Integrated Rotating Fibrous Bed Bioreactor-Ultrafiltration Process for Xanthan Gum Production from Whey Lactose

DISSERTATION

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

By

Ching-Suei Hsu

Graduate Program in Chemical and Biomolecular Engineering

The Ohio State University

2011

Dissertation Committee:

Professor Shang-Tian Yang, Advisor

Professor Jeffrey J. Chalmers

Professor Kurt W. Koelling

Copyright by

Ching-Suei Hsu

2011

Abstract

Biopolymer fermentation is an environmentally friendly process compared to petroleum-based polymer production. The goal of this study was to evaluate the feasibility of producing xanthan gum, an important biopolymer widely used in food and oil-recovery industries, from whey lactose, a low-value byproduct from cheese manufacturing in the dairy industry, in an integrated fermentation-ultrafiltration process.

First, the fermentation kinetics of xanthan gum production from glucose, galactose, and their mixture, respectively, were studied with Xanthomonas campestris in stirred tank fermentors. In general, comparable fermentation performance in terms of productivity, product yield, and final product titer and quality (rheological properties) was obtained with these various carbon sources. Further batch fermentations with hydrolyzed whey permeate (lactose) showed the feasibility of xanthan gum production using whey permeate as an alternative low-cost feedstock.

However, the high broth viscosity due to product accumulation can cause serious mixing and mass (especially oxygen) transfer problems in conventional stirred-tank bioreactors, resulting in low product yields and poor product quality. A rotating fibrous bed bioreactor (RFBB) operated under a high gravity field can increase mass transfer in viscous xanthan gum fermentation due to the shear-thinning property of xanthan gum broth, thus increasing reactor productivity and final product titer. Furthermore, cells immobilized in the RFBB would allow continuous production of xanthan gum in a low-

ii cell or cell-free broth that can be readily concentrated by ultrafiltration (UF) before alcohol precipitation, thus reducing the amount of alcohol and energy used in the downstream processing by 10-fold. Ultrafiltration also allows the recycle of the fermentation spent medium in subsequent batch xanthan gum fermentations. The effects of recycling the ultrafiltration permeate on xanthan gum fermentation were thus studied, and the results showed no significant changes in productivity, yield, titer, and product quality when the fermentation medium consisted of 75% of recycled permeate, confirming the feasibility of recycling the fermentation medium through the integrated

RFBB-UF process.

To evaluate the scalability of the RFBB, xanthan gum fermentations in 20-liter

RFBB operated in a repeated-batch mode were studied, and the results showed comparable performance to those obtained with 5-liter RFBB. A mathematical model for predicting the oxygen transfer rate, which affects the xanthan gum productivity, in the

RFBB was also developed for process scale up. Finally, process and economic analyses were performed using SuperPro Designer, and the results confirmed that the integrated

RFBB-UF process can reduce the xanthan gum production cost significantly, largely due to the improved reactor productivity and reduced raw materials and energy costs.

Overall, the integrated RFBB-UF process is environmentally friendly and cost effective in producing xanthan gum from whey permeate.

iii

Dedication

This document is dedicated to my family.

iv

Acknowledgments

My grateful acknowledgment goes to my advisor, Dr. Shang-Tian Yang, for his invaluable guidance, patience, kindness and support throughout my Ph.D. study. I have benefited greatly from his expertise in science, deep insights in our research as well as his great personality.

I wish to thank Dr. Robin Ng, Dr. Yunling Bai, and Dr. Liping Wang for their help in my early years of Ph.D. study, and to thank department staff members, Paul

Green and Leigh Evrard, for their assistance in customizing and fabricating the bioreactor used in this study. I also want to thank all the members in my advisor’s research group for their friendship during these years.

I would also like to thank Prof. Jeffrey Chalmers, Prof. Kurt Koelling, and Prof.

Karen Mancl for taking time to serve on my dissertation committee, as well as their valuable suggestions and advices to my research project.

Financial support from the United State Department of Agriculture Small

Business Innovation Research (SBIR) Program via a subcontract from Bioprocessing

Innovative Company is deeply appreciated.

Finally, I wish to thank my family for their unconditionally love and support, and

Dr. Xudong Zhang for his encouragements and company throughout my PhD study.

v

Vita

2002...... B.S. Chemical Engineering and Material

Engineering, National Central University

(Taiwan)

2004...... M.S. Chemical Engineering, National Tsing

Hua University (Taiwan)

2005 to present ...... Graduate Research Associate, Chemical and

Biomolecular Engineering, The Ohio State

University

Fields of Study

Major Field: Chemical and Biomolecular Engineering

vi

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... vi

Fields of Study ...... vi

Table of Contents ...... vvii

List of Tables ...... xiv

List of Figures ...... xvi

Chapter 1: Introduction ...... 1

1.1. Introduction ...... 1

1.2. Objectives ...... 3

1.3. References ...... 6

Chapter 2: Literature Review ...... 9

2.1. Xanthan gum structure and rheology property ...... 9

2.2. Xanthan gum biosynthesis...... 13 vii 2.3. Xanthan gum production ...... 14

2.3.1. Xanthan gum fermentation using inexpensive substrates ...... 14

2.3.2. Xanthan gum fermentation using whey related substrates ...... 15

2.3.3. Effect of sugar mixtures on xanthan gum production ...... 17

2.4. Reactor design for xanthan gum production ...... 18

2.4.1. Rotating fibrous bed bioreactor (RFBB) ...... 19

2.4.2. Xanthan gum fermentation using RFBB ...... 20

2.4.3. Oxygen transfer in RFBB ...... 22

2.4.4. Mean residence time and liquid holdup in RFBB ...... 23

2.4.5. Cell adsorption in RFBB ...... 24

2.4.6. Other research on RFBB ...... 26

2.4.7. Other oxygen transfer study in xanthan gum fermentation ...... 27

2.5. Ultrafiltration in biopolymer separation ...... 27

2.6. Water reuse in fermentation ...... 28

2.7. References ...... 29

Chapter 3: Xanthan gum fermentation using hydrolyzed whey permeate (HWP) as an alternative carbon source ...... 54

3.1. Summary ...... 54

3.2. Introduction ...... 54

viii 3.3. Materials and methods...... 57

3.3.1. Culture and media ...... 57

3.3.1.1. Glucose or galactose-based medium ...... 57

3.3.1.2. Lactose hydrolysate-based medium ...... 57

3.3.1.3. HWP-based medium...... 58

3.3.2. Stirred tank bioreactor ...... 58

3.3.3. Analytical methods ...... 58

3.3.3.1. Cell density ...... 58

3.3.3.2. Glucose and galactose ...... 59

3.3.3.3. Xanthan gum concentration...... 59

3.3.3.4. Apparent viscosity, intrinsic viscosity and rheology of xanthan gum ...... 60

3.4. Results and discussion ...... 61

3.4.1. Fermentation kinetics of xanthan gum production from different carbon

sources ...... 62

3.4.2. Characteristic of xanthan gum produced from glucose, galactose, lactose

hydrolysate, and HWP...... 63

3.5. Conclusions and recommendation...... 64

3.6. References ...... 66

Chapter 4: Water Recycle in RFBB-Ultrafiltration Integrated Process ...... 77

4.1. Summary ...... 77 ix 4.2. Introduction ...... 77

4.3. Materials and methods...... 79

4.3.1. Culture and media ...... 79

4.3.2. Shake flask culture for replacement percentage experiment ...... 79

4.3.3. Fermentation using a stirred tank ...... 80

4.3.4. Ultrafiltration ...... 80

4.3.5. Analytical methods ...... 81

4.3.5.1. Cell density ...... 81

4.3.5.2. Xanthan gum concentration...... 81

4.3.5.3. Viscosity of xanthan gum ...... 81

4.4. Results and disccusion...... 82

4.4.1. Replacement percentage ...... 82

4.4.2. Sequential fermentation using spent medium from ultrafiltration ...... 83

4.4.3. Charateristic of xanthan gum produced from spent medium ...... 85

4.5. Conclusions and recommendation...... 88

4.6. References ...... 89

Chapter 5: Xanthan Gum Fermentation using Rotating Fibrous Bed Bioreactor (RFBB)

...... 103

5.1. Summary ...... 103

x 5.2. Introduction ...... 103

5.3. Materials and methods...... 111

5.3.1. Culture and media ...... 111

5.3.2. Rotating fibrous bed bioreactor ...... 112

5.3.2.1. Liquid continuous repeated-batch ...... 112

5.3.2.2. Gas continuous repeated-batch ...... 113

5.3.3. Analytical methods ...... 113

5.3.3.1. Cell density ...... 113

5.3.3.2. Xanthan gum concentration...... 114

5.3.3.3. Viscosity of xanthan gum ...... 114

5.4. Results and discussion ...... 114

5.4.1 Xanthan gum production in repeated-batch RFBB under liquid continuous mode ...... 114

5.4.2 Xanthan gum production in repeated-batch RFBB under gas continuous mode .

...... 116

5.4.3 Oxygen transfer coefficient in RFBB under gas continuous mode ...... 116

5.4.4 RFBB scale up ...... 118

5.5. Conclusions and recommendation...... 121

5.6. Nomenclature ...... 122

5.7. References ...... 123 xi Chapter 6: Scale-up and Economic Evaluation of an Integrated Rotating Fibrous Bed

Bioreactor-Ultrafiltration Process for Viscous Biopolymer Fermentation ...... 137

6.1. Summary ...... 137

6.2. Introduction ...... 137

6.3. Materials and methods...... 138

6.3.1. Process flowsheet ...... 138

6.3.2. Economic analysis ...... 139

6.3.3. Equipment cost ...... 140

6.4. Results and discussion ...... 140

6.4.1. Major equipment specificationand cost ...... 140

6.4.2. Cost of raw material and utility to produce 3000 ton/year xanthan gum ...... 141

6.4.3. Economic analysis ...... 142

6.4.3.1. Direct fixed capital (DFC) ...... 142

6.4.3.2. Annual operating cost ...... 142

6.4.3.3. Profitability analysis ...... 142

6.4.4. Reducing environmental impact from liquid, solid waste disposal and emission

by using integrated RFBB-UF process for xanthan gum production ...... 143

6.5. Conclusions and recommendation...... 144

Chapter 7: Conclusions and Recommendations ...... 161

7.1. Conclusions ...... 161 xii 7.2. Recommendations ...... 163

Bibliography ...... 165

Appendix ...... 176

xiii

List of Tables

Table 2.1. Literature reviews of xanthan gum production from whey ...... 40

Table 2.2. Literature reviews of xanthan gum production from glucose or sucrose ...... 41

Table 2.3. Literature reviews of xanthan gum production from other substrates ...... 42

Table 2.4. Literature review of bioreactor studies in xanthan gum fermentation ...... 43

Table 2.5. Litermaure reviews of water recycle in fermentation ...... 45

Table 3.1. Summary fermentation using different carbon sources ...... 69

Table 3.2. Summary rheology index of 5 g/L xanthan gum ...... 70

Table 3.3. Literature review of rheology index of xanthan gum ...... 70

Table 4.1. Medium composition in replacement percentage study. each flask contained 50 mL of medium at pH 6.5...... 90

Table 4.2. Volume, viscosity and concentration of xanthan gum before and after ultrafiltration...... 91

Table 4.3. Summary of fermentation using ultrafiltration permeate...... 92

Table 4.4. Summary of appearant viscosity and intrinsic viscosity of xanthan gum from fresh and spent medium...... 93

Table 5.1. Summary of RFBB fermentation setups and operation modes in literatures. 125

Table 5.2. Comparison of xanthan gum production in free-cell and RFBB fermentations

...... 126

xiv Table 5.3. Comparison of RFBB fermentation in 5 L liquid continuous repeated-batches

...... 127

Table 5.4. Comparison of RFBB fermentation in gas continuous repeated-batches ...... 128

Table 6.1. Process performance based on historical experimental data...... 145

Table 6.2. Major equipment specification and cost of STR-D and STR-UF-D processes

(production rate of 3000 tons/year)...... 146

Table 6.3. Major equipment specification and cost of RFBB-D and RFBB-UF-D processes (production rate of 3000 tons/year) ...... 147

Table 6.4. Raw material cost in STR-D and STR-UF-D processes (production rate of

3000 tons/year)...... 148

Table 6.5. Raw material cost in RFBB-D and RFBB-UF-D processes (production rate of

3000 tons/year)...... 149

Table 6.6. Annual utilities requirement for 3000 tons xanthan gum production ...... 150

Table 6.7. Fixed capital estimate of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (production rate of 3000 tons/year) ...... 151

Table 6.8. Annual operating cost of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (production rate of 3000 tons/year) ...... 152

Table 6.9. Profitability analysis of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (production rate of 3000 tons/year) ...... 153

Table 6.10. Executive summary of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (production rate of 3000 tons/year) ...... 154

xv

List of Figures

Figure 1.1. An integrated fermentation-ultrafiltration process for xanthan gum production...... 7

Figure 1.2. Research objectives and scopes of this study ...... 8

Figure 2.1. Chemical structure of xanthan gum (Garcia-Ochoa et al., 2000) ...... 46

Figure 2.2. Conformations of xanthan gum in solution (Margaritis and Pace, 1985) ...... 47

Figure 2.3. Effect of shear rate on the apparent viscosity of xanthan gum solution at 20ºC

(Bandalusena et al., 2009) ...... 48

Figure 2.4. Metabolic pathway of X. campestris producing xanthan gum (Hsu and Lo,

2003) ...... 49

Figure 2.5. Mechanism of xanthan gum synthesis in X. campestris (Vorhölter et al., 2008)

...... 50

Figure 2.6. Conventional xanthan gum production process(Rosalam and England, 2006)

...... 51

Figure 2.7. A rotating fibrous bed bioreactor (arrows indicate liquid flow direction)

(Yang et al., 1996) ...... 52

Figure 2.8. A Rotating fibrous bed bioreactor system (Doma, 1999)...... 53

xvi Figure 3.1. Batch fermenation kinetics of xanthan gum production in a stirred-tank fermentor at pH 7 and 30oC from various carbon sources, such as glucose (A), galactose

(B), and lactose hydrolysate (C), with 0.3% yeast extract...... 71

Figure 3.2. Batch fermenation kinetics of xanthan gum production in a stirred-tank fermentor at pH 7 and 30oC from HWP with 0.3% (A), and 0.1% (B) yeast extract ...... 72

Figure 3.3. Effect of carbon source on cell growth (A) and broth viscosity (B) ...... 73

Figure 3.4. Rheology of 0.5 % solution of xanthan gum produced from glucose, galactose, lactose hydrolysate, and HWP. (Spindle S) ...... 74

Figure 3.5. Shear thinning properties of xanthan gum from different carbon sources ..... 75

Figure 3.6. Effect of carbon sources on intrinsic viscsoity of xanthan gum ...... 76

Figure 4.1. Diagram of ultrafiltration system ...... 94

Figure 4.2. Effects of ultrafiltration permeate replacement percentage on cell growth, xanthan gum production, and broth viscosity (n=2) ...... 95

Figure 4.3. Fermentation kinetics of xanthan gum production in fresh medium STR-UFP0

(A), and spent medium STR-UFP1 (B) an d STR-UFP2 (C)...... 96

Figure 4.4. Comparision of cell growth (A), glucose consumption (B), and broth viscosity

(C) in xanthan gum fermentations with fresh and recycled media ...... 97

Figure 4.5. Comparision of apparent viscosity of xanthan gum (1% w/v) produced from fresh and recyceled media in sequential batch fermentations. Spindle 5, 100 rpm (A), and

60 rpm (B)...... 98

Figure 4.6. Rheology of 0.5 % xanthan gum from fresh and spent medium in terms of shear stress (Spindle S)...... 99

xvii Figure 4.7. Apparent viscosity of 1 % solution of xanthan gum produced from fermentation with fresh and spent media (spindle 5)...... 100

Figure 4.8. Effect of ultrafiltration permeate on intrinsic viscsoity of xanthan gum...... 101

Figure 4.9. The accumulation of possible inhibitors by recycling the spent medium at

75% replacement rate ...... 102

Figure 5.1. A Rotating fibrous bed bioreactor (arrows indicate liquid flow direction).

(Yang et al., 1996) ...... 129

Figure 5.2. A rotating fibrous bed bioreactor system (Doma, 1999) ...... 130

Figure 5.3. Repeated-batch fermentations for xanthan gum production in a rotating fibrous bed bioreactor operated under liquid continuous conditions ...... 131

Figure 5.4. Repeated-batch fermentations for xanthan gum production in a rotating fibrous bed bioreactor operated under gas continuous conditions ...... 132

Figure 5.5. S Solubility of oxygen in xanthan gum solution (Doma, 1999) ...... 133

Figure 5.6. Effect of recirculation rate on kLa at different rotational rate, with 20 g/L xanthan gum, air flow rate 3.5 L/min, and liquid volume of 3 L (Doma, 1999) ...... 134

Figure 5.7. Average residence time at different rotational rate, liquid flow rate and xanthan gum concentration (Doma, 1999)...... 135

Figure 5.8. Power consumption at different rotational rate and different volumetric air flow rate.(Doma, 1999) ...... 136

Figure 6.1. Process diagram of a conventional xanthan gum production process (STR-D).

...... 155

Figure 6.2. Process diagram of a conventional xanthan gum production process with ultrafiltration (STR-UF-D)...... 156 xviii Figure 6.3. Process diagram of a xanthan gum production process using RFBB (RFBB-D)

...... 157

Figure 6.4. Process diagram of an integrated RFBB-UF xanthan gum production process

(RFBB-UF-D)...... 158

Figure 6.5. Comparison of annual water usages for different xanthan gum production processes with an annual produtction rate of 3000 tons...... 159

Figure 6.6. Comparison of solid biomass wastes generated from free-cell (STR) and immobilized-cell (RFBB) fermentations with an annual produtction rate of 3000 tons. 160

xix

Chapter 1: Introduction

1.1. Introduction

Xanthan gum is the most successful microbial extracellular heteropolysaccharide applied in various industries, such as additive to drilling mud in oil recovery, and thickening and suspending agent in food, cosmetic and pharmaceutical industries. Because of its unique rheological properties and stable characteristics in wide ranges of pH and temperature, it has a worldwide market over 75,000 tons and $500 million annually (Wyatt and

Liberatore, 2009).

Xanthan gum is produced by a family of plant pathogen bacteria, Xanthomonas sp. The strain commonly used in manufacture production was originally isolated from cabbage.

In 1960s, xanthan gum production was commercialized and used glucose or sucrose as carbon source in stirred-tank fermentors (Garcia-Ochoa et al., 2000). After fermentation, xanthan gum was recovered by alcohol precipitation, using either ethanol or isopropanol.

Then, the product was dried and grinded into powder. The color of xanthan gum powder is beige white to light yellow.

The raw materials of fermentation account for a significant part of the process cost, especially with the increasing price of glucose and water today. In this case, using waste stream from other industries, such as whey permeate from cheese manufacturing, for xanthan gum production would not only solve the problem of the waste stream disposal but also reduce the cost for xanthan gum production.

1 Cheese whey is a byproduct in cheese production. The main components of cheese whey are protein, lactose, fat and lactic acid. The protein in cheese whey can be collected and concentrated via ultrafiltration and sold as whey protein. However, the remaining lactose in the ultrafiltration permeate has not been fully utilized. How to utilize the waste stream instead of disposing it becomes a critical challenge of cheese manufacturing today. For decades, application of whey and whey permeate in fermentation has been investigated extensively. However, only a few microorganisms can utilize lactose as the carbon source, and to date only lactic acid and ethanol production from whey or whey permeate has been successfully developed for industrial application (Converti et al., 1991;

Medvedeva et al., 1989; Hsiao and Glatz, 1996). On the other hand, whey disposal is costly with a price about $307 per ton due to its high biological oxygen demand (BOD).

Similarly, the disposal of the enormous volume of fermentation spent media after xanthan gum recovery is also costly because of the residual sugars and nutrients present in the spent medium. In this study, the feasibility of using whey permeate and ultrafiltration permeate for xanthan gum production was thus investigated.

However, current xanthan gum fermentation is facing several challenges. The viscosity increase during fermentation limits oxygen transfer in fermentation and affects the xanthan gum yield, productivity and final product concentration. The cost of product could be reduced by increasing yield, productivity and final concentration using a rotating fibrous bed bioreactor (RFBB) integrated with an ultrafiltration (UF) process shown in Figure 1.1. This integrated RFBB-UF process could increase the product yield to >75% (w/w), productivity to >1 g/L/h, and final concentration to 35 g/L as compared to a low productivity of <0.5 g/L/h and final concentration of <30 g/L with conventional

2 processes in stirred tanks (Yang et al., 1996). Integrating fermentation with ultrafiltration to concentrate xanthan gum about 10 times before alcohol precipitation can also reduce the volume of alcohol used in precipitation and the energy used in alcohol recovery.

Using this approach, xanthan gum production can be greatly improved (Yang et al., 1996).

However, further scale up study with this integrated fermentation-ultrafiltration process design is needed.

1.2. Objectives and scope of study

The overall goal of this study was to develop an economical and sustainable fermentation-ultrafiltration process for producing xanthan gum from whey lactose. The specific objectives and associated work completed in this study are briefly described below:

1) To study the feasibility of using lactose present in the waste stream from the dairy

industry as the fermentation substrate for xanthan gum production

Lactose from cheese whey permeate cannot be used by most of microorganisms.

Fortunately, after enzymatic hydrolysis, lactose can be decomposed to glucose and galactose. Thus, whey permeate can be used more efficiently. In this study, the effects of different carbon sources including glucose and galactose on xanthan gum fermentation were studied. The performance of these carbon sources was evaluated in terms of fermentation productivity, product yield and final product concentration. Finally, the rheology of xanthan gum produced from different carbon sources was also compared.

The results are presented in Chapter 3.

3 2) To study the feasibility of recycling the ultrafiltration permeate for xanthan gum

fermentation

The booming of human population, industrialization and urbanization increases the demand for water. As water source is becoming scarce and insufficient, it may raise the tension among various applications. To reduce the stress, water recycle demands immediate attention. Water is the major component in the fermentation medium. Tons of water are used in each batch of industrial fermentation, with another few times more required in temperature control, facility cleaning, Cleaning-In-Place (CIP), and

Sterilization-In-Place, etc. After product recovery, nutrient residues and metabolic wastes remain in the fermentation broth. Usually, it is disposed into the sewer system. According to the United States regulation, the maximum concentration of BOD discharges is 250 mg/L, with a surcharge fee varying in different cities. For example, the surcharge of BOD in Columbus, Ohio, in 2011 is $0.66-0.74/pound. If the spent medium can be recycled, it not only can reduce the amount of water used in fermentation, but also can reduce the amount of the waste water. To date, there is not much research in fermentation water recycle and no report on the study of recycling water in biopolymer fermentation.

Therefore, in this study, the feasibility and effects of recycling ultrafiltration permeate on xanthan gum fermentation were investigated in shake flasks. The optimal percentage of the spent medium in the medium for each subsequent batch fermentation was determined and then further evaluated in stirred-tank fermentors. The productivity, yield, final product concentration, and rheological properties of xanthan gum from fermentations with recycled spent media were compared with those from fermentations with fresh media, and the results are given in Chapter 4.

4 3) To optimize and scale up rotating fibrous bed bioreactor (RFBB) for xanthan gum

production

Mass transfer limitation due to high broth viscosity in xanthan gum fermentation can be alleviated by using a rotating fibrous bed bioreactor. In the previous study, due to the equipment capacity, operational parameters including fermentation volume, rotational rate and liquid recirculation rate were limited. In this study, repeated-batches with a higher broth volume using higher recirculation liquid flow rates were conducted in both gas and liquid continuous modes. The results were used to validate an empirical model of oxygen transfer that can be used in the scale up design of the RFBB for xanthan gum fermentation. Detailed discussion on the RFBB scale up for xanthan gum fermentation is presented in Chapter 5.

4) To evaluate the economic feasibility of the integrated RFBB-UF process for

xanthan gum production

Ultrafiltration (UF) is an energy efficient operation widely used in recovering and concentrating proteins and other biopolymers from a dilute solution. UF can reduce the volume of xanthan gum fermentation broth before alcohol precipitation by about 10-fold.

With cells immobilized in the RFBB, ultrafiltration can be readily integrated with the fermentation process without suffering from membrane fouling by the microorganisms.

Therefore, the integrated RFBB-UF process should reduce the xanthan gum production cost significantly. In Chapter 6, economic analysis of xanthan gum production using the integrated process was studied and compared to the conventional xanthan gum fermentation process at various production scales.

5 Figure 1.2 illustrates the four specific objectives and the scope of this dissertation research. In addition, Chapter 2 provides detailed literature review focusing on different strategies for enhancing xanthan gum production, including the use of alternative substrates and novel reactor designs and operation modes. Finally, Chapter 7 provides conclusions and recommendations for future work.

1.3. References

Converti, A., Perego, P., Lodi, A., Fiorito, G., Borghi, M.D., Ferraiolo, G. In-situ ethanol recovery and subst4rate recycling during continuous alcohol fermentation. Bioprocess Engineering, (1991), 7, 3-10.

Garcia-Ochoa, F., Gomez Castro, E., Santos V.E., Oxygen transfer and uptake rates during xanthan gum production. Enzyme and Microbial Technology (2000), 27, 680-690.

Hsiao, T-Y. and Glatz, C.E., 1996, Water Reuse in the L-Lysine Fermentation Process, Biotechnology and Bioengineering, 49, 341-347.

Medvedeva, E.I., Oleinik, T.P., Panchenko, K.A., Petrenko, E. B. Multipurpose use of liquid wastes in microbial lysine synthesis technology. Biotekhnologiya, (1989), 6, 761-767.

Wyatt, Nicholas B. and Liberatore, Matthew W. Rheology and viscosity scaling of the polyelectrolyte xanthan gum. Journal of Applied Polymer Science. (2009), 114, 4076-4084.

Yang, Shang-Tian; Lo, Yang-Ming; Min, David B. Xanthan gum fermentation by Xanthomonas campestris immobilized in a novel centrifugal fibrous-bed bioreactor. Biotechnology Progress. (1996), 12(5), 630-637

6

Fermentation Ultrafiltration Alcohol Distillation

Filtration

2.5% 15%

Heating Drying o 60 C Alcohol precipitation

Xanthan Gum Spent medium

Figure 1.1. An integrated fermentation-ultrafiltration process for xanthan gum production. The fermentation carried out in a rotating fibrous bed bioreactor can produce cell-free xanthan broth that can be concentrated by ultrafiltration without suffering from membrane fouling. The concentrated broth can greatly reduce the amount of alcohol used in the xanthan gum recovery process and the energy cost. The process can also reuse much of the fermentation medium and reduce the spent medium discharged for wastewater treatment.

7

Figure 1.2. Research objectives and the scope of this study.

8

Chapter 2: Literature Review

2.1. Xanthan gum structure and rheology property

Since xanthan gum was discovered in the 1950s, it has been studied thoroughly for its structure and properties. Xanthan gum is a large molecule with a molecular weight ranging from 2 to 50 million Dalton, which can be affected by fermentation conditions.

Xanthan gum can be classified as a branched heteropolysaccharide, which is a cellulose- like backbone with three-unit-long side chains at the C-3 position of the backbone

(Figure 2.1) produced by Xanthomonas bacteria. The side chain residuals contain two D- mannose and a D-glucuronic acid. The mannose residue linked to the backbone has D- acetyl substituents on the C-6 position and the pyruvic acetal substituents are occasionally added on the terminal of D-mannosyl residues. The stiff side chains make xanthan a semirigid polyelectrolyte (Tinland et al., 1990). The frequency of this addition depends on bacterial strains, nutrient supply, and fermentation conditions. Generally, the pyruvate content is between 0.31 and 0.56 (Zirnsak et al., 1999). The branching structure of xanthan gum contributes to the great stability of xanthan gum in a wide range of temperature, pH and pK (ionic concentration), and good resistance to ultrasonic treatment in food process (Tiwari et al., 2010).

The conformation of xanthan gum has three different forms in aqueous solution, such as single-stranded, double-stranded, or partly dissociated double-stranded structures

9 (Stokke, 1986). They are favored by different acetate to pyruvate ratios, temperatures, and ionic strengths (Figure 2.2). For example, at high temperature, above 40 ºC, xanthan molecules shift from the ordered form to a disordered structure (Casas et al., 2000). The negative charged side chains also make xanthan gum considered as a polyelectrolyte. The relationship of zero shear rate viscosity to xanthan gum concentration can be used to

* indicate the overlap (c ), entanglement (ce), and final critical (cD) concentrations for isotropic polyelectrolyte solutions. Below overlap concentration, xanthan gum molecules are too diluted and can be considered with no interaction between the polymers. The rheological property observed below overlap concentration represents a single polymer molecule responding to the stress. When the polymer concentration is above overlap concentration, interaction between polymers is no longer negligible. The exact concentration of critical point between dilute and semidilute regimes depends on the xanthan sample and solvent conditions. Between overlap concentration and entanglement concentration is semidilute unentangled regime. Semidilute entangled regime is between entanglement concentration and final critical concentration. Above final critical concentration is considered as concentrated regime. Wyatt and Liberatore (2009) found that there were four concentration regimes in a salt-free xanthan gum solution, but only three concentration regimes were found when salt was present in the solution. In a salt-

* free solution, the associated critical concentrations of xanthan gum (c ≈70 ppm, ce≈400

* ppm, cD≈ 2000 ppm) were smaller than those (c ≈200 ppm, ce≈800 ppm) in a 50 mM

NaCl solution, and no final critical concentration was found in the salt solution (Wyatt and Liberatore, 2009). The higher critical concentration in salt solution could be because the ions screen the electrostatic interactions between the polymer chains and therefore 10 affect the interactions between the polymer chains. Their result shows that xanthan polymers behaved like neutral polymers in salt solution. They also found that in a salt free environment, diluted (below 20 ppm) xanthan gum solution was close to Newtonian fluid instead of shear thinning fluid. Because most of the paper studied the rheology of xanthan gum for its application, the concentrations studied were mostly within dilute and semidilute regimes. In fermentation process, the concentration of xanthan gum is much higher, which is within concentrated regimes. In the concentrated regime, Argin-Soysal and Lo (2004) studied the effect of hydrogen bonding within xanthan molecules and found that hydrogen bonding affected xanthan rheology more significantly at low xanthan gum concentration. In their study, xanthan gum solutions transited from isotropic to anisotropic between the concentration of 1 and 2 wt %.

A notable rheological property of xanthan gum is that it is a non-Newtonian

(pseudoplastic) fluid. The viscosity of xanthan gum decreases with increasing shear rate, so that it is easy to operate with mixing and pumping. The rheology of xanthan gum can be described by a modified power law:

    K() n 0 (2.1) where τ is the shear stress (Pa, or dyne/cm2),   is the shear rate (s-1), n is the flow

n index, K is the consistency index (Pa∙s ), and τ0 is the yield stress, which is the stress required to initiate the movement of the fluid. The yield stress depends on xanthan concentration, so it can be neglected in dilute xanthan solution. Most of the xanthan gum application papers studied rheology of xanthan gum solution in dilute concentration, and the rheology model can be simplified to the Power law. Song et al. (2006) studied the

11 rheology of concentrated xanthan gum, from 1 wt % to 4 wt %. Based on their results, xanthan gum within this concentration range showed a finite magnitude of yield stress, which could be attributed to the large numbers of hydrogen bonds in the helix structure of xanthan gum. Since xanthan is a pseudoplastic liquid, the n here is less than one. The smaller the n is, the more shear thinning the fluid is. The rheological properties of xanthan macromolecules are affected by the molecular weight and the content of pyruvate residues. The larger molecule results in higher viscosity and the pyruvate content facilitates the aggregated form of xanthan gum, which also results in a higher viscosity. Based on Psomas et al.’s finding (2007), molecular weight was not sensitive to the operational conditions studied in their research, and all xanthan gum produced in various conditions was around 5×105 Da. The fermentation ranges studied in their paper were often used in xanthan gum production with agitation rate between 100-600 rpm, temperature between 25-35 ºC, and cultivation time between 24-72 h. According to

Bandalusena et al., (2009) density of xanthan gum solution under 20 ºC was nearly constant (998.22 to 999.11 kg/m3) between 0.025 to 0.2 wt %. Their result of shear viscosity of xanthan gum solution at 20 ºC was in Figure 2.3. Rocha-Valadez et al. (2007) also found that n decreased from ~1 to 0.1 and K increased from 0 to 30 during fermentation within 72 h. The n decreased from 0.76 to 0.46 and K increased from

0.0068 to 0.138 Pa∙sn, when the concentration of xanthan solution in distilled water increased from 0.05 to 0.1 g/L. This data range is too narrow for fermentation, but it covers the concentration range in xanthan gum application as a thickening agent

(Bandalusena et al., 2009). Tiwari et al. (2010) reported the n and k as 0.265 and 6.787 of

1 % xanthan gum solution. For the United States Pharmacopeia–National Formulary 12 (USP-NF) grade xanthan gum, the minimum apparent viscosity should not be less than

600 mPa, under a standard protocol measuring 1% (w/w) xanthan gum (with additional

1.0 % w/w KCl) via a rotating spindle viscometer at 60 rpm, 24±1ºC (Thacker, et al.,

2010). Thacker et al. (2010) compared the rheological characteristics of 6 different grades and lots of xanthan gum, in terms of Power law coefficients and exponents of apparent viscosity (K, n), storage modulus (K’, n’), and loss modulus (K”, n”). According to their finding, although all grades and lots of xanthan gum met the standard of USP-NF’s lower limit of viscosity, the n and K were significantly different among lots of the xanthan gum, both inter- and intra-manufacture. This could mean that the distribution of molecular weight or side chains content varies in each lot and grade. To better understand what cause the variations, a distribution of molecular weight and the pyruvate content need to be determined. Zirnsak et al. (1999) summarized the steady shear and dynamic rheological properties of xanthan gum solution. According to their study, shear thinning properties of xanthan gum were more significant in aqueous solution than in higher viscous solution, such as corn syrup. When xanthan gum was dissolved in high viscous solvent, the shear viscosity did not depend on shear rate.

2.2. Xanthan gum biosynthesis

Xanthan gum can be produced by various Gram-negative Xanthomonas bacteria, including X. campestris, X. fragaria, X. gummisudans, X. juglanids, X. phaseoli, and X. vasculorum (Kennedy and Bradshaw, 1984), among which X. campestris is currently used in commercial production. Many strains of Xanthomonas have been sequenced

(Qian et al., 2005; Thieme et al., 2005; Lee et al., 2005; Ochiai et al., 2005). The

13 metabolic pathway of X. campestris producing xanthan gum is shown in Figure 2.4 (Hsu and Lo, 2003). In 2008, Vorhölter et al. proposed the mechanism of xanthan gum synthesis, which suggested that xanthan gum was polymerized near cell membranes

(Figure 2.5). The cell-cell signaling or quorum sensing that triggers X. campestris to form biofilm, which consists of xanthan gum, has also been studied (Vojnov et al. 2001). In addition, several unstructured (Weiss and Ollis, 1980; Garcia-Ochoa et al., 1995) and structured (Garcia-Ochoa et al., 1998 and 2004; Faria et al., 2010) kinetic models have been developed to predict the production of xanthan gum.

2.3. Xanthan gum production

Currently, xanthan gum is primary produced in complex medium with glucose or sucrose in stirred tanks, followed by alcohol precipitation to recovery xanthan gum (Figure 2.6).

Garcia-Ochoa et al. (2000) summarized the maximum xanthan gum concentration and yield from literatures, using stirred tank, bubble column, plugging jet reactor and airlift reactor. The final concentration of xanthan gum production was less than 30 g/L in all operating conditions and reactors used. More literatures using glucose and sucrose are summarized in Tables 2.1, 2.2, and 2.3.

2.3.1. Xanthan gum fermentation using inexpensive substrates

Finding alternative carbon sources for xanthan gum production is still attractive, especially due to the increasing price of glucose today. To reduce the cost of raw materials in fermentation and to solve the waste and by product pollution from other industries, there were several xanthan gum fermentation processes using agricultural by- product, waste stream, such as whey from cheese industry, and hydrolysates (Salah et al., 14 2010; Hamilton et al., 2009; Bilanovic, et al., 2008; Ashraf et al., 2008; Zhang and Chen,

2010) has been investigated. The varieties of agricultural starch or waste studied was regional and interesting, such as sugarcane fluid (Hamilton et al., 2009; Faria et al.,

2010), chestnut extract (Liakopoulou-Kyriakides, 1999), melon, watermelon, cucumber and tomato (Moreno et al., 1998), sugar beet molasses (Afendra et al., 2002), and sago

(Janaun et al., 2005). Using raw starch to produce xanthan gum can increase the utilization of biomass in agriculture; however, it requires efficient α-amylase in the microorganism to break down the starch. The α-amylase could be native or recombinant

(Konsoula et al., 2008). Most of the literatures only reported the feasibility of using these inexpensive substrates. Faria et al. (2010) further proposed an unstructured kinetic model to predict yield and productivity of xanthan gum production from sugar cane broth. The yield and productivity could reach 0.35 g xanthan/g sucrose and 0.57 g/L/h, respectively.

However, the final concentration of xanthan gum was only around 14 g/L, when initial sucrose concentration was 40 g/L. Overall, although using alternative carbon sources from agricultural or industrial waste could reduce the cost of raw material and processing waste stream, and further reduce the pollution, the yield and final concentration of xanthan gum was not competitive to those using glucose as the carbon source. Therefore, the optimization of these processes is required. A review paper (Rosalam and England,

2006) summarized the efforts of researchers on alternative carbon source search.

2.3.2. Xanthan gum fermentation using whey related substrates

Whey has been proposed as an alternative carbon source for many years. Americans consume 48 million tons of cheese every year. The abundance of liquid byproduct, whey,

15 from cheese production (contains about 4.5 - 5% lactose, 0.6 - 0.8% soluble proteins, 0.4

- 0.5% lipids and 8 - 10% mineral salts) has yet been fully utilized. Due to the high

Biological Oxygen Demand (BOD) (30,000 to 50,000 mg O2 per liter wastewater), whey disposal could cause severe environmental problems and high cost to cheese manufactures. Today, whey is filtrated to extract whey protein for food, nutrient supply or feed. Whey permeate is still rich of lactose, which leaves the high BOD problem unsolved.

Researchers tried to apply whey or whey permeate as carbon sources in fermentation to produce ethanol, lactic acid, acetic acid, propionic acid, single cell protein production

(Coté et al., 2004; González Siso, 1996; Bogaert, 1997; Lamboley et al., 1997; Nicolas et al., 2004; Nykanen et al., 1998; Schepers et al., 2006; Silveira et al., 2005; Tyagi and

Kluepfel, 1998). There were also many attempts to use the lactose in whey and whey permeate to produce lactic acid. The demand of whey permeate for lactic acid production is much less than whey permeate generated, so exploration of more applications of excessive whey permeate is highly desired in current cheese industry.

Xanthan gum is one of the potential products from whey permeate. However, most wild- type Xanthomonas campestris strains do not consume lactose, because of low galactosidase activity. To solve the problem, several attempts have been made: 1) to isolate X. campestris strains which can utilize lactose (Schwartz and Bodie, 1985; Yang et al., 2003; Ashraf, et al., 2008; Mesomo et al., 2009); 2) to clone β-galactosidase gene or other genes for lactose utilization into X. campestris (Thorne, et a., 1988; Walsh et al.,

1984); 3) to hydrolyze lactose before fermentation (Maldonado, 1992). There were some

16 lactose-utilizing strains isolated and reported, but their productivity was not comparable to those using glucose or other easily utilized carbon sources (Yang, 2002 and 2003). In addition, isolated strains utilizing lactose as carbon source were not stable after certain generation (Schwartz and Bodie, 1985). Before 2008, the fermentation scale using whey in xanthan gum production was only in flasks with the highest final concentration of 22-

25 g/L. From 2008, Ashraf et al. and Mesomo et al. started to product xanthan gum from whey in fermentors, with a working volume of 2 and 2.5 L. The highest production from whey was 36 g/L by Mesomo et al. in 2009. However, the concentration of whey used in fermentation was not reported, so the yield was unknown. Table 2.1 summarized the xanthan gum production using whey.

2.3.3. Effect of sugar mixtures on xanthan gum production

In our study, xanthan gum was produced from hydrolyzed whey permeate. Unlike the major carbon source, hydrolyzed lactose-rich whey permeate contains 1:1 of glucose and galactose. How efficiently the carbon sources are used in X. campestris determines the productivity. One interesting study used a glucose/xylose mixture as carbon source in xanthan gum production (Zhang and Chen, 2010). It showed that the product concentration was strongly related to glucose concentration. However, xylose affected the molecular weight, pyruvate and acetate content of xanthan gum, which means xylose could be used to improve the quality of xanthan gum. In this research, glucose and xylose were used simultaneously, which was similar to our study using glucose/galactose mixture. Comparing to xylose, galactose was more efficiently used in X. campestris. In this study, the author claimed that xanthan gum production rate was determined by the

17 glucose concentration in the beginning of fermentation. When the ratio of glucose to xylose was higher than 1:1, inhibition occurred. Intrinsic viscosity was used to compare the average molecular weight of xanthan gum from different ratios of carbon sources.

The intrinsic viscosity reported here was between 0.7 to 1.043 L/g, which was much lower than our data 3.44 to 6.87 L/g. Overall, in terms of hydrolysate fermentation, whey permeate hydrolysate seems to be more promising for xanthan gum production comparing to lignocelluloses hydrolysate.

2.4. Reactor design for xanthan gum production

The key issue for xanthan fermentation is that X. campestris is strictly aerobic bacteria, so sufficient oxygen supply is critical during fermentation. Conventional batch fermentation in stirred tank reactors could not provide enough oxygen for X. campestris, because the accumulated xanthan gum dramatically increased the viscosity of the broth and it became a barrier for oxygen transfer during the fermentation. The limitation of oxygen severely restricted the yield, productivity and even quality of xanthan gum (Yang, 1996; Herbst,

1992; Hsu and Lo, 2003). For example, the dissolved oxygen concentration during fermentation was an important factor for xanthan molecular weight (Flores, 1994).

Researchers have tried to overcome this oxygen limitation problem in viscous, non-

Newtonian fermentation broth by studying the fermentation parameters (feeding technique, temperature, pH, agitation, aeration, and antifoam adding), and the reactor systems (airlift bioreactors and bubble columns) and configurations (impeller design).

(Rosalam and England, 2006; Roukas and Mantzouridou, 2001; Kawase and Hashimoto,

1996; Godbole et al., 1984; Rocha-Valadez et al., 2007; Brahmbhatt, 2007; Lo et al.,

18 2001). The bubble column had an upper viscosity limit, which the oxygen transfer decreased dramatically beyond this point (Van Reit and Tramper, 1991). For a given superficial air velocity, the oxygen transfer was even lower than in bubble columns

(Stanbury, 1995). Another system tested was the water-in-oil (w/o) system (Zhao, 1999).

The w/o system could provide a large number of small compartments for fermentation; therefore, xanthan gum was separated into small particles which reduced the apparent viscosity of the overall system. Although, this system had high productivity, it was not easy for downstream purification to remove the oil phase and surfactant. Therefore, it was hard to be commercialized. The xanthan gum fermentation in different reactor design was summarized in Table 2.4, which is expended from a review paper by Garcia-Ochoa et al. (2000). Commercially, Stirred tanks are stilled used in xanthan gum production, and they can be used under high viscosity of up to 2000 cp (Shuler and Kargi, 1992).

2.4.1. Rotating fibrous bed bioreactor (RFBB)

To overcome the problem of oxygen transfer in viscous biopolymer fermentation, we propose to use a RFBB (Figure 2.7), which is a modified stirred tank composed with a rotating fibrous bed and a recirculation system (Yang et al., 1996). The rotating fibrous bed composed of two concentric stainless steel cylinders was added onto the stir impeller.

Between the two stainless steel cylinders was a spiral-packed cotton towel packed with a stainless steel wire mesh. This part was for microorganism immobilization. After attaching on the matrix, cells were able to stay in the fibrous bed under a high gravitational field generated by bed rotation. The recirculation system was to pump the bottom xanthan broth out and then spray back to the fibrous bed from top via a nozzle.

19 The rotating packed bed (RPB) in a reactor was patented by Ramshaw and Mallison in

1981 to increase gas liquid mass transfer in distillation, adsorption, and stripping (Rao et al., 2004). The bed can be packed with different matrices, such as glass spheres, plastic beads, metal wiremesh, and disks. The range of porosities and surface areas of these dry packing materials were 0.4-0.95 and 750-5000 m2/m3. RPB has been commercially used by companies, such as Glitsch Inc. and the Dow Chemical Company (Rao et al., 2004).

2.4.2. Xanthan gum fermentation using RFBB

In 1996, Yang et al. first applied this concept to overcome the oxygen limitation problem in xanthan gum fermentation and design a rotating fibrous bed bioreactor (RFBB), it was also called a Centrifugal Packing Bed Reactor (CPBR). This design has been studied in two modes, one is centrifugal packing bed reactor under liquid continuous mode (CPBR-

LC), and the other is centrifugal packing bed reactor under gas continuous mode (CPBR-

GC). In a CPBR-LC, fibrous bed was totally embedded in the liquid medium, while in a

CPBR-GC, it was partially embedded in the broth. By using CPBR-GC, a cell-free xanthan gum broth could be produced in a volumetric productivity of ~1 g/L/h, which was higher than 0.5 g/L/h in stirred tanks (Yang et al., 1996). A higher cell density was obtained in CPBR-LC (6.8 g/L) and CPBR-GC (15.36 g/L), comparing to in STR (~2 g/L). However, the viability was only 60 %, which might due to medium change between batch runs and the insufficient oxygen transfer rate with such a high cell density.

Therefore, to increase oxygen transfer rate by increasing the rotational rate and recirculational rate could help improve the performance of CPBR. In addition, because cells were immobilized, a series of repeated batch fermentation could be conducted. The

20 time and labor in reactor cleaning and inoculation, in stirred tank batch fermentation, could be eliminated.

A similar design to RFBB was reported (Zhu et al., 2006). It was a modified stirred tank with an external loop. The external loop started after 24 h inoculation. It pumped out the broth from the bottom of the fermentor and circulated the broth back to the top. This design could increase final product concentration from 22.4 g/L to 24.3 g/L and decrease fermentation time from 48h to 42 h, using corn starch, corn steep liquor and sodium glutamate as substrates (total sugar 32 g/L). It was equivalent to overall productivity of

0.75 g/L/h. However, the increasing final concentration was not as high as the performance of RFBB (> 35 g/L). The authors claimed the external loop could eliminate the dead spaces of the fermentor. However, they found that the external loop did not increase the dissolved oxygen in the broth. Furthermore, the dissolved oxygen in the experiment with external loop was lower than without it.

Rosalam et al. (2008) developed a continuous recycled packed fibrous-bed bioreactor- membrane to produce cell free xanthan gum. The setup was similar to the concept of fibrous bed bioreactor developed in our lab. In addition, it was integrated with a dead-end microfiltration to further eliminate the cells in the broth. In this design, the matrix for cell attachment, cotton, was packed in a glass tube (470 mm in length and 22 mm in diameter). Cells were immobilized on cotton, while medium and gas passed through the fibrous bed from bottom to the top. The liquid flow rate controlled at the dilution rates of

1.44, 2.88, and 4.32 per day, and the gas flow rate was 30 mL/min, which was equilibrant to 0.17 vvm of fibrous bed volume. The maximum xanthan gum concentration obtained from this design was 20 g/L, which was slightly higher than the 18 g/L of free cell stirred 21 tank fermentation. This result was not as impressive as the final concentration of 35 g/L in our RFBB design, which could be due to the limited oxygen supply in their fibrous bed. Rosalam et al. recovered cell free xanthan gum through dead-end microfiltration.

However, the daily replacement of membrane could be costly in a production scale.

2.4.3. Oxygen transfer in RFBB

* The oxygen transfer rate (OTR) can be defined by OTR= kLa(CO2 -CO2), where kLa is

* oxygen transfer coefficient, CO2 is saturated oxygen concentration, and CO2 is oxygen concentration in solution. OTR depends on reactor design, liquid properties, and operation conditions, such as air flow rate and stirrer speed. With the success of xanthan gum production in CPBR, (kLa) in a CPBR was then compared with the kLa in stirred tanks using different impellers and water-in-oil system (Lo et al., 2001). Although the

-1 water-in-oil system in stirred tank reactor had higher kLa (0.020 s at 3.5 % w/v xanthan gum) than in a CPBR (0.018 s-1 at 3.5 % w/v xanthan gum), the recovery of product was difficult in the water-in-oil system, which limited its scalability (Lo et al., 2001). The gas continuous mode was designed to provide extensive gas-liquid contact, and the higher kLa in CPBR-GC than in CPBR-LC confirmed the more effective of CPBR-GC in terms of oxygen transfer. The authors also evaluated the oxygen transfer with and without recirculation loop in CPBR-LC. It shows that the recirculation loop was more effective at low rotational rate (100-300 rpm) at low xanthan concentration (1.2 and 2.2 %), and at high xanthan concentration (3.5 %). The value of recirculation loop is more significant when the agitation is poor. For a CPBR-GC, the recirculation loop is critical for the operation.

22 Up to this point, the CPBR study was in a 5 L modified reactor. Although a cell free xanthan gum broth could be produced in high productivity, yield and final concentration, the kLa based on these fermentation conditions may not reach the maximum. The rotational rate was limited by the severe vibration over 500 rpm, and the liquid flow rate was limited by the capacity of available pump. In addition, in the gas continuous mode, the fibrous bed was still partially embedded in the broth, not truly gas continuous.

Therefore, we proposed to use a larger reactor with a circulation tank and a higher capacity pump (Figure 2.8). In this design, fibrous bed was equipped in an empty vessel while the broth was collected in another vessel. In this way, the fibrous bed was more exposed to the air. The recirculation loop provided the contact of the broth to cells and air. Using this setup, Doma (1999) studied the kLa based on the operation parameters and proposed a model of kLa in a RFBB in terms of gas flow rate (VG), liquid flow rate (VL), rotational rate (ω), and liquid volume (L).

~0.4 0.109 2.4 -1 kLa  (ω )( VG )(VL )(L ) (2.2)

2.4.4. Mean residence time and liquid holdup in RFBB

Mean residence time based on different operational parameters was also investigated to calculate the liquid hold up in the fibrous bed. Residence time distribution (RTD) can be used to understand complex flow patterns in non ideal reactors. It is useful to clarify the mixing and mass transfer limitation in RFBB. RTD study in RFBB was carried out at 25

°C by adding 3 ml of 2.5% (w/w) KCl tracer solution into the inlet of the rotor and measuring the conductivity of the tracer solution at the exit of the rotor via a platinum-

23 coated dip cell. The cell is linked to a digital conductivity meter. The mean residence time distribution is calculated by

tE(t) t   E(t)  . (2.3)

 t     It can be normalized by a dimensionless transformation of t into θ  t  and E(t) into

 tE(t) E(θ)   . With data of mean residence time, liquid hold up can be estimated by

Lt  100 L 60V bed , (2.4) where εL is liquid hold up (%), L is volumetric flow rate of liquid (l/min), t is mean residence time (s) and Vbed is volume of packed bed (L). According to Doma (1999), it shows that at higher gas flow rate and rotational rate, less liquid hold up resulted in a drier fibrous bed during operation, which means the liquid passed through fibrous bed in droplets instead of as liquid film. In sum, RFBB can provide high oxygen transfer rate, and the advantages of scale up feasibility determine the value of RFBB used in industry.

2.4.5. Cell adsorption in RFBB

One of the advantages of RFBB is to absorb cells in a fibrous bed; hence the broth is almost cell free. This cell free broth reduces the difficulty of the downstream purification process, because it is hard to separate cells from highly viscous biopolymer broth by conventional industrial separation techniques, such as microfiltration, flocculation, and centrifugation. Moreover, by immobilizing the cells on the matrix, a series of repeated batch fermentations can be achieved. The repeated batches give high biopolymer

24 productivity, because it can reduce the long period (about 24 h) for cell growth in conventional batch fermentation. Issues that need to be addressed are that those cells may aggregate on the matrix to further reduce mass transfer efficiency, and that dead cells may not properly leave the fermentor. Therefore, understanding the factors of cell attachment and detachment will be helpful for selecting viable cells from dead cells during repeated-fed-batch fermentation. Cell adsorption can be classified by specificity: specific or nonspecific adsorption. The two major factors in selecting matrix for cell adsorption are the adsorption capacity and strength of binding. The binding force between cells and matrix depends on the surface properties of the matrix and cell types.

Generally, the cell surface is negatively charged. The dominant forces are different between the cell surfaces and different charged surfaces. For example, cells can adhere on a positively charged surface by electrostatic force, on a negatively charged surface by covalent binding or H bonding, and on neutral surface by covalent binding, H bonding, or van der Waals force. Covalent binding to immobilize cells is not desired, because it requires additional surface treatment on matrix. In order to provide a better mass transfer, a higher agitation rate is preferred. In this case, weak binding forces in cell adsorption can result in the removal of cells from the matrix by hydrodynamic shear. It is necessary to determine the upper limit shear stress in a RFBB. According to Daniels (1980), nonspecific cell adsorption is affected by the character of microorganisms, adsorbent, and environment factors. The microorganism effects are based on species, culture age, and concentration. The adsorbent effects depend on hydrophobicity, surface roughness, ionic form, charge, and size The environment factors are culture medium, pH, inorganic salt concentration, organic compounds, liquid flow rate, contact time, and temperature. In the 25 previous study, almost all suspended X. campestris could be absorbed onto a cotton towel especially when xanthan gum existed (Yang et al., 1998). Different types of materials were tested, such as cotton towel, cotton fabric, and 50 % cotton and 50 % polyester. The result showed that the cotton towel had the best performance on cell adsorption, which may be due to the hydrophilic surface property and the surface roughness. Then, Hsu et al. (2004) studied the effect of fibrous surface on cell immobilization in fibrous bed.

Despite the effort put on modifying the fibrous surface by creating roughness and hairiness surface, untreated cotton towel still showed the best immobilization efficiency.

This also confirmed the result in Yang et al., 1998 that cotton towel was the best material for fibrous bed so far.

2.4.6. Other research on RFBB

A study of intracellular biosynthesis of xanthan gum in CPBR-GC and CPBR-LC by metabolic flux analysis was reported by Hsu and Lo in 2003. The results showed that although gas continuous mode was more efficient in terms of kLa study, in the microbial energy-efficiency study, 5.0 % (w/v) glucose in a CPBR-LC gave the highest energy- efficiency among all operation modes, including stirred tank reactors and CPBR-GC (Hsu and Lo, 2003). The better production by using higher carbon and nitrogen (C/N) ratio was due to the higher glucose uptake rate. The authors speculated that the configuration of

CPBR provided higher oxygen transfer efficiency, which may cause the transformation of

Glc-6-P to sugar nucleotides and then to form xanthan gum.

26 2.4.7. Other oxygen transfer study in xanthan gum fermentation

Oxygen transfer can be increased by improving oxygen transfer coefficient, kLa, or increasing dissolved oxygen. Because kLa is a function of reactor design and operational condition, it is necessary to understand the correlation of kLa. In 1995, García-Ochoa et

- al. reported an empirical correlation for oxygen transfer in stirred tank: kLa = 3.08×10

4 0.43 1.75 -0.39 Vs N μ , where Vs is the air flow rate, N is the rotational speed of impeller, and μ is the apparent viscosity. In 2000, they further integrated oxygen transfer and uptake rate to develop model for scaling up xanthan gum production. To solve oxygen transfer problem, Rocha-Valadez et al. (2007) evaluated hydrodynamic and oxygen tension effects separately for rheological-complex fermentation, such as xanthan production. In this study, power drawn and dissolved oxygen tension were controlled, instead of manipulating agitation rate. Therefore, aeration (oxygen limitation) can be studied exclusively from agitation. Brahmbhatt (2007) introduced pure oxygen through multiple places along the height of the reactor in viscous product fermentation. In both researches, pure oxygen was injected into fermentors to meet the oxygen demand.

2.5. Ultrafiltration in biopolymer separation

Recovery xanthan gum from highly viscous broth containing 10-30 g/L xanthan gum, 1-

10 g/L cells, and 3-10 g/L residual nutrients is difficult (Garcia-Ochoa et al., 2000).

Generally, broth is sterilized to kill cells by thermal treatment under 80-130 °C for10-20 min with pH 6.3-6.9 (Smith and Pace, 1982) instead of removing cells. Then, the heat treated broth was added, 2 to 3 times volume of alcohol (eg., ethanol or isopropanol) to precipitate xanthan gum. Due to the low concentration (2 to 3 % w/v) of xanthan gum

27 after fermentation, the volume of alcohol required is massive. Ultrafiltration has been proposed to concentrate xanthan gum broth before alcohol precipitation, because it is well established and widely applied in protein concentration and waste stream clean-up.

By using a tubular polysulfone hollow fiber membrane with 500,000 molecular weight cut off (MWCO), xanthan gum could be concentrated from ~2.5 % to 13.5 % or higher with a recovery yield larger than 95 % (Yang et al., 1996). Generally, process broth with cells could block ultrafiltration membrane and cause membrane fouling. By using the cell free broth from a RFBB can prevent blocking in the ultrafiltration, which makes ultrafiltration separation more applicable in xanthan gum recovery. Ultrafiltration was also applied in γ-PGA recovery, which is a viscous microbial polypeptide (Do et al.,

2001).

2.6. Water reuse in fermentation

In a fermentation process, massive amounts of water is usually used once and then discharged. The wastewater contains residuals of nitrogen source, mineral salts, and metabolic wastes. Attempts to recycle fermentation broth have been investigated in several fermentation processes, such as intracellular penicillin amidase production. The recycled broth was obtained based on the process of primary product separation.

Normally, cells were removed via centrifugation or filtration, and the supernatant or permeate was then processed for recycle. By applying ultrafiltration in xanthan gum recovery, the permeate becomes a candidate for water reuse. The problems of spent medium reuse are the accumulated inhibitive compounds during the fermentation.

Possible inhibitors in the spent medium are acetic acid, unfermentable polysaccharides

28 and ions. The inhibition could be eliminated by diluting the spent broth with fresh medium. The replacement percentage could be determined by the inhibition of cell growth, productivity and product quality. To maximize the replacement percentage and the numbers of sequential batches of recycling were the main goal of water reuse studies.

The research related to water reuse was summarized in Table 2.5. The highest recycle ratio is in continuous ethanol production (Converti, A. et al., 1991). The inhibition was solved by enzymatic hydrolysis and ion removal. In batch fermentation, Hsiao, et al.

(1994) reported to produce intracellular lipid in 7 batches in series with 75 % of recycle water. To date yet no study on recycling water in biopolymer fermentation has been reported.

2.7. References

Afendra, Amalia S.; Yiannaki, Efthalia E.; Palaiomylitou, Maria A.; Kyriakidis, Dimitrios A.; Drainas, Constantin. Co-production of ice nuclei and xanthan gum by transformed Xanthomonas campestris grown in sugar beet molasses. Biotechnology Letters (2002), 24(7), 579-583.

Amanullah, A.; Satti, S.; Nienow, A. W. Enhancing xanthan fermentations by different modes of glucose feeding. Biotechnology Progress (1998c), 14(2), 265-269.

Argin-Soysal, S. and Lo, Y.M., 2004, Characterization of intermolecular interactions of xanthan biopolymers based on conformational changes evidenced by rheo-optical approaches, 2004 IFT Annual Meeting, July 12-16, Las Vegas, NV. 99C-10

Ashraf S; Soudi M R; Sadeghizadeh M Isolation of a novel mutated strain of Xanthomonas campestris for xanthan production using whey as the sole substrate. Pakistan journal of biological sciences: PJBS (2008), 11(3), 438-42

Bandalusena, H. C. Hemaka; Zimmerman, William B.; Rees, Julia M. Microfluidic rheometry of a polymer solution by micron resolution particle image velocimetry: a model validation study From Measurement Science and Technology (2009), 20(11), 115404/1-115404/9 29

Babu, P.S.R.. Panda, T. 1991. Effect of recycling of fermentation broth fro the production of penicillin amidase. Process Biochemistry, 26, 7-14.

Bogaert, J.C., Production and novel applications of natural L (+) lactic acid: Food pharmaceutics and biodegradable polymers. Cerevis (1997), 22, 46-50.

Brahmbhatt, Sudhir. Use of pure oxygen in viscous fermentation processes. U.S. Pat. Appl. Publ. (2007), 6pp

Cadmus MC, Knutson CA, Lagoda AA, Pittsley JE, Burton KA. Synthetic media for production of quality xanthan gum in 20 liter fermentors. Biotechnol Bioeng (1978), 20:1003-14

Casas, J.A., Mohedano, A.F., Garcia-Ochoa, F. Viscosity of guar gum and xanthan/guar gum mixture solutions. Journal of the Science of Food and Agriculture (2000), 80, 1722-1727.

Chang, Chun; Xu, Gui-zhuan; Li, Hong-liang; Chen, Jun-ying; Ma, Xiao-jian. Study on fermentation condition of Xanthomonas campestris QH79 for xanthan gum. Zhengzhou Gongye Daxue Xuebao (2000), 21(1), 92-94

Converti, A., Perego, P., Lodi, A., Fiorito, G., Borghi, M.D., Ferraiolo, G. 1991, In-situ ethanol recovery and subst4rate recycling during continuous alcohol fermentation. Bioprocess Engineering, 7, 3-10.

Cote, A.; Brown, W. A.; Cameron, D.; Van Walsum, G. P. Hydrolysis of lactose in whey permeate for subsequent fermentation to ethanol. Journal of Dairy Science (2004), 87(6), 1608-1620

Daniels, S. L. Mechanisms involved in sorption of microorganisms to solid surfaces. In Adsorption of microorganisms to surfaces; Bitton, G., Marshall, K. C., Eds.; John Wiley: New York, 1980, pp 7-58.

Daniels, M.J., Barber, C.E., Turner, P.C., Cleary, W.G.and Sawczyc, M.K., Isolation of mutants of Xanthomonas campestris pv. campestris Showing Altered Pathogenicity Journal of General Microbiology (1984) 130, 2447-2455

De Vuyst L, Vermeire A, Van Loo J, Vandamme EJ. Nutritional, physiological and process Ð technological improvements of xanthan fermentation process. Mec Fac Landbouwwet Rijkuniv Gent (1987a), 52: 1881-900.

De Vuyst L, Van Loo J, Vandamme EJ. Two-step fermentation process for improved xanthan production by Xanthomonas campestris NRRL B-1459. J Chem Technol Biotechnol (1987b), 39:263-73. 30

Doma, B. T., A rotating fibrous bed bioreactor for xanthan gum fermentation. (1999), Doctoral dissertation at University of the Philipines, Diliman, Quezon City, Philipines.

Do, Jin Hwan; Chang, Ho Nam; Lee, Sang Yup. Efficient recovery of γ -poly (glutamic acid) from highly viscous culture broth. Biotechnology and Bioengineering (2001), 76(3), 219-223.

Faria, Sandra; Vieira, Patricia Angelica; Resende, Miriam Maria; Ribeiro, Eloizio Julio; Cardoso, Vicelma Luiz. Application of a model using the phenomenological approach for prediction of growth and xanthan gum production with sugar cane broth in a batch process. LWT--Food Science and Technology (2010), 43(3), 498- 506.

Funahashi, H,m Machara, M,m Taguchi, H., Yoshida, T. effect of glucose concentration on xanthan gum production by Xanthomonas campestris. J. Chem. Eng. Jpn. (1987a), 65, 603-606

Garcia-Ochoa, F.; Santos, V. E.; Alcon, A. Xanthan gum production: an unstructured kinetic model. Enzyme Microbiol Technol, 1995, 17, 206-17.

Garcia-Ochoa, F.; Santos, V. E.; Alcon, A. Metabolic structured kinetic model for xanthan production. Enzyme and Microbial Technology (1998), 23(1/2), 75-82.

Garcia-Ochoa, F., Gomez Castro, E., Santos V.E., Oxygen transfer and uptake rates during xanthan gum production. Enzyme and Microbial Technology (2000), 27, 680-690.

Garcia-Ochoa, Felix; Santos, Victoria E.; Alcon, Almudena. Chemical structured kinetic model for xanthan production. Enzyme and Microbial Technology (2004), 35(4), 284-292.

Gonzalez Siso, M. I. The biotechnological utilization of cheese whey a review. Bioresource Technology (1996), 57(1), 1-11

Godbole SP, Shumpe A, Shah YT, Carr NL. Hydrodynamics and mass transfer in non- Newtonian solutions in a bubble column. AIChE J 1984;30:213–20.

Hamilton, Michelle A. Hamilton; Dawkins, Garth S.; Mellowes, Winston A. Xanthan gum production from sugarcane fluids. U.S. Pat. Appl. Publ. (2009), 5 pp

Hsiao, T.Y., Glatz, C.E., Glatz, B.A. 1994, Broth recycle in a yeast fermentation. Biotechnol. Bioeng. 44, 1228-1234.

31 Hsiao, T-Y. and Glatz, C.E., 1996, Water Reuse in the L-Lysine Fermentation Process, Biotechnology and Bioengineering, 49, 341-347.

Hsu, C. H.; Chu, Y. F.; Argin-Soysal, S.;Hahm, T. S.; Lo, Y. M. Effects of surface characteristics and xanthan polymers on the immobilization of Xanthomonas campestris to fibrous matrices. Journal of Food Science (2004), 69(9), E441-E448

Hsu, Chia-Hua; Lo, Y. Martin. Characterization of xanthan gum biosynthesis in a centrifugal, packed-bed reactor using metabolic flux analysis. Process Biochemistry (2003), 38, 1617-1625.

Janaun, J.; Wong, K. W.; Chu, C. M.; Prabhakar, A. Evaluation of xanthan production from Xanthomonas campestris using sago starch as carbon source. World Congress of Chemical Engineering, 7th, Glasgow, United Kingdom, July 10-14, 2005, (2005), 83515/1-83515/9

Ji, Lina; Mo, Xiaoyan; Zhan, Guyu. Effect of reduced pressure distillation on production rate and relative molecular weight of xanthan gum. Xi'an Jiaotong Daxue Xuebao (2001), 35(5), 549-550.

Kalogiannis, Stavros; Iakovidou, Gesthimani; Liakopoulou-Kyriakides, Maria; Kyriakidis, Dimitrios A.; Skaracis, George N. Optimization of xanthan gum production by Xanthomonas campestris grown in molasses. Process Biochemistry (Oxford, United Kingdom) (2003), 39(2), 249-256.

Kawase and Hashimoto, 1996 Y. Kawase and N. Hashimoto, Gas hold-up and oxygen transfer in three-phase external loop airlift bioreactors: non-Newtonian fermentation broths, Journal of Chemical Technology and Biotechnology 65 (1996), pp. 325–334

Kennedy JF, Jones P, Barker SA, Banks GT. Factors affecting microbial growth and polysaccharide production during the fermentation of Xanthomonas campestris cultures. Enzyme Microbiol Technol (1982), 4:39-43.

Kennedy J.F. and Bradshaw, I.J. Production properties and applications of xanthan. Prog. Ind. Microbiol (1984), 19, 319-71

Kongruang, Sasithom. Quantitative analysis of exopolysaccharide production in A stirred tank bioreactor. AIChE Annual Meeting, Conference Proceedings, Cincinnati, OH, United States, Oct. 30-Nov. 4, 2005 (2005), 438l/1-438l/15.

Konsoula, Zoe; Liakopoulou-Kyriakides, Maria; Perysinakis, Angelos; Chira, Panayiota; Afendra, Amalia; Drainas, Constantin; Kyriakidis, Dimitrios A. Heterologous expression of a hyperthermophilic α-amylase in xanthan gum producing

32 Xanthomonas campestris cells. Applied Biochemistry and Biotechnology (2008), 149(2), 99-108.

Lee, Byoung-Moo; Park, Young-Jin; Park, Dong-Suk; Kang, Hee-Wan; Kim, Jeong-Gu; Song, Eun-Sung; Park, In-Cheol; Yoon, Ung-Han; Hahn, Jang-Ho; Koo, Bon- Sung; Lee, Gil-Bok; Kim, Hyungtae; Park, Hyun-Seok; Yoon, Kyong-Oh; Kim, Jeong-Hyun; Jung, Chol-hee; Koh, Nae-Hyung; Seo, Jeong-Sun; Go, Seung-Joo. The genome sequence of Xanthomonas oryzae pathovar oryzae KACC10331, the bacterial blight pathogen of rice. Nucleic Acids Research (2005), 33(2), 577-586

Lamboley, L., Lacroix, C., Chmpagne, C.P., Vuillemard, J.C., Continuous mixed strain mesophilic starter production in supplemented whey permeate medium using immobilized cell technology. (1997), Biotechnol. Bioeng. 56, 502-516

Liakopoulou-Kyriakides, M.; Psomas, S. K.; Kyriakidis, D. A. Xanthan gum production by Xanthomonas campestris w.t. fermentation from chestnut extract. Applied Biochemistry and Biotechnology (1999), 82(3), 175-183.

Lo, Y.M., A Novel immobilized cell bioreactor and membrane ultrafiltration for xanthan gum production. (1996), Doctoral dissertation at the Ohio State University

Lo, Y. M.; Hsu, C. H.; Yang, S. T.; Min, D. B. Oxygen transfer characteristics of a centrifugal, packed-bed reactor during viscous xanthan fermentation. Bioprocess and Biosystems Engineering (2001), 24(3), 187-193

Lopez, M. J.; Moreno, J.; Ramos-Cormenzana, A. The effect of olive mill wastewaters variability on xanthan production. Journal of Applied Microbiology (2001), 90(5), 829-835

Maldonado, A., Effect of medium composition on xanthan gum production from whey permeate using Xanthomonas campestris NRRL B-1459 (1992), Master thesis at the Ohio State University

Margaritis, A. and G. W.Pace. Microbial polysaccharides. In Comprehensive biothenology vol. 3/ Moo-Young, M. (Ed.), (1985), Pergamon Press, new Yourk, NY; pp. 1005-1043

Medvedeva, E.I., Oleinik, T.P., Panchenko, K.A., Petrenko, E. B., 1989. Multipurpose use of liquid wastes in microbial lysine synthesis technology. Biotekhnologiya, 6, 761-767

Mesomo, Michele; Silva, Marceli Fernandes; Boni, Gabriela; Padilha, Francine Ferreira; Mazutti, Marcio; Mossi, Altemir; de Oliveira, Debora; Cansian, Rogerio Luis; Di Luccio, Marco; Treichel, Helen. Xanthan gum produced by Xanthomonas campestris from cheese whey: production optimisation and rheological 33 characterization. Journal of the Science of Food and Agriculture (2009), 89(14), 2440-2445

Milas, M.; Rinaudo, M.; Tinland, B. The viscosity dependence on concentration, molecular weight and shear rate of xanthan solutions. Polymer Bulletin (Berlin, Germany) (1985), 14(2), 157-64

Moraine RA, Rogovin P. Kinetics of polysaccharide B-1459 fermentation. Biotechnol Bioeng (1966), 8:511-24

Moraine RA, Rogovin P. Xanthan biopolymer production at increased concentration by pH control. Biotechnol Bioeng (1971),13:381-91.

Moraine RA, Rogovin P. Kinetics of the xanthan fermentation. Biotechnol Bioeng (1973), 15:225-37.

Moreno, J.; Lopez, M. J.; Vargas-Garcia, C.; Vazquez, R. Use of agricultural wastes for xanthan production by Xanthomonas campestris. Journal of Industrial Microbiology & Biotechnology (1998), 21(4/5), 242-246.

Nery, Tatiana Barreto Rocha; Brandao, Lillian Vasconcellos; Esperidiao, Maria Cecilia Azevedo; Druzian, Janice Izabel. Biosynthesis of xanthan gum from the fermentation of milk whey: productivity and viscosity. Quimica Nova (2008), 31(8), 1937-1941

Nicolas, G., Auger, I., Beaudoin, M., Halle, F., Morency, H,m LaPointe, G., Lavoie, M.C., Improved methods for mutacin detection and production. J. Microbiol. Methods, (2004), 59, 351-361.

Nykanen, A., Lapvetelainen, A., Hietnen, R.M., Kallio, H., The effect of lactic acid, nisin, whey permeate, sodium chloride and related combinations on aerobic plate count and the sensory characteristics of rainbow trout. Lebensm. Wiss. Technol., (1998), 31, 286-290.

Ochiai, Hirokazu; Inoue, Yasuhiro; Takeya, Masaru; Sasaki, Aeni; Kaku, Hisatoshi. Genome sequence of Xanthomonas oryzae pv. oryzae suggests contribution of large numbers of effector genes and insertion sequences to its race diversity. JARQ (2005), 39(4), 275-287

Papoutsopoulou, S. V.; Ekateriniadou, L. V.; Kyriakidis, D. A. Genetic construction of Xanthomonas campestris and xanthan gum production from whey. Biotechnology Letters (1994), 16(12), 1235-40

34 Peters H-U, Herbst H, Hesselink PGM, LuÈndsdorf H, Schumpe A, Deckwer W-D. The influence of agitation rate on xanthan production by Xanthomonas campestris. Biotechnol Bioeng (1989), 34:1393-7.

Pinches A, Pallent LJ. Rate and yield relationships in the production of xanthan gum by batch fermentations using complex and chemically defined growth media. Biotechnol Bioeng (1986), 28:1484-96.

Pons A, Dussap CG, Gros JB. Xanthan batch fermentations: compared performance of a bubble column and a stirred tank fermentor. Bioprocess Eng (1990), 5:107-14.

Psomas, S.K., Liakopoulou-Kyriakides, M., and Kyriakidis, D.A., Optimization study of xanthan gum production using response surface methodology. Biochemical Engineering Journal (2007)35, 273-280.

Qian, Wei; Jia, Yantao; Ren, Shuang-Xi; He, Yong-Qiang; Feng, Jia-Xun; Lu, Ling-Feng; Sun, Qihong; Ying, Ge; Tang, Dong-Jie; Tang, Hua; Wu, Wei; Hao, Pei; Wang, Lifeng; Jiang, Bo-Le; Zeng, Shenyan; Gu, Wen-Yi; Lu, Gang; Rong, Li; Tian, Yingchuan; Yao, Zhijian; Fu, Gang; Chen, Baoshan; Fang, Rongxiang; Qiang, Boqin; Chen, Zhu; Zhao, Guo-Ping; Tang, Ji-Liang; He, Chaozu. Comparative and functional genomic analyses of the pathogenicity of phytopathogen Xanthomonas campestris pv. campestris. Genome Research (2005), 15(6), 757- 767

Rao, D. P.; Bhowal, A.; Goswami, P. S. Process Intensification in Rotating Packed Beds (HIGEE): An Appraisal Ind. Eng. Chem. Res. 2004 43 1150

Robinson, D. K. and Wang, D.I. C. A transport controlled bioreactor for the simultaneous production and concentration of xanthan gum. Biotechnol. Prog. (1988), 4, 231- 241

Rocha-Valadez, J. A.; Albiter, V.; Caro, M. A.; Serrano-Carreon, L.; Galindo, E. A fermentation system designed to independently evaluate mixing and/or oxygen tension effects in microbial processes: development, application and performance. Bioprocess and Biosystems Engineering, (2007), 30(2), 115-122

Rogovin P, Albrecht W, Sohns V. Production of industrial grade polysaccharide B-1459. Biotechnol Bioeng (1965), 7:161-9.

Rosalam, S. and England, R., Review of xanthan gum production from unmodified starches by Xanthomonas campestris sp. Enzyme and Microbial Technology, (2006), 39(2), 197-207

35 Rosalam, S., Krishnaiah, D. and Bono, A. Cell free xanthan gum production using continuous recycled packed fibrous-bed bioreactor-membrane. Malaysian Journal of Microbiology, (2008), 4(1), pp 1-5

Roukas, T. and F. Mantzouridou, 2001. Effect of the aeration rate on pullulan production and fermentation broth rheological properties in an airlift reactor. J. Chem. Technol. Biotechnol., 76: 371-376.

Schepers, A.W., Thibault, J., Lacroix, C., Continuous lactic acid production in whey permeate/yeast extract medium with immobilized Lactobacillus helveticus in a two-stage process: model and experiments. Enzyme Microb. Technol., (2006), 38, 324-337.

Schwartz, Robert D.; Bodie, Elizabeth A. Production of high-viscosity whey broths by a lactose - utilizing Xanthomonas campestris strain. Applied and Environmental Microbiology (1985), 50(6), 1483-5

Serrania, Javier; Vorhoelter, Frank-Joerg; Niehaus, Karsten; Puehler, Alfred; Becker, Anke. Identification of Xanthomonas campestris pv. campestris galactose utilization genes from transcriptome data. Journal of Biotechnology, (2008), 135(3), 309-317

Shu Ch-H, Yang Sh-T. Effects of temperature on cell growth and xanthan production in batch cultures of Xanthomonas campestris. Biotechnol Bioeng (1990), 35:454-68.

Shuler, M. L., Kargi, F. Bioprocess Engineering basic concepts, second edition, (1992), pp 286

Silva, M.F., Fornari, R.C.G., Mazutti, M.A., Oliveira, D., Padilha, F.F., Cichoski, A.J., Cansian, R.L., Luccio, M.D., Treichel, H.. Production and characterization of xanthan gum by Xanthomonas campestris using cheese whey as sole carbon source. Journal of Food Engineering (2009), 90, 119-123.

Silveira, W.B., Passos, F.J.V., Mantovani, H.C., Passos, F.M.L., Ethanol production from cheese whey permeate by Kluyveromyces Marxianus UFV-3: a flux analysis of oxido-reductive metabolism as a function of lactose concentration and oxygen levels. Enzyme Microb. Technol. (2005), 36, 930-936.

Smith, Ian H.; Pace, Gary W. Recovery of microbial polysaccharides. Journal of Chemical Technology and Biotechnology (1979-1982) (1982), 32(1), 119-29. Herbst, Holger; Schumpe, Adrian; Deckwer, Wolf Dieter. Xanthan production in stirred-tank fermentors: oxygen transfer and scale-up. Chemical Engineering & Technology (1992), 15(6), 425-34

36 Song, K-W., Kim, Y-S., and Chang, G-S., Rheology of concentrated xanthan gum solutions: steady shear flow behavior, Fibers and Polymers, (2006), 7(2), 129-138.

Souw P, Demain AL. Role of citrate in xanthan production by Xanthomonas campestris. J Ferment Technol (1980), 58:411-6.

Stanbury PF, Whitaker A, Hall SJ. 1995. Principles of fermentation technology, 2nd ed. New York: Pergamon. p 243–276.

Stokke, Bjoern T.; Elgsaeter, Arnljot; Smidsroed, Olav. Electron microscopic study of single - and double -stranded xanthan. International Journal of Biological Macromolecules (1986), 8(4), 217-25

Thacker, A., Fu, S., Boni, R.L., and Block, L.H., Inter- and intra-manufacturer variability in pharmaceutical grades and lots of xanthan gum. AAPS PharmSciTech, (2010) 11, 4, pp1619-1626.

Thieme, Frank; Koebnik, Ralf; Bekel, Thomas; Berger, Carolin; Boch, Jens; Buettner, Daniela; Caldana, Camila; Gaigalat, Lars; Goesmann, Alexander; Kay, Sabine; Kirchner, Oliver; Lanz, Christa; Linke, Burkhard; McHardy, Alice C.; Meyer, Folker; Mittenhuber, Gerhard; Nies, Dietrich H.; Niesbach-Kloesgen, Ulla; Patschkowski, Thomas; Rueckert, Christian; Rupp, Oliver; Schneiker, Susanne; Schuster, Stephan C.; Vorhoelter, Frank-Joerg; Weber, Ernst; Puehler, Alfred; Bonas, Ulla; Bartels, Daniela; Kaiser, Olaf. Insights into genome plasticity and pathogenicity of the plant pathogenic bacterium Xanthomonas campestris pv. vesicatoria revealed by the complete genome sequence. Journal of Bacteriology (2005), 187(21), 7254-7266

Thorne, Linda; Tansey, Lisa; Pollock, Thomas J. Direct utilization of lactose in clarified cheese whey for xanthan gum synthesis by Xanthomonas campestris. Journal of Industrial Microbiology (1988), 3(5), 321-8

Tinland, B., Maret, G., and Rinaudo, M., Reptation of semidilute solutions of wormlike polymers. Macromolecules (1990), 23, 596-602.

Tiwari, B.K., Muthukumarappan, K., O’Donnell, C.P., Cullen, P.J., Rheological properties of sonicated guar, xanthan and pectin dispersions. International Journal of Food Properties (2010) 13, 223-233.

Tyagi, R.D., Kluepfe, D., Bioconversion of cheese whey to organic acids. In: Martin, A.M. (Ed.), Bioconversion of Waste Materials to Industrial Products. Blackie Academic, Glasgow, UK. (1998)

Van’t Riet, K. and Tramper, J., 1991, Basic Bioreactor Design (New York: Marcel Dekker Inc.). 37

Vojnov, Adrian A.; Slater, Holly; Daniels, Michael J.; Dow, J. Maxwell. Expression of the gum operon directing xanthan biosynthesis in Xanthomonas campestris and its regulation in planta. Molecular Plant-Microbe Interactions (2001), 14(6), 768-774

Vorhölter, Frank-Joerg; Schneiker, Susanne; Goesmann, Alexander; Krause, Lutz; Bekel, Thomas; Kaiser, Olaf; Linke, Burkhard; Patschkowski, Thomas; Rueckert, Christian; Schmid, Joachim; Sidhu, Vishaldeep Kaur; Sieber, Volker; Tauch, Andreas; Watt, Steven Alexander; Weisshaar, Bernd; Becker, Anke; Niehaus, Karsten; Puehler, Alfred. The genome of Xanthomonas campestris pv. campestris B100 and its use for the reconstruction of metabolic pathways involved in xanthan biosynthesis. Journal of Biotechnology (2008), 134(1-2), 33-45.

Walsh, Patricia M.; Haas, Michael J.; Somkuti, George A. Genetic construction of lactose - utilizing Xanthomonas campestris. Applied and Environmental Microbiology (1984), 47(2), 253-7.

Weiss, Rebecca M.; Ollis, David F. Extracellular microbial polysaccharides. I. Substrate, biomass, and product kinetic equations for batch xanthan gum fermentation. Biotechnology and Bioengineering (1980), 22(4), 859-73.

Wyatt, Nicholas B. and Liberatore, Matthew W. Rheology and viscosity scaling of the polyelectrolyte xanthan gum. Journal of Applied Polymer Science. (2009), 114, 4076-4084.

Yang, Shang-Tian; Lo, Yang-Ming; Chattopadhyay, Devamita. Production of cell-free xanthan fermentation broth by cell adsorption on fibers. Biotechnology Progress (1998), 14(2), 259-264.

Yang, Shang-Tian; Lo, Yang-Ming; Min, David B. Xanthan gum fermentation by Xanthomonas campestris immobilized in a novel centrifugal fibrous- bed bioreactor. Biotechnology Progress (1996), 12(5), 630-637

Yang, T.C., Wu, G.H., and Tseng, Y.H., Isolation of a Xanthomonas campestris strain with elevated β-galactosidase activity for direct use of lactose in xanthan gum production. Letters in Applied Microbiology. (2002), 35, 375-379.

Yang, T.C., Hu, R.M., Hsiao, Y.M., Weng, S.F., Tseng, Y.H., Molecular genetic analyses of potential β-galactosidase genes in Xanthomonas campestris. J. Mol. Microbiol. Biotechnol. (2003), 6, 145-154.

Yano, T., Mori, H., and Kobayashi, T., 1980, Reusability of broth supernatant as medium. J. Ferment. Technol., 58, 3, 259-266

38 Zhang, Zhiguo; Chen, Hongzhang. Fermentation performance and structure characteristics of xanthan produced by Xanthomonas campestris with a glucose/xylose mixture. Applied Biochemistry and Biotechnology (2010), 160(6), 1653-1663.

Zhao S, Kuttuva SG, Ju L-K (1999) Oxygen transfer characteristics of multiple-phase dispersions simulating water-in-oil xanthan fermentations. Bioprocess Eng 20:313–323

Zhu, Shengdong; Wu, Yuanxin; Yu, Ziniu; Tong, Haibao; Cheng, Dachang; Xie, Decheng. Improving xanthan fermentation in a mechanically stirred aerated fermenter via external loop. Research Journal of Microbiology (2006), 1(1), 70-75.

Zirnsak, M.A., Boger, D.V., Tirtaatmadja, V. Steady shear and dynamic rheological properties of xanthan gum solutions in viscous solvents. J. Rheol., (1999), 43(3), 627-650.

39

Fermentation techniques Process performance

Separation Reference techniques Product Volumetric Final product Microbial strains Substrates Bioreactor yield productivity concentration

50 wt % cheese 1.2 g/100 whey supply with 0.125 g/L/h Papoutsopoulou et X. campestris XLM 1521 2L flasks -- mL Cheese 12 g/L 1.5% lactose or (96 h) al., 1994 whey glucose

X. campestris mutant X. 10 % v/v whey 2 L Biostat B 72 h 10 g/L Ashraf, et al., 2008 campestris b82

X. campestris pv Centrifugation 250 mL flasks (80 mangiferaeindicae 1230 and followed by 80 mL cheese whey mL working -- 0.34 g/L/h 25 g/L Silva et al., 2009 X. campestris pv manihotis ethanol 40 volume) 1182 precipitation Whey from 80 mL medium in isolated X. campestris 120 h 21.91 g/L Nery et al., 2008 mozzarella cheese 250 mL flasks Recombinant X. campestris Whey lactose (2 % 200 mL shake ------24 h 1.8 % (w/w) Thorne, et al., 1988 X59-1232 w/v) flasks Centrifugation X. campestris pv. 900 mL cheese followed by 36 g/L Mesomo et al., mangiferaeindicae IBSBF 2.5 L Biostat-B N/A 0.5 g/L/h whey ethanol (1831.34 cp) 2009 1230 precipitation Lactose based Isolated Xc17L Flask ------1.852 g/L Yang, et al., 2002 medium Table 2.1. Literature reviews of xanthan gum production from whey

40

Fermentation techniques Process performance Separation Microbial Product Volumetric Final product Reference Substrates Bioreactor techniques strains yield productivity concentration

Controlled the glucose 35 g/L Robinson and concentration between Wang, 1988 30 and 40 g/L Immobilized 130 h 55 g/L pore Robinson and cells volume Wang, 1988 X. campestris Fed batc, initial 40 g/L 20 L stirred 0.82 0.65 62 g/L Amanullah, et al., NRRL B-1459 glucose, followed by tank (14 L 1998 E2 continuous addition from working 41 30 to 82 h at a rate of 1.3 volume)

g/L/h

X. campestris 15 g/L sucrose 50 L -- 1.8 g/g 0.375 g/L/h 27 g/L (8000 Chang et al., 2000 QH79 fermentor (72h) cp) X. campestris 40 g/L sucrose, 2.3 g/L 14 L stirred -- 0.42 0.36 g/L/h 26 g/L Rocha-Valadez et NRRL-B1459 citric acid monohydrate, tank (10L g/g (72h) al., 2007 after 24h of culture time, working 20g/L sucrose was added volume) to the culture medium X. campestris 10 g/L glucose 2 L LSL Isopropanol 0.7 g/g 0.097 g/L/h 7 g/L Psomas et al., 2007 ATCC 33913 Biolite repeated (72h) Bioreactor precipitation (working volume: 1L) Table 2.2. Literature reviews of xanthan gum production from glucose or sucrose

41 Fermentation techniques Separation techniques Process performance

Microbial Substrates Bioreactor Product yield Volumetric Final product Reference strains productivity concentration

X. campestris 20 % (v/v) Flasks Isopropanol precipitation added 1 -- 0.058 g/L/h 7 g/L Lopez et al., 2001 NRRL B1459 Olive mill % KCl (120 h) S4LII wastewaters

X. campestris 40 g/L corn Flask Reduce pressure distillation 0.748 g/g 0.415 g/L/h 29.9 g/L Ji et al., 2001 J-12 starch before alcohol precipitation (72h)

X. campestris 175 g/L 500 mL flask -- 0.3028 g/g 2.2 g/L/h 53 g/L Kalogiannis et al., ATCC 1395 molasses, 4 g/L (100 mL (24 h) 2003 K2HPO4, working volume)

42 X. campestris Coconut juice 3.7 L -- 6.407 g/g/ 0.665 g/L/h 23.29 g/L Kongruang et al., 2005

TISTR 1100 fermentor (35 h) 0.883 (877.5 cp) (2.7 L g/g/h working volume) X. campestris 9 L pH Stirred tank -- 0.2 to 0.36 48 to 168 h 20 to 36 g/L Hamilton et al., 2009 NRRL B- adjusted bioreactor g/g (600 to 1600 1459 sugarcane fluid cp) X. campestris Glucose/xylose Flask Adjust pH to 3, then diluted for 0.33 to 0.61 0.21 to 0.04 11.69 to 18.16 Zhang and Chen, 2010 CGMCC mixture in centrifugation followed by gXG/g g/L/h g/L (lower than 1.1781 different ratio alcohol precipitation glucose;0.09 20 cp) to 0.45gXG/g total sugar X. campestris 40 g/L sucrose 4 L Biostat-B Broth was diluted 1:1 with DI 0.35 0.57 14 g/L Faria et al.,2010 pv. from sugar water, and then centrifuged and campestris cane broth filtered to remove cells. before NRRL B- ethanol precipitation, saturated 1459 KCl solution was added Table 2.3. Literature reviews of xanthan gum production from other substrates

42 Bioreactor Operational conditions Fermentation results Reference Temp- pH Volume Aeration N (rpm) Yield Time Max. erature (°C) (L) rate (L/L min) (% w/w) (h) conc. (g/L) Stirred 20-30 6.8 10 1.5 225-300 46 72-96 14 Cadmus et al. (1978) tank (controlled) 28 7 227 0.5 90-290 67 96 15 Rogovin et al. (1965) 2268 30-250 28 7 - 1 1000 70 96 15 Moraine and Rogovin (1966) 28 7, 7.1 8 1 500-1000 73 40 14.6 Moraine and (controlled using NH4OH) Rogovin (1971) 28 7 - - - 75 96 29 Moraine and (controlled) Rogovin (1973) 25 7 - 0.5 500 50 60 10.5 Souw and (controlled) Demain (1980) 30 7 10 0.4 600 76 45 17 Pinches and

(controlled) 2.5 1000 Pallent (1986)

43 28 7 6 1 250-700 75 144 27.9 De Vuyst et al.

(1987a,b 30 7 6 1 350-1200 66 96 30 Funahashi et al. (1987) 28 7 - 0.3 200-800 34 90 18.5 Peters et al. (1989) (controlled) 20-34 7 - 1.16 800 81 52 19 Shu and Yang (1990) (controlled) 29 6.9 3.6 0.3, 0.6 500-900 65 50 13 Pons et al. (1990) (controlled) 30 7 3.5 0.5 400-600 45 69 22.5 Kennedy et al. (1982) (controlled) Table 2.4. Literature review of bioreactor studies in xanthan gum fermentation (Expended from a review paper by Garchia-Ochoa et a., 2000)

43

Table 2.4 Continued

Bioreactor Operational conditions Fermentation results Temp- pH Reference Reference N (rpm) Yield Time Max. Reference erature (% (h) conc. (°C) w/w) (g/L) 26 7 Reference Reference 300-1300 50 96 12.5 Schweikart and (controlled) Quinlan (1989) 28 7 Reference Reference 210-1200 75 70 30 Garcia-Ochoa et al. (1997) 35 7 Reference Reference 600 0.71 72 7 Psomas et al. (2007)

Bubble 29 6.9 Reference Reference - 50 120 20 Pons et al. (1989) column (controlled) Airlift 28 7 Reference Reference 50 80 25 Suh et al. (1992) (controlled) 27 7 Reference Reference 45 49 (2nd 25 Kessler et al. (1993) 44 (controlled) step) Plugging 28 7 Reference Reference <50 100 18 Zaidi et al. (1991) jet reactor (controlled) Reference Reference

Continuous 30 7 Reference Reference - 180 20 Rosalam et al. (2008) recycled packed fibrous-bed bioreactor- membrane

44

Strain Product Recycle process Recycle Numbers of Comments Reference percentage Sequential batches N/A Intracellular N/A 40 to 60 % 3 Without inhibiting Babu et al., 1991 penicillin production amidase Saccharomyces Ethanol N/A Complete Continuous Inhibition was solved by Converti et al., 1991 cerevisiae recycle fermentation enzymatic hydrolysis and ion removal. Corynebacterium L-Lysine Cation 75 3 Defined medium; without Hsiao and Glatz, 1996 glutamicum exchange affecting cell mass and ATCC 21253 chromatography lysine production 50 N/A Complex medium containing beet molasses; 45 lower lysine production and higher cell mass production yeast Intracellular N/A 75 7 Defined medium; without Hsiao et al., 1994 lipid affecting cell mass and lysine production N/A L-lysine Ultrafiltration 10% 4 consecutive Without inhibition on Medvedeva et al., 1989 recycle batches lysine production E. coli N/A Centrifugation N/A N/A Affecting cell growth Yano et al., 1980

Table 2.5. Literature reviews of water recycle in fermentation (N/A: not available)

45

Figure 2.1. Chemical structure of xanthan gum (Garcia-Ochoa et al., 2000)

46

Figure 2.2. Conformations of xanthan gum in solution (Margaritis and Pace, 1985)

47

Figure 2.3. Effect of shear rate on the apparent viscosity of xanthan gum solution at 20ºC

(Bandalusena et al., 2009)

48

Figure 2.4. Metabolic pathway of xanthan gum biosynthesis in X. campestris (Hsu and

Lo, 2003).

49

Figure 2.5. Mechanism of xanthan gum synthesis in X. campestris. (Vorhölter et al.,

2008)

50

Fig.1. Conventional xanthan production process (Rosalam and England, 2006) Figure 2.6. Conventional xanthan gum production process (Rosalam and England, 2006)

51

Figure 2.7. A rotating fibrous bed bioreactor (arrows indicate liquid flow direction).

(Yang et al., 1996)

52 Controller Motor

Thermometer

pH probe

Air DO probe

Motor Temperature controller Recirculating RFBB vessel Peristaltic pumps

Figure 2.8. A Rotating fibrous bed bioreactor system (Doma, 1999)

53

Chapter 3: Xanthan gum fermentation using hydrolyzed whey permeate (HWP) as an alternative carbon source

3.1. Summary

The research of whey recycling in cheese industry has been studied for decades, yet has not reached a solution. Today, although the whey protein is extracted and sold as a value added product, the lactose rich whey permeate still causes environmental problems after disposal. Therefore, we proposed to convert whey permeate to a value-added polysaccharide, xanthan gum. Because of the low affinity of β-galactosidase in

Xanthomonas campestris, before using whey permeate, whey lactose should be hydrolyzed into glucose and galactose for the microorganism to consume. The results showed that whey permeate could be used as an alternative carbon source for xanthan gum production, yet the rheological properties of xanthan gum from whey permeate hydrolysate were not the same as from glucose. The final xanthan gum concentration, yield, and productivity were 32.16 g/L, 0.73 g xanthan/g carbon source, 0.41 g/L/h, respectively.

3.2. Introduction

Cheese is one of the favorite foods in the world. There are 48 million tons of cheese consumed in the U.S. every year. During production, after the removal of the firm curd, the liquid stream is separated as the byproduct whey. This byproduct contains about

54 4.5-5% lactose, 0.6-0.8% soluble proteins, 0.4-0.5% lipids and 8-10% mineral salts. In the past, most whey was disposed of directly and some was used as feed. The protein and lactose-rich liquid were not fully utilized. Due to the high biological oxygen demand

(BOD) (30,000 to 50,000 mg O2 per liter wastewater), whey disposal can cause severe environmental problems and high cost to cheese manufacturers. Today, whey is filtered to extract whey protein for food, nutrient supply or feed. However, whey permeate is still rich in lactose, which leaves the high BOD problem unsolved. The effluent disposal of whey costs cheese manufacturers about $307 per ton.

Researchers have investigated the application of lactose-rich whey and whey permeate for years (Taron et al., 1995;Yang et al., 2003). The digest systems of animal and human limit the application of lactose-rich whey permeate as food and feed. Therefore, researchers used whey or whey permeate as the carbon source in fermentation to produce ethanol, lactic acid, acetic acid, propionic acid, and single cell protein (Coté et al., 2004;

González Siso, 1996; Bogaert, 1997; Lamboley et al., 1997; Nicolas et al., 2004;

Nykanen et al., 1998; Schepers et al., 2006; Silveira et al., 2005; Tyagi and Kluepfel,

1998). Although, there are many attempts to use the lactose in whey and whey permeate to produce lactic acid, the demand of whey permeate for lactic acid production is much less than whey permeate generated. The exploration of more applications of excessive whey permeate is thus highly desired in the current cheese industry.

Xanthan gum is one potential product from whey permeate with increasing demand every year. Using whey permeate in xanthan gum production can not only largely reduce the cost of whey permeate disposal, but also make profits. Wild-type Xanthomonas

55 campestris strain does not consume lactose, because of its low galactosidase activity. To solve the problem, several attempts have been made: 1) to isolate X. campestris strains which could utilize lactose (Schwartz and Bodie, 1985; Yang et al., 2003; Ashraf, et al.,

2008; Mesomo et al., 2009); 2) to clone β-galactosidase gene or other genes for lactose utilization into X. campestris (Thorne, et a., 1988; Walsh et al., 1984); 3) to hydrolyze lactose before fermentation (Maldonado, 1992). There were some lactose-utilizing strains isolated and reported, but their productivity is not comparable to those using glucose or other easily utilized carbon sources (Yang et al., 2002 and 2003). In addition, isolated strains utilizing lactose as carbon source were not stable after certain generation

(Schwartz and Bodie, 1985). More information about xanthan gum production using whey and other substrates has been summarized in Tables 2.1, 2.2 and 2.3 (see Chapter

2).

In this research, lactose was hydrolyzed into glucose and galactose as fermentation substrates. Xanthomonas campestris can use glucose and galactose simultaneously to produce xanthan gum. The feasibility of using hydrolyzed whey permeate (HWP) as an alternative carbon source to produce xanthan gum was studied in terms of the fermentation efficiencies, such as growth kinetics, final concentration, yield, and productivity; and product properties, such as product viscosity, rheology (consistency index and flow index), intrinsic viscosity and average molecular weight. The results were compared to those using glucose, galactose and lactose hydrolysate.

56 3.3. Materials and methods

3.3.1. Culture and media

A wild type xanthan gum production strain, Xanthomonas campestris NRRL B1459, was used in this study. Unless otherwise noted, the medium contained 40 g/L carbon source, 3 g/L yeast extract, 2 g/L K2HPO4, and 0.1 g/L MgSO4∙7H2O in tap water at pH 6.5 with

500 ppm antifoaming agent (MURNC MCA250 or MCA 270, Murnc, Inc., Libertyville,

IL). Four different carbon sources, glucose, galactose, lactose hydrolysate and HWP were used. They were sterilized and added into the medium with a final concentration of 40 g/L before inoculation.

3.3.1.1. Glucose or galactose-based medium

Glucose or galactose was dissolved in tap water. Then, the volume and pH were adjusted to 500 mL and 6.5 before autoclave.

3.3.1.2. Lactose hydrolysate-based medium

400 g/L lactose solution was hydrolyzed with β-galactosidase from Aspergillus oryzae at pH 4.5, 40 ˚C. After the complete hydrolysis, the enzyme reaction was terminated by increasing the temperature to 70 ˚C. Then, the lactose hydrolysate was sterilized using autoclave. Before inoculation, 300 mL lactose hydrolysate was added into the fermentor to meet the final concentration of 40 g/L. The concentration of lactose, glucose and galactose was determined by using an HPLC (Refractive Index Detector, Shimadzu,

RID-10A) on an organic acid analysis column (BIO-RAD, Aminex® HPX-87H ion

57 exclusion column, 300 mm×7.8mm) at 45 ºC. The concentrations of glucose and galactose after hydrolysis are given in the Appendix.

3.3.1.3. HWP-based medium

40% whey permeate (Brewster Dairy Inc, Brewster, OH) was hydrolyzed by

β-galactosidase from Aspergillus oryzae at pH 6, room temperature overnight. For the concentrations of glucose and galactose after hydrolysis, refer to the Appendix. To prepare a 40 g/L HWP for fermentation, 300 ml HWP was clarified through a 1.0 μm microfiltration (Polycap 36 AS, aqueous solution filter capsule, sterile and non-pyrogenic with polypropylene housing, 400 cm2 EFA, 0.45μm pore size, nylon membrane, 1/4-3/8” stepped barbs, Waterman, Springfield Mill, UK) and then sterilized by autoclaving.

3.3.2. Stirred tank bioreactor

The reactor systems used in this study were B. Braun Biostat B, New Brunswick BioFLO

II and New Brunswick BioFLO3000. The working volume was 3 L. The production of xanthan gum was two-phase fermentation, with a temperature of 25 ˚C and pH of 6.5 in the first day after inoculation and 30˚C and pH 7.0 thereafter. The first stage was for cell growth and the second stage was for xanthan gum production.

3.3.3. Analytical methods

3.3.3.1. Cell density

After taking the sample, the broth was diluted to a proper concentration, from 1 to 20 times. Then, the diluted samples were centrifuged to remove the supernatant, and the

58 cells were resuspended into 1 mL distilled water. The optical density was measured to determine the cell density by comparing to a standard curve of optical density with dried cell weight (see Appendix).

3.3.3.2. Glucose and galactose

The fermentation sample was centrifuged and the supernatant was diluted 20 times to measure the glucose using a glucose analyzer (YSI 2700 Select Biochemistry analyzer,

YSI Inc., Yellow Spring, Ohio). After measuring the glucose concentration, the cell-free samples were kept in a -20 °C freezer for further analysis. After the batch, all cell-free samples were diluted into proper concentrations to determine the galactose concentration by HPLC (Refractive Index Detector, Shimadzu, RID-10A) on an organic acid analysis column (BIO-RAD, Aminex® HPX-87H ion exclusion column, 300 mm×7.8mm) at 45

ºC. The samples containing xanthan gum were filtered through 30 MWCO membranes

(Amicon ultra centrifugal filters, 0.5 mL 30 kDa, regenerated cellulose, Millipore,

Billerica, MA) before running HPLC to prevent column failure.

3.3.3.3. Xanthan gum concentration

After fermentation, viscous broth was centrifuged (Beckman Model J2-21 Centrifuge,

JLA-10.500 rotor at 10,000 rpm for 30 min, Beckman Coulter Inc., Brea, CA) to remove cells. Next, 15 mL broth was precipitated with 30 mL ethanol. Then, the precipitant was washed with 5 mL ethanol once. To determine the xanthan gum concentration, the precipitant was dried by lyophilization (LABCONCO Freeze dry system with a

LABCONCO Stopping tray dryer) for one day. The weight difference of the tube divided

59 by the broth volume before precipitation was xanthan gum concentration. To determine the xanthan gum concentration during fermentation, a standard curve of xanthan gum concentration vs. viscosity was used.

3.3.3.4. Apparent viscosity, intrinsic viscosity and rheology of xanthan gum

The broth was diluted into a proper concentration and the viscosity was determined by using a Brookfield viscometer (RVTD II) with SS spindle (Brookfield, Raynham, MA), at 100 rpm and room temperature. 16 ml of the solution were used for each measurement.

The apparent viscosity was then calculated by multiplying the dilution ratio. In the shear stress and shear rate study, 0.5 % (w/v) xanthan gum solution was prepared by dissolving the freeze-dried xanthan gum powder into distilled water. Shear stress was measured at different shear rates at room temperature using the same viscometer and spindle.

A common method to determine molecular weight of polymers is to analyze the intrinsic viscosity of the polymer and then convert the intrinsic viscosity to molecular weight

a through the Mark-Houwink-Sakurada equation: [η]C=0 = KM , where K and a are

1.7×10-4 mL/g and 1.14 (Milas et al., 1985). A commonly-used calibrated kinematic capillary viscometer tube (Cannon-Fenske type, ASTM size 25 B384, viscosity range: 0.5 to 2 cS, Approx. constant: 0.002, Fisher Scientific, Waltham, MA) was applied to determine the intrinsic viscosity. By measuring the time required for solution and solvent to pass through the levels on the capillary viscometer, relative viscosity ηrelative (or ηrel) was obtained by η/η0= t/t0. Then, the intrinsic viscosity was determined by plotting inherent viscosity (lnηrel/C) against concentration. Samples were prepared in a series of concentrations (5, 2.5, 1.25, 0.125 g/L). Each sample was measured three times.

60 Rheological behavior of 0.5 % xanthan solution was measured using a Brookfield viscometer (RVTD II) with SS spindle at various shear rates and at room temperature.

The rheological data was fitted to the Ostwald-de-Waele Equation: τ = Kγn to obtain the flow consistency index, K, and flow behavior index, n.

3.4. Results and discussion

3.4.1. Fermentation kinetics of xanthan gum production from different carbon sources

The goal was to study the feasibility of using HWP as an alternative carbon source for xanthan gum production. Because HWP is a mixture of glucose, galactose, and some residual proteins, the kinetics and characteristics of xanthan gum produced using glucose, galactose, or lactose hydrolysate, which contained glucose and galactose and mimicked

HWP, were also studied in the fermentation and the results were compared to those using

HWP.

The viscosity kinetics in Figure 3.1 show that xanthan gum started to accumulate 24 h after inoculation in all batches. The fermentation kinetics curves were similar using glucose and galactose (Figure 3.1A and B). In the batch using lactose hydrolysate (Figure

3.1C), in which glucose and galactose coexisted, this strain used both glucose and galactose simultaneously with a higher consumption rate of glucose; however, the consumption rate was similar when glucose or galactose was the only carbon source in the medium. Table 3.1 summarizes the fermentations using different carbon sources. In general, comparable yields could be achieved using glucose and galactose for xanthan gum production, with similar final product concentration and productivity.

61 The results show a much higher biomass growth using HWP (Figure 3.2A). It reached more than 5 g/L biomass during the fermentation, which could cause higher oxygen demand, less production, and the problem to separate the cells in downstream process.

The excessive biomass could be due to the residual proteins in the HWP providing the extra nitrogen source for cell growth. In addition, a lower broth viscosity (141.6 cp) in the fermentation using HWP was observed. The final xanthan gum concentration, yield, and productivity were 30.24 g, 0.84 g xanthan/g carbon sources, and 0.43 g/L/h.

Optimally, carbon source is converted to product instead of biomass. To control the biomass formation, the concentration of yeast extract in the medium was decreased from

3 g/L to 1 g/L. Figure 3.2B shows a less biomass formation from 1 g/L yeast extract as compared to the previous batch. Less biomass requires less oxygen, which is easier for process scale up. However, unlike the batch using 3 g/L yeast extract, there was some galactose left in the batch with less yeast extract. The final xanthan gum concentration, yield, and productivity in the batch of 1 g/L yeast extract were 32.16 g, 0.73 g xanthan/g carbon source, and 0.41 g/L/h. As shown in Table 3.1, comparable final product concentration, yield, and overall productivity were obtained when yeast extract was reduced from 0.3 g/L to 0.1 g/L. In other words, yeast extract supplement can be reduced by 67 % when using HWP to produce xanthan gum.

The doubling time of cells using glucose and galactose was 2.98 and 4.13 h, respectively

(Figure 3.3), which was similar to the 3 to 4 h in the literature (Daniels et al., 1984).

However, the doubling time was longer in cell growth using glucose and galactose

62 mixture. The doubling time of cells using lactose hydrolysate and HWP was 6.75 and

7.08 h, respectively.

3.4.2. Characteristic of xanthan gum produced from glucose, galactose, lactose

hydrolysate, and HWP

Figure 3.3B shows that a higher final viscosity was reached when glucose was the only carbon source in the medium. This could be due to the xanthan gum production from galactose not being as consistent as from glucose. Therefore, it affected xanthan gum production when both glucose and galactose existed in the medium. Again, Figures 3.4 and 3.5 show the rheology of xanthan gum produced from different carbon sources. From both figures, it shows that xanthan gum from glucose had higher shear-thinning properties compared to the xanthan gum from sugar mixtures. This could be due to the rheology data of galactose being varied in different batches, which did not have the same consistency as the xanthan gum from glucose. Therefore, it affects the xanthan gum quality when both glucose and galactose are present in the medium. This result agreed with the finding by Thorne et al. (1988), they claimed that different carbon sources could cause the difference in xanthan gum chemical structure. The consistency index K and flow index n of xanthan gum from different carbon sources are listed in Table 3.2. The indexes in literatures are summarized in Table 3.3. However, the xanthan gum concentration used in this study (5 g/L) was different from literatures (30-62 g/L), so the n and K cannot be compared directly between Tables 3.2 and 3.3.

Intrinsic viscosity is a function of molecular shape, size, and shear rate. It is commonly measured via a capillary viscometer, which is simple and low in cost. However, it is

63 limited by uncontrollable shear rate. In this study, the capillary viscometer was operated under atmosphere. According to Rochefort & Middleman (1987), in distilled water at 25

°C, the backbone of the xanthan molecule could not hold a helix structure in a random and disorder form. This is because of the electrostatic repulsions from the charged side chains. Therefore, in our viscosity measurements, xanthan molecules were likely in this random form. From Figure 3.6, the intrinsic viscosity of the xanthan gum produced from glucose, lactose hydrolysate and HWP was similar. The xanthan gum from galactose was slightly higher. According to Thorne et al. (1988), the viscosity of xanthan gum produced by lactose-utilized mutant, X. campestris X59-1232, was lower with the medium containing clarified whey compared to using lactose and glucose. It suggests that the clarified whey could affect the chemical composition or molecular weight of the xanthan gum. The corresponding average molecular weights were calculated with the

a -4 Mark-Houwink-Sakurada equation: [η]C=0 = KM , where K and a are 1.7×10 mL/g and

1.14, respectively (Milas et al., 1985). From the corresponding average molecular weight data in Table 3.1, it suggests that the average molecular weight of xanthan gum from

HWP was slightly lower than that from glucose. However, it still falls within the range of

2 to 50 million Dalton reported in literatures.

3.5. Conclusions and recommendation

In conclusion, xanthan gum started to accumulate 24 h after inoculation in all batches. In the batches using lactose hydrolysate and HWP, glucose was preferred over galactose. It is interesting that this strain used both glucose and galactose simultaneously with different consumption rates.

64 By using HWP, a higher final product concentration, yield, and productivity were achieved compared to the fermentation using glucose. About 67 % of yeast extract could be reduced to save raw material cost when HWP was used as the fermentation substrate.

Further study is necessary to understand the lower broth viscosity from HWP. According to the rheology curve, xanthan gum produced from HWP showed less shear-thinning characteristics, which could be attributed to the pyruvate content affected the viscosity of xanthan gum (Cheetham et a., 1989; Candia and Deckwer, 1999).

Autoclaving whey permeate hydrolysate generates denatured proteins and debris of precipitant, because of the coexisting of proteins and carbon source. One advantage of using RFBB in fermentation with HWP is that the cotton towel could ease downstream process by adsorbing the debris of the precipitant. Alternatively, milder methods, such as microfiltration and pasteurization, could be used for hydrolyzed whey permeate sterilization. In this case, the nutrient degradation due to high temperature autoclaving would be avoided.

65 3.6. References

Amanullah, A.; Satti, S.; Nienow, A. W. Enhancing xanthan fermentations by different modes of glucose feeding. Biotechnology Progress (1998), 14(2), 265-269.

Ashraf S; Soudi M R; Sadeghizadeh M Isolation of a novel mutated strain of Xanthomonas campestris for xanthan production using whey as the sole substrate. Pakistan journal of biological sciences: PJBS (2008), 11(3), 438-42.

Bogaert, J.C., Production and novel applications of natural L (+) lactic acid: Food pharmaceutics and biodegradable polymers. Cerevis (1997), 22, 46-50.

Candia, J.-L. Flores; Deckwer, W.-D. Effect of the Nitrogen Source on Pyruvate Content and Rheological Properties of Xanthan. Biotechnology Progress (1999), 15(3), 446-452

Cheetham, Norman W. H.; Norma, N. M. Nik. The effect of pyruvate on viscosity properties of xanthan. Carbohydrate Polymers (1989), 10(1), 55-60.

Coté, A., Brown, W.A., Cameron, D., and van Walsum, G.P., Hydrolysis of lactose in whey permeate for subsequent fermentation to enthanol. J. Dairy Sci., (2004), 87, 1608-1620.

Daniels, M. J.; Barber, C. E.; Turner, P. C.; Sawczyc, M. K.; Byrde, R. J. W.; Fielding, A. H. Cloning of genes involved in pathogenicity of Xanthomonas campestris pv. campestris using the broad host range cosmid pLAFR1. EMBO Journal (1984), 3(13), 3323-8

Lamboley, L., Lacroix, C., Chmpagne, C.P., Vuillemard, J.C., Continuous mixed strain mesophilic starter production in supplemented whey permeate medium using immobilized cell technology. (1997), Biotechnol. Bioeng. 56, 502-516

Maldonado, Antonio., Effect of medium composition on xanthan gum production from whey permeate using Xanthomonas campestris NRRL B-1459. Master Thesis. The Ohio State University. 1992

Mesomo, Michele; Silva, Marceli Fernandes; Boni, Gabriela; Padilha, Francine Ferreira; Mazutti, Marcio; Mossi, Altemir; de Oliveira, Debora; Cansian, Rogerio Luis; Di Luccio, Marco; Treichel, Helen. Xanthan gum produced by Xanthomonas campestris from cheese whey: production optimisation and rheological characterization. Journal of the Science of Food and Agriculture (2009), 89(14), 2440-2445

66 Milas, M., Rinaudo, M. and Tinland, B. The viscosity dependence on concentration, molecular weight and shear rate of xanthan solutions, Polymer Bulletin, (1985), 14,157-164

Nicolas, G., Auger, I., Beaudoin, M., Halle, F., Morency, H,m LaPointe, G., Lavoie, M.C., Improved methods for mutacin detection and production. J. Microbiol. Methods, (2004), 59, 351-361.

Nykanen, A., Lapvetelainen, A., Hietnen, R.M., Kallio, H., The effect of lactic acid, nisin, whey permeate, sodium chloride and related combinations on aerobic plate count and the sensory characteristics of rainbow trout. Lebensm. Wiss. Technol., (1998), 31, 286-290.

Rochefort, W. E. Middleman, S. (1987). Rheology of xanthan gum: salt, temperature, and strain effects in oscillatory and steady shear experiments. Journal of Rheology, 31(4), 337-369.

Schepers, A.W., Thibault, J., Lacroix, C., Continuous lactic acid production in whey permeate/yeast extract medium with immobilized Lactobacillus helveticus in a two-stage process: model and experiments. Enzyme Microb. Technol., (2006), 38, 324-337.

Schwartz, Robert D.; Bodie, Elizabeth A. Production of high-viscosity whey broths by a lactose - utilizing Xanthomonas campestris strain. Applied and Environmental Microbiology (1985), 50(6), 1483-5

Silva, Marceli Fernandes; Fornari, Rejane C. G.; Mazutti, Marcio A.; de Oliveira, Debora; Padilha, Francine Ferreira; Cichoski, Alexandre Jose; Cansian, Rogerio Luis; Di Luccio, Marco; Treichel, Helen. Production and characterization of xanthan gum by Xanthomonas campestris using cheese whey as sole carbon source. Journal of Food Engineering (2009), 90(1), 119-123.

Silveira, W.B., Passos, F.J.V., Mantovani, H.C., Passos, F.M.L., Ethanol production from cheese whey permeate by Kluyveromyces Marxianus UFV-3: a flux analysis of oxido-reductive metabolism as a function of lactose concentration and oxygen levels. Enzyme Microb. Technol. (2005), 36, 930-936.

Taron, C.H., Benner, J.S., Hornstra, L.J., Guthrie, E.P., A novel β-galactosidase genes isolated from the bacterium Xanthomonas campestris exhibits strong homology to several eukaryotic β-galactosidase. Glycobiol., (1995), 5, 603-610

Thorne, Linda; Tansey, Lisa; Pollock, Thomas J. Direct utilization of lactose in clarified cheese whey for xanthan gum synthesis by Xanthomonas campestris. Journal of Industrial Microbiology (1988), 3(5), 321-8

67

Tyagi, R.D., Kluepfe, D., Bioconversion of cheese whey to organic acids. In: Martin, A.M. (Ed.), Bioconversion of Waste Materials to Industrial Products. Blackie Academic, Glasgow, UK. (1998)

Walsh, Patricia M.; Haas, Michael J.; Somkuti, George A. Genetic construction of lactose – utilizing Xanthomonas campestris. Applied and Environmental Microbiology (1984), 47(2), 253-7.

Yang, T.C., Wu, G.H., and Tseng, Y.H., Isolation of a Xanthomonas campestris strain with elevated β-galactosidase activity for direct use of lactose in xanthan gum production. Letters in Applied Microbiology. (2002), 35, 375-379.

Yang, T.C., Hu, R.M., Hsiao, Y.M., Weng, Sh.F. and Tseng, Y.H., Molecular genetic analysis of potential β-galactosidase genes in Xanthomonas campestris. J. Molecul. Microbial. Biotechnol., (2003), 6, 145-154.

68

Max. cell dry Final viscosity Final product Yield Overall Intrinsic Average weight (g/L) (cp) conc. (g/L) productivity viscosity molecular (g/L/h) (L/g) weight (106 Da) 40 g/L glucose 2.8 ±0.14 234.6 ±26 22.65 ±5.57 0.755±0.007 0.26 ±0.084 5.345 ±1.138 4.47±0.97

40 g/L galactose 2.58±0.31 167.63±76 21.23±4.56 0.68±0.19 0.28±0.088 6.87 4.7

40 g/L lactose 2.60 182 ±31 35.025 ±13.18 0.77 ±0.091 0.29 ±0.042 4.40 3.18 hydrolysate

40 g/L HWP 4.23 158 ±24 31.2 ±1.35 0.79 ±0.07 0.42 ±0.014 5.15 3.65 +0.3% yeast extract 40 g/L HWP 2.74 188-144 32.16 0.73 0.41 3.44 2.56 +0.1%

6 yeast extract

9

Table 3.1. Summary fermentation using different carbon sources

69

n K R2 (mPa∙sn) Commercial XG 0.152 3474 0.98 XG from glucose 0.206 3254 0.93 XG from galactose 0.348 1759 0.97 XG from lactose 0.351 1014 0.99 hydrolysate XG from HWP 0.632 165 0.99

Table 3.2. Summary rheology index of 5 g/L xanthan gum

Substrat Xanthan Strains n K Reference es concentration

X. campestris pv. Cheese 30 g/L 0.28 18.59a Mesomo, et al., Mangiferaeindicae whey 2009 IBSBF 1230 X. campestris 1230 Cheese 30 g/L 0.571 1.898 Silva et al., whey mPa∙sn 2009 X. campestris 1182 Cheese 30 g/L 0.475 1.926 Silva et al., whey mPa∙sn 2009 X. campestris NRRL Glucose 33-62 g/L 0.11-0.22 49000-60 Amanullah et B-1459 E2 000 al., 1998 mPa∙sn Table 3.3. Literature review of rheology index of xanthan gum (n and K are the index of

Oswald-de Waele model, τ = Kγn) a: unit unknown

70

Figure 3.1. Batch fermenation kinetics of xanthan gum production in a stirred-tank fermentor at pH 7 and 30oC from various carbon sources, such as glucose (A), galactose

(B), and lactose hydrolysate (C), with 0.3% yeast extract.

71

Figure 3.2. Batch fermenation kinetics of xanthan gum production in a stirred-tank fermentor at pH 7 and 30oC from HWP with 0.3% (A), and 0.1% (B) yeast extract.

72

Figure 3.3. Effect of carbon source on cell growth (A) and broth viscosity (B).

73

Figure 3.4. Rheology of 0.5 % solution of xanthan gum produced from glucose, galactose, lactose hydrolysate, and HWP. (Spindle S)

74

Figure 3.5. Shear thinning properties of xanthan gum from different carbon sources

75

Figure 3.6. Effect of carbon sources on intrinsic viscsoity of xanthan gum

76

Chapter 4: Water Recycle in a RFBB-Ultrafiltration Integrated Process

4.1. Summary

To make use of excessive waste water and reduce costs of water and raw materials in fermentation, this research studied the feasibility of the recycling and reuse of fermented spent media from xanthan gum production after ultrafiltration. It was demonstrated that ultrafiltration permeate could be used to replace up to 75% of tap water in the fermentation medium without significantly affecting the final product concentration and rheological properties. The kinetics of fermentation and characteristics of xanthan gum produced from media with fresh water and spent media were compared. The capability to produce xanthan gum from spent media was demonstrated in three sequential batches.

Based on the results, ultrafiltration proves promising in the primary recovery of xanthan gum production. It can not only reduce the amount of alcohol required for xanthan gum precipitation, but also reduce the wastewater treatment cost and raw materials required, including tap water and minerals. The quality of xanthan gum produced could meet the requirements for industrial applications.

4.2. Introduction

In a fermentation process, massive amounts of water are usually used once and then discharged. The wastewater contains residuals of nitrogen source, mineral salts, and metabolic wastes. The feasibility of recycling fermentation broth has been investigated in

77 several fermentation processes, including intracellular penicillin amidase, L-lysine, and ethanol production (Babu, et al., 1991; Hsiao and Glatz, 1996; Converti, et al., 1991). The recycled broth was obtained based on the process of primary product separation.

Normally, cells were removed via centrifugation or filtration, and the supernatant or permeate was then processed for recycling.

One major problem in reusing the spent medium is the accumulation of inhibitive compounds during the fermentation. Possible inhibitors in the spent medium are acetic acid, unfermentable polysaccharides and salts (Hsiao and Glatz, 1996). The inhibition could be eliminated or reduced by diluting the spent broth with fresh medium, and the replacement percentage could be determined by minimizing its negative effects on cell growth, productivity and product quality. The main goal of water reuse study was to maximize the replacement percentage and the number of sequential batches with spent medium recycling. The research related to water reuse is summarized in Table 2.4. To date, no study on recycling water in biopolymer fermentation has been reported.

The objective of this chapter was to study the feasibility of recycling ultrafiltration permeate for use in xanthan gum fermentation. The percentage of spent medium in medium preparation was optimized and three sequential batches were conducted using a stirred tank bioreactor. Biomass concentration, xanthan gum production and xanthan gum properties using different media were studied.

78

4.3. Materials and methods

4.3.1. Culture and media

Xanthomonas campestris NRRL B1459 was maintained on agar plate at 4 °C and subcultured every two weeks. Cells from an agar plate were inoculated into 5 mL medium. One day later, 2 mL broth was transferred to two flasks each with 50 mL medium. After one day culture, cells were used to inoculate a 4 L fermentor. Medium preparation was based on the percentage of ultrafiltration permeate used in the fermentation. Details are described later. All media used in bioreactor scale fermentation containd 500 ppm antifoaming agent, MURNC MCA250 or MCA 270 (Murnc, Inc.,

Libertyville, IL).

4.3.2. Shake-flask culture for replacement percentage experiment

The replacement percentage experiment was conducted in 250-mL flasks with 50 mL working volume. Ultrafiltration permeate in this experiment was the primary permeate, which the fermentation broth was cultured from fresh medium. Media were prepared based on the ratio of tap water and ultrafiltration permeate (Table 4.1). Glucose was sterilized separately from the other components and mixed before inoculation to prevent browning reaction. After 72 h of inoculation, broth in each condition was collected to measure biomass, viscosity and final xanthan gum concentration.

79

4.3.3. Fermentation using a stirred tank

The reactor system used in this study was New Brunswick BioFLO 3000 equipped with a

5 L glass vessel. The production of xanthan gum was two-phase fermentation, in which the temperature and pH was 25 ˚C, and 6.5 during the first day after inoculation and changed to 30 ˚C and pH 7.0 for the rest of the culture.

4.3.4. Ultrafiltration

After the fermentation, cells were removed by centrifugation at 10,000 rpm, 4ºC for 30 min (Beckman Model J2-21 Centrifuge, JLA-10.500 rotor, Beckman Coulter Inc., Brea,

CA). Then, xanthan gum was concentrated through a tubular hollow fiber ultrafiltration cartridge (Model No. 0720199, Koch Membrane Systems Inc., Ann Arbor, MI) equipped with a peristaltic pump drive (Voltage: 115 VAC, frequency: 50/60 Hz, capacity: 6,600 rpm, full load: 143 Watts, Millipore, Billerica, MA) and a standard pump head (Model

7024-20, Cole-Palmer Instrument Company, Vernon Hills, IL) for L/S 24 tubing at room temperature. The pressure drop in the ultrafiltration system was monitored with two pressure gauges installed at the inlet and outlet of the filtration cartridge, and the pressure drop was controlled at no larger than 25 pisg via a valve (Figure 4.1). After ultrafiltration, permeate was collected and used to replace tap water for the subsequent batches. If the ultrafiltration permeate was not used in the same day after ultrafiltration, it was store in a

4 ºC cold room before use.

80

4.3.5. Analytical methods

4.3.5.1. Cell density

Cells were centrifuged and resuspended in distilled water. The optical density was determined with a spectrophotometer (Shimadzu UV-Visible Spectrophotometer) at 650 nm. The measured optical density was converted to dried cell weight according to a calibration curve.

4.3.5.2. Xanthan gum concentration

At the end of the fermentation, samples were centrifuged to remove cells and 15 mL broth was precipitated with 30 mL ethanol. After centrifugation to remove the supernatant, precipitated xanthan gum was washed with 5 mL ethanol and then freeze-dried via a freeze-dryer (LABCONCO Freeze dry system with a LABCONCO

Stopping tray dryer) overnight. It was then weighted to calculate final product concentration.

4.3.5.3. Viscosity of xanthan gum

Broth was diluted to a proper concentration before measuring the viscosity. The viscosity was obtained via a Brookfield viscometer (RVTD II) (Brookfield, Raynham, MA) at a shear rate of 122.3 1/s (spindle S at 100 rpm) under room temperature. The broth viscosity was the measured value multiplied by the dilution ratio.

The United States Pharmacopeia/National Formulary (USP/NF) monograph viscosity specification for xanthan gum requires that the apparent viscosity of a 1.0 % w/w xanthan gum solution (with the addition of 1.0 % w/w KCl), measured at 24±1 ºC via a rotating 81

spindle viscometer at 60 rpm, should not less than 600 mPa·s. Therefore, recovered xanthan gum from fermentation was prepared in 1.0 w/w, with 1.0 % w/w KCL in distilled water to measure its viscosity at 60 rpm using spindle 5.

Intrinsic viscosity was measured using a calibrated kinematic capillary viscometer tube

(Cannon-Fenske type, ASTM size 25 B384, viscosity range: 0.5 to 2 cS, Approx. constant: 0.002, Fisher Scientific, Waltham, MA) and determined as described in the previous chapter.

4.4. Results and discussion

4.4.1. Replacement percentage

To study the feasibility of using ultrafiltration permeate in xanthan gum production, different amounts of ultrafiltration permeate was used to replace tap water for xanthan gum production in flasks in order The replacement percentages were 25%, 50%, 75%, and 100%. The amounts of K2HPO4 and MgSO4·7H2O were reduced according to the replacement percentages. The control shake flasks used fresh medium. The medium pH was adjusted to 6.5 before sterilization. One mL cells were inoculated into each flask and cultured at 25 °C. After 72 h, xanthan gum was harvested by removing cells through centrifugation and precipitated with 67% v/v ethanol.

As shown in Figure 4.2A, the biomasses from the shake-flasks ran with replacement percentages of 25, 50, and 75% were higher than the control. The reduced biomass with

100 % permeate replacement indicates that the total replacement of tap water to ultrafiltration permeate inhibited the cell growth and xanthan gum production. 100 %

82

permeate replacement is excluded from the following discussion. The biomass from spent medium was slightly higher than the control, and there was no significant difference between the replacement percentages. The viscosity of broth in this study showed that a higher viscosity of broth could be achieved by using a higher replacement percentage.

The final xanthan gum concentration from media with various percentages of permeate were not significantly different from each other, but higher than the control. The results also indicated that the higher viscosity was resulted from the higher molecular weight instead of higher xanthan gum concentration.

In conclusion, a 75 % replacement of ultrafiltration permeate could be used in future fermentation and further optimization of medium supply would reduce the medium cost.

4.4.2. Sequential fermentations using spent medium from ultrafiltration

After determining the replacement percentages of ultrafiltration permeate in each batch, the feasibility of spent medium reuse in sequential fermentations was studied by comparing the biomass concentration, final product concentration, yield, productivity and product properties, such as broth viscosity and intrinsic viscosity.

The fermentation started with 4 L fresh medium, which contained 3 g/L yeast extract, 2 g/L K2HPO4, 0.1 g/L MgSO4·7H2O, and 25 g/L glucose in tap water at pH 6.5, which

(STR-UFP0) was also used as the control (Figure 4.3A). After fermentation, cells were removed via centrifugation and the xanthan gum was concentrated by ultrafiltration. The permeate from ultrafiltration (Table 4.2) was collected for next batch fermentation

(STR-UFP1).

83

As shown in Table 4.2, in the first ultrafiltration batch, about 75% of the broth volume was removed by ultrafiltration. A lower recovery yield in this table was due to the oversize ultrafiltration column, which held up 400-500 mL of xanthan gum broth. The loss of this liquid hold up could be neglected when the process is scaled up to a larger working volume. Therefore, the 75 % replacement is practical in logistics.

A 75 % replacement of permeate was then used in a 3 L stirred tank fermentation with extra 25 g/L glucose, 3 g/L yeast extract and complimentary mineral (1.5 g K2HPO4 and

0.075 g MgSO4‧7H2O). In Figure 4.3B, the glucose was consumed within 48 h and 18.3 g/L xanthan gum was produced using the primary ultrafiltration permeate. After the fermentation, the broth was again centrifuged to remove cells before ultrafiltration. The ultrafiltration permeate from STR-UPF1 fermentation was used in preparing the medium for STR-UFP2 batch (1.8 L) with a replacement percentage of 75 %. At the end of

STR-UFP2, 22.55 g/L xanthan gum was produced within 72 h using secondary ultrafiltration permeate (Figure 4.3C). The results of the sequential batches are summarized in Table 4.3. Due to the volume loss after ultrafiltration, the permeate from the third batch was not enough for a new batch, so only three batches of fermentation were conducted.

Generally, the doubling time of Xanthomonas campestris in literature is 3 to 4 h (Daniels et al., 1984). In Figure 4.4A, the doubling time of X. campestris in fresh medium batch

(STR-UFP0) was 3.92 h; while the doubling time of cells in 75 % replacement ultrafiltration permeate was 2.98 h (STR-UFP1). The doubling time was higher using 75

% primary ultrafiltration permeate, which suggests that the permeate was not toxic to

84

xanthan gum fermentation. However, the doubling time of biomass using 75 % secondary ultrafiltration permeate was longer (5.5 h). A lag phase occurred in STR-UFP2, but the culture recovered and reached the same biomass concentration as STR-UPF1. In this figure, it shows that peak biomass concentration was higher using spent medium than using fresh medium, which could be due to the residual nitrogen source in ultrafiltration permeate.

In the first two batches, the glucose was used up in 48 h, but in the third batch,

STR-UFP2, the glucose was consumed in 72 h (Figure 4.4B), which can be attributed to the long lag phase in STR-UFP2. This lag phase could be eliminated by using a better seeding culture. Basically, the glucose consumption rates in these three batches were similar.

According to the broth viscosity during fermentation (Figure 4.4C), the xanthan gum production was lower when using 75% permeate than when using fresh medium.

However, at the end of fermentation, the broth viscosity using the primary permeate and the secondary permeate was similar to the control. The yield of all three sequiential batches were slimiar to the yield of xanthan gum fermentation using fresh medium. A slightly higher productivity from primary and secondary permeate could be due to the higher biomass, which could be contributed from the remaining nitrogen source in the permeate.

4.4.3. Characteristic of xanthan gum produced from spent medium

In fermentation, the broth viscosity increased with increasing the xanthan gum concentration and its molecular weight. A common method to characterize the

85

rheological property of xanthan gum is to measure the apparent viscosity under a constant shear rate of 6 s-1 (Brookfield viscometer (RVTD II) spindle #5, at 100 rpm). Generally, the apparent viscosity of 1 % xanthan gum in water ranges from 1200 to 1600 cp for the application of hydrocolloid and thickening agent. In this study, after fermentation, xanthan gum was concentrated via ultrafiltration and precipitated by enthanol. Next, xanthan gum was dried by an oven and was then dissolved in water to prepare 1% solution. The results showed that the xanthan gum from first subsquential batch

(STR-UFP1) had a similar viscosity as compared to xanthan gum produced from fresh medium (STR-UFP0), and both met the application industrial requirement. However, the viscosity of xanthan gum from STR-UFP2 was lower than those from the previous two batches (Figure 4.5A). The reason for a lower viscosity of xanthan gum from STR-UFP2 will be discussed later.

Another specification for xanthan gum is United States Pharmacopeia/National

Formulary (USP/NF) monograph viscosity. Xanthan gum 1.0 % w/v was prepared in 1.0

% KCl solution and the viscosity was measured at 60 rpm (spindle 5) at 24±1 ºC. Figure

4.5B shows that all xanthan gum samples matched the standard criteria of USP/NF.

Next, we compared the shear stress of xanthan gum from sequential batches under different shear rates (Figure 4.6). When the shear rate increased, the slope of shear stress over shear rate decreased dramatically with xanthan gum produced from the fresh medium. However, the slope gradually decreased with xanthan gum produced from the spent medium. There could be a phase or shape change of xanthan gum when the shear rate increased. The transition was sharper in xanthan gum from the fresh medium than the

86

spent medium. In other words, xanthan gum from fresh medium is more shear-thinning.

In Figures 4.6A, 4.6B, and 4.7, the curves show the shear thinning property of xanthan gum solution.

In the previous section, it seems that the xanthan gum produced from STR-UFP2 was less viscous as compared to the previous two batches. However, the fermentation results showed a compatible broth viscosity with similar productivity between STR-UFP0 and

STR-UFP2. Therefore, we studied the molecular weight of xanthan gum from these three batches. In this study, the molecular weight of xanthan gum from each batch was compared by its intrinsic viscosity measured via a Cannon-Fenske capillary viscometer.

According to Figure 4.8, the intrinsic xanthan gum viscosity of each batch (STR-UFP0,

STR-UFP1, and STR-UFP2) was 6.1561, 5.0177, and 4.4499 L/g, respectively. The corresponding average molecular weights were calculated by the

a -4 Mark-Houwink-Sakurada equation: [η]C=0 = KM , where K and a are 1.7×10 mL/g and

1.14 (Milas et al., 1985). Although there seems to be a slight decrease of intrinsic viscosity of xanthan gum in sequential batches, the average molecular weight were not significantly different (Table 4.4). Although the intrinsic viscosity of xanthan gum depends on microorganism and nutrient, the spent medium showed little effect on the average molecular weight of the final product. Based on the final concentration and final viscosity of xanthan gum in each batch, the higher viscosity in STR-UFP2 batch was attributed to the higher concentration of xanthan gum instead of a higher molecular weight. Since the average molecular weight was similar among the three batches, the lower viscosity of xanthan gum from STR-UFP2 (Figures 4.5) could be due to the

87

structure difference of xanthan gum, such as the pyruvate content difference on the xanthan molecules. More experiments relating to the side chains should be done to answer the question.

4.5. Conclusions and recommendation

In conclusion, ultrafiltration can not only recover and concentrate biopolymers, such as xanthan gum, after fermentation, but also recycle the medium for the next fermentation run. Ultrafiltration permeate (up to 75 % replacement) from xanthan gum fermentation broth could be reused in at least one sequential fermentation without interfering with cell growth and xanthan gum production. In the second sequential fermentation batch, there was a lag phase right after seeding. However, the culture was recovered and produced a compatible amount of xanthan gum within 72 h. Xanthan gum fermentation was similar using fresh medium, primary permeate and secondary permeate in terms of final viscosity, final product concentration, yield, and productivity. The molecular weight of xanthan gum produced in the spent media was similar to those in the fresh medium.

In this study, fermentation batches were conducted in a stirred tank, and centrifugation is required to remove cells before ultrafiltration. Future study will be conducted with the fermentation using a RFBB, which could eliminate the centrifugation process, by cell immobilization within RFBB. Ultrafiltration permeate was sterilized via autoclave before fermentation in this study. An alternative method is to use microfiltration and pasteurization, which is commonly used in industrial processes. In this case, the nutrient degradation due to high temperature autoclaving would be avoided.

88

If the bottleneck of the fermentation is not the limitation of metabolites with 75% replacement, the metabolites will eventually reach 300 % of those at the end of the first batch (Figure 4.9). According to this curve, the accumulation of metabolites increases dramatically in the first 10 batches. Therefore, if the inhibition did not occur in the first

10 batches, the water could be recycled in infinite batches. Based on this calculation, 10 sequential batches are required to verify the infinite- batch recycling.

4.6. References

Babu, P.S.R.. Panda, T. 1991. Effect of recycling of fermentation broth fro the production of penicillin amidase. Process Biochemistry, 26, 7-14.

Converti, A., Perego, P., Lodi, A., Fiorito, G., Borghi, M.D., Ferraiolo, G. 1991, In-situ ethanol recovery and subst4rate recycling during continuous alcohol fermentation. Bioprocess Engineering, 7, 3-10.

Daniels, M.J., Barber, C.E., Turner, P.C., Cleary, W.G.and Sawczyc, M.K., Isolation of mutants of Xanthomonas campestris pv. campestris Showing Altered Pathogenicity Journal of General Microbiology (1984) 130, 2447-2455

Hsiao, T-Y. and Glatz, C.E., 1996, Water Reuse in the L-Lysine Fermentation Process, Biotechnology and Bioengineering, 49, 341-347.

Milas, M.; Rinaudo, M.; Tinland, B. The viscosity dependence on concentration, molecular weight and shear rate of xanthan solutions. Polymer Bulletin (Berlin, Germany) (1985), 14(2), 157-64

89

Medium Replacement percentage (%) Control Components 25 50 75 100 Glucose (g) 1.2 1.2 1.2 1.2 1.2 Yeast extract 0.15 0.15 0.15 0.15 0.15 (g)

K2HPO4 (g) 0.1 0.075 0.05 0.025 0

MgSO47H2O 0.005 0.00375 0.0025 0.00125 0 (g) Tap water 50 37.5 25 12.5 0 (mL) Ultrafiltration 0 12.5 25 37.5 50 permeate (mL) Table 4.1. Medium composition in replacement percentage study. each flask contained 50 mL of medium at pH 6.5.

90

Before After Volume Recovery UF batch Volume Viscosity Concentration (ml) Viscosity Concentration yield (ml) (cp) (g/L) (permeate/ (cp) (g/L) concentrate) 1 3250 215.6 18.71 2300/500 815 72.57 ~60 % 2 2250 194.6 18.3 1350/400 N/A 51.65 ~50 % Table 4.2. Volume, viscosity and concentration of xanthan gum before and after ultrafiltration. (N/A: not available)

91

91

Specific Final Final Yield Recycle Biomass Productivity growth rate viscosity product (g XG/g ratio (%) (g/L) (g/L/h) (h-1) (cp) conc. (g/L) glucose) Fresh medium 0 2.7 0.28 215.6 18.71 0.75 0.23 4 L (STR-UFP0) Primary permeate 75 3.7 0.26 194.6 18.3 0.732 0.38 3 L (STR-UFP1) Secondary Permeate 92 75 3.8 0.25 231.6 22.55 0.8165 0.334

1.8 L (STR-UFP2) Fresh medium 0 3.1±0.63 0.28±0.02 205.4±14 20.64±2.72 0.82±0.091 0.26±0.05 (average) Table 4.3. Summary of fermentation using ultrafiltration permeate

92

Average molecular Apparent viscosity (cp) Intrinsic viscosity (L/g) weight (106 Da)

Fresh medium 1356±17 5.3±1.1 3.80±1.0

Primary permeate 1270±14 5.0 3.6 3 L (STR-UFP1) Secondary Permeate 956±6 4.4 3.2 1.8 L (STR-UFP2) Table 4.4. Summary of appearant viscosity and intrinsic viscosity of xanthan gum from fresh and spent

93 media

93

Figure 4.1 Diagram of ultrafiltration system

94

Figure 4.2. Effects of ultrafiltration permeate replacement percentage on cell growth, xanthan gum production, and broth viscosity (n=2)

95

Figure 4.3. Fermentation kinetics of xanthan gum production in fresh medium STR-UFP0

(A), and spent medium STR-UFP1 (B) an d STR-UFP2 (C).

96

Figure 4.4. Comparision of cell growth (A), glucose consumption (B), and broth viscosity

(C) in xanthan gum fermentations with fresh and recycled media.

97

Figure 4.5. Comparision of apparent viscosity of xanthan gum (1% w/v) produced from fresh and recyceled media in sequential batch fermentations. Spindle 5, 100 rpm (A), and

60 rpm (B).

98

Figure 4.6. Rheology of 0.5 % xanthan gum from fresh and spent medium in terms of shear stress (Spindle S).

99

1.0E+06 STR-UFP0 STR-UFP1 1.0E+05 STR-UFP2

1.0E+04 Viscosity (cp) Viscosity 1.0E+03

1.0E+02 0.01 0.1 1 10 100 1000 Rotational rate (rpm)

Figure 4.7. Apparent viscosity of 1 % solution of xanthan gum produced from fermentation with fresh and spent media (spindle 5).

100

10 STR-UFP0 STR-UFP1 8 STR-UFP2

y = -5.0032x + 6.1561

6 /C

rel y = -4.4576x + 5.0177 ƞ ln 4

2 y = -3.7818x + 4.4499

0 0 0.1 0.2 0.3 0.4 0.5 0.6 Xanthan concentration (g/L)

Figure 4.8. Effect of ultrafiltration permeate on intrinsic viscsoity of xanthan gum

101

350 300 250 200  150 0.75n  3 100 n0

50 Inhibitor accumulation (%) accumulation Inhibitor 0 0 5 10 15 20 25 30 Batch #

Figure 4.9. The accumulation of possible inhibitors by recycling the spent medium at

75% replacement rate.

102

Chapter 5: Xanthan Gum Fermentation using Rotating Fibrous Bed Bioreactor (RFBB)

5.1. Summary

Rotating Fibrous Bed Bioreactor (RFBB) could improve oxygen transfer in viscous xanthan gum fermentation. It could also increase productivity and efficiency in repeated-batch cultures once started by saving the time and cost required to prepare additional inoculations. In the previous study, operational parameters, including rotational rate, liquid recirculation rate, and working volume, were limited by equipment capacity. Repeated-batches with a larger broth volume using higher rational rate and recirculation liquid flow rate were conducted in both gas and liquid continuous modes in this study. In addition, all experiments conducted in RFBB were reviewed and compared as to the effect of operational parameters on volumetric productivity, yield, and final concentration. The result indicated that the final concentration was not sensitive to oxygen transfer coefficient, while productivity was sensitive to the oxygen transfer in xanthan gum production.

5.2. Introduction

Xanthan gum is currently produced by Xanthomonas campestris fermentation in stirred tanks. Because of the increasing viscosity during fermentation, oxygen transfer becomes critical in the fermentation and affects final product concentration, yield, productivity, and product quality. Because of the low product concentration, it requires a large volume 103 of the alcohol to precipitate the xanthan gum. Even though the ethanol is recycled by distillation, the ethanol and the energy in distillation are still costly in the production.

To overcome the problem of oxygen transfer in viscous biopolymer fermentation, we proposed to use a rotating fibrous bed bioreactor (RFBB), which is a modified stirred tank composed of a rotating fibrous bed and a recirculation system (Figure 5.1). The rotating fibrous bed composed of two concentric stainless steel cylinders is added onto the impeller. Between the two stainless steel cylinders is a spiral-packed cotton cloth, which is supported by a stainless steel wire mesh. This part is for microorganism immobilization. After attaching to the matrix, cells are able to stay in the fibrous bed under a high gravitational field when the bed is rotating. The recirculation system is to pump xanthan broth out from the bottom and then spray the broth back to the fibrous bed via a nozzle. Rotational rate, gas flow rate, recirculation liquid flow rate, and liquid volume are important operational parameters in RFBB.

There are three important factors of RFBB for solving the oxygen transfer limitation problem: creating a high gravitational field through rotating, pumping during recirculation, and spraying during recirculation. First, since the rotation of the fibrous bed can create a HIGEE (high “g”) field, this design can force gas and liquid to move through the bed. In other words, it forces gas to contact with fermentation broth and the broth to contact with the cells on the matrix. Therefore, it can increase gas-liquid and liquid-solid volumetric mass transfer. For xanthan gum production, HIGEE field also reduces the viscosity of shear-thinning xanthan gum by providing centrifugal force as shear stress, so the lower apparent viscosity further increases mass transfer in the broth. Continuous

104 pumping and circulating of the broth can minimize heterogeneous distribution in the vessel. In addition, the stress provided from pumping is also beneficial to lowering biopolymer viscosity. Finally, the spray of broth back to the RFBB breaks the broth into liquid drops in the head space, which can greatly increase the surface for gas-liquid transfer. When the broth is pumped out and re-sprayed on the top of the matrix, the broth with low dissolved oxygen at the bottom of the RFBB can be exposed to an oxygen-rich phase on the top of the matrix.

The HIGEE concept and design were patented by Ramshaw and Mallison in 1981 (Rao et al., 2004). They built a HIGEE environment by introducing a rotating packed bed in a reactor. The bed could be packed with different matrices, such as glass spheres, plastic beads, metal wire mesh, and disks. The range of porosities and surface areas of these dry packing materials were 0.4-0.95 and 750-5000 m2/m3. The advantage of a rotating packed bed (RPB) reactor in process intensification was to increase gas liquid mass transfer in distillation, adsorption, and stripping. Companies (e.g., Glitsch Inc. and the Dow

Chemical Company) reported the commercial use of RPB (Rao et al., 2004).

Yang et al. (1996) applied this concept to overcome the oxygen limiting problem in xanthan gum fermentation and design a rotating fibrous bed bioreactor (RFBB), which was also called a Centrifugal Packed Bed Reactor (CPBR). This design has been studied in two modes: centrifugal packed bed reactor under a liquid continuous mode (RFBB -LC) and centrifugal packed bed reactor under a gas continuous mode (RFBB -GC). In RFBB

-LC, fibrous bed was totally immersed in the liquid medium, while in RFBB-GC, it was partially immersed in the broth. By using RFBB-GC, a cell-free xanthan gum broth could

105 be produced with a volumetric productivity of ~1 g/L/h, which was higher than 0.5 g/L/h in traditional stirred tanks (Yang et al., 1996). At the end of the fermentation, a much higher cell density was obtained in RFBB-LC (6.8 g/L) and RFBB-GC (15.36 g/L), as compared to in traditional stirred tanks (~2 g/L). However, the viability at the end of the repeated-batches was only 60 %, which might be due to medium changes between batches and/or insufficient oxygen transfer for such high cell densities. Therefore, enhancing OTR by increasing the rotational rate and recirculational rate could help improve the performance of RFBB.

The kinetics and factors of cell adsorption in fibrous matrix have been studied (Yang et al., 1998). It showed that among all the fabrics (cotton towel, cotton fabric sheet without looping, 50 % cotton-50% polyester fabric sheet, and polyester fabric sheet), cotton towel with xanthan gum on it had the best adsorption result. It required ~50 h for cell attachment at the beginning of the fermentation and cell immobilization allowed very high density growth within the bioreactor through a series of batches in the repeated-batch mode. The time and labor for reactor cleaning and inoculation are greatly reduced compared to batch fermentation. Volumetric oxygen transfer coefficient (kLa) in a 5-L RFBB was compared with kLa in stirred tanks using different impellers and water-in-oil system (Lo et al., 2001). Although the water-in-oil system in a stirred tank

-1 -1 reactor had a higher kLa (0.020 s at 3.5 % w/v xanthan gum) than in RFBB (0.018 s at

3.5 % w/v xanthan gum), the recovery of product was difficult in the water-in-oil system, which limited its scalability (Lo et al., 2001). The gas continuous mode was designed to provide extensive gas-liquid contact, and the kLa measurement of RFBB-GC confirmed

106 that RFBB-GC has a better oxygen transfer capacity compared to RFBB-LC. The authors also evaluated the oxygen transfer with and without a recirculation loop in RFBB-LC.

The results showed that the impact of a recirculation loop on oxygen transfer efficiency was more significant when the agitation was low in RFBB-LC. For RFBB-GC, the recirculation loop was critical for the operation.

Although a cell free xanthan gum broth was produced with high productivity, yield and final concentration, the process was not fully optimized due to rotation rate and recirculation rate limitations, which limited oxygen transfer efficiency. The rotational rate was limited by severe vibrations over 500 rpm, and the recirculation liquid flow rate was limited by the capacity of pump head available. In addition, in the gas continuous mode, the fibrous bed was not truly gas continuous with the part of the fibrous bed embedded in the broth. Therefore, a larger RFBB with a separated circulation tank and a higher capacity pump was proposed (Figure 5.2). In this design, fibrous bed was equipped in an empty vessel while the broth was collected in another vessel. In this way, the fibrous bed was more exposed to the air and the oxygen transfer increased. The recirculation loop provided the contact of the broth to cells and air. Using this setup, Doma (1999) studied the kLa and established a kLa model in RFBB in terms of gas flow rate, liquid flow rate, rotational rate and liquid volume. He also studied the mean residence time based on different operational parameters to calculate the liquid hold-up in the fibrous bed. He claimed that at higher gas flow rates and rotational rates, liquid hold-up was less and the fibrous bed during operation was drier, which led to droplets instead of a liquid film

107 when the liquid passed through fibrous bed. In sum, the RFBB can provide a high oxygen transfer rate and demonstrates scale up feasibility.

There were other studies related to RFBB reported (Hsu and Lo, 2003; Hsu et al., 2004).

The first one was a study of intracellular biosynthesis of xanthan gum in RFBB-GC and

RFBB-LC by metabolic flux analysis (Hsu and Lo, 2003). Although gas continuous mode was more efficient according to kLa study, 5.0 % (w/v) glucose in RFBB-LC gave the highest energy-efficiency among three operation modes, including stirred tank reactor,

RFBB-LC, and RFBB-GC (Hsu and Lo, 2003). The better production with a higher C/N ratio in RFBB-LC was due to higher glucose uptake rate. The authors speculated that the configuration of RFBB provided higher oxygen transfer efficiency, which might cause the transformation of Glc-6-P to sugar nucleotides and then to formation of xanthan gum.

The second paper studied the effect of fibrous surfaces on cell immobilization in fibrous bed (Hsu et al., 2004), in which untreated cotton towel showed the best immobilization efficiency. It agreed with the result from Yang et al. (1998) that cotton towel was the best choice thus far.

In early-stage researches related to RFBB, shaft vibration at high rotational rate, limited liquid flow rate due to the pump capacity, and partial gas continuous mode due to limited vessel space were considered as the three most critical issues (Doma, 1999). The vibration problem was later solved by using a larger shaft and motor equipped in a larger fermentor. The liquid flow rate was increased by using a larger pump. The addition of a circulation tank created a true gas continuous environment for the fibrous bed (Doma,

1999). Then, kLa model was established covering most of the important operational

108 parameters, including rotational rate, liquid flow rate, gas flow rate, and xanthan gum concentration (Doma, 1999). Although a high rotational rate had been considered to be beneficial to mass transfer, a rotational rate higher than 400 rpm resulted in cell detachment due to the weak bonding between cells and the fibrous matrix (Doma, 1999).

A similar design of RFBB was reported by modifying a stirred tank with an external loop

(Zhu et al., 2006). The external loop started 24 h after inoculation, which pumped out the broth from the bottom of fermentor and circulated the broth back to the top of the fermentor. This design increased final product concentration from 22.4 g/L to 24.3 g/L and decreased fermentation time from 48 h to 42 h, using corn starch, corn steep liquor and sodium glutamate as substrates with a total sugar concentration of 32 g/L. It was equivalent to an overall productivity of 0.75 g/L/h. However, the increased final concentration was not as high as the performance of RFBB (> 35 g/L). The authors claimed that the external loop could eliminate the dead spaces in the fermentor. They mentioned that the external loop did not increase or even decreased the dissolved oxygen in the broth, which was different from the results from our lab (Lo et al., 2001). The authors also claimed that the increase of production was due to better mixing. The overall energy consumption in their study was measured via a kilo-watt-hour meter, and the result showed that the addition of external loop reduced the energy consumption by 18 % per unit xanthan gum production. The RFBB setups and operation modes in literatures are summarized in Table 5.1.

Generally, the scale-up of bioprocess was based on the following parameters (Najafpour,

2007):

109  Similar Reynolds number or momentum factors

 Constant power consumption per unit volume of liquid, P/V

 Constant impeller tip velocity, NDi

 Equal liquid mixing time and recirculation time

 Constant volumetric of mass transfer coefficient, kLa

 Keep all environmental factors for the microorganism constant

Similar Reynolds number or momentum factors were generally used in the scale-up of the unit operations, such as piping. A unit operation requiring agitation or mixing could be scaled up based on constant power consumption per unit volume. A constant impeller tip velocity was also used, especially for the agitation of shear sensitive materials, or viscous broth to ensure the sufficient mixing. Maintaining similar liquid mixing time and recirculation time is also important during scale-up. Constant volumetric of mass transfer coefficient was the most commonly used parameter in aerobic fermentation and cell culture, because oxygen transfer not only affects the biomass production, but also the quality of many products from fermentation and cell cultures (Trujillo-Roldán et al.,

2001). Finally, all environmental factors are required to keep constant to ensure the product quality.

In this study, we scaled up the process based on the most commonly used factors in bioprocess, kLa and P/V. The most critical issue in viscous biopolymer fermentation is oxygen transfer. Therefore, applying the same oxygen transfer coefficient in the lab-scale and the large-scale can provide the same level of oxygen transfer to microorganism culture. To maintain a constant kLa, large-scale process typically requires a much higher

110 sparging rate and/or a more intensive agitation. The enormous power for agitation in viscous biopolymer fermentation often limits cell growth in large scale bioreactor.

In the study of Yang et al. (1996), the liquid volume in a 5 L fermentor was reduced from

5 L to 2.5 L for a gas continuous mode. Even at a liquid volume of 2.5 L, the fibrous bed was partially submerged in the broth, which was not completely gas continuous. The maximum rotational rate in Lo et al. (2001) was 350 rpm due to vibration at higher rotational rates. In Doma’s design, a true gas continuous mode was achieved. Because a

20 L fermentor was used to install the fibrous bed, the shaft and motor were much more stable at high rotational rates. The advantage of Doma’s design over the original design

(Figure 5.1) is that it could hold more liquid in the circulation tank and maintain the fibrous bed in a gas continuous mode. However, in Doma’s dissertation, only 2 L liquid broth was circulated between fibrous bed and circulation tank. In order to scale up the

RFBB, 4 L fermentation broth was circulated through the same size of fibrous bed in this study. Then, the scaled up results were compared to those from literatures and discussed in terms of oxygen transfer, mixing and shear in RFBB.

5.3. Materials and methods

5.3.1. Culture and media

Xanthomonas campestris NRRL B1459 was used in this study. The medium contained 25 g/L or 50 g/L glucose, 3 g/L yeast extract, 2 g/L K2HPO4, and 0.1 g/L MgSO4‧7H2O in tap water at pH 6.5 with 500 ppm antifoaming agent (MURNC MCA250 or MCA 270;

111 Murnc, Inc., Libertyville, IL). 5 % (v/v) inoculum was added to the fermentor at the beginning of the fermentation.

5.3.2. Rotating fibrous bed bioreactor

The feasibility and performance of RFBB in xanthan gum fermentation were studied under a mode of liquid continuous repeated-batch in a 5 L reactor and a mode of gas continuous repeated-batch in a 20 L reactor with a 5 L circulation tank. The accumulating xanthan gum helped cell attach to fibrous matrix, which started 24 h post-inoculation.

The rotational rate was low at 150 rpm until 48 h post-inoculation, which allowed cell attachment. After the cell immobilization, rotational rate was raised to 300 - 350 rpm, medium was changed, and the recirculation started for the rest of the fermentation.

5.3.2.1. Liquid continuous repeated batch

The rotating fibrous bed bioreactor was modified from a 5 L New Brunswick BioFLOII reactor by adding a stainless bed (OD = 9 cm, ID = 2.7 cm, height = 15 cm; the thickness of the fibrous bed is 3.15 cm) and a circulation system using a peristaltic pump

(Masterflex drive 7549-32 with Easyload pump head, Cole Parmer, IL) (Figure 5.1). The working volume was 4 L. Cotton towel with a stainless mesh was packed spirally in the fibrous bed as the matrix for cell adsorption. The RFBB fermentation was operated in two phases. The temperature and pH were 25 ºC and 6.5, respectively, in the first 24 h and shifted to 30 ºC and 7.0 for the rest of the fermentation, which was the same two phase process as the stirred tank fermentation in the previous chapters. The same two phase process was also used for all the following fermentations. The rotational rate was

112 150 rpm in the first 48 h for cell immobilization and increased to 350 rpm later. A ~500 mL/min circulation was started after 48 h. The aeration was 1.5 vvm. In the repeated-batches, viscous broth was pumped out at the end of batches 2, 3, and 4, and fresh medium was added into the fermentor. Medium in batch 1 was not pumped out and extra 2.5 % glucose was added into the fermentor to start batch 2.

5.3.2.2. Gas continuous repeated-batch:

An improved gas continuous repeated-batch (Figure 5.2) was conducted to obtain a larger working volume. There were two vessels in this design, including a circulation tank containing culture broth of 4 L and a stirred tank bioreactor (20 L) installed with a fibrous bed (OD = 9 cm, ID = 2.7 cm, height =15 cm). The rotor and the circulation tank were autoclaved twice at 121 ºC, 15 psig for 45 min and cooled to room temperature. Then, the medium was pumped into the circulation tank. After inoculated with 150 mL seed, the

RFBB was operated under a rotational rate of 150 rpm in the first 48 h and increased to

300 rpm after the immobilization phase. The circulation also started at 48 h post-inoculation at a liquid recirculation rate of ~500 mL/min. The aeration was 1.5 vvm.

At the end of each batch, viscous broth was pumped out from the circulation tank before fresh medium pumped in.

5.3.3. Analytical Methods

5.3.3.1. Cell density

Cells were centrifuged to be removed from spent medium and resuspended to a proper dilution in distilled water. The optical density was measured by using a 113 spectrophotometer (Shimadzu UV-Visible Spectrophotometer) at 650 nm. The measured optical density was converted to dried cell weight according to a calibration curve.

5.3.3.2. Xanthan gum concentration

At the end of the fermentation, samples were centrifuged to remove cells and 15 mL broth was precipitated with 30 mL ethanol in a pre-weighed 50 mL centrifuge tube. After centrifugation to remove the supernatant, precipitated xanthan gum was washed with 5 mL ethanol and then freeze-dried via a freeze dryer (LABCONCO Freeze dry system with a LABCONCO Stopping tray dryer) over night. The tube containing dried xanthan gum was then weighed to calculate final product concentration.

5.3.3.3. Viscosity of xanthan gum

Broth was diluted to a proper concentration before measuring the viscosity. The viscosity was obtained via a Brookfield viscometer (RVTD II) (Brookfield, Raynham, MA) at a shear rate of 122.3 s-1 (spindle S at 100 rpm) under room temperature. The broth viscosity was the measured value multiplied by the dilution factor.

5.4. Results and discussion

5.4.1. Xanthan gum production in repeated-batch RFBB under liquid continuous mode

The liquid continuous repeated-batch used the same setup in the previous study (Yang et al., 1996) except a 1.85 times higher recirculational rate, due to a pump with higher capacity used. Figure 5.3 shows that the cells were grown and immobilized in the fibrous matrix. A very stable production could be repeated in at least 5 batches. The fermentation

114 was stopped after the fifth batch because of a mechanical problem, which led to a lower productivity. The production of each batch is summarized in Table 5.2. The results were then compared with the previous study (Table 5.3). It showed a lower final concentration, yield and productivity, which could be due to an uneven spray of xanthan gum broth to the fibrous bed. In this setup, a new nozzle (BETE MP125N 316, BETE Fog Nozzle, Inc.,

Greenfield, MA) was installed to replace the tubular sparger used in the previous study.

The tubular sparger was mainly a stainless tube with small holes on it. When the broth passed through the holes, it could not provide a satisfying and even distribution of broth on the inner surface of a fibrous bed. This new nozzle was intended to create a spray layer to cover most of the inner area of a fibrous bed. However, when viscosity increased in xanthan gum broth, the pressure provided by the pump could not support the fluid sprayed out of the nozzle. By operating with different pumps, such as peristaltic pumps and a gear pump, it only created a stream instead of a broth spray layer. When the viscosity increased, the broth was dripping out of the nozzle, instead of spraying. The stream (or dripping) contacted only a small area of the fibrous bed and resulted in an uneven distribution. Although gear pump is designed for viscous fluid transportation, the fermentation broth with bubbles could stop the operation of gear pump by the interruption of air in the fluid. Even with poor distribution of xanthan gum to the fibrous bed, the production was compatible to fermentations in stirred tanks (Tables 5.2 and 5.3).

Due to the poor distribution, the effect of increasing recirculation rate in the RFBB could not be correctly evaluated. If the nozzle is more effective, a higher productivity, yield, and final concentration are expected.

115 5.4.2. Xanthan gum production in repeated-batch RFBB under gas continuous mode

In the gas continuous repeated-batch, a 20 L fermentor was installed with a fibrous bed and connected to a 5 L circulation tank (Figure 5.2). Overall, three batches were conducted in this repeated-batch fermentation (Figure 5.4). The results were summarized and compared with the previous study in Table 5.4. The productivity in each batch was between 0.22 to 0.32 g/L/h (Table 5.2). Although the productivity here was lower than the productivity in Doma, 1999 and Yang et al, 1996 (Table 5.4), the process was scalable with doubling the liquid volume.

5.4.3. Oxygen transfer coefficient in RFBB under gas continuous mode

According to an empirical kLa model (Doma, 1999), kLa was proportional to rotational

~0.4 0.109 2.4 rate, gas flow rate, liquid flow rate, and reciprocal of liquid volume by ω , VG , VL ,

-1 and L . Therefore, based on the operational parameters (Table 5.4), the apparent kLa was

2.24 times larger in Doma’s experiment than in this study, which resulted in a higher productivity. The specific oxygen uptake rate affected the specific productivity of xanthan gum significantly, which was described in the follow empirical equation (Yang et al., 1996).

dCP  0.0312  0.0012q , q 2 > 5 mg O /g cell·h (5.1) O2 O 2 dt CX

When qO2 is less than 5 mg O2/g cell·h, the specific productivity is assumed to decrease to zero linearly. Compared to the gas continuous repeated-batch result and the gas continuous results from Doma (1999), the two fibrous beds were the same size, while the

116 liquid volume of this study was doubled. The overall cell density (Cx) in this study was thus only half of the cell density in Doma’s study.

Generally, the dissolved oxygen rate in a bioreactor can be expressed in the following equation: dC O2  k a(C*  C )  q C (5.2) dt L O2 O2 O2 X

Cell density (Cx) could be considered as a constant, because cells were in the stationary phase in repeated-batch fermentation. At this phase, oxygen mainly supports cell maintenance and xanthan gum biosynthesis. In addition, the dissolved oxygen changes in

dC a period of time, O2 , was negligible. Therefore, the equation can be simplified as dt below: k a(C*  C )  q C , (5.3) L O2 O2 O2 X

Which means oxygen transfer rate is equal to oxygen uptake rate.

Because kLa and overall cell density (Cx) in this study were about half of Doma’s study, the specific oxygen uptake rate in these two studies could be considered the same.

Because Cx from Doma was twice that of this study, productivity was calculated to be twice that of this study (Equation 5.1). Since productivity from Doma was actually 4 times of this study (Table 5.4), other aspects of the process should be investigated. Again, the ineffective spray from the new nozzle could be one of the reasons resulting in a low productivity in this study.

However, xanthan gum final concentration was less sensitive to kLa in RFBB (Table 5.4).

A similar final xanthan gum concentration at 2.5 % glucose was achieved although a 117 decrease of apparent kLa by 2.24 folds was used (Table 5.4). In conclusion, the decreased apparent kLa more significantly affected the productivity than the final concentration. By extrapolating the curves in Figure 5.6 to meet the operational parameters used in Doma’s and this study (Table 5.4), the apparent kLa in both fermentation batches were below 50 h-1.

5.4.4. RFBB scale up

The three most important things in scale up are: mass transfer, mixing, and shear. Mass transfer, especially oxygen transfer, is the most critical issue in xanthan gum fermentation.

Mass transfer can be addressed by reactor design, such as using gas sparger systems that balance oxygen mass-transfers efficiency with carbon dioxide removal, or in this case, using a rotating fibrous bed to increase the contact of gas, liquid and microorganisms.

The high gravity field created by the rotation of fibrous bed reduced the apparent viscosity of fermentation broth; it also contributed to increase in the oxygen transfer.

According to equation 5.1, the specific productivity of xanthan gum production depends on specific oxygen uptake rate.

To ensure a sufficient supply, the oxygen uptake rate considered in this chapter is based on the maximum specific oxygen uptake rate of X. campestris reported by Amanullah et al (1998), which was 30.4 mg O2/g cell·h. According to Yang et al. (1996), at the end of the repeated-batch of RFBB fermentation, ~7 g/L cells were immobilized on the fibrous matrix. When it multiplied by the maximum specific oxygen uptake rate of X. campestris reported by Amaullah et al., we can estimate the oxygen uptake rate (~210 mg O2/h) for the right hand side of equation 5.3. 118 In the left hand side of the equation 5.3, saturated oxygen concentration in xanthan gum solution is a function of temperature and xanthan gum concentration. Assume the final xanthan gum concentration was 75 g/L as reported, the saturated oxygen concentration can be calculated (~3.71 mg O2/L) based on a correlation curve in Figure 5.5 (Doma,

1999):

C*  7.5553 0.0513C . (5.4) O2 P

Assume dissolved oxygen tension (DOT) is maintained at 20 % of saturated oxygen in water at 25 ºC, which is 20 % of 8.6 mg O2/L (=1.72 mg O2/L).

Applying the saturated oxygen concentration in xanthan gum solution, the oxygen concentration in the xanthan gum solution and the maximum oxygen uptake rate (~210 mg O2/h), we can get an estimation of kLa required in RFBB to support a 75 g/L xanthan gum production, which should not be less than 105.64 h-1. Based on Figure 5.6, to reach a

-1 kLa larger than 105.64 h , the liquid flow rate should not be less than 2 L/min. To date, the repeated batch fermentation were still operated under a circulation rate of 500 mL/min. To improve the production, a higher liquid flow rate should be used.

Generally, mixing time can be addressed by optimizing the addition location and use of multipoint additions, increasing agitation rates and optimizing the impeller location and setup. When the mixing time is too long, the local environmental conditions in a fermentor could be too harsh for microorganism to survive. In viscous biopolymer fermentation, such as xanthan gum, mixing takes a much longer time than the fermentation processes with lower viscosity. The mixing time of a fibrous bed under gas continuous mode with a liquid flow rate of 500 mL/min and rotation rate of 300 rpm

119 could be calculated by extrapolation using Figure 5.7. The calculated mixing times are in a range of 20 to 45 seconds, which is acceptable, according to the accumulation of xanthan gum during fermentation.

Shear can be addressed through the use of shear protecting reagents and agitation.

Because of viscous xanthan gum forming a slimy protection layer surrounding the microorganisms, damage from shear is not observed in this kind of biopolymer fermentation. On the other hand, shear can reduce the apparent viscosity of a shear-thinning biopolymer and further increase oxygen transfer in the fermentation. In this case, a sufficiently high shear is desired to increase the production. In RFBB, shear is provided basically by the rotation of the fibrous bed and the circulation of the broth via a pump through a nozzle. Xanthan gum is a shear-thinning biopolymer, and its rheological properties could be modeled by Power law.

n1 a  K() (5.5)

The K and n are the function of xanthan gum concentration, and could be expressed by following equations (Doma, 1999).

0.364 n  0.551C p (5.6)

2 K  0.00187Cp 1.85Cp , 2.4 g/L ≤ Cp ≤ 75.4 g/L (5.7)

Applying Equations 5.6 and 5.7 into 5.5, we can get

0.364 2 (0.551C p 1) a  (0.00187C p 1.85C p )() (5.8) where  is proportional to the rotational rate of RFBB.

120 Therefore, the higher rotational rate could result in a higher oxygen transfer and thus improve the production. However, power consumption increased dramatically when the rotational rate was above 400 rpm (Figure 5.8). In an optimal rate should not only provide the maximum production, but also be energy efficient.

In addition, Doma (1999) simulated xanthan gum fermentation in RFBB under different rotational rates and recirculational rates. Based on his finding, xanthan gum fermentation had not reached the maximum theoretical product concentration, and could be improved by operating under higher rotational rate and liquid flow rate. For example, xanthan gum production could reach 100 g/L at 400 rpm rotation rate and 2.5 L/min liquid flow rate.

5.5. Conclusions and recommendation

Compared to free-cell fermentation, the immobilized cell fermentation in RFBB gave higher xanthan gum yield and volumetric productivity. The increased xanthan yield and productivity were attributed to the higher density cell growth through refreshing medium and retaining cells inside the bioreactor via cell immobilization. The improvement in xanthan yield and productivity due to cell immobilization was also observed within each repeated-batch by comparing the data from the second batch with the data from the initial batch. RFBB can be operated in repeated-batch and fed-batch modes with relatively similar performance.

This study demonstrated the feasibility and scalability of the RFBB for xanthan gum production. The process can be further improved by optimizing the reactor design and operation. Long-term performance and stability of RFBB needs to be tested with more consecutive batches. It is desirable to analyze the rheological properties of the xanthan 121 gum products from different fermentation systems. According to Amanullah et al. (1998), when the biomass was sufficient, glucose-based xanthan gum production could be operated without any nitrogen source. This strategy can also be used in our RFBB fermentation to reduce the usage of yeast extract and the fermentation cost.

5.6. Nomenclature

C : Oxygen concentration in xanthan gum solution (mg/L) O2

C* O2 : Saturated oxygen concentration in xanthan gum solution (mg/L)

CP : Xanthan gum concentration (g/L)

CX : Cell density in the rotating fibrous bed bioreactor (g cell/L)

K: Fluid Consistency coefficient (Nm-2·sn) n: Flow behavior index kLa : Volumetric oxygen transfer coefficient (hr-1)

qO2 : Specific oxygen uptake rate (mg O2/g cell·h)

3 VL : Recirculation rate of liquid (m /s)

L : Total liquid volume in RFBB (m3)

 : Shear rate (s-1)

μa: Apparent viscosity of xanthan gum (Pa·s)

ω : Rotational rate (rad/s)

122 5.7. References

Amanullah, A.; Satti, S.; Nienow, A. W. Enhancing xanthan fermentations by different modes of glucose feeding. Biotechnology Progress (1998), 14(2), 265-269.

Doma, B. T., A rotating fibrous bed bioreactor for xanthan gum fermentation. (1999), Doctoral dissertation at University of the Philipines, Diliman, Quezon City, Philipines.

Hsu, Chia-Hua; Lo, Y. Martin. Characterization of xanthan gum biosynthesis in a centrifugal, packed-bed reactor using metabolic flux analysis. Process Biochemistry (2003), 38(11), 1617-1625.

Hsu, C. H.; Chu, Y. F.; Argin-Soysal, S.; Hahm, T. S.; Lo, Y. M. Effects of surface characteristics and xanthan polymers on the immobilization of Xanthomonas campestris to fibrous matrices. Journal of Food Science (2004), 69(9), E441-E448

Lo, Y. M.; Hsu, C. H.; Yang, S. T.; Min, D. B. Oxygen transfer characteristics of a centrifugal, packed-bed reactor during viscous xanthan fermentation. Bioprocess and Biosystems Engineering (2001), 24(3), 187-193

Najafpour, G.D., Bioprocess scale-up. Biochemical Engineering and Biotechnology, Chapter 13 Elsevier B.V. 2007

Ramshaw, C.; Mallinson, R. H. Mass Transfer Process. U.S. Patent 4,283,255, 1981.

Rao, D. P.; Bhowal, A.; Goswami, P. S. Process Intensification in Rotating Packed Beds (HIGEE): An Appraisal. Industrial & Engineering Chemistry Research (2004), 43(4), 1150-1162 Trujillo-Roldán, M. A., Peña, C., Ramirez, O.T., Galindo, E., Effect of Oscillating Dissolved Oxygen Tension on the Production of Alginate by Azotobactervinelandii, Biotechnology Progress, (2001), 17, 6, 1042-1048

Yang, Shang-Tian; Lo, Yang-Ming; Min, David B. Xanthan gum fermentation by Xanthomonas campestris immobilized in a novel centrifugal fibrous- bed bioreactor. Biotechnology Progress (1996), 12(5), 630-637

Yang, Shang-Tian; Lo, Yang-Ming; Chattopadhyay, Devamita. Production of cell-free xanthan fermentation broth by cell adsorption on fibers. Biotechnology Progress (1998), 14(2), 259-264

Zhu, Shengdong; Wu, Yuanxin; Yu, Ziniu; Tong, Haibao; Cheng, Dachang; Xie, Decheng. Improving xanthan fermentation in a mechanically stirred aerated 123 fermenter via external loop. Research Journal of Microbiology (2006), 1(1), 70-75.

124

Design Figure 5.1 Figure 5.2 Modes Liquid continuous Gas continuous Gas continuous

Literatures Yang et al., 1996 Yang et al., 1996 Doma, 1999 Yang et al., 1998 Lo et al., 2001 (This study) Lo et al., 2001 Hsu and Lo, 2003 Hsu and Lo, 2003 (This study)

Table 5.1. Summary of RFBB fermentation setups and operation modes in literatures.

125

Cell density Xanthan gum Xanthan gum Productivity # (g/L) conc. (g/L) yield (%) (g/L/h) Free cell 1.2 – 2.0 17.0 – 25.0 0.69 – 0.8 0.18

5 L RFBB B1 0.30 6.9 0.39 0.15 Liquid B2 0.12 22.0 0.79 0.37 continuous B3 0.22 29.6 0.74 0.46 (Figure 5.1) B4 0.49 29.4 0.75 0.35 B5 0.47 31.8 0.82 0.21 20 L RFBB B1 0.29 16.1 0.68 0.22 Gas B2 0.41 22.3 0.72 0.30 continuous B3 0.23 19.7 0.85 0.32 (Figure 5.2) FB 0.81 75.4 0.83 0.71 Table 5.2. Comparison of xanthan gum production in free-cell and RFBB fermentations FB: fed-batch

126

Yang et al., 1996 This study Liquid volume 5 L 4 L Liquid flow rate ~270 mL/min ~500 mL/min Rotational rate 350 rpm 350 rpm Gas flow rate 1.5 vvm 1.5 vvm productivity 0.7 g/L/h (5 % glucose); 0.21 g/L/h (5 % glucose) 0.5 g/L/h (2.5 % glucose) 0.39 g/L/h (2.5 % glucose)* Yield 0.9 (5 % glucose); 0.82 (5 % glucose) (g xanthan/g 0.85 (2.5 % glucose) 0.76 (2.5 % glucose)* glucose) Final 35 g/L (5 % glucose); 31.8 g/L (5 % glucose) concentration 25 g/L (2.5 % glucose) 27 g/L (2.5 % glucose)*

Table 5.3. Comparison of RFBB fermentation in 5 L liquid continuous repeated batches. *: average of batch 2, 3, and 4.

127

Yang et al., 1996 Doma, 1999 This study Setup Figure 5.1 (5 L Figure 5.2 (20 L Figure 5.2 (20 L reactor) reactor) reactor) Liquid 2.5 L 2 L 4 L volume Liquid flow ~270 ml/min ~500 ml/min ~500 ml/min rate Rotational 350 rpm 400 rpm 300 rpm rate Gas flow rate 1.5 vvm 1-2 vvm 1.5 vvm

Productivity ~1 (5 % glucose); ~1.2 (5 % glucose); 0.30 (5 % glucose); ~0.9 (2.5 % glucose) ~0.96 (2.5 % glucose) 0.32 (2.5 % glucose) Yield ~0.8 (5 % glucose); ~0.96 (5 % glucose); 0.72 (5 % glucose); (g xanthan/ g ~0.9 (2.5 % glucose) ~0.9 (2.5 % glucose) 0.85 (2.5 % glucose) glucose) Final 35 g/L (5 % glucose); 35 g/L (5 % glucose); 22.3 g/L (5 % glucose); concentration 25 g/L (2.5 % 20 g/L (2.5 % glucose) 19.7 g/L (2.5 % glucose) glucose) Table 5.4. Comparison of RFBB fermentation in gas continuous repeated batches.

128

Figure 5.1. A Rotating fibrous bed bioreactor (arrows indicate liquid flow direction). (Yang et al., 1996)

129

Controller Motor

Thermometer

pH probe

Air DO probe

Motor Temperature controller Recirculating RFBB vessel Peristaltic pumps

Figure 5.2. A Rotating fibrous bed bioreactor system (Doma, 1999)

130

45 5 5-liter RFBB, Liquid continuous Xanthan gum 40 Repeated Batch Glucose Cell 35 4

30 3 25

20 2 (g/L) Cell 15

Glucose, Xanthan Gum (g/L) XanthanGum Glucose, 10 1

5

0 0 0 50 100 150 200 250 300 350 Time (h)

Figure 5.3. Repeated-batch fermentations for xanthan gum production in a rotating fibrous bed bioreactor operated under liquid continuous conditions.

131

60 5 20-liter RFBB, gas continuous Repeated batch Xanthan gum 50 Glucose 4 Cells

40 3

30

2 (g/L) Cell 20

Glucose, Xanthan Gum (g/L) XanthanGum Glucose, 1 10

0 0 0 20 40 60 80 100 120 140 160 180 200 Time (h)

Figure 5.4. Repeated-batch fermentations for xanthan gum production in a rotating fibrous bed bioreactor operated under gas continuous conditions.

132

Figure 5.5. Solubility of oxygen in xanthan gum solution (Doma, 1999)

133

Fermentation in Table 5.4 using 20 L reactor were operated at 0.5 L/min recirculation rate

Figure 5.6. Effect of recirculation rate on kLa at different rotational rate, with 20 g/L xanthan gum, air flow rate 3.5 L/min, and liquid volume of 3 L (Doma, 1999).

134

Figure 5.7. Average residence time at different rotational rate, liquid flow rate and xanthan gum concentration (ungassed; 200 rpm □, 400 rpm ○, 600 rpm Δ, 800 rpm ■)

(Doma, 1999)

135

Figure 5.8. Power consumption at different rotational rate and different volumetric air flow rate. (volumetric air flow rate 5 L/min □, 7 L/min■) (Doma, 1999).

136

Chapter 6: Scale-up and Economic Evaluation of an Integrated Rotating Fibrous Bed Bioreactor-Ultrafiltration Process for Viscous Biopolymer Fermentation

6.1. Summary

In this chapter, a rotating fibrous bed bioreactor integrated with ultrafiltration process was scaled up and compared to the commercial stirred tank reactor process. The economic analysis was conducted via SuperPro Designer software. Xanthan gum production was used as a representative of viscous biopolymer fermentation. The operational and scale-up parameters were based on our previous experimental data in lab-scale reactors and ultrafiltration. The results show that the integrated RFBB-UF process can reduce the production cost up to 65% compared to the commercial process. At 3000 tons xanthan gum per year, the capital investment is reduced from $19,882,563 to $19,028,481,

Equipment cost (smaller size fermentor), raw material cost (mainly from alcohol reduction) and the energy cost is reduced by 2.5%, 52%, and 6.6%, respectively. In addition, water usages and waste water generation can be reduced by up to 80%, and 90% reduction in solid biomass waste can be achieved with the integrated process.

6.2. Introduction

The market of an important food thickening additive, xanthan gum, is larger than

75,000,000 kg/yr worldwide. Currently, this biopolymer is produced in stirred tanks and recovered by alcohol precipitation. Because of the highly viscous broth during

137 fermentation, the current process with poor oxygen transfer limits the yield, concentration and productivity. To solve this problem, a RFBB-UF process has been studied in the previous chapters. The integrated process has been demonstrated to increase the final product concentration, yield, and productivity. To further estimate the feasibility and benefits of producing xanthan gum using RFBB-UF integrated process, an economic analysis was conducted based on a plant with an annual production of 3000 tons.

The objective of this chapter was to apply the optimized operational parameters of RFBB and ultrafiltration in large-scale xanthan gum production to meet the annual worldwide demand. To study the advantage of RFBB-UF integrated process, an economic analysis was conducted to compare large-scale xanthan gum production using the integrated process with the current industrial process.

6.3. Materials and methods

6.3.1. Process flowsheet

There are four processes compared in this chapter, including stirred tank reactor-distillation (STR-D) (Figure 6.1), stirred tank reactor-ultrafiltration-distillation

(STR-UF-D) (Figure 6.2), rotating fibrous bed bioreactor-distillation (RFBB-D) (Figure

6.3), and rotating fibrous bed bioreactor-ultrafiltration-distillation (RFBB-UF-D) (Figure

6.4). STR-D is a stirred tank bioreactor followed by alcohol precipitation and alcohol recycling using a distillation column. STR-UF-D is a process similar to STR-D but an extra ultrafiltration is inserted between fermentation and precipitation. In addition, spent medium is recycled from ultrafiltration permeate. RFBB-D is similar to STR-D but using a rotating fibrous bed bioreactor instead of a stirred tank. RFBB-D with an extra 138 ultrafiltration between the fermentor and precipitation is RFBB-UF-D. In RFBB-UF-D process, spent medium is also recycled. In all four processes, a distillation column was used to recover the alcohol to reduce the cost of alcohol. The process flowsheet was drawn via SuperPro (Intelligen, Inc.).

6.3.2. Economic analysis

The economic analysis was conducted via SuperPro Designer (Intelligen, Inc.). The operational parameters were determined by the optimal conditions for cell adsorption, oxygen transfer and power consumption in the 5 L lab scale stirred tank bioreactor and

RFBB. In this study, Xanthomonas campestris NRRL 1459 was cultured in the medium containing 30 g/L glucose, 3 g/L yeast extract, 2 g/L K2HPO4, 0.1 g/L MgSO4, 500 ppm antifoam A, in tap water at pH 7.

The parameters and assumptions in the economic analysis are as follows.

 Temperature - ~28oC

 Aeration rate – 1 vvm

 Agitation rate – 500 to 1000 rpm

 Fermentation time – 50 h (RFBB) to 80 h (STR)

 Final xanthan gum concentration – 25 g/L (STR) to 35 g/L (RFBB)

 Product yield - ~85% from glucose

 Reactor volumetric productivity – 0.5 g/L/h (STR) to 1.0 g/L/h (RFBB)

Table 6.1 summarizes the process performance used in this study based on historical experimented data.

139 6.3.3. Equipment cost

Equipment cost was estimated according to the following equation:

CapacityA EquipmentcostA EquipmentcostB( )n , CapacityB where index n varies with equipment type. When n equals to 0.6, it is called six-tenth-factor rule. Equipment costs of fermentor and blending tank were estimate based on a 200,000 L stirred tank of $250,000, and the index was 0.53. Ultrafiltration was estimated based on a capacity of 370 ton/day system of $750,000. The membrane was estimated by $200/m2. The base cost of precipitation tank was 120,000 L tank costing

$100,000. Cost of distillation column was estimated according to a 7323 L distillation column costing $250,000. The cost of a spray dryer with 358 kg/h capacity was estimated to be $267,100. Pumps used to transport liquids were estimated at $25,000 each. Other costs including piping and supporting equipments.

6.4. Results and discussion

6.4.1. Major equipment specification and cost

Tables 6.2 and 6.3 summarize the major equipment sizes and costs in four processes.

Because of the higher productivity and final concentration achieved in RFBB, fewer fermentors are required to produce 3000 tons/year xanthan gum. When ultrafiltration is introduced to the process to concentrate xanthan gum before alcohol precipitation, it can reduce the working volume in precipitation and also reduce the size of precipitation tank.

The advantage of reducing the working volume by ultrafiltration also shows in the size reduction of distillation column. 140 6.4.2. Cost of raw material and utility to produce 3000 tons/year xanthan gum

Tables 6.4 and 6.5 compare the cost of raw material in four different processes. Because the productivity of processes using RFBB is 2-fold of the processes using stirred tanks, the working volume of RFBB is half and less raw materials, such as yeast extract, salts, and water are required. Processes using ultrafiltration before alcohol precipitation can recycle the spent medium from ultrafiltration permeate. By recycling the spent medium, the cost of water can be reduced ~75%.

Annual utility in four processes are compared in Table 6.6. Electricity is based on a 30 million kWh annual requirement of a 3000 tons/yr xanthan gum production in a conventional plant. The energy required to operate ultrafiltration is 800,000 kWh. The electricity cost is based on current market price of $0.05/kWh. Steam provides the energy for ethanol evaporation in distillation. The energy required for ethanol evaporation is about 836.98 kJ/kg. The latent heat of evaporation of steam is 970.6 BTU/lb (equals to

2255.58 kJ/kg). Concentrating xanthan gum before precipitation can significantly reduce the alcohol required in distillation and the energy consumption in distillation by more than 75%-80%. Natural gas provides the heat to remove the water in the drying of xanthan gum product. About 20% w/v xanthan gum could be obtained after precipitation.

The energy required to remove the moisture in the xanthan gum product is based on the heat required to evaporate 12,000 tons water annually.

141 6.4.3. Economic analysis

6.4.3.1. Direct fixed capital (DFC)

Direct fixed capital was estimated by the sum of total plant cost (direct and indirect), contractor's fee, and contingency, and is summarized in Table 6.7. The direct fixed capital

(DFC) of the integrated RFBB-UF process plant producing 3000 tons/year xanthan gum is ~$17.7 million, which is about 97% of the DFC of a conventional xanthan gum plant with the same production capacity.

6.4.3.2. Annual operating cost

The major operating costs include raw materials, utilities, labors, and facility-dependent costs, such as depreciation and maintenance charges. As can be seen in Table 6.8, the raw material cost is the most important factor accounting for 43% - 63% of the product cost.

Using the integrated RFBB-UF process can reduce the operating cost by ~35% compared to the conventional process (STR-D).

6.4.3.3. Profitability analysis

As shown in Table 6.9, the total investment of an integrated RFBB-UF process

(RFBB-UF-D: $19 millions) is about 95% of the conventional process (STR-D: ~$19.9 millions). The unit production costs of the four processes are $4.07 (STR-D), $3.12

(STR-UF-D), $3.29 (RFBB-D), and $2.65 (RFBB-UF-D). With a selling price of $10/kg, one month working capital, and 10 years depreciation, the return on investment (ROI) is

47-78%. If the ROI is set to be 20%, we can suppress the selling price to $4.77/kg in an integrated RFBB-UF process.

142 6.4.4. Reducing environmental impact from liquid, solid waste disposal and emission by

using integrated RFBB-UF process for xanthan gum production

The integrated RFBB-UF process gives a great opportunity to reuse the high density biomass in repeated batch fermentations and spent medium recycle. The high-concentration xanthan gum broth going into the alcohol precipitation reduces the amount of alcohol required in the recovery step, and lowers the emission level in distillation in alcohol recovery.

Generally, fermentation generates an enormous amount of waste liquid stream after product recovery. The waste stream from spent medium contains high BOD and pollutes the environment to dispose it. Recycling the ultrafiltration permeate from xanthan gum recovery allows us to reuse the water and salts in the medium and reduce the amount of waste liquid stream. Based on our finding, 75% replacement could result in comparable final xanthan gum concentration, yield and higher productivity. As shown in Figure 6.5, by recycling the UF permeate, we can save 75% of water used in the fermentation and therefore also reduce 75% of the waste water. Compared to STR, the RFBB also can reduce water use by ~28% because high xanthan gum concentration can be produced in the fermentation. With the integrated RFBB-UF, water use can be reduced by ~82% compared to the conventional process.

Generally, biomass in stirred-tank fermentation is about 4 g/L and in RFBB most of the cells were immobilized on the fibrous matrix and the cell in broth was about 0.5 g/L.

When repeated batch is performed in a RFBB, the biomass stays in the RFBB and can be reused. The solid waste disposal can thus be greatly reduced by 87 % (Figure 6.6).

143 6.5. Conclusions and recommendation

The process of integrated RFBB-UF for xanthan gum production was economically favorable with a production rate of 3000 ton/yr. By adding ultrafiltration, the cost reduction in ethanol is about 5.7-fold, and in utilities is about 17%. The number of fermentor required can be reduced by half when using the RFBB, which is a more efficient oxygen transfer design compared to stirred tanks. The unit production cost is

35% less in integrated RFBB-UF-D process compared to the conventional STR-D process. In addition, the cell-free broth from the RFBB can eliminate the biomass in the xanthan gum, resulting in less cell debris present in the broth. Therefore, a higher quality and purity of xanthan gum can be achieved. Overall, RFBB-UF-D is the most economic and favorable process for large-scale production of xanthan gum.

144

STR-D STR-UF-D RFBB-D RFBB-UF-D Fermentation: Xanthan gum yield from glucose 0.85 g/g 0.85 g/g 0.85 g/g 0.85 g/g Reactor productivity 0.5 g/L/h 0.5 g/L/h 1 g/L/h 1 g/L/h Final xanthan concentration 25 g/L 25 g/L 35 g/L 35 g/L Ultrafiltration Product recovery yield (%) - 95 -- 95 Final xanthan concentration 150 g/L 150 g/L Alcohol Precipitation Product recovery yield (%) 90 90 90 90 Overall product yield (%) 76.5 72.7 76.5 72.7

Table 6.1 Process performance based on historical experimental data.

145 Equipment cost STR-D Number Size Unit cost Cost Blending tank 2 158,730 L 176,900 353,800 Fermentor 8 200,000 L 250,000 2,000,000 Precipitation tank 1 119,048 L 100,000 100,000 Distillation tower 1 7323 L 250,000 250,000 Pump 2 25,000 50,000 Spray Dryer 1 358 kg/h 267,100 267,100 Others $ 1,057,315 Total $ 4,078,215 Equipment cost STR-UF-D Number Size Unit cost Cost Blending tank 2 167,084 L 181,800 363,600 Fermentor 9 200,000 L 250,000 2,250,000 Ultrafiltration 1 47.7 m2 1,050,000 1,050,000 Precipitation tank 1 20,886 L 39,600 39,600 Distillation tower 1 1,285 L 99,400 99,400 Pump 2 25,000 50,000 Spray Dryer 1 358 kg/h 267,100 267,100 Others $ 1,441,895 Total $ 5,561,595 Table 6.2. Major equipment specifications and costs of STR-D and STR-UF-D processes (Production rate of 3000 tons/year).

146 Equipment cost RFBB-D Number Size Unit cost Cost Blending tank 2 70,862 L 115,400 230,800 Fermentor 4 200,000 L 375,000 1,500,000 Precipitation tank 1 85,034 L 83,300 83,300 Distillation tower 1 5,231 L 209,200 209,200 Pump 2 25,000 50,000 Spray Dryer 1 358 kg/h 267100 267,100 Others $ 819,140 Total $ 3,159,540 Equipment cost RFBB-UF-D Number Size Unit cost Cost Blending tank 2 74,591 L 118,600 237,200 Fermentor 4 200,000 L 375,000 1,500,000 Ultrafiltration 1 34.1 m2 750,000 750,000 Precipitation tank 1 20,886 L 39,600 39,600 Distillation tower 1 1,285 L 99,400 99,400 Pump 2 25,000 50,000 Spray Dryer 1 358 kg/h 267100 267,100 Others $ 1,030,155 Total $ 3,973,455 Table 6.3. Major equipment specifications and costs of RFBB-D and RFBB-UF-D processes (Production rate of 3000 tons/year).

.

147 Raw Material (per Year) STR-D STR-UF-D Annual amount Unit cost Cost ($/yr) Annual amount Unit cost Cost ($/yr) Glucose 3,921,569 kg 0.5 /kg 1,960,784 4,127,967 kg 0.5 /kg 2,063,983 Yeast extract 400,000 kg 2 /kg 800,000 421,053 kg 2 /kg 842,105 K2HPO4 266,667 kg 0.5 /kg 133,333 280,702 kg 0.5 /kg 140,351 MgSO4.7H2O 13,333 kg 0.1 /kg 1,333 14,035 kg 0.1 /kg 1,404 NaOH 177,778 kg 0.45 /kg 80,000 187,135 kg 0.45 /kg 84,211 Antifoam 66,667 L 0.015 /L 1,000 70,175 L 0.015 /L 1,053 Water 133,333,333 L 0.000544 /kg 72,533 140,350,877 L 0.00054 /kg n/a Water consumed n/a L n/a /kg n/a 35,087,719 L 0.00054 /kg 19,088 14 Ultrafiltration

8 2 2

membrane n/a n/a n/a 95.47 m 200 /m 19,095

KCl 1,333,333 kg 0.5 /kg 666,667 233,918 kg 0.5 /kg 116,959

Alcohol (EtOH) 105,200,000 kg 0.75 /kg n/a 18,456,140 kg 0.75 /kg n/a Alcohol consumed 5,260,000 kg 0.75 /kg 3,945,000 922,807 kg 0.75 /kg 692,105 Total $7,660,651 $3,980,354 Table 6.4. Raw material costs in STR-D and STR-UF-D processes (Production rate of 3000 tons/year).

148

Raw Material (per Year) RFBB-D RFBB-UF-D Annual amount Unit cost Cost ($/yr) Annual amount Unit cost Cost ($/yr) Glucose 3,921,569 kg 0.5 /kg 1,960,784 4,127,967 kg 0.5 /kg 2,063,983 Yeast extract 285,714 kg 2 /kg 571,429 300,752 kg 2 /kg 601,504 K2HPO4 190,476 kg 0.5 /kg 95,238 200,501 kg 0.5 /kg 100,251 MgSO4.7H2O 9,524 kg 0.1 /kg 952 10,025 kg 0.1 /kg 1,003 NaOH 126,984 kg 0.45 /kg 57,143 133,668 kg 0.45 /kg 60,150 Antifoam 47,619 L 0.015 /L 714 50,125 L 0.015 /L 752 Water 95,238,095 L 0.000544 /kg 51,810 100,250,627 L 0.00054 /kg n/a Water

14 recycled n/a L n/a /kg n/a 25,062,657 L 0.00054 /kg 13,634

9 Ultrafiltration membrane n/a n/a n/a 34.09 m2 200 /m2 6,820 KCl 952,381 kg 0.5 /kg 476,190 233,918 kg 0.5 /kg 116,959 Alcohol (EtOH) 75,142,857 kg 0.75 /kg n/a 18,456,140 kg 0.75 /kg n/a Alcohol recycled 3,757,143 kg 0.75 /kg 2,817,857 922,807 kg 0.75 /kg 692,105

Total $6,032,118 $3,657,161 Table 6.5. Raw material costs in RFBB-D and RFBB-UF-D processes (Production rate of 3000 tons/year).

149 STR-D STR-UF-D RFBB-D RFBB-UF-D

Electricity (KWH) 30,000,000 30,800,000 30,000,000 30,800,000 $0.05/KWH $1,500,000 $1,540,000 $1,500,000 $1,540,000

Distillation steam (kg) 39,036,654 6,848,536 27,883,324 6,848,536 $0.005/kg $195,183 $34,243 $139,417 $34,243

Natural gas (MMBTU) 25,654 25,654 25,654 25,654 $4.52/MMBTU $115,958 $115,958 $115,958 $115,958

Table 6.6. Annual utilities requirement and costs for 3000 tons xanthan gum production

1

50

150 FIXED CAPITAL ESTIMATE SUMMARY (2011 prices) A. TOTAL PLANT DIRECT COST (TPDC) (physical cost) STR-D STR-UF-D RFBB-D RFBB-UF-D 1. Equipment Purchase Cost $ 4,078,215 $ 5,561,595 $ 3,159,540 $ 3,973,455 2. Installation $ 1,631,286 $ 2,224,638 $ 1,263,816 $ 1,589,382 3. Process Piping $ 1,223,465 $ 1,668,479 $ 947,862 $ 1,192,037 4. Instrumentation $ 611,732 $ 834,239 $ 473,931 $ 596,018 5. Insulation $ 203,911 $ 278,080 $ 157,977 $ 198,673 6. Electricals $ 815,643 $ 1,112,319 $ 631,908 $ 794,691 7. Buildings $ 1,631,286 $ 2,224,638 $ 1,263,816 $ 1,589,382 8. Yard Improvement $ 407,822 $ 556,160 $ 315,954 $ 397,346 9. Auxiliary Facilities $ 2,446,929 $ 3,336,957 $ 1,895,724 $ 2,384,073 TPDC = $ 13,050,288 $ 17,797,104 $ 10,110,528 $ 12,715,056

B. TOTAL PLANT INDIRECT COST (TPIC) 10. Engineering $ 1,631,286 $ 2,224,638 $ 1,263,816 $ 1,589,382 11. Construction $ 1,427,375 $ 1,946,558 $ 1,105,839 $ 1,390,709 TPIC = $ 3,058,661 $ 4,171,196 $ 2,369,655 $ 2,980,091

C. TOTAL PLANT COST (TPDC+TPIC) TPC = $ 16,108,949 $ 21,968,300 $ 12,480,183 $ 15,695,147

12. Contractor's fee $ 407,822 $ 556,160 $ 315,954 $ 397,346 13. Contingency $ 1,631,286 $ 2,224,638 $ 1,263,816 $ 1,589,382 (12+13) = $ 2,039,108 $ 2,780,798 $ 1,579,770 $ 1,986,728 D. DIRECT FIXED CAPITAL (DFC) TPC+12+13 = $ 18,148,057 $ 24,749,098 $ 14,059,953 $ 17,681,875 Table 6.7. Fixed capital estimate of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (Production rate of 3000 tons/year).

151 ANNUAL OPERATING COST - SUMMARY (2011 prices) Cost Item STR-D STR-UF-D RFBB-D RFBB-UF-D $/Year % $/Year % $/Year % $/Year %

Raw Materials 7,660,651 62.72 3,980,354 42.50 6,032,118 61.09 3,657,161 46.01

Labor-Dependent 201,600 1. 65 231,000 2.47 117,600 1.19 126,000 1.59

Facility-Dependent 2,540,728 20.80 3,464,874 36.99 1,968,393 19.94 2,475,462 31.14

Laboratory/QC/QA - 0.00 - 0.00 - 0.00 - 0.00

Consumables - 0.00 - 0.00 - 0.00 - 0.00 Waste Treatment/Disposal - 0.00 - 0.00 - 0.00 - 0.00

Utilities 1,811,142 14.83 1,690,201 18.05 1,755,375 17.78 1,690,201 21.26

Transportation - 0.00 - 0.00 - 0.00 - 0.00

Miscellaneous - 0.00 - 0.00 - 0.00 - 0.00 Advertising and Selling - 0.00 - 0.00 - 0.00 - 0.00

Running Royalties - 0.00 - 0.00 - 0.00 - 0.00 Failed Product Disposal - 0.00 - 0.00 - 0.00 - 0.00

TOTAL 12,214,121 100 9,366,429 100 9,873,486 100 7,948,824 100 Table 6.8. Annual operating costs of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (Production rate of 3000 tons/year).

152 PROFITABILITY ANALYSIS (2011 prices) STR-D STR-UF-D RFBB-D RFBB-UF-D A. DIRECT FIXED CAPITAL $ 18,148,057 $ 24,749,098 $ 14,059,953 $ 17,681,875 B. WORKING CAPITAL $ 827,104 $ 498,196 $ 674,767 $ 462,513 C. STARTUP COST $ 907,403 $ 1,237,455 $ 702,998 $ 884,094 D. UP-FRONT R&D $ 0 $ 0 $ 0 $ 0 E. UP-FRONT ROYALTIES $ 0 $ 0 $ 0 $ 0 F. TOTAL INVESTMENT (A+B+C+D+E) $ 19,882,563 $ 26,484,748 $ 15,437,718 $ 19,028,481 G. INVESTMENT CHARGED TO THIS PROJECT $ 19,882,563 $ 26,484,748 $ 15,437,718 $ 19,028,481 H. REVENUE STREAM FLOWRATES kg/year 3,000,000 3,000,000 3,000,000 3,000,000 I. PRODUCTION (UNIT) COST ($/kg) $ 4.07 $ 3.12 $ 3.29 $ 2.65 J. SELLING/PROCESSING PRICE ($/kg) $ 10 $ 10 $ 10 $ 10 K. REVENUES ($/year) $ 30,000,000 $ 30,000,000 $ 30,000,000 $ 30,000,000 L. ANNUAL OPERATING COST $ 12,214,121 $ 9,366,429 $ 9,873,486 $ 7,948,824 M. GROSS PROFIT (K-L) $ 17,7 85,879 $ 20,633,571 $ 20,126,514 $ 22,051,176 N. TAXES (40 %) $ 7,114,352 $ 8, 253,429 $ 8,050,606 $ 8,820,470 O. NET PROFIT (M-N + Depreciation ) $ 12,659,784 $ 12,380,143 $ 12,075,908 $ 13,230,705 GROSS MARGIN 145.62% 220.29% 203.84% 277.41% RETURN ON INVESTMENT 63.67% 46.74% 78.22% 69.53% PAYBACK TIME(years) 1.57 2.14 1.28 1.44 Table 6.9. Profitability analysis of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (Production rate of 3000 tons/year).

153

EXECUTIVE SUMMARY (2011 prices) STR-D STR-UF-D RFBB-D RFBB-UF-D TOTAL CAPITAL INVESTMENT $19,882,563 $26,484,748 $15,437,718 $19,028,481 CAPITAL INV. CHARGED TO THIS PROJECT $19,882,563 $26,484,748 $15,437,718 $19,028,481 OPERATING COST ($/year) $12,214,121 $9,366,429 $9,873,486 $7,948,824 PRODUCTION RATE (kg/year) 3,000,000 3,000,000 3,000,000 3,000,000 UNIT PRODUCTION COST ($/kg) $4.07 $3.12 $3.29 $2.65 TOTAL REVENUES $30,000,000 $30,000,000 $30,000,000 $30,000,000 GROSS MARGIN 145.62 % 220.29 % 203.84% 277.41% RETURN ON INVESTMENT 63.67 % 46.74% 78.22% 69.53% PAYBACK TIME (year) 1.57 2.14 1.28 1.44 Table 6.9. Executive summary of STR-D, STR-UF-D, RFBB-D, and RFBB-UF-D processes (Production rate of 3000 ton/year).

154 medium P-3 / ST-101 S-104 Heat Sterilization gas out water P-4 / AF-101 S-105 Air Filtration S-103 P-1 / V-101 S-106 Blending / Storage P-2 / PFF-101 P&F Filtration air

P-6 / G-101S-108 P-7 / AF-102 S-101 Gas Compression S-109 Air Filtration

P-5 / V-102 P-8 / V-103 STR Blending / Storage

S-115 P-10 / MX-102 Mixing S-110 S-112 P-12 / C-101 S-125 Distillation S-102

S-111

S-126 P-13 / SDR-101 S-116 Spray Drying S-114 S-107 P-11 / MX-101 EtOH DS-103 Mixing

DS-104

DS-101 DS-102 DS-106 DS-107 P-19 P-15 / FL-101 P-16 / LB-101 Filling P-18 / BX-101 Truck (Discrete) Labeling DS-105 P-9 / CR-101 Packaging precipitation

Figure 6.1. Process diagram of a conventional xanthan gum production process (STR-D).

155 medium P-3 / ST-101 S-104 Heat Sterilization gas out water P-4 / AF-101 S-105 Air Filtration S-103 P-1 / V-101 S-106 Blending / Storage P-2 / PFF-101 P&F Filtration air

P-6 / G-101S-108 P-7 / AF-102 S-101 Gas Compression S-109 Air Filtration

P-5 / V-102 P-8 / V-103 STR Blending / Storage

S-115 P-10 / MX-102 S-110 Mixing S-112 P-12 / C-101 S-125 Distillation S-102

S-111 S-117

S-126 P-13 / SDR-101 S-118 S-116 Spray Drying S-114 S-107 P-14 / UF-101 P-11 / MX-101 EtOH Ultrafiltration DS-103 Mixing

DS-104

DS-101 DS-102 DS-106 DS-107 P-19 P-15 / FL-101 P-16 / LB-101 Filling P-18 / BX-101 Truck (Discrete) Labeling DS-105 P-9 / CR-101 Packaging precipitation

Figure 6.2. Process diagram of a conventional xanthan gum production process with ultrafiltration (STR-UF-D).

156 Medium P-3 / ST-101 S-104 Heat Sterilization gas out water P-4 / AF-101 S-105 Air Filtration S-103 P-1 / V-101 S-106 Blending / Storage P-2 / PFF-101 P&F Filtration air

S-108 P-6 / G-101 S-101 P-7 / AF-102 S-109 Gas Compression Air Filtration

P-5 / V-102 P-8 / V-103 RFBB Blending / Storage

S-115 P-10 / MX-102 Mixing S-110 S-112 P-12 / C-101 S-125 Distillation S-102

S-111

S-126 P-13 / SDR-101 S-116 Spray Drying S-114 S-107 P-11 / MX-101 EtOH DS-103 Mixing

DS-104

DS-101 DS-102 DS-106 DS-107 P-19 P-15 / FL-101 P-16 / LB-101 Filling P-18 / BX-101 Truck (Discrete) Labeling DS-105 P-9 / CR-101 Packaging precipitation

Figure 6.3. Process diagram of a xanthan gum production process using RFBB

(RFBB-D).

157 medium

water S-104 S-103 P-1 / V-101 gas out Blending / Storage P-3 / ST-101 P-4 / AF-101 S-105 S-106 P-2 / PFF-101 Heat Sterilization Air Filtration P&F Filtration

air S-108 P-7 / AF-102 S-101 S-109 P-6 / G-101 Air Filtration Gas Compression P-5 / V-102 P-8 / V-103 RFBB Blending / Storage

S-116 P-10 / MX-102 Mixing S-112 P-12 / C-101 S-125 Distillation S-102 S-110

S-111

P-13 / SDR-101

S-118 Spray Drying S-114 S-107 P-11 / MX-101 DS-103 Mixing S-117 EtOH S-115 DS-104

DS-101 DS-102 DS-106 DS-107 P-19 P-15 / FL-101 P-16 / LB-101 P-17 / UF-101 Filling P-18 / BX-101 Truck (Discrete) Labeling DS-105 P-9 / CR-101 Ultrafiltration Packaging precipitation Figure 6.4. Process diagram of an integrated RFBB-UF xanthan gum production process

(RFBB-UF-D).

158

140 kg)

6 120

100

80

60

40

20 Annual amount of water (10 water of amountAnnual 0 STR STR-UF RFBB RFBB-UF

Figure 6.5. Comparison of annual water usages for different xanthan gum production processes with an annual produtction rate of 3000 tons.

159 600

500

400

300

200

100 Solid waste from biomass (ton/yr)biomassfrom waste Solid 0 STR RFBB

Figure 6.6 Comparison of solid biomass wastes generated from free-cell (STR) and immobilized-cell (RFBB) fermentations with an annual produtction rate of 3000 tons.

160

Chapter 7: Conclusions and Recommendations

7.1. Conclusions

The overall goal of this study was to develop an economical and sustainable fermentation-ultrafiltration process for producing xanthan gum from whey lactose.

In the first aim, the feasibility of using whey lactose as an alternative carbon source for xanthan gum production was studied by comparing the final product concentration, yield, productivity, and average molecular weight of xanthan gum from glucose, galactose, lactose hydrolysate, and hydrolyzed whey permeate (HWP). The results showed that

HWP could be used as an alternative carbon source for xanthan gum production. By using HWP, a higher final product concentration, yield and productivity were achieved compared to the fermention using glucose. The amount of yeast extract could be reduced by 67 % when HWP was used as the fermentation substrate. Based on our investigation, the yield of xanthan gum produced from hydrolyzed whey permeate was about 80%. The

BOD of 40 % whey permeate is about 256,000 mg O2/L. According to the United States regulation, the maximum concentration of BOD discharges is 250 mg/L, with a surcharge fee varying in different cities. For example, the surcharge of BOD in Columbus, Ohio, in

2011 is $0.66-0.74/pound. Thus, to produce 3,000,000 kg/yr of xanthan from hydrolyzed whey permeate, it would use ~9,375,000 L/yr of 40 % whey permeate (=2,476,613 gallon/yr). We can save the chesse industry ~$3,800,000 annually from disposing 2.4 million gallons of whey permeate on BOD surcharge. The calculation is shown bellow: 161 2.4million gallon/yr× (256,000-250) mg O2/L ×8.34 lb/gallon ×$0.74/lb = $3,788,128/yr

Instead of paying large amount of money to dispose whey permeate, we succesfully developed the process to convert the whey permeate into a value-added product, xanthan gum.

The feasibiltiy of recycling spent medium from xanthan gum fermentation was demonstrated in Chpater 4. The results showed that ultrafiltration can not only recover and concentrate xanthan gum, but also recycle the medium for next fermentation run.

Ultrafiltration permeate (up to 75 % replacement) from xanthan gum fermentation broth could be reused in at least three sequential fermentation. Xanthan gum fermentation was similar using fresh medium, primary permeate and secondary permeate in terms of final viscosity, final product concentration, and yield. A higher productivity was observed when the spent medium was reused in the fermentation. The molecular weight of xanthan gum produced in the fermentation using recycled spent media was similar to those from the fresh medium. In addition, xanthan gum produced from the recycled spent medium meets the standard criteria of United States Pharmacopeia/National Formulary (USP/NF).

The scalability of RFBB was studied by comparing the operational parameters and fermentation kinetics of gas- and liquid-continuous repeated batch fermentations. The results showed that the immobilized-cell fermentations in the RFBB gave higher xanthan gum yield and volumetric productivity compared to those from stirred tank fermentations.

The increased xanthan yield and productivity can be attributed to the high density cell growth through refreshing medium and retaining cells inside the bioreactor by cell immobilization. The improvement in xanthan yield and productivity due to cell immobilization was also observed within each repeated-batch by comparing the data 162 from the second batch with the data from the initial batch. The RFBB can be operated in repeated-batch and fed-batch modes with relatively similar performance.

In the economic analysis study, four processes were compared, including STR-D, STR-

UF-D, RFBB-D, and RFBB-UF-D. Processes using ultrafiltration for xanthan gum production is economically favorable, because ethanol and energy usages are reduced greatly. Therefore, ultrafiltration is a promising primary recovery unit operation between fermentation and alcohol precipitation. The number of fermentors can be reduced by half when using RFBB, because of the higher productivity and final xanthan gum concentration. The unit production cost was 35 % lower in the integrated RFBB-UF-D process as compared to the conventional STR-D process. In addition, the cell-free broth from a RFBB can eliminate the biomass in the xanthan gum, resulting in less cell debris in the broth, and thus a higher quality and purity of xanthan gum product. The study shows that RFBB-UF-D is the most economical process among the four processes studied for large-scale xanthan gum production.

Overall, this study shows that xanthan gum can be produced environmental-friendly and economically from whey lactose in an integrated RFBB-UF process.

7.2. Recommendations

In medium preparation, microfiltration or pasteurization could be introduced to sterilize hydrolyzed whey permeate and ultrafiltration permeate. In this case, the nutrient degradation due to high temperature could be avoided.

As for fermentation, the RFBB process can be further improved by optimizing the reactor design and operation. Long-term performance and stability of RFBB need to be tested

163 with more consecutive batches. Because cells on fibrous bed can be reused in repeated batches, feeding strategy supplying glucose without the nitrogen source, when the biomass was sufficient, could be investigated to further reduce the raw material cost of yeast extract. In addition, to study the inhibition of recycling spent medium and the maximum number of recycling batch, at least 10 sequential batches are required to verify the infinite batch recycling. If the bottleneck of the fermentation is not the limitation of metabolites with 75% replacement, the metabolites will eventually reach 300 % of those at the end of the first batch.

In terms of xanthan gum properties, further study is necessary to understand the lower broth viscosity from HWP, such as investigating the pyruvate content and molecular weight distribution. It is also necessary to analyze the rheological properties of the xanthan gum produced using RFBB.

164

Bibliography

Afendra, Amalia S.; Yiannaki, Efthalia E.; Palaiomylitou, Maria A.; Kyriakidis, Dimitrios A.; Drainas, Constantin. Co-production of ice nuclei and xanthan gum by transformed Xanthomonas campestris grown in sugar beet molasses. Biotechnology Letters (2002), 24(7), 579-583.

Amanullah, A.; Satti, S.; Nienow, A. W. Enhancing xanthan fermentations by different modes of glucose feeding. Biotechnology Progress (1998c), 14(2), 265-269.

Argin-Soysal, S. and Lo, Y.M., 2004, Characterization of intermolecular interactions of xanthan biopolymers based on conformational changes evidenced by rheo-optical approaches, 2004 IFT Annual Meeting, July 12-16, Las Vegas, NV. 99C-10

Ashraf S; Soudi M R; Sadeghizadeh M Isolation of a novel mutated strain of Xanthomonas campestris for xanthan production using whey as the sole substrate. Pakistan journal of biological sciences: PJBS (2008), 11(3), 438-42

Bandalusena, H. C. Hemaka; Zimmerman, William B.; Rees, Julia M. Microfluidic rheometry of a polymer solution by micron resolution particle image velocimetry: a model validation study From Measurement Science and Technology (2009), 20(11), 115404/1-115404/9

Babu, P.S.R.. Panda, T. 1991. Effect of recycling of fermentation broth from the production of penicillin amidase. Process Biochemistry, 26, 7-14.

Bogaert, J.C., Production and novel applications of natural L (+) lactic acid: Food pharmaceutics and biodegradable polymers. Cerevis (1997), 22, 46-50.

Brahmbhatt, Sudhir. Use of pure oxygen in viscous fermentation processes. U.S. Pat. Appl. Publ. (2007), 6pp

Cadmus MC, Knutson CA, Lagoda AA, Pittsley JE, Burton KA. Synthetic media for production of quality xanthan gum in 20 liter fermentors. Biotechnol Bioeng (1978), 20:1003-14

165

Candia, J.-L. Flores; Deckwer, W.-D. Effect of the Nitrogen Source on Pyruvate Content and Rheological Properties of Xanthan. Biotechnology Progress (1999), 15(3), 446-452

Casas, J.A., Mohedano, A.F., Garcia-Ochoa, F. Viscosity of guar gum and xanthan/guar gum mixture solutions. Journal of the Science of Food and Agriculture (2000), 80, 1722-1727.

Chang, Chun; Xu, Gui-zhuan; Li, Hong-liang; Chen, Jun-ying; Ma, Xiao-jian. Study on fermentation condition of Xanthomonas campestris QH79 for xanthan gum. Zhengzhou Gongye Daxue Xuebao (2000), 21(1), 92-94

Cheetham, Norman W. H.; Norma, N. M. Nik. The effect of pyruvate on viscosity properties of xanthan. Carbohydrate Polymers (1989), 10(1), 55-60.

Converti, A., Perego, P., Lodi, A., Fiorito, G., Borghi, M.D., Ferraiolo, G. 1991, In-situ ethanol recovery and subst4rate recycling during continuous alcohol fermentation. Bioprocess Engineering, 7, 3-10.

Cote, A.; Brown, W. A.; Cameron, D.; Van Walsum, G. P. Hydrolysis of lactose in whey permeate for subsequent fermentation to ethanol. Journal of Dairy Science (2004), 87(6), 1608-1620

Daniels, M.J., Barber, C.E., Turner, P.C., Cleary, W.G.and Sawczyc, M.K., Isolation of mutants of Xanthomonas campestris pv. campestris Showing Altered Pathogenicity Journal of General Microbiology (1984) 130, 2447-2455

Daniels, M. J.; Barber, C. E.; Turner, P. C.; Sawczyc, M. K.; Byrde, R. J. W.; Fielding, A. H. Cloning of genes involved in pathogenicity of Xanthomonas campestris pv. campestris using the broad host range cosmid pLAFR1. EMBO Journal (1984), 3(13), 3323-8

Daniels, S. L. Mechanisms involved in sorption of microorganisms to solid surfaces. In Adsorption of microorganisms to surfaces; Bitton, G., Marshall, K. C., Eds.; John Wiley: New York, 1980, pp 7-58.

De Vuyst L, Vermeire A, Van Loo J, Vandamme EJ. Nutritional, physiological and process Ð technological improvements of xanthan fermentation process. Mec Fac Landbouwwet Rijkuniv Gent (1987a), 52: 1881-900.

De Vuyst L, Van Loo J, Vandamme EJ. Two-step fermentation process for improved xanthan production by Xanthomonas campestris NRRL B-1459. J Chem Technol Biotechnol (1987b), 39:263-73.

166 Doma, B. T., A rotating fibrous bed bioreactor for xanthan gum fermentation. (1999), Doctoral dissertation at University of the Philipines, Diliman, Quezon City, Philipines.

Do, Jin Hwan; Chang, Ho Nam; Lee, Sang Yup. Efficient recovery of γ -poly (glutamic acid) from highly viscous culture broth. Biotechnology and Bioengineering (2001), 76(3), 219-223.

Faria, Sandra; Vieira, Patricia Angelica; Resende, Miriam Maria; Ribeiro, Eloizio Julio; Cardoso, Vicelma Luiz. Application of a model using the phenomenological approach for prediction of growth and xanthan gum production with sugar cane broth in a batch process. LWT--Food Science and Technology (2010), 43(3), 498- 506.

Funahashi, H,m Machara, M,m Taguchi, H., Yoshida, T. effect of glucose concentration on xanthan gum production by Xanthomonas campestris. J. Chem. Eng. Jpn. (1987a), 65, 603-606

Garcia-Ochoa, F.; Santos, V. E.; Alcon, A. Xanthan gum production: an unstructured kinetic model. Enzyme Microbiol Technol, 1995, 17, 206-17.

Garcia-Ochoa, F.; Santos, V. E.; Alcon, A. Metabolic structured kinetic model for xanthan production. Enzyme and Microbial Technology (1998), 23(1/2), 75-82.

Garcia-Ochoa, F., Gomez Castro, E., Santos V.E., Oxygen transfer and uptake rates during xanthan gum production. Enzyme and Microbial Technology (2000), 27, 680-690.

Garcia-Ochoa, Felix; Santos, Victoria E.; Alcon, Almudena. Chemical structured kinetic model for xanthan production. Enzyme and Microbial Technology (2004), 35(4), 284-292.

Garcia-Ochoa, F., Gomez Castro, E., Santos V.E., Oxygen transfer and uptake rates during xanthan gum production. Enzyme and Microbial Technology (2000), 27, 680-690. Gonzalez Siso, M. I. The biotechnological utilization of cheese whey a review. Bioresource Technology (1996), 57(1), 1-11

Godbole SP, Shumpe A, Shah YT, Carr NL. Hydrodynamics and mass transfer in non- Newtonian solutions in a bubble column. AIChE J 1984;30:213–20.

Hamilton, Michelle A. Hamilton; Dawkins, Garth S.; Mellowes, Winston A. Xanthan gum production from sugarcane fluids. U.S. Pat. Appl. Publ. (2009), 5 pp

167 Hsiao, T.Y., Glatz, C.E., Glatz, B.A. 1994, Broth recycle in a yeast fermentation. Biotechnol. Bioeng. 44, 1228-1234.

Hsiao, T-Y. and Glatz, C.E., 1996, Water Reuse in the L-Lysine Fermentation Process, Biotechnology and Bioengineering, 49, 341-347.

Hsu, C. H.; Chu, Y. F.; Argin-Soysal, S.;Hahm, T. S.; Lo, Y. M. Effects of surface characteristics and xanthan polymers on the immobilization of Xanthomonas campestris to fibrous matrices. Journal of Food Science (2004), 69(9), E441-E448

Hsu, Chia-Hua; Lo, Y. Martin. Characterization of xanthan gum biosynthesis in a centrifugal, packed-bed reactor using metabolic flux analysis. Process Biochemistry (2003), 38, 1617-1625.

Janaun, J.; Wong, K. W.; Chu, C. M.; Prabhakar, A. Evaluation of xanthan production from Xanthomonas campestris using sago starch as carbon source. World Congress of Chemical Engineering, 7th, Glasgow, United Kingdom, July 10-14, 2005, (2005), 83515/1-83515/9

Ji, Lina; Mo, Xiaoyan; Zhan, Guyu. Effect of reduced pressure distillation on production rate and relative molecular weight of xanthan gum. Xi'an Jiaotong Daxue Xuebao (2001), 35(5), 549-550.

Kalogiannis, Stavros; Iakovidou, Gesthimani; Liakopoulou-Kyriakides, Maria; Kyriakidis, Dimitrios A.; Skaracis, George N. Optimization of xanthan gum production by Xanthomonas campestris grown in molasses. Process Biochemistry (Oxford, United Kingdom) (2003), 39(2), 249-256.

Kawase and Hashimoto, 1996 Y. Kawase and N. Hashimoto, Gas hold-up and oxygen transfer in three-phase external loop airlift bioreactors: non-Newtonian fermentation broths, Journal of Chemical Technology and Biotechnology 65 (1996), pp. 325–334

Kennedy JF, Jones P, Barker SA, Banks GT. Factors affecting microbial growth and polysaccharide production during the fermentation of Xanthomonas campestris cultures. Enzyme Microbiol Technol (1982), 4:39-43.

Kennedy J.F. and Bradshaw, I.J. Production properties and applications of xanthan. Prog. Ind. Microbiol (1984), 19, 319-71

Kongruang, Sasithom. Quantitative analysis of exopolysaccharide production in A stirred tank bioreactor. AIChE Annual Meeting, Conference Proceedings, Cincinnati, OH, United States, Oct. 30-Nov. 4, 2005 (2005), 438l/1-438l/15.

168 Konsoula, Zoe; Liakopoulou-Kyriakides, Maria; Perysinakis, Angelos; Chira, Panayiota; Afendra, Amalia; Drainas, Constantin; Kyriakidis, Dimitrios A. Heterologous expression of a hyperthermophilic α-amylase in xanthan gum producing Xanthomonas campestris cells. Applied Biochemistry and Biotechnology (2008), 149(2), 99-108.

Lee, Byoung-Moo; Park, Young-Jin; Park, Dong-Suk; Kang, Hee-Wan; Kim, Jeong-Gu; Song, Eun-Sung; Park, In-Cheol; Yoon, Ung-Han; Hahn, Jang-Ho; Koo, Bon- Sung; Lee, Gil-Bok; Kim, Hyungtae; Park, Hyun-Seok; Yoon, Kyong-Oh; Kim, Jeong-Hyun; Jung, Chol-hee; Koh, Nae-Hyung; Seo, Jeong-Sun; Go, Seung-Joo. The genome sequence of Xanthomonas oryzae pathovar oryzae KACC10331, the bacterial blight pathogen of rice. Nucleic Acids Research (2005), 33(2), 577-586

Lamboley, L., Lacroix, C., Chmpagne, C.P., Vuillemard, J.C., Continuous mixed strain mesophilic starter production in supplemented whey permeate medium using immobilized cell technology. (1997), Biotechnol. Bioeng. 56, 502-516

Liakopoulou-Kyriakides, M.; Psomas, S. K.; Kyriakidis, D. A. Xanthan gum production by Xanthomonas campestris w.t. fermentation from chestnut extract. Applied Biochemistry and Biotechnology (1999), 82(3), 175-183.

Lo, Y.M., A Novel immobilized cell bioreactor and membrane ultrafiltration for xanthan gum production. (1996), Doctoral dissertation at the Ohio State University

Lo, Y. M.; Hsu, C. H.; Yang, S. T.; Min, D. B. Oxygen transfer characteristics of a centrifugal, packed-bed reactor during viscous xanthan fermentation. Bioprocess and Biosystems Engineering (2001), 24(3), 187-193

Lopez, M. J.; Moreno, J.; Ramos-Cormenzana, A. The effect of olive mill wastewaters variability on xanthan production. Journal of Applied Microbiology (2001), 90(5), 829-835

Maldonado, A., Effect of medium composition on xanthan gum production from whey permeate using Xanthomonas campestris NRRL B-1459 (1992), Master thesis at the Ohio State University

Margaritis, A. and G. W.Pace. Microbial polysaccharides. In Comprehensive biothenology vol. 3/ Moo-Young, M. (Ed.), (1985), Pergamon Press, new Yourk, NY; pp. 1005-1043

Medvedeva, E.I., Oleinik, T.P., Panchenko, K.A., Petrenko, E. B., 1989. Multipurpose use of liquid wastes in microbial lysine synthesis technology. Biotekhnologiya, 6, 761-767

169 Mesomo, Michele; Silva, Marceli Fernandes; Boni, Gabriela; Padilha, Francine Ferreira; Mazutti, Marcio; Mossi, Altemir; de Oliveira, Debora; Cansian, Rogerio Luis; Di Luccio, Marco; Treichel, Helen. Xanthan gum produced by Xanthomonas campestris from cheese whey: production optimisation and rheological characterization. Journal of the Science of Food and Agriculture (2009), 89(14), 2440-2445

Milas, M., Rinaudo, M. and Tinland, B. The viscosity dependence on concentration, molecular weight and shear rate of xanthan solutions, Polymer Bulletin, (1985), 14,157-164

Moraine RA, Rogovin P. Kinetics of polysaccharide B-1459 fermentation. Biotechnol Bioeng (1966), 8:511-24

Moraine RA, Rogovin P. Xanthan biopolymer production at increased concentration by pH control. Biotechnol Bioeng (1971),13:381-91.

Moraine RA, Rogovin P. Kinetics of the xanthan fermentation. Biotechnol Bioeng (1973), 15:225-37.

Moreno, J.; Lopez, M. J.; Vargas-Garcia, C.; Vazquez, R. Use of agricultural wastes for xanthan production by Xanthomonas campestris. Journal of Industrial Microbiology & Biotechnology (1998), 21(4/5), 242-246.

Najafpour, G.D., Bioprocess scale-up. Biochemical Engineering and Biotechnology, Chapter 13 Elsevier B.V. 2007

Nery, Tatiana Barreto Rocha; Brandao, Lillian Vasconcellos; Esperidiao, Maria Cecilia Azevedo; Druzian, Janice Izabel. Biosynthesis of xanthan gum from the fermentation of milk whey: productivity and viscosity. Quimica Nova (2008), 31(8), 1937-1941

Nicolas, G., Auger, I., Beaudoin, M., Halle, F., Morency, H,m LaPointe, G., Lavoie, M.C., Improved methods for mutacin detection and production. J. Microbiol. Methods, (2004), 59, 351-361.

Nykanen, A., Lapvetelainen, A., Hietnen, R.M., Kallio, H., The effect of lactic acid, nisin, whey permeate, sodium chloride and related combinations on aerobic plate count and the sensory characteristics of rainbow trout. Lebensm. Wiss. Technol., (1998), 31, 286-290.

Ochiai, Hirokazu; Inoue, Yasuhiro; Takeya, Masaru; Sasaki, Aeni; Kaku, Hisatoshi. Genome sequence of Xanthomonas oryzae pv. oryzae suggests contribution of large numbers of effector genes and insertion sequences to its race diversity. JARQ (2005), 39(4), 275-287 170

Papoutsopoulou, S. V.; Ekateriniadou, L. V.; Kyriakidis, D. A. Genetic construction of Xanthomonas campestris and xanthan gum production from whey. Biotechnology Letters (1994), 16(12), 1235-40

Peters H-U, Herbst H, Hesselink PGM, LuÈndsdorf H, Schumpe A, Deckwer W-D. The influence of agitation rate on xanthan production by Xanthomonas campestris. Biotechnol Bioeng (1989), 34:1393-7.

Pinches A, Pallent LJ. Rate and yield relationships in the production of xanthan gum by batch fermentations using complex and chemically defined growth media. Biotechnol Bioeng (1986), 28:1484-96.

Pons A, Dussap CG, Gros JB. Xanthan batch fermentations: compared performance of a bubble column and a stirred tank fermentor. Bioprocess Eng (1990), 5:107-14.

Psomas, S.K., Liakopoulou-Kyriakides, M., and Kyriakidis, D.A., Optimization study of xanthan gum production using response surface methodology. Biochemical Engineering Journal (2007)35, 273-280.

Qian, Wei; Jia, Yantao; Ren, Shuang-Xi; He, Yong-Qiang; Feng, Jia-Xun; Lu, Ling-Feng; Sun, Qihong; Ying, Ge; Tang, Dong-Jie; Tang, Hua; Wu, Wei; Hao, Pei; Wang, Lifeng; Jiang, Bo-Le; Zeng, Shenyan; Gu, Wen-Yi; Lu, Gang; Rong, Li; Tian, Yingchuan; Yao, Zhijian; Fu, Gang; Chen, Baoshan; Fang, Rongxiang; Qiang, Boqin; Chen, Zhu; Zhao, Guo-Ping; Tang, Ji-Liang; He, Chaozu. Comparative and functional genomic analyses of the pathogenicity of phytopathogen Xanthomonas campestris pv. campestris. Genome Research (2005), 15(6), 757- 767

Ramshaw, C.; Mallinson, R. H. Mass Transfer Process. U.S. Patent 4,283,255, 1981.

Rao, D. P.; Bhowal, A.; Goswami, P. S. Process Intensification in Rotating Packed Beds (HIGEE): An Appraisal Ind. Eng. Chem. Res. 2004 43 1150

Robinson, D. K. and Wang, D.I. C. A transport controlled bioreactor for the simultaneous production and concentration of xanthan gum. Biotechnol. Prog. (1988), 4, 231- 241

Rocha-Valadez, J. A.; Albiter, V.; Caro, M. A.; Serrano-Carreon, L.; Galindo, E. A fermentation system designed to independently evaluate mixing and/or oxygen tension effects in microbial processes: development, application and performance. Bioprocess and Biosystems Engineering, (2007), 30(2), 115-122

171 Rochefort, W. E. Middleman, S. (1987). Rheology of xanthan gum: salt, temperature, and strain effects in oscillatory and steady shear experiments. Journal of Rheology, 31(4), 337-369.

Rogovin P, Albrecht W, Sohns V. Production of industrial grade polysaccharide B-1459. Biotechnol Bioeng (1965), 7:161-9.

Rosalam, S. and England, R., Review of xanthan gum production from unmodified starches by Xanthomonas campestris sp. Enzyme and Microbial Technology, (2006), 39(2), 197-207

Rosalam, S., Krishnaiah, D. and Bono, A. Cell free xanthan gum production using continuous recycled packed fibrous-bed bioreactor-membrane. Malaysian Journal of Microbiology, (2008), 4(1), pp 1-5

Roukas, T. and F. Mantzouridou, 2001. Effect of the aeration rate on pullulan production and fermentation broth rheological properties in an airlift reactor. J. Chem. Technol. Biotechnol., 76: 371-376.

Schepers, A.W., Thibault, J., Lacroix, C., Continuous lactic acid production in whey permeate/yeast extract medium with immobilized Lactobacillus helveticus in a two-stage process: model and experiments. Enzyme Microb. Technol., (2006), 38, 324-337.

Schwartz, Robert D.; Bodie, Elizabeth A. Production of high-viscosity whey broths by a lactose - utilizing Xanthomonas campestris strain. Applied and Environmental Microbiology (1985), 50(6), 1483-5

Serrania, Javier; Vorhoelter, Frank-Joerg; Niehaus, Karsten; Puehler, Alfred; Becker, Anke. Identification of Xanthomonas campestris pv. campestris galactose utilization genes from transcriptome data. Journal of Biotechnology, (2008), 135(3), 309-317

Shu Ch-H, Yang Sh-T. Effects of temperature on cell growth and xanthan production in batch cultures of Xanthomonas campestris. Biotechnol Bioeng (1990), 35:454-68.

Shuler, M. L., Kargi, F. Bioprocess Engineering basic concepts, second edition, (1992), pp 286

Silva, M.F., Fornari, R.C.G., Mazutti, M.A., Oliveira, D., Padilha, F.F., Cichoski, A.J., Cansian, R.L., Luccio, M.D., Treichel, H.. Production and characterization of xanthan gum by Xanthomonas campestris using cheese whey as sole carbon source. Journal of Food Engineering (2009), 90, 119-123.

172 Silveira, W.B., Passos, F.J.V., Mantovani, H.C., Passos, F.M.L., Ethanol production from cheese whey permeate by Kluyveromyces Marxianus UFV-3: a flux analysis of oxido-reductive metabolism as a function of lactose concentration and oxygen levels. Enzyme Microb. Technol. (2005), 36, 930-936.

Smith, Ian H.; Pace, Gary W. Recovery of microbial polysaccharides. Journal of Chemical Technology and Biotechnology (1979-1982) (1982), 32(1), 119-29. Herbst, Holger; Schumpe, Adrian; Deckwer, Wolf Dieter. Xanthan production in stirred-tank fermentors: oxygen transfer and scale-up. Chemical Engineering & Technology (1992), 15(6), 425-34

Song, K-W., Kim, Y-S., and Chang, G-S., Rheology of concentrated xanthan gum solutions: steady shear flow behavior, Fibers and Polymers, (2006), 7(2), 129-138.

Souw P, Demain AL. Role of citrate in xanthan production by Xanthomonas campestris. J Ferment Technol (1980), 58:411-6.

Stanbury PF, Whitaker A, Hall SJ. 1995. Principles of fermentation technology, 2nd ed. New York: Pergamon. p 243–276.

Stokke, Bjoern T.; Elgsaeter, Arnljot; Smidsroed, Olav. Electron microscopic study of single - and double -stranded xanthan. International Journal of Biological Macromolecules (1986), 8(4), 217-25

Taron, C.H., Benner, J.S., Hornstra, L.J., Guthrie, E.P., A novel β-galactosidase genes isolated from the bacterium Xanthomonas campestris exhibits strong homology to several eukaryotic β-galactosidase. Glycobiol., (1995), 5, 603-610

Thacker, A., Fu, S., Boni, R.L., and Block, L.H., Inter- and intra-manufacturer variability in pharmaceutical grades and lots of xanthan gum. AAPS PharmSciTech, (2010) 11, 4, pp1619-1626.

Thieme, Frank; Koebnik, Ralf; Bekel, Thomas; Berger, Carolin; Boch, Jens; Buettner, Daniela; Caldana, Camila; Gaigalat, Lars; Goesmann, Alexander; Kay, Sabine; Kirchner, Oliver; Lanz, Christa; Linke, Burkhard; McHardy, Alice C.; Meyer, Folker; Mittenhuber, Gerhard; Nies, Dietrich H.; Niesbach-Kloesgen, Ulla; Patschkowski, Thomas; Rueckert, Christian; Rupp, Oliver; Schneiker, Susanne; Schuster, Stephan C.; Vorhoelter, Frank-Joerg; Weber, Ernst; Puehler, Alfred; Bonas, Ulla; Bartels, Daniela; Kaiser, Olaf. Insights into genome plasticity and pathogenicity of the plant pathogenic bacterium Xanthomonas campestris pv. vesicatoria revealed by the complete genome sequence. Journal of Bacteriology (2005), 187(21), 7254-7266

173 Thorne, Linda; Tansey, Lisa; Pollock, Thomas J. Direct utilization of lactose in clarified cheese whey for xanthan gum synthesis by Xanthomonas campestris. Journal of Industrial Microbiology (1988), 3(5), 321-8

Tinland, B., Maret, G., and Rinaudo, M., Reptation of semidilute solutions of wormlike polymers. Macromolecules (1990), 23, 596-602.

Tiwari, B.K., Muthukumarappan, K., O’Donnell, C.P., Cullen, P.J., Rheological properties of sonicated guar, xanthan and pectin dispersions. International Journal of Food Properties (2010) 13, 223-233.

Trujillo-Roldán, M. A., Peña, C., Ramirez, O.T., Galindo, E., Effect of Oscillating Dissolved Oxygen Tension on the Production of Alginate by Azotobactervinelandii, Biotechnology Progress, (2001), 17, 6, 1042-1048

Tyagi, R.D., Kluepfe, D., Bioconversion of cheese whey to organic acids. In: Martin, A.M. (Ed.), Bioconversion of Waste Materials to Industrial Products. Blackie Academic, Glasgow, UK. (1998)

Van’t Riet, K. and Tramper, J., 1991, Basic Bioreactor Design (New York: Marcel Dekker Inc.).

Vojnov, Adrian A.; Slater, Holly; Daniels, Michael J.; Dow, J. Maxwell. Expression of the gum operon directing xanthan biosynthesis in Xanthomonas campestris and its regulation in planta. Molecular Plant-Microbe Interactions (2001), 14(6), 768-774

Vorhölter, Frank-Joerg; Schneiker, Susanne; Goesmann, Alexander; Krause, Lutz; Bekel, Thomas; Kaiser, Olaf; Linke, Burkhard; Patschkowski, Thomas; Rueckert, Christian; Schmid, Joachim; Sidhu, Vishaldeep Kaur; Sieber, Volker; Tauch, Andreas; Watt, Steven Alexander; Weisshaar, Bernd; Becker, Anke; Niehaus, Karsten; Puehler, Alfred. The genome of Xanthomonas campestris pv. campestris B100 and its use for the reconstruction of metabolic pathways involved in xanthan biosynthesis. Journal of Biotechnology (2008), 134(1-2), 33-45.

Walsh, Patricia M.; Haas, Michael J.; Somkuti, George A. Genetic construction of lactose - utilizing Xanthomonas campestris. Applied and Environmental Microbiology (1984), 47(2), 253-7.

Weiss, Rebecca M.; Ollis, David F. Extracellular microbial polysaccharides. I. Substrate, biomass, and product kinetic equations for batch xanthan gum fermentation. Biotechnology and Bioengineering (1980), 22(4), 859-73.

Wyatt, Nicholas B. and Liberatore, Matthew W. Rheology and viscosity scaling of the polyelectrolyte xanthan gum. Journal of Applied Polymer Science. (2009), 114, 4076-4084. 174

Yang, Shang-Tian; Lo, Yang-Ming; Chattopadhyay, Devamita. Production of cell-free xanthan fermentation broth by cell adsorption on fibers. Biotechnology Progress (1998), 14(2), 259-264.

Yang, Shang-Tian; Lo, Yang-Ming; Min, David B. Xanthan gum fermentation by Xanthomonas campestris immobilized in a novel centrifugal fibrous- bed bioreactor. Biotechnology Progress (1996), 12(5), 630-637

Yang, T.C., Wu, G.H., and Tseng, Y.H., Isolation of a Xanthomonas campestris strain with elevated β-galactosidase activity for direct use of lactose in xanthan gum production. Letters in Applied Microbiology. (2002), 35, 375-379.

Yang, T.C., Hu, R.M., Hsiao, Y.M., Weng, S.F., Tseng, Y.H., Molecular genetic analyses of potential β-galactosidase genes in Xanthomonas campestris. J. Mol. Microbiol. Biotechnol. (2003), 6, 145-154.

Yano, T., Mori, H., and Kobayashi, T., 1980, Reusability of broth supernatant as medium. J. Ferment. Technol., 58, 3, 259-266

Zhang, Zhiguo; Chen, Hongzhang. Fermentation performance and structure characteristics of xanthan produced by Xanthomonas campestris with a glucose/xylose mixture. Applied Biochemistry and Biotechnology (2010), 160(6), 1653-1663.

Zhao S, Kuttuva SG, Ju L-K (1999) Oxygen transfer characteristics of multiple-phase dispersions simulating water-in-oil xanthan fermentations. Bioprocess Eng 20:313–323

Zhu, Shengdong; Wu, Yuanxin; Yu, Ziniu; Tong, Haibao; Cheng, Dachang; Xie, Decheng. Improving xanthan fermentation in a mechanically stirred aerated fermenter via external loop. Research Journal of Microbiology (2006), 1(1), 70-75.

Zirnsak, M.A., Boger, D.V., Tirtaatmadja, V. Steady shear and dynamic rheological properties of xanthan gum solutions in viscous solvents. J. Rheol., (1999), 43(3), 627-650.

175

Appendix

176 Glucose and galactose concentration in lactose hydrolysate and hydrolyzed whey permeate

Lactose hydrolysate HWP

GOS 29.57 N/A

lactose 60.79 28.025

glucose 177.32 144.83

galactose 130.45 144.60

N/A: not available

Calibration curve of X. campestris concentration vs. optical density at 650 nm.

0.5 y = 0.5464x - 0.0047 R² = 0.9998 0.4

0.3

0.2 weight (g/L) weight

0.1

0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 OD 650

177