High-Performance Liquid Chromatograph (HPLC)

HPLC Determination of Caffeine in Soft Drinks


pp. 542-544, 595-604, and 611-613 in Harris text



Chromatographic techniques are widely used in all areas of science because they allow the analyst to separate and quantify the components of a mixture.  In this experiment you will use high-performance liquid chromatography (HPLC) to determine the amount of caffeine in soft drinks. 

The principle on which chromatographic methods are based is simple.  Compounds in a mixture are separated from each other based on their preferences for one of two different solvents (or phases) in contact with each other.  For example if a polar solvent and a nonpolar solvent are brought into contact, polar molecules will prefer to be in the polar solvent, while nonpolar molecules prefer to be in the nonpolar solvent.  In a chromatography experiment one of these solvents is stationary (the stationary phase), while the other solvent (the mobile phase) flows over it.  HPLC utilizes a small-diameter column packed with small particles coated with the stationary phase.  A diagram of the basic components of an HPLC system is shown in the figure below.

A piston-based pump continuously delivers liquid mobile phase at a steady flow rate from the mobile phase reservoir through the system.  A very small volume of sample is injected with a syringe into the flowing mobile phase where it is carried to the column.  In the column the various components of the sample mixture are separated from each other into bands.  These bands “elute” from the column and move toward the detector.  Several different types of HPLC detectors are available; we will use a variable-wavelength UV-vis spectrophotometric detector that measures the absorbance of the flowing stream of liquid. When a species with an absorbance different from the mobile phase passes the through the detector, a peak is generated.  The result is a plot of detector signal (absorbance) vs. time.  This plot is called a chromatogram.

The chromatographic column you will use in this experiment is a reversed-phase column, meaning that the stationary phase is nonpolar (this is the reverse of older methods).  The column consists of small, porous silica particles that have hydrocarbon chains eighteen carbon atoms long (a C-18 column) chemically bonded to their surfaces.  Mobile phases used in reversed-phase columns are polar.  In this experiment you will use a mobile phase which is 55% 0.025 M phosphate buffer (pH 3.0) and 45% methanol.  Separation of components in a sample relies on different substances having different preferences for the mobile and stationary phases.  Molecules that are very polar will elute from the column early, since they prefer to stay in the mobile phase. Nonpolar molecules, preferring the environment of the stationary phase, will be retained on the column for longer periods of time.

For our experiments we will be using a Varian ProStar HPLC system, shown in the picture above.  You’ll notice there are two pumps and two mobile phase containers.  This allows us to pump the buffer from one pump and the methanol from the other.  The computer software will allow us to mix the two liquids in our desired 55:45 ratio.  If this ratio stays constant throughout the experiment, we are performing an isocratic separation.  Sometimes, we can improve a separation by changing the ratio of mobile phases over the course of the experiment.  This is referred to as gradient elution.

You’ll notice also that our system uses an autosampler instead of a simple injector.  The autosampler will allow us to run many samples in a row automatically.  A picture of the autosampler with the lid raised is shown below.


The column is housed in a column heater for precise temperature control.  The detector is an absorbance detector as described above.  It actually allows us to monitor two separate wavelengths in the UV and/or visible region simultaneously.

In this experiment you will prepare a number of caffeine standard solutions, along with two soft drink samples, and then set up a sequence on the autosampler to inject each one in order.  From your results you’ll generate a calibration curve based on the peak areas for the standards, then fit the peak area of caffeine from your samples to the calibration curve to calculate the concentration of caffeine in the soft drinks.