Batter Rheology and Crumb Softness in Sorbitol Free Sponge Systems with Erythritol

Industrial sponge systems formulated without sorbitol demand a different stabilisation strategy for the aqueous phase and the gas-cell interface, because erythritol does not reproduce sorbitol’s moisture-sorption profile or plasticising interaction with starch at equal mass addition. Sorbitol, listed as E 420 in Regulation (EC) No 1333/2008, is used in high-ratio sponge batters at 2–6 g/100 g total batter mass to retain crumb moisture and delay firming during ambient distribution. Erythritol, listed as E 968, has a relative sweetness of approximately 60–70% of sucrose, an energy value of 0.2 kcal/g, an aqueous solubility of about 37 g/100 g at 25 °C, and an endothermic heat of solution of approximately −180 J/g; it is non-reducing, does not participate in Maillard browning, and remains crystalline at relative humidities below 90%. In a sponge formulation, these properties shift the thermodynamic state of water, the rheology of the continuous serum phase, and the post-bake recrystallisation behaviour of both starch and polyol. Consequently, batter density, yield stress, storage modulus, and crumb compression force cannot be expected to remain within process capability limits when sorbitol is removed without compensatory changes in water addition, starch selection, or emulsifier concentration. Published data for the fully formulated erythritol sponge system remain limited; the following discussion therefore combines standard batter rheology methods with established physicochemical properties of erythritol.

What Rheological Transitions Occur When Sorbitol Is Replaced by Erythritol at Equivalent Sweetness Contribution?

The replacement ratio between sorbitol and erythritol is not a fixed mass ratio because sorbitol provides approximately 60% of sucrose sweetness and erythritol provides approximately 70%; a formulation targeting equal sweetness therefore requires slightly less erythritol than sorbitol, but the functional demand for crumb softness may require a different mass. Sponge batter is a four-phase system comprising gas cells, starch granules, protein-stabilised interfacial films, and a continuous sugar-polyol serum phase. The apparent viscosity at 10 s⁻¹ and 20 °C in industrial sponge batters typically falls between 5 and 15 Pa·s, with the flow behaviour index derived from a Herschel-Bulkley fit ranging from 0.60 to 0.85; these values depend on air volume fraction, protein level, fat type, and sugar crystal size. The presence of undissolved erythritol crystals with a particle size above 50 µm acts as a suspended filler that raises low-shear viscosity and may introduce a yield stress component, whereas fully dissolved erythritol reduces serum viscosity relative to sorbitol at equal molal concentration because of its lower molecular weight and reduced hydrodynamic volume. Dynamic oscillatory measurements using a rheometer fitted with a 50 mm parallel plate geometry and a 1 mm gap under a Peltier-controlled temperature of 20 °C can capture the linear viscoelastic region; sponge batters typically show a linear region below 0.1% strain, above which the breakdown of gas-cell networks produces strain-dependent loss of storage modulus. In systems where erythritol is pre-dissolved in warm water at 40 °C and then cooled to 20 °C, the serum phase often exhibits lower apparent viscosity than the sorbitol reference because erythritol has a molecular weight of 122.12 g/mol versus 182.17 g/mol for sorbitol and forms a less extended hydration shell. The resultant batter may require longer mixing time at constant mixer speed to reach a specific gravity of 0.45 g/cm³, because the lower continuous-phase viscosity accelerates drainage of the protein-stabilised film unless emulsifier concentration is increased. When erythritol is not fully dissolved, the residual crystals act as particulate fillers; this condition changes the Herschel-Bulkley consistency coefficient upward and reduces the flow behaviour index, but published data for sponge-specific magnitudes are limited. Dynamic frequency sweeps from 0.1 Hz to 10 Hz show that sorbitol-containing sponge batters generally have tan δ values of 0.25–0.35 at 1 Hz; erythritol replacement shifts the response through a combination of reduced serum viscosity, altered water activity, and direct polyol-starch interaction. The direction of the tan δ shift depends on whether erythritol is in solution or suspension, so process control should track both batter temperature and microscopy-based crystal counts.

During continuous high-speed aeration, the phase volume of air increases from about 0.05 to 0.45–0.50, and the power draw of a vertical planetary mixer can rise as the batter transitions from a viscous liquid to a structured foam. In production-scale vertical planetary mixers with bowl capacities of 100–250 L, the mixing time required to reach target density is normally 3–8 min at tool speeds of 150–300 rpm; the addition of erythritol above 4 g/100 g as non-dissolved crystals can extend this time by up to 2 min and can produce batch-to-batch density variation of ±0.03 g/cm³ when incoming crystal size distribution is not controlled. The heat of solution of erythritol, approximately −180 J/g, produces a measurable batter temperature reduction of 2–4 °C when crystalline erythritol is added directly to an aqueous phase at 30 °C; this cooling effect can stabilise air-cell size but may also reduce the solubility limit and increase the undissolved crystal fraction. A temperature-controlled aeration head with jacket setpoint 18–22 °C and a mass flow meter on the air line are necessary to maintain constant density when sorbitol is removed. In-line rotational viscometry at 10 s⁻¹ with a bypass loop is preferred to periodic sampling because erythritol-containing batters can undergo rapid viscosity changes during the first 60 s after mixing due to crystal dissolution and temperature equilibration.

When Batter Water Activity Falls Below 0.75 Without Sorbitol Humectancy

Aerated sponge batters containing sorbitol normally reach a water activity between 0.85 and 0.90 at 25 °C; sorbitol’s hygroscopicity then retards moisture loss from the baked crumb during distribution. Erythritol, at equal solids, can depress water activity through solute concentration because its lower molecular weight generates more moles per unit mass, but it does not retain water at ordinary atmospheric humidities; measured water activity according to ISO 18787:2017 may therefore be equal or lower in the erythritol batter without predicting a softer crumb at day 7. The critical difference appears after baking: sorbitol remains dispersed in the amorphous starch-gluten matrix and plasticises the cell walls, whereas erythritol can either remain as dispersed crystals or recrystallise during storage, contributing limited plasticisation and potentially increasing perceived crumbliness. Dynamic vapor sorption analysis shows that sorbitol begins measurable water uptake above 60% relative humidity, whereas erythritol remains essentially non-hygroscopic up to 90% relative humidity; this divergence is the main reason that sorbitol-free sponge systems with erythritol require a lower baking moisture target or a separate humectant system when crumb softness must be maintained for more than 5–7 days. Published data for the specific configuration of erythritol sponge stored at 25 °C and 50% relative humidity are limited, but the physical chemistry indicates that moisture migration from crumb to crust will proceed at a faster rate than in sorbitol-containing controls if water activity gradients are not reformulated.

Because erythritol has a melting point of approximately 121.5 °C, undissolved crystals can persist during the early oven phase and may influence starch gelatinisation as a dispersed solid rather than as a dissolved solute. Differential scanning calorimetry according to ISO 11357-1:2023 at a heating rate of 10 °C/min typically shows wheat starch gelatinisation onset temperatures in sponge formulations between 55 °C and 65 °C, with peak temperatures near 70 °C; polyol addition generally shifts these values upward by reducing water activity and competing for water. Erythritol, when dissolved, may shift the gelatinisation peak upward more than sorbitol at equal mass because its higher molality produces stronger water-binding in the thermodynamic sense. The practical consequence is that the sponge crumb may set later or require higher internal temperature, increasing the risk of a coarse crumb if leavening gases escape before the starch network has fully formed. Crumb firmness measured by AACC 74-09.01 with a 25 mm cylindrical probe, 25% compression, crosshead speed 1.0 mm/s, and trigger force 0.05 N is the standard method for tracking staling; industrial sponge products commonly show fresh firmness values of 4–8 N and day-7 values of 10–15 N when stored in closed packaging at 20–25 °C. In sorbitol-free erythritol sponge systems, firmness progression may show a steeper increase between day 1 and day 7 if no starch-modifying system is present, although the published comparative data needed to establish a universal rate constant are incomplete.

Thermal Events During Baking and Ambient Storage

The thermal events linking batter rheology to crumb softness occur in three overlapping stages: gas expansion and water evaporation in the first oven phase, starch gelatinisation and protein coagulation in the centre, and post-bake starch retrogradation during storage. In sorbitol-containing sponge systems, the polyol remains in the amorphous phase and reduces the mobility of water molecules, which slows the recrystallisation of amylopectin and contributes to longer softness. Erythritol does not form a similar amorphous humectant phase; once the batter temperature exceeds the erythritol melting point near 121.5 °C, crystalline erythritol can melt, but on cooling it tends to recrystallise rapidly unless the surrounding matrix viscosity is high enough to inhibit nucleation. This recrystallisation releases water that was not bound as a hydration shell, creating local moisture gradients that may accelerate starch retrogradation in adjacent crumb regions. Differential scanning calorimetry of stored crumb samples at 4 °C can measure the amylopectin retrogradation endotherm, typically observed between 45 °C and 65 °C at a heating rate of 10 °C/min; an increase in endotherm enthalpy over 7 days correlates with crumb firming. In sponge systems containing erythritol, the amorphous starch fraction may be lower because erythritol competes for water during gelatinisation but does not plasticise the rubbery cell wall after baking. Formulators should therefore expect to compensate through higher batter moisture, addition of an alternative plasticising humectant such as glycerol, or use of pregelatinised starch or starch-modifying enzymes if extended softness is required.

Although erythritol does not denature egg albumin or interact with lipid interfaces in the same way as sodium chloride or sugars, its effect on interfacial film rheology becomes measurable when the polyol is partly crystalline. In sponge formulations stabilised with emulsifier blends such as distilled monoglycerides and propylene glycol monostearate, the drainage rate of the lamellae is affected by continuous-phase viscosity and particle concentration. A batter with erythritol crystals demonstrates different foam stability than a batter with fully dissolved erythritol, even when the specific gravity is identical; the crystals can puncture or bridge the thin films during expansion in the oven. This distinction is critical for high-ratio sponge systems where the batter contains more than 100 parts sugar and polyol per 100 parts flour and the gas-cell films are already under high strain during baking.

Oscillatory Strain Sweep and Frequency-Dependent Gelation Points in Erythritol Substitution

Strain sweeps on a controlled-stress or controlled-strain rheometer identify the linear viscoelastic limit of aerated batters before gelatinisation. A typical sponge batter at 20 °C and 1 Hz shows a linear region up to 0.1% strain; beyond this limit, the storage modulus G′ decreases as air-cell walls yield under increasing deformation. In frequency sweeps from 0.1 Hz to 10 Hz within the linear region, the slope of log G′ versus log frequency for sponge batters is usually between 0.05 and 0.15, indicating a weak gel structure, while tan δ values of 0.20–0.35 confirm the predominantly elastic character required for gas retention. Replacement of sorbitol with fully dissolved erythritol is expected to reduce G′ relative to the sorbitol reference when batter density is held constant, because lower serum viscosity accelerates film drainage; however, the magnitude of this shift has not been reliably quantified in published sponge batter studies for erythritol. Replacement with partially crystalline erythritol may increase G′ and introduce a distinct shear-history dependence, but the effect depends on particle size distribution and the fraction of undissolved polyol. Frequency-dependent gelation points during baking are more relevant to crumb structure because starch swelling and protein coagulation transform the batter from a viscous foam into a solid foam. In-line monitoring of G′ at 1 Hz during a temperature ramp from 20 °C to 95 °C at 2 °C/min provides a gelation trace; the crossover of G′ and G″ occurs earlier in sorbitol-containing batters than in erythritol batters with undissolved crystals, suggesting delayed structure setting. This shift must be accounted for in oven profile design, particularly when erythritol addition exceeds its solubility limit in the available water at 20 °C.

In production-scale continuous aerators, the switch from sorbitol to erythritol creates a different failure mode near the depositor because erythritol crystals can sediment when the batter is held at low shear. Continuous batter aerators using rotor-stator heads and compressed air or nitrogen injection typically operate at overrun settings corresponding to a final specific gravity of 0.42–0.48 g/cm³; if the batter is held in a buffer tank for more than 10–15 min without gentle agitation, undissolved erythritol particles with a density near 1.45 g/cm³ settle and generate non-uniform density in the deposited batter. This settling produces a measurable increase in deposit weight variation across a multi-piston depositor, with the first deposits from the bottom of the hopper showing higher density and lower bake volume. The problem is reduced by pre-dissolving erythritol in the formulation water at 40–45 °C, using erythritol with a particle size D90 below 100 µm as determined by laser diffraction according to ISO 13320:2020, and applying low-shear recirculation at 5–10 rpm in the buffer tank. Published data for erythritol sedimentation in sponge batters are limited, but the Stokes settling velocity can be estimated from the density difference between erythritol and the continuous phase; this calculation supports the need for particle size reduction and controlled holding time.

Can Erythritol Maintain Crumb Softness Without Sorbitol's Plasticizing Capacity?

The retention of crumb softness in a sorbitol-free sponge system is not determined solely by water activity; it also depends on the mechanical properties of the cell wall matrix at the storage temperature. Sorbitol has a glass transition temperature that is depressed by water and acts as a plasticiser, allowing the starch-gluten matrix to remain ductile at low moisture and low temperature. Erythritol has a higher melting point and a greater tendency to crystallise, so it does not provide the same plasticising function. In sponge systems where sorbitol is replaced with erythritol, the crumb can become firmer and more brittle after 48–72 h at 20 °C, particularly if the equilibrium moisture content of the crumb is below 28 g/100 g. Texture profile analysis performed with a texture analyser using 25% strain, 1 mm/s crosshead speed, and a 25 mm probe can separate the contributions of firmness, springiness, and cohesiveness; in erythritol sponge systems, the loss of cohesiveness often appears earlier than the increase in firmness, which is consistent with brittle cell walls rather than uniform starch retrogradation. Published data for this specific configuration are limited, but the mechanical response indicates that erythritol cannot be used as a direct sorbitol substitute when softness over a shelf life longer than 5 days is mandatory unless a secondary plasticiser or starch anti-staling system is introduced.

Formulation boundaries for erythritol in sponge systems are governed by the available water in the batter and the storage temperature profile. The aqueous solubility of erythritol is approximately 37 g/100 g at 25 °C; if the formulation water is insufficient to dissolve the added erythritol, the undissolved crystals can be detected as roughness on the tongue and can act as nucleation sites for further crystallisation in the crumb. A practical control is to dissolve erythritol in the free water component at 40–45 °C, then cool the solution to 20–22 °C before combining with flour, starch, emulsifier, and aerated fat. This process reduces the undissolved crystal fraction but increases the heat removal load on the mixer because of the endothermic heat of solution. When erythritol is combined with high-intensity sweeteners, the bulk sweetness can be maintained at lower polyol mass, reducing the crystallisation risk but also removing the body and water-binding contribution of the sugar alcohol. In such systems, the batter serum viscosity can fall below the range required to stabilise air cells, and the formulation may require addition of a hydrocolloid or modified starch. The use of erythritol above 5 g/100 g in total batter mass without particle size control and without dissolution in warm water is not recommended for industrial sponge lines because of the combined risk of depositor weight variation, grittiness, and post-bake hardening.

Analytical Verification and Compliance Matrix

The following matrix defines the minimum measurements for comparing sorbitol-containing and sorbitol-free sponge systems during process development and production qualification. Each measurement is linked to a standard method or a defined instrument configuration; acceptance values are provided as industrial targets rather than universal specifications because sponge formulations differ in flour protein, batter fat, emulsifier system, and finished moisture.

To reduce the risk of caking and cross-contamination at transfer points, sorbitol-free sponge batters containing erythritol should not be blended with residual sorbitol fractions in volumetric feeders; sorbitol begins measurable moisture uptake above 60% relative humidity and can cause bridging in hoppers, whereas erythritol remains non-hygroscopic up to 90% relative humidity. Pre-drying of erythritol is required only when free-flowing crystal handling is necessary under elevated relative humidity or when dilute-phase pneumatic conveying is used, because even non-hygroscopic crystals can acquire surface moisture in steam-cleaned transfer lines. The combination of erythritol with amine-based leavening salts should be evaluated in the context of ammonia release and batter pH, because erythritol itself does not participate in Maillard reactions but can alter the dissolution pattern of ammonium bicarbonate in a water-limited batter. Published data for this specific interaction in aerated sponge batters are limited, so process qualification should include batter pH, density stability, and crumb firmness after 24 h and 168 h of storage at 20–25 °C.