Characterisation of aroma active compounds of Spanish saffron by gas chromatography–olfactometry: Quantitative evaluation of the most relevant aromatic compounds

Food Chemistry 127 (2011) 1866–1871

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Characterisation of aroma active compounds of Spanish saffron by gas chromatography–olfactometry: Quantitative evaluation of the most relevant aromatic compounds Laura Culleré ⇑, Felipe San-Juan, Juan Cacho Laboratory for Aroma Analysis and Enology, Aragón Institute of Engineering Research (I3A), Department of Analytical Chemistry, Faculty of Sciences, University of Zaragoza, 50009 Zaragoza, Spain

a r t i c l e

i n f o

Article history: Received 12 November 2010 Received in revised form 25 January 2011 Accepted 2 February 2011 Available online 24 February 2011 Keywords: Active aroma compounds Saffron GC–olfactometry Multidimensional chromatography Safranal Isophorone

a b s t r a c t The aromatic composition of a saffron sample from Valle de Jiloca (Teruel, Spain) was evaluated for the first time by gas chromatography–olfactometry (GC–O). Volatiles released by 10 g of saffron sample were collected in a trapping system consisting of LiChrolut EN resins and eluted with dichloromethane/methanol (95:5). GC–O revealed that the aroma emitted by this kind of saffron is due to at least twenty different aroma molecules. From an olfactometric point of view, the most important aroma compounds of this saffron sample were safranal (modified frequency value [MF] 93%), followed by 2,3-butanedione, hexanal, E-2-nonenal and an odorant with a characteristic aroma of burnt curry that could not be identified. All of them had MF values higher than 70%. An estimate was made of the levels of these aromatic molecules detected by GC–O. Safranal and isophorone, both volatiles with aromatic descriptors of ‘‘saffron’’ were quantified using a headspace microextraction (HS-SPME) method. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Saffron is the commercial name of the dried stigmas of Crocus sativus L. flowers. Saffron is considered to be one of the most expensive spices and is cultivated in many countries, including Spain, Greece, India, China and Iran. Saffron is used mainly as a spice, food colorant and herbal medicine, in which it is used as an analgesic (Ríos, Recio, Giner, & Manez, 1996). In recent years it has been demonstrated that saffron has distinctive anticancer activities (Nair, Pannikar, & Pannikar, 1991; Ríos et al., 1996). Many extraction processes have been used to extract the chemical components of saffron, such as hydrodistillation (HD), micro-simultaneous hydrodistillation–extraction (MSDE) (Rodel & Petrzika, 1991; Tarantilis & Polissiou, 1997), vacuum headspace (VHS) (Tarantilis & Polissiou, 1997), supercritical fluid extraction (SFE) (Zougagh, Rios, & Valcarcel, 2006), thermal desorption (TD) (Alonso, Salinas, EstebanInfantes, & SanchezFernandez, 1996; Carmona et al., 2006), extraction with organic solvent (Tarantilis, Polissiou, & Manfait, 1994), solid-phase microextraction (SPME) (D’Auria, Mauriello & Rana, 2004) and ultrasonic solvent extraction (USE) (Jalali-Heravi, Parastar, & Ebrahimi-Najafabadi, 2009; Kanakis, Daferera, Tarantilis, & Polissiou, 2004).

⇑ Corresponding author. Tel.: +34 976762076; fax: +34 976761292. E-mail address: [email protected] (L. Culleré). 0308-8146/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodchem.2011.02.015

Many papers have been published on saffron volatile composition (Carmona et al., 2006; Carmona, Zalacain, Salinas, & Alonso, 2007; Du, Wang, Hu, & Yao, 2008; Kanakis et al., 2004; Maggi et al., 2009; Rodel & Petrzika, 1991; Tarantilis & Polissiou, 1997; Zougagh et al., 2006). All of them have focused on identifying the major compounds present in this spice using gas chromatography coupled to mass spectrometry (GC–MS). However, information on the range of the levels of some of these major compounds is provided in only some papers (Alonso et al., 1996; Jalali-Heravi et al., 2009; Kanakis et al., 2004). In the paper published by Kanakis et al. (2004), 2,6,6-trimethyl-1,3-cyclohexadiene-1-carboxaldehyde, or safranal, and 4-hydroxy-2,6,6-trimethyl-1-cyclohexene1-carboxaldehyde, HTCC, were quantified in Greek saffron. Alonso et al. (1996) quantified only the safranal content present in saffron samples from La Mancha (Spain), whereas Jalali-Heravi et al. (2009) quantified forty compounds identified by GC–MS in Iranian saffron. The information obtained by GC–MS analysis is interesting and useful for knowing the qualitative profile, but incomplete because it is necessary to evaluate which compounds present are odour-active and contribute to the saffron aroma. In order to locate and rank the odorants that are most important aromatically, olfactometric study (GC–O) is required. Few papers have undertaken GC–O studies of saffron samples. Rodel & Petrzika (1991) were the first to use GC–O to evaluate saffron aroma. In their first paper on the topic, they only confirmed the highly intense and characteristic odour associated with the peak

L. Culleré et al. / Food Chemistry 127 (2011) 1866–1871

generated by safranal, which is considered to be the impact compound of saffron. However, this study revealed that many other unidentified compounds contribute to the complete aroma of saffron. Cadwallader, Baek, and Cai (1997) carried out the second and most important olfactometric study of saffron samples using aroma extract dilution analysis (AEDA). They concluded that a compound tentatively identified as 2-hydroxy-4,4,6-trimethyl2,5-cyclohexadien-1-one has an extremely important role in saffron aroma, even greater than that of safranal. Recently, Maggi et al. (2009) analysed different Spanish saffron samples by GC– O. The aim was to assign aroma descriptors to the nineteen volatile saffron constituents identified by GC–MS. In conclusion, the only study that reveals complete olfactometric information in detail (gas chromatographic retention data estimated on both polar and non-polar columns, olfactory descriptors, chemical identities of the compounds responsible for these aromas and a concentration range of the most important odorant) was that of Cadwallader et al. (1997). Therefore, the first aim of this study was to examine the aromatic profile of a saffron sample from Teruel (Spain) by GC–O for the first time and determine the key odorants. The second aim was to provide quantitative or semi-quantitative information about all of the compounds considered relevant in saffron according to the previous olfactometric study.

2. Materials and methods 2.1. Samples The saffron sample used in this study was obtained from a saffron farming area in the province of Teruel (Aragón, Spain). A large amount of saffron was provided by La Carrasca S.A. The samples were kept at ambient temperature and protected from light until analysis. 2.2. Reagents Solvents: dichloromethane and methanol were purchased from Merck (Darmstadt, Germany); water was purified in a MilliQ system from Millipore (Bedford, MA). Resins: LichrolutÒ EN resins (non-polar resins) and polypropylene cartridges (0.8 cm internal diameter, 3 ml internal volume) were supplied by Merck (Darmstadt, Germany). Standards: The standards used for identifications were supplied by Aldrich (Steinheim, Germany), Fluka (Buchs, Switzerland), PolyScience (Niles, IL), Lancaster (Strasbourg, France) and Panreac (Barcelona, Spain). An alkane solution (C8–C28), 20 mg L1 in dichloromethane, was used to calculate the linear retention index (LRI) of each analyte. A standard of (Z)-2-nonenal was not found, but this aldehyde is present in commercial (E)-2-nonenal at a concentration of 5–10% (Ferreira et al., 2009; Valim, Rouseff, & Lin, 2003). 2.3. Gas chromatography–olfactometry 2.3.1. Preparation of extracts Saffron volatiles were collected using a purge and trap system (Campo, Ferreira, Escudero, & Cacho, 2005). The LichrolutÒ EN cartridge (400 mg) was placed on top of a bubbler flask containing 10 g of saffron. This sample was purged by a stream of nitrogen at ambient temperature for 4 h. Volatile saffron constituents released in the headspace were trapped in the cartridge containing the sorbent and were further eluted with 3.2 mL of dichloromethane containing 5% methanol.

1867

2.3.2. GC–O analysis All sniffing experiments were carried out in a Trace gas chromatograph from ThermoQuest (San Jose, CA), equipped with a flame ionisation detector (FID) and a sniffing port (ODO-1 from SGE, Ringwood, Australia) connected by a flow splitter to the column exit. The columns used for this study were a DB-Wax column from J&W (Folsom, CA), 30 m  0.32 mm ID, 0.5-lm film thickness, and an HP-5MS column from Agilent (Santa Clara, CA), 30 m  0.25 mm ID, 0.25-lm film thickness. A constant pressure of 52 kPa was maintained throughout the analysis time. The carrier was H2. One microlitre was injected in splitless mode for 1 min splitless time. Injector and detector were both kept at 250 °C. The temperature program was 40 °C for 5 min, which was then raised by 4 °C min1 to 100 °C, followed by 6 °C min1 to 220 °C, and finally kept at 220 °C for 20 min. To prevent condensation of high-boiling compounds on the sniffing port, the port was heated with a laboratory-made rheostat. The olfactometric strategy used in this study combined measurements of intensity and frequency of detection and has been widely used in previous papers published by members of our laboratory (Cullere et al., 2010; Escudero, Campo, Farina, Cacho, & Ferreira, 2007). In this GC–O study, assessments were carried out by a panel of six expert judges belonging to our laboratory. The saffron extract was concentrated to 200 lL, and the extract was smelled once by each panellist. Sniffing time was approximately 40 min and each judge took part in one session per day. Panellists were asked to score the intensity of each aromatic stimulus using a 4-point scale (0 = not detected, 1 = weak, 2 = clear but not intense note, 3 = intense note). The signal obtained was modified frequency (MF(%)), a parameter that was calculated with the formula proposed by Dravnieks (1985):

MFð%Þ ¼ ðFð%Þ  Ið%ÞÞ1=2 where F(%) is the detection frequency of an aromatic attribute expressed as a percentage of total number of judges and I(%) is the average intensity expressed as percentage of the maximum intensity. 2.4. Gas chromatography–mass spectrometry (GC–MS) 2.4.1. Quantification of safranal and isophorone (HS-SPME-GC–MS) The quantification method was designed and published by D’Auria et al. (2004). This methodology is based on HS-SPMEGC–MS analysis and uses a 100-lm PDMS-SPME fibre, purchased from Supelco-Spain (Madrid, Spain). The fibre was maintained over the sample (0.1 g) in a 20-mL vial at 36 °C for 20 min. The analyses were performed with a CP-3800 chromatograph coupled to a Saturn 2200 ion trap mass spectrometric detection system from Varian (Palo Alto, CA). A DB-Wax-ETR capillary column (J&W Scientific) of 60 m  0.25 mm I.D., film thickness 0.25 lm, was used, preceded by a 3 m  0.25 mm uncoated pre-column from Supelco (Bellefonte, PA). Helium was the carrier gas at a flow rate of 1 mL min1. The oven temperature was initially 40 °C for 5 min and was then raised at 4 °C min1 to 100 °C, followed by a rate of 6 °C min1 to 220 °C and finally held at this temperature for 30 min. The MS parameters were: MS transfer line and chamber ionisation temperature 200 °C and trap emission current 80 lA. The global run time was recorded in full scan mode (m/z 45–200 mass range). The chromatographic data were analysed with Varian Saturn GC–MS Version 5.2 software. The injection was in split mode 1/80 at a temperature of 250 °C. A desorption time of 0.4 min was used. The detector was held at 230 °C.

1868

L. Culleré et al. / Food Chemistry 127 (2011) 1866–1871

b-Damascenone was added as an internal standard because it has a chemical structure similar to that of the targeted analytes and is not present in the saffron sample studied. Fifty microlitres of the internal standard solution (b-damascenone 1 mg L1 in water) were added to the mass of 0.1 g placed in a 20-mL SPME vial. The linearity of the method was studied by standard addition of known amounts of analytes (20, 50, 100 and 200 lg) placed in the 20-mL vial of SPME containing 0.1 g of saffron sample. Reproducibility was studied by evaluating the signal obtained in nine determinations performed on different days and at different levels of concentration (an unspiked sample and spiked samples at two different concentration levels: 20 and 100 lg). 2.4.2. Identification and semi-quantitative evaluation of the rest of odorants (GC–O–GC–O–MS) This multidimensional chromatographic system was used to identify and carry out a semi-quantitative estimation of the most relevant odorants detected previously by GC–O. Fifty microlitres of the extract, obtained by the purge and trap system and subsequently concentrated to 200 lL, was injected in a multidimensional GC–O–GC–O–MS system from Varian. The system consisted of two independent gas chromatographs interconnected by a thermoregulated transfer line kept at 200 °C and equipped with a Deans valve switching system (Valco Instruments, Houston, TX), two olfactory ports and FID and MS detectors, as described previously (Campo, Cacho, & Ferreira, 2006; Cullere, Escudero, Perez-Trujillo, Cacho, & Ferreira, 2008). Chromatograph 1 was equipped with a DB-Wax column (polyethylene glycol) from J&W, 30 m  0.32 mm I.D. with 0.5-lm film thickness. The oven temperature program was 40 °C for 5 min, which was then raised by 4 °C min1 to 100 °C, followed by 6 °C min1 to 220 °C, and finally held at this temperature for 40 min. Initially, the GC–O extract (50 lL) was monitored by olfactometry in the first chromatograph to select the fraction containing the target odorant. In further chromatographic runs, selective heart-cuttings were made to isolate these odorants, which were transferred to the second chromato-

graph equipped with a FactorFour VF-5MS column from Varian (30 m  0.32 mm; 1-lm film thickness). In this second oven, isolated odorant was trapped in a CO2 cryotrapping unit and monitored by olfactometry with simultaneous MS detection. Two minutes after the heart-cutting, CO2 flow was removed at the same time that the temperature program (4 °C min1 up to 200 °C and then 50 °C min1 up to 300 °C) of the second oven was activated. MS parameters were: transfer line 170 °C; ion trap 150 °C, and trap emission current 30 lA. The global run time was recorded in full scan mode (m/z 40–250 mass range). FID and MS data were registered and processed with Workstation 6.30 software equipped with the NIST 98 (US National Institute of Standards and Technology) MS library (NIST, Gaithersburg, MD). The programmable temperature vaporising injector (PTV) conditions, delay time and heart-cutting interval were the same as those used in a previous paper (Campo et al., 2006). The identity of the odorants was determined from the odour description, mass spectrum and linear retention indices on both columns (DB-Wax and VF-5MS). The identity was confirmed by injection of the pure reference standard, when available. Semi-quantitative estimation was carried out by preparing synthetic solutions containing known amounts of the odorants and analysing them using the same instrument. Some recovery studies were necessary for estimation. The chromatographic signals produced by spiked saffron (with a mass of 5 lg of each odorant) were compared to the signals resulting from the addition of equivalent analyte masses to dichloromethane solutions.

3. Results and discussion 3.1. Aromatic profile study of saffron by GC–O The results of these experiments with this saffron sample are summarised in Table 1. The table shows the chromatographic retention data of the different odour zones detected in the olfactometric experiments, the odour descriptions of the zones given by the trained panel, the chemical identity of the odorant

Table 1 GC–O study: gas chromatographic retention data, olfactory description, chemical identity and modified frequency percentage MF(%) for each compound. LRI VF5-MSDBWax

967 1078 1215 1288 1301 1336 1400 1453 1505 1535 1585 1630 1645 1671 1701 1812 1828 1917 2040 2071

LRI DBWax HP5-MS

600 775 <900 1004 993 1007 1058 600 1170 1182 1180 878 1131 <900 1270 1390 1134 1131 1060 1159

Odor descriptor

Butter, cream Grass Grass, geranium Lemon Mushroom Clove, spicy Saffron Vinegar Green, metallic Melon, aldehydic Cucumber Cheese Saffron Cheese Rancid oil Fatty, deep-fried Burnt, curry Roses Cotton candy Cotton candy

Supplierd

Identity

b

2,3-Butanedione Hexanalb 3-Hexen-2-one unknown 1a Octanalb,a 1-Octen-3-oneb 6-Methyl-5-hepten-2-oneb,a Isophoroneb Acetic acidb (Z)-2-Nonenalb (E)-2-Nonenalb (E,Z)-2,6-Nonadienalc Butyric acidb,a Safranalb Isovalerianic acidb (E,E)-2,4-Nonadienalc,a (E,E)-2,4-Decadienalb Unknown 2a b-Phenylethanolb Furaneolc Homofuraneolc,a

1 1 1 4 1 1 5 1 1 1 3 1 2 1 1 2 1 1

%MF

73 71 65 61 55 65 61 65 50 71 67 51 93 61 59 43 75 33 45 48

Odorants previously reported In GC–O studies

In GC–MS studies

A Aunknown None None A None A, B, C A Aunknown Aunknown A None A, B, C A None A None A,C A None

None D, E None D, E None D, E D, E, F, G D, E None None None None D, E, F, G None None None None D,E,F None None

A: (Cadwallader et al., 1997); B; (Rodel & Petrzika, 1991); C: (Maggi et al., 2009); D: (Du et al., 2008); E: (D’Auria, Mauriello, Racioppi & Rana, 2006); F: (Tarantilis & Polissiou, 1997); G: (Kanakis et al., 2004). a Odorants reported for the first time by GC–O in saffron. b Identification based on coincidence of gas chromatographic retention in two different columns and mass spectrometric data with those of the pure compounds available in the laboratory. c As for footnote ‘a’, but these compounds did not produce any clear signal in the mass spectrometer because of its low concentration. d Suppliers: (1) Aldrich; (2) Fluka; (3) PolyScience; (4) Lancaster; (5) Panreac.

1869

L. Culleré et al. / Food Chemistry 127 (2011) 1866–1871

responsible for the odour and the olfactometric scores (as modified frequency in %) assigned by the panel of experts. Only the odorants with MF higher than 30% are included in this table. Therefore, the MF (%) of the odorants given in the table should give a reasonable idea of the potential importance of the compound in the aroma profile of this saffron sample. As shown in the table, up to 20 different important odour zones were detected in the GC–O experiments. The identity of all the chemicals responsible for these odours could be established with different degrees of certainty, as can be seen in Table 1. Only two of them were not identified. The high modified frequency value of safranal (93%) was noteworthy. 2,3-Butanedione, hexanal, (E)2-nonenal and a fourth unknown compound with a characteristic odour of burnt curry also presented high modified frequencies. Most of the remaining odorants had modified frequency values that were quite high, 50–70%. It is also important to emphasise the role of carbonyl compounds in saffron aroma. Only six out of the twenty compounds were not carbonyl compounds (Table 1). The odorant compounds included three acids (acetic, butyric and isovaleric), an alcohol (b-phenylethanol) and two furans (furaneol and homofuraneol). The carbonyls included two aliphatic aldehydes (hexanal and octanal), three ketones, five alkenals, and two carbonyls with a cyclohexane base (safranal and isophorone). Thirteen of these odorants were previously considered relevant (Cadwallader et al., 1997). Three of them could not be identified then, but were found to be hexanal, (E)-2-nonenal and (Z)-2-nonenal in the present study. Seven compounds were identified for the first time as relevant odorants in a GC–O study of saffron aroma (Table 1). These compounds were octanal, butyric acid, (E,E)-2,4-nonadienal, homofuraneol, 6-methyl-5-hepten-2-one and two unknown compounds. Octanal and 6-methyl-5-hepten-2-one have been detected previously by GC–MS analysis in saffron samples (D’Auria, Mauriello, Racioppi & Rana, 2006; Du et al., 2008). Most of these compounds (except butyric acid and homofuraneol) had MF values higher than 55%. This may indicate that some of these newly identified components of saffron might be important compounds in its aroma. On the other hand, some odorants detected in a previous GC–O study (Cadwallader et al., 1997), such as 2-acetylpyrroline and three unknown compounds with RI (DB-Wax) 1682, 1734 and 1980, respectively, were not identified in our aromatic profile. The odorant with RI (DB-Wax) 1734, identified as 2-hydroxy-4,4,6-trimethyl-2,5-cyclohexadien-1-one by Cadwallader et al. seems to play a special role in the saffron sample from La Mancha (Spain) because it was identified as the most important aromatic compound, even more important than safranal. However, 2-hydroxy4,4,6-trimethyl-2,5-cyclhexadien-1-one was detected in our saffron sample with a MF of less than 30% (exactly 19%) and is thus not listed in Table 1. Therefore, it could be concluded that despite the presence of this odorant in both kinds of saffron, its concentration was only high enough to be relevant in one of the saffron varieties. 3.2. HS-SPME-GC–MS analysis of safranal and isophorone The first aim of this study was to develop a unique methodology for quantifying all of the odorants considered to be relevant in

saffron aroma by GC–O study. We considered applying the method based on HS-SPME recently developed by D’Auria et al. (2004), due to a series of obvious advantages, such as ease of automation, simple management, and the absence of any need for organic solvents. Moreover, the D’Auria et al. method combines extraction and preconcentration in one step, so this technique provides significantly more rapid sample preparation than the majority of traditional methods. However, only seven compounds presented chromatographic signals when we applied this method on a saffron sample: two carbonyls (hexanal and 6-methyl-5-hepten-2-one), two acids (acetic and butyric acid), b-phenylethanol, isophorone and safranal. Isophorone and safranal merit special attention because of their characteristic aroma of saffron, which makes them key odorants in this spice. 3.2.1. Method validation Quality parameters, i.e., linearity, limits of detection and precision, were evaluated. The linearity of the method was obtained by plotting the calibration graphs of the corresponding ion peak areas (Table 2) against the mass of each compound added to the saffron sample (0.1 g). The linearity was satisfactory, with coefficients of determination of more than 0.99 for safranal and isophorone. The sensitivity of the method was evaluated in terms of the limit of detection (LOD). This parameter was defined as the amount of analyte that produced a signal three times greater than the baseline noise. Under the experimental conditions, LODs were between 0.03 and 0.3 mg/kg (Table 2). The precision of this method was evaluated as reproducibility by performing replicated analyses of saffron samples at three concentration levels: unspiked and spiked at two different concentration levels (low and high). These analyses were carried out on two days. The relative standard deviation (RSD%) of the results was 66% for safranal and isophorone (Table 2). 3.2.2. Safranal and isophorone quantification The levels of concentration of these impact odorants in the saffron sample analysed are shown in Table 2. There is a limited amount of information in the scientific literature about the levels of these aroma molecules emitted by saffron. The concentration of safranal shown in Table 2 is much lower than the concentration reported in Iranian saffron (Jalali-Heravi et al., 2009) (436 mg/100 g) and much higher than the concentration of 20.6 mg/100 g found in a saffron sample from La Mancha (Spain) (Alonso et al., 1996). However, the level of safranal present in our saffron sample was relatively consistent with the wide range of concentrations (40–700 mg/100 g) previously reported in different Greek saffron samples (Kanakis et al., 2004). Isophorone has been quantified in Iranian saffron at 21.0 mg/100 g (Jalali-Heravi et al., 2009), more than the double the level found in our study. These differences reflect a wide variability in the aromatic composition, at least in these two odorants, that could be attributed to differences in the origin of the samples analysed and above all in the extraction methodology applied. In Cadwallader et al. (1997) volatile compounds were isolated by different methodologies. The first strategy consisted of simultaneous steam distillation–solvent extraction (SDE) and the second

Table 2 Quality parameters of HS-SPME-GC–MS methodology developed to quantify safranal and isophorone. Masses of the ions selected for the, equations of calibration graphs, linear dynamic ranges, coefficients (r2), limits of detection (LODs) and reproducibility, R.S.D. (%).

a b

Compound

m/z

Equationa

r2

Range (mg/kg)

LOD (mg/kg)

R.S.D. (%)b

mg/kg saffron

Safranal Isophorone

150 138

y = 3802.5X + 519195 y = 2467.1 + 24447

0.9993 0.9991

0.03–3300 0.3–2000

0.03 0.3

6 4

1365 ± 81.9 99.1 ± 4.0

n = 5 calibration points constructed from saffron sample (0.1 g) and with different spiked levels (20, 50, 100 and 200 lg). Mean of nine determinations on two different days using saffron samples spiked with different levels of these compounds.

1870

L. Culleré et al. / Food Chemistry 127 (2011) 1866–1871

Table 3 Estimated concentrations of aroma compounds in saffron sample.

Identity

(lg/kg) By GC–O–GC–O–MS system

2,3-Butanedione Hexanal Unknown 1 Octanal 1-Octen-3-one 6-Methyl-5-hepten-2-one Isophorone Acetic acid (Z)-2-Nonenal (E)-2-Nonenal (E,Z)-2,6-Nonadienal Butyric acid Safranal Isovalerianic acid (E,Z)-2,4-Nonadienal (E,E)-2,4-Decadienal Unknown 2 b-Phenylethanol Furaneol Homofuraneol

0.50 150 n.e. 30 0.006 600 – 400 0.15 14 0.40 120 – 0.40 0.050 0.50 n.e. 20 0.004 0.016

(–): Concentration level not estimated because these compounds have been quantified previously by HS-SPME-GC–MS.

was a direct solvent extraction (DE). As a result of this study, different volatile profiles were obtained by these different strategies. This example shows the influence of the extraction methodology employed on the aroma composition obtained. Moreover, it is already known that the mode of production varies according to the geographical area and probably is responsible for the final saffron aroma (Carmona et al., 2007). Therefore, analysis of the aroma would be a useful tool for assessing the origin of the sample, making it possible to obtain a fingerprint of each variety of saffron, to avoid fraudulent labelling. 3.3. Semi-quantitative estimation of other aroma active compounds In order to estimate the rest of the important odorants in saffron, an extract obtained by means of the purge and trap system was used. This extract, which was previously concentrated to 200 lL, was injected in a multidimensional chromatograph. This strategy allowed analytes of interest to be captured selectively and taken to a second dimension. As many injections as needed were performed to capture the 16 odour zones (Table 3). The concentrations of safranal and isophorone were not estimated this way because these compounds were already quantified by HSSPME-GC–MS. On the other hand, known masses of the pure references were injected and captured as before with the saffron extract. Using the percentage recovery of each compound in the extraction process (data not shown), the response factors were estimated and the masses of emitted odorants were evaluated. The estimated values are summarised in Table 3. According to the results shown in Table 2, it was concluded that safranal and isophorone were the two compounds with the highest concentration. 6-Methyl-5-hepten-2-one, acetic and butyric acid, and hexanal followed them in importance, all at levels higher than 100 lg/kg. 4. Conclusions The aromatic profile of saffron from Valle de Jiloca (Teruel, Spain) was evaluated by gas chromatography–olfactometry (GC–O) to investigate the volatile molecules that constitute the aroma of this kind of saffron. As a result of this study, twenty

compounds were reported as relevant odorants. Notably, most of them (except eight odorants) were carbonyl compounds. Only two odorants were not identified. Five odorants had modified frequency values of more than 70%. Safranal was the most important odorant (with MF of 93%), followed by an unknown compound with a burnt curry aroma, (75%), and by 2,3-butanedione, (E)-2nonenal and hexanal. To complete the characterisation of saffron from Valle de Jiloca (Teruel, Spain), a quantitative estimate was made of nearly all the major odorants detected in GC–O. Acknowledgement This work was funded by a ‘‘Cheque tecnológico’’ from Govern of Aragón (D.G.A.). References Alonso, G. L., Salinas, M. R., EstebanInfantes, F. J., & SanchezFernandez, M. A. (1996). Determination of safranal from saffron (Crocus sativus L.) by thermal desorption gas chromatography. Journal of Agricultural and Food Chemistry, 44(1), 185–188. Cadwallader, K. R., Baek, H. H., & Cai, M. (1997). Characterization of saffron flavor by aroma extract dilution analysis. In S.J. Riach, & C.T. Ho (Eds.), Spices: Flavor Chemistry and Antioxidant Properties (pp. 66–79). American Chemical Society Symposium Series 660: Washington, DC. Campo, E., Cacho, J., & Ferreira, V. (2006). Multidimensional chromatographic approach applied to the identification of novel aroma compounds in wine – Identification of ethyl cyclohexanoate, ethyl 2-hydroxy-3-methylbutyrate and ethyl 2-hydroxy-4-methylpentanoate. Journal of Chromatography A, 1137(2), 223–230. Campo, E., Ferreira, V., Escudero, A., & Cacho, J. (2005). Prediction of the wine sensory properties related to grape variety from dynamic-headspace gas chromatography–olfactometry data. Journal of Agricultural and Food Chemistry, 53(14), 5682–5690. Carmona, M., Martinez, J., Zalacain, A., Rodriguez-Mendez, M. L., de Saja, J. A., & Alonso, G. L. (2006). Analysis of saffron volatile fraction by TD-GC–MS and enose. European Food Research and Technology, 223(1), 96–101. Carmona, M., Zalacain, A., Salinas, M. R., & Alonso, G. L. (2007). A new approach to saffron aroma. Critical Reviews in Food Science and Nutrition, 47(2), 145–159. Cullere, L., Escudero, A., Perez-Trujillo, J. P., Cacho, J., & Ferreira, V. (2008). 2-Methyl3-(methyldithio)furan: A new odorant identified in different monovarietal red wines from the Canary Islands and aromatic profile of these wines. Journal of Food Composition and Analysis, 21(8), 708–715. Cullere, L., Ferreira, V., Chevret, B., Venturini, M. E., Sanchez-Gimeno, A. C., & Blanco, D. (2010). Characterisation of aroma active compounds in black truffles (Tuber melanosporum) and summer truffles (Tuber aestivum) by gas chromatography– olfactometry. Food Chemistry, 122(1), 300–306. D’Auria, M., Mauriello, G., Racioppi, R., & Rana, G. L. (2006). Use of SPME-GC–MS in the study of time evolution of the constituents of saffron aroma: Modifications of the composition during storage. Journal of Chromatographic Science, 44(1), 18–21. D’Auria, M., Mauriello, G., & Rana, G. L. (2004). Volatile organic compounds from saffron. Flavour and Fragrance Journal, 19(1), 17–23. Dravnieks, A. (1985). Atlas of odor character profiles. PA.: In A. Ed. Philadelphia. p. 354. Du, H. Y., Wang, J., Hu, Z. D., & Yao, X. J. (2008). Quantitative structure–retention relationship study of the constituents of saffron aroma in SPME-GC–MS based on the Projection Pursuit Regression method. Talanta, 77(1), 360–365. Escudero, A., Campo, E., Farina, L., Cacho, J., & Ferreira, V. (2007). Analytical characterization of the aroma of five premium red wines. Insights into the role of odor families and the concept of fruitiness of wines. Journal of Agricultural and Food Chemistry, 55(11), 4501–4510. Ferreira, V., San Juan, F., Escudero, A., Cullere, L., Fernandez-Zurbano, P., SaenzNavajas, M. P., et al. (2009). Modeling quality of premium Spanish Red wines from gas chromatography–olfactometry data. Journal of Agricultural and Food Chemistry, 57(16), 7490–7498. Jalali-Heravi, M., Parastar, H., & Ebrahimi-Najafabadi, H. (2009). Characterization of volatile components of Iranian saffron using factorial-based response surface modeling of ultrasonic extraction combined with gas chromatography–mass spectrometry analysis. Journal of Chromatography A, 1216(33), 6088–6097. Kanakis, C. D., Daferera, D. J., Tarantilis, P. A., & Polissiou, M. G. (2004). Qualitative determination of volatile compounds and quantitative evaluation of safranal and 4-hydroxy-2,6,6-trimethyl-1-cyclohexene-1-carboxaldehyde (HTCC) in Greek saffron. Journal of Agricultural and Food Chemistry, 52(14), 4515–4521. Maggi, L., Carmona, M., del Campo, C. P., Kanakis, C. D., Anastasaki, E., Tarantilis, P. A., et al. (2009). Worldwide market screening of saffron volatile composition. Journal of the Science of Food and Agriculture, 89(11), 1950–1954.

L. Culleré et al. / Food Chemistry 127 (2011) 1866–1871 Nair, S. C., Pannikar, B., & Pannikar, K. R. (1991). Antitumor activity of saffron (Crocus sativus). Cancer Letters, 57, 109–114. Ríos, J. L., Recio, M. C., Giner, R. M., & Manez, S. (1996). An update review of saffron and its active constituents. Phytotherapy Research, 10, 189–193. Rodel, W., & Petrzika, M. (1991). Analysis of the volatile components of saffron. HrcJournal of High Resolution Chromatography, 14(11), 771–774. Tarantilis, P. A., Polissiou, M., & Manfait, M. (1994). Separation of picrocin, cis-transcrocins and safranal of saffron using high-performance liquid-chromatography with photodiode-array detection. Journal of Chromatography A, 664(1), 55–61.

1871

Tarantilis, P. A., & Polissiou, M. G. (1997). Isolation and identification of the aroma components from saffron (Crocus sativus). Journal of Agricultural and Food Chemistry, 45(2), 459–462. Valim, M. F., Rouseff, R. L., & Lin, J. M. (2003). Gas chromatographic–olfactometric characterization of aroma compounds in two types of cashew apple nectar. Journal of Agricultural and Food Chemistry, 51(4), 1010–1015. Zougagh, M., Rios, A., & Valcarcel, M. (2006). Determination of total safranal by in situ acid hydrolysis in supercritical fluid media: Application to the quality control of commercial saffron. Analytica Chimica Acta, 578(2), 117–121.