of 15 klx. Tomato plants with a height of about 30 cm were selected for monitoring. A 2 mm incision was carefully made with a scalpel on the tomato stem, and the microsensor was inserted into the incision to complete the implantation. Osmotic stress was induced by adding 30% PEG to the hydroponic nutrient solution. The tomato stem was clipped with hemostatic forceps to produce mechanical injury stress. Ultraviolet (UV) stress for the tomato plants was induced using UV lamps (λ = 368 nm).
2.6. Detection of H2O2 in tomato plants via UV spectroscopy
A tomato stem frozen in liquid nitrogen was ground immediately. Then 0.2 mL of TCA (0.1%) solution and 0.8 mL of KI (1 mol∙L
−1) solution were added to 200 mg of the crushed stem and mixed well with a vortex mixer. The mixture was centrifuged at 12 000
g for 15 min. Then, 200 μL of supernatant was put in a 96-well plate to record the absorbance at 350 nm [
32], [
33].
3. Results and discussion
3.1. Sensing system design
The core of the system is the LIG-based microsensor, which is composed of three layers: a PI substrate, a conductive and sensing layer, and a PI mask layer (
Fig. 1(d); Fig. S1(a) in Appendix A). The conductive and sensing layer is made of 3D porous LIG, which is prepared on the surface of PI film using CO
2 LIG technology [
13], [
34]. The 3D porous LIG provides a conductive substrate with a large surface area for the modification of PtNPs with electrocatalytic properties. A circular working area with a diameter of 500 μm is covered by electrodeposited PtNPs as an electrochemical sensing layer and a Nafion membrane as an anti-interference layer. To prevent mechanical damage to the sensing layer when the microsensor is inserted into the stem, a mask layer (50 μm thick) made by laser cutting PI tape is used to cover the surface of the PI substrate (Fig. S1(b) in Appendix A).
Traditional wearable plant sensors are powered by wired power sources or batteries. However, the presence of a large number of power lines and frequent battery replacements are not conducive to the continuous and stable operation of wearable plant sensors. Thus, providing a sustainable power supply without power lines or the need for battery replacement is both a crucial and challenging issue in the design of wearable plant sensors [
1]. Sufficient light conditions in the crop-growing environment provide the possibility of a PV power supply. Therefore, a PV power supply module was designed, fabricated, and then combined with a data-acquisition and -transmission module to form a self-powered sensing system (
Fig. 1(e)). The PV module, which has an output voltage of 6.0 V, can charge a 3.0 V lithium-ion (Li) battery by harvesting the available light energy in the agricultural environment. As shown in Fig. S2(a) in Appendix A, the red indicator light on the circuit board shows that the PV module can power the battery even under artificial light conditions (15 klx). In the data-acquisition module, the bias potential required for electrochemical measurements is derived from the voltage generated by the digital-to-analog converter (DAC) in the chip and the external reference voltage. The current of the microsensor is amplified and converted into a voltage signal through a transimpedance amplifier (TIA), and the voltage signal is converted into a digital voltage value after being read by the analog-to-digital converter (ADC) inside the microcontroller. In the data-transmission module, electrochemical data can be transmitted to the user interface by a long range radio (LoRa) network within 1000 m. Finally, the received voltage value is converted into the current value via software. Figs. S2(b) and (c) in the Appendix A illustrates the design of the whole electronic circuit and provides a picture of the entire system used for H
2O
2 monitoring in tomato plants.
3.2. Characterizations
As an electrochemical sensing platform, the 3D porous LIG provides a conductive substrate with a large surface area for the modification of noble-metal nanoparticles with electrocatalytic activity toward H
2O
2 and plays a crucial role in the sensitive detection of H
2O
2 [
35], [
36], [
37]. By adjusting the laser power and scanning rate, LIG morphologies with fibers (LIGF), LIGS, and sheets mixed with fibers (LIGSF) can be obtained (
Fig. 2(a); Fig. S3 in Appendix A) [
38], [
39], [
40].
We found that more uniform PtNPs could be electrodeposited on the surface of LIGS (
Fig. 2(b); Fig. S3). To explain this difference, LIGF, LIGS, and LIGSF were characterized using Raman spectroscopy and XRD. As shown in
Fig. 2(d), the main differences according to Raman spectroscopy are the 2D band at 2658 cm
-1, which corresponds to 3D multilayer graphene structures, and the D′ band at 1616 cm
-1, which corresponds to the edge defects of graphene. The LIGS have a higher 2D band and a more obvious D′ band than the LIGSF and LIGF, which indicates that the presence of fiber reduces the lamellar structure and edge defects of the graphene. As shown in
Fig. 2(e), the LIGS have the lowest (100) peak at 43.1°, which reveals that the LIGS have a high-quality graphene crystal structure. Since electrochemical electron transfer commonly occurs at the edge defects, the presence of fibers in LIGF may induce a slower heterogeneous electron transfer rate, which is not conducive to the electrodeposition of PtNPs. The electrochemical performance of the LIG microelectrodes electrodeposited with PtNPs also confirmed that there were more PtNPs on the LIGS (Fig. S4 in Appendix A). Therefore, LIGS were chosen as the modification platform of PtNPs. Interestingly, we found that the electrochemical performance of LIGS could be improved by means of a simple electrochemical activation treatment (Fig. S5 in Appendix A), and the best electrochemical performance was achieved when the LIGS were electroactivated in PBS using CV within five cycles in a voltage range from +0.3 to +1.4 V. This phenomenon can be attributed to the fact that electrochemical activation can remove organic impurities at the LIGS surface to improve the electron transfer rate. In addition, we investigated the electrodeposition conditions of the PtNPs, including the electrolyte and the number of electrodeposition cycles, by means of SEM and electrochemical characterizations. Based on the SEM (Fig. S6 in Appendix A) and electrochemical (Fig. S7 in Appendix A) results, the optimal electrodeposition conditions for the PtNPs on LIGS are the CV method within seven cycles in Na
2SO
4 electrolyte.
After preparing the PtNPs on the surface of the LIGS (denoted as LIGS/PtNPs), an anti-interference layer of Nafion film was coated onto the surface of the LIGS/PtNPs (denoted as LIGS/PtNPs/Nafion) to obtain a microsensor that could be implanted. The Nafion film, which has a negative charge, repels anionic interferents, and its self-adhesion enhances the stability of the interface [
41], [
42]. Prior to investigating the detection performance of the LIGS/PtNPs/Nafion microsensor, we examined the electrocatalytic behavior of the microsensor toward H
2O
2. As shown in
Fig. 2(f), the CV curves for different concentrations of H
2O
2 indicate that H
2O
2 undergoes electrochemical oxidation at +0.6 V and electrochemical reduction at −0.4 V. To avoid interference from dissolved oxygen, the oxidation potential was selected as the applied potential for the CA method [
36]. The CA responses of H
2O
2 under different potentials were investigated, and the highest response value was obtained at +0.6 V (Fig. S8 in Appendix A). Therefore, +0.6 V was selected as the optimal potential for the CA method. As shown in
Figs. 2(g) and
(h), under the optimal detection conditions, the LIGS/PtNPs/Nafion microsensor exhibited a good linear relationship with H
2O
2 in the range of 2-200 μmol∙L
−1. The limit of detection (LoD) was calculated to be 0.35 μmol∙L
−1, according to the equation LoD = 3Sb/
m, where Sb and
m represent the standard deviation of the blank signals and the slope of the calibration curve, respectively.
3.3. Anti-interference performance and repeatability
The anti-interference performance of the microsensor against chemical interferents was investigated. The H
2O
2 produced in plants can cause an increase in the content of ascorbic acid (AsA), which has antioxidant properties. AsA has a wide potential window and is the main chemical interferent in the electrochemical detection of H
2O
2 (Fig. S9(a) in Appendix A) [
43]. As shown in Fig. S9(b) in Appendix A, the LIGS/PtNPs/Nafion microsensor exhibited no response to AsA (50 μmol∙L
−1), while maintaining the highest response to H
2O
2 (Fig. S9(c) in Appendix A). In addition, K
+, Na
+, Ca
2+, glucose (Glu), and sucrose (Suc) were selected as interferents, and the response of the microsensor to the individual interferents and to a mixture of these interferents was investigated. As shown in
Figs. 3(a) and
(b), the interference had little effect on the response of the microsensor. Five microsensors were selected to detect the same concentration of H
2O
2, and the relative standard deviation (RSD) of the response was 3.5% (
Fig. 3(c)), indicating satisfactory repeatability.
It has been reported that the pH of tomato plants is commonly in the range of 5.0-8.0 [
44]. Therefore, it was necessary to investigate the effect of pH fluctuation on the response of the microsensor. In the range of pH 5.0-8.0, with an interval of 0.5, the CA responses of the LIGS/PtNPs/Nafion microsensor to H
2O
2 with a concentration of 100 μmol∙L
−1 were investigated. As shown in
Fig. 3(d), the response fluctuate of the LIGS/PtNPs/Nafion microsensor fell within a certain range (RSD = 6.2%). The pH tolerance of the microsensor can be attributed to the Nafion layer on the Ag/AgCl reference electrode, as a bare Ag/AgCl reference electrode can be affected by the pH, resulting in a shift in the oxidation potential. The use of a Nafion film has been demonstrated to reduce the potential shift to 13 mV over a pH range of 4.0-10.0 [
45].
Normal plant growth is inseparable from temperature difference. Therefore, it was necessary to investigate the effect of temperature difference on the response of the LIGS/PtNPs/Nafion microsensor. The CA responses of the microsensor at 17, 20, 25, 30, and 35 were investigated in PBS electrolyte. As shown in
Fig. 3(e), the response increased with an increase in temperature, due to the increased conductivity caused by the increased electron-phonon scattering and thermal velocity of the electrons [
34], [
46]. A relationship curve between the microsensor response and the ambient temperature was established for temperature calibration.
Air flow and mechanical disturbance commonly exist in the planting environment. Therefore, the response of the LIGS/PtNPs/Nafion microsensor implanted into the tomato stem to an air flow of 1.5 m∙s
-1, shaking the stem, and touching the leaves was investigated [
12]. As shown in
Figs. 3(f) and
(g), the microsensor exhibited a certain degree of fluctuation. After the disturbance was eliminated, the response of the microsensor immediately recovered and became stable. The fluctuation might be caused by a relative displacement between the sensor and the plant tissue, which can be avoided by means of a stable fixation method or an interface with strong tissue adhesion.
3.4. In vitro simulation
As the microsensor is a mini-invasive sensing tool, its sensing interface comes into direct contact with the plant’s phloem and xylem sap. The complex contents of the sap may block the normal operation of the microsensor. Thus, it was necessary to study the blocking effect of plant sap on the H
2O
2 signal, in order to obtain a calibration method for the accurate quantitative analysis of H
2O
2 in vivo. The bleeding sap of tomato plants was collected according to the steps shown in
Fig 4(a), and different concentrations of H
2O
2 were added to the bleeding sap to be analyzed by the LIGS/PtNPs/Nafion microsensor. When H
2O
2 was directly added to the bleeding sap without any pretreatment, the lowest concentration that could be measured was about 100 μmol∙L
−1 (
Figs. 4(b) and
(c)). This is likely because the peroxidase in the bleeding sap catalyzes the decomposition of the added H
2O
2. This result indirectly demonstrated that the LIGS/PtNPs/Nafion microsensor is specific to H
2O
2 and has the potential to detect peroxidase activity. Next, the fresh bleeding fluid was heated to inactivate the peroxidase, and H
2O
2 was added to the treated bleeding sap. The response of the microsensor to the bleeding sap is shown in
Fig. 4(d); the minimum concentration of H
2O
2 detected was 5 μmol∙L
−1.
Fig. 4(e) shows the calibration curve based on the absolute response of the microsensor and H
2O
2 concentration. The results show that the microsensor is sensitive to 5-200 μmol∙L
−1 H
2O
2, with a good linear relationship. However, due to biodiversity, the baseline responses may differ in actual monitoring. It is difficult to perform a quantitative analysis using the absolute response. Thus, in order to solve quantitative problems, a calibration curve was drawn using the change rate of the current (Δ
I/I0) and the H
2O
2 concentration (
Fig. 4(f)). This made it possible to effectively deduct the difference in background signals between plants and improve the quantitative accuracy of the actual monitoring results.
3.5. In vivo monitoring
The various environmental factors causing harm to plants, such as drought, salinity, radiation, pests, weeds, or rodents, are collectively referred to as stress [
16], [
47]. Stress can be divided into abiotic stress and biological stress. Abiotic stress, which includes drought, salt, UV radiation, and mechanical injury caused by insect feeding, can limit crop production and threaten national food security [
19]. Thus, the self-powered sensing system was used to monitor the dynamic H
2O
2 levels in tomato plants in real time under osmotic, UV, and mechanical injury stress.
Drought and salt mainly lead to hyperosmotic stress, which is usually referred to as osmotic stress [
18]. Osmotic stress can increase the abscisic acid (ABA) content in plants. SnRK2s (protein kinase) activated by ABA can phosphorylate RbOHF (NADPH oxidase) to generate O
2 
¯ in the exosome space and then form H
2O
2. As shown in
Fig. 5(a), the microsensor was implanted into a tomato stem to monitor the H
2O
2 signal. Tomato plants were cultured in hydroponic nutrient solution containing 30% PEG to induce osmotic stress [
48], [
49]. The monitoring results are shown in
Fig. 5(b). An H
2O
2 signal peak was obtained at 12.5 h (after 7.5 h of osmotic stress), which was denoted as the oxygen burst. The H
2O
2 signal then continued for about 6.5 h. In the control group, the monitoring signals of the tomato plants without osmotic stress remained stable. As shown in
Fig. 5(c), another tomato plant was subjected to osmotic stress after 5 h of cultivation; the H
2O
2 signal increased and reached a peak at 18 h, at which time the top of the tomato plant was obviously withered. At 19 h, the PEG solution was extracted, and fresh hydroponic nutrient solution was injected to relieve the osmotic stress. It was observed that the H
2O
2 signal then began to decline, with another intense oxygen burst occurring at 23 h. This oxygen burst lasted for about 19 h. At 42 h, the signal began to stabilize. At this time, the tomato plant had recovered its fresh state. Furthermore, the microsensor monitoring the tomato plants without osmotic stress exhibited a stable response for 50 h (the fluctuations around 24 and 40 h might be caused by vibration during manual farming operations). After monitoring for 50 h, the microsensor was removed and used to detect 100 μmol∙L
−1 H
2O
2 added to bleeding sap. The H
2O
2 concentration was quantitatively calculated based on the calibration curve in
Fig. 4(f) and was found to be 88.33 μmol∙L
−1, indicating that the microsensor still maintained adequate analytical performance after working for 50 h (Fig. S10 in Appendix A).
Mechanical injury can induce a rapid H
2O
2 signal [
50], [
51]. Therefore, it is convenient to use mechanical injury stress monitoring to study the spatial and temporal distribution of H
2O
2 signals in plants. As shown in
Fig. 5(d), the microsensor was implanted into a tomato stem at about 10 cm above the root. The side branches, leaves, and main stem were clamped by hemostatic forceps to form mechanical injuries. As shown in
Fig. 5(e) and Figs. S11(a)-(c) in Appendix A, the injured main stem and side branches generated several H
2O
2 signals within minutes, which is difficult to capture using conventional detection methods. The microsensor was unable to obtain the H
2O
2 signals caused by the mechanical injury applied to the leaves, similar to a previous study [
19]. The H
2O
2 produced in the leaves might have been decomposed during the transmission process, leaving insufficient to be transmitted to the microsensor implantation site.
The distance between the position of the injury and the microsensor and the time required for the H2O2 signal to appear were recorded in order to calculate the transmission rate of H2O2. As shown in Fig. S11(d) in Appendix A, the transmission rate of the H2O2 signal generated above the microsensor was the fastest (1.58 mm∙s-1), while the transmission rate of the H2O2 signal generated by the side branches was the lowest (0.23 mm∙s-1). To demonstrate that the generated signal originated from H2O2, active oxygen inhibitors were injected into the stem during the monitoring, including peroxidase (CAT), diphenyleneiodonium (DPI), and LaCl3. As shown in Fig. S12 in Appendix A, no H2O2 signal was observed during the monitoring.
Light is important for crop growth. In recent years, due to the destruction of the ozone layer, the UV light reaching the earth’s surface has gradually increased [
52], [
53]. Too much UV light causes stress effects on crop growth [
52], [
54]. As shown in
Fig. 5(f), UV light radiation (365 nm) was applied to form stress on tomatoes. As shown in
Fig. 5(g), an H
2O
2 signal lasting for a few minutes appeared three times. It is worth noting that the response curve had some noise, which might be caused by interference from UV light with electromagnetic wave properties.
It can be seen from the above results that an H
2O
2 signal is an early response of plants to stress that can be generated within a few minutes. The wave characteristics of the H
2O
2 signal under various stresses were compared. As shown in
Fig. 5(h), the H
2O
2 signal caused by osmotic stress can last for tens of hours, while the H
2O
2 signal caused by mechanical injury stress only lasts for tens of seconds, and that caused by UV stress lasts for tens of minutes. Finally, according to the quantitative calibration curve in
Fig. 4(f), we calculated the concentration of the H
2O
2 signal generated by the above stresses. The results are shown in
Fig. 5(i). The H
2O
2 concentration generated by various stresses basically ranged within 10-100 μmol∙L
−1, which is consistent with the conclusions reported in the literature [
19], [
20]. It is clear that the H
2O
2 waves generated by plants in response to various stresses are different in terms of both time dimension and concentration. To further verify the accuracy of this method, the H
2O
2 content in tomato plants before and after osmotic stress was analyzed by means of UV spectrophotometry (Fig. S13 in Appendix A). The results showed that the H
2O
2 content after osmotic stress increased by 89 μmol∙L
−1, in comparison with the result obtained from the sensing system of 78 μmol∙L
−1. The difference in these values might be caused by biodiversity and the sample treatment.
In conclusion, we found that the tomato plants have the fastest response but weakest response intensity to mechanical damage. The response speed to osmotic stress was the slowest, but the response intensity was the largest. These results may demonstrate the time and concentration specificity of the H2O2 signal in response to abiotic stress, indicating that tomato plants can easily cope with some mechanical injury in daily growth. However, under osmotic stress and UV light, once the stress intensity exceeds the tolerance limit of the tomato plants, the plants produce a high-intensity H2O2 signal to resist stress damage.
4. Conclusions
It is necessary to analyze the breeding target traits and crucial biochemical molecules in plants via micro, dynamic, continuous, accurate, and intravital methods. H2O2 is a key early signal molecule for plants dealing with various stressors and has the characteristics of fast, long-distance, and automatic transmission. In this work, we developed an implantable and self-powered sensing system for the in vivo monitoring of the dynamic H2O2 level in a plant. Using this system, the transmission process of the H2O2 signal was monitored and analyzed in vivo, and the time and concentration specificity of the H2O2 signal for abiotic stress were investigated. The system has the following advantages:
(1)It can convert the H2O2 signal into a readable electrochemical signal both in vivo and in real time;
(2)The sensing system has a high time resolution (0.1 s);
(3)There is no need to inject various marking materials into the plant being monitored;
(4)The sensing system possesses anti-interference capability against environmental changes;
(5)Power from sunlight or artificial light in the agricultural environment can be fully utilized to continuously power the sensing system.
The results showed that this system can continuously monitor the H2O2 level in tomato plants under abiotic stress, including osmotic stress, mechanical damage, and UV radiation. The transmission process of the H2O2 signal was monitored and analyzed in vivo, and the content and waveform differences in H2O2 caused by different abiotic stresses were compared. We also investigated the time and concentration specificity of the H2O2 signal for abiotic stress. The system provides an advanced analysis tool for plant stress-resistance breeding and the early sensing diagnosis of plant stress in facility agriculture. In future, a more stable fixation method, stronger tissue adhesion, a reduction of baseline drift and noise, and the effect of device implantation on plant growth will be researched.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This research was supported by the Joint Funds of the National Natural Science Foundation of China (U23A20173) and the Fundamental Research Funds for the Central Universities.
Acknowledgments
This research was supported by the Joint Funds of the National Natural Science Foundation of China (U23A20173) and the Fundamental Research Funds for the Central Universities.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2023.11.021.
PDF
(2985KB)
