Effects of multi-functional additives during foam extrusion of wheat gluten materials

The study addresses the need to replace fossil-based polymer foams with bio-based alternatives by exploiting wheat gluten (WG) as a low-cost, renewable protein capable of forming cohesive, processable materials. Prior work has shown WG can be extruded into foams using blowing agents such as ammonium bicarbonate (ABC), with ABC outperforming sodium bicarbonate due to earlier decomposition and gas generation at lower temperatures. Foam structure in WG can be tuned by composition and processing conditions, and crosslinkers can enhance strength, rigidity, cohesion, and water resistance. Here, the authors investigate how three non-toxic multifunctional additives—genipin (GNP), gallic acid (GA), and citric acid (CA)—influence the properties of glycerol-plasticised, ABC-foamed WG foams. Each additive can act in multiple roles (crosslinker/grafting agent, radical scavenger, potential plasticiser), but the dominant effect in the WG/ABC/glycerol matrix was unknown. The objective is to determine the primary functional roles of these additives in foam-extruded WG and to relate structure to properties relevant to absorbency, cushioning, and sealing applications.

WG is a promising bio-based alternative for polymer foams due to its foaming ability (well known from bread making) and capacity to aggregate/crosslink via disulfide bonds, enabling standard thermoplastic processing (extrusion, injection molding, thermoforming). Previous studies demonstrated extrusion foaming of WG and that ABC produces more uniform, porous WG foams than sodium bicarbonate because ABC decomposes at lower temperatures to ammonia and carbon dioxide. Foam structures vary with composition and processing; crosslinking agents improve mechanical strength and water resistance in WG materials. Literature indicates: genipin is a non-toxic natural crosslinker/grafting agent (alternative to glutaraldehyde/formaldehyde), capable of polymerization and crosslinking with proteins via primary amines, and can self-polymerize at high pH. Gallic acid, a dietary polyphenol, can crosslink proteins via quinone chemistry but is also a radical scavenger that can inhibit peroxyl-radical reactions and potentially reduce radical-induced degradation or crosslinking during extrusion. Citric acid is a cheap, bio-based polycarboxylic acid used to graft/crosslink polymers (starch, chitosan, cellulose) and proteins (soy, fish, collagen), and at higher concentrations can plasticize matrices. These prior findings motivate testing CA, GA, and GNP in WG foams to tune porosity, crosslinking, and performance.

Materials: Vital wheat gluten (WG) powder (82 wt% protein) from Lantmännen Reppe AB; glycerol; ammonium bicarbonate (ABC, ≥98%); gallic acid (GA, ≥97.5%); citric acid (CA, ≥99.5%); genipin (GNP, ≥98%); a commercial nitrile butadiene rubber (NBR) foam (density 120 kg/m³) used for benchmarking.

Formulation and extrusion: WG was manually mixed with glycerol (mass ratio 70/30, WG/G). ABC at 5 wt% (relative to WG+glycerol) was used as blowing agent for foamed samples. GA, CA, or GNP were added at 1 or 5 wt% per 100 g WG/G. ABC and CA were ground (ZM 200, ring sieve 0.25, 6000 rpm) before mixing. Mixtures were extruded in a Brabender Do-Corder C3 single-screw extruder (L/D 20, screw compression ratio 2.5), barrel temperatures 50–60–70 °C (from hopper to die), screw speed 120 rpm, circular die 6.5 mm. Extrudates were dried overnight at 40 °C and stored ≥1 week in desiccator (RH ≤10%). Sample nomenclature: WG/G, WG/G/ABC, WG/G/ABC/1GA, /5GA, /1CA, /5CA, /1GNP, /5GNP.

Gas generation study and TGA: ABC decomposition/foaming power assessed by placing pure ABC, WG/G, and WG/G/ABC in latex-sealed bags submerged in silicon oil; heated in oil bath at 70 °C; gas volume via oil displacement recorded up to 20 min to compute decomposition rate (DR) and total gas volume. Additional sealed-bag heating of pure ABC at 70 and 90 °C assessed decomposition kinetics (30 min at 90 °C approached ~90% decomposition). Thermogravimetric analysis (Mettler Toledo TGA/SDTA851) measured mass loss at 70, 80, and 90 °C for 30 min in N2 (50 mL/min).

Density and porosity: Density and open/closed porosity determined via modified Archimedes method using limonene (ρ=842 kg/m³) or n-heptane (ρ=684 kg/m³). Solid matrix density ρWG/G=1290 kg/m³ computed from WG and glycerol densities. Measurements of mass in air, immersed mass, and wet mass after 1 s immersion allowed calculation of open and closed pore volumes and total porosity; expansion ratio (ER) also measured.

Morphology (SEM): Foams were cryo-fractured (liquid N2), mounted, Pd/Pt sputter-coated, and imaged by FE-SEM (Hitachi S-4800, 3 kV, 10 µA). Additional SEM (Hitachi TM-1000, 10 kV) used for samples after compression. Pore sizes measured from ≥50 measurements using ImageJ.

Swelling capacity (SC): ~200 mg specimens in tea bags immersed in 0.9 wt% NaCl (saline) from 1 s to 24 h. After withdrawal and blotting, weights recorded; SC (%)=(Wt−Wa)/Wa×100.

FTIR and 1H NMR of mixtures: ATR-FTIR (PerkinElmer Spectrum 100, 4000–600 cm⁻1, 16 scans, 4 cm⁻1 resolution; normalized to CH2 band at 2926 cm⁻1) used to probe WG foams and binary/ternary mixtures of glycerol, ABC, and each additive (simulating extrusion: 70 °C, 5 min). 1H NMR (Bruker Avance III HD 400 MHz) performed in D2O or DMSO-d6 (for GA and GNP systems) on pure and mixed additives to detect interactions.

Protein solubility and size (SE-HPLC): Three-step extraction in 0.5 wt% SDS phosphate buffer (pH 6) with sonication to sequentially break noncovalent and disulfide bonds (Ext.1–Ext.3). UV detection at 210 nm; polymeric proteins (1–15 min) and monomeric proteins (15–26 min) quantified; total extractability normalized to WG powder.

Crosslinking degree: Lowry’s method quantified soluble protein after denaturation; crosslinking normalized to WG/G (0% reference), noting genipin’s color interferes with detection.

Mechanical tests: Cyclic compression per ISO 844:2007 on 5 mm-long cylinders (Ø 6.5 mm extrudates), 10–50% strain in steps, 10 mm/min; 5 min recovery between cycles. Elastic modulus from <5% strain (ASTM D1621-16). Yield/strength at 10% strain or stress at 10% if no yield. Hysteresis loss rate computed as unloading/loading energy ratio. NBR foam tested similarly for comparison.

Compression set: ASTM D395-18. WG foams cut to 10 mm length, compressed to 40% strain for 1, 5, 24, 48 h and 1 week; recovery tracked from 1 s up to 1 month at 23 °C, 50% RH. NBR specimen compressed to ~76% strain due to its lower density. Compression set CS=(t0−t1)/t0×100.

Statistics: Student’s t-test (p<0.05) using Statgraphics 18; results reported as mean ± SD with significance letters.

• Visual/structural differences: GA and GNP foams were dark brown, CA foams and references were light brown/beige, indicating reactions with WG; color gradients across cross-sections were observed.
• Expansion ratio (ER): References WG/G and WG/G/ABC had ER ≈1.5 and 1.42–1.45. With additives: GA (1 and 5 wt%) ER≈1.21–1.24; CA (1 and 5 wt%) ER≈1.46–1.60; GNP 1 wt% ER≈1.00, 5 wt% ER≈0.86. At 10 wt% GA or CA, ER decreased to 0.7, indicating plasticization at higher contents.
• Density and porosity (Table 1): • WG/G: 883 kg/m³; total porosity 31.6% (open 1.9%, closed 29.7%); avg pore ~65 µm. • WG/G/ABC: 720 kg/m³; total porosity 44.1% (open 7.7%, closed 36.4%); avg pore ~215 µm. • GA systems: 1 wt% GA 840 kg/m³ (35.0% total), 5 wt% GA 820 kg/m³ (36.7% total); open porosity 16–20% and closed 17–19%; pores ~145–190 µm. • CA systems: Lowest densities 641–650 kg/m³; highest total porosity ~50% with mainly closed cells (closed 39–42%, open 7–11%); pores ~183–195 µm. • GNP systems: 1 wt% GNP highest density 950 kg/m³ (lowest total porosity 26.1%); 5 wt% GNP 804 kg/m³ (37.5% total). 5 wt% GNP had the largest open-cell content (open 26.5%, closed 11.0%); pores ~154–190 µm.
• Gas generation from ABC: Theoretical gas (for 50 g batch) 2530 mL vs measured 30 mL after 10 min at 70 °C for WG/G/ABC; DR for pure ABC ~20 mL/min vs WG/G/ABC ~4 mL/min (first 20 min). Increasing temperature to 90 °C greatly increased decomposition, requiring ~30 min for ~90% decomposition.
• Chemical interactions (FTIR/NMR): • CA with glycerol showed merged C=O peaks at 1720 cm⁻1, consistent with ester formation (CA–glycerol) with residual carboxyl; disappearance of CA 778 cm⁻1 band indicated interactions. • Glycerol and ABC exhibited new/changed bands at ~1626 and 1643 cm⁻1, indicating interactions. • GA mixtures showed color changes and NMR evidence of interactions with glycerol, though FTIR changes were subtle. • GNP mixtures turned brown; NMR showed a broad condensation/esterification-associated signal (P1), consistent with genipin reactions.
• Protein solubility/aggregation (SE-HPLC): WG/G had lowest solubility after noncovalent-bond extraction (Ext.1), indicating higher aggregation; 5 wt% GA had highest Ext.1 solubility and highest total extractability, indicating least crosslinking/aggregation. 5 wt% GNP also showed high total solubility after sonication/SDS. Lowest total extractability in WG/G/ABC and WG/G/ABC/5CA indicated higher aggregation/crosslinking. Monomers exceeded polymers in all but 5 wt% CA; 5 wt% GA showed largest monomer–polymer difference (≈65% vs 35%).
• Crosslinking degree (Lowry): 5 wt% GA had the lowest apparent crosslinking (−121% vs WG/G reference), consistent with radical scavenging reducing crosslinking; GA at 1 wt% also reduced crosslinking (−8%). CA samples were near the reference (± few %). GNP values could not be measured due to interference from blue coloration in the assay.
• Swelling/uptake (saline): Rapid capillary uptake was highest in GNP foams (≈50% in 1 s), attributed to higher open-cell content (notably 5 wt% GNP). The highest 24 h uptake (~130%) occurred for low GA content; overall 24 h uptake spanned ~100–130% across formulations.
• Mechanical properties (Table 2): • Yield strength at 10% strain across foams: 17–63 kPa (WG/G/ABC/5GA highest 63 kPa; GA-5%; WG/G/ABC/1GNP 60 kPa; references 52 kPa for WG/G and 18 kPa for WG/G/ABC). • Elastic modulus at 10% strain: 0.35–4.4 MPa (lowest WG/G/ABC/1GA 0.35 MPa; highest WG/G/ABC/1GNP 4.4 MPa; WG/G and WG/G/ABC/5GA ~4.1 MPa). • At 50% strain, modulus 0.008–0.23 MPa depending on formulation. • Hysteresis loss rate increased with strain; at 50% strain it ranged ≈84–92.5% across foams (highest for WG/G/ABC/5GA ≈92.5%). • Stiffness generally decreased with repeated cycling, indicating minor non-reversible damage; energy loss increased with strain/number of cycles.
• NBR comparison (Table 3, Fig. 4c): WG/G/ABC/5CA had much higher stress/stiffness than low-density NBR (120 kg/m³). NBR showed lower energy loss rate and increasing modulus with cycles (less structural damage), while WG foam modulus decreased with cycling.
• Compression set and recovery (Fig. 5): WG foams exhibited high recovery after 1 h compression (WG/G recovered 96% at 48 h; 99% at 1 month; WG/G/ABC/5CA recovered 94% at 48 h; 97% at 1 month). Longer compression times increased plastic deformation. After 1 week at 40% strain: WG/G CS ≈31% vs WG/G/ABC/5CA ≈40%. NBR showed much higher compression set after long periods (up to ~84% after 48 h), indicating greater permanent deformation.
• Dominant additive effects: CA primarily promoted crosslinking/grafting, yielding low-density, largely closed-cell foams (~50% porosity). GA primarily acted as a radical scavenger, reducing aggregation/crosslinking and increasing solubility; low GA maximized 24 h saline uptake. GNP acted predominantly as a crosslinker at 1 wt% (highest density/stiffness), while 5 wt% GNP increased open-cell content enabling rapid capillary uptake.

The central question was how multifunctional additives CA, GA, and GNP modulate the structure–property relationships of WG foams extruded with ABC and plasticised with glycerol. The data show distinct dominant roles for each additive under the mild extrusion conditions (≤70 °C). Citric acid promoted crosslinking/grafting with WG (and possibly reactions with ABC), producing the lowest-density foams dominated by closed cells and relatively low open porosity, which supports good sealing behavior and stable mechanical response. Gallic acid, despite its potential as a crosslinker, functioned primarily as a radical scavenger under these conditions, suppressing disulfide-mediated aggregation/crosslinking. This yielded higher protein extractability and, at low content, the highest equilibrium saline uptake (~130% in 24 h), suiting absorbent applications. Genipin strongly crosslinked WG at low loading (1 wt%), producing the densest, stiffest foam (highest modulus and yield strength) with the lowest overall porosity; at higher loading (5 wt%), increased open-cell content led to very rapid capillary uptake (≈50% in 1 s), advantageous for fast absorption. The incomplete decomposition of ABC at 70 °C constrained expansion, but variations in additive chemistry balanced crosslinking, plasticization, and gas–matrix interactions to tailor cell openness, pore size (65–215 µm), and mechanical damping (84–92% energy loss at 50% strain). Compared with a commercial NBR foam, the WG foams exhibited higher stresses and damping at a much higher density, and lower compression set over extended times, indicating competitiveness for cushioning and sealing. Collectively, the results validate protein biofoams as a sustainable alternative to conventional polymer/rubber foams where liquid uptake, energy dissipation, and low compression set are key.

Naturally derived multifunctional additives enable substantial tuning of foam-extruded wheat gluten properties under low-temperature processing. Citric acid primarily acts as a crosslinker/grafting agent, yielding the lowest-density (~640–650 kg/m³), largely closed-cell foams with total porosity ~50%. Gallic acid acts predominantly as a radical scavenger, producing the least aggregated/crosslinked networks (highest extractability) and the highest 24 h saline uptake at low loading (~130%). Genipin at 1 wt% yields the densest and stiffest foam (950 kg/m³; highest modulus/yield), whereas 5 wt% increases open-cell content enabling very rapid capillary uptake (~50% in 1 s). All foams exhibit high damping in cyclic compression (84–92% hysteresis loss at 50% strain) and low compression set compared with a commercial NBR foam, supporting potential in cushioning, sealing, and absorbent applications. Future work could optimize ABC decomposition and extrusion temperature profiles to increase expansion while limiting protein over-aggregation, and map additive concentrations/processing windows to target specific combinations of cell openness, absorption kinetics, and mechanical damping.

• ABC decomposition at 70 °C during extrusion was incomplete (measured gas far below theoretical), limiting expansion and influencing cell structure.
• FTIR detection sensitivity limited observation of some reactions (small extent below detection); esterification and genipin ring-opening reactions may have been incomplete at higher additive loadings (5 wt%).
• Lowry crosslinking assay could not be applied to genipin-containing samples due to color interference; crosslinking assessment relied on indirect measures for GNP.
• SE-HPLC quantification at 210 nm may underestimate crosslinking for gluten-based systems due to possible absorbance shifts to shorter wavelengths.
• Mechanical comparison with NBR was confounded by large density differences (WG foams were much denser), and NBR was tested at a different maximum compressive strain in compression set due to fixture limitations.
• The study focuses on high-density foams (26–52% porosity); ultra-low-density regimes were not explored.