Originalarbeiten á Originals

Holz als Roh- und Werkstoff 59 (2001) 413±421 Ó Springer-Verlag 2001

Honeymoon MUF adhesives for exterior grade glulam M. Properzi, A. Pizzi, L. Uzielli

An exclusively MUF- (melamine-urea-formaldehyde) based honeymoon adhesive for glulam and ®ngerjoints is presented in which one component is composed of a high performance MUF resin, while the second separate application component is based on just slightly acid water thickened to the same viscosity of the ®rst component by the addition of 1.5% carboxymethyl cellulose (CMC). MUF-based, honeymoon-type, fast-setting, separate application adhesive systems, which do not need any resorcinol, are shown here to be capable of performing as adhesives for structural exterior-grade joints and glulam and to satisfy all the requirements of the relevant adhesive speci®cations for such an application. The parameters which are shown to be determining are mainly the performance of the MUF resin, if and once an excellent resin formulation is available, both the ratio of melamine to urea and the molar ratio having a lesser effect. The performance only starts to drop lower than the requirements of relevant standards from M:U weight ratios well below 20:80 and of the order of 10:90. Addition of resorcinol at these failing levels while improving slightly the performance do not solve the problem, and resorcinol addition does not allow satisfaction regarding speci®cation requirements. At the higher M:U ratios such as M:U ˆ 47:53, but even lower, addition of resorcinol does not improve at all the results, its addition again revealing itself super¯uous. The reasons for such a behaviour are presented and explained. The MUF honeymoons present all the other usual advantages associated with honeymoon adhesives, namely high curing rate, long pot-life, tolerance to higher moisture content of the substrate, and tolerance to even quite severe imbalances in viscosity and proportions between the two components.

MUF-Typ. Die andere besteht aus dem gleichen, jedoch leicht angesaÈuerten MUF-Harz, das durch Zusatz von 1,5%iger Carboxymethylcellulose (CMC) auf die gleiche ViskositaÈt wie die erste Komponenten eingestellt ist. Solche rasch haÈrtenden MUF-Harze vom Typ `Honeymoon' benoÈtigen keinen Zusatz von Resorcin. Es wird hier gezeigt, daû sie fuÈr Glulam und Keilzinken in Auûenverwendung geeignet sind und allen Anforderungen an solche Konstruktionszwecke genuÈgen. Wenn eine geeignete Zusammensetzung einmal gefunden ist, spielen Abweichungen vom VerhaÈltnis der Komponenten nur eine geringe Rolle. Die von den zustaÈndigen Normen geforderten QualitaÈtsstandards werden nur dann unterschritten, wenn das VerhaÈltnis M:U unter einen Wert von 20:80 bis zur GroÈûenordnung von 10:90 faÈllt. In diesem Bereich erhoÈht ein Zusatz von Resorcin zwar leicht die QualitaÈt, loÈst aber nicht das Problem, d.h. die Normwerte werden dann auch durch Resorcin-Zusatz nicht erreicht. Bei M:U-VerhaÈltnissen von 47:53, sogar darunter, wird durch ResorcinZusatz keine Verbesserung erzielt; Resorcin erweist sich also auch hier als uÈber¯uÈssig. Die GruÈnde fuÈr dieses Verhalten werden vorgestellt und erlaÈutert. Die MUF-Kleber vom Typ `Honeymoon' zeigen alle anderen Vorteile dieses Klebertyps, wie schnelle AushaÈrtzeit, lange Topfzeit, Toleranz hoÈherer Feuchtigkeitswerte der Materialen und sogar eine Toleranz von erheblichen ViskositaÈtsunterschieden und VerhaÈltnissen der beiden Komponenten.

1 Introduction Honeymoon fast-set adhesives for exterior-grade structural glulam and ®ngerjointing have now been used industrially for about twenty years (Pizzi 1983, 1989, 1994). From the original concept of the honeymoon fast-set MUF-Kleber vom Typ `Honeymoon' fuÈr Glulam zur system and the initial systems which were eventually inAuûenverwendung dustrialized (Pizzi 1989, Pizzi and Cameron 1984, Pizzi Der hier vorgestellte Zwei-Komponenten-Kleber (Typ et al. 1980), and are still extensively used commercially, `Honeymoon') fuÈr Glulam und Keilzinken enthaÈlt ausseveral relevant variations were derived over the years. schlieûlich Melamin, Harnstoff und Formaldehyd. Die eine Thus, to the original PRF/PRF systems were added fairly Komponente besteht aus einem hochwirksamen Harz vom soon afterwards the PRF/Tannin systems, both being fast adhesives of structural marine grade capability able of setting and curing at elevated rates but at ambient and A. Pizzi (&), M. Properzi even lower winter temperatures, and even at higher ENSTIB, University of Nancy 1, 88051 Epinal, France moisture contents of the timber (Pizzi 1989, 1994, Pizzi e-mail: [email protected] and Cameron 1984). Variations of the two same types but Tel.: +33-3-29811150, Fax: +33-3-29340975 more apt to automatic ®ngerjointing equipment and capable of pot-lives of 48 hours or longer, while maintaining M. Properzi, L. Uzielli the same curing rate (Pizzi 1989, 1994) were then added. Á DISTAF, Universita degli Studi di Firenze, Firenze, Italy Variations based on TRF/TRF systems (Stephanou and

413

414

Pizzi 1994, Pizzi et al. 1980), were also commercialized earlier, and all of these variations are now in industrial use in a few countries, and have been so for a long time. All these systems have been aimed at the decrease of the expensive resorcinol content on total resin solids while conserving the structural, exterior bonding performance well over the requirements of the standards. Some other variation on the same theme have been presented in the literature, although some of them might not as yet be extensively commercial, such as for instance (i) substituting one inexpensive component (for instance tannin) with another inexpensive component (for instance soya protein hydrolysates) with, however, the need to use heating (Steele et al. 1998) to achieve performance and thus partly denying the advantages of a separate application fast set adhesive system; (ii) the use of ``pre-branched'' resins to produce fast-set adhesive systems of extremely low resorcinol content (Scopelitis and Pizzi 1993, Pizzi 1994), and capable of bonding very high moisture content timber, and (iii) the use of a melamine resin and a resorcinol separate components system (Greenweld, US patent 5,674,338 ± 1997), the alleged greater tolerance of which to the bonding of higher moisture content timber being in reality on par with most of the other systems already listed above. However, for all the improvements apported to the commercial resins of this type and all the different variations on the honeymoon-type fast-set adhesives, all these systems are still based on some resorcinol or resorcinol-aided component. Systems (ii) and (iii) above are the two possibly using the lowest amount of resorcinol. System (iii) using as one component a melamine-urea-formaldehyde (MUF) resins of high melamine content and resorcinol as a second component is just unusual in its use of a MUF rather than a PRF resin, a very acceptable resin concept, except for the fact that it is coupled with a phenol such as resorcinol. The coupling of an acid-setting MUF adhesive and of resorcinol might well present no advantages or even some potentially serious disadvantages. It has been shown, for example, that thermosetting PMUF resins do not present a better performance than equivalent MUF resins and that often, depending on their sequence of manufacture, they present instead a much worse performance. There are very well de®ned technical and chemical reasons for this (Higuchi et al. 1994, Cremonini et al. 1996) that boil down to the relevant differences in reactivity of the two materials, namely the phenol (here resorcinol) and melamine. The reactivity of melamine and even urea at the acid setting pHs they need is much greater than that of any phenol, even resorcinol, as this pH range is the one of lowest rectivity of any phenol. Thus, even resorcinol runs the risk to be very little linked to the MUF matrix, especially in a fast setting system, such as a honeymoon, and at best it will remain as a by-passed pendant side group not able to ful®ll the function for which it has been added. There have also been, in the past, equally valid trade reasons for their use, their introduction having been originally prompted by the then perceived need to upgrade the exterior performance of MUF wood adhesives. The idea that the addition of small percentages of phenol to a MUF resin yields resins of better exterior durability is an

incorrect myth today traditionally perpetuated in the wood panels industry. Newer formulations of MUF resins always outperform the corresponding PMUF. PMUFs are not bad resins, they are simply resins in which one of the materials, phenol or resorcinol, is often wasted for no purpose. This is not too important for the relatively inexpensive phenol, but it is rather grave in the case of the expensive resorcinol. The idea that resorcinol under the conditions of operation of a MUF resin can be used to upgrade its performance, or at least to accelerate its rate of curing can indeed be challenged. This paper then deals with the development of a new type of exterior-grade, structural honeymoon-type fast-set MUF resins for glulam/®ngerjointing of excellent performance, without the need to use any resorcinol, equally bonding higher moisture content timber, of totally clear/ transparent colour of the hardened glue-line, and on achieving the lowest possible melamine content at parity of performance by simply upgrading MUF resin formulation.

2 Experimental 2.1 Resin preparation MUF resins of urea to melamine weight ratios of 47:53, 40:60, 30:70, 20:80 and 10:90, and of (M + U):F molar ratios of 1:2, 1:1.9, 1:1.7 and 1:1.5 were prepared according to a resin formulation already reported (Pizzi 1994). The MUF resins were prepared to a degree of advancement as de®ned by a water tolerance of 150%, while some resins of much lesser advancement and of water tolerance of 300% were also prepared. These are indicated in the text and in the relevant ®gure legends. Solid contents and viscosity of the resins were maintained in the range 66±70% and 2000± 2500 mPa.s, unless otherwise indicated. Gel permeation chromathography (GPC) was carried out on a Waters GPC equipped of both UV and refractometer detectors, in dimethylformamide as the mobile phase. 2.2 Glue-mix preparation and testing The two components of the separate application honeymoon adhesive were composed as follows: Component A: the MUF resin at a pH of approximately 10, solids content of 72%±73%, and viscosity of 2000± 2500 mPa.s . No ®llers were added. Component B: 1.01 parts by weight carboxymethyl cellulose is dissolved in 55 parts water and the solution is left to hydrate well for 24 hours. 27 parts by weight of formic acid solution is then added. The two components were then each spread on the surface of separate beech (Fagus silvatica) strips of dimensions 120 ´ 25 ´ 3 mm, and these were then assembled to have a bonded overlap of 25 ´ 25 mm, then clamped and left in the clamp for 16 hours and then aged for 7 days, all at 20 °C. The bonded specimens so prepared were then tested for tensile strength dry (10 specimens), after 24 hours cold water soak (10 specimens), and after 2 hours boil (10

specimens), tested wet, according to the relevant British Standard BS 1204, Part 1, SABS, ASTM and European Norm EN. For the tests of strength increase as a function of time the specimens were bonded and assembled according to the same procedure, but they were clamped and tested at 15 minutes intervals to build the curves shown in the ®gures.

2.3 Thermomechanical analysis (TMA) The MUF resins above, after addition of 2% amonium chloride hardener, were tested dynamically by thermomechanical analysis (TMA) on a Mettler apparatus. Triplicate samples of beech wood alone, and of two beech wood plys each 0.6 mm thick bonded with each system, for a total samples dimensions of 21´6´1.2 mm were tested in non-isothermal mode between 40 °C and 220 °C at a heating rate of 10 °C/min with a Mettler 40 TMA apparatus in three points bending on a span of 18 mm exercising a force cycle of 0.1/0.5 N on the specimens with each force cycle of 12 seconds (6 s/6 s). The classical mechanics relation between force and de¯ection E ˆ [L3/(4bh3)][F/ (f)] allows the calculation of the Young's modulus E for each case tested. 2.4 13 C NMR spectrum The liquid 13C NMR spectrum of the MUF resins used were obtained on a BruÈker MSL 300 FT-NMR spectrometer. Chemical shifts were calculated relative to (CH3)3Si(CH2)3SO3Na dissolved in D2O for NMR shifts control (Breitmeier and Voelter 1987) The spectra were done at 62.90 MHz for a number of transients of approximately 1000. All the spectra were run with a relaxation delay of 5 s and chemical shifts were accurate to 1 ppm. 2.5 Model All the curves of increase of strength as a function of time shown in the Figures were obtained according to the Ramberg-Osgood model (Ramberg and Osgood 1943) which indicates that all the curves are composed of two sections almost rectilinear. At the beginning the curve is represented by the equation f ˆ Kt, if f £ Re, where t is the time in hours, K is a constant representing the slope of the curve, Re is the elastic limit of the system, and f is the tensile strength recorded. The second half of the curve responds instead to the equation f ˆ C(t±Re+E0)n if f > Re and where Re ˆ Felastic/K. The values obtained for all these parameters are reported elsewhere (Properzi 2000). 3 Discussion The ``honeymoon''-type, fast-setting, separate application adhesive presented and developed here is rather unusual for a series of features never used before. Thus, (i) the two components to be applied to the two wood faces to be mated are composed of a resin (component A) and of a very diluted solution of a thickener (component B) to which a diluted acid catalyst has been added, and not by two resins as it is normally the case for these adhesives.

Component B is nothing else than 1% carboxymethyl cellulose (CMC) in water, hence component B is really nothing more than ``thickened'' water carrying the catalyst. (ii) Component A is just formed of a pure top of the range melamine-urea-formaldehyde (MUF) resin, in itself this not being new as one case of a fast-set adhesive based on a MUF resin is on record, although helped along by the use of resorcinol in component B (Greenweld). Apart from this what is new here is instead the lack of any resorcinol or resorcinol-like material as an accelerator, upgrader, forti®er, performance or durability enhancer of the MUF resin in either components A and B. What is dealt with here is a MUF resin formulation of such a top performance to be able to act alone as a honeymoon structural exteriorgrade fast-set adhesive for glulam delivering resorcinollike durability and performance without any trace of resorcinol whatsoever. The 13C NMR spectrum characterizing one of the resins used, namely the sequential MUF 47:53 M:U, of molar ratio F/(M+U) ˆ 1.9, is shown in Fig. 1. The spectrum indicates that this presents a rather classical sequential MUF formulation in which the melamine is well copolymerized in the resin (Pizzi 1994), but has also the characteristic to present a rather noticeable proportion of urons (signals at 156±158 ppm and at 80± 81 ppm) (Soulard et al. 1998, 1999), bound in the skeleton of the resin itself yielding both a markedly higher strength and low formaldehyde emission (Soulard et al. 1998, 1999). The results shown in Tables 1, 2, 3, and 4 of tests carried out according to several international standards for exterior-grade, structural glulam adhesives clearly indicate that such a pure MUF resin can be succesfully used as a fast-set honeymoon adhesive for ®ngerjointing and glulam. They also indicate that the concept that a resin is only needed to be applied on a single pro®le of the two surfaces to be joined can indeed hold true: all what is needed on the second pro®le is nothing else than a very diluted solution of an acid catalyst (not too acid, here the pHs obtained were around 4) and a thickener in a concentration just suf®cient to maintain component B at a viscosity not too different from that of component A (the MUF resin). The basic MUF resin used (Pizzi 1994) is a sequential formulation developed especially to be a top performer at low adhesive content, according to logical reactivity principles, and has been shown many times to yield top of the range MUF thermosetting adhesives. Its effectiveness is rather independent of the molar ratio used, within the 1.5 to 1.9 range. It is, however, interesting to peruse in depth the results obtained. In Table 1 the results obtained on close contact joints bonded with a series of high molar ratio MUF resins, namely (M+U):F ˆ 1:1.9, but in which the relative weight proportions of melamine and urea varies from M:U of 47:53, down to weight ratios of 10:90, indicate that all the resins down to a 20 percentage weight of melamine are able to satisfy relevant international speci®cations for exterior grade structural adhesives. Only the M:U 10:90 MUF adhesive does fail the requirements of the American, British, South African, Australian and European Norm standards, and this only in the boiling test. A decrease in melamine content is shown in Table 1 to yield

415

416

Fig. 1. Quantitative 13C NMR spectrum of a sequentially prepared MUF resin of 47:53 M:U weight ratio, of 1:1.9 (M+U):F molar ratio used for honeymoon application. Note the intense uron signals at 154±157 ppm and at 80±81 ppm

Bild 1. Quantitatives 13C NMR-Spectrum eines MUF-Harzes fuÈr Zweikomponentenkleber von Typ `Honeymoon'. Die VerhaÈltnisse M:U und (M+U)/F betragen 47:53 bzw. 1:1,9. Man beachte das intensive Uron-Signal bei 154±157 ppm und 80±81 ppm

Table 1. Effect on tensile strength and wood failure percentage of the M:U weight ratio of the MUF resin used in a honeymoon adhesive system to bond beech strips. Tested according to British standard BS 1204 Part 1 Tabelle 1. Ein¯uû des VerhaÈltnisses M:U eines MUF-Harzes vom Typ `Honeymoon' auf Zugfestigkeit und Bruchrate von verleimten Buchenbrettern. Tests nach BS 1204, Teil 1 M:U Weight ratio

47:53 40:60 30:70 20:80 10:90 BS1204, part 1 EN 301 SABS 1349-1982

F/(M+U) Molar ratio

1.9 1.9 1.9 1.9 1.9

Dry

24 h soak

2 h boil

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

3360 2906 2980 2961 2862 ³2500 ³2000 ³2500

100 100 100 100 100 ± ± ³75

2820 2610 2550 2390 2036 ³2000 ³1200 ³2000

100 100 100 90 70 ± ± ³75

2308 2220 1950 2025 400 ³1500 ³1200 ³1500

100 100 80 70 0 ± ± ³75

a decrease in wet performance of the bonded joint, but this decrease, although noticeable is ®rst of all rather small and does not impede the adhesive to pass well the relevant strength and percentage wood failure requirements of the various standards. This means that relatively inexpensive MUF formulations of 20% melamine on resin solids and

TMA max MOE (MPa)

3354 2944 3043 2999 3002

containing no resorcinol can be used to good effect for glulam and ®ngerjointing application. The thermomechanical analysis in bending of the same resins, according to methods already described (Garcia and Pizzi 1997, Kamoun et al. 1998, Kamoun and Pizzi 2000), supported what was observed by tensile test, indi-

cating as well that in dry tests the MUF 47:53 gives better results than the other, but already from the MUF 40: 60 one cannot notice much difference between the dry results obtained with the different resins. Equally interesting is the dependance of total performance on the (M+U):F molar ratio, which is shown in Table 2. As would be expected, the performance shows an overall decrease as the molar ratio decreases from 1:2 down to 1:1.5. Notwithstanding this fact, however, even at the low molar ratio of 1:1.5 all the joints pass the relevant standard speci®cations. It is clear then that for standard durability tests one can, within limits decrease the molar ratio or the melamine content, and still obtain very acceptable results. Of these two options the latter is more economically attractive, the former more performance attractive. Strength development as a function of time is an important parameter to consider and to judge the adhesive system by, as one deals here with fast-set adhesives. In Fig. 2 are shown the rates of strength development of the MUF systems of Table 1. The rate of strength development decreases from the maximum observed for the 47:53 M:U case to a minimum as one passes to the 40: 60 case and remains approximately at this minimum up to the 30:70 M:U case. It starts to increase again with the 20:80 M:U case and slightly increases further in the case of

the 10: 90 M:U case. The changes in slope might not appear much but they yield very different rates of strength development. Thus, as a comparison, the top of the range 47:53 MUF adhesive system reaches the dry strength standard requirement after 2 hours from catalysis, the 40:60 and the 30:70 MUF adhesive systems after approximately 4 hours, the 20:80 MUF after slightly less than three hours and the lowest cost 10:90 MUF after approximately the same 2 hours of the top of the range resin (Fig. 2). It is then evident that two effects, both effects working against each other are present and that the relative proportions of melamine and urea have an important effect on the rate of strength development of the system, probably more marked than on the ultimate aged strength (Table 1), with the 47:53 and 10:90 MUF resin systems de®netely behaving as very acceptable honeymoon fast-set systems. Equally interesting is that the maximum dry strength for the curves in Fig. 2 passes from 5.5 kN (5500 MPa) for the 47:53 MUF down to 4700 MPa for the 40: 60 and 30:70 MUF resins and back up again to 5000 and 5100Mpa, respectively, for the 20:80 and 10:90 MUF systems. Here the initial maximum strength decrease, and the increase that follows as the weight ratio M:U varies as a consequence of the decrease in the relative proportion of melamine can be explained by the presence of two effects operating against each other. Thus, as regards the ®rst effect, the higher the

Table 2. Effect on tensile strength and wood failure percentage of the F/(M+U) molar ratio of the MUF resin used in a honeymoon adhesive system to bond beech strips. Tested according to British standard BS 1204 Part 1 Tabelle 2. Ein¯uû des VerhaÈltnisses F/(M+U) eines MUF-Harzes vom Typ `Honeymoon' auf Zugfestigkeit und Bruchrate von verleimten Buchenbrettern. Tests nach BS 1204, Teil 1 M: U Weight ratio

F/(M+U) Molar ratio

47:53 2.0 47:53 1.9 47:53 1.7 47:53 1.5 Standards requirements

Dry

24 h soak

2 h boil

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

3400 3360 2764 2761 ³2500

100 100 100 100 ³75

2920 2820 2526 2493 ³2000

100 100 100 100 ³75

2360 2310 2139 2116 ³1500

100 100 90 80 ³75

TMA max MOE (MPa)

± 3432 3245 3153

Fig. 2. Tensile strength increase as a function of time of beech joints (BS 1204, Part 1) bonded with MUF based honeymoon adhesive systems: effect of the variation of the M:U weight ratio Bild 2. Zeitlicher Anstieg der Zugfestigkeit von Buchenverbindungen, verklebt mit einem ZweikomponentenMUF-Harz vom Typ `Honeymoon' sowie Ein¯uû des M:U-VerhaÈltnisses. Tests nach BS 1204, Teil 1

417

418

proportion of linked melamine in the resin the higher the average functionality of the system, the lower the degree of conversion at which gelling of the resin occurs, and the tighter and hence stronger is the ®nal network formed by the hardened resin. Using accepted gel theory forecasting formulas (Odian 1994, Durand/Bruneau 1982 a,b) one can calculate the degree of conversion at the gel point for the ®ve resins in Fig. 2 as respectively 0.52, 0.55, 0.61, 0.67 and 0.73 indicating the preponderant contribution of melamine to the level of cross-linking of the system. The second effect instead depends on the relative residual ¯exibility of the hardened network formed. The more ¯exible the network the better glueline stress distribution. Thus, within well-de®ned limits, as one decreases the proportion of melamine, the segments between the network cross-linking nodes becomes longer and more ¯exible (Pizzi 1997, 1999, Pizzi et al. 1997), hence the density of cross-linking decreases resulting in better distribution of stress on testing and in better strength results (Ebewele et al. 1991a, b, Pizzi 1994). The combination of the two effects yields the type of results observed in Fig. 2 as regards maximum strength, with this parameter going through a minimum between the two maxima corresponding to the optimum conditions for one of the two effects.

In Fig. 3 instead, are shown the rates of strength development of MUF systems of different F/(M+U) molar ratio. As could be expected, a decrease in the F/(M+U) molar ratio slows down the rate of strength development of the MUF system. Thus, while the 1.9 molar ratio MUF system reaches the dry strength standard requirement after 4 hours from catalyst addition, the 1.7 molar ratio resin reaches the same strength after slightly less than 5 hours and the 1.5 MUF after slightly more than 6 hours. It must be pointed out that the MUF 47:53, 1.9 in Fig. 3 (4 hours to reach dry standard) and the equivalent one in Fig. 2 (2 hours to reach dry standard) are not the same resin, the difference between them being the grade of advancement of the resin. These being re¯ected in the 150% and 300% water tolerance for the resins in Fig. 2 and 3 respectively. The point is well illustrated in Fig. 4 were two identical MUF 47:53, 1.9 resins differ only in water tolerance (the slower one at 300% and the faster one at 150%). The difference in water tolerance corresponds to well de®ned differences at the molecular level. Thus, gel permeation chromatography of the two resins shown in Fig. 5a and b indicates that the two resins present marked differences at the level of their number average (Mn) and weight average (Mw) molecular masses. The resin at 300% water tolerance

Fig. 3. Tensile strength increase as a function of time of beech joints (BS 1204, Part 1) bonded with MUF based honeymoon adhesive systems: effect of the variation of the F/(M+U) molar ratio Bild 3. Zeitlicher Anstieg der Zugfestigkeit von Buchenverbindungen, verklebt mit einem ZweikomponentenMUF-Harz vom Typ `Honeymoon' sowie Ein¯uû des VerhaÈltnisses F/(M+U). Tests nach BS 1204, Teil 1 Fig. 4. Tensile strength increase as a function of time of beech joints (BS 1204, Part 1) bonded with MUF based honeymoon adhesive systems: effect of the variation of resin advancement as represented by percentage water tolerance. Faster resin, water tolerance ˆ 150%; slower resin, water tolerance ˆ 300% Bild 4. Zeitlicher Anstieg der Zugfestigkeit von Buchenverbindungen, verklebt mit einem ZweikomponentenMUF-Harz vom Typ `Honeymoon' (Tests nach BS 1204, Teil 1) sowie Ein¯uû der Feuchtetoleranz verschieden vorpolymerisierter Harze. Schneller haÈrtendes Harz: 150% Feuchtetoleranz; langsamer haÈrtendes Harz: 300% Feuchtetoleranz

419

Fig. 5a, b. Mol mass distribution (By GPC) of MUF resins (M:U ˆ 47:53, F/(M+U) ˆ 1.9) resins of different advancement as represented by percentage water tolerance; (a) water tolerance ˆ 300%. (b) water tolerance ˆ 150% Bild 5a, b. Molmassenverteilung mittels GPC von unterschiedlich vorpolymerisierten MUF-Harzen (MolverhaÈltnisse M: = 47:53 und F/M+U) ˆ 1,9); (a) Feuchtetoleranz ˆ 300%. (b) Feuchtetoleranz ˆ 150%

level presents Mn ˆ 509 and Mw ˆ 792, with a Mw/Mn ˆ 1.56 (less disperse and less advanced / polymerized), while the resin at 150% water tolerance level presents Mn ˆ 614 and Mw ˆ 1179, with a Mw/Mn ˆ 1.92 (more polymerized/ advanced). In Table 3 are shown the limits of tolerance of the top range MUF resin system, namely the 47:53, 1.9 molar ratio MUF, as regards its tolerance to viscosity differences between the components A and B of the separate application adhesive system, and to the moisture content of the substrate. All the cases in Table 3 were bonded at a component B viscosity of 2500 mPa.s while the viscosity of component A was varied between 360 and 2500 mPa.s, indicating that like all the honeymoon adhesive types of the past (Pizzi 1989, Pizzi et al. 1980) this type too is very tolerant to strong differences in components A and B

proportions, applied to the two wood pro®les. The results in this table were obtained with beech wood bonded at the high moisture content of 18%: the results obtained, well above the requirements of the standard, indicate here, too, that the system will be able to perform well even at much higher moisture contents of the substrate, in the true tradition of most types of honeymoon adhesives (Pizzi 1989, Pizzi et al. 1980). All the results discussed above were obtained in absence of resorcinol. To check what the in¯uence of resorcinol would be, namely to check, at parity of all other conditions, if its addition in the component B would yield better glue-line performance or faster gelling of the system, both ultimate strength after 7 days ageing, as well as rate of strength increase trials as a function of time, were carried out. In Fig. 6 are shown the rate of strength increase

Table 3. Effect of MUF resin viscosity variation on tensile strength and wood failure percentage of a MUF resin used in a honeymoon adhesive system to bond beech strips ± all resins are 47:53 M:U and molar ratio 1.9 Tabelle 3. Ein¯uû der ViskositaÈtsstreuung eines MUF-Harzes vom Typ `Honeymoon' auf Zugfestigeit und Bruchrate von verleimten Buchenbrettern. Alle Harze haben ein M:U-VerhaÈltnis von 47:53 und ein MolverhaÈltnis von 1,9 Viscosity (mPa.s)

Solids Content (%)

2500 73 1500 71 450 67 360 66 Standards requirements

Dry

24 h soak

2 h boil

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

3360 3350 2931 2624 ³2500

100 100 100 100 ³75

2820 2881 2506 2623 ³2000

100 100 100 100 ³75

2310 2125 2000 2462 ³1500

100 100 90 80 ³75

420

curves at ambient temperature of the MUF 47:53, 1.9 alone, and of the same resin in which 4% resorcinol on resin solids has been added to component B (where acid and CMC are located. The two curves are coincident indicating that at least at low amount of resorcinol the latter does not accelerate at all the curing of the MUF adhesive. In Table 4 are shown the results of tensile strength and % wood failure of MUF honeymoon adhesive systems of different F/(M+U) molar ratio and different M:U weight ratios without resorcinol and when 4% resorcinol (based on resin solids) is added to component B of the system. Resorcinol addition does not appear to have any signi®cant bearing on adhesive performance for resins of higher proportion of melamine (Table 4). An upgrading effect of resorcinol is rather evident only in the case of the MUF resin of lowest melamine content, namely the MUF 10:90 in Table 4, but as in this case the resin fails rather miserably to comply with most relevant standards both with and without resorcinol, the use of resorcinol has no bearing even in this case. It could be possible that the use of higher proportions of resorcinol might well be bene®cial, but this would be so at the expenses of having an

overengineered and overpriced resin for this application. In conclusion, resorcinol is not needed in honeymoon fastset adhesives when a top of the range, well formulated MUF resin, is used, neither as a performance support nor as an accelerator of curing. This con®rms again (Cremonini et al. 1996, Higuchi et al. 1994), with another type of phenol and for a different application, that MUF/phenols copolymers in general are not an effective way to improve an adhesive performance.

4 Conclusions In conclusion MUF-based, honeymoon-type, fast-setting, separate application adhesive systems, which do not need any resorcinol, are capable to perform as adhesives for structural exterior-grade joints and glulam and to satisfy all the requirements of the relevant adhesive speci®cations for such an application. The parameters which are shown to be determining are mainly the method of preparation of the MUF resin, if and once an excellent resin formulation is available, both the ratio of melamine to urea and the molar ratio having a lesser effect.

Fig. 6. Tensile strength increase as a function of time of beech joints (BS 1204, Part 1) bonded with MUF based honeymoon adhesive systems: effect of presence or absence of 3% resorcinol on total liquid resin. The two curves are coincident and cannot be distinguished Bild 6. Zeitlicher Anstieg der Zugfestigkeit von Buchenverbindungen, verklebt mit einem ZweikomponentenMUF-Harz vom Typ `Honeymoon' (Tests nach BS 1204, Teil 1) sowie Ein¯uû des Zusatz oder der Abwesenheit vom 4% Resorcin (bezogen aud Festharzanteil). Die beiden Kurven fallen zusammen und koÈnnen nicht voneinander unterschieden werden Table 4. Effect of resorcinol addition in component B, if any, on tensile strength and wood failure percentage of MUF resins used as honeymoon adhesive systems to bond beech strips Tabelle 4. Ein¯uû eines Resorcin-Zusatzes zu Komponente B auf Zugfestigkeit und Bruchrate von Buchenbrettern, die mit einem Zweikomponenten-MUF-Kleber vom Typ `Honeymoon' verleimt waren M:U Weight ratio + Resorc.

F/(M+U) Molar ratio

47:53 1.7 47:53+4% R 1.7 47:53 1.5 47:53+4% R 1.5 30:70 1.9 30:70+4% R 1.9 20:80 1.9 20:80+4% R 1.9 10:90 1.9 10:90+4% R 1.9 Standard requirements

Dry

24 h soak

2 h boil

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

Strength (N)

Wood failure (%)

2764 2812 2762 2688 2980 2938 2958 2925 2862 2875 ³2500

100 100 100 100 100 100 100 100 100 100 ³75

2526 3150 2423 2462 2550 2700 2390 2538 2036 2875 ³2000

100 100 100 100 100 100 90 100 70 100 ³75

2139 2325 2116 2188 1950 2000 2025 1875 400 1725 ³1500

90 100 80 100 80 80 60 60 0 30 ³75

References

Breitmaier E, Voelter W (1987) Carbon-13 NMR Spectroscopy, 3rd Edition, VCH, Weinheim Cremonini C, Pizzi A, Tekely P (1996) In¯uence of PMUF resins preparation method on their molecular structure and performance as adhesives for plywood. Holz Roh- Werkstoff 54(2): 85±88 Durand D, Bruneau C-M (1982a) Polymer 23: 69±73 Durand D, Bruneau C-M (1982b) Makromol. Chem. 183: 1021±1025 Ebewele RO, River BH, Myers GE, Koutsky JA (1991a) Flexibilising effects of methylene amines in UF resin adhesives, Part I. J. Appl. Polymer Sci. 42: 2997±3002 Ebewele RO, River BH, Myers GE, Koutsky JA (1991b) Flexibilising effects of methylene amines in UF resin adhesives, Part II. J. Appl. Polymer Sci. 43: 1483±1489 Garcia R, Pizzi A (1997) Cross-linked and entanglement networks in thermomechanical analysis of polycondensation resins. J. Appl. Polymer Sci. 70(6): 1111±1119 Higuchi M, Roh J-K, Tajima S, Irita H, Honda T, Sakata I (1994) Polymeric structures of melamine-based composite adhesives, in Proceedings No. 4735 of the Adhesives and bonded wood symposium, Forest Products Society, Madison, pp 429± 449 Kamoun C, Pizzi A (2000) Particleboard I.B. forecast by TMA bending in MUF adhesives curing. Holz Roh- Werkstoff 58: 288± 289 Kamoun C, Pizzi A, Garcia R (1998) The effect of humidity on cross-linked and entanglement networking of formaldehydebased wood adhesives. Holz Roh- Werkstoff 56: 235±243 Odian G (1994) Principles of Polymerization, Wiley, New York Pizzi A (1983) Wood Adhesives Chemistry and Technology Vol. 1 (A. Pizzi Ed.). Marcel Dekker Inc., New York Pizzi A (1989) Wood Adhesives Chemistry and Technology Vol. 2 (A. Pizzi Ed.). Marcel Dekker Inc., New York Pizzi A (1994) Advanced Wood Adhesives Technology. Marcel Dekker Inc., New York Pizzi A (1997) On the correlation of some theoretical and experimental parameters in polycondensation cross-linked network. J. Appl. Polymer Sci. 63: 603±617 Pizzi A (1999) Molecular mechanics/dynamics and adhesion, chapter 3 in Adhesion Promotion Techniques (K. L. Mittal and A. Pizzi Eds.), Marcel Dekker Inc., New York

Pizzi A, Cameron F-A (1984) Fast-set adhesives for glulam. Forest Prod. J. 34(9): 61±68 Pizzi A, Probst F, Deglise X (1997) Molecular mechanics modelling of interfacial energy and ¯exibility. J. Adh. Sci. Techn. 11(4): 573±590 Pizzi A, Rossouw DduT, Knuffel W, Singmin M (1980) ``Honeymoon'' phenolic and tannin-based fast setting adhesive systems for exterior grade ®ngerjoints. Holzforschung Holzverwertung 32(6): 140±151 Properzi M (2002) Impiego di resine amminoplastiche MUF nella realizzazione di sistemi di incollaggio di tipo ``Luna di Miele''. Doctoral thesis in cotutele, University of Nancy 1, France, and University of Firenze, Italy Ramberg W, Osgood WR (1943) Description of stress-strain curves by three parameters. NACA tech. Note 902 Scopelitis E, Pizzi A (1993) The chemistry and development of branched PRF wood adhesives of low resorcinol content. J. Appl. Polymer Sci. 47: 351±360 Soulard C, Kamoun C, Pizzi A (1998) Uron and uron-ureaformaldehyde interior wood adhesives. Holzforschung Holzverwertung 50(5): 89±94 Soulard C, Kamoun C, Pizzi A (1999) Uron and uron-ureaformaldehyde resins. J. Appl. Polymer Sci. 72(2): 277±289 Steele PH, Kreibich RE, Steynberg PJ, Hemingway RW (1998) Finger jointing green southern yellow pine with a soy-based adhesive. Adhesives Age. 41(10): 49±53 Stephanou A, Pizzi A (1994) Some drawbacks of an alternative type of honeymoon fast-set adhesive system for ®ngerjointing. Holzforschung Holzverwertung. 45(4): 73±74 US patent 5,674,338 ± 1997 British Standard BS 1204, parts 1 and 2. 1982. Speci®cation for synthetic resins adhesives European Norm EN 301 (1992) Adhesifs de nature phenolique et aminoplast pour structures portants en bois: classi®cation et exigences de performance. EN 302-1. Adhesifs pour structures portants en bois: methodes d'essais South African Bureau of Standards SABS 1349 ± 1981. Phenolic and aminoplastic resin adhesives for the laminating and ®ngerjointing of timber

421