Eur. J. Wood Prod. (2012) 70:621628 DOI 10.

1007/s00107-011-0593-4

O R I G I NA L S O R I G I NA L A R B E I T E N

Inter-laboratory comparison of formaldehyde emission from particleboard using ASTM D 6007-02 method
Mohamed Z.M. Salem M. Bhm . Barck J. Srba

Received: 14 April 2011 / Published online: 19 January 2012 Springer-Verlag 2012

Abstract An inter-laboratory comparison of formaldehyde emission (FE) from laminated (PL) and uncoated (P2) particleboard (PB) with 16 mm thick-E1, used for non-structural and interior applications such as furniture materials and supplied from a commercial plant in the Czech Republic, was performed in two laboratories using the American small test chamber (ASTM D 6007-02). The results showed highly signicant differences between the laboratories for PL (P < 0.001), and not signicant for P2 (P = 0.33), and the differences between samples were signicant for P2 and PL (P < 0.05). The FE values ranged between 0.056 and 0.090 and from 0.088 to 0.099 ppm for PL and from 0.129 to 0.223 and from 0.175 to 0.182 ppm for P2 as measured in laboratories 1 and 2, respectively. There was a good correlation (R 2 = 0.84) between the formaldehyde values measured for both types of particleboards in the two laboratories. Most of the formaldehyde values were in agreement with the requirements of the California Air Resources Board (CARB) regulation. Furthermore, the results suggested that there was a higher homogeneity between the samples in laboratory 2

than in laboratory 1. The variation was related to the chamber conditions and sampling as well as inter-panel variations. Vergleich der in zwei Prfstellen gem ASTM D 6007-02 gemessenen Formaldehydabgabe aus Spanplatten Zusammenfassung Die Formaldehydabgabe (FE) aus beschichteten (PL) und unbeschichteten (P2) 16 mm dicken E1-Spanplatten (PB) fr nicht tragende Verwendungen im Innenbereich, wie zum Beispiel Mbel, wurde in zwei Prfstellen verglichen. Die Platten wurden von einer Firma in Tschechien geliefert und in der kleinen Prfkammer nach ASTM D 6007-2 geprft. Die Ergebnisse der zwei Prfstellen unterschieden sich bei den beschichteten Platten stark signikant (P < 0,001) und bei den unbeschichteten Platten nicht signikant (P = 0,33). Bei beiden Prfkrpertypen P2 und PL unterschieden sich die einzelnen Proben signikant (P < 0,05). Die Werte der Formaldehydabgabe lagen in Labor 1 bei den beschichteten Platten (PL) zwischen 0,056 und 0,090 ppm und bei den unbeschichteten Platten (P2) zwischen 0,129 und 0,223 ppm. In Labor 2 wurden Werte zwischen 0,088 und 0,099 ppm fr PL und fr P2 zwischen 0,175 und 0,182 ppm gemessen. Die in den beiden Laboren gemessenen Formaldehydwerte beider Plattentypen korrelierten gut (R 2 = 0,84). Die meisten der Werte erfllten die Anforderungen der Luftreinhaltungskommission Kaliforniens (California Air Resources Board CARB). Daneben zeigen die Ergebnisse, dass die Proben in Prfstelle 2 homogener waren als die Proben in Prfstelle 1. Diese Schwankung wurde auf Unterschiede zwischen den Prfkammern, die Probennahme sowie Unterschiede zwischen den Platten zurckgefhrt.

M.Z.M. Salem M. Bhm . Barck Faculty of Forestry and Wood Sciences, Czech University of Life Sciences Prague, Kamck 1176, Praha 6-Suchdol, 165 21 Prague, Czech Republic M.Z.M. Salem ( ) Faculty of Agriculture (ELShatby), Forestry and Wood Technology Department, Alexandria University, Alexandria, Egypt e-mail: zidan_forest@yahoo.com M.Z.M. Salem e-mail: salemmohamed@d.czu.cz J. Srba Timber Research and Development Institute, Prague, Czech Republic

622

Eur. J. Wood Prod. (2012) 70:621628 Table 1 Properties and composition of melamineureaformaldehyde adhesive resin Tab. 1 Eigenschaften und Zusammensetzung des MelaminHarnstoffFormaldehyd-Klebstoffes Parameter Resin content (%)b Solid resin content (%) Viscosity (mPa s at 20 C) pH at 20 C Density Wax (%)c Hardener % (NH4 NO3 Free formaldehyde )c F/(M + U) molar ratio MUFa Face: 11, Core: 7 62 150600 9.011 11501250 kg/m3 0.5 3 (57% urea as solid content) 1.1 (1.1:1) <0.2%

1 Introduction Formaldehyde is now a widely present indoor air pollutant chemical that has been found at low levels in homes, ofces, and the urban environment. Many consumer products including particleboards (PB) containing formaldehyde-based resins release formaldehyde vapor, leading to consumer dissatisfaction and health hazard-related complaints in the indoor air (Pickrell et al. 1986; Salthammer et al. 2010). PB is one of the major wood-based panels used in interior applications as furniture and the main source of formaldehyde emitted inside buildings causing health hazard effects. The International Agency for Research on Cancer (IARC) conducted an evaluation of formaldehyde and concluded that there is sufcient evidence that formaldehyde is toxic and is associated with possible health hazards causing nasopharyngeal cancer in humans (i.e., in the region of the throat behind the nose). At concentrations above 0.1 ppm in air, formaldehyde can irritate the eyes and mucous membranes resulting in watery eyes. Formaldehyde inhaled at this concentration may cause headaches, a burning sensation in the throat, and difculties breathing, as well as triggering or aggravating asthma symptoms (IARC 2004; Kim et al. 2010). Simultaneously, good laboratory methods were of the highest importance for controlling the formaldehyde emission (FE) from exposed areas of the nal product. Standard methods for measuring FE from wood based panels use test chambers. In Europe, the chamber (0.225, 1 and 12 m3 in volume) according to EN 717-1 (2004) and in North America the large chamber (volume 22 m3 ) according to ASTM E 1333-96 (2002) were proposed as primary reference methods. To measure the FE in a chamber takes time and requires special and expensive equipment. Simpler laboratory methods that can be used for inhomogeneous products with good correlation to the chamber methods are therefore needed. Numerous methods have been mainly used for the determination of FE from PB and other wood products such as perforator (EN 120 1993), 24-h Japanese desiccator (JIS A1460 2001), ask (EN 717-3 1996), and Field and Laboratory Emission Cell (FLEC) (Risholm-Sundman 1999; Kim et al. 2010) methods, and good correlations have been found between the FE values of the chamber methods and those from other methods (Risholm-Sundman 1999; Sundin et al. 1987; Que et al. 2007; Que and Furuno 2007; Park et al. 2011). However, the lack of certied reference material made it difcult to establish an inter-calibration between test methods (Bulian et al. 2003). Moreover, the signicant variations in the FE measurements between laboratories were

Melamine content (wt % to MUF resin) 4 (in powder form)

a Letter M represents melamine, letter U represents urea and letter F represents formaldehyde b Percent c Percent

based on oven dry weight of wood particles based on solid content of MUF

due to the heterogeneities of chamber conditions such as volume, materials, sampling air, and specic differences in test conditions, and it was reported that comparison of FE measurements carried out by different laboratories becomes important in view of their utilization for material emission databases (Wiglusz et al. 2000; Risholm-Sundman et al. 2007; Yrieix et al. 2010; Risholm-Sundman and Wallin 1999). The California Air Resources Board (CARB) adopted a new regulation in April 2007 to reduce the FE from composite wood products (CARB 2007). The modied version of this regulation was nally approved in April 2008. This regulation proposed the reduction of FE standards from PB in two phases: Phase 1 limits the effective date to 2009 (0.18 ppm) and Phase 2 limits the effective date to 2010 2011 (0.09 ppm). This new regulation has established the most stringent FE limits on wood products in the United States and requires that wood panels and products manufactured from wood panels be certied by a third party laboratory (Third Party Certier, TPC) approved by CARB. On top of that, the secondary method, ASTM D 6007-02 (2002) which is a small-scale chamber ranging in size from 0.02 to 1 m3 with additional conditions has been specied extensively to measure the FE for the requirements of CARB and the applicable CARB limits. In this article an inter-laboratory comparison of FE measurements from PB was conducted in two different laboratories using ASTM D 6007-02 as a quality control test method.

Eur. J. Wood Prod. (2012) 70:621628

623

2 Material and methods 2.1 Test material The emission of formaldehyde was measured from freshly produced PB from a commercial plant in the Czech Republic with 16 mm thick-E1 (May, 2010) bonded with melamine ureaformaldehyde (MUF) adhesive resin (Table 1). The boards were classied into two types; the rst type from 15 was laminated PB (PL) and the second type from 6 10 was uncoated PB (P2). The boards were manufactured in three layers with Norway spruce (Picea abies L.) particles (17.5% of ne particles for each face and back layer with 65% of coarse particles for the core layer). The laminated boards were coated with melamine-impregnated decorative paper on both sides with grammage 80 g/m2 . The temperature, pressure and specic press time used to manufacture the boards were 195 C, 3.43 MPa and 14 s/mm. For the melamine-impregnated papers, alpha cellulose decorative papers were impregnated with MUF adhesives (62% solid content). The press conditions for coating the melamine-impregnated decorative papers to the PB surfaces were 180 C press temperature, 2.6 MPa pressure and 25 s press time. The randomly selected panels (10 panels for each type) with nominal dimensions of 2000 mm 1100 mm 16 mm, which accounted for the full number of samples, were wrapped with polyethylene lm. The delivered panels were cut into test samples in order to measure the FE according to ASTM D 6007-02. Nine samples, representing randomly distributed portions of an entire panel, were tested in three groups of three (Fig. 1), resulting in three test results, which were averaged to represent one data point for the panel. The samples were edge-sealed with aluminum tape and conditioned at 23 C 0.5 C and 50% 5% relative humidity (RH) for 10 days prior to analysis. 2.2 Small-scale chamber (ASTM D 6007-02) The inter-laboratory experiment was performed in a 1 m3 and 0.225 m3 small-scale chamber system in laboratory 1 (Lab1) and laboratory 2 (Lab2), respectively. The dimensions of the test samples in the chambers were as follows: 3 pieces of 220 mm 220 mm and 3 pieces of 200 mm 80 mm for Lab1 and Lab2, respectively. Table 2 lists the test conditions and the differences between the two chambers. The chamber system consisted of the chamber assembly, an air control unit, and a clean air supply system. Filtered air from a clean air supply system was adjusted to a desired temperature, RH, and an air exchange rate. The controlled air was gathered in a mixing tank which was placed in the temperature-controlled climate chamber with the chamber assembly.

Fig. 1 Schematic sampling and formaldehyde emission measurements from particleboard using small-scale chamber Abb. 1 Versuchsdesign und Messung der Formaldehydabgabe einer Spanplatte in der kleinen Prfkammer

The specimens were located in the chamber so that the conditioned air stream circulated all over the panel surfaces. The chamber was operated and the temperature, RH, and barometric pressure were recorded during the testing period. The chamber test was conducted at a given Q/A ratio (Table 2). The specimens remained in the operating chamber until a steady state formaldehyde concentration was reached. 2.2.1 Quantication of formaldehyde The emission of formaldehyde was quantied according to the modied National Institute of Occupational Safety and Health (NIOSH) Test Method 3500. The method was standardized in the USA by NIOSH 3500 (1994) and has been used for the determination of formaldehyde in large-scale chambers, small-scale chambers, and when using the desiccator method. The method is summarized as follows: air was sampled from the chamber at 0.21 dm3 /min through two impinger bottles, the rst one containing 20 ml of 1% aqueous sodium bisulte (NaHSO3 ) and the second one remained empty as a blank. Portions (4 ml) of the impinger solutions were mixed with 0.1 ml of 1% chromatropic acid (1,8-dihydroxynaphthalene-3,6-disulfonic acid) solution and 6 ml concentrated H2 SO4 (96%) and left for the reaction to proceed, and then cooled. In the rst step, the chromatropic acid reacts with formaldehyde to give a redviolet hydroxydiphenylmethane derivative. In the second step of the reaction, a violet quinoid oxidation product was formed with atmospheric oxygen. The concentrated sulfuric acid is a catalyst for dehydration and oxidation (Salthammer et al. 2010). Solution absorbencies were read through a UV/vis spectrophotometer set at 580 nm. The reaction is specic for formaldehyde when the pH value is <1.0. The formaldehyde concentrations are expressed in ppm.

624 Table 2 Differences between ASTM D 6007-02 chambers and their experimental conditions Tab. 2 Unterschiede zwischen den Prfkammern gem ASTM D 6007-2 und den Versuchsbedingungen Lab Chambers V (m3 ) 1 2 1 0.225 Wall material Al Al L (m2 /m3 ) 0.43 0.43 Conditions Fan speed (m/s) (25) (25) Temp. (C) 24 3 25 1

Eur. J. Wood Prod. (2012) 70:621628

RH (%) 50 5 50 4

N (h1 ) 2 0.5016

Q/A (m/h) 4.651 1.167

Al: Aluminum Q/A = N/L = Q/V , where V : volume of closed systemthe interior volume of the test chamber (m3 ). N : air change rate (N is equal to Q/V ) A/V the ratio of conditioned and ltered air that enters or is replaced in the small chamber in one hour divided by the interior volume of the small chamber, air changes per hour (ACH). L: loading ratio (L is equal to A/V )the total exposed surface area, excluding panel edges, of the product being tested divided by the test chambers interior volume (m2 /m3 ). Q: makeup air owthe quantity of conditioned and ltered air fed into the chamber per unit time (m3 /h). N/L ratio: (N/L is equivalent to Q/A)the ratio of air ow through the chamber to sample surface area (m/h) Table 3 Precision statistics for ASTM D 6007-02 results for formaldehyde emission from the two types of particleboard Tab. 3 Ergebnisse der gem ASTM D 6007-2 bestimmten Formaldehydabgabe aus den zwei Spanplattentypen Parameter Formaldehyde value (ppm) PL Lab1 x Median SD CV % N min. max. SEM Lower quartile Upper quartile Variance 0.071 0.073 0.019 26.18 14 0.044 0.099 0.005 0.054 0.086 3.51 104 Lab2 0.093 0.095 0.008 8.79 15 0.080 0.109 0.002 0.086 0.098 6.7 105 P2 Lab1 0.170 0.171 0.04 24.23 15 0.101 0.273 0.01 0.139 0.199 1.71 103 Lab2 0.179 0.182 0.009 5.36 15 0.160 0.192 0.002 0.173 0.188 9.3 105

2.3 Statistical evaluation of results The statistical analyses of the results were carried out employing the descriptive statistical procedure, which included the calculation of means, medians, minimum and maximum values, quartiles, standard deviations (SD), standard error of mean (SEM), variance and relative coefcient of variations (CV%). The test for data normality was done using a normal quantile plot. Identication of outliers between quartiles, dened as the values exceeding 1.5 times the interval between quartiles, i.e., the interval between the 75% value and the 25% value, was also made. Moreover, analysis of variance (ANOVA) was done to investigate the signicant effect of the laboratory and sample on FE values. The comparison between means was done employing a Duncans multiplerange test at 0.05 level of probability. Moreover, the correlation was done between the FE values resulting from the two laboratories. All analyses were performed using SAS version 8.2 (2001).

3 Results and discussion An evaluation of the differences between the values of FE observed from the two types of PB; PL and P2 measured in two laboratories is presented in Table 3. The summary of statistics for the FE shows average values of observed parameters between both laboratories. These data are presented graphically in Figs. 24. Figure 2 shows the variation between the average of three measurements for all samples except for sample 5c, which was measured only by Lab2; the other measurement from Lab1 was not available. The relative coefcient of variation for the FE values from PL was 26.18% and 8.79%, and for P2 24.23% and 5.36%, for Lab1 and Lab2, respectively. Furthermore, the outliers of the measured FE values presented in Fig. 3 from the two laboratories as in PL and P2 showed that there were differences between the values from the two laboratories, which indicated that Lab1 had a high number of data numerically distant from the rest of the data

Eur. J. Wood Prod. (2012) 70:621628

625

compared to Lab2. The normal quantile plot shown in Fig. 4 indicates that the distribution of FE values measured from Lab2 had a higher homogeneity than those from Lab1. The ANOVA results in Table 4 for PL and P2, show that the differences between samples were signicant for both types (P < 0.05). Differences between laboratories were highly signicant for PL (P < 0.001) and not signicant for P2 (P = 0.33). However, the interactions between the

Fig. 2 Concentration of formaldehyde emissions from particleboard (16 mm) measured in the two laboratories Abb. 2 In den beiden Prfstellen gemessene Formaldehydabgabe aus den beiden Plattentypen (16 mm)

samples and laboratories for the two types of PB were not signicant. The variability between the values of FE due to sample heterogeneity is presented in Table 5. The FE values from PL ranged between 0.056 and 0.090 and from 0.088 to 0.099 ppm for PL and from 0.129 to 0.223 and 0.175 to 0.182 ppm for P2 as measured in Lab1 and Lab2, respectively. Moreover, it was seen that the sample number 5 PL had a high amount of FE measured in both laboratories, also there were no differences between the values of FE from the PL samples measured in Lab2. On the other hand, there were differences between FE values for the same samples when measured in Lab1. In addition, the same measurement trends were found between the FE values from P2 samples measured in both laboratories, even though there were no significant differences between the two laboratories (P = 0.33). Thus, these results suggested that there was a higher homogeneity between the samples in Lab2 than in Lab1, and the most important source of variation is inter-laboratory bias due to heterogeneity of chamber conditions, and probably sampling air as well (Wiglusz et al. 2000; RisholmSundman et al. 2007). In addition, inter-laboratory studies provide a better understanding of the behavior measurements of the FE from the same samples of PB. Considering that the samples received in both laboratories were distributed and selected randomly throughout the PB bundle, which makes it unlikely that the observed variations were due to inter-panel variability, there were some variations in the formaldehyde amount through the tested panel, for example; the panel PL number 1 observed differences in the samples 1a, 1b and 1c (0.065, 0.06 and 0.044 ppm, respectively as measured in Lab1) and (0.08, 0.092 and 0.093 ppm, respectively as measured in Lab2) as shown in Fig. 2. In addition, there were some variations in the average formaldehyde concentrations among the boards, (Table 5) for example the PL panels (0.056, 0.090, 0.062,

Fig. 3 BoxWhiskers plot of formaldehyde concentrations variation from particleboard (16 mm) between the two laboratories Abb. 3 Unterschiede zwischen der in beiden Prfstellen gemessenen Formaldehydabgabe aus den Spanplatten

626 Fig. 4 Normal quantile plot of small-scale chambers data Abb. 4 Normal-Quantil-Plot der in den kleinen Prfkammern gemessenen Werte

Eur. J. Wood Prod. (2012) 70:621628

Table 4 Signicance levels for the effects of sample and laboratory on the overall ASTM D 6007-02 values (ppm) Tab. 4 Signikanzniveaus fr den Einuss der Probennahme und der Prfstellen auf die nach ASTM D 6007-2 bestimmten Werte (ppm) Type PL Source of variation Laboratory (Lab) Sample Lab Sample P2 Lab Sample Lab Sample
ns P

F value 22.14*** 3.40* 2.32ns 0.99ns 3.32* 2.78ns

*** P

Pr. > F 0.0002 0.03 0.09 0.33 0.03 0.055

Whole R 2 0.71

Whole CV % 14

0.56

13.62

> 0.05 (not signicant);

*P

< 0.05 (signicant);

< 0.001 (highly signicant)

0.065 and 0.090 ppm, for panels 1, 2, 3, 4 and 5, respectively as measured in Lab1). The obvious results of high, intermediate or low emission variations suggested that there were inter-panel variations in the formaldehyde release from PB, and these results were in agreement with previous investigations (Roffael et al. 1979). It should be noted that the results from this interlaboratory comparison are intended to identify TPCs that may need to work with CARB staff to enhance the precision or accuracy of their chamber testing (CARB 2010). In addition, for Lab1 which had outlier points, CARB recommends that this laboratory should calibrate its chamber as described in the recommendations of CARB, and CARB will follow up with this laboratory as necessary (CARB 2010). Despite these differences between formaldehyde values in Lab1 and Lab2, a strongly positive acceptable linear correlation between the formaldehyde values from the two types of particleboards with 16 mm thickness had an R 2

value of 0.84 with two measurements outside the 95% condence intervals (Fig. 5). In Fig. 3, the mean values of FE measured from PL in Lab1 (0.072 ppm) were lower than in Lab2 (0.093 ppm) and correspondingly, there were signicant differences between the means, while there were no signicant differences between the means of the formaldehyde emitted from P2 in Lab1 (0.171 ppm) and Lab2 (0.179 ppm). In addition, the emission of formaldehyde from particleboards was dramatically reduced due to the lamination of the boards with melamine-impregnated paper (Fig. 3). This result was in agreement with previous results obtained in many studies relating to the effect of coating of composite wood products with different coatings like decorative vinyl lm and melamine-impregnated paper on the FE values (Grigoriou 1987; Kim et al. 2010; Nemli and ztrk 2006; Kim 2010; Salem et al. 2011). Simultaneously, the formaldehyde concentration values from most of the PL samples measured in Lab1 and Lab2

Eur. J. Wood Prod. (2012) 70:621628

627

Table 5 Differences between formaldehyde emission values of two types of particleboard measured by ASTM D 6007-02 (ppm) in the two laboratories Tab. 5 Unterschiede zwischen den Formaldehydabgabewerten der beiden Spanplattentypen, die gem ASTM D 6007-2 (ppm) in den beiden Prfstellen gemessen wurden Sample PL* Lab1 1 2 3 4 5 Mean of lab* Sample 0.056 0.01b 0.090 0.007a 0.062 0.02b 0.065 0.01b 0.090 0.01a 0.072 0.02B P2* Lab1 6 7 8 9 10 Mean of lab* 0.129 0.009c 0.174 0.03b 0.163 0.05bc 0.165 0.02bc 0.223 0.02a 0.171 0.04A Lab2 0.175 0.01b 0.177 0.02b 0.182 0.008ab 0.182 0.007ab 0.181 0.009ab 0.179 0.009A 0.152 0.03b 0.175 0.03ab 0.172 0.04ab 0.174 0.02ab 0.201 0.03a Lab2 0.088 0.007a 0.090 0.009a 0.093 0.005a 0.095 0.009a 0.099 0.01a 0.093 0.008A Sample mean* 0.072 0.02c 0.090 0.007ab 0.078 0.02bc 0.080 0.02bc 0.095 0.01a Sample mean*

Values (mean SD). Different letters represent statistical differences between the averages of the values
* Means with the same small letter within the same column are not signicantly different (P < 0.05). Means with the same capital letter within the same row are not signicantly different (P < 0.05)

Fig. 5 Correlation between the formaldehyde values for 16 mm particleboards measured in the two laboratories Abb. 5 Korrelation zwischen den in den beiden Prfstellen gemessenen Formaldehydabgabewerten der 16 mm Spanplatten

ranged from 0.056 to 0.099 ppm, were in agreement with the CARB regulation for PB Phase 2 limit (0.09 ppm). Furthermore, most of the FE values from P2 ranged from 0.129 to 0.182 and these values met the requirements of the CARB Phase 1 limit (0.18 ppm), except for sample 10 P2 which had a high amount of formaldehyde (0.223 ppm). The CARB Phase 1 was roughly equivalent to or greater than the European E1-emission class (0.1 ppm), as was the Japanese F class (1.5 mg/L). Phase 2 (0.09 ppm) limits

were comparable to the Japanese F limits (0.07 ppm), the so-called E0 levels in Europe. 4 Conclusion The inter-laboratory results showed signicant variations between the two laboratories, most of them caused by sample heterogeneity, and the inconsistency of the values was related to the chamber conditions and sampling. These re-

628

Eur. J. Wood Prod. (2012) 70:621628 JIS A1460 (2001) Building boards. Determination of formaldehyde emissiondesiccator method. Japanese Industrial Standard Kim S (2010) Control of formaldehyde and TVOC emission from wood-based ooring composites at various manufacturing processes by surface nishing. J Hazard Mater 176(13):1419 Kim K-W, Kim S, Kim H-J, Park JC (2010) Formaldehyde and TVOC emission behaviors according to nishing treatment with surface materials using 20 L chamber and FLEC. J Hazard Mater 177(1 3):9094 Nemli G, ztrk (2006) Inuences of some factors on the formaldeI hyde content of particleboard. Build Environ 41(6):770774 NIOSH 3500 (1994) Formaldehyde by VIS; manual of analytical methods. 4th ed. National Institute of Occupational Safety and Health: Washington, DC Pickrell JA Grifs LC Mokler BV et al (1986) Formaldehyde release rate coefcients from selected consumer products. In: Meyer, B, Andrews, B, Reinhardt, RM (eds), Formaldehyde release from wood products. American Chemical Society Symposium Series No. 385, Washington, DC, pp 4051 Park B-D, Kang E-C, Park S-B, Park JY (2011) Empirical correlations between test methods of measuring formaldehyde emission of plywood, particleboard and medium density berboard. Eur J Wood Prod 69:311316 Que Z, Furuno T (2007) Formaldehyde emission from wood products: relationship between the values by the chamber method and those by the desiccator test. Wood Sci Technol 41:267279 Que Z, Furuno T, Katoh S, Nishino Y (2007) Evaluation of three test methods in determination of formaldehyde emission from particleboard bonded with different mole ratio in the ureaformaldehyde resin. Build Environ 42(3):12421249 Risholm-Sundman M (1999) Determination of formaldehyde emission with eld and laboratory emission cell (FLEC)recovery and correlation to the chamber method. Indoor Air 9:268272 Risholm-Sundman M, Wallin N (1999) Comparison of different laboratory methods for determining the formaldehyde emission from three-layer parquet oors. Holz Roh- Werkst 57:319324 Risholm-Sundman M, Larsen A, Vestin E, Weibull A (2007) Formaldehyde emissioncomparison of different standard methods. Atmos Environ 41(5):31933202 Roffael E, Greubel D, Mehljorn L (1979) ber die Bestimmung der Formaldehydabgabe von Spanplatten nach dem PerforatorVerfahren und nach der WKI-Methode. Holz-Zentbl 104(24):396 397 Salem MZM, Bhm M, Bernkov J, Srba J (2011) Effect of some manufacturing variables on formaldehyde release from particleboard: relationship between different test methods. Build Environ 46(10):19461953 Salthammer T, Mentese S, Marutzky R (2010) Formaldehyde in the indoor environment. Chem Rev 110(4):25362572 SAS (2001) Users Guide: Statistics (Release 8.02). SAS Inst. Inc., Cary, NC, USA Sundin EB, Mansson B, Endrody E (1987) Particleboard with different contents of releasable formaldehyde: a comparison of the board properties including results from four different formaldehyde tests. In: Proc of the 21st international particleboard/composite materials symposium. Washington State University, Pullman, pp 139186 Wiglusz R, Jarnuszkiewicz I, Sitko E, Nikel G (2000) Interlaboratory comparison experiment on the determination of formaldehyde emitted from mineral wood board using small test chambers. Build Environ 35(1):5357 Yrieix C, Dulaurent A, Laffargue C, Maupetit F, Pacary T, Uhde E (2010) Characterization of VOC and formaldehyde emissions from a wood based panel: results from an inter-laboratory comparison. Chemosphere 79:414419

sults point to the necessity of having identical experimental conditions in the chambers. Moreover, there was a good correlation (R 2 = 0.84) between the formaldehyde values for both types of particleboards as measured according to the method of ASTM D 6007-02 in the two laboratories. The reduction in the emission of formaldehyde from particleboards E1-16 mm measured in both laboratories resulted from coating the boards with melamine-impregnated decorative papers. Furthermore, the comparison showed that laboratory 2 gave better accuracy than laboratory 1 and suggested that laboratory 2 has sufcient experience in formaldehyde emission analysis. As a general rule for further studies in laboratories deemed to be outliers, or whose results were identied as a concern, CARB recommends that these laboratories calibrate their small chamber in accordance with the large chamber. In this way the inter-laboratory comparisons are now very important for determination of formaldehyde emission, in addition regulations and trends need to become clearer within Central and Eastern Europe.
Acknowledgements This work was nancially supported by grants from the Internal Grant Agency, Faculty of Forestry and Wood Sciences, Czech University of Life Sciences, Prague. The authors acknowledge the contribution of the participating laboratories.

References
ASTM E 1333-96 (2002) Standard test method for determining formaldehyde concentrations in air and emission rates from wood products using a large chamber. American Society for Testing and Materials, West Conshohocken, PA, p 12 ASTM D 6007-02 (2002) Standard test method for determining formaldehyde concentration in air from wood products using a small-scale chamber. American Society for Testing and Materials, West Conshohocken, PA, p 8 Bulian F, Battaglia R, Ciroi S (2003) Formaldehyde emission from wood based panels: the importance of intercalibrating the test methods. Holz Roh- Werkst 61:213215 CARB (2007) Proposed airborne toxic control measure (ATCM) to reduce formaldehyde emissions from composite wood products. http://www.arb.ca.gov/toxics/compwood/compwood.htm CARB (2010) Results of international interlaboratory comparison of composite wood product third party certiers (TPC). State of California, California Environmental Protection Agency, Air Resources Board EN 120 (1993) Wood-based panelsdetermination of formaldehyde contentextraction method called perforator method. European Standard EN 717-3 (1996) Wood-based panelsdetermination of formaldehyde release. Part 3. Formaldehyde release by the ask method. European Standard EN 717-1 (2004) Wood-based panelsdetermination of formaldehyde release. Part 1. Formaldehyde emission by the chamber method. European Standard Grigoriou A (1987) Formaldehyde emission from the edges and faces of various wood based materials. Holz Roh- Werkst 45(2):6367 IARC (International Agency for Research on Cancer) (2004). Overall evaluations on carcinogenicity to humans. IARC Monographs, Vol. 1. International Agency for Research on Cancer, Lyon, France