Material Properties of Films from Enzymatically Tailored Arabinoxylans

Rye arabinoxylan was first investigated by Preece and Hobkirk(14) and thereafter by Aspinall and Sturgeon(15) and Aspinall and Ross.(16) They conclude...
0 downloads 0 Views 165KB Size
Download PDF
2042

Biomacromolecules 2008, 9, 2042–2047

Material Properties of Films from Enzymatically Tailored Arabinoxylans Anders Ho¨ije,† Erik Sternemalm,† Susanna Heikkinen,‡ Maija Tenkanen,‡ and Paul Gatenholm*,† Department of Chemical and Biological Engineering, Biopolymer Technology, Chalmers University of Technology, SE-412 96 Go¨teborg, Sweden, and Department of Applied Chemistry and Microbiology, University of Helsinki, FIN-00014, Helsinki, Finland Received March 20, 2008; Revised Manuscript Received May 7, 2008

Rye arabinoxylan, with an initial arabinose to xylose (Ara/Xyl) ratio of 0.50, was enzymatically modified with R-L-arabinofuranosidase. Different enzyme dosages were used to prepare arabinoxylan samples with a gradient of arabinose content varying from Ara/Xyl ratio 0.50 to 0.20. The degree of polymerization of the arabinoxylans was not affected by the enzymatic treatment, as detected with SEC-MALLS. Arabinoxylan samples with an Ara/ Xyl ratio of 0.30 and below agglomerated in a water solution as seen by changes in light scattering. All samples, however, formed cohesive films upon drying, without addition of external plasticizers. The film from untreated arabinoxylan was completely amorphous; whereas films of the enzyme-treated arabinoxylans were semicrystalline with an increasing degree of crystallinity with decreasing arabinose content as determined by WAXS. Oxygen permeability measurements of the films showed that decreased arabinose content also resulted in lower oxygen permeability of the films. All films were strong and relatively stiff, but showed variations in strain at break. The moderately debranched film with an Ara/Xyl ratio of 0.37 had highest strain at break among all the films tested, yet was stiff and strong. This material also exhibited yielding and had stress/strain behavior similar to synthetic semicrystalline polymers, with a tendency to strain-induced crystallization. Such a combination of mechanical properties combined with oxygen barrier properties is very attractive for packaging applications.

Introduction Arabinoxylans are heteropolysaccharides found in almost all annual plants and also in softwood, which make them one of the most abundant biopolymers on earth. Xylans are polymers with a (1f4)-linked β-D-xylopyranosyl backbone with varying degree and type of substitution. The main chain can be substituted at positions C2 or C3 or both with R-L-arabinofuranosyl, D-glupyranosyluronic acid (or its 4-O-methyl ether), or 1 D-xylopyranosyl residues. Some naturally occurring xylans also carry O-acetyl and feruloyl substituents. These esterified groups are removed when alkali is used during the xylan isolation process.2 Biosynthesis of arabinoxylan can be divided into four steps: chain or backbone initiation, propagation, side chain addition, and termination.3 Our understanding of these different steps in biosynthesis is still very incomplete. The biosynthesis of the xylans occurs in the golgi apparatus. The synthesized polymer is packed into vesicles and transported to the plasma membrane where they become integrated with the newly formed cellulose microfibrils.3 The degree of arabinofuranosyl substitution varies significantly between arabinoxylans from different species, but also between different parts of the plant and between plant material grown in different seasons.4 There are also differences in substitution patterns; the xylopyranosyl units can be mono or disubstituted and the arabinofuranosyl units can be further substituted with other sugars.5,6 This variation in arabinose substitution has a great impact on the solubility and material properties of the polymer. Increased degree of substitution on * To whom correspondence should be addressed. Phone: +46 (0)317723407. Fax: +46 (0)31-7723418. E-mail: [email protected]. † Chalmers University of Technology. ‡ University of Helsinki.

arabinoxylan increases the solubility of the polymer. It has been suggested that the plant uses the degree of substitution as a regulatory mechanism for solubility.7 According to this, xylan could be transported in the cell sap as a soluble arabinose enriched polymer and be deposited as this in the grain or as the insoluble arabinose depleted xylan in other parts of the plant. The side groups also prevent association between molecules as would happen with a purely linear molecule such as cellulose.8 Highly substituted arabinoxylans are usually amorphous. With a lower degree of substitution, unsubstituted sequences of the xylan backbone may approach each other and form stable interchain associations and consequently form crystalline regions.8 Highly substituted arabinoxylans have also been shown to be able to crystallize if dried slowly enough.9 The semicrystalline material is likely to exhibit different properties from the amorphous. Understanding of the substitution pattern and its relation to material properties is essential when designing new materials based on biopolymers such as arabinoxylan. Various processes have been proposed to isolate arabinoxylans from different biomass sources. The arabinoxylans in lignin containing resources such as wood, straw and husks generally require alkali to render them soluble. The extraction can be facilitated by use of a number of pretreatments such as chlorite delignification, organosolv pulping, steam explosion,10 enzyme treatment,11 ultrasound treatment,12 or twin screw extrusion.13 In addition to the alkali-soluble arabinoxylans, cereal endosperms contain water- soluble arabinoxylan, which can be extracted with water. Pretreatment and extraction methods affect the structure and Ara/Xyl ratio, and the isolated product will, therefore, have different properties. Rye arabinoxylan was first investigated by Preece and Hobkirk14 and thereafter by Aspinall and Sturgeon15 and

10.1021/bm800290m CCC: $40.75  2008 American Chemical Society Published on Web 06/24/2008

Material Properties of Films from Arabinoxylans

Aspinall and Ross.16 They concluded that water extractable arabinoxylan from rye has about every second xylopyranosyl residue substituted with a R-L-arabinofuranosyl unit attached at position C3 on the xylopyranosyl residues and distributed randomly along the chain. Later, Bengtsson and Åman17 suggested that the branched xylopyranosyl residues were mainly distributed as isolated units or small blocks of two residues rather than randomly. They also suggested that water extractable arabinoxylan consists of at least two types of structures either as different polymers or as regions in the same polymer. Arabinoxylan I has only monosubstituted xylose residues, while arabinoxylan II has mainly disubstituted residues. A similar distribution pattern was also reported by Vinkx et al.18 The arabinose substituents can be partially removed using mild acid in combination with heat. In our previous work, we studied the effect of acid-catalyzed debranching of rye arabinoxylan.19 One disadvantage was the concomitant reduction of chain length with the acid treatment. The more sophisticated method for the controlled tailoring of arabinoxylan stucture is to use specific enzymes.20,21 R-L-Arabinofuranosidases (EC 3.2.1.55) are enzymes which cleave terminal arabinofuranosyl residues from different arabinose containing polysaccharides or oligosaccharides. The R-L-arabinofuranosidases acting on polymeric xylans (arabinoxylan arabinofuranohydrolases, AXH) are further divided according to their substrate specificities as some act on (1f2)- and (1f3)-linked R-L-arabinofuranosyl units on monosubstituted xylopyranosyl residues (AXH-m), whereas others release solely (1f3)-linked R-L-arabinofuranosyl units from disubstituted xylopyranosyl residues (AXH-d3).22 In this study, the AXH-m type R-L-arabinofuranosidase was used to selectively remove arabinose residues from rye arabinoxylan. Arabinoxylan from barley husk has been shown to form films without any addition of plasticizers,23 in contrast to glucuronoxylan from wood, which requires an external plasticizer.24 The current study was undertaken to gain more understanding of the relationship between the chemical structure of arabinoxylans, specifically the Ara/Xyl ratio, and material properties of films. The aim was to evaluate the possibility to produce arabinoxylan with tailor-made properties for a desired application by controlling the Ara/Xyl ratio of the biopolymer using enzymatic treatment.

Materials and Methods Enzyme Treatments. The arabinoxylan substrate was a high viscosity rye flour arabinoxylan (lot 20601) from Megazyme International (Wicklow, Ireland). The enzyme used was R-L-arabinofuranosidase kindly obtained from Novozymes (Bagsvaerd, Denmark). The R-arabinosidase activity of the enzyme preparation was determined according to Poutanen.25 The arabinoxylan was dissolved in water (2% w/v) according to instructions from the supplier. An equal volume of arabinoxylan solution and 50 mM sodium acetate buffer containing an appropriate amount of R-arabinosidase were mixed and incubated for 5 h at 40 °C with stirring. The enzyme dosages used (200, 1000, and 10000 nkat/g of arabinoxylan) were chosen according to preliminary screening experiments in small scale. A reference sample (0 nkat/g) was treated in the same way but without an enzyme addition. After incubation, the enzyme action was terminated by keeping the solutions in boiling water bath for 15 min. A small sample (100 µL) was taken aside for further analysis of the released arabinose, and the arabinoxylan solutions were dialyzed for 48 h (MWCO 12-14000 Da) against water to remove buffer ions and released arabinose. Chemical Analysis. The amount of released arabinose was measured with a lactose/galactose assay kit according to manufacturer’s instruc-

Biomacromolecules, Vol. 9, No. 7, 2008

2043

tions (K-LACGAR, Megazyme) from the enzyme-treated arabinoxylan sample taken prior to dialysis. The samples were centrifuged for 10 min at 10000 rpm before analysis. A solution of L-arabinose (0.1 g/L) was used as a control. The monosaccharide composition of the untreated and enzymetreated arabinoxylans was analyzed by GC after acid methanolysis26 using a 6890N Network GC system (Agilent Technologies, Santa Clara, California) with a HP-5 capillary column (30 m × 0.32 mm × 0.25 µm, Agilent Technologies). The temperature program was as follows: oven temperature 70 °C, 2 °C/min to 175 °C, 12 °C/min to 290 °C. D-Arabinose, D-xylose, and D-glucose were used as standards. The samples and the standards were analyzed in triplicate. The areas of all the peaks resulting from different anomers of each monosaccharide were summed for quantification. The molecular weight distributions of the samples were examined using a high performance SEC system (Waters Corporation, Milford, MA). The system had three serially connected columns (Shodex OHpak SB-803 HQ, SB-804 HQ, and SB-806 M HQ) controlled at 50 °C, and a refractive index (RI) detector (Optilab DSP, Wyatt Technology Corp., Santa Barbara, CA). The eluent used was 0.1 M sodium nitrate with 0.02% sodium azide with a flow rate of 0.4 mL/min. Samples were prepared at a concentration of 2 mg per mL and filtered through a 0.45 µm GHP membrane filter before injection of 100 µL in the SEC-system. The 1H spectra were obtained on a Varian Unity 500 spectrometer (Varian NMR Systems, Palo Alto, CA) operating at 500 MHz for 1H. The measurements were performed at 50 °C and 1H chemical shifts (ppm) were referenced to an internal acetone signal at 2.225 ppm. Water-soluble samples were exchanged three times with D2O and finally dissolved in 1 mL of pure D2O for the NMR analysis Film Preparation and Analysis. The arabinxylan solutions were after dialysis degassed by ultrasonication under vacuum for 5 min, after which they were cast into polystyrene Petri dishes (PS, Ø 14 cm). Two films were cast from each arabinoxylan sample. The dry matter and thickness of the films were adjusted by taking into account the amount of removed arabinose i.e. more solution was used for more treated samples. About 0.5 g of arabinoxylans was used for each film (film thickness 20-25 µm, Ø 14 cm). The films were dried in a climate room at 23 °C and 50% RH and then conditioned for at least 7 days before analysis. Tensile testing was performed at 23 °C and 50% RH using an Instron 4465 universal testing machine with a load cell of 100 N, an initial grip distance of 40 mm and the strain rate 5 mm/min. The samples were cut in the rectangular specimens with a width of 10 mm and length of approximately 100 mm; 10 replicate specimens were tested from each film type. The thickness of the specimens was measured with a micrometer (Lorentzen & Wettre, precision 1 µm) at five points and an average was calculated. The water absorption of films was determined at 50.0% RH in a climate room and at 75.5% and at 98.0% RH by equilibrating pieces of films in desiccators with saturated NaCl solution and water, respectively, according to ASTM standard E 104. The humidity inside the desiccators was monitored with a moisture analyzer. The water content of the films was analyzed at equilibrium on a thermogravimetric analyzer (Mettler Toledo TGA 850 thermo gravimetric analyzer equipped with STARe software). Three replicate samples of each film were heated from 25 to 120 °C at a heating rate of 50 °C/min. The temperature was then maintained at 120 °C for 10 min, during which a steady sample weight was reached. Weight loss during the heat treatment was taken as water content. The degree of crystallinity of the films was determined by wideangle X-ray scattering. Pieces of the films were investigated with a Siemens diffractometer D5000. The 2θ angle was varied between 15 and 45 degrees. The oxygen transmission rate was measured according to ASTM standard 3985-95. Measurements were performed on a Systech Instruments 8001 permeation analyzer. The measurements were done in duplicate at 50% RH and 23 °C; the sample area was 5 cm2. The

2044

Biomacromolecules, Vol. 9, No. 7, 2008

Höije et al.

Figure 2. Shematic presentation of stuctures of (a) untreated (AX0.50) and (b) enzyme-treated (AX-0.20) rye arabinoxylans.

Figure 1. Anomeric region of the 1H NMR spectra of (a) untreated (AX-0.50) and (b) enzymatically-treated (AX-0.37) rye arabinoxylans measured at 500 MHz in D2O at 50 °C and referenced to internal acetone (2.225 ppm). R-Araf-(1f3)m ) (1f3)-linked R-L-arabinofuranosyl unit on monosubstituted xylopyranosyl residues, R-Araf(1f3)di and R-Araf-(1f2)di ) (1f3)- and (1f2)-linked R-L-arabinofuranosyl units on disubstituted xylopyranosyl residues, respectively.

permeability was calculated from the thickness of the film, which was calculated as an average of five measurements.

Results and Discussion Enzymatic Modification of Arabinoxylans. Preliminary experiments showed that a treatment with an enzyme dosage of 10 000 nkat/g for 5 h yielded the maximum liberation using the R-L-arabinofuranosidase on rye arabinoxylan. Almost 70% of the theoretical arabinose could be removed based on the analysis of the liberated arabinose. Thus, 10000 nkat/g was chosen as the highest enzyme concentration. The other concentrations chosen according to the preliminary tests were 1000 and 200 nkat/g, with the aim to release 40-45% and 20-25% of arabinose, respectively. The aim was to prepare four samples with decreasing arabinose content. The arabinose-to-xylose ratio of the arabinoxylans after enzyme treatments was analyzed by GC. In addition, the amount of released arabinose was measured and used to calculate the Ara/Xyl ratio. Reported values are the average of these two methods. The untreated sample had Ara/Xyl ratio of 0.50. The enzyme treatments with 200, 1000, and 10000 nkat/g resulted in modified arabinoxylans with Ara/Xyl ratios of 0.37, 0.30, and 0.20, respectively. The arabinoxylan samples are, thus, later referred to as AX-0.50, AX-0.37, AX-0.30, and AX-0.20. The R-arabinofuranosidase used was not able to remove all arabinose substituents from the rye arabinoxylan. By comparing the structures of arabinoxylans before (AX-0.50) and after enzyme treatment (AX-0.37) by the 1H NMR spectroscopy, it can be concluded that the enzyme was able to release only arabinofuranosyl residues from the monosubstituted xylopyranosyl units (Figure 1). The signals of anomeric protons at 5.39 ppm indicating the presence of R-L-arabinofuranosyl (1f3)linkage to monosubstituted β-D-xylopyranosyl residue decreased significantly due to the enzyme action, whereas the signals at 5.28 and 5.22 ppm originated from the (1f3)-linked and (1f2)linked R-L-arabinofuranosyl substituents on the doubly substituted β-D-xylopyranosyl unit in the main chain remaining.27

Figure 3. Molecular weight distributions of water-soluble fractions of untreated (AX-0.50) and enzymatically-treated (AX-0.37) arabinoxylan. The other samples were not soluble enough to be analyzed with the eluent used.

Thus, the enzyme used was AXH-m type and, according to its action, the sample of rye arabinoxylan used carried about 2/3 and 1/3 of the arabinofuranosyl substituent on mono- and disubstituted xylopyranosyl residues, respectively. After the enzyme incubation, the AX-0.30 and AX-0.20 samples showed visible agglomeration. This is in contrast with what we have seen in the previous study19 where mild acid treatment was used to remove the arabinose substituents and agglomeration was seen first below Ara/Xyl 0.1 on the same substrate. The reason for this behavior can probably be found in the specificity of the enzyme. The enzyme acted only on arabinose residues on monosubstituted xylose residues. It may also possess a higher activity on less-substituted segments due to the steric effects. This could lead to a more rapid debranching of certain sections of the polymer chain. Unsubstituted sections of the xylan backbones may form associations that would lead to agglomeration. Compared to mild acid treatment, where a more random debranching mechanism is likely, this will not happen as frequently. Also, the inevitable reduction in molecular weight due to main chain scissions in mild acid treatment leads to an increased solubility. The suggested structure of the untreated and enzyme-modified arabinoxylans are shown schematically in Figure 2. The molecular weight distributions of the AX-0.50 and AX0.37 samples are shown in Figure 3. The AX-0.37 sample was less soluble and, consequently, less material is detected. It also shows a higher molecular weight fraction peak, which is likely to be aggregates formed between debranched chain segments. Although it is difficult to compare samples of different degrees of substitution, because degree of substitution is likely to affect polymer conformation in solution, no notable degradation of the main chain is seen based on the chromatograms. It would be expected that shorter arabinoxylan fragments, if formed, would be the most water-soluble part of the sample. So it could

Material Properties of Films from Arabinoxylans

Biomacromolecules, Vol. 9, No. 7, 2008

2045

Table 2. Stress, Modulus, and Strain at Break for Films Prepared from Untreated (AX-0.50) and Enzymatically-Treated (AX-0.37, AX-0.30, AX-0.20) Rye Arabinoxylansa sample

stress at break (MPa)

strain at break (%)

modulus (MPa)

AX-0.50 AX-0.37 AX-0.30 AX-0.20

52.4 ( 5.9 57.7 ( 5.9 38.6 ( 4.9 36.8 ( 3.1

4.7 ( 1.8 10.4 ( 4.2 7.1 ( 0.9 7.1 ( 0.9

1750 ( 740 1160 ( 300 630 ( 125 680 ( 200

a

The averages are based on 10 measurements.

Figure 4. X-ray diffractogram of films prepared from untreated (AX0.50) and enzymatically-treated (AX-0.37, AX-0.30, AX-0.20) rye arabinoxylans. Table 1. Water Content of Films Prepared From Untreated (AX-0.50) and Enzymatically-Treated (AX-0.37, AX-0.30, AX-0.20) Rye Arabinoxylans and Equilibrated at Different Relative Humiditiesa sample

water content (%) 50% RH

76% RH

98% RH

AX-0.50 AX-0.37 AX-0.30 AX-0.20

10.97 ( 0.23 12.47 ( 0.19 13.42 ( 0.19 13.29 ( 0.30

14.70 ( 0.30 16.48 ( 0.45 16.68 ( 0.80 15.0 ( 0.65

54.58 ( 2.39 48.95 ( 1.56 48.67 ( 0.96 40.75 ( 6.38

a

Figure 5. Typical stress-strain curves of films prepared from untreated (AX-0.50) and enzymatically-treated (AX-0.37, AX-0.30, AX0.20) rye arabinoxylans.

Three parallel analyses were done.

be concluded that the enzyme preparation did not contain any contaminating xylan backbone-degrading enzymes, and the R-arabinofuranoside treatment was selectively affecting only arabinose substitution. The AX-0.30 and AX-0.20 samples agglomerated to such extent that they could not be analyzed with the water-based eluent used in the SEC. Material Properties of Enzyme-Treated Arabinoxylans. All arabinoxylan samples formed cohesive self-supporting films upon drying in Petri dishes. The thickness of the film was between 20 and 25 µm. The AX-0.30 and AX-0.20 films were translucent and had some cracks compared to the AX-0.50 and AX-0.37 films, which were completely transparent. Figure 4 shows a WAXS diffractogram of the four different films. The amount of amorphous and crystalline material in a sample can be estimated by integrating the crystalline peaks and the amorphous background signal. The AX-0.50 film is an amorphous material with no peaks, indicating crystallinity. The AX-0.37, AX-0.30, and AX-0.20 films have clear crystalline peaks between 18 and 27 degrees. They also show a higher intensity of the signal for angles above 27 degrees, indicating a higher degree of crystallinity of those samples compared to the untreated sample. Figure 3 also shows that a distinction can be made between the AX-0.37 and the AX-0.30 and AX-0.20 films in terms of crystallinity, where the two latter seem to have a higher degree of crystallinity. It can thus be concluded that crystallinity increases with decreasing arabinose substitution, which also explains the increasing translucent appearance of the films. The degree of crystallinity of the different films was not determined in this study. The water absorption of films was studied in three different relative humidities (Table 1). At 50% RH, a trend of increasing water content with decreasing arabinose content can be seen.

At 76% RH, the films from AX-0.37, AX-0.30, and AX-0.20 had still higher water content than AX-0.50. The increase of water sorption between 50% RH and 76% RH was, however, lowest for AX-0.20. The first part of water sorption isotherm, which is due to water condensation, reflects the sorption sites availability. The differences in water content are not very large among the samples and might be due to the interactions between water and hydroxyl groups of unsubstituted units, which are not organized in crystalline regions. The second part of water sorption isotherm is an exponential branch and an increase in water sorption. There is a clear difference in the behavior of the films as a function of debranching. The sample AX-0.50 shows much higher water uptake than the samples that have been debranched. During sorption, water is acting as plasticizer and allows more molecular mobility, which contributes to further chain separation and more water enter the intermolecular space. At 98% RH, the crystallites that are present in the debranched samples prevent the swelling effect by water as chemical crosslinks do. This effect is most pronounced for the most crystalline sample (AX-0.20). The average values of tensile stress, modulus, and strain at break of the films in Table 2 and the typical stress-strain diagrams in Figure 5 show how the enzymatic treatment of arabinoxylan affected the mechanical properties of the cast films. The AX-0.50 sample, prepared from the untreated rye arabinoxylan, shows an average stress at break of 52 MPa and a strain at break of 4.7%. As stated above, the film from the untreated arabinoxylan is an amorphous material. As the films are deformed, the xylan chains are disentangled and orientated in the direction of the applied force. In the case of the untreated arabinoxylan, the arabinose substituents prevent molecular interaction and crystallization between the xylan chains. The AX-0.37 film shows both higher stress at break and strain at break compared to the AX-0.50 film. One possible explanation

2046

Biomacromolecules, Vol. 9, No. 7, 2008

Höije et al.

(EVOH).32 Xylan films are good oxygen barriers because hydrogen bonds contribute to dense packing of the xylan chains.

Conclusions

Figure 6. Oxygen permeability of films prepared from untreated (AX-0.50) and enzymatically-treated (AX-0.37, AX-0.30, AX-0.20) rye arabinoxylans. Value shown is an average of two parallel measurements.

is strain-induced crystallization. Regions of unsubstituted xylan chains may be able to interact at the yield point, orientation, and crystallization will contribute to a material of higher stress at break, which is seen as the s-shaped curve for the AX-0.37 films (Figure 5). The AX-0.30 and AX-0.20 films also show this behavior, however, the result is a lower strength of film compared to the untreated AX-0.50 film. This may be explained by macroscopic phase separation due to agglomeration. As mentioned above, the AX-0.30 and AX-0.20 solution, from which films were cast, showed visible agglomeration of arabinoxylan. As the film forms, the result is a two-phase system with one water-soluble and one insoluble fraction of arabinoxylan, which results in fewer overlaps between chains. The AX0.30 and AX-0.20 films possessed higher strain at break than the films from the untreated AX-0.50. They are also less stiff. This might be due to higher water content, which results in plasticization. Similar behavior was previously observed with glycerol plasticized films from rye and wheat arabinoxylans after the extensive R-arabinofuranosidase treatment.21 It can thus be concluded that a moderate removal of arabinose substituents from the arabinoxylan chain will yield a material of higher toughness. Removal of too many substituents will cause formation of insoluble fractions, which may act as stress concentrators and lead to a material of lower stress at break. Controlled enzymatic modification of guar gum galactomannan to reduce galactose substitution was recently also found to result in films with higher tensile strength and strain at break compared to the highly substituted starting material.28 The oxygen permeability of the films was investigated to evaluate the effect of chemical structure on oxygen barrier properties. All films had low oxygen permeability with values between 1.1 and 2.0 units. A clear trend can be seen in with decreasing permeability for films of declining degree of arabinose substitution (Figure 6). As the number of crystalline domains increases, so does the density of the material and the length of the diffusion path around these domains, hence, the decrease in permeability. Other parameters such as temperature, humidity, and chain orientation may also affect the permeability.30 Barley arabinoxylan films have previously been reported to have a oxygen permeability value of 0.16.29 The oxygen permeability was considerably lower than glycerol plasticized amylose and amylopectin films31 but slightly higher than the commercially used barrier plastic ethylene vinyl alcohol

Enzymatic treatment of rye arabinoxylan has been designed and performed to prepare arabinoxylans with a gradient of arabinose content. All samples have been evaluated to find the effect of molecular structure on material properties. We have observed agglomeration in water solution, crystallization in films, reduced oxygen permeability, reduced water sorption at high RH, and yielding behavior of films with debranched arabinoxylan. All of these findings point toward enzymatic debranching, resulting in sequences of unsubstituted xylose units, which act as physical cross-linking. Enzymatic treatment seems to be a very attractive option for tailor-making xylans with specific properties. Future work will focus on agglomeration behavior in water solution with light scattering studies. Acknowledgment. Nordic Forest Research Co-operation Committee, Lyckeby Sta¨rkelsen Research Foundation, and the Knowledge Foundation are gratefully acknowledged for funding. Novozymes A/S is acknowledged for supplying the R-Larabinofuranosidase. The authors also want to thank Dr. Vratislav Langer for help with WAXS and Dr. Liisa Virkki and Ndegwa Maina for the NMR analysis.

References and Notes (1) Aspinall, G. O.; Ferrier, R. J. J. Chem. Soc. 1957, 4188–94. (2) Gabrielii, I.; Gatenholm, P.; Glasser, W. G.; Jain, R. K.; Kenne, L. Carbohydr. Polym. 2000, 43, 367–374. (3) Porchia, A. C.; Sorensen, S. O.; Scheller, H. V. Plant Phys. 2002, 130, 432–41. (4) Izydorczyk, M. S.; Biliaderis, C. G. Carbohydr. Polym. 1995, 28, 33– 48. (5) Ebringerova´, A.; Heinze, T. Macromol. Rapid Commun. 2000, 21, 542–556. (6) Ho¨ije, A.; Sandstro¨m, C.; Roubroeks, J. P.; Andersson, R.; Gohil, S.; Gatenholm, P. Carbohydr. Res. 2006, 341, 2959–2966. (7) Perlin, A. S. Cereal Chem. 1951, 28, 382–393. (8) Andrewartha, K. A.; Phillips, D. R.; Stone, B. A. Carbohydr. Res. 1979, 77, 191–204. (9) Nieduszynski, I. A.; Marchessault, R. H. Biopolymers 1972, 11, 1335– 1344. (10) Glasser, W. G.; Kaar, W. E.; Jain, R. K.; Sealey, J. E. Cellulose 2000, 7, 299–317. (11) Buchanan, C. M.; Buchanan, N. L.; Debenham, J. S.; Shelton, M. C.; Wood, M. D.; Visneski, M. J. Corn fiber for the production of advanced chemicals and materials: Separation of monosaccharides and methods thereof U.S. Patent 6352845, March 5, 2002. (12) Sun, R. C.; Sun, X. F.; Ma, X. H. Ultrason. Sonochem. 2002, 9, 95– 101. (13) Marechal, P.; Jorda, J.; Pontalier, P. Y.; Rigal, L. ACS Symp. Ser. 2004, 864, 38–51. (14) Preece, I. A.; Hobkirk, R. J. Inst. Brew. 1953, 59, 385–92. (15) Aspinall, G. O.; Sturgeon, R. J. J. Chem. Soc. 1957, 4469–71. (16) Aspinall, G. O.; Ross, K. M. J. Chem. Soc. 1963, 1681–1686. (17) Bengtsson, S.; Åman, P. Carbohydr. Polym. 1990, 12, 267–277. (18) Vinkx, C. J. A.; Stevens, I.; Gruppen, H.; Grobet, P. J.; Delcour, J. A. Cereal Chem. 1995, 72, 411–18. (19) Sternemalm, E.; Hoije, A.; Gatenholm, P. Carbohydr. Res. 2008, 343, 753–757. (20) Tenkanen, M. In Hemicelluloses: Science and Technology; Gatenholm, P., Tenkanen, M., Eds.; ACS Symposium Series 864; American Chemical Society: Washington, DC, 2004; pp 292-311. (21) Tenkanen, M.; Soovre, A.; Heikkinen, S.; Jouhtima¨ki, S.; Talja, R.; Helen, H. Production and properties of films from cereal arabinoxylans Proceedings of Science and Technology of Biomass: AdVances and Challenges, Roma, Italy, May 8-10, 2007; pp 7073. (22) Van Laere, K.; Voragen, C.; Kroef, T.; Van den Broek, L.; Beldman, G.; Voragen, A. Appl. Microbiol. Biotechnol. 1999, 51, 606–613. (23) Ho¨ije, A.; Gro¨ndahl, M.; Tømmeraas, K.; Gatenholm, P. Carbohydr. Polym. 2005, 61, 266–275.

Material Properties of Films from Arabinoxylans (24) Gro¨ndahl, M.; Eriksson, L.; Gatenholm, P. Biomacromolecules 2004, 5, 1528–1535. (25) Poutanen, K. J. Biotechnol. 1988, 7, 271–281. (26) Sundberg, A.; Sundberg, K.; Lillandt, C.; Holmbom, B. Nord. Pulp Pap. Res. J. 1996, 11, 216–219, 226. (27) Hoffmann, R. A.; Leeflang, B. R.; de Barse, M. M. J.; Kamerling, J. P.; Vliegenthart, J. F. G. Carbohydr. Res. 1991, 221, 63–81. (28) Mikkonen, K. S.; Rita, H.; Hele´n, H.; Talja, R.; Hyvo¨nen, L.; Tenkanen, M. Biomacromolecules 2007, 8, 3198–3205.

Biomacromolecules, Vol. 9, No. 7, 2008

2047

(29) Gro¨ndahl, M.; Gatenholm, P. ACS Symp. Ser. 2007, 954, 137–152. (30) Lu, Y.; Zhang, L.; Xiao, P. Polym. Degrad. Stab. 2004, 86, 51–57. (31) Rindlav-Westling, A.; Stading, M.; Hermansson, A.-M.; Gatenholm, P. Carbohydr. Polym. 1998, 36, 217–224. (32) McHugh, T. H.; Krochta, J. M. Edible Coat. Films ImproVe Food Qual. 1994, 139–87.

BM800290M