*Article* **Characterisation of Wood Particles Used in the Particleboard Production as a Function of Their Moisture Content**

**Dorota Dukarska 1,\* , Tomasz Rogozi ´nski <sup>2</sup> , Petar Antov <sup>3</sup> , Lubos Kristak <sup>4</sup> and Jakub Kmieciak <sup>1</sup>**


**Abstract:** The properties of particleboards and the course of their manufacturing process depend on the characteristics of wood particles, their degree of fineness, geometry, and moisture content. This research work aims to investigate the physical properties of wood particles used in the particleboard production in dependence on their moisture content. Two types of particles currently used in the production of three-layer particleboards, i.e., microparticles (MP) for the outer layers of particleboards and particles for the core layers (PCL), were used in the study. The particles with a moisture content of 0.55%, 3.5%, 7%, 10%, 15%, and 20% were tested for their poured bulk density (*ρp*), tapped bulk density (*ρt*), compression ratio (*k*), angle of repose (*αR*), and slippery angle of repose (*αs*). It was found that irrespective of the fineness of the particles, an increase in their moisture content caused an increase in the angle of repose and slippery angle of repose and an increase in poured and tapped bulk density, while for PCL, the biggest changes in bulk density occurred in the range up to 15% of moisture content, and for MP in the range above 7% of moisture content, respectively. An increase in the moisture content of PCL in the range studied results in a significant increase in the compression ratio from 47.1% to 66.7%. The compression ratio of MP increases only up to 15% of their moisture content—a change of value from 47.1% to 58.7%.

**Keywords:** wood particles; moisture content; angle of repose; slippery angle of repose; poured bulk density; tapped bulk density

#### **1. Introduction**

The physicomechanical properties of particleboards, except their technological parameters, depend significantly on the characteristics of the raw materials used in their production [1]. These are the type and amount of the adhesive system used for their bonding, the type of the raw material, its degree of fragmentation and geometry (size and shape), and moisture content [2–12]. The right choice of the moisture content of the particles, independently of their fineness degree, is also necessary for assuring the correct industrial process of particleboards. The excessive moisture content of particles may cause delamination of the particleboards during their pressing [13]. In turn, overly dry particles increase the risk of fires during drying and also contribute to the formation of wood dust, which disturbs the process of particles bonding or mat densification during hot pressing. One of the stages of wood particles preparation for particleboards production, independently of the material they originate from and bonding agents used, is the energy-consuming drying operation. Therefore, the possibility of producing boards from particles with higher moisture content is an opportunity to optimise energy consumption in production plants, e.g., by reducing the work of dryers or selecting appropriate drying conditions and thereby reducing the production costs [14,15]. Furthermore, the aspect of wood particles' moisture

**Citation:** Dukarska, D.; Rogozi ´nski, T.; Antov, P.; Kristak, L.; Kmieciak, J. Characterisation of Wood Particles Used in the Particleboard Production as a Function of Their Moisture Content. *Materials* **2022**, *15*, 48. https:// doi.org/10.3390/ma15010048

Academic Editor: Tomasz Sadowski

Received: 2 December 2021 Accepted: 20 December 2021 Published: 22 December 2021

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

**Copyright:** © 2021 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/).

content also seems to be of interest due to the trend of using isocyanate adhesives in the production process. Conventional thermosetting aminoplastic resins show deteriorated water resistance, which results from the hydrolysis of the methylene bridges [16,17]. However, polymeric 4,40 -diphenylmethane diisocyanate (pMDI) adhesives show high reactivity and the ability to a chemical reaction with wood components and the water they contain. It profitably influences the bonding quality of the particles and thereby the physicomechanical properties of particleboards [4,18–20]. These properties were studied by Jiang and Lu [21] for producing boards meeting the requirements of the EN 312 standard [22] prescribed for P2 type boards from particles with a moisture content of 25%, bonded with melamine–urea–formaldehyde resin (MUF) modified with different additional proportions of polyurethane prepolymer. As studies showed, the use of pure MUF resin for bonding particles with moisture content above 20% leads to the blowout of particleboards [13].

While preparing particles for the production of particleboards, it should also be considered that the moisture content of the mat increases together with the moisture content of the particles, which influences the parameters of the bonding process and the quality of particles bonding and, as a result, the properties of the finished boards. The moisture content of the mat is one of the most important factors influencing the heat transfer in the mat [23]. The rate of heat penetration in the mat determines the pressing time, which is critical for the efficiency of the production process [24]. It influences the increase in the temperature in the mat, which determines the cure rate of the adhesive resin. The combination of high temperature, moisture content, and time may cause an excessive increase in the vapour pressure in the mat and trigger an explosion when the press is opened [23,25]. However, the technological problems connected with the excessive moisture content of wood particles and vapour pressure during pressing can be effectively solved. Murayama et al. [26] investigated the temperature variability and vapour pressure during pressing of the particleboards and concluded that this problem can be solved by choosing an optimal moisture content of board layers. The increased moisture content of the face-layer and the lower face-layer thickness was expected to reduce the time of reaching the required temperature in the hot-pressing process. The usage of air-injection during the pressing of boards can also be a solution [19]. In the developed method, the air-injection press, which has holes punched in the heating plates, injects high-pressure air into the board through the holes of one plate and releases the air through the holes of the other plate. The advantage of this way of board pressing is allowing for reducing the pressing time required for manufacturing boards from high-moisture-content particles. Unfortunately, the air-injection press could not improve the properties of the particleboards [13].

As it results from the above considerations, the issue of the influence of particle moisture on the properties of finished particleboards and the course of their production process is practically well known. However, an often overlooked issue is the effect of particle moisture content on their physical properties such as poured bulk density, tapped bulk density, angle of repose, and slippery angle of repose. These properties are relevant because they influence the method and conditions of their storage and transport. They also influence the course of the production process, including the operation of transport and dosing devices and the choice of proper technological parameters during the bonding and pressing of particles, and consequently the properties of finished particleboards. Moisture content affects characteristics of bulk solids including wood and lignocellulosic particles such as particle-size distribution and bulk density [10,27,28]. The range of changes in these properties depending on the moisture content of the particles should be known in order to be able to assess its possible influence on the course of the production process and the properties of particleboards. Considering the above aspects, this work aimed to evaluate the physical properties of wood particles commonly used in the production process of three-layer particleboards depending on their moisture content.

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

#### *2.1. Materials 2.1. Materials*

Two types of wood particles used in the production of three-layer particleboards were used in the study, i.e., so-called microparticles intended for outer layers of particleboards and particles of the core layer (Figure 1). Two types of wood particles used in the production of three-layer particleboards were used in the study, i.e., so-called microparticles intended for outer layers of particleboards and particles of the core layer (Figure 1).

**Figure 1.** Images of wood particles: (**a**) microparticles (MP), (**b**) particles of the core layer particleboards (PCL). **Figure 1.** Images of wood particles: (**a**) microparticles (MP), (**b**) particles of the core layer particleboards (PCL).

These particles have been produced under industrial conditions mainly from middleand small-sized softwoods and selected low-density hardwood species. The raw material was also residues from the sawmilling industry in the form of shavings, sawdust, and chips. The initial moisture content of wood particles, determined by the drying–weighing method, was 7 ± 0.5%. To achieve the intended goal, the particles were submitted to drying or moisturising to the moisture content of 0.55%, 3.5%, 7%; 10%, 15%, and 20% ± 0.5%. In effect, the moisturising of different types of particles can be conducted directly or indirectly by increasing the moisture content of the environment [29]. In the present study, the first method was used i.e., wetting of the particles by spraying them with an appropriate amount of water and in order to homogenise the moisture content in the whole mass by seasoning for a period of 72 h. The moisture analyser MA R. 50 (Radom, Radwag, Poland) was used for the control of the moisture content of the particles. It determines the moisture content of the material on the basis of weight losses of the tested sample during its heating at a determined drying temperature (105°C was used in the study). For preliminary characterisation of the raw materials used in this study, their fractional composition was determined for particles with 0.55% moisture content and additionally for particles with 20% moisture content. In the case of PCL particles, the fractional composition was determined based on sieve analysis, with the use of flat sieves made of mesh with square perforations of 6.3, 5.0, 4.0, 2.5, 1.6, 1.0, and 0.5 mm. In turn, for MP particles, sieves with mesh sizes of 3.15, 1.25, 1.0, 0.63, 0.4, and 0.315 mm were used. For PCL particles, due to the greater differences in the shape and size of individual particles, the additional dimensional analysis was carried out, for which 250 particles of the predominant fraction (from a 2.5 mm mesh sieve) with the moisture content of 0.55% were drawn. Their length, width, and thickness were determined. This allowed the estimation of the basic shape factors of this type of particles, i.e., the degree of slenderness (௦), flatness (*ψ*), width coefficient (*m*), and specific surface area (௪) estimated according to the equations shown below [10]: These particles have been produced under industrial conditions mainly from middle- and small-sized softwoods and selected low-density hardwood species. The raw material was also residues from the sawmilling industry in the form of shavings, sawdust, and chips. The initial moisture content of wood particles, determined by the drying–weighing method, was 7 ± 0.5%. To achieve the intended goal, the particles were submitted to drying or moisturising to the moisture content of 0.55%, 3.5%, 7%; 10%, 15%, and 20% ± 0.5%. In effect, the moisturising of different types of particles can be conducted directly or indirectly by increasing the moisture content of the environment [29]. In the present study, the first method was used i.e., wetting of the particles by spraying them with an appropriate amount of water and in order to homogenise the moisture content in the whole mass by seasoning for a period of 72 h. The moisture analyser MA R. 50 (Radom, Radwag, Poland) was used for the control of the moisture content of the particles. It determines the moisture content of the material on the basis of weight losses of the tested sample during its heating at a determined drying temperature (105 ◦C was used in the study). For preliminary characterisation of the raw materials used in this study, their fractional composition was determined for particles with 0.55% moisture content and additionally for particles with 20% moisture content. In the case of PCL particles, the fractional composition was determined based on sieve analysis, with the use of flat sieves made of mesh with square perforations of 6.3, 5.0, 4.0, 2.5, 1.6, 1.0, and 0.5 mm. In turn, for MP particles, sieves with mesh sizes of 3.15, 1.25, 1.0, 0.63, 0.4, and 0.315 mm were used. For PCL particles, due to the greater differences in the shape and size of individual particles, the additional dimensional analysis was carried out, for which 250 particles of the predominant fraction (from a 2.5 mm mesh sieve) with the moisture content of 0.55% were drawn. Their length, width, and thickness were determined. This allowed the estimation of the basic shape factors of this type of particles, i.e., the degree of slenderness (*λs*), flatness (*ψ*), width coefficient (*m*), and specific surface area (*Fw*) estimated according to the equations shown below [10]:

$$
\lambda\_s = \frac{l}{h} \tag{1}
$$

$$
\psi = \frac{b}{h} \tag{2}
$$

$$
\hbar m = \frac{l}{b} \tag{3}
$$

$$F\_{\rm w} = \frac{0.002n}{w}(lh + lb + bh) \tag{4}$$

where *l*—mean length of wood particles (mm), *h*—mean thickness of wood particles (mm), *b*—mean width of wood particles (mm), *w*—mean weight of dry wood particles (g), *n*—number of wood particles selected for analysis, 0.002—coefficient taking into consideration the fact that wood particles have two surfaces and a unit converter from mm to m.

Such analysis in the case of particles with a diverse geometry (Figure 1) is justified because of the fact that the biomass particles are mostly inhomogeneous in terms of size and shape [30]. As a result, two particles going through the same sieve with the same mesh size may differ in shape. Therefore, the information obtained from the sieving process may not fully reflect the geometry of the biomass particles with such an irregular shape [31]. The parameters characterising the geometry of the PCL particles are presented in Table 1.



\* mean value, \*\* standard deviation.

Subsequently, the prepared material was tested for the influence of the moisture content of the particles on their poured bulk density (*ρp*), tapped bulk density (*ρt*), compression ratio (*k*), angle of repose (*αR*), and slippery angle of repose (*αs*).

#### *2.2. Poured Bulk Density and Tapped Bulk Density of Wood Particles*

The first parameter, the poured bulk density, was expressed as the ratio of the weight of loosely poured wood particles to their volume. To determine the effect of the moisture content of the tested particles on their tapped bulk density, particles loosely poured into a pot equipped with a volume scale were densified on a lab electromagnetic vibratory sieve shaker AS200 (Retsch GmbH, Haan, Germany) in the time of 10 min and with the vibration amplitude of 2 mm. The tapped bulk density was expressed as the ratio of the weight of poured wood particles to their volume recorded after the tapping. Based on the obtained results of the poured bulk density and tapped bulk density, the compression ratio (*k*) of the tested particles was determined, depending on their moisture content according to the equation [10]:

$$k = \left(\frac{\rho\_l}{\rho\_p} \cdot 100\right) - 100\tag{5}$$

where *k*—compression ratio (%), *ρp*—poured bulk density (kg/m<sup>3</sup> ), and *ρt*—tapped bulk density (kg/m<sup>3</sup> ).

The average values of poured bulk density and tapped bulk density for each tested variant were determined based on five unitary measurements.

#### *2.3. Angle of Repose and Slippery Angle of Repose of Wood Particles 2.3. Angle of Repose and Slippery Angle of Repose of Wood Particles*

variant were determined based on five unitary measurements.

*Materials* **2021**, *14*, x FOR PEER REVIEW 5 of 14

In general, the angle of repose is defined as the angle between the slant height and the base of a cone created during the loose falling of bulk material at right angles to the ground. However, due to the differences in the fineness and therefore the size of the tested particles, other procedures were used to determine the influence of the moisture content on the angle of repose of the particles. In the case of microparticles, the analysis of the angle of repose (according to PN-74 Z-04002.07 standard [32]) was based on pouring them in a steel discharge hopper with a calibrated hole with the diameter of 22 mm, which was attached to the base with a gear train (Figure 2a). Next, the hopper was being lifted with a linear movement to the moment of pouring of the microparticles on a plate with the diameter of 120 mm. After piling up a stable cone, its height was measured. In contrast, in the case of PCL-type particles, the angle of repose was determined by pouring them into a cylinder with the diameter of 120 mm and the height of 100 mm, and then by lifting it, a cone from the particles was formed whose height was decoded from the millimetre scale attached to the base of the device (Figure 2b) [33]. In general, the angle of repose is defined as the angle between the slant height and the base of a cone created during the loose falling of bulk material at right angles to the ground. However, due to the differences in the fineness and therefore the size of the tested particles, other procedures were used to determine the influence of the moisture content on the angle of repose of the particles. In the case of microparticles, the analysis of the angle of repose (according to PN-74 Z-04002.07 standard [32]) was based on pouring them in a steel discharge hopper with a calibrated hole with the diameter of 22 mm, which was attached to the base with a gear train (Figure 2a). Next, the hopper was being lifted with a linear movement to the moment of pouring of the microparticles on a plate with the diameter of 120 mm. After piling up a stable cone, its height was measured. In contrast, in the case of PCL-type particles, the angle of repose was determined by pouring them into a cylinder with the diameter of 120 mm and the height of 100 mm, and then by lifting it, a cone from the particles was formed whose height was decoded from the millimetre scale attached to the base of the device (Figure 2b) [33].

The average values of poured bulk density and tapped bulk density for each tested

**Figure 2.** Scheme of the device measuring the angle of repose of: (**a**) microparticles, (**b**) particles of the core layer of particleboards. **Figure 2.** Scheme of the device measuring the angle of repose of: (**a**) microparticles, (**b**) particles of the core layer of particleboards.

The angle of repose for both types of wood particles tested was estimated based on the equation: The angle of repose for both types of wood particles tested was estimated based on the equation:

$$\text{tg } \alpha\_R = \frac{2h}{D - d} \tag{6}$$
 
$$\text{mo biocht } \text{Imml} \text{ and } D \text{---cone } \text{base\'i} \text{m} \\ \text{s\'el\'as}$$

where ோ—angle of repose [°], *h*—cone height [mm], and *D*—cone base diameter [mm]. where *αR*—angle of repose [◦ ], *h*—cone height [mm], and *D*—cone base diameter [mm]. In the case of MP (Figure 2a), the value of *d* = 0.

In the case of MP (Figure 2a), the value of *d* = 0. The analysis of the slippery angle of repose was based on evenly pouring particles with the weight of about 200 g (MP and PCL) on a levelled and flat surface of the measuring device and then by lifting its side edge, finding the minimal rake angle that causes pouring of the layer of material. The slippery angle of repose value was directly decoded The analysis of the slippery angle of repose was based on evenly pouring particles with the weight of about 200 g (MP and PCL) on a levelled and flat surface of the measuring device and then by lifting its side edge, finding the minimal rake angle that causes pouring of the layer of material. The slippery angle of repose value was directly decoded from a protractor pitch attached to the base of the device (Figure 3). *Materials* **2021**, *14*, x FOR PEER REVIEW 6 of 14

ture as a qualitative factor.

*3.1. Fractional Composition of Wood Particles*

**3. Results** 

**Figure 3.** Scheme of the device measuring the slippery angle of repose of tested particles.

*2.4. Statistical Analysis*  with the larger ones, which escalated with the increase in their moisture content [34]. **Figure 3.** Scheme of the device measuring the slippery angle of repose of tested particles. The mean values of angle of repose and slippery angle of repose were determined based on five unitary measurements.

The obtained results of the tests selected for wood particles were statistically analysed using the STATISTICA v.13.1 software (StatSoft Inc., Tulsa, OK, USA). The mean values of the determined parameters were compared in the one-way analysis of vari-

analysis of variance was used, assuming the size of the fraction and wood particle mois-

The wood raw materials used in the study, independently of the degree of their fineness, posed a mixture of particles with diverse shapes and sizes. It is proclaimed by images of particles (Figure 1), the estimated coefficient of PCL shape (Table 1), and their fractional compositions (Figure 4), which were determined for dry particles (0.55%) and additionally for particles with the moisture content of 20%. Based on these diagrams, it can be concluded that in the case of MP independently of their moisture content, the particles of fractions 0.63 mm and 0.4 mm had the largest weight share. At a moisture content of 0.55%, the share of 0.63 mm fraction was 41.9%, and at a moisture content of 20%, it was slightly more, i.e., about 46.8%. In contrast, for the MP of fraction 0.4 mm, the share is greater by approximately 25%, which was observed for the moisture content of 0.55% in relation to the moisture content of 20% (respectively 30.4% and 24.2%). A similar dependence can be observed in the case of MP of a finer fraction, i.e., < 0.4 mm. In contrast, the tests of the PCL particles showed that the particles of fractions 2.5 and 1.4 mm had the biggest share in the whole mixture. The content of dry particles (i.e., with the moisture content of 0.55%) of fraction 2.5 mm remained on the level of about 30%, whereas dry particles of fraction 1.4 mm were on the level of about 25.6%. Analogically to MP, for the particles of this type, it was also observed that the increase in their moisture content caused a significant differentiation of their fractional composition. In general, it can be stated that the increase in the moisture content of particles resulted in a decreased share of the finer fractions (smaller than the predominant fraction) and increased share of the larger fractions. The observed differences were statistically significant, which was confirmed by the two-factor post hoc analysis, which in the case of MP of fractions 0.315–1.0 mm allowed for distinguishing six homogenous groups. In the case of PCL, statistically the greatest differences were observed for the fractions 1.4; 2.5, and 6.3 mm. The observed changes in the fractional composition of the tested particles can be explained by the fact that at higher moisture content of the particles, an insufficient separation of MP occurs due to their adhesion to larger particles. Finer particles show the ability to agglomerate

from a protractor pitch attached to the base of the device (Figure 3).

#### *2.4. Statistical Analysis*

The obtained results of the tests selected for wood particles were statistically analysed using the STATISTICA v.13.1 software (StatSoft Inc., Tulsa, OK, USA). The mean values of the determined parameters were compared in the one-way analysis of variance—Tukey's post hoc test, in which homogeneous groups of mean values for each parameter were identified for *p* = 0.05. In the case of the fractional composition, a two-factor analysis of variance was used, assuming the size of the fraction and wood particle moisture as a qualitative factor.

#### **3. Results**

#### *3.1. Fractional Composition of Wood Particles*

The wood raw materials used in the study, independently of the degree of their fineness, posed a mixture of particles with diverse shapes and sizes. It is proclaimed by images of particles (Figure 1), the estimated coefficient of PCL shape (Table 1), and their fractional compositions (Figure 4), which were determined for dry particles (0.55%) and additionally for particles with the moisture content of 20%. Based on these diagrams, it can be concluded that in the case of MP independently of their moisture content, the particles of fractions 0.63 mm and 0.4 mm had the largest weight share. At a moisture content of 0.55%, the share of 0.63 mm fraction was 41.9%, and at a moisture content of 20%, it was slightly more, i.e., about 46.8%. In contrast, for the MP of fraction 0.4 mm, the share is greater by approximately 25%, which was observed for the moisture content of 0.55% in relation to the moisture content of 20% (respectively 30.4% and 24.2%). A similar dependence can be observed in the case of MP of a finer fraction, i.e., <0.4 mm. In contrast, the tests of the PCL particles showed that the particles of fractions 2.5 and 1.4 mm had the biggest share in the whole mixture. The content of dry particles (i.e., with the moisture content of 0.55%) of fraction 2.5 mm remained on the level of about 30%, whereas dry particles of fraction 1.4 mm were on the level of about 25.6%. Analogically to MP, for the particles of this type, it was also observed that the increase in their moisture content caused a significant differentiation of their fractional composition. In general, it can be stated that the increase in the moisture content of particles resulted in a decreased share of the finer fractions (smaller than the predominant fraction) and increased share of the larger fractions. The observed differences were statistically significant, which was confirmed by the two-factor post hoc analysis, which in the case of MP of fractions 0.315–1.0 mm allowed for distinguishing six homogenous groups. In the case of PCL, statistically the greatest differences were observed for the fractions 1.4, 2.5, and 6.3 mm. The observed changes in the fractional composition of the tested particles can be explained by the fact that at higher moisture content of the particles, an insufficient separation of MP occurs due to their adhesion to larger particles. Finer particles show the ability to agglomerate with the larger ones, which escalated with the increase in their moisture content [34].

Tukey test).

**Figure 4.** Fractional composition of microparticles (**a**) and particles of the core layer of particleboards (**b**) depending on their moisture content (a, b, c …—homogeneous groups as determined by the **Figure 4.** Fractional composition of microparticles (**a**) and particles of the core layer of particleboards (**b**) depending on their moisture content (a, b, c . . . —homogeneous groups as determined by the Tukey test).

#### *3.2. Poured Bulk Density and Tapped Bulk Density of Wood Particles 3.2. Poured Bulk Density and Tapped Bulk Density of Wood Particles*

The poured bulk density of particles of different biomass types are dependent on their shape, size, the way of their forming in the mass, and the friction between the particles [35,36]. In the case of material such as wood particles, wood chips, or wood dust, the poured bulk density also depends on the absolute density of the wood itself [37]. Figure 5 presents the results of tests on the influence of particle moisture content on the formation of their poured bulk density and tapped bulk density. It was determined that in the case of MP particles, the increase in moisture content from 0.55% to 7% did not cause statistically significant differences in the values of the poured bulk density. The observed mean values in Tukey's test were classified in the same homogenous group (a). The further increase in the moisture content resulted in a gradual increase in the poured bulk density. The poured bulk density of particles of different biomass types are dependent on their shape, size, the way of their forming in the mass, and the friction between the particles [35,36]. In the case of material such as wood particles, wood chips, or wood dust, the poured bulk density also depends on the absolute density of the wood itself [37]. Figure 5 presents the results of tests on the influence of particle moisture content on the formation of their poured bulk density and tapped bulk density. It was determined that in the case of MP particles, the increase in moisture content from 0.55% to 7% did not cause statistically significant differences in the values of the poured bulk density. The observed mean values *ρ<sup>p</sup>* in Tukey's test were classified in the same homogenous group (a). The further increase in the moisture content resulted in a gradual increase in the poured bulk density.

**Figure 5.** Poured bulk density and the tapped bulk density of microparticles (**a**) and wood particles of the core layer of particleboard (**b**) depending on their moisture content (a, b, c …—homogeneous groups as determined by the Tukey test). **Figure 5.** Poured bulk density and the tapped bulk density of microparticles (**a**) and wood particles of the core layer of particleboard (**b**) depending on their moisture content (a, b, c . . . —homogeneous groups as determined by the Tukey test).

The MP-type particles with a moisture content of 20% showed the value higher by about 14% than the ones with the lowest moisture content, whereby the highest increase in density occurred by changing the moisture content from 15% to 20%. In the case of PCL, the increase in poured bulk density was observed just by 9% but to the moisture content of 15%. Regarding the PCL particles with a moisture content of 20%, a small but statistically significant decrease was observed. Previous studies on the influence of particle moisture content of different types of biomass (almond nut, sunflower seed, flaxseed, straw) on their physical properties showed that this relationship can develop in different ways [2,10,38–40]. For example, Dukarska et al. [10] and Aviara et al. [40] proved that the bulk density of seeds (*Moringa oleifera)* or straw particles of selected grain species increased with respect to an increase in moisture content. In contrast, other researchers such as Littelield et al. [41] showed that the bulk density of pecan shells decreased with an increase in moisture content. According to the authors, the decrease in the poured bulk density of the particles is caused by the increase in their size under the influence of the moisture content and thereby their volume, which was following on faster than the weight The MP-type particles with a moisture content of 20% showed the *ρ<sup>p</sup>* value higher by about 14% than the ones with the lowest moisture content, whereby the highest increase in density occurred by changing the moisture content from 15% to 20%. In the case of PCL, the increase in poured bulk density was observed just by 9% but to the moisture content of 15%. Regarding the PCL particles with a moisture content of 20%, a small but statistically significant decrease *ρ<sup>p</sup>* was observed. Previous studies on the influence of particle moisture content of different types of biomass (almond nut, sunflower seed, flaxseed, straw) on their physical properties showed that this relationship can develop in different ways [2,10,38–40]. For example, Dukarska et al. [10] and Aviara et al. [40] proved that the bulk density of seeds (*Moringa oleifera)* or straw particles of selected grain species increased with respect to an increase in moisture content. In contrast, other researchers such as Littelield et al. [41] showed that the bulk density of pecan shells decreased with an increase in moisture content. According to the authors, the decrease in the poured bulk density of the particles is caused by the increase in their size under the influence of the moisture content and thereby their volume, which was following on faster than the weight increase as a result of increasing the moisture content. As expected, along with the increase in the moisture content of particles,

their tapped bulk density also increased. Considering the statistical analysis, the influence of the moisture content of particles on their tapped bulk density was greater than in the case of poured bulk density. It can be observed especially with regard to the PCL particles. Analysing the results obtained for MP, it can be stated that this density increased more with the increase in the moisture content from 0.55 to 3.5% and from 10% to 15%. However, the increase in the moisture content of MP from 0.55% to 20% caused the increase in their tapped bulk density by about 12% and in the case of PCL by about 22%. It results from the fact that the higher moisture content of MP (similar to wood dust) contributes to an increase in their volume, which causes the decrease in free space around them and the increase in consistency of the whole mass of particles [10,27,28]. In practice, the raw material with higher moisture content, independently of their degree of fineness, requires an increase in the volume essential to their storage or transport [41]. Moreover, it can be concluded that at higher moisture content, the PCL particles are more susceptible than MP to compaction by vibration, which can cause some difficulties in the technological process of producing particleboards and during their transport.

Comparing the results from Figure 5, it was also established that in the tested range of the moisture content, MP were marked by higher poured and tapped bulk density than the particles used for the core layers of particleboards. The literature shows that fine particles (such as MP characterised by their small size) are better at filling empty spaces during their pouring than larger particles. As the sizes of the particles increase, larger particles cannot sufficiently fill the empty spaces during tapping, which causes the decrease in their tapped bulk density [41]. This corresponds with the works of other authors who, studying the effect of particle size of different types of biomass, also observed that the value *ρ<sup>t</sup>* increases with decreasing particle size of cereal straw, corn straw, switchgrass, and nutshells [35,41,42]. As demonstrated by Dukarska et al. [10], in the case of cereal straw, stem morphology is also an important factor that is related to the size and geometry of the particles, in particular, their thickness and degree of slenderness.

#### *3.3. Compression Ratio*

The changes in poured and tapped bulk density caused by the changes in the moisture content of the tested wood particles are reflected also in their compression ratio. A graphical representation of the compression ratio of wood particles depending on their moisture content is presented in Figure 6. It can be stated that in the case of PCL, along with the increase in their moisture content from 0.55% to 20%, a gradual and significant increase in the value of this parameter by about 48% was observed. The increase in the degree of fineness of the wood raw material to MP caused the decrease in vulnerability to its compressibility. It was observed that the increase in their moisture content resulted in the increase in the compression ratio from 44.9% to 58.7%, so by about 25% in comparison to the dry particles, however just in the moisture content range to 15%. By 20% moisture content of the MP, the decrease in its moisture content was set down to the level, which had been set down for dry particles. This phenomenon probably results from the small sizes of MP and their significant compaction in the whole mass. Moreover, increasing their moisture content resulted in reduced free spaces between particles and increased their adhesion to each other. When a particulate system becomes damp, the cohesion increases due to the creation of the liquid bridge bonds between the particles. The system remains stable until the moisture content is too high for strong bridges. When the particulate system becomes more and more wet, the material reaches the state of a slurry. During drying out the dump particulate system, the solid bonds and bridges between the particles can be formed. As a result of this, the material will become cohesive again. Changes in moisture content and resulting changes in other properties of a particulate system can cause serious problems in handling particles in an industrial installation [43]. In addition, biological materials (which may also include wood particles) become softer; thus, deformation is greater with an increase in moisture content [41]. In contrast to MP, increasing the moisture content of PCL chips from 15% to 20% results in a further increase in their compression ratio up

to 66.7%. This might be attributed to the changes in the dimensions of the particles on account of their swell influenced by the increasing moisture content. Along with increasing dimensions of the particles, the volume of the space between them in a layer also increased and can be filled with the particles during tapping. compression ratio up to 66.7%. This might be attributed to the changes in the dimensions of the particles on account of their swell influenced by the increasing moisture content. Along with increasing dimensions of the particles, the volume of the space between them in a layer also increased and can be filled with the particles during tapping.

logical materials (which may also include wood particles) become softer; thus, deformation is greater with an increase in moisture content [41]. In contrast to MP, increasing the moisture content of PCL chips from 15% to 20% results in a further increase in their

*Materials* **2021**, *14*, x FOR PEER REVIEW 10 of 14

**Figure 6.** The compression ratio of wood particles depending on their moisture content. **Figure 6.** The compression ratio of wood particles depending on their moisture content.

#### *3.4. Angle of Repose and Slippery Angle of Repose of Wood Particles 3.4. Angle of Repose and Slippery Angle of Repose of Wood Particles*

The angle of repose and slippery angle of repose are important physical parameters characterising wood particles. Graphical representation of the angle of repose and the slippery angle of repose of MP and PCL depending on their moisture content is presented analytically in Figure 7. Independently of the fineness of the particles, along with increasing their moisture content, the angle of repose and slippery angle of repose also increased. However, it can be observed that the biggest changes in values were recorded for the angle of repose of PCL particles. In this case, the increased moisture content from 0.55% to 20% resulted in increased ோ from 36.6° to 47.6°, i.e., by about 30%, whereby, as the variance analysis showed, statistically significant differences were set down for the moisture content higher than 3.5%. This is evidenced by the results of the analysis of variance, which identified five different homogeneous groups for each PCL moisture content above this value. Increasing the moisture content had much less effect on the angle of repose of the MP, for which the maximum value of ோ totals just 7%. Markedly, the greatest differences between MP and PCL particles were observed in the range of lower moisture content (up to 7%). This is an important observation in view of the fact that in the industrial practice, depending on the technology, the moisture content of the particles for the outer layers of the three-layer particleboards varies from 2% to 8% and in the case of the core layers, it varies from 1% to 6%. The results of the multivariate significance tests ANOVA presented in Table 2 are a confirmation that the lack of interactions between the tested effects (the type and moisture content of particles) were in the range of the higher moisture content, over 10% (*p* > 0.05). The results obtained for the angle of repose and the slippery angle of repose can be explained by the differences in their sizes and shape as well as by their consistency and looseness. The physical properties of wood particles, e.g., the size, shape, or roughness of the surface significantly affect the looseness of the particles of different types of biomass [30,31,36]. According to the subject literature, the increased particle size is connected with the decrease in their tenacity and therefore the value of the angle of The angle of repose and slippery angle of repose are important physical parameters characterising wood particles. Graphical representation of the angle of repose and the slippery angle of repose of MP and PCL depending on their moisture content is presented analytically in Figure 7. Independently of the fineness of the particles, along with increasing their moisture content, the angle of repose and slippery angle of repose also increased. However, it can be observed that the biggest changes in values were recorded for the angle of repose of PCL particles. In this case, the increased moisture content from 0.55% to 20% resulted in increased *α<sup>R</sup>* from 36.6◦ to 47.6◦ , i.e., by about 30%, whereby, as the variance analysis showed, statistically significant differences were set down for the moisture content higher than 3.5%. This is evidenced by the results of the analysis of variance, which identified five different homogeneous groups for each PCL moisture content above this value. Increasing the moisture content had much less effect on the angle of repose of the MP, for which the maximum value of *α<sup>R</sup>* totals just 7%. Markedly, the greatest differences between MP and PCL particles were observed in the range of lower moisture content (up to 7%). This is an important observation in view of the fact that in the industrial practice, depending on the technology, the moisture content of the particles for the outer layers of the three-layer particleboards varies from 2% to 8% and in the case of the core layers, it varies from 1% to 6%. The results of the multivariate significance tests ANOVA presented in Table 2 are a confirmation that the lack of interactions between the tested effects (the type and moisture content of particles) were in the range of the higher moisture content, over 10% (*p* > 0.05). The results obtained for the angle of repose and the slippery angle of repose can be explained by the differences in their sizes and shape as well as by their consistency and looseness. The physical properties of wood particles, e.g., the size, shape, or roughness of the surface significantly affect the looseness of the particles of different types of biomass [30,31,36]. According to the subject literature, the increased particle size is connected with the decrease in their tenacity and therefore the value of the angle of repose [30]. In turn, fine particles, which can include wood MP, have a higher specific surface area, which increases the contact and cohesiveness among the particles, which may cause difficulties in their flow in technological conditions [31].

repose [30]. In turn, fine particles, which can include wood MP, have a higher specific surface area, which increases the contact and cohesiveness among the particles, which

may cause difficulties in their flow in technological conditions [31].

**Figure 7.** The angle of repose and the slippery angle of repose of microparticles (**a**) and particles of the core layer of particleboard (**b**) depending on their moisture content (a, b, c …—homogeneous groups as determined by the Tukey test). **Figure 7.** The angle of repose and the slippery angle of repose of microparticles (**a**) and particles of the core layer of particleboard (**b**) depending on their moisture content (a, b, c . . . —homogeneous groups as determined by the Tukey test).

**Table 2.** Multivariate significance tests of the MP and PCL particle angle of repose. **Table 2.** Multivariate significance tests of the MP and PCL particle angle of repose.


With regard to the slippery angle of repose, no significant differences were determined. Independently from the degree of fineness of the particles the increase in the value *α<sup>s</sup>* was on the same level, i.e., 17% on average. The increase in the slippery angles of repose along

with the increased moisture content of particles may indicate an increase in their traction and thereby restriction of their mobility. Comparing the data obtained for both tested types of wood particles, it can be concluded that the values of MP's slippery angles of repose were greater than the PCL particles of particleboards.

#### **4. Conclusions**

Based on the conducted research, the influence of the moisture content variation on the physical properties of PCL and MP was determined. It was demonstrated that the increase in the moisture content of the particles resulted in increased dimensions of wood particles, independently of their degree of fineness, and increased values of the slippery angle of repose and the angle of repose. Furthermore, the increased moisture content of wood particles led to enhanced tapped bulk density for both types of studied wood particles for the production of three-layer particleboards. Regarding the poured bulk density, increased values were determined at 15% moisture content and over 7% moisture content for the PCL particles and MP, respectively. The PCL particle compression ratio was also increased. With regard to the MP particles, this increase was determined only at moisture content values up to 15%. The results of the present study can be utilised in the industrial practice of the wood-based panel industry to optimise the technological parameters and production costs related to particleboard manufacturing.

**Author Contributions:** Conceptualisation, D.D. and T.R.; methodology, D.D. and T.R.; validation, D.D., T.R., P.A. and L.K.; formal analysis, P.A. and L.K.; investigation, D.D., T.R. and J.K; resources, D.D. and T.R.; writing—original draft preparation, D.D. and T.R.; writing—review and editing P.A. and L.K.; visualisation, D.D., T.R. and J.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** The study was supported by the funding for statutory R&D activities as the research task No. 506.224.02.00 and 506.227.02.00 of the Faculty of Forestry and Wood Technology, Pozna´n University of Life Sciences. This research was also supported by the project No. HИC-Б-1145/04.2021 "Development, Properties and Application of Eco-Friendly Wood-Based Composites" carried out at the University of Forestry, Sofia, Bulgaria. This research was also supported by the Slovak Research and Development Agency under the contracts No. APVV-18-0378, APVV-19-0269.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

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

#### **References**


**Anita Wronka <sup>1</sup> , Eduardo Robles <sup>2</sup> and Grzegorz Kowaluk 1,\***


**Abstract:** This research aimed to confirm the ability to reduce carbon dioxide emissions by novel composite production using plantation waste on the example of lignocellulosic particles of black chokeberry (*Aronia melanocarpa* (Michx.) Elliott) and raspberry (*Rubus idaeus* L.). Furthermore, to characterize the particles produced by re-milled particleboards made of the above-mentioned alternative raw materials in the light of further recycling. As part of the research, particleboards from wooden black chokeberry and raspberry were produced in laboratory conditions, and select mechanical and physical properties were examined. In addition, the characterization of raw materials (particles) on the different processing stages was determined, and the fraction share and shape of particles after re-milling of the produced panels was provided. The tests confirmed the possibility of producing particleboards from the raw materials used; however, in the case of boards with raspberry lignocellulose particles, their share cannot exceed 50% so as to comply with the European standards regarding bending strength criterion. In addition, the further utilization of chips made of re-milled panels can be limited due to the significantly different shape and fraction share of achieved particles.

**Keywords:** biopolymer; wood; upcycling; composite; recycling; mechanical properties; physical properties; carbon storage; raspberry; black chokeberry; bio waste

#### **1. Introduction**

Every year, society's awareness of caring for the Earth is growing. The growing amount of waste is a problem, with its storage and greater carbon dioxide emissions. In the case of fruit bushes grown in Poland, which are pruned each year, their branches are often left in the field or are burned. It can be used as a biofuel to avoid wasting energy, but it is not yet a common practice in Poland. Another way to use orchard waste is to produce three-layer particleboards for the furniture industry. Even though the tree species used for wood products are renewable, it should not be limited only to it because renewable does not mean that it is infinite. Because of this, it is necessary to explore using other lignocellulosic materials that will fully or partially replace the wood raw material. This attempt to move into the broad utilization of renewable biopolymers was also suggested by Bari and collaborators [1]. Some attempts have already been made that have been proven to be more or less effective, for this purpose, materials such as pepper stalks [2], sugarcane [3,4], almond shell [5], apple and plum branches [6], bamboo chips [7], straw [8], wheat straw and corn pith [9], kiwi prunings [10], coffee husk [11], flax shiv [12], acai (*Euterpe oleracea* Mart.) fruit [13], oil palm empty fruit bunch [14], and kenaf [15] were used. The use of wooden lignocellulosic parts of fruit plant waste allows for the binding of carbon dioxide in the form of particleboards, without emitting it into the atmosphere. In this field, good examples are raspberry (*Rubus idaeus* L.) and chokeberry (*Aronia melanocarpa* (Michx) Elliott) plantation waste. These represent substantial waste in Polish fields, as

**Citation:** Wronka, A.; Robles, E.; Kowaluk, G. Upcycling and Recycling Potential of Selected Lignocellulosic Waste Biomass. *Materials* **2021**, *14*, 7772. https:// doi.org/10.3390/ma14247772

Academic Editors: L'uboš Krišt'ák, Petar Antov and Réh Roman

Received: 23 November 2021 Accepted: 12 December 2021 Published: 16 December 2021

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

**Copyright:** © 2021 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/).

Polish production accounts for 60–70% of the world's production potential. The cultivation area of chokeberry is about 40 km<sup>2</sup> per year and the annual harvest of fruits is from 40 to 60 thousand tonnes. The main recipients of chokeberry fruits are China, Japan, and South Korea [16]. Whereas the area of raspberry cultivation in Poland is over 290 km<sup>2</sup> , placing it at fifth place in the world's raspberry producers and third in Europe (after Russia and Serbia) [17].

The fruits of these shrubs are cultivated for their taste and health benefits. Raspberry fruits are rich in anthocyanins and have anti-inflammatory and anticancer properties, so it is often recommended to drink raspberry juice during colds [18]. Medicinal values also characterize black chokeberry fruits; just like raspberries, they have an antioxidant effect, and their consumption is recommended to prevent chronic diseases [19]. This added value for the fruits allows for assuming that the potential availability of lignocellulosic resources of those above-mentioned alternative raw materials will grow shortly. Therefore, it seems worth researching the development of long-term storage regarding the carbon fixed in these raw materials, such as producing particleboards and attempting to upcycle these wooden wastes and recycle the produced composites.

This investigation aimed to determine the ability to utilize raspberry and chokeberry lignocellulosic particles to produce particleboards for furniture purposes and to characterize the wooden particles produced by the re-milling particleboards mentioned regarding further recycling. As a result, the following hypothesis has been investigated: the lignocellulosic particles of raspberry and black chokeberry are valuable raw materials to produce the particleboards and obtain particles from re-milled panels, which can potentially be re-used in particleboard production.

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

#### *2.1. Materials*

Raspberry (*Rubus idaeus* L.) (Figure 2) and black chokeberry (*Aronia melanocarpa* (Michx.) Elliott) (Figure 3) wooden stalks were used for the current work. Two year old raspberry stalks, as well as four year old chokeberry rods, were collected from Polish fields. The raw materials were dried in a chamber drier under 70 ◦C to air-dry the moisture content (about 10–12%), and the bark content (*w/w*) was measured by manual debarking about 2 kg of each tested material. The wooden branches of the chokeberry and raspberries were shredded on saw blade in separate batches (50 mm long chips) and then grounded into a fine fraction using a laboratory three-knife drum mill (laboratory prototype) with an outlet equipped with 6 <sup>×</sup> 12 mm<sup>2</sup> mesh to form particles. The bulk density of the particles was calculated as the weight of a selected fraction, divided by the measuring cylinder's capacity (in volume). The measurement was repeated five times for every fraction. The produced particles were sorted on mesh of size 0.5 and 1 mm (face layers), and 8 mm and 2 mm (core layer) to exclude the oversized and undersized particles. The pictures of the cross-cuts of the investigated raw materials were taken with a NIKON SMZ 1500 (Kabushiki-gaisha Nikon, Minato, Tokyo, Japan) optical microscope.

#### *2.2. Elaboration of Composites*

Three-layer composites were produced as particleboards (PB) with different black chokeberry and raspberry contents. The lignocellulosic particles were dried to a moisture content of 5%. As a result, particleboards with a nominal density of 600 kg m−<sup>3</sup> , 32% (*w/w*) of face layer content, and a total thickness of 16 mm were produced. The following content (*w/w*) of alternative raw materials was applied: 0% (reference panels, 100% of industrial (coniferous) particles), 10%, 25%, 50%, and 100%. The industrial urea-formaldehyde resin Silekol S-123 (Silekol Sp. z o.o., K˛edzierzyn—Ko´zle, Poland) was used to resinate the particles, where the resination of particles for the face layer was 12% and the core layer was 10%. No hydrophobic agent (like paraffin emulsion) was added. The curing was done for 82 s inside an oven at 100 ◦C. Panels were pressed on a hydraulic press (ZUP-NYSA PH-1P125) at a maximum pressure of 2.5 MPa, with a temperature of 200 ◦C, and a time

factor of 20 s mm−<sup>1</sup> . The produced boards were conditioned before the tests in a climatic chamber (producer: Research and Development Centre for Wood-Based Panels Sp. z o. o. in Czarna Woda, Poland) at 20 ◦C and 65% air humidity, until a constant mass was obtained. The main steps of the material flow and samples preparation are presented in Figure 1.

**Figure 1.** The process of material flow and samples preparation chart.

#### *2.3. Characterization of the Elaborated Panels*

All of the elaborated PB were conditioned at 20 ◦C, and the test specimens were cut on a saw blade, as required by European standards EN-326-2 [20] and EN-326-1 [21]. The modulus of rupture (MOR) and elasticity (MOE) were determined according to EN 310 [22], and the internal bond (IB) was determined according to EN 319 [23]. All the mechanical properties were examined with an INSTRON 3369 (Instron, Norwood, MA, USA) laboratory-testing machine, and, whenever applicable, the results were referred to standards [24]. Board density was determined according to EN 323 [25], thickness swelling (TS) to EN 317 [26], and surface water absorption was done following EN 381-1 [27]. The density profiles of the tested PB (three types: reference, 100% of raspberry, and 100% of chokeberry) were measured on a GreCon DAX 5000 device (Fagus-GreCon Greten GmbH and Co. KG, Alfeld/Hannover, Germany).

#### *2.4. Raw Material Recycling and Characterization*

The composites were re-milled on a laboratory knife mill (laboratory prototype delivered by Research and Development Centre for Wood-Based Panels Sp. z o. o. in Czarna Woda, Poland) equipped with three knives, two contra-knives, and a 6 <sup>×</sup> 12 mm<sup>2</sup> mesh. The fraction of chips taken from the re-milled particleboards was tested with an IMAL (Imal s.r.l., San Damaso (MO), Italy) vibrating laboratory sorter with seven sieves. The selected sieve sizes were 8, 4, 2, 1, 0.5, 0.25, and <0.25 mm. The amount of tested material for each fraction was about 100 g, and the set time of continuous vibrating was 5 min. As many as five repetitions were done for every tested material.

#### *2.5. Statistical Analysis*

Analysis of variance (ANOVA) and t-tests calculations were used to test (α = 0.05) for significant differences between factors and levels, where appropriate, using IBM SPSS statistic base (IBM, SPSS 20, Armonk, NY, USA). A comparison of the means was performed when the ANOVA indicated a significant difference by employing the Duncan test. The statistically significant differences for the achieved results are given in the Results and Discussion paragraph whenever the data were evaluated.

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

#### *3.1. Materials Characterization*

The bark content (*w/w*) was 7.4% for raspberry and 18.0% for chokeberry. According to [28], the average bark content of pine (*Pinus sylvestris* L.), which is the main raw material for particleboard production in Poland, is about 6.7% (*w/w*). Significant differences in bark content of the investigated materials were found. Such a high content of bark in the case of chokeberry could influence the mechanical properties of the produced PB [29]. It was found by Kowaluk et al. [6] that the bark density of orchard trees can be remarkably lower than the density of the wood. As confirmed in the case of single-layer particleboards produced from *Quercus cerris* bark [30], these panels had remarkably low mechanical properties when compared to the commercial particleboards. What was also confirmed by the mentioned researchers, is that the panels produced from *Quercus cerris* bark had low TS. The bark particles, being highly brittle, could also raise the fine particles/dust production when milled, which could negatively influence the mechanical properties of the panels. On the other hand, a fine bark particle can be upcycled and utilized, as was proven by Mirski and collaborators [31].

Concerning the anatomy of the investigated raw materials, raspberry (Figure 2) has a large amount (volumetric) of foamy parenchyma pith. However, this part of the material can be easily disintegrated mechanically, and it is not easy to separate the particles produced from the remaining particles. Furthermore, as a brittle and soft tissue, it produces a large amount of fine particles, characterized by a large specific surface. This feature is not desirable in PB production, since this fraction requires a high amount of resin to be added. If the resination is not tuned regarding these fine fractions, the mechanical parameter of produced PB drops down.

**Figure 2.** Cross-cut: (**a**) across and (**b**) along the fibers of a raspberry stalk.

The cross-cut of chokeberry (Figure 3) can be referred to as broadleaf plants. The year rings (Figure 3a) are clearly visible, and wood rays are going horizontally between bark and pith (Figure 3b; bark on the left, pith on the right). It is worth pointing out that the pith is also in foam form, which was found for raspberry stalks, but here the amount of foam pith was significantly lower than for raspberry.

**Figure 3.** Cross-cut: (**a**) across and (**b**) along the fiber of the chokeberry stalk.

In Figure 4, the results of the measurement of the bulk density of particles used to produce the tested composites and those produced by re-milling of the tested composites are presented. In the case of the face layers' intended particles, the highest bulk density was found for chokeberry particles (164 kg m−<sup>3</sup> ). A 2.4% lower bulk density (when referred to highest value) was found for industrial face layer particles (160 kg m−<sup>3</sup> ). The lowest bulk density value among the tested particles was registered for raspberry particles (83 kg m−<sup>3</sup> ), which means an almost 50% lower density for chokeberry. When analyzing the core layer purpose particles, the results were as follows (descending order): industrial (157 kg m−<sup>3</sup> ), chokeberry (121 kg m−<sup>3</sup> ), and raspberry (89 kg m−<sup>3</sup> ). The results of the measurement of the bulk density of re-milled particles show that the bulk density of these particles was higher than for the primary particles, and, what should be pointed out, is that the differences between the tested materials were less than 2% when considering the lowest value. The achieved average bulk density values were statistically significantly different when compared within the same group (face, core, and re-milled).

**Figure 4.** Bulk density of particles used to produce composites, considering the particles produced by the re-milling of composites.

The achieved density values of raspberry and chokeberry particles were low compared to other alternative lignocellulosic raw materials [6]. This is promising information, as, in the case of compressed lignocellulosic composites, a low bulk density leads to better densification, creating more spots where separate particles are connected. Thus, the produced composite structure is more even, less porous, and has higher mechanical properties. This can also lead to lower water absorption. However, it was confirmed by Papadopoulos et al. [7] that a lower bulk density can reduce the mat permeability due to densification during hot pressing, and the heat transfer through such a mat can be significantly slower.

#### *3.2. Modulus of Rupture and Modulus of Elasticity*

As shown in Figure 5, the modulus of rupture values decreased when the content of alternative raw materials increased. The MOR decrease was higher for composites produced of raspberry particles (from over 15 N mm−<sup>2</sup> when 0% of raspberry particles to less than 10 N mm−<sup>2</sup> for 100% raspberry composite). In the same conditions, MOR decreased for chokeberry composites that had reached over 12.1 N mm−<sup>2</sup> . When compared within the same raw materials, the only statistically significant differences between average MOR values were found for the highest and lowest content of raw materials. When referring to the EN standard [24], it was found that in the case of raspberry, the content of alternative raw material should not exceed 50% for meeting the standard requirements.

**Figure 5.** Modulus of rupture of the tested composites.

Similar tendencies for a reduction of MOE when the alternative raw material contend grows are presented in Figure 6. The reduction of MOE from the value of reference composite, 2805 N mm−<sup>2</sup> , was to 1800 N mm−<sup>2</sup> for the fully chokeberry composite and to 1617 N mm−<sup>2</sup> for the 100% raspberry composite. It is worth adding that when referring to the EN standard [24], in the case of raspberry, the content of alternative raw material should not exceed about 80% for meeting the standard requirements. Furthermore, statistically significant differences for the average MOR values for chokeberry were found for all composite types excluding 25 and 50% alternative raw material particles share, when, in the case of raspberry, there were no statistically significant differences between the composites of 10 and 25%, as well as between 50 and 100%.

**Figure 6.** Modulus of elasticity of the tested composites.

Raspberry panels were expected to have the best mechanical properties because of their lower bulk density; however, they presented low mechanical features for MOR and MOE. The reason for this can be the content of the core, as can be appreciated in Figure 2. This part of the raw material can influence the production of particles with a low bulk density, but these particles do not allow for carrying a high mechanical load when the samples are bent, the face layers are strained/compressed, and core layers are under shear stress. Moreover, the geometry of the particles used to produce the composites can play a role here. As Wronka and Kowaluk demonstrated [32], the raspberry particles are shorter and have blunt (wide) ends when compared to industrially used softwood particles. Furthermore, because of the structure of the raspberry stalk, where the region of higher mechanical properties is on the external zone of the rod, the particles produced from this raw material are of a lower length to thickness ratio (slenderness), which is not desirable for particle composites. It has been confirmed [33] that the best mechanical properties for composites are achievable with a high length to thickness ratio.

#### *3.3. Internal Bond*

The positive effect of a low bulk density of raspberry particles has been found when analyzing the IB values of the tested composites. As shown in Figure 7, the IB was significantly raised when the content raspberry particles rose. The reference composite IB value was 0.72 N mm−<sup>2</sup> , and, for 100% raspberry composite, the IB was 1.04 N mm−<sup>2</sup> , while for 100% chokeberry composite, it was 0.53 N mm−<sup>2</sup> . It should be pointed out that when comparing the achieved results of IB, all of the tested composites met the requirements of a specified European standard [24]. Furthermore, the statistical analyses within the alternative raw materials mentioned show no statistically significant differences between IB average values of 10% and 25%, 25%, and 50% for raspberry, as well as between 10% and 50%, and between 25% and 100% for chokeberry.

#### *3.4. Thickness Swelling and Water Absorption*

The results of the measurement of thickness swelling of the tested composites after 2 h and 24 h of soaking in water are presented in Figure 8. As can be seen, in the case of raspberry composites, the swelling in thickness significantly grew with the alternative raw material content increase. After 2 h of soaking, the lowest TS for the reference composite (0% of raspberry particles) was 18%, while for the 100% raspberry panel, the TS was 33%, which is an increase of more than 83%. After 24 h of soaking, the TS of the reference composite was below 20%, and for the 100% raspberry composite, the TS was over 36% (89% growth). In the case of chokeberry, the increase in alternative raw material particles content caused

a decrease in thickness swelling. After 24 h of soaking of the chokeberry composites, the TS was 16%, which was an almost 16% reduction of TS. The only statistically significant differences for the average values of TS after 2 h of raspberry composites were found between the 0%, 50%, and 100% panels and the same composites after 24 h of soaking. Regarding chokeberry, statistically significant differences after 2 h of soaking were found for composites of 0% and 100%, and the same after 24 h of soaking. It should be highlighted that when referring to the achieved results of TS, none of the tested composites met the requirements of the European standard [24].

**Figure 7.** Internal bond of the tested composites.

**Figure 8.** Thickness swelling of the tested composites.

Such a significant rise of TS of composites made of raspberry particles can be explained by the low bulk density of raspberry particles, as presented in Figure 4. Although a low bulk density helps with better densification of the pressed mat, this highly compressed mat can be easily decompressed by water penetrating the composite in light of the swelling in thickness. Thus, the material, which was more densified during hot pressing (composite preparation), has a potential of higher TS. On the other hand, the opposite situation was found in the case of chokeberry composites, where the mat densification was lower due to the higher bulk density of the chokeberry particles.

The WA values of the tested composites of different contents of alternative raw materials are presented in Figure 9. The high water absorption values after 24 h of soaking for the raspberry samples, from over 77% for the reference composite to over 108% for 100% raspberry composite, can be explained by the presence of low-density core particles, which can react with water like a sponge. The higher increase of WA for the samples with a higher content of raspberry particles after 24 h compared to WA after 2 h of soaking means that the structure of the samples is less penetrative (tighter) against water, and more time is needed to reach the deeper zones of the samples. This can be explained by the higher densification of the mat built by low bulk density particles. When evaluating the WA of chokeberry composites, it can be found that with the rising content of chokeberry particles, the WA slightly rose after 2 h of soaking, whereas, after 24 h of soaking, the WA decreased with the increase in chokeberry particles content. This means that chokeberry particles cause lower water absorption. A specific type of composite here can be the 100% chokeberry panel, where the maximum WA was reached after 2 h of soaking and did not raise even after 24 h of total soaking. One of the reasons could be the high bulk density of chokeberry particles, which lead to lower compression of particles during pressing, and leave more unfilled (empty) zones in the composite structure. These zones can be filled with water in a short time. Another reason is that the deciduous wood has a five times higher potential to transfer the water due to the larger dimensions of the vessels [28]. Statistically significant differences of average WA for raspberry composites after 2 h of soaking were between 0% and 100%, and between 0%, 10%, and 25% against 50% and 100% composites after 24 h. For chokeberry, these differences were found between samples of 0%, 10%, and 25% against 100% after 2 h, and 0%, 10%, and 25% against 50% and 100% composites after 24 h.

**Figure 9.** Water absorption of the tested composites.

#### *3.5. Density and Density Profiles*

The results of the density profile measurement of the tested composites are presented in Figure 10. Since the tested composites are symmetrical, half of the density profile is shown to improve the readability of the plots. As can be seen, the highest values of density in the face zone, over 950 kg m−<sup>3</sup> , located about 0.7 mm in deep from the surface, were found for the industrial particles composite. On the other hand, the highest density zone of the raspberry composite, about 815 kg m−<sup>3</sup> , was found about 1.8 mm under the composite surface. A similar zone for the chokeberry composite, but with a lower density, about 780 kg m−<sup>3</sup> , was found at 0.4 mm under the surface. In the case of the lowest density in the core layers (middle of the thickness), the lowest value, about 550 kg m−<sup>3</sup> , was registered for the raspberry composite, when the remaining composites had a similar core layer density,

which was about 590 kg m−<sup>3</sup> . It should be mentioned here that all of the tested samples were of the same average density of about 600 kg m−<sup>3</sup> .

A high density of face layers, which was found for the reference composites (industrial particles), can significantly influence the bending properties of composites as these face layers are generally responsible for tension and compression stresses when the material is bent. This remark can be confirmed on figures presenting MOR (Figure 5) and MOE (Figure 6) values. It was also confirmed for fibrous composites of different density profiles [34]. However, as the differences between the density values of core layers of the tested panels are low, it can be hard to refer to the remaining features of the tested composites.

#### *3.6. Recycled Material Characterization*

The pictures of different particles produced by re-milling the tested composites and industrial (not re-milled) particles are shown in Figure 11. It can be found that in the case of industrial particles, for all fractions excluding dust (< 0.25 mm), the particles had a high length to width ratio, which can be estimated on the level of 4:1 for 8 mm fraction and even higher, about 20:1 for 2 mm fraction. On the other hand, the pictures of re-milled particles of fractions 8 and 4 mm show that the particles were not elongated anymore, and these were more rounded or square, with a length-to-width ratio of about 1:1. For a fraction of 2 mm, a significant difference was found for chokeberry particles, which are more similar to industrial particles. In addition, the smaller chokeberry particles (1 mm and below) are closer to industrial (not re-milled) particles. When re-milled, industrial and raspberry particles are short and of higher width. These remarks can be valuable in the light of further use of re-milled particles, since, as confirmed, the shape of the particles can significantly influence the properties of the produced particle composites [35,36].

**Figure 10.** Density profiles of the tested composites.


**Figure 11.** Pictures of the morphology of particles produced by re-milling the tested composites compared to industrial particles (each picture dimension is 50 mm × 50 mm). **Figure 11.** Pictures of the morphology of particles produced by re-milling the tested composites compared to industrial particles (each picture dimension is 50 mm × 50 mm).

The mass fraction share of particles produced by the re-milling of investigated composites and the fraction share of primary industrial particles is shown in Figure 12. As can be seen, in the case of raw industrial particles, the largest share is for particles of size 1 and 2 mm (over 74%), and 17% of size 4 mm. The remaining fractions are less than 9%. When analyzing the fraction share of re-milled particles, it can be stated that the fraction share of industrial and chokeberry particles is similar. The difference is between the distribution of fractions smaller and larger than 1 mm: for industrial re-milled particles, a more significant amount of fractions smaller than 1 mm were found, and a smaller amount of fractions were larger than 1 mm. The opposite distribution was found for chokeberry particles. Significant differences in fraction share were found for raspberry re-milled particles. These particles had many fractions of bigger dimensions, where the content of fractions of 4 mm + 8 mm was about 47%. This type of material also provides a large amount of smaller fractions: the sum of fractions 0.5 mm, 0.25 mm, and < 0.25 mm was about 30%, whereas, in the case of remaining re-milled materials, it was about 23% for industrial re-milled, 18% for re-milled chokeberry, and 8% for primary industrial particles. It should be pointed that a high amount of small fractions is not profitable when considering the achieved particles to be used as a raw material to produce similar particle composites. Since the specific surface of the particles grows with the particle size decrease, and, thus, a larger amount of binder is needed to cover the particle surface adequately, such small fractions should be separated and subjected to alternative processing/utilization.

**Figure 12.** Mass fraction share of particles produced by re-milling of composites considering industrial particles.

#### **4. Conclusions**

According to the conducted research and the analysis of the achieved results, the following conclusions and observations can be drawn:


**Author Contributions:** A.W. took part in designing the experiments and performed the experiments, analyzed the data, and wrote the first draft of the paper; E.R. analyzed data statistically and wrote the final version of the paper; and G.K. designed the experiments, and analyzed the data. All the authors assisted in writing and improving the paper. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research was financed by the own funds of the Institute of Wood Sciences and Furniture at the Warsaw University of Life Sciences—SGGW. E.R. wants to acknowledge the tenure track position "Bois: Biobased materials", part of E2S UPPA supported by the "Investissements d'Avenir" French program managed by the ANR (ANR-16-IDEX-0002).

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** Some of the mentioned tests have been completed within the Student Furniture Scientific Group (Koło Naukowe Meblarstwa), Faculty of Wood Technology, Warsaw University of Life Sciences, SGGW.

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

#### **References**

