Valorizing the Organic Fraction of Municipal Solid Waste (OFMSW) as Composite Panels for Construction or Furniture

Abstract

Residual lignocellulosic biomass represents a major resource to be incorporated into the circular economy, with up to 1400 Mt/y in EU27. Due to its complex composition of three biopolymers (cellulose, hemicellulose and lignin) combined with its seasonal and regional variability and high water content, its valorization involves manifold challenging aspects. Herein a three-step procedure is presented to transform this type of biomass into solid composite panels: hydrothermal carbonization (HTC), dry thermal treatment and curing a phenolic resin. HTC triggers chemical dehydration of the polysaccharide part of the lignocellulose and breaks up the cell structure of the plants. This facilitates the diffusion of the water and its separation by filtration, which is more energy efficient than evaporation. HTC and thermal treatment induce chemical changes that concentrate the carbon content and make the material suitable for crosslinking with a phenolic resin, achieving a 90% renewable content. The composite panels are competitive with products of the particle and fiberboard sector with respect to tensile strength and screw withdrawal resistance. Hence, the products can be employed for construction or in the furniture industry.

1. Introduction

The European Green Deal was proposed by the European Commission in 2019 and approved by the European Parliament in 2020 and consists of a disrupting policy declaration for the subsequent decades [1,2]. This document was not only designed as a climate change mitigation action but also for increasing the well-being and health of citizens by demanding access to clean water, construction of energy efficient buildings, production of healthy and affordable food, support for public transport and many more benefits.

With respect to climate change mitigation, the aim is to reduce net greenhouse gas emissions by at least 55% by 2030 compared to 1990 levels [1]. Fossil-based heat and fuel in industry, buildings and the transport sector have to be replaced by renewables, as they are the cleanest energy source available. Furthermore, resources should be saved by increasing the circular economy, in addition to using biomass preferentially, i.e., by re-directing materials from the end of the value chain, such as textiles, electronics or biomass, towards other industrial sectors.

The residual lignocellulosic biomass potential in Europe is huge and adds up to 1400 Mt/year for EU-27 and Switzerland, with an energy content of 8500 PJ [3]. The main resources are straw (290 Mt/year) and forest residues (320 Mt/year) followed over a certain distance by the organic fraction of municipal solid waste (OFMSW; 90 Mt/year). Other feedstocks are pruning residues (15 Mt/year), agro-industrial food process waste (14 Mt/year), roadside vegetation (6.3 Mt/year) and biomass from urban green areas (2.4 Mt/year) [3].

Many lignocellulosic feedstocks have a rather high water content. Water evaporation consumes energy, which is a burden for economic and ecological balance. The high water content makes transport more expensive and poses an obstacle to valorization by classical thermochemical processes such as pyrolysis or gasification. An alternative is the processing of the biomass in the presence of water, as is performed in hydrothermal carbonization (HTC) [4,5]. This process mainly induces a chemical dehydration of polysaccharides and degrades the plant structure. The dehydration implies a concentration of the carbon content and an increase in the hydrophobicity of the material. These changes facilitate the separation of water by mechanical filtration, which is more energy efficient than water evaporation. The obtained carbonaceous solid product is called hydrochar.

Depending on the composition of the starting material of the HTC process, the hydrochar can be valorized in multiple applications. When employing refined (purified) biomass such as glucose, materials for energy storage can be produced when hydrochar is used for the production of electrodes in rechargeable batteries or supercapacitors [6]. Valorization as activated carbon is always an option for carbonaceous materials [7]. Pruning materials can be converted into a renewable solid fuel [8]. Application is still limited when heterogeneous waste materials such as the OFMSW are employed for HTC with variable compositions and involving inaccurately sorted contents [9]. In this case, hydrochar has been made suitable for its use as soil amendment material by the elimination of phytotoxic properties [10,11,12,13,14]. However, the economic revenues are rather low.

Hydrochar has been used as a filler to reinforce polymers. For instance, hydrochar was produced starting from rice husk and up to 20 wt% was incorporated into a polylactic acid polymer. Thereby, the tensile modulus of the composite material was increased by more than 60% compared to the neat polylactide [15]. A similar result has been reported for the reinforcement of poly(butylene adipate-co-terephthalate) (PBAT) with hydrochar produced from wood chips: the elastic modulus was almost doubled by incorporating 20 wt% of hydrochar produced under mild conditions (180 °C) [16]. In another study, carbon black was substituted partially when used as filler for rubber. It has been found by vulcanization rheology measurements that hydrochar influences vulcanization chemistry [17]. Again, up to approximately 20 wt% of hydrochar induced positive effects and this route might be a valorization pathway for hardwood waste. From these three examples, it can be seen that hydrochar is a promising material for crosslinking with polymers. However, the incorporation of pristine hydrochar, i.e., without further thermal refinement, is limited to amounts up to approximately 20 wt%.

Refined biomass has been incorporated into phenolic resins to produce composite materials. Phenolic resins were the first fossil-derived polymers invented more than a century ago and were commercialized under the brand name Bakelite [18,19]. In recent years, phenol has been partially substituted by biomass components such as lignin [20,21]. This replacement provides a renewable-based resin material and is a clear step forward from an ecological point of view. However, the relatively high price of lignin, caused by its separation process from lignocellulosic biomass, is a commercial drawback [22].

This issue can be overcome by employing refined hydrochar obtained from biomass without commercial value or with a negative one (saving disposal costs), such as, for instance, rice straw, orange peel waste or the OFMSW. Herein, we will show that the appropriate adaption processes, namely a second thermal treatment and suitable curing conditions, make these residues suitable for the new value chain (Figure 1). The final resin composite material will be commercialized as panels for the construction and furniture sectors. Panel models (Figure 1) are prepared and characterized by means of their tensile strength and screw withdrawal resistance, following standards established for particleboards and fiberboards. The first results are promising and it is demonstrated that suitable materials are produced.

2. Materials and Methods

2.1. Production of Pristine Hydrochar in an Industrial Reactor

For the production of pristine hydrochar in an industrial pilot reactor from OFMSW in 2016, see ref. [14], and from orange peel waste, see ref. [23].

Residual biomass, i.e., the organic fraction of municipal solid waste (OFMSW), was supplied by a municipality in the north of Spain, from separate collection, and contained minor amounts of impurities, including biodegradable plastic bags, glass and plastic bottles. A total of 12.6 t of this biomass, with a humidity of 75% (3180 kg of dry mass), was fed continuously into the reactor for 9.25 h. The temperature was maintained between 205 and 215 °C under an autogenous pressure of 18 to 20 bar, resulting in an approximate residence time of 4 h. Solid materials with high inorganic content, such as stones, sand, glass and metals, were separated from the hydrochar in a post-treatment by physical means (density). After removing the process water by filtration, hydrochar with approximately 50% humidity was dried thermally and pelletized. A total of 1563 kg (49.1% mass yield based on dry input) was obtained, together with 181 kg of other solid materials (5.7% yield). On a dry basis, the pellets contained 14.6% ash, 19.6% fixed carbon and 65.8% volatile matter. Ash, fixed carbon and volatile contents were determined as described below. On average, the carbon content was 65%, and on a dry and ash-free base, the hydrogen content and nitrogen content were 7.3% and 2.6%, respectively (Table S1). Sulphur content was below the detection limit of the elemental analysis CHNS.

Rice straw was received from the Albufera area, Valencia, Spain. A total of 3009 kg of this biomass with a humidity of 33% (2031 kg of dry mass) was fed continuously to the reactor and the temperature was maintained between 205 and 215 °C, which generated an autogenous pressure of 18 to 20 bar. Solid material with a high content of inorganic material (SiO₂) was separated from the hydrochar by a post-treatment based on physical means (density). After removing the process water by filtration, hydrochar with approximately 50% humidity was dried thermally and pelletized. A total of 1010 kg (49.7% mass yield based on dry input) was obtained, together with 95.4 kg of other solid materials (4.7% yield). On a dry base, the pellets contained 17.5% of ash, 22.0% of fixed carbon and a remaining 59.9% of volatile matter. Ash content, fixed carbon content and volatiles were determined as described above. On average, the carbon content was 60%, and on a dry and ash-free base, hydrogen content and nitrogen content were 6.3% and 1.6%, respectively. Sulfur content was below the detection limit of the elemental analysis CHNS.

2.2. Feedstock Composition

Orange peels are not a typical lignocellulosic biomass as they contain a high water-soluble fraction [24]. Hence, glucose, fructose and sucrose account for 46% of the dry matter. Insoluble polysaccharides represent 20% of the dry matter, consisting primarily of cellulose (14%) and hemicellulose (6%). The non-saccharide component of lignocellulose, i.e., the lignin, is only present in marginal amounts of below 1 wt% of the dry matter. The ash content is low with values below 3 wt%. In addition, orange peels contain a significant amount of proteins (7%) and of pectin (19%).

Rice straw consists of cellulose (40–46%), hemicellulose (25–28%) and lignin (4–10%). The inorganic part accounts for 15 wt%, determined as ash content. The latter consists mainly of 68% SiO₂ and 12% K₂O [25].

2.3. Thermal Post-Treatment of Hydrochar

The thermal treatment of hydrochar was carried out in a vertical, tubular quartz reactor (Darío Lorusso, Cuarzo & Vidrio, Hontoria, Segovia, Spain, Figure S8a) at 600 °C for one hour, placing the thermocouple at the outer wall of the reactor within the heating mantle (Watlow, Ceramic fiber Heater, VC402A06A-0000R, Madrid, Spain; Figure S8b). Hydrochar pellets (50–100 g) were placed in the reactor and heated with a heating rate of 6 K/min to the desired temperature, which was maintained for one hour. During the treatment, a nitrogen flow of approximately 20 mL/min was applied. Liquids were condensed at the outlet under ambient conditions. After cooling down to room temperature, the hydrochar was recovered. Before employing the hydrochar for the resin preparation, it was milled in a Pulverisette 11 (Fritsch) (FRITSCH, Idar-Oberstein, Germany) for one minute at 7000 rpm. The particle size distribution was determined by sieving: 53% < 0.2 mm and 40% 0.2–0.6 mm.

2.4. Characterization of Hydrochar and Thermally Treated (Refined) Hydrochar

The ash content was determined as described in UNE 32004:1984 [26]. The sample was crushed in a mortar to a particle size of <0.2 mm and dried (105 °C). Approximately 1 g (weighed exactly) was placed in a crucible and heated in a muffle furnace from room temperature to 815 °C at a heating rate of 3 K·min⁻¹. The temperature (815 °C) was maintained for one hour. The residue obtained was the ash of the sample. The ash content was calculated by the following formula:

Ash content/% = Mass/g (ash)/Mass/g (dry sample) × 100

The volatile content was measured following the standard UNE 32019:1984 [27], namely by heating a crushed, sieved and dried (105 °C) sample to 900 °C in a closed vessel for seven minutes. The volatile content was eliminated by the procedure and was calculated by the following formula:

Volatile content/% = (Mass/g (dry sample) − Mass/g (remaining))/Mass/g (dry sample) × 100

The fixed carbon content was calculated from the volatile content and the ash content by the following formula:

Fixed carbon/% = 100 − volatile content/% − ash content/%

For elemental analysis (CHNS), the samples were crushed in a mortar to a particle size < 0.2 mm. The samples were analyzed on a Thermo Scientific Flash 2000 Organic Element Analyzer apparatus (Thermo Fisher Scientific, Dreieich, Germany). Values are stated on a dry basis and, in addition, the carbon value on a dry-and-ash-free (daf) basis (Table S1). The composition of the ashes (Table S2), involving the elements Na, K, Mg, Al, Fe, Si and P, was determined by inductively coupled plasma optical emission spectroscopy (ICP OES). Therefore, a sample (20–30 mg) was disaggregated in a HNO₃/HF/HCl mixture (1:1:3) and the solution analyzed on a Thermo Scientific iCAP PRO apparatus (Thermo Fisher Scientific, Dreieich, Germany). The calcium content (CaO and CaSO₄) was measured by X-ray fluorescence of the solid ashes on a PANalytical MiniPal 4 apparatus (Malvern Panalytical, Kassel, Germany).

The textural analysis was performed by measuring CO₂ isotherms at 0 °C on volumetric Micromeritics ASAP 2020 and 2420 devices (BSD instruments, Großröhrsdorf, Germany) after activation at 300 °C and under vacuum for the HC-x_600 samples and after activation at 150 °C for the hydrochar samples HC-x without any thermal post-treatment. The Dubinin–Astakhov (DA) method was used to calculate the surface area and the pore volume [28].

FT-IR spectra were obtained using the crushed powder on a JASCO FT/IR-4700 apparatus (JASCO Applied Sciences, Schwentinental, Germany) equipped with an ATR PRO ONE accessory (Figures S2 and S3). The ¹³C MAS NMR spectra were measured on a Bruker AVIII HD 400WB apparatus (Bruker, Bremen, Germany). Characterization of the surface morphology of the samples by means of scanning electron microscopy (SEM) was carried out on a ThermoFisher Scientific Quattro S field emission environmental scanning electron microscope (ESEM) (Thermo Fisher Scientific, Dreieich, Germany) operated by the Helmholtz-Zentrum Hereon and Jülich Center for Neutron Science (JCNS) at MLZ (Figures S4–S7). The SEM micrographs were taken at a working distance of 10 mm with a probe current of 7 pA and an acceleration voltage of 5 kV using an Everhart–Thornley detector (ETD) collecting surface near-secondary electrons to obtain high topography information. Energy-dispersive X-ray spectroscopy (EDS) was performed with a ThermoFisher EDS UltraDry Si-drift detector (Thermo Fisher Scientific, Dreieich, Germany) (60 mm²) at a take off angle of 40° with an energy resolution of 127 eV Mn Kα. The acceleration voltage was set to 15 kV and probe current to 2 pA, resulting in a count rate of 2–10 kcps and a detector dead time of 5%.

For the elemental mapping, the dwell time and repetitions were adjusted depending on the image size ranging from 2.3 × 1.6 mm² (overview scans) to 150 × 97 µm² (single particles), leading to a total acquisition time of 30 min and 10 min, respectively.

X-ray diffraction (XRD) patterns of the thermally treated hydrochar were obtained with synchrotron radiation at the HEMS beamline of PETRA III at the DESY side station (Figure S9) [29] by using a photon energy of 87.1 keV corresponding to a wavelength of 0.01423 nm. A PerkinElmer XRD1621 (PerkinElmer, Rodgau, Germany) image detector with a pixel size of 200 × 200 mm² and a resolution of 2048 × 2048 pixels was used to record the XRD patterns. The sample-to-detector distance was 1674.6 mm, allowing us to record complete Debye–Scherrer rings. The calibration was performed by using a LaB₆ standard sample (NIST standard reference material SRM-660a) for calibration. Fit2D (V17.052) [30] software was used to treat the 2D patterns from synchrotron measurements and the Full-Prof (V 5.20) software [31] was applied for Rietveld refinement [32] of the acquired synchrotron diffraction patterns.

2.5. Preparation of the Composite Panel

In a round-bottom flask equipped with a reflux condenser, the milled carbon powder (20 g) obtained from the two-step procedure of HTC and thermal post-treatment was placed together with phenol (2.22 g; 10 wt% of phenol with respect to the total carbon powder and phenol), ethanol (26.6 mL) and sodium hydroxide (1.20 g). The mixture was stirred and heated to 80 °C. Formaldehyde (37% in water, 30 mL) was added drop-wise. After 4 h the mixture was allowed to cool down and the solvent was removed at the rotavapor. A viscous paste was obtained which was placed onto an aluminum pressing mold (Figure S10, left-hand side). The mold was placed between two plates of a press and heated to 130 °C (Figure S11). A minimum pressure of 10 MPa was applied for 15 min. Afterwards, the pressure was released and the produced polymer unmolded (Figure S10, right-hand side).

Material properties were determined following the standards for particleboards and fiberboards (Figure 2): UNE-EN 319 [33] for the tensile strength perpendicular to the plane of the board and UNE-EN 320 [34] for the resistance to axial withdrawal of screws.

3. Results

3.1. Hydrothermal Carbonization of Wet Lignocellulosic Biomass

Wet lignocellulosic biomass is not suitable for the direct production of phenolic resin composites, and its properties have to be adapted and the required functionalities have to be generated. This is achieved by the first two steps, namely hydrothermal carbonization (HTC) and subsequent thermal treatment in the absence of oxygen. The second process might be characterized as a (dry) pyrolytic treatment of the hydrochar, which is the solid product of the first transformation.

Three different biomass resources have been selected: the OFMSW, rice straw and orange peel waste. The HTC was carried out on industrial pilot scale, and the results are summarized in

Table 1. HTC trials at an industrial pilot plant at 205 to 215 °C and a residence time of 4 h.

Entry	Sample Name	Biomass Feed	Input Wet Biomass [t]	Water Content [%]	Dry Material [t]	Output Hydrochar [t]	Mass Yield [a] [%]	Average Ash Content [b] [%]
1	HC-1a [c]	OFMSW	9.24	75	2.31	1.05	45.5 [d]	13.3
2	HC-1b	OFMSW	12.6	75	3.18	1.56	49.1 [e]	14.6
3	HC-2	rice straw	3.01	33	2.03	1.01	49.7 [f]	17.5
4	HC-3 [g]	orange peel waste	12.3	84	1.96	0.723	36.9	6.6

^[a] Mass yield is defined as the mass ratio of dry output hydrochar to dry input material and expressed as a percentage. ^[b] Ash content of the hydrochar measured following UNE 32004:1984. ^[c] From reference [14]. ^[d] Additional solid material with higher ash content was obtained at a 12% yield. ^[e] Additional solid material with higher ash content was obtained at a 5.7% yield; in addition to the solid and the aqueous phase, an organic liquid was obtained which counted for 1.3 wt% of the dry matter of the starting material. ^[f] In addition, 4.7% was obtained and separated (by density) after the process with high inorganic content. ^[g] From reference [23].

. The mass yield related to the dry content of the biomass was almost 50%. Therefore, it has to be taken into account that approximately 20 wt% of the dry mass was converted into water by chemical dehydration and 5 wt% into carbon dioxide. Hence, the carbon yield was much higher than the mass yield. Only for orange peel waste was the mass yield significantly lower at only 37%. The lower HTC mass yield for this feedstock is a general feature and has been reported in several cases [35,36]. The reason for this might be the elevated non-cellulosic content such as proteins (7%) and pectin components (19%) [24]. Both have a lower mass yield in the HTC process [37,38,39].

As a general rule, for the mass yield, it can be stated that a higher polysaccharide content decreases the mass yield as this part is dehydrated. In contrast, the lignin content is hardly modified, and its mass yield is nearly quantitative. Water-soluble organic matter is lost in the process water, resulting in mass yield deterioration. The inorganic content is incorporated almost entirely into the hydrochar and remains unchanged under the reaction conditions. An exception is the alkali content, especially the potassium, which is dissolved into the process water. Changes in the ash (inorganic) content of the hydrochar can be introduced on an industrial scale by mechanical separation of particles of higher density, such as stones or sand (cf. Table 1, footnotes [d] and [e]).

The OFMSW has been processed twice in different years. The hydrochar from the first trial has been used in earlier studies on the elimination of the phytotoxicity [14] and on phosphorus recovery [40]. When the results of both HTC trials are compared, it can be stated that the characteristics of the hydrochar were quite similar: both carbonization products were high-ash materials with 13.3 wt% and 14.6 wt% for hydrochar HC-1a and HC-1b, respectively (Table 1, entries 1 and 2), with similar ash compositions (Table S2, entries 1 and 2). Especially in these trials, a significant amount of improper (non-carbonic) materials such as stones, sand, glass, cans, etc., were separated from the hydrochar during the HTC pilot trials by physical means. The second hydrochar sample, HC-2, obtained from rice straw with a mass yield of 49.7%, also possessed a high ash content of almost 18%, even higher than HC-1 (Table 1, entry 3). Hydrochar HC-3 derived from orange peel waste was taken from a former trial [23]. It possessed a relatively low ash content of only 6.6 wt% (Table 1, entry 4).

3.2. Thermal Treatment as Refinement Process for Hydrochar

The second carbon concentration in the carbonaceous material and further structural transformations were induced by a pyrolysis-like thermal treatment of the hydrochar at 600 °C. A downflow nitrogen atmosphere avoided the presence of air, and the samples HC-1_600, HC-2_600 and HC-3_600 were produced. In

Table 2. List of samples of renewable resin precursors prepared.

Sample	Starting Material	Yield [%]	Ash Content [a] [% dry]	Volatiles [b] [% dry]	Fixed Carbon [b] [% dry]	Pore Volume [c] [cm3/g]	Surface Area [c] [m2/g]	Molar Ratio [d]
								O/C	H/C
HC-1	OFMSW	45.5–49.1	13.3–14.6	65.8–68.4	18.3–19.6	n.d. [e]	n.d. [e]	0.26 [f]	1.44 [f]
HC-1_600	HC-1	38.1 [g]	34.3	13.5	52.2	58.9	300	0.08	0.35
HC-2	Rice straw	49.7	17.5	59.9	22.0	n.d. [e]	n.d. [e]	0.40	1.27
HC-2_600	HC-2	47.1	31.2	17.1	51.7	n.d. [e]	n.d. [e]	0.14	0.45
HC-3	Orange peel waste	36.9	6.6	73.6	19.8	17.3	87	0.39	1.09
HC-3_600	HC-3	46.2	14.1	20.0	65.9	68.0	342	0.08	0.30

^[a] Determined following standard UNE 32004:1984. ^[b] Determined following standard UNE 32019:1984. ^[c] Measured with carbon dioxide at 273 K. ^[d] Detailed information on the CHN analysis is listed in Table S1. ^[e] Not determined. ^[f] Value of the HC-1a sample. ^[g] Up to 30 wt% of a liquid was condensed during the thermal treatment.

, it can be seen that the mass yield of the pyrolytic treatment varied from 35 wt% up to 50 wt%, depending on the initial biomass employed for the hydrochar production. In the case of a lower yield, an organic liquid was co-produced during the thermal treatment (Table 2, footnote f). In general, ash content was increased by this treatment, and approximately doubled (Table 2). Volatile organic content (considered as such for solid fuels determined at 900 °C following standard UNE 32019:1984) dropped below 20% and the fixed carbon content increased to at least 50% of the dry matter (Table 2).

The two-step process for the transformation of the biomass into the resin precursor can be visualized in the van-Krevelen diagram (Figure 3). The positions of polysaccharides, such as cellulose or hemicellulose, are marked in the upper-left quadrant (0.83 and 1.67 for O/C and H/C molar ratios, respectively, for (C₆H₁₀O₅)_n). During the HTC process, the biomass is mainly dehydrated and to a small extent decarboxylated [23,41]. The hydrochar is located at the center of the diagram with O/C and H/C approximate molar ratios of 0.4 and 1.2, respectively, with the exception of HC-1, which involves a higher aliphatic content. With the pyrolytic treatment, the lower-left corner (0.08 and 0.3 for O/C and H/C molar ratios, respectively) is reached, indicating a high carbonaceous character of the material with a high energy density. Again, the main chemical transformation is dehydration. In addition, the small contribution of a dehydrogenation-type transformation is proposed. The latter may occur in cyclohexane rings which are transformed into aromatic rings.

A slightly different global transformation method is observed when starting from the OFMSW. The hydrochar HC-1 position is deviated from the point 0.4/1.2 to the point 0.25/1.4 (O/C and H/C molar ratios). This displaced localization can be rationalized by the adsorption of significant fatty acid content [14]. Fatty acids are abundant in the OFMSW and have a high aliphatic character with a low O/C ratio and a H/C of 2. Hence, the location of the data point for HC-1 has to be interpreted as an average value of the solid material and the adsorbate. During the heat treatment of HC-1 at temperatures up to 300 °C, the fatty acids are desorbed [14] and the “purified” hydrochar pyrolyzed when raising the temperature further, in the same way as in the other two cases.

The change in the chemical structure during the pyrolytic treatment is monitored by IR and ¹³C MAS NMR spectroscopy. The IR spectra document a de-functionalization of the material: typical C–H, C=O and C–O bands disappear after the treatment (Figure S2) [42]. Interestingly, the IR spectra are different from the ones of lignin, which show well-shaped bands for the aliphatic C–H bands (below 3000 cm⁻¹) and C-O and C-C single and double bonds (1000–1700 cm⁻¹; Figure S3) [42].

The ¹³C MAS NMR spectra support the same transformation pattern (Figure 4): the carbonyl signal (215–190 ppm) disappears as well as signals from the C–O region (90–60 ppm) and from the aliphatic carbon region (50–10 ppm). The remaining (broad) signal (140–110 ppm) is consistent with aromatic moieties. Interestingly, all thermally treated hydrochars give very similar ¹³C MAS NMR spectra (Figure 4). The transformation of the functional groups can also be monitored by FT-IR (Figure S2). Hydrochar samples involve narrow signals in the C–H, carbonyl or C–O regions. In contrast, pyrolyzed hydrochar samples show only broad hills or resemble flat lines. Interestingly, the FT-IR spectra of pyrolyzed hydrochar are quite different from those of organosolv lignin or Kraft lignin (Figure S3).

For the HC-1_600 and HC-2_600 samples, ashes accounted for one third of the material, and only the HC-3_600 was below 20 wt% (Table 2). Elemental composition (bulk; measured on the ashes by ICP-OES after acid digestion) showed a high calcium and phosphorus content for the HC-1-derived materials and a high silicon content for the HC-2-derived carbons (Table S2). When having a closer look at the elemental distribution on the surface by means of energy-dispersive X-ray spectroscopy (EDS; Table S3), it was observed that the locations of calcium and phosphorus were matching, indicating the presence of calcium-phosphate-type domains for the HC-1_600 sample (Figure S4). By XRD obtained from synchrotron radiation (Figure S9), a Ca₁₀(PO₄)₆(OH)₂, hydroxyapatite, crystallographic phase accounted for almost half of the total amount (Table S4).

When comparing the silicon distribution for the HC-2_600 (60 wt%; Table S2) and the HC-3_600 (22 wt%, Table S2), a different allocation was detected: in the case of the rice-straw-derived material, HC-2_600, small particles (often below µm scale) were found with a high dispersion over the carbon material (Figure S5), whereas in HC-3_600, bigger particles were detected with a diameter of up to 50 µm (Figure S6).

3.3. Preparation of the Composite Panels

The carbon material produced by the two-step procedure was employed for the synthesis of (resole-type) phenolic resin composites. Thereby, the aim was to substitute 90 wt% of fossil-derived phenol and to keep only 10 wt% to produce a resin composite from 90% renewable material. For the synthesis, refined hydrochar with phenol (90:10 wt%/wt%) was used together with formaldehyde in ethanol under basic conditions. Thereby, formaldehyde was considered a renewable chemical, as it can be obtained from a renewable source such as biomass-derived syngas [43,44] or renewable methanol [45,46].

For curing the resin precursor, the mixture of refined hydrochar with the synthesized adhesive was filled into molds (Figure S10) and pressurized with 10 MPa at 130 °C for 15 min (Figure S11). The resulting panel model is shown in Figure 1 and Figure S10. Specific material properties were determined following the standards for particleboards and fiberboards. Tensile strength was measured corresponding to UNE-EN 319:1994 and values of 1.1 to 1.6 MPa were obtained (

Table 3. Determination of material characteristics applying standards for particleboards and fiberboards: tensile strength and resistance to axial withdrawal of screws.

Entry	Starting Material	Density [kg/m3]	Tensile Strength [a] UNE-EN 319 [MPa]	Screw Withdrawal Resistance [b] UNE-EN 320 [N]
1	HC-1_600	1300	1.55	2590
2	HC-2_600	1130	1.13	not determined
3	HC-3_600	1203	1.58	2030

^[a] Particleboards and fiberboards—Determination of tensile strength perpendicular to the plane of the board; 0.4–1.5 MPa for commercial boards, cf. Table S5. ^[b] Particle boards and fiberboards—Determination of resistance to axial withdrawal of screws; 700–1300 N for commercial boards; cf. Table S5.

). A deformation was not observed and the failure was brittle (Figure 5). The screw withdrawal resistance (SWR) was higher at 2000 N, following UNE-EN 320:1994 (Table 3). Both values were in the range of commercial products in the particle- and fiberboard sector, i.e., 0.4–1.5 MPa and 700–1300 N for tensile strength and SWR, respectively (Table S5).

4. Discussion

The present study is based on the working hypothesis that lignocellulosic biomass can be made suitable for its incorporation into phenolic resins and, thereby, the lignin content is not relevant. Suitable aromatic domains with a similar reactivity as phenols or polyphenols are created by a two-step process involving HTC and pyrolytic treatment. Chemical transformation of the organic content supports this hypothesis as can be seen below.

4.1. Feedstock Properties and Selection Criteria

Three different biomass resources have been selected, guided by the following criteria: availability, composition and economic benefits. Straw, as already stated before, is one of the most abundant resources and involves a lignocellulosic composition that varies only within a small range. Therefore, it can be predicted that the properties of the final material will be reproducible independently from regions or changing climate conditions during its growth from year to year. Hence, election criteria are availability and composition in this case.

The interest in the next feedstock, orange peel waste (OPW), is mainly due to its composition; it provides substantial proof for the working hypothesis, although its use on an industrial scale is realistic. Considerable amounts (30 Mt/year worldwide; 3 Mt/year in Europe) [47] of this feedstock are available and it is generated at hot spots, namely juice producers, avoiding transportation costs. Its composition makes it singular: it is a lignocellulosic material with low lignin content (<1 wt%) [24,48]. In addition, in the same way as straw, its composition is constant with only small variations. For this reason, the resin precursor produced from this feedstock will have a low lignin-derived content. A successful production of the resin will demonstrate the insignificance of the lignin content of the residual biomass for the resin synthesis.

The third feedstock is the organic fraction of municipal solid waste (OFMSW), which is interesting for all three criteria: availability, composition and economic benefits. With respect to the availability, its production volume is significant in all European regions [49,50]. It is a resource that is already concentrated in municipal areas and requires only a low logistic effort. From the composition point of view, it represents the biggest challenge. It involves a high lignocellulosic part from vegetable waste. However, this matter is mixed with other organic biogenic leftovers such as proteins, and it contains inorganics such as eggshells or bones or inorganic waste that does not belong in the organic waste. If this resource can be converted into a resin composite, then it can be supposed that the procedure will accomplish its mission for a broad range of biogenic residues with a certain lignocellulosic content. In case of success, the economic balance of the process will be supplemented by the negative price of this resource, since current management of the OFMSW is quite expensive and does not fulfill European sustainability standards in many cases.

The OFMSW is very heterogeneous and therefore difficult to simulate at a laboratory scale. Hence, it is decided to carry out the hydrothermal carbonization on a ton-scale in an industrial pilot plant and to refrain from providing a detailed characterization in terms of moisture content and composition [8]. Thus, this first step will produce a certain homogenization for the OFMSW feedstock. At a later stage of this study, a detailed characterization of the hydrochar and the refined hydrochar will provide insights into the inorganic content of the raw materials, enabling conclusions to be drawn regarding their composition. The other two raw materials are processed in the same way, i.e., on a ton-scale at the pilot plant facility. As a side effect and additional benefit, this approach also initiates the pathway toward industrial implementation of the value chain.

4.2. Transformation of the Organic Content

Lignocellulose consists of three different biopolymers, namely cellulose, hemicellulose and lignin, all of which are highly oxy-functionalized. Basic functional groups are hydroxy groups and carbonyl groups, which build different functional groups such as ketones, aldehydes, carboxylic acids, aliphatic alcohols, phenols, cyclic and acyclic ethers, esters and lactones. Only lignin would have suitable functional groups, namely phenolic moieties, which are suitable for undergoing phenolic resin chemistry. However, in general, lignin is embedded into a dense three-dimensional structure in lignocellulose and the potential connection sites are not accessible. Therefore, a chemical transformation is required to relax the structural constraints and to generate additional suitable functional groups.

The two-step transformation process of HTC and thermal treatment eliminates the oxygen content to a great extent, as can be easily seen in the van-Krevelen diagram (Figure 3). This occurs mainly by aldol-type condensations of carbonyl groups, forming carbon–carbon bonds and eliminating water molecules. The dehydration is evidenced in the van-Krevelen diagram as displacement along a straight line with a slope of two, towards the origin, indicated by the blue arrow (Figure 3). All hydrochars produced involve a molar O/C ratio of 0.4 or lower. These ratios indicate an acceptable transformation of the lignocellulosic biomass into hydrochar, although under more severe conditions, O/C molar ratios of 0.3 can be achieved. However, as can be seen in the following, an O/C molar ratio of 0.4 is sufficient for the subsequent transformation (crosslinking to a phenolic resin after the second thermal treatment). The final O/C ratio is relatively independent from the initial composition of the lignocellulose and can be achieved with any lignocellulosic biomass.

In the same way, the elimination of oxygen functionalities is confirmed by the ¹³C MAS NMR (Figure 3) during the pyrolytic treatment. Hence, pristine hydrochar involves carbonyl groups and hydroxy groups, evidenced by signals around 200 ppm and in the range from 50 to 90 ppm, whereas the refined hydrochar does not involve these signals anymore. Instead, a broad singlet centered at 125 ppm is observed, characteristic of aromatic moieties. Due to the broad shape of the signal, oxygen-substituted carbon atoms cannot be identified. However, taking into account the elemental analysis indicating an oxygen content of 5 to 10 wt%, it can be concluded that phenol moieties must be present. On an atomic scale, the spectra indicate that carbon atoms, which have an sp³ configuration in the biomass and bear one hydroxy group, are transformed and adopt an sp² configuration, forming plenty of double bonds and aromatic units. The transformation of the functional groups is also supported by FT-IR spectroscopy: narrow bands are transformed into broad hills (Figure S2). Furthermore, it can be concluded from the FT-IR that the generated material from the two-step process is different from typical lignin samples obtained by the organosolv process or the Kraft process; the obtained spectra are different (Figure S3).

In summary, it can be stated that the two-step procedure consisting of HTC and a pyrolytic treatment of the hydrochar product converts polysaccharides into carbon materials with a low oxygen content and low hydrogen content; carbon atom configurations change from an sp³ configuration to an sp² one. By NMR spectroscopy, the transformation and homogenization of the organic part is best documented, resulting in an aromatic structure. The latter may be suitable for incorporation into phenolic resins by interlinkage to phenolic resin domains.

4.3. Ash Compositions

The ash content or inorganic content is a significant part of the final resin precursor: it ranges from 15 wt% up to one third of the “carbon” material (Table 2). In addition, it behaves like a footprint for the original type of biomass employed for the hydrochar production. During the two-step process, the inorganic content of the original biomass is accumulated due to the dehydration of the organic part and the elimination of the volatile organic part. In contrast, only a minor part of the initial inorganic material is dissolved, mainly involving alkali cations such as sodium or potassium. Other elements, such as mixed oxides of silicon, aluminum, iron, magnesium and calcium, are insoluble under the process conditions of HTC and non-volatile under the pyrolysis conditions.

Having a closer look into the compositions of the ashes, each carbon material turns out to be unique. As stated before for HC-1_600 and HC-2_600, a hydroxyapatite crystallographic phase accounts for almost half of the inorganic part (Table S4). It is a form of calcium phosphate and is the main inorganic mineral component of teeth and bones [50]. This finding is consistent with the origin of the carbon material, the OFMSW, which probably contains some bones from food waste. Ash composition shows a high silica content for hydrochar HC-2 (>70 wt%, Table S2, entry 5). This is to be expected as rice straw is considered a biobased silica source owing to its high silica content (ash content of 15 wt% with 68% SiO₂) [25,51,52,53]. Also, the distribution of silicon and oxygen in small domains as observed by SEM/EDS (Figure S5) is consistent with the rice straw origin of the material.

Silicon oxide was also found in the orange-peel-waste-derived material (HC-3_600). However, the distribution was quite different to that observed for the rice-straw-derived hydrochar HC-2_600: the oxide was in bigger particles of a size of up to 0.05 mm (Figure S6). The shape was consistent with contamination of the orange peel waste with sand or dust. The ash content opened up certain variability for the incorporation of further elements into the carbon materials and, therewith, into the renewable resin composite.

For a final evaluation of the inorganic content, two conclusions can be drawn. Most importantly, the ash content, even up to one-third of the material, has no negative effect on the final objects; tensile strength or screw withdrawal resistance are in the range of those produced from low-ash material (Table 3). On the other hand, the inorganic part of biomass can be used to introduce certain elements such as calcium or phosphorus or to homogenously distribute silicon oxide domains into the phenolic resin without any effort. This might be interesting for special applications.

4.4. Production of Composite Panels

The carbon material produced by the two-step procedure is converted successfully into a small panel using 10 wt% of phenol together with formaldehyde. The preparation process can be considered a suspension of the carbon material, i.e., of the refined hydrochar, into a phenol–formaldehyde resin precursor which is then cured. It is supposed that the synthesis of the resin precursor in the presence of the carbon material permits the modification of the surface of the carbon material by phenol and formaldehyde reacting with suitable moieties such as phenolic groups. These surface “extensions” are then incorporated into the phenolic resin providing a strong chemical interconnection. A chemical crosslinking between the embedded carbon particles and the resin is substantiated by the excellent tensile strength and screw withdrawal resistance.

However, the preparation of the resin in the presence of the carbon material could be considered a bottleneck for upscaling the process. Given that the carbon material accounts for 90% of the mass, it is estimated that preparing the resin separately would reduce the required volume by a factor of 10. This issue must be given greater attention in future.

On the other hand, if the chemical bonds are formed and this is also the case for lignin-poor materials, then suitable groups are present in the carbon material. This is in line with the hypothesis of the transformation of the polysaccharide part into subunits suitable for phenolic resin chemistry and with the interpretation of the results of the solid-state NMR. For lignin-rich materials, this biopolymer part can add further strength to the structure as composite materials from lignin and phenolic resins are well known in the literature [20,21].

During the curing process, elevated temperature and pressure are necessary to achieve the desired material properties of the composite panel. Heat is necessary for providing the required activation energy for the chemical formation of the phenolic resin. In addition, the composite model of embedded carbon particles explains the need for pressure during curing: the carbon interparticle space has to be minimized so that it can be filled with the smallest possible amount of phenolic resin. An artistic representation of the model is depicted in Figure 6.

In summary, the composite model of carbon particles embedded within a phenolic resin network is fully consistent with the observations from NMR spectroscopy and the mechanical properties of the composite panels, provided that chemical crosslinking exists between the two components.

5. Conclusions

Lignocellulosic resources with high moisture content were transformed through a three-step process involving hydrothermal carbonization (HTC), thermal treatment and dispersing the carbon material in a phenolic resin. The first two steps primarily converted the polysaccharide parts of the lignocellulose, i.e., hemicellulose and cellulose, into aromatic, oxygen-containing moieties. These new-formed functionalities exhibited similar reactivity as lignin, the third major component of lignocellulose, allowing the material to be surface-functionalized and subsequently crosslinked with a phenolic resin matrix. Applying pressure during the curing process minimized the interparticle space and enhanced the mechanical properties of the resulting panels. Small-scale panel models produced in this way demonstrated mechanical properties that were competitive with commercial particle- and fiberboards. These panels involve a mass ratio of refined hydrochar to phenol of 9:1. Herewith, the incorporation is far larger than in the case when hydrochar (without refinement) is used as a filler for polymers (when the optimum is approximately 20 wt%).

The novel procedure presented here provides a promising route toward real-world applications of innovative composite materials derived from residual lignocellulosic biomass. It provides an alternative to more classical valorization routes such as incineration or composting. Life cycle assessment is mandatory for evaluating the real sustainability of the overall process and will be carried out in the near future.

6. Patents

This work forms the basis of a Spanish patent application with reference number P202430723.