Comparative Study of Water Absorption and Dimensional Stability Between Bamboo Nodes and Internodes

Abstract

Bamboo often suffers from moisture-induced cracking, in which the structural and dimensional differences between nodes and internodes may be key contributing factors. Taking Phyllostachys edulis (Carrière) J. Houz. as an example, this study systematically examined the water absorption behavior and dimensional stability of bamboo nodes and internodes, and further analyzed their pore structure and chemical composition to provide a comprehensive understanding of their moisture response. This study systematically compared nodes and internodes of Phyllostachys edulis in water absorption behavior, dimensional stability, pore architecture, and vascular structure. Results showed that internodes exhibited higher water absorption rates and capacities in both short- and long-term tests, whereas nodes displayed lower water uptake and were prone to cracking during drying, indicating reduced dimensional stability. Anatomical and infrared analyses revealed that diaphragms, transverse vascular bundles, and spiral networks in nodes increased fluid path tortuosity, reducing longitudinal permeability. Pore structure analysis further indicated that internodes contained abundant pores facilitating rapid liquid transport, while node pores were mainly medium to large, favoring liquid retention but limiting permeability. Higher cellulose crystallinity and lignin content in nodes enhanced hydrophobicity, further restricting water penetration. Additionally, the complex fiber orientation in nodes induced anisotropic swelling and internal stress, increasing the risk of twisting and cracking. This multi-scale investigation elucidates the structural and compositional mechanisms underlying the observed differences in water absorption behavior and dimensional stability between nodes and internodes. These findings offer valuable insights for improving the moisture resistance, dimensional stability, and overall performance of bamboo materials in engineered applications, and provide a solid foundation for their targeted modification and optimization.

1. Introduction

Plastic pollution is rapidly emerging as a global environmental crisis, drawing widespread attention from governments and researchers alike. Consequently, the development of recyclable and biodegradable alternative materials has become an urgent scientific challenge [1,2,3,4]. Among various candidate resources, Bambusoideae plants have attracted considerable interest due to their outstanding specific strength, specific stiffness, and rapid regenerative capacity, making them promising renewable alternatives to conventional engineering plastics and broadleaf timber [5,6,7,8]. At the same time, with the full-scale implementation of the “Bamboo as Plastic Substitute” initiative by the International Bamboo and Rattan Organization, the bamboo industry is entering a new phase characterized by refined processing and high-value utilization [9,10,11,12,13].

Nevertheless, the low-dimensional stability of bamboo under humid and hot conditions, coupled with its high susceptibility to cracking, has become a critical bottleneck limiting its large-scale application in outdoor structural materials, high-end furniture, and composite engineering components [14,15]. Existing studies generally attribute this limitation to the pronounced structural heterogeneity of bamboo [16]. Specifically, the bamboo culm consists of periodically distributed nodes and internodes, which systematically differ in tissue architecture, cell morphology, and multi-scale structural features [17,18,19,20]. Internodal regions are dominated by longitudinal vascular bundle sheaths, which significantly reduce moisture resistance but also increase the risk of cracking during hygroscopic swelling and drying shrinkage [21,22]. In contrast, nodal regions, with their diaphragms and circumferential vascular bundle sheaths, act as effective barriers to moisture transport. However, even in these regions, rapid moisture absorption or drying can induce localized stress concentrations, leading to crack formation [17,23,24,25]. Previous studies have shown that cracks in bamboo are predominantly concentrated at nodes, highlighting the critical role of nodes in the overall cracking behavior [26]. Furthermore, techniques such as low-field nuclear magnetic resonance have revealed differences in moisture absorption rates and water content between nodes and internodes, indicating that internodal regions absorb water more readily while nodal structures help mitigate local water content and delay crack initiation [27].

Despite these findings, existing studies still lack a systematic investigation linking the hygrothermal behavior–structure–performance relationships of bamboo [28,29,30]. To fill this gap, the present work aims to clarify the intrinsic differences in water absorption and dimensional stability between the nodes and internodes of Phyllostachys edulis (Carrière) J. Houz. (Moso bamboo). Particular attention is given to how structural features, pore characteristics, and chemical composition collectively influence moisture transport and deformation behavior. By establishing these relationships, this study provides a scientific basis for improving the crack resistance and dimensional stability of bamboo, thereby supporting its high-value utilization in engineered and functional applications.

2. Materials and Methods

Bamboo specimens used in this study were collected from small-diameter (approximately 3 cm) 3~4-year-old Phyllostachys edulis culms in Sichuan, China. Five bamboo blocks with axial lengths of 15 ± 0.2 mm and 8 ± 0.2 mm were cut from the internode and node of Moso bamboo, respectively. The 15 ± 0.2 mm rings were used for water absorption tests, while the 8 ± 0.2 mm rings were used for dimensional stability measurements. Prior to testing, all samples were oven-dried at 103 °C and weighed.

2.1. Water Permeability Test

Water absorption tests were conducted to quantify the differences in moisture diffusion behavior between bamboo nodes and internodes. To further clarify these differences, both short-term and long-term water absorption rates were examined. The detailed procedure is illustrated in Figure 1. The experiments were carried out at room temperature using the atmospheric-pressure permeation method, which relies on the intrinsic capillary infiltration of bamboo. This approach allows the results to directly reflect the material’s water absorption behavior under practical conditions, providing valuable reference for the design, processing, and application of bamboo-based products [15].

Six samples from bamboo nodes and internodes (15 ± 0.2 mm in length) were used for the long-term water absorption test. After being dried at 103 °C and weighed, the samples were vertically immersed on a metal wire mesh filled with deionized water (the liquid level was controlled to cover the bottom of the samples by 3 ± 0.2 mm) (Figure 1). The samples were taken out at 1 h, 2 h, 5 h, 7 h, 9 h, 12 h, 24 h, 48 h, 72 h, 96 h, 120 h, 144 h, 168 h, 192 h, 216 h and 240 h. The excess surface moisture was gently absorbed with filter paper, and the mass was immediately weighed and recorded. The water absorption and rate of the bamboo samples are calculated through Formulas (1) and (2), respectively:

$$\omega (\%)={\displaystyle \frac{m_t-m_0}{m_0}}\ast 100$$

(1)

$$v(\%/\mathrm{h})={\displaystyle \frac{m_t-m_0}{m_0\ast t}}\ast 100$$

(2)

Here, w represents the water absorption percentage (%), v represents the water absorption rate (%/h), m₀ represents the mass of the sample (g) before water absorption, mt represents the mass of the sample (g) after water absorption hours, and t represents the time of water absorption (h).

To more intuitively reveal the differences in water absorption behavior between bamboo nodes and internodes, an infrared thermography (IRT) system was employed during the short-term (about 20 min) water absorption test to track the axial migration of moisture in bamboo (Figure 1 and Figure 2). IRT is based on the principle that all objects above absolute zero emit infrared radiation, the intensity of which correlates with their surface temperature. The higher the temperature, the stronger the radiation detected by the imager, and the brighter or redder the corresponding area appears. This non-contact and real-time technique is widely used to analyze the moisture distribution and thermal properties of bamboo and wood materials. Prior to testing, each specimen was split longitudinally into two halves along the horizontal diameter, and the diaphragms within the nodes were removed to obtain semi-tubular samples of both node and internode. AB adhesive was then applied to the freshly exposed inner surfaces to simulate the hydrophobic characteristics of the outer layer of bamboo. Subsequently, the treated specimens were vertically immersed in deionized water to perform an end-pressure water absorption test, following the same procedure as used in the long-term water absorption test. For the left halves, infrared thermography was employed to qualitatively monitor short-term water uptake, whereas for the right halves, the gravimetric method was simultaneously applied. Finally, water absorption ratio and absorption rate were calculated according to Equations (1) and (2).

2.2. Dimensional Stability Test

Dimensional stability experiments were carried out to evaluate stability differences between nodes and internodes. The experimental reference standard is GB/T 15780-1995 “Test Methods for Physical and Mechanical Properties of Bamboo”, which studies the dimensional change behavior of bamboo under different environmental conditions (including changes in humidity and temperature) and assesses its dimensional stability. Six samples with a size of 8 ± 0.2 mm were prepared from the bamboo internodes and nodes. First, the moisture content was balanced in a constant temperature and humidity chamber at 20 ± 2 °C and a relative humidity of 65% ± 5%, and the initial size (L₀) was measured. The samples were then immersed in deionized water at room temperature, and their dimensions were measured every 12 h until the dimensions no longer changed. The saturated-state dimensions (L₁) were obtained, and the wet-expansion rate (S₁) was calculated. Next, the saturated sample was placed back into the same temperature and humidity environment, adjusted to a 12% air-dry state, and measured every two days until it stabilized; the air-dry state dimensions (L₂) were obtained, and the air-dry shrinkage rate (S₂) was calculated. Finally, the air-dried samples were dried in an oven at 103 °C until mass equilibrium was achieved. The dimensions of the absolute dry state (L₃) were recorded, and the absolute dry shrinkage rate (S₃) was calculated. The specific formula is shown in Equation (3) below: To further understand the dimensional change characteristics of bamboo under alternating dry and wet conditions, after completing the dry shrinkage and wet expansion experiment in this study, the eccentricity (Formula (4)) was introduced as a supplementary indicator to more comprehensively evaluate the dimensional stability of bamboo.

$$S_n(\%)=|{\displaystyle \frac{L_n-L_i}{L_i}}|\ast 100$$

(3)

where n = 1 represents the swelling rate in saturated water, where i = 0; n = 2 represents the shrinkage rate in air drying, where i = 1; n = 3 indicates the absolute dry shrinkage rate, where i = 1.

$$e=\sqrt{1-{{\displaystyle \frac{b}{a}}}^2}$$

(4)

where e denotes the eccentricity, with a value range of 0 < e < 1 (when e = 0, the ellipse degenerates into a circle; when e approaches 1, the ellipse becomes very flat); a denotes the length of the major half-axis of the ellipse (Figure 3B); b denotes the length of the shorter half-axis of the ellipse (Figure 3B).

2.3. X-Ray Diffraction (XRD) Test

X-ray diffraction (XRD) combined with the Segal method was used to calculate cellulose crystallinity (CrI), revealing how the ratio of crystalline to amorphous regions regulates moisture transport and dimensional changes. Samples were taken from the internodes and nodes with a size of 8 ± 0.2 mm. The green and yellow parts of the bamboo were removed from the samples. The samples were then ground into powder with a mesh size of less than 100 and dried to an absolutely dry state. All measurements were performed on Bruker D8 Advance X-ray diffractometer (Bruker, Germany). The tube voltage was set at 40 kV, the tube current at 30 mA, the wavelength at 0.54056 nm (Cu-Kα target), the scanning Angle range was 5° to 45°, and the scanning rate was 4°/min. The relative crystallinity was calculated by the Segal [31] method, and the specific algorithm is shown in Formula (5):

$$CrI={\displaystyle \frac{I_{002}-I_{am}}{I_{002}}}\ast 100\%$$

(5)

where: CrI—relative crystallinity (%), I002—maximum intensity of (002) lattice diffraction (arb. units), Iam—scattering intensity of amorphous background. (I002 and Iam share identical units when 2θ ≈ 18°).

2.4. Mercury Intrusion Porosimetry

Mercury intrusion porosimetry was employed to characterize pore size distribution, porosity, and tortuosity. Samples were taken from the bamboo internodes and nodes, and the samples were processed into blocks smaller than 1 × 1 × 1 cm. Pore size distribution, porosity, median pore diameter, and pore volume of internode and node specimens were quantitatively measured using mercury intrusion porosimetry (MIP) with an AutoPore IV 9500 porosimeter (Micromeritics, Norcross, GA, USA). During the test, the specimens were immersed in mercury, and the intrusion pressure was stepwise increased from 0 MPa to 420.58 MPa according to the preset data acquisition points of the instrument. The extrusion process was then performed in a stepwise manner from 420.58 MPa back to 0 MPa. The pore diameters of the specimens were calculated using the Washburn equation [32,33]:

$$d={\displaystyle \frac{-4\gamma \mathit{cos}\theta }{p}}$$

(6)

where d represents the pore diameter, θ represents the mercury contact angle with bamboo/wood materials (130°), γ represents the mercury surface tension (0.485 N/m), p represents the pressure (MPa) applied to mercury.

2.5. Chemical Composition Test

Analyses of the three major chemical components were conducted to clarify their distribution patterns in nodes versus internodes. The cellulose, hemicellulose, and lignin contents of internode and node specimens were determined following the National Renewable Energy Laboratory (NREL) procedure. The samples were first ground into powder, dried to a constant weight, and sieved through a 60-mesh screen to ensure uniform particle size. A precise amount of oven-dried bamboo powder was weighed and subjected to acid hydrolysis with 72% sulfuric acid, followed by dilution and hydrolysis at elevated temperature. Acid-soluble lignin and acid-insoluble lignin were separated through filtration and washing. The monosaccharide content in the hydrolysate was analyzed using high-performance liquid chromatography (HPLC), and the cellulose and hemicellulose contents were subsequently calculated. Finally, the residue was dried and ignited to determine the acid-insoluble lignin content.

3. Results and Discussion

3.1. Water Absorption Behavior

The longitudinal permeability of bamboo internode and node was compared through atmospheric pressure end-absorption tests, and the curves of water absorption ratio and absorption rate as a function of time are presented in Figure 2a,b. The results revealed that both internode and node exhibited similar water absorption patterns, characterized by a rapid uptake at the initial stage followed by gradual saturation, with the absorption ratio possibly reaching a steady state. However, the absorption ratio of the internode was significantly higher than that of the node. Specifically, during the long-term experiment (250 h), the internode showed a sharp increase in absorption ratio at the early stage, which stabilized at around 90%, whereas the node stabilized at a much lower value of approximately 70% (Figure 2a). Further analysis of the absorption rate (Figure 2b) demonstrated that the internode displayed a markedly higher rate during the early stage (0–25 h), with an initial value of about 27.5%. This rate gradually declined and eventually reached equilibrium. In contrast, the node consistently exhibited a lower absorption rate with only slight variations over time, starting from an initial level of ~10%, which was substantially lower than that of the internode. These findings clearly indicate that the longitudinal permeability of the internode surpasses that of the node. The superior permeability of the internode can be attributed to the abundance of vascular bundles aligned in the longitudinal direction, particularly vessels and sieve tubes in the metaxylem, which provide direct pathways for water transport and thereby enhance longitudinal permeability [34,35]. By comparison, the nodal region is structurally denser and reinforced with diaphragms and transverse vascular bundles, which effectively obstruct lateral water migration [36,37]. To verify whether the above macroscopic differences were statistically significant, independent-samples t-tests were conducted to quantitatively compare the water-absorption rate and water-absorption speed between internode and node samples, as shown in

Table 1. t-test of water absorption rate and water absorption velocity.

Type	Mean Difference	df	Std. Error Difference	t	Sig
Water-absorption rate	0.23662	166	0.03891	6.082	0
Water-absorption velocity	0.06335	166	0.01135	5.583	0

. The results revealed significant differences in both parameters. Therefore, the conclusion that the longitudinal permeability of internodes is significantly superior to that of nodes has been statistically validated and is fully consistent with the macroscopic features shown in Figure 2a,b.

To further verify this trend, short-term water absorption tests were carried out on internodes and nodes, recording changes within 40 min (Figure 2c,d). Infrared thermography was also used to visualize longitudinal water penetration, with images shown in Figure 2e. Since nodes absorbed water more slowly, the imaging time was extended to 160 min. Results showed that the internode reached a stable absorption of 8%, while the node stabilized at only 2% (Figure 2c). The absorption rate analysis (Figure 2d) confirmed that the internode absorbed water rapidly at first and then slowed, whereas the node maintained a consistently low rate.

Thermal imaging further supported these results (Figure 2e). The color scale indicates that temperature decreased as water was absorbed. The internode showed rapid and uniform color change, with water migrating from bottom to top within 17 min. By contrast, the node changed slowly, with only slight color variation even after 160 min, suggesting diffusion-controlled penetration. Moreover, uneven color distribution around the diaphragm indicated structural restriction.

Figure 3A shows high-resolution scanning images of the cross-sections of bamboo nodes at different distances (0–6 mm) from the nodal diaphragm. It can be seen that at a distance of 6 mm, the arrangement of the vascular bundles in the cross-section is neat and the same as that in the internodes. As the distance decreases, the morphology of the vascular bundles begins to change, accompanied by the appearance of radial gaps (red circles). Within the radial gaps, special morphology of vascular bundles was observed, which might be sections of the bending vascular bundles. Other axial vascular bundles around them were compressed and deformed, and the arrangement position changed due to the radial gaps and bending of the vascular bundles. This indicates a reorientation of vessels at the node, consistent with Li et al. and Zhang et al. [38,39], who showed that vascular bundles form spiral branching networks near the diaphragm. Such structures increase pathway tortuosity, explaining the lower longitudinal permeability of nodes compared with internodes.

3.2. Dimensional Stability

The dimensional stability of nodes and internodes was compared by measuring the water-saturated swelling, air−dried shrinkage, oven-dried shrinkage, and eccentricity of the bamboo specimens, and the results are presented in Figure 4. Notably, during the drying process, nodes exhibited varying degrees of cracking, whereas the internodal rings remained intact, as shown in Figure 4b. Similar observations were reported by Chen et al. [40], who found that cracks in round, laminated, and flattened bamboo occurred at nodes in 72%, 77%, and 89% of cases, respectively. Overall, the complex structure, higher density, and greater anisotropic shrinkage of nodes lead to stress concentration during drying, making nodes the primary sites for crack initiation. These findings indicate that nodes are less dimensionally stable than internodes during the drying process [15,23].

The shrinkage and swelling data indicate that nodes exhibit lower dimensional stability than internodes under varying moisture content. In the water-saturated state (Figure 4a), the radial swelling of nodes reached 5.99%, significantly higher than the 2.66% observed for internodes. The difference was even more pronounced in the wall-thickness direction, with nodes swelling by 12.22% compared to 9.12% for internodes. In the air−dried state (Figure 4b), radial shrinkage of nodes was 5.24%, also higher than the 3.41% for internodes; however, in the wall−thickness direction, nodes showed slightly lower shrinkage (9.30%) than internodes (10.23%). Upon reaching the oven-dried state (Figure 4c), nodes exhibited greater shrinkage in all directions, with wall-thickness shrinkage reaching 15.71% compared to 13.12% for internodes. Overall, nodes showed significantly higher swelling and shrinkage in both radial and wall-thickness directions, indicating that the nodal region is the weak link in dimensional stability during moisture cycling.

These results are in good agreement with previous studies. Mou et al. [41] confirmed that during bamboo moisture absorption, nodes generally expand more in the radial and wall-thickness directions than internodes. Similarly, Anokye et al. [42] reported that nodes of different bamboo species often exhibit higher shrinkage than internodes during drying. From a microstructural perspective, Zhang et al. and Wang et al. [39,43] pointed out that the complex vascular arrangement, transverse vessels, and larger microfibril angles in nodes make them more sensitive to water migration and volumetric strain, resulting in greater swelling and shrinkage. Multi-scale experiments by Huang et al. [44] further revealed that differences in density and fiber orientation between nodes and internodes at various layers (inner, middle, outer) significantly affect the rate and magnitude of drying shrinkage, which aligns well with the observed high wall-thickness shrinkage of nodes in the oven-dried state.

Furthermore, as shown in Figure 4d, nodes and internodes exhibited differences in eccentricity across different moisture states. The eccentricity of nodes was slightly higher than that of internodes in the water-saturated (S−M−E), air-dried (A−D−S), and oven-dried (O−D−S) conditions. For instance, in the S−M−E state, node eccentricity was 0.32 versus 0.30 for internodes; in the A-D-S state, it increased to 0.34 for nodes while remaining 0.30 for internodes. Smaller eccentricity values indicate higher cross-sectional symmetry and better dimensional stability, whereas increased eccentricity reflects structural asymmetry, which can lead to localized stress concentration and uneven deformation under mechanical load or moisture variation. The higher eccentricity of nodes reflects lower structural symmetry, which is associated with the complex distribution of vascular bundles, irregular fiber arrangement, and heterogeneous tissue density and composition at the nodal region. Wang et al., 2024 [45] showed that irregular nodal structures lead to stress concentration and uneven deformation in radial stiffness and bending strength.

Moreover, Figure 3B shows that nodes are prone to microcracking during drying, which further disrupts the symmetry of their cross-sectional structure, leading to increased eccentricity and a marked reduction in structural stability in the oven-dried state. In contrast, the eccentricity of internodes remained around 0.30 across all three moisture conditions, indicating a more balanced structure with better symmetry and dimensional stability.

3.3. X-Ray Diffraction (XRD) Analysis

The results of XRD indicated that both internodes and nodes exhibited distinct diffraction peaks in the 2θ range of approximately 15–35° (Figure 5a), demonstrating that they share the same crystalline structure. Analysis of the diffraction patterns revealed that the crystallinity of internode fibers was 49.31%, while that of node fibers was 52.22% (Figure 5b). The slightly higher crystallinity of nodes compared to culms is consistent with the findings of Keisuke Toba et al., who reported differences in crystallinity among bamboo of different ages and regions [46]. Since water molecules primarily adsorb onto the free hydroxyl (-OH) groups in the amorphous regions of cellulose, the densely packed crystalline regions are less permeable, resulting in higher water absorption in internodes than in nodes [47].

Furthermore, the higher crystallinity of nodes, along with their greater rigidity and stress concentration, contributes to an increased tendency for cracking during dimensional stability tests under varying moisture conditions [48,49]. From a growth and physiological perspective, nodes serve as partitioning structures in bamboo and require enhanced structural stability to support the culm height and resist external forces. The higher crystallinity provides superior mechanical properties for nodes but correspondingly reduces their permeability. In contrast, the relatively lower crystallinity of internodes may facilitate water and nutrient transport, supporting bamboo growth and material exchange [48,49,50,51].

3.4. Pore Structure Analysis

Bamboo, as a typical porous medium [52], exhibits a distinct multiscale pore structure. According to the classification system of the International Union of Pure and Applied Chemistry (IUPAC), pores are generally divided into three categories: micropores (pore diameter < 2 nm), mesopores (2–50 nm), and macropores (>50 nm) [53,54]. Micropores are mainly located within the spaces between cellulose microfibrils and the lignin–hemicellulose matrix [55]. These structures primarily govern the adsorption capacity of bamboo for water molecules and other small substances. In contrast, mesopores are predominantly distributed in the middle lamella and in local indentations of the cell wall. They are closely associated with capillary condensation and thus play a crucial role in liquid retention and slow infiltration. Macropores, on the other hand, correspond to anatomical features such as vessel lumina, parenchyma cavities, and intercellular spaces [56]. These pores serve as the main channels for convective flow and rapid fluid transport. Therefore, the hierarchical pore system collectively determines both the sorption and permeability behavior of bamboo.

In this study, mercury intrusion porosimetry (MIP) was employed to analyze the pore structure differences between bamboo nodes and internodes, with the aim of elucidating their distinct water absorption behaviors. The results are summarized in

Table 2. MIP results of nodes and internodes.

Samples	Total Intrusion Volume (mL/g)	Total Surface Area (m2/g)	Median Pore Diameter (nm)	Mean Pore Diameter (nm)	Porosity (%)
Internode	0.6388 ± 0.03	85.97 ± 0.26	30.92 ± 0.11	29.72 ± 0.12	46.6736 ± 0.06
Node	0.6774 ± 0.02	95.08 ± 0.37	35.92 ± 0.16	28.50 ± 0.09	46.9996 ± 0.07

. According to the MIP measurements, the node samples exhibited slightly higher values of total intrusion volume (0.6774 ± 0.02 mL·g⁻¹), specific surface area (95.08 ± 0.37 m²·g⁻¹), and median pore diameter (35.92 ± 0.16 nm) compared with the internodes (0.6388 ± 0.03 mL·g⁻¹, 85.97 ± 0.26 m²·g⁻¹, and 30.92 ± 0.11 nm, respectively). By contrast, the average pore diameter of the nodes (28.50 ± 0.09 nm) was marginally lower than that of the internodes (29.72 ± 0.12 nm), while the overall porosity difference between the two was negligible. It is worth noting that the average pore size and median pore size represent different structural characteristics: the former indicates the overall scale of the pore system, whereas the latter highlights the dominant transport channels. The average pore size thus reflects the extent of structural development, while the median pore size governs capillary adsorption and liquid retention. The larger median pore size of the nodes indicates a reduced capillary driving force within individual pores, which in turn explains their slower initial water uptake rate (capillary absorption) compared with the internodes. This tendency is further evidenced by the results presented in Figure 2c–e.

In addition, although the nodes exhibit higher porosity and total pore volume than the internodes, their overall water absorption is lower. This phenomenon may be attributed to the partially transverse orientation of pores within the nodes. As shown in Figure 3A, the vascular bundle structures along the longitudinal direction of the nodes are partially bent transversely, and their connectivity is relatively poor. Consequently, water cannot easily permeate through these pores. This may explain why the nodes possess abundant porosity yet exhibit lower water uptake. This inference is supported by the mercury intrusion data in Figure 6a: under low initial pressures, the penetration of mercury in the nodes is limited, whereas at higher pressures, the intruded mercury volume surpasses that of the internodes.

Indeed, the node structure contains diaphragms (node plates/septa) and more complex partitions, which can slightly increase the total pore volume by creating additional micro- or mesopores, yet at the same time reduce the connectivity of the pore network. In addition, cellular blockages, such as closed lumina, fiber layer deposits, or tyloses, further obstruct the channels and limit effective pathways for rapid fluid transport. Moreover, the nodes have a higher specific surface area, implying the presence of more micro- and mesopore surfaces capable of adsorbing bound water [29]. Swelling and shrinkage of the cell wall induced by bound water can cause micro-scale dimensional changes. If local water exchange is restricted due to limited connectivity, stress can accumulate, ultimately leading to cracking. This may explain why nodes are more prone to cracking.

Figure 6b,c illustrate the differences in pore size distribution between nodes and internodes. From the ΔV curve (Figure 6c), it can be seen that the internodes have significantly higher pore volume in the micropore range (<26 nm) compared with nodes, while their pore volume in the mesopore range (26–675 nm) is lower. In the macropore and super-macropore range (>105 nm), internodes again exhibit larger pore volume. These results suggest that internodes are particularly rich in micropores and super-macropores, whereas nodes have more developed structures in the mesopore range.

Correspondingly, the dV/dlogd curve (Figure 6b) shows a sharper distribution peak around 23 nm for the internodes, indicating a more concentrated pore size distribution. The nodes also exhibit a notable peak near 21 nm, but it is lower and more broadly distributed. Taken together, these analyses indicate that internodes are especially rich in micropores, implying denser fibers or more developed cell wall microstructures. These micropores are primarily associated with capillary effects. In contrast, macropores and super-macropores correspond to anatomical features such as vessel lumina, parenchyma cavities, and intercellular spaces, serving as main channels for convective flow and rapid fluid transport. Consequently, micropores and super-macropores favor small-molecule adsorption and fast transport, respectively. Moreover, the broader mesopore size distribution in nodes suggests structural heterogeneity, which may contribute to localized stress concentration during drying and swelling.

3.5. Analysis of Chemical Composition

Table 3. Determination Results of the three major elements of internodes and nodes.

Samples	Cellulose	Hemi- Cellulose	Acid-Insoluble Lignin	Acid-Soluble Lignin	Lignin	The Three Major Substances
Internode	39.4 ± 0.33%	20.1 ± 0.29%	25.0 ± 0.37%	1.4 ± 0.11%	26.4 ± 0.19%	85.9 ± 0.27%
Node	37.3 ± 0.21%	23.5 ± 0.31%	25.4 ± 0.41%	2.4 ± 0.09%	27.8 ± 0.23%	88.6 ± 0.31%

shows the contents of the three major elements in bamboo nodes and internodes. The results show that there are significant differences in the composition of the three major elements between internodes and nodes. Specifically, the cellulose content of internodes (39.4 ± 0.33%) is slightly higher than that of nodes (37.3 ± 0.21%), while the hemicellulose (23.5 ± 0.31%) and lignin (27.8 ± 0.23%) contents of bamboo nodes are significantly higher than those of internodes (20.1 ± 0.29% and 26.4 ± 0.19%). This discovery is highly consistent with the research results of Li et al. In their research on the cytochemical components of different structures of Moso bamboo, Li et al. also found that the cellulose content in bamboo nodes was lower than that in internodes [56], while the hemicellulose content was higher than that in bamboo culms. This difference in chemical composition is a structural adjustment made by bamboo nodes to adapt to their role as plant support points and branching points. The higher total sum of the three major elements (88.6 ± 0.31%) in bamboo nodes also indicates that the cell wall material is denser, which lays the foundation for its unique physical and mechanical properties.

In practical applications, bamboo nodes often exhibit a contradictory phenomenon of low water absorption rate but poor dimensional stability. This characteristic can be explained from both its unique chemical composition and anatomical structure. Firstly, the high lignin content of bamboo nodes endows them with stronger hydrophobicity. Lignin encapsulates hydrophilic cellulose, effectively hindering the penetration of water molecules, thereby resulting in a relatively low overall water absorption percentage. Secondly, the poor dimensional stability is due to its complex anatomical structure. Unlike the basically parallel arrangement of vascular bundles on internodes, the vascular bundles inside nodes are arranged in a complex cross and spiral pattern. When water enters the material, hydrophilic cellulose and hemicellulose will expand. Due to the different arrangement directions of the fibers inside the bamboo nodes, this expansion is uneven in different directions, that is, anisotropic expansion occurs. This uneven expansion will generate huge internal stress, making the bamboo nodes more prone to twisting, cracking and deformation during the moisture–drying cycle, thus showing poor dimensional stability.

4. Conclusions

This study systematically compared the water absorption, dimensional stability, and pore structure of nodes and internodes in Phyllostachys edulis. The results show that nodes exhibit significantly lower water absorption than internodes, with pores concentrated in the medium-to-large range and longitudinal permeability limited by diaphragms, transverse vascular bundles, and spiral branch networks. Higher cellulose crystallinity and lignin content further enhance hydrophobicity, restricting water transport. Interestingly, despite their low water uptake, nodes display poorer dimensional stability during wetting–equilibrium–oven-drying cycles, often developing cracks and deformations, which is attributed to the complex, crossed arrangement of fibers and vascular bundles and uneven pore distribution.

This apparent contradiction indicates that low water absorption does not guarantee high dimensional stability. The anisotropic swelling of fibers and heterogeneous pore structure in nodes generate localized internal stresses, promoting twisting, cracking, and macroscopic deformation. In contrast, internodes, with more uniform fiber alignment and interconnected pores, can rapidly absorb water while maintaining superior dimensional stability. These findings reveal the intrinsic structural–functional disparities between nodes and internodes, providing a scientific basis for bamboo functionalization and “bamboo-for-plastic” material applications.