Elemental analyses to determine the percentages of carbon (C), hydrogen (H), nitrogen (N), and sulfur (S) of particleboard waste, uncoated particleboard, melamine coated sheets, fir wood, biochar, and activated biochars were performed by a CHNS Analyzer (Perkin Elmer 2400 Series II, Waltham, Massachusetts, USA). These analyses were carried out to evaluate the efficiency of the nitrogen removal processes. The level of nitrogen devolatilization was deduced from the results of the elemental analyses and their degradation level after thermal treatment according to Eq. (1):(1) Nremoved=Ni ∗ mi−Nf∗ mf Ni ∗ mi∗100 where N_removed: Mass percentage of nitrogen removed during each thermal treatment test; N_i: Nitrogen content in the sample before thermal treatment; m_i: Initial mass of each sample before thermal treatment; N_f: Percentage of nitrogen in each sample after thermal treatment; m_f: Residual mass of each sample after thermal treatment.

The thermal stability and degradation behavior of particleboard samples were analyzed using a Jupiter Netzsch STA 449 F5 thermogravimetric analyzer (Exton, PA, USA). In each experiment, a fixed amount of sample was placed in a ceramic cup on a microbalance and the temperature was increased from 35°C to 450°C at 20 °C/min. Helium was maintained at 50 mL/min to keep the atmosphere inert. However, the pilot scale CarbonFX could not provide a completely inert medium and 5 mL/min of pure oxygen was thus added to achieve a system that was as close as possible to the CarbonFX process.

The gases released during the thermal treatment of particleboard at 450°C were transported from the TGA to the GC/MS (Agilent Technologies 789B, Santa Clara, California, USA), whereas the gases produced during activation were recovered with gas bags and manually injected with a syringe into the GC/MS system. The gases generated during the thermal treatment in the Netzsch analyzer were transported through an He carrier gas (1 mL/min) to the chromatography HP-5 type capillary column (30 m × 0.25 mm × 0.25 μm). The volume injected into the column was 1 μl at 100°C every minute, i.e., every 20°C increase. The identification of the gas compounds released during the TGA or during the activation in the in-house furnace was carried through the NIST02 spectra library. Blank GC/MS tests were performed before each test to ensure that the GC column was purged.

X-ray photoelectron spectroscopy (XPS) using a Kratos Axis Ultra DLD spectrometer (Manchester, UK) equipped with Al anode (hν = 1486.6 eV) and operated at 225 Watts was used to analyze the wood panel waste, biochar and activated biochars. The spectrum was recorded with a pass energy of 80 eV, and a step size of 1 eV. Atomic concentrations were obtained by the integration of each elemental peak. A scan from 0 to 1400 eV was acquired and the elemental composition of the surface of each sample was calculated using CasaXPS software.

The specific surface area (S_BET) and pore size distribution were performed using a high-resolution physical surface analyzer (Micromeritics ASAP 2460, Norcross, Georgia, USA). The biochar and activated biochars were degassed at 105°C and 250°C, respectively, under vacuum for 48 h. Nitrogen and carbon dioxide were used as adsorbates at −196°C and 0°C, respectively. The isotherms obtained from the N₂ adsorption tests were processed using the Brunauer-Emmett-Teller (BET) method to determine: (i) S_BET [45]; (ii) micropore volume (V_μ, cm³/g) according to the Dubinin-Radushkevich method [46]; (iii) total pore volume (V_t, cm³/g) calculated at the relative pressure of 0.97 from the amount of nitrogen adsorbed [47]; (iv) mesopore volume (V_m, cm³/g) calculated by the difference V_t−V_μ; and (v) pore size distribution by applying density functional theory to N₂ adsorption isotherms [48].

To study the thermal behavior of samples, the thermogravimetric (TG) curves and their derivatives (DTG) for particleboard and fir wood (control) are presented in Fig. 3. The TG analyses were performed using a temperature range between 35°C and 450°C at 20°C/min. The pyrolysis process can be divided into three stages for both samples. For the first stage, the mass loss occurred between 35°C and 200°C, i.e., 8% and 2% for wood and particleboard, respectively, due to the evaporation of water and volatile compounds. The second stage, between 200°C and 400°C, is where cellulose, hemicellulose, lignin, urea-formaldehyde (UF), and melamine-formaldehyde (MF) resins were thermally degraded. The mass loss at this stage was about 54% and 58% for wood and particleboard samples, respectively.

The particleboard reached its highest degradation rate of 14% at 356°C, while fir wood reached 19% at 366°C. This indicates that UF and MF resins accelerate the chemical reactions causing particle board degradation, which is in agreement with results reported by Feng et al. [6]. The third stage, between 400°C and 450°C, is when only lignin continues to pyrolyze. The TG curve also shows that the final mass of the pyrolyzed samples was about 29% and 32% for wood and particleboard, respectively. Moreno et al. [50] stated that the degradation temperature range for particleboard is close to that of wood, as particleboard contains approximately 90% wood. However, there is a slight difference in the appearance of a peak at 308°C due to the decomposition of urea-formaldehyde (UF) resin [51]. The results of the thermogravimetric analysis show that the highest degradation rate for particleboard is 356°C. Indeed, the pyrolysis process of particleboard from this study was performed using CarbonFX technology at 450°C for 2 s. The choice of such a temperature was to avoid any kind of tar formation during the production of the activated biochar in the in-house activation furnace.

Elemental analyses performed on wood panels and their components separately, and on biochar and activated biochars are summarized in Table 1. Specifically, nitrogen was present in all samples, but at a higher percentage compared to fir wood (>2.45% vs. 0.13%). The high nitrogen content is explained by the presence of urea-formaldehyde glue and melamine-formaldehyde coating as shown by Girods et al. [43]. This result is consistent with Feng et al. [20] who reported content of 7.4% nitrogen in the particleboard. This difference is probably due to the varying percentages and types of glue used in the panels. However, the nitrogen content present in BW, BWP and activated BWP decreased from 4.29% to 2.45% (BWP950). Indeed, the resins associated with wood degrade at a lower temperature range compared to fir wood, resulting in the removal of a greater amount of resin initially present in the wood panels. This statement is in agreement with other studies from the available literature [50,52,53].

After pyrolysis, the carbon content of the biochar and activated biochars increased and the hydrogen content decreased compared to their precursors. This is due to the release of volatiles during pyrolysis which results in the removal of non-carbon species and promotes carbon enrichment [54]. Thus, the activation step was responsible for an increased carbon content from 69.17% to 81.18%, while the percentage of hydrogen decreased from 3.23% (BWP) to 1.46% (BWP950). The results also revealed that the percentages of C and H slightly increased and decreased, respectively, with an increase in the activation temperature from 750°C to 950°C (78.99% and 1.85% vs. 81.18% and 1.46%). This is due to the progressive consumption of carbon and hydrogen atoms by the CO₂ gas during physical activation and by the simultaneous formation of oxygenated sites on the walls of the activated carbon [55]. Nitrogen was present in all activated carbon samples, ranging from 2.45% to 3.82%. The two-step thermochemical treatment, fast pyrolysis followed by physical activation in the presence of CO₂, allowed the activated biochars to acquire a stable amount of nitrogen content. According to the literature, activated biochars prepared from wood panels have an excellent adsorption capacity for acid gases and phenolic compounds due to their nitrogen-containing surface groups [19]. To evaluate the efficiency of the thermal treatment, the percentage of nitrogen removed during the pyrolysis of different samples was calculated from the results of the elemental analyses and the residual mass. The results showed that approximately 66% of nitrogen initially present in the different panel components was removed during the pyrolysis process. Furthermore, the efficiency of the process in terms of nitrogen removal does not appear to be influenced by the percentage of nitrogen initially present in the samples.

The percentages of nitrogen removed by the pyrolysis treatment in this study are in agreement with those reported in the literature [17–18,43,51,53]. Girods et al. [17] reported that an appropriate pyrolysis temperature (250°C–300°C) and residence time (8–15 min) could effectively remove about 70% of the initial nitrogen content in wood panels. Lemonon et al. [53] showed that the nitrogen removal efficiency was about 66% for the pyrolysis of melamine flooring at a temperature of 275°C and a residence time of 11 min. In another study, Zhan et al. [56] demonstrated that proper pyrolysis at low temperatures (250°C–300°C) could effectively achieve 63% nitrogen removal from medium-density fiberboard (MDF). In the present study, the fast pyrolysis process at 450°C for 2 s was also efficient for nitrogen removal. As previously mentioned, 450°C was chosen to avoid tar formation during biochar activation in the continuous in-house activation furnace. In addition, fast pyrolysis using the CarbonFX technology (250 kg/h) is a pilot scale process compared to laboratory scale furnaces (few grams in a static furnace) used in the studies mentioned above.

Textural characteristics such as specific surface area and pore size distribution of biochar and activated biochars are presented in Table 2. The N₂ adsorption and desorption isotherms of activated biochars pyrolyzed at 450°C and activated at 750°C, 850°C and 950°C are presented in Fig. 4. According to the International Union of Applied Chemistry (IUPAC) classification, the isotherms obtained for the BWP750 and BWP850 are of type I, which reflects the presence of micropores of 2 nm or less. On the other hand, the isotherm obtained for the activated biochar at 950°C belongs to types I and IV, which indicates that the activated biochar produced contains a mixture of micropores and mesopores. As expected, the use of CO₂ as an activating agent during physical activation mainly produces microporosity in the carbon material [57]. All three activated carbons are predominantly microporous at proportions of 85%, 78%, and 66% for the activated biochars prepared at 750°C, 850°C, and 950°C, respectively (the remaining being mesopores).

Nitrogen adsorption isotherms were used to determine the specific surface area, micropore, mesopore, and total pore volume of the different activated biochars, as seen in Fig. 4. The biochar developed low porosity due to the low pyrolysis temperature (450°C) and short residence time (approximately 2 s) in the CarbonFX fast pyrolysis reactor [44]. In addition, attempts to use N₂ gas for biochar adsorption analysis were unsuccessful, as biochar is generally very microporous, with ultramicropores (smaller than 0.7 nm). These limitations in ultramicroporosity analysis using N₂ are due to diffusion restrictions at low pressure, which prevent nitrogen from entering the narrowest micropores. The use of CO₂, rather than N₂ (−196°C), for adsorption analysis is suggested when the analysis is performed at moderate pressures and higher temperatures (0°C) [58].

The specific surface area of biochar derived from pyrolyzed particleboard at 450°C was 99 m²/g. However, the activated biochars had much larger specific surface areas, ranging from 450 to 866 m²/g. It can also be seen that the micropore, mesopore, and total pore volume increased with increasing temperatures from 0.155 to 0.272 cm³/g, 0.028 to 0.139 cm³/g, and 0.183 to 0.411 cm³/g, respectively. This increase in pore volume and specific surface area is due to the penetration of an oxidizing agent (CO₂) into the internal structure of the biochar, thereby resulting in the removal of carbon atoms, and consequently, the opening and widening of inaccessible pores [59]. Among all the activated biochars obtained, the biochar activated at the highest temperature (950°C) had the highest specific surface area, micropore, mesopore and total pore volume, 866, 0.272, 0.139 and 0.411 cm³/g, respectively. A study of the pyrolysis of particleboard waste carried out using a laboratory-scale quartz tube reactor at 250°C, 300°C, and 400°C, followed by activation in the presence of steam at 800°C, 900°C, and 1000°C, presented specific surface areas between 800 and 1300 m²/g and pore volumes between 0.37 and 0.73 cm³/g [19]. Thus, the results from the present study using a larger scale are closer to the ones presented in the literature (at laboratory scale).

XPS analysis was performed to obtain a clear understanding of the nitrogen functionalities in the different samples as well as to study the chemical changes that occur in the biochar and activated biochars prepared by pyrolysis and further activation, respectively. By identifying the energies associated with each peak, it was possible to determine the nature of the chemical elements present in these materials. Carbon (C 1s, 284 eV), oxygen (O 1s, 532 eV) and nitrogen (N 1s, 400 eV) were clearly identified in all spectra. Relatively weak peaks of other elements such as sulfur (S 2p), sodium (Na 1s) and calcium (Ca 2p) were also observed. The integration of the area of each peak after the removal of the baseline (Appendix, Fig. S1) allowed the determination of the atomic concentration of the different elements which are summarized in Table 3.

The elemental composition obtained by XPS is different from the one obtained by elemental analysis. This means that the elemental composition at the surface of the material is different from that within the sample. This is because the distribution of surface functional groups in the sample is not uniform [60]. This difference between the elemental composition obtained by XPS and that obtained by elemental analysis was also observed by Xu et al. [61] on biochar derived from fiberboard. The carbon percentage is always higher when measured by XPS, while the nitrogen percentage is always lower. Despite this, the trend of increasing carbon percentage and decreasing nitrogen percentage after pyrolysis and activation is the same as that observed for elemental analysis. The carbon percentage of the particleboard increased from 74.9% to 82.2% after pyrolysis and 93.5% after activation performed at 950°C. While the percentage of nitrogen decreased from 3.4% to 2.2% and 1.4%, respectively. The N 1s peak of each sample was deconvoluted to identify and estimate the relative proportions of the nitrogenous functional groups. The deconvoluted XPS N 1s spectra of the particleboard, biochar, and activated biochars, and the relative contributions of each nitrogen species are also presented in Fig. S2 (Appendix) and Table 3. The peaks can be attributed to pyridine nitrogen (N-6, 398.7 ± 0.3 eV), pyrrole nitrogen (N-5, 400.3 ± 0.3 eV), and nitrogen oxides (N-X, 402–405 eV) [52]. The dominant nitrogen species in the particle panel, biochar, and activated biochars were N-5, N-6 and N-X, respectively. These results are in good agreement with the activated biochars derived from medium density fiberboard reported by Wu et al. [62]. N-5 accounted for 94.1% of the nitrogenated species in the particleboard (5.9% N-X), but after activation, N-6 appeared and N-5 was reduced for all samples. These results reveal that the chemical state of nitrogen sensitively changed after the thermochemical treatments.

To obtain information on the release of volatile organic compounds during the pyrolysis of particleboard, a thermal degradation process was performed by TGA coupled with GC/MS gas analysis. The sample was studied in quasi-continuous mode, i.e., the evolved gas was ejected every minute into the GC at 100°C. The peaks were characterized and a total of 11 compounds were identified: methane, nitric oxide, trimethylamine, ammonia, carbon dioxide, 1,1-dimethylhydrazine, acetic acid, methylamine, 8-hydroxy-2-octanone, 1-propanol, and 5-hydroxy-2-pentanone. Based on the areas under the chromatographic peaks associated with the different compounds, CO₂ was the major compound emitted during pyrolysis and was detected at temperatures above 250°C. The formation of CO₂ can be mainly attributed to the cracking and rearrangement reactions of carbonyl and carboxyl groups in organic compounds and to the decomposition of carbonates from inorganic compounds [63]. Indeed, during pyrolysis, hemicellulose and cellulose are converted to carbonyls by cracking, while lignin is converted to phenols by dehydration, depolymerization, and decarboxylation [21]. In addition, CH₄ was emitted at low temperatures from 56°C to 156°C due to the cracking of methoxy, methyl, and methylene groups [64].

The analysis of mass spectroscopy data revealed the release of several nitrogenous compounds during pyrolysis, mainly nitric oxide, ammonia, and trimethylamine, which were observed in the temperature ranges of 176°C–255°C, 239°C–301°C, and 287°C–385°C, respectively. These nitrogenous compounds were produced by the degradation of their main precursors, urea-formaldehyde and melamine-formaldehyde [53]. Indeed, nitrogenous gases from the pyrolysis of wood-based panels are significantly influenced by the addition of UF and MF resin [65]. These results are in agreement with the available literature given that most of the compounds were already identified by Girods [66] and Moreno et al. [50]. However, Feng et al. [6] identified isocyanic acid (HNCO) and hydrogen cyanide (HCN) during the pyrolysis of particleboard at 800°C with a heating rate of 20 °C/min using a thermogravimetric analyzer coupled with Fourier transform infrared analysis (TG-FTIR). Neither compound was identified in this study. The composition of the pyrolytic gas depends on several factors such as the pyrolysis reactor, temperature, residence time, heating rate and feedstock [21]. The present study is only based on the fast pyrolysis of particleboard at a pilot scale, while the studies available in the literature used slow pyrolysis furnaces at a laboratory scale.

A semi-qualitative analysis of the gases emitted during the activation of the particleboard-derived biochar was performed using a GC/MS analyzer to identify the volatile organic compounds emitted at 750°C, 850°C, and 950°C. A condenser was used to condense the gases. A sample was taken before and after condensation to evaluate the efficiency of the condenser. The analyzed gases were recovered through the activation process using gas bags. The main gases generated during activation before and after condensation were nitric oxide, ethylenediamine, carbonyl sulfide, benzene, toluene, p-Xylene, styrene, 1,3-cyclopentadiene, acetylene, hydrogen sulfide and acetic acid. The emissions resulting from the activation of particleboard biochars were dominated by aromatic hydrocarbons and, in particular, toluene and benzene, which were detected at the three different activation temperatures. While other aromatic hydrocarbons, such as p-Xylene and styrene, were detected at very low concentrations. The emission of two nitrogen species was also noted for all gas samples, including nitric oxide and ethylenediamine, probably due to the decomposition of pyridinic-N, pyrrolic-N and nitrogen oxides present in the biochar. Also, two sulfur compounds were released during activation given that the percentage of sulfur present in biochar was 0.87%, namely carbonyl sulfide (COS) which is present in all samples and hydrogen sulfide (H₂S) detected only at 850°C after gas condensation. The gases released before and after condensation were almost the same.