Biomacromolecules 2004, 5, 1262-1269
1262
Designing New Materials from Wheat Protein Dara L. Woerdeman,*,† Wim S. Veraverbeke,‡ Richard S. Parnas,§ Dave Johnson,§ Jan A. Delcour,‡ Ignaas Verpoest,† and Christopher J. G. Plummer| Metallurgy and Materials Engineering Department and Laboratory of Food Chemistry, Katholieke Universiteit Leuven, B-3001 Heverlee, Belgium, Institute of Materials Science, University of Connecticut, 06269 Storrs, Connecticut, and Laboratoire de technologie des composites et polyme` res (LTC), Ecole Polytechnique Fe´ de´ rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland Received December 15, 2003; Revised Manuscript Received February 19, 2004
We recently discovered that wheat gluten could be formed into a tough, plasticlike substance when thiolterminated, star-branched molecules are incorporated directly into the protein structure. This discovery offers the exciting possibility of developing biodegradable high-performance engineering plastics and composites from renewable resources that are competitive with their synthetic counterparts. Wheat gluten powder is available at a cost of less than $0.5/lb, so if processing costs can be controlled, an inexpensive alternative to synthetic polymers may be possible. In the present work, we demonstrate the ability to toughen an otherwise brittle protein-based material by increasing the yield stress and strain-to-failure, without compromising stiffness. Water absorption results suggest that the cross-link density of the polymer is increased by the presence of the thiol-terminated, star-branched additive in the protein. Size-exclusion high performance liquid chromatography data of molded tri-thiol-modified gluten are consistent with that of a polymer that has been further cross-linked when compared directly with unmodified gluten, handled under identical conditions. Remarkably, the mechanical properties of our gluten formulations stored in ambient conditions were found to improve with time. Introduction There have been numerous studies of gluten in the form of a thin film.1-5 Films have been cast from gluten protein dispersions in water under different pH conditions or in ethanol.6 It has also been demonstrated that plasticizing agents can improve film flexibility and decrease brittleness.5-7 The Tg of wheat gluten decreased from 183 °C in its dry state to about 66 °C when plasticized with water at a fractional weight content of 0.1. Water may have reduced the Tg by mechanisms other than those commonly associated with plasticization.7 To improve the toughness of the biobased polymer without compromising its other mechanical properties (e.g., strength and stiffness), we employ a thiolterminated, star-branched additive, a highly mobile reactive modifier that can covalently bond with reactive sites in the gluten protein. In the thiol-terminated, star-branched molecule (Figure 1), the poly(ethylene glycol) branches serve as spacer arms. Plasticizer efficiency is known to increase with arm length.8,9 Most of the studies presented thus far have relied on plasticizers that, at most, form only hydrogen bonds with the gluten polypeptide chains.2,7-12 The use of a chemical cross-linker to modify properties of a protein material has * Corresponding author. Current address: R&D Green Materials, 100 Harvest Circle, Williamsburg, VA 23185, U.S.A. E-mail:
[email protected]. † Metallurgy and Materials Engineering Department, Katholieke Universiteit Leuven. ‡ Laboratory of Food Chemistry, Katholieke Universiteit Leuven. § University of Connecticut. | Ecole Polytechnique Fe ´ de´rale de Lausanne (EPFL).
Figure 1. Schematic of the thiol-terminated, star-branched molecule (m + n + o ) 20) used to improve the mechanical properties of the wheat gluten protein polymer. Unlike other low-molecular-weight thiolbased compounds, the odor produced by the thiol-terminated, starbranched molecule depicted above is minimal because of its low vapor pressure.
been reported as well, but in these other cases, induction of a cross-linking reaction with the protein chains requires the use of either a catalyst5 or an aggressive radiation treatment.13 In a more recent study,14 Pommet et al. explored the use of fatty acids as a plasticizing agent for wheat gluten. During the past few decades, the cereal science community has made a concerted effort to elucidate the properties of wheat gluten proteins, because it is these proteins that possess the unique properties required for making the vital dough used in bread making and pasta making.15,16 The rheological properties of wheat gluten strongly affect the behavior of bread and pasta doughs and are expected to play an important role in any process developed to produce plastics and composites from wheat gluten. The first major wheat gluten study was published in 1907,17 which described gluten’s unique properties and how it forms. Several studies have demonstrated the existence of a polymeric glutenin fraction (with molecular weights
10.1021/bm034530+ CCC: $27.50 © 2004 American Chemical Society Published on Web 04/08/2004
New Materials from Wheat Protein
Biomacromolecules, Vol. 5, No. 4, 2004 1263
Figure 2. Proposed model of glutenin structure depicting the disulfide bridges between the individual glutenin polymer subunits. Each bar represents a different polypeptide chain or glutenin subunit. Adapted from and reprinted with permission from ref 16. Copyright 2002 Taylor & Francis Group.
ranging from ca. 80 000 to several million) comprised of high-molecular-weight glutenin subunits (HMW-GS) and low-molecular-weight glutenin subunits (LMW-GS; with molecular weights between 30 000 and 80 000) and monomeric gliadin proteins (MW < 50 000).15-19 A portion of the polymeric fraction can only be solubilized with the help of a reducing agent or by sonication.15,19 Results have shown that it is this polymeric fraction that gives rise to some of gluten’s unusual properties.15,20 Glutenin is typically modeled as a heterogeneous mixture of disulfide (S-S)-stabilized polymers, with a complex polymeric structure and broad molecular size distribution.21-23 Conversely, the gliadins act as plasticizers and weaken interactions between glutenin chains.24,25 The structure of the main classes of glutenin subunits, the relationship of the glutenin subunits to the gliadin monomers, and the pattern of disulfide linkages in the glutenin polymers were elaborated in a recent review article by Veraverbeke and Delcour.16 These subunits are linked together by disulfide bonds between cysteine residues to form the very large glutenin polymers (Figure 2). The main point to note is the very large fraction of cysteine residues bound through intramolecular disulfide linkages. This intramolecular organization hinders covalent bonding between separate glutenin polymers, leaving larger scale organization to weaker bonds such as hydrogen bonds. Many attempts have been made in the past to reveal the basic structure of gluten and to understand the origin of its
viscoelastic properties upon hydration. However, it is evidently impossible to completely solubilize gluten proteins without disrupting their native structure.16 As a result, it has been difficult to unequivocally determine the structure of gluten. In more fundamental studies on gluten proteins, a reducing agent, such as β-mercapto-ethanol or dithiothreitol (DTT), is often used to break the disulfide bonds that link the various glutenin subunits to each other (Figure 2). Results and Discussion To test whether the thiol-terminated, star-branched molecule used in the present work (Figure 1) readily behaved as a reducing agent when mixed into a gluten protein slurry, we performed a series of mixograph tests to compare the rheological behavior of a wheat flour dough doped with the reducing agent, DTT, with that of a wheat flour dough doped with the thiol-terminated, star-branched molecule, at identical molar ratios. (Converting existing intramolecular disulfide bonds to free sulfhydryl groups is a preliminary requirement for success of an intermolecular disulfide cross-linker.) In the Mixograms (Figure 3), the torque required to turn a mixing paddle through the dough is plotted versus time, the point at which the dough is first developed from flour and water. The Mixogram in Figure 3a shows that in the control (unmodified dough), the torque remains at a high and constant level for a relatively long time. The Mixogram in Figure 3b shows that DTT, the known reducing agent,
1264
Biomacromolecules, Vol. 5, No. 4, 2004
Woerdeman et al.
Figure 4. Tensile testing data for native gluten and tri-thiol-modified gluten [gluten modified with 5.8% (w/w) thiol-terminated, star-branched additive].
Figure 3. Dough mixgraphs using Legat as a base flour. Hydrated with (a) plain deionized water (control), (b) the DTT reducing agent solution, and (c) the thiol-terminated, star-branched molecule solution.
decreases the resistance of the dough to mixing. The Mixogram in Figure 3c (thiol-terminated, star-branched additive) is clearly similar to that of Figure 3b, which allows us to deduce that the thiol-terminated, star-branched molecule indeed acts as a reducing agent in breaking glutenin disulfide bonds. The effect that DTT or the thiol-terminated, starbranched molecule had on the dough was consistent with that observed in earlier studies,26 where DTT was found to reduce optimal dough mixing time and resistance to breakdown. An important next step was to understand the role of the thiol-terminated, star-branched molecule in governing the mechanical properties of gluten after it was converted into an engineering plastic. Conversion was achieved by applying heat and pressure to gluten powder, which yielded a rigid plasticlike material. The standard tensile test was used to gauge the material stiffness, strength, and work of fracture. Figure 4 shows tensile stress-strain curves from two specimens, one of which is representative of tri-thiolmodified gluten and the other of which is typical of unmodified gluten. In both cases, a very low constant strain rate of 0.01/min was applied, and the resulting load on the specimen was measured. The stress rose to the breaking point, and the sample failed. The specimens of gluten in which the thiol-terminated, star-branched molecule was
incorporated failed at roughly 4.3% strain, compared with the 2.0% failure strain of the unmodified gluten. One standard of comparison particularly important for structural application is the work of fracture per unit volume of material, which is simply the integral under the stressstrain curve. The integral under the plain gluten curve (blue) is roughly 0.3 MPa. However, the integral under the tri-thiolmodified gluten curve (pink) is approximately 1.3 MPa, a fourfold increase improvement. Furthermore, one can note from the initial slope of the stress-strain curve that the stiffness was not reduced by the addition of the thiolterminated, star-branched molecule, a point of great improvement relative to previous work where plasticizers were added. As an aside, Smith et al. have proposed a mechanism behind the strength and toughness of natural adhesives, fibers, and composites.27 Their atomic force microscopy studies on individual titin protein molecules yielded sawtooth-likeextension curves, which, according to the authors, might have reflected the “successive opening of intrachain loops or folded domains within a single molecule, or the successive release of sacrificial interchain bonds holding a cross-linked multichain matrix together”.27 In future work, we will explore if the descriptive model presented by Smith and co-workers could to some extent be applied to a protein-based polymer as heterogeneous and polydisperse as commercial wheat gluten. The modified and unmodified samples showed similar water absorption properties up to about 80 h (Figure 5). At about 300 h, the unmodified specimens had absorbed more than twice as much water as the tri-thiol-modified specimens. The water absorption measurements are consistent with an increase in the cross-link density of the gluten protein polymer, because the tri-thiol-modified gluten specimen was found to absorb less than half the amount of water as the unmodified gluten specimen. One can anticipate this result if the thiol-terminated, star-branched additive is in some way incorporated into the protein network. Recent size-exclusion high performance liquid chromatography (SE-HPLC) studies (Figure 6) further substantiate our hypothesis that the thiol-terminated, star-branched additive serves first as a reducing agent and later as a cross-
New Materials from Wheat Protein
Figure 5. Normalized water uptake as a function of time. After 300 h, the plain gluten specimen (top) absorbed more than twice as much water as the tri-thiol-modified gluten specimen (bottom), indicating increased cross-linking in the protein network.
Figure 6. SE-HPLC chromatographs illustrating the function of the thiol-terminated, star-shaped molecule before and after molding. The largest peak around 21 min is believed to be a signature of the lowmolecular-weight gliadins (or glutenin subunits) in the protein. (a) Molding appears to reduce the material’s overall solubility. Moreover, the data of molded tri-thiol-modified gluten are consistent with those of a polymer that has been further cross-linked. (b) Similar samples that were first treated with 1% (w/v) DTT to cleave all S-S bonds prior to injecting the material into the HPLC column. Because SEHPLC is a highly quantitative technique, the fact that the two molded samples exhibit reduced peak areas at 21 min could be an indication that chemical bonds apart from S-S bonds form at elevated pressures and temperatures.
linking agent, once the modified gluten powder is in the heated mold. In the first set of SE-HPLC experiments (Figure 6a), the effect of thiol-modification was studied for protein samples before and after molding. Increasing time on the x axis inversely corresponds to decreasing molecular weight
Biomacromolecules, Vol. 5, No. 4, 2004 1265
species eluting from the column. Gliadins and LMW-GS emerge from the column at 20-22 min, oligomeric proteins and HMW-GS emerge at 18-20 min, and extractable polymeric proteins (the highest molecular weight species) emerge from the column starting at 13 min. A comparison of the molecular weight distributions of sodium dodecyl sulfate (SDS)-extractable proteins from unmodified (Figure 6a, black curve) and thiol-modified (Figure 6a, green curve) gluten indicates that before compression molding, the thiol-terminated, star-branched additive decreases the molecular weight of the polymers and increases the levels of monomers. This result is to be expected if the additive reduces a significant fraction of the disulfide bonds in the protein, and, consequently, LMW-GS and possibly HMW-GS are liberated. We believe that the molding process alone induces some level of interchain bonding among the protein chains28 because the level of SDS-extractable protein decreases dramatically and the monomers become incorporated into polymers (Figure 6a, red curve). However, the effect is amplified in the presence of the thiol-terminated, starbranched additive, as evidenced by the differences in the amplitude and the molecular weight distribution of the “unmodified (Figure 6a, red curve) gluten specimen after molding” relative to that of the “thiol-modified (Figure 6a, yellow curve) gluten specimen after molding”. This result is to be expected if the additive induces further cross-linking during the molding process. A second SE-HPLC experiment (Figure 6b) was performed on reduced protein samples, which were obtained by adding 1.0% (w/v) DTT to the extraction fluid. (Hence, Figure 6a illustrates nonreduced material, while Figure 6b illustrates reduced material.) The purpose for performing this study was to further examine the effects of molding (high pressure and temperature conditions) on both the unmodified gluten and the tri-thiol-modified gluten. A comparison of the SE-HPLC profiles of reduced unmodified (Figure 6b, black curve) and thiol-modified (Figure 6b, green curve) gluten before molding indicates that a similar number of glutenin subunits were liberated as a result of the DTT treatment and that the presence of the thiol-terminated, star-branched additive did not appear to interfere with the DTT reducing agent in any way. Comparing the SE-HPLC profiles of reduced unmodified gluten before (Figure 6b, black curve) and after (Figure 6b, red curve) molding indicates that the molding resulted in the liberation of fewer glutenin subunits. A likely explanation is that interchain chemical interactions apart from disulfide bond formation took place in the gluten polymer matrix at elevated pressures and temperatures. Another possible explanation is that the DTT was unable to reduce all of the S-S bonds in the molded material, even though the molded materials had been crushed to 400-µm-sized particles prior to testing. Comparison of the SE-HPLC profiles of reduced unmodified (Figure 6b, red curve) and thiol-modified (Figure 6b, yellow curve) gluten after molding once again indicates that the tri-thiol-modified gluten polymer is less soluble than the unmodified gluten polymer. Selecting an appropriate modifier for gluten proved to be only one of the many challenges in this study. A variety of
1266
Biomacromolecules, Vol. 5, No. 4, 2004
Woerdeman et al.
Figure 7. (A) Optical micrograph depicting two gluten specimens compression molded at 150 °C/25 bar for 5 min. When the plain gluten polymer is compression molded from a doughlike state (left), with an H2O content of roughly 60%, a rubbery material will result. Molding from the dry powder state (right) will result in a plaque that is stiff but brittle. (B) Optical microscopy (a) and transmission electron microscopy (b) of a molded gluten specimen with micrometer-sized inclusions believed to be starch granules. (C) Optical microscopy of a rather atypical region of a molded gluten specimen showing a locally high concentration of inclusions in an unusual arrangement.
new hurdles associated with the molding step became evident, all having to do with the increased complexity of protein-based polymers in comparison with their synthetic counterparts. For example, even seemingly “dry” gluten powder contains low levels of moisture (roughly 10%). And we have observed firsthand that if wheat gluten contains significant amounts of water (approaching a doughlike consistency) at the time that it is molded into a specimen, the material will retain much of that water even when molding temperatures are set above 100 °C (Figure 7A). In hindsight, however, the complication introduced by water should not come as a surprise, particularly because standard
epoxy is known to interact with water in a variety of complex ways, as others have illustrated spectroscopically.29 Other unknowns include the extent to which the gluten proteins interact with the other constituents present in commercial wheat gluten. In general, commercial wheat gluten contains roughly 75% protein, 10% starch and nonstarch polysaccharides,
