Abstract
This work aimed to evaluate the effect of the precipitation of CaCO3 via subsequential in-situ mineral formation based on a solution-exchange process of two solution-exchange cycles via impregnation with CaCl2 in ethanol and NaHCO3 in water. The effects were investigated in terms of the structure of the wood and the thermal, physical, mechanical, and decay resistance properties of nine species commonly used in commercial reforestation in Costa Rica. The thermogravimetric analysis results showed that the woods with the highest formation of CaCO3 showed a more pronounced signal at 200 °C in relation to untreated/wood; therefore, they were more thermostable. The fire-retardancy test showed that flame time in CaCO3/wood composites was longer than for untreated/wood in half of the species tested, presenting a positive effect of mineralization. Wood density, decay resistance, modulus of rupture (MOR), modulus of elasticity (MOE) in flexion, and MOR in compression were slightly affected by mineralization. Water absorption increased, but it had no negative effect on the dimensional stability. In general, mineralization can be a chemical treatment to increase the dimensional stability and fire resistance of hardwood species without modifying the wood’s physical and mechanical properties.
Download PDF
Full Article
Effect of CaCO3 on the Wood Properties of Tropical Hardwood Species from Fast-growth Plantation in Costa Rica
Roger Moya,a,* Johanna Gaitán-Álvarez,a Alexander Berrocal,a and Fabio Araya b
This work aimed to evaluate the effect of the precipitation of CaCO3 via subsequential in-situ mineral formation based on a solution-exchange process of two solution-exchange cycles via impregnation with CaCl2 in ethanol and NaHCO3 in water. The effects were investigated in terms of the structure of the wood and the thermal, physical, mechanical, and decay resistance properties of nine species commonly used in commercial reforestation in Costa Rica. The thermogravimetric analysis results showed that the woods with the highest formation of CaCO3 showed a more pronounced signal at 200 °C in relation to untreated/wood; therefore, they were more thermostable. The fire-retardancy test showed that flame time in CaCO3/wood composites was longer than for untreated/wood in half of the species tested, presenting a positive effect of mineralization. Wood density, decay resistance, modulus of rupture (MOR), modulus of elasticity (MOE) in flexion, and MOR in compression were slightly affected by mineralization. Water absorption increased, but it had no negative effect on the dimensional stability. In general, mineralization can be a chemical treatment to increase the dimensional stability and fire resistance of hardwood species without modifying the wood’s physical and mechanical properties.
Keywords: Wood treatment; Wood chemical treatment; Wood plantation; Fast-growth plantation; Wood properties
Contact information: a: Instituto Tecnológico de Costa Rica, Escuela de Ingeniería Forestal, Apartado 159-7050, Cartago, Costa Rica; b: Centro de Investigación y de Servicios Químicos y Microbiológicos (CEQIATEC), Escuela de Química, Instituto Tecnológico de Costa Rica, Cartago 159-7050, Costa Rica;
* Corresponding author: rmoya@itcr.ac.cr
INTRODUCTION
Minerals are common in living organisms, where they maintain rigidity and hardness of the structures. The process of mineral formation by living organisms is known as biomineralization (Krajewska 2018). Minerals are also found in different shapes and sizes in natural deposits in the Earth’s crust (Krajewska 2018). Biomineralization can be biologically controlled in the cell process (Benzerara et al. 2011; Gadd et al. 2012, 2014; Li et al. 2014, 2015; Kumari et al. 2016). There are around 60 different biological minerals (Anbu et al. 2016). The Earth’s crust minerals constitute over 4% of biological minerals, and they are commonly found in rocks such as chalk, marble, travertine, and tuff, among others (Krajewska 2018).
Among the wide range of minerals from biomineralization and deposits of the Earth’s crust, calcium carbonate (CaCO3) is the main product and is considered the most useful today (Rodriguez-Navarro et al. 2012; Dhami et al. 2013; Kumari et al. 2016). Additionally, CaCO3 appears as different polymorphs (calcite, aragonite, and vaterite); two hydrated polymorphs (monohydrocalcite and ikaite), and several amorphous phases (Sáãnchez-Román et al. 2007; Dhami et al. 2014). Calcite is the primary CaCO3 product and it is the most thermodynamically stable polymorph of CaCO3 (Okwadha and Li 2010; Ganendra et al. 2014). Vaterite, in turn, is considered a minor, metastable and transitional phase to calcite formation (Tourney and Ngwenya 2009). This mineral has been used for decades in construction and, more recently, it has been implemented in the agricultural, medical and engineering industries (Hoque 2013). Specifically, it is applied in the paper, paint, food, ceramics, construction, ink, adhesives, drugs, cable, and plastic industries (Ozen et al. 2013). In the plastics industry for example, it is used as a loading agent to substitute high value polymers (Ozen et al. 2013). For water purification it can be used to eliminate ions of heavy metals such as Cu+2, Pb+2, Cd+2, Zn+2, and Cr+5 (Hong et al. 2011). In the paper industry, it is used for producing high gloss, greater opacity and whiteness (Barhoum et al. 2014; Wu et al. 2016).
In contrast, wood has a porous network that poses a series of limitations, namely, high moisture content and hygroscopicity, dimensional instability, and high flammability (Uribe and Ayala 2015). However, this porous network is an optimal platform for the deposition of inorganic matter such as CaCO3 (Merk et al. 2016); therefore, minerals can be introduced in its hierarchical structure to help improve the material characteristics (Tampieri et al. 2009; Hübert et al. 2010; Shabir Mahr et al. 2012). Nevertheless, the precipitation process of crystals of CaCO3 is highly complex, because it depends on factors that affect the nucleation process and the subsequent crystal growth, which are often difficult to control (Declet et al. 2016).
Nevertheless, the first experiences on mineralization or CaCO3 formation inside the wood were reported a few years ago (Tsioptsias and Panayiotou 2011; Merk et al. 2015; Uribe and Ayala 2015). Tsioptsias and Pahayictou (2011) introduced CaCO3 using aqueous solutions and supercritical carbon dioxide in Picea abies, corroborating that under controlled conditions the treatment has fire-retardant effects. They concluded that the fire retardation is achieved in both glowing and smolding combustion and may be due to different mechanisms. Seeking fire retardancy, Merk et al. (2015, 2016) used a modern and simple method based on a subsequential in-situ mineral formation based on a solution-exchange process in beech and spruce woods. Those studies showed that the formation of crystals of polycrystalline calcite and vaterite inside the wood occurs in lumina vessels and to a lesser extent in adjacent cell walls. Recently, using the methodology proposed by Merk et al. (2016), Gaitán-Alvarez et al. (2020) studied CaCO3 precipitation in tropical hardwoods and determined that it occurs in the form of calcite and vaterite, mainly in the vessels and rays of the species. They also determined that in-situ crystal formation is difficult to control.
Despite the mentioned studies, the knowledge about the effect of CaCO3 precipitation on the wood properties is limited; besides, the studies are limited to a few softwoods, such as beech, spruce, and Picea, and the processes of wood impregnation have been performed at a low scale (Tsioptsias and Panayiotou 2011; Klaithong et al. 2013; Burgert et al. 2016). Because of the great number of tropical timber species and the large variety of anatomical structures (Tenorio et al. 2016; Liu et al. 2018), studying the effect of CaCO3 precipitation on wood properties becomes highly relevant to solve the problems of high moisture content, dimensional stability, and natural decay, among others (Mantanis 2017).
Given this context, the present work aimed at evaluating the effects of CaCO3 precipitation on wood structure, on the thermal properties (thermal stability and fire retardancy), physical properties (density, moisture content, water absorption by immersion, swelling, and moisture equilibrium), mechanical properties (static flexure and grain parallel compression), and durability properties (natural decay) of nine species commonly used in commercial reforestation in Costa Rica. Studying these effects, it will be possible to treat tropical species with CaCO3 to increase their fire retardancy and characterize the principal changes in wood properties.
EXPERIMENTAL
Materials
Sapwood from nine species from fast growth forest plantations in Costa Rica presenting good permeability was used (Moya et al. 2015). The species were Cedrela odorata (Co), Cordia alliodora (Ca), Enterolobium cyclocarpum (Ec), Gmelina arborea (Ga), Hieronyma alchorneoides (Ha), Samanea saman (Ss), Tectona grandis (Tg), Vochysia ferruginea (Vf), and Vochysia guatemalensis (Vg). The age of the plantations that contributed the material ranged between 4 and 8 years. Three trees per species were cut down for sampling. Then, the samples were cut into 1-m-long logs and sawn into 7.5 cm wide × 2.5-cm-thick boards. These boards were air dried until 12 to 15% moisture content was reached. Then, pieces 46 cm (length) × 7.5 cm (width) × 2 cm (thickness) were extracted, making sure that the pieces were composed of sapwood.
The reagents used were ETI SODA (Ciner Group, Istanbul, Turkey), solid-state sodium bicarbonate (NaHCO3), and solid-state calcium chloride (CaCl2) from CASO FCC FLAKES Solvay (Missouri, Il, USA). Absolute ethyl alcohol 99% m/m of the brand Reactivos Químicos Gamma (San José, Costa Rica), distributed by Laboratorios Químicos ARVI S.A, was used.
Mineralization process
The in-situ mineralization process was performed in 20 samples (46 cm long × 7.5 cm wide × 2 cm thick), using the vacuum and pressure equipment shown in the diagram in Fig. 1a. The samples were placed into a tank (25 cm diameter × 48 cm length × 27 L capacity) for impregnation with salts. Calcium carbonate (CaCO3) formation (Eq. 1) was performed by means of two impregnation cycles. Figure 1b shows impregnation with CaCl2 (Eq. 2) and impregnation with NaHCO3 (Eq. 3). In the CaCl2 cycle, the wood was impregnated with CaCl2 in an ethanol solution at 1 mol-L-1 concentration (Eq. 2), by first applying vacuum at -70 kPa (gauge) for 20 min and then immersing the samples totally into the solution and applying 690 kPa pressure for 30 min (Fig. 1a). In the NaHCO3 cycle, the samples were impregnated with an aqueous solution at 1 mol-L-1 concentration (Eq. 2), by applying vacuum for 20 min, followed by 690 kPa pressure for 50 min. After finishing the first cycle, the samples were washed in distilled water and then were left to dry inside a chamber under controlled conditions (22 °C and 66% relative humidity) until reaching 12% moisture content. After the second cycle, the samples were washed in ethanol and then dried at 40 °C for 4 h. The details of this process can be consulted in Gaitán-Alvarez et al. (2020). Equations 1 through 3 are as follows:
Fig. 1. Flow chart of equipment pressure (a) and flow chart, temperature, and time for in-situ mineralization (b) used for in-situ mineralization
Methods
Evaluation of the mineralization process
In the CaCl2 cycle, the samples were weighed before and after impregnation with CaCl2. Thus, absorption of the CaCl2 solution (Eq. 4) and salt retention (Eq. 5) were obtained. In the NaHCO3 cycle, again the sample was weighed before and after impregnation with NaHCO3 and again absorption of the NaHCO3 solution (Eq. 4) and salt retention (Eq. 5) were obtained. The details regarding salt retention and CaCO3 formation can be found in Gaitán-Alvarez et al. (2020). Salt absorption and retention were calculated according to Eqs. 4 and 5:
Thermogravimetric analysis
The thermogravimetric analysis (TGA) was realized for CaCO3/wood composites and 5 mg previously dried sawdust from untreated/wood was used. The heating rate used was 20 °C min-1 from 50 °C to 800 °C, in an ultra-high purity nitrogen atmosphere at a flow of 100 mL min-1. A thermogravimetric analyzer from TA Instruments (Lukens drive, New Castle, USA), model SDT Q600, was used. The TGA gave values of weight loss relative to temperature, these values were used to perform the derivative thermogravimetric analysis (DTG), which was used to obtain the temperature at which the sample was degraded. The TGA data and its derivatives were analyzed using TA Instruments (TA Instruments, Lukens Drive New Castle, USA) Universal Analysis 2000 software.
Physical properties
The following physical properties were determined: density, moisture content (MC), water absorption of immersed wood in water, tangential and radial swelling of wood, and moisture absorption due to changes in the condition of equilibrium moisture content (EMC) (from 12% to 18%). Sample density was determined for 20 CaCO3/wood composites samples and 20 untreated/wood samples; the samples’ volume and weight were measured, and then the density (weight/volume) was determined. The MC was calculated for 20 CaCO3/wood composites samples and 20 untreated/wood samples, following the procedures in ASTM D4442-42 (2016). To determine the tangential and radial swelling and the percentage of water absorption, 20 samples 5 × 5 × 2 cm from each treatment were conditioned and weighed at 12% EMC and then conditioned for a period of 3 to 4 weeks at 18% EMC. After conditioning, the samples were weighed and measured again. This procedure was based on the ASTM D4933-99 (1999) standard but modified to the conditions in Costa Rica, where environmental moisture conditions are approximately 18%. Tangential and radial swelling were calculated using Eq. 6. Moisture absorption was determined through Eq. 8 by obtaining first 12% and 18% moisture content with respect to dry weight (Eq. 7). The difference between both moisture measurements (12% and 18%) represented sample moisture absorption (Eq. 8):
For weight gain from water immersion, 20 samples for each species/treatment previously weighed were immersed in water for 24 h. After this period, the samples were weighed and the weight gain was determined following Eq. 7, according to ASTM D4446-13 (1985).
Mechanical properties
The mechanical properties analyzed were resistance to static flexure and grain parallel compression. In both tests, procedure B described in ASTM D143 (2016) was used. Twenty samples were taken from each species/treatment for each one of the tests. The tests were performed in a testing machine Tinius Olsen model H10KT (Tinius Olsen Test Machine Company, Horsham, PA, USA) in the case of static flexure and Tinius Olsen L60 (Tinius Olsen Test Machine Company, Horsham PA, USA) for compression.
Fire retardancy
For this test, 10 CaCO3/wood composite samples and 10 untreated/wood samples were used for each species. The sample size was 15 cm in length × 10 cm in width × 9 mm in thickness, in accordance with the methodology proposed by Taghiyari (2012). The samples were placed in the device specially designed for this test. Once the sample was placed at an angle of 45° and 27 mm from the flame, the test timing began, as follows: (i) dark time, the moment when the back of the sample exposed to the flame began to form a dark spot; (ii) hole time, the time when a hole begins to appear on the back of the sample exposed to fire; (iii) flame time, the moment when the flame begins to appear on the surface exposed to fire; (iv) ember time, when the embers begin to appear around the burning hole; and (v) end time, the time when the test stops, when most of the surface of the sample has already been consumed by fire. Weight loss was also recorded during the curfew test, recording the weight before and after the test.
Durability
The methodology in ASTM D2017-81 (1995) was used to perform the test of resistance to natural decay. In each treatment/species, 40 samples, 2 cm wide × 2 cm long × 2 cm thick were prepared. Two types of fungi were used for this test, Trametes versicolor and Lenzites acuta, corresponding to white and brown rot, respectively. Both fungi are of university collection, usually used in assessment of decay resistance wood (TEC, Cartago, Costa Rica). Twenty samples per species/treatment were subjected to degradation by each one of the fungi, for a period of 12 weeks.
Statistical Analysis
The statistical analysis consisted primarily of checking the normality and homogeneity of the data and the elimination of outliers. Then, the descriptive analysis consisted of determining the mean, standard deviation, and coefficient of variation for each variable studied for each species and treatment. For each variable evaluated, an analysis of variance (ANOVA) was performed with a level of statistical significance of p < 0.05 to determine variability in response to the mineralization treatment. Tukey’s test was used to determine the statistical significance of the differences between the means of the variables. This analysis was done with the SAS 9.4 program (SAS Institute Inc., Cary, NC, USA).
RESULTS AND DISCUSSION
Absorbed amounts of the solutions of CaCl2 and NaCO3 in the different species ranged from 42.2 to 168.0 L/m3 and 69.5 to 214.6 L/m3 (Table 1), respectively. The ANOVA for these two parameters showed that absorption of the CaCl2 solution in-situ was higher in Ec, Ha, Ss, Vf, and Vg, while lower absorptions were observed in Co, Ga, and Tg (Table 1). Meanwhile, absorption of the NaCO3 solution appeared in greater quantities in Ha, Ss, and Vf; and in lesser quantities in Co, Ga, and Tg (Table 1). Additionally, CaCO3 retention in-situ ranged from 2.8 to 9.2 kg/m3 and was greater in Ec, Ha, Ss, Vf, and Vg, and smaller in Co, Ga, and Tg, with statistical differences between them (Table 1). Details of the formation and location of CaCO3 crystals can be found in Gaitán-Alvarez et al. (2020), and Fig. 2 shows the salt formation in E. cyclocarpum, with greater formation and C. odorata with lower salt formation (Table 1). As for unreacted salts in the wood, surplus ions of Ca2+ and CO32+ could be observed within the same species, reaching up to 40% of the treated species (Table 1). A surplus of Ca2+ ions was observed in Ca, Ec, Ss, and Vg, while in Ca, Ga, Ha, Tg, and Vf the surplus ions were CO32+. As for the amount of unreacted salt relative to volume, Ec presented the highest values with 3.8 and 6.0 g/m3 of CO32- and Ca2+, respectively, followed by Vf and Vg. The wood with less amounts of unreacted salts were Co and Ca wood (Table 1).
Table 1. Absorption of CaCl2 and NaCO3, and Retention of CaCO3 Estimated in the in-situ Mineralization Process of Nine Woods from Fast-growth Tropical Species in Costa Rica
The numbers in parentheses represent variation coefficients, and different letters in each parameter represent significant statistical differences (p < 0.05).
Fig. 2. SEM images showing in-situ CaCO3 formation in different anatomical features in E. cyclocarpum with highest salt formation and C. odorata with lowest salt formation: High quantity of CaCO3 crystals formed in lumina of vessels (a) and in lumina ray parenchyma (b) of E. cyclocarpum and lower of CaCO3 crystals formed in lumina of vessels (a) and in lumina ray parenchyma (b) of C. odorata. Arrow shows CaCO3 crystals.
Although adequate absorptions were achieved to have the number of ions (Ca2+ or CO32+) sufficient for CaCO3 formation, the stoichiometry is not suitable for the formation of this salt (Table 1). Salt formation is a complex mechanism that includes the polymorphism exhibited by crystals of calcium carbonate, calcium oxalate, and different polymorphs of the same compound that can transform into each other. Crystal formation also influences the environmental temperature, concentration, pH, viscosity, additives, and so on (Zeng et al. 2018), which are conditions difficult to control in processes and equipment used in commercial preservation intended for application in wood mineralization.
Thermogravimetric Analysis
The TGA of the two different salts used for in-situ CaCO3 formation showed that for CaCl2 used in the first cycle, decomposition occurred in two phases: the first phase with the highest peak at 149 °C and the second phase with the highest peak at 200 °C (Fig. 3a).
Fig. 3. TGA and DTG analysis of CaCO3/wood composites and untreated/wood of nine fast growth tropical species in Costa Rica
Regarding NaHCO3, decomposition occurred in one phase, with the highest peak showing at 160 °C (Fig. 3b). In CaCO3, decomposition occurred in only one phase, with the maximum peak at 98 °C (Fig. 3c).
In relation to CaCO3/wood composites, thermal decomposition of the nine woods showed the same pattern of untreated wood in the TGA; however, the DTG curve showed differences in the CaCO3/wood composites of the various species (Fig. 3a through l). The DTG curve presented three important decomposition stages: the first one appeared as a signal in the form of a slight curve after 200 °C; the second stage showed at the highest peak of decompositions between 320 °C and 360 °C; and the third as a slight signal at the end after 380 °C (Fig. 3d through l). All of the woods studied showed that the maximum decomposition peak of the CaCO3/wood composites occurred before untreated/wood (Fig. 3d through l). The DTG curve also showed that in the woods with higher CaCO3 formation (Ec, Ss, Vf, and Vg) the signal at 200 °C was more pronounced than the signal of untreated/wood at 200 °C (Fig. 3f, i, k, and l). Meanwhile, in the species with average retention (15.2 to 17.4 kg/m3), such as Ca and Ha wood, the pronunciation of the curve at 200 °C was slight, almost similar to untreated/wood (Fig. 3c and h). Additionally, CaCO3/wood composites of Co, Ga, and Tg wood that presented the smallest values of CaCO3 retention, followed the same behavior of untreated/wood (Fig. 3e, g, and j).
The TGA results showed that, except for Co and Ss, at the beginning CaCO3/wood composites of the species lost weight faster than untreated/wood; however, after 350 °C the opposite behaviour occurred: untreated/wood lost weight faster (Fig. 3d through 3l). This effect was more pronounced in CaCO3/wood composites above 16.29 kg/m3 (Ec, Ha, Ss, Vf, and Vg) (Fig. 3f, 3h, 3i, 3k, and 3l). These results agree with those obtained by Tsioptsias and Pnayiotou (2011), who found the same behavior in CaCO3/wood composites of Picea abies. Fast initial degradation of CaCO3/wood composites was due to early endothermal degradation of CaCO3, water, and CO2 (Dash et al. 2000), where most degradation occurred before reaching 100 °C (Fig. 3c). However, at higher temperatures (> 350 °C), CaCO3/wood composites turned more stable to combustion due to aragonite transforming into calcite at 387 °C (Brown and Gallagher 2003), which decomposed into CaO and CO2 at temperatures above 850 °C (Singh and Singh 2007).
Fire Resistance
In relation to flame time, CaCO3/wood composites of Ca, Ec, Ha, and Vg needed more time than untreated/wood of the same species (Table 2). The remaining species (Ca, Ga, Sa, Tg, and Vf) showed no statistical differences between CaCO3/wood composites and untreated/wood (Table 2). Ember time varied again between species and was statistically higher in CaCO3/wood composites of Co, Ec, Ha, and Ss in relation to untreated/wood. In CaCO3/wood composites of Ha and Tg, ember time was shorter than in untreated/wood. No statistical differences appeared in the remaining species (Ca, Ga, Vf, and Vg) (Table 2). Dark time and hole time were not statistically affected between CaCO3/wood composites and untreated/wood of Ca, Ec, Ga, Ha, Vf, and Vg. Dark time increased statistically in CaCO3/wood composites of Ca and Ss but diminished statistically in CaCO3/wood composites of Tg (Table 2). Lastly, the total testing time varied with the type of wood (Fig. 4b). Ca, Ga, Vf, and Vg woods showed no statistical differences between CaCO3/wood composites and untreated/wood (Fig. 4b). The end time in CaCO3/wood composites of Co, Ec, and Ss was statistically higher than in untreated/wood. In contrast, in CaCO3/wood composites of Ha and Tg, the end time was statistically lower than in untreated/wood (Fig. 4b).
Table 2. Different Times in the Fire-retardancy Test of CaCO3/Wood Composites and Untreated/Wood of Nine Fast-growth Tropical Species in Costa Rica
The numbers in parentheses represent variation coefficients and different letters in each parameter represent significant statistical differences (p < 0.05)
The evaluation of the weight loss at the end of the flame test demonstrated that all the mineralized woods showed less weight loss relative to untreated/wood; however, in CaCO3/wood composites of Ss wood the weight loss was statistically higher than in CaCO3/wood composites (Fig. 4b).
The fire resistance analysis confirmed the previous results. CaCO3/wood composites of species showing greater in-situ CaCO3 formation, such as Ec, Ss, Vf, and Vg (Table 1), resisted fire better (Fig. 4; Table 2). These results agree with those obtained by Merk et al. (2015, 2016) and Tsioptsias and Panayiotou (2011) in CaCO3/wood composites of beech and spruce. Merk et al. (2015) explained that impregnating the wood with CaCO3 affects the thermal decomposition of the cellulose by reducing the formation of volatiles. Similarly, Yao et al. (2008) indicated that minerals embedded in the tissues dilute the amount of fuel material and constitute a barrier to heat transmission and weight transport during pyrolysis. However, other authors (Walters and Lyon 2003; Hull et al. 2011) indicated that CaCO3 limits the supply of oxygen and volatile material. Therefore, incrementing the quantity of embedded CaCO3 crystals in Ec, Ss, Vf, and Vg reduces the flammability of the mineralized material and increases its thermal stability and fire resistance (Fig. 3; Table 2).