logo

HPLC Caffeine Analysis in Beverages: Wanyi LC3500 DAD Method Validated per GB 5009.139-2014

2026-10-21

Latest company case about HPLC Caffeine Analysis in Beverages: Wanyi LC3500 DAD Method Validated per GB 5009.139-2014
Case Detail
Project Background

Caffeine is a naturally occurring methylxanthine alkaloid found in coffee beans, tea leaves and cocoa, and it is one of the most widely used functional additives and characteristic flavour compounds in carbonated soft drinks, energy drinks, tea beverages and ready-to-drink coffee. Moderate intake sharpens alertness and relieves fatigue, yet excessive or long-term consumption can trigger palpitations, insomnia and nervous excitement, and sensitive groups such as pregnant women, children and people with cardiovascular conditions tolerate far lower doses. National food safety standards therefore set explicit caffeine limits for beverages.

In China, GB 5009.139-2014, the National Food Safety Standard for the determination of caffeine in beverages, is the statutory basis for caffeine quantification in beverage matrices. High performance liquid chromatography (HPLC) has become the preferred technique for complex matrices such as composite drinks, milk-containing beverages and herbal functional drinks, thanks to its excellent separation, strong resistance to matrix interference and highly accurate quantification.

Customer Challenge

A beverage manufacturer needed to verify caffeine levels in two very different product lines: a carbonated cola-type drink and a milk tea product. The two matrices need different preparation strategies: carbonated samples must be degassed and cleaned up, while milk-containing products need protein precipitation before injection. Both had to sit comfortably inside the calibration range, and the chromatographic system had to stay free of matrix build-up over long injection sequences.

Instrumentation and Reagents
ItemConfiguration
HPLC systemWanyi LC3500 with diode array detector (DAD)
Analytical columnC18, 4.6 x 250 mm, 5 um
Mobile phaseMethanol / Water = 24 / 76
Flow rate1.0 mL/min
Column temperature25 degrees C
Detection wavelength272 nm
Injection volume10 uL
Auxiliary equipmentUltrasonic cleaner, vortex mixer, 0.45 um aqueous microporous membrane

Reagents included a caffeine reference standard, HPLC-grade methanol, magnesium oxide, trichloroacetic acid solution and ultrapure water.

Method
Standard solution preparation
  • Stock solution (2.0 mg/mL): 20 mg of caffeine standard, weighed to 0.1 mg, dissolved in methanol and made up to 10 mL; stored at 4 degrees C with a six-month shelf life.
  • Intermediate solution (200 ug/mL): 5.0 mL of stock diluted to 50 mL with water; stored at 4 degrees C with a one-month shelf life.
  • Calibration working solutions: 0.5, 1.0, 2.0, 5.0 and 10.0 mL of intermediate solution diluted to 10 mL, giving 10.0, 20.0, 40.0, 100 and 200 ug/mL; prepared fresh on the day of analysis.
Sample preparation

Cola-type beverage: degas the sample in an ultrasonic bath at 40 degrees C for 5 minutes. Weigh 5 g (accurate to 0.001 g), dilute to 5 mL with water so that the caffeine concentration falls within the calibration range, shake well, add 0.5 g magnesium oxide, shake again and allow to settle. Filter the supernatant through a 0.45 um membrane before injection.

Milk-containing coffee and tea drinks: weigh 1 g (accurate to 0.001 g), dilute to 10 mL with trichloroacetic acid solution, shake well and let stand so that proteins precipitate, then filter the supernatant through a 0.45 um membrane.

Validation Results
Performance parameterResult
Calibration range10.0 to 200.0 ug/mL, correlation coefficient r > 0.9999
Repeatability (n = 6)Retention time RSD and peak area RSD both < 0.3%
Caffeine content, cola sample95.051 mg/kg
Caffeine content, milk tea sample260.0456 mg/kg
Spike recovery at 40 ug/mLGreater than 94% for both matrices
Limit of detection0.0446 mg/kg (cola); 0.75 mg/kg (milk-containing tea drink)

Peak shapes were symmetric and theoretical plate counts were high. Blank solvent injections showed no interference at the caffeine retention time, confirming that magnesium oxide clean-up and trichloroacetic acid protein precipitation remove matrix components effectively. Six consecutive injections gave stable retention time and peak area, showing excellent instrument precision, while recovery above 94% in both matrices proved minimal sample loss and a robust preparation procedure.

Conclusion

Following GB 5009.139-2014, the Wanyi LC3500 HPLC system with diode array detection delivered a fully validated workflow for caffeine determination in beverages. Linearity across 10.0 to 200.0 ug/mL with r > 0.9999, repeatability below 0.3% RSD and recovery above 94% confirm that the method is accurate, precise and reproducible for both carbonated and milk-containing products. Detection limits of 0.0446 mg/kg and 0.75 mg/kg show the low noise and high sensitivity of the system, making it suitable for routine quality control, label verification and regulatory compliance testing in beverage manufacturing and third-party laboratories.

Why This Configuration Works
  • Diode array detection at 272 nm provides selective, low-noise quantification of caffeine in coloured and turbid matrices.
  • A simple methanol and water mobile phase on a C18 column keeps method transfer between laboratories fast and inexpensive.
  • Matrix-specific clean-up, magnesium oxide for carbonated drinks and trichloroacetic acid for milk-based drinks, protects the column and extends its service life.
  • Stable retention times and symmetric peaks reduce re-injection and repeat work in high-throughput QC laboratories.