Eur. J. Wood Prod. (2012) 70:679–688 DOI 10.1007/s00107-012-0595-x
O R I G I NA L S O R I G I NA L A R B E I T E N
Influence of ageing on mechanical properties of wood to wood bonding with wheat flour glue Stefano D’Amico · Marta Hrabalova · Ulrich Müller · Emmerich Berghofer
Received: 22 August 2011 / Published online: 25 February 2012 © Springer-Verlag 2012
Abstract Earlier research into native wheat flour for wood to wood bonding showed excellent bonding properties comparable to synthetic adhesives, but no data about ageing behaviour is available. Short and long term effects on mechanical properties were analysed by lap joint testing and modified DCB-specimens. Results showed no significant reduction in bonding properties, but a trend to lower adhesive strength after 12 months of storage was noticeable. Changes in wheat polymers were observed by means of DSC and FTIR-ATR. Soluble degradation products of starch were analysed by GC-FID after methanolysis and derivatisation. FTIR measurements indicated changes in the structure of starch, but no appreciable alteration of proteins. Investigations by DSC showed increasing crystallinity durS. D’Amico () · U. Müller Competence Center for Wood Composites and Wood Chemistry (Wood K plus), Altenberger Strasse 69, 4040 Linz, Austria e-mail:
[email protected] S. D’Amico Competence Center for Wood Composites and Wood Chemistry (Wood K plus), St. Peter Straße 25, 4021 Linz, Austria S. D’Amico Universität für Bodenkultur, UFT—Universitäts- u. Forschungszentrum Tulln, Konrad Lorenzstrasse 24, 3430 Tulln, Austria M. Hrabalova Institute for Natural Materials Technology, Department of Agrobiotechnology, IFA-Tulln, University of Natural Resources and Applied Life Sciences, Konrad Lorenz Strasse 20, 3430 Tulln, Austria E. Berghofer Institute of Food Technology, Department of Food Science and Technology, University of Natural Resources and Applied Life Sciences, Muthgasse 18, 1190 Vienna, Austria
ing 3 months of storage. After 6 months more degradation products were detected. Results indicated that hydrolysis of starch is responsible for a moderate decrease of bonding performance; wheat proteins seem to be less affected. Einfluss der Alterung auf die mechanischen Eigenschaften von Weizenmehlleim für die Verklebung von Holz Zusammenfassung Frühere Untersuchungen mit unbehandeltem Weizenmehl für die Verklebung von Holz zeigten sehr gute Festigkeiten vergleichbar mit synthetischen Klebstoffen, aber es sind keine Informationen über das Alterungsverhalten bekannt. Kurz- und Langzeiteffekte wurden mittels Längszugscher- und modifzierten DCB-Proben untersucht. Die Ergebnisse zeigten keine signifikante Verschlechterung der Verklebungsfestigkeit nach 12 Monaten, aber eine Tendenz zu geringeren Festigkeiten war erkennbar. Veränderungen in den Polymeren im Weizenmehl wurden mit Hilfe von DSC und FTIR analysiert. Die FTIR Messungen zeigten Strukturveränderungen der Stärke, aber keine sichtlichen Veränderungen der Proteine. Untersuchungen mittels DSC ergaben, dass die Kristallinität der Stärke während den ersten drei Monaten anstieg. Nach 6 Monaten wurden immer mehr Abbauprodukte der Stärke gemessen. Die Ergebnisse lassen den Rückschluss zu, dass die Hydrolyse der Stärke für den moderaten Verlust der Festigkeiten verantwortlich ist, während die Proteine weitaus weniger von der Lagerung beeinflusst werden. Abbreviations DCB double cantilever beam DSC differential scanning calorimetry FF furfural HMF 5-(Hydroxymethyl)furfural
680
GC FID ATR FTIR PCA
Eur. J. Wood Prod. (2012) 70:679–688
gas chromatography flame ionization detector attenuated total reflectance Fourier transform infrared spectroscopy principal component analysis
1 Introduction The rising price of petroleum and environmental concerns about some current wood adhesives, e.g., reducing VOC’s (volatile organic compounds) and toxic cross-linkers (formaldehyde, isocyanates etc.) has led to increasing interest in renewable polymeric materials (Zhang et al. 2010). Several regulations within the European Union were issued to limit the negative impact of formaldehyde (Däumling et al. 2005). Although bio-based adhesives have been used for thousands of years, they play only a very minor role in current wood-bonding markets. The most important application area is the usage as cheap filler in synthetic resins (Dix 1987; Plath 1972). Only little research was done to develop starch glues for wood bonding, mainly in combination with other natural or synthetic raw materials (Imam et al. 1999; Moubarik et al. 2010; Tondi et al. 2011). Of the bio-based feeds, wheat flour is of great interest because of its large production volume and low cost. Compared to isolated starches no costs for extracting and processing arise which can limit its profitability. Furthermore wheat flour contains proteins (wheat gluten), which are responsible for the unique viscoelastic properties of wheat based materials due their flexible cross-links and fibrous structure. Wheat gluten consists of two protein fractions, gliadins and glutelins, which can build up a three dimensional network with a very large size. Previous results have revealed a big potential for wheat flour as bio-based adhesive. Only by controlling the press temperature outstanding bonding properties were gained for wheat flour glue without any modifications. Under optimal conditions tensile shear strength according to EN 302-1 (2004) of over 6 MPa and nearly complete wood failure were reached. The performance of the wheat flour adhesive is similar to that of other synthetic resins and glues for spruce wood in a dry environment (D’Amico et al. 2010; Konnerth et al. 2006; Veigel et al. 2011). The excellent bonding properties and extreme low price of wheat flour makes it very interesting for many indoor applications. Furthermore, no harmful or toxic compounds were emitted to the environment. In general, biological materials are much more affected by humidity and degradation processes than synthetic ones (Ramis et al. 2004). Especially natural polymers are susceptible to biodegradation accelerated by enzymes and low or high pH-levels because of their hydrolysable cross-links. Speed of decomposition is related to degree of polarity,
therefore hydrophobic materials are more inert (Zhang et al. 2010). Also in grain-based products (e.g., bread, cereals) ageing behaviour, called staling, is of great interest. Staling is a very complex process, mainly related to recrystallisation of starch. Due to retrogradation mechanical properties are affected; usually they become firmer and more brittle (Baik and Chinachoti 2000; Kizil et al. 2002). Besides time and temperature, also water plays a critical role in this process. During retrogradation water evaporates and the amount of adsorbed water is diminished (Lionetto et al. 2005; Zhou et al. 2009). For analysis of staling, several techniques were developed. Thermal and spectroscopic methods like DSC and IR-spectroscopy belong to the most important ones (Baik and Chinachoti 2000; Cocchi et al. 2005; Ji et al. 2007; Lionetto et al. 2005). Although a lot of research was done to characterize ageing behaviour of wheat based foods, little known is known about wheat based materials for technical applications. Some studies of ageing behaviour of pure starch films were made (Kuutti et al. 1998; Viguié et al. 2007), but no data in respect to wood bonding is available. The aim of this work was to analyse the influence of ageing on the bonding capacity of native wheat flour in respect to wood to wood bonding. Furthermore, the research was focussed on determining reasons on a molecular basis which affect bonding properties.
2 Materials and methods 2.1 Materials All reagents were of at least analytical grade and from Sigma–Aldrich (Steinheim, Germany) unless otherwise specified. Commercial wheat flour type 405 was obtained from VonWiller (Schwechat, Austria) with the following non-starch contents: water 12.34 ± 0.09, proteins 12.42 ± 0.15 and ash 0.38 ± 0.03 (w/w % of flour mass). The moisture content was determined by the mass-loss after 48 hours in a drying chamber at 103°C, while the protein content was determined according to the Kjehldal method (N × 5.68). 2.2 Sample preparation Lap joint specimens for the determination of the tensile shear strength of the bond line were produced according to the European Standard EN 302-1 (2004). Sample preparation was performed with a wheat flour suspension as adhesive according to D’Amico et al. (2010). The samples were stored for different periods, 14 days, 1, 2, 3, 4, 5, 6, 8, 9, 11 and 12 months at 20 ± 2°C and 65% relative humidity. Shear strength of lap joint testing was used predominantly to evaluate short term effects.
Eur. J. Wood Prod. (2012) 70:679–688
Fig. 1 Geometry of modified DCB-specimens. All dimensions in mm Abb. 1 Geometrie der modifizierten DCB-Prüfkörper. Alle Abmessungen in mm
Additionally modified double cantilever beam-(DCB) specimens (Fig. 1) were produced to analyse changes in the bonding strength for long period storage. By this testing method, the specific fracture energy of the bonding was detected. Samples with a length of 60 mm and a width of 20 mm were prepared out of the glued joints. The samples were fixed with a melamine-formaldehyde-resin, Prefere 4535 and 20% hardener Prefere 5046 (Dynea, Austria), on beech-ledgers with dimensions of 20 × 16 × 80 mm3 (width × thickness × length). To initialise the breakage in the bonding line a kerf of 5±0.5 mm was cut. The remaining length of the bond-line was 55 ± 1 mm. To attach the test grips, holes with a diameter of 6.0 mm were drilled about 10 mm away from the end of the specimens. 2.3 Mechanical testing For mechanical testing a universal testing machine (Zwick/ Roell) with a 20 kN load cell and an average speed of 0.5 mm/min was used to attain breakage of lap joints within 60 ± 30 s. The percentage of wood failure was evaluated visually. The testing of modified DCB specimens was performed according to Veigel et al. (2011). An initial load with a speed of 1 mm/min was put on the samples until maximum load was reached. After a decrease of 50% in the load, testing speed was increased progressively to 10 mm/min. The fracture energy was calculated from the load-displacement curve by integrating the load in relation to the displacement. Afterwards the energy was purchased to the bonding area in order to calculate the specific fracture energy (J/m2 ). The generated data were analysed statistically with one way ANOVA (analysis of variance) at P < 0.05. 2.4 Spectroscopic, thermal and chemical analysis For analytical measurements, a nut with a depth of 2 mm ± 0.5 mm was cut into small spruce boards with an area
681
of about 10 cm2 . The boards with nuts were glued together under equal conditions (concentration of wheat flour glue, press time, temperature, pressure). Notwithstanding the preparation of samples for mechanical testing an excess of flour suspensions was put on the boards to gain an area with enriched adhesive. After same storage periods and conditions analogue samples for mechanical testing (see 2.1 Sample preparation) of the cured glue was prepared out of the nut. The isolated glue was milled with a vibration ball mill (Perkin-Elmer, Germany) for two minutes, dried at 40°C in vacuum for 24 hours and used afterwards for chemical analysis. To avoid higher temperatures during milling a short break was taken after 60 s of milling. Attenuated total reflectance Fourier-transform Infrared Spectroscopy (ATR-FTIR) was performed to analyse changes in the chemical structure. The measurements were carried out with a Helios FT-IR-microscope (Bruker Optik GmbH, Germany) equipped with a diamond ATR which was connected to a Tensor 27 IR-spectrometer (Bruker Optik GmbH, Germany). The samples were set directly to the diamond crystal. Scanning was conducted from 4000 to 600 cm−1 with a resolution of 2 cm−1 and 128 scans for each spectrum. At least six spectra were taken for each sample. Post-spectroscopic manipulation between 1800 to 600 cm−1 (CO2 compensation, vector standardisation, ground line correction) was performed with OPUS software program (Version 6.5). The gained data were analysed qualitatively with PCA (Principal Component Analysis). Thermal properties of the wheat flour slurry were analysed by means of differential scanning calorimeter (DSC 200 F3, Netzsch, Germany). About to 10 mg of the flour 20 µl of water was added, hence the corresponding rate of suspensions used for wood joint bonding. The measurements were carried out in medium pressure crucibles at a temperature range of 10–150°C and a heating rate of 10°C/min. A second heating scan was performed to control if complete gelatinization or retrogradation was accomplished. An empty pan was used as reference. The measurements were done in triplicate. It is well known that the Hm of amylopectin corresponds to the material crystallinity. The degree of retrogradation was calculated according to Liu et al. (2010): g
Degree of retrogradation (%) = (Hmr /Hm ) · 100 where Hmr is the melting enthalpy of wheat flour after retg rogradation and Hm is the value of melting enthalpy obtained from initial gelatinisation of the native wheat flour. For analysis of degradation products about 10 mg of the prepared glue were extracted with 1 ml deionised water for 1 hour at 30°C during shaking. The co-extracted proteins were precipitated with Carrez I (Solution of potassium hexacyanoferrate (II) in water with a concentration of
682
Eur. J. Wood Prod. (2012) 70:679–688
150 g/L) and Carrez II (zinc sulfate in water with a concentration of 300 g/L) solutions and the supernatant desiccated in vacuum at 40°C over night after centrifugation (5300 U/min and 5 min). The dry extract was hydrolysed with 1 ml 2 molar hydrochloric acid in methanol for 5 hours at 103°C. After cooling 1 ml sorbitol-solution (0.1 g/l) was added as internal standard. Afterwards 1 ml of each sample was dried in vacuum at 40°C. The dried methanolysed extracts were sylilated for GC (gas chromatography) analysis with 150 µL HMDS (1,1,1,3,3,3-hexamethyldisilazane) and 70 µL TMCS (trimethylchlorosilane) in 200 µL pyridine for 5 hours at 25°C. Gas chromatography (GC) was performed on an Agilent Technologies 6890N GC instrument equipped with a flame ionization detector (FID). Samples were separated on a HP5-MS (30 m × 250 µm × 0.25 µm) column. The temperature program of the column oven was 150°C (2 min), 10°C min−1 –280°C, 20°C min−1 –300◦ C (3 min). Sorbitol was used as internal standard to quantify amounts of monosaccharides. Due to the different chemical structure the amount of FF and HMF was calculated by means of a straight calibration line (R 2 = 0.989). All measurements were done in triplicate.
3 Results 3.1 Mechanical properties In Fig. 2a box plots of shear strength and in Fig. 2b average amount of wood failure of lap-joints after different periods of ageing are shown. Lap joint testing was primarily used to evaluate short term effects reflected in the bigger quantity of samples for ageing until 6 months. Values for shear strength of about 6 MPa confirmed previous results. Anymore samples aged for one month are showing the highest and older samples (11 and more months) the lowest shear strength. All in all, no significant differences in the bonding strength were detected, only between one and 11 resp. 12 months of storage time. But a trend to lower bonding properties can be seen with increasing time of storage. The amount of wood failure showed similar findings. Until 9 months rate of wood failure remained on a high level and decreased not until 11 months curtly below 50%. Also here a trend to less bonding performance after 11 resp. 12 months was found. But the high bonding performance of samples stored for one month was not affirmed. The amount of wood failure is even lower compared to samples stored for 14 days. This indicates that density of the used wood boards was higher, and therefore improved shear strength was measured. However, maximum shear strength of lap joints is also influenced by wood quality and sample geometry (D’Amico et al. 2010; Konnerth et al. 2006). This effect should be minimized by
Fig. 2 (a) Comparison of shear strength and (b) average amount of wood failure of lap joints tested according to EN 302-1 (2004) after different storage periods (the samples tested after 14 days of storage were set to zero months). Circles are displaying outliners and bars represent standard deviation Abb. 2 Vergleich (a) der Festigkeiten und (b) des durchschnittlichen Holzbruchanteils der Längszugscherproben gemäß EN 302-1 (2004) nach verschiedenen Lagerungsintervallen (die Proben nach 14 Tagen Lagerung wurden auf Null gesetzt). Die Kreise zeigen Ausreißer an und die Balken stehen für die Standardabweichungen
preparing all boards from the same piece of wood as mentioned in the European Standard EN 302-1 (2004). Anyway little variations in the density cannot be avoided.
Eur. J. Wood Prod. (2012) 70:679–688
Effect of long term storage was additionally evaluated by means of fracture energy testing. The same glued boards were used to fabricate jointed samples. These samples were fixed on beech-ledgers with a melamine-formaldehyde resin. Due to the low flexibility of this adhesive influence on testing the fracture energy should be minimized. In Fig. 3, fracture energy of modified DCB-specimens calculated by means of the area method is shown. The area method was chosen because of the higher explanatory power and facile
683
calculation scheme (Veigel et al. 2011). The determined fracture energies of 200–300 J/m2 confirm the good bonding properties of the wheat flour glue and are comparable to relatively brittle synthetic resins (Veigel et al. 2011). Here as well loss of bonding strength was not significant, but the drift to lower mechanical properties was confirmed. In contrast to the shear strength of lap joints, no significant difference between one and 12 months of storage was detected. 3.2 Spectroscopic, thermal and chemical analysis
Fig. 3 Comparison of specific fracture energies after one and 12 months of ageing. Circles are displaying outliners Abb. 3 Vergleich der spezifischen Bruchenergie nach einem und zwölf Monaten Lagerung. Die Kreise zeigen Ausreißer an
Infrared spectroscopy was used to study changes in the physicochemical structure of aged wheat flour polymers, primarily starch and proteins. By means of infrared spectroscopy it is possible to observe the retrogradation of stored starch materials by changes in peak shapes and intensities of their characteristic peaks in the range between 1150 and 900 cm−1 (Cocchi et al. 2005; Sevenou et al. 2002). Furthermore, structural information on wheat gluten can be obtained. Especially the amide I band at about 1650 cm−1 is related to protein conformation (Li et al. 2006). In Fig. 4(a) the mean spectra of samples with different age and in (b) the loadings of PCA are illustrated. Loadings show only little variations in the range of amide I band indicating that proteins are marginally affected during storage. Main changes were observed in the region between 1100 and 950 cm−1 , which are related to starch. First rise of intensity in the peak at 996 cm−1 and the flattening of peak at 1019 cm−1 indicate increasing crystallinity (Cocchi et al. 2005). Until 3 months crystallinity strongly increased, afterwards only little variations of the mentioned areas were detected. Due to other processes occurring during ageing, e.g., changes in water content by evaporation, hydrolysis of starch and formation of degradation products, it was not possible to calculate the rate
Fig. 4 (a) Mean spectra and (b) loadings plot 1 & 2 of manipulated signals between 1750 and 400 cm−1 calculated for each storage period (14 days, 1, 3, 6 and 12 months) Abb. 4 (a) Gemittelte Spektren und (b) Loading Plot 1 & 2 der bearbeiteten Spektren zwischen 1750 and 400 cm−1 berechnet für das jeweilige Alterungsintervall (14 Tage, 1, 3, 6 und 12 Monate)
684
Eur. J. Wood Prod. (2012) 70:679–688
Fig. 5 Extent of retrogradation calculated from DSC data by the relative change in melting enthalpy stored for 14 days, one and three months. The melting enthalpy of native wheat flour was used as basis for entire crystallinity. All measurements were done in triplicate and the bars represent standard deviation Abb. 5 Ausmaß der Retrogradation basierend auf den DSC Daten durch Berechnung der relativen Schmelzenthalpie nach einer Lagerung von 14 Tagen, einem und drei Monaten. Die Schmelzenthalpie des nativen Weizenmehls wurde als Grundlage für die ursprüngliche Kristallinität verwendet. Alle Messungen wurden dreimal wiederholt und die Balken entsprechen der Standardabweichung
of retrogradation exactly. Especially the water content has an influence on the degree of retrogradation (Livings et al. 1997). CO and CC stretching and COH bending vibrations for crystalline as well as for amorphous starch are affected by water (van Soest et al. 1995). Therefore, FTIR spectra were only used to demonstrate the approximate progress of ageing. Retrogradation was observed in more detail by DSC. The used conditions (water content and temperature) should ensure complete gelatinisation. Above 65% water in the sample all granules absorb water without limitation and only one endothermic peak of melting occurs (D’Amico et al. 2010). Additionally, a second heating cycle was performed to control if gelatinisation or retrogradation was finished. But no remarkable melting enthalpies were detected during the 2nd heating cycle. The gelatinization enthalpy of native wheat flour was used as absolute amount of origin crystallinity. Due to the high enthalpies after 3 months of ageing in the range of native wheat flour no more measurements were performed. Results of DSC measurements show increasing crystallinity during ageing (Fig. 5). First, enthalpies rose rapidly within one month and then slowed down with longer storage time. After three months a crystalline amount like in the native wheat flour was determined. This indicates that a maximum of retrogradation was accomplished. Analysis of water-soluble degradation products was performed by means of GC with FID after methanolysis and sylilation. The chromatograms for native wheat flour and after respective storage periods are presented in Fig. 6(a)– (c). Due to the plurality of hydroxyl-groups of sugars and the way of sample preparation more than one peak arises for each monosaccharide (Reimerdes and Rothkitt 1984). In respect of glucose three different peaks can be seen. Additionally, products of decomposition can be detected, e.g.
sugar acids, FF and HMF. However it was not possible to separate FF and HMF because of the similarity of these molecules and the temperature program used, which was established for sugar and polysaccharide analysis. Therefore, the sum of FF and HMF is quoted. In none of the samples prepared from the glue line glucoronic acid could be detected. Development of starch fragments (in glucose equivalents) and FF/HMF content during ageing are shown in Fig. 7. In the native wheat flour (type 405) as well as in the heated glue samples only small amounts of pentoses were detected, which occur in native wheat flour. Contrary to grain flours with very low fineness degree, wood contains high amounts of pentoses in the form of hemicelluloses. Therefore, the presence of wooden particles in the prepared glue can be excluded. Acetic acid was not detectable because of its extremely volatile character, and therefore probably lost during sample preparation. In 14 days stored samples no FF or HMF was measured. After one month marginal contents arise and after 6 months of storage time the amount of FF/HMF increased strongly. However the content of glucose equivalents remained constant. After one year, enhanced hydrolysis of starch was ascertained by the significant increase of water extractable starch snatchings and therefore higher glucose content. Simultaneously, also amounts of FF/HMF amplified.
4 Discussion With lap joint and modified DCB-specimen testing it was possible to evaluate the bonding power during ageing. Furthermore, both methods confirmed earlier results and indi-
Eur. J. Wood Prod. (2012) 70:679–688
685
Fig. 6 Chromatograms of water-soluble degradation products of wheat flour glue after different ageing intervals (a: native; b: one month; c: 12 months). Samples were analysed by means of GC-FID after methanolysis and sylilation Abb. 6 Chromatogramme der wasserlöslichen Abbauprodukte des Weizenmehlleims nach verschiedenen Lagerungszeiten (a: nativ; b: ein Monat; c: 12 Monate). Die Proben wurden mittels GC-FID nach Methanolyse und Sylilierung analysiert
Fig. 7 Quantity of water-extractable degradation products of starch (in glucose equivalents) and sum of FF and HMF during ageing of 14 days, 1, 3, 6 and 12 months. Amounts below detection limit were set to zero. All measurements were done in triplicate and the bars represent standard deviation Abb. 7 Menge der wasserlöslichen Abbauprodukte der Stärke (in Glukose Äquivalenten) und Summe an FF und HMF während der Lagerung über 14 Tage, 1, 3, 6 und 12 Monaten. Konzentrationen unterhalb der Nachweisgrenze wurden auf Null gesetzt. Alle Messungen wurden dreimal wiederholt und die Balken entsprechen der Standardabweichung
cated that the used adhesive ensures almost stable properties over a period of one year in a moderate climate. Only a trend to lower bonding strength was observed. Values of lap joints showed little variations because of the influence of wood. Therefore, it is necessary to investigate concurrently the amount of wood failure. Especially when good adhesion and therefore high rates of wood failure are gained the validity of lap joint testing is limited (D’Amico et al. 2010; Konnerth et al. 2006; Veigel et al. 2011). Although wood with vertical annual growth rings instead of a grain angle of 3 ± 1◦ was used for modified DCB-specimen, no considerable amount of wood failure
was noted. Only one specimen showed wood failure and was therefore excluded in the results. However, testing of bonding performance by means of modified DCB-specimens seems to be superior. Compared to lap joint testing a bigger bonding area was examined and influence of wood did not affect the bonding capacity detected. By means of this test method, the variation in wood mechanical properties is left out, and measured values arise from fundamental adhesive and interface properties (Lavisci et al. 2003; Veigel et al. 2011). Furthermore, adhesives can be characterised in a more detailed way; conclusions about the rupture-behaviour are possible. Therefore, glues can be as-
686
signed to brittle and flexible fracture behaviour (Veigel et al. 2011). One disadvantage is that sample preparation is more complicated. With FTIR it was possible to allocate which polymers in wheat flour were affected by ageing. By means of PCA and loading plots degree of changes were visualized. The results showed only marginal changes for proteins, but strong changes for starch molecules. Due to its hydrophobic character gluten is less suitable for hydrolysis. Additionally, DSC was used to investigate retrogradation. DSC measurements showed increasing extent of retrogradation and enthalpies from melting of starch crystalline structures. First, speed of retrogradation was quite fast, after 14 days about 25% and after one month nearly half of origin crystallinity was accomplished. Afterwards growth of crystalline structures was slower. Two months later retrogradation was finished and a similar crystalline amount of starch like in the native flour was reached. Thermal and spectroscopic measurements showed similar results, increasing crystallinity of starch during time of storage. Recently heated wheat flour suspension is an elastic system containing swollen starch granules that are mainly deformed and crushed. For starches containing both amylose and amylopectin, a composite gel network is gained consisting of amylopectin-enriched granular fragments inside an amylose gel matrix. During storage staling occurs and mechanical properties are altered. An important cause of staling is retrogradation of starch, where starch molecules re-associate into an ordered structure and crystalline areas are formed. The starch retrogradation is mainly caused by amylopectin since the retrogradation of amylose is generally finished a few hours after heating. This process is accompanied by an increase in firmness and loss of water (Miles et al. 1984). Usually reorganisation of amylopectin occurs much faster,;in starch gels and products usually within a few days or weeks. Also the speed of retrogradation is higher (Cocchi et al. 2005; Lionetto et al. 2005; Liu et al. 2010). The slow process can be explained by the reduced mobility of starch due to the adhesion on the wooden surface in the glue line. Furthermore, an almost sealed system is generated in the glue line, and therefore mobility of water is limited. Generally, water evaporates and the amount of adsorbed water is reduced during ageing of starch-based materials, but this process is retarded due to the surrounding wood. Measured crystallinity after only three months is in the range of the native flour, which suggests that the recrystallisation process has finished. However, the calculated retrogradation rate seems to be quite high, but is in the range of the results from other researches (Del Nobile et al. 2003; Lu et al. 1997). Little differences can be explained by shorter storage periods or even measurement parameters.
Eur. J. Wood Prod. (2012) 70:679–688
Glucose and glucose oligosaccharides arise by hydrolysis of starch (Nagamori and Funazukuri 2004). The main degradation products of monosaccharide in a moderate acidic environment are furfural FF, HMF and acetic acid (Chheda et al. 2007; Qi et al. 2008). During baking of grain based products also FF and HMF are formed due to the reaction of monosaccharides and proteins, called Maillard reaction (Ameur et al. 2006). However the fresh glue contained no FF and HMF, and therefore these compounds can be addressed to degradation processes. Furthermore, glucoronic acid, an oxidation product of glucose, could not be detected even for aged samples. Therefore, oxidation processes seem to have little influence. Acetic acid is also generated by thermal treatment (Tjeerdsma et al. 1998) and ageing of wood from hemicelluloses (Ters et al. 2011). Due to higher temperatures the formation of acetic acid from hemicelluloses was promoted during sample preparation and could catalyse depolymerisation of carbohydrates. The lower pH-value caused by acetic acid could enhance degradation reactions and products like FF and HMF (Nagamori and Funazukuri 2004). Although acetic acid could not be quantified, it likely plays an important role due its catalytic attribute. Staling of heated wheat flour products is a complex process involving many factors, but the recrystallisation of starch during storage is believed to be a main contributor (Wang et al. 2004). Determined fracture energies affirmed the suggestion that the containing starch has retrogradated, and therefore the adhesive bond became more brittle. After one month already half the amount of native crystallinity was reached. Between one month and one year of ageing fracture energies decreased somewhat, which can be influenced by the higher rigidity due the advanced recrystallisation. Results showed correlation between bonding performance and concentration of extractable starch fragments measured in glucose equivalents. During ageing till six months bonding strength as well as water-soluble starch fragments remained constant. The bonding strength of longer aged samples decreases and the measured amount of glucose increased. However, starch hydrolysis seems to be responsible for less adhesive strength. The amount of gluten seems to have no considerable influence on bonding properties. Generally, the disulfidebonds and the fibrous construction in gluten are responsible for its flexible character and high elasticity. However fracture energies of flexible adhesives, e.g. polyurethanes, are much higher (Veigel et al. 2011). This can be explained by the relation between starch and gluten in wheat flour, which is approximately 8:1. Therefore, starch is mainly responsible for the character of the adhesive system due to the much higher amount. Similar conclusions were found for wheat flour based foods like bread (Lionetto et al. 2005; Wang et al. 2004). Numerous factors influence the durability and properties of adhesive bonds. It must also be considered that a sizeable
Eur. J. Wood Prod. (2012) 70:679–688
part of the loss in bond strength can be assigned to the decrease of mechanical properties of the adherent wood (Follrich et al. 2010). However results of lap joint testing showed a decrease in shear strength as well as in degree of wood failure. Consequently, it can be assumed that shear strength falls because of less adhesive and cohesive power. However, used smooth storage conditions of 20 ± 2◦ C and 65% relative humidity should not affect wood properties significantly.
5 Conclusion Used mechanical methods allowed for examining the bonding performance of the wheat flour adhesive during storage of one year. The good results from earlier research were confirmed with both test configurations. However, evaluation of bonding strength for wooden samples by means of DCBspecimens seems to be superior compared to lap joints, because influence of wood is avoided. FTIR analysis indicated that ageing affects starch much more than wheat proteins. The analytical data correlates well with adhesive and cohesive properties. First, a rapid rise of crystallinity was determined by DSC and infrared spectroscopy. The adhesive properties remained extensively constant. After 6 months of storage an increase in degradation products was detected, but water-soluble hydrolysis products of starch snatches remained constant. Advanced increase of hydrolysis and degradation products was detectable after one year of storage, concurrently bonding power of wheat flour decreased. Although no significant loss in adhesive properties was measured, both methods for mechanical testing indicated a trend to lower performance. However, it can be assumed that degradation of starch is mainly responsible for the moderate loss of adhesive power during ageing. Acknowledgements The authors gratefully acknowledge the financial support by the Competence Centre for Wood Composites and Wood Chemistry, Wood K plus. We also like to express our thanks to Holzindustrie Schweighofer GmbH and Austrian Science Fund (FWF) (Project no. L319-B16) for supplying the research on renewable woodstarch composites.
References Ameur LA, Trystram G, Birlouez-Aragon I (2006) Accumulation of 5hydroxymethyl-2-furfural in cookies during the backing process: Validation of an extraction method. Food Chem 98(4):790–796 Baik MY, Chinachoti P (2000) Moisture redistribution and phase transitions during bread staling. Cereal Chem 77(4):484–488 Chheda JN, Romàn-Leshkov Y, Dumesic JA (2007) Production of 5hydroxymethylfurfural and furfural by dehydration of biomassderived mono- and poly-saccharides. Green Chem 9:342–350 Cocchi M, Foca G, Marchetti A, Sighinolfi S, Tassi L, Ulrici A (2005) Use of multivariate analysis of MIR spectra to study bread staling. Ann Chim 95(9–10):657–666
687 D’Amico S, Hrabalova M, Müller U, Berghofer E (2010) Bonding of spruce wood with wheat flour glue—Effect of press temperature on the adhesive bond strength. Ind Crop Prod 31(2):255–260 Däumling C, Brenske KR, Wilke O, Horn W, Jann O (2005) Healthrelated evaluation procedure of VOC and SVOC emissions from building products—A contribution to the European construction products directive. Gefahrst Reinhalt Luft 65(3):90–92 Del Nobile MA, Martoriello T, Mocci G, La Notte E (2003) Modeling the starch retrogradation kinetic of durum wheat bread. J Food Eng 59(2–3):123–128 Dix B (1987) Bonding veneer plywood with modified diisocyanates. Holz Roh- Werkst 45(12):487–494 EN 302-1 (2004) Klebstoffe für tragende Holzbauteile—Prüfverfahren—Teil 1: Bestimmung der Längszugscherfestigkeit. Österreichisches Normungsinstitut, Wien Follrich J, Stöckel F, Konnerth J (2010) Macro- and micromechanical characterization of wood-adhesive bonds exposed to alternating climate conditions. Holzforschung 64(6):705–711 Kizil R, Irudayaraj J, Seetharaman K (2002) Characterization of irradiated starches by using FT-Raman and FTIR spectroscopy. J Agric Food Chem 50(14):3912–3918 Konnerth J, Gindl W, Harm M, Müller U (2006) Comparing dry bond strength of spruce and beech wood glued with different adhesives by means of scarf- and lap joint testing method. Holz Roh- Werkst 64(4):269–271 Kuutti L, Peltonen J, Myllärinen P, Teleman O, Forssell P (1998) AFM in studies of thermoplastic starches during ageing. Carbohydr Polym 37(1):7–12 Imam SH, Mao LJ, Chen L, Greene RV (1999) Wood adhesive from crosslinked poly(vinyl alcohol) and partially gelatinized starch: Preparation and properties. Starch - Stärke 51(6):225–229 Ji Y, Zhu K, Qian H, Zhou H (2007) Staling of cake prepared from rice flour and sticky rice flour. Food Chem 104(1):53–58 Lavisci P, Pizzo B, Gagliano JM, Triboulot P, De Ciechi M (2003) Fracture energy testing of thick joints with structural wood adhesives. Holz Roh- Werkst 61(5):355–357 Li W, Dobraszczyk BJ, Dias A, Gil AM (2006) Polymer conformation structure of wheat proteins and gluten subfractions revealed by ATR-FTIR. Cereal Chem 83(4):407–410 Lionetto F, Maffezzoli A, Ottenhof MA, Farhat IA, Mitchell JR (2005) The retrogradation of concentrated wheat starch systems. Starch Stärke 57(1):16–24 Livings SJ, Breach C, Donald AM, Smith AC (1997) Physical ageing of wheat flour-based confectionery wafers. Carbohydr Polym 34(4):347–355 Liu H, Yu L, Tong Z, Chen L (2010) Retrogradation of waxy cornstarch studied by DSC. Starch - Stärke 62(10):524–529 Lu TJ, Jane JL, Keeling PL (1997) Temperature effect on retrogradation rate and crystalline structure of amylose. Carbohydr Polym 33(1):19–26 Miles MJ, Morris VJ, Ring SG (1984) Some recent observations on the retrogradation of amylose. Carbohydr Polym 4(1):73–77 Moubarik A, Pizzi A, Allal A, Charrier F, Khoukh A, Charrier B (2010) Cornstarch-mimosa tannin-urea formaldehyde resins as adhesives in the particleboard production. Starch - Stärke 62(3–4):131–138 Nagamori M, Funazukuri T (2004) Glucose production by hydrolysis of starch under hydrothermal conditions. J Chem Technol Biotechnol 79:229–233 Plath L (1972) Role of starch products in gluing of wood materials by synthetic resin adhesives. Starch - Stärke 24(9):306–309 Qi W, Zhang SP, Xu QL, Ren ZW, Yan YJ (2008) Degradation kinetics of xylose and glucose in hydrolysate containing dilute sulfuric acid. Chin J Process Eng 8(6):1132–1137 Ramis X, Cadenato A, Salla JM, Morancho JM, Vallés A, Contat L, Ribes A (2004) Thermal degradation of polypropylene/starchbased materials with enhanced biodegradability. Polym Degrad Stab 86(3):483–491
688 Reimerdes EH, Rothkitt KD (1984) Qualitative and quantitative analysis of carbohydrates in food. Fresenius J Anal Chem 318(34):220–224 Sevenou O, Hill SE, Farhat IA, Mitchell JR (2002) Organisation of the external region of the starch granule as determined by infrared spectroscopy. Int J Biol Macromol 31(1–3):79–85 Ters T, Follrich J, Hinterstoisser B (2011) Artificially ageing of softwood and its effects on mechanical properties and chemistry. Wood Mater Sci Eng 6:58–68 Tjeerdsma B, Boonstra M, Pizzi A, Tekely P, Militz H (1998) Characterisation of thermally modified wood: molecular reasons for wood performance improvement. Holz Roh- Werkst 56(3):149– 153 Tondi G, Wieland S, Wimmer T, Schnabel T, Petutschnigg A (2011) Starch-sugar synergy in wood adhesion science: basic studies and particleboard production. Eur J Wood Prod. doi:10.1007/ s00107-011-0553-z van Soest JJG, Tournois H, de Wit D, Vliegenthart JFG (1995) Short-range structure in (partially) crystalline potato starch deter-
Eur. J. Wood Prod. (2012) 70:679–688 mined with attenuated total reflectance Fourier-transform IR spectroscopy. Carbohydr Res 279:201–214 Veigel S, Follrich J, Gindl-Altmutter W, Müller U (2011) Comparison of fracture energy testing by means of double cantilever beam-(DCB)-specimens and lap joint testing method for the characterization of adhesively bonded wood. Eur J Wood Prod. doi:10.1007/s00107-010-0499-6 Wang X, Choi SG, Kerr WL (2004) Effect of gluten content on recrystallisation kinetics and water mobility in wheat starch gels. J Sci Food Agric 84(4):371–379 Zhang X, Gozukara Y, Sangwan P, Gao D, Bateman S (2010) Biodegradation of chemically modified wheat gluten-based natural polymer materials. Polym Degrad Stab 95(12):2309–2317 Zhou YG, Li D, Wang LJ, Li Y, Yang BN, Bhandari B, Chen XD, Mao ZH (2009) Effect of water content on thermal behaviors of common buckwheat flour and starch. J Food Eng 93(2):242–248
