Porosity analysis of three types of balsa ( Ochroma pyramidale ) wood depending on density

Balsa ( Ochroma pyramidale ), which is the lowest density wood, is a useful species in various industries. In general, balsa can be divided into three types (low density: LD, middle density: MD, high density: HD). In this study, we classified the porosities of the three balsa types into through‑pore porosity, blind‑pore porosity, and closed‑pore porosity. As a result, the total porosity of balsa showed a high positive ( + ) correlation with closed‑pore porosity, but total porosity of balsa showed negative ( − ) correlations with blind‑pore porosity and through‑pore porosity. Such information can be useful when using balsa as a natural porous material.


Introduction
Balsa (Ochroma pyramidale), a typical low-density wood, is used widely in industrial applications that require lightweight materials such as thermal insulation, wind turbine blades, and manufacturing of aircraft [1,2]. Moreover, balsa is the preferred material for high-rise sandwich panels [3,4].
When balsa is used as a core material, its elastic properties are very important. From a prior study, the axial compressive Young's modulus and strength of balsa tend to increase linearly depending on density. In other words, the mechanical capability of balsa is dependent on its density [4][5][6].
Balsa's density range is wide, from approximately 0.06 to 0.38 g cm −1 . Typically, balsa can be separated based on density: low density (LD) less than 0.1 g cm −1 , middle density (MD) of 0.1-0.2 g cm −1 , and high density (HD) greater than 0.2 g cm −1 [3]. Normally, wood density is constant, and the lower is the density of the wood, the higher is its porosity [7].
The pore analysis of wood is critical as a basis for studying the macro-physical properties of wood [8]. Methods for analyzing wood's pore structure include an anatomical method using a microscope and physical methods using fluid or gas [8,9].
Borrega et al. [10] analyzed the anatomical features of LD (0.064 g cm −1 ), MD (0.163 g cm −1 ), and HD (0.274 g cm −1 ) balsa by electron microscopy. The lengths of the major axes of the vessels were approximately 220.4 for LD, 320.8 for MD, and 258.8 μm for HD, respectively, and the lengths of the minor axes were approximately 156.6 for LD, 251.0 for MD, and 206.6 μm for HD, respectively. In addition, balsa's fiber volume fraction was 76.3 for LD, 73.5 for MD, and 66.4 for HD. Pan et al. [11] analyzed the pore structure of balsa wood by mercury intrusion porometry and reported that the pore size was 0.1-100 μm, with porosity of approximately 89.5%.
However, few previous studies have analyzed pore structure of the three types of balsa (LD, MD, and HD) using physical methods. Therefore, we focused on the analysis of pore structure of three types of balsa using a physical method.
According to the International Union of Pure and Applied Chemistry (IUPAC) [12], pores are classified as through pores when both ends are connected to the external environment, as blind pores when only one side is connected to the outside, or as closed pores when there is no external connection.
Wood is an anisotropic material with different anatomical characteristics depending on the cut direction [13]. The surfaces of tangential and radial planes are almost non-porous; however, the surface of the transverse plane displays a pore structure. Thus, permeability of the transverse section is much higher than that in the other two directions [14]. However, most timber industries use tangential and radial planes. From the point of view of porosity, when wood is used in the tangential and radial planes, the pore structure of the wood is closed. Unlike other wood species, balsa is used in the transverse plane in sandwich panels. Therefore, classification of porosity is important for understanding balsa's porous properties.
Thus, in this study, the pore structure of the balsa transverse section was investigated. Accordingly, three types of balsa were prepared according to density, and porosity analysis (through pore, blind pore, and closed pore) was performed. It is expected that this study will be of great utility to wood researchers. Figure 1 depicts the preparation of specimens in this study. Three types of North American Balsa (Ochroma pyramidale) logs were prepared. Their densities were 0.13 ± 0.03 (± standard deviation) for LD, 0.22 ± 0.01 for MD, and 0.39 ± 0.01 g cm −3 for HD. The logs were milled into 29-mm-diameter rods and cut into 10-mm-thick cylindrical specimens with a table saw. In order to conduct porosity analysis, 13 specimens for each type were selected. In addition, for morphology analysis, one specimen of each type was selected. It was confirmed by visual inspection that none of the specimens contained knots. Balsa logs were supplied from the Woodworker's Emporium (Las Vegas, NV, USA). Specimen production was carried out in Saehan Timber Co., Ltd. (Gyeonggi-do, Ilsan, Korea). We kept the samples in a standard laboratory environment at 20 °C and 50% humidity for 2 weeks. The moisture content (MC) of the specimens was 6.67% for LD, 6.94% for MD, and 7.38% for HD.

Morphological analysis by SEM (scanning electron microscopy)
The transverse planes of the three types of balsa were observed with SEM (model: Genesis-1000, Emcrafts, Korea) at 100× magnification for vessels and 10,000× magnification for cell walls. The specimen Fig. 1 Specimen preparation of balsa based on density pretreatment process was as follows: Specimens were cut to a size of about 5 (W) × 5 (D) × 5 (H) mm cube. They were immersed in 10% ethanol and decompressed to -0.1 MPa in a vacuum chamber for 30 min. Next, the surface was shaved by a microtome, and the cube specimens were dried in an oven at 40 °C for 5 h. Finally, they were coated with gold ions. We measured vessel diameter, vessel area, and cell wall thickness using an image software (model: Virtuoso ver 1.1, Emcraft, Korea).

Pore size measurement
We measured the pore sizes of three types of balsa using a capillary flow porometer (model: CFP-1200AEL, PMI Inc, Ithaca, NY, USA) based on the bubble point method of ASTM F-316 [15]. The specimens were immersed in wetting liquid (model: Galwick, PMI Inc, Ithaca, NY, USA), and air pressure was applied to each. The initial pressure at which the wetting liquid was extruded from the specimen was detected. The pore size is calculated as in Eq. 1: where D = pore size (μm), τ = surface tension of wetting liquid (dyne cm −1 ), p = pressure (psi), and C = constant (0.415).

Porosity classification
Porosity refers to the pore volume ratio of a porous material. Therefore, the total porosity ( φ total ) of a cylindrical specimen is defined by Eq. 2: where, for convenience, ρ = density of the cylindrical specimen, and ρ true = true density of the wood (assumed to be 1.5 g cm −3 ) [7].
The closed-pore porosity ( φ closed ) and open-pore porosity ( φ open ) of the cylindrical specimen can be measured with a gas pycnometer (PYC-100A-1, Porous Materials Inc. USA) [16]. Open-pore porosity is the sum of blind-pore porosity ( φ blind ) and through-pore porosity ( φ through ) (Eq. 3): In our previous work, we proposed a method to distinguish between blind and through pores [17][18][19][20]. We followed the same procedure in this study. First, the cylindrical specimen was wetted in Galwick solution (Porous Materials Inc. USA), which is assumed to penetrate the open pores. Galwick solution has a low surface tension (15.9 dyne cm −1 ) and is absorbed very quickly.
Next, the cylindrical specimen was placed in a hollowbottom chamber, and air pressure was applied longitudinally. It was assumed that only Galwick solution in the through pores of the cylindrical specimen was extruded. Finally, the cylindrical specimen was weighed. The blindpore volume (V blind ) of the cylindrical specimen was obtained by calculating the difference in cylindrical specimen weight before and after wetting with Galwick solution (Eq. 4): where m 0 = specimen weight (g), m 1 = specimen weight after Galwick solution extruded (g), and ρ Galwick = density of Galwick solution (1.79 g cm −3 ).
The blind-pore porosity of a cylindrical specimen can be calculated by Eq. 5: where V total = specimen volume (cm 3 ).
The through-pore porosity of a cylindrical specimen can be calculated by Eq. 6: In addition, we performed a one-way analysis of variance (ANOVA), and Duncan's multiple range tests determined the significance of the difference at a 10% level (p < 0.1). Figure 2 shows SEM images of the balsa transverse plane depending on specimen density, while Table 1 provides vessel diameter, vessel area, and cell wall thickness in three types of balsa wood.

SEM images of balsa depending on density
Their vessel arrangement was in the shape of diffuseporous wood. As for the major axis of the vessel, LD and MD were larger than HD. In minor axis, MD was larger than LD and HD. The vessel area also showed that MD was larger than LD and HD. The cell wall thickness was thinner in LD than in MD and HD. Figure 3a shows the pore sizes of the three types of balsa. The pore size of LD was 148.57 ± 13.47 c , 138.69 ± 12.27 b for MD, and 117.33 ± 3.98 a μm for HD. As a result, pore size of the HD is higher in all three types of balsa (different letters denote significant differences between groups at a 10% level). Figure 3b shows the three types of balsa's porosity analysis. The total porosity of the balsa calculated from   Eq. 2 was 91.54 ± 1.73 c for LD, 85.56 ± 0.14 b for MD, and 74.36 ± 0.20 a % for HD.

Results of pore size and porosity analysis
In the porosity analysis, the through-pore porosities of HD and MD were greater than that of LD (p < 0.01), and the difference between HD and MD was not statistically significant (p = 0.433). However, the closed-pore porosity was in the order of LD > MD > HD (p < 0.01). Table 2 provides results of Pearson's correlation analysis. Total porosity showed a strong positive (+) correlation with closed-pore porosity. However, total porosity showed negative (-) correlations with blind-pore porosity and through-pore porosity.
In solid porous materials, closed-pore porosity is related to thermal conductivity [21,22], which indicates that the higher is the closed-pore porosity, the greater is the thermal insulation of the porous material. Therefore, to predict the thermal characteristics of the transverse plane of balsa, it is possible to make a prediction based only on total porosity without needing to calculate closed-pore porosity.
The longitudinal permeability of wood is related to properties such as drying, heat transfer, and sound absorption [23][24][25][26][27][28]. According to our previous studies, gas permeability increases when through-pore porosity is high [17][18][19][20]. Therefore, when predicting these properties of balsa, through-pore porosity rather than total porosity should be considered as the main parameter.
Consequently, HD can be more suitable if the balsa transverse plane is to be used as a sound absorbing material. However, if it is to be used as an insulator, LD would be more suitable. These findings need to be confirmed in the future.

Conclusions
1. The porosity of balsa can be classified into throughpore porosity, blind-porosity, and closed-pore porosity. 2. The density of balsa showed the same trend as closed-pore porosity. 3. Through-porosity of HD was the greatest among the three types of balsa.

Funding
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2019R1I1A3A02059471) and was supported under the framework of an international cooperation program managed by the NRF of Korea (NRF-2020K2A9A2A08000181).

Availability of data and materials
Not applicable. Through-pore porosity Blind-pore porosity