The materials were selected from those commonly used in the packaging industry. These materials were UF resin, natural wood (larch and poplar), and two types of the PB, one of which was made from larch (PBL) and the other from poplar (PBP). The 30-year-old larch timber and 10-year-old poplar timber both came from the Da Hinggan Mountains (Inner Mongolia, China). The UF adhesive was provided by Beijing Chemical Co. The ratio of formaldehyde to urea was 1.1꞉1.0 and the solid content was 53%. The PB samples were made in the authorso laboratory with 5%, 10% and 20% UF resin respectively. The condition to prepare the PB is controlled precisely to ensure the resin cured. The results of an elemental analysis of the experimental materials are shown in Table 1. The data were obtained from air-drying tests, and the oxygen content was determined by subtraction. The component analysis results of the two timbers are shown in Table 2.
Sample C H O N Larch wood 46.15 6.31 47.36 0.12 Poplar wood 45.24 6.30 48.36 0.10 UF 33.45 5.02 29.31 32.22 PBL 44.15 6.11 46.28 3.46 PBP 43.35 5.96 47.55 3.14 Notes: element analyzer was used to analyzing C, H and N. O% was calculated by subtraction method.
Table 1. Elemental analysis of samples (%)
Sample Cellulose Hemicellulose Lignin pH Ash Larch wood 42.76 20.79 24.03 4.90 0.905 Poplar wood 45.87 28.44 19.12 5.01 1.169
Table 2. Component analysis of samples (%)
The pyrolysis experiment was performed using a pre- programmed temperature ramp on coupled thermal gravimetric-Fourier transform infrared (TG-FT-IR) instruments. Nitrogen did not react under the chosen experimental conditions, and therefore the use of nitrogen as a carrier gas had no effect on the experimental nitrogen production results.
The coupled TG-FT-IR experimental devices included a NETZSCH STA449F3 simultaneous thermal gravimetric analyzer (NETZSCH, Germany) and Bruker TENSOR 27 Type FT-IR spectroscopy (Bruker, Germany). The dedicated connector and tubing connecting the TG analyzer to the infrared spectrometer and the piping connection interface were insulated to maintain a maximum constant temperature of 350℃, which would enable a slow online real-time analysis of the pyrolysis products.
The slow preset temperature-program experiments were conducted in the following manner. The sample was placed in an aluminum oxide (Al2O3) crucible and compacted. Once the crucible was positioned in the experimental setup, the environment was purged with nitrogen to obtain an inert atmosphere. The HR was then set, though the temperature range for pyrolysis was automatically controlled. The products from pyrolysis were carried with the carrier gas into the FT-IR spectroscopy, which was connected to the thermal analyzer for online real-time detection.
A computer-controlled linear heating mode was used to control the HR (10℃/min, 20℃/min, and 40℃/min), sample size (10-25 mg), pyrolysis temperature (25℃- 1000℃), and pyrolysis residence time (10 min). High- purity nitrogen (99.999%) was used during the experiments to ensure an inert atmosphere within the reaction system, and the carrier gas flow rate was 50 mL/min. This flow effectively ensured an adequate residence time and served to prevent secondary cracking of the small port and connecting pipe. The IR spectrometer and TG analyses were conducted at 180℃, which prevented the condensation of tar and absorption of NH3. After completion of each experiment, a blank was read to correct for the phenomenon of virtual thermogravimetric weight- gain and to eliminate any systematic errors. The real pyrolysis curve for each set of experimental conditions was then obtained by subtracting the blank curve. The FT-IR wavenumber range was 500-4000 cm-1, and the scanning frequency was 4 s-1 with a resolution of 1 cm-1.
The initial slow-pyrolysis nitrogen compound products are mainly gaseous nitrogen, carbon nitrogen, and tar nitrogen (Chen et al., 2017). Because this study was limited to the FT-IR analysis, only the gaseous nitrogen was analyzed. The initial slow-pyrolysis products of gaseous nitrogen are mainly NH3, HCN, and HNCO compounds. Most studies reported that the NH3 and HCN are representative materials of gaseous nitrogen (Becidan et al. 2007; Wang et al., 2016). However, there is some controversy over HNCO. While its formation during the heat treatment process is expected, there are some uncertainties in the literature over its detection and manner of generation. The main reasons for this uncertainty are the unreliability of the test method and inaccuracy of the HNCO analysis. Many researchers have used solution absorption to detect the nitrogen compounds, but when the HNCO is released during an experiment, ammonium (NH4+) will form during hydrolysis; therefore, the HNCO can not be measured by this method. Some studies have discussed the presence and release characteristics of the HNCO when using FT-IR technology to analyze nitrogen compounds in the product (Dejong et al., 2003; Hansson, 2003; Girods et al. 2008a, 2008b; Wang et al., 2011; Ren et al., 2012). However, researchers thought that the use of the HNCO line detection methods, such as FT-IR, is not viable because the biomass pyrolysis process is bound to generate a little amount of water. Additionally, hydroxyl ions generated during pyrolysis would instantly react with any generated HNCO, which would produce NH3. In view of this, it was assumed in this study that the slow pyrolysis of the PB waste and its components generated only two gaseous nitrogen compounds, NH3 and HCN, and that these compounds alone were detected by the FT-IR analysis.
Figures 1 and 2 are the standard spectra for the NH3 and HCN, respectively, from the US National Institute of Standards and Technology (NIST). Referring the previous literatures, the NH3 and HCN windows were determined (Dejong et al., 2007; Peng et al., 2009; Nola et al., 2009, 2010; Giuntoli et al., 2009).