Effects of Blueberry (Vaccinium corymbosum) Juice on Lipid Oxidation During Spray Drying of Microencapsulated Menhaden Oil

Abstract

This study investigated the effect of blueberry juice on menhaden oil lipid oxidation during microencapsulation. Oil in water emulsions containing menhaden oil with 0, 5, or 10% blueberry juice were spray dried to produce control-M, 5% BJ-M, and 10% BJ-M microencapsulated powders, respectively. All microencapsulated powders had similar encapsulation efficiencies with low surface oil content. Peroxide value (meq/kg of oil) was 4.50, 4.31, and 3.38 for control-M, 5% BJ-M and 10% BJ-M, respectively. Ten percent BJ-M had lower (P < 0.05) anisidine value, and totox values than 5% BJ-M and control-M. This indicated that 10% blueberry juice reduced lipid oxidation in menhaden oil during microencapsulation to a greater extent than the other formulations.

Keywords:

• Microencapsulation
• Menhaden oil
• Spray drying
• Blueberry juice
• Lipid oxidation

INTRODUCTION

Blueberries (Vaccinium corymbosum) are considered to be a good source of phenolic compounds.^[1] The polyphenols and anthocyanins present in blueberries are known to have a positive role in human health and in disease prevention.^[1,^2] Blueberries have received much attention due to their high antioxidant activity compared to many other fruits and vegetables.^[1,^3] Antioxidant compounds, such as phenolic acids, polyphenols, and flavonoids, present in berries are able to scavenge free radicals such as peroxide, hydro-peroxide, and lipid peroxide, and thus inhibit oxidative mechanisms that lead to lipid oxidation.^[2]

Microencapsulation is defined as a process by which small amounts of core material such as fish oil and flavors are enveloped within a wall material to form micro size capsules.^[4,^5] Although many methods are available for microencapsulation, spray drying is the most common technique employed to produce microcapsules for food related applications because equipment is readily available for producing high quality microcapsule powders.^[4] Spray drying is an economical and versatile process for producing microencapsulated powders.

Successful microencapsulation of fish oils result preventing direct contact of oil with oxygen or metal ions, avoiding direct light exposure and in the trapping of off flavors.^[6] Although spray drying is the most popular microencapsulation technique used in the food industry, it is known to induce lipid oxidation and produce powders with porous structure.^[5] Also, low spray drying temperatures are advantageous to maintain fish oil quality, but microencapsulation requires temperatures as high as 170°C to produce a product with low (1–2%) moisture content.^[6] Research related to incorporation of natural antioxidants from berries into fish oil emulsions is scant. Blueberry antioxidants might reduce the amount of lipid oxidation occurring during emulsification (ultra-sonication) and spray drying; therefore, providing a greater storage stability of microencapsulated powders (even with a high load of core material). Tocopherol, a lipophilic antioxidant, and rosemary extract (a hydrophilic antioxidant) are commonly used as natural antioxidants in food products. It has been reported that tocopherol is more effective in reducing oxidization in oil-in-water emulsions than rosemary extract.^[7,^8] Blueberry extract consists of both lipophilic and hydrophilic antioxidants, which may reduce lipid oxidation during preparation and drying of the emulsion. The blueberry juice (BJ) antioxidants scavenge free radicals in fish oils, and also can chelate pro-oxidants such as metal ions.^[9] This could result in reduced lipid oxidation and increased fish oil stability during emulsification, spray drying, and subsequent storage. Hence, the objectives of this study were, (1) to determine the effect of BJ addition on lipid oxidation during emulsification and spray drying of microencapsulated menhaden (Brevoortia patronus) oil, and (2) to evaluate the emulsion and powder properties.

MATERIALS AND METHODS

Preparation of BJ

Three batches of locally obtained fresh high bush blueberries (Vaccinium corymbosum L.) were steam blanched using a steamer (Welco, Minneapolis, MN, USA) for 2 min and then rapidly cooled down to 10°C using an ice water bath. Afterwards, the blueberries were blended using a mechanical blender (Magic bullet, Mainland, China) for 1 min and the resulting mash was centrifuged at 2037.6 × g using a centrifuge (J2-HC-TB-002 model, Beckman instruments Inc., CA, USA) for 20 min at 4°C.

Physicochemical Properties of BJ

The yield (%) in BJ was calculated as: (weight of juice/initial blueberry weight)×100. Total monomeric anthocyanin content (TAC) of the BJ was determined by the pH differential method of AOAC 2005.02.^[10] The results were expressed as mg of cyanidin-3-glucoside equivalents per liter of BJ. Total phenol content (TPC) of BJ was determined according to the Folin-ciocalteau method.^[11] The results were expressed as milligrams of gallic acid equivalent (GAE) per liter of BJ. The anti-scavenging activity (AA) of BJ was determined by employing the DPPH radical scavenging assay (1,1-Diphenyl-2-picryl-hydrazyl) according to the method of Locatelli et al.^[12] The results were expressed as μmol of trolox equivalent (TE) per 100 mL of BJ.

The pH of BJ was measured using a bench top pH meter (Symphony, VWR Scientific, USA). The titratable acidity was determined using a titration method, and results were expressed as percent total organic acid on the basis of citric acid which is the predominant acid present in blueberries.^[13] The Brix of BJ was measured at 20°C using a hand held digital refractometer (Model AR 200, Reichart Analytical Instruments, USA). Color of the BJ was measured using a Lab Scan XE Colorimeter (Hunter Associates Laboratory, Inc. Resbon, VA) and the results were reported in CIE Lab color scales (L* value is the degree of lightness to darkness, a* value is the degree of redness to greenness, and b* value is the degree of yellowness to blueness). Chroma and hue angle values were calculated using Eqs. (1) and (2), respectively.

(1)

(2)

Physicochemical Properties of Fish Oil

Purified menhaden (Brevoortia patronus) oil (without added antioxidants) was provided by OmegaPure^TM (Omega Protein Inc., Houston, TX, USA). The menhaden oil (MO) was analyzed for peroxide value (PV) according to the AOCS official method cd 8–53.^[14] PV was calculated and reported as milliequivalents of peroxide per 1 kg of MO. The anisidine value (AV), which represents secondary lipid oxidation products in oils, was determined by using the AOCS official method cd 18–90.^[14] The totox values of the oils were calculated as described by AOCS cg 3-91,^[14] as totox = AV + 2PV. Totox provides the overall oxidation state of the oils. The color of MO was measured using a Lab Scan XE Colorimeter (Hunter Associates Laboratory, Inc. Resbon, VA, USA) and the results were reported in CIE Lab color scales (L*, a*, and b* values). The chroma and the hue angle were calculated as described previously.

Preparation of Emulsions

Oil in water emulsions of 500 mL was prepared with 20% (w/w) MO containing 0, 5, or 10% of BJ by weight. These emulsions were designated as control-E, 5% BJ-E and 10% BJ-E, respectively. All of the emulsions were prepared using 5% (w/w) octenyl succinic anhydride (OSA) modified starch and 10% (w/w) corn syrup of dextrose equivalent of 36/43 (ADM company, Decateur, IL, USA) as wall materials. The wall materials for microencapsulation were selected based on preliminary emulsion stability experiments. Wall materials were dissolved in water and then BJ was added to the solutions. Then, MO was added to the wall material solution and the mixture was stirred before further emulsification using an ultrasonic processor (Model CPX 500, Cole Palmer Instruments, Vernon Hills, IL, USA) having a 20 KHz probe. Emulsification was done at the pulse rate of 2 to 1 for 5 min. The amplitude level was set to 80% for sonicating the emulsions. During the emulsification process the temperatures of the sample containers were kept at 20 ± 1°C using an ice bath. All of the emulsions were prepared in triplicates.

Emulsion Stability, pH, and Color

The stability of the emulsions was evaluated according to a method described by Min et al.^[15] The oil recovery percentage was calculated using Eq. (3).^[16] And the emulsion stability was determined using Eq. (4).

(3)

(4)

The pH of the emulsions was measured using a bench top pH meter (Symphony, VWR Scientific, USA). Emulsion color was measured with a Lab Scan XE Colorimeter (Hunter Associates Laboratory, Inc., VA, USA) and was reported in CIE Lab color scales (L*, a*, and b* values).

Emulsion Flow Properties

Emulsion flow properties were determined using an AR 2000ex Rheometer (TA Instruments, New Castle, DE, USA). The rheometer was fitted with a plate geometry employing 40 mm diameter plates and a 400 μm sample gap. An emulsion sample was placed on the temperature controlled plate and allowed to equilibrate at temperatures of 5, 15, and 25°C. The shear stress of the emulsions was measured at shear rates of 1 to 200 s⁻¹. The emulsion flow properties were analyzed using the power law (Eq. 5) model where σ represents shear stress (Pa); symbolizes shear rate (s⁻¹); k represents consistency index (Pa.sⁿ), and n symbolizes flow behavior index.

(5)

Preparation of Spray Dried Microcapsules

The emulsions were transformed into microcapsules using a pilot plant spray drier (FT 80 Tall Form Spray Dryer, Armfiled Inc., Jackson, NJ, USA) employing a co-current drying procedure. The operational conditions were: feed temperature, 30°C, inlet air temperature, 170°C, and nozzle air gauge pressure, of 70 kPa. The control-E, 5%BJ-E and 10% BJ-E emulsions were spray dried to produce microencapsulated powders of control-M, 5% BJ-M and 10% BJ-M, respectively. The dried powders were collected and stored in amber bottles at 4°C for further analysis.

Lipid Oxidation of Emulsions and Microcapsules

The oil was extracted from emulsions and microcapsules according to a method described by Wanasundara and Shahidi.^[17] The oil recovered from the emulsions and microcapsules were analyzed for PV, AV, and Totox as previously described.

Fatty Acid Methyl Esters (FAMEs) Profile Analysis

FAMEs were determined for MO and oil extracted from the microcapsules according to the method described by Maxwell and Marmer.^[18] The fatty acid methyl esters were analyzed using a gas chromatograph (model 7890A Agilent) coupled with a flame ionization detector (Agilent Technologies Inc, Santa Clara, CA, USA). The peaks were identified using retention times of standards: Supelco 37, PUFA #1, PUFA #3, and cod liver oil from Supelco (Bellefonte, PA, USA). Data was expressed as percent of total integrated area.

Microencapsulation Efficiency (ME)

ME designates the amount of core material encapsulated within powder particles. The ME of spray dried powders was calculated using surface oil content and total oil content according to Eq. (6).

(6)

Total Oil Content

Surface oil content refers to the free or extractable oil around microcapsules. The surface oil content of microencapsulated MO powders was determined using petroleum ether as the extraction solvent and a method described by Jafariet al.^[4] The surface oil content was measured based on the weight difference between the initial clean flask and the flask with the extracted oil residue. The results were expressed as amount of surface oil content per gram of microcapsules. The total oil content of the microcapsules which include both encapsulated and surface oil was determined as described by Kosaraju et al.^[19] The total oil content was calculated by weight difference and expressed as the oil content per gram of microcapsules.

Color and Morphology of Microencapsulated MO Powder

The color of the microcapsules was measured using a Lab Scan XE Colorimeter (Hunter Associates Laboratory, Inc., Resbon, VA, USA) and was reported in CIE Lab color scales (L*, a*, and b* values). The morphology of microcapsules was observed using a scanning electron microscope (SEM) using an acceleration voltage of 5 kV. Microcapsules were mounted onto an aluminum SEM specimen stub using a double sided sticky tape, and coated with gold: palladium (60:40) using S150 sputter coater (Edwards High Vacuum International, Wilmington, MA, USA) to make the samples conductive. The coated stubs were examined and imaged with SEM.

Particle Size Distribution

Particle size of the microcapsules was measured using a Microtrac S3500 light scattering system (MicroTrac, Largo, FL, USA). The system consists of three fixed 780 nm solid state lenses with a computerized single lens alignment. The measurement capacity was from 0.24 to 2800 mm diameter. The distribution of particle size was calculated using light scattering pattern. Results were noted as 50th (median) and 90th percentiles of the microcapsules sizes.

Statistical Analysis

Mean values and standard deviations of triplicate experiments were reported. Analysis of variance (ANOVA) was carried out to determine the difference among treatment means (SAS Version 8.2, SAS Institute Inc., Cary, NC, USA) using the post hoc Tukey’s Studentized Range Test.

RESULTS AND DISCUSSION

Physicochemical Properties and Yield of BJ

Yield of BJ was 61% (w/w) of the berries. Blueberry seeds, mucilage, and pectin were contained in the residue. BJ contained 106 of TAC (mg/L cya-3-glu) and 3273 of TPC (mg/L GAEs; Table 1). Blueberries have higher anthocyanin and phenols content than many types of berries such as cranberries, black berries, and strawberries.^[1,^3] The BJ was dark reddish in color (a* = 32.11 and L* = 6.65±0.03) and was acidic with a pH value of 3.23 and titratable acidity of 0.83 (citric acid equivalents). Blueberries had high AA due to the presence of anthocyanins and polyphenols. Moreover, blueberries contain the antioxidant phenolic compounds myricetin, quercetin, chlorogenic acid, and kaempferol, of which chlorogenic acid contributes the most of the antioxidant activity.^[1]

TABLE 1 Physicochemical properties and yield (%) of BJ*

BJ yield (%)	61 ± 4.0
TAC^a	106 ± 10
TPC^b	3273 ± 22
AA^c	3021.70 ± 232.22
^0Brix^e	11.3 ± 1.76
pH^e	3.2 ± 0.3
Percent Titratable acidity^d	0.83 ± 0.03
Color L*	6.65 ± 0.03
a*	32.11 ± 0.04
b*	11.32 ± 0.05
Hue angle (degrees)	19.42 ± 1.33
Chroma	34.05 ± 0.06

^aExpressed in mg/L of cyanindin-3-glucoside equivalents;

^bExpressed in mg/L gallic acid equivalents;

^cExpressed in μmol/100 mL Trolox equivalents;

^dExpressed based on citric acid;

^eThe values are taken at 20°C;

*Values are means and SD of triplicate determinations. BJ = blueberry juice; TAC = total anthocyanin content; TPC = total phenol content; AA = anti-scavenging activity.

According to the US Department of Agriculture,^[20] blueberries ranked as number one in antioxidant activity based on ORAC assay (oxygen radical absorbing capacity) compared to 40 common fresh fruits and vegetables. Fresh blueberries contain 2400 ORAC units/100 grams,^[21] which is higher than other berries including grapes, strawberries, raspberries, and cranberries. Several methods are available to determine antioxidant activity; in this study DPPH (2, 2-diphenyl-1-picrylhydrazyl) assay was used to measure the AA of BJ. The DPPH assay mainly addresses the radical scavenging ability of an antioxidant. This method is appropriate to indicate antioxidant activity in a lipid system in which lipid oxidation is carried out by a series of free radical reactions.^[22] The value obtained (3022 μmol TE/100g; Table 1) was lower than the values reported in the previous study related to blueberry fruit based on the DPPH assay.^[23] Dragović-Uzelac et al.^[23] reported a total antioxidant activity (5630–7600 μmol TE/100g) based on DPPH assays of many blueberry cultivars.

Emulsion Stability, Color, Physicochemical, and Rheological Properties of Emulsions

ME is thought to be affected by emulsion stability since higher emulsion stability corresponds to higher efficiency.^[24] All of the prepared MO emulsions (with or without the addition of BJ) had >98% emulsion stability (Table 2). Emulsions with BJ added significantly increased stability over control-E. This may be due to the lipophilic and hydrophilic compounds present in the juice which might aid in further emulsion stabilization.^[25] The pH of the emulsions decreased with the juice addition (pH = 3.2; Table 1). Emulsion stability was not affected by the pH of the emulsions, indicating that the starch emulsifier (OSA starch) was not affected by the pH of the mixture. It is noted that emulsions stabilized using protein emulsifiers are sensitive to the pH of the solution. Blueberry extract significantly changed the emulsion color from off white to light pink. 10% BJ-E and 5% BJ-E had higher a* and lower L* and b* values than control-E.

TABLE 2 Emulsion stability, pH, and color of MO emulsions*

	Control-E	5% BJ-E	10% BJ-E
Percent emulsion stability	98.76 ± 0.36^b	99.93 ± 0.09^a	99.64 ± 0.33^a
Moisture content (%)	68.90 ± 0.11^a	66.47 ± 0.87^b	64.58 ± 1.10^c
pH	4.93 ± 0.02^a	4.55 ± 0.03^b	4.32 ± 0.02^c
Color L*	88.42 ± 0.62^a	84.92 ± 0.24^b	81.63 ± 0.18^c
a*	−2.80 ± 0.08^c	1.07 ± 0.02^b	2.68 ± 0.19^a
b*	2.74 ± 0.04^a	0.76 ± 0.17^b	0.53 ± 0.05^c
Hue angle (degrees)	139.55 ± 0.14° ^a	34.09 ± 0.57° ^b	19.57 ± 0.97° ^c
Chroma	3.78 ± 0.10^a	1.52 ± 0.02^c	3.36 ± 0.18^b

*Values are means and SD of triplicate determinations;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05); MO = menhaden oil; control-E = MO emulsion; 5% BJ-E = MO emulsion with 5% BJ; 10% BJ-E = MO emulsion with 10% BJ.

TABLE 3 Flow behavior properties of control-E, 5% BJ-E, and 10% BJ-E*

		Temperature (°C)
	TRT	5	15	25
n	Control-E	0.770 ± 0.018^cA	0.827 ± 0.024^bA	0.912 ± 0.005^aA
	5% BJ-E	0.695 ± 0.008^cB	0.717 ± 0.002^bC	0.763 ± 0.003^aC
	10% BJ-E	0.770 ± 0.015^cA	0.797 ± 0.002^bB	0.819 ± 0.003^aB
k (Pa.s^n)	Control-E	0.137 ± 0.009^aB	0.047 ± 0.005^bC	0.024 ± 0.001^cC
	5% BJ-E	0.264 ± 0.014^aA	0.183 ± 0.005^bA	0.118 ± 0.002^cA
	10% BJ-E	0.137 ± 0.007^aB	0.094 ± 0.002^bB	0.065 ± 0.003^cB
Apparent viscosity** (Pa.s)	Control-E	0.023 ± 0.002^aC	0.019 ± 0.000^bC	0.011 ± 0.001^cC
	5% BJ-E	0.040 ± 0.001^aB	0.032 ± 0.000^bB	0.014 ± 0.000^cB
	10% BJ-E	0.053 ± 0.000^aA	0.041 ± 0.001^bA	0.016 ± 0.000^cA

*Values are means and SD of triplicate determinations;

**Apparent viscosity reported at shear rate of 200 s⁻¹;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05); ^ABCMeans with different exponents in the same column indicate significant difference (p ≤ 0.05);

Control-E = MO emulsion; 5% BJ-E = MO emulsion with 5% BJ; 10% BJ-E = MO emulsion with 10% BJ;

n = flow index; k = consistency index.

The chroma value indicated strength of surface color which reflected the degree of visual difference from neutral grey.^[26] Chroma value of 5% BJ-M was lower than the other emulsions (Table 2). The 5% BJ-M emulsion color was closer to neutral grey than the other emulsions where control-E was more yellow in color and 10% BJ-E was more reddish in color. The hue angle value expresses the closest match with the spectral and non-spectral colors, in which, a sample with hue angle 0° is purplish red, 90° is yellow, 180° is bluish green, and 270° is blue.^[26] The hue angle value obtained for 5%BJ-E and 10%BJ-E was close to 0° because of purplish red color, while control-E had a hue angle value close to 90° because of its light yellow color. In most cases emulsion color is not a critical parameter since the powders are not individually consumed and the quantity of microencapsulated fish oil powders in a food product is low. The American Heart Association daily intake recommendation of EPA and DHA for coronary heart disease patients with high triglyceride levels, at levels of 1 g per day and 2–4 g/day, respectively.^[27] Although 10% BJ addition slightly changed the powder color, this amount would be less likely to cause an effect in practical applications.

Flow properties of emulsions are important in spray drying operations. Emulsion viscosity is important during microencapsulation since it has a direct effect on emulsion droplet size during atomization.^[16] More viscous emulsions produce larger droplet size resulting in powders with large particle size.^[28] Apparent viscosity of the emulsions decreased with increasing temperature (Table 3). The apparent viscosity at a shear rate of 200 s⁻¹ was higher in 10% BJ-E than 5% BJ-E and control-E samples. The flow indexes (n) of the emulsions control-E, 5% BJ-E, and 10% BJ-E were less than 1 at all of the temperatures used, indicating a pseudo plastic fluid behavior (Table 3). The flow index of all emulsions increased with increased temperature. Thus emulsions attempt to behave as Newtonian fluids at higher temperatures. Control-E had a higher flow index than either 5% BJ-E or 10% BJ-E at all temperatures other than 5°C. At 5°C the flow index of control-E and 10% BJ-E were equal. This may be due to the fact that 5% BJ-E and 10% BJ-E have lower moisture content compared to control-E. The percent moisture content of control-E, 5% BJ-M, and 10% BJ-M were 68.90, 66.47, and 64.58, respectively (Table 2).

Conversely, the emulsion consistency index (k) decreased with increased temperature in all the emulsions. The k value is related to the viscous nature of the emulsions, in that it increases with increased viscosity.^[16] Among the different emulsions, 5% BJ-E had a higher k value than 10% BJ-E. Control-E had the lowest k value at all temperatures other than 5°C (Table 3). At 5 °C, Control-E and 10% BJ-E had the same k value. A higher k value may be due to the solids present in the juice, but with 5% BJ-E having the highest k value, the data does not show any clear relationship between total solids with k value.

Lipid Oxidation During Emulsification and Microencapsulation

The PV, AV, and Totox values of MO, control-E, 5% BJ-E and 10% BJ-E after emulsifications are shown in Table 4. Compared to the initial MO, the emulsification process resulted in significantly increase of both PV and AV in the control-E and 5% BJ-E emulsions. The PV increased from an initial value of 2.10 meq/kg for MO to 2.78 and 2.56 for control-E and 5%BJ-E respectively. However, 10%BJ-E did not show a significant difference in PV and AV compared to MO (Table 4). The AV of these emulsions increased in a similar manner as the PV, which indicated that BJ, at 10% addition, effectively prevented oxidation occurring during the emulsification process using the ultra-sonication technique.

TABLE 4 PV, AV, Totox value of MO and MO emulsions*

	MO	Control-E	5% BJ-E	10% BJ-E
PV**	2.10 ± 0.02^c	2.78 ± 0.02^a	2.56 ± 0.01^b	2.15 ± 0.03^c
AV	5.16 ± 0.06^c	5.87 ± 0.09^a	5.45 ± 0.04^b	5.22 ± 0.11^c
Totox	9.36 ± 0.38^c	11.43 ± 0.13^a	10.57 ± 0.06^b	9.52 ± 0.17^c

*Values are means and SD of triplicate determinations;

**meq peroxide/kg oil;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05); MO = menhaden oil; control-E = MO emulsion; 5% BJ-E = MO emulsion with 5% BJ; 10% BJ-E = MO emulsion with 10% BJ; PV = peroxide value; p-AV = anisidine value.

The emulsification method used in the experiment, i.e., ultra-sonication, is known to cause lipid oxidation. Chemat et al.^[29] reported an increase in PV value from 5.38 meq/kg to 6.33 meq/kg in sonicated sunflower oil samples processed for 2 min (20 KHz, 150 W). The ultrasound assisted lipid oxidation is attributed to the cavitation phenomenon, which affects structural and functional components leading to lipid oxidation.^[29] The collapse of cavitation bubbles in the emulsions resulted in formation of hot spots having high temperatures and pressure of 500 MPa. These hot spots resulted in lipid oxidation caused by various mechanisms such as thermal effects, and from free radicals generated during sonication.

Baik et al.^[30] reported an increase of PV of fish oil emulsions from 9.5 ± 0.4 mmol/Kg to 17 ± 4.0 mmol/kg in spray dried powders. Drusch et al.^[31] explained that during the spray drying process, fish oil exposed to air, high temperature, and high pressure lead to increased lipid oxidation. Similarly, our experimental results showed an increase in PV and AV in microencapsulated MO powders compared to pure oil as shown in Fig. 1. Lipid oxidation during spray drying is a function of drying temperature. Higher drying temperatures are related to increased lipid oxidation.^[32] Furthermore, the degree of lipid oxidation is also related to the fraction of un-encapsulated oil (i.e., surface oil) which might undergo rapid oxidation during the spray drying process.^[31]

FIGURE 1 PV, AV, Totox value of control-M, 5%BJ-M and 10%BJ-M powders; Values are means and SD of triplicate determinations; MO = menhaden oil; control-M = microencapsulated MO; 5%BJ-M = microencapsulated MO with 5% BJ; 10%BJ-M = microencapsulated MO with 10% BJ; PV = peroxide value; AV = anisidine value.

Lower PV, AV, and Totox values were obtained for 10% BJ-M than for 5% BJ-M or for control-M powders (Fig. 1). This indicated that 10% BJ-M prevented lipid oxidation during spray drying to a greater extent than the other formulations and 5% BJ-E had significantly lower PV values compared to control-M. AV and Totox values of 5% BJ-M do not show significant difference with those of control-M. This indicated that increased juice addition reduced lipid oxidation during spray drying. The results indicated BJ antioxidants effectively reduced lipid oxidation during spray drying.

During microencapsulation MO undergoes multiple changes in physical properties such as the amount of bulk oil, dispersed oil in the aqueous phase (emulsion), and dispersed oil in dry matrix (microencapsulated powders). Although microencapsulation reduces lipid oxidation during storage, addition of antioxidants is required to prevent lipid oxidation that occurs during processing and to further reduce lipid oxidation during storage.^[31]

A study conducted using antioxidants such as tocopherol and ascorbyl palmitate showed a decrease in lipid oxidation during the emulsification process compared to control samples.^[30] Another study showed that emulsions with tocopherol and rosemary extract had reduced auto-oxidation during the process of spray drying.^[8] Oil-in-water emulsions are better stabilized by lipophilic antioxidants such as tocopherol compared to hydrophilic antioxidants; such as rosemary extract.^[30] Conversely, hydrophilic antioxidants may be more effective in bulk oil.^[30] Since BJ consists of both lipophilic and hydrophilic antioxidants^[25,^1] it has the possibility of preventing lipid oxidation during the emulsification and spray drying processes. Moreover, it can be advantageous to employ antioxidants having larger quantities of hydrophilic antioxidant activity compared to lipophilic, such as BJ does, since such antioxidants can effectively prevent lipid oxidation in the microencapsulated powders during spray drying and subsequent storage period.^[25]

FAME Composition of MO and Microencapsulated Powders

Table 5 summarizes the percent of fatty acid methyl esters present in the initial MO and in the oil extracted from control-M, 5% BJ-M, and 10% BJ-M. EPA and DHA omega-3 type fatty acids are the most important and can undergo rapid oxidation due to their highly unsaturated nature. The experimental values reported for MO EPA and DHA content is in accordance with the literature.^[33,^6] All the microencapsulated powder formulations contained around 14% EPA (Table 5). The EPA content of MO and 10% BJ-M were not significantly different; however, the EPA content of control-M and 5% BJ-M was significantly different than that of MO. No significant difference in EPA content was obtained between 5% BJ-M and control-M. This could be associated to the PV, AV, and Totox values obtained for microencapsulated powders, where 10% BJ-M and MO did not have significant difference, while 5% BJ-M and control-M had significant differences in the previously mentioned lipid oxidation parameters. The percent DHA contents of all the microencapsulated powder formulations were around 13.5%. DHA content also exhibited a similar pattern to EPA, where MO was not significantly different from 10% BJ-M, while MO was significantly different from 5% BJ-M and control-M, and 5% BJ-M had no significant difference with control-M.

TABLE 5 Fatty acid methyl ester composition (% w/w) of MO, control-M, 5% BJ-M, and 10% BJ-M*

	MO	Control-M	5% BJ-M	10% BJ-M
C14:0 Myristic	8.73 ± 0.07^a	8.86 ± 0.06^a	8.81 ± 0.06^a	8.88 ± 0.07^a
C15:0	0.85 ± 0.00^a	0.86 ± 0.00^a	0.86 ± 0.00^a	0.87 ± 0.00^a
C16:0 Palmitic	19.31 ± 0.10^b	19.54 ± 0.06^a	19.25 ± 0.23^b	19.68 ± 0.28^a
C16:1n7 palmitooleic	11.87 ± 0.07^a	11.97 ± 0.07^a	12.07 ± 0.04^a	11.90 ± 0.05^a
Unknown	1.60 ± 0.02^a	1.60 ± 0.03^a	1.43 ± 0.06^b	1.59 ± 0.03^a
C17:0	0.72 ± 0.00^a	0.74 ± 0.00^a	0.73 ± 0.00^a	0.74 ± 0.00^a
Unknown	1.22 ± 0.01^a	1.23 ± 0.01^a	1.24 ± 0.01^a	1.22 ± 0.01^a
C17:2n6	0.75 ± 0.00^a	0.73 ± 0.00^a	0.74 ± 0.00^a	0.74 ± 0.00^a
C18:0 Stearic	3.69 ± 0.01^a	3.63 ± 0.01^a	3.63 ± 0.03^a	3.67 ± 0.04^a
C18:1n9c Oleic	6.29 ± 0.03^a	6.35 ± 0.02^a	6.40 ± 0.04^a	6.36 ± 0.02^a
C18:1n7	3.11 ± 0.02^b	3.14 ± 0.01^b	3.17 ± 0.01^a	3.14 ± 0.01^b
C18:2n6c Linoleic	1.52 ± 0.01^a	1.53 ± 0.01^a	1.54 ± 0.01^a	1.53 ± 0.01^a
C18:3n3 Alpha-Linolenic	1.69 ± 0.02^a	1.67 ± 0.01^a	1.69 ± 0.01^a	1.67 ± 0.02^a
C18:4n3	3.30 ± 0.02^a	3.28 ± 0.03^a	3.31 ± 0.04^a	3.26 ± 0.02^a
C20:1n9	0.99 ± 0.00^a	1.02 ± 0.01^a	1.03 ± 0.03^a	1.01 ± 0.02^a
C20:3n3	1.80 ± 0.00^b	1.80 ± 0.02^a	1.83 ± 0.02^a	1.80 ± 0.01^a
C20:4n6 arachidonic	1.15 ± 0.06^a	1.15 ± 0.01^a	1.17 ± 0.01^a	1.16 ± 0.02^a
EPA (20:5n3)	14.21 ± 0.03^a	13.94 ± 0.01^b	14.04 ± 0.03^b	14.13 ± 0.05^a
C22:5n3	2.83 ± 0.01^a	2.77 ± 0.05^a	2.74 ± 0.04^a	2.79 ± 0.03^a
DHA (22:6n3)	13.75 ± 0.06^a	13.42 ± 0.08^b	13.51 ± 0.03^b	13.64 ± 0.02^a
Total omega 3	37.58 ± 0.05^a	36.76 ± 0.06^c	37.12 ± 0.08^bc	37.29 ± 0.03^ab
Total omega 6	3.42 ± 0.02^a	3.41 ± 0.02^a	3.45 ± 0.03^a	3.43 ± 0.01^a
Saturates (S)	33.30 ± 0.05^a	33.63 ± 0.08^a	33.28 ± 0.07^a	33.84 ± 0.08^a
Mono unsaturates (M)	22.26 ± 0.19^a	22.48 ± 0.13^a	22.67 ± 0.18^a	22.41 ± 0.13^a
Poly unsaturates (P)	41.00 ± 0.03^a	40.17 ± 0.02^c	40.57 ± 0.02^b	40.72 ± 0.02^ab
ω3/ω6	10.99 ± 0.03^a	10.78 ± 0.01^c	10.76 ± 0.02^c	10.87 ± 0.01^b
P/S	1.23 ± 0.02^a	1.19 ± 0.03^a	1.22 ± 0.05^a	1.20 ± 0.02^a
Total fatty acids identified	99.38 ± 0.12	99.11 ± 0.12	99.19 ± 0.11	99.78 ± 0.08

*Values are means and SD of triplicate determinations;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05);

MO = menhaden oil; control-M = microencapsulated MO; 5% BJ-M = microencapsulated MO with 5% BJ; 10% BJ-M = microencapsulated MO with 10% BJ.

ME of Microencapsulated MO Powders

Table 6 shows the surface oil content, total oil content, and ME of microencapsulated MO powders. A high oil load was used in the experiment in order to study the effect of antioxidants on lipid oxidation. The total oil content of the MO powders ranged from 55 to 59%. Surface oil content of control-M was significantly higher than 10% BJ-M and 5% BJ-M. The greater surface oil content of control-M could be related to its lower emulsion stability compared to 5% BJ-M and 10% BJ-M (Table 2).

TABLE 6 Surface oil content, total oil content, and microencapsulation efficiency of microencapsulated MO powders

Powder property	Control-M	5% BJ-M	10% BJ-M
Surface oil content %	7.2 ± 0.1^a	3.6 ± 0.3^c	5.9 ± 0.1^b
Total oil content %	59.20 ± 2.11^a	57.31 ± 4.23^a	55.12 ± 1.42^a
Microencapsulation efficiency %	87.80 ± 0.92^b	93.68 ± 0.66^a	89.27 ± 0.60^b

Values are means and SD of triplicate determinations;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05);

MO = menhaden oil; control-M = microencapsulated MO; 5% BJ-M = microencapsulated MO with 5% BJ; 10% BJ-M = microencapsulated MO with 10% BJ.

TABLE 7 Color of microencapsulated MO powders

	Control-M	5% BJ-M	10% BJ-M
L*	97.03 ± 0.03^a	89.33 ± 0.02^b	83.41 ± 0.02^c
a*	−1.04 ± 0.01^c	3.42 ± 0.01^b	6.13 ± 0.01^a
b*	5.24 ± 0.03^a	1.27 ± 0.03^b	1.07 ± 0.01^c

Values are means and SD of triplicate determinations;

^abcMeans with different exponents in the same row indicate significant difference (p ≤ 0.05);

MO = menhaden oil; control-M = microencapsulated MO; 5%BJ-M = microencapsulated MO with 5% BJ; 10% BJ-M = microencapsulated MO with 10% BJ.

All of the MO powders had high ME. 5% BJ-M had significantly higher encapsulation efficiency than did control-M and 10% BJ-M. Successful microencapsulation of MO should result in high ME and low surface oil content on the powder particles. A higher ME and lower surface oil content was observed in 5% BJ-M compared to 10% BJ-M. This could be attributed to the higher viscosity of 10% BJ-E. Higher viscosity emulsions tend to produce larger particles.^[28] According to Jafariet al.^[28] microencapsulated powders with large particle size contain more surface oil than powders with small particle size. Powders having more surface oil resulting in lower ME. Surface oil not only affects the wettability and dispensability of the powder but also readily undergoes lipid oxidation.^[4] ME can be affected by many factors such as properties of wall and core materials, emulsion characteristics, and spray drying parameters.^[4] Generally microencapsulated powder has oil load percentages ranging from 30–50%.^[28] Having higher oil load in the microcapsules can be advantageous. However, high oil load in microcapsules is associated with reduced encapsulation efficiencies and increased surface oil contents. It has been suggested that a core to wall material ratio of 1:4 is optimal for wall materials like gum arabic and modified starches.^[28] The microcapsules produced using n-OSA emulsifying starch and corn syrup with an oil load of around 57% produced powders during the present study with high ME.

Color, SEM, and Particle Size of Microencapsulated MO Powders

The L*, a*, and b* color values of microencapsulated MO powders (control-M, 5% BJ-M, and 10%BJ-M) are shown in Table 7. The color of the emulsions with or without the BJ correlated with the color of their microencapsulated powders. The addition of BJ increased the a* value of the powders and reduced the L* and b* values. The powders produced without the juice had a light yellow color (higher b*) due to the color of the oil. Microencapsulated powder enriched products are becoming increasingly popular in the market. There are a number of products exist with enriched fish oil powders such as meat and poultry products, bread and bakery products, fats, eggs and egg products, milk and dairy products, pasta, sauces, juices, soft drinks, instant concentrates, and spreads.^[34] The quantity of microencapsulated fish oil powders in a food product is very low and this amount mostly followed the recommended daily intake of EPA and DHA levels. Although BJ slightly changed the powder color, it would not likely to cause an effect on practical applications related to fish oil powders.

SEM images of the MO powders control-M, 5% BJ-M, and 10% BJ-M are shown in Fig. 2. All of the powders had individual spherical particles with some dents and shrinkage that could be attributed to the early stage of drying under high drying rates.^[28] This is the typical appearance of a spray dried product. There is some aggregation of particles that can be seen in all three powder samples which may be due to surface oil.^[35] The three powders do not differ greatly, indicating that extract addition does not change the morphology of the powders. There are no distinct cracks on the powder surface observed on any of the three menhaden powder samples.

FIGURE 2 SEM image of microencapsulated control-M (a), 5%BJ-M (b), and 10%BJ-M (c) powders. Control-M = microencapsulated MO; 5%BJ-M = microencapsulated MO with 5% BJ; 10%BJ-M = microencapsulated MO with 10% BJ.

The particle size distributions of control-M, 5% BJ-M, and 10% BJ-M are illustrated in Fig. 3. All powders had mean particle size ranging from 20 to 30 μm. Generally, the particle size of microencapsulated powders produced by spray drying is 1–400 μm.^[28] As previously explained, addition of BJ increased the emulsion concentration and viscosity, and produced powders with high mean particle size (Fig. 3). The mean particle size of 10% BJ-M was significantly higher than both control-M and 5% BJ-M, and 5% BJ-M had greater mean particle size than control-M. Also the SEM microstructure (Fig. 2) of control-M, 5% BJ-M, and 10% BJ-M were consistent correlated with the particle size distribution where 10% BJ-M SEM images showed bigger particles compare to 5%BJ-M and control-M.

FIGURE 3 Particle size distribution of control-M, 5%BJ-M and 10%BJ-M; values are means and SD of triplicate determinations. Control-M = microencapsulated MO; 5%BJ-M = microencapsulated MO with 5% BJ; 10%BJ-M = microencapsulated MO with 10% BJ.

Some researchers reported that higher particle size increased the amount of oil retention. However, Chang et al.^[36] found that total oil retention was highest in intermediate size particles while large particles showed the lowest oil retention. Also, it was found that increased oil load in emulsions will result in larger oil droplets and an increase in mean particle size of the microcapsules. Particle size can be affected by emulsion characteristics (viscosity and solids concentration) and drying conditions.^[28]

CONCLUSION

MO emulsions produced with or without BJ had high emulsion stability. All of the emulsions showed pseudo plastic (n < 1) shear thinning flow behavior. The addition of BJ changed the color of both the emulsions and the microencapsulated powders by increasing a* and reducing L* and b* values. During the emulsification and spray drying process BJ retarded lipid oxidation of MO. Increased juice addition was directly related to decreased lipid oxidation in emulsions and powders. Powders, with or without BJ, did not have significant differences in powder morphology. Mean particle size of the powders ranged between 20–30 μm. This study indicated that BJ could be added to MO emulsions in order to prevent potential lipid oxidation during emulsification and spray drying without significantly altering emulsion or powder characteristics.