Ionic Liquid Lignosulfonate: Dispersant and Binder for Preparation of Biocomposite Materials

Abstract Ionic liquid lignins are prepared from sodium lignosulfonate by a cation exchange reaction and display glass transition temperatures as low as −13 °C. Diethyleneglycol‐functionalized protic cations inhibit lignin aggregation to produce a free‐flowing “ionic liquid lignin”, despite it being a high‐molecular‐weight polyelectrolyte. Through this approach, the properties of both lignin and ionic liquids are combined to create a dispersant and binder for cellulose+gluten mixtures to produce small microphases. Biocomposite testing pieces are produced by hot‐pressing this mixture, yielding a material with fewer defects and improved toughness in comparison to other lignins. The use of unmodified lignosulfonate, acetylated lignosulfonate, or free ionic liquid for similar materials production yields poorer substances because of their inability to maximize interfacial contact and complexation with cellulose and proteins.


Materials
Sodium lignosulfonate (SLS; 93%) was purchased from Roth and used as received. Tris-[2-(2methoxyethoxy)ethyl]amine (TrisEG; 94%) was purchased from Merck and used as received. Imidazole (≥99 %), 1-vinylimidazole (≥99 %), trioctylamine (98 %), pyridine (99.8 %), gluten from wheat, and cellulose fibers (medium) were purchased from Sigma Aldrich and used as received. Amberlite® IR120 was purchased from Sigma Aldrich and rinsed with deionized water prior to use. 1 H and 13 C{ 1 H}-NMR spectra were collected on a Bruker DPX-400 spectrometer. Thermogravimetric analysis (TGA) experiments were conducted using a Netzsch TG209-F1 Libra. An aluminum crucible was used for the measurement of ~10 mg of sample under flow of nitrogen (10 mL/min). The samples were heated at a rate of 2.5 K/min to 600 °C. Differential scanning calorimetry (DSC) experiments were performed on a PerkinElmer DSC-1 instrument at a heating/cooling rate of 10 K/min under nitrogen flow and cycled five times. Glass transition temperatures (T g ) was acquired by determining the temperature on the thermal curve corresponding to half the heat flow differences between the extrapolates onset and extrapolated end. This was performed using Netzsch Proteus Thermal Anlaysis software. Melting points were acquired from the final heating cycle. Inductively coupled plasma optical emission spectrometry (ICP-OES) was performed on a Perkin Elmer Optima 8000. Infrared spectroscopy was conducted on a Nicolet iS 5 FT-IR spectrometer.

Preparation of cation-exchanged lignosulfonate
Two different procedures were employed depending on the amount of desired material. To prepare cation-exchanged lignosulfonates on the order of 1-2 g, Amberlite® IR120 (5.00 g) was added to a 10 mL vial and rinsed with deionized water (3 x 5 mL) followed by the addition of fresh water (6 mL) and base (11.75 mmol; Imidazole, 0.80 g; 2-methylimidazole, 0.96 g; 1-allylimidazole, 1.27 g; 1vinylimidazole, 1.10 g, 1-methylimidazole, 0.96 g; pyridine, 0.92 g; TrisEG, 3.79 g.). The vials were shaken overnight using a wrist-action shaker before being rinsed by water, ethanol, acetone, and then water on a fritted filter to remove excess base. The resin was then added to a SLS solution (0.40 g, 5 mL H 2 O) and shaken overnight using a wrist-action shaker. The resin was then separated from solution by gravity filtration and the water evaporated in vacuo with heating leaving a dark power in quantitative yields. The isolated material when TrisEG was used was a dark viscous oil.
Preparation of TrisEG:LS on a larger scale was performed using a column. Amberlite® IR120 (157.43 g) was added to a column (3 cm diameter) fitted with a stopcock and rinsed with deionized water (3 L). TrisEG (119.35 g, 367.88 mmol) was then dissolved in 1 L deionized water and passed through the column slowly. The resin was then rinsed with water (3 L) to remove excess TrisEG. Then 85.69 g SLS was dissolved in water (500 mL) and slowly passed through the column and the cation-exchanged lignin isolated in a 1 L beaker. Volatiles were evaporated by heating in an oven at 90 °C for 12 hours which removed most of the water. More complete drying was achieved by heating the oil at 90 °C in vacuo for 24 hours, which resulted in a material with ~1 wt% H 2 O content as determined by KFtitration.

Method discussion
Salt-metathesis is routinely performed to synthesize a variety of ILs and is high yielding when the organic salt is sufficiently insoluble and precipitates from solution. In this case however the exchanged product was too soluble in water and no precipitate was formed. To achieve the cation exchange and isolation of the lignin product, we employed the use of a resin. A strong acidic resin (Amberlite® IR120; sulfonic acid functionalized) was treated with different nitrogen bases ( Figure 1) to form the protic salt, then cation exchanged with SLS. The spent resin was separated by gravity filtration and water evaporated to isolate the cation-modified lignin in quantitative yields. The resin was regenerated with a strong acid and reused multiple times. This methodology is simple, amenable to larger scale production, and requires no heating or toxic reagents in order to modify the thermal and chemical properties of lignosulfonates. Unlike salt metathesis, where the precipitation of one product helps to drive the reaction forwards, the exchange of cations between two identical anions and no precipitation leads to incomplete conversion; however the use of excess ammoniumfunctionalized resin can favour high exchange yields to produce the desired product.
The T g range from as low as -13 °C to 115 °C for the modified lignins and the absence of an obervable T g for SLS indicates a change in the intermolecular forces between the lignin macromolecules as a result of the organic cation and that the cation structure is significant. While some sodium ions remained in the product after exchange, ion-exchange reactions with macromolecules are usually less complete in comparison to small molecules. In this case the vast majority (>80%) of sodium ions were exchanged with different organic cations and removed from solution using the exchange resin.

Preparation of the biocomposite
Water (17.0 mL) was added to the desired amount of lignin and stirred until complete dissolution. Gluten (6.0 g) and cellulose (10.0 g) were then mixed together as powders, either with the help of a rod or a mixer, until the powders obtained a homogeneous colouration. The lignin solution was poured into the gluten-cellulose powder mixture and the resulting dough kneaded by hand or with a mixer until all the liquid was absorbed by the powder. The wet dough was then placed into a custom made aluminum mold (5x5x0.5 cm) and placed into a hot-press previously heated to 130°C. The pressing of the specimen was performed gradually (every 5 minutes) to allow a gentle removal of the water, until the pressure of 10 bar was recorded on the instrument. Once the final pressure was reached, the specimens were cooled to room temperature under pressure and then recovered.
Sample Preparation: A band saw was used to cut the specimen in different thickness for flexural (0,5 cm thick) and tensile tests (0,2 cm thick) .

Mechanical Testing
All mechanical tests were measured using a Zwick mechanical tester zwickiLine Z2.5 equipped with a loadcell of 1 kN. Elastic modulus was manually calculated at 0,05-0,25% of strain.
All the measurements were recorded using the software TestXpert II V3.71.
Flexural test: A three-point-bending test was performed with a 2.5 kN mechanical testing machine (zwicki, Zwick Roell), equipped with a 1 kN load cell. The strains were measured by cross-head travel. The samples were loaded with a gap of 3 cm, a maximum force of 800 N, a preload of 0,5 N and a test speed of 0,025 mm s -1 .
Tensile test: Tensile test was performed with a 2.5 kN mechanical testing machine (zwicki, Zwick Roell), equipped with a 1 kN load cell. The strains were measured by cross-head travel. The samples were loaded with a gap of 3 cm, a maximum force of 800 N, a preload of 1 N and a test speed of 0,02 mm s -1 .
Cyclic test: The cycling test was performed until 9MPa of maximum stress and recovered until 1 N (on specimens of 30 mm) before the stress was removed to restart the cycle (100 times).
Small angle X-ray scattering (SAXS) SAXS was carried using a Nanostar (Bruker AXS, Karlsruhe, Germany) device, equipped with a SIEMENS KFF CU 2K-90 X-ray tube, operating at 40 kV and 100 mA and generating an X-ray beam with a wavelength of 1.5418 Å (Cu Kα radiation) and a focal spot size of 550 µm. Silver behenate standard was used to calculate beam center and the exact sample-to-detector distance. 2D scattering plots were corrected for background and transmission.

Discussion of the Bending and Tensile Tests
Composites prepared using SLS exhibited mechanical properties typical of brittle materials, where the bending modulus increased with increasing SLS content up to 27 wt% (Table S1, 3-SLS), but also broke at low deformation (<2%). This brittleness is likely a result of the defects caused by phaseseparated lignin which causes points of rupture in the material. Specimens prepared using TrisEG:MsOH displayed the opposite effect, with a progressively decreasing elastic modulus and rupture at 24 % of deformation in bending tests. (Table S1, entry 4-TrisEG:MsOH) This can be ascribed to the lack of cohesion between grains and the presence of TrisEG:MsOH which simply fills the the composite's free-volume rather than dispersing and binding cellulose+gluten phases.
Composites containing TrisEG:LS were much more stiff to initial deformation in comparison, but also exhibited a higher deformaton at break at 38 wt% content (~13% for 4-TrisEG:LS compared to ~1% for 4-SLS). The elastic modulus is comparable to the SLS samples, but a higher maximum stress was measured when 27 wt% TrisEG:LS was employed (3-TrisEG:LS vs 3-SLS). We attribute this difference to the fewer defects in the specimen. The plasticizing effect becomes dominant at 38 wt% TrisEG:LS where the stress-strain curve appears more similar to a ductile material rather than a typical particle board with a decrease of both Emod and stress maximum, and a dL jump from 3 to 12.6% yielding the toughest material of the materials here reported (~15 MJ/m 3 ). Tensile tests displayed a similar trend to the 3 point bending tests (Table S2). An increase in elastic modulus from 664 to 1812 MPa was observed for SLS samples with greater lignin content and a low deformation at break of <1 %. Composite 1-TrisEG:LS again shows that TrisEG:LS is a good dispersant and binder for the preparation of such fiberboard, yielding materials with good resistance to initial deformation (elastic modulus of GPa range), high maximum stress (up to 10 MPa), and particularly high toughness in comparison to SLS containing samples (15 MJ/m 3 , 4-TrisEG:LS vs 0,57 MJ/m 3 , (4-SLS). The much superior maximum stress for 4-TrisEG:LS compared to 4-SLS (~10 MPa vs ~3 MPa in Figure 4A) was achieved by the inhibition of defects in the composite, leading to a higher force required for the rupture of the specimens.