Chemical Modification of Lignocellulosic Fibers To Produce High-Performance Composites

Transcription

1 In: Glass, J. Edward; Swift, Graham, eds. Agricultural and synthetic polymers Biodegradability and utilization. ACS symposium series 433. American Chemical Society197th national meeting; 1989 April 9-14; Dallas, TX. Washington, DC: American Chemical Society; 199. Chapter 21. Chapter 21 Chemical Modification of Lignocellulosic Fibers To Produce High-Performance Composites Roger M. Rowell Forest Products Laboratory, Forest Service, U.S. Department of Agriculture, Madison, WI The performance properties of composites made from wood and other lignocellulosic materials can be greatly improved by changing the basic chemistry of the cell wall polymers. This paper reviews published research on reducing dimensional instability and susceptibility to degradation by biological organisms, heat, and ultraviolet radiation to produce high-performance lignocellulosic composites based on acetylation of the furnish before product formation. Wood and other lignocellulosic materials are three-dimensional, polymeric composites made up primarily of cellulose, hemicelluloses, and lignin. These polymers make up the cell wall and are responsible for most of the physical and chemical properties of these materials. Wood and other lignocellulosic materials have been used as engineering materials because they are economical, renewable, and strong and have low processing-energy requirements. They have, however, several undesirable properties, such as dimensional instability due to moisture sorption with varying moisture contents, biodegradability, flammability, and degradability by ultraviolet light, acids. and bases. These properties are all the result of chemical reactions involving degradative environmental agents. Because these types of degradation are chemical in nature, it should be possible to eliminate them or decrease their rate by modifying the basic chemistry of the lignocellulosic cell wall polymers. This chapter not subject to U.S. copyright Published 199 American Chemical society

2 21. ROWELL Chemical Modification of Lignocellulosic Fibers 243 Most research on chemical modification of lignocellulosic materials has focused on improving either the dimensional stability or the biological resistance of wood. This paper reviews the research on these properties for wood and other lignocellulosic composites and describes opportunities to improve fire retardancy and resistance to ultraviolet degradation. Reaction Chemistry Reactive organic chemicals can be bonded to cell wall hydroxyl groups on cellulose. hemicelluloses, and lignin. Much of our research has involved simple epoxides (1-3) and isocyanates (4). but most of our recent effort has focused on acetylation. Acetylation studies have been done using fiberboards (5,6). hardboards (7-11). particleboards (12-2). and flakeboards (21-23). using vapor phase acetylation (8,24,-26), liquid phase acetylation (1,27), or reaction with ketene (28). If this acetylation system does not include a strong catalyst or cosolvent, only the easily accessible hydroxyl groups will be acetylated. We developed an acetylation system that uses no strong catalyst or cosolvent and probably acetylates only easily accessible hydroxyl groups (27). Several lignocellulosic fibers were acetylated using this procedure: reaction times from 15 min to 4 h were used on Southern Pine, aspen, bamboo (29). bagasse (3), jute (31), pennywort, and water hyacinth (32). All the lignocellulosic materials used were easily acetylated. Acetyl content resulting from acetylation plotted as a function of time shows all data points fitting a common curve (Fig. 1). A maximum weight percent gain (WPG) of about 2 was reached in a 2-h reaction time, and an additional 2 h increased the weight gain only by 2 to 3 percent. Without a strong catalyst, acetylation using acetic anhydride alone levels off at approximately 2 WPG for the softwoods, hardwoods, grasses, and water plants. Moisture Sorption Sorption of moisture is due mainly to hydrogen bonding of water molecules to the hydroxyl groups in the cell wall polymers. By replacing some of the hydroxyl groups on the cell wall polymers with acetyl groups, the hydroscopicity of the lignocellulosic material is reduced. Table I shows the equilibrium moisture content (EMC) of several lignocellulosic materials at 65 percent relative humidity (RH). Reduction in EMC at 65 percent RH of acetylated fiber referenced to unacetylated fiber plotted as a function of the bonded acetyl content is a straight line (Fig. 2). Although the points shown in Figure 2 come from many different lignocellulosic materials. they

3 244 AGRICULTURAL AND SYNTHETIC POLYMERS Figure 1--Rate of acetylation of various lignocellulosic materials. O, Southern Pine; +, aspen:, bamboo;, bagasse:, jute: +, pennywort:, water hyacinth.

4 21. ROWELL Chemical Modification of Lignocellulosic Fibers 245 Figure 2--Reduction in equilibrium moisture content (EMC) as a function of bonded acetyl content for various acetylated lignocellulosic materials. O, Southern Pine; +, aspen;, bamboo:, bagasse; jute: +, pennywort;, water hyacinth.

5 246 AGRICULTURAL AND SYNTHETIC POLYMERS Table I. Equilibrium Moisture Content (EMC) of Various Acetylated Lignocellulosic Materials (65 Percent RH, 27 C) Reaction weight Acetyl gain content EMC Material (percent) (percent) (percent) Southern Pine Aspen Bamboo Bagasse Jute Pennywort Water hyacinth all fit a common line. A maximum reduction in EMC is achieved at about 2 percent bonded acetyl. Extrapolation of the plot to 1 percent reduction in EMC would occur at about 3 percent bonded acetyl. Because the acetate group is larger than the water molecule, not all hygroscopic hydrogen-bonding sites are covered. The fact that EMC reduction as a function of acetyl content is the same for many different lignocellulosic materials indicates that reducing moisture sorption and. therefore, achieving cell wall stability are controlled by a common factor. The lignin, hemicellulose, and cellulose contents of all the materials plotted in Figure 2 are different (Table II). Earlier results showed that the bonded acetate was mainly in the lignin and hemicelluloses (33) and that isolated wood cellulose does not react with uncatalyzed acetic anhydride (34). Acetylation may be controlling the moisture sensitivity due to the lignin and hemicellulose polymers in the cell wall but not reducing the sorption of moisture in the cellulose polymer because

6 21. ROWELL Chemical Modification of Lignocellulosic Fibers 247 Table II. Chemical Composition of Some Lignocellulosic Materials Composition (percent) Sugars Material Lig- Cellu- Glu- Xylose Galac- Arabi- Mannose Uronic Extrac- Ash Acetyl nin a lose cose tose nose acids tives b Southern Pine Aspen Bamboo Bagasse Jute Pennywort Water hyacinth a Klason. b 6 h of reflux, benzene ethanol.

7 248 AGRICULTURAL AND SYNTHETIC POLYMERS 1. these materials vary widely in their lignin, hemicellulose, and cellulose content, 2. acetate is found mainly in the lignin and hemicellulose polymer, and 3. isolated cellulose does not acetylate by the procedure used. Dimensional Stability Dimensional instability, especially in the thickness direction. is a greater problem in lignocellulosic composites than in solid wood because composites undergo not only normal swelling (reversible swelling) but also swelling caused by the release of residual compressive stresses imparted to the board during the composite pressing process (irreversible swelling). Water sorption causes both reversible and irreversible swelling, with some of the reversible shrinkage occurring when the board dries. Dimensional instability of lignocellulosic composites has been the major reason for their restricted use. We are in the process of producing fiberboards from various types of acetylated lignocellulosic fibers. Most of our research has been on pine or aspen particleboards or flakeboards. so the data presented here on dimensional stability and biological resistance come mainly from these types of boards. The rate of swelling in liquid water of an aspen flakeboard made from acetylated flakes and phenolic resin (3) is shown in Figure 3. During the first 6 min, control boards swelled 55 percent in thickness, while the board made from flakes acetylated to 17.9 WPG swelled less than 2 percent. During 5 days of water soaking, the control boards swelled more than 66 percent, while the 17.9-WPG board swelled about 6 percent. Control boards made from bamboo particles using a phenolic adhesive swelled about 1 percent after 1 h, 15 percent after 6 h, and 2 percent after 5 days. Particleboards made from acetylated bamboo particles swelled about 2 percent after 1 h and only 3 percent after 5 days (35). Thickness changes in a six-cycle water-soaking/ovendrying test for an acetylated aspen flakeboard are shown in Figure 4 (27). Control boards swelled more than 7 percent in thickness during the six cycles, compared with less than 15 percent for a board made from acetylated flakes. Acetylation greatly reduced both irreversible and reversible swelling. In a similar five-cycle water-soaking/ovendrying test on bamboo particleboards. control boards swelled more than 3 percent, while boards made from acetylated particles swelled about 1 percent. Figure 5 (27) shows changes in thickness of aspen flakeboards made from control and acetylated flakes using a phenolic adhesive at different relative humidities. After four cycles of 3 to 9 percent RH, control boards swelled 3 percent in thickness, while acetylated boards at 17.9 WPG swelled about 5 percent. The results of both liquid water and water vapor tests show that acetylation of lignocellulosic materials greatly improves dimensional stability of composites made from these materials.

8 21. ROWELL Chemical Modification of Lignocellulosic Fibers 249 Figure 3--Rate of swelling in liquid water of aspen flakeboard made from acetylated flakes. O = control; + = 7.3 WPG; = 11.5 WPG; = 14.2 WPG; + = 17.9 WPG.

9 25 AGRICULTURAL AND SYNTHETIC POLYMERS Figure 4--Changes in ovendry (OD) thickness in repeated water-soaking test of aspen flakeboard made from acetylated flakes. O = control; + = 7.3 WPG: = 11.5 WPG; = 14.2 WPG; + = 17.9 WPG.

10 21. ROWELL Chemical Modification of lignocellulosic Fibers 251 Figure 5--Changes in ovendry (OD) thickness at 3 percent and 9 percent relative humidity of aspen flakeboard made from acetylated flakes (27 C). O = control; + = 7.3 WPG; = 11.5 WPG; = 14.2 WPG; + = 17.9 WPG.

11 252 AGRICULTURAL AND SYNTHETIC POLYMERS Biological Resistance Chemical modification of wood composite furnish for biological resistance is based on the theory that the potentially degrading enzymes must directly contact the substrate and that the substrate must have a specific chemical configuration and molecular conformation. Reacting chemicals with the hydroxyl groups on cell wall polymers chemically changes the substrate so that the highly selective enzymatic reactions cannot take place. Chemical modification also reduces the moisture content of the cell wall polymers to a point where biological degradation cannot take place. Particleboards and flakeboards made from acetylated flakes have been tested for resistance to several different types of organisms. In a 4-week termite test using Reticulitermes flavipes (subterranean termites), boards acetylated at 16 to 17 WPG were very resistant to attack, but not completely so (17,36,37). This may be attributed to the severity of the test. However, since termites can live on acetic acid and decompose cellulose to mainly acetic acid, perhaps it is not surprising that acetylated wood is not completely resistant to termite attack. Chemically modified wood composites have been experimentally exposed to decay fungi in several ways. Untreated aspen and pine particleboards and flakeboards exposed to white-, soft-. and brown-rot fungi and tunneling bacteria in a fungal cellar were destroyed in less than 6 months, while particleboards and flakeboards made from furnish acetylated to greater than 16 percent acetyl weight gain showed no attack after 1 year (Table III) (36-38). In a standard 12-week single-culture soil-block test, control aspen flakeboards exposed to the white-rot fungus Trametes versicolor lost 34 percent weight. while acetylated flakeboards (17 percent acetyl weight gain) lost no weight (36-38). In exposure to the brown-rot fungus Tyromyces palustris, control aspen flakeboards lost only 2 percent weight when a phenol-formaldehyde adhesive was used but lost 3 percent weight when an isocyanate adhesive was used. If the flakeboards were water leached before the soil-block test was run. control boards made with phenol-formaldehyde adhesive lost 44 percent weight with the brown-rot fungus Gloeophyllum trabeum (34-36). This shows that the adhesive can influence results of fungal toxicity, especially in a small, closed test container. Weight loss resulting from fungal attack is the methodmost used to determine the effectiveness of a preservative treatment to protect wood composites from decaying. In some cases, especially for brown-rot fungal attack, strength loss may be a more important measure of attack since large strength losses are known to occur in solid wood at very low wood weight loss (39). Adynamic bending-creep test has been developed to determine strength losses when wood composites are exposed to a brown- or white-rot fungus (4). Using this bending-creep test on aspen flakeboards. control boards made with phenol-formaldehyde adhesive failed in an average of 71 days with T. palustris and 212 days with T. versicolor (41). At failure, weight losses averaged 7.8 percent for T. palustris and 31.6 percent for T. versicolor. Isocyanate-bonded control

12 21. ROWELL Chemical Modification of Lignocellulosic Fibers 253 Table III. Fungal Cellar Tests on Aspen Flakeboards Made from Control and Acetylated Flakes a,b Rating at intervals d Acetyl WPG C 2 mo 3 mo 4 mo 5 mo 6 mo 12 mo S/2 S/3 S/3 S/3 S/ S/ S/1 S/1 S/2 S/3 S/ S/ S/1 S/2 S/ S/ S/ a Nonsterile soil containing brown-, white-. fungi and tunneling bacteria. and soft-rot b Flakeboards bonded with 5 percent phenol-formaldehyde adhesive. c Weight percent gain. d Rating system:, no attack; 1, slight attack; 2, moderate attack; 3, heavy attack; 4, destroyed; S, swollen. flakeboards failed in an average of 2 days with T. palustris and 118 days with T. versicolor, with average weight loss at failure of 5.5 percent and 34.4 percent, respectively (41). Very little or no weight loss occurred with either fungi in flakeboards made using phenol-formaldehyde or isocyanate adhesive with acetylated flakes. None of these specimens failed during the test period. Deflection-time curves for flakeboards are shown in Figure 6. The curves show an initial increase of deflection for both control and acetylated flakeboards. then a stable zone, and finally, for control boards, a steep slope to failure. The study showed that less than 5 mm of deflection was caused by creep due to moisture alone (41). Mycelium fully covered the surfaces of isocyanate-bonded control flakeboards within 1 week, but mycelial development was significantly slower in phenol-formaldehyde-bonded control flakeboards. Both isocyanate- and phenol-formaldehyde-bonded acetylated flakeboards showed surface mycelium colonization during the test time, but the fungus did not attack the acetylated flakes. so little strength was lost. In similar bending-creep tests, both control and acetylated pineparticleboardsmadeusingmelamine-urea-formaldehydeadhesive failed because T. palustris attacked the adhesive in the glueline (42). Mycelium-invaded the inner part of all boards. colonizing in both wood and glueline in control boards but only in the glueline in acetylated boards. After a 16-week exposure to T. palustris, the internal bond strength of control aspen flakeboards made with phenol-formaldehyde

13 254 AGRICULTURAL AND SYNTHETIC POLYMERS Figure 6--Deflection-time curves of phenol-formaldehyde- (PF-) and isocyanate- (IS-) bonded flakeboards in bending-creep tests under progressive fungal attack by T. palustris (upper) and T. versicolor (lower). U = PF control: + = PF acetylated: c = IS control; O = IS acetylated.

14 21. ROWELL Chemical Modification of Lignocellulosic Fibers 255 adhesive was reduced more than 9 percent and that of flakeboards made with isocyanate adhesive was reduced 85 percent (43). After 6 months of exposure in moist unsterile soil, the same control flakeboards made with phenol-formaldehydeadhesive lost 65 percent of their internal bond strength and those made with isocyanate adhesive lost 64 percent internal bond strength. Failure was due mainly to great strength reductions in the wood caused by fungal attack. Acetylated aspen flakeboards lost much less internal bond strength during the 16-week exposure to T. palustris or 6-month soil burial. The isocyanate adhesive was somewhat more resistant to fungal attack than thephenol-formaldehydeadhesive. In the case of acetylated composites, loss in internal bond strength was mainly due to fungal attack in the adhesive and moisture, which caused a small amount of swelling in the boards. Acetylated pine flakeboards have also been shown to be resistant to attack in a marine environment (44). Control flakeboards were destroyed in 6 months to 1 year, mainly because of attack by Limnoria tripunctata, while acetylated boards showed no attack after 2 years. All laboratory tests for biological resistance conducted to this point show that acetylation is an effective means of reducing or eliminating attack by soft-, white-, and brown-rot fungi, tunneling bacteria, and subterranean termites. Tests are presently underway on several lignocellulosic composites in outdoor environments. Other Properties Other properties of lignocellulosic composites can be improved by changing the basic chemistry of the furnish (45). Acetylation has been shown to improve ultraviolet resistance of flakeboards. Table IV shows that acetylation greatly reduced weight loss and errosion rate due to loss of surface fiber from aspen flakeboards. The rate of errosion for boards made from acetylated flakes is half that of control boards. In outdoor tests, flakeboards made from acetylated pine flakes were still light yellow in color when control boards had turned dark orange to light gray. Table IV. Weight Loss and Rate of Erosion for Acetylated Aspen Flakeboard After Accelerated Weathering Weight loss (percent) after weathering period Erosion rate at 2,4 h Sample 14 h 5 h 8 h 1,7 h (µm/h) Control Acetylated

15 256 AGRICULTURAL AND SYNTHETIC POLYMERS Acetylation does not change the fire properties of lignocellulosic materials. In thermogravametric analysis, acetylated and control pine sawdust pyrolyzed at the same temperature and rate (46). The heat of combustion and rate of oxygen consumption were also the same for control and acetylated specimens. showing that the acetyl group added to the cell wall has approximately the same carbon. hydrogen. and oxygen content as the cell wall polymers. Reactive fire retardants can be bonded to the cell wall hydroxyl groups. The effect would be an improvement in dimensional stability, biological resistance, and fire retardancy. Conclusions Dimensional instability and susceptibility to degradation by biological organisms, heat. and ultraviolet radiation can be greatly reduced by modification of lignocellulosic cell wall polymers. These modifications result in a furnish that can be converted into composites of any desired shape, density, and size and provide an opportunity for a manufacturer to distinguish a product line based on quality. uniformity, and performance. Of the various reaction systems studied to date. a simple noncatalyzed acetylation process appears to be closest to implementation on a commercial basis. Composites made from lignocellulosic materials have been restricted from many markets because of their moisture sorption, dimensional instability, and to a lesser extent, biological degradation. These negative properties can be overcome, allowing flakes. particles, and fiber from wood and agricultural residues to find markets related to high-performance composites. For some applications, a combination of materials may be required to achieve a composite with the desired properties and performance. Property-improved lignocellulosic fibers can be combined with materials such as metal, glass, plastic, natural polymers, and synthetic fiber to yield a new generation of composite materials. New composites will be developed that utilize the unique properties obtainable by combining many different materials. This trend will increase significantly in the future. Literature Cited 1. Rowell, R. M.; Tillman, A-M.; Zhengtian, L. Wood Sci. Technol. 1986, 2, Rowell, R. M.; Ellis, W. D. Reaction of Epoxides with Wood. Res. Pap. FPL 451. U.S. Department of Agriculture, Forest Serv., Forest Prod. Lab., p. 3. Tillman, A-M. J. Wood Chem. Technol. (In press). 4. Rowell, R. M.: Ellis, W. D. Am. Chem. Soc. Symp. Ser. 1981, 172, Bekere, M.; Shvalbe, K.; Ozolinya, I. Latvijas Lauksaimniecibas Akademija. Raksti 1978, 163, Bristow, J. A.; Back, E. L. Svensk Papperstid. 1969, 72, Sudo, K. Mokuzai Gakkaishi 1979, 25, Klinga, L. O.; Tarkow, H. Tappi 1966, 49,

16 21. ROWELL Chemical Modification of Lignocellulosic Fibers Zhang, G.; Yin, S.; Wang, W.; Xu, R.; Chen, C. J. Nanjing Technol. College Forest Products 1981, 3, Ozolinya. I. O.; Shvalbe. K. P.: Bekere. M. R.: Shnyutsinsh, F. A.; Mitsane, L. V.; Karlsone, I. M. USSR Patent , Shvalbe, K. P.; Ozolinya, I. O.; Bekere, M. R.; Hitsane, L. V.; Karlsone, I. M.; Dudin'sh, M. M. USSR Patent , Rowell, R. M. In Wood Modification; Lawniczak, M., Ed.; Polish Acad. Sci.: Posnan, Poland, 1985; Tillman, A.-M.; Simonson, R.; Rowell, R. M. In Wood Modification; Lawniczak, M., Ed.; Polish Acad. Sci.: Posnan, Poland, 1985; Nishimoto, K.; Imamura, Y. Mokuzai Kogyo 1985, 4, Yoshida, Y.: Kawai, S.; Imamura, Y.; Nishimoto, K.; Satou, T.; Nakaji, M. Mokuzai Gakkaishi 1986, 32, Rowell, R. M.; Simonson, R.; Tillman, A.-M. Paperi ja Puu 1986, 68, Imamura, Y.; Nishimoto, K.; Yoshida, Y.; Kawai, S.; Sato, T.; Nakaji, M. Wood Res. 1986, 73, Rowell, R. M.; Imamura, Y.; Kawai, S.; Norimoto, M. Wood Fiber Sci. 1989, 21, Kiguchi, M.; Suzuki, M. Mokuzai Gakkaishi 1985, 31, Imamura, Y.; Nishimoto, K. Mokuzai Gakkaishi 1987, 33, Youngquist, J. A.; Krzysik, A.; Rowell, R. M. Wood Fiber Sci , Youngquist, J. A.; Rowell, R. M.; Krzysik, A. Holz als Roh- und Werkstoff 1986, 44, Rowell, R. M.; Plackett, D. New Zealand J. Forest Sci. (In press.) 24. Arora, M.; Rajawat, J. S.; Gupta, R. C. Holzforschung and Holzverwertung 1981, 33, Rowell, R. M.; Tillman. A.-M.; Simonson, R. J. Wood Chem. Technol. 1986, 6, Rowell, R. M.; Simonson, R.; Tillman, A.-M. Nordic Pulp and Paper Res. J. 1986, 2, Rowell, R. M.; Tillman. A.-M.; Simonson, R. J. Wood Chem. Technol. 1986, 6, Rowell, R. M.; Wang, R.H.S.; Hyatt, J. A. J. Wood Chem. Technol. 1986, 6, Rowell, R. M.; Norimoto, M. J. Jap. Wood Res. Soc. 1987, 33, Rowell, R. M.; Keany, F. Wood and Fiber Sci. (In press.) Rowell, R. M.; Simonson, R.; Tillman, A. M. European Patent Application , Rowell, R. M.; Rowell, J. S. In Cellulose and Wood; C. Schurch, Ed.; Wiley, New York, 1989; Rowell, R. M. Wood Sci. 1982, 15, Rowell, R. M. Unpublished data. 35. Rowell, R. M.; Norimoto, M. Mokuzai Gakkaishi 1988, 34, Rowell, R. M.; Esenther, G. R.; Nicholas, D. D.; Nilsson, T. J. Wood Chem. Technol. 1987, 7,

17 258 AGRICULTURAL AND SYNTHETIC POLYMERS 37. Rowell, R. M.; Esenther, G. R.; Youngquist, J. A.; Nicholas, D. D.; Nilsson, T.; Imamura, Y.; Kerner-Gang, W.; Trong, L.; Deon, G.; Proc. Symp. Protection of Wood-Based Composite Products. 1988, Nilsson. T.; Rowell, R. M.; Simonson, R.; Tillman, A.-M. Holsforschung 1988, 42, Cowling, E. B. Comparative Biochemistry of the Decay of Sweetgum Sapwood by White-Rot and Brown-RotFungus, Technol. Bull. 1258, U.S. Department of Agriculture, Forest Serv., p Imamura, Y.; Nishimoto, K. J. Soc. Materials Sci. 1985, 34, Rowell, R. M.; Youngquist, J. A.; Imamura, Y. Wood Fiber Sci. 1988, 2, Imamura, Y.; Rowell, R. M.; Simonson, R.; Tillman, A.-M. Paperi ja Puu 1988, 9, Imamura, Y.; Nishimoto, K.; Rowell, R. M. Mokuzai Gakkaishi 1987, 33, Johnson, B. R.; Rowell, R. M. Mater. und Org. 1988, 23, Rowell, R. M. In Chemistry of Solid Wood; Rowell, R. M., Ed.; Advances in Chemistry Series 27; American Chemical Society: Washington, DC, 1984; Chapter 4. pp Rowell, R. H.; Susott, R. A.; De Groot, W. G; Shafizadeh, F. Wood and Fiber Sci. 1984, 16, RECEIVED January 22, 199 Printed on recycled paper