INTRODUCTIONFormation of the flavor profile of cognac and brandyis a complex multistage process. Their aroma, taste,and color depend on too many factors, including thequality of raw materials, the technology of fermentationand distillation, etc. One of the most important factorsis the aging in oak casks: its time and conditions areresponsible for the numerous transformations of organiccompounds, such as extraction, synthesis, biosynthesis,oxidation, etc. [1].Different classes of compounds contribute tothe formation of the sensory profile of cognacs withdifferent aging periods (Tables 1 and 2) [2–9].Figure 1 shows descriptors that make up the sensoryprofile of cognac [15].The gustatory sensation formation is a complexprocess, where a single shade of flavor may result froma whole complex of compounds [16]. People are ableto perceive five basic tastes: sweet, sour, bitter, salty,and “umami”, which was discovered in the early XXcentury.In fact, the taste sensation forms in the brain asprotein structures trigger its response to a combinationof external stimuli. Several sensory stimuli shapeperceptions from several descriptors. For instance,spicy tones are formed by compounds of mustard andpepper because carbon dioxide is responsible for thistaste. Fresh tones depend on several compounds of plantraw materials, e.g. mint, or on individual substances,e.g. xylitol. A sense of astringency appears when salivaproteins interact with food polyphenols. How panelistsevaluate one particular descriptor depends on a complexof organic compounds that enhance or minimizetheir effect on taste receptors due to spatial stereoisomerism,etc. [17].Table 1 Compounds that affect the sensory profile of cognacsCompounds Source Effect on cognac qualityFatty alcohols Amino acids of raw materialsduring fermentationResinous, honey, floral, and ripe fruity tonesEthers Raw materials; fermentationand aging in oak casksFruity tones; ethyl acetate is responsible for floraland anis aromaAldehydes and ketones Raw materials; fermentationand aging in oak casksUnpleasant unripe tones; nutty and floral tonesNorizoprenoids and terpenes Fermentation of plantraw materialsResinous tones, e.g. myrcene; fruity and floral tones;caryophylenes are responsible for the tone of cedar pine nutsPolyphenols and phenols Oak wood during aging A wide range of flavors and colors Therefore, the present research objective was tostudy the effect that compounds in cognacs of differentages produce on the intensity of perception of individualdescriptors in order to reveal extra quality assessmentcriteria.STUDY OBJECTS AND METHODSThe present research featured cognac samples ofvarious ages purchased in a network supermarket.Cognacs were stored in a dark room at 20 ± 1°С.The reduced extract was analyzed by distillationfollowed by a pycnometric analysis of solids [18].The pH of the samples was measured insevenplicates using a pH meter (METTLER TOLEDO,USA).The list of phenolic and furan compounds includedgallic, syringic, vanilla and sinapic acids, vanillin,syringaldehyde, coniferaldehyde, sinapaldehyde,5-hydroxymetifurfural, furfural, and 5-methylfurfural.Their content was analyzed by high performance liquidchromatography (HPLC) using a diode array detectorAgilent Technologies 1200 (Agilent, USA). We also useda Hypersil 5 um C18 250×4.6 mm column (Thermo,USA) with wavelengths of 270 and 310 nm. The test  samples and standards (0.02 cm3) were introduced in areversed-phase column at 40°C. The mobile phase wasrepresented by a 0.025 mol/dm3 solution of potassiumdihydrogen phosphate (A) with pH = 2.5, and a solutionof acetonitrile (B) in the ratio of A:B = 87:13. The elutionrate was 1.3 cm3/min.The mass concentration of sugars, i.e. fructose,glucose, and sucrose, was analyzed by HPLC using anAgilent Technologies 1200 diode array detector (Agilent,USA). A Hypersil 5 um C18 250×4.6 mm column(Thermo, USA) had wavelengths of 440 and 540 nm.The test samples and standard solutions were injectedin a volume of 0.02 cm3 of a reversed-phase column at40°C. The mobile phase was represented by distilledwater (A) and acetonitrile solution (B) in the ratio ofA:B = 87:13. The elution rate was 600 cm3/min.The mass concentration of higher alcohols,ethers, and hydrocarbons was assessed using gaschromatography (HPHC). A flame ionizationdetector (GC-FID) was used to detect variousvolatile components, including methanol, ethanol, 1-,2-propanol, 1-, 2-butanol, isobutanol, isoamilol, hexanol,phenylethyl alcohol, acetaldehyde, isobutyl aldehyde,acetone, ethyl formate, diethyl formate, ethyl acetate,isoamyl acetate, ethyl caproate, ethyl lactate, ethylcaprylate, ethyl caprate, guaiacol, and eugenol. Theanalysis also involved such non-volatile components aso-cresol, tyrosol, myrcene, and β-terpineol.All measurements were conducted in sevenplicates,standard deviation ≤ 5%. Each sample in the volumeof 5 cm3 (40% vol.) was added to 0.25 cm3 of internalstandard solution and placed in 2 cm3 vials. Eachcomponent was introduced at a concentration of2 g/dm3 in absolute alcohol. The vials were hermeticallysealed. A sample of 0.002 cm3 was introduced intothe chromatograph inlet. The column thermostat temperature was 220°C, and the carrier gas velocity was1.3 cm3/min.The sensory evaluation of the cognac samplesinvolved seven panelists with an extensive experiencein cognac industry and sensory tests. The panelistsworked in separate booths, isolated from externalfactors. The cognac samples were served chilledto 18 ± 1°C in testing glasses at room temperature20 ± 1°C under white diffused light. The samples wereevaluated according to set of descriptors in comparisonwith the reference sample. The result was expressedin points from 0 to 10 (0 – impossible to evaluate;1–2 – unsatisfactory (demonstrates a severe flaw);3–4 – satisfactory (demonstrates an obvious flaw);5–6 – satisfactory (violates the quality standard); 7–8 –very satisfactory (slightly violates the quality standard);9–10 – excellent (corresponds with the quality standard).The statistical analysis was performed insevenplicates. The descriptive statistics and valueswere expressed as mean ± standard deviation (SD). TheStudent-Fisher method provided multivariate modelsof the correlation and regression dependence of theparameters. The reliability limit of the obtained data(P ≥ 0.95) was used to assess various factors thataffected the content of polyphenols in all theexperiments. The obtained statistical data wereprocessed using the Statistics program (MicrosoftCorporation, Redmond, WA, USA, 2006).RESULTS AND DISCUSSIONTables 3–5 show the content of ethanol, reducedextract, carbohydrates, volatiles, furans, and phenols inthe cognac samples of different ages.The data are representatives of seven independentexperiments, and values are expressed in mean (± SD).Table 3 shows that the content of ethanol stayedwithin the permissible values for cognac productsspecified in State Standard 31732-2014 “Brandy. Generalspecifications” and did not fall below 40.0 ± 0.3% or4.00 ± 0.03 g/dm3. The content of volatile compoundsin the samples increased together with the aging time,which correlates with the previously published scientificdata [19, 20].Table 3 clearly demonstrates that the active aciditydecreased insignificantly as the aging time increased.The total acidity index depended on the origin of thewood. However, it increased as a result of long-termaging in oak casks due to the oxidation of ethanol ascompounds passed from the wood to the cognac [21, 22].The pH value is known to depend on the amount ofacids and the strength of the distillate. As the contentof alcohol in the distillate increases, the dissociationof carboxyl groups decreases, and acidity drops. Astannins dissolve, volatile acids appear, and the strengthdecreases during aging, the pH decreases [23]. ThepH value also depends on the amount of dissolvedtannins with an acidic pH, which increases the acidityof the distillates [23]. The experimental data in Table 5confirmed these trends.The data are representatives of seven independentexperiments, and values are expressed in mean (± SD). Table 4 shows that the content of volatile fractionsin the cognac samples increased together with theaging period, as reported in [19]. The total of higheralcohols was 1.774–2.092 mg/dm3. A longer aging periodtriggered the process of oxidation in higher alcohols(Table 4). Since the content of these alcohols in thecognac distillate was low, the oxidation of each alcoholwas insignificant, in comparison with the oxidativeprocesses of ethyl alcohol. The amounts of aldehydes,acids, and ethers formed by higher alcohols were alsoinsignificant. Nevertheless, even in such small quantitiesthat are elusive for conventional analysis methods,these substances still affect the taste of cognac due tothe sheer fact of their existence [24–26]. If the cognaccomposition is well-balanced, higher alcohols form thebasis of its sensory profile [27].Undesirable tones may result from excessiveacetaldehyde that form during oxidation, especiallyin the samples with a longer aging period, dependingon the characteristics of oak wood [22, 28]. However,if other volatile compounds are present, the excessiveacetaldehyde in these samples does not disrupt the tastebalance.Ethers also affect the flavor profile of cognacs. Theircontent depends on the aging time [29]. If ethyl acetateexceeds the sensitivity threshold (180 mg/dm3), it affectsthe sensory profile of the distillate, giving it undesirabletones [30].The data are representatives of seven independentexperiments, and values are expressed in mean (± SD).Table 5 shows that the cognac samples containedtypical phenolic acids and aldehydes in quantities thatdid not exceed those featured in research publications forcognacs of 2.5–15 years of aging [31–33]. The content ofsyringaldehyde is a marker of aging time. It was in therange of 2.5–7.7 mg/dm3 and increased with aging time,which corresponded with scientific publications on thischemical substance and other simple phenolic acids andaldehydes [31].Table 5 illustrates the ratio of syringaldehyde andvanillin, which is also a marker of aging time. This ratiostayed within the range of 2–4, established for collectionsamples, and was 2.4–2.5 [31, 34].Phenolic acids are involved in the complexbiochemical processes of aging and affect thesensory profile of cognacs [35]. For instance, gallicacid, a product of hydrolysis of soluble gallotanninsand eluggotannins of oak wood, affects the agingprocesses, acts as an oxidation catalyst, and removessulfides [36, 37]. As alcohol comes in contact with oak bark duringaging, it triggers solubilization with the subsequentcleavage of the covalent alkylaryl ether. This reactionleads to the cleavage of lignins and produces vanillin,syringaldehyde, and their acids, which affect the tasteprofile of cognac distillates [38]. Table 5 shows thatphenolic acids and aldehydes increased with aging,which is consistent with the previously publishedresearch data [38].Furan compounds appear as the temperatureincreases during the decomposition of non-starchpolysaccharides of oak bark or during distillationfrom five-membered sugars [39]. The amount offuran compounds is known to affect the number ofdistillations [26]. The content of furan compoundsincreased after a prolonged contact of oak bark andcognac distillate.Reducing sugars, i.e. glucose, arabinose, andfructose, were also registered in the distillate samples.During aging, the contact of alcohol and oak wood ledto the hydrolysis of hemicelluloses and hydrolyzabletannins [40]. Sugars affected the sensory profile ofcognacs, and their quantity increased with aging(Table 5).Volatile phenolic compounds, phenols, and terpenecompounds are responsible for some characteristictones in the cognac bouquet. The content ofphenolic compounds increases with aging, while theconcentration of terpene compounds decreases as aresult of their lability (Table 5).The cognac samples underwent a sensory evaluation(Table 6) using the descriptors presented in Fig. 2, whichdemonstrates how certain organic compounds composeparticular descriptors.In low alcohol drinks, bitterness is known to dependon alcohol content [41]. This study proved that bitternessdepends not only on aliphatic alcohols, but also onphenolic compounds.Aldehydes are responsible for mildness [42].However, aliphatic alcohols with their different tonesalso might help make the taste of cognac milder, and thecontent of o-cresol might also produce a certain effecton the mildness [2]. Astringency appears when phenoliccompounds are released during aging as a result ofcontact with oak wood, depending on the aging time andpH [5, 6].The resinous tones result from the combined actionof organic compounds in the distillate; it defines thequality of the finished product [33]. This descriptor isformed during fermentation, distillation, and aging[1, 33]. As a result, resinousness may depend on thecontent of aliphatic alcohols, phenolic compounds, andterpenoids [5, 7, 33].Oiliness, another cognac descriptor, appears mainlydue to secondary fermentation products that remainafter distillation, and partly due to the contact of alcoholwith oak [33]. Fruity tones depend on such secondaryfermentation products as aldehydes and alcohols, as wellas on terpene compounds, which is associated with thefermentation of fruit raw materials [9].Chocolate tones are more difficult to form than therest of the descriptors. Chocolate tones are known todepend on secondary fermentation products, volatilephenolic compounds, vanillin, and methylfurfural, thelatter also being responsible for sweet-nutty tones [26].The intensity indicators for each descriptor werequantitatively correlated with the results of the sensoryevaluation (Tables 3–5). They were processed in orderto obtain correlation and regression equations that madeit possible to calculate the dependence of the tones onparticular compounds (Table 7).The values of the coefficients were analyzed inmodulus in each group of the dependencies (Y1, Y2,Y3, Y4, Y5, Y6 and Y7) and the variables. In group Y1, thevariables at X3 had a larger coefficient because o-cresolhad a greater effect on descriptor Y1; in groups Y2 and Y3,phenolic alcohols contributed; in Y4 – volatile phenoliccompounds and aldehydes; in Y5 – oxymethylfurfural;inY6 – terpene compounds, and in Y7 – vanillin.Table 8 demonstrates equations for the dependenceof the compounds (X) on the aging period (Yx) obtainedby the method of pair linear correlation.The greatest value belonged to variable X3.Therefore, the change in the content of volatilephenolic compounds affected the sensory profile ofthe cognac samples more than other compounds.Probably, descriptor groups Y4 (resinousness) andY7 (chocolate tone) had a greater affect on the tasteperception in comparison with other descriptor groups.Phenolic compounds, i.e. acids, aldehydes, alcohols,and volatile compounds, were especially important forthe development of the sensory profile of the cognacsamples.CONCLUSIONThe correlation and regression analysis made itpossible to assess the role of v Figure 2 Descriptors for sensory evaluation of cognac samples arious organic compoundso-cresol[X3]Ethers[X3]Ethers[X2]Fruity tone [Y6]Chocolate tone [Y7]in the development of taste profiles for cognac samplesof different ages. The paper introduced equations ofmultivariate models that describe the effect of organiccompounds on the descriptors of cognac products.Linear regression equations revealed that phenoliccompounds of various classes played a major role in thetaste profile formation. The obtained data will make itpossible to form a list of additional criteria for sensoryevaluation of cognac products.CONTRIBUTIONM.N. Eliseev supervised the research project.O.A. Kosareva developed the research plan,I.N. Gribkova and O.M. Alexeyeva performed theexperimental research, obtained the data, and analyzedthem.CONFLICT OF INTERESTThe authors declare that there is not conflict ofinterests regarding the publication of this article.