Musty Off-Flavor in

Oct 7, 2016 - (9) Reported concentrations were also similar in both studies,(8, 9) but odor activity values (OAVs; ratio of concentration to the respe...
0 downloads 0 Views 948KB Size
Download PDF
Article pubs.acs.org/JAFC

Characterization of Key Odorants Causing a Fusty/Musty Off-Flavor in Native Cold-Pressed Rapeseed Oil by Means of the Sensomics Approach Katrin Matheis and Michael Granvogl* Department für Chemie, Lehrstuhl für Lebensmittelchemie, Technische Universität München, Lise-Meitner-Straße 34, D-85354 Freising, Germany ABSTRACT: The sensomics approach was used to clarify the formation of the fusty/musty off-flavor of native cold-pressed rapeseed oil. A “positive control” (PC) showing the desired sensory attributes and an oil eliciting a fusty/musty off-flavor (OF) were analyzed. Comparative aroma extract dilution analysis (cAEDA), identification experiments, quantitation by stable isotope dilution assays (SIDAs), calculation of odor activity values (OAVs), and aroma recombination resulted in 11 odorants with an OAV ≥ 1 in PC. Main differences between both oils were obtained for compounds caused by microbial influence revealing significantly higher concentrations in OF, e.g., for ethyl 2-methylbutanoate, 2-methoxyphenol, 3-hydroxy-4,5-dimethylfuran2(5H)-one (sotolon), 2- and 3-methylbutanoic acid, and 4-methylphenol. Comparison of the key odorants in OF with those of the rapeseeds (OFS), from which it was pressed, showed the same 18 compounds proving that the grade of the seeds and their storage conditions are important criteria for the quality of the final oil. Finally, a further 7 native cold-pressed rapeseed oils, eliciting the same sensory defect, were analyzed to confirm aroma-active marker compounds responsible for the fusty/musty offflavor. KEYWORDS: rapeseed oil, off-flavor, aroma extract dilution analysis, stable isotope dilution analysis, odor activity value, aroma recombination, sensomics concept

■

INTRODUCTION

changes of volatiles in refined rapeseed oil stored at an elevated temperature (60 °C).6 Recently, Tynek et al.7 determined the total amount of the oxidation markers propanal, hexanal, and nonanal in six different varieties of cold-pressed rapeseed oils. The molecular sensory science concept was applied for the first time by Pollner and Schieberle8 showing 2-isopropyl-3methoxypyrazine, (E,E)-2,4-nonadienal, (E,Z)-2,6-nonadienal, 3-methylbutanoic acid, 3-methylbutanal, hexanal, and octanal as key odorants in native cold-pressed rapeseed oil from unpeeled rapeseeds. Very recently, mostly the same compounds were characterized by the sensomics approach in a commercial native cold-pressed rapeseed oil produced from unpeeled seeds revealing 2-isopropyl-3-methoxypyrazine, dimethyl trisulfide, dimethyl sulfide, octanal, butanoic acid, 2-isobutyl-3-methoxypyrazine, 2-sec-butyl-3-methoxypyrazine, (E,E)-2,4-nonadienal, hexanal, 3-methylbutanal, and 3-hydroxy-4,5-dimethylfuran2(5H)-one as important aroma compounds.9 Reported concentrations were also similar in both studies,8,9 but odor activity values (OAVs; ratio of concentration to the respective odor threshold) varied because of significant differences in the determined odor thresholds in oil. Although it is known that native cold-pressed rapeseed oils can elicit a fusty/musty off-flavor, up to now, no studies are available elucidating the compounds, which are responsible for its development. Thus, the aim of the present study was first to apply the sensomics concept on two native cold-pressed

In regard to the annual worldwide consumption of vegetable oils, rapeseed oil (27.2 million metric tons in 2014/15) was always third after palm oil (58.6) and soybean oil (48.0) during the last 10 years, based on data from the US Department of Agriculture.1 In the mid-1970s, the cultivation of the so-called “0-rape” represented the first step to launch rapeseed oil as edible oil by replacement of harmful erucic acid by oleic acid.2 Successful establishment of rapeseed oil for human and animal consumption was achieved ten years later due to lowered levels of glucosinolates resulting in the loss of bitterness (“00-rape”). Consequently, rapeseed oil gained more importance for human nutrition because of its positive physiological properties, which are correlated with their unsaturated fatty acids in the triglycerides, and worldwide consumption increased by 70% within the last 10 years.1,3 Besides health benefits and nutritional value, flavor attributes are also one of the most important criteria for consumers to buy a certain food. Thus, more knowledge about the aroma composition and the possibility to avoid off-flavor formation in rapeseed oils during processing or storage conditions will lead to a further increase of acceptance. However, until now, only some reports on volatiles, and especially on odorants, of rapeseed oil are available. In 1995, Snyder et al.4 reported about an increase of volatiles, including eight aldehydes and three alkanes in canola oil, corn oil, soybean oil, and sunflower oil, comparing the original state with that after 8 and 16 days of accelerated storage (60 °C). Jeleń et al.5 compared volatiles in refined and cold-pressed rapeseed oils investigating the influence of storage time and temperature on the composition of volatiles. Later, the same group analyzed the © 2016 American Chemical Society

Received: Revised: Accepted: Published: 8168

August 5, 2016 October 6, 2016 October 7, 2016 October 7, 2016 DOI: 10.1021/acs.jafc.6b03527 J. Agric. Food Chem. 2016, 64, 8168−8178

Article

Journal of Agricultural and Food Chemistry

[2H2]-γ-nonalactone;25 [2H2−4]-octanal;27 [2H2−4]-1-octen-3-one;28 and [2H5]-2-phenylethanol.29 Syntheses. [2H2−4]-2-Acetylpyridine. [2H2−4]-2-Acetylpyridine was synthesized according to Sabot et al.30 To a solution of 1,5,7triazabicyclo[4.4.0]dec-5-ene (TBD) (11.5 mg, 0.083 mmol) in deuterated chloroform (CDCl3; 3 mL), 2-acetylpyridine (0.1 g, 0.83 mmol) was added. The reaction mixture was stirred at room temperature for 12 h and the reaction was quenched with hydrochloric acid (1 mol/L; 1 mL). The organic layer was washed with water and brine (both 2 × 2 mL), dried over anhydrous sodium sulfate, and was, finally, filtered. A high vacuum distillation using solvent assisted flavor evaporation (SAFE) technique31 was performed to remove nonvolatile constituents, e.g., colorants. The incorporation of the labeling was checked by GC-MS: after 12 h, only 0.6% remained unlabeled, only 2.3% were [2H1]-labeled, 21.6% [2H2]-labeled, 69.0% [2H3]-labeled, and 6.1% [2H4]-labeled. RI (DB-FFAP) = 1554, RI (OV-1701) = 1117, RI (DB-5) = 1031. MS (EI), m/z (%): 96 (100), 78 (93), 80 (82), 124 (75, M+), 51 (72), 46 (45), 52 (44), 50 (38), 95 (27), 123 (20), 79 (15), 45 (12), 125 (10), 106 (8). MS (CI, methanol), m/z (%): 125 (100, [M+H]+), 124 (26), 126 (8). [4,5- 2 H 2 ]-trans-4,5-Epoxy-(E)-2-decenal. [4,5- 2 H2 ]-trans-4,5Epoxy-(E)-2-decenal was synthesized according to Lin et al.32 with a slightly modified cleanup by column chromatography using a diol phase (40−63 μm) filled in a glass column (20 cm × 2 cm i.d.). The crude product was dissolved in pentane, applied onto the column and elution was carried out with pentane (150 mL) and a mixture of pentane/diethyl ether (95/5, v/v; 150 mL) collecting 15 fractions (each 20 mL) until no odor impression was perceived. The target compound was found in fractions 11 to 14. RI (DB-FFAP) = 1997, RI (OV-1701) = 1553, RI (DB-5) = 1382. MS (NCI, isobutane), m/z (%): 99 (100), 168 ([M−]; 45). Isolation of Volatiles. PC and OF (each 90 g) were diluted with diethyl ether (each 20 mL) and the volatiles were separated from the nonvolatiles by means of thin-film distillation,33 followed by high vacuum distillation using solvent assisted flavor evaporation (SAFE) technique31 as recently described.9 The concentrate (100 μL) was used for comparative aroma extract dilution analysis (cAEDA). OFS (215 g) were frozen with liquid nitrogen and ground to a fine powder by a commercial blender. Diethyl ether (2 × 400 mL) was added and the powder was extracted by stirring for 2 × 90 min at room temperature. After filtration and SAFE distillation, the workup procedure was continued as described for the oils.9 For identification, the distillates were fractionated as reported recently.9 Comparative Aroma Extract Dilution Analysis (cAEDA) and Identification Experiments. cAEDA as a screening method was applied to PC, OF, and OFS to get a first idea which odorants should be important for the overall aroma. To enable direct comparability, the 2.4-fold amount of OFS compared to the oils was used, due to the fat content of the seeds (42%). In addition, the samples were concentrated to the same final volume. The concentrated aroma extracts were diluted stepwise 1 + 1 (v + v) with diethyl ether. In this way, flavor dilution (FD) factors were determined. The original distillates (PC, OF, and OFS) as well as each dilution were analyzed via high-resolution gas chromatography-olfactometry (HRGC-O) until no odor impression was detected at the sniffing port. The samples were analyzed by three experienced panelists leading to the so-called FD factors (highest dilution, in which the odorant was last detected at the sniffing port) for each aroma-active compound. Odorants with FD factors ≥8 were identified as recently described.9 High-Resolution Gas Chromatgraphy-Olfactometry (HRGCO). HRGC-O and determination of linear retention indices (RIs)34 for each odorant were performed as recently described.9 Comparative Static Headspace Aroma Dilution Analysis (cHS-ADA) and Static Headspace High-Resolution Gas Chromatography−Mass Spectrometry (HS-HRGC-MS). Especially for highly volatile aroma-active compounds, which often coelute with the solvent by application of liquid injection technique, headspace

rapeseed oils, one showing the desired sensory attributes (“positive control”; PC) and the other evoking a fusty/musty off-flavor (OF), to characterize the aroma-active compounds, which are responsible for the off-flavor. In addition, the seeds (OFS), from which OF was pressed, were investigated to check the influence of the raw material as well as of the pressing process on the off-flavor formation.

■

MATERIALS AND METHODS

Oil Samples. The native cold-pressed rapeseed oil showing the desired sensory attributes was purchased from a commercial supplier. The oil was pressed from unpeeled rapeseeds at a maximum of 28 °C and was only sedimented and, finally, decanted. The off-flavor rapeseed oil was manually pressed by Max Rubner-Institut (Detmold, Germany) from rapeseeds provided by an oil mill. Chemicals. The following reference compounds for identification and quantitation experiments were obtained from commercial sources: acetaldehyde, acetylpyrazine, 2-aminoacetophenone, 2-sec-butyl-3methoxypyrazine, citral (mixture of geranial and neral), (E,E)-2,4decadienal, γ-decalactone, dimethyl sulfide, ethyl 3,5-dimethylpyrazine, ethyl 3,6-dimethylpyrazine, ethyl 2-methylbutanoate, hexanal, (Z)-3hexen1-ol, 3-hydroxy-4,5-dimethylfuran-2(5H)-one, indole, 2-isobutyl3-methoxypyrazine, 2-isopropyl-3-methoxypyrazine, linalool, 2-methoxyphenol, 3-methylbutanal, 2-methylbutanoic acid, 3-methylbutanoic acid, methyl 2-methylbutanoate, 4-methylphenol, methylpropanal, (E,E)-2,4-nonadienal,, (E,Z)-2,6-nonadienal, γ-nonalactone, δ-nonalactone, (E)-2-nonenal, γ-octalactone, octanal, (E)-2-octenal, 1octen-3-ol, phenylacetic acid, 2-phenylethanol (Aldrich; Sigma-Aldrich, Taufkirchen, Germany); 2-methoxy-4-vinylguaiacol, (Z)-6-nonenal, δoctalactone, 1-octen-3-one, pentanoic acid, 2-propionylthiazole (Alfa Aesar, Karlsruhe, Germany); 2-acetylpyridine, butanoic acid, (E)-2decenal, heptanoic acid, 4-hydroxy-2,5-dimethylfuran-3(2H)-one (Fluka, Sigma-Aldrich); 2-methylbutanal, 4-vinylphenol (Lancaster, Mühlheim/Main, Germany); dimethyl trisulfide (Acros Organics, Geel, Belgium); 1-octanol (Carl Roth, Karlsruhe, Germany); 3hydroxy-4-methoxybenzaldehyde, myrcene (Sigma-Aldrich); and (E)β-damascenone (kindly provided by Symrise, Holzminden, Germany). (Z)-1,5-Octadien-3-one10 and trans-4,5-epoxy-(E)-2-decenal11 were synthesized according to the literature. The following chemicals were commercially obtained: nitrogen (liquid) (Linde, Munich, Germany); acetone, dichloromethane, diethyl ether, ethanol, hydrochloric acid, LiChroprep Diol (40−63 μm), silica gel 60, sodium chloride, sodium hydrogen sulfate, anhydrous sodium sulfate, sulfuric acid (Merck, Darmstadt, Germany); argon, helium, hydrogen, nitrogen (Westfalen, Münster, Germany); nalkanes (C5−C26), chloroform-d, lithium aluminum deuteride, 2octyn-1-ol, 1,5,7-triazabicyclo[4.4.0]dec-5-ene (Sigma-Aldrich); (formylmethylene)triphenylphosphorane and manganese(IV) oxide (Alfa Aesar). Dichloromethane and diethyl ether were of p.a. grade and freshly distilled prior to use. All chemicals were at least of analytical grade. Stable Isotopically Labeled Standards. [2H6]-Dimethyl sulfide (Sigma-Aldrich), [2H7]-4-methylphenol (C/D/N Isotopes, Quebec, Canada), [2H9]-2-methylbutanoic acid (EQ Laboratories, Augsburg, Germany), and [2H6]-myrcene (Santa Cruz Biotechnology, Dallas, TX) were commercially obtained. The following stable isotopically labeled compounds were synthesized as previously described: [2H3]acetylpyrazine;12 [2H3]-2-aminoacetophenone; 13 [2H2]-butanoic acid;14 [2H3]-2-sec-butyl-3-methoxypyrazine analogue to 2-isobutyl-3methoxypyrazine;15 [2H2−4]-(E,E)-2,4-decadienal;16 [2H6]-dimethyl trisulfide;17 [2H3]-2-ethyl-3,5-dimethylpyrazine;18 [2H3]-ethyl 2-methylbutanoate;19 [2H2−4]-hexanal;20 [13C2]-3-hydroxy-4,5-dimethylfuran2(5H)-one; 21 [ 13 C 2 ]-4-hydroxy-2,5-dimethylfuran-3(2H)-one; 22 [ 2 H 3 ]-2-isobutyl-3-methoxypyrazine and [ 2 H 3 ]-2-isopropyl-3methoxypyrazine;15 [2H3]-2-methoxyphenol;18 [2H2]-2-methylbutanal and [2H2]-3-methylbutanal;23 [2H2]-3-methylbutanoic acid;24 [2H2](E,E)-2,4-nonadienal;20 [2H2]-(E,Z)-2,6-nonadienal;16 [2H2]-γ-nonalactone;25 [2H2]-(E)-2-nonenal;26 [2H2]-γ-octalactone analogue to 8169

DOI: 10.1021/acs.jafc.6b03527 J. Agric. Food Chem. 2016, 64, 8168−8178

Article

Journal of Agricultural and Food Chemistry

Table 1. Stable Isotopically Labeled Standards, Selected Ions (m/z), Response Factors (Rf), and Instruments including Capillary Column(s) Used in the Stable Isotope Dilution Assays ion (m/z)a compound acetylpyrazine 2-acetylpyridine 2-aminoacetophenone butanoic acid 2-sec-butyl-3-methoxypyrazine (E,E)-2,4-decadienal dimethyl sulfide dimethyl trisulfide trans-4,5-epoxy-(E)-2-decenal 2-ethyl-3,5-dimethylpyrazine 2-ethyl-3,6-dimethylpyrazinee ethyl 2-methylbutanoate hexanal 3-hydroxy-4,5-dimethylfuran-2(5H)-one 4-hydroxy-2,5-dimethylfuran-3(2H)-one 2-isobutyl-3-methoxypyrazine 2-isopropyl-3-methoxypyrazine 2-methoxyphenol 2-methylbutanal 3-methylbutanal 2-methylbutanoic acid 3-methylbutanoic acid 4-methylphenol myrcene (E,E)-2,4-nonadienal (E,Z)-2,6-nonadienal γ-nonalactone (E)-2-nonenal γ-octalacton octanal 1-octen-3-one 2-phenylethanol

isotope label 2

[ H3] [2H2−4]d [2H3] [2H2] [2H3] [2H2−4]d [2H6] [2H6] [2H2] [2H3] -e [2H3] [2H2−4]d [13C2] [13C2] [2H3] [2H3] [2H3] [2H2] [2H2] [2H9] [2H2] [2H7] [2H6] [2H2] [2H2] [2H2] [2H2] [2H2] [2H2−4]d [2H2−4]d [2H5]

analyte

internal standard

Rfb

systemc

capillary column(s)

123 122 135 89 167 153 63 127 97 137 137 131 101 129 129 167 153 125 87 87 103 103 109 137 139 139 157 141 143 129 127 105

126 124−126d 138 91 170 155−157d 69 133 99 140 140e 134 103−105d 131 131 170 156 128 89 89 112 105 116 143 141 141 159 143 145 131−133d 129−131d 110

0.69 0.97 0.87 1.00 0.65 0.97 0.71 0.95 0.99 0.76 0.94 0.83 1.00 1.00 1.00 0.87 0.96 1.00 0.96 0.89 0.69 1.00 0.75 1.00 0.91 0.89 0.71 0.75 0.66 0.88 0.61 0.70

II II II I II II III II IV II II II I II II II II II II II II II II II II II I II I II II I

DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-5 DB-FFAP/OV-1701 DB-FFAP DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/BGB-174E DB-FFAP/BGB-174E DB-FFAP/BGB-176 DB-FFAP/BGB-176 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP DB-FFAP/OV-1701 DB-FFAP DB-FFAP/OV-1701 DB-FFAP/OV-1701 DB-FFAP

a

Ions used for quantitation. bResponse factor determined by analyzing defined mixtures of analyte and internal standard. cSystem used for quantitation: I: GC-MS, II: GC/GC-MS, III: HS-GC-MS, IV: GC-MS (NCI). dInternal standard was used as a mixture of isotopologues. e2-Ethyl3,6-dimethylpyrazine was quantitated using [2H3]-2-ethyl-3,5-dimethylpyrazine as internal standard. technique was performed as previously reported.35,36 Odor quality and intensity, retention index and mass spectra (EI mode at 70 eV and CI mode at 70 eV, methanol as reactant gas) obtained were compared to the data of the in-house database. Quantitation by Stable Isotope Dilution Analysis (SIDA). The stable isotopically labeled standards (0.5−5 μg; dissolved in dichloromethane or diethyl ether; amounts depending on the concentrations of the respective analytes determined in a preliminary experiment) were added either to PC or OF (5−100 g) and the mixture was stirred for 15 min at room temperature for equilibration. Afterward, workup procedure was performed as described above for the isolation of volatiles prior to HRGC-MS or HRGC/HRGC-MS, respectively. OFS (5−50 g) were frozen with liquid nitrogen and ground to a fine powder by a commercial blender. Diethyl ether (25−100 mL; twice) and aliquots of the internal standards (0.5−5 μg) were added. The mixture was stirred for 2 × 90 min at room temperature for equilibration, filtrated, and workup was continued as described above. Dimethyl sulfide was quantitated by static HS-HRGC-MS (CI mode). PC, OF (1 g each), or OFS (2.5 g) were weighed into a headspace vial (20 mL) and [2H6]-dimethyl sulfide (0.2 μg; dissolved in refined sunflower oil) was added. After equilibration (1 h at 30 °C), an aliquot (1 mL) of the headspace above the oil was injected by a gastight syringe.

To obtain the response factors (Rf) of each odorant, defined binary mixtures of analytes and standards in five different ratios (5 + 1, 3 + 1, 1 + 1, 1 + 3, 1 + 5) were analyzed (Table 1). Quantitation was performed either by HRGC-MS, a two-dimensional setup (HRGC/ HRGC-MS), or HS-HRGC-MS depending on the amount and the volatility of the odorant. Calculation of Concentrations in OFS. Concentrations of aromaactive compounds in OFS were calculated on the basis of their fat content to be able to compare the results with those obtained for PC and OF. Therefore, the fat content in OFS was determined by acidic hydrolysis according to Weibull-Stoldt prior to Soxhlet extraction revealing a fat content of 42%. Based on the hypothesis that all odorants are, due to mostly lipophilic properties, in the fat of the rapeseeds, a factor of 2.4 has to be applied in comparison to the oils. High-Resolution Gas Chromatography−Mass Spectrometry (HRGC-MS). HRGC-MS was performed as recently described.35 The peak areas of the analyte and labeled standard were determined separately by using the respective mass traces of the protonated molecular masses or selected fragments (Table 1). Two-Dimensional High-Resolution Gas Chromatography− Mass Spectrometry (HRGC/HRGC-MS). In case of an overlapping of an analyte by major volatiles, a two-dimensional setup was applied. HRGC/HRGC-MS was performed as recently described35 applying the column combinations summarized in Table 1. For separation and quantitation of (R)-2-, (S)-2- and 3-methylbutanal (BGB-174E 8170

DOI: 10.1021/acs.jafc.6b03527 J. Agric. Food Chem. 2016, 64, 8168−8178

Article

Journal of Agricultural and Food Chemistry column) as well as (R)-2-, (S)-2-, and 3-methylbutanoic acid (BGB176 column; both 30 m × 0.25 mm i.d., 0.25 μm film thickness; BGB Analytik, Böckten, Switzerland) chiral columns were applied in the second dimension.37 High-Resolution Gas Chromatography−Mass Spectrometry in Negative Chemical Ionization (NCI) Mode. For HRGC-MS (NCI), a gas chromatograph 7890B GC equipped with a DB-FFAP capillary column (30 m × 0.32 mm i.d., 0.25 μm film thickness; J&W Scientific; Agilent Technologies, Waldbronn, Germany) and a cold oncolumn injector was coupled to a mass spectrometer 240 Ion Trap GC-MS (both Agilent Technologies). Mass spectra were recorded in NCI mode by low-energetic, thermal secondary electrons of 1 eV generated by collision of primary electrons (70 eV) with the reactant gas isobutane. Aroma Profile Analysis (APA). For APA, the intensities of selected odor attributes describing the overall aroma of both oils and the rapeseeds were rated on a seven-point linear scale from 0 (not perceivable) to 3 (strongly perceivable) by the sensory panel consisting of at least 20 experienced assessors, participating in weekly sensory sessions intended to train their abilities to recognize and describe different aroma qualities. Sensory analyses were performed in a sensory room equipped with single booths at 21 ± 1 °C. The samples (5 g) were presented in covered glass vessels (40 mm i.d., total volume = 45 mL). Determination of Orthonasal Odor Thresholds (OTs). OTs were determined following the procedure recently described.9,38 Aroma Recombination. For aroma recombination of PC and OF, odorless refined rapeseed oil was used. For OFS, the seeds (50 g) were frozen with liquid nitrogen, ground with a commercial blender, and stepwise extracted with methanol, diethyl ether, hexane, and pentane (each 300 mL), filtered, and finally dried, resulting in an odorless powder. Subsequently, the seeds were dried to remove solvent residues and spiked with the natural amount of fat content (42%) using refined rapeseed oil. All quantitated aroma compounds with an OAV ≥ 1 were prepared in ethanolic solutions and were added to the refined rapeseed oil (for PC and OF recombinants) or to the deodorized powder (42% of fat) in the naturally occurring concentrations. Thereby, the amount of ethanol did not exceed its odor threshold in oil (850 μg/kg). The recombinants and the original native cold-pressed rapeseed oils (PC and OF) as well as the corresponding rapeseeds (OFS) were evaluated by the sensory panel as explained above for APA.

Figure 1. Aroma profile analyses of PC (solid line) and OF (broken line).

in-house database containing >1000 aroma-active reference volatiles. Odorants with the highest FD factors were identified as 2-isopropyl-3-methoxypyrazine (18; earthy, pea-like), 2isobutyl-3-methoxypyrazine (22; bell pepper-like), 2-aminoacetophenone (55; foxy; all 2048), octanal (10; citrus-like, green), 2-propionylthiazole (34; roasty), trans-4,5-epoxy-(E)-2decenal (46; metallic; all 512), 1-octen-3-one (11; mushroomlike), dimethyl trisulfide (12; cabbage-like), and γ-decalactone (52; peach-like, coconut-like; all 128) (Table 2). Dimethyl sulfide (2; asparagus-like) was identified with the highest FD factor during headspace ADA (Table 3).9 Identification of Key Aroma Compounds in OF and OFS. Again, both distillates elicited the typical overall aroma of either the original oil or the corresponding seeds when they were evaluated on a strip of filter paper. In OF, 42 aroma-active areas were detected with FD factors between 8 and 2048. The highest FD factor of 2048 was obtained for 22 (bell pepperlike), 42 (gammon-like, smoky), and 43 (flowery, honey-like), followed by 18 (earthy, pea-like) and 21 (pea-like, roasty) with an FD factor of 1024 (Table 2). Comparison of the FD chromatograms of PC and OF already provided a first indication of differences, illustrating that compounds no. 7, 21, 28, 31, 32, 33, 42, 43, 50, and 56 showed clearly increased FD factors in OF (Figure 2). Also in HS-ADA, differences were found between PC and OF. PC showed the highest FD factor of 8 for 2 (asparagus-like), whereas 3 (malty) had the highest FD factor of 4 in OF (Table 3). Applying the above-mentioned procedure of the sensomics concept,39 the following compounds could be unequivocally identified in OF: 3-methylbutanal (4; malty), 2-methylbutanal (5; malty), ethyl 2-methylbutanoate (7; fruity), octanal (10; citrus-like, green), dimethyl trisulfide (12; cabbage-like), 2isopropyl-3-methoxypyrazine (18), 2-ethyl-3,5-dimethylpyrazine (19; earthy, pea-like), 2-ethyl-3,6-dimethylpyrazine (20; earthy, pea-like), 2-sec-butyl-3-methoxypyrazine (21), 2-isobutyl-3-methoxypyrazine (22), 2-acetylpyridine (25; nutty, roasty), 2-methylbutanoic acid (31), 3-methylbutanoic acid (32), (E,E)-2,4-nonadienal (33; fatty, green), 2-methoxyphenol (42), 2-phenylethanol (43), 4-methylphenol (50; fecal), 3hydroxy-4,5-dimethylfuran-2(5H)-one (54; seasoning-like), and 2-aminoacetophenone (55; foxy) (Table 2). In contrast to the significant differences between PC and OF, the comparison between OF and the corresponding rapeseeds OFS, from which OF was pressed, showed only small differences in FD factors confirming the sensory results of the APA (Figure 3). For example, only FD factors of 17, 25, 43,

■

RESULTS AND DISCUSSION To get a first impression of the differences in the overall aroma between the native cold-pressed rapeseed oil with desired sensory attributes (“positive control”; PC) and the native coldpressed rapeseed oil eliciting the fusty/musty off-flavor (OF), aroma profile analyses were performed. Both oils showed a similar intensity for the earthy/pea-like attribute, whereas the cabbage-like odor impression was dominating only in PC. Additionally, a nutty/fatty impression was evaluated as a characteristic odor of PC. In contrast, a seasoning-like aroma was predominating in OF beside malty and sweaty impressions (Figure 1). Identification of Key Aroma Compounds in PC. Volatiles were extracted by thin-film distillation and for efficient removal of oil residues, additional, a SAFE distillation31 was applied. On a strip of filter paper, the concentrated distillate showed the typical aroma of the original oil. Afterward, the distillate obtained was subjected to AEDA using HRGC-O. As previously reported, 53 odor-active areas were detected with FD factors ranging between 8 and 2048.9 According to the sensomics concept,39 retention indices on two GC columns of different polarity as well as the aroma quality and intensity of the odorants in the distillate were determined. For an unequivocal identification, mass spectra (EI and CI mode) were recorded. All results were compared to data available in an 8171

DOI: 10.1021/acs.jafc.6b03527 J. Agric. Food Chem. 2016, 64, 8168−8178

Article

Journal of Agricultural and Food Chemistry

Table 2. Comparison of FD Factors of the Most Aroma-Active Compounds in Native Cold-Pressed Rapeseed Oils (PC and OF) and the Correspoding Rapeseeds (OFS), of which OF was Pressed RId no.

a

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

compound

b

3-methylbutanal 2-methylbutanal methyl 2-methylbutanoate ethyl 2-methylbutanoate hexanal myrcene octanal 1-octen-3-one dimethyl trisulfide (Z)-1,5-octadien-3-one (Z)-3-hexen-1-ol (E)-2-octenal 1-octen-3-ol (Z)-6-nonenal 2-isopropyl-3-methoxypyrazine 2-ethyl-3,5-dimethylpyrazine 2-ethyl-3,6-dimethylpyrazine 2-sec-butyl-3-methoxypyrazine 2-isobutyl-3-methoxypyrazine (E)-2-nonenal linalool 2-acetylpyridine 1-octanol (E,Z)-2,6-nonadienal acetylpyrazine butanoic acid (E)-2-decenal 2-methylbutanoic acid 3-methylbutanoic acid (E,E)-2,4-nonadienal 2-propionylthiazole geranial pentanoic acid (E,Z)-2,4-decadienal unknown (E,E)-2,4-decadienal (E)-β-damascenone trans-2,3-epoxyundecanal 2-methoxyphenol 2-phenylethanol γ-octalactone heptanoic acid trans-4,5-epoxy-(E)-2-decenal γ-nonalactone 4-hydroxy-2,5-dimethylfuran-3(2H)-one δ-nonalactone 4-methylphenol unknown γ-decalactone 2-methoxy-4-vinylphenol 3-hydroxy-4,5-dimethylfuran-2(5H)-one 2-aminoacetophenone 3-propylphenol 4-vinylphenol indole phenylacetic acid vanillin

odor quality

c

malty malty fruity fruity green hop-like, geranium-like citrus-like, green mushroom-like cabbage-like hop-like, geranium-like fatty, cucumber-like nutty, roasty mushroom-like citrus-like, earthy, pea-like earthy earthy pea-like, roasty bell pepper-like cucumber-like citrus-like nutty, roasty citrus-like cucumber-like popcorn-like sweaty fatty, nutty sweaty sweaty fatty, green roasty citrus-like sweaty deep-fried pungent deep-fried fruity citrus-like, metallic gammon-like, smoky flowery, honey-like coconut-like sweaty metallic coconut-like caramel-like coconut-like fecal phenolic peach-like, coconut-like smoky, clove-like seasoning-like foxy smoky, leather-like phenolic fecal honey-like, beeswax-like vanilla-like

8172

FD factorse

DB-FFAP

DB-5

PC

OF

OFS

941 943 1001 1013 1020 1137 1291 1299 1367 1377 1390 1394 1402 1423 1428 1457 1459 1491 1516 1527 1541 1550 1564 1577 1624 1636 1643 1665 1667 1686 1705 1720 1728 1740 1770 1795 1815 1843 1865 1916 1922 1949 2008 2031 2047 2075 2078 2097 2154 2183 2206 2229 2284 2368 2472 2533 2540

652 652 755 847 802 986 1009 980 968 983 858 1053 980 nd 1095 1078 1078 1164 1095 1147 1105 nd 1073 nd 1022 821 1261 860 860 1223 nd 1272 nd 1223 nd 1318 nd 1245 1086 1116 1264 nd 1382 1350 1069 1366 1077 nd 1476 1318 1109 1300 nd 1237 1287 1261 1406

32 32 8