Solubility of spruce galactoglucomannans determines interfacial morphology and emulsion stability

Specific supramolecular interactions in polysaccharides change their solubility rendering diverse interfacial properties in oil-in-water (O/W) emulsions. We studied the effect of solubility of softwood hemicelluloses, spruce 34 galactoglucomannans (GGMs) isolated by pressurized hot-water extraction methods or recovered as side soluble or insoluble fraction dominates the interfacial morphology. Under the condition of sufficient interfacial 46 coverage, insoluble fractions complement the emulsion stability with a filling effect in the continuous phase of 47 emulsions. The findings improve our understanding of bio-b ased polysaccharides’ solubility , their emulsion 48 stability mechanisms, and strategies to tailor via biorefining approaches.


51
Stabilization of dispersed systems such as emulsions by biopolymers and bio-based colloidal particles is an 52 emerging area of research in life science applications, including food and pharmaceuticals (Dickinson 2003).

53
Polysaccharides are an important class of biopolymers obtained from plants, microbes, seaweed, and algae, and 54 utilized with or without physical/chemical modification (Dickinson 2003). Their key functions as emulsifiers, 55 stabilizers, or gelators are governed by solubility, which is a result of complex inter-/intramolecular interactions 56 between inherent characteristics of polysaccharides (molar mass, degree of branching/substitution, functional 57 groups, monosaccharide composition), solvent and solvent environment (pH, ionic strength, temperature) 58 (Whistler 1973). When soluble, polysaccharides attain a molecular dispersed state with individual molecules 59 commonly in a random coil conformation, while insoluble polysaccharides display colloidal particle behavior.

60
Cellulose and chitin are examples of the latter. Molecular associations result from H-bonding, van der Waals and/or 61 hydrophobic interactions between functional groups in polysaccharides, e.g., hydroxyl, acetyl, carbonyl groups 62 (Dumitriu 2004) or covalently bound non-polysaccharide components such as bound proteins or phenolic groups, 63 e.g., proteins in gum Arabic (Sanchez et al. 2018), ferulic acid in pectins (Zhang et al. 2015), and arabinoxylans 64 (Ebringerova et al. 1994). GGMs. However, the macromolecular state of GGMs in an aqueous solution and eventually at the oil-water 83 interface is largely unknown. We have previously noted that GGMs isolated from pressurized hot water extraction 84 (PHWE) were semi-soluble and solubility depended on the purity of the extracts. There were evidences of 85 supramolecular aggregates, agglomerates, and colloidal particles (Bhattarai et al. 2020) suggesting the need to 86 investigate the multiphase emulsion system stabilized by GGMs by comprehensively studying their solubility.

87
In this study, we evaluate the solubility of GGMs recovered as industrial side-streams from the thermomechanical 88 pulping process (TMP) and hot water extraction of wood chips using the BLN process (named after inventors) 89 (Schoultz 2015). Recovered GGMs from these industrially relevant extraction approaches have previously 90 exhibited well to satisfactory emulsion stabilization abilities. These GGMs differ in molar mass and total content 91 of phenolic compounds; therefore, we hypothesize these affect their solubility. To understand the effect of recovery 92 processes on GGM solubility, we compare these results with our recent findings on GGMs obtained from the 93 PHWE process (Fig. 1). In the next step, we characterize the oil-water interface stabilized by GGMs

98
This study contributes to our understanding of the colloidal properties of polysaccharides, especially those from 99 plant extracts with varying purity, and the possibility to tailor an application-based recovery approach to isolate 100 them. As the demand for bio-based hydrocolloids is rising with an emphasis given to those obtained from 101 unused/waste resources, knowledge on potentially valuable wood hemicelluloses will facilitate their future use in 102 pharmaceuticals, food, and cosmetics sectors, for example.

115
GGMs were recovered from the TMP process in a Finnish TMP mill on an industrial-scale following the method 116 of Willför et al. (2003) Mechanical shearing during the TMP process releases water-soluble hemicelluloses from 117 wood into the process water, (Thornton 1994;Willför et al. 2003) which was collected after a series of filtration 118 and ultra-filtration. The resulting isolate was used after spray drying and after its purification using ethanol added 5 to the isolate at a 9:1 v/v followed by drying in a vacuum oven at 40 °C (Xu et al. 2007). The spray-dried GGM 120 and ethanol-precipitated GGM from this process are to as sTMP GGM and eTMP GGM, respectively.

121
Another GGMs was obtained by hot water extraction namely, BLN method patented by Schoultz (2015) and 122 PHWE method based on Kilpeläinen et al. (2014) Water at 150-170 °C was used to recover GGMs from wood 123 chips (BLN) or sawdust (PHWE) in a flow-through extractor. Using the PHWE method, the extract was collected 124 for 70 min whereas using the BLN method, the extract was recirculated multiple times through the biomass before 125 collection. By doing so, aromatic residues in the extract such as lignins, lignin-derived compounds, and extractives 126 were readsorbed back to the biomass to obtain high purity. The extract was further purified by adding ethanol at     190 where p is the power-law coefficient. A p < 3 indicates structures described as mass fractals (e.g., wrinkled paper).

191
A p-value between 3 and 4 arises from surface fractals with a dense and homogenous inner structure, whereas p = 192 4 indicates smooth and well-defined interfaces. A special case of p = 1 and p = 2 correspond to thin cylinders or 193 strings and thin planes or disks, respectively (Schmidt 1991). The measurement was performed on shear-treated samples within 1-2 days of the treatment.

199
In the second part of the study, emulsions were prepared using the GGMs. All emulsions were prepared using 1 200 wt% hemicellulose and 2.5 wt% rapeseed oil to ensure an emulsifier-rich regime (  intensities. An example of data generated in the MC procedure is shown in Fig. S2B

306
To understand the nanostructural characteristics of supramolecular fractions, SAXS experiments were performed 307 (Fig. 3). The samples were also studied after mechanical shearing treatment using a microfluidizer to observe any 308 shear-induced nanostructural changes. Selected aqueous phases were also measured after being subjected to 309 heating and freeze-thawing treatment. The scattering data from aqueous phases of sTMP GGM did not exhibit any 310 correlation peak in the studied length scale (2-200 nm) with no difference in the scattering feature before and 311 after high-shear treatment (Fig. 3A), and even after heating and freeze thawing (scattering data not shown). The  Table S4) (Dumitriu 2004). Further studies are required to elucidate the structure-solubility 362 relationship in GGMs.

363
In the next step, using the relatively insoluble GGMs from the TMP process (sTMP GGM and eTMP GGM),

376
We investigated the interfacial structures of emulsions by combining SAXS (Fig. 4) with cryo-scanning electron 377 microscopy (Fig. 5). Scattering from emulsions comprises contributions from their aqueous phases (stabilizers + 378 buffer), oil, and oil/water interface. Scattering data from GGM-based emulsions did not exhibit any correlation 379 peak ( Fig. 4A-C). The broad correlation peak in BGX and T20 emulsion in the length scale ≈ 40 nm (  (Table S3).

395
Emulsions from both sTMP GGM and eTMP GGM exhibited p > 4, BLN GGM and ePHWE GGM exhibited 396 p ≈ 4, and BGX exhibited p ≈ 3 indicating diffused, smooth, and heavy surface fractal characteristics of interfaces, 397 respectively. The first point to note is, oil-water interface from stabilizers with a highly aggregated supramolecular fraction (sTMP GGM, eTMP GGM, and BGX) differed from less aggregated ones (BLN GGM and ePHWE GGM).

399
The second point to note is the distinction of the fractal oil-water interface of BGX emulsion from the rest.

400
Scattering data from sPHWE GGM and T20 emulsions could not be fitted as their aqueous phases had correlation 401 peaks that were not present in the same length scale as in their emulsions, therefore, their subtraction was not

409
Further, similar objects of varying morphology such as rod-or platelet-like were visualized in the continuous phase 410 of this emulsion (Fig. 5, 1d). This organization of supramolecular structures of different sizes and shapes in the 411 interfacial/near-interfacial regions is the proposed explanation behind the diffused interfaces in these emulsions.

425
Some dents were observed on the surface of oil droplets (Fig. 5, panel 4c)   were evaluated during a 2-week storage time (Fig. 6A-B). The dynamic changes from droplets' 433 coalescence/flocculation in emulsions during 4-week storage were compared using the Turbiscan Stability Index 434 (TSI) (Fig. 6C-D).