Production of high-oleic palm oil nanoemulsions by high-shear homogenization (microfluidization)

Innovative Food Science and Emerging Technologies 35 (2016) 75–85

Contents lists available at ScienceDirect

Innovative Food Science and Emerging Technologies journal homepage: www.elsevier.com/locate/ifset

Production of high-oleic palm oil nanoemulsions by high-shear homogenization (microfluidization) Ricaurte Leidy a, Perea-Flores María de Jesús b, Martinez Anamaria a, Quintanilla-Carvajal María Ximena a,⁎ a

Facultad de Ingeniería, Universidad de La Sabana, km 7 vía autopista Norte, Bogotá, Colombia Laboratorio de Microscopía Confocal, Centro de Nanociencias y Micro y Nanotecnología, Unidad Profesional ‘Adolfo López Mateos’, IPN, Luis Enrique Erro S/N, Zacatenco, C. P. 07738, Mexico City, Mexico

b

a r t i c l e

i n f o

Article history: Received 5 February 2016 Received in revised form 13 March 2016 Accepted 7 April 2016 Available online 13 April 2016

a b s t r a c t Nanoemulsions present benefits such as an increase in the bioavailability, solubility and targeted delivery of encapsulated substances, and thus, they are a method of incorporating high nutritional value oils, such as high-oleic palm oil (HOPO). In this work, O/W nanoemulsions were obtained by microfluidization using HOPO (1–20% w/w) in the oily phase along with whey (1–20% w/w), Tween 20 (1:1 w/w ratio) and water in the aqueous phase following a surface response design methodology. The response variables were the average drop size (ADS), the polydispersity index (PDI), the zeta potential (ζ), CIELAB color parameters and viscosities of the fresh nanoemulsions (0 days) and nanoemulsions stored at two temperatures (5 and 19 °C) for 4 days. The ADS, PDI and ζ values varied between 163 and 2268 nm, 0.2 and 1, and −29.7 and −47.2 mV, respectively. The viscosity was affected by the storage temperature; after 4 days at 19 °C, it increased almost 6-fold compared to the viscosity of the fresh sample. With regards to the color parameters, significant changes were observed based on the concentrations of HOPO and whey. In addition, the prediction equations only presented errors below 7% for the evaluated variables, with R2 values above 0.85. Finally, the influence of whey protein denaturation at 60 °C on the stabilities of the two most stable nanoemulsions, according to the optimization process, was observed. Industrial relevance: Among its many benefits, nanoencapsulation is characterized by increasing the bioavailability of the encapsulated active compound and by the protection that it grants against environmental and processing effects, as micronutrients (for example vitamins) can be susceptible to chemical, enzymatic and/physical instability prior, during and after the processing of food products that contain. One of the techniques studied in recent decades for obtaining nanoemulsions is microfluidization. Microfluidization is a high-energy method that uses high pressure to force the fluid through microchannels that have a specific configuration, emulsifying the fluid by the combined effects of cavitation, shear and impact, thus showing an excellent emulsifying efficiency. However, in food industries the use of microfluidization is not popular and other kind of high shear homogenization are used. In this work, the development of stable emulsions using microfluidization, calls for the use of other types of materials that can provide emulsifying characteristics, such as whey, a compound that is currently one of the main effluents of dairy processes, depending on the type of product. Obtaining nanoemulsions for encapsulation purposes has been studied in many functional products, but to the best of our knowledge, it has not been reported with high-oleic palm oil. This oil contains approximately 50% saturated, 10% diunsaturated and 40% monounsaturated fatty acids, with oleic acid in sn-2 position in triacylglycerols. This composition makes palm oil as soluble as olive oil. In addition, high-oleic palm oil (HOPO), in particular, has a high stability because it is an oleic acid-rich oil, which has been introduced to replace trans fats and has presented a healthy alternative to such fats in food formulations and the fried food industry. It is also important to highlight that oleic acid has a range of health benefits, such as a decrease in the total cholesterol, an increase in HDL (high-density lipoprotein) and a decrease in LDL (low-density lipoprotein). Oleic acid also retards the development of heart diseases, promotes the formation of antioxidants in the body and reinforces the integrity of the cell wall. In addition, red palm oil (crude) contributes significant nutritional value because it is rich in β-carotenes, α-tocopherol and tocotorienols, supplying vitamins and provitamins that are important for but not produced by the human body. Finally, this work demonstrates that emulsion drop size does not affect the stability of the nanoemulsion if it formulation is designed. Therefore, the goal of this work was to evaluate the most favorable conditions for the microfluidization, formulation and storage of HOPO nanoemulsions using whey powder to produce stable nanoemulsions. © 2016 Elsevier Ltd. All rights reserved.

⁎ Corresponding author at: Campus Puente del Común, km7 vía Autopista Norte Bogotá, Colombia. E-mail address: [email protected] (M.X. Quintanilla-Carvajal).

http://dx.doi.org/10.1016/j.ifset.2016.04.004 1466-8564/© 2016 Elsevier Ltd. All rights reserved.

76

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

1. Introduction Nanoencapsulation is a process by which substances are packed in a miniature-sized vessel, and bioactive packing is performed at the nanoscale (Quintanilla-Carvajal et al., 2009). Among its many benefits, nanoencapsulation is characterized by increasing the bioavailability of the encapsulated active compound and by the protection that it grants against environmental and processing effects, as micronutrients (for example vitamins) can be susceptible to chemical, enzymatic and/physical instability prior, during and after the processing of food products that contain them (Joye, Davidov-Pardo, & McClements, 2014). In many cases, nanoencapsulation begins with the production of nanoemulsions, which are systems formed by an oily and an aqueous phase, in which they are emulsified through the use of an emulsifier. In addition, the nanoemulsions are formed with small drop sizes and high surface areas (Kim, Oh, Lee, Song, & Min, 2014). Such properties grant them potential advantages over conventional emulsions, such as a good physical stability and higher bioavailability (Tabibiazar et al., 2015). One of the techniques studied in recent decades for obtaining nanoemulsions is microfluidization. Microfluidization is a high-energy method that uses high pressure to force the fluid through microchannels that have a specific configuration, emulsifying the fluid by the combined effects of cavitation, shear and impact, thus showing an excellent emulsifying efficiency (Shen & Tang, 2012). Furthermore, the microfluidization method requires high-energy inputs and dedicated equipment, but could produce ultrafine emulsions at much lower surfactant-to-oil ratio (SOR b 0.1) (Komaiko & McClements, 2015; Yang, Marshall-Breton, Leser, Sher, & McClements, 2012). It is also important to highlight the role of the surfactant when generating an emulsion that is stable over time. However, the development of stable emulsions calls for the use of other types of materials that can provide this type of emulsifying characteristic, such as whey, a compound that is currently one of the main effluents of dairy processes, depending on the type of product (Das, Sarkar, Sarkar, Bhattacharjee, & Bhattacharjee, 2015). Thus, this residue has been studied in the production of emulsions or as a coating material for encapsulation processes at different scales (Esfanjani, Jafari, Assadpoor, & Mohammadi, 2015; Wang et al., 2016) because it contains approximately 2–18% proteins, 8–14% minerals (ash) and 77–80% carbohydrates (Muangrat, Onwudili, & Williams, 2011). On the other hand, the characteristics related to the stability of an emulsion are the drop size, the viscosity and the zeta potential (ζ). However, the application of nanoemulsions in diverse food matrices must also ensure the acceptance of parameters such as the color. These parameters are related to the quality and the consumer perception, both visual as well as sensory, towards the product because the color and appearance of foods are important factors that contribute to their selection by the consumer (Silva et al., 2011). Obtaining nanoemulsions for encapsulation purposes has been studied in many functional products, but to the best of our knowledge, it has not been reported with high-oleic palm oil. This oil contains approximately 50% saturated, 10% di-unsaturated and 40% monounsaturated fatty acids, with oleic acid in sn-2 position in triacylglycerols. This composition makes palm oil as soluble as olive oil (Akoh, 2005). In addition, high-oleic palm oil (HOPO), in particular, has a high stability because it is an oleic acid-rich oil, which has been introduced to replace trans fats and has presented a healthy alternative to such fats in food formulations and the fried food industry (Kodali, 2014). It is also important to highlight that oleic acid has a range of health benefits, such as a decrease in the total cholesterol, an increase in HDL (high-density lipoprotein) and a decrease in LDL (low-density lipoprotein). Oleic acid also retards the development of heart diseases, promotes the formation of antioxidants in the body and reinforces the integrity of the cell wall (Munévar, 2010). In addition, red palm oil (crude) contributes significant nutritional value because it is rich in βcarotenes, α-tocopherol and tocotorienols (Akoh, 2005), supplying

vitamins and provitamins that are important for but not produced by the human body. Therefore, the goal of this work was to evaluate the most favorable conditions for the microfluidization, formulation and storage of HOPO nanoemulsions using whey powder to produce stable nanoemulsions. 2. Methodology and materials 2.1. Materials The nanoemulsions were prepared with crude high-oleic palm oil donated by the National Federation of Palm Oil Growers (Federación Nacional de Cultivadores de Palma de Aceite) (Colombia), whey powder donated by the Alpina Corporation, Tween 20 (Scharlau, Spain) and Milli-Q water. 2.2. Preparation of coarse emulsions The coarse emulsions were homogenized in an Ultra-Turrax (IKA, USA) at 9500 rpm, incorporating Tween 20 followed by the sequential addition of whey powder and HOPO to the Milli-Q water over 10 min. Subsequently, such emulsions were processed to obtain the nanoemulsions. 2.3. Nanoemulsion preparation The nanoemulsions were obtained following the methodology of (Quintanilla-Carvajal et al., 2014), with some modifications. They were homogenized in an LM10 microfluidizer (Microfluidics, England) following a response optimization design obtained from the software Design Expert Version 7.1.0 (Stat-Ease Inc., MN, USA), in which the following three numerical and one categorical factors were varied: HOPO concentration (1–20% w/w), whey concentration (1–20% w/w), and microfluidization pressure (10,000–20,000 psi), as well as the microfluidization cycles as a three-level categorization factor (1, 2 and 3). The Tween 20 concentration was held constant at 1% w/w with respect to the HOPO concentration. This emulsifier was chosen for allowing rapidly adsorb to the surface of the oil droplets, reduce interfacial tension to prevent droplet coalescence (Degner, Chung, Schlegel, Hutkins, & McClements, 2014; Jo & Kwon, 2014; Teo et al., 2016) and has shown good results in small particles for various applications including nanoemulsions (Silva et al., 2011). Table 1 shows the conditions provided by the response optimization design for the preparation of HOPO nanoemulsions. 2.4. Storage of the HOPO nanoemulsions The nanoemulsions obtained (Table 1) were stored for 4 days (t = 4) (Adjonu, Doran, Torley, & Agboola, 2014; Y. Li, Zheng, Xiao, & McClements, 2012). These samples were evaluated at two different conditions: room temperature (19 °C) and refrigeration temperature (5 °C). 2.5. Nanoemulsion characterization The fresh nanoemulsions were characterized with respect to the average drop size (ADS), the polydispersity index (PDI), the ζ, the color according to the CIELAB method (L*, a*, b*, Cab* and h°ab) and the apparent viscosity (η). This last response variable was measured for the fresh nanoemulsions and t = 4 at room and refrigeration temperatures. 2.5.1. Analysis of the drop size, PDI and ζ The ADS, the PDI and the ζ were measured in the nanoemulsions reported in Table 1 by the use of a Zetasizer NanoZS (Malvern Instruments, England) dynamic light scattering (DLS) device with a laser diffractometer using a water dilution of 1:100 v/v (Qian, Decker, Xiao, & McClements, 2012; Sessa et al., 2014). The measurements were performed in triplicate with a scattering angle of 173°.

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

77

Table 1 Response optimization surface experimental design methodology for preparation of nanoemulsions and ajusted variables to the model: ADS, ζ, PDI, L*, a*, b*, Cab*, h°ab, viscosity for fresh and stored nanoemulsions of HOPO. Run

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

HOPO concentration [A] (% wt/wt)

20.0 20.0 1.0 20.0 7.1 9.0 13.9 7.1 1.0 1.0 5.2 5.2 20.0 9.0 15.5 1.0 13.9 20.0 20.0 1.0 20.0 8.7 15.5 1.0 12.0 12.0 1.0 20.0

Whey concentration [B] (% wt/wt)

20.0 11.0 1.0 1.0 10.9 20.0 12.6 10.9 9.9 10.4 20.0 20.0 10.5 20.0 1.0 1.0 12.6 20.0 1.0 20.0 20.0 18.1 1.0 20.0 1.0 19.9 1.0 7.08

Pressure [C] (psi)

15,700 10,000 14,203 20,000 16,752 10,000 14,500 16,752 20,000 10,000 17,700 17,700 20,000 10,000 12,400 10,000 14,500 20,000 10,000 10,000 10,000 20,000 12,400 20,000 20,000 13,731 20,000 15,900

Cycles [D]

2 1 2 1 2 2 3 2 1 3 3 3 3 2 3 1 3 1 2 1 3 1 3 2 2 1 3 2

Fresh nanoemulsions, t = 0

Stored nanoemulsions, t = 4

ADS (nm)

ζ (mV)

PDI

L*

a*

b*

Cab*

2261.0 2268.0 186.3 1417.0 389.0 359.3 1026.0 405.9 264.9 232.3 308.4 279.6 2083.7 715.9 1630.0 174.9 1276.0 1322.0 1469.7 280.6 1779.7 947.3 1653.0 232.9 1170.3 1295.0 163.7 2014.0

−30.5 −33.9 −34.9 −43.1 −33 −32.4 −34.4 −33.3 −34.3 −34.7 −29.7 −31.8 −35.3 −31.3 −44.7 −37.9 −32.8 −33.5 −40.5 −34.1 −34.9 −33.4 −43.2 −42.7 −47.2 −34.3 −34.5 −38.8

0.4 0.5 0.3 0.8 0.4 0.4 0.8 0.4 0.3 0.2 0.3 0.3 0.6 0.7 1 0.2 0.5 0.2 0.9 0.2 0.8 0.6 1 0.2 1 0.5 0.2 0.7

50.7 55.9 54.5 56.4 54.1 50.7 52.7 52 52.6 54.5 51.8 50.8 53.6 50.8 57.9 59 52.9 53.6 59 54.8 54 53 57.4 47.5 56.4 53.4 55.9 52.8

8.7 3.9 −7.9 3.7 3.1 4.4 5.3 0.8 −4.6 −5.7 2.7 5.1 5.8 6.4 0.4 −9.5 6.5 9.9 −3.2 −3.7 3.2 −1.3 0.6 −0.4 −0.5 1.4 −8.4 1.6

57.2 59.7 55.5 61.2 59.8 57.9 59.5 58.2 57.5 56.1 56.5 58.3 59.1 59.3 65.1 58.7 59.2 58.9 66.9 54.6 58.2 58.8 61.7 49.9 62 57.9 52 55.8

57.8 59.8 56.0 61.4 60.0 58.0 59.7 58.2 57.7 56.3 56.6 58.6 59.4 59.6 65.1 59.4 59.6 59.7 67.0 54.7 58.3 58.9 61.7 49.9 62.0 57.9 52.7 55.9

2.5.2. Analysis of the drop size and ζ as a function of temperature A temperature sweep was performed between 20 and 70 °C for the two optimum nanoemulsions (1-opt and 2-opt) with the lowest and highest drop sizes but with minimum ζ values (more stable) given by the response optimization design in the Design Expert software Version 7.1.0 (Stat-Ease Inc., MN, USA) with the formulations shown in Table 3 in the Zetasizer NanoZS device (Malvern Instruments, England) using a water dilution of 1:100 v/v. The measurements were performed in duplicate.

2.5.3. Confocal laser scanning microscope (CSLM) Confocal Laser Scanning Microscopy (CLSM) was performed using a LSM 710 NLO microscope (Carl Zeiss, Germany) for emulsions following the methodology of Quintanilla-Carvajal et al. (2014). Samples were mounted on glass slides and observed at a laser wavelength of 405, 561, 488 and 633 nm with a 100, 2, 2 and 2% capacity, respectively, and 63 ×/1.40 plan-apochromat objective. This capture mode used a spectral imaging technique that automatically outputs separated channels of the multiple labeled samples (Hernández-Hernández et al., 2014). The autofluorescence signals from samples were observed by using the built-in software ZEN (Carl Zeiss, Germany).

Viscosity Troom, (mPa·s)

Viscosity Trefrig, (mPa·s)

81.3 86.6 98.1 85.7 86.4 85.7 84.9 89.2 94.6 94.7 87.3 85.0 84.4 83.7 89.7 98.5 83.8 80.5 92.8 93.9 89.8 91.3 89.4 90.5 90.5 88.6 99.2 88.3

93.5 8.6 66.8 13.5 34.9 89.0 13.8 34.5 14.1 13.7 8.3 11.9 12.9 92.8 1.5 0.9 16.4 55.9 17.5 40.5 55.9 15.1 14.5 40.0 13.2 34.4 0.9 87.2

545.5 179.1 66.8 15.9 1.9 182.8 260.6 107.9 14.0 13.8 193.5 130.2 260.5 327.3 2.9 3.1 252.7 155.4 16.7 12.3 553.3 107.9 11.8 20.3 11.6 54.0 31.3 138.5

74.4 31.9 0.9 1.8 2.2 7.4 19.7 1.5 1.4 1.4 5.7 4.6 28.3 21.4 12.9 1.0 12.9 77.6 18.8 80.3 109.0 17.2 12.9 23.0 13.6 112.2 1.0 58.8

2.5.5. Color determination The images acquisition for the color quantification was performed following the methodology of Quintanilla-Carvajal, Arenas-Ocampo, Campos-Mendiola, Camacho-Díaz, and Jiménez-Aparicio (2015) using the J V1.34 software (National Association of Health, USA) and the installation of the Color Space Converter plugin. The HOPO nanoemulsions with a thickness of 2 mm were placed on a matte white background. Images were captured under indirect daylight using a Nikon Digital reflex camera (model D50) with a Kikon lens (model Af-S) of 18–55 mm/f4.5–5.0. The focal distance was kept constant at 10 cm and perpendicular between the camera and the nanoemulsions. The images were saved in the. BMP format, and the image size was 840 × 840 pixels in RGB color. The chrome values Cab* were estimated following the equation of Sui, Bary, and Zhou (2016): C
ab ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 a
2 þ b :

ð1Þ

Additionally, the h°ab (hue) was determined in degrees (°) using the following equation of Poynton (2003) as a function of the values a* and b*: 

2.5.4. Determination of the η for fresh and stored nanoemulsions The η was measured for fresh and stored nanoemulsions (viscosity, viscosity Troom and viscosity Trefrig) in a MCR 502 rheometer (Anton Paar, Germany) with a parallel plate geometry (PP50) of 49.97 mm using a 0.5 mm gap within a shear velocity range between 1 and 100 s−1. Approximately 1.2 mL of the sample was carefully placed in the lower plate. Due to the nanoemulsion Newtonian behavior (the linear fit of the shear stress as a function of the shear rate was performed with an average R2 of 0.995 (data not shown)), an arbitrary shear rate of 100 s− 1 was chosen to fit the η in the Design Expert software.

h°ab

Viscosity (mPa·s)

hab ¼ tan−1

  b : a

ð2Þ

2.6. Statistical analysis The statistical analysis was performed using the response optimization surface experimental design methodology in the Design Expert software Version 7.1.0 (Stat-Ease Inc., MN, USA). Quadratic and linear models were used to express the response variables as a function of the independent factors, where A, B, C and D are the coded values of the oil concentration, the whey concentration, the pressure and the cycle number, respectively. A statistical

78

Table 2 ANOVA for the ajusted variables to response optimization design: ADS, ζ, a*, b*, L, c*, h°ab, viscosity for fresh and stored nanoemulsions of HOPO. Factor independiente

ζ, mV

ADS, nm

PDI

L

a*

b*

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

3643.88 3593.93 1.21 1,73 13,81

5 1 1 1 2

b 0.0001 b 0.0001 0,79 0,75 0,66

1,86 0,64 0,29 0,00 0,02 0,11 0,01 0,12 0,02 0,00 0,06 0,18 0,04 0,01 0,04 0,09 0,94 0,83

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

0,0007 b 0.0001 0001 0,83 0,48 0,01 0,40 0,04 0,20 0,87 0,14 0004 0,10 0,43 0,80

b 0.0001 0,08 b 0.0001 0001 0001

566.79 325,55 240,08 22,10 22,44

5 1 1 1 2

b 0.0001 b 0.0001 b 0.0001 0,07 0,18

0,16

0,0004 0,01 b 0.0001 0,13 0,70 0001 0,89 0,28 0,61 0,50 0001 0,58 0003 0,03 0,05

5 1 1 1 2

17 5

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

169,54 4,37 125,41 19,94 27,55

316.25 37,71 0,91 0,89

533,97 30,31 216,52 8,30 2,25 59,76 0,06 8,97 0,84 4,64 100,45 1,01 48,65 19,43 25,82 5,26 0,94 0,85

25,73 2,98 0,86 0,82

17 5

0,15

125,78 8,23 0,81 0,77

17 5

0,05

268.74 63.43 65.51 12,36 2,32 1,72 0,25 4,34 13,91 11,09 9,04 56,40 0,10 6,80 32,80 9,83 0,86 0,63

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

Table 2 (continued) Factor independiente

Model A B C D AB AC AD BC BD CD A^2 B^2 C^2 Lack of Fit Pure Error R2 R2- Adjusted

Fresh nanoemulsions, t = 0

Stored nanoemulsions, t = 4 h°ab

b*

Cab*

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

SS

df

p-Value

0,02 0003 0003 0,12 0,77 0,54 0,81 0,62 0,10 0,31 0,38 0005 0,88 0,24 0,11

258.04 62,27 64,86 11,54 2,11 0,27 0,70 3,98 14,98 10,06 10,07 52,70 0,25 6,51 34,07 10,60 0,85 0,60

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

0,03 0.004 0.003 0,14 0,79 0,81 0,70 0,65 0,10 0,36 0,36 0,01 0,82 0,26 0,11

619.08 254.87 155.83 14,44 9,83 3,60 24,57 13,82 1,48 29,00 3,53 19,60 4,89 4,49 42,09 9,31 0,92 0,79

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

0002 b 0.0001 0,0003 0,12 0,42 0,42 0,05 0,30 0,60 0,11 0,72 0,08 0,35 0,37 0,06

174,27 3,02 51,02 2,72 72,57 0,62 3,92 3,05 9,42 2,47 12,83 13,31 0,68 9,51 5,15 3,53 0,95 0,87

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5 5

0,0002 0,09 b 0.0001 0,11 b 0.0001 0,42 0,06 0,22 0,01 0,29 0,01 0003 0,40 0,01 0,34

993.68 227.16 414.94 0,09 91,96 162.86 2.41 5,34 12.01 47.57 30.64

14 1 1 1 2 1 1 2 1 2 2

0001 0001 b 0.0001 0,93 0,04 0002 0,64 0,78 0,31 0,14 0,27

81,41 55,54 0,88 0,75

8 5

0,57

232,40 79,77 79,01 2,59 7,66 0,40 0,08 9,28 4,58 29,50 14,28 10,19 10,76 1,25 9,68

17 1 1 1 2 1 1 2 1 2 2 1 1 1 5

0,0002 b 0.0001 b 0.0001 0,17 0,08 0,57 0,80 0,06 0,08 0002 0,02 0,02 0,01 0,33 0,07

0,86 0,63

Viscosity

Viscosity Tamb

Viscosity Tcool

0,95 0,87

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

Model A B C D AB AC AD BC BD CD A^2 B^2 C^2 Lack of Fit Pure Error R2 R2- Adjusted

Fresh nanoemulsions, t = 0

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

significance test was used in the total error criteria with a confidence level of 95%. The significant terms in the model were found through analysis of variance (ANOVA). The fit of the model was evaluated by the R2 value. The values of the R2 coefficients were higher than 0.85 for all of the response variables to achieve a good fit of the model, as the R2 must not be lower than 80% for any process that uses biological materials (Rodríguez-Bernal et al., 2015). The graphic and numerical optimization of the Design Expert software was used for response optimization. The variables that statistically responded to a linear model were the ADS, L* and a*; those that responded to a quadratic model were the ζ, the PDI, b*, c*, h°ab, the fresh viscosity and the viscosity at Tcool; and the viscosity at Troom responded to a 2FI model, as observed in Table 2. All of the models gave 3 prediction equations of the response variables for each one of the microfluidization cycles (Table 4). 3. Results and discussion 3.1. Effect of the formulation and microfluidization conditions on the ADS, PDI and ζ of the HOPO emulsions The ADS, the PDI and the ζ obtained for the response optimization design are shown in Table 1. For the ADS, the minimum value was 163.7 nm, and the maximum value was 2268.0 nm, where only the oil concentration (A) presented influence on this variable (see Table 3). It was observed that the HOPO composition did not cause interactions with other compounds because the HOPO contains approximately 59% oleic acid, which has emulsifying characteristics. However, as it is highly hydrophobic (HLB = 1.0), it tends to preferentially accumulate in the oil, thus inhibiting its surfactant activity (D J McClements, 2004). In addition, 12 processing conditions were obtained, which generated emulsions with ADS values lower than 500 nm and higher than 100 nm, the range considered for the nanoemulsions to be an intermediate size between normal emulsions and microemulsions (Xin et al., 2013). Fig. 1 shows the ADS isoplots, presenting a square root fitted to the response and linear model with a R2 value of 0.91, which changes as a function of the HOPO concentration, where the pressure and the microfluidization cycles did not have an effect. In the ADS response variable, the oil concentration was decisive for possibly two reasons: the oil viscosity and the amount of energy supplied in the homogenization. (See Fig 2.) The homogenization energy is understood as the intense shear that occurs at high pressure in high-shear homogenizers such as the microfluidizer™, in which the following two processes are promoted: drop breakage and re-coalescence (Wooster, Golding, & Sanguansri, 2008). In Fig. 1, it can be observed that the ADS values at HOPO concentrations between 18 and 20% w/w and any whey concentration present a very intense red zone for 1 cycle, followed by 3 cycles and 2 microfluidization cycles with lower proportions of such color. Therefore, the re-coalescence effect was observed in the production of HOPO nanoemulsions. On the other hand, when the HOPO concentration was increased, the number of oil drops increased, allowing the drops more possibilities of escaping the high-shear zone (Sadeghpour Galooyak & Dabir, 2015). With regards to the oil viscosity, the combined effects of high shear and pressure generated in the microfluidization process could be enough to reduce the ADS because the drop breakage force is higher

79

for high-viscosity oils than for low-viscosity oils (Qian & McClements, 2011). Additionally, as the HOPO concentration increased, regardless of the influence of the whey concentration, the ADS linearly increased with respect to the significant variable (A). A trend towards increased drop size is also observed when increasing the HOPO concentration (see Fig. 1). Additionally, according to the prediction equations of Table 4, the variable that contributes to the increase in the ADS is the HOPO concentration, which is related to the effect analyzed above. Regarding the PDI, this index is a dimensionless measurement of the drop size distribution amplitude (Malvern Instruments Limited, 2015), which can take values between 0.05 for monodispersed distributions and above 0.7 when the sample has a very wide size distribution (Malvern Instruments Limited, 2011). As observed in Table 2, the independent factors or significant manageable variables for the PDI were the HOPO concentration (A) and whey concentration (B), as well as the interactions between these two (AB), the oil and cycles (AD) and the HOPO concentration squared (A2). The minimum and maximum values obtained for the PDI were 0.2 and 1, respectively (see Table 1). This variable fit the response optimization design with a R2 value of 0.94. For the value of 0.2 in the polydispersion, a relatively narrow drop size distribution is observed, which suggests that the high-shear homogenization method for obtaining nanoemulsions was very successful (Barradas, de Campos, Senna, Coutinho, & Tebaldi, 2014). However, the nanoemulsions with a PDI of 1 were those that contained the highest HOPO concentration and the minimum whey concentrations at 2 and 3 microfluidization cycles, and with regards to the ADS, these nanoemulsions were actually the ones that presented larger sizes, that is, as mentioned above, the energy supplied in the form of pressure and microfluidization cycles was not enough to homogeneously decrease the nanoemulsion drop size and distribution. Another of the variables determined was the ζ, which is understood as the electro-kinetic potential difference between the dispersion medium and the sliding plane (stationary fluid layer bonded to the dispersed drop) of the drops that move (Liu, Shim, Wang, & Reaney, 2015), that is, the difference in the electro-kinetic charge of the drop surface with respect to its dispersing medium. The ζ varied within the range of values between − 29.7 and − 47.2 mV (see Table 1), which indicates good electro-kinetic stability. This variable fit with an R2 value of 0.94 showed that the HOPO concentration (A) and the whey concentration (B) had significant influence, in addition to the interaction between them (AB) and the interaction between the pressure and cycles (CD) (see Table 2). Such formulation and processing variables have a high influence on the ζ value. With regards to the processing variables, the energy supply provided to the nanoemulsion with the cycles and the high pressure causes the ζ to increase or decrease. The ζ increases when the supplied energy generates electrostatic charges strong enough to avoid coalescence because at values of ζ higher than +30 mV or more negative than −30 mV, they are usually considered stable due to the repulsive forces that are predominant in the nanoemulsion system (Salvia-Trujillo, Rojas-Graü, SolivaFortuny, & Martín-Belloso, 2015a). A decrease of the ζ could be attributed to an over-processing or over-energization of the system, in which recoalescence of the new drops formed takes place (Sadeghpour Galooyak & Dabir, 2015). This over-processing phenomenon is observed in Fig. 1, where ζ values closest to zero (less stable) were obtained at 3 microfluidization cycles with low oil concentrations and high whey concentrations (see

Table 3 Experimental optimum conditions obtained by for the response optimization design for ADS and ζ: experimental values versus values of prediction equations. ζ (mV)

ADS (nm) Run 1-opt 2-opt

HOPO concentration (% w/w) 474 20

Whey concentration (% w/w) 1 1

Pressure (psi) 20,000 10,000

Cycles 2 3

Experimental

Model

Experimental

Model

356.5 1787.0

354.0 1938.2

−44.1 −47.2

−43.8 −49.1

80

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

Fig. 1. Isoplots for ajusted variable ADS to the response optimization design: (a) 1 cycle, (b) 2 cycle y (c) 3 cycles of microfluidization.

Fig. 3c). This result could be related with whey denaturation at 3 cycles of microfluidization due to the accumulation of heat in the system, which can cause aggregation, flocculation, phase separation or destabilization of the emulsions (Dissanayake & Vasiljevic, 2009) so the variation might be due to changes in chemical alterations such as formation or decomposition of the droplets (P.–H. Li & Lu, 2016). Regarding the formulations, in general, the ζ is more negative when the oil concentration increases and the whey concentration decreases for all the microfluidization cycles, and this effect can be observed in Table 4 because such factors positively affect the equation decreasing the ζ values. In addition, when Tween 20 is used for production of nanoemulsions, it generates a negative charge due to the adsorption of hydroxyl ions at the O/W interfaces and to hydrogen bonds that form EO groups between the Tween 20 and hydroxyl ions (Xin et al., 2013). Finally, the minimum PDI and ADS values were obtained when working with 2 microfluidization cycles according to the prediction equations of Table 4, which suggest that 2 microfluidization cycles generate optimum conditions in these two response variables. Sadeghpour Galooyak and Dabir (2015) studied the emulsification by microfluidization of O/W emulsions through the response surface methodology, and they found that both the drop diameter and the PDI improve when the number of steps was increased from 2 to 4 cycles. However, in our study, in 3 microfluidization cycles, the nanoemulsions presented similar ADS and PDI values, but those containing high concentrations of whey, due to the over-processing and initiation of whey protein denaturation, had lower stabilities because the nanoemulsion exit temperature increases with an increase in the number of steps by the microfluidizer, even up to 60 °C. Such a phenomenon was discussed in Section 3.2. 3.2. Trends of ADS (a) y ζ (b) in function of the temperature for the two optimum nanoemulsions The temperature effects on the ADS and the ζ were measured for the two optimum emulsions with regards to the stability for the minimum

and maximum ADS. These effects were investigated because the drop size determines properties such as palatability in food products (Kimura, Uchida, Kanada, & Namiki, 2015); in addition, temperature is one of the most important variables with regards to the processing and quality of foods (W. R. Kim, Aung, Chang, & Makatsoris, 2015). Therefore, it is important to establish the behavior of the HOPO nanoemulsions when this processing variable changes because in commercial applications, the influence of the thermal treatments is indicated in the nanoemulsion stability (Komaiko & McClements, 2014). The sweep was performed up to 70 °C because that is the maximum average temperature to which most foods are subjected (Rynne, Beresford, Kelly, & Guinee, 2004). Fig. 4 shows the ADS trends for the two optimum nanoemulsions with values in ranges from 350 to 430 nm and 890 to 1229 nm for the 1-optimum and 2-optimum runs, respectively. Additionally, a trend was observed for the ADS, where the two emulsions slowly varied in size up to a temperature of 50 °C. Above this temperature, they presented a significant decrease in the drop size, up to a temperature of 60 °C, where the minimum ADS was observed for the two runs. However, a trend towards increasing the particle size up to 2-fold, for the average of the minimum ADS, was observed at temperatures above 60 °C. Finally, these nanoemulsions were observed by CLSM immediately after its preparation; images showed differences between drop sizes due to HOPO concentration and conditions of the microfluidization process (Fig. 4). With regards to the ζ, as observed in Fig. 3a and b, the optimum runs showed a linear trend that varied between ranges from − 40.5 to − 12.9 mV and from − 48.9 to − 25 mV for the 1-optimum and 2-optimum runs, respectively. Therefore, the behavior of both ADS and ζ are related to whey protein denaturation and the change in emulsifier solubility. Constant increases in the ADS and ζ could be related to the denaturation of whey proteins because such denaturation takes place at temperatures close to 72 °C, as reported by (Rynne et al., 2004), who studied the denaturation of whey proteins throughout a range of

Fig 2. Isoplots for ajusted variable ζ to the response optimization design: (a) 1 cycle, (b) 2 cycle y (c) 3 cycles of microfluidization.

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

81

Fig. 3. Trends of ADS (a) y ζ (b) in function of the temperature for the two optimum nanoemulsions.

temperatures (72, 77, 82 and 87 °C) and obtained a partial denaturation of 2.79% of the total protein at 72 °C. However, the denaturation effect can be analyzed from some of the globular proteins, which constitute 20% of the whey proteins (Delavari et al., 2015), and especially from α-lactalbumin, which can present refolding of its conformation at temperatures b90 °C (Lam & Nickerson, 2015). Lam and Nickerson (2015) studied the effects of the pH and temperature in several types of α-lactalbumin, concluding that temperature increases allowed a higher hydrophobicity of the surface, causing changes in the surface charge (ζ). On the other hand, another major protein in whey is β-lactoglobulin, which contains two intramolecular disulfide bonds and a free thiol group that is hidden in the native state. This thiol group becomes accessible during heating due to the unfolding of the native whey proteins (Wolz & Kulozik, 2015). In addition, Dybowska (2011) studied the effect of thermal treatment on the physical properties of the whey protein concentrate dispersions, and the resulting protein was 30% stabilized in the O/W emulsion. It was observed that most of the changes in the molecular mass of the aggregates formed under heating took place around the denaturation temperature of the whey proteins at around 75 °C. In addition, the stability percentage of the emulsions presented the lowest value at 60 °C. According to the above information, it is possible to claim that the initiation of the whey powder denaturation occurred at temperatures above 60 °C, causing agglomeration of the drops, and that the change in ζ is due to the unfolding or refolding of the proteins that constitute the whey, mainly the globular proteins. However, the Tween 20, by being an ethoxylated surfactant, tends to change its hydrophilicity because it is soluble in water at low temperatures, but its solubility changes towards oil and increases with higher temperature (Hasenhuettl &

Hartel, 2008), which indicates that aggregation was generated in the drops containing HOPO, whey and Tween 20; therefore, two phases could be observed in the nanoemulsions subjected to temperatures above 60 °C. It is also important to highlight the use of surfactant concentrations that are lower than those reported in the literature, which is described by (Yang et al., 2012) for homogenization processes that use microfluidization. Finally, the experimental error by comparison with prediction equations, as observed in Table 3, the ADS and ζ values were close to the values estimated by the equations, with a maximum error of 7.8% and a minimum error of 0.6% for ADS, respectively. The error associated with the maximum ADS is due to the high PDI presented by such an emulsion (PDI = 1) in comparison to the PDI for the minimum ADS (PDI = 0.42).

3.3. Change of the η due to the HOPO nanoemulsion storage temperature The η values for the fresh nanoemulsions and stored (t = 4) are shown in Table 1. For the fresh nanoemulsions, the minimum and maximum η were 0.9 and 93.5 mPa·s, respectively. The nanoemulsions with low viscosity were prepared with the same minimum concentrations of HOPO and whey (1% w/w for both compounds), that is, approximately 98% was water, with a viscosity was of 0.89 mPa·s (Zhang, Bing, & Reineccius, 2015). The highest viscosities were obtained for fresh nanoemulsions with high contents of whey or HOPO. A study reported by Panagopoulou et al. (2015) showed that the viscosities of olive oil emulsions, stabilized with bacterial cellulose and isolated whey protein, with 0.25% wt and 3% wt, respectively, were approximately 100 mPa·s at a shear rate of 100 s−1.

Fig. 4. CSLM of optimum HOPO nanoemulsions with average drop size: (a) minimum (1-opt) and (b) maximum (2-opt).

82

Table 4 Equations for ADS, ζ, a*, b*, L, c*, h°ab, viscosity for fresh and stored nanoemulsions of HOPO.

ADS (nm)= ζ (mV)=

PDI=

L*= a*= b*=

Cab*

h°ab =

Viscosity (mPa.s)=

Viscosity Tamb (mPa.s)= Viscosity Tcool (mPa.s)=

1 cycle

2 cycles −5

2

(14. 50 + 1. 54 × [A]− 0. 027 × [B] − 6. 3E × [C]) −57 . 83 − 0 . 468 × [A] +0 .841 × [B] +0 . 003 × [C] + 0 . 026 × [AB] − 1 .55E−5 × [AC]− 5 . 80 × [BC]+ 0 . 006 × [A2] − 0 . 033 × [B2]− 9 .88E−8 × [C2] 0.33+0.08×[A]−0.005×[B]−4.05E−5 ×[C]−0.001× [AB]−6.21E−7 ×[AC]−9.75E−7 ×[BC]−0.002×[A2]+0.001× [B2]+2.09E−9 ×[C2] 61 . 051 + 0 .053 × [A]− 0 . 275 × [B]− 0 .0002 × [C] 13 . 193 + 0 .464 × [A]+ 0 . 380 × [B]+ 0 .0002 × [C] 73 . 426 + 1 .118 × [A]− 0 . 565 × [B]− 0 .002 × [C]− 0 . 004 × [AB] +3 .109E−6 × [AC] +2 . 359E−5 × [BC]− 0 . 045 × [A2] + 0 . 0015 × [B2] +5 .847E−8 × [C2] 74 . 202 + 1 .027 × [A]− 0 . 629 × [B] − 0 .002 × [C]− 0 . 001 × [AB] +5 .199E−6 × [AC] +2 . 449E−5 × [BC]− 0 . 043 × [A2] + 0 .002 × [B2]+ 5 .718E−8 × [C2] 106.465−0.742×[A]−0.311×[B]−0.001×[C]+0.006× [AB]+3.070E−5 ×[AC]−7.691E−6 ×[BC]+0.026×[A2]+0.010× [B2]+4.747E−8 ×[C2] (−15. 16 − 0. 65 × [A]+ 0. 41 × [B] +0. 002 × [C]+ 0. 002 × [AB] +1. 22E−5 × [AC] − 1. 94E−5 × [BC] + 0. 021 × [A2] + 0. 003 × [B2]− 6. 91E−8 × [C2])2 (0. 19 +0. 04 × [A] + 0. 13 × [B] +0. 003 × [C] + 0.04 × [AB] − 9. 58E−6 × [AC] − 2. 08E−5 × [BC])2 (−3. 73 − 0. 27 × [A] +0. 28 × [B]+ 6. 86E−4 × [C]+ 0. 002 × [AB] +1. 22E−5 × [AC] − 1. 35E−5 × [BC] + 0. 019 × [A2] +0. 015 × [B2] − 2. 50E−8 × [C2])2

3 cycles −5

2

(12.80+1.542×[A]−0.027×[B]−6.3E ×[C]) −46.06−0.482×[A]+0.844×[B]+0.002×[C]+0.026×[AB]−1.55E−5 × [AC]−5.80×[BC]+0.006×[A2]−0.033×[B2]−9.88E−8 ×[C2] 0.204+0.10×[A]−0.009×[B]−3.51E−5 ×[C]−0.001×[AB]−6.21E−7 × [AC]−9.75E−7 ×[BC]−0.002×[A2]+0.001×[B2]+2.09E−9 ×[C2] 58 . 553 + 0 . 053 × [A] − 0 . 275 × [B] − 0 . 0002 × [C] −11 . 264 + 0 . 464 × [A] + 0 . 380 × [B] + 0 . 0002 × [C] 77.512+1.243×[A]−0.625×[B]−0.002×[C]−0.004×[AB]+3.109E−6 × [AC]+2.359E−5 ×[BC]−0.045×[A2]+0.0015×[B2]+5.847E−8 ×[C2] 78.629+1.1540×[A]−0.679×[B]−0.002×[C]−0.001×[AB]+5.199E−6 × [AC]+2.449E−5 ×[BC]−0.043×[A2]+0.002×[B2]+5.718E−8 ×[C2] 105.608−0.524×[A]−0.648×[B]−0.001×[C]+0.006×[AB]+3.070E−5 × [AC]−7.691E−6 ×[BC]+0.026×[A2]+0.010×[B2]+4.747E−8 ×[C2] (−4. 82 − 0. 69 × [A]+ 0. 34 × [B] + 0. 001 × [C]+ 0. 002 × [AB]+ 1. 22E−5 × [AC] − 1. 94E−5 × [BC] + 0. 021 × [A2] + 0. 003 × [B2]− 6. 91E−8 × [C2])2 (5. 22 + 0. 09 × [A]+ 0. 37 × [B] − 2. 47E−5 × [C] + 0. 04 × [AB] − 9. 58E−6 × [AC] − 2. 08E−5 × [BC])2 (−9.57−0.11×[A]−0.008×[B]+0.001×[C]+0.002×[AB]−1.73E−6 × [AC]−1.35E−5 ×[BC]+0.019×[A2]+0.015×[B2]−2.50E−8 ×[C2])2

(13.96+1.542×[A]−0.027×[B]−6.3E−5 ×[C])2 −64 . 38 − 0 . × [A]+ 0 . 844 × [B]+ 0 . 002 × [C]+ 0 . 026 × [AB] − 1 . 55E−5 × [AC] − 5 .80 × [BC] + 0 . 006 × [A2] − 0 . 033 × [B2] − 9 . 88E−8 × [C2] 0.57+0.10×[A]−0.007×[B]−6.51E−5 ×[C]−0.001× [AB]−6.21E−7 ×[AC]−9.75E−7 ×[BC]−0.002×[A2]+0.001× [B2]+2.09E−9 ×[C2] 59.741+0.053×[A]−0.275×[B]−0.0002×[C] −11 . 139 + 0 . 464 × [A] + 0 . 380 × [B] + 0 . 0002 × [C] 73.526+1.2224×[A]−0.430×[B]−0.002×[C]−0.004× [AB]+3.109E−6 ×[AC]+2.359E−5 ×[BC]−0.045× [A2]+0.0015×[B2]+5.847E−8 ×[C2] 74 . 092 + 1 . 125 × [A] − 0 . 495 × [B] −0 . 002 × [C] − 0 . 001 × [AB]+ 5 . 199E−6 × [AC]+ 2 . 449E−5 × [BC] − 0 . 043 × [A2]+ 0 . 002 × [B2]+ 5 . 718E−8 × [C2] 107 . 920 − 0 . 546 × [A] − 0 . 482 × [B] − 0 . 001 × [C]+ 0 . 006 × [AB]+ 3 . 070E−5 × [AC] − 7 . 691E−6 × [BC] + 0 . 026 × [A2]+ 0 . 010 × [B2]+ 4 . 747E−8 × [C2] (−10. 03 − 0. 57 × [A] + 0. 32 × [B]+ 0. 001 × [C] + 0. 002 × [AB]+ 1. 22E−5 × [AC] − 1. 94E−5 × [BC]+ 0. 021 × [A2]+ 0. 003 × [B2] − 6. 91E−8 × [C2])2 (−7. 61 + 0. 19 × [A]+ 0. 57 × [B] − 6. 86E−4 × [C] + 0. 04 × [AB]− 9. 58E−6 × [AC] − 2. 08E−5 × [BC])2 (−4. 23 − 0. 101 × [A] + 0. 008 × [B] − 7. 66E−4 × [C] + 0. 002 × [AB]− 1. 73E−6 × [AC] − 1. 35E−5 × [BC]+ 0. 019 × [A2]+ 0. 015 × [B2] − 2. 50E−8 × [C2])2

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

Equations for…

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

The independent factors were the whey concentration, the microfluidization cycles, the HOPO concentration (A2) and the squared pressure (C2), as well as the interactions between the whey and pressure (BC) and the pressure and cycles (CD). The fresh nanoemulsion viscosity increased when the whey concentration increased. However, the microfluidization cycles showed that for 1 and 3 cycles, the maximum viscosities obtained were 60 mPa·s for high whey concentrations. This effect was more prominent for 2 microfluidization cycles because the maximum viscosities obtained were higher than 80 mPa·s. Li, Ma, and Cui (2014) studied curcumin nanoemulsions stabilized by isolated whey proteins and observed that the viscosity increased at higher protein concentrations. The storage temperature of nanoemulsions in ranges between 5 and 40 °C has been studied by many authors (Hategekimana, Chamba, Shoemaker, Majeed, & Zhong, 2015a; Silva, Cerqueira, & Vicente, 2015). The nanoemulsions stored at room temperature (Troom) showed η values between 1.9 and 553.3 mPa·s, for which the independent factors were the HOPO and whey concentrations, the cycles and the interaction between the concentrations of HOPO and whey (AB). The viscosity showed a proportional increase at high concentrations of whey and oil; this trend was similar for all microfluidization cycles, and the effects of the high concentrations become more prominent when the nanoemulsions were run through more cycles. The ADS is directly related to the viscosity obtained because when the HOPO concentration increases to where the microfluidizer becomes inefficient, larger drops that are susceptible to destabilization (Ozturk, Argin, Ozilgen, & McClements, 2014) are generated, and therefore, the viscosity increases. The nanoemulsions stored at Trefrig showed η values between 0.88 and 112.2 mPa·s, for which the significant factors were the concentrations of HOPO and whey, these concentrations squared and the interactions between the whey and cycles and the cycles and pressure. The viscosities for the nanoemulsions stored at Trefrig increased with increasing HOPO and whey concentrations. However, a low-viscosity zone is obtained for whey and HOPO concentrations below 10% w/w. These results showed the temperature effects on the aggregation and stabilization of the nanoemulsions during their storage for 4 days. The viscosity values increased approximately 5.9 and 1.2-fold at Troom and Tcool, respectively, with respect to the fresh nanoemulsions, which indicates that removing heat and maintaining the nanoemulsion at a refrigeration temperature improves the stability because the lack of drop aggregation or coalescence could probably be related to the absence of electrostatic projection, thus maintaining the repulsion between drops (Hategekimana, Chamba, Shoemaker, Majeed, & Zhong, 2015b). In addition, at low storage temperatures, the rate and frequency of collisions between drops is lower, which reduces the growth of the nanoemulsion drop size (Adjonu et al., 2014). This temperature effect on the storage at 4 and 25 °C was studied by Adjonu et al. (2014) in O/W nanoemulsions stabilized with whey protein and its hydrolysates, in which a greater cremation was observed in the O/W nanoemulsions stored at 25 °C compared to those stored at 4 °C. In general, all of the independent factors significantly contribute to the viscosities, whether the samples are fresh or stored at Troom or Tcool because the total equation is elevated to the second power for these response variables, where the concentrations of HOPO and whey contribute more significantly, as observed in Table 4. 3.4. Microfluidization effect on the color parameters of the HOPO nanoemulsions The color of an object can be described by the coordinates L*, a*, b*, Cab* and h°ab, where L* is the luminosity, a* is the shift between red and green, b* is the shift between yellow and blue (Khan et al., 2009), Cab* describes the purity of a color, that is its saturation (Gil-Muñoz, Gómez-Plaza, Martı́nez, & López-Roca, 1997) and h°ab is the color hue (Tuberoso et al., 2014). Taking into account the CIELAB coordinates

83

above, the color of a nanoemulsion is generated when a beam of light strikes the oil drops and scatters the light. This depends to a great extent on the oil concentration (Salvia-Trujillo, Rojas-Graü, Soliva-Fortuny, & Martín-Belloso, 2015b); however, it is not the only factor that influences its appearance. According to the prediction theory of McClements, 2002b, L* is influenced by the drop size, giving higher luminosity values when the drop diameter increases, generating a greater scattering of light. In contrast, the smaller particles scatter less light. This phenomenon was observed in the study by Salvia-Trujillo, Rojas-Graü, Soliva-Fortuny, and MartínBelloso (2013), where lemon essential oil was used as lipid base and sodium alginate was used as an aqueous phase. However, in our results (Table 1), it is observed that the L* values presented parabolic behavior, that is, when increasing the drop size, the luminosity increases until reaching a maximum point, where it starts to decrease. This phenomenon was studied by McClements, 2002a, where it was concluded that the Cab* has an inversely proportional behavior with respect to L*. The significant independent factors were determined for all of the color response variables, which were the HOPO and whey concentrations; this result occurs because palm oil presents a dark red to brown red color, and whey is a cream color. Therefore, the qualitative color obtained for the nanoemulsions was yellow. The a* values shown in Table 1 ranged from 9.87 to −9.53. Because this variable linearly varied when increasing the HOPO and whey concentrations, the values tended to be more positive, which represents a shift towards red. Regarding b*, the range in which it varied was relatively small, between 49.90 and 66.91 (see Table 1), due to the qualitatively yellow color of the nanoemulsions. The b* values shift towards the blue (minimum values of b*) when the whey concentrations are high and the HOPO concentrations are low. On the other hand, the Cab* coordinate presents a behavior similar to b* because the variable indicates the contribution of a* and b* (Gil-Muñoz et al., 1997), as observed in Eq. (1) where the b* values predominate. These values represent a mean yellow saturated color for the HOPO nanoemulsions. The response variable L* presented values between 47.46 and 59.00, which are considered intermediate luminosity values because the L* interval can take values between 0 and 100 (León, Mery, Pedreschi, & León, 2006), due to the effect of the homogenization between the HOPO and whey. L* decreased when the whey concentration increased for any HOPO concentration. This effect is generated by the interaction between the HOPO and the whey during microfluidization, where the initial color characteristics are lost because the L* values for the unmicrofluidized HOPO and the whey are 30.66 and 66.12, respectively. Finally, h°ab presented values between 80.47° and 99.21°, which indicate high purity in the yellow hue because the values are close to 90° (Tuberoso et al., 2014); in addition, the HOPO concentration generates a shift of the h°ab towards a more red hue. According to the prediction equations of Table 4, it was determined that the factors that significantly contribute to all of the color parameters are the HOPO and whey concentrations. However, for the coordinates b* and Cab*, the HOPO concentration predominates, while the whey concentration predominates for L*, a* and h°ab. In addition, it was verified that the maximum value for the coordinates b* and Cab* can be obtained in 2 microfluidization cycles, with more yellow and saturated nanoemulsions acquired. Furthermore, the optimum minimum value of h°ab can be obtained in 2 cycles, reaching a more pure hue of the yellow color. However, with 3 microfluidization cycles, a maximum value for the coordinate a* is achieved, with a tendency towards a red color. Finally, a higher L* value, whose optimum condition is that of highest clarity (a higher value), is obtained from 1 cycle. 4. Conclusions This work was focused on obtaining the most favorable microfluidization, formation and storage conditions for the nanoemulsions obtained from HOPO and whey for future applications in the food

84

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85

industry. Through the use of response optimization methodologies, this study demonstrates a useful tool for obtaining reliably predictable variable behaviors and equations. The results showed that minimum ADS, PDI and ζ values can be generated through microfluidization. The optimum number of microfluidization cycles to obtain the minimum ADS was 2 at high pressure because the coalescence effect is avoided. In addition, low concentrations of HOPO and whey helped produce stable nanoemulsions. On the other hand, a low storage temperature improved the nanoemulsion stability because the viscosity did not significantly change with respect to the viscosities of the fresh nanoemulsions. Finally, denaturation was indicated at 60 °C because both the ADS and the ζ increased, causing the destabilization and aggregation of the optimized nanoemulsion drops. Acknowledgments The authors would like to thank to Cenipalma, Colombia for kindly supplying the High Oleic Palm Oil used in this study and to Alexandra Mondragón Serna, Leader of the project of Health and Nutrition of Cenipalma. References Adjonu, R., Doran, G., Torley, P., & Agboola, S. (2014). Formation of whey protein isolate hydrolysate stabilised nanoemulsion. Food Hydrocolloids, 41, 169–177. http://dx.doi. org/10.1016/j.foodhyd.2014.04.007. Akoh, C. C. (2005). Handbook of functional lipids. CRC Press Retrieved from https://books. google.com.co/books?id=BIjPC4xwBKsC Barradas, T. N., de Campos, V. E. B., Senna, J. P., Coutinho, C. dos S. C., Tebaldi, B. S., Silva, K. G. de H. e, & Mansur, C. R. E. (2014). Development and characterization of promising o/w nanoemulsions containing sweet fennel essential oil and non-ionic sufactants. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 480, 214–221. http://dx.doi.org/10.1016/j.colsurfa.2014.12.001. Das, B., Sarkar, S., Sarkar, A., Bhattacharjee, S., & Bhattacharjee, C. (2015). Recovery of whey proteins and lactose from dairy waste: A step towards green waste management. Process Safety and Environmental Protection. http://dx.doi.org/10.1016/j.psep. 2015.05.006 (0), In press, Accepted Manuscript. Degner, B. M., Chung, C., Schlegel, V., Hutkins, R., & McClements, D. J. (2014). Factors influencing the freeze-thaw stability of emulsion-based foods. Comprehensive Reviews in Food Science and Food Safety, 13(2), 98–113. http://dx.doi.org/10.1111/ 1541-4337.12050. Delavari, B., Saboury, A. A., Atri, M. S., Ghasemi, A., Bigdeli, B., Khammari, A., ... Goliaei, B. (2015). Alpha-lactalbumin: A new carrier for vitamin D3 food enrichment. Food Hydrocolloids, 45, 124–131. http://dx.doi.org/10.1016/j.foodhyd.2014.10.017. Dissanayake, M., & Vasiljevic, T. (2009). Functional properties of whey proteins affected by heat treatment and hydrodynamic high-pressure shearing. Journal of Dairy Science, 92(4), 1387–1397. http://dx.doi.org/10.3168/jds.2008-1791. Dybowska, B. E. (2011). Whey protein-stabilized emulsion properties in relation to thermal modification of the continuous phase. Journal of Food Engineering, 104(1), 81–88. http://dx.doi.org/10.1016/j.jfoodeng.2010.11.030. Esfanjani, A. F., Jafari, S. M., Assadpoor, E., & Mohammadi, A. (2015). Nano-encapsulation of saffron extract through double-layered multiple emulsions of pectin and whey protein concentrate. Journal of Food Engineering, 165, 149–155. http://dx.doi.org/10. 1016/j.jfoodeng.2015.06.022. Gil-Muñoz, R., Gómez-Plaza, E., Martı́nez, A., & López-Roca, J. (1997). Evolution of the CIELAB and other spectrophotometric parameters during wine fermentation. Influence of some pre and postfermentative factors. Food Research International, 30(9), 699–705. http://dx.doi.org/10.1016/S0963-9969(98)00029-5. Hasenhuettl, G. L., & Hartel, R. W. (2008). Food emulsifiers and their applications. Springer New York Retrieved from https://books.google.com.co/books?id=Bs1euQO8ZJUC Hategekimana, J., Chamba, M. V. M., Shoemaker, C. F., Majeed, H., & Zhong, F. (2015a). Vitamin E nanoemulsions by emulsion phase inversion: Effect of environmental stress and long-term storage on stability and degradation in different carrier oil types. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 483, 70–80. http:// dx.doi.org/10.1016/j.colsurfa.2015.03.020. Hategekimana, J., Chamba, M. V. M., Shoemaker, C. F., Majeed, H., & Zhong, F. (2015b). Vitamin E nanoemulsions by emulsion phase inversion: Effect of environmental stress and long-term storage on stability and degradation in different carrier oil types. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 483, 70–80. http:// dx.doi.org/10.1016/j.colsurfa.2015.03.020. Hernández-Hernández, H. M., Chanona-Pérez, J. J., Calderón-Domínguez, G., Perea-Flores, M. J., Mendoza-Pérez, J. a., Vega, A., ... Farrera-Rebollo, R. R. (2014). Evaluation of agave fiber delignification by means of microscopy techniques and image analysis. Microscopy and Microanalysis, 20(05), 1436–1446. http://dx.doi.org/10.1017/ S1431927614012987. Jo, Y. J., & Kwon, Y. J. (2014). Characterization of β-carotene nanoemulsions prepared by microfluidization technique. Food Science and Biotechnology, 23(1), 107–113. http:// dx.doi.org/10.1007/s10068-014-0014-7.

Joye, I. J., Davidov-Pardo, G., & McClements, D. J. (2014). Nanotechnology for increased micronutrient bioavailability. Trends in Food Science & Technology. http://dx.doi.org/ 10.1016/j.tifs.2014.08.006. Khan, M. A. I., Ueno, K., Horimoto, S., Komai, F., Someya, T., Inoue, K., ... Ono, Y. (2009). CIELAB color variables as indicators of compost stability. Waste Management, 29(12), 2969–2975. http://dx.doi.org/10.1016/j.wasman.2009.06.021. Kim, I. -H., Oh, Y. A., Lee, H., Song, K. Bin, & Min, S. C. (2014). Grape berry coatings of lemongrass oil-incorporating nanoemulsion. LWT - Food Science and Technology, 58(1), 1–10. http://dx.doi.org/10.1016/j.lwt.2014.03.018. Kim, W. R., Aung, M. M., Chang, Y. S., & Makatsoris, C. (2015). Freshness gauge based cold storage management: A method for adjusting temperature and humidity levels for food quality. Food Control, 47, 510–519. http://dx.doi.org/10.1016/j.foodcont.2014. 07.051. Kimura, S., Uchida, S., Kanada, K., & Namiki, N. (2015). Effect of granule properties on rough mouth feel and palatability of orally disintegrating tablets. International Journal of Pharmaceutics, 484(1–2), 156–162. http://dx.doi.org/10.1016/j.ijpharm. 2015.02.023. Kodali, D. R. (2014). Trans Fats: Replacement Solutions. AOCS Retrieved from https:// books.google.com.co/books?id=hTZfCgAAQBAJ Komaiko, J., & McClements, D. J. (2014). Optimization of isothermal low-energy nanoemulsion formation: Hydrocarbon oil, non-ionic surfactant, and water systems. Journal of Colloid and Interface Science, 425, 59–66. http://dx.doi.org/10.1016/j.jcis. 2014.03.035. Komaiko, J., & McClements, D. J. (2015). Low-energy formation of edible nanoemulsions by spontaneous emulsification: Factors influencing particle size. Journal of Food Engineering, 146, 122–128. http://dx.doi.org/10.1016/j.jfoodeng.2014.09.003. Lam, R. S. H., & Nickerson, M. T. (2015). The effect of pH and temperature pre-treatments on the structure, surface characteristics and emulsifying properties of alphalactalbumin. Food Chemistry, 173, 163–170. http://dx.doi.org/10.1016/j.foodchem. 2014.09.078. León, K., Mery, D., Pedreschi, F., & León, J. (2006). Color measurement in L∗a∗b∗ units from RGB digital images. Food Research International, 39(10), 1084–1091. http://dx. doi.org/10.1016/j.foodres.2006.03.006. Li, M., Ma, Y., & Cui, J. (2014). Whey-protein-stabilized nanoemulsions as a potential delivery system for water-insoluble curcumin. LWT - Food Science and Technology, 59(1), 49–58. http://dx.doi.org/10.1016/j.lwt.2014.04.054. Li, P. -H., & Lu, W. -C. (2016). Effects of storage conditions on the physical stability of dlimonene nanoemulsion. Food Hydrocolloids, 53, 218–224. http://dx.doi.org/10. 1016/j.foodhyd.2015.01.031. Li, Y., Zheng, J., Xiao, H., & McClements, D. J. (2012). Nanoemulsion-based delivery systems for poorly water-soluble bioactive compounds: Influence of formulation parameters on polymethoxyflavone crystallization. Food Hydrocolloids, 27(2), 517–528. http://dx.doi.org/10.1016/j.foodhyd.2011.08.017. Liu, J., Shim, Y. Y., Wang, Y., & Reaney, M. J. T. (2015). Intermolecular interaction and complex coacervation between bovine serum albumin and gum from whole flaxseed (Linum usitatissimum L.). Food Hydrocolloids, 49, 95–103. http://dx.doi.org/10.1016/j. foodhyd.2015.02.035. Malvern Instruments Limited (2011). Dynamic Light Scattering Common Terms Defined. Retrieved from http://www.biophysics.bioc.cam.ac.uk/wp-content/uploads/2011/02/ DLS_Terms_defined_Malvern.pdf Malvern Instruments Limited. (2015). Studying the stability of coating particles using the Zetasizer Nano. McClements, D. J. (2002a). Colloidal basis of emulsion color. Current Opinion in Colloid and Interface Science, 7(5–6), 451–455. http://dx.doi.org/10.1016/S1359-0294(02)00075-4. McClements, D. J. (2002b). Theoretical prediction of emulsion color. Advances in Colloid and Interface Science, 97(1–3), 63–89. http://dx.doi.org/10.1016/S0001-8686(01)00047-1. McClements, D. J. (2004). Food Emulsions: Principles, Practice, and Techniques. Taylor & Francis Retrieved from https://books.google.com.co/books?id=huCSbPwyvBoC Muangrat, R., Onwudili, J. A., & Williams, P. T. (2011). Alkaline subcritical water gasification of dairy industry waste (whey). Bioresource Technology, 102(10), 6331–6335. http://dx.doi.org/10.1016/j.biortech.2011.02.056. Munévar, L. E. Z. (2010). Situación y perspectivas del aceite de Palma alto oleico O x G en Colombia. Revista Palmas, 31, 349–353. Ozturk, B., Argin, S., Ozilgen, M., & McClements, D. J. (2014). Formation and stabilization of nanoemulsion-based vitamin E delivery systems using natural surfactants: Quillaja saponin and lecithin. Journal of Food Engineering, 142, 57–63. http://dx.doi.org/10. 1016/j.jfoodeng.2014.06.015. Panagopoulou, E., Tsouko, E., Kopsahelis, N., Koutinas, A., Mandala, I., & Evageliou, V. (2015). Olive oil emulsions formed by catastrophic phase inversion using bacterial cellulose and whey protein isolate. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 486, 203–210. http://dx.doi.org/10.1016/j.colsurfa.2015.09.056. Poynton, C. (2003). Digital video and HDTV. Digital Video and HDTV: Elsevier. http://dx. doi.org/10.1016/B978-155860792-7/50087-2. Qian, C., & McClements, D. J. (2011). Formation of nanoemulsions stabilized by model foodgrade emulsifiers using high-pressure homogenization: Factors affecting particle size. Food Hydrocolloids, 25(5), 1000–1008. http://dx.doi.org/10.1016/j.foodhyd.2010.09.017. Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2012). Physical and chemical stability of β-carotene-enriched nanoemulsions: Influence of pH, ionic strength, temperature, and emulsifier type. Food Chemistry, 132(3), 1221–1229. http://dx.doi.org/10.1016/j. foodchem.2011.11.091. Quintanilla-Carvajal, M. X., Arenas-Ocampo, M. lucía, Campos-Mendiola, R., CamachoDíaz, B. H., & Jiménez-Aparicio, A. R. (2015). Caracterización morfométrica de estructuras florales de Tagetes erecta L. Y Tagetes patula L. (Asteraceae) utilizando análisis digital de imágenes y dimensión fractal. Gayana Botánica, 72(1), 137–144. Quintanilla-Carvajal, M. X., Camacho-Díaz, B. H., Meraz-Torres, L. S., Chanona-Pérez, J. J., Alamilla-Beltrán, L., Jimenéz-Aparicio, A., & Gutiérrez-López, G. F. (2009).

L. Ricaurte et al. / Innovative Food Science and Emerging Technologies 35 (2016) 75–85 Nanoencapsulation: A new trend in food engineering processing. Food Engineering Reviews, 2(1), 39–50. http://dx.doi.org/10.1007/s12393-009-9012-6. Quintanilla-Carvajal, M. X., Hernández-Sánchez, H., Alamilla-Beltrán, L., Zepeda-Vallejo, G. , Jaramillo-Flores, M. E., de Jesús Perea-Flores, M., ... Gutiérrez-López, G. F. (2014). Effects of microfluidisation process on the amounts and distribution of encapsulated and non-encapsulated α-tocopherol microcapsules obtained by spray drying. Food Research International, 63, 2–8. http://dx.doi.org/10.1016/j.foodres.2014.05.025. Rodríguez-Bernal, J. M., Tello, E., Flores-Andrade, E., De Jesús Perea-Flores, M., VallejoCardona, A., Gutiérrez-López, G. F., & Quintanilla-Carvajal, M. X. (2015). Effect of borojo (Borojoa patinoi Cuatrecasas) three-phase composition and gum arabic on the glass transition temperature, (january). http://dx.doi.org/10.1002/jsfa.7190. Rynne, N. M., Beresford, T. P., Kelly, A. L., & Guinee, T. P. (2004). Effect of milk pasteurization temperature and in situ whey protein denaturation on the composition, texture and heat-induced functionality of half-fat Cheddar cheese. International Dairy Journal, 14(11), 989–1001. http://dx.doi.org/10.1016/j.idairyj.2004.03.010. Sadeghpour Galooyak, S., & Dabir, B. (2015). Three-factor response surface optimization of nano-emulsion formation using a microfluidizer. Journal of Food Science and Technology, 52(5), 2558–2571. http://dx.doi.org/10.1007/s13197-014-1363-1. Salvia-Trujillo, L., Rojas-Graü, A., Soliva-Fortuny, R., & Martín-Belloso, O. (2015a). Physicochemical characterization and antimicrobial activity of food-grade emulsions and nanoemulsions incorporating essential oils. Food Hydrocolloids, 43(0), 547–556. http://dx.doi.org/10.1016/j.foodhyd.2014.07.012. Salvia-Trujillo, L., Rojas-Graü, M. A., Soliva-Fortuny, R., & Martín-Belloso, O. (2013). Effect of processing parameters on physicochemical characteristics of microfluidized lemongrass essential oil-alginate nanoemulsions. Food Hydrocolloids, 30(1), 401–407. http://dx.doi.org/10.1016/j.foodhyd.2012.07.004. Salvia-Trujillo, L., Rojas-Graü, M. A., Soliva-Fortuny, R., & Martín-Belloso, O. (2015b). Use of antimicrobial nanoemulsions as edible coatings: Impact on safety and quality attributes of fresh-cut Fuji apples. Postharvest Biology and Technology, 105, 8–16. http://dx. doi.org/10.1016/j.postharvbio.2015.03.009. Sessa, M., Balestrieri, M. L., Ferrari, G., Servillo, L., Castaldo, D., D'Onofrio, N., ... Tsao, R. (2014). Bioavailability of encapsulated resveratrol into nanoemulsion-based delivery systems. Food Chemistry, 147, 42–50. http://dx.doi.org/10.1016/j.foodchem.2013.09.088. Shen, L., & Tang, C. -H. (2012). Microfluidization as a potential technique to modify surface properties of soy protein isolate. Food Research International, 48(1), 108–118. http://dx.doi.org/10.1016/j.foodres.2012.03.006. Silva, H. D., Cerqueira, M. A., Souza, B. W. S., Ribeiro, C., Avides, M. C., Quintas, M. A. C., ... Vicente, A. A. (2011). Nanoemulsions of β-carotene using a high-energy emulsification–evaporation technique. Journal of Food Engineering, 102(2), 130–135. http://dx.doi.org/10.1016/j.jfoodeng.2010.08.005.

85

Silva, H. D., Cerqueira, M. A., & Vicente, A. A. (2015). Influence of surfactant and processing conditions in the stability of oil-in-water nanoemulsions. Journal of Food Engineering, 167, 89–98. http://dx.doi.org/10.1016/j.jfoodeng.2015.07.037 Part. Sui, X., Bary, S., & Zhou, W. (2016). Changes in the color, chemical stability and antioxidant capacity of thermally treated anthocyanin aqueous solution over storage. Food Chemistry, 192, 516–524. http://dx.doi.org/10.1016/j.foodchem.2015.07.021. Tabibiazar, M., Davaran, S., Hashemi, M., Homayonirad, A., Rasoulzadeh, F., Hamishehkar, H., & Mohammadifar, M. A. (2015). Design and fabrication of a food-grade albuminstabilized nanoemulsion. Food Hydrocolloids, 44, 220–228. http://dx.doi.org/10.1016/ j.foodhyd.2014.09.005. Teo, A., Goh, K. K. T., Wen, J., Oey, I., Ko, S., Kwak, H. -S., & Lee, S. J. (2016). Physicochemical properties of whey protein, lactoferrin and tween 20 stabilised nanoemulsions: Effect of temperature, pH and salt. Food Chemistry, 197, Part, 297–306. http://dx.doi.org/10. 1016/j.foodchem.2015.10.086. Tuberoso, C. I. G., Jerković, I., Sarais, G., Congiu, F., Marijanović, Z., & Kuś, P. M. (2014). Color evaluation of seventeen European unifloral honey types by means of spectrophotometrically determined CIE L*Cab*h(ab)° chromaticity coordinates. Food Chemistry, 145, 284–291. http://dx.doi.org/10.1016/j.foodchem.2013.08.032. Wang, L., Gao, Y., Li, J., Subirade, M., Song, Y., & Liang, L. (2016). Effect of resveratrol or ascorbic acid on the stability of α-tocopherol in O/W emulsions stabilized by whey protein isolate: Simultaneous encapsulation of the vitamin and the protective antioxidant. Food Chemistry, 196, 466–474. http://dx.doi.org/10.1016/j.foodchem.2015.09.071. Wolz, M., & Kulozik, U. (2015). Thermal denaturation kinetics of whey proteins at high protein concentrations. International Dairy Journal, 49, 95–101. http://dx.doi.org/10. 1016/j.idairyj.2015.05.008. Wooster, T. J., Golding, M., & Sanguansri, P. (2008). Impact of oil type on nanoemulsion formation and Ostwald ripening stability. Langmuir : The ACS Journal of Surfaces and Colloids, 24(10), 12758–12765. http://dx.doi.org/10.1021/la801685v. Xin, X., Zhang, H., Xu, G., Tan, Y., Zhang, J., & Lv, X. (2013). Influence of CTAB and SDS on the properties of oil-in-water nano-emulsion with paraffin and span 20/tween 20. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 418, 60–67. http:// dx.doi.org/10.1016/j.colsurfa.2012.10.065. Yang, Y., Marshall-Breton, C., Leser, M. E., Sher, A. A., & McClements, D. J. (2012). Fabrication of ultrafine edible emulsions: Comparison of high-energy and low-energy homogenization methods. Food Hydrocolloids, 29(2), 398–406. http://dx.doi.org/10. 1016/j.foodhyd.2012.04.009. Zhang, J., Bing, L., & Reineccius, G. A. (2015). Formation, optical property and stability of orange oil nanoemulsions stabilized by Quallija saponins. LWT - Food Science and Technology, 64(2), 1063–1070. http://dx.doi.org/10.1016/j.lwt.2015.07.034.