Feasibility Studies for

Encapsulated Cell Bioaugmentation of Contaminated Aquifers

by

Peyman Moslemy

Department of Chemical Engineering

McGill University, Montreal

January 2002

A thesis submitted to the Faculty of Graduate Studies and Research in partial

fulfillment of the requirements of the degree of

Doctor of Philosophy

© Peyman Moslemy 2002 National Library Bibliothèque nationale 1+1 of Canada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395, rue Wellington Ottawa ON K1A ON4 Ottawa ON K1 A ON4 canada canada

The author bas granted a non­ L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library ofCanada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies ofthis thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format électronique.

The. author retains ownership ofthe L'auteur conselVe la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts from it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son pemnsslOn. autorisation.

0-612-78740-0

Canada Preface

The manuscript-based thesis option was chosen according to the following thesis preparation guideline given by the Faculty of Graduate Studies and Research:

As an alternative to the traditional thesis format, the dissertation can consist of a collection ofpapers ofwhich the student is an author or co-author. These papers must have a cohesive, unitary character making them a report of a single program of research. The structure for the manuscript-based thesis must conform to the following:

1. Candidates have the option ofincluding, as part ofthe thesis, the text ofone or more papers submitted, or to be submitted, for publication, or the clearly-duplicated text (not the reprints) ofone or more published papers. These texts must conform to the "Guidelines for Thesis Preparation" with respect to font size, line spacing and margin sizes and must be bound together as an integral part ofthe thesis. (Reprints of published papers can be included in the appendices at the end ofthe thesis.)

2. The thesis must be more than a collection of manuscripts. All components must be integrated into a cohesive unit with a logical progression from one chapter to the next. In order to ensure that the thesis has continuity, connecting texts that provide logical bridges between the different papers are mandatory.

3. The thesis must conform to aU other requirements of the "Guidelines for Thesis Preparation" in addition to the manuscripts. The thesis must include the following:

(a) a table ofcontents (b) an abstract in English and French (c) an introduction which clearly states the rational and objectives ofthe research (d) a comprehensive review of the literature (in addition to that covered in the introduction to each paper) (e) a final conclusion and summary

4. As manuscripts for publication are frequently very concise documents, where appropriate, additional material must be provided (e.g., in appendices) in sufficient detail to allow a clear and precise judgement to be made of the importance and originality ofthe research reported in the thesis.

5. In general, when co-authored papers are included in a thesis the candidate must have made a substantial contribution to aU papers included in the thesis. In addition, the candidate is required to make an explicit statement in the thesis as to who contributed to such work and to what extent. This statement should appear in a single section entitled "Contributions ofAuthors" as a preface to the thesis. Preface 11

This thesis is prepared as a result of the experimental studies performed by P. Moslemy to evaluate the feasibility of use of encapsulated cell technology for in situ bioaugmentation of contaminated aquifers. Chapter 1 presents an introduction to environmental pollution with petroleum hydrocarbons and various site bioremediation strategies including in situ bioaugmentation. Chapter 1 alsodescribes the concept of bioencapsulation and the application of encapsulated microbial cells to biological degradation of hazardous compounds. An introduction to gellan gum, a natural polymer used for encapsulation of bacteria in this study, along with its physical and chemical characteristics, methodof gelation, and production of gel microbeads is also covered. In Chapter 2 the objectives of this study are presented. The contents of Chapters 3 to 6 are adopted from the manuscripts submitted or to be submitted to scientific journals for publication. The investigations performed to develop a two-phase dispersion technique for encapsulation of bacteria in gellan gum microbeads are presented in Chapter 3. The influence of various process parameters on size distribution of microbeads, and the repeatability in the microbead formation process and particle size measurement are demonstrated. Chapter 4 presents the experiments performed to evaluate the feasibility of transport of encapsulated cell microbeads through porous soil media. Transport of gellan gum microbeads was studied by pulse injection of a microbead suspension into soil columns packed with different grain size classes of gravel and sand. The effect of grain size distribution on the extent of transport and the variation of microbead dispersion with injection time across the soil matrix are illustrated. The transport of gellan gum microbeads was further investigated by intermittent injection of a suspension of microbeads into sand columns. Uniform dispersion of bacterial carriers across the contaminated area of an aquifer is crucial to the successful formation of a bioactive zone. In Chapter 5 the effect of grain size and sorne operation parameters such as injectant concentration and injection time on distribution patterns of microbeads are presented. Preface 11l

The performance of gellan gum-encapsulated bacteria in the biodegradation of gasoline is demonstrated in Chapter 6. The capacity of the encapsulated cells to degrade gasoline under aerobic conditions was evaluated in comparison with free (non­ encapsulated) cells in liquid suspension and saturated soil microcosms. The influence of initial gasoline concentration and encapsulated cell mass loading on the extent and the rate of biodegradation are also illustrated. Chapter 7 highlights the main conclusions of this experimental study. In Chapter 8 the major contributions to the existing knowledge are emphasized. Chapter 9 will end this text with the author' s recommendations for future work.

Contributions of Authors The contents of Chapters 3 to 6 of this thesis are adopted from the manuscripts submitted or to be submitted to scientific journals. The research was conducted by P. Moslemy, under the supervision of Prof. Ronald J. Neufeld and Dr. Serge R. Guiot who are named as co-authors. Dr. Denis Millette provided insight into the design of the soil column experiments in work described in Chapters 4 and 5, and is listed as a co-author. iv

Abstract

Encapsulated cell bioaugmentation is a novel approach to in situ bioremediation of contaminated aquifers. This study was carried out to develop encapsulated cell microbeads of appropriate size suitable for in situ bioaugmentation, and to evaluate the feasibility of such a remediation strategy based on the performance of entrapped cells in the biodegradation of a common groundwater contaminant and the transport of cell carriers through porous soil media. A two-phase dispersion technique, termed emulsification-internal gelation, was developed to encapsulate a gasoline-degrading bacterial consortium in gellan gum microbeads. The influence of emulsion parameters including stirring rate, disperse phase volume fraction, emulsifier concentration, emulsification time, and cell mass loading on size distribution of microbeads was studied. The microbead diameter was controlled within a narrow range of 10 - 50 ~ by selection of appropriate emulsion conditions. A high degree of repeatability in the microbead formation process and particle size measurement was demonstrated. Transport experiments were conducted in horizontal soil columns (5.2 cm id x 110 cm long) packed with different grain size classes of gravel (2 - 16 mm) and sand (0.125 - 2 mm). The transport of cell-free microbeads through soil was first investigated in pulse injection experiments carried out with gravel- and sand-packed columns. The feasibility of the formation of a bioactive zone consisting of encapsulated cells was then evaluated in intermittent injection experiments performed with cell-free microbeads in columns packed with different grain size classes of sand. A suspension of microbeads in artificial groundwater (AGW) was pulsed into a column for 6 h, followed by injection of bead-free AGW for 42 h. In general, the total amount of microbeads traveling across a given section of the column decreased with the decrease of mean grain size. The results of this study demonstrated the feasibility of transport of gellan gum microbeads through a wide range of soil grains, i.e. medium sand to medium gravel (0.25 - 16 mm), across distances up to 110 cm. Abstract v

Intermittent injection experiments were performed to obtain the information on distribution patterns of microbeads across llO-cm sand matrices during extended injection periods. In these studies, a suspension of microbeads in AGW was injected at 0.5 L h- l during intermittent 12-h periods. The extent of transport increased with injection time, varying as a descending function of travel distance. The dispersion of microbeads across the length of sand columns suggested a potential for the formation of a bioactive zone of encapsulated cells across a sandy aquifer with similar grain size distribution and hydrodynamic properties. Based on the quantities of dispersed microbeads, this potential would be higher for a sandy aquifer mainly consisted of medium (0.25 - 0.5 mm), coarse (0.5 - 1 mm), and very coarse (l - 2 mm) sand classes of soil. An enriched bacterial consortium, isolated from a gasoline-polluted site was encapsulated in gellan gum microbeads. The capacity of encapsulated cells to degrade gasoline under aerobic conditions was evaluated in comparison with free (non­

encapsulated) cells. Encapsulated cells (2.6 mgcells g-l bead) degraded over 90% gasoline 1 hydrocarbons (initial gasoline concentration 50 - 600 mg L- ) within 5 - 10 days at 10 oC. Equivalent levels of free cells removed comparable amounts of gasoline (initial

1 concentration 50 - 400 mg L- ) within the same period, but required up to 30 days to

1 degrade the highest level of gasoline tested (600 mg L- ). The reduction of encapsulated

cell mass loading below 2.6 mg cells g-l bead caused a substantial decrease in the extent of biodegradation within a 30-day incubation period. Encapsulated cells dispersed within the porous soil matrix of saturated soil microcosms, demonstrated a reduced performance

1 in the removal of gasoline (initial concentrations of 400 and 600 mg L- ), removing 30 ­ 50% gasoline hydrocarbons, compared to 40 - 60% by free cells, within 21 days of incubation. The results of this study suggested that gellan gum-encapsulated bacterial cells have the potential to be used for biodegradation of gasoline hydrocarbons in contarninated aquifers. VI

Résumé

La bioaugmentation par des cellules microbiennes encapsulées constitue une nouvelle approche de biorestauration in situ des aquifères contaminés. Cette étude a été menée dans le but de développer des microcapsules de cellules avec une dimension appropriée à la bioaugmentation in situ et d'évaluer la faisabilité de cette stratégie de restauration en fonction du transport des billes dans le sol et du rendement des cellules encapsulées dans la dégradation d'un contaminant couramment retrouvé dans les eaux souterraines. Une technique à deux phases de dispersion, nommée émulsion interne-gélification, a été développée pour encapsuler,en microbilles de gomme gellane, un consortium bactérien capable de dégrader l'essence. On a étudié l'influence sur la distribution de taille des microbilles, de paramètres d'émulsion, comme la vitesse de mélange, la fraction volumétrique de la phase dispersée, la concentration d'émulsificateur, le temps d'émulsification ainsi que la quantité de biomasse. Le diamètre des microbilles a pu être maintenu dans une gamme de 10 à 50 !lm en sélectionnant adéquatement les conditions d'émulsion. L'étude du processus de formation des microbilles et la mesure de leur taille ont montré un fort taux de répétabilité. Les expériences de transport ont été réalisées avec des colonnes de sol horizontales (5,2 cm de diamètre interne x 110 cm de longueur) remplies de gravier et de sable de différentes classes (2 - 16 mm et 0,125 - 2 mm, respectivement). Le transport des microbilles à travers le sol a tout d'abord été analysé lors d'expériences d'injection des microbilles sans cellules par impulsion, effectuées avec les colonnes de sable et gravier. La faisabilité de la formation d'une zone bioactive, i.e. une zone contenant des cellules encapsulées, a été évaluée ensuite lors d'expériences d'injection intermittente réalisées avec des microbilles sans cellules dans les colonnes remplies de sable de différente taille. Après une phase d'injection de 6 heures dans la colonne, d'eau souterraine artificielle avec des microbilles en suspension, de l'eau sans microbilles a été introduite dans la colonne pendant 42 h. En général, la quantité totale de microbilles qui traverse une section donnée de la colonne décroît avec la diminution de la taille du grain. Les résultats de cette étude ont montré la faisabilité du transport de microbilles de gomme gellane à Résumé VIl

travers divers types de grains, i.e. du sable moyen au gravier moyen (0,25 - 16 mm), le long d'une colonne de 110 cm. Des expériences d'injection intermittente ont été réalisées afin d'obtenir des données sur les profils de distribution des microbilles dans une matrice de sable sur une distance de 110 cm pendant des périodes d'injection plus longues. Pour cette étude, une suspension de microbilles dans une eau souterraine artificielle a été injectée à 0,5 L h- I pendant des périodes intermittentes de 12 h. La quantité de billes transportées s'accroît avec la durée d'injection et décroît avec l'augmentation de la distance de transport. La dispersion des microbilles le long des colonnes de sable suggère la possibilité de formation d'une zone bioactive de cellules encapsulées dans un aquifère sableux avec une distribution de taille de grains et des propriétés hydrodynamiques similaires. En se basant sur les quantités de microbilles dispersées, ce potentiel serait plus important pour un aquifère sableux principalement composé de sable à grains moyens (0,25 - 0,5 mm), à grains grossiers (1 - 2 mm) et à grains très grossiers Cl - 2 mm). Un consortium enrichi de bactéries, isolées de sites contaminés par de l'essence, a été encapsulé dans des microbilles de gomme gellane. La capacité des cellules encapsulées à dégrader l'essence dans des conditions aérobies a été évaluée en comparant l'activité de dégradation des cellules encapsulées à celle de cellules libres (non encapsulées). Les cellules encapsulées (2,6 mgcellules g-I bille) peuvent dégrader plus de 90 % des hydrocarbures de l'essence (concentration initiale d'essence: 50 - 600 mg LI) en 5 à 10 jours à 10 oc. Pour une même concentration de micro-organismes, les bactéries libres ont été capables de dégrader des quantités comparables d'essence (concentration initiale d'essence: 50 - 400 mg LI) dans le même temps, mais ont requis jusqu'à 30 jours pour dégrader la plus haute concentration d'essence testée (600 mg L- I). La réduction de la masse cellulaire encapsulée en dessous de 2,6 mgcellules g-I a engendré une diminution significative de la biodégradation de l'essence avec une période d'incubation de 30 jours. Les cellules encapsulées dispersées dans la matrice du sol poreux dans des microcosmes de sol saturé a mis en évidence une diminution de l'efficacité de dégradation de l'essence (concentrations initiales de 400 et 600 mg L- I), transformant 30 à 50 % des hydrocarbures de l'essence, contre 40 à 60 % pour les cellules libres en 21 jours d'incubation. Les Résumé Vlll

résultats de cette étude suggèrent que les bactéries encapsulées dans la gomme gellane peuvent être utilisées à des fins de biodégradation des aquifères contaminés à l'essence. IX

Acknowledgements

1 would like to express my sincere gratitude to my supervisor Professor Ronald Neufeld and my co-supervisor Dr. Serge Guiot. Without their constructive supervision and insight this work would never have been possible. 1 wish to appreciate their dedication, motivation, guidance, patience, support, and encouragement throughout this research.

1 would also like to glVe my smcere thanks to the following people at the Environmental Biotechnology Sector of the Biotechnology Research Institute of the National Research Council of Canada (Montreal, Quebec, Canada) for their help and support during the course of this work:

• Mr. Adrian Pilon (Director of Environmental Biotechnology Sector) for admitting me in the sector, and for his continuaI interest with my research. • Dr. Jalal AI-Hawari (Analytical Chemistry Group Leader) for providing the gas chromatography instruments, and for his continuaI encouragement. • Ms. Chantale Beaulieu and Mr. Stephane Deschamps (Analytical Chemistry Group) for their technical assistance in gas chromatographie analysis. • Dr. Geoffrey Sunahara (Applied Ecotoxicology Group Leader) for providing the Coulter Counter instrument, and Ms. Sylvie Rocheleau (Applied Ecotoxicology Group) for her technical assistance in particle size analysis. • Ms. Marie-Josée Levesque, Mr. Jean-Claude Frigon, and Mr. Jerome Breton (Environmental Bioengineering Group) for their technical assistance in preparing laboratory equipment and experimental setups. • Dr. Laleh Yerushlami, Dr. Boris Tartakovsky, Dr. Denis Rho, Dr. Odile Tresse, Dr. Darwin Lyew, and Ms. Sandra Cimpoia (Environmental Bioengineering Group) for their fruitful discussions on several scientific matters, and for their support and persistent encouragement.

Finally, 1 would like to extend my everlasting appreciation to my wife Azin whose love is manifested in so many ways, but particularly in her unstinting patience, understanding, encouragement and support. Without her, none of this would have been possible. x

Table of Contents

Preface

Abstract IV

Résumé VI

Acknowledgements IX

Table of Contents X List of Figures xiv List of Tables xix List of Publications and Conferences xxi

1. Introduction 1 1.1. Environmental Pollution 1 1.2. Environmental Bioremediation 2 1.3. Bioaugmentation 3 lA. Transport of Bacteria through Soil 5 1.5. Bioencapsulation 7 1.6. Use of Bioencapsulation in Environmental Research 8 1.7. Encapsulated Cell Bioaugmentation Il 1.8. Oellan Oum 14 1.9. Oellan Oum Oelation 16 1.10. Microbial Encapsulation via Emulsification-Intemal Oelation 16 References 18

2. Objectives 33

3. Production of Size-Controlled Gellan Gum Microbeads Encapsulating Gasoline Degrading Bacteria 34 3.1. Abstract 35 3.2. Introduction 35 Table ofContents Xl

3.3. Material and Methods 37 3.3.1. Reagents 37 3.3.2. Microorganisms and Growth Medium 38 3.3.3. Microbead Production 39 3.3.4. Gasoline Biodegradation 41 3.4. Results 42 3.5. Discussion 49 3.6. Conclusions 52 Acknowledgements 53 References 54

4. Transport of Gellan Gum Microbeads in Soil Columns of Various Grain Size Distributions 57 4.1. Abstract 58 4.2. Introduction 58 4.3. Materials and Methods 61 4.3.1. Microbead Production 61 4.3.2. Artificial Groundwater 62 4.3.3. Granular Media and Columns 62 4.3.4. Hydrodynamic Properties 65 4.3.5. Microbead Transport in Granular Media 66 4.3.6. Analysis of Gellan Gum Microbeads 68 4.4. Results 68 4.5. Discussion 76 4.6. Conclusions 80 Acknowledgements 80 References 82 Table ofContents XlI

5. Transport of Gellan Gum Microbeads through Sand: An Experimental Evaluation for Encapsulated Cell Bioaugmentation 86 5.1. Abstract 87 5.2. Introduction 87 5.3. Materials and Methods 89 5.3.1. Chemicals 89 5.3.2. Microbead Production 90 5.3.3. Artificial Groundwater 90 5.3.4. Porous Media and Columns 91 5.3.5. Hydrodynamic Properties 93 5.3.6. Transport in Porous Media 94 5.3.7. Analysis of Gellan Gum Microbeads 97 5.4. Results and Discussion 97 5.5. Conclusions 108 Acknowledgements 109 References 110

6. Biodegradation of Gasoline by Gellan Gum-Encapsulated Bacterial Cells 113 6.1. Abstract 114 6.2. Introduction 114 6.3. Materials and Methods 116 6.3.1. Materials 116 6.3.2. Microbial Culture and Growth Medium 116 6.3.3. Encapsulation 118 6.3.4. Biodegradation in Liquid Suspension Microcosms 119 6.3.5. Biodegradation in Soil Microcosms 120 6.3.6. Analytical Techniques 121 6.3.7. Adsorption of Gasoline on Gellan Gum 122 6.3.8. Adsorption of Gasoline on Soil 123 6.3.9. Chemical Stability 123 Table ofContents Xlll

6.4. Results 123 6.5. Discussion 130 Acknowledgements 134 References 135

7. Conclusions 138

8. Summary 141

9. Contributions 148

10. Recommendations 149 xiv

List of Figures

Figure 1.1. Gellan gum repeating unit (Jansson et al., 1983; O'Neill et al., 1983). 15

Figure 1.2. Schematic representation of proposed mode! for the conformation of gellan gum on cooling in the presence of cations (e) that promote gel formation (Robinson et al., 1991). 17

Figure 3.1. Design of reactor and impeller: a) reactor and baffle arrangement, b) quarter-circular paddle. (drawing is not to scale). 40

Figure 3.2. Frequency (e) and cumulative (0) size distribution of gellan gum­ encapsulated mixed bacterial culture microbeads formed using

0.75% gellan gum, 0.06% CaCI2, 8 g wet cells per liter sol, 0.1 % (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm. 43

Figure 3.3. Effect of the emulsifier concentration on mean diameter (e) and span (0) of gellan gum microbeads using disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rp~ M

Figure 3.4. Effect of the disperse phase volume fraction (<1» on mean diameter (e) and span (0) of gellan gum microbeads using 0.1 % (w/w) emulsifier, emulsification time 10 min, and stirring rate 4500 rpm. 45

Figure 3.5. Effect of the cell mass loading on mean diameter (e) and span (0) of gellan gum microbeads using 0.1 % (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm. 46 List ofFigures xv

Figure 3.6. Normal distribution of mean diameter of gellan gum microbes using 0.1 % (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm. 47

Figure 3.7. Biodegradation of gasoline by: (e) encapsulated, (0) free, and (0) deactivated encapsulated mixed bacterial culture at 10 oC; TPH is total petroleum hydrocarbons. Data are average of duplicate mns, and error bars show the average deviation. The mean diameter of encapsulated cell microbeads was 23 /lm. 49

Figure 4.1. Schematic of the packed column setup: (1) Scale, (2) Magnet Stirrer, (3) Microbead Suspension Flask, (4) AGW Recharge Tank, (5) Shut-offValves, (6) Masterflex Pump, (7) Digital ControIler, (8) Sampling Ports, (9) Piezometric Tube, (10) Packed Column, (11) Anti-siphon Tube, (12) Collecting Tank. 64

Figure 4.2. Concentration histories for gellan gum microbeads injected into column C packed with 2 - 4 mm grave!. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient from the granular bed inlet. 70

Figure 4.3. Concentration histories for gellan gum microbeads injected into column D packed with 1 - 2 mm sand. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient from the granular bed inlet. 71

Figure 4.4. Concentration histories for gellan gum microbeads injected into column E packed with 0.5 - 2 mm sand. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient List ofFigures XVI

from the granular bed inlet. 72

Figure 4.5. Variation of total breakthrough (TB) of gellan gum microbeads in the effluent of packed columns vs. time. Curves represent columns A (e), B (_), C (À), D (T), E (0), and F (0). 73

Figure 4.6. Variation offractional collector efficiency (l1t/l1+) for diffusion

(110+)' interception (l1t), and sedimentation (l1s+) mechanisms for soil columns; estimated based on RT model using partic1e density

3 20 of 1,008 kg m- , Hammaker constant of 10- J, fluid properties of water at 10 oC, and flow rates given in Table 4.3. 79

Figure 5.1. Schematic of the packed column setup: (1) Scale, (2) Magnet Stirrer, (3) Inflow Reservoir, (4) Shut-offValves, (5) Masterflex Pump, (6) Digital Controller, (7) Sampling Ports, (8) Piezometric Tubes, (9) Packed Column, (10) Anti-siphon Tube, (11) Outflow Reservoir. 92

Figure 5.2. Concentration histories of gellan gum microbeads in columns A, B, and C during various injection phases. Breakthrough curves are for 5 cm (e), 50 cm (0), and 110 cm (À) downgradient from the bed inlet. 98

Figure 5.3. Variation of hydraulic gradient (Llh/LlI) with time for columns A, B, and C. Curves represent the travel distance within the various sections: 5 - 27.5 cm (e) and 27.5 - 50 cm (0) for column A, 5 - 20 cm (e) and 20 - 35 cm (0) for columns Band C. 101

Figure 5.4. Variation of total breakthrough (TB) of gellan gum microbeads with time for columns A, B, and C during various injection phases. Curves are for 5 cm (e), 20 cm (_),50 cm (À), 80 cm (T), and List ofFigures xvii

115 cm (0) downgradient from the bed inlet. 102

Figure S.S. Variation of retention percentage (RP) of gellan gum microbeads with time for columns A, B, and C during various injection phases. Curves represent the travel distance within the various sections: 5 - 20 cm (e), 20 - 50 cm (.), 50 - 80 cm (À), and 80 - 110 cm (T). 105

Figure 5.6. Variation of collector efficiency (11+) with grain size, estimated

1 based on RT model using Darcy velocity of 0.23 m h- , particle

3 20 density of 1,008 kg m- , Hammaker constant of 10- J, and fluid properties of water at 10°C. Curves represent the particle diameter (in flm): 1 (.),5 (e), 10 (V), 20 (i1), 30 (0), and 40 (0). 107

Figure 5.7. Variation of fractional collector efficiency (11t/11+) for diffusion (llD+/11+), interception (11t/11+), and sedimentation (11s+/11+) mechanisms with particle diameter and grain size. 108

Figure 6.1. Biodegradation of gasoline in liquid medium by free bacteria (open symbols) and encapsulated bacteria (closed symbols) for initial

1 gasoline concentrations of (mg L- ): 100 (0 e), 200 (0.),400 (LlÀ), and 600 (V T). Data are mean of duplicate fUns, and error bars show the average deviation from the mean value. Encapsulated cell mass

1 loading is 2.6 mgcells g-l bead. Cell concentration is 0.26 gcells L- MSM in both free and encapsulated cell systems. 124

Figure 6.2. Specifie degradation rate (sdr) of gasoline for free (0) and eneapsulated (e) baeteria as a function of initial total petroleum hydrocarbons (TPH) concentration in the liquid phase. Encapsulated cell mass

1 loading is 2.6 mgcells g-l bead. Cell concentration is 0.26 gcells L- MSM in both free and encapsulated cell systems. 125 List ofFigures xviii

Figure 6.3. Biodegradation of gasoline in liquid medium by encapsulated

bacteria for cell mass loading values of (mgcells g-l bead): 10.4 (e), 2.6 (.), 1.0 (.Â.), and 0.26 (T). Data are mean of duplicate runs, and error bars show the average deviation from the mean value.

1 Initial gasoline concentration is 400 mg L- • Cell concentration is 1.04,0.26,0.10, and 0.026 gcells L- 1 MSM in systems inoculated with encapsulated cells at 10.4,2.6, 1.0, and 0.26 mgcells g-l bead, respectively. 126

Figure 6.4. Biodegradation of gasoline in saturated soil by free bacteria (0)

and encapsulated bacteria for cell mass loading values of (mgcells g-l bead): 2.6 (e), 1.3 (.), and 0.26 (.Â.). Data are mean of duplicate runs, and error bars show the average deviation from the mean

1 value. Initial gasoline concentration is (mg L- ): 400 (a) and 600 (b). Cell concentration is 0.52 gcells L- 1 MSM in free cell systems, and is 0.52, 0.26, and 0.052 gcells L- 1 MSM in systems inoculated with

encapsulated cells at 2.6, 1.3, and 0.26 mgcells g-l bead, respectively. 129

Figure 6.5. Adsorption of gasoline on gellan gum gel (e) and on soil (0). 130 xix

List of Tables

Table 1.1. Characteristics of microbial supports for environmental applications. 9

Table 1.2. Examples of the use of encapsulated microbial cells in biodegradation studies. 10

Table 1.3. Advantages and limitations of using encapsulation for biodegradation of hazardous compounds. 12

Table 1.4. Soil classification based on grain size (Morris and Johnson, 1967). 13

Table 3.1. Characteristics of a commercial gasoline. 38

Table 3.2. Variations in size distribution of gellan gum microbeads with stirring rate, using disperse phase volume fraction 0.143 and emulsification time 10 min. 43

Table 3.3. Variations in size distribution of gellan gum microbeads with emulsification time, using disperse phase volume fraction 0.143 and stirring rate 4500 rpm. 46

Table 3.4. Statistical characteristics of triplicate batches of encapsulated MBC microbeads; (~) population mean, and (cr) standard deviation. 48

Table 4.1. Grain size distribution in packed columns based on the standard soil classification of the US Geological Survey (USGS). 63

Table 4.2. Characteristics of the granular media in packed columns. 65 List afTables xx

Table 4.3. Operational conditions of injection experiments in packed columns. 67

Table 4.4. Total breakthrough (TB) and sectional total breakthrough (sTB) of gellan gum microbeads in soil colurnns after 48-h injection. 74

Table 4.5. Retention (RET) and sectional retention (sRET) of gellan gum microbeads in soil columns after 48-h injection. 75

Table 5.1. Grain size distribution in packed colurnns based on the standard soil classification of the US Geological Survey (USGS). 91

Table 5.2. Characteristics of the porous media in sand colurnns. 94

Table 5.3. Operational conditions of injection experiments in packed columns. 96

Table 5.4. Characteristic size parameters of suspended microbeads across column C. 103

Table 6.1. Characteristics of a commercial gasoline. 117

Table 6.2. Initial phase distribution of gasoline hydrocarbons in liquid and saturated soil microcosms. 120

Table 6.3. Variations of volumetrie degradation rate (vdr) and specifie degradation rate (sdr) of gasoline with encapsulated cell mass Ioading and initial gasoline concentration. 127 XXI

List ofPublications and Conferences

I. Publications Moslemy, P., Guiot, S. R, and Neufeld, R J. (2002) Production of size-controlled gellan gum microbeads encapsulating gasoline-degrading bacteria. Enzyme and Microbial Technology 30: 10-18. Moslemy, P., Neufeld, R J., Millette, D., and Guiot, S. R (submitted July 2001). Transport of gellan gum microbeads through sand: An experimental evaluation for encapsulated cell bioaugmentation. Advances in Environmental Research. Moslemy, P., Neufeld, R J., and Guiot, S. R (submitted October 2001) Biodegradation of gasoline by gellan gum-encapsulated cell cultures. Biotechnology and Bioengineering. Moslemy, P., Millette, D., Guiot, S. R, and Neufeld, R J. (to be submitted). Transport of gellan gum microbeads in soil columns of various grain size distributions. Journal of Environmental Engineering.

II. Conference Presentations Moslemy, P., Millette, D., Guiot, S. R., and Neufeld, R J. (1998). Production and application of gellan gum microbeads for in-situ bioremediation of groundwater. In: International Workshop on Bioencapsulation VII, November 20 - 23, Easton, Maryland, USA. Moslemy, P., Millette, D., Guiot, S. R., and Neufeld, R J. (1999). Transport of gel microbeads through columns of aquifer material for in-situ bioaugmentation. In: International Workshop on Bioencapsulation VIII, September 13 - 15, Trondheim, Norway. Moslemy, P., Guiot, S. R., Millette, D., and Neufeld, R J. (2000). Laboratory investigation of forced-gradient transport of microbeads through natural granular media. In: 50th Canadian Chemical Engineering Conference, October 15 - 18, Montreal, Quebec. 1

1. Introduction

1.1. Environmental Pollution Release of crude oil and various petroleum products from leaking underground storage tanks, accidentaI spills, distribution systems, industrial operations, and routine or illegal discharges are common sources of contamination to aquatic and terrestrial environments (Atlas, 1981; Blumer et al., 1971; Chapelle, 1999; Day et al., 2001; McKee et al., 1972). It was estimated that 1.7 - 8.8 million metric tons of petroleum, equivalent to 0.08 - 0.4% of the world annual production, pollute the oceans each year (Bartha, 1986; Brown, 1987). Although no comparable estimates were available for terrestrial petroleum hydrocarbon pollution, it was stated that the extent of contamination is likely to equal if not to exceed the figure cited for the marine environment since the major part of petroleum production, refining, distribution, and utilization activities take place on land. Petroleum liquid products including gasoline, naphtha, kerosene, fuel oils, and lubricating oils are comprised of a variety of light and heavy aliphatic, alicyclic, and aromatic hydrocarbon components (Speight, 1999). The release of petroleum products such as gasoline from leaking underground storage tanks is a ubiquitous source ofcontamination to soil and groundwater (Atlas and Cerniglia, 1995; Cherry, 1987; Day et al., 2001). In Canada alone, there were approximately 70,000 underground storage tanks located at petroleum retail outlets, of which about 20 - 25% were found to be leaking (Cherry, 1987). In the United States, more than 10% of the 3.5 million petroleum product storage tanks were leaking and had caused at least 300,000 environmental accidents (Dowd, 1984). Gasoline is a light distillate and may contain C4 to Cu alkane, cycloalkane, alkene, and aromatic hydrocarbons (Hancock, 1985). These compounds are sufficiently soluble in water to pose a major treat to groundwater quality. Aromatic components such as benzene, toluene, ethylbenzene and xylenes (ortho-, meta-, and para-xylene), collectively referred to as BTEX, comprise about 20 - 30% of commercial gasoline. The BTEX compounds are typically the targets of regulatory concern among the many constituents of petroleum products. The US EPA has 1. 1ntroduction 2

classified them as hazardous waste (Cookson, 1995). The BTEX compounds are relatively water-soluble and mobile compared with other components of petroleum (Fan and Krishnamurthy, 1995). Therefore, even a small discharge of gasoline has the potential to render large portions of groundwater unfit for drinking. Gasoline hydrocarbon components may be transformed from the fuel liquid phase to the air, water, and soil phases by a variety of physical, chemical, and biological processes that can occur in the subsurface. Upon the release of gasoline into a soil environment, a minor fraction of gasoline hydrocarbons may be subject to evaporative and photodegradative losses. However, the majority of gasoline hydrocarbons will be subject to vertical and horizontal infiltration at a rate determined primarily by the stratification and permeability of the soil materials. Infiltrated hydrocarbons migrate by gravity downward through unsaturated soillayers (vadose zone) and may eventually reach the soil water table (capillary zone) where they form a floating light non-aqueous phase liquid (LNAPL). A small portion of LNAPL is dissolved in groundwater (saturated zone), forming a contaminant plume which spreads within the direction of bulk groundwater flow. Residual LNAPL may spread out laterally and act as a long-term source of groundwater contamination.

1.2. Environmental Bioremediation Bioremediation is the use of biological processes to degrade or destroy hazardous chemical contaminants in soil, water, and air systems. Bioremediation is applied principaIly to raise intrinsic degradation rates to significantly higher values by mitigating environmental rate­ limiting factors. It had been known since the 1940s that sorne microorganisms are capable of degrading petroleum hydrocarbons (ZobeIl, 1946). However, it was only after the accidentaI release of massive volumes of gasoline and crude oil into aquatic environments in the late 1960s (Atlas, 1981) and the early 1970s (Chapelle, 1999) that bioremediation was actively considered as a remedial strategy to clean up contaminated groundwater and marine systems. The largest bioremediation project involved the removal of oil pollutants from hundreds of kilometers of impacted rocky shorelines, following the spill of more than 32,000 cubic meters (200,000 barrels) of crude oil from the oil tanker Exxon Valdez in 1. Introduction 3

Prince William Sound, Alaska in March 1989 (Prince, 1993). Biodegradation of petroleum hydrocarbons has been studied extensively in both laboratory and field trials. Several investigators have reviewed various physicochemical, microbiological, and technological aspects of petroleum hydrocarbon biodegradation observed in these trials (Atlas, 1981; Bartha, 1986; Chapelle, 1999; Gruiz and Kriston, 1995; Head and Swannell, 1999; Leahy and Colwell, 1990; Morgan et al., 1993; Morgan and Watkinson, 1989). Over the years, a wide variety of engineered in situ and ex situ bioremediation technologies have been deve10ped (Cookson, 1995; Suthersan, 1997). These technologies are mainly designed to increase the availability of electron acceptors (oxygen or nitrate), electron donors (organic substrates), nutrients (nitrogen, phosphorous, or potassium), or degrading microbial communities. The biodegradation of gasoline hydrocarbons as the common environmental pollutants, have been explored in a multitude of laboratory investigations (Gupta and Tao, 1996; Solano-Serena et al., 1998; Solano-Serena et al., 2000a; Solano-Serena et al., 2000b; Solano-Serena et al., 2000c; Solano-Serena et al., 1999; Yerushalmi and Guiot, 1998; Yerushalmi et al., 1999; Zhou and Crawford, 1995) and field studies (Lahvis et al., 1999; Maura Jutras et al., 1997; Phelps and Young, 1999; Walters et al., 1994; Wright et al., 1997). Aromatic components of gasoline (BTEX) have also been the main focus of several biodegradation studies (Chang et al., 1993; Gibson et al., 1998; Kelly et al., 1996; Mallakin and Ward, 1996; Morgan et al., 1993; Salanitro et al., 1997; Shim and Yang, 1999). The microbial degradation of hydrocarbons is strongly influenced by physical and chemical factors such as temperature, oxygen, nutrients, salinity, water activity, pH, and concentration and availability of the contaminant, and also by biological factors such as the composition, adaptability, and physiological capabilities of the microbial populations.

1.3. Bioaugmentation Although it is weIl established that petroleum hydrocarbon compounds will be biologically degraded under optimal environmental conditions, rate-limiting factors may reduce the biodegradation rate of hydrocarbons. These factors can be classified as environmental and microbiological factors. Accordingly, there are two principal approaches for 1. Introduction 4

bioremediation. The first relies on the metabolic capacities of indigenous microbial populations, and the mitigation of environmental rate-limiting factors such as oxygen or nutrient supply, temperature, and contaminant bioavailability. Bioremediation is therefore accomplished by environmental modification or biostimulation, for example through aeration or fertilizer application, increasing the temperature, or use of surfactants to overcome factors that limit the rate of hydrocarbon biodegradation by the indigenous microorganisms. The second approach is generally referred to as bioaugmentation, which is defined as the process of adding specifieally adapted exogenous microorganisms to a system to stimulate biodegradation and enhance the rate and/or extent of biodegradation. The rationale for this approach is that the indigenous mieroorganisms may not be capable of degrading the wide range of potential substrates present in such complex mixtures as petroleum (Atlas, 1977), or that the development of an indigenous population of degrading mieroorganisms in a heavily contaminated environment may be suppressed by the high concentration of toxic contaminants (Aamand et al., 1995; Forsyth et al., 1995). Bioaugmentation has been proved to be a complementary remediation approach where the biostimulation alone is not suffieient to trigger the biological activity of indigenous microbial populations, or where the time to generate sufficient populations is uncertain and possibly long (Baud-Grasset and Vogel, 1995; Ellis et al., 2000; Salanitro et al., 2000). Furthermore, in situations where a bioremediation process depends on cometabolism, introducing microorganisms with unique and specialized metabolic capabilities can enhance the rates and extents of cometebolism (Bouchez et al., 1995). For in situ bioaugmentation of contaminated groundwater systems, the added microorganisms should be maintained strictly at activity levels that are high enough to enable control of degradation pathways (Dybas et al., 1998; Witt et al., 1999). Intrinsic biodegradation of contaminants may result in incomplete removal of contaminants or transformation to toxic byproducts. Therefore, a potential advantage of bioaugmentation with active microorganisms is to control metabolic pathways such that the biodegradation leads to formation of non-toxie metabolites (Ellis et al., 2000; Witt et al., 1999). One of the drawbacks to bioaugmentation with exogenous microorganisms is competition with indigenous microfiora for available substrate and nutrient resources. The 1. Introduction 5

rate and extent of biodegradation can be increased if the augmented microbial populations are able to survive and express their hydrocarbon-degradation activities in the environments to which they are added. The added microorganisms should outcompete the indigenous microorganisms and have superior oil-degrading capacities (Atlas, 1977). Addition of a microbial suspension may mobilize and leach contaminants from the unsaturated zone into the water within the weIl casing and into the groundwater, leading to a dramatic increase in dissolved contaminant concentration within the immediate vicinity of an injection weIl (Maxwell and Baqai, 1995). Consequently, the high levels of toxic contaminants may impose an inhibitory effect, causing death or deactivation of injected microorganisms at early stages of bioaugmentation. Another potential drawback to bioaugmentation largely reflects the hydrologic difficulties inherent in delivering microorganisms efficiently to contaminated aquifer sediments (Maxwell and Baqai, 1995; Thomas and Ward, 1994). In one field-scale bacterial transport study in a sandy aquifer, a large number (> 99%) of radiolabelled indigenous bacterial strain PL2W31 of low adhesion was retained within 0.5 m from the injection weIl, and only a small percentage of bacteria traveled across the 4-m flow field (DeFlaun et al., 1997). In another study, it was reported that the transport and recovery of a specific strain of Pseudomonas fluorescens 5R bacteria injected into a sandy aquifer, was more erratic than that of a dissolved non-reactive tracer (Burlage et al., 1995). It was suggested that this observed behavior was due to complex patterns of bacterial attachment to sediment grain surfaces.

1.4. Transport of Bacteria through Soil Bacterial transport is a fundamental process in delivery of degrading microorganisms to contaminated sites for bioremediation. Dissemination of added bacteria is often intervened by a microbial sorption process, leading to the attachment of cells to soil grains. The extent of microbial attachment is a function of cell type and its adhesion properties (Simoni et al., 1998), cell concentration (Camesano and Logan, 1998), chemical composition and ionic strength of suspending fluid (Simoni et al., 2000), fluid velocity (Camesano and Logan, 1998), soil mineralogy (Knapp et al., 1998) and heterogeneity (Bolster et al., 1999), and i. introduction 6

temperature (McCaulou et al., 1995). It has been shown that attachment to sand surfaces stimulates synthesis of exopolymers in subsurface bacterial isolates (Dennis and Turner, 1998; Sharp et al., 1999; Vandevivere and Kirchman, 1993). The growth of attached bacteria and development of interstitial biofilms within short distances of the injection point can cause a permanent decrease in soil permeability (Jennings et al., 1995; Taylor and Jaffé, 1990), leading to pore clogging and failure of the bioaugmentation process. In view of the proposed use of bacteria to block pore spaces in oil reservoirs in order to enhance oil recovery (Brown et al., 1985), the risk of blockage and consequent inhibition of water movement is a critical aspect of bioaugmentation. A number of remedial strategies have been undertaken by researchers to overcome the microbial attachment and to enhance the transport of microorganisms through soil. The reduction of ionic strength of the carrier fluid and the addition of nonionic surfactants substantially reduced the bacterial attachment to glass, quartz, and soil surfaces for Alcaligenes paradoxus culture, and to a lesser degree for a subsurface isolate, within 1-cm long mini-columns (Li and Logan, 1999). However, these treatments were stated to be insufficient to increase the transport distance of cells for field applications. In another study, reporting the influence of an anionic biosurfactant on the transport of three strains of Psuedomonas aeruginosa with different cell hydrophobicity, about 50 - 80% of the injected bacteria were immobilized within 5-cm long sand columns due to sorption (Bai et al., 1997). Alternatively, the use of specialized degrading bacteria with limited adhesion to aquifer solids has been practiced for in situ bioaugmentation. In a field-scale trial, a specialized bacterial strain, Burkholderia cepacia ENV435, capable of degrading chlorinated ethenes, traveled the 2 m distance from the injection to the recovery well in a semi-confined silty sand aquifer, leading to high groundwater cell concentrations of 2.2 x

7 8 1 10 - 1.9 x 10 cfu mL- (Steffan et al., 1999). The transport of B. cepacia cells was, however, subject to severe filtration effect by the soil matrix, leading to several order-of­ magnitude reductions in suspended cell concentration within longer distances. Despite these promising results, the use of exogenous microorganisms for bioaugmentation is subject to local regulations and may not be a feasible bioremediation approach in general. 1. Introduction 7

Furthermore, the isolation of and application of indigenous strains of low adhesion with high degrading capacity and high survival rate characteristics may not be always possible.

1.5. Bioencapsulation Bioencapsulation technology has revolutionized the traditional methods of application of chemical and biological formulations in a broad spectrum of science and engineering fields. The concept of encapsulation has originated perhaps with an inspiration from the morphology and fine structure of singie-celled organisms. Natural cell-capsular membranes are remarkably successful in fulfilling specifie functions. Among the most important functions are protection of the interior material (core) and control of the flow of materials (permeation) across the cell membrane. The earliest encapsulated product was developed by Barrett Green, a chemist from Dayton, Ohio (USA) who invented the first carbonless copy paper in 1940. Green's investigations on 'coacervation' techniques led him to the idea of coupling the preparation technique of solid gelatin spheres with inclusion of an oil phase within the gelatin coacervate. Thereby, Green prepared the first gelatin microcapsules in 1942. Through his leadership, the first marketable copy paper that required no carbon interleaves was developed for business forms in the early 1950's (Fanger, 1974).' Since then, the concept of encapsulation has been developed in different ways and for many different applications. Bioencapsulation can be defined as the physical confinement (immobilization) of biologically active materials (e.g. cells and enzymes) to a certain defined region of polymerie supports with the preservation of sorne desired activity or reactivity. Encapsulation is classified into a number of categories (Karel et al., 1985; Willaert and Baron, 1996) among which the 'gel entrapment' has been widely applied to environmental applications (Cassidy et al., 1996; McLoughlin, 1994; Trevors et al., 1992). Gel entrapment refers to the confinement of cells (encapsulants) within or throughout a polymerie matrix. A comprehensive review of materials and methods, mass transfer and modeling, microbial physiology, and applications of gel entrapment systems is presented in a previous report (Willaert and Baron, 1996). An encapsulated cell system can be divided into three components: the ceIls, the support material, and the solution that fills the porous 1. Introduction 8

matrix of the supporL The space around the encapsulated cells is called the 'microenvironment' since the chemical properties of the interstitial solution may be quite different from those of the bulk solution. Both technical and economic advantages of bioencapsulation technology have resulted in an enormous amount of research for its scientific and industrial applications during the past decades (Cassidy et al., 1996; Champagne et al., 1994; Groboillot et al., 1994; Hutchison, 1993; Park and Chang, 2000).

1.6. Use ofBioencapsulation in Environmental Research Increased awareness of profound effects of environmental problems associated with subsurface contamination and industrial wastewater disposaI has stimulated investigations of technologies, which avoid, reduce or eliminate these problems. Bioaugmentation of subsurface polluted environments (soil and groundwater) or surface bioreactors with pre­ isolated microorganisms capable to degrade certain pollutants is an alternate remedy for decontamination of sites or industrial waste effluents (Pritchard, 1992; Vogel, 1996). Successful application of living microorganisms to bioaugmentation schemes depends on inoculum density and formulation (i.e. liquid vs. solid), mode of application (single or multiple introductions), mode of operation, biotic and abiotic environmental effects, rate of survival, and the extent of distribution and transport of added microorganisms (through the soil matrix). Both biotic and abiotic factors play critical roles in determining the survival of introduced microorganisms. In subsurface applications, biological factors include predation by protozoans, lower level of starvation resistance of introduced microbes, and lack of suitable soil niches for extended cell survival. Moisture content, pH, texture, and oxygen and nutrients availability are abiotic factors controlling the survival of microorganisms. Encapsulation has emerged as a promising solution to overcome practical limitations of using free cell formulations. The polymerie matrix of the support material provides a defined, stable, consistent, and protective microenvironment, where cells can survive and metabolic activity can be maintained for extended periods of time without the immediate release of large number of cells. The entrapped cells can better tolerate numerous environmental stresses, and may be released after adaptation to surrounding environmental conditions. The main characteristics of microbial supports for 1. Introduction 9

environmental applications are summarized in Table 1.1. A variation of these characteristics may be considered in the selection of an appropriate microbial support for a specifie application. For instance, micrometer-sized carriers are required for subsurface injection (Petrich et al., 1995) while macrometer-sized supports may be desired for surface bioreactor applications or for constructing subsurface permeable reactive barriers (Razavi­ Shirazi and Veenstra, 2000).

Table 1.1. Characteristics of microbial supports for environmental applications.

Shape Size (surface area-to-volume ratio) Porosity Toxicity Biocompatibility Cell mass loading capacity Retention capacity Performance quality Stability (chemical, biological, mechanical, and thermal) Ease of production Reusability Longevity Cost (economic feasibility)

The use of encapsulated cells for biodegradation of hazardous compounds has been explored in numerous studies, as listed in Table 1.2. However, most of these investigations have been performed at the laboratory scale, and practical uses of encapsulated cells in the open environment have yet to be realized. Both natural and synthetic polymer gels have been used for microbial encapsulation. Criteria for gel supports to be used for soil applications are different from those required for controlled bioreactor systems. Although the long-term stability, low porosity, and small pore sizes afforded by synthetic polymers can be effective for use in bioreactors, they may not be desirable in open environmental 1. Introduction 10

Table 1.2. Examples of the use of encapsulated microbial cells in biodegradation studies.

Compound Microorganism Carrier Reference

Petroleum

Crude oil Mixed culture Urea-formaldehyde (Mohn, 1997) Crude oil Yarrowia lipolytica Agar-Alginate (Zinjarde & Pant, 2000) Crude oil Yarrowia lipolytica Polyurethane (Oh et al., 2000)

Aliphatic Hydrocarbons

Hexadecane Mixed culture Urea-formaldehyde (Mohn, 1997)

n-Alkanes (CI4-C I6) Prototheca zopfi Alginate (Suzuki et al., 1998)

Monoaromatic Hydrocarbons (MAHs)

Benzene Pseudomonas putida Polyacrylamide (Somerville et al., 1977) Benzene Rhodococcus sp. Alginate (Paje et al., 1998) 2,4-Diaminotoluene Mixed culture Chitosan (Veenstra & Subramanian, 1994)

Polyaromatic Hydrocarbons (PAHs)

Naphthalene Pseudomonas sp. Alginate (Manohar & Karegoudar, 1998) Phenanthrene Pseudomonas sp. Alginate (Weir et al., 1995) Phenanthrene Pseudomonas sp. Alginate (Weir et al., 1996) Phenanthrene Mixed culture Urea-formaldehyde (Mohn, 1997)

Phenolic Compounds

Phenol Candida tropicalis Alginate, Polystyrene, (Hackel et al., 1975) Polyacrylamide Phenol Candida tropicalis Polyacrylamide, (Klein & Schara, 1981) Polymethacrylamide Phenol Pseudomonas sp. Alginate, Polyacrylamide (Bettmann & Rehm, 1984) Phenol Mixed culture Agar (Dwyer et al., 1986) Phenol Mixed culture Alginate (Lee et al., 1994) Phenol Rhodococcus sp. Alginate (Pai et al., 1995) Phenol Mixed culture Alginate (Shishido et al., 1995) Phenol Pseudomonas putida Alginate (Aksu & Bülbül, 1998) Phenol Mixed culture Tetraethoxysilane (Bninyik et al., 1998) Phenol Pseudomonas putida Alginate (Mordocco et al., 1999) Phenol Pseudomonas putida Chitosan (Annadurai et al., 2000) Phenol Trichosporon cutaneum Polyacrylonitrile (Godjevargova et al., 2000) Phenol Pseudomonas putida Alginate (Banerjee et al., 2001) p-Cresol Pseudomonas sp. Alginate (O'Reilly et al., 1988) p-Cresol Pseudomonas sp. Alginate (O'Reilly and Crawford, 1989) Phenol, 0- & p-Cresol Anaerobie sludge Alginate (Tawfiki Hajji et al., 2000) p-Nitrophenol Moraxella sp. K-Carrageenan (Errampalli et al., 1999) 1.1ntroduction Il

applications. The pore size of the support applied to a bioreactor process should be much smaller than the encapsulated cell to avoid or minimize cell release and washout from the bioreactor. Thus, cells remain inside the support while substrates and nutrients can diffuse in and products can diffuse out. In field applications, however, supports with high porosity are desirable since they provide high cell mass loading as weIl as high diffusional mass transfer rates. Therefore, the use of natural polymers with relatively larger pore size and greater degree of biodegradability may be suitable for sorne subsurface applications. It may be desired to release the encapsulated cells after positioning them in the target zone, and after adaptation to the surrounding environment (McLoughlin, 1994). In general, natural polymers are recommended for use in soil (Trevors et al., 1992). The characteristics of encapsulated inoculants and their advantages over free cells for use in soil have been discussed comprehensively in earlier reports (Cassidy et al., 1996; Trevors et al., 1992). A list of advantages and limitations of encapsulated cell systems with respect to their importance in bioaugmentation applications is outlined in Table 1.3.

1.7. Encapsulated Cell Bioaugmentation It was mentioned that the attachment of bacteria to soil grain surfaces is a major impediment to in situ bioaugmentation of contaminated aquifers. Encapsulation of bacteria within polymeric gel microbeads may enhance transport distances for bioaugmentation to occur over a large region of the subsurface. Encapsulation provides a secluded microenvironment to the bacteria, preventing thei( attachment to soil grains while protecting them from the exterior biotic and abiotic stresses. Applying the encapsulated microorganisms to groundwater requires production of microbeads that can be introduced hydraulically into the porous matrix of a contaminated aquifer. A suspension of microbeads is injected into the saturated zone of aquifer through the use of an appropriate number of injection/withdrawal wells installed across the path of the contaminant plume. The number of and the distance between these wells are to be determined as a function of various parameters including contaminant plume width, site geological formation and hydrodynamic characteristics, injectant concentration, injection time, and flow rate. The induced gradient formed by the network of wells forces the microbeads to transport 1. Introduction 12

Table 1.3. Advantages and limitations of using encapsulation for biodegradation of hazardous compounds.

Advantages: Protection from biotic and abiotic environmental stresses Increased microbial survival and metabolic activity Increased cell mass loading Co-encapsulation of nutritional additives Preferential cell growth through the aerobic and anaerobic zones within the gel matrix Ability to catalyze a linked series of reactions Sustained release of cells as a form of continuous inoculant Sustained release of chemicals such as nutrients and oxidizers Prevention or reduction of off-site drift during field applications Reduced possibility of inoculum contamination during storage, transport and application Use of non-toxic, biodegradable, and non-polluting carriers Producible in large quantities Storable for extended periods Usable with existing mechanical application equipment Limitations: CeIlleakage from the gel matrix Possible restriction of gas and solute diffusion Possible reduced oxygen consumption rates Possible cell morphological or metabolic alterations Possible partialloss of initial activity upon encapsulation Effects of changes in water activity on survival of microbes Initial cost of production

through the soil interstices, leading to the dispersion of microbeads across the contaminant plume, and thus creating a biological barrier of highly concentrated and active microorganisms. Hence, the contamination passing through the barrier is degraded by the encapsulated microorganisms, leading to in situ bioremediation of the groundwater. Transport of microbeads is controlled by the filtration effect of the soil matrix, which is influenced by several factors including the carrier fluid, the suspension of particles, and the porous medium characteristics. A number of previous reports (BaIes et al., 1997; Harmand et al., 1996; Harvey et al., 1989; Harvey et al., 1993; Jegatheesan and 1. Introduction 13

Vingeswaran, 1997; Kau and Lawler, 1995; Moran et al., 1993a; Moran et al., 1993b; Petrich et al., 1995; Petrich et al., 1998; Reddi and Bonala, 1997) have discussed the effect of grain and particle size distributions, and granular media heterogeneity and depth on transport of particles through granular media. Monodisperse micrometer- and submicrometer-sized synthetic polymer microspheres have been mostly used in deep-bed filtration studies, or as tracers of bacteria in laboratory or field transport studies. However, there has been little effort to investigate the transport of cell encapsulating gel carriers for subsurface bioaugmentation. Grain size distribution is one of the important parameters influencing the particle transport and dispersion through the soil matrix. A standard soil classification system devised by the US Geological Survey (USGS) is presented in Table lA (Morris and Johnson, 1967).

Table 1.4. Soil classification based on grain size (Morris and Johnson, 1967).

Material Grain Size (mm) Clay < 0.004 Silt 0.004 - 0.062 Very fine sand 0.062 - 0.125 Fine sand 0.125 - 0.25 Medium sand 0.25 - 0.5 Coarse sand 0.5 - 1 Very coarse sand 1 - 2 Very fine gravel 2-4 Fine gravel 4-8 Medium gravel 8 - 16 Coarse gravel 16 - 32 Very coarse gravel 32 - 64

A semi-empirical model developed by Arya and his coworkers, translates the particle size distribution of an assemblage of uniformly sized spherical particles in cubic 1. Introduction 14

packing into its pore size distribution (Arya and Dierolf, 1989; Arya and Paris, 1981; Reddi and Bonala, 1997):

r;~( ~:;) (LI) where ri is the pore radii, Ri is the particle radii, e is the porosity, and a* is a parameter representing the effective pore length associated with each particle. It was found that an average value of 9.11 mm for a* provides an excellent agreement between predicted and measured values of pore radii for soils ranging from very coarse sands to silts (Arya and Dierolf, 1989). An unconsolidated aquifer may be composed of severallayers of gravel and sand within the 2 - 16 mm and 0.125 - 2 mm ranges, respectively. According to this model, the interstitial pore size may range from 8 !lm for a grain size of 0.125 mm to 480 !lm for a grain size of 2 mm. Therefore, bacterial carriers lying within this range should be able to be transported through the sand and gravel media.

1.8. Gellan Gum Gellan gum is an extracellular polysaccharide produced through aerobic fermentation processes by the microorganism Sphingomonas paucimobilis ATCC 31461 (Giavasis et al., 2000), earlier referred to as Pseudomonas elodea (Kang and Veeder, 1982; Kang and Veeder, 1983), Auromonas elodea (Robinson et al., 1991), and Sphingomonas elodea (Sworn and Kasapis, 1998). The chemical structure of gellan gum is shown in Figure 1.1. Gellan gum is made up of repeating tetrasaccharide units consisting of a linear sequence of D-glucose, D-glucuronic acid, D-glucose, and L-rhamnose (Jansson et al., 1983; O'Neill et al., 1983). The use of gellan gum as a safe, non-toxic gelling material with no physiological effects and high resistant to enzymatic breakdown has been proposed for a variety of encapsulation-based applications. Complex coacervation of gelatin-gellan gum mixtures was described for the microencapsulation of sunflower oil, paraffin oil, and solid materials such as aluminum powder (Chilvers and Morris, 1987). Gellan gum was used as a thermoresistant gelling polysaccharide for the fermentation of whey or whey permeate with encapsulated bacteria Thermus aquaticus at high temperatures of 60 - 80 oC (Norton and 1. Introduction 15

Lacroix, 1990). In a later study, the feasibility of using gellan gum as an entrapment matrix for a mesophilic lactic acid bacterium, Bifidobacterium longum ATCC 15707, was investigated (Camelin et al., 1993). High mechanical stability as weIl as high biocatalyst activity was obtained. Gellan gum has been also used to formulate sustained release drug delivery as weIl as herbicide delivery systems (Santucci et al., 1996). The in vitro release of drugs such as theophylline and benzamide, and a weed control agent, metribuzin, from gellan gum beads was studied.

COOH o o o o

OH OH OH OH OH

...3).~.D.Glcp.(1-4).~.D.GlcpA.(1-4).~.D.Glcp.(1-4).a.L.Rhap.(1...

Figure 1.1. Gellan gum repeating unit (Jansson et al., 1983; O'Neill et al., 1983).

The mechanism of gelation and texture of gellan gum suggests a strong similarity with agar and carrageenans. However, gellan gum gel has superior rheological properties to agar and carrageenan gels at equivalent concentrations (Sanderson et al., 1989), and therefore, it can be used at substantially lower concentrations. The gel has been stable over the wide pH range of 2 - 10 (Ashtaputre and Shah, 1995), suggesting its suitability for use in both acidic and basic environments. The application of gellan gum for encapsulation of viable cells has been addressed at considerably lower concentrations of both gel and gelling agent compared to K-carrageenan, agar, and alginate (Buitelaar et al., 1988; Nilsson et al., 1983). Unlike sorne other ion-sensitive gelling polysaccharides such as alginate and 1. Introduction 16

K-carrageenan, the reactivity between gellan gum and ions is non-specifie and gels can be formed with a wide variety of cations including alkaline and alkaline-earth cations (Moorehouse et al., 1981; Sanderson and Clark, 1983). A combination of these characteristics makes gellan gum a premium natural polymer for encapsulation of active microorganisms with promising performance in environmental subsurface applications.

1.9. Gellan Gum Gelation It has been proposed that upon gelation of gellan gum, the polymerie chains form extended, intertwined, three-fold, left-handed, parallel double-helical structures (Chandrasekaran et al., 1988). At high temperatures between 75 and 90 oC, gellan gum exists in solution (sol) as disordered coils (Figure 1.2). On cooling in the presence of appropriate cations, such as sodium or calcium, they convert to double-helical structures in which a proportion of helices are able to associate into cation-mediated aggregates which crosslink the gel network (Robinson et al., 1991). The sol-gel transition temperature varies between 20 and 60 oC, shifting to higher temperatures with increasing gellan gum and gelling agent concentrations (Miyoshi et al., 1994; Moritaka et al., 1991; Moritaka et al., 1992). Divalent cations such as calcium and magnesium ions are more efficient than monovalent cations such as sodium and potassium ions, and provide improved rheological properties (Sanderson and Clark, 1983).

1.10. Microbial Encapsulation via Emulsification-Internal Gelation Emulsion techniques have been utilized for the encapsulation of viable cells within micro­ (0.1 - 1 mm) and macrobeads (1 - 3 mm) of natural thermotropic gel polymers such as agar (Nilsson et al., 1983), agarose (Knaebel et al., 1996; Nilsson et al., 1983), and K­ carrageenan (Audet and Lacroix, 1989; Nilsson et al., 1983). The bead formation process involves the dispersion of two immiscible liquid phases resulting in a water-in-oil (W/O) emulsion. A suspension of viable cells in an aqueous solution of the polymer (disperse phase) is emulsified in a hydrophobie phase such as a vegetable oil (continuous phase). .The gelation of the small droplets of the dispersed phase is subsequently initiated by decreasing the emulsion temperature below the sol-gel transition temperature. 1. Introduction 17

Coils

Figure 1.2. Schematic representation of proposed model for the conformation of gellan gum on cooling in the presence of cations (e) that promote gel formation (Robinson et al., 1991).

Other researchers have used emulsion techniques, termed as emulsification-internal gelation, to encapsulate biocatalysts (Poncelet et al., 1992; Poncelet et al., 1995), and DNA (Alexakis et al., 1995) in alginate microspheres. The gelation of alginate droplets was triggered by gentle acidification (to pH 6.5) of the water-oil dispersion through adding an oil-soluble acid and releasing soluble calcium ions from a salt complex. In the present study, the emulsification-internal gelation method was used to encapsulate a bacterial consortium in gellan gum microbeads. Calcium is used as the gelling agent and is available inside gellan gum sol droplets to induce an internaI gelation. 1. 1ntroduction 18

References Aamand, J., Bruntse, G., Jepsen, M., J0rgensen, C. and Jensen, B. K. (1995). Degradation of PAHs in Soil by Indigenous and Inoculated Bacteria. In: Bioaugmentation for Site Remediation. Hinchee, R. E., Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 121-127. Aksu, Z. and Bülbül, G. (1998). Investigation of the combined effects of external mass transfer and biodegradation rates on phenol removal using immobilized P. putida in a packed-bed column reactor. Enzyme and Microbial Technology 22: 397-403. Alexakis, T., Boadi, D. K., Quong, D., Groboillot, A., O'Neill, 1., Poncelet, D. and Neufeld, R. J. (1995). Microencapsulation of DNA within alginate microspheres and crosslinked chitosan membranes for in vivo application. Applied Biochemistry and Biotechnology 50: 93-106. Annadurai, G., Rajesh Babu, S., Mahesh, K. P. O. and Murugesan, T. (2000). Adsorption and bio-degradation of phenol by chitosan-immobilized Pseudomonas putida (NICM 2174). Bioprocess Engineering 22: 493-501. Arya, L. M. and Dierolf, T. S. (1989). Predicting soil moisture characteristics from particle-size distributions: An improved method to calculate pore radii from particle radii. International Workshop on Indirect Methods for Estimating the Hydraulic Properties of Unsaturated Soils, The University of California, Riverside, CA, 115­ 124. Arya, L. M. and Paris, J. F. (1981). A physicoempirical model to predict the soil moisture characteristic from particle-size distribution and bulk density data. Soil Science Society of America Journal 45: 1023-1030. Ashtaputre, A. A. and Shah, A. K. (1995). Studies on viscous, gel-forming exopolysaccharide from Sphingomonas paucimobilis GS 1. Applied and Environmental Microbiology 61: 1159-1162. Atlas, R. M. (1977). Stimulated petroleum biodegradation. CRC Critical Reviews in Microbiology 5: 371-386. Atlas, R. M. (1981). Microbial degradation of petroleum hydrocarbons: An environmental perspective. Microbiological Reviews 45: 180-209. 1. Introduction 19

Atlas, R. M. and Cerniglia, C. E. (1995). Bioremediation of petroleum pollutants: Diversity and environmental aspects of hydrocarbon biodegradation. BioScience 45: 332-338. Audet, P. and Lacroix, C. (1989). Two-phase dispersion process for the production of biopolymer gel beads: Effect of various parameters on bead size and their distribution. Process Biochemistry December: 217-226. Bai, G., Brusseau, M. L. and Miller, R. M. (1997). Influence of a rhamnolipid biosurfactant on the transport of bacteria through a sandy soil. Applied and Environmental Microbiology 63: 1866-1873. BaIes, R. c., Li, S., Yeh, T.-C. J., Lenczewski, M. E. and Gerba, C. P. (1997). Bacteriophage and microsphere transport in saturated porous media: Forced-gradient experiment at Bordon, Ontario.Water Resources Research 33: 639-648. Banerjee, L, Modak, J. M., Bandopadhyay, K., Das, D. and Maiti, B. R. (2001). Mathematical model for evaluation of mass transfer limitations in phenol biodegradation by immobilized Psuedomonas putida. Journal of Biotechnology 87: 211-223. Bartha, R. (1986). Biotechnology of Petroleum Pollutant Biodegradation. Microbial Ecolog'y 12: 155-172. Baud-Grasset, F. and Vogel, T. M. (1995). Bioaugmentation: Biotreatment of Contaminated Soil by Adding Adapted Bacteria. In: Bioaugmentation for Site Remediation. Hinchee, R. E., Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 39-48. Bettmann, H. and Rehm, H. J. (1984). Degradation of phenol by polymer entrapped microorganisms. Applied Microbiology and Biotechnology 20: 285-290.. Blumer, M., Sanders, H. L., Grassle, J. F. and Hampson, G. R. (1971). A small oil spill. Environment 13: 2-12. Boister, C. H., Mills, A. L., Hornberger, G. M. and Herman, J. S. (1999). Spatial distribution of deposited bacteria following miscible displacement experiments in intact cores. Water Resources Research 35: 1797-1807. Bouchez, M., Blanchet, D. and Vandecasteele, J.-P. (1995). Degradation of polycyclic aromatic hydrocarbons by pure strains and by defined strain associations: inhibition 1. Introduction 20

phenomena and cometabolism. Applied Microbiology and Biotechnology 43: 156­ 164. Branyik, T., Kuncova, G., Paca, J. and Demnerova, K. (1998). Encapsulation of microbial cells into silica gel. Journal of Sol-Gel Science and Technology 13: 283-287. Brown, L. R. (1987). Oil-degrading microorganisms. Chemical Engineering Progress: 35­ 40. Brown, M. J., Moses, V., Robinson, J. P. and Springham, D. G. (1985). Microbial enhanced oil recovery: Progress and prospects. CRC Critical Reviews in Biotechnology 3: 159-197. Buitelaar, R. M., Hulst, A. C. and Tramper, J. (1988). Immobilization of biocatalysts in thermogels using the resonance nozzle for rapid drop formation and an organic solvent for gelling. Biotechnology Techniques 2: 109-114. Burlage, R. S., Palumbo, A. V. and McCarthy, J. (1995). Signal Quantification of Bacteria for a Groundwater Transport Experiment. In: Bioaugmentation for Site Remediation. Hinchee, R. E., Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 139-144. Camelin, 1., Lacroix, c., Paquin, c., Prevost, H., Cachon, R. and Divies, C. (1993). Effect of chelants on gellan gum rheological properties and setting temperature for immobilization of living Bifidobacteria. Biotechnology Progress 9: 291-297. Camesano, T. A. and Logan, B. E. (1998). Influence of fluid velocity and cell concentration on the transport of motile and nonmotile bacteria in porous media. Environmental Science and Technology 32: 1699-1708. Cassidy, M. B., Lee, H. and Trevors, J. T. (1996). Environmental applications of immobilized microbial cells: A review. Journal of Industrial Microbiology 16: 79­ 101. Champagne, C. P., Lacroix, C. and Sodini-GaIlot, 1. (1994). Immobilized cell technologies for the dairy industry. Critical Reviews in Biotechnology 14: 109-134. Chandrasekaran, R., Millane, R. P. and Arnott, S. (1988). The crystal structure of gellan. Carbohydrate Research 175: 1-5. i. introduction 21

Chang, M.-K., Voice, T. C. and Criddle, C. S. (1993). Kinetics of competitive inhibition and cometabolism in the biodegradation of benzene, toluene, and p-xylene by two Pseudomonas isolates. Biotechnology and Bioengineering 41: 1057-1065. Chapelle, F. H. (1999). Bioremediation of petroleum hydrocarbon-contaminated ground water: The perspectives of history and hydrology. Ground Water 37: 122-132. Cherry, J. A (1987). Groundwater Occurrence and Contamination in Canada. Canadian Aquatic Resources 215: 387-426. Chilvers, G. R. and Morris, V. J. (1987). Coacervation of gelatin-gellan gum mixtures and their use in microencapsulation. Carbohydrate Polymers 7: 111-120. Cookson, J., J. T. (1995). Bioremediation Engineering: Design and Application. McGraw­ Hill, New York. Day, M. J., Reinke, R. F. and Thomson, J. A. M. (2001). Fate and transport of fuel components below slightly leaking underground storage tanks. Environmental Forensics 2: 21-28. DeFlaun, M. F., Murray, C. J., Holben, W., Scheibe, T., Mills, A, Ginn, T., Griffin, T., Majer, E. and Wilson, J. L. (1997). Preliminary observations on bacterial transport in a coastal plain aquifer. FEMS Microbiology Reviews 20: 473-487. Dennis, M. L. and Turner, J. P. (1998). Hydraulic conductivity of compacted soil treated with biofilm. Journal of Geotechnical and Geoenvironmental Engineering 124: 120­ 127. Dowd, R. M. (1984). Leaking underground storage tanks. Environmental Science and Technology 18: 309A Dwyer, D. F., Krumme, M. L., Boyd, S. A and Tiedje, J. M. (1986). Kinetics of phenol biodegradation by an immobilized methanogenic consortium. Applied and Environmental Microbiology 52: 345-351. Dybas, M. J., Barcelona, M., Bezborodnikov, S., Davies, S., Forney, L., Heuer, H., Kawka, O., Mayotte, T., Sepulveda-Torres, L., Smalla, K., Sneathen, M., Tiedje, J., Voice, T., Wiggert, D. c., Witt, M. E. and Criddle, C. S. (1998). Pilot-scale evaluation of bioaugmentation for in-situ remediation of a carbon tetrachloride-contaminated aquifer. Environmental Science and Technology 32: 3598-3611. 1. Introduction 22

Ellis, D. E., Lutz, E. J., Odom, J. M., Buchanan, R. J. J., Bartlett, C. L., Lee, M. D., Harkness, M. R. and Deweerd, K. A. (2000). Bioaugmentation for accelerated in situ anaerobic bioremediation. Environmental Science and Technology 34: 2254-2260. Errampalli, D., Tresse, O., Lee, H. and Trevors, J. T. (1999). Bacterial survival and mineralization of p-nitrophenol in soil by green fluorescent protein-marked Moraxella sp. G21 encapsulated cells. FEMS Microbiology Ecology 30: 229-236. Fan, C. y. and Krishnamurthy, S. (1995). Enzymes for enhancing bioremediation of petroleum contaminated soils: A brief review. JAPCA - The Journal of the Air & Waste Management Association 45: 453-460. Fanger, G. O. (1974). Microencapsulation, A Brief History and Introduction. In: Microencapsulation: Processes and Applications. Vandegaer, J. E. (Ed.), Plenum Press, New York, 1-20. Forsyth, J. V., Tsao, Y. M. and Bleam, R. D. (1995). Bioremediation: When Is Augmentation Needed? In: Bioaugmentation for Site Remediation. Hinchee, R. E., Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 1-14. Giavasis, 1., Harvey, L. M. and McNeil, B. (2000). Gellan gum. Critical Reviews in Biotechnology 20: 177-211. Gibson, T. L., Abdul, A. S. and Chalmer, P. D. (1998). Enhancement of in situ bioremediation of BTEX-contaminated ground water by oxygen diffusion from silicone tubing. Ground Water Monitoring and Remediation 18: 93-104. Godjevargova, T., Aleksieva, Z. and Ivanova, D. (2000). Cell immobilization of Trichosporan cutaneum strain with phenol degradation ability on new modified polymer carriers. Process Biochemistry 35: 699-704. Groboillot, A., Boadi, D. K., Poncelet, D. and Neufeld, R. J. (1994). Immobilization of cells for application in the food industry. Critical Reviews in Biotechnology 14: 75­ 107. Gruiz, K. and Kriston, E. (1995). In situ bioremediation of hydrocarbon in soil. Journal of Soil Contamination 4: 163-173. Gupta, G. and Tao, J. (1996). Bioremediation of gasoline-contaminated soil using poultry litter. Journal of Environmental Science and Health-Part A 31: 2395-2407. 1. Introduction 23

Hackel, D., Klein, J., Megnet, R. and Wagner, F. (1975). Immobilisation of Microbial Cells in Polymeric Matrices. European Journal of Applied Microbiology 1: 291-293. Hancock, E. G. (1985). The Manufacture of Gasoline and the Chemistry of Its Components. In: Technology of Gasoline. Hancock, E. G. (Ed.), Blackwell Scientific Publications, Oxford, 20-56. Harmand, B., Rodier, E., Sardin, M. and Dodds, J. (1996). Transport and capture of submicron particles in natural sand: Short column experiments and a linear mode!. Colloids and Surfaces. A: Physicochemical and Engineering Aspects 107: 233-244. Harvey, R. W., George, L. H., Smith, R. L. and LeBlanc, D. R. (1989). Transport of microspheres and indigenous bacteria through a sandy aquifer: Results of natural­ and forced-gradient tracer experiments. Environmental Science and Technology 23: 51-56. Harvey, R. W., Kinner, N. E., MacDonald, D., Metge, D. W. and Bunn, A. (1993). Role of physical heterogeneity in the interpretation of small-scale laboratory and field observations of bacteria, microbial-sized microsphere, and bromide transport through aquifer sediments. Water Resources Research 29: 2713-2721. Head, 1. M. and Swannell, R. P. J. (1999). Bioremediation of petroleum hydrocarbon contaminants in marine habitats. CUITent Opinion in Biotechnology 10: 234-239. Hutchison, K. (1993). Encapsulation in softgels for pharmaceutical advantage. Special Publications of the Royal Society of Chemistry 138: 86-97. Jansson, P. E., Lindberg, B. and Sandford, P. A. (1983). Structural studies of gellan gum, an extracellular polysaccharide elaborated by Pseudomonas elodea. Carbohydrate Research 124: 135-139. Jegatheesan, V. and Vingeswaran, S. (1997). Interaction between organic substances and submicron particles in deep bed filtration. Separation and Purification Technology 12: 61-66. Jennings, D. A., Petersen, J. N., Skeen, R. S., Hooker, B. S., Peyton, B. M., Johnstone, D. L. and Yonge, D. R. (1995). Effects of slight variations in nutrient loadings on pore plugging in soil columns. Applied Biochemistry and Biotechnology, Part A­ Enzyme Engineering and Biotechnology 51/52: 727-734. 1. Introduction 24

Kang, K. S. and Veeder, G. T. (1982). Polysaccharide S-40 and bacterial fermentation process for its preparation. US Pat. 4,326,053. Kang, K. S. and Veeder, G. T. (1983). Fermentation process for its preparation of polysaccharide S-40. US Pat. 4,377,636. Karel, S. F., Libicki, S. B. and Robertson, C. R. (1985). The immobilization of whole cells: Engineering principles. Chemical Engineering Science 40: 1321-1354. Kau, S. M. and Lawler, D. F. (1995). Dynamics of deep-bed filtration: Velocity, depth, and media. Journal of Environmental Engineering 121: 850-859. Kelly, W. R., Hornberger, G. M., Herman, 1. S. and Mills, A. L. (1996). Kinetics of BTX biodegradation and mineralization in batch and column systems. Journal of Contaminant Hydrology 23: 113-132. Klein, J. and Schara, P. (1981). Entrapment of living cells in covalent polymeric networks. II. A quantitative study on the kinetics of oxidative phenol degradation by entrapped Candida tropicalis cells. Applied Biochemistry and Biotechnology 6: 91-117. Knaebel, D. B., Stormo, K. E. and Crawford, R. L. (1996). Immobilization of bacteria in macro- and microparticles. In: Bioremediation Protocols. Sheehan, D. (Ed.), Humana Press, Inc., Totowa, New Jersey, 67-78. Knapp, E. P., Herman, J. S., Hornberger, G. M. and Mills, A. L. (1998). The effect of distribution of iron-oxyhydroxide grain coatings on the transport of bacterial cells in porous media. Environmental Geology 33: 243-248. Lahvis, M. A., Baehr, A. L. and Baker, R. J. (1999). Quantification of aerobic biodegradation and volatilization rates of gasoline hydrocarbons near the water table under natural attenuation conditions. Water Resources Research 35: 753-765. Leahy, J. G. and Colwell, R. R. (1990). Microbial degradation ofhydrocarbons in the environment. Microbiological Reviews 54: 305-315. Lee, C.M., Lu, C. J. and Chuang, M. S. (1994). Effects of immobilized cells on the biodegradation of chlorinated phenols. Water Science and Technology 30: 87-90. Li, Q. and Logan, B. E. (1999). Enhancing bacterial transport for bioaugmentation of aquifers using low ionic strength solutions and surfactants. Water Research 33: 1090­ 1100. 1. Introduction 25

Mallakin, A. and Ward, O. P. (1996). Degradation of BTEX compounds in liquid media and in peat biofilters. Journal of Industrial Microbiology 16: 309-318. Manohar, S. and Karegoudar, T. B. (1998). Degradation of naphthalene by cells of Pseudomonas sp. strain NGK 1 immobilized in alginate, agar and polyacrylamide. Applied Microbiology and Biotechnology 49: 785-792. Maura Jutras, E., Smart, C. M., Rupert, R., Pepper, 1. L. and Miller, R M. (1997). Field­ scale biofiltration of gasoline vapors extracted from beneath a leaking underground storage tank. Biodegradation 8: 31-42. Maxwell, C. Rand Baqai, H. A. (1995). Remediation of Petroleum Hydrocarbons by Inoculation with Laboratory-Cultured Microorganisms. In: Bioaugmentation for Site Remediation. Hinchee; RE., Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 129-137. McCaulou, D. R, BaIes, R C. and Arnold, R G. (1995). Effect of temperature-controlled motility on transport of bacteria and microspheres through saturated sediment. Water Resources Research 31: 271-280. McKee, J. E., Laverty, F. B. and Hertel, R M. (1972). Gasoline in Groundwater. Journal of the Water Pollution Control Federation 44: 293-302. McLoughlin, A. J. (1994). Controlled release of immobilized cells as a strategy to regulate ecological competence of inocula. Advances in Biochemical EngineeringIBiotechnology 51: 1-45. Miyoshi, E., Takaya, T. and Nishinari, K. (1994). Gel-sol transition in gellan gum solutions. II. DSC studies on the effects of salts. Food Hydrocolloids 8: 529-542. Mohn, W. W. (1997). Indirect bioremediation: biodegradation of hydrocarbons on a commercial sorbent. Biodegradation 8: 15-19. Moorehouse, R, Colegrove, G. T., Stanford, P. A., Baird, J. K. and Kang, K. S. (1981). PS-60: A new gel-forming polysaccharide. In: Solution Properties of Polysaccharides. Brant, D. A. (Ed.), American Chemical Society, Washington DC, 111-124. 1. Introduction 26

Moran, D. C., Moran, M. c., Cushing, R. S. and Lawler, D. F. (1993a). Particle behavior in deep-bed filtration: Part l-Ripening and breakthrough. American Water Works Association Journal 58: 69-81. Moran, D. c., Moran, M. c., Cushing, R. S. and Lawler, D. F. (1993b). Particle behavior in deep-bed filtration: Part 2-Particle detachment. American Water Works Association Journal 85: 82-93. Mordocco, A., Kuek, C. and Jenkins, R. (1999). Continuous degradation of phenol at low concentration using immobilized Pseudomonas putida. Enzyme and Mierobial Technology 25: 530-536. Morgan, P., Lewis, S. T. and Watkinson, R. J. (1993). Biodegradation ofbenzene, toluene, ethylbenzene and xylenes in gas-condensate-contaminated ground-water. Environmental Pollution 82: 181-190. Morgan, P. and Watkinson, R. J. (1989). Hydrocarbon degradation in soils and methods for soil biotreatment. CRC Critical Reviews in Biotechnology 8: 305-333. Moritaka, H., Fukuba, H., Kumeno, K., Nakahama, N. and Nishinari, K. (1991). Effect of monovalent and divalent cations on the rheological properties of gellan gels. Food Hydrocolloids 4: 495-507. Moritaka, H., Nishinari, K., Nakahama, N. and Fukuba, H. (1992). Effects of potassium chloride and sodium chloride on the thermal properties of gellan gum gels. Bioscience, Biotechnology and Biochemistry 56: 595-599. Morris, D. A. and Johnson, A. 1. (1967). Summary of Hydrologie and Physical Properties of Rock and Soil Materials, as Analyzed by the Hydrologic Laboratory of the D.S. Geological Survey 1948-1960, D.S. Geological Survey Water Supply Paper 1839-D. Nilsson, K., Birnmaum, S., Flygare, S., Linse, L., Schroder, u., Jeppsson, D., Larsson, P.­ O., Mosbach, K. and Brodelius, P. (1983). A general method for the immobilization of cells with preserved viability. European Journal of Applied Microbiology and Biotechnology 17: 319-326. Norton, S. and Lacroix, C. (1990). Gellan gum gel as entrapment matrix for high temperature fermentation processes-Rheological study. Biotechnology Techniques 4: 351-356. 1. Introduction 27

O'Neill, M. A., Selvendran, P. R and Morris, V. J. (1983). Structure of the acidic extracellular gelling polysaccharide produced by Pseudomonas elodea. Carbohydrate Research 124: 123-133. O'Reilly, K. T. and Crawford, R L. (1989). Kinetics of p-Cresol degradation by an immobilized Pseudomonas sp. Applied and Environmental Mierobiology 55: 866­ 870. O'Reilly, K. T., Kadakia, R, Korus, R A. and Crawford, R. L. (1988). Utilization of immobilized-bacteria to degrade aromatic compounds common to wood-treatment wastewaters. Water Science and Technology 20: 95-100. Oh, Y.-S., Maeng, J. and Kim, S.-J. (2000). Use of microorganism-immobilized polyurethane foams to absorb and degrade oil on water surface. Applied Microbiology and Biotechnology 54: 418-423. Pai, S.-L., Hsu, Y.-L., Chong, N.-M., Sheu, c.-S. and Chen, C.-H. (1995). Continuous degradation of phenol by Rhodococcus sp. immobilized on granular activated carbon and in calcium alginate. Bioresource Technology 51: 37-42. Paje, M. L., Marks, P. and Couperwhite, 1. (1998). Degradation of benzene by a Rhodococcus sp. using immobilized cell systems. World Journal of Microbiology and Biotechnology 14: 675-680. Park, J. K. and Chang, H. N. (2000). Microencapsulation of microbial cells. Biotechnology Advances 18: 303-319. Petrieh, C. R, Stromo, K. E., Knaebel, D. B., Ralston, D. R and Crawford, R L. (1995). A Preliminary Assessment of Field Transport Experiments Using Encapsulated Cells. In: Bioaugmentation for Site Remediation. Hinchee, R, Fredrickson, J. and Alleman, B. C. (Eds.), Battelle Press, Columbus, OH, 237-244. Petrich, C. R, Stromo, K. E., Ralston, D. R and Crawford, R L. (1998). Encapsulated cell bioremediation: Evaluation on the basis of particle tracer tests. Ground Water 36: 771-778. Phelps, C. D. and Young, L. Y. (1999). Anaerobie biodegradation ofBTEX and gasoline in various aquatic sediments. Biodegradation 10: 15-25. 1. Introduction 28

Poncelet, D., Lencki, R, Beaulieu, c., Halle, J. P., Neufeld, R J. and Fournier, A. (1992). Production of alginate beads by emulsification/internal gelation. 1. Methodology. Applied Microbiology and Biotechnology 38: 39-45. Poncelet, D., Poncelet De Smet, B., Beaulieu, c., Huguet, M. L., Fournier, A. and Neufeld, R J. (1995). Production of alginate beads by emulsification/internal gelation. II. Physicochemistry. Applied Microbiology and Biotechnology 43: 644-650. Prince, R C. (1993). Petroleum spill bioremediation in marine environments. Critical Reviews in Microbiology 19: 217-242. Pritchard, P. H. (1992). Use of inoculation in bioremediation. Current Opinion in Biotechnology 3: 232-243. Razavi-Shirazi, F. and Veenstra, J. N. (2000). Development of a biological permeable barrier to remove 2,4,6-trichlorophenol from groundwater using immobilized cells. Water Environment Research 72: 460-468. Reddi, L. N. and Bonala, M. V. S. (1997). Analytical solution for fine partic1e accumulation in soil filters. Journal of Geotechnical and Geoenvironmental Engineering 123: 1143-1152. Robinson, G., Manning, C. E. and Morris, E. R. (1991). Conformation and Physical Properties of the Bacterial Polysaccharides Gellan, Welan, and Rhamsan. In: Food Polymers, Gels, and Colloids. Dickinson, E. (Ed.), The Royal Society of Chemistry, Cambridge, England, 22-33. Salanitro, J. P., Johnson, P. C., Spinnler, G. E., Maner, P. M., Wisniewski, H. L. and Bruce, C. (2000). Field-scale demonstration of enhanced MTBE bioremediation through aquifer bioaugmentation and oxygenation. Environmental Science and Technology 34: 4152-4162. Salanitro, J. P., Wisinewski, H. L., Byers, D. L., Neaville, C. C. and Schroder, R A. (1997). Use of aerobic and anaerobic microcosms to access BTEX biodegradation in aquifers. Ground Water Monitoring and Remediation 17: 210-221.

Sanderson, G. R, Bell, V. L. and Ortega, D. (1989). A comparison of gellan gum, agar, l(­ carrageenan, and algin. Cereal Foods World 34: 991-'998. Sanderson, G. R and Clark, R C. (1983). Gellan gum. Food Technology 37: 62-70. 1. Introduction 29

Santucci, E., Alhaique, F., Carafa, M., Coviello, T., Murtas, E. and Riccieri, F. M. (1996). Gellan for the formulation of sustained delivery beads. Journal of Controlled Release 42: 157-164. Sharp, R. R., Cunningham, A. B., Komlos, J. and Billmayer, J. (1999). Observation of thick biofilm accumulation and structure in porous media and corresponding hydrodynamic and mass transfer effects. Water Science and Technology 39: 195­ 201.

Shim, H. and Yang, S.-T. (1999). Biodegradation of benzene, toluene, ethylbenzene, and 0­ xylene by a coculture of Pseudomonas putida and Pseudomonas fluorescens immobilized in a fibrous-bed bioreactor. Journal of Biotechnology 67: 99-112. Shishido, M., Kojima, T., Araike, Y.-K. and Toda, M. (1995). Biological phenol degradation by immobilized activated sludge in gel bead with three-phase fluidized­ bed bioreactor. Transactions of the Institute of Chernical Engineers 73A: 719-726. Simoni, S. F., Bosma, T. N. P., Harms, H. and Zehnder, A. J. B. (2000). Bivalent cations increase bath the subpopulation of adhering bacteria and their adhesion efficiency in sand columns. Environmental Science and Technology 34: 1011-1017. Simoni, S. F., Harms, H., Bosma, T. N. P. and Zehnder, A. J. B. (1998). Population heterogeneity affects transport of bacteria through sand columns at low flow rates. Environmental Science and Technology 32: 2100-2105. Solano-Serena, F., Marchal, R., Blanchet, D. and Vandecasteele, J.-P. (1998). Intrinsic capacities of soil microflora for gasoline degradation. Biodegradation 9: 319-326. Solano-Serena, F., Marchal, R., Casarégola, S., Vasnier, c., Lebeault, J.-M. and Vandcasteele, J.-P. (2000a). A Mycobacterium strain with extended capacities for degradation of gasoline hydrocarbons. Applied and Environmental Microbiology 66: 2392-2399. Solano-Serena, F., Marchal, R., Huet, T., Lebealt, J.-M. and Vandecasteele, J.-P. (2000b). Biodegradability of volatile hydrocarbons of gasoline. Applied Microbiology and Biotechnology 54: 121-125. 1. Introduction 30

Solano-Serena, F., Marchal, R, Lebeault, J.-M. and Vandcasteele, J.-P. (2000c). Distribution in the environment of degradative capacities for gasoline attenuation. Biodegradation 11: 29-35. Solano-Serena, F., Marchal, R., Ropars, M., Lebeault, J.-M. and Vandecasteele, J.-P. (1999). Biodegradation of gasoline: kinetics, mass balance and fate of individual hydrocarbons. Journal of Microbiology 86: 1008-1016. Somerville, H. 1., Mason, 1. Rand Ruffell, RN. (1977). Benzene degradation by bacterial cells immobilized in polyacrylamide gel. European Journal of Applied Microbiology 4: 75-85. Speight, J. G. (1999). The Chemistry and Technology of Petroleum. Marcel Dekker, Inc., New York. Steffan, R J., Sperry, K. L., Walsh, M. T., Vainberg, S. and Condee, C. W. (1999). Field­ scale evaluation of in situ bioaugmentation for remediation of chlorinated solvents in groundwater. Environmental Science and Technology 33: 2771-2781. Suthersan, S. S. (1997). Remediation Engineering: Design Concepts. CRC Press, Inc., Boca Raton, FL. Suzuki, T., Yamaguchi, T. and Ishida, M. (1998). Immobilization of Prototheca zopfi in calcium-alginate beads for the degradation of hydrocarbons. Process Biochemistry 33: 541-546. Sworn, G. and Kasapis, S. (1998). Effect of conformation and molecular weight of co­ solute on the mechanical properties of gellan gum gels. Food Hydrocolloids 12: 283­ 290. Tawfiki Hajji, K., Lépine, F., Biasaillon, J.-G., Beaudet, R, Hawari, J. and Guiot, S. R (2000). Effects of bioaugmentation strategies in UASB reactors with a methanogenic consortium for removal of phenolic compounds. Biotechnology and Bioengineering 67: 417-423. Taylor, S. W. and Jaffé, P. R (1990). Biofilm growth and the related changes in the physical properties of a porous medium: 1. Experimental investigation. Water Resources Research 26: 2153-2159. 1. Introduction 31

Thomas, J. M. and Ward, C. H. (1994). Introduced Organisms for Subsurface Bioremediation. In: Handbook of Bioremediation. Norris, R. D., Hinchee, R. E., Brown, R., McCarty, P. L., Semprini, L., Wilson, J. T., Kampbell, D. H., Reinhard, M., Bouwer, E. J., Borden, R. C., Vogel, T. M., Thomas, J. M. and Ward, C. H. (Eds.), Lewis Publishers, Boca Raton, FL, 227-244. Trevors, J. T., van Elsas, J. D., Lee, H. and van Overbeek, L. S. (1992). Use of alginate and other carriers for encapsulation of microbial cells for use in soil. Microbial Releases 1: 61-69. Vandevivere, P. and Kirchman, D. L. (1993). Attachment simulates exopolysaccharide synthesis by a bacterium. Applied and Environmental Microbiology 59: 3280-3286. Veenstra, J. N. and Subramanian, S. (1994). Biological treatment of a simulated industrial wastewater using chitosan-immobilized activated sludge. The 49th Purdue Industrial Waste Conference, Lewis Publishers, Chelsea, MI, 519-532. Vogel, T. M. (1996). Bioaugmentation as a soil bioremediation approach. CUITent Opinion in Biotechnology 7: 311-316. Walters, M. D., Harrison, J. c., Ott, D. E. and Reiter, P. F. (1994). In-situ bioremediation of gasoline contaminated groundwater and soils: A practical approach. The 49th Purdue Industrial Waste Conference, Lewis Publishers, Chelsea, MI, 57-69. Weir, S. c., Dupuis, S. P., Providenti, M. A., Lee, H. and Trevors, J. T. (1995). Nutrient­ enhanced survival of and phenanthrene mineralization by alginate-encapsulated and free Pseudomonas sp. UG14Lr cells in creosote-contaminated soil slurries. Applied Microbiology and Biotechnology 43: 946-951. Weir, S. c., Providenti, M. A., Lee, H. and Trevors, J. T. (1996). Effect of alginate encapsulation and selected disinfectants on survival of and phenanthrene mineralization by Pseudomonas sp. UG14Lr in creosote-contaminated sail. Journal of Industrial Microbiology 16: 62-67. Willaert, R. G. and Baron, G. V. (1996). Gel entrapment and micro-encapsulation: Methods, applications and engineering principles. Reviews in Chemical Engineering 12: 1-205. 1. Introduction 32

Witt, M. E., Dybas, M. J., Wiggert, D. C. and Criddle, C. S. (1999). Use of bioaugmentation for continuous removal of carbon tetrachloride in model aquifer columns. Environmental Engineering Science 16: 475-485. Wright, W. F., Schroeder, E. D., Chang, D. P. Y. and Romstad, K. (1997). Performance of a pilot-scale compost biofilter treating gasoline vapor. Journal of Environmental Engineering 123: 547-555. Yerushalmi, L. and Guiot, S. R. (1998). Kinetics of biodegradation of gasoline and its hydrocarbon constituents. Applied Microbiology and Biotechnology 49: 475-481. Yerushalmi, L., Manuel, M. F. and Guiot, S. R. (1999). Biodegradation of gasoline and BTEX in a microaerophilic biobarrier. Biodegradation 10: 341-352. Zhou, E. and Crawford, R. L. (1995). Effects of oxygen, nitrogen, and temperature on gasoline biodegradation in soi!. Biodegradation 6: 127-140. Zinjarde, S. S. and Pant, A. (2000). Crude oil degradation by free and immobilized cells of Yarrowia lipolytica NCIM 3589. Journal of Environmental Science and Health-Part A-Toxic Hazardous Substances 35: 755-763. Zobell, C. E. (1946). Action of microorganisms on hydrocarbons. Bacteriological Reviews 10: 1-49. 33

2. Objectives

The application of encapsulated cells to in situ bioaugmentation of contaminated aquifers necessitates the production of micrometer-sized polymerie carriers with high biocatalytic activity. Transport of cell carriers through the porous matrix of soil is also required for successful bioaugmentation of the contaminated aquifer. Therefore, the objectives of this research were defined as follows:

(1) A gasoline-degrading bacterial consortium was to be encapsulated in size-controlled gellan gum microbeads using a two-phase dispersion (emulsion) technique. The effect of various emulsion parameters on size distribution of microbeads was to be studied. These parameters included stirring rate, disperse phase volume fraction, emulsifier concentration, emulsification time, and cell mass loading. The precision or repeatability of the microbead formation process and particle size measurement was to be determined. (2) Transport of gellan gum microbeads through porous soil media was to be evaluated in horizontal columns packed with various soil grain-size distributions. The soil matrices were to be characterized in terms of their hydrodynamic properties as well as grain size distributions. The extent of distribution of microbeads was to be assessed as a function of travel distance, injectant concentration, and injection time. (3) Biodegradation of gasoline by encapsulated cells was to be assessed in comparison with free, non-encapsulated cells. The performance of encapsulated bacteria in the removal of gasoline hydrocarbons was to be studied in liquid suspension and saturated soil microcosms. The effects of initial gasoline concentration and encapsulated cell mass loading on the extent of biodegradation were to be examined. 34

3. Production of Size-Controlled Gellan Gum Microbeads Encapsulating Gasoline Degrading Bacteria

Peyman Moslemy a,b , Serge R. Guiot b , and Ronald J. Neufeld C

a Department of Chemical Engineering, McGill University, 3610 University Street, Montreal, Quebec, Canada H3A 2B2

b Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

C Department of Chemical Engineering, Queen's University, Dupuis Hall, Kingston, Ontario, Canada K7L 3N6

A two-phase dispersion technique was developed in this study to encapsulate a gasoline­ degrading bacterial consortium in gellan gum microbeads. The influence of emulsion parameters on size distribution of microbeads was elucidated, and the precision of this technique to produce size-controlled microbeads was demonstrated. The candidate has carried out aIl the work in this chapter. This article has been published in the journal of Enzyme and Microbial Technology. 3. Production ofSize-Controlled Gellan Gum Microbeads 35 Encapsulating Gasoline Degrading Bacteria

3.1. Abstract Controlling the mean diameter ofpolymeric carriers is crucial to the successful application of encapsulated cells for in situ bioaugmentation of contaminated aquifers. The cell carriers should be small enough to be transported through a granular soil matrix, thus an emulsification-internal gelation technique for production of cell-encapsulating gellan gum microbeads is proposed. Mean diameter and size distribution ofmicrobeads were investigated as a function of the water-in-oil emulsion parameters. The mean diameter of the microbeads ranged from 12 to 135/lIIl, varying as a descending function of the stirring rate (1000 - 5500 rpm) and emulsifier concentration (0 - 0.20% w/w), and as an ascending function of the disperse phase volume fraction (0.08 - 0.25). A bacterial consortium encapsulated within the microbeads (23 Ilm mean diameter) showed improved biodegradation activity in the removal 1 of gasoline (400 mg L- ), as compared to free cells. A high degree of repeatability in the microbead formation process and particle size measurements was demonstrated. The results of this study suggest that the emulsification process can potentially be used for the large-scale production of controlled-diameter gellan gum-encapsulated cell microbeads for subsurface bioremediation applications.

Keywords: Gellan gum; Encapsulation; Emulsification; Microbead; Mixed bacterial culture

3.2. Introduction Encapsulation of microbial cells involves entrapping living microorganisms within a semi­ permeable polymeric gel structure. The use of encapsulated cells for environmental applications has many advantages over free cell formulations, including protection from biotic stresses such as predation by protozoa and bacteriophage (Smit et al., 1996), protection from abiotic stresses such as the inhibitory effect of toxic compounds (Cassidy et al., 1997), enhanced survival and improved physiological activity (Weir et al., 1995), supply of co­ encapsulated nutritional additives (Trevors et al., 1993), increased cell densities (Lee et al., 1994), and preferential cell growth in various internaI aerobic and anaerobic zones of encapsulating gel (Beunink and Rehm, 1990). Encapsulation may also increase transport distances of degrading cells in subsurface bioaugmentation schemes. Microbial attachment to soil grain surfaces followed by formation of extracellular polymers (exopolymers) is a major impediment to successful distribution of free active cells through saturated soils. The various physical, chemical, and biological mechanisms by which microorganisms clog aquifer materials are discussed extensively in an earlier report (Baveye et al., 1998). Encapsulation in polymeric matrices isolates the microbes from the exterior environment and eliminates 3. Production ofSize-Controlled Gellan Gum Microbeads 36 Encapsulating Gasoline Degrading Bacteria chemical and biological patterns of the clogging, leading to enhanced microbial transport through porous soil media. The application of encapsulated degrading cells to bioaugmentation of a contaminated aquifer requires large-scale production of sub-millimeter diameter carriers (microbeads smaller than 50 /lm) to enable hydraulic distribution of encapsulated cells into the granular matrix of the aquifer. Transport of mierobeads through the soil pores can be induced towards the target contaminant plume by a forced-gradient circulating flow system. The encapsulated cells should be eventually immobilized within the contaminated zone and degrade mobile contaminants. The porous matrix of gel microbeads permits adequate diffusion of the contaminants, dissolved gases, nutrients, and metabolic byproducts, leading to decontamination of groundwater. Phase dispersion techniques have been described for encapsulation of living cells within micro- (0.1 - 1 mm) and macrobeads (l - 3 mm) of natural thermotropic gel polymers such as agar (Nilsson et al., 1983), agarose (Knaebel et al., 1996; Nilsson et al., 1983), and K-carrageenan (Audet and Lacroix, 1989; Nilsson et al., 1983). Encapsulation in gelled natural polymers is occasionally favored for various reasons such as high porosity of natural gels, non-toxicity ofthe medium, simplicity of the encapsulation process, and high viability of the encapsulated cells (Cassidy et al., 1996; Kolot, 1981; Leenen et al., 1996; McLoughlin, 1994; Trevors et al., 1992). The bead formation process involves the dispersion of two immiscible liquid phases resulting in a water-in-oil (W/O) emulsion. A suspension of viable cells in an aqueous solution of the polymer (disperse phase) is emulsified in a hydrophobie phase such as vegetable oils (continuous phase). The gelation of the small droplets of the dispersed phase is subsequently initiated by decreasing the emulsion temperature. In the present study, an emulsification-internal gelation method was developed for the encapsulation of a mixed bacterial culture in gellan gum microbeads. Gellan gum is produced by the microorganism Sphingomonas elodea (ATCC 31461), earlier referred to as Pseudomonas elodea, through fermentation processes (Kang et al., 1982). The mechanism of gelation and texture of gellan gum suggest a strong similarity with agar and carrageenans. However, gellan gum has been found to have superior rheological properties to agar and carrageenans at equivalent concentration (Sanderson et al., 1989). The use of gellan gum for encapsulation of viable cells has been addressed at considerably lower concentrations of both gel and gelling agent comparing to K-carrageenan, agar, and alginate described in previous reports (Buitelaar et al., 1988; Nilsson et al., 1983). Unlike sorne other ion-sensitive gelling polysaccharides such as alginate, the reactivity between gellan gum and ions is non-specifie 3. Production ofSize-Control/ed Gel/an Gum Microbeads 37 Encapsulating Gasoline Degrading Bacteria and gels can be formed with a wide variety of cations including alkaline and alkaline-earth cations (Moorehouse et al., 1981; Sanderson and Clark, 1983). For a given gellan gum concentration, however, divalent calcium and magnesium ions have been used at substantially lower levels than monovalent sodium and potassium ions to achieve gelation leading to strong gels (Sanderson and Clark, 1983). Use of gellan gum as an entrapment matrix has been recommended in fermentation processes due to its mechanical and thermal stability (Camelin et al., 1993; Norton and Lacroix, 1990). The present work was also based on the use of gellan gum for microbial encapsulation, due to its promising superior properties to other natural polymers. Although various physicochemical parameters of the emulsification technique for encapsulation purposes have been studied (Audet and Lacroix, 1989; Oyez et al., 1997; Poncelet et al., 1992; Poncelet et al., 1995), the results are strongly dependent on the individual characteristics of the particular polymer used. Therefore, this study was directed toward developing an emulsification method for encapsulation of viable and active bacterial consortia in gellan gum microbeads. Furthermore, the effect of various emulsion parameters on size distribution of microbeads was investigated in order to produce microbeads smaller than 50 /lm, which could be used as cell carriers for in situ bioaugmentation of contaminated aquifers.

3.3. Material and Methods 3.3.1. Reagents Gellan gum (Kelcogel®) was generously provided by the CP Kelco US, Inc., formerly NutraSweet Kelco Co. (San Diego, CA, USA). Canola oil was purchased from a local food distributor. Gasoline was purchased from a commercial fuel retailer. The physical and chemical properties of gasoline, and its chemical composition are summarized in Table 3.1.

Gasoline is a complex mixture of C4 to CIl hydrocarbons. The chemical composition of gasoline may vary, depending on refining process conditions, bIending of stock, source of crude oil, season, and geographic region. The variation of gasoline composition may prevent achieving similar experimental results within an extended time frame. The present study was carried out using samples from a single batch of gasoline. The main objective was to demonstrate and compare the performance of free and encapsulated cells in the removal of a commercially available gasoline product with a complex entity rather than a laboratory­ prepared hydrocarbon mixture with a limited number of hydrocarbons (artificial gasoline). AlI other chemicals used in this study were of reagent grade. 3. Production ofSize-Controlled Gellan Gum Microbeads 38 Encapsulating Gasoline Degrading Bacteria

Table 3.1. Characteristics of a commercial gasoline.

Chemical Composition Representative Concentrations (% w/w) Hydrocarbon Group:

n-Alkanes (C4 - Cll) 10 - 30

Branched Alkanes (C4 - C9) 18 - 60

Cycloalkanes (C6 - C9) 3 - 14

Alkenes (C4 - C6) 5 - 14

Branched Alkenes (Cs & C6) <1

Monoaromatics (BTEX, C3- & C4-benzenes) 18 - 40 Polyaromatics <2.5

Other Ingredients: Lead, Pb Max. 5 mg/L Manganese,Mn Max. 18 mg/L Sulfur, S Max. 0.10 % mass RSH Max. 0.0030 % mass Anti-Oxidant Min. 5.7 g/m; Corrosion Inhibitor Min. 6.0 g/m; MTBE Max. 2.7 % mass Particulate Matter Max. 2.2 mg/L

Physical and Chemical Properties Boiling Point 35 - 220 oC Density 0.798 g/mL @ 23 oC 0.811 g/mL @ 10 oC Vapor Density (Air =1) 3.5 Octane No. 87

3.3.2. Microorganisms and Growth Medium The mixed bacterial culture (MBC) was previously isolated from the top layers of a gasoline­ contaminated soil of an industrial site (Montreal, Quebec, Canada). The ceUs were preserved as a frozen stock culture at -80 oc. After thawing, the culture was suspended in an enrichment I mineral salts medium (MSM), supplied with gasoline (160 mg L- ) as the sole source of carbon. The culture was then harvested by centrifugation at 12,000 x g for 15 min at 4 oC. The ceU pellet was re-suspended in 2 mL of a sterile 0.1 % (w/v) calcium chloride solution in preparation for encapsulation. The enrichment culture was a consortium of spherical (cocci) 3. Production ofSize-Controlled Gellan Gum Microbeads 39 Encapsulating Gasoline Degrading Bacteria and rod-like (bacilli) as revealed by microscopic observations. The average dimensions of cocci and bacilli were 0.91lm and 0.7 x 2.4/lffi, respectively. The average density of wet cells I determined by drying five replicates ofthe cell pellet at 105°C, was 1.14 ± 0.04 g mL- . The enrichment had gone through twelve successive transfers before this study. l The MSM contained (g L- ): KH2P04 0.87; K2HP04 2.26; (NH4)2S04 1.1; and MgS040.047. To this solution was added 1 mL per liter trace metal solution composed of (g l L- ): CO(N03)2.6H20 0.291; AlK(S04)2.12H20 0.474; CuS04 0.16; ZnS04.7H20 0.288; FeS04.7H20 2.78; MnS04.H20 1.69; Na2Mo04.2H20 0.482; and Ca(N03)2.4H20 2.36. The final pH of the medium was 7.0 ± 0.1.

3.3.3. Microbead Production Gellan gum microbeads were produced by emulsification-intemal gelation. A 0.75% (w/v) dispersion of gellan gum in sterile de-ionized water was prepared and heated to 75 oC to dissolve and form the pregel solution (sol). Calcium chloride was added at 0.06% (w/v) and the sol was left at room temperature to cool to 45 oc. The pH was adjusted to between 6.9 and 7.2 with 0.1 N NaOH. The sol was then emulsified in 330 mL sterile canola oil aided by a non-ionic oil soluble surfactant, Span 80 (sorbitan monooleate), at 45 oc. The W/O emulsion was formed in a l-L round-bottomed cylindrical glass reaction vessel equipped with four baffle blades as illustrated in Figure 3.1. The emulsion was vigorously stirred using a quarter­ circular paddle impeller assembled with a T-Line laboratory stirrer (model 102, Talboys Engineering Corp., Montrose, PA, USA). The impeller was adjusted at one-third of the liquid height from the reactor bottom in aIl experiments. There was a 0.2-cm gap between the baffle blades and the reactor wall in order to prevent stagnant zones and to minimize the accumulation ofgel microbeads behind the blades. To initiate gelation, the reaction vessel was cooled rapidly to 15 oC by means of an ice bath with continued stirring for 2 h. The oil­ microbead dispersion was transferred with gentle mixing into 500 mL of a sterile 0.1 % (w/v) calcium chloride solution. The oil was removed by aspiration after partitioning of microbeads into the aqueous phase, and the microbeads were washed repeatedly with a sterile 0.1 % (v/v) Tween 80 (polyoxyethylene (20) sorbitan monooleate) solution. For encapsulation of cells, a 2-mL suspension of cells in sterile 0.1% (w/v) calcium chloride solution was mixed with the sol after pH adjustment at 45 oC, and the mixture was then emulsified in canola oil. The effect of various emulsion parameters including the stirring rate (1000 - 5500 rpm), disperse phase volume fraction (0.077 - 0.250), emulsifier concentration (0 - 0.20% w/w), emulsification time (3 - 30 min), and cell mass loading (0 - 20 gcells L- l sol) on size 3. Production of Size-Controlled Gellan Gum Microbeads 40 Encapsulating Gasoline Degrading Bacteria

(a) E u o N

E t u l!"l ,...C\l

.11 .. 0.8 cm ~ ~1 1... 10.0 cm 1

(b)

1" 5.0 cm .1

Figure 3.1. Design of reactor and impeller: a) reactor and baffle arrangement, b) guarter­ circular paddle. (drawing is not to scale)

distribution of microbeads was studied. The mean diameter and Slze distribution of mi~robeads were measured by means of a Malvem particle and size analyzer (series 2600, Malvem Instruments, Inc., Southborough, MA, USA) based on the laser light diffraction technique. The size distribution was demonstrated by frequency and cumulative volume distribution curves. The mean diameter was obtained from the 50% point of the cumulative distribution curve. The spread between the 10 and 90% points was expressed by the span (eg. 3.1), d90 - dlO Span = (3.1 ) dsü 3. Production ofSize-Controlled Gellan Gum Microbeads 41 Encapsulating Gasoline Degrading Bacteria

where dlo' dso and d90 are diameters at 10, 50 and 90% points of the cumulative distribution curve, respectively. The disperse phase volume fraction (<1» was defined as the ratio of the volume of disperse phase and the total emulsion volume (eq. 3.2),

<1> = VD (3.2) VD+VC where Vc and VD are the volumes ofthe continuous and disperse phases, respectively. The precision or repeatability of the microbead formation process and particle size measurements was studied by parametric statistical analysis of the mean diameter of fifteen batches of cell-free microbeads, produced under a particular set of emulsion conditions. The distribution of mean diameter of microbeads was examined by constructing the probability density function for a normal distribution, and estimating the 95% confidence interval (CI) for the population of mean diameters. Furthermore, the repeatability of the cell encapsulation process was studied on triplicate batches at various cell-Ioadings.

3.3.4. Gasoline Biodegradation Biodegradation activity of the encapsulated MBC in the removal of gasoline hydrocarbons was investigated in 160-mL serum bottles (in duplicate) on a rotary shaker (100 rpm, 10°C), and compared to that of free cells. Each serum bottle contained 20 mL of MSM, and was inoculated with either free or encapsulated MBC (2 gcells L- I sol) to give a final concentration of 0.26 gcells L- I MSM. The bottles were capped with Mininert standard gas-tight valves (Supelco, Sigma-Aldrich Canada Ltd., Oakville, ON, Canada). Encapsulated ceIls, deactivated by autoclaving at 121°C for 20 min, were used in control serum bottles under similar conditions. Gasoline was injected directly into the liquid phase in each bottle using a micro­ I syringe to give final substrate concentration of 400 mg L- . Headspace hydrocarbons were sampled with a gas-tight micro-syringe (Hamilton, no. 1705) 6 h after injection of gasoline into shaking bottles, and analyzed by gas chromatography (GC). The variation of total peak area of gasoline hydrocarbons determined by GC analysis was monitored during the incubation period (3 weeks). The gasoline removal efficiency was then estimated using the time-based gasoline hydrocarbons content measured in the gas phase, reported to its initial amount in the gas phase. As a mIe, in a closed vessel with fixed liquid and headspace volumes, the content of a volatile chemical species in the system gas phase is linearly proportional to its total content in the system. Thus the depletion percentage of such chemical species in the gas phase is identical to the total depletion percentage upon biodegradation in the liquid phase. However, 3. Production ofSize-Controlled Gellan Gum Microbeads 42 Encapsulating Gasoline Degrading Bacteria gasoline is consisted of many different chemicals whose degradation occurs at unequal rates. Although the initial partition of gasoline between the gas and liquid phases is independent of the total gasoline content, the composition ofresidual gasoline changes through the course of biodegradation, becoming more hydrophobie at the end as compared to the initial gasoline. This results in a relatively higher amount of gasoline hydrocarbons to be displaced to the gas phase, as compared to the initial instance. Thus the gasoline depletion percentage in the gas phase is an under-estimation of the overall biodegradation percentage. However the difference is slight: measurements of gasoline in both liquid and gas at the beginning and the end of the biodegradation test showed no more than 10% difference between degradation based on gas phase measurements and that based on total amounts. Hence the gasoline biodegradation estimation based on gas phase measurements is conservative, though yet quite acceptable. Furthermore, as such under-estimation applies to both encapsulated and free cell systems, the differential between the two degradation profiles should not be affected. Headspace samples (50~) were analyzed using a HP gas chromatograph (Hewlett­ Packard, mode! HP 6890) equipped with a flame ionization detector (FID), and a packed column (Supelco 1-2485, 1% SP-lOOO, 3 mm x 2 m, 60/80 mesh Carbopack B). Both the injector and detector temperatures were set at 250 oC, while the column temperature was kept I at 225 oc. Helium was used as the carrier gas at a flow rate of 50 mL min- .

3.4. Results Typical size distribution curves for gellan gum microbeads produced by emulsification­ internaI gelation are presented in Figure 3.2. The single peak of the frequency distribution curve represents the total microbead volume, and is characterized by the mean diameter corresponding to 50% on the cumulative distribution curve, and span which indicates the degree of polydispersity of the microbeads. In general, the size distributions were unimodal and followed the log-normal distribution, and microbeads were spherical, when observed microscopically. Diameters in this particular sample ranged from 16 to 34 /lm with a mean diameter of 21 /lm and a span of 0.2. Microbeads were formed at various stirring rates to evaluate the effect of impeller rotational speed on size distribution. As shawn in Table 3.2, the mean diameter decreased (by 63 %) from 88 to 31 lJffi with an increase in the rotational speed from 1000 to 3500 rpm, and remained relatively constant at a higher stirring rate of4500 rpm in the absence of emulsifier. With emulsifier present, the agitation rate for size reduction was limited at higher impeller rotational speeds. Although the use of the emulsifier resulted in further reduction of mean diameter to 17 lJffi at 4500 rpm, the diameter remained unchanged at higher stirring rates. The 3. Production ofSize-Controlled Gellan Gum Microbeads 43 Encapsulating Gasoline Degrading Bacteria

80 100 70 90 80 .- 60 .- ;:!!. ;:!!. 0 70 0 50 -->- 60 --Q) u > c +:i Q) 40 50 ctj ::J ::J 0" 40 Q) 30 E l... ::J U. 30 20 Ü 20 10 10 0 0 10 15 20 25 30 35 40 Diameter (IJm)

Figure 3.2. Frequency (.) and cumulative (0) size distribution of gellan gum-encapsulated mixed bacterial culture microbeads formed using 0.75% gellan gum, 0.06% CaC12, 8 g wet cells per liter sol, 0.1 % (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm.

Table 3.2. Variations in size distribution of gellan gum microbeads with stirring rate, using disperse phase volume fraction 0.143 and emulsification time 10 min.

Impeller Speed (rpm) Emulsifier (% w/w) Mean Diameter (/lm) Span 1000 0 87.6 1.0 1500 0 67.0 1.1 1500 0.1 43.5 1.0 1500 0.2 40.6 1.7 2500 0 52.5 1.1 3500 0 31.2 0.8 4500 0 32.3 0.7 4500 0.1 17.2 0.3 5000 0.1 19.5 0.5 5500 0.1 18.5 0.6 3. Production ofSize-Controlled Gellan Gum Microbeads 44 Encapsulating Gasoline Degrading Bacteria effect of the emulsifier on size reduction was similarly observed at 1500 rpm. The addition of Span 80 at 0.1 % caused a reduction of mean diameter (by 35%) to 44 /-lm while doubling the emulsifier concentration to 0.2% only resulted in a slight decrease in mean diameter (by 7%) to 41 /-lm. In general, the mean diameter and span of the size distributions decreased with the increase of the impeller rotational speed. The influence of emulsifier concentration on size distribution is examined in Figure 3.3. The mean diameter dropped to 17 f-lill with an increase of the emulsifier concentration to 0.15%. The span decreased in the emulsifier-aided emulsions, compared to that without emulsifier, but remained approximately constant for the entire range of emulsifier concentration.

40 1.6 1.4 ..- 35 E :::l. 1.2 --'- 30 Q) 1 +-' Q) c E 25 0.8 CO CO c- (5 Cf) c 0.6 CO 20 Q) 0.4 ~ 15 0 0 0 0 0.2 10 0 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 Emulsifier (% w/w)

Figure 3.3. Effect of the emulsifier concentration on mean diameter (.) and span (0) of gellan gum microbeads using disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm.

The variations in size distribution of microbeads with disperse phase volume fraction,

<1> are presented in Figure 3.4. The mean diameter ranged from 31 to 17 /-lm with the decrease

in <1> from 0.250 to 0.077. The span was noticeably higher for values of <1> greater than 0.143. 3. Production ofSize-Controlled Gellan Gum Microbeads 45 Encapsulating Gasoline Degrading Bacteria

The impact of volume fraction on mean diameter was diminished with the decrease in the disperse phase fraction indicating a lower limit for effectiveness of this parameter in reducing the microbead size. In experiments carried out at 1500 rpm and within the same range of <1>, the mean diameter decreased from 135 to 67 /lm and span remained constant at about 1.0. The variations in microbead size distribution with the emulsification time are summarizéd in Table 3.3, which indicate that the emulsion reaches an equilibrium in less than 10 min. The mean diameter decreased to 25 Il-m with an increase of the emulsification time to 7 min but climbed to 32 Il-m when the stirring was extended ta 10 min, and remained constant for longer emulsification times.

40 1.6 1.4 .- 35 E 1.2 :::I.

~ 30 --CI) 1 ...... CI) c <ù E 25 0.8 c.. <ù en (5 0.6 c 20 <ù CI) 0.4 ~ 15 0.2 10 0 0 0.05 0.1 0.15 0.2 0.25 0.3

Disperse Phase Volume Fraction, <1>

Figure 3.4. Effect of the disperse phase volume fraction (<1» on mean diameter (.) and span (0) of gellan gum microbeads using 0.1 % (w/w) emulsifier, emulsification time 10 min, and stirring rate 4500 rpm.

The effect of cell mass loading on size distribution is exarnined in Figure 3.5. It was noted that increasing the concentration ofwet cells to 20 g L- I did not cause a real change in the size distribution. However, a slight increase in the mean diameter of the microbeads was observed 3. Production ofSize-Controlled Gellan Gum Microbeads 46 Encapsulating Gasoline Degrading Bacteria compared to that ofcell-free microbeads. The mean diameter and span ranged from 21 to 26 /lm and 0.2 to 004, respectively.

Table 3.3. Variations in size distribution of gellan gum microbeads with emulsification time, using disperse phase volume fraction 0.143 and stirring rate 4500 rpm.

Emulsification Time (min) Mean Diameter (Ilffi) Span 3 27.8 0.3 5 26.1 0.3 7 24.7 0.3 10 32.3 0.7 20 32.0 0.9 30 31.9 0.7

40 1.6 1 1 1 1 1 1 1 1 1 1

- 1.4 ..- 35 - E - ::::l. 1.2 10- 30 f- --Q) - ..- 1 Q) c ctS E 25 f-~ ~ - 0.8 ctS c. i5 - • CI) . • - 0.6 c f- - • • ctS 20 • • Q) a - 0.4 ~ l- a 15 F'- a a a a a -u - 0.2

10 1 1 1 1 1 1 1 1 1 1 0 0 2 4 6 8 10 12 14 16 18 20 22 1 Cel! Loading (g L- )

Figure 3.5. Effect of the cell mass loading on mean diameter (.) and span (0) of gellan gum microbeads using 0.1% (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm. 3. Production ofSize-Controlled Gellan Gum Microbeads 47 Encapsulating Gasoline Degrading Bacteria

Fifteen batches of cell-free microbeads were produced to study the precision of the formulation process. The mean (fl) and standard deviation (cr) of the population of mean diameters were 19.2 flm and 3.7 flill respectively. The average deviation of mean diameter of various batches ranged from 0 to 2.0 flill for duplicate sample to sample measurements. The absolute value of the skewness of the population was 0.16, indicating that the probability distribution of the mean diameters could have a symmetrical bell shape (Metcalfe, 1994). Accordingly, a probability density function for normal (Gaussian) distribution of mean diameter was constructed as illustrated in Figure 3.6. The mean diameter of microbeads ranged between 11.5 and 25.6 flm. The 95% confidence interval for the fl and cr of the population was 17.4 - 21.0 flill and 2.7 - 5.9 flill, respectively, according to the standard normal distribution method (Bethea and Rhinehart, 1991). These results demonstrate the repeatability of the emulsification process for production of microbeads with consistent mean diameter below 50 flill.

0.12

0.1

>. 0.08 ~ ..c Cù ..c 0.06 0 lo... a.. 0.04

0.02

0 0 5 10 15 20 25 30 35 40 Mean Diameter (~m)

Figure 3.6. Normal distribution of mean diameter of gellan gum microbeads using 0.1 % (w/w) emulsifier, disperse phase volume fraction 0.143, emulsification time 10 min, and stirring rate 4500 rpm. 3. Production ofSize-Controlled Gellan Gum Microbeads 48 Encapsulating Gasoline Degrading Bacteria

The statistical data for triplicate batches ofencapsulated MBC rnicrobeads for various cell-ioading levels are summarized in Table 3.4. The population mean (fl) and standard deviation (cr) of mean diameter of rnicrobeads ranged from 21.1 to 24.4 flm and 0.3 to 3.1 fllIl, respectively, for triplicate batch preparations of different cell-ioading rates. The average deviation of mean diameter ranged between 0 and 0.4 flm for duplicate sample to sample measurements for different batches. The average cell number loading was estimated based on the population mean (Il) of triplicate batches, average dimensions of cell species, and the density of wet cells measured experimentally (Table 3.4). The number of cells encapsulated per unit volume (mL) of the gel support in this study ranged from 2.6 x 109 to 1.0 X lOlo (corresponding to 15 to 78 cells per rnicrobead), which is comparable with the quantities used in other previous studies with alginate and K-carrageenan macrobeads (Beunink and Rehm, 1990; Cassidy et al., 1997; Greenberg et al., 1996).

Table 3.4. Statistical characteristics of triplicate batches of encapsulated MBC microbeads; (Il) population mean, and (cr) standard deviation.

Cell Mass Loading fl±cr Ave. Cell Number Loading (g per L sol) (flm) (per no. rnicrobead) (per mL rnicrobead) 2 22.1 ± 1.0 15 2.6 x 109 4 21.1 ± 0.3 25 5.1 x 109 8 24.4 ± 3.1 78 1.0 x lOlO

The activity of encapsulated MBC (2 gcells L- I sol) was compared with free cells for biodegradation of gasoline at 10 oc. The biodegradation profiles are illustrated in Figure 3.7 I for an initial gasoline concentration of 400 mg L- . The encapsulated cells degraded gasoline hydrocarbons immediately after incubation, and more than 98% removal was achieved within approximately 10 days. In comparison, the biodegradation by free cells started at a lower rate after a 3-day lag phase such that the time to 80% biodegradation was much longer than encapsulated cells, but the biodegradation accelerated progressively and the same level of removal was reached within the above period. The encapsulation of MBC in gellan gum microbeads was advantageous in reduction of an adaptation period required by free cells. Deactivated encapsulated cells showed no degradation activity, indicating that the biological 3. Production ofSize-Controlled Gellan Gum Microbeads 49 Encapsulating Gasoline Degrading Bacteria activity ofliving cells in the liquid phase is responsible for the depletion of hydrocarbons in the gas phase. The biodegradation of gasoline hydrocarbons by free and gellan gum-encapsulated cells has been further investigated in both liquid suspension and saturated soil systems, and will be the subject of Chapter 6.

100

.- 80 ~ 0

--ct! > 60 0 E ID CI: 40 I a.. 1- 20

0 0 3 6 9 12 15 18 21 lime (days)

Figure 3.7. Biodegradation of gasoline by: (e) encapsulated, (0) free, and (0) deactivated encapsulated mixed bacterial culture at 10 oC; TPH is total petroleum hydrocarbons. Data are average of duplicate mns, and error bars show the average deviation. The mean diameter of encapsulated cell microbeads was 23 /-lm.

3.5. Discussion The objective ofthe present study was to develop an emulsification-intemal gelation technique for encapsulation of viable degrading cells in gellan gum microbeads. It is important that the microbeads be formulated with high levels of biodegradation activity, and a diameter appropriate for injection and distribution into a contaminated aquifer. The gellan gum-encapsulated cells demonstrated a high biodegradation activity in the removal of gasoline. Although the overall removal of gasoline by free and encapsulated cells 3. Production ofSize-Controlled Gellan Gum Microbeads 50 Encapsulating Gasoline Degrading Bacteria was comparable (over 98%), the encapsulated cells degraded gasoline at a much higher initial rate, and without an extended lag period as encountered with free cells. Encapsulation provided the cells with a protective solid barrier, reducing their bioavailable concentration in the microbeads inner space with respect to that of the bulk liquid. It is likely that the concentration gradient within the gel matrix alleviated the need of the entrapped cells for adaptation to high gasoline concentration and reduced the lag period, while the free cell organisms had to adapt to the high concentration of toxie gasoline hydrocarbons near their surface in the liquid phase. In general, an aquifer may consist of several layers of silt, sand and gravel mixture with porosities ranging from 30 to 40%. The dimension of soil pores can be estimated from a pore channel model described in a previous report (Reddi and Bonala, 1997). Typically for soils composed of spherical grains of 130 flm (silty sand) to 700 flm (gravelly sand) mean diameter, the pore channel diameter ranges from 8 to 100 flm. This range of pore dimension should allow the transport of microbeads smaller than 50 flm. The microbeads within this range will further satisfy the particle straining criteria (McDowell-Boyer et al., 1986), which considers a minimal value of 10-20 for the ratio of the grain to suspended particle diameter for sufficient transport. Therefore, the influence of emulsion parameters on size distribution of microbeads was investigated in order to produce microbeads of minimal size (below 50 flm), adequate for subsurface applications. During emulsification, the gellan gum sol droplets are deformed giving rise to droplet disruption and dispersion. The deformation is opposed by the Laplace pressure (Walstra, 1983),

~P = 2y (3.3) r where 'Y is the interfacial tension and r is the radius of a spherical droplet. The deformation takes place when the shear stresses supplied by mechanieal agitation overcome the Laplace pressure. It is evident from equation 3.3 that higher stirring rates should be applied to disrupt smaller droplets. At higher rotational speeds, the sol is more finely dispersed due to the higher shear forces applied, leading to formation of smaller microbeads upon gelation. The reduction in mean diameter with the impeller rotational speed has been similarly demonstrated in other studies on use of the emulsification technique for production of alginate (Poncelet et al., 1992) and K-carrageenan (Audet and Lacroix, 1989) beads (0.2 - 4 mm). The mean diameter of gellan gum microbeads was approximately constant for the stirring rates above 4500 rpm (Table 3.2). Therefore, the impeller speed was set at 4500 rpm in subsequent formulations. The formation of smaller microbeads corresponded to a decrease 3. Production ofSize-Controlled Gellan Gum Microbeads 51 Encapsulating Gasoline Degrading Bacteria in the span of the size distributions. Microbeads with a lower degree of polydispersity are desired to facilitate the uniform distribution of injected microbeads through the contaminated zone of an aquifer. The use of an emulsifier favored size reduction. A large amount of mechanical energy should be dissipated to maintain the large interfacial area provided by small disrupted droplets. In general, emulsifier lowers interfacial tension, facilitating disruption of the droplets due to decrease in the Laplace pressure (eq. 3.3), leading to the formation of smaller microbeads. Therefore, it can be expected that at higher concentrations of emulsifier, less mechanical energy is needed. The same emulsifier (Span 80) was used at concentrations as high as 2% to decrease the mean diameter of alginate beads (Poncelet et al., 1995) in an emulsification process. In the present work, the emulsifier concentration was limited to 0.2% due to its potential toxic effect. The dispersed internal phase has a much larger interfacial area than the respective individual phases in the emulsion. In dilute dispersions (<1> < 0.167), droplets are weIl separated, reducing the chance ofcollision and coalescence. Furthermore, the viscosity of the W/0 emulsion decreases when the volume fraction of disperse phase decreases (Becher, 1977). The decrease of viscosity in turn causes a decrease in the droplet size under constant shear rate conditions. The mean diameter of K-carrageenan beads was similarly shown to be a function ofthe aqueous volumetric fraction when K-carrageenan concentration was equal to or lower than 3% (Audet and Lacroix, 1989). The variations of <1> affect microbead yield and must be considered as a design parameter of the emulsification process. For a given volume of the continuous phase, decreasing the value of <1> will reduce both the microbead diameter and the amount of final microbead product per batch. Generally, a <1> value of 0.143 was chosen for the production of microbeads in this study as a compromise between the desired size distribution and yield of the microbeads. Increasing the emulsification time was found to be a less effective way to reduce the mean diameter, compared to the stirring rate, surfactant concentration, and <1>. A similar conclusion was reached on the effect of stirring time on size distribution of K-carrageenan sol droplets in a soybean oïl dispersion (Audet and Lacroix, 1989), and gelatin sol droplets in paraffin oil (Ovez et al., 1997). The disruption of droplets in an emulsion often follows first­ order kinetics (Walstra, 1983). The rate of disruption rapidly decreases when the emulsification time increases, and the droplet size approaches a lower limit. The breakup of a disrupted droplet is repeated in several consecutive steps. In each step the broken droplet requires less deformation time than the previous steps (Walstra, 1983). Therefore the emulsion reaches an equilibrium condition within a few minutes. In our experiments, the 3. Production ofSize-Controlled Gellan Gum Microbeads 52 Encapsulating Gasoline Degrading Bacteria emulsification seems to be complete within less than 10 min and the prolongation of the stirring time had no effect on further size reduction. The mean diameter and span of microbeads were increased for emulsification times beyond 7 min (Table 3.3). This could be explained by an increase in the coalescence rate of the large number of sol droplets formed during a prolonged emulsification. Hence, the rate of coalescence of droplets could increase more rapidly as a function of the number of droplets (Walstra, 1983), leading to formation of larger microbeads. The variations of size distribution of microbeads with the cell mass loading were small over the range of 1 - 20 gcells L- I sol. However, the mean diameter of the microbeads increased slightly compared to those without cells. This difference could be neglected considering the 95% CI of the mean diameter of cell-free microbeads. Statistical analysis demonstrated the precision and reliability of the emulsification technique for production of encapsulated cell microbeads with a defined narrow range of size distribution. The ability to formulate encapsulated cell microbeads with controlled diameter, mainly smaller than 50 1JIIl, is the key to large-scale subsurface bioremediation applications. The present results establish the role of the most important operational parameters of a W/0 emulsification process for encapsulation of microbial cells. The method utilized here is simple and uses non-toxic biocompatible materials. The oil is reusable, lowering the operational costs. Moreover, since emulsions can be produced in large-scale equipment, this process has potential for field applications. The transport of microbeads through a wide range of soil grain distributions (0.125 - 16 mm) will be the subject of subsequent chapters.

3.6. Conclusions The emulsification-intemal gelation procedure developed in this study illustrates a novel technique for producing biologically active gellan gum-encapsulated cells. High 9 1 concentrations of degrading cells Cl 0 - 1010 cells mL- support) were encapsulated within gellan gum microbeads of as small as 21 - 26 flm mean diameter. A high degree of gasoline removal (over 98%) was achieved by the encapsulated mixed culture. The encapsulation gave an advantage to entrapped cells in reducing the lag phase experienced in biodegradation by free cells. It was shown that the mean diameter and size distribution of the microbeads can he controlled by changing the emulsification conditions. Size measurement results indicated that the mean diameter varied as a descending function of the impeller rotational speed, and surfactant concentration, but as an ascending function of the volume fraction of the disperse phase. The span of the size distributions was decreased concurrently with the mean diameter in all formulations. Among the different techniques employed to reduce the microbead 3. Production ofSize-Controlled Gellan Gum Microbeads 53 Encapsulating Gasoline Degrading Bacteria diameter, high stirring rates (> 3500 rpm) and low disperse phase volume fraction « 0.143) along with the use of an emulsifier such as Span 80 were found to be the most effective. The statistical analysis demonstrated a high degree of precision in formulation of microbeads, 21 ­ 26 ~m mean diameter. The emulsification-intemal gelation is a promising technique for production of encapsulated cell microbeads for in situ bioaugmentation of sorne contaminated aquifers.

Acknowledgements The authors thank the Natural Sciences and Engineering Research Council of Canada for financial support and McGill University for the awarding of the Max Stern Fellowship to P.M. This research was supported by the National Research Council of Canada, NRC paper no. 44601. 3. Production ofSize-Controlled Gellan Gum Microbeads 54 Encapsulating Gasoline Degrading Bacteria

References Audet, P. and Lacroix, C. (1989). Two-phase dispersion process for the production of biopolymer gel beads: Effect of various parameters on bead size and their distribution. Process Biochemistry December: 217-226. Baveye, P., Vandevivere, P., Hoyle, B. L., DeLeo, P. C. and Sanchez de Lozada, D. (1998). Environmental impact and mechanisms of the biological clogging of saturated soils and aquifer materials. Critical Reviews in Environmental Science and Technology 28: 123-191. Becher, P. (1977). Physical Properties of Emulsions. Robert E. Krieger Publishing Company, New York. Bethea, R. M. and Rhinehart, R. R. (1991). Statistical inference and estimations. In: Applied Engineering Statistics, Marcel Dekker, Inc., New York, 61-80. Beunink, J. and Rehm, H.-J. (1990). Coupled reductive and oxidative degradation of 4­ chloro-2-nitrophenol by a co-immobilized mixed culture system. Applied Microbiology and Biotechnology 34: 108-115. Buitelaar, R. M., Hulst, A. C. and Tramper, J. (1988). Immobilization of biocatalysts in thermogels using the resonance nozzle for rapid drop formation and an organic solvent for gelling. Biotechnology Techniques 2: 109-114. Camelin, 1., Lacroix, c., Paquin, c., Prevost, H., Cachon, R. and Divies, C. (1993). Effect of chelants on gellan gum rheological properties and setting temperature for immobilization ofliving Bifidobacteria. Biotechnology Progress 9: 291-297. Cassidy, M. B., Lee, H. and Trevors, 1. T. (1996). Environmental applications of immobilized microbial cells: A review. Journal ofindustriai Microbiology 16: 79-101. Cassidy, M. B., Shaw, K. W., Lee, H. and Trevors, J. T. (1997). Enhanced mineralization of pentachlorophenol by K-carrageenan-encapsulated Pseudomonas sp. UG30. Applied Microbiology and Biotechnology 47: 108-113. Greenberg, N., Tartakovsky, B., Yinne, G., Ulitzur, S. and Sheintuch, M. (1996). Observations and modeling of growth of immobilized microcolonies of luminous E. Coli. Chemical Engineering Science 51: 743-756. Kang, K. S., Veeder, G. T., Mirrasoul, P. J., Kaneko, T. and Cottrell, 1. W. (1982). Agar-like polysaccharide produced by a Pseudomonas species: Production and basic properties. Applied and Environmental Microbiology 43: 1086-1091. Knaebel, D. B., Stormo, K. E. and Crawford, R. L. (1996). Immobilization of bacteria in macro- and microparticles. In: Bioremediation Protocols. Sheehan, D. (Ed.), Humana Press, Inc., Totowa, NI, 67-78. 3. Production ofSize-Control/ed Gel/an Gum Microbeads 55 Encapsulating Gasoline Degrading Bacteria

Kolot, F. B. (1981). Microbial carriers-Strategy for selection-Part 1. Process Biochemistry August/September: 2-9. Lee, S.-T., Rhee, S. K. and Lee, G. M. (1994). Biodegradation of pyridine by freely suspended and immobilized Pimelobacter sp. Applied Microbiology and Biotechnology 41: 652-657. Leenen, E. J. T. M., Dos Santos, V. A P., Grolle, K. C. P., Tramper, J. and Wijffels, R H. (1996). Characteristics of and selection criteria for support materials for cell immobilization in wastewater treatment. Water Research 30: 2985-2996. McDowell-Boyer, L. M., Hunt, J. R and Sitar, N. (1986). Particle transport through porous media. Water Resources Research 22: 1901-1921. McLoughlin, A J. (1994). Controlled release of immobilized cells as a strategy to regulate

ecological competence of inocula. Advances III Biochemical Engineering/ Biotechnology 51: 1-45. Metcalfe, A V. (1994). Justifying Engineering Decisions. In: Statistics in Engineering-A Practical Approach, Chapman & Hall, London, 32-62. Moorehouse, R, Colegrove, G. T., Stanford, P. A, Baird, J. K. and Kang, K. S. (1981). PS­ 60: A new gel-forming polysaccharide. In: Solution Properties of Polysaccharides. Brant, D. A (Ed.), American Chemical Society, Washington DC, 111-124. Nilsson, K., Birnmaum, S., Flygare, S., Linse, L., Schroder, u., Jeppsson, u., Larsson, P.-O., Mosbach, K. and Brodelius, P. (1983). A general method for the immobilization of cells with preserved viability. European Journal of Applied Microbiology and Biotechnology 17: 319-326. Norton, S. and Lacroix, C. (1990). Gellan gum gel as entrapment matrix for high temperature fermentation processes-Rheological study. Biotechnology Techniques 4: 351-356. Oyez, B., Çitak, B., Oztemel, D., Balbas, A, Peker, S. and Çakir, S. (1997). Variation of droplet sizes during the formation of microcapsules from emulsions. Journal of Microencapsulation 14: 489-499. Poncelet, D., Lencki, R, Beaulieu, c., Halle, J. P., Neufeld, R J. and Fournier, A (1992). Production of alginate beads by emulsificationlinternal gelation. 1. Methodology. Applied Microbiology and Biotechnology 38: 39-45. Poncelet, D., Poncelet De Smet, B., Beaulieu, c., Huguet, M. L., Fournier, A and Neufeld, R J. (1995). Production of alginate beads by emulsificationlinternal gelation. II. Physicochemistry. Applied Microbiology and Biotechnology 43: 644-650. 3. Production ofSize-Controlled Gellan Gum Microbeads 56 Encapsulating Gasoline Degrading Bacteria

Reddi, L. N. and Bonala, M. V. S. (1997). Analytical solution for fine particle accumulation in soil filters. Journal of Geotechnical and Geoenvironmental Engineering 123: 1143­ 1152. Sanderson, G. R., Bell, V. L. and Ortega, D. (1989). A comparison of gellan gum, agar, K­ carrageenan, and algin. Cereal Foods World 34: 991-998. Sanderson, G. R. and Clark, R. C. (1983). Gellan gum. Food Technology 37: 62-70. Smit, E., Lee, H., Trevors, J. T. and van Elsas, J. D. (1996). Interaction between a genetically engineered Pseudomonas jluorescens and bacteriophage sz)R2f in soil: Effect of nutrients, alginate encapsulation, and the wheat rhizosphere. Microbial Ecology 31: 125-140. Trevors, J. T., van Elsas, J. D., Lee, H. and van Overbeek, L. S. (1992). Use of alginate and other carriers for encapsulation of microbial cells for use in soil. Microbial Re1eases 1: 61-69. Trevors, J. T., van Elsas, J. D., Lee, H. and Wolters, A C. (1993). Survival of alginate­ encapsulated Pseudomonas jluorescens cells in sail. Applied Microbiology and Biotechnology 39: 637-643. Walstra, P. (1983). Formation of Emulsions. In: Encyclopedia of Emulsion Technology. Becher, P. (Ed.), Marcel Dekker, Inc., New York, 57-127. Weir, S. c., Dupuis, S. P., Providenti, M. A, Lee, H. and Trevors, J. T. (1995). Nutrient­ enhanced survival of and phenanthrene mineralization by alginate-encapsulated and free Pseudomonas sp. UG14Lr cells in creosote-contaminated soil slurries. Applied Microbiology and Biotechnology 43: 946-951. 57

4. Transport of Gellan Gum Microbeads in Soil Columns of Various Grain Size Distributions

Peyman Moslemy a,b, Denis MUlette b,t, Serge R. Guiot b , and

Ronald J. Neufeld C

a Department of Chemical Engineering, McGill University, 3610 University Street,

Montreal, Quebec, Canada H3A 2B2

b Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

C Department of Chemical Engineering, Queen's University, Dupuis Hall, Kingston, Ontario, Canada K7L 3N6

t CUITent address: Hydrogéo Plus Inc., 3333 Queen-Mary Road, Suite 501, Montreal, Quebec, Canada H3V 1A2

This study was conducted to evaluate the feasibility of the transport of cell-encapsulating gellan gum microbeads through the gravel and sand fractions of soil. The candidate has caITied out aIl the work in this chapter. Dr. Denis Millette contributed by providing his expertise in the field of Hydrogeology, and by offering insight to the design of soil columns. This article is to be submitted to the Journal of Environmental Engineering for publication. 4. Transport ofGellan Gum Microbeads in Soit Columns 58 ofVarious Grain Size Distributions

4.1. Abstract Application of gel-encapsulated bacteria to in situ bioaugmentation of a contaminated aquifer mainly depends on adequate transport and dispersion of injected capsules through the aquifer. The feasibility of encapsulated cell bioaugmentation was evaluated based on transport characteristics of gellan gum microbeads (12.6 - 18.0 /lm mean dia.) through soil. Transport experiments were conducted in six horizontal columns (5.2 cm id x 110 cm long) packed with different grain sizes of gravel (2 - 16 mm) and sand (0.25 - 2 mm). A suspension of microbeads (about 40 mg beads LI) in artificial groundwater was pulsed into each column for 6 h, followed by injection of bead-free water for 42 h. In general, the total amount of microbeads traveling across a given section of the column increased with time but decreased towards the column outlet as a function of mean grain size. For instance, the overall percentage of microbeads discharged from the column after 48 h was 25% for the sand column with a mean grain size of 0.77 mm as compared to 44% for the gravel column with a mean grain size of 3.2 mm. The average degree of retention within the columns varied with grain size distribution and travel distance, ranging from 0 - 0.06% of injected microbeads per centimeter for the columns with a mean grain size larger than 3.2 mm and 0.08 - 1.60% per centimeter for the columns with a mean grain size equal to or smaller than 3.2 mm. Results of this study demonstrate the transport of gellan gum microbeads through medium sand to medium gravel (0.25 - 16 mm) across distances up to 110 cm.

4.2. Introduction Bioaugmentation of a contaminated aquifer by active degrading microbial cells is one approach for in situ remediation of the aquifer, when indigenous microorganisms are unable to satisfactorily degrade the hazardous contaminants (Pritchard, 1992; Vogel, 1996). However, the augmented cells may be limited in their biodegradation capacity for various reasons including inhibitory effect of high concentrations of the contaminant, predation by protozoa, and insufficient transport of microorganisms through the porous matrix of the contaminated aquifer (Goldstein et al., 1985). During transport, microbial cells attach to available grain surfaces in the form of micro-colonies and produce exopolysaccharides, leading to formation of a plugging biofilm which can cause a 4. Transport ofGellan Gum Microbeads in Soil Columns 59 ofVarious Grain Size Distributions significant decrease in pore permeability (Jennings et al., 1995; Taylor and Jaffé, 1990), and an early interruption in the injection process. Encapsulation of metabolically active microbia} cells within insoluble polymeric gel carriers is an emerging technique to overcome practical limitations in the use of conventional free cell formulations. Encapsulation isolates the entrapped cells from the exterior environment, preventing cell attachment and formation of biofilm within aquifer interstices. Advantages in the use of encapsulated cells for environmental applications have been discussed comprehensively in a review paper (Cassidy et al., 1996). Applying the encapsulation technique to bioremediation of a contaminated aquifer requires production of gel microbeads that can be introduced hydraulically into the porous matrix of the aquifer. Transport of microbeads is controlled by the filtration effect of the soil matrix, which is influenced by various factors including the carrier fluid, the suspension of particles, and the porous medium characteristics. A number of previous reports (BaIes et al., 1997; Harmand et al., 1996; Harvey et al., 1989; Harvey et al., 1993; Jegatheesan and Vingeswaran, 1997; Moran et al., 1993a; Moran et al., 1993b; Petrich et al., 1998; Reddi and Bonala, 1997) have discussed the effect of grain and particle size distributions, and granular media heterogeneity and depth on transport of particles through granular media. Monodisperse micrometer- and submicrometer-sized synthetic polymer microspheres have been mostly used in deep-bed filtration studies, or as tracers of bacteria in transport studies under laboratory or field conditions. With the exception of an earlier study (Petrich et al., 1998), in which polystyrene microspheres were used to resemble the cell carriers, there has been little effort to investigate the transport of cell encapsulating gel microspheres larger than 10 /lm for subsurface remediation applications. Transport of 2-, 5-, and 15-/lm polystyrene microspheres with neutral surface characteristics was investigated through forced-gradient tracer experiments in a confined unconsolidated heterogeneous aquifer to evaluate the feasibility of using encapsulated cell microbeads for in situ bioremediation (Petrich et al., 1998). Microsphere concentrations decreased substantially within a 1.5 to 6 m distance from the injection wells. Particle filtration through the aquifer porous matrix was attributed mainly to a sedimentation mechanism, assuming that the chemical interactions between the uncharged microspheres and soil grains had little or no influence 4. Transport ofGellan Gum Microbeads in Soi! Columns 60 ofVarious Grain Size Distributions on particle removal. It was stated that the encapsulated celI microspheres could be applied to some contaminated aquifers subject to the degree of aquifer heterogeneity and flow path characteristics. In general, microencapsulation of cells within polymerie carriers involves production of microbeads inevitably consisting of a wide range of sizes (Brandenberger and Widmer, 1998; Hammill and Crawford, 1997; Ogbonna et al., 1989; Poncelet et al., 1992; Poncelet et al., 1995). Therefore, any attempt to study the transport behavior of celI carriers through granular media for in situ bioremediation should consider the effect of polydispersity. In the present study, gellan gum microbeads were produced with an emulsification technique, suitable for large-scale microbial encapsulation. Gellan gum is a polysaccharide produced by the microorganism Sphingomonas paucimobilis (ATCC 31461), earlier referred to as Pseudomonas elodea, through fermentation processes (Giavasis et al., 2000). The gelation of gelIan gum can be induced by cooling a warm

2 aqueous dispersion of gelIan gum in the presence of appropriate cations such as Ca +. The setting temperature is in the range of 20 - 60 oC, depending on the reagents type and concentration. Gellan gum gel has a number of functional properties that can be readily modified (Buitelaar et al., 1988; Kang et al., 1982; Nilsson et al., 1983; Sanderson et al., 1989; Sanderson and Clark, 1983). These include: 1) versatile texture and rheological characteristics, 2) stability to pH variations, 3) compatibility upon combination with other gums and polymers, 4) low concentration requirements compared to other polymers, 5) non-specifie reactivity with monovalent and divalent cations, and 6) high resistance to enzymatic degradation. A combination of these properties makes gelIan gum a superior medium to other natural polymers for encapsulation of active microorganisms. The purpose of this study was to evaluate the feasibility of in situ encapsulated celI bioaugmentation based on the transport characteristics of gellan gum microbeads through natural granular media with defined size distributions covering gravel (2 - 16 mm) and sand (0.25 - 2 mm) classes of soil. Grain size distribution is one of the important parameters influencing the particle transport and dispersion through the soil matrix, and therefore was the primary subject of our study. An unconsolidated aquifer may be composed of severallayers of gravel and sand within the above size ranges. The interstitial 4. Transport ofGellan Gum Microbeads in Soil Columns 61 ofVarious Grain Size Distributions pore size corresponding to a grain size of 0.25 - 2 mm is estimated to be between 20 and 480 /lm (Arya and Dierolf, 1989; Arya and Paris, 1981; Reddi and Bonala, 1997), suggesting that gellan gum microbeads lying within this range would not be strained from the carrying fluid. To verify this hypothesis, the transport of microbeads was investigated by conducting forced-gradient experiments under single pass uniform-flow conditions with pulse injection of microbeads into the gravel and sand columns. A suspension of microbeads in artificial groundwater was injected into each column, and microbead transport through the granular matrix was monitored by analysis of the pore water samples collected from sampling ports placed along the length of the column. The effect of grain size distribution on transport and deposition of gellan gum microbeads through the packed columns was evaluated.

4.3. Materials and Methods 4.3.1. Microbead Production Microbeads of gellan gum (Kelcogel®, the CP Kelco US, Inc., formerly Nutrasweet Kelco Company, San Diego, CA, USA) were produced by a technique termed emulsification­ internaI gelation. A dispersion of gellan gum in de-ionized water was prepared at 0.75% (w/v), and heated to 75 oC to dissolve. Calcium chloride was added at 0.06% (w/v) and the solution cooled to 45 oc. The pH was adjusted to 7 ± 0.1 with 0.1 N NaOH. The resulting pregel solution (disperse phase) was emulsified in 330 mL of canola oil (continuous phase) in a l-L stirred baffled cylindrical vessel (10 cm dia.) held at 45 oc. The pregel solution can be inoculated with a cell culture prior to emulsification for microbial encapsulation. The disperse phase volume fraction in the emulsions varied from 0.08 - 0.14. Span 80 (sorbitan monooleate) was used as an emulsifier at 0.1 % (w/w) in the emulsification process. The resulting water-in-oil emulsion was stirred vigorously at 4500 rpm using a quarter-circular paddle (5 cm dia.) for 10 min. The gelation of gellan gum droplets was initiated by cooling the reaction vessel to 15 oC in an ice bath with continued stirring. Microbeads were separated from the oil dispersion after 2 h by partitioning into 0.1 % (w/v) calcium chloride solution (0.5 L). The oil phase was removed by aspiration, and the microbeads were washed repeatedly with a 0.1 % (v/v) Tween 80 (polyoxyethylene-20- 4. Transport ofGellan Gum Microbeads in Soil Columns 62 ofVarious Grain Size Distributions sorbitan monooleate) solution. The size distribution and mean diameter of microbeads were measured by means of a Malvern particle size analyzer (series 2600, Malvern Instruments, Inc., Southborough, MA, USA).

4.3.2. Artificial Groundwater An artificial groundwater (AGW) was used in aIl transport experiments. The AGW was prepared with a chemical composition similar to a natural groundwater (McCaulou et al.,

1995) as foIlows (mg per liter of distilled water): MgS04.7H20 69.0; NaHC03 50.0;

CaC12.2H20 14.5; Ca(N03)2.4H20 64.0; and KF 2.0. The pH was adjusted to 7.5 with 1 N l NaOH. The total dissolved solids (TDS) of the AGW was 96.5 mg as CaC03 L- .

4.3.3. Granular Media and Columns The soil material was coIlected from the surface soillayers of two locations along a weIl­ documented esker in Ville-Mercier (Quebec, Canada). The coIlected material was thoroughly mixed and analyzed for determination of grain size distribution. Prior to packing the columns, the granular material was screened using standard sieves (US Standard Testing Sieve, VWR Scientific Products, West Chester, PA, USA) and a mechanical sieve shaker (model RX-29, W. S. Tyler, Mentor, OH, USA) to obtain one of the pre-defined grain-size classes of the US Geological Survey (USGS) standard soil classification system (Morris and Johnson, 1967). Transport of gel microbeads was studied in six columns (A to F) covering the different size ranges of gravel and sand of the USGS soil classification system (Table 4.1). A schematic description of the packed column system is given in Figure 4.1. Each column consisted of a Schedule-40 transparent PVC tube (5.2 cm id x 110 cm long) and was plugged with PVC fittings at both ends. Glass beads of 0.6 cm in diameter were placed at the inlet and outlet of each column to a depth of 2.5 cm to support and immobilize the granular bed and to ensure proper liquid distribution and discharge at the inlet and outlet of the column. The effective length of the granular media (excluding glass beads) was 110 cm. 4. Transport ofGellan Gum Microbeads in Soil Co/umns 63 ofVarious Grain Size Distributions

Table 4.1. Grain size distribution in packed columns based on the standard soil classifica- tion of the US Geological Survey (USGS).

Column A B C DE F Grain Type Size (mm) Weight (%) Medium Gravel 8 - 16 56.7 Fine Gravel 4-8 32.8 75.9 Very Fine Gravel 2-4 10.5 24.1 100 Very Coarse Sand 1-2 100 34.4 26.6 Coarse Sand 0.5 - 1 65.6 50.8 Medium Sand 0.25 - 0.5 22.6

Mean grain size, Dso (mm) 9.2 5.8 3.2 1.21 0.88 0.77

Eight sampling ports were installed at 15-cm intervals across the column. One sampling port was installed before the inlet of the column, and another one after the outlet to allow verification of the inflow and monitoring of the outflow concentration. Each sampling port consisted of a male adapter (type B, 1/4" x 1/4", Cole-Parmer Instrument Co., Vernon Hills, IL, USA), a 12.5-mm rubber septum, and a steel tube (2 mm id x 40 mm long). The adapter was screwed to the column before packing with grains. After packing, a reusable hypodermic Luer-Lock needle (B-D Yale, 15G x 3 112", VWR Canlab, Ville Mont Royal, QC, Canada) connected to a standard Mininert valve (Supelco, Sigma-Aldrich Canada Ltd., Oakville, ON, Canada) and a 3-mL plastic syringe, was inserted into the septum and passed through the adapter and steel tube to reach the surface of packed grains at the centerline of the column. This configuration allowed the collection of interstitial samples from the centerline of each column. Each column was wet-packed in a vertical position. The granular material was emplaced sequentially as individual homogenized 5­ cm layers under standing water to minimize air entrapment and was tamped down with a steel piston during filling to attain uniform packing. A Masterflex variable speed modular digital drive/dispenser (model 77300, Cole-Parmer Instrument Co., Vernon Hills, IL, USA) was used to circulate water across the columns. 4. Transport ofGellan Gum Microbeads in Soil Columns 64 ofVarious Grain Size Distributions

9 7

11 8 6

12 (il)

' Rubber Male / ~ Septum 4 Adapter,\ lII1lI' Washer t 2_ 1

Figure 4.1. Schematic of the packed column setup: (1) Scale, (2) Magnet Stirrer, (3) Microbead Suspension Flask, (4) AGW Recharge Tank, (5) Shut-off Valves, (6) Masterflex Pump, (7) Digital Controller, (8) Sampling Ports, (9) Piezometrie Tube, (10) Packed Column, (11) Anti-siphon Tube, (12) Collecting Tank.

Each column was equipped with three inlet and outlet ports to ensure uniform distribution and discharge of liquid medium over the packed bed cross-section. The column effluent discharged from the end of a tube that was held at a fixed reference elevation and dropped through an air gap (anti-siphon tube) into a tube directed into an effluent 4. Transport ofGellan Gum Microbeads in Soil Columns 65 ofVarious Grain Size Distributions collecting tank. Vertical piezometric tubes were installed at both ends, and along the column for measuring the differential head across the granular bed. The entire column system was housed in a laboratory refrigerator to maintain a constant temperature of 10°C, which represents the temperature typically measured in groundwater. The column was placed in a horizontal position and operated in a horizontal-flow mode.

4.3.4. Hydrodynamic Properties The hydrodynamic properties of the granular media in columns A to F determined by conducting a non-reactive radiotracer (tritium) test and a hydraulic conductivity test are summarized in Table 4.2.

l The tracer test was initiated by injecting tritium-amended AGW at 0.44 - 0.49 L h- . At regular intervals, pore water samples were collected from sampling ports located at 5, 20, 50, 80, and 110 cm downgradient from the packed bed inlet. The radioactivity of the samples was determined by means of a liquid scintillation analyzer (model TRI-CARB 2100 TR, Packard Instrument Company, Meriden, CT, USA). The experimental data were used to calculate the average water velocity and longitudinal dispersivity based on the theory of the advective-dispersive transport for a non-reactive solute in an isotropic homogeneous saturated porous medium (Ogata, 1970), and using a Simplex Optimization computer program (Devlin, 1994). The effective porosity was then obtained using the definition of average linear velocity in porous media (Freeze and Cherry, 1979).

Table 4.2. Characteristics of the granular media in packed columns.

Column Parameter ABC DEF Effective Porositya 0.58 0.41 0.46 0.45 0.43 0.43 Hydraulic Conductivity (cm sol) 27.78 4.71 2.11 0.34 0.14 0.08 2 Permeability (cm ) 3.7e-04 6.3e-05 2.8e-05 4.5e-06 1.8e-06 1.1e-06 Dispersivity (cm) 8.51 2.46 1.36 0.56 0.34 0.43

3 aTotal volume of soil in column =2337 cm . 4. Transport ofGellan Gum Microbeads in Soil Columns 66 ofVarious Grain Size Distributions

The hydraulic conductivity test was carried out to obtain the values of hydraulic conductivity and permeability of the granular matrices. The test was initiated by pumping AGW into the column and measuring the flow rate and differential head across the column at regular intervals. The hydraulic conductivity of the granular matrix was obtained from the slope of the linear curve of the specifie discharge versus hydraulic gradient data (Freeze and Cherry, 1979). The linear relationship illustrates Darcy's law for laminar fluid flow through a porous medium. Permeability of the granular matrix was then estimated from its mathematieal relationship with hydraulic conductivity (Freeze and Cherry, 1979). Proportional increase of differential head with the column length and comparable values of porosity estimated from the radiotracer breakthrough data measured at various sampling distances along each column (data not shown) indicated that the columns were uniformly packed. The Gaussian shape of the tritium breakthrough curves suggested that there was minimal flow along the walls of the columns during experiments.

4.3.5. Microbead Transport in Granular Media Forced-gradient experiments with pulse input of gellan gum mierobeads were used to study the transport of mierobeads through the granular matrix in each horizontal column. Following the hydraulic conductivity test, at least 2 pore volumes of water were displaced by injection of pure AGW at 0.50 L h- 1 prior to injection of microbeads. A suspension of microbeads in AGW was then pulsed through the columns for 6 h, under the conditions given in Table 4.3. The measured concentration of the injectant slightly fluctuated over the injection period. These fluctuations can be attributed to the mierobead settlement in the injection tubes, or disturbance caused in the inflow stream during sampling. The influent was switched back to bead-free AGW after 6 h, and injection was maintained for 42 h. Samples of pore water were collected from the inlet and outlet sampling ports, and from sampling ports located at 5, 20, 50, 80, and 110 cm downgradient from the packed bed inlet. The outlet sampling port was considered to be 115 cm downgradient from the granular bed inlet. The flow rate was monitored by weighing the inflow reservoirs during the entire course of the experiment. 4. Transport ofGellan Gum Microbeads in Soil Columns 67 ofVarious Grain Size Distributions

Table 4.3. Operational conditions of injection experiments in packed columns.

Column Parameter AB C D E F

Microbeads Size Distribution: a Range (/lm) 4 - 192 12- 34 12 - 40 12 - 40 12 - 34 12 - 22

dlO (/lm) 5.1 15.4 15.7 15.7 15.7 15.3 dso (/lm) 12.6 16.9 17.8 17.8 18 16.7

d90 (/lm) 32.3 18.6 21.2 21.2 21.9 18.1 I Microbeads Conc. (mg L- ) 40±5 44±8 37 ±4 42± 5 36±6 44±6 I Flow rate (L h- ) 0.465 0.443 0.446 0.499 0.433 0.477 I Ave. Pore Velocity (m h- ) 0.38 0.51 0.46 0.52 0.47 0.52 a dx is microbead diameter corresponding to the 10, 50, or 90% point of the cumulative distribution curve.

Transport of microbeads was evaluated from observed dimensionless concentration histories, CICo (the ratio of microbead concentration at sampling points downgradient to average injectant concentration vs. time), total breakthrough (TB), sectional total breakthrough (sTB), retention (RET), and sectional retention (sRET). The TB for microbeads was defined as the percentage of total (accumulative) mass of microbeads in samples collected downgradient at a given time relative to that in the injectant. Likewise, the sTB was defined as the percentage of total mass of microbeads in samples collected downgradient of a given section at a given time relative to that in samples collected upgradient of that section. The TB and sTB were numerically calculated from the following equations, J,rI Q.C(l, t)dt ] TB(l,t)= i~o=O x 100 (4.1) [ Q.C(O,t)dt 10 =0

f =0 Q.C(12,t)dt] sTB(12' t) = i~ x 100 (4.2) [ Q.C(ll't)dt 10=0 4. Transport ofGel/an Gum Microbeads in Soil Columns 68 ofVarious Grain Size Distributions where Q is the flow rate, t is the elapsed injection time, C(O,t) and C(l,t) are the time concentrations of microbeads in the injectant and at the distance l, respectively. The RET indicates the average percentage of microbeads immobilized per unit length (cm) of the soil matrix across a given (axial) distance after a certain time. Similarly, the sRET is defined as the relative amount of microbeads immobilized per unit length (cm) of a given section of the column. The RET and sRET were calculated as follows,

RET(l t) = TB (lI' t) - TB(l2' t) (4.3) 2' 1 -1 2 1

(4.4)

4.3.6. Analysis of Gellan Gum Microbeads The collected interstitial samples were analyzed using a total carbohydrate assay (Dubois et al., 1956). The method involved the reaction of gellan gum with a 5 g L- 1 solution of hydrazine sulfate in concentrated sulfuric acid which causes hydrolysis of the gellan gum polysaccharides to form hydroxymethyl furfurals from hexoses. The solution of this product was then treated with 5% (w/v) phenol to produce a colored compound. The treated sample was left at room temperature for one hour. The color intensity, which is a function of gellan gum concentration, was then measured at 490 nm and compared to a suspension of gellan gum microbeads with a known concentration which served as a standard.

4.4. Results The transport of microbeads was investigated in six horizontal columns (Table 4.1) packed with different size range classes of gravel (A: Dso =9.2 mm, B: Dso =5.8 mm, and C: Dso =

3.2 mm) and sand (D: Dso = 1.21 mm, E: Dso = 0.88 mm, and F: D so = 0.77 mm). A suspension of gellan gum microbeads in artificial groundwater (AGW) was introduced into the columns at about 0.43 - 0.5 L h- 1 for 6 h, followed by injection of bead-free AGW for 42 h (Table 4.3). The microbeads were produced by the emulsification-internal gelation method. Microscopically, the microbeads were uniformly spherical. Mean diameter of microbeads, 4. Transport ofGel/an Gum Microbeads in Soil Columns 69 ofVarious Grain Size Distributions which corresponds to the 50% point of the cumulative distribution curve, varied from 12.6 - 18.0 Ilm depending on the experimental conditions. The microbeads injected into column A had a wider range of size distribution compared to those injected into other columns. Nevertheless, these microbeads were used for transport experiment in column A since they were smaller than 480 Ilm, estimated as the pore size corresponding to 2-mm grains (Reddi and Bonala, 1997). Dimensionless concentration histories of gellan gum microbeads in columns C, D, and E are shown in Figures 4.2 - 4.4. The transport of microbeads through columns A, B, and F is compared with these columns, but the pertinent concentration profiles are not shown for the conciseness of this report. The breakthrough curves in columns C and D occurred within 0.25 h after injection and consisted of a major elementary peak at 5, 20 and 50 cm as weIl as in 80 cm (data not shown) travel distances. This elementary peak was considerably attenuated towards the column outlet and was no longer observed at llO-cm distance. Sorne individual minor peaks were observed in columns C and D at later stages of injection. The concentration histories in columns A and B exhibited a major elementary peak, which was observed at aIl sampling points across the columns. Concentration histories in columns E and F differed substantially from those in the other columns. The breakthrough curves demonstrated a major concentration peak at 5-cm distance, which was replaced by several discrete lower concentration peaks at downgradient distances. The frequency of appearance of such individual peaks was higher at 5- and 20-cm distances compared to 50- and llO-cm distances, indicating the filtration of microbeads upon travelling across the columns. The variations of total breakthrough (TB) of microbeads in the effluent with injection time are depicted in Figure 4.5. The TB profiles in columns A and B were similar during the 48-h injection, indicating 100 and 97% breakthrough of microbeads within this period, respectively. The TB of microbeads in column C was slightly lower than that of column D during the first 12 h. Afterwards, the breakthrough profiles gradually diverged and the TB of microbeads observed in the effluent of column C was about 18% less than that of column D after 48 h of injection. This can be attributed to the average pore water velocity in column C that was lower by Il%, as compared to column D (Table 4.3). 4. Transport ofGellan Gum Microbeads in Soil Co/umns 70 ofVarious Grain Size Distributions

1.0 "V

0.8 (a) -

0.6 0 - Ü Ü 0.4 -

0.2

0.0 1.0

0.8 (b)

o 0.6 Ü...... Ü 0.4

0.2

0.0 1.0

0.8 (c)

o 0.6 Ü...... Ü 0.4

0.2

0.0 1.0

0.8 (d)

o 0.6 Ü...... Ü 0.4

0.2

0.0 ~~WoII~.....~~~~.uJ.~~.L3W o 6 12 18 24 30 36 42 48 Time (h) Figure 4.2. Concentration histories for gellan gum microbeads injected into column C packed with 2 - 4 mm gravel. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient from the granular bed inlet. 4. Transport ofGellan Gum Microbeads in Soil Columns 71 ofVarious Grain Size Distributions

1.a ...... r-r"'T""T....,...... ,,....,...... ,.....,...... -r...,....,-.-,...,..T""T""'T""T....,...... ,.....,

0.8

o 0.6

§ 0.4

0.2

0.0 ~.J.....J...-ilIIIl."...... M4~~'-I-...... 1.a ...... T""T""'T""T....,...... ,,....,...... ,.....,...... -r...,....,-.-,...,..T""T""'T""T....,...... ,.....,

0.8 (b)

o 0.6 Ü...... Ü 0.4

0.2

0.0 -'-'- ....,r.;.-+-F-"_~_ 1.a r-T"""T""T-r-T""T"'""1r-T""T""T"...-r""""""'-'-"""'"T""T-r-T"-.....,

0.8 (c)

o 0.6 Ü...... Ü 0.4

0.2

O. a ...... L-.L...... L..iII--'L.J-...L...... L

0.8 (d)

o 0.6 Ü...... Ü 0.4

0.2

0.0 ...~WIlIiIWJ.l~...... ,..aJ4et~.....J.-.I-~ o 6 12 18 24 30 36 42 48 Time (h) Figure 4.3. Concentration histories for gellan gum microbeads injected into column D packed with 1 - 2 mm sand. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient from the granular bed inlet. 4. Transport ofGellan Gum Microbeads in Soil Columns 72 ofVarious Grain Size Distributions

1.0

0.8 (a)

0 0.6 Ü Ô 0.4

0.2

0.0 1.0

0.8 (b)

o 0.6 Ü...... Ü 0.4

0.2

0.0 1.0

0.8 (c)

o 0.6 Ü...... Ü 0.4

0.2

0.0 1.0

0.8 (d)

o 0.6 Ü...... Ü 0.4

0.2

0.0 llloij,.....~~....oUîl-'-l...4~~....~~.1...J o 6 12 18 24 30 36 42 48 lime (h) Figure 4.4. Concentration histories for gellan gum microbeads injected into column E packed with 0.5 - 2 mm sand. Breakthrough curves are for 5 cm (a), 20 cm (b), 50 cm (c), and 110 cm (d) downgradient from the granular bed inlet. 4. Transport ofGellan Gum Microbeads in Soil Columns 73 ofVarious Grain Size Distributions

Higher velocities are expected to yield greater breakthrough for a given travel distance. As the water velocity increases, the removal of microbeads by straining and sedimentation processes decreases due to the shorter retention times and greater fluid shear forces at higher velocities. Increased shear forces would likely result in a decrease in capture efficiency because of the greater fluid drag on partic1es near the media surface. The TB of microbeads in columns E and F was equal during the first 24 h of injection. However, by the end of experiment, the amount of microbeads discharged into the effluent from column E was slightly higher than that discharged from column F.

100

..-.. ~ 0 80 --..c 0> ~ 0 60 :lo... ..c...... ~ CO Q) 40 :lo... III ...... CO 0 20 1-

0 0 10 20 30 40 50 lime (h)

Figure 4.5. Variation of total breakthrough CTB) of gellan gum microbeads in the effluent of packed columns vs. time. Curves represent columns A ce), B C-), CC....), D CT), ECO), and F (0).

The TB of microbeads was a direct function of the travel distance, as illustrated by the data in Table 4.4. More than 97% of injected microbeads traveled through the entire 4. Transport ofGellan Gum Microbeads in Soil Columns 74 ofVarious Grain Size Distributions

1ength of both co1umns A and B after the 48-h injection. The value of TB in co1umn C was greater than 99% at 20 cm distance but the degree of attenuation progressive1y increased towards the co1umn outlet and on1y 44% of microbeads were discharged into the effluent after 48 h of injection. The TB of microbeads in co1umn D reduced a10ng the 1ength of the co1umn and reached 62% at the effluent after 48 h of injection. Likewise, the TB in co1umn E showed a decreasing trend as a function of the trave1 distance, and was about ha1f of that obtained in co1umn D after 48 h of injection. The experimental results indicated that microbeads trave1ed to a higher extent a10ng the first 50 cm of co1umn F, as compared to co1umn E. However, lower amount of microbeads discharged into the effluent of column F after 48-h injection.

Table 4.4. Total breakthrough (TB) and sectional total breakthrough (sTB) of gellan gum microbeads in soil columns after 48-h injection.

Column Section Distance, 1 Column (cm-cm) (cm) ABCD E F

TB a (%)

0-5 5 100 100 99.6 97.3 92.8 95.4 0-20 20 100 99.9 99.4 87.6 68.8 0-50 50 99.8 99.2 88 72.8 60 70.1 0-80 80 99.2 98.9 76.6 68.3 53.8 48.6 0-115 115 99.8 96.9 44.2 62.1 31.3 24.9

sTB b (%)

0-5 5 100 100 99.6 97.3 92.8 95.4 5 - 20 20 100 99.9 99.8 90.0 74.2 20 - 50 50 99.8 99.3 88.5 83.1 86.7 73.5 c 50 - 80 80 99.5 99.8 87.1 93.9 90.1 69.3 80 - 115 115 100 98 57.7 90.9 58.2 51.2 a Equation 1. b Equation 2.

C Calculated for the 5 - 50 cm section. 4. Transport ofGellan Gum Microbeads in Soil Co/umns 75 ofVarious Grain Size Distributions

Table 4.5. Retention (RET) and sectional retention (sRET) of gellan gum microbeads in soil columns after 48-h injection.

Column Section Length Column (cm-cm) (cm) A BC D EF

RET a (% cm'!)

0-5 5 0 0 0.076 0.538 1.444 0.914 5 - 20 15 0.002 0.006 0.013 0.648 1.597 20 - 50 30 0.007 0.025 0.381 0.494 0.306 0.563 c 50 - 80 30 0.018 0.008 0.380 0.148 0.196 0.717 80 - 115 35 0 0.057 0.926 0.178 0.642 0.677

sRET b (% cm'!) 0-5 5 0 0 0.076 0.538 1.444 0.914 5 - 20 15 0.002 0.006 0.013 0.666 1.721 20 - 50 30 0.007 0.025 0.384 0.564 0.444 0.590 c 50 - 80 30 0.018 0.008 0.431 0.203 0.329 1.022 80 - 115 35 0 0.057 1.209 0.261 1.195 1.394 a Equation 3. b Equation 4.

C Calculated for the 5 - 50 cm section.

The estimation of sectional total breakthrough (sTB) revealed that microbeads traveled through the major parts ofeach column at more or less equallevels, based on the total mass of microbeads entering a defined section of the column (Table 4.4). The exception to this general statement is the extent of transport through the 80 - 115 cm distance of columns C, E, and F, where the sTB of microbeads is considerably lower than 50 - 80 cm distance. Although the radiotracer and hydraulic conductivity tests indicated overall homogeneity ofthe soil matrix of each column, the creation of local heterogeneity due to the polydispersity and non-uniformity of the natural grains used for packing, may lead to non-uniform distribution of microbeads through the porous media. In fact, as it is shown in Table 4.5, the variations of retention (RET) or sectional retention (sRET) of microbeads with the differential travel distance after 48-h injection did not follow any 4. Transport ofGellan Gum Microbeads in Soil Columns 76 ofVarious Grain Size Distributions particular trend. However, the RET had a tendency to increase with a decrease in the mean grain size for any given differential distance. There was no evident decrease in hydraulic conductivity of the packed columns during and after injection of microbeads, suggesting that the use of a more concentrated suspension of microbeads is possible. Besides, the detection of microbeads in the effluent of aU columns suggests that microbeads could travel much longer distances through the granular media with a similar size distribution and hydrodynamic characteristics.

4.5. Discussion Successful application of the encapsulated ceU microbeads to in situ bioaugmentation depends strongly on uniform dispersion of microbeads through the soil matrix. Encapsulated cells should be evenly distributed at high densities throughout the contaminated zone of an aquifer to bring about the desired biodegradation activity. Using the encapsulation technique described earlier, we were able to encapsulate a bacterial culture at concentrations as high as 20 g cells per liter of gellan gum solution, corresponding to 2.54 x 107 cells mg- I beads (results not shown). Based on those results, the RET data suggest that under operational conditions used in this study (Table 4.3),5.5 x

5 5 10 to 7 X 10 encapsulated cells per gram of soil can be fairly distributed from an injection point across the llO-cm distance of a soil with characteristics similar to that in column F (Tables 4.1 and 4.2). The increase of cell mass loading in gel microbeads as well as injectant concentration, and prolongation of the injection process may provide an opportunity to deliver higher concentrations of microbial cells to the contaminated soil matrix. The TB and RET of microbeads are important design parameters as their variation with grain size distribution, travel distance, and injection time determines the distance required between the injection wells and the duration of injection for uniform distribution of encapsulated cells in a field application. Despite reduction of microbead concentration along the length of soil columns, the variation of RET with travel distance followed dissimilar trends. The RET values obtained at farther distances were comparable to those obtained for the first 50 cm. This can be partly due to several scattered concentration peaks, which appeared at later stages of 4. Transport ofGellan Gum Microbeads in Soil Columns 77 ofVarious Grain Size Distributions injection at such distances. The discontinuous concentration profiles can be attributed to one or combination of the following phenomena: 1) reduction in particle capture efficiency as the result of spontaneous release, 2) migration of particles through less direct flow paths, and 3) exclusion of particles from larger blocked pores. Secondary breakthrough of microbeads may be caused by the spontaneous release of previously captured microbeads due to normal flow of suspension through the soil bed. The spontaneous release may occur if local variations in pressure or flow rate (pore water velocity) change the flow in the neighborhood of retained microbeads or if a moving microbead collide with a retained microbead. The released microbeads as well as those freely suspended may be excluded from the larger blocked pores, being directed through unblocked but more tortuous flow paths formed by smaller pores. Therefore the transport of microbeads through lower permeability zone of the soil matrix may cause a delay in their movement, leading to their appearance at later stages of injection. The secondary peaks appearing at various distances in columns C, D, E, and F were more frequent than those at the same distances of columns A and B. The occurrence of release phenomena such as the increase in local interstitial velocity and the size exclusion effect were more likely to happen in the finer packing of columns C, D, E, and F. The experimental design did not allow differentiation between the breakthrough caused by the free passage (attenuated or unattenuated) of microbeads and that caused by the capture and release mechanism. Nevertheless, there should be a direct relationship between attenuation and capture of microbeads. The increased contact of microbeads with solid surfaces in columns C, D, E, and F occurring as a result of increased attenuation allows for greater possibilities of capture by grain collectors. The interactions between the microbeads and grains lead to filtration of microbeads due to several simultaneous capture mechanisms including mechanical (sedimentation, interception, and diffusion) and physico-chemical (sorption). Diffusion and sorption are relatively more important parameters in transport of microorganisms and microbial-sized particles « 5 /lm dia.), and may be less determinant in the transport of gellan gum microbeads, which were mostly greater than 12 /lm in this study. This statement can be supported by evaluating the single-collector efficiency (11+) 4. Transport ofGelian Gum Microbeads in Soil Columns 78 ofVarious Grain Size Distributions and comparing the magnitude of its elements representing diffusion (Tl0+) , interception (Tlt), and sedimentation (Tls+). The single-collector efficiency (Tl+) indieates the fraction of particles flowing toward a grain of porous medium, or collector, that actually collides with the collector, and is removed from suspension. A semi-empirieal relation referred to as the RT model (Logan, 1999; Logan et al., 1995; Rajagopalan and Tien, 1976), is used to calculate the above parameters based on the mean particle (dso) and grain (Dso) sizes for each column:

Tl+ = Tl~ + Tl7 + Tl; (4.5) Tl~ = 4.04y2b~3Pe -2/3 (4.6)

2 8 Tl7 b L 1f R*IS/8 (4.7) =y H o 2 Tl; =0.00338y bHS*1.2 (4.8) where y = (1-8)113, bH = 2(1-y)/(2-3y+3y-2Y'), Pe = Peclet No., Lo = London No., R* = dsoID so , S* = settling velocity/Darcy velocity, 8 = effective porosity, dso = particle mean diameter, and Dso =collector mean diameter. The variations of fractional collector efficiencies (Tli+/Tl+) for soil columns are shown in Figure 4.6. It is evident that diffusion has the least contribution to overall collector efficiency for aIl columns. Sedimentation is the major capture process in the gravel columns (A, B, and C) while the particle capture by interception dramatically increases in sand columns (D, E, and F). Microorganisms exhibit a natural tendency to adhere to solid surfaces depending on their hydrophobie characteristics and attachment capabilities. Therefore, bacterial transport through porous media can be limited to short distances by sorption and attachment to grain surfaces. For instance, transport of bacteria Klebsiella oxytoca and Burkholderia cepacia G4PR1 in columns of medium and coarse silica sand (0.25 - 1 mm) was described by reversible and irreversible sorption processes (Hendry et al., 1999). Maximum recoveries of bacteria in the effluent of lIA-cm columns were 32 and 12.6% for K. oxytoca and B. cepacia, respectively. In the present study, 70% of mierobeads were recovered at a distance of 50 cm in column F packed with 0.25 - 2 mm sand. Although a direct comparison is not warranted since the experimental conditions of the present study differed from those involved with free bacteria (Hendry et al., 1999; Jennings et al., 1995; Taylor 4. Transport ofGellan Gum Microbeads in Soil Columns 79 ofVarious Grain Size Distributions and Jaffé, 1990), the extensive transport of microbeads through the granular media observed in this study presumes that encapsulation of cells within polymeric gel carriers would avoid cell sorption to grain surfaces. This enhances the potential for the use of encapsulated cells for subsurface remediation applications.

0 Diffusion • Interception • Sedimentation 100 90 80 70 ..- ~ --0 60 + 50 ~ + 40 ~ 30 20 10 0 A B C D E F Column

Figure 4.6. Variation of fractional collector efficiency (11i+/11+) for diffusion (110+)' interception (11t), and sedimentation (11s+) mechanisms for soil columns; estimated based

3 20 on RT model using particle density of 1,008 kg m- , Hammaker constant of 10- J, fluid properties of water at 10 oC, and flow rates given in Table 4.3.

Although transport behavior of particles through the aquifer can markedly vary in space and depth due to the heterogeneity normally associated to the aquifer matrix (Harvey et al., 1993), laboratory investigation of particle transport in packed columns gives useful information on physical and chemical parameters that affect dispersion and migration of 4. Transport ofGellan Gum Microbeads in Soil Columns 80 ofVarious Grain Size Distributions particles through porous media. The present study was performed to gain an insight into transport characteristics of cell encapsulating gel microbeads through natural granular media. Each column was packed with a gravel or sand fraction of soil instead of mono-size grains to reproduce a heterogeneous porous media. The gravel and sand columns of this study may resemble grain size distribution of adjacent layers of an aquifer, and reflect the effect of aquifer heterogeneity on transport of cell encapsulating gel microbeads. It appears likely that the encapsulated cell microbeads can be implemented for in situ bioaugmentation of sorne contaminated aquifers mainly composed of the medium gravel to medium sand fractions of soil represented by the USGS standard classification system.

4.6. Conclusions Gellan gum gel microbeads ranging from 12 - 40 /lm, suitable for encapsulation of high concentrations of microbial ceIls, were produced by the emulsification-internal gelation technique. The breakthrough of microbeads pulsed into the soil columns of various grain size distributions increased with the injection time, reaching a maximum value of 25% for sand (0.25 - 2 mm), as compared to 100% for gravel (2 - 16 mm) after 48-h injection. Microbeads were dispersed at different levels (0.56 - 0.91 % cm-lof porous media) for various sections of sand column (0.25 - 2 mm) with the smallest mean grain size (0.77 mm), suggesting the feasibility of distribution of encapsulated cells along the length of soil porous media of similar characteristics. Detection of microbeads in the effluent of aIl columns suggests that transport of the tested microbeads through the granular media consisting of medium gravel to medium sand fractions of the USGS standard soil classification system is feasible, and the maximum achievable travel distance by the microbeads is greater than 110 cm. Based on the results obtained from this study, transport and deposition of gellan gum microbeads is influenced by grain size distribution, travel distance, and injection time.

Acknowledgments The authors thank the Natural Sciences and Engineering Research Council of Canada and National Research Council of Canada (NRC paper no. 00000) for financial support, and 4. Transport ofGellan Gum Microbeads in Soil Columns 81 ofVarious Grain Size Distributions

McGill University for the awarding of the Max Stern Fellowship ta P.M. The technical support of Anne-Marie LeBlanc, Gilles Grimal, and Laurent Munier is appreciated. 4. Transport ofGellan Gum Microbeads in Soit Columns 82 ofVarious Grain Size Distributions

References Arya, L. M. and Dierolf, T. S. (1989). Predicting soil moisture characteristics from particle-size.distributions: An improved method to calculate pore radii from particle radii. International Workshop on Indirect Methods for Estimating the Hydraulic Properties of Unsaturated Soils, The University of California, Riverside, CA, 115­ 124. Arya, L. M. and Paris, J. F. (1981). A physieoempirical model to prediet the soil moisture characteristic from particle-size distribution and bulk density data. Soil Science Society of America Journal 45: 1023-1030. BaIes, R. C., Li, S., Yeh, T.-C. J., Lenczewski, M. E. and Gerba, C. P. (1997). Bacteriophage and mierosphere transport in saturated porous media: Forced­ gradient experiment at Bordon, Ontario. Water Resources Research 33: 639-648. Brandenberger, H. and Widmer, F. (1998). A new multinozzle encapsulation/ immobilization system to produce uniform beads of alginate. Journal of Biotechnology 63: 73-80. Buitelaar, R. M., Hulst, A. C. and Tramper, J. (1988). Immobilization of biocatalysts in thermogels using the resonance nozzle for rapid drop formation and an organie solvent for gelling. Biotechnology Techniques 2: 109-114. Cassidy, M. B., Lee, H. and Trevors, J. T. (1996). Environmental applications of immobilized microbial cells: A review. Journal of Industrial Microbiology 16: 79­ 101. Devlin, J. F. (1994). A simple and powerful method of parameter estimation using Simplex Optimization. Ground Water 32: 323-327. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A. and Smith, F. (1956). Colorimetrie methods for determination of sugars and related substances. Analytieal Chemistry 28: 350-356. Freeze, R. A. and Cherry, J. A. (1979). Groundwater. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Giavasis, 1., Harvey, L. M. and MeNeil, B. (2000). Gellan gum. Critical Reviews in Bioteehnology 20: 177-211. 4. Transport ofGellan Gum Microbeads in Soil Co/umns 83 ofVarious Grain Size Distributions

Goldstein, R. M., Mallory, L. M. and Alexander, M. (1985). Reasons for Possible Failure of Inoculation to Enhance Biodegradation. Applied and Environmental Microbiology 50: 977-983. Hammill, T. B. and Crawford, R. L. (1997). Bacterial microencapsulation with three algal polysaccharides. Canadian Journal of Microbiology 43: 1091-1095. Harmand, B., Rodier, E., Sardin, M. and Dodds, J. (1996). Transport and capture of submicron particles in natural sand: Short column experiments and a linear model. Colloids and Surfaces. A: Physicochemical and Engineering Aspects 107: 233-244. Harvey, R. W., George, L. H., Smith, R. L. and LeBlanc, D. R. (1989). Transport of microspheres and indigenous bacteria through a sandy aquifer: Results of natural­ and forced-gradient tracer experiments. Environmental Science and Technology 23: 51-56. Harvey, R. W., Kinner, N. E., MacDonald, D., Metge, D. W. and Bunn, A. (1993). Role of physical heterogeneity in the interpretation of small-scale laboratory and field observations of bacteria, microbial-sized microsphere, and bromide transport through aquifer sediments. Water Resources Research 29: 2713-2721. Hendry, M. J., Lawrence, J. R. and Maloszewski, P. (1999). Effects of velocity on the transport of two bacteria through saturated sand. Ground Water 37: 103-112. Jegatheesan, V. and Vingeswaran, S. (1997). Interaction between organic substances and submicron particles in deep bed filtration. Separation and Purification Technology 12: 61-66. Jennings, D. A., Petersen, J. N., Skeen, R. S., Hooker, B. S., Peyton, B. M., Johnstone, D. L. and Yonge, D. R. (1995). Effects of slight variations in nutrient loadings on pore plugging in soil columns. Applied Biochemistry and Biotechnology, Part A­ Enzyme Engineering and Biotechnology 51/52: 727-734. Kang, K. S., Veeder, G. T., Mirrasoul, P. J., Kaneko, T. and CottreIl, 1. W. (1982). Agar­ like polysaccharide produced by a Pseudomonas species: Production and basic properties. Applied and Environmental Microbiology 43: 1086-1091. Logan, B. E. (1999). Environmental Transport Processes. John Wiley & Sons, Inc., New York. 4. Transport ofGellan Gum Microbeads in Soil Columns 84 ofVarious Grain Size Distributions

Logan, B. E., Jewett, D. G., Arnold, R. G., Bouwer, E. J. and O'Melia, C. R (1995). Clarification of clean-bed filtration models. Journal of Environmental Engineering 121: 869-873. McCaulou, D. R, BaIes, R C. and Arnold, R G. (1995). Effect of temperature-controlled motility on transport of bacteria and microspheres through saturated sediment. Water Resources Research 31: 271-280. Moran, D. c., Moran, M. c., Cushing, R S. and Lawler, D. F. (1993a). Particle behavior in deep-bed filtration: Part l-Ripening and breakthrough. American Water Works Association Journal 58: 69-81. Moran, D. c., Moran, M. c., Cushing, R S. and Lawler, D. F. (1993b). Particle behavior in deep-bed filtration: Part 2-Particle detachment. American Water Works Association Journal 85: 82-93. Morris, D. A. and Johnson, A. 1. (1967). Summary of Hydrologic and Physical Properties of Rock and Soil Materials, as Analyzed by the Hydrologic Laboratory of the U.S. Geological Survey 1948-1960, U.S. Geological Survey Water Supply Paper 1839­ D. Nilsson, K., Birnmaum, S., Flygare, S., Linse, L., Schroder, U., Jeppsson, U., Larsson, P.­ O., Mosbach, K. and Brodelius, P. (1983). A general method for the immobilization of cells with preserved viability. European Journal of Applied Microbiology and Biotechnology 17: 319-326. Ogata, A. (1970). Theory of dispersion in a granular medium. Fluid Movement in Earth Materials-U.S. Geological Survey Professional Paper 411-1. Washington, United States Government Printing Office. Ogbonna, J. c., Matsumura, M., Yamagata, T., Sakuma, H. and Kataoka, H. (1989). Production of micro-gel beads by a rotating disk atomizer. Journal of Fermentation and Bioengineering 68: 40-48. Petrich, C. R, Stromo, K. E., Ralston, D. R and Crawford, R L. (1998). Encapsulated cell bioremediation: Evaluation on the basis of particle tracer tests. Ground Water 36: 771-778. Poncelet, D., Lencki, R, Beaulieu, c., Halle, J. P., Neufeld, R J. and Fournier, A. (1992). 4. Transport ofGel/an Gum Microbeads in Soil Columns 85 ofVarious Grain Size Distributions

Production of alginate beads by emulsificationlinternal gelation. 1. Methodology. Applied Microbiology and Biotechnology 38: 39-45. Poncelet, D., Poncelet De Smet, B., Beaulieu, C., Huguet, M. L., Fournier, A. and Neufeld, R J. (1995). Production of alginate beads by emulsification/internal gelation. II. Physicochemistry. Applied Microbiology and Biotechnology 43: 644-650. Pritchard, P. H. (1992). Use of inoculation in bioremediation. Current Opinion in Biotechnology 3: 232-243. Rajagopalan, R and Tien, C. (1976). Trajectory analysis of deep-bed filtration with the sphere-in-cell porous media model. AIChE Journal 22: 523-533. Reddi, L. N. and Bonala, M. V. S. (1997). Analytical solution for fine particle accumulation in soil filters. Journal of Geotechnical and Geoenvironmental Engineering 123: 1143-1152.

Sanderson, G. R, Bell, V. L. and Ortega, D. (1989). A comparison of gellan gum, agar, l(­ carrageenan, and algin. Cereal Foods World 34: 991-998. Sanderson, G. R and Clark, R C. (1983). Gellan gum. Food Technology 37: 62-70. Taylor, S. W. and Jaffé, P. R (1990). Biofilm growth and the related changes in the physical properties of a porous medium: 1. Experimental investigation. Water Resources Research 26: 2153-2159. Vogel, T. M. (1996). Bioaugmentation as a soil bioremediation approach. CUITent Opinion in Biotechnology 7: 311-316. 86

5. Transport of Gellan Gum Microbeads through Sand: An Experimental Evaluation for Encapsulated Cell Bioaugmentation

Peyman Moslemy a,b, Ronald J. Neufeld c, Denis MUlette b,t, and

Serge R. Guiot b

a Department of Chemical Engineering, McGill University, 3610 University Street,

Montreal, Quebec, Canada H3A 2B2

b Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

C Department of Chernical Engineering, Queen's University, Dupuis Hall, Kingston, Ontario, Canada K7L 3N6

t Current address: Hydrogéo Plus Inc., 3333 Queen-Mary Road, Suite 501, .Montreal, Quebec, Canada H3V lA2

This study was carried out to obtain the information on distribution patterns of gellan gum microbeads injected into porous sand media, and to elucidate the influence of sand grain size distribution and injectant concentration on the extent of transport during extended periods of injection. The candidate has carried out all the work in this chapter. Dr. Denis Millette contributed by providing his expertise in the field of Hydrogeology, and by offering insight to the design of soil columns. This article has been submitted to the journal ofAdvances in Environmental Research for publication. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 87 for Encapsulated Cell Bioaugmentation

5.1. Abstract Transport of 10 - 40 /lm gellan gum microbeads was studied in horizontal sand columns to evaluate the feasibility of using gel-encapsulated bacteria for bioaugmentation of contaminated aquifers. Three 5.2 x 110 cm columns were packed with sand (A: 0.5 - 2 mm, B: 0.25 - 2 mm, C: 0.125 - 2 mm). Microbeads in artificial groundwater were injected at 0.5 L h- 1 during intermittent 12-h periods. Breakthrough increased with injection time, varying as a descending function of travel distance. After 72 h of injection, about 75% of microbeads were dispersed across a 5 - 110 cm distance in column A, compared to 78% across a 5 - 50 cm in column B, and 76% across a 5 - 20 cm in column C. The wider dispersion of microbeads across the length of column A compared to those observed in columns Band C, suggests a higher potential for the formation of a uniform bioactive zone of encapsulated cells across a sandy aquifer with such grain size distribution and hydrodynamic properties.

Keywords: Gellan gum; Encapsulation; Emulsification; Microbead; Transport, Soil Column

5.2. Introduction Bioaugmentation of a contaminated aquifer involves direct injection of active degrading bacteria to enhance the biodegradation of contaminants (Vogel, 1996). A major impediment to bioaugmentation is the filtration of injected bacteria during their transport through soil porous media. The concentration of suspended cells can be reduced, mainly due to attachment to soil grains within short distances of the injection weIl. It has been shown that attachment to sand surfaces stimulates synthesis of exopolymers in subsurface bacterial isolates (Vandevivere and Kirchman, 1993). The growth of attached bacteria and development of interstitial biofilms can cause a permanent decrease in soil permeability, leading to weIl c10gging and failure of the bioremediation process. Bacterial attachment can be substantially altered through changes in carrier fluid chemistry, cell hydrophobicity and surface charge. The reduction of ionic strength of the carrier fluid and the addition of nonionic surfactants substantially reduced the bacterial attachment to glass, quartz, and soil 5. Transport ofGel/an Gum Microbeads Through Sand: An Experimental Evaluation 88 for Encapsulated Cel/ Bioaugmentation surfaces for Alcaligenes paradoxus culture, and to a lesser degree for a subsurface isolate, within 1-cm long mini-columns (Li and Logan, 1999). However, these treatments were stated to be insufficient to increase the transport distance of cells for field applications. In another study, reporting the influence of an anionic biosurfactant on the transport of three strains of Psuedomonas aeruginosa with different cell hydrophobicities, about 50 - 80% of the injected bacteria were immobilized within 5-cm long sand columns due to sorption (Bai et al., 1997). Encapsulation of bacteria within polymeric gel microbeads may enhance transport distances for bioaugmentation to occur over a large region of the subsurface. Encapsulation provides a secluded microenvironment ta the bacteria, preventing their attachment to soil grains while protecting them from the exterior biotic and abiotic stresses. Several studies have previously addressed the transport of suspended polymeric microspheres as a function of size distribution of particles, and grain size distribution and heterogeneity of porous media (BaIes et al., 1997; Harmand et al., 1996; Harvey et al., 1989; Harvey et al., 1993; McCaulou et al., 1995; Petrich et al., 1998; Silliman, 1995). Monodisperse micrometer­ and submicrometer-sized latex microspheres have been mostly used in deep-bed filtration studies using cm-long columns, or as tracers of bacteria in transport studies under laboratory or field conditions. Cell carriers should be small enough to be capable of being transported to the site of contamination, and yet be large enough to encapsulate an adequate number of bacteria to provide a high biocatalytic activity within the microbead matrix. For instance, if the number of encapsulated spherical cells (cocci) is assumed to be proportional to the cubic ratio of microbead and cell diameters, then the number of cells that can be potentially entrapped in a lO-llm microbead will be three orders-of-magnitude greater than that in a 1-llm particle. The main objective of the present study was to evaluate the feasibility of in situ encapsulated cell bioaugmentation on the basis of transport characteristics of gel microbeads through fine to very coarse fractions of sand (0.125 - 2 mm), which generally comprise the main body of an aquifer material. As mentioned above, microbeads of at least 10 Ilm (dia.) are better candidates for subsurface bioremediation applications. Therefore, in this study, gellan gum microbeads of 10 - 40 Ilm (dia.) were produced with a two-phase 5. Transport ofGeUan Gum Microbeads Through Sand: An Experimental Evaluation 89 for Encapsulated eeu Bioaugmentation dispersion technique, suitable for large-scale production of encapsulated cells. According to a semi-empirical model (Reddi and Bonala, 1997), describing the relationship between the interstitial pore and grain radii for an assemblage of uniformly sized spherical particles in a cubic packing, the pore size corresponding to the above grain size ranges from 8 to 480 Ilm. This wide range of pore size suggests that gellan gum microbeads can be potentially transported through the fine to very coarse sand porous media. To verify this potential, the transport of microbeads was studied by conducting forced-gradient once­ through experiments with intermittent injection of microbeads into relatively long sand columns (110 cm) of various grain size distributions. The use of long columns in this study rather than mini-columns may provide results that can be more representative of large­ scale bioaugmentation schemes. The transport and retention of microbeads were assessed as a function of grain size distribution, travel distance, and injection time.

5.3. Materials and Methods 5.3.1. Chemicals Gellan gum (Kelcogel®) was a donation from the CP Kelco US, Inc., formerly NutraSweet Kelco Company (San Diego, CA, USA). Gellan gum is produced by the microorganism Sphingomonas paucimobilis (ATCC 31461), earlier referred to as Pseudomonas elodea, through fermentation processes (Giavasis et al., 2000). Gellan gum gel has superior rheological properties to agar and carrageenan gels at equivalent concentrations (Sanderson et al., 1989). The gel has been stable over the wide pH range of 2 - 10 (Ashtaputre and Shah, 1995), suggesting its suitability for use in both acidic and basic environments. The application of gellan gum for encapsulation of viable cells has been addressed at considerably lower concentrations of both gel and gelling agent comparing to K­ carrageenan, agar, and alginate (Buitelaar et al., 1988; Nilsson et al., 1983). Unlike sorne other ion-sensitive gelling polysaccharides such as alginate, the reactivity between gellan gum and ions is non-specific and gels can be formed with a wide variety of cations including alkaline and alkaline-earth cations (Moorehouse et al., 1981; Sanderson and Clark, 1983). Use of gellan gum as an entrapment matrix has been recommended in fermentation processes due to its mechanical and thermal stability (Camelin et al., 1993; 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 90 for Encapsulated Cell Bioaugmentation

Norton and Lacroix, 1990). A combination of these characteristics makes geIlan gum a premium natural polymer for encapsulation of active microorganisms with promising performance in subsurface applications. Canola oil was purchased from a local food distributor. AIl other chemicals were of reagent-grade, purchased from commercial suppliers.

5.3.2. Microbead Production Microbeads of geIlan gum were produced by a two-phase dispersion technique termed emulsification-internal gelation. A 0.75% (w/v) dispersion of geIlan gum was dissolved in 55 mL deionized water by heating to 75 - 80 oc. The solution was then cooled to 45 oC after adding CaClz at 0.06% (w/v), and its pH was adjusted to 7 with 0.1 N NaOH. The pregel solution was then emulsified in 330 mL of canola oil in a 1-L stirred baffled cylindrical vessel (10 cm dia.) held at 45 oC. For bacterial encapsulation, a ceIl culture can be dispersed in the pregel solution before emulsification. The resulting water-in-oil emulsion aided by 0.1 % (w/w) Span 80 as an emulsifier, was stirred vigorously at 4500 rpm using a quarter-circular paddle (5 cm dia.) for 10 min. The gelation of geIlan gum droplets was triggered by cooling the reaction vessel with chilled water to 15 oC with continued stirring. Microbeads were separated from the oil dispersion after 2 h by partitioning into 0.1 % (w/v) calcium chloride solution (0.5 L). The oil phase was removed by aspiration and the microbeads were rinsed repeatedly with a 0.1 % (v/v) Tween 80 solution. The size distribution and mean diameter of microbeads were measured by means of a Malvern particle size analyzer (series 2600, Malvern Instruments, Inc., Southborough, MA, USA). A fresh batch of microbeads was produced for each column experiment.

5.3.3. Artificial Groundwater Artificial groundwater (AGW) was prepared with a chemical composition similar to a natural groundwater (McCaulou et al., 1995), containing (mg per liter of distilled water):

MgS04.7HzO 69.0; NaHC03 50.0; CaClz.2HzO 14.5; Ca(N03)z.4HzO 64.0; and KF 2.0. The pH was adjusted to 7.5 with 1 N NaOH. 5. Transport ofGel!an Gum Microbeads Through Sand: An Experimental Evaluation 91 for Encapsulated Cel! Bioaugmentation

5.3.4. Porous Media and Columns The porous media was derived from near-surface sediments of a location along a well­ documented esker in Ville-Mercier (Quebec, Canada). The collected material was thoroughly mixed, and then screened using standard sieves (US Standard Testing Sieve, VWR Scientific Products, West Chester, PA, USA) and a mechanical sieve shaker (model RX-29, W. S. Tyler, Mentor, OH, USA) to obtain one of the pre-defined grain-size classes of the US Geological Survey (USGS) standard soil classification system (Morris and Johnson, 1967). Transport of gel microbeads was studied in three columns (A, B, and C) covering the different size range categories of sand of the USGS soil classification system (Table 5.1). A schematic description of the packed column system is given in Figure 5.1. Each column consisted of a Schedule-40 transparent PVC tube (5.2 cm id x 110 cm long), plugged with PVC fittings at both ends. Glass beads of 0.6 cm in diameter were placed at the inlet and outlet of each column to a depth of 2.5 cm to support and immobilize the granular bed, and to ensure proper liquid distribution and discharge at the inlet and outlet of the column. The effective length of the sand matrix was 110 cm.

Table 5.1. Grain size distribution in packed columns based on the standard soil classification of the US Geological Survey (USGS).

Column A B C Grain Type Size (mm) Weight (%) Very Coarse Sand 1-2 34.4 26.6 22.2 Coarse Sand 0.5 - 1 65.6 50.8 42.3 Medium Sand 0.25 - 0.5 22.6 18.8 Fine Sand 0.125 - 0.25 16.7

Mean grain size, Dso (mm) 0.88 0.77 0.68 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 92 for Encapsulated Cell Bioaugmentation

8 6

10

11 •, Rubber TSeptum Sampling Port ..• Washer 2_ 1

Figure 5.1. Schematic of the packed column setup: (1) Scale, (2) Magnet Stirrer, (3) Inflow Reservoir, (4) Shut-off Valves, (S) Masterflex Pump, (6) Digital Controller, (7) Sampling Ports, (8) Piezometrie Tubes, (9) Packed Column, (10) Anti-siphon Tube, (11) Outflow Reservoir.

Eight sampling ports were installed at 15-cm intervals across the column. One sampling port was installed before the inlet of the column, and another after the outlet to allow verification of the inflow, and monitoring of the outflow concentration. Each sampling port consisted of a male adapter (type B, 1/4" x 1/4" NPT, Cole-Parmer Instrument Co., Vernon Hills, IL, USA), a 12.S-mm rubber septum, and a steel tube (2 mm 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 93 for Encapsulated Cell Bioaugmentation

id X 40 mm long). The adapter was screwed into 114" NPT-threaded openings spired along the column before packing with grains. After packing, a reusable hypodermic Luer-Lock needle (B-D Yale, 15G x 3 112", VWR Canlab, Ville Mont Royal, QC, Canada), connected to a standard Mininert valve (Supelco, Sigma-Aldrich Canada Ltd., Oakville, ON, Canada), and a 3-mL plastic syringe, was inserted into the septum and passed through the adapter and steel tube to reach the surface of packed grains at the centerline of the column. This configuration allowed the collection of interstitial samples from the centerline of each column. The granular material was emplaced under standing water with the column held in a vertical position to minimize air entrapment, and was tamped down with a steel piston during filling to attain a uniform packing. A Masterflex variable speed modular digital drive/dispenser (model 77300, Cole­ Parmer Instrument Co., Vernon Hills, IL, USA) was used to circulate water across the columns. Bach column was equipped with three inlet and outlet ports to ensure uniform distribution and discharge of the liquid medium over the packed bed cross-section. The column effluent was directed through an air-gap anti-siphon tube hold at a fixed elevation and then discharged into the outflow reservoir. Vertical piezometrie tubes were installed at both ends, and along the column for measuring the differential head across the bed. The entire column system was enclosed in a laboratory refrigerator to maintain a constant temperature of 10 oC, which represents the temperature typically measured in groundwater. The column was placed in a horizontal position and operated in a horizontal-flow mode.

5.3.5. Hydrodynamic Properties The hydrodynamic properties of the porous media in the sand columns were determined by conducting a non-reactive radiotracer (tritium) test and a hydraulic conductivity test. Data are summarized in Table 5.2. The tracer test was initiated by injecting tritium-amended AGW at 0.46 L h- 1 for columns A and B, and 0.52 L h- 1 for column C. At regular intervals, 0.5 mL pore water samples were collected from sampling ports located at 5, 20, 50,80, and 110 cm downgradient from the packed bed inlet. The radioactivity of the samples was determined using a liquid scintillation analyzer (model TRI-CARB 2100 TR, Packard Instrument 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 94 for Encapsulated Cell Bioaugmentation

Company, Meriden, CT, USA). The experimental data were used to calculate the longitudinal dispersivity based on the theory of the advective-dispersive transport for a non-reactive solute in an isotropie homogeneous saturated porous medium (Ogata, 1970), and using a Simplex Optimization computer program (Devlin, 1994). The effective porosity was then obtained using the definition of average linear velocity in porous media (Freeze and Cherry, 1979).

Table 5.2. Characteristies of the porous media in sand columns

Column Parameter A B C Effective Porosity 0.44 0.43 0.41 Hydraulic Conductivity (cm S-l) 0.147 0.109 0.021 2 Permeability (cm ) 2.21e-06 1.46e-06 2.73e-07 Dispersivity (cm) 0.387 0.397 0.674 Collector Efficiency a 6.20e-03 7.82e-03 1.22e-02 a Calculated from equation 5.2 based on the grain mean size (Dso) and the microbead mean

diameter (dso).

The hydraulic conductivity test was carried out to obtain the values of hydraulic conductivity and permeability of the sand matrices. The test was initiated by pumping AGW into the column and measuring the flow rate and differential head across the column at regular intervals. The hydraulic conductivity of the sand matrix was obtained from the slope of the linear curve of the specifie discharge versus hydraulic gradient data (Freeze and Cherry, 1979). Permeability of the granular matrix was then estimated from its mathematical relationship with hydraulic conductivity (Freeze and Cherry, 1979).

5.3.6. Transport in Porous Media Forced-gradient experiments with intermittent input of gellan gum mierobeads were used to study the transport of microbeads through the sand matrix in each column. The size 5. Transport ofGeUan Gum Microbeads Through Sand: An Experimental Evaluation 95 for Encapsulated eeu Bioaugmentation distribution of microbeads injected into the columns is summarized in Table 5.3. The microbeads had a unimodal distribution, and microscopically were observed to be spherical. At least 2 pore volumes of water were displaced by injection of pure AGW at about 0.50 L h-! prior to injection of microbeads. A suspension of microbeads in AGW was then introduced into the columns during successive injection phases, under the conditions given in Table 5.3. Despite continuous agitatiqn of the microbead suspension in the inflow reservoir during injection, the concentration of the injectant slightly fluctuated during the injection period. These fluctuations can be attributed to the microbead settlement in the injection tubes, or disturbance caused in the inflow stream during sampling. Samples of water (3 mL) were collected from the inlet and outlet sampling ports, and from the sampling ports located at 5, 20, 50, 80, and 110 cm downgradient from the packed bed inlet. The outlet sampling port was considered to be 115 cm downgradient from the porous bed inlet. The flow rate was monitored by weighing the inflow reservoir during the entire course of the experiment. Transport of microbeads was evaluated from observed concentration histories (microbead concentration vs. time at sampling points downgradient), total breakthrough (TB), and retention percentage (RP). The TB for microbeads was calculated as the percentageof total (accumulative) mass of microbeads in samples collected downgradient at a given time normalized to that in the injectant. The RP indicates the percentage of microbeads retained within a given section of the sand matrix after a certain time, calculated as RP = (TB)! - (TB)2 (5.1) where (TB)! and (TB)2 are the values of total breakthrough at given travel distances of the column. Size distribution of microbeads in the samples collected from 20, 50, 80, and 110 cm travel distances of column Cafter 50, 60, 70, and 80 h of injection was measured by using a Coulter particle and size analyzer (model Z2, Coulter Particle Characterization, Hialeah, FL, USA). 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 96 for Encapsulated Cell Bioaugmentation

Table 5.3. Operational conditions of injection experiments in packed columns.

Column Parameter A B C

Size Distribution (Size in /lm): a Size Range 12 - 40 12 - 30 10 - 40

dlO 15.4 15.2 15.7 dso 17.0 16.4 18.5

d90 19.0 17.7 26.0 dmode 18.9 18.9 18.9

1 Suspension Concentration, Co (mg L- ): b Phase 1 42.1 ± 6.3 43.3 ± 9.2 52.0 ± 8.3 Phase II 77.0 ±4.8 75.5 ± 6.2 70.0 ± 5.8 Phase III 100.5 ± 7.0 Phase IV 0 Phase V 99.9 ± 6.3

1 Flowrate, Q (L h- ): Phase 1 0.474 0.486 0.498 Phase II 0.474 0.486 0.498 Phase III 0.498 Phase IV 0.438 Phase V 0.270

1 Darcy Velocity (m h- ): Phase 1 0.22 0.23 0.23 Phase II 0.22 0.23 0.23 Phase III 0.23 Phase IV 0.21 Phase V 0.13

a d lO, dso , and d90 are microbead diameters corresponding to the 10, 50, and 90% point of the cumulative distribution curve, respectively. dmode is mode diameter representing the most frequent microbead diameter. b Phase 1: 0 - 48 h, Phase II: 48 - 72 h, Phase III: 72 - 87 h, Phase IV: 87 - 119 h, Phase V: 119 - 128 h 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 97 for Encapsulated Cell Bioaugmentation

5.3.7. Analysis of Gellan Gum Microbeads The collected interstitial samples were analyzed using a total carbohydrate assay (Dubois et al., 1956). The method involved the reaction of gellan gum with a 5 g L- 1 solution of hydrazine sulfate in concentrated sulfuric acid which causes hydrolysis of the gellan gum polysaccharides to form hydroxymethyl furfurals from hexoses. The solution of this product was then treated with 5% (w/v) phenol to produce a colored compound. The treated sample was left at room temperature for one hour. The color intensity, which is a function of gellan gum concentration, was then measured at 490 nm using a spectrophotometer and compared to a suspension of gellan gum microbeads with a known concentration that served as standard.

5.4. Results and Discussion The transport of gellan gum microbeads was investigated in three horizontal columns packed with different size range classes of sand (A: Dso =0.88 mm, B: Dso =0.77 mm, and C: Dso = 0.68 mm) to study the compositional effect of each class on the transport and distribution of microbeads through a sandy porous media (Table 5.l). A suspension of microbeads in artificial groundwater (AGW) was introduced into the columns at about 0.5 L h- 1 during intermittent l2-h periods. The experiments on columns A and B were performed in two injection phases (Table 5.3), which lasted for 48 h (phase 1) and 24 h (phase II). The injectant concentration was doubled in phase II to examine the variations of microbead breakthrough in response to a step change in the inlet concentration. The injection in column C was carried out in five steps (Table 5.3), subsequently lasting for 48 h (phase 1), 24 h (phase II), 15 h (phase III), 32 h (phase IV), and 9 h (phase V). The injectant concentration was increased by approximately 40% in phase II as well as in phase III of injection in column C. The concentration histories of microbeads at trave1 distances of 5, 50, and 110 cm through sand columns A, Band C are shown in Figure 5.2. The breakthrough curves in columns A and B had similar features although lower concentrations were generally observed in column B during phase II of the experiment. The microbead concentration at a

1 5 cm distance increased to the injection levels (40 mg beads L- ) within 3 h of injection for 5. Transport-ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 98 for Encapsulated Cell Bioaugmentation

(1) (II) •• 100 A ...

~ ...J 80 Cl 5 • "ct! 60 • • ..cal

e() 40 ~ 20 O~~~~~~L~.J 120 0r-r--1,.....,12--,-----;_2T4---r-.-_3,6-,---,-~4,8....-....-6,0----;---r---,72 (1) (II) 100 B

~ ...J 80 Cl 5 , . -g 60 • ..cal e() 40 ~ 20 o .~~~~;;!i~~~ o 12 24 36 48 160 (1) •• (II) (III).~ (IV) (V) 140 c • ~ 120 '...J •• Cl 100 5 80

60

40

20

12 24 36 48 60 72 84 96 108 120 132 Time (h)

Figure 5.2. Concentration histories of gellan gum microbeads in columns A, B, and C during various injection phases. Breakthrough curves are for 5 cm ce), 50 cm CD), and 110 cm CA) downgradient from the bed inlet. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 99 for Encapsulated Cell Bioaugmentation column A and 1 h for column B in phase l, and remained at injection levels for the entire period of this phase. During phase II, the microbead concentrations at a 5 cm distance reached the injectant concentration within 12 h for column A as compared to 4 h for column B. The prolonged transient stage experienced in phase II, compared to phase l, could be due to the increased filtration effect of the sand matrix at higher concentration of injectant. The concentrations of microbeads at a travel distance of 50 and 110 cm along column A generally increased with time but were lower compared to the 5-cm distance. The concentration of microbeads at 50 and llO-cm distances of column B varied mainly between 0 and 30 mg beads L- I during injection. The breakthrough curves in column C followed different patterns in comparison with columns A and B (Figure 5.2). The concentration of microbeads remained mainly below 40 mg beads L- I at a 5 cm distance during phase 1 but increased to higher levels at the beginning of phase II, and exceeded the injectant concentration in several pore samples. The concentration varied around the injection levels (100 mg beads LI) during phase III and then decreased to close-to-zero values when the inflow was switched to bead-free AGW after 87 h of injection (phase IV). During phase III, the hydraulic gradient was progressively increased within a 20 cm distance from the bed inlet. Accordingly, the injection of microbead suspension was terminated after 87 h due to the limitation of piezometric measurements by the experimental set-up. The concentration rose to about 80 mg beads L- I when the injection of suspension was restored in phase V. The concentration of several pore water samples collected at a 5 cm distance was higher than the injectant concentration during the earlier stages of phases 1 and II in columns A and B, and phase II in column C. The capture and accumulation (deposition) of the injected microbeads within a short distance from the sand bed inlet is the likely explanation for the high concentration of microbeads observed near the bed inlet. Subsequent increase in pore water velocity may cause secondary release of previously captured particles, leading to microbead re-mobilization resulting in a concentration higher than the injectant. The increase in deposition of microbeads within a small volume of the sand matrix may also contribute to ~he re-entrainment of microbeads into the flow stream due to fluid shear. This phenomenon was more pronounced in column C than other 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 100 for Encapsulated CeU Bioaugmentation columns, indicating an intense filtration effect by the sand matrix of this column. The average pore velocity, which is the quotient of the division of Darcy velocity by the effective porosity, was higher for column B than column A by 7%. Lower average pore velocity contributes to greater accumulation of microbeads within 5 cm distance of column A, leading to liberation of higher concentration of microbeads during the transient stage in phase II for this column compared to column B. The variation of hydraulic gradient with injection time is depicted in Figure 5.3. The hydraulic gradient within a 5 - 27.5 cm and 27.5 - 50 cm distance of column A increased from about 0.04 to 0.06 cm H20 cm-! of porous media during 72 h of injection. The hydraulic gradient within 5 - 20 cm and 20 - 35 cm distances of column B showed an overall increase from about 0.07 to 0.16 and 0.15 cm H20 cm-! of porous media, respectively. The hydraulic gradient within a 5 - 20 cm distance of column C increased from 0.37 to 0.55 cm H20 cm-! of porous media during phases 1 and II, and then increased substantially to 3.03 cm H20 cm-! of porous media in phase III. The injection of particle­ free AGW (Phase IV, Table 5.3) had minimal effect on displacement of the captured particles from the 5 - 20 cm distance since no decrease in the hydraulic gradient was observed. Overall, the hydraulic gradient within a 5 - 20 cm distance of column C increased by one order-of-magnitude from its initial value during 128 h of injection, forcing termination of the experiment. The hydraulic gradient within a 20 - 35 cm distance of column C was nearly constant during the injection experiment. The slight decrease of the hydraulic gradient in phase V was due to a decrease in the injectant flow rate (Table 5.3). The size distribution analysis of suspended microbeads in the pore AGW samples collected from column C, indicated small variations in the mean diameter (dso) of microbeads independent of travel distance and injection time (Table 5.4). However, the mode diameter (dmode) of such microbeads, which represents the most frequent particle size, was lower than that of the injectant, indicating a variation in size distribution of interstitial suspended microbeads compared to the injectant. This difference can be attributed to the filtration of larger microbeads through the sand matrix. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 101 for Encapsulated Cel! Bioaugmentation

0.2 0.18 (1) (II) 0.16 A -E 0.14 u o 0.12 I '" 0.1 E ~ 0.08 ~ ~ 0.06 <] 0.04 ~~~...... ~M-W""~""rr­ 0.02

O'--'--'--'--'--L..o---...... '---J----l----l----l----l----l--..l--..l--..l----' o 12 24 36 48 60 72 0.2 ,.....-.---.---.---.---.---...... ,...... ,...... ,...... ,--,--,--,--,--,--,--,--,:0 0.18 (1) • (II) 0.16 8 • "7E 0.14 u ° o 0.12 I '" 0.1 E ~ 0.08 ~ ~ 0.06 <] F--.jJliI­ ° 0.04 •• 0.02

OL..o---L-L-L-...... '---J'---J--..l--..l--..l--..l--..l--..l--..l--..l--..l--..l--l o 12 24 36 48 60 72 4r-r---r-r--r-r--1---r-r--r--r--t----r-""T"-,-.--r--t----r----r--r-...--,--, 3.5

- 3 E u o 2.5 • I '" 2 E ~ 1.5 ~ ~ <]

12 24 36 48 60 72 84 96 108 120 132 Time (h)

Figure 5.3. Variation of hydraulic gradient (~h/~l) with time for columns A, B, and C. Curves represent the travel distance within the various sections: 5 - 27.5 cm (e) and 27.5 ­ 50 cm (0) for column A, 5 - 20 cm (e) and 20 - 35 cm (0) for columns Band C. 5. Transport ofGelian Gum Microbeads Through Sand: An Experimental Evaluation 102 for Encapsulated Celi Bioaugmentation

80

60 ~ ~ l:C ~ 40

20

72

80

60

l:C ~ 40 2:~~::::::J

o 12 24 36 48 60 72 100 ,---,.-,--,-.,.-,.---r----r-..,.-.--.-,--,-.,.-,,...,..,...---r-""-'--'-'---=""1

80 C

60 ~ ~ l:C ~ 40

20 o~~~~~~~~EJ o 12 24 36 48 60 72 84 96 108 120 132 Time (h)

Figure 5.4. Variation of total breakthrough (TB) of gellan gum microbeads with time for columns A, B, and C during various injection phases. Curves are for 5 cm (e), 20 cm (_), 50 cm (....), 80 cm (....), and 115 cm (0) downgradient from the bed inlet. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 103 for Encapsulated Cel! Bioaugmentation

Table 5.4. Characteristic size parameters of suspended microbeads across column C.

Diameter (Ilm) a

Time (h) Distance (cm) ~ode dso drnax 50 20 12.8 16.2 31.4 50 14.5 20.1 36.9 80 n.d. b n.d. n.d. 110 n.d. n.d. n.d.

60 20 13.7 18.7 39.7 50 12.2 18.0 37.7 80 12.2 20.1 37.1 110 13.2 20.0 35.4

70 20 12.8 17.8 32.6 50 14.1 19.8 39.7 80 12.8 17.8 32.0 110 13.7 21.2 34.7

80 20 12.8 17.2 35.0 50 15.9 20.2 39.7 80 16.8 19.8 39.2 110 12.2 18.3 31.4

a drnode' dso, and drnax represent mode, mean, and maximum diameters, respectively. b not determined.

The variation of total breakthrough (TB) of microbeads with travel distance and injection time is shown in Figure 5.4. The TB of microbeads was generally an ascending function of injection time and descending function of travel distance. However, sorne discrepancies were observed in the variation of TB with travel distance. For instance, the TB observed at 80 and 115-cm distances of column C were slightly higher than that at 50 cm (phases II-V). This effect was also observed with column B (phase 1), and can be due to selective movement of microbeads through the porous media. Low permeability zones of the sand matrix surrounding the inserted sampling needle may cause selective movement of microbeads through the more permeable zones farther from the sampling point, which 5. Transport ofGel/an Gum Microbeads Through Sand: An Experimental Evaluation 104 for Encapsulated Cel/ Bioaugmentation consequently limits the observation of migrating microbeads. The less permeable zones could be created during packing the columns or caused by deposited microbeads through blockage of the interstitial pores. Accordingly, zero or negative values of retention percentage (RP) were obtained for microbeads travelling through such distances, as illustrated in Figure 5.5. The RP profiles demonstrate the distribution of the microbeads through various sections of each column. The highest and lowest RP values obtained after 72 h injection for both columns A and B were within 20 - 50 and 80 - 110 cm distances, respectively. In column B, the RP of microbeads within a 80 - 110 cm distance increased during the first 24 h of injection but eventually reached below zero values, indicating the complete passage of microbeads through this section of the column. The RP data for column C indicated that (after 128 h of injection) more than 85% of microbeads were captured within a 5 - 20 cm distance, as compared to 5 and 10% retained within 20 - 50 and 80 - 110 cm distances, respectively. The accumulation of microbeads within a 5 - 20 cm distance of column C was intensified at the beginning of phase II when the injectant concentration was doubled. The RP profiles in Figure 5.5 demonstrate that microbeads were more widely spread through the llO-cm length of column A than columns Band C. After 72 h of injection, about 75% of microbeads injected within this period, were distributed across a 5 - 110 cm distance in column A, compared to 78% across a 5 - 50 cm in column B, and 76% across a 5 - 20 cm in column C. The most non-uniform distribution of microbeads was obtained in column C, where the major portion of microbeads was immobilized within 5 ­ 20 cm downgradient from the bed inlet. These results imply that the degree of distribution of microbeads through the sand matrices was strongly influenced by the grain size range. The wider dispersion of microbeads across the length of column A compared to those observed in columns Band C, suggests a higher potential for the formation of a uniform bioactive zone of encapsulated cells across a sandy aquifer with such grain size range (0.5 ­ 2 mm) under the employed operating conditions. Increasing the injection time and use of a low injectant concentration throughout the injection process may alleviate the filtration effect of the sand matrices, enhancing the distribution of microbeads through greater distances of sand media. 5. Transport ofGelian Gum Microbeads Through Sand: An Experimental Evaluation 105 for Encapsulated Celi Bioaugmentation

(1) (II) 40 A

30

a..cr: 20

10 o

-1 0 L----l----l_....I------l-_"'-----I-_'---L----l._...l-----l-----I o 12 24 36 48 60 72 50 ...... --.----,-...,---,--,----,--,..--.----,.---r---,----..,

40

30

a..cr: 20

10 o

-1 0 L----l----l._....I------l-_"'-----I---J'---'----'-_...l----l-----l o 12 24 36 48 60 72 90 ,.-,---,--r-,...... ,--,--.-.-,.--,--,--r-,...... ,...,....,..-.-.-,.--,...."....~ 80 70 60 50 §: 40 30 20

10 ~~=~::==:::= o '-HIll- .---.~~~ ~---,-A---.k-7jt--*i -1 0 ...... o 12 24 36 48 60 72 84 96 108 120 132

Time (h)

Figure 5.5. Variation of retention percentage (RP) of gellan gum microbeads with time for columns A, B, and C during various injection phases. Curves represent the travel distance within the various sections: 5-20 cm (e), 20-50 cm (_), 50-S0 cm (....), and SO-11O cm (....). 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 106 for Encapsulated Cell Bioaugmentation

The filtration of microbeads through porous media is governed by gravitational sedimentation, interception, and diffusion mechanisms. Among these, the latter is more important in transport of micrometer-sized particles « 5 /lm dia.) and has less influence on the removal of microbeads used in this study. This fact can be clearly demonstrated by plotting the variation of collector efficiency ('If) against the suspended particle and collector (grain) dimensions. The collector efficiency is calculated based on a semi­ empirical model (Logan, 1999; Rajagopalan and Tien, 1976) as follows, 2 8 2 11+ =4.04y2bHI/3pe-2/3 +y bHLoI/ R*15/8 +0.00338y bHS*1.2R*-0.4 (5.2) where y = (1_8)1/3, bH = 2(1-y)/(2-3y+3y-21), Pe = Peclet No., Lo = London No., R* = ~/de' S* = settling velocity/Darcy velocity, ~ = particle diameter, de = collector diameter, and 8 =effective porosity.

The variations of 11 + as a function of grain and microbead sizes are depicted in Figure 5.6. The Rajagopolan-Tien (RT) model describes the overall collision rate (collector efficiency) of a suspended particle by superposition of removal terms due to diffusion (110+)' interception (11/), and gravitational sedimentation (11s+) mechanisms, assuming that collisions are additive by each mechanism. It can be observed that 11+ decreases with the decrease of particle diameters larger than 5 /lm for any given grain size. However, for the grain size range of 0.375 - 2 mm the value of 11+ is greater for l-/lm particles compared to 5-/lm particles, suggesting that somewhat different mechanisms govern the transport of micrometer-sized particles. The profiles for 1- and 5-/lm particles are presented in Figure 5.6 for comparison purposes only to illustrate the differences in filtration mechanisms of particles larger and smaller than 5 /lm. As illustrated in Figure 5.7, the variations of fractional collector efficiencies corresponding to diffusion, interception, and gravitational sedimentation weIl demonstrate that diffusion is the main mechanism controlling the removal of particles smaller than 5 /lm. Interception and gravitational sedimentation mechanisms contribute to the removal of larger particles while the importance of the former varies inversely with the latter with the increase of grain size. The major portion of microbeads (50 - 90%) produced in this study, laid within a range of 10 - 20 /lm (Table 5.3). It is evident from Figure 5.6 that the value of 11+ for such particles is of the same order-of-magnitude as 1 - 5 /lm particles, at least for the 5. Transport ofGel/an Gum Microbeads Through Sand: An Experimental Evaluation 107 for Encapsulated Cel/ Bioaugmentation most of the grain size range examined. On the other hand, a small microbead is limited in its capacity for encapsulation of a large number of bacteria, which is required to biodegrade high concentrations of contaminants. The small carriers are also more subject to sorption to solid collectors due to increased London-van der Waals attractive forces (Logan, 1999). Therefore the microbeads size range used in the present study (10 - 40 /...lm dia.) constitutes a fair trade-off between the extent of filtration and the potential for dense cell encapsulation and conveying.

+~ >- g 10-1 Q) '(3 ::E w lo- -§ 10-2 Q) o Ü

10-3 L--I-J...... l...--I-.J...... J.--l.-~~b::==:=..e...... l-..l.--J.....J o 0.5 1 1.5 2 2.5 Grain Size (mm)

Figure 5.6. Variation of collector efficiency (11+) with grain size, estimated based on RT

1 3 model using Darcy velocity of 0.23 m h- , particle density of 1,008 kg m- , Hammaker constant of 10-20 J, and fluid properties of water at 10 oc. Curves represent the particle diameter (in /...lm): 1 (_),5 (e), 10 (V), 20 (d), 30 (0), and 40 (0). 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 108 for Encapsulated Cel! Bioaugmentation

• Diffusion • Interception D Sedimentationl

Particle Diameter (IJm) 1 5 10 20 30 40 100

80

60

40

20

o

Grain Diameter (mm)

Figure 5.7. Variation of fractional collector efficiency (11t/11+) for diffusion (11D+/11+), interception (11t/11+), and sedimentation (11s+/11+) mechanisms with particle diameter and grain size.

5.5. Conclusions The present study was performed to investigate the feasibility of creating a bioactive zone using cell-encapsulating microbeads through sand matrices. Adequate transport and uniform distribution of cell carriers through the contaminated region of an aquifer are important parameters for the successful application of the encapsulated bacteria to in situ bioaugmentation. The encapsulated cells should provide a bioactive zone across a given contaminated region of the aquifer porous media. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 109 for Encapsulated Cell Bioaugmentation

A two-phase dispersion technique suitable for large-scale encapsulation of microbial cells was used to produce gellan gum microbeads ranging from lOto 40 /lm (dia.). Microbeads were injected intermittently into horizontal sand columns packed with various size distributions of sand, which may collectively resemble the overlaying formations of the saturated zone of an aquifer. The results suggest that under the employed injection conditions, gellan gum microbeads can be distributed through certain zones of a soil matrix, essentially composed of fine, medium, coarse and very coarse fractions of sand. The breakthrough of microbeads increased with injection time but decreased with travel distance. The deposition of microbeads within short distan~es from the porous media inlet was accompanied by an increase in the hydraulic gradient across these sections. The widest distribution of microbeads was achieved in a column packed with coarse and very coarse sand fractions ranging from 0.5 to 2 mm. The highest filtration effect was observed in the column packed with fine to very coarse sand fractions ranging from 0.125 to 2 mm. Microbeads were recovered at the outlet of aIl columns suggesting that they can be transported across greater distances, subject to the adoption of appropriate injection conditions. Prolonged injection of a low concentration suspension of microbeads (e.g. 40 mg beads L'I) would provide a better opportunity to extend the longitudinal distribution of microbeads for in situ bioaugmentation applications.

Acknowledgements. The authors thank the Natural Sciences and Engineering Research Council of Canada and the National Research Council of Canada (NRC paper no. 00000) for financial support, and McGill University for the awarding of the Max Stem Recruitment Fellowship to P.M. 5. Transport ofGel/an Gum Microbeads Through Sand: An Experimental Evaluation 110 for Encapsulated Cel/ Bioaugmentation

References Ashtaputre, A. A. and Shah, A. K. (1995). Studies on viscous, ge1-forming exopolysaccharide from Sphingomonas paucimobilis GS 1. Applied and Environmental Microbiology 61: 1159-1162. Bai, G., Brusseau, M. L. and Miller, R. M. (1997). Influence of a rhamnolipid biosurfactant on the transport of bacteria through a sandy soil. Applied and Environmental Microbiology 63: 1866-1873. BaIes, R: C., Li, S., Yeh, T.-C. J., Lenczewski, M. E. and Gerba, C. P. (1997). Bacteriophage and microsphere transport in saturated porous media: Forced­ gradient experiment at Bordon, Ontario. Water Resources Research 33: 639-648. Buitelaar, R. M., Hulst, A. C. and Tramper, J. (1988). Immobilization of biocatalysts in thermogels using the resonance nozzle for rapid drop formation and an organic solvent for gelling. Biotechnology Techniques 2: 109-114. Camelin, 1., Lacroix, C., Paquin, c., Prevost, H., Cachon, R. and Divies, C. (1993). Effect of chelants on gellan gum rheological properties and setting temperature for immobilization of living Bifidobacteria. Biotechnology Progress 9: 291-297. Devlin, J. F. (1994). A simple and powerful method of parameter estimation using Simplex Optimization. Ground Water 32: 323-327. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A. and Smith, F. (1956). Colorimetric methods for determination of sugars and related substances. Analytical Chemistry 28: 350-356. Freeze, R. A. and Cherry, J. A. (1979). Groundwater. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Giavasis, 1., Harvey, L. M. and McNeil, B. (2000). Gellan gum. Critical Reviews in Biotechnology 20: 177-211. Harmand, B., Rodier, E., Sardin, M. and Dodds, J. (1996). Transport and capture of submicron particles in natural sand: Short column experiments and a linear mode!. Colloids and Surfaces. A:Physicochemical and Engineering Aspects 107: 233-244. Harvey, R. W., George, L. H., Smith, R. L. and LeBlanc, D. R. (1989). Transport of microspheres and indigenous bacteria through a sandy aquifer: Results of natural- 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 111 for Encapsulated Cell Bioaugmentation

and forced-gradient tracer experiments. Environmental Science and Technology 23: 51-56. Harvey, R W., Kinner, N. E., MacDonald, D., Metge, D. W. and Bunn, A (1993). Role of physical heterogeneity in the interpretation of small-scale laboratory and field observations of bacteria, microbial-sized microsphere, and bromide transport through aquifer sediments. Water Resources Research 29: 2713-2721. Li, Q. and Logan, B. E. (1999). Enhancing bacterial transport for bioaugmentation of aquifers using low ionic strength solutions and surfactants. Water Research 33: 1090-1100. Logan, B. E. (1999). Particle Transport in Porous Media. In: Environmental Transport Processes (Ed.), John Wiley & Sons, Inc., New York, 564-613. McCaulou, D. R, BaIes, R C. and Arnold, R G. (1995). Effect of temperature-controlled motility on transport of bacteria and microspheres through saturated sediment. Water Resources Research 31: 271-280. Moorehouse, R, Colegrove, G. T., Stanford, P. A, Baird, J. K. and Kang, K. S. (1981). PS-60: A new gel-forming polysaccharide. In: Solution Properties of Polysaccharides. Brant, D. A (Ed.), American Chemical Society, Washington DC, 111-124. Morris, D. A. and Johnson, A I. (1967). Summary of Hydrologie and Physical Properties of Rock and Soil Materials, as Analyzed by the Hydrologie Laboratory of the V.S. Geological Survey 1948-1960, V.S. Geologieal Survey Water Supply Paper 1839­ D. Nilsson, K., Birnmaum, S., Flygare, S., Linse, L., Schroder, V., Jeppsson, V., Larsson, P.­ O., Mosbach, K. and Brodelius, P. (1983). A general method for the immobilization of cells with preserved viability. European Journal of Applied Microbiology and Biotechnology 17: 319-326. Norton, S. and Lacroix, C. (1990). Gellan gum gel as entrapment matrix for high temperature fermentation processes-Rheological study. Biotechnology Techniques 4: 351-356. 5. Transport ofGellan Gum Microbeads Through Sand: An Experimental Evaluation 112 for Encapsulated Cell Bioaugmentation

Ogata, A. (1970). Theory of dispersion in a granular medium. Fluid Movement in Earth Materials-U.S. Geological Survey Professional Paper 411-1. Washington, United States Government Printing Office. Petrich, C. R, Stromo, K. E., Ralston, D. R and Crawford, R L. (1998). Encapsulated cell bioremediation: Evaluation on the basis of particle tracer tests. Ground Water 36: 771-778. Rajagopalan, R and Tien, C. (1976). Trajectory analysis of deep-bed filtration with the sphere-in-cell porous media mode!. AIChE Journal 22: 523-533. Reddi, L. N. and Bonala, M. V. S. (1997). Analytical solution for fine particle accumulation in soil filters. Journal of Geotechnical and Geoenvironmental Engineering 123: 1143-1152.

Sanderson, G. R, Bell, V. L. and Ortega, D. (1989). A comparison of gellan gum, agar, K­ carrageenan, and algin. Cereal Foods World 34: 991-998. Sanderson, G. R and Clark, R. C. (1983). Gellan gum. Food Technology 37: 62-70. Silliman, S. E. (1995). Particle transport through two-dimensional, saturated porous media: Influence of physical structure of the medium. Journal of Hydrology 167: 79-98. Vandevivere, P. and Kirchman, D. L. (1993). Attachment simulates exopolysaccharide synthesis by a bacterium. Applied and Environmental Microbiology 59: 3280-3286. Vogel, T. M. (1996). Bioaugmentation as a soil bioremediation approach. Current Opinion in Biotechnology 7: 311-316. 113

6. Biodegradation of Gasoline by Gellan Gum­ Encapsulated Bacterial Cells

Peyman Moslemy a,b ,Ronald J. Neufeld c, and Serge R. Guiot b

a Department of Chemical Engineering, McGill University, 3610 University Street, Montreal, Quebec, Canada H3A 2B2

b Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

C Department of Chemical Engineering, Queen's University, Dupuis Hall, Kingston, Ontario, Canada K7L 3N6

In this study, the capacity of gellan gum-encapsulated bacteria to degrade gasoline was evaluated in comparison with free (non-encapsulated) cells. The biodegradation of gasoline was explored in both liquid suspension and saturated soil microcosms, and the effect of initial gasoline and cell concentration on the rate and the extent of biodegradation was investigated. The candidate has carried out all the work in this chapter with the exception of Headspace-GC analysis of hydrocarbon containing liquid samples, which was conducted by Mr. Stephane Deschamps. This article has been submitted to the journal of Biotechnology and Bioengineering for publication. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 114

6.1. Abstract Encapsulated ceIl bioaugmentation is a novel alternative solution to in situ bioremediation of contaminated aquifers. This study was conducted to evaluate the feasibility of such a remediation strategy based on the performance of encapsulated ceIls in the biodegradation of gasoline, a major groundwater contaminant. An enriched bacterial consortium, isolated from a gasoline-poIluted site was encapsulated in geIlan gum microbeads (16 - 53 /lm dia.). The capacity ofthe encapsulated ceIls to degrade gasoline under aerobic conditions was evaluated in comparison with free (non-encapsulated) ceIls. Encapsulated ceIls (2.6 mgcells g-I bead) degraded over 90% gasoline hydrocarbons (initial concentration 50 - 600 mg L-1) within 5 ­ 10 days at 10°C. Equivalent levels of free ceIls removed comparable amounts of gasoline 1 (initial concentration 50 - 400 mg L- ) within the same period, but required up to 30 days to 1 degrade the highest level of gasoline tested (600 mg L- ). Free ceIls exhibited a lag phase in biodegradation, which increased from 1 to 5 days with an increase in gasoline concentration 1 (200 - 600 mg L- ). Encapsulation provided ceIls with a protective barrier against toxic hydrocarbons, eliminating the adaptation period required by free ceIls. The reduction of encapsulated ceIl mass loading from 2.6 to 1.0 mgcel1s g-I bead caused a substantial decrease in the extent of biodegradation within a 30-day incubation period. Encapsulated ceIls dispersed within the porous soil matrix of saturated soil microcosms, demonstrated a reduced 1 performance in the removal of gasoline (initial concentrations of 400 and 600 mg L- ), removing 30 - 50% gasoline hydrocarbons, compared to 40 - 60% by free ceIls, within 21 days of incubation. The results of this study suggest that geIlan gum-encapsulated bacterial ceIls have the potential to be used for biodegradation of gasoline hydrocarbons in aqueous systems.

Keywords: Biodegradation; Gasoline; Bacterial consortium; Encapsulation; GeIlan gum; Mierobead.

6.2. Introduction Biological degradation of hazardous waste compounds by augmented encapsulated rnicroorganisms is a new line of approach to in situ biorernediation of sorne contaminated aquifers. Microbial encapsulation is one of the emerging avenues of immobilized ceIl technology, which involves entrapment of viable active microorganisms within an insoluble polymerie support (carrier). The applications of encapsulated microbial ceIls for environmental bioprocesses have been discussed extensively in previous reports (Cassidy et al., 1996; McLoughlin, 1994; Trevors et al., 1992). The use of encapsulated ceIls may offer 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 115

several advantages over free ceIl suspensions in bioremediation schemes. Encapsulation may potentially reduce biotic and abiotic stresses, providing a number of advantages including protection of ceIls from toxic effects of hazardous compounds (Bettmann and Rehm, 1984; Manohar and Karegoudar, 1998; Paje et al., 1998; Somerville et al., 1977), and increasing their survival and metabolic activity (Hall et al., 1998; Suzuki et al., 1998; Trevors et al., 1993; Weir et al., 1996). The present study was performed to address the biological activity of an encapsulated bacterial consortium in the removal of gasoline from aqueous systems. Gasoline is a complex mixture of C4-Cll aliphatic and aromatic hydrocarbons that are sufficiently soluble in water to pose a major threat to groundwater quality. The release of gasoline from leaking underground storage tanks, accidentaI spills, distribution systems, and other industrial operations are common sources of contamination to soil and groundwater (Atlas and Cerniglia, 1995; Cherry, 1987; Day et al., 2001). Although numerous researchers have explored the biodegradation of gasoline by free ceIls in both laboratory (Solano-Serena et al., 1998; Solano-Serena et al., 2000a; Solano-Serena et al., 2000b; Solano-Serena et al., 20ü0c; Solano-Serena et al., 1999; Yerushalmi and Guiot, 1998; Yerushalmi et al., 1999; Zhang and Bouwer, 1997; Zhou and Crawford, 1995) and field (Gruiz and Kriston, 1995; Maura Jutras et al., 1997; Morgan et al., 1993), little effort has been made to use encapsulated cells. The use of encapsulated cells has been investigated for the biodegradation of sorne aliphatic and aromatic hydrocarbons. Prototheca zopfi was encapsulated in 3-mm alginate beads to degrade n-alkanes, C14-C16 (Suzuki et al., 1998). Pseudomonas putida encapsulated in 0.29 x 0.50 cm polyacrylamide gel strands (Somerville et al., 1977), and Rhodococcus sp. cells encapsulated in alginate beads (Paje et al., 1998), were applied to biodegradation of benzene. Naphthalene was degraded by Pseudomonas sp. cells encapsulated in agar and 2­ mm alginate beads, and in 1-mm polyacrylamide granules (Manohar and Karegoudar, 1998). Pseudomonas sp. cells encapsulated in 2 - 3 mm alginate beads, and polyurethane pellets were used for biodegradation of phenanthrene (Weir et al., 1995). The application of encapsulated cells to in situ bioaugmentation of a contaminated aquifer necessitates the formulation of polymeric micro-particles of 10 - 50 ~m diameter (Moslemy P, Neufeld RJ, Millette D, Guiot SR. 2001. Transport of Gellan Gum Microbeads through Sand: An Experimental Evaluation for Encapsulated Cell Bioaugmentation. submitted to the journal of Advances in Environmental Research.), carrying viable cells with high levels of biodegradation activity. Upon hydraulic injection of a suspension of cell carriers into the saturated zone of the aquifer, these carriers should be directed through the soil porous matrix to be emplaced ahead of or across the contaminant plume. The porous structure of the carrier 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 116

would allow the diffusion of dissolved contaminants into the polymer matrix where they would be consumed by the entrapped cells, leading to the decontamination of groundwater. The main objective ofthis study was to evaluate the potential of an enriched bacterial consortium encapsulated in gellan gum microbeads to degrade gasoline. Such information is needed to define appropriate implementation strategies for in situ bioaugmentation schemes. The biodegradation of gasoline was investigated under aerobic conditions in both liquid suspension and saturated soil microcosms. The performance of the encapsulated cells at various concentrations of gasoline was evaluated based on the extent and the rate of biodegradation, and compared to that of free, non-encapsulated cells. The influence of encapsulated cell mass loading on biodegradation was also evaluated.

6.3. Materials and Methods 6.3.1. Materials Gellan gum (Kelcogel®) was donated by the CP Kelco US, mc., formerly NutraSweet Kelco Co. (San Diego, Califomia, USA). Gasoline was purchased from a commercial fuel retailer. The physical properties and chemical composition are summarized in Table 6.1. The present study was carried out using samples from a single batch of gasoline. Canola oil was purchased from a local food distributor. AlI other chemicals used in this study were of reagent grade.

6.3.2. Microbial Culture and Growth Medium A bacterial consortium isolated from a gasoline-contaminated soil of an industrial site (Montreal, Quebec, Canada), was used in this study. The bacteria were preserved at -80 oC in lO-mL ampules containing 5 %(v/v) glycerin as a cryoprotectant. After thawing, the culture was suspended in an enrichment mineral salts medium (MSM), supplied with gasoline (160 1 mg L- ) as the sole source of carbon. The enri~hment culture was a consortium of spherical cocci and bacilli as revealed by microscopie observation. The enrichment underwent twelve successive transfers before the encapsulation and biodegradation experiments. The culture was harvested by centrifugation at 12,000 x g for 15 min at 4 oC and re-suspended in 2 mL of a sterile 0.1 % (w/v) calcium chloride solution in preparation for encapsulation. 1 The MSM solution contained (g L- ): KH2P04 0.87; K2HP04 2.26; (NH4)2S04 1.1; 1 and MgS040.047; amended with a trace metal solution (1 mL per liter) composed of (g L- ): CO(N03)2.6H20 0.291; AIK(S04)2.12H20 0.474; CuS04 0.16; ZnS04.7H20 0.288; FeS04.7H20 2.78; MnS04.H20 1.69; Na2Mo04.2H20 0.482; and Ca(N03)2.4H20 2.36. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 117

The mineraI solution was prepared initially without MgS04, and sterilized by autoclaving at

121°C for 20 min. The sterile mineraI solution was then combined with a sterile MgS04 solution under aseptic conditions to give the MSM solution with final desired concentrations. The pH of the medium was 7.0 ± 0.1.

Table 6.1. Characteristics of a commercial gasoline.

Chemical Composition Representative Concentrations (% w/w) Hydrocarbon Group:

n-Alkanes (C4 - Cu) 10 - 30

Branched Alkanes (C4 - C9) 18 - 60

Cycloalkanes (C6 - C9) 3 - 14 - ) Alkenes (C4 C6 5 - 14

Branched Alkenes (Cs & C6) <1

Monoaromatics (BTEX, C3- & C4-benzenes) 18 - 40 Polyaromatics <2.5

Other Ingredients: Lead, Pb Max. 5 mglL Manganese,Mn Max. 18 mglL Sulfur, S Max. 0.10 % mass RSH Max. 0.0030 % mass Anti-Oxidant Min. 5.7 g/mJ Corrosion Inhibitor Min. 6.0 g/mJ MTBE Max. 2.7 % mass Particulate Matter Max. 2.2 mglL

Physical and Chemical Properties Boiling Point 35 - 220 oC Density 0.798 g/mL @ 23 oC 0.811 g/mL @ 10 oC Vapor Density (Air =1) 3.5 Chemical Oxygen Demand, COD a (g 02 g-l) 3.42 Octane No. 87 a Estimated based on the average concentration of main components of a typical gasoline product (Fan and Krishnamurthy, 1995). 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 118

6.3.3. Encapsulation Gellan gum-encapsulated cell microbeads were produced by emulsification-intemal gelation. A 0.75% (w/v) dispersion of gellan gum in sterile de-ionized water was prepared and heated to 75 oC to dissolve and form the pregel solution (sol). Calcium chloride was added at 0.06% (w/v) and the sol was left at room temperature to cool to 45 oc. The pH was adjusted to 6.9 ­ 7.2 with 0.1 N NaOH. A 2-mL suspension of cells in sterile 0.1 % (w/v) calcium chloride solution was mixed with the sol to a final concentration of 0.2, 0.8, 1, 2, and 8 gcells L- 1 sol, and the mixture was then emulsified in 330 mL sterile canola oil aided by a 0.1 %(w/w) non­ ionic oil soluble surfactant, Span 80 (sorbitan monooleate), at 45 oC. The disperse phase volume fraction was 0.143 in all preparations. The emulsion was formed in a l-L round­ bottomed cylindrical glass reaction vessel equipped with four baffle blades. The emulsion was vigorously stirred at 4500 rpm, using a quarter-circular paddle impeller assembled with a T­ Line laboratory stirrer (model 102, Talboys Engineering Corp., Montrose, PA, USA). The rotational speed of the impeller was monitored regularly by means of a Fisherbrand tachometer (Fisher Scientific, Nepean, Ontario, Canada), and rectified if necessary. After 10 min, gelation was initiated by cooling the reaction vessel to 15 oC using an ice bath, leading to entrapment of cells in gellan gum microbeads. The oil-microbead dispersion was transferred with gentle mixing into 500 mL of a sterile 0.1 % (w/v) calcium chloride solution. The oil was removed by aspiration after partitioning of microbeads into the aqueous phase, and the microbeads were rinsed repeatedly with a sterile 0.1 % (v/v) Tween 80 (polyoxyethylene (20) sorbitan monooleate) solution. The encapsulated cell microbeads were incubated in a sterile 0.1 % (w/v) calcium chloride solution at 4 oC before the biodegradation experiments. Size distribution of microbeads was measured by means of a Malvem particle and size analyzer (series 2600, Malvem Instruments, Inc., Southborough, MA, USA). Microbead sizes ranged from 16 - 53 f.lm with mean diameters of 23 - 30 f.lm. The density (at 25 OC) and dry percentage of wet encapsulated cell microbeads were 1.0088 ± 0.0028 g mL- 1 and 1.08 ± 0.16%, respectively, as determined by drying nine pre­ weighed samples of wet microbeads at 105 oC, followed by measuring the weight loss, and assuming that the volume of the sample is equal to that of the evaporated water. The microbead samples were separated from the aqueous dispersion by centrifugation at 12,400 x g for 10 min at 4 oC. The density (at 25 OC) and dry percentage of wet cells (five samples) were similarly determined to be 1.1397 ± 0.0411 g mL- 1 and 12.43 ± 3.59%. Based on these values, the encapsulated cell mass loading for microbeads prepared at 0.2, 0.8, 1, 2, and 8 gcells 1 L- sol, was estimated to be 0.26, 1.0, 1.3,2.6, and 10.4 mgcells g'l bead, respectively. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial CeUs 119

6.3.4. Biodegradation in Liquid Suspension Microcosms Biodegradation activity of free and encapsulated cells in the removal of gasoline hydrocarbons was investigated in 160-mL serum bottles (in duplicate) on a rotary shaker (100 rpm, 10 OC). Each serum bottle contained 20 mL of sterile MSM, and was inoculated with either free or encapsulated cells (2.6 mgcells g-I bead) to give a final concentration of 0.26 gcells L-I MSM. The bottles were capped with standard Mininert gas-tight valves (Supe1co, Sigma-Aldrich Canada Ltd., Oakville, ON, Canada). Gasoline was injected directly into the liquid phase in each bottle using a micro-syringe to give total concentrations of 50, 100, 200, I 400, and 600 mg L- . Suspension of encapsulated cells (2.6 mgcells g-I bead), deactivated by autoclaving at 121°C for 20 min, as well as cell-free sterile aqueous medium were used in control serum bottles under similar conditions to evaluate the removal of gasoline by abiotic processes. AH biodegradation experiments were carried out in the dark to prevent any photocatalytic reactions. The influence of encapsulatedcell mass loading on biodegradation was evaluated by conducting experiments with microbeads of various cell loading (0.26, 1.0, 2.6, and 10.4 mgcells g-I bead) at an initial gasoline concentration of 200, 400, and 600 mg LI. Gas samples (50 ilL) were collected with a gas-tight micro-syringe (Hamilton, no. 1705) equipped with a gas-tight syringe valve (Hamilton, GTS valve) 4 - 6 h after injection of gasoline into shaking bottles, and analyzed by gas chromatography (GC). The variation of the total peak area (tpa) of total petroleum hydrocarbons (TPH) deterrnined by GC analysis, was monitored during the incubation period (30 days). The gasoline removal efficiency was then estimated using the time-based tpa values obtained for the gas samples as, (tpa) - (tpa) TPH Removal (%) = t X 100 (6.1) t)tpa to where (tpa)to and (tpa)t correspond to the gas samples collected first (t =to) and at a given time (t = t) after gasoline injection, respectively. The gas chromatographic analysis of both gaseous and aqueous phase concentrations in control serum bottles indicated that more than 90% of the injected gasoline hydrocarbons were present in the headspace of free cell bottles while over 79% distributed into the headspace ofencapsulated cell bottles (Table 6.2). The oxygen level in the gas phase at the end of biodegradation experiments was deterrnined by the GC analysis. Gas samples (50 J1L) were collected with a gas-tight micro­ syringe (Hamilton, no. 1705) equipped with a GTS valve. Appropriate standard curves, correlating the total initial concentration of gasoline with initial concentrations of gasoline hydrocarbons in the gas, liquid, and solid (gel microbeads) phases, were constructed. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial eeus 120

Table 6.2. Initial phase distribution of gasoline hydrocarbons in liquid and saturated soil microcosms

TPH a Distribution in Liquid Medium (%)

Free Cells Encapsulated Cells b Gasoline (mg L- I) Gasoline (mg L- I) Phase 50 100 200 400 600 50 100 200 400 600 Vapor 92.3 91.4 90.8 90.1 90.0 79.3 84.3 85.4 88.7 89.4 Liquid 7.7 8.6 9.2 9.9 10.0 11.4 8.5 7.7 6.1 5.6 Gel 9.3 7.2 6.9 5.2 5.0

TPH a Distribution in Saturated Soil (%)

Free Cells Encapsulated Cells b Gasoline (mg LI) Gasoline (mg LI) Phase 400 600 400 600 Vapor 75.8 69.2 75.2 70.4 Liquid 1.6 2.7 1.5 2.1 Gel 2.6 3.6 Soil 22.6 28.1 20.7 23.9 a TPH: Total Petroleum Hydrocarbons b Cell mass loading 2.6 mgcells g-I bead

6.3.5. Biodegradation in Soil Microcosms Non-polluted soil from near-surface sediments was collected locally. Soil grain sizes ranged from 0.125 to 2 mm. The soil was dried in an oven at 105 oC for 24 h. The dry soil density was 1.898 ± 0.031 g mL-l, as measured in triplicate by a differentiai-volume method. Soil microcosms were prepared with 25 g of dry soil in 160-mL serum bottles (in duplicate), sterilized by autoclaving at 121°C for 20 min on three consecutive days. Each serum bottle was inoculated with either free or encapsulated cells (2.6 mgcells g-I bead) under aseptic conditions. The soil was gently mixed to disperse the inocula throughout the soil matrix. The serum bottle was filled with 10 mL of sterile MSM, and capped with a Mininert valve. The final concentration ofbiomass in each bottle was 0.52 gcells L- I MSM. Gasoline was injected directly into the liquid phase in each bottle to obtain a totaI concentration of 400 and 600 mg 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 121

LoI. Sterile non-inoculated soil media were used as control. The serum bottles were incubated in the dark at 100 rpm and 10 oc. The biological activity ofencapsulated celI microbeads with a celI mass loading of 2.6 mgcells gol bead was compared to those with 0.26 and 1.3 mgcells gol bead at an initial gasoline concentration of 400 and 600 mg LoI in soil microcosms under similar experimental conditions. Gas samples (50 !-JL) were analyzed by GC, and the variations of total peak area of TPH were monitored during the incubation period (30 days). The gasoline removal efficiency was estimated in a similar fashion to liquid suspension microcosms (Eq. 6.1). At the end of each experiment, a liquid sample was colIected from the serum bottle, and analyzed with a headspace gas chromatography technique (Headspace-GC). The oxygen level in the gas phase at the end of biodegradation experiments was determined by the GC analysis. Gas samples (250 ilL) were colIected with a gas-tight micro­ syringe (Hamilton, no. 1725) equipped with a GTS valve. Appropriate standard curves were constructed to estimate the initial concentrations of gasoline hydrocarbons in the gas, liquid, and solid (gel microbeads and soil) phases, with respect to the total initial concentration of gasoline.

6.3.6. Analytical Techniques Gasoline in headspace samples was analyzed using a HP gas chromatograph system (model HP 6890, Hewlett-Packard Co., Wilmington, Delaware, USA) equipped with a flame ionization detector (FID), and a packed column (Supelco 1-2485, 1% SP-1000, 3 mm x 2 rn, 60/80 mesh Carbopack B). Both the injector and detector temperatures were set at 250 oC, while the column temperature was kept at 225 oc. Helium was used as the carrier gas at a flow rate of 50 mL/min. Concentration of gasoline in liquid samples was deterrnined by the Headspace-GC method on a Varian gas chromatograph (model Chrompack CP-3800, Varian Associates, mc., Walnut Creek, Califomia, USA) equipped with a Tekmar headspace autosampler (series 7050 Carrousel, Rosemount Analytical, Inc., Tekmar Co., Cincinnati, Ohio, USA). The chromatograph used a flame ionization detector (FID), and a capillary column (Supelco SPB­ 1,60 m x 0.25 mm X 11lm film) with temperature programming (35 oC for 3 min, 5 oC/min up to 150 oC, 15 OC/min up to 240 oC, folIowed by 16 oC/min up to 320 OC). The injector and detector temperatures were set at 325 and 330 oC, respectively. Helium was used as the carrier gas at 430 KPa. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial CeUs 122

The composItIOn of gas (Oz' Nz' and COz) was detennined by a HP gas chromatograph system (model HP 6890, Hewlett-Packard Co., Wilmington, Delaware, USA) equipped with a thermal conductivity detector (TCD), and a packed column (Supelco, 3.2 mm x 3.7 m, 60/80 mesh Chrom 102). Both the injector and detector temperatures were set at 125 oc. The column temperature was first held at 35 oC for 7.5 min, then programmed to 100 oC at a rate of 75 OC/min, and finally held at 100 oC for 2 min. Helium was used as the carrier gas at a flow rate of60 mL/min.

6.3.7. Adsorption of Gasoline on Gellan Gum The adsorption capacity of gellan gum gel for gasoline was detennined by estimating the quantity of gasoline hydrocarbons that could be adsorbed by the unit mass of gel at 10°C. A gellan gum sol was prepared according to the encapsulation procedure. The sol was molded in 5-mL plastic syringes, and left at room temperature to harden. The syringes were refrigerated ovemight. The gel cylinder was removed from the syringe pistol using the syringe plunger, after cutting off the syringe tip. The gel was then cut into smaller I-mL cylinders, which were transferred into a sterile 0.1 % (w/v) calcium chloride solution and refrigerated for 7 days. The gel cylinders were sterilized by autoclaving at 121°C for 20 min before the experiment. The adsorption experiment was performed in 20-mL glass vials, containing 10 mL of a sterile MSM solution. Each vial received 4 mL of gel under aseptic conditions. The vials (in duplicate) were spiked with gasoline to yield total concentrations of I 16,32,64,80, 160,200,320,400,480, and 600 mg L- , and were sealed and cramped with sterile Teflon-lined rubber septa and aluminum caps immediately. The vials were placed in horizontal position on an Orbit shaker in the dark, and were shaken at 100 rpm for 3 days. The concentration of gasoline hydrocarbons in the liquid phase of each vial was detennined by the Headspace-GC technique, in comparison with the control vials containing no gel cylinders, which received 10 - 690 mg L- I gasoline. The amount of gasoline hydrocarbons adsorbed on the gel was found by difference of the final and initial concentrations. The adsorption data were analyzed by the Freundlich isotherm relationship defined by the following equation: q = Kic~/n (6.2) where q is the amount of solute adsorbed per unit mass of adsorbent, Ce is the equilibrium concentration of solute in the bulk solution, Ki is the adsorption equilibrium constant, and n is an empirical constant. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial eeus 123

6.3.8. Adsorption of Gasoline on Soil The adsorption of gasoline on soil was investigated at 10°C. The adsorption experiment was conducted in 20-mL glass vials (in duplicate) each containing lOg of dry soil. The vials were autoclaved at 121°C for 20 min on three consecutive days. Each vial was then filled with 10 mL of a sterile MSM solution. Gasoline was injected into the liquid phase to yield total concentrations of 16, 32, 40, 48, 64, 80, 160, 200, 320, 400, and 600 mg Loi. The vials were hermetically closed with sterile Teflon-lined rubber septa, and cramped with aluminum caps immediately. The vials were shaken in the dark in a horizontal position at 100 rpm. After 3 days, the concentration of liquid phase was determined by the Headspace-GC technique, in comparison with control vials. The soil adsorption data were correlated based on the Freundlich isotherm in a similar fashion to gellan gum gel.

6.3.9. Chemical Stability The chemical stability of gellan gum gel in the presence of phosphate-based anions of MSM solution was evaluated. Gellan gum sol was prepared as described in the encapsulation procedure. The sol was molded in a plastic tube, and left at room temperature to gel. The tube was then refrigerated before the experiment. The gel was removed from the plastic mold using a plunger, and was cut into small cylinders (1.1 cm dia. x 1.5 cm length) by means of a surgical blade. After weighing, three gel cylinders were transferred into small Erlenmeyer flasks (one cylinderper flask), each containing 20 mL of MSM. The flasks were placed on an Orbit shaker (150 rpm) first at 10 oC for 3 days, and then at room temperature for 7 days. Three other cylinders were similarly incubated in the MSM solution but were left at room temperature for 21 days.

6.4. Results A gasoline-degrading enriched bacterial consortium was encapsulated in gellan gum microbeads using the emulsification-intemal gelation method. The biodegradation of gasoline hydrocarbons by free and encapsulated cultures was assessed in liquid suspension and saturated soil microcosms under aerobic conditions at 10°C. Figure 6.1 presents the biodegradation profiles in liquid suspension microcosms.

Encapsulated cells (2.6 mgcells g-I bead) began to degrade gasoline hydrocarbons immediately after inoculation at all gasoline concentration levels (50, 100, 200, 400, and 600 mg LoI). Although free cells exhibited no lag phase in biodegradation of 50 and 100 mg LoI gasoline, an approximate delay of 1, 2, and 5 days was observed for concentrations of 200, 400, and 600 mg LoI, respectively. Encapsulated cells demonstrated a reduction in the adaptation period 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial eeus 124

100

.- 80 -;:R"0

-CO > 60 0 E CD CC 40 I a.. ~ 20

0 0 3 6 9 12 15 100

.- 80 0-;:R"

-CO > 60 0 E CD CC 40 I a.. ~ 20

0 0 3 6 9 12 15 18 21 24 27 30 lime (days)

Figure 6.1. Biodegradation of gasoline in liquid medium by free bacteria (open symbols) and 1 encapsulated bacteria (closed symbols) for initial gasoline concentrations of (mg L- ): 100 (0 e), 200 (0 .),400 (L1 "'), and 600 (V T). Data are mean of duplicate mns, and error bars show the average deviation from the mean value. Encapsulated cell mass loading is 2.6 mgcells g-l bead. Cell concentration is 0.26 gcells L- 1 MSM in both free and encapsulated cell systems. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 125

required by free ce11s when exposed to high concentrations of gasoline hydrocarbons. Fo11owing this adaptation period, biodegradation by free ce11s acce1erated, 1eading to comparable degrees of total petroleum hydrocarbons (TPH) removal, however at rates lower than in the case of encapsulated ce11s. No gasoline disappearance was observed in the control non-inoculated liquid medium or with the deactivated encapsulated ce11 suspensions. Over 90% TPH were removed by encapsulated ce11s within 5 days at 50, 100 and 200 mg L-I, as compared to 76% at 400 mg L- Iand 67% at 600 mg L- I. In comparison, within the same period, free ce11s removed over 90% TPH at 50 and 100 mg L-I, but 78, 33, and 7% TPH at 200, 400, and 600 mg LI, respectively. The experimental data for 50 mg L- I gasoline were similar to 100 mg L- I in both free and encapsulated ce11 systems, and therefore are not shown. The average values of specifie degradation rate (sdr) of gasoline were estimated based on 50% TPH removal and the time required to reach this removal level after the lag period when any. The variations of sdr for free and encapsulated ce11s are depicted against the initial TPH concentration in the liquid phase (as calculated from the data of Table 6.2) in Figure 6.2.

350 e 300 .,.....-. '-0 250 .,...., !!!. Q) 200 () 0) o 0) 150 E -6-- 100 (f) 50 0 0 10 20 30 40 50 60 70 1 TPH (mg L- )

Figure 6.2. Specifie degradation rate (sdr) of gasoline for free (0) and encapsulated (e) bacteria as a function of initial total petroleum hydrocarbons (TPH) concentration in the liquid phase. Encapsulated ce11 mass loading is 2.6 mgcells g-I bead. Ce11 concentration is 0.26 gcells L-I MSM in both free and encapsulated ce11 systems. 6. Biodegradation ofGasoline by Gel/an Gum-Encapsulated Bacterial Cel/s 126

The sdr increased linearly with initial TPH concentration for encapsulated cells, whereas it reached a maximum at an approximate TPH concentration of 40 mg L- I, corresponding to an initial gasoline concentration of 400 mg L- I, and then declined as the concentration was increased. The decrease of cell mass loading below 2.6 mgcells g-I bead substantially decreased the rate ofbiodegradation at all gasoline concentration levels (200,400, and 600 mg L- I). This is illustrated in Figure 6.3 for an initial gasoline concentration of 400 mg L- I. The encapsulated cells at 0.26 mgcells g-I bead removed gasoline hydrocarbons immediately after inoculation at 200 mg L- I but a 2 - 3 day lag phase was observed at higher concentrations. The increase of cell mass loading to 1.0 mgcells g-I bead slightly increased the extent of biodegradation, and the encapsulated cells demonstrated a short one-day lag phase at 600 mg LI. Within 30 days of incubation, the encapsulated cells at 0.26 mKells g-l bead removed 39 -

100

80 ~ -0

-CO > 60 0 E ID 0: 40 I a.. 1-- 20

0 0 3 6 9 12 15 18 21 24 27 30 lime (days)

Figure 6.3. Biodegradation of gasoline in liquid medium by encapsulated bacteria for cell mass loading values of (mgcells gol bead): 10.4 (e), 2.6 (_), 1.0 (....), and 0.26 (T). Data are mean of duplicate mns, and error bars show the average deviation from the mean value. Initial gasoline concentration is 400 mg LI. Cell concentration is 1.04, 0.26, 0.10, and 0.026 gcells L­ I MSM in systems inoculated with encapsulated cells at 10.4, 2.6, 1.0, and 0.26 mgcells g-l bead, respectively. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 127

54% TPH at different gasoline concentration levels, compared to 62 - 77% degraded by encapsulated cells at 1.0 mgcells g-l bead. The extent of TPH removal by encapsulated cells at 2.6 and 10.4 mgcells g-l bead reached considerably higher levels of 95 - 99% within only 7 ­ 10 days of incubation. The average values of volumetrie degradation rate (vdr) and specifie degradation rate (sdr) of gasoline were estimated based on 50% TPH removal, as summarized in Table 6.3. The encapsulated cells (0.26 mgcells g-l bead) removed only 39% TPH at 600 mg L- 1 gasoline within 30 days of incubation. Therefore, the vdr and sdr were calculated based on this value. The vdr varies increasingly with both cell mass loading and gasoline concentration. However, the increase of cell mass loading from 2.6 to 10.4 mgcells g-l bead has a small effect if any on vdr. In general, the sdr is an ascending function of gasoline concentration.

Table 6.3. Variations of volumetrie degradation rate (vdr) and specifie degradation rate (sdr) of gasoline with encapsulated cell mass loading and initial gasoline concentration.

Cell Mass Loading Gasoline vdr a sdr a 1 1 1 (mgcells g-l bead) (mg L- ) (mg L- day-') (mg geells- day-') 10.4 200 30 28 400 58 55 600 n.a. n.a.

2.6 200 28 107 400 50 192 600 67 256

1.0 200 3 32 400 13 129 600 11 108

0.26 200 3 116 400 5 197 b 600 8 300 b a Estimated based on 50% TPH removal. b Calculated based on 30 days incubation period. en.a. =not available 6. Biodegradation of GasoLine by GelLan Gum-EncapsuLated BacteriaL Cells 128

The biodegradation of gasoline in saturated soil microcosms followed somewhat different patterns than in the case of liquid suspensions. As it is shown in Figure 6.4, TPH removallevels for encapsulated cells of various cell mass loading (0.26, 1.0, and 2.6 mgcells g-l bead) were similar. The removal of TPH by microbeads loaded with 0.26 mgcells g-l bead started after 6 days at 600 mg L- 1 gasoline, and the extent of biodegradation gradually increased after this period. Free cells presented a better performance in the removal of TPH at 1 a gasoline concentration of 400 mg L- • The biodegradation was initiated at a higher rate and improved within 7 days, but eventually slowed down and leveled off after 10 days at about 55% TPH removal. However, when exposed to a gasoline concentration of 600 mg LI, free cells demonstrated more or less similar activity to that of encapsulated cells. The same levels of biodegradation were achieved by both free and encapsulated cells after 30 days, while no further improvement in the extent of removal was observed after this period. At the end of the biodegradation experiments, the TPH concentration in the aqueous phase was within a range of 0.4 - 0.7 mg L- 1 and 0.1 - 0.9 mg L- 1 for soil microcosms inoculated with free and encapsulated cells, respectively. The experiments performed with control saturated soil systems indicated no removal of TPH. The biodegradation of gasoline was not limited by oxygen. The analysis of gas samples collected at the end of experiments confirmed the theoretical calculations, verifying the aerobic biodegradation of gasoline. The theoretical chemical oxygen demand (COD) of gasoline was calculated to be 3.42 mg 02 mg-l, based on the 25 most representative hydrocarbons of a commercial gasoline (Fan and Krishnamurthy, 1995). The solubility of 1 1 oxygen in MSM (salinity 4,336 mg L- ) is 8.5 mg L- at 22°C and barometric pressure (Metcalf & Eddy, 1979). From the ideal gas law, the concentration of oxygen in the serum bottle headspace was calculated to be about 2,000 mg L- 1 MSM. Thus, based on the theoretical COD of gasoline, the total amount of oxygen available in each bottle was sufficient 1 for complete biodegradation at highest initial gasoline concentration (i.e. 600 mg L- ). Indeed, the analysis of the gas samples by the Ge technique indicated considerable levels of residual oxygen in the system for the entire range of gasoline concentration tested (data not presented). The variations of the mass of gasoline adsorbed per unit mass of gellan gum gel and soil, q (mg g-l) are depicted against the equilibrium gasoline concentration in solution, Ce (mg 1 L- ) in Figure 6.5. The adsorption isotherms for gasoline were evaluated according to the Freundlich relationship (eq. 6.2). The isotherm for the adsorption of gasoline on gellan gum 3 was linear with an equilibrium constant (Kj ) of 8.86 X 10- • Gasoline was adsorbed on soil to 3 a lesser degree, yielding a Kj of 9.58 x 10- and an empirical constant (n) of 1.40. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial CeUs 129

100

.- 80 (a) ~ 0

-CO > 60 0 E Q) CC 40 I a.. 1- 20

0 100

.- 80 (b) ~ 0

-CO > 60 0 E Q) CC 40 I a.. 1- 20

0e.-...... =-.Jl J...... L...... L.....J...... l.-J.-l.-J.....I...... L.....J..-l--L-l.-L.....L...... L.....J..-l--L.....J...-L...... L....I 036 9 12 15 18 21 24 27 30 lime (days)

Figure 6.4. Biodegradation of gasoline in saturated soil by free bacteria (0) and encapsulated bacteria for cell mass loading values of (mgcells g-I bead): 2.6 (e), 1.3 (_), and 0.26 (À). Data are mean of duplicate runs, and error bars show the average deviation from the mean value. Initial gasoline concentration is (mg LI): 400 (a) and 600 (b). Cell concentration is 0.52 gcells L-I MSM in free cell systems, and is 0.52, 0.26, and 0.052 gcells L-I MSM in systems inoculated with encapsulated cells at 2.6, 1.3, and 0.26 mKells g-I bead, respectively. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 130

The adsorption experiments revealed that in addition to biological degradation, physical adsorption on the polymerie carrier as weIl as on soil is responsible for the removal of gasoline hydrocarbons.

1.4 e 1.2 e 1 ..- 0> 0.8 0> -E 0.6 --0- 0.4 0.2 0 0 40 80 120 160 200 C (mg/L) e

Figure 6.5. Adsorption of gasoline on gellan gum gel (e) and on soil (0).

Gellan gum gel was verified to be chemically stable in the presence of phosphate ions in the MSM solution. The experiments carried out with gel cylinders showed no measurable change in weight or visible change in the geometry of these cylinders after incubation in the MSM solution for up to 21 days. Moreover, gellan gum cylinders remained intact when exposed to high concentrations of gasoline in the MSM solution during adsorption experiments.

6.5. Discussion The experimental results demonstrated high biodegradation activity of gellan gum­ encapsulated cells in the removal of gasoline from liquid suspension microcosms, in comparison with free cells. The encapsulated cells degraded gasoline at a much higher initial 6. Biodegradation ofGasoZine by GelZan Gum-EncapsuZated BacteriaZ CeUs 131

rate, and without an extended lag period as encountered with free cells at equivalent microbial concentrations. The superior performance ofencapsulated cells compared to free cells can he explained by the protective effect of gellan gum against the toxicity of high gasoline concentrations. The protective action of the gel matrix may be attributed to one or a combination of: (l) adsorption of gasoline hydrocarbons by the gel matrix, which lowers the concentration of the dissolved hydrocarbons in the cell microenvironment within the porous microbeads; (2) interaction of hydrocarbons with phospholipid compounds constituting the cell membrane, which may cause a damage to the membrane structure, increasing its permeability, and leading to the loss of intracellular ions and metabolites. Suspended free cells are readily exposed to high levels of toxic hydrocarbons, which inhibit their biological activity. In contrast to free cells, the release of intracellular material from the encapsulated cells is lower and potentially retarded by the gel matrix; (3) formation of micro-colonies within the gel matrix such that the cells located at the outer layers of these colonies provide a diffusion barrier for the cells located at the internaI layers, limiting their exposure to high levels of toxie hydrocarbons; or (4) formation of micro-colonies within the gel matrix may also limit the loss of intracellular material from the damaged cells. Upon the growth of cell aggregates the ratio of external colony surface area to colony volume decreases, minimizing the rate of the loss of intracellular compounds accumulated inside the colony. It is noteworthy that only a small fraction of initial gasoline remains as dissolved in the aqueous phase (Table 6.2). Upon injection of gasoline into the liquid phase in a serum bottle, a large fraction is flashed into the headspace. The residual hydrocarbons in the aqueous phase are partitioned between the solid (i.e. gellan gum mierobeads and soil) and liquid phases. As revealed by adsorption experiments, a considerable fraction of gasoline is adsorbed on gellan gum gel, reducing the concentration of hydrocarbons in the liquid phase. This effect may improve the performance of encapsulated cells in comparison with free cells, when suspended in the liquid medium. In liquid suspension microcosms, gasoline hydrocarbons are distributed between the gas and liquid phases during biodegradation. Hydrocarbons in the gas phase are not available to the free or encapsulated microorganisms. However, by depletion of the hydrocarbons in the liquid phase, more hydrocarbons diffuse from the gaseous into the aqueous phase. Since the serum bottles were shaked during the biodegradation experiments, it may be assumed that the diffusion and dispersion of hydrocarbons from the gas phase are not limiting to biodegradation reactions. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 132

The plots of specifie degradation rate (sdr) of gasoline as a function of the initial TPH concentration in the liquid phase (Figure 6.2) indicated that free and encapsulated ce11s present different biodegradation kinetic profiles. The kinetic profile of free ce11s is clearly of the Andrews type (substrate inhibition) with a maximum sdr of about 185 mg gcells- I dol at TPH concentration in the 35 - 40 mg L- I range, while sdr values decrease at higher TPH concentrations. In contrast, the kinetic of biodegradation by encapsulated ce11s appeared as a first-order function of the TPH concentration, at least for the TPH concentration range studied, without any sign of inhibition. It is worthwhile to notice also that the sdr of free and encapsulated ce11s are quite comparable in the low concentration range, indicating that they likely have similar half-saturation or affinity constant (KJ This means that there was no substrate limitation within the ge11an gum microbeads, likely because oftheir very sma11 size. The decrease of cell mass loading below 2.6 mgcells g-l bead caused an abrupt decrease in the volumetrie degradation rate (vdr) of gasoline (Table 6.3), but the increase of ce11 mass loading above this level did not enhance the vdr considerably. The variation of vdr and sdr with cell mass loading and gasoline concentration indicates that at higher encapsulated ce11 levels (2.6 and 1004 mgcells g-I bead) the biodegradation is limited by gasoline concentration. However, at lower encapsulated cell levels (0.26 and 1.0 mgcells g'l bead), the biodegradation is rather limited by cell mass loading, as indicated by specifie rates being in the same range. This can explain the lag phase observed in biodegradation with microbeads at 0.26 and 1.0 mgcells g-l bead. In the saturated soil rnicrocosms, a large arnount of the gasoline is adsorbed on soil, as indicated in Table 6.2. The concentration of hydrocarbons in the bulk liquid and accordingly in the pore solution within the gel matrix is lower, thus more limiting. In contrast to liquid suspension microcosms, free and encapsulated ce11s were trapped within the porous soil media in soil microcosms. The agitation of a thin layer of liquid standing over the bulk soil matrix may not provide sufficient rnixing with the gas at the interface, lirniting the mass transfer of hydrocarbons to the bulk aqueous phase, and leading to inferior biological activity in the removal ofTPH. The possibility of the release of a sma11 fraction of free ce11s into the liquid layer, and subsequent growth of such ce11s in suspended forrn may explain the higher TPH removal by free cells, compared to encapsulated cells, at 400 mg L- I gasoline (Figure 604a). The adsorption of hydrocarbons on ge11an gum and soil, along with poor mass transfer to the bulk aqueous phase may expIain the slow removal of TPH in saturated soil microcosms. The final concentration of TPH was less than 1 mg L- I MSM in all saturated soil microcosms. Ifthe adsorption on soil is assumed to be a reversible process, then the TPH 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial Cells 133

concentration in the soil will be substantially lower than the initial corresponding values. Thereby, the overall removal of TPH from the soil microcosms would be higher than the values indicated in Figure 6.4. This method of preparation of soil rnicrocosms was to simulate an in situ biodegradation scenario following the bioaugmentation of a contaminated aquifer. It was preferred to use a fixed soil bed system rather than a soil suspension rnixed with microbeads. A suspension system would have caused undesired breakage of rnicrobeads, leading to the release of entrapped cells. The use of gellan gum as a bacterial carrier for in situ bioaugmentation of a contarninated aquifer could be also advantageous regarding its sorptive characteristics. The adsorption of dissolved gasoline hydrocarbons by geIlan gum enhances the removal of hydrocarbons, limiting the dispersal of contarninants in the aquifer. Chernical stability is also one of the advantages of geIlan gum gel over other commonly used natural polysaccharide gels such as calcium alginate. In earlier studies, calcium alginate dissolved in the presence of calcium-chelating agents such as potassium phosphate (Lee et al., 1994), or disintegrated upon contact with a medium containing benzene (Paje et al., 1998). The increase ofencapsulated ceIl mass loading was an efficient method to enhance the biodegradation of gasoline hydrocarbons. It should be noted, however, that the efficiency of this method could be restricted by geometrical as weIl as biological constraints. The capacity of a rnicrobead « 50 llm dia.) to encapsulate bacterial ceIls (l - 3 llm) is lirnited. Besides, very high concentrations of ceIls may not be desired since in such instances the ceIls entrapped within the surface layers of a bead may consume all the available substrate, leaving the interior cells under a starving condition, and therefore, rendering a considerable fraction of the bead ineffective. The results showed that the increasing of encapsulated ceIl mass loading beyond 2.6 mgcells g-! bead to achieve higher degradation rates was unnecessary. The present study was conducted to evaluate the biodegradation activity of geIlan gum-encapsulated bacterial ceIls in the removal of gasoline. The results obtained with encapsulated ceIls indicated a protective effect by the gel matrix, shortening an adaptation period required by free ceIls when exposed to high contaminant concentrations. The results demonstrate the successful performance of encapsulated ceIls in the removal of gasoline hydrocarbons in liquid suspension and saturated soil systems. High degrees of biodegradation were achieved with encapsulated ceIls in liquid suspensions, although the biodegradation in saturated soil microcosms was likely to be lirnited by insufficient level of dissolved hydrocarbons available to encapsulated ceIls. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial eeus 134

Acknowledgements The authors would like to thank the Natural Sciences and Engineering Research Council of Canada and the National Research Council of Canada (NRC paper no. 00000) for financial support, and McGill University for the awarding of the Max Stem Fellowship to P.M. The technical assistance of Stephane Deschamps is gratefully acknowledged. 6. Biodegradation ofGasoline by Gellan Gum-Encapsulated Bacterial eeus 135

References Atlas, R. M. and Cerniglia, C. E. (1995). Bioremediation of petroleum pollutants: Diversity and environmental aspects of hydrocarbon biodegradation. BioScience 45: 332-338. Bettmann, H. and Rehm, H. J. (1984). Degradation of phenol by polymer entrapped microorganisms. Applied Microbiology and Biotechnology 20: 285-290. Cassidy, M. B., Lee, H. and Trevors, J. T. (1996). Environmental applications of immobilized microbial cells: A review. Journal ofIndustrial Microbiology 16: 79-1Ol. Cherry, J. A. (1987). Groundwater Occurrence and Contamination in Canada. Canadian Aquatic Resources 215: 387-426. Day, M. J., Reinke, R. F. and Thomson, J. A. M. (2001). Fate and transport of fuel components below slightly leaking underground storage tanks. Environmental Forensics 2: 21-28. Fan, C. Y. and Krishnamurthy, S. (1995). Enzymes for enhancing bioremediation of petroleum contaminated soils: A brief review. The Journal of the Air & Waste Management Association 45: 453-460. Gmiz, K. and Kriston, E. (1995). In situ bioremediation of hydrocarbon in soil. Journal of Soil Contamination 4: 163-173. Hall, B. M., McLoughlin, A. J., Leung, K. T., Trevors, J. T. and Lee, H. (1998). Transport and survival of alginate-encapsulated and free lux-lac marked Pseudomonas aeruginosa UG2Lr cells in soil. FEMS Microbiology Ecology 26: 51-61. Lee, S.-T., Rhee, S. K. and Lee, G. M. (1994). Biodegradation of pyridine by freely suspended and immobilized Pimelobacter sp. Applied Microbiology and Biotechnology 41: 652-657. Manohar, S. and Karegoudar, T. B. (1998). Degradation of naphthalene by cells of Pseudomonas sp. strain NGK 1 immobilized in alginate, agar and polyacrylamide. Applied Microbiology and Biotechnology 49: 785-792. Maura Jutras, E., Smart, C. M., Rupert, R., Pepper, I. L. and Miller, R. M. (1997). Field-scale biofiltration of gasoline vapors extracted from beneath a leaking underground storage tank. Biodegradation 8: 31-42. McLoughlin, A. J. (1994). Controlled release of immobilized cells as a strategy to regulate ecological competence of inocula. Advances in Biochemical Engineering / Biotechnology 51: 1-45. Metcalf & Eddy, 1. (1979). Wastewater Engineering Treatment, DisposaI, and Reuse. McGraw-Hill, New York. 6. Biodegradation ofGasoline by GeUan Gum-Encapsulated Bacterial eeUs 136

Morgan, P., Lewis, S. T. and Watkinson, R. J. (1993). Biodegradation of benzene, toluene, ethylbenzene and xylenes in gas-condensate-contaminated ground-water. Environmental Pollution 82: 181-190. Paje, M. L., Marks, P. and Couperwhite, I. (1998). Degradation ofbenzene by a Rhodococcus sp. using immobilized cell systems. World Journal of Microbiology and Biotechnology 14: 675-680. Solano-Serena, F., Marchal, R, Blanchet, D. and Vandecasteele, J.-P. (1998). Intrinsic capacities of soil microfiora for gasoline degradation. Biodegradation 9: 319-326. Solano-Serena, F., Marchal, R, Casarégola, S., Vasnier, c., Lebeault, J.-M. and Vandcasteele, J.-P. (2000a). A Mycobacterium strain with extended capacities for degradation of gasoline hydrocarbons. Applied and Environmental Microbiology 66: 2392-2399. Solano-Serena, F., Marchal, R, Huet, T., Lebealt, J.-M. and Vandecasteele, J.-P. (2000b). Biodegradability of volatile hydrocarbons of gasoline. Applied Microbiology and Biotechnology 54: 121-125. Solano-Serena, F., Marchal, R, Lebeault, J.-M. and Vandcasteele, J.-P. (2000c). Distribution in the environment of degradative capacities for gasoline attenuation. Biodegradation 11: 29-35. Solano-Serena, F., Marchal, R, Ropars, M., Lebeault, J.-M. and Vandecasteele, J.-P. (1999). Biodegradation of gasoline: kinetics, mass balance and fate of individual hydrocarbons. Journal of Microbiology 86: 1008-1016. Somerville, H. l, Mason, J. R. and Ruffell, R N. (1977). Benzene degradation by bacterial cells immobilized in polyacrylamide gel. European Journal of Applied Microbiology 4: 75-85. Suzuki, T., Yamaguchi, T. and Ishida, M. (1998). Immobilization of Prototheca zopfi in calcium-alginate beads for the degradation of hydrocarbons. Process Biochemistry 33: 541-546. Trevors,l T., van Elsas, 1 D., Lee, H. and van Overbeek, L. S. (1992). Use of alginate and other carriers for encapsulation of microbial cells for use in soil. Microbial Releases 1: 61-69. Trevors, J. T., van Eisas, J. D., Lee, H. and Wolters, A. C. (1993). Survivai of aiginate­ encapsulated Pseudomonas fluorescens cells in soil. Applied Microbiology and Biotechnology 39: 637-643. Weir, S. c., Dupuis, S. P., Providenti, M. A., Lee, H. and Trevors, J. T. (1995). Nutrient­ enhanced survival of and phenanthrene mineralization by alginate-encapsulated and 6. Biodegradation ofGasoZine by GelZan Gum-EncapsuZated BacteriaZ CeUs 137

free Pseudomonas sp. UG14Lr cells in creosote-contaminated soil slurries. Applied Microbiology and Biotechnology 43: 946-951. Weir, s. c., Providenti, M. A., Lee, H. and Trevors, J. T. (1996). Effect of alginate encapsulation and selected disinfectants on survival of and phenanthrene mineralization by Pseudomonas sp. UG14Lr in creosote-contaminated sail. Journal of Industrial Microbiology 16: 62-67. Yerushalmi, L. and Guiot, S. R. (1998). Kinetics of biodegradation of gasoline and its hydrocarbon constituents. Applied Microbiology and Biotechnology 49: 475-481. Yerushalmi, L., Manuel, M. F. and Guiot, S. R. (1999). Biodegradation of gasoline and BTEX in a microaerophilic biobarrier. Biodegradation 10: 341-352. Zhang, W.-X. and Bouwer, E. 1. (1997). Biodegradation of benzene, toluene and naphthalene in soil-water slurry microcosms. Biodegradation 8: 167-175. Zhou, E. and Crawford, R. L. (1995). Effects of oxygen, nitrogen, and temperature on gasoline biodegradation in sail. Biodegradation 6: 127-140. 138

7• Conclusions

The main conclusions of this study were as follows:

Cl) The experimental results of this study demonstrated the feasibility of application of encapsulated cells to in situ bioaugmentation of sorne contaminated aquifers. (2) The emulsification-internal gelation procedure can be used to encapsulate high levels of viable gasoline-degrading bacteria in size-controlled gellan gum microbeads (10 ­ 50 /lm). The statistical analysis demonstrated a high degree of precision or repeatability in formulation and size-distribution of gellan gum microbeads. (3) Gellan gum microbeads pulsed into soil columns as a suspension in artificial groundwater, were transported through a wide range of gravel (2 - 16 mm) and sand (0.25 - 2 mm) porous media. Substantiallevels of pulsed microbeads traveled through llO-cm length of soil consisting of very fine, fine, and medium gravel (2 - 16 mm) as weIl as soil consisting of very fine and fine gravel (2 - 8 mm). However, finer granular media including very fine gravel (2 - 4 mm), very coarse sand Cl - 2 mm), soil consisting of coarse (0.5 - 1 mm) and very coarse sand Cl - 2 mm), and soil consisting of medium (0.25 - 0.5 mm), coarse (0.5 - 1 mm) and very coarse sand Cl - 2 mm) imposed a considerable filtration effect on suspended microbeads. (4) The transport of microbeads pulsed into gravel and sand porous media, demonstrated the feasibility of dispersion of encapsulated cells through the soil matrix of a contaminated aquifer with similar grain size distribution and hydrodynamic characteristics. Detection of microbeads in the effluent of aIl columns suggested that the transport of microbeads through medium gravel to medium sand fractions of soil is feasible, and that the maximum achievable travel distance by the microbeads is greater than 110 cm. Therefore, in situ bioaugmentation of the aquifer may be carried out by installing the microbead injection wells at a minimum distance of 110 cm apart from each other. 7. Conclusions 139

(5) Gellan gum microbeads injected intermittently into soil columns as a suspension in artificial groundwater were transported through a wide range of sand (0.125 - 2 mm) porous media. The largest amount of microbeads was discharged from the column packed with coarse and very coarse sand. After 72 h of injection, about 74% of injected microbeads dispersed across the llO-cm column packed with coarse and very coarse sand. Within the same period, comparable amounts of microbeads were dispersed within the first 50 cm of the column packed with medium, coarse, and very coarse sànd, and within the first 20 cm of the column packed with fine (0.125 - 0.25 mm), medium, coarse, and very coarse sand. These results confirmed that the microbeads can be transported across distances greater than 110 cm, and also suggested that the formation of a bioactive zone of dispersed encapsulated cell microbeads can be feasible in a soil consisting of medium to very coarse sand. (6) Under the employed operating conditions it was not possible to readily mobilize the microbeads across the column packed with fine to very coarse sand, and about 87% of microbeads were immobilized within the first 20 cm of this column after 128 h injection. However, the detection of microbeads at the outlet of this column suggested that the microbeads could travel across a soil matrix with such grain size distribution and hydrodynamic properties.

1 (7) Gellan gum-encapsulated cells (2.6 mgcells t bead) demonstrated a high biological 1 activity in the removal of gasoline (50 - 600 mg L- ) in aqueous systems at 10 oc. High degrees of biodegradation were achieved with encapsulated cells in liquid suspensions, although the biodegradation in saturated soil microcosms was limited by insufficient level of dissolved hydrocarbons available to encapsulated cells. (8) The results obtained with encapsulated cells indicated a protective effect by the gel matrix, shortening an adaptation period required by free cells when exposed to high gasoline concentrations.

(9) The reduction of encapsulated cell mass loading below 2.6 mgcells g-l bead caused a substantial decrease in the extent and the rate of biodegradation within the incubation period, whereas the increase of cell mass loading had a minor effect towards the enhancement of these values. 7. Conclusions 140

(10) Gellan gum gel adsorbed gasoline hydrocarbons but remained stable in the presence of hydrocarbons. The adsorption effect can be correlated using the Freundlich isotherm. 141

8. Summary

Bioaugmentation is one of the major processes for remediation of contaminated sites. This process involves the addition of specifically adapted microorganisms to a contaminated soil or groundwater system, which is lacking microbial populations with appropriate biological activity for the degradation of contaminants. Despite the prospects for beneficial effects of bioaugmentation, a number of field studies have occasionally reported the failure of such a remediation process in contaminated aquifers. Inhibition of augmented microorganisms by high levels of toxic contaminants, outcompetition by indigenous microflora, and particularly, insufficient transport of microbes through soil from the point of injection are among the drawbacks of bioaugmentation. Bacterial transport through porous soil media is a fundamental process in delivery of degrading cells to contaminated sites. Transport of suspended cells is often obstructed by a sorption process, leading to attachment of cells to soil grains, followed by synthesis of exopolymers by the attached cells. The growth of attached bacteria and development of interstitial biofilms within short distances of the injection point can cause a permanent decrease in soil permeability, resulting in the failure of the bioaugmentation process. Encapsulation of viable active bacteria in polymeric gel microbeads may be an alternative solution to overcome practicallimitations in the use of conventional free cell suspensions for bioaugmentation. The polymeric matrix of microbeads provides a defined, stable, consistent, and protective microenvironment for the entrapped cells to both thrive and degrade the toxic contaminants. Encapsulation may also increase transport distances of bacteria since the polymeric capsules isolate the cells from the exterior environment, eliminating chemical and biological patterns of clogging. The transport of encapsulated cell microbeads is controlled by the filtration effect of porous soil media, which is influenced by suspended particles and suspension fluid as weIl as soil characteristics including its grain size distribution. The latter parameter, which can be translated into pore size distribution, is one of the important parameters affecting the particle transport. Empirical models, describing particle straining criteria, suggest that microbeads smaller 8. Summary 142

than 50 um may freely travel through a wide range of sand and gravel without being sequestered from the carrier fluid. It is evident that the application of encapsulated cells to in situ bioaugmentation of a contaminated aquifer requires production of micrometer-sized cell carriers to enable hydraulic distribution of encapsulated cells into the granular matrix of the aquifer. Therefore, the main objectives of this study were: 1) to develop a technique for encapsulation of active cells in polymerie gel mierobeads; 2) to evaluate biodegradation of gasoline by encapsulated cells in comparison with free (non-encapsulated) cells; and 3) to investigate the feasibility of encapsulated cell bioaugmentation based on the transport of gel microbeads through porous soil media. A consortium of gasoline-degrading bacteria isolated from a contaminated site, was encapsulated in gellan gum microbeads using an emulsifieation-internal gelation method. Gellan gum is a natural polymer (polysaccharide), which forms a gel structure by cooling a warm aqueous solution of polymer in the presence of divalent cations such as calcium ions. The encapsulation technique was based on the emulsification of a dispersion of cells in 0.75% gellan gum solution within a food-grade vegetable oil. The removal of heat from the reaction vesselleads to the gelation of dispersed gellan gum droplets, entrapping the cells within porous gel microbeads. Investigation of the influence of emulsion parameters including stirring rate, disperse phase volume fraction, emulsifier concentration, and emulsification time on size distribution revealed that all but the latter parameter had a favorable effect towards the reduction of microbead size. Variation of the emulsifier concentration within a range of a - 0.15% and disperse phase volume fraction within a range of 0.077 - 0.2, along with a high stirring rate, 4500 rpm, enabled production of microbeads with a narrow size range, la - 50 um. Increasing the emulsifier concentration lowers the interfacial tension, facilitating the disruption and breakage of gellan gum droplets, while increasing the impeller rotational speed provides higher mechanieal energy for the breakage of droplets. Decreasing the disperse phase volume fraction reduces the probability of collision and coalescence of droplets, thus allowing formation of smaller microbeads. It was shown that high concentrations of cells can be encapsulated in gellan 9 gum mierobeads (2.6 x 10 - 1.0 X lOlO cells mL,1 mierobead), and that the cell loading level does not have any effect on size distribution of mierobeads. The statistical analysis of 8. Summary 143

mean diameter of multiple batches of microbeads suggested that gellan gum-encapsulated cells can be produced with a high degree of precision by the emulsification-internal gelation technique. The transport of gellan gum microbeads was studied in horizontal soil columns (5.2 cm id x 110 cm long) packed with various combinations of very fine (2 - 4 mm), fine (4 - 8 mm), and medium (8 - 16 mm) classes of gravel as weIl as fine (0.125 - 0.25 mm), medium (0.25 - 0.5 mm), coarse (0.5 - 1 mm), and very coarse (l - 2 mm) classes of sand. The feasibility of transport of microbeads was first explored in pulse injection experiments, in

1 which a suspension of microbeads in artificial groundwater, AGW (40 mg beads L- ) was injected at about 0.5 L h- 1 into the column for 6 h, followed by injection of bead-free AGW for 42 h. The experimental results showed that microbeads traveled across longer distances in the columns packed with gravel and very coarse sand than in the columns packed with coarse and medium sand. In general, the microbead concentration profiles consisted of a major peak, which appeared along the entire length of the columns packed with medium and fine gravel and the first 50 cm of the columns packed with very fine gravel or very coarse sand within the first 12 h of injection. However, this major concentration peak was not observed in the other columns except at a 5-cm travel distance in the column packed with coarse-very coarse sand. The filtration of microbeads in soil columns is governed by a combination of diffusion, interception, and sedimentation mechanisms. The profound filtration of microbeads in the columns packed wi~h a mixture of medium, coarse, and very coarse sand, and the employment of a low concentration of microbead suspension may have led to the disappearance of the major concentration peak in these columns, compared to what was observed in the columns packed with coarser grains. However, the transport of microbeads along the llO-cm length of sand columns was evident as the discrete concentration peaks were detected in the sampling ports across the entire length as weIl as in the effluent of these columns. Previous studies on transport of bacteria have shown substantial filtration of free cells within short distances of sand media. Although different experimental conditions prevent a direct comparison with these studies, the high degree of recovery of gellan gum microbeads at comparably much longer travel distances of sand matrices suggest that gel microbeads may be used for long-range delivery of bacteria in 8. Summary 144

sorne contaminated aquifers. The experimental results indicated that the gravel classes of soil have a minimal filtration effect on the injected microbeads but the sand classes of soil restrict their transport. Hence, the potential for creation of a biologicaIly active zone of encapsulated ceIls was further investigated in columns packed with fine, medium, and coarse sand classes of soil. A suspension of gellan gum microbeads in AGW was injected at about 0.5 L h- I during 12-h intermittent periods into the sand columns packed either with coarse-very coarse sand, with medium-coarse-very coarse sand, or with fine-medium-coarse-very coarse sand. The concentration of suspension was 40 - 50 mg L- I during the first 48 h of injection but it was raised to 70 - 80 mg L- I after this period, and kept constant for 24 h in aIl columns. The concentration was increased to 100 mg L- I for the column containing fine sand after 72 h of injection, and kept at this level for 15 h. The profiles of microbead concentration at different travel distances indicated that the degree of filtration of microbeads through the sand matrices was a direct function of grain size distribution. After 72 h of injection, about 75% of microbeads were dispersed within a 5 - 110 cm distance in the column packed with coarse-very coarse sand, whereas the comparable amounts of microbeads were captured within a 5 - 50 cm and 5 - 20 cm distance of the columns packed with medium-coarse-very coarse and fine-medium-coarse-very coarse sand, respectively. In general, increasing the concentration of injectant to 70 - 80 mg L- I enhanced the accumulation of microbeads within the short distances from the sand bed inlet, while this effect was more pronounced in the column packed with fine-medium-coarse-very coarse sand. The accumulation of microbeads was accompanied by the increase of hydraulic gradient. Although the hydraulic gradient in the proximity of bed inlet was changed by only two-fold in aIl columns, the increase of concentration to 100 mg L- I in the column containing fine sand, showed a dramatic increase in hydraulic gradient by one order-of­ magnitude within the first 20 cm of this column. These results suggested that injection of a low concentration of microbeads may provide a better opportunity to deliver the encapsulated cells across long distances. The use of suspensions with high concentrations may cause the early clogging of the porous soil media in the vicinity of an injection weIl, leading to failure of the bioaugmentation process. The analysis of size distribution of 8. Summary 145

microbeads collected at different travel distances of the column containing fine grains indicated the filtration of larger microbeads through the porous media, as the mode diameter of these microbeads (12 - 17 um) was smaller than that of the injectant (19 um). The results suggested that under the injection conditions employed in this study, the creation of a bioactive zone may be more feasible in aquifers consisting of medium, coarse, and very coarse sand. Encapsulated cell microbeads are delivered to the contaminated zone of an aquifer through a network of injection/withdrawal wells. The suspension of microbeads can be first injected through an injection weIl, while water is withdrawn through a proximate withdrawal weIl. The process can be later reversed by inverting the direction of flow through the two wells. This procedure may provide a uniform distribution of microbeads through a given distance of porous soil media. Biodegradation of gasoline by free and gellan gum-encapsulated bacteria was the subject of the last part of this study. The main objectives were: 1) to demonstrate the biological activity of specifie gasoline degrading bacteria after encapsulation; and 2) to compare the extent and the rate of biodegradation by encapsulated bacteria with those for free bacteria. The experiments performed in liquid suspension microcosms, demonstrated high activity of encapsulated cells (2.6 mgcells g-I bead) in the removal of gasoline at overall initial concentrations of 50 - 600 mg L- I, degrading over 90% of gasoline hydrocarbons within 7 - 10 days of incubation (100 rpm, 10 oC, dark). The decrease of encapsulated cell mass loading below 2.6 mgcells g-I bead substantially decreased the (volumetrie) rate and the extent of biodegradation, whereas a four-fold increase of entrapped cell concentration had a small affect if any on these parameters. In comparison with encapsulated cells, free cells exhibited a lag phase in biodegradation of 200 - 600 mg L- I gasoline, increasing from 1 to 5 days with the increase of gasoline concentration. Free cells degraded comparable amounts of hydrocarbons within 7 - 10 days at aIl gasoline levels but 600 mg Cl, at which they required up to 30 days. The superior performance of encapsulated cells can be explained by the protective effect of gellan gum gel against the toxicity of high gasoline concentrations. The protective action of gel may be attributed to several phenomena including: 1) adsorption of hydrocarbons on gellan gum, lowering the dissolved hydrocarbon concentrations; 2) formation of microcolonies within the gel matrix, 8. Summary 146

providing a diffusion barrier and protecting cells 10cated in the internaI layers of these mierocolonies; and 3) reduction of the loss of intracellular material from the damaged cells by either the gel matrix or microcolonies. The experiments verified the adsorption of hydrocarbons by gellan gum, however, the verifieation of other hypotheses demands further investigations. The plots of specifie degradation rate of gasoline as a function of initial dissolved TPH concentration suggested that free and encapsulated cells may follow different biodegradation kineties. Free cells exhibited an Andrews-type inhibition model within a range of 5 - 60 mg L- I, while encapsulated cells followed a first-order kinetic model within a range of 5 - 35 mg L- I. The specific degradation rates of free and encapsulated cells were quite comparable at concentrations below 20 mg LI, indicating that there was no substrate limitation within the gel mierobeads. The specifie degradation rate for encapsulated cells was higher than those for free cells at concentrations above 20 mg L- I. Overall, the biodegradation experiments in liquid suspension microcosms demonstrated the superior activity of encapsulated cells as compared to free cells at high levels of toxic gasoline hydrocarbons. AIso, it was concluded that there was an optimum cellloading level (2.6 mgcells g-I) for the effective performance of encapsulated cells. Regarding the promising results of the biodegradation studies in liquid suspension microcosms, the performance of encapsulated cells was further investigated in saturated soil mierocosms. The soil mierocosms were designed to simulate an in situ bioremediation process followed by bioaugmentation. Free cells and encapsulated cell mierobeads were gently dispersed throughout the soil matrix emplaced in a serum bottle. The soil was then saturated with the liquid medium and was spiked with gasoline at initial concentrations of 400 and 600 mg L-I. Free and encapsulated cells degraded comparable amounts of gasoline hydrocarbons. However, both the rate and the extent of biodegradation were much lower, compared to those in liquid suspension microcosms. Phase distribution analyses showed that a relatively smaller fraction of hydrocarbons was initially dissolved in the liquid phase and larger fractions were adsorbed on gellan gum and sail. The gas chromatographic analysis of liquid samples collected at the end of incubation time (30 days) indicated that a very low concentration (below 1 mg L- I) of hydrocarbons was left in the liquid phase. Based on these data and regardless of reversibility of the adsorption process, it was 8. Summary 147

concluded that biodegradation was probably slowed down due to limited transfer of hydrocarbons from the gaseous to the liquid phase as the result of particular design of saturated soil microcosms. The interfacial mixing of bulk liquid phase with the gaseous phase was minimal due to the immobilized saturated soil matrix. Therefore, the growth of bacteria was limited because of lack of sufficient substrates in the liquid phase. Despite the overall poor results obtained with saturated soil microcosms, with respect to biodegradation of hydrocarbons initially dissolved in the liquid phase, it can be speculated that encapsulated cells may demonstrate a high biological activity when exposed to a continuous source of dissolved hydrocarbons, as is the case in a contaminated aquifer. Therefore, it is claimed that gellan gum-encapsulated bacteria may have the potential to be used for in situ bioaugmentation of sorne contaminated aquifers. 148

9. Contributions

The following are claimed as contributions to knowledge:

(1) Gellan gum obtained from a commercial producer was used to encapsulate a gasoline­ degrading bacterial consortium. (2) A two-phase dispersion technique, termed emulsification-internal gelation, was developed to encapsulate bacteria in small gellan gum microbeads (10 - 50 /lm). The influence of emulsion conditions on size distribution of microbeads was demonstrated, and those with substantial impact on microbead size were identified. (3) The transport of gellan gum microbeads through a wide range of porous soil media packed in horizontal columns was demonstrated. The compositional influence of each soil class on microbead transport was presented. (4) The breakthrough data were obtained across a comparatively long travelling distance and injection time, and the feasibility of distribution of microbeads was demonstrated. (5) The aerobic biodegradation of gasoline by encapsulated bacteria was evaluated in liquid suspension and saturated soil microcosms at 10°C. It was shown that encapsulated cells could be used to achieve high degrees of biodegradation within shorter periods of time, as compared to free cells, when exposed to high concentrations of hydrocarbons. The effect of initial gasoline concentration and encapsulated cell mass loading on the rate and the extent of biodegradation was demonstrated. (6) The adsorption of gasoline hydrocarbons by gellan gum gel was evaluated and the extent of adsorption was correlated with the equilibrium concentration of aqueous phase. (7) An encapsulated cell bioaugmentation process was simulated with encapsulated cell microbeads dispersed in saturated soil microcosms. 149

10. Recommendations

The following studies are recommended towards the development and implementation of encapsulated cell bioaugmentation technology:

(1) The emulsification-internal gelation method used in this study enables production of small microbeads suitable for subsurface injection. The technique is based on the dispersion of an aqueous polymer solution in an oil phase, and requires the separation of produced microbeads from the oil phase at the end of the encapsulation process. The elimination of the oil separation step can be accomplished by using other encapsulation techniques such as spray atomization where gelation and hardening processes take place in an aqueous medium. Therefore it is recommended that this technique be employed in future work and be compared with the emulsion technique. (2) Survival of encapsulated bacteria should be investigated as a function of time and environmental conditions. (3) Gellan gum microbeads can be engineered to enable the release of microbial cells after being placed in the contaminated zone of an aquifer. Therefore, the chemical structure of gellan gum gel should be tailored to optimize the rate of cell release if desired. (4) The encapsulation technology may be utilized to deliver substrates, nutrients, and oxygen release compounds to a subsurface contaminated environment. The gel microbeads can be designed to allow sustained release of such compounds during a long period of time if desired. (5) The transport of microbeads through porous soil media should be investigated under a wide range of operational conditions. The delivery of encapsulated cell microbeads can be influenced by injectant concentration, carrier fluid velocity, and mode of operation (continuous vs. intermittent). AIso, further experiments should be performed to evaluate the transport of microbeads through soil of a wide grain size distribution (0.125 - 16 mm). 10. Recommendations 150

(6) The biodegradation of gasoline hydrocarbons by encapsulated cells should be investigated in soil column bioreactors packed with sterile and non-sterile soil to simulate an in situ bioremediation system. (7) Although the aerobic biodegradation of gasoline was the main focus of this study, the encapsulated cell bioaugmentation technology can be used in the case of other environmental contaminants. This technology may be especially suitable to deliver bacteria and nutrients to deep contaminated aquifers where the lack of degrading microorganisms, oxygen, and other nutrients limit natural biodegradation processes.