Freezing point temperature is in favor of not‐from‐concentrate apple juice storage

Abstract The short storage period is still a problem hindering the promotion of not‐from‐concentrate (NFC) apple juice, despite the fact it possesses higher nutritional value and more attractive taste compared with its concentrated counterparts. In this study, we compared the effects of temperature range including room temperature (25°C), refrigerator temperature (4°C), freezing point temperature (−1.5°C), and frozen temperature (−18°C), respectively, on the quality of NFC apple juices during a long storage period (150 days). The results suggested that all the juices exhibited good safety during the storage, and the juice stored at −1.5°C possessed higher polyphenol contents, physicochemical properties, less color alteration, and less loss of aroma and taste than 25 and 4°C. Besides, although an exceedingly low temperature (−18°C) could greatly retard the juice deterioration, the loss of aroma and taste was significant. Overall, our results indicated that the NFC juice was most favored by storage at freezing point temperature (−1.5°C), with the highest similarity to the freshly squeezed apple juice.

result, the appropriate storage temperature is a perquisite for maintaining juice qualities. Freezing point temperature, slightly higher than supercooling point, has been used for the storage of animal organs and fresh fish (Wang et al., 2018;Zhu, Ma, Yang, Xiao, & Xiong, 2016). It has been reported that the nectarine stored at freezing point (−1.5°C) exhibited preferably physiological qualities and better antioxidant capacity than at 0 or 5°C (Zhao, Shu, Fan, Cao, & Jiang, 2018). Also, storing green beans at near its freezing point has been proven to retard the deterioration of the qualities of the beans (Guo, Ma, Sun, & Wang, 2008).
By far, it has been well established that the freezing point temperature can exert a more significant effect on the maintenance of the qualities of fruits. However, the reports concerning the temperature effects on NFC apple juice quality are very limited. Thus, this study was used to evaluate the effects of the storage temperature on the quality of NFC apple juices, including room temperature (25°C), low temperature (4°C), freezing point temperature (determined to be around −1.5°C) and frozen temperature (−18°C). Additionally, hierarchical cluster analysis (HCA) was applied to pinpoint the difference between the stored NFC apple juice and the freshly squeezed apple juice based on their polyphenolic contents. Furthermore, principal component analysis (PCA) and linear discriminant analysis (LDA) were used to distinguish the juices so as to emphasize their differences in taste.

| Juice preparation and storage
Fresh "Fuji" apples were purchased from a local market (Xi'an, China).
The apples were washed, peeled, and then the seeds and cores were also removed, after which the apple flesh was cut into slices (6 mm in diameter and 5 mm in thickness). Immediately, the slices were immersed into 0.6% ascorbic acid solution to avoid the enzymatic browning. Afterward, the slices were squeezed by using a juice extractor (HU-780WN, Hurom, Korea) and the juice was collected and then filtered with a 100-mesh filter. After sterilization at 98°C for 30 s and canning, the juice was cooled to room temperature for further storage. Four temperatures including room temperature (25°C), refrigerator temperature (4°C), freezing point temperature, and frozen temperature (−18°C) were selected to store the juice. For each temperature, physicochemical properties and microorganisms of the juice were investigated after a time interval of 30 days. Besides, the polyphenol content, aroma, and taste were also determined after 150 days of storage.

| Measurement of freezing point temperature
The freezing point temperature was measured according to a reported method by Jie, Lite, Lite, and Yang (2003) with some modifications (Jie et al., 2003). When the juice was cooled at a rate of 0.5°C/min, the temperature of the juice was recorded per 2 s using a digital thermometer (testo 106, Testo, Germany) and plotted as a function versus temperature. The corresponding temperature plateau at which the juice did not exhibit discernible temperature change was defined as the freezing point temperature of the juice.

| Detection of microorganisms
Aerobic bacterial counts were determined according to the national standard (GB 4789.2-2016) in China. After plates were incubated at 36°C for 48 hr, the amounts of microorganic colonies were recorded. Besides, yeasts and molds were also detected according to the method defined by the national standard (GB 4789.15-2016) in China. Briefly, 1 ml of the 10-fold diluted apple juice was inoculated into the Rose Bengal Agar (RBA) medium. After RBA plates were incubated at 28°C for 5 days, the amounts of yeast and mold colonies were calculated. The results were expressed as log (CFU/ml).
According to the national standard (GB 4789.3-2016) in China, coliforms were determined as follows: Juice samples were inoculated into the lauryl sulfate tryptose (LST) broth tubes, and subsequently, the matrixes were cultivated at 28°C for 48 hr to check whether there appeared the bubbles in the broth. Generally, the generation of the bubbles can be deemed to be indicative of coliforms.

| Detection of physicochemical properties
The pH value was measured using a digital pH meter (FE20 Plus, Mettler-Toledo, China). The titratable acidity (TA) was determined by using 0.1 M NaOH solution to titrate the 20-fold diluted juice, and phenolphthalein served as the pH indicator. The results were expressed as g/L, with malic acid as the reference. The total soluble solid (TSS) was measured using a refractometer (PAL-1, Atago, Japan), and the result was expressed as °Brix. Besides, the TSS/TA is calculated as the ratio of TSS content to TA content. The turbidity was determined using a turbidimeter (ET76910, Lovibond, Germany).
Before testing, all the juices were diluted 10-fold with distilled water.
Based on the CIE color system, the color values of the juices were measured using a colorimeter (NS800, 3nh, China). The L*, a*, and b* values correspond to lightness, greenness (−a*) or redness (+a*), and blueness (−b*) or yellowness (+b*), respectively. The total color difference (ΔE) was calculated using the following equation:

| UPLC analysis of polyphenolic compositions
The mixture of the apple juice (3 ml) and 80% methanol (6 ml, v/v) was sonicated at 25°C for 15 min, subsequently centrifuged at 7,104 g at 4°C for 10 min. The supernatant was collected and then filtrated through a 0.22-μm Teflon membrane for UPLC analysis. The qualification and quantification analysis of polyphenols was performed according to our previous method (Sun et al., 2017) using a UPLC system equipped with a Thermo ® Syncronis C18 column (250 × 4.6 mm, 5 μm, I.D., USA) and a Thermo ® Ultimate 3000 UPLC system (Thermo Electron Co. USA).

| Electronic nose analysis
The aroma of the NFC juices was analyzed using an electronic nose (Smartnose, ISENSO, China) containing 10 metal-oxide semiconductor chemical sensors. The detailed information is given in Table S1.
Prior to analysis, the gas path of e-nose was purged by clean air for 30 min to normalize the sensor signals. After that, 10 ml of the apple juice was transferred into a 50-ml glass vial equipped with a plastic septum and the juice was equilibrated for 5 min to ensure the volatiles were fully collected. Then, the volatiles lingering in headspace were absorbed by a Luer-lock needle connected to a Teflon tubing (3 mm) at a flow rate of 300 ml/min. The data were collected at a time interval of 0.1 s. The response of sensor was expressed as the ratio of conductance G/G 0 (G 0 and G represent the conductance of sensor before and after being exposed to the gas samples, respectively).

| Electronic tongue analysis
An electronic tongue (Smartongue, ISENSO, China) composed of six metallic disk electrodes (platinum, gold, palladium, titanium, tungsten, and silver) was applied for the taste analysis of the NFC juices.
Prior to detection, the electronic tongue (e-tongue) was warmed up for 30 min using 0.01 M KCl solution and the metallic sensors were subsequently washed with distilled water for 2-3 min, so as to minimize the nose interference. When analyzed, the juices were filtrated with a filtrating membrane to avoid the artifact signals caused by the solid particles and the temperature was 25°C.

| Statistical analysis
SPSS 18.0 software (SPSS Inc., USA) was used to analyze the obtained data, and the results were expressed as the means ± standard deviations. The significant difference was calculated using Duncan's test at a level of p < 0.05. Besides, PCA, LDA, and HCA were also performed to pinpoint the difference between the NFC apple juices.

| Freezing point temperature of NFC apple juice
The temperature curve of the NFC juice is displayed in Figure S1. As expected, the temperature of the juice was observed to decrease to its supercooling point temperature (−1.8°C), and then, the temperature was slightly increased and formed a distinct plateau. Thus, the temperature of the juice at the plateau region (−1.5°C) was regarded as the freezing point .

| Detection of microorganisms in NFC juices
The microorganism amounts of the juices during storage were regularly detected to reflect the effects of temperature on the juice safety. No aerobic bacterial counts, and yeast and mold counts were observed in the initial NFC juice, as reflected in Figure 1a,b. Regarding the juice stored at 25°C, both the aerobic plate counts (APC) and the yeast and mold counts (YMC) showed higher growing rates than at other storage temperatures. In de-

| Physicochemical properties of NFC juices
During the storages at different temperatures, the pH values were observed to be decreased, and the TA contents were increased, which may be due to the acidic metabolites produced by yeasts and lactobacillus. As discussed above, low temperature can inhibit the growth of microorganisms, and thus, the juice stored at −18°C showed the lowest fluctuations of pH and TA. Moreover, it can be seen that the TSS of juices was significantly increased during the early storage period and then followed by a decreasing trend (Table 1A). Notably, the juice stored at −1.5 and −18°C displayed a delayed increase in TSS compared with the higher temperature (i.e., 4 and 25°C), indicating that the maintenance of TSS can be favored by a lower temperature. Additionally, the ratio of TSS to TA is usually calculated to assess the overall taste of juice because the TSS/TA ratio has been reported to have a stronger correlation to the taste of juice than TSS and/or TA alone (Crisosto & Crisosto, 2005 (Tian et al., 2018). In UFC juices, the flesh particles were mainly composed of proteins, cellulose, hemicellulose, and pectin. It can be found that the turbidity was continuously deceased during the storage. And the decrease in temperature could facilitate the retention of turbidity, as indicated by the higher turbidity in the juice stored at −1.5°C. That can be explained by the fact that the viscosity of the juice was higher at a lower temperature, which may reduce the precipitation rate of the suspended flesh particles. As expected, the juice stored at −18°C exhibited the greatest turbidity, which may be because the frozen juice immobilized the flesh particles. Thus, the precipitation of the particles was largely inhibited.

| Color quality of NFC juices
Compared with the concentrated apple juice, NFC juice usually exhibits a color more close to the flesh of fresh apples, owing to the omitted concentration process during which the browning of the juice inevitably occurs. However, during a long term of storage, the color of NFC juice may be deepened due to the enzymatic browning.
For example, it has been reported that both the oxidation of polyphenolic compounds and Maillard reaction can lead to the deepened color of the juice (Lee, Seo, Rhee, & Kim, 2016). In this study, L*, a*, b*, and ΔE values were applied to fully characterize the juice color, and the results are summarized in Table 1B. As listed in the table, the L*, a*, and b*values of all the juices were found to be continuously decreased during the whole storage period, suggesting that the juices had a deeper color as the storage proceeded. In addition, it can be found that the total color difference (∆E), which is generally used to characterize the comprehensive color difference between the test group and the control group, was continuously increased. The result suggested that the color of the stored juices exhibited a widened difference as the storage was prolonged. Moreover, the juice color could be better maintained as the storage temperatures decreased, which is in agreement with a previous study reporting that the decrease in storage temperature was associated with a lower ΔE value (Buvé et al., 2018). Generally, when ΔE value is ranged from 3.0 to 6.0, the color difference can be discerned by unaided eye (Wibowo et al., 2015). In our study, the ΔE values of all the juices were over 3.0 after 150 days of storage, indicating that the changes in color would be visible.

| Qualification and quantification of polyphenols in NFC juices
UPLC was used to identify and quantify the individual phenolics in the NFC juices. Based on the peaks, fourteen polyphenolic compounds including thirteen known polyphenols and one unknown compound were identified, and the result is given in Table 2.  Table 2. It can be observed that the major phenolic compounds were 3-caffeoylquinic acid, procyanidin B2, and epicatechin, which is consistent with a previous study (Heinmaa et al., 2017). Among the 13 identified individual polyphenols, epigallocatechin, catechin, procyanidin B2, epicatechin, and epicatechin gallate belong to procyanidins, and gallic acid, protocatechuic acid, chlorogenic acid, and 4-hydroxybenzoic acid belong to the family of phenolic acids (Tian et al., 2018). It has been reported that the most abundant polyphenols in apples are procyanidins, followed by phenolic acids (Renard et al., 2011). In the NFC juice, however, chlorogenic acid (3-caffeoylquinic acid) was found to account for the highest content of polyphenols, as listed in Table 2. It may be because the removal of peels during the NFC preparation greatly reduced the procyanidins and phloridzin contents, which mostly exist in apple peels (Chen, Zhang, Wang, Li, & Ma, 2012;Heinmaa et al., 2017).
By contrast, chlorogenic acid is most abundant in apple flesh (Zhao et al., 2015). Moreover, it was found that the contents of individual polyphenols were roughly decreased during the storage period, suggesting that the polyphenols in the juices may be slightly degraded. Also, the increased storage temperature caused more pronounced polyphenol degradation, which is in line with the reports by Kim et al. (2018) and Teleszko et al. (2016). Due to significant antioxidant capacity, plant polyphenols are often involved into the free radical scavenging reaction to reduce the oxidative stress of organisms, and thus, the polyphenol content is commonly regarded as one crucial parameter for assessing the nutritional value of juices. At low temperatures (i.e., −1.5 and −18°C), the polyphenol degradation could be retarded and therefore the polyphenol contents were well maintained after 150 days of storage. However, the difference in molecular architecture allows the individual polyphenols to exhibit different stability. As an example, when the juice was stored at 25°C, the contents of epigallocatechin, procyanidin B2, epicatechin, and phloridzin exhibited a larger extent of decrease, as reflected in Table 2, demonstrating the instability of procyanidin (Dai et al., 2018). On the contrary, the contents of 3-caffeoylquinic acid and protocatechuic acid were relatively stable during the whole storage, which is in accordance with a previous study (Rekas, Scibisz, Siger, & Wroniak, 2017).

| E-nose analysis of NFC juices
The e-nose response signals of the juices including the control (the freshly squeezed apple juice) and the NFC juices stored at different temperatures are given in Figure 3. It was found that the response signals of sensor 2 (S2, the following abbreviations are the same), S5, and S3 in all the juices were significantly increased and then remained at a relatively high levels, indicating that the main aroma constituents in NFC apple juice may include ketones, aldehydes, and alkane organic and NO X compounds. Besides, the TA B L E 2 Retention time (Rt), wavelengths of maximum absorption (λmax), tentative identification, and quantification of the phenolic compounds of juices stored at different temperatures response signals of the freshly squeezed apple juice were significantly higher than the juices stored for 150 days, suggesting that the aroma constituents in apple juice were decreased during the storage. Notably, the juice stored at −1.5°C showed the highest response levels compared with 25, 4, and −18°C, indicating a maximum retention of aroma. Additionally, the radar chart was yielded based on the average response signals of the 10 sensors during 100-105 s. As shown in Figure 3f, the different juice samples showed a similar shape but significantly different response strengths, which further confirmed the decrease in aroma constituents in NFC juice.

| E-tongue analysis of NFC juices
The tastes of the juices were detected by e-tongue, and the results were further analyzed by means of multivariate statistical analysis methods including PCA and LDA to pinpoint the possible alterations of the juices in taste. PCA is a classical unsupervised algorithm of pattern recognition to analyze, classify, and reduce the dimensionality of numerical data sets in a multivariate problem (Costache, Corcoran, & Puslecki, 2009). It can decompose the experimental data matrix into latent variables and explain the variables in data by two or three principal components, so as to elucidate the relationships between the original variables and their influence on the matrix (Tian, Deng, & Chen, 2007). LDA is a classical machine learning technique, which mainly uses the statistical information of image pixels while ignoring their scattering characteristics (Mahdianpari et al., 2018). As shown in Figure 4, all the juices were well separated into individual regions. Figure 4a suggests that the score plot of PCA explained 86% of the total variance, which consisted of PC1 (62.3%) and PC2 (23.7%). As a consequence, PC1 and PC2 can be considered adequate to represent the original data set. In general, the juice samples that can be grouped into a cluster were characterized by a high similarity (Tian et al., 2018). As shown in Figure 4a, the juices stored at different temperatures had significant difference in taste, whereas the juice stored at −1.5°C was found to have a more similar taste to the freshly squeezed apple juice.
To verify the result of PCA, LDA was performed to analyze the same data matrix. Figure 4b suggests that LD1 and LD2 could explain 72.5% and 16.9% of the total variance, respectively, and the  Guo et al. (2008), who reported that compared with room temperature and low F I G U R E 3 E-nose response signals of NFC apple juices. (a) the freshly squeezed apple juice (control); (b) the juice stored at 25°C; (c) the juice stored at 4°C; (d) the juice stored at −1.5°C; and (e) the juice stored at −18°C, and the radar chart (f) temperature (4°C), controlled freezing point storage showed preferably higher taste qualities. However, exceedingly low temperature (i.e., −18°C) may also cause significant changes in the taste of the juices, as shown in Figure 4b. It may be due to the fact that the process of freezing and thawing would affect the texture of the pulp cells.

| CON CLUS ION
In this work, the NFC juices were prepared and then stored at a wide temperature range, as represented by 25, 4, −1.5, and −18°C, respectively. Through the detection of microorganisms, it was found that the low storage temperatures were positively related to the decreased growth rate of the yeast and mold. After 150 days of storage, both the aerobic plate counts and the yeast and mold counts were lower than the standard defined by Institute of Food Science and Technology (IFST), demonstrating good safety during the storage. Besides, compared with 4 and 25°C, the freezing point temperature (−1.5°C) could retard the decrease of the physicochemical properties and polyphenol contents. Also, the low temperature was favorable for maintaining the original color of the juice. However, an exceedingly low temperature such as −18°C could cause the significant loss of aroma and taste in NFC apple juice. The further PCA and LDA suggested that after 150 days of storage, the juice stored at −1.5°C exhibited the highest similarity to the freshly squeezed apple juice, which may indicate that −1.5°C was more favorable for storing NFC apple juice.

ACK N OWLED G M ENT
This research is sponsored by the China Agriculture Research System (CARS-27) from the Ministry of Agriculture of the People's Republic of China.

CO N FLI C T O F I NTE R E S T S
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

E TH I C A L S TATEM ENT
This study does not involve any human or animal testing.