See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/242271024

Determination and Characterization of Caffeine in Tea, Coffee and Soft Drinks

by Solid Phase Extraction and High Performance Liquid Chromatography (SPE -
HPLC)

Article · November 2006

CITATIONS READS

28 11,134

4 authors, including:

Abdul Mumin
The University of Western Ontario
18 PUBLICATIONS   149 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Abdul Mumin on 14 October 2014.

The user has requested enhancement of the downloaded file.

 Malaysian Journal of Chemistry, 2006, Vol. 8, No. 1, 045 - 051

Determination and Characterization of Caffeine in Tea, Coffee and Soft

Drinks by Solid Phase Extraction and High Performance Liquid
Chromatography (SPE – HPLC)
Md. Abdul Mumin, Kazi Farida Akhter, Md. Zainal Abedin*, Md. Zakir Hossain.
Department of Chemical Engineering and Polymer Science,
Shah Jalal University of Science and Technology,
Sylhet-3114, Bangladesh.
*Correspondence should be addressed: abedin@sust.edu.

Abstract : Caffeine (1,3,5-trimethylxanthine), a mild addicting drug was isolated, purified and
characterized from tea (black and green) and coffee. Isolation was done by liquid-liquid extraction
using chloroform as an extracting solvent. Extraction was carried out in four steps such as leaching, dye
removal, liquid extraction and recrystallization. Crude caffeine was purified by solid phase extraction
(SPE) method. The solvent used for recrystallization were toluene and petroleum ether. The purity of
caffeine was ascertained by the determination of melting temperature. Pure caffeine obtained from
different samples was characterized by UV-Visible spectrophotometer, TLC, FT-IR and HPLC. In
HPLC, 50mM KH2PO4 (pH=2), acetonitrile and methanol (40:8:2) was used as solvent as well as
mobile phase. Amount of caffeine in various soft drinks (Cola) commercially available in Bangladesh
was also determined by HPLC method.

Key Words: • Caffeine, • Chloroform, • SPE, • HPLC, • KH2PO4, • Acetonitrile.

Received : 15.02.06; accepted : 9.11.06

Introduction
Caffeine is a naturally occurring substance found weight is 194.19g, melting point is 236 0C, point at
in the leaves, seeds or fruits of over 63 plants species which caffeine sublimes is 1780C at atmospheric
worldwide and is part of a group of compounds pressure, pH is 6.9 (1% solution), specific gravity is
known as methylxanthines. The most commonly 1.2, volatility is 0.5%, vapor pressure is 760 mm Hg
known sources of caffeine are coffee, cocoa beans, at 178 0C, solubility in water is 2.17%, vapor density
cola nuts and tea leaves. Caffeine is a 6.7 [3]
pharmacologically active substance and depending on Decaffeination is a popular term in present
the dose, can be a mild central nervous system modern world to optimize the caffeine contents in
stimulant. Caffeine does not accumulate in the body various sources. This is simply use of a solvent,
over the course of time and is normally excreted which extract caffeine. For this purpose, the currently
within several hours of consumption [1]. available solvents are chloroform, methyl chloride,
Caffeine is an alkaloid of the methylxanthine ethyl acetate, super critical carbon dioxide etc.
family. In its pure state, it is an intensely bitter white Methylene chloride is used to decaffeinate a high
powder. Its chemical formula is C8H10N4O2, its proportion of conventional teas. As a solvent,
systematic name is 1,3,5-trimethylxanthine [2] and its methylene chloride is highly effective, but also
chemical formula is shown below. potentially dangerous under certain circumstances. It
can cause faintness, dizziness, and headache if
inhaled at high concentrations. Ethyl acetate is
another compound used to extract caffeine from tea.
It removes caffeine from tealeaves effectively; it can
also extract other chemical components as well.
Studies on green tea decaffeinated with ethyl acetate
have shown the potential for up to 30% of
epigallocathechin gallate (EGCG-considered to be
the primary beneficial component in green tea) and
Structure of caffeine other beneficial antioxidant compounds to be
extracted along with the caffeine [4-6]. Ethyl acetate
Pure caffeine occurs as odorless, white, fleecy is also moderately toxic.
masses, glistening needles of powder. Its molecular
046 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

Super critical carbon dioxide (CO2) is the fancy (25ml) was added with the filtrate, boiled for 5 min
name for using highly pressurized carbon dioxide— and filtered again. The purpose of the addition of
the gas that adds bubbles to mineral water—to 10% lead ethnoate is to convert any extracted tannins
dissolve caffeine from tealeaves. The advantages of or other acids into anions. As electrically neutral
CO2 are that it does not leave a chemical residue and polar molecules, the acids tend to be soluble in both
it has a minimal effect on the flavor and beneficial water and CHCl3, complicating the purification and
compounds inherent to the tea [6]. increasing the tendency to form an emulsion in the
Though chloroform is toxic, in laboratory next step. As anions they are not soluble in CHCl3
purposes it is best caffeine extracting solvent and the and this helps to avoid an emulsion. The 10% lead
results obtained by characterization becomes ethnoate seems to cause precipitation of some
satisfactory and reproducible always [7]. substances that tend to clog the filter paper in the
In the present study, we have extracted caffeine next step and also helps avoid formation of
from green tea, black tea, coffee and then emulsions. A 500 ml separatory funnel was put into
characterized by melting point, max (UV- Visible), IR an iron ring on a ring stand. Pouring the tea solution
absorption bands, Rf (TLC) and RT (HPLC). One of in the separatory funnel and adding about 30 ml of
the major objectives is to develop an improved CHCl3, the solution was shaken uniformly while
purification method based on Solid phase extraction. stopcock was opened to expel vapors. The layers
We have also developed an easily adaptable HPLC were allowed to separate and the lower layer
method for both qualitative and quantitative (chloroform) was collected into a 100ml beaker and
determination of Caffeine. In Bangladesh, there is no the separation procedure was repeated for the second
authentic data about caffeine content in soft drinks. time to collect into the beaker. Anhydrous sodium
So, we also determined the concentration of caffeine sulfate was added in the beaker containing the
in various available soft drinks, especially cola combined extracts. The anhydrous sodium sulfate
drinks. The proposed method is simple, rapid and has would act to remove any water and water-soluble
significant advantages over spectrophotometric salts that were retained in the chloroform (organic
methods as well as other HPLC methods. layer) or accidentally transferred during decantation.
The beaker containing the extract was then heated a
Experimental short period for dryness using a water bath and the
temperature was controlled low enough at 70-900C to
Materials: avoid caffeine decomposition. After 24 hours, white
KH2PO4 (Merck, Germany); Phosphoric acid (85 crude caffeine obtained at the bottom of the beaker.
%, Merck); Petroleum ether (BDH, England); The crude caffeine obtained from the above method
Standard Caffeine (Merck, Germany); Lead Ethnoate, was purified and recrystallized by solid phase
Chloroform, Toluene, Methanol, Acetonitrile(Merck, extraction method using chloroform, toluene and
Germany). petroleum ether.
Black tea (Ispahani Mirzapur tea), Coffee In case of coffee, 1 gm coffee was taken into a
(Nescafe) and various brand soft drinks were 500 mL beaker and subsequently 250mL hot distilled
collected from local market and green tea from water was added. Coffee was properly dissolved in
Bangladesh Tea Research Institute (BTRI). the hot distilled water. Then the whole procedure was
same as that of isolation of caffeine from tea.
Apparatus:
HPLC; Model: Waters 515, with UV Detector & Solid Phase Extraction (SPE):
Vacuum filter, pH Meter (REX, Model pHS-25), Visi Commercial tea and coffee consists of many
TM
-1 SPE Single Sample Processor (Supelco) and components that cause chromatographic interferences
UV-Visible spectrophotometer (UV-1601,Shimadzu, with caffeine. For this reason the sample treatment
Japan), Melting Electrical melting point apparatus proposed consists of SPE with Sep-Pak C18 (500 mg)
(Electrothermal, IA9100), FT-IR (SHIMADZU, cartridges that enable separation of caffeine and
Model-IR Prestige-21), TLC (Whatman, 250 m remove most of the interfering components. The SPE
layer, 20X20 cm). method of Lyold et.al [8], Mottaleb et.al [9] and Cho
et al.[10] was used for the extraction of caffeine. The
Methods stationary phase was activated with 3mL 50%
methanol and pre-equilibrated with 3mL 1%
Isolation of caffeine from tea and coffee: methanol. The columns were air-dried by drawing air
5 gm solid tea was taken in 500 ml beaker and through them for 10 minutes. 0.5gm homogenized
subsequently distilled water (225 ml) was added to it. crude caffeine was dissolved with 100mL double
The mixture was boiled for 10 min and filtered by distilled water and then solution (10 mL) was loaded.
using a Buchner Funnel. 10% lead ethnoate solution The adsorbed caffeine was eluted twice with 2mL
047 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

chloroform. The pure caffeine was dried dihydrogen phosphate (pH=2), acetonitrile and
subsequently and then dissolved in another solvent methanol (40:8:2), which was also mobile phase.
such as toluene and a small amount of petroleum Standard solutions were injected in HPLC pump by
ether was added for recrystallization. In case of using a syringe after setting proper flow rate
HPLC analysis pure caffeine was dissolved in mobile (0.5mL/min), attenuation (64) and chart speed (0.5
phase. Before HPLC analysis all samples were cm/min) for which a typical pressure is about 1328-
filtered through a 0.45 m pore size FP 30/45 CA-S 1331psi. UV detector was used at a maximum
filters ( Schleicher and Schuell, Darmstadt, Germany) wavelength of 254nm and chromatograms were
at 7 bar max. Samples (20 L) of solutions of the obtained with almost same retention time. Three
samples were injected into the HPLC column. replicates of each standard were taken and a plot of
relative peak area vs. concentration made to obtain
Method of Spiking: calibration curve.
Different amount of black tea samples were
spiked with a known amount (0.5g) of analytical Determination of caffeine in soft drinks:
standards of caffeine. The standard caffeine solution Each soft drink was degassed properly by
in water (20mL) was thoroughly mixed with tea placing it in a vacuum flask and connecting the flask
solution (after leaching). To ensure homogenization, to a vacuum pump for 30 minutes. It was kept under
the spiked samples were shaken by using mechanical vacuum until no more bubbles appear. Then each
shaker for 30min. Then the solvent extraction and sample was filtered through a 0.45µm syringe filter
SPE processes were carried out as discussed before. with a 5mL syringe. 2 mL of filtered drink sample
The whole process was also carried out for only was 10 times diluted by using mobile phase (solvent).
standard caffeine (0.5g) for the method validation. 20 µL of each diluted sample was injected into the
Unspiked tea samples were also treated similarly. column and recorded the trace. The relative peak
areas were determined for three replicates of each
Characterization of pure Caffeine: dilute sample. Then the concentration of each dilute
Different physical methods were employed to sample and finally the real concentration of caffeine
characterize the crystalline caffeine. in soft drinks samples were calculated from
calibration curve. [11]
Determination of melting point:
The melting point of different extracted pure Determination of caffeine in Green Tea, Black Tea
samples (after SPE purification) was carried out in a and Coffee:
digital melting point apparatus (Electrothermal, 1.0 mg of isolated purified caffeine was
IA9100). The average melting points of the samples dissolved in 100mL mobile phase (solvent). Then the
was 2350C. sample was filtered and determined by HPLC as
previous section.
Thin Layer Chromatography: (Whatman, 250 m
layer, 20X20 cm) Reproducibility:
By the purified product (crystalline caffeine) a The reproducibility of this method was also
plate was developed using chloroform as mobile checked by determining the percentage recovery of
phase and visualized under UV- lamp. The Rf value known amount of standard caffeine in the sample.
was measured and it was found 0.63. For example if X be the actual content of caffeine in
IR Spectroscopy: the sample and 5ppm standard caffeine was added
The IR –spectrum of extracted purified caffeine then the content become (X+5)ppm and if the
was carried out by using a FT-IR spectroscopy observed concentration is X’, then the percentage
(SHIMADZU, Model-IR Prestige-21). recovery is given by
X′
% Recovery = × 100 %
UV Spectrometry:
An UV- absorption spectrum of extracted
(X + 5)
purified crystalline caffeine was prepared at different
Results & Discussion
absorbance against different wavelength using a UV-
For preparing calibration curve in HPLC 10 to
absorption spectrophotometer from SHIMADZU
60 ppm caffeine standards were used to identify peak
Corporation, Japan; Model: UV -1601. The max was
at retention time around 6.18-6.22 min. The peak area
found to be 275 nm.
increases from the lowest standard to the highest. The
HPLC Optimization (calibration): slope of the curve was 56659, which was used for
The different known concentration of caffeine in determination of caffeine in tea, coffee and soft
a solvent, which consisted of 50 mM potassium drinks.
048 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

40

35

30

25

20
15

10

0
0 20 40 60 80

Figure 1: Calibration curve for caffeine in 50 mM KH2PO4 (pH=2), acetonitrile and methanol
(40:8:2)

Table 1: Isolation and Purification of Caffeine from Tea and Coffee. (n=3)

Sample Amount of Amount of % of crude Amount of % of Standard

Sample (g) Caffeine Caffeine Caffeine Caffeine Deviation
after solvent before SPE after SPE after SPE (%)
extraction (g) Purification (g)
Black Tea 5.0 0.352 7.04 0.167 3.34 0.88
Green Tea 5.0 0.244 4.88 0.122 2.44 0.79
Coffee 1.0 0.137 13.7 0.052 5.20 0.83

Table2: Percentage recovery of caffeine from black tea spiked with standard caffeine.

Amount of Amount of
Amount of
No. of standard Caffeine Spiked Result
Black tea
observations Caffeine added obtained after (%)
(g)
(g) SPE (g)
1 5 0.612 97
2 5 0.610 96
0.5
3 4 0.578 96
4 3 0.545 95
049 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

Table 3 : Characteristics of purified Caffeine isolated from Tea and Coffee.

Sample Melting point max Retention Time Rf

O
( C) (min)
Black Tea 235 270 6.19 0.63
Green Tea 236 270 6.20 0.63
Coffee 235 269 6.20 0.64

Table 4 : Characteristic IR-absorption bands of samples and standard.

Bonds (cm-1) (cm-1) (cm-1) (cm-1) (Coffee)

(standard) (Black Tea) (Green Tea)
C=H 2860 2855 2853 2853
C-H 3100 3105 3105 3108
C=C 1645 1648 1652 1651
C=O 1690 1692 1696 1696
C-N 1240 1238 1237 1238
C=N 1590 1597 1595 1596

Table 5 : Caffeine determination in green tea, black tea and coffee. (by HPLC) (n=3)

Sample Avg. RT (min) Avg. Peak Area Concentration of Standard

(a.u.) Caffeine (mg/L) Deviation (%)
Green Tea 6.20 559791 9.88 1.03
Black Tea 6.19 555258 9.80 0.98
Coffee 6.20 562723 9.93 0.97

Table 6 : Caffeine determination in various soft drinks (Cola-drinks) (by HPLC) (n=3)

Sample Avg. RT Avg. Peak Area Concentration of Standard

(min) (a.u.) Caffeine (mg/L) Deviation (%)
Coca cola 6.22 739402 130.5 0.93
Double cola 6.22 920712 162.5 0.88
Uro cola 6.18 708240 125 0.89
RC cola 6.20 753567 133 0.97
Suncrest cola 6.20 617585 109 0.85
Pepsi cola 6.22 572258 101 0.87
050 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

Extraction of caffeine from tea and coffee was obtained caffeine showed similar Rf value. This
achieved by using chloroform as an extracting confirms the purity of caffeine. The IR-spectrum of
solvent. It was observed that the extraction efficiency isolated caffeine showed similar absorption bands
of caffeine from various sources by using chloroform when compared with that of literature [7]. The IR-
was much higher than other solvents. Table 1 shows spectrum of Table 4 indicates the absolute purity of
extraction efficiency of crude caffeine from black tea, the purified caffeine.
green tea and coffee. The amount of crude caffeine We have developed a HPLC method for the
was found to be 7.04% from black tea, 4.88% from determination of caffeine, which was carried by High
green tea and 13.7% from coffee (Table-1). We Performance Liquid Chromatography (HPLC)
observed that black tea contained a higher percentage instead of using UV-Visible spectrophotometer. We
of crude caffeine than that of green tea and coffee chose HPLC method for the determination of
contained a higher percentage of crude caffeine than caffeine, because HPLC is the most widely used
that of both black and green tea. To purify the crude qualitative and quantitative determination and
caffeine, SPE purification method was used. The separation method. The method is popular because it
amount of pure caffeine after SPE was found to be is non-destructive and may be applied to thermally
3.34% from black tea, 2.24% from green tea and labile compounds (unlike GC); it is also a very
5.20% from coffee (Table-1) After purification we sensitive technique since it incorporates a wide
have seen that coffee also contained a higher choice of detection methods. With the use of post-
percentage of pure caffeine than that of both black column derivatization methods to improve selectivity
and green tea. From the result of previous works it is and detection limits, HPLC can easily be extended to
evident that the SPE purification method resulted in a trace determination of compounds that do not usually
percentage of yields with greater purity. provide adequate detector response. The wide
5 g of black tea samples after solvent extraction applicability of HPLC as a separation method makes
and SPE purification was found to contain an average it a valuable separation tool in many scientific fields.
value of 0.167g of caffeine that corresponds to about By using HPLC method, we determined the
3.34% which is in good agreement with the literature retention time and the relative peak area of extracted
value of 3.32 to 3.81[1,15]. However, the caffeine purified caffeine. The retention time of the purified
content of various sources varies with soil conditions caffeine and that of the standard caffeine were almost
and climate. similar, which confirms the identity of caffeine. From
From the method of validation in case of only relative peak area by using calibration curve the
0.5 g standard caffeine the amount of caffeine after amount of extracted purified caffeine for different
solvent extraction and SPE was found as 0.45 g. It samples were also determined (Table 5).
has been found also that the percent recovery of Determination of the amount of caffeine was also
caffeine (black tea) spiked with standard caffeine was carried out in several soft drinks that are available in
95 to 97% (Table-2). Bangladesh. The commercial names of the analyzed
The pure white crystalline caffeine isolated from soft drinks were Coca Cola, Double Cola, Uro Cola,
sources was found to melt at 2350C. The average RC cola, Suncrest Cola and Pepsi cola. The caffeine
melting points of extracted crude caffeine of different present in these soft drinks amounted to 130.5,
samples were about 230OC. It confirms the presence 162.5,125,133,109 and 101 ppm respectively (Table
of impurities. Purification performed by SPE 6). The caffeine contents of the soft drinks marketed
procedure gave the average melting point of about in Bangladesh were found to be similar with that of
235OC. It can be concluded that the extracted the literature value [1].
product after purification was pure. This confirms
that SPE purification was an essential step to obtain Conclusion
pure caffeine. A method has been developed for the extraction,
An UV-absorption spectrum of extracted purified purification of caffeine from tea, coffee and some
caffeine was prepared by using UV-Visible soft drinks. Caffeine from the tea and coffee was
spectrophotometer. The max was found for various extracted by liquid-liquid extraction and interferences
sources to be 271nm which is similar to that found in were removed by employing solid-phase extraction
literature [16,17]. (SPE).
By the purified product (crystalline caffeine) a The purified caffeine was then analyzed by
Thin Layer Chromatography was carried out on HPLC. The soft drinks were directly analyzed by
plates (Stationary phase: Silica gel coated on HPLC for their caffeine content. Solid phase
Aluminium) and the chromatographic process was extraction prior to HPLC was found to be an essential
developed using chloroform as mobile phase and step for determining the amount of caffeine by
visualized under UV-lamp. The Rf value was HPLC. Characterization of caffeine was achieved by
measured and it was found that commercially determining melting temperature, Rf value, IR
 051 M Z Abedin, et. al Determination and Characterization of Caffeine in
Tea, Coffee and Soft Drinks by SPE – HPLC

spectrum and UV spectrophotometry. 9. Mottaleb, M.A., Abedin, M.Z., Islam, M.S.

(2004) Determination of volatile organic
References compounds in river water by SPE and GC, J.
1. Barone, J.J., Roberts, H.R. ( 1996) “Caffeine Envi. Sci., C16, 497
Consumption”, Food Chemistry and Toxicology, 10. Cho, C.M., Ko, J.H., Cheong, W.J. (2000)
McGraw-Hill, Newyork, 34, 119 Talanta, 51, 799
2. Arnaud, M. J. (1987) The Pharmacology of 11. Yushas, B. (2002) Determination of Caffeine in
Caffeine, Prog Drug, 31, 273. Beverages with HPLC, Chem., 384
3. Clementz, G. L., Dailey, J. W. Psychotropic 12. Arnaud, D.J.F., Watson, I.D., Williums, J. and
effects of caffeine, Amer. Fan Physician. 37 167 Berry, D.J. (1988) The Clinical use and
(1988) measurement of theophylline. Ann Clin Biochem.
4. Kirmer, D.A. (1988) Caffeine use and abuse in 25, 4
psychiatric clients. J. Psychosoc Nurs Ment 13. Tondor, L., Rudas, N. (1991) The course of a
Health Serv., 26, 20 seasonal bipolar disorder influenced by caffeine.
5. Johnson, G.D., Fatis, M., Sonnek, D. (1988) A J Affect Disor., 22, 249
survey of caffeine use and associated side effects 14. Lee, M.A., Flegel, P., Greden, J. F. (1988)
in a college population. J Drug Educ., 18(3), 211 Anxiogenic effects of caffeine on panic and
Jefferson, J.W. (1998) Lithium tremor and depressed patients. American Journ Psychiatry.
caffeine intake: two cases of drinking less and 145, 632
shaking more, J.Clin Psychiatry., 49, 72 15. Baruah, S., Hazarika, M., Mahanta, P.K., Horita,
Islam, M. S., Rahman, M. M., Abedin, M. Z. H. and Murai, T. (1986) Effect of plucking
(2002) Isolation of caffeine from commercially intervals on the chemical constituents of CTC
available tea and tea waste, Jahangirnagar Uni. black teas, Agri. Biol. Chem., 50, 1039
J. Sci., 25, 9. 16. Aeschbacher, H.V. et al. (1975) The Effect of
8. Lyold, L.L., Warner, F.P., Kennedy J.F.and Caffeine on Barbiturate Sleeping Time and Brain
White, C.A. (1988) Journal Chromatography, Level. J. Pharmacol. Exp. Ther, 192(3), 635
437, 447 17. Christian, G. D. (1994) Analytical Chemistry,
John Wiley, 5th Edi 505.

View publication stats