Encapsulation of Anthocyanins in Alginate-Pectin Hydrogel Particles and Modeling the Release at Low and High Ph

Encapsulation of anthocyanins in alginate-pectin hydrogel particles and modeling the release at low and high pH

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Jingxin Guo, M.S.

Graduate Program in Food, Agricultural and Biological Engineering

The Ohio State University

2017

Dissertation Committee:

Dr. Gonul Kaletunc, Advisor

Dr. Dennis Heldman

Dr. Sudhir Sastry

Dr. Monica Giusti

Copyrighted by

Jingxin Guo

2017

Abstract

Due to the possible relationship between the consumption of synthetic dyes in food and their adverse behavioral and neurological effects in some children, there is a growing interest for using natural colorants to replace these artificial dyes.

Anthocyanins (ACNs), for instance, are a group of natural pigments carried in many fruits and vegetables (blueberry, purple sweet potato, red cabbage and many others).

Anthocyanins provide red or purple to blue color at different pH levels. In addition to their visual appearance, ACNs were proposed to have multiple health benefits including anti-inflammatory and anti-carcinogenic activity, preventing cardiovascular disease, and facilitating weight management. However, ACNs have been reported to be sensitive to their ambient environments, including pH, light, oxygen and heat.

One way to protect ACNs against these environmental factors is encapsulation.

Encapsulation is a delivery system for sensitive bioactive compounds (e.g. ACNs) which is usually applied to improve the stability, to maximize the retention, and to control the release at the target locations of the encapsulated bioactive agents.

Encapsulation in biopolymer based hydrogels provides excellent opportunities for the food industry because these carbohydrates are from natural resources and are approved by U.S. Food and Drug Administration for use in food materials.

The major objective of this dissertation was to develop a pH responsive hydrogel system using food-grade polymers to encapsulate ACNs, and to achieve further release of the bioactive compounds at simulated target conditions.

To start with, a biopolymer mixture composed of alginate (Al) and pectin (P) that can form hydrogel were used to generate particles using extrusion method in low pH ii

buffer. We also produced novel disc shaped particles which can potentially enhance the particle adhesion in intestines, our ultimate target location. As the pH increases, Al-P hydrogels go through a gel-sol transition and the dissolution kinetics of the hydrogel dominates the bioactive compound release. According to our study, the volume change of spherical and disc shaped particles showed that the hydrogel particles would be stable in low pH beverages (pH below 3.0) during storage. At pH 5.0 and 7.0, hydrogel particle dissolution due to increased electronic repulsive forces, measured as the particle size change followed a zero-order kinetic model. The 2.8% TGC 43-57 wt%

Al-P disc particles had the fastest and the 2.2% TGC 82-18 wt% Al-P spherical particles had the slowest volume dissolution rate at pH 7.0 and 37°C. Activation energies of hydrogel particles were significantly affected by pH, particle shape and Al to P ratio.

Such a smart biopolymer system which responds to pH provides an opportunity to use food as a vehicle for targeted delivery of bioactive compounds.

Then, encapsulation of purple corn (PC) and blueberry (BB) extracts in alginate- pectin hydrogel particles was achieved to protect the bioactive compound— anthocyanins (ACNs) from degradation. The alginate-pectin hydrogel particles containing PC or BB extracts were produced by extrusion method in pH 1.2 buffer.

Factors including initial ACNs concentration in particle, shaple of particles (disc and sphere), alginate-pectin ratio, total gum concentration (TGC), ACN source, and curing bath conditions were chosen to evaluate their influences on encapsulation efficiency

(EE) of ACNs using mass balance. In general, the initial ACN concentration and particle shape showed no influence on EE, while the alginate-pectin ratio, TGC, ACN source and the pH of the curing bath showed major impact on EE.

iii

Loss of ACNs during short-time storage after encapsulated in Al-P hydrogel particles was also investigated in pH 3.0 buffer. The effect of alginate to pectin ratio,

TGC, particle shape, particle weight to solution volume (wt/vol), ACN source and storage temperature on the EE at equilibrium during storage was evaluated. The higher alginate content particles showed higher EE at equilibrium during storage using PC or

BB extracts, compared with low alginate content particles. High TGC improved the percent EE at equilibrium during storage significantly for both ACN sources, and higher amount of ACN was retained in spherical shaped particles than disc shaped particles.

Also, ACN stability in particles or in solution was evaluated during storage under light or dark condition at 20⁰ C. The degradation of ACN in solution showed first-order kinetics while the degradation of ACN in particle followed zero-order kinetics. The stability of ACN encapsulated in alginate-pectin particles was significantly improved over ACN in solution during storage in both light and dark conditions.

Lastly, the diffusion study of ACNs from the Al-P gel was evaluated. Using a well- mixed and temperature-controlled system, the diffusion behavior of PC and BB ACNs from Al-P hydrogel particles were investigated. The diffusion coefficients for ACN were determined based on the mathematical approach using Fick’s second law. Various experimental conditions were evaluated to assess the behavior of the diffusivity of different initial ACN concentration in particles (20.75, 87.78, 114.96 and 173.23 μg/ml), alginate to pectin ratios (82-18 and 43-57 Al-P), total gum concentrations (2.2% and

2.8%), particle shapes (disc and spherical shaped), ACN sources (PC and BB ACN) and temperatures (4, 24 and 37⁰ C). Results showed the diffusion coefficient of ACNs was significantly affected by the particle shape and the operating temperature, while independent from the initial ACN concentration in particle, alginate to pectin ratio, total gum concentration and ACN source. In all scenarios, the diffusion coefficient

calculated was noticeably smaller than the diffusion coefficient of ACN in pure water calculated by using Wilke & Chang’s equation (1955). Partition coefficient was evaluated for all the studied conditions, and showed strong dependent all the evaluated parameters except the initial ACN concentration in particles. These results can be used to design the hydrogel particles to achieve different equilibrium state during diffusion, and the significant higher activation energy using alginate-pectin hydrogel particles suggested the potential of using this matrix for processes those undergoes strong temperature fluctuation.

Acknowledgement

I would like to thank my advisor, Dr. Gonul Kaletunc. Thank you for providing the opportunity for me to work with you, and trained me to become a scientist capable of designing, conducting and publishing food engineering related researches. I am proud to say that I have spent five and meaningful years to challenge myself and get my work done. I would also like to offer my thanks to my committee members, Dr.

Sudhir Sastry, Dr. Dennis Heldman and Dr. Monica Giusti, thanks very much for your patience, guidance and encourage.

I would also like to thank my family members. My parents, Yuming Guo and

Shuqin Na who raised me and comforted me when I was down. Also, I would like to say thank you to my husband Xu Yang for your unconditional support.

Last but not the least, I am so grateful for my baby girl—Amy Yang, you complete my life as a mother and I want you know that I will love you for a lifetime.

Vita

December 09, 1985……………………………………….…Born, Beijing, China

2004-2008...………………………………….…B.S., Food Science and Engineer China Agricultural University, China

2010-2012………….…..………………………M.S., Food Science and Engineer Illinois Institute of Technology

2012-present………………..………………………Graduate Research Associate The Ohio State University

Publications

Guo, J., & Kaletunç, G. (2016). Dissolution kinetics of pH responsive alginate-pectin hydrogel particles. Food Research International, 88, 129-139.

Fields of Study

Major Field: Food, Agricultural and Biological Engineering

vii

Table of contents

Abstract ...... ii Acknowledgement ...... vi Vita ...... vii Table of contents ...... viii List of Tables ...... xii List of Figures ...... xiii Chapter 1: Literature Review ...... 1 1.1 Anthocyanins (ACNs) ...... 1 1.1.1 Anthocyanins as natural food colorants ...... 1 1.1.2 Structure of Anthocyanins...... 2 1.1.3 The pH effect on Anthocyanin chemical structure ...... 3 1.1.4 Anthocyanins stability ...... 4 1.1.5 Anthocyanin sources ...... 6 1.1.5.1 Purple corn anthocyanin ...... 6 1.1.5.2 Blueberry anthocyanin ...... 7 1.1 Anthocyanin encapsulation systems ...... 7 1.1.1 Spray-drying ...... 8 1.1.2 Ionic gelation ...... 9 1.1.3 Others ...... 11 1.2 Hydrogel based encapsulation delivery system ...... 11 1.3.1 Calcium-alginate hydrogel encapsulation ...... 12 1.3.2 Pectin hydrogel ...... 13 1.3.3 Alginate and pectin hydrogels ...... 15 References ...... 17 Chapter 2: Dissolution Kinetics of pH Responsive Smart Alginate-Pectin Hydrogel Particles ...... 29 Abstract ...... 29 2.1 Introduction ...... 30 2.2 Materials and methods ...... 33 2.2.1 Materials ...... 33 2.2.2 Preparation of gel solutions ...... 33 2.2.3 Particle fabrication ...... 33 2.2.4 Particle size and shape characterization ...... 34 2.2.5 Gel mechanical properties...... 34 2.2.5.1 Sample preparation ...... 34

viii

2.2.5.2 Gel mechanical properties measurement ...... 35 2.2.6 Particle stability study ...... 36 2.2.7 Data analysis ...... 36 2.2.7.1 Statistical analysis ...... 36 2.2.7.2 Calculation of activation energy ...... 37 2.2.7.3 Dependence of activation energy on pH, Al to P ratio, and shape of the hydrogel particles ...... 38 2.3 Results ...... 39 2.3.1 Fabrication of particles ...... 39 2.3.2 Mechanical properties of hydrogel particles ...... 41 2.3.3 Stability of hydrogel particles as a function of pH and temperature ...... 43 2.3.3.1 Stability of hydrogel particles at low pH ...... 44 2.3.3.2 Stability of hydrogel particles above pH 5.0 ...... 45 2.3.3.3 Modeling dissolution of hydrogel particles at high pH as a function of time and temperature...... 46 2.3.4 Modeling of the effect of alginate-pectin ratio, pH level and hydrogel particle shape on activation energy ...... 47 2.4 Conclusions ...... 50 Reference ...... 51 Chapter 3. Extraction of Anthocyanins encapsulated in alginate-pectin hydrogel particles in pH 5.0 buffer solution ...... 68 Abstract ...... 68 3.1 Introduction ...... 68 3.2 Materials and Methods ...... 70 3.2.1 Materials ...... 70 3.2.2 Preparation of ACN stock solutions ...... 70 3.2.3 Preparation of alginate-pectin hydrogel particles ...... 71 3.2.4 Extraction of ACN from hydrogel particles ...... 72 3.2.5 Data analysis ...... 72 3.3 Results and discussion ...... 73 3.4 Conclusions ...... 76 Reference ...... 77 Chapter 4: Encapsulation of purple corn and blueberry extracts in alginate-pectin hydrogel particles ...... 85 Abstract ...... 85 4.1 Introduction ...... 86 4.2 Materials and methods ...... 89 4.2.1 Materials ...... 89

4.2.2 Preparation of ACN stock solutions ...... 90 4.2.3 Preparation of gel solution with anthocyanins ...... 90 4.2.4 Hydrogel particle production ...... 91 4.2.5 Encapsulation efficiency of ACN in hydrogel ...... 92 4.2.6 ACN retention in hydrogel particles during subsequent storage ...... 92 4.2.7 ACN loss during gelation and particle curing ...... 93 4.2.8 Improvement of encapsulation efficiency ...... 93 4.2.9 Effect of light on free and encapsulated ACN during storage ...... 94 4.2.10 Porosity measurement ...... 94 4.3 Result and discussion ...... 95 4.3.1 Evaluation of factors affecting encapsulation efficiency of ACN ...... 95 4.3.1.1 Effect of initial ACN concentration ...... 97 4.3.1.2 Effect of particle shape ...... 98 4.3.1.3 Total gum concentration ...... 99 4.3.1.4 Alginate to pectin ratio ...... 100 4.3.1.5 Effect of ACN botanical source ...... 100 4.3.2 Improvement of encapsulation efficiency of ACN ...... 101 4.3.3 Effect of storage conditions on ACN release ...... 102 4.3.3.1 Particle weight to solution volume ratio ...... 102 4.3.3.2 Storage temperature ...... 103 4.3.3.3 Effect of light ...... 104 4.3 Conclusion ...... 106 Reference 108 Chapter 5. Determination of purple corn and blueberry ACN diffusion coefficients in alginate-pectin hydrogel particles ...... 124 Abstract ...... 124 5.1 Introduction ...... 125 5.2 Theory ...... 129 5.2.1 Mass Transfer inside Particles ...... 129 5.2.2 External mass transfer resistance ...... 131 5.3 Materials and methods ...... 132 5.3.1 Materials 132 5.3.2 Preparation of ACN stock solution ...... 132 5.3.3 Preparation of hydrogel particles ...... 133 5.2.2.4 Sample analysis ...... 133 5.3.5 Diffusion study ...... 134 5.3.6 ACN release from particles above pH 5.0 ...... 135

5.2.3 Data analysis ...... 136 5.2.3.1 Statistical analysis ...... 136 5.2.3.2 Calculation of activation energy ...... 136 5.4 Results and discussion ...... 136 5.4.1 Modeling parameters at pH 3.0 ...... 136 5.4.2 Effect of initial ACN concentration using PC ACN spherical particle ...... 137 5.3.3 Effect of particle structure...... 138 5.3.3.1 Alginate to pectin ratio ...... 138 5.3.3.2 Effect of total gum concentration ...... 138 5.3.2.3 Effect of particle shape ...... 139 5.4.4 Effect of ACN source ...... 140 5.4.5 Effect of temperature on ACNs diffusion ...... 140 5.4.6 Release of ACN at pH above 5.0 ...... 142 5.4.6.1 Effect of temperature ...... 144 5.4.6.2 Effect of gel composition ...... 144 5.5 Conclusion ...... 145 Reference ...... 147 Bibliography ...... 164

List of Tables

Table 1. Effect of hydrogel fabrication parameters on particle size and shape. Curing solution pH is 1.2 buffer...... 54 Table 2. Mechanical properties of the alginate-pectin gels prepared in pH 1.2 curing bath...... 55 Table 3. Volume dissolution rate constants under various pH and temperature conditions...... 56 Table 4. Comparison of the dimensional dissolution rate constants for the spherical and disc shaped hydrogel particles...... 57 Table 5. Activation energy and frequency factor determined by nonlinear fit for various pH, alginate-pectin ratio and particle shapes...... 58 Table 6. Extraction of PC and BB ACN with different SV to PW ratios in 3ml pH 5.0 buffer solution...... 79 Table 7. Effect of curing bath augmentation with ACN on encapsulation efficiency. TGC: 2.2%. Al-P: 82-18%. Spherical particles...... 113 Table 8. Particle weight to solution volume ratio on ACN loss during storage ...... 114 Table 9. Percent ACN weight change in solution, particle and particle in solution (PIS) samples under light and dark as a function of time...... 115 Table 10. Values of the effective diffusion coefficient De and partition coefficient Kp of PC ACN of 2.2% TGC 82-18 Al-P particles for different concentrations in particles at 24⁰C...... 151 Table 11. Effective diffusion coefficient De and partition coefficient of PC and BB ACN of 2.2% TGC 82-18 and 43-57 Al-P particles at 24⁰C...... 152 Table 12. Effective diffusion coefficient De for PC and BB ACN in 2.2% TGC 82-18 Al-P spherical particles and PC ACN in 2.2% TGC 43-57 Al-P spherical particles as a function of temperature...... 153 Table 13. Kinetic rate constant (k) of release and dissolution study using 2.2% TGC 82- 18 Al-P spherical particles under 4, 24 and 37 ºC at pH 5.0 and 7.0...... 154 Table 14. Rate constants of release of PC ACN at 37 ºC using 2.2% TGC 43-57 Al-P, 2.8% TGC 43-57 Al-P and 2.2% TGC 82-18 Al-P spherical particles at pH 5.0 and 7.0...... 155

xii

List of Figures

Figure 1. Anthocyanin aglycone basic structure...... 25 Figure 2. Anthocyanin chemical structure transformation under different pH conditions in aqueous phase (adapted from: Houbiers et al., 1998), where R1 and R2 are H, OH or OCH3 groups...... 26 Figure 3. Alginate structure at low pH range...... 27 Figure 4. High methoxyl pectin chemical structure at low pH condition...... 28 Figure 5. Shapes of Al-P hydrogel particles a. 2.8% TGC 43:57 wt% Al-P spherical particle, b. 2.8% TGC 43:57 wt% Al-P disc particle top view, c. 2.8% TGC 43:57 wt% Al-P (F2) disc particle side view...... 59 Figure 6. Alginate and pectin structure at low and high pH...... 60 Figure 7. Total work as a function of successive compression- decompression cycles for 2.2% TGC 82:18 wt% Al-P hydrogels at 10% ( ○ ) 15% (□ ) and 20% (◊) deformation levels and for 2.8% TGC 43:57 wt% Al-P hydrogels at 10% (●) 15% (■) and 20% (◆) deformation levels...... 61 Figure 8. Percent recoverable work as a function of successive compression- decompression cycles for 2.2% TGC 82:18 wt% Al-P gels at 10% (○) 15% (□) and 20% (◊) deformation levels and for 2.8% TGC 43:57 wt% Al-P gels at 10% (●) 15% (■) and 20% (◆) deformation levels...... 62 Figure 9. Volume change of hydrogel particles as a function of time for disc shape hydrogel particles at pH 3.0 (●), 5.0 (■) and 7.0 (◆) at 37 °C for a. 2.2% TGC 82:18 wt% Al to P ratio, b. 2.8% TGC 43:57 wt% Al to P ratio. Solid lines represent the fitted data to zero-order kinetic model...... 63 Figure 10. Volume change of hydrogel particles as a function of time over a three week storage period at 4 °C. 2.8% TGC 43:57 wt% Al-P, disc shape (●), 2.8% TGC 43:57 wt% Al-P, spherical shape (■), 2.2% TGC 82:18 wt% Al-P, disc shape (◆) and 2.2% TGC 82:18 wt% A l-P, spherical (▲)...... 64 Figure 11. Volume change of hydrogel particles as a function of time at pH 7.0 and at 4 °C (●), 24 °C (■), and 37 °C (◆) for a. 2.8% TGC 43:57 wt% Al-P, disc shape hydrogel particles b. 2.8% TGC 43:57 wt% Al-P, spherical hydrogel particles. Solid lines represent the fitted data to zero-order kinetic model...... 65 Figure 12. Dissolution behavior of a hydrogel, a. Proposed explanation for release behavior of uniformly distributed hypothetical bioactive compound (red dot) during dissolution of the hydrogel particle at high pH, b. The experimental observation of dissolution behavior of 2.2% TGC 82:18 wt% Al-P spherical hydrogel particles at pH 5.0 and at 4°C as a function of time. The red color was used for demonstration of hydrogel particle size change...... 66 Figure 13. Calculated volume change using Eq. (2.9) versus experimental volume change for 43:57 Al-P spherical hydrogel particles in pH 7.0 (○), 43:57 Al-P spherical hydrogel particles in pH 5.0 (□), 43:57 Al-P disc hydrogel particles in pH 7.0 (◊), 43:57 Al-P disc hydrogel particles in pH 5.0 (∆), 82:18 Al-P spherical hydrogel particles in pH 7.0 (●), 82:18 Al-P spherical hydrogel particles in pH 5.0 (■), 82:18 Al-P disc hydrogel particles in pH 7.0 (◆), 82:18 Al-P disc hydrogel particles in pH 5.0 (▲). 67 Figure 14. Extraction of ACN encapsulated by alginate-pectin hydrogel particles. a) Particles containing ACN was placed in 3 ml pH 5.0 buffer; b) ACN released to pH 5.0 buffer after particles dissolution...... 80 Figure 15. Visible absorbance spectrum of dissolved blank 2.2% Al-P particles at pH

xiii

5.0 (green) dissolved ACN containing (5.36 μg/ml) 2.2% Al-P particles at pH 5 (red), and 0.1wt% alginate-pectin solution (blue) (overlapped with green line)...... 81 Figure 16. Visible absorbance spectrum of ACN solution at a concentration of 4.56 μg/ml: PC ACN solution at pH 5.0 (red), PC ACN solution at pH 1.2 (blue), BB ACN solution at pH 5.0 (green) and BB ACN solution at pH 1.2 (black)...... 82 Figure 17. Visible absorbance spectrum of PC ACN solution in pH 5.0 at ACN concentrations of 0.89 μg/ml (blue), 1.14 μg/ml (red), 1.96 μg/ml (black) and 3.69 μg/ml (green)...... 83 Figure 18. The mass of ACN measured by direct method versus solvent volume to particle weight ratio: PC ACN (●) and BB ACN (■)...... 84 Figure 19. Percent encapsulation efficiency after curing (EEc) (●) and percent ACN retention during storage (ARs) (■) for PC ACN encapsulation in 2.2% TGC 82-18% Al-P spherical particles as a function of initial total monomeric ACN concentration in solution...... 116 Figure 20. Mass of ACN in particle versus mass of ACN in solution at equilibrium for purple corn after curing (●) and after storage (■). 2.2% TGC, 82-18% Al-P, spherical particles, particle weight to solution volume ratio is 0.136...... 117 Figure 21. Encapsulation efficiency during curing and storage of PC and BB ACN using 2.2% TGC 82-18 Al-P spherical particles (red), 2.2% TGC 43-57 Al-P spherical particles (blue) and 2.8% TGC 43-57 Al-P spherical particles (black) with initial PC and BB ACN concentration at 26.9 and 27.7 µg/ml, respectively...... 118 Figure 22. Comparison of improvement of EEc during curing for PC and BB encapsulation. Percent EEc after curing as a function of normalized ACN concentration in curing bath by ACN concentration in droplet for purple corn (●) and for blueberry (■). Initial PC and BB ACN concentrations are 233.9 and 27.7 µg/ml respectively...... 119 Figure 23. Overall encapsulation efficiency (EEo) of PC ACN at equilibrium in storage at different particle weight to solution volume ratio. 2.2% TGC 82-18% Al-P spherical particles. Initial PC ACN concentration is 713.5 µg/ml...... 120 Figure 24. Overall encapsulation efficiency (EEo) as a function of temperature for PC ACN in 43-57% Al-P (●) and 82-18% Al-P (■) spherical particles. 2.2% TGC. Particles stored at pH 3.0. Initial PC ACN concentration is 713.5 µg/ml...... 121 Figure 25. Stability of PC ACN as a function of exposure time to fluorescent light. a) Stability for samples exposed to fluorescent light at 20 ⁰C; b) Stability for samples stored in dark at 20 ⁰C. ACN aqueous solution (●), ACN encapsulated in hydrogel particle (♦), ACN encapsulated in hydrogel particle dispersed in solution (■)...... 122 Figure 26. Gelation study of 2.2% TGC 82-18 Al-P encapsulated BB ACN with disc (■) and spherical (●) shape and encapsulated PC ACN with disc (□) and spherical (○) shape in pH 1.2 buffer solution over time...... 123 Figure 27. Geometry of spherical and cylindrical geometry of the particles built in COMSOL using 2D-axisymmetry dimension...... 156 Figure 28. PC ACN uptake by solution as a function of time in 2.2% TGC 82-18 Al-P spherical particles at 24⁰C using initial ACN concentration in particle at 60.2 (●), 85.9 (), 112.5 () and 172.0 μg/ml (▲). Symbols represent normalized ACN concentration obtained from experiments while the curves represent predicted bulk ACN concentrations from the best fit using COMSOL optimization module with global least square method for the corresponding experimental data...... 157 Figure 29. Diffusion of PC ACN from 2.2% TGC 82-18 Al-P spherical particles into pH 3.0 buffer solution at 4 (○), 24 (◇) and 37⁰C (□). Symbols represent normalized

xiv

ACN concentration obtained from experiments while the curves represent predicted bulk ACN concentrations from the best fit using COMSOL optimization module with global least square method for the corresponding experimental data...... 158 Figure 30. Diffusivity as a function of temperature of PC ACN using 2.2% TGC 82-18 Al-P spherical particles (●), BB ACN using 2.2% TGC 82-18 Al-P spherical particles () and PC ACN using 2.2% TGC 43-57 Al-P spherical particles ()...... 159 Figure 31. PC ACN release in pH 5.0 media at 4, 24 and 37⁰C. Absorbance ratio of PC ACN at 4⁰C (▲), 24⁰C (♦) and 37⁰C (∎); and dissolution of area ratio of particles at 4⁰C (∆), 24⁰C (◊) and 37⁰C (□) (obtained from previous work (Guo & Kaletunc, 2016))...... 160 Figure 32. PC ACN release in pH 7.0 media at 4, 24 and 37⁰C. Absorbance ratio of PC ACN at 4⁰C (▲), 24⁰C (♦) and 37⁰C (∎); and dissolution of area ratio of particles at 4⁰C (∆), 24⁰C (◊) and 37⁰C (□) (obtained from previous work (Guo & Kaletunc, 2016))...... 161 Figure 33. The PC ACN release rate constants with 2.2% TGC 82-18 Al-P spherical particles as a function of temperature at pH 5.0 (■) and 7.0 (●)...... 162 Figure 34. PC ACN release in pH 5.0 and 7.0 media at 37⁰C. a) Absorbance ratio of PC ACN at pH 5.0 using 2.2% TGC 43-57 Al-P spherical particle (●), 2.8% TGC 43-57 Al-P spherical particle (∎) and 2.2% TGC 82-18 Al-P spherical particle (▲); and b) Absorbance ratio of PC ACN at pH 7.0 using 2.2% TGC 43-57 Al-P spherical particle (●), 2.8% TGC 43-57 Al-P spherical particle (∎) and 2.2% TGC 82-18 Al-P spherical particle (▲)...... 163

Chapter 1: Literature Review

1.1 Anthocyanins (ACNs)

1.1.1 Anthocyanins as natural food colorants

Anthocyanins (ACNs) are a group of water-soluble phenolic compounds responsible for the coloration of many fruits, vegetables and flowers, contributing to the attractive red, orange, purple and blue colors. They are extracted from plant materials including leaves, fruits, roots and flower. The natural colors provided from ACNs change from red or purple to colorless as pH increases from 1.0 to 4.0-5.0, and the color changes again to blue while pH increases to 9.0 (Cevallos-Casals & Cisneros-Zevallos, 2004). ACNs have been consumed for centuries without any reported adverse effects and have pH-dependent color as described earlier (Cevallos-Casals & Cisneros-Zevallos, 2004). There is now a growing interest in using anthocyanins as natural food colorant as a promising alternative for artificial/synthetic dyes (Giusti & Wrolstad, 2003). In 2011, the global sales of natural colorants grew to $600 million with 7% annual growth rate, while the sales of artificial and synthetic dye market growth were less than 4% from 2007 to 2011 (Lao & Giusti, 2016).

Several reasons account for this growing trend of applying ACNs as food colorants.

Research has shown possible relationships between the consumption of synthetic dyes and adverse behavioral/neurological effects in some children (McCann et al., 2007; Stevens et al., 2013; Weiss et al., 1980). In addition to the variety of the color ACN provides, research has proposed that ACNs have health benefits properties, including anti-inflammatory and anti-carcinogenic activity, preventing cardiovascular disease, and facilitating weight

management (He & Giusti, 2010). Other health benefits of ACNs have also been reported including antimutagenic (Ga̧siorowski et al., 1997; Peterson et al, 1998), antioxidant (Noda et al., 2000; Prior, 2003), anti cell mutation (Aoki et al., 2004) and prevention of obesity and diabetes (Tsuda et al, 2003).

1.1.2 Structure of Anthocyanins

Anthocyanins are secondary metabolites from higher plants. They belong to flavonoids, as their aglycones share a typical C 6 (A ring)-C 3 (C ring)-C 6 (B ring) carbon skeleton in the phenolics family (Fig. 1) (Harborne 1998). This most basic form was referred as anthocyanidin. Due to the different hydroxylation and methylation on different locations of the B-ring, it was reported that almost 25 different aglycones identified in nature (Andersen & Jordheim, 2006). Among the 25 aglycones, 95% of them are consist of six anthocyanidins (aglycones): Pelargonidin (Pg), Malvidin (Mv), Peonidin (Pn),

Petunidin (Pt), Cyanidin (Cy), Delphinidin (Dp) (Clifford, 2000). The different colors are a presentation of the structure difference: more hydroxyle groups on the B-ring lead the color falling on the end of blue spectrum, while more methoxyl groups lead the color falling on the red end (Delgado-Vargas & Paredes-Lopez 2003; Heredia et al., 1998). In fruits and vegetables, it was reported that the highest amount was Cy (50%), followed by Pg, Dp and

Pn (12%), Pt (7%) and lowest Mv (7%) (Castañeda-Ovando et al., 2009; Kong et al., 2003).

The non-methylated anthocyanidins glycoside was mostly reported, which was found in the colored leaves (80%), in fruits (69%), and in flowers (50%), while the most abundant one is Cy-3-glucoside among them (Kong et al., 2003).

Besides of the substitution on the B-ring of aglycones, various anthocyanins are consist of numerous glycosylations and acylations combinations (Wrolstad, 2004). It has

been suggested that the combinations of varying the subsititions on the B-ring, glycosylations and acylations affected the stability (Giusti, Rodríguez-Saona, & Wrolstad

1999), which could be due to the intramolecular hydrogen bonds between the sugar moiety, acylating group and the anthocyanidin from the complex anthocyanins (Giusti, Ghanadan,

& Wrolstad 1998; Borkowski et al., 2005). This intramolecular hydrogen bonds can be referred as folding, copigmentation or stacking, which is preferable for anthocyanin stability. The proposed explanaition for improved stability by intramolecular bonds was due to either the radical formation or an electron loss was spread out through the entire network of phynolic groups and hydrogen bonds, which as a result, weakened the degradation of the anthocyanin; or the intramolecular bonds formed spatial reorientation, and new confirguration was created which was potentially more stable (Giusti, Ghanadan,

& Wrolstad 1998; Borkowski et al., 2005).

1.1.3 The pH effect on Anthocyanin chemical structure

Anthocyanins have been reported to be a group of flavonoids which structure transformation under different pHs aqueous solution is reversible (Fig. 2). According to

Brouilard & Delaporte (1977), there are four major anthocyanin structures existing in the aqueous solution at equilibrium, including two colorless forms (carbinol pseudobase and chalcone), one blue form (quinonoidal base) and one red form (flavylium cation).

Flavylium cation is the predominant anthocyanin form when pH below 2.0, while the colorless carbinol pseudobase is generated when pH varies from 3.0 to 6.0. This is occurred by rapid hydration of the falvylium cation by nucleophilic attack of water as pH increased from below 2.0 to a higher pH ranged from 3.0 to 6.0. The hydration process was reported to be depending on the pH change, and the time required to reach equilibrium was ranged

from 30 to 103 s (Brouillard & Dubois, 1977). The reversion of this process (from carbinol pseudobase to flavylium cation) was reported to be almost instant when anthocyanin was acidified (Brouillard & Delaporte, 1977).

Meanwhile, at pH 3.0 to 6.0, the colorless carbinol pseudobase can transform to another open ring form (Fig. 2), which is the colorless chalcon pseudobase. The transformation process from carbinol pseudobase to chalcon was reported to be slow, while the transformation from chalcon to flavylium cation was even slower (Francis, 1989).

As pH increases further to low acidic or a neutral condition, deprotonation of the flavylium cation occurs and the purple to violet quinonoidal base forms (Giusti & Wrolstad,

2001). This reaction was extremely fast and was favored when competing with hydration reactions (Brouillard & Dubois, 1997). When pH is above 8.0, the blue anionic quinonoidal base develops and can carry one or two negative charges (Asenstorfer et al., 2003; Chen &

Hrazdina, 1982; Jing, 2006; Mateus & de Freitas, 2009).

1.1.4 Anthocyanins stability

ACNs in fruits or vegetable extracts are proved to be affected by pH (Cevallos-Casals

& Cisneros-Zevallos, 2004; Kirca et al., 2007; Reyes & Cisneros-Zevallos, 2007), the presence of light (Carlsen & Staelfeldt, 1997; Cevallos-Casals & Cisneros-Zevallos, 2004), ascorbic acid (Choi et al., 2002; Skrede et al, 1992), enzyme (Rossi et al., 2003; Skrede et al., 2000), and temperature (Patras et al., 2010). The sensitivity of ACN to environmental conditions was evaluated by measuring the % color retention, monomeric ACN concentration, Hunter color parameters (Cemeroğlu et al., 1994; Cevallos-Casals &

Cisneros-Zevallos, 2004; Wang & Xu, 2007; Yang et al., 2008). The effect of pH on the color and monomeric ACN concentration of various fruit and vegetable extracts was

evaluated (Cevallos-Casals & Cisneros-Zevallos, 2004; Kirca et al., 2007; Reyes &

Cisneros-Zevallos, 2007): the half-lives of black carrot extract was reduced from 25.1 hours at pH 2.5 to 12.6 hours pH 6.0 at 70⁰ C (Kirca et al., 2007). Reyes and Cisneros-

Zevallos (2007) found the rate of color loss of red-flesh potato, purple-flesh potato and grape at pH 3.0 was two times faster than those stored in pH 1.0 at 25⁰ C. A wider range of pH from 4.0 to 11.7 was used to store ACN at 20⁰ C, and the percent color retention of red sweet potato extract was reported to be 99% at pH 2.0 and only 55% at pH 4.0 after

138 days storage protected from light and oxygen, while the % color retention of purple corn and red grape extract was not detectable after storage (Cevallos-Casals & Cisneros-

Zevallos, 2004). The light treatment on elderberry extract using UV-visible at 313, 366 and

436 nm showed under pH 3.0-3.8, the degree of color bleaching was ten times higher in

UV light than visible light by cleavage of covalent bonds (Carlsen & Staelfeldt, 1997). Co- existing compound such as ascorbic acid was reported to accelerate the rate of monomeric

ACN concentration degradation in strawberry syrup, shortening the half-life from 27 days to 18 days when stored at 20ºC (Skrede et al, 1992). Choi and co-workers (2002) found the presence of ascorbic acid showed slightly faster degradation rate of ACN in blood orange juice, with 0.3% difference in monomeric ACN concentration decrease per week at refrigeration temperature. Polyphenol oxidase was reported to be the main cause of monomeric ACN degradation during juice processing—enzyme contained non-blanched blueberry ACN degraded 45% more than the blanched ones when incubating at 40⁰ C for

3 hours (Skrede et al., 2000). Similar result was also reported that the presence of enzyme in blueberry fruits resulted in a 20% lower % retention of ACN when processing blueberry fruits to juice (Rossi et al., 2003). The food processing treatments were also reported to

cause ACN irreversible structure change due to high temperature. The degradation of monomeric ACN during food processing was reported as 43% loss in blueberry puree during blanching, 80% loss of raspberry purees during jam manufacuring, and 59%, 41% and 29% of red cabbage products during blanching, boiling and steaming respectively

(Patras et al., 2010). Therefore, the instability of ACNs to their surrounding environments limited the application in food industry.

1.1.5 Anthocyanin sources

Anthocyanins exist in a variety of fruits or vegetables those are of blue-purple-red color. As defined in USA regulations, ACNs extracted from the edible portion of fruit

(21CFR73.250) or vegetable (21CFR73.260) can be used. ACN sources have been reported to be from highly pigmented fruits and vegetables. The exact profile of the ACNs varies significantly from one to another. Fruits usually contain simple ACN with mon- or diglycosylations, while vegetables are consist of ACNs with multiple glycosylations and acylations on the anthocyanidin. The composition of the anthocyanidin types also varies greatly from species to species. Therefore, it is important to acknowledge the composition of anthocyanins from the plant extracts.

1.1.5.1 Purple corn anthocyanin

Purple corn (Zea mays L.), also called purple maize, is originally from Peru and widely cultivated and consumed. Due to its deep purple color, its extracts have been widely used for coloring two of the most popular food: chichi morada and mazamorra morada (FAO,

2013). ACN as one of the major color source in purple corn is reported to be high in content ranged from 6.8 to 82.3 mg/g FW in purple corn (Cevallos-Casals & Cisneros-Zevallos,

2003; Li et al., 2008; Wu et al., 2006). The profile of ACN in purple corn has been reported

to contain six major anthocyanins with 17 minor anthocyanins (Barry, 2013; De Pascual-

Teresa et al., 2002; Jing & Giusti, 2005; Jing et al., 2007; Li et al., 2008). The six major

ACNs include cyanidin, pelargonidin, peonidin glucosylated at position 3 and the malonic acid substitution on the 6’’ position of the glucose molecule.

1.1.5.2 Blueberry anthocyanin

The anthocyanin composition in blueberry varies by the cultivars. The majority of them are the glycosidic anthocyanidins including cyanidin, malvidin, petunidin, delphinidin and peonidin with monoarabinosides, monoglucosides and monogalactosides (Norberto et al.,

2013). A small portion of acylated anthocyanins were also found in blueberry (Yousef et al., 2013). Work from Yousef and co-workers (2013) reported that 70% of the total anthocyanins content of the commercial blueberry cultivars were malvidin-3-O-galactoside, malvidin-3-O-arabinoside, delphinidin-3-O-galactoside, delphinidin-3-O-arabinoside and cyanidin-3-O-arabinoside. Similarly, Barry (2013) reported these five anthocyanin composition contributed 62.6% of the total anthocyanin content in blueberry concentrated.

1.1 Anthocyanin encapsulation systems

The instability of the natural colorants greatly limited their application in food matrix.

Variety of research in the past decades has been done to investigate approaches to increase the stability of the natural colorants. Besides of studying the stability of diffrerent sources of anthocyanins, encapsulation system was intensively studied due to its advantage of protection of natural colorants from external environment. Spray-drying, ionic gelation and some other methods are discussed in this section.

1.1.1 Spray-drying

Spray-dyring is a technique utilizing the heat transfer between the feed fluid and hot air, resulting in a dry powder product. It has been widely applied in food industry due to its advantages including low operating cost, high quality of capsules, good yield, small sized powder and continuous operation (Mahdavi et al., 2014). The particle size is usualy ranged from 10 to 400 µm (Mahadavi et al., 2014). Since ACNs are water soluble compounds, they are suitable for spray-drying approach.

Typically, natural gums (alginate, gum Arabic, and carrageenans), proteins (whey protein, dariry protein, and gelatin), carbohydrates (maltodextrins and cellulose derivatives) and lipids (wax and emulsifiers) (Zuidam & Shimoni, 2010) can be used as the encapsulating materials. Maltodextrin (MD) is mostly frequently used in encapsulating anthocyanins (Shahidi & Han, 1993). The encapsulated ACNs by MD were reported to include black carrot extract (Ersus & Yurdagel, 2007), Garcinia indica extract (Nayak &

Rastogi, 2010), berberis extract (Huang & Yang, 2011), Andes berry and Tamarillo extract

(Olaya et al., 2009), Roselle-pinapple juice (Osman & Endut, 2009) and purple weet potato flour (Ahmed et al., 2010). Tonon and co-workers (2010) produced spray-dried particles with encapsulating acai juice by MD, gum Arabic and tapioca starch, and results showed the best protection of ACN was using MD as the encapsulation material. A mixture of MD with other materials was also used to encapsulate ACNs, such as mixture of MD with soybean protein isolate and ß-cyclodextrin-MD-arabic gum (Ying-chang et al., 2010), modified starch-MD (Nogueira et al., 2011), MD-gum Arabic and MD-gelatin (Mahdavi et al., 2016). The reported encapsulation efficiency (EE) of ACN in MD among those research was ranged from 89% to 95% (Huang & Yang, 2011; Ying-chang et al., 2010;

Robert et al., 2010). The stability of ACNs extracted from Berberis kaschgarica showed an increase towards temperature, light, oxidants and metal ions after encapsulated in MD

(Huang & Yang, 2011); pomegranate juice extract showed significant improvement on stability with lower degradation of ACN when encapsulated with MD than in soybean protein isolates (Robert et al, 2010); and black carrot extract capsules showed 3 times extended shelf-life compared with no encapsulation (Ersus & Yurdagel, 2007).

Though the described advantage of using spray-drying, there are some disadvantages of this technique. First, there is a high requirement on the solubility of the shell material in water which would limit the selection of encapsulant (Desai & Park, 2005). Second, due to the high temperature processing for spray-drying (> 100⁰ C), this method cannot be used for heat-sensitive material. The high temperature processing also reported to cause micro- cracks on the particle surfaces (Jones et al., 2013), which may accelerate the loss of encapsulated materials from the capsules when submerged in liquid food product.

1.1.2 Ionic gelation

Anthocyanins were also reported to be encapsulated in gels which were formed by ionic gelation. The gel formation occurs due to ionic interaction between the wall material and the ionic solution, usually by extrusion method. On a small scale, this technique often involved with using a syringe to extrude gel solution through a needle and dropping the gel droplets into an ionic solution. The size of the particles which is depending on the diameter of the needle, the viscosity of the gel solution and the flow rate of the feed is typically 0.5 to 6 mm (Blandino et al., 1999; Ouwerx et la., 1998). Combined with other technique such as electrostatic or ultrasonic technique, the size of the particle can be further reduced. When using electrostatic technique, due to the charge on the ionic solution and the droplet, smaller

droplets were produced, therefore 40 to 2500 µm particles can be achieved (Burey et al,

2008). Ultrasonic was applied to breakup the gel solution stream by vibration, which generated particle size ranging from hundreds of microns (Cellesi et al., 2004) up to 5 mm

(Hunik & Tramper, 1993). The ionic gelation method has been widely used in pharmaceutical and food industry due to their advantages of using organic solvents. It also allows a wider range of viscosity of the gel solution to use when comparing to spray-drying method. In addition, by avoiding using extreme pH or temperature condition, this technique is especially beneficial to encapsulate bioactive compounds which are sensitive to temperature and pH.

It has been reported that alginate was used as wall material and formed Ca-alginate microparticles to encapsulate haskap berries extract (Celli et al., 2016). Results showed the highest EE was achieved at 63.12% and the EE was significantly affected by the concentration of alginate solution. Chokeberry extract was also reported to be encapsulated with Ca-alginate capsules and compared with Ca-alginate using inulin as filler (Ćujić et al.,

2016). Particle size generated using these two different wall materials was ranged from 800 to 1340 µm and decreases as reducing the needle size. They found the inulin filler only showed an effect on the release behavior, while showed no influence on EE or particle size.

Ca-alginate microbeads layered by chitosan encapsulating raspberry extract was investigated and high EE was achieved (80 to 89%) (Belščak-Cvitanović, 2011). Ca-pectin-

CMC-whey protein hydrogels encapsulated with Roselle extract showed an improvement on antioxidant activities (Serrano-Cruz et al., 2013).

1.1.3 Others

Besides of spray-drying and extrusion techniques, other approaches have also been assessed to encapsulated ACNs. Pullulan was mixed with hibiscus extracts which is rich in

ACNs, and the mixture was freeze-dried to obtain dry powders (Gradinaru et al., 2003). It showed that the stability of the encapsulated ACN was directly affected by the water activity of the storage condition. MD and gum Arabic were also reported to encapsulate

Roselle extract by freeze-drying (Selim et al., 2008). Their result showed the encapsulation significantly elongated the half-life of the ACN up to 433.1 days after storage.

Thermo-set gels were applied to encapsulate ACNs, such as whey protein isolate, glucan gel and curdlan gel (Betz & Kulozik, 2011; Ferreira et al., 2009; Xiong et al., 2006).

Bilberry ACN was reported to be encapsulated by acidic whey protein isolate (Betz &

Kulozik, 2011). The microparticles generated by emulsion/heat gelation method were smaller than 70 µm, and the encapsulation showed delay release of ACN when placed under simulated gastric condition. Curdlan was reported to form hard gel under low pH and used to encapsulate blackberry extract, which resulted in an EE ranged from 80.3 to

96.7% (Ferreira et al., 2009).

1.2 Hydrogel based encapsulation delivery system

A number of delivery systems based on hydrogel encapsulation of sensitive bioactive compounds are designed to improve the stability, to maximize the retention, and to control the release at the target locations in human body. Hydrogels used in food applications are formed from biopolymers such as alginate, pectin, chitosan, modified starch or carrageenan

(Burey et al., 2008). These hydrogel particles have been produced to encapsulate hydrophilic bioactive compounds including antioxidants in Ca-alginate and Ca-alginate-

chitosan (Deladino et al., 2008), lactose and ethanol in Ca-alginate (Gabardo et al., 2011), and glucose in carrageenan (Nguyen & Luong, 1986). Hydrogels were also used to entrap lipophilic materials such as casein in Ca-alginate (Zhang et al., 2015) and curcumin in Ca- pectin (Nguyen et al, 2014). In addition, encapsulation in biopolymer based hydrogels provides improved opportunities for the food industry because they are from natural resources and approved for food use by U.S. Food and Drug Administration for use in food materials (Burey et al., 2008; FDA, 2016). In this literature review, three types of hydrogels

(alginate-calcium, pectin and alginate-pectin) will be summarized and reviewed.

1.3.1 Calcium-alginate hydrogel encapsulation

The term alginate is not a scientific term which does not correspond to certain chemical structure. Actually, alginates refer to salts/derivatives of alginic acid extracted from brown seaweed (Phaeophyceae) (King, 1983). Even though alginate does not specify a certain structure, but it is composed of M and G residues, which are β-(1→4)-linked D- mannuronic acid and α-(1→4)-linked L-guluronic acid, respectively. The ratio between M and G residues of alginate will be dependent on the species of the seaweed. It was suggested that the acidic condition caused rapidly discharging of the crosslinking calcium ions in Ca- alginate hydrogel particles, and therefore the Ca-alginate gel particles converted to alginic acid (Østberg et al., 1994). As shown in Figure 3, it is the fundamental structure of alginate.

Alginate is one of the most commonly used hydrogel polymers to encapsulate bioactive compounds, due to its straight-foward and most investigated gelation mechanisms.

Alginate can form hydrogel through ionotropic mechanisms with the presence of calcium, the divalent cations (Biswal and Singh, 2004). To be more specific, the gelation happens when there is interaction between the G residue and calcium (Douglas and Tabrizian, 2005;

Glicksman, 1983). Several factors can affect the gelation speed, including temperature, ion concentration and alginate concentration (Draget, 2000).

ACN extracts of various sources were reported to be encapsulated by hydrogels.

Examples of using calcium alginate to encapsulate bioactive compounds (e.g. ACNs) are listed. Ca-alginate or Ca-alginate-inulin hydrogel particles encapsulated with chokeberry extracts (Ćujić et al., 2016) and carqueja extract (Balanč et al., 2016) were prepared with electrostatic extrusion method. Without the electrostatic addition, extrusion method was used to generate Ca-alginate hydrogel particles to encapsulate Yerba mate (Córdoba et al.,

2013; Deladino et al, 2008) and haskap berries (Celli et al., 2016). Chokeberry extract was encapsulated in Ca-alginate and Ca-alginate-inulin hydrogel particles prepared with 1.5% w/v low and medium viscosity sodium alginate solution with EE at 37% (Ćujić et al., 2016).

Ca-alginate particles prepared with 1.6% and 1.8% w/v low viscosity sodium alginate solution were also used to encapsulate wine waste extract containing ACN, obtained 30% and 20% EE using different sizes needles for extrusion (Aizpurua-Olaizola et al., 2016).

Celli and co-authors (2016) reported the highest EE of haskap berries extracts could be achieved at 63.12% with 10% w/w low viscosity sodium alginate solution. Their results also showed that with 1% w/w reduction of alginate concentration, the EE was decreased dramatically to 17.97%.

1.3.2 Pectin hydrogel

Pectin is of natural abundance: it can be found in many plant cells, especially plant cell walls (Glicksman, 1983). The most commonly used pectin, especially for the commercially used pectin is extracted from apple pomace or citrus peel by using mildly acidic solvents. Also, sometimes, sugar beet and sunflower are also used as commercially

pectin source (Hoefler, 2004).

The chemical structure of pectin is mainly composed of galacturonic acid units, with on average, one free acid group in every 3-4 methyl esters of the galacturonic acid units, and the average molecular weights can vary from 110,000-150,000 Dalton (Christensen,

1983; Hoefler, 2004). The methyl esters are corresponded to the degree of esterification

(DE), which in pectin can be up to 70%-80%. The chemical structure of pectin determines its gelling capacity. To be more specific, the main structural factors which determine the gelling capacity include pectin’s degree of esterification, and the methyl ester position on the pectin backbone (Hoefler, 2004). In Figure 4, common structure of pectin was shown, which represents all significant structures. Pectin, also based on their DE, can be classified as low ester (LE) pectin (DE less than 50%) and high ester (HE) pectin (DE equal or more than 50%). The gelation mechanisms are quite different between LE and HE pectins: LE pectin gelates with divalent cations while HE pectin forms gel at low pH with high soluble solids (e.g. jam) (Hoefler, 2004). The soluble solids in the system play a significant part in

HE pectin gelation, since it can alter the HE pectin structure (Nussinovitch, 1997). The mechanism for LE pectin gelation can be described as “egg-box” model (Rees, 1982) which shares similarity with the gelation structure as alginate gels. The similarity can be described as the following: the basis of the gelation system are chains of “calcium bridge” chelated by -OH and -CO groups (Rees, 1982). Ion concentration, gel preparation and hydrocolloid concentration are also the factors which affect LE pectin gel, the same as for alginate gel.

1.3.3 Alginate and pectin hydrogels

Sodium alginate and pectin are anionic polysaccharides which form biodegradable hydrogels (Wang et al., 2012). Alginate comprises both guluronic acid and mannuronic acid, and pectin only galacturonic acid. The abundant carboxylate and hydroxyl groups in both alginate and pectin result in a high hydrophilicity, which leads to polymer chain extension by the charge repulsion and to a high absorption of water (Wang et al., 2012).

Synergistic gelation of alginate-pectin mixtures occurs at pH below 3.4 due to the hydrophobic effect (Higuita-Castro et al., 2012; Toft et al., 1986). Alginate-pectin (Al-P) mixtures have been used to encapsulate vitamin C, anthocyanins (Higuita-Castro et al.,

2012) and folic acid (Madziva et al., 2005). Madziva and co-workers (2005) used an alginate and pectin gel mixture dropping into calcium chloride to fabricate hydrogel particles to encapsulate folic acid. Higuita-Castro and co-workers (2012) showed Al-P hydrogel particles were formed while pH dropped from 6.0 to 3.0 by adding glucono-δ- lactone to biopolymer solution. Though Al-P hydrogel potentially can provide more opportunities for applications of environmental stimuli responsive drug and nutrient delivery, the factors influencing their fabrication and degradation characteristics have not been elucidated.

The pH responsive characteristic distinguishes Al-P hydrogel from alginate or pectin only hydrogels since neither of them alone form a pH responsive gel. As the pH increases,

Al-P hydrogels are expected to go through a gel-sol transition and the kinetic of this transition can be a key factor in release of bioactive compound. Recently, the stability of alginate-pectin hydrogel particle at different pHs was reported (Guo & Kaletunç, 2016).

The research showed that the particles were formed at pH below 3.0 and dissolved above

pH 5.0 over a time period of 3 min to 20 min depending on the gel formulation and particle shape at 37⁰ C. As reported by Kong and Singh (2008) and Guerra and co-workers (2012), pH 5.0 and 7.0 media simulates a full stomach and small intestine and colon. Since the absorption location for ACN was reported to be in the intestines (He & Giusti, 2010), this unique pH responsive hydrogel—alginate-pectin particles can demonstrate the potential for controlled release of ACN in the intestines.

References

Ahmed, M., Akter, M. S., Lee, J. C., & Eun, J. B. (2010). Encapsulation by spray drying of bioactive components, physicochemical and morphological properties from purple sweet potato. LWT-Food Science and Technology, 43(9), 1307-1312.

Aizpurua-Olaizola, O., Navarro, P., Vallejo, A., Olivares, M., Etxebarria, N., & Usobiaga, A. (2016). Microencapsulation and storage stability of polyphenols from Vitis vinifera grape wastes. Food chemistry, 190, 614-621.

Andersen, Ø. M., Jordheim M. (2006). The anthocyanins. In Ø. M. Andersen, K. R. Markham (Eds.), Flavonoids: Chemistry, Biochemistry and Applications (pp. 471–552). Boca Raton, FL: CRC Press.

Aoki, H., Wada, K., Kuze, N., Ogawa, Y., & Koda, T. (2004). Inhibitory effect of anthocyanin colors on mutagenicity induced by 2-amino-1-methyl-6-phenylimidazo [4, 5-b] pyridine (PhIP). Foods and Food Ingredients Journal of Japan, 209, 240-246.

Asenstorfer, R. E., Iland, P. G., Tate, M. E., & Jones, G. P. (2003). Charge equilibria and pKa of malvidin-3-glucoside by electrophoresis. Analytical biochemistry, 318(2), 291-299.

Augst, A. D., Kong, H. J., & Mooney, D. J. (2006). Alginate hydrogels as biomaterials. Macromolecular bioscience, 6(8), 623-633.

Balanč, B., Kalušević, A., Drvenica, I., Coelho, M. T., Djordjević, V., Alves, V. D. Sousa, I., Moldão‐ Martins, M., Rakić, V., Nedović, V & Bugarski, B. (2016). Calcium– Alginate–Inulin Microbeads as Carriers for Aqueous Carqueja Extract. Journal of food science, 81(1), E65-E75.

Belščak-Cvitanović, A., Stojanović, R., Manojlović, V., Komes, D., Cindrić, I. J., Nedović, V., & Bugarski, B. (2011). Encapsulation of polyphenolic antioxidants from medicinal plant extracts in alginate–chitosan system enhanced with ascorbic acid by electrostatic extrusion. Food research international, 44(4), 1094-1101.

Betz, M., & Kulozik, U. (2011). Whey protein gels for the entrapment of bioactive anthocyanins from bilberry extract. International dairy journal, 21(9), 703-710.

Biswal, D. R., and Singh, R. P. (2004). The flocculation and rheological characteristics of hydrolyzed and unhydrolyzed grafted sodium alginate in aqueous solutions. J Appl Polym Sci., 94, 1480–1488.

Blandino, A., Macias, M., & Cantero, D. (1999). Formation of calcium alginate gel capsules: influence of sodium alginate and CaCl2 concentration on gelation kinetics. Journal of bioscience and bioengineering, 88(6), 686-689.

Brouillard, R., & Delaporte, B. (1977). Chemistry of anthocyanin pigments. 2. Kinetic and thermodynamic study of proton transfer, hydration, and tautomeric reactions of malvidin 3-glucoside. Journal of the American Chemical Society, 99(26), 8461-8468.

Brouillard, R., & Dubois, J. E. (1977). Mechanism of the structural transformations of anthocyanins in acidic media. Journal of the American Chemical Society, 99(5), 1359- 1364.

Burey, P., Bhandari, B. R., Howes, T., & Gidley, M. J. (2008). Hydrocolloid gel particles: formation, characterization, and application. Critical reviews in food science and nutrition, 48(5), 361-377.

Carlsen, C., & Stapelfeldt, H. (1997). Light sensitivity of elderberry extract. Quantum yields for photodegradation in aqueous solution. Food chemistry, 60(3), 383-387.

Castaneda-Ovando, A., de Lourdes Pacheco-Hernández, M., Páez-Hernández, M. E., Rodríguez, J. A., & Galán-Vidal, C. A. (2009). Chemical studies of anthocyanins: A review. Food chemistry, 113(4), 859-871.

Cellesi, F., Weber, W., Fussenegger, M., Hubbell, J. A., & Tirelli, N. (2004). Towards a fully synthetic substitute of alginate: Optimization of a thermal gelation/chemical crosslinking scheme (“tandem” gelation) for the production of beads and liquid-core capsules. Biotechnology and Bioengineering, 88(6), 740-749.

Celli, G. B., Ghanem, A., & Brooks, M. S. L. (2016). Optimized encapsulation of anthocyanin-rich extract from haskap berries (Lonicera caerulea L.) in calcium-alginate microparticles. Journal of Berry Research, 6(1), 1-11.

Cemeroğlu, B., Velioglu, S., & Isik, S. (1994). Degradation kinetics of anthocyanins in sour cherry juice and concentrate. Journal of Food Science, 59(6), 1216-1218.

Cevallos-Casals, B. A., & Cisneros-Zevallos, L. (2004). Stability of anthocyanin-based aqueous extracts of Andean purple corn and red-fleshed sweet potato compared to synthetic and natural colorants. Food Chemistry, 86, 69-77.

Chen, L. J., & Hrazdina, G. (1982). Structural transformation reactions of anthocyanins. Experientia, 38(9), 1030-1032.

Chen, X. D., & Özkan, N. (2007). Stickiness, functionality, and microstructure of food powders. Drying technology, 25(6), 959-969.

Choi, M. H., Kim, G. H., & Lee, H. S. (2002). Effects of ascorbic acid retention on juice color and pigment stability in blood orange (Citrus sinensis) juice during refrigerated storage. Food Research International, 35(8), 753-759.

Christensen, S. H. (1983). Pectins. In Glicksman, M. (Eds.), Food Hydrocolloids (pp. 205– 229). Boca Raton, FL: CRC Press Inc.

Clifford, M. N. (2000). Anthocyanins–nature, occurrence and dietary burden. Journal of the Science of Food and Agriculture, 80(7), 1063-1072.

Córdoba, A. L., Deladino, L., & Martino, M. (2013). Effect of starch filler on calcium-

alginate hydrogels loaded with yerba mate antioxidants. Carbohydrate polymers, 95(1), 315-323.

Deladino, L., Anbinder, P. S., Navarro, A. S., & Martino, M. N. (2008). Encapsulation of natural antioxidants extracted from Ilex paraguariensis.Carbohydrate Polymers, 71(1), 126-134.

Delgado-Vargas, F., Paredes-Lopez, O. (2003). Anthocyanins and betalains. In Natural Colorants for Food and Nutraceutical Uses, eds. F Delgado-Vargas, O Paredes-Lopez, pp. 167–219. Boca Raton, FL: CRC Press.

Desai, K. G. H., & Jin Park, H. (2005). Recent developments in microencapsulation of food ingredients. Drying technology, 23(7), 1361-1394.

Douglas, K. L., and Tabrizian, M. (2005). Effect of experimental parameters on the formation of alginate-chitosan nanoparticles and evaluation for their potential application as DNA carrier. J Biomat Sci-Polym E., 16, 43–56.

Draget, K. I. (2000). Alginates. In Phillips, G. O., and Williams, P. A. (Eds.) Handbook of Hydrocolloids (pp. 379–395). Boca Raton, FL: CRC Press.

Ersus, S., & Yurdagel, U. (2007). Microencapsulation of anthocyanin pigments of black carrot (Daucus carota L.) by spray drier. Journal of Food Engineering, 80(3), 805-812.

Ferreira, D. S., Faria, A. F., Grosso, C. R., & Mercadante, A. Z. (2009). Encapsulation of blackberry anthocyanins by thermal gelation of curdlan. Journal of the brazilian chemical society, 20(10), 1908-1915.

Francis, F. J., & Markakis, P. C. (1989). Food colorants: anthocyanins. Critical Reviews in Food Science & Nutrition, 28(4), 273-314.

Ga̧siorowski, K., Szyba, K., Brokos, B., & Oszmiański, J. (1997). Antimutagenic activity of anthocyanins isolated from Aronia melanocarpa fruits. Cancer Letters, 119(1), 37-46.

Giusti, M. M., & Wrolstad, R. E. (2001). Characterization and measurement of anthocyanins by UV-visible spectroscopy. In Wrolstad, R.E., Acree, T.E., An, H., Decker, E.A., Penner, M.H., Reid, D.S., Schwartz, S.J., Shoemaker, C.F., & Peter, S. (Eds.), Current protocols in food analytical chemistry (pp. F.1.2.1–F1.2.13). New York: John Wiley & Sons.

Giusti, M. M., & Wrolstad, R. E. (2003). Acylated anthocyanins from edible sources and their applications in food systems. Biochemical Engineering Journal, 14, 217-225.

Giusti, M. M., Ghanadan, H., & Wrolstad, R. E. (1998). Elucidation of the structure and

conformation of red radish (Raphanus sativus) anthocyanins using one-and two- dimensional nuclear magnetic resonance techniques. Journal of Agricultural and Food Chemistry, 46(12), 4858-4863.

Giusti, M. M., Rodríguez-Saona, L. E., & Wrolstad, R. E. (1999). Molar absorptivity and color characteristics of acylated and non-acylated pelargonidin-based anthocyanins. Journal of Agricultural and Food Chemistry, 47(11), 4631-4637.

Glicksman, M. (1983a). Food Hydrocolloids–Volume 1. Boca Raton, FL: CRC Press Inc.

Glicksman, M. (1983b). Food Hydrocolloids–Volume 3. Boca Raton, FL: CRC Press Inc.

Gradinaru, G., Biliaderis, C. G., Kallithraka, S., Kefalas, P., & Garcia-Viguera, C. (2003). Thermal stability of Hibiscus sabdariffa L. anthocyanins in solution and in solid state: effects of copigmentation and glass transition. Food Chemistry, 83(3), 423-436.

Guerra, A., Etienne-Mesmin, L., Livrelli, V., Denis, S., Blanquet-Diot, S., & Alric, M. (2012). Relevance and challenges in modeling human gastric and small intestinal digestion. Trends in biotechnology, 30(11), 591-600.

Harborne, J.B. (1998). Phenolic compounds. In JB Harborne (Eds.). Phytochemical Methods—A Guide to Modern Techniques of Plant Analysis (pp. 66–74). New York: Chapman & Hall.

He, J., & Giusti, M. M. (2010). Anthocyanins: natural colorants with health-promoting properties. Annual review of food science and technology, 1, 163-187.

Heredia, F. J., Francia-Aricha, E. M., Rivas-Gonzalo, J. C., Vicario, I. M., & Santos-Buelga, C. (1998). Chromatic characterization of anthocyanins from red grapes—I. pH effect. Food Chemistry, 63(4), 491-498.

Higuita-Castro, N., Gallego-Perez, D., Love, K., Sands, M. R., Kaletunç, G., & Hansford, D. J. (2012). Soft lithography-based fabrication of biopolymer microparticles for nutrient microencapsulation. Microencapsulation. Industrial Biotechnology, 8(6), 365-371.

Hoefler, A. C. (2004). Hydrocolloids. Minnesota: Eagen Press.

Huang, J. D., & Yang, L. (2011). Microencapsulation of anthocyanins from fruits of Berberis kaschgarica Rupr. Journal of Food Science, 16, 5-14.

Hunik, J. H., & Tramper, J. (1993). Large-scale production of κ-carrageenan droplets for gel-bead production: theoretical and practical limitations of size and production rate. Biotechnology progress, 9(2), 186-192.

Jing, P. (2006). Purple corn anthocyanins: chemical structure, chemoprotective sctivity and structure/function relationships (Doctoral dissertation, The Ohio State University). Retrieved from https://etd.ohiolink.edu

Jones, J. R., Prime, D., Leaper, M. C., Richardson, D. J., Rielly, C. D., & Stapley, A. G. (2013). Effect of processing variables and bulk composition on the surface composition of spray dried powders of a model food system. Journal of food engineering, 118(1), 19-30.

King, A. H. (1983). Brown Seaweed Extracts (Alginates). In Glicksman, M. (Eds.). Food Hydrocollids (pp. 115–182). Boca Raton, FL: CRC Press.

Kırca, A., Özkan, M., & Cemeroğlu, B. (2007). Effects of temperature, solid content and pH on the stability of black carrot anthocyanins. Food chemistry, 101(1), 212-218.

Kong, F., & Singh, R. P. (2008). Disintegration of solid foods in human stomach. Journal of food science, 73(5), R67-R80.

Kong, J. M., Chia, L. S., Goh, N. K., Chia, T. F., & Brouillard, R. (2003). Analysis and biological activities of anthocyanins. Phytochemistry, 64(5), 923-933. Lao, F., & Giusti, M. M. (2016). Quantification of Purple Corn (Zea mays L.) Anthocyanins Using Spectrophotometric and HPLC Approaches: Method Comparison and Correlation. Food Analytical Methods, 9(5), 1367-1380.

Madziva, H., Kailasapathy, K., & Phillips, M. (2005). Alginate-pectin microcapsules as a potential for folic acid delivery in foods. Journal of microencapsulation, 22(4), 343-351.

Mahdavi, S. A., Jafari, S. M., Assadpoor, E., & Dehnad, D. (2016). Microencapsulation optimization of natural anthocyanins with maltodextrin, gum Arabic and gelatin. International journal of biological macromolecules, 85, 379-385.

Mahdavi, S. A., Jafari, S. M., Ghorbani, M., & Assadpoor, E. (2014). Spray-drying microencapsulation of anthocyanins by natural biopolymers: A review. Drying technology, 32(5), 509-518.

Mateus, N., & de Freitas, V. (2008). Anthocyanins as food colorants. In Gould, K., Davies, K., Winefield, C. (Eds.), Anthocyanins (pp. 284-304). New York: Springer.

McCann, D., Barrett, A., Cooper, A., Crumpler, D., Dalen, L., Grimshaw, K., Kitchin, E., Lok, K., Porteous, L., Prince, E., Sonuga-Barke, E., Warner, J. O., & Stevenson, J. (2007). Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: a randomised, double-blinded, placebo-controlled trial. The lancet, 370(9598), 1560-1567.

Nayak, C. A., & Rastogi, N. K. (2010). Effect of selected additives on microencapsulation of anthocyanin by spray drying. Drying Technology, 28(12), 1396-1404.

Noda, Y., Kneyuki, T., Igarashi, K., Mori, A., & Packer, L. (2000). Antioxidant activity of nasunin, an anthocyanin in eggplant peels. Toxicology, 148(2), 119-123.

Nogueira, R. I., Santiago, M. C. P. I., dos Santos Gomes, F., Pontes, S. M., Lopes, P. R. C., & Freitas, S. P. (2011, October). Microencapsulation of bioactive compounds from

pomegranate (punica granatum l) juice using spray drying. In European Drying Conference, Balearic Island, Spain.

Norberto, S., Silva, S., Meireles, M., Faria, A., Pintado, M., & Calhau, C. (2013). Blueberry anthocyanins in health promotion: A metabolic overview. Journal of Functional Foods, 5(4), 1518-1528.

Nussinovitch, A. (1997). Hydrocolloid Applications. London: Chapman and Hall..

Olaya, C. M., Castaño, M. P., & Monroy, G. A. G. (2009). Stability of anthocyanins from Rubus glaucus and Solanum betaceum as affected by temperature and water activity. Acta Biológica Colombiana, 14(3), 141.

Østberg, T., Lund, E. M., & Graffner, C. (1994). Calcium alginate matrices for oral multiple unit administration: IV. Release characteristics in different media. International Journal of Pharmaceutics, 112(3), 241-248.

Ouwerx, C., Velings, N., Mestdagh, M. M., & Axelos, M. A. V. (1998). Physico-chemical properties and rheology of alginate gel beads formed with various divalent cations. Polymer Gels and Networks, 6(5), 393-408.

Patras, A., Brunton, N. P., O'Donnell, C., & Tiwari, B. K. (2010). Effect of thermal processing on anthocyanin stability in foods; mechanisms and kinetics of degradation. Trends in Food Science & Technology, 21(1), 3-11.

Peterson, J., & Dwyer, J. (1998). Flavonoids: dietary occurrence and biochemical activity. Nutrition Research, 18(12), 1995-2018.

Prior, R. L. (2003). Fruits and vegetables in the prevention of cellular oxidative damage. The American journal of clinical nutrition, 78(3), 570S-578S.

Rees, D. A. (1982). Polysaccharide conformation in solutions and gels–recent results on pectins. Carbohyd Polym., 2, 254–263.

Reyes, L. F., & Cisneros-Zevallos, L. (2007). Degradation kinetics and colour of anthocyanins in aqueous extracts of purple-and red-flesh potatoes (Solanum tuberosum L.). Food Chemistry, 100(3), 885-894.

Robert, P., Gorena, T., Romero, N., Sepulveda, E., Chavez, J., & Saenz, C. (2010). Encapsulation of polyphenols and anthocyanins from pomegranate (Punica granatum) by spray drying. International Journal of Food Science & Technology, 45(7), 1386-1394.

Rossi, M., Giussani, E., Morelli, R., Scalzo, R. L., Nani, R. C., & Torreggiani, D. (2003). Effect of fruit blanching on phenolics and radical scavenging activity of highbush blueberry juice. Food Research International, 36(9), 999-1005.

Selim, K. A., Khalil, K. E., Abdel-Bary, M. S., & Abdel-Azeim, N. A. (2008). Extraction, encapsulation and utilization of red pigments from roselle (Hibiscus sabdariffa L.) as

natural food colourants. In Alexandria Journal for Food Science and Technology. Special Volume Conference (pp. 7-20).

Serrano-Cruz, M. R., Villanueva-Carvajal, A., Rosales, E. J. M., Dávila, J. F. R., & Dominguez-Lopez, A. (2013). Controlled release and antioxidant activity of Roselle (Hibiscus sabdariffa L.) extract encapsulated in mixtures of carboxymethyl cellulose, whey protein, and pectin. LWT-Food Science and Technology, 50(2), 554-561.

Shahidi, F., & Han, X. Q. (1993). Encapsulation of food ingredients. Critical Reviews in Food Science & Nutrition, 33(6), 501-547.

Skrede, G., Wrolstad, R. E., & Durst, R. W. (2000). Changes in anthocyanins and polyphenolics during juice processing of highbush blueberries (Vaccinium corymbosum L.). Journal of Food Science, 65(2), 357-364.

Skrede, G., Wrolstad, R. E., Lea, P., & Enersen, G. (1992). Color stability of strawberry and blackcurrant syrups. Journal of Food Science, 57(1), 172-177.

Stevens, L. J., Kuczek, T., Burgess, J. R., Stochelski, M. A., Arnold, L. E., & Galland, L. (2013). Mechanisms of behavioral, atopic, and other reactions to artificial food colors in children. Nutrition reviews, 71(5), 268-281.

Toft, K., Grasdalen, H., & Smidsrod, O. (1986). Synergistic gelation of alginates and pectins. In ACS Symposium series-American Chemical Society (USA).

Tonon, R. V., Brabet, C., & Hubinger, M. D. (2010). Anthocyanin stability and antioxidant activity of spray-dried açai (Euterpe oleracea Mart.) juice produced with different carrier agents. Food Research International, 43(3), 907-914.

Tsuda, T., Horio, F., Uchida, K., Aoki, H., & Osawa, T. (2003). Dietary cyanidin 3-O-β-D- glucoside-rich purple corn color prevents obesity and ameliorates hyperglycemia in mice. The Journal of nutrition, 133(7), 2125-2130

U.S. FDA. (2016). Generally recognized as safe (GRAS). Retrieved from http//www.fda.gov/Food/IngredientsPackaging/GenerallyRecognizedasSafeGRAS/defaul t.htm.

Walther, B., Hamberg, L., Walkenström, P., & Hermansson, A. M. (2004). Formation of shaped drops in a fast continuous flow process. Journal of colloid and interface science, 270(1), 195-204.

Wang, W. D., & Xu, S. Y. (2007). Degradation kinetics of anthocyanins in blackberry juice and concentrate. Journal of Food Engineering, 82(3), 271-275.

Wang, Y., Bamdad, F., Song, Y., & Chen, L. (2012). Hydrogel particles and other novel protein-based methods for food ingredient and nutraceutical delivery systems. Encapsulation technologies and delivery systems for food ingredients and nutraceuticals, 412-450.

Weiss, B., Williams, J. H., Margen, S., Abrams, B., Caan, B., Citron, L. J., Cox, C., McKibben, J., Ogar, D. & Schultz, S. (1980). Behavioral responses to artificial food colors. Science, 207(4438), 1487-1489.

Wrolstad, R. E. (2004). Anthocyanin pigments—Bioactivity and coloring properties. Journal of Food Science, 69(5). Xiong, S., Melton, L. D., Easteal, A. J., & Siew, D. (2006). Stability and antioxidant activity of black currant anthocyanins in solution and encapsulated in glucan gel. Journal of agricultural and food chemistry, 54(17), 6201-6208.

Yang, Z., Han, Y., Gu, Z., Fan, G., & Chen, Z. (2008). Thermal degradation kinetics of aqueous anthocyanins and visual color of purple corn (Zea mays L.) cob. Innovative Food Science & Emerging Technologies, 9(3), 341-347.

Ying-chang, L., Chun-mao, L., Xian-jun, M., Yong, M., & Na21, Y. (2010). Study on the Microencapsulation of blueberry anthocyanins. Food and Fermentation Industries, 6, 24- 29.

Yousef, G. G., Brown, A. F., Funakoshi, Y., Mbeunkui, F., Grace, M. H., Ballington, J. R., Loraine, A., & Lila, M. A. (2013). Efficient quantification of the health-relevant anthocyanin and phenolic acid profiles in commercial cultivars and breeding selections of blueberries (Vaccinium spp.). Journal of agricultural and food chemistry, 61(20), 4806- 4815.

Zuidam, N. J., & Shimoni, E. (2010). Overview of microencapsulates for use in food products or processes and methods to make them. In Encapsulation technologies for active food ingredients and food processing (pp. 3-29). New York: Springer.

Figure 1. Anthocyanin aglycone basic structure.

Neutral and slightly acidic pH

pH < 2

pH from 3 to 6

Figure 2. Anthocyanin chemical structure transformation under different pH conditions in aqueous phase (adapted from: Houbiers et al., 1998), where R1 and R2 are H, OH or OCH3 groups.

Figure 3. Alginate structure at low pH range.

Figure 4. High methoxyl pectin chemical structure at low pH condition.

Chapter 2: Dissolution Kinetics of pH Responsive Smart Alginate-Pectin Hydrogel Particles Abstract

Encapsulation is used for protection of bioactive compounds during processing, storage, and passage through the upper gastrointestinal (GI) tract and delivery to the small intestine.

A number of pH responsive synthetic polymers are approved for drug delivery but are not allowed for food applications. We developed a biopolymer mixture composed of alginate and pectin that can form hydrogel when the pH is below 3.4. We also produced novel disc shaped particles which can potentially enhance the particle adhesion in intestines. As the pH increases, Al-P hydrogels go through a gel-sol transition and the dissolution kinetics of the hydrogel dominates the bioactive compound release. The goals of this study are to investigate the relative effects of factors contributing to the dissolution kinetics of Al-P hydrogel and to develop mathematical models characterizing the degradation behavior of the hydrogels under product storage and lower GI tract conditions. The volume change of spherical and disc shaped particles showed that at pH 3.0, the hydrogel particles would be stable in low pH beverages during storage. At pH 5.0 and 7.0, hydrogel particle dissolution followed a zero-order kinetic model. The 2.8% TGC 43:57 wt% Al-P disc particles had the fastest and the 2.2% TGC 82:18 wt% Al-P spherical particles had the slowest volume dissolution rate at pH 7.0 and 37 ᵒC. Activation energies of hydrogel particles were significantly affected by pH, particle shape and Al to P ratio. Such a smart biopolymer system which responds to pH provides an opportunity to use food as a vehicle for targeted delivery of bioactive compounds.

2.1 Introduction

Hydrogels have been investigated for delivery of nutrients and drugs in food and pharmaceutical industries and building scaffolds and sensors in medical applications. A variety of synthetic and natural polymers of hydrophilic properties were used as encapsulant to form hydrogels (Peppas et al, 2006). In food applications, FDA requires that encapsulant materials should have the “Generally Regarded As Safe (GRAS)” status.

Typically, encapsulation occurs by physical entrapment of bioactive materials within the polymer networks of hydrogel, thereby inhibiting interactions between encapsulated material and external environment (Kuang et al., 2010). Then, hydrogel particles of biopolymer origin, containing bioactive materials can be incorporated in the food material for delivery of beneficial compounds to the human body. Encapsulant provides a protection to the bioactive molecules by isolating them from external environment encountered during food during processing, storage, and human stomach (Anal & Singh, 2007; Cabane et al.,

2012; Drusch, 2007). It is also expected that upon reaching intestines, hydrogels should swell or dissolve so that beneficial compounds can be released in human body. Such “smart polymer” systems which are sensitive to physical and chemical stimulants were developed for controlled delivery of nutrients and drugs.

A number of protein and polysaccharides were used alone or in combination to form hydrogels (Wang et al., 2012). Alginate and pectin blend was shown to form hydrogels when the solution pH decreases to pH 3.0 (Higuita-Castro et al, 2012). Numerous natural polysaccharides including agar, alginate, carrageenan, pectin, gelatin can form hydrogels under different conditions, and are considered as GRAS (FDA, 2015). Their biocompatibility and biodegradability are advantageous as delivery media applied in the

food industry. The hydrogels formed by these materials have been produced (Burey et al.,

2008; R푒́ et al., 2009; Toft et al., 1986) by different encapsulation techniques. Droplet extrusion is a fabrication method where in a biopolymer solution is extruded through an opening into a curing solution. The morphology and dimension of the resultant Ca-alginate particles were shown to depend on the flow rate of the feed solution, the diameter of the opening, surface tension and the viscosity of the polymer solution (Burey et al., 2008; Chan et al., 2009). The typical dimension of the hydrogel particles fabricated by this technique is 0.5-6 mm with or without combining other techniques (Burey et al., 2008). Smaller hydrogel particles were reported to be generated with the droplet extrusion method combined with other techniques, such as electrostatic dripping, laminar jet breakup, jet cutting, jet nebulizer and disk nebulizer (R푒́ et al., 2009).

Hydrogel can be formed by addition of counter ions to alginate or pectin solutions. The gelation of alginate or pectin requires the presence of divalent cations or sugar, respectively

(Walther et al., 2004); however synergistic gelation of alginate-pectin mixtures occurs at pH below 3.4 (Higuita-Castro et al., 2012; Toft et al., 1986). Alginate-pectin mixtures have been used to encapsulate vitamin C, anthocyanins (Higuita-Castro et al., 2012) and folic acid (Madziva et al., 2005). Madziva and co-workers (2005) used an alginate and pectin gel mixture dropping into 0.1 to 1.0M of calcium chloride to fabricate hydrogel particles.

Higuita-Castro and co-workers (2012) showed alginate-pectin hydrogel particles were formed while pH dropped from 6.0 to 3.0 by adding Glucono-δ-lactone to biopolymer solution. However, the controlled degradation of hydrogels is very important for release of incorporated bioactive compounds. Degradation of alginate hydrogels was shown to be controlled by the biopolymer molecular weight, composition or the conditions of surrounding environment (Augst et al., 2006). Although the findings indicated the possibility of controlling the degradation kinetics of alginate gels, biopolymer blend hydrogels such as Al-P can provide more opportunities for applications of environmental stimuli responsive drug and nutrient delivery and the factors influencing their fabrication and degradation characteristics have not been elucidated.

The pH responsive characteristic distinguishes Al-P hydrogel from alginate or pectin only hydrogels since neither of them alone form a pH responsive gel. As the pH increases,

Al-P hydrogels are expected to go through a gel-sol transition and the kinetic of this transition can be a key factor in release of bioactive compound. Several factors including temperature, alginate to pectin ratio, shape of the particles, and the pH of the medium contribute to the dissolution of the hydrogel. The goals of the current study are to optimize the encapsulation parameters including the alginate-pectin ratio, flow rate, dripping distance, and pH of curing solution for varying the morphology and mechanical properties of the hydrogel particles and to investigate the relative effects of the factors contributing to dissolution kinetics of Al-P hydrogel particles. The findings are used to develop mathematical models characterizing the degradation behavior of the hydrogels at pH 5.0 in a full stomach and 7.0 in the small intestine and colon (Kong & Singh, 2008; Guerra et al.,

2012) so that the release behavior of bioactive compounds can be predicted.

2.2 Materials and methods

2.2.1 Materials

High guluronic acid content alginate (SF 120) was obtained from FMC Biopolymer

(Philadelphia, PA). High methoxyl content pectin (Pretested® Pectin, rapid and slow-set) were provided by TIC Gum (White Marsh, MD). Reagents (hydrochloric acid, potassium chloride, citric acid, sodium citrate, sodium phosphate monobasic and sodium phosphate dibasic) to prepare buffer solutions of pH 1.2, 3.0, 5.0 and 7.0 were purchased from Fisher

Scientific (Waltham, MA).

2.2.2 Preparation of gel solutions

Each of 2% (w/w) alginate (Al) and 4% (w/w) slow and rapid pectin (P) solutions were prepared by dispersing the powders in deionized water with a high shear mixer for 30 minutes at ambient temperature. After incubating at 4 °C for 8 hours to allow bubbles to rise, the solutions were mixed in various ratios to obtain 2.2% to 3.0% total gum concentration (TGC). Mixtures with four Al to P ratios were prepared at 50, 60, 80 and 90 wt% of alginate solution, the remainder being equal amount of slow and rapid pectin solutions. The mixtures were stored at 4 °C for an additional 12 hours for bubbles to rise.

2.2.3 Particle fabrication

The Al-P solutions were extruded through a 0.337 mm diameter (23 G) needle

(Hamilton, Nevada) with a peristaltic pump (Cole Parmer, IL) at volumetric flow rates of

0.017 or 0.022 ml/s. The droplets extruded through the needle formed hydrogel particles when they came into contact with gently agitated in low pH buffered solution. 0.1 m pH

1.2 HCl/KCl buffer or pH 3.0 citric acid/sodium citrate buffer were used as curing medium for hydrogel particles. The pH of the curing buffers (pH 1.2 or 3.0), dropping distances

(1.5-20 cm), Al-P solution flow rates (0.017 or 0.022 ml/s), TGC (2.2-3.0 wt%) and Al to

P ratio in gel (33:67, 43:57, 67:33 or 82:18 wt% ratio) were varied to evaluate their effects on particle size and shape. Al-P hydrogel particles were cured at 4 °C for 2 hours prior to characterization of the dimensions.

2.2.4 Particle size and shape characterization

A digital microscope (Amoeba Dual Purpose Digital Microscope, Celestron, CA) was used to photograph hydrogel particles placed on a microscope slide at 10X magnification.

Top and side view images of the hydrogel particles were obtained to determine the shape.

The size was determined using Image J software (National Institutes of Health, Bethesda,

MD).

2.2.5 Gel mechanical properties

2.2.5.1 Sample preparation

Hydrogel particles at 43:57 and 82:18% (wt) Al to P ratio and at both 2.2 and 2.8%

(wt) TGC were prepared to measure gel strength and elasticity. The gel solutions were mixed and poured into a plexiglass tube of 1 cm internal diameter and 1 cm height. One end of the tube was sealed with parafilm, and the tube was filled with the mixture from the open side. The tube was placed vertically inside a beaker filled with 40 ml of pH 1.2 buffer solution to promote gel formation. After 1.5 hours, an adequately thick gel layer formed on the open side. Then the parafilm was removed and the tube was inverted to expose the other end to the curing solution for 1.5 hours. The tube was placed horizontally in the curing solution to allow the buffer solution to diffuse into the gel from both ends. After 2 hours, the gels were ready for mechanical property tests.

2.2.5.2 Gel mechanical properties measurement

Mechanical properties of cylindrical gels with the dimensions of 1 cm diameter and 1 cm height were determined using Texture Analyzer (TA.XTPLUS, Hamilton, MA). A stainless steel cylinder probe with 40 mm diameter was used at a speed of 0.2 mm/s. Gel samples were compressed to determine gel strength and strain at the point of failure. Force versus deformation data were recorded. Then the data were converted to corrected stress, 𝜎퐶푂푅, and Hencky’s strain, 휖퐻, as defined below (Kaletunç et al., 1991b):

퐹(푡)[퐻0 − ∆퐻(푡)] (2.1) 𝜎퐶푂푅(푡) = 퐴0퐻0

퐻0 (2.2) 휖퐻(푡) = ln [ ] 퐻0 − ∆퐻(푡) where 퐻0 is the height of the gel specimen at time zero, ∆퐻(푡) is the absolute deformation at time t, and 퐴0 is the cross-sectional area of the gel specimen at time zero.

Gel strength, 𝜎퐶푂푅 at failure, and 휖퐻 at failure were determined from the stress- strain curve. The deformability modulus, 퐸퐷, was determined as the slope of the initial linear portion of the stress-strain curve as defined by Eq. (2.3):

𝜎퐶푂푅(푡) (2.3) 퐸퐷 = 휖퐻(푡)

Elasticity of the gels was determined from four successive compression- decompression cycles at deformation levels of 10, 15 and 20%. The deformation levels for the elasticity test were selected to prevent the gel from reaching the breaking point. The threshold force was set at 10 g to determine the starting point of the compression experiments. The degree of elasticity was defined as the ratio of recoverable work to total work for each compression-decompression cycle. The total work was calculated as the area under the stress-strain curve of the compression cycle and the recoverable work was

calculated as the area under the stress-strain curve of the decompression cycle (Kaletunç et al., 1991b).

2.2.6 Particle stability study

The stability of hydrogel particles was investigated as a function of temperature (4, 24 and 37 °C), pH (3.0, 5.0 and 7.0), Al to P ratio (82:18 wt% Al-P at 2.2 wt% TGC and 43:57 wt% Al-P at 2.8 wt% TGC) and hydrogel particle shape (spherical and disc). Throughout this manuscript, 82:18 wt% Al-P mixture at 2.2 wt% TGC is referred as formulation 1 (F1) and 43:57 wt% Al-P mixture at 2.8 wt% TGC is referred as formulation 2 (F2).

For each experiment, hydrogel particles were separated using a filtration funnel with vacuum and washed with a pH 1.2 buffer solution to remove the residual gel solution. To assess stability, ten particles were selected and characterized for shape and dimension by microscope. The hydrogel particles were transferred into 6ml of pH 5.0 or 7.0 buffers and agitated at 400 rpm. Particle size was monitored as a function of time over a 20min period in the pH 5.0 or 7.0 buffers. At selected time intervals, buffer was replaced with a pH 1.2 buffer to stop further dissolution. Then, the particles were removed from the solution and the hydrogel particle size was measured. Experiments in pH 3.0 buffer were conducted over a 23 day period. All beakers and solutions were sterilized using UV light to prevent microbial growth during the long term storage. Each experiment was performed in triplicate.

2.2.7 Data analysis

2.2.7.1 Statistical analysis

The statistical analysis was conducted to assess the variance among the replicates using ANOVA (analysis of variance), and the comparison of volume dissolution rate

constants at different pHs, gel ratios and hydrogel particle shapes using the t test and

Tukey’s HSD (Honest significant difference) test using JMP for Windows, release 10 (SAS

Institute Inc., Cary, N.C.).

2.2.7.2 Calculation of activation energy

The kinetics of hydrogel particle dissolution on a volume basis were evaluated by fitting both zero-order (Eq. (2.4)) and first-order (Eq. (2.5)) kinetic models to the data collected at pH 5.0 and 7.0.

푉(푡) = 푉0−푘0푡 (2.4)

퐿푛(푉(푡)) = 퐿푛(푉0) − 푘1푡 (2.5)

3 where 푘0 (mm /s) is the rate constant for zero order rate expression and 푘1 is the rate constant (1/s) for the first order rate expression. 푉(푡) and 푉0 are the volumes of the hydrogel particles at time t and zero, respectively. The dependence of the dissolution rate constant on temperature was described by the Arrhenius equation. The overall equation expressing volume change of hydrogel particles as a function of time and temperature for a zero-order kinetic model is defined as in Eq. (2.6).

퐸 (2.6) 푉 = 푉 − 퐴푡 exp (− 푎 ) (푡) 0 푅푇

3 where 퐴 is frequency factor (mm /s), Ea is activation energy (J), R is universal gas constant

(8.314 퐽 ∙ 푚표푙−1 ∙ 퐾−1) and T is absolute temperature (K).

The 퐸푎 and 퐴 values were calculated at each condition of pH, Al to P ratio and shape by the nonlinearfit function in MATLAB (R2015b, The MathWorks Inc., MA).

2.2.7.3 Dependence of activation energy on pH, Al to P ratio, and shape of the hydrogel particles

A second order model was applied to analyze the activation energy dependence on pH, alginate content, hydrogel particle shape and their interaction terms (Montgomery, 2008).

In the model, the hydrogel particle shape was considered as a discrete factor (spherical was level 1 and disc was level 2), while alginate content and pH were considered as continuous factors and were coded as -1 and 1. The 82 wt% alginate content (for F1 gel ratio) was coded as -1, while 43 wt% alginate content (for F2 gel ratio) was coded as 1. pH 5.0 was coded as -1 and 7.0 as 1. The final model is expressed by Eq. (2.7):

휏푆1(푠푝ℎ푒푟𝑖푐푎푙 푠ℎ푎푝푒) 휏푋1푆1 ∗ 푋1 휏푋2푆1 ∗ 푋2 (2.7) 푦 = 훽0 + 훽1 ∗ 푋1 + 훽2 ∗ 푋2 + { + { + { 휏푆2(푑𝑖푠푐 푠ℎ푎푝푒) 휏푋1푆2 ∗ 푋1 휏푋2푆2 ∗ 푋2

+ (훽12) ∗ 푋1 ∗ 푋2 + 휀 where 푦 is the predicted variable (in this case, it is the activation energy); 푋1 and 푋2 are

the independent continuous variables (pH and alginate content); 휏푆1 and 휏푆2 are coefficients for the discrete factor of spherical and disc hydrogel particle shape, respectively; 훽0 is the offset term; 훽1 and 훽2 are coefficients for pH and alginate

content, 휏푋1푆1 and 휏푋1푆2 are the coefficients for interaction term of pH and hydrogel

particle shape (S1 for spherical and S2 for disc shape); 휏푋2푆1and 휏푋2푆2 are the coefficients for the interaction term of alginate content and hydrogel particle shape; 훽12 is the coefficient for pH and alginate content interaction; and 휀 is the statistical error due to other sources of variability such as the system error. Each pair of discrete factor coefficients is constrained to a sum equal to zero. Equation 2.7 was solved by using JMP software (SAS

Institute, 2015) based on minimizing the sum of the squares of the difference between the

calculated and the predicted values.

2.3 Results

2.3.1 Fabrication of particles

Preliminary experiments were conducted at various Al to P ratio (33:67, 43:57, 67:33 and 82:18), TGC (2.2, 2.4, 2.8, and 3.0 %), flow rates (0.017 and 0.022 ml/s) and drop heights (1.5, 5, 10, 20 cm) by using pH 1.2 and pH 3.0 buffer solutions as curing baths.

Both hydrogel particle size and shape were found to be affected by the feed flow rate, Al to P ratio, the drop height, and curing bath pH. Conditions that generated irregularly shaped hydrogel particles were not included in further studies (Table 1). Gel samples cured in pH

1.2 buffer solution had considerably greater gel strength (1.5-2 times) than those cured in pH 3.0 buffer solution (data not shown). Therefore, only the hydrogel particles cured in pH

1.2 buffer were used for further hydrogel particle characterization. Spherical and disc shaped hydrogel particles were produced by using both F1 and F2 formulations (Table 1).

Specifically, the disc shaped hydrogel particles were produced by using F2 at both 0.017 and 0.022 ml/s flow rates, at 5 cm drop height and using F1 at both flow rates, at 20.0 cm drop height. The spherical hydrogel particles were generated using F1 at both 0.017 and

0.022 ml/s flow rates at 5 cm drop height and using F2 at 0.022 ml/s flow rate and 1.5 cm drop height (Table 1). Photographs of representative samples for both spherical and disc shaped hydrogel particles are shown in Figure 5a—c.

At a flow rate of 0.022 ml/s, for both F1 and F2, it was possible to produce both disc and spherical shaped hydrogel particles by varying the drop height (Table 1). Spherical hydrogel particles formed using F1 and F2 at 0.022 ml/s flow rate were similar in size, with a diameter of 2.68±0.24 mm and 2.76±0.10 mm respectively; while at 0.017 ml/s with F1

formulation, the diameter was 2.35±0.07 mm, an average of 15% smaller. The disc hydrogel particles generated by using F1 had 2.88±0.17 mm diameter and 1.40±0.12 mm thickness while F2 produced disc hydrogel particles of 3.22±0.15 mm diameter and

1.63±0.15 mm thickness. The disc hydrogel particles produced using F2 were 12% larger in diameter and 16% larger in thickness than the ones produced from F1. The disc hydrogel particles produced at low flow rate had larger diameters than their counterparts produced at high flow rate. The four conditions shown in bold in Table 1were selected for comparative analysis of hydrogel particle stability because they all can be produced at the same extrusion flow rate.

The shape of the hydrogel particles studied in this research were affected by a combination of factors including the distance between needle and curing bath surface, alginate-pectin ratio and the flow rate. At the same flow rate and Al to P ratio, as the drop height increased the shape of the hydrogel particles produced changed from sphere to a disc for both F1 and F2 formulations. The change of shape to disc occurred at a higher drop height for F1 hydrogel particles than F2 hydrogel particles. The forces affecting the hydrogel particle shape for the formation of Ca-alginate particles were described by Chan et al. (2009) as the viscous forces and the surface tension forces versus impact and drag forces. At the total gum concentration range investigated in this study (2.2-2.8%), surface tension values are expected to stay approximately constant. Therefore, viscous forces should be the determining factors for the hydrogel particle shape. The F1 formulation had nearly two times greater viscosity than the F2 formulation, 2 Pa∙s versus 1.2 Pa∙s. For the higher viscosity F1 formulation, the longer distance increases the momentum of a falling drop so that impact force between the drop and the surface of the curing bath is greater

creating a deformation on droplet surface resulting in the change in shape from a sphere to a disc. Depending on the physical properties of solutions, drop height can be used to control the shape of the hydrogel particles.

2.3.2 Mechanical properties of hydrogel particles

Mechanical properties of gels at Al to P ratios of 43:57 and 82:18 at each TGC of 2.2% and 2.8% were determined to investigate the effect of TGC and Al to P ratios on the gel strength, strain at failure, modulus and total work at failure (Table 2). The TGC did not show any significant effect on gel strength for the same Al to P gel ratio, while increasing

TGC significantly (p < 0.05) reduced the strain at failure and increased the modulus for both Al to P ratios. The extent of increase of modulus by TGC depended on the Al to P ratio. For 43:57 Al-P gels, the modulus did not increase significantly with TGC, while for high alginate content gels, the effect of TGC on modulus was significant (p < 0.05). The work required to break the gels decreased with increasing TGC. The higher strength of the high pectin gels can be related to the chemical interaction between alginate-pectin chain and hydrogen ions. It was reported that the gelation of alginate-pectin solution at low pH condition was affected by the content of methyl esterified D-galacturonic acid in pectin and L-guluronic acid residues in alginate (Thom et al., 1982). Because pectin used in this study had high methyl ester content and alginate contained high guluronic acid residues,

F2 gel may have a greater inter-chain interaction between alginate and pectin chains than the lower pectin content gel prepared using F1 formulation. A proposed explanation for the interaction between alginate and pectin chains at low and high pH conditions is shown in

Figure 6. For high guluronic acid content alginate and high methoxylated pectin chains, the pH decrease does not greatly affect pectin chains because of the high COOCH3 groups,

but decreases electrostatic repulsion due to the protonation of COO- groups to COOH in both alginate and pectin chains. Decreasing pH does not affect either hydrogen bonding between OH groups or hydrophobic interactions and attractive forces become dominant due to reduced repulsive electrostatic forces. The effect of interaction between alginate and pectin chains on gel strength was higher than the influence of increasing total gum concentration. On the other hand, a greater strain at failure associated with a higher work observed for F1 gels indicates that F1 gels have less resistant to deformation than F2 gels

(Table 2). This finding is consistent with the result that for the F2 formulation a shorter distance is needed to form a disc shaped particle. For F2 hydrogels, a higher stiffness

(modulus—material’s resistance to deformation) can be attributed to greater interaction between the alginate-pectin chains.

Elasticity of the hydrogels was studied as it can be related to the ability of hydrogel particles to return to their original shape after potential deformation during food processing

(Fig. 7 and 8). The total work is used to analyze the elasticity of the hydrogel particles because even if the hydrogel returns to the original dimensions, the total mechanical work may not be recoverable (Kaletunç et al., 1991b). The total work required increased with increasing deformation levels for both formulations. For 10 and 15% deformation levels, both formulations exhibited a similar total work (Fig. 7). However, at 20% deformation level, total work was higher for F1 than F2. During the successive compression- decompression cycles, the total work remained unchanged at 10 and 15% deformation levels. At 20% deformation level, the total work increased after the first compression, but thereafter remained constant for both formulations. The degree of elasticity of gels is defined as the percent of recoverable work after a compression (Fig. 8). Recoverable work

was approximately 80% at a 10% deformation level for both F1 and F2. The two formulations exhibited different response of elasticity to increasing deformation levels.

While the elasticity of gels steadily decreased at 15 and 20% deformation for F2 gels, elasticity of the F1 gels decreased at 15% deformation and remained constant at 20% deformation. In all the gels, the magnitude of the percent recoverable work decreased with the applied deformation increased indicating that increasing deformation weakens the structure of the gels and causes a loss of elasticity. Statistical analysis showed that the percent recoverable work was constant over the four cycles for both formulations at 10% deformation level.

The total work stayed approximately constant for four successive compression- decompression cycles suggesting that the elasticity of the gels do not change when they are subjected to small deformation levels. Similar behavior, characteristic of an elastic material, was reported for agar, alginate and kappa-carrageenan gels of similar concentration

(Kaletunç et al., 1991b). Non-elastic materials, such as cheese or ripe banana showed a large decrease of total work at even deformation levels of 12.5% (Kaletunç et al., 1991a).

2.3.3 Stability of hydrogel particles as a function of pH and temperature

Stability of hydrogel particles is quantified in terms of volume change under various pH conditions. The volume change of hydrogel particles were investigated under various pH and temperature combinations, simulating low pH beverages (pH 3.0) at refrigeration

(4°C), a full stomach (pH 5.0) and intestines (pH 7.0) at 37°C. Comparison of hydrogel particle volume changes at the three pH conditions showed that while the volume of hydrogel particle remained constant over a period of 3 weeks at pH 3.0 and 4 °C, the hydrogel particle volume decreased over a significantly shorter time scale of less than 10

minutes at pH 5.0 and 7.0 at 37°C (Fig. 9a and b).

2.3.3.1 Stability of hydrogel particles at low pH

Low pH beverages are appropriate delivery vehicles for hydrogel particles containing bioactive compounds. Therefore, stability of hydrogel particles in terms of volume change at pH 3.0 was investigated for 23 days. Experimental results did not show any dissolution of encapsulant material at pH 3.0. Swelling of hydrogel particles was observed with the volume change depending on the formulation of the hydrogel particles. Figure 10 shows the hydrogel particle volume changes for both F1 and F2 formulations and for both spherical and disc shaped hydrogel particles at 4°C and at pH 3.0. Similar results were observed for the hydrogel particles at all temperatures investigated. A maximum swelling of 15% for F1 formulation was observed for spherical hydrogel particles at 24°C. F2 hydrogel particles exhibited as high as 50% swelling for disc shaped hydrogel particles at

24°C. For all other conditions, F2 hydrogel particles exhibited a swelling of 30 percent or less measured after 8 days of storage with a residual swelling of 15 percent or less after 23 days (data not shown). Swelling was more significant for F2 hydrogel particles than F1 hydrogel particles. The results are in agreement with the scheme shown in Figure 8. As the pH increases the hydrogen ion concentration decreases thereby deprotonating carboxyl groups and leading to swelling of high pectin gels, more than high alginate gels. The stability of the hydrogel particles in pH 3.0 buffer under all three temperatures confirmed their possible application in low pH beverages under refrigeration conditions and at room temperature. Particles were also subjected to a temperature of 72°C for 15 seconds to simulate pasteurization conditions, which resulted in a minimal 0.3~4.7 percent volume decrease.

2.3.3.2 Stability of hydrogel particles above pH 5.0

Hydrogel particles dissolved as a function of time above pH 5.0 as shown in Figure

11a and b for F2 hydrogel particles at pH 7.0 and 37 °C. Above pH 5.0, the hydrogel dissolution rate was not highly depended on pH, as dissolution rates at pH 7.0 were approximately 1.1-1.5 higher than at pH 5.0. The dissolution rate of both disc and spherical hydrogel particles increased with increasing temperature. Comparison of Figure 11a and b indicates that disc hydrogel particles dissolved faster than spherical hydrogel particles at all temperatures investigated. Dissolution rate constants were found to be affected by pH, temperature, hydrogel particle shape and Al to P ratio as outlined in Table 3. The key factor affecting the release of encapsulated bioactive material above 5.0 is proposed to be by dissolution of the hydrogel particles as shown in Figure 12a. The release of the bioactive compound is controlled by the dissolution rate of the encapsulant material. This proposed explanation is consistent with experimental observation of decreasing volume as a function of time as shown in Figure 12b for spherical F1 hydrogel particles at pH 5.0 and 4 °C.

Volume dissolution rate constants changed between 3.8 × 10−3푚푚3 ∙ 푠−1 and

65.5 × 10−3푚푚3 ∙ 푠−1 depending on the formulation and environmental conditions.

Under lower gastrointestinal (GI) tract conditions of pH 7.0 and 37°C, F2 hydrogel particles are expected to have 2~2.5 times higher volume dissolution rate constants than

F1 hydrogel particles of the same shape due to the effect of formulation. The dissolution of the hydrogel particles at pH above 5.0 may be attributed to disruption of interactions between alginate and pectin because dissolution happened at a faster rate for high pectin

F2 hydrogels than high alginate F1 hydrogels. Lowering the total gum concentration of high pectin gels to 2.2% increased the volume dissolution rate constant of spherical

hydrogel particles to 17.6 mm3/s and 18.6 mm3/s at pH 5.0 and 7.0 respectively. Although both TGC and Al to P ratio affected the dissolution rate, Al to P ratio has a relatively higher influence on the rate constant than TGC (Table 3).

For the same formulation, spherical hydrogel particles had lower volume dissolution rate constants than their disc counterparts under all pH and temperature conditions. A comparison of diameter and thickness change for disc hydrogel particles showed that the diameter dissolution rate constants were significantly (p < 0.05) higher than the thickness dissolution rate constants (Table 4). Similarly, a comparison of diameter dissolution rate constants of disc and spherical hydrogel particles show that at the same pH, formulation and temperature, the diameter change of disc hydrogel particles had a higher rate constant than that of spherical hydrogel particles (Table 4). The observed results indicate that the geometry of the hydrogel particles affected the dissolution rate constant.

2.3.3.3 Modeling dissolution of hydrogel particles at high pH as a function of time and temperature

Dissolution behavior of the disc and spherical hydrogel particles produced from F1 and F2 at pH 5.0 and 7.0 followed a zero order kinetics at 4, 24 or 37°C (Fig. 11a and b).

The volume dissolution rate increased significantly (p < 0.05) with temperature for each

Al to P ratio, hydrogel particle shape and pH (Table 3).

The effect of the temperature on hydrogel particle dissolution rate constant followed the Arrhenius Equation at both pH 5.0 and 7.0. The fitting of the equation describing volume change of hydrogel particles as a function of time and temperature (Eq. (2.6)) showed that activation energies of hydrogel particles varied between 27.11 and 39.02 kJ/mol indicating that hydrogel particles with different formulations and shapes, and in

different pH environments had different sensitivities to temperature (Table 5). F2 hydrogel particles in general had higher sensitivity to temperature than F1 hydrogel particles. The higher sensitivity of the F2 gels to temperature may be due to the higher pectin content, because pectin is reported to be sensitive to temperature change as hydrogen bonds and hydrophobic interactions between pectin molecules are believed to decrease as temperature increases (Oakenfull & Scott, 1984).

2.3.4 Modeling of the effect of alginate-pectin ratio, pH level and hydrogel particle shape on activation energy

The activation energy of hydrogel particle dissolution depended on Al to P ratio, pH and hydrogel particle shape as shown in Table 5. For the same pH and hydrogel particle shape, alginate-pectin gel ratio affected the activation energy of F2 hydrogel particles dissolution significantly (p < 0.05) more than for F1 hydrogel particles. Effect of pH on the activation energy did not follow a specific trend as only the F1 disc hydrogel particles showed a higher activation energy at pH 7.0 compared with pH 5.0. The activation energy of hydrogel disc hydrogel particles dissolution was higher than that for spherical hydrogel particles at the same pH, except for the F1 hydrogel particles.

The activation energy data were fitted to a general second order model including a first order dependence of individual variables and their interaction terms. Fitting of the experimental data to Eq. (2.7) followed by F tests demonstrated that interaction terms of pH*alginate content and shape*alginate content were not statistically significant because the probability values were 0.10 and 0.77, respectively. After omission of the statistically non-significant terms, the general equation simplifies to Eq. (2.8):

−4.59 푆1 −1.15 푆1 (2.8) 퐸푎 = 39.29 + 0.99푝퐻 − 19.96퐴푙 + { − 5.13퐴푙 ∗ { 4.59 푆2 1.15 푆2

The F test was used to determine the probability that the full model (Eq. (7)) and the simplified model (Eq. (2.8)) yield different residuals—the differences between the calculated and predicted 퐸푎 values as described by Kaletunç (2007). The probability of equivalent fits by full and simplified models to the 퐸푎 calculated values is 0.64, indicating the omitted terms do not contribute significantly and the simplified model can be used to predict the 퐸푎 values for this particular data set. The model for prediction of 퐸푎 also provides means to determine the significance of each factor (pH, hydrogel particle shape and Al to P ratio) and their interaction terms on the activation energy. The probability that the factors are highly significant in affecting activation energy was 0.0024 for hydrogel particle shape, less than 0.0001 for Al to P ratio, and 0.001 for pH, which are all significantly smaller than α = 0.05.

The proposed simplified model can be used to predict the activation energy for pH between 5.0 and 7.0, alginate fraction from 0.43 to 0.82 and for either spherical or disc shaped particles. Equations 2.6 and 2.8 were combined into Eq. 2.9 to express the volume change of hydrogel particles as a function of time, temperature, particle shape, pH and alginate fraction.

1 −4.59 푆1 (2.9) 푉(푡) = 푉0 − 퐴푡푒푥푝(− (39.29 + 0.99푝퐻 − 19.96퐴푙 + { 푅푇 4.59 푆2

−1.15 푆 − 5.13퐴푙 ∗ { 1 )) 1.15 푆2

Eq. 2.9 was used to determine the calculated volume change and the calculated values were plotted against the corresponding experimental volume change (Fig. 13). The

relationship between the calculated and experimental volume change are described by a linear equation with slope of 0.98, intercept of 0.004 and a correlation coefficient of 0.91.

All data points appeared to be closely scattered around the perfect predictive equation shown by the green solid line. The average absolute error between the calculated and experimental values was determined to be 11.9% within the two green dashed lines displayed for the entire data set in Fig. 13.

These results can be used to predict the release of encapsulated bioactive compound.

The release rate of encapsulated bioactive compound is expected to be controlled by the dissolution rate of the pH responsive Al-P hydrogel particles above pH 5.0. The release rate can be predicted based on the position of the hydrogel particles in the GI tract. For fastest release of bioactive compounds at pH 7.0 and 37°C, F2 hydrogel particles with disc shape should be used, because such hydrogel particles are expected to release 70% of bioactive compounds within 2.5 min. For slowest release of bioactive compounds at pH

7.0 and 37°C, F1 hydrogel particle with spherical shape should be used, because hydrogel particles are expected to release 70% of the bioactive compounds within 7 min.

Based on the kinetic model, 40% of the bioactive compounds are expected to be released within 3-7 min after hydrogel particle exposure to pH 5.0 at 37 °C which is the expected environment of a full stomach. Given that the residence time reported for liquid products in the stomach is around 20 min (Collins et al., 1991) and can be up to 3 hours for solid products (Guerra et al., 2012), hydrogel particles in a beverage at pH 3.0 or lower should be consumed with an empty stomach so that hydrogel particles can be delivered intact to the intestine. To this end, Al-P hydrogel particles can be used as pH responsive drug or nutrient delivery vehicles controlled release of bioactive compounds from depends

on the dissolution kinetics of the smart biopolymer blend above pH 5.0.

2.4 Conclusions

Alginate- pectin hydrogel particles can be prepared either in spherical or disc shape by extrusion of mixed biopolymer solution into a low pH buffer for use in food applications.

The Al to P ratio, solution flow rate, and dropping distance affected the hydrogel particle shape and size. Al-P hydrogel particles maintained their integrity at pH 3.0 or below but dissolved as a function of time above pH 5.0 as described by a zero order kinetic model.

The dissolution was accelerated with temperature as defined by the Arrhenius relationship.

An empirical modeling of activation energies showed that hydrogel dissolution is highly affected by particle shape, Al to P ratio, and pH. The shape of the hydrogel particles is expected to influence the attachment of the hydrogel particles to the intestinal wall. The findings also revealed that the Al-P hydrogel dissolution rates were highly influenced by the shape of the particles thereby controlling the delivery of an encapsulated bioactive compound.

The dissolution rate of hydrogel is proposed to be the key factor controlling the release rate of the bioactive material in the GI tract. Further studies are ongoing in our laboratory to compare the release rates of bioactive compounds by diffusion versus by dissolution at pH above 5.0. The kinetic modeling as a function of time and temperature, and further characterization of kinetic parameters as a function of biopolymer ratio, pH and hydrogel particle shape will allow us to achieve more precise design of pH responsive hydrogel delivery systems with desired mechanical and dissolution properties necessary for encapsulation and controlled release of bioactive compounds.

Reference

Anal, A. K., & Singh, H. (2007). Recent advances in microencapsulation of probiotics for industrial applications and targeted delivery. Trends in Food Science & Technology, 18(5), 240-251.

Augst, A. D., Kong, H. J., & Mooney, D. J. (2006). Alginate hydrogels as biomaterials. Macromolecular bioscience, 6(8), 623-633.

Cabane, E., Zhang, X., Langowska, K., Palivan, C. G., & Meier, W. (2012). Stimuli- responsive polymers and their applications in nanomedicine. Biointerphases, 7(1), 9.

Chan, E. S., Lee, B. B., Ravindra, P., & Poncelet, D. (2009). Prediction models for shape and size of ca-alginate macrobeads produced through extrusion–dripping method. Journal of colloid and interface science, 338(1), 63-72.

Collins, P. J., Houghton, L. A., Read, N. W., Horowitz, M., Chatterton, B. E., Heddle, R., & Dent, J. (1991). Role of the proximal and distal stomach in mixed solid and liquid meal emptying. Gut, 32(6), 615-619.

Drusch, S. (2007). Sugar beet pectin: A novel emulsifying wall component for microencapsulation of lipophilic food ingredients by spray-drying. Food Hydrocolloids, 21(7), 1223-1228.

Higuita-Castro, N., Gallego-Perez, D., Love, K., Sands, M. R., Kaletunç, G., & Hansford, D. J. (2012). Soft Lithography-Based Fabrication of Biopolymer Microparticles for Nutrient Microencapsulation. Industrial Biotechnology, 8(6), 365-371.

Kaletunç, G., Normand, M. D., Johnson E.A., & Peleg, M. (1991a). “Degree of elasticity” determination in solid foods. Journal of food science, 56(4), 950-953.

Kaletunç, G., Normand, M. D., Nussinovitch, A., & Peleg, M. (1991b). Determination of elasticity of gels by successive compression–decompression cycles. Food hydrocolloids, 5(3), 237-247.

Kaletunç, G. (2007). Prediction of specific heat of cereal flours: A quantitative empirical correlation. Journal of food engineering, 82(4), 589-594.

Kong, F., & Singh, R. P. (2008). Disintegration of solid foods in human stomach. Journal of food science, 73(5), R67-R80.

Kuang, S. S., Oliveira, J. C., & Crean, A. M. (2010). Microencapsulation as a tool for incorporating bioactive ingredients into food. Critical reviews in food science and nutrition, 50(10), 951-968.

Madziva, H., Kailasapathy, K., & Phillips, M. (2005). Alginate-pectin microcapsules as a potential for folic acid delivery in foods. Journal of microencapsulation, 22(4), 343-351.

Montgomery, D. C. (2008). Design and analysis of experiments. John Wiley & Sons.

Oakenfull, D., & Scott, A. (1984). Hydrophobic interaction in the gelation of high methoxyl pectins. Journal of Food Science, 49(4), 1093-1098.

Peppas, N. A., Hilt, J. Z., Khademhosseini, A., & Langer, R. (2006). Hydrogels in biology and medicine: from molecular principles to bionanotechnology. Advanced Materials- Deerfield Beach Then Weinheim-, 18(11), 1345.

R 푒́ M. I., Santana, M. H. A. & d’Avila M. A. (2009). Chapter 7: Encapsulation Technologies for Modifying Food Performance. In Passos, M. L., & Ribeiro, C. P. Editor (Eds.). Innovation in food engineering: New techniques and products. CRC Press.

SAS Institute. (2015). JMP® 12 fitting linear models. SAS Institute Inc.

Thakur, B. R., Singh, R. K., Handa, A. K., & Rao, M. A. (1997). Chemistry and uses of pectin—a review. Critical Reviews in Food Science & Nutrition, 37(1), 47-73.

Thom, D., Dea, I. C. M., Morris, E. R., & Powell, D. A. (1982). Interchain associations of alginate and pectins. Prog. Food Nutr. Sci, 6, 97-108.

Toft, K., Grasdalen, H., & Smidsrod, O. (1986). Synergistic gelation of alginates and pectins. In ACS Symposium series-American Chemical Society (USA).

U.S. Food and Drug Administration (FDA). (2015). Generally Recognized as Safe (GRAS). Retrieved from http://www.fda.gov/Food/IngredientsPackagingLabeling/GRAS/.

Walther, B., Hamberg, L., Walkenström, P., & Hermansson, A. M. (2004). Formation of shaped drops in a fast continuous flow process. Journal of colloid and interface science, 270(1), 195-204.

Table 1. Effect of hydrogel fabrication parameters on particle size and shape. Curing solution pH is 1.2 buffer.

Drop Volumetric Flow: 0.017 ml/s Volumetric Flow: 0.022 ml/s

Height 2.8%TGC 43:57 2.2% TGC 82:18 2.8%TGC 43:57 2.2% TGC 82:18

(cm) wt% Al-P (F2) wt% Al-P (F1) wt% Al-P (F2) wt% Al-P (F1)

1.5 Ellipsoid Tear Drop Sphere Tear Drop (D: 2.76mm)

5.0 Disc Sphere Disc Sphere (D: 3.93mm; (D:2.35mm) (D:3.22mm; (D:2.68mm) T: 1.79mm) T:1.63mm)

10.0 Partial Dome Ellipsoid Partial Dome Ellipsoid

20.0 Complete Dome Disc Complete Dome Disc (D: 2.99mm; (D: 2.88mm; T: 1.59mm) T: 1.40mm)

Table 2. Mechanical properties of the alginate-pectin gels prepared in pH 1.2 curing bath.

Gel ratio TGC Strength (kPa) Strain at Modulus Work

failure (kPa) (kJ/m3)

a a a a 43:57 wt% 2.8% 480.50±8.41 0.79±0.02 93.63±7.20 131.24±5.12

Al-P a b a b 2.2% 466.05±3.04 1.00±0.01 86.55±8.41 162.38±1.61

b a b a 82:18 wt% 2.8% 418.53±12.25 0.76±0.01 120.95±15.77 119.13±2.18

Al-P b b a b 2.2% 408.76±14.52 0.85±0.05 83.67±10.86 169.65±3.04 a, b Levels not connected by the same letter are significantly different at α = 0.05.

Table 3. Volume dissolution rate constants under various pH and temperature conditions.

3 3 Volume dissolution rate constant X10 (mm /s) 2.2%TGC 82:18 wt% 2.8%TGC 43:57 wt%

Temperature Al-P (F1) hydrogel Al-P (F2) hydrogel

Shape pH (°C) particle particles

Disc pH 5.0 4 4.79±0.14 7.95±0.10 24 14.19±1.64 29.54±1.86

37 17.96±0.60 43.92±1.27

pH 7.0 4 6.49±0.15 10.50±0.43

24 16.37±0.22 45.12±1.77

37 25.97±1.48 65.47±1.11

Sphere pH 5.0 4 3.84±0.38 5.55±0.43 24 9.68±1.67 15.44±0.34

37 15.10±0.49 27.94±0.51

pH 7.0 4 4.24±0.76 5.74±0.19

24 11.95±2.85 16.29±0.54

37 15.95±0.83 33.11±0.76

Table 4. Comparison of the dimensional dissolution rate constants for the spherical and disc shaped hydrogel particles.

Dissolution rate constant × 103 (mm/s)

2.2%TGC 82:18 Al-P (F1) particles 2.8%TGC 43:57 Al-P (F2) particles

Disc particles Spherical Disc particles Spherical particles particles T

pH (°C ) Thickness Diameter Diameter Thickness Diameter Diameter

0.47±0.00 0.66±0.05*, 0.48±0.05* 0.51±0.01 0.90±0.04*,

5.0 4 6* ** * * ** 0.72±0.05**

2.07±0.32*, 1.26±0.08* 2.20±0.03 3.46±0.24*,

24 1.16±0.08* ** * * ** 1.71±0.04**

3.30±0.14*, 1.69±0.07* 3.21±0.10 4.60±0.68*,

37 1.78±0.04* ** * * ** 3.58±0.07**

0.90±0.01*, 0.59±0.05* 0.69±0.06 0.97±0.15*,

7.0 4 0.42±0.04* ** * * ** 0.77±0.02**

2.06±0.02*, 1.49±0.22* 3.01±0.06 5.35±0.39*,

24 1.35±0.02* ** * * ** 1.73±0.02**

3.29±0.17*, 2.00±0.04* 4.53±0.49 8.37±0.39*,

37 2.61±0.05* ** * * ** 3.38±0.09**

* Significant difference existed between the thickness and diameter dissolution rate constants of disc particles at α = 0.05.

** Significant difference between the diameter dissolution rate constants existed for the disc and spherical particles at α = 0.05.

Table 5. Activation energy and frequency factor determined by nonlinear fit for various pH, alginate-pectin ratio and particle shapes.

Volume frequency

Activation factor (mm3/s)

Energy

Level (kJ/mol)

2.8%TGC 43:57 wt% Al-P (F2) disc 39.02±0.94a 27686±1183

particle pH 7.0

2.8%TGC 43:57 wt% Al-P (F2) disc 38.35±1.60a,b 13464±4805

particle pH 5.0

2.8%TGC 43:57 wt% Al-P (F2) 35.38±0.83b,c 244678±12334

spherical particle pH 7.0

2.8%TGC 43:57 wt% Al-P (F2) 33.81±1.23c,d 133406±10292

spherical particle pH 5.0

2.2%TGC 82:18 wt% Al-P (F1) disc 30.45±1.75d,e 2261±304

particle pH 7.0

2.2%TGC 82:18 wt% Al-P (F1) 30.09±0.94e 727±101

spherical particle pH 5.0

2.2%TGC 82:18 wt% Al-P (F1) 27.75±1.53e 2956±290

spherical particle pH 7.0

2.2%TGC 82:18 wt% Al-P (F1) disc 27.11±0.83e 664±37

particle pH 5.0

Levels not connected by same letter are significantly different at α = 0.05.

Figure 5. Shapes of Al-P hydrogel particles a. 2.8% TGC 43:57 wt% Al-P spherical particle, b. 2.8% TGC 43:57 wt% Al-P disc particle top view, c. 2.8% TGC 43:57 wt% Al-P (F2) disc particle side view.

Figure 6. Alginate and pectin structure at low and high pH.

4000 4000

3500 3500

3000 3000

2500 2500

2000 2000

1500 1500

1000 1000

500 500

0 0 1 2 3 4 Number of Cycle

Figure 7. Total work as a function of successive compression- decompression cycles for

2.2% TGC 82:18 wt% Al-P hydrogels at 10% (○) 15% (□) and 20% (◊) deformation levels and for 2.8% TGC 43:57 wt% Al-P hydrogels at 10% (●) 15% (■) and 20% ( ◆) deformation levels.

100 100

80 80

60 60

40 40

20 20

0 0 1 2 3 4 Number of Cycle

Figure 8. Percent recoverable work as a function of successive compression- decompression cycles for 2.2% TGC 82:18 wt% Al-P gels at 10% (○) 15% (□) and 20%

(◊) deformation levels and for 2.8% TGC 43:57 wt% Al-P gels at 10% (●) 15% (■) and

20% (◆) deformation levels.

a b

y = 1 + 0x R= 0 82:18 Al:P disc 37C pH 3 y = 1 + 0x R= 0 43:57 Al:P disc 37C pH 3 y = 0.9728 - 0.0019808x R= 0.9853 y = 0.9898 - 0.00328x R= 0.9953 43:57 Al:P disc 37C pH 5 y = 0.98435 - 0.0027342x R= 0.99577 82:18 Al:P disc 37C pH 5 y = 1.0151 - 0.0050213x R= 0.99531 82:18 Al:P disc 37C pH 7 43:57 Al:P disc 37C pH 7

1.2 1.2

1 1

0.8 0.8

0.6 0.6

V/Vo

0.4 0.4

0.2 0.2

0 0 0 100 200 300 400 500 0 100 200 300 400 500 time (s) time (s) Figure 9. Volume change of hydrogel particles as a function of time for disc shape hydrogel particles at pH 3.0 (●), 5.0 (■) and 7.0 (◆) at 37 °C for a. 2.2% TGC 82:18 wt% Al to P ratio, b. 2.8% TGC 43:57 wt% Al to P ratio. Solid lines represent the fitted data to zero- order kinetic model.

1.4

1.2

V/Vo

0.8

0.6

0 5 10 15 20 25 Time (days)

Figure 10. Volume change of hydrogel particles as a function of time over a three week storage period at 4 °C. 2.8% TGC 43:57 wt% Al-P, disc shape (●), 2.8% TGC 43:57 wt%

Al-P, spherical shape (■), 2.2% TGC 82:18 wt% Al-P, disc shape (◆) and 2.2% TGC 82:18 wt% A l-P, spherical (▲).

a b

1.2 1.2

1 1

0.8 0.8

0.6 0.6

V/Vo

0.4 0.4

0.2 0.2

0 0 0 200 400 600 800 0 300 600 900 1200 1500 time (s) time (min)

Figure 11. Volume change of hydrogel particles as a function of time at pH 7.0 and at 4 °C

(●), 24 °C (■), and 37 °C (◆) for a. 2.8% TGC 43:57 wt% Al-P, disc shape hydrogel particles b. 2.8% TGC 43:57 wt% Al-P, spherical hydrogel particles. Solid lines represent the fitted data to zero-order kinetic model.

0min 5min 15min 25min

Figure 12. Dissolution behavior of a hydrogel, a. Proposed explanation for release behavior of uniformly distributed hypothetical bioactive compound (red dot) during dissolution of the hydrogel particle at high pH, b. The experimental observation of dissolution behavior of 2.2% TGC 82:18 wt% Al-P spherical hydrogel particles at pH 5.0 and at 4°C as a function of time. The red color was used for demonstration of hydrogel particle size change.

0.8

0.6

0.4

0.2

0 0 0.2 0.4 0.6 0.8 1 Experimental V/Vo

Figure 13. Calculated volume change using Eq. (2.9) versus experimental volume change for 43:57 Al-P spherical hydrogel particles in pH 7.0 (○), 43:57 Al-P spherical hydrogel particles in pH 5.0 (□), 43:57 Al-P disc hydrogel particles in pH 7.0 (◊), 43:57

Al-P disc hydrogel particles in pH 5.0 (∆), 82:18 Al-P spherical hydrogel particles in pH 7.0 (●), 82:18 Al-P spherical hydrogel particles in pH 5.0 (■), 82:18 Al-P disc hydrogel particles in pH 7.0 (◆), 82:18 Al-P disc hydrogel particles in pH 5.0 (▲).

Chapter 3. Extraction of Anthocyanins encapsulated in alginate-pectin hydrogel particles in pH 5.0 buffer solution

Abstract

Anthocyanins (ACN) are added to food to enhance color and health benefits.

Anthocyanins are known to be sensitive to heat, light and pH. Their delivery can be improved and their losses are reduced during food processing and storage by encapsulation. A direct method for quantification of encapsulated anthocyanin is essential for accurate determination of encapsulation efficiency and to assess the stability of encapsulated ACN during processing and storage. The proposed method comprises extraction of encapsulated ACN in alginate-pectin hydrogel by dissolving the particles in pH 5.0 buffer to determine the amount encapsulated. The gel-solution transformation of alginate-pectin hydrogel occurs as a response to pH change. The absorbance of ACN in the extract solution was measured and converted to the total monomeric ACN concentration by means of a pH 5.0 calibration curve. The optimum extraction efficiency was achieved at a solution volume to particle weight ratio of approximately 28 ml/g (10-11 particles in 3 ml of pH 5.0 buffer) hydrogel particles containing either purple corn or blueberry extract. The proposed approach can be used to estimate the amount of ACN encapsulated in the hydrogel particles directly and efficiently, and the method can be extended to other pH-responsive hydrogels.

3.1 Introduction

ACN are a group of natural colorants which provide color to fruits and vegetables from red to blue (Cevallos-Casals & Cisneros-Zevallos, 2004). Anthocyanins are proposed to have health benefits (He & Giusti, 2010). ACN is shown to be sensitive to 68

pH (Castaneda-Ovando et al., 2009), light (Carlsen & Stapelfeldt, 1997), and heat

(Patras et al., 2010). Encapsulation is utilized to protect bioactive materials from external environment and to achieve targeted delivery. Alginate-pectin (Al-P) blends have been developed to produce pH responsive hydrogel to encapsulate vitamin C and anthocyanins (Barry, 2013; Higuita-Castro et al., 2012).

Calculation of encapsulation efficiency requires accurate, direct determination of encapsulated material. For direct determination, it is necessary to use solvents to extract encapsulated material. Solvents such as acidified ethanol, water, acetone and methanol were used to extract anthocyanins from plant materials (Azmir et al., 2013).

Anthocyanins were extracted from glucan and whey protein gels by homogenizing the gel at pH 1.2 (Betz & Kulozik, 2011; Xiong et al., 2006) and from curdlan gels by acidified methanol solution (Ferreira et al., 2009). ACN encapsulated in alginate- chitosan (Belščak-Cvitanović et al., 2011) and alginate/inulin gels formed in CaCl2 solution (Ćujić et al., 2016) was extracted upon chelation by sodium citrate. Santos et al. (2013) estimated encapsulation efficiencies of ACN in polyethylene glycol by extraction with pH 1.0 buffer and in Ca-alginate with phosphate buffer at pH 7.4 or sodium citrate solution and by a mass balance approach. ACN encapsulated in Al-P hydrogel was quantified by ultrasound assisted acidified acetone extraction which is a long process with multiple extraction steps and using an organic solvent (Barry, 2013).

The objective of the study was to develop an efficient and green method to extract

ACN from Al-P hydrogel particles by using an aqueous solution as solvent based on the dissolution of the hydrogel above pH 5.0 (Guo & Kaletunc, 2016). The total monomeric ACN concentration within the particles was calculated by using the absorbance at 526 nm and a calibration curve at pH 5.0. The optimum extraction

conditions were investigated by varying solvent volume to hydrogel particle weight ratio.

3.2 Materials and Methods

3.2.1 Materials

Alginate (SF 120) was obtained from FMC Biopolymer (Philadelphia, PA) with guluronic to mannuronic acid ratio of 1.7. Slow and rapid set pectin (Pretested® Pectin) was provided by TIC Gum (White Marsh, MD) with degree of esterification of 63-67% and 71-75% respectively. Anthocyanin sources included purple corn extract (PC) and blueberry juice concentrate (BB). The blueberry (Vaccinium sp.) juice concentrate was provided by SVZ (Othello, WA) while the purple corn extract (Zea mays L.) as powder was obtained from Alicorp S.A.A. (Lima, Peru). Chemicals used to make buffer solutions (pH 1.2, pH 3.0 and pH 5.0) including hydrochloric acid, potassium chloride, sodium citrate and citric acid were purchased from Fisher Scientific (Waltham, MA).

3.2.2 Preparation of ACN stock solutions

The BB stock solution was prepared as 1:2 (w/w) ratio by mixing 3 grams of BB concentrate with 6 grams of 0.1 m pH 3.0 citrate buffer. The PC stock solution was prepared by mixing 3 grams of PC powder with 13 grams of pH 3.0 citrate buffer, followed by sonication for 10 min for complete dissolution and filtration with a syringe filter with a pore size of 0.45 μm (Fisher Scientific, Waltham, MA). The concentrations for PC and BB stock solution were measured by pH differential method (Giusti &

Wrolstad, 2001). The total monomeric ACN concentration of both solutions were determined based on the major component in the extract: for PC extract, cyanidin-3- glucoside was the major component (55.3%) while the malvidin-3-galatoside was the major component (35.0%) in BB concentrate (Barry, 2013). Therefore the total

monomeric ACN concentration of PC ACN was calculated as cyanidin-3-glucoside with using molecular weight of 449.2 g/mol and molar extinction coefficient of

26,900L ∙ 푚표푙−1 ∙ 푐푚−1 (Giusti & Wrolstad, 2001). Due to the lack of information for malvidin-3-galactoside (molar extinction coefficient), the total monomeric ACN concentration of BB ACN was calculated and reported as malvidin-3-glucoside equivalent with using molecular weight of 493.43 g/mol and molar extinction coefficient of 28,000 L ∙ 푚표푙−1 ∙ 푐푚−1 (Giusti & Wrolstad, 2001).

3.2.3 Preparation of alginate-pectin hydrogel particles

0.979 grams of alginate, 0.107 grams of rapid set pectin and 0.107 grams of slow set pectin were blended together and dispersed in deionized water with a high shear mixer at ambient temperature for 30 minutes. PC or BB stock solution was added into the alginate-pectin mixture slowly to obtain a 7.83% (w/w) ACN solution in the final gel solution and was stirred for additional ten minutes. Then the gel solution was degassed in a sonicator bath for 10 minutes and was left at ambient temperature for 2 hours to obtain a bubble-free solution.

The alginate-pectin (Al-P) hydrogel particles were produced by extruding the gel solution into pH 1.2 buffer as described by Guo & Kaletunc (2016). The Al-P-ACN gel mixture was extruded through a 23G (inner diameter of 0.337 mm) needle (Hamilton,

Nevada) by using a peristaltic pump (Masterflex model 7016-21, Cole Parmer, IL).

Droplets were generated at a volumetric flow rate of 0.022 ml/s and were collected in a gently agitated 0.1 m HCl/KCl pH 1.2 buffer. The drop distance between the needle and surface of the curing bath was maintained at 5 cm to produce spherical particles with an average diameter of 2.67 mm. Hydrogel particles were further cured for 2 hours at 4 ⁰ C. The total droplet weight to buffer weight ratio was kept at 0.136.

3.2.4 Extraction of ACN from hydrogel particles

ACN was extracted from hydrogel particles based on the dissolution properties of

Al-P hydrogel at pH 5.0 (Guo & Kaletunc, 2016). A 0.2 m pH 5.0 buffer solution was used as solvent. Solvent volume to particle weight ratio (SV/PW) was varied to find the optimum ratio for extraction. Six SV to PW values ranging from 56.3 to 18.8 (5 to 15 particles) for purple corn or four SV/PW values ranging from 37.5 to 23.1 (8 to 13 particles) for blueberry. Experiments were performed by placing particles in 3 ml of pH

5.0 buffer at 24⁰ C with vigorous stirring. After dissolution of hydrogel particles, the absorbance of the solution was recorded from 700 to 400 nm. The absorbance at 526 nm in pH 5.0 buffer was used to calculate the total monomeric ACN content in the extracted solution. A calibration curve was constructed by diluting an ACN solution prepared in pH 5.0 buffer of which total monomeric ACN concentration was determined by pH differential method (Giusti & Wrolstad, 2001).

Encapsulated ACN was also quantified by a material balance approach. The mass of ACN within the hydrogel particles after curing was calculated by subtracting the total mass of ACN in the curing bath at pH 1.2 from the total mass of ACN in the Al-

P-ACN solution. The total monomeric ACN concentration in curing bath was determined from the absorbance at 520 nm and a calibration curve developed at pH 1.2.

ACN stability at pH 5.0 was studied by measuring absorbance spectra of ACN solutions as a function of time. The contribution of Al-P solution to absorbance of ACN at pH 5.0 was also investigated by measuring absorbance spectra of dissolved blank Al-

P particles at the lowest SV/PW corresponding to the highest particle weight.

3.2.5 Data analysis

The statistical analysis was conducted to assess the variance among the different 72

SV/PW conditions using ANOVA (JMP for Windows, release 10, SAS Institute Inc.,

Cary, N.C.). The comparison of the amount of ACN in particles calculated by direct and indirect methods at each condition was determined by using the t test (JMP for

Windows, release 10, SAS Institute Inc., Cary, N.C.).

3.3 Results and discussion

Effect of pH on ACN concentration:

The developed extraction method uses 0.2 m pH 5.0 buffer for extraction of

ACN from Al-P hydrogel particles to provide sufficient buffer capacity. At pH 5.0, the

Al-P particles dissolved within approximately 15 minutes (Guo & Kaletunc, 2016).

Figure 14 shows the hydrogel particles before and after dissolution in pH 5.0 buffer.

Initially, the discrete particles are observed in a colorless solution (Fig. 14a), while after dissolution, a homogenous solution with a pink color was observed due to release of

ACN from the particles (Fig. 14b). The amount of ACN released was determined by measuring the absorbance of the final solution.

Absorbance spectra of the dissolved blank particles in pH 5.0 buffer were recorded from 700 to 400 nm to determine the contribution of pectin-alginate and ACN to the total absorbance. The concentration of alginate and pectin after dissolution of blank particles in pH 5.0 buffer at the highest number of particles (15 particles) was

0.1% (w/w). An Al-P solution of 0.1% (w/w) was prepared and its absorbance was recorded. The spectra of 0.1% Al-P solution and of dissolved blank particles are indistinguishable and do not contribute to the absorbance of dissolved particles containing ACN (Fig. 15).

The method developed used pH 5.0 buffer as a solvent for ACN extraction. The

ACN solution at the concentration of 4.56 μg/ml prepared in pH 1.2 and 5.0 buffer

solution had corresponding absorbance values of 0.412 at 520 nm and 0.173 at 526 nm respectively. Similarly, the absorbance values of PC ACN in pH 1.2 buffer was 0.721, and in pH 5.0 buffer was 0.351 at ACN concentration of 4.56 μg/ml. A shift in the wavelength corresponding to peak absorbance was observed for both BB and PC anthocyanins from 520 nm at pH 1.2 to 526 nm at pH 5.0 with different corresponding absorbance values (Fig. 16).

The calibration curves at pH 1.2 and 5.0 were constructed by using the absorbance spectra of known concentrations of ACN. Figure 17 shows the absorbance spectra at pH 5.0 for total monomeric PC-ACN concentrations of 0.89, 1.14, 1.96 and

3.69 μg/ml. The wavelength corresponding to maximum absorbance was determined to be 526 nm at all concentrations and absorbance increased linearly with increasing concentration confirming that the proposed approach of determining ACN concentration by dissolving hydrogel particles at pH 5.0 can be used as a direct measurement of encapsulated ACN. For extraction, the selection of solvent is critical as it should be able to dissolve the shell material effectively, be compatible with core material so it will not cause degradation and be an environmentally friendly, green solvent. Santos et al. (2013) reported that phosphate buffer at pH 7.4 or sodium citrate solution used to dissolve the Ca-alginate beads caused degradation of ACN and the

ACN concentration could only be determined by an indirect mass balance approach.

Control experiments with PC and BB ACN solutions were conducted to monitor the stability of ACN solutions by means of absorbance at pH 5.0 as a function of time over a period of 30 min. The rate of the concentration change for PC ACN was 0.0009

µg/ml per min, which resulted in a concentration change of 0.4% after 30 min; while for BB ACN it was 0.0004 µg/ml per min, which resulted in a concentration change of

0.2% after 30 min. Statistical analysis showed that the concentration change was

insignificant for both PC and BB ACN (p > 0.05) over 30 min. Therefore, the proposed direct method, which takes less than 30 min, can be used for extraction of ACN from hydrogel and for accurate and efficient determination of ACN.

Effect of SV/PW: SV/PW was varied to determine the optimal extraction condition.

For PC ACN, six SV to PW ratios between 56 and 19 ml/g; and for BB, five SV to PW ratios between 38 and 23 ml/g were used. The amount of ACN in hydrogel particles calculated by direct and indirect methods for PC and BB are shown in Table 6. Although the mass of ACN calculated by direct and indirect method was significantly different

(p < 0.05) for some cases, in general the results determined by each method were in good agreement. The difference between the results of the two approaches may be attributed to potential errors inherent to each method. The indirect method utilizes the material balance concept; calculating the mass of ACN within particles after curing indirectly from the mass of ACN in the curing bath. During curing, the ACN diffuses out from particles and water diffuses into the particles causing final curing bath volume to be lower than initial. The use of initial curing bath volume may cause a lower estimation of ACN mass in particles.

The amounts of ACN per gram of particle calculated with the direct method are plotted as a function of SV/PW (Fig. 18). The results show that the highest ACN extraction from particles was obtained at a SV/PW ratio of 28.1 ml/g (10 particles) for

PC ACN and 27.3 ml/g (11 particles) for BB ACN with a corresponding extracted ACN of 130.0 µg/g particle for PC ACN and 52.5 µg/g particle for BB ACN, respectively.

The lower estimation of ACN amount at high SV/PW can be attributed to very dilute

ACN samples while at low SV/PW, to the presence of high concentration of dissolved shell material.

Furthermore, the indirect method can only be used to estimate the initial amount of

ACN encapsulated in hydrogel particles and it cannot be used to determine the stability of ACN encapsulated in particles due to environmental effects and the degradation as a function of time. The direct method is essential to determine the effect of processing or storage conditions on ACN encapsulated within the hydrogel so that the amount of

ACN delivered to the human body can be accurately calculated.

3.4 Conclusions

The extraction of ACN from Al-P hydrogel particles provides a new approach to quantify directly the encapsulated ACN amount in particles. This method allows extraction of ACN from particles at pH 5.0 in less than 30 min during which the ACN concentration remains constant. This method is used to determine the amount of encapsulated ACN in pH responsive Al-P hydrogel. The protocol can be expanded to include other ACN sources and other pH responsive hydrogels.

Reference

Azmir, J., Zaidul, I. S. M., Rahman, M. M., Sharif, K. M., Mohamed, A., Sahena, F., Jahurul, M. H. A., Ghafoor, K., Norulaini, N. A. N., & Omar, A. K. M. (2013). Techniques for extraction of bioactive compounds from plant materials: a review. Journal of Food Engineering, 117, 426- 436.

Barry, A. M. (2013). Encapsulation, Color Stability, and Distribution of Anthocyanins from Purple Corn (Zea mays L.), Blueberry (Vaccinium sp.), and Red Radish (Raphanus sativus) in a Cold-Setting Pectin-Alginate Gel (Master thesis, The Ohio State University).

Betz, M., & Kulozik, U. (2011). Whey protein gels for the entrapment of bioactive anthocyanins from bilberry extract. International dairy journal, 21(9), 703-710.

Carlsen, C., & Stapelfeldt, H. (1997). Light sensitivity of elderberry extract. Quantum yields for photodegradation in aqueous solution. Food chemistry, 60(3), 383-387.

Castaneda-Ovando, A., Pacheco-Hernandez, M., Paez-Hernandez, M., Rodriguez, J.A & Galan-Vidal, C.A. ( 2009). Chemical studies of anthocyanins: A review. Food chemistry, 113, 859-871.

Cevallos-Casals, B. a, & Cisneros-Zevallos, L. (2004). Stability of anthocyanin-based aqueous extracts of Andean purple corn and red-fleshed sweet potato compared to synthetic and natural colorants. Food Chemistry, 86(1), 69–77.

Guo, J., & Kaletunç, G. (2016). Dissolution kinetics of pH responsive alginate-pectin hydrogel particles. Food Research International, 88(A), 129-139.

He, J., & Giusti, M. M. (2010). Anthocyanins: natural colorants with health-promoting

properties. Annual Review of Food Science and Technology, 1, 163–87.

Santos, D.T., Albarelli, J.Q., Beppu, M.M. & Meireles, M.A.A. (2013). Stabilization of anthocyanin extract from jabuticaba skins by encapsulation using supercritical CO2 as solvent. Food Research International, 50, 617-624.

Xiong, S., Melton, L. D., Easteal, A. J., & Siew, D. (2006). Stability and antioxidant activity of black currant anthocyanins in solution and encapsulated in glucan gel. Journal of agricultural and food chemistry, 54, 6201-6208.

Table 6. Extraction of PC and BB ACN with different SV to PW ratios in 3ml pH 5.0 buffer solution.

Purple corn ACNa Blueberry ACNb

SV/PW Initial Initial t test SV/PW Initial Initial t test (ml/g) mass of mass of (α=0.05) (ml/g) mass of mass of (α=0.05) ACN by ACN by ACN by ACN by extraction mass extraction mass method in balance method in balance droplets method droplets method (μg) in droplets (μg) in droplets (μg) (μg) 56.3 8.60±0.06 8.67±0.01 0.360 37.5 6.69±0.01 7.29±0.02 0.016* 40.2 12.35±0.11 12.13±0.03 0.203 30.0 8.46±0.09 9.02±0.02 0.045* 35.2 14.29±0.01 13.87±0.02 0.006* 27.3 10.45±0.20 10.38±0.20 0.639 31.3 16.07±0.10 15.60±0.02 0.047* 25.0 11.03±0.04 10.97±0.01 0.136 28.1 18.05±0.01 17.34±0.01 0.004* 23.1 11.31±0.06 11.87±0.01 0.028* 18.8 25.39±0.02 23.41±0.01 0.003* *The difference between the direct and indirect method results was significant (p < 0.05). a particle size: 2.68±0.26 mm b particle size: 2.67±0.23 mm

a b

Figure 14. Extraction of ACN encapsulated by alginate-pectin hydrogel particles. a)

Particles containing ACN was placed in 3 ml pH 5.0 buffer; b) ACN released to pH 5.0 buffer after particles dissolution.

ACN dissolved particle dissolved blank particle blank 0.25

0.2

0.15

0.1

0.05

-0.05 400 450 500 550 600 650 700 Wavelength (nm)

Figure 15. Visible absorbance spectrum of dissolved blank 2.2% Al-P particles at pH

5.0 (green) dissolved ACN containing (5.36 μg/ml) 2.2% Al-P particles at pH 5 (red), and 0.1wt% alginate-pectin solution (blue) (overlapped with green line).

Figure 16. Visible absorbance spectrum of ACN solution at a concentration of 4.56

μg/ml: PC ACN solution at pH 5.0 (red), PC ACN solution at pH 1.2 (blue), BB ACN solution at pH 5.0 (green) and BB ACN solution at pH 1.2 (black).

1.14 ug/ml 3.69 ug/ml 0.89 ug/ml 1.96 ug/ml

0.15

0.1

0.05

0 400 450 500 550 600 650 700 Wavelength (nm)

Figure 17. Visible absorbance spectrum of PC ACN solution in pH 5.0 at ACN concentrations of 0.89 μg/ml (blue), 1.14 μg/ml (red), 1.96 μg/ml (black) and 3.69

μg/ml (green).

Figure 18. The mass of ACN measured by direct method versus solvent volume to particle weight ratio: PC ACN (●) and BB ACN (■).

Chapter 4: Encapsulation of purple corn and blueberry extracts in alginate- pectin hydrogel particles

Abstract

Encapsulation of purple corn (PC) and blueberry (BB) extracts in alginate-pectin hydrogel particles was achieved to protect anthocyanins (ACNs) from degradation.

Combinations of alginate to pectin ratios at 82 to 18% and 43 to 57% and total gum concentrations (TGC) at 2.2% and 2.8% TGC were prepared to encapsulate both PC and BB ACN. The alginate-pectin hydrogel particles containing PC or BB extracts were produced by dripping solution into pH 1.2 buffer. Blueberry extract encapsulation efficiency was significantly higher than that of purple corn extract due to ACN chemical structure differences and the compatibility between the ACN structures and alginate- pectin hydrogel structure at the low pH environment. Effect of initial ACN concentration in droplets, particle shape, alginate to pectin ratio, TGC, ACN source, and curing bath conditions on percent encapsulation efficiency after curing (EEc) was investigated. The initial ACN concentration and particle shape didn’t influence the EEc, while the alginate to pectin ratio, TGC, ACN source and the pH of the curing bath showed significant effect on the EEc. The EEc was improved from 26% to 65% for PC

ACN and from 48% to 116% for BB ACN by augmenting curing bath with ACN at various concentrations. The ACN retention during storage (ARs) in hydrogel particles stored in pH 3.0 buffer was improved at low temperature and high particle weight to solution volume ratio. Higher amount of ACN was retained in the hydrogel particles when spherical particles were used. Encapsulation in hydrogel particles significantly reduced the anthocyanin degradation upon exposure to light. The degradation of ACN

was described with a first-order kinetics with half-life values of 1.7 years for encapsulated PC ACN and 2 months for PC ACN in solution. Hydrogel production and subsequent storage conditions can be optimized to increase the anthocyanin delivered to human body using the low pH beverages as a delivery vehicle.

4.1 Introduction

There is a growing interest to reduce the use of synthetic dyes in food and beverages by replacing with natural colorants because of the public concern of their adverse effects. Recent studies showed that synthetic dyes used in food products can cause adverse behavioral and neurological effects including hyperactivity in children

(McCann et al., 2007; Stevens et al., 2013; Weiss et al., 1980). Anthocyanins (ACNs) are natural food colorants present in fruits and vegetables with red, purple or blue colors and are utilized in the beverage industry. The color of ACN is most intense at low pH and intensity decreases and even color changes as pH increases and ACN does not have color around pH 4.0-5.0 ( Cevallos-Casals & Cisneros-Zevallos, 2004).

ACNs also have been recognized due to their health benefits in chronic diseases mainly attributed to their antioxidant properties. In vitro and in vivo models in the literature show that anthocyanins have anti-inflammatory and anti-carcinogenic activity and they have the potential to prevent cardiovascular disease, to control obesity and diabetes (He & Giusti, 2010).

In spite of the benefits, the limitation of ACN use in food products results from sensitivity of ACN color to pH, light, oxygen, heat and chemical structure (Rein,

2005). A color change of ACN with pH is a manifestation of a reversible chemical structure transformation among four anthocyanin forms between pH 1 to 6 (He &

Giusti, 2010). Above pH 7, irreversible structural changes occur and ACN is degraded. Reyes and Cisneros-Zevallos (2007) reported that the rate of color loss of

red-flesh potato, purple-flesh potato and grape at pH 3.0 was two times faster than those stored in pH 1.0 at 25⁰C. Therefore, ACN stability kinetics may be different at each pH. ACN stability is affected by high temperatures (70-90⁰C) during processing causing a 40-50% total anthocyanin losses (Patras et al., 2010). Studies of exposure of

ACN to light showed that the degree of color bleaching was ten times higher in UV light than visible light at pH 3.0-3.8 for elderberry extract and the effect was attributed to cleavage of covalent bonds (Carlsen & Stapelfeldt, 1997). Anthocyanin molecules with glycosylation and acylation naturally existing in radishes, red potatoes, red cabbage, black carrots were reported to have much higher ACN color stability than the non-acylated ACNs present in plants such as red grapes, blueberries, black currents, and purple corn (Cevallos-Casals & Cisneros-Zevallos, 2004; Giusti &

Wrolstad, 2003).

Encapsulation serves an important function in the delivery of bioactive compounds to human body. A number of delivery systems based on encapsulation of sensitive bioactive compounds are designed to improve the stability, to maximize the retention, and to control the release at the target locations in human body.

Encapsulation in biopolymer based hydrogels provides excellent opportunities for the food industry because they are from natural resources and approved for food use by

U.S. Food and Drug Administration for use in food materials (Burey et al., 2008;

FDA, 2006).

Hydrogels to be used in food applications were produced by using biopolymers such as alginate, pectin, chitosan, modified starch or carrageenan (Burey et al., 2008).

These hydrogel particles have been investigated to encapsulate hydrophilic bioactive compounds including antioxidants in Ca-alginate and Ca-alginate-chitosan (Deladino et al., 2008), lactose and ethanol in Ca-alginate (Gabardo et al., 2011), and glucose in

carrageenan (Nguyen & Luong, 1986). ACN extracts of various sources were reported to be encapsulated by hydrogels. Chokeberry extract was encapsulated in Ca-alginate and Ca-alginate-inulin hydrogel particles were prepared with sodium alginate solution with an encapsulation efficiency of 37% (Ćujić et al., 2016). Ca-alginate particles prepared with low viscosity sodium alginate solution were used to encapsulate wine waste extract containing ACN (Aizpuruza-Olaizola et al., 2016). The highest encapsulation efficiency of 63% within alginate gel was reported for haskap berries extracts by using 10% low viscosity sodium alginate solution (Celli et all., 2016)

However, the same investigators indicated that with 1% w/w reduction of alginate concentration, the encapsulation efficiency was decreased greatly to 18%.

Alginate and pectin blends were used to produce hydrogel with calcium ions for encapsulation of folic acid to improve encapsulation efficiency and to reduce leakage

(Madziva et al., 2005). Alginate and pectin gel blends were reported to form hydrogels at pH below 3.4 (Higuita-Castro et al., 2012; Toft et al., 1986, Guo &

Kaletunc, 2016). Vitamin C and anthocyanins were encapsulated by alginate-pectin hydrogels by adding glucono-δ-lactone to biopolymer solution to reduce slowly the pH from 6.0 to 3.0 to induce gelation (Higuita-Castro et al., 2012). Guo & Kaletunç (2016) reported that the gel-sol transformation of alginate-pectin hydrogel above pH 5.0 is kinetically controlled over a time period of 3 min to 20 min depending on the gel formulation, particle shape, temperature and pH. The absorption of ACN occurs in the intestines where the pH is above 5.0 (He & Giusti, 2010; Kong & Singh, 2008; Guerra et al., 2012). Therefore, the unique gel-sol transformation attribute of alginate-pectin can be utilized for design of hydrogel gel formulations for targeted and controlled release of bioactive compounds in human body. Encapsulation in alginate-pectin hydrogel is also expected to protect against pigment degradation of non-acylated ACNs

sources such as purple corn and blueberry which have limited food application opportunities.

The aim of this research was to determine the encapsulation efficiency of ACN within alginate-pectin hydrogel as a function of gel formation conditions including pH, alginate to pectin ratio, total hydrocolloid concentration, and ACN chemical structure. Blueberry and purple corn extracts were selected because they contain non- acylated ACN molecules known to have limited color stability. The storage stability of encapsulated ACNs in alginate-pectin hydrogel was assessed and compared with

ACN solution as a function of temperature and exposure time to light.

4.2 Materials and methods

4.2.1 Materials

Alginate with guluronic to mannuronic acid ratio of 1.7 (SF 120) was provided by

FMC Biopolymer (Philadelphia, PA). Two types of pectin (Pretested® Pectin, rapid set and slow-set) with degree of esterification of 71-75% and 63-67% respectively were obtained from TIC Gum (White Marsh, MD). Purple corn and blueberry extracts were used as anthocyanin source. Purple corn (PC) extract (Zea mays L.) was in powder form provided by Alicorp S.A.A. (Lima, Peru). Blueberry (BB) (Vaccinium sp.) juice concentrate was supplied by SVZ (Othello, WA). Reagents to prepare pH 1.2, 3.0, 5.0 and 7.0 buffer solution (hydrochloric acid and potassium chloride, sodium citrate and citric acid, sodium monobasic and dibasic) were obtained from Fisher Scientific

(Waltham, MA).

4.2.2 Preparation of ACN stock solutions

The PC stock solution was prepared by mixing 3 grams of PC powder with 13 grams of pH 3.0 buffer. The mixture was sonicated for 10 min and filtered by a syringe filter (Fisher Scientific, Waltham, MA) with pore opening of 0.45 μm. The

BB stock solution was prepared by mixing 3 grams of BB concentrated extract with 6 grams of pH 3.0 buffer. The concentrations of the final stock solutions for PC and BB were measured using pH differential method (Giusti & Wrolstad, 2001). Visible absorbance spectrum of PC and BB solutions at pH 1.0 and 4.5 were recorded between 700 and 400 nm by using a spectrophotometer (Cary 5000, Agilent, Santa

Clara, CA). The total monomeric ACN concentration of both solutions were calculated based on cyanidin-3-glucoside with a maximum absorbance peak at 520 nm using equations 4.1 and 4.2:

퐴 = (퐴520 − 퐴700)푝퐻 1.0 − (퐴520 − 퐴700)푝퐻 4.5 (4.1)

퐴 × 푀푊 × 퐷퐹 × 1000 (4.2) 퐶표푛푐푒푛푡푟푎푡𝑖표푛 표푓 퐴퐶푁 (푚푔⁄퐿) = 휀 × 퐿

where A is the absorbance, MW is the molecular weight of cyanidin-3-glucoside

(449.2 g/mol), DF is the dilution factor, ε is the molar absorptivity of cyanidin-3- glucoside at 520 nm (26,900 L∙cm-1∙mol-1), and L is the path length (1 cm).

4.2.3 Preparation of gel solution with anthocyanins

Alginate-pectin (Al-P) solution was prepared by blending alginate together with equal amounts of slow and rapid pectin powders and dispersing in deionized water with a high shear mixer for 30 minutes at ambient temperature. The PC solution or BB solution was added into the gel mixture slowly to obtain a 7.83% (w/w) ACN solution in the final mixture. The final BB ACN concentration in Al-P-ACN solution was 27.4 90

μg/ml. The PC ACN concentration was varied from 26.9 to 713.5 μg/ml in the final Al-

P-ACN solution to evaluate the effect of initial ACN concentration on encapsulation efficiency. Two total gum concentrations (TGC) of 2.2% and 2.8% were investigated.

Al-P solutions with Al to P ratios of 43:57 or 82:18 percent were prepared. After the

ACN stock solution addition, the gel mixture was stirred for ten minutes and degassed for 10 minutes in a sonicator bath, and was kept at ambient temperature for an additional

2 hours to obtain a bubble-free solution.

4.2.4 Hydrogel particle production

The Al-P hydrogel particles were produced by extruding the Al-P-ACN solution into a low pH buffer bath as described by Guo & Kaletunc (2016). The Al-P solutions were extruded through a 0.337 mm inside diameter (23G) needle (Hamilton, Nevada) using a peristaltic pump (Masterflex model 7016-21, Cole Parmer, IL) at a volumetric flow rate of 0.022 ml/s to generate droplets. The hydrogel particles were formed when droplets came into contact with gently agitated buffer. Buffers used were 0.1 M pH 1.2

HCl/KCl or 0.1M pH 3.0 citric acid/sodium citrate. The total droplet weight to buffer solution weight ratio was kept at 0.136. Either spherical or disc shaped hydrogel particles were produced by changing the dropping distance between the needle and surface of the buffer bath at different Al-P ratios (Guo & Kaletunc, 2016). For 2.2%

TGC and 82-18 Al-P solution, the spherical hydrogel particles were produced at a dropping distance of 5 cm while the disc particles were formed at a distance of 20 cm.

For 2.2 and 2.8% TGC and 43-57 Al-P solution, spherical particles were formed at 1.5 cm distance, and disc particles were at 5 cm. Al-P hydrogel particles were further cured at 4 °C for 2 hours.

4.2.5 Encapsulation efficiency of ACN in hydrogel

Hydrogel particles were removed from curing bath using filtration funnel with vacuum. The absorbance of the curing bath was measured with a spectrophotometer.

The total monomeric ACN concentration was calculated by using the absorbance at 520 nm and the calibration curve developed at pH 1.2 based on pH differential method. The mass of ACN in hydrogel particles after curing was calculated by subtracting the total mass of ACN in curing bath from the total mass of ACN in the solution extruded to form droplets to determine percent encapsulation efficiency after curing (EEc) (Eq. 4.3).

푚𝑖푛𝑖푡𝑖푎푙 퐴퐶푁 𝑖푛 푑푟표푝푙푒푡푠 − 푚퐴퐶푁 𝑖푛 푐푢푟𝑖푛푔 푏푎푡ℎ (4.3) 퐸퐸푐 = × 100% 푚𝑖푛𝑖푡𝑖푎푙 퐴퐶푁 𝑖푛 푑푟표푝푙푒푡푠

Encapsulation efficiency of ACN in hydrogels were investigated as a function of several parameters including, initial ACN concentration, botanical source of ACN, total gum concentration, alginate to pectin ratio, and curing bath pH.

4.2.6 ACN retention in hydrogel particles during subsequent storage

Loss of ACN from spherical and disc shaped hydrogel particles at 2.2% TGC and

82-18% Al-P ratio were studied in pH 3.0 buffer during storage. 0.34g of hydrogel particles were stored at a constant temperature. A quartz cuvette (21-Q-10, Starna Cells,

Atascadero, CA) containing 2.5ml of pH 3.0 buffer was also stored at the same constant temperature. Buffer temperature in the cuvette was monitored by using a thermocouple.

When the buffer reached to desired experimental temperature, hydrogel particles were added to the buffer which was stirred using an 8 mm long micro stir bar (Fisher

Scientific, Waltham, MA). The cuvette containing hydrogel particles was sealed with parafilm and placed inside the closed chamber to avoid light exposure. The ACN loss from hydrogel particles into surrounding buffer was determined by recording the absorbance spectra from 700 to 400 nm until an equilibrium was established between

ACN in particles and in surrounding solution. The study was conducted at 4, 24 or

37⁰ C for all hydrogel particles at a 0.34g particles/ 2.5ml buffer (w/v) ratio. The effect of particle to buffer ratio was also investigated at additional five ratios of 0.06, 0.116,

0.2, 0.27 and 0.33 by using spherical particles containing PC extract. The total monomeric ACN concentration in the solution was calculated using the absorbance at

520 nm and the calibration curve developed at pH 3.0 based on pH differential method.

Percent of ACN remaining in the particles at equilibrium was calculated based on the

ACN amount in the droplets. All experiments were carried out in duplicate.

푚퐴퐶푁 𝑖푛 푝푎푟푡𝑖푐푙푒 푎푓푡푒푟 푐푢푟𝑖푛푔 − 푚퐴퐶푁 𝑖푛 푏푢푓푓푒푟 푑푢푟𝑖푛푔 푠푡표푟푎푔푒 (4.4) AR푠 = × 100% 푚퐴퐶푁 𝑖푛 푝푎푟푡𝑖푐푙푒 푎푓푡푒푟 푐푢푟𝑖푛푔

An overall encapsulation efficiency (EEo) then was defined to take into account the loss of ACN in beverage and to determine the actual amount of ACN delivered to the human body.

4.2.7 ACN loss during gelation and particle curing

The loss of ACN from Al-P solution droplets during gelation and from hydrogel particles during curing were investigated. ACN total monomeric concentration in curing bath was monitored at 24⁰ C as a function of time for 30 min by measuring absorbance spectra from 700 to 400 nm. The experiments were conducted with 2.2%

TGC 82-18% Al-P ratio spherical and disc shaped particles containing PC and BB extract.

4.2.8 Improvement of encapsulation efficiency

The loss of ACN from hydrogel droplets were observed during gel formation in pH 1.2 buffer. ACN solution at various concentrations was added to the curing buffer bath to reduce diffusion loss from droplets/particles to surrounding solution. The effect of ACN augmentation on encapsulation efficiency was investigated at ACN concentrations of 0, 85, 130, 250 and 500 μg/ml total monomeric ACN concentration for purple corn extract and 0, 8, 18.9, 21.7 and 178.8 μg/ml for blueberry extract in the

curing solution. Encapsulation efficiency after curing in curing bath augmented with

ACN by considering the initial amount of ACN added to the curing bath.

4.2.9 Effect of light on free and encapsulated ACN during storage

The effect of light on total monomeric ACN content of free and encapsulated purple corn extract was investigated. Samples studied were PC extract solution in pH

3.0 buffer, hydrogel particles containing PC extract, and hydrogel particles containing

PC extract in pH 3.0 buffer. All samples were placed in glass vials with a rubber cap and sealed with parafilm. The light treatment was performed by exposing the sample to white fluorescent cool light (General electric company, F15T8), while the dark treatment was performed by covering the vials containing samples with aluminum foil at 20⁰ C.

The total monomeric ACN content in solution was determined as a function of time by recording the absorbance of ACN solution from 700 to 400 nm with a UV- visible spectrophotometer. The total monomeric ACN concentration in solution was calculated by using a calibration curve developed at pH 3.0 based on pH differential method. The amount of anthocyanin in hydrogel particles was determined by pH 5.0 extraction method (Guo & Kaletunc, 2017).

4.2.10 Porosity measurement

Porosity of hydrogel particles was determined based on measured total particle volume and estimated void volume. The total particle volume was measured by volume displacement method. A known mass of particles was added into a 25 ml volumetric flask filled with pH 1.2 buffer and the volume of the particles was calculated by dividing the mass of pH 1.2 buffer displaced by the density of pH 1.2 buffer. The void volume of hydrogel particles was estimated from the liquid volume based on the modified method described by Eiselt and co-workers (2000). Eiselt et al (2000) assumed that

void volume is equal to the liquid volume of the gel which can be determined by drying.

However, not all the water in a gel structure exists in the pores. Some of the water is present to hydrate the structure. Therefore, a known amount of particles were freeze- dried until a constant mass was reached. Then dried particles were placed in a dessicator over a saturated salt solution of potassium sulfate with a water activity of 0.976 at 20

⁰ C until an equilibrium was reached. The mass change between the wet particles and the hydrated particles was considered as the mass of liquid in the pores of the gel structure. The corresponding pore volume was determined by dividing the mass of liquid by the density of liquid in pores.

The porosity was calculated by using the equation (4.5):

(푚푤푒푡 푝푎푟푡𝑖푐푙푒 − 푚푟푒ℎ푦푑푟푎푡푒푑 푝푎푟푡𝑖푐푙푒)/𝜌푤푎푡푒푟 (4.5) ε = × 100% 푉푤푒푡 푝푎푟푡𝑖푐푙푒

4.3 Result and discussion

4.3.1 Evaluation of factors affecting encapsulation efficiency of ACN

Encapsulation efficiency and protection of the encapsulated material from environment are important characteristics to describe the effectiveness of encapsulation process. Hydrogel formation occurs when alginate-pectin droplets containing ACN comes in contact with low pH aqueous solution (Guo & Kaletunc, 2016). During gel formation and curing steps, ACN loss from droplets first and then from particles occur until an equilibrium is established between ACN inside the particles and in the curing solution which defines the encapsulation efficiency during gel formation and curing process (EEc). After curing, the particles are transferred to another low pH aqueous solution simulating a low pH beverage. Depending on the structural characteristics of the particles, ACN leakage from particles occurred until an equilibrium is established between ACN in the particles and in the surrounding aqueous solution. Anthocyanin 95

retention during storage (ARs) was calculated to consider ACN losses to the simulated beverage used as a delivery vehicle. An overall efficiency (EEo) was defined to account for the losses during gelation and curing, and storage, so the ACN delivered to human body can be determined. Various factors including total monomeric ACN concentration in initial hydrogel solution, botanical source of ACN extracts (chemical structure of

ACN), total gum concentration, alginate to pectin ratio, particle shape, pH of curing bath were investigated to evaluate their influences on EEc. Effects of light and temperature during storage on the ARs and on the EEo were evaluated to determine the effectiveness of encapsulation to protect ACN from environment.

Curing bath pH and temperature affected the EEc based on the studies conducted for forming the hydrogel particles at pH 1.2 or pH 3.0. For 2.2% TGC 82-18% Al-P spherical particles, EEc was 24.1% for particles produced at pH 1.2 while it was 17.0% for particles formed using pH 3.0 solution. The significant difference (p < 0.05) of EEc due to curing bath pH may be attributed to the proton concentration difference in the curing bath corresponding to two pH values. There is almost 100 times higher proton concentration at pH 1.2 compared to at pH 3.0 which may have resulted in an increased gelation rate of alginate-pectin mixture. The gelation mechanism between alginate and pectin chains was considered to be due to hydrogen bond (Thom et al., 1982; Guo &

Kaletunc, 2016) and the proton concentration available in curing bath is expected to affect the rate of gelation process. Morris & Chilvers (1984) reported that increasing the glucono--lactone concentration to decrease the pH of alginate-pectin solution, reduced the lag time for gelation and increased the gelation rate. Similarly,

WalkenstrÖm et al. (2003) suggested that the gelation speed was affected by the pH of the alginate-pectin gel solution which are in agreement with our findings. For longer gelation time, it is expected to have lower encapsulation efficiency. A similar

conclusion for the encapsulation efficiency was affected by the gelation speed was reported by Yeo and Park (2004).

The effect of curing temperature on the hydrogel particle formation was evaluated at 4⁰ C and at 24⁰ C for spherical hydrogel particles at 2.2% TGC and 82-18% Al-P ratio. The EEc values of 24.3% at 4⁰ C and 24.1% at 24⁰ C for PC ACN were observed indicating no significant difference in encapsulation efficiency at two temperatures. Gel formation happens due to diffusion of hydrogen ions in the surrounding solution to Al-

P droplets. Concurrently, ACN is expected to diffuse out from the droplets to surrounding pH 1.2 solution. Although, the diffusivity is expected to be increased by temperature, because both diffusion rates are affected by temperature leading to a negligible influence of temperature change on the net diffusion rate. The control experiment with pure PC ACN solution in pH 1.2 under the same temperature condition showed the total monomeric ACN concentration was stable over the studied time period of 2hrs.

4.3.1.1 Effect of initial ACN concentration

Encapsulation efficiency of PC ACN in 2.2% TGC 82-18 Al-P spherical particles were evaluated by varying the initial total monomeric ACN concentration in solution from 26.9 to 713.5 μg/ml. Variation of initial ACN concentration of Al-P-ACN solution up to 26 fold did not show any effect on EEc while both ARs and EEo increased (Fig.

19). Results showed that the EEc of ACN was ranged from 24.1 to 26.0% (Fig. 19), and the EE was not significantly (p > 0.05) affected by the initial total monomeric ACN concentration in particles. Although EEc remained almost constant, the actual amount of ACN encapsulated inside the particles increased. During subsequent storage, the diffusion of ACN from solid hydrogel particles was observed. The percent ACN retained (ARs) in particles increased with increasing initial amount of ACN in the

particles (Fig. 19). The increase of initial total monomeric ACN concentration in particle resulted in a higher concentration in curing bath solution and a proportionally improved amount in particle. Within the studied initial total monomeric ACN concentration range, the relationship showed linear increasing without reaching to a plateau. Result suggested that with increasing the initial total monomeric ACN concentration in particle, higher amount of ACN can be encapsulated within the investigated range.

The differences between EEc and ARs values suggest that rates of ACN loss from liquid droplets and from hydrogel particles are different. In curing solution, ACN loss can occur prior to gel formation from liquid droplets to surrounding solution while after gel formation from solid porous hydrogel particles to surrounding solution. It is expected that diffusivity of biological solutes in aqueous solutions will be 2-3 times larger than in biological gels (Geankoplis, 2003). The equilibrium curve between the mass of ACN inside the particle and in the curing bath showed a linear relationship with regression coefficient R2 at 0.996 while a nonlinear relationship existed at equilibrium between the mass of ACN in particle and in liquid during the storage (Fig. 20). The results shown in Fig. 20 indicates the differences in rates of mass transfer between liquid-liquid and solid-liquid systems.

4.3.1.2 Effect of particle shape

The spherical and disc shaped particles were produced at 2.2% TGC and 43-57%

Al-P ratio by using PC extract. The EEc values were determined to be similar, 21.6% for spherical particle and 21.4% for disc particles while the ARs values were 36.6% for spherical and 6.5% for disc shaped particles. The result indicated that although shape did not affect EEc, the retention of ACN in hydrogel particles changed significantly between spherical and disc particles during storage. Particle shape forms due to the

difference in forces encountered during droplet contact with the curing bath solution

(Guo & Kaletunc, 2016). Initially, droplet was in liquid state until the gelation process was completed and the hydrogel particle was formed. The loss of ACN from hydrogel particles occurs by diffusion. Amount of ACN diffused out will be directly proportional to the surface area of the particles which will be defined by the shape of the particle.

The surface area of the 2.2% TGC 82-18 Al-P spherical particle was 22.6 mm2 while for disc shaped particle, it was 25.7 mm2. The larger area of the disc particles therefore is expected to have a larger amount of ACN diffusion to the surrounding storage solution, and resulting in a lower percent ARs during storage.

4.3.1.3 Total gum concentration

The total gum concentration was increased from 2.2% to 2.8% to investigate the effect of TGC on encapsulation efficiency of ACN. Both 2.2% and 2.8% TGC 43-57%

Al-P spherical particles were generated by using PC and BB ACN. With increasing

TGC, the EEc of PC ACN was increased from 21.6% to 31.1%, and from 47.9% to 61.0% for BB ACN (Fig. 21). The major loss of ACN occurs during curing prior to gelling from liquid droplet to curing bath solution. Higher total polysaccharide concentration may increase the gelation rate leading to higher EEc value observed during curing. It is expected that increasing total gum concentration will reduce the porosity and the available water leading to higher retention (ARs) observed during the storage. The porosity of 2.8% TGC particles was 0.91 while it was 0.94 for 2.2% TGC particles for the same Al-P ratio of 43-57%.

The increase in EEc by increasing the TGC was also reported by Yeo & Park

(2004) and suggested as the higher concentration of polymer (PLGA) resulted in a faster gelation. Morris & Chilvers (1984) also reported increasing gelation rates with increasing TGC of alginate-pectin blends which is in agreement with the higher EEc

achieved in this study when TGC was increased from 2.2% to 2.8%.

4.3.1.4 Alginate to pectin ratio

The alginate to pectin ratio investigated was 82-18% and 43-57% alginate to pectin.

The two ratios were selected to be able to generate both spherical and disc shaped particles (Guo & Kaletunc, 2016). At the same TGC (2.2%), both 82-18 and 43-57 Al-

P spherical particles were produced encapsulating with PC and BB ACN and cured in pH 1.2 buffer solution. Results showed that increasing alginate fraction at the same

TGC increased the EEc for encapsulation of PC ACN (p < 0.05), while for BB ACN,

EEc was not significantly different for both alginate to pectin ratios during curing (p >

0.05). For storage studies, increasing alginate concentration resulted in higher ACN retention in hydrogel particles for both PC and BB ACN with increasing ARs values from 36.6 to 45.6% for PC ACN and from 20 to 57% for BB ACN.

The percent ACN diffused out during storage with either PC or BB ACN was lower by using 82-18% Al-P particles than 43-57% Al-P particles at the same TGC of

2.2%. This result suggested that the gel structure of 82-18 Al-P was relatively less porous. The diffusion of ACN within the hydrogel particles can be considered as diffusion in porous media, therefore related to the porosity of the gel particles.

Experiments showed that the porosity of 2.2% TGC 82-18 Al-P spherical particles was

0.89 while it was 0.94 for 2.2% TGC 43-57 Al-P spherical particles which can lead to higher ARs values observed for higher alginate ratios.

4.3.1.5 Effect of ACN botanical source

Comparison of the EEc and ARs values reported in Fig. 21 reveals that under all of the conditions investigated BB ACN encapsulation efficiency was higher than that of PC ACN. Blueberry extract contains a number of anthocyanins including delphinidin, cyanidin, petunidin, peonidin, and malvidin with the largest fraction

100

being malvidin-3-galactoside (Barry, 2013). Purple corn extract has cyanidin, pelargonidin, and peonidin with more than 50% being cyanidin-3-glucoside (de

Pascual-Teresa el al., 2002; Jing & Giusti, 2005; Lao & Giusti, 2016). Malvidin-3- galactoside contains one hydroxyl group and two methoxy groups while cyanidin-3- glucoside has two hydroxyl groups on the anthocyanidin ring. Therefore, malvidin-3- galactoside is relatively more hydrophobic and cyanidin-3-glucoside is relatively more hydrophilic. When each of these anthocyanins is encapsulated at low pH in alginate-pectin hydrogel which is expected to be uncharged (Guo & Kaletunc, 2016), alginate-pectin hydrogel is expected to favor hydrophobic over hydrophilic encapsulates, greater encapsulation of malvidin-3-galactoside is expected as is observed in this study. Under the same conditions, blueberry EEc and ARs values were higher than those of purple corn.

It is also observed that as the alginate-pectin ratio changed from 43-57% to 82-

18%, the EEc increased for PC ACN but did not change for BB ACN. Under low pH conditions, as the alginate fraction increases carboxylate groups in the gel increase.

Potential formation of ester linkages with hydroxyl groups of anthocyanins in the low pH environment (pH 1.2) of the curing bath, may result in higher EEc values for cyanidin-3-glucoside containing PC ACN. In addition to physical factors such as porosity, encapsulation efficiency is affected by physico-chemical interactions and potential formation of covalent linkages between the anthocyanins and alginate-pectin.

4.3.2 Improvement of encapsulation efficiency of ACN

Encapsulation efficiency values after curing were approximately 26% for PC ACN and was 47% for BB ACN for spherical hydrogel particles produced at 2.2% TGC, 82-

18% Al-P. It was apparent that most ACN loss occurred prior to gel formation from liquid droplets to curing bath. Therefore, curing bath was augmented with ACN prior

101

to dropping of particles to reduce the diffusion of ACN from droplets to curing bath solution.

Curing baths containing 8, 13, 19 and 22 μg/ml of BB ACN and 85, 130 and 250

μg/ml of PC ACN were prepared for investigation of improving encapsulation efficiency (Table 7). For both ACN studies, EEc values increased as the concentration of corresponding ACN in curing bath increased potentially due to reduction of diffusion loss prior to hydrogel formation. Correspondingly, EEo values also increased indicating that the overall delivery to human body can be improved two times for BB ACN and three times for PC ACN by enrichment of curing bath with ACN solution. When the concentration of added ACN in curing bath was normalized by the ACN concentration in droplets for comparison of both data set, it was apparent that addition of ACN to curing bath had a higher influence on PC than BB. The higher influence may be attributed to the presence of two hydroxyl ions in PC ACN in comparison to one hydroxyl group in BB ACN.

4.3.3 Effect of storage conditions on ACN release

4.3.3.1 Particle weight to solution volume ratio

In previous studies, particle weight to solution volume (PW/SV) ratio was kept constant at 0.136 for both curing bath and storage studies. It is considered that during the storage, equilibrium condition can be shifted to favor the reduction of ACN loss from hydrogel particles if a larger amount of particles were added to storage solution.

Spherical hydrogel particles containing PC ACN were prepared at 2.2% TGC 82-18%

Al-P and were added to storage solution at PW/SV values ranging from 0.06 to 0.33

(Table 8). The percent overall encapsulation efficiency (EEo) increased first and reached to a constant value around PW/SV=0.3 (Fig. 22). As the amount of particles in solution increased, a larger amount of anthocyanin can be retained within the particles at

102

equilibrium.

4.3.3.2 Storage temperature

The effect of storage temperature in retaining the anthocyanins within the hydrogel particles were investigated by using particles produced at 2.2% TGC, 43-57% and 82-

18% Al-P. The spherical particles were placed inside storage solution at pH 3.0 at temperatures of 4, 24 and 37⁰C. The percent EEo values decreased with temperature for both particles studied (Fig. 24). During storage ACN loss from hydrogel particles occurs due to diffusion. The diffusivity of ACN is expected to increase with temperature based on the Arrhenius relationship which will lead lower encapsulation efficiencies with increasing temperature. Fig. 24 shows that EEo for 82-18% Al-P hydrogel particles is higher than that of 43-57% Al-P particles, indicating a higher expected diffusivity for lower alginate content particles which have higher porosity values. The dependence of

EEo on temperature for higher alginate (88-12%) containing particles is more pronounced than lower alginate (43-57%) containing particles which may be attributed to enhancement of anthocyanin stability by pectin. Buchweitz and co-workers (2013a, b) reported enhancement of stability anthocyanins from black currant and strawberry upon addition of pectin during 18 weeks of storage at 20 ⁰C. However, the stabilization effect depended on both the anthocyanin structure and the pectin source and structure.

Investigators reported that the strawberry anthocyanin mainly, pelargonidin-glycosides, was stabilized by apple pectins and black current anthocyanins, cyanidin- and delphinidin-glycosides were stabilized by citrus pectins and was attributed to the number of hydroxyl groups. The observation of higher stability with temperature of higher pectin formulations in this study may be an indication of stronger interaction between the hydroxyl groups of cyanidin- glycosides and pectin through hydrogen bonding (Buchweitz et al., 2013a).

103

4.3.3.3 Effect of light

Anthocyanins were reported to be sensitive to light and exposure to light causes degradation of anthocyanins (Carlsen & Stapelfeldt, 1997). The effect of encapsulation to protect ACN from light was investigated by comparing the ACN losses for ACN aqueous solution, for ACN encapsulated in hydrogel particles and for ACN encapsulated in hydrogel particles dispersed in solution (PIS) exposed to fluorescence cool white light. The samples stored in the dark were used as control samples for each data set. Total monomeric ACN content monitored as a function of time at 20 ⁰C reveals anthocyanin degradation due to fluorescence cool white light exposure with varying degrees and protective ability of hydrogel against fluorescence cool white light (Fig.

24). Control samples stored in the dark also showed some degradation.

Anthocyanin degradation was described by first-order kinetics. The corresponding kinetic parameters are given in Table 9. The comparison of rate constants show that stability of ACN to fluorescence cool white light exposure from highest to lowest are given as ACN encapsulated in hydrogel particles, ACN encapsulated in hydrogel particles dispersed in solution, and ACN aqueous solution. Half-life values calculated for each case demonstrated that while ACN encapsulated in hydrogel have a half-life value of 1.7 years, the half-life of ACN aqueous solution was 2 months. Hydrogel particles are desirable because the color will be visible due to transparent properties of hydrogel. Fluorescence cool white light exposure studies showed that although transparent, hydrogel can act as a light barrier and delay the anthocyanin degradation well beyond the expected shelf-life of fruit juices. In applications, hydrogel particles are expected to be dispersed in a beverage. The fluorescence cool white light exposure of hydrogel particles in solution showed a half-life value of 9 months for anthocyanins which is almost five times longer than ACN in solution. The faster degradation

104

observed for hydrogel particles dispersed in solution in comparison with the hydrogel particles can be due to degradation of ACN present in solution which leads to diffusion of ACN from particle to solution due to ACN concentration difference between the particle and solution. All the control samples stored in the dark also exhibited loss of anthocyanins although at much lower levels which may be attributed to oxygen exposure. Similar to light exposure studies, half-life value for ACN encapsulated in hydrogels was the highest which may be speculated for hydrogel being a potential oxygen barrier. This finding will need to be further studied in a controlled environment.

Under light, the fastest degradation was found when using ACN aqueous solution sample with the rate constant at 0.012 hr-1. A similar result was found also with purple corn ACN in pH 3.0 buffer under fluorescent light at 20 ⁰C (Cevallos-Casals &

Cisneros-Zevallos, 2004). By extracting the color retention of the purple corn ACN solution at maximum wavelength as a function of time, the analysis showed the color retention of purple corn ACN followed first-order kinetics with a rate constant of 0.013 hr-1, which was similar with the result obtained in this research (0.012 hr-1).

The protection provided by alginate-pectin hydrogel particle may be due to both the physical barrier and the intermolecular association between polyuronic acids from pectin and ACN. Pectin-ACN interactions leading to anthocyanin stability were suggested by several investigators in the literature (Hubbermann et al., 2006; Maier et al., 2009; Buchweitz et al., 2013 a, b). Maier and co-workers (2009) reported that the stability improvement of grape pomace extract encapsulated in pectin gel was better than in gelatin gel when stored under neon light and dark at 20 ⁰ C. Their results showed that ACN retained in pectin gel after 2 weeks was 73% for the samples exposed to light and was 85% for the samples stored under dark conditions. The findings in this study was in agreement Maier et al., (2006) results as the ACN remained in the hydrogel

105

particles was 70% under light and 90% under dark conditions after 15 days.

Hubbermann and co-workers (2006) found that the ACN-rich elderberry stability was improved by alginate aqueous solution, while the ACN-rich black currant concentrate was stabilized by both pectin or alginate aqueous solution during storage in pH 3.9 acetic buffer under daylight at ambient temperature. Comparison of the model gels produced with pectin, agar-agar and gelatin showed that the stability of ACN in both the elderberry and black currant samples during storage were significantly better in pectin model gels than in agar and gelatin samples.

4.3 Conclusion

Alginate-pectin hydrogel particles were used to encapsulate anthocyanins extracted from blueberry and purple corn. The encapsulation efficiency during curing increased with increasing total gum concentration, lower pH, increasing relative alginate content.

The physico-chemical interaction between the alginate-pectin and anthocyanin resulted in different encapsulation efficiencies of blueberry or purple corn anthocyanins with different chemical structures. During gel formation step, by augmenting the ACN concentration in the curing bath, encapsulation efficiency can be increased.

During storage of the hydrogel particles in solution, anthocyanin losses were reduced by storing the particles at low temperatures and at high particle weight to solution ratios. Higher ACN losses during the storage was also observed for disc shaped particle in comparison with the spherical shaped particles. Alginate-pectin hydrogel acted as a barrier to flurescent light to significantly reduce the anthocyanin degradation rate. Alginate-pectin hydrogel protects anthocyanin from detrimental effects of surrounding environment and also have a stabilization effect on anthocyanins due to chemical interaction between the alginate-pectin system and various forms of anthocyanidins present in the edible plant sources. A number of parameters can be

106

optimized during hydrogel production and subsequent storage to increase the anthocyanin delivered to human body by using the low pH beverages as a delivery vehicle.

107

Reference

Buchweitz, M., Speth, M., Kammerer, D. R., & Carle, R. (2013). Stabilisation of strawberry (Fragaria x ananassa Duch.) anthocyanins by different pectins. Food chemistry, 141(3), 2998-3006.

Celli, G. B., Ghanem, A., & Brooks, M. S. L. (2016). Optimized encapsulation of anthocyanin-rich extract from haskap berries (Lonicera caerulea L.) in calcium- alginate microparticles. Journal of Berry Research, 6(1), 1-11.

Cemeroğlu, B., Velioglu, S., & Isik, S. (1994). Degradation kinetics of anthocyanins in sour cherry juice and concentrate. Journal of Food Science, 59(6), 1216-1218.

Carlsen, C., & Stapelfeldt, H. (1997). Light sensitivity of elderberry extract. Quantum yields for photodegradation in aqueous solution. Food chemistry, 60(3), 383-387.

108

de Pascual-Teresa, S., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (2002). LC–MS analysis of anthocyanins from purple corn cob. Journal of the Science of Food and Agriculture, 82(9), 1003-1006.

Deladino, L., Anbinder, P. S., Navarro, A. S., & Martino, M. N. (2008). Encapsulation of natural antioxidants extracted from Ilex paraguariensis. Carbohydrate Polymers, 71(1), 126-134.

Eiselt, P., Yeh, J., Latvala, R. K., Shea, L. D., & Mooney, D. J. (2000). Porous carriers for biomedical applications based on alginate hydrogels. Biomaterials, 21(19), 1921- 1927.

Francis, F. J. & Markakis, P. C. (1989). Food colorants: Anthocyanins, critical reviews in food science and nutrition. Food Science and Nutrition, 28(4), 273-314.

Gabardo, S., Rech, R., & Ayub, M. A. Z. (2011). Determination of lactose and ethanol diffusion coefficients in calcium alginate gel spheres: predicting values to be used in immobilized bioreactors. Journal of Chemical & Engineering Data, 56(5), 2305-2309.

Giusti, M. M., & Wrolstad, R. E. (2001). Characterization and measurement of anthocyanins by UV‐ visible spectroscopy. Current protocols in food analytical chemistry. F:F1:F1.2.

Giusti, M. M., & Wrolstad, R. E. (2003). Acylated anthocyanins from edible sources and their applications in food systems. Biochemical Engineering Journal, 14(3), 217- 225.

Guo, J., & Kaletunç, G. (2016). Dissolution kinetics of pH responsive alginate-pectin hydrogel particles. Food Research International, 88(A), 129-139.

He, J., & Giusti, M. M. (2010). Anthocyanins: natural colorants with health-promoting properties. Annual Review of Food Science and Technology, 1, 163–87.

Hubbermann, E. M., Heins, A., Stöckmann, H., & Schwarz, K. (2006). Influence of acids, salt, sugars and hydrocolloids on the colour stability of anthocyanin rich black currant and elderberry concentrates. European Food Research and Technology, 223(1), 83-90.

109

Jing, P., & Giusti, M. M. (2005). Characterization of anthocyanin-rich waste from purple corncobs (Zea mays L.) and its application to color milk. Journal of agricultural and food chemistry, 53(22), 8775-8781.

Kırca, A., Özkan, M., & Cemeroğlu, B. (2007). Effects of temperature, solid content and pH on the stability of black carrot anthocyanins. Food chemistry,101(1), 212-218.

Kong, F., & Singh, R. P. (2008). Disintegration of solid foods in human stomach. Journal of food science, 73(5), R67-R80.

Kuo, C. K., & Ma, P. X. (2001). Ionically crosslinked alginate hydrogels as scaffolds for tissue engineering: part 1. Structure, gelation rate and mechanical properties. Biomaterials, 22(6), 511-521.

Lao, F., & Giusti, M. M. (2016). Quantification of Purple Corn (Zea mays L.) Anthocyanins Using Spectrophotometric and HPLC Approaches: Method Comparison and Correlation. Food Analytical Methods, 9(5), 1367-1380.

Madziva, H., Kailasapathy, K., & Phillips, M. (2005). Alginate-pectin microcapsules as a potential for folic acid delivery in foods. Journal of microencapsulation, 22(4), 343- 351.

Maier, T., Fromm, M., Schieber, A., Kammerer, D. R., & Carle, R. (2009). Process and storage stability of anthocyanins and non-anthocyanin phenolics in pectin and gelatin gels enriched with grape pomace extracts. European Food Research and Technology, 229(6), 949.

McCann, D., Barrett, A., Cooper, A., Crumpler, D., Dalen, L., Grimshaw, K., Stevenson, J. (2007). Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: a randomised, double-blinded, placebo-controlled trial. Lancet, 370, 1560–1567.

Morris, V. J., & Chilvers, G. R. (1984). Cold setting alginate‐ pectin mixed gels. Journal of the Science of Food and Agriculture, 35(12), 1370-1376.

Nguyen, A. L., & Luong, J. H. T. (1986). Diffusion in κ‐ carrageenan gel beads. Biotechnology and bioengineering, 28(8), 1261-1267.

Nguyen, A. T. B., Winckler, P., Loison, P., Wache, Y., & Chambin, O. (2014). Physico- chemical state influences in vitro release profile of curcumin from pectin beads. Colloids and Surfaces B: Biointerfaces, 121, 290-298.

Ouwerx, C., Velings, N., Mestdagh, M. M., & Axelos, M. A. V. (1998). Physico- chemical properties and rheology of alginate gel beads formed with various divalent cations. Polymer Gels and Networks, 6(5), 393-408.

110

Rao, M. A., & Cooley, H. J. (1995). Rates of structure development during gelation and softening of high-methoxyl pectin—sodium alginate—fructose mixtures. Food hydrocolloids, 9(4), 229-235.

Skrede, G., Wrolstad, R. E., Lea, P., & Enersen, G. (1992). Color stability of strawberry and blackcurrant syrups. Journal of Food Science, 57(1), 172-177.

Thom, D., Dea, I. C. M., Morris, E. R., & Powell, D. A. (1982). Interchain associations of alginate and pectins. Progress in Food and Nutrition Science, 6(1-6), 97-108.

U.S. FDA. (2016). Generally recognized as safe (GRAS). Retrieved from http//www.fda.gov/Food/IngredientsPackaging/GenerallyRecognizedasSafeGRAS/def ault.htm.

Walkenström, P., Kidman, S., Hermansson, A. M., Rasmussen, P. B., & Hoegh, L. (2003). Microstructure and rheological behaviour of alginate/pectin mixed gels. Food Hydrocolloids, 17(5), 593-603.

Wang, W. D., & Xu, S. Y. (2007). Degradation kinetics of anthocyanins in blackberry juice and concentrate. Journal of Food Engineering, 82(3), 271-275.

Weiss, B., Williams, J. H., Margen, S., Abrams, B., Caan, B., Citron, L. J. & Schultz, S. (1980). Behavioral responses to artificial food colors. Science, 207(4438), 1487- 1489.

Yang, Z., Han, Y., Gu, Z., Fan, G., & Chen, Z. (2008). Thermal degradation kinetics of

111

aqueous anthocyanins and visual color of purple corn (Zea mays L.) cob. Innovative Food Science & Emerging Technologies, 9(3), 341-347.

Yeo, Y., & Park, K. (2004). Control of encapsulation efficiency and initial burst in polymeric microparticle systems. Archives of pharmacal research, 27(1), 1.

Zhang, Z., Zhang, R., Decker, E. A., & McClements, D. J. (2015). Development of food-grade filled hydrogels for oral delivery of lipophilic active ingredients: pH- triggered release. Food Hydrocolloids, 44, 345-352.

112

Table 7. Effect of curing bath augmentation with ACN on encapsulation efficiency.

TGC: 2.2%. Al-P: 82-18%. Spherical particles.

ACN type ACN Percent Percent ACN Percent concentration encapsulation retention overall in curing bath efficiency during encapsulation (g/ml) during curing storage efficiency (% EEc ) (% ARs) (% EEo ) BB 0 47.6 57.3 27.1 8 73.8 50.3 37.4 13 81.4 47.1 38.5 19 95.4 40.6 39.1 22 116.3 42.3 49.4 PC 0 26.0 39.1 10.2 85 51.7 46.9 24.3 130 56.1 44.2 24.8 250 64.8 46.6 30.3

113

Table 8. Particle weight to solution volume ratio on ACN loss during storage

TGC: 2.2%. Al-P: 82-18%. Spherical particles. PC ACN.

PW/SV Percent Percent ACN Percent overall mg/ml encapsulation retention during encapsulation efficiency during storage efficiency curing (% ARs) (% EEo ) (% EEc ) 0.06 24.1 75.8 8.1 0.116 24.1 46.2 9.5 0.136 24.1 46.0 11.0 0.2 24.1 43.5 15.4 0.27 24.1 42.5 20.3 0.33 24.1 37.0 21.7

114

Table 9. Percent ACN weight change in solution, particle and particle in solution (PIS) samples under light and dark as a function of time.

Aqueous solution Particle Particles suspended in solution k (1/hr) t1/2 (hr) k (1/hr) t1/2 (hr) k (1/hr) t1/2 (hr)

light 0.0120 58 0.0011 630 0.0025 277

dark 0.0018 385 0.0003 2,310 0.0007 990

115

Figure 19. Percent encapsulation efficiency after curing (EEc) (●) and percent ACN retention during storage (ARs) (■) for PC ACN encapsulation in 2.2% TGC 82-18%

Al-P spherical particles as a function of initial total monomeric ACN concentration in solution.

116

Figure 20. Mass of ACN in particle versus mass of ACN in solution at equilibrium for purple corn after curing (●) and after storage (■). 2.2% TGC, 82-18% Al-P, spherical particles, particle weight to solution volume ratio is 0.136.

117

40 2.2% 82-18 sph 2.2% 43-57 sph 30 2.8% 43-57 sph

20 Percent Percent encapsulationefficiency 10

0 PC curing PC storage BB curing BB storage

Figure 21. Encapsulation efficiency during curing and storage of PC and BB ACN using 2.2% TGC 82-18 Al-P spherical particles (red), 2.2% TGC 43-57 Al-P spherical particles (blue) and 2.8% TGC 43-57 Al-P spherical particles (black) with initial PC and BB ACN concentration at 26.9 and 27.7 µg/ml, respectively.

118

Figure 22. Comparison of improvement of EEc during curing for PC and BB encapsulation. Percent EEc after curing as a function of normalized ACN concentration in curing bath by ACN concentration in droplet for purple corn (●) and for blueberry

(■). Initial PC and BB ACN concentrations are 233.9 and 27.7 µg/ml respectively.

119

Figure 23. Overall encapsulation efficiency (EEo) of PC ACN at equilibrium in storage at different particle weight to solution volume ratio. 2.2% TGC 82-18% Al-P spherical particles. Initial PC ACN concentration is 713.5 µg/ml.

120

Figure 24. Overall encapsulation efficiency (EEo) as a function of temperature for PC

ACN in 43-57% Al-P (●) and 82-18% Al-P (■) spherical particles. 2.2% TGC. Particles stored at pH 3.0. Initial PC ACN concentration is 713.5 µg/ml.

121

a) 0 b 0

-0.2 -0.2 -0.4 -0.4

-0.6 change) change) -0.6

-0.8 ln(Percent ln(Percent weight ACN

ln(Percent ln(Percent weight ACN -0.8 -1 0 200 400 600 0 200 400 600 Time (hr) Time (hr)

Figure 25. Stability of PC ACN as a function of exposure time to fluorescent light. a)

Stability for samples exposed to fluorescent light at 20 ⁰C; b) Stability for samples stored in dark at 20 ⁰C. ACN aqueous solution (●), ACN encapsulated in hydrogel particle (♦), ACN encapsulated in hydrogel particle dispersed in solution (■).

122

6 (ug/ml) 4

Concentrationof ACN curingbath in 0 0 5 10 15 20 25 30 35 Time (min)

Figure 26. Gelation study of 2.2% TGC 82-18 Al-P encapsulated BB ACN with disc

(■) and spherical (●) shape and encapsulated PC ACN with disc (□) and spherical (○) shape in pH 1.2 buffer solution over time.

123

Chapter 5. Determination of purple corn and blueberry ACN diffusion coefficients in alginate-pectin hydrogel particles

Abstract

Using a well-mixed and temperature-controlled system, the diffusion behavior of purple corn (PC) and blueberry (BB) anthocyanins (ACNs) from alginate-pectin hydrogel particles were experimentally investigated. The diffusion coefficients for ACN were determined based on the mathematical approach using Fick’s second law. Various experimental conditions were evaluated to assess the behavior of the diffusivity of different initial ACN concentration in particles (20.75, 87.78, 114.96 and 173.23 μg/ml), alginate to pectin ratios (82-18 and 43-57 Al-P), total gum concentrations (2.2% and

Partition coefficient was evaluated for all the studied conditions, and showed strong dependent all the evaluated parameters except the initial ACN concentration in particles.

These results can be used to design the hydrogel particles to achieve different equilibrium state during diffusion, and the significant higher activation energy using alginate-pectin hydrogel particles suggested the potential of using this matrix for processes those undergoes strong temperature fluctuation.

124

5.1 Introduction

Anthocyanins (ACNs) are a group of water-soluble phenolic compounds responsible for color of many fruits, vegetables and flowers, with red, orange, purple and blue colors. The color of ACN changes with pH (Cevallos-Casals & Cisneros-

Zevallos, 2004). There is a growing interest in using anthocyanins as natural food colorant as a promising alternative to synthetic dyes (Giusti & Wrolstad, 20030. The global sales of natural colorants grew to $600 million with 7% annual growth rate at

2011, while the sales of artificial and synthetic dye market growth were less than 4% from 2007 to 2011 (Mintel and Leatherhead Food Research, 2013). ACNs were also reported to have various health benefits including antioxidant (Noda et al., 2000; Prior,

2003), inducing cell mutation (Aoki et al., 2004), anti-carcinogenesis (Hagiwara et al.,

2001) and prevention of obesity and diabetes (Tsuda et al, 2003). ACNs extracts are sensitive by exposure to light (Carlsen & Staelfeldt, 1997; Cevallos-Casals & Cisneros-

Zevallos, 2004), high pH (Cevallos-Casals & Cisneros-Zevallos, 2004; Kirca et al.,

2007; Reyes & Cisneros-Zevallos, 2007) and temperature (Patras et al., 2010), which limit their application in food industry.

ACNs can be protected from the environment by encapsulation. ACNs can be protected, delivered and released at a desired location in the gastrointestinal tract by designing encapsulation formulations. The encapsulation process was reported to affect the stability of ACNs and their antioxidant capacity (Nicoli et al, 1999), therefore selection of encapsulation process is also an important factor. Polysaccharide hydrogels were preferred due to their gentle gelation mechanism to encapsulate fruit extracts which are rich in ACNs (Balanč et al., 2016; Belščak-Cvitanović et al, 2011; Celli et al., 2016; Córdoba et al, 2013; Ćujić et al., 2016; Deladino et al, 2008; Oidtmann et al,

2012). A relatively high percent release of the encapsulated extracts from hydrogels in

125

simulated gastric fluid, simulated intestinal fluid or water were observed. Carqueja extract encapsulated in Ca-alginate-chitosan was completely released in acidic condition (Balanč et al., 2016). Oidtmann et al. (2011) reported more than 80% encapsulated bilberry extract was released within 20 min in the simulated gastric fluid; and chokeberry ACN was released almost 100% after 10min immersed in water or acidic medium (pH 2.2) when encapsulated in Ca-alginate or Ca-alginate-inulin hydrogel particles (Ćujić et al., 2016). The higher rate of release from Ca-alginate hydrogel under acidic condition was attributed to release of calcium ions (Østberg et al., 1994). The chitosan added Ca-alginate formulation did not dissolve in acidic, but dissolution of chitosan in ascorbic acid system limited the application (Belščak-

Cvitanović et al, 2011). Chitosan added hydrogel formulations will be limited to delivery of bioactive compounds absorbed in stomach.

Alginate-pectin hydrogel blends was reported to form pH-responsive hydrogel particles at low pH and maintained their integrity under acidic condition such as empty stomach, but dissolved in intestinal condition where pH is greater than 5.0 (Guo &

Kaletunc, 2016). Alginate and pectin are natural polysaccharides and are considered as generally recognized as safe (GRAS) product (FDA, 2015). They have been largely employed in drug delivery system due to their biocompatibility and biodegradability.

Alginate is extracted from seaweed, containing free carboxylic groups which react with divalent cations to form gels, while high methylated pectin containing abundant hydroxyl groups forms hydrogels with the presence of sugar (Walther et al., 2004). The alginate-pectin hydrogel forms gels when pH below 3.4 (Higuita-Castro et al., 2012;

Toft et al., 1986) and has been used to encapsulate vitamin C, anthocyanins (Higuita-

Castro et al., 2012). The release of the encapsulated materials from gels can be different depending of the environment surrounding the hydrogel and the state of the hydrogel

126

(Zhang et al., 2015). The release of the encapsulated material depends on both the properties of the material and the properties of the hydrogel and its structure. Models that used to describe the release of encapsulated anthocyanins mainly using either

Peppas equation which is simplified from Fick’s second law or using first-order kinetics

(Arifin et al., 2006; Higuchi, 1963; Lin & Metters, 2006; Ritger & Peppas, 1987a;

Ritger & Peppas, 1987b). However, this semi-empirical mathematical model has its limits—as reported by Ritger and Peppas (1987a), this model was limited with only using the first 15% of the fractional release to determine the diffusional parameter. This greatly restricted the amount of values that can be used to determine diffusivity values.

Instead, Fick’s second law has been numerically solved to determine the diffusivity of various solutes previously (Arnaud & Lacroix, 1991; Gabardo et al, 2011; Ha et al.,

2008; Longo et al., 1992; Merchant et al., 1987). The rate of encapsulated glucose diffusion was determined with unsteady-state diffusion which sphere immersed in infinite and finite liquid phase (Merchant et al., 1987). A 2 cm Ca-alginate sphere bead was used to encapsulate glucose with four rotation speeds. Using the solute diffused into and out from the sphere which was calculated by mass balance, diffusivity was determined using analytical solution and computer regression analysis. The diffusivity was increased as bead rotational speed increased, and around 6 X 10-10 m2/s. The Biot number was also evaluated at each rotation speed and at all conditions was larger than

100, which suggested negligible external film formation. The diffusion of lactose in κ- carrageenan/locust bean gum gel beads was evaluated with and without entrapped bacteria (Arnaud & Lacroix, 1991). They reported the effective diffusivity of lactose was not affected by the initial lactose concentration, and the diffusivity was around 5 X

10-10 m2/s. Lactose and ethanol diffusion were also evaluated by Gabardo et al. (2011) in Ca-alginate spheres. Their results suggested the diffusion coefficients were affected

127

by initial lactose and ethanol concentration which was agreed with the work from Ha and co-workers (2008) who studied the diffusion of chlorferon and diethylthiophophate in Ca-alginate beads, but opposite to the conclusion drew by Arnaud & Lacroix (1991).

Gabardo and co-workers (2011) also reported that the diffusion coefficient was affected by alginate concentration in particle and temperature. B. subtilis and S. marcescens proteases diffusion encapsulated in Ca-alginate beads were evaluated and the corresponding diffusivity was depending on the type of the protease: B. subtilis was higher than S. marcescens which was proposed to be due to steric reasons or interactions within the proteases. They also proposed a linear relationship between the diffusion coefficients and the molecular weight—the higher molecular weight, the lower diffusivity.

In this study, alginate-pectin hydrogel particles were reported to maintain constant size during storage in pH 3.0 solution (Guo & Kaletunc, 2016), therefore, in low pH the release of ACN was expected to be diffusion-controlled, following Fick’s second law. Diffusivity of a solute in a hydrogel matrix was estimated by solving based on numerically the mathematical model Fick’s second law (Arnaud & Lacroix, 1991;

Gabardo et al, 2011; Ha et al., 2008; Longo et al., 1992). At pH 5.0 and 7.0, the release of ACN was controlled both by diffusion and dissolution due to the dissolution of the particle matrix (Guo & Kaletunc, 2016). The corresponding mathematical model was built and solved numerically. The diffusivity of blueberry and purple corn anthocyanin encapsulated in alginate-pectin hydrogel particles was estimated as a function of various initial ACN concentration, alginate to pectin ratio, total gum concentration,

ACN botanic source, particle shape, and storage temperature by numerical analysis using release models being diffusion controlled at low pH or diffusion-dissolution controlled release models.

128

5.2 Theory 5.2.1 Mass Transfer inside Particles

Spherical and cylindrical hydrogel particles were considered as porous solid containing ACN as a solute, placed inside a finite volume of solution. A schematic of a spherical and disc particle model built by COMSOL software is shown in Fig. 27.

5.2.1.1 Spherical particles in pH 3.0

The ACN diffused from the spherical particles to the medium was described with a mass transfer by diffusion equation (Eq. (5.1)). De was assumed to be uniform inside the particles and not affected by solution concentration:

휕퐶 1 휕 휕퐶 (5.1) 푠 = 퐷 ( (r 푠)) 휕푡 푒 푟 휕푟 휕푟

3 퐶푠(r,t) is the ACN (solute) concentration in the particles liquid phase (mol/m ), t is diffusion time (s), r is the radial position in particle (m), and De is the effective diffusivity (m2/s). Effective diffusivity also include the contribution due to tortuosity inside the particle. The initial and boundary conditions are defined as:

푡 = 0, 0 > 푟 > 푅, 퐶푠 = 퐶0 (5.2)

휕퐶 푡 > 0, 푟 = 0, 푠| = 0 (5.3) 휕푟 푟=0

휕퐶푙 휕퐶푠 (5.4) t > 0, 푟 = 푅, 푉푙 = −푁퐴퐷푒 | 휕푡 휕푟 푟=푅 where R is the particle radius (m), 퐶0 is the initial ACN concentration in particle

3 3 liquid phase (mol/m ), 퐶푙 is the ACN concentration in solution (mol/m ), 푉푙 is the volume of solution (m3), N is the number of particles, and A is the area of particle (m2).

The solution is initially free from ACN.

5.2.1.2 Disc shaped particles in pH 3.0

The diffusion behavior in disc particles was described as mass transfer within the cylindrical geometry. The partial differential equation describing the change of ACN 129

concentration as a function of time and radial and axial dimensions as follows:

(5.5) 휕퐶 1 휕 휕퐶 휕2퐶 푠 = 퐷 (( (r 푠)) + 푠) 휕푡 푒 푟 휕푟 휕푟 휕푧2 where z is the thickness position in particles.

The system initial condition was the same, while one boundary condition was added:

휕퐶 (5.6) 푡 > 0, 푧 = 0, 푠 = 0 휕푧

휕퐶 (5.7) 푡 > 0, 푧 = 0, 푠 = 0 휕푧

휕퐶 (5.8) t > 0, 푟 = 푅 푎푛푑 푧 = 푍, 푉 푙 푙 휕푡

휕퐶푠 휕퐶푠 = −푁퐷푒(퐴푠𝑖푑푒 | + 2 × 퐴푡표푝 | ) 휕푟 푟=푅 휕푧 푧=푍 where Z is the half thickness of the particle (m), Atop is the top area, and Aside is the side area (m2).

Computational analysis of the diffusion of ACN was performed using COMSOL

Multiphysics 5.2 model. A 2D axisymmetric geometry was used to construct a quarter of the spherical and cylindrical geometry. One particle was modeled as the domain and the boundary conditions were setup as described in Eq. (5.3), (5.4), (5.6), (5.7) and (5.8).

The parameter De was determined by using parametric sweep with finding the minimum

2 objective function of (Clp-Cle) , where Clp is the predicted average concentration of

ACN in particle liquid phase and Cle is the experimental average concentration of ACN in particle liquid phase.

The percent deviation (Merchant et al., 1987) and average absolute deviation (AAD)

(Ditudompo & Takhar, 2015) were calculated for each condition as defined:

130

퐶 퐶 2 ( 푡⁄ ) −( 푡⁄ ) (5.9) 퐶푒푞 퐶푒푞 푁 푒푥푝 푝푟푒푑 ∑푛+1[ 퐶 ] ( 푡⁄ ) 퐶푒푞 Percent deviation= 표푟푒푑 × 100 푁−1

1 퐶 −퐶 (5.10) 퐴퐴퐷 = ∑푛 |( 푙,푒푥푝 푙,푝푟푒푑) | 푛 𝑖=1 퐶 푙,푒푥푝 𝑖

5.2.1.3 ACN release at pH above 5.0

The ACN release at pH above 5.0 was modeled using continuity equation (Eq.

5.11):

푑푚 (5.11) 푚̇ − 푚̇ + 푅푒푎푐푡𝑖표푛 = 𝑖푛 표푢푡 푑푡 where 푚̇ 𝑖푛 is the mass flow rate of ACN entering to the controlled volume (g/s), 푚̇ 표푢푡 is the mass flow rate of ACN leaving the controlled volume (g/s), Reaction is the rate

푑푚 of ACN release due to particle dissolution (g/s), and is the rate of mass change over 푑푡 the controlled volume (g/s). Considering no ACN was added into the system,

푚̇ 𝑖푛equals to zero; assuming the ACN release was mainly due to particle dissolution, then 푚̇ 표푢푡 can be considered as zero.

5.2.2 External mass transfer resistance

The above system was based on one important assumption that the external mass transfer resistance is negligible. To justify the assumption, Biot (Bi) number was calculated as Eq. (5.12). It was suggested that with Biot number greater than 100

(Blanch and Clark, 1996), the influence of external mass transfer resistance is insignificant.

푘 푅 (5.12) 퐵𝑖 = 퐿 퐷푒 where 푘퐿 is the mass transfer coefficient at the boundary between particle and media

(m/s). The determination of 푘퐿 was followed the procedure previously introduced by

Harriot (1962).

131

5.3 Materials and methods

5.3.1 Materials

High guluronic acid content (mannuronic to guluronic acid ratio at 1.7) alginate

(SF 120) and pectin containing high methoxyl content (Pretested® Pectin, rapid set with degree of esterification of 71-75% and slow-set with degree of esterification of 63-67%) were provided by FMC Biopolymer (Philadelphia, PA) and TIC Gum (White Marsh,

MD), respectively. Purple corn (PC) rich in ACN powder (Zea mays L.) was supported byAlicorp S.A.A (Lima, Peru). Blueberry (BB) juice concentrates (Vaccinium sp.) was obtained by SVZ (Othello, WA). Reagents including potassium chloride, hydrochloric acid, citric acid, sodium citrate, sodium phosphate dibasic and monobasic to prepare buffers were purchased from Fisher Scientific (Waltham, MA).

5.3.2 Preparation of ACN stock solution

Both PC and BB ACN stock solutions were prepared in pH 3.0 buffer solution.

Three grams of PC powder was mixed with 13 grams of pH 3.0 buffer to make PC ACN stock solution. The powder was further dissolved by sonication for 10 min and the mixture was filtered by a 0.45 μm opening syringe filter (Fisher Scientific, Waltham,

MA). Three grams of BB concentrates was mixed with 6 grams of pH 3.0 buffer to make BB ACN stock solution. The concentration of the PC and BB ACN stock solution was measured by pH differential method (Giusti & Wrolstad, 2001). Scans of PC and

BB solutions diluted by pH 1.0 and pH 4.5 were recorded from 700 to 400 nm by a spectrophotometer (Cary 5000, Agilent, Santa Clara, CA). The total monomeric ACN concentration of each stock solution was calculated based on cyaniding-3-glucoside at maximum absorbance peak (520 nm) using Equations 5.10 and 5.11:

퐴 = (퐴520 − 퐴700)푝퐻 1.0 − (퐴520 − 퐴700)푝퐻 4.5 (5.10)

132

퐴 × 푀푊 × 퐷퐹 × 1000 (5.11) 퐶표푛푐푒푛푡푟푎푡𝑖표푛 표푓 퐴퐶푁 (푚푔⁄퐿) = 휀 × 퐿

5.3.3 Preparation of hydrogel particles

Alginate-pectin (Al-P) hydrogels with distinct alginate to pectin percentage weight ratios were prepared (82-18 and 43-57), while the total gum concentration (TGC) was

2.2 and 2.8 wt% in final gel solution. ACN solution was added when a uniform Al-P gel solution was achieved after mixing for 30 min. For every 10 g of gel solution, 0.85 g of either purple corn or blueberry ACN solution was added slowly to avoid surface film formation. After all ACN mixed in, the Al-P-ACN solution was degassed for 10 min in a sonicator and placed under 4℃ to allow bubbles to rise. Al-P particles were produced by dripping Al-P-ACN solution through a 0.337 mm diameter (23G) needle using a peristaltic pump with a flow rate at 0.022ml/s. The droplets were collected in a beaker containing 0.1 m HCl/KCl pH 1.2 buffer which is constantly agitated. The ratio between dropped gel solution to curing bath volume was kept at 0.136. Both spherical and disc particles were produced with varying the dropping distance, which is previously determined (Guo & Kaletunc, 2016). 2.2% TGC 82-18 Al-P spherical particles were dropped at drop distance of 5 cm, while the disc particles were dropped at 20 cm. 2.2% and 2.8% TGC 43-57 Al-P spherical particles were obtained by dropping at 1.5 cm and disc particles were produced at 5 cm distance. All particles were allowed to harden for 2 hours at 4℃ before removed from curing bath for diffusion study.

5.2.2.4 Sample analysis

Particles were filtered out from curing bath with vacuum, and the absorbance spectrum of the curing bath was measured using UV-vis-spectrophotometer (Cary 5000,

Agilent, Santa Clara, CA). The total monomeric ACN concentration was calculated by using the calibration curve developed at pH 1.2 based on pH differential method, and

133

the absorbance used to calculate the ACN concentration was measured at 520 nm. The mass of ACN in particles after curing was determined by subtracting the mass of ACN in curing bath from the mass of ACN in droplets, which determined the initial ACN concentration in particle before diffusion study. During diffusion study, the particles were placed in the buffer solution, and the absorbance spectrum of the solution was measured at 520 nm using spectrophotometer, and its concentration was calculated according to previously obtained calibration curve developed at pH 3.0.

5.3.5 Diffusion study

5.3.5.1 Experimental Setup

Experiments to estimate diffusion coefficients were performed using a 2.5 ml buffer solution at pH 3.0, 5.0 or 7.0 in a quartz cuvette with 0.34 g of particles. Particle size prior to the diffusion study was characterized (Guo & Kaletunc, 2016). A microscope (Amoeba Dual Purpose Digital Microscope, Celestron, CA) was used to take photograph of hydrogel particles placed on a microscope slide. The size of the particle was determined using Image J software (National Institutes of Health, Bethesda,

MD). ACN contained particles were added to the 0.1 m pH 3.0 buffer solution, which was initially free from ACN and mixed with a 7 mm long stir bar (Fisher Scientific,

Waltham, MA). The temperature of the buffer solution was controlled by a peltier temperature controlled cuvette holder (Agilent, Santa Clara, CA). Absorbance spectra of the buffer solution containing particles were recorded between 700 and 400 nm as a function of time over a 240 min period. Absorbance values were converted to ACN concentration by using a calibration curve prepared at 520 nm at pH 3.0.

134

5.3.5.2 Estimation of Diffusion Coefficient

A series of experiments with different alginate to pectin ratio (82-18 and 43-57),

TGC (2.2% and 2.8%), particle shape (spherical and disc) and temperature (4, 24 and

37⁰ C) were carried out in pH 3.0 buffer solution to study diffusion characteristics of

PC and BB ACN in Al-P particles. Spherical particles with different initial PC ACN concentrations in particles ranged from 60.2 to 172.0 μg/ml were used to evaluate the influence of ACN concentration on the diffusion behavior. Effects of ACN botanic source using BB and PC ACN were evaluated by using the same initial ACN concentration of approximately 170 μg/ml in particle. Each experiment was carried out with a new batch particles, and all conditions were performed in duplicate.

5.3.6 ACN release from particles above pH 5.0

Al-P hydrogel used in this study is a pH responsive gel and dissolves above pH 5.0

(Guo & Kaletunc, 2016). The release of ACN from particles at pH 5.0 and 7.0 were investigated as a function of time. 0.34g particles were placed in 2.5ml pH 5.0 or 7.0 buffer inside a quartz cuvette, capped and sealed with parafilm. The cuvettes were placed inside a temperature controlled chamber at 4, 24 or 37⁰ C with vigorously stirring throughout the experiment. The absorbance of the solution was monitored from

700 to 400 nm using the spectrophotometer as a function of time. The maximum wavelength for pH 5.0 was at 526 nm and for pH 7.0 was at 554 nm. At pH 5.0, the absorbance was recorded over a period of 60min at 4⁰ C, 30min at 24⁰ C and 24min at

37⁰ C, while at pH 7.0, over a period of 44min at 4⁰ C, 20min at 24⁰ C and at 37⁰ C for 15min. The effect of particle type was evaluated using 2.2% TGC 43-57 Al-P, 2.8%

TGC 43-57 Al-P and 2.2% TGC 82-18 Al-P spherical particles, and the particles were placed in pH 5.0 or 7.0 at 37⁰ C.

Control experiments were carried out at pH 5.0 and 7.0 over 30 min time period at

135

4⁰ C, 20min at 24⁰ C, and 10min at 37⁰ C. PC ACN solution in pH 5.0 and 7.0 buffer was prepared, and the absorbance from 700 to 400 nm was monitored at targeted temperature. The degradation of ACN presented as absorbance was calculated as a function of time, using first order kinetic. The degradation of PC ACN over time kinetics was used to correct the absorbance obtained in the ACN release results.

5.2.3 Data analysis

5.2.3.1 Statistical analysis

The statistical analysis was conducted to assess the variance among the replicates using ANOVA (analysis of variance), and the comparison of diffusion coefficient and partition coefficient with various alginate to pectin ratios, total gum concentrations, particle shapes, ACN sources and temperatures was performed by t test with JMP for

Windows, release 10 (SAS Institute Inc., Cary, N.C.).

5.2.3.2 Calculation of activation energy

The dependence of the diffusion coefficient on temperature was described by the

Arrhenius equation defined as in Eq. (5.12).

퐸 (5.12) 퐷 = 퐷 exp (− 푎 ) 푒 0 푅푇

2 where 퐷0 is the pre-exponential factor (m /s), Ea is activation energy (J/mol), R is universal gas constant (8.314 퐽 ∙ 푚표푙−1 ∙ 퐾−1) and T is absolute temperature (K).

5.4 Results and discussion

5.4.1 Modeling parameters at pH 3.0

Biot numbers were calculated for agitation speed at 400 rpm to determine the extent of external mass transfer resistance for ACN. While the mass transfer

-5 coefficient kLwas around 1.5 X 10 m/s, the calculated Biot number was ranged from

113 to 2178 at different experimental condition. The Biot number at 24ºC was around

200 for all conditions, while the low temperature at 4ºC reached a high Biot number

136

due to lower diffusion coefficients. For all conditions, the Biot number was larger than 100, therefore, the external film resistance can be ignored (Blanch & Clark,

1996) and the experimental condition met the boundary conditions.

The ADD was calculated to evaluate the goodness fit of the model using the determined diffusivity value. ADD was the lowest at 0.002% and highest at 0.18%.

5.4.2 Effect of initial ACN concentration using PC ACN spherical particle

The diffusion characteristics of PC ACN into pH 3.0 buffer solution from particles containing different initial PC ACN concentrations ranging from 60.2 to 172.0 μg/ml are presented in Figure 28. Such figure showed that the ACN concentration of the bulk solution increased rapidly immediately after the addition of particles into the solution.

The equilibrium was approached after 100 min.

The effective diffusion coefficients calculated for PC ACN were ranged from 0.961 to 1.04 X 10-10 m2/s with initial ACN concentration in particles from 60.2 to 172.0

μg/ml. The effective diffusion coefficients showed no significant influence (p < 0.05) with PC ACN concentration initially in particles within the studied range. The insignificance of the initial concentration on diffusion coefficient were also found and reported by Arnard and Lacroix (1991) using κ-Carrageenan/Locust bean gum-lactose

(12.5 to 50.0 g/L) system, Hannoun and Stephanopoulos (1986) using Ca-alginate- glucose (2 to 100 g/L) and ethanol (10 to 80 g/L) systems. Using equation developed by Wilke & Chang (1955), the diffusion coefficient of the ACN in pure water at 24⁰ C was calculated as 2.2 X 10-10 m2/s, which is twice higher than the diffusivity calculated in particles. In particle, the diffusion was slowed due to a longer path length produced by the impermeable segments of alginate-pectin molecules. Therefore the diffusion coefficient in particle was expected to be lower than the diffusion coefficient in pure water.

137

5.3.3 Effect of particle structure

5.3.3.1 Alginate to pectin ratio

At the same TGC amount (2.2%), both 82-18 and 43-57 Al-P ratios were used to produce particles containing PC or BB ACN to evaluate the effect of alginate to pectin ratio on the diffusion behavior. These two ratios were selected due to the possibility of generating both spherical and disc shaped particles as previously reported (Guo &

Kaletunc, 2016). When using PC ACN, the diffusion coefficient of ACN in 82-18 Al-P spherical particles was 1.04 ± 0.11 X 10-10 m2/s while it was 1.17 ± 0.12 X 10-10 m2/s for 43-57 Al-P spherical particles (Table 9). When using BB ACN, the diffusion coefficient of ACN in 82-18 Al-P disc particle was 2.01 ± 0.09 X 10-10 m2/s while it was 2.09 ± 0.09 X 10-10 m2/s for 43-57 Al-P disc particles. Meanwhile, the diffusion coefficient of ACN in 82-18 Al-P spherical particle was 1.53 ± 0.11 X 10-10 m2/s while it was 1.29 ± 0.10 X 10-10 m2/s for 43-57 Al-P spherical particles. Statistical analysis showed that either using PC or BB ACN, the diffusion coefficient of ACN was not significantly (p > 0.05) affected by the alginate to pectin ratio.

5.3.3.2 Effect of total gum concentration

The effect of total gum concentration on the diffusional parameters was evaluated using 2.2 and 2.8% TGC 43-57 Al-P spherical particles. Results showed that the diffusion coefficient was not significantly (p > 0.05) affected by the increase of total gum concentration (Table 10). Similar conclusions were drawn by several other research: lactose encapsulated in 0.8, 1.0 and 1.2% Ca-alginate microcapsules showed similar diffusion coefficients (Chai et al., 2004); similarly, the diffusion coefficients of glucose, fructose, sucrose and lactose in Ca-alginate membrane liquid-core particles were not influenced by the variation of alginate concentration (Dembczynski &

Jankowski, 2000); and Gabardo and co-worker (2011) found the concentration of Ca-

138

alginate at 3, 4 and 6% showed no significant effect on the diffusion coefficient of ethanol.

5.3.2.3 Effect of particle shape

The effect of particle shape on the diffusional characteristics was evaluated using spherical and disc shaped particles using BB ACN. The diffusion coefficient showed significant influence (p < 0.05) by the particle shape: as shown in Table 8, the diffusion coefficient of disc particles was higher than the ones of spherical particle at each ACN source and Al-P ratio. The effect of particle shape on the diffusivity may related to the difference of particle internal structure. The diffusion coefficient was 1.4 times higher with using 2.2% TGC 82-18 Al-P disc particle than the spherical particle, which was

1.6 times higher with using 2.2% TGC 43-57 Al-P disc particles than the spherical particle. The higher difference of the diffusion coefficient between the spherical and disc particles using the 43-57 Al-P particle suggested that the shape effect was more significant than using the 82-18 Al-P particles.

Research has shown that the ACN retained in particles during storage was 5.6 times when using spherical particles compared with disc particles (Guo & Kaletunc, 2017).

This may be a combination result of higher surface area and diffusivity of disc particles than spherical particles. The surface area of the spherical and disc shaped particles using these two formulations was calculated as 22.6 mm2 for 2.2% TGC 82-18 Al-P spherical particles and 25.7 mm2 for disc particles, which was 1.14 times higher than the spherical particles; it was 23.9 mm2 for 2.2% TGC 43-57 Al-P spherical particles and 32.8 mm2 for disc particles which was 1.37 times higher than the spherical particles. The two factors combining together may be the reason of much higher ACN retained in particles during storage.

139

5.4.4 Effect of ACN source

In order to investigate the effect of the structure difference between the PC and BB

ACN, both ACNs were encapsulated in the 2.2% TGC -82-18 and 43-57 Al-P spherical particles and 2.8% TGC 43-57 Al-P spherical particles. As the effective diffusion coefficient described the rate of diffusion of ACN from the particle to the surrounding solution, the results showed that the release rate of ACN from each particle was higher when using BB ACN then PC ACN, however statistically insignificantly (p > 0.05) affected by the ACN sources (Table 2), except when using 82-18 Al-P spherical particles. Possible interactions may affect the diffusion behavior using these particles: one is related to the interaction between the hydroxyl group on B-ring of anthocyanidin and the carboxyl group and hydroxyl group of pectin by hydrogen bond; one may related to the different hydrophobicity of major component in each ACN extracts; it can also related to the steric hindrance caused by the macromolecules.

5.4.5 Effect of temperature on ACNs diffusion

The effect of temperature on the diffusion coefficients of ACN was investigated by using 2.2% TGC 82-18 Al-P spherical particles encapsulated with PC or BB ACN and

43-57 Al-P spherical particles encapsulated with PC ACN at temperature 4, 24 and

37⁰ C. The initial concentrations of PC or BB ACN in particles were approximately similar at 170 μg/ml. The ANOVA test showed that diffusion coefficient values were significantly (p < 0.05) affected by the temperature studied (Table 9): a variation of diffusion coefficient of PC ACN from 0.13 X1010 m2/s to 1.48 X1010 m2/s was observed when using 2.2% TGC 82-18 Al-P spherical particles as the temperature increased from

4 to 37⁰ C; similarly, the diffusion coefficients of BB ACN were varied from 0.29

X1010 m2/s to 2.22 X1010 m2/s when encapsulated in 2.2% TGC 82-18 Al-P spherical particles when temperature increasing from 4 to 37⁰ C. A similar increase from 0.10

140

X1010 m2/s to 1.85 X1010 m2/s for diffusion coefficient of PC ACN was also found when using 2.2% TGC 43-57 Al-P spherical particles. Diffusion coefficient increase with temperature is expected and can be modeled by Arrhenius equation (Fig. 30).

Nguyen and Luong (1986) reported the glucose diffusion using κ -Carrageenan particles showed the diffusion coefficient increased exponentially from 2.24 X1010 m2/s to 3.66 X1010 m2/s when the temperature changed from 25 to 34⁰ C. The diffusion coefficients of lactose in Ca-alginate beads was increased from 5.16 X1010 m2/s to 6.32

X1010 m2/s with temperature increasing from 25 to 30⁰ C (Gabardo et al., 2011).

The diffusion of PC ACN using 2.2% TGC 82-18 Al-P spherical particle as a function of temperature is shown in Figure 4 as a presentation, similar behavior was found for the other two cases. The influence of temperature on the ACN diffusion clearly presented effect on the mass transfer, increasing accordingly the diffusion coefficient.

Diffusion coefficients of ACN with three conditions as a function of temperature were described using Arrhenius equation and the diffusivity of ACN was exponentially

(R2 > 0.94) decreased as temperature decreased for all three cases. The values of the activation energy Ea was 55.54 kJ/mol, 66.65 kJ/mol and 44.99 kJ/mol for 2.2% TGC

82-18 Al-P spherical particles encapsulated with PC ACN, 2.2% TGC 82-18 Al-P spherical particles encapsulated with BB ACN, and 2.2% TGC 43-57 Al-P spherical particles encapsulated with PC ACN, respectively. The change of viscosity of 2.2%

TGC 82-18 Al-P solution respect to temperature change was lower than the 2.2% TGC

43-57 Al-P gel solution (result not shown), which suggested that the structure change with temperature change was more obvious with 2.2% TGC 43-57 Al-P spherical particles than 2.2% TGC 82-18 Al-P spherical particles. Therefore, a lower activation energy when using 82-18 Al-P spherical particles was expected due to less structural

141

change. Activation energies of the diffusion coefficients of lactose and ethanol encapsulated in Ca-alginate beads were reported to be 21.8 and 28.5 kJ/mol (Gabardo et al., 2011); the energy of activation for lactose diffusion in Ca-alginate beads was 20.0 kJ/mol (Oyaas et al., 1995); and the activation energy was found to be 14.7 kJ/mol and

13.8 kJ/mol for glucose and maltose when using polyacrylamide gels (Yankov, 2004).

Compared with the reported activation energies using different gel matrix, the energy of activation of alginate-pectin hydrogel particles was almost twice higher than the other gel matrix, suggested that the alginate-pectin hydrogel was more sensitive to temperature change.

5.4.6 Release of ACN at pH above 5.0

Al-P hydrogel particles dissolved when pH is above 5.0 Guo & Kaletunc (2016).

Therefore, the release of ACN is expected to be due to hydrogel dissolution in addition to diffusion.

As ACN is very sensitive to temperature (Patras et al., 2010), control experiments were carried out to evaluate the absorbance change over time as ACN in pH 5.0 or pH

7.0 buffer solution over time. The change of absorbance at each pH was used to correct the absorbance reading for the release experiments. At pH 5.0 and 7.0, the degradation of PC ACN monitored as the absorbance of the solution followed first order kinetics with R2 > 0.89 at 24⁰ C. Previous research (Cevallos-Casals & Cisneros-Zevallos, 2004) also showed that PC ACN degradation followed a first order kinetics. The degradation rate constants were used to correct the absorbance measured at pH 5.0 and 7.0.

The corrected absorbance was first calculated using the degradation kinetic rate constants obtained from control experiment at each temperature and pH, then the absorbance ratio (A/Ainf) was plotted as a function of time (Fig. 31 and 32). While the absorbance represented the ACN release from the hydrogel particle to surrounding

142

media, the S/S0 represented the dissolution results of the particle surface change over time (Guo & Kaletunc, 2016). As the release of ACN at pH above 5.0 described by continuity equation, the release rate constant was assumed to be due to dissolution. At the same temperature, the ACN release followed zero order kinetics. It was observed that the absorbance change had a higher slope than the particle surface change, suggesting a higher rate constant of the ACN release than the particle dissolution. By fitting to zero order kinetic for the ACN absorbance ratio to time, at all temperatures, at both pH 5.0 and 7.0, the rate constants of the ACN release were determined to be higher than the particle surface dissolution, except for at 37ºC which the rate constants were similar. Therefore, the diffusion term was not negligible, as a result, the ACN release was a combination of ACN released by the particle dissolution as well as the diffusion of ACN from the hydrogel particle to the surrounding solution. The comparison between the release of ACN and particle dissolution was using the particle surface dissolution result due to the diffusion of ACN from the hydrogel particles was related to the particle area.

Comparing the rate constants obtained at the same temperature, the rate constant of the release of ACN was higher at pH 7.0 than pH 5.0, except at 37⁰C which was similar. At 4ºC, the release rate constant was 1.25 times higher at pH 7.0 than pH 5.0, while the dissolution rate constant was 1.23 times higher; and at 24 ºC, the release rate constant was 1.13 times higher at pH 7.0 than pH 5.0, while the dissolution rate constant was 1.09 times higher. The difference of the release rate constant between the two pH levels was higher than the difference of the dissolution rate constant between the two pH levels suggested that the higher rate constant was contributed by both dissolution and diffusion. At higher pH, the gel dissolution was faster due to the deprotonation from the alginate-pectin chains. As the alginate-pectin chain dissociated, the structure was

143

not maintained, therefore the channel for ACN to diffuse were increased, which resulted in a higher diffusion of the ACN. As a result, the release of ACN at pH 7.0 was higher than pH 5.0 at 4 and 24⁰C. The activation energy results (25.95 kJ for pH 5.0 and 24.72 kJ for pH 7.0) suggested a lower sensitivity to temperature change when particles exposed to pH 7.0 than pH 5.0.

5.4.6.1 Effect of temperature

The 2.2% TGC 82-18 Al-P spherical particles were placed into temperature at 4,

24 and 37ºC at pH 5.0 and 7.0 to evaluate the temperature effect during ACN release.

At pH 5.0 and 7.0, the release of ACN in particle as a function of time followed zero- order kinetics (R2 > 0.97) (Fig. 31 and 32). The rate constants of absorbance ratio change at pH 5.0 were determined as 0.0205, 0.0541, and 0.0652 per minute for 4, 24 and 37ºC. The rate constants were also determined for pH 7.0 as 0.0256, 0.0614, and

0.0778 per minute for 4, 24 and 37ºC (Table 12). The rate constants of ACN release at pH 5.0 and 7.0 were increased as temperature increase, and were plotted as a function of temperature (Fig. 33). Using Arrhenius equation, the activation energy of pH 5.0 and

7.0 was calculated as 25.95 kJ and 24.72 kJ (R2 > 0.95). As the release of ACN was a comination of diffusion and dissolution, it is expected as temperature increase, the release rate constant increased, as both diffusion and dissolution were faster at a higher temperature due to gel structure dissociation.

5.4.6.2 Effect of gel composition

The effect of particle type on the release of PC ACN at pH above 5.0 at 37 ºC was investigated using three particles including 2.2% TGC 43-57 Al-P, 2.8% TGC 43-57

Al-P and 2.2% TGC 82-18 Al-P spherical particles (Fig. 34). Results showed that at the same pH, the release rate constant was highest when using 2.2% TGC 43-57 Al-P spherical particles, followed by 2.8% TGC 43-57 Al-P spherical particles and lowest

144

with 2.2% TGC 82-18 Al-P spherical particle (Table 13). The rate constant of the release of ACN of 2.2% TGC 43-57 Al-P spherical particles was almost twice higher than the one of the 2.8% TGC 43-57 Al-P spherical particles, which was almost twice higher than the rate constant of the release of ACN when using 2.2% TGC 82-18 Al-P spherical particles. The higher rate constant of release ACN when using 2.2% TGC 43-

57 Al-P spherical particles can be related to the gel structure dissociation during pH increasing, which was presented by the higher rate constant of the dissolution of 2.8%

TGC 43-57 Al-P spherical particles compared with 2.2% TGC 82-18 Al-P spherical particles (Table 13).

The rate constants of the release of ACN were higher compared with the rate constants of dissolution of the particles except the rate constants for 2.2% TGC 82-18

Al-P spherical particles were similar, suggesting the release of ACN was a combination of the release from diffusion of ACN from the particles and the release of ACN along with the particle dissolution.

The results showed that the release of ACN at pH above 5.0 significantly affected by the media pH, temperature, total gum concentration and alginate to pectin ratio, and the release was a combination of both diffusion and dissolution behavior. Due to the particle characteristics at high pH, full release of ACN from the particles can be reached, and therefore, alginate-pectin hydrogel can be used as a delivery method to achieve targeted delivery.

5.5 Conclusion

Mass transfer is the core issue in the design of controlled release of bioactive materials in food processing and storage. With the interests of using natural colorants— anthocyanins in food product, the comprehension of diffusivity characteristics and partition coefficient of anthocyanins in alginate-pectin hydrogel particles could help to

145

design the application of encapsulated ACNs in food matrix. In this work, the effective diffusion coefficients of ACNs from alginate-pectin hydrogel particles based on the

Fick’s second law were evaluated. Important variables such as temperature, alginate to pectin ratio, total gum concentration, particle shape and ACN source were tested, with only the temperature and particle shape being of significance for diffusivity. Full release of ACN encapsulated in the alginate-pectin hydrogel particles at pH above 5.0 suggested that the hydrogel formula can be used to achieve targeted and controlled release of ACN by varying the particle type, media temperature and pH.

146

Reference

Arifin, D. Y., Lee, L. Y., & Wang, C. H. (2006). Mathematical modeling and simulation of drug release from microspheres: Implications to drug delivery systems. Advanced drug delivery reviews, 58(12), 1274-1325.

Arnaud, J. P., & Lacroix, C. (1991). Diffusion of lactose in κ −carrageenan/locust bean gum gel beads with or without entrapped growing lactic acid bacteria. Biotechnology and bioengineering, 38(9), 1041-1049.

Balanč, B., Kalušević, A., Drvenica, I., Coelho, M. T., Djordjević, V., Alves, V. D. Sousa, I., Moldão‐ Martins, M., Rakić, V., Nedović, V & Bugarski, B. (2016). Calcium–Alginate–Inulin Microbeads as Carriers for Aqueous Carqueja Extract. Journal of food science, 81(1), E65-E75.

Betz, M., & Kulozik, U. (2011). Whey protein gels for the entrapment of bioactive anthocyanins from bilberry extract. International dairy journal, 21(9), 703-710.

Blanch, H.W., & Clark, D.S. (1996). Biochemical engineering. New York: Marcel Dekker, Inc. p. 137.

Chai, Y., Mei, L. H., Lin, D. Q., & Yao, S. J. (2004). Diffusion coefficients in intrahollow calcium alginate microcapsules. Journal of Chemical & Engineering Data, 49(3), 475-478.

Córdoba, A. L., Deladino, L., & Martino, M. (2013). Effect of starch filler on calcium- alginate hydrogels loaded with yerba mate antioxidants. Carbohydrate polymers, 95(1), 315-323.

Deladino, L., Anbinder, P. S., Navarro, A. S., & Martino, M. N. (2008). Encapsulation of natural antioxidants extracted from Ilex paraguariensis.Carbohydrate Polymers, 71(1), 126-134.

147

Dembczynski, R., & Jankowski, T. (2000). Characterisation of small molecules diffusion in hydrogel-membrane liquid-core capsules. Biochemical engineering journal, 6(1), 41-44.

Ditudompo, S., & Takhar, P. S. (2015). Hybrid mixture theory based modeling of transport mechanisms and expansion‐ thermomechanics of starch during extrusion. AIChE Journal, 61(12), 4517-4532.

Ferreira, D. S., Faria, A. F., Grosso, C. R., & Mercadante, A. Z. (2009). Encapsulation of blackberry anthocyanins by thermal gelation of curdlan.Journal of the brazilian chemical society, 20(10), 1908-1915.

Gabardo, S., Rech, R., & Ayub, A. Z. (2011). Determination of lactose and ethanol diffusion coefficients in calcium alginate gel spheres: predicting values to be used in immobilized bioreactors. Journal of Chemical and Engineering Data, 56, 2305-2309.

Ga̧siorowski, K., Szyba, K., Brokos, B., & Oszmiański, J. (1997). Antimutagenic activity of anthocyanins isolated from Aronia melanocarpa fruits. Cancer Letters, 119(1), 37-46.

Geankoplis, C. (2003). Transport processes and unit operations. Allyn & Bacon, Boston.

Giusti, M. M., & Wrolstad, R. E. (2001). Characterization and measurement of anthocyanins by UV‐ visible spectroscopy. Current protocols in food analytical chemistry. F:F1:F1.2.. Ha, J., Engler, C. R., & Lee, S. J. (2008). Determination of diffusion coefficients and diffusion characteristics for chlorferon and diethylthiophosphate in Ca‐ Alginate gel beads. Biotechnology and bioengineering, 100(4), 698-706.

Hagerman, A. E., & Butler, L. G. (1978). Protein precipitation method for the quantitative determination of tannins. Journal of Agricultural and Food chemistry, 26(4), 809-812.

Hannoun, B. J., & Stephanopoulos, G. (1986). Diffusion coefficients of glucose and ethanol in cell‐ free and cell‐ occupied calcium alginate membranes. Biotechnology and bioengineering, 28(6), 829-835.

Harriott, P. (1962). Mass transfer to particles: Part I. Suspended in agitated tanks. AIChE Journal, 8(1), 93-101.

Higuchi, T. (1963). Mechanism of sustained-action medication. Theoretical analysis of rate of release of solid drugs dispersed in solid matrices. Journal of Pharmaceutical Sciences, 52, 1145-1149.

Li, L., Fang, Y., Vreeker, R., Appelqvist, I., & Mendes, E. (2007). Reexamining the egg-box model in calcium−alginate gels with X-ray diffraction. Biomacromolecules, 8(2), 464-468. 148

Lin, C. C., & Metters, A. T. (2006). Hydrogels in controlled release formulations: network design and mathematical modeling. Advanced drug delivery reviews, 58(12), 1379-1408.

Longo, M. A., Novella, I. S., García, L. A., & Díaz, M. (1992). Diffusion of proteases in calcium alginate beads. Enzyme Microbial Technology, 14, 586-590.

Merchant, F. J. A., Margaritis, A., Wallace, J. B., & Vardanis, A. (1987). A novel technique for measuring solute diffusivities in entrapment matrices used in immobilization. Biotechnology and bioengineering, 30(8), 936-945.

Nguyen, A. & Luong, J. H. T. (1986). Diffusion in ᴋ-carrageenan gel beads. Biotechnology and Bioengineering, 28, 1261-1267.

Noda, Y., Kneyuki, T., Igarashi, K., Mori, A., & Packer, L. (2000). Antioxidant activity of nasunin, an anthocyanin in eggplant peels. Toxicology, 148(2), 119-123.

Oidtmann, J., Schantz, M., Mäder, K., Baum, M., Berg, S., Betz, M., Kulozik, U., Leick, S., Rehage, H., Schwarz, K., & Richling, E. (2012). Preparation and comparative release characteristics of three anthocyanin encapsulation systems. Journal of agricultural and food chemistry, 60(3), 844-851.

Østberg T., Lunda E.M., & Graffnera C. (1994). Calcium alginate matrices for oral multiple unit administration: IV. Release characteristics in different media. International Journal of Pharmaceutics, 112(3), 241–248.

Øyaas, J., Storrø, I., Svendsen, H., & Levine, D. W. (1995). The effective diffusion coefficient and the distribution constant for small molecules in calcium‐ alginate gel beads. Biotechnology and bioengineering, 47(4), 492-500.

Peppas, N. A. (1985). Analysis of Fickian and non-Fickian drug release from polymers. Pharmaceutica Acta Helvetiae, 60, 110–111.

Peterson, J., & Dwyer, J. (1998). Flavonoids: dietary occurrence and biochemical activity. Nutrition Research, 18(12), 1995-2018.

Pothakamury, U. R., & Barbosa-Cánovas, G. V. (1995). Fundamental aspects of controlled release in foods. Trends in Food Science & Technology, 6(12), 397-406.

Prior, R. L. (2003). Fruits and vegetables in the prevention of cellular oxidative damage. The American journal of clinical nutrition, 78(3), 570S-578S.

Ritger, P. L., & Peppas, N. A. (1987a). A simple equation for description of solute release I. Fickian and non-Fickian release from non-swellable devices in the form of slabs, spheres, cylinders or discs. Journal of controlled release, 5(1), 23-36.

Ritger, P. L., & Peppas, N. A. (1987b). A simple equation for description of solute release II. Fickian and anomalous release from swellable devices. Journal of controlled release, 5(1), 37-42.

Santos, D. T., Albarelli, J. Q., Beppu, M. M., & Meireles, M. A. A. (2013). Stabilization of anthocyanin extract from jabuticaba skins by encapsulation using supercritical CO 2 149

as solvent. Food Research International, 50(2), 617-624.

Toft, K., Grasdalen, H., & Smidsrod, O. (1986). Synergistic gelation of alginates and pectins. In ACS Symposium series-American Chemical Society (USA).

U.S. Food and Drug Administration (FDA). (2015). Generally Recognized as Safe (GRAS). Retrieved from http://www.fda.gov/Food/IngredientsPackagingLabeling/GRAS/.

Ventura, I., Jammal, J., & Bianco-Peled, H. (2013). Insights into the nanostructure of low-methoxyl pectin–calcium gels. Carbohydrate polymers, 97(2), 650-658.

Walkenström, P., Kidman, S., Hermansson, A., Rasmussen, P. B., & Hoegh, L. (2003). Microstructure and rheological behavior of alginate/pectin mixed gels. Food Hydrocolloids, 17, 593-603.

Walther, B., Hamberg, L., Walkenström, P., & Hermansson, A. M. (2004). Formation of shaped drops in a fast continuous flow process. Journal of colloid and interface science, 270(1), 195-204.

Yankov, D. (2004). Diffusion of glucose and maltose in polyacrylamide gel. Enzyme and microbial technology, 34(6), 603-610.

Zhang, Z., Zhang, R., Chen, L., Tong, Q., & McClements, D. J. (2015). Designing hydrogel particles for controlled or targeted release of lipophilic bioactive agents in the gastrointestinal tract. European Polymer Journal, 72, 698-716.

150

Table 10. Values of the effective diffusion coefficient De of PC ACN of 2.2% TGC 82-

18 Al-P particles for different concentrations in particles at 24⁰ C.

Initial PC ACN concentration (μg/ml) De X1010 m2/s

20.75 0.96±0.02a

87.78 1.01±0 .09a

114.96 0.99±0.11a

173.23 1.04±0.10a

151

Table 11. Effective diffusion coefficient De of PC and BB ACN of 2.2% TGC 82-18 and 43-57 Al-P particles at 24⁰ C.

ACN source Particle shape TGC (%) A-P ratio De X1010 m2/s

PC ACN spherical 2.2 82-18 1.04±0.10

2.2 43-57 1.17±0.12

2.8 43-57 0.97±0.05

BB ACN disc 2.2 82-18 2.01±0.09

2.2 43-57 2.09±0.09

spherical 2.2 82-18 1.53±0.11

2.2 43-57 1.29±0.10

2.8 43-57 1.17±0.18

152

Table 12. Effective diffusion coefficient De for PC and BB ACN in 2.2% TGC 82-18

Al-P spherical particles and PC ACN in 2.2% TGC 43-57 Al-P spherical particles as a function of temperature.

Temperature (⁰ C)

Particle ACN source 4 24 37

2.2% TGC 82-18 Al-P PC ACN 0.13±0.04 1.04±0.10 1.48±0.04

2.2% TGC 82-18 Al-P BB ACN 0.29±0.01 1.41±0.06 2.22±0.15

2.2% TGC 43-57 Al-P PC ACN 0.10±0.01 1.17±0.12 1.85±0.21

153

Table 13. Kinetic rate constant (k) of release and dissolution study using 2.2% TGC 82-

18 Al-P spherical particles under 4, 24 and 37 ºC at pH 5.0 and 7.0. k (min-1) Release Dissolution (Guo & Kaletunc,

2016)

Temperature pH 5.0 pH 7.0 pH 5.0 pH 7.0

4 ºC 0.021 0.026 0.018 0.022

24 ºC 0.054 0.061 0.051 0.055

37 ºC 0.065 0.078 0.066 0.078

154

Table 14. Rate constants of release of PC ACN at 37 ºC using 2.2% TGC 43-57 Al-P,

2.8% TGC 43-57 Al-P and 2.2% TGC 82-18 Al-P spherical particles at pH 5.0 and

7.0. k (min-1) Release Dissolution (data obtained

by previous research)

Particle type pH 5.0 pH 7.0 pH 5.0 pH 7.0

2.2% TGC 43-57 Al-P 0.348 0.411 NA NA

2.8% TGC 43-57 Al-P 0.169 0.187 0.125 0.127

2.2% TGC 82-18 Al-P 0.065 0.078 0.066 0.078

155

Figure 27. Geometry of spherical and cylindrical geometry of the particles built in

COMSOL using 2D-axisymmetry dimension.

156

1.2

1 60.2 ug/ml 0.8 85.9 ug/ml

inf 112.5 ug/ml

/C 0.6 t

C 172.0 ug/ml

0.4 60.2 ug/ml 85.9 ug/ml 0.2 112.5 ug/ml 172.0 ug/ml 0 0 50 100 150 200 250 time (min)

Figure 28. PC ACN uptake by solution as a function of time in 2.2% TGC 82-18 Al-P spherical particles at 24⁰C using initial ACN concentration in particle at 60.2 (●), 85.9

(), 112.5 () and 172.0 μg/ml (▲). Symbols represent normalized ACN concentration obtained from experiments while the curves represent predicted bulk ACN concentrations from the best fit using COMSOL optimization module with global least square method for the corresponding experimental data.

157

0.8 4C

nf 0.6 4C

/Ci t

C 24C 0.4 24C 0.2 37C 37C 0 0 50 100 150 200 250 time (min)

Figure 29. Diffusion of PC ACN from 2.2% TGC 82-18 Al-P spherical particles into pH 3.0 buffer solution at 4 (○), 24 (◇) and 37⁰ C (□). Symbols represent normalized

158

2.5

2 /s

2 1.5

m 10

1 De X 10 X De

0.5

0 0.0031 0.0033 0.0035 0.0037 1/Temperature (K-1)

Figure 30. Diffusivity as a function of temperature of PC ACN using 2.2% TGC 82-18

Al-P spherical particles (●), BB ACN using 2.2% TGC 82-18 Al-P spherical particles

() and PC ACN using 2.2% TGC 43-57 Al-P spherical particles ().

159

1.2

1 so

/A 0.8

A -

and 1 and 0.6

inf /Abs

t 0.4 Abs

0.2

0 0 10 20 30 40 50 60 Time (min)

Figure 31. PC ACN release in pH 5.0 media at 4, 24 and 37⁰C. Absorbance ratio of PC

ACN at 4⁰C (▲), 24⁰C (♦) and 37⁰C (∎); and dissolution of area ratio of particles at

4⁰C (∆), 24⁰C (◊) and 37⁰C (□) (obtained from previous work (Guo & Kaletunc, 2016)).

160

1.2

so /A

st 0.8

A -

and 1 and 0.6

inf /Abs

t 0.4 Abs 0.2

0 0 10 20 30 40 50 Time (min)

Figure 32. PC ACN release in pH 7.0 media at 4, 24 and 37⁰ C. Absorbance ratio of

PC ACN at 4⁰C (▲), 24⁰C (♦) and 37⁰C (∎); and dissolution of area ratio of particles at 4⁰C (∆), 24⁰C (◊) and 37⁰C (□) (obtained from previous work (Guo & Kaletunc,

2016)).

161

1) -

0.1 k (min k

0.01 0.0031 0.0033 0.0035 0.0037 1/Temperature (K-1)

Figure 33. The PC ACN release rate constants with 2.2% TGC 82-18 Al-P spherical particles as a function of temperature at pH 5.0 (■) and 7.0 (●).

162

a 1.2 b 1.2

1 1 inf

inf 0.8 0.8 /Abs

/Abs 0.6 0.6

t t Abs Abs 0.4 0.4 0.2 0.2 0 0 0 5 10 15 20 0 5 10 15 Time (min) Time (min)

Figure 34. PC ACN release in pH 5.0 and 7.0 media at 37⁰ C. a) Absorbance ratio of

PC ACN at pH 5.0 using 2.2% TGC 43-57 Al-P spherical particle (●), 2.8% TGC 43-

57 Al-P spherical particle (∎) and 2.2% TGC 82-18 Al-P spherical particle (▲); and b)

Absorbance ratio of PC ACN at pH 7.0 using 2.2% TGC 43-57 Al-P spherical particle

(●), 2.8% TGC 43-57 Al-P spherical particle (∎) and 2.2% TGC 82-18 Al-P spherical particle (▲).

163

Bibliography

Anal, A. K., & Singh, H. (2007). Recent advances in microencapsulation of probiotics for industrial applications and targeted delivery. Trends in Food Science & Technology, 18(5), 240-251.

Augst, A. D., Kong, H. J., & Mooney, D. J. (2006). Alginate hydrogels as biomaterials. Macromolecular bioscience, 6(8), 623-633.

Balanč, B., Kalušević, A., Drvenica, I., Coelho, M. T., Djordjević, V., Alves, V. D. Sousa, I., Moldão‐ Martins, M., Rakić, V., Nedović, V & Bugarski, B. (2016). Calcium–Alginate–Inulin Microbeads as Carriers for Aqueous Carqueja Extract. Journal of food science, 81(1), E65-E75.

164

Betz, M., & Kulozik, U. (2011). Whey protein gels for the entrapment of bioactive anthocyanins from bilberry extract. International dairy journal, 21(9), 703-710.

Blanch, H.W., & Clark, D.S. (1996). Biochemical engineering. New York: Marcel Dekker, Inc. p. 137.

Buchweitz, M., Speth, M., Kammerer, D. R., & Carle, R. (2013). Stabilisation of strawberry (Fragaria x ananassa Duch.) anthocyanins by different pectins. Food chemistry, 141(3), 2998-3006.

Cabane, E., Zhang, X., Langowska, K., Palivan, C. G., & Meier, W. (2012). Stimuli- responsive polymers and their applications in nanomedicine. Biointerphases, 7(1), 9.

Carlsen, C., & Stapelfeldt, H. (1997). Light sensitivity of elderberry extract. Quantum yields for photodegradation in aqueous solution. Food chemistry, 60(3), 383-387.

Cemeroğlu, B., Velioglu, S., & Isik, S. (1994). Degradation kinetics of anthocyanins in sour cherry juice and concentrate. Journal of Food Science, 59(6), 1216-1218.

Chai, Y., Mei, L. H., Lin, D. Q., & Yao, S. J. (2004). Diffusion coefficients in intrahollow calcium alginate microcapsules. Journal of Chemical & Engineering Data, 49(3), 475-478.

Christensen, S. H. (1983). Pectins. In: Food Hydrocolloids. pp. 205–229. Glicksman, 165

M., Eds., CRC Press Inc., Boca Raton

Córdoba, A. L., Deladino, L., & Martino, M. (2013). Effect of starch filler on calcium- alginate hydrogels loaded with yerba mate antioxidants. Carbohydrate polymers, 95(1), 315-323.

Ćujić, N., Trifković, K., Bugarski, B., Ibrić, S., Pljevljakušić, D., & Šavikin, K. (2016). Chokeberry (Aronia melanocarpa L.) extract loaded in alginate and alginate/inulin system. Industrial Crops and Products, 86, 120-131. de Pascual-Teresa, S., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (2002). LC–MS analysis of anthocyanins from purple corn cob. Journal of the Science of Food and Agriculture, 82(9), 1003-1006.

Deladino, L., Anbinder, P. S., Navarro, A. S., & Martino, M. N. (2008). Encapsulation of natural antioxidants extracted from Ilex paraguariensis.Carbohydrate Polymers, 71(1), 126-134.

Dembczynski, R., & Jankowski, T. (2000). Characterisation of small molecules diffusion in hydrogel-membrane liquid-core capsules. Biochemical engineering journal, 6(1), 41-44.

Ditudompo, S., & Takhar, P. S. (2015). Hybrid mixture theory based modeling of transport mechanisms and expansion‐ thermomechanics of starch during extrusion. AIChE Journal, 61(12), 4517-4532.

Draget, K. I. (2000). Alginates. In: Handbook of Hydrocolloids. pp. 379–395. Phillips, G. O., and Williams, P. A., Eds., CRC Press, Boca Raton.

Drusch, S. (2007). Sugar beet pectin: A novel emulsifying wall component for microencapsulation of lipophilic food ingredients by spray-drying. Food Hydrocolloids, 21(7), 1223-1228.

Eiselt, P., Yeh, J., Latvala, R. K., Shea, L. D., & Mooney, D. J. (2000). Porous carriers for biomedical applications based on alginate hydrogels. Biomaterials, 21(19), 1921- 1927.

Ferreira, D. S., Faria, A. F., Grosso, C. R., & Mercadante, A. Z. (2009). Encapsulation of blackberry anthocyanins by thermal gelation of curdlan.Journal of the brazilian chemical society, 20(10), 1908-1915.

Francis, F. J. & Markakis, P. C. (1989). Food colorants: Anthocyanins, critical reviews in food science and nutrition. Food Science and Nutrition, 28(4), 273-314.

166

Ga̧siorowski, K., Szyba, K., Brokos, B., & Oszmiański, J. (1997). Antimutagenic activity of anthocyanins isolated from Aronia melanocarpa fruits. Cancer Letters, 119(1), 37-46.

Geankoplis, C. (2003). Transport processes and unit operations. Allyn & Bacon, Boston.

Giusti, M. M., & Wrolstad, R. E. (2001). Characterization and measurement of anthocyanins by UV‐ visible spectroscopy. Current protocols in food analytical chemistry. F:F1:F1.2.. Giusti, M. M., & Wrolstad, R. E. (2003). Acylated anthocyanins from edible sources and their applications in food systems. Biochemical Engineering Journal, 14, 217-225.

Glicksman, M. (1983). Food Hydrocolloids–Volume 1. CRC Press Inc, Boca Raton.

Guo, J., & Kaletunç, G. (2016). Dissolution kinetics of pH responsive alginate-pectin hydrogel particles. Food Research International, 88(A), 129-139.

Ha, J., Engler, C. R., & Lee, S. J. (2008). Determination of diffusion coefficients and diffusion characteristics for chlorferon and diethylthiophosphate in Ca‐ Alginate gel beads. Biotechnology and bioengineering, 100(4), 698-706.

Hagerman, A. E., & Butler, L. G. (1978). Protein precipitation method for the quantitative determination of tannins. Journal of Agricultural and Food chemistry, 26(4), 809-812.

Harriott, P. (1962). Mass transfer to particles: Part I. Suspended in agitated tanks. AIChE Journal, 8(1), 93-101.

He, J., & Giusti, M. M. (2010). Anthocyanins: natural colorants with health-promoting properties. Annual Review of Food Science and Technology, 1, 163–87.

Higuchi, T. (1963). Mechanism of sustained-action medication. Theoretical analysis of rate of release of solid drugs dispersed in solid matrices. Journal of Pharmaceutical Sciences, 52, 1145-1149.

167

Hoefler, A. C. (2004). Hydrocolloids. Eagen Press, Minnesota.

Kaletunç, G. (2007). Prediction of specific heat of cereal flours: A quantitative empirical correlation. Journal of food engineering, 82(4), 589-594.

Kaletunç, G., Normand, M. D., Johnson E.A., & Peleg, M. (1991a). “Degree of elasticity” determination in solid foods. Journal of food science, 56(4), 950-953.

Kaletunç, G., Normand, M. D., Nussinovitch, A., & Peleg, M. (1991b). Determination of elasticity of gels by successive compression–decompression cycles. Food hydrocolloids, 5(3), 237-247.

King, A. H. (1983). Brown Seaweed Extracts (Alginates). In: Food Hydrocollids. pp. 115–182. Glicksman, M., Eds., CRC Press, Boca Raton.

Kırca, A., Özkan, M., & Cemeroğlu, B. (2007). Effects of temperature, solid content and pH on the stability of black carrot anthocyanins. Food chemistry,101(1), 212-218.

Kong, F., & Singh, R. P. (2008). Disintegration of solid foods in human stomach. Journal of food science, 73(5), R67-R80.

Kuang, S. S., Oliveira, J. C., & Crean, A. M. (2010). Microencapsulation as a tool for incorporating bioactive ingredients into food. Critical reviews in food science and nutrition, 50(10), 951-968.

Kuo, C. K., & Ma, P. X. (2001). Ionically crosslinked alginate hydrogels as scaffolds for tissue engineering: part 1. Structure, gelation rate and mechanical properties. Biomaterials, 22(6), 511-521.

Li, L., Fang, Y., Vreeker, R., Appelqvist, I., & Mendes, E. (2007). Reexamining the egg-box model in calcium−alginate gels with X-ray diffraction. Biomacromolecules, 8(2), 464-468.

Lin, C. C., & Metters, A. T. (2006). Hydrogels in controlled release formulations: network design and mathematical modeling. Advanced drug delivery reviews, 58(12), 1379-1408.

Longo, M. A., Novella, I. S., García, L. A., & Díaz, M. (1992). Diffusion of proteases in calcium alginate beads. Enzyme Microbial Technology, 14, 586-590. 168

Madziva, H., Kailasapathy, K., & Phillips, M. (2005). Alginate-pectin microcapsules as a potential for folic acid delivery in foods. Journal of microencapsulation, 22(4), 343- 351.

Montgomery, D. C. (2008). Design and analysis of experiments. John Wiley & Sons.

Morris, V. J., & Chilvers, G. R. (1984). Cold setting alginate‐ pectin mixed gels. Journal of the Science of Food and Agriculture, 35(12), 1370-1376.

Nguyen, A. & Luong, J. H. T. (1986). Diffusion in ᴋ-carrageenan gel beads. Biotechnology and Bioengineering, 28, 1261-1267.

Noda, Y., Kneyuki, T., Igarashi, K., Mori, A., & Packer, L. (2000). Antioxidant activity of nasunin, an anthocyanin in eggplant peels. Toxicology, 148(2), 119-123.

Nussinovitch, A. (1997). Hydrocolloid Applications. Chapman and Hall, London.

Oakenfull, D., & Scott, A. (1984). Hydrophobic interaction in the gelation of high methoxyl pectins. Journal of Food Science, 49(4), 1093-1098.

Ouwerx, C., Velings, N., Mestdagh, M. M., & Axelos, M. A. V. (1998). Physico- chemical properties and rheology of alginate gel beads formed with various divalent cations. Polymer Gels and Networks, 6(5), 393-408.

Øyaas, J., Storrø, I., Svendsen, H., & Levine, D. W. (1995). The effective diffusion 169

coefficient and the distribution constant for small molecules in calcium‐ alginate gel beads. Biotechnology and bioengineering, 47(4), 492-500.

Peppas, N. A. (1985). Analysis of Fickian and non-Fickian drug release from polymers. Pharmaceutica Acta Helvetiae, 60, 110–111.

Peterson, J., & Dwyer, J. (1998). Flavonoids: dietary occurrence and biochemical activity. Nutrition Research, 18(12), 1995-2018.

Pothakamury, U. R., & Barbosa-Cánovas, G. V. (1995). Fundamental aspects of controlled release in foods. Trends in Food Science & Technology, 6(12), 397-406.

Prior, R. L. (2003). Fruits and vegetables in the prevention of cellular oxidative damage. The American journal of clinical nutrition, 78(3), 570S-578S.

Rao, M. A., & Cooley, H. J. (1995). Rates of structure development during gelation and softening of high-methoxyl pectin—sodium alginate—fructose mixtures. Food hydrocolloids, 9(4), 229-235.

R푒́ M. I., Santana, M. H. A. & d’Avila M. A. (2009). Chapter 7: Encapsulation Technologies for Modifying Food Performance. In Passos, M. L., & Ribeiro, C. P. Editor (Eds.). Innovation in food engineering: New techniques and products. CRC Press.

Rees, D. A. (1982). Polysaccharide conformation in solutions and gels–recent results on pectins. Carbohyd Polym., 2:254–263.

Ritger, P. L., & Peppas, N. A. (1987b). A simple equation for description of solute release II. Fickian and anomalous release from swellable devices. Journal of controlled release, 5(1), 37-42.

Santos, D. T., Albarelli, J. Q., Beppu, M. M., & Meireles, M. A. A. (2013). Stabilization of anthocyanin extract from jabuticaba skins by encapsulation using supercritical CO 2 as solvent. Food Research International, 50(2), 617-624.

SAS Institute. (2015). JMP® 12 fitting linear models. SAS Institute Inc.

Skrede, G., Wrolstad, R. E., Lea, P., & Enersen, G. (1992). Color stability of strawberry and blackcurrant syrups. Journal of Food Science, 57(1), 172-177.

Thakur, B. R., Singh, R. K., Handa, A. K., & Rao, M. A. (1997). Chemistry and uses of pectin—a review. Critical Reviews in Food Science & Nutrition, 37(1), 47-73.

Thom, D., Dea, I. C. M., Morris, E. R., & Powell, D. A. (1982). Interchain associations of alginate and pectins. Progress in Food and Nutrition Science, 6(1-6), 97-108.

Toft, K., Grasdalen, H., & Smidsrod, O. (1986). Synergistic gelation of alginates and pectins. In ACS Symposium series-American Chemical Society (USA).

Tsuda, T., Horio, F., Uchida, K., Aoki, H., & Osawa, T. (2003). Dietary cyanidin 3-O- β-D-glucoside-rich purple corn color prevents obesity and ameliorates hyperglycemia in mice. The Journal of nutrition, 133(7), 2125-2130