1. Introduction
Almost 4 million deaths were associated with the inadequate consumption of fruits and vegetables in 2017. In fact, last year, the World Health Organization (WHO) recommended a healthy diet low in fat, sugar and sodium, and rich in fruits and vegetables. Specifically, WHO suggests that an intake of more than 400 g per day of fruits and vegetables improves overall health and may reduce the risk of non-communicable diseases [
1].
Recently, there has been a worldwide increase in interest in pomegranate juice due to numerous health benefits of its consumption [
2]. Because of its high antioxidant capacity, pomegranate and its components such as juice, seeds and peel, have favorable health effects, including antibacterial and anticarcinogenic properties [
2].
Bioactive compounds of pomegranate include phenolic components such as hydrolysable tannins (ellagitannins and gallotannins), phenolic acids (gallic acid) and flavonoids (anthocyanins) [
3]. The major anthocyanins in pomegranate juice are as follows: delphinidin-3,5-diglucoside; cyanidin-3,5-diglucoside; cyanidin-3-O-glucoside and delphinidin-3-O-glucoside, pelargonidin-3-0-glucoside and the 3,5-diglucoside [
4]. Its high polyphenol content implies remarkable anti-inflammatory, antioxidant, antitumor, antimicrobial, anti-obesity, antidiabetic, diuretic and depurative qualities, among other, so it can be used as a preventive measure against cardiovascular diseases, cancer, and neurodegenerative diseases. Its ability to slow down ageing has also been demonstrated [
3,
5,
6,
7].
Pomegranate juice has a low caloric value due to its light fat and protein levels. The sugar is represented by glucose and fructose, with the content of the last one being normally higher than that of glucose. It also contains potassium, as well as phosphorus, magnesium, calcium and iron [
4].
The natural appearance of pomegranate juice is turbid, which makes it difficult to preserve. Therefore, it must undergo a clarification process before being placed on the market. In this process, suspended solids are removed to produce a clear juice with a better taste and to avoid the appearance of turbidity after bottling. One method to carry out this process is through membrane ultrafiltration (UF) [
8,
9,
10,
11].
Different techniques have been used to conditionate and clarify pomegranate juice, such as pasteurization, thermal concentration, and the use of fining agents [
12]. However, these techniques modify the bioactive compounds and the antioxidant activity of the juice.
An increased consumer demand for high-quality pomegranate juice, as well as a growing industrial interest in the production of different products (functional food, nutraceuticals, etc.), have promoted interest in minimal-processing technologies [
13,
14]. In this context, membrane processes represent an innovative approach to improve the quality of pomegranate juice.
Microfiltration and UF membrane processes have proven to be comparable to pasteurization in guaranteeing the microbiological stability of juice, and avoiding the deterioration of the final product. In addition, UF and nanofiltration membranes offer new perspectives in juice fractionation with the aim of recovering and purifying bioactive compounds of interest to produce functional ingredients [
12].
For all the above reasons, the introduction of the UF technology represents a turning point in the production of high-quality, naturally flavored juice [
2]. Through UF, it is possible to obtain an additive-free juice, avoid temperature-induced degradation, and maintain a constant pH for the final product [
8].
Different authors use UF membranes to clarify pomegranate juice. For example, Conidi et al. [
15] tested UF and nanofiltration flat-sheet membranes with nominal molecular weight cut-off (MWCO) ranging from 1000 to 4000 Da to biologically purify active compounds from clarified pomegranate juice. Other authors showed that clarified juice obtained by membrane processing has a more attractive color than fresh juice, which can improve the marketability of the product [
10].
However, the main drawback of these processes is the fouling of membranes [
15]. Therefore, the expansion of membrane technology in the juice clarification industry has been limited by membrane fouling [
16,
17,
18]. Consequently, there is still a need for an in-depth study of the factors affecting the process from the perspective of membrane selection and process conditions.
The objective of this study is, therefore, to screen different membranes in the clarification process of pomegranate juice. The selectivity of the membranes and the efficiency of the process have been compared by means of the rejection coefficient and the permeate flux. Relevant physicochemical parameters such as transmittance, degrees Brix and turbidity, among others, have also been measured. Finally, analysis of the influence of the different fruit varieties and the degree of maturity on the pomegranate juice clarification process by membranes has been carried out.
3. Results and Discussion
The ultrafiltration membranes were pre-treated prior to the filtration of the pomegranate juice. This treatment consisted of the immersion of the membranes in distilled water for 10 min and their passing through the membranes on the membrane module system. Similarly, after the filtration of the different varieties of juice, the equipment was cleaned with distilled water again.
3.1. Screening of the Optimal Membrane for Pomegranate Juice Clarification
This study was carried out using two pomegranate varieties, Mollar and Wonderful, in the initial stage of the season, which involved the low-ripening stage of the fruit. Three types of membranes, with two different chemical compositions, were used, two of them having the same MWCO. The parameters studied in order to choose the optimal membrane were the rejection coefficient and the permeate flux, which represented the selectivity of the membranes and the efficiency of the process, respectively.
Figure 2 shows the rejection coefficient and permeate flux obtained from the different UF membranes for the different varieties of the pomegranate juice used.
The rejection coefficients obtained from the GR-60PP membrane using different pomegranate varieties were around 90%. However, the permeate fluxes were extremely low. The FS-40 membrane obtained a high rejection coefficient from the Mollar variety, but it was low from Wonderful. Although a high permeate flux was obtained from the two varieties, this membrane was discarded because of the rejection coefficient values.
The GR-40PP membrane was selected because of its excellent results; the values of the rejection coefficient were 90% and 80% from the
Mollar and
Wonderful varieties, respectively, whereas the permeate flux values were similar to those obtained from the FS-40 membrane. These results were similar to those obtained by other authors using a PVDF membrane of a similar cut-off size [
20].
Different samples of feed, permeate and concentrate were obtained from the UF process, and parameters such as °Bx, pH, Turbidity, Na
+ and K
+ ions were analyzed.
Table 2 and
Table 3 show the values of different physicochemical parameters for the varieties
Mollar and
Wonderful, respectively.
The absorbance values in the feed, permeate and concentrate samples evolved in the same way for the pomegranate juice of the Mollar variety as for the Wonderful variety when using the FS-40PP membrane. An increase in the absorbance values in the retentate samples with respect to the feed samples, and a decrease in the permeate samples, could be observed. The transmittance of the samples in the concentrate stream was higher than that of the feed for all the membranes tested, with particularly high values for both varieties of pomegranate juice in the GR60-PP membrane, and in the GR-40PP membrane for the Mollar variety.
Regarding sugars measured as °Bx, a decrease in content was observed in the permeate samples, obtaining higher values in the concentrate samples for the GR-40PP and FS-40PP membranes. A similar behavior was observed for both varieties of pomegranate juice with respect to the degree of acidity. The salt content (sodium and potassium) decreased more considerably in the case of the GR-60PP membrane, which could be explained due to the smaller molecular cut-off size of this membrane. The pH value of the pomegranate juice was not affected by the UF process. In the case of the
Mollar variety, pH values oscillated around 4, while for the
Wonderful variety, these values were slightly higher than 3, similar to the results obtained by other authors [
10].
Turbidity showed a significant decrease when comparing the feed and permeate samples, being more pronounced for the GR-60PP and GR-40PP membranes. This may be because both membranes have the same chemical composition (polysulphone). The physicochemical parameters shown in
Table 2 and
Table 3 confirm that the GR-40PP membrane performs adequately for both varieties, confirming the proposed choice.
The above results were compared with the study carried out by Mirsaeedghazi et al. [
20] in which turbidity, pH, °Bx and acidity were measured after the clarification of pomegranate juice with different membranes. It was found that the results for these parameters were similar to those obtained in this study [
10].
In addition, in
Table 4, the absolute quality requirements of the Reference Guide for pomegranate juice of the European Fruit Juice Association (AIJN) have been reviewed and compared with the values obtained in our study [
4].
The GR-60PP and FS-40PP membranes were able to retain a higher amount of sugars for the Mollar variety, and a value below minimum for the °Bx in the permeate. The values for sodium and the formaldehyde index (FI) were within the limits in all samples, while potassium was outside the accepted range for both varieties in the GR-60PP membrane.
3.2. Influence of the Chemical Composition of Membranes on Clarification
The GR-60PP and GR-40PP membranes have the same chemical composition but different MWCO, 25 and 100 kDa, respectively. To compare how the chemical composition affects the clarification of pomegranate juice, FS-40PP and GR-40PP membranes were chosen as they have the same molecular cut-off size (100 kDa), so the only difference between them is the chemical composition. While GR-40PP is composed of polysulphone, the FS-40PP membrane is composed of fluoropolymer. Both materials confer hydrophobic characteristics to membranes, but polysuphone shows oxygen and sulphur dioxide subunits, providing some hydrophobicity to the GR-40PP membrane. Contrarily, fluoropolymer shows CH
2 and CF
2, providing high hydrophobicity to this membrane. Thus, hydrophilic subunits of polysulphone could provide higher capacity to create hydrogen bonds and Van der Waals interaction with phenolic compounds in the pomegranate juice [
12], which confers better characteristics for pomegranate juice clarification.
In
Figure 2A,C, it is observed that, for the
Mollar variety, high rejection coefficients (88 and 96% for GR-40PP and FS-40PP membranes, respectively) and high permeate fluxes (around 0.17 and 0.195 10
−3 kg/sm
2 for FS-40PP and GR-40PP membranes, respectively) were obtained. However, when studying the pomegranate juice of the
Wonderful variety (
Figure 2B,D), the rejection coefficient was higher for the GR-40PP membrane and the permeate flux was higher for the FS-40PP, but both parameters were lower in this variety.
3.3. Influence of the Degree of Ripeness on Pomegranate Juice Clarification
The ripeness stage of pomegranate affects the initial point for clarification. Fernandes et al. [
21] described that the highest flavonoid, phenolic compounds and antioxidant activity were found in juices from the ripeness stage; meanwhile, they were reduced from skin and pellicles within the ripeness process. The amount of sugars is proportional to the ripeness stage, so higher amounts are found in ripe fruits. Something similar was found by Ydjedd et al. [
22] in carob, where the flavonoids and phenolic compound were reduced during ripening.
Figure 3 shows the selectivity and efficiency values for the two types of samples and two maturity stages. In both cases, the membrane selected as optimal in the previous section (GR-40PP) was used.
As it can be observed, the rejection coefficient for both the
Mollar and
Wonderful varieties was higher for the late maturity samples than for the early maturity ones. However, higher permeate fluxes were obtained for the early maturity samples. In the case of the
Wonderful variety, these fluxes were found to be very similar. These expected results were similar to those found by other researchers [
23]. The GR-40PP membrane rejected sugars and phenolics in juice, whose content was higher in the late maturity stage.
Although phenolic compounds were rejected, the clarification processes showed beneficial effects since sugars were also rejected, a finding similar to that of other authors [
24]. Furthermore, the reduction of these substances, and other solids, could decrease turbidity and sensory properties, as described by several authors [
12,
25]. According to Baklouti [
16], the application of UF technology to clarified pomegranate juice decreased the amount of phenolic compounds that cause astringency and bitterness, thus improving clarity but reducing the natural red color.
Table 5 shows the values of the physicochemical parameters for
Mollar and
Wonderful varieties.
The behavior observed in these assays was similar to that obtained from the analysis of the different membranes. The pH value did not undergo significant modifications after passing through the membrane, while the percentage of acidity and the FI decreased in the permeate samples and increased in the samples extracted from the retentate stream. Absorbance values decreased in the permeate samples and increased in the retentate samples, while the opposite behavior was found when studying the transmittance. Sugars decreased after juice clarification, especially for the late samples. When studying the sugar content, it can be found that early samples of the Mollar variety had higher levels of sugars than the late ones, while in the Wonderful variety, the late samples had a greater sugar content. The salt content (sodium and potassium) decreased in the permeate samples, being higher in the early maturity samples than in the late maturity ones. In the retentate streams, the salt content behaved differently, depending on the variety and the degree of maturity.
When comparing the results obtained in the present study on the influence of the ripening degree in the clarification process, it was found that other authors showed similar results (pH values and °Bx) [
26]. The study carried out by Onsekizoglu [
27] highlighted that the organic acid content was maintained during the clarification process due to the low molecular weights of these compounds.
Table 6 shows the values of the Reference Guide for pomegranate juice of the European Fruit Juice Association (AIJN) and the values obtained for both degrees of maturity [
4].
The late Mollar variety failed to obtain a below-minimum value for °Bx and potassium, while the late Wonderful variety obtained an above-maximum value for the FI.
3.4. Fouling Behavior of the Membranes
In order to study the fouling in each membrane, the filtration of water was carried out both before and after the filtration of the pomegranate juice. In this way, the permeate fluxes obtained in both situations could be compared. As the water fluxes are represented as J
w, they are shown in
Figure 4.
It is evident that the permeability of the membranes decreased after the clarification process for all the membranes used. The rapid decrease in the permeate flux was a clear indicator that membrane fouling may have occurred, which also meant a decrease in membrane efficiency.
In
Table 7, fouling index values divided by pressure ranges for the different membranes are shown.
It is observed that the GR-40PP membranes obtained a lower percentage of fouling than GR-60PP and FS-40PP. In general, as the applied pressure increased, the fouling was higher. The GR-40PP1 membrane was used for early varieties and the GR-40PP2 membrane for late varieties. In the tests carried out with the GR-40PP membrane, a higher soiling was obtained for the late samples.
When comparing the results with other research, which used a different ultrafiltration membrane to study the fouling and quality of pomegranate juice, the results showed that the main limiting factor in the use of this type of membrane in the clarification of pomegranate juice was fouling. This phenomenon causes a decrease in permeate flows and, in turn, a decrease in the efficiency of the process [
15]. This occurs because particles larger than the molecular cut-off size of the membranes accumulate on the membrane surface, forming a layer that prevents the filtration of smaller particles.
4. Conclusions
The following conclusions can be drawn from the results obtained in this research.
Comparing the results from the study carried out for the Mollar and Wonderful varieties with the early degree of maturity, it was determined that the optimal membrane was GR-40PP, since it was the one that offered the best results in terms of membrane selectivity and process efficiency. In the same way, the physicochemical parameters studied indicated that this membrane showed excellent results.
In relation to the study of the maturity degree, it was found that the selectivity of the membrane was higher for samples with a late maturity degree, but the process efficiency was higher in early samples. Concerning the study of the chemical composition of the membranes, it could be seen that similar and higher results were obtained for the Mollar variety, whereas for the Wonderful variety, they were lower, with the membrane selectivity being higher for the GR-40PP membrane and the process efficiency being higher for the FS-40 PP membrane.
Finally, after the pomegranate juice clarification process, all the membranes suffered a high decrease in permeability due to the fouling that occurred during the process, which led to a reduction in the process efficiency. The membrane with the lowest percentage of fouling was the GR-40PP, which has been called the optimal membrane and, in particular, the one used with the early samples.