Enzymatic oxidation as a potential new route to produce polysaccharide aerogels

Abstract

Specific enzymatic oxidation of terminal galactosyl-containing polysaccharides (guar galactomannan (GM), and tamarind seed galactoxyloglucan (XG)) was used to prepare hydrogels. The hydrogels were lyophilized to form novel types of polysaccharide aerogels: biobased and biodegradable, lightweight, and stiff materials. The compressive moduli of the aerogels were greatly dependent on the oxidation, polysaccharide type, freezing method, and ambient moisture. Ice crystal templated, oriented aerogels from oxidized XG (XG-OX) showed the highest compressive modulus, 359 kPa, when determined parallel to the freezing and drying direction (i.e., the vertical direction). The water vapor sorption of freeze-dried GM and XG was not significantly affected by oxidation, even though the oxidized GM (GM-OX) and XG-OX aerogels were no longer water-soluble. GM-OX and XG-OX aerogels absorbed liquid water 40 and 20 times their initial weight, respectively. Focused ion beam scanning electron microscopy showed that the inner structure of oriented aerogels from GM-OX consisted of a honeycomb architecture with a pore diameter of some tens of micrometers. On the other hand, corresponding aerogels from XG-OX seemed to contain longer capillaries oriented in the freezing direction. This observation was supported by imaging the XG-OX aerogels using high-resolution synchrotron X-ray microtomography. The enzymatic hydro- and aerogel preparation method is considered a green way to obtain novel, functional products from polysaccharides.

---

Introduction

Modified plant polysaccharides represent a highly interesting area of development toward sustainable solutions for modern materials science. Polysaccharides are a versatile group of polymers that can be functionalized by a vast number of chemical modification techniques. Furthermore, enzymatic tailoring of polysaccharides is a site-specific way of introducing new properties to polysaccharide-based materials. In comparison to chemical processing, enzymatic modification has various advantages. Protecting groups are not needed; the reactions are highly selective and can be performed in aqueous solutions. Toxic and/or hazardous chemicals are avoided.

Galactose oxidase (GaO, EC 1.1.3.9) is a single copper metalloenzyme catalyzing the oxidation of the primary C-6 hydroxyl group of galactose to a carbonyl, thus producing aldehydes (Fig. 1).†¹ GaO is secreted by the fungus Fusarium. The corresponding gene has been isolated and expressed, e.g., in Pichia pastoris and Escherichia coli to improve enzyme production.² GaO accepts oligo- and polysaccharides with terminal galactosyls as substrates, as well as some small alcohols with primary hydroxyl groups, such as glycerol. The catalytic reaction of GaO comprises oxidative and reductive half-reactions, using molecular oxygen as an electron acceptor and producing hydrogen peroxide. Recently, we have optimized the reaction conditions for the GaO-catalyzed oxidation of polysaccharides.³ GaO was also utilized as part of a chemo-enzymatic route to obtain anionic cellulose-interacting polymers⁴ as well as further functionalized polysaccharides.^5,6

Fig. 1 The partial structures of (A) guar gum galactomannan, with an average Galp : Manp ratio of 1 : 1.5,⁷ and (B) tamarind galactoxyloglucan, with an average Galp : Xylp : Glcp ratio of 1 : 2.3 : 2.8,⁹ and (C) the selective oxidation of galactosyls at C-6 with galactose oxidase, followed by (D) the formation of hemiacetal crosslinks.

Guar galactomannan (guar gum, GM) is a polysaccharide used as a thickener and stabilizer in various food products due to its water-binding and thickening ability; it is also used in pharmaceutical and cosmetic applications and in the textile, paper, mining, and oil industries.⁷ GM is obtained from guar bean seeds (Cyamopsis tetragonoloba). It consists of a linear β(1→4)-D-mannopyranosyl backbone with α-D-galactosyl side units, the galactosyls attached to the backbone via (1→6)-bonds (Fig. 1). According to Wielinga,⁷ native GM does not form gels. However, at concentrations above 0.5 wt%, the G′ and G′′ curves of aqueous GM solutions showed a crossover point as a function of frequency, with a small difference in the moduli, indicating a weak gel-like structure.⁸

Tamarind (Tamarindus indica) seed galactoxyloglucan (XG) is a commonly used food additive in Asia, particularly in Japan.⁹ It has a more branched structure than GM, having the terminal galactosyl units at the end of xylosyl branches: β-D-Galp(1→2)-α-D-Xylp. The xylosyl units are attached to the β(1→4)-D-glucopyranosyl backbone via (1→6)-bonds (Fig. 1). Native XG can form hydrogels in very high concentrations.¹⁰

Due to their terminal galactosyl units, both GM and XG are suitable substrates for GaO. The reactive carbonyl groups obtained in the oxidation react with hydroxyl groups, thus forming inter- and/or intramolecular hemiacetal bonds between the polysaccharide chains (Fig. 1). Due to this enzymatic crosslinking, highly elastic hydrogels are formed, their properties depending on the degree of oxidation and concentration.⁸ For example, oxidized GM was found to form a gel already at a 0.2% w/v concentration, and oxidized XG changed the behavior from viscoelastic liquid to gel with increasing concentrations. At higher concentrations, both oxidized GM and XG did not vary significantly as a function of frequency.⁸

The gel-forming ability of polysaccharides can be utilized in the preparation of aerogels, which are highly porous and lightweight, but mechanically strong, thermal insulator materials. Generally, aerogels are defined as gels in which the liquid phase has been replaced by air, with very moderate shrinkage of the solid network.^11,12 Polysaccharide-based aerogels were recently envisioned as novel innovative solutions for drug release¹³ and active packaging, sustainable thermal insulators, and food applications.¹⁴ Until now, the most widely studied polysaccharide aerogels have been prepared from alginate or nanofibrillated cellulose (NFC). To form an aerogel, it is necessary to first form a gel, which, in the case of polysaccharides, most often is a hydrogel. The hydrogel formation can be achieved through physical gelling mechanisms, such as with NFC, agar, starch, or β-glucan; by chemical crosslinking with di- or trivalent cations, which is applicable with anionic polysaccharides; or by using organic crosslinking molecules.¹⁴ Recently, Köhnke et al.¹⁵ crosslinked glucuronoarabinoxylan from oat spelts by oxidation to aldehydes, using sodium periodate, to induce hydrogel formation. To prepare aerogels, the hydrogel is dried, i.e., the liquid phase is removed, maintaining its three-dimensional, porous structure. The high mechanical stiffness and large surface area of the resulting aerogels can be utilized, e.g., in the sorption of desired components on the aerogel surface.¹⁴

Sehaqui et al.¹⁶ added 10–30 wt% of XG on the NFC matrix to produce biomimetic aerogels. To the best of our knowledge, no studies on aerogels from plain GM and XG have been published. Hydrogel formation by enzymatic oxidation provides a new, previously unexplored route for the preparation of aerogels from polysaccharides. Due to the safety and acceptability of GM and XG for food use and their structural suitability as a substrate in the environmentally favorable enzymatic processing method using GaO, GM and XG were used as the aerogel-forming materials in the present study. Our aim was to develop novel polysaccharide aerogels by applying enzymatic oxidation technology and to characterize the physical and morphological properties of the obtained aerogels.

Results

Hydrogel and aerogel formation

GM and XG were first enzymatically oxidized at a 1% w/v concentration to the galactoaldehyde derivatives forming the hemiacetal crosslinks. The conditions for the enzymatic oxidation were optimized in an earlier study.³ The reaction was carried out in an aqueous solution without a buffer. This approach has been found feasible with regard to further reactions or utilization, as there is no need to adjust the pH and salts from the buffers are not present. The degree of oxidation (DO) of GM and XG was somewhat varied due to their different galactose content. The DO of GM samples' total carbohydrates was high, 20–25%, due to the high galactose content of GM (50–63% of Gal oxidized). The DO of XG samples' total carbohydrates was 10–12% (63–75% of Gal oxidized). Thus the GM-OX samples contained 1.3–1.6 mmol g⁻¹ carbonyl groups and XG-OX had 0.63–0.75 mmol g⁻¹. The degree of polymerization (DP) of the polysaccharides was not affected by the oxidation.^3,4

The GM-OX and XG-OX hydrogels were frozen and lyophilized to form novel types of aerogels (Fig. 2). After lyophilization, no shrinkage of the outer dimensions of the cubic specimens was observed.

Fig. 2 Photograph of ice crystal templated (oriented) and lyophilized aerogels from oxidized (A) guar galactomannan (GM-OX) and (B) tamarind galactoxyloglucan (XG-OX). The average side length of the cubic specimens was 17 mm in all directions.

Density and mechanical properties

The aerogels were lightweight, with densities of 0.016 ± 0.0002 g cm⁻³ for the GM-OX and 0.012 ± 0.0003 g cm⁻³ for the XG-OX.

The compressive modulus of the cubic specimens was determined to evaluate the aerogel preparation method's effect on the material's mechanical stiffness. In compression testing, the specimens collapsed, but did not fracture. Therefore, the compressive strength could not be determined.

Native XG and GM, which formed a viscous aqueous solution, but not a highly elastic hydrogel, were lyophilized and their compressive modulus was determined (Fig. 3A). Those specimens were very soft and showed low compressive moduli between 2 and 22 kPa. The compressive moduli of aerogels from enzymatically oxidized XG-OX and GM-OX were notably higher and varied between 32 and 360 kPa, depending on the polysaccharide type, the orientation of the aerogel, and relative humidity (RH) (Fig. 3A). In general, the XG-OX aerogels were stiffer than the corresponding GM-OX aerogels. The compressive moduli of conventionally frozen, unoriented GM-OX aerogels was 33 kPa in a vertical direction (casting direction of the cubic samples), and 58 kPa in a horizontal direction. Correspondingly, the compressive moduli of unoriented XG-OX aerogels were 153 and 123 kPa. In comparison, the oriented (by unidirectional freezing, carried out by immersing the gels into a bath containing CO₂ ice and ethanol) specimens showed greatly higher compressive moduli (227 and 359 kPa for GM-OX and XG-OX, respectively) when the compression test was performed in a vertical direction (Fig. 3B). In the horizontal direction, the compressive moduli were 53 (GM-OX) and 45 kPa (XG-OX) (Fig. 3). Interestingly, at increased RH (98%), the vertical compressive moduli of aerogels were clearly lower than at 50% RH, but the horizontal moduli were not affected by the increased RH (Fig. 3B).

Fig. 3 (A) Compressive modulus of conventionally frozen, lyophilized guar galactomannan (GM) and tamarind galactoxyloglucan (XG) as well as aerogels prepared from oxidized (OX) GM-OX and XG-OX, in horizontal (H) and vertical (V) directions, analyzed at 50% RH. (B) Compressive modulus of ice crystal templated GM-OX and XG-OX aerogels at 50% and 98% RH. The analyses were based on 10 replicates and the error bars indicate standard deviations.

Water and water vapor uptake

The water vapor absorption and desorption of aerogels were studied to determine the ability of the materials to bind and release moisture at different RHs. The shapes of the sorption curves were sigmoidal, as commonly reported for hygroscopic materials (Fig. 4). The hysteresis phenomenon was observed for all studied samples. Enzymatic oxidation and consequent gel formation slightly decreased the moisture uptake and moisture release of both polysaccharide matrices (GM and XG) at medium RH levels. The sorption behavior of GM compared to XG as well as that of GM-OX and XG-OX was rather similar, the only significant difference being the higher moisture uptake of GM and GM-OX at 90% RH (34%) than that of XG and XG-OX (29%) (Fig. 4).

Fig. 4 Moisture uptake of (A) lyophilized guar galactomannan (GM) and oriented aerogels prepared from oxidized GM-OX and (B) lyophilized tamarind galactoxyloglucan (XG) as well as aerogels from oxidized XG-OX. The data are average from two parallel analyses.

To evaluate the capacity of the aerogels to absorb liquid water, the specimens were immersed in water and weighed after certain periods of time (1, 3 and 24 h). The aerogels maintained their shape and did not dissolve, but floated on top of the water, which was magnetically stirred at RT. The weighing results, shown in Table 1, demonstrate that the GM-OX aerogels absorbed water as much as 40 times their initial weight. The XG-OX aerogels absorbed less, 10 times their initial weight after one hour, but the absorption increased to 20 times their initial weight after 24 h of stirring in water (Table 1).

Table 1 Water absorption of oriented aerogels from oxidised guar galactomannan (GM-OX) and galactoxyloglucan (XG-OX) (average ± standard deviation from three parallel specimens)

Aerogel	Time (h)	Weight (g)	Water uptake (g)	Water uptake (× aerogel initial weight)
GM-OX	0	0.06 ± 0.006	0	0
	1	2.5 ± 0.4	2.4 ± 0.4	40
	3	2.7 ± 0.2	2.6 ± 0.2	44
	24	2.6 ± 0.5	2.5 ± 0.5	42
XG-OX	0	0.06 ± 0.005	0	0
	1	0.7 ± 0.04	0.6 ± 0.04	10
	3	0.9 ± 0.06	0.9 ± 0.05	13
	24	1.3 ± 0.2	1.3 ± 0.2	20

---

Morphology

The aerogels prepared by ice crystal templating differed from those conventionally frozen, in terms of their visual appearance. The unidirectionally frozen aerogels had a pattern of aligned vertical structures visible with the naked eye. A photograph of the samples is presented in Fig. 2. In comparison, the conventionally frozen aerogels did not show uniform orientation.

To obtain information on the inner structure of the aerogels, they were viewed with focused ion beam scanning electron microscopy (FIB-SEM). Fig. 5A shows a magnified view of the side surface of GM-OX aerogel, in which the vertical structures are clearly visible. At some, apparently random parts, the surface of the GM-OX was covered by a thin film with a thickness of approximately a few micrometers. At other locations, the film was discontinuous and revealed the pore structure underneath (Fig. 5B). The latter consisted of a honeycomb structure, i.e., vertical capillaries divided into cells or compartments with horizontal walls. These cells, i.e., pores in the aerogel structure, were some tens of micrometers in diameter.

Fig. 5 Focused ion beam scanning electron microscopic images of oriented aerogels from oxidised guar galactomannan (GM-OX) (A and B) and galactoxyloglucan (XG-OX) (C and D). In D, the surface of XG-OX was milled open at a 52° angle and the view was tilted in the opposite direction. The scale bars are 50 μm (A and D), 200 μm (B), and 1 mm (C).

The film covering the side surface of XG-OX aerogel was more continuous than that on the GM-OX aerogel. The porous inner structure was only visible by FIB-SEM close to the bottom of the intact cubic XG-OX specimen (Fig. 5C). To view the XG-OX aerogel in more detail, the surface was milled, revealing the structure beneath the film (Fig. 5D). The milling was done at an angle of 52°; in addition, the viewing angle was tilted in the opposite direction. Thus the structures shown in Fig. 5D consist of the mouths of capillaries running towards the top of the aerogel specimen. The XG-OX aerogel did not show as many horizontal walls as the GM-OX aerogel, but instead the capillaries in the XG-OX sample seemed to be taller.

In order to visualize the structure away from the surfaces of the aerogel blocks, samples were cut from near the center of the cubic specimens and imaged with X-ray microtomography (XMT). In 3D renderings of the XMT reconstructions (Fig. 6), the aerogels appear to have a foam-like structure with pore sizes in the order of a few tens of micrometers. For a better visualization, one XG-OX sample was also scanned using high-resolution phase contrast tomography at a beamline ID19 of the European Synchrotron Radiation Facility (ESRF). Three orthogonal slices through the reconstruction are shown in Fig. 7. The inner structure appears very similar to the one close to the surface, seen in Fig. 5D, with the pore size in the order of 10 μm horizontally, and slightly higher in the vertical direction.

Fig. 6 X-ray microtomography reconstructions of oriented aerogels from oxidized guar galactomannan (GM-OX) (A) and tamarind galactoxyloglucan (XG-OX) (B), voxel size 1 × 1 × 1 μm³. The scale bar is 0.5 mm.

Fig. 7 Synchrotron phase-contrast X-ray microtomography reconstruction of an aerogel from oriented, oxidized galactoxyloglucan (XG-OX), voxel size 0.355 × 0.355 × 0.355 μm³. The scale bar is 150 μm.

Discussion

Aerogel properties

The native GM and XG formed viscous solutions in water, but not highly elastic gels.⁸ Correspondingly, the lyophilized, unmodified GM and XG samples could not be considered aerogels, but soft and fluffy, easily water-dispersible and soluble matter. The enzymatic oxidation and gel formation completely changed the characteristics of the materials. The hydrogel formation was visually observed but not confirmed, for example, by rheological measurements, as we previously studied identical preparation method and hydrogel products having similar degrees of oxidation.⁸

The GM-OX and XG-OX aerogels showed similar or slightly lower density values than aerogels prepared from 1–15% (w/v) solutions of sodium alginate, cellulose, β-glucan, or starch, as reviewed by Mikkonen et al.¹⁴ GM-OX and XG-OX form elastic gels through enzymatic oxidation even at low concentrations.⁸ In the present work, 1% solutions of the polysaccharides were used for gel formation, leading to low-density dried aerogels.

The dried GM-OX and XG-OX aerogels were much more rigid than the native lyophilized GM and XG, which was also shown in the compression testing (Fig. 3A). The mechanical stiffness of the GM-OX and XG-OX aerogels was high, particularly considering their low density. In comparison, Ca-alginate aerogels with densities between 0.039 and 0.067 g cm⁻³, studied by Cheng et al.,¹⁷ were brittle and suddenly collapsed upon compression. Those aerogels yielded a compressive strength between 20 and 70 kPa, depending on the aerogel density and the use of reinforcements. The present GM-OX and XG-OX aerogels were tough rather than brittle, and instead of fracturing, they were squeezed into denser structures during compression. Most likely the pore walls bended or collapsed, and the pores condensed. Previously, Sehaqui et al.¹⁸ prepared NFC aerogels with varying densities. The lyophilized NFC aerogels with a similar density to that of the present GM-OX and XG-OX aerogels, 0.014 g cm⁻³, showed a compressive modulus of 35 kPa, which was at the same level of magnitude as that of the unoriented GM-OX aerogels in the present study. On the other hand, NFC aerogels with higher density (0.105 g cm⁻³) showed greatly higher compressive moduli, up to 2800 kPa.¹⁸

The aerogel formation method, i.e., the conventional vs. unidirectional freezing technique of hydrogels greatly affected the mechanical properties of the aerogels. Unidirectional freezing increased the compressive moduli of aerogels determined in the vertical direction (i.e., freezing direction), but decreased the modulus in the horizontal direction, in comparison to aerogels frozen conventionally in a freezer. Previously, Köhnke et al.¹⁹ compared the compressive moduli of lyophilized aerogels from glucuronoarabinoxylan and cellulose nanocrystals either plunged in liquid nitrogen or unidirectionally frozen using liquid nitrogen. They found that the compressive modulus of aerogels was higher (285 kPa) for the unidirectionally frozen samples, measured in the freezing direction, than for the samples frozen by being plunged into liquid nitrogen (158 kPa). This corresponds well with our present data, in which the oriented XG-OX aerogels were slightly stiffer and the GM-OX aerogels nearly as stiff as the oriented xylan-based aerogels studied by Köhnke et al.,¹⁹ and the compressive modulus of the conventionally frozen XG-OX aerogels was at the same magnitude, whereas that of the corresponding GM-OX aerogels was lower than the modulus of xylan-based aerogels plunged into liquid nitrogen.¹⁹ Thus the unidirectional freezing of polysaccharide gels and the consequent orientation of the polymers can result in direction-dependent mechanical performance of the materials, in analogy to phenomena well known among synthetic polymer chemistry, e.g. oriented polypropylene²⁰ as well as paper technology.²¹ In addition to the freezing method, the drying of aerogels via lyophilization may affect the self-assembly or aggregation of molecular structures, as demonstrated with basic copper(II) salts by Wei et al.²² Such self-assembled microstructure is assumed to contribute to the mechanical properties of aerogels. In future, we aim at exploring the detailed morphology of polysaccharide aerogels to understand their mechanical properties.

Köhnke et al.¹⁵ oxidized glucuronoarabinoxylan (GAX) by sodium periodate to contain carbonyl groups 1.7–6.0 mmol g⁻¹, i.e., more than the enzymatically oxidized GM-OX (1.3–1.6 mmol g⁻¹) and XG-OX aerogels (0.63–0.75 mmol g⁻¹) in the present study. The chemically oxidized GAX-based aerogels showed clearly lower compressive moduli than the untreated control sample.¹⁵ The moduli values, determined in the freezing direction of the aerogels, were also lower than those of the oriented GM-OX and XG-OX aerogels (vertical direction) in the present study. Köhnke et al.¹⁵ explained their results by the opening of the xylan pyranose ring due to the sodium periodate oxidation. When the C6 of galactosyls is selectively oxidized by GaO, this undesirable outcome is avoided and the enzymatic oxidation of GM and XG results in enhanced material properties. In addition, during the sodium periodate oxidation, a significant decrease of DP may occur,²³ which also is avoided by the utilization of GaO.

The aerogels showed an interesting, selective softening behavior due to increased moisture content. The water vapor absorption analysis showed that the moisture uptake of aerogels at 50% RH was about 10%, while at 90% RH the moisture uptake of the samples was clearly higher (Fig. 4). The compressive moduli of the aerogels were compared at 50% and 98% RH. Interestingly, the aerogels softened (i.e., their compressive moduli decreased) only in the vertical direction, but not in the horizontal direction (Fig. 3). Further studies are needed to better understand the distribution of polysaccharide chains and water molecules in the oriented aerogels as well as the effect of the molecular arrangement on the mechanical properties of aerogels.

The enzymatic oxidation and subsequent gel formation changed the behavior of GM and XG with regard to moisture. Native GM and XG were water-soluble at room temperature (RT), but the aerogels from oxidized GM-OX and XG-OX maintained their integrity even after 24 hours under magnetic stirring in water. The starting solutions for sample preparation contained 1% polysaccharide in water; thus, the oxidized hydrogels bound 100 g of water per 1 g of polysaccharides. Lyophilization of the hydrogels into aerogels decreased their water-absorption capacity: the GM-OX and XG-OX aerogels absorbed water 40 and 20 times their dry weight, respectively (Table 1). However, that amount of bound water is still considerable and comparable to results obtained for alginate aerogels, 20 g of water per 1 g of aerogel.²⁴ Citric acid–crosslinked xylan–chitosan aerogels absorbed even more, 80 g of water per 1 g of aerogel.²⁵ The velocity of water absorption was much faster for GM-OX aerogels (1 h) than for XG-OX aerogels, whose water sorption seemed to continue for at least 24 h.

The enzymatic oxidation did not, however, greatly affect the water vapor absorption on the lyophilized specimens (Fig. 4). This can be explained by the fact that the GaO-catalyzed reaction did not significantly affect the hydrophilic characteristics of the materials. Oxidizing the primary hydroxyl group in some of the terminal galactosyl residues of GM and XG is not supposed to affect their hydrophilicity. The resulting hemiacetal bonds between the created carbonyl groups and existing hydroxyl groups are assumed to be the cause for the formation of a crosslinked polysaccharide network, i.e., hydrogel. Due to the crosslinking, some of the functional groups (carbonyl, hydroxyl) were occupied and not available for binding moisture. This could explain the minor differences observed in the water vapor sorption of native versus oxidized GM and XG (Fig. 4). On the other hand, the decreased water solubility and increased integrity of the aerogels in an aqueous environment was most likely due to the crosslinking and network formation of GM-OX and XG-OX by the hemiacetal bonds.

Morphology of aerogels

Ice-crystal template freezing was used to affect the heat transfer and liquid phase (water) crystallization within the cubic specimens. Freezing kinetics favor crystal growth directions parallel to the temperature gradient.¹⁹ The visually observed patterns in the XG-OX and GM-OX aerogels most likely resulted from the growth of ice crystals and the simultaneous orientation of polysaccharide chains from the bottom to the top of the hydrogel samples. We suggest that parts of the polysaccharide chains were immobilized during the freezing of the bottom of the hydrogel, while other parts of the chains were still mobile and able to reach or stretch towards the higher temperature at the upper part of the hydrogel sample. The matrix formed thus could consist of oriented blocks of polysaccharide chains organized vertically.

According to Köhnke et al.,¹⁹ the aerogel pores can also be oriented depending on the freezing conditions. We studied the pore structure of unidirectionally frozen XG-OX and GM-OX aerogels by FIB-SEM (Fig. 5) and microtomography (Fig. 6). The FIB-SEM images support the idea that the freezing water shaped the inner structure of the aerogels, creating organized honeycomb structures in the GM-OX and capillaries in the XG-OX aerogels. In the XMT images (Fig. 6), the structure appears to be more foam-like and the pore size seems larger than when viewed with FIB-SEM. In addition, the XMT images of oriented GM-OX aerogels showed some agglomerates, which could either originate from poorly dissolved GM or from crosslinked structures formed through GaO oxidation (Fig. 6). On the other hand, the XMT images of XG-OX only showed the pore walls. Due to the low X-ray attenuation in the samples, conventional absorption XMT had very poor contrast, and was only able to visualize the thickest structures. It is therefore possible that the observed large pores could contain a finer, thin-walled, pore structure, which is lost in the noise due to poor contrast. This hypothesis is supported by Fig. 7, where the higher brilliance, monochromatic beam, and lower X-ray energy of the synchrotron source combined with phase retrieval techniques result in a vastly improved signal-to-noise ratio.

Experimental

Materials

Galactose oxidase used was from Fusarium spp., which was produced recombinantly in Pichia pastoris, and donated by Dr Sybe Hartmans (specific activity of 770 U mg⁻¹) measured by chromogenic ABTS assay as described by Spadiut et al.² Guar galactomannan (GM, 14K0168, M_w 2600 kDa (ref. 3)) horseradish peroxidase (HRP, P8250, type II, 181 U mg⁻¹), and catalase (C30, from bovine liver, 22 000 U mg⁻¹) were purchased from Aldrich (St. Louis, MO, U.S.A.) Tamarind seed galactoxyloglucan (XG, M_w 1300 kDa as analysed in dimethyl sulfoxide by the method of Pitkänen et al.²⁶) was obtained as a donation from Dainippon Sumitomo Pharma (Fukushima, Japan).

Enzymatic oxidation and aerogel preparation

Polysaccharide (GM, XG) was stirred in Milli-Q water (1% w/v) at least 1 h to ensure dissolution. The enzymes GaO, HRP, and catalase were added (HRP to activate GaO and catalase to remove H₂O₂ formed in the oxidation). The enzyme dosages and reaction conditions were based on previous studies.³ The amount of GaO was related to the approximate amount of terminal galactose present in the polymer (1.53 U of GaO per mg of galactose). The dosages of HRP and catalase were 1 U mg⁻¹ and 115 U mg⁻¹, respectively. For example, 612 U of GaO, 400 U of HRP and 46 kU of catalase were used in the oxidation of 1 g of GM containing ca. 400 mg of galactose (ca. 40% of total sugars). After stirring the reaction at RT for 48 h the enzymes were inactivated by heating the solution in a boiling water bath for ca. 10 min.

Air bubbles were removed from the heated hydrogels in vacuo until no bubbles were visually detected. The hydrogels were allowed to cool down and then transferred into cubical Petri dishes utilizing a 20 ml syringe. To prepare unidirectionally frozen samples using ice crystal template, the Petri dishes were placed on top of a CO₂ ice plate in an EtOH bath. After 0.5–1 h the hydrogel was frozen, and the samples were placed in −70 °C. For comparison, samples were prepared by conventional freezing by placing the hydrogels into the Petri dishes as described, freezing them at −20 °C, and then placing them at −70 °C. Water was removed by lyophilization.

Degree of oxidation

Samples (ca. 1 ml; 10 mg of polysaccharide) were taken from the hydrogels after the enzymes had been inactivated. A previously developed GC-MS method was used to determine the degree of oxidation.³ In the method, the aldehydes formed in the oxidation were reduced back to alcohols with NaBD₄ introducing, along with the reduction, deuterium labels. The samples were then methanolyzed, silylated and analyzed by GC-MS. The degree of oxidation was then calculated from the difference between the m/z 361 and m/z 362 ions in the mass spectra of unreacted starting material galactosyls and the oxidized, deuterated product galactosyls.³

Physical characterization of aerogels

Density. The density of the oriented GM-OX and XG-OX aerogels was determined utilizing glass beads (0.2 mm diameter, 1.68 g cm⁻³). An aerogel cube (weight = w_s) was placed into a glass container of known volume and weight (w_d), and the container filled with the beads. The weight was recorded (container, sample, and beads = w_dsb) and the weight of the beads (w_b) calculated by subtracting: w_b = w_dsb − w_d − w_s. The volume of the beads (v_b) was then obtained by v_b = w_b/1.68 g ml⁻¹. The density of the aerogel sample (ρ_s) was calculated by dividing w_s by the reminder of v_b subtracted from v_d(ρ_s = w_s/(v_d − v_b)). The determination was conducted for two cubic samples and the average represented the density.

Compression testing. The compressive moduli of the aerogels was determined at 23 °C and 50% RH (climate room) using an Instron 33R4465 universal testing machine with a load cell of 100 N. The specimens were compressed for 6 mm at the rate of 1.3 mm min⁻¹. Ten replicate cubic specimens with a width, length, and thickness of approximately 17 mm were tested in vertical and horizontal directions. The dimensions of the specimens were measured with a digital caliper (MarCal 16 ER, Mahr, Göttingen, Germany, precision 10 μm) at two points, and an average was calculated. The samples were conditioned at the climate room for at least three days before analysis. In addition to 50% RH, unidirectionally frozen (oriented) GM-OX and XG-OX aerogels were conditioned at 98% RH, in desiccators containing pure deionized water.

Dynamic vapor sorption. A DVS Intrinsic sorption microbalance (Surface Measurement Systems, Alperton, Middlesex, UK) was used to collect water absorption and desorption isotherms of lyophilized GM and XG as well as oriented GM-OX and XG-OX aerogels. The experiments were carried out in duplicate at 25 °C and RH values from 0 to 90% and back to 0%. The sample was hydrated stepwise in 10% RH steps by equilibrating the sample weight at each step. The moisture uptake was calculated according to eqn (1):where W_moist is the sample weight equilibrated at the chosen relative humidity and W_dry is the weight of the dry sample. The specimens used for the sorption studies were conditioned in vacuum desiccators over anhydrous P₂O₅ (RH 0%) before analysis.

Water absorption. The oriented GM-OX and XG-OX aerogel samples were weighed and placed into decanter glasses containing 50 ml of water. The water was magnetically stirred at RT. To weigh the amount of absorbed water in the cubes after 1, 3, and 24 h, they were taken from the glass, using a spatula, and rolled quickly on a wipe to remove excess water from the surface. The weight of the dry cubes was subtracted from the weight of the wet cubes to obtain the amount of absorbed water. Three parallel experiments were conducted, and an average was calculated.

Morphology of aerogels

Focused ion beam scanning electron microscopy. Cubic, oriented aerogel specimens from oxidized GM-OX and XG-OX were coated with an Au/Pd film, using a Cressington sputter coater, to increase surface conductivity and to improve surface mechanical stability during electron beam imaging. The surfaces of the specimens were viewed and the XG-OX aerogel was milled open with 30 kV Ga+ ions, using focused ion beam scanning electron microscopy (Quanta 3D 200i FIB-SEM).

Microtomography. For the microtomography experiments, small pieces of approximately 1 × 1 × 5 mm³ were cut from near the center of the oriented GM-OX and XG-OX aerogel specimens and glued on top of carbon fiber rods for scanning.

The absorption XMT scans were performed with a Nanotom 180NF nanofocus XMT scanner (Phoenix|X-ray Systems and Services GmbH, Germany, presently part of GE Measurement and Control Solutions), using 50 kV X-ray tube voltage and 400 μA current. The scan geometry resulted in an effective pixel size of 0.5 × 0.5 μm² in the transmission images. The reconstructions were calculated with 1 × 1 × 1 μm³ voxel size using datos|x software supplied by the equipment manufacturer.

The synchrotron experiment was performed at beamline ID19, ESRF, using 18 keV X-ray energy and a voxel size of 0.355 × 0.355 × 0.355 μm³. Single-distance phase retrieval (Paganin method) was applied to the transmission images prior to reconstruction.

Conclusions

We presented a novel method to prepare aerogels from polysaccharides, using specific enzymatic oxidation as a route to form hydrogels. Enzymatic modification is considered an environmentally favorable method to convert polysaccharides into advanced materials with beneficial properties. To the best of our knowledge, this is the first report on the use of plain GM and XG as aerogel matrices. Enzymatic oxidation using GaO was shown to be a suitable technique to produce hydrogels and further aerogels from GM and XG. Lyophilization of the GM-OX and XG-OX hydrogels resulted in the formation of aerogels with low density, high water absorption capacity, and high mechanical stiffness. An additional benefit of the enzymatic oxidation was that the GM-OX and XG-OX aerogels were no longer soluble in water, in contrast to their starting materials. Ice crystal template was used to shape the aerogels and create an oriented structure, which was visualized by FIB-SEM and microtomography. In the future, the demonstrated concept could be used to prepare various types of functional materials, with reinforcements or active components mixed with the polysaccharides before gel formation, entrapping them in the matrix. The obtained aerogels are biobased and biodegradable, and prepared without hazardous chemicals in an aqueous environment. Both GM and XG are acceptable for food and packaging use, which could be potential future applications for such aerogels.

Acknowledgements

We thank Dr Sybe Hartmans for the galactose oxidase and Dainippon Sumitomo Pharma for tamarind galactoxyloglucan. We acknowledge the European Synchrotron Radiation Facility for providing synchrotron radiation facilities, and Dr Paul Tafforeau for his assistance in using beamline ID19. The Academy of Finland is acknowledged for funding KSM (project number 268144) and KP (project numbers 133027 and 252183), the National Doctoral Programme in Materials Physics for funding J-PS, and Magnus Ehrnrooth Foundation for funding AG.

Notes and references

1. J. W. Whittaker, Chem. Rev., 2003, 103, 2347.
2. O. Spadiut, L. Olsson and H. Brumer, Microb. Cell Fact., 2010, 9, 68.
3. K. Parikka, A.-S. Leppänen, L. Pitkänen, M. Reunanen, S. Willför and M. Tenkanen, J. Agric. Food Chem., 2010, 58, 262.
4. K. Parikka, A.-S. Leppänen, C. Xu, L. Pitkänen, P. Eronen, M. Österberg, H. Brumer, S. Willför and M. Tenkanen, Biomacromolecules, 2012, 13, 2418.
5. C. Xu, O. Spadiut, A. C. Araújo, A. Nakhai and H. Brumer, ChemSusChem, 2012, 5, 661.
6. A.-S. Leppänen, C. Xu, K. Parikka, P. Eklund, R. Sjöholm, H. Brumer, M. Tenkanen and S. Willför, Carbohydr. Polym., 2014, 100, 46.
7. W. C. Wielinga, Galactomannans, in Handbook of Hydrocolloids, ed. G. O. Phillips and P. A. Williams, Woodhead Publishing, Cambridge, UK, 2nd edn, 2009, pp. 228–251.
8. K. Parikka, F. Ansari, S. Hietala and M. Tenkanen, Food Hydrocolloids, 2012, 26, 212.
9. K. Nishinari, M. Takemasa, K. Yamatoya and M. Shirakawa, Xyloglucan, in Handbook of Hydrocolloids, ed. G. O. Phillips and P. A. Williams, Woodhead Publishing, Cambridge, UK, 2nd edn, 2009, pp. 535–566.
10. D. R. Nisbet, K. E. Crompton, S. D. Hamilton, S. Shirakawa, R. J. Prankerd, D. I. Finkelstein, M. K. Horne and J. S. Forsythe, Biophys. Chem., 2006, 121, 14.
11. S. S. Kistler, Coherent expanded aerogels, J. Phys. Chem., 1932, 36, 52.
12. A. Du, B. Zhou, Z. Zhang and J. Shen, Materials, 2013, 6, 941.
13. C. A. García-González, M. Alnaief and I. Smirnova, Carbohydr. Polym., 2011, 86, 1425.
14. K. S. Mikkonen, K. Parikka, A. Ghafar and M. Tenkanen, Trends Food Sci. Technol., 2013, 34, 124.
15. T. Köhnke, T. Elder, H. Theliander and A. J. Ragauskas, Carbohydr. Polym., 2014, 100, 24.
16. H. Sehaqui, M. Salajkova, Q. Zhou and L. A. Berglund, Soft Matter, 2010, 6, 1824.
17. Y. Cheng, L. Lu, W. Zhang, J. Shi and Y. Cao, Carbohydr. Polym., 2012, 88, 1093.
18. H. Sehaqui, Q. Zhou and L. A. Berglund, Compos. Sci. Technol., 2011, 71, 1593.
19. T. Köhnke, A. Lin, T. Elder, H. Theliander and A. J. Ragauskas, Green Chem., 2012, 14, 1864.
20. A. J. Peacock and A. Calhoun, Polymer Chemistry Properties and Applications, Hanser Publishers, Munich, Germany, 2006.
21. M. J. Kirwan, Paper and paperboard packaging, in Food Packaging Technology, ed. R. Coles, D. McDowell and M. J. Kirwan, Blackwell Publishing, Oxford, UK, 2003, pp. 241–281.
22. W. Wei, P. Gao, J. Xie, S. Zong, H. Cui and X. Yue, J. Solid State Chem., 2013, 204, 305.
23. C. K. Simi and T. E. Abraham, Food Hydrocolloids, 2010, 24, 72.
24. R. R. Mallepally, I. Bernard, M. A. Marin, K. R. Ward and M. A. McHugh, J. Supercrit. Fluids, 2013, 79, 202.
25. A. Salam, R. A. Venditti, J. J. Pawlak and K. El-Tahlawy, Carbohydr. Polym., 2011, 84, 1221.
26. L. Pitkänen, L. Virkki, M. Tenkanen and P. Tuomainen, Biomacromolecules, 2009, 10, 1962.

---

Footnote

† The GaO structure presented in the graphical abstract was based on data published by Ito et al., Nature, 1991, 350, 87.

---