Investigation of glucose hydrate formation and loss: Parameters, mechanisms, and physical stability

Glucose is used extensively in the food industry, often sold as a bulk ingredient or as an ingredient in powdered drink and dessert mixes. Commercially, glucose is crystallized as either α-anhydrous glucose (α-AG) or glucose monohydrate (GM). The β-anhydrous glucose (β-AG) form is also possible, but is more difficult to crystallize and not sold as an ingredient. Conversion between the anhydrous and monohydrate structure may occur during long-term storage, however, hydrate formation and loss parameters of glucose were not found in the current literature. As crystalline materials, α-AG and GM are considered reasonably stable during storage below their literature reported deliquescence points (RH0) of 89-91%RH and 91%RH, respectively, at 25°C. However, many other crystalline sugars, such as sucrose, have been observed to cake below their RH0. The flowability of ingredients is a top priority in the food industry, since unacceptable or caked product will result in lost production time and decreased customer satisfaction. Thus, it is important to determine the chemical and physical stability of α-AG and GM during storage. This research investigated hydrate formation in α-AG and hydrate loss in GM. Under dynamic conditions, hydrate formation occurs before deliquescence in both α-AG and GM. Therefore, this research introduces a new term, dynamic deliquescence (RH0d), to report deliquescence influenced by additional water-solid interactions, such as hydrate formation, under dynamic conditions. To enhance stability, GM is dried during production to a moisture content below its full stoichiometric monohydrate moisture content, and therefore hydrate formation is still possible in GM. Furthermore, X-ray powder diffraction detected a small amount of β-AG in the commercially available GM samples. Hydrate formation under equilibrium conditions occurs in α-AG at 68%RH and hydrate loss occurs in GM at 11%RH at 25°C. Hydrate formation is possible during storage at 64%RH, however, the conversion is very slow and beyond the realistic time frame for the industrial storage of glucose. Hydrate formation in α- AG and hydrate loss in GM both follow random nucleation and diffusion mechanisms during α-AG equilibrium storage at 75, 80, and 85%RH at 25°C and GM equilibrium storage at 0%RH and 35, 40, and 45°C. Raman spectroscopy was used to confirm these mechanisms and was introduced as a new tool for such analysis. Since caking is a major problem in powdered ingredients, the physical stability of α-AG and GM was also studied in this research. Using a qualitative caking scale from free flowing with minimal clumping (1) to fully caked (5), the stability of α-AG and GM during relative humidity storage at 25°C was investigated. The critical relative humidity for caking was determined to be 68%RH for ‘as-is’ α-AG and 53%RH for ‘as- is’ GM. Deliquescence was not observed during the storage of α-AG and GM from 0 to 84%RH, therefore an additional mechanism of caking was used to describe the caking of crystalline materials stored below their RH0. Capillary condensation between particles leads to the formation of liquid bridges, which over time solidify due to dissolution and mass transfer across the liquid bridge without a change in relative humidity or temperature. The critical relative humidity for liquid bridge formation, RHcc, is dependent on particle size and temperature. As particle size decreases, capillary condensation increases due to the formation of smaller capillaries between particles. The small particles are able to form additional liquid bridges and also fill with condensation at lower relative humidity values, which leads to caking at a lower relative humidity compared to large particles. Hydrate formation was not found to influence caking in α-AG or GM and the presence of β-AG was not found to influence caking in GM. Storage of α-AG as a binary mixtures with sucrose (AG:S) decreased the storage stability of α-AG compared to α-AG alone. The addition of sucrose shifted the particle size distribution toward smaller particle sizes and therefore increased capillary condensation and caking was observed at a lower storage relative humidity, 64%RH at 25°C. The addition of sucrose to glucose (GM:S) did not change the storage stability of GM. The particle size distributions of GM and GM:S were very similar and caking was observed at the same relative humidity, 53%RH, at 25°C in both samples; however, the rate of caking occurred faster in GM:S compared to GM stored alone. This research significantly contributes to the literature in the areas of crystalline glucose storage parameters, hydrate formation and loss mechanisms, and physical stability during storage. Previous handling recommendations suggested storage of both α-AG and GM at or below 55%RH at 30°C. However, this research has shown hydrate loss in GM to occur at 11%RH at 25°C and caking to occur at 53%RH at 25°C. Particle size greatly influenced caking and, therefore, may be useful for future product formulation of stable powdered ingredients and ingredient mixes