Fiber morphology, chemical composition, and properties of kraft pulping handsheet made from four Thailand bamboo species

ABSTRACT The microstructural, morphological, chemical, pulp, and paper properties of four fast growing Thai bamboo species, Bambusa vulgaris, Bambusa longispiculata, Dendrocalamus membranaceus, and ×Thyrsocalamus liang, were characterized to evaluate their potential as alternative fibrous raw materials in pulping and papermaking. It was found that the chemical composition of the four bamboo species was not significantly different. The alpha cellulose, lignin, and ash contents were between 46% and 50%, 27% and 29%, and 2% and 3%, respectively. The average fiber length ranged between 1.74 and 2.16 mm and had long fibers similar to softwood fibers. The bamboo fibers were thick-walled, narrow, had a small lumen, and were rigid. The pulp yield of all the bamboo species, prepared by kraft pulping, was high between 45% and 50% and had a low rejection of 0.12–0.33%. The handsheets were prepared from the bamboo pulp using various beating time to develop strength properties. It was observed that the handsheet quality improved with increasing beating time. We conclude that these properties indicate a great potential for these bamboo species for pulp and paper production.


Introduction
Lignocellulosic biomass includes both wood and non-wood biomass. Wood-based lignocellulosic biomass is derived from both hardwood and softwood species and is the main raw material in pulp production. Non-wood lignocellulosic biomass, obtained from rice straw, corn stalks, canola stalks, sugarcane bagasse, and bamboo, is an important fiber resource for pulp production in countries with a shortage of wood-based raw material. Although woody species are a major source of fiber supply in papermaking, the lack of forest resources in many areas of the world makes non-wood fiber an important alternative fibrous source for pulp and papermaking industries (Enayati et al. 2009;Gülsoy and Şimşir 2018).
Bamboo belongs to the family of grasses and is considered as an alternative source of non-wood raw materials due to its easy propagation, wide distribution, availability, fast-growing nature, shortrotation, and high productivity. Over the decades, many researchers have studied the potential of different bamboo species as pulp material resources (Azeez, Andrew, and Sithole 2016;Chitbanyong et al. 2018;Kamthai 2007;Nitisoravut and Malinen 2007;Pitiphatharaworachot et al. 2019). Bamboo is found abundantly in Thailand (Leksungnoen 2017) with its forests covering approximately 261,000 ha or 0.51% of the total land area (1.60% of total forest area).
The properties of bamboo vary with species, age, location, and external factors (Grosser and Liese 1971;Wang et al. 2016). In general, bamboo can be considered as having a longfibered or semilong-fibered fibrous material as the length of fibers in many bamboo species is comparable to that of softwood fibers, with average values being in the range of 1.5 to 4.4 mm. Knowledge about the morphological and chemical properties of bamboo is necessary for assessing its suitability for various end products and then extensive utilization of this resource.
The most popular pulping process in the world is the kraft pulping process because of the excellent strength of the obtained pulp and easy recovery of chemicals (Kamthai and Puthson 2005). The kraft pulping provides satisfactory delignification as well as a high yield and viscosity. The kraft process is generally preferred for bamboo, given that the fiber dimensions and primary chemical constituents of bamboo typically bear a close resemblance to woody species.
Information about bamboo as a pulping raw material is still limited in Thailand. Thus, the main objective of this study was to investigate the pulp and papermaking properties of four bamboo species. For this purpose, the chemical composition, fiber morphology, and kraft pulp properties of four bamboo species, namely Bambusa vulgaris, Bambusa longispiculata, Dendrocalamus membranaceus, and ×Thyrsocalamus liang, were determined. Beating process was used in conjunction with kraft pulping to improve the strength of the bamboo handsheets.

Preparation of raw material
Three-year-old bamboo culms from the four species of bamboo (B. vulgaris (a), B. longispiculata (b), D. membranaceus (c), and ×T. liang (d)) were collected from a bamboo plantation located in the Forest Species Central Park, Wang Nam Khiao district, Nakhon Ratchasima, Thailand ( Figure 1). The culms of each species were cut at a height of 30 cm above the ground, labeled, and weighed on site. For pulping, the culms were manually chopped to the approximate size of industrial chips with dimensions of 30 × 20 × 2 mm. The moisture content of the chips was determined according to the TAPPI T258 om-11 standard. The chemicals and solvents such as sodium hydroxide (NaOH), sodium sulfide (Na 2 S⋅9 H 2 O), hydrochloric acid (HCl), glacial acetic (CH 3 COOH), hydrogen peroxide (H 2 O 2 ), and other such chemicals were of laboratory grade. All chemicals were purchased from Merck Co., Ltd. (Bangkok, Thailand) and were used without further purification.

Microstructure of the bamboo culm
Bamboo culm blocks of length 2 cm were cut from the culm at the diameter at breast height (DBH). Using stereo microscopy (Stemi 508, Zeiss, Germany) and a digital camera (Axiocam 208color, Zeiss, Germany), the transverse sections of the blocks were observed to determine the distribution of vascular bundles. Transverse sections of the culm of thickness 30-40 µm were obtained from the blocks using a sliding microtome (SM2000 R, Leica, Germany). The obtained samples were stained with safranin-o, dehydrated in alcohol, and mounted on a glass. The microstructures were then observed under a light microscope (Axioscope 7, Zeiss, Germany) and photographed with a digital camera (Axiocam506, Zeiss, Germany).

Basic density
The basic density was determined in accordance with the TAPPI T258 om-16 standard. The bamboo chips were soaked in distilled water for 24 h to ensure full cell saturation. The green volume of each sample was then measured through the water displacement method after which the chips were oven dried at a temperature of 103 ± 2°C to determine the oven dry weight. The basic density was calculated using the formula: basic density = oven dry weight of sample/green volume of sample.

Fiber morphology
Macerated fibers were prepared from the bamboo culms by cutting at the DBH. The bamboo culms were split longitudinally with the grain to small stick (25 × 2 × 2 mm). The pieces of bamboo sticks were macerated with a mixture of glacial acetic acid and 30% hydrogen peroxide (1:1 ratio) at 75°C for a duration of 48 h in a water bath until the color of the stick became white, according to the Franklin method (Franklin 1945). The macerated samples were rinsed with distilled water, then disintegrated, and stained with 1% safranin-o solution. The lumen width and cell wall thickness of the fibers were examined under a light microscope and photographed with a digital camera. The fiber length and width were examined with a fiber quality analyzer (FQA-360, OpTest, Canada). The slenderness ratio (fiber length/fiber width), Runkel ratio ([2 × cell wall thickness]/lumen width), and flexibility (lumen width/fiber width) were then calculated.

Determination of chemical composition
The bamboo culms cut at the DBH were further split into small pieces and ground into powder using a laboratory mill (Thomas-WILEY Model 4; Arthur H. Thomas Company, Philadelphia, PA, USA). The obtained air dried bamboo powder was screened through a 40-mesh screen and retained on a 60mesh screen. The collected bamboo powder was then subjected to a chemical composition analysis. The holocellulose content was analyzed according to the extractive-free wood method described by Wise, Murphy, and D'Addieco (1946). The other chemical compositions determined were alpha cellulose (TAPPI T203 cm-09), acid-insoluble lignin (TAPPI T222 om-11), ash (TAPPI T211 om-12), alcohol-benzene extractives (TAPPI T204 cm-07), water solubility (TAPPI T207 cm-08), and 1% NaOH solubility (TAPPI T212 om-07).

Kraft pulping
The kraft pulping was operated using a laboratory rotating batch reactor (7-L Digester, SEW-Eurodrive, Bruchsal, Germany), to obtain the target kappa number of 20-25 of pulp. The kraft pulping condition is shown in Table 1 (Jansiri et al. 2021;Nitisoravut and Malinen 2007). After cooking, the brown stock obtained was washed, disintegrated, and screened with a screen plate having a 0.15 mm wide slot. The pulp yield and reject content were calculated in percentage. Kappa number was determined based on TAPPI standard T236 om-06. The screened bamboo pulp was collected and kept in a refrigerator to prepare the handsheets.

Physical and mechanical properties of the handsheet
The bamboo pulp was beaten in a valley beater (Laurentzen and Wettress, Stockholm, Sweden) using deionized water for four different durations of 0, 15, 30, and 45 min, respectively. The pulp drainability was measured using the Canadian standard method (TAPPI T227 om-09). Standard handsheets weighing 60 g/m 2 were made using both unbeaten and beaten bamboo pulp for each of the beating time durations according to the TAPPI T205 sp-06 standard. All handsheets were conditioned at a relative humidity of 50 ± 2% and a temperature of 23 ± 1°C for a week before testing. The handsheets were measured for their physical properties such as apparent density (according to the TAPPI T220 sp-10 standard) and thickness (according to the TAPPI T411 om-08 standard) using a precision micrometer (Laurentzen and Wettress, Sweden). The mechanical properties, i.e., tensile strength (TAPPI T494 om-06), tearing strength (TAPPI T414 om-04), bursting strength (TAPPI T403 om-10), and folding endurance (TAPPI T511 om-08), were determined by using a tensile tester (EJA-series; Thwing-Albert Instrument Co. Ltd., West Berlin, USA), tearing resistance tester (Thwing-Albert Instrument Co. Ltd., West Berlin, USA), bursting strength tester (Laurentzen and Wettress, Stockholm, Sweden), and folding endurance tester (Kumagai Riki Kogyo Co. Ltd., Tokyo, Japan), respectively, with 10 replicates for each condition.

Surface morphological properties of handsheet
The surface morphology of the handsheets derived from bamboo pulp was observed under a scanning electron microscope (SEM). The unbeaten and beaten pulp handsheets were observed under a fieldemission scanning electron microscope (FE-SEM) (SU8020; Hitachi, Tokyo, Japan) at 10 kV and 15.0 mm observation distance, after coating through Pt-sputtering using the Quorum Q150RES apparatus (Quorum Technologies, Lewes, UK) for 120 s.

Microscopic characteristics, basic density, and fiber morphology of bamboo
The structure of a bamboo culm cross section is illustrated in Figure 2. The culm of bamboo was hollow and of a circular shape, resulting in a cylindrical tissue. This comprised scattered vascular bundles embedded in parenchymatous ground tissue across the culm-wall. The density of the vascular bundle increased gradually from the inner to the outer layer, as shown in Figure 2a. The vascular bundles of bamboo consisted of phloem, two metaxylem vessels, fiber strands, parenchyma cells, and sclerenchyma sheaths, as shown in Figure 2b (Chaowana 2013;Grosser and Liese 1971;Liese and Grosser 2000;Nitisoravut and Malinen 2007). The fibers were distributed around the protoxylem, phloem, and metaxylem vessels, and there were many fibers found inside the fiber strands of the vascular bundles (Figure 2b-2c). The vascular bundles in the middle of the culm-wall were relatively larger than the ones located on the outside. The vascular bundle types of the four bamboo species (B. vulgaris (a), B. longispiculata (b), D. membranaceus (c), and ×T. liang (d)) were of the broken-waist type, which is composed of two parts (central vascular strand and one fiber strand) (Grosser and Liese 1971), as shown in Figure 3. The basic density of bamboo culm was between 671 and 712 kg/m 3 as shown in Table 2 and is denser than softwood (380 to 490 kg/m 3 ) and hardwood (520 to 560 kg/m 3 ) (Petráš et al. 2020;Saranpaa 2003). This is probably because bamboo culms had a high number of vascular bundles and the fibers had thick walls with small lumen, resulting in a higher basic density (Chaowana 2013;Nitisoravut and Malinen 2007). Theoretically, denser bamboo chips are more difficult to penetrate with cooking liquor, which is the reason for a higher required load of the digester for bamboo cooking. Therefore, cooking time must be extended for chemical liquor penetration into bamboo chips (Nitisoravut and Malinen 2007).
Fiber morphological properties are important for evaluating the suitability of pulp and paper production of a raw material. The microphotographs of bamboo fiber are shown in Figure 4 and fiber properties of bamboos are listed in Table 2. The fiber length was observed to vary between the four bamboo species. The length of bamboo fibers ranged between 1.74 and 2.16 mm, which is between that of softwood (2.7 to 4.6 mm) and hardwood (0.7 to 3.0 mm) fiber length (Mousavi et al. 2013). The average fiber length of bamboo species estimated in this study is comparable to bamboo species Bambusa vulgaris (2.10 mm), Thyrsostachys siamensis    (Azeez, Andrew, and Sithole 2016;Cao et al. 2014;Deniz et al. 2017;Nitisoravut and Malinen 2007;Wang et al. 2016). However, fiber lengths of the four bamboo species were slightly shorter than D. asper (2.89 mm) (Jansiri et al. 2021). The fiber width, lumen width, and wall thickness of the four bamboo species ranged between 22 and 24 µm, 4.8 and 6.5 µm, and 15 and 17 µm, respectively. Hence, it can be concluded that the bamboo fibers had a thick fiber wall, with a narrow and small lumen, and that such characteristics of bamboo fiber have been reported previously (Jansiri et al. 2021;Kamthai 2007;Nitisoravut and Malinen 2007;Wang et al. 2016;Zhan et al. 2015). The slenderness ratio, Runkel ratio, and flexibility ratio, derived from fiber dimensions, have been used to determine the suitability of the fibrous raw material in pulp and paper production. All bamboo fibers had a high slenderness ratio ranging between 76 and 91. These values are close to the slenderness ratio of softwood pulp fiber (89) but higher than that of hardwood pulp fiber (36) (Boonpitaksakul et al. 2019). This can lead to a stronger fiber to fiber bonding during paper formation and increase the strength of the paper thus obtained (Gülsoy and Şimşir 2018).
All bamboo fibers had a high Runkel ratio (>1) and was in the range of 2.8 to 4.2. This indicated that bamboo fibers had rigid fibers and were not easily collapsed when compared to fibers with lower Runkel ratio (<1). This might lead to a higher beating time to develop the strength of paper (Gülsoy and Şimşir 2018).
Fiber flexibility ratio is the ratio of lumen diameter and fiber width. All bamboo fibers had a low flexibility ratio (<30) within a range of 19.8 to 27.0. (Gülsoy and Şimşir 2018) classified highly elastic fibers as those with a fiber flexibility ratio >75, elastic fibers between 50 and 70, rigid fibers between 30 and 50, and very rigid fibers (had a fiber flexibility ratio <30). As such, based on the fiber flexibility ratio, the bamboo fibers were classified as very rigid fibers, relative to the softwood pulp fiber (38) and hardwood pulp fiber (58) (Boonpitaksakul et al. 2019). Table 3 presents the chemical composition of the four bamboo species. The chemical composition of lignocellulose is an important parameter to evaluate the suitability for pulp production. The holocellulose content and alpha cellulose of the four bamboo species were between 72% and 74% and 46% and 50%, respectively. The results indicate that all bamboo species have a potential to be used as pulp material in pulp production since they have a cellulose content higher than 34% (Syed, Zakaria, and Bujang 2016). The cellulose content of bamboo species in our study is comparable to that of bamboo species and softwoods reported previously in the literature (Gülsoy and Şimşir 2018;Kamthai 2007;Rowell et al. 2012).

Chemical composition
The lignin content of all bamboo species was high and between 27% and 29%, similar to softwood (Mansouri et al. 2012;Rowell et al. 2012). Therefore, a longer cooking time and chemicals for delignification would be required during the pulping process. The ash content of all bamboo species was between 2% and 3%, which is higher than that reported for softwood and hardwood (González et al. 2013;Mansouri et al. 2012;Rowell et al. 2012). A high ash content can cause problems such as refining and in the recovery system during the pulping process (Gülsoy and Şimşir 2018;Kamthai 2007).
There were about 21-24% 1% NaOH extractives, 6-7% hot water extractives, 5-6% cold water extractives, and 2-3% ethanol-benzene extractives in the four bamboo species. This indicates that the bamboo species contained a high amount of low molecular weight carbohydrates, starch, inorganic compounds, and other constituents. This result is similar to that found for other bamboo species (Kamthai 2007;Moradbak et al. 2016;Nitisoravut and Malinen 2007).

Kraft pulping
Pulping conditions and cooking chemical dosages were optimized to produce unbleached pulp of kappa number between 20 and 25. Due to the high basic density and low lumen volume, to preserve the quality of the original fiber for morphological study, tradeoffs were severe as high chemical consumption, high temperature, and a long cooking time were required. Figure 5 shows the kraft pulp properties of bamboos. The pulp yield and reject were between 45% and 50% and 0.12% to 0.33%, respectively, and this result is closely related to the alpha cellulose of the bamboos (Table 3). The kappa number of the four bamboo species was between 18 and 24.

Surface morphological properties of the handsheets
The SEM micrographs of surface morphology of the bamboo pulp handsheets before and after the beating process with a beating time of 45 min is shown in Figure 6. All the unbeaten bamboo pulp handsheets had a high porosity, were bulky, weak bonding, and poor formation (Figure 6 upper row) compared to the beaten bamboo pulp handsheets. This result indicates that the beating caused a clear external fibrillation in all bamboo pulp handsheets and the fibrillation increased with a longer beating time. This might be able to improve the mechanical properties of the bamboo pulp handsheets (Khantayanuwong et al. 2021;Mohd-Hassan, Muhammed, and Ibrahim 2014). However, the bamboo fibers preserved their specific properties, i.e., long, narrow, thick-walled, small lumen, and very rigid fibers, as mentioned earlier.
Thus, a non-collapsed beaten bamboo pulp was still able to maintain its handsheet structure, as shown in Figure 6. Compared with the collapsibility of wood fibers, the bamboo fibers had a lower collapsibility (Khantayanuwong et al. 2021;Nitisoravut and Malinen 2007).

Physical properties of the handsheets
The properties of unbeaten and beaten bamboo pulps and their handsheet properties are shown in Figure 7. The freeness of all the bamboo pulps gradually decreased with increasing beating time. At the same time, the freeness levels and the freeness of all bamboo pulps decreased faster than that of bleached hardwood and softwood pulps (Khantayanuwong et al. 2021). This is probably because bamboo pulps fiber had a high thick-walled, resulting in higher fine formation, which enhanced the drainage resistance (Nitisoravut and Malinen 2007). The density, tensile index, burst index, and folding endurance of handsheet increased with increasing beating time (Figure 7b-f). This is probably because of the development of internal delamination and fibrillation in the bamboo pulp fibers, increased bonding ability, and collapsibility during beating process. The tensile index of bamboo pulps exhibited a slow development compared to that of softwood pulps at the same freeness levels. Thick-walled bamboo pulps tend to result in resistant internal fibrillation and therefore tensile index becomes lower than softwood pulps (Khantayanuwong et al. 2021). As a result, the thickwalled bamboo pulps would need more beating energy in order to develop sufficient bonding. Figure 7f shows the tear index of the bamboo pulps. The tear index values rapidly increased with increasing beating time from 0 to 15 min, and gradually decreased with increasing beating time from 30 to 45 min.

Conclusions
This study explores the potential of using the fibers and pulp obtained from fast-growing bamboo species as alternative fibrous raw material for paper production. The results indicate that the basic densities of bamboo species are very high, above 650 kg/m 3 . The bamboo fibers have long fibers similar to that of softwood fibers, and the fibers are thick-walled and narrow and have a small lumen. The chemical composition of bamboo species includes a high content of alpha cellulose, lignin, watersoluble matters, extractives, and ash. All the four bamboo species had a high pulp yield and low reject. When handsheets were prepared from the bamboo pulps with various beating time, apparent density, tensile index, burst index, and folding endurance of the handsheets improved with increasing beating time. Additionally, tear index rapidly improved during the early beating stage and gradually reduced after an extended beating time. All the bamboo pulps exhibited a good tearing resistance.