Spectrophotometric analysis was carried out using the Folin–Ciocalteu reagent to determine the mass concentration of phenolic substances. The Folin–Ciocalteu reagent was prepared by dissolving 100 g of sodium tungstate and 25 g of sodium molybdate in 700 mL of distilled water. Adding 50 mL of phosphoric acid 85% ($\rho$20 = 1.71 g/mL) and 100 mL of concentrated hydrochloric acid ($\rho$20 = 1.19 g/mL). Bring to the boil and reflux for 10 hours. Then, 150 g of lithium sulfate and a few drops of bromine were added, and the mixture was boiled for 15 minutes. After cooling, the solution volume was brought to 1 liter using distilled water [24].

Gallic acid, caffeic acid, malvidin-3-O-glucoside chloride, (+)-D-catechin, quercetin dihydrate, isoquercetin were purchased from Fluka Chemie (GmbH, Buchs, Switzerland), trans-resveratrol, and (–)-epicatechin, syringic acid were used as standard substances in determining the qualitative and quantitative compositions of phenolic compounds by high-performance liquid chromatography acid, p-coumaric acid, kaempferol, and ferulic acid were purchased from Sigma-Aldrich (GmbH, Munich, Germany).

The main types of white and red wines were prepared by complete or partial fermentation of pulp sugars, as well as the use of grape stems and stopping fermentation by alcoholization of various harvest years (1996–2018). A total of 215 samples were used as material in this study. Their production technology and age largely determined the choice of samples for analysis. Notably, the study included both red wines and white wines prepared by alcoholization with ethyl alcohol (fortified wines) and contact of the must with the solid parts of the grape bunch (grape stems and pulp), providing a complete extraction of biologically valuable compounds with a phenolic structure, and as a result, a higher extractivity of wines.

High absorption values of phenolic compounds contained in wines in the UV wavelength range by light energy determine the effectiveness of spectrophotometric determination of the concentration of phenolic substances in wines of various types and ages. Before measuring the absorption rate of wines, they were diluted 15, 30, and 60 times. Considering that the maximum accuracy of measurements is achieved with absorption values of 0.3–0.7 units [36] and the maximum value of the measured absorption value should not exceed 3 units—a weakening of the radiation intensity by 1000 times, and the concentration of phenolic substances in wines is the bulk of measurements varying from 200 mg/dm${}^3$ to 1500 mg/dm${}^3$, then in a sensor that records the absorption value, the optimal optical path length should be 0.05 mm $÷$ 0.07 mm${}^3$, the equivalent to diluting the sample 15 times. Absorption spectra were recorded in quartz cuvettes at wavelengths of 220–430 nm, with an optical path length of 1 cm after 10–20 nm, by a Specord 40 spectrometer (Analytic Jena, Jena, Germany) single-beam scanning spectrophotometer in automatic mode.

All phenolic substances contained in wine are oxidized by the Folin–Ciocalteu reagent. This reagent is formed from a mixture of phosphotungstic acid and phosphomolybdic acid, which, after oxidation of the phenols, is reduced to a mixture of blue oxides of tungsten and molybdenum.

The proportionality principle of the mass concentration of phenolic substances to the intensity of the absorption of the prepared reaction mixture, together with the properties of the Folin–Ciocalteu reagent that oxidizes phenolic substances, is the basis for determining the total content of phenolic compounds (TCPCs). The TCPCs were measured using the colorimetric method, as described in [24, 37], with the absorbance measured using a KFK-3M spectrophotometer (GazoAnalit, Smolensk, Russia).

Individual phenolic substances in the studied wine samples were determined by HPLC using an Agilent Technologies chromatographic system (model 1100) (Agilent Technologies, Waldbronn, Germany) with a diode array detector. To separate the substances, a Zorbax SB – C18 chromatographic column with a size of 2.1 $\times$ 150 mm was filled with silica gel containing a grafted octadecylsilyl phase with a sorbent particle size of 3.5 µm; chromatography was carried out in the gradient mode [37]. Mobile phase A contained methanol, while mobile phase B contained an aqueous solution of 0.6% trifluoroacetic acid. The volume of the injected sample was 10 µL. The gradient composition, reported as the proportion of mobile phase B, was programmed as follows: 0 min, 8%; 0–8 min, 8–38%; 8–24 min, 38–100%; 24–30 min, 100%. The mobile phase flow rate was 0.25 cm${}^3$/min.

Chromatograms were recorded at the following wavelengths:

- 280 nm: gallic acid, (+)-D-catechin, (-)-epicatechin

- 313 nm: derivatives of hydroxycinnamic acids

- 371 nm: quercetin

- 525 nm: anthocyanins

Individual compounds were identified by comparing their spectral characteristics with the spectra described in the literature [38, 39, 40, 41, 42, 43] and by comparing the retention calibration graphs of the dependence of the peak area on the concentration of the substance, constructed from solutions of standard substances. The content of anthocyanins was determined in terms of malvidin-3-O-glucoside chloride, caftaric acid, 2-S glutathionylcaftaric acid; in terms of caffeic acid and coutaric acid; in terms of p-coumaric acid and fertaric acid; in terms of ferulic acid and quercetin-3-O-glucuronide; in terms of isoquercitrin, polymeric, and oligomeric procyanidins; in terms of (+)-D-catechin.

The mass concentrations of phenolic substances determined by HPLC were summed to determine the total concentration of the monomeric forms of the phenolic compounds (TCMFPCs).

Distilled water was used as a reference solution to dilute the wine samples. This was considered appropriate since the pH of the solution in wine samples shows little change with dilution due to the presence of buffer components in wines. All analyses were carried out in triplicate.

The measurement results were processed using standard mathematical statistical methods using the “Data Analysis” package in Microsoft Excel. The reproducibility of measurements was at least 10%, with a confidence level of p = 0.95.

After examining the UV spectral data, it was found that most of the studied wines had a characteristic peak within 265–285 nm (peak I), although its exact position and shape depended on their extractivity and age. The characteristic absorption spectra for the different types of wine included in this study are shown in Fig. 1.

The absorption curves analysis (Fig. 1) showed that with an increase in the extractive and age of the wine, the position of the maxima of the absorption peaks in this area shifted towards higher wavelengths, demonstrated in the inset.

Table 1 shows the experimentally determined values of the absorption maximum for wines of various types and ages, along with their 95% confidence intervals.

The UV spectra data also demonstrated that as the age of wine increased, the shape of the peak at ca. 280 nm became less pronounced, while the absorbance in the wavelength region above 300 nm was increased. When grouped by age, there was a strong, positive correlation between the absorbance at a wavelength of 280 $\pm$ 2 nm and the concentration of the total phenolic compounds for wines of various types. This correlation is shown in Fig. 2 for young wines (1–2 years) and wines aged for 8–10 years.

Even though the peak shape of the maximum absorption of wine with aging becomes less pronounced (Fig. 1), the amount of light energy absorption at a wavelength of 280 nm per unit mass of phenolic substances increases, which can be seen in the greater slope between the phenolic concentration and absorbance (Fig. 2). The correlation coefficient between the absorption value and the total concentration of phenolic substances was 0.98. This can be explained by the different qualitative composition of phenolic substances in young and aged wines and by the fact that, despite the degradation of anthocyanins and a decrease in the total concentration of phenolic substances during aging, grape phenolic compounds and their derivatives remain in the aged wine and show strong absorbance in the 280 nm region.

This observation suggests the use of absorbance at 280 nm as a quality indicator by measuring the concentration of total phenolic compounds in young wine materials and estimating the chemical age of the wine material (based on the slope of the absorbance/concentration curve).

To generate absolute measurements of the light-absorbing ability of phenolic compounds, it is advisable to introduce the concept of the “absorption coefficient of phenolic substances of grapes and wine”. In analytical chemistry, the concept of the molar absorption (extinction) coefficient is usually used, which is equal to the absorbance of a solution with a concentration of 1 M in a cell with an optical path length of 1 cm [16]. However, because grape and wine matrices are not pure solutions and contain mixtures of various phenolic substances with various degrees of polymerization, it is necessary to use a different metric. We suggest using the absorbance of a solution at 280 nm and an optical path length of 1 cm, with the solution containing a concentration of 1 g/dm${}^3$ of total phenolic substances as determined by a certified method using the Folin–Ciocalteu reagent. This indicator of phenolic substances in wine materials should be measured in the absorbance range close to 0.434 by diluting the sample with distilled water. This value was chosen empirically by preliminary experimental trials.

The obtained absorbance value should then be multiplied by the wine dilution factor and divided by the concentration of the phenolic substances. Table 2 shows the calculated values of the absorption coefficient of phenolic substances in grapes, obtained based on statistical processing of spectrophotometric data on young wine materials at a 95% confidence level.

As mentioned, the physical meaning of this value is the absorbance of a solution containing 1 g/dm${}^3$ of phenolic substances determined using the Folin–Ciocalteu reagent according to the specified protocol [15], with an optical path length of 1 cm at a wavelength of 280 nm.

In this case, based on the above definition, the concentration of phenolic substances in the test sample will be equal to

$$\mathrm{C}=\frac{{\mathrm{D}}_{280}\times \mathrm{n}}{{\epsilon }_{280}}\times 1000,\mathrm{mg}/{\mathrm{dm}}^3$$

where D${}_{\mathrm{280}}$ is the absorbance of the wine material solution, measured at a wavelength of 280 nm in a quartz cell with an optical path length of 1 cm; n is the dilution factor of the wine material; $\epsilon$${}_{280}$ is the absorption coefficient of phenolic substances in grapes; 1000 is the conversion factor for obtaining data in mg/dm${}^3$.

Next, we considered the possibility of directly measuring the absorbance of wines without their preliminary dilution to create a rapid sensor for calculating the concentration of phenolic substances in grapes and wine based on the absorbance at 275–280 nm. The maximum accuracy of spectrophotometric measurements is achieved at absorbance values between 0.3 and 0.7 units, and the maximum measurable absorbance value should not exceed 3 units. With respect to a 1 cm path length cuvette, this is equivalent to diluting the sample by 15 times.

In the next stage of the study, we investigated the qualitative and quantitative compositions of phenolic substances in the main types of white and red grape wines, from young to aged. The total concentrations of monomeric forms of phenolic substances (measured by HPLC) and the total concentration of phenolic substances (measured using the Folin–Ciocalteu method) in the main types of white grape wines are presented in Table 3, Fig. 3.

From these results (Table 3, Fig. 3), it can be seen that the average total phenolic content in the white wines ranged from 0.26 g/dm${}^3$ in dry white wines less than 1 year old to 1.64 g/dm${}^3$ in dry white wines made using every component from a bunch of grapes (must, skins, seeds, and stems). For the monomeric phenolic content, the lowest value of 0.06 g/dm${}^3$ was also found for dry white wines less than 1 year old, while the highest monomeric content of 0.16 g/dm${}^3$ was identified in dry white wines made with pomace and stems. When analyzed by the proportion of phenolic substances that were monomeric forms, the lowest proportion (9.76%) was found in wines made using all components of the grape bunch (must, skins, seeds, and stems), while the highest proportion (24.10%) was seen in semi-dry and semi-sweet white wines. This difference appears to be due to the technology used in their production. The average total phenolic content in red wines ranged from 1.65 g/dm${}^3$ in red fortified wines over 8 years old to 3.33 g/dm${}^3$ in red fortified liqueur wines aged 1–2 years (Table 4).

For the sum of the monomeric phenolic substances, the lowest content (0.12 g/dm${}^3$) was found in fortified red wines (1–2 years), and the highest (0.31 g/dm${}^3$) in dry red wines less than 1 year of age (Fig. 4).

Further investigation into the monomeric phenolic constituents in the main types of white wines revealed the presence of phenolic acids (hydroxybenzoic and hydroxycinnamic acids), flavanols, and flavonols. The highest concentration of phenolic acids (118.31 mg/dm${}^3$) was found in dry, white wines prepared using every component of the grape bunch (must, skins, seeds, and stems), while the lowest value of 30.01 mg/dm${}^3$ was found in white fortified liqueur wines over 8 years of age (Table 5).

The flavanol content varied widely, from 12.36 mg/dm${}^3$ (in white fortified liqueur wines aged 1–2 years) to 36.07 mg/dm${}^3$ (in dry white wines made using the pomace and stem).

Flavonols were found in the lowest concentrations of all the monomeric phenolic substances, with contents ranging from trace amounts to 4.00 mg/dm${}^3$.

The same four classes of monomeric phenolics (hydroxybenzoic acids, hydroxycinnamic acids, flavanols, flavonols) were also found in red wines and anthocyanins (Table 6). The highest concentrations of phenolic acids (123.67 mg/dm${}^3$) were found in aged red dry wines (1–2 years of age), whereas the lowest concentrations (59.65 mg/dm${}^3$) were in red fortified liqueur wines, the age of which exceeded 8 years. The flavanol contents varied widely from 21.54 mg/dm${}^3$ (in red fortified liqueur wines over 8 years old) to 84.97 mg/dm${}^3$ (in dry red wines over 2 years old). The flavonol contents were generally higher than in white wines, with concentrations ranging from 1.81 mg/dm${}^3$ (in red fortified liqueur wines over 8 years old) to 23.52 mg/dm${}^3$ (in dry red wines aged no more than 1 year). Finally, anthocyanins, which were unique to the red wines, were found across the widest range of concentrations, from 2.13 mg/dm${}^3$ (in fortified red wines over 8 years old) to 184.26 mg/dm${}^3$ (in dry red wines less than 1 year old).