Mechanisms of biogenic formaldehyde generation in wood

TR Number
Journal Title
Journal ISSN
Volume Title
Virginia Tech

This work addresses biogenic formaldehyde (CH₂O) generated by wood during the manufacture of non-structural wood-based composites, from which CH₂O emissions are regulated. The target for regulation has been anthropogenic CH₂O released from hydrolytically unstable amino resins like urea-formaldehyde. However, current regulations (the Formaldehyde Standards for Composite Wood Products Act, signed into law in 2010 and implemented in 2016) restrict allowable emissions to such low levels that biogenic CH₂O may affect regulation compliance. The industry has met the latest regulations with new amino resin technologies. Nevertheless persistent anecdotal reports suggest that biogenic CH₂O complicates regulation compliance. This work represents an industry/university cooperation to seek a more thorough understanding of biogenic CH₂O, to begin documentation of biogenic CH₂O levels in wood, and to study the conditions and chemical mechanisms of its formation.

Efforts began by establishing CH₂O analysis using the fluorimetric acetylacetone determination. A custom 12-liter chamber with controlled temperature and relative humidity, and "ultrapure" nitrogen (N₂) ventilation was created to measure CH₂O emissions from flakes sampled from four Virginia pine (Pinus virginiana) trees. Emissions from never-heated specimens varied significantly among the four trees, ranging from 0.02 – 0.19 µg CH₂O/m³g dry wood. Heating (200°C, 1 hour), followed by chamber equilibration, resulted in significantly increased emissions on the order of 50%. Sequential heating, followed by chamber equilibration (in other words, heat/equilibrate/measure emission/repeat), resulted in declining emissions suggesting that a finite chemical source of CH₂O was being depleted by the sequential heat treatment. Flake specimens were stored in the open laboratory, and over 2-3 months laboratory storage, initially high emitting specimens gradually emitted less CH₂O, and initially low emitters gradually emitted more CH₂O. Concerns over laboratory contamination were perhaps allayed when background levels of laboratory CH₂O were determined to be similar to the background levels in the ultrapure N₂ used to ventilate the chamber. Measurement of emissions was abandoned, and thereafter a simple water extraction technique (~ 94% CH₂O recovery) was used to measure the CH₂O content of never-heated and heated wood specimens, where the difference was identified as CH₂O generated due to heating.

Increment cores from living Virginia pine (Pinus virginiana), yellow-poplar (Liriodendron tulipifera), and radiata pine (Pinus radiata) trees were used to measure CH₂O content and CH₂O generation due to heating (200°C, 10 min). Significant variations within and between trees of the same species were observed. Tissue types (juvenile/mature, heartwood/sapwood) sometimes correlated to higher CH₂O contents and greater heat-generation potential; but sometimes not depending upon species. Heating increased CH₂O levels 3-60 fold. Heating with high moisture levels caused greater CH₂O generation than for dry specimens. This moisture effect and a separate serendipitous observation suggested that CH₂O generation is acid catalyzed. Radiata pine generated extraordinarily high CH₂O levels when heated, far exceeding the other two species. It was suggested that pine extractives might catalyze CH₂O generation, perhaps in lignin.

Pinus virginiana wood was heated (200°C, 10 or 60 min) while dry or after aqueous/acid or base pretreatment in order to reveal mechanisms of formaldehyde (CH₂O) generation. Among wood structural polymers, lignin was the overwhelming source of biogenic CH₂O, consistent with prior reports. The effects of wood extractives are mentioned below. The selection of acid catalyst strongly affected CH₂O generation as predicted in the acidolysis literature of lignin model compounds and isolated lignins. Lignin methoxyl cleavage was also observed, but was considered an unlikely source of thermochemical CH₂O. Alkaline pretreatments suppressed CH₂O generation. Regarding wood-based composite manufacture, the implications are that lignin reactions can be manipulated during hot-pressing. Potential benefits include reduced product emissions, and/or novel crosslinking strategies using biogenic CH₂O.

Heat generation of CH₂O in Virginia pine and radiata pine was substantially reduced by extractives removal, but there was no such effect in yellow-poplar wood. Results suggested that pine extractives promote CH₂O generation by catalyzing or otherwise promoting C2 cleavage (acidolysis) in lignin. Thioacidolysis demonstrated that pine lignin reactions were strongly dependent upon the presence or absence of the extractives. When present, pine extractives seemed to promote C2 cleavage (CH₂O generation), but otherwise reduced the overall extent of lignin degradation. When pine extractives were removed, lignin suffered substantial degradation, but apparently less C2 cleavage since CH₂O generation was reduced. In contrast, thioacidolysis showed that yellow-poplar extractives appeared to promote lignin degradation, but extractives removal had no detectable impact on CH₂O generation. Implications exist for biorefinery research because it was shown that lignin reactions can be strongly affected by wood extractives.

Two dimensional, proton-carbon, correlation NMR spectroscopy (2D NMR), and solvent submersion dynamic mechanical analysis (DMA) was used to investigate wood changes caused by heating in the presence or absence of external acid catalysis. 2D NMR was relatively insensitive to fine lignin changes that were detected using thioacidolysis. 2D NMR was effective for observing lignin changes under more extreme heating conditions, and evidence was found for lignin crosslinking reactions that probably occurred through substitution into lignin aromatic rings. DMA showed that most heating conditions caused an increase in the lignin glass transition temperature (Tg), consistent with heat-induced lignin crosslinking. Under one experimental condition of wood heating, DMA showed a reduction in the lignin glass transition temperature (Tg). This suggested that lignin cleavage without subsequent repolymerization might be promoted by carefully controlled conditions, and this has implications for biorefinery research where lignin repolymerization can be problematic.

Finally, this work strongly supported the hypothesis that lignin generates CH₂O through well-known acidolysis pathways where CH₂O is borne from the lignin gamma-methylol group. Therefore, it was predicted that upon heating corn (Zea mays L.) stalk should generate less CH₂O than wood because corn stalk lignins exhibit a high degree of coumaric acid esterfication at the gamma-methylol group. This hypothesis was perhaps verified- it was found that in 4 out of 6 experimental heating conditions that corn stalk generated significantly less CH₂O than Virginia pine.

Wood, biogenic formaldehyde, polymer, lignin, extractives