A Novel Automated Liquid/Liquid Extraction Technique for the Determination of Caffeine in Coffee

Significance of the Topic

Accurate measurement of caffeine content in coffee is critical for product quality control, labeling compliance and consumer information. Variability in caffeine concentration arises from bean type, roast level and brewing method, making reliable analytical techniques essential for producers and laboratories.

Aims and Overview

This application study evaluates an automated liquid–liquid extraction (LLE) approach to quantify caffeine in various coffee blends and brands. The objectives include comparison of light, medium, dark and flavored roasts from Brand A, assessment of medium roast differences across Brands A, B and C, and examination of brewing method effects using Brand C.

Used Instrumentation

• EST Analytical FLEX autosampler with 250 µL syringe
• Agilent 7890 gas chromatograph
• Agilent 5975 mass spectrometer
• Titan PTV large-volume inlet
• Restek Rxi-5Sil MS capillary column (30 m × 0.25 mm × 0.25 µm)

Methodology and Experimental Parameters

Caffeine was extracted automatically by adding methylene chloride to 2 mL coffee samples, agitating at 25 °C for five minutes, allowing phase separation and transferring the organic layer directly into the GC inlet. A 5000 ppm caffeine standard in water was used to generate a calibration curve. Key sampler parameters included dual rinse cycles, 250 µL solvent volume and controlled air gaps. GC oven programming ranged from 40 °C to 310 °C with helium carrier gas at 1.4 mL/min. MS scan mode covered m/z 40–500 after a 3 min solvent delay.

Main Results and Discussion

The caffeine levels measured per single serving cup were:

• Decaf: 27.4 mg
• Morning Roast: 96.5 mg
• Medium Roast A: 110.6 mg
• Dark Roast: 110.8 mg
• Vanilla Flavored: 116.5 mg
• Mocha Flavored: 119.8 mg
• Medium Roast B: 123.5 mg
• Medium Roast C: 102.6 mg
• Brewed Pot C: 181.4 mg

The dark roast showed similar caffeine to the medium roast, while flavored coffees had slightly higher levels due to added compounds. Brand B yielded the highest medium roast caffeine, and Brand C the lowest. Brewing a full pot significantly increased caffeine compared to a single cup. The automated extraction also effectively removed flavoring agents, enabling clear chromatographic separation and reproducible quantification.

Benefits and Practical Applications

The automated LLE workflow reduces manual handling, improves throughput and enhances reproducibility. It is well suited for routine quality assurance in coffee production, formulation studies, and regulatory compliance testing. The method’s precision and ease of use make it a strong candidate for high-volume laboratory environments.

Future Trends and Potential Applications

Further integration with hyphenated techniques such as LC-MS/MS or GC-IR can expand analyte coverage and enhance identification. Development of microextraction formats or green solvent systems will improve sustainability. Adaptation to other beverage matrices or biofluids could broaden the method’s scope, and coupling with real-time process monitoring may enable in-line quality control in industrial settings.

Conclusion

The described automated liquid–liquid extraction using the FLEX autosampler combined with GC/MS provides a robust, reproducible and efficient approach for caffeine determination across diverse coffee samples. It offers high accuracy, minimal manual intervention and clear potential for scaling in analytical laboratories.

References

No external literature references were cited in the original application note.