determination of caffeine in beverages by high performance liquid

determination of caffeine in beverages by high performance liquid
Skoog D. A., Holler F. J., and Crouch S. R., Principles of Instrumental Analysis, 6th
edition, Harcourt Brace College Publishers, 2007. Chapters 26 and 28.
Liquid chromatography (LC) refers to chromatography in which the mobile phase
is a liquid. The four basic types of LC are partition chromatography; adsorption or
liquid-solid chromatography; ion exchange chromatography and size exclusion/ gel
chromatography. They differ in the exact nature of the stationary phase, thus the
mechanism/ processes insuring differential retention of analytes. Early LC utilized long
glass columns with wide diameters (1 to 5 cm), and solid support particles of diameter in
the 150- to 250-µm range to insure reasonable flow rates (still less than 1mL per minute).
Under these conditions, separations could take up to several hours.
High-performance liquid chromatography (HPLC) was developed to increase
speed and efficiency in liquid chromatography. Decreasing the size of the solid support
material increased efficiency; i.e. decreases the height of theoretical plates. In the van
Deemter equation [1], covered in the GC module,
H = A+
+ Cu = A + + (C S + C M )u
the C coefficient which relates the linear velocity of the mobile phase to mass transfer
between phases, can be expressed as a sum of two coefficients C S and C M , related to
the stationary and mobile phase respectively. The C M coefficient is directly
proportional to the square of the diameter of the particles, leading to the conclusion that a
decrease in the size of the particles of the stationary phase supporting material will result
in the decrease of the theoretical plate height ( H ). However, use of smaller size
particles (3-10 µm) requires high pumping pressures (several thousands psi) for achieving
separation within reasonable time periods.
Whereas in gas chromatography the mobile phase does not interact with the
analytes and serves only to transport analytes through the column, in liquid
chromatography, the mobile phase interacts with the analyte, thus plays a very important
role in affecting separation parameters such as retention times and resolution. The role of
the mobile phase on separation adds versatility, flexibility and the range of forces that can
be exploited to achieve separation of various complex mixtures by liquid
In bonded phase liquid chromatography, the stationary phase is composed of a
relatively low molecular weight solvent-like molecule covalently bound to a solid support
particle that is packed in the column. The solid support in modern high performance
columns typically consists of 3 to 10 µm-diameter porous silica gel particles (5 µm in this
experiment). In reverse phase bonded phase (RPBP) chromatography, the bonded phase
is non-polar and the mobile solvent phase is polar. And, in normal phase bonded phase
(NPBP) chromatography, the bonded phase is polar, and the mobile solvent phase is nonpolar. The column in this experiment is packed with an octadecyl (C18) reverse-phase
bonded stationary phase.
In chromatography, a component injected onto the column will be distributed
between the mobile, M, and the stationary, S, phases according to its affinity for both
phases. Thus, in reverse phase HPLC, a polar molecule that interacts more strongly with
the M phase will elute quickly from the column. In contrast, a non-polar molecule would
interact more strongly with the S phase and so would elute more slowly from the column.
The retention time of a component on the column is related to the capacity factor, k i , for
a column, which is in turn related to the distribution coefficient, K i , of the analyte i
between the S and M phases [1].
Ki =
ni , s
k i' =
ni , M
= Ki
W s = weight of S phase in column (g)
V M = volume of M phase in column (L)
[i ]s = concentration of i in s, mol/g
[i ]M = concentration of i in M, mol/L
Experimentally, k i is determined from the retention time for component i, t R,i ,
and from the void time, t 0 . The void time is usually obtained from the time required for
either the solvent peak, or an unretained component of the mixture, to elute from the
k i' =
t R,i − t 0
Use of solvent mixtures
Ideally, in a separation, the capacity factors (k’) for all components should lie
between 2 and 5 to effect good baseline resolution of the peaks in a limited period of
time. For a given stationary phase the k' of a particular component can be controlled by
changing the polarity of the mobile phase. Tables of solvent polarity or elution strength
have been prepared [2, 3]. However, the exact order of elution strength will depend on
the stationary phase and the components examined. Fine control of elution strength of
the mobile phase is obtained by using binary and ternary mixtures of solvents. In reverse
phase HPLC the most common solvent mixtures are H2O and methanol (CH3OH) or H2O
and acetonitrile (CH3CN).
Use of solvent gradient elution
Unfortunately, in complex mixtures k' can vary from zero to, more than twenty
for a single solvent strength. This leads to the "general elution problem" where no one
set of conditions is effective in removing all components from a column in a reasonable
time period, while still attaining resolution of each component. Since solvent polarity has
a strong effect on k' in reverse-phase chromatography, it is convenient to use a binary
solvent mixture (e.g. H2O and CH3OH) consisting of two solvents of differing polarity,
and change the percentage of each in the mixture during the elution of the sample. This
is known as gradient elution chromatography. In gradient elution, the solvent
composition changes over time as shown in Figure 1.
of A
in A&B
Figure 1. Solvent gradient of solvent A with solvent B.
Thus, the mobile phase polarity, or elution strength, varies during the chromatogram.
Note that various gradient shapes can be applied. Gradient elution will not be used this
semester, however an example of its use is discussed below.
Gradient elution is customarily used to separate components of a homologous
series. A typical experiment in Instrumental Analysis laboratory involves separation of a
homologous series of para-hydroxybenzoate esters. A mixtures containing the methyl-,
ethyl-, propyl-, butyl…nonyl hydroxybenzoates is separable within reasonable time by
applying gradient elution. Use of gradient elution allows good resolution of the lower
molecular weight components, while eluting the octyl and nonyl esters in a reasonable
period of time.
The solubilization of a solute in a solvent involves intermolecular interactions of
the solute (i) and solvent (j) with energy Eij [4, 5]. For an organic molecule consisting of
several different functional groups, e.g. CH3-(CH)n-Y, it is possible to break the overall
interaction energy up into the sum of individual interaction energies,
E ij = E CH , j + nE CH , j + E y , j
Martin first showed [6] that the log of the distribution coefficient, K i , for a solute
between two phases j and l is approximately proportional to the sum of the differences of
the individual interaction energies as illustrated below for CH 3 − (CH 2 ) n − Y .
) (
) (
log K iα E CH , j − E CH ,l + n E CH , j − E CH ,l + EY , j − EY ,l
The Martin equation thus predicts that if E CH 2, j > E CH 2,l , log K i (and log k i )
will increase as the number, n, of a certain functional group in a homologous series
increases. In reverse phase chromatography, the E CH 2, S > E CH 2, M since the mobile
phase is polar and the octadecyl phase is not. Thus the hydroxybenzoate methyl ester will
be less retained than the hydroxy-benzoate nonyl ester.
In this experiment, high performance reversed phase liquid chromatography is used
1) Separate the various substances contained in four common beverages from caffeine.
2) Determine the concentration of caffeine in these beverages.
Separation will be conducted under ISOCRATIC conditions (i.e. elution at constant
solvent composition). This experiment illustrates the power of HPLC in analyzing
components in complex mixtures. Traditional method for the determination of caffeine is
via extraction followed by spectrophotometric quantitation. HPLC allows for rapid
separation and quantitation of caffeine from the many other substances found in these
beverages including tannic acid, caffeic acid and sucrose.
A Varian Model 5000 Liquid Chromatograph interfaced to a variable wavelength
UV/ VIS detector (Varian UV-50) and a strip chart recorder will be used. Reading of this
section may be facilitated by consulting the diagrams of important components appended
at the end of this hand out.
The column is a commercially packed 5-µm octadecyl (C18) bonded phase column
with a length of 25 cm.
The UV-50 Variable Wavelength detector is a double beam manual
spectrophotometer, equipped with a deuterium and a tungsten lamp, and a
monochromator, to provide wavelengths from 200nm to 720 nm. The optical path length
is 1 cm and the sample flow cell volume is 8-µL. The wavelength of measurement will
be 254 nm, the optimum wavelength for caffeine determination.
Solvents and Pump
Two solvent reservoirs, one containing HPLC grade methanol and the other
HPLC grade water, are already connected to the pump in the chromatograph. A two way
valve (Purge valve) permits routing of the solvent upstream the column (for purging)
when open (clockwise rotation), or to the column (counterclockwise rotation) when
The chromatograph is fitted with an automatic VALCO sample injector
comprised of a sampling valve, an external 10-µL sample loop and a fill-port fitting
assembly. The injector is controlled by the microprocessor. A programmed event 4
injects the content of the sample loop onto the column, and a programmed event 0 returns
the valve to the load position.
Operation: general starting procedures
1. Turn power on the HPLC module, the UV/VIS detector, the recorder and the
attenuator (unit above the recorder).
2. Select the wavelength (254 nm) and the bandwidth (4 nm) on the UV/VIS detector.
3. Use the CRT display and Keyboard to check and change the current conditions of the
instruments and to build operating programs. After turning on the HPLC, the display
should read, Power recovered, Programs saved. Press the DSPL key. The Flow rate
(1.5-mL/ min.), solvent composition (%B 100), and EVNT (0) settings will be
FLOW, %, PRGM, RSVR and EVNT are always used with the TIME key.
4. Always use the instrument gradient program in order to gradually change to the
desired composition.
5. The instrument should be on for at least 30 minutes to equilibrate.
D.1 Mobile phase, Standard and Samples preparation
1. Mobile phase/ solvent
1 L of a 25%: 75% Methanol: Water mixture (% volume).
The TA should have prepared the mobile phase/ solvent before the beginning of the
lab. HPLC solvents should always be of high purity (HPLC grade or freshly distilled)
in order to prevent alteration of the column chemistry by strongly adsorbing
impurities. The TA should have filtered the solvents to remove particulate larger than
0.45 µm, in order to preserve pump and column life.
2. Standards
Pipette 5-, 10-, 15- and 20-mL portions of the stock 10-ppm caffeine solution
provided into 25-mL volumetric flasks, and dilute to volume with the solvent mixture.
Pipette 1-mL of the 100 ppm solution of acetaminophen provided into 25-mL
volumetric flask, and dilute to volume with the 10-ppm solution of caffeine. The
stock solutions are already filtered so the standards do not need filtration.
3. Instant Coffee
Put ~150-mL of hot tap water into a 250-mL Erlenmeyer and set the flask on a hot
plate to boil (heat the indicated amount of water for the other sample at the same
time). Spoon out the amount of instant coffee appropriate for one cup, about one
rounded teaspoon (1.5 to 2.5 g), onto a tarred weighing paper and record the mass of
coffee to the nearest 10 mg. Do not use more than 2.5 g. Once the water has boiled,
stir in the coffee and set aside to cool. When the coffee has cooled, transfer it
quantitatively to a 200-mL volumetric flask and make up to the mark with water.
4. Tea
Bring to boil about 400-mL of hot tap water in a 500-mL flask. Meanwhile, weigh
one tea bag and one emptied tea bag. Once the water has boiled, prepare the tea as
you normally would, recording the time you steep the tea to the nearest 0.5-minute.
Once brewed, remove the tea bag. When the tea has cooled, transfer quantitatively to
a 500-mL volumetric flask and dilute to the mark with tap water.
5. Dilution and Preparation for Injection
Prepare a 20-fold dilution of the tea and the two degassed cola beverages, by
pipetting 5-mL of each into 100-mL volumetric flasks, and diluting to volume with
the solvent. Pipette 5-mL of the coffee solution into a 250-mL volumetric flask and
dilute to volume with the solvent to obtain 1:50 dilutions.
Before injection, the dilute solutions of tea and coffee must be filtered to remove
particulate! Filter only ~10-mL of your samples (for cleaning the sample loop and for
injections). The filtered sample can be kept in 4-dram vials. The vials should be very
clean (absolutely no particles). Rinse twice with a few drops of filtered samples.
1. Before beginning, check the solvent composition (must be 25 % CH3OH, indicated by
%B = 25 on the CRT screen) and the flow rate (must be 1.5-mL/ min.). Open the
purge-valve (clockwise rotation) and purge the system for five minutes. Remember
to close it after flushing (counterclockwise rotation). Make sure the entry for EVNT
is 0. Press start on the chromatograph to start pumping the solvent to the column for
5 to 10 minutes.
If the composition of the solvent is different from 25 % (e.g. 100 %), set a gradient
program to change the composition. Enter TIME = 20 min., % = 25, followed by
ENTER. Check that at TIME = 0, the value for EVNT is 0, and the value for % is
100 %. Start the pump. When the solvent composition is 25 %, do not stop the
pump, just move to the next two steps.
2. While pumping, use the COARSE ZERO Knob on the detector to move the pen to the
zero position on the recorder, with the attenuator set at 8. This sets the zero
absorbance reading of the detector using our mobile phase as the blank.
3. Cleaning the sample loop before injection.
Rinse the syringe with several aliquots of the solvent then fill it with the solvent.
Make sure the sample injector is in the load setting (i.e. EVNT = 0). Insert the
syringe needle all the way into the port. Do not push too hard on the syringe; just
make sure it is bottomed.
Flush about 250-µL of mobile phase through to clean the loading passages (do the
same to prevent cross-contamination between runs). Refill the syringe, wipe clean,
reinsert, and flush in another 250-µL, making sure not to inject any air bubbles. The
flushing syringe is removed while flushing is being completed so as to fill the special
teflon sleeve filling to top. This procedure prevents any bubbles from being pumped
into the loop.
Rinse the syringe with several aliquots of the sample to be injected. Before injection,
be sure there are no air bubbles in the syringe. Point the needle upwards, tap the
syringe with your fingers, and then push the plunger to expel the air bubbles.
4. Flush and fill the sample loop with one of your standard solutions.
5. To practice filling the sample loop and injecting use the following program
6. In order to optimize reproducibility; use one continuous fill of the sample loop.
Therefore, use the 1000-µL syringe so as to have sufficient sample in the syringe for
refilling the sample loop between injections; and use the following program to make
multiple injections at set intervals without stopping the pump.
TIME 0.0
TIME 1.0
TIME 1.2
TIME 2.0
TIME 2.2
TIME 3.0
TIME 3.2
TIME 4.0
TIME 4.2
(1st injection after 1 minute)
(close injector valve)
(2nd injection, 1 minute after first injection)
(close injector valve)
(3rd injection, 1 minute after 2nd injection)
(close injector valve)
(4th injection, 1 minute after 3rd injection)
(close injector valve)
7. Press start to run the program
(you have 48 seconds to do this). Check that the retention times are reproducible
(±0.05 min) to ensure that the column is equilibrated. Then, use the same procedure
to make four replicate injections of all the standards. Remember to reset the program
between runs.
8. Use a single injection program to obtain one 'clean' chromatogram of the Coca-Cola
9. Next, make four replicate injections of the Coca-Cola sample using two minutes
intervals between injections in the program.
10. Repeat steps 8 and 9 to obtain one 'clean' chromatogram and four replicate runs for
the other beverages.
11. When you have finished your injections, program the instrument to gradually change
the mobile phase composition to 100 % CH3OH
TIME = 0
% = 25
TIME = 20
% = 100
To erase previous values, use the same commands, then press DEL instead of
ENTER. Make sure you have erased all previous entries or switch to a new program.
12. Flush the sample loop with the solvent.
13. Turn off the detector, the recorder, the attenuator and the chromatograph.
1. Plot the peak areas and peak heights calibration curves for caffeine using units of ppm
for concentration. Use the triangulation method to obtain peak areas (peak height
times peak width at one-half peak height).
2. Use linear regression analysis to obtain the best straight line through the experimental
points [1,7].
3. Calculate the standard deviation on the slopes and intercepts as described in
Appendix 1 of your text Book [1] or reference [7].
4. From the calibration curves report the concentration of caffeine in the injected
beverage samples, and in the original samples in ppm. Evaluate the overall standard
deviation for each of your results.
5. Report the % by weight of caffeine in each of the original solid materials.
6. Calculate the capacity factor of caffeine using the retention time of acetaminophen as
t0 .
7. Explain the rationale for using a reverse-phase C18 column for the determination of
8. Derive the relationship between retention time and capacity factor from the following
v = L / t 0 = average linear velocity of mobile phase
vi = L / t R,i = φ i,M × v = average linear velocity of i in mobile phase
ni , M
φ i,M =
ni , s + ni , M
1. Skoog D. A., Holler F. J., and Crouch S. R., Principles of Instrumental Analysis, 6th
Ed., Harcourt Brace & Company, Orlando, Florida, 2007, Chap. 26 and Chap.28
2. Snyder L. R., Kirkland J. J., Introduction to Modern Liquid Chromatography, John
Wiley, Toronto, 1979, Pp. 218-225.
3. Karger B. L., Snyder L. R., Horvath C., An Introduction to Separation Science, John
Wiley, Toronto, 1973, Pp. 271-274.
4. Ibid. Pp. 55-57, 275-276.
5. Giddings J. C., Unified Separations Science, John Wiley & Sons Inc., New York,
1991, Pp. 24-30.
6. A.J.P. Martin A. J. P., Biochem. Soc. Symp. 3, 4 (1949).
7. Skoog D. A., West D. M., Holler, Analytical Chemistry: An Introduction, 5th Ed.,
Saunders College, Philadelphia, 1990, Chapter 4, Sec. 4B-3 to 22-6.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF