Effect of juice concentration on storage stability, betacyanin degradation kinetics, and sensory acceptance of red-fleshed dragon fruit (Hylocereus polyrhizus) juice

ABSTRACT

The aim of this study was to evaluate the effect of red-fleshed dragon fruit (Hylocereus polyrhizus) juice concentration on betacyanin degradation kinetics and sensory acceptance. The samples stored at 4°C were comparatively stable compared to 25°C in terms of betacyanin retention, and yeast and mold growth. The loss of betacyanin in red-fleshed dragon fruit juice and concentrate at 25, 37, and 45°C were of first-order reaction. The drink reconstituted from red-fleshed dragon fruit concentrate showed better acceptability in terms of sweetness, flavor, and overall acceptability. The current study demonstrated improved betacyanin retention and sensory acceptance of red-fleshed dragon fruit juice via juice concentration.

KEYWORDS:

• Pitaya
• Concentrate
• Kinetics
• Sensory
• Storage, Betacyanin

Introduction

The red-fleshed dragon fruit (Hylocereus polyrhizus), also known as pitaya, pitahaya, and strawberry pear,^[1] contains phenolic compounds and betacyanin that contribute to antioxidant activity.^[2,3] Examples of other nutrients and minerals found in red-fleshed dragon fruit are protein, carbohydrate, fat, vitamin B1 (thiamin), vitamin B2 (riboflavin), vitamin B3 (niacin), vitamin B6 (pyridoxine), vitamin B12 (cobalamin), vitamin C, carotene, flavonoid, phosphorus, iron, and phytoalbumin.^[2] According to Jaafar et al.,^[2] dragon fruit aids digestion, reduces cholesterol levels and high blood pressure, prevents diabetes and colon cancer, as well as combats against cough and asthma.

Betacyanin contributes to the red-purple color of the red-fleshed dragon fruit.^[4] The stability of betacyanin is affected by heat, oxygen, light, pH, as well as moisture.^[5] Ascorbic acid has been reported to stabilize betacyanin in both juice and concentrate at agitated or non-agitated conditions. In contrast, light degraded betacyanin in both juice and concentrate.^[6]

Concentration reduces the moisture content to a level whereby the water activity is low enough to support degradation mechanisms, thus serving as an effective preservation method for juices. It is commonly practiced to increase the relative solids concentration. The concentrated product reduces transportation and storage costs as a result of the reduced storage volume that is involved. The objective of the current study is to determine the effect of juice concentration on the storage stability, shelf life, and sensory acceptance of red-fleshed dragon fruit (Hylocereus polyrhizus) juice. The current work is of great industrial significance to reduce the cost of transportation and increase the shelf life of red-fleshed dragon fruit juice.

Materials and methods

Chemicals

The following were used in the current study: red-fleshed dragon fruit (Great Sun Pitaya Sdn. Bhd., Telok Panglima Garang, Kuala Langat, Selangor), hydrochloric acid 98%, HCl (Fisher Chemicals, USA), ascorbic acid 99.7% C₆H₈O₆ (Merck, Darmstadt, Germany), di-sodium hydrogen phosphate/sodium phosphate dibasic anhydrous (R&M Chemicals, Essex, UK), gallic acid C₇H₆O₅ (R&M Chemicals, Essex, UK), ethanol C₂H₅OH (HmbG® Chemicals, Hamburg, Germany), Folin–Ciocalteu’s phenol reagent (R&M Chemicals, Essex, UK), sodium carbonate Na₂CO₃ (Merck, Darmstadt, Germany), bacteriological saline 85% (Oxoid, Hampshire, England), plate count agar (PCA; BD, Sparks, USA), and Sabouraud’s dextrose agar (SDA; Merck, Darmstadt, Germany).

Preparation of juice and juice concentrate

Red-fleshed dragon fruits were obtained from Great Sun Pitaya Sdn. Bhd (the orchard was located at Telok Panglima Garang, Kuala Langat, Selangor). Fresh fruits were stored at 4°C for 1 week before use. The fruits were then halved and the peels were manually removed. The flesh was juiced using a juicer (Panasonic Juicer model MJ-70M, Malaysia) to remove coarse cloud particles and seeds as carried out by Nur ‘Aliaa et al.^[7] All the juices was ensured to be within 13.5 ± 1°brix.

Filtration

Dragon fruit pulp was treated according to Nur ‘Aliaa et al.^[7] The treated pulp was centrifuged (Hettich Zentrifugen Universal 320R, Germany) at 3000 × g for 10 min, and the supernatant was collected. The juice was then filtered through a filter paper (CHMLAB GROUP, Barcelona, Spain) using a Buchner funnel attached to an oil-less vacuum pump (Rocker 300, Rocker Scientific Company Limited, Taiwan).

Juice concentration

The method used for juice concentration was according to Schweiggert et al.^[8] and Woo et al.,^[9] with slight modification. Filtered juice was concentrated in a rotary evaporator (EYELA, model N-1100S-WD, Fisher Scientific, USA) with a water bath (EYELA OSB-2100, Fisher Scientific, USA) attached to an aspirator (EYELA, model A-1000S, Fisher Scientific, USA) to a total soluble solids (TSSs) of 60 ± 5 at 40°C followed by frozen storage at –80°C until further analysis.

Storage stability

Juice or concentrate samples were added with 0.25% (w/w) ascorbic acid. Samples were subsequently adjusted to pH 4.0 with 1M HCl. These samples were then pasteurized at 65°C for 30 min. Juice and concentrates were stored at 4 (refrigerator) and 25°C (room temperature), respectively. Samples were analyzed weekly for betacyanin and ascorbic acid content (AAC) throughout 8 weeks of storage as adapted from Abbo et al.^[10] Bacteria as well as yeast and mold growth were also monitored weekly.

pH and titratable acidity (TA)

pH was measured using a portable pH meter (SG2, Mettler Toledo, Schwerzenbach, Switzerland). TA was determined using the official method 942.15 of AOAC.^[11] The juice (5 mL) was added to 20 mL Milli-Q water and the sample was titrated slowly with 0.1 N NaOH until the pH was brought to about 8.1. The samples were stirred continuously with a magnetic stirrer on a hot-plate stirrer. The pH of the sample was checked regularly. Volume of NaOH required to bring the pH to 8.1 was recorded. The TA was expressed in g/L of citric acid (Eq. 1).

(1)

*= equivalent weight of citric acid

Determination of betacyanin content (Bc)

For spectrophotometric analysis of samples, dragon fruit juice was diluted with McIlvaine buffer (pH 6.5) until a maximum absorption of 1.00 ± 0.05 was reached.^[5] McIlvaine buffer was prepared from 0.1 M citric acid (29.65 mL) and 0.2 M sodium phosphate dibasic (70.35 mL).^[12] Bc was expressed as betanin equivalents and was calculated as described by previous studies:^[5,13]

(2)

where A is absorption value at λ_max (537 nm) corrected by the absorption at 600 nm; F is dilution factor; MW is molecular weight of betanin (550 g/mol); ε is molar extinction coefficient of betanin (60,000 L/mol cm); and l is path length of the cuvette (1 cm).

Betacyanin retention (photometrical) was calculated as modified by Schweiggert et al.:^[8]

(3)

where Bc₀ represents the initial Bc and Bc₁ represents the final Bc. Since F, MW, ε, and l are all constants, only the corrected absorbance (A) was used to calculate the proportion of betacyanin retained.

Determination of AAC

The AAC was determined using the iodine titration method modified from Suntornsuk et al.^[14] One milliliter of juice was transferred into a 125 mL conical flask and diluted with 4 mL distilled water. Five milliliters of 2 N sulfuric acid was added, mixed and 2 mL of 1% starch was added as an indicator. The solution was directly titrated with 0.01 N iodine. A blank titration was performed prior to titration of each sample. Each milliliter of 0.01 N iodine is equivalent to 0.8806 mg of ascorbic acid.

Bacteria, yeast, and mold counts

Microbiological analysis was modified from Abbo et al.^[10] For microbial counts, serial dilutions were performed with sterile 0.85% bacteriological saline. Samples were spread plated onto microbiological media. Yeast and mold counts were determined using SDA, whereas aerobic mesophilic bacterial counts were determined using PCA. Plates were incubated inverted for yeast and mold at 30°C for 3 days while for aerobic mesophilic bacteria it was incubated at 37°C for 24 h.

Reaction order of betacyanin degradation

Juice and concentrate samples were stored in universal bottles in incubator ovens of 25°C,^[15] Thirty-seven degrees Celsius (PERCIVAL model AR-66L2, Percival Scientific, Perry, Iowa) and 45°C (UFE400, Memmert, Schwabach, Germany). The proportion of betacyanin retained was analyzed at several intervals. Juice stored at 25, 37, and 45°C were measured over a period of 33, 26, and 19 days, respectively, whereas concentrate stored at 25, 37, and 45°C were measured over a period of 68, 68, and 36 days, respectively. Juice and concentrate samples stored at 25°C were measured over a longer span of time due to the expected slower rate of reaction compared to storage at higher temperatures.^[15]

To determine the reaction order of betacyanin degradation, zero- and first-order models were evaluated to see which model was better fitted with the betacyanin degradation data. For the zero-order model, linear trend lines were plotted, whereas for first-order model, exponential trend lines were plotted. The model which gave a higher coefficient of determination (R²) value would be selected.^[16]

The degradation of betacyanin at each temperature applied were subjected to separate regression analysis using first-order degradation model Liu et al.^[17] and Tonon et al.:^[15]

(4)

where is the proportion of betacyanin retained after the reaction time t. To obtain the rate constant (k) at each temperature, a plot of -ln was plotted against time (days). This would give a straight line passing through the origin with slope k (rate constant). The temperature dependence for k, the reaction rate constant was assumed to adhere to the Arrhenius relationship.^[18] The Arrhenius equation is shown in Eq. (5).

(5)

k is the rate constant; k₀ is the pre-exponential factor; Ea is the activation energy in kcal mol⁻¹; R is the gas constant (1.987 kcal mol⁻¹ K⁻¹), and T is the absolute temperature in K.^[19] A linear equation can be formed by taking the logarithms of both sides and rearranging the equation to become:

(6)

Hence, a plot of ln(k) versus would give a straight line with k₀ as the y-intercept and as the slope. The best fit straight line was drawn and extrapolated to get the rate constant value at 4°C. By setting the failure point (= 0.5) and using the rate constant value at 4°C, Eq. (4) was applied to calculate the time for failure if the product were to be stored at 4°C.

Consumer preference test

Juice and concentrate samples were added with 1% (w/w) of 25% (w/w) citric acid to bring the pH down to approximately pH 4.0. Consumer preference test was modified from Walkling-Ribeiro et al.^[20] and Palgan et al.^[21] This test included 126 untrained panelists ranging from 18 to 26 years of age. Prior to sensory evaluation the red-fleshed dragon fruit concentrate was reconstituted to a drink of ~14°brix, similar to that of the juice with no reconstitution (juice without prior concentration). Both the reconstituted drink and juice with no reconstitution (juice without prior concentration) were stored in a refrigerator at 4°C before serving to panelists. Both samples were given random 3-digit codes and served (~15 mL) together with a glass of still water to panelist. A 9-point hedonic scale was used, where 1 was the lowest (least liked) and 9 was the highest score (most liked). Flavor attributes were also surveyed by asking panelists to evaluate the samples regarding the occurrence of cooked, bland, artificial, earthy, or metallic flavor, if present.

Statistical analysis

Measurements were performed in triplicates (n = 3) and were reported as mean ± standard deviation. The results were tested for normality and homogeneity of variance using the Statistics Package for the Social Sciences (SPSS) version 16.0. When the data is of normal distribution and has equal variance, analysis of variance (ANOVA) followed by post-hoc Tukey’s test were performed to determine significant differences (p < 0.05). For non-normal data and data with unequal variance, the data was transformed and retested for normality and homogeneity of variance. If the transformed data was normal with equal variance, ANOVA and Tukey’s test were performed. If data still remained non-normal with unequal variance, Kruskal Wallis test was performed using MedCalc software to determine significant differences (p < 0.05). For sensory evaluation data, similar statistical methods were carried out. A radar chart was drawn to show the mean scores (n = 126) for sensory attributes of dragon fruit juice and concentrate.

Results and discussion

Physicochemical properties and Bc

The juice sample showed °brix of 14.2 ± 0.0, pH of 3.98 ± 0.01, and TA of 4.35 ± 0.13 g/L. The concentrate sample has a higher °brix of 54.0 ± 0.5, pH of 4.05 ± 0.04, and TA of 3.16 ± 0.07 g/L. The pH of the juice and concentrate samples were within the range of orange juice,^[22,23] persimmon juice,^[24] and Spanish pomegranate juices,^[25] and the TA is within the standards for apple and passion fruit juice.^[26] Overall, betacyanin was most stable in the concentrate sample stored at 4°C where no significant (p ≥ 0.05) decrease in betacyanin retention after 8 weeks of storage (Fig. 1). This was followed closely by the juice sample stored at 4°C, which had betacyanin retention of 0.93 ± 0.02 after 8 weeks of storage (Fig. 1). Both samples stored at 4°C had higher betacyanin retentions in comparison to samples stored at 25°C (Fig. 1). After 8 weeks of storage, the proportion of betacyanin left in the concentrate sample stored at 25°C was 0.67 ± 0.01, while the juice sample stored at 25°C showed the lowest betacyanin retention of 0.35 ± 0.01 (Fig. 1). These findings suggest that storage at 4°C showed better betacyanin retention compared to storage at 25°C, and betacyanin was more stable in concentrate samples compared to juice samples (concentrate 4°C > juice 4°C > concentrate 25°C > juice 25°C). In the current study, concentrate that has a lower water activity (0.832 ± 0.007) compared to juice (0.983 ± 0.001) improves the stability of betacyanin.^[27,28] In the current study, juice stored at 4°C increased in betacyanin retention from 1.00 to 1.07 ± 0.02 in week 1 (Fig. 1). This could possibly indicates betacyanin regeneration as previously described by several authors who found a higher color retention in betanin solutions^[29] and red beet^[30] after 1 day of cool storage in comparison to immediately after heating.

Figure 1. Betacyanin retention for juice and concentrate samples stored at 4 and 25°C throughout 8 weeks (mean ± standard deviation; n = 3). Small letters mean significant difference (p < 0.05) across the samples with storage time (weeks).

AAC

It was observed that ascorbic acids content of concentrate samples stored at 4°C throughout 8 weeks of storage remained at approximately 250 mg/100 mL (Fig. 2). At a storage temperature of 25°C, the concentrate samples had a slight decrease in AAC after 8 weeks of storage (Fig. 2). Ascorbic acid in juice stored at 4°C decreased after 8 weeks, while a more drastic drop was observed for juice stored at 25°C (Fig. 2). These current results indicate that juice samples generally exhibited a more drastic reduction in AAC compared to concentrate samples, and ascorbic acid was better retained when stored at 4°C compared to storage at 25°C (concentrate 4°C > concentrate 25°C > juice 4°C > juice 25°C). The concentrate samples with lower water activity favored ascorbic acid retention as compared to the juice sample. This is in consistent with Lee and Chen,^[18] ascorbic acid degradation was slower at 4°C compared to 24°C.

Figure 2. Ascorbic acid, mg/100 mL for juice and concentrate samples stored at 4 and 25°C throughout 8 weeks (mean ± standard deviation; n = 3). Small letters mean significant difference (p < 0.05) across the samples with storage time (weeks).

Microbiological analysis

The juice samples showed higher initial bacteria, yeast and mold count compared to the concentrate samples (Table 1). This could be because of the concentrate samples that had undergone concentration in a rotary evaporator at 40°C, thus inactivating some heat-sensitive microorganisms. In addition, once the water activity was reduced from 0.983 ± 0.001 (juice) to 0.832 ± 0.007 (concentrate), it serves as a mild preservative, which inhibits growth of most microorganisms. For both bacteria, as well as yeast and mold counts, pasteurization at 65°C for 30 min managed to reduce microbial counts to non-detectable levels immediately after treatment (Table 1). Rivas et al.^[31] reported that thermal treatment at 98°C for 21 s managed to reduce total microbial contamination in blended carrot and orange juice to <1 cfu/mL. Bull et al.^[32] found that Valencia orange juice (pasteurized at 65°C for 60 s) and navel orange juice (pasteurized at 85°C for 25 s) were microbiologically stable when stored at 4°C for 4 and 12 weeks, respectively. The juice sample stored at 25°C showed yeast and mold growth from week 7 onward (430 ± 26 cfu/mL), while yeast and mold counts in the other three samples were not detectable (Table 1). By week 8, the juice stored at 25°C had a yeast and mold count of 2433 ± 416 cfu/mL (Table 1). Under the Good Manufacturing Practices, the total yeast count should be less than 10, while a maximum acceptable yeast count at any point during the shelf life of the food product was set at 10⁶ for pasteurized fruit juices.^[33] Therefore, the juice stored at 25°C after 7 weeks was considered undesirable.

Table 1. Bacteria, yeast, and mold counts, expressed as cfu/mL for juice and concentrate samples before and after pasteurization and during storage at 4 and 25°C throughout 8 weeks (mean ± standard deviation; n = 3).

Week			Before pasteurization (cfu/mL)	After pasteurization (cfu/mL)	1	2	3	4	5	6	7	8
Juice	4°C	Bacteria	4900 ± 300	ND	ND	ND	ND	ND	ND	ND	ND	ND
		Yeast & mood	5567 ± 1274	ND	ND	ND	ND	ND	ND	ND	ND	ND
	25°C	Bacteria	4767 ± 306	ND	ND	ND	ND	ND	ND	ND	ND	ND
		Yeast & mood	5733 ± 551	ND	ND	ND	ND	ND	ND	ND	430 ± 26	2433 ± 416
	4°C	Bacteria	590 ± 80	ND	ND	ND	ND	ND	ND	ND	ND	ND
		Yeast & mood	947 ± 51	ND	ND	ND	ND	ND	ND	ND	ND	ND
	25°C	Bacteria	623 ± 15	ND	ND	ND	ND	ND	ND	ND	ND	ND
		Yeast & mood	1037 ± 74	ND	ND	ND	ND	ND	ND	ND	ND	ND

ND: Non detectable levels.

Reaction order of betacyanin degradation

A higher determination coefficient R² was found when an exponential trend line was plotted compared to when a linear trend line was plotted (except for concentrate samples stored at 25°C; Table 2). Thus, the loss of betacyanin in red-fleshed dragon fruit juice and concentrate at the three temperatures can be described as a first-order reaction. The current study is in accordance with previous reports by Saguy^[34] and Herbach et al.^[35] in which thermal degradation of betalain pigments in the pH range of 3.0 to 7.0 adheres to first-order degradation kinetics.

Table 2. Prediction functions and coefficient of determination (R²) values for zero- and first-order models fitted to the thermal degradation of betacyanin data.

		Determination coefficients of first-order betacyanin degredation		Determination coefficients of zero-order betacyanin degredation
		Prediction functions	R^2	Prediction functions	R^2
	25°C	y = exp(-0.02x)	0.9968	y = -0.0155x + 1	0.9878
Juice	37°C	y = exp(-0.115x)	0.9906	y = -0.0476x + 1	0.8097
	45°C	y = exp(-0.199x)	0.9905	y = -0.0665x + 1	0.6967
	25°C	y = exp(-0.01x)	0.9805	y = -0.0078x + 1	0.9914
Concentrate	37°C	y = exp(-0.04x)	0.9966	y = -0.0183x + 1	0.6922
	45°C	y = exp(-0.079x)	0.9984	y = -0.0289x + 1	0.4168

Using Eq. (4) (), graphs of – were plotted against time (t) for each of the juice and concentrate samples stored at 25, 37, and 45°C, respectively. The slope of the linear regression line was the rate constant, k (juice 25°C, y = 0.0198x; juice 37°C, y = 0.1147x; juice 45°C, y = 0.1985x; Concentrate 25°C, y = 0.0102x; Concentrate 37°C, y = 0.04x; concentrate 45°C, y = 0.0828x). The higher rate constant values for juice samples showed that betacyanin in juice samples degrade faster than in concentrate samples.

So far, rate constants at several high temperatures are known, as well as a general relationship between rate constant and temperature as depicted by the Arrhenius equation where k = k₀e^(-Ea/RT). Zheng and Lu^[36] reported that the Arrhenius equation is most suitable when determining the kinetic rate constants of food quality attributes affected by temperature. When Arrhenius plots were constructed by plotting ln(k) against 1/T (K⁻¹) for the juice and concentrate samples, these plots gave a straight line with a y-intercept of k₀ and a slope of –Ea/R. Estimated k value at 4°C was calculated from the Arrhenius plots and reported in Table 3. It was observed that the red-fleshed dragon fruit concentrate required a longer time compared to that with the red-fleshed dragon fruit juice to reach half of the betacyanin retention (Table 3).

Table 3. Prediction functions, R², rate constant at 4°C and the calculated time to reach failure point if both juice and concentrate samples were to be stored in 4°C refrigerator.

	Sample
	Juice	Concentrate
Prediction functions	y = -11164x + 33.627	y = -9977.6x + 28.916
R^2	0.9742	0.9983
Κ at 4°C*	0.0012	0.0008
Approximate time to reach failure point; betacyanin retention = 0.5 (years)*	1.51	2.31

Consumer preference test

Sensory evaluation was carried out to determine consumer acceptance of drink reconstituted from the red-fleshed dragon fruit concentrate through hedonic evaluations at 95% confidence level. The mean scores (n = 126) for overall preference, aroma, sweetness, flavor, and color are plotted in Fig. 3. A total of 126 consumers evaluated both products (juice vs. drink reconstituted from concentrate) and the results showed that there was no significant difference (p ≥ 0.05) for aroma and color; whereas there was a significant difference (p < 0.05) for sweetness, flavor, and overall preference of the two products.

Figure 3. Radar chart of mean scores for sensory attributes of dragon fruit juice and concentrate. Attributes with a significant difference (p < 0.05) are marked with an asterisk (*).

For aroma, some consumers liked the fresh, grass-like smell of the juice, while some preferred the more neutral, honey-like smell of the drink reconstituted from concentrate. Results showed no significant difference in aroma likeability (p ≥ 0.05; Fig. 3), thus indicating that panelist accepted the aroma of both samples in general. For color, the juice sample had a more reddish tone compared to drink reconstituted from concentrate sample that had a red-violet colour. Since no significant difference in color was found (p ≥ 0.05; Fig. 3), this showed that panelist generally accepted the color of both samples. Moreover, results showed that consumers preferred the sweetness and flavor of the drink reconstituted from concentrate compared to the juice sample (p < 0.05; Fig. 3). The drink reconstituted from concentrate had a significantly higher overall acceptability score compared to the juice sample (p < 0.05; Fig. 3). Therefore, the drink reconstituted from the red-fleshed dragon fruit concentrate was more well-liked compared to the red-fleshed dragon fruit juice (Fig. 3).

Conclusion

Yeast and mold growth was only observed in juice stored at 25°C in week 7. Samples stored at 4°C had a better betacyanin retention than samples stored at 25°C (concentrate 4°C > juice 4°C > concentrate 25°C > juice 25°C). As for ascorbic acid, the concentrate samples had a better retention than juice samples (concentrate 4°C > concentrate 25°C > juice 4°C > juice 25°C). The loss of betacyanin in red-fleshed dragon fruit juice and concentrate at 25, 37, and 45°C is of first-order reaction. Red-fleshed dragon fruit concentrate required a longer time to reach half of the betacyanin retention compared to that with the red-fleshed dragon fruit juice. Overall, it was found that the drink reconstituted from red-fleshed dragon fruit concentrate was more well-liked compared to the juice without prior concentration.

Funding

The authors would like to thank the School of Science, Monash University Malaysia, for providing research funding for this project.

Additional information

Funding

The authors would like to thank the School of Science, Monash University Malaysia, for providing research funding for this project.