Cold-concentration of food extracts through application of gas hydrates (clathrates)

Gas hydrates are solid structures made of water and gas that remind in their configuration of water cages surrounding a gas molecule. With CO2, a cheap gas that is often used in the food industry for various applications, they form under moderately high pressure and low temperature, yet above the water freezing point. Thanks to their high water-binding capacity, they offer prospect to a gentle concentration technology by removing water from the concentrate in solid form. With the use of established CO2 hydrate state diagrams, target pressure and temperature levels were defined for the experiments. CO2 gas hydrate formation was triggered in a stirred-tank batch reactor by applying temperature levels of 1 and 4 °C, pressure levels of 36 or 41 bar, and coffee concentrations varying between 7.5 and 40 wt.%. The viscosity of the liquids before gas hydrate formation was analyzed to understand the influence of flow conditions on gas hydrate formation. Based on the trial trajectories using different temperature and pressure combinations and the corresponding gas hydrate induction time, optimal process conditions were determined. The induction time corresponds to the time period between the start of the process and the formation of first gas hydrate crystals and can be prolonged through the presence of certain molecules, the gas hydrate inhibitors. Since gas hydrate formation is an exothermic reaction, the induction time can be determined by a sudden temperature increase. The microscopic and macroscopic structures of the obtained gas hydrate, its gas content, growth over time and solute inclusions set the base for gas hydrate characterization. By measuring the total solids content (TS) of the concentrate and the total solids recovery, the process concentration efficiency was determined. Key coffee aroma compounds of concentrates and hydrates were analyzed. Additionally, coffee solutions were replaced by sugar solutions of 7.5 wt.% to test the concentration potential and process robustness with simplified model systems containing isolated coffee constituents. The viscosity of the coffee solutions increased with increasing concentration. By fast CO2 injection, gas hydrate formation was successfully triggered at all chosen process conditions and coffee concentrations with high reproducibility. Induction times were not influenced by the applied pressure but increased at higher temperatures and coffee concentrations, also owing to the presence of gas hydrate inhibitor molecules in coffee. At concentrations of up to 20 wt.% TS and by applying 1 °C target temperature, gas hydrate formation was triggered so quickly that induction times could not be quantified. Various gas hydrate structures were observed, ranging from large, continuous macroscopic structures to small gas hydrate crystals dispersed in the coffee concentrate. Particularly the large gas hydrate structure is highly desired for simple separation from the concentrate, and its formation was in general correlation with a low process temperature and extremely short induction times. Gas hydrate structure growth occurred linearly over the tested time range of five hours and consisted of around 18 wt.% gas. A more detailed analysis using cryo scanned electron microscopy revealed a combination of larger or smaller dense microscopic structures surrounded by porous ice layers. The applied target pressure did not affect the quality of the microstructure. Using an initial coffee concentration of 13 wt.% TS, the formed hydrate consisted of around 3 wt.% coffee solids, while the concentrate contained 21 wt.% solids. 30 key aroma compounds for coffee were identified in hydrate and concentrate and were non-selectively distributed in both phases, directly proportional to the TS in the respective phase. This shows that the presence of coffee aroma in the gas hydrate phase mostly results from aroma-containing coffee liquid adhering to the hydrate. No difference in aroma profiles between the initial 7.5 wt.% coffee solution prior concentration and hydrate/concentrate after concentration was observed. By triggering gas hydrate formation in 7.5, 13 and 21 wt.% starting TS, a proportional increase in TS of both the hydrate and the concentrate phase in correlation to the initial concentration of the coffee solution was observed, revealing the possibility of a stepwise batch concentration. Eventually, a concentration of 40 wt.% TS was obtained, a value of high industry relevance. By replacing the coffee solutions with 7.5 wt.% sugar solutions (arabinose, galactose, mannose) but using identical process parameters, the same degree of concentration was obtained. In summary, gas hydrate concentration technology bears the genuine potential to replace conventional concentration technologies fully or at least partially. In order to further improve concentration efficiency, continuous processing options could be explored in future works.