Introduction

Wood fiber is made of multi-layered cell walls which approximately consist of two-phase structure, i.e., rigid crystal of cellulose microfibrils (CMFs) and flexible matrix substances (MTs) of lignin, hemicellulose, and noncrystalline cellulose. The CMF, the size of which is a few nanometers in width [1], consists of dozens of β-1,4-glucan chains. Particularly in the secondary wall, which occupies most part of the cell wall, the CMF accumulates in a direction nearly parallel to the longitudinal axis of the wood fiber, while MT fills in the gaps between CMFs in each layer.

X-ray diffraction (XRD) technique is available for collecting information on the microstructure of crystalline materials [2, 3]. For example, some authors have adopted the XRD technique to detect the crystal lattice deformation induced by tensile or bending stress [4–6], moisture adsorption [7–15], cell wall maturation [16–18], and so forth. Among them, change in lattice spacing due to moisture adsorption is a quite interesting phenomenon. Toba et al. [15] pointed out the possibility that the CMF is laterally expanded by the shrinking MT during drying, and that binding force between CMF and MT is reduced after giving repeated drying and re-swelling treatments to the wood cell wall. Recently, some authors showed that repeated drying and re-swelling increase the crystallinity in wood cell wall [13, 19, 20]. These studies contain the information on the microscopic properties of the wood cell wall; thus, it is rather natural to consider that crystal lattice deformation is more or less correlated with increasing of the crystallinity, which is induced by repeated moisture change. However, no studies have focused on this problem yet, thus whole picture of fine structure in wood cell wall is still unaddressed.

It is well known that chemical treatments induce substantial change in the microscopic region of wood cell wall. For example, Jungniki et al. [21] investigated the influences of delignification and subsequent alkali treatments on the spruce wood cell wall in terms of nanostructural organization, and they reported that visible size of CMF aggregation changed in the courses of chemical treatments. Fahlén and Salmén [22] observed that kraft pulp processing increased formation of nanoscale pores in wood cell wall, and at the same time, aggregation of cellulose fiber also became larger as observed by atomic force microscope. These studies imply that chemical treatment induces the changes in wood cell wall at the microscopic level, thus providing important information on the microscopic structure and behaviors of the interface between CMF and MT.

The purpose of the present study is to further understand the effects of repeated drying and re-swelling on the mechanical interaction between CMF and MT in wood cell wall, following our previous reports [18, 20]. Thus, we prepared specimens which were subjected to various types of pretreatments, that is, chemical extraction as well as hygrothermal treatment to give some damages to principal constituents of wood cell wall, followed by repeated dry-and-wet treatments. Based on the measurements of d-spacing of (200) lattice plane of cellulose Iβ (d 200), the crystallinity, and the crystal size, we tried to clarify the microscopic behaviors of cell wall constituents, as well as interfacial structure between CMF and MT.

Materials and methods

Tested tree

The specimens studied were prepared from sap wood of 39-year-old sugi (Cryptomeria japonica D. Don), grown in Nagoya University Forest, Japan. The diameter at breast height was 24.0 cm. Oven-dried density was 0.31 ± 0.02 g/cm3 (mean ± SD). Quarter-sawn specimen with dimensions of 2 × 16 × 15 mm (thickness × length × width) was made from the tested tree. Figure 1 showed the process flow of subsequent treatments in the present study. For the purpose of inducing selective damages to the microscopic region in the cell wall, prior to repeated dry-and-wet (RDW) treatments, never-dried specimens were divided into eight groups without drying and were subjected to differing pretreatments: untreated (control), ethanol and benzene extraction, light and intensive delignification (2 or 6 h), light and intensive alkali extraction (soaked in 3 or 10 % of aqueous NaOH solution), and light and intensive hygrothermal treatments (120 or 140 °C). The details are as follows.

Fig. 1
figure 1

Process flow for each type of pretreatment and subsequent RDW treatment. Rectangles with dashed lines indicate the processes that were repeated

Full size image

Ethanol and benzene extraction

Ethanol and benzene extracted specimens were prepared using Soxhlet extractor with refluxing a mixture of ethanol and benzene (1:2 v/v) for 24 h. After the extraction, mixed solvent was replaced by pure ethanol and then was replaced by distilled water in stages every second days (pure ethanol, 95, 90, 70, 50, 30, 10, and 0 %).

Delignification

Delignification specimens were prepared based on improved acid-chlorite method [23], which uses acetic acid buffer solution for the purpose of inhibiting outflow of polysaccharide during treatment. After 2 or 6 h of treatments, specimens were rinsed with running water for 1 week.

Alkali extraction

Alkali extracted specimens were prepared by soaking in 3 or 10 % concentration of aqueous NaOH solution for a month at room temperature. After the extraction, specimens were rinsed with running water for 1 week.

Hygrothermal treatment

Hygrothermal treated specimens were prepared using an autoclave at 120 or 140 °C for an hour. After the treatment, specimens still in the autoclave were cooled to room temperature.

Repeated dry-and-wet treatment

After giving each pretreatment, all specimens were conditioned in a sealed desiccator with distilled water, 100 % RH, for about 1 month until they reached equilibrium, which was considered as the fiber saturation point (FSP). The FSP is defined as the moisture content of wood specimen when cell wall is filled with bound water [24].

Subsequently, all specimens were given the RDW treatment. They were seasoned to the dry condition in a sealed desiccator with phosphorous (V) oxide for 1 week without heating, and then, they were soaked in distilled water at room temperature to make them re-swell again. This process of drying and re-swelling was repeated 5 times.

XRD measurements

XRD measurements were performed to assess the d-spacing of (200) lattice plane of cellulose Iβ (d 200), the crystallinity, and the crystal size of (200) lattice plane of crystalline cellulose in wood cell wall at the FSP before and after pretreatments, and at each dry condition in the process of RDW treatments. Figure 2 describes a schematic of d 200 of the crystalline cellulose in the wood cell wall. In the present study, detections of the peak positions of 110 and \( 1\bar{1}0 \) diffractions were not performed. The authors consider that peak separation between 110 and \( 1\bar{1}0 \) diffractions might cause some errors because these two peaks are too close to each other in wood cell wall. In addition, 200 diffraction peak is sharper than 110 and \( 1\bar{1}0 \). Thus, we consider it is better to measure the d 200 as an indicator of the transverse directional change of the cellulose crystals.

Fig. 2
figure 2

Cellulose microfibrils which compose of wood cell wall and d 200 in crystalline structure

Full size image

An X-ray diffractometer (XD-D1w, Shimadzu Co., Japan) was used to measure the crystalline properties under the following conditions. All XRD measurements were made with the reflection technique. The incident X-ray radiation was the Cu Kα characteristic X-ray passing through a nickel filter with a power of 30 kV and 30 mA. Both the air-scattering prevention slit and the divergence slit were 1°. The width of the detection slit was 0.1 mm. Scanning speed and integration time in the measurement of d 200 were 0.125°/min and 0.4 s, respectively. On the other hand, those were 2.0°/min and 2.0 s in the measurement of the crystallinity and the crystal size, respectively. Scanning ranges of 2θ were 21.5°–23.2° in the d 200 and 5.0°–40.0° in the crystallinity and the crystal size. Quarter-sawn specimens were vertically fixed on the sample holder of XRD device to irradiate radial section. Moisture-conditioned air was circulated through the specimen cabinet mounted on the turntable to prevent the specimen from changing moisture content during measurements.

The d 200 was calculated using the Bragg equation, as follows:

$$ d_{200} = \frac{\lambda }{{2\sin \theta_{200} }} $$
(1)

where \( \lambda \) is the incident X-ray wavelength (0.1542 nm), and \( \theta \) 200 is the Bragg angle (°) on the basis of the peak position of 200 diffraction obtained from the diffraction diagram. The peak position of silicon powder was used to remove the error in 2θ angle measurements. To confirm the significant figure in the measurement of diffraction angle, we performed repeated measurements using silicon powder (2θ ≅ 28.445°). As a result, significant value in 2θ angle measurement was limited in the range of 0.0001°.

Since wood contains a large amount of crystalline cellulose, we can regard the wood as a crystalline material. In the present study, the degree of crystallinity in wood cell wall was calculated using Segal method [25], as follows:

$$ C_{\text{S}} = \frac{{I_{200} - I_{\text{Am}} }}{{I_{200} }} \times 100\,(\% ) $$
(2)

where C S was the crystallinity (%) of wood cell wall, I 200 is the reflection intensity of 200 diffraction, and I Am is the intensity at the minimum near 18.5° of 2θ angle. Segal method is simple to be applied and do not require peak separation between 110 and \( 1\bar{1}0 \) diffraction. We consider that Segal method is sufficient to estimate the crystallinity in the case of investigating not the absolute content of crystal but the qualitative changes of cellulose in wood cell wall.

The size of cellulose crystal was calculated on the basis of the Scherrer equation [3], as follows:

$$ D_{200} = \frac{K\lambda }{{B\,\cos \,\theta_{200} }} $$
(3)

where D 200 is the crystal size (nm) for (200) plane diffraction, K the Scherrer constant (0.9), λ the incident X-ray wavelength (0.1542 nm), B the half bandwidth in radians, and θ 200 is the Bragg angle (°) for (200) plane.

Results and discussion

d-Spacing of (200) lattice plane of cellulose Iβ (d 200)

Figure 3 shows variations of d 200 induced by each pretreatment and subsequent RDW treatments. Based on the changes of d 200 due to treatment, crystal strain (ε 200) was calculated from the following equation:

$$ \varepsilon_{200} = \frac{{d_{200}^{'} - d_{200}^{\text{Green}} }}{{d_{200}^{\text{Green}} }} \times 100\,(\% ) $$
(4)

where \( d_{200}^{\text{Green}} \) is d 200 (nm) of green specimen at the FSP and d′200 is the one after giving each pretreatment or subsequent RDW treatments. Table 1 shows the results in ε 200.

Fig. 3
figure 3

Variations of d 200 induced by each pretreatment and subsequent repeated dry-and-wet (RDW) treatments. In this figure, “A” indicates that there was significant difference <5 % level between before and after pretreatment. “a” indicates that there were significant differences <5 % level compared to the fifth RDW treatment. “b” shows no significance compared to the fifth RDW treatment. A gray solid line indicates the level of d 200 after the fifth RDW treatment in each case

Full size image
Table 1 Crystal strains ε 200 (%) induced by each pretreatment and subsequent RDW treatments
Full size table

First, slight increases in d 200 were observed after each pretreatment (Fig. 3b–h). The ε 200 was 0.09–0.19 % in each pretreatment, as shown in Table 1. In our previous paper, we reported that boiling treatment changed the d 200 in the green wood cell wall, which was caused by relaxing the growth stress remaining in microscopic region of wood cell wall; as a result, 0.15–0.18 % of ε 200 was observed [18]. This coincides with ε 200 observed after pretreatments in the present study. These suggest that each pretreatment in the present study would act in a similar manner to boiling treatment, leading to the increase in d 200.

Figure 3a shows the results of untreated specimen. The d 200 largely expanded due to initial drying. However, it gradually decreased and then became constant in the process of RDW treatments. This trend of variation in d 200 was reported also in our previous study [15]. In the present study, it was found that light pretreated specimens (Fig. 3b, c, e, g) showed similar trend of variation in d 200 as untreated specimens, while intensive pretreated specimens (Fig. 3d, f, h) showed constant value at a relatively small level throughout the whole process of RDW treatments. It should be noted that d 200 in all specimens became uniform 0.3988 ± 0.0002 nm after giving 5 times of RDW treatments.

We consider that shrinking MT in untreated or light pretreated specimens has an ability to give a lateral tensile stress to the CMF in the drying process (Fig. 3b, c, e, g). As a result, d 200 largely expands with drying. However, this tensile stress was gradually reduced in the process of RDW treatments. On the other hand, tensile stress in intensive pretreated specimens has been already weakened from the beginning of RDW treatment (Fig. 3d, f, h). This reduction of tensile stress to the CMF at drying suggests that repeated moisture changing produced the same effect as that of giving intensive pretreatment.

On comparing the average value of ε 200 before and after giving 5 times of RDW treatments, it was reduced up to 26 %. If the changes in d 200 would be a simple elastic deformation, it is natural to consider that RDW treatments decreased the ability of the MT to give lateral tensile stress to the CMF at the drying process. Concretely to say, each pretreatment or subsequent RDW treatments gave substantial damage to the cell wall in microscopic region, such as outflow or degradation of the constituents in the MT, interfacial separation between CMF and MT, and so forth. These damages weakened mechanical interaction between CMF and MT. It is considered that the above differences in values of d 200 in the earlier stages of RDW treatments depend on the extent of the defect caused by pretreatments.

Crystallinity and crystal size

Figures 4 and 5 show variations of the crystallinity and the crystal size induced by each pretreatment and subsequent RDW treatments, respectively. The crystallinity and the crystal size at the FSP increased after each pretreatment (Figs. 4b–h, 5b–h). Outflow or degradation of the MT is one of the possible factors to cause increases in the crystallinity and the crystal size. Increases of the crystallinity and the crystal size were also shown in Figs. 4b and 5b (24 h of ethanol and benzene extraction treatment). However, sugi contains at most 3 % of extractives, which are soluble in mixture of ethanol and benzene. The amounts are much lower than those of cellulose, hemicellulose, and lignin. This indicates that increases in the crystallinity and the crystal size, as shown in Figs. 4b and 5b, were not merely attributed to removal of ethanol and benzene soluble extractives, but also to changing in the relative ratio of crystalline component in whole wood cell wall.

Fig. 4
figure 4

Variation of crystallinity of wood cell wall induced by each pretreatment and subsequent repeated dry-and-wet (RDW) treatments. “A”, “a” and “b” show same meaning as in Fig. 3. A gray solid line indicates the level of the crystallinity after the fifth RDW treatment in each case

Full size image
Fig. 5
figure 5

Variation of crystal size induced by each pretreatment and subsequent repeated dry-and-wet (RDW) treatments. “A”, “a” and “b” show same meaning as in Fig. 3. A gray solid line in each case indicates the level of the crystal size after the fifth RDW treatment

Full size image

Thereafter, both the crystallinity and the crystal size decreased due to subsequent drying in all specimens. After that, RDW treatments were performed on the specimens. On the whole, the crystallinity showed two types of variations in the process of RDW treatments. Some showed smaller values of the crystallinity in the early stage of RDW treatments, and then, it became higher in the later stage (Fig. 4a, b, c, e, g). Others kept higher values from the beginning throughout 5 times of RDW treatments (Fig. 4d, f, h). Latter type was observed in specimens which were given intensive pretreatments. However, the crystallinity in both types became constant at the end of RDW treatments. It is considered that differences in the crystallinity between the above two types reflect the extent of damage of MT caused by each pretreatment. On the other hand, the crystal size kept increasing even after the crystallinity became constant in all specimens (Fig. 5).

The results of crystalline properties suggest that RDW treatments have similar effects as pretreatments, which induce substantial damages to the microscopic region in wood cell wall. In addition, repeated changing of moisture content in wood cell wall creates the condition to promote the crystallization of cellulose in wood cell wall, as reported in our previous paper [20]. We consider that non-crystalline cellulose, which contains a large amount of free hydroxyl groups, forms additional hydrogen bonding in the process of RDW treatments. Extent of hydrogen bonding increased after repeated moisture changing. Intensive pretreatments, which produced certain damages in microscopic region of the cell wall (e.g., interfacial separation between the CMF and MT), create the condition to accelerate hydrogen bonding formation in noncrystalline cellulose. This is the reason why the crystallinity of wood cell wall showed higher values from the beginning of RDW treatment in intensive pretreated specimens (Fig. 4d, f, h).

Our idea remains a matter of speculation. However, it would be natural to consider that some sorts of substantial damages arise in microscopic region in wood cell wall because absolute dried weight of specimen remained constant throughout 5 times of RDW treatments. For total understanding of the mechanism of phenomena observed in this study, step-up investigation will be required, e.g., comparing the physical properties of RDW-treated wood with thermal-dried wood, aged wood, and so forth [26, 27].