We have studied the oxidation of furfural by hydrogen peroxide in the presence of sodium molybdate in detail for the first time. The reactions of furfural with hydrogen peroxide under the conditions of autocatalysis by the acids that accumulate [2][3][4][5][6][7] and catalysis by compounds of selenium and vanadium [8, 9] were reported earlier. The authors of these papers isolated ~-formylacrylic acid (I) or its esters [8, 9] and 2(5H)-furanone (II) and succinic acid (III) [6, 7] as the main products. It was assumed that oxidation took place either according to the classical Baeyer-Villiger mechanism [2, 7] or by hydroxylation [9], depending on the conditions. It was reported that compound (I), maleic acid (IV), and fumarie acid (V) and also a certain amount of malic acid (VI) were formed during the oxidation of furfural by a 5 % solution of hydrogen peroxide in the presence of NaHMoO 4 [10]. The authors consider that the reaction in the presence of sodium hydromolybdate takes place similarly to the reaction in the furfural-hydrogen peroxide-selenium compound system.As seen from the published literature, in papers on the peroxide oxidation of furfural [2-10] the possibility of using the peroxo complexes of molybdenum, which are easily formed in the aqueous hydrogen peroxide-molybdenum vl system, has not been realized [11]. In the light of known data on the epoxidizing ability of such ~umplexes [11][12][13][14][15][16][17][18][19][20] and in order to discover new synthetic possibilities in the reaction of furfural with hydrogen peroxide we studied its characteristics in the presence of sodium molybdate.The oxidation of furfural was conducted at 60~ with a 30% aqueous solution of hydrogen peroxide in the presence of catalytic amounts of Na2MoO 4 (Table 1, version 1). The pH of the initial reaction mixture was 5 on account of the stabilizer contained in the hydrogen peroxide. In the first 5-10 min its color changed from red-brown to lemon-yellow, which remained to the end of the process.The end of the reaction was detected by the transformation of the initial reagents and the formation of the initial and final compounds by means of a set of analytical methods, i.e., chromatography with authentic samples (GLC, HPLC, TLC, and paper chromatography), polarography, and titrimetric analysis (for analysis of the acidic and peroxide compounds). Kinetic curves for the consumption of furfural and hydrogen peroxide and the accumulation of furanone (II) and acids were obtained (Fig. 1).As seen from the obtained results (Table 1, version 1, and Fig. 1), the furanone (II) is mainly formed in the investigated system together with tartaric acid (VII), malic acid (VI), and succinic acid (III). A distinctive feature of the process *For Communication 3, see [1].

Three high quality red wines – Merlot, Cabernet Sauvignon and Pinot Noir – were used for development of an optimized formulation of a new blended wine called Zvezda Kubani (“The Star of Kuban”). The experimental plan was implemented with the mixture designs and triangular surfaces module in the STATISTICA package. According to the experimental plan, we made and studied 31 variants of wines, including 3 monovariants, 3 mixtures of 2 wines and 25 mixtures of 3 wines. In addition, highly qualified specialists have studied the changes in the mixtures according to the results of a sensory assessment to model the connection between the sensory perception of wine mixtures and the new blended wine formulation. As a result, we developed a mathematically proved formulation of a new blended wine, Zvezda Kubani , containing 48% Merlot, 35% Cabernet Sauvignon and 17% Pinot Noir. The experimental verification of the suggested composition of the blend proved to be a strong indicator of the experts' sensory assessment.

The objects of study were 150 samples of natural dry red and white grape wines of Russian origin, obtained by traditional technologies from European and hybrid grape varieties grown in wineries in Krasnodar Krai in 2010-2016. Natural red (Cabernet, Merlot) and white (Aligote, Riesling, Pinot Noir) (alcohol content of 9-13 % by volume, acidity of 4-7 g/dm3), as well as blend wines based on Cabernet Sauvignon, Merlot and Pinot Noir wines made under experimental conditions were analyzed. Chromatographic and electrophoretic methods were used to determine the content of volatile components and amino acids in the studied samples. A sensory assessment of wine quality was carried out by wine specialists working in the wine industry and having professional experience in the field of sensory analysis. Using statistical modeling we carried out a comparative assessment of the role of amino acids-threonine, proline, arginine and volatile compounds-methanol, acetic acid, furfural in the perception of taste and aromatic properties of wines, a general indicator of which is the average tasting rating. High adequacy of the regression model constructed using covariance analysis indicates that mainly amino acids and volatile compounds determine the sensory properties of wines. The dominant role of amino acids in the perception of taste and aromatic characteristics compared to other wine components is mathematically justified in accordance with the criterion of one-dimensional significance. It has been shown that more than 82% of the sensory characteristics of the analyzed wines group fall on the amino acids and volatile compounds under consideration, and less than 18%-on all the others, including titrated acids, free amino acids, mineral components, phenols, etc.

Поступила в редакцию 14 июня 2014 г., после исправлений -2 сентября 2014 г.Обсуждены вопросы оценки интегральной характеристики вин, получаемой в ходе уста-новления их качества, указаны достоинства и недостатки существующих процедур. Рассмо-трены способы инструментальной оценки качества вин, которые часто приводят к получению неоднозначных результатов. Показана возможность более простых решений оценки качества вин с использованием методов математического моделирования по определенным параме-трам дегустации. Проанализированы и установлены параметры, формирующие качество ви-ноградного вина, с использованием которых математическая модель способна дать дегуста-ционную оценку вина, позволяющую решать оперативные задачи и существенно сократить продолжительность исследования. Методами математической статистики установлено нали-чие взаимосвязей между дегустационной оценкой и содержаниями летучих компонентов, по-строено адекватное линейное статистически значимое уравнения регрессии, при помощи ко-торого возможно предсказать дегустационные оценки для вин высокого, среднего и низкого качества. С целью апробации предложенной модели предсказаны дегустационные оценки не-которых виноградных вин.Ключевые слова: вино, корреляция, статистика, модель, компонент, регрессия Якуба Юрий Федорович -канд. техн. наук, доцент, заведующий ЦКП ГНУ СКЗНИ-ИСиВ.Область научных интересов: высокоэффективный капиллярный электрофорез, капиллярная газовая хроматография, химия пищевых производств.Более 100 опубликованных работ, в том числе автор ряда монографий и патентов.Темердашев Зауаль Ахлоович -докт. хим. наук, профессор, зав. кафедрой анали-тической химии КубГУ.Область научных интересов: анализ объектов окружающей среды, разработка аналитических схем контроля.Более 180 опубликованных работ, в том числе автор ряда монографий и патентов.Халафян Алексан Альбертович -докт. техн. наук, профессор кафедры приклад-ной математики КубГУ.Область научных интересов: компьютерный анализ данных, построение и анализ вероятностно-статистических моделей.Более 100 опубликованных работ, в том числе автор нескольких монографий, учебников, свидетельств о регистрации программ. ВведениеОрганолептические свойства вин формируют-ся за счет взаимодействия ароматических и вкусо-вых характеристик, причем ароматические созда-ют по большей части летучие соединения, с другой стороны, вкусовые компоненты оказывают влияние на содержание летучих веществ в паровой фазе, их распределение и соотношение, что свидетель-ствует о неразрывности процесса дегустации [1]. Летучие компоненты вин, образующиеся в про-цессе алкогольного брожения виноградного сус-ла (спирты, альдегиды жирного и ароматического рядов, летучие кислоты, простые и сложные эфи-

No abstract