Accepted Manuscript Structural and oxidative stabilization of spray dried fish oil microencapsulates with gum arabic and sage polyphenols: characterization and release kinetics
P.K. Binsi, Natasha Nayak, P.C. Sarkar, A. Jeyakumari, P. Muhamed Ashraf,
George Ninan, C.N. Ravishankar PII:
S0308-8146(16)31521-7 DOI: http://dx.doi.org/10.1016/j.foodchem.2016.09.126
Reference:
FOCH 19906 To appear in:
Food Chemistry Received Date:
18 April 2016 Revised Date:
13 August 2016 Accepted Date:
19 September 2016 Please cite this article as: Binsi, P.K., Nayak, N., Sarkar, P.C., Jeyakumari, A., Muhamed Ashraf, P., Ninan, G.,
Ravishankar, C.N., Structural and oxidative stabilization of spray dried fish oil microencapsulates with gum arabic and sage polyphenols: characterization and release kinetics, Food Chemistry (2016), doi: http://dx.doi.org/10.1016/ j.foodchem.2016.09.126
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.
1
Structural and oxidative stabilization of spray dried fish oil microencapsulates
1 with gum arabic and sage polyphenols: characterization and release kinetics
2
3 Binsi PK1*, Natasha Nayak1, Sarkar PC2., Jeyakumari A3, Muhamed Ashraf P1.,George
4 Ninan1, Ravishankar CN1 5
6 1. ICAR- Central Institute of Fisheries Technology (CIFT), Matsyapuri, Willington Island,
7 Cochin, India – 682 029 8 2. ICAR - Indian Institute of Natural Resins and gums, Namkum, Ranchi-834 010
9 3. ICAR-Mumbai Research Centre of CIFT, Sector-1, Vashi, Navi Mumbai- Maharashtra
10 400 703 11
12
13
14 * Corresponding author 15 Tele: ++ 91-484 - 2412300
16 Fax: ++ 91 - 484 – 2668212 17 E-mail: binsipk@yahoo.com
18
19 Stable fish oil-sage extract microencapsulates ..
20
2
Abstract 21 The synergistic efficacy of gum arabic and sage polyphenols in stabilising capsule wall
22 and protecting fish oil encapsulates from heat induced disruption and oxidative deterioration
23 during spray drying was assessed. The emulsions prepared with sodium caseinate as wall
24 polymer, gum arabic as wall co-polymer and sage extract as wall stabiliser was spray dried using
25 a single fluid nozzle. Fish oil encapsulates stabilised with gum arabic and sage extract (SOE)
26 exhibited significantly higher encapsulation efficiency compared to encapsulates containing gum
27 arabic alone (FOE). Scanning electron microscopic and atomic force microscopic images
28 revealed uniform encapsulates with good sphericity and smooth surface for SOE, compared to
29 FOE powder. In vitro oil release of microencapsulates indicated negligible oil release in buffered
30 saline whereas more than 80% of the oil loaded in encapsulates were released in simulated GI
31 fluids. The encapsulates containing sage extract showed a lower rate of lipid oxidation during
32 storage.
33 Key words: Fish oil encapsulates, oxidation, sage extract, polyphenol
34
35
36
37
38
39
40
41
42
3
1.
Introduction 43 The health benefits conferred by Omega-3 fatty acids, such as eicosapentaenoic acid
44 (EPA) and docosahexaenoic acid (DHA), arise mainly from its anti-inflammatory and anti- 45 arrhythmic properties that are beneficial to cardiac functioning (Endo, & Arita, 2016). The
46 epidemiological studies have suggested that high fish consumption is inversely associated with
47 cognitive impairment, depression and development of dementia or Alzheimer’s disease (Freund- 48
Levi et al., 2014). In spite of all these health benefits, there are certain human food safety
49 concerns associated with dietary intake of omega-3 rich fish oil. Omega-3 fatty acids are highly
50 susceptible to oxidative changes, producing toxic secondary and tertiary compounds. This inturn
51 restricts the amount of fish oil that can be directly consumed or added to the foods, particularly
52 in infant formula.
53 Microencapsulation of fish oil in a stable wall matrix has been recognised as a suitable
54 technique for the delivery of fish oil supplements, as it reduces oxidative deterioration and
55 improves bioavailability. Among the various techniques used for fish oil microencapsulation, the
56 cheapest one is spray-drying and its cost is 30-50 times lesser than freeze-drying (Gharsallaoui,
57 Roudaut, Chambin, Voilley, & Saurel, 2007). However, the high operational temperature
58 coupled with the mechanical shearing during atomisation is a major challenge in the
59 encapsulation process, as these forces may destabilise the emulsion, and promote capsule
60 collapse as well as oxidation of fish oil during the drying process. Consequently, use of cross- 61 linking agents very often becomes necessary as the ionic nature of the interactions between the
62 wall forming materials alone does not guarantee the structural integrity of the resulting
63 microcapsules (Koupantsis, Pavlidou, & Paraskevopoulou, 2016). Presently, formaldehyde,
64 glutaraldehyde as well as the enzyme transglutaminase have been employed as protein cross- 65
4
linking agents. However, owing to the adverse health and environmental effects associated with
66 aldehydes, and the low economic viability of enzyme mediated processes, the demand for natural
67 and economically sustainable alternatives is high. Hence, incorporation of natural cross-linkers
68 and antioxidants in the emulsion may be highly advisable, especially when spray drying is
69 involved.
70 Controlled release of food ingredients at the right place and right time is another key
71 functionality that can be provided by microencapsulation. Generally, carbohydrate and protein- 72 based microcapsules are water soluble and hence not suitable for controlled-release applications
73 (Cho, Shim, & Park, 2003). However, if the proteins are cross-linked into stable forms, the
74 application of proteins as wall material would be greatly increased for targeted and controlled
75 delivery of sensitive supplements. Cross-linking changes the net charge of protein molecules,
76 and hence can be explored to alter the solubility pattern of protein in a given medium. This has
77 the added technical advantage that the microcapsules act as supplement for omega-3 fish oil and
78 protein, simultaneously.
79 Plant essential oils rich in polyphenols are multifunctional and are ideal for encapsulating
80 fish oil, as they are effective protein cross-linkers, and at the same time serve as potent
81 antioxidant for fish oil by acting as reducing agents or singlet oxygen scavengers. Polyphenols
82 are known to react under oxidising conditions with side chain amino groups of peptides, leading
83 to the formation of protein cross-links, as shown in gelatin–pectin microparticles which
84 possessed enhanced lipophilicity and resistance to thermal degradation after cross-linking with
85 polyphenols in coffee and grape juice (Strauss, & Gibson, 2004). Similarly, significant cross- 86 linking ability of tannic acid on myofibrillar protein was reported (Prodpran, Benjakul, &
87 Phatcharat, 2012). Plant polyphenolic compounds show strong interaction with milk proteins,
88
5
such as casein, as well. Previously, strengthening of the microcapsule walls through reticulation
89 with the use of glycerol and tannic acid for sodium caseinate or whey protein isolate- 90 carboxymethylcellulose microcapsules was reported (Koupantsis et al., 2016). The interaction
91 between proteins and plant phenols proceeds with polymerization mainly by the formation of
92 non-covalent interactions bridged through numerous hydrogen bonds (Frazier et al., 2010) as
93 well as through covalent C–N or C–S bonds (Strauss, & Gibson, 2004). Herbs, such as sage
94 (Salvia sp.), are included under ‘spices and other natural seasonings and flavourings’ that are
95 generally recognized as safe, by USFDA. The antioxidant activity of sage is mainly related to
96 two phenolic diterpenes, carnosic acid and carnosol, which are effective free-radical scavengers
97 (Dorman, Peltoketo, Hiltunen, & Tikkanen, 2003). Gum arabic is a commonly employed agent
98 for encapsulating fish oil and other vegetable oils, because of its colloidal functionality and
99 compatibility with most carbohydrates and proteins that are commonly employed as wall
100 polymers. Nevertheless, the efficacy of herbal extracts on structural stabilisation of fish oil
101 encapsulate is not evaluated so far. Hence, the objective of the present study was to evaluate the
102 efficacy of sage extract and gum arabic on structural and oxidative stabilisation of fish oil
103 encapsulates with an emphasise on minimising the adverse effect of elevated temperature and
104 mechanical shearing during spray drying. Apart from that, the oxidative stability of the
105 encapsulates during accelerated oxidative atmosphere and oil release pattern in buffered saline
106 and gastrointestinal environment was evaluated.
107 2. Materials and methods 108 2.1. Raw materials
109
6
Gum arabic from Acacia Senegal with an average molecular weight of 285 kDa (containing <1%
110 protein and < 1% insoluble matter), pepsin (activity of 800–2500 ml units/mg of protein) and
111 pancreatin 4NF, both of porcine origin, were procured from Hi-Media Pvt Ltd. (Mumbai, India).
112 Sodium caseinate of extrapure grade was purchased from Sisko Research Laboratories Pvt. Ltd.
113 (Mumbai, India). Fish oil extracted from Indian oil sardine (Sardinella longiceps) with known
114 fatty acid composition was used for microencapsulate preparation (Myristic acid: 11.98%, stearic
115 acid: 4.88%, palmitic acid: 23.66%, palmitoleic acid:13.22, oleic acid:8.55%, EPA:13.55%,
116 DHA:10.42%). Sage (Salvia officinalis) extract containing 45% (w/w) essential oil was
117 purchased from Synthite Industries Ltd. (Cochin, India). The composition of sage extract as
118 determined by Gas Chromatography-Mass Spectroscopic (GC-MS-MS) analysis indicated the
119 following components. 1,8, Cineole (12.8%), α-humulene (11.8%), viridiflorol (11.1%), β- 120 caryophyllene (5.8%), α-thujone (4.2%), β-thujone (4.1%), carvacrol (3.6%), thymol (3.4%), β- 121 pinene (3.3%), α-pinene (3.1%) , camphor (3.1%), limonene (2.8%), borneol (2.5%), linalool
122 (2.1%), viridiflorene (2.1%), Myrcene (1.9%), p-cymene (1.2%), linalylacetate (1.1%),
123 camphene (0.9%), 14-hydroxy-9-epi- (E) –caryophyllene (0.8%), caryophyllene oxide (0.7%), γ- 124 muurolene (0.7%),δ -cadinene (0.6%), Aromadendrene (0.4%).
125 2.2. Preparation of emulsions 126 The composition of the emulsions for spray drying was established based on our previously
127 reported results (Jeyakumari, Janarthanan, Chouksey, & Venkateshwarlu, 2014). The stability of
128 the emulsions were confirmed by quantifying phase separation over 24h and microscopic
129 examination of emulsion droplets. Accordingly, emulsions with 7.5% total solids was prepared
130 with sodium caseinate, gum arabic and fish oil in the ratio of 2:2:1 (ie. 3g sodium caseinate, 3g
131 gum arabic and 1.5g fish oil for 100 ml of emulsion). Two different emulsion formulations for
132
7
encapsulation were prepared namely, fish oil (FO emulsion) and fish oil containing 1% sage
133 extract (w/w of fish oil) keeping the total solid content constant at a level of 7.5% (SO
134 emulsion). Emulsions were prepared by dissolving sodium caseinate in a water bath held at 40ºC
135 with intermittent stirring. After proper dissolution, the solution was allowed to cool and gum
136 arabic was added with continuous stirring to avoid lump formation. After the dissolution of wall
137 material, fish oil previously mixed with 1% sage extract was added to the solution. The mixture
138 was homogenized with a high speed homogenizer (Poly system PT 2100, Kinematica, AG) at
139 25000rpm for 5 min. Prior to spray drying, the emulsions were allowed to stabilize at 4ºC for 1 h.
140 2.3. Spray drying of emulsions 141 Spray drying was employed using a pilot-scale spray dryer (Hemraj Pvt Ltd, Mumbai, India)
142 with an atomizer nozzle of 0.5 mm diameter at 450 KPa. The inlet and outlet air temperatures
143 were maintained at 160˚C and 80˚C, respectively and feed rate was adjusted to 15-22 g/min. The
144 encapsulate powder prepared from FO emulsion, SO emulsion are hereafter designated as FOE
145 and SOE, respectively. Three independent trials of encapsulations were carried out for both the
146 emulsions.
147 2.4. Characterization of emulsion 148 2.4.1. Emulsion stability index (ESI)
149 About 150 ml aliquots of each sample were transferred to graduated cylinders of 250 ml and kept
150 at 4oC for 24 h. The volume of the bulk unseparated phase was measured after 24 h and the
151 stability was expressed in terms of stability index:
152
153
8
Where: H0 represents the emulsion initial volume and H1 is the unseparated phase volume.
154 2.4.2. Emulsion microstructure 155 The microstructure of the emulsion was analysed by smearing the emulsion sample onto a
156 microscope slide and observing under an inverted microscope (Leica Microsystems, Wetzlar,
157 Germany) and an atomic force microscope (XE-100, Park Systems, Korea) in non-contact mode
158 employing silicon tips. Emulsions were prepared freshly and analysed at room temperature
159 (25ºC). For inverted microscopy, the images were obtained at 200x magnification and the
160 average size of 5 droplets each of predominant dimensions was determined using image
161 processing software (Leica Microsystems Imaging Solutions, Cambridge, UK) with a CCD
162 camera. For AFM analysis, 10 µl of the emulsion sample was directly pipetted over silica wafers,
163 air dried and analysed over 4µ scanning area.
164 2.5. Analysis of microencapsulates 165 2.5.1. Fourier transform-infrared spectroscopic analysis (FTIR)
166 The FTIR analysis of both the encapsulates were carried out immediately after spray drying
167 using a Thermo Fisher Scientific FT-IR spectrometer (Model NicoletTM iSTM 10, Thermo Fisher
168 Scientific, Waltham, MA), by KBr pellet method in the wavelength range of 4000-400 cm-1. The
169 spectra was analysed using OMNIC software (Thermo Fisher Scientific).
170 2.5.2. Determination of encapsulation efficiency and percentage loss (EE) of oil
171 Surface oil from the spray dried powder was extracted by pentane according to the method
172 described elsewhere (Dieffenbacher & Lüthi, 1986) with slight modifications. 1 g of powder was
173 mixed with 10 ml pentane in a 100 ml Erlenmeyer flask with a stopper. The mixture was shaken
174 for 2.5 h at 25ºC and then passed through a Whatmann No. 4 filter paper. Solvent collected was
175
9
evaporated and surface oil present was determined gravimetrically. The total oil was extracted
176 from the original powder, whereas encapsulated oil was extracted from the residual dried powder
177 after extracting the free oil. About 0.5g of original powder and residual dried powder,
178 respectively, was mixed with aqueous hydrochloric acid solution (2 M, 4 ml) and boiled at 95ºC
179 for 30 min. After cooling, ethanol (2 ml) was added to the mixture and shaken vigorously. To
180 this, 10 ml of petroleum ether was added and the mixture was centrifuged at 9000 rpm for 5 min
181 at 25 ºC. The upper phase was transferred dried in an oven at 105ºC and the oil present in the
182 sample was determined gravimetrically.
183 Encapsulation efficiency (EE %) was determined by the following formula as described by
184 Wang, Liu, Chen, & Selomulya (2016) 185
186 Where, TO is the total oil per gram of encapsulate powder on dry weight basis, SO is the surface
187 oil per gram of encapsulate powder on dry weight basis.
188 Percentage loss during spray drying was calculated based on the total oil content of encapsulates
189 and loaded oil per gram of wall polymer in the emulsion on dry weigh basis.
190
191 Where, LO is the total oil loaded per total solid content in the emulsion (total weight of sodium
192 caseinate, gum arabic and oil), ie, here LO = 0.20 g oil/g solid weight.
193 2.5.3 Morphology of microencapsulates.
194 2.5.3.1 Scanning electron microscopy (SEM) 195
10
The surface appearance and morphology of FOE and SOE microencapsulates were examined by
196 SEM (XL 30 Philips, Netherlands). Samples were fixed onto double-sided adhesive carbon tabs
197 mounted on SEM stubs, coated with gold in a sputter coater, and examined by SEM.
198 2.5.3.2. Atomic force microscopy (AFM) 199 The morphology of the FOE and SOE microencapsulates was investigated by atomic force
200 microscope in non-contact mode employing silicon tips, in the similar way the emulsion droplet
201 was analysed. The samples (100 mg) were dispersed in distilled water (10 ml) by sonicating with
202 a probe sonicator. A 10 µl of the dispersed sample was pipetted over silica wafers, air dried and
203 analysed for its surface characteristics.
204 2.5.4. Physical properties of microencapsulates
205 2.5.4.1. Determination of moisture content 206
Moisture content of microencapsulates was determined by AOAC method (AOAC, 2000).
207 2.5.4.2. Determination of bulk density 208 Briefly, 2g powder was loosely packed in 10 ml graduated cylinder and the volume was
209 recorded. Bulk density of the powder was calculated by dividing weight of the sample by its
210 volume.
211 2.5.4.3. Hygroscopicity 212 One gram of sample was placed in a desiccator with a saturated solution of NaCl (relative
213 humidity of 75.3%). After one week, samples were weighed and hygroscopicity was expressed
214 as the amount of absorbed moisture per 100 g of sample (g/100 g).
215
11
2.5.5. Oil release properties of microencapsulates
216 2.5.5.1. Release characteristics of encapsulates in buffered saline
217 The release kinetics of microencapsulates in buffered saline was analysed by the method given
218 by Hosseini, Zandi, Rezaei, & Farahmandghavi (2013). Microencapsulates (20 mg) were placed
219 in a centrifuge tube containing 5 ml of 60% phosphate buffer saline (pH 7.4) and 40% ethanol.
220 The mixture was incubated at ambient temperature (27±2 ºC) under gentle agitation. At specific
221 time interval of 1h, samples were centrifuged at 9000 rpm for 5 min at 25 ºC. A specific volume
222 of supernatant was sucked out for analysis, and was replaced with an equivalent volume of fresh
223 media. The cumulative amount of oil released in the medium at various sampling time intervals
224 was determined from the standard curve derived separately for fish oil at 215 nm (derived from
225 the absorbance spectra obtained for the fish oil) and fish oil containing 1% sage extract at 275
226 nm (derived from the absorbance spectra obtained for fish oil containing 1% sage extract), using
227 a UV–VIS spectrophotometer. Cumulative percentage of fish oil released was obtained by
228 dividing the cumulative amount of oil released at each sampling time point (Mt) to the total
229 weight of the oil present in the encapsulates (Mo)
230
231 2.5.5.2. Oil release kinetics during in vitro digestion using simulated GI fluids
232 Simulated gastric fluid (SGF) containing pepsin and simulated intestinal fluid (SIF) containing
233 pancreatin were prepared according to the methods given in the US Pharmacopeia (2000). In the
234 case of SIF, a 10-fold higher concentration of pancreatin was used since the sample used was dry
235
12
powder, as suggested previously by Kosaraju, D'ath, & Lawrence (2006) for encapsulated fish
236 oil. Microcapsules (5.0 g) were initially subjected to gastric digestion at 37oC at 100rpm for 2 h.
237 Further, SIF was added and intestinal digestion was continued under similar conditions for
238 another 3h. The quantity of oil released from the microcapsules after exposure to SGF and SIF
239 was determined separately by the method given by Patten, Augustin, Sanguansri, Head, &
240 Abeywardena (2009). Microstructure of the encapsulates was monitored prior and after gastric
241 and intestinal digestion separately using an inverted microscope (Leica Microsystems, Wetzlar,
242 Germany) by directly mounting the emulsion sample at ambient temperature (25 ºC) at 200x
243 magnification.
244 2.5.6. Measurement of lipid oxidation in microencapsulates by accelerated storage study
245 2.5.6.1. Determination of lipid oxidation products
246 The oxidative stability of the microcapsules was evaluated by accelerated storage at 60ºC for 7
247 days, using a hot air oven. The microencapsulates (about 30 g) was placed in sealed glass bottles
248 and covered with aluminum foil to avoid exposure to light. A 30 g sample of pure fish oil used
249 for the encapsulation (PFO) was also analyzed to compare the result. The peroxide values (PV)
250 of microencapsulates were determined during definite intervals according to the method
251 described by Shantha and Decker (1994). Hydroperoxide concentration in the sample was
252 determined using a standard curve made from cumene hydroperoxide and expressed as mEq
253 O2/kg of oil loaded. Changes in secondary oxidation products of the emulsion were determined
254 by measuring Thiobarbituric acid reactive substances (TBARS) according to the method
255 described by McDonald and Hultin (1987) and expressed as mg of malonaldehyde / kg powder.
256 2.5.6.2. Colour 257
13
The colour of microencapsulates were determined by Hunter lab color meter (Color Flex, Hunter
258 Lab Inc., Reston, VA, USA). The samples were filled in a 64 mm glass sample cup to a pre- 259 determined level and L*, a* and b* parameters were determined.
260 2.5.7. Statistical analysis 261 The data obtained were analyzed by one way analysis of variance (ANOVA) using SPSS
262 software version 16.0 (SPSS Inc, Chicago, Illinois, USA). All mean separations were carried out
263 at significance level of 95% (p<0.05).
264 3. Results and discussion 265 3.1. Characterization of emulsion
266 3.1.1. Emulsion stability index 267 Obtaining a stable liquid emulsion is a prerequisite for proper encapsulation in spray-dried
268 powders. Our previous studies indicated that sodium caseinate along with the surface active gum
269 arabic formed stable emulsions with fish oil, without any visible phase separation (Jeyakumari et
270 al., 2014). In the present study, the stability of emulsions prepared with and without sage extract
271 for a period of 24 h was confirmed prior to spray drying. The results indicated that both the
272 emulsions were kinetically stable, with an unseparated phase fraction of 96.66% and 98.67%,
273 respectively, for FO and SO emulsions (Table 1). Several studies have inferred that properties
274 and stability of fish oil emulsions are much affected by the level of minor ingredients, especially
275 emulsifier (Jiménez-Martín, Gharsallaoui, Pérez-Palacios, Carrascal, & Rojas, 2015; Komaiko,
276 Sastrosubroto, & McClements, 2016). It appears that, in SO emulsion sage polyphenols acted as
277 an emulsifier, by bridging between the hydrophobic oil phase and the hydrophilic caseinate-gum
278
14
arabic polymer phase entrapping oil molecules firmly inside the hydrophobic core. On the
279 otherhand, the slight phase separation observed in FO emulsion might be related to the random
280 protein–protein interaction induced by the shearing during homogenization (Koh, Chandrapala,
281 Zisu, Martin, Kentish, & Ashokkumar, 2014). In the present study, a single concentration of sage
282 extract was chosen for encapsulation studies, as 0.5% was found to be effective in protecting
283 emulsion from oxidation under accelerated storage studies (based on linoleic acid peroxidation
284 value: data not shown). However, considering the volatile nature of the extract, 1% level was
285 opted for encapsulation.
286 3.1.2. Emulsion microstructure 287 The emulsions prepared with and without sage extract were observed under an inverted
288 microscope and atomic force microscope. Both emulsions showed a homogeneous distribution of
289 oil droplets with good stability to aggregation or coalescence, when observed under an inverted
290 microscope (Figure 1A &B). It appears that there was an adequate number of wall polymer
291 (sodium caseinate and gum arabic) and oil molecules in the aqueous phase to form stabilized
292 emulsion droplets. Generally, formation of aggregates/flocculates of the wall polymers in
293 emulsions with high polymer content is expected. On the other hand, in cases where the wall
294 polymer is limiting, the excess oil comes out as coalescence (Dickinson, 2003). In the present
295 study, there was no evidence of these destabilizing phenomenons over a period of 1h when
296 observed under microscope. The SO emulsion showed more uniform sized particles with
297 predominance of smaller average droplet size (less than 2 µm) compared to a wide size range of
298 FO emulsion (3-10 ± 1.2 µm), which confers better interfacial stability to SO emulsion.
299 The Atomic Force Microscopic (AFM) image of emulsion droplets over as small
300 scanning area of 4 µm showed well-structured projections in SO emulsion, whereas FO emulsion
301
15
droplet showed more or less a flat image with uneven and less defined projections (Figure 1B &
302 D). Moreover, the mean diameter of the droplets in the scanned area was much less (0.5 nm) for
303 the SO emulsion compared to 13.44 nm for the FO emulsion. It is generally agreed that, small oil
304 droplets will be enclosed and embedded more efficiently within the wall matrix of the
305 microcapsules; hence, the resultant emulsion will be more stable during the spray drying process
306 which is one of the critical parameters to have the optimum efficiency (Sari et al., 2015). It
307 appears that presence of sage polyphenols created a compact configuration of oil droplets with
308 ordered network of wall polymers around the droplets, which could be densely packed at the
309 emulsion interface, as evidenced from microscopic images of fresh emulsion.
310 3.2. Characterisation of oil encapsulates 311 3.2.1. FT-IR spectra of encapsulates
312 In order to understand the nature of the interaction between wall polymers and polyphenols in
313 sage extract, FT-IR spectroscopic analysis of both the encapsulates were carried out (Figure 2).
314 Spectral behaviour of the samples was recorded in the region from 4000 - 400 cm–1, however, the
315 bands in the region 3600 – 1400 cm–1 were analyzed in detail, since they are characteristic of
316 bending and stretching vibrations of OH groups and NH groups that contribute significantly to
317 protein-protein/polyphenol-protein interaction through hydrogen bonding. Another region of
318 critical evaluation was 1800 - 1400 cm–1, characteristic of the bending vibrations of the same
319 groups. In the present study, the typical profile of casein was not observed as gum arabic was
320 also included as a co-polymer. The amide Ι (mainly C=O stretch) and II (C–N stretching coupled
321 with N–H bending modes) peaks for both FOE and SOE appeared at similar frequency of 1681- 322
1606 cm−1 and 1508 cm−1, respectively. However, amide-1 was slightly condensed in SOE. The
323
16
decrease in the intensity of the amide I band in the spectra suggests a major conformational
324 change introduced by the modification of terminal carboxyl groups of casein, in the presence of
325 polyphenols. Similar infrared spectral changes were observed for protein amide I band in casein- 326 tea polyphenol complexes where major protein conformational changes occurred (Hasni et al.,
327 2011). Coincidentally, a more pronounced ester group was detectable at 1744 cm−1 in SOE
328 compared to that in FOE, that might have resulted from the interaction of terminal carboxyl
329 group of casein/gum arabic with phenolic components. A major difference in the spectra
330 appeared near the NH stretching regions of amide-A and amide–B region, where only a single
331 distinct peak was evident in FOE at 3195 cm-1, which disappeared almost completely in SOE.
332 Apart from this, two minor peaks appeared at 2887 and 2969 cm-1 in FOE, were also present in
333 SOE with slight band shift towards lower wave numbers. A separate phenolic band could not be
334 detected in SOE, as it could be merged with that of phenolic amino acids in casein, however a
335 band shift from 1377 in FOE to 1396 cm-1 in SOE was observed, which might be attributed to
336 the O-H in plane deformation in polyphenol. Similarly, a minor band shift from 3585cm-1 in
337 FOE towards a higher wavenumber of 3604cm-1 was observed in SOE, suggesting a possible
338 modification of phenolic components. From the spectral changes observed in the present study,
339 the interaction between the wall polymers and sage polyphenols was evident in SOE sample. The
340 interactions between protein and polyphenols are mediated mainly through non-covalent
341 hydrogen and hydrophobic interactions, which alter the tertiary and secondary structure of
342 complex proteins as evidenced by the changes in characteristic absorption bands (Hasni et al.,
343 2011). In the present study, polymer cross linking through polyphenol complexing was expected
344 as these polyphenolic components contains multiple functional sites, which can introduce inter
345 and intrachain hydrogen bonds at many sites in polymer chains, in turn act as bridging molecules
346
17
for the protein/ carbohydrate polymers in the wall matrix. In addition to that, a covalent
347 interaction was postulated (Strauss, & Gibson, 2004), in which the diphenol moiety of a phenolic
348 acid or other polyphenols is readily oxidized to an orthoquinone in the presence of molecular
349 oxygen. The orthoquinone in turn forms a dimer in a side reaction, or reacts with amino or
350 sulfhydryl side chains of polypeptides to form covalent C–N or C–S bonds with the phenolic
351 ring, with regeneration of hydroquinone. The latter can be reoxidized and bind a second
352 polypeptide, resulting in a cross-link. Alternatively, two quinones, each carrying one chain, can
353 dimerize, also producing a cross-link. In SOE, the spectral change associated with amide-1,
354 appearance of a well-defined ester band and spectral shift in the phenolic band towards higher
355 wavenumber, all suggest wall polymer cross-linking mediated through polyphenol interactions.
356 3.2.2. Encapsulation efficiency (EE) and percentage loss during drying
357 The encapsulation efficiency reflects the degree of protection offered by the wall material
358 to oil droplets embedded within the wall material. As expected, SOE sample indicated higher EE
359 of 73.27% compared to 68.96% for FOE sample, which confirms the structural stabilization of
360 the wall matrix (network formed by wall material polymers) by polyphenols (Table 1). The
361 encapsulation efficiency of the process can be increased by applying higher inlet temperature
362 during drying, which leads to rapid crust formation around drying droplets and thereby giving
363 higher core retention. However, high temperature quite often results in collapse of encapsulates
364 leading to oil migration to the surface. This can be resolved by adding a suitable cross-linking
365 agent for the wall polymer, such as polyphenols for protein, which hastens the formation and
366 integrity of polymer crust. Several researchers (Liang, Shoemaker, Yang, Zhong, & Huang,
367 2013; Soottitantawat, Bigeard, Yoshii, Furuta, Ohkawara, & Linko, 2005) have observed an
368
18
inverse relationship between the emulsion droplet size and the retention of core material.
369 Accordingly, the higher encapsulation efficiency obtained for SOE could be related to the higher
370 emulsion stability, stronger crust formation and the smaller emulsion droplet size as compared to
371 that of FOE encapsulates.
372 The surface oil represents non-encapsulated oil and has been used as an important
373 parameter determining the quality of encapsulated products. In the present study, it was evident
374 from the FT-IR profile that even though the emulsions were kinetically stable at ambient
375 temperature, the elevated temperature and mechanical shearing during spray drying affected the
376 behaviour of wall polymers during atomisation. Accordingly, FOE showed significantly higher
377 surface oil content of 3.87 g/100 g of dry powder (31.01% of total oil content) as compared to
378 3.4 g/100 g of dry powder (26.79% of total oil content) for SOE (Table 1). Total oil is the total
379 amount of oil present in the encapsulate powder after spray drying. The TO registered a slightly
380 higher value of 12.72% of dry powder for SOE compared to 12.5% for FOE (Table 1), however
381 not statistically different indicating certain loss of oil during spray drying in both the emulsions
382 as compared to the total loaded quantity of oil in the emulsion (20% of total solid content). There
383 was only a negligible difference in the percentage loss of oil during spray drying between
384 encapsulates (Table 1), indicating the meagre loss of sage extract by volatilization during drying
385 phase. The slightly higher oil loss observed in FOE might be related to the higher content of free
386 oil, which is more susceptible to volatilisation during spray drying.
387 3.2.3. Morphology of microencapsulates 388 3.2.3.1. Scanning Electron Microscopy (SEM)
389
19
SEM images of encapsulates revealed distinct differences in size and surface regularity. A wide
390 range of particles having varying dimensions were observed for both the samples (Figure 3).
391 Among the two samples, the size variability was higher for FOE samples which suggests that the
392 atomized FO emulsion droplet was a mixture of intact polymer (which was not participated in oil
393 encapsulation), free oil droplets and oil encapsulates. On the otherhand, the microcapsules
394 incorporated with sage extract were more uniform in size and shape with good sphericity (Figure
395 3B), compared to FOE powder (Figure 3A). This indicates the stable structural interactions
396 between polyphenols in sage extract and the Na-caseinate-guar gum matrix polymers, and the
397 presence of higher proportion of oil encapsulates rather than intact polymer and free oil droplets.
398 Previously, fish oil co-encapsulated with phytosterols and limonene using whey protein isolate
399 and sodium caseinate as wall materials yielded good quality microcapsules, with higher retention
400 of EPA and DHA (Chen, Mcgillivrav, Wen, Zhong, & Quek, 2013). On the otherhand, in the
401 FOE sample, most of the capsules appeared to be collapsed with many large internal voids,
402 probably due to the mechanical stress induced by uneven drying at different parts of the droplets.
403 Wrinkles or dimples on the surface were observed in both the SOE and FOE encapsulates but
404 comparatively more in FOE encapsulates. This could possibly be due to the lower encapsulated
405 oil content of FOE allowing extensive shrinkage during the early stage of the drying process and
406 vacuole formation during later drying period. In SOE, the interaction between polyphenols in
407 sage extract and the polymer matrix resulted in fast crust formation, which was resistant to the
408 mechanical stress during spray drying. This inturn confirms the lower surface oil content
409 observed in SOE powder, as there was less opportunity for oil in the core to come out of the
410 capsules. Similarly, small agglomerates were visible in FOE, which could be due to the high
411 surface oil content that adhered the capsules together.
412
20
3.2.3.2. Atomic force microscopy (AFM) 413 The AFM images of SOE powder (Figure 3D) confirmed the higher encapsulation
414 efficiency and regular surface morphology of encapsulates compared to that of FOE (Figure 3C),
415 as represented by the evenly distributed projections/peaks in the SOE image. The spherical
416 encapsulates will be visible as peaks or elevations from the base, whereas the intact polymer and
417 the free oil will be visible as base aligned depressions. Conversely, the AFM image of FOE
418 indicated sparsely distributed uneven peaks on a continuous matrix/base, suggesting lesser
419 number of encapsulates in the composition. The size distribution obtained by AFM indicated the
420 formation of encapsulates near to nanoscale in both FOE and SOE samples. Even though, the
421 emulsion droplet size was considerably higher for FOE sample, the dried particles in the given
422 scanned area of 4µm were smaller in size (107 nm) than SOE (115 nm). This may be related to
423 the change in size composition of atomized droplets as well as the behaviour of wall polymers
424 during different phases of drying. The larger diameter of SOE particles confirms the formation of
425 proper orientation of wall polymers during spraying and formation of a thicker and denser crust
426 during drying with maximum retention of oil in the core, which increased the effective diameter
427 of the encapsulates. Accordingly in FOE, even though the droplet size was higher, the poor
428 stability of droplets during spray drying might have resulted in the shrinkage/collapse of the
429 capsule. Similar to the SEM image, a huge number of small aggregates were visible in the
430 topographical image of FOE samples, where as SOE powder remained as particles without
431 visible aggregates.
432 The height profile provides an estimate of the nature of the surface roughness. The
433 surface roughness of the encapsulates were evaluated and expressed in terms of their root-mean
434 square (RMS), Ra (average roughness) and Rsk (skewness) values extracted from height images
435
21
(Figure 3 C, D, inset). RMS indicates the deviation in the height of the evaluated object and can
436 be used to represent the roughness of a continuous surface (Yu & Ivanisevic, 2004). In the case
437 of encapsulated powder, the peaks or elevations from the base represent the presence of particles
438 or encapsulates in the scanned area. Hence, higher roughness values represent the presence of
439 more encapsulated particles in the given area. Both RMS and Ra indicated higher values for SOE
440 samples than FOE samples. Rsk values of both the samples yielded negative values, with higher
441 absolute value for FOE samples. In particular, a negative Rsk value of continuous surface means
442 that a larger number of valleys with respect to peaks are present on the surface profile, while a
443 positive Rsk value means that the presence of peaks is prevailing. The negative Rsk values
444 observed for samples may be due to the large discontinuity of the scanned area. Accordingly, the
445 lower EE and the formation of a lower number of regular capsules gives a lesser number of
446 elevations or peaks in the FOE samples, which results in the lower roughness values on
447 topographical scanning. The use of roughness values to characterize encapsulates is not
448 widespread, because of the large variability of roughness values of particulate surface calculated
449 from the height data. AFM generally gives accurate measurements for a material surface, rather
450 than for individual particles. The observations in the present study suggest AFM as a useful tool
451 for confirming the efficiency of encapsulation process along with other quantification
452 techniques.
453 3.2.4. Physical properties of encapsulates 454
3.2.4.1. Moisture content 455 Moisture content of fish oil encapsulates is an important parameter, as higher water
456 activity enhances lipid oxidation. Moreover, at higher moisture levels, the wall material changes
457
22
from the glassy state to amorphous rubbery state with a high molecular mobility, leading to the
458 release of encapsulated oil during storage (Velasco, Dobarganes, & Marquez-Ruiz, 2003). The
459 moisture content of FOE and SOE was found to be 5.91% and 5.01% respectively (Table 1). The
460 slightly lower moisture content observed in SOE encapsulates could be attributed to the higher
461 hydrophobicity imparted by sage extract to the encapsulate wall material. Furthermore, the
462 ordered network of wall material induced by the polyphenols in the extract facilitated the
463 distribution of water molecules in definite voids, facilitating easy moisture diffusion during the
464 drying process.
465 3.2.4.2. Bulk density 466 Bulk density is an important property of powdered products owing to functional and
467 economic reasons. High bulk density is desirable to reduce shipping and packaging costs while
468 very low bulk density influences other powder properties, such as flowability. The bulk density
469 of FOE and SOE encapsulates was 0.39 g/ml and 0.42 g/ml, respectively (Table 1), which is
470 higher than the values reported in literature for fish oil encapsulated with skimmed milk powder
471 (Aghbashlo, Mobli, Madadlou, & Rafiee, 2013). The difference in the bulk density value of
472 encapsulates can be related to difference in morphology of encapsulates, such as sphericity and
473 surface regularity of encapsulates. Spherical shaped particles with regular surface without any
474 dents or wrinkles can be more closely packed in a given volume, resulting in higher bulk density.
475 Accordingly, in the present study, the higher bulk density of SOE confirms the formation of
476 smooth and more regular encapsulates in SOE compared to that in FOE. Moreover, the higher
477 bulk density of SOE also suggests a better stability for SOE powder against oxidative
478 degradation compared to FOE powder. The higher the bulk density, the lesser the presence of
479
23
occluded air within the encapsulates and therefore, lesser the possibility of oxidative
480 deterioration of the core oil phase.
481 3.2.4.3. Hygroscopicity 482 Spray-dried particles can easily absorb water from the surrounding environment,
483 developing stickiness and caking during storage, unless necessary precautions are taken. Among
484 the two encapsulates, FOE showed a lower hygroscopicity value as compared to SOE (Table 1).
485 This might be due to the high surface oil content in the FOE as compared to the SOE. The
486 presence of non-encapsulated oil on the particle surface forms a hydrophobic layer thus reducing
487 the water absorption by encapsulates.
488 3.2.4.4. Colour 489 Accumulation of non-volatile decomposition products such as oxidized triacylglycerols
490 and free fatty acids during oxidation can lead to colour changes which indicate the extent of oil
491 deterioration in high fat containing foods. The colour of freshly encapsulated FOE was slightly
492 off-white in colour mainly imparted by gum arabic which is light brown in colour. Similarly,
493 freshly encapsulated SOE was slightly darker in colour due to the presence of sage extract which
494 is originally pale yellow-green in colour. There was distinct difference in the colour parameters
495 of FOE and SOE during different days of accelerated storage study (Table 1). Both the
496 encapsulated samples showed reduction in L* values during storage, however reduction was
497 minimum in SOE compared to FOE throughout the accelerated storage study at 60ºC.
498 The chroma parameters such as a* and b* values of all the samples showed marked
499 changes during storage. Both the freshly encapsulated powders showed positive a* values
500 indicating the nearness towards red chroma with significantly higher value for FOE sample. An
501
24
increase in redness of the fat containing formulation indicates the presence of fat oxidation
502 products. The b* values of all the samples showed positive values indicating the yellow chroma,
503 the intensity of which increased during storage. The higher yellowness values exhibited by FOE
504 during storage indicates the rapid oxidation of surface oil which yielded coloured secondary and
505 tertiary oxidation products.
506 3.2.5. In vitro release pattern of encapsulates in buffered saline
507 The in vitro release pattern of encapsulates in buffered saline containing alcohol
508 simulates the fate of encapsulates when incorporated in to a food system, especially when it
509 contains a nonpolar component like oil (emulsion) or protein precipitating agents like alcohol
510 (fermented products). The SOE showed a constant rate of oil release, which was less than 5% in
511 buffered saline for samples drawn at every I h interval, compared to 11-14% in FOE after first
512 hour of incubation (Figure 4), indicating the formation of wall matrices capable of conferring
513 better protection to SOE against surface disintegration and diffusion of oil into the food system.
514 The cumulative oil release over 4 h of incubation accounted 55.22% of total oil in FOE
515 compared to 19.74% in SOE. As there was a huge difference in the surface oil content of FOE
516 with the cumulative oil content in the buffered saline, the significantly higher rate of oil released
517 in FOE might be primarily related to the poor structural integrity of FOE encapsulates, resulting
518 in rapid diffusion of oil into the medium. Conversely in SOE, crosslinking reduced the mobility
519 of the polymer structure and enhanced its water resistance. Both the encapsulates showed
520 maximum rate of oil release during the initial 1 h followed by a linear pattern of release with
521 time. These results are consistent with the structural inferences derived from SEM and AFM
522 images.
523
25
3.2.6. In vitro gastro-intestinal digestion profile of microencapsulates
524 3.2.6.1. Effect of SGF and GIF on oil release 525
In order to assess the efficacy of microencapsulated fish oil for oral delivery or as a supplement
526 for targeted delivery of bioactives to specific parts of the gastro-intestinal tract (GIT), it is
527 imperative to test the stability and release behaviour of microencapsulates during GIT transit in
528 vitro before proceeding to in vivo trials (Kosaraju et al., 2006). Moreover, protein-polyphenol
529 interaction can also influence the digestion of proteins with the enzymes of the gastrointestinal
530 tract. The oil released after exposure to simulated gastric fluid was 71.71% and 64.26% of total
531 oil for FOE and SOE samples, respectively (Table 1). The release of oil from encapsulates
532 depends mainly on the extent of surface disintegration of wall matrix by acids and enzymes
533 present in the gastric fluids. The higher quantity of oil released from FOE might be attributed to
534 the presence of a comparatively higher number of broken encapsulates that allowed easy access
535 of enzymes and faster capsule degradation. For the same reason, the strong polymer matrix
536 formed in SOE encapsulates retarded the release of oil from microcapsules even in the gastric
537 fluid. On further exposure to simulated intestinal fluid, 15.41% of FOE and 14.98% of total oil of
538 SOE were released (Table 1). It is apparent that casein might have undergone structural
539 rearrangement under the low pH environment of gastric fluid containing the protease pepsin and
540 released most of the oil. However, the oil entrapped in the disintegrated/partially digested
541 polymer network is expected to release further at the intestine by the action of pancreatin, which
542 contains amylase to disintegrate gum arabic, and protease to hydrolyse casein. In-vivo human
543 clinical studies are required to understand the actual bioavailability of the oil from any designed
544 formulation. Also, release of oil to certain extent is expected to occur at colon by colonic fluid as
545 well as by the bacteria normally inhabiting the colon. In the present study, colonic release from
546
26
encapsulates has not been studied, as more than 80% of the oil loaded in encapsulates were
547 released in the GI tract itself. Moreover, the colon is one of the most difficult parts of the GI tract
548 to simulate in the laboratory as there exists a wide variation in the bacterial populations and it
549 requires cultivation of colonic bacteria under anaerobic conditions. Hence, many researchers
550 prefer to directly carry out tests on animal models (McClements, 2014).
551 3.2.6.2. Microscopic analysis of gastrointestinal digests
552 The encapsulates before and after digestion were observed using an inverted microscope
553 to confirm the inferences on the fate of the encapsulates in gastrointestinal tract. Both the
554 encapsulates showed aggregation and flocculate formation in the simulated fluids before
555 undergoing digestion (Figure 5 A&D). The structure of both encapsulates were destroyed
556 following in vitro digestion. After passing through the stomach phase, a reduction in droplet size
557 was observed due to hydrolysis of casinate-gum complex and weakening of electrostatic
558 repulsion (Figure 5 B&E). The digestion profiles revealed lower rate of aggregate formation in
559 gastric fluid for FOE (Figure 5 B) compared to SOE (Figure 5E), indicating the intense rupture
560 of capsules in FOE. However, as the digestion proceeded through the intestinal phase, both the
561 digest appeared more or less similar (Figure 5 C&F). The results suggest that fish oil
562 encapsulated within cross-linked microspheres would be digested more or less similar to those in
563 conventional fish oil encapsulates.
564 3.2.7. Measurement of lipid oxidation in microencapsulates
565 3.2.7.1. Measurement of PV 566 Inorder to ascertain the protective effect of sage extract against fat oxidation, the oxidative
567 pattern of encapsulates were compared against pure fish oil used for encapsulation (PFO) and
568
27
fish oil containing 1% sage extract (SFO). The protective effect of sage extract in standard
569 antioxidant in vitro assays was reported in our earlier report (Binsi et al., 2016). The PV of pure
570 fish oil was 4.56 meq of O2/kg oil, which increased considerably to 10.89 and 12.32 meq of
571 O2/kg oil in FOE and SOE, respectively (Figure 6A). This clearly shows that during spray
572 drying, oil particles underwent certain extent of oxidation before getting encapsulated, and
573 addition of antioxidant could not impart protection during this stage. However, the protective
574 effect of encapsulation was clearly visible during storage, as PFO showed more than ten-fold
575 increase in peroxide content after 24 h of storage. SFO also followed similar oxidative pattern of
576 PFO, however showed lower PV than PFO during storage. Both FOE and SOE samples showed
577 a progressive increase up to the 5thday with lower absolute values for SOE at any day of
578 sampling. Previously, sage extract at 0.02% was reported to be highly effective in protecting the
579 stability of rapeseed oil during accelerated oxidation storage conditions and was comparable
580 with that of butylated hydroxytoluene (BHT) at the same concentration (Bandoniene, Pukalskas,
581 Venskutonis, & Gruzdien., 2000). However, the results of the present study suggests a physical
582 mechanism of protection rather than the expected radical scavenging activities of polyphenols, as
583 the protection was more visible during storage rather than during spray drying. It is logical to
584 infer from the results of other parameters that the wall of SOE was more dense and protective
585 than that of FOE, which prevented further diffusion of bulk oil from the core to the surface apart
586 from offering barrier property against oxygen from surface to the core.
587 3.2.7.2. Measurement of Thiobarbituric acid reactive substances (TBARS)
588 The pure fish oil and encapsulated samples showed significant variation (p < 0.05) in TBARS
589 values under accelerated storage (Figure 6B). Both the encapsulates presented comparable initial
590
28
TBARS values of 1.03 and 0.92 for FOE and SOE respectively, whereas that of PFO and SFO
591 were significantly lower (p < 0.05). Homogenization of the emulsion involves intense
592 mechanical stress and turbulences, which on further exposure to high temperature during spray
593 drying leads to rapid oil oxidation. The TBARS value crossed the limit of 2 after 24 h of storage
594 in all the samples and showed almost similar values during the initial two days of storage. The
595 initial high values showed by all the samples may be related to the generation of secondary
596 oxidation products from the already existing peroxides in the samples. However, during further
597 storage the lowest rate of increase in TBARS value was observed in SOE sample, which was
598 even lower than SFO. Regarding the use of antioxidants in spray drying, Serfert, Drusch, &
599 Schwarz (2009) have inferred that antioxidants efficient in stabilizing liquid systems do not
600 necessarily increase the stability of oil encapsulated by spray-drying. However, the results of
601 present study shows that the protective effect can be fortified by incorporating antioxidants in the
602 oil, but the differences between oxidation rates might be related to the formation of an efficient
603 physical barrier at the interface. In other words, the enhancement of the stability is due to the
604 formation of a glassy state crust of sodium caseinate and gum arabic reducing the molecular
605 mobility of oxygen and thus slowing down the rate of lipid oxidation.
606 4. Conclusion 607 The purpose of the present study was to increase the structural and oxidative stability of spray
608 dried fish oil encapsulates by incorporating polyphenol rich sage extract as protein cross-linker
609 and as antioxidant to fish oil so as to minimize the effect of high temperature during spray
610 drying. FT-IR profile of the encapsulates confirmed the interaction of polyphenols with wall
611 polymers. Incorporation of sage extract improved the surface morphology and size uniformity of
612
29
encapsulates. Higher encapsulation efficiency and lower surface oil content in the encapsulated
613 products was achieved by incorporating sage extract at 1% level in the emulsion prior to spray
614 drying. The results of lipid oxidation parameters during accelerated storage study suggested a
615 physical mechanism of protection. Eventhough the total oil released was higher for FOE sample
616 during in vitro studies, the extent of protection offered by the sage extract against fat oxidation
617 and stability in normal food environment was significant. Hence, current findings shows that
618 incorporation of sage extract in fish oil before spray drying may be advocated. However, the
619 complete pattern of fish oil release in gastrointestinal tract could not be satisfactorily elucidated.
620 Further studies are required to enhance the release properties of encapsulates at targeted sites of
621 gastrointestinal tract.
622 Acknowledgment 623 This work was carried out under the research project funded by Science and Engineering
624 Research Board SERB (Department of Science and Technology, India – Ministry of Food
625 Processing Industries, India).
626 References 627 Aghbashlo, M., Mobli, H., Madadlou, A., & Rafiee, S. (2013). Influence of wall material and
628 inlet drying air temperature on the microencapsulation of fish oil by spray drying. Food
629 Bioprocess Technology, 66, 1561-1569.
630 AOAC, (2000). Official methods of analysis of the association of official analytical chemists
631 17th ed. Washington, DC.
632
30
Bandoniene, D., Pukalskas, A., Venskutonis, P. R., & Gruzdien, D., 2000. Preliminary
633 screening of antioxidant activity of some plant extracts in rapeseed oil. Food Research
634 International, 339, 785-791 635 Binsi, P. K., Ravishankar, C. N., Mathew, S. S., Athira, V. P., Biji, K. B., Joshy, C. G., &
636 Srinivasa Gopal, T. K. (2016). Comparative Evaluation of Natural Antioxidant Coating and
637 Vacuum-Packaging Technique on Quality Characteristics of Chill Stored Giant Trevally
638 Fillets. Journal of Aquatic Food Product Technology, 25(2), 197-209.
639 Chen, Q., McGillivray, D., Wen, J., Zhong, F., & Quek, S. Y. (2013). Co-encapsulation of
640 fish oil with phytosterol esters and limonene by milk proteins. Journal of Food
641 Engineering, 117(4), 505-512.
642 Cho, Y. H., Shim, H. K., Park, J., 2003. Encapsulation of fish oil by an enzymatic gelation
643 process using transglutaminase cross-linked proteins. Journal of Food Science, 689, 2717- 644
2723.
645 Dieffenbacher, A., & Lüthi, B. (1986). Die direktekolorimetrischebestimmung der
646 peroxidzahl POZ in milchprodukten. MitteilungenausdemGebiete der Lebensmittel- 647 untersuchungun Hygiene. 77, 544-553.
648 Dorman, H. J. D., Peltoketo, A., Hiltunen, R., & Tikkanen, M. J. (2003). Characterisation of
649 the antioxidant properties of de-odourised aqueous extracts from selected Lamiaceae
650 herbs. Food Chemistry, 832, 255-262.
651 Endo, J., & Arita, M. (2016). Cardioprotective mechanism of omega-3 polyunsaturated fatty
652 acids. Journal of cardiology, 67(1), 22-27.
653
31
Frazier, R. A., Deaville, E. R., Green, R. J., Stringano, E., Willoughby, I., Plant, J., &
654 Mueller-Harvey, I. (2010). Interactions of tea tannins and condensed tannins with
655 proteins. Journal of pharmaceutical and biomedical analysis, 51(2), 490-495.
656 Freund Levi, Y., Vedin, I., Cederholm, T., Basun, H., Faxén Irving, G., Eriksdotter, M.,
657 Hjorth, E., Schultzberg, M., Vessby, B., Wahlund, L.O., & Salem, N., (2014). Transfer of
658 omega‐3 fatty acids across the blood–brain barrier after dietary supplementation with a
659 docosahexaenoic acid‐rich omega‐3 fatty acid preparation in patients with Alzheimer's
660 disease: the OmegAD study. Journal of internal medicine,275(4), 428-436
661 Gharsallaoui, A., Roudaut, G., Chambin, O., Voilley, A., & Saurel, R. (2007). Applications
662 of spray-drying in microencapsulation of food ingredients: An overview. Food Research
663 International, 409, 1107-1121.
664 Hasni, I., Bourassa, P., Hamdani, S., Samson, G., Carpentier, R., & Tajmir-Riahi, H. A.
665 (2011). Interaction of milk α-and β-caseins with tea polyphenols. Food Chemistry, 126(2),
666 630-639.
667 Hosseini, S. F., Zandi, M., Rezaei, M., & Farahmandghavi, F. (2013). Two-step method for
668 encapsulation of oregano essential oil in chitosan nanoparticles: preparation, characterization
669 and in vitro release study. Carbohydrate Polymers, 951, 50-56.
670 Jeyakumari, A., Kathari, D.C., & Venkateshwarlu, G. (2014). Microencapsulation of fish oil- 671 milk based emulsion by spray drying: Impact on Oxidative Stability. Fishery Technology, 51
672 (1), 31-37 673
32
Jiménez-Martín, E., Gharsallaoui, A., Pérez-Palacios, T., Carrascal, J. R., & Rojas, T. A.
674 (2015). Volatile compounds and physicochemical characteristics during storage of
675 microcapsules from different fish oil emulsions. Food and Bioproducts Processing, 96, 52- 676
64.
677 Koh, L. L. A., Chandrapala, J., Zisu, B., Martin, G. J. O., Kentish, S. E., & Ashokkumar,
678 M. (2014). A comparison of the effectiveness of sonication, high shear mixing and
679 homogenization on improving the heat stability of whey protein solutions. Food and
680 Bioprocess Technology, 7(2), 556–566.
681 Komaiko, J., Sastrosubroto, A., & McClements, D. J. (2016). Encapsulation of ω-3 fatty
682 acids in nanoemulsion-based delivery systems fabricated from natural emulsifiers: Sunflower
683 phospholipids. Food chemistry, 203, 331-339.
684 Kosaraju, S. L., D'ath, L., & Lawrence, A. (2006). Preparation and characterisation of
685 chitosan microspheres for antioxidant delivery. Carbohydrate Polymers, 642, 163-167.
686 Koupantsis, T., Pavlidou, E., & Paraskevopoulou A. (2016). Glycerol and tannic acid as
687 applied in the preparation of milk proteins – CMC complex coavervates for flavour
688 encapsulation. Food Hydrocolloids, 57, 62–71.
689 Liang, R., Shoemaker, C. F., Yang, X., Zhong, F., & Huang, Q. (2013). Stability and
690 bioaccessibility of β-carotene in nanoemulsions stabilized by modified starches. Journal of
691 agricultural and food chemistry, 61(6), 1249-1257.
692 McClements, D. J. (2014). Nanoparticle-and microparticle-based delivery systems:
693 Encapsulation, protection and release of active compounds. CRC Press.
694
33
McDonald, R. E., & Hultin, H. O. (1987). Some characteristics of the enzymic lipid
695 peroxidation system in the microsomal fraction of flounder skeletal muscle. Journal of Food
696 Science, 521, 15-21.
697 Patten, G. S., Augustin, M. A., Sanguansri, L., Head, R. J., & Abeywardena, M. Y. (2009).
698 Site specific delivery of microencapsulated fish oil to the gastrointestinal tract of the
699 rat. Digestive Diseases Science, 543, 511-521.
700 Prodpran, T., Benjakul, S., & Phatcharat, S. (2012). Effect of phenolic compounds on protein
701 cross-linking and properties of film from fish myofibrillar protein. International journal of
702 biological macromolecules, 51(5), 774-782.
703 Sari, T. P., Mann, B., Kumar, R., Singh, R. R. B., Sharma, R., Bhardwaj, M., & Athira, S.
704 (2015). Preparation and characterization of nanoemulsion encapsulating curcumin. Food
705 Hydrocolloids, 43, 540-546.
706 Serfert, Y., Drusch, S., & Schwarz, K. (2009). Chemical stabilisation of oils rich in long- 707 chain polyunsaturated fatty acids during homogenisation, microencapsulation and storage.
708 Food Chemistry, 1134, 1106-1112.
709 Shantha, N.C., & Decker, E.A. (1994). Rapid, sensitive, iron-based spectrophotometric
710 methods for determination of peroxide value of food lipid. Journal of AOAC. International,
711 772, 421–424.
712 Soottitantawat, A., Bigeard, F., Yoshii, H., Furuta, T., Ohkawara, M., Linko, P. 2005.
713 Influence of emulsion and powder size on the stability of encapsulated D-limonene by spray
714 drying. Innovative Food Science and Emerging Technologies, 61, 107-114.
715
34
US Pharmacopeia. (2000). National Formulary USP 24 NF 19, Rockville, MD
716 Velasco, J., Dobarganes, C., & Marquez-Ruiz, G. (2003). Variables affecting lipid oxidation
717 in dried microencapsulated oils. Grasas y Aceites, 543, 304-314.
718 Yu, M., & Ivanisevic, A. (2004). Encapsulated cells: an atomic force microscopy
719 study. Biomaterials, 2517, 3655-3662.
720 Wang, Y., Liu, W., Chen, X. D., & Selomulya, C. (2016). Micro-encapsulation and
721 stabilization of DHA containing fish oil in protein-based emulsion through mono-disperse
722 droplet spray dryer. Journal of Food Engineering, 175, 74-84.
723
724
725 726
35
LEGEND TO FIGURES 727
728
729
730 Figure 1.
A-B
Microscopic images of fish oil emulsions Inverted microscopic image: (A) FOE (B) SOE
Atomic force microscopic image: (C) FOE (D) SOE
Figure 2.
FT-IR profile of fish oil encapsulates
Figure 3.
A-D
SEM images of fish oil encapsulates (A) FOE and (B) SOE
AFM images of fish oil encapsulates (C) FOE and (D) SOE
Figure 4.
A-B
In vitro release profile of fish oil encapsulates
Figure 5 A-F
Optical microscopic images of gastro-intestinal digests of fish oil encapsulates (A) FOE before digestion (B) FOE after gastric digestion (C) FOE after gastric and intestinal digestion (D) SOE before digestion (E) SOE after gastric digestion (F) SOE after gastric and intestinal digestion
Figure 6 A-B Changes in PV and TBARS values of fish oil encapsulates and fish oil during accelerated storage study
FOE SOE
FOE SOE PFO SFO FOE SOE
Figure 1
A B C D
1509 3195 1744 2969 2966 1508 Figure 2 1745 Absorbance (%)
Wave number (cm-1) 3604 3585 1607 1606 FOE SOE
0 500 1000 1500 2000 2500 3000 3500 4000 3501
A B Figure 3 C RMS = 3.41; Ra= 2.87;Rsk =-0.84 D RMS = 5.15;Ra = 4.43; Rsk = -0.24
0 3 6 9 12 15 0 1 2 3 4 Cumulative release of fish oil (%)
Time (hr) FOE SOE Figure 4
C B A D E F Figure: 5
0 40 80 120 160 200 A 0 10 20 30 0 2 4 6 8 Storage period (Days)
PV (mEq O2/kg oil ) TBARS value (mg malonaldehyde/kg oil)
FOE SOE PFO SFO
Figure 6
42
881 Parameter FOE SOE Emulsion stability index (%)
96.66a* ±0.89 98.67 b±0.72 Microencapsulation efficiency (%)
68.99 a ±1.08 73.21 b ±1.10 Surface oil ( g/100 of dry powder )
3.87 a ±0.05 3.40 b ±0.06 Total oil (g/100 of dry powder)
12.50 a ±0.04 12.72 a ±0.03 Percentage loss during spray drying (as
% of loaded oil in emulsion) 33.57 a ±0.05 33.04 a ±0.09
Bulk density(g/ml) 0.39 a ±0.06 0.42 a ±0.11 Hygroscopicity (g/100g)
7.70 a ±0.15 8.39 b ±0.21 Moisture content (%) 5.91 a ±0.06
5.01 b ±0.04 Oil released by gastric fluid (% of total oil)
71.71 a ±0.88 64.26 b ±0.83 Oil released by intestinal fluid (% of total oil)
15.41 a ±0.21 14.98 b ±0.22 Total oil released (as % of total oil)
87.12 a ±0.93 79.24 b ±0.75 Unreleased oil (as % of total oil)
12.88 a ±0.16 20.76 b ±0.14 Storage Day Hunter-lab colour parameters of fish oil encapsulates during accelerated storage
FOE SOE 0 82.76±0.98 1.37±0.10 18.97±0.78 82.7±0.95
0.73±0.11 16.32±0.28 1 82.61±0.91 1.74±0.12 22.48±0.41
81.82±0.82 1.12±0.08 20.19±0.31 2 81.55±0.81 2.26±0.11
24.16±0.53 81.24±0.85 1.45±0.09 21.57±0.58 3 82.38±0.88 2.02±0.08
23.84±0.84 82.18±0.88 1.37±0.11 21.98±0.32 4 82.35±0.92 1.97±0.12
23.74±0.79 81.78±0.90 1.41±0.12 22.14±0.40 Table 1: Physicochemical and oil release properties of fish oil encapsulates
43
5 82.45±0.73 2.24±0.16 24.62±0.58 81.49±0.91 1.71±0.08
23.09±0.33 6 79.83±0.58 2.54±0.18 24.97±0.62 80.41±0.85
1.93±0.08 23.49±0.28 7 78.01±0.43 2.17±0.16 24.23±0.73
80.58±0.88 1.73±0.07 23.36±0.43
882
883
884 885 Values in parenthesis represents standard deviation for n=3;
*Treatment mean values with same letters are not significantly different from each other (p < 0.05).
44
Highlights 886 • Sage polyphenols along with gum arabic strengthened shell matrix
887 • Sage polyphenols improved encapsulation efficiency
888 • The stabilised structure reduced oil oxidation
889 • In vitro release kinetics confirmed controlled release of oil in modified encapsulates.
890
891