Volatile Compounds of Dry Beans (Phaseolus vulgaris ... - Springer Link

6 downloads 0 Views 140KB Size Report
Oct 10, 2007 - Abstract Volatile compounds of uncooked dry bean. (Phaseolus vulgaris L.) cultivars representing three market classes (black, dark red kidney ...
Plant Foods Hum Nutr (2007) 62:177–183 DOI 10.1007/s11130-007-0059-3

ORIGINAL PAPER

Volatile Compounds of Dry Beans (Phaseolus vulgaris L.) B. Dave Oomah & Lisa S. Y. Liang & Parthiba Balasubramanian

Published online: 10 October 2007 # Springer Science + Business Media, LLC 2007

Abstract Volatile compounds of uncooked dry bean (Phaseolus vulgaris L.) cultivars representing three market classes (black, dark red kidney and pinto) grown in 2005 were isolated with headspace solid phase microextraction (HS-SPME), and analyzed with gas chromatography mass spectrometry (GC–MS). A total of 62 volatiles consisting of aromatic hydrocarbons, aldehydes, alkanes, alcohols and ketones represented on average 62, 38, 21, 12, and 9× 106 total area counts, respectively. Bean cultivars differed in abundance and profile of volatiles. The combination of 18 compounds comprising a common profile explained 79% of the variance among cultivars based on principal component analysis (PCA). The SPME technique proved to be a rapid and effective method for routine evaluation of dry bean volatile profile.

Keywords Beans (Phaseolus vulgaris L) . Headspace solid-phase microextraction . GC–MS . Principal component analysis . Cultivars . Volatiles . Pinto . Dark red kidneys . Black beans . Uncooked beans

B. D. Oomah (*) : L. S. Y. Liang National Bioproducts and Bioprocesses Program, Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, 4200 Highway 97, P.O. Box 5000, Summerland, British Columbia, Canada V0H 1Z0 e-mail: [email protected] P. Balasubramanian Sustainable Agriculture Program, Lethbridge Research Centre, Agriculture and Agri-Food Canada, 5403-1 Avenue South, P.O. Box 3000, Lethbridge, Alberta, Canada T1J 4B1

Introduction Dry beans are important crop in Canada, one of the five main producing (0.32 Mt in 2005–2006) countries and the third largest exporter of dry beans, accounting for 11% of world exports [1]. The low Canadian per capita dry bean consumption at 2.45 kg per annum in 2005 [2], may be due to the unsophisticated taste and flavor consumers generally associate with dry bean products despite their nutritional and health benefits. Flavor, an important factor in overall acceptability, and cooking time were two main characteristics used by consumers to select a given type of bean according to a survey conducted in Mexico [3, 4]. Studies have shown that hard to cook beans have sensorial differences (changes in flavor, color and texture) compared to fresh beans that make them unacceptable to consumers and processors. Flavor is most often ascertained on cooked or canned beans. For instance, cooked dark red kidney beans were rated superior in flavor and overall score (assessed by sensory evaluation) to pinto beans when investigated for quickcooking quality [5]. However, later studies of cooked bean from thirty-six cultivars showed that pinto and red kidney beans were most and least acceptable, respectively because of their predominant flavor characteristics [6]. Presently, flavor of canned beans is judged subjectively for flat, dull, bitterness, acid, sweet and off-flavor by a trained and experienced sensory evaluation team with scores ranging between 2.4 to 3.2 (2 = fair, 3 = good and 4 = very good) with high (24%) coefficient of variation for white bean [7]. Studies of flavor in dry beans [8–10] were preceded by identification of about 90 volatile components of canned whole beans as early as 1967 [11]. Twenty six volatile compounds have been identified in uncooked dry bean (red and white beans) seeds isolated by vacuum steam distillation

178

at 50 °C with continuous solvent (hexane) extraction [9]. 1-octen-3-ol (1.4 μg/kg) and (Z)-3-hexenol were presumed to be the most important aroma components of uncooked bean [9]. Considerable variation in unusual C8 unsaturated compounds was observed with different bean samples indicating a possible relation to variation in flavor. Twentythree compounds, mainly aliphatic and aromatic alcohols and aldehydes, have been identified in powdered dry beans [10]. Solid phase microextraction (SPME) is a simple, sensitive, robust, reliable, low cost and very popular fast screening sampling technique based on analyte diffusion that combines the advantages of both static and dynamic head space for qualitative volatile analysis [12]. Recently, 50 compounds were detected in French bean using headspace solid phase micro extraction (HS-SPME) with one of the most frequently used fiber divinylbenzene/carboxen/polydimethylsiloxane (DVB/ CAR/PDMS) [13]. SPME showed greater extraction efficiencies of alcohols and odor active compounds—hex-3-enol, 1-hexanol, oct-1-en-3-ol and 3-octanol. The DVB/CAR/ PDMS fiber generally gives high recoveries with the majority of analytes investigated [14]. The origin of flavor compounds found in dry bean seeds is not definitive; however, the presence of aliphatic saturated and unsaturated alcohols suggests enzymatic oxidation of bean lipids [15] similar to that shown for raw peas. Information currently available on the volatile components of dry beans is deficient and full investigations of the uncooked beans are long overdue. In the present study, HS-SPME with DVB/CAR/PDMS fiber was applied, as a solvent-free sample preparation method, with gas chromatography–mass spectrometry (GC–MS) analysis, to provide the initial investigation of the volatile profile of dry beans from Manitoba. This is the first step in unraveling and elucidating bean volatiles as a prerequisite in developing new cultivars for increased economic value and novel bean ingredients for the health and functional food sectors.

Materials and Methods Dry beans (Phaseolus vulgaris) grown and harvested in 2005 from the same field plot in Morden, southern Manitoba dry bean production area were obtained from Agriculture and Agri-Food Canada (Morden, MB, Canada). Three market class of beans; black (cv. AC Harblack), pinto (cv. CDC Rio, Onyx, AC Pintoba, Maverick) and dark red kidney (cv. ROG 802 and Redhawk) were used in this study. The beans were stored in a dry room (23 °C, 15–20% relative humidity) prior to analysis. Extraction by Solid Phase Microextraction (SPME) Dry bean seed sample (10 g) freshly ground in a small coffee mill was transferred into a 125 ml precleaned

Plant Foods Hum Nutr (2007) 62:177–183

Erlenmeyer flask hermetically capped with PTFE/silicon septum (Supelco, Inc., Bellefonte, PA, USA). Each sample was allowed to equilibrate in the flask at 50 °C for 1 h using a heating module (Heidolph MR 3002, Germany). These conditions were found to be appropriate for the sample headspace to equilibrate, as confirmed by three successive GC–MS analyses. Volatiles were extracted by exposing a 1 cm-50/30 μm divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) Stable Flex SPME fiber (Supelco, Bellefonte, PA) at 50 °C for 1 h. Prior to each extraction, the fiber was conditioned in the injection port of the GC as recommended by the supplier. This SPME method was adopted based on the selection of the most effective fiber in adsorbing a wide range of polar and nonpolar analytes and sampling conditions during preliminary investigations. Grinding and collection of headspace volatiles were performed in triplicate from the same lot for each sample. Gas Chromatography Mass Spectrometry (GC–MS) After extraction, the analytes were thermally desorbed at 250 °C for 2 min in the injection port of an Agilent 6890/ 5973 GC–MS (Agilent Technologies Wilmington, DE) and separated on a Supelcowax 10 polar column (Supelco, Bellefonte, PA), 60 m×0.25 mm with a 0.50 μm film thickness. The injection port was operated in a splitless mode subjected to an initial flow rate of 1.7 ml/min of ultrahigh-purity helium. The initial oven temperature of 40 °C was ramped at 1 °C/min to 70 °C, then 5 °C/min increase rate up to 200 °C, followed by 50 °C/min increase rate to 250 °C and held at that temperature for 15 min for a total run time of 74 min. Before sampling, the fiber was reconditioned for 15 min in the GC injection port at 250 °C, and a blank experiment was performed. The Agilent 5973 quadrupole mass spectrometer was operated in the electron ionization mode at 70 eV, a source temperature of 230 °C, quadrupole at 150 °C, in the scan range m/z 35 to 350. Data were collected with Agilent enhanced ChemStation software (standard MSD version) and searched against the NIST (v. 02) and Wiley (v. 138) libraries (Palisade Corp., Newfield, NY). Compounds were identified by preliminarily library search, and identities were confirmed by comparison of their GC retention time with eight internal standard solutions of C7–C22 n-alkanes and MS ion spectra. Levels of flavour components were determined from the average of three replicate chromatograms, calculated and expressed as the area units of their abundance (total area counts, TAC). Data were subjected to analysis of variance by the general linear models (GLM) procedure, means comparison by Duncan’s test, and Principal component analysis according to Statistical Analysis System [16].

Plant Foods Hum Nutr (2007) 62:177–183

Results and Discussion A total of 62 headspace volatile compounds isolated by SPME were tentatively identified in beans by GC–MS. Cultivars differed in relative abundance of extracted volatiles with low (190×106) (AC Pintoba, CDC Rio, Maverick and Onyx) abundance. The volatile profile of AC Harblack and ROG 802 consisted of only 32 and 25 compounds, respectively (Table 1), and in addition had the lowest content of 13 compounds. The low volatile compounds detected in AC Harblack and ROG 802 probably indicates suppression of their lipoxygenase enzyme, particularly the lipoxygenase 3 enzyme involved in aroma biosynthesis [17]. The volatile classes (Table 1) included 14 alkanes (4.6 to 42.6×106 TAC for AC Harblack and AC Pintoba, respectively), 11 aldehydes (14.4 to 55.8×106) and ten aromatic hydrocarbons (29.1 to 83.4×106 for AC Harblack and Onyx, respectively), ten alcohols (3.6 to 19.3×106 for AC Harblack and AC Pintoba, respectively), seven ketones (4.3 to 12.7×106 for ROG 802 and CDC Rio, respectively), three terpenes (0.8 to 14.7×106) and two furans (0.4 to 2.7×106 for AC Harblack and AC Pintoba, respectively). The aromatic hydrocarbons, aldehydes, alkanes, alcohols and ketones represented 62.2±23.7, 37.8±16.4, 21.1±15.2, 11.5±6.5 and 8.8±3.6×106 TAC, respectively. Styrene, the most abundant compound, varied significantly among bean cultivars. The high level of styrene despite its natural occurrence in beans, generally present at levels of less than 10 μg/kg [18], was alarming at first since no obvious source of contamination could be found. A bean sample designated as “organic” obtained from a local health food store produced results within the range of the original bean cultivars confirming the high styrene levels. Analysis of blank samples also ensured that the specific signals were originating only in the presence of bean samples. The detection of styrene, sometimes at extremely high concentrations, in food products with no apparent source of contamination has been reported previously [19]. A recent study of aroma compounds in a cashew apple-based alcoholic beverage also found high levels of styrene (27%) with no reasonable explanation [20]. Furthermore, artifact peaks are not normally observed with SPME [21]. Therefore, the high styrene level was assumed not to affect or compromise the analysis. Hexanal was the major aldehyde detected in beans, a result consistent with those observed in soybean and winged beans [22]. Significant differences were observed in other common aldehydes, benzaldehyde, nonanal and 2-heptenal levels among bean cultivars. The highest and lowest abundance of hexanal in Onyx and AC Harblack, respectively suggests major differences in their linoleate compositions since

179

production of hexanal depends on linoleate content [23]. Production of aliphatic unsaturated aldehydes, particularly 2hexenal, 2-heptenal, 2,4-heptadienal and 2-nonenal has been related to physical damage inflicted by blending raw dry been seeds in water [24]. Hexenal and heptenal are flavor compounds commonly identified in many fruits and vegetables. These aliphatic unsaturated aldehydes occurred at the lowest (1.4 and 1.3×106) for AC Harblack and ROG 802, respectively and highest levels in CDC Rio and Maverick (9.8 and 8.5×106, respectively). The high alkane, especially for cultivars AC Pintoba, Maverick, CDC Rio and Onyx (>20×106) may indicate the beginning of lipid oxidation thereby generating characteristic aroma of dry beans since alkanes are mostly derived from oxidative decomposition of lipids [25]. Most of the ten identified alcohols were also products of lipid oxidation. In dry beans, only two alcohols were related to specific aromas, e.g., 1-penten-3-ol originating from linoleate oxidation has a penetrating, grassy etheral odor [25] and 3-methyl-1-butanol with penetrating green aroma. Both these alcohols were most abundant in AC Pintoba. The flavor of alcohols in dry beans may be considered unimportant due to their relatively high threshold value [26]. Seven saturated ketones were present in dry beans (Table 1) with abundance of acetone, 6-methyl-5-hepten-2-one and 3,5octadien-2-one. Two furans were identified in the volatiles of dry beans, although as a chemical class their contribution to the overall odor may be negligible. Three terpenes, α-pinene, 3-carene, and limonene were detected in beans. CDC Rio, AC Pintoba and Maverick contained the most terpenes, while AC Pintoba had the highest α-pinene and limonene contents. The presence of limonene (1–4%) in vacuum steam volatile oil of dry red beans has been reported [9]. Compounds derived from lipids (2-butanone, 2-ethylfuran, pentanal, decane, toluene, hexanal, ethylbenzene, 1-penten-3-ol, 1-methylethylbenzene, o-xylene, heptanal, 1-pentanol, styrene, 1,3,5-trimethylbenzene, octanal, 2-heptenal, 1,2,3-trimethylbenzene, 6-methyl-5-hepten-2-one, 1-hexanol, nonanal, 1-octen-3-ol, 2,4-heptadienal, decanal, 3,5-octadien-2-one, benzaldehyde and 2-nonenal) predominated in all cultivars (50×106 for AC Harblack to 161×106 for Onyx). Abundant representatives in this category included decane, hexanal, 1-penten-3-ol, styrene, 2-heptenal, 6-methyl-5-hepten-2-one, nonanal, 3, 5-octadien-2-one and benzaldehyde. Statistically significant differences were observed among cultivars for several of the lipid-derived compounds including 1-penten-3-ol, 1-methylethylbenzene, O-xylene, 2-heptenal and 3, 5-octadien-2-one. The variation in the volatile compounds of bean cultivars may be related to their main fatty acid components. Hexanal, pentanal, 1-octen-3-ol has been reported as the lipoxygenase catalyzed oxidation products of polyunsaturated fatty acids [27].

180

Plant Foods Hum Nutr (2007) 62:177–183

Table 1 Volatile compounds identified in bean cultivars using SPME Cultivar Black Peak

Compounds

Pinto

Dark red kidney

AC Harblack

CDC Rio

Onyx

AC Pintoba

Maverick

ROG 802

Redhawk

Aromatic hydrocarbons 15 Toluene 19 Ethylbenzene 20 p-Xylene 24 1-Methylethylbenzene 25 o-Xylene 30 Propylbenzene 34 Styrene 35 1,3,5-Trimethylbenzene 38 α-Methylstyrene 39 1,2,3-Trimethylbenzene Subtotal

nd nd 0.89c 0.67e 0.69f nd 26.83c nd nd nd 29.08

1.83b 0.96a 2.44a 1.50bc 2.41b nd 69.02a 1.65b 0.69ab 1.15b 81.65

1.28b 0.86a 2.54a 1.70ab 1.91c 0.70a 72.57a 1.01c 0.86a nd 83.43

3.98a nd nd 1.85a 3.69a nd 65.79a 3.94a nd 2.21a 81.46

1.87b 0.86a 2.33a 1.64ab 2.22b 0.78a 61.62a 1.59b 0.75ab 1.05b 74.71

nd nd nd 1.15d 0.99e nd 32.35c nd nd nd 34.49

0.85b nd 1.35b 1.28cd 1.31d nd 44.57b Nd 0.57b nd 49.93

Aldehydes 11 Pentanal 17 Hexanal 26 Heptanal 32 2-Hexenal 36 Octanal 37 (E)-2-Heptenal 42 Nonanal 45 (E,E)-2,4-Heptadienal 48 Decanal 52 Benzaldehyde 53 (E)-2-Nonenal Subtotal

nd 8.44c nd nd nd 1.02e 1.82b 0.39c 0.59d 2.15c nd 14.41

1.26b 27.07b 1.59a 3.40a 1.33a 3.73b 4.48a 1.21a 3.39a 5.02a 1.44a 53.92

1.17b 33.26a 1.33ab 1.35b 0.98b 3.04c 5.39a 1.20a 1.62bc 5.06a 1.35a 55.75

1.70a 20.82b nd nd nd 4.30a 5.16a 1.27a nd 5.51a nd 38.76

1.17b 21.87b 1.61a 3.25a nd 2.99c 5.45a 1.12a 2.07b 5.06a 1.14a 45.73

0.76c 10.80c nd nd nd 1.28e 1.45b nd 0.49d 2.75c nd 17.53

1.19b 21.39b 1.10b 2.02b 0.79b 1.75d 4.14a 0.82b 1.14cd 3.99b nd 38.33

Alkanes 1 Hexane 2 Heptane 3 Octane 5 Nonane 7 2,6-Dimethyloctane 8 4-Methylnonane 10 3-Methylnonane 12 Decane 13 4-Methyldecane 16 Butylcyclohexane 18 Undecane 21 Octylcyclopropane 27 2,6,10-Trimethyl dodecane 29 Dodecane Subtotal

0.65c 0.33b 1.57d 0.51d nd nd nd 1.57d nd nd nd nd nd nd 4.63

3.27a 2.44a 3.69bc 2.02b 1.02b 2.00b 1.11b 9.07b 1.61b 1.05a 1.86b nd 1.68b 2.39a 33.21

1.49bc 1.09b 4.26a 1.51bc 0.81b 1.21c 0.69c 5.79c 0.88b nd 1.19c 1.93a nd nd 20.85

2.43ab 2.02a nd 2.93a 1.78a 3.50a 1.72a 13.96a 2.87a nd 3.22a 2.73a 2.72a 2.72a 42.60

3.35a 2.08a 3.39c 1.95b 0.91b 1.85b 1.10b 9.28b 1.26b 1.00a 2.07b nd 1.26b 2.37a 31.87

0.59c 0.52b 1.29d 0.53d nd nd nd 1.09d nd nd nd 1.39a nd nd 5.41

0.96bc 0.66b 4.19ab 1.35c nd nd nd 2.15d nd nd nd nd nd nd 9.31

Alcohols 23 1-Penten-3-ol 31 1-Pentanol 33 3-Methyl-1-butanol 41 1-Hexanol 43 1-Octen-3-ol 46 2-Ethyl-1-hexanol 54 1-Octanol 56 1-Nonanol

1.10e nd nd 1.10cd 0.70d 0.27c nd 0.25d

2.11cd nd 1.70a 3.64a 2.85ab 0.64b 0.71b 2.77a

3.43b 3.22a nd 3.16ab 2.86ab 0.58bc 0.49bc 1.64c

4.02a nd 1.84a 2.98b 2.62bc 1.42a 1.32a 1.96bc

2.60c 1.21b 1.14b 3.11b 3.12a 0.58bc 0.58bc 2.15b

1.80d nd nd 0.85d 2.39c nd nd nd

2.65c nd nd 1.43c nd 0.46bc 0.34c 0.50d

Plant Foods Hum Nutr (2007) 62:177–183

181

Table 1 (continued) Cultivar Black Peak

Compounds

Pinto

Dark red kidney

AC Harblack

CDC Rio

Onyx

AC Pintoba

Maverick

ROG 802

Redhawk

nd 0.20a 3.62

0.52b 1.01a 15.95

nd nd 15.38

2.38a 0.73a 19.27

0.51b 0.62a 15.62

nd nd 5.03

nd nd 5.38

Ketones 4 Acetone 6 2-Butanone 40 6-Methyl- 5-hepten-2-one 51 3,5-Octadien-2-one 55 1-(2-furanyl)-1-propanone 57 Acetophenone 60 Geranylacetone Subtotal

0.87e nd 1.45b 1.17d 0.39d 0.58d 0.41b 4.87

1.42bcd 0.89a 4.19a 2.64b 0.95b 1.35ab 1.29a 12.73

1.76b 0.71a 4.27a 2.27bc 0.96b 1.52a nd 11.49

2.39a nd 3.77a 2.80b nd 1.14bc nd 10.10

1.65bc 0.75a 1.84b 4.69a 1.52a 1.51a nd 11.96

1.19de nd 1.66b 0.97d 0.47cd nd nd 4.29

1.34cd nd 1.47b 1.68cd 0.62c 0.95c 0.35b 6.41

Monoterpenes 14 α-Pinene 22 Δ3-Carene 28 Limonene Subtotal

0.81e nd nd 0.81

4.87b 1.08a 2.89b 8.84

3.38d 0.98a 1.44c 5.80

8.36a nd 6.32a 14.68

4.10c 0.94a 2.56b 7.60

0.73e nd nd 0.73

0.87e nd nd 0.87

Furans 9 2-Ethylfuran 47 2-Propyl furan Subtotal

nd 0.42b 0.42

1.36ab 1.18a 2.54

1.11b 1.07a 2.18

1.59a 1.06a 2.65

1.21b 1.19a 2.40

nd 0.51b 0.51

0.67c 0.98a 1.65

Others 44 Bromoacetic acid decyl ester 49 3-Phenylindole 50 3-Hydroxymandelic acid ethyl ester 58 Methoxy-phenyl-oxime 59 2,4-Dimethylbenzenamine Subtotal

nd 0.50b 0.25c 0.52c nd 1.27

0.69a nd 0.60b 1.09a 0.45ab 2.83

0.72a 1.00b 0.61b 0.87b 0.52a 3.72

nd 2.52a 0.97a nd nd 3.49

0.63a 0.96b 0.66b 1.21a 0.50a 3.96

nd 0.49b nd nd nd 0.49

0.40b 0.84b 0.49bc 0.83b 0.41b 2.97

Total area X106

59

212

199

213

194

68

115

61 Benzyl alcohol 62 Phenol Subtotal

Peak number corresponds to elution order in chromatograms. Values are mean area units (divided by 106 ) from three replicate samples. Means with different letter(s) within a row for each volatile component of cultivars are significantly different at p