Study on the release and sensory perception of encapsulated d‐limonene flavor in crystal rock candy using the time–intensity analysis and HS‐GC/MS spectrometry

Abstract This research was conducted to evaluate encapsulated d‐limonene perception and release in rock candy. Microcapsules with wall materials of 75/25 of gum Arabic/Maltodextrin by 20% of wall materials) were produced for using in rock candy. To evaluate the flavor release from rock candy by time–intensity method, a model system was developed and time–intensity sensory evaluation was conducted by trained sensory panelists in order to determine the effect of three different matrices (water, water and flavored rock candy, and water with flavored rock candy and citric acid (pH = 3) at three serving temperatures (10, 45, and 75°C) on the perception of d‐limonene release. Results showed that release of d‐limonene from flavored rock candy with acid citric (pH = 3) at 75°C had the highest perceived sensation whereas the matrix of microcapsule in water at 10°C had the lowest perception. On the other hand, increasing the temperature from 10 to 75°C had significant effects on the release and perception of d‐limonene (p < .05). Headspace gas chromatography–mass spectrophotometry confirmed results from time–intensity sensory evaluation, which indicated that the release of d‐limonene increased in the presence of sucrose and citric acid (pH = 3).

effective method for protecting flavoring agent overprocessing and preserving the final product.
Microencapsulation is the packaging of small particles, which make up a core, with a film of continuous polymer. Flavoring agents are often sensitive to oxygen, heat, light or acid, and can be preserved through microencapsulation, which also allows a refcontrolled release of the core contents (Krishnan, Kshirsagar, & Singhal, 2005). This resultant increase in flavoring agent chemical stability is of great importance in the food industry, where hydrophobic flavors are incorporated into powders by encapsulation (Yoshii et al., 2001). The way in which these flavors are released, including how they can be released under control in foods, comprises an important aspect in the estimation of the potential storage period (Pszczola, 1998;Reineccius, 1995;Whorton, 1995). d-limonene is a major component of various citrus oils including lemon, lime and orange, and has a lemon-like smell which is commonly consumed with wide range of soft and alcoholic drinks. It is classified as a monocyclic monoterpene and emits a pleasing citrus scent. Being both a fragrant and a flavoring agent, d-limonene has widespread usage in soaps, foods, perfumes, chewing gum, and beverages (Li & Chiang, 2012). Over the production process of rock candy, microencapsulation of d-limonene ensures preservation against heat and other destructive factors.
A number of studies have shown the effects of tastants on perceived flavor intensity and most of them suggest an enhancement of flavor perception by sweetness and sourness (Bonnans & Noble, 1993;McBride & Johnson, 1987;Pfeiffer, Hort, Hollowood, & Taylor, 2006). Interactions occur both within and between modalities. An example of the former is how sugar subdues the sour taste of citric acid (Curtis, Stevens, & Lawless, 1984), and two examples of the latter are the impact of aroma on the perception of flavor (Pfeiffer et al., 2006).
Throughout the consumption of food, sensory perception changes due to its dynamic nature (Cliff & Heymann, 1993).
Processes such as breathing, salivation, chewing, tongue movement, and swallowing impact the sensory perception of food, moment by moment (Lawless & Heymann, 2010). In the past, conventional methods of measuring sensation were based on a static perception of food for a specific moment. However, methods to capture moment by moment sensory perception by employing dynamic techniques have been developed and are much closer to reality (Dijksterhuis & Piggott, 2000). So as a new product which contains d-limonene microcapsule in its structure, investigation of the time and place of interactions must be identified as they can affect how flavor is perceived, in addition to other characteristics, thereby modifying the products' sensory profile (Hewson, Hollowood, Chandra, & Hort, 2008). Such interactions may take place at different levels, including physical interactions between constituents which lead to fluctuations in volatile release (Da Porto, Cordaro, & Marcassa, 2006).
Sugars are able to affect how flavor is released from some flavoring agents, altering their volatility; furthermore, the effect of a change in pH on flavoring agent release has also been investigated (Friel, Linforth, & Taylor, 2000;Hansson, Andersson, & LeufveÂn, 2001a). The aim of the present study were (a) to study d-limonene release from crystal rock candy, (b) to develop a model system to study flavor release from encapsulated molecules, and (c) to develop time-intensity method to evaluate flavor perception from a novel confectionery product.

Highlights
• The present study considered time-intensity as a dynamic sensory evaluation method to evaluate perception of encapsulated d-limonene in flavored rock candy.
• In this research, a model system containing sucrose and citric acid was developed to better perception of microencapsulated d-limonene release.
• Headspace GC/MS was applied to scrutinize the release of encapsulated d-limonene in different matrices and temperatures.
• Higher temperature was shown to be more effective on perception and release of d-limonene.
F I G U R E 1 Stick Rock candies | 935 VATANKHAH LOTFABADI eT AL.

| Materials
d-limonene (ρ = 0.84g/cm 3 at 20°C, Molar Mass: 136.24 g/mol) was purchased from Nacalai Tesque. Gum Arabic and Maltodextrin .5) were supplied from SDFCL and Sigma-Aldrich, respectively. Sugar was prepared from Paniz Shahd Binalood Co. Citric acid was obtained from Sigma-Aldrich Chemistry. Distilled water was used for preparing all solutions. All organic chemicals used in the analyses were of the analytical grade.

| Preparation of flavored rock candy
The production of rock candy was performed by using sugar and water; first, water was heated up to 100°C, and then sugar was gradually added up to a level of total solution by constant agitating; Heating would be continue until the brix of solution reach to 83; then, microcapsule (d-limonene was encapsulated with mixture of gum Arabic and maltodextrin at ratio of 3:1 by spray drying; total solid content of wall materials was 20% w/w; Vatankhah Lotfabadi, Mortazavi, Yeganehzad, & Sadeghian, 2018) by amount of 0.04 g for each rock candy added to the syrup and then flavored syrup was incubated so that the crystallization occurred over 16 hr inside the oven by gradual descending temperature from 80 to 40°C. The final product has 10 g weights for following sensory evaluation.

| Training session
In this research, time-intensity method was developed to evaluate flavor release from rock candy; therefore, to better perform the test, a model system was developed (as described in 2.3.2). Five 1-hr sessions were conducted in order to train and acquaint panelists with the time-intensity methodology in developed model system.
For data acquisition and data analysis, the Sensomaker program (Kiumarsi et al., 2019;Nunes & Pinheiro, 2012) was used. Through a graphical interface in the form of 10-point scale, with 0 meaning no perception and 10 signifying an extreme perception of d-limonene, each panelist indicated the intensity of the attribute of sample with monadic presentation, using complete block design (Wakeling & Macfie, 1995).

| Flavor perception using time-intensity (TI) analysis
The sensory panel consisted of ten trained judges (five males and five females; age ranging 26-40 years) from the Research and Development Department of Saharkhiz Company and Research Institute of Food Science and Technology (RIFST), who were all skilled in food and beverage sensory evaluation. The attribute evaluated was d-limonene flavor, and the samples were presented in a monadic way, using a balanced complete block design (Wakeling & MacFie, 1995).

| Experimental procedure
The nine samples comprised d-limonene microcapsule in water at each of 10, 45, and 75°C, flavored rock candy in water at each of 10, 45, and 75°C, and flavored rock candy in water and citric acid at each of 10, 45, and 75°C. All samples were labeled with randomized three-digit codes and were presented in 30 ml proportions of solutions; sensory evaluation was conducted in an air-conditioned room (20°C). The panelists began evaluating by clicking on a "start" button and consumed 10 ml of sample (a gulp of the sample over 2 s), during the next 60 s, the panelists began evaluating by clicking on a "start" button and consumed a gulp of the sample over 2 s. During the next 20 s, using the mouse, the panelists indicated the perceived intensity of d-limonene on the scale (Table 1). At the end of the analysis, the evaluation stopped after the panelists reached the left end of the scale after 60 s. a message indicating the end of the test appeared and the panelists rinse their mouths with mineral water in order to prepare for the next sample. All samples were presented in randomized order and two replications. The time-intensity course was characterized by time when intensity is 90% of Imax at increasing part of the curve, area under the curve, and plateau which indicates on time interval which the intensity is ≥90% of Imax.

| Headspace gas chromatography-mass spectrometry
Static headspace analysis followed by gas chromatography-mass spectrometry (GC-MS) analyses was done on flavored rock candy at 10°C, 45°C, and 75°C.
All analyses were carried out in duplicates; samples were prepared in 30-ml volumetric flasks with valve caps; After completely dissolve crystal rock candy, 1,000 µl of the vapor phase from each sample flask at each temperature was injected into GC-Agilent

| Statistical analysis
An analysis of variance (ANOVA) was applied for the scores of each panelist and for selected parameters consisting of maximum intensity (Imax), running time of maximum intensity (plateau), and area under the curve (area); the probability level was p = .95. Tukey's test was applied to compare the averages of samples using the Sensomaker software (Pinheiro, Nunes, & Vietoris, 2013). The data of mean TI curves were presented in graphic form by using the Microsoft Excel 2012.

| Effect of matrix and temperature on dlimonene perception
The effect of sucrose, citric acid, and water as blank sample on the release of d-limonene was investigated at three serving temperatures which are usual temperatures for consumption of tea and cold drinks; intensity of d-limonene flavor was the only attribute analyzed by assessors, and time-intensity curves for d-limonene microcapsule in three matrixes and three temperatures (10, 45, and 75°C) are presented in Figure 2. Table 2 shows the analysis of variance (ANOVA) and Tukey's mean test. Figure 2 shows d-limonene intensity over time for samples with different temperatures and matrices. As observed in Figure 2, all of the samples had the same release trends and time-intensity profile for d-limonene but differed in relation to their temporal profiles which are presented in Table 2.
Analysis of variance on samples reveals significant differences among parameters including maximum intensity and area under the curve. As shown in Table 2, we found that the maximum intensity of d-limonene varied significantly (p < .05) among samples, sample 3 (Imax = 8.28), sample 2 (Imax = 7.20), and sample 6 (Imax = 6.34) put in order from first to last, respectively. Also, sample 7 had the lowest Imax (1.75). Furthermore, a similar trend was obtained for the area under the curve. However, regarding the plateau parameter, which indicates the duration of d-limonene flavor, there were no significant differences among them. The results thereby showed that across all the evaluated samples, sugar, citric acid, and higher temperature had a significant effect on d-limonene perception intensity.
Investigating the area under the curve, the samples' trend of time intensity was observed ( Table 2) The previous studies have suggested (Frank & Byram, 1988;Schifferstein, 1995) the compatibility of flavor-tastant pairing had a significant effect on predicting influences on perception. By immersing rock candy to the water solutions, an increase in the d-limonene release in gas phase was extensively observed (p < .05). This probably happened as a reaction to the "salting out" effect of sucrose (Voilley, Simatos, & Loncin, 1977) by which sucrose interacts with water, causing an increase in the concentration of flavor compounds in the remaining "free water" (Voilley et al., 1977). The "salting out" effect of glucose or sucrose leads to the release of aroma compounds coming from sugar solutions (Friel et al., 2000;Hansson et al., 2001a;Hansson, Andersson, LeufveÂn, & Pehrson, 2001b;Nahon, Koren, Roozen, & Posthumus, 1998;Voilley et al., 1977). Interactions of sugars and water causes the increase of the concentration of aroma compounds in the vapor phase (Lubbers & Guichard, 2003). The concentration of carbohydrate directly affects the viscosity of the system, and it also affects the retention and release of flavor compounds (Naknean & Meenune, 2010). As shown by the results, at any specifically determined temperature, by adding sucrose in a form of rock candy in the media, a more pronounced release of d-limonene was significantly observed (p < .05). This was probably due to the "salting out" effect of sucrose (Voilley et al., 1977).  (Lindsay, 1985).

As Log p-values
Changes in pH might also affect the flavor compounds themselves.
Increasing the pH will shift the equilibrium so that larger amounts of the citric acid are in the dissociated form, whereas lowering the pH leads to a larger amount of the nondissociated form (Bennett, 1992).
A number of studies have presented various effects of tastants on perceived flavor intensity and mostly express an enhancement of flavor perception by sweetness and sourness (Bonnans & Noble, 1993;McBride & Johnson, 1987;Pfeiffer et al., 2006), but this may depends on the congruency or in other words, the predicted pro-  Frank & Byram, 1988;Schifferstein, 1995).
As Horn (1981) stated on evaluating sweetness and several factors influencing its perception, it was suggested that sweetness of sucrose can somehow be suppressed by acidic ingredients such as citric acid.
Oral temperature also affected sensory attributes, but to a lesser extent. This suggests that the physical/chemical characteristics are dominating in stimulating sensations of flavor and texture properties, and that these characteristics are readily altered by a change in temperature (Engelen et al., 2003).
The various treatments had variant effects on the volatile release amounts that appear in the headspace, and thus, the samples may be perceived diversely during consumption (Patana-anake & Barringer, 2015). Various temperatures were tested to evaluate different conditions for product consumption. Patana-anake and Barringer (2015) suggested that temperature had a proportional effect on volatile  Ventanas, Mustonen, Puolanne, and Tuorila (2010a) assessed the effect of increasing temperature on flavor perception in aromatic foods and found that increasing the temperature led to enhanced perceived flavor intensity. Also, Kähkönen, Tuorila, and Hyyönen (1995) and Ryynänen, Tuorila, and Hyvönen (2001) have stated that the odor intensity of cheese soup was stronger at 63°C than at 33°C, and the odor and flavor intensity of carrot, meat patty, and mashed potato increased as their serving temperatures increased from 25 to 65°C.
Sensitivity to NaCl was significantly higher in solution temperatures of 22°C and 37°C than at 0 or 55°C (Pangborn, Chrisp, & Bertolero, 1970). The perceived sweetness of sucrose solutions of low concentrations was reported to vary in direct proportion with solution temperature (Bartoshuk, Rennert, Rodin, & Stevens, 1982;Calvino, 1986;Green & Frankmann, 1987), where the sweetness was greater at higher temperatures. Hansson et al. (2001a), Hansson et al. (2001b) showed that an increase in the concentration of sucrose (from 20% to 60% w/w) resulted in a significant increase (p < .05) in the release of isopenthyl acetate, ethyl hexanoate, cis-3-hexenyl acetate, linalool, and L-menthone to the gas phase above the soft drink. Results also showed that the same amounts of added citric acid had no effect on flavor release when pH was moderated by sodium hydroxide.
The result of headspace GC-MS is shown in Table 3 which indicate the release of d-limonene among three matrices as well as three various temperatures were significantly different. It is clear that temperature had a significant effect on the release of d-limonene, as proven by time-intensity. d-limonene release from flavored rock candy in water and citric acid (75°C) was much higher that the release from other matrices; moreover, flavored rock candy in water and d-limonene microcapsule in water stand in second and third places. These results are therefore in agreement with sensory analysis; and as Godshall (1997) claimed, different sugars (sucrose, invert sugar, and glucose syrup) interact with water to different degrees, increasing various water activity values, and therefore impacting the release of flavor compounds.

| CON CLUS ION
In this study, flavor release from crystal rock candy was evaluated by

ACK N OWLED G M ENTS
Authors would like to thank Dr. Maryam Kiumarsi for her assistance with sensory methodology and comments that greatly improved the manuscript.

CO N FLI C T O F I NTE R E S T
None declared.

E TH I C A L A PPROVA L
This study was approved by Department of Food Science and Technology, Faculty of Agriculture, Ferdowsi University of Mashhad.