Article pubs.acs.org/JAFC
Effects of Distillation System and Yeast Strain on the Aroma Profile of Albariño (Vitis vinifera L.) Grape Pomace Spirits Y. Arrieta-Garay,† P. Blanco,‡ C. López-Vázquez,‡ J. J. Rodríguez-Bencomo,† J. R. Pérez-Correa,§ F. López,*,† and I. Orriols‡ †
Departament d’Enginyeria Química, Facultat d’Enologia, Universitat Rovira i Virgili, Avinguda dels Països Catalans 26, Campus Sescelades, 43007 Tarragona, Spain ‡ Estación de Viticultura y Enología de Galicia, Instituto Galego da Calidade Alimentaria, Ponte San Clodio, 32427 Leiro, Spain § ASIS, Departamento de Ingeniería Química y Bioprocesos, Pontificia Universidad Católica de Chile, Avenida Vicuña Mackenna 4860, Casilla 306, Santiago 22, Chile ABSTRACT: Orujo is a traditional alcoholic beverage produced in Galicia (northwest Spain) from distillation of grape pomace, a byproduct of the winemaking industry. In this study, the effect of the distillation system (copper charentais alembic versus packed column) and the yeast strain (native yeast L1 versus commercial yeast L2) on the chemical and sensory characteristics of orujo obtained from Albariño (Vitis vinifera L.) grape pomace has been analyzed. Principal component analysis, with two components explaining 74% of the variance, is able to clearly differentiate the distillates according to distillation system and yeast strain. Principal component 1, mainly defined by C6−C12 esters, isoamyl octanoate, and methanol, differentiates L1 from L2 distillates. In turn, principal component 2, mainly defined by linear alcohols, linalool, and 1-hexenol, differentiates alembic from packed column distillates. In addition, an aroma descriptive test reveals that the distillate obtained with a packed column from a pomace fermented with L1 presented the highest positive general impression, which is associated with the highest fruity and smallest solvent aroma scores. Moreover, chemical analysis shows that use of a packed column increases average ethanol recovery by 12%, increases the concentration of C6−C12 esters by 25%, and reduces the concentration of higher alcohols by 21%. In turn, L2 yeast obtained lower scores in the alembic distillates aroma profile. In addition, with L1, 9% higher ethanol yields were achieved, and L2 distillates contained 34%−40% more methanol than L1 distillates. KEYWORDS: S. cerevisiae strains, spirits, packed column, alembic, chemical composition, sensory analysis
■
INTRODUCTION Orujo is a traditional alcoholic beverage produced in Galicia (northwest Spain) from distillation of grape pomace, a byproduct of the winemaking industry.1 The importance of this spirit is widely recognized; hence, European Regulation EC 110/082 of spirits drinks includes the geographical indication Orujo de Galicia, the only one from Spain.3 Different studies have been carried out evaluating the effect of pomace storage conditions,4−6 grape variety,7−11 yeast strain used during fermentation,12−17 and systems of distillation on the final characteristics of grape pomace distillates.18−22 Moreover, the distillation technique is a very important factor in the production of this alcoholic drink. In Galicia, the grape pomace distillate is produced in artisanal distilleries by use of copper charentais alembics and in industrial distilleries using steam entrainment. However, packed-column distillation systems are not commercially used and have not been studied before to produce orujo distillates. A packed column with partial condenser, designed and built in our group,23 has been applied before to produce pear and kiwi distillates.24,25 This system yielded better distillates than those obtained with traditional alembics. On the other hand, the inoculated yeast used to ferment the grape pomace can influence notably the production of volatile compounds and the aroma quality of the final distillate.15 Therefore, in this work, we applied two distillation systems (copper charentais alembic and packed © 2014 American Chemical Society
column) and two yeast strains (native and commercial) to study their influence on chemical and sensory characteristics of the obtained distillates.
■
MATERIALS AND METHODS
Grape Pomace and Fermentation. This study was carried out in the experimental distillery of the Estación de Viticultura e Enoloxı ́a de Galicia (EVEGA) using white grape pomace from Vitis vinifera cv. Albariño supplied by an industrial winery from Rı ́as Baixas (Galicia, Spain). The grape pomace was homogeneously distributed into 12 plastic containers of 50 kg and inoculated with two different yeast strains: six containers with Saccharomyces cerevisiae XG3 from EVEGA yeast collection (named L1) and the remaining six containers with the commercial active dry yeast (ADY) Uvaferm BDX from Lallemand (named L2). The native strain was chosen because it has been used before to produce a high organoleptic quality wine,26,27 whereas the commercial strain showed good results in the fermentation of other fruits, such as pear.28 S. cerevisiae L1 was grown in YEPD medium [1% (w/v) yeast extract, 2% (w/v) peptone, and 2% (w/v) glucose] at 28 °C for 24 h, and the cells were recovered by centrifugation, washed with sterile water, and added to six of the grape pomace containers at a final concentration of 106 cells/mL. L2 was rehydrated in 250 mL of Received: Revised: Accepted: Published: 10552
June 20, 2014 September 24, 2014 October 12, 2014 October 13, 2014 dx.doi.org/10.1021/jf502919n | J. Agric. Food Chem. 2014, 62, 10552−10560
Journal of Agricultural and Food Chemistry
Article
sugared water at 37 °C for 20 min and added to the remaining six containers at a concentration of 25 g of yeast/hL, as suggested by the manufacturer. In both cases, the yeast inoculum was mixed with marc and spread at the bottom, middle, and top of the containers to ensure its homogeneous distribution in the pomace. Then the containers were hermetically closed and the fermentation and storage were performed at room temperature until distillation. Distillation. Each 50 kg fermentation tank was divided in two equal portions, which were distilled separately by one method or the other. Six distillations per yeast were performed in both distillation systems. Alembic Distillations. Fermented Albariño grape pomace (22−23 kg) was distilled in a 50 L copper charentais alembic. The base of the boiler was heated by direct fire, with natural gas as a heating source, and tap water was used to cool the total condenser. The heating power was set to obtain an initial average distillation rate in the range 1.3−1.8 L/h. On the basis of sensorial analysis, distillation products were separated into three fractions: head, heart, and tail. The head fractions were defined as the first 220 mL; the following 2320 and 1740 mL were collected as heart fraction for L1 and L2 distillates, respectively; and the following 600 and 880 mL were collected as tail fraction for L1 and L2 distillates, respectively. Packed-Column Distillations. Fermented grape pomace of Albariño variety (22−23 kg) was distilled in a packed column with a 50 L copper boiler, previously described by Garcia-Llobodanin et al.23 The base of the boiler was heated by direct fire, and the heating and partial condenser cooling rates of the distillation column were adjusted to obtain a distillate flow rate of 1.03 L/h. The first 240 and 280 mL was collected as head fraction for L1 and L2 distillates, respectively; the heart fraction was defined as the following 2320 and 2290 mL for L1 and L2 distillates, respectively; and the tail fraction was defined as the following 210 and 310 mL for L1 and L2 distillates, respectively. Microbiological Control. Samples from grape pomace before fermentation and from each container before distillation were taken for yeast isolation. Twenty grams of marc was randomly collected from each container, dispersed in 100 mL of YEPD, and incubated in a rotary shaker at 28 °C for 2 h. Then the appropriate decimal dilutions in sterile water were spread on WL nutrient agar (Scharlau Microbiology, Barcelona, Spain) supplemented with 100 μg/mL chloramphenicol to inhibit bacterial growth. The plates were incubated at 28 °C until visible colonies appeared. After counting, 20 colonies from each sample were randomly selected, isolated as pure cultures on fresh YEPD plates, and stored in the same medium with 15% (v/v) glycerol at −70 °C for further characterization. All isolated colonies were plated on lysine agar medium (Scharlau Microbiology, Barcelona, Spain) to distinguish between Saccharomyces and non-Saccharomyces yeasts. Saccharomyces yeasts do not grow in this medium. Yeast isolates were characterized at species level by polymerase chain reaction (PCR) amplification of the 5.8S rRNA gene and the two ribosomal internal transcribed spacers (ITS) and subsequent restriction analysis with the endonucleases HinfI, HaeIII, and CfoI.29 PCR products and their corresponding restriction fragments were separated on 2% and 3% agarose gels in 1× Tris−acetate− ethylenediaminetetraacetic acid (TAE) buffer, respectively. After staining with ethidium bromide (0.5 μg/mL), the DNA pattern bands were visualized under UV light and photographed with a Gelprinter Plus system (TDI). Additionally, PCR amplification and sequencing of D1/D2 region of 26S rDNA gene were used to confirm yeast identification. The D1 and D2 domains were amplified by use of the primers NL-1 and NL-4 as described by Kurtzman and Robnett.30 PCR products were purified by use of the PCR Extract Mini kit (5PRIME) according to the supplier’s instructions. The same primers were used for sequencing with a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) and a 3130xl genetic analyzer (Applied Biosystems). Sequence similarities were obtained by use of GenBank BLASTN search facilities.31 The identification was considered as correct when the identity of gene sequences was 98% or higher. Those yeasts identified as S. cerevisiae were characterized at strain level by analysis of mitochondrial DNA restriction fragment length polymorphism (mtDNA-RFLP). Total DNA was obtained as
described by Querol et al.32 DNA samples were digested with the restriction endonuclease HinfI (Promega) and the fragments obtained were separated by electrophoresis on 0.8% (w/v) agarose gels and visualized as described above. Chemical Analysis of Distillates. Ethanol content (%, v/v) was measured by electronic densimetry according to the official method.33 Chemical composition of heart fractions was determined by gas chromatography coupled with flame ionization detector (GC-FID), with direct injection of the distillate, previously adjusted to 40% (v/v). Analyses were carried out on two different columns. Macroconstituents (methanol, higher alcohols, acetaldehyde, 1,1-diethoxyethane, ethyl acetate, ethyl lactate, 1-hexanol, isobutyraldehyde, ethyl formate, methyl acetate, 2-propenal, 2-butanol, and allylic alcohol) were determined on a capillary column CP WAX-57 CB (Varian) (50 m × 0.32 mm i.d. × 0.2 μm film thickness) on an Agilent 6890 chromatograph (Agilent Technologies, Waldbronn, Germany) equipped with a split/splitless injector following the methodology reported by López-Vázquez et al.9 Separation of the remaining compounds was done on a Supelcowax 10 capillary column (30 m × 0.32 mm × 0.25 μm film thickness; Supelco Inc., Bellefonte, PA) in a Varian CP3900 GC chromatograph as described by López-Vázquez et al.34 Samples were analyzed in triplicate. Sensory Analysis. Evaluation of the distillates was carried out in two steps: first a ranking preference test and then a descriptive analysis. All distillates (heart fractions) were diluted with Milli-Qtreated water (Millipore Corp., Bedford, MA) to an ethanol content of 40% (v/v). The samples were served at room temperature in Association Française de Normalisation (AFNOR) glasses labeled with random numbers. The first evaluation was carried out by a group of 10 females and 27 males, 18−60 years old (potential consumers) but not trained in specific evaluation of distillates. They were asked to rank by preference the distillates, evaluating separately aroma and taste of the spirits. The results were analyzed by the Friedman statistical test. The second evaluation was a descriptive sensory analysis performed by a trained panel (five people) with experience in distillates assessment. Previously, the most representative descriptive attributes of the distillates were defined in an initial tasting session according to the method described in the ISO 13299:2003 standard.35 The chosen attributes were aroma and taste descriptors. This test consisted of a preliminary round-table discussion, where the panelists defined the terms: 11 aroma descriptors (floral, fruity, vegetal/herbaceous, spicy, positive general impression, solvent, rancid, oxidized, pungent, burnt/ smoky, negative general impression) and eight taste descriptors (astringency, mellowness, sweet, harmony, alcoholic, persistence, prickling, and general quality in mouth). Then an individual smelling and tasting of the samples was carried out by the panelists in order to qualify the distillates. The intensities of the descriptors were rated on a discontinuous scale from 0 to 5. A score of 0 indicated that a descriptor was not perceived, and a score of 5 indicated the highest intensity. Statistical Analysis. One-way analysis of variance (ANOVA) was applied to the data obtained from the GC analysis. The aim was to ascertain whether the distillation system and/or yeast strain led to any significant differences (at 5% level) in volatile composition of the spirits. The variables that showed significant differences were used for principal component analysis (PCA). All statistical analyses were performed with the SPSS statistical package (version 17.0).
■
RESULTS AND DISCUSSION Microbiological Control. Albariño grape pomace presented a yeast population of 2.3 × 105 colony-forming units (cfu)/mL before inoculation and fermentation. This value was similar to that described by Bovo et al.14 for Italian marcs but lower than the counts reported in more recent studies.12,15 The distribution of yeast species is the following: Metschnikowia pulcherrima and Hanseniaspora uvarum were the predominant species found, at 43.5% and 52.2%, respectively, whereas S. 10553
dx.doi.org/10.1021/jf502919n | J. Agric. Food Chem. 2014, 62, 10552−10560
Journal of Agricultural and Food Chemistry
Article
Table 1. Alcohol Balances in Both Distillation Systems for Albariño Grape Pomace Spirita copper charentais alembic units % (v/v) % L p.a./100 kg of grape pomaceb g/100 L of real distillate g/100 kg of grape pomace g/100 L of real distillate mg/100 kg of grape pomace g/100 L of real distillate mg/100 kg of grape pomace g/100 L of real distillate mg/100 kg of grape pomace
L1 (n = 5)
packed column
L2 (n = 6)
L1 (n = 5)
Heart-Fraction Ethanol Alcoholic Strength 61.92 ± 4.50 a,a 50.68 ± 3.14 a,b Heart-Fraction Ethanol Yield 81.2 ± 8.33 a,a 65.66 ± 6.5 a,b 7.89 ± 0.59 a,a 5.87 ± 0.61 a,b Total Higher Alcohols Recovered 469.3 ± 48.3 a,a 367.2 ± 61.8 a,b 40.46 ± 2.9 a,a 32.8 ± 7.19 a,b Total Linear Alcohols Recovered 0.30 ± 0.75 a,a 0.23 ± 0.04 a,b 29.2 ± 10.1 a,a 20.55 ± 8.2 a,b Total Monoterpenols Recovered 1.35 ± 0.24 a,a 1.05 ± 0.16 a,b 131.7 ± 18.29 a,a 93.92 ± 13.04 a,b Total 1-Linalool Recovered 0.45 ± 0.04 a,a 0.39 ± 0.06 a,a 43.4 ± 2.26 a,a 35.2 ± 5.45 a,b
L2 (n = 6)
75.80 ± 3.86 b,a
75.2 ± 4.55 b,a
86.7 ± 2.91 a,a 8.28 ± 0.60 a,a
83.91 ± 2.6 b,a 8.93 ± 0.46 b,a
409.7 ± 51.4 a,a 37.4 ± 7.03 a,a
397.1 ± 60.4 a,a 42.2 ± 5.18 a,a
0.21 ± 0.04 b,a 19.97 ± 9.04 b,a
0.21 ± 0.03 a,a 22.33 ± 5.11 a,a
1.0 ± 0.12 b,a 100.2 ± 13.85 b,a
1.16 ± 0.20 a,b 124.13 ± 19.2 b,b
0.37 ± 0.02 b,a 37.6 ± 4.45 b,a
0.40 ± 0.04 a,a 42.9 ± 4.6 b,b
a Different letters before the comma in the same row indicate a significant difference (p ≤ 0.05) with respect to the distillation system (within the same yeast strain). Different letters after the comma in the same row indicate a significant difference (p ≤ 0.05) with respect to the yeast strain (within the same distillation system). bp.a.:pure alcohol.
Table 2. Macroconstituent Composition of Distillates (Heart Fraction) Obtained from Albariño Grape Pomace Fermented with Different Yeast Strainsa macroconstituent (g/hL p.a.) copper charentais alembic compd
packed column
L1 (n = 5)
L2 (n = 6)
L1 (n = 5)
L2 (n = 6)
432.87 ± 30.04 a,a 221.00 ± 58.60 a,a
579.99 ± 33.03 a,b 126.43 ± 63.41 a,b
438.15 ± 35.88 a,a 109.44 ± 91.5 b,a
616.15 ± 83.9 a,b 105.10 ± 78.7 a,a
acetaldehyde 1,1-diethoxyethane ∑ acetaldehyde + 1,1-diethoxyethane
31.96 ± 10.54 a,a 30.95 ± 14.05 a,a 62.92 ± 24.41 a,a
23.76 ± 5.84 a,a 12.84 ± 4.64 a,b 36.61 ± 10.47 a,b
21.73 ± 9.54 b,a 34.39 ± 22.06 a,a 56.12 ± 31.5 a,a
21.21 ± 4.74 a,a 30.86 ± 7.79 b,a 52.07 ± 11.82 b,a
1-propanol
63.26 ± 4.72 a,a
77.43 ± 4.93 a,b
57.24 ± 8.09 a,a
72.89 ± 5.96 a,b
methanol ethyl acetate
2-methyl-1-propanol 1-butanol 2-butanol allylic alcohol 2-methyl-1-butanol 3-methyl-1-butanol ∑ total higher alcoholsb ethyl lactate 1-hexanol isobutyraldehyde ethyl formate methyl acetate 2-propenal
134.56 1.06 0.33 0.11 95.75 282.90 577.98
± ± ± ± ± ± ±
4.01 a,a 1.27 a,a 0.02 a,a 0.05 a,a 4.89 a,a 12.19 a,a 18.75 a,a
152.03 0.14 0.32 0.03 84.93 244.5 559.3
± ± ± ± ± ± ±
23.15 a,a 0.04 a,a 0.04 a,a 0.02 a,b 22.92 a,a 71.44 a,a 117.5 a,a
104.09 0.34 0.24 0.05 72.02 218.64 452.62
± ± ± ± ± ± ±
25.01 b,a 0.13 b,a 0.06 a,a 0.05 b,a 14.79 b,a 36.89 b,a 91.2 a,a
129.50 0.60 0.26 0.03 69.57 200.4 473.2
± ± ± ± ± ± ±
28.0 a,a 1.01 a,a 0.06 a,a 0.02 a,a 18.1 a,a 53.2 a,a 101.3 a,a
33.28 7.47 0.79 2.17 2.15 0.62
± ± ± ± ± ±
7.89a,a 0.47 a,a 0.66 a,a 1.22 a,a 0.58 a,a 0.30 a,a
53.89 7.68 0.27 0.93 1.62 0.32
± ± ± ± ± ±
12.26a,b 1.10 a,a 0.09 a,b 0.36 a,b 0.76 a,a 0.10 a,a
19.74 5.58 0.27 1.36 1.15 0.61
± ± ± ± ± ±
9.44b,a 0.88 b,a 0.14 b,a 0.35 b,a 0.96 b,a 0.35 a,a
19.53 6.22 0.25 1.033 1.30 0.67
± ± ± ± ± ±
8.45b,a 0.84 b,a 0.08 a,a 0.33 a,a 0.86 a,a 0.16 b,a
a Different letters before the comma in the same row indicate a significant difference (p ≤ 0.05) with respect to the distillation system (within the same yeast strain). Different letters after the comma in the same row indicate a significant difference (p ≤ 0.05) with respect to the yeast strain (within the same distillation system). b∑ 1-propanol, 2-Methyl-1-propanol, 1-butanol, 2-butanol, allylic alcohol, 2-methyl-1-butanol, and 3-methyl-1butanol.
cerevisiae was not detected. An isolate of Pichia kluyveri was also identified at 4.3%. These results partially agree with those reported by Bovo et al.14 for Italian marcs. They found that Hanseniaspora sp. and M. pulcherrima were the main yeasts,
although their distribution was different, with dominance of the apiculate yeasts, and S. cerevisiae was present below the detection limits. Other yeast species have been reported in Prosecco marcs14 and marcs from Portugal.13 10554
dx.doi.org/10.1021/jf502919n | J. Agric. Food Chem. 2014, 62, 10552−10560
Journal of Agricultural and Food Chemistry
Article
Table 3. Microconstituent Ester Content of Distillates (Heart Fraction) Obtained from Albariño Grape Pomace Fermented with Different Yeast Strainsa esters (g/hL p.a.) copper charentais alembic compd isobutyl acetate butyl acetate isoamyl acetate hexyl acetate 2-phenylethyl acetate ∑ acetates of higher alcohols ethyl butyrate ethyl hexanoate ethyl octanoate ethyl decanoate ethyl dodecanoate ∑ ethyl esters C6−C12c
L1 (n = 5) 0.18
