Академический Документы
Профессиональный Документы
Культура Документы
a r t i c l e i n f o a b s t r a c t
Article history: Seventy-three saffron samples belonging to three different storage time (<1 year, 34 and 89 years)
Received 25 November 2009 were analysed using ultrasound assisted extraction-gas chromatographymass spectrometry and ultra-
Accepted 23 March 2010 sound assisted extraction-gas chromatographyolfactometry. Nineteen volatile compounds were identi-
ed and their aromatic notes were assigned. In addition to safranal, the main compound found, (30.14
43.94% in mass of total volatiles), other major compounds were 4-hydroxy-2,6,6-trimethyl-1-cyclohex-
Keywords: en-1-carboxaldehyde and 3,5,5-trimethyl-2-cyclohexene-1-one. These compounds were signicantly dif-
Aroma
ferent for less than 1 year storage when compared with the 34 and 89 years of storage, although the
Olfactometry
Saffron (Crocus sativus L.)
minor constituents 2-hydroxy-3,5,5-trimethylcyclohex-2-en-1,4-dione and isomer of 4-hydroxy-3,5,5-
Volatile prole trimethyl-2-cyclohex-1-one varied signicantly for all three harvests. Positive and negative aroma
descriptors were related with the overall saffron chemical volatile composition. Saffron with less than
1 year storage contained a higher proportion of saffron, ower and spicy descriptors, while the oldest saf-
frons (34 and 89 years of storage) contained volatiles with vegetal, caramel and citrus notes. The aro-
matic notes that contributed most to saffron storage differentiation are spicy, freshly cut grass and
vegetable.
2010 Elsevier Ltd. All rights reserved.
0963-9969/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodres.2010.03.025
1330 L. Maggi et al. / Food Research International 43 (2010) 13291334
Numerous papers have been written on saffron volatile content tively. One microlitre of the sample was injected manually in the
(Alonso, Salinas, Esteban-Infantes, & Snchez-Fernndez, 1996; splitless mode. The identication of volatile compounds was per-
Alonso, Salinas, & Garijo, 1998; Carmona, Zalacain, Salinas, & Alon- formed as described in previous works (Anastasaki et al., 2009;
so, 2007; Du, Wang, Hu, & Yao, 2008; Kanakis, Daferera, Tarantilis, Maggi et al., 2009). The quantication of volatile compounds pres-
& Polissiou, 2004; Maggi et al., 2009; Zougagh, Ros, & Valcrcel, ent in saffron was carried out using safranal as reference as has
2006), but few on saffron aroma contribution. In addition, these been reported previously (Anastasaki et al., 2009; Maggi et al.,
references are limited by the number of samples collected, a max- 2009).
imum of three (Cadwallader, 2002; Narasimhan, Chand, & Raja- One sample per harvest was subjected to gas chromatography
lakshmi, 1992; Rdel & Petrzika, 1991). In general, these studies olfactometry (GCO) under the same chromatographic conditions
on saffron sensory evaluation remarked that a good aroma quality (column, temperature, gas ow rate) as above on a Thermo Scien-
was given by sweet, oral and spicy notes. tic Focus GC (Thermo Electron SpA, Milan, Italy). The chromatog-
The main objective of this study was to analyse 73 samples with rapher was equipped with a splitter dividing efuent between a
different storage time in order to determine how saffron volatile ame ionization detector and an olfactometry port with humidied
proles and its sensory characteristics change during storage. Aro- air as a make-up (sniffer 9000 system, Brechbhler, Scientic Anal-
matic descriptors obtained by GCMS/FID-O will be related with ysis Solution, Switzerland). The identication was based upon
overall saffron chemical volatile composition which has been ana- retention times, retention indices. Mass spectrometry was used to
lysed by USAEGCMS. conrm the identications. Snifng of the column efuent was per-
formed by three panellists that were familiarised with saffron aro-
2. Materials and methods ma. When an odour was detected, the panellist would slide a lever
on a potentiometer for which the amount of movement reected
2.1. Samples the intensity of the odour. This information, in the form of peaks
outputted to the instrument software, could be overlaid over the
A total of 73 Spanish samples with three different times of stor- data from the FID detector and used to determine which peaks were
age distributed as follows were analysed: 50 samples with less aroma descriptors. Samples were run three times by each panellist.
than 1 year of storage, 13 between 3 and 4 storage years, and 10
with 89 years of storage. All samples were obtained directly from 2.4. Statistical analysis
the producers to guarantee their origin and harvest. The samples
were kept at 4 C in absence of light until their analysis. Evaluation of the signicance of differences was performed
using analysis of variance (ANOVA). A discriminant analysis was
2.2. Chemicals and reagents also carried out. The statistical program was SPSS 17.0 for Win-
dows (SPSS Inc., Chicago, USA).
Standard: safranal with a purity of 98% was supplied by Sigma
Aldrich (Madrid, Spain).
3. Results and discussion
Solvents: diethyl ether was purchased from Panreac (Barcelona,
Spain).
3.1. Chemical volatile content
2.3. Procedure and instrumentation The eight main volatile components represent 95%, 89% and 87%
for saffron with less than 1 year, 34 and 89 years of storage,
2.3.1. Isolation of the volatile compounds (USAE) respectively (Fig. 1). In all samples, safranal (compound 1) repre-
Ultrasound assisted extraction was performed in an ultrasound sented the main constituent of volatile composition, contributing
water bath Sonorex (Berlin, Germany) Super RK 255H type highly to saffron aroma due to its high impact factor (Cadwallader,
(300 150 150 mm internal dimensions) at the xed-frequency 2002). The trend observed in Fig. 1 in relation with safranal is in
of 35 kHz. The temperature of the sonicated water bath was 25 C. agreement with the previous bibliography, since it is known that
The sample ask was charged with 4 g of saffron stigmas. The sol- during the rst 2 years of storage, safranal content increased and
vent system extractant was 40 mL of diethyl ether. Each saffron then decreased again (Carmona et al., 2005, 2006; Carmona, Zalac-
sample was sonicated twice for 15 min (two fractions per saffron ain, Salinas, & Alonso, 2007). The same behaviour was observed for
sample). For each sonication a new volume of the solvent system the samples included in this study, even if the mean value for saf-
extractant was added in the sample ask. The organic extract fron with 89 years of storage went down it was not signicantly
(80 mL) was concentrated using a rotary evaporator Laborata different from samples with less than 1 year.
4000 Efcient (Heidolph Instruments GmbH & Co. KG, Schwabach, The presence of 4-hydroxy-2,6,6-trimethyl-3-oxocyclohexa-
Germany). The volume of the nal organic extract in diethyl ether 1,4-diene-1-carboxaldehyde (compound 2) (Fig. 1) in the volatile
was 4 mL. fraction extracts depends greatly on the extraction procedure, as
it was conrmed previously by Kanakis et al. (2004). While several
2.3.2. Volatile analysis (GC/MSFID-O) authors report its content as fewer than 2.30% (Rdel & Petrzika,
In order to tentatively identify the volatile compounds of saf- 1991), for others (Semiond et al., 1996) it was found with a range
frons organic extract, GCMS was used. A Hewlett Packard 5890 of 7.533.8% when supercritical uid extraction with CO2 was per-
Series II chromatograph (Palo Alto, CA, USA) equipped with a formed. A substantial difference was found for the same saffron
5972 Series mass selective detector (MS) in the electron impact sample (more than 15% difference) when it is extracted with other
mode (70 eV) and a HP-5 ms capillary column (30 m, 0.25 mm extraction techniques such as the microsimultaneous hydrodistil-
i.d., 0.25 lm lm thickness) with helium as carrier gas at 1 mL/ lationextraction (MSDE) and ultrasound assisted extraction
min was used for the analysis of the organic extract. Column tem- (Kanakis et al., 2004). Together with its erratic behaviour (Fig. 1),
perature was initially kept for 3 min at 50 C, then gradually in- the previous information suggests that compound 2 is not a suit-
creased to 180 C with a rate of 3 C/min, and nally increased to able marker for volatile differentiation.
250 C with a rate of 15 C/min and kept for 5 min. The injector HTCC (4-hydroxy-2,6,6-trimethyl-1-cyclohexen-1-carboxalde-
and detector temperatures were set at 220 and 290 C, respec- hyde, compound 3) and isophorone (3,5,5-trimethyl-2-cyclohex-
L. Maggi et al. / Food Research International 43 (2010) 13291334 1331
O O
OH OH O O O O OH
Aromatic notes Saffron Fresh Spicy Flower Vegetable Flower Vegetable Caramel
cut grass
For each compound, different letters among bars indicate significant differences at 0.05% level.
Fig. 1. Mean value (expressed % in mass of total volatiles content %) of the rst eight volatile compounds detected in saffron belonging to different harvests.
ene-1-one, compound 4), were signicantly different for 1 year ences were observed for one and 89 years of storage, while this
storage saffron when compared with the other two times (34 compound showed a decrease within samples storaged during 3
and 89 storage years). Its disappearance from 1 to 34 storage 4 years in comparison with the other two groups of samples. On
years was congruent with the increase in safranal content, thus the other hand, the content of 2,6,6-trimethyl-2-cyclohexene-1,4-
supporting the theory that it could act as a reservoir for safranal dione, known also as 4-ketoisophorone (compound 7) and 2-hy-
generation during storage (Carmona, Zalacain, & Alonso, 2006). droxy-4,4,6-trimethyl-2,5-cyclohexadien-1-one (compound 8)
The values for compound 4, which was the main compound were structurally related to compound 5 (C9 structure com-
belonging to the C9 group, were according to the scientic bibliog- pounds). These results were consistent with the mechanism pro-
raphy (Rdel & Petrzika, 1991; Tarantilis & Polissiou, 1997). This posed by Carmona et al. (2006) and updated in this paper
compound plays a noticeable role in the generation of new com- (Fig. 2). As previously mentioned, compound 4 is degraded to gen-
pounds during the aging process (Carmona et al., 2006). For erate compounds 7 and 8, having a common pathway (Fig. 2) and
2,2,6-trimethyl-1,4-cyclohexanedione (compound 5), no differ- different from compound 5. For 4-hydroxy-3,5,5-trimethyl-2-
Fig. 2. Proposed generation mechanism of the most relevant saffron aroma compounds.
1332 L. Maggi et al. / Food Research International 43 (2010) 13291334
4
c
b
3
% content
2
b
b
a b b b
1 a a a a a
a c
b b a c b b b
a b a b a a a a
a a
0
compound 9 10 11 12 13 14 15 16 17 18 19
OH OH O
O OH O
O O
OH O O OH
Isomer of Isomer of
compound 9 compound 6
Aromatic notes Citrus Citrus Flower Caramel Vegetable Flower Citrus Vegetable Caramel Flower Vegetable
For each compound, different letters among bars indicate significant differences at 0.05% level.
Fig. 3. Mean value (expressed % in mass of total volatiles content %) of the minor volatile compounds detected in saffron belonging to different harvests.
cyclohex-1-one (compound 6), no differences have been found for (compound 12), the safranal isomer, decreased as safranal did since
the 3 groups of samples (Fig. 3), although this compound is an probably they shared the same mechanism of formation (Carmona,
intermediate of isomer of 4-hydroxy-3,5,5-trimethyl-2-cyclohex- Zalacain, Salinas, & Alonso, 2007). On the contrary, isomer of
1-one (compound 19) which increases its content signicantly 4-hydroxy-2,6,6-trimethyl-3-oxocyclohex-1-en-1carboxaldehyde
with storage time. (compound 13) and compound 17 increased their content during
With regard to the minor constituents (Fig. 3), those able to dis- storage probably due to the oxidation of safranal proposed in
tinguish different storage times were isophorone-4-methylene Fig. 4. Also, the mean value of compound 15 increased during stor-
(compound 10), 2-hydroxy-3,5,5-trimethylcyclohex-2-en-1,4- age. Furthermore, it could be an intermediary of the C9 or C10 com-
dione (compound 11), 2-phenylethanol (compound 14), 3-[(E)- pounds, such as compounds 7 and 19, which may explain their
but-1-enyl]-2,4,4-trimethyl-cyclohex-2-enol (compound 15), increment. Compound 19 signicantly increased with the storage
2,6,6-trimethyl-3-oxo-1-cyclohexen-1-carboxaldehyde (com- time, thus it could be considered as a warning signal and an indi-
pound 17), 2,2-dimethyl-cyclohexane-1-carboxaldehyde (com- cator of old harvests. Other minor compound such as 14 is charac-
pound 18) and compound 19, especially compounds 11 and 19. teristic of Greek saffron due to the dehydration process used in this
Compound 11 varied signicantly among the three harvests. The country (Carmona, Zalacain, Salinas, & Alonso, 2007; Goliaris,
content of 2,6,6-trimethyl-1,4-cyclohexadiene-1-carboxaldehyde 1999). Its high concentration could be considered negative when
CHO
CHO O
OH
CHO (13)
? (1) CHO
*
CHO
OH
(3) * O
(17)
(12)
*
* These compound pathways come from Carmona et al. (2006a).
Fig. 4. Possible evolution of compounds during the storage and aging step.
L. Maggi et al. / Food Research International 43 (2010) 13291334 1333
the presence of stamens or an excess of pollen are indicated in saf- M.R. Salinas for her analysis supervision. Thanks to Ms. K. Walsh
fron (Goliaris, 1999). On the other hand, its presence could be for proofreading the English text.
positive and used as a genuineness indicator for saffron with a
determined production origin (Carmona, Zalacain, Salinas, &
Alonso, 2007). References
Tarantilis, P. A., Polissiou, M., & Manfait, M. (1994). Separation of picrocrocin, cis- Winterhalter, P., & Straubinger, R. M. (2000). Saffron. Renewed interest in an ancient
trans-crocins and safranal of saffron using high-performance liquid spice. Food Reviews International, 16, 3959.
chromatography with photodiode-array detection. Journal of Chromatography Zarghami, N. S., & Heinz, D. E. (1971). Monoterpene aldehydes and isophorone-
A, 664(1), 5561. related compounds of saffron. Phytochemistry, 10, 27552761.
Tarantilis, P. A., Tsoupras, G., & Polissiou, M. (1995). Determination of saffron Zougagh, M., Ros, A., & Valcrcel, M. (2006). Determination of total safranin by
(Crocus sativus L.) components in crude plant extract using high-performance in situ acid hydrolysis in supercritical uid media: Application to the quality
liquid chromatographyUVvisible photodiode-array detection mass control of commercial saffron. Analytica Chimica Acta, 578(2), 117121.
spectrometry. Journal of Chromatography A, 699(12), 107118.