Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes

Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes

Accepted Manuscript Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes Carmela ...

876KB Sizes 13 Downloads 148 Views

Accepted Manuscript Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes

Carmela Conidi, Alfredo Cassano, Francesca Caiazzo, Enrico Drioli PII:

S0260-8774(16)30340-5

DOI:

10.1016/j.jfoodeng.2016.09.017

Reference:

JFOE 8662

To appear in:

Journal of Food Engineering

Received Date:

27 July 2016

Revised Date:

16 September 2016

Accepted Date:

19 September 2016

Please cite this article as: Carmela Conidi, Alfredo Cassano, Francesca Caiazzo, Enrico Drioli, Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes, Journal of Food Engineering (2016), doi: 10.1016/j.jfoodeng. 2016.09.017

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Research highlights 

Flat-sheet membranes were studied to purify phenolic compounds from clarified pomegranate juice Productivity and separation capability of selected membranes were investigated



Desal GK membrane displayed high productivity, low fouling index and a good separation

NU SC RI PT



efficiency

The retentate fraction displayed high antioxidant activity



Glucose and fructose were mainly recovered in permeate and diafiltrate streams

AC

CE

PT

ED

MA



ACCEPTED MANUSCRIPT

1 2 3

Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes

5

SC RI PT

4

Carmela Conidi, Alfredo Cassano*, Francesca Caiazzo, Enrico Drioli

6 7

Institute on Membrane Technology, National Research Council, ITM-CNR, via P. Bucci, 17/C

8

University of Calabria, I-87036 Rende, Cosenza, Italy

NU

9 10

MA

11 12 13

ED

14 15

18 19

CE

17

PT

16

----------------------------------------------------

21 22 23 24

* Corresponding Author. Tel.: +39 0984 492067; Fax: +39 0984 402103 E-mail address: [email protected]

AC

20

1

ACCEPTED MANUSCRIPT

Abstract

26

Pomegranate juice is well recognized for its nutritional and health benefits due to the presence of

27

phenolic compounds, including anthocyanins, ellagic acid, phytoestrogenic flavonoids and

28

tannins. Therefore, the demand for the production of functional foods containing bioactive

29

compounds isolated from the juice has remarkably increased in the last decade.

30

In this study ultrafiltration (UF) and nanofiltration (NF) flat-sheet membranes, with nominal

31

molecular weight cut-off (MWCO) ranging from 1,000 to 4,000 Da, were tested to purify

32

biologically active compounds from clarified pomegranate juice. The filtration process was

33

evaluated in a crossflow pilot unit equipped with a Sepa CFII Membrane Cell System featuring

34

an effective membrane area of 0.014 m2.

35

A first screening was made in order to evaluate the performance of selected membranes in terms

36

of productivity, fouling index and retention towards sugars, phenolic compounds and total

37

antioxidant activity. Among these membranes the Desal GK membrane, with a MWCO of 2,000

38

Da, displayed higher permeate fluxes, lower fouling index and a good separation efficiency of

39

sugars from phenolic compounds in comparison with the other tested membranes. Therefore

40

further experiments were addressed to evaluate the separation capability and the productivity of

41

this membrane at different transmembrane pressure (TMP) values. Concentration/diafiltration

42

experiments were also performed in order to obtain a retentate fraction enriched in phenolic

43

compounds and a permeate stream mainly containing glucose and fructose.

44

According to the proposed process the yields of polyphenols and anthocyanins in the retentate

45

stream were of the order of 84.8% and 90.7%, respectively. The diafiltration step allowed to

46

obtain a recovery efficiency in the permeate side for glucose and fructose up to 90% and 93%,

47

respectively.

AC

CE

PT

ED

MA

NU

SC RI PT

25

48 2

ACCEPTED MANUSCRIPT

49

Keywords: Pomegranate juice; ultrafiltration; nanofiltration; anthocyanins; membrane processing.

50

1. Introduction

52

Pomegranate is one of the first five cultivated foods in the world widely grown in many countries

53

including Iran, India, Turkey, Tunisia, Pakistan, China, USA and Spain. Popular in Eastern as

54

well as Western parts of the world, pomegranate thrives well in regions with semi-arid and sub-

55

tropical climatic conditions but is also naturally adapted in regions with cold winters and hot

56

summers (Ozgen et al., 2008). The total world production is estimated currently at 2 million

57

tons/year (Erkan, 2011). In recent years, the interest for pomegranate fruit and its derivatives has

58

increased remarkably as evidenced by hundreds of publications on their chemical composition,

59

potential uses and proven health-promoting effects (Gumienna et al., 2016; Jurenka, 2008;

60

Lansky and Newman, 2007). Several studies have focused on the ability of different components

61

of the fruit, including the juice, seed oil, peel, flower extracts or their derivatives to protect

62

against several diseases such as cancer (Dai and Mumper, 2010), type 2 diabetes (Banihani et al.,

63

2013), atherosclerosis (Al-Jarallah et al., 2013) and cardiovascular diseases (Aviram et al., 2008)

64

providing the scientific basis for some use of pomegranate in traditional medicine. In addition,

65

these products have been shown to posses antimicrobial, anti-hepatoxic and antiviral properties

66

(Faria and Calhau, 2011). These health benefits have been attributed to the high antioxidant

67

capacity that is strongly correlated with the high concentration and chemical composition of

68

phenolic anthocyanins and hydrolysable tannins such as punicalagins, punicalin, peduncalagin,

69

and ellagic acid (Vegara et al., 2013). Different studies have also shown that the antioxidant

70

activity of pomegranate juice is much higher than most other fruit juices and beverages (Gil et al.,

71

2000; Seeram et al., 2008).

AC

CE

PT

ED

MA

NU

SC RI PT

51

3

ACCEPTED MANUSCRIPT

Given the wide spectrum of health promoting activities exerted by pomegranate and the

73

enormous interest that bioactive compounds isolated from this fruit have raised in the scientific

74

community, the interest of researchers has been addressed in recent years to the optimization of

75

the extraction and purification procedures of these compounds for the development of functional

76

foods meeting the consumer requirements.

77

Solvent organic extractions (SOEs) are the most commonly used procedures to extract bioactive

78

compounds from pomegranate fruits (Sood and Gupa, 2015). It is generally known that the yield

79

of chemical extraction depends on type of solvent (polarity), extraction time and temperature

80

(Singh et al., 2014). However, solvents commonly employed such as methanol, ethanol, acetone

81

and ethyl acetate are not always “food friendly” and not suitable or safe for their utilization in the

82

food industry (Amyrgialaki et al., 2015). In addition, long extraction times and high temperatures

83

may produce an oxidation of phenolics leading to a decreased yield of phenolics in the extracts. It

84

has been also shown that high temperatures (>70 °C) cause a rapid degradation of anthocyanins

85

(Havlíková and Miková, 1985). Alternative methods, such as microwave extraction (Zengh et al.,

86

2011), ultrasound-assisted extractions (Tabaraki et al., 2012) and supercritical fluid extractions

87

(He et al., 2012), have been also applied in the extraction of phenolic compounds from

88

pomegranate peels and seeds. However, low extraction efficiency, partial oxidation and

89

degradation of compounds of interest, high requirements of istrumentations and costs on

90

industrial scale are typical drawbacks which often outweigh the technical benefits. Therefore, it is

91

of critical importance to select efficient extraction procedures in order to maintain the stability of

92

phenolic compounds.

93

In this contest, membrane separation processes (MSPs) represent a valid alternative to traditional

94

technologies due to their low operating and maintenance costs, mild operating conditions of

95

temperature and pressure, easy control and scale-up and highly selective separations. They do not

AC

CE

PT

ED

MA

NU

SC RI PT

72

4

ACCEPTED MANUSCRIPT

require any extraction mass agents or chemical additives, avoiding product contaminations and

97

preserving the biological activity of the compounds of interest (Drioli and Romano, 2001).

98

Pressure-driven membrane operations, such as microfiltration (MF), ultrafiltration (UF),

99

nanofiltration (NF) and reverse osmosis (RO) are today well-established technologies in food and

100

beverage industries for the treatment of several products and by-products (Daufin et al., 2001;

101

Patsioura et al., 2011; Tylkowski and Tsibranska, 2015). Other membrane processes, such as

102

osmotic distillation (OD), membrane distillation (MD) and pervaporation (PV) have been also

103

investigated in recent years for selected applications in the same area. Moreover, the development

104

of hybrid processes based on the combination of different membrane unit operations and

105

conventional separation technologies offers new and much more opportunities in terms of

106

competitiveness, improvement of quality, process or product novelty and environmental

107

friendliness (Cassano, 2016; Conidi et al., 2014). The use of membrane technology in the

108

treatment of pomegranate juice has been recently investigated. In particular, MF and UF

109

processes have been studied to clarify pomegranate juice as alternative technologies to the

110

traditional use of fining agents (gelatin, bentonite, diatomaceous earth, silica sol, etc.) and other

111

techniques including centrifugation, decantation, depectinization and filtration (Baklouti et al.,

112

2012; Cassano et al., 2015; Mirsaeedghazi et al., 2010a; Mirsaeedghazi et al., 2010b); MD and

113

OD have been evaluated for their potential in the concentration of the juice as alternative to the

114

thermal evaporation (Cassano et al., 2011; Onsekizoglu, 2013).

115

In recent years UF and NF operations have gained a great interest for the separation and

116

concentration of bioactive compounds from plant extracts and by-products of agro-food

117

industries (Cassano et al., 2014; Cissé et al., 2011; Diaz-Reinoso et al., 2009; Galanakis et al.,

118

2013; Giacobbo et al., 2013; Li and Chase, 2010; Mello et al., 2010; Murakami et al., 2013;

119

Tsibranska and Tylkowski, 2013; Tylkowski et al., 2010). According to the so-called "5-Stages

AC

CE

PT

ED

MA

NU

SC RI PT

96

5

ACCEPTED MANUSCRIPT

Universal Recovery Processing" approach, the production of target compounds from food wastes

121

includes the following steps: (i) macroscopic pre-treatment, (ii) separation of macro- from micro-

122

molecules, (iii) extraction, (iv) purification and (v) product formation (Galanakis, 2012). UF and

123

NF are considered key physicochemical and non-destructive techniques applied in the second,

124

third and fourth step of the above downstream processing (Galanakis, 2015).

125

NF membranes have a nominal pore size in the range of 0.5-1 nm; the typical range of MWCO

126

levels is between 200 and 1,000 Dalton. UF involves the use of membranes with a MWCO in the

127

range of 1-300 kDa and a pore size of about 0.01 m (Baker, 2004). The separation capabilities

128

of UF and NF membranes are mainly related to size exclusion but interactions between solutes

129

and membrane like charge interactions, bridging and hydrophobic interactions may play an

130

important role in the formation of fouling layers at the membrane surface (or within the

131

membrane pores) which will exert some influence on passage of solutes through the membrane.

132

The formation of fouling layers due to macromolecules like proteins and dietary fibres has been

133

also reported in literature (Galanakis et al., 2014; Patsioura et al., 2011). Moreover, diafiltration

134

conditions can be employed in order to remove contaminants with low molecular weight (MW)

135

from valuable products with higher molecular weight in order to increase the product yield of the

136

process (Aspelund and Glatz, 2010; Teixeira et al., 2014).

137

No literature is readily available on the performance of UF and NF membranes for the separation

138

and purification of phenolic compounds from sugars in pomegranate juice. In the light of these

139

considerations, this work investigated the performance of flat-sheet UF and NF membranes, with

140

different membrane material and molecular weight cut-off (MWCO), for separating and

141

concentrating phenolic compounds from clarified pomegranate juice. The performance of the

142

selected membranes was compared in terms of permeate fluxes, retention towards sugars, total

AC

CE

PT

ED

MA

NU

SC RI PT

120

6

ACCEPTED MANUSCRIPT

antioxidant activity and biologically active compounds (mainly total polyphenols and

144

anthocyanins). To fulfil the final aim to purify the selected bioactive compounds from sugars, the

145

membrane process was also studied in a diafiltration mode.

146 147

2.

148

2.1. Pomegranate juice extraction and clarification

149

Pomegranates, of Calabria origin, were purchased from a local open market (Cosenza, Italy).

150

Fruits were washed in cold tap water and drained. They were manually cut-up into two halves

151

and then squeezed by using an electric juicer (Aristalco S.r.l., Treviso, Italy). The obtained juice,

152

having a deep-red color, was pre-filtered with a cotton fabric filter. The extracting procedure

153

gave an average juice yield of 40% (w/w).

154

The raw juice was clarified by using a laboratory unit supplied by Verind SpA (Milan, Italy)

155

equipped with a cellulose triacetate UF membrane module (FUC 1582, Microdyn Nadir,

156

Wiesbaden, Germany) in hollow fiber configuration with a nominal MWCO of 150 kDa ad a

157

membrane surface area of 0.26 m2. The juice filtration was conducted according to the batch

158

concentration mode (the retentate stream was flowed back to the feed tank while the permeate

159

stream was collected separately) up to a weight reduction factor (WRF) of 4.8. The WRF is

160

defined as the ratio between the initial feed weight and the weight of the resulting retentate

161

according to the following equation:

162

𝑊𝑅𝐹 = 𝑊 = 1 + 𝑊

163

where Wf, Wp and Wr are the weight of feed, permeate and retentate, respectively.

𝑊𝑝

AC

𝑊𝑓

CE

PT

ED

MA

NU

Material and methods

SC RI PT

143

𝑟

𝑟

(1)

7

ACCEPTED MANUSCRIPT

The UF system was operated at a transmembrane pressure (TMP) of 0.6 bar, an axial feed

165

flowrate (Qf) of 400 L/h and a temperature (T) of 25±2 °C. The produced clarified juice was

166

stored at -18 °C and defrosted before use

SC RI PT

164

167

2.2. Treatment of clarified juices with UF and NF membranes

169

2.2.1. Setup

170

Membrane filtration experiments were performed by using a lab crossflow membrane filtration

171

unit (Sepa CF II, GE Water and Process Technologies, Canada, USA) equipped with a stainless

172

steel rectangular cross-flow cell. This cell is designed to simulate flow dynamics of commercially

173

available spiral membrane elements, by using a combination of stainless steel shims, feed spacers

174

and permeate carriers. The dimensions of the cell are 14.6 cm, 9.5 cm and 0.86 mm for channel

175

length, width, and height, respectively. These channel dimensions provide an effective membrane

176

area of 140 cm2 and a cross-sectional flow area of 0.82 cm2. The test cell of the unit is rated for

177

operating pressures up to 69 bar.

178

The assembled cell body was inserted into a cell holder and compressed against the cell holder

179

top through a piston. The feed stream (clarified pomegranate juice) was pumped from a stainless

180

steel vessel with a capacity of 2 L to the feed inlet located on the cell body bottom through a high

181

pressure pump (SSE1507 - Interpump 63SS Series Pump). The permeate flowing to the center of

182

the cell body top was collected in a manifold and through a permeate outlet connection into a

183

permeate vessel. Two manometers placed before and after the cell were used to measure the inlet

184

and the outlet pressure and, consequently, the applied TMP. The feed flow rate (Qf) and the TMP

185

values were regulated by a pressure control valve on the retentate side. The temperature of the

186

juice was controlled by using a cooling coil placed into the feed vessel fed with tap water. The

AC

CE

PT

ED

MA

NU

168

8

ACCEPTED MANUSCRIPT

187

permeate flux was periodically monitored by acquisition of its weight, using an electronic

188

balance (with an accuracy of 0.1 g) placed under the permeate vessel.

SC RI PT

189

2.2.2. Membranes

191

Four commercial flat-sheet membranes from different manufacturers were studied in this work,

192

namely Etna 01PP from Alfa-Laval (Lund, Sweden), PES 004H from Mycrodin-Nadir

193

(Wiesbaden, Germany), SelRO MPF-36 from Koch Membrane Systems (Wilmington, USA) and

194

Desal GK from GE Water & Process Technologies (Trevose, USA). Their typical characteristics

195

according to the manufacturers’ data sheet are reported in Table 1.

NU

190

196

2.2.3. Batch concentration experiments

198

A first set of experiments was performed according to the batch concentration configuration in

199

which the permeate stream was collected separately while the retentate was recycled bach to the

200

feed reservoir up to a WRF of 1.5.

201

All experiments were performed at a TMP of 10 bar and an operating temperature of 25±1 °C.

202

The membrane performance was evaluated in terms of productivity (permeate flux), solute

203

rejection and fouling index.

204

The permeate flux (Jp) was determined by measuring the collected permeate weight in a given

205

time through the membrane surface area by using the following equation:

206

𝐽𝑝 = 𝐴 ∙ 𝑡

207

where Jp is the permeate flux (kg/m2h), Wp the permeate weight (kg) at time t (h) and A the

208

membrane surface area (m2).

209

The retention (R) of selected membranes towards specific compounds was determined as:

AC

𝑊𝑝

CE

PT

ED

MA

197

(2)

9

ACCEPTED MANUSCRIPT

(

Cp

)

(3)

R = 1 ‒ C ∙ 100

211

where Cp and Cf are the concentration of a specific component in the permeate and feed,

212

respectively.

213

The fouling index (FI) was calculated by comparing the pure water permeability before and after

214

the juice filtration according to the following equation:

215

𝐹𝐼 =

216

where WP1 is the pure water permeability after the pomegranate juice filtration and WP0 the

217

water permeability of the virgin membrane.

218

The water permeability of each membrane was determined by the slope of the straight line

219

obtained by plotting the water flux values, measured in fixed conditions of temperature (25±1

220

°C), versus the applied TMP. After the treatment with the clarified juice, membranes were rinsed

221

with tap water for 30 min and their pure water permeability was measured; then, the fouled

222

membranes were submitted to a cleaning procedure by using a 0.125 M NaOH solution, at 40±1

223

°C for 60 min. At the end of the chemical cleaning procedure, the pure water permeability was

224

measured again.

225

The cleaning efficiency (CE) was evaluated according to the following equation:

226

𝐶𝐸 =

227

where WP3 is the water permeability after the chemical cleaning and WP0 is the water

228

permeability of the virgin membrane.

230

𝑊𝑃1

( ) ∙ 100

CE

𝑊𝑃0

PT

𝑊𝑃3

( ) ∙ 100

ED

MA

NU

(4)

𝑊𝑃0

(5)

AC

229

f

SC RI PT

210

2.2.4. Experiments with the membrane Desal GK

10

ACCEPTED MANUSCRIPT

After the first set of experiments, according to the obtained results, the Desal GK membrane was

232

selected to perform experimental runs in total recycle configuration (both permeate and retentate

233

were continuously recycled to the feed reservoir) in order to study the effect of TMP on the

234

permeate flux and selectivity towards the compounds of interest. The TMP value was varied in

235

the range 5-25 bar maintaining the operating temperature at 25±1 °C.

236

In order to improve the removal of glucose and fructose from the UF retentate constant volume

237

diafiltration experiments were performed in a discontinuos way. In particular, the clarified juice

238

was firstly concentrated in batch concentration mode in selected operating conditions (TMP, 15

239

bar; T, 25±1 °C) until a minimum retentate hold-up of 500 mL was reached in the filtration unit,

240

corresponding to a WRF of 5. Then the retentate was diluted with the same amount of purified

241

water and the permeate was removed separately. The filtration and dilution procedure was

242

repeated three times, thus approaching diafiltration conditions. The last filtration run was

243

operated until the minimum retentate hold-up was reached.

244

ED

MA

NU

SC RI PT

231

2.3. Analytical evaluations

246

Permeate and retentate samples coming from different experiments were immediately frozen and

247

kept at -18 °C until analysed. Samples were analysed for total phenols, total soluble solids (TSS),

248

anthocyanins, sugars and total antioxidant activity (TAA).

CE

249

PT

245

2.3.1. Total phenols

251

Total phenols were estimated colorimetrically by using the Folin-Ciocalteu method (Singleton et

252

al., 1999). The method is based on the reduction of tungstate and/or molybdate in the Folin-

253

Ciocalteu reagent by phenols in alkaline medium resulting in a blue colored product. A sample

254

aliquot (0.2 mL) was mixed with 1 mL of a 10 fold diluted Folin-Ciocalteu reagent and 0.8 mL of

AC

250

11

ACCEPTED MANUSCRIPT

7.5% sodium carbonate. Then the mixture was allowed to stand for 30 min at room

256

temperature. The absorbance was measured at 760 nm by using an UV-visible spectrophotometer

257

(Shimadzu UV-160A, Japan). Gallic acid solutions with concentrations ranging from 10 to 100

258

mg/L were used for calibration. A dose response linear regression was generated by using the

259

gallic acid standard absorbance and results were expressed as mg/L gallic acid equivalent (GAE)

260

All measurements were performed in triplicate in triplicate and the results were averaged.

SC RI PT

255

261

2.3.2. Anthocyanins

263

Anthocyanins were assessed by high-performance liquid chromatography (HPLC) using an

264

Agilent 1100 series system (Agilent Technologies, Waldbronn, Germany) equipped with a

265

vacuum degasser, a quaternary pump, an autosampler and an UV-Vis diode array detector.

266

Chromatographic separation was performed by using a Luna C18(2) column (250 mm x 4.60

267

mm, 5m, Phenomenex, Torrance, CA, USA) according to the following conditions: V = 1

268

mL/min; T = 25 °C;  = 518 nm. The mobile phase was a mixture of H2O/HCOOH (9:1) as

269

solvent A and H2O/HCOOH/CH3CN (4/1/5) as solvent B. Anthocyanins separation was achieved

270

by using the following linear gradient: starting condition, 88% A, 12% B; 26 min, 70% A, 30%

271

B; 35 min, 100% B; 43 min, 88% A, 12% B; 46 min 88% A, 12% B. Anthocyanins were

272

identified by matching the retention time and their spectral characteristics against those of

273

standards (cyanidin 3,5-diglucoside, delphinidin 3-glucoside, cyanidin 3-glucoside, pelargonidin

274

3-glucoside). Quantification was made according to the linear calibration curves of standard

275

compounds (Mondello et al., 2000).

AC

CE

PT

ED

MA

NU

262

276 277

2.3.3. Sugars

12

ACCEPTED MANUSCRIPT

The quantitative determination of glucose and sucrose was carried out by an HPLC system

279

(Thermo Scientific Accella 600, USA) equipped with a binary pump, an autosampler, a

280

thermostated column compartment and refractometer index detector. Separation was achieved

281

with a Luna NH2 100A column (250 mm x 4.60 mm, 5 Phenomenex, Torrance, CA, USA).

282

Samples were eluted in isocratic mode by using a mixture of acetonitrile/water (80:20). Operating

283

conditions were as follows: V = 1 mL/min, T = 40 °C, pressure = 85 bar (Ruiz-Rodríguez et al.,

284

2011).

285

Prior to HPLC analysis all samples were diluted with acetonitrile (9:1) and filtered by using 0.45

286

m nylon filters. A sample volume of 20 L was used. The peak areas in the chromatograms

287

were plotted against calibration curves obtained from standard solutions (external standard

288

method), in a concentration range of 0.5-2 mg/mL for each compound.

289

MA

NU

SC RI PT

278

2.3.4. Total antioxidant activity (TAA)

291

TAA

292

ethylbenzothiazoline-6-sulfonic acid) (ABTS) free radical decolourisation assay in which the

293

radical cation is generated by reaction with potassium persulfate before the addition of the

294

antioxidant (Re et al., 1999). This method gives a measure of the antioxidant activity of pure

295

substances and of mixtures by monitoring the reduction of the radical cation as the percentage

296

inhibition of absorbance at 734 nm. Spectrophotometric measurements were performed by using

297

an UV–Visible recording spectrophotometer (UV-160 A, Shimadzu Scientific Instruments, Inc.,

298

Japan) at 30 °C. ABTS was dissolved in water at 2 mM concentration: ABTS radical cation was

299

produced by reacting 10 mL of ABTS stock solution with 100 L of 70 mM potassium persulfate

300

solution (ABTS:K2S2O8 = 1:0.35 molar ratio) and allowing the mixture to stand in the dark at

determined

by

an

improved

version

of

the

of

the

2,2’-azino-bis-(3-

AC

CE

PT

was

ED

290

13

ACCEPTED MANUSCRIPT

room temperature for 12–16 h before use. The work solution was prepared diluting 1 mL of the

302

ABTS radical cation solution to 25 mL with PBS buffer (5 mM Na2HPO4, 5 mM NaH2PO4, NaCl

303

9 g/L, pH= 6.8) to a final UV absorbance of 0.70 ± 0.02 at 734 nm. After addition of 10 L of

304

sample to 10 mL of ABTS work solution, the absorbance at 734 nm was recorded every min for a

305

total of 6 min. The value at 5 min was used to calculate the results reported as TAA, expressed in

306

terms of mM Trolox equivalent. Each determination was performed in triplicate. Results were

307

expressed as mean ± SD of three samples.

SC RI PT

301

308

2.3.5.Total soluble solids

310

Total soluble solids, expressed as °Brix, were measured by using a hand refractometer (Atago

311

Co., Tokyo, Japan) with a scale range of 0-32 °Brix.

MA

NU

309

312

3. Results and discussions

314

3.1. Pomegranate juice clarification and juice composition

315

Figure 1 shows the dependence of permeate flux and WRF on time observed in the clarification

316

of the raw juice in the selected operating conditions. The initial permeate flux of about 37 kg/m2h

317

decreased gradually with operating times due to concentration polarization and fouling

318

phenomena to reach a steady-state value of about 7 kg/m2h. As reported in Table 2, the UF

319

treatment produced a complete removal of suspended solids with the production of a clear juice

320

with a red brilliant colour.

321

The raw juice was characterized by a total soluble solids (TSS) content of 17.03±0.04 °Brix

322

which is in agreement with data reported by other Authors (Dafny-Yalin et al., 2010; Ferrara et

323

al., 2011). This value appeared to be higher in comparison to the clarified juice: this phenomenon

AC

CE

PT

ED

313

14

ACCEPTED MANUSCRIPT

can be attributed to the presence of suspended solids in the raw juice that can interfere with the

325

measurement of the refractive index.

326

As previously reported, polyphenols from pomegranate juice have taken great attention in the last

327

years for their health-promoting properties (Viuda-Martos et al., 2010; Sreeja et al., 2014). The

328

content of polyphenols in the clarified juice (2457.50±15.30 mg/L) was well preserved in

329

comparison to the raw juice and resulted higher than that reported for typical cultivars grown in

330

southern Turkey, Greece and Chile (Ferrara et al., 2011). The predominant polyphenols detected

331

in both raw and clarified juices were anthocyanins; in the clarified juice cyanidin 3,5-O-

332

diglucoside resulted the most representative compound (136.10±5.30 mg/L), followed by

333

cyanidin 3-O-glucoside (53.71±2.06 mg/L), delphinidin-3-O-glucoside (14.61±0.48 mg/L) and

334

pelargolidin 3,5-O-diglucoside (4.77±0.43 mg/L).

335

The amount of phenolic compounds is directly correlated with the antioxidant activity of the

336

clarified juice (Mousavinejad et al., 2009) which, according to the ABTS method, was of

337

26.0±2.8 mM Trolox.

338

According to Tezcan et al. (2009) glucose and fructose were the only sugar types detected in the

339

juice. The fructose content (19.45±0.62 g/L) resulted higher in comparison with the glucose one

340

(12.53±0.48 g/L). The absence of sucrose in the fruit can be explained by the enzymatic

341

hydrolysis of this sugar into glucose and fructose during the ripening process (Zarei et al., 2011).

342

In the clarified juice the content of glucose and fructose resulted 2.8% and 8.3% lower than that

343

measured in the raw juice, respectively.

NU

MA

ED

PT

CE

AC

344

SC RI PT

324

345

3.2. Performance of selected membranes

346

The pure water permeability (WP0) values measured for each membrane are summarized in Table

347

3. According to the results the WP values decreased in the following order: 15

ACCEPTED MANUSCRIPT

PES 004H > Etna 01PP> MPF-36 > Desal GK.

349

These values are not correlated with the MWCO of the selected membranes: indeed, the PES004

350

H membrane, with the highest MWCO (4,000 Da) showed the highest permeability value

351

(WP0=47.81 kg/m2hbar); Etna 01PP and MPF-36 membranes with the same MWCO (1,000 Da),

352

but different membrane material (thin-film composite and fluoro polymer, respectively) showed

353

WP0 values of 22.16 kg/m2hbar and 14.54 kg/m2hbar, respectively. The lowest WP0 value (10.5

354

kg/m2hbar) was found for the Desal GK membrane with a MWCO of 2,000 Da. These results can

355

be attributed to the internal structure of the selected membranes which is strongly related to their

356

composition, morphology and hydrophilicity/hydrophobicity (Benítez et al., 2009).

357

A different behavior was observed in the treatment of the clarified pomegranate juice. In the

358

selected operating conditions the permeate flux decreased gradually up to reach a steady-state

359

value. The initial decrease of permeate flux is generally explained by the effect of concentration

360

polarization phenomena whereas the second phase of decrease is due to the accumulation of

361

molecules and particles on the membrane surface or inside the pores of the membrane as far as

362

the concentration of the feed solution increases (Conidi et al., 2015).

363

The Desal GK membrane presented the highest initial and steady-state permeate flux values (22.6

364

kg/m2h and 11.3 kg/m2h, respectively) when compared with the other tested membranes (Figure

365

2). On the other hand, the PES 004H membrane presented the lowest permeate flux values (2.4

366

kg/m2h and 0.7 kg/m2h, respectively). These results confirm that the permeate flux is affected not

367

only by TMP but also by membrane material and structure as well as by interactions between

368

solute and membranes.

369

The obtained results in term of permeate fluxes, are in agreement with experimental data reported

370

by other Authors. Cissé et al. (2007) tested different UF and NF membranes to separate and

371

concentrate anthocyanins compounds from Hibiscus sabdariffa L. extracts. They found that

AC

CE

PT

ED

MA

NU

SC RI PT

348

16

ACCEPTED MANUSCRIPT

Desal thin film composite membranes, with a MWCO in the range 500-1,000 Da, presented

373

higher permeate fluxes when compared with polyethersulphone membranes with higher MWCO

374

(in the range 5-50 kDa) working in the same operating conditions (operating pressure 20 bar).

375

Similary, the Desal GK membrane exhibited permeate fluxes values of about 40 kg/m2h at 30 bar

376

in the concentration of ellagitannins from blackberry juice which were higher than those

377

measured with a polyethersulphone membrane of 5000 Da (UP005, Microdyn Nadir) at the same

378

operating pressure (Acosta et al., 2014).

379

The tested membranes were also compared in terms of fouling index (FI) and cleaning efficiency

380

(CE) (Table 3). The lowest FI value was measured for the Desal GK membrane (51%), followed

381

by MPF-36 (77%) and PES 004H (88%) membranes; on the other hand the highest FI value was

382

detected for the Etna 01PP membrane (95%). The CE resulted higher than 90% for all

383

investigated membranes.

MA

NU

SC RI PT

372

384

3.3. Retention of UF and NF membranes towards analyzed compounds

386

The analyses of anthocyanin compounds in samples coming from batch concentration

387

experiments revealed that the rejection of the selected membranes towards these compounds were

388

higher than 80%, with the exception of the ETNA 01PP membrane (Figure 3a). The MPF-36

389

membrane exhibited the highest rejection towards the more representative anthocyanins of the

390

juice; the observed retention for the PES 004H membrane was in the range 60-99%, while the

391

Etna 01PP membrane showed a lower retention (in the range 60-80%). These results were

392

confirmed by analyses of phenolic compounds and TAA. In particular, the rejection of all

393

selected membranes towards phenolic compounds was

394

membrane exhibiting the highest retention values (Figure 3b). As expected, with the exception of

395

the Etna 01PP membrane, a strict correlation was observed between the rejection towards

AC

CE

PT

ED

385

higher than 80% with the MPF-36

17

ACCEPTED MANUSCRIPT

phenolic compounds and TAA, since phenolic compounds mainly contribute to the TAA of the

397

juice (Tezcan et al., 2009). In particular, the observed retention of Desal GK membrane towards

398

TAA and phenolic compounds was of 78% and 88%, respectively; similar results were obtained

399

with the PES 004H membrane (retention values were of 86% and 95%), while the retention of the

400

MPF-36 membrane was higher than 95% for both components. The lowest retention value

401

towards TAA was measured with the Etna 01PP membrane (retention of about 57%); adversely,

402

the retention towards polyphenols was of about 85%. As reported by Galanakis et al. (2015) the

403

high concentration of phenols in samples obtained from the filtration of Cypriot wine with the

404

Etna 01PP membrane did not reflect higher antioxidant capacity, probably due to the observed

405

antagonistic effect between different phenolic compounds and anthocyanins.

406

All selected membranes exhibited a low rejection towards sugar compounds. In particular, with

407

the exception of the MPF-36 membrane, the rejection towards soluble solids was of the order of

408

30%. For glucose and fructose the observed rejections were lower than 10%, with the Desal GK

409

exhibiting the lowest retention towards both compounds (in the range 1-4%) (Figure 3c).

ED

MA

NU

SC RI PT

396

410

3.4. Experiments with Desal GK membrane

412

On the basis of the preliminary results related to membrane productivity, membrane fouling and

413

membrane selectivity the Desal GK membrane was selected to perform other experiments both in

414

total recycle configuration (in order to evaluate the effect of TMP on the rejection of compounds

415

of interest and productivity) and in batch concentration configuration at higher WRF, in order to

416

increase the concentration of phenolic compounds in the retentate stream. Diafiltration was also

417

used to improve the removal of sugar compounds from phenolics.

418

Figure 4 shows the effect of TMP on the steady-state permeate flux and retention towards TSS,

419

TAA and total polyphenols for the Desal GK membrane. A linear increase of the permeate flux,

AC

CE

PT

411

18

ACCEPTED MANUSCRIPT

from 7 to 40 kg/m2h, was observed by increasing the operating pressure in the range of

421

investigated values (5-25 bar): the absence of a limiting flux can be attributed to the preliminary

422

treatment of the raw juice by UF which removed suspended solids that, as well known are the

423

main responsible compounds in the formation of fouling layers. On the other hand, the retention

424

of phenolic compounds was not affected by TMP: indeed, the retention of phenolic compounds

425

was higher than 90%, independently by the applied pressure. Accordingly, the TAA retention

426

showed a similar trend. Adversely, an increasing of TMP induced higher retention of TSS with a

427

significant retention (higher than 40%) at 25 bar, suggesting that this pressure should be avoided

428

in order to reach high separation factors between sugars and phenolic compounds. The increased

429

retention levels can be explained not only by steric considerations but also through the

430

interactions between solutes and membrane material and the association of the solutes with

431

retained compounds.

432

Data of steady-state permeate fluxes at different TMP were in agreement with those related to the

433

fouling index. In particular, in the range of investigated pressures, a little increase of the fouling

434

index (from 30 to 35%) was observed when TMP was increased at 25 bar. Data of water

435

permeability before and after cleaning procedures, confirmed a greater contribution of the fouling

436

resistance to the total resistance if compared to the cake layer resistance (Table 4).

437

The effect of TMP on the anthocyanins rejection for the Desal GK membrane is illustrated in

438

Figure 5. For all the investigated compounds the retention was higher than 90% independently by

439

the operating pressure. In addition, a strict correlation between the retention value and the

440

molecular weight of each compound was observed, with cyanidin-3,5-O-diglucoside, exhibiting

441

the highest rejection.

442

Experimental runs in selected operating conditions (15 bar and 26 °C) were performed according

443

to the batch concentration configuration up to WRF 5. In these conditions the initial permeate

AC

CE

PT

ED

MA

NU

SC RI PT

420

19

ACCEPTED MANUSCRIPT

flux of 30 kg/m2h was reduced up to 10 kg/m2h when the final WRF was reached. The analyses

445

of permeate and retentate streams revealed that, according to the high retention of the Desal GK

446

membrane towards phenolic compounds, phenolic compounds and anthocyanins were

447

concentrated in the retentate stream. In particular, the content of single anthocyanins, total

448

phenols and TAA in the retentate stream was in agreement with the reached WRF. On the other

449

hand, the content of soluble solids, glucose and fructose in the permeate stream was of the same

450

order of the clarified juice due to the low rejection of the membrane towards these compounds

451

(Table 5).

452

In Table 6 the mass balance of the process at WRF 5 for total anthocyanins, polyphenols and

453

sugars, is reported. This balance is referred to an experimental run in which starting from 3.3 kg

454

of clarified pomegranate juice, 2.64 kg of the permeate and 0.65 kg of retentate, were obtained

455

(final WRF 5, recovery factor 80.3%). The recovery factor of glucose and fructose in the

456

permeate stream was of the same order of the recovery factor of the process (about 80%); 86% of

457

phenolic compounds and more than 89% of anthocyanins were recovered in the retentate stream.

458

In order to improve the removal of glucose and fructose from the retentate a diafiltration step was

459

applied. The process was operated until to WRF 5; then the retentate was filled up with purified

460

water to its original volume. The filtration and dilution procedure was repeated thus approaching

461

diafiltration conditions. The last filtration run was operated until the minimum retentate hold-up

462

was reached. The redilution of the retentate had a positive impact on the transmembrane flux. As

463

soon as the retentate was filled up with water a significant increase in the flux, up to 20 kg/m2h

464

was observed (Figure 6). The addition of water reduces the osmotic pressure of the retentate and

465

a higher driving force filtration is obtained (Schütte et al., 2015).

466

The composition of the collected diafiltrate in terms of glucose and fructose revealed a similar

467

content of the initial retentate solution indicating that these compounds are still removed in the

AC

CE

PT

ED

MA

NU

SC RI PT

444

20

ACCEPTED MANUSCRIPT

permeate side. On the other hand, the observed rejection for all investigated anthocyanin

469

compounds was higher than 97%. High retentions of phenolic compounds and TAA were also

470

detected (Table 7).

471

Consequently, the efficiency of the Desal GK membrane in the purification of bioactive

472

compounds from clarified pomegranate juice is enhanced by combining the concentration step

473

with diafiltration due to the different selectivity of the membrane towards these compounds: the

474

water addition increases the quantity of sugars in the diafiltrate increasing at the same time the

475

purification factor of biologically active compounds in the retentate. The water addition may also

476

reduce the amount of membrane fouling and improve the solute separation by reducing the

477

solute-solute interactions in the retentate (Almanasrah et al., 2015; Nguyen et al., 2016).

478

MA

NU

SC RI PT

468

3.5. Mass balance of the process

480

The growing interest of health promoting activities exerted by pomegranate juice has addressed

481

the researchers towards the optimization of new extraction and separation techniques in order to

482

produce extracts enriched in biologically active compounds for their use in other markets such as

483

those of functional ingredients, nutraceuticals, cosmeceuticals and food colourings. On the basis

484

of the obtained results, a mass balance of the membrane fractionation process was carried out in

485

order to quantify the amount of biologically active compounds and sugars recovered in the

486

different permeate and retentate fractions. The mass balance, illustrated in Figure 7, was

487

estimated for an initial volume of clarified juice of 1000 L. According to the final volume

488

reduction factor of the process, about 200 L of concentrated solution are obtained. In these

489

conditions the yields of polyphenols and anthocyanins in the retentate stream are of the order of

490

84.8% and 90.7%, respectively. By applying the diafiltration step the efficiency of glucose and

491

fructose recovery can be increased up to 90% and 93%, respectively.

AC

CE

PT

ED

479

21

ACCEPTED MANUSCRIPT

The final retentate of the process exhibits very high antioxidant activity: it can be reused for the

493

formulation of nutraceutical products or as a natural colorant in alternative to the use of synthetic

494

ones; the residual permeate and diafiltrate streams, with high content of sugars, can be reused as

495

food additives or as bases for soft drinks.

496

SC RI PT

492

Conclusions

498

The potential of commercial flat-sheet membranes in the separation and purification of bioactive

499

compounds from sugars in clarified pomegranate juice was evaluated. Preliminary experiments

500

were carried out to assess the performance of the membranes in terms of productivity, fouling

501

index, cleaning efficiency and rejection coefficients. All the tested membranes showed high

502

retention towards biologically active compounds and low retention towards sugars. Among the

503

investigated membranes the Desal GK exhibited higher productivity, lower fouling index and

504

good cleaning efficiency. Permeate fluxes and rejection coefficients of phenolic compounds were

505

not affected by the operating pressure in the range of investigated values. The phenolic content,

506

the content of single anthocyanins and the antioxidant activity of the retentate stream was in

507

agreement with the reached WRF when the clarified juice is treated according to the batch

508

concentration configuration, confirming the suitability of the selected membrane for

509

concentration purposes. At the same time the low retention of the Desal GK membrane towards

510

fructose and glucose allows the recovery of these compounds in the permeate stream. The sugar

511

yield can be improved significantly if diafiltration is applied.

512

The retentate stream, enriched in phenolic compounds, exhibits very high antioxidant activity

513

suggesting its reuse for the formulation of nutraceutical products. It can be also reused as natural

514

colorant as alternative to the use of synthetic ones due to the presence of anthocyanins.

AC

CE

PT

ED

MA

NU

497

22

ACCEPTED MANUSCRIPT

515

Permeate and diafiltrate fractions enriched in sugar compounds can be reused as food additives or

516

as bases for soft drinks.

SC RI PT

517 518

References

519

Almanasrah, M., Brazinha, C., Kallioinen, M., Duarte, L.C., Roseiro, L.B., Bogel-Lukasik, R.,

520

Carvalheiro, F., Mänttäri, M., & Crespo, J.G. (2015). Nanofiltration and reverse osmosis as a

521

platform for production of natural botanic extracts: The case study of carob by-products.

522

Separation and Purification Technology, 149, 389-397.

Al-Jarallah, A., Igdoura, F., Zhang, Y., Tenedero, C.B., White, E.J., MacDonald, M.E., Igdoura,

524

S.A., & Trigatti, B.L. (20139. The effect of pomegranate extract on coronary artery

525

atherosclerosis in SR-BI/APOE double knockout mice. Atherosclerosis, 228, 80-89.

MA

NU

523

Amyrgialaki, E., Makris, D.P., Mauromoustakos, A., & Kefalas, P. (2014). Optimisation of the

527

extraction of pomegranate (Punica granatum) husk phenolics using water/ethanol solvent

528

systems and response surface methodology. Industrial Crops and Products, 59, 216-222.

530

Aspelund, M.T., & Glatz, C.E. (2010). Purification of recombinant plant-made proteins from corn extracts by ultrafiltration. Journal of Membrane Science, 353, 103-110.

PT

529

ED

526

Aviram M., Volkova, N., Coleman, R., Dreher, M., Reddy, M.K., Ferreira, D., & Rosenblat, M.

532

(2008). Pomegranate phenolics from the peels, arils, and flowers are antiatherogenic: Studies in

533

vivo in atherosclerotic apolipoprotein E-deficient (E0) mice and in vitro in cultured

534

macrophages and lipoproteins. Journal of Agricultural and Food Chemistry, 56, 1148-1157.

536

AC

535

CE

531

Baker, R.W. (2004). Membrane Technology and Applications, John Wiley and Sons, Ltd., England.

23

ACCEPTED MANUSCRIPT

Baklouti, S., Ellouze-Ghorbel, R., Mokni, A., & Chaabouni, S. (2012). Clarification of

538

pomegranate juice by ultrafiltration: study of juice quality and of the fouling mechanism. Fruits,

539

67, 215-225.

540 541

SC RI PT

537

Banihani, S., Swedan, S., & Alguraan, Z. (2013). Pomegranate and type 2 diabetes. Nutrition Research, 33, 341-348.

542

Benítez, F.J., Acero, J.L., Leal, A.I., & Gonzáles, M. (2009). The use of ultrafiltration and

543

nanofiltration membranes for the purification of cork processing wastewater. Journal of

544

Hazardous Materials, 162, 1438-1445.

Cassano, A., Conidi, C., & Drioli, E. (2011). Clarification and concentration of pomegranate

546

juice (Punica granatum L.) using membrane processes. Journal of Food Engineering, 107, 366-

547

373.

549

MA

548

NU

545

Cassano, A., Conidi, C., & Ruby-Figueroa, R. (2014). Recovery of flavonoids from orange press liquor by an integrated membrane process. Membranes, 4, 509-524. Cassano, A., Conidi, C., & Tasselli, F. (2015). Clarification of pomegranate juice (Punica

551

granatum L.) by hollow fibre membranes. analyses of membrane fouling and performance.

552

Journal of Chemical Technology and Biotechnology, 90, 859-866.

PT

ED

550

Cassano, A. (2016). Integrated membrane processes in the food industry. In A. Basile, C.

554

Charcosset (Eds.), Integrated Membrane Systems and Processes (pp. 34-60). John Wiley &

555

Sons, Chichester.

CE

553

Cissè, M., Vaillant, F., Pallet, D., & Dornier, M. (2011). Selecting ultrafiltration and

557

nanofiltration membranes to concentrate anthocyanins from roselle extract (Hibiscus sabdariffa

558

L.). Food Research International, 44, 2607-2614.

559 560

AC

556

Conidi, C., Cassano, A., & Garcia-Castello, E. (2014). Valorization of artichoke wastewaters by integrated membrane process. Water Research, 48, 363-374. 24

ACCEPTED MANUSCRIPT

Conidi, C., Rodriguez-Lopez, A.D., Garcia-Castello, E.M., & Cassano, A. (2015). Purification of

562

artichoke polyphenols by using membrane filtration and polymeric resins. Separation and

563

Purification Technology, 144, 153-161.

SC RI PT

561

564

Dafny-Yalin, M., Glazer, I., Bar-Ilan, I., Kerem, Z., Holland, D., & Amir, R. (2010). Color,

565

sugars and organic acids composition in aril juice and peel homogenates prepared from different

566

pomegranate accessions. Journal of Agricultural and Food Chemistry, 58, 4342-4352.

567 568

Dai, J., & Mumper, R.J. (2010). Plant phenolics: extraction, analysis and their antioxidant and anticancer properties. Molecules, 15, 7313-7352.

Daufin, G., Escudier, J.P., Carrère, H., Bérot, S., Fillaudeau, L., & Decloux, M. (2001). Recent

570

and emerging applications of membrane processes in the food and dairy industry. Food and

571

Bioproducts Processing, 79, 89-102.

MA

NU

569

Díaz-Reinoso, B., Moure, A., Domínguez, H., & Parajó, J.C. (2009). Ultra- and nanofiltration of

573

aqueous extracts from distilled fermented grape pomace. Journal of Food Engineering, 91, 587-

574

593.

ED

572

Drioli E., Romano, M., 2001. Progress and new perspectives on integrated membrane operations

576

for sustainable industrial growth. Industrial & Engineering Chemistry Research, 40, 1277-1300.

577

Erkan, M. (2011). Pomegranate (Punica granatum L.). In E.M. Yahia (Ed.), Postharvest biology

578

and technology of tropical and subtropical fruits (pp. 287-311). Woodhead Publishing Limited,

579

Oxford.

581

CE

Faria, A., & Calhau, C. (2011). The bioactivity of pomegranate: impact on health and disease.

AC

580

PT

575

Critical Reviews in Food Science and Nutrition, 51, 626-634.

582

Ferrara, G., Cavoski, I., Pacifico, A., Tedone, L., & Mondelli, D. (2011). Morpho-pomological

583

and chemical characterization of pomegranate (Punica granatum L.) genotypes in Apulia region,

584

Southeastern Italy. Scientia Horticulturae, 130, 599-606. 25

ACCEPTED MANUSCRIPT

Galanakis, C.M. (2012). Recovery of high added-value components from food wastes:

586

conventional, emerging technologies and commercialized applications. Trends in Food Science

587

& Technology, 26, 68-87.

SC RI PT

585

588

Galanakis, C.M., Markouli, E., & Gekas, V. (2013). Recovery and fractionation of different

589

phenolic classes from winery sludge using ultrafiltration. Separation and Purification

590

Technology, 107, 245-251.

591 592

Galanakis, C.M., Chasiotis, S., Botsaris, G., & Gekas, V. (2014). Ultrafiltration optimization for the recovery of -glucan from oat mill waste. Food Research International, 65, 477-483.

Galanakis, C.M. (2015). Separation of functional macromolecules and micromolecules: From

594

ultrafiltration to the border of nanofiltration. Trends in Food Science & Technology, 42, 44-63.

595

Galanakis, C.M., Kotanidis, A., Dianellou, M., & Gekas, V. (2015). Phenolic content and

MA

596

NU

593

antioxidant capacity of Cypriot wines. Czech Journal of Food Science, 33, 126-136. Giacobbo, A., Bernardes, A.M., & de Pinho, M.N. (2013). Nanofiltration for the recovery of low

598

molecular weight polysaccharides and polyphenols from winery effluents. Separation Science

599

and Technology, 48, 2524-2530.

ED

597

Gil, M.I., Tomas-Barberan, F.A., Hess-Pierce, B., Holcroft, D.M., & Kader, A.A. (2000).

601

Antioxidant activity of pomegranate juice and its relationship with phenolic composition and

602

processing. Journal of Agricultural and Food Chemistry, 50, 4581-4589.

CE

PT

600

Gumienna, M., Szwengiel, A., & Górna, B. (2016). Bioactive components of pomegranate fruit

604

and their transformation by fermentation processes. European Food Research and Technology,

605

242, 631-640.

606 607

AC

603

Havlíková, L., & Miková, K. (1985). Heat stability of anthocyanins. Zeitschrift fuer Lebensmittel-Untersuchung und –Forschung, 181, 427-432.

26

ACCEPTED MANUSCRIPT

He, L., Zhang, X., Xu, H., Xu, C., Yuan, F., Knez, Z., Novak, Z., & Gao, Y. (2012). Subcritical

609

water extraction of phenolic compounds from pomegranate (Punica granatum L.) seed residues

610

and investigation into their antioxidant activities with HPLC–ABTS +assay. Food and

611

Bioproducts Processing, 90, 215-223.

612 613

SC RI PT

608

Jurenka, J.S., 2008. Therapeutic applications of pomegranate (Punica granatum L.): a review. Alternative Medicine Review, 13, 128-144.

Lansky, E.P., & Newman, R.A. (2007). Punica granatum (pomegranate) and its potential for

615

prevention and treatment of inflammation and cancer. Journal of Ethnopharmacology, 109,

616

177-206.

618

Li, J., & Chase, H.A. (2010). Applications of membrane techniques for purification of natural products. Biotechnology Letters, 32, 601-608.

MA

617

NU

614

Mello, B.C.B.S., Cunha Petrus, J.C., & Hubinger, M.D. (2010). Concentration of flavonoids and

620

phenolic compounds in aqueous and ethanolic propolis extracts through nanofiltration. Journal

621

of Food Engineering, 96, 533-539.

ED

619

Mondello, L., Cotroneo, A., Errante, G., Dugo, G., & Dugo, P. (2000). Determination of

623

anthocyanins in blood orange juices by HPLC analysis. Journal of Pharmaceutical and

624

Biomedical Analysis, 23, 191-195.

PT

622

Mirsaeedghazi, H., Emam-Djomeh, Z., Mousavi, S.M., Ahmadkhaniha, R., & Shafiee, A.

626

(2010a). Effect of membrane clarification on the physicochemical properties of pomegranate

627

juice. International Journal of Food Science and Technology, 45, 1457-1463.

AC

CE

625

628

Mirsaeedghazi, H., Emam-Djomeh, Z., Mousavi, S.M., Aroujalian A., & Navidbakhsh, M.

629

(2010b). Clarification of pomegranate juice by microfiltration with PVDF membranes.

630

Desalination, 45, 1457-1463.

27

ACCEPTED MANUSCRIPT

Mousavinejad, G., Emam-Djomeh, Z., Rezaei, K., & Khodaparast, M.H.H. (2009). Identification

632

and quantification of phenolic compounds and their effects on antioxidant activity in

633

pomegranate juices of eight Iranian cultivars. Food Chemistry, 115, 1274-1278.

SC RI PT

631

634

Murakami, A.N.N., Amboni, R.D.D.C., Prudencio, E.S., Amante, E.R., Fritzen-Freire, C.B.,

635

Boaventura, B.C.B., Munoz, I.D., Branco, C.D., Salvador, M., & Maraschin, M. (2013).

636

Concentration of biologically active compounds extracted from Ilex paraguariensis St. Hil. by

637

nanofiltration. Food Chemistry, 141, 60-65.

Nguyen, D.T.N.N., Lameloise, M.L., Guiga, W., Lewandowski, R., & Bouix, M. (2016).

639

Optimization and modeling of diananofiltration for the detoxification of lingo-cellulosic

640

hydrolysates- Study at pre-industrial scale. Journal of Membrane Science, 512, 111-121.

642

Onsekizoglu, P. (2013). Production of high quality clarified pomegranate juice concentrate by

MA

641

NU

638

membrane processes. Journal of Membrane Science, 442, 264-271. Ozgen, M., Durgaç, C., Serçe, S., & Kaya, C. (2008). Chemical and antioxidant properties of

644

pomegranate cultivars grown in the Mediterranean region of Turkey. Food Chemistry, 111, 703-

645

706.

647

Patsioura, A., Galanakis, C.M., & Gekas, V. (2011). Ultrafiltration optimization for the recovery

PT

646

ED

643

of beta-glucan from oat mill waste. Journal of Membrane Science, 373, 153-163. Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C.A. (1999).

649

Antioxidant activity applying and improved ABTS radical cation decolorization assay. Free

650

Radical Biology and Medicine, 26, 1231-1237.

AC

CE

648

651

Ruiz-Rodríguez, B.M., Morales, P., Fernández-Ruiz, V., Sánchez-Mata, M.C., Cámara, M., Díez-

652

Marqués, C., Pardo-de-Santayana, M., Molina, M., Tardío, J., 2011. Valorization of wild

653

strawberry-tree fruits (Arbutus unedo L.) through nutritional assessment and natural production

654

data. Food Research International, 44, 1244-1253. 28

ACCEPTED MANUSCRIPT

655 656

Schütte, T., Niewersch, C., Wintgens, T., & Yüce, S. (2015). Phosphorus recovery from sewage by nanofiltration in diafiltration mode. Journal of Membrane Science, 480, 74-82. Seeram, N.P., Aviram, M., Zhang, Y., Henning S.M., Feng, L., Dreher, M. & Heber, D. (2008).

658

Comparison of antioxidant potency of commonly consumed polyphenol-rich beverages in the

659

United States. Journal of Agriculture and Food Chemistry, 56, 1415-1422.

SC RI PT

657

Singh, M., Jha, A., Kumar, A., Hettiarachchy, N., Rai, A.K., & Sharma, D. (2014). Influence of

661

the solvents on the extraction of major phenolic compounds (punicalagin, ellagic acid and gallic

662

acid) and their antioxidant activities in pomegranate aril. Journal of Food Science and

663

Technology-Mysore, 51, 2070-2077.

NU

660

Singleton, V.L., Orthofer, R., & Lamuela-Raventós, R.M. (1999). Analysis of total phenols and

665

other oxidation substrates and antioxidants by means of Folin-Ciocalteau reagent. Methods in

666

Enzymology, 299, 152-178.

668

Sood, A., & Gupta, M. (2015). Extraction process optimization for bioactive compounds in pomegranate peel. Food Bioscience, 12, 100-106.

ED

667

MA

664

Sreeja, S., Hima, S., Parvathy, M., Juberiya, M.A., & Sreeja, S. (2014). Pomegranate fruit as a

670

rich source of biologically active compounds. Biomed Research International, Article

671

ID 686921, 1-12.

PT

669

Tabaraki, R., Heidarizadi, E., & Benvidi, A. (2012). Optimization of ultrasonic-assisted

673

extraction of pomegranate (Punica granatum L.) peel antioxidants by response surface

674

methodology. Separation and Purification Technology, 98, 16-23.

AC

CE

672

675

Teixeira, A.R.S., Santos, J.L.C., & Crespo, J.G. (2014). Solvent resistant diananofiltration for

676

production of steryl esters enriched extracts. Separation and Purification Technology, 135, 243-

677

251.

29

ACCEPTED MANUSCRIPT

Tezcan, F., Gültekin-Özgüven, M., Diken, T., Özçelic, B., & Erim, F.D. (2009). Antioxidant

679

activity and total phenolics, organic acid and sugar content in commercial pomegranate juices.

680

Food Chemistry, 115, 873-877.

SC RI PT

678

681

Tsibranska, I., & Tylkowski, B. (2013). Concentration of ethanolic extracts from Sideritis ssp. L.

682

by nanofiltration: Comparison of dead-end and cross-flow modes. Food and Bioproducts

683

Processing, 91 (2013) 169-174.

Tylkowski, B., Trusheva, B., Bankova, V., Giamberini, M., Peev, G., & Nikolova, A. (2010).

685

Extraction of biologically active compounds from propolis and concentration of extract by

686

nanofiltration. Journal of Membrane Science, 348, 124-130.

688

Tylkowski, B., & Tsibranska, I. (2015). Overview of main techniques used for membrane characterization. Journal of Chemical Technology and Metallurgy, 50, 2015, 3-12.

MA

687

NU

684

Vegara, S., Mena, P., Martí, N., Saura, D., & Valero, M. (2013). Physicochemical characteristics,

690

polyphenol compositions and antioxidant potential of pomegranate juices from 10 Chinese

691

cultivars and the environmental factors analysis. Food Chemistry, 141, 1630-1636.

ED

689

Viuda-Martos, M., Fernandez-Lopez, J., & Perez-Alvarez, J.A. (2010). Pomegranate and its

693

many functional components as related to human health: a review. Comprehensive Reviews in

694

Food Science and Food Safety, 9, 635-654.

696

Zarei, M., Azizi, M., & Bashir-Sadr, Z. (2011). Evaluation of physicochemical characteristics of

CE

695

PT

692

pomegranate (Punica granatum L.) fruit during ripening. Fruits, 66, 121-129. Zheng, X., Liu, B., Li, L., Zhu, X., 2011. Microwave-assisted extraction and antioxidant activity

698

of total phenolic compounds from pomegranate peel. Journal of Medical Plants Research, 5,

699

1004-1011.

AC

697

30

ACCEPTED MANUSCRIPT Figure captions

Figure 1 - Clarification of pomegranate juice by UF. Time course of permeate flux and WRF.

NU SC RI PT

(Operating conditions: TMP = 0.6 bar; Qf = 400 L/h; T = 25±1 °C).

Figure 2 - Time course of permeate flux for clarified pomegranate juice processed with: (a) Etna 01PP, MPF-36 and PES 004H membranes and (b) Desal GK membrane (operating conditions: TMP, 10 bar; T, 25±1 °C; WRF, 1.5).

Figure 3 - Retention of selected membranes towards: (a) anthocyanins; (b) phenolic

MA

compounds and TAA; (c) sugars.

Figure 4 - Ultrafiltration of clarified pomegranate juice with Desal GK membrane. Effect of TMP on steady-state permeate flux and retention towards polyphenols, total antioxidant

ED

activity (TAA) and total soluble solids (TSS).

PT

Figure 5 - Ultrafiltration of clarified pomegranate juice with Desal GK membrane. Effect of

CE

TMP on anthocyanins retention.

Figure 6 - Ultrafiltration of pomegranate juice. Time course of permeate flux and WRF

AC

(Operating mode: batch concentration/diafiltration; operating conditions: TMP, 15 bar; T, 25±1 °C)

Figure 7 – Mass balance of the fractionation process of the pomegranate clarified juice with Desal GK membrane (GLU, glucose; FRU, fructose; POL, polyphenols; ANT, anthocyanins).

ACCEPTED MANUSCRIPT

40

6 Jp

35

5

WRF

3

15

2

10

1

5 0

0

0

100

200

300

400

MA

Operating time (min)

AC

CE

PT

ED

FIGURE 1

500

600

WRF

20

NU SC RI PT

4

25

2

Jp (kg/m h)

30

ACCEPTED MANUSCRIPT

8 ETNA 01PP MPF 36 PES 004H

NU SC RI PT

2

Jp (kg/m h)

6

4

2

(a) 0 0

200

400

600

800

1000

Operating time (min)

MA

25

15

ED

2

Jp (kg/m h)

20

5

(b)

0

AC

CE

0

PT

10

FIGURE 2

100

200

Operating time (min)

300

400

ACCEPTED MANUSCRIPT

100 Cyanidin 3,5 diglucoside Cyanin 3, glucoside Pelargolidin 3-glucoside Delphinidin-3-glucoside

60

40

20

(a) 0 PES 004

ETNA 01PP

MPF 36

Membranes

100

NU SC RI PT

Retention (%)

80

DESAL GK

Total polyphenols TAA

60

MA

Rejection (%)

80

40

(b)

ED

20

0

ETNA 01PP

MPF 36

DESAL GK

PT

PES 004

100

TSS glucose fructose

80

Rejection (%)

(c)

AC

CE

FIGURE 3

60

40

20

0 PEES 004

ETNA 01PP

MPF 36

DESAL GK

ACCEPTED MANUSCRIPT

50

40

60

30

40

20

20

10

0

0

0

5

10

15

TMP (bar)

20

25

AC

CE

PT

ED

MA

FIGURE 4

Jp (kg/m2h)

80

Polyphenols TAA TSS Jp

NU SC RI PT

Retention (%)

100

30

ACCEPTED MANUSCRIPT

100

60

cyanidin 3,5-diglucoside cyanidin 3-glucoside pelargolidin 3-glucoside delphinidin-3-glucoside

40

20 5

10

15

NU SC RI PT

Rejection (%)

80

20

MA

TMP (bar)

AC

CE

PT

ED

FIGURE 5

25

30

ACCEPTED MANUSCRIPT

35

6 Jp

30

VRF

5

4

WRF

20

NU SC RI PT

Jp (kg/m2h)

25

15

3

10

2

5

0 0

200

400

600

800

1000

MA

Operating time (min)

AC

CE

PT

ED

FIGURE 6

1200

1 1400

RI PT

Vfeed = 1000 L GLU = 11.09 kg FRU = 19.18 kg POL = 2.704 kg ANT = 0.183 kg

UF, 150 kDa

Clarified juice

UF retentate

NU

UF, 2kDa VRF = 5

SC

Raw juice

Vret = 200 L

GLU = 1.855 kg FRU = 3.560 kg POL = 2.295 kg ANT = 0.166 kg

Nutraceutical products

UF retentate

Vperm = 300 L

GLU = 8.815 kg FRU = 15.86 kg POL = 0.301 kg ANT = 0.018 kg

UF permeate

GLU = 1.117 kg FRU = 2.14 kg POL = 0.071 kg ANT = 0.004 kg

AC

CE PT

ED

UF diafiltrate Food additives

FIGURE 7

Water (300 L)

MA

Vperm = 800 L

Diafiltration

V = 200 L

GLU = 1.15 kg FRU = 1.17 kg POL = 2.22 kg ANT = 0.16 kg

Final product

RI PT

Table 1- Characteristics of selected flat-sheet membranes Etna 01PP

PES 004H

Manufacturer

Alfa Laval

Mycrodin-Nadir

MWCO (Da)

1,000

pH operating range

1-11 1-11.5

Max. operating temperature (°C)

60

Max. operating pressure (bar)

10

AC

CE PT

ED

pH range in cleaning conditions

Polyethersulphone

Desal GK

Koch

GE

Thin-film

Thin-film

Composite

composite

4,000

1,000

2,000

0-14

1-13

1-11

1-12

0-14

1-12

95

70

70

10

35

27

NU

Composite fluoro polymer

MA

Membrane material

SelRO MPF-36

SC

Membrane Type

ACCEPTED MANUSCRIPT Table 2 - Chemical composition of raw and clarified pomegranate juice Raw juice Clarified juice

Suspended solids (%w/w)

4.00±0.08

n.d.

Total soluble solids (TSS) (°Brix)

17.03±0.04

14.06±0.04

Glucose (g/L)

12.89±0.51

12.53±0.48

21.22±0.42

19.45±0.62

150.90±3.20

136.10±5.30

57.66±4.50

53.71±2.06

17.80±0.11

14.61±0.48

5.60±0.95

4.77±0.43

26.80±2.90

26.00±2.80

2636.80±12.80

2457.50±15.30

Fructose (g/L) Cyanidin 3,5-O-diglucoside (mg/L) Cyanidin 3-O-glucoside (mg/L) Delphinidin 3-O-glucoside (mg/L) Pelargolidin 3,5-O-diglucoside (mg/L) TAA (mM Trolox)

AC

CE

PT

ED

MA

Total polyphenols (mg GAE/L)

NU SC RI PT

Parameter

ACCEPTED MANUSCRIPT Table 3 – Water permeabilities, fouling index and productivity of selected membranes in the treatment of clarified pomegranate juice

Membrane type PES 004H

MPF-36

Desal GK

WP0 (kg/m2hbar)

22.16

47.86

14.58

10.62

WP1 (kg/m2hbar)

1.17

5.8

4.04

5.20

(kg/m2hbar)

20.88

43.08

14.07

10.34

88

72.3

51

90

96.50

97.36

2.4

10

34

0.7

2

17

WP3

FI (%)

94.72

CE (%)

94.22

J0 (kg/m2h)

6.5

Jstaz (kg/m2h)

3

NU SC RI PT

Etna 01PP

AC

CE

PT

ED

MA

Legend: WP0 =water permeability before the treatment with pomegranate juice; WP1 =water permeability after the treatment with pomegranate juice; WP3 =water permeability after chemical cleaning; FI = fouling index; CE = cleaning efficiency; Jo = initial permeate flux; Jstaz = steady-state permeate flux

ACCEPTED MANUSCRIPT Table 4 – Water permeabilities and fouling index of Desal GK membrane for experimental runs performed at different TMP values WP0

WP1

WP2

WP3

FI

(bar)

(kg/m2hbar)

(kg/m2hbar)

(kg/m2hbar)

(kg/m2hbar)

(%)

5

10.22

6.29

10

10.19

15

NU SC RI PT

TMP

6.8

10.19

30.45

6.79

7.46

10.63

33.36

10.63

7.28

7.40

9.89

31.51

20

9.89

6.59

6.70

9.93

33.30

25

9.93

6.40

7.04

10.02

35.50

AC

CE

PT

ED

MA

Legend: WP0 =water permeability before the treatment with pomegranate juice; WP1 =water permeability after the treatment with pomegranate juice; WP2 =water permeability after cleaning with water; WP3 =water permeability after chemical cleaning; FI = fouling index

ACCEPTED MANUSCRIPT Table 5 - Analyses of TSS, sugars, total polyphenols, anthocyanins and TAA in juice samples from UF with Desal GK membrane (WRF 5) Feed

Permeate

Retentate

Glucose (g/L)

11.09±1.7

11.02±0.5

9.29±0.4

Fructose (g/L)

19.27±1.27

19.83±1.6

17.80±1.5

TSS (°Brix)

14.66±0.05

11.73±0.23

21.03±0.05

Total polyphenols (mg/L)

2704.4±14.1

295.0±5.9

11478.0±17.2

TAA (mM Trolox)

27.74±1.15

6.24±0.12

118.27±4.6

Cyanidin 3,5-O-diglucoside (mg/L)

120.2±10.2

14.4±1.2

562.8±9.7

43.5±1.5

6.7±0.5

185.1±3.4

15.24±0.25

1.68±0.07

71.35±3.2

4.1±0.1

0.74±0.03

16.2±1.2

Cyanidin 3-O-glucoside (mg/L) Delphinidin 3-O-glucoside (mg/L)

AC

CE

PT

ED

MA

Pelargolidin 3,5-O-diglucoside (mg/L)

NU SC RI PT

Parameter

ACCEPTED MANUSCRIPT Table 6 - Mass balance of the UF process with Desal GK membrane (WRF 5) Feed

Total permeate

Final retentate

Balance

3.300 kg

2.65 kg

80.3%

0.65 kg

19.7 %

100.0%

Glucose

36.6 g

29.2 g

79.8%

6.03 g

16.5%

96.3%

Fructose

63.3 g

52.5 g

Total anthocyanins

0.604 g

0.062 g

Total polyphenols

8.924 g

0.966 g

NU SC RI PT

Volume

11.57 g

18.0%

100.0%

10.0%

0.542 g

89.7%

99.7%

11.0%

7.634 g

86.0%

96.0%

AC

CE

PT

ED

MA

82.0%

ACCEPTED MANUSCRIPT Table 7 - Analyses of TSS, sugars, total polyphenols, anthocyanins and TAA in samples of pomegranate juice coming from the diafiltration process Rejection

Feed

Diafiltrate

Glucose (g/L)

10.38±0.06

11.02±0.5

0

Fructose (g/L)

15.37±0.05

15.42±0.32

0

TSS (°Brix)

14.66±0.05

11.73±0.23

19.9±0.2

Total polyphenols (mg/L)

112704.4±14.1

234.6±0.3

97.9±0.4

TAA (mM Trolox)

115.76±15.01

2.67±0.42

97.6±0.6

Cyanidin 3,5-O-diglucoside (mg/L)

540.8±8.3

7.61±0.80

98.5±1.5

Cyanidin 3-O-glucoside (mg/L)

180.1±2.4

4.24±0.58

97.6±1.5

Delphinidin 3-O-glucoside (mg/L)

73.3±1.9

1.32±0.11

98.1±0.5

Pelargolidin 3,5-O-diglucoside (mg/L)

15.1±1.2

0.41±0.4

97.2±1.4

MA ED PT CE AC

(%)

NU SC RI PT

Parameter