Determination of Vitamin E in Coffee Beans by HPLC

Using a Micro-extraction Method

R.C. Alves,* S. Casal and M.B.P.P. Oliveira
REQUIMTE/Serviço de Bromatologia, Faculdade de Farmácia, Universidade do Porto
Rua Anı´bal Cunha 164, 4099-030 Porto, Portugal

This work reports a solid–liquid micro-extraction method for vitamin E quantification in coffee beans
(before and after roasting) with normal-phase HPLC/diode-array/fluorescence detection. The compounds
were extracted after protein precipitation and overnight maceration (4 8C) in n-hexane, in the presence of
butylated hydroxytoluene, using tocol as internal standard. The extraction method precision was inferior
to 5% with mean recoveries near 100%. Chromatographic resolution from co-eluting interferences was
achieved within 8 min with a 75  3.0 mm (3 mm) silica column by using an isocratic elution with n-hexane/
1,4-dioxane (98 : 2), at a flow rate of 0.7 mL/min. The diode-array detector was a valuable tool in the
detection of co-eluting interferences and quantification was based on the fluorescence measurements. Only
two vitamers, a- and b-tocopherol, were confirmed and quantified, being the latter generally the major
compound in both arabica and robusta coffees. The present analytical method proved to be simple,
sensitive, reproducible, accurate, allowing a fast quantification with low organic solvent consumption.

Key Words: coffee beans, vitamin E, tocopherols, HPLC

INTRODUCTION age-related cataract, Parkinson’s disease, atherosclero-

sis, low-density lipoproteins oxidation, coronary heart
Vitamin E is the common name given to a group of disease, and some immunologic diseases (Bramley et al.,
eight naturally occurring and structurally related 2000; Gama et al., 2000; Nelis et al., 2000).
vitamers. Chemically, they contain a chromanol ring, Only plants have the ability to biosynthesize
with a distinct substitution pattern of methyl groups vitamin E. As lipophilic compounds, they appear in
(a-, b-, -, and -), and a 16-carbon phytyl side chain, high concentration in edible vegetable oils being their
saturated or unsaturated in tocopherols and toco- distribution pattern related with the oil type (Gama
trienols, respectively (Schneider, 2005). et al., 2000; Hammond, 2003). Other food sources of
Depending on their structure and stereochemistry, the vitamin E are described in literature, namely, hazelnuts
eight forms show different biological potencies, being (Amaral et al., 2005), cereals (Panfili et al., 2003) green
RRR-a-tocopherol the most active one, with 100% vegetables, fruits (Woollard and Indyk, 2003) and coffee
vitamin E activity (Eitenmiller and Landen, 1999). Their (Speer and Kölling-Speer, 2006).
biological effects are mostly derived from their anti- Despite being vitamin E deficiency a rare event, its
oxidant properties. While protecting polyunsaturated presence in coffee is of importance, especially when this
fatty acids from peroxidation, they contribute to cellular beverage is reported as the single greatest contributor to
membranes integrity (Nelis et al., 2000). In humans, the total antioxidant intake (Svilaas et al., 2004).
vitamin E acts as an integral part of the primary The presence of tocopherols in the unsaponifiable
intracellular defense system and it has been correlated matter of green coffee beans oil was described for the
with the prevention of several diseases: cancer, first time by Folstar et al. (1977), using thin-layer
chromatography and gas chromatography. Since then,
some studies about vitamin E in coffee have been
published, mainly using HPLC with fluorescence detec-
tion (Aoyama et al., 1988; González et al., 2001;
*To whom correspondence should be sent Kölling-Speer et al., 2006; Jham et al., 2007). As in
(e-mail: rita.c.alves@gmail.com).
other food matrices, normal-phase chromatographic
Received 16 August 2007; revised 8 January 2008.
separation is advantageous allowing a complete separa-
Food Sci Tech Int 2009;15(1):0057–63 tion of all the eight vitamin E compounds, especially
ß SAGE Publications 2009 b- and -tocopherol (Gama et al., 2000; Kamal-Eldin
Los Angeles, London, New Delhi and Singapore
ISSN: 1082-0132 et al., 2000; Rupérez et al., 2001), together with a higher
DOI: 10.1177/1082013208102695 solubility for lipidic compounds when operating with
Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015 57
58 R.C. ALVES ET AL.

organic solvents (Kamal-Eldin et al., 2000). In order to (Madrid, Spain). Absolute ethanol was from Panreac,
take advantage of their UV characteristics and native Spain. Butylated hydroxytoluene (BHT) was used as
fluorescence, the detection of these compounds is mainly antioxidant and was obtained from Aldrich (Madrid,
performed by UV and/or fluorescence (Nelis et al., Spain). A 10 mg/mL working solution of BHT was
2000). prepared in n-hexane. All other reagents were of
Aoyama et al. (1988) analyzed the tocopherols analytical grade.
amount in dry and roasted coffee beans showing that Arabica and robusta coffee beans, both green and
a higher content of total tocopherols could be found in after a standard industrial roasting procedure (210 8C,
the hexane soluble components of an ethanol extract. 10 min), were obtained from a local coffee roaster.
More recent studies reported the extraction step using
oil saponification (Kölling-Speer et al., 2006) or Soxhlet Methods
extraction (González et al., 2001; Jham et al., 2007).
Nevertheless, some vitamin E loss may occur during Sample Preparation
these procedures due to the compounds sensibility to
alkaline pH and heat (Eitenmiller and Landen, 1999). All samples were mechanically powdered to pass
Generally, the results confirm the presence of two through a 0.75 mm sieve. Sample moisture was deter-
main tocopherols, a- and b-, with -tocopherol only in mined by drying at 103  2 8C until constant weight.
trace amounts. The reported amounts are quite variable,
a fact that might be attributed to their natural biological Vitamin E Extraction
variation in the coffee beans, but also to the different
methodologies applied. Recently, Jham et al. (2007) The antioxidant (BHT solution, 75 mL), internal
quantified also -tocopherol in roasted coffee. standard (190 mg/mL, 50 mL) and absolute ethanol
The simultaneous use of diode-array and fluorescence (1 mL) were added to a 150 mg amount, exactly weighed,
detection has already been successfully applied to other of each ground coffee sample, and mechanically
food matrices (Casal et al., 2001; Amaral et al., 2005) homogenized for 30 min in a orbital vortex mixer
revealing to be essential, by allowing identification of (VV3, VWR Int.) with a multiple sample support.
co-eluting interferences (based in their UV spectra). Afterwards, 2 mL of n-hexane were added and the
Mass spectrometry, when available, is the most advan- mixture was mixed again for more 30 min and left
tageous tool in this field (Bustamante-Rangel et al., overnight at 4 8C.
2007; Schwartza et al., 2008). After a third mixing period, 1 mL of NaCl 1% (m/v)
The aim of this work was to develop a simple and was added. The mixture was briefly vortexed and
nonaggressive micro-method to extract vitamin E from centrifuged (2 min, 5000 rpm). The organic phase was
coffee, raw and roasted, with the minimum consumption separated and the residue reextracted twice with 2 mL of
of organic solvents. Besides, the chromatographic n-hexane. The organic phases were combined and a
conditions (HPLC) were optimized in order to obtain sufficient amount of anhydrous Na2SO4 was added to
a faster, simple, sensitive, and selective way to identify the total extract. The mixture was vortexed and
and quantify these compounds using both diode-array centrifuged in order to collect the n-hexane layer. The
and fluorescence detection systems. extract was taken to dryness under a nitrogen stream,
at ambient temperature, re-suspended with 500 mL of
n-hexane and injected into the HPLC system.
MATERIALS AND METHODS
Analytical Determination
Tocopherols (a, b, , and ) and tocotrienols (a, b, ,
and ) were purchased from Calbiochem (La Jolla, CA). The chromatographic analysis was carried out in a
Standard solutions were prepared in n-hexane and HPLC integrated system equipped with an AS-950
kept at 20 8C. Their concentrations were periodically automated injector, a PU-980 pump, an MD-910
evaluated by UV, according with their molar absortiv- multiwavelength diode array detector (DAD) and an
ities in ethanol (Eitenmiller and Landen, 1999), and FP-920 fluorescence detector (Jasco, Japan).
further dilutions performed as required for calibration The chromatographic separation of the compounds
or other purposes. was achieved on a normal phase SupelcosilTM LC-SI
The internal standard tocol (2-methyl-2-(4,8,12- (3 mm) 75  3.0 mm (Supelco, Bellefonte, PA), operating
trimethyltridecyl)chroman-6-ol) was obtained from at constant room temperature (21 8C). A mixture of
Matreya Inc., PA, USA. A 10 mg/mL solution was n-hexane and 1, 4-dioxane (98 : 2) was used as eluent, at
prepared in n-hexane, kept at 20 8C, and diluted to 0.7 mL/min. The detection was performed by both the
working solutions as necessary. DAD connected in series with the fluorescence detector,
HPLC grade n-hexane was obtained from Merck programmed for excitation at 290 and emission at
(Darmstad, Germany) and 1,4-dioxane from Fluka 330 nm (gain 10).
Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015
 Vitamin E in Coffee Beans 59

The compounds under study were identified by coffee, and the higher standard deviations represent an
chromatographic comparisons with authentic standards, ineffective extraction, probably due to the reduced
by co-elution and by their UV spectral characteristics. solvent contact time. Similar a-tocopherol amounts
Quantification was performed on the basis of the were found through both methods C and A. However,
internal standard method using the fluorescence signals. it seems that a slight loss of b-tocopherol, the main
Data were analyzed using a Borwin-PDA Controller compound, occurred during saponification. Besides,
Software (JMBS, France). saponification did not allow the use of
tocol as internal standard due to its degradation
Statistical Analysis in alkaline conditions and conversion to co-eluting
interferences. For that reason it was only added after the
Data are reported as mean  standard deviation, neutralization step.
always in triplicate. Data were analyzed by the one-
way ANOVA. All analyses were carried out with Optimization of Chromatographic Conditions
Microsoft Excel statistical software (Microsoft Office
Excel 2003, Microsoft Corp., Redmond, WA). Several normal-phase HPLC conditions are described
in literature for different food matrices, usually achieving
a good peak separation of the eight compounds of
RESULTS AND DISCUSSION vitamin E between 15 and 30 min (Gama et al., 2000;
Kamal-Eldin et al., 2000; Panfili et al., 2003;
Choice of the Extracting Method Hewavitharana et al., 2004; Amaral et al., 2005).
Beyond time consuming it also represents an elevated
In order to choose the best conditions to extract waste of organic solvents, environmentally harmful.
vitamin E from coffee beans, with a minimum con- Thus, chromatographic conditions allowing a good
sumption of organic solvents, three methods were tested peak separation in less time will be, certainly, beneficial
(Table 1). An arabica coffee, raw and roasted, was used in all points of view.
for these extraction studies. All operations were per- In order to optimize the chromatographic conditions
formed in the absence of direct sunlight, using amber in this work, two normal-phase columns were tested and
glassware. compared for vitamin E separation.
As described in Table 2, method B gave the better A Inertsil 5 SI normal phase column (5 mm, 250  3 mm)
results, allowing a higher extraction of tocopherols and from Varian (Middelburg, The Netherlands) was
smaller standard deviations. The amounts obtained with initially used, with a mobile phase of n-hexane and
method A were the lowest, for both green and roasted 1,4-dioxane (95.5 : 4.5), and a flow rate of 0.7 mL/min

Table 1. Description of the three extraction methods tested.

Method Brief description References
A BHT, tocol, and ethanol; Casal et al. (2001)
Extraction with n-hexane (3, 2 mL, vortex: 2 min); Amaral et al. (2005)
Organic phases combined, taken to dryness (nitrogen stream) and reconstituted for
HPLC analysis.
B Similar to method A, but an overnight extraction step was added, and the vortex Proposed method
mixing times increased to 30 min.
(Detailed in section Materials and Methods)
C Ethanol and pyrogallol ISO 9936 : 2006
Saponification
Extraction with diethyl ether, cleaning with water and HCl 0.01M
Addition of anhydrous sodium sulphate and tocol; taken to dryness (nitrogen stream)
and reconstituted for HPLC analysis.

Table 2. Tocopherol contents achieved with the three extractive methods tested.
Raw coffee Roasted coffee
a-Tocopherol b-Tocopherol a-Tocopherol b-Tocopherol
Extraction method mg/100 g (Mean  SD, n ¼ 3)
A 2.32  0.07 6.60  0.17 2.29  0.13 6.03  0.57
B 2.45  0.02 7.00  0.01 2.45  0.07 6.44  0.14
C 2.43  0.05 6.35  0.01 2.41  0.20 5.73  0.18

Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015
60 R.C. ALVES ET AL.

TM
(Amaral et al., 2005). A Supelcosil LC-SI (3 mm, mobile phase (A) and SupelcosilTM LC-SI column using
75  3 mm), from Supelco (Bellefonte, PA), was also n-hexane/1,4-dioxane (98 : 2) (B), both at 0.7 mL/min.
tested under the same chromatographic conditions. It was In order to evaluate the columns performance in what
possible to observe that the last column was able to concerns to the coffee extracts in study, several samples
separate the standards in58 min, compared to 17 min in of raw and roasted coffees were prepared and injected in
the first, but the peaks resolution was not as good as in the both columns. An example for an arabica sample can be
former. Therefore, in order to take advantage of the observed in Figures 2 (raw) and 3 (roasted). Both
shorter run time and reduced solvent consumption with columns allowed a good separation of a- tocopherol but
Supelcosil column, different eluent mixtures were tested, in the first column b-tocopherol co-eluted with an
always in isocratic mode, in order to get a better interfering compound formed during roasting
separation of some critical vitamers: n-hexane/1,4-diox- (Figure 3(a)). With the Supelcosil column it was possible
ane (97 : 3, 98 : 2, and 98.5 : 1.5) and n-hexane/1,4- to obtain not only cleaner chromatograms in a shorter
dioxane/isopropanol (97.9 : 2 : 0.1 and 97.7 : 2 : 0.3). The time, but also to avoid this co-elution. Since tocopherols
best separation was achieved with 2% of 1,4-dioxane in and tocotrienols show maximum absorption between
n-hexane. Different flow rates (from 0.4 to 0.8 mL/min) 280 and 300 nm (Eitenmiller and Landen, 1999), by
were also tested and the best compromise was obtained analyzing the co-eluting compound UV spectrum, it was
with a flow rate of 0.7 mL/min. possible to conclude that it did not belong to the vitamin
In Figure 1 it is possible to observe the fluorescence E family. Besides, its position changed when the
chromatograms of a standard mixture using the Inertsil chromatographic conditions were modified, while toco-
5 SI column with n-hexane/1,4-dioxane (95.5 : 4.5) as pherols kept their elution order.
Some authors (Folstar et al., 1977; Cros et al., 1985;
Aoyama et al., 1988; Speer et al., 2006) also mentioned
(a)
the presence of small amounts of -tocopherol in
µV coffee. However, in our studied samples, only a- and b-
1 34
tocopherol were confirmed by the DAD. Despite having
3.0E+05 5

2
9 (a)
7 µV 2
2.0E+05 8
4.0E+05

6 3.0E+05
1.0E+05

3
2.0E+05 1
5.00 10.00 15.00 20.00 (min)
(b) 1 1.0E+05
µV 34
5
3.0E+05
2 5.00 10.00 15.00 (min)
9
7 (b) µV 2
8
2.0E+05 4.0E+05

6
3.0E+05
1.0E+05
3
2.0E+05 1
2.00 4.00 6.00 8.00 (min)
Figure 1. HPLC fluorescence chromatograms of a 1.0E+05
standard mixture (approximately concentration for
each vitamin E compound: 2.8 mg/mL) with Inertsil
5 SI column (a) and with SupelcosilTM LC-SI (b); (1) 2.00 4.00 6.00 8.00 (min)
a-tocopherol, (2) a-tocotrienol, (3) b-tocopherol, (4) Figure 2. HPLC fluorescence chromatograms of a
-tocopherol, (5) b-tocotrienol, (6) -tocotrienol, (7) raw arabica coffee with Inertsil 5 SI column (a) and
-tocopherol, (8) internal standard (tocol), (9) with SupelcosilTM LC-SI (b); (1) a-tocopherol, (2)
- tocotrienol. b-tocopherol, (3) internal standard (tocol).
Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015
 Vitamin E in Coffee Beans 61

small chromatographic peaks with retention times were calculated based on the average ratio between the
corresponding to other vitamin E compounds, its UV peak areas (arbitrary units) of each compound and the
spectra did not confirm its identity. Only González et al. internal standard versus the standard concentration
(2001) described a significant increase of -tocopherol (mg/mL). The relative average deviations of triplicates
after the roasting procedure, probably a misidentifica- were always minor than 2%. The correlation factors (r)
tion, avoidable by using a DAD or mass detector. were higher than 0.999 for all vitamers. These param-
eters, together with the limits of detection (LOD) and
Method Validation quantification (LOQ), are detailed in Table 3. The
detection and quantification limits were calculated as 3.3
The linearity range was tested with diluted standard and 10 timed the standard deviation of the response/the
solutions subjected to the entire extraction method, as slope of the calibration curve, respectively.
described in experimental. The regression equations The precision of the extraction method was deter-
mined by repeatability (intra-day) and intermediate
(a) precision (inter-day). Intra-day precision was evaluated
µV
by assaying a raw (robusta) and a roasted (arabica)
8.0E+05 coffee extracted six times during the same day. The
mean coefficients of variation were 2.19 and 2.30%
6.0E+05 for a- and b-tocopherol, respectively. Analyzing the
same samples in six different days (inter-day precision),
2 the mean coefficients of variation were 4.94% for
4.0E+05
a-tocopherol and 5.05% for b-tocopherol.
3 In the absence of a reference matrix, the method
2.0E+05 1
accuracy was evaluated, for both raw (robusta) and
roasted (arabica) coffees, by the standard addition
2.00 10.00 15.00 (min) procedure (percentage of recovery). The standards (a-,
b-, and -tocopherol) were added to the samples in three
(b) µV concentration levels before the extraction procedure.
The results are listed in Tables 4 (raw robusta) and 5
2 (roasted arabica). The average recovery values for a-, b-,
4.0E+05
and -tocopherol were, respectively, 100  4%, 100  1%,
and 103  4% in the raw sample and 100  3%,
3.0E+05
104  7%, and 105  7% in the roasted one.
3
2.0E+05 1
Determination of Vitamin E in Coffee Samples
1.0E+05 In order to test the developed method 24 coffee beans
samples were analyzed (6 arabica and 6 robusta, each
one raw and roasted). The results can be observed in
2.00 4.00 6.00 8.00 (min)
Table 6. The extraction method was performed
Figure 3. HPLC fluorescence chromatograms of the in triplicate and the final extract injected three times
same arabica coffee after roast, with Inertsil 5 SI into the HPLC system.
column (a) and with SupelcosilTM LC-SI (b); (1) The average amounts found in arabica raw coffees
a-tocopherol, (2) b-tocopherol, (3) internal standard were 2.58  0.45 and 7.55  1.09 mg/100 g d.w., for a-
(tocol). and b-tocopherol, respectively, with a total average

Table 3. Analytical validation parameters.

Compound Retention time (min) Correlation factor (r) Linearity range (mg/mL) LOD (mg/100 g) LOQ (mg/100 g)
a-Tocopherol 1.91 0.9998 0.09–26.5 0.006 0.018
b-Tocopherol 3.11 0.9998 0.10–27.0 0.005 0.015
-Tocopherol 3.43 0.9999 0.10–27.0 0.004 0.012
-Tocopherol 5.09 0.9999 0.08–25.0 0.016 0.049
a-Tocotrienol 2.30 0.9998 0.09–26.0 0.005 0.015
b-Tocotrienol 3.83 0.9999 0.10–27.0 0.008 0.025
-Tocotrienol 4.32 0.9999 0.10–26.5 0.006 0.019
-Tocotrienol 6.83 0.9999 0.11–28.0 0.006 0.018
Tocol (IS) 5.99 – – – –

Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015
62 R.C. ALVES ET AL.

Table 4. Recovery of a-, b-, and -tocopherols from a spiked sample of a raw robusta coffee.
Compound Present (mg/100 g) Added (mg/100 g) Found (mg/100 g) Recovery (%) CV
a-Tocopherol 2.23  003 0.17 2.40  0.02 101.1 1.50
0.33 2.57  0.00 103.1 0.10
0.70 2.90  0.07 95.7 0.40
b-Tocopherol 2.29  0.02 0.18 2.47  0.05 100.5 2.90
0.35 2.64  0.13 98.8 3.60
0.71 3.00  0.08 100.0 1.10
-Tocopherol – 0.19 0.19  0.02 107.7 0.10
0.36 0.36  0.10 100.7 2.90
0.72 0.72  0.04 99.4 0.50

Table 5. Recovery of a-, b-, and -tocopherols from a spiked sample of a roasted arabica coffee.
Compound Present (mg/100 g) Added (mg/100 g) Found (mg/100 g) Recovery (%) CV
a-Tocopherol 3.82  0.07 0.34 4.16  0.02 99.7 2.2
0.70 4.54  0.00 103.5 0.4
1.22 5.02  0.07 98.1 1.3
b-Tocopherol 9.03  0.11 0.35 9.39  0.05 105.2 2.4
0.71 9.82  0.13 110.1 0.8
1.25 10.24  0.08 97.2 0.1
-Tocopherol – 0.35 0.39  0.02 112.6 2.3
0.72 0.76  0.10 105.1 0.3
1.26 1.25  0.04 98.6 0.6

Table 6. Vitamin E in coffee beans (dry basis).

Raw1 (mg/100 g) Roasted1 (mg/100 g)
Coffee sample a-Tocopherol b-Tocopherol Total a-Tocopherol b-Tocopherol Total
Arabica
Hawaii 2.18  0.02 7.01  0.02 9.20 2.15  0.03 5.78  0.00 7.93
Brazil 3.33  0.01 9.28  0.12 12.61 2.89  0.04 6.90  0.21 9.79
Costa Rica 2.66  0.01 7.77  0.05 10.43 2.51  0.07 6.65  0.16 9.17
Jamaica 2.45  0.02 7.00  0.01 9.45 2.43  0.03 6.54  0.01 8.97
Honduras 2.77  0.07 8.13  0.23 10.90 2.78  0.06 7.83  0.26 10.60
Ethiopia 2.11  0.08 6.13  0.35 8.24 2.11  0.11 5.66  0.06 7.77
Mean 2.58  0.45 a 7.55  1.09 b 10.14 2.48  0.32 a 6.56  0.79 b 9.04
Robusta
Ivory Coast 2.10  0.02 2.17  0.04 4.27 1.92  0.17 1.85  0.11 3.77
Vietnam 1.56  0.02 1.90  0.02 3.45 1.38  0.02 1.24  0.01 2.62
Uganda 1.51  0.17 1.99  0.12 3.50 1.32  0.13 1.44  0.13 2.76
India 1.66  0.03 1.86  0.01 3.51 1.65  0.06 1.68  0.09 3.32
Cameroon 1.70  0.02 2.05  0.04 3.75 1.71  0.00 1.63  0.04 3.34
Indonesia 1.93  0.01 2.49  0.03 4.43 1.85  0.25 1.84  0.06 3.69
Mean 1.74  0.23 c 2.08  0.23 d 3.82 1.64  0.24 c 1.61  0.24 c 3.25
1
Mean  SD, n ¼ 3. Data followed by different letters within each column and line are significantly different according to ANOVA at p50.05.

(a þ b) of 10.14 mg/100 g. In roasted arabica coffees, with 3.82 mg/100 g of total tocopherols. Roasted robustas
2.48  0.32 and 6.56  0.79 mg/100 g were found for coffee showed a- and b-tocopherol mean amounts of
a- and b-tocopherol, respectively, with a total average 1.64  0.24 and 1.61  0.24 mg/100 g, correspondingly,
vitamin E amount of 9.04 mg/100 g. Despite the medium with a total average of 3.25 mg/100 g. During roast,
roast loss of 11%, no significant differences ( p50.05) robusta coffees lost around 15% of total tocopherols.
were found for tocopherols amount, before and after the Contrarily to arabica coffees, significant differences
roast. ( p50.05) were found between raw and roasted robustas.
In what concerns to robusta raw coffees, average Our average results are in accordance with those
amounts of 1.74  0.23 and 2.08  0.23 mg/100 g published by Kölling-Speer et al. (2006), Ogawa et al.
were found for a- and b-tocopherol, respectively, (1989), and Aoyama et al. (1988), with a higher
Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015
 Vitamin E in Coffee Beans 63

prevalence of b-tocopherol found, especially in arabica Folstar P., Van der Plas H.C., Pilnik W. and De Heus J.G. (1977).
Tocopherols in the unsaponifiable matter of coffee bean oil.
coffees. However, these results were different from those Journal of Agricultural and Food Chemistry 25: 283–285.
of González et al. (2001) that reports high -tocopherol Gama P., Casal S. and Oliveira B. (2000). Development of
amounts in roasted coffee. The presence of - toco- an HPLC/diode-array/fluorimetric detector method for
pherol, recently published by Jham et al. (2007) was not monitoring tocopherols and tocotrienols in edible oils.
Journal of Liquid Chromatography & Related Technologies 23:
confirmed. 3011–3022.
Based on the preliminary results achieved, with González A.G., Pablos F., Martı́n M.J., Léon-Camacho M. and
apparent differences between arabica and robusta Valdenebro M.S. (2001). HPLC analysis of tocopherols and
triglycerides in coffee and their use as authentication parameters.
coffees ( p50.05), it will be interesting to study a Food Chemistry 73: 93–101.
possible discrimination between coffee varieties with a Hammond E.W. (2003). Vegetable oils - composition and analysis.
higher number of samples. Their different roasting In: Caballero B., Trugo L.C. and Finglas P.M. (eds),
susceptibilities will also be studied in detail. Encyclopedia of Food Sciences and Nutrition. Oxford, United
Kingdom: Elsevier Science Ltd., pp. 5916–5921.
The micro-extraction method presented proved to be
Hewavitharana A.K., Lanari M.C. and Becu C. (2004) Simultaneous
efficient for a rapid and accurate quantification of determination of vitamin E homologs in chicken meat by liquid
vitamin E in coffee beans, with low organic solvent chromatography with fluorescence detection. Journal of
Chromatography A 1025: 313–317.
consumption both during extraction and chromato-
ISO (2006). Animal and Vegetable Fats and Oils- Determination of
graphic separation. Tocopherol and Tocotrienol Contents by High-performance Liquid
Chromatography. ISO 9936, Geneve, Switzerland: ISO.
Jham G., Winkler J., Berhow M. and Vaughn S. (2007). -Tocopherol
as a marker of Brazilian coffee (Coffea arabica L.) adultera-
tion by corn. Journal of Agricultural and Food Chemistry 55:
ACKNOWLEDGMENTS 5995–5999.
Kamal-Eldin A., Görgen S., Pettersson J. and Anna-Maija L. (2000).
R.C.A. is grateful to Fundação para a Ciência e a Normal-phase high performance liquid chromatography
of tocopherols and tocotrienols: comparison of different
Tecnologia for a Ph.D. grant (SFRH/BD/22449/2005). chromatographic columns. Journal of Chromatography A 881:
The authors thank to BICAFÉ for providing raw and 217–227.
roasted coffee samples. Kölling-Speer I., Kurzrock T. and Speer K. (2006). Tocopherols in
green coffees harvested in different years. In: Proceedings of the
21st International Scientific Colloquium on Coffee. Montpellier,
France.
Nelis H.J., D’Haese E. and Vermis K. (2000). Vitamin E. In:
REFERENCES De Leenheer A.P., Lambert W.E. and Bocxlaer J.F.V. (eds),
Modern Chromatographic Analysis of Vitamins. New York:
Amaral J.S., Casal S., Torres D., Seabra R.M. and Oliveira B.P.P. Marcel Dekker Inc., pp. 143–228.
(2005). Simultaneous determination of tocopherols and tocotrie- Ogawa M., Kamiya C. and Iida Y. (1989). Contents of tocopherols in
nols in hazelnuts by a normal phase liquid chromatographic coffee beans, coffee infusions and instant coffee. Nippon Shokuhin
method. Analytical Sciences 21: 1545–1548. Kogyo Gakkaishi 36: 490–494.
Aoyama M., Maruyama T., Kanematsu H., Niiya I., Tsukamoto M., Panfili G., Fratianni A. and Irano, M. (2003). Normal phase
Tokairin S. and Matsumoto T. (1988). Studies on the improve- high-performance liquid chromatography method for the
ment of antioxidant effect of tocopherols. XVII. Synergistic effect determination of tocopherols and tocotrienols in cereals.
of extracted components from coffee beans. Yukagaku 37: Journal of Agricultural and Food Chemistry 51: 3940–3944.
606–612. Rupérez F.J., Martı́n D., Herrera E. and Barbas C. (2001).
Bramley P.M., Elmadfa I., Kafatos A., Kelly F.J., Manios Y., Chromatographic analysis of a-tocopherol and related
Roxborough H.E., Schuch W., Sheehy P.J.A. and Wagner K.-H. compounds in various matrices. Journal of Chromatography A
(2000). Vitamin E. Journal of the Science of Food and Agriculture 935: 45–69.
80: 913–938. Schneider C. (2005). Chemistry and biology of vitamin E. Molecular
Bustamante-Rangel M., Delgado-Zamarreño M.M., Sánchez-Pérez A. Nutrition & Food Research 49: 7–30.
and Carabias-Martı́nez R. (2007). Determination of tocopherols Schwartza H., Ollilainenb V., Piironenb V. and Lampib A. (2008).
and tocotrienols in cereals by pressurized liquid extraction-liquid Tocopherol, tocotrienol and plant sterol contents of vegetable
chromatography-mass spectrometry. Analytica Chimica Acta 587: oils and industrial fats. Journal of Food Composition and Analysis
216–221. 21: 152–161.
Casal S., Macedo B. and Oliveira M.B.P.P. (2001). Simultaneous Speer K. and Kölling-Speer I. (2006). The lipid fraction of the coffee
determination of retinol, b-carotene and a-tocopherol in adipose bean. Brazilian Journal of Plant Physiology 18: 201–216.
tissue by high performance liquid chromatography. Journal of Svilaas A., Sakhi A.K., Andersen L.F., Svilaas T., Ström E.C.,
Chromatography B 763: 1–8. Jacobs Jr D.R., Ose L. and Blomhoff R. (2004). Intakes of
Cros E., Fourny G. and Vincent J.C. (1985). Tocopherols in coffee. antioxidants in coffee, wine, and vegetables are correlated with
High-pressure liquid chromatographic determination. In: plasma carotenoids in humans. The Journal of Nutrition 134:
Proceedings of the 11th International Colloquium on the 562–567.
Chemistry of Coffee. Paris. Woollard D.C. and Indyk H.E. (2003). Tocopherols – properties and
Eitenmiller R.R. and Landen Jr W.O. (1999). Vitamin E. Vitamin determination. In: Caballero B., Trugo L.C. and Finglas P.M.
Analysis for the Health and Food Sciences –3. Boca Raton: CRC (eds), Encyclopedia of Food Sciences and Nutrition Oxford,
Press, pp. 109–148. United Kingdom: Elsevier Science Ltd., pp. 5789–5796.

Downloaded from fst.sagepub.com at UNIV CALIFORNIA SAN DIEGO on April 14, 2015