*Article* **Improved On-Site Characterization of Arsenic in Gypsum from Waste Plasterboards Using Smart Devices**

**Masamoto Tafu 1,\* , Juna Nakamura <sup>2</sup> , Momoka Tanii <sup>2</sup> , Saori Takamatsu <sup>1</sup> and Atsushi Manaka <sup>1</sup>**


**Abstract:** The impurities in waste plasterboards, a product of ethical demolition, are a serious problem for their recycling. Plasterboards, the wall materials used in old buildings, are often recycled into gypsum powder for various applications, including ground stabilization. However, this powder contains various chemical impurities from the original production process of the gypsum itself, and such impurities pose a risk of polluting the surrounding soil. Here, we present a simple method for verifying the presence of arsenic, a harmful element in recycled gypsum that is suitable for use at demolition sites. First, we developed a simple pretreatment method using a cation-exchange resin to dissolve insoluble gypsum suspended in water by exploiting a chemical equilibrium shift, and we estimated the quantity suitable for releasing the arsenic from arsenic-containing gypsum. This pretreated solution could then be tested with a conventional arsenic test kit by observing the color changes in the test paper using the image sensor of a smart device. This simple method could determine a wide range of arsenic quantities in the gypsum, which would be helpful for monitoring arsenic in recycled gypsum powder, thereby supporting the development of a safe circular economy for waste plasterboards.

**Keywords:** plasterboard; arsenic; recycling; on-site determination

#### **1. Introduction**

Plasterboards consisting of solidified gypsum (calcium sulfate dihydrate, CaSO4·2H2O) between paper sheets are widely used as wall materials in houses constructed using the 2 × 4 (two-by-four) method. In Japan, the lifetime of houses is approximately 40 years, and when houses are demolished, the plasterboards are collected for recycling. Specifically, the reclaimed gypsum is pulverized, treated, and used in new plasterboards. However, this recycling process is limited, and most gypsum in plasterboards is derived from mining (natural gypsum) as well as the byproducts of various chemical plant processes (chemical gypsum) and flue gas desulfurization (FGD), as shown in Figure 1. Specifically, chemical gypsum originates from phosphate and fluoride production and smelting, whereas FGD gypsum is a byproduct of thermal power plants using coal and heavy oil. Moreover, it potentially contains chemical impurities, including fluoride, arsenic, and cadmium, derived from the raw materials used in these chemical processes. As shown in this figure, gypsum is also is widely used as a component of Portland cement, as well as a ground stabilizer to improve ground hardness, which has been exhaustively studied. However, because gypsum in plasterboards is supplied from various sources, the recycled gypsum from waste plasterboards poses a risk of soil pollution by potentially releasing fluoride, arsenic, and cadmium into the surrounding soil.

**Citation:** Tafu, M.; Nakamura, J.; Tanii, M.; Takamatsu, S.; Manaka, A. Improved On-Site Characterization of Arsenic in Gypsum from Waste Plasterboards Using Smart Devices. *Materials* **2022**, *15*, 2446. https:// doi.org/10.3390/ma15072446

Academic Editors: Ana Mladenovic and Claudio Ferone

Received: 26 January 2022 Accepted: 22 March 2022 Published: 26 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Figure 1.** Material flow of gypsum in Japan. FGD: flue gas desulfurization. **Figure 1.** Material flow of gypsum in Japan. FGD: flue gas desulfurization.

In particular, the arsenic in waste gypsum has severe environmental effects. As a hazardous waste product of the metallurgical industry, arsenic-bearing gypsum (ABG) is derived from the lime neutralization of waste acid liquor [1,2]. Some amount of ABG was used in plasterboards in Japan from 1973 to 1997, and in 2017, approximately 1.5 × 104 tons of ABG were separated from abandoned buildings in Japan [3]. Waste plasterboards containing ABG must be carefully collected during demolition, and ABG-containing waste plasterboards must be identified to safely recycle gypsum for ground stabilization to avoid polluting the soil. Ideally, this identification is carried out directly at construction In particular, the arsenic in waste gypsum has severe environmental effects. As a hazardous waste product of the metallurgical industry, arsenic-bearing gypsum (ABG) is derived from the lime neutralization of waste acid liquor [1,2]. Some amount of ABG was used in plasterboards in Japan from 1973 to 1997, and in 2017, approximately 1.5 <sup>×</sup> <sup>10</sup><sup>4</sup> tons of ABG were separated from abandoned buildings in Japan [3]. Waste plasterboards containing ABG must be carefully collected during demolition, and ABG-containing waste plasterboards must be identified to safely recycle gypsum for ground stabilization to avoid polluting the soil. Ideally, this identification is carried out directly at construction sites, which could lead to the development of a safer circular economy for recycled gypsum from waste plasterboards.

sites, which could lead to the development of a safer circular economy for recycled gypsum from waste plasterboards. One on-site determination method for analyzing the arsenic in gypsum is X-ray fluorescence (XRF) [4,5], but this method requires skilled handling and/or a license to operate the radiation apparatus. To overcome this problem, we focused on adapting facile commercial test kits for determining the arsenic contents in an aqueous solution, a method that can be employed in the field. Because of the low solubility of gypsum, determining its arsenic levels requires a pretreatment to dissolve the gypsum into a homogeneous solution through a pyrolysis process involving harmful chemicals (such as hydrochloric acid [6] or perchloric acid [4]). After pretreatment, arsenic is released into the solution in a form suitable for detection, and conventional analytical methods can then be applied. We previously demonstrated that gypsum was easily dissolved in water containing cation- and anion-exchange resins and that the fluoride content in gypsum could be successfully de-One on-site determination method for analyzing the arsenic in gypsum is X-ray fluorescence (XRF) [4,5], but this method requires skilled handling and/or a license to operate the radiation apparatus. To overcome this problem, we focused on adapting facile commercial test kits for determining the arsenic contents in an aqueous solution, a method that can be employed in the field. Because of the low solubility of gypsum, determining its arsenic levels requires a pretreatment to dissolve the gypsum into a homogeneous solution through a pyrolysis process involving harmful chemicals (such as hydrochloric acid [6] or perchloric acid [4]). After pretreatment, arsenic is released into the solution in a form suitable for detection, and conventional analytical methods can then be applied. We previously demonstrated that gypsum was easily dissolved in water containing cationand anion-exchange resins and that the fluoride content in gypsum could be successfully determined in the resulting solution using a simple colorimetric method [7]. Based on these earlier findings, we hypothesized that the arsenic in waste gypsum could also be released by a pretreatment technique using only a cation-exchange resin because arsenic forms arsenate anions.

termined in the resulting solution using a simple colorimetric method [7]. Based on these earlier findings, we hypothesized that the arsenic in waste gypsum could also be released by a pretreatment technique using only a cation-exchange resin because arsenic forms arsenate anions. Therefore, in this study, we aimed to develop a simple pretreatment method for determining the arsenic levels in the gypsum from waste plasterboards to facilitate its use as a ground stabilizer. We adopted the following approach. First, we determined the suitable quantity of cation-exchange resin required to dissolve gypsum and release the arsenic it contains into water. The volume of arsenic in the resultant solution could then be determined using a conventional arsenic determination test kit based on Gutzeit′s method. We Therefore, in this study, we aimed to develop a simple pretreatment method for determining the arsenic levels in the gypsum from waste plasterboards to facilitate its use as a ground stabilizer. We adopted the following approach. First, we determined the suitable quantity of cation-exchange resin required to dissolve gypsum and release the arsenic it contains into water. The volume of arsenic in the resultant solution could then be determined using a conventional arsenic determination test kit based on Gutzeit0 s method. We also endeavored to interpret the color change in the test paper from the arsenic test kit based on the concentration of arsenic using data from the image sensors in smart devices, such as smartphones and/or tablets. The results suggest that the proposed method can rapidly determine the amount of arsenic in gypsum. We expect this innovative technique to facilitate the monitoring of harmful pollutants in recycled gypsum powder obtained from waste plasterboards for environmental safety.

also endeavored to interpret the color change in the test paper from the arsenic test kit based on the concentration of arsenic using data from the image sensors in smart devices,

to facilitate the monitoring of harmful pollutants in recycled gypsum powder obtained

from waste plasterboards for environmental safety.

#### **2. Materials and Methods**

#### *2.1. Materials and Samples*

First, we prepared arsenic-containing gypsum samples for characterization instead of using existing ABG from waste plasterboards. Calcium sulfate dihydrate (FUJI FILM Wako Pure Chemical, Bellwood Rd, VA, USA) was used to prepare this arsenic-containing gypsum by mixing 0.3 g of reagent gypsum and 10 cm<sup>3</sup> of aqueous solutions containing various amounts of sodium arsenite. Each mixture was then ultrasonicated for 5 min and dried in a convection oven at 80 ◦C for 24 h. The water used in all experiments was prepared via ion exchange and ultrapurification using a Milli-Q water purification system (Milli-Q A10, Merck-Millipore, Burlington, MA, USA).

#### *2.2. Dissolution of Gypsum by Cation-Exchange Resin*

A total of 300 mg of the arsenic-containing gypsum samples was mixed with 20 cm<sup>3</sup> of water, and various amounts of a cation-exchange resin were added (Amberlite IR120 H, DuPont Water Solutions, Wilmington, DE, USA). The mixture was shaken at 200 strokes per minute for 5 min using a reciprocal shaker. The temperature was adjusted to 298 K. The liquid phase was separated via pressure filtration through a cartridge membrane filter (pore size: 0.45 µm). The amount of gypsum dissolved in the water was analyzed by determining the calcium and sulfur content using inductively coupled plasma atomic emission spectrometry (ICP-AES, 720ES, Agilent Technologies, Inc., Santa Clara, CA, USA) with argon plasma.

#### *2.3. Determination of Arsenic Content in the Gypsum*

In order to measure the amount of arsenic present in an aqueous solution, two types of arsenic test kits based on Gutzeit's colorimetric method for lower and higher contents (MQuant Arsenic tests, model 1.01747 and 1.17927, respectively, Merck KgaA, Darmstadt, Germany) were selected. The arsenic-containing gypsum samples were dissolved by using the method described above. After the pretreatment, the obtained water samples were tested using the arsenic test kits, which indicate the arsenic contents through changes in the color of the test paper. This color change was determined using the image sensor of a tablet device (ZenPad 8.0, ASUSTeK Computer, Taipei, Taiwan).

In order to confirm these results, the volume of arsenic released from the arseniccontaining gypsum samples was also characterized by ICP-AES, as follows: each sample was mixed with water, and the cation-exchange resin using the method above, and the arsenic content in the obtained solution was analyzed. The determination limit of arsenic by ICP-AES was approximately 0.05 mg/L.

#### **3. Results and Discussion**

#### *3.1. Suitable Volume of Cation-Exchange Resin for Gypsum Dissolution*

First, the required amount of cation-exchange resin that adequately dissolves gypsum was determined. Figure 2 shows the sulfur and calcium concentrations in water, as measured by ICP-AES after treating the pristine gypsum reagent with the cation-exchange resin. Because of the low solubility of gypsum in water, the calcium and sulfur concentrations after mixing them in water without the cation-exchange resin differed from the values obtained by dissolving all the gypsum in water using the resin (blue and red lines in the figure, respectively). Specifically, adding the cation-exchange resin increased the sulfur concentration and decreased the calcium concentration. This phenomenon indicates that shifting the chemical equilibrium (Equation (1)) to the right successfully dissolved the gypsum in water, which was attributed to a decrease in the calcium concentration as a result of using the ion-exchange resin (Equation (2)).

$$\text{CaSO}\_4 \cdot 2\text{H}\_2\text{O} \rightleftharpoons \text{Ca}^{2+} + \text{SO}\_4^{2-} + 2\text{H}\_2\text{O} \tag{1}$$

$$\text{R-H}^+ + \text{Ca}^{2+} \rightarrow [\text{R-Ca}^{2+}]^+ \tag{2}$$

Concentration (mgdm**-**3)

In Equation (2), R represents the cation-exchange resin. The experimental results demonstrated that 3.0 g or more of the cation-exchange resin in 20 cm<sup>3</sup> of water was required to dissolve 0.3 g of gypsum reagent. demonstrated that 3.0 g or more of the cation-exchange resin in 20 cm3 of water was required to dissolve 0.3 g of gypsum reagent. 15,000

In Equation (2), R represents the cation-exchange resin. The experimental results

In Equation (2), R represents the cation-exchange resin. The experimental results

demonstrated that 3.0 g or more of the cation-exchange resin in 20 cm3 of water was

R-H+ + Ca2+ → [R-Ca2+]+ (2)

R-H+ + Ca2+ → [R-Ca2+]+ (2)

*Materials* **2022**, *15*, x FOR PEER REVIEW 4 of 8

*Materials* **2022**, *15*, x FOR PEER REVIEW 4 of 8

required to dissolve 0.3 g of gypsum reagent.

**Figure 2.** Sulfur and calcium concentrations in the water sample, as measured by ICP-AES after treating the gypsum reagent with a cation-exchange resin. Closed circles (●): sulfur concentration, triangles (▲): calcium concentration. Colored lines: sulfur (red) or calcium (blue) after dissolving 100% of gypsum in water using the resin. **Figure 2.** Sulfur and calcium concentrations in the water sample, as measured by ICP-AES after treating the gypsum reagent with a cation-exchange resin. Closed circles (•): sulfur concentration, triangles (N): calcium concentration. Colored lines: sulfur (red) or calcium (blue) after dissolving 100% of gypsum in water using the resin. triangles (▲): calcium concentration. Colored lines: sulfur (red) or calcium (blue) after dissolving 100% of gypsum in water using the resin. The release of arsenic from the gypsum sample was then examined under the

The release of arsenic from the gypsum sample was then examined under the pretreatment conditions. In this study, we prepared arsenic-containing gypsum samples with predetermined amounts of arsenic instead of using existing ABG; thus, the arsenic contents were known and did not require further determination. Figure 3 shows the The release of arsenic from the gypsum sample was then examined under the pretreatment conditions. In this study, we prepared arsenic-containing gypsum samples with predetermined amounts of arsenic instead of using existing ABG; thus, the arsenic contents were known and did not require further determination. Figure 3 shows the change in the arsenic concentration in water, as determined by ICP-AES after dissolving the gypsum samples containing various amounts of arsenic. pretreatment conditions. In this study, we prepared arsenic-containing gypsum samples with predetermined amounts of arsenic instead of using existing ABG; thus, the arsenic contents were known and did not require further determination. Figure 3 shows the change in the arsenic concentration in water, as determined by ICP-AES after dissolving the gypsum samples containing various amounts of arsenic.

**Figure 3.** Arsenic concentrations in the water sample, as measured by ICP-AES, after treating the 0.0 0 50 100 150 200 Arsenic content (mgkg**-**1) **Figure 3.** Arsenic concentrations in the water sample, as measured by ICP-AES, after treating the suspensions of the arsenic-containing gypsum samples with a cation-exchange resin as a function of the known arsenic content in the prepared samples. Experimental condition: gypsum: 0.3 g, **Figure 3.** Arsenic concentrations in the water sample, as measured by ICP-AES, after treating the suspensions of the arsenic-containing gypsum samples with a cation-exchange resin as a function of the known arsenic content in the prepared samples. Experimental condition: gypsum: 0.3 g, cation-exchange resin: 4 g in 20 mL of water.

suspensions of the arsenic-containing gypsum samples with a cation-exchange resin as a function of the known arsenic content in the prepared samples. Experimental condition: gypsum: 0.3 g,

Figure 3 shows a strong linear relationship between the arsenic contents in the

in the gypsum is successfully released into the water after the cation-exchange resin pretreatment, thus eliminating the need for conventional treatment techniques that

gypsum sample and the arsenic concentration in the treated water. Evidently, the arsenic in the gypsum is successfully released into the water after the cation-exchange resin pretreatment, thus eliminating the need for conventional treatment techniques that

Figure 3 shows a strong linear relationship between the arsenic contents in the

cation-exchange resin: 4 g in 20 mL of water.

Figure 3 shows a strong linear relationship between the arsenic contents in the gypsum sample and the arsenic concentration in the treated water. Evidently, the arsenic in the gypsum is successfully released into the water after the cation-exchange resin pretreatment, thus eliminating the need for conventional treatment techniques that employ harmful chemicals. However, the amounts of arsenic released into the treated water were slightly lower than the values estimated from the arsenic used to prepare the gypsum samples. This result suggests that some arsenic ions were adsorbed on the ion-exchange resin. The pretreatment conditions, including the selection of ion-exchange resin, must therefore be optimized in future research. the gypsum is successfully released into the water after the cation-exchange resin pretreatment, thus eliminating the need for conventional treatment techniques that employ harmful chemicals. However, the amounts of arsenic released into the treated water were slightly lower than the values estimated from the arsenic used to prepare the gypsum samples. This result suggests that some arsenic ions were adsorbed on the ion-exchange resin. The pretreatment conditions, including the selection of ion-exchange resin, must therefore be optimized in future research.

#### *3.2. Improved Determination of Arsenic Concentration Using Conventional Tests and Image Processing 3.2. Improved Determination of Arsenic Concentration Using Conventional Tests and Image Processing*

*Materials* **2022**, *15*, x FOR PEER REVIEW 5 of 8

Next, the amount of arsenic was determined using the arsenic test kit for higher arsenic content by analyzing the arsenic content in the solution obtained after the resin pretreatment, as described in the previous section. Figure 4 shows a photograph of the test paper from the kit after detecting arsenic in different water samples. As shown in the figure, the color change is not easily recognizable without skilled observation. In order to better quantify these results, changes in the output signal from a tablet image sensor after capturing images of the test paper color were plotted as a function of the arsenic concentration, as shown in Figure 5. Evidently, the blue output value from the image sensor strongly correlates with the arsenic content in the solution over a concentration range from 0.05 to 0.15 mgdm−<sup>3</sup> . A similar relationship was obtained using the test kit for lower arsenic contents (data not shown). Next, the amount of arsenic was determined using the arsenic test kit for higher arsenic content by analyzing the arsenic content in the solution obtained after the resin pretreatment, as described in the previous section. Figure 4 shows a photograph of the test paper from the kit after detecting arsenic in different water samples. As shown in the figure, the color change is not easily recognizable without skilled observation. In order to better quantify these results, changes in the output signal from a tablet image sensor after capturing images of the test paper color were plotted as a function of the arsenic concentration, as shown in Figure 5. Evidently, the blue output value from the image sensor strongly correlates with the arsenic content in the solution over a concentration range from 0.05 to 0.15 mgdm−3. A similar relationship was obtained using the test kit for lower arsenic contents (data not shown).

**Figure 4.** Photograph of the test paper for higher arsenic contents in the presence of different levels of arsenic in the water samples. **Figure 4.** Photograph of the test paper for higher arsenic contents in the presence of different levels of arsenic in the water samples.

Based on these findings, we further investigated the arsenic levels in the simulated arsenic-containing gypsum. First, each gypsum sample was dissolved in water containing the cation-exchange resin, and the obtained solutions were tested using the arsenic test kit and the tablet for imaging, as described above. The results shown in Figure 6 indicate that the imaging with the tablet enables accurate determination of the arsenic concentration in the solution over the range of 10–100 mg kg−<sup>1</sup> (Figure 6a) and 4–80 mg kg−<sup>1</sup> (Figure 6b) using the arsenic test kits for higher and lower arsenic contents, respectively. Based on the results of the kit for lower contents, the arsenic content was thereafter determined using the arsenic test kit for higher arsenic contents, as the results obtained from the arsenic test kit for lower arsenic contents were quantitatively limited. However, these results indicated a higher sensitivity for detecting low arsenic contents in the gypsum.

60

80

100

Blue value

120

0

50

100

Blue value

150

*Materials* **2022**, *15*, x FOR PEER REVIEW 6 of 8

**Figure 5.** Change in the blue value of the image sensor as a function of the arsenic concentration in the solution. **Figure 5.** Change in the blue value of the image sensor as a function of the arsenic concentration in the solution. kit for lower arsenic contents were quantitatively limited. However, these results indicated a higher sensitivity for detecting low arsenic contents in the gypsum.

80 **Figure 6.** Arsenic levels in arsenic-containing gypsum, as determined using the proposed method, using the test kits for (**a**) high; (**b**) low arsenic contents. **Figure 6.** Arsenic levels in arsenic-containing gypsum, as determined using the proposed method, using the test kits for (**a**) high; (**b**) low arsenic contents.

#### 60 Blue value*3.3. Benefits of the Study Results in Gypsum Recycling 3.3. Benefits of the Study Results in Gypsum Recycling*

(**a**) (**b**) **Figure 6.** Arsenic levels in arsenic-containing gypsum, as determined using the proposed method, using the test kits for (**a**) high; (**b**) low arsenic contents. *3.3. Benefits of the Study Results in Gypsum Recycling*  The results of this study were used to evaluate the benefits of recycling gypsum. Certain properties of gypsum make it suitable for use as a fertilizer [8]; therefore, recycled 0 50 100 150 Arsenic (mgkg**-**1) 0 20 40 1 10 100 Arsenic (mgkg**-**1) The results of this study were used to evaluate the benefits of recycling gypsum. Certain properties of gypsum make it suitable for use as a fertilizer [8]; therefore, recycled gypsum from waste plasterboards could be used in agricultural applications if the impurity concentrations are controlled. Because arsenic is harmful to agricultural The results of this study were used to evaluate the benefits of recycling gypsum. Certain properties of gypsum make it suitable for use as a fertilizer [8]; therefore, recycled gypsum from waste plasterboards could be used in agricultural applications if the impurity concentrations are controlled. Because arsenic is harmful to agricultural activities [9], arsenic-containing gypsum should not be recycled for this purpose to avoid soil pollution. However, conventional methods for monitoring arsenic require specific analytical methods, skills, and equipment and involve lengthy processes. Consequently, it can be difficult to determine the volume of arsenic-containing gypsum in waste plasterboards at intermediate waste treatment facilities. Indeed, for the safe recycling of gypsum, determining the presence or absence of arsenic is more significant than quantifying the exact amount of arsenic present. The results described in the previous section suggest that arsenic-bearing gypsum can be easily identified using simplified pretreatment and conventional arsenic test kits. Further, this novel, simple method can be used for the on-site determination of the arsenic content in waste gypsum. Our results could, therefore, be useful in identifying ABG before accepting it for further reprocessing based on its arsenic content.

gypsum from waste plasterboards could be used in agricultural applications if the impurity concentrations are controlled. Because arsenic is harmful to agricultural Fluoride is also an important impurity in waste plasterboards. We previously reported a method for the on-site determination of the fluoride content in waste gypsum by

pretreatment with cation- and anion-exchange resins [7]. We also developed a simplified stabilization method for fluoride in gypsum by adding dicalcium phosphate dihydrate (DCPD, CaHPO4·2H2O) [10,11] using the transformation reaction of DCPD to stable fluorapatite (FAp, Ca10(PO4)6F2) [12]. rapatite (FAp, Ca10(PO4)6F2) [12]. The flowchart in Figure 7 shows the benefits of our combined findings in making gypsum safe for various recycling applications. If the arsenic content is sufficiently low, the gypsum can be saved from being landfilled, but the fluoride content must be checked.

ABG before accepting it for further reprocessing based on its arsenic content.

*Materials* **2022**, *15*, x FOR PEER REVIEW 7 of 8

arsenic-containing gypsum should not be recycled for this purpose to avoid soil pollution. However, conventional methods for monitoring arsenic require specific analytical methods, skills, and equipment and involve lengthy processes. Consequently, it can be difficult to determine the volume of arsenic-containing gypsum in waste plasterboards at intermediate waste treatment facilities. Indeed, for the safe recycling of gypsum, determining the presence or absence of arsenic is more significant than quantifying the exact amount of arsenic present. The results described in the previous section suggest that arsenic-bearing gypsum can be easily identified using simplified pretreatment and conventional arsenic test kits. Further, this novel, simple method can be used for the on-site determination of the arsenic content in waste gypsum. Our results could, therefore, be useful in identifying

Fluoride is also an important impurity in waste plasterboards. We previously reported a method for the on-site determination of the fluoride content in waste gypsum by pretreatment with cation- and anion-exchange resins [7]. We also developed a simplified stabilization method for fluoride in gypsum by adding dicalcium phosphate dihydrate (DCPD, CaHPO4·2H2O) [10,11] using the transformation reaction of DCPD to stable fluo-

The flowchart in Figure 7 shows the benefits of our combined findings in making gypsum safe for various recycling applications. If the arsenic content is sufficiently low, the gypsum can be saved from being landfilled, but the fluoride content must be checked. If the fluoride content is also sufficiently low, the gypsum can be used in agricultural applications. Alternatively, the fluoride can be stabilized using DCPD, making the gypsum suitable for ground stabilization. Thus, our approach leads to a reduction in the amount of waste gypsum disposed of in landfills, safer use of recycled gypsum in agricultural applications, and the efficient use of DCPD to stabilize fluoride in the recycled gypsum used for ground stabilization, which could prevent the release of fluoride into the surrounding soil. If the fluoride content is also sufficiently low, the gypsum can be used in agricultural applications. Alternatively, the fluoride can be stabilized using DCPD, making the gypsum suitable for ground stabilization. Thus, our approach leads to a reduction in the amount of waste gypsum disposed of in landfills, safer use of recycled gypsum in agricultural applications, and the efficient use of DCPD to stabilize fluoride in the recycled gypsum used for ground stabilization, which could prevent the release of fluoride into the surrounding soil.

**Figure 7.** Potential advantages of our findings for the safe reprocessing of recycled gypsum obtained from waste plasterboards. DCPD: dicalcium phosphate dihydrate. **Figure 7.** Potential advantages of our findings for the safe reprocessing of recycled gypsum obtained from waste plasterboards. DCPD: dicalcium phosphate dihydrate.

#### **4. Conclusions**

**4. Conclusions**  The results of this study suggest that the arsenic content in gypsum recycled from waste plasterboards could be determined via a pretreatment method employing an ionexchange resin, which facilitates the release of arsenic into the solution. Then, the results The results of this study suggest that the arsenic content in gypsum recycled from waste plasterboards could be determined via a pretreatment method employing an ionexchange resin, which facilitates the release of arsenic into the solution. Then, the results of conventional colorimetric arsenic test kits can be monitored using a tablet to better quantify the arsenic concentration.

of conventional colorimetric arsenic test kits can be monitored using a tablet to better The key points can be summarized as follows:


The safety of recycled gypsum powder from the waste plasterboard is essential for various applications, particularly when the gypsum is used in soil. The results of this research are thus expected to be readily applied to the construction of a safe waste plasterboard recycling system that adheres to the concept of a circular economy.

**Author Contributions:** Conceptualization, M.T. (Masamoto Tafu), A.M. and J.N.; Methodology, M.T. (Masamoto Tafu), J.N. and M.T. (Momoka Tanii); Software for the tablet, A.M.; Writing—original draft preparation, M.T. (Masamoto Tafu) and J.N.; Writing—review and editing, M.T. (Masamoto Tafu), A.M. and S.T.; Project administration, S.T. and M.T. (Masamoto Tafu); Funding acquisition, M.T. (Masamoto Tafu). All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Gypsum Board Association of Japan.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data that support the findings of this study are available from the corresponding author, M.T. (Masamoto Tafu), upon reasonable request.

**Acknowledgments:** We are grateful to Haruka Tsunekawa for providing technical support with the arsenic monitoring and Nobuhito Motomura and Takayuki Motomura from the Recycle Factory Co. Ltd., Hokkaido, Japan for various useful suggestions on recycling waste plasterboards for agricultural use.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Akihiro Takahashi 1,\*, Naoyuki Yamamoto <sup>1</sup> , Yu Ooka <sup>2</sup> and Toshinobu Toyohiro <sup>1</sup>**


**Abstract:** With the crisis awareness of global warming and natural disasters, utilization of local wood has drawn increasing attention in achieving the Sustainable Development Goals (SDGs). It is necessary to investigate the deformation and fracture of the structural tissue in wood in order to improve the safety and reliability of wood application. However, deformation and fracture mechanisms of the structural tissue in each annual ring are unknown. The mechanical characteristics of wood are reflected in the properties of earlywood and latewood. In the present study, microstructural observation and tensile tests were conducted to examine the relationship between the mechanical properties and fracture behavior of latewood in the growth direction in Japanese cedar. Brittle fracture behavior of the latewood specimen was confirmed based on the tensile stress–strain curve and features of the fracture surface. Moreover, two fracture modes, tensile fracture and shear fracture, were recognized. Weibull analysis of tensile strength in each fracture mode was performed to evaluate the reliability and utility of brittle latewood. Lastly, two fracture mechanisms were discussed based on the failure observation findings by a scanning electron microscope.

**Keywords:** Japanese cedar; latewood; mechanical property; fracture surface observation

#### **1. Introduction**

The increasing demand for materials made from renewable sources with a small environmental load has increased recently. This is due to several driving forces such as shortages of natural resources, changes in consumers and their concerns over social environmental issues, and the SDGs [1]. As a naturally grown material with carbon sequestration properties, wood has significant appeal as a sustainable material. The use of wood in industry and construction can reduce carbon in nature [2]. Therefore, wood is an environmentally friendly material that has been used for the construction of houses [3], marine environments [4], bridges [5–7], and wooden goods for many centuries. Regardless of species, engineered wood is a valuable construction material because of their highly desirable strength/density index. Significant progress in technology has been made for the last several decades to push the limit of wood construction with an advantage of higher strength/density index than other materials. As a result, there has been a noteworthy shift in public perception in terms of the acceptance of wood as a material for high-rise buildings by engineered wood such as glued laminated timber, GLT laminated veneer lumber, LVL, and cross-laminated timber, CLT [8]. There is already a growing list of highrise wooden buildings that have been constructed in different countries [9], and the trend is expected to continue. In general, buildings up to 10 stories tend to use the CLT as the primary structure [10–15]. The longitudinal elastic modulus and tensile strength of the GLT beam and CLT wall panel are one of the important characteristic values that determine the suitability of the high-rise buildings.

**Citation:** Takahashi, A.; Yamamoto, N.; Ooka, Y.; Toyohiro, T. Tensile Examination and Strength Evaluation of Latewood in Japanese Cedar. *Materials* **2022**, *15*, 2347. https://doi.org/10.3390/ ma15072347

Academic Editor: Hideyuki Kanematsu

Received: 25 January 2022 Accepted: 3 March 2022 Published: 22 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Japanese cedar (*Cryptomeria japonica*) called Sugi, a kind of conifer, is the most produced wood in Japan, accounting for 57% in 2017 [15] and is expected as domestic lamina in all layers for the GLT and CLT frames [16] in Japan. The main supply of raw Sugi tree is shifting from medium-diameter logs (with a diameter between 140 and 220 mm) to largediameter logs due to Japanese government's policies with respect to forestry management, and the production of large logs (over 300 mm in diameter) is increasing significantly within the timber manufacturing industry [17]. In an earthquake-prone country such as Japan, anti-seismic buildings with large-sized CLT are desired. For that reason, facilitating the collection of wide laminae from outside in a large-diameter log for large-sized laminated timbers is efficient, as shown in Figure 1. Japanese cedar (*Cryptomeria japonica*) called Sugi, a kind of conifer, is the most produced wood in Japan, accounting for 57% in 2017 [15] and is expected as domestic lamina in all layers for the GLT and CLT frames [16] in Japan. The main supply of raw Sugi tree is shifting from medium-diameter logs (with a diameter between 140 and 220 mm) to large-diameter logs due to Japanese government's policies with respect to forestry management, and the production of large logs (over 300 mm in diameter) is increasing significantly within the timber manufacturing industry [17]. In an earthquake-prone country such as Japan, anti-seismic buildings with large-sized CLT are desired. For that reason, facilitating the collection of wide laminae from outside in a large-diameter log for largesized laminated timbers is efficient, as shown in Figure 1.

GLT beam and CLT wall panel are one of the important characteristic values that deter-

*Materials* **2022**, *15*, x FOR PEER REVIEW 2 of 16

mine the suitability of the high-rise buildings.

**Figure 1.** Cutout arrangement of lamina from different log diameters. **Figure 1.** Cutout arrangement of lamina from different log diameters.

Although some information on Sugi timber quality has been obtained for medium Sugi logs, little is known about the quality in the large logs, the supply of which is expected to increase imminently. Studies have been carried out to evaluate mechanical properties of engineered woods such as GLT and CLT made of medium Sugi logs [18–22] or large Sugi logs [23–27] by Japanese researchers. According to research by Ido et al. [19] using lamina taken from large Sugi logs, the estimated tensile strength of CLT, as calculated using the Young's modulus of lamina of each layer, and the tensile strength of a lamina unit were found to be in good agreement with the measured tensile strength of CLT. On the other hand, the modulus of elasticity, MOE, and the modulus of rupture, MOR, in a Sugi sample by bending tests have widely been researched related to wood properties such as density, moisture content, microfibril angle (MFA), and so on [28–34], where these characteristics in medium-diameter Sugi logs were examined. However, tensile properties in the growth direction with respect to the microtome sample in the outer region where the lamina is cut have not been investigated yet. Wood is a hierarchically structured material with levels that might be termed as the tree structure, macroscopic (annual rings), microscopic (cellular), ultrastructural (cell walls), and biochemical levels (polymers such as cellulose, hemicellulose, and lignin). Thus, to utilize the functions of large-diameter Sugi logs, knowledge of their physical and mechanical properties is necessary as a whole. As above, for the engineered wood structure design such as large-sized GLT and CLT, knowledge of wood strength and rigidity is fundamental. However, previous studies have mostly focused on wood properties related to macro cross-sectional characteristics. Considering the current state of knowledge, it is a truism to claim that the mechanical strength of a wood material depends on its microstructure. Moreover, the microscopic mechanisms underlying the mechanical performance of wood need to be explained to meet the wood demand and improve technical skills and structural design tech-Although some information on Sugi timber quality has been obtained for medium Sugi logs, little is known about the quality in the large logs, the supply of which is expected to increase imminently. Studies have been carried out to evaluate mechanical properties of engineered woods such as GLT and CLT made of medium Sugi logs [18–22] or large Sugi logs [23–27] by Japanese researchers. According to research by Ido et al. [19] using lamina taken from large Sugi logs, the estimated tensile strength of CLT, as calculated using the Young's modulus of lamina of each layer, and the tensile strength of a lamina unit were found to be in good agreement with the measured tensile strength of CLT. On the other hand, the modulus of elasticity, MOE, and the modulus of rupture, MOR, in a Sugi sample by bending tests have widely been researched related to wood properties such as density, moisture content, microfibril angle (MFA), and so on [28–34], where these characteristics in medium-diameter Sugi logs were examined. However, tensile properties in the growth direction with respect to the microtome sample in the outer region where the lamina is cut have not been investigated yet. Wood is a hierarchically structured material with levels that might be termed as the tree structure, macroscopic (annual rings), microscopic (cellular), ultrastructural (cell walls), and biochemical levels (polymers such as cellulose, hemicellulose, and lignin). Thus, to utilize the functions of large-diameter Sugi logs, knowledge of their physical and mechanical properties is necessary as a whole. As above, for the engineered wood structure design such as large-sized GLT and CLT, knowledge of wood strength and rigidity is fundamental. However, previous studies have mostly focused on wood properties related to macro cross-sectional characteristics. Considering the current state of knowledge, it is a truism to claim that the mechanical strength of a wood material depends on its microstructure. Moreover, the microscopic mechanisms underlying the mechanical performance of wood need to be explained to meet the wood demand and improve technical skills and structural design technologies. To this end, in-depth studies on large-diameter Sugi logs for CLT have just begun.

nologies. To this end, in-depth studies on large-diameter Sugi logs for CLT have just begun. An annual growth ring in Japanese cedar is shown in Figure 2. In general, latewood and earlywood are determined based on the plane cut of a tree log. The latewood region has a narrow wall thickness of 0.1 to 0.4 mm.

has a narrow wall thickness of 0.1 to 0.4 mm.

**Figure 2.** Annual growth ring in Japanese cedar used in present study. **Figure 2.** Annual growth ring in Japanese cedar used in present study.

At present, there are no material testing standards for tensile tests using miniature wood test pieces. However, a suitable tensile examination method for thin latewood in Japanese cedar was previously proposed in [35]. Moreover, it was clarified that the failure behavior of latewood follows two macroscopic fracture patterns [36]. However, their fracture mechanisms remain unclear. Fewer studies have shown the tensile properties and fracture behavior in latewood taken from the outer region in large-diameter Sugi logs and to accumulate fundamental data on wood properties. The aim of the present study is to investigate the tensile stress–strain behavior of a latewood as a simple substance specimen collected from the outer side in heartwood in a Sugi log. In addition, Weibull analysis is performed on the fracture strength of latewood, and different failure behaviors of the latewood are discussed based on experimental results in SEM observation. There is a lack of research into learning the from microtome sector to surely understand the tensile behavior in latewood. The present study will improve our knowledge and skills of holistic wood utilization and structural design techniques on manufacturing engineered woods such as At present, there are no material testing standards for tensile tests using miniature wood test pieces. However, a suitable tensile examination method for thin latewood in Japanese cedar was previously proposed in [35]. Moreover, it was clarified that the failure behavior of latewood follows two macroscopic fracture patterns [36]. However, their fracture mechanisms remain unclear. Fewer studies have shown the tensile properties and fracture behavior in latewood taken from the outer region in large-diameter Sugi logs and to accumulate fundamental data on wood properties. The aim of the present study is to investigate the tensile stress–strain behavior of a latewood as a simple substance specimen collected from the outer side in heartwood in a Sugi log. In addition, Weibull analysis is performed on the fracture strength of latewood, and different failure behaviors of the latewood are discussed based on experimental results in SEM observation. There is a lack of research into learning the from microtome sector to surely understand the tensile behavior in latewood. The present study will improve our knowledge and skills of holistic wood utilization and structural design techniques on manufacturing engineered woods such as large-sized GLT and CLT by Japanese cedar.

An annual growth ring in Japanese cedar is shown in Figure 2. In general, latewood and earlywood are determined based on the plane cut of a tree log. The latewood region

#### **2. Materials and Methods**

large-sized GLT and CLT by Japanese cedar.

#### **2. Materials and Methods**  *2.1. Experimental Material*

*2.1. Experimental Material*  Wood samples in this study were cedar from Nichinan city in southern Kyushu, Japan. The tree age was 40 years. The cut log had a body diameter of 300 mm (up to 4 m above the ground) and was sun-dried. The moisture content (*MC*) defined as the weight Wood samples in this study were cedar from Nichinan city in southern Kyushu, Japan. The tree age was 40 years. The cut log had a body diameter of 300 mm (up to 4 m above the ground) and was sun-dried. The moisture content (*MC*) defined as the weight of water in the cut log was given as a percentage of the oven-dried weight [37]:

$$\text{MC} = \frac{\text{moist weight} - \text{oven} - dried}{\text{oven} - dried} \times 100(\%), \tag{1}$$

The *MC* in the cut log was 12%. The wood density calculated from the green volume and air-dry weight was 370 kg/m3 [37]. The *MC* in the cut log was 12%. The wood density calculated from the green volume and air-dry weight was 370 kg/m<sup>3</sup> [37].

Latewood samples were collected from rings 25 to 35 starting from the pith (i.e., outside heartwood). The manufacturing process and water immersion of the latewood tensile specimens in the growth direction taken from the cut log is shown in Figure 3. The specimen for tensile testing was cut in a straight grain orientation using a cutter tool and sandpaper. After cutting, the specimen was immersed in water for 24 h to remove residual strain on the specimen after cutting. After immersion, the specimen was dried in ambient air (relative humidity of 65% and room temperature of 298 K) for 72 h. Based on [38–40], Latewood samples were collected from rings 25 to 35 starting from the pith (i.e., outside heartwood). The manufacturing process and water immersion of the latewood tensile specimens in the growth direction taken from the cut log is shown in Figure 3. The specimen for tensile testing was cut in a straight grain orientation using a cutter tool and sandpaper. After cutting, the specimen was immersed in water for 24 h to remove residual strain on the specimen after cutting. After immersion, the specimen was dried in ambient air (relative humidity of 65% and room temperature of 298 K) for 72 h. Based on [38–40], the specimen thickness (radial) was 0.2 ± 0.05 mm, width (tangential) was 3.0 ± 0.5 mm, and length (longitudinal) was 130 mm.

the specimen thickness (radial) was 0.2 ± 0.05 mm, width (tangential) was 3.0 ± 0.5 mm,

the specimen thickness (radial) was 0.2 ± 0.05 mm, width (tangential) was 3.0 ± 0.5 mm,

**Figure 3.** Manufacturing process and orientation of the latewood tensile specimen. **Figure 3.** Manufacturing process and orientation of the latewood tensile specimen. **Figure 3.** Manufacturing process and orientation of the latewood tensile specimen.

and length (longitudinal) was 130 mm.

and length (longitudinal) was 130 mm.

*Materials* **2022**, *15*, x FOR PEER REVIEW 4 of 16

Microstructures of the latewood specimen in two orientations, T-R plane (a) and L-T plane (b), are shown in Figure 4. Tracheids with square or slightly rectangular cells of differing thickness were observed in the T-R plane. The thickness of the cell wall of tracheids ranged from 1 to 5 μm. The other region (three broken-line frames) running from the top to the bottom of the microstructure was the ray tissue. The tracheids and ray cells had a specific tilt angle, *θ* to the L direction in the L-T plane (b). The range was approximately 0 to 20°, measured by a protractor qualitatively. The density and *MC* in the latewood specimen measured by [37] were 897 kg/m3 and 10%, respectively. Microstructures of the latewood specimen in two orientations, T-R plane (a) and L-T plane (b), are shown in Figure 4. Tracheids with square or slightly rectangular cells of differing thickness were observed in the T-R plane. The thickness of the cell wall of tracheids ranged from 1 to 5 µm. The other region (three broken-line frames) running from the top to the bottom of the microstructure was the ray tissue. The tracheids and ray cells had a specific tilt angle, *θ* to the L direction in the L-T plane (b). The range was approximately 0 to 20◦ , measured by a protractor qualitatively. The density and *MC* in the latewood specimen measured by [37] were 897 kg/m<sup>3</sup> and 10%, respectively. Microstructures of the latewood specimen in two orientations, T-R plane (a) and L-T plane (b), are shown in Figure 4. Tracheids with square or slightly rectangular cells of differing thickness were observed in the T-R plane. The thickness of the cell wall of tracheids ranged from 1 to 5 μm. The other region (three broken-line frames) running from the top to the bottom of the microstructure was the ray tissue. The tracheids and ray cells had a specific tilt angle, *θ* to the L direction in the L-T plane (b). The range was approximately 0 to 20°, measured by a protractor qualitatively. The density and *MC* in the latewood specimen measured by [37] were 897 kg/m3 and 10%, respectively.

**Figure 4.** Microstructures in the two orientations in the latewood specimen: T-R plane (**a**) and L-T plane (**b**). **Figure 4.** Microstructures in the two orientations in the latewood specimen: T-R plane (**a**) and L-T plane (**b**). **Figure 4.** Microstructures in the two orientations in the latewood specimen: T-R plane (**a**) and L-T plane (**b**).

#### *2.2. Tensile Examination Procedure 2.2. Tensile Examination Procedure 2.2. Tensile Examination Procedure*

Tensile testing of the latewood specimen was performed in accordance with Japanese industrial standards, JIS Z2241 [41], at an initial loading speed of 1.0 mm/min. The universal testing system (EZ-SX, Shimadzu) was used to assess mechanical properties in the Tensile testing of the latewood specimen was performed in accordance with Japanese industrial standards, JIS Z2241 [41], at an initial loading speed of 1.0 mm/min. The universal testing system (EZ-SX, Shimadzu) was used to assess mechanical properties in the Tensile testing of the latewood specimen was performed in accordance with Japanese industrial standards, JIS Z2241 [41], at an initial loading speed of 1.0 mm/min. The universal testing system (EZ-SX, Shimadzu) was used to assess mechanical properties in the present study. Tensile load, *F*, was measured by a one-side loaded cell with a capacity of 500 N, and displacement and tensile strain of the tensile specimen in the growth direction were measured using a noncontact extensometer (DVE-101/201, Shimadzu, Kyoto, Japan)

and one-side strain gauge (FLK-1-11, Tokyo Measuring Instruments Lab., Tokyo, Japan) [42]. For area measurement to calculate tensile stress, *σ*, and elastic modulus, *E*, the average value of three measured cross-sectional areas, *A*, of the specimen in the longitudinal axis before tensile examination was used. Therefore, engineering tensile stress, *σ = F/A*. *E*, was determined as a proportion of a regression line fitted to the stress–strain chart between 10 and 30 MPa. The 6 specimens attached to the strain gauge were turned out and 39 specimens without the gauge were prepared for tensile examination using a noncontact type extensometer, as shown in Figure 5. In advance, a tensile test was conducted using a thin metallic lead (*Pb*, thickness of 0.2 mm and elastic modulus of 16 GPa) attached to a strain gauge, and it was confirmed that the elastic modulus was measurable without the influence of the adhesive. A microphotograph of the specimen that failed along the frame of the attached strain gauge is shown in Figure 6. This failure was considered to be due to the hardness of the adhesive. As a result, the deformation in the elastic region was evaluated by a strain gauge and the noncontact type extensometer was used for fracture strain evaluation. tween 10 and 30 MPa. The 6 specimens attached to the strain gauge were turned out and 39 specimens without the gauge were prepared for tensile examination using a noncontact type extensometer, as shown in Figure 5. In advance, a tensile test was conducted using a thin metallic lead (*Pb*, thickness of 0.2 mm and elastic modulus of 16 GPa) attached to a strain gauge, and it was confirmed that the elastic modulus was measurable without the influence of the adhesive. A microphotograph of the specimen that failed along the frame of the attached strain gauge is shown in Figure 6. This failure was considered to be due to the hardness of the adhesive. As a result, the deformation in the elastic region was evaluated by a strain gauge and the noncontact type extensometer was used for fracture strain evaluation. An optical microscope (OM, VHX-2000, KEYENCE, Osaka, Japan) and electrical scanning microscopy (SEM) were used to observe the microstructural features. SEM (S-4800, Hitachi High-Technologies Corp., Ibaraki, Japan) was also used to examine the fractographic features. For interior views of the SEM observation, sputter coating with the Pt-Pd target was conducted. tween 10 and 30 MPa. The 6 specimens attached to the strain gauge were turned out and 39 specimens without the gauge were prepared for tensile examination using a noncontact type extensometer, as shown in Figure 5. In advance, a tensile test was conducted using a thin metallic lead (*Pb*, thickness of 0.2 mm and elastic modulus of 16 GPa) attached to a strain gauge, and it was confirmed that the elastic modulus was measurable without the influence of the adhesive. A microphotograph of the specimen that failed along the frame of the attached strain gauge is shown in Figure 6. This failure was considered to be due to the hardness of the adhesive. As a result, the deformation in the elastic region was evaluated by a strain gauge and the noncontact type extensometer was used for fracture strain evaluation. An optical microscope (OM, VHX-2000, KEYENCE, Osaka, Japan) and electrical scanning microscopy (SEM) were used to observe the microstructural features. SEM (S-4800, Hitachi High-Technologies Corp., Ibaraki, Japan) was also used to examine the fractographic features. For interior views of the SEM observation, sputter coating with the Pt-Pd target was conducted.

present study. Tensile load, *F*, was measured by a one-side loaded cell with a capacity of 500 N, and displacement and tensile strain of the tensile specimen in the growth direction were measured using a noncontact extensometer (DVE-101/201, Shimadzu, Kyoto, Japan) and one-side strain gauge (FLK-1-11, Tokyo Measuring Instruments Lab., Tokyo, Japan) [42]. For area measurement to calculate tensile stress, *σ*, and elastic modulus, *E*, the average value of three measured cross-sectional areas, *A*, of the specimen in the longitudinal axis before tensile examination was used. Therefore, engineering tensile stress, *σ = F/A*. *E*, was determined as a proportion of a regression line fitted to the stress–strain chart be-

present study. Tensile load, *F*, was measured by a one-side loaded cell with a capacity of 500 N, and displacement and tensile strain of the tensile specimen in the growth direction were measured using a noncontact extensometer (DVE-101/201, Shimadzu, Kyoto, Japan) and one-side strain gauge (FLK-1-11, Tokyo Measuring Instruments Lab., Tokyo, Japan) [42]. For area measurement to calculate tensile stress, *σ*, and elastic modulus, *E*, the average value of three measured cross-sectional areas, *A*, of the specimen in the longitudinal axis before tensile examination was used. Therefore, engineering tensile stress, *σ = F/A*. *E*, was determined as a proportion of a regression line fitted to the stress–strain chart be-

*Materials* **2022**, *15*, x FOR PEER REVIEW 5 of 16

*Materials* **2022**, *15*, x FOR PEER REVIEW 5 of 16

**Figure 6.** Failure along the edge of the strain gauge occurred during tensile testing. **Figure 6.** Failure along the edge of the strain gauge occurred during tensile testing. **Figure 6.** Failure along the edge of the strain gauge occurred during tensile testing.

*2.3. Tensile Strength Evaluation by Weibull Statistics 2.3. Tensile Strength Evaluation by Weibull Statistics*  An optical microscope (OM, VHX-2000, KEYENCE, Osaka, Japan) and electrical scanning microscopy (SEM) were used to observe the microstructural features. SEM (S-4800, Hitachi High-Technologies Corp., Ibaraki, Japan) was also used to examine the fractographic features. For interior views of the SEM observation, sputter coating with the Pt-Pd target was conducted.

#### *2.3. Tensile Strength Evaluation by Weibull Statistics* is expressed as:

*Materials* **2022**, *15*, x FOR PEER REVIEW 6 of 16

 ൌ ቀ ఙబ ቁ ቀ <sup>ఙ</sup> ఙబ ቁ

symbol *m* is the Weibull modulus and *σ0* is the scale parameter.

The Weibull analysis was employed to assess a wide range of issues, including the mechanical properties of brittle materials and life-time testing [43]. The two-parameter continuous probability density function for tensile strength variables is expressed by: ൌ 1 െ ቂെ ቀ <sup>ఙ</sup> ఙబ ቁ ቃ, (3) where *P* is the probability of failure at stress, *σB*. In this study, the tensile strength of late-

The Weibull analysis was employed to assess a wide range of issues, including the mechanical properties of brittle materials and life-time testing [43]. The two-parameter continuous probability density function for tensile strength variables is expressed by:

The mean density function is asymmetrical and will assume only positive values. The

The cumulative distribution function that gives the probability of failure *P* at stress *σ*

ିଵ ቂെ <sup>ቀ</sup> <sup>ఙ</sup>

$$P = \left(\frac{m}{\sigma\_0}\right) \left(\frac{\sigma}{\sigma\_0}\right)^{m-1} \exp\left[-\left(\frac{\sigma}{\sigma\_0}\right)^m\right],\tag{2}$$

ఙబ ቁ 

ቃ, (2)

The mean density function is asymmetrical and will assume only positive values. The symbol *m* is the Weibull modulus and *σ<sup>0</sup>* is the scale parameter. The two dominant fracture modes observed in the present study, tensile fracture (a) and shear fracture (b), are shown in Figure 7. The macroscopic crack path of tensile frac-

The cumulative distribution function that gives the probability of failure *P* at stress *σ* is expressed as: ture was perpendicular to the loading direction. On the other hand, the shear crack grew across to both ends of the specimen, causing a shear fracture, (b). In addition, the specimen

$$P = 1 - \exp\left[-\left(\frac{\sigma}{\sigma\_0}\right)^m\right] \tag{3}$$

where *P* is the probability of failure at stress, *σB*. In this study, the tensile strength of latewood was assessed by Weibull analysis. Tensile stress–strain curves, which correspond to the two fracture modes, are shown in Figure 8. The curve of the tensile fracture was an approximately linear mechanical re-

#### **3. Results and Discussion** sponse, whereas that of the shear fracture followed a gentle curve from the elastic region.

latewood.

The two dominant fracture modes observed in the present study, tensile fracture (a) and shear fracture (b), are shown in Figure 7. The macroscopic crack path of tensile fracture was perpendicular to the loading direction. On the other hand, the shear crack grew across to both ends of the specimen, causing a shear fracture, (b). In addition, the specimen with a *θ* of 0 to 11◦ exhibited a tensile fracture and another *θ* of 12 to 20◦ exhibited a shear fracture. This suggested a strong relationship between the *θ* and the fracture behavior of latewood. The applied stress was at a maximum (i.e., tensile strength, *σB*) and the tensile specimen quickly broke. This brittle fracture behavior also occurred at a low loading speed of 0.01 mm/min. Reiterer et al. reported that stress–strain curves of a latewood tensile specimen with MFA <5° and =20° in *Prica abies* [44]. Results on the fracture observation photograph for both were unknown in [44], but it is interesting that they were close to those in our experimental findings. The measurement of MFA in testing latewood was our next research task.

**Figure 7.** Classified failure modes were tensile fracture (**a**) and shear fracture (**b**). **Figure 7.** Classified failure modes were tensile fracture (**a**) and shear fracture (**b**).

Tensile stress–strain curves, which correspond to the two fracture modes, are shown in Figure 8. The curve of the tensile fracture was an approximately linear mechanical response, whereas that of the shear fracture followed a gentle curve from the elastic region. The applied stress was at a maximum (i.e., tensile strength, *σB*) and the tensile specimen quickly broke. This brittle fracture behavior also occurred at a low loading speed of 0.01 mm/min. Reiterer et al. reported that stress–strain curves of a latewood tensile specimen with MFA <5◦ and =20◦ in *Prica abies* [44]. Results on the fracture observation photograph for both were unknown in [44], but it is interesting that they were close to those in our experimental findings. The measurement of MFA in testing latewood was our next research task.

Several procedures have been suggested to determine Weibull parameters such as linear regression (the Weibull plot), weighted linear regression, and maximum likelihood [45–47]. The Weibull plot is the most common and simplest method to determine Weibull parameters. The tensile strength values are ranked from the minimum to the maximum and each value is assigned a probability of failure (*P*) based on its ranking, *i*, with *i* ranging from 1 to *n*,

where *n* is the number of measurements of (tensile fracture, *n* = 19, and shear fracture, *n* = 20). The cumulative probability of failure (*P*) is calculated using the following equation:

$$P\_i = \frac{i - 0.3}{n + 0.4} \,\text{s}\tag{4}$$

where *i* is the rank and *n* is the total number of data. In this study, *n* was the total number of data for the measured tensile strength. *Materials* **2022**, *15*, x FOR PEER REVIEW 7 of 16

**Figure 8.** Stress–strain curves obtained from tensile examination. The curves correspond to the two respective failure modes. **Figure 8.** Stress–strain curves obtained from tensile examination. The curves correspond to the two respective failure modes.

Several procedures have been suggested to determine Weibull parameters such as linear regression (the Weibull plot), weighted linear regression, and maximum likelihood [45–47]. The Weibull plot is the most common and simplest method to determine Weibull The results of Weibull analysis by plotting *lnln(1/1*−*P)* versus *lnσ<sup>B</sup>* for the tensile strength of the latewood specimen in the growth direction are shown in Figure 9. In the case, Equation (3) can be rearranged to give the following equation:

$$
ln \ln \left(\frac{1}{1 - P}\right) = m \ln \sigma - m \ln \sigma\_0 \tag{5}
$$

fracture, *n* = 20). The cumulative probability of failure (*P*) is calculated using the following equation: ൌ ି.ଷ ା.ସ, (4) where *i* is the rank and *n* is the total number of data. In this study, *n* was the total number and the experimental tensile strength data plotted in Equation (5) give an approximate straight line from whose equation the parameters *m* and *σ*<sup>0</sup> can be estimated based on the linear regression. The Weibull parameter *m* of a tensile fracture was *m* = 4.8, with a correlation coefficient of *R* = 0.98, whereas that of a shear fracture was *m* = 5.5, with an *R* = 0.95. The mean tensile strength *σmean* can be expressed as [48]:

$$
\sigma\_{\text{mean}} = \sigma\_0 \Gamma \left( 1 + \frac{1}{m} \right) . \tag{6}
$$

ቁ, (6)

case, Equation (3) can be rearranged to give the following equation: ቀ <sup>ଵ</sup> ଵି<sup>ቁ</sup> ൌ െ , (5) and the experimental tensile strength data plotted in Equation (5) give an approximate where Γ(·) is a gamma function. The mean tensile strength *σmean* calculated by Equation (6) in the shear fractured specimen was approximately 29% lower than that in the tensilefractured specimen. Weibull parameters and other mechanical properties are summarized in Table 1.

straight line from whose equation the parameters *m* and *σ0* can be estimated based on the linear regression. The Weibull parameter *m* of a tensile fracture was *m* = 4.8, with a correlation coefficient of *R* = 0.9*8*, whereas that of a shear fracture was *m* = 5.5, with an *R* = 0.95.

ൌ ቀ1 <sup>ଵ</sup>

where ሺ∙ሻ is a gamma function. The mean tensile strength *σmean* calculated by Equation (6) in the shear fractured specimen was approximately 29% lower than that in the tensilefractured specimen. Weibull parameters and other mechanical properties are summarized

in Table 1.

The mean tensile strength *σmean* can be expressed as [48]:

**Figure 9.** Weibull analysis plots corresponding to the two failure modes. **Figure 9.** Weibull analysis plots corresponding to the two failure modes.

**Table 1.** Weibull parameters, elastic modulus, *E*, and fracture strain, *εf*, determined by tensile examination in the present study. **Table 1.** Weibull parameters, elastic modulus, *E*, and fracture strain, *εf*, determined by tensile examination in the present study.


ture 20 126 5.5 0.95 116 0.014 ± 0.006 The simple composite mechanics rule that can be utilized to take into account the tracheids' orientation is the maximum stress criterion, assuming that the plane is modeled as a thin mat with fibers (analogous to the tracheids) oriented at an angle *θ* (analogous to the angle *θ* in Figure 4) to the fiber axis. Failure of the mat occurs either at a critical (local) stress value ଵ ଵ௨ parallel to the fibers, ଶ ଶ௨ perpendicular to the fibers, or at a shear stress ଵଶ ଵଶ௨ along the fibers. The (local) in-plane stresses working parallel and The simple composite mechanics rule that can be utilized to take into account the tracheids' orientation is the maximum stress criterion, assuming that the plane is modeled as a thin mat with fibers (analogous to the tracheids) oriented at an angle *θ* (analogous to the angle *θ* in Figure 4) to the fiber axis. Failure of the mat occurs either at a critical (local) stress value *σ*<sup>1</sup> ≥ *σ*1*<sup>u</sup>* parallel to the fibers, *σ*<sup>2</sup> ≥ *σ*2*<sup>u</sup>* perpendicular to the fibers, or at a shear stress *τ*<sup>12</sup> ≥ *τ*12*<sup>u</sup>* along the fibers. The (local) in-plane stresses working parallel and perpendicular to the fibers (*σ*1, *σ*2, *τ*12) can then be expressed as the (global) stresses applied in the x- and y-directions of the fiber mat *σx*, *σy*, *τxy* according to [49]:

$$\left\{ \begin{array}{c} \sigma\_{1} \\ \sigma\_{2} \\ \tau\_{12} \end{array} \right\} = [T] \left\{ \begin{array}{c} \sigma\_{x} \\ \sigma\_{y} \\ \tau\_{xy} \end{array} \right\},\tag{7}$$
  $\sigma\_{1} = \begin{array}{cccc} \dots & \dots & \dots & \dots & \dots & \dots \end{array}$ 

௫௬ൡ, (7)

൝ ଶ ଵଶ ൡ ൌ ሾሿ൝ where the transformation matrix is given by:

Shear frac-

as:

$$\begin{bmatrix} T \end{bmatrix} = \begin{bmatrix} \cos^2\theta & \sin^2\theta & 2\cos\theta\sin\theta\\ \sin^2\theta & \cos^2\theta & -2\cos\theta\sin\theta\\ -\cos\theta\sin\theta & \cos\theta\sin\theta & \cos^2\theta - \sin^2\theta \end{bmatrix} \tag{8}$$

If uniaxial tension ሺ ଶ ൌ ଵଶ ൌ 0ሻ is assumed, the stress ௫௨ (i.e., applied tensile െ ଶ െ ଶ If uniaxial tension ( *σ*<sup>2</sup> = *τ*<sup>12</sup> = 0) is assumed, the stress *σxu* (i.e., applied tensile stress) to cause failure in the material can be expressed for each of the three failure modes as:

$$
\sigma\_{xu} = \frac{\sigma\_{1u}}{\cos^2 \theta} \,\,\,\tag{9}
$$

$$
\sigma\_{\rm xu} = \frac{\sigma\_{\rm 2u}}{\sin^2 \theta} \tag{10}
$$

, (9)

, (10)

௦ఏ௦ఏ, (11)

$$
\sigma\_{xu} = \frac{\pi\_{12u}}{\cos \theta \sin \theta} \tag{11}
$$

where *σxu* in Equations (9)–(11) indicate axial stress, transverse stress, and shear stress, respectively. Under the assumption of independent modes of failure with no interaction between each other and using experimental data for *σ*1*u*, *σ*2*u*, and *τ*12*u*, Equations (9)–(11) can be used to calculate the maximum tensile strength of latewood material with tracheids oriented at a given angle, *θ*. Conversely, data of the maximum (global) tensile strength can be predicted together with information on *θ* to calculate the local properties *σ*1*u*, *σ*2*u*, and *τ*12*u*, for a material. As shown in Figure 10, the predicted critical tensile strength for tensile fracture (solid line, AB) is given by ௨௫ ൌ ଵହ ௦మఏ ; ሺ0° ൏ 12°ሻ, (12) σmean is 165 MPa when indicating tensile fracture, and the angle, *θ =* 12°, is a specific tilt angle when transiting from tensile fracture to shear fracture. The two dashed lines of Equations (9) and (10) cross at point B. The stress value at point B is obtained at 173 MPa from Equation (12). Next, the shear fracture value, ଵଶ௨ at point B can be calculated by substituting ௨௫ ൌ 173 MPa in Equation (11), and a certain value, ଵଶ௨ ൌ 35 MPa, is determined. Therefore, the predicted critical tensile strength for shear fracture (solid line,

where ௫௨ in Equations (9)–(11) indicate axial stress, transverse stress, and shear stress, respectively. Under the assumption of independent modes of failure with no interaction between each other and using experimental data for ଵ௨, ଶ௨, and ଵଶ௨, Equations (9)– (11) can be used to calculate the maximum tensile strength of latewood material with tra-

௫௨ ൌ ఛభమೠ

௫௨ ൌ ఙభೠ ௦మఏ

௫௨ ൌ ఙమೠ ௦మఏ

$$
\sigma\_{\text{ux}} = \frac{165}{\cos^2 \theta}; \ (0^\circ \le \theta < 12^\circ)\_\prime \tag{12}
$$

*Materials* **2022**, *15*, x FOR PEER REVIEW 9 of 16

**Figure 10.** Predicted critical tensile strength depending on angle, *θ*, of the applied stresses for the onset of two failure modes. **Figure 10.** Predicted critical tensile strength depending on angle, *θ*, of the applied stresses for the onset of two failure modes.

A photograph of the nearby crack path in the tensile fracture observed from the L-T plane is shown in Figure 11. Microscopic crack deflection was frequent during fracture propagation of the tensile fracture. The fracture surface observed from above is shown in Figure 12. Fracture modes of tracheids were observed, and the enlarged view of frame A shows fibrils in the tip of the fractured tracheid (Fs indicated by an arrow), the separation of cells at the middle lamella (IC indicated by an arrow), a crack path cut through the tracheids (TW indicated by an arrow), and an hierarchical crack propagation within the cell wall (IW indicated by two arrows). According to Côté and Hanna [50], three kinds of cell fractures in many species are recognized: intercell failure (IC), transwall failure (TW), σmean is 165 MPa when indicating tensile fracture, and the angle, *θ =* 12◦ , is a specific tilt angle when transiting from tensile fracture to shear fracture. The two dashed lines of Equations (9) and (10) cross at point B. The stress value at point B is obtained at 173 MPa from Equation (12). Next, the shear fracture value, *τ*12*<sup>u</sup>* at point B can be calculated by substituting *σux* = 173 MPa in Equation (11), and a certain value, *τ*12*<sup>u</sup>* = 35 MPa, is determined. Therefore, the predicted critical tensile strength for shear fracture (solid line, BC) for angles from 12 to 20◦ can be drawn, as shown in Figure 10. Further, it can be seen that Figure 10 qualitatively corresponds to tensile strength depending on the angle of the tracheid and therewith the prediction of the fracture behavior.

A photograph of the nearby crack path in the tensile fracture observed from the L-T plane is shown in Figure 11. Microscopic crack deflection was frequent during fracture propagation of the tensile fracture. The fracture surface observed from above is shown in Figure 12. Fracture modes of tracheids were observed, and the enlarged view of frame A shows fibrils in the tip of the fractured tracheid (Fs indicated by an arrow), the separation of cells at the middle lamella (IC indicated by an arrow), a crack path cut through the tracheids (TW indicated by an arrow), and an hierarchical crack propagation within the cell wall (IW indicated by two arrows). According to Côté and Hanna [50], three kinds of cell fractures in many species are recognized: intercell failure (IC), transwall failure (TW), and intrawall failure (IW). Intercell failure takes place at the middle tracheid lamella and is simply the interfacial debonding between tracheids at these junctions. Transwall failure is the complete rupture when the fracture path cuts across the wall. Intrawall failure occurs within the secondary wall and, in most instances, it is at the S1/S<sup>2</sup> interface or close to it. These fracture characteristics were confirmed in the fracture surface in the tensile fracture. These fracture modes tended to produce a highly rough fracture surface, as shown in Figure 11. Figure 12 also shows fracture of the ray cell observed on the tensile fracture surface. It is well known that the structure and distribution of the ray cell have a strong relationship with the compressive mechanical property and its fracture behavior in Sugi timber [51]. In present study, the influence of the ray cell on the tensile fracture behavior is unknown. However, it is suggested that the tensile fracture is partly related because the fracture of the ray cell was included in a part of the crack path. is simply the interfacial debonding between tracheids at these junctions. Transwall failure is the complete rupture when the fracture path cuts across the wall. Intrawall failure occurs within the secondary wall and, in most instances, it is at the S1/S2 interface or close to it. These fracture characteristics were confirmed in the fracture surface in the tensile fracture. These fracture modes tended to produce a highly rough fracture surface, as shown in Figure 11. Figure 12 also shows fracture of the ray cell observed on the tensile fracture surface. It is well known that the structure and distribution of the ray cell have a strong relationship with the compressive mechanical property and its fracture behavior in Sugi timber [51]. In present study, the influence of the ray cell on the tensile fracture behavior is unknown. However, it is suggested that the tensile fracture is partly related because the fracture of the ray cell was included in a part of the crack path. is the complete rupture when the fracture path cuts across the wall. Intrawall failure occurs within the secondary wall and, in most instances, it is at the S1/S2 interface or close to it. These fracture characteristics were confirmed in the fracture surface in the tensile fracture. These fracture modes tended to produce a highly rough fracture surface, as shown in Figure 11. Figure 12 also shows fracture of the ray cell observed on the tensile fracture surface. It is well known that the structure and distribution of the ray cell have a strong relationship with the compressive mechanical property and its fracture behavior in Sugi timber [51]. In present study, the influence of the ray cell on the tensile fracture behavior is unknown. However, it is suggested that the tensile fracture is partly related because the fracture of the ray cell was included in a part of the crack path.

and intrawall failure (IW). Intercell failure takes place at the middle tracheid lamella and

and intrawall failure (IW). Intercell failure takes place at the middle tracheid lamella and is simply the interfacial debonding between tracheids at these junctions. Transwall failure

*Materials* **2022**, *15*, x FOR PEER REVIEW 10 of 16

*Materials* **2022**, *15*, x FOR PEER REVIEW 10 of 16

**Figure 11.** Serrated crack path in the tensile fracture observed in the L-T plane. **Figure 11.** Serrated crack path in the tensile fracture observed in the L-T plane. **Figure 11.** Serrated crack path in the tensile fracture observed in the L-T plane.

**Figure 12.** SEM image of the T-R fracture surface indicating the behavior of the tensile fracture. The enlarged view (**b**) is section A surrounded by the dashed lines in (**a**). **Figure 12.** SEM image of the T-R fracture surface indicating the behavior of the tensile fracture. The enlarged view (**b**) is section A surrounded by the dashed lines in (**a**). **Figure 12.** SEM image of the T-R fracture surface indicating the behavior of the tensile fracture. The enlarged view (**b**) is section A surrounded by the dashed lines in (**a**).

Ifju et al. [52] and Kifetew et al. [53] also observed fractures in the latewood region where the fibrils appeared at the tips of tracheids, as shown in Figure 12b. Approximately 97% of the Japanese cedar's cells are tracheids. Tracheids in the latewood have thick-walled cells, which consist of the S<sup>2</sup> layer of 86% in the cedar [54]. The microfibrils of the S<sup>2</sup> layer are almost fully aligned along the direction of the tracheid [53] and, therefore, its internal structure plays a key role in determining the mechanical properties, especially under applied load parallel to the grain. The microfibril angle (MFA) is the angle between helical windings of microfibrils in the S<sup>2</sup> layer of the tracheid and the longitudinal cell axis; on research in Sugi, MFA was found to be a crucial factor in obtaining mechanical properties such as stiffness [31–33,54–57] and bending load–deflection behavior [34,58]. A large MFA shows low stiffness, on the other hand, and a small MFA in wood shows high stiffness. In general, each cell was considerably stiffer and stronger parallel to its axis than perpendicular. Therefore, the elastic modulus and tensile strength for the specimen with the tensile fracture were higher than those of the specimen with the shear fracture.

A photograph of a single shear crack in the shear fracture observed from the L-T plane is shown in Figure 13. Macroscopic shear fractures occurred due to brittleness and at an angle of 12 to 20◦ in the tensile direction. This corresponds with the growth direction of tracheids being tilted at *θ* of 12 to 20◦ in the L direction. A microphotograph taken perpendicular to the shear fracture surface is shown in Figure 14. The shear crack propagated through the tracheid interface and intercellular layer into the ray tissue. We confirmed that the crack propagated in a stepwise manner, as shown in the upper part of Figure 14. In addition, the fracture of tracheids was mainly due to interfacial debonding along the lamellar structure of the tracheid. It is known that the dry wood cell interface between tracheids is filled with deposits such as lignin, gum, resin, and tylose [59]. The shear strength was significantly lower than the fibril strength in the tracheid. For this reason, interfacial debonding at the tracheid interface readily occurred. As shown in Figure 4b, the ray tissue configuration in the L-T plane had a high aspect ratio and its tips were sharp. Although the ray tissue and cell structures were not clarified, the shear resistance (i.e., shear modulus and shear strength) of the intercellular layer of ray tissue was weak based on fracture surface observation in this study. Miyoshi et al. measured the breaking length of ray tissue after the lateral tensile test [60] and reported that the mechanical properties of wood in the lateral direction are significantly affected by the structural features such as deformation of cell shapes and arrangement of ray tissue or tracheids [61]. Figure 15 shows another microphotograph of the shear fracture surface. Intercell failure predominated in the entire fracture surface in the tracheid region. Several twisting and tearing fibrils (indicated by arrows) were observed in the region of failed tracheids above and below the microphotograph in Figure 15a. Those of the fibrils were spread out in response to the shear direction, while an open plane without fibrils was viewed in the region of the ray cell (see Figure 15b). The strength characteristics in the region of the ray cell were lower than those of tracheids with fragments spiraling out. The observational finding is obviously evidence to decide the existence of different fracture strength levels on the interface tracheids and ray cell. The crack origin site of the shear fracture in the present study is unknown, but considering the earlier occurrence of shear cracks in latewood, ray cells may be involved. This result suggests that nonlinearity in the stress–strain curve due to shear loading as illustrated in Figure 8 was caused by accumulation fracture at ray cells occurring during testing.

Based on this study, the mechanical properties of Sugi latewood were closely related to the tilt of tracheid grains. Many previous studies demonstrated the influence of the slope of wood grain on the MOE and MOR as follows: Hankinson's formula is well known as a prediction equation for the strength as a function of the grain angle [62]. Xavier et al. [63] and Bilko et al. [64] reported that grain deviation in the testing force axis causes a degradation in shear strength using the shear arcan test, in which the angle of the grain ranges from 0 to 90◦ . Gupta et al. revealed the effects of the grain angle on the shear strength by the shear block test [65]. Mania et al. demonstrated that the grain deviation angle has the greatest influence on mechanical parameters, such as elastic energy and work until maximum load, using the bending test with different wood species [66]. Although these studies were conducted using samples containing earlywood and latewood, their

conclusions are consistent with the present study in that the tilt of tracheids, *θ*, is important for mechanical properties. *Materials* **2022**, *15*, x FOR PEER REVIEW 12 of 16 *Materials* **2022**, *15*, x FOR PEER REVIEW 12 of 16

*Materials* **2022**, *15*, x FOR PEER REVIEW 12 of 16

**Figure 13.** A single shear crack path in the shear fracture observed in the L-T plane. **Figure 13.** A single shear crack path in the shear fracture observed in the L-T plane. **Figure 13.** A single shear crack path in the shear fracture observed in the L-T plane. **Figure 13.** A single shear crack path in the shear fracture observed in the L-T plane.

cellular layer into the ray tissue.

dashed lines in (**a**).

**Figure 14.** Microfractography of the shear fracture behavior along the tracheid interface and inter-**Figure 14.** Microfractography of the shear fracture behavior along the tracheid interface and intercellular layer into the ray tissue. **Figure 14.** Microfractography of the shear fracture behavior along the tracheid interface and intercellular layer into the ray tissue. **Figure 14.** Microfractography of the shear fracture behavior along the tracheid interface and intercellular layer into the ray tissue.

**Figure 15.** SEM image showing the behavior of the shear fracture. Arrows indicate tearing fibril fragments in the intercell-failed tracheids. The enlarged view (**b**) is section B surrounded by the **Figure 15.** SEM image showing the behavior of the shear fracture. Arrows indicate tearing fibril fragments in the intercell-failed tracheids. The enlarged view (**b**) is section B surrounded by the dashed lines in (**a**). **Figure 15.** SEM image showing the behavior of the shear fracture. Arrows indicate tearing fibril fragments in the intercell-failed tracheids. The enlarged view (**b**) is section B surrounded by the dashed lines in (**a**). **Figure 15.** SEM image showing the behavior of the shear fracture. Arrows indicate tearing fibril fragments in the intercell-failed tracheids. The enlarged view (**b**) is section B surrounded by the dashed lines in (**a**).

In the present study, although the fracture mechanisms were not clarified in detail, our results suggest that the low mechanical property in ray cells have a distinct effect on fracture morphologies in Sugi latewood. These results might also influence the other properties such as cutting and processing of timber and durability in wood products in large-diameter Sugi logs because GLT and CLT beams are subject to shearing. On the other hand, large-diameter Sugi trees are equivalent to aged Sugi. For effective utilization of Sugi wood resources in the future, it is important to understand the properties of wood derived from aged-Sugi trees.

#### **4. Conclusions**

In this paper, the tensile stress–strain behavior of latewood as a simple substance specimen collected from the outer side in heartwood in a large-diameter Sugi log was investigated. Based on this study, including tensile examination, Weibull statistics analysis, and fracture surface observation by SEM, we made the following conclusions:


**Author Contributions:** Conceptualization, A.T. and Y.O.; methodology, A.T. and N.Y.; formal analysis, A.T. and Y.O.; performing and experiments, A.T., Y.O. and N.Y.; data curation, A.T. and T.T.; writing—original draft preparation, A.T.; writing—review and editing, A.T. and T.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** The present work was partially supported by the ICHIJU Industrial Science and Technology Promotion Foundation, Japan of FY2021.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data will be made available upon reasonable request.

**Acknowledgments:** The authors would like to thank S. Yuki and H. Yamamoto for technical assistance with the experiments.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

