Food Chemistry 160 (2014) 266–280

Contents lists available at ScienceDirect

Food Chemistry

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

Analytical Methods

A powerful methodological approach combining headspace solid phasemicroextraction, mass spectrometry and multivariate analysis forprofiling the volatile metabolomic pattern of beer starting raw materials

http://dx.doi.org/10.1016/j.foodchem.2014.03.0650308-8146/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +351 291705112; fax: +351 291705149.E-mail address: jsc@uma.pt (J.S. Câmara).

João L. Gonçalves a, José A. Figueira a, Fátima P. Rodrigues a, Laura P. Ornelas b, Ricardo N. Branco b,Catarina L. Silva a, José S. Câmara a,⇑a CQM – Centro de Química da Madeira, Universidade da Madeira, Centro de Ciências Exactas e da Engenharia da Universidade da Madeira,Campus Universitário da Penteada, 9000-390 Funchal, Portugalb ECM – Empresa de Cervejas da Madeira, PEZO, Parque Empresarial Zona Oeste, 9304-003 Câmara de Lobos, Funchal, Portugal

a r t i c l e i n f o

Article history:Received 30 October 2012Received in revised form 9 March 2014Accepted 12 March 2014Available online 25 March 2014

Keywords:BeerRaw materialsGlobal fingerprintVolatile metabolitesSolid phase microextraction (SPME)Multivariate statistical analysis

a b s t r a c t

The volatile metabolomic patterns from different raw materials commonly used in beer production,namely barley, corn and hop-derived products – such as hop pellets, hop essential oil from Saaz varietyand tetra-hydro isomerized hop extract (tetra hop), were established using a suitable analytical proce-dure based on dynamic headspace solid-phase microextraction (HS-SPME) followed by thermal desorp-tion gas chromatography–quadrupole mass spectrometry detection (GC–qMS). Some SPME extractionparameters were optimized. The best results, in terms of maximum signal recorded and number of iso-lated metabolites, were obtained with a 50/30 lm DVB/CAR/PDMS coating fiber at 40 �C for 30 min. A setof 152 volatile metabolites comprising ketones (27), sesquiterpenes (26), monoterpenes (19), aliphaticesters (19), higher alcohols (15), aldehydes (11), furan compounds (11), aliphatic fatty acids (9), aliphatichydrocarbons (8), sulphur compounds (5) and nitrogen compounds (2) were positively identified. Eachraw material showed a specific volatile metabolomic profile. Monoterpenes in hop essential oil and corn,sesquiterpenes in hop pellets, ketones in tetra hop and aldehydes and sulphur compounds in barley werethe predominant chemical families in the targeted beer raw materials. b-Myrcene was the most dominantvolatile metabolite in hop essential oil, hop pellets and corn samples while, in barley, the predominantvolatile metabolites were dimethyl sulphide and 3-methylbutanal and, in tetra hop, 6-methyl-2-penta-none and 4-methyl-2-pentanone. Principal component analysis (PCA) showed natural sample groupingamong beer raw materials.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Beer is one of the most popular beverages worldwide. It is avery complex matrix containing volatile and semi-volatile metab-olites, many of them contributing to its flavour, originating fromraw materials, namely malted barley and hops or hop-derivedproducts such as hop pellets, CO2 hop extract and tetra hop extract,as well as from the brewing process (Bamforth, 2003; Hughes,2003; Preedy, 2009; Silva, Augusto, & Poppi, 2008). Often, otherraw materials are used, like unmalted barley, wheat or corn(Preedy, 2009).

In traditional brewing, barley (Hordeum vulgare) is the grain ofchoice. It is readily available, fairly inexpensive, and presentsnumerous health benefits. In addition, it is a source of starch,

proteins, and cytolytic, proteolytic and amylolytic enzymes, whichare necessary for the efficient production of wort (Goode,Wijngaard, & Arendt, 2005). Typically, barley is subject to hydra-tion, partial germination and then kilning, which arrests germina-tion but ensures that specific enzyme activities survive. Thisprocess starts to release simple sugars, amino acids and otherlow molecular weight compounds that, under the influence of heat,yield a complex portfolio of Maillard reaction products (Preedy,2009). Malted barley can have an impact on beer stability due tothe presence of some metabolites with antioxidant properties,namely phenolic compounds, ascorbic acid, melanoidins, andseveral enzymes. Although barley malt is the most importantcereal, corn is also used as starch-containing adjuncts or extendersand sources for fermentable sugars (Erbe & Brückner, 2000).

Hops (Humulus lupulus L.) and hop-derived products includinghop pellets and tetra hop also impart attractive aromas as wellas the typical bitter taste. Hops are an economically important crop

J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280 267

for the brewing industry where they are used to impart flavour andaromas such as floral, spicy, herbal, woody and fruity characteris-tics. There are a large number of hop varieties commercially avail-able with distinct odour characteristics. Geographical location,climate and agronomical factors are the main factors that affectvolatile composition. Generally, hops are added during or afterthe wort-boiling process to provide the bitter taste, characteristicaromas, and allow the chemical isomerization of a-acids to themore bitter iso-a-acids. To minimize evaporation of essential oiland retain aroma compounds, premium aroma hops are added atthe end of boiling (late hopping) (Fernandes, Passos, Medeiros, &da Cunha, 2007; Preedy, 2009).

Understanding the differences in volatile metabolomic patternsfrom natural products requires reliable and sensible analyticalmethods. Several extraction and concentration methods, includingliquid–liquid extraction (LLE) (Wei, Mura, & Shibamoto, 2001), stirbar sorptive extraction (SBSE) (Horák et al., 2009; Perestrelo,Nogueira, & Câmara, 2009) and solid-phase extraction (SPE)(Mendes, Gonçalves, & Câmara, 2012) have been reported for theanalysis of a great number of volatile compounds belonging toheterogeneous chemical classes, such as esters, higher alcohols,fatty acids, aldehydes, ketones, hydrocarbons, ethers, sulphur-,alicyclic-, aromatic-and heterocyclic compounds, among others.These techniques, however, present some drawbacks such as theuse of organic solvents and expensive devices with a limitedlifetime as well as carryover or cross-contamination problems.Consequently, in order to overcome these drawbacks, solid phasemicroextraction (SPME) and more recently microextraction bypacked sorbent (MEPS) (Jönsson, Hagberg, & van Bavel, 2008) haveemerged as efficient extraction/pre-concentration methods andreliable alternatives to traditional sample preparation techniquesbecause of their simplicity, low cost, selectivity, and sensitivitywhen combined with appropriate detection modes (Caldeira,Rodrigues, Perestrelo, Marques, & Câmara, 2007; Câmara et al.,2007; Dietz, Sanz, & Cámara, 2006). SPME, developed by Pawliszynand co-workers (Arthur, Killam, Buchholz, Pawliszyn, & Berg,1992), is selective, rapid, simple and solvent-free allowing thepre-concentration of volatile compounds. Moreover, SPME canintegrate sampling, extraction, concentration and injection into asingle uninterrupted process, resulting in high sample throughput.Since its introduction, publications of using this promisingtechnique have described the analysis of volatiles in alcoholicbeverages (Caldeira et al., 2007; Perestrelo, Caldeira, Rodrigues, &Camara, 2008; Rodrigues, Caldeira, & Camara, 2008) and in differ-ent food samples, such as fruit (Pereira, Pereira, & Câmara, 2011),meats (Théron et al., 2010), vinegars (Cirlini, Caligiani, Palla, &Palla, 2011) and fish (Iglesias, Gallardo, & Medina, 2010).

The volatile metabolomic profile of raw materials contains valu-able information for brewers. However, no literature can be founddescribing this. Thus, in the present study, a simple and solvent-free technique, based on volatile extraction using SPME in head-space mode (HS-SPME) combined with gas chromatography–massspectrometry (GC–qMS) for identification, was used to establishthe metabolomic profile of volatile compounds in different rawmaterials used in Coral beer production (Empresa de Cervejas daMadeira, Portugal), mainly barley, corn, hop pellets, hop essentialoil from Saaz variety and tetra hop. The procedure was optimizedby selection of the appropriate fibre, and extraction temperatureand time. Potential differences in the sample composition wereinvestigated. The data obtained were subjected to correlation anal-ysis to determine the relationships amoung the different beer rawmaterials samples analyzed.

To our knowledge, the metabolomic patterns of volatile com-pounds in the raw materials for beer have not been investigatedpreviously. The results could provide useful information regardingthe distinctive metabolomic profile among the raw materials, some

of which contribute directly to flavour while others are importantin building up the background flavour of the final product.

2. Materials and methods

2.1. Reagents and materials

The SPME fibre coated with divinylbenzene/carboxen on poly-dimethylsiloxane (DBV/CAR/PDMS; StableFlex, 50/30 lm), SPMEholder for manual sampling, temperature controlled six-vial agita-tor tray and clear glass screw cap vials for SPME with PTFE/silica(film thickness 1.3 mm) septa were purchased from Supelco (Belle-fonte, PA, USA).

2.2. Samples

The samples of milled barley and corn, hop pellets and hopessential oil, obtained by supercritical CO2 extraction, and tetra-hydro isomerized hop extract (tetra hop) were kindly providedby Empresa de Cervejas da Madeira (ECM), Madeira Island(Portugal). Samples were transported under refrigeration(ca. 2–5 �C) to the laboratory and stored at �20 �C until analysis.

2.3. HS-SPME extraction procedure

The HS-SPME experimental parameters were established andoptimized previously by Gonçalves, Figueira, Rodrigues, andCâmara (2012). The SPME holder for manual sampling and fibrewere purchased from Supelco (Aldrich, Bellefonte, PA, USA). TheSPME device included a fused silica fibre coating partially cross-linked with 50/30 lm DVB–CAR–PDMS. The advantage of use a tri-phasic fibre, such as the DVB/CAR/PDMS chosen for this study, isthe recovery of volatile metabolites with both high and low polar-ity (Cirlini et al., 2011; Théron et al., 2010). The CAR-phase is highlyadsorptive (Iglesias et al., 2010) and, thus, increases the retentioncapacity of the fibre (Silva et al., 2008). Prior to use, the SPME fibrewas conditioned at 270 �C for 60 min in the GC injector, accordingto the manufacturer’s instructions. Blank runs were completed, be-fore sampling, each day to ensure no carry-over of analytes fromprevious extractions.

For the HS-SPME assay, aliquots of 0.5 ± 0.001 g of milled bar-ley, milled corn, hop pellets or hop essential oil were placed intoa 4 mL glass vial. The vial was closed and placed in a thermostaticcontrolled water bath adjusted to 40 ± 0.1 �C. The SPME fibre wasmanually inserted into the sample vial headspace during 30 min.After extraction, the fibre was retracted prior to removal fromthe sample vial and immediately inserted into the GC injection portfor desorption at 250 �C for six minutes in splitless mode.

For headspace extraction of volatile compounds from tetra hopextracts, aliquots of 0.58 mL (�0.5 g) were placed into a 4 mL glassvial and subjected to the same procedure to determine the meta-bolomics profile.

2.4. GC–qMS conditions

GC-qMS analysis of the SPME-collected volatile metaboliteswere performed on an Agilent Technologies 6890N Network gaschromatograph equipped with a BP-20 fused silica capillary col-umn (30 m length � 0.25 mm i.d.; film thickness 0.25 lm, SGE)and connected to an Agilent 5973N quadrupole mass-selectivedetector. Helium (Air Liquid, Portugal) was used as the carriergas at a flow rate of 1.1 mL min�1 (column head pressure of12 psi). The injections were performed in the splitless mode(5 min). The GC temperature program was from 40 �C (held for1.0 min) up to 200 �C at a rate of 1.7 �C min�1 (held for 1.0 min)

268 J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280

and further up to 220 �C at a rate of 30 �C min�1 (held for 1 min).For the MS system, the temperatures of the transfer line, quadru-pole and ionization source were 250, 180 and 230 �C, respectively;electron impact mass spectra were recorded at 70 eV and the ion-ization current was about 20 lA. The acquisitions were performedin full scan mode (30–300 m/z). The GC peak area of each metabo-lite was obtained from the ion extraction chromatogram (IEC) byselecting target ions for each. Reproducibility was expressed as rel-ative standard deviation (RSD). Signal acquisition and data pro-cessing were performed using MSD Chemstation (version E.02.00,Agilent Technologies, Palo Alto, CA).

Volatile metabolites were identified by comparison of GC reten-tion times and mass spectra with those obtained for pure standardcompounds (when available) using the data-system library (NIST,2005 software, Mass Spectral Search Program V.2.0d; NIST 2005,Washington, DC) and Kovat’s retention indices values (RI) calcu-lated according to Van Den Dool and Kratz (1963):

RI ¼ 100� ðA� CðnÞ=Cðnþ 1Þ � CðnÞÞ þ 100n

where A is the retention time of analyte; C(n) is the retention timeof n-alkane with n carbon atoms eluting before A; C(n + 1) is theretention time of the next n-alkane with (n + 1) carbon atoms elut-ing after A; and n is the number of carbon atoms in the n-alkanes.

For determination of RI, a homologous series of straight-chainn-alkanes (C8-C24, 40 mg L�1 in n-hexane) was used and the val-ues compared (when available) with values reported in the litera-ture for similar chromatographic columns and the experimentalconditions described above. All samples were analysed in triplicateand the results averaged.

2.5. Multivariate data analysis

Significant differences among beer raw materials were deter-mined by one-way analyses of variance (Anova) using SPSS version18.0. Principal component analysis (PCA) was also performed usingSPSS. This technique was applied to the normalized total peak areasof the volatile metabolites identified. PCA is an unsupervised tech-nique that reduces the dimensionality of the original data matrixretaining the maximum amount of variability. It is, therefore, possi-ble to explain the differences between several raw materials bymeans of factors obtained from the datasets and, at the same time,to determine which variables contribute most to such differences.

Stepwise Linear Discriminant Analyses (SLDA), probably the mostwidely applied supervised pattern recognition method, searchesfor directions (discriminant functions) that achieve maximum sep-aration among categories by maximising the between-class vari-ance relative to the within-class variance. SLDA renders anumber of orthogonal linear discriminant functions equal to thenumber of categories minus one. This method minimises the vari-ance within categories and maximises the variance between cate-gories. The variables included in the analyses are determined witha stepwise-LDA using a Wilk’s Lambda as a selection criterion andan F statistic factor to establish the significance of the changes inLambda when a new variable is tested. The prediction capacity ofthe discriminant models was studied by ‘‘cross-validation’’ todetermine the stability of the model.

3. Results and discussion

The combination of SPME with gas chromatography–mass spec-trometry is a powerful tool with superb sensitivity for establishingmetabolomic fingerprints of volatiles, in both qualitative and quan-titative terms. However, the efficiency of SPME extraction methodsdepend on the distribution constant of analytes partitioned be-tween the sample and fibre coating material. The chromatographic

profiles for volatile metabolites (peaks labelled on the chromato-grams) from beer raw materials obtained using HS-SPMEDVB/CAR/

PDMS/GC–qMS are presented in Fig. 1.A total of 152 volatile metabolites (91% of the extracted volatile

metabolites), 25 in barley, 20 in corn, 75 in hop pellets, 71 in hopessential oil and 39 in tetra hop, were identified based on compar-ison of mass spectra with the reference database (MS) and calcu-lated retention indices (RIcalc) with values reported in theliterature (RIlit) for the polyethylene glycol (or equivalent) capillarycolumn (Table 1). The retention indices of the experimental datawere in good agreement with those reported on the literature.Up to 65 min, a DRI = RIcalc � RIlit| ranged between 1 and 51 (RSD<5%, on average).

Up to approximately 65 min, a linear correlation between reten-tion index and retention time was obtained. The average relativestandard deviations (RSD, %) for the retention indices ranged from0.6% to 4.3%.

The metabolites analysed by HS-SPMEDVB/CAR/PDMS/GC–qMS arelisted in Table 1 according to elution order on a BP-20 capillary col-umn. Table 1 also includes Kovat’s retention indices (RI), odourdescriptor (El-Sayed, 2012; Terry & Arn, 2012), molecular formula,chemical class, and the corresponding average peak areas (n = 3).

The metabolites identified were grouped into different chemicalclasses including ketones (27), sesquiterpenes (26), monoterpenes(19), aliphatic esters (19), higher alcohols (15), aldehydes (11), fur-an compounds (11), aliphatic fatty acids (9), aliphatic hydrocar-bons (8), sulphur compounds (5) and nitrogen compounds (2). Itis worth noting the amounts of monoterpenes and sesquiterpenesin hop essential oil and hop pellets (expressed as GC peak areas)were significantly higher than those found in the other raw mate-rials. The structures of the major metabolites identified in the tar-get raw materials are shown in Fig. 2.

Significant differences (p < 0.05) in the average peak areas of al-most all of the metabolites identified were found among the fivebeer raw materials. Only b-myrcene was identified in all fivesamples.

3.1. Barley

A total of 25 volatile metabolites belonging to distinct chemicalgroups, mainly aldehydes (35%), sulphur compounds (28%), higheralcohols (22%), and in lower amounts furan compounds (11%),monoterpenes (3%), nitrogen compounds (1%) and ketones (0.3%)(Fig. 3) were identified in milled barley. A typical total ion chro-matogram (TIC) of volatile metabolites from barley using HS-SPMEDVB/CAR/PDMS/GC–qMS is shown in Fig. 1. The distribution ofvolatile metabolites chemical classes is presented in Fig. 3.

Considering the individual metabolites (Table 1), dimethyl su-fide (2; 3.05 � 107 ± 2.3%) was the main component found in bar-ley samples followed by 3-methylbutanal (10; 2.36 � 107 ± 2.9%),3-methyl-1-butanol (50; 9.65 � 106 ± 0.8%), 1-hexanol (74;7.23 � 106 ± 3.6%) and ethyl alcohol (13; 3.79 � 106 ± 3.7%).

In cereal grains, the presence of aldehydes and alcohols is oftenrelated to lipid oxidation (Preedy, 2009). As described by Yang,Schwarz, and Vick (1993), barley contains lipoxygenases andhydroperoxide isomerases that are responsible for the oxidationof unsaturated fatty acids leading to the production of volatilecompounds such as hexanal. Consequently, it is highly probablethat most odorants of barley result from lipid oxidation.

Beer raw materials could also be discriminated based on theircharacteristic volatile metabolites. For instance, 2-methylpropanal,3-hexanone, 2,5-dimethylfuran, 2-pentylfuran, 2-hexenal, 3-meth-allyl cyanide, 3-methyl-1-butanol, 1-pentanol, 2-methylpyrazine,1-octen-3-ol, furfural, (E)-2-nonenal, 5-methylfurfural, phenyl-acetaldehyde, furfuryl alcohol, and 2-phenylethanol only occurredin barley.

Fig. 1. Typical chromatograms (GC–qMS) of volatile metabolites from beer raw material samples, milled barley, milled corn, hop pellets, hop essential oil and tetra-hydroisomerized hop extract, extracted using HS-SPMEDVB/CAR/PDMS (peak assignments and identification, see Table 1).

J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280 269

3.2. Corn

Corn (Zea mays) is sometimes used as a (partial) substitute forbarley malt due to price, availability, and influence on the finalproduct. Generally, it gives the beer a lighter colour. There are noprevious studies reporting the use of SPME to extract volatilemetabolites from corn samples. According to the experimental re-sults, 20 metabolites were identified in the headspace of milledcorn samples with b-myrcene (38; 3.99 � 107 ± 25.2%) the majorcomponent (Table 1). Ethyl alcohol (13; 2.07 � 106 ± 18.2%) wasthe second most abundant metabolite followed by acetone (4;1.11 � 106 ± 2.1%), b-caryophyllene (107; 9.03 � 105 ± 17.7%), 3-methylbutyl-2-methylpropanoate (43; 7.87 � 105 ± 2.6%), b-pinene (31; 5.35 � 105 ± 5.7%) and b-terpinene (32;5.08 � 105 ± 1.6%). The monoterpenes were found to be the mostabundant chemical group followed by higher alcohols, sesquiter-penes, ketones and, in low amounts, aliphatic hydrocarbons, alde-hydes, aliphatic esters and aliphatic fatty acids. Fig. 3 presents thedistribution of volatile metabolites from different chemical groups.In corn, sesquiterpenes, ketones, aldehydes and monoterpenes ac-counted for 96.7% of total GC peak area.

3.3. Hop derived products

3.3.1. Hop pelletsIt was possible to identify 75 volatile metabolites in hop pellets

using HS-SPMEDVB/CAR/PDMS/GC–qMS, which included 23 sesquiter-penes, 15 monoterpenes, 11 ketones, seven aliphatic fatty acids, six

aliphatic esters, six higher alcohols, three aldehydes, two furancompounds, one aliphatic hydrocarbon and one sulphur com-pound. Hop pellets were characterized by a high content of, basedon peak area (Table 1), b-myrcene (38; 6.12 � 108 ± 6.9%) followedby b-caryophyllene (107; 5.15 � 108 ± 5.8%), b-farnesene (118;2.85 � 108 ± 4.0%), a-humulene (113; 2.24 � 108 ± 7.2%), 2-undecanone (109; 2.71 � 107 ± 0.3%) and linalool (104;2.63 � 107 ± 7.0%).

3.3.2. Hop essential oil derived from Saaz varietyHops may contain 0.5–3.0% essential oils and much of the fla-

vour in a beer depends on the aroma compounds added by thehops (Preedy, 2009). According to method developed, the hopessential oil derived from Saaz comprised of two major fractions:the first belonging to a group of hydrocarbons of which terpenehydrocarbons accounted for about 89.4% (total peak area) andthe remaining 10.6% aliphatic hydrocarbons, sulphur compoundsand compounds containing oxygen (oxygenated fraction, which isgenerally more aromatic and less volatile) such as aliphatic esters,aldehydes, ketones, furan compounds, higher alcohols and ali-phatic fatty acids (Fig. 4). As with the hop pellets, b-myrcene(38; 2.16 � 109 ± 5.2%) was the major component identified inthe Saaz hop essential oil followed by a-humulene (113;6.94 � 108 ± 13.3%), b-caryophyllene (107; 2.92 � 108 ± 12.6%), 3-methylbutyl-2-methylpropanoate (43; 8.03 � 107 ± 8.2%), methyl-4-decenoate (111; 6.99 � 107 ± 12.8%) and b-pinene (31;5.68 � 107 ± 16.4%).

Table 1Volatile metabolites identified in beer raw materials by HS-SPMEDVB/CAR/PDMS/GC-qMS (extraction temperature 40 �C for 30 min), the corresponding retention indices, odour descriptors, molecular formula (MF) and total peak area.

N� RIcala Volatile metabolite IDb Odour descriptor MFc Family Total peak area (�105 ± r)

Barley Corn Hop pellets Hop essential oil Tetrahop

1 907 2-Methyl-1,3-butadiene S, MS C5H8 Aliphatichydrocarbon

32.79 ± 2.91 25.92 ± 1.36

2 921 Dimethyl sulfide S, MS Gas, Sulphury, Cabbage, Moldy C2H6S Sulphurcompound

305.26 + 7.01 31.95 ± 3.62

3 934 2-Methylpropanal S, MS Green, Pungent, Burnt, Malty, Toasted C4H8O Aldehyde 30.65 ± 0.854 937 2-Propanone S, MS Glue, Fruity C3H6O Ketone 11.10 ± 0.23 80.26 ± 6.75 92.11 ± 11.84 162.05 ± 29.295 954 2-Methylfuran MS Sweet-gassy, Metallic-burnt, musty note C5H6O Furan

compound26.10 ± 0.71

6 957 2-Butenal MS Ethereal, Pungent C4H6O Aldehyde 7.94 ± 0.52 7.39 ± 1.507 962 3-Methylfuran MS C5H6O Furan

compound9.74 ± 1.75

8 964 2-Butanone S, MS Chocolate, Gas, Ethereal, Cheese, Butter C4H8O Ketone 12.55 ± 1.639 968 2-Methylbutanal S, MS Almond, Strong burnt, Malty, Cocoa C5H10O Aldehyde 4.27 ± 0.10

10 970 3-Methylbutanal MS Almond-like, Toasted, Malty, Green C5H10O Aldehyde 235.82 ± 6.93 11.67 ± 1.8311 971 2,6-Dimethyl octane MS C10H22 Aliphatic

hydrocarbon323.80 ± 30.07

12 975 3-Methylbutanone MS C5H10O Ketone 143.19 ± 26.0113 977 Ethyl alcohol S, MS Ethanol-like, Pungent, Sweet C2H6O Higher

alcohol37.87 ± 1.41 20.71 ± 3.80 10.37 ± 1.39 16.70 ± 3.85

14 987 2,5-Dimethylfuran MS C6H8O Furancompound

30.75 ± 0.71

15 998 2-Pentanone S, MS Fruity, Thinner, Acetone C5H10O Ketone 2.25 ± 0.1516 1007 Decane MS Fusel-like, Fruity, Sweet C10H22 Aliphatic

hydrocarbon30.60 ± 2.43

17 1012 4-Metyl-2-pentanone S, MS Ether C6H12O Ketone 8.24 ± 0.48 7.35 ± 0.71 589.81 ± 81.8018 1016 3-Metyl-2-pentanone S, MS C6H12O ketone 166.04 ± 26.4119 1016 a-Pinene S, MS

RTerpeny C10H16 Monoterpene 26.93 ± 3.08 110.77 ± 7.35

20 1024 MMECHd MS C10H20 Aliphatichydrocarbon

9.67 ± 1.11

21 1029 2-Methyl-3-buten-2-ol MS Sweet, Oily-fruity, Herbaceous-earthy C5H10O Higheralcohol

97.82 ± 8.43 40.75 ± 3.65

22 1035 3-Hexanone S, MS Grape, Fresh C6H12O Ketone 2.84 ± 0.0723 1036 5,6-Undecadiene MS C11H20 Aliphatic

hydrocarbon12.56 ± 0.21

24 1038 Camphene S, MS Camphoraceous, Pine, Oily, Herbal C10H16 Monoterpene 6.00 ± 0.79 7.93 ± 0.19 3.95 ± 0.4225 1039 2-Butylfuran MS C8H12O Furan

compound2.66 ± 0.23

26 1045 Dimethyl disulfide S, MS Sulphur, Cabbage, Ripened cheese, Putrid C2H6S2 Sulphurcompound

5.86 ± 0.68 33.14 ± 0.40

27 1047 5-Methyl-3-hexanone S, MS C7H14O Ketone 13.73 ± 1.8428 1051 2-

MethylpropylpropanoateS, MS C7H14O2 Aliphatic

ester14.21 ± 1.07

29 1051 Hexanal S, MS Grassy, Herbal, Leafy C6H12O Aldehyde 36.66 ± 0.79 4.32 ± o.15 2.19 ± 0.3830 1056 MMPPe MS C8H16O2 Aliphatic

ester0.81 ± 0.10 56.64 ± 3.82

31 1059 b-Pinene S, MS Musty, Green, Sweet, Pine, Resin,Turpentine, Woody

C10H16 Monoterpene 4.33 ± 0.28 5.35 ± o.30 126.75 ± 0.07 567.86 ± 93.40

32 1059 b-Terpinene S, MS C10H16 Monoterpene 5.08 ± 0.0833 1077 2-Methylbutyl acetate MS Herbaceous, Ethereal rum-like, Fermented-

fruity odor, BananaC7H14O2 Aliphatic

ester27.41 ± 2.94

34 1078 3,4-Dimethyl-2-pentene MS C7H14 Aliphatichydrocarbon

2.47 ± 0.28

270J.L.G

onçalveset

al./FoodChem

istry160

(2014)266–

280

35 1082 5-Methyl-2-hexanone MS C7H14O Ketone 23.80 ± 2.6036 1106 2-Heptanone S, MS Toasted, Nutty, Gravy, Soapy, Fruity C7H14O Ketone 15.23 ± 0.6837 1110 Dodecane MS Fusel-like C12H26 Aliphatic

hydrocarbon9.67 ± 1.20

38 1113 b-Myrcene S, MS Musty, Sweet, Lemon, Spicy, Woody C10H16 Monoterpene 23.07 ± 2.19 398.89 ± 100.64 6119.19 ± 422.97 21625.13 ± 1116.35 24.59 ± 1.0039 1126 MMBPf MS Mixed fruit odor C9H18O2 Aliphatic

ester25.13 ± 0.81

40 1127 2-Pentylpropanoate MS C8H16O2 Aliphaticester

1.07 ± 0.03 224.13 ± 12.55

41 1130 Limonene S, MS Citrus-like, Orange, Fruity, Peely C10H16 Monoterpene 4.75 ± 0.43 87.65 ± 7.53 191.69 ± 15.2042 1135 3-Methyl

cyclopentanoneMS Roasted beef C6H10O Ketone 7.39 ± 1.16

43 1135 MBMPg MS Mixed fruit odor C9H18O2 Aliphaticester

7.87 ± 0.20 802.55 ± 65.69 7.42 ± 0.71

44 1137 b-Phellandrene S, MS Terpeny, Fruity, Minty, Herbal C10H16 Monoterpene 39.46 ± 2.87 214.37 ± 15.9545 1139 2-Pentylfuran MS Buttery, Green bean-like C9H14O Furan

compound7.84 ± 0.21

46 1143 2-Methyl-1-butanol S, MS Malty, Balsamic, Wine, Ripe onion C5H12O Higheralcohol

60.10 ± 1.57 106.02 ± 2.12

47 1143 2-Hexenal S, MS Apple-like, Fruity, Green, Herbal, Leafy C6H10O Aldehyde 13.47 ± 0.8048 1147 Methallyl cyanide MS C5H7N Nitrogen

compound11.77 ± 1.2

49 1148 Methyl-3-methylbutanethioate

MS C6H12OS Sulphurcompound

42.41 ± 0.17

50 1155 3-Methyl-1-butanol S, MS Whiskey, Pungent, Balsamic, Alcohol,Fruity, Malty, Ripe onion, Burnt,

C5H12O Higheralcohol

96.55 ± 0.77

51 1163 6-Methyl-2-heptanone S, MS Fresh, Sweet C8H16O Ketone 2547.37 ± 448.5852 1170 (E)-b-Ocimene S, MS Herbaceous, Mild, Citrus, Orange, Lemon C10H16 Monoterpene 3.91 ± 0.68 17.27 ± 1.0853 1173 c-Terpinene S, MS Citrus-like, Herbaceous, Fruity, Sweet C10H16 Monoterpene 8.42 ± 1.13 14.33 ± 0.6854 1176 5-Methylethylhexanoate MS C8H16O2 Aliphatic

ester2.10 ± 0.25 0.92 ± 0.15 273.50 ± 20.92

55 1182 m-Cymene S, MS C10H14 Monoterpene 10.06 ± 1.4856 1184 (Z)-b-Ocimene MS Citrus-like, Herbaceous C10H16 Monoterpene 1.64 ± 0.13 28.84 ± 2.97 279.25 ± 24.5557 1190 1-Pentanol S, MS Fruity, Green, Sweet, Pungent C5H12O Higher

alcohol15.34 ± 1.47

58 1195 o-Cymene S, MS C10H14 Monoterpene 8.21 ± 1.06 8.75 ± 0.1059 1199 2-Methyl pyrazine MS Nutty, Roasty, Cocoa, Chocolate C5H6N2 Nitrogen

compound2.32 ± 0.15

60 1205 a-Terpinolene MS Woody, Fruity, Sweet, Piney C10H16 Monoterpene 6.06 ± 0.44 13.30 ± 0.2361 1210 2-Octanone S, MS Stewed, Fatty, Green, Fruity, Cheese odor C8H16O Ketone 4.09 ± 0.5562 1211 2-Methylbutyl-2-

methylbutyrateS, MS Fruity, Apple, Rum, Berry odor C10H20O2 Aliphatic

ester58.92 ± 7.70

63 1218 Methyl-2-methylheptanoate

MS C9H18O2 Aliphaticester

13.32 ± 1.05

64 1221 Methylheptanoate MS C8H16O2 Aliphaticester

2.39 ± 0.10 7.44 ± 0.18 205.47 ± 16.90

65 1225 2-Methylbutylpentanoate

MS C10H20O2 Aliphaticester

1.91 ± 0.05

66 1225 2,7-Dimethyl-4-octanone

MS C10H20O Ketone 28.14 ± 1.13

67 1229 Pentyl-3-methylbutanoate

MS C10H20O2 Aliphaticester

43.51 ± 4.64

68 1234 2,6-Dimethyl-4-hepten-3-one

MS C9H16O Ketone 62.37 ± 2.77

69 1258 3-Methyl-2-buten-1-ol S, MS Fruity, Green C5H10O Higheralcohol

14.29 ± 0.94 6.33 ± 0.60

70 1267 6-Methyl-5-hepten-2-one

S, MS Mushroom, Earthy, Rubber, Woody C8H14O Ketone 43.13 ± 3.38 41.71 ± 2.36

71 1276 3,4-Dimethyl-2- S, MS C8H16O Ketone 7.75 ± 0.50

(continued on next page)

J.L.Gonçalves

etal./Food

Chemistry

160(2014)

266–280

271

Table 1 (continued)

N� RIcala Volatile metabolite IDb Odour descriptor MFc Family Total peak area (�105 ± r)

Barley Corn Hop pellets Hop essential oil Tetrahop

hexanone72 1277 Methyl-6-

methylheptanoateS, MS C9H18O2 Aliphatic

ester17.83 ± 1.57 237.20 ± 26.94

73 1285 HDMCPFh MS C10H16O Furancompound

13.03 ± 0.69 7.41 ± 0.79

74 1290 1-Hexanol S, MS Green C6H14O Higheralcohol

72.33 ± 2.62 3.45 ± 0.46 11.00 ± 1.83

75 1299 Dimethyl trisulfide S, MS Rotten food, Fishy, Cauliflower, Cabbage,Onion

C2H6S3 Sulphurcompound

3.11 ± 0.64

76 1315 5-Methyl-4-hepten-3-one

MS C8H14O Ketone 30.06 ± 7.47

77 1319 3-Hexen-1-ol S, MS C6H12O Higheralcohol

1.61 ± 0.33 1.84 ± 0.13

78 1323 3-Nonanone MS Fruity, Soapy, Fatty, Green, Baked C9H18O Ketone 97.25 ± 8.67 34.82 ± 3.4379 1326 Nonanal S, MS Green, Fruity, Waxy, Soapy, Lavender, C9H18O Aldehyde 1.93 ± 0.2680 1327 Methyl octanoate MS Orange, Fruity, Green C9H18O2 Aliphatic

ester246.02 ± 23.30

81 1332 Propyl hexanethioate MS C9H18OS Sulphurcompound

69.08 ± 8.51

82 1335 2,4,4-Trimethyl-2-pentene

MS C8H16 Aliphatichydrocarbon

115.38 ± 26.93

83 1356 3-(4-methyl-3-pentenyl)furan

MS Woody C10H14O Furancompound

42.71 ± 4.27 40.13 ± 6.04

84 1361 2-Octanol S, MS Mushroom, Fat C8H18O Higheralcohol

3.36591

85 1366 2,8-Dimethyl-5-nonanone

MS C11H22O Ketone 17.14 ± 3.34

86 1374 Tetrahydrolinalool MS Floral, Citrus, Woody, Herbal C10H22O Monoterpene 331.33 ± 66.9087 1379 Heptyl propanoate MS C10H20O2 Aliphatic

ester7.37 ± 0.78

88 1382 a-Copaene S, MS Woody, Earthy C15H24 Sesquiterpene 12.94 ± 1.0389 1389 Acetic acid S, MS Sour, Vinegar, Pungent C2H4O2 Fatty acid 2.23 ± 0.44 49.16 ± 2.70 31.15 ± 0.2590 1390 1-Octen-3-ol S, MS Mushroom, Spicy, Rubbery, Herbaceous C8H16O Higher

alcohol6.35 ± 0.15

91 1400 Furfural S, MS Woody, Almond, Sweet, Fruity C5H4O2 Furancompound

32.82 ± 2.1

92 1401 Ylangene S, MS Fruity C15H24 Sesquiterpene 50.59 ± 5.16 50.34 ± 6.9093 1413 a-Cubebene S, MS Mild waxy, Woody C15H24 Sesquiterpene 131.29 ± 12.34 157.93 ± 25.85 10.72 ± 0.4394 1418 Muurolane MS C15H28 Sesquiterpene 40.41 ± 1.2895 1431 2-Decanone MS Citrus, Orange-like C10H20O Ketone 143.21 ± 12.6396 1433 2-Ethyl-1-hexanol S, MS Mild, Oily, Slightly floral rosy odor C8H18O Higher

alcohol1.52 ± 0.06

97 1434 Methyl nonanoate MS Fruity, Nut-like, Coconut-like C10H20O2 Aliphaticester

179.31 ± 19.40

98 1437 Decanal S, MS Burnt, Waxy, Orange skin-like, Floral C10H20O Aldehyde 1.71 ± 0.0599 1457 Benzaldehyde S, MS Burnt sugar, Almond, Woody C7H6O Aldehyde 5.85 ± 0.21 3.19 ± 0.19 7.07 ± 1.51

100 1468 (E)-2-Nonenal S, MS Green, Cucumber-like, Soapy, Floral C9H16O Aldehyde 33.10 ± 2.13101 1469 Methyl-5-nonenoate MS C10H18O2 Aliphatic

ester66.32 ± 8.23

102 1482 Propanoic acid S, MS Pungent, Rancid, Soy, Fruity, Cheese-like C3H6O2 Fatty acid 5.30 ± 0.47103 1488 Octyl-2-

methylpropanoateMS C12H24O2 aliphatic ester 26.34 ± 3.76

104 1497 Linalool S, MS Green, Floral, Lemon, Lavender-like C10H18O monoterpene 262.56 ± 18.39105 1510 2-Methylpropanoic acid S, MS Rancid, Cheese, Phenolic, Fatty C4H8O2 Fatty acid 23.99 ± 0.77

272J.L.G

onçalveset

al./FoodChem

istry160

(2014)266–

280

106 1514 5-Methylfurfural S, MS Caramel, Burnt sugar, Spicy, Acid, Coffee C6H6O2 Furancompound

12.77 ± 1.89

107 1521 b-Caryophyllene S, MS Oily, Fruity, Woody C15H24 Sesquiterpene 9.03 ± 1.59 5153.74 ± 297.68 2924.93 ± 370.07108 1543 5,5-Dimethyl-2(5H)-

furanoneS, MS C6H8O2 Furan

compound40.40 ± 3.87

109 1543 2-Undecanone MS Fruity, Musty, Dusty, Green C11H22O Ketone 270.71 ± 0.79 201.70 ± 24.53110 1551 DMOMIEi MS C14H22O Ketone 19.69 ± 0.20111 1573 Methyl-4-decenoate MS C11H20O2 Aliphatic

ester698.85 ± 89.41

112 1578 Phenylacetaldehyde MS Honey-like, Sweet, Rose, Green, Floral C8H8O Aldehyde 17.70 ± 1.39113 1595 a-Humulene S, MS Musty, Spicy, Woody, Terpene-like, Fruity C15H24 Sesquiterpene 4.42 ± 0.96 2244.95 ± 160.65 6944.19 ± 920.53 42.78 ± 8.39114 1602 c-Selinene S, MS Wood C15H24 Sesquiterpene 19.13 ± 0.45 23.32 ± 1.06115 1613 3,7-Dimethyl-1-octanol MS Floral, Rose, Waxy, Petal C10H22O Higher

alcohol22.97 ± 5.01

116 1613 Furfuryl alcohol S, MS Burnt sugar, Creamy, Caramellic note C5H6O2 Furancompound

5.72 ± 0.18

117 1618 c-Muurolene S, MS Oily, Herbaceous C15H24 Sesquiterpene 260.70 ± 30.23118 1620 b-Farnesene S, MS Oily, Fruity, Citrus-like, Woody C15H24 Sesquiterpene 2849.49 ± 114.76119 1632 2-Pentyl-2-cyclopenten-

1-oneMS C10H16O Ketone 7.86 ± 1.18

120 1635 (Z)-b-Famesene MS Woody, Green C15H24 Sesquiterpene 3.78 ± 0.42121 1639 Borneol MS Camphoraceous, Musty C10H18O Monoterpene 14.25 ± 2.62122 1640 (E,Z)-4-

EthylidenecyclohexeneMS C8H12 Aliphatic

hydrocarbon27.60 ± 0.10

123 1645 Methylgeranate MS Green, Fruit, Floral C11H18O2 Monoterpene 177.98 ± 37.08 442.20 ± 60.55124 1654 a-Selinene MS Pepper-like, Orange C15H24 Sesquiterpene 58.77 ± 3.75 65.57 ± 9.40125 1656 2-Dodecanone MS C12H24O Ketone 40.55 ± 4.71126 1664 a-Muurolene MS Woody C15H24 Sesquiterpene 39.22 ± 0.97 35.03 ± 5.26127 1674 b-Bisabolene MS Balsamic, Herbaceous C15H24 Sesquiterpene 48.41 ± 0.69128 1687 (Z, E)-a-Farnesene MS Wood, Sweet C15H24 Sesquiterpene 40.76 ± 1.35129 1696 a-Amorphene MS C15H24 Sesquiterpene 125.98 ± 2.48 109.90 ± 19.11130 1700 d-Cadinene S, MS Wood, Herbaceous C15H24 Sesquiterpene 184.26 ± 1.75 191.14 ± 31.50131 1711 a-Farnesene S, MS Woody C15H24 Sesquiterpene 22.00 ± 0.51132 1714 Valencene S, MS Green, Oil, Pepper-like, Orange C15H24 Sesquiterpene 22.33 ± 1.12133 1725 Cubenene S, MS C15H24 Sesquiterpene 18.18 ± 0.21 35.01 ± 4.04134 1729 a-Curcumene S, MS Herb C15H22 Sesquiterpene 13.02 ± 0.09135 1739 a-Cadinene S, MS Dry-wood, Weak medicinal C15H24 Sesquiterpene 18.26 ± 0.26136 1739 NHj MS C15H24 Sesquiterpene 15.51 ± 0.52137 1782 2-Tridecanone MS Spicy, Herbaceous C13H26O Ketone 41.64 ± 6.88 18.70 ± 3.78138 1788 Calamenene S, MS Weak Spicy, Weak floral C15H22 Sesquiterpene 21.28 ± 0.16139 1828 Hexanoic acid S, MS Sweaty, Pungent, Cheese, Goat-like C6H12O2 Fatty acid 15.80 ± 0.35 31.32 ± 4.36140 1835 Nerol S, MS Floral, Rose, Citrus, Marine C10H18O Monoterpene 19.20 ± 1.65 29.74 ± 5.79141 1857 Benzyl alcohol S, MS Aromatic, Floral, Fruity C7H8O Higher

alcohol3.36 ± 0.24 8.92 ± 1.38

142 1883 a-Calacorene MS Wood C15H20 Sesquiterpene 13.24 ± 0.02143 1884 Methyl-3,6-

dodecadienoateMS C13H22O2 Aliphatic

ester45.70 ± 11.24

144 1894 2-Phenylethanol S, MS Honey-like, Spicy, Herbal, Rose, Pollen C8H10O Higheralcohol

4.31 ± 0.15

145 1906 2-Tetradecanone S, MS C14H28O Ketone 2.81 ± 0.16146 1945 Caryophyllene oxide MS Sweet, Fruity, Sawdust, Herbaceous C15H24O Sesquiterpene 12.43 ± 0.47147 1951 Heptanoic acid S, MS Fatty, Sour-sweat-like, Rancid odor C7H14O2 Fatty acid 35.67 ± 3.07 24.14 ± 3.25148 1998 Perilla alcohol MS C10H16O Monoterpene 5.69 ± 0.49149 2108 Humulene oxide MS Creamy, Fatty, Buttery C15H24O Sesquiterpene 56.99 ± 6.52150 2177 Octanoic acid S, MS Fatty acid, Cheese, Fresh, Moss C8H16O2 Fatty acid 5.42 ± 0.04151 2301 Nonanoic acid S, MS Green, Fat, Musty, Sweaty, Sour C9H18O2 Fatty acid 5.34 ± 0.11152 2305 Hexadecanoic acid S, MS Oily C16H32O2 Fatty acid 2.91 ± 0.06Total volatile

metabolites25 20 75 71 39

(continued on next page)

J.L.Gonçalves

etal./Food

Chemistry

160(2014)

266–280

273

Tabl

e1

(con

tinu

ed)

N�

RI c

ala

Vol

atil

em

etab

olit

eID

bO

dou

rde

scri

ptor

MFc

Fam

ily

Tota

lpe

akar

ea(�

105

±r

)

Bar

ley

Cor

nH

oppe

llet

sH

opes

sen

tial

oil

Tetr

ahop

Tota

lPe

akA

rea

(�10

8)

1.07

0.49

19.2

638

.44

5.08

%R

.S.D

.(n

=3)

2.62

22.0

5.84

8.24

15.3

aK

ovat

’sre

ten

tion

inde

xre

lati

ven-

alka

nes

(C8–C

20)

ona

BP-

20ca

pill

ary

colu

mn

.b

IDm

eth

od:

MS,

ten

tati

vely

iden

tifi

edm

etab

olit

es;

S,m

etab

olit

espo

siti

vely

iden

tifi

edu

sin

gau

then

tic

stan

dard

.c

MF:

Mol

ecu

lar

form

ula

.d

MM

ECH

:C

is-1

-met

hyl

-4-(

1-m

eth

ylet

hyl

)-cy

cloh

exan

e.e

MM

PP:

2-M

eth

yl-2

-met

hyl

prop

ylpr

opan

oate

.f

MM

BP:

2-M

eth

yl-2

-met

hyl

buty

lpro

pan

oate

.g

MB

MP:

3-M

eth

ylbu

tyl-

2-m

eth

ylpr

opan

oate

.h

HD

MC

PF:

Hex

ahyd

ro-1

,1-d

imet

hyl

-4-m

eth

ylen

e-4H

-cyc

lope

nta

[c]f

ura

n.

iD

MO

MIE

:1-

(8,8

-dim

eth

yloc

tah

ydro

-3a,

6-m

eth

anoi

nde

n-1

-yl)

-eth

anon

e.j

NH

:[1

R-(

1-al

pha,

4a.a

lph

a,8a

.alp

ha)

]-1,

2,4a

,5,6

,8a-

hex

ahyd

ro-4

,7-d

imet

hyl

-1-(

1met

hyl

eth

yl)-

nap

hth

alen

e.

274 J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280

3.3.3. Tetra hopA total of 39 volatile metabolites belonging to distinct chemical

groups, mainly ketones (76.7%, total peak area from volatile frac-tion) aliphatic hydrocarbons (10.2%), monoterpenes (7.6%), higheralcohols (2.8%), and in low amounts sesquiterpenes (1.8%), furancompounds (0.8%) and aliphatic esters (0.2%) (Fig. 3), were identi-fied in the tetra-hydro isomerized hop extract. The major constitu-ents were, 6-methyl-2-heptanone (51; 2.55 � 108 ± 17.6%) and4-metyl-2-pentanone (17; 5.90 � 107 ± 13.9%), followed by tetra-hydrolinalool (86; 3.31 � 107 ± 20.1%), 2,6-dimethyloctane (11;3.24 � 107 ± 9.3%), 3-methyl pentanone (18; 1.66 � 107 ± 15.9%)and acetone (4; 1.62 � 107 ± 18.1%), which together accountedfor 81.1% of the volatile metabolite fraction.

3.4. Potential contribution of beer raw materials volatile metabolitesto the final product

Based on the volatile metabolites identified in beer raw materi-als, it would be interesting to predict the volatile metabolites thatcould potentially contribute to the aroma of the final product. Fig. 4shows the heat map for the main aroma descriptors found in theraw materials investigated.

Nykänen and Suomalainen (1983) reported the majority ofterpenes in alcoholic beverages originate in raw materials,although some compounds can be produced in biosynthesis by mi-cro-organisms. However, it is generally accepted that terpenes inbeer aroma are derived mainly from hops (Takoi et al., 2010).The monoterpene b-myrcene (38) and the sesquiterpenes b-caryo-phyllene (107), a-humulene (113) and b-farnesene (118) are wellknown to be major components of hop or hop-derived products(Preedy, 2009). Other monoterpenes, including a- and b-pinene(19, 31), camphene (24), limonene (41), (E)- and (Z)-b-ocimene(52, 56), b- and c-terpinene (32, 53), o-cymene (58) and a-terpino-lene (60), as well as sesquiterpenes c- and a-selinene (114, 124),c- and a-muurolene (117, 126), ylangene (92), d- and a-cadinene(130, 135), are also commonly found in hops (Moir, 2000). Despitethe great importance of terpenes in the quality of the final product,several studies (Fritsch & Schieberle, 2005; Kishimoto, Wanikawa,Kagami, & Kawatsura, 2005; Kishimoto, Wanikawa, Kono, & Shibata,2006) have suggested most of the hydrophobic terpenes from hopcones or pellets do not remain in the finished beer. Autoxidationand subsequent hydrolysis, and rearrangement of sesquiterpenehydrocarbons, lead to a large number of reaction products, whichincrease during storage of hops (Preedy, 2009) or can be synthe-sized during the boiling step. These include the epoxides withcaryophyllene oxide (146) and humulene epoxide II (149) beingthe most abundant. Preedy (2009) reported that sesquiterpeneepoxides and diepoxides undergo further hydrolysis and rear-rangements to form various ketones and alcohols.

Monoterpenols are, generally, biosynthetic products related tothe biosynthesis of b-myrcene. These include linalool (104) andtetrahydrolinalool (86), which are particularly important as floralodorants in hop pellets and tetra-hydro isomerized hop extract,respectively. Other monoterpenols found in the raw materials in-cluded borneol (121), nerol (140) and perilla alcohol (148).

3.4.1. Aliphatic estersAliphatic esters can be obtained by ethanolysis of acyl-coen-

zyme A (AcylCoA), which is formed during fatty acid synthesis ordegradation. Esters are characterized by pleasant fruit odours thatcontribute to the aromatic finesse of beer (Briggs, Boulton, Brookes,& Stevens, 2004). We determined many aliphatic esters in the rawmaterials including: linear methyl alkanoates (e.g. methylheptanoate, 64); branched methyl alkanoates such as methyl2-methyl-alkanoates and the methyl isoalkanoates (e.g. methyl-6-methylheptanoate, 72); unsaturated methyl alkenoates (e.g.

S

Dimethyl sulphide(2)

O

3-Methyl butanal(10)

O

2-Propanone(4)

OH

Ethyl alcohol(13)

O

4-Methyl-2-pentanone(17)

β-Myrcene(38)

HO

3-Methyl butanol(50)

6-Methyl-2-heptanone(51)

O HO

Tetrahydrolinalool(86)

β-Caryophyllene(107)

α-Humulene(113)

β-Farnesene(118)

Fig. 2. Structures of the major volatile metabolites identified in beer raw materials (numbering refers to Table 1).

J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280 275

methyl-3,6-dodecadienoate, 143); alkyl propanoates (e.g. heptylpropanoate, 87) and alkyl isobutanoates (e.g. octyl-2-methylpro-panoate, 103). Briggs et al. (2004) suggested the homologous seriesof straight chain methyl esters most likely originated from fattyacid biosynthesis, whereas the branched chain esters are derivedfrom amino acid biosynthesis.

The volatile fraction composed of C7–C13 aliphatic esters fromfatty acid were predominant in the hop essential oil of the differentvarieties. Perpète, Mélotte, Dupire, and Collin (1998) used volatilesfrom several varieties to establish a classification flowchart for hopessential oil, and the quantification of 3-methylbutyl-2-methylpro-panoate (43) and methyl-4-decenoate (111) can be used to dis-criminate the origins of hops. Furthermore, high level of methyl-3,6-dodecadienoate (143) can help to identify the Galena cultivar.

3.4.2. Higher alcoholsHigher alcohols are products of amino acids catabolism via

deamination of the a-keto acids (2-oxo acids) that are decarboxy-led to aldehydes and, finally, reduced to alcohols. Ethanols, 2-methyl-1-butanol, 3-methyl-1-butanol, and 1-octanol, identifiedin some of the raw materials, were also found in beer samples.The higher alcohols (i.e. of higher molecular weight than ethanol)are important as the immediate precursors of more flavour-activeesters, so higher alcohol formation needs to be controlled to en-sure ester production is optimal. The most important alcoholsare the amyl alcohols: 2-methyl-1-butanol (46; found in hopessential oil and tetra-hydro isomerized hop extract), and 3-methyl-1-butanol (50), and the aromatic alcohol, 2-phenylethanol(144). High concentrations of higher alcohols can result in a pun-gent and fusel-like smell and taste, whereas optimal levels impartfruity character (Nykänen, 1986). Increased concentration of amyl

alcohols (having an aroma described as sweet and choking) cancontribute negatively to the aroma of the final product, while2-phenylethanol contributes to a pleasant aroma, described asflowery, sweet and perfume-like notes (Falqué, Fernández, &Dubourdieu, 2001).

Dragone, Mussatto, Oliveira and Teixeira (2009) reported that1-hexanol has a positive influence on the aroma of the alcoholicbeverage when it occured at concentrations up to 20 mg L�1. Incontrast, increased concentrations of 1-hexanol, being describedas coconut-like, harsh and pungent, can contribute negatively tothe product aroma. Nykänen and Suomalainen (1983) reportedbenzyl alcohol was a minor compound in alcoholic beverages. Itis usually derived from the reduction of benzaldehyde, introducinga sweet and delicate floral aroma in alcoholic beverages.

3.4.3. Carbonyl compoundsCarbonyl compounds have an important role in beverage aroma

since they may affect flavour of beers (Saison, De Schutter,Delvaux, & Delvaux, 2008; Vesely, Lusk, Basarova, Seabrooks, &Ryder, 2003) and other beverages (Ferreira, Culleré, Loscos, &Cacho, 2006; Ledauphin, Barillier, & Beljean-Leymarie, 2006;Schmarr et al., 2008; Sowinski, Wardencki, & Partyka, 2005).Aldehydes (17) and ketones (13) were identified in the beer rawmaterial. They may have originated from a wide range of chemicalreactions such as lipid oxidation, Maillard reactions, Streckerdegradation and aldol condensation (Saison, De Schutter, Uyttenhove,Delvaux, & Delvaux, 2009).

Aldehydes are derived from the autoxidation and enzymolysisoxidation of the double carbon–carbon bond of unsaturated fattyacids present in cereals. Generally, aldehydes have a greater impacton the aroma of cereal products because of their low odour

(A)

(B)

0,0E+00 1,0E+07 2,0E+07 3,0E+07 4,0E+07 5,0E+07 6,0E+07

AE

HA

AH

FA

SC

NC

FC

Average Total Peak Area (a.u.)

0,0E+00 3,0E+08 6,0E+08 9,0E+08 1,2E+09 1,5E+09

Mon

Ses

Ald

Ket

Average Total Peak Area (a.u.)

Tetrahop Hop essen�al oil Hop pellets Corn Barley

Fig. 3. Distribution of volatile metabolites chemical classes by beer raw material.Chemical class code: Ket: ketone; Ald: aldehyde; Ses: sesquiterpenes: Mon:monoterpenes; FC: furan compounds; NC: nitrogen compounds; SC: sulphurcompounds; FA: fatty acids; HA: higher alcohols; AE: aliphatic esters.

276 J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280

threshold values. Aldehydes detected in the beer raw materialsbelonged to the linear aldehyde (n-alkanal), branched-chain aldehyde(methyl-alkanal), unsaturated aldehyde (2-alkenal) and aromaticaldehyde classes, respectively. The linear aldehydes, from hexanalto decanal, provide grassy, green, citrus and fatty odour character-istics. Branched-chain aldehydes, namely 2-methylbutanal (9) and3-methylbutanal (10) are described as potent flavour compounds(Smit, Engels, & Smit, 2009). They are, generally, perceived as maltyand chocolate-like. Unsaturated aldehydes (2-alkenals), such as(E)-2-butenal (6), (E)-2-hexenal (47) and (E)-2-nonenal (100) havevery low odour thresholds. With increasing C-chain length, theodour becomes less citrusy and fruity, and more fat-like. (E)-2-Nonenal (100) is most frequently cited as the cause of an unpleas-ant ‘cardboard’ flavour in beer (Preedy, 2009) or described as anunpleasant off-flavour and odour of rancid butter in stored beer(Svoboda et al., 2011). It was found solely in barley, whichaccounted 3.1% for total volatile fraction.

Benzaldehyde (99), with almond/acre odour, and phenyl acetal-dehyde (112), with a floral odour, were detected in millet corn.These aromatic aldehydes have been found in extruded cereals(Pfannhauser, 1993) and in model systems using corn starch, zeinand corn oil (Huang, Bruechert, Hartman, Rosen, & Ho, 1987).According to Pfannhauser (1993), benzaldehyde is a thermal reac-tion product of 2,4-decadienal and hexanal. However, Christophand Bauer-Christoph (2007) reported that benzaldehyde originatedfrom the enzymatic degradation of amygdalin in the stones of thefruits. It is considered an important food-flavouring agent and is akey ingredient in natural fruit flavours.

Ketones play an important role in the sensorial profile of beerraw materials and probably the final product. In general, ketonesare compounds associated with the pleasant fruity aromas. In beerraw materials, especially in hop-derived products, appearedmainly to be aliphatic monoketones (2-alkanones), branched-chainketones (methyl- and dimethyl-alkanones) and, in low amounts,unsaturated branched-chain ketones (e.g. 6-methyl-5-hepten-2-one, 70). Hashimoto and Eshima (1979) revealed that someketones and aldehydes, such as C3 to C11 2-alkanones, C2 to C10

alkanals, C4 to C7 2-alkenals and C6 to C7 2,4-alkedienals, arevolatile degradation products of iso-a-acids in beer model systemswith varying chain lengths. In an earlier study (Hashimoto &Kuroiwa, 1975), 2-methylpropanal (2), 2-propanone (4), 3-methyl-2-butanone (12), 4-methyl-2-pentanone (17) and 2-methyl-3-buten-2-ol (21) were also identified as oxidation products.4-Methyl-2-pentanone (17), 3-methyl-2-pentanone (18), 2-hepta-none (36), 2-octanone (61), 6-methyl-5-hepten-2-one (70),2-nonanone (78), 2-decanone (95) and 2-undecanone (109) arealso, usually, found in beer. Surburg and Panten (2006) reportedthat 2-alkanones (C3–C15), besides being present in the volatilefraction of many fruits and foodstuffs, do not contributed signifi-cantly to aroma with the exception of odd-numbered methylketones C7, C9 and C11, which possess a characteristic nutty note(Surburg et al., 2006).

3.4.4. Aliphatic fatty acidsThese compounds constitute an important group of aroma

compounds that can contribute fruity, cheesy and fatty odours(Rodrigues et al., 2008). Acids are also present in hop or hopderived products and are, usually, associated with aged hop asdegradation products of the a-acids and b-acids (Preedy, 2009).Williams and Wagner (1979) showed that degradation of thea- and b-acids released 2-methylpropanoic acid (105), 2-methylb-utanoic acid and 3-methylbutanoic acid. According to Williamsand Wagner (1979), these acids are precursors in the formationof staling esters and, within this family, acetic acid (89) and hepta-noic acid (147) are notable for their contribution.

The composition of these raw materials includes many othercompounds, such as aliphatic hydrocarbons, sulphur and nitrogencompounds. Aliphatic hydrocarbons were present at low levelsand are represented by a series of linear alkanes (e.g. decane, 16and dodecane, 37), a number of branched alkanes (e.g. 2,6-dimetyloctane, 11) and several trace alkenes (e.g. 5,6-undecadiene, 23).

Recently, Lermusieau and Collin (2003) noted the occurrenceand origins of sulphur compounds in hops and beer. Besides beingpresent at trace levels, sulphur compounds can contribute to the fi-nal odour of beer due mainly to their very low odour thresholds.Among sulphur compounds, methyl thioesters have been com-monly identified in hop essential oil (Lermusieau & Collin, 2003).The concentration of thioesters depends on variety and local grow-ing conditions, and increase considerably upon kiln drying, inde-pendent of sulphur dioxide (SO2) application. Dimethyl sulphide(2) is also generated during steam distillation and wort boilingby thermal degradation of S-methylcysteine sulphoxide. The levelsof dimethyl suphide and polysulphides (e.g. dimethyl and tri-methyl sulphides; 26 and 75) also increase with concentrationsof sulphur and when kilning is performed without SO2 (Preedy,2009). These sulphide compounds have characteristic cooked veg-etable, onion, rubbery and sulphous odours, which may impact thearoma of beers.

As regards to nitrogenous compounds in beer, they are derivedmainly from barley malt and its adjuncts (Preedy, 2009). Accordingto Preedy (2009), nitrogen compounds may also have an importantrole in influencing foam quality and stability. The content of thesecompounds in barley depends on variety and on environmentalcondition during cultivation.

Odor descriptor Odor descriptor

2-Methyl-1,3-butadiene 1 77 3-Hexen-1-olDimethyl sulfide X X x 2 78 X X X 2-Nonanone

2-Methylpropanal X X X 3 79 X X X X X Nonanal2-Propanone X 4 80 X X X Methyl octanoate2-Methylfuran X 5 81 S-Propyl hexanethioate

2-Butenal X 6 82 2,4,4-Trimethyl-2-pentene3-Methylfuran 7 83 X 3-(4-methyl-3-pentenyl)furan

2-Butanone X X X 8 84 X X 2-Octanol2-Methylbutanal X X 9 85 2,8-Dimethyl-5-nonanone3-Methylbutanal X X X 10 86 X X X X Tetrahydrolinalool

2,6-Dimethyl octane 11 87 Heptyl propanoate3-Metylbutanone 12 88 X α-Copaene

Ethyl alcohol X X 13 89 X Acetic acid2,5-Dimethylfuran 14 90 X X 1-Octen-3-ol

2-Pentanone X X 15 91 X X FurfuralDecane X X 16 92 X Ylangene

4-Metyl-2-pentanone X 17 93 X X α-Cubebene3-Metyl-2-pentanone 18 94 Muurolane

α-Pinene 19 95 X 2-DecanoneMMECHa 20 96 X X X 2-Ethyl-1-hexanol

2-Methyl-3-buten-2-ol X X X 21 97 X X Methyl nonanoate3-Hexanone X X 22 98 X X X X Decanal

5,6-Undecadiene 23 99 X X X BenzaldehydeCamphene X X x 24 100 X X X X X (E)-2-Nonenal

2-Butylfuran 25 101 Methyl-5-nonenoateDimethyl disulfide X 26 102 X X Propanoic acid

5-Methyl-3-hexanone 27 103 Octyl-2-methylpropanoate2-Methylpropyl propanoate 28 104 X X X X Linalool

Hexanal X X X 29 105 X 2-Methylpropanoic acidMMPPb 30 106 X X X 5-Methylfurfuralβ-Pinene X X X X X 31 107 X X X β-Caryophyllene

β-Terpinene 32 108 5,5-Dimethyl-2(5H)-furanone2-Methylbutyl acetate X X X 33 109 X X X X 2-Undecanone

3,4-Dimethyl-2-pentene 34 110 DMOMIEf

5-Methyl-2-hexanone 35 111 Methyl-4-decenoate2-Heptanone X X 36 112 X X X Phenylacetaldehyde

Dodecane X 37 113 X X X X α-Humuleneβ-Myrcene X X X X X 38 114 X γ-Selinene

MMBPc X 39 115 X X 3,7-Dimethyl-1-octanol2-Pentylpropanoate 40 116 X X Furfuryl alcohol

Limonene X X 41 117 X X γ-Muurolene3-Methylcyclopentanone X 42 118 X X X X β-Farnesene

MBMPd X 43 119 2-Pentyl-2-cyclopenten-1-oneβ-Phellandrene X X 44 120 X X (Z)-β-Famesene

2-Pentylfuran X X 45 121 X Borneol2-Methyl-1-butanol X X 46 122 (E, Z)-4-Ethylidenecyclohexene

2-Hexenal X X X 47 123 X X X MethylgeranateMethallyl cyanide 48 124 α-Selinene

S-Methyl 3-methylbutanethioate 49 125 2-Dodecanone3-Methyl-1-butanol X X X 50 126 X α-Muurolene

6-Methyl-2-heptanone X X 51 127 X β-Bisabolene(E)-β-Ocimene X X X 52 128 X X (Z,E)-α-Farnesene

γ-Terpinene X X X 53 129 α-Amorphene5-Methyl-methyl ester hexanoic acid 54 130 X X δ-Cadinene

m-Cymene 55 131 X α-Farnesene(Z)-β-Ocimene X X 56 132 X X X Valencene

1-Pentanol X X X 57 133 Cubeneneo-Cymene 58 134 X α-Curcumene

2-Methyl-pyrazine X 59 135 X X α-Cadineneα-Terpinolene X X 60 136 NHg

2-Octanone X X X 61 137 X 2-Tridecanone2-Methylbutyl 2-methylbutyrate X X 62 138 X X Calamenene

Methyl 2-methylheptanoate 63 139 X X X Hexanoic acidMethyl heptanoate 64 140 X X Nerol

2-Methylbutyl pentanoate 65 141 X X Benzyl alcohol2,7-Dimethyloctan-4-one 66 142 X α-CalacorenePentyl 3-methylbutanoate 67 143 Methyl-3,6-dodecadienoate

2,6-Dimethyl-4-hepten-3-one 68 144 X X X Phenylethyl alcohol3-Methyl-2-buten-1-ol X X 69 145 2-Tetradecanone

6-Methyl-5-hepten-2-one X X 70 146 X X Caryophyllene oxide3,4-Dimethyl-2-hexanone 71 147 X Heptanoic acid

Methyl-6-methylheptanoate 72 148 Perilla alcohol HDMCPFe 73 149 X Humulene oxide

1-Hexanol X 74 150 X X X Octanoic acidDimethyl trisulfide X X 75 151 X X X X Nonanoic acid

5-Methyl-4-hepten-3-one 76 152 X Hexadecanoic acid

Fatty

Mus

ty

Mal

ty

Ethe

r

Her

bal

Sulfu

r

Woo

dy

Gas

Flor

al

Frut

y

Citr

us

Gre

en

Gas

Flor

al

Frut

y

Mal

ty

Ethe

r

Her

bal

Sulfu

r

Woo

dy

Citr

us

Gre

en

Fatty

Mus

tyVolatile metabolite Volatile metabolitePeak Number

Fig. 4. Heat map of the odour descriptors of volatile metabolites found in beer raw materials. Results based on the literature data (El-Sayed, 2012; Terry et al., 2012).aMMECH: Cis-1-methyl-4-(1-methylethyl)-cyclohexane; bMMPPA: 2-Methyl-2-methylpropylpropanoate; cMMBPA: 2-Methyl-2-methylbutylpropanoate; dMBMPA: 3-Methylbutyl-2-methylpropanoate; eHDMCPF: Hexahydro-1,1-dimethyl-4-methylene-4H-cyclopenta[c]furan; fDMOMIE: 1-(8,8-dimethyloctahydro-3a,6-methanoinden-1-yl)-ethanone; gNH: [1R-(1-alpha, 4a.alpha, 8a.alpha)]-1,2,4a,5,6,8a-hexahydro-4,7-dimethyl-1-(1methylethyl)-naphthalene.

J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280 277

3.5. Multivariate data analysis

3.5.1. Unsupervised pattern recognition analysisThe dataset consisted of a 15 � 152 matrix. Rows represented

the raw material analysed (15 objects) while columns representedthe peak areas of 152 volatile metabolites determined usingHS-SMPEDVB/CAR/PDMS/GC–qMS. A preliminary evaluation of theinformation in the data matrix was performed using principalcomponent analysis (PCA) (unsupervised pattern recognition)(Berrueta, Alonso-Salces, & Héberger, 2007). PCA was performed

on normalised data. Only the principal components (PCs) witheigenvalues P1 were considered for statistic treatment. Theresults are presented as a bi-dimensional plot of sample scores(in the space defined by the first two PCs) in Fig. 5. Five separateclusters were observed, showing the natural separation accordingto the raw material (Fig. 5A). Two groups, hop essential oil andhop pellets were clearly separated from one another, and the otherraw materials, barley, corn and tetra hop.

The two main PCs accounted for 98.1% of the total data variabil-ity (Table 2). The first component (PC1), explaining 52.6% of the

(A) (B)

Hop pellets

Hop essen�al oil

Tetrahop

Barley

Corn

Fig. 5. Principal component analysis (PCA) scores scatter plot (A) and loadings weight plot for the first and second principal components (B) of beer raw materials using thevolatile metabolites identified by GC-qMS. Each number corresponds to a particular volatile metabolite, as indicated in Table 1. ZA – normalized variables values.

Table 2Percentage of total variance explained.

PCa Initial eigenvalues Extraction sums of squared loadings (%) Rotation sums of squared loadings

Total % of Variance Cumulative (%) Total % of Variance Cumulative (%) Total % of Variance Cumulative (%)

1 52.603 55.961 55.961 52.603 55.961 55.961 49.487 52.646 52.6462 39.578 42.104 98.065 39.578 42.104 98.065 42.694 45.419 98.0653 1.129 1.202 99.267

a PC – Principal Component.

Table 3Classification Results for Supervised Pattern Recognition Technique SLDA Applied to Beer Raw Material Samples.

Ident Predicted Group Membership Total

Barley Corn Hop pell Hop esse Tetrahop

Classification resultsb,c

Original Cont Barley 3 0 0 0 0 3Corn 0 3 0 0 0 3Hop pell 0 0 3 0 0 3Hop esse 0 0 0 3 0 3Tetrahop 0 0 0 0 3 3

% Barley 100.0 .0 .0 .0 .0 100.0Corn .0 100.0 .0 .0 .0 100.0Hop pell .0 .0 100.0 .0 .0 100.0Hop esse .0 .0 .0 100.0 .0 100.0Tetrahop .0 .0 .0 .0 100.0 100.0

Cross-validateda Cont Barley 3 0 0 0 0 3Corn 0 3 0 0 0 3Hop pell 0 0 2 1 0 3Hop esse 0 1 0 0 2 3Tetrahop 0 0 0 0 3 3

% Barley 100.0 .0 .0 .0 .0 100.0Corn .0 100.0 .0 .0 .0 100.0Hop pell .0 .0 66.7 33.3 .0 100.0Hop esse .0 33.3 .0 .0 66.7 100.0Tetrahop .0 .0 .0 .0 100.0 100.0

a Cross validation is done only for those cases in the analysis. In cross validation, each case is classified by the functions derived from all cases other than that case.b 100.0% Of original grouped cases correctly classified.c 73.3% Of cross-validated grouped cases correctly classified.

278 J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280

variance, separated mainly hop essential oil from the tetra hop,barley and corn. PC2 (accounting for 45.4% of the total variability)clearly separated hop pellets from all the other raw materials.Analysis of the loadings plot revealed the compounds responsible

for the separation between samples (Fig. 5B). Observing the load-ing of the variables (Fig. 5B), the ones that contributed most toPC1 were dimethyl disulfide (26), -methylpropylpropanoate (28),2-methylbutanal (9), 5,6-undecadiene (23) and b-pinene (31).

J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280 279

Dominant features in PC2 were 2-pentanone (15), 2-methyl-2-methylbutylpropanoate (39), 2-octanone (61) and methyl-2-meth-ylheptanoate (63).

3.5.2. Supervised pattern recognition analysisA stepwise linear discriminant analysis (SLDA) (supervised

analysis) was carried out to obtain classification rules for predict-ing composition of beer raw materials, based on the volatilemetabolome. A normalised data matrix for the volatile metabolitesfrom beer raw materials was the input dataset. SLDA producedclassification rules that were 100% successful when assigning sam-ples to classes for all the raw material investigated (Table 3). Whenchecking the model using cross-validation, the percentage of suc-cess was 100% for barley but lower for corn and hop essential oil(66.7%) (Table 3). The worse prediction percentages were observedfor hop pellets and tetra hop (33.3%) (Table 3).

4. Conclusion

In this paper, we report the composition of volatile metabolitesfrom different raw materials commonly used in beer productionand analysed using HS-SMPEDVB/CAR/PDMS/GC–qMS. This approachallowed the creation of volatile metabolomic patterns derived fromraw materials of beer, some of which may be influence in the aro-ma of the final product. An important advancement in raw materi-als metabolite analysis was demonstrated, as the method allowedfor the simultaneous analysis of a significant number of metabo-lites found in the headspace of beer raw materials compared withother conventional methods.

Acknowledgments

The authors acknowledge the Empresa de Cervejas da Madeira(ECM) for the supply of beer raw material samples and PortugueseFoundation for Science and Technology (FCT) through the StrategicPlan PEst-OE/QUI/UI0674/2011, and the Portuguese Mass Spec-trometry Network – RNEM2011.

References

Arthur, C. L., Killam, L. M., Buchholz, K. D., Pawliszyn, J., & Berg, J. R. (1992).Automation and optimization of solid-phase microextraction. AnalyticalChemistry, 64(17), 1960–1966.

Bamforth, C. (2003). Beer: Tap into the art and science of brewing. New York: OxfordUniversity Press.

Berrueta, L. A., Alonso-Salces, R. M., & Héberger, K. (2007). Supervised patternrecognition in food analysis. Journal of Chromatography A, 1158(1–2), 196–214.

Briggs, D. E., Boulton, C. A., Brookes, P. A., & Stevens, R. (2004). Brewing: Science andpractice. Cambridge, UK: CRC Press.

Caldeira, M., Rodrigues, F., Perestrelo, R., Marques, J. C., & Câmara, J. S. (2007).Comparison of two extraction methods for evaluation of volatile constituentspatterns in commercial whiskeys: Elucidation of the main odour-activecompounds. Talanta, 74(1), 78–90.

Câmara, J. S., Marques, J. C., Perestrelo, R. M., Rodrigues, F., Oliveira, L., Andrade, P.,et al. (2007). Comparative study of the whisky aroma profile based onheadspace solid phase microextraction using different fibre coatings. Journalof Chromatography A, 1150(1–2), 198–207.

Christoph, N., & Bauer-Christoph, C. (2007). Flavour of spirit drinks: Raw materials,fermentation, distillation, and ageing. In R. G. Berger (Ed.), Flavours andfragrances: Chemistry, bioprocessing and sustainability (pp. 219–239). Berlin:Springer.

Cirlini, M., Caligiani, A., Palla, L., & Palla, G. (2011). HS-SPME/GC–MS andchemometrics for the classification of Balsamic Vinegars of Modena ofdifferent maturation and ageing. Food Chemistry, 124(4), 1678–1683.

Dietz, C., Sanz, J., & Cámara, C. (2006). Recent developments in solid-phasemicroextraction coatings and related techniques. Journal of Chromatography A,1103(2), 183–192.

Dragone, G., Mussatto, S. I., Oliveira, J. M., & Teixeira, J. A. (2009). Characterisation ofvolatile compounds in an alcoholic beverage produced by whey fermentation.Food Chemistry, 112(4), 929–935.

El-Sayed, A. (2012). The Pherobase: Database of Pheromones and Semiochemicals,2012. <http://www.pherobase.com>. Accessed 08.03.12.

Erbe, T., & Brückner, H. (2000). Chromatographic determination of amino acidenantiomers in beers and raw materials used for their manufacture. Journal ofChromatography A, 881(1–2), 81–91.

Falqué, E., Fernández, E., & Dubourdieu, D. (2001). Differentiation of white wines bytheir aromatic index. Talanta, 54(2), 271–281.

Fernandes, E. S., Passos, G. F., Medeiros, R. M., da Cunha, F., Ferreira, J., Campos, M.M., et al. (2007). Anti-inflammatory effects of compounds alpha-humulene and(�)-trans-caryophyllene isolated from the essential oil of Cordia verbenacea.European Journal of Pharmacology, 569, 228–236.

Ferreira, V., Culleré, L., Loscos, N., & Cacho, J. (2006). Critical aspects of thedetermination of pentafluorobenzyl derivatives of aldehydes by gaschromatography with electron-capture or mass spectrometric detection:Validation of an optimized strategy for the determination of oxygen-relatedodour-active aldehydes in wine. Journal of Chromatography A, 1122(1–2),255–265.

Fritsch, H. T., & Schieberle, P. (2005). Identification based on quantitativemeasurements and aroma recombination of the character impact odourantsin a Bavarian pilsner-type beer. Journal of Agricultural and Food Chemistry,53(19), 7544–7551.

Gonçalves, J., Figueira, J., Rodrigues, F., & Câmara, J. S. (2012). Headspace solid-phasemicroextraction combined with mass spectrometry as a powerful analyticaltool for profiling the terpenoid metabolomic pattern of hop-essential oil derivedfrom Saaz variety. Journal of Separation Science, 35, 2282–2296.

Goode, D. L., Wijngaard, H. H., & Arendt, E. K. (2005). Mashing with unmalted barley– Impact of malted barley and commercial enzyme (Bacillus spp.) additions.Master Brewers Association of the Americas, 42(3), 184–198.

Hashimoto, N., & Eshima, T. (1979). Oxidative degradation of isohumulones inrelation to flavour stability of beer. Journal of the Institute of Brewing, 85,136–140.

Hashimoto, N., & Kuroiwa, Y. (1975). Proposed pathways for the formation ofvolatile aldehydes during storage of bottled beer. Proceedings of the AmericanSociety of Brewing Chemists, 33, 104–111.

Horák, T. s., Culík, J. i., Cejka, P., Jurková, M., Kellner, V. r., Dvorák, J., et al. (2009).Analysis of free fatty acids in beer: Comparison of solid-phase extraction, solid-phase microextraction, and stir bar sorptive extraction. Journal of Agriculturaland Food Chemistry, 57(23), 11081–11085.

Huang, T. C., Bruechert, L. J., Hartman, T. G., Rosen, R. T., & Ho, C. T. (1987). Effect oflipids and carbohydrates on thermal generation of volatiles from commercialzein. Journal of Agricultural and Food Chemistry, 35(6), 985–990.

Hughes, P. (2003). Encyclopedia of food sciences and nutrition. Amsterdam: Elsevier.Iglesias, J., Gallardo, J. M., & Medina, I. (2010). Determination of carbonyl

compounds in fish species samples with solid-phase microextraction with on-fibre derivatization. Food Chemistry, 123(3), 771–778.

Jönsson, S., Hagberg, J., & van Bavel, B. (2008). Determination of 2,4,6-trichloroanisole and 2,4,6-tribromoanisole in wine using microextraction inpacked syringe and gas chromatography–mass spectrometry. Journal ofAgricultural and Food Chemistry, 56(13), 4962–4967.

Kishimoto, T., Wanikawa, A., Kagami, N., & Kawatsura, K. (2005). Analysis of hop-derived terpenoids in beer and evaluation of their behavior using the stirBarâ’Sorptive extraction method with GC–MS. Journal of Agricultural and FoodChemistry, 53(12), 4701–4707.

Kishimoto, T., Wanikawa, A., Kono, K., & Shibata, K. (2006). Comparison of theodour-active compounds in unhopped beer and beers hopped withdifferent hop varieties. Journal of Agricultural and Food Chemistry, 54(23),8855–8861.

Ledauphin, J., Barillier, D., & Beljean-Leymarie, M. (2006). Gas chromatographicquantification of aliphatic aldehydes in freshly distilled Calvados and Cognacusing 3-methylbenzothiazolin-2-one hydrazone as derivative agent. Journal ofChromatography A, 1115(1–2), 225–232.

Lermusieau, G., & Collin, S. (2003). Volatile sulphur compounds in hops and residualconcentrations in beer – A review. Journal of the American Society of BrewingChemists, 61(3), 109–113.

Mendes, B., Gonçalves, J., & Câmara, J. S. (2012). Effectiveness of high-throughputminiaturized sorbent- and solid phase microextraction techniques combinedwith gas chromatography–mass spectrometry analysis for a rapid screening ofvolatile and semi-volatile composition of wines – a comparative study. Talanta,88, 79–94.

Moir, M. (2000). Hops – A millennium review. Journal of the American Society ofBrewing Chemists, 58(4), 131–146.

Nykänen, L. (1986). Formation and occurrence of flavour compounds in wine anddistilled alcoholic beverages. American Journal of Enology and Viticulture, 37(1),84–96.

Nykänen, L., & Suomalainen, H. (1983). Aroma of beer, wine and distilled alcoholicbeverages. Berlin: Kluer Academic Publishers.

Pereira, J., Pereira, J., & Câmara, J. S. (2011). Effectiveness of different solid-phasemicroextraction fibres for differentiation of selected Madeira island fruits basedon their volatile metabolite profile-identification of novel compounds. Talanta,83(3), 899–906.

Perestrelo, R., Caldeira, M., Rodrigues, F., & Camara, J. S. (2008). Volatile flavourconstituent patterns of Terras Madeirenses red wines extracted by dynamicheadspace solid-phase microextraction. Journal of Separation Science, 31(10),1841–1850.

Perestrelo, R., Nogueira, J. M., & Câmara, J. S. (2009). Potentialities of two solventlessextraction approaches – Stir bar sorptive extraction and headspace solid-phasemicroextraction for determination of higher alcohol acetates, isoamyl estersand ethyl esters in wines. Talanta, 80(2), 622–630.

280 J.L. Gonçalves et al. / Food Chemistry 160 (2014) 266–280

Perpète, P., Mélotte, L., Dupire, S., & Collin, S. (1998). Varietal discrimination of hoppellets by essential oil analysis I. Comparison of fresh samples. Journal of theAmerican Society of Brewing Chemists, 56(3), 104–108.

Pfannhauser, W. (1993). Volatiles formed during extrusion cooking of cereals.Flavour and Fragrance Journal, 8(2), 109–113.

Preedy, V. (2009). Beer in health and disease prevention. London: Academic Press.Rodrigues, F., Caldeira, M., & Camara, J. S. (2008). Development of a dynamic

headspace solid-phase microextraction procedure coupled to GC-qMSD forevaluation the chemical profile in alcoholic beverages. Analytica Chimica Acta,609(1), 82–104.

Saison, D., De Schutter, D. P., Delvaux, F., & Delvaux, F. R. (2008). Optimisation ofa complete method for the analysis of volatiles involved in the flavourstability of beer by solid-phase microextraction in combination with gaschromatography and mass spectrometry. Journal of Chromatography A,1190(1–2), 342–349.

Saison, D., De Schutter, D. P., Uyttenhove, B., Delvaux, F., & Delvaux, F. R. (2009).Contribution of staling compounds to the aged flavour of lager beer by studyingtheir flavour thresholds. Food Chemistry, 114(4), 1206–1215.

Schmarr, H.-G., Sang, W., Ganß, S., Fischer, U., Köpp, B., Schulz, C., et al. (2008).Analysis of aldehydes via headspace SPME with on-fiber derivatization to theirO-(2,3,4,5,6-pentafluorobenzyl)oxime derivatives and comprehensive 2D-GC–MS. Journal of Separation Science, 31(19), 3458–3465.

Silva, G. A. d., Augusto, F., & Poppi, R. J. (2008). Exploratory analysis of the volatileprofile of beers by HS-SPME–GC. Food Chemistry, 111(4), 1057–1063.

Smit, B., Engels, W., & Smit, G. (2009). Branched chain aldehydes: Production andbreakdown pathways and relevance for flavour in foods. Applied Microbiologyand Biotechnology, 81(6), 987–999.

Sowinski, P., Wardencki, W., & Partyka, M. (2005). Development and evaluation ofheadspace gas chromatography method for the analysis of carbonyl compoundsin spirits and vodkas. Analytica Chimica Acta, 539(1–2), 17–22.

Surburg, H., & Panten, J. (2006). Common fragrance and flavour materials: Preparation,properties and uses. Germany: Wiley-VCH.

Svoboda, Z., Mikulíková, R., Beláková, S., Benešová, K., Márová, I., & Nesvadba, Z.(2011). Optimization of modern analytical SPME and SPDE methods fordetermination of &lt;i&gt;Trans&lt;/i&gt;-2-nonenal in Barley, Malt and Beer.Chromatographia, 73, 157–161.

Takoi, K., Koie, K., Itoga, Y., Katayama, Y., Shimase, M., Nakayama, Y., et al. (2010).Biotransformation of hop-derived monoterpene alcohols by lager yeast andtheir contribution to the flavour of hopped beer. Journal of Agricultural and FoodChemistry, 58(8), 5050–5058.

Terry, A., & Arn, H. (2012). Flavournet and human odour space. Vol. 2012:Flavournet and human odour space. <http://www.flavournet.org>, AccessedMarch 8.

Théron, L., Tournayre, P., Kondjoyan, N., Abouelkaram, S., Santé-Lhoutellier, V., &Berdagué, J.-L. (2010). Analysis of the volatile profile and identification ofodour-active compounds in Bayonne ham. Meat Sci., 85(3), 453–460.

van Den Dool, H., & Kratz, P. D. (1963). A generalization of the retention indexsystem including linear temperature programmed gas–liquid partitionchromatography. Journal of Chromatography A, 11, 463–471.

Vesely, P., Lusk, L., Basarova, G., Seabrooks, J., & Ryder, D. (2003). Analysis ofaldehydes in beer using solid-phase microextraction with on-fiberderivatization and gas chromatography/mass spectrometry. Journal ofAgricultural and Food Chemistry, 51(24), 6941–6944.

Wei, A., Mura, K., & Shibamoto, T. (2001). Antioxidative activity of volatile chemicalsextracted from beer. Journal of Agricultural and Food Chemistry, 49(8),4097–4101.

Williams, R. S., & Wagner, H. P. (1979). Contribution of hop bitter substances to beerstaling mechanisms. Journal of the American Society of Brewing Chemists, 37, 13–19.

Yang, G., Schwarz, P. B., & Vick, B. A. (1993). Purification and characterization oflipoxygenase isoenzymes in germinated barley. Cereal Chemistry, 70, 589–595.