Volatile Composition and Sensory Attributes of Smoothies Based on Pomegranate Juice and Mediterranean Fruit Purées (Fig, Jujube and Quince)

To increase the intake of fruits and vegetables—especially among young people—the food industry is trying to develop new, easy-to-eat and long-shelf-life products, such as smoothies. Nowadays, consumers are choosing their foods based not only on nutritional/functional properties (content of polyphenols, vitamins, minerals, among others), but also on sensory attributes. The aim of this study was to investigate the volatile composition by HS-SPME and the sensory profile by descriptive sensory analysis of novel smoothies prepared by blending fig, jujube or quince purée with pomegranate juices (cv. Mollar de Elche or Wonderful) at two ratios purée:juice (40:60 or 60:40). Twenty-three volatile compounds were identified by GC-MS and classified as alcohols, aldehydes, esters, furans, ketones, terpenes and terpenoids. Among volatile compounds, the five predominant ones in the studied smoothies were: (i) 5-HMF (30.6%); (ii) 3-hexen-1-ol (9.87%); (iii) hexanal (9.43%); (iv) 1-hexanol (8.54%); and (v) 3-octanone (7.67%). Fig smoothies were sweet and had flavor and volatiles related to fig, pomegranate, and grape. While jujube products were bitter and had jujube and pear notes. Finally, quince smoothies were consistent, sour and had quince, apple and floral notes. Thus, the type of fruit purée used clearly determined the flavor of the final product. The smoothies prepared with Mollar de Elche pomegranate juice were characterized by having high intensity of pear odor/aroma and consistency, and the Wonderful smoothies were characterized by lower consistency and more intense pomegranate aroma and sourness.


Introduction
Nutrition is the most important external factor influencing children's development. Its influence is essential from birth through childhood. Consumption of fruit and vegetables is important as it plays an essential role in preventing childhood obesity, and preventing many diseases, including certain cancers, osteoporosis, diabetes, coronary heart disease, stroke, neuronal degeneration, and type II diabetes [1][2][3]. Therefore, the World Health Organization (WHO) recommends eating a minimum of 400 g of fruits and vegetables per day to improve health and prevent the above-mentioned chronic diseases. Because current consumption is lower than the recommended intake, the development of easy-to-eat fruit-based products such as smoothies could be a good option. Despite of the natural sugar content, smoothies could reach the promotion into the children's diet [4,5]. Smoothies are beverages containing a blend of fruit pulp, fruit juice, ice, yoghurt, and/or milk. They are becoming a so popular way to consume fruits, especially among young people. These products are typically purchased freshly prepared from juice bars or as a processed product (mildly pasteurized) from the refrigerated section of retail outlets. Even after the economic crisis of 2007-2008, smoothies remained a popular and convenient way of consuming fruit [6]. Fruit components of smoothies could be considered as natural foods because of their nutrient profile or health-protecting qualities [7].
The southeastern part of Spain is one of the most intensively Mediterranean agricultural areas dominated by fruit orchards and vegetables fields. Some of these crops grown in this area, include fruits with appropriate characteristics for developing new products (organoleptic and functional properties), but underutilized, such as figs (Ficus carica), jujubes (Ziziphus jujuba) and quinces (Cydonia oblonga). It was recently found that the blend of pomegranate juice with the above mentioned fruits purée seemed a great opportunity to promote their use in an easy, sustainable and healthy way [5]. A positive effect of the addition of fig, jujube, and quince purée was observed in the nutritional and functionality of the novel pomegranate smoothies. For example, the addition of jujube contributed to an enrichment of the final smoothies in terms of vitamin C and organic acid content, while an increase of pectin content was found in fig and quince pomegranate-based smoothies [5].
On the other hand, the high interest in consuming fruit and vegetable products clearly shows that consumers are choosing products based not only on nutritional/functional properties (content of polyphenols, vitamins, minerals, among others), but also on sensory attributes (taste, smell, appearance, or even satisfaction) [8]. For this reason, fruit smoothies have become popular among health-conscious consumers and are among the major sources of bioactive compounds in daily diet [9,10]. Therefore, after knowing the nutritional and functional quality of the smoothies previously developed, the next steps should be to evaluate their volatile compositions and sensory profiles.
Taking all above mentioned into account, the aim of this study was to study the volatile composition and sensory profile of 12 different smoothies prepared using pomegranate juice (from 2 cultivars, cv., Mollar de Elche and Wonderful) and purée of Mediterranean crops (figs, jujubes and quinces) at different ratios purée:juice.

Plant Material
Pomegranates (Punica granatum) cv. Mollar de Elche and Wonderful, figs (Ficus carica) cv. Colar, jujubes (Ziziphus jujuba), cv. Grandes de Albatera and quinces (Cydonia oblonga) cv. Gigante de Vranja were hand-harvested in between mid-August and mid-October 2016 at a commercial maturity stage. The different stages of the smoothie preparation process as well as the ratio of purée:juice were previously described by Cano-Lamadrid et al. [5]. Briefly, the stages of the smoothie preparation process were: i.
Purée preparation: figs (F), jujubes (J), or quinces (Q) were peeled, ground, and heated at 80 • C in a Thermomix device (Vorwerk, Wuppertal, Germany); 10 mL of rhubarb juice per 1 kg of fruit were added to prevent enzymatic browning of the fruit due to the high oxalic acid concentration which chelates copper from the active site of polyphenol oxidase (PPO) [11,12]. After, the particle size of the mixture was reduced in a blender (Symbio, Zelmer, Rzeszów, Poland) until getting a thin purée. Then, the samples were cooled to room temperature. ii.
Pomegranate juice preparation: pomegranate fruits (Mollar de Elche and Wonderful, PM and PW, respectively) were cut in halves, and arils were manually separated from the husk and ground in a Thermomix to obtain the pomegranate juices. iii.

Volatile Compounds
Volatile composition of the samples under analysis was obtained by headspace solid phase microextraction (HS-SPME). Five g of each smoothie + 10 mL ultrapure water were placed into 50-mL vials with polypropylene caps and PTFE/silicone septa. A magnetic stirring bar was added, together with NaCl (15%) and the vial was placed in a water bath with controlled temperature and automatic stirring. A 50/30 µm DVB/CAR/PDMS fiber (Supelco, Bellefonte, PA, USA) (high capacity of trapping fruit volatile compounds) was exposed to the sample headspace for 50 min at 40 • C to simulate the mouth temperature during the chewing process. Desorption of the volatile compounds from the fiber coating was carried out in the injection port of the GC-MS for 3 min.
The identification and semiquantification of the volatile compounds was performed on a gas chromatograph (GC-MS), Shimadzu GC-17A (Shimadzu Corporation, Kyoto, Japan), coupled with a Shimadzu mass spectrometer detector GC-MS QP-5050A. The chromatographic set up and conditions were identical to those recently reported, with the only exception that the column used was a Restek Rxi-1301 Sil MS (Restek Corporation, Palo Alto, USA) of 30 m × 0.25 mm internal diameter, 0.25-µm film thickness. Analyses were carried out using helium as carrier gas at a flow rate of 6 mL min −1 in a split ratio of 6 and a program: (a) initial temperature 80 • C; (b) rate of 3.0 • C min −1 to 210 • C and hold for 1 min; (c) rate of 25 • C min −1 from 210 to 300 • C and hold for 3 min. Injector and detector temperatures were held at 230 and 300 • C, respectively.
Most of the compounds were simultaneously identified by using 3 different analytical methods: (1) retention indices, (2) GC-MS retention times (authentic chemicals), and (3) mass spectra (authentic chemicals and Wiley spectral library collection). Identification was considered tentative when it was based only on mass spectral data. The volatile composition analysis was run in triplicate and results were expressed as percentage of the total area represented by each one of the volatile compounds.

Sensory Evaluation with Trained Panel
Eight trained panelists (aged 30 to 55 years; 4 females and 4 males) from the Escuela Politécnica Superior de Orihuela (EPSO), Universidad Miguel Hernández de Elche (UMH) with more than 500 h of training in sensory testing participated in this study. The panel was selected and trained following the ISO standard 8586-1 (1993), and it is specialized in descriptive sensory evaluation of pomegranate products [13][14][15][16]. For the present study, the panel worked during 2 orientation sessions (90 min for each one) discussing the main organoleptic characteristics of commercial smoothies and fruit-based: pomegranate, figs, jujube, and quinces. The lexicon used for describing the flavor and texture attributes was based on the previously developed by other authors [17,18]. Both lexicons were adapted for smoothies based on pomegranate during the orientation sessions (Table 1). Samples were served into odor-free, disposable 90 mL covered plastic cups, at room temperature and were coded using 3 digit numbers as previous studies indicated [15]. Unsalted crackers and distilled water were provided to panelists to clean their palates between samples. The panel used a continuous numeric scale (0-10) for quantifying the intensity of smoothie attributes, where 0 represents none and 10 extremely strong, with 0.5-unit increments.

Statistical Analysis
Data were subjected to analysis of variance (ANOVA), after checking the normality and homogeneity of the variance, and later to Tukey's multiple-range test to compare the means. Differences were considered statistically significant at p < 0.05. All statistical analyses were performed using Statgraphics Plus 5.0 software (Manugistics, Inc., Rockville, MD, USA). Instrumental parameters correlated with sensory descriptors were used to perform a principal component analysis (PCA regression map) and a dendrogram analysis using XLSTAT Premium 2016 (Microsoft Corporation, Redmond, WA, USA). Euclidean distance by Ward method was performed for the dendrograms of clusters.

Volatile Profile and Composition
HS-SPME is a standard method used for the isolation of volatile compounds; and it is considered as an environmentally friendly technique (due to no solvents are used), selective and very sensitive [19]. This technique has been successfully used to establish the volatile profiles of different matrices such as herbs, wines, vegetables, and fruits [20]. Volatile composition has been investigated in different pomegranate products, such as pomegranate juice [21,22] and dehydrated pomegranate arils [16]. It was also studied in different fruits such as jujube [23], quinces [24] and figs [25]. However, the present study is the first one evaluating the combined effects of two factors (pomegranate cultivar and ratio purée:juice) on the volatile profile of smoothies blended with different Mediterranean fruits. Table 1 shows the retention indices used for the identification of the compounds, together with the main sensory descriptors of each of the volatiles. Twenty-three volatile compounds were isolated, identified, and their relative abundance determined in the pomegranate smoothies blended with fig, jujube and quince purée samples using this method. Previously, 12 and 14 different compounds were identified in the PW and PM juices [22], indicating that the addition of Mediterranean fruits increased the volatile profile on pomegranate products. Among identified volatile compounds (Table 2), several common compounds were previously detected in Mollar de Elche and Wonderful pomegranate juices (V1, V3, V4, V7, V11, V13, V15, V17, V18, V19 and V20) [22], and in heat treated pomegranate-based products (dried arils) (V1, V3, V4, V13, V17, V18 and V21) [16]. On the other hand, V4, V6, V11, V12, V13, V14, V15, V18 and V19 were previously identified in quinces fruits [24], while V1, V3, V4, V8, V11, V12,V14, V14, V17 and V19 were identified in previous studies of jujube fruits [23,26,27]. As to figs fruits and dried figs, V1, V2, V4, V5, V6, V8, V10, V11, V13, V14, V15, V16 and V18 were already detected [25,28]. The combinations of different fruit matrix and heat treatments for pasteurization could generate other volatile compounds not previously described and identified (V9, V22 and V23).  Eleven compounds were common in all 3 types of smoothies (fig, jujube, and quince), including for instance hexanal, furfural, 3-heptanone, hexyl acetate, linalool and HMF (Table 2).
At the beginning, statistics were preformed individually for each type of smoothie because of the completely different nature of the products under study, and later, the effect of purée fruit was also analyzed. Table 3 shows the relative abundance of the volatile compounds, grouped by chemical family, in the smoothies prepared with the 2 pomegranate cultivars (Mollar de Elche and Wonderful) and 3 fruits purée (figs, jujubes and quinces) for 2 ratios purée:juice (40:60 and 60:40). To make the discussion of this section easier, the pomegranate volatile compounds have been grouped into 7 chemical families: The main chemical groups of the pomegranate smoothies were: (i) furans, representing (32.8% ± 6.8%) of the total concentration of aroma compounds, followed by (ii) aldehydes (20.2% ± 4.0%), (iii) alcohols (19.3% ± 3.8%), (iv) ketones (10.8% ± 4.4%), (v) terpenoids (6.6% ± 2.7%), (vi) terpenes (5.9% ± 1.2%), and (vii) esters (4.6% ± 1.4%). It can be observed that volatile profile differed between the fresh pomegranate juices in which ALCs were the predominant family (67%, mainly 1-hexanol and 3-hexen-1-ol) in PW juices, while ALDs (30%) played an important role and were the most abundant chemical family in the PM juices [22].
The 5 predominant compounds in the studied smoothies were: (i) 5-HMF (mean for all samples 30.6%); (ii) 3-hexen-1-ol (9.87%); (iii) hexanal (9.43%); (iv) 1-hexanol (8.54%); and (v) 3-octanone (7.67%). The fact that the predominant compound was 5-HMF was unexpected. The furanic compound 5-HMF forms as an intermediate in the Maillard reaction between hexoses and amino components, and from direct dehydration of sugars under acidic conditions (caramelization) during thermal treatments applied to foods [29]. For instance, this compound is used as an indicator of the intensity of thermal treatment in honey [30]. In a previous study, 5-HMF and furfural were even found in the optimized dehydrated pomegranate arils [16]. However, the novel pomegranate smoothies highlighted by having less content than the above mentioned product. Without any doubt this compound is generated during the two heating steps of the smoothie preparation. The other 4 compounds are typical of fruits and fruit-based products; for instance, 3-hexen-1-ol, hexanal and 1-hexanol are key compounds of the peach flavor [31]. As 3-octanone was a predominant compound in the volatile composition of fresh wild mushrooms [32].
Hexanal, 3-hexen-1-ol and 3-octanone were more abundant in Mollar de Elche samples, while linalool and 5-HMF predominated in Wonderful smoothies (Table 3). It is worth mentioning that certain compounds can be used as an indicator of the fruit purée used in the smoothies. For instance, the volatile compounds exclusively identified in figs smoothies were: α-pinene, 1-octanol, ethyl octanoate, terpinen-4-ol and α-gurjunene. On the other hand, linalool oxide was exclusively present in the jujube smoothies. Quinces did not provide any exclusive compound to the smoothies. Formulation with a higher percentage of pomegranate juice (60%) led to higher abundance of 3-hexen-1-ol and linalool, while 60% of fruit purée increased the content of 3-octanone (Table 3).  followed by the same letter, within the same row for the same fruit purée, were not significantly different (p < 0.05), according to Tukey's least significant difference test; υ nd: not detected, meaning below the quantification limit.
As a brief summary of this section, it can be stated that Mollar de Elche pomegranate juice and smoothies prepared using fig and jujube purées were less sensitive to heat treatment than Wonderful and quince smoothies, as reflected by lower 5-HMF contents.
Recently, consumers' overall liking ("drivers of liking") of pomegranate-based products (dehydrated arils) was positively linked with the presence of aldehydes, esters, aliphatic alcohols and terpenes [16] which were presented in the novel developed products. Industry could use these liking drivers as quality indicators for improving their commercial and future novel smoothies.

Descriptive Sensory Analysis
An appropriate performance of the panel was observed with a good reproducibility by the end of the orientation sessions. Sixteen attributes (odor, basic tastes, flavor, texture, and defects) were used to fully describe the smoothies and are presented in Table 4. The smoothies prepared with Mollar de Elche pomegranate juice presented higher intensity of the fruit purée (F, J or Q) and pear odor, while the Wonderful smoothies had more intense pomegranate odor. The use of figs intensified the grape and pomegranate odor notes, while jujubes increased the pear notes. Regarding the basic tastes, Mollar de Elche pomegranate smoothies led to sweeter notes compared to the scores of Wonderful which were defined as sourer samples. With respect to the fruit purée, the use of figs intensified the sweetness (Table 4).
Regarding to flavor attributes, similar trends as odor attributes was observed. Wonderful samples led to higher scores of pomegranate compared to Mollar de Elche smoothies that presented higher intensities of fruit purée (F, J or Q). In accordance with our results, previous study indicated that the use of Wonderful juices in pomegranate products enhanced pomegranate notes [16].
The fiberness and consistency were the texture attributes evaluated and highly significant differences were observed, especially in the product fiberness. The use of jujubes and quinces led to the highest intensities of fiberness and consistency, respectively. Therefore, the type of fruit purée used had an important effect on fiberness, with jujubes leading to the highest intensity and quinces to intermediate values. Finally, it is important to highlight that the factor having the lowest influence on the sensory profile was the ratio purée:juice, which only influenced the consistency (with the 60:40 products being the most consistent ones). Recently, it was observed that the pectin content increased when using the highest content of fruit purée in smoothies [5]. Pectin has techno-functional characteristics which enhance texture of the smoothies by the reaction of certain water-soluble pectic substances with Ca ions to form some Ca pectates [5]. The purée:juice ratio can, therefore, be adjusted accordingly to the consistency preferred by the consumers. The price of the fruits used for the purée should be also be taken into consideration in order to produce a smoothie with high qualitative characteristics, but at affordable price.

Principal Component Analysis and Pearson's Correlations
A principal component analysis (PCA) was conducted to clearly see the relationships among the 12 smoothies, their volatile composition and sensory profile. Figure 1 shows that the first principal component (F1) explained 31.40% of the total data variance and the second one (F2) explained 18.56% of the total variance. Foods 2020, 9, x; doi: FOR PEER REVIEW www.mdpi.com/journal/foods

Principal Component Analysis and Pearson's Correlations
A principal component analysis (PCA) was conducted to clearly see the relationships among the 12 smoothies, their volatile composition and sensory profile. Figure 1 shows that the first principal component (F1) explained 31.40% of the total data variance and the second one (F2) explained 18.56% of the total variance. As can be seen, samples were grouped mainly according to the type of fruit purée in the smoothies, regardless of the purée:juice ratio and the pomegranate cultivar. Figure 1B shows how As can be seen, samples were grouped mainly according to the type of fruit purée in the smoothies, regardless of the purée:juice ratio and the pomegranate cultivar. Figure 1B shows how samples are grouped into 3 main clusters. Fig smoothies were characterized by high intensity of fig, pomegranate, cranberry, green, grape notes and sweetness, and with the following volatile compounds: V6, V14, V15, V19 V20, and V23, which present the following odor descriptors: woody, floral, citrus, and green. The relationship between volatile compounds and fig odor and aroma were backed up by significant values of the Pearson's correlations (Table 4). Jujube smoothies were characterized by high intensity of jujube and pear notes, fiberness and bitterness, and were associated with V8, V10, V13, and V16, being citrus and floral the main descriptors. Values above 0.7 for Pearson's correlations were found among jujube and these volatile compounds (Table 4). Finally, quince smoothies were characterized by high intensity of consistency and quince, green, apple and floral notes, and sourness, and were linked with V2, V4, V17 and V21. The descriptors of these volatiles are caramel, woody and floral.

Conclusions
The volatile composition and the sensory profile of novel smoothies prepared blending fig, jujube or quince purée with pomegranates juices (cv. Mollar de Elche or Wonderful) at two ratios of purée:juice (40:60 or 60:40) were studied. Twenty-three volatile compounds were identified, the five predominant ones: (i) 5-HMF (mean for all samples 30.6%); (ii) 3-hexen-1-ol (9.87%); (iii) hexanal (9.43%); (iv) 1-hexanol (8.54%); and (v) 3-octanone (7.67%). Fig smoothies were sweet and had flavor and volatiles related to fig, pomegranate, and grape. Meanwhile, jujube products were bitter and had jujube and pear notes. Finally, quince smoothies were sour and had quince, apple and floral notes. Thus, the type of fruit used clearly determined the flavor of the final product. The smoothies prepared with Mollar de Elche pomegranate juice were characterized by having high intensity of pear odor/aroma and consistency. While Wonderful smoothies were characterized by lower consistency and more intense pomegranate aroma and sour. However, further research is still needed to fully optimize these novel products. Two ways of improvement can be researched: (i) increasing pomegranate notes; and (ii) avoiding undesirable compounds after the Maillard Reaction. After knowing the volatile compounds and sensory profiles of these developed smoothies, it is worth to continue researching in this area studying the volatile composition of other types of smoothies and other types of pasteurization treatments to avoid undesirable compounds and aromas. To know the active odor compounds of these products also can be necessary in the future studies. Moreover, consumer studies should be carried out to know the drivers of smoothie's consumer acceptance.