Caffeine is one of the relatively soluble compounds of interest in instant-coffee production. But saying caffeine is “retained” or “lost” along the line requires distinguishing between absolute caffeine amount and caffeine concentration per unit of product mass. When water is removed from the extract, concentration can increase even though total caffeine hasn't changed.

How does extraction determine the amount of caffeine entering the liquid?

To avoid confusing these two phenomena, data needs to be converted to absolute caffeine amount at each stream. Only then can you say a step causes caffeine loss versus just making the solution more concentrated.

At the extraction step, caffeine tends to transfer into the aqueous phase along with other soluble matter. The recovery level depends on contact conditions between water and roasted-ground coffee. Temperature, time, particle size, and extraction-system design can change the degree of extraction. Since caffeine isn't representative of all soluble matter, using caffeine alone to assess overall extraction efficiency can be misleading.

Roasting is a step before the instant line but affects the chemical baseline that extraction receives. Roast level changes bean structure and chemical composition. However, when tracking caffeine, you should avoid the simple inference that darker roasting always increases or decreases caffeine in one fixed direction. Measurement needs to be on the same mass basis with controlled conditions.

Concentration and drying change concentration differently from causing loss

After extraction, concentration mainly removes water. If total caffeine in the product stream is conserved, caffeine concentration per gram of solids can change according to the water-to-solids ratio. When the concentrate is dried, water removal continues to increase the solids ratio. So a good tracking table needs both stream mass, solids, and total caffeine amount at each stage.

Studies on instant coffee show caffeine can be quantified by liquid chromatography and used together with indicators like 5-CQA to describe product composition. A study developing chlorogenic-acid-enriched instant coffee also reported caffeine across different formulas and showed the final composition depends on raw material and blending, not the drying step alone.

On what basis should sampling and results be reported?

Retention can only be calculated once the start and end points are clearly defined. If caffeine in the raw material is used as the reference, you need to know the initial caffeine mass and the caffeine amount obtained in the product. If you only compare mg caffeine per gram of powder before and after drying, you can confuse a concentration change with a real loss. This is a common error in interpreting processing data.

To track the line, sampling should be done at roasted raw material, extract, concentrate, and final powder. Each sample needs recorded stream mass and analytical method. When data is converted to absolute caffeine amount, you can determine which stage actually causes loss. If the total amount stays nearly constant but concentration changes, that's mainly a concentration effect.

The ability to retain caffeine is therefore more a mass-balance problem than an indicator specific to drying technology. To conclude whether a step causes caffeine loss, you need to track absolute amount, control the calculation basis, and place results within the extraction–concentration–drying chain. This approach helps avoid interpreting a concentration change as a chemical loss.

Another issue is the basis for reporting results. Caffeine can be reported as mg/g raw material, mg/g dry matter, or mg/g finished powder. These units answer different questions. When tracking retention, you should keep both stream mass and content per unit mass so you can convert between bases without losing information.

When does a process change actually change caffeine recovery?

If the extract is blended from multiple raw-material streams, each stream's caffeine also needs recording separately before blending. A change in blend ratio can change the final product's caffeine without relating to extraction or drying efficiency. This matters especially when the line uses multiple coffee types or roast levels.

The analytical method needs to be validated for accuracy and repeatability in the relevant coffee matrix. A multi-compound matrix can affect the analytical signal if the method isn't controlled. So retention data is only trustworthy when sampling, sample preparation, and quantification procedures stay stable.

Tracking caffeine stage by stage also has value when studying a process change. If a change increases solid recovery but caffeine recovery doesn't increase correspondingly, that's a sign the extract composition has shifted. This can be more useful information than a single overall efficiency number.

If the line has a step blending multiple raw-material streams, each stream's caffeine needs recording separately before blending. A change in blend ratio can change the final product's caffeine without relating to extraction or drying efficiency.

Some notes when reading the data

For the final product, caffeine per serving also depends on the amount of powder used. So production data and consumption data are two different layers.

References

• Nóbrega et al. (2023), Classification of instant coffees based on caffeine content and roasting degree using NIR spectrometry and multivariate analysis, Microchemical Journal, 190, 108624.
• Development of an instant coffee enriched with chlorogenic acids (2016), Journal of Food Science and Technology, PMCID: PMC4984699.
• Moroney et al. (2020), Trends in Food Science & Technology, 96, 45–60.
• NCBI Bookshelf, Coffee: extraction and instant coffee manufacture.