Wood-borne formaldehyde varying with species, wood grade, and cambial age.
While the formaldehyde issue primarily focuses on adhesive systems used in wood-based panels, natural wood itself contains detectable formaldehyde. Potentially, this wood-borne formaldehyde is emitted over time; therefore, even with wood alone no "zero emission" is evident. In this work, the variation of formaldehyde contents in important commercial wood species that are dried and converted to wood particles for wood-based panel production was studied. Furthermore, whether wood grade or juvenile vs. mature wood have any effect on the formaldehyde content was determined. Results indicate that formaldehyde varied up to 4-fold across commercial softwood and hardwood species, but remained at low concentrations (under 1 mg/100 g). Softwoods generally had higher formaldehyde contents than hardwoods, while wood grade seemed to have no effect. The differences between juvenile and mature wood, however, were much more expressed. The lowest formaldehyde content was seen with juvenile wood from beech (under 0.15 mg/100 g), and the highest concentration was found in mature pine wood (approximately 0.70 mg/100 g).
Formaldehyde emission has been a subject of public concern in the formaldehyde-based resin-bonded wood-based panel industry for decades. Wood-based panel products, such as plywood, particleboard, oriented strandboard (OSB), and medium density fiberboard (MDF), are most commonly manufactured using either urea-formaldehyde or phenolformaldehyde adhesives. According to the European standard EN 312 (CEN 2003), particleboards are classified as E 1 for formaldehyde contents up to 8 mg/100 g (o.d. [oven dry]), and as E 2 for concentrations between 8 mg/100 g and 30 mg/100 g (o.d.), respectively. Currently, industry has been targeting even lower formaldehyde concentrations. Consumer products containing wood-based panels might release gaseous formaldehyde depending on temperature, relative humidity (RH), air change rates, and specific product properties such as the vertical density profile (Marutzky 1989, Marutzky et al. 1992), which has resulted in a significant amount of health-related issues (Bernstein et al. 2008). No irritation can be found up to 0.370 mg/[m.sup.3] (0.30 ppm) formaldehyde air loading (Schupp et al. 2005), while loadings of 0.625 mg/[m.sup.3] (0.50 ppm) have shown irritations of eyes and the upper respiratory tract of human beings in case of acute impact (Schupp et al. 2005). Formaldehyde has also been found to produce nasal carcinomas after long-term chronic exposure to 14.1 ppm and 5.6 ppm of formaldehyde in rats and mice, respectively (Kim and Kim 2005). While the issue of formaldehyde is primarily focused on adhesive systems used in wood-based panels, natural wood itself also emits detectable formaldehyde. While there are numerous investigations into wood-based product emissions, little data are available for natural wood emissions from different species and wood types.
Meyer and Boehme (1997) found increased formaldehyde emissions after 14 days of kiln-drying from green conditions to a target of 8 percent relative moisture content (MC) with the species European beech (Fagus sylvatica L.), Douglas-fir (Pseudotsuga menziesii Mirb. (Franco)), Norway spruce (Picea abies L. (Karst.)), and Scots pine (Pinus sylvestris L.). Roffael (1988) showed for pine particles that wood inherent formaldehyde increased with drying time and temperature. Roffael (1982) described potential hydrolytic processes leading to formaldehyde formation due to degradation of lignin and hemicelluloses; thermally induced formation of formic and acetic ions were responsible for delignification. Two positions within lignin are prone to formaldehyde separation under acidic conditions: 1) side-chain units containing a primary alcohol, based on either the retro-aldol or the retroPrins reaction, and 2) carbonyl groups in the [beta]-position. Marutzky and Roffael (1977) discussed the conversions of [omega]-hydroxypropiovanillon into acetovanillon (retro-aldol reaction), [alpha]-(3-methoxy-4-hydroxyphenyl)-glycerin-[beta]-guaiacylether into 3-methoxy-4-hydroxy-styryl-[beta]-guaiacylether (retro-Prins reaction) as well as [beta]-1-dilignol-units toward styryl derivatives (retro-Prins reaction) as potential delignification reactions, which also lead to the formation of formaldehyde. Colakoglu et al. (1998) found a decrease in natural acetyl groups due to steaming and drying with the species Oriental beech (Fagus orientalis L.) and Turkish pine (Pinus brutia Tenure), which was also associated with higher formaldehyde emission rates of the manufactured plywood. Regarding the disintegration of hemicellulose, Marutzky and Roffael (1977) described reactions of hexoses to oxymethylfurfural and further toward furfural and formaldehyde. All of the reactions discussed by Marutzky and Roffael (1977) have' in common that the transformation of methyl alcoholgroups leads to the formation of formaldehyde.
In addition to the existence of wood-borne formaldehyde, lignocellulosic substances may also contain specific chemical structures such as phenolics (e.g., A-tings of the phloroglucinol-type from tannins) that are capable of decreasing formaldehyde emissions of wood-based panels (Hashidia et al. 2006), which is then followed by a polycondensation mechanism such as in the case of phenol-formaldehyde glues. Lower formaldehyde emissions of boards made solely from heartwood compared to sapwood is well documented (Lelis et al. 1993, 1994; Dix and Roffael 1995, 1997). Lelis and Roffael (1995) proved higher reactivity of formaldehyde with heartwood extractives of Douglas-fir (Pseudotsuga menziesii Mirb. (Franco)) compared to sapwood by the Stiasny reaction as well as by UV absorption at 280 nm. Since bark has a higher phenolic content, formaldehyde emissions from wood-based panels may be lowered by adding bark to the wood furnish prior to consolidation (e.g., Roffael and Dix 1988, Buchholzer 1990).
Many detection methods for formaldehyde evaluation have been developed by analyzing either the total amount of formaldehyde through extraction and titration or determining the amount of emitted formaldehyde gas. The most common perforator method (EN 120, CEN 1993) based on extraction is also capable of detecting low concentrations, as found in natural wood. A number of methods exist for the analyses of emitted formaldehyde (e.g., cubic chamber method, WKI bottle method, flask method, gas analysis; see, Roffael 1975, Meyer and Boehme 1997).
Wood-based panels have been developed using formaldehyde-free binders, which may result in very low but detectable formaldehyde contents. This wood-borne content of formaldehyde is a regularly formed product; therefore, "zero emission" is intrinsically, under conventional wood processing conditions, not achievable. The objective of this work was to determine how formaldehyde contents vary across important commercial wood species that were dried and converted to wood particles. In addition to species-specific effects, differences in woodborne formaldehyde from different wood grades as well as juvenile vs. mature wood are also discussed.
Materials and methods
Samples were collected from the softwood species Norway spruce (Picea abies (Karst.) L.) and Scots pine (Pinus sylvestris L.), as well as from the hardwood species European oak (Quercus sp.), European beech (Fagus sylvatica L.), and poplar (Populus sp.). For each species two grades were chosen, a superior one according to class A/B, and an inferior one according to class C and worse, both following the Austrian timber grading rules (Anon. 2006). Class A/B was of good to medium quality, low in knottiness, no marked reaction wood, and low in spiral grain. Fungal decay, insect attack, and discoloration were widely excluded. Class A/B wood was assigned as "ideal." The low grade was selected according to class C of the Austrian timber grading rules, which defines this wood as below-average quality, showing considerable knottiness, high reaction wood and spiral grain, pitch pockets, and other acceptable defects. Class C was assigned as "real." For each grade class, juvenile and mature wood were also identified. A composite sample was prepared for each subgroup, totalling 20 different samples.
Wood particles with a mesh size between 2 mm and 4 mm were produced using a disk flaker. Particles were dried to an average moisture content of 5 percent dry mass in a 24 m laboratory flow dryer at 200 [degrees]C to pass one or two times each for approximately 2 to 4 seconds. Drying medium was electrically heated and free from combustion products. Particles were stored in sealed plastic bags under cooled conditions (4 [degrees]C). Average thickness of the wood particles was 0.35 mm, with an aspect ratio of about 5. This distinct particle geometry resulted in an average "bulk density" of 88 kg/[m.sup.3] for poplar, 103 kg/[m.sup.3] for spruce, 113 kg/[m.sup.3] for pine, 129 kg/[m.sup.3] for beech, and 153 kg/[m.sup.3] for oak.
According to the EN 120 (CEN 1993) standard, samples were individually stored for 1 Week at 23 [degrees]C and 50 percent relative humidity leading to an equilibrium moisture content of 9 percent. Wood-borne formaldehyde was determined with the perforator method. Following EN 120 (CEN 1993) procedures, each of the 20 different samples was analyzed on the basis of double measurements, with the lower value being not more than 20 percent lower than the higher one. Since formaldehyde concentration of natural wood is lower compared to wood-based panels, flask size was expanded from 1 to 2 liters to accommodate the required sample mass of 100 to 110 g. Consequently, the extraction solution amount (toluene) was increased from 600 to 900 mL or even 1300 mL. This was in accordance with Roffael (1989), who found that a reduced amount of toluene did not affect the results obtained with the EN 120 (CEN 1993) method. Distillation was done for 2 hours as described in EN 120 (CEN 1993), but the heating phase until the first bubbles passed the filter had to be extended from the 20 to 30 minute range up to 1 hour. All other conditions were in accordance with EN 120 (CEN 1993).
In addition to the perforator measurements, wood chemical analyses were done on the same samples. The wood was ground using an ultra centrifugal mill to pass to an 80-mesh sieve, hydrolyzed after extraction in 72 percent sulphuric acid for 2 hours, diluted with water to a 3 percent solution, boiled for 4 hours, filtrated, and the precipitate rinsed and dried. Acid insoluble lignin was gravimetrically determined (TAPPI 1991 ). Acid soluble lignin was measured through determining the absorption of the precipitate in a UV spectrometer at 205 nm, relative to the absorption of a 3 percent sulphuric acid (TAPPI 1988). Holocellulose was obtained by extraction in an accelerated solvent extractor (ASE), by adding a mixture of toluene and ethanol (2:1) followed by a reaction for 15 minutes at 75 [degrees]C with 10 percent peracetic acid, filtration, rinsing with a mixture of acetone and ethanol (1:1), and drying at 103 [degrees]C. The results obtained for all of the raw materials were based on double measurements.
Results and discussion
Across the measured species, formaldehyde concentration varied between 0.06 mg/100 g (dry wood mass) and 0.71 rag/ 100 g, with the mean value of 0.37 mg/100 g (Fig. 1). These values are more than 10-fold lower than the threshold value of 8 mg/100 g for particleboards classified as E 1. The two coniferous species, spruce and pine, showed significantly higher formaldehyde compared to the three hardwood species. The average formaldehyde value for pine was 0.54 mg/ 100 g, followed by spruce at 0.47 mg/100 g. Beech was lowest at 0.27 mg/100 g, followed by poplar as second lowest at 0.29 mg/100 g, and oak at 0.36 mg/100 g. Meyer and Boehme (1997) found slightly lower perforator values for dried wood (0.16 mg/100 g for beech and 0.28 mg/100 g for pine). The lower values have most likely evolved from the fact that formaldehyde was determined with solid wood cubes and not with ground wood particles.
Although formaldehyde levels from the "ideal" graded wood were slightly but constantly above the lower "real" grading (approximately 0.04 mg/100 g; Figs. 2 and 3), the wood grading effect was not statistically significant. Significant differences within the species were found between mature and juvenile wood: for juvenile wood, beech was lowest and spruce highest (Figs. 1 and 2), and in mature wood, the oak was lowest and pine highest (Figs. 1 and 3). For beech, poplar, and pine, the mature wood concentrations in formaldehyde were higher than that of juvenile wood (Fig. 1). The opposite was the case for oak and spruce. Dix and Roffael (1997) examined particleboards produced with pine heartwood vs. sapwood and also found higher concentrations in the outer sapwood and in more mature wood, respectively.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Linear regression analysis was performed with formaldehyde as the dependent variable and total lignin as well as holocellulose as independent variables (Table 1). The standardized model is written as follows (adjusted [r.sup.2] = 0.25):
Formaldehyde content = 0.48 total lignin content - 0.27
holocellulose content + error
The model shows a stronger impact of lignin than holocellulose affecting the formaldehyde content. The regression coefficients indicate the amount of change expected in formaldehyde relative to its standard deviation (SD), for a change in one SD in the independent variable lignin or holocellulose. A one SD change in lignin will cause a 0.48 times SD change for formaldehyde.
It is known that wood-borne formaldehyde is formed from the main wood components cellulose, hemicellulose, and lignin as well as its extractives (Schafer and Roffael 2000). Each species has its own characteristic formaldehyde emission potential due to the existing chemical composition. Lignin releases considerably more formaldehyde than the carbohydrate components in wood (Roffael 2006), which explains the higher coefficient in the model. It is obvious that the cambial age effect superimposes with heartwood and sapwood. The species oak and pine, both forming significant amounts of extractives in their heartwood, show greater formaldehyde differences between juvenile and mature wood than spruce and poplar. Although formaldehyde contents of juvenile and mature beech wood are significantly different, absolute values for beech are among the lowest obtained in this study. Beech "red heartwood" is a facultative oxidation caused type of heartwood established at later tree ages (Knoke 2002). This heartwood type and the low extractive content of beech wood might further explain its low wood-borne formaldehyde content.
The results of this study show that wood-borne formaldehyde of dried wood particles varied up to 4-fold across commercial softwood and hardwood species, but remained at low concentrations under 1 mg/100 g. Generally, softwoods have higher formaldehyde contents than hardwoods. While different wood grades seem to have no effect on formaldehyde content, the differences between juvenile and mature wood were much more expressed. The lowest formaldehyde content was found with juvenile wood from beech (under 0.15 mg/100 g) and the highest was mature wood from pine (around 0.70 mg/ 100 g). The presented data are important because consumers need to understand that formaldehyde emissions occur through natural degradation processes. Thus, "zero emission" wood products are simply not achievable under today's processing techniques.
This work was funded by the Austrian Government as well as the Provincial Governments of Upper Austria, Lower Austria, and Carinthia. The authors would like to thank Thomas Ters, BOKU Vienna, for fruitful discussions, and the two anonymous reviewers for very constructive comments on the earlier versions of the manuscripts.
Anonymous. 2006. Osterreichische Holzhandelsusancen. 310 pp.
Bernstein, J.A., N. Alexis, H. Bacchus, I.L. Bernstein, P. Fritz, E. Horner, N. Li, S. Mason, A. Nel, J. Oulette, K. Reijula, T. Reponen, J. Seltzer, A. Smith, and S.M. Torlo. 2008. The health effects of nonindustrial indoor air pollution. J. Allergy Clin. Immunol. 121(3):58-591.
Buchholzer, P. 1990. Einfluss von Pappelrinde auf einige physikalischelastomechanische und chemische Eigenschaften von Pappelholzspanplatten. Holz als Roh- und Werkstoff 48:30.
Colakoglu, G., S. Colak, and M. Tufekci. 1998. Einfluss der Dampfung und Trocknung von Furnieren auf den Acetylgruppengehalt und die Formaldehydabgabe yon Sperrholz. Holz als Roh- und Werkstoff 56:121-123.
Dix, B. and E. Roffael. 1995. Zum Verhalten des Splint- und Kernholzes der Larche (Larix decidua) bei der Herstellung von feuchtebestandigen Spanplatten unter Einsatz verschiedener Bindemittel. Holz als Roh- und Werkstoff 53:357-367.
-- and --. 1997. Einfluss der Verkemung und des Baumalters auf die Eigenschaften von Spanplatten aus Kiefernholz (Pinus sylvestris)--Teil 2: Physikalisch-technologische Eigenschaften und Formaldehydabgabe von Spanplatten aus Kern- und Splintholz der Kiefer. Holz als Roh- und Werkstoff 55:103-109.
European Committee for Standardization (CEN). 1993. EN 120. 1993. Wood-based panels--Determination of formaldehyde content--Extraction method called the perforator method. 1 Feb. 1993.
--. 2003. EN 312. Particleboard--Specifications. August 2003.
Hashidia, K., S. Ohara, and R. Makino. 2006. Improvement of formaldehyde-scavenging ability of condensed tannins by ammonia treatment. Holzforschung 60:178-183.
Kim, S. and H.J. Kim. 2005. Comparison of standard methods and gas chromatography method in determination of formaldehyde emission from MDF bonded with formaldehyde-based resins. Bioresour. Technol. 96(13): 1457-1464.
Knoke, T. 2002. Value of complete information on red heartwood formation in beech (Fagus sylvatica). Silva Fenn.36(4):841-851.
Lelis, R., E. Roffael, and G. Beeker. 1993. Zum Verhalten yon Splintund Kernholz der Kiefer bei Verleimung mit Harnstoffformaldehydharzen (UF-Harzen). Holz-Zentralbl 7:120-121.
--, and --, and --. 1994. Zur Verleimbarkeit von Splint- und Kernholz der Douglasie mit Phenolformaldehydharzen (PF-Harze), Melamin-Harnstoff-Phenol-Formaldehydharzen (MUPF-Harzen) und Diisocyanat-Klebstoffen (PMDI). Holz-Zentralbl 129:2144, 2146; 131/132:2181-2182.
-- and --. 1995. Uber die Reaktivitat yon Douglasiensplint- und -kernholz und deren Heisswasserextrakte gegentiber Formaldehyd. Holz als Roh- und Werkstoff 53:12-16. Marutzky, R. and E. Roffael. 1977. Uber die Abspaltung von Formaldehyd bei der thermischen Behandlung. Teil 1: Modell-Versuche. Holzforschung 31:8-12.
--. 1989. Moglichkeiten der Formaldehydminderung in belasteten Innenraumen. Holz als Roh- und Werkstoff 47: 207-21l.
A. Flentge, and C. Boehme. 1992. Abh/ingigkeit der Formaldehydabgabe von MDF vom Rohdichteprofil. Holz als Rohund Werkstoff 50:239-240.
Meyer, B. and C. Boehme. 1997. Formaldehyde emission from solid wood. Forest Prod. J. 47(5):45-48.
Roffael, E. 1975. Messung der Formaldehydabgabe---Praxisnahe Methode zur Ermittlung der Formaldehydabgabe harnstoffgebundener Spanplatten fur das Bauwesen. Holz-Zentralbl 111:1403-1404.
--. 1982. Die Formaldehydabgabe von Spanplatten trod anderen Werkstoffen. DRW-Verlag Weinbrenner-KG, Stuttgart, Germany. 154 pp.
--. 1988. Perforator- und WKI-Wert von Kiefemspanen, abhangig von den Trocknungsbedingungen. Holz als Roh- und Werkstoff 46:106.
-- and B. Dix. 1988. Zur Bedeutung von schnellwuchsigen Baumarten als Rohmaterial fur die Holzwerkstoffherstellung unter besonderer Berticksichtigung von Pappelholz fur Spanplatten. Holz als Roh- und Werkstoff 46:245-252.
--. 1989. Einfluss der Toluolmenge bei der Bestimmung des Perforatorwertes yon El-Spanplatten. Holz als Roh- und Werkstoff 47:248.
--. 2006. Volatile organic compounds and formaldehyde in nature, wood and wood based panels. Holz als Roh- und Werkstoff 64:144-149.
Schafer, M. and E. Roffael. 2000. On the formaldehyde release of wood. Holz als Roh- und Werkstoff 58:259-264.
Schupp, T., H.M. Bolt, and J.G. Hengstler. 2005. Maximum exposure levels for xylene, formaldehyde and acetaldehyde in cars. Toxicology 206:461-470.
TAPPI 222 om-88. 1988. Acid-insoluble lignin in wood and pulp. TAPPI Standard.
TAPPI UM 250. 1991. Acid-soluble lignin in wood and pulp. TAPPI Useful Methods.
Rupert Wimmer *
The authors are, respectively, Researcher, Competence Center for Wood Composites and Wood Chemistry, Linz, Austria (m.weigl@ kplus-wood.at); Professor, Dept. for Material Sciences and Process Engineering; BOKU--Univ. of Natural Resources and Applied Life Sciences, Vienna, Austria, as well as Professor, Inst. for Wood Biology and Wood Technology, Georg-August-Universitat, Gottingen (RWimmer@tmi-goettingen.de); Researcher, Inst. for Natural Materials Technology, BOKU--Univ. of Natural Resources and Applied Life Sciences, Tulln, Austria (firstname.lastname@example.org); and Head of Research and Development, Fritz Egger GmbH and Co. Holzwerkstoffe, Unterradlberg, Austria (email@example.com). This paper was received for publication in June 2008. Article No. 10498.
* Forest Products Society Member.
Table 1.--Mean values of total lignin and holocellulose content of all 20 different samples based on double measurements. Species Grade Cambial age Total lignin Holocellulose (%) Oak Ideal Mature 20.3 60.1 Oak Real Mature 20.0 61.9 Oak Ideal Juvenile 21.1 62.9 Oak Real Juvenile 23.6 63.5 Beech Ideal Mature 22.8 71.8 Beech Real Mature 20.5 69.5 Beech Ideal Juvenile 22.8 69.3 Beech Real Juvenile 22.7 70.5 Poplar Ideal Mature 22.8 74.1 Poplar Real Mature 23.9 68.9 Poplar Ideal Juvenile 20.6 72.1 Poplar Real Juvenile 25.5 67.7 Pine Ideal Mature 25.2 64.2 Pine Real Mature 24.8 65.3 Pine Ideal Juvenile 26.4 60.1 Pine Real Juvenile 24.5 60.8 Spruce Ideal Mature 26.8 67.0 Spruce Real Mature 25.6 66.5 Spruce Ideal Juvenile 27.3 65.8 Spruce Real Juvenile 25.5 66.8
|Printer friendly Cite/link Email Feedback|
|Author:||Weigl, Martin; Wimmer, Rupert; Sykacek, Eva; Steinwender, Martin|
|Publication:||Forest Products Journal|
|Date:||Jan 1, 2009|
|Previous Article:||Measuring resin-adhesive spray characteristics using a laser diffraction analyzer.|
|Next Article:||Coming events.|