Clarification of citrus fruit juices using microfiltration technique equipped with indigenously developed novel ceramic membrane

Abstract

Microfiltration of citrus fruit juices using membrane technology is a promising method for clarification without losing their inherent properties to extend their shelf life. The present work discusses the development of a tubular ceramic microfiltration membrane and its performance in clarifying two kinds of citrus fruit juices, mandarin and sweet orange. The membrane was prepared by the extrusion method from indigenous bentonite clay, exhibited a porosity of 37% with 0.11 μm pore size, and possessed adequate flexural strength of 18 MPa. The fabricated membrane's potential was evaluated by conducting the tangential filtration of both centrifuged and enzyme-treated centrifuged fruit juices. Also, the applied pressure (68.94–344.7 kPa) and crossflow rate (110–150 Lph) were varied to study the clarified juice properties. At low operating conditions, the highest clarity of the juices was identified despite low permeate flux. The desired properties of juices, including pH, citric acid content, and total soluble solids, were unaffected by pretreatment and tangential membrane filtration, whereas the pectin content, which reduces the juice quality, was eliminated entirely. Furthermore, fouling analysis was carried out using Hermia's models, and cake filtration was identified to be dominant for both juices.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13197-023-05734-y.

Introduction

Citrus fruits are being consumed worldwide due to their several health benefits, and consequently, their worldwide production reached 143.7 million tons in 2019. In India, citrus fruit production increased from 6.87 million tons in 2011 to 13.31 million tons in 2019 (FAO 2021). However, Indian markets have a seasonal supply of citrus fruits, especially mandarin and sweet orange. Mandarins are rich in organic sugars, amino acids, and several vitamins, including vitamin C (14–42 mg/100 gFW), provitamin A (β-carotene: 0.5–3.7 μg/gFW and β-cryptoxanthin: 3.0–13.3 μg/gFW), and folate (16–36 mg/100 gFW). In addition, they contain various beneficial minerals, including potassium (140 mg/100 gFW), calcium (14–47 mg/100 g), magnesium (11–12 mg/100 gFW), iron (0.15–0.60 mg/100 gFW), zinc (0.07–0.50 mg/100 g), and mangansese (0.05 mg/100 gFW) (Lado et al. 2015). Besides, sweet oranges are also an excellent source of vitamin C (53.2 mg). Furthermore, they contain calcium (40 mg), potassium (14 mg), thiamin (0.087 mg), niacin (0.282 mg), and magnesium (10 mg) (Etebu and Nwauzoma 2014; Sarah O et al. 2021). These citrus fruits also contain pectin, fiber, and hemicellulose. The peptides in pectin, polyphenols, and proteins make the juice viscous and form a haze that impacts long-term storage (Rai et al. 2005). In general, the shelf life of citrus fruit juices depends on several factors, including storage conditions, processing conditions, and packaging material, etc. (Corrêa De Souza et al. 2004). Suppose only the constituents of juices are considered. In that case, pectin and polyphenolic components are mainly responsible for the haze and sediment formation, high turbidity, and browning during the more extended storage of juices. In addition, their presence leads to intense fouling during membrane filtration (Baklouti et al. 2012). Therefore, removing pectin and polyphenols from fresh juice is anticipated to reduce the viscosity and improve the juice's browning index, clarity, and shelf life.

Membrane technologies are preferably suitable for the clarification processes over conventional methods owing to their numerous advantages, including low energy consumption and no requirement of any additional additives (Laorko et al. 2010; Gomes et al. 2013). Membrane processes can efficiently eliminate the suspended solids from fresh juice and while preserving its flavour (Laorko et al. 2010). Besides, ceramic membranes are receiving intense research on fruit juice clarification due to their extensive properties, such as operability at higher temperatures and pressures and excellent resistance towards corrosiveness which facilitate them to provide longer shelf life (Nandi et al. 2009; Gomes et al. 2013; Qin et al. 2015). However, high manufacturing costs have limited their applications due to highly-priced inorganic precursors. Therefore, researchers are working towards identifying relatively low-priced raw materials such as kaolin, dolomite, bentonite, sugarcane bagasse ash, etc., making economically feasible ceramic membranes (Vinoth Kumar et al. 2016; Mestre et al. 2019; Sandhya Rani and Kumar 2021; Satyannarayana et al. 2022). However, the main disadvantage of any kind of membrane separation process is fouling, which results in flux reduction. As this phenomenon is unavoidable, pretreatment of juice is usually done to prevent fouling and enhance the permeate quantity. Various pretreatment methods, including centrifugation and enzymatic treatment, followed by centrifugation (ETCJ), are in practice. These methods increase the clarity while reducing the juice's color, viscosity, and pectin content. Also, these methods are industrially practical for longtime storage as they eliminate the pectin content without losing its essential natural properties, such as pH, citric acid content, and total soluble solids (TSS). Specifically, enzyme treatment enhances cluster formation and viscosity reduction, which results in an increased permeate flux and clarity (Maktouf et al. 2014).

Recent research studies have reported applying low-cost ceramic membranes (prepared from low-cost alternative raw materials) in citrus fruit juice clarification. A study fabricated a kaolin-based flat ceramic membrane having 23.6% porosity and 0.285 μm pore size, utilized for the mosambi juice clarification, which was previously ETCJ. It was reported that for a transmembrane pressure of 137.9 kPa, 21.45 × 10⁻⁶ m³/m² s permeate flux was obtained (Nandi et al. 2009). Furthermore, another study elaborated a low-cost ceramic membrane for mosambi juice clarification using dry-compaction method. They have reported that ET before centrifugation is highly suggested as it has dramatically reduced the alcohol insoluble solids (AIS) and irreversible fouling (Emani et al. 2013). These works clarified juices using a dead-end feed flow and studied the effect of applied pressure on clarified juice properties. However, it will result in more concentration polarization and fouling. On the other hand, the cross/tangential filtration mode is more promising over dead-end flow filtration as it eliminates the formation of thick cake layers and thus creates less fouling. A few studies were reported on clarifying citrus fruit juices using low-cost ceramic membranes operated in crossflow filtration mode (Hubadillah et al. 2019). However, they were not considered for the comprehensive characterization of fruit juice clarification and rational discussion on various critical influencing parameters, i.e., operating pressure and crossflow rates. Also, among the various membrane configurations, the tubular configuration could be especially suitable for fruit juice clarification where the feed stream contains a relatively high portion of large particles. Hence, to sustain heavy particles load and severe membrane fouling, tubular configured membranes are effective. The literature shows that no work has been addressed for fruit juice clarification using bentonite-based tubular ceramic membranes so far. In this context, using inexpensive bentonite tubular ceramic membrane to clarify citrus fruit juices would be an effective technique in terms of its low cost and efficiency.

In the present study, the main focus was on clarifying two citrus fruit juices namely mandarin and sweet orange, using a low-cost tubular ceramic membrane operated in crossflow mode. The low-cost ceramic membrane used in this study was fabricated from inexpensive and abundant raw material bentonite clay. In addition to great local abundance and low cost, bentonite clay provides good plasticity and lubricity, which are required during extrusion process. Due to these excellent advantages of bentonite clay, no additional binder or plasticizer is required while fabricating the membrane. The characterization of chosen raw material, the membrane’s fabrication procedure and its characterization were elaborately discussed in the author’s earlier published work (Satyannarayana et al. 2022). In the present work, an indigenously fabricated tubular membrane was utilized for clarification of fruit juices in crossflow mode. Here, initially, the citrus juices namely mandarin and sweet orange, were extracted from their respective fruits and systematically characterized by a particle size analyzer and other critical analytical methods, including pH, color, clarity, citric acid content, total soluble solids, and alcohol insoluble solids. Then the juices were pre-treated by centrifugation and enzyme-treated centrifugation to minimize the extensive fouling. Further, the performance of the fabricated ceramic membrane on the clarification of pre-treated juices is evaluated by analyzing the influence of operating pressure and crossflow rates on the properties of the clarified juices. In addition, the fouling mechanism involved during clarification is intensely discussed using four Hermia’s models.

Materials and methods

Membrane fabrication

Bentonite clay (70 wt%) was used as a significant raw precursor in the preparation of the ceramic membrane, and it was procured locally in India. In addition to bentonite clay, two additives, including rice husk ash (25 wt%) and charcoal (5 wt%), were also added to enhance the membrane's mechanical strength and porosity. The characterization of these three raw materials and the detailed fabrication procedure were deliberated in detail in our recently published work (Satyannarayana et al. 2022). The extrusion method was used for the membrane fabrication, followed by sintering at a temperature of 900 °C. The raw material mixtures were well mixed with distilled water to prepare paste for extruding membranes. The paste was fed into an extruder to get the tubular membrane possessing 100 mm in length, 10 mm outside, and 5 mm inside diameters. Then, the membrane was subjected to thermal treatment in several steps and finally sintered at 900 °C. The fabricated membrane characteristics are presented in Table S1.

Preparation of mandarin and sweet orange juices

Mandarin and sweet orange fruits were procured from a local market in Tadepalligudem (Andhra Pradesh), India. These citrus fruits were gently cleaned with tap water in order to remove extraneous particles from the peel. Then, fresh juice (FJ) was obtained by manually extracting juice from fruits. Both fresh juices were pre-treated by centrifugation at 4000 rpm (Make: REMI, Model: R-24, Rotor type: Angular, Capacity: 4 × 100 ml) for 20 min to get centrifuged juice (CJ), and enzymatic treatment was followed by centrifugation to obtain the enzyme-treated centrifuged juice (ETCJ). In the present study, Pectinase (Aspergillus niger) (SRL, India) with an enzyme activity of 3.5 units/mg was utilized for the enzymatic treatment. This enzyme with 0.0004 w/v%. concentration was used for the enzymatic treatment of juices. After adding the enzyme, the juices were heated in a water bath at a temperature of 40 °C for 100 min. Several studies reported that the temperature used for enzyme treatment did not significantly alter the desired parameters, including pH, citric acid (titratable acidity), and TSS (Rai et al. 2005). In addition to the mentioned parameters, heat treatment’s effect on vitamin C should also be considered as it is an essential nutritional composition of citrus fruit juices. In a study, enzymatic treatment using pectinex enzyme was done on the pulp of umbu fruit which contain vitamin C. They reported that vitamin C was reduced from 10.06 mg/100 g to 7.43 mg/100 g after the enzymatic treatment at 45 °C for 2 h (Souza et al. 2017). In another study, Litchi fruits were clarified using pectinase enzyme at 40 °C for 2 h. It was reported that the ascorbic acid content was observed to be decreased from 17.6 mg/100 g to 11.8 mg/100 g (Vijayanand et al. 2010). From this, it can be inferred that vitamin C content has reduced a little, but a significant amount is present even after the heat treatment done during enzyme incubation. After enzyme treatment, the juices were kept at 90 °C in a hot water bath for 5 min for leftover enzyme deactivation. Further, the temperature of juices was brought to ambient temperature naturally. Then, centrifugation at 4000 rpm was carried out for about 20 min to obtain enzyme-treated centrifuged juices. This enzyme deactivation process did not affect the juice's essential parameters, including pH, citric acid content, and total soluble solids (Nandi et al. 2009; Emani et al. 2013). Indeed, this enzyme deactivation process is essential to maintain the stability and nutritional values during its shelf life. The deactivation of enzymes can be done in two ways: low temperature long time (LTLT) and high temperature short time (HTST). LTLT indicates the heating of juice at a low temperature (63–65 °C) for a long time, whereas HTST indicates the heating of juice at high temperature (90–95 °C) for shorter period. Research studies suggest that HTST helps in preserve pigments and vitamins in the juice (Aghajanzadeh et al. 2021). Accordingly, the current study followed the same. The parameters used in centrifugation, including rpm and time, were chosen by referring to the published literature (Nandi et al. 2009; Emani et al. 2013). Similarly, as enzymatic treatment is affected by several factors, including enzyme concentration, incubation time, and temperature etc., the present work has chosen the operating conditions based on the earlier published works, which have shown better results (Nandi et al. 2009; Emani et al. 2013).

Microfiltration of fruit juices

A crossflow filtration setup, shown in Fig. S1, was utilized for the clarification studies. It mainly includes a feed tank, a membrane module, a diaphragm pump for the transportation of feed, a pressure gauge, and three control valves. The fabricated bentonite-based tubular ceramic membrane having 0.11 µm pore size and 7.899 × 10⁻⁷ m² membrane area was used for the juice clarification. Before utilizing every fresh membrane, the system was operated at 482 kPa with distilled water to open any blocked pores of the membrane. The membrane was placed correctly in the membrane module before the experiment. Then the feed (CJ/ETCJ) was directed into the membrane module using the pump, and the retentate was subjected to recirculation. The obtained permeate was weighed with the help of an electronic weighing balance.

Microfiltration experiments were conducted for both CJ and ETCJ of mandarin and sweet orange juices at five different operating pressures ranging from 68.94 to 344.7 kPa as well as three cross flow rates at 110, 130, and 150 Lph, respectively, to analyze the effects of operating conditions on the characteristics of permeated juice. Three control valves, namely the inlet, retentate, and by-pass, were carefully adjusted to get desired pressure and crossflow rate.

The permeate flux (mandarin and sweet orange juices) was evaluated using the standard expression, Eq. (1).

$$\text{Flux}=\frac{\text{Volume}\text{of}\text{permeate}\left({\text{m}}^3\right)}{\text{Area}\left({\text{m}}^2\right)\times \text{time}\left(\text{s}\right)}$$ 1

All the microfiltration experiments were carried out carefully to get accurate average membrane performance properties.

Analytical methods

A laser particle size analyzer evaluated the particle size distribution for FJ, CJ, and ETCJ (Make: Malvern Panalytical, Model: Zetasizer Ver 7.13). Also, the juice samples were examined for color, clarity, TSS, pH, citric acid content, density, viscosity, and AIS. The absorbance peaks at 420 nm and 660 nm in the UV–Visible spectrophotometer (Make: Shimadzu, Model: UV-1800), giving the color and clarity respectively for both juices. TSS was measured with the help of a digital refractometer (Make: Milwaukee; Model: MA871) (Hubadillah et al. 2019). Furthermore, density and viscosity were measured using the 25 ml pycnometer and Ostwald viscometer, respectively. Citric acid content was measured using the titration method. This method added a few drops of phenolphthalein indicator to a 5 ml juice sample. The titration of this solution was carefully done against 0.1N NaOH until the pink color was observed. The citric acid content was measured based on the volume of 0.1N NaOH utilized (Rai et al. 2005; Nandi et al. 2009). Pectin content present in the juices was measured by observing the AIS. For this, 10 g of juice was mixed with 150 ml of 80% methanol solution. This solution mixture was heated below the boiling point of methanol for 30 min. After simmering, the solution mixture was filtered. The residual solids were dried at 100 °C for about 2 h. Finally, the weight of dry residue was properly measured to obtain AIS in terms of weight percentage (Nandi et al. 2011).

Membrane cleaning

After every experimental run, membrane cleaning was conducted to regain its original permeability (Kumar et al. 2016). It was observed that after each experimental run, the membrane possessed a thick gel layer on its surface. Therefore, the membrane's surface was wiped, and cleaned with distilled water about 15 min in sonication bath. Following this, 0.1 wt% of HCl solution was passed through the membrane for 15 min, followed by cleaning with distilled water. Later, the reduction in the membrane's permeability was evaluated using MF setup. The reduction percentage in permeability after each run was observed to be less than 3% for all the experimental runs. The membrane was utilized to carry out further experiments.

Fouling analysis

Hermia's models interpreted the phenomenon of fouling for crossflow filtration of citrus juices in this work (Kumar et al. 2016). Four types of Hermia's models, standard, complete, intermediate pore blocking, and cake filtration, are often used to predict fouling behavior in membrane separation processes. These four models were pictorially illustrated in Fig. S2. Standard pore blocking raises due to the tortuous nature of the pores, a decrease in pore volume results in the reduction of permeate volume. In case of complete pore blocking, solute particles are considerably bigger than the membrane's pore size and thus eventually block the membrane's surface. In intermediate pore blocking, solute particles and membrane pores are similar in size. The cake filtration model assumes that the solute particles are often relatively bigger than pores, forming a cake layer on the membrane's surface, hence, permeate flux reduces due to added resistance. The four Hermia models can be represented by equations given below:

$$\left(\text{a}\right)\text{Standard pore blocking},J^{-0.5}=J_0^{-0.5}+k_{s}t$$ 2

$$\left(\text{b}\right)\text{Complete pore blocking},\text{ln}{J}^{-1}=\text{ln}{J}_0^{-1}+k_{b}t$$ 3

$$\left(\text{c}\right)\text{Intermediate pore blocking},J^{-1}=J_0^{-1}+k_{i}t$$ 4

$$\left(\text{d}\right)\text{Cake filtration},J^{-2}=J_0^{-2}+k_{c}t$$ 5

Plotting graphs between $J^{-0.5}$ and t, $\ln J^{-1}$ and t, $J^{-1}$ and t, and $J^{-2}$ and t will result in a linear line with $k_s$, $k_b$, $k_i$, and $k_c$ as slope and $J_0^{-0.5}$, $\ln J_0^{-1}$, $J_0^{-1}$, and $J_0^{-2}$ as y-intercepts for best fitting of standard, complete, intermediate pore blocking, and cake filtration model, respectively. Accordingly, the best fit was achieved by interpreting the coefficient of correlation (R²) value.

Several studies used Hermia’s models to determine the flux decline mechanisms during fruit juice clarification. For example, a study observed the fouling mechanism during the filtration of kiwi fruit juice using three membranes possessing different pore sizes. The membrane possessed 0.3 µm pore size, showing a succession of intermediate pore blocking and cake filtration models, respectively (Qin et al. 2015). Another study has used a mixed cellulose ester MF membrane possessing 0.45 µm pore size to clarify bitter orange juice. They concluded that the cake filtration is dominant mechanism for all the applied pressures and feed velocities (Mirsaeedghazi and Emam-Djomeh 2017). Similarly, the present work discussed the fouling mechanism during the filtration of centrifuged and enzyme-treated centrifuged mosambi and sweet orange juices using four of Hermia’s fouling models.

Statistical analysis

In the present study, all the experiments were done three times, and the mean values were given for all the parameters. ANOVA was utilized to compare the means by using SPSS 12.0 (SPSS, Inc., Chicago, IL, USA) software. The differences were considered significant at P < 0.05 (Zhao et al. 2014; Aghajanzadeh et al. 2016).

Results and discussions

The properties of fresh, centrifuged, and enzyme-treated centrifuged juices

Particle size distributions of fresh, centrifuged, and enzyme-treated centrifuged juices were identified and presented in Fig. S3. Also, Table 1 shows that both the juices have nearly similar particle sizes, and it was significantly reduced after pretreatment. The particle size of fresh mandarin juice and sweet orange juice was estimated as 2.92 μm and 2.12 μm, respectively; these values were reduced to 0.79 μm and 0.73 μm, respectively, after centrifugation. In the same way, after enzymatic pretreatment followed by centrifugation, it was reduced to 0.68 μm and 0.64 μm, respectively. From this, it can be observed that juice pretreatment significantly altered the average particle size. The bigger-sized particles were significantly eliminated by centrifugation. Also, the addition of enzyme resulted in the hydrolysis of pectic molecules that facilitates splitting of bigger particles to yield smaller particles (Nandi et al. 2009). However, even after enzyme treatment followed by centrifugation, the average particle size of both juices was more than the hydraulic pore size of the membrane.

Table 1.

Physicochemical properties of mandarin and sweet orange juices (FJ, CJ, ETCJ)

Property of juice	FJ		CJ		ETCJ
	Mandarin	Sweet orange	Mandarin	Sweet orange	Mandarin	Sweet orange
Average particle size (μm)	2.95	2.12	0.79	0.73	0.68	0.64
pH	3.89 ± 0.4	3.9 ± 0.3	4.1 ± 0.2	3.94 ± 0.2	4.0 ± 0.4	3.85 ± 0.5
Density (g/cm3)	1.16 ± 0.12	1.15 ± 0.15	1.15 ± 0.09	1.14 ± 0.10	1.14 ± 0.13	1.14 ± 0.08
Viscosity (Pa. s)	1.67 ± 0.13	3.03 ± 0.02	1.44 ± 0.02	2.181 ± 0.03	1.320 ± 0.02	1.735 ± 0.02
TSS (°Brix)	7.0 ± 0.4	9.0 ± 0.2	6.8 ± 0.5	8.9 ± 0.3	6.9 ± 0.2	8.9 ± 0.3
Color (A420)	1.65 ± 0.3	3.03 ± 0.4	1.13 ± 0.2	1.14 ± 0.3	0.46 ± 0.4	0.58 ± 0.3
Clarity (%T660)	0.92 ± 1.1	0.3 ± 1.3	30.1 ± 2.2	28.8 ± 2.5	65.43 ± 2.1	61.83 ± 2.3
Citric acid (g/L)	0.63 ± 0.05	0.34 ± 0.04	0.62 ± 0.05	0.33 ± 0.06	0.62 ± 0.03	0.32 ± 0.02
AIS (wt%)	0.71 ± 0.06	0.84 ± 0.04	0.42 ± 0.04	0.55 ± 0.03	0.22 ± 0.03	0.21 ± 0.03

The other physicochemical properties of mandarin and sweet orange juices include pH, density, viscosity, TSS, color (A₄₂₀), clarity (%T₆₆₀), citric acid content, and AIS for all the juices, i.e., FJ, CJ, ETCJ were measured and their characteristics are summarized in Table 1. From the obtained results, it can be inferred that color, viscosity, and AIS were significantly reduced for pre-treated juices whereas the clarity was improved (P < 0.05). This reduction in viscosity and AIS after pre-treatment is a desirable consequence that minimizes the gel layer formation on the membrane surface and improves the permeate flux. Furthermore, an improvement in clatrity and a reduction in color after pre-treatment were observed due to the elimination of suspended particles by the pre-treatment process. In addition, density was observed slightly decreased for the enzyme pre-treated juices. Therefore, it can be incidental that pretreatment has significantly eliminated the pectin content and colloidal particles in the juices. However, as the pectinase enzyme chemically reacts with the pectin molecules and degrades them efficiently, better performance was observed in ETCJ when compared to CJ (P < 0.05) (Rai et al. 2007). It is noteworthy to mention that the desirable properties of juices, including pH, TSS, and citric acid content, remained nearly constant even after enzyme pretreatment (P > 0.05).

Microfiltration of mandarin and sweet orange juices

Microfiltration of both centrifuged as well as enzyme-treated centrifuged juices of mandarin and sweet orange was done using a fabricated microfiltration membrane. The physico-chemical properties of clarified juices are presented in Table 2. It can be observed that AIS and viscosity were significantly reduced when compared with CJ and ETCJ (Table 1). This is due to the efficient removal of suspended solids and pectin material during microfiltration process. In addition, a considerable improvement in clarity was observed owing to the elimination of colloidal and suspended particles present in the pre-treated juices.

Table 2.

Physicochemical properties of mandarin and sweet orange juice after clarification

Type of juice	Property of juice	Influence of pressure at constant cross flow rate of 150 Lph (kPa)										Influence of crossflow rate at constant pressure of 206.8 kPa (Lph)
		CJ					ETCJ					CJ			ETCJ
		68.9	137.8	206.8	275.8	344.7	68.9	137.8	206.8	275.8	344.7	110	130	150	110	130	150
Mandarin	pH	3.9	4.1	4.0	4.1	3.9	3.9	4.0	3.9	3.9	3.9	4.1	4.1	4.0	3.9	4.0	4.0
	Density (g/cm3)	1.14	1.14	1.14	1.14	1.15	1.13	1.14	1.14	1.14	1.14	1.14	1.14	1.14	1.13	1.14	1.14
	Viscosity (Pa. s)	1.22	1.22	1.23	1.26	1.29	1.14	1.17	1.18	1.21	1.23	1.21	1.22	1.23	1.15	1.17	1.18
	TSS (°Brix)	6.2	6.0	6.1	6.2	6.1	6.1	6.0	6.2	6.1	6.1	6.0	6.1	6.1	6.0	6.1	6.2
	Color (A420)	0.10	0.11	0.14	0.17	0.19	0.08	0.1	0.12	0.14	0.15	0.11	0.13	0.14	0.09	0.11	0.12
	Clarity (%T660)	93.2	92.9	92.1	91.4	90.8	95.1	94.2	93.6	93.1	92.3	93.18	92.9	92.1	94.6	94.2	93.6
	Citric acid (g/L)	0.63	0.62	0.63	0.63	0.62	0.61	0.63	0.64	0.63	0.63	0.62	0.63	0.63	0.62	0.63	0.64
	AIS (wt%)	0.01	0.01	0.02	0.03	0.03	–	–	–	–	–	0.01	0.01	0.02	–	–	–
Sweet orange	pH	3.8	3.9	3.8	3.9	3.9	3.8	3.8	3.8	3.9	3.8	3.9	3.9	3.8	3.8	3.8	3.8
	Density (g/cm3)	1.14	1.14	1.14	1.14	1.15	1.13	1.14	1.14	1.14	1.14	1.14	1.14	1.14	1.13	1.14	1.14
	Viscosity (Pa s)	1.22	1.22	1.23	1.26	1.29	1.14	1.17	1.18	1.21	1.23	1.21	1.22	1.23	1.15	1.17	1.18
	TSS (°Brix)	8.9	9.1	9.0	9.0	8.9	8.9	9.0	9.1	8.9	8.9	8.9	9.0	9.0	8.9	9.0	9.0
	Color (A420)	0.12	0.13	0.14	0.17	0.21	0.11	0.12	0.13	0.15	0.17	0.11	0.13	0.14	0.1	0.11	0.13
	Clarity (%T660)	91.9	91.5	90.5	90.1	89.2	94.5	93.6	93.2	92.5	91.5	91.2	90.9	90.5	94.1	93.7	93.2
	Citric acid (g/L)	0.32	0.32	0.33	0.32	0.33	0.32	0.31	0.33	0.32	0.33	0.31	0.32	0.33	0.32	0.31	0.33
	AIS (wt%)	0.02	0.03	0.03	0.03	0.04	–	–	–	–	–	0.02	0.02	0.03	–	–	–

Furthermore, the crossflow rate and applied pressure strongly influence the characteristics of clarified juices. Accordingly, the attributes of permeated clarified juices were observed by varying applied pressure and crossflow rate.

Influence of applied pressure

As shown in Fig. 1, for all the pressures, the flux has achieved a steady state value within 55 min of operation for both juices. When the pressure was raised from 68.9 to 344.7 kPa, the flux was increased from 5.3–2.47 × 10⁻⁶ to 15.95–5.08 × 10⁻⁶ m³/m² s for mandarin and 3.97–1.9 × 10⁻⁶ to 13.26–4.64 × 10⁻⁶ m³/m² s for sweet orange centrifuged juices respectively. As pressure is the driving force, similar trends, i.e., an increase in permeate flux with applied pressure, were observed in other research studies (Emani et al. 2013; Hubadillah et al. 2019). Enzymatic treatment noticeably enhanced the permeate flux from 7.95–3.44 × 10⁻⁶ to 11.93–5.99 × 10⁻⁶ m³/m² s for mandarin, and 7.16–3.11 × 10⁻⁶ to 10.61–5.03 × 10⁻⁶ m³/m² s for sweet orange respectively for similar operating conditions (pressure: 206.84 kPa; crossflow rate: 150 Lph). In both cases, a few physiochemical properties, including color, and clarity index, were changed considerably with applied pressure (P < 0.05). For example, in the clarification of mandarin CJ, the color index was increased from 0.1 to 0.19, and accordingly clarity index decreased from 93.2 to 90.8 when the pressure rose from 68.9 kPa to 344.7 kPa. The pigments responsible for color can pass through the membrane at high pressures. As a result, the color index was increased, and consequently, the clarity index was reduced with pressure (Baklouti et al. 2012; Mondal et al. 2016; Le et al. 2021). However, the change in desirable properties, including TSS, pH, and citric acid content, was almost negligible with the pressure (P > 0.05), as shown in Table 2.

Fig. 1.

Influence of applied pressure on permeated juice flux through the fabricated membrane at 150 Lph cross flow rate

Influence of crossflow rate

The influence of crossflow rates on the juice clarification was analyzed using the microfiltration system at 110, 130, and 150 Lph. The flux decline with time at different crossflow rates for CJ and ETCJ was graphically represented in Fig. 2. Permeate flux was identified to be increased at higher crossflow rates owing to the decrease in concentration polarization. In addition, cake layer thickness was also reduced at increased cross flow rates. This is because of induced shear stress on the membrane's surface. From Table 2, it can be observed that the clarity of the clarified juices is slightly reduced with an increase in crossflow rate (P < 0.05). This might be due to the reduced cake layer thickness at higher crossflow rates because the cake layer serves as an extra porous resistance and improves clarity at lower crossflow rates (Kumar et al. 2016). In addition, enzymatic treatment enhanced the clarity compared with centrifugation. For enzyme-treated juice, the clarity was enhanced from 93.2 to 95.1 for mandarin and 91.9 to 94.5 for sweet orange at 68.9 kPa and 110 Lph. On the other hand, change in crossflow rate also did not significantly affect juice's desirable properties, including pH, citric acid content, and TSS (P > 0.05).

Fig. 2.

Influence of crossflow rate on permeated juice flux through the fabricated membrane at 206.8 kPa

Significant difference in clarity was observed at different applied pressures and cross flow rates (P < 0.05). From the results, it can be inferred that the highest clarity of juices was obtained at low applied pressures and crossflow rates. In contrast, greater permeate flux values were obtained at higher operating conditions. But increased pressure demands more power consumption which leads to high costs. Considering these factors, the authors have chosen 206.8 kPa and 150 Lph as the best-operating conditions and analyzed the fouling mechanism for the same.

Here, the low cost tubular ceramic membrane was observed to have membrane porosity 37%, average pore size of 0.11 µm, water permeability of 5.805 × 10⁻⁸ m³/m² s kPa, and corrosion resistance (< 1% weight loss in acidic medium) that are significant to be employed for fruit juice clarification. In addition, the performance of the fabricated bentonite-based membrane was compared with the membranes reported in other research studies. Table S3 shows that the clarity of 93.6% and 93.2% were obtained at an applied pressure of 206.7 kPa for the clarified ETCJ of mandarin and sweet orange. In addition, almost complete removal of AIS content was noticed. Table S3 shows that these results are on par with the results obtained in the earlier published works that clarified other fruit juices. This denotes the significant effieciency of the fabricated low-cost tubular bentonite-based membrane.

Analysis of fouling mechanism involved in the clarification of juices

Plots between J^−0.5, J⁻¹, ln(J⁻¹), and J⁻² versus t were drawn according to Eqs. (2–5), respectively, to observe the fitting of experimental data in four kinds of fouling models. The mechanisms involved during flux declination in clarifying CJ and ETCJ of mandarin and sweet orange juices can be observed from the plots shown in Fig. 3. It can be observed that the flux declination is more during the first 20 min of operation (Figs. 1, 2). Therefore, the flux decline mechanism was identified in two steps: in the initial stage (up to 20 min of operation) and the final stage (after 20 min). The fitness was analyzed using the coefficient of correlation (R²), and these values for both juices were given in Table S2. From this, intermediate pore blocking (R² ranges from 0.960 to 0.989) and cake filtration models (R² ranges from 0.980–0.998) seemed to be more dominant mechanisms in the initial stage for clarifying both mandarin and sweet orange CJ and ETCJ juices. However, the cake filtration mechanism was significantly dominating during the later stage. As there is not much difference between intermediate and cake filtration mechanisms in the first stage, it can be determined that, as a whole, cake filtration is the best fit with an R² value of 0.971 and 0.979 for mandarin and sweet orange CJ juices, respectively (Table 3).

Fig. 3.

Linearized plots of various fouling models for the clarification of sweet orange and mandarin CJ and ETCJ juices

Table 3.

Summary of parameters concerning four fouling models plotted for mandarin and sweet orange juice clarification (CJ and ETCJ)

Juice type	Standard pore blocking			Complete pore blocking			Intermediate pore blocking			Cake filtration
	R2	ks	J0−0.5	R2	kb	ln (J0−1)	R2	ki × 10−4	J0−1 × 10−5	R2	kc × 10−9	J0−2 × 10−10
Mandarin CJ	0.891	3.05	377.6	0.851	0.013	11.88	0.924	0.28	1.39	0.971	1.22	1.57
Sweet orange CJ	0.906	3.18	397.5	0.867	0.013	11.98	0.938	0.30	1.53	0.979	1.50	1.89
Mandarin ETCJ	0.908	1.95	303.1	0.876	0.011	11.44	0.935	0.13	0.90	0.975	0.36	0.71
Sweet orange ETCJ	0.916	2.35	320.9	0.885	0.012	11.55	0.941	0.18	1.00	0.976	0.55	0.84

Similarly, for the clarification of the enzyme-treated centrifuged juices, the best fit for the entire operation was obtained for cake filtration with an R² value of 0.975 and 0.976 for mandarin and sweet orange juices, as mentioned in Table 3. As shown in Table S3, the observed fouling mechanism is similar to the mechanism followed during the clarification of citrus fruit juices in other research studies. As cake filtration is reversible in nature, this fouling analysis suggests that the fabricated membrane possesses an antifouling character and could be regenerated after proper cleaning. In addition, k_s, k_i, k_b, and k_c values were estimated for all the fouling models. This value denotes the severity of fouling that occurred in the membrane during the filtration process (Vinoth Kumar et al. 2015). This value was observed to be decreased for the enzyme-treated juices compared to centrifuged ones. For instance, the k_c value was significantly reduced for enzymatically treated juices, as shown in Table 3, revealing that the fouling has been considerably reduced for enzyme pre-treated juices.

Conclusion

A low-cost ceramic membrane was prepared successfully from indigenous and inexpensive bentonite clay via extrusion, followed by sintering of 900 °C. The prepared membrane's performance was effectively assessed on clarifying citrus fruit juices, namely mandarin and sweet orange. Microfiltration studies revealed that the clarity of juices was identified to be more at low operating conditions. On the other hand, as pressure and crossflow rates were raised, the clarity was reduced a little despite high permeate flux. Enzymatic treatment significantly enhanced the permeate flux as well as clarity of the juices and reduced the fouling of the membrane. Also, complete removal of AIS content was obtained for both juices at all the operating pressures and crossflow rates. In all the cases, the clarified juices retained their original pH, citric acid content, and TSS, which signifies the advantage of clarification by ceramic membrane over conventional thermal and chemical treatment methods. The fouling analysis was done for all the clarification experiments by applying four Hermia models to identify the appropriate fouling mechanism. Among four, cake filtration appeared as the best fit for clarifying both juices, which recommended that the fabricated membrane possesses antifouling nature and could be regenerated after proper cleaning.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

KVVS conceptualized, carried out the experiments, analyzed and wrote the manuscript; SLSR contributed to the analysis and writing of the manuscript; RVK conceived the idea, supervised, examined the work, and reviewed and edited the manuscript.

Funding

This work was financially supported by Science and Engineering Research Board, Department of Science and Technology, Government of India (File No: EEQ/2018/001432).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.