1070-4272/05/7809-1427C2005 Pleiades Publishing, Inc.
Russian Journal of Applied Chemistry, Vol. 78, No. 9, 2005, pp. 1427!1429. Translated from Zhurnal Prikladnoi Khimii, Vol. 78, No. 9,
2005, pp. 1451!1454.
Original Russian Text Copyright + 2005 by Khasanov, Dychko, Kuryaeva, Ryzhova, Mal’tseva.
AND ION-EXCHANGE PROCESSES
A New Procedure for Caffeine Determination
V. V. Khasanov, K. A. Dychko, T. T. Kuryaeva, G. L. Ryzhova, and E. V. Mal’tseva
Tomsk State University, Tomsk, Russia
Received August 24, 2004; in final form, February 2005
Abstract-The content of caffeine in multicomponent natural objects (tea leaves, coffee beans, and instant
coffee) and caffeine-containing beverage foods was determined by ion-exchange chromatography.
Since caffeine is an inherent component of some
plants and foodstuffs produced from them , availa-
bility of the reliable procedures of its determination is
important for various-purpose analyses (such as tech-
nological, umpire, etc.). In accordance with GOST
(State Standard) , the main method of caffeine
analysis is a photometric determination based on hy-
drolytic oxidation of caffeine into tetramethylpurpuric
acid followed by photometric registration of the ab-
sorption density at l 540 nm. The main disadvantage
of this procedure is underestimation of the caffeine
content in the samples .
The most general procedure of caffeine determina-
tion used in laboratories beyond Russia is reversed-
phase high-performance liquid chromatography (RP
HPLC). Kusch and Knupp  compared the analytical
data on the content of caffeine in beverage foods deter-
mined by high-performance thin-layer chromatography
and RP HPLC. They found that, despite some restric-
tions, each procedure provides reliable results with
small measurement errors. As for caffeine determina-
tion in vegetable extracts of tea and coffee, the prob-
lem is more complex. The content of caffeine in tea
and coffee is so high that no additional procedures of
its preconcentration (such as solid-phase extraction
often used in the analysis of caffeine in biological
liquids ) are required. With solid-phase extraction
and RP HPLC, the detection limit of the caffeine
in protein-containing samples was determined as
200 ng cm
. However, wide assortment of compounds
extracted with water from tea and coffee hinders direct
determination of caffeine by RP HPLC, which is prob-
ably due to the short retention times of the polar caffe-
ine molecules, especially in sorbents with small and
intermediate content of grafted hydrocarbon chains.
As a result, the caffeine peaks overlap with peaks of
concomitant compounds even in the case of gradient
elution. Another technique recently tested for deter-
mination is capillary electrophoresis . Though this
method provides very high separation efficiency, it is
not widely used yet.
Nishitani and Sagesaka  performed an HPLC
analysis of green tea components (catechins, phenolic
glycosides, acids, and caffeine) on a C18 Wakosil-II
5C18 HG (30150 mm) reversed-phase column (Wako
Pure Chemical Industries, Ltd., Japan) using complex
gradient profile with changes in the flow rate. The
analysis time was 40 min. Smaller amount of compo-
nents (8) can be separated with this system in 20 min.
The detection limit of the components varied from 1.4
to 3.5 ng and the concentration dependence was linear
up to charges of 1500 ng; the introduced/found ratio
In this study we used a procedure more selective to
caffeine, based on two separation mechanisms through
ion-exchange and hydrophobic interactions.
Analysis was performed on an FPLC liquid chro-
matograph (LKB Pharmacia, Sweden) equipped with
a5005 mm column packed with Mono S sorbent
(strong cation-exchanger with grafted sulfopropyl
groups) produced by the same company. This sorbent
is a hydrophilic polymer consisting of spherical par-
ticles (10 mm size) with a pore size of about 100 nm.
A ready-to-use column was supplied by the company.
The elution was performed with a gradient in
the organic modifier content. Two solutions were
used: (A) 0.025 M sodium acetate (pH 6.8) and
(B) 0.025 M sodium acetate (pH 6.8)3acetonitrile
(70 : 30 volume ratio). The column was equilibrated
for 4 min with a mixture A : B = 95 : 5; then the sam-
ple was injected, and after 4 min the gradient was
switched on, up to the composition A : B = 15 : 85
in 10 min; the flow rate was 1 ml min
(300 cm h