This work investigates the feasibility of producing low density particleboards using an adhesive system based on sour cassava starch, taking advantage of its adhesive and self-expansion properties. Relevant properties of the produced particleboards were evaluated according to European Standards including: density, internal bond, moisture content and thickness swelling. Low density particleboards were produced with densities between 207 kg/m³ and 407 kg/m³. The best performance corresponded to particleboard with a density of 318 kg/m³, an internal bond strength of 0.67 N/mm², and a thickness swelling of 8.7%. These values meet the standard requirements of general purpose lightweight boards for use in dry conditions. Heat post-treatment (24 h at 80 °C) led to lower internal bond strength, due to retrogradation (recrystallization of amylose and amylopectin chains upon cooling) causing higher rigidity of the starch binder. However, it showed to have a significant effect on decreasing the thickness swelling.

It has been shown that wood particleboards bonded with sour cassava starch can display low density combined with good physico-mechanical performance, thanks to starch being able to produce a strong foam that fills the interparticular space. Here we optimize the pressing conditions for the production of these panels. The procedure involved hot-plate pressing in two stages: (1) lowering the top platen to a specified thickness for a duration designated as pressing time, followed by (2) raising the top platen to allow panel expansion for a duration designated as hold time. The parameters studied were the pressing time (10 to 150 s), the hold time (290 to 890 s), and the top platen temperature (80 to 190 °C). The hold time and pressing time showed to be crucial parameters. The best operating conditions corresponded to 600 s of press cycle time, comprising 60 s of pressing time and 540 s of hold time. The top platen temperature used was 190 °C. The particleboards produced had a density of 405 kg·m⁻³, an internal bond strength of 0.44 N·mm⁻², and a thickness swelling of 13.2%. This can be considered as very good performance, taking into account the panels’ low density.

In the present work, and for the first time, totally biosourced low-density particleboards were produced using cardoon particles (a no added value by-product from the Portuguese cheese making industry), bound with a potato starch adhesive. Different starch/cardoon ratios (0.6, 0.8, 1 and 1.2) were tested and the effect of different bio-based additives (chitosan, wood fiber and glycerol) on the performance of the adhesive system was evaluated. The best result was obtained for a formulation with a starch/cardoon mass ratio of 0.8, a chitosan/starch mass ratio of 0.05 and a water/starch mass ratio of 1.75. The particleboards produced had a density of 323 kg·m⁻³, internal bond strength of 0.35 N·mm⁻² and thickness swelling of 15.2%. The values of density and internal bond strength meet the standard requirements of general-purpose lightweight boards for use in dry conditions according to CEN/TS 16368 specification. Moreover, the susceptibility of the formulations with best results was established against subterranean termites and one decay fungi.

This work analyzes the life cycle environmental impacts of producing a particleboard based on cardoon fibers and a starch/chitosan adhesive from a “cradle-to-gate” perspective, considering the following life cycle steps: raw material production, adhesive preparation (component mixing and heating), cardoon fiber preparation (crushing and sieving), adhesive and fiber mixing, hot-pressing and final processing. The functional unit is a particleboard with the dimensions of 220 × 220 × 16 mm³. For the life cycle inventory, experimental data obtained from the production of particleboard on a pilot scale were used. The Aspen Plus V9 software was used to simulate the heating process in the manufacture of the biological adhesive and obtain the data associated with this stage. Portuguese or European conditions were considered for the background processes, using data from the EcoInvent V3.5 LCI database. The environmental impacts were quantified using the RECIPE methodology. To complement the study, the VOCs present in the panel were analyzed using the “active headspace” technique. The results show that for most of the environmental impact categories, energy consumption is dominant, followed by starch and chitosan production. Using fully renewable electricity produced in photovoltaic panels, instead of the Portuguese electricity mix, significantly reduces the impacts in most of the environmental impact categories, for example, the carbon footprint is reduced by 34%. Future studies will analyze how the environmental impacts can be further reduced, and how process scale-up may influence them.