- Email: [email protected]
Sequential fractionation of value-added coconut products using membrane processes
G Model
JIEC-2269; No. of Pages 6 Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Sequential fractionation of value-added coconut products using membrane processes Ching Yin Ng a, Abdul Wahab Mohammad a,b,*, Law Yong Ng a, Jamaliah Md Jahim a,b a
Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia b Research Centre for Sustainable Process Technology (CESPRO), Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia
A R T I C L E I N F O
Article history: Received 3 March 2014 Received in revised form 7 October 2014 Accepted 13 October 2014 Available online xxx Keywords: Fractionation Ultrafiltration Nanofiltration Skimmed coconut milk Value-added products
A B S T R A C T
The coconut waste-skimmed coconut milk was employed for sequential fractionation using UF and NF membranes to produce value-added products (coconut proteins, plant hormones – kinetin and zeatin). The retention factors achieved by UF membrane (PS10): albumin (0.9822 0.0799) and globulin (0.9975 0.0783); NF membrane (NF1): kinetin (0.9238 0.0345) and zeatin (0.9511 0.0355). Coconut protein powder was obtained after spray-drying process using concentrated coconut protein (UF retentate). SDS-PAGE showed that molecular weights of the coconut proteins were 17, 34, 55 and 150 kDa. Proximate and HPLC analyses revealed that the obtained samples were enriched with basic nutrients and well-balanced amino acids composition, respectively. ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Coconut (Cocos nucifera L.) is an important palm species around the region of Asia Pacific and South East Asia. It is commonly used as the cooking ingredient in these countries. Besides, coconut is also famous with its versatile applications [1]. One of the renowned coconut products is virgin coconut oil (VCO). VCO has been recognized as the healthiest crop oil and can be extensively employed in various fields such as food, beverage, medicinal, pharmaceutical, nutraceutical, cosmetic, etc. [2]. The high-valued fatty acids and anti-viral properties of the VCO have increased the demand of VCO worldwide and thus its production [3]. Large amount of valuable by-products have been discarded during the production of VCO. This has been considered as a wastefulness which also creates environmental issue in some areas near to the coconut processing plant. These discarded by-products still contain some high-valued nutritional components such as coconut protein, plant hormones, vitamins, minerals, amino acids and so
* Corresponding author at: Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia. Tel.: +603 89216102; fax: +603 89252546. E-mail addresses: [email protected] (C.Y. Ng), [email protected], [email protected] (A.W. Mohammad), [email protected] (L.Y. Ng), [email protected] (J.M. Jahim).
forth, which are highly beneficial to the consumers [4]. These valuable components can be potentially extracted by suitable separation techniques or other appropriate processes to fully utilize the raw materials employed during the VCO production. Full utilization of these by-products would definitely increase the profit of the coconut industry. The skimmed coconut milk is the main by-product after the coconut oil extraction process. The skimmed coconut milk consists of approximately 70% of the total proteins, carbohydrates, sugars, vitamins and minerals [5]. Through fractionation and concentration processes, the by-products from skimmed coconut milk can be employed in functional food, food supplement and food formulation. In addition to the compounds mentioned above, there are some other smaller-size minor components exist in the skimmed coconut milk. For example, there is a group of plant hormone called cytokinins which consists of kinetin, zeatin and traces of other compounds [6,7]. The kinetin and zeatin have been recognized as valuable compounds which possess anti-viral, anti-bacteria and anti-aging properties. They can be employed to stimulate the cell division of human body [1,8]. If the kinetin and zeatin from skimmed coconut milk can be separated and concentrated using appropriate processes, they will have huge potential to be applied in many applications such as medicinal, pharmaceutical and nutraceutical sectors [9,10]. The great versatility of these valuable compounds from the
http://dx.doi.org/10.1016/j.jiec.2014.10.028 1226-086X/ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 2
C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
coconut has attracted the attentions of many researchers to produce more profitable value-added coconut products. Many attempts have been tested to separate the valuable compounds in agricultural or food products such as distillation, evaporation, chemical extraction and so forth [11,12]. However, the conventional separation methods which involve heat treatment will deteriorate some thermal sensitive compounds like proteins [13,14]. Membrane filtration is an environmental friendly process compared to conventional methods [15]. The advantages of membrane filtration technique including low labor cost, low maintenance cost, less space consuming, energy saving, feasible to scale up, no chemicals used, no heat treatment, able to prolong the shelf-life of product, can retain the natural flavor and aroma of raw materials [16,17]. Due to the numerous advantages possessed by membrane filtration, it is used widely from resource recovery to wastewater treatment [18–21]. Therefore, the membrane separation technique is proposed in this study as one of the most probable ways to fractionate the skimmed coconut milk to produce those value-added products. In food technology field, the fractionation method using membrane technology has been well developed since past decades. For example, the fractions of fat and volatile compounds from goats’ milk cheese has been successfully obtained using the fractionation method by sequential ultrafiltration and nanofiltration processes [16]. Compounds with varying molecular sizes managed to be retained by different membrane categories such as microfiltration, ultrafiltration and nanofiltration. In the fractionation of dairy products, ultrafiltration [22,23] and nanofiltration [24–26] are always employed in order to obtain the valuable milk components. The protein hydrolysates can be efficiently separated by nanofiltration process due to the size exclusion mechanism and Donnan effect. In beverage industry, the membrane filtration technique is a common method to clarify and concentrate the fruit juice. Combination of various membrane types have been employed to enhance the quality of the produced juice [27–29]. However, there is not much study about the fractionation of coconut milk using membrane technology to obtain the valuable compounds. If the membrane processes can be employed to fractionate and concentrate the valuable nutrients from coconut waste effectively, this will contribute to a great breakthrough in the coconut industries. Besides that, implementation of spray-drying process could further enhance the quality of value-added products produced (such as coconut protein). Spray-drying is the most widely used commercial dehydration method owing to its short duration of heat-contact and high rate of evaporation, which can produce high quality products. Powders with precise specifications can be generated using spray-drying process in continuous operation [30]. The objective of present study is to determine the capability of sequential ultrafiltration and nanofiltration processes to fractionate the value-added products from skimmed coconut milk. In addition, the membrane performance will be evaluated and discussed in terms of normalized flux decline and solute retentions for protein, kinetin and zeatin compounds. Materials and methods Materials Coconut of Malayan Tall variety has been selected for this research work. To determine the concentration of albumin and globulins proteins in UF streams, standard bovine serum albumin (BSA), standard immunoglobulin (IgG) and protein assay dye reagent concentrate were supplied by Bio-Rad (USA). During the UF process, polysulfone (PS) membrane with 10 kDa (supplied by Koch, USA) was employed. The employed NF membrane was
manufactured by Amfor Inc., China. High performance liquid chromatography (HPLC) analysis was used to determine the concentration of kinetin and zeatin in NF streams. The standards of kinetin and zeatin were supplied by Sigma-Aldrich (Steinheim, Germany). Kinetin and zeatin solutions with the concentration ranging from 10 to 500 mM were prepared by dissolving the kinetin and zeatin standards in HPLC-grade methanol. These varying concentrations of kinetin and zeatin standards were stored at or below temperature of 4 8C. The chemicals used for HPLC analysis including HPLC-grade methanol (Tedia, USA), formic acid (Tedia, USA) and triethylamine (TEA) (Merck, Germany). Buffer solutions employed in the HPLC analysis has been prepared using methanol and 0.1% formic acid with pH 3.2. Preparation of skimmed coconut milk The milky white color fresh coconut milk was produced from solid grated coconut endosperm using a coconut extraction machine. Then, the extracted fresh coconut milk was filtered through micro-size sieving cloth to remove the bigger-size particles. Simple pasteurization process was conducted prior to the membrane filtration process in order to prolong the shelf-life of extracted coconut milk. The fresh coconut milk was heated at 60 8C for 15 min. This pasteurization method was conducted in order to reduce the microbial loading up to 10% according to the Hagenmaier method [31]. After that, the pasteurized coconut milk was poured into 125 L capacity cream separator machine (Elecrem, France) to separate the coconut cream/fat from the fresh coconut milk. There are two outputs from the cream separator: concentrated coconut cream and skimmed coconut milk. The produced skimmed coconut milk needs to be stored at a cold condition (0–4 8C) before it is used as a feed solution for the membrane filtration process. Membrane filtration processes (UF and NF membranes) were then employed to fractionate the skimmed coconut milk to produce the desired value-added coconut products. Membrane filtration process The polysulfone (PS) membrane was employed in the UF process. Polysulfone membrane was employed owing to its several superior properties such as high material toughness, good stability at high temperature, high resistant to various solvent, good resistant to wide pH range (pH 2–13) and low protein binding tendency [32]. The molecular weight cut-off (MWCO) value of this PS membrane is 10 kDa. The abbreviation of the UF membrane used in this experiment is PS10. NF1 membrane was employed in the NF filtration process. The membrane material of NF 1 is polyamide. Polyamide provides the desired properties such as high rejection of undesired materials (like salts), good mechanical strength and high filtration rate at low pressure. Prior to the filtration process, the membranes were soaked overnight to remove any preservative layer and dirt particle. Membrane compaction step was carried out at a higher pressure value (greater than operating pressure) for each employed membrane in order to enhance the permeate flux and membrane permeability [33]. Ultra-pure water was used during the membrane permeability test. In this study, a cross-flow system was used in the UF process. The cross-flow system was equipped with a 4 L stainless steel jacketed feed tank (embedded with a mixer) and a variable feed pump (Hydra-cell Pump, Mn, USA). The temperature of feed solution was controlled by the circulating water within the jacketed feed tank. The cross-flow velocity (CFV) of feed and the trans-membrane pressure (TMP) was controlled using a feed flowmeter (F-400, Blue-White, USA) and a permeate needle valve
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
(Swagelok, UK), respectively. A digital balance (GF6100, A&D, Japan) was connected to a computer to record the permeate flux in the continuous mode. The active surface area of PS10 membrane in cross-flow filtration system is 32.06 cm2. The operating pressure of 2 bar and temperature of 60 8C were used during the UF process of skimmed coconut milk. The CFV of 2.726 L/min and a stirring rate of 42 rad/s were maintained throughout the whole process of UF using PS10 membrane. In this study, a volume reduction factor (VRF) of 2 was set in each experimental run under batch concentration mode (the retentate was recycled back to the feed tank). The mass of collected permeate was recorded every 60 s with the help of the digital balance for permeate flux calculation. The collected UF permeate was stored for further fractionation process using NF1 membrane. Dead-end filtration module (Sterlitech HP4750 stirred cell, WA, USA) with a membrane active surface area of 15.20 cm2 was used to perform the NF process. Constant operating pressure of 5 bar and temperature of 25 8C were set in the NF process. A compressed nitrogen cylinder was connected to the dead-end filtration cell to supply the desired pressure. Duration employed for the NF process was 360 min. The mass of NF permeate was recorded by a connected digital balance for every 60 s throughout the filtration process for permeate flux calculation. Evaluation of membrane performance The performances of the employed UF and NF membranes were evaluated in terms of solution fluxes (expressed by the unit L/m2 h) and membrane retention capabilities towards particular solutes (albumin protein, globulin protein, kinetin and zeatin). In order to determine these parameters, Eq. (1) has been employed. Solution flux; Jw ¼
Q ADT
(1)
where Q is determined by the quantity of collected permeate in unit liter (L); A is the active surface area of used membrane (m2) and DT is the time interval of sampling (h). The normalized flux of produced permeate was calculated according to the following Eq. (2): Normalized flux ¼
Jw Jo
(2)
where Jw (L/m2 h) is the actual permeate flux at steady state and Jo (L/m2 h) is the initial flux of pure water. To determine the retention capabilities of PS10 and NF1 membranes towards coconut proteins (albumin and globulin) and cytokinins components (kinetin and zeatin), respectively, the following Eq. (3) has been used. Retention factor; Ret:F ¼ 1
Q pf QR
(3)
where Qpf (mg/L) is the concentration of solute in permeate and QR (mg/L) is the solute concentration in retentate stream. To calculate the volume reduction factor of the feed solution in UF process, the following Eq. (4) was employed. Volume reduction factor; VRF ¼
Qo Q o Q p Q p0
(4)
Where Qo (L) is the initial feed solution quantity; Qp (L) is the permeate quantity and Qp0 (L) is the permeate quantity loss during the filtration. All of these performance indicators are important in order to understand the overall methodology of filtration processes and the results will be discussed.
3
Spray-drying process After the UF process using PS10 membrane, the produced coconut protein solution was sent for spray-drying process in order to produce coconut protein powder. The Lab Plant brand spraydyer (Filey, UK) was used to perform the spray-drying process. Prior to the spray-drying process, the concentrated coconut protein solution was added with 5% of malto-dextrin and 20 Brix has been measured by a hand-held refractometer (Erma, Japan). The homogenized and concentrated coconut protein solution was fed into the spray-dryer by a peristaltic pump at a constant flowrate of 1.75 ml/min. The solution was then atomized through a 2 mm diameter nozzle into small droplets at 7 bar air pressure within a co-current air-flow system. The temperature of feed solution was set at 25 1 8C. The inlet and outlet temperatures of spray-dryer were 170 1 8C and 85 1 8C, respectively. The spraydried coconut protein powders were stored for further analyses. Analysis In this work, the targeted products throughout the UF and NF processes are: albumin proteins (measured as bovine serum albumin (BSA)), globulin proteins (measured as immunoglobulin (IgG)) [34,35], kinetin and zeatin. The dominant proteins in coconut milk are albumin and globulin proteins; whereas the kinetin and zeatin components are recognized as plant hormone or cytokinins. The Bradford method was applied in order to determine the concentration of albumin and globulin proteins in the feed, retentate and permeate streams after the UF process. BSA and IgG with varying concentrations were prepared and analyzed by a UVspectrophotometer (Genesys, Moston, MA) prior to real sample analysis. The purpose of this step was to obtain the standard plot of BSA and IgG. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis was also performed in order to determine the molecular weights of proteins which were presence in the UF streams (feed, retentate and permeate). Besides, the samples from UF streams (feed, retentate and permeate) and spray-dried coconut protein powders were sent for proximate analysis to determine the presence of basic nutrients (including protein, fat, carbohydrate, moisture and ash). Association of Official Analytical Chemists (AOAC) method was used throughout the proximate analysis [36]. In addition, the amino acids profile analysis using high performance liquid chromatography (HPLC) instrument was carried out in order to determine the amino acids compositions within the concentrated coconut protein solution. An AccQ Tag Column with a dimension of 3.9 150 mm was used during the HPLC analysis. The employed HPLC instrument is equipped with a Waters 410 Scanning Fluorescence. The unit used to report the amino acids composition is in mass (g) of amino acid per 100 g of protein, (g/100 g). The kinetin and zeatin were the targeted products after the NF process. To determine the concentrations of kinetin and zeatin in the NF streams (feed, retentate and permeate), a high performance liquid chromatography (HPLC) system (Agilent Technologies 1200 Series, Santa Clara, USA) was used. To detect and separate the kinetin and zeatin from other components within the sample, a UV detector was employed. During the analysis, the data was recorded and processed by the accompanying system software (ChemStation for LC 3D System). The samples were preliminarily filtered by a 0.45 mm Whatman glass microfiber filter prior to HPLC analysis. To initiate the analysis, 10 ml of filtered sample was injected into a C18 reverse phase column (Zorbax SB-C18 100A´˚ , 150 mm in length, 2.1 mm in diameter, Agilent Technologies, Santa Clara, USA). Initially, the HPLC system running condition was: methanol (0.1%)–formic acid buffer (10:90, v/v). The column thermostat was set at 25 8C. A flowrate of 0.3 ml/min was set
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
4
throughout the whole separation process. To achieve high detection accuracy of kinetin and zeatin components, the wavelength was fixed at 265 nm. Results and discussion Normalized flux declines during the ultrafiltration and nanofiltration processes PS10 membrane was employed in the cross-flow system to fractionate the skimmed coconut milk to produce concentrated coconut protein solution during UF process. For the NF process, NF1 membrane was employed to further fractionate the UF permeate in order to obtain kinetin and zeatin compounds. To achieve the VRF of 2 in the UF process, the filtration period spent by the cross-flow system was 360 min. Time taken by the NF1 membrane to complete the filtration was also roughly 360 min. To evaluate the performances of PS10 and NF1 membranes, the following normalized flux decline curves (Fig. 1) were plotted. Prior to the plotting of normalized flux curves, simple calculations of solution fluxes and normalized fluxes were made according to the Eqs. (1) and (2). According to Fig. 1, the UF process showed higher normalized fluxes compared to the NF process. However, at the initial stage of filtration processes, the normalized fluxes of UF process dropped more drastically than the NF process, which were followed by gradual decline along the filtration time. All of these phenomena can be clearly seen from Fig. 1. The main factor contributed to the sharp decline of normalized fluxes at the initial stage is dominated by concentration polarization phenomenon [22,37–39]. The steady states of normalized fluxes were achieved for both UF and NF processes as the durations of filtration were extended. Generally, feed sample contains considerable amount of proteins will eventually lead to membrane fouling [40]. During the filtration processes, the foulants may build up thus forming the fouling cake layers on top of the membrane surfaces. This may in turn affect the selectivity of the membrane towards certain desired solutes [38,41]. Retention factors of albumin and globulin proteins by PS10 membrane During the UF process, PS10 membrane was used to concentrate the proteins of skimmed coconut milk using cross-flow system. Predominant types of coconut proteins are albumin and globulin types. Thus, the concentrations of albumin and globulin proteins in the UF streams (permeate and retentate) were determined
Fig. 2. Retention factors of albumin, globulin proteins (by PS 10 membrane), kinetin and zeatin (by NF 1 membrane).
according to the Bradford method. The obtained results were substituted into Eq. (3) to calculate the retention factors of albumin and globulin proteins. Fig. 2 shows the retention factors chart of albumin and globulin proteins for UF process. As it can be seen from Fig. 2, both retention factors (albumin and globulin) achieved very high reading. The retention factor results for albumin and globulin were 0.9822 0.0799 and 0.9975 0.0783, respectively. It can be concluded that the albumin and globulin proteins in the skimmed coconut milk had been successfully retained and concentrated by PS10 membrane during the UF process. The results obtained were comparable to several other works in which the retention factors for albumin and globulin were 0.83–0.97 and 0.96–0.98, respectively [42,43]. Retention factors of kinetin and zeatin by NF1 membrane After the fractionation process of skimmed coconut milk using UF membrane (PS10), the collected UF permeate still contained numerous high nutritional components. The UF permeate was further fractionated by NF1 membrane to obtain additional valueadded products. One of the high-valued components is cytokinins. It consists of kinetin and zeatin components. These components have been identified to possess anti-aging properties and have high potential to be used in pharmaceutical, medicinal and cosmetic products. Thus, kinetin and zeatin were set as target products for the NF process. After the NF process, the concentrations of kinetin and zeatin in the NF streams (feed, retentate and permeate) were determined by HPLC analysis. The obtained results were calculated using Eq. (3). Fig. 2 also shows the retention factors of kinetin and zeatin in the NF process using NF1 membrane. Both components (kinetin and zeatin) had been successfully retained by NF1 membrane with their retention factors greater than 0.90. The retention factors achieved by NF process for kinetin and zeatin were 0.9238 0.0345 and 0.9511 0.0355, respectively. Thus, the results show that the kinetin and zeatin can be effectively retained by employing NF1 membrane during the NF process. SDS-PAGE of coconut proteins
Fig. 1. Normalized flux decline plots of UF and NF processes by PS10 and NF1 membranes, respectively.
The gel electrophoresis method using SDS-PAGE to determine the molecular weights of proteins is presented in Fig. 3. The molecular markers with molecular weights ranging from 10 kDa to 160 kDa were used in this analysis. The molecular markers can be clearly seen at the left lane in Fig. 3. UF feed, retentate and permeate samples were employed for the SDS-PAGE analysis for protein molecular weight identification. By analyzing the SDSPAGE profile (Fig. 3), it can be seen that no protein bands were detected in the UF permeate stream. Thus, it can be postulated that no protein with molecular weight greater than 10 kDa had
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
5
Table 2 Compositions of amino acids in the concentrated coconut protein.
Fig. 3. Molecular weights of coconut proteins by SDS-PAGE analysis.
successfully passed through the PS10 membrane during the UF process. It is obvious to see that the protein bands appeared in the UF feed were almost identical to the UF retentate. The intensity of protein bands in UF retentate is higher than UF feed. Again, this has proved that the coconut protein in the skimmed coconut milk had been successfully concentrated by PS10 membrane during the UF process. The major bands appeared in the SDS-PAGE profiles represented the dominant protein fractions in the UF streams. These dominant protein fractions possessed molecular weights of 17, 34, 55 and 150 kDa. The identified protein fractions are categorized as albumin and globulin types, which are accounting for 60% and 30% of total protein, respectively [11,44]. In the previous study, the total coconut protein was found to have a relatively similar pattern as shown in Fig. 3 [35]. Proximate analysis and amino acids profile of concentrated coconut proteins Table 1 shows the results of proximate analysis for the UF streams (feed, retentate and permeate) and spray-dried coconut protein powders. The compositions of some basic nutrients (protein, fat, carbohydrate, ash, moisture and energy) were reported in the unit of percentage or g/100 g of sample. According to the results obtained, each of the UF feed, retentate and permeate is mostly composed of moisture or water content. The moisture content for each of the UF stream is approximately 90%. After Table 1 Proximate composition of UF streams (feed, retentate and permeate) and coconut protein powder. Parameter (g/100 g)
UF feed
UF retentate
UF permeate
Coconut protein powder
Protein Fat Carbohydrate Ash Moisture Energy (kcal/100 g)
2.10 0.80 4.50 1.00 91.60 34.00
6.50 0.30 3.50 0.80 88.90 19.00
0.20 0.00 2.90 0.80 96.10 12.00
32.85 0.30 58.35 5.10 3.40 368.00
Amino acids
Composition (g/100 g protein)
Aspartic acid Serine Glutamic acid Glycine Arginine Alanine Valine Lysine Isoleucine Leucine Phenyl alanine Histidine Threonine Proline Tyrosine Methionine
8.95 2.04 21.50 3.55 14.20 2.19 2.99 3.74 2.07 3.17 3.86 1.80 2.50 3.60 1.80 1.80
the spray-drying process of concentrated coconut protein solution, the water content within the coconut protein powder left only 3.40%. The protein content in the UF retentate was 6.50%. The protein content in the UF retentate was higher than the UF feed (2.10%) and UF permeate (0.20%). This shows that the protein in the feed solution (which is skimmed coconut milk) has been successfully concentrated by the UF membrane (PS10). It has been identified that there are approximately 32.85% of protein in the coconut protein powder. UF permeate showed zero fat content, while UF feed consisted of 0.80%, UF retentate consisted of 0.30% and coconut protein powder consisted of 0.30%. The proximate analysis results proved that the spray-dried coconut protein powders produced in this work possessed high content of protein (32.85%) and carbohydrate (58.35%). Table 2 displays the compositions of amino acids in the concentrated coconut protein. The concentrated coconut protein solution showed a well-balanced essential amino acids profile. Compositions of glutamic acid, arginine and aspartic acids were found to be higher in comparison to other fractions of amino acids. However, the concentrated coconut protein is deficient in the sulfur amino acid—methionine. Comparable findings were reported by other researchers [5,11]. In brief, the produced concentrated coconut protein and coconut protein powder are highly suitable to be employed for human consumption and used as functional food in the related industries. The economical and nutritional values of the coconut by-products will be worth for future development especially in the food industries. Conclusion In the present work, UF and NF processes have been successfully conducted using cross-flow system and dead-end module, respectively. According to the normalized flux decline curves performed by the UF (PS10) and NF (NF1) membranes, the normalized fluxes displayed typical declining trend over the filtration time. It has been verified that these membranes have been successfully employed to concentrate the coconut proteins (albumin and globulin) and separate the valuable components (kinetin and zeatin) from the skimmed coconut milk. Most of the albumin and globulin proteins in the skimmed coconut milk can be retained by the employed UF membrane (PS10). The filtration process using NF membrane (NF1) had been successfully employed to further fractionate the valuable compounds (kinetin and zeatin) within the UF permeate. After the concentration process using UF (PS10) membrane in the cross-flow system, the achieved retention factors of albumin and globulin proteins were 0.9822 0.0799 and 0.9975 0.0783, respectively. For
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 6
C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
the NF process, the retention factors obtained for both kinetin and zeatin were greater than 0.90. The retention factor achieved by kinetin and zeatin were 0.9238 0.0345 and 0.9511 0.0355, respectively. The success in fractionating the kinetin and zeatin by employing membrane filtration processes in this study will encourage the future employment of these compounds in the medicinal, pharmaceutical and nutraceutical fields due to their own anti-viral and anti-aging properties. Through the solute retention test, it has been verified that the albumin and globulin proteins in the skimmed coconut milk can be successfully concentrated by the UF process. By analyzing the SDSPAGE profiles, four major protein bands were identified in the UF samples (feed and retentate) having molecular weights of 17, 34, 55and 150 kDa. Besides, SDS-PAGE profiles also show that the intensity of protein bands in the UF retentate is higher than UF feed and permeate. Thus, the SDS-PAGE profiles have concluded that the coconut protein was successfully concentrated in this study. In addition, the results of proximate analysis which also verified that the protein content in the UF retentate was higher than the UF feed and permeate. Coincidently, this observation is in good agreement with the results obtained through SDS-PAGE analysis. Relatively well-balanced amino acid compositions in the concentrated coconut protein were obtained through the HPLC test. To make the coconut protein easier for handling, spray-drying process had been conducted using concentrated coconut protein solution. Coconut protein in powder form has been successfully produced through this work. Proximate analysis shows the coconut protein powder consists of 32.85% protein, 58.35% carbohydrate, 0.30% fat, 5.10% ash and 3.40% moisture. As a conclusion, the sequential UF and NF processes in this work have successfully produced the targeted or desired value-added coconut products (coconut protein, kinetin and zeatin). Last but not least, full utilization of the discarded coconut byproducts by UF and NF processes to produce highly valuable coconut value added products will bring a great breakthrough to Coconut industries and food technology field. Acknowledgment
[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39]
The authors of this work wish to gratefully acknowledge the financial support of the Ministry of Science and Technology (ScienceFund) through Project no. 02-01-02-SF1021. The authors would also like to acknowledge Malaysia Ministry of Education for the MyBrain15 Scholarship provided.
[40] [41] [42]
References
[43]
[1] J.W.H. Yong, L. Ge, Y.F. Ng, S.N. Tan, J. Mol. 14 (2009) 5144.
[44]
A.M. Marina, Y.B. Che Man, I. Amin, Trends Food Sci. Technol. 20 (2009) 481. B.J. Villarino, L.M. Dy, M.C.C. Lizada, LWT – Food Sci. Technol. 40 (2007) 193. J.L. McCullough, G.D. Weinstein, Cosmet. Dermatol. 15 (2002) 29. C.C. Seow, C.N. Gwee, Int. J. Food Sci. Technol. 32 (1997) 189. L. Ge, J.W.H. Yong, N.K. Goh, L.S. Chia, S.N. Tan, E.S. Ong, J. Chromatogr. B 829 (2005) 26. N.C. Yin, A.W. Mohammad, N.L. Yong, Adv. Sci. Lett. 19 (2013) 3620. Z. Ma, L. Ge, A.S.Y. Lee, J.W.H. Yong, S.N. Tan, E.S. Ong, Anal. Chim. Acta 610 (2008) 274. T.M. Kajs, R. Hagenmaier, C. Vanderzant, K.F. Mattil, J. Food Sci. 41 (1976) 352. J. Voller, M. Zatloukal, R. Lenobel, K. Dolezal, T. Be´res, V. Krystof, L. Spı´chal, P. Niemann, P. Dzuba´k, M. Hajdu´ch, M. Strnad, Phytochemistry 71 (2010) 1350. S.J. A.S. Samson, R.N. Khaund, C.M. Cater, K.F. Mattil, J. Food Sci. 36 (1971) 725. B.-K. Chen, L.L. Diosady, Int. J. Appl. Sci. Eng. 1 (2003) 55. K. Watanabe, T. Matsuda, R.Y.O. Nakamura, J. Food Sci. 50 (1985) 507. A. Tansakul, P. Chaisawang, J. Food Eng. 73 (2006) 276. P.-C. Chiu, C.-C. Chan, H.-M. Lin, H.-C. Chiu, J. Cosmet. Dermatol. 6 (2007) 243. E. Engel, J.-B. Lombardot, A. Garem, N. Leconte, C. Septier, J.-L. Le Que´re´, C. Salles, Int. Dairy J. 12 (2002) 609. L. Sua´rez, M.A. Dı´ez, R. Garcı´a, F.A. Riera, J. Ind. Eng. Chem. 18 (2012) 1859. C. Kao, Int. J. Environ. Sci. Technol. 11 (2014) 149. G. Masmoudi, R. Trabelsi, E. Ellouze, R. Amar, Int. J. Environ. Sci. Technol. 11 (2014) 1007. A.W. Mohammad, P.T. Yap, T.Y. Wu, Desalin. Water Treat. 10 (2009) 332. T.Y. Wu, A.W. Mohammad, J. Md Jahim, N. Anuar, J. Chem. Technol. Biotechnol. 84 (2009) 1390. H.G.R. Rao, A.S. Grandison, M.J. Lewis, J. Sci. Food Agric. 66 (1994) 563. W. Youravong, M.J. Lewis, A.S. Grandison, Food Bioprod. Process. 81 (2003) 303. A. Martinez-Ferez, S. Rudloff, A. Guadix, C.A. Henkel, G. Pohlentz, J.J. Boza, E.M. Guadix, C. Kunz, Int. Dairy J. 16 (2006) 173. E. Sua´rez, A. Lobo, S. A´lvarez, F.A. Riera, R. A´lvarez, Desalination 198 (2006) 274. B. Balannec, G. Ge´san-Guiziou, B. Chaufer, M. Rabiller-Baudry, G. Daufin, Desalination 147 (2002) 89. A. Cassano, L. Donato, E. Drioli, J. Food Eng. 79 (2007) 613. A. Cassano, L. Donato, C. Conidi, E. Drioli, Innov. Food Sci. Emerg. Technol. 9 (2008) 556. B. Koroknai, Z. Csanadi, L. Gubicza, K. Belafi-Bak, Desalination 228 (2008) 295. N. Rastogi, K. Raghavarao, in: Production of Dehydrated Coconut Milk, 1994. R. Hagenmaier, Coconut Aqueous Processing, Publications, University of San Carlos, Cebu City,Philippines, 1980. L.Y. Ng, C.P. Leo, A.W. Mohammad, J. Appl. Polym. Sci. 121 (2011) 1804. K.M. Persson, V. Gekas, G. Tra¨ga˚rdh, J. Membr. Sci. 100 (1995) 155. A. Naik, S.N. Raghavendra, K.S.M.S. Raghavarao, Appl. Biochem. Biotechnol. 167 (2012) 1290. K. Kwon, K.H. Park, K.C. Rhee, J. Agric. Food Chem. 44 (1996) 1741. AOAC, Official Method of Analysis, 15 edn., The Association of Official Agricultural Chemists, Virginia, 1990. S.-H. Lin, C.-L. Hung, R.-S. Juang, Chem. Eng. J. 145 (2008) 211. A. Grenier, M. Meireles, P. Aimar, P. Carvin, Chem. Eng. Res. Des. 86 (2008) 1281. T. Mansor, Y. Che Man, M. Shuhaimi, M. Abdul Afiq, F. Ku Nurul, Int. Food Res. J. 19 (2012) 837. T.Y. Wu, A.W. Mohammad, J. Md Jahim, N. Anuar, Biochem. Eng. J. 35 (2007) 309. J.P.F. De Bruijn, F.N. Salazar, R. Borquez, Food Bioprod. Process. 83 (2005) 211. G. Daufin, J.P. Escudier, H. Carrere, S. Berot, L. Fillaudeau, M. Decloux, Food Bioprod. Process. 79 (2001) 89. J. Leberknight, B. Wielenga, A. Lee-Jewett, T.J. Menkhaus, J. Membr. Sci. 366 (2011) 405. K.S. Kwon, D. Bae, K.H. Park, K.C. Rhee, J. Food Sci. 61 (1996) 753.
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
JIEC-2269; No. of Pages 6 Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Sequential fractionation of value-added coconut products using membrane processes Ching Yin Ng a, Abdul Wahab Mohammad a,b,*, Law Yong Ng a, Jamaliah Md Jahim a,b a
Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia b Research Centre for Sustainable Process Technology (CESPRO), Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia
A R T I C L E I N F O
Article history: Received 3 March 2014 Received in revised form 7 October 2014 Accepted 13 October 2014 Available online xxx Keywords: Fractionation Ultrafiltration Nanofiltration Skimmed coconut milk Value-added products
A B S T R A C T
The coconut waste-skimmed coconut milk was employed for sequential fractionation using UF and NF membranes to produce value-added products (coconut proteins, plant hormones – kinetin and zeatin). The retention factors achieved by UF membrane (PS10): albumin (0.9822 0.0799) and globulin (0.9975 0.0783); NF membrane (NF1): kinetin (0.9238 0.0345) and zeatin (0.9511 0.0355). Coconut protein powder was obtained after spray-drying process using concentrated coconut protein (UF retentate). SDS-PAGE showed that molecular weights of the coconut proteins were 17, 34, 55 and 150 kDa. Proximate and HPLC analyses revealed that the obtained samples were enriched with basic nutrients and well-balanced amino acids composition, respectively. ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Coconut (Cocos nucifera L.) is an important palm species around the region of Asia Pacific and South East Asia. It is commonly used as the cooking ingredient in these countries. Besides, coconut is also famous with its versatile applications [1]. One of the renowned coconut products is virgin coconut oil (VCO). VCO has been recognized as the healthiest crop oil and can be extensively employed in various fields such as food, beverage, medicinal, pharmaceutical, nutraceutical, cosmetic, etc. [2]. The high-valued fatty acids and anti-viral properties of the VCO have increased the demand of VCO worldwide and thus its production [3]. Large amount of valuable by-products have been discarded during the production of VCO. This has been considered as a wastefulness which also creates environmental issue in some areas near to the coconut processing plant. These discarded by-products still contain some high-valued nutritional components such as coconut protein, plant hormones, vitamins, minerals, amino acids and so
* Corresponding author at: Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia. Tel.: +603 89216102; fax: +603 89252546. E-mail addresses: [email protected] (C.Y. Ng), [email protected], [email protected] (A.W. Mohammad), [email protected] (L.Y. Ng), [email protected] (J.M. Jahim).
forth, which are highly beneficial to the consumers [4]. These valuable components can be potentially extracted by suitable separation techniques or other appropriate processes to fully utilize the raw materials employed during the VCO production. Full utilization of these by-products would definitely increase the profit of the coconut industry. The skimmed coconut milk is the main by-product after the coconut oil extraction process. The skimmed coconut milk consists of approximately 70% of the total proteins, carbohydrates, sugars, vitamins and minerals [5]. Through fractionation and concentration processes, the by-products from skimmed coconut milk can be employed in functional food, food supplement and food formulation. In addition to the compounds mentioned above, there are some other smaller-size minor components exist in the skimmed coconut milk. For example, there is a group of plant hormone called cytokinins which consists of kinetin, zeatin and traces of other compounds [6,7]. The kinetin and zeatin have been recognized as valuable compounds which possess anti-viral, anti-bacteria and anti-aging properties. They can be employed to stimulate the cell division of human body [1,8]. If the kinetin and zeatin from skimmed coconut milk can be separated and concentrated using appropriate processes, they will have huge potential to be applied in many applications such as medicinal, pharmaceutical and nutraceutical sectors [9,10]. The great versatility of these valuable compounds from the
http://dx.doi.org/10.1016/j.jiec.2014.10.028 1226-086X/ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 2
C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
coconut has attracted the attentions of many researchers to produce more profitable value-added coconut products. Many attempts have been tested to separate the valuable compounds in agricultural or food products such as distillation, evaporation, chemical extraction and so forth [11,12]. However, the conventional separation methods which involve heat treatment will deteriorate some thermal sensitive compounds like proteins [13,14]. Membrane filtration is an environmental friendly process compared to conventional methods [15]. The advantages of membrane filtration technique including low labor cost, low maintenance cost, less space consuming, energy saving, feasible to scale up, no chemicals used, no heat treatment, able to prolong the shelf-life of product, can retain the natural flavor and aroma of raw materials [16,17]. Due to the numerous advantages possessed by membrane filtration, it is used widely from resource recovery to wastewater treatment [18–21]. Therefore, the membrane separation technique is proposed in this study as one of the most probable ways to fractionate the skimmed coconut milk to produce those value-added products. In food technology field, the fractionation method using membrane technology has been well developed since past decades. For example, the fractions of fat and volatile compounds from goats’ milk cheese has been successfully obtained using the fractionation method by sequential ultrafiltration and nanofiltration processes [16]. Compounds with varying molecular sizes managed to be retained by different membrane categories such as microfiltration, ultrafiltration and nanofiltration. In the fractionation of dairy products, ultrafiltration [22,23] and nanofiltration [24–26] are always employed in order to obtain the valuable milk components. The protein hydrolysates can be efficiently separated by nanofiltration process due to the size exclusion mechanism and Donnan effect. In beverage industry, the membrane filtration technique is a common method to clarify and concentrate the fruit juice. Combination of various membrane types have been employed to enhance the quality of the produced juice [27–29]. However, there is not much study about the fractionation of coconut milk using membrane technology to obtain the valuable compounds. If the membrane processes can be employed to fractionate and concentrate the valuable nutrients from coconut waste effectively, this will contribute to a great breakthrough in the coconut industries. Besides that, implementation of spray-drying process could further enhance the quality of value-added products produced (such as coconut protein). Spray-drying is the most widely used commercial dehydration method owing to its short duration of heat-contact and high rate of evaporation, which can produce high quality products. Powders with precise specifications can be generated using spray-drying process in continuous operation [30]. The objective of present study is to determine the capability of sequential ultrafiltration and nanofiltration processes to fractionate the value-added products from skimmed coconut milk. In addition, the membrane performance will be evaluated and discussed in terms of normalized flux decline and solute retentions for protein, kinetin and zeatin compounds. Materials and methods Materials Coconut of Malayan Tall variety has been selected for this research work. To determine the concentration of albumin and globulins proteins in UF streams, standard bovine serum albumin (BSA), standard immunoglobulin (IgG) and protein assay dye reagent concentrate were supplied by Bio-Rad (USA). During the UF process, polysulfone (PS) membrane with 10 kDa (supplied by Koch, USA) was employed. The employed NF membrane was
manufactured by Amfor Inc., China. High performance liquid chromatography (HPLC) analysis was used to determine the concentration of kinetin and zeatin in NF streams. The standards of kinetin and zeatin were supplied by Sigma-Aldrich (Steinheim, Germany). Kinetin and zeatin solutions with the concentration ranging from 10 to 500 mM were prepared by dissolving the kinetin and zeatin standards in HPLC-grade methanol. These varying concentrations of kinetin and zeatin standards were stored at or below temperature of 4 8C. The chemicals used for HPLC analysis including HPLC-grade methanol (Tedia, USA), formic acid (Tedia, USA) and triethylamine (TEA) (Merck, Germany). Buffer solutions employed in the HPLC analysis has been prepared using methanol and 0.1% formic acid with pH 3.2. Preparation of skimmed coconut milk The milky white color fresh coconut milk was produced from solid grated coconut endosperm using a coconut extraction machine. Then, the extracted fresh coconut milk was filtered through micro-size sieving cloth to remove the bigger-size particles. Simple pasteurization process was conducted prior to the membrane filtration process in order to prolong the shelf-life of extracted coconut milk. The fresh coconut milk was heated at 60 8C for 15 min. This pasteurization method was conducted in order to reduce the microbial loading up to 10% according to the Hagenmaier method [31]. After that, the pasteurized coconut milk was poured into 125 L capacity cream separator machine (Elecrem, France) to separate the coconut cream/fat from the fresh coconut milk. There are two outputs from the cream separator: concentrated coconut cream and skimmed coconut milk. The produced skimmed coconut milk needs to be stored at a cold condition (0–4 8C) before it is used as a feed solution for the membrane filtration process. Membrane filtration processes (UF and NF membranes) were then employed to fractionate the skimmed coconut milk to produce the desired value-added coconut products. Membrane filtration process The polysulfone (PS) membrane was employed in the UF process. Polysulfone membrane was employed owing to its several superior properties such as high material toughness, good stability at high temperature, high resistant to various solvent, good resistant to wide pH range (pH 2–13) and low protein binding tendency [32]. The molecular weight cut-off (MWCO) value of this PS membrane is 10 kDa. The abbreviation of the UF membrane used in this experiment is PS10. NF1 membrane was employed in the NF filtration process. The membrane material of NF 1 is polyamide. Polyamide provides the desired properties such as high rejection of undesired materials (like salts), good mechanical strength and high filtration rate at low pressure. Prior to the filtration process, the membranes were soaked overnight to remove any preservative layer and dirt particle. Membrane compaction step was carried out at a higher pressure value (greater than operating pressure) for each employed membrane in order to enhance the permeate flux and membrane permeability [33]. Ultra-pure water was used during the membrane permeability test. In this study, a cross-flow system was used in the UF process. The cross-flow system was equipped with a 4 L stainless steel jacketed feed tank (embedded with a mixer) and a variable feed pump (Hydra-cell Pump, Mn, USA). The temperature of feed solution was controlled by the circulating water within the jacketed feed tank. The cross-flow velocity (CFV) of feed and the trans-membrane pressure (TMP) was controlled using a feed flowmeter (F-400, Blue-White, USA) and a permeate needle valve
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
(Swagelok, UK), respectively. A digital balance (GF6100, A&D, Japan) was connected to a computer to record the permeate flux in the continuous mode. The active surface area of PS10 membrane in cross-flow filtration system is 32.06 cm2. The operating pressure of 2 bar and temperature of 60 8C were used during the UF process of skimmed coconut milk. The CFV of 2.726 L/min and a stirring rate of 42 rad/s were maintained throughout the whole process of UF using PS10 membrane. In this study, a volume reduction factor (VRF) of 2 was set in each experimental run under batch concentration mode (the retentate was recycled back to the feed tank). The mass of collected permeate was recorded every 60 s with the help of the digital balance for permeate flux calculation. The collected UF permeate was stored for further fractionation process using NF1 membrane. Dead-end filtration module (Sterlitech HP4750 stirred cell, WA, USA) with a membrane active surface area of 15.20 cm2 was used to perform the NF process. Constant operating pressure of 5 bar and temperature of 25 8C were set in the NF process. A compressed nitrogen cylinder was connected to the dead-end filtration cell to supply the desired pressure. Duration employed for the NF process was 360 min. The mass of NF permeate was recorded by a connected digital balance for every 60 s throughout the filtration process for permeate flux calculation. Evaluation of membrane performance The performances of the employed UF and NF membranes were evaluated in terms of solution fluxes (expressed by the unit L/m2 h) and membrane retention capabilities towards particular solutes (albumin protein, globulin protein, kinetin and zeatin). In order to determine these parameters, Eq. (1) has been employed. Solution flux; Jw ¼
Q ADT
(1)
where Q is determined by the quantity of collected permeate in unit liter (L); A is the active surface area of used membrane (m2) and DT is the time interval of sampling (h). The normalized flux of produced permeate was calculated according to the following Eq. (2): Normalized flux ¼
Jw Jo
(2)
where Jw (L/m2 h) is the actual permeate flux at steady state and Jo (L/m2 h) is the initial flux of pure water. To determine the retention capabilities of PS10 and NF1 membranes towards coconut proteins (albumin and globulin) and cytokinins components (kinetin and zeatin), respectively, the following Eq. (3) has been used. Retention factor; Ret:F ¼ 1
Q pf QR
(3)
where Qpf (mg/L) is the concentration of solute in permeate and QR (mg/L) is the solute concentration in retentate stream. To calculate the volume reduction factor of the feed solution in UF process, the following Eq. (4) was employed. Volume reduction factor; VRF ¼
Qo Q o Q p Q p0
(4)
Where Qo (L) is the initial feed solution quantity; Qp (L) is the permeate quantity and Qp0 (L) is the permeate quantity loss during the filtration. All of these performance indicators are important in order to understand the overall methodology of filtration processes and the results will be discussed.
3
Spray-drying process After the UF process using PS10 membrane, the produced coconut protein solution was sent for spray-drying process in order to produce coconut protein powder. The Lab Plant brand spraydyer (Filey, UK) was used to perform the spray-drying process. Prior to the spray-drying process, the concentrated coconut protein solution was added with 5% of malto-dextrin and 20 Brix has been measured by a hand-held refractometer (Erma, Japan). The homogenized and concentrated coconut protein solution was fed into the spray-dryer by a peristaltic pump at a constant flowrate of 1.75 ml/min. The solution was then atomized through a 2 mm diameter nozzle into small droplets at 7 bar air pressure within a co-current air-flow system. The temperature of feed solution was set at 25 1 8C. The inlet and outlet temperatures of spray-dryer were 170 1 8C and 85 1 8C, respectively. The spraydried coconut protein powders were stored for further analyses. Analysis In this work, the targeted products throughout the UF and NF processes are: albumin proteins (measured as bovine serum albumin (BSA)), globulin proteins (measured as immunoglobulin (IgG)) [34,35], kinetin and zeatin. The dominant proteins in coconut milk are albumin and globulin proteins; whereas the kinetin and zeatin components are recognized as plant hormone or cytokinins. The Bradford method was applied in order to determine the concentration of albumin and globulin proteins in the feed, retentate and permeate streams after the UF process. BSA and IgG with varying concentrations were prepared and analyzed by a UVspectrophotometer (Genesys, Moston, MA) prior to real sample analysis. The purpose of this step was to obtain the standard plot of BSA and IgG. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis was also performed in order to determine the molecular weights of proteins which were presence in the UF streams (feed, retentate and permeate). Besides, the samples from UF streams (feed, retentate and permeate) and spray-dried coconut protein powders were sent for proximate analysis to determine the presence of basic nutrients (including protein, fat, carbohydrate, moisture and ash). Association of Official Analytical Chemists (AOAC) method was used throughout the proximate analysis [36]. In addition, the amino acids profile analysis using high performance liquid chromatography (HPLC) instrument was carried out in order to determine the amino acids compositions within the concentrated coconut protein solution. An AccQ Tag Column with a dimension of 3.9 150 mm was used during the HPLC analysis. The employed HPLC instrument is equipped with a Waters 410 Scanning Fluorescence. The unit used to report the amino acids composition is in mass (g) of amino acid per 100 g of protein, (g/100 g). The kinetin and zeatin were the targeted products after the NF process. To determine the concentrations of kinetin and zeatin in the NF streams (feed, retentate and permeate), a high performance liquid chromatography (HPLC) system (Agilent Technologies 1200 Series, Santa Clara, USA) was used. To detect and separate the kinetin and zeatin from other components within the sample, a UV detector was employed. During the analysis, the data was recorded and processed by the accompanying system software (ChemStation for LC 3D System). The samples were preliminarily filtered by a 0.45 mm Whatman glass microfiber filter prior to HPLC analysis. To initiate the analysis, 10 ml of filtered sample was injected into a C18 reverse phase column (Zorbax SB-C18 100A´˚ , 150 mm in length, 2.1 mm in diameter, Agilent Technologies, Santa Clara, USA). Initially, the HPLC system running condition was: methanol (0.1%)–formic acid buffer (10:90, v/v). The column thermostat was set at 25 8C. A flowrate of 0.3 ml/min was set
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
4
throughout the whole separation process. To achieve high detection accuracy of kinetin and zeatin components, the wavelength was fixed at 265 nm. Results and discussion Normalized flux declines during the ultrafiltration and nanofiltration processes PS10 membrane was employed in the cross-flow system to fractionate the skimmed coconut milk to produce concentrated coconut protein solution during UF process. For the NF process, NF1 membrane was employed to further fractionate the UF permeate in order to obtain kinetin and zeatin compounds. To achieve the VRF of 2 in the UF process, the filtration period spent by the cross-flow system was 360 min. Time taken by the NF1 membrane to complete the filtration was also roughly 360 min. To evaluate the performances of PS10 and NF1 membranes, the following normalized flux decline curves (Fig. 1) were plotted. Prior to the plotting of normalized flux curves, simple calculations of solution fluxes and normalized fluxes were made according to the Eqs. (1) and (2). According to Fig. 1, the UF process showed higher normalized fluxes compared to the NF process. However, at the initial stage of filtration processes, the normalized fluxes of UF process dropped more drastically than the NF process, which were followed by gradual decline along the filtration time. All of these phenomena can be clearly seen from Fig. 1. The main factor contributed to the sharp decline of normalized fluxes at the initial stage is dominated by concentration polarization phenomenon [22,37–39]. The steady states of normalized fluxes were achieved for both UF and NF processes as the durations of filtration were extended. Generally, feed sample contains considerable amount of proteins will eventually lead to membrane fouling [40]. During the filtration processes, the foulants may build up thus forming the fouling cake layers on top of the membrane surfaces. This may in turn affect the selectivity of the membrane towards certain desired solutes [38,41]. Retention factors of albumin and globulin proteins by PS10 membrane During the UF process, PS10 membrane was used to concentrate the proteins of skimmed coconut milk using cross-flow system. Predominant types of coconut proteins are albumin and globulin types. Thus, the concentrations of albumin and globulin proteins in the UF streams (permeate and retentate) were determined
Fig. 2. Retention factors of albumin, globulin proteins (by PS 10 membrane), kinetin and zeatin (by NF 1 membrane).
according to the Bradford method. The obtained results were substituted into Eq. (3) to calculate the retention factors of albumin and globulin proteins. Fig. 2 shows the retention factors chart of albumin and globulin proteins for UF process. As it can be seen from Fig. 2, both retention factors (albumin and globulin) achieved very high reading. The retention factor results for albumin and globulin were 0.9822 0.0799 and 0.9975 0.0783, respectively. It can be concluded that the albumin and globulin proteins in the skimmed coconut milk had been successfully retained and concentrated by PS10 membrane during the UF process. The results obtained were comparable to several other works in which the retention factors for albumin and globulin were 0.83–0.97 and 0.96–0.98, respectively [42,43]. Retention factors of kinetin and zeatin by NF1 membrane After the fractionation process of skimmed coconut milk using UF membrane (PS10), the collected UF permeate still contained numerous high nutritional components. The UF permeate was further fractionated by NF1 membrane to obtain additional valueadded products. One of the high-valued components is cytokinins. It consists of kinetin and zeatin components. These components have been identified to possess anti-aging properties and have high potential to be used in pharmaceutical, medicinal and cosmetic products. Thus, kinetin and zeatin were set as target products for the NF process. After the NF process, the concentrations of kinetin and zeatin in the NF streams (feed, retentate and permeate) were determined by HPLC analysis. The obtained results were calculated using Eq. (3). Fig. 2 also shows the retention factors of kinetin and zeatin in the NF process using NF1 membrane. Both components (kinetin and zeatin) had been successfully retained by NF1 membrane with their retention factors greater than 0.90. The retention factors achieved by NF process for kinetin and zeatin were 0.9238 0.0345 and 0.9511 0.0355, respectively. Thus, the results show that the kinetin and zeatin can be effectively retained by employing NF1 membrane during the NF process. SDS-PAGE of coconut proteins
Fig. 1. Normalized flux decline plots of UF and NF processes by PS10 and NF1 membranes, respectively.
The gel electrophoresis method using SDS-PAGE to determine the molecular weights of proteins is presented in Fig. 3. The molecular markers with molecular weights ranging from 10 kDa to 160 kDa were used in this analysis. The molecular markers can be clearly seen at the left lane in Fig. 3. UF feed, retentate and permeate samples were employed for the SDS-PAGE analysis for protein molecular weight identification. By analyzing the SDSPAGE profile (Fig. 3), it can be seen that no protein bands were detected in the UF permeate stream. Thus, it can be postulated that no protein with molecular weight greater than 10 kDa had
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
5
Table 2 Compositions of amino acids in the concentrated coconut protein.
Fig. 3. Molecular weights of coconut proteins by SDS-PAGE analysis.
successfully passed through the PS10 membrane during the UF process. It is obvious to see that the protein bands appeared in the UF feed were almost identical to the UF retentate. The intensity of protein bands in UF retentate is higher than UF feed. Again, this has proved that the coconut protein in the skimmed coconut milk had been successfully concentrated by PS10 membrane during the UF process. The major bands appeared in the SDS-PAGE profiles represented the dominant protein fractions in the UF streams. These dominant protein fractions possessed molecular weights of 17, 34, 55 and 150 kDa. The identified protein fractions are categorized as albumin and globulin types, which are accounting for 60% and 30% of total protein, respectively [11,44]. In the previous study, the total coconut protein was found to have a relatively similar pattern as shown in Fig. 3 [35]. Proximate analysis and amino acids profile of concentrated coconut proteins Table 1 shows the results of proximate analysis for the UF streams (feed, retentate and permeate) and spray-dried coconut protein powders. The compositions of some basic nutrients (protein, fat, carbohydrate, ash, moisture and energy) were reported in the unit of percentage or g/100 g of sample. According to the results obtained, each of the UF feed, retentate and permeate is mostly composed of moisture or water content. The moisture content for each of the UF stream is approximately 90%. After Table 1 Proximate composition of UF streams (feed, retentate and permeate) and coconut protein powder. Parameter (g/100 g)
UF feed
UF retentate
UF permeate
Coconut protein powder
Protein Fat Carbohydrate Ash Moisture Energy (kcal/100 g)
2.10 0.80 4.50 1.00 91.60 34.00
6.50 0.30 3.50 0.80 88.90 19.00
0.20 0.00 2.90 0.80 96.10 12.00
32.85 0.30 58.35 5.10 3.40 368.00
Amino acids
Composition (g/100 g protein)
Aspartic acid Serine Glutamic acid Glycine Arginine Alanine Valine Lysine Isoleucine Leucine Phenyl alanine Histidine Threonine Proline Tyrosine Methionine
8.95 2.04 21.50 3.55 14.20 2.19 2.99 3.74 2.07 3.17 3.86 1.80 2.50 3.60 1.80 1.80
the spray-drying process of concentrated coconut protein solution, the water content within the coconut protein powder left only 3.40%. The protein content in the UF retentate was 6.50%. The protein content in the UF retentate was higher than the UF feed (2.10%) and UF permeate (0.20%). This shows that the protein in the feed solution (which is skimmed coconut milk) has been successfully concentrated by the UF membrane (PS10). It has been identified that there are approximately 32.85% of protein in the coconut protein powder. UF permeate showed zero fat content, while UF feed consisted of 0.80%, UF retentate consisted of 0.30% and coconut protein powder consisted of 0.30%. The proximate analysis results proved that the spray-dried coconut protein powders produced in this work possessed high content of protein (32.85%) and carbohydrate (58.35%). Table 2 displays the compositions of amino acids in the concentrated coconut protein. The concentrated coconut protein solution showed a well-balanced essential amino acids profile. Compositions of glutamic acid, arginine and aspartic acids were found to be higher in comparison to other fractions of amino acids. However, the concentrated coconut protein is deficient in the sulfur amino acid—methionine. Comparable findings were reported by other researchers [5,11]. In brief, the produced concentrated coconut protein and coconut protein powder are highly suitable to be employed for human consumption and used as functional food in the related industries. The economical and nutritional values of the coconut by-products will be worth for future development especially in the food industries. Conclusion In the present work, UF and NF processes have been successfully conducted using cross-flow system and dead-end module, respectively. According to the normalized flux decline curves performed by the UF (PS10) and NF (NF1) membranes, the normalized fluxes displayed typical declining trend over the filtration time. It has been verified that these membranes have been successfully employed to concentrate the coconut proteins (albumin and globulin) and separate the valuable components (kinetin and zeatin) from the skimmed coconut milk. Most of the albumin and globulin proteins in the skimmed coconut milk can be retained by the employed UF membrane (PS10). The filtration process using NF membrane (NF1) had been successfully employed to further fractionate the valuable compounds (kinetin and zeatin) within the UF permeate. After the concentration process using UF (PS10) membrane in the cross-flow system, the achieved retention factors of albumin and globulin proteins were 0.9822 0.0799 and 0.9975 0.0783, respectively. For
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028
G Model
JIEC-2269; No. of Pages 6 6
C.Y. Ng et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx
the NF process, the retention factors obtained for both kinetin and zeatin were greater than 0.90. The retention factor achieved by kinetin and zeatin were 0.9238 0.0345 and 0.9511 0.0355, respectively. The success in fractionating the kinetin and zeatin by employing membrane filtration processes in this study will encourage the future employment of these compounds in the medicinal, pharmaceutical and nutraceutical fields due to their own anti-viral and anti-aging properties. Through the solute retention test, it has been verified that the albumin and globulin proteins in the skimmed coconut milk can be successfully concentrated by the UF process. By analyzing the SDSPAGE profiles, four major protein bands were identified in the UF samples (feed and retentate) having molecular weights of 17, 34, 55and 150 kDa. Besides, SDS-PAGE profiles also show that the intensity of protein bands in the UF retentate is higher than UF feed and permeate. Thus, the SDS-PAGE profiles have concluded that the coconut protein was successfully concentrated in this study. In addition, the results of proximate analysis which also verified that the protein content in the UF retentate was higher than the UF feed and permeate. Coincidently, this observation is in good agreement with the results obtained through SDS-PAGE analysis. Relatively well-balanced amino acid compositions in the concentrated coconut protein were obtained through the HPLC test. To make the coconut protein easier for handling, spray-drying process had been conducted using concentrated coconut protein solution. Coconut protein in powder form has been successfully produced through this work. Proximate analysis shows the coconut protein powder consists of 32.85% protein, 58.35% carbohydrate, 0.30% fat, 5.10% ash and 3.40% moisture. As a conclusion, the sequential UF and NF processes in this work have successfully produced the targeted or desired value-added coconut products (coconut protein, kinetin and zeatin). Last but not least, full utilization of the discarded coconut byproducts by UF and NF processes to produce highly valuable coconut value added products will bring a great breakthrough to Coconut industries and food technology field. Acknowledgment
[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39]
The authors of this work wish to gratefully acknowledge the financial support of the Ministry of Science and Technology (ScienceFund) through Project no. 02-01-02-SF1021. The authors would also like to acknowledge Malaysia Ministry of Education for the MyBrain15 Scholarship provided.
[40] [41] [42]
References
[43]
[1] J.W.H. Yong, L. Ge, Y.F. Ng, S.N. Tan, J. Mol. 14 (2009) 5144.
[44]
A.M. Marina, Y.B. Che Man, I. Amin, Trends Food Sci. Technol. 20 (2009) 481. B.J. Villarino, L.M. Dy, M.C.C. Lizada, LWT – Food Sci. Technol. 40 (2007) 193. J.L. McCullough, G.D. Weinstein, Cosmet. Dermatol. 15 (2002) 29. C.C. Seow, C.N. Gwee, Int. J. Food Sci. Technol. 32 (1997) 189. L. Ge, J.W.H. Yong, N.K. Goh, L.S. Chia, S.N. Tan, E.S. Ong, J. Chromatogr. B 829 (2005) 26. N.C. Yin, A.W. Mohammad, N.L. Yong, Adv. Sci. Lett. 19 (2013) 3620. Z. Ma, L. Ge, A.S.Y. Lee, J.W.H. Yong, S.N. Tan, E.S. Ong, Anal. Chim. Acta 610 (2008) 274. T.M. Kajs, R. Hagenmaier, C. Vanderzant, K.F. Mattil, J. Food Sci. 41 (1976) 352. J. Voller, M. Zatloukal, R. Lenobel, K. Dolezal, T. Be´res, V. Krystof, L. Spı´chal, P. Niemann, P. Dzuba´k, M. Hajdu´ch, M. Strnad, Phytochemistry 71 (2010) 1350. S.J. A.S. Samson, R.N. Khaund, C.M. Cater, K.F. Mattil, J. Food Sci. 36 (1971) 725. B.-K. Chen, L.L. Diosady, Int. J. Appl. Sci. Eng. 1 (2003) 55. K. Watanabe, T. Matsuda, R.Y.O. Nakamura, J. Food Sci. 50 (1985) 507. A. Tansakul, P. Chaisawang, J. Food Eng. 73 (2006) 276. P.-C. Chiu, C.-C. Chan, H.-M. Lin, H.-C. Chiu, J. Cosmet. Dermatol. 6 (2007) 243. E. Engel, J.-B. Lombardot, A. Garem, N. Leconte, C. Septier, J.-L. Le Que´re´, C. Salles, Int. Dairy J. 12 (2002) 609. L. Sua´rez, M.A. Dı´ez, R. Garcı´a, F.A. Riera, J. Ind. Eng. Chem. 18 (2012) 1859. C. Kao, Int. J. Environ. Sci. Technol. 11 (2014) 149. G. Masmoudi, R. Trabelsi, E. Ellouze, R. Amar, Int. J. Environ. Sci. Technol. 11 (2014) 1007. A.W. Mohammad, P.T. Yap, T.Y. Wu, Desalin. Water Treat. 10 (2009) 332. T.Y. Wu, A.W. Mohammad, J. Md Jahim, N. Anuar, J. Chem. Technol. Biotechnol. 84 (2009) 1390. H.G.R. Rao, A.S. Grandison, M.J. Lewis, J. Sci. Food Agric. 66 (1994) 563. W. Youravong, M.J. Lewis, A.S. Grandison, Food Bioprod. Process. 81 (2003) 303. A. Martinez-Ferez, S. Rudloff, A. Guadix, C.A. Henkel, G. Pohlentz, J.J. Boza, E.M. Guadix, C. Kunz, Int. Dairy J. 16 (2006) 173. E. Sua´rez, A. Lobo, S. A´lvarez, F.A. Riera, R. A´lvarez, Desalination 198 (2006) 274. B. Balannec, G. Ge´san-Guiziou, B. Chaufer, M. Rabiller-Baudry, G. Daufin, Desalination 147 (2002) 89. A. Cassano, L. Donato, E. Drioli, J. Food Eng. 79 (2007) 613. A. Cassano, L. Donato, C. Conidi, E. Drioli, Innov. Food Sci. Emerg. Technol. 9 (2008) 556. B. Koroknai, Z. Csanadi, L. Gubicza, K. Belafi-Bak, Desalination 228 (2008) 295. N. Rastogi, K. Raghavarao, in: Production of Dehydrated Coconut Milk, 1994. R. Hagenmaier, Coconut Aqueous Processing, Publications, University of San Carlos, Cebu City,Philippines, 1980. L.Y. Ng, C.P. Leo, A.W. Mohammad, J. Appl. Polym. Sci. 121 (2011) 1804. K.M. Persson, V. Gekas, G. Tra¨ga˚rdh, J. Membr. Sci. 100 (1995) 155. A. Naik, S.N. Raghavendra, K.S.M.S. Raghavarao, Appl. Biochem. Biotechnol. 167 (2012) 1290. K. Kwon, K.H. Park, K.C. Rhee, J. Agric. Food Chem. 44 (1996) 1741. AOAC, Official Method of Analysis, 15 edn., The Association of Official Agricultural Chemists, Virginia, 1990. S.-H. Lin, C.-L. Hung, R.-S. Juang, Chem. Eng. J. 145 (2008) 211. A. Grenier, M. Meireles, P. Aimar, P. Carvin, Chem. Eng. Res. Des. 86 (2008) 1281. T. Mansor, Y. Che Man, M. Shuhaimi, M. Abdul Afiq, F. Ku Nurul, Int. Food Res. J. 19 (2012) 837. T.Y. Wu, A.W. Mohammad, J. Md Jahim, N. Anuar, Biochem. Eng. J. 35 (2007) 309. J.P.F. De Bruijn, F.N. Salazar, R. Borquez, Food Bioprod. Process. 83 (2005) 211. G. Daufin, J.P. Escudier, H. Carrere, S. Berot, L. Fillaudeau, M. Decloux, Food Bioprod. Process. 79 (2001) 89. J. Leberknight, B. Wielenga, A. Lee-Jewett, T.J. Menkhaus, J. Membr. Sci. 366 (2011) 405. K.S. Kwon, D. Bae, K.H. Park, K.C. Rhee, J. Food Sci. 61 (1996) 753.
Please cite this article in press as: C.Y. Ng, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.10.028