Kinetika Perubahan Kapasitas Antioksidan dan Mutu Fisik Tempe Selama Pemanasan

KINETIC CHANGES OF ANTIOXIDANT CAPACITY AND

PHYSICAL QUALITY OF TEMPE DURING HEATING

ANNA AMANIA KHUSNAYAINI

GRADUATE SCHOOL

BOGOR AGRICULTURAL UNIVERSITY BOGOR

DECLARATION OF THESIS, SOURCE OF INFORMATION,

AND COPYRIGHT

With this I declare that this thesis Kinetic Changes of Antioxidant Capacity and Physical Quality of Tempe During Heating is my own work under the direction of an advisory committee. It has not yet been presented in any form to any education institution. The sources of information published or not yet published by other researchers have been mentioned and listed in the references.

With this I give the copyright of my thesis to Bogor Agricultural University.

Bogor, March 2016 Anna Amania Khusnayaini NIM F251120231

RINGKASAN

ANNA AMANIA KHUSNAYAINI. Kinetika Perubahan Kapasitas Antioksidan dan Mutu Fisik Tempe Selama Pemanasan. Dibimbing oleh PURWIYATNO HARIYADI dan ELVIRA SYAMSIR.

Tempe merupakan produk pangan tradisional khas Indonesia yang dihasilkan dari proses fermentasi kacang kedelai dengan kapang Rhizopus spp. dan beberapa jenis kapang lainnya, seperti Mucor spp. Jika dibandingkan dengan kacang kedelai, tempe memiliki kandungan gizi yang lebih baik tetapi masa simpannya lebih pendek. Proses termal dapat diaplikasikan sebagai metode alternatif untuk memperpanjang umur simpan tempe. Akan tetapi, proses ini dapat menyebabkan perubahan pada aspek fisik, gizi, dan sensori. Dalam hal ini, model kinetika dapat digunakan untuk memprediksi perubahan mutu pada beberapa kondisi proses untuk mendapatkan produk yang aman dan memaksimalkan mutu.

Tujuan utama penelitian ini adalah untuk mempelajari kinetika perubahan kapasitas antioksidan dan mutu fisik selama pemanasan tempe dalam kondisi isotermal pada suhu 75, 85 dan 95o_{C. Potongan tempe (d: 12 mm, t: 20 mm)}

dimasukkan ke dalam tabung vial dengan larutan garam (NaCl) 2% sebagai medium pemanas. Perubahan kapasitas antioksidan (kapasitas penangkapan radikal DPPH, kandungan total fenol (TPC), dan total flavonoid (TFC)) serta mutu fisik (analisis profil tekstur (TPA), warna dengan kromameter) dianalisis dan dimodelkan sebagai reaksi ordo satu.

Perlakuan panas dapat mempengaruhi kapasitas antioksidan dan mutu fisik tempe secara signifikan. Selama proses pemanasan laju perubahan kedua parameter tersebut meningkat dengan semakin meningkatnya suhu dan lamanya waktu pemanasan. Kapasitas antioksidan berubah dengan kecenderungan yang sama dengan perubahan TPC. Hal ini mengindikasikan bahwa TPC berpengaruh signifikan terhadap kapasitas antioksidan tempe (p < 0.05). Namun demikian, TFC relatif konstan selama pemanasan.

Tidak semua atribut mutu dapat dimodelkan secara baik sebagai parameter kinetika termal. Hal ini disebabkan karena beberapa parameter tersebut tidak dipengaruhi secara signifikan oleh kondisi proses pemanasan, seperti TFC, hue, elastisitas, kelengketan, kepaduan, dan daya kunyah. Parameter mutu tempe yang sensitif terhadap panas, seperti kapasitas antioksidan, TPC, kekerasan, kecerahan, kekuningan, kehijauan, dan kroma, dapat dimodelkan secara baik sebagai reaksi ordo pertama dengan kecenderungan kurva yang berbeda-beda.

Energi aktivasi menunjukkan tingkat sensitivitas parameter mutu terhadap perubahan suhu. Kemiringan kurva untuk degradasi mutu fisik terlihat lebih tajam daripada kapasitas antioksidan. Hal ini berarti bahwa laju perubahan mutu fisik lebih sensitif terhadap panas daripada parameter lainnya. Berdasarkan perhitungan, energi aktivasi perubahan warna (Ea kehijauan=181.44 kJ/mol, Ea kekuningan=82.75

kJ/mol, Ea kroma=73.95 kJ/mol, dan Ea kecerahan=54.98 kJ/mol) lebih besar daripada

perubahan tekstur kekerasan (44.32 kJ/mol). Energi aktivasi untuk perubahan kapasitas antioksidan lebih kecil (14.68 kJ/mol) daripada TPC (30.16 kJ/mol). Kata kunci: kandungan fenol, kapasitas menangkap DPPH, kekerasan, produk

SUMMARY

ANNA AMANIA KHUSNAYAINI. Kinetic Changes of Antioxidant Capacity and Physical Quality of Tempe During Heating. Supervised by PURWIYATNO HARIYADI and ELVIRA SYAMSIR.

Tempe is an Indonesia traditional food made by fermentation of soybean with Rhizopus spp. and some other moulds, such as Mucor spp. As compared to raw soybean, tempe has improved nutritional value, but has shorter shelf-life. Thermal process might be applied as an alternative method to extend the self-life. However, this process induces some changes to the physical, nutritional, and sensory properties. Kinetic models are required to predict quality changes at different processes to provide a safe product and maximize the quality retention.

The main objective of this research was to study the kinetics of antioxidant capacity and physical changes during isothermal heating of tempe at different temperature (75, 85 and 95oC). Tempe was cut (d: 12 mm, h: 20 mm) and placed in vial tubes with 2% of aqueous salt (NaCl) solution as heating medium. Changes of antioxidant capacity (DPPH scavenging capacity, total phenolic content (TPC), total flavonoid content (TFC)) and physical quality (texture profile analysis (TPA), color by chromamater) of tempe during heating were analyzed and modeled as the first order reaction.

Thermal treatments affected significantly on the antioxidant capacity and physical quality of tempe. During heating the rate of changes in antioxidant capacity and physical quality of tempe increased with increase of time and temperature process. The changes of antioxidant capacity had similar trend to TPC indicating that TPC strongly influenced the antioxidant power of tempe. In the other hand, TFC was relatively constant during heating.

Not all tested quality attributes could be modeled well as thermal kinetic parameters because some of them were not affected significantly by process conditions, such as the total flavonoid content, hue, springiness, stickiness, cohesiveness, and chewiness. The heat sensitive quality parameter of tempe, such as antioxidant capacity, TPC, hardness, lightness, yellowness, greenness, and chroma, could be modeled well as first-order reaction with different tendency.

The activation energy indicated the parameter sensitivity to temperature changes. The slope for degradation of physical properties was sharper than antioxidant capacity indicating that the rate of physical change was more heat sensitive than other parameter. Determination of activation energy from physical attributes showed that activation energies for color changes (Ea greenness=181.44

kJ/mol, Ea yellowness=82.75 kJ/mol, Ea chroma=73.95 kJ/mol, Ea lightness=54.98 kJ/mol)

were greater than textural changes (44.32 kJ/mol). The activation energy for antioxidant capacity (14.68 kJ/mol) was smaller than TPC (30.16 kJ/mol).

Keywords: color, DPPH scavenging, fermented soybean, hardness, phenolic content.

© Copyright, IPB 2016

All rights reserved

No part or all of this work may be excerpted without inclusion or mentioning the sources. Exception only for research and education use, writing for scientific papers, reporting, critical writing or reviewing of problems; and this exception does not inflict a financial loss in the proper interest of Bogor Agricultural University.

No part or this entire work may be transmitted and reproduced in any form without written permission from Bogor Agricultural University.

A Thesis

As a Part of The Requirements for Achieving The Master Degree of Science

in

Food Science Program

KINETIC CHANGES OF ANTIOXIDANT CAPACITY AND

PHYSICAL QUALITY OF TEMPE DURING HEATING

GRADUATE SCHOOL

BOGOR AGRICULTURAL UNIVERSITY BOGOR

2016

ACKNOWLEDGEMENTS

In the name of Allah, the Most Gracious and the Most Merciful

Firstly, my humble praise and gratitude to Allah Almighty; the Most Merciful, the Most Beneficent, for the strengths and His blessing for providing me this opportunity and granting me the capability to proceed successfully.

Thanks for everyone who helped me in completing this work. Special appreciation goes to my main supervisor, Prof Dr Purwiyatno Hariyadi, for his excellect supervision, guidance, advice, and financial supports. Not forgotten, my appreciation to my co-supervisor, Dr Elvira Syamsir, for her support and detailed review during the preparation of this thesis. I would like to express my appreciation to the coordinator of Food Science Program, Prof Ratih Dewanti, for her support and help towards my postgraduate affairs. My acknowledgement also goes to all the technicians and office staffs in Department of Food Science and Technology for their co-operations. Special thanks go to all my classmates of Food Science Program for the friendship and memories.

Last but not least, my deepest gratitude goes to my beloved parents; Mr. Mahmud S. and Mrs. Sri Martiyah and my sisters; Sofia Nabila, and Samara Rahma Dania. They were always supporting and encouranging me with their best wishes. Also not forgetting my husband, Eka Bobby Febrianto, and my sons, Adam Zindani Alfayyad and Azzam Zanki Alfaruq, for their love and care. To those who indirectly contributed in this research, your kindness means a lot to me.

Bogor, March 2016 Anna Amania Khusnayaini

CONTENTS

LIST OF TABLES vi

LIST OF FIGURES vi

LIST OF APPENDICES vi

1 INTRODUCTION 1

1.1 Background 1

1.2 Objectives 2

1.3 Hypothesis 2

1.4 Benefits 2

2 LITERATURE REVIEW 3

2.1 Soybean 3

2.2 Tempe 4

2.3 Antioxidant in Tempe 5

2.4 Antioxidant, Phenolics, and Flavonoids in Food 5

2.5 Effects of Heat on Quality Attributes 6

2.6 Kinetics of Quality Degradation 7

3 METHODS 10

3.1 Materials 10

3.2 Tempe Preparation 10

3.3 Sample Preparation 11

3.4 Temperature Measurement 11

3.5 Thermal Treatments 11

3.6 Extraction of Samples 11

3.7 Radical DPPH Scavenging Capacity 12

3.8 Total Phenolic Content (TPC) 12

3.9 Total Flavonoid Content (TFC) 12

3.10 Textural Analysis 13

3.11 Color Analysis 13

3.12 Kinetic Modeling 13

4 RESULTS AND DISCUSSION 15

4.1 Determination of Lag Time 15

4.2 Effects of Heating on Antioxidant Properties 15

4.3 Effects of Heating on Physical Quality 23

4.4 Kinetic Modeling of Quality Changes 25

4.5 Activation Energy of Quality Changes 29

4.6 Application of Kinetic Data 31

5 CONCLUSIONS AND SUGGESTION 35

5.1 Conclusions 35

REFERENCES 36

LIST OF TABLES

1 Chemical composition of soybean and tempe 3

2 Requirements of tempe quality based on Indonesian national standard

(SNI 3144:2009) 4

3 Characteristics of fresh tempe 16

4 Correlations between antioxidant capacity and total

phenolic/flavonoid content 19

5 Activation energy for antioxidant properties and physical quality of

tempe during thermal treatments 31

LIST OF FIGURES

1 a) First order reaction and b) Arrhenius curve 8

2 Diagram of t-T _{relationship for microbial destruction (F) and cooking}

(C) 9

3 The scheme of tempe research 10

4 Thermocouple position in the vial tube 11

5 Position of sample during TPA 13

6 Heating profiles of tempe center at 75, 85, and 95o_{C 15}

7 Effects of heating on DPPH scavenging capacity of tempe at 75, 85

and 95oC 17

8 Effects of heating on total phenolic content of tempe at 75, 85, and

95oC 18

9 Effects of heating on total flavonoid content of tempe at 75, 85 and

95oC 20

10 Force versus time curves of texture profile analysis during thermal

treatments of tempe 21

11 Effects of heating on hardness, springiness, stickiness, cohesiveness,

and chewiness of tempe at 75, 85, and 95o_{C 22}

12 Effects of heating on lightness, greenness, yellowness, chroma, and

hue of tempe at 75, 85, and 95oC 24

13 Kinetic modeling of DPPH scavenging capacity of tempe during

heating at 75, 85, and 95o_{C 25 26}

14 Kinetic modeling of total phenolic content of tempe during heating at

75, 85, and 95oC 27

15 Kinetic modeling of hardness of during heating at 75, 85, and 95o_{C 28}

16 Kinetic modeling of lightness, greenness, yellowness, and chroma

during heating at 75, 85, and 95oC 29

17 Arrhenius plot for antioxidant properties and physical changes of

tempe at 75, 85, and 95oC 30

LIST OF APPENDICES

1 Standard curve to analysis: a) DPPH scavenging capacity, b) total phenolic content, and c) total flavonoid content 43 2 Correlation analysis between antioxidant capacity and total phenolic

content 44

3 Correlation analysis between antioxidant capacity and total flavonoid

content 48

1

INTRODUCTION

1.1Background

Tempe is an Indonesia traditional food made by fermentation of soybean with Rhizopus spp. and some other moulds, such as Mucor spp. Some enzymes produced during fermentation such as proteases, lipases, carbohydrases, and phytases degrade macromolecules into micromolecules (Nout and Rombouts 1990) contributing to the appearance, aroma, texture, flavor, and nutritional quality of tempe (Wiesel et al. 1997). Fungal biomass growth binds the bean cotyledons together tightly, thus tempe resulted has a firm texture and unique appearance.

After soybean fermentation, the hardness and lightness of tempe decreased, but the antioxidative capacity became higher than the raw soy beans (De Reu et al. 1997, Chang et al. 2009, Handoyo and Morita 2006). Some subtances contributed to improve antioxidant power of tempe, such as free amino acids, peptides, isoflavonoid aglycones, 3-hydroxyanthranilic acid (HAA), and gamma-aminobutyric acid (GABA) (Esaki et al. 1996, Watanabe et al. 2007). Sheih et al. (2000) reported that a thirds of antioxidant capacity in tempe were contributed by aglycone isoflavones and the other two-thirds were contributed by peptides liberated during fermentation.

Furthermore, as compared to soybeans tempe had improved bioactive compounds, such as glucosamine, GABA, folate, vitamin B12, vitamin B6, ergosterol, aglycone isoflavones, and some minerals (Wiesel et al. 1997, Eklund-Jonsson et al. 2006, Feng et al. 2007, Koh et al. 2012, Mo et al. 2013). However, the great healthy benefits of tempe were not balanced with the short shelf-life of tempe. For this reason, thermal process might be applied as an alternative methods to extend the shelf-life because it destroyed microorganisms effectively. On the other hand, thermal treatment induced some changes to the physical, nutritional, and sensory properties of food.

Some researchers stated that thermal treatments affected on quality attributes of soybean products. Boiling yellow soybean flours for 120 min could decrease the lightness and increase the chroma intensity (Xu and Chang 2008). Moreover, thermal processing significantly affected on the antioxidant capacity, total phenolic content (TPC), total flavonoid content (TFC), and total isoflavones in soymilk and yellow soybeans (Huang et al. 2006, Xu and Chang 2008b, Xu and Chang 2009).

Otherwise, kinetic models are required to predict quality changes at different processes to provide a safe product and maximize the quality retention. But, a study of quality evaluation of tempe subjected to the thermal processing conditions is very limited. This study was conducted in model systems because it would be esier to explore some changes of tempe during thermal process. Therefore, the changes that occur during thermal processing of tempe can be predicted for optimal product quality.

2

1.2Objectives

The following objectives were formulated for this study:

1. to evaluate of the effects of process variables (time and temperature) on the antioxidant power (antioxidant capacity, total phenolic content, and total flavonoid content) and physical quality (texture and color) of tempe during heating.

2. to develop a kinetic model of antioxidant capacity and physical quality changes of tempe during heating.

1.3Hypothesis

1. Thermal treatments could change antioxidant capacity, total phenolic content, total flavonoid content, textural, and color quality of tempe. 2. The rate of change in antioxidant capacity, total phenolic content, total

flavonoid content, textural, and color quality of tempe could be modeled well by first order kinetic reaction.

1.4Benefits

1. Understanding how combination of time and temperature could affect on antioxidant power (antioxidant capacity, total phenolic content, total flavonoid content) and physical quality (texture, color) of tempe.

2. Determination of process parameters to maintain and optimize quality retention of tempe.

3

2

LITERATURE REVIEW

2.1Soybean

Soybean (Glycine max (L.) Merrill) is a species of legume native to China which is member of the family Leguminosae, subfamily Papilionoideae, and the genus Glycine, L. (Liu 1997). There are two kinds of soybean: yellow soybean coat usually used in Indonesia as raw material of tempe, tofu, bean sprout, and soymilk, and black soybean as main material of soy sauce production (Purwanti 2004). Besides, soy bean is processed into soy protein ingredients including soy flour, concentrates, isolates, and textured soy proteins applied in various types of bakery, dairy, and meat products, infant formulas, and the so-called new generation soyfoods (Liu 1997). Comparation of chemical composition between soybean and tempe is tabulated in Table 1.

Table 1 Chemical composition of soybean and tempe (Kwon et al. 2010).

Component Soybean Tempe

Protein (%) 20-42 23-55

Lipid (%) 18-22 14-23

Carbohydrate (%) 35 10-30

Isoflavone (mg/100 g) Genistin 36-86 6-19

Daidzin 15-57 2-3

Glycitin 2-6 0.1-0.4

Malonyl genistin 123-186 39-42

Genistein 0.2-5 7-10

Daidzein 0.3-5 7-8

Glycitein 0.1-0.6 0.5-0.7

Asam amino (g/100 g) Aspartate 1.4-1.7 1.5

Glutamate 3.8-4.8 2.3

Serine 1.1-1.4 0.8

Glycine 0.7-0.9 0.6

Arginine 1.5-1.9 0.9

Alanine 0.8-1.0 0.6

Proline 1.1-1.3 0.6

Histidine 0.7-0.9 0.4

Valine 0.8-1.0 0.6

Methionine 0.2-0.3 0.2

Cystein 0.5-0.6 0.2

Isoleucine 1.0-1.3 0.7

Leucine 1.6-2.0 1.1

Phenylalanine 1.1-1.4 0.5

Tyrosine 0.8-1.0 0.5

Lysine 1.4-1.7 0.8

Threonine 0.8-1.0 0.5

Mineral (g/100 g) Na 0.02 0.006

4

2.2Tempe

Tempe is a product made from fermented soybean with Rhizopus spp. mould having compact solid form, grey-white color, and specific aroma of tempe (BSN 2009). The main mold used for the preparation of tempe in Indonesia is Rhizopus oligosporus (Feng 2006). More strains identified in tempe, including Rhizopus formosaensis and Rhizopus oryzae (Babu et al. 2009). Various microorganisms, including filamentous fungi, yeasts and bacteria, are found in the traditional tempe. The growth of lactic acid bacteria (LAB), coexisting with R. oligosporus, is performed to improve the safety and could inhibit the growth of pathogenic bacteria (Ashenafi and Busse 1991, Feng et al. 2005). Table 2 shows the requirements of tempe quality regulated by National Standardization Agency of Indonesia (BSN).

Table 2 Requirements of tempe quality based on Indonesian national standard (SNI 3144:2009).

Parameters Unit Requirements

Conditions  Aroma  Color  Taste - - - normal, specific normal normal

Moisture (w/w) % max. 65

Ash (w/w) % max. 1.5

Lipid (w/w) % min. 10

Protein (N × 6.25) (w/w) % min. 16

Crude fiber (w/w) % min. 2.5

Metal poisoning  Cd  Pb  Sn  Hg  As mg/kg mg/kg mg/kg mg/kg mg/kg max. 0.2 max. 0.25 max. 40.0 max. 0.03 max. 0.25 Microorganisms  Coliform  Salmonnella MPN/g - max. 10 negative/25 g

Tempe is made by soaking soybeans in water, removal of the seed coat, boiling the cotelydons, cooling, inoculating them with pure or traditional mixture of starters, and incubating as a solid bed at 25-30o_{C for 24-72 h (Nout and Kiers}

2005). During fungal fermentation the enzymes degrades some substrates leading to an increase of free amino acids, water-soluble nitrogen compounds, free fatty acids, and the development of specific flavor (Jeleń et al. 2013). Anti nutritional factors, such as trypsin inhibitors, tannin, and phytic acids, were also leached out during cooking, dehulling, and fermentation. Besides, tempe processing eliminated almost stachyose, the most flatulent oligosaccharide in soybean (Egounlety and Aworh 2003).

5 In addition to being a source of proteins, tempe has several interesting functional properties from its bioactive compounds. Soy isoflavones are recognized for their benefits to human health, reducing menopause symptoms and lowering the risk of chronic diseases, such as cardiovascular ones, osteoporosis, and breast, prostate and colon cancer (Matsuo et al. 1997, Alekel et al. 2000, Badger et al. 2005, Lee et al. 2005, Ma et al. 2008, Dong and Qin 2011). Tempe also contains gamma amino butyric acid (GABA) which has antihipertensive effects (Aoki et al. 2003, Tsai et al. 2006).

2.3 Antioxidant in Tempe

An antioxidant is any substance that significantly delays or inhibits the oxidation of an oxidizable substrate (Halliwell and Gutteridge 1989, Sies 1993, Halliwell 1995). Antioxidant compounds can scavenge free radicals and increase shelf life by retarding the process of lipid peroxidation, which is one of the major reasons for deterioration of food products during processing and storage (Halliwell 1997). Antioxidants protect the human body from free radicals and ROS effects that can retard the progress of many chronic diseases as well as lipid peroxidation (Gülçin 2012).

Natural antioxidants are found in almost all plants, microorganisms, fungi, and animal tissues (Pokorny 1999). Phenolic compounds are secondary plant metabolites that are thought to be an integral part of both human and animal diets (Gülçin 2006). The predominant of natural antioxidants are phenolic compounds such as tocopherols, flavonoids, phenolic acids, and lignans, as well as carotenoids and ascorbic acid (Ames et al. 1993, Chen et al. 1996, Gülçin 2012). Nitrogeneous compounds such as oligopeptides, proteins, and melanoidins also have antioxidative activity (Chen et al. 1996).

According to Hoppe et al. (1997), tempe is reported to be antioxygenic. Fermentation of tempe causes increased antioxidative capacity (Chang et al. 2009). The antioxidant components of soybean flour have been shown to be isoflavones, phospholipids, tocopherols, amino acids, and peptides (Gyӧrgy et al. 1964, Naim et al. 1974, Naim et al. 1976, Hayes et al. 1977, Murakami et al. 1984). The predominant isoflavones in tempe were genistein (5,7,4’-trihydroxyisoflavone) and daidzein (7,4’-dihydroxyisoflavone), produced by digesting genistin and daidzin with β-glucosidase during fermentation (Murakami et al. 1984). Isoflavones contributed only one-thirds of the antioxidative capacity in tempe and the other two-thirds were contributed by the peptides derived from proteolyzing soy proteins (Sheih et al. 2000). Esaki et al. (1996) found 3-hydroxyanthranilic acid (HAA) as another antioxidative substance formed in tempe.

2.4 Antioxidant, Phenolics, and Flavonoids in Food

The terms of antioxidant activity and antioxidant capacity have different meanings. Activity refers to the rate constant of a reaction between a specific antioxidant and a specific oxidant. Capacity is a measure of the amount (as a mole) of a given free radical scavenged by a sample (MacDonald-Wicks et al.

6

2006). Quantification of antioxidant capacity in the foods can be determined by DPPH (2,2-diphenyl-1-picrylhydrazyl) scavenging capacity assay. This method is based on the reduction of DPPH in alcoholic solution in the presence of a hydrogen-donating antioxidant due to the formation of the non-radical form DPPH-H in the reaction. The antioxidants were able to reduce the stable free radical DPPH to the yellow coloured diphenyl-picrylhydrazine showing a maximum absorbance at 517 nm (Gülçin 2012). When DPPH radicals encounter a proton-donor substrate such as an antioxidant, the radicals would be scavenged and the absorbance is reduced (Blois 1958, Gülçin et al. 2009).

Plant phenolics have been one of the major groups of primary antioxidant compounds acting as reducing agents, hydrogen atom donators, singlet oxygen scavengers, and transition metal ion chelators (Rice-Evans et al. 1996, Karaman et al. 2009, Gülçin 2012). The FCR (Folin-Ciocalteu Reducing) assay has been used to measure a total phenolics in the products (Prior et al. 2005), relying on the transfer of electrons from phenolic compounds to the FCR in alkaline medium forming blue complexes that can be detected spectrophotometrically at 750–765 nm (Singleton and Rossi 1965, Singleton et al. 1999, Cai et al. 2004, Song et al.

2010). FCR or Folin’s phenol reagent is a mixture of phosphomolybdate and

phosphotungstate used for the colorimetric assay of phenolic and polyphenolic antioxidants (Singleton et al. 1999).

Flavonoids are naturally phenolic antioxidants occuring in living cells as glycosides and may break down to their respective aglycone and sugar by enzymes or acid-heat treatments (Shahidi et al.1992, Yao et al. 2004). Antioxidant mechanisms of flavonoids were divided into free radical chain breaking, metal chelating, and singlet oxygen quenching, with the inhibition of enzymatic activity (Clifford and Cuppett 2000). The total flavonoid contents (TFC) was estimated by colorimetric assay involving Al(III)-flavonoid complexes formed in solution (Deng and Van Berkel 1998).

2.5 Effects of Heat on Quality Attributes

Term quality is refered to the composite of the characteristics that differentiate individual units of a product and have significance in determining the consumer acceptability (Kramer and Twigg 1966), such as color, texture, nutritional

value, and sensory attributes. Quality loss involves both subjective factors like taste that cannot be readily quantified, and quantifiable factors such as nutrient degradation (Awuah et al. 2007). Despite destroying microorganisms effectively, thermal processing can induce permanent changes to the nutritional and sensory attributes of foods depending on the process severity.

External apperances, both color and texture, are the most important factors that have to be considered before accepting or rejecting a food product. Some researchers studied the effect of thermal processing on the color and texture of food products: peas (Smout et al. 2003, Garrote et al. 2008), faba beans (Revilla and Vivar-Quintana 2008), carrot (Rastogi et al. 2008, Peng et al. 2014), potatoes (Alvarez and Canet 2002, Moyano et al. 2007, Rattan and Ramaswamy 2014), green asparagus (Lau et al. 2000), red chili puree and paste (Ahmed et al. 2002),

7 soybean (Gandhi and Bourne 1991, Xu and Chang 2008b_{), tempe (Handoyo and}

Morita 2006).

Thermal softening can be due to changes in cell wall matrix polysaccharides (celluloses, hemicelluloses, pectins etc.) which depend on many factors such as pH, types and amounts of various salts present in the plant cell wall. Water uptake by polysaccharides results in reduction of cohesiveness of cell wall matrix thereby resulting in lower adhesion. Other reason for softening is due to loss of turgor pressure which is pressure of cell components against the cell wall and it is due to the water content inside the cell (Lund 1982). The thermal treatments can result in plasmolysis which reduces the turgor pressure and it is responsible for softening of food (Rao and Lund 1986).

Thermal processing influences the nutritional value of some food products by changing the total of antioxidant capacity. Boiled and steamed eclipse black beans exhibited significantly lower antioxidant activities than raw beans in total phenolic content (TPC), DPPH free radical scavenging activity (DPPH), and oxygen radical absorbing capacity (ORAC) (Xu and Chang 2008a). As compared to the raw soybeans, all processing methods caused significant decreases in TPC, DPPH, ORAC, total flavonoid content (TFC), condensed tannin content (CTC), monomeric anthocyanin content (MAC), and ferric reducing antioxidant power (FRAP), and in black soybeans (Xu and Chang 2008b).

In the other hand, thermal processing elevated total antioxidant activity and bioaccessible lycopene content in tomatoes and produced no significant changes in TPC and TFC. The increase in total antioxidant activity of the heat-processed tomatoes was due to the increased amount of lycopene as a major phytochemical in tomatoes and other bound phytochemicals released from the cell matrix (Dewanto et al. 2002). However, high pressure processed tomato and carrot purées had significantly higher antioxidant capacities when compared to thermally treated samples (Patras et al. 2009).

2.6 Kinetics of Quality Degradation

Some of quality degradation reaction in food product can be described well by first order kinetic reaction. This model was characterized by a straight line when the logarithm of the texture property was plotted against heating time (Rizvi and Tong 1997). Generally, thermal softening in fruits and vegetables was suited to the first order rate with the firmness as the primary texture attribute (Rao and Lund 1986, Bourne 1987, Rizvi and Tong 1997). The heat-induced degradation of natural pigments and browning reaction also followed the first order reaction (Villota and Hawkes 2007).

The equations for a degradation reaction for first order are (Van Boekel 2008):

-dC_dt =kC

-ln(_CC

8

where Ct is the concentration at time (t), C0 is the initial concentration, t is time,

and k is rate constant.

(a) (b)

Figure 1 a) First order reaction curve (Villota and Hawkes 2007) and b) Relationship between rate constant and temperature (Berk 2009). All chemical reactions are accelerated when the temperature is elevated (Berk 2009). The temperature dependence of a reaction rate constant can be expressed by the Arrhenius equation (Villota and Hawkes 2007, Berk 2009):

k=koexp - Ea⁄RT

lnk_ko=_RTEa

where ko is the frequency or collision factor, Ea is the activation energy, R is the

gas constant (8.314 J/K mol or 1.987 cal/K.mol), and T is the absolute temperature (K).

The design of optimal thermal processes relies on relevant kinetic data for bacterial inactivation and quality changes (Van Loey et al. 1995). Generally, every 10oC rise in temperature the rate of chemical reaction increases two-fold, while the rate of microbial destruction rises ten-fold or otherwise quality degradation is less temperature sensitive than microbial destruction (Holdsworth 1985). There are differences in thermal behaviour between safety and quality factors during heating. Therefore, optimization is needed to ensure the product safety and maximize retention of quality attributes.

Greenwood et al. (1944) were the first to observe thermal behaviour between safety and quality attributes. They studied the destruction of thiamin in cured pork luncheon meat at three levels, 50%, 20%, and 10%, compared with microbial destruction. The rate of thiamine destruction is doubled with an 18oF increase in temperature as contrasted with a tenfold increase in the rate of destruction of heat-resistant bacteria. They assumed that rapid heating method at the higher processing temperatures were more favorable to thiamine retention. Since then many publications have referred to this technique to optimize processing conditions (Holdsworth 1985).

9 Figure 2 represents the effect of different time–temperature combinations on quality retention and microbial inactivation for thermal process. It can be seen from the graph that the acceptable combinations of time and temperature for thermal process fall within the sterile area. Heating of a product in non-sterile area was unacceptable because the microbiological requirements was not achievable.

Figure 2 Diagram of t-T _{relationship for microbial destruction (F) and cooking}

10

3

METHODS

3.1Materials

Yellow-seeded soybean (Glycine max L.) was obtained from Organization of Tempe and Tofu Producers (KOPTI) Bogor, Indonesia. A commercial powdered tempe starter (Raprima brand, Bandung, Indonesia) was used for tempe making. This was stored at 5o_{C, dry and in the dark. All chemicals used were}

pro-analysis grade purchased from Sigma Chemical Co. (St. Louis, Mo., U. S. A) and Merck KgaA (Darmstadt, Germany). The scheme of research is given in Figure 3.

Figure 3 The scheme of tempe research. 3.2Tempe Preparation

Clean soybean grains (300 g) were soaked with an acetis acid solution pH 3.5 (650 mL) for 22 hours at 27o_{C. The soybean was then dehulled manually, and}

the separated cotyledons were boiled (95oC) in tap water (2 L) for 40 min. After cooling down to room temperature, solid state fermentation was performed by inoculating the mixed culture Raprima (0.03 %) into cooked soybean in perforated polypropylene bags (25 cm × 12 cm), and then incubated for 40 hours at 30oC.

Production of Tempe _{Determination of lag time} Heat Penetration Test during heating Thermal Treatments

Heating tempe at 75, 85, 95oC (isothermal condition) for specific interval of time

depending on the tempe properties

Chemical & Physical Analysis Antioxidant capacity, total penolic content, total flavonoid

content, texture, color

Kinetic Modeling Determination of kinetic

parameters (k and Ea)

Information about kinetic changes of antioxidant capacity and physical quality of tempe during heating as the basic

11 3.3Sample Preparation

Fresh tempe was cut into cylindrical shape with diameter of 12 mm and height of 20 mm. Three tempe cuts (± 6 g) were placed in vial tubes (d: 20 mm, h: 75 mm) considering to minimize lag time. Before closing the tube, 2% of aqueous salt (NaCl) solution (± 20 mL) was added, which was usually used as heating medium in thermal process.

Figure 4 Thermocouple position in the vial tube. 3.4Temperature Measurement

Coper thermocouple wires were introduced into the center of the cylinders such that the thermocouple junction was placed at the geometrical center of the vial where it was a coldest point (Figure 4). Thermocouples were connected to a thermorecorder to record the temperature readings at regular intervals of 1 min for 1 h. A water bath (Gesellschaft für Labortechnik, Germany) set at three different temperatures (75, 85, and 95o_{C) was used. Heating profiles of instrument and} samples were used to evaluate the process lag time.

3.5Thermal Treatments

Samples were heated in water bath (Gesellschaft für Labortechnik, Germany) set at three different temperature (75, 85, and 95oC) after lag time in isothermal condition. Process time for chemical, textural, and color analysis was 120, 30, and 12 min, respectively. Following the heat treatments, samples were immediately cooled in ice water to minimize quality deteration during cooling. For chemical analysis tempe and the salt solution were separated and freeze-dried, whereas for physical analysis only tempe were determined.

3.6Extraction of Samples

The freeze-dried samples were accurately weighed into centrifuge tubes and extracted with 50% acetone and 80% acetone (1:5, w/v). Solvents were selected according to a solvent comparison study by Xu and Chang (2007), which

Thermocouple junction 2% of Salt solution

12

found 50% acetone extracts exhibited the highest TPC values and 80% acetone the highest DPPH scavenging capacity and TFC values for yellow soybean. The sample flour-solvent mixtures were vortexed 16 times during 4 h of extraction according to Ferreira et al. (2011). The tubes were then centrifuged by 5810R Centrifuge (Hamburg, Germany) at 3000 rpm for 10 min, and the extracts were stored at 5oC in the dark for use.

3.7Radical DPPH Scavenging Capacity

The DPPH (2,2-diphenyl-1-picrylhydrazyl) power of samples were evaluated according to the method of Xu and Chang (2007). The sample extract (0.2 mL) was added to 3.8 mL ethanol solution of DPPH radical (0.1 mM). The mixture was vortexed vigorously for 1 min and then kept at room temperature in the dark for 30 min. The absorbance of samples was measured using a UV-Visible Spectrophotometer (U-2900, Hitachi, Tokyo, Japan) at 517 nm against ethanol as a blank. A negative control was mixture of DPPH solution and extraction solvent. The inhibitory percentage of DPPH was calculated according to the following equation:

Inhibition (%) = [1 – (Asample/Acontrol)] × 100

The DPPH free radical scavenging capacity was expressed as millimoles of ascorbic acid equivalent per gram of freeze-dried sample (mmol of AAE/g). The standard calibration curve range was 10 to 1000 μM (R2 = 0.996).

3.8Total Phenolic Content (TPC)

Total soluble phenolics in the extracts were determined using a Folin-Ciocalteu assay as described by Xu and Chang (2007). The sample extract (50 μL), distilled water (3 mL), Folin-Ciocalteu’s reagents (250 μL), and 7% NaCO3 (750μL) were vortexed and allowed to stand for 8 min at room temperature. Thereafter, distilled water (950 μL) was added to the mixture and the absorbance was read using the U-2900 Spectrophotometer at 765 nm after 2 h of incubation against distilled water blank. The TPC was expressed as milligrams of gallic acid equivalents per gram of freeze-dried sample (mg of GAE/g). The TPC was expressed as milligrams of gallic acid equivalents per gram of freeze-dried sample (mg of GAE/g). The standard calibration curve range was 10 to 1000 μM (R2 ₌ 0.998).

3.9 Total Flavonoid Content (TFC)

The TFC was determined using a colorimetric method described by Xu and Chang (2007). The sample extract (0.25 mL), distilled water (1.25 mL), and 5% NaNO2 solution (75μL) were mixed in a test tube and kept for 6 min of

incubation. Then, 150 μL of 10% AlCl3·6H2O solution was added to the mixture.

13 vortexed well. The absorbance was measured at 510 nm using an UV-visible spectrophotometer (U-2900, Hitachi). The TFC was expressed as milligrams of catechin equivalents per gram of freeze-dried sample (mg of CAE/g). The standard calibration curve range was 5 to 500 μM (R2 _{= 0.999).}

3.10 Textural Analysis

The texture measurement was accomplished by using TA-XT2i Texture Analyzer (Stable Micro Systems Ltd., Godalming, Surrey, UK) with cylindrical probe (d: 30 mm, h: 50 mm), by performing texture profile analysis (TPA). Tempe were compressed to 50% of their original height by two compressions. Texture profile was generated by keeping pre-test, test, and post-test speeds of 2, 1, and 2 mm/s respectively. The waiting time between the two-cycles of the TPA test was 5 s. Textural parameters such as hardness, springiness, stickiness, cohesiveness, and chewiness were evaluated by using Texture Expert 1.22 software.

Figure 5 Position of sample during TPA. 3.11 Color Analysis

Analysis of visual color was performed by using CR310 Chromameter (Konica, Minolta, Tokyo, Japan). Hunter’s color parameters (L, a, and b values) for the surface of treated tempe were recorded. The L value indicated lightness (0 to 100), a, the red (+) or green (−) coordinate, and b, the yellow (+) or blue (−) coordinate. (−). The limits for a and b are + or −80. The hue (h) and chroma (C) were calculated according to the following equations (Wrolstad and Smith 2010):

h = tan-1 b a_{⁄ C =}1/2_√(a2_+b2₎

3.12 Kinetic Modeling

In this study the changes of quality attributes during thermal process were described as first-order reaction characterized by logarithmic relationship between concentration of food quality and time that is expressed mathematically as follows (Van Boekel 2008):

Force

14

-ln(_CCt

0)=kt

where Ct is the concentration at time (t), C0 is the initial concentration t is time,

and k is rate constant.

The temperature dependence of a reaction rate constant can be expressed by the Arrhenius equation (Van Boekel 2008):

k=koexp - Ea⁄RT

where ko is the frequency or collision factor, Ea is the activation energy, R is the

15

4

RESULTS AND DISCUSSION

4.1Determination of Lag Time

The heat transfer of tempe took place from water (waterbath heating medium) to salt solution through the vial glass surface. Convection heat transfer was at the outside and inside surface of the vial and conduction was through the vial wall. Then, the heat traveled through the 2% of salt solution to the tempe surface via convection and furthermore, the conductive heat transfer took place from the surface to the center of tempe.

Figure 6 shows the heating profiles recorded at 75, 85, and 95o_{C in the}

cold spot of the vial. Measurement of the lag time was to ensure the isothermal condition of heating. It can be seen from the experimental data gathered that the temperature profiles of tempe center lagged during heating at three different temperature which it was needed heating time of 6 min to achieve the targeted heating temperature of 75oC, whereas at 85 and 95oC the lag time was 7 min.

Figure 6 Heating profiles of tempe center at 75, 85, and 95oC. 4.2Effects of Heating on Antioxidant Properties 4.2.1 Antioxidant Capacity

The DPPH assay measures against the 2,2-diphenyl-1-picryhydrazyl radical which is reduced to the yellow coloured in alcoholic solution in the presence of a hydrogen-donating antioxidant due to the formation of the non-radical form DPPH-H (Gülçin 2012). The fresh tempe used in this study had antioxidant capacity of 2.00 mM of AAE/g (see Table 3). The effect of heating at three different temperatures (75, 85, and 95o_{C) on radical DPPH scavenging}

capacity of fresh tempe is shown in Figure 7. Heating samples increased the total amount of antioxidant capacity in total of tempe-salt solution. The increase of antioxidant power also occured only in salt solution, whereas antioxidant capacity decreased in tempe during heating period of 120 min.

0 20 40 60 80 100

0 10 20 30 40 50 60

T

em

pera

ture

(

oC)

Time (min)

75ºC 85ºC 95ºC

16

A similar pattern had been reported for ultra high temperature (UHT) process of soymilk (Xu and Chang 2009), pasteurization of tea extracts (Manzocco et al. 1998), and steaming of several vegetables such as carrots, spinach, mushrooms, asparagus, broccoli, cabbage, red cabbage, green and red peppers, potatoes and soybeans (Halvorsen et al. 2006), which showed an increase in the antioxidant capacity. The steaming process significantly increased about 75-140% DPPH values of antioxidant capacity in yellow soybeans as compared to the raw soybeans (Xu and Chang 2008).

Table 3 Characteristics of fresh tempea_.

Parameters Unit Value

Antioxidant capacity mM of AAE/g 2.00 ± 0.03 Total phenolic content mg of GAE/g 1.45 ± 0.01 Total flavonoid content mg of CAE/g 0.77 ± 0.04

Hardness kg 1.19 ± 0.04

Springiness - 0.38 ± 0.01

Stickiness - 0.21 ± 0.02

Cohesiveness - 0.22 ± 0.01

Chewiness - 0.14 ± 0.01

Lightness - 64.00 ± 0.05

Greenness - -0.30 ± 0.01

Yellowness - 3.69 ± 0.03

Chroma - 3.70 ± 0.03

Hue - -88.93 ± 0.09

a_{Data are expressed as average ± standard error of mean (SEM).}

According to Sheih et al. (2000) that two-thirds of antioxidant capacity in tempe was contributed by peptides. Tempe containted 17 amino acids (Kwon et al. 2010), which tyrosine, phenylalanine, cysteine, lysine, arginine, aspartate, glutamate, histidine, glycine and proline were suggested playing important role in antioxidant power of soybean (Chen et al. 1996, Fang et al. 2002, Hu et al. 2003, Je et al. 2004, Rajapakse et al. 2005, Riison et al. 1980 Saiga et al. 2003). Soybean protein isolate conducted denaturation at 76.5oC with high water content (Kitabatake et al. 1990). However, antioxidant capacity of protein and amino acids was relatively constant during thermal treatments (Arcan and Yemeniciog˘lu 2007).

Isoflavones in soybean have been considered to be the source of antioxidant. Genistein, daidzein, and their glycosides had a radical DPPH scavenging power, ferric reducing-antioxidant power (FRAP), and suppression of low-density lipoprotein (LDL) oxidation (Lee et al. 2005). Heating soybean product at 70-90oC caused degradation of glucoside isoflavones to aglycone form (Eisen et al. 2003) which the aglycones had stronger antioxidant capacity than the glucosides (Hayes et al. 1977, Pratt and Stafforini 1979). As compared to raw soybean, boiled yellow soybean had significant increases of β-glucoside (daidzin, glycitin, and genistin) and aglycone (daidzein, glycitein, and genistein)

17 isoflavones, but significantly decreased the content of the malonylglucoside forms (malonyldaidzin, malonylglycitin, and malonylgenistin) (Xu and Chang 2008). Therefore, this research indicated that isoflavones from phenolic group obtained from 80% acetone extract were the responsible components affecting the changes of antioxidant capacity of tempe during heating.

On the other hand, the decline of antioxidant compounds in tempe was due to the increase of water soluble antioxidant subtances to 2% of salt solution. Thermal process might decrease firmness and adhesion of cell walls (Van Buren 1979), thus it induced releasing bound phenolic compounds, such as flavonoids, accumulated in the vacuoles (Brecht et al. 2008). Briefly, as longer heating time and higher temperature, the bound antioxidant components were more liberated from cell and leached into salt solution.

Figure 7 Effects of heating on DPPH scavenging capacity of tempe at 75, 85, and 95o_C.

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

DP P H sca v en g in g ca p a ci ty (m M AA E/g ) Time (min) Tempe

75ºC 85ºC 95ºC

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

DPPH sc a v en g in g c a p a city (m M o f AA E/g ) Time (min) Salt Solution

75ºC 85ºC 95ºC

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

DPPH sc a v en g in g c a p a city (m M AA E/g ) Time (min)

Total of Tempe and Salt Solution

18

4.2.2 Total Phenolic Content

The total phenolic contents (TPC) were measured using the Folin-Ciocalteu assay, which was based on the transfer of electrons from phenolic compounds to the Folin-Ciocalteu reagent in alkaline medium forming blue complexes that can be detected spectrophotometrically at 750–765 nm (Singleton et al. 1999, Cai et al. 2004). As can be seen from Table 3, the fresh tempe contained phenolic compounds of 1.45 mg of GAE/g. When it was subjected to the thermal processing, the TPC changed with the same trend as antioxidant capacity (Figure 8).

Figure 8 Effects of heating on total phenolic content of tempe at 75, 85, and 95oC.

The increase of TPC occured in salt solution, whereas TPC of tempe decreased during heating period of 120 min indicating that heat treatments caused water-soluble phenolic compounds leaching into heating medium. Totally heating

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

To ta l p h en o li c co n te n t (m g of G AE/g ) Time (min) Tempe

75ºC 85ºC 95ºC

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

To ta l p h en o li c co n te n t (m g o f G AE/g ) Time (min) Salt Solution

75ºC 85ºC 95ºC

0,00 1,00 2,00 3,00 4,00

0 30 60 90 120

To ta l p h en o li c co n te n t (m g of G AE/g ) Time (min)

Total of Tempe and Salt Solution

19 of samples increased the total amount of TPC in tempe and salt solution. Heating of tempe might cause the degradation of polyphenols and release of bound phenolic compositions from the vacuoles (Brecht et al. 2008). According to correlation analysis (see Table 4), there were significant correlations between TPC and DPPH scavenging capacity (p < 0.05) at 75, 85, and 95o_{C in all components}

of samples. It can be assumed that phenolic compounds strongly influenced the antioxidant capacity of tempe.

The changes of phenolic content in tempe was affected by composition of phenolic acids. Raw and boiled yellow soybean contained free phenolic acids both benzoic type, such as gallic acid, 2,3,4-trihydroxybenzoic acid, vanillic acid and protocatechualdehyde, and cinnamic type, such as chlorogenic, sinapic, and trans-cinnamic acid. Moreover, conjugated phenolic acids were also detected in both raw and boiled yellow soybean, such as benzoic type (gallic, protocatechuic, 2,3,4-rihydroxybenzoic, p-hydroxybenzoic, gentistic, syringic, vanillic acid, protocatechualdehyde and vanillin) and cinnamic type (caffeic, p-coumaric, m-coumaric, o-m-coumaric, sinapic and trans-cinnamic acid) (Xu and Chang 2008).

Thermal processing caused complex variations in phenolic acid profiles of soy products. For instance, boiling soymilk caused significant increases in free gallic, protocatechuic, 2,3,4-trihydroxybenzoic, sinapic acid, and subtotal benzoic acids (Xu and Chang 2009), whereas boiling treatment of yellow soybean significantly increased 2,3,4-trihydroxybenzoic acid (Xu and Chang 2008). In addition, pressure steaming of yellow soybeans caused the increase of benzoic acid about 23% and also the TPC value about 35%. The elevated total phenolic values were correlated to the release of phenolic substances with reactivity toward Folin-Ciocalteu reagent from polymerized structural substances in cell walls upon thermal processing (Xu and Chang 2008).

Table 4 Correlations between antioxidant capacity and total phenolic/ flavonoid content.

Correlation Coefficienta (r)

TPC TFC

75oC 85oC 95oC 75oC 85oC 95oC DPPH

scavenging capacity

Tempe 0.77* 0.89* 0.68* 0.09 0.83* 0.51* Salt solution 0.81* 0.69* 0.89* 0.16 0.32 0.04 Total of tempe

and salt solution 0.81* 0.49* 0.79* 0.10 0.21 0.05

a _{*correlation is significant at the 0.05 level.}

4.2.3 Total Flavonoid Content

The total flavonoid contents (TFC) in tempe was estimated by colorimetric assay involving Al(III)-flavonoid complexes formed in solution (Deng and Van Berkel 1998). The results of spectrophotometric analysis showed that before heating the flavonoid compounds of tempe was 0.77 mg of CAE/g (see Table 3). According to Figure 9, the TFC was relatively constant during thermal treatments at three different temperatures (75, 85, and 95oC). Similar heat effects were found

20

in heating of tomatoes at 88o_{C for 2, 15, and 30 min, which resulted no significant}

changes in the total flavonoid content (Dewanto et al. 2002). The results of correlation analysis (see Table 4) indicated that no significant correlations existed between total flavonoid content and DPPH scavenging capacity (p < 0.05) except tempe at 85 and 95o_C.

Figure 9 Effects of heating on total flavonoid content of tempe at 75, 85 and 95o_C.

In the other hand, Xu and Chang (2009) have studied that traditional stove cooking increased by 20-23% TFC, steam injection cooking increased by 28-65% TFC, direct Ultra High Temperature (UHT) processing increased by 48-90% TFC, and indirect UHT processing increased by 74-113% TFC. The effects of heating at different temperature on changes of TFC in this research can not be observed distinctly. The spectrophotometric method used were less sensitive to analyze the changes of TFC during thermal treatments. The flavonoid compounds were poorly occuring in yellow soybean (Xu and Chang 2007), thus the more sensitive

0,00 0,30 0,60 0,90 1,20 1,50

0 30 60 90 120

To ta l fla v o n o id c o n te n t (m g o f C A E/g ) Time (min) Tempe

75ºC 85ºC 95ºC

0,00 0,30 0,60 0,90 1,20 1,50

0 30 60 90 120

To ta l fla v o n o id c o n te n t (m g o f C A E/g ) Time (min) Salt Solution

75ºC 85ºC 95ºC

0,00 0,30 0,60 0,90 1,20 1,50

0 30 60 90 120

To ta l fla v o n o id c o n te n t (m g of C A E/g ) Time (min)

Total of Tempe and Salt Solution

21 methods were required to study the trend of changes in total flavonoid components, such as (High Performance Liquid Chromatoraphy) HPLC.

The predominant flavonoid compounds in soybean are isoflavones which its retention and distribution were significantly affected by thermal processing (Xu and Chang 2008, Xu and Chang 2009). Although the TFC in this study was relatively constant, the antioxidant capacity from 80% acetone extract elevated during heating at 75, 85, and 95oC (see Figure 7). These findings indicated that thermal treatments transformed some compounds which had lower antioxidant status to higher antioxidant power (Xu and Chang 2009). The aglycone forms of isoflavones had stronger antioxidant capacity than the glucosides (Hayes et al. 1977, Pratt and Stafforini 1979).

75oC 85oC

95oC

Figure 10 Force versus time curves of texture profile analysis during thermal treatments of tempe.

22

Figure 11 Effects of heating on hardness, springiness, stickiness, cohesiveness, and chewiness of tempe at 75, 85, and 95oC.

0,00 0,20 0,40 0,60 0,80 1,00 1,20

0 10 20 30

H a rd n ess (k g ) Time (min) Hardness

75ºC 85ºC 95ºC

0,00 0,20 0,40 0,60 0,80 1,00 1,20

0 10 20 30

S p rin g in ess Time (min) Springiness

75ºC 85ºC 95ºC

0,00 0,20 0,40 0,60 0,80 1,00 1,20

0 10 20 30

S tick in ess Time (min) Stickiness

75ºC 85ºC 95ºC

0,00 0,20 0,40 0,60 0,80 1,00 1,20

0 10 20 30

Co h esi v en ess Time (min) Cohesiveness

75ºC 85ºC 95ºC

0,00 0,20 0,40 0,60 0,80 1,00 1,20

0 10 20 30

Ch ew in ess Time (min) Chewiness

23 4.3Effects of Heating on Physical Quality

4.3.1 Textural Profiles

Texture profile analysis (TPA) basically simulates the mastication process of humans to give force versus time curves as a function of deformation. Before heating the fresh tempe had textural characteristics as followed: 1.19 kg of hardness, 0.38 of springiness, 0.21 of stickiness, 0.22 of cohesiveness, and 0.14 of chewiness. Thermal processing effectively impacted on hardness, whereas other parameters, such as springiness, stickiness, cohesiveness, and chewiness, were remainly constant when subjected to heating for 30 min (Figure 11).

The main texture attribute of vegetables is hardness or firmness (Bourne 1989) because it has the best correlation to the consumer acceptance (Bourne 1982). According to force-time curves (see Figure 10), hardness is the maximum force needed to compress the sample. Most published studies have described thermal degradation of texture based on hardness characteristic and it is possible to model changes based on other properties, but their relation to consumer perception has not been well established (Rizvi and Tong 1997).

The firmness of various bean was strongly correlated with soluble pectin content increasing due to thermal treatment (Huang and Bourne 1983). Intercellular adhesive material, such as pectin, held the firm texture of cell wall structure. When broken by heat, the pectin was depolymerized, so the resistance of plant tissue to fracturability loss decreased during heating (Loh and Breene 1982). In addition, thermal treatments resulted in plasmolysis which reduces the turgor pressure of cell wall and caused softening texture (Rao and Lund 1986). The presence of salt (NaCl) in the medium accelerated the solubility of pectin (Van Buren 1986). Hence, the more salt dissolved in solution, the texture of tempe became softer.

4.3.2 Visual Appearance

Appearance of the food product is the first thing influencing on buying ability of consumers. This is characterized by the color which Hunter L a b model is often used to analyze the color changes. Another color system recommended by CIE is the L C h system (Wrolstad and Smith 2010). Figure 12 shows typical results on the changes of color parameters of tempe surface for 12 min at different temperatures. Visual inspection could not differ color changes of samples, but the instrumental analysis using Minolta Chromameter was able to detect the color differences from the surface of samples. The color parameters of fresh tempe were L = 64.00 (lightness), a = -0.30 (greenness), b = 3.69 (yellowness), c = 3.70 (chroma), and H = -88.93 (hue) (see Table 3). During thermal processing, the lightness, yellowness, chroma, and hue value significantly increased, but the greenness tended to decline slightly.

24

Figure 12 Effects of heating on lightness, greenness, yellowness, chroma, and hue of tempe at 75, 85, and 95oC.

64,00 64,50 65,00 65,50 66,00

0 2 4 6 8 10 12

Lig h tn ess (L) Time (min) Lightness

75ºC 85ºC 95ºC

-2,00 -1,50 -1,00 -0,50 0,00

0 2 4 6 8 10 12

G re en n ess (a ) Time (min) Greenness

75ºC 85ºC 95ºC

3,00 3,50 4,00 4,50 5,00

0 2 4 6 8 10 12

Ye ll o w n ess (b ) Time (min)

75ºC 85ºC 95ºC

3,00 3,50 4,00 4,50 5,00

0 2 4 6 8 10 12

Ch ro m a (C) Time (min)

75ºC 85ºC 95ºC

-83,00 -82,50 -82,00 -81,50 -81,00

0 2 4 6 8 10 12

H u e (h ) Time (min)

75ºC 85ºC 95ºC

-83,00 -82,50 -82,00 -81,50 -81,00

0 2 4 6 8 10 12

H u e (h ) Time (min)

25 According to Clydesdale and Ahmed (1978), object-light interactions may affect on color measurement of samples, such as reflection from the surface, refraction into the object, transmission through the object, diffusion, and absorption within the object. Heating tempe in salt solution caused the increase of water content entering into the tissue. When the source light from chromameter came on the tempe surface, it would be resulted increasing intensity of object-light interactions. In this way, increasing water content of tempe produced the more reflected light from the tempe surface. This might increase the lightness of tempe and change the other color propeties.

The change of chroma and hue parameters corresponded to the intensity of yellowness and greenness. The elevation of hue angle was influenced by the increase of yellowness and the decrease of greenness intensity. The chroma had almost similar value to yellowness. Hence, the reduction of greenness value did not significantly affected on the chroma of tempe.

A food system which contain a carbonyl group of reducing sugar and an amine group of free amino acids subjected to high temperature treatment should experience the Maillard reaction involving the formation of brown pigmen (Kim and Lee 2009). It is well known that tempe contains high protein (23-55%) and also some sugars (Kwon et al. 2010). When tempe was heated for certain time, the carbonyl group and amine group might interact to form brown color of Maillard reaction. But, the presence of salt in heating medium could decrease the rate of Maillard Reaction due to the decrease of water activity value (BeMiller and Huber 2008). Therefore, it can be assumed that the use of 2% salt solution as heating medium might be effectively retain the visual appearance of tempe.

4.4Kinetic Modeling of Quality Changes

Kinetic modeling is one of the quantitative modeling used as indicators of food quality degradation. A systematic way is done to stimulate an appropriate design process, thus the quality retention of product can be managed well (Van Boekel 2008). According to Hindra and Baik (2006), the rate of quality change strongly relates with to two main factors, product composition (moisture content, water activity, pH, etc.) and environmental factors (time, temperature, etc.). Furthermore, kinetic order reaction is determined by the characteristics of chemical reaction in the product. Most heat induced reactions are assumed to obey first order kinetic which the reaction rate is proportional to the reactant concentration or quality properties (Holdsworth and Simpson 2007).

Determination of quality indicators induced by heating was based on a logaritmic relationship describing the quality attributes as a function of variable process (time and temperature) obeying first order kinetic reaction. However, in this study not all tested properties could be modeled well as thermal kinetic parameters because some of them were not affected significantly by process conditions, such as the total flavonoid content, hue, springiness, stickiness, cohesiveness, and chewiness. The heat sensitive quality parameters of tempe, such as antioxidant capacity, TPC, hardness, lightness, yellowness, greenness, and chroma, could be modeled well as first-order reaction with different tendency.

26

Figure 13 Kinetic modeling of DPPH scavenging capacity of tempe during heating at 75, 85, and 95oC.

The effects of heating at three different temperatures (75, 85, and 95oC) on radical DPPH scavenging capacity of tempe are shown in Figure 13. The results of spectrophotometric analysis indicated an increase in antioxidant capacity of total of tempe-salt solution and only salt solution due to thermal treatments. In contrast, antioxidant capacity of tempe decreased for 120 min during heating period, but the rate of decreasing antioxidant capacity in tempe increased with increase of temperature (k75=0.0050, k85=0.0051, and k95=0.0052 min-1). On the

other hand, the rate of increasing antioxidant capacity in salt solution (k75=0.0145,

k85=0.0150, and k95=0.0174 min-1) was greater than in total of tempe-salt solution

(k75=0.0025, k85=0.0026, and k95=0.0033 min-1). Totally the thermal treatments of

tempe at 75 and 85oC gave almost same effects on total of tempe-salt solution for 120 min. However, heating at both of temperature was significantly different as

-0,8 -0,4 0,0 0,4 0,8 1,2 1,6 2,0

0 30 60 90 120

ln

(Ct/Co

)

Time (min) Tempe

75ºC: y = -0.0050x, r² = 0.791 85ºC: y = -0.0051x, r² = 0.895 95ºC: y = -0.0052x, r² = 0.828

-0,8 -0,4 0,0 0,4 0,8 1,2 1,6 2,0

0 30 60 90 120

ln ( C t/ C o) Time (min) Salt Solution

75ºC: y = 0.0145x, r² = 0.951 85ºC: y = 0.0150x, r² = 0.939 95ºC: y = 0.0174x, r² = 0.885

-0,8 -0,4 0,0 0,4 0,8 1,2 1,6 2,0

0 30 60 90 120

ln ( C t/ C o) Time (min)

Total of Tempe and Salt Solution

75ºC: y = 0.0025x, r² = 0.887 85ºC: y = 0.0026x, r² = 0.849 95ºC: y = 0.0033x, r² = 0.880

27 compared to at 95o_{C. It can be assumed that the rate constant of antioxidant}

capacity significantly changed at temperature up to 85oC for 120 min.

Figure 14 Kinetic modeling of total phenolic content of tempe during heating at 75, 85, and 95oC.

The changes of total phenolic contents during heating expressed as mg of galic acid equivalent/g of sample are shown in Figure 14. When samples were subjected to the thermal processing, the TPC changed with the same trend as antioxidant capacity showing the decrease in tempe and the increase in salt solution and total of tempe-salt solution. The rate of decreasing TPC in tempe increased with increase of temperature (k75=0.0034, k85=0.0036, and k95=0.0041

min-1_{). Moreover, the TPC of salt solution elevated more quickly (k}

75=0.0090, -0,5 0,0 0,5 1,0 1,5

0 30 60 90 120

ln

(Ct/Co

)

Time (min) Tempe

75ºC: y = -0.0034x, r² = 0.809 85ºC: y = -0.0036x, r² = 0.954 95ºC: y = -0.0041x, r² = 0.888

-0,5 0,0 0,5 1,0 1,5

0 30 60 90 120

ln

(Ct/Co

)

Time (min) Salt Solution

75ºC: y = 0.0090x, r² = 0.908 85ºC: y = 0.0111x, r² = 0.878 95ºC: y = 0.0134x, r² = 0.805

-0,5 0,0 0,5 1,0 1,5

0 30 60 90 120

ln ( C t/ C o ) Time (min)

Total of Tempe and Salt Solution

75ºC: y = 0.0029x, r² = 0.778 85ºC: y = 0.0043x, r² = 0.953 95ºC: y = 0.0051x, r² = 0.910

28

k85=0.0111, and k95=0.0134 min-1) than in total of tempe-salt solution

(k75=0.0029, k85=0.0043, and k95=0.0051 min-1).

Figure 15 Kinetic modeling of hardness of during heating at 75, 85, and 95o_C.

Figure 15 presents hardness parameter (kg) of the samples at three different temperatures. Generally thermal softening in fruits and vegetables was expressed as the first order rate with the firmness as the primary texture attribute (Bourne 1987). The force required to compress tempe decreased during 30 min of thermal treatments. As temperature evelated, the rate of softening tempe increased (k75=0.0075, k85=0.0126, and k95=0.0165 min-1) which at 95oC textural

degradation occured most rapidly. Thermal process affected on turgor cell and changes in wall-pectin composition leading to a decrease in firmness (Revilla and Vivar-Quintana 2008, Van Buren 1986).

Typical results on the changes of color parameters of tempe surface for 12 min at different temperatures are shown in Figure 16. The lightness (L), yellowness (b) and chroma (C) value increased, but the greenness (a) tended to decline during thermal processing. It can be seen from the graphs that the rate of changes in color attribute at 95oC occured most rapidly, followed by heating at 85oCand lastly at 75oC. The increase rate of yellowness (k75=0.0028, k85=0.0076,

and k95=0.0132 min-1) and chroma (k75=0.0033, k85=0.0078, and k95=0.0132 min -1_{) occured more rapidly as compared to lightness (k}

75=0.0005, k85=0.0010, and

k95=0.0014 min-1) for 12 min of heating. On the other hand, the rate of decreasing

in greenness was the greatest (k75=0.0070, k85=0.0126, and k95=0.0207 min-1)

among other color parameters. The chroma influenced by yellowness and greenness had almost similar value to yellowness intensity. Hence, the reduction of greenness value did not significantly affected on the chroma of tempe.

-0,60 -0,40 -0,20 0,00 0,20

0 5 10 15 20 25 30

ln

(H

t/Ho

)

Time (min) Hardness

75ºC: y = -0.0075x, r² = 0.938 85ºC: y = -0.0131x, r² = 0.870 95ºC: y = -0.0172x, r² = 0.921

29

Figure 16 Kinetic modeling of lightness, greenness, yellowness, and chroma during heating at 75, 85, and 95oC.

4.5Activation Energy of Quality Changes

The Arrhenius equation described the effects of typical temperature ranges on tempe attribute changes during thermal process. This model is to regress the natural logarithm of the reaction rate constant, Ink, versus the reciprocal of the absolute temperature, 1/T, to obtain the estimates of Ink0, the intercept and Ea/R,

the slope index of the regression line (Van Loey et al. 1995). The Arrhenius plot of antioxidant activity and physical changes of tempe is given in Figure 17. It can be seen from the curves that the slope for degradation of physical properties was

-0,30 -0,15 0,00 0,15

0 2 4 6 8 10 12

ln

(Lt/Lo

)

Time (min) Lightness

75ºC: y = 0.0005x, r² = 0.947 85ºC: y = 0.0010x, r² = 0.898 95ºC: y = 0.0014x, r² = 0.942

-0,30 -0,15 0,00 0,15

0 2 4 6 8 10 12

ln (a t/a o ) Time (min) Greenness

75ºC: y = -0.0070x, r² = 0.962 85ºC: y = -0.0126x, r² = 0.974 95ºC: y = -0.0207x, r² = 0.997

-0,30 -0,15 0,00 0,15

0 2 4 6 8 10 12

ln (b t/b o ) Time (min) Yellowness

75ºC: y = 0.0028x, r² = 0.952 85ºC: y = 0.0076x, r² = 0.970 95ºC: y = 0.0132x, r² = 0.944

-0,30 -0,15 0,00 0,15

0 2 4 6 8 10 12

ln (C t/Co ) Time (min) Chroma

75ºC: y = 0.0033x, r² = 0.938 85ºC: y = 0.0078x, r² = 0.897 95ºC: y = 0.0132x, r² = 0.939

30

sharper than antioxidant capacity indicating that the rate of physical change was more heat sensitive than other parameter.

Key: AC= antioxidant capacity, TPC= total phenolic content, h= hardness, L= lightness, G= greenness, Y= yellowness, C= chroma, T= tempe, S= salt solution, TS= total of tempe and salt solution.

Figure 17 Arrhenius plot for antioxidant properties and physical changes of tempe at 75, 85, and 95oC.

The activation energy (Ea) (Table 5), which represents the least amount of

energy needed for a chemical reaction to take place, was calculated by the slope of curves (Ea/R). The activation energy also indicated the parameter sensitivity to

temperature changes. Determination of activation energy from physical attributes showed that activation energies for color changes were greater than textural changes meaning that color properties were more heat sensitive than textural attributes. The activation energy for antioxidant capacity was smaller than total phenolic content meaning that TPC was more heat sensitive than antioxidant

-8 -6 -4 -2 0

0,0027 0,0028 0,0029

ln

k

(1

/m

in

)

1/T (1/K)

Antioxidant Properties

AC (T) AC (S) AC (TS) TPC (T) TPC (S) TPC (TS)

-8 -6 -4 -2 0

0,0027 0,0028 0,0029

ln

k

(1

/m

in

)

1/T (1/K)

Physical Quality

H (T) L (T) G (T) Y (T) C (T)

31 capacity of tempe. A positive value of Ea means that the reaction rate increases

with increasing temperature. The information from Arrhenius parameters can be used to optimize thermal process and maximize quality retention of tempe by choosing an appropriate time-temperature combination.

Table 5 Activation energy for antioxidant properties and physical quality of tempe during thermal treatments.

Parameters Component Equations r² Ea

(kJ/mol) Antioxidant

capacity

Tempe y = -251.1x – 4.5769 0.999 2.09 Salt solution y = -1160.3x – 0.9188 0.873 9.65 Total y = -1765.4x – 0.952 0.842 14.68 Total phenolic

content

Tempe y = -1194.1x – 2.2653 0.945 9.93 Salt solution y = -2549.3x + 2.6167 0.999 21.19 Total y = -3627.2x + 4.6135 0.957 30.16 Hardness Tempe y = -5330.3x + 10.467 0.968 44.32 Lightness Tempe y = -6612.4x + 11.453 0.968 54.98 Greenness Tempe y = -21823x + 58.122 0.868 181.44 Yellowness Tempe y = -9952.8x + 22.787 0.978 82.75 Chroma Tempe y = -8894.3x + 19.892 0.985 73.95

4.6Application of Kinetic Data

Design of optimal thermal processing relies on kinetic of bacterial inactivation and quality changes (Van Loey et al. 1995). The destruction of microorganisms and quality factors does not proceed at the same rate (Stumbo 1973, Holdsworth 1985). Commonly, every 10oC rise in temperature the rate of chemical reaction elevates two-fold, while the rate of microbial destruction rises ten-fold or in other words quality deterioration is less heat sensitive than microbial destruction (Holdsworth 1985). Because of the differences in thermal behaviour between safety and quality factors, optimization is required to ensure the product safety and maximize retention of quality attributes.

A conventional thermal process (blanching, pasteurization, and sterilisation) often leads to undesirable changes in food quality, such as color, texture, flavor, nutrition, and functionality (Ndiaye et al. 2009). An accurate kinetic data of microbial inactivation and quality loss can be applied to estimate the effects of time-temperature process on product quality desired. Heating at higher temperatures is relatively lower destruction of quality factors. In addition, when the heat transfer rate is enhanced then the process will be shorter resulting in lesser destruction in quality factors and this forms HTST (High Temperature Short Time) concept (Rattan 2012).

Figure 18 illustrates the application of kinetic parameters of tempe to determine the appropriate time-temperature combination during thermal process. Tempe belongs to low acid food products, thus the thermal process considers both public health (Clostridium botulinum) and spoilage thermophilic organisms

48

Appendix 3 Correlation analysis between antioxidant capacity and total flavonoid

content.

Regression Analysis: Antioxidant versus Flavonoid 75 C TEMPE

The regression equation is

Antioxidant = 0,744 + 0,99 Flavonoid Predictor Coef SE Coef T P Constant 0,7440 0,6301 1,18 0,276 Flavonoid 0,992 1,180 0,84 0,428 S = 0,225800 R-Sq = 9,2% R-Sq(adj) = 0,0% Analysis of Variance

Source DF SS MS F P Regression 1 0,03605 0,03605 0,71 0,428 Residual Error 7 0,35690 0,05099

Total 8 0,39295 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 1 0,630 1,7450 1,3692 0,1400 0,3758 2,12R 5 0,380 1,3350 1,1211 0,1924 0,2139 1,81 X R denotes an observation with a large standardized residual. X denotes an observation whose X value gives it large leverage.

Regression Analysis: Antioxidant versus Flavonoid 75 C SALT

SOLUTION

The regression equation is

Antioxidant = 0,726 + 2,40 Flavonoid Predictor Coef SE Coef T P Constant 0,7259 0,8515 0,85 0,422 Flavonoid 2,404 2,034 1,18 0,276

S = 0,645550 R-Sq = 16,6% R-Sq(adj) = 4,7% Analysis of Variance

Source DF SS MS F P Regression 1 0,5819 0,5819 1,40 0,276 Residual Error 7 2,9171 0,4167

Total 8 3,4991 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 1 0,165 0,505 1,123 0,534 -0,618 -1,70 X 9 0,380 2,975 1,639 0,221 1,336 2,20R R denotes an observation with a large standardized residual.

49

X denotes an observation whose X value gives it large leverage.

Regression Analysis: Antioxidant versus Flavonoid 75 C TOTAL

The regression equation is

Antioxidant = 1,38 + 1,70 Flavonoid Predictor Coef SE Coef T P Constant 1,379 1,836 0,75 0,477 Flavonoid 1,701 1,955 0,87 0,413

S = 0,496390 R-Sq = 9,8% R-Sq(adj) = 0,0% Analysis of Variance

Source DF SS MS F P Regression 1 0,1865 0,1865 0,76 0,413 Residual Error 7 1,7248 0,2464

Total 8 1,9114 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 9 0,91 4,020 2,927 0,173 1,093 2,35R R denotes an observation with a large standardized residual.

Regression Analysis: Antioxidant versus Flavonoid 85 C TEMPE

The regression equation is

Antioxidant = 0,118 + 2,56 Flavonoid Predictor Coef SE Coef T P Constant 0,1177 0,1973 0,60 0,570 Flavonoid 2,5645 0,4434 5,78 0,001

S = 0,117630 R-Sq = 82,7% R-Sq(adj) = 80,2% Analysis of Variance

Source DF SS MS F P Regression 1 0,46293 0,46293 33,46 0,001 Residual Error 7 0,09686 0,01384

Total 8 0,55979

Regression Analysis: Antioxidant versus Flavonoid 85 C SALT

SOLUTION

The regression equation is

Antioxidant = - 1,07 + 10,5 Flavonoid Predictor Coef SE Coef T P Constant -1,072 1,401 -0,77 0,469 Flavonoid 10,503 5,729 1,83 0,109

50

Analysis of Variance

Source DF SS MS F P Regression 1 1,3733 1,3733 3,36 0,109 Residual Error 7 2,8607 0,4087

Total 8 4,2340 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 2 0,275 0,655 1,816 0,286 -1,161 -2,03R R denotes an observation with a large standardized residual.

Regression Analysis: Antioxidant versus Flavonoid 85 C TOTAL

The regression equation is

Antioxidant = 4,79 - 3,07 Flavonoid Predictor Coef SE Coef T P Constant 4,786 1,515 3,16 0,016 Flavonoid -3,075 2,222 -1,38 0,209

S = 0,493290 R-Sq = 21,5% R-Sq(adj) = 10,3% Analysis of Variance

Source DF SS MS F P Regression 1 0,4662 0,4662 1,92 0,209 Residual Error 7 1,7033 0,2433

Total 8 2,1695

Regression Analysis: Antioxidant versus Flavonoid 95 C TEMPE

The regression equation is

Antioxidant = 0,475 + 2,00 Flavonoid Predictor Coef SE Coef T P Constant 0,4753 0,3132 1,52 0,173 Flavonoid 1,9989 0,7384 2,71 0,030

S = 0,169915 R-Sq = 51,1% R-Sq(adj) = 44,2% Analysis of Variance

Source DF SS MS F P Regression 1 0,21160 0,21160 7,33 0,030 Residual Error 7 0,20210 0,02887

Total 8 0,41369 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 1 0,630 1,7450 1,7346 0,1670 0,0104 0,33 X 2 0,380 1,6050 1,2349 0,0630 0,3701 2,35R R denotes an observation with a large standardized residual. X denotes an observation whose X value gives it large leverage.

51

Regression Analysis: Antioxidant versus Flavonoid 95 C SALT

SOLUTION

The regression equation is

Antioxidant = 1,43 + 0,37 Flavonoid Predictor Coef SE Coef T P Constant 1,4328 0,7814 1,83 0,109 Flavonoid 0,370 2,103 0,18 0,865 S = 0,735909 R-Sq = 0,4% R-Sq(adj) = 0,0% Analysis of Variance

Source DF SS MS F P Regression 1 0,0168 0,0168 0,03 0,865 Residual Error 7 3,7909 0,5416

Total 8 3,8077 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 2 0,610 0,745 1,659 0,594 -0,914 -2,10R R denotes an observation with a large standardized residual.

Regression Analysis: Antioxidant versus Flavonoid 95 C TOTAL

The regression equation is

Antioxidant = 3,75 - 1,14 Flavonoid Predictor Coef SE Coef T P Constant 3,752 1,419 2,64 0,033 Flavonoid -1,142 1,829 -0,62 0,552 S = 0,523458 R-Sq = 5,3% R-Sq(adj) = 0,0% Analysis of Variance

Source DF SS MS F P Regression 1 0,1068 0,1068 0,39 0,552 Residual Error 7 1,9181 0,2740

Total 8 2,0249 Unusual Observations

Obs Flavonoid Antioxidant Fit SE Fit Residual St Resid 2 0,990 2,350 2,621 0,439 -0,271 -0,95 X X denotes an observation whose X value gives it large leverage.

52

Appendix 4 Example of thermal process design calculation.

The changes of hardness has reaction rate constant:

k

75

= 0.0075 min

-1

k

85

= 0.0131 min

-1

k

95

= 0.0172 min

-1

Process is designed to experience 10% reduction of hardness, thus

-ln

(

_C

C

t

0

)

=k

1

t

-ln

(

₁₀₀

90

)

=k

1

t

0.1054=k

1

t

t

75

= 14.05 min



log (t

75

) = 1.15

t

85

= 8.04 min



log (t

85

) = 0.90

t

95

= 6.12 min



log (t

95

) = 0.79

The logaritmic values of time are plotted in linear regression as followed

y = -0.018x + 2.4787

where x is temperature (

o

_{C) and y is time (min) in logaritmic scale.}

If temperature is set at 100

o

C, the process time should be

log t = -0.018(100) + 2.4787

log t = 0.6787

t = 4.77 min

53

CURRICULUM VITAE

The author of this thesis was born on the 16

th

June 1988 in Surakarta,

Central Java. She is the older daughter of teacher Mahmud S and Sri Martiyah.

She completed her secondary education in 2006 at Al-Mukmin Boarding School,

Sukoharjo. From 2006 to 2011 she attended food science and technology at the

Bogor Agricultural University. The title of her bachelor thesis was

“Effects of

Sterilization Level in Canning Process

on Physical Properties of Gudeg Product”.

In 2012, she undertook a Masters in Food Science at the Bogor

Agricultural University, graduating in 2016. Her research was concern for thermal

kinetic of tempe.

Paper of this thesis was submitted for publication at

International

Food Research Journal.