The microstructure of the leaves is observed using the Phenom ProX scanning electron microscope (SEM) manufactured by Phenom. Samples of three different roasted tobacco leaves, processed into cured leaves using the same curing method, measuring 2 mm × 2 mm, are immersed in a 2.5% glutaraldehyde phosphate buffer (pH 7.2) and subjected to vacuum pumping at room temperature until they sank. The samples are then stored in a 4°C refrigerator for more than 72 h to ensure thorough cleaning and drying. The preparation of the samples aim to maintain intact tissue and cell morphology, fully expose the intended observation area, and ensure they remain fully dry upon entering the vacuum chamber for optimal conductivity and high secondary electron yield. These samples are subsequently observed at an accelerating voltage of 15 kV (Zhang et al., 2020; Cai et al., 2008).

The primary apparatus used in this experiment is a contact angle measuring instrument (SDC-100S). The tobacco leaf sample is meticulously positioned at the center of the sample stage. An electric injection system, located directly above the stage, facilitates the release of droplets. Adjacent to this, an LED light source is installed. The stage is then elevated to allow droplets to impinge on the surface of the tobacco leaf. Once stabilized, the stage is lowered to capture the droplet’s morphology. The contact angle is subsequently obtained from the captured image of the droplet’s shape. The volume of the liquid used (water and diiodomethane) is 1 μL, the humidity of the test environment is 45%, and the drops are set to take pictures 15 s after deposition. The corresponding image size is 4,000 μm × 3,000 μm. For measuring the contact angle on the venation, different areas on the same tobacco leaf are selected. When measuring the contact angle on the foliage, three tobacco leaves are selected, and different areas on each leaf are measured.

In order to obtain the roughness of Roast tobacco leaves, an optical profilometer with Vertical Scanning Interferometry (NT9100) manufactured by Veeco company is used. It provides fast and high-precision 3D surface topography measurement in the vertical scanning range of 0.1 nm–1 mm, suitable for micron measurement. The minimum measurable piece size is 50 μm × 50 μm, with a maximum measurement range extending to the centimeter level. The tall bumps on the Roast tobacco leaves fall within the Z-range limitation of the optical profilometer, allowing the entire vertical range of the leaves to be measured. The measuring surface of the corresponding tobacco leaf is kept in a horizontal state. The light used is a broad-spectrum LED, the pixel number is 640 × 480 and the pixel resolution is 972.5 nm, and the total tobacco leaf measurement area is 622.4 μm × 466.8 μm. The tobacco sample is divided into 2 mm × 2 mm sections for analyzing surface features using the Profilm 3D Image Acqusition.

Upon utilizing scanning electron microscope (SEM), we discern two unique microstructures within tobacco leaves, subsequently classifying the leaf tissue into two primary components: the venation and the foliage. We choose three specimens of tobacco leaves, specifically Roast tobacco leaves 1, 2, and 3, to further investigate the interplay between these microstructures and their influence on the wettability and overall functionality of the leaf surfaces. The preliminary phase of our study entails the documentation of the surface microscopic morphology of tobacco leaves. In our investigation of the selected tobacco samples, we employ SEM to capture representative surface morphologies, which are depicted in Figure 1. The findings reveal the microscopic morphology of the venation in tobacco leaves, characterized by a predominantly regular groove-ridge pattern like Figure 1A. In contrast, Figure 1B illustrates the microscopic morphology of the foliage, which diverges from the venation by exhibiting an irregular, bulging topography.

To investigate the wetting properties of the venation and foliage of tobacco leaves, we conduct experiments to measure the contact angle of water droplets on their respective surface morphologies. However, on these microstructured surfaces, the droplets exhibit varying degrees of anisotropy. Specifically, the shape and contact angle of water droplets perpendicular to the venation differ from those along the long axis of the veins. This phenomenon occurs because the water droplets spread along the long axis of the venation, resulting in a smaller contact angle in that direction (Calvimontes et al., 2012; Timoshenko et al., 2012; Wang et al., 2018). In our experiments, we primarily focused on the contact angle perpendicular to the venation direction. Figure 2 presents some results from our experiments, showing the measured contact angles perpendicular to the venation direction. We selected one Roast tobacco and measured the contact angles of different areas on the venation. The average contact angle is 117.2° ± 2.3°. This result demonstrates that the venation of tobacco leaves primarily exhibit hydrophobic properties. The results of the contact angles for the other venations are included in Supplementary Figure S1, as the measurement process is consistent across all samples.

After measuring the contact angles on the venation, we similarly assess the contact angle on the foliage. Figure 3 exhibits one of experimental contact angles for Roast tobacco leaf 1, Roast tobacco leaf 2, and Roast tobacco leaf 3. Numerous experiments are conducted to obtained the average contact angles for the different tobaccos, which are found to be 103.578°, 99.290°, 108.113°, respectively. These results indicate that different tobacco leaves exhibit varying degrees of hydrophobicity. However, our experimental results demonstrate that the variation in contact angle across different areas of the same tobacco leaves is minimal, it can be saw in Supplementary Figures S2–S4. The experimental results indicate that the foliage also possesses a contact angle suggestive of hydrophobic properties, thereby confirming that the tobacco leaves exhibit hydrophobicity in our study.

In order to establish the prediction model of tobacco leave contact angle, we also need to measure Young’s contact angle of plants. Therefore, in the following chapters, we will describe in detail how to measure and obtain an approximate value of Young’s contact angle of tobacco leaves.

The concept of the contact angle is first introduced by Thomas Young in 1805, which is expressed as follows (Lin et al., 2014): cos ⁡ θ Y = γ sg − γ sl γ lg (1) where γ sg is the interfacial tension between solid and gas, γ sl is the interfacial tension between solid and liquid, γ lg is the interfacial tension between liquid and gas and θ Y corresponds to the Young’s contact angle.

To obtain Young’s contact angle, it is necessary to know the values of three interfacial tensions from Equation 1 (Zhu et al., 2007a; Zhu et al., 2007b). Using these surface tension values, we can calculate Young’s contact angle. To ensure the accuracy of Young’s contact angle calculation, we reviewed varous works conducted by numerous scholars to calculate the surface tension and ultimately arrive at a method that yields more satisfactory results (Chindam et al., 2015; Chindam et al., 2016; Zenkiewicz et al., 2006). We selected three different types of tobacco leaves for the experiment. The method used to measure these leaves is called Neumann method and the team applied this method to calculate the surface energy of paper and found it has minor error (Fan et al., 2022; Chindam et al., 2015; Chindam et al., 2016; Zenkiewicz et al., 2006). The relationship between interfacial tensions and Young’s contact angle is as follows: 2 γ S / γ L 1 2 e − 0.000115 γ L − γ S 2 = 1 + cos ⁡ θ (2)

Where γ S is the surface tension of solid, γ L is the surface tension of liquid. Combined with the Owens-Wendt-Rabel-Kaelble (OWRK) method, researchers use these approaches to determine partial surface tension, and the results are closed to reality, both formulas are extensively used in the literature (Tavana et al., 2005; Cui et al., 2004; Liu et al., 2013; Pak and Kim, 2023; Chindam et al., 2015; Chindam et al., 2016; Zenkiewicz et al., 2006): 2 γ S d γ L d 1 2 + 2 γ S p γ L p 1 2 = γ L 1 + cos ⁡ θ (3) where γ S d is the nonpolar component of the solid’s surface tension, γ L d is the nonpolar component of the liquid’s surface tension, γ S p is the polar component of the solid’s surface tension, γ L p is the polar component of the liquid’s surface tension. The methods involve two types of droplets: water and diiodomethane. The polar component of water’s surface tension is 51.000 mN/m, and the nonpolar component is 21.800 mN/m (Asai et al., 2022; Tavana et al., 2005; Strobel and Lyons, 2011; Kim et al., 2015; Decker et al., 1999; Pak and Kim, 2023). For diiodomethane, the polar component is 0 mN/m, and the nonpolar component is 50.800 mN/m (Cui et al., 2004). Researchers use a video-optical contact angle meter to measure the contact angles of three varieties of Yunnan tobacco leaves with these probing liquids, and calculate the surface tension and their components (Wang et al., 2014; Liu et al., 2013).

In this experiment, two physical quantities are required: the interfacial tension at solid-gas interface, the interfacial tension at liquid-gas interface. The measurement procedure is as follows: First, diiodomethane is used to measure the contact angles of the three different smooth tobacco leaves as shown in Figure 4, three tobacco leaves’ contact angles are approximately 60°. Multiple measurements are taken for each type of tobacco leaf to calculate the average values, which are found to be 62.801°, 61.923°, and 60.000°, respectively. The contact angles show small differences among different tobacco leaves. Additionally, when measuring various areas of the same tobacco leaf, only minor differences in the contact angles are observed, as shown in Supplementary Figures S5–S7. Then, the average contact angles of diiodomethane are substituted into Equation 2 to obtain the surface tension of solid γ S , substituting the known tension component of diiodomethane into Equation 3 and we can obtain the nonpolar component and polar component of solid (Marmur, 1996; Du et al., 2024). The results of tobacco leaves are similar to those reported by other researchers (Wang et al., 2014). Table 1 provides a detailed presentation of the results, indicating that the final surface tension values for the three types of tobacco leaves are comparable, reflecting both polar and non-polar forces.

Finally, by substituting the polar and nonpolar forces of water into Equation 3, we obtain an approximate Young’s contact angle between water and an ideal solid, and we consider both venation and foliage have the same Young’s contact angle, the results are presented in Table 2. The results indicate that the Young’s contact angles for the three types of tobacco leaves are closely aligned, approximately 90°. However, the experimental contact angle significantly deviates from the Young’s contact angle, primarily due to variations in chemical composition and surface microstructure.

Figure 1 depicts a regular groove-ridge pattern on the surface of venation. As a result, we propose a two-dimensional model composed of a flat base, semi-cylindrical venation, and wetting droplets, as demonstrated in Figure 5. In this theoretical model, the radius of the semi-cylindrical venation is denoted by R, and h represents the height difference between the radius of the tobacco venation and the depth of droplet infiltration into the venation. The angle α subtended by the circle is the central angle corresponding to the infiltrated portion.

For a single venation, the solid and liquid contact area is expressed as S 1 and the projected area of venation is S 2 (AuthorAnonymous et al., 2018; Jiang et al., 2011; Hsieh et al., 2008; Milne and Amirfazli, 2012; Shim et al., 2014; Lim et al., 2016). Assume the length of the venation is L, we can calculate the contact area S 1 , and the projected area of the venation S 2 S 1 = 2 L R 2 − h 2 (5) S 2 = 2 R L (6)

Within the framework of the Cassie–Baxter model, which considers the two previously mentioned physical properties, the ratio of solid-liquid contact is acknowledged as being of utmost importance. This critical physical parameter reflects the proportion of the liquid’s contact area with the leaf venation compared to the venation’s projected area (AuthorAnonymous et al., 2018; Jiang et al., 2011; Hsieh et al., 2008; Milne and Amirfazli, 2012). This ratio is critical for understanding the wetting behavior and adhesion characteristics of a droplet on the leaf surface (Ye and Mizutani, 2023; Xue et al., 2012). Combined the Equation 5 and Equation 6, the fraction f can be calculated as follows: f = S 1 S 2 = 2 L R 2 − h 2 2 R L = 1 − h R 2 (7)

Based on the Cassie–Baxter model, by substituting Equation 7 into Equation 4, we derive an explicit equation that can be used to predict the contact angle on venation of tobacco leave cos ⁡ θ c = 1 − h R 2 cos ⁡ θ Y + 1 − 1 (8) where θ c is contact angle at Cassie–Baxter state. It can be observed that the contact angle is intricately linked to the ratio of h/R, as well as the Young’s contact angle. Using Equation 8, we plot the curves illustrating how the ratio of h/R and Young’s contact angle influence the contact angle. By adjusting these two factors, we can modify the wettability.

Contrary to the venation of leaves, the foliage presents an irregular shape, complicating the development of either two-dimensional or three-dimensional models. To overcome this obstacle, we employ an innovative approach to calculate the contact angle of foliage by introducing the ratio of the actual surface area to its projected area.

We use a profilometer and select a slice of tobacco to calculate the surface area and its projected area. Figure 6 shows the roughness of three different types of tobacco under a profiler. The projected area of the samples is 0.289 mm². Areas of higher elevation are marked in red, transitioning to green and blue with decreasing height, indicating lower elevations. The total surface area is calculated as the sum of these surface folds. Table 3 presents the surface area, projected area, and the ratio of surface area to projected area for three different types of tobacco. It is evident that the surface of Roast tobacco 3 exhibits greater smoothness compared to Roast tobacco 1.

The solid-liquid contact area is denoted as S 1 , and the roughness A can be defined as the ratio of surface area to projected area. The surface area is expressed as S S and the projected area is expressed as S P . According to the Cassie–Baxter model, factors affecting contact include Young’s contact angle and the proportion of the solid-liquid contact area, as indicated by Equation 4. When the Young’s contact angle for a solid is held constant, the proportion of the solid-liquid contact area becomes the sole determinant influencing the contact angle. Given the uncertainty regarding the values of S 1 and S S , we introduce the concept of roughness. By interpreting the meaning of roughness and its associated physical quantities, we can derive an expression for roughness. A = S S S P (9)

Since the contact angle depends on the solid-liquid contact area, we can link Equations 4, 9 with percentage of contact area f 1 . This allows us to derive the contact angle expression for foliage, as presented in Equation 10. cos ⁡ θ C = S 1 S S A cos ⁡ θ Y + 1 − 1 (10)

This expression captures the interplay between the contact angle and the specific geometry of the solid-liquid interface on the foliage surface. It can be seen that the contact angle expression for foliage can be characterized by a single unknown parameter S 1 / S S . This equation establishes a correlation between the contact angle and two key factors: the surface roughness and the ratio of the solid-liquid contact surface area to the surface area. This mathematical relationship provides a quantitative framework for understanding and predicting the wettability of foliage surfaces, which is essential for a wide range of applications, from agricultural practices to the design of biomimetic materials.

During the process of SEM imaging, we discover an interesting phenomenon: for some foliage surfaces that are relatively smooth and flat, the solid-liquid ratio in the Cassie–Baxter model can be directly obtained by binarizing the partially observed foliage images. This occurs because, when a liquid contacts a solid surface with a micro-nano structure, the liquid does not completely fill all the grooves and holes but instead forms a continuous layer of air that supports the liquid along with the solid surface, allowing water droplets to rest on the solid surface (Bormashenko, 2010; Shi et al., 2012). Based on this mechanism, the air in the surface voids of the flat tobacco leaf supports the water droplets, leading to an observable solid-liquid contact area. This phenomenon provides a new method to calculate the contact angle.

We employ a method that involves characterizing the micromorphology of the foliage using an electron microscope and analyzing it with ImageJ software. The characterized area is divided into regions representing the solid-liquid contact area and the non-contact area. The software then computes the ratio of these areas, which is subsequently utilized in the Cassie–Baxter model to determine the corresponding contact angle. In this study, we examine various types of tobacco leaves and different foliage blocks. Figure 7 demonstrates the alterations resulting rom image binarization and segmentation. We select a part of area of tobacco leaves under SEM as shown in Figure 7A. Specifically, in Figure 7B, the solid-liquid contact area is represented by the white regions, while the non-contact areas are depicted in black. This visual representation aids in understanding of the distribution and extent of contact between the solid and liquid phases on the surface being analyzed. Each cell is individually separated, labeled, and its area calculated. Upon completing these calculations for all cells, the sum of the areas corresponding to the white regions, which represent the solid-liquid contact area, can be determined. The ratio of this sum to the selected area provides the percentage of the solid-liquid contact area.

Table 4 presents the contact angles calculated through the method in Section 4.3, the solid-liquid area of the contact and non-contact are identified by ImageJ, and the percentage of the two are substituted into Equation 4, theoretically predicted contact angles for three types of tobacco leaves. For each identical Roast tobacco leaf, we take 10 different foliage parts to obtain the contact area of solid-liquid. As shown in this Table, the average contact angle for the different types of tobacco are 105.130°, 101.750°, and 102.250°, respectively. The observed variations in error are minor across different tobacco types, resulting in an overall negligible deviation. For roasted tobaccos, the average theoretical contact angle closely approximates the experimental contact angle, with an error margin of approximately 2%. This indicates that the theoretical predictions for contact angles closely align with the experimental data.

From the Equation 10, it can be seen that the value of the contact angle can be adjusted by modifying the ratio of solid-liquid contact area to surface area S 1 / S S and Young’s contact angle. We plotting some curves about the relationship of contact angle between the ratio S 1 / S S and Young’s contact angle through the Equation 10 as shown in Figure 8. As observed in Figure 8A, the contact angle decreases from 180° to 90° as the ratio S 1 / S S increases. This observation suggests that a larger solid-liquid contact area correlates with a higher degree of hydrophilicity on the plant surface. In other words, as the contact area between the plant surface and the liquid increases, the surface exhibits a greater affinity for water, indicating a more hydrophilic nature. Conversely, the contact angle increases with the increasing of Young’s contact angle, as shown in Figure 8B. Interestingly, despite Young’s contact angle being less than 90°, by manipulating the ratio S 1 / S S , it is possible to induce a contact angle on the plant surface that exceeds 90°, thereby exhibiting hydrophobicity. This demonstrates the plant’s ability to alter its surface properties from hydrophilic to hydrophobic through the adjustment of the contact area ratio.

Some researchers have conducted extensive simulations to replicate the contact angle of droplets on various microstructures. For instance, they employed Evolver to simulate the infiltration process on Nepenthesand discovered that the contact angle ranged from 155° to 155.400° under an infiltration ratio of 0.700–0.800, providing theoretical insights into the superhydrophobic mechanism of Nepenthes (Wang et al., 2022). To validate the contact angle prediction model for the venation, we also simulate the contact angle of tobacco leaves by using COMSOL Multiphysics. This approach allows us to test and refine our understanding of the hydrophobic properties exhibited by intricate structures found in nature.

We establish a two-dimensional venation model in COMSOL using the two-phase flow module in fluid flow, employing phase filed method. As calculated in previous sections, the intrinsic contact angles of various tobacco leaves are approximately 90°. Thus, within our simulation, we set the intrinsic contact angle of the wetting wall to 90° to accurately represent these tobacco leaves. A corresponding micro-morphology is constructed on this basis. We apply phase field and laminar flow boundary conditions, designating outlets, inlets, and open boundaries accordingly, while ensuring of mass conservation. The grid type is a free triangular grid, and the grid size is very fine, with the software automatically dividing it. Upon stabilization of the droplet, we measure the contact angles and the ratio of h/R. The results of the venation’ theoretical model is shown in Figure 9, which illustrates the simulated contact angle of a water droplet in comparison to the predictions of the model for the Cassie–Baxter state. For the venation of tobacco leaves, the contact angle also increases with the increasing of the ratio h/R. The contact angle of the droplet is time-dependent during the dynamic simulation, and the contact angle decreases until it remains constant with the passage of time. Based on this behavior, we extend the simulation time until the droplet at a stable state (Dezellus et al., 2002; Wang et al., 2022). Both the simulation and the model exhibit reasonably good agreement, with the model accurately capturing the trend observed in the simulation. This congruence between the simulation and the model further validates our understanding of the hydrophobic properties of the venation structure and its interaction with water droplets.

From the Equation 8 of venation, it can be seen that the value of the contact angle can be regulated by changing the ratio of h/R and the Young’s contact angle. Figure 9 presents the trend of contact angle variations in accordance with the theoretical model. Figure 9 demonstrates that an increase in the h/R ratio leads to an enlargement of the contact angle. The ratio h/R quantifies the solid-liquid contact area, with a higher value indicating a decreased contact area between the solid and liquid phases. As this ratio approaches zero, the contact angle reaches its minimum, indicative of complete liquid infiltration into the solid surface, achieving the Wenzel state. Conversely, as the ratio h/R approaches 1, the contact angle reaches its maximum, indicating an almost non-existent contact between the liquid and the solid surface. This condition corresponds to the Cassie state, which is an ideal scenario for superhydrophobic surfaces. In this state, the liquid appears to hover above the surface rather than making direct contact. This relationship highlights the direct influence of geometric parameters on the wetting properties of surfaces, as predicted by our theoretical model. Figure 10 presents the trend of contact angle variations in accordance with the theoretical model of venation from Equation 8. From Figure 10, we find the same phenomenon with foliage that the larger of the Young’s contact angle, the larger the contact angle. Similarly, by modifying the microstructural morphology, plants displaying Young’s contact angles less than 90° can be transformed into hydrophobic plants with contact angles greater than 90°. This adjustment demonstrates the impact of surface structure on the hydrophobic properties of plant surfaces, highlighting the practical application of altering wetting characteristics through morphological changes.

The results indicate that both the foliage and venation of tobacco leaves exhibit hydrophobic properties. The variations in contact angles among different tobacco leaves are primarily attributed to the magnitude of the Young’s contact angle and the percentage of the liquid droplet’s wetted area. By manipulating the percentage of the wetted area on the venation and foliage parts and adjusting their respective proportions, one can achieve the desired wetting effects. This approach underscores the potential for targeted modifications of plant surface hydrophobicity through alterations of the surface micromorphology.