université d'ottawa university of · pdf fileconcentration de la n-padhe, ......
TRANSCRIPT
Université d'Ottawa University of Ottawa
FEASIBILITY OF THE BIO-CONVERSION OF EIYDROCARBON FEEDSTOCK TO CITRIC ACID
Anna Maria Crolla B.A.%. Chemical Engineering
B.Sc. Biochemistry
A thesis submitted to the School of Graduate Studies in partial fulfillment of the requirements for the degree of
Master of Applied Science Chernical Enginetring
in the
Department of Chemical Engincering University of Ottawa
Ottawa, Canada
National Library of Canada
Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bibliographie Services services bibliographiques
395 Wellington Street 395, nie Wellington OitawaON K1AON4 OnawaON K l A O N 4 Canada Canada
Your hk Votre milirenœ
Our fi& Noire rekrenœ
The author has granted a non- exclusive licence ailowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of this thesis in microform, paper or electronic formats.
The author retains ownership of the copyright in this thesis. Neither the
L'auteur a accordé une Licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la fome de microfiche/film, de reproduction sur papier ou sur format électronique.
L'auteur conserve la propriété du droit d'auteur qui protège cette thèse.
thesis nor substantial exûacts fiom it Ni la thèse ni des extraits substantiels may be printed or othewise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
ABSTRACT r
Currently, the rnajority of worldwide microbial production of citric acid utilizes
Aspegillics niger in a carbahydrate based submerged fmentation. An innovative
alternative to utilin'ng carbohydrates an hydrocarbon feedsfocks. Due to their hi@
carbon content, hydrocarôons theoretically have the potential of producing high yields of
citric acid. With the increasing prices in the sugars market, it is of economical interest to
evaluate the feasibility of u t i b g hydrocarbom in citric acid production. It is thenfore
the focus of this shidy to examine the feasibility of using hydrocarbon oils as the sole
source of carbon for the production of citnc acid and biomass.
Iuitial lab experimmts were conducted using 1875 ml batch flask fe~llentations to
determine the yeast strain of Candida 1ipoIyticca and hydrocarbon feedstock of choice.
The two yeast strains investigated were Candida Iipolyticco NRRL-Y-IO94 and Candida
lipolyn'ca NRRL-Y-1095. The two feeâstocks evaluated were kaosene (Km 1-K) and n-
paraffin solvent (Norpar IS), both obtained from Impexial Oil Ltd. From these tests it
was found that Candido lipolyticcn M W - Y 4 0 9 5 assirnilated both fccdsfocks more
effeaively than Candida lipolytica NRRL-Y-1094. A citrk acid productivity of 47 mg/L-
h was obtained when Candi& 1ipoStica NRRL-Y-IO95 was uscd with the n - p a d k
solvent, while a citric acid productivity of 23 m@-h was obtained when using kmsene.
Therefore dl subsequent studies were conducted using Candida lipI'*cu NRRL-Y-IO95
and n-paraffin solvent.
To determine the optimum level of initial biomaw concentration, n-parafh
concentration, iron concentration and temperature for the production of citric acid, a
central composite design was developed using 200 ml batch flask fcnnentations. The
design involved conducting 31 batch flask fermentations under various combinations of
high and low values of these four parameters. From this investigation it was found that
the optimum levels of each parameter for citric acid production wen, 10-12 % vohmie for
initiai biomass inoculum, 10-1 5% volume for n-par- concentration, 10 mg& for f e c
nitrate concentration, and 26-30°C for temperature.
These optimum levels of the above operating parameters were implemented in larger
scde (7litre) fermentations. These famatations were conducted in a 14L Chemap
fmentor. The fermentations investigateâ the effécts of aeration, agitation (mechanical
agitation h m the fmnentor and an extemaî agitator), and batch vasus fed-batch systems
on biomass and citric acid production. It was determined that an &on increase h m
0.5 to 2.0 wm in batch systems did not a f b t overaîî citric acid yield and productivity.
An agitation rate increase h m 400 to 800 rpm in batch systcms increased average citric
acid productiviîy by 39%. In fed-batch systems, an agitation increase h m 400 to lûûû
rpm resulted in a citnc acid productivity increase of 125 %. The utilkation of an extemai
agitator running at 1400 rpm (while the fermentor agitator is ruMing at 400 rpm)
increased ciûic acid productivity by 75% h m when an extemal agitator is absent h m
the system. However, an extemai agitation of 1600 rpm resulted in a 20-30 % increase in
ce11 lyses. Citric acid yields wae increased by 100% when rnoving h m batch to fed-
batch systems. When a thm cycle fed-batch system was implcmented, citric acid yields
were increased by 100% h m fed-batch systems and 200% h m batch systems.
Actuellement, la majorité de la production microbienne mondiale d'acide citrique utilise
le Aspergillus nzger dans une fermentation submergée a base de glucide. L'utilisation de
charges d'alimentation à base d'hydrocarbure plutôt que de glucides est une alternative
innovatrice. Vu leur tmew élevée en carbone, les hydrocarbures ont, en théorie, le
potentiel de produire de hauts rendements d'acide citrique. Face à l'augmentation du prix
des sucres sur le marché, on peut avoir intérêt, sur le plan économique, à évaluer la
faisabilité d'utiliser des hydrocarbures dans la production d'acide citrique. La présente
étude a donc pour objet d'examiner la faisabilité d'utiliser des huiles d'hydrocarbures
comme source exclusive de carbone pour la production d'acide citrique et la biomasse.
Les premières expériences en laboratoire ont été effectuées au moyen de fmentations
par lots en ballon de 1875 mL pour détenniner la souche de l e m de Candido l@o&tica
et la charge dliyârocarbure optimale. Les deux souches de levure étudiées ont été la
Candida lipolytica NRRL-Y-1094 et la Candida lipoiytica NRRL-Y-1095. Les deux
charges d'alimentation évaluées ont été le kérosène (Km 1-K) et le solvant n-
parafane(Nofpar 15), tous deux obtenus de la Compagnie pétrolière impériale Ltée. Ces
tests ont révélé que la Candida Iipoytica NRRL-Y-1095 assimilait les deux charges plus
efficacement que la Candida lipoi'ica NRRL-Y-1094. On a obtenu um productivité
d'acide citrique de 47 mgA-h en utilisant la Ca~dida 1ipoIytica NRRL-Y-1095 avec la n-
paraffine, comparativement à une productivité d'acide citrique de 23 mg/i-H avec le
kérosène. Toutes les études subséquentes ont par conséquent été effectuées avec la
Candida li@o&ica NRRL-Y-1095 et la n-paraffine.
Pour déterminer le niveau optimum de concentration initiale de la biomasse, de
concentration de la n-padhe, de concentration du fer et de la tcmpératurc pour la
production d'acide citrique, on a mis au point un modèle composite central a l'aide de
fmentaîions par lots m balion de 200 mL. Nous avons ainsi effcctui 31 fermentations
en utilisant diverses combinaisons de valeurs f~bles et élevées de ces quatre paramitreS.
A partir de cette étude, nous avons constaté que les niveaux optimums de chaqae
paramètre pour la production d'acide citrique étaient de 10 à 12 % pour le volume de
concentration initiale de la biomasse, de 10 à 15 % pour le volume de concentration de la
n-paraffine, de 10 mgA pour la concentration de nitrate de fa , et de 26 à 30°C pour la
température.
Ces niveaux optimums pour chacun des paramètres cidessus ont ensuite été utilisés dans
des fermentations à plus grande échelle (7 litres) effectuées dans un fermenteur Chemap
de 14L. On a alors étudié les effets de I'aération, de l'agitation (agitation mécanique du
fermenteur et d'un agitateur exteme) et des systèmes par lots (batch systems) par rapport
aux systèmes a écoulement discontinu (fod-batch) sur la biomasse et la production d'acide
citrique. Il a été déterminé qu'une augmentation de l'aération de 0,5 à 2,O wm dans les
systèmes par lots ne modifiait pas la production globale d'acide citrique ni la productivité.
Une augmentation du taux d'agitation de 400 à 800 tours/minute dans les systèmes par
lots a augmenté la productivité moyeane d'acide citrique de 39 %. Dans les systèmes à
écoulement discontinu, l'augmentation de l'agitation de 400 à 1000 toUfS/minute s'est
soldee par une augmentation de la productivité d'acide citrique de 125 %. L'utilisation
â'un agitateur externe fonctionnant a 11400 tourdminute (pendant que l'agitateur du
fermenteur fonctionnait à 400 todmiaute) a augmenté la productivité d'acide citrique de
75 % par rapport à la productivité du système sans agitateur externe. Mais, une agitation
externe de 1 6 0 tours-minute a produit une augmentation de 20 a 30 % de lyse cellulaire.
La production d'acide citrique s'est accrue de 100 % lorsque l'on est passé du système par
lots au système à écoulement discontinu. Lorsque l'on a mis en oeuvre un système 1
écoulement discontinu en trois cycles, la production d'acide citrique a augmenté de 100 %
dans le système à &oulement discontinu, a de 200 % dans le système par lots.
AKNOWLEDGEMENTS
1 would Wre to thank my supetvisor, Dr. Kevin Kennedy, for his unlimited guidance and
support, and extensive knowledge in the world of biotechnology. I wouid iike to express
my sincm appmiation to Mr. David Kmsky, of Touchstone Contracting, Management
& Engineering Services Ltd., whose e t d encouragement and training is fomer
appreciated. 1 would also like to thank Mr. Francisco Aposaga for his technical
assistance and support.
I would aiso Iüce to take this opportunity to acknowledge the hancial support of
Touchstone Contracting, Management & Engineering Services Ltd. and the Natural
Sciences and Engineering Research Council.
Lady, but certainly not least, 1 would like to thadc my family for their patience and
relentless support without which this would not have been possible. I would especially
like to thank my mother and fatha for their continuous guidance and encouragement, and
al1 the ways they have taught me.
To my parents, Iolanda and Gaetano CroUa, the type of people I aspire of becoming.
TABLE OF CONTENTS
Abstract
List of Tables
List of Figures
List of Abbreviations
Nomenclature
Chapter 1 : Introduction
1.1 Citric Acid Production Worldwide 1.2 Objectives and Scope of Research 1.3 Layout of Thesis
Chapter 2: Prduction of Citric Acid: A Fermentation Approach
2.1 Citric Acid Background 2.1.1 Physical Roperties 2.1.2 Chemical Properties 2.1.3 Occurrence of Citric Acid
Page
ii
iv
xiii
2.2 Citric Acid Production Processes-Fermentation 10 2.2.1 Rocesses using Aspegilus nzger
2.2.1.1 SuTfice Fermentation 2.2.1.2 Submerged Fermentation
2.2.1 Roccsses ushg Yeast 2.2.2.1 Submerged Fermentation using Carbohydratcs 2.2.2.2 Submerged Fermentation using n-Alkanes
2.3 Citric Acid Produciiig Yeast: Condida Iipolyfjcu 2.3.1 Candida lipoi)tica-The organism 2.3.2 Using Candi& lip&tica in the Production of
Citcic Acid 2.4 Composition of Medium
2.4.1 Carbon Source 2.4.2 Nitmgen Source 2.4.3 Minerai Source 2.4.4 pH and Temperature
2.5 Acration and Agitation
Chapter 3: Selection of Hydrocarbon Feedstock and Yeast Cwltiire
3.1 Introduction 3.2 Material and Methods
3.2.1 Yeast Culture 3.2.2 Hydrocarbon Feedstock 3.2.3 Yeast Ceil Growth and Fexmentation Mtdia 3.2.4 Experimcntal Protocols
3.2.4.1 Reparation of Yeast Seed Culture 3 U . 2 Reparation of Rc-Fermentation Stage
-8iomass Growîh 3.2.4.3 Reparation of Fermentation Stage
-Citric Acid Production 3.2.5 Analyîical Methods
3.3 Resuits and Discussion 3.3.1 Effect of Hydrocarbon Feedstock on
Citric Acid Production 3.3.1.1 4 Litre Fermentation with Kmsene Feeàstock 3.3.1.2 4 Litre Fermentation with n - P a r e Feedstock
3.3.2 Effect of Yeast Culture on Citric Acid Production 3.3.2.1 Fermentation with Cmdidu lipol'ca-Y- 1 094 3.3.2.2 Fermentation with Candida lijdytica-Y-1095
3.4 Conciusions
Page 27
.- 32
32 33 33 33 33 33 33
34
34 35 35
35 35 37 40 40 40 41
Chapter 4: Optimilntion of Biomass Inocul~m, n -Padfh Concentration, Iron Concentration and Temperatun for the Production of CiMc Acid ushg Cundfda üjwlyaèa Y-1095 42
4.1 Intraduction 4.2 Material and Methods
4.2.1 Yeast Culture 4.2.2 FttdStock 4.2.3 Yeast Celi Growth and Fermentation Media 4.2.4 Expaimcntal Protocols
4.2.4.1 Repmtion of Yeast Sesd Culture 4.2.4.2 Reparation of Re-Fermentation Stage
- InOCUlum Biomass Growth 4.2.4.3 RepaTafion of Fermentation Stage
- Citric Acid Production
4.2.5 Analytical Metho& 4.3 Results and Discussion 4.4 Conclusions
Page 48 48 57
Chapter 5: Citric Acid Prodiidion - Investigation of Aeration and Agftation E f f i in Batch and Fed-Batch Systems 58
5.1 Introduction 5.2 Material and Methods
5.2.1 Yeast Culture 5.2.2 Feedstock 5.2.3 Yeast Ce11 Growth and Fermentation Media
5.2.3.1 Batch Systerns 5.2.3.2 Fed-Batch Systems
5.2.4 Experimental Protocols 5.2.4.1 Prepaxation of Yeast Seed Culture for
Batch and Fed-Batch Systems 5.2.4.2 Reparation of Re-Fermentation Stage for
Batch and Fed-Batch Systems - Biomass Gmwth 5.2.4.3 Reparation of Fermentation Stage for
Batch Systems - Citric Acid Roduction 5.2.4.4 Reparation of Fexmentation Stage for
Fed-Batch Systems - Citric Acid Roductiou 5.2.5 Analytical Methods for Batch and Fed-Batch S ystems
5.3 Results and Discussion 5.3.1 Effect of Aeration on Biomass and Ciûic Acid
Production - Batch Systems 5.3.2 Efféct of Agitation on Biomass and Cihic Acid
Production - Batch Systems 5.3.3 Effect of Agitation on Biomass and Citric Acid
Roduction - Fed- Batch Systems 5.3.4 Efféct of Gney Lightnin In-Line Agitator on
Biomass anci Ciûic Acid Roduction - Fed-Baîch Systems 5.3.5 Effect ofAgitationand3 CycleFed-Batch
Fermentation on Biomass and Citric Acid Yields 5.4 Conclusions
Chapter 6: Conclusions and Recommendations Page 94
Chapter 7: References
Appendix A: Anaïyticrl Methods
A.l Yeast Culture A.2 Cultivation of Yeast h m Lyophilized Reparation A.3 Hydrocarbon Feedstock A.4 Yeast CeU Gmwth and Fermentation Media AS Analytical Methods
AS. 1 Biomass Concentration A S .2 Hydrocabn Concentration A.5.3 Citric Acid Concentration
A.5.3.1 Enzymatic Assay Analysis AS .3.2 High Performance Liquid Chromatography
Anaiysis
Appendlr B: Raw Data B. 1 Screening of Hydrocarbon Feeàstock and Yeast Strain B.2 Raw Data of Central Composite Design Flask Batch Tests B.3 Study of Effccts of Aeration on Citric Acid and
Biomass Production B.4 Batch Fermentations - Effécts of Agitation on Citrie Acid
and Biomass Production B.5 Fed-Batch Fermentations - Effects of Agitation on Cihic
Acid and Biomass Production B.6 Fed-Batch Fermentations - Effect of Feeding Rate on
Citric Acid and Biomass Production B.7 Batch Famentations - Effcct of Extenial Agitator B.8 Fed-Batch Famentations - Effect of E x t d Agitator B.9 Fed-Batch Fermentations - Effect of 3 Cycle Feeâing System B. 1 0 Standard Curve
Appcndi. C: Centrai Composite Esperimenbl Design
C.1 Background C.2 Four-Factor Centrai Composite Design C.3 Anaiysis of Moâei
D. 1 Citric Acid Production D.2 Biomass Production Mode1
Appendix E: Ce11 Lyses and Cell Viabiiity
E.l Cell Lyses E.2 Ce11 Viability
Appendix F: Simple Cdcilations
xii
LIST OF TABLES
Page
Table 1 : Yeast Culture and Hydrocarbon Feedstodc Composition in Re-Fermentation Stage Biornass Growth Flasks
Table 2: Parameter Levels of Central Composite Design
Table 3: Quantitative Values of Parameter Levels
Table 4: Summary of Empirical Models Developed for Citric Acid and Biomass Production Mode1 Response Parameter Estimates
Table 5: S m m m y of Aeration EEècts on Biomass and Citric Acid Production in Batch Fermentation
Table 6: Summary of Agitation Effects on Biomass and Citric Acid in Batch n-Parafan Fermentation
Table 7: Summary of Agitation Effects on Biornass and Citric Acid Production in Fed-Batch Fermentation
Table 8: Summary of Agitation Effects with and without Greey Lightnin In-Line Agitator on Biomass and Citric Acid Pmduction in Fed-Batch Fermentation
Table 9: Summary of Agitation Effects with and without Greey Lightnin In-Line Agitator on Biomass and Citric Acid noduction in 3 Cycle Fed Batch Fexmentation
LIST OF FIGURES
Figure 1: Worldwide Citric Acid Production
Figure 2: A Simplined Metabolic Scheme Showing Citric Acid Production h m Carbohydrates and n-Aikanes
Figure 3: Schematic ofCbemap Famentor Utilized for For Fermentation Ruas
Figure 4: Schematic of Greey Lightnin Exterd Agitatar In-Line with fermentor
Figure 5: Cihic Acid Fermmtation using Kerosene and Candida lipoi'ytica Y-1094
Figure 6: Citric Acid Fermentation using Kerosene and Candido lipofytica Y-1095
Figure 7: Citric Acid Fermentation using n-Paraffin and Candido lipoiytica Y- IO94
Figure 8: Citric Acid Fermentation using n-Par* and Candi& lipolytica Y- 1095
Figure 9: Surfice Plots of the Ernpirical Model of the Effêcts Initial Biomass Iwcuium, n-Paraffin Concentraiion, Fmic Nitrate Concentration and Temperature on the Production of Citric Acid
Figure 10: SUTface Plots of the Ernpirical Model of the Effccts Initial Biomass hoculun, n-Paraffin Concentration, Ferric Nitrate Concentration and Tanperature on the Production of Biomass
Figure 1 1 : Normal RobabiIity Plot for the Rcsiduals fhm the Citnc Acid Production Model
Figun 12: Normal Probability Plot for the Residuals âom the Biomass Production Modtl
Figure 13: Residual Plot for Citric Acid Production Mode1
Page
2
xiv
Figure 14: Residual Plot for Biomass Production Mode1
Figure 15: Aeration Effects on Biomass Production in Batch n-Paraflh Fermentation
Figure 16: Aeration Effects on Citric Acid Roduction in Batch n-Paraffin Fementation
Figure! 17: Agitation Effects on Biomass Production in Batch n-Parafnn Fermentation
Page 56
Figure 18: Agitation Effects on Citric Acid Production in Batch n-Parafb Fermentation 76
Figure 19: Agitation Bffccts on Biomass Roduction in Fed-Batch n-Paraifin Fermentation 81
Figun 20: Agitation Effects on Citric Acid Production in Fed-Batch n-ParafEn Fermentation 82
Figure 21 : Agitation Effects on Biomass Production in Fed-Batch n-Parafb Fermentation using Greey Lightnin Agitator 85
Figure 22: Agitation Effccts on Citnc Acid Production in Fed-Batch n-Pafna Fermentation using Greey Lightnin Agitator 86
Figure 23: Agitation E f f ' on Biomass Production in 3 Cycle Fed Batch n-Paraffh Fermentation with and without Greey Lightain Agitator
Figure 24: Agitation Effects on Citric Acid Production in 3 Cycle Fed Batch n-ParafEn Fennentation with and without Gney Lightnin Agitator
FDA
L-LDH
NADH
NRRL
M M
LIST OF ABBREVIATIONS
Arialysis of variance
American Type Culture Collection
Citric acid
Citrate lyase
Food and h g Administration
L-lactate dehydrogenase
Malate dehydrogenase
Nicoinamide-adenine denucleotid
Northem Regional Research Laboratory
Response sdâce me-
Volume of air per volume of medium per minute
Biomass concentration, g/L
Coded Initiai biomass concentration, % volume
Citric acid concentration, g L
Expected value of Y
Depees of Fmdom
Coded Ferric nitrate concentration, mg/L
Codcd n-Paraffin concentration, % volume
Number of factors being studied
Total number of design points
Number of center points
Number of coefficients
Coefficient of Detexmination
Sum of squans
Coded Temperature, O C
Coded factor
Initial biomars concentration, % volume
n-Paratnn concentration, % volume
Fmic nitrate concenttation, mg/L
Tempexahue, OC
Observai value
Greek Letrers
a Distance from star point to center
l3 Mode1 parameter
CHAPTER 1: Introduction
-
Citric acid is widely used in food, phaceutical and chernical indusûies due to its
distinctive properties as an aciduient, flavoring agent, antioxidant and chelating agent.
Commercially, citric acid is produced fiom molasses, sucrose, or glucose syrups by
AspergiLIus niger in sUTface or submerged oxidative fermentation processes. Several
yeast have also been show11 to produce citric acid h m a wide variety of carbon sources,
incluallig glucose, n-*es, and edible oik, in submerged oxidative fermentation
processes. This thesis focuses on the evaluation of producing citric acid through the
utilization of yeast and n-alkanes, as a potmtial alternative to the more haditionai use of
carbohydrate substrates. The results of the study will be considered for evenhial
commercialization possibilities by Touchstone Contracting, Management and
Engineering Services Ltd.
Q 1.1 Citrie Acid Production Worldwide
Ciüic acid is rnanufactured in over 20 counûies with 1994 production at approximately
550 thousand metric tons with a value at over $750 m o n (Blair and Staal, 1991;
Brady, 1996). The distribution of worldwide production is show11 in Figure 1. Most of
this ptoduction is used for foods and ôevcrages, with industrial applications (e.g.
detergents, metaI cleaniiig) becoming mon important on a worldwide basis. The world's
largest proàucing companies are Bayer and ADM (each of which has a 17% market
sâare) followed by Jungbunzlauer, Cargill and Citrique Berge (Bradly, 1996).
Figure 1 : Worldwide Citric Acid Roduction (Blair and Staal, 199 1)
Western Europe produces more citric acid than any other region, while the United States
is the m e r up. The Unitcd States is a net importer of citric acid, while Westcm Europe
is a net exporter. Altbough domestic consumption volumes in Western E m p e and
United States in 1994 were mughly equal at 192-200 thouand metric tons, end-use
market demaad in the two regions varies (Brady, 1996). For hsîance, Western European
demand for citric acid use in household detergents and cleanem was 30% higher than that
in the United States, even though the United States was still transitionhg out of
phosphate builders for laundry detergent formulations in 1994 ( B d y , 1996). Similarly,
the United States utiiized more citric acid in food and beverages applications, particuiarly
in beverages, than did Western Empe.
Canada is cmently without a citnc acid producer, and was estimateci in 1995 to have
imported 1 1.9 thousand metric tons of citric acid h m the United States and Europe at
approximately $2.40/kg (Touchstone Engineering Ltd., 1996). Similarly to the United
States, the largest market in Cariada for citric acid is as a food acidulent. The U.S. market
is relatively matun and is growing at a rate of 305% annually. The Canadian food
acidulent market for ciûic acid is said to have expaimced measufable growth in ment
years; there has been some shiAing of the dry drink mixtures business to Canada h m the
U.S. to take advantage of lower sugar prices (Touchstone Engineering Ltd., 1996). The
major ciûic acid buyers in Canada are food and pharmaceutical companies situated in
Toronto and Montreai. For this reason, Touchstone Engineering Ltd. is proposhg to
establish a Canadian facility in eastem Ontario, to best s m e the major consumers of the
two ngions. Touchstone Engineering Ltd. has proposed the use of hydrocarbons as a
cheap fermentation feedsfock for the production of citric acid As thae is c ~ e n t l y w
plant in production utilinng hydrocarfions, the Company required an investigation, by the
University of ûttawa, into the feasibility of utilizhg hydrocsrbon f b k s .
1.2 Objectives and Scope of Research
The o v d purpose of the research was to evaluate the technical feasibility of the
production of citric acid through a fermentation process utilihg yeast and n-alhes.
SpecificaUy the goals of the research were:
1. To determine a strain of the Candi& l@oZlytica species and hydrocarbon feedstock for
use in the study of citnc acid production;
2. To study the effects of ternpmture, initial biomass inoculmn, initial hydrocarbon
concentration and iron sait concentration on the production of citric acid;
3. To observe the effects of aeration and agitation on the production of citric acid, while
also comparing the effkcts of mechanicd agitation in the ftnnentor itself and
mechanical agitation e x t d y ;
4. To compare citric acid production between batch and fed-batch fermentations.
In selecting the strain of Candida Iipolytica to be used throughout the study, two sûains
were obtained b m the United States Department of Agriculture, C. ZipoZyticco MZRL-
Y1094 and C. lipolytica NRRL-YI095 The two strains were used in batch fermentations
using two hydrocarbon feedstocks, n-paraffin and kerosene. The Càndih lipoytica
strain and feedstock uscd in the subsequent studies of the thesis was the yeast and
hycitocarbon fecdstock fcrmcntation mixture with optimum ciûic acid pmduction; C.
l ipol~ca NRRL-Y1095 and n-par& f d t o c k .
A ceneal composite design was implemented to determine the effeçts of four pararneters,
temperature, initiai biomass inoculum, initial hydrocarbon concentration and iron salt
concentration, on the production of ciûic acid. From the design, the level of each
parameter for optimal citric acid production was deteRnind. These opaating levels for
each parameter were then used for the subsequent shidy of larger scale fermentations.
The effects of aeration and agitation on ciûic acid production were investigated using a
Chcmap famentor. An extemal agitator was also introduced to the famcator system and
citric yields compand betwem fermentations under only mechanical agitation h m the
fermentor and fermentations under mechanical agitation h m both the fmentor and
extemai agitator. Citric acid yields were also compared between batch and fed-batch
fmentations.
Q 13 Layoat of Thesis
The thesis has been divided into three sections:
the selection of a yeast strain and hydrocarbon feedsfock;
the setting of operating values for temperature, initiai biomass inoculum, initial
hydracarbon concentration and iron sait concentration;
the assesment of the technicd faiibiity of the production of citric acid
tbrough hydrocarbon fermentation, unâer the influaces of d o n , mechanical
agitation (hm bot. a Chemap fcrmentor and an aannal agitator).
The bais for this approach is presented in Chapter 2. This chapter includes a critical
analysis of the development and evolution of citric acid production. Chapter3 presents
the expaimental method and aiialytical basis for Candida ZipZytico strain selection and
the selection of a n-paraffin solvent as the hydrocarbon f d t o c k . Chapter 4 presents the
design and d y s i s of the central composite design for the selection of operating values
of temperature, initial biomass inoculurn, initial hydrocarbon concentration and iron salt
concentration. Chapter 5 includes the experimental method and comprehensive analysis
of the effects of aeration and agitation on citric acid production under the infiuence of
both batch and fed-batch systems. This chapter also focuses on examining the effects of
an extemal agitator, added to aie fermentation apparatus, on citric acid production. A
sumxnary of the hdings h m the entire work and recommendations for firrther nsearch
are presented in Chapter 6. Refmnces are provided in Chapter 7.
CHAPTER 2: Production of Citric Acid: A Fermentation Approach
(Citric Acid, Anhydrous)
- - -- - --
5 2.1 Citric Acid Background
2.1.1 Physicd Propertîes
Citric acid (2-hydroxy-l,2,3-propanetricarboxylic acid), is a naturai camponent and
common metabolite of plants and animais. It is the most versatile and widely used
organic acid in fw&, beverages, and pharmaceuticals.
CH,-COOH I
HO-C-COOK I
CH,-COOH
Because of its hctionaiity and environmental acceptability, citric acid and its salts
(primarily sodium and potassium) are used in many industrial applications for chelation,
buffering, pH adjustment, and derivatization. These uses include laundry detergents,
shampoos, cosmetics, aihanced oil recovcry, and chemical cleanllig (Blair and Staal,
1991). Aqueous solutions of citric acid are exceîlent b u f f i g systans when partidy
neutralized because citric acid is a weak acid and has thnc dmxylic pups, haice thrte
pK's. At 20°C pK, = 3.15, pK, = 4.77, and pK, = 6.39 (Weast, 1989), while the
buffering range for citrate solutions is pH 2.5 to 6.5 (Blair and Staal, 1991).
2.1.2 Chemicai Properties
Citric acid undergoes most of the reactions typical of organic hydroxy polycar&xylates:
Decomposition: When heated above 17S°C, citric acid decomposes to fonn aconitic
acid, citraconic acià, itaconic acid, acetonedicariKIxyiic acid, carbon dioxide, and
water (Blair and Staal, 1991).
Esterification: Citric acid is easily esterified with many alcohols under azeotropic
conditions in the presence of a catalyst such as sulfure acid, p-toluenesulfonic acid,
or dfonic acid-type ion-exchange min. Alcohols typically used in citric acid
esterification are methyl, ethyl, butyl, and allyl aicohols (Blair and Staal, 199 1).
Chidation: Citric acid is eafily oxidized by a variety of oxidipng agents such as
peroxides, hypochlorite, p d a t e , permanganate, peziodate, hypobromite, chromate,
manganese dioxide, and nit& acid. The products of oxidation are usually
acetonedicarboxylic acid, oxalic acid, carbon dioxide, ami water (Blair and Steai,
1991).
Salt Formation: Citnc acid forms mono-, di-, and tribasic salts with many cations
such as alkalies, arnmonia, and amines. Salts may be prepareâ by dwct
neuealuation of a solution of citrie acid in water using the appropriate base, or by
double decomposition using a citrate salt and a soluble metai salt (Blair and Staal,
199 1). Trisodium ci- is more widely used than any of the other salts of citric acid.
Chelate Fornation: Citric acid complexes with many multivalent metai ions to fonn
chdates. This imwrtant chcmicaî ~ronertv maka citric acid and citrates useM in
controlling metal con tamination Uiat can affect the colour, stabiiity, or appuirance of
a product or the efficiency of a process (Blair and Staal, 199 1).
2A3 Occurrence of Citrîc Acid
Citric acid occurs widely in the plant and animal kingdom. It is found most abundantly
in the fhits of the citm species, but is also present as a fiee acid or as a salt in the fruit,
seeds, or juices of a wide variety of flowers and plants. The citrate ion occurs in al1
animal tissues and fiuids (Blair and Staal, 1991).
Citric acid occurs in the terminal oxidative metabolic systezn of vimially al1 organisms.
This oxidative metabolic system (Figun 2), variously called the Krebs Cycle (for its
discoverer, H. A. Krebs), the ûicarboxylic acid cycle, or the cikic acid cycle, is a
metabolic cycle involving the conversion of carbohydrates, fats, or proteins to carbon
dioxide and water. This cycle releases energy necessary for an organism's growth,
movemenh luminescence, chemosynthesis, ami reproduction (Blair and Staal, 1991). The
cycle also provides the carbon containing materials h m which celis synthesize amho
acids and fats. Many yaists, mol&, and bacteria conduct the citric acid cycle, and cm be
selected for theK ability to maximize ciûic acid production in the process. This is the
bais for the commacial fermentation processes wed today to pmduce citric acid.
Figure 2: A Simplified Metabolic Schane Showing Citric Acid Production fkom Carbohyârates and n-Alkanes (Milsom and Meers, 1985)
5 2.2 Citric Acid Production Processes: Fermentation
From among the historically used processes for the production of cihic acid the following
are still important: (MilSom and Meas, 1985)
(a) Aspeqtiüus neet
i. Sinface fmentation using bect molasses;
ii. Submeqed fermentation using beet or cane molasses or glucose syrup.
@) Y'èmt
i. Submaged fermentation using beet molasses or glucose symp;
ii. Submerged fermentation using n-aikanes.
2.2.1 Processes using AsprrgUlus niger
2.2.1.1 Surf ce Fermentation
The surface fermentation using A. niger with beet molasses as raw material is stili being
used by some manufacturers (Milsom and Meers, 1985). The power rcquirements,
however somewhat labour intensive, are much less than in the submerged fermentation.
Because ciûic acid manufachirers keep their methods a secret, linle authoritative material
has been published. However it is lmown that when using beet molasses, an initial pH
must be adjusted betwcen 5 to 7 as A. niger will not germinate at higher hydrogen ion
concentrations. The lack of germination in molasses at low pH is ascribed to the presence
of acetic acid, which is a nomal constituent of molasses (Fend and Lmpold, 1957).
Unionized acetic acid is the sptcies haî prevents the germination.
Once the molasses based medium is sterilizcd and cooled, it is run down into a series of
tmys supporteci on racks in a vnitilated chamba. The trays, which are usually made of
very high purity aluminum, arr fiiled to a depth of betwan 0.05 and 0.2 m (Mihm and
Meers, 1985). Spores of A. niger an thea distributed over the surfiace of the medium in
the trays. Staile air is suppliai to the fcxmcntation chambcr. The air perfonns the dual
fimction of supplying oxygen and cprrying away fermentation heat. A tempaatun of
30°C is often employed. After a period of 7 to 15 &ys the trays are emptied, while the
mycelirmi is separateci at the same time h m the fermentation broth. The fermentation
liquor is then pumped forward where the citric acid is recovd.
2.2.1.2 Submerged Fermentation
Althou& the production of citric acid Uuough surface fermentation using strains A. niger
is still being used today, it is being replaced by a submerged process hown as deep tank
fermentation (Blair and Staal, 1991). In submerged fermentation a vegetative inoculum
stage is used to ailow spores of A. niger to genninate in an inocuium medium before
being transfmed to the main fmentation medium in a larger fermentation vessel. This
initial growth stage is of the utmost importance to the success of the fennatation. "The
morphology of the mycelium at this point is crucial according to many reports, not only
in relation to the shape of the hyphae themselves, but also ui the aggregation of the
growth into srnail sphencal pellets. nius the hyphae should be abnomally short, saibby,
forked and bulbous." (Snell and Schweiger, 1951; Kisser et al., 1980)
When a separate inoculum stage is employed, a suspension of spores of A. niger is
introduccd into a sterilizcd medium in the inoculum fmmtor. The medium is acrated
and agitated at a tempcrature of about 30°C for a paiod of 18 to 30 hours (Hustede and
Rudy, 1976). At which point the stcrilized fcnncntation medium is transferred to the
main famentor and the p w n inoculum iacoprated at the ratio 1 L of inoculum to 10 L
of fmentation medium. In molasses based iaoculum and femicntation media the initial
pH is n o d y in the range of 5 to 7. As pmiously mentioned, A. niger strains do not
12
gmninate or gmw at lower pH vaiues in this type of medium. However, lower pH values
cm be tolerated in glucose or sucrose based media, and are offen used with Avantage
(Milsorn and Meers, 1985). At lower pH values th= is a lesser chance of infection by
adventitious organisms.
When the rate of increase in cieic acid concentration has reached a point where it is
uneconornical to proceed, the fcnnentation is discontinueci and the broth pumped to a
filter or centrifuge where the mould is separateci h m the liquor which is passeci forward
for citric acid recovery (Parton and Willis; 1990).
2.23 Processes nsing Yeast
2.2.2.1 Submerged Fermentation usntg Carbohydrates
In 1965 Tabuchi and Abe (1968) filed a patent for the manufacture of citric acid &om
glucose and molasses in which the organisms used were eight genera of ycasts including
Candida. The fermentation was canied out at a neutral pH by being buffaed by calcium
carbonate incosporated in the medium. In 1971, a firrther patent by Iizuka (1971) reports
yields of about 65% h m blackstrap molasses containhg calcium carbonate ushg strains
of Candida oleopküo. However. it was soon found thai a limitation on citric acid yitld
was the pduction of quantities of L-(+)-isocittic acid (Kyowa Hakko Kogyo LM.,
1972). Kimura et al. (1974) selecttd a subspecies of Candi& piiLiennondii producing
only srnail quantities of iso-citric acid. A continuous fermentation pmcss for the
production of citric acid h m blaclrsbap molasses using Condida guzlliennondii has ban
d d b e d by Miall and P p k a (1975).
An inoculum culture is prepared in a smaller fmentor. Whm the inoculum is
sufficiently grown it is transfened to a fennentation medium in the larger main famentor.
The fennentation is conducted at a temperature between 25 end 37°C and a pH that is not
too low (dcpending on the organism), or yeast growth will be impaired. The pH cm
subsequentiy be allowed to faU When citric acid accumulation becornes imcconomicaily
low, the broth is hamesteci for citric acid and yeast.
The advantages of using yeasts ratha than A. niger, is not only that yeasts are more
tolerant to infection by adventitious organisms, but are also capable of using very high
initiai sugar concentrations and have faster fermentation cycles. This combination resuits
in high productivity rates per m.
2.2.2.2 Submerged Femtentation uring n-Alkanés
It has been known for some time that some migoorganisms are able to gmw on
hydrocarbon substrates (Bushnell and Haas, 1941; Milsom and Meers, 1985). Johnson
(1964) noted that organisms prefened straight chah alkanes to branched compounds and
that the prefmed chah length depended on the strain. Takeda Chernical Industries Ltd.
(1970) first reporteci the production of citric acid by yeasts h m n-aliranes in the patent
literature. The patent claims the use of Condido species, in particular C. ZipoSiceo, C.
tropicafis, C. intemedia, CC. pampsiiosis and C. guiIZiennondii. The patent has a priority
date in Japan of June 7,1967 and swns to be a master patent whac the use of Cmidda is
concerned in the production of citric acid fÎurn n - b c s (Mîlsom and Meers, 1985).
n-Ahanes are insoluble in aqueous sohtions, thaefore it is essential during fermentation
to maintain a fine dispersion of the hydrocarbons in the aqueous phase. To aid this
dispersion mechanicd agitation must be implernented during the fernatation proces.
The subrnergeà fermentation using n-alkanes, as that ushg carbohydraes, has an initiai
inoculum p w t h phase. Once the inocdm growth is sufficiait it is then transfd to
the main fermentor containing the fermentation medium. The fermentation is conducted
at a temperature between 25 and 37OC and a pH that is not too low, between 4 and 7
(depending on the organism), that yeast growth will be impaircd. The pH cm
subsequently be allowad to fall, anywhere between 2.8 and 4 (again depending on the
organism), as citric acid accumulates (Miall and Parker, 1975).
The advantages of using n-aikanes over carbohydrates are lower substrate costs and
potentially higher citric acid yields. Starting h m a highly reduced fonn of carbon such
as n-aikanes, very high weight yields of citric acid are possible. The 'theoretical' yield is
about 250% but allowing for biomass and carbon dioxide production a reasonable value
would be 175% (Milsom and Meers, 1985).
8 23 Citric Acid Producing Yeast: Candida I ~ o f ~ c t z
The ability of some microorganissns, incluâing yeasts, to utiîize ailcanes as a source of
carbon has beai known for many years, and during the Second World War these
orgpnisms were identifiai as causative agents in the biodcterioration of oil supplies
(Rose, 1987). The intercst in alkane-utifizing microorganisms grcw in the early 1970s as
they were considered to be potential sources of single-ce11 protein using oil as a
feedstock. Research became intense when they were seen as potentially very acceptable
food and fodder mimofganisms (Rose, 1987). There are many kinds of yeasts capable
of producing citric acid fmm various hydrocarbon sources (Tabuchi and Igoshi., 1978).
A historical m e y of growth of yeasts on aikanes came h m Sherman and Levi in 1974
who reviewed production of yeast biomass h m alkanes. At least seven genaa of
ascomycetous yeasts (Debaryomyces, Lodderomyca, Metschikowza, Pichia,
Sacclraromycopssis, Schwanniomyces and Wingea) contain species that can grow at the
expense of hydrocarbons (Rose, 1987). But undoubtedly the most studied yeast genus
with species able to grow on hydrocarbons is Candida, while the genera Rhodotonila,
Selenotila (now considenxi to be Candida species), S p ~ r i d i ~ b ~ l ~ ~ , Sporobolomyces,
Tonclopsis and Trichosporm complete the list of yeast gcnera with species able to utilize
n-allranes among the Deutmmycetes (Rose, 1987).
A considerabie number of members of the genus Condido will accumulate citric acid
when they are aerobically propagated in a nutrient meàium comprishg an appropriate
carbon source, intimaîely mixed with an quebus phase containing essential rninaals and
p w i h factors. Cadidido ZipZyhca, Candida guilIiermondii and Gzndidu tropicaZis have
successfuily shown the ability to assimilate hydrocarbons for the production of citric acid.
The thne Condida specics have demo~l~trated citric acid yielâs in the range of 464 12g&
( P h r hc., 1975; Takcda Chunical Industries Ltd., 1975). Candi& Iipolyh*ca has
shown to gcncrally producc higher citric acid yitlds than C&da guilliemondii
(Tetaaishi et al., 1974; Shimiai et al., 1974), with a P h Inc. patent (1975) iflustrahg
citric acid yields 20 times bigher when using Candida l@oi'yticu. However, Candida
1ipoIyticco has not consistently shown higher citric acid yielàs than Candi& tropiculis.
Takeda Chernical Industries Ltd. demonstrateci in its two patents (1975,1972) 5% higher
citric acid yields when usiag Candido lipojmca over Candida tropcaks, while
Matsumoto et al. (1984) illustrateci 25% higher citric acid yields when using Candida
tropicah over Candida li>oZyticco. Teranishi et al. (1974) and Shimuu et al. (1974)
studied the respiration (oxidation of substrate) behavior of Candida lipolytica, Candi&
guilliennondii and Candida tropicolis on n-alkane substrates. Consistently they found
that despite the fact that Candida iropiculis cells showed a higher n-alhe-oxidizing
capacity than either Candida lipoljtica or Candida g~illiennondii~ the ce11 yield (based
on the amount of substrate consumed) was relatively low. In their stuàies, Candida
tropicalis demo~l~trated a significantly greater sensitivity to respiratory inhibitors,
cyanide and &de, when grown on n-aUcanes where respiratory activities were lowered by
about 50%. Teranishi et al. (1974) and Shimiai et al. (1974) concluded h m th& snidies
that each Candida yeast possess a cornplex respiratory system that is significantly
changeable depending on the growth and fermentation conditions that are applied, iike
the carbon substrate used or the cell-cycle phase.
From these studies, al1 thne Candida species have shown both succ~~~fù î biomass
gmwth and ciûic acid production rates while assimilating n-alkane substrates, and were
considered as potential candidates for the study. However, Candida pillientlondii and
Condido tropicaiis are classined as human pathogms @ougias, 1987; Hmley et al.,
1987). W e the laôoratory useci for the fenaentation shidies in this thesis was not
equipped to handle pathogenic cultures, C. guilliennondii and C. tropiculis could not be
considered for the shidy. Therefore the fmentation studies was wnducted using
Candida lipolyticu qecies.
23.1 CllltdIda lwfpacu-The Oginbm
The discovery in the late 1960s that high yields of citric acid were pduced by Candida
lipofytica when grown on n-paraffins signaied a potential dnimatic change in the
fmentation industry because this organic acid had been commercially produced since
19 19 by Aspergillus nzger (Kurtpnan, 1988). The hdings occumd at a t h e when sugar
substrates were being increasingly diverted to food use and petroleum was still
inexpensive. Whether C. lipolytica replaces A. niger depends upon the price of
petroleum.
Food and Drug Administration (FDA) regulations (1995) document C. lipoytica as a
food additive that may be safely used as the organism for fermentation production of
citnc acid. The food additive is the enyme system of îhe organism C. lipol'ca and its
associated maabolites produced during the fermentation process. Candida liplytica is
the oniy species of the gmus Condida recognized by the FDA regulations as an additive
intended for the use in the fennentatîon pmess for the production of citric acid h m n-
akanes (FDA Regulations, 1995).
The noqathogaiic Condida lipdytica organism is ciassifiad as foilows (FDA
Reguiations, 1995):
Class: Deutaromycetes Order: Moniliales Family : C~tococcaceae Genus: Candida Species: lipolytica
2.3.2 Using Cundr'da l@t#y&a in the Production of Citric AcY
Many studies utilizing Candi& lipoiytica in the production of cilric acid have been
conducted on both carbohydrate and hydrocarbon feedstocks. Reccntly, Rane and Sims
(1996) showed ciîric acid yields of Candida lzpolytica (NRRL Y-1095) grown on glucose,
in a 2 litre mechanically agitated batch system, in the range of 0.38-0.77 g citric acidlg
glucose consmed. While in 1994 Rane and Sims (1995) described citric acid yields of
Candida iipoi'ytica ( N . Y-1095) p w n on glucose, in a 2 litre mechanicaliy agitated
fed-batch system, in the range of 0.48-0.59 g citric acidg glucose consumeci. Wentworth
and Cooper (1996) investigated citric acid production ushg Candida lipoi'ytica (ATCC
20390) and glucose in a self-cycling fermentation system. The self-cycling fennentor
consisted of a cyclone column famaitor in which the broth is continuously recirculated.
In this case Wentworth and Cooper describsd citric acid yielâs in the range of 0.1 1-0.23 g
citric acid/g glucose consumcd. A few years early, BrifEaud and Engassa (1979-0 alm
describexi a citric acid yield for Candida lipoytica (D 1805) grown on glucose, in a 2 liîre
mechanically agitated batch systcm, of 0.75 g ciûic acid/g glucose consumecl. They
hther described a trickic-flow fermentation (Briffaud and Engasser, 1979-II) with
Cadida IipoIMca (D 1805) gmwn on glucose, in a 3 litre sphcrical flask, with a slightly
lower ciûic acid yieid of 0.63 g citric acid/g glucose c o ~ l ~ ~ l ~ l c d . The tricklc-flow
fenncntor wnsisted of a g las column Gxed in a stimd sphcxicai flask. The colunm was
packed with cylindncal wood shavings. The culture medium was pumped from the flask
to the top of the columa, then it trickled d o m the wood packing. Air was introduced at
the top of the column, and was operated in a cocurrent tnckling moâe. Simultaneously
in 1979, Aiba and Matsuoka (1979) showed citric acid yiel& in the range of 0.14-0.33 g
cihic acid/g glucose consumed in a 4 litre continuow fmentation process using
Candida lipolyticca.
Although research in the a m of citric acid production has concentrateci on the utilization
of carbohydrate feedstocks, there have been numerou studies conducted on the
utilization of hydrocarbon feedstocks. Most of the work on hydrocarbon fermentation for
citric acid production took place predominantly in the early to mid seventies. Much of
the work in hydrocarbon fermentation was cornmaiad by the Japanese, and soon
followed by the North Americans. This interest in hyârocarbon utiiization was driven by
falling petroleum prices, and was quickly abandoncd as petrolem prices drastically
increased in the late seventies, early eighties. Today with moderate petroleum prices,
there is a resuffacing of interest in hydrocarbon utilization for the production of citric
acid, and other organic acids.
Much of the initial work in the production of citric acid by yeast growing on
hydrocarbons was started by Takeda Chernical Industries Ltd. With s w d of their
patents (Takeda Chernicd Industries Ltd., 1970, 1972, 1975), the company developed a
process for the production of chic acid using n-paraffins @aving 13 to 15 carbon atoms)
and C&& lipoZMca (ATCC 20114). The process involves an 8 &y baich fmentation
resulting in citric acid yields in the range of 0.38-1.2 g ciûic acidg a-parafi consumeci.
Takeda Chemical Industries Ltd. stated in their patent that the yeast, while able to utilize
the hydrocarbons, was not able to consume the cihic acid being produced, and isocitric
acid was produced in minute quantities (Takeda Chernical Industries Ltd., 1975). In
1974, Pker Limited also issued a patent for the production of citric acid using Candida
lipolytica (ATCC 20228 and NRRL Y-1094) grown on a n-paraffin feedstock mvhg 14
to 16 carbon atoms) (Mid1 and Parker, 1974; Pfizer Ltd., 1975). This process involves a
6 litre conthuous fermentation nmning for 304 hours (71 hours batch process plus 233
hours continwus process). The patent describes a citric acid yield of 1.27 g citric acid/g
n-parafEn consumeà. Also in 1974 Hattori et ai. illustrated citric acid yields in the range
of 0.51-0.86 g citric acidlg n-para£fin with Candida lipoiytica (KY622I) grown on n-
paraffins (having 12 to 15 carbon atoms) in a 80 hour 3 litre continuous fmentation
process. Gledhill et al. (1973) demonstrated a 7.5 litre semi-continuous ce11 recycle
system utilizing Candida lipoiytica (ATCC 8661) aiid n-paranins (contaiaing 12 to 15
carbon atoms). The ce11 recycle systcm involved nmoving a portion of the fermentor
broth and separating the yeast cells under non aseptic conditions by use of a cenûifbge.
Yeast cells fiom the centrifbged pellet and hydrocarbon iayer werc recombineci, dilutexi to
the original volume in a non sterile recycle medium, and retumed to the fennentor.
Gltdhill et al. (1973) describeci citric acid yielâs of 0.75-0.8 g citric acidg n-parafEn
consumed.
Foliowing these earlicr works, that set the fomdation
Finukawa et al. (1977, 1982) scnaied mutants of the
found a mutant which they classikd as a new species
for hydmcarbon fermentation,
Càndida liplNca parent and
called Condida citnca. AAer
conducting 7 &y 500 mL batch fmentations on a n-paraffin feedstock (containing 14-
17 carbon atoms), they iliust~ated citric acid yields in the range of 0.77-0.86 g citric
acid/g n-parafnn consumed. In 1974 Hat& et al. (1974) also reviewed the citric acid
yields of a mutant and wild-type stralli of the Candida lipoljtica parent strain. The wild-
type strain was classified as ATCC 20346 and the mutant strain was classified as ATCC
20324. It was dernoIlstrateci that the citric acid yield obtained while usiag ATCC 20346
and a n-par& feedstock hydrocarbon in a 7 &y 2 litre fermentation was 0.66 g citric
acid/g n-parafh consumed. When using ATCC 20324 under the same conditions a 38%
higher yield was obtaineâ, 0.89 g citric acidlg n-parafi. In 1991, Avchieva and Vinarov
(1993) also screen the Candzda l@oi)tica (917) parent strah for a higher citric acid
producing mutant. They found mutant 91 7-20 which producecl a citric acid yield of 0.71
g citric acidlg n-paraffin consumeci, when allowed to grow on a n-parafh feedstock for
96 hours in a 750 rnL flask fermentation. The 917-20 mutant produced a higher citric
acid yield than the parent s a by about 9 folds.
Mer miewing the literature survey and a verbal discussion with Dr. Kurtzman, a yeast
taxonomist with the U.S. Department of Agriculture, Candida lipol'ytica NRRL Y-1094
and Candido lipolyticu NRRL Y-1095 were chosen as the candidates for this shidy. NRRL
Y-IO94 was suggestcd by the Pfizer Limited patent (Miail and Parka, 1974) as a good
culture, and this was confkmed by Dr. Kurt~nan. Dr. Kiirt~nan also suggcstcd NRRL. Y-
1095, as the ycast had a successfiil reptation when used on glucose for citric acid
production and should be as succasfiil on hydtocarbons.
Q 2.4 Composition of Medium
Medium formulation is an essential stage in the design of successfbi laboratory
experiments, pilot-sale development and manuf'acturing processes. The wnstituents of a
medium must satisfy the elemental requirements for celi biomass and metabolite
production and there must be an adequate supply of energy for biosynthesis and ce11
maintenance (Stanbury, 1995). This section will focus on the components of the medium
that will be used in this study on ciûic acid fermentation. These components include,
carbon source, nitmgen source, mineral sources, biomass concentration, pH and
temperature.
2.4.1 Carbon Source
Energy for p w t h cornes fiom either the oxidation of medium components or h m light.
Since most industrial micro-organisms are chemooganotropbs (Stanbury, 1995), the
most common source of energy is the carbon source. The most common carbon sources
are carbohydrates, hydrocarbons, lipids, methmol and proteins.
It is now recognwd that the rate at which the carbon source is metaboiized can oAen
influence the formation of biomass or production of primary or seconâary metabolites.
The main product of a fcrmcntation process will offen determine the choice of carbon, as
alternative carbon sources wili affect the yield of product and thus the wst of producing
biornass a d o r metabolites.
In reviewiag the literature s w e y the most evident feedstock to utilize is n-paraffUl
(Hattori et al., 1974; Glendhill et al., 1973; Takeda Chemical Industries Ltd., 1970, 1972,
1975; Miail and Parker, 1974; Pfizer Ltd., 1975; Avchieva and V h v ; 1991). The n-
parattlli feedstock of choice should contain 9 to 19 carbon atoms, particularly n-parafhs
containhg 12 to 16 carbon atoms or 14 to 19 carbon atoms (Miall and Parker, 1974).
However, the Pfizer Ltd. (1975) patent suggests utiiizing kaosene as a potential
feedstock. Although kmsene is a mixture of aiicanes and alkenes, its significantly
cheaper cost may prove it to be the more economicaily viable choice. Thmfore, the two
feedstocks chosen for meening in this study were n-paratnn (Noqat 15) and kerosene
(Kero 1-K) solvents obtained tiom Imperial Oil Ltd., Toronto, Canacla. Norpar 15
contains more than 97% mixed nomal parafnns composed of predominantly 14 to 17
carbon atoms. K m L X is compaseâ of a 46053% total paraffins mixture, including both
branched (iso) and normal paraffllis having predominantly 9 to 13 carbon atoms. Less
than half of the total parafnns contait is normal parafnas. Approximately 18% of the
kerosene solvent is composod of aromatic molecules. The main interest in K m 1-K is its
market value which is half that of Norpar 15. The amount of hydrocarbon feedstocks
utilized in the fermentation process of this study will be d e t d e d by Nnaing batch
flask fermentation tests and performing a central composite design on the tests to waluate
hydrocarbon levels for optimum citric acid production. However it has bmi stated by
Takda Chemical Industries Ltd. (1970, 1972, 1975) and Pfizer Ltd. (Mid aml Parker,
1974) that operathg hydrocarbon levels shouid range betwecn 5-20 % by volume,
aithough they do stiûe a lessa or higba: amount can be used.
2.4.2 Nitrogen Source
An essential nutrient for biomass reproduction is nitrogen. However citric acid
accumulation occurs in a nitrogen limited environment. Most rnicmrganisms can utilize
inorganic or organic sources of nitmgen. Inorganic aitrogen may be supplied as ammonia
gas, ammonium salts or nitrates. Ammonium saits such as ammonium stdphate will
usually produce acid conditions as the ammonium ion is utilized and the fke acid is
liberated. ûrganic nitmgen can be supplied as amino acids, protein or una In rnany
instances p w t h will be fhier with a supply of organic nitrogen (Stanbury et al., 1995).
The nitrogen source and content in this study has been modelecl afta the Pfizer Ltd.
patent (Miall and Parker, 1974). Urea was chosen as the nitmgni source in the
femientation broth as it can be consumed by the organisms quicker than an inorganic
source, therefore reaching a nitrogen limited enviro~lrnent sooner for citric acid initiation.
The concentration of urea utilued in the onidy was 2.0 fi.
2.4.3 Minerai Sources
Al1 micmorganisms rquire certain mincral elements for p w t h and metabolian (Parton
and Willis, 1990; Staubuyy et al., 1995; Midl and Parker, 1974). In mimy media,
magnesium, phosphorus, potassium, suîphur, calcium and chlorine are essential
components. ûthcrs such as copper, imn, mangrnese and zinc arc essential as trace
elements. The mincral componaits useâ in the study have baa modelcd after the P k c r
Ltd. patent (Mid and Parka, 1974). However the iron content is to be detcrmined by
batch fiask fermentation tests where the iron salt level will be detenniried for optimal
citric acid production.
Iron concentration in the fermentation medium for citnc acid production is essential
because it determines the ratio of citric acid to isocitric a d being produced. Too high an
iron concentration wili result in high isocitric acid levels, while too low an iron
concentration will result in low citnc acid yields (Akiyama et al., 1973). Iton is essential
for activating the citrate synthetase enzyme for citric acid production, however high iton
levels will activate the aconitate hydratase (aconitase) enzyme for isocitric acid
production (Akiyama et al., 1973, Shu and Johnson, 1948). The recommended level of
iron salt concentrations is between 10 to 100 mg/L (Stanbury et al., 1995).
2.4.4 pH and Temperatun
Traditionally ce11 culture medium have becn buffimd with a bicarbonate/car&n dioxide
system. The normal starting pH range of the most commonly employed culture medium
is between 6.8 and 7.8 (Lavery, 1990). The optimum for each ce11 type will vary
depending on the physiology of the culture. In this study, the pH will be aiiowed to drop
as citric acid is being produced. The pH will be maintained at 4.0 with a 1 .O N NaOW1 .O
N HCl system. Takeda Chernical IndusaTies Ltd. (1970, 1972, 1975), P M Ltd. (Mal1
and Parker, 1974) state that the pH in the fermentation medium during citric acid
production can be mahtained bctwccn 2.8 and 3.5. They state that the yeast can fiinction
in such am acidic mvllonment. They dso mgpst kcepiag the pH this low so that
contamination can be kept at a minimum. Howevcr, Rane and Sims (1993,1994, 1995)
suggested that the pH of the medium be kept at 4.0 when using Candi& lipoI'*ca lVRRL
Y-1095.
Maintenance of temperature is crucial throughout a fermentation pracess, as biomass
metabolism is highiy variable with changes in tempaature. For the production of citnc
acid, the operating temperature of the fermentation system is to be controlled at a set
point in the range of 2530°C (Midl and Parker, 1974; Funikawa et al., 1977), while
Takeda Chemical Industries Ltd. (1970, 1972, 1975) suggests a temperaturc in the range
of 25-370C. The temperature utilized throughout the fermentation systems in this study is
to be determineci by batch fermentation flask tests and a central composite design, when
the temperature for optimal citric acid production is evaluated.
8 2.5 Aeration and Agitation
The majority of fermentation processes are aerobic and, therefore, wuire the provision
of oxygen (Stanbury et ai., 1995). The oxygen danand of a fermentation process is
usually satisfied by acration and agitation of the fmentation broth.
In hydrocarban fermentation there are at least two physicai problems of engineering
interest which are crucial for hydrocarbon fermentations. The first relates to the very
high oxygem demand per unit ce11 mass and the otha aises h m the îimited solubility of
hydrocarbon oils in aqueous media (Mm-Young et ai., 1971). Stanbury et al. (1995)
states tint a hydrocarbon fermentation rcquires tbree timcs the amount of oxygcn to
produce the same amount of biomass h m carbohydrates. Furthemore, Johnson (1964)
and Aiba et ai.(1969), have suggested that dissolved hydrocarbon oil may not be the main
source of substnue supply, and that rather uptake occurs by ceU attachment to droplets of
oiI.
ûxygen is supplied to the culture in the fom of air sparging through the fermentation
media Stanbury et ai. (1995) represented the tramfer of oxygen h m air to the ce11 as
occumhg in a number of steps:
i. The transfa of oxygen h m an air bubble into solution.
ii. The transfer of the dissolved oxygen through the fermentation medium to the
microbial cell.
iii. The uptake of the dissolved oxygen by the d l .
Both aeration and agitation affect the extent of dissolved oxygen in the medium.
Aeration rates serve as the initial volume of air bubbles introduced to the system whereas
agitation rates sene to assist oxygen transfw in the following ways (Stanbury et al.,
1995):
i. Agitaiion incrcases the ar#i available for oxygen transfer by dispershg air
in the culture fluid in the form of smaîi bubbles.
ii. Agitation delays the escape of air bubbles frorn the liquid
iii. Agitation prevents coalescence of air bubbles.
iv. Agitation demeases the thichiess of the iiquid film at the gas-liquid
interface by crrating turbulence in the culture fluid.
Agitation in hydrocarbon fmentations is also crucial to the dispersion of oil droplets
tbroughout the medium. Mm-Young et ai. (1971) found that smaller oil droplets, the
result of higha agitation rates, resdted in m e r assimilation of the hydrocarbon source
by the yeast. However agitation rates running too high can impose a dangerously large
shear stress on ce11 walls, ultimately decnasing or destroying celî viability.
For the purposes of this study, a Chanap femientor (Figure 3) will be used in the
investigation of the effects of aefation and agitation on citric acid and biomass
production. In this fe~llentor, the air wili be introduced into the medium h m the bottom
of the fennentor to prolong the detention of air in the medium; aiiowing for a longer
suff ie contact between the medium and air bubbles during oxygen tntnsfa between the
two phases.
Air Swsa
pH Probe
b
Figure 3: Schanatic of Chemap Fennentor Utilizcd for Femientation Runs
The air liae is fiîted with a filter to collect any grease that may be present. The fnmaitor
is equip with 4 stainless steel baffie impellers for agitation. The impellen are driven by a
bottom belt dnvt. A pH probe, temperature sensor and dissolved oxygen probe are aiso
present for analy sis and control.
The study will also investigate the potential of ut i lhg an e x t d agitator in the
production of citric acid. Due to the high oxygen demand of the process, an extanal
agitator will be ~ ~ ~ e ~ t e d in-he whh the fmnitor (Figure 4).
Figure 4: Schematic of Greey Lightnin Extemal Agitator In-line with Fermentor
The fermentation broth will be re-circulated through the e x t d agitator (Grcey Lightnin
In-Line Agitaîor) and back into the fcnnentor. Agitation will be implancnted in bpth the
extemal agitator and in the fermentor. The extanal agitator has the potential of serving
thne purposes:
i. As the chamber of the agitator is 5 by 3.8 cm, with an impeïler span of 2.5 cm, it has
the potential of producing smaller, more uniformly dispased air bubbles throughout
the medium. Then will be a more d o m interaction betweea air bubbles
(introduced by the sparger in the fmentor) and impellers. These smaller air bubbles
increase the surface area for interaction betweai the gaseous and aqueous phases;
ii. By rexirculating the medium, thae wiU be a longer detention time of air bubbles in
the medium, therefore allowing for a larger traasfer of oxygen between the two
phases.
iii. The smailer, more unifonn air bubbles produced by the extemal agitator will aiso
allow for more sdace area for the yeast to adhere to for hydrocarbon consumption.
From the literature review th= has w t been any mention of utilizing an external agitator
in a fermentation process for the proâuction of cihic acid h m n-alkanes. Therefore this
is a novel appmach to a solution for the high oxygen demand in citric acid production
h m hydrocarbons. Shce there has not barn any work in this area, the agitator will have
to be compared to the agitation effectiveness of the fermentor. Several fermentations will
be carried out to detemiine the full potaitid of an extemai agitator.
CHAPTER 3: Selection of Hydrocarbon Feedstock and Yeast Culture
93.1 Introduction
Many Candida species are capable of utilizing hydmcarb0n.s as the priaciple source of
assimilable carbon. Candida lipolytica has been shown to metaboiize hydrocarbon
solvents whiie accumuiaîing citic acid. Concurring with the U.S.
Agriculture, Candida lipolytica NRRL-Y-1094 and Candi& 1ipolMca
have demonstrated successfiil citric acid accumulation (Kurtzman, 1996).
The hydrocarbon forming the principle source of assimilable carbon should consist
essentially of one or more straight chah dkmes or 1-aikenes from 9 to 19 carbons,
partidarly n-alkanes containing h m 12 to 16 carbon atoms or âom 14 to 19 carbon
atoms (Miali and Parker, 1974). From the literature review, as reviewed in Chapter 2, it
was possible to narrow feedstock candidates to kcroscne and n-paranin solvents. The
intaest in kerosene is governeci by its market value which is half that of n-paraffia
(Imperia1 Oil, 1995).
This chapter focws on s d g ciûic acid production with two publicly availabie
strains of Candi& li@ytica and two industridy available h y c h a b n iecdsfacks. The
selected yeast strain and hydrocarbon fesdstock was then uscd in the subst~ucnt work on
the production of citric acid h m hydrocarbons, Chapters 4md 5.
83.2 Materiais and Methods
3.2.1 Yeast Culture
The yeast cultuns used in this part of the sîudy are Candida lipolyticu NRRL-Y1094 and
Candido IipoIyntico NRRL-Y1095 desctibed in Appendix A, Section A. 1. The lyophilized
yeast cells were propagated as d d b e d in Appendix A, Section A.2.
3.2.2 Hydroeubon Feedstock
The hydrocarbon f d t o c k s u s d in this part of the study are Ken, 1-K and Norpar 15
dcscribed in A p p d i x A, Section A.3.
3.2.3 Yeast CeU Growth and Fermentation Media
The media describecl in Tables A2, A3 and A4 in Appeadix A, Section A.4 were used for
the inoculation yeast seed culture, the pre-fermentation stage for yeast biomass growth
and the fmentation stage for ciûic acid production, respectively.
3.2.4 Experlwntal Protocols
R2.4.l Prepurution of Yeapt Seed Culture
The yeast seed culture was pfcpared by dispeiising 645 mL quantitics of the medium
desCnbed in Table A2 (Appaidix A) into two 2.0 îiire Erlenmeyer flash. The medium in
each flask was iwculated with one of the two ymst cultures msintained on agar plates;
one flask with Candido Iipoiyticu-Y-1094 and the oîher with Candi& l@ofpca-Y-1095.
These flaslrs wae fittd with foam bugs, to mhimk contamhtion of media, and then
incubated at 25 k 1°C for 48 hours in a Lab-Line Instruments 3597 Environ-Shaker rotary
shaker nimùng at 90 rpm. .-
3.2.4.2 Preparution of Pre-Fementation Stage-Bomass Growth
At the end of the seed culture propagation, the seed cultures weie used as inoculation
cultures for both Candida culture fermentations. A 1725 mL volume of pn-fermentation
stage medium described in Table A3 (Appendix A) was dispensai into each of four 4.0
litre Erlenmeyer fiasks. The four flasks w m inocuiated under aseptic conditions with
different yeast culture and hydrocarbon solvent combinations, as described in Table 1,
and then fitted with foam bungs.
Table 1: Yeast Culture and Hydrocarbon Feedstock Composition in Re-Fermentation Stage Biomass Growth Flasks
Bottle Number Yeast Culture Hydmcarbon Feedstock 1 Candida lipolytica- Y-1094 Keroscne 2 Candida l@olytica-Y4 094 n-Paraisliin 3 Candida lipolytica- Y-1095 Kerosene 4 Candida lipolytica- Y4 095 n-Parafi
The pre-fmentation stage flasks were then incubated at 25 f 1°C in the same Lab-Line
shaker at 50 rpm for 70 hours while being aerated by sparging with filtercd air at a rate of
1 .O Litres pa minute (0.5 8 volume air/volume medium/minute [ml).
3.2.4.3 heparation of Fermentation Stage-Cihic Acid Production
The chic acid producing stage of the fermentation was startcd by dispc~lsing 1875 mL of
the fermentation media deScnbed in Table A4 (Appcndix A) into each of four 4.0 litre
Erlemeyer fiasks. Each flask contained the same yeast culture and hydrocarbon
feedstock combination descnbed in Table 1; the yeast culture inoculations were h m the
respective pre-fermentation growth. The flasks were fitted with foam bugs and then
incubated at 25 f lac for 120 hours in the Lab-Line shaker nmning at 100 rpm. Using
filtercd air, each flask was aerated at a rate of 2.5 litres pa minute (1.33 wm). Samples
were taken daily for analysis of biomass concentration, hydrocarbon feedstock
co~lsumption and citric acid pmduction.
3.2.5 Anaiyticd Methods
The analyticai methods used are described in Appendix A, Section AS. The biomass
concentration was andyzed using the method outlined in Section A.5.1. The
hydrocarbon concentrations were analyzed usiug the method outlincd in Section A.5.2.
The citric acid concentrations were analyzed using the method outlined in Section
A.5.3.1.
933 Results and Discussion
AU aw data for the foiiowing expiments can be f m d in Appendix B.
3.3.1 Effcet of Hydrourbon Fcedstoek on Citric Acid Production
3.3.1.1 4 Litre Fermentation wirh Kerosene FeedstocR
Figures 5 and 6 describe cihic acid fcrmcntations utilizuig kcrosene as a f d o c k ova a
120 hour famatation pcriod. Figure 5 illustrates the use of Chdi& lipI'*ca-Y-1094
with the kerosene substrate. It was obsaved that at the completion of the femientation
14.5 mgL of ciûic acid was produceci with an overall productivity of 0.12 mfl-h.
Kerosene co~lsumption was 18%. The biomass concentration increased 4.5 folds fiom
initial concentration.
Figure 5: Citric Acid Fermentation using Kerosme and Candida lipoljticu-Y-1094
In Figure 6, w h m Candida lipolytica-Y4095 was used, citic acid production was 20
mg/L with an overaîi productivity of 0.17 mg/L-h. Kerosene consumption was 20 %.
Biomass concentration resuited in a 5 fold increase initial concentration.
In cornparhg Figures 5 and 6, Condido lipolyèu-Y-1095 seems to assimilate the
kerosene feedstock more effcaively than Candido IipoINca-Y-1094, with a 40 % higher
citric acid yield.
Figure 6: Citric Acid Fermentation using Kerosaie and Candida lipolytica-Y-1095
However, the ha1 biomass concentration is 20 % higher with Candida lipoljtica-Y-1094
than Candida lipolytica-Y-1095 using the kaosene substrate. However the production
trends portrayed in both fermentations are similar with only slight shifts in &ta Both
yeast cultures go tbrough a 20 hour lag phase in citric acid production.
3.3.1.2 4 Litre Fennentorion with n-Parafin Feedrack
Figures 7 and 8 illustrate chic acid production utiliziiig the n-paraffin fttdstock ovcr a
120 hour fermentation pcrioâ. Figure 7 d c s c r i i the fermentation of n-parafnn with
Candida lipo&tica-Y-1094. In this fcnnentaîion the citric acid yield was 2.79 gR. with an
ovedl productivity of 0.023 g/L-h or 23 mg/L-h. n-Paraffin co~lsumption was 35 %.
Biomass concentration increased by 86 fol& h m the initial concentration.
la, 5
Figure 7: Ciîric Acid Fermentation using n-Paraffin and Candido lipolytica-Y-1094
Figure 8 illustrates the fermentation of n-par& with Candida lipolyticu-Y-1095 over a
120 hour fermentation period. Citric acid production was 5.69 g/L with a productivity of
0.047 k/L-h or 47 mg&-b n-Paraffin coasumption was 43 %. Biomass concentration
increased by a 171 fold of initial concentration.
Comparing Figures 7 and 8, Candida 1ipoIytz'ca-Y-1095 secms to do a bena job at
assimilating n-parafnn to citnc acid than Candi& lipoNca-Y-1094. Candido lipolyrca-
Y-1095 species produceci a 104 % higher citric acid yield than the fcmientation involving
Candida Y-1094. The use of Gzndida lipolytica-Y-1095 also resulted in a 99 % higher
biomass concentration over Candida lipolyticco-Y-1094. Both figures show the same
trend in biomass, citric acid and n - p a d k concentrations.
Figiire 8: Citric Acid Fermentation using n-Paraffin and Conddu lipolytica-Y-1095
There is a similar lag phase with no significant acid production until hour 40-50 when
citric acid production begins.
In summary, it seems that the n-paraffin fadstock rather than the Lerosene is more
effectively assimilated by both Condidu cultures. Higher citric acid and biomass yields
were associated with n-padk solvent. As the kcrosene solvent contains 20% ammatics,
which are more soluble in aqueous liquids than long straight chah n-aikanes, and the
yeast cannot readi1y break these aromatic carbon-carbon bonds, the kaosene may have
been a toxic element to the yeast during growth and citric acid production.
3.3.2 Eff;ect of Yeast Culture on Citric Acid Production
Figures 5 to 8 describe fermentations utilizing Candi& !ipoZyticua-Y-1094 and Candida
1ipoIytzca-Y-1095 under the influence of kerosene and n-parafnn feedstocks.
3.3.2.1 Fermentation with Candida lipoljtica-Y-1094
Figures 5 and 7 show citric acid production using Candida l@oZytica-Y-1094 on kerosene
and n-paraflnn, respectively. Ciûic Acid yield with kaosene was 6.2 x IO4 g citric acidfg
kerosene consumed with a biomass yield of 8.2 x 1W3 g biomasdg kerosene consumcd.
Slightly higher yields were found when using n-parafIin as a feedstock, citric acid yield
was 0.07 g citric acidlg n-paraffin consumai and biomass concentration was 0.1 1 g
biomassfg n-paraffin consumed.
3.3.2.2 Fermentation wiih Candida lipo'a-Y-1095
Figures 6 and 8 describe citric acid production using Candida IipoIyticu-Y-IO95 on
kerosene and n-parafnn, respectively. The kmscne fermentation resulted in a citric acid
yield of 8.3 x IO4 g chic acidlg kerosene consumcd and a biomass yield of 6.7 x 10" g
biomasdg kerosene consumed. The n-parafnn fenncntation yielded higher
concentrations for both citric acid and biomass, with citric acid yield of 0.11 g citric
acid/g n-parafIin consumeâ and biomass yield of 0.19 g biomasdg n-paiaffin c~nsumtd.
In sunmiary, Candida lipOi"ca-Y-1095 assimilated both hydrocarbon fecdstocks more
effectively for citric acid production than Candida li@l@'cu-Y-I 094.
93.4 Conclusions
Based on the results discussed in this chapter, it was found that the yeast of choice was
Candida l@o&tica-Y1095 and the desirad feedstock was the n-paraffin solvent. In
particular, it was found that:
1. Fermenthg n-paraffin solvent with Candida liplyticca-Y-1095 resuited in citric acid
yield of 0.1 1 g citric acidlg n-paraffin consumeci, biomass yield of 0.19 g biomasdg
noparanin consumecl with a citric acid productivity of 47 mg/L-h;
2. Fermenting n-parafnn solvent with Candida lipoiytica-Y494 resulted in citric acid
yield of 0.07 g citric acid/g n-parafb, biomass yield of 0.1 1 g biomasdg n - p m f k
consumed with a citric acid productivity of 23 mfl-h;
3. Fermenting kerosene solvent with Candida lipoiytica-Y-1095 nsulted in a citric acid
yield of 8.3 x 1 O4 g citric acid/g kerosene consumed, biornass yield of 6.7 x lu3 g
biomasdg kerosene consumed with a citric acid productivity of 0.17 m&h;
4. Fermenthg kerosene solvent with Cadido lipol'èa-Y-1094 resuited in a chic acid
yield of 6.2 x IO4 g citic acidfg kaosene consume& biomass yield of 8.2 x 1 (P g
biomasdg kerosene consumeci with a citric acid productivity of 0.12 mgL-h.
CHAPTER 4: Optimization of Biomass Inoculum, n-Paraffin Concentration, Iron Salt Concentration and Temperature for the Production of Citric Acid using Candida l@ulytica Y-1095
Q 4.1 Introduction
The mechanism by which Candida IipoIytica produce citric acid in large amomt under
certain conditions is still a problem of F a t interest. In generai, unlimited p w t h of C.
lipolytica in a rich medium results in minimai citric acid production. However, citric acid
could be pmduced in grat quantity if the gmwth of C. lipoljticu was properly nseicted
(Kapoor et al., 1982; Milsom, 1987) during the citric acid producing phase of the
fermentation cycle.
Growth restriction by carbon, nitrogen or phosphorus for high citric acid accumulation
has been the subject of controversy. Some researchers say rn i~~~~rganisms need a fairly
high initial concentration of carbon feedstock (15-1 8%) to induce citric acid
accumulation in batch fmentations (Kapoor, et al., 1982). Xu et al. (1989) observed the
optimal ciûic acid concentration at a sugar concentration of 10%. Honecker d al. (1989)
rcported thai an initial mgar concentration of 20% led to maiamal citric acid production
by fk cefls. Kubicek and Rohr (1977) have shown citric acid production is possible in
the presence of an excess of nitmgen providai that phosphate is limiting. Kristiaasen and
Sinclair (1978), on the other hand, c o n s i d d that exhaution of nitmgen to be a
nccessary pnquisite for citric acid production in batch fcrmcntation. Kristiansa and
Sinclair (1979) reported that less citic acid was produced during phosphate-limited
steady states than was obtained during the comparable nitrogen-limited states in a
The temperatun of the fermentation process shouid be kept in the range of 26 to 3S°C
(Miall and Parker, 1974). The specific temperature is dependent on the organisni used
and equipment used. An optimal temperature range of * 1°C should be determineci in
flask batch fermentations prior to initiating large scde fmentatiom.
Iron salts are essentid to the production of citric acid as it activates the production of
Acetyl CO-enzyme A, essential for the production of citric acid (Milsom and Meers,
1985). Howeva, an excess of iron wiil activate the production of aconitate hydratase
(aconitase), the enzyme which catalyzes the production of iso-citric a d . Isu-ciüic-acid
is a by-product of the fermentation that is not desired, as a an isomer of citric acid, it
takes away h m potential yield.
Initial biomass iwculum is usually in the range of 1042% volume (Moo Young et al.,
1971; Rane and Sims, 1996; Blair et al., 1991). An excess in the biomass concentration
will lead to high yields in biomass concentrations and lower chic acid concentrations
(Whitworth d Moo Young, 1973), while a deficiency in biomas concentration wili
lead to a long lag in citric acid production.
Most of the previous works related to citric acid fermentation were conductod using "one-
factor-at-a-time" technique. Unfortunately, it frequently fails to locate the region of
optimum response, since the production does not take into account any joint effects of the
factors on the response. An altemative approach is to lay out a "matrk" of di interesthg
combinations of the operating variables. The matxix method has the advantage of
thoroughly explorhg the experimental surface, but it requires a number of measurements
too large to be exploreci in a realistic amount of tirne and resources. The rcsponse surface
method (RSM) is prefcmd, because it can simultaneously consider several factors at
many different levels and comsponding interactions among these factors using a small
number of observations. Furthmore, statistical inferaice techniques can be used to
assess the importance of the individual factors, the appropriateness of their hctional
form, and the sensitivity of the response to each factor (McLean and Bums, 1997).
Recently, RSM has been employed to solve the multivariate problems and optimize
several responses in many types of eqerimentation (Mddox and Richcrt, 1977;
Giovanni, 1983; King and Zall, 1992).
In this study the RSM approach is adopted to locate the co-optimum levels of
temperature, initial hydrocarbon concentration, initial biomass inoculum and iron salt
concentration and to gain an insight of the interactions among these fxtors during the
ciûîc acid fermentation of Candida lijwlj&u l i M 1 0 9 5 on n-pamih.
Q 4.2 Materinl and Methods
4.2.1 Yeast Culture
The yeast culture used in this part of the study was Candida Zipolyrica 1095
described in Appendix A, Section A.1. The lyophilwd yeast was propagated by the
method described in Appendix A, Section A.2.
43.2 Hydroeubon Feedstock
The hydrowbon feedstock used in this part of the study was the N o m 15 solvent
describeû in Appendix A, Section A.3.
4.23 Yeast Ce1 Growth and Fermentation Media
The media describeci in Tables A2, A3 and A4 in Appendix A, Section A.4 were used for
the inoculation yeast seed culture, the pre-fmentation stage for yeast biomass gtowth
and the fermentative stage for cihic acid production, respectively. However, in this part
of the study the hydrocarbon feedstock and yeast seed culture growth inoculation
concentrations in the medium described in Table A3 (Appcndix A) become 77 mLn and
154 mL/L, respectively. For the medium described in Table A4 (Appendix A), the
concentrations of f d c nitrate, hydrocarbon feedsfock and pn-fcnnentation yeast p w t h
inoculation are varied in the central composite design batch f~l~llentations.
4.2.4 ExperiwnW Protoeds
4.2.4.l Preparation of Yemt Seed Culture
A yeast seed culture was p r e p d by dispensing 600 mi of sterilized medium, as
desmibed in Table A2 (Appendix A), into a 2.0 litn Erlenmeyer flask. The medium was
inoculateci aseptically with three loopfûls, using a stdlized inoculating loop, of yeast
culture that was maintaineci on glucose agar plates. The flasL was fitted with foam bungs
to minimize contamination of culture. Incubation o c c d at 26 I 1°C for 48 hours in a
LaboLine Instruments 3597 EnWon-Shaker rotary shaker running at 120 rpm.
4.2.4.2 Preparation of Re-Fermentation Stage - Inoculum Biomass Growth
The 48 hour pre-fermentation stage began by dispensing 2.0 litres of sterilized media,
described in Table A3 (Appendix A), into a 4.0 L Erlenmeyer fiask. Yeast culture
hocdation was by way of aseptically dispensing the yeast seed culture growth h o the
flask. The inocuiated medium was agitated at 26 1°C in a Lab-Line lnstnunents 3597
Environ-Shaker rotary shaker Mming at 250 rpm and acrated with f i l t d air at 1 .O wm.
4.2.4.3 Preparation of Fermentation Stage - Cilric Acid Production
Thirty one batch fmentation bottle tests wen perfonned to satisfy a centrai composite
design. The central wmpositc design investigated four parameters, initial biomass
inuculum, n-paraflh concentration, f h c nitrate concentration, and tempaahin. Tables
2 and 3 describe how each parameta was variai in the ôotile tests. Fermentation began
by diqeming 2 0 mL of sttrifizeâ medium, d c s c r i i in Table A4 (Appendix A) and
adjusted as per each parameter level deScnbed in Tables 2 and 3, into each of the thirty
one 500 mL Erlenmeyer flasks. Yeast culture inoculation was accomplished by
aseptically decanting the desired volume of pre-femientation culture into the yeast-free
fenaentation medium. The batch fermentations were carrieci out at an agitation speed of
200 rpm in a Lab-Line Instruments 3597 EnWon-Shakcr rotary shaker and a New
Brunswick Scientific Environmental Incubation Shaker, and at an aeration rate of 1.0
wm. Fermentations were conducted for the duration of 5 days.
Tabie 2: Parameter Levels of Central Composite Iksign (Coded Values)
Bottle Initial Biomass n - P a d h Ferric Nitrate Tempaature Number hoculuni Concentration Concentration
Table 3: Quantitative Values of the Coded Parameter Levels
Parameta -2 - 1 O +1 +2 Ijnitial .Biomass
r
Inocuium, %volume 5 7.5 10 12.5 15 n-ParaflFin Concentration, % volume 5 7.5 10 12.5 15
Ferric Nitrate Concentration, mg/L 2 6.0 10 14 18
Temperature, OC 22 25 28 31 34
Sarnples were takm for biomass production, n-paraffin consumption and citnc acid
production analysis.
4.23 Analyticai Methods
The auaiytical methods used are describeci in Appendix A, Section AS. The biomass
concentration was analyzed using the method outlined in Section A.5.1. The
hydrocarbon concentrations were analyzed using the method outlined in Section A.5.2.
The citric =id concentrations were analyzed using the mcthod outlined in Section
A.5.3.2.
8 4.3 Results and Discussion
The raw &ta for the following experiments cm be found in Appendix B. Detaiis of the
design can be found in Appcndix C and D. The foilowing discussion focuses on the
centrai composite design and the effeds of initial biomaas inocuium, n-paraffm
concentration, iron nitrate concentration and tempaatun on ciûic acid and biomass
production.
RSM is a sequentid procedm with an initial objective to lead the expexhenter rapidly
and efficiently to the general vicinity of the optimum. Since the location of the optimum
is unknown pnor to nmning the RSM expriment, it is essential to use a design that
provides equal precision of estimation in all direction. In 0 t h words, rotability is a very
important property in the selection of a respome sudace design. Although two-level
factorial experiments d l ouly yield data to fit a limitai model, linear in al1 façtors with
some product ternis, equation [l] (Myers, 1990), it and its functioas are the most cornmon
initiai experiments in the study of response surfaces, because orthoganility of the design
minimizes the variance of the regmion coefficients and any first-order (two-bel)
orthogonal design is rotatahle (McLean and Burns, 1997; Thompson, 1982). The
expeirimental raw &ta is pnsented in Appendix B and the model design in Appeadix C.
The data collected fiom the batch nins were used to develop empincal models desaibing
the experimental system. The models were generated using the method of teast squares.
The technique involves the estimation of model parameters for second order models of
the fonn:
E(Y) is the expectcâ value of the rcsponsc variable Po, pi, arc the mode1 parameters X, and ?$ are the coded factors king studied k is the number of factors bang studied
The development of the empirical models for citric acid production and biornass
production are jmesented in Appendix D i t h sample caiculatious presented in Appendix
F. Table 4 surnmarizes the parameter estimates obtained by least-squares analysis for
citric acid and biornass pmduction.
Table 4: Summary of Empirical Models Developed for Citric Acid and Biomass Production Model Response Panuneter Estimates (Quantitative Values)
Parameter Citric Acid, g/L Biomass, fi @O 8.85 i 0.39 4.61 f 0.11 81 0.99 k 0.2 1 0.46 + 0.06 b 1.33 + 0.2 1 0.48 -+ 0.06 b 0.49 + 0.2 1 0.18 f 0.06 8 4 0.52 f 0.2 1 0.29 k 0.06 Pi 1 -1.03 f 0.20 -0.23 + 0.06 &2 -0.71 f 0.20 -0.2 1 + 0.06 Bn -0.91 k 0.20 -0.10 f 0.06 Pa -1.27 f 0.20 -0.40 f 0.06 a2 - -0.18 + 0.07 - 0.17 f 0.07 R~ 0.900 0.925
F (regrtssion) 19.1 22.89 F.MA .vt (rmession) 2.40 2.35
F flack of fit) 10.29 2 .O F.05.vl .v2 Oll~k of fit) 3.93 3.96
Bl2, a;, pz3, h4, @34 were not considerd in the citric acid mode! since the p-levels
for these parameters were greater than 0.05. While &,, Bl4, b4, b4 were not considered
in the biomass mode1 since the plevels were grrater than 0.05, see Appendix D for
details. Therefore, substituting the wnesponding B parameter values found in Table 4
into equation [I l , the equations for the citric acid and biornass models become:
Citric Acid Prduction Modef:
Biomuss Production Model:
The responses to equations [2] and [3], at various coded levels of temperature, initial
biomass inoculum, initial ngarafnn concentration and fmic nitrate concentration, are
describecl in Figures 9 and 10 for citric acid and biomass production, respectively. Both
Figures 9 and 10 show quantitative values for cihc acid and biomass production, whenas
the biomass inoculum, n-padfh concentration, f d c nitrate concentration and
temperature are tepnsented as their coded values.
Figure 9 indicates that optimal citric acid production occurs at an initial biomass
inoculrnn of 10- 12 % volume, an initial n-par& concentration of 10-15 % volume, a
f-c nitrate concentration of 10 mgL and a temperature of 26-30 OC. The figure shows
how cruciai the iron salt ( f d c nitrate) is to the production of citric acid. Levels as low
as 2 mg/L and as high as 18 mg/L can inhibit the production of cihic acid at a low n-
parafhn level of 5% volume. However a level of 10 mg,, at the same level of n-
parafnn, can produce approximately 3.4 g/L citric acid.
Figure 10 shows that the optimal level of biomass production occurs at an initial biomass
concentration betwecn 12-15 % volume, an initial n - p a r e concentration of 10-15 %
volume, a fmic nitrate concentration of 10-18 mg/L and a temperature of 2630 OC.
Ha, very low concentrations of noparanin and f d c nitrate do not inhibit biornass
production as significantly as in citnc acid production.
n-Paraffin Concentration w- 0 9
n-Paraffin Concentration W C - - - - 0-9
Figures 11 and 12 show normal pmbabiiity plots for both the citnc acid and biomass
production models. The residuais for both models appear to foîîow a normal distribution.
The data, in both whs, follow tht line representing a nomial distribution.
Residuals
Figure 11: Nomal Robability Plot for the Residuals h m the Cikic Acid Production Modtl
-2.5 L---- J -0.5 -0.3 -0.1 0.1 0.3 0.5
Residuals
Figure 12: Normal Probability Plot for the Residuals from the Biomass Production Modei
Figures 13 and 14 show the residual plots versus the predicted values of the response.
These figures indicate a constant variance. The equal scatter of the residual data above
and below the x-axis indicates that the variance was independent of the value of the citric
acid and biomass production.
1 2 3 4 5 6 7 8 9 10
Predided Values
Figure 13: Residual Plot for Citric Acid Roduction Mode1
Predicted Values
Figure 14: Residual Plot for Biomass Production Modcl
8 4.4 Conclusions
From results obtained during this study, it was found that:
Ferric nitrate and initial n-parafnn concentrations had significant effects on the
production of citric acid;
The optimal level of temperature was found to be 26-30°C for both citric acid and
biomass models;
The optimal level of initial biomass inocuium was found to be 10-12 % volume for
citric acid production and 12-1 5 % volume for biornass production;
The optimai level of initial n - p 6 concentration was found to be 10-1 5 % volume
for both models;
The optimal level of f&c nitrate concentration was found to be 10 mgiL for citric
acid production and 1 O- 1 8mg/L for biomass production;
Both the citnc acid and biomass production models have normal distributions and
constant variances.
CHAPTER 5: Citric Acid Production - Investigation of Aeration and Agitation Effects in Batch and Fed-Batch Systems
55.1 Introduction
The fermentation pattern for ciûic acid production demonstrates a biomass growth phase
followed by a citnc acid accumulation phase, the transition being brought about by
mimbial nitmgen limitation (Milsom and Meers, 1985). Whcn yeast ceils find
themselves in 3 nitrogai limited condition and carbon excess environment, they are
forced to store carbon through the Krebs Cycle. The carbon substrate is oxidued to CO2
via the tricarboxylic acid cycle, Figure 2 in Chapta 2, where ciûic acid is accumuiated.
When n-paranins are used as substrate they are converted by P-oxidation to acetyl
coenzyme A which is combined with oxaIoacetic acid to produce citric acid by meam of
citrate synthetase. Many yeasts have the ability to accumulate high concentrations of
citric acid during tricarboxylic acid cycle respiration. Candida lipoiytica is one species
known for m;ÿ9minng citric acid production while assimilating either carbohydrate or
hydrocarbon substrates (Milsum and Meers, 1985; Miali and Parker, 1972).
Men producing cihic acid through n-perafnn f~~~~lcntation, as in ail femicntations,
pnparation of culture iaoculum shoufd not be taken for p t c d . The developmnit of the
inocuium, through various stages, has a definite effet on the subsequent performance of
the process. In some cases poor management of the inoculum can nsult in an aimost
complete loss of productivity. A fermentation process requires biomass which is highiy
efficient and predictable in performance (Stanbury et al., 1995; P a o n and Willis, 1990).
In a laboratory fermentation the inoculum stages rarely exceed two stages, whereas as
many as six may be tequired for certain production processes (Parton and Willis, 1990).
nie n - p a d h batch fermentations studied in this chapter employ a two stage inoculum
gmwth. The preparaîion of the primary inoculum (seed culture) occurs in a shaker h k
whae it provides a biomass growth which is used to staTt the secondary inoculum culture.
The secondary iaoculum @re-fermentation stage) takes place in a fenncntor (same
fermentor vesse1 as in fermentation stage) to provide a fennentor-stimd culture p w n
under controiied parameters, such as medium composition, agitation speed, amtion rate,
temperature and pH. Subsequently the secondary inoculum gmwth is used to start the
fermentation stage, where biomass production is continueci but quickly h i n d d as
nitrogen is soon depleted and citric acid production/accumulation begins.
The scaled-up pre-fermentation stage (scaled-up to Chemap fmnentor) inoculum
preparation saves two purposcs; to supply an iwculum of optimum size and to pravide a
viable inocuium culture with a nproducible and reliable baseline of operoting parametm.
To satis* these aiteria, a fmentor is implementeâ in the cultivation of a sccondary
inoculum as the vesse1 size and control over operathg parameters is significantly grcater
than shaker flask cultivation. Controlling fcnnentor parametas in both pfermentation
and fermentation stages at inoculum transfa assists in nducing the Lag phase in the
fmcntation stage, since culture adaptation under chsnging famentor environment can
be kept to a minimum. Lincoln (1960) states that the lag phase in a fcnnentation is
minimbed by growing the culture in the 'final type' medium. Lincoln's argument is that
the inoculum development medium shouid be sufficientiy similar in composition, pH and
temperature to the production m d u m to minimize any period of adaptation of the culture
to the production medium. Stanbury et al. (1 995) dds that a major diffncnce in pH and
osmotic pressures may cause sudden changes in uptake rates wbich, in tirni, may affkt
viabiliîy .
In hydrocarbon fermentation there are at lest two physical problans of engineering
interest which are crucial for hydrocarbon fe~llentations. The nrst relates to the vay
high oxygen demand p a unit ce11 mass and the other arises h m the Mted solubility of
hydrocarbon oils in aqueous media (Moo-Young et al., 1971). Furtherrnore, Johnson
(1964) and Aiba et aL(1969). have suggested that dissolved hydrocarbon oil may not be
the main source of substrate supply, and that rather uptake occurs by ce11 attachment to
âroplets of oil.
Oxygen is supplied to the culture in the fonn of air sparging through the fmentation
media Staabury et al. (1995) represented the transfer of oxygen h m air to the ce11 as
occurring in a thne steps:
i. The transfe~ of oxygen k m an air bubble into solution.
ü. The transfkr of the dissolved oxygcn through the fmmtation medium to the
mimbiaî cell.
iii. The uptake of the dissolved oxygcn by the ceL
Both aeration and agitation a k c t the extent of dissolved oxygen in the medium.
Aeration rates serve as the initial volume of air bubbles introduced to the system whereas
agitation rates serve to wist oxygen -fer in the following ways (Stanbury et al.,
1995):
i. Agitation increases the area available for oxygm transfer by dispefsing air
in the culture fiuid in the form of mal1 bubbles.
ii. Agitation delays the escape of air bubbles h m the liquid
iii. Agitation prevents coalescnice of air bubbles.
iv. Agitation decreases the thickiess of the liquid film at the gar-liquid
interface by creating turbulence in the culture fluid.
Agitation in hydrocarbon fermentations is also crucial to the dispersion of oil droplets
throughout the medium. Moo-Young et al. (1971) found that mnaller oil droplets, the
result of higher agitation rates, resulted in greater assimilation of the hydrocarbon source
by the yeast. However agitation rates nmning too high can impose an dangerously large
shear stress on ce11 walls, ultimately decrcasing or destroying ce11 viability.
The purpose of this study is to detmnine the effects of aeration and agitation on ce11
p w t h and citric acid production. These parameters wae expnimeatally evaluated in a
batch and fed-baich system utiluing n-peranin as the hydrocarbon ftcdsfock and
Candido lipiNca MZRL-Y-1095 as the organism.. The optimal opaPing levels
detemincd in Chapte 4 are implcmcnted in these fmentations.
$5.2 Materials and Methods
5.2J Yeast Culture .-
The yeast culture used in this part of the sîudy was Candi& Zipolyticu MU1L-Y1095
described in Appendix A, Section A.1. The lyophilized yeast was propagated by the
methcd desCnbed in Appendix A, Section A.2.
52.2 Feedstock
The hydrocarbon feedstock used in this part of the study was the Norpar 15 solvent
described in Appendix A, Section A.3.
5.2,3 Yeast CeU Growth and Fermentation Media
5.2.3.1 Batch Systems
The media described in Tables A2, A3 and A4 in Appendix A, Section A.4, wen used
for the inoculation yeast sead culture, the pre-fermentation stage for yeast biomass
growth and the fermenkitive stage for citric acid production, respcctively. However, in
this part of the study the hydracarbon feedstock and yeast seed culture p w t h
inoculation concentrations in Table A3 (Appendk) become 77 mUL and 154 mUL,
respectively. For the medium d&M in Table A4 (Appendix A), the concentrations of
f d c nitrate, hydrocafbon feedstock and pn-femientation yeast growtb inoculation
become 10 mgL, 143 mUL and 143 mL/L rcspccitvely.
5.2.3.2 Fed-Batch Systems
The propagation of a yeast seed culture, serWlg as the inoculation culture, was conducted
using the de- medium described in Table A2 (Appendix A). The media used for the
pre-fermentation stage biomass growth and fermentation stage citric acid production are
d d b e d in Tables A3 and A4, respectively. The medium used for the fcading of the
fermentation stage in the 7 day fed-batch systems is described in Table A5 in Appendix
A, Section A.4. The medium used for the feeàing of the fermentation stage in the 18 &y
fed-batch systems is described in Table A6 in Appendix A, Section A.4. This medium is
similar to that described in Table A5, however a maii amount of nimgen is added to
meâim to help maintain yeast ce11 viability throughout the fcnnentation.
5.2.4 ExperimenW Protoeols
5.Z.4.1 Prepatation of Yeart Seed Culture for Batch and Fed-Botch Systems
A ycast seed culture was preparad by dispc~lsing 600 mi of sterilized medium, as
describeci in Table A2, into a 2.0 litn Erlenmeyer flask. The medium was inoculated
asepticaîiy with three loopfuls, using a sterilized inoculahg loop, of yeast culture that
was maintained on glucose agar plates. The fia& was firted with foam bungs to
minimize contamination of culture. Incubation occurred at 26 f 1°C for 48 hom in a
Lab-Line Instruments 3597 Environ-Shaker rotary shaker runnîng at 120 rpm.
5.2.4.2 Preparation of Pre-Fennntation Stage fur Batch and Fed-Boch Systenr- Biomass Growth
The 48 hour pre-fermentation stage began by dispensing 3.9 Litres of sterilued medium,
describeà in TableW, into a 14 L Chemap AG Fermentor (24.7 x 31.1 cm cylindrical
glas vessel). Yeast culture inoculation was by way of asepticdy dispensing the yeast
seed culture growth into the fermentor. The inoculateci medium was agitated at 28 f 1 OC
with a standard four bladed (3.8 x 0.2 x 18.4 cm staidess steel impe11er) open turbine at
various speds (400, 800, 1000. 1200 rpm) and aerateû with filtmd air at various rates
(OS, 1.0,2.0 wm).
5.2.4.3 Preparation of Fennentation Stage for Batch System - Cimè Acid Production
Fennentation began by dispensing 7 litres of sterilized medium, as descrlbed h Table A3,
into the 14 L Chemap Fennentor. Yeast culture inoculation was accomplished by
aseptically draining 3/4 of the pre-fermentation culture fmm the fenntlltor and ldding
yeast-fixe fermentation medium. The fmentation was carried out at 28 i 1°C for 5
days. Agitation, with a four bladed open turbine, and aeration, with fiItered air, were
applied at various rates, see section 5.2.4.2.
5.2.4.4 Prepurutiun of Fmentation Stage for Fed-Batch System - Cilric Acid Production
Fermentation began by dispashg 4 litres of sterilized mtdium, as d t s c n i in Table A3,
into the 14 L Chanap Fermentor. Yeast culture hocdation was accumpiished by
aseptically draining 314 of the pre-fcnacntation culture h m the f m o r and aâding
yeast-ke fmentation medium described in Table A4. The fermentation was d e d out
at 28 f 1°C for 5 days. Feeding of the fed-batch f~~llentation with feading medium,
described in Tables A5 and A6 (Appendix A), was done using a Cole-Pamer paistaltic
pump, where the final volume in the fmentor was taken to 7 litres. The 7 &y fed-batch
system was fed at the 96 hou mark with medium d e s c r i i in Table M. The 18 &y fad-
batch system was fad at the 96,216,336 hour marks with the medium desaibed in Table
A6. The feading rates varied as 0.75, 1 .O and 1.5 Uday.
Agitation, with a four blade open turbine, was supplied at various rates, 400, 800, 1000,
1200 rpm and aeration supplied with filtered air at 1 .O wm. When externai agitation was
implemented into the design, a Greey Lighbiin In-Lhe Agitator was used to agitate the
broth being n-cuculated fiom the fermentor, into the agitator, then back into the
fnmentor, see Figrues 3 and 4 in Chapter 2. Extemal agitation was investigated at 1400
and 1600 Ipm at a ncirculation rate of 4.0 L/&. Samples wen taken evay 12 hours
for biomass production, n-parafk consumption and citnc acid production analysis.
5.2.5 Anaiytical Methads for Batch and Fed-Batcb Systems
The analytical methods used an described in Appendix A, Section A 5 The biomass
concentration was analyzcd using the method outfincd in Section A.5.1. The
hydrocatbon concentrations wcre analyzed using the mahod outIined in Section A.5.2.
The citric acid concentrations wae d y z e d Usmg the methd outlinai in Section
A.5.3.2.
55.3 Results and Discussion
Details of the experiments and associaîed raw data cm be found in Appendix A and B,
respectively. All fermentations involved a " p w t h phase" for biomass p w t h and
"production phase" for chic acid praduction (as nitmgen is depleted the yeast go into
citric acid production). The following discussion focuses on aeration and agitation
innuences on ôoth phases in batch and fed-batch n-parafnn fermentations. This chapter
also demonstrates the effécts of agitation, nsulting h m mechanical agitation h m the
fameator and in-line agitator, on biomass and citric acid production in fed-batch systans.
It should be notexi that 'veld" refemd to in the discussion is the amount of citric acid
produceci per n-paraffin consumed at a certain point in time in the fermentation. However
"overall yieiâ" referred to in the discussion nfers to the total amount of citnc acid
produced pa the total n-parafb consumed at the end of the fermentation.
5.3.1 Effkct of Aeration on Biomass and Cltric Acid Production - Batch Systems
Thm different aeration rates were investigated, 0.5, 1 .O and 2.0 wm while maintainhg
an agitation rate of 400 rpm. Figuns 15 and 16 describe the production of biomass and
citric acid under the influence of the various aeration rates. Table 5 summarizes ovcrall
biomass and citric acid yields and productivity as @&cd by acration. Rtvicwing Figure
lS(a), in the fkst 48 hours (growth phase) aeration of 1.0 wm produccd the highest
biomass concentration, 8 g/L. While the subsequcnt 120 hour chic acid praduction
phase results in the 0.5 wm acration producing the highest biomars concentration of 10.9
fi. Considering biomass yields, Figure 15(b) shows that in the growth phase an aclitfion
of 0.5 w m resulted in highest biomass production with a biomass yield of 0.5 g
biomasdg n-parafh-consumed. In the citrie acid production phase, the aezation of 0.5
wm generated the highest production with a biomass yield of 0.5 g biomasdg n -padh-
consumed about 25 hours into the production phase, tben dropping to 0.25 g biomasdg n-
parafan-comumed by the end of the phase @OUT 168). While overaii biomass yield,
shown in Table 5, supports that an aeration of 1.0 w m generated the highest biomass
production with cm o v d l biomass yield value of 0.23 g biornasdg n-pafaffin-consumed.
The 2.0 wm fermentaticn has the lowest production of biomas in both growth and
production phase. A possible explmation is that if the impelier is imable to disperse the
incoming air, then low oxygen transfer rates are achieved due to the impella becoming
'flooded' (Stanbury et al., 1995). Stanbury et al. describe flooduig as the phewmenon
where the air-flow dominates the flow pattern and is due to an inappropriate combination
of air flow rate and speed of agitation. Even though the dissolved oxygen probe rexorded
dissolved oxygen values in the range of 74-78%, the probe rnay have been in the line of
the dominating air flow and not recording representative dissolved oxygen values of the
dispersed medium.
Moo-Young et al. (1971) rcporied biomass production in the range of 4.5 to 5.0 g/L under
an acration of 1.0 wm and agitation of 500 rpm duRag a 48 hour hydrocarbon
fermentation. Biomass production (Figure lS(a)) in tbis study, at 1.0 wm and 400 rpm,
shows a slightly higher biomass concentration of 8 JJ/L 48 hours into the production
phase of the fcrmcntation. Aitexnativcly Furukawa et al. (1977) reportcd a biomass
concentration of 12.1 g/L, at 0.5 wm and 600 xpm, during a 93 hour hydracarbon
fet~llentation, while Figure 1 S(a) of this sbidy shows a Iowa biomass production of 10.9
g L , at 0.5 w m and 400 rpm 93 hours uito the production phase. Aithough the cumnt
study on biomass concentrations during fermentation differ slightly fiom previous studies
the growth profiies are consistent. The production of biomass enters an exponential
phase, after a short lag phase, pnor to reaching a stabilization in growth.
The cumnt &ta observations of an increase in biomass production during the
fermentation stage is consistent with Knstiaiwn et al. (1978) findiags. Knstiansen et al.
found that even under nitmgen limitecl environments (imücated as production phase in
this discussion) biomass production continueci. He attributed this growth to carbon
storing cells and concluded that cells under excess carbon not oniy store carbon for citric
acid production but also for biomass production.
Figure 16(a) shows the production of citric acid during the fermentation stage. No citric
acid production occurs in the p w t h phase. During the initiai 92 hours of the production
phase the 1 .O wm aeratcd fermentation resuits in the highest citric acid production, 4.25
fi. However in the final 28 hours the 2.0 wm aeratcd fmentation surpasses this to a
fimi citric acid production of 4.8 g,L. Considering citric acid yields, Figure 16(b), the
highest chic acid yieid of 0.23 g citric acid/g n-paraffin-cotlsullcd is producecl by the 1 .O
w m aerated fcnncntation during the initiai 48 hours of the production phase, while it
ends the fmentation with a lower pduction yield of 0.15 g ciûic acid/g n-parafan-
consurnad. Incrcasing acration fkom 0.5 to 1.0 wrn d t c â in incmsiag citric acid
production. The hîgher acration rate i n d the rate of oxygcn transfcf hughout the
medium and therefore increased the pduction of citric acid. The possible 'fîoodhg' of
the impeller at an d o n of 2.0 wm affected citric acid production resulbg in the
lowest citric acid yield of approximately 0.1 g citric acid/g n-paraflh consumed.
Citcic acid concentrations found in this study are lower than some published values.
Funikawa et al. (1977) reported citric acid production of 65 g/L while ushg a mutant
strain of the pannt Candida Iipoiytica in a n-paraffin fermentation for 93 hours at 0.5
wm and 600 rpm. Akiyama et al. (1973) shows cieic acid production to be 60 gL using
the parent Candida 1Qoiytica in a n-paraffin fermentation for 72 hours in a 200 ml shake
flask at 200 rpm on a rotary shaker. W h m a ~ in Uiis saidy the highest cihic acid
concentration achieved was 4.8 g/L in a 120 hour production phase fmentation. W l e
the magnitude of ci& acid production is different fiam the published values, the
production profile is still consistent. The lag phase is virtuaily non existent, and the
exponential phase is sharp followed by a slow stabilization of production.
Al1 the fermentations experience peaking of ciaic acid yields in the fmt 24 to 48 houn of
the production phase. Table 5 shows the highest overall citric acid yield at an aeration
rate of 1 .O wm with 0.13 g cihic acidlg n-padfh-consumed and an average pductivity
of 0.04 g/L-h. Theoretically, a peaking citric acid yield shouid be mauitained for more
than a 12-24 hout period and not expericnce such a drastic decline in production yields,
as is seai in ail the fermentations of the cumnt study.
a) Biomass Production
b) Biomass Yield
Figure 15: Aeration Effects on Biomass Production in Batch n-Parafb Fermentation at Soorpm and 28 t 1°C
a) Citric Acid Roduction 5
I
b) Citrfc Acid Yidd
Figure 16: Acration Effects on Citric Acid Roduction in Batch n-Parafnn Fmcntation at 400 rpm and 28 f 1°C
Table 5: Sirmmary of Aetation Effects on Biornass and Citnc Acid Production in Batch Fermentation (120 hour production phase)
Parameter 0.5 vvm* 1.0 vvm* 2.0 vvm* Overall Biomass Yield, g biomasdg n-parafiin consumed 0.2 1 0.23 0.14
Overall Citric Acid Yield, g ciûic acid/g n-paraffin consumed 0.10 0.13 0.1 1
û v d Cihic Acid Productivity, g1L-h 0.03 0.04 0.04 * Ail ferm~~ltatiom under agication rate of 400 rpm
5.3.2 Effect of Agitation on Biomass and CiMc Acid Production - Batch Systems
Four agitation rates were investigated, 400, 800, 1000 and 1200 rpm. Figures 17 and 18
describe biomass and citric acid production under the influence of these agitation rates,
respectively, while maintainhg aeration rates of 1.0 wm. Table 6 Summarizes overail
biomass and citric acid yields dong with citric acid productivity. Figure 17(a) shows an
increase in biomass production with agitation, in the growth phase. Biomass production
in the 1000 and 1200 rpm femientations is highest at 8 g/L at the end of the growth phase.
In the initial 48 hours of the production phase the agitation rate is not effecting the
production of biomass, while in the final 20 hours of production phase the highly agitatcd
fe~~~lentations, 1000 rpm and 1200 rpm, produce the highest biomass concentrations of
14.8 and 14.6 g/L, respectively. Considering biomass yield, ail the f ~ t a t i o n s have
decrtasing yields during the growth phase. The 800 and 1ûûû rpm fermentation produces
the highest biomass yield of 0.57 g b i o m d g n-pard'bconsumed in the pwth phase;
AU fermentations eXpenence a biomass field peelr in the initial 12 to 48 hom of the
production phase with a subsequently decreasing biomass yield. In the production phase,
the 800 rpm agitated fermentation produces the highest peaking biomass yield of 0.5 g
biomasdg n-pa,-consurned. The overall biomass yields, Table 19, show consistent
values between 0.22-0.27 g biomasdg n-para-consumed. The observed increasing
biornass production with agitation can be attribut4 to the fact that as agitation increases
two things occur, oxygen ûansfer from air bubbles into solution incnases as irnpellers
efficiently disperse incoming air and n-paraffin oil droplets are smaller, p a t e r in numba
and more adequately disperseci throughout the medium. Hence the higher agitation rates
allow for higher transfer rates of oxygen and n-parafb carbon into the cells for biomass
production.
Moo-Young et al. (1971) reported biomass production of 4.5 g/L 48 hours into two n-
paraf'fïn fermentation at 500 rprn and 1 .O wm and 800 rpm and 1 .O wm. However, in the
cumnt study biomass production 48 hours into the growth phase of each agitated
fermentation is significantly hi-; the 400, 1000 and 1200 rpm apitatcd fenacntations
demonstrate a biomass concentration of 8.0 g/L, while the 800 rprn agitatcd fcrmcntaîion
demonstrates a biomw concentration of 7 g/L. Each fermentation in the study goes on
to produce even higher biomass concentrations in the production phase, with the 1OOO
and 1200 rprn conditions having the highest biomass production in the range of 14.644.8
fi. Al1 four fermentation nuis seem to be going through a lag in biomass production
during the initial 48 hours of the production phase.
Figure 18(a) shows citric acid concentrations increasing with agitation. The 1200 rpm
agitated fermentation produces the highest citric acid concentration of 9.5 glL.
Considering citric acid yield, Figure 18(b), the 1200 rpm agitated fermentation results in
the highest yield of 0.3 g citric acidg n-paraflïn-consumed. AU fermentations cxperience
peaking citric acid yields in the initial 48 to 72 houn of the production phase, followed
by a decrease in citric acid yield. The highest overall citnc acid yield, Table 6, of 0.2 1 g
citric acidg n-padlin-consumeci is shared by both the 1000 rpm and 1200 rpm agitated
fermentations, respectively. However the 1000 rpm fermentation has the highest overall
prod~ctivity of 0.07 &oh.
Citric acid production is highest at higher agitation levels. As discussed in biomass
production high agitation results in gmiter oxygen and carbon transfer into the yeast
cells, hmce increasing the rate of respiration (citric acid production). Again cieic acid
concentrations are lower than published values. Both Funikawa et aL(1977) and
Akiyama et aL(1973) rcport citric acid concentrations of 65@ (93 hours at 0.5 wm and
600 rpm) and 60 gL (72 hours in a 200 mL flask at 200 rpm on a rotary shaker),
respectively. While the highest citric acid concentration achieved in this study was 9.5
g/L in a 120 hour production phase fermentation. As discussed in section 5.3.1 the citric
acid production profile is consistent with published work, howeva the magnitudes of
production differ. The lag phase is Wtually non existent, and the cxponential phase is
sharp followed by a slow stabiluation of production.
O 50 100 150 m i
lime, hour I
Figure 17: Agitation Effects on Biomass Roduction in Batch n-Paratfin Fermentation at 1.0wmand28f toc
Figure 18: Agitation Effects on Citric Acid Production in Batch a-Parafin Fermentation at 1.0 wmand28 f 1°C
Table 6: Summary of Agitation Effects on Biomass and Citric Acid Production in Batch n-Para* Fermentation (1 20 hour production phase)
Parameter 4ûûrpm* 800rpm* lûûûrpm* lZOOrpm* O v d Biomass Yield, g biomasdg n-paranin consumed 0.23 0.22 0.27 0.23 Overall Chic Acid Yield, g citric acid/g n-paranin consumed 0.13 0.19 0.21 0.2 1 û v d l Citric Acid Productivity, a& 0.04 0.06 0.07 0.05 * AU fermentations under aeration rate of 1 .O wm
5.3.3 Effect of Agitatioo on Cihic Acid Production - Fed-Batch Systems
Four agitation rates were investigated in this study, 400, 800, 1000, 1200 rpm while
maintaining an aeration rate of 1.0 m.. Figures 19 and 20 describe biomass and citric
acid production under these agitation rates, respectively. Table 7 summarizes o v d l
biomass and citric acid yields dong with citric acid productivity. Figure 19(a) shows
biomass production to be fairly consistent with changing agitation rates, with biomass
concentrations in the range of 10.5-1 1.5 g/L suspended soli&. Considering biomass
yields, Figure 19(b), ail fermentations experience a peak in biomass yield in the p w t h
phase, where the 400 and 800 rpm agitated fermentations experience the highest biomass
yield of about 0.625 g biomasdg n-paraffin consumeâ. Wbile the 1000 and 1200 rpm
agitated fermentations have the lowest biomass yield of 0.3 g biomass/g n-paiafnn
consumed in the growth phase. In the production phase, al1 fcnnmtations experience
increasing biomass yields with the 400 and 800 rpm agitatcd fcrmentations being highest
with a value of 0.45 g biomdg n-paraffin consumcd, and the 1ûûû and 1200 rpm
agitated femiatations being lowest with a value of 0.30 g biornasslg n-parafnn
consumed. The highest overd1 biomass yield was 0.45 g b iomdg n-paraffin consumed
by the 400 rprn agitated fermentation. The observeci increasing biomass production with
agitation can be amibuted to the fact that as agitation increases two things happa
oxygen transfer fiom air bubbles into solution increases as impellas efficiently disperse
incoming air and n-parafi oil droplets are smaller, greater in number and more
adequately dispersed throughout the medium. Hence the higher agitation rates allow for
higher transfer rates of oxygen and n-paraffin carbon into the cells for biomass
production. This theory has been consistently supported with the expaimental data h m
the fermentation studies conducted.
Moo-Young et ai. (1971) reported biomass production of 4.5 g/L 48 hours into two n-
parafin fermentation at 500 rprn and 1 .O wm and 800 rprn and 1 .O wm. However, in the
cment study biomass production 48 hours Uito the production phase of each agitated
fermentation is significantly higher, with the 400, 800 and 1 0 0 rpm agitated
fmentations having similar concentrations h m 7.0 to 8.0 g/L, whiie that of the 1200
rprn agitated fermentation is at 5 g/L. Both this fed-batch fermentation and the prcvious
batch femenbtions produced higher biomass concentrations than what Moo-Young
reported. Each fermentation in the study goes on to produce even higha biomass
concentrations, in fact each fermentation has a biomass concentration that stabilizes in the
range of 10.5 to 1 1.5 g 5 . The 1 200 rprn fermentation seans to be going h u g h a lag in
biomass production during the initial 48 hours of the production phase. This could be
atnniuted to the yeast ce11 adapting to its surr0~11dings of higher &car sûesscs. If this is
the case, the lag in biomass production would also be cxpectcd in the 1000 rpn
fmentation, however no apparent lag in biomass production is evident in the initial
phase of the production phase.
Figure 20(a) shows citric acid concentrations increasing with agitation witb the lûûû rprn
agitated fennmtation having the highest final citric acid concentration of 11 g/L and the
400 rpm agitated fmentation having the lowest fM citric acid concentration of 5 g/L.
Considering citric acid yields, Figure 20(b), the 800 rpm agitated fermentation has the
highest yield of 0.4 g citric acid/g n - p d n consumed. The highest o v d citnc acid
yield, Table 7, of 0.41 g chic acid/g n-parafün consumed belongs to the 800 rpm agitated
fermentation. However, the 1Oûû rpm agitated fermentation had the highest citric acid
productivity of 0.09 g/L-h.
The citric acid concentrations found in this study are lower than some published values.
Furukawa et d.(1977) and Akiyama et aL(1973) report cihic acid concentrations of 65gL
(93 hours at 0.5 wm and 600 rpm) and 60 g/L, (72 hours iB a 200 ml flask at 200 rprn on
a rotary shaker), respectively. While the highest citric acid concentration achieved in this
study was 12.0 g/L in a 120 hour fmentation. While the magnitude of citric acid
production is not consistent with published values, the production profile was consistent.
The lag phase in ciûic acid production is vimially non existent, and the exponential phase
is sharp followed by a slow plateauing of production. Another significant trend obsmed
is the sharp inmase in citric acid production afta the 96 hour feeding of 1.0 Uday. The
800 and 1 0 rprn agitated fermentations expdcnce a 75 % inc11c8sc in citric acid yield
24 hours into the fed-batch feeding. This feeding significantly shifts the yeast h m its
stationary production phase, prior to the 96 hour feeding, to an exponential production
phase following feeding.
Table 7: Summary of Agitation Effects on Biomass and Citxic Acid naduction in Fed- Batch Fermentation
P arameter 400 rpm* 800 rpm8 lOOOrpm* 1200 rpm* Overd1 Biomass Yield, g biomasdg n-parafnn consumed 0.45 0.43 0.3 1 0.32
Overall Citric Acid Yield, g citric acid/g n-pa- consumed 0.22 0.41 0.37 0.32
O v d l Citric Acid Productivity, a - h 0.04 0.07 0.09 0.07 * Al1 femcntations under aeration rate of 1 .O wm
a) Biomass Production
-400 rpm 8 0 0 rpm 4 1000 rpm +1200 rpm
b) Biomass Yield
Figure 19: Agitation E f f ~ on Biomass Production in Fed-Batch noparaffin Fermentation at 1 .O wm and 28 1°C
a) Citric Acid Production
) I \
Fcadng at 1 .O Uday
b) Citric Acid Yield
Figure 20: Agitation Effects on Citric Acid Production in Fed-Batch n-Paraftin Fermentation at 1 .O wm and 28 5 I°C
53.4 Effcet of Grey Lightnin In-Line Agitator on Biomw and Citric Acid Production - Fd-Batcb Systems
.- Two different external agitation rates were investigated, 1400 and 1600 rpm. While the
extemal agitator was operating at these speeds the fmentor agitation was maintained at
400 rprn with an aeration of 1.0 wm. Figure 21 and 22 describe the production of
biomass and citric acid under the infiuence of the various external agitation rates,
respectively. Table 8 describes the overall biomass and citric acid yield, dong with ciûic
acid productivity. Biomass production, Figure 21(a), shows that the fermentations with
no extemal agitator and the 1400 rpm agitator have fairly consistent biomass
concentrations of 11.5 g/L. Biomass yields, Figure 21(b), show the femientation with no
extemal agitator has the highest yield of 0.45 g biomasdg n - p a d h consumed. The
highest overalt biomass yield, Table 21, is 0.45 g biornasdg n-parafi belonging to the
fmenbtion with no extemal agitation.
Figure 22(a) shows the production of chic acid to be highest with the fe~llentation with
the external agitator nuining at 1400 rpm, 9.5 fi. Both the fermentations without the
extmial agitator and with the agitator at 1600 rpm have the same citnc acid production of
5 g/L. The highest o v d l citric acid yield is 0.37 g citric acidg n - p a r e consurnad
îxom the fermentation with an extemal agitation of 1400 rprn with the highest
productivity of 0.07 g/L-h belonging to this same fmentation. The fermentation
nmning with extemal agitation of 1600 rprn expeIienced 20-30 % of c d lyses, see
Appendix E, and this is the most likely factor attributing to citric acid fields king as Iow
as the control femiatation (without externa1 agitation).
In this study it was found that the external agitator, operathg at 1400 rpm, contributecl to
an increased citric acid production, a 90 % increase from the control fermentation was
observed. However, operathg the extemal agitator at a speed of 1600 rpm did d t in a
significant degree of ce11 lyses (Appendix E), that resulted in a loss in citnc acid
production. The increase in extemal agitation does follow the principle put forwani by
Moo-Young et al. (1971), that higher levels of agitation result in n-paraff'ul oil droplets
being smaller, greater in number and more adequately disperseâ throughout the medium
allowing for a more efficient contact with yeast cells to adhere to the droplets for carbon
metabolism. While the extemal agitator may in fact have accomplisheà this, the physical
shear on biornass cells resulted in ce11 lyses being the limiting factor in ciûic acid
production.
Again citric acid concentrations are lower than published values nported by Furukawa et
d(1977) and Akiyama et d.(1973), citric acid concentrations of 65g/L (93 hours at 0.5
wm and 600 rpm) and 60 g/L (72 houn in a 200 ml flask at 200 rpm on a rotary shaker),
respectively. While the highest citric acid concentration achieved in this study was 9.5
gR. in a 120 hour fermentation with the 1400 rpm e x t d agitated fmentation. As
discussed in section 5.3.1 the citric acid production profile is consistent with published
work. The lag phase is virtually non existent, and the exponcntial phase is sharp foliowed
by a slow stabilinng of production and &er the 96 hour feoding of 1.0 Uday there is
another expondal phase in citric acid production. However this was prcdombmt in the
fermentation wioi the external agitator nirining at 1400 rpm.
a) Biomass Production
b) Biomass Yield
Figure 21: Agitation Eff- on Biomass Produaion in Fed-Batch n-Paraffin Fermentation using Greey Lightnin in-Line Agitator at 1 .O wm and 28 1 OC
a) Citrie Acid Produetion
10 1
9 . - G j a - . m i
! i 9 7 - - - 1 l
t 1 5 - -
1
:: 1
1
Y 4-• .Y 2 3.- O
2 - - I F-at 1.0 Uùay
1 - -
O - , I
b) Citric AcWl Yield 1
Figure 22: Agitation Effkcts on Citric Acid Roduction in Fed-Batch n-Paraffin Fermentation using Greey Lightnin In-Line Agitator at 1.0 wm and 28 A 1°C
Table 8: Summary of Agitation EEects with and without Greey Lightnin In-Line Agitator on Biomass and Citric Acid Production in Fed-Batch Fermentation
Parameter 400 xpm 400 rpm 4 o o r ~ m (no agiutor)* + 1400 rprn + 1600 rprn
(agitator)* (agitator)* O v d Biomass Yield, g biomasdg n-parafnn consumed 0.45 0.38 0.30
ûverall Citric Acid Yield, g citnc aciâ/g n-par- consumed 0.22 0.37 0.2 1
Ovaall Citric Acid Productivity, ah 0.04 0.07 0.04 * Al1 fermentations under aeration rate of 1 .O wm
5.3.5 Effect of Agitation and 3 Cycle Fed-Batch Fermentation @n Biomw and Citric Acid Yields
From the two previous shidies on batch and fed-batch systems, there was a 1ûû% incnase
in ciûic acid yield moving fiom batch to fed-batch fermentations. Howeva, thexe was no
significant improvement in ciaic acid productivity. With this in mind, the next study
investigates the potential of improving citric acid yield and productivity under a 3-cycle
fed-batch system. The 3cycle system involves feeding the fmmtation at threc diffaent
points in the fermentation. Figure 23 and 24 describe biomass and citric acid production,
respectively, under the infhences of agitation and a 3 cycle fed batch system. Table 9
summarizes the overall biomass and citric acid yields and citric acid productivity. Figure
23(a) describes the highest biomass production to be with the 800, 1ûûû and 1200 rprn
agitated fermentations with biomass concentrations in the range of 16-18 gL. This is re-
iterated in Figure 23@) with biomass yield values in the range of 0.4 to 0.6 g biomasd g
n-paraffin consumed. The highest overall biomass yield of 0.59 g biornasdg n-paranin
consumed occurs with the 1000 rpm agitated fermentation.
Figure 24(a) shows the highest ciûic acid concentration occur for the 800,100, and 1200
rpm agitated fermentations with citric acid concentrations in the range of 30 to 40 g/L.
The lowest citric acid concentration was produceci h m the 400 ~ p m agitated
fermentation, 9 g/L. Figure 24(b) indicates that the highest citric acid yiel& of 0.8 to 1.0
g citric acid/g n-para& consumed resulted from fermentations at 800, 1000, and l2ûû
rpm. The highest overalI citric acid yield (0.99 g citric acidlg n-parafnn consumed) and
highest citric acid productivity (0.1 g/L-h) occmed ail while fermenthg at 1000 rpm.
The most obvious trend is after each feeding cycle of 1 .O Uday, at 96,216 and 336 hours,
there is an exponential increase in citnc acid production. However this sharp increase in
citric acid production is most evident after the 96 and 216 hour feedings while afkr the
336 hour feeding the citric acid production siightly increases but quickly stabiiizes.
Ciûic acid yields are still lower than published values. The 3 cycle fed-batch system did
increase the magnitude of ciûic acid concentration but it did not a f k t the citric Md
productivity of the fermentations. Funikawa et al. (1977) and Akiyama et al. (1973)
reporteci citric acid productivities of 0.7 g/L-h and 0.8 gLh, respectively. The highest
chic acid productivity in this study was 0.1 gL-h by the 1000 rpm agitatcd fermentation.
Again these differences between published and the nportcd data can be attributai to
various factors such as n-para* components not being consistent over the past 20 years
and a different Candida lipolylica strai. andh mutant of strain was used.
iomass Production
Feed 1 Feed 2 Feed 3 I
+400 rpm +a00 prn +1000 rpm rpm +400 rpm + 14ûû rpm ( ô m ) i
1) Biomass Yield Feed t F88d 2 Feed 3
Fire 23: Agitation Effects on Biornass Production in 3 Cycle Fed-Batch n-Pataffin Fermentation with and without Grrey Lightnin ui-Line Agitator at 1 .O w m and 28 * 1 O C
a) Citric Acid Ploduction Feed 1 Feed 2 Feed 3
5) Citric Acid Yield Feed 1 F88d 2 Feed 3
1
Figure 24: Agitation Effkcts on Citric Acid Roduction in 3 Cycle Fed-Batch n-Paraffin Fermentation with and without Greey Lightnin In-Line Agitaor at 1.0 wm and 28 k 1°C
Table 9: Summary of Agitation Effects with and without Greey Lightnin In-Line Agitator on Biomass and Citric Acid Production in 3 Cycle Fed-Batch Fermentation
ipm* (agitator)* rpm* qm* rpm* Overall Biomass Y ield, g biomasdg n-parafi consumed 0.30 0.36 0.39 0.59 0.44
O v d l Citric Acid Y ield, g citric acid/g n-para- consumed 0.22 0.49 0.76 0.99 0.88
ûverall Productivity, %La 0.02 0.05 0.08 0.10 0.09 * AU fermentations under aeration of 1 .O vvm
The citric acid yields in the 3sycle fed-batch system where 100% higher than the yields
obtained fiom the single cycle fed-batch system and where 200% highei Uian the yitlàs
obtained fiom the batch system. Appendix E illustrates the ce11 viability during the 3-
cycle fermentation.
95.4 Conclusions
Batch Systems
Based on the results discussed in this chapter, it was found that:
1. An aeration rate increase fkom 0.5 to 2.0 vvm resdted in a 46 % decrease in
overall biomass yield, howeva overall citric acid yield and pductivity was not
affec ted;
2. An agitation rate increase frorn 400 to 800 rpm resuîted in a 58% i n m e in ,
overall citric acid yield and a 39% increase in citric acid productivity, howevex
overall biomass yield was not affected;
3. Optimum citric acid production, under the iduence of various aeration rates, was
4.8 g/L at 2.0 w m and 400 rpm;
4. Optimum citric acid production, under the influence of various agitation rates,
was 9.5 g/L at 1200 rprn and 1.0 wm;
5. Biomass continued to be produceci in the nitrogen limited femiatation stage;
6. Citnc acid yields peaked in the first 12 to 24 hours of the fermentation stage with
subsequently decreasing yields.
Fed-Batc h Systems
Based on the nsults discussed in this chapter, it was found that:
1. An agitation rate increase fiom 400 to 1200 rprn resdted in a biomass yield decrease
of 0.1 5 g biomasdg n-paraffin;
2. An agitation rate increase fiom 400 to 1000 rprn resulted in a ciûic acid ykld incrcase
of 0.21 g citric acid/g n-paraffin;
3. The addition of an extemal agitator (Greey Lightnin In-Line Agitator) nmning at
1400rpm decreased the biomass yield by 0.07 g biomasdg n-parafiin ficm the control
fermentation of no extemal agitation;
4. The addition of an extemal agitator (Greey Lightnin In-LUie Agitator) nmning at
1400 rprn increased the citric acid yield by 0.15 g citric acidlg n-paraf@
The addition of an extemal agitator (Greey Lightnin In-Line Agitaîor) nuining at
1600 rpm experienced 20-30 % of cell lyses;
An agitation rate increase of 400 to 1000 rpm resulted in a biomass yield bcnare of
0.29 g biomasdg n-paraffin;
An agitation rate increase of 400 to 1 O00 rpm resuited in a citric acid yield increase of
0.77 g citric acid.g n-paraffin;
The 3 cycle fed-batch fermentations increased citric acid yields by 0.56 g cihic acidlg
n-paraffin, but the citric acid productivity was not affected;
Following fed-batch feedings the citric acid production sbarply increased; taking
production fiom a stationary phase to an exponential phase;
10. The 3-cyle fed-batch system increased overall citnc acid yields by 100% h m those
found in the single cycle fed-batch system and increased overali citric acid yields by
200% fiom those found in the batch system.
CHAPTER 6: Conclusions and Recommendations
From this thesis study it was found that Candida lipoi'ytica can assimilate hydrocarbons
for the production of citric acid and biomass. In particular, throughout the study it was
found tbat:
1. Candida lipolytica N U - Y-I 095 can assimilated a n-plranin solvent but cannot
utilize kerosene;
2. The optimum levels of the following parameten for citric acid production are:
Initial Biomass Inocuium = 1042% volume n-Paraffh Concentration = 10- 15% volume Ferric Nitrate Concentration = 1 OmgL Temperature = 26-30°C
3. Under Batch Systems:
i. An aeration rate increase fiom 0.5 to 2.0 wm resuited in a 46 % decrtase in
overall biomass yield, however overall citric acid yield and productivity was
not affected;
ü. An agitation rate increase fiom 400 to 800 rpm resulted in a 58% increase in
o v d l citric acid yield and a 39% increase in citric acid productivity,
however o v d biomass yield was not affacted;
iii. Optimum citric acid production, under the influence of various d o n rates,
was 4.8 g/L at 2.0 wm and 400 rpm;
iv. Optimum citric acid production, under the influence of various agitation rates,
was 9.5 giL at 1200 rpm and 1.0 wm;
4. Under Fed-Batch Systems:
1.
. . Il.
iii.
iv.
v.
vi.
An agitation rate increase from 400 to 1200 rprn nsulted in a biomass yield
decrease of 0.15 g biomasdg n-parfi;
An agitation rate increase from 400 to 1 OOO rprn resulted in a ciûic acid yield 8
increase of 0.2 1 g citric acicüg n-par&;
The addition of an extemal agitator (Greey Lightnin In-Line Agitator) nimiing
at 14ûûrpm decreased the biomass yield by 0.07 g biomasdg n-parafnn h m
the control fermentation of no external agitation;
The addition of an external agitator (Greey Lightnin In-Line Agitator) running
at 1400 rpm increased the citric acid yield by 0.15 g citric acidfg n-paraffin;
The addition of an external agitator (Greey Lightnin In-Line Agitator) Mining
at 1600 rpm experienced 20-30 % of cell lyses;
An agitation rate increase of 400 to 1000 rprn resulted in a biomass yield
increase of 0.29 g biomasdg n - p d n ;
vii. An agitation rate increase of 400 to 1000 rpm resuited in a chic acid yield
increase of 0.77 g citric acid/g n-paraffin;
viii.The 3 cycle fed-batch fermentations increased citric acid yields by 0.56 g
citric acidlg n-paraffin, but the citric acid productivity was not afkted;
Ur. The fcyle fed-batch system increased overd1 citric acid yields by 1W/o h m
those found in the single cycle fed-batch system and i n d overail citric
acid yields by 200% fiom those found in the batch system.
Although the feasibility of utiiizing n-&anes as a feedstock for citric acid production hes
been demonstrated, M e r investigations are required:
Studies on coniinuous fermentation systems to M y understand the
advmtages/disadvantages on citric acid production;
Investigate the potential of utilizhg a mixture of carbohydrate and hydrocarbon
feedstocks to assess how carbohydrates can influence the assimilation of
hydrocarbons (lag phase may be shortened significantly);
Mutation of the yeast, to isolate higher citric acid producing populations;
hvestigate the use of surfactants to allow hydrocarbons to be kept in solution longer,
Further nins should be established to improve on citnc acid yields;
Based on present and future studies, conduchg a cost/benefit anaiysis would asses
the advantages/disadvantages of utilizing hydrocarbon feedstocks over carbohydrate
feedstocks (using the conventional method with A. niger) in an industrial scale
process in today's market.
CHAPTER 7: References
Aiba, S. and M. Matsuoka, "Identification of Metabolic Model: Citrate Production h m Glucose by Condidu lipolytica", Biotechnology and Bioengineering, 21, p. 1373- 1386 (1 979).
Aiba, S ., A.E. Humphre y and N.F. Millis, Biochemical Enginee~g-2d Edition, Academic Press Inc., p.3-190 (1973).
Aiba, S., K. L. Huang, V. Moritz, and J. Someya, Journal of Fermentation Technology, 47, p. 2 1 1 (1 969).
Akiyarna, S., T. Suzuki, Y. Sumino, Y. Nakao and H. Fukuda, "Induction and Citric Acid Productivity of Fluoroacetate-sensitive Mutant Strallis of Candida lipolyrica", Agriculturd and Biological Chemistry, 37, p.879-884 (1973).
Avchieva P.B. and A.Yu. VUiarov, "Obtaining Citric Acid Roducing Mutants of the Yeast Candidu lipolytica", Microbiology(Trans1ated fiom Mikrobiologiya), 62, p.161- 165 (1993).
Blair, G., P. Staal and H a m a n n & Reimer Corporation, "Citric Acid", in Encyclopedia of Chemcial Technology edited by Jacqueline I. Kroschwitz and Mary Howe-Grant, John Wiley & Sons Inc., p.354-380 (1991).
Boehringer and Mannheim, "Enzyrnatic BioAnalysis W-method for Citric Acid Boehringer Mannheim Method Procedures (outline accompanies product), Boehringer Mannheim (1 992).
Boehringer and Mannheim, %nzymatic BioAndysis W-method for D-Isocitric Acid", Boehringer Mannheim Method Procedures (outline accompanies product), Boehringer Mannheim (1 989).
Brady, R., "Citric Acid", in Chemicd Economics Handbook (CEH) at Intemet Site: w w w . c b r d . ~ c o n s u l t i n g . c o r n / C I N / J a n u a r y F e b 6 / ~ 2 . h , SRI Consdting, p.636.50ûûA (1996).
Briffauâ, J. and M. Engasser, "Citric Acid Production h m Glucose. 1. Growth and Exaction Kinetics in a S h e d Fermentor", Biotechnology and Bioengineeing, 21, p.2083-2092 (1979).
Briffaud, J- and M. Engasser, "'Citric Acid Production fiom Glucose. II. Growth and Excretion Kinetics in a Trickle-Flow Fermentor", Biotechnology and Bioengineering, 21, p.2093-2111 (1 979).
Bushenll, L.D. and H.E. Haas, "The Utilization of certain hydrocarbons by microocganisms, Journal of Bacteriology, 41, p.653675 (1941).
Douglas, L.J., "Adhesion to Sdaces", in The Yeasts Vol. 2 edited by AH. Rose and J.S. Harrison, Academic Press, p.239-275 (1 987).
FDA Regulations, "Code of Federal Regulations Title 21", Food and Bug Administration(Washington, DC), 21 CFR Cb.l(4-135) Edition, p. 114-1 15, 1995.
Fend, 2. and J. Leopold, "Mechanism of inhibition by acetic acid of the germination of spores of A.nzger", Nature (London), 179, p.922 (1957).
Funikawa, T., T. Ogino, T. Matsuyoshi, "Fermentative Production of Citric Acid fiom n- Paraffins by Sacchromycopsis lipdytica, Journal of Fermentation Technology, 60, p.28 1-286 (1982).
Furukawa, T., T. Matsuyoshi, Y. Minoda and K. Yamada, ''Fennentative Production of CiÛic Acid fkom n-Parafnns by Yeast", J o d of Fermentation Technology, 55, p.356- 363 (1 977).
Giovanni, M., "Response Surface Methodology and Product ûptimization", Food Technology, p.41-45 ( November 1983).
Gledhill, W.E., LD. Hill and P.H. Hodson, "Citrate Production fiom Hydrocarbons by Use of a Nonsterile, Semicontinuous Cell Recycle System", Biotechnology and Bioengineering, 15, p.963-972 (1 973).
Hattori, K., S. Yokoo and 0. hada, 'Effect of Ammonium Ion on the Ratio of Citric Acid to d-Isociûic Acid fomed fiom n-Paraffin", Journal of Fermentation Technology, 52, p.542-550 (1974).
Honecker, S., B. Bisping, Z. Yang, and H-J. Rehm, "Infiuemce of sucrose concentration and phosphate limitation on citric acid production by immobilized cells of RrpergiIIus niger. Applied Microbiology Biotechnology, 3, p. 1 7-24, (1 989).
Hurley, Re, J. de Louvois and A. M~lhall, 'Yeasts as Human and Animal Pathogens", in The Yeasts Vol. 1 edited by AH. Rose and J.S. Harrison, Academic Ress, p.207-273 (1 987).
Hustede, H. and H. Rudy, "M~ufactuire of citric acid by submergeâ fermentationy', US Patent 3 940 3 15,4p. (1976).
Iinika, H., 'Treparation of Citric Acid by a Fermentation Process", British Patent 1 218 476,Sp. (1971).
Imperia1 Oil, "Norpar 15 and Kero 1-K Product Sheets", Exxon Chemical Company (1 995).
Johnson, M. J., 'Vtilisation of Hydrocarbons by Microorganisms", Chemical Indu~Q, 78, p. 1532-1 537 (1964).
Kapoor, K. K, K. Chaudhary and P. Tauro, 'Titric Acid", in Prescott and Dunn's Industrial Microbiology edited by G. Reed, AVI Publishing Co. Inc., p. 709-747 (1982).
Kunura, K., K. Takayama and T. Nakanishi (Kyowa, Hakko Kogyo Ltd.), "Citric Acid Production", British Patent 1366 526,4p. (1 974).
King, V.A.-E. and R.R. Zall, "A Response Surface Methodology Approach to the Optimization of Controlled Low-Temperature Vacuum Dehydration", Food Research International, 25, p. 1-8 (1 992).
Kisser, M., C.P. Kubicek and M. Rohr, "Influence of Manganese on Morphology and CeU Wall Composition of A.niger during Citric Acid Fmentation", Arch. Microbiology, 128, p.26-33 (1980).
Kristinasen B. and C.G. Sinclair, "Production of Citric Acid in Continuous Culture", Biotechnology and Bioengineering, 21, p.297-3 15 (1 979).
Kristiansen B. and C.G. Sinclair, "Production of Citric Acid in Batch Culture", Biotechnology and Bioengineering, 20, p. 17 1 1 - 1 722 (1 978).
Kubicek, C. P. and M. Rohr, 'The Influence of manganese on enzyme synthesis and citric acid accumulation by Aspergi'I1u.s niger. European Journal of Applied Environmental Microbiology, 4, p. 167-1 75 (1977).
Kurtmian, C. P., Verbal telephone convmation h m his US. Department of Agriculture Lab (1996).
Kurtzman, CS., ''Identification and Taxonomy", in Yeasts edited by B.E. Kirsop and C.P. KurtPnan? Cambridge University Press, p.99-141 (1988).
Kyowa Hakko Kogyo Co. Ltd., "Process for Producing Citric Acid and Isocimc Acià", British Patent 1 278 01 3,4p. (1972).
Lavery, M., "Animal Cell Fermentation", in Fermentation A Ractical Approach edited by B. McNeil and LM. Harvey, Oxford Press, p.205-218 (1990).
Lincoln, R. E-, "Control of Stock Culture Preservation and Inoculum Build-up in Bacterial Fermentation", Journal of Biochem. Microbiology Tech. Eng., p. 481-500 (1 960).
Maddox, I.S. and S.H. Richert, " Use of Response Surface Methodology for the Rapid ûptimization of MicrobioIogical Mediay7, Journal of Applied Bacteriology, 43, p. l97-2M (1 977).
Matsumoto, Tey Y. Ichikawa and T. Nagata, %mess for Producing Citric Acid from Hydrocarbons by Fnmentation", U.S. Patent 4 424 274,7p. (1 984).
McLean, David and N.J. Burns, "CHG 3337 Data Collection and Interpretation-Course Notes", University of Onawa (1997).
Metcal f and Eddy, Inc . , Wastewater Engineering-Treatment, Disposal, and Reuse revised by G. Tchobanoglous and F. L. Burton, McGraw-Hill, Inc., 3" Edition (1991).
M i d , L.M. and G.F. Parker (Pfizer Limited), "Citric Acid Production", British Patent 1 418 561,4p. (1975).
M i d , LM. and G.F. Parker (Pfizer Limited), "Citric Acid Production", British Patent 1 369 295,4p. (1974).
Milsom, P.E., "Organic acids by fermentation, especially citric acid", in Food Biotechnology-1 edited by R. D. King and P. S. J. Cheetham, Elsevier Applied Science, London and New York, p. 273-306 (1987).
Miisom, P.E. and J.L. Meers, "Citric Acid", in Comprehensive Biotechnology Vol. 3 edited by M. Moo-Young, Pergamon Ress Ltd., p.665680 (1985).
Moo-Young, M., T. Shimizu and D.A. Whitworth, 6'Hydrocarôon Fermentations Using Candida I@olyticu. 1: Basic Growth Parameters for Batch and Continuous Culture Conditions", Biotechnology and Bioengineering, 13, p.74 1-760 (1 97 1).
Moo-Young, M. and T. Shimizu, "Hydrocarbon Fermentations Using Candida lipoiyfica. II: A Model for CeII Growth Kinetics", Biotechnology and Bioengineering, 13, p.76 1-778 (1 97 1).
Myers, R. H., 'Response surfàce desigas and the prediction variance huiction", in Statinical Design and Analysis of Industrial Experiments d t e d by S. Ghosh, Marcel Dekker I.C., p. 21 1-229 (1990).
Parton, C. and P. Willis, "Strain Presexvation, hocuium Reparation and Inoculurn Development", in Fermentation A Practical Approach edited by B. McNeil and L.M. Harvey, Oxford Ress, p.3964 (1990).
Pfizer Inc., " Fermentation Process for the Production of Citric Acid", Canadian Patent 978 120,23p. (1975).
Rane, K.D. and KA. S h s , "Cihic Acid Production by Yarrowiu Zzpolytica: Effect of Nitrogen and Biomass Concentration on Yield and Productivity", Biotechnology Letters, 18, p.ll39-ll44 (1996).
Rane, KD. and K.A. Sims, "Citic Acid Production by Candida lipot'ytica Y1095 in Ce11 Recylce and Fed-Batch Fermentors", Biotechnology and Bioengineering, 46, p.325-332 (1995).
Rane, KD. and KA. Sims, "Oxygen Uptake and Citric Acid Production by Candida lipoiytica Y 1 O W ' , Biotechnology and Bioengineering, 43, p. 1 3 1 - 1 3 7 (1 994).
Rane, K.D. and K.A. Sims, "Production of Citric Acid by Candida lipolytica Y1095: E ffect of Glucose Concentration on Yield and Productivity", Enzyme Microbial Technology, 15, p.646-65 1 (1 993).
Rose, A.H., "Responses in the Chernical Environment", in nie Yeasts Vol. 2 edited by A.H. Rose and J.S. Harrison, Academic Press, p.5-35 (1987).
Shennan, J. L. and J. D. Levi, "Progress in Industrial Microbiology", 13, v (1 974).
Shimizu, S., Y. Teranishi, A. Tanaka and S. Fukui, "Comparative Studies on Respiratory Activity and Cytochrome Content of Condido tropicalis pK 233 Grown on Hycirocarbon and on Glucose", Agricultural and Biological Chemistry, 38, p. 1 5 8 1 - 1 5 87 (1 974).
Shu, P. and M.J. Johnson, "The interdependence of Medium Constituents in Citric Acid Production by Submerged Fermentation", Journal of Bacteriology, 56, p.577-5 85 (1 948).
Snell, R.L. and L.B. Schweiger (Miles Laboratories Inc.), "Citnc Acid by Fermentation", British Patent 653 808,Sp. (1951).
Stanbury, P.F., A. Whitaker and S.J. Hall, Rinciples of Fermentation Technology, Elsevier Science Ltd., p. 1-2 10,243-272 (1 995).
Tabuchi, T. and M. Abe (Takeda Chernical Industries Ltd.), ' k e s s for the Production of Citric Acid", Japanese Patent 68 20 707 (1968).
Tabuchi, T. and K. Igoshi, 'Xeguiation of Enyme Synthesis of the Glyoxylate, the Citric Acid, and the Methylcitric Acid Cycles in Candida [ i p o ' a " , Agricultural and Biological Chemistry, 42, p.238 1-2386 (1978).
Takeda Chemical Industries Limited (Takeda Yakuhin Kogyo Kabushiki Kaisha), "Production of Citric Acid from Hydrocarbons by Candida Yeasts", Canadian Patent 973824, 14 p. (1975).
Takeda Chemical Industries Limited (Takeda Y a k u h Kogyo Kabushiki Kaisha), "Method for Producing Citric Acid", British Patent 1 297 243,8p. (1 972).
Takeda Chemical Industries Limited (Takeda Yakubin Kogyo Kabushiki Kaisha), 'Uethod of Producing Ciûic Acid by Fermentation", British Patent 1 21 1 246, 9p. (1 970).
Teranishi, Y., A. Tanaka and S. Fukui, "Some Specific Features of the Respiration of Hydrocarbon-utilizable Candida Yeasts", Agrkultural and Biological Chemistry, 38, p.1779-1783 (1974).
Thompson, D., "Response Surface Experimentation", Journal of Food Rocessing and Reservation, 6, p. 155-1 88 (1 982).
Touchstone Engineering Ltd., 'The Feasibility of Citric Acid koduction through Hydrocarbon Fermentation", Touchstone Contracting, Management & Engineering Services Ltd.-Internal publications, 2, 1996).
Weast, R.C., CRC Handbwk of Chemistry and ~hysics-69' Ed., CRC Press, p.163 (1989).
Wentworth, S.D. and D.G. Cooper, "Self-Cychg Fermentation of a Citric Acid Pmducing Strain of Candida lipol'yticu", Journal of Fermentation and Bioengineering, 81, p.400-405 (1 996).
Whitworth, D.A. and M. Moo-Young, "Hydrocarbon Fermentations: Oxidation Mechanism and Nonionic-Sufâctant Effêcts in a Culture of Candidu lipolytica", Biotechnology and Bioengineering, 15, p.649-675 (1 973).
Xu, Da., C. P. Madrid, M. Rohr, and C. P. Kubicek, 'The influence of type and concentration of carbon source on the production of citric acid by Aspergillus niger", Applied Microbiology Biotechnology, 30, p. 553-558 (1 989).
APPENDM A:
Analytical Methods
This appendix provides a detailed listing of aU analytical methods used throughout the experirnental program.
A1 Yeast Culture .-
Two yeast cultures were obtained fiom the Agriculhiral Research S e ~ c e s Culture
Collection Division of the U.S. Department of Agriculture, Peona, United States,
Candida Zipolytlica NRRL-Y4 094 and Candida iipoiytica NRRL-Y - 1 0%. Yeast
colonies were maintained on glucose agar plates and transferred monthly to fkesh agar
plates to insure ce11 viability.
A2 Cultivation of Yeast from Lyopbilized Preparation
To cultivate the yeasts from lyophilized preparation, yeast pellets were dissolved in the
sterilized nutrient media described in Table A 1.
Tabk Al: Nutrient Medium for the Reconstitution of Yeast Pellets fiom Lyophilized Preparation (Miall and Parker, 1974)
Components Concentration
Yeast Extract 3 .O Malt Extract 3 .O Peptone 5.0 Glucose 10.0
Hvdrocarbon Feedstock (mUL1 70
Loopfuls of each yeast and nutrient suspension were streaked, using aseptic techniques,
on to agar plates composed of the media described in Table 1 with the addition of 1 O@
of Bacto agar. The agar plates were incubated at 25°C for 2-3 days or until yeast colonies
populated the plates. The yeast plates were stored in a refiigerator held between 4 k 1°C.
These yeast colonies smed as parent strains for the inoculation of subsequent
fermentations conducted throughout the study. The yeast plates were transferred to fiesh
agar plates monthly to h u r e yeast ceii viability.
A3 Hydrwarbon Feeàstoek
The two hydrocarbon feedstocks used as the principle carbon source were kerosene and
normal parafxn solvents. Both feedstock were obtained from Imperia1 Oil Ltd., Toronto,
Canada. The kerosene solvent (Kerol-K) was composed of 80% total parafnns,
including both branchai (iso), cyclo and normal paraffins having predominmtiy 9 to 13
carbon atoms. Less than half of this total paranin content is normal p a r a h .
Approximately 20% of the kerosene solvent was composed of 9-10 carbon atom aromatic
molecules. The second hydrocarbon feedstock was a normal paraffin solvent (Norpar 15),
containhg more than 97% mixed normal parafnns composed of predorninantly 14 to 17
carbon atoms. Specifically, the Norpar 15 solvent (density of 77 1 g/L) was composed of
32.34% 14 carbon atom molecules, 4 2 4 % of 15 carbon atom molecules, 16-1 8% of 16
carbon atom molecules, 4-6% of 17 carbon atom molecules, and 1-3% 18 carbon atom
molecules.
114 Yeast Cell Growtb and Fermentation Media
The media descnbed in Tables A2, A3 and A4 were used for the inoculation yeast seed
culture, the pre-fermentation stage for yeast biomass growth and the femientative stage
for citric acid production, respectively.
Table A2: Medium for Yeast Seed Culture Growth (Midi and Parker, 1974)
Components Concentration Yeast Extract (g/L) 3.0 Malt Extrac t (fl) 3 .O peptone 5.0 Glucose (@) 10.0
Hydrocarbon Feedstock (mL/L) 70.0
Table A3: Medium for Pre-Fermentation Stage used for Yeast Biomass Growth (Mid and Parker, 1974)
Components Concentration Corn Steep Liquor (g/L) 2.6 Ammonium Sulphate (g/L) 4.0 Calcium Carbonate @IL) 15.0
Hydrocarbon Feedstock (mL/L) 87.0
Yeast Seed Culture Growth Solution (mL/L) 43 .O
Table A4: Medium for Fermentation Stage used for Citric Acid Production (Miail and Parker, 1974)
Components Concentration u=a ( g f u 2.0 Magnesium Sulphate Heptahydrate ( g L ) 0.4 Calcium Carbonate (g/L) 6.0 Potassium Dihydrogen Ortho-Phosphate (fi) 0.75 Thiamine Hydrochloride (rng/L) 0.25 Ferric Nitrate Nonahydrate (mg/L,) 0.65 Zinc Sulphate Heptahydrate (m@) 1.2 Copper Suiphate Pentahyhte (mg/L) 0.3 1 Manganese Sulphate Monohydrate (ma) 0.27
Hydrocarbon Feedstock (mLL) 157.0
Re-Fermentation Yeast Growth Solution (mWL1 43.0
Media described in Tables A5 and A6 were used for feeding in fed-batch fermentation
s ystems.
Table AS: Medium for Feeding of Fermentation Stage in 7 Day Fed-Batch Systerns - Citric Acid Production (Miall and Parker, 1974)
Components Concentration Magnesium sulphate heptahydrate ( g l ' ) 0.4 Potassium dihydrogen ortho-phosphate (gL) Thiamine hydrochloride (mgL) F&c nitrate nonahydrate (ma) Zinc sulphate heptahydrate (mg/L) Copper sulphate pentahyûrate (mgL) Manganese sulphate monohydrate (mg@)
Table A6: Medium for Feeding of Fermentation Stage in 1 8 day Fed-Batch S ystems - Citric Acid Production
Components Concentration WL) 0.25
Magnesium sulphate heptahydrate (a) 0.4 Potassium dihydrogen ortho-phosphate (a) 0.75 Thiamine hydrochloride (m&) 0.25 Femc nitrate nonahydrate (rng/L) 0.65 Zinc sulphate heptahydrate (mg/L) 1.2 Copper sulphate pentahydrate (mg/L) 0.3 1 Manganese sulphate monohydrate (mglL) 0.27
Ali media was sterilized by autoclaving at 121°C and pressure of 15 p.s.i.g. for 20
minutes, followed by a slow exhaust, prior to the addition of the hydrocarbon feedstock
and the yeast ce11 inoculation.
A 5 Annlytical Methods
A.5.1 Biomass Concentration
Yeast concentration was detennined by dry c d weight arialysis. 40 mL samples taken
over time of the fermentation broth were centrihged at 12,000 ~pm at 5OC for 15 minutes
using a Dupont Sorvall Instniment RCSC centrifuge. Supernatant was carefully removed
nom the centrifuge vials using Pasteur pipettes, making sure not to disturb the yeast
pellet at the base of the vial. The supernatant was set aside for M e r anafysis of ciaic
acid and hydrocarbon concentrations. The yeast peliet was removed nom the vial and
placed in a pretreated and pre-weighed duminum dish. (The aluminum dish, prior to
being used, had been placed in a muffie fumace at 500 f 50°C for 1 hour and cooled in a
desiccator to remove any volatile residue.) The aiuminum dish containing the yeast pellet
was placed in a drying oven at 104 f 1°C for 2 hours. The aluminum dish plus the dricd
yeast pellet was cooled in a desiccator and then weighed. The cycle of drymg, cooling
and weighing was repeated until the change in weight was less than 4% of the previous
weight or OSmg. The dried yeast sample was then ignited by piacing the dish in a mufile
f h a c e at 500 f 50°C for 20 minutes and then cooled in a desiccator and then weighed.
Once again, the cycle of ignition, cooling and weighing was repeated until the change in
weight was less than 4% of the previous weight or 0.5mg. The concentration of dried
yeast cells was calculatad by the followhg equation:
yconc = (mbef - nhft) / Vsamp2e
Ycm, = concentration of dried yeast pellet in rng/mL (fi) rnbq = weight of dried yeast pellet plus dish beforr ignition, mg
mafi = weight of dned yeast pellet plus dish after ignition, mg vsumple = volume of the fermentation broth sample, mL
This method measures the organic composition of the yeast cells, and does not take into
account the trace inorganic compounds in the yeast. It should also be noted that although
the fermentation broth also contains calcium carbonate, this inorganic salt is stable up to a
temperature of 825°C (Metcalf and Eddy, Inc., 1991) and will not influence the yeast
concentration.
A.S.2 Hydrocarùun Concennatbn
Hydrocarbon concentration was determineci by extracting the residual hydrocarbons from
the fermentation broth supernatant with hexane as the extraction solvent. The volume of
the supernatant was recorded with a graduated cylinder. A known volume of hexane
solvent was decanted into a separatory h e l , at lest twice the volume of fermentation
supernatant, followed by the measured volume of supernatant. The final addition was
done slowly to avoid emulsification. The separatory fùnnel was then shaken gently and
placed on a retort stand to allow the organic and aqueous phases to separate. The shaking
and settling steps are repeated several tirnes, mtil there are no visible hydrocarbon
droplets left in the aqueous phase iiquid. The aqueous phase iiquid is decanted through
the base port of the separatory funriel. Anhydrous magnesium sulphate was added to the
organic phase liquid rexnaining in the funne1 to remove any residuals aqueous material.
The magnesium sulphate was then filtered out of the organic phase liquid. The volume of
the organic phase liquid was then recorded with a graâuated cylinder. The concentration
of residual hyârocarbon in the fermentation broth sample cm then be rneasured as
follows:
Hcconc = concentration of hydrocarbon feedstock in supematant of centrifuged fmentation sample, rng/mL (gL)
VOrg = volume of organic phase liquid fiom extraction containing both hexane solvent and hydrocarbon feedstock, mL
Vhex = volume of hexane solvent added to extraction process, mL V-= volume of supematant of cenûifbged fermentation sample, mL PH(: = density of hydrocarbon feedstock, mg/mL
A.5.3 Citric Acid Concentration
A. 5.3.1 Ertzymatic Assay Analysis
Citric acid concentration was detemine by enzymatic assay. The eazymatic test kit was
obtained fiom Boehringer Mannheim, Montréal, Canada. The test, as outlined by
Boehringer Mannheim (1992), is based on the prïnciple that citric acid (citrate) is
converted to oxaloacetate and acetate in the reaction catalyzed by the enzyme citrate lyase
(CL) :
CL Citric Acid - ûxaioacetate + Acetate
In the presence of the emyme malate dehyrdmgenase (MDH) and L-lactate
dehydrogenase (L-LDH), oxaloacetate and its decarboxylation product pyruvate are
reduced to L-malate and L-lactate, respectively, by reduced nicotinamide-adenine
dinucleotide (NADH):
The amount of NADH oxidized in reactions [A41 and [As] is stoichiometnc with the
amount of citric a d . NADH is determined by means of its absorbance at 340 nm.
The enzymatic test kit is composed of two solutions, containhg NADH and Citrate lyase
solution. The NADH was a 12 mL solution composed of 1.4 g of lyophilisate, consisting
oE glycylglycine bufYer, pH 7.8; malate dehydrogenase, 136 U; L-lactate dehydrogenase,
280 U; NADH, 6.0 mg; stabilizers. The Citrate lyase Solution was a 0.3 mL solution
composed of 50 mg of lyophilisate citrate lyase, 12 U. NADH concentration was
determinecl using a Beckman DU40 spectrophotometer to measure absorbance readings
at a wavelength of 340 m. Glass cuvettes with a 1 cm light path were used with the
spectrophotometer.
For each citric acid concentration measurement two absorbance readings need to be
taken, one before the oxiâation of NADH in the sample solution and the other &er the
reaction is complete. A blank reading must always be taken to avoid the influence of the
absorbance of any contamhmts present. Table A7 describes the solutions that need to be
prepared.
Table A7: Reparation of Solutions for Citric Acid Assay (Boehringer and Mannheim, 1992)
Pipette into Cuvettes Blank Solution Cuvette Samle Solution Cuvette Mixture 1: NADH Solution Distilled Water Fermentation Supematant * Mixture 2: Mixture 1 + Citrate lyase Solution 0.02 mL 0.02 mL * Supematant a* hydrocarbons had ben extracted
Once Mixture 1 is pipetmi into the Blank solution cuvette and the Sample solution
cuvette, each cuvette was covered with parafilmm and mixed by gentle swirling. The
absorbame at 340111x1 of these solutions (A,) was read after approxhately 5 minutes. The
reaction in each cuvette was started by addition of the Citrate lyase Solution, as described
by Mixture 2. Absorbance at 340nm was then taken on completion of the reaction (AJ in
each cuvette, after 5 minutes. The absorbance differences (AA = A, - A& for both the
blank and sample cuvettes were detennined. The absorbance diffmnce of the blank was
then subtracted h m the absorbance difference of the sample, as follows:
where :
The citric acid concentration c m then be calculateci by the foliowing equation:
where:
Cconc = citnc acid concentration, g/L
Vf = final volume in cuvettes, 3.02 mL V, = hydrocarbon-fm fe~nentation supanatant volume, mL
MW = molecular weight of citric acid, grno1 = 192.1 g/mol
d = light path, cm = 1.0 cm
E = absorption coefficient of NADH at 340 m, L x mmor' x cm*' = 6.3 L x mmol-' x cm-'
A.5.3.2 High Pe fonnance Liquid Chromatograph (HPLC) Anabsis
Cihic Acid concentration was determined by isocratic HPLC analysis using a Hewlett
Packard 1090 Liquid Chromatograph and Waters p Bondapak Cl8 3.9 x 300 mm
column. The mobile phase consisted of 0.1 M KH,PO, in distilled deionized water
adjusted to a pH of 2.5 with concentrated &PO4. Analysis consisted of a mobile phase
fiow rate of 0.6 dhn.in, ambient column temperature (25OC), and injection volume of 20
PL. 1 .O ml of each n-paraffin fiee supernatant sample was fikered using 0.2 pm
Millipore GV-13 filters prior to injection into column. Absorbante readings were taken
at a wavelength of 215 nm and citric acid concentrations determined using a standard
curve of absorbance readings at various known citric acid concentrations. However, in
this citric acid concentration there is also trace amouts of iso-citric acid. The HPLC
retention tirne c m o t differentiate between the two isomm. Thmfore to get the acnial
citric acid values minus the isocitnc acid, isocitric acid was determined using an
enzymatic assay.
The enymatic asay kit was obtained by Boehringa and Mannheim, Montréal, Canada.
The kit contained solutions of imidazole buffer, iyophilisate (consisting of NADP) and
manganese sulphate. The assay involved operations at a wavelength of 340 nm and a 1
cm light path using a Beckman DU40 spectrophotometer. Once the concentration was
determulecl it was subtracted from the HPLC concentration and citric acid was
det ermined.
The principle behind the test is that D-isocitric acid @-isocitrate) is oxidatively
decarboxylated by nicotinamide-adenine dinucleotide phosphate (NADP) in the presence
of the enzyme isocitate dehydrogenase (ICDH):
D-isocitrate + NADP' ICDH 2-oxoglutarate + C O + NADPH + K [A81 I
The amount of NADPH f o d in reaction [Ag] is stoichiometric with the arnount of D-
isocitrate. NADPH is determined by means of its absorbante at 340 nrn. The bound D-
isocitric acid (esters, lactones) is detennined after alkaline hydrol ysis, [Ag] and [A 1 O]:
D-isocitric acid ester + H,O pH 9-10 D-isocitrate + alcohol
D-isocitric acid lactone + H,O pH 9- 10 D-isocitrate
APPENDIX B:
Raw Data
This appendix lists ali of the raw data collected throughout the experiments presented in this thesis
B.l Screening of Hydrocarbon Feedstock and Yeast Strain
Table B.l: Citric Acid and Biomass Concentrations When Using Kerosene and Candidrr J@o&tica Y-1094
- -- --
Time Yeast Conc. itric Acid Conc Kerosene Conc. ( b w (Sn) (fi) (%LI
Table B.2: Citric Acid and Biomrss Concentrations When Using n-Parafiin and C~~ @oI@ca Y4094
T h e Yeut Conc. Citric Acid n-Paraffin @W W) (gn) (%LI
Table B.3: Citric Acid and Biomass Concentrations When Using Kerosene and CandUa l@o&tica Y4095
Time Yeast Conc. itric Acid Conc Keroseue Conc. (hm) (fm 6m en)
Table B.4: Citric Acid and Biomass concentrations When Using a-Parnffin nid Candido l@o&tàca Y-1095
B.2 Raw Data of Central Composite Design Flask Batch Tests - Table 33.5: Biomass Mode1 - Raw Data of Flask Batch Tests
BIO-G-L BIOMASS% HC-% FE-MGJ TEMP-C
Table B.6: Citric Acid Model - Raw Data of F'lask Batch Tests
CITR-G-L BIOMASS% HC-O/a FE-MGJ TEMPC
B3 Study of the Effects of Aeration on Citric Acid and Biomass Production
Table B.7: Citric Acid and Biomass Production at 0.5 vvm and 400 rpm
T h e CiMc Acid Biomass n-par- g CA/ g biomassf Prodnctivity hour %L %L g/L g n-pairnia g n - p a W i gRrh CA
Table B.8: Citric Acid and Biomass Production rt 1.0 wm and 400 rpm
Time Citric Acid Biomass n-pinma g CA/ g biomusl Productivity hour fi@ fl glL g n-pardh g n-par* W h CA
Table B.9: Citic Acid aad Biomass Production at 2.0 wm and 400 rpm
T h e Citric AcY Biomass n-par- g CA/ g biomasd Productivity hour fi %L g/L g n-paraffin g n-paraffin gRrh CA
B.4 Batch Fermentations - Effects of Agitation on Citric Acid and Biomass Production
Table B.lO: Citric Acid and Biomass Production at 400 rpm and 1.0 vvm
T h e Citric Aci Biomass n-Paraffin gCA/ gbiomasd Productivity ( h o w (%LI (pn) (%L) g n-panaffin g n-parfiin (%Gb) CA
Table B.ll: Citric Acid and Biomass Production at 800 rpm and 1.0 vvm
Time Citric Aci Biomass n-Par& gCA/ gbiomasd Productivity W u 0 (%LI (fi) (%L) g n-paraffin g n-parfiin (gn-h) CA
Table B.12: Citric Acid and Biomass Production at lûûû rpm and 1.0 vvm
Time Citrie Aci Biomass n - P a r e gCN gbiomassl Praductivity mur) (gn) (pn) (gL) g n-paraffin g n-parnin @&h) CA
Table B.13: Citric Acid and Biomrss Proàuctio~~ at 12 O0 rpm and 1.0 w m
Time Citric Aci Biomus n-Paraffin gCAI gbiomissl Prodiictivity (hou0 (fi) WL) (sn) g n-paraffin g n-parib (g/~-b) CA
B.5 Fed-Batch Fermentations - Effects of Agitation on Citric Acid and Biomass Production
Table B.14: Cihic Acid and Biomass Proàucüoa at 400 rpm and 1.0 wm at a fading rate of 1.OLlday
Time Citric Acid Biomass n-paraffin &A/ gbiomwd Prductivity hour %L fl g/L g n-parafh g n-parfiin (g/L-h) CA
Tabk B.15: Citric Acid and Blomass Proàuction at 800 rpm and 1.0 wm at a feeding rate of l.OL/day
Time Citric ~ c i d ~iomass O-par- gCM gbiomwsl Productivity hour fi fi glL g a-parnfllin g n-parfhn (dLh) CA
Table B.16: Citric Acid and Biomass Production at 1000 rpm and 1.0 wm at a fading rate of 1.OLlday
Time Ci- Acid Biomass n-paraffin gCA/ gbiomrrd Productivity hour fi 8n g/L g n-paraffin g n-parnin (bn-h) CA
Table B.17: Cihic Acid and Biomass Production at 1200 rpm and 1.0 w m at a fceding rate of 1.OLlday
T h e CiMc Acid Biomass n-pa- gCA/ gbiomassl Productivity hour fl %L g/L g n-pad!fh g n-pPrnin (%L-h) CA
B.8 Fed-Batch Fermentations - Effet of Extemai Agîtator
Table B.20: Citrie Acid and Biomus Production aï an Extemai Agitation of 1400 rpm, 1.0 wm and fotding rate of 1.OL/day
Time itrk Aci Biomass n-paraffin gCA/ gbiomud Productivity hour fi@ f i g/L I I - p d g n-parnin W h ) CA
Note: Fermentor Agitation is kept at 400 rpm througinout fementatioas
Tibk B.21: Citric Acid and Biomass Production at an Extemal Agitation of 1600 rpm, 1.0 w m and feeàing rate of l.OL/day
Time itric Aci Biomass nparnffin gCN g b i o m d ~roduc6viG hour en %L g/L n-para@ g n-parfllin (gn-h) CA
Note: Fermentor Agitation k kept at 400 rpm throughaut fermentations
B.9 Fed-Batch Fermentations - Effeet of a 3 Cycle Feeding System
Table B.22: Citric Acid and Biomass Production at 400 rpm, 1.0 wm and feeding rate of 1.0 Uday
Time Citric Acid Biomass n-paraffin gCN gbiomass/ ~ r o d u c t i ~ ~ hour %L ln g/L g n-paraffin g n-par* (gn-h) CA
Table B.23: Cibic Acid and Biomus Production at 800 rpm, 1.0 wm and feeding rate of 1.OLIday
Time Citric Acid Biomass n-paraffin gCAI gbiomassi Productivity hoor fi fl g/L g n-paraffin g n-parffin (dL-h) CA
Tabk B.24: Citnc Acià and Biomass Production at 1000 rpm, 1.0 wm and feeding rate of l.OL/day
Time Citric Acid Biomws a-parifnn gCA/ gbiomad Productivity hour fi fl g/L g O-paraffin g i-parfni (gn-h) CA
Tabb B.25: Citric Acid and Biomass Production at 1200 rpm, 1.0 wm and feeding rate of 1.0 L/day
Time Cibie Acid Biomass nparaffîn gCN gbiomassl Productivity hour f i fi g / L gn-pannii g n-poilfia (dGb) CA
Table B.26: Citric Acid and Biomass Production at an External Agitation of 1400 rpm, 1.0 wm and feeding rate of 1.0 Wday
T h e Ciaie Acid . Biomass noparaffin gCA/ gbiomassl Prductivity hour !ivL fl g /L g n-paraffin g n-parfb (gn-b) CA
Nore: Fermentor Agitation i( kept at 400 rpm thughout fermentations
APPENDIX C:
Central Composite Experimental Design
This appendix describes the experimental strategy, the central composite design, and the model-building techniques
C.l Background
A response d a c e methodology in chernical process development is demonstrated by the
work of Thompson (1982). The centrai composite design for second order desigm is as
follows:
where
E(Y) is the expected value of the nsponse variable Po, pi, Pj are the mode1 parameters X, and X, are the codai factors being studied k is the number of factors being studied
The design consists of a 2k factorial or hctional factorial design augmented by 2k axial
(star) points at (fa, O, O,. . ., O), (O, O, ka,. . ., O),. . ., (O, O, 0,. . ., ka) and n, center points at
(0, 0,. . ., O), where a is the distance of the star point fiom the center. The choice of a
establishes the rotatability of a central composite design. For a four factor design, a
should be set at 2.0, as discussed in McLean et al. (1997). it is the recommended a value
for a four factor design.
The cenûai composite design that was used for this study was a four factor design for
ciûic acid production and bioma~s production. The design used is describeci as a Box-
Wilson central composite circumscribed design for four factors (McLtan et al., 1997).
The design if fùrther described in seaion C2.
C.2 Four-Factor Centrai Composite Design
The pmposed mode1 for the four-factor design is :
The run conditions used for the batch flask tests for citic acid and biomass production
are described in Tables C 1.
Tabk C.l: Run Conditions for the Four-Factor Centrai Composite Design
Biomass HC FE Tem~ Coded Codcd Coded Coded XI,% Xb% Kmgk &,Oc X , xt x3 X1 7.5 7.5 6 25 -1 -1 * 1 - 1
C.3 Anaiysis of Model
Using lest squares anaîysis, four critical assumptiom are made: (McLean et al., 1997)
1. The values of the expairnenital factors are known exactly; the associated
variance is zero. In other words, the uncertainty associated with the vaiue of
the experimental factor has less of an effect on the response value than the
uncertainty associated with the measured value of the mponse itself.
2. The expected vaiue of the random aror is zero.
3. The variance of the random mor is constant over the range of the
experimental factors that are used to collect the data
4. T'here is no association of the random enor for any one &îa point with the
random emr for any other &ta point.
It is necessary to perfom an analysis of the nsidual h m the model, to detamine the
adequacy of the least squares fit. Plots of the residuals, both normal probability plots and
plots of the residuals versus pndicted values of the response variable, are constructed to
ver@ that the random error follows the underlying assumption.
ANOVA (Analysis of Variance) tables are consiructed as weli. In an ANOVA d y s i s ,
the variation is separated hto regession and residual variance, as show in Table C2.
Table C2: ANOVA Table for Multiple Rcgression Modeis (Thompson, 1982)
Source Degrees of Sum of Mean F Variance Freedom wu=- squares
Regression P hss m=ReSS/p Residual N p 1 MR=RSS/N-pl
w RSS
Oadc of fit) N-p-1-( II,,- 1) RLSS m=RLSSI N-pl-( q-1)
mm m r ) %-1 %SS MRp=RpSS/ 4-1
Total N- 1 SS
In multiple regression models, there are p degrees of fmdom for the sum of squares due
to regression, because p coefficients, p,, p,, . . ., p , m u t be estimated to obtain RSS. The
degrees of M o m for the entire system is dehed as the total number of experimental
nuis minus 1 (N-1).
nierefore,
Sum of Squares
ss=Z(yi-y)2 P l
whae SS is the s u m of squares of deviations of the obsmed values h m their
mean value;
RSS= C ( y i - y d 2
where RSS is the sum of squares of the residuals;
RSS = ss - RSS
whcre &SS is the sum of squares due to the regession and,
yi is the obscnted value of lzsponse variable at design point i
y is the mean value of the observcd responses
y,, is the predicted response variable at point i
The lack fit test checks whether the order of the mode1 is correct. The test
involves dividing the variability of the residuals into two components,.pure error
and lack of fit. The pure error sum of squares is caiculated using the values of the
nsponse variables at the replicate (center) points (Thornpson, 1982):
%SS = z (y, - y$ [CS1
The corresponding degrees of freedom is the number of center points minus 1
1). The lack of fit F-statistic shown in Table B2 provides the basis for the
evaluation of lack of fit.
The coefficient of detemination, R', is another measure of the aâequacy of the fit
where,
R2 = 1-RSS/SS = (SS - RSS)/SS = &SS/SS
Values range fiom O to 1, with values closer CO 1 indicating high consistencies
between the preâicted and observeci values. R2 also provides a measure of the
fiaction of the total variability in the data explaineci by the regression.
Mode1 Results
This appendix contains the empirical modeling data and results for the control.
D.l Citric Acid Production
- Table D.1: Parameter Estimates for Citric Acid Production Model
1 83 Pt 1 822 Estimate 8.849 0.995 1.329 0.493 0.520 -1.027 -0.711 Std. Err- 0.421 0-227 0.227 0.227 0.227 0.208 0.208 1(w 21.028 4.380 5,847 2. t 69 2.286 -4.932 -3.4 13 plevel 0.000 0.00 1 0.000 0,046 0.036 0.000 0.004
Table Do1 cont'd : Parameter Estimates for Citric Acid Production Model
Std En. 0.208 0.208 0.278 0.278 0.278 0.278 0.278 0.278 t( 16) -4.367 -6.114 -0.734 0.325 0.325 1.219 0.882 -0.518 plevel 0.000 0.000 0.474 0.748 0.748 0.240 0.390 0.61 1 Variance explained 89.7% R4.947
Table D.2: Parameter Estimates for Modified Citric Acid Productioa Model*
Bo 1 B 2 81 i Estimate 8.849 0.995 1.329 0.493 0.520 -1.027 -0.711 Std. Err. 0.393 0.2 12 0.2 12 0.2 12 0.212 O. 194 0.195 t(22) 22.492 4.685 6.254 2.32 2.445 -5.275 -3.65 1 plevel 0.000 0.000 0.000 0.030 0.023 0.000 0.00 1
Table D.2 ewt'd: Pairmeter Estima- for Modifd Citric Acid Production Model
&3 Bu Estimate -0.909 - 1.273 Std. En. 0.194 0.195 a9 4.671 4.541
Variance explained 87.5% R= 0.935
* Modificd Model has parameters with plevels greater uian 0.05 in Table DI mnoved
Table D.3: ANOVA Table for Citric Acid Production Mode1
Source of Dqrees of Sum of M a n Squares F ~ , v I . v ~
Variance Freeùom Squares Regession 8 166.6 20.8 19.1 2.40 Residuai -- 77 24.0 1.09 (Lack of Fit) 16 23-18 1 -44 10.29 3.93 Pure Error) 6 0.82 O. 14 Total 30 1W.6
Table D.4: Pure Error Variance for Citric Acid Production Mode1 - - - . .
For Reblicate Set Mean S m of Residuals Squared Degrees of Freedom 6 Pure Error Variance 0.14
Table D.5: Correhtion Matrù for Parameter Estimates for Citric Acid Production
81 B2 D3 81 i Pz A3 Bo 1.00 0.00 0.00 0.00 0.00 4.51 -0.51 -0.51 -0.51 Bi 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 82 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 & 0.00 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 8, 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00 0.00 al -0.51 0.00 0.00 0.00 0.00 1.00 0.11 0.11 0.11
4 .51 0.00 0.00 0.00 0.00 0.11 1.00 0.11 0.11 a3 -0.51 0.00 0.00 0.00 0.00 0.11 0.11 1.00 0.1 1
-0.51 0.00 0.00 0.00 0.00 0.11 0.11 0.11 1.00
D.2 Biomass Production Model
- Table D.6: Parameter Estimates for Biomass Production Mdel
1 1 I 822 Estimate 4.613 0.457 0.48 1 0.181 0.289 -0.232 -0.215 Std En- 0.1 16 0.063 0.063 0.063 0.063 0.057 0.057 Y 16) 39.474 7.236 7.632 2.865 4.58 1 4.019 -3.717 plevei 0.000 0.000 0.000 0.01 1 0.000 0.000 0.002
Table D.6 coot'd: Parameter Estimates for Biomass Production Model
A3 BI t BI O 814 3 4 4
Estimate -0.106 -0,404 -1.175 0.017 4.050 0.170 -0.062 -0.055 Std. Err. 0.057 0.057 0.077 0.077 0.077 0.078 0.078 0.077 t( 16) -1.792 4.981 -2,264 0,226 -0.647 2.199 -0.808 -0.712 plevel 0.092 0.000 0.038 0.823 0.526 0.043 0.431 0.487 Variance explained 93.2% R= 0,965
Table D.7: Parameter Estima tes for Modified Biomass Production Model * I B 3 81 i
Estimate 4.6 1309 0.457 0.48 1 0.181 0.289 10.232 -0.215 Std. Err. 0.109 0.059 0.059 0.059 0.059 0.054 0.054 @O) 42.044 7.706 8.128 3.052 4.880 -4.281 -3.959 plevel 0.000 0.000 0,000 0.006 0.000 0.000 0.00 1
Table D.7 cont9d: Parameter Estimates for Modified Biomass Production Model*
&3 B i 2
Estimate 4.106 -0.404 -1,175 0.170 Std, Err, 0.054 0.054 0.072 0.072 t(20) -1.910 -7.436 -2.412 2.343 plevel 0,071 0.000 0.025 0.029 Variance explained 92.5% R= 0.962
* Modified Model has B parameten with plevels greater than 0.05 in Table D6 removed
Table D.8: ANOVA Table for Biomass Production Model
Source of DegrCa of Suaof MeaiSquares F F.s,~ 1 .V2
Variance Frœdom Squares Regression 10 20.64 2.06 22.89 2.35 Residual 20 1.69 0.09 (Lack of Fit) 14 1.38 O. 10 2.0 3.96 (Pure Error) 6 0.3 1 0.05 Total 30 22.33
Table D.9: Pure Error Varirince for Biomass Production Model --
For Replicate Set Mean 3 -87 Surn of Residuals Squared 0.3 1 Degrees of Freeàorn 6 Pure Error Variance 0.05
Table D.10: Correiation Matru for Parameter Estimates for Biomass Production
Tabk D.10 wat9d: Comlation M a t h for Paramter Estimates for Biomass Praducti011
APPENDIX E
Cell lyses and CeU Viability
This appendix presents the data on the cell lyses and ceU viabiiity in fermentation systems
E.l Cell Lyses
Table-E.1 : Cell Lyses during Fermentation* using Greey Lightnin In-Line Agitator
The, Ce11 Lyses at 1400 rpm Ceil Lyses at 1600 rpm hrs (% ceh) (% cells) O 5.4 4.7 42 6.9 10.8 84 7.8 16.9 126 10.5 21.3 168 14.1 28.3
* Analysis was done in a febbatch systcxn at 1.0 wm, 400ipm f-tor agitation, 28% and fetdmg rate of 1 .O Uday
Measunment of Celî Lyses
Ce11 lyses was determined using a hemocytometa. The hemocytometer allows for a
direct microscopie count of yeast cells. The percentage of lysed ceils was daemiined as
the number of cells lysed per the total number of cells. A hemocytometer is a slide with a
counting chamber 0.1 nmi deep. On the bottom of the couting chamber there is an
etched square divided into nine squares, each 1 mm2. The central square is divided into
16 squares. Thmefore, within the central square millimeter thae are a total of 400 small
squares. To count the number of cells, count only the number of cciis in 5 of the squares
(central square and the 4 corner squares). Multiply the numba of cells counted in the 5
squares by 5 to estimate the number of cells/mm2 of SUrfkce.
The procedure involves placing a drop of the yeast suspension in each of the two
comting chambers. The hemocytometa is then placed on the microscope stage and the
yeast cells are countcâ using the 40x lem.
E.2 CeU Viability
- Tabk E.2: Cell Viability of Yeast during 3 Cycle Fed-Batch System
Time, Ce11 Viability at Ce11 Viability at Cell Viability at hrs 800 rpm, % 1ûûû rpm, % 1200 rpm, %
O 61 58 65 108 75 69 72 216 82 78 80 324 72 77 76 432 59 63 49
Measurement of ceM viabiiity
Ce11 Viability was measwed ushg a rnethylene blue staiaing procedure. Blue cells
indicate dead cells. Therefore, ce11 viability was taken as the numba of unstained cells
dinded by the total number of cells. The ce11 count follows the hemocytometer
procedure described in section E. 1
Sam~le Calculations
This appendix illustrates the sample caIculations
F.l Y ield rad Productivîty Calculatiois
Note: using data from Table B.7
OveraH Citric Acid Yield
= total g citrïc acid produced / total g n-paraffin consumed
= 3.56 1 (1 14-79) = 0.102 g citnc acid/g n-paraffin
Overall Biomass Yield
= totai g biomass produced / total g n-paraflin consumed
= (10.88-3 32) l(114-79) = 0.210 g biomass / g n - p d n
Citric Acid Produetivity
= citric acid produced l time duration of fermentation
= 3.56 / (16848) = 0.03 @-h
F.2 Central Composite Design Cdculatioas
Note: Biomass Production Model is Considered here
Table F. 1 shows the Predicted and Residual values of the rnodel.
Calcuiating the mean,
S m of Residual Squared = (-0.065)' + (-0.1 53)' + . . . + (-0.097)~ + (-0 -092)' + (-0.4 1 212 = 1.69
Sum of squares = (2.03401)' + (3.27-4.01)' +...+ (2.12401)' + (4.99401)~ = 22.33
S m of squares of regession = 22.33 - 1.69 = 20.64
Pure Error Variance = (0.233)~ + (0.023)~ +. . . +(0.4 1 212 + (0.092)' = 0-3 1
Table F.1: Redicted and Residual Values for Biomass Production Mdel Predicted Residual
2.095 0.065
There fore,
S m of Squares for jack of fit = 1.69-0.3 1 = 1.38
Mean Squares for regression = 20.64/10 = 2.06
Mean Squares for residual = 1-69/20 = 0.09
Mean Squares for lack of fit = 1.38/14 = 0.10
Mean Squares for pure error = 0.3 1 /6 = 0.05
F (regression) = 2.06/0.09 = 22.89
F (residual) = 0.1 /O .O5 = 2 .O