The experimental setup recreated the procedure currently used in the whisky industry and was fundamentally different from the setup used by Matthew Ferguson Stewart in his earlier, non-representative study. Instead of filtering the micelles with cellulose filters (coffee/tee filters), for this study a filling of kieselguhr powder with a depth of 20mm to 40mm was put into a chromatography column (capacity: 0.5 litres) over a cellulose filter (coffee filter).
The German Wikipedia describes kieselguhr as a whitish, powdery substance predominantly made of the silicon dioxide shells of fossilised diatoms (algae). They have a porous structure. One millilitre of kieselguhr contains the shells of approximately one billion diatoms. The inner surface is very large, which makes it ideal for filtering liquid foodstuffs such as wines, juices and spirits. Kieselguhr is sterilised in ovens at temperatures ranging from 600° to 800°C, so that liquids are not organically contaminated when filtered.
A customary double neck round bottom flask with a capacity of 1 litre was put under the chromatography column as a collection container. The second neck of the round bottom flask was sealed with a core containing a PTFE valve in order to prevent the alcohol from evaporating during filtration. At the same time pressure was applied to the chromatography column from above via an identical core with valve through a PTFE hose, using an aquarium air pump (EHEIM air pump 100), in order to accelerate the filtering process. To prevent the building up of counterpressure in the round bottom flask the valve at the round bottom flask was slightly opened during filtration.
The experimental equipment was chosen in such a way that it could fit into a customary refrigerator without freezing compartment (Siemens KI 18RA20). The refrigerator was set to maximum cooling capacity, resulting in an average temperature of 0° to 4°C. A remote-reading thermometer (Oregon Scientific) was used to monitor the temperature.
The bottles designated for filtration were also cooled down to 0° to 4°C in the same refrigerator. Before the start of the experiment the bottle temperature was monitored with a no-touch remote-reading thermometer with fixing laser (Voltcraft IR-270L).