Isolation of Caffeine from Tea

The purpose of this experiment was to practice fundamental laboratory techniques by isolation of caffeine from tea leaves.

Camellia sinensis (tea) Caffeine

Analysis (Isolation of Caffeine)

This experiment was conducted to study the simple isolation of an alkaloid and to train with fundamental laboratory equipment like the separatory funnel and distillation apparatus. Ten commercial tea bags (23.06 grams) were used for the extraction. Boiling water (~275 mL) was used to extract caffeine from the solid leaves. This was the most efficient solvent because water causes tea to swell, allowing for a caffeine solubility of 670 mg/mL. Sodium carbonate was added to the tea/water solution to ensure that caffeine remained in its free base form and to keep acidic tannins water soluble. The mixture was decanted after seven minutes and cooled.

In a separatory funnel, the water solution and dichloromethane (30 mL) were gently mixed. The dichloromethane solvent was used because caffeine is basically the only substance from the water solution that will dissolve in it. After mixing, the two layers began to separate. At the point of separation an emulsion began to form. The emulsion looked foamy and grew quite large (1/8 of the volume). The emulsion dwindled after half an hour. The solutions were separated and the water solution was added to dichloromethane (30 mL) in the funnel once more to ensure maximum recovery of caffeine. In an effort to prevent another emulsion, a highly saturated solution of salt water was also added to the solutions. After mixing the separatory funnel, an emulsion formed, but it was not nearly as large as before. The second volume of dichloromethane was separated from the water solution and added to the first volume.

To ensure that no water remained in the dichloromethane solution, calcium chloride pellets were added. These pellets are very hydrophilic. The pellets left the dichloromethane solution crystal clear. Now the solution should have contained only the solvent and caffeine. To isolate caffeine, the solution was filtered through silicon paper (another drying agent) into a distillation apparatus. The solvent boiled away leaving a fine white powder, caffeine.

Table A: Percent Recovery of Caffeine

Williamson, author of Organic Experiments, anticipates that this exercise should yield a mg/gram of tea (see Table A). At this standard, the experiment had 91% recovery of the caffeine. However, Williamson also describes the caffeine as greenish in appearance, which was not seen in the experiment. The experiment seemed highly successful in isolating the alkaloid.

Editor’s note (October 22, 2026): The student’s procedure and results are left as written. A few points need clarifying.

• The expected-yield figure is missing. The sentence above says Williamson’s book expects “a mg/gram of tea,” but the number itself is missing from the original write-up. We couldn’t confirm the figure, so the gap is left as is. The 91% recovery is measured against the table’s “~23 milligrams” (21 mg ÷ 23 mg ≈ 91%).
• The calcium chloride, not the “silicon paper,” was the drying agent. Drying agents like anhydrous calcium chloride remove the trace water left in an organic solution by forming hydrates, and they are filtered or decanted off afterward (LibreTexts, Reaction Work-Ups). “Silicon paper” most likely means silicone-treated (phase-separator) filter paper. This paper is water-repellent: it lets the organic solvent through and holds back droplets of the water layer (Cytiva, Whatman 1PS Phase Separator). It can catch stray water and it removed the calcium chloride pellets, but it doesn’t absorb dissolved water the way a drying agent does. If plain filter paper was used, it only removed the pellets.
• The 670 mg/mL figure is for boiling water. One gram of caffeine dissolves in about 1.5 mL of boiling water (roughly 670 mg/mL), but in only about 46 mL of water at room temperature (about 22 mg/mL at 25 °C) (PubChem, Caffeine). The high number reflects the hot water, not swelling of the leaves.

How the liquid-liquid extraction works

When a solution is shaken with a second liquid that doesn’t mix with it, each dissolved compound divides itself between the two layers in a fixed ratio called the partition (or distribution) coefficient, K = (concentration in the organic layer) ÷ (concentration in the water layer) (LibreTexts, Extraction Theory). Caffeine dissolves in dichloromethane at about 140 mg/mL, compared with about 22 mg/mL in room-temperature water, so most of it moves into the dichloromethane (SUNY Oneonta, Isolation of Caffeine from Tea Leaves). Extracting with two or more fresh portions of solvent, as in this lab, is more efficient than a single extraction (LibreTexts).

• Why sodium carbonate is added: tannins and gallic acid also dissolve in dichloromethane. The base converts these acidic compounds into negatively charged salts that stay in the water, so the caffeine is extracted nearly pure (LibreTexts, Extraction of Caffeine From Tea Leaves).
• Why emulsions form: the tannin salts act like soap and help the two layers stay mixed, so the funnel should be swirled or rocked gently rather than shaken hard (SUNY Oneonta). Adding salt (brine) makes organic compounds less soluble in the water layer, which can help the layers separate (“salting out”) (LibreTexts, Step-by-Step Procedures for Extractions).
• Which layer is which: dichloromethane is denser than water (specific gravity about 1.33), so it is the bottom layer. It is drained out through the stopcock, and the water layer stays in the funnel for the next portion of solvent (LibreTexts; NIOSH).

For another student’s write-up of a similar extraction, which uses sodium sulfate as the drying agent, see Isolation of Caffeine from Tea Leaves.

Safety note on dichloromethane

Dichloromethane (methylene chloride) is volatile; it boils at about 40 °C (104 °F). It can be absorbed by breathing the vapor or through the skin. Exposure can irritate the eyes and skin and cause drowsiness or dizziness. NIOSH lists it as a potential occupational carcinogen (NIOSH Pocket Guide, Methylene chloride; PubChem, Dichloromethane). OSHA also treats it as a potential occupational carcinogen (OSHA, Methylene Chloride). The International Agency for Research on Cancer classifies it as probably carcinogenic to humans (Group 2A). The body also converts part of it to carbon monoxide, and NIOSH lists the cardiovascular and central nervous systems among its target organs (PubChem; NIOSH). Work with it in a fume hood, keep it off your skin and out of your eyes, and vent the separatory funnel often.