MSc dissertation FINAL Timea Grego

CIVIL, ENVIRONMENTAL &

GEOMATIC ENGINEERING

DEPARTMENT

EVALUATION OF FINE SAND FILTRATION FOR

LARGE SCALE ALGINATE PURIFICATION FOR

CLINICAL USE

M.SC. DISSERTATION

TIMEA GREGO

M.SC. ENVIRONMENTAL SYSTEMS ENGINEERING

SUPERVISED BY: DR LUIZA CINTRA CAMPOS

DR CLARE SELDEN

London, 31st August 2009

EVALUATION OF FINE SAND FILTRATION FOR LARGE SCALE ALGINATE PURIFICATION FOR CLINICAL USE

TIMEA GREGO II

MSc DISSERTATION SUBMISSION

Student Name: TIMEA GREGO

Programme: M.Sc. Environmental Systems Engineering

Supervisors: Dr Luiza C Campos

Dr Clare Selden

Dissertation Title:

EVALUATION OF FINE SAND FILTARTION FOR LARGE SCALE ALGINATE

PURIFICATION FOR CLINICAL USE

DECLARATION OF OWNERSHIP

I confirm that I have read and understood the guidelines on plagiarism, that I understand the

meaning of plagiarism and that I may be penalised for submitting work that has been

plagiarised.

I declare that all material presented in the accompanying work is entirely my own work except

where explicitly and individually indicated and that all sources used in its preparation and all

quotations are clearly cited.

Should this statement prove to be untrue, I recognise the right of the Board of Examiners to

recommend what action should be taken in line with UCL‟s regulations.

Signature: Date:


TIMEA GREGO III

This M.Sc. Thesis was a collaboration between the Centre for Hepatology at the Royal Free

University College London (UCL) - Royal Free Hospital Campus, and the Civil,

Environmental and Geomatic Engineering Department, UCL.


TIMEA GREGO IV


TIMEA GREGO V

THIS THESIS IS DEDICATED TO THE LOVING MEMORY OF MY FATHER,

JÁNOS GREGÓ (1955-1989)

DIPLOMAMUNKÁMAT ÉDESAPÁM EMLÉKÉNEK AJÁNLOM

GREGÓ JÁNOS (1955-1989)


TIMEA GREGO VI

ABSTRACT

Alginate is a natural polysaccharide, harvested from brown seaweed. Tissue engineering and

medical science mainly utilise its gel-forming “ability in the presence of various divalent

cations” [1]. In this project, alginate is used to microencapsulate HepG2 cells for use in a

Bioartificial Liver machine. Therefore, biocompatibility of alginates for immobilising

hepatocytes is crucial. Several in house purification methods were published, which tried to

achieve a suitable biocompatibility of the polymer by removing the known impurities i.e.

proteins and endotoxins, but failed to maintain the physical properties of alginate solution.

In this study, an attempt was made to purify research grade sodium alginate solution from

“unknown” (micron particulate) impurities by fine sand filtration [2]. For clinical application

through regulatory, authority approval particulates must be removed to prevent transit to the

patient.

The hypothesis was that the sand filtration will remove particulates between 1µm to 10µm

from the solution without inducing physical changes in the polymer solution. The specific

aims were

1. To evaluate the effect of fine sand filtration on the sodium alginate solution‟s

properties.

2. To determine, measure any changes in the physical properties of the non-Newtonian

fluid by rheology.

3. Encapsulate empty alginate beads to assess morphology of beads after filtration.

4. To immobilise encapsulated HepG2 cells in filtered alginate solution and compare

with non filtered alginate encapsulated HepG2 cells using cell growth, viability and

function as comparators.

The results showed that micron particulates removal efficiency of dual media filtration was

higher compared to mono medium filter. However, the purification method of the polymer

had negatively influenced the dynamic viscosity of the solution, thereby altering the

functional properties of the hydrogels.

As can be seen, the hypothesis was not proven in the large scale purification of alginate for

clinical use. The research outcomes highlighted the need for an alternative approach to purify

alginate solution, which could be achieved by utilising the gas-solid separation method of dry

alginate powder.


TIMEA GREGO VII

ACKNOWLEDGEMENTS

My sincerely appreciation goes to my supervisors, Dr Luiza Campos and Dr Clare Selden, for

their support and encouragement during my research work. I highly appreciate both of you for

giving me the opportunity to work on a multidisciplinary research project. Thank you very

much.

I would also like to thank to Amir Gander, for his continuous support, patient and guidance.

Dear Amir, I really appreciate all the help you gave me. You have never let me down and I

thank you very much for this.

I extend my appreciation to the people at the Environmental Engineering Laboratory and

Department of Pharmacology at University College London (UCL), and Hepatology

Laboratory at Royal Free Hospital for their support. I also want to thank to Dr Arnold

Darbyshire for giving the opportunity to use the rotational viscometer at University College

Medical School at the Department of Surgery, Royal Free Hospital.

I would like to express my appreciation to Matt Kite and Graeme Oakes at PAMAS

Partikelmess- und Analysesysteme GmbH for providing the Liquid particle counting

apparatus. I would also like to thank to Sibelco UK Ltd. for supplying the sand samples for

the experiment.

Special acknowledgement goes to, Mr Ian Sturtevant, at University College London, Civil,

Environmental Engineering and Geomatic Engineering Department for his support, time and

technical advice.

I would like to record my appreciation to University of Miskolc, Hungary for providing me

with a solid engineering and science knowledge.

I would like to thank to all my friends for encouraging and helping me during the M.Sc.

course.


TIMEA GREGO VIII

Dear Mummy, I would like to thank you for bringing me up with the greatest love and care. I

really appreciate your support and care. I am so proud of you! Without you, I would not be

the person who I am. Also, I extend my gratitude to my sisters, Erika and Marianna, for their

love and understanding.

As well, I must not forget my loved ones who cannot be with me. Dear Daddy, I have

inherited the strength and passion for science and engineering from you. Your love and

support was always by my side. I miss you so much!

Drága Édesanya, szeretném megköszönni, hogy a legnagyobb szeretetben és gondoskodásban

neveltél fel. Büszke vagyok rád! Merhetettlenül hálás vagyok, a mindenkori támogatásodért

és megértésedért. Nélküled nem lehetnék az, aki vagyok.


TIMEA GREGO IX


TIMEA GREGO X

CONTENTS

ABSTRACT ............................................................................................................................................. VI

LIST OF FIGURES .................................................................................................................................. XII

LIST OF TABLES ................................................................................................................................. XVI

NOMENCLATURE ............................................................................................................................... XVII

LIST OF ABBREVATIONS................................................................................................................... XVIII

1 INTRODUCTION .............................................................................................................................. 1

2 RESEARCH GOALS AND APROACH ................................................................................................ 5

2.1 AIMS OF THE INVESTIGATION AND APPROACH ......................................................................... 5

3 LITERATURE REVIEW ................................................................................................................... 6

3.1 ALGINATE - NATURAL POLYMER ............................................................................................. 6

3.2 STRUCTURE AND PROPERTIES OF ALGINATE ........................................................................... 8

3.3 CHARACTERISTIC AND PROPERTIES OF ALGINATE ................................................................... 9

3.3.1 MOLECULAR WEIGHT ...................................................................................................... 9

3.3.2 SOLUBILITY ................................................................................................................... 10

3.3.3 GEL FORMATION ............................................................................................................ 10

3.3.4 VISCOSITY ..................................................................................................................... 11

3.4 PURIFICATION METHODS OF ALGINATE FOR CLINICAL USE ................................................. 13

3.5 FINE SLOW SAND FILTRATION PROCESS ................................................................................ 11

3.5.1 PHYSICAL PARAMETERS AFFECT THE OPERATION OF FILTER MEDIA ............................ 12

3.5.2 MECHANISM OF FILTRATION ......................................................................................... 13

3.5.3 TYPES OF GRANULAR MEDIA FILTERS ........................................................................... 21

3.5.4 HYDRAULICS OF SAND FILTER ....................................................................................... 24

3.5.5 TIME-INDEPENDENT NON-NEWTONIAN FLUIDS ............................................................ 26

4 MATERIAL AND METHODS ......................................................................................................... 30

4.1 PREPARATION OF 2% HEPES BUFFERED ALGINATE SOLUTION ............................................. 31

4.2 TESTING OF FILTER MEDIA ..................................................................................................... 32

4.2.1 CHEMICAL ANALYSIS OF FILTER MEDIA ........................................................................ 32

4.2.2 PHYSICAL ANALYSIS OF FILTER MEDIA ......................................................................... 32

4.2.3 SETTLING VELOCITY OF INDIVIDUAL GRAINS................................................................ 33

4.2.4 SHAPE (SPHERICITY) OF FILTER MEDIA ......................................................................... 33

4.2.5 POROSITY OF FILTER MEDIA .......................................................................................... 34

4.2.6 DETERMINATION OF HEADLOSS IN CLEAN GRANULAR MEDIA FILTER .......................... 34

4.3 SYSTEMS PREPARATION FOR FILTRATION .............................................................................. 35

4.3.1 TREATMENT OF FILTER MEDIA BEFORE USE .................................................................. 35

4.3.2 TREATMENT OF THE FILTRATION SYSTEMS BEFORE USE ............................................... 35

4.4 FILTRATION SYSTEMS INSTALLATION .................................................................................... 35


TIMEA GREGO XI

4.5 SAND FILTRATION EXPERIMENT ............................................................................................. 40

4.5.1 FILTERED AQUEOUS SODIUM-ALGINATE SAMPLES ........................................................ 41

4.5.2 SODIUM ALGINATE VISCOSITY MEASUREMENT ............................................................. 43

4.5.3 TURBIDITY ..................................................................................................................... 43

4.6 LYOPHILISATION OF FILTERED SODIUM-ALGINATE ................................................................ 43

4.7 RECONSTRUCTION OF SODIUM-ALGINATE ............................................................................. 45

4.7.1 ANALYTICAL CENTRIFUGATION .................................................................................... 45

4.7.2 DETERMINATION OF THE SIZE DISTRIBUTION OF THE PARTICULATES ........................... 46

4.7.3 ENCAPSULATION OF FILTERED SODIUM-ALGINATE ....................................................... 46

4.8 ENCAPSULATION OF HEPG2 CELLS ........................................................................................ 47

5 EXPERIMENTAL RESULTS AND DISCUSSION............................................................................... 48

5.1 TESTING OF FILTER MEDIA ..................................................................................................... 48

5.1.1 PHYSICAL ANALYSIS OF FILTER MEDIA ......................................................................... 48

5.1.2 SPHERICITY (SHAPE) OF FILTER MEDIA ......................................................................... 51

5.1.3 SETTLING VELOCITY OF INDIVIDUAL GRAINS................................................................ 51

5.1.4 POROSITY OF FILTER MEDIA .......................................................................................... 52

5.2 HYDRAULIC OF FILTRATION ................................................................................................... 55

5.2.1 DETERMINATION FLOW REGIMES AND DRAG COEFFICIENT .......................................... 55

5.2.2 DETERMINATION OF CLEAN-WATER HEAD LOSS IN GRANULAR-MEDIA FILTER .......... 57

5.2.3 TURBIDITY ..................................................................................................................... 61

5.3 EVALUATION OF THE VISCOSITY MEASUREMENTS OF SODIUM ALGINATE SOLUTIONS .......... 62

5.3.1 EFFECT OF DIFFERENT CONCENTRATION ON SOLUTION VISCOSITY .............................. 63

5.3.2 EFFECT OF AUTOCLAVING ON SOLUTION VISCOSITY..................................................... 65

5.3.3 EFFECT OF TEMPERATURE ON SOLUTION VISCOSITY..................................................... 66

5.3.4 EFFECT OF FILTRATION ON THE SOLUTION VISCOSITY .................................................. 67

5.4 RECONSTRUCTION OF SODIUM-ALGINATE ............................................................................. 69

5.4.1 EVALUATION OF THE EFFECT OF FINE SAND FILTRATION ON THE SIZE DISTRIBUTION OF

THE PARTICULATES ....................................................................................................................... 69

5.4.2 EVALUATION OF PARTICLE SIZE MEASUREMENT OF FILTERED SODIUM ALGINATE

SOLUTIONS .................................................................................................................................... 73

5.4.3 ENCAPSULATION OF FILTERED SODIUM-ALGINATE ....................................................... 74

5.5 ENCAPSULATION OF HEPG2 CELLS ........................................................................................ 76

5.5.1 MORPHOLOGY OF CELL CULTURED ALGINATE BEADS .................................................. 76

5.5.2 PROTEIN PRODUCTION FROM ENCAPSULATED HEPG2 CELL ......................................... 77

5.5.3 ASSESSMENT OF CELL VIABILITY IN ALGINATE BEADS ................................................. 78

6 CONCLUSIONS .............................................................................................................................. 81

7 RECOMMENDATIONS ................................................................................................................... 84

7.1 OPERATION OF THE GAS-SOLID CYCLONE .............................................................................. 85

REFERENCES ........................................................................................................................................ 88


TIMEA GREGO XII

LIST OF FIGURES

PHOTOGRAPHS: All of the photographs for this dissertation were taken by Timea Grego, unless

otherwise stated.

FIGURE 1.1 THE THREE PHASES OF THE BAL SYSTEM

FIGURE 3.1 LAMINARIA GENERA

FIGURE 3.2 MONOMER SALTS OF ALGINATE

FIGURE 3.3 SHEAR RATE VERSUS VISCOSITY OF DIFFERENT ALGINATE

SOLUTION CONCENTRATION

FIGURE 3.4 CORRELATION BETWEEN TEMPERATURE AND THREE

DIFFERENT 1% ALGINATE SOLUTION

FIGURE 3.5 SOLUTION CONCENTRATION EFFECTS ON THE POLYMER

VISCOSITY

FIGURE 3.6 HYDRAULICS OF A FLUID THROUGH PERMEABLE MEDIA

FIGURE 3.7 DRAG COEFFICIENT (CD) FOR SPHERES AS A FUNCTION OF

RE OVER LAMINAR-, TRANSITION- AND TURBULENT

FLOW REGIMES. IT CAN BE SEEN FROM THE FIGURE, THAT

RE INCREASES WITH DECREASING CD.

FIGURE 3.8 GRAVITATION FORCE (FG), BUOYANT FORCE (FB) AND

DRAG FORCE (FD) AFFECT THE MOTION OF A PARTICLE IN

WATER DURING SEDIMENTATION

FIGURE 3.9 TRANSPORT MECHANISMS DURING FILTRATION

FIGURE 3.10 ELECTRICAL DOUBLE LAYER

A) ELECTROSTATIC FIELD AROUND AN ANIONIC PARTICLE

B) ILLUSTRATION OF ZETA POTENTIAL, SHEAR

PLANE, AND ELECTROSTATIC POTENTIAL

FIGURE 3.11 DIFFERENT SEPARATION PROCESSES IN ACCORDANCE

WITH THE SIZE DISTRIBUTION OF POSSIBLE IMPURITIES

PRESENT IN WATER

FIGURE 3.12 DIFFERENT SAND FILTER MEDIA SYSTEMS

FIGURE 3.13 VISCOSITY OF NEWTONIAN FLUIDS

FIGURE 3.14 FLOW CURVE OF NEWTONIAN FLUIDS


TIMEA GREGO XIII

FIGURE 3.15 FLOW CURVES OF VARIOUS TIME-INDEPENDENT NON

NEWTONIAN FLUIDS

FIGURE 4.1 LIFE CYCLE OF THE EXPERIMENTAL WORK

FIGURE 4.2 PHASE I. FILTRATION: SMALL SCALE SYSTEM BEFORE

THE FILTER MEDIA INSTALLATION

FIGURE 4.3 PHASE IV. FILTRATION: LARGE SCALE, SINGLE MEDIA

FILTRATION SYSTEM

FIGURE 4.4 PHASE IV. FILTRATION: DUAL MEDIA FILTRATION

SYSTEM

FIGURE 4.5 SHELL FREEZING PROCESS OF FILTERED NA-ALGINATE

SOLUTION

FIGURE 4.6 DRIED ALGINIC ACID AFTER LYOPLISIATION

FIGURE 4.7 ANALYTICAL CENTRIFUGE ANALYSIS BY EPPENDORF

CENTRIFUGE 5415R

FIGURE. 5.1 GRAIN SIZE DISTRIBUTION GRAPH OF RH-70

PARTICLE SIZES (µM) VERSUS CUMULATIVE PASSING OF

THE GRAINS (%)

FIGURE. 5.2 GRAIN SIZE DISTRIBUTION GRAPH OF RH-110 PARTICLE

SIZE (µM) VERSUS CUMULATIVE PASSING OF THE

PARTICLES

FIGURE 5.3 MASS –VOLUME RELATIONSHIP OF A SEDIMENT SAMPLE

FIGURE 5.4 (A-F) VOLUMETRIC FLOW RATES VERSUS SAMPLING

DATES DURING PHASE I-III AND V. FILTRATION PERIODS

FIGURE 5.5 A-D FLOW CURVES OF NEWTONIAN AND TIME-

INDEPENDENT NON-NEWTONIAN FLUIDS

FIGURE 5.6 DYNAMIC VISCOSITY OF 0.2% NAALG SOLUTION AS A

FUNCTION OF SHEAR STRESS (PA)

FIGURE 5.7 FLOW CURVES OF AUTOCLAVED NAALG (CONTROL)

SOLUTIONS

FIGURE 5.5 A-D FLOW CURVES OF NEWTONIAN AND TIME-

INDEPENDENT NON-NEWTONIAN FLUIDS

FIGURE 5.6 DYNAMIC VISCOSITY OF 0.2% NAALG SOLUTION AS A

FUNCTION OF SHEAR STRESS (PA)

FIGURE 5.7 FLOW CURVES OF AUTOCLAVED NAALG (CONTROL)

SOLUTIONS


TIMEA GREGO XIV

FIGURE 5.8 VISCOSITY OF 2% NAALG SOLUTION AT DIFFERENT

TEMPERATURES

FIGURE 5.9 FLOW CURVES OF 2% NAALG SOLUTION AT DIFFERENT

TEMPERATURES

FIGURE 5.10 FLOW CURVES OF FINE SAND FILTERED 0.2% NA-

ALGINATE SOLUTIONS

FIGURE 5.11 FLOW CURVES OF MONO MEDIUM AND DUAL MEDIA

FILTERED 0.2% NA-ALGINATE SOLUTIONS

FIGURE 5.12 PARTICLE SIZE DISTRIBUTION OF NON-PURIFIED 2%

SODIUM ALGINATE SOLUTION

FIGURE 5.13 LENGTH DISTRIBUTION (µM) OF PARTICLES IN RH-70

SAND FILTERED NAALG SOLUTION. (A) PARTICLES

LENGTH UNDER SLOW FLOW RATE (B) PARTICLES

LENGTH UNDER INCREASED FLOW RATE

FIGURE 5.14 LENGTH DISTRIBUTION (µM) OF PARTICLES IN RH-110

SAND FILTERED NAALG SOLUTION. (A) PARTICLES

LENGTH UNDER SLOW FLOW RATE (B) PARTICLES

LENGTH UNDER INCREASED FLOW RATE

FIGURE 5.15 PARTICLES LENGTH DISTRIBUTION (µM) OF LARGE SCALE

FINE SAND FILTERED NAALG SOLUTION

FIGURE 5.16 DISTRIBUTION OF PARTICLES SIZES (µM) AND NUMBERS

OF NA-ALGINATE POWDER

FIGURE 5.17 ALGINATE BEADS FROM NON-PURIFIED 2% SODIUM

ALGINATE SOLUTION

FIGURE 5.18. EFFECT OF THE NA-ALGINATE SOLUTION

CONCENTRATION ON THE MORPHOLOGY AND SIZE OF

EMPTY ALGINATE BEADS. BEADS WERE PREPARED USING

1.25% (A) AND 1.5% (B) DMF FILTERED SOLUTION; C

1.75% MMF FILTERED SOLUTION AND D 2% FINE SAND

FILTERED AQUEOUS NA-ALGINATE.

FIGURE 5.19 PHASE 4: ENCAPSULATED HEPG2 CELLS; MONO MEDIUM

FILTERED 1.875% SODIUM ALGINATE SOLUTION WAS

USED FOR CELL IMMOBILISATION

FIGURE 5.20 PHASE 4: ENCAPSULATED HEPG2 CELLS; DUAL MEDIA

FILTERED 1.875% SODIUM ALGINATE SOLUTION WAS

USED FOR CELL IMMOBILISATION


TIMEA GREGO XV

FIGURE 5.21 ALBUMIN (PROTEIN) PRODUCTION FROM ENCAPSULATED

HEPG2 CELLS IN A NON FILTERED, MONO MEDIUM, AND

DUAL MEDIA FILTERED ALGINATE SOLUTION

FIGURE 5.22 VIABILITY OF ENCAPSULATED HEPG2 CELLS IN A NON

FILTERED AND MONO MEDIUM, AND DUAL MEDIA

FILTERED ALGINATE SOLUTION

FIGURE 5.23 CELLS NUMBER OF PURIFIED AND NON PURIFIED

ALGINATE ENCAPSULATED HEPG2 CELLS

FIGURE 5.24 LIVE AND DEAD ENCAPSULATED HEPG2 CELLS IN MONO

MEDIUM AND DUAL MEDIA FILTERED ALGINATE

MICROCAPSULES AT DAY 3 STAINED WITH FDA AND PI.

FIGURE 7.1 PARTICLE SIZE DISTRIBUTION OF SEVERAL MATERIALS IN

A FUNCTION OF SEPARATION TECHNIQUE; CYCLONES

SEPARATION PARTICLE SIZE 1µM-1,000µM [66].

FIGURE 7.2 REVERSE FLOW CYCLONE SEPARATOR. SUSPENDED

PARTICLES IN GAS ENTER IN THE CYCLONE AND

CENTRIFUGAL FORCES (VORTEX) DRIVE THEM TO THE

CENTRAL PART, WHERE THE PARTICLES WILL BE

SEPARATED.


TIMEA GREGO XVI

LIST OF TABLES

TABLE 1.1 SPECIFICATION OF SODIUM-ALGINATE

TABLE 3.1 APPLICATION OF ALGINATE BY ITS CHARACTERISTICS

AND FUNCTIONALITY

TABLE 3.2 M, G AND MG FRACTION COMPOSITION OF

COMMERCIALLY AVAILABLE ALGAL ALGINATES

TABLE 3.3 SPHERICITY OF SAND

TABLE 3.4 POROSITY OF SAND

TABLE 3.5 DESIGN PARAMETERS OF SINGLE-MEDIUM, DUAL-, AND

MULTI-MEDIA FILTERS

TABLE 4.1 WATER QUALITY PARAMETERS OF MILLI-Q WATER

TABLE 4.2 FILTERED NA-ALGINATE SAMPLES DURING THE FINE

SAND FILTRATION PERIODS

TABLE 5.1 REDUCED DATA OF GRAIN SIZE DISTRIBUTION ANALYSIS

OF RH-70 (COARSE SAND) AND RH-110 (FINE SAND)

TABLE 5.2 DETERMINED DENSITY OF RH-70 FILTER MEDIUM

COMPARED TO DATA AVAILABLE IN LITERATURE

TABLE 5.3 ESTIMATED SPHERICITY OF RH-70 AND RH-110

COMPARED TO DATA AVAILABLE IN LITERATURE

TABLE 5.4 AVERAGE SETTLING VELOCITY OF COARSE AND FINE

SAND GRAINS

TABLE 5.5 ESTIMATED POROSITY OF RH-70 AND RH-110

TABLE 5.6 PHYSICAL PARAMETERS OF THE FILTER MEDIA

TABLE 5.7 ESTIMATED SPHERICITY VALUE AND CALCULATED

EQUIVALENT HYDRAULIC DIAMETER

TABLE 5.8 VOLUMETRIC FLOW RATES, APPROACH VELOCITIES AND

HEADLOSSES DURING FILTRATION

TABLE 5.9 TURBIDITY LEVEL AND CONCENTRATION (NTU) OF THE

EFFLUENT SAMPLES AFTER BACKWASHING

TABLE 5.10 PREPARED SAMPLES FOR VISCOSITY ANALYSIS

TABLE 5.11 EFFECT OF AUTOCLAVING AT 121°C FOR 10 AND 20MINS

ON THE VISCOSITY OF NAALG SOLUTIONS


TIMEA GREGO XVII

NOMENCLATURE

DEFINITION SYMBOL UNIT

GRAVITY g

s

m81.9

SETTLING VELOCITY tv

s

m

APPROACH VELOCITY OF FILTRATION v

s

m

HYDRAULIC DIAMETER d m

FILTER DIAMETER fd m

SIEVE SIZE Sd m

REYNOLDS‟ NUMBER R DIMENSIONLESS

DRAG COEFFICIENT DC DIMENSIONLESS

HEAD LOSS OF THE SAND h m

HEIGHT OF THE SETTLING COLUMN SCh m

BED DEPTH L m

AVERAGE TIME tav [s]

POROSITY OF THE GRANULAR MEDIA n [%]

GREEK SYMBOLS

DENSITY OF WATER AT 20°C1 )20(2 CTOH

998.23

3m

kg

ABSOLUTE VISCOSITY OF WATER AT

20 °C )20(2 CTOH

sPa

sm

kg3100087.1

ABSOLUTE VISCOSITY OF SODIUM

ALGINATE (NaAlg) lgNaA

sPa

sm

kg

GRAIN SPHERICITY OR SHAPE FACTOR Ψ DIMENSIONLESS

DENSITY OF SAND S

32650

m

kgS


TIMEA GREGO XVIII

LIST OF ABBREVATIONS

AMC-BAL Amsterdam Medical Centre-bioartificial liver

ASTM American Standards for Testing

BAL BioArtificial Liver

DMF Dual Media Filter (RH-70&RH-110)

ELAD

EMEA European Medicines Agency

FDA Food and Drugs Administration

FDA Fluorescein diacetate

GRAS Generally Recognised as Safe

MELS Modular extracorporeal liver system

MHRA Medicines and Healthcare products Regulatory Agency

MMF Mono Medium Filter (RH-110)

NaAlg Sodium alginate

PI Propidium iodide

TEMPS Tissue-Engineered Medical Products


TIMEA GREGO XIX


TIMEA GREGO 1

1 INTRODUCTION

This dissertation focuses on the large scale purification treatment of research-grade sodium

alginate (NaAlg) for clinical use. Alginate is a natural biopolymer extracted from brown

seaweed, and used for encapsulation of living cells, thereby allowing their use within a

bioartificial (BAL) device.

Liver disease is the only major cause of death which increases every year in the UK. The

Office of National Statistics reported in September 2008 that liver disease caused an

estimated 161 deaths/million population of male and 78 deaths/million population of

female deaths [3]. Acute Liver failure is often classified as severe liver disease symptoms

(jaundice, encephalopathy) with an 80-90% reduction in liver cell numbers, which has

many causes including alcohol abuse, infection, genetic causes, with a mortality rate of

between 70-90%. Chronic liver disease is often attributed to cirrhosis of the liver,

developed over time often due to alcohol abuse, non-alcoholic steatohepatitis, and chronic

hepatitis [4]. For patients today the only possible treatments are an organ transplantation,

or spontaneous liver regeneration.

Scientists and researchers are continuously seeking to develop non-biological or biological

liver support systems to treat patient with hepatic failure. Non-biological systems classified

as “plasma exchange, albumin dialysis, hemo(dia)filtration, and adsorbent based devices”

[5] provide only blood detoxification. Functions of liver are so complex when hepatic

failure occurs; the accumulated endotoxin, cell growth inhibitors and various known and

unknown materials cause “neurological abnormalities to the liver” and other internal

organs [5]. It is difficult to purify the blood from these compounds by using conventional

treatment methods. Therefore, to fulfil the complex function more advanced hepatocyte

based devices have been developed, such as

ELAD, which is a hollow-fibre bioreactor and contains human C3A cell line,

Modular extracorporeal liver system (MELS) a complex system in which “four

interwoven hollow-fibre capillary systems carry out independent functions”;

Amsterdam Medical Centre-bioartificial liver (AMC-BAL) loaded with “10-14

billion immobilised porcine hepatocytes on a spirally wound polyester fabric”;

HepatAssistTM

device was the first BAL system in “FDA-approved clinical trials”

[5].


TIMEA GREGO 2

The Liver Group has developed a Bioartificial Liver Device (BAL) that aims to bridge the

gap between diagnosis and transplantation or aid spontaneous regeneration. This BAL

system consists of three phases (Figure 1.1) Phase 1 HepG2 (a human hepatocyte immortal

cell line) cells are encapsulated in alginate beads achieving a 3D state, which provides

better cell growth and function [6]. For immobilisation of HepG2 cells purified alginate is

required; Phase 2 Encapsulated cells are proliferated in a Fluidised Bed Bioreactor over 8-

10 days; Phase 3 patient plasma is passed over beads by fluidisation [7].

Figure 1.1 THE THREE PHASES OF THE BAL SYSTEM

PHASE 1 HEPG2 CELLS ARE ENCAPSULATED IN ALGINATE BEADS; PHASE 2 ENCAPSULATED CELLS

ARE PROLIFERATED IN A FLUIDISED BED BIOREACTOR OVER 8-10 DAYS; PHASE 3 PATIENT PLASMA

IS PASSED OVER BEADS BY FLUIDISATION.

“(i.) FDA STAINED HEPG2 CELLS AFTER ENCAPSULATION. (ii.) FDA STAINED HEPG2 CELLS AFTER

8 DAYS OF PROLIFERATION. (iii.) A PHASE CONTRAST ALGINATE BEADS 8 DAYS AFTER INITIAL

IMMOBILISATION. (iv.) FLUIDISED BED BIOREACTOR (FBB) COMPETENCE CULTURE SETUP. (v.)

FBB CONTAINING ALGINATE” [8].

Phase I. Phase II. Phase III.


TIMEA GREGO 3

The biomedical application of alginate for regulatory approval requires the production of a

highly purified polysaccharide. Sodium alginate (Table 1.1) is listed as “direct food

materials affirmed as Generally Recognised as Safe (GRAS)” by the American Food and

Drug Administration (FDA) [9]. In addition to the Food and Drug Administration, USA

(FDA), the European Medicines Agency (EMEA) and the Medicines and Healthcare

products Regulatory Agency (MHRA), and The American Society for Testing and

Materials (ASTM) have also established standards and guidelines for biopolymers in

Tissue-Engineered Medical Products (TEMPs). „ASTM Alginate and Chitosan Standard

Guide‟ provides information on the physical and chemical properties for the biomaterial

[10]. Moreover, Alginate is referenced in the European, British, and United States

Pharmacopoeia and The Complete Drug Reference Book [11-12].

TABLE 1.1 SPECIFICATION OF SODIUM ALGINATE [13]

SODIUM – ALGINATE (NAALG)

CHEMICAL FORMULA n

NaOHC 676

STRUCTURE OF ALGINATE

PROPERTIES OF THE POLYMER Stabilizer, thickener and ability to form gels in the presence of

divalent cations

As an existing scaffold material for immobilising living liver cells, the purity level of the

biopolymer is crucial. According to recent research, alginate may contain “unknown” and

known “contaminants” [2]. The former can be classified as micron particulates, and the

latter as heavy metals, endotoxins, proteins, pyrogens, and polyphenols [14].

Numerous works have been published on the efficient reduction of “known” impurities

from alginates such as free flow electrophoresis (FFE) and multi-step chemical

purification, which utilises chemical treatments [15]. Alginate purified by FFE treatment is

not suitable for clinical use because the process is time consuming and prohibitively

expensive. An advantage of the multi-step chemical approach compared to FFE is that

alginate can be created with different mannuronic (M)- and guluronic (G) acid ratios [15].


TIMEA GREGO 4

The most commonly detected impurities (endotoxin, proteins and polyphenols) can be

eliminated by these purification methods [2]. However, it is important to note that viscosity

of the polymer changes during the purification process for both FFE and multi-step

chemical treatment, which would impact on the functionality within the Liver Group BAL.

An alternative treatment method (Depth Charged Filtration) of 0.2% Na-alginate solution

was tested at UCL Medical School, Royal Free Campus. Zeta Plus Maximizer Dual-Zone

Depth Filters were used, which utilise a cellulose based filtration system. The filter media

is positively charged thus allowing the removal of anionic impurities such as cellular

debris, endotoxins and viruses from the solution [16]. Although it removed larger

particulates, viscosity of the filtered Na-alginate solution was changed as a result of the

filtration process, and made the purified biopolymer unsuitable for cell encapsulation. As

can be seen, during the depth filtration the same problem emerged compared to the

previously described purification processes.

It has become clear after an interdisciplinary literature review, that there is little

precedence for removal of micron particulates („unknown‟ contaminants) using sand

filtration, from analytical grade material for use in a clinical environment. Therefore, fine

sand filtration process was chosen because it does not require pre-treatment of the influent

and it is easy to operate compared to the published in-house alginate purification methods

[2, 17, 14, 18-20].


TIMEA GREGO 5

2 RESEARCH GOALS AND APROACH

2.1 Aims of the investigation and approach

The aim of this dissertation was to explore the hypothesis that sand filtration will remove

particulates from alginate solutions without altering the viscosity of the resultant solution

The specific aims of the research project were:

To remove micro particulates from 1µm to less than 10µm.

To evaluate the effect of filter media size on the micro-particulates removal.

To determine the effect of flow rate and sand depth on particulates removal efficiency.

To measure the viscosity of the non-Newtonian fluid during the treatment process life-

cycle, to indicate changes in alginate composition.

To determine the particle size distribution of micron particulates, which validates the

efficiency of filtration.

To use the filtered alginate for the encapsulated HepG2 cells.

To evaluate the effects of different sodium-alginate concentrations on the bead shape,

if bead integrity is compromised.

To evaluate the growth and function of the encapsulated cells over 15 days.


TIMEA GREGO 6

3 LITERATURE REVIEW

The present research is a multidisciplinary project incorporating medical science and

engineering. To understand the complexity of the research project, the holistic nature of the

project was investigated. There were three dimensions of this view

Spatial aspect, in which an investigation was undertaken to determine the link

between this research environment in a global context, by establishing the reason

for the study.

Temporal view, which requires investigating the research project in terms of the

past, present and future.

System dimension, which aimed to demonstrate the interaction between the sub –

system and the super system and helped to build up a critical picture of the problem

statement [21].

3.1 Alginate - Natural Polymer

Sodium Alginates (polysaccharide) are salts of alginic acid which is a natural polymer

abundant in nature. The polysaccharide is commonly extracted from marine brown algae of

class Phaeophyceae, providing a structural component of the plant and is also synthesised

by the soil bacteria and more environmentally common Pseudomonas aeruginosa cellular

polyose in soil bacteria [22-23]. Commercially available alginate is extracted from a

particular brown seaweed of genus of Macrocystis, and Laminaria (Figure 3.1).


TIMEA GREGO 7

FIGURE 3.1. LAMINARIA GENERA [24]

In recent decades, the demand for alginate production has increased rapidly. The medical,

pharmaceutical, food and industries utilise the functional properties of alginate, such as

solubility in water, gelling, thickener and stabilising effects (Table 3.1). Most relevantly,

biomedical and pharmaceutical industries often use alginate as scaffolds in tissue

engineered medical products [10] and for cell implantation [25], using medical grade

alginate, which is a highly purified form.

TABLE 3.1 APPLICATIONS OF ALGINATES BY ITS CHARACTERISTICS AND FUNCTIONALITY [26]

CHARACTERISTICS FUNCTIONALITY BENEFITS APPLICATIONS

MOLECULAR WEIGHT Viscosity Thickening, film-

forming

Suspensions, coatings

COMPOSITION AND

SEQUENCE

Cross-linking Gelling Immobilization,

encapsulation

DISSOCIATION, PKA Soluble at pH,

precipitate at pH,

swelling capability

Solutions, fibers, films,

adsorption

Solutions, pastes,

scaffolds, dressings,

membranes, tablet

disintegration

POLYANION Cation affinity Chelation Gelation, drug, metal

binding


TIMEA GREGO 8

3.2 Structure and Properties of Alginate

Alginates are non-branched binary copolymers consisting of two residues namely, β-D-

mannuronic acid (M) and α-L- guluronic acid (G), linked by a (1→4) glycoside bonds [10,

26, 27]. During the extraction of the carbonhydrate polymers, the mannuronic and

guluronic acids are converted into mannuronate (M) and guluronate (G) salts (Figure 3.2).

It has been reported that the structural differences of the two uronic acids monomers are

minor thus they form different chair conformations [27].

Physical properties of the hydrogel depend on the block structure and composition of

alginate. Determination of diad (FGG, FGM, FMG and FMM), triad (FGGG, FGGM, FMGG, FMGM,

FMMM, FMMG, FGMM, and FGMG) and higher order frequencies of the monomer composition

by NMR spectroscopy allows the calculation of average block length [28].

The G- content, the length of the G-block and molecular weight determine the gelling

properties of alginate. Table 3.2 shows the composition and block structure in different

sources of alginate. Na-alginate is used as an encapsulation medium for HepG2 cells,

which seeks to provide the cell component for the BAL. The main advantage of the

polymer is to provide supporting environment for the encapsulated cells therefore,

promoting growth and function of HepG2 cells compared with monolayer.


TIMEA GREGO 9

FIGURE 3.2 MONOMER SALTS OF ALGINATE [29]

Alginate is a block copolymer, hence monomer rings are distributed within the polymer in

diverse sequence and can be described by three block structures. These are homo-

polymeric consisting of mannuronate (M)- and guluronate (G)-rich alginate, and hetero-

polymeric with different sequences of M and G (MG-blocks) [30, 23, 27].

3.3 Characteristic and properties of Alginate

3.3.1 Molecular weight

Alginates are polydisperse systems as a result of a broad range of composition, sequential

structure and consist of molecules of different molecular weight (MW). Due to this

property, the average molecular weight of alginate describes the average number of

molecule species in the whole distribution of the polymer [26, 22].


TIMEA GREGO 10

Two most common ways to express the average molecular weight of a polymer is the

number average

___

nM and weight average

___

wM methods [26-27].

Alginate polymers are anionic (polymer is negatively charged) polyelectrolytes at pH 7

[31-32]. The molecular characterisations of polyelectrolytes are based on the Mark-

Houwink-Sakurada equation (intrinsic viscosity), membrane osmometry, sedimentation

equilibrium, gel permeation chromatography (GPC) and light-scattering measurement

[33, 26].

3.3.2 Solubility

Sodium-alginate is soluble in cold water while alginic acid and Ca-alginate are insoluble

[34]. This is important for encapsulation as it allows formation of beads during the

polymerisation reaction. Solubility of the polysaccharide depends on the pH, ion strength

and presence of divalent cations in the solute. pH determines the electrostatic charges on

the two uronic sugars [22, 30].

3.3.3 Gel formation

The primary functional property of the carbohydrate polymer is the ability to form gels in

the presence of multivalent or divalent cations such as Mg2+

< Ca2+

< Sr2+

< Br2+

[35].

Gelling properties of alginate depend on the ratio of the two uronic acids residual (M/G).

High G content and molecular weight will produce a strong less flexible gel with heat

resistant capability but prone to syneresis on freeze-thaw, whereas M rich alginates are

weaker but more flexible and can form gels either at very high or low Ca2+

concentration

[35]. Table 3.2 gives more information about the different M, G and MG fraction

composition of algal alginates. M/G ratio, molecular weight and viscosity influence the

physicochemical properties of alginate.


TIMEA GREGO 11

TABLE 3.2 M, G AND MG FRACTION COMPOSITION OF COMMERCIAL AVAILABLE ALGAL ALGINATES [22]

SOURCE FG FM FMG

Laminaria japonica 0.35 0.65 0.17

Laminaria digitata 0.41 0.59 0.16

Laminaria

hyperborean

Leaf 0.55 0.45 0.17

Stipe 0.68 0.32 0.12

Outer cortex 0.75 0.25 0.09

Macrocystis pyrifera 0.39 0.61 0.23

Durviella Antarctica 0.29 0.71 0.14

Ascophyllum

nodosum

Fruiting body 0.10 0.90 0.06

Old tissue 0.36 0.64 0.20

Other valuable properties of alginates are swellability and film-forming, which are

important to biomedical applications [26]. This may cause the volume of the beads to

increase over time as water moves into the bead causing it to swell. Whether this has an

effect on function or growth is yet to be established.

3.3.4 Viscosity

In general, alginate solutions are pseudoplastic systems or non-Newtonian fluids.

Pseoudoplasticity refers to the decrease in viscosity of a solution with the increase of shear

rate (Figure 3.3). Temperature (Figure 3.4), molecular weight, concentration of the solution

(Figure 3.5) and pH may influence the viscosity. Autoclaving of alginate solution at 121°C

for 10 or 20mins can change the viscosity of the solution and probably causing changes in

the M/G chain length of the polymer and providing a less stable material for cell

encapsulation [22]. Lyophilisation (freeze drying) of alginate solution at the freezing

temperature can also effect the physical properties of polymer. During freeze-drying

operation, both “freezing and drying” stresses are produced which could alter the viscosity

of the polymer [36].


TIMEA GREGO 12

FIGURE 3.3 VISCOSITY OF DIFFERENT ALGINATE CONCENTRATIONS VERSUS SHEAR RATE.

VISCOSITY OF THE SOLUTION DECREASES WITH INCREASING SHEAR RATE [30].

FIGURE 3.4 TEMPERATURE PROFILE OF THREE DIFFERENT 1% ALGINATE

SOLUTION VISCOSITY [30]


TIMEA GREGO 13

FIGURE 3.5 SOLUTION CONCENTRATION EFFECTS ON THE POLYMER VISCOSITY [30]

The higher the concentration of the solution is, the greater its viscosity. During

encapsulation, cell media fluid mixtures pass through a 200µm nozzle thus surface force

increases. Higher concentrations of solution could results in non-spherical beads, which

alters the biocompatibility of the polymer providing a less suitable matrix for cells

development (Figure 3.5).

3.4 Purification Methods of Alginate for Clinical Use

Biomedical industry such as tissue engineering and cell transplantation is continuously

seeking biocompatible and non-toxic materials for cell encapsulation to be used in

bioartificial organs. Alginate hydrogels have been used as immobilisation matrix for

human and animal cells. Therefore, purification level of the biopolymer is crucial. During

harvesting and extraction of the biopolymer, it can be contaminated. Presence of various

impurities may initiate an immune response. Therefore, different purification methods have

been established but these unfortunately often resulting in changes to the viscosity of the

solution. As explained, successful encapsulation and maintenance of cell function and

growth through bead shape or morphology is affected by solution viscosity [2]. For this

reason a method is required that removes the presence of contaminants for regulatory

approval, without altering the chemical composition or viscosity.


TIMEA GREGO 14

This section is concerned with the published insitu (clinical based) alginate purification

methods that aim to make the polymer suitable for human implantation. These processes

have significantly reduced endotoxins, proteins, polyphenols, ash and heavy metals from

alginates [26]. However, the methods were not primarily concerned with micron

particulate removal which affects the biocompatibility of the polysaccharide, as their

removal would have occurred along side removal of these other compounds. However, the

dramatic effects on the alginate composition and subsequent properties from these

processes made them unsuitable for Bioartificial Liver application. Therefore, this section

gives a brief overview of the different purification methods and their efficiency.

Firstly, it has become clear, after conducting a broad ranging literature review, that

currently there is no recognised standard purification method available for large scale

contamination free alginate production. Secondly, there are no medical standards that

would limit the concentration of impurities in the polymer. However, ATSM has published

„Standard and Guidelines for Biopolymers in Tissue-Engineered Medical Products; ASTM

Alginate and Chitosan Standard Guide‟. It states the major impurities of concern such as

endotoxin, protein, heavy metals and microbiological burden but it is limited only the

endotoxin level below 5 EU/kg of body weight [10] most of these components will be

removed insitu as part of the BAL quality assurance system, which is an inline filtration

system.

The first purification method was published in 1992 using free-flow electrophoresis (FFE).

FFE treatment contains two purification steps, a combined physicochemical pre-treatment

of alginate and a chemical purification procedure [19]. The FFE treatment is a complex,

time consuming, and expensive treatment. FFE method for purifying alginate endotoxin

and mitogenic contaminants level was successful, although the authors had pointed out that

“some unidentified residues were exhibited in the purified alginate after chemical

extraction” [19].

An alternative sodium alginate purification method, hollow fiber dead-end ultrafiltration,

was used by van de Ven, W. J. C. et al. [14] Membrane processes such as micro, nano and

ultra filtrations are utilised in drinking water treatment to remove visible materials and

substances with different particulate sizes [14].


TIMEA GREGO 15

In 1997, De Vos, P. et al [18] published a report on „Improved biocompatibility but limited

graft survival after purification of alginate for microencapsulation of pancreatic islets‟. The

study focused on the correlation between the alginate purification methods and its impact

on the biocompatibility of the polymer microcapsules [18]. A multi – step (filtration,

extraction and precipitation) process was installed to purify raw sodium alginate, which

was filtered over four different pore sized filters (5.0, 1.2, 0.8 and 0.45µm). For removing

non-precipitated purities, Na-alginate solution was filtered over Buchner funnel three

times, and followed by a chloroform-butanol extraction step for the removal of the protein

content [18]. The final step was ethanol precipitation, which follows a Buchner funnel

filtration and at the end of the process, alginate is preserved by lyophilisation.

Biocompatibility of alginate capsules were improved after the multi-step purification but

graft abnormality and limited life-cycle was observed [18]. More details about the recently

published purification methods can be found in Appendix A.


TIMEA GREGO 11

3.5 Fine Slow Sand Filtration Process

In drinking water treatment sand filtration (biological treatment), which is one of the oldest

water purification methods, is the fourth unit operation process in a water treatment plant.

Prior to this coagulation, flocculation and sedimentation remove large suspended

particulates and materials. Further removal of the particulate impurities is done by

percolating water through a permeable media (sand or anthracite) to remove fine

suspended particulates.

During the removal mechanism, suspended organic and inorganic materials are captured

within the pore spaces of the filter media, thereby removed from the water [37]. Numerous

factors affect the efficient operation of the filter media such as [38]:

Filter depth,

Grain size and sphericity (shape) of the filter,

Composition of the filter bed,

Filtration rate,

Influent concentration,

Characteristics of the suspended materials,

Water temperature, and

Porosity of the filter bed.

Fine sand filtration was chosen as an alternative purification method of Na-alginate

solution because the system is efficient in the removal of fine particles at low flow rate,

cost effective and easy to operate in comparison to the previous in house purification

methods (Section 3.4.).

In this chapter, the theory of sand filtration and flow of Newtonian and non-Newtonian

fluids is discussed.


TIMEA GREGO 12

3.5.1 Physical parameters affect the operation of filter media

3.5.1.1 Shape of the filter media

Sphericity is defined as “the ratio of the equivalent hydraulic size (dh) to the sieve size

(geometric mean of passing and retaining sieve openings) (ds)”. It is a dimensionless value

and can be calculated by Eq. (3.1). Table 3.3. shows the calculated sphericity of filter

media by different researchers.

s

h

d

d

(3.1)

TABLE 3.3 SPHERICITY OF SAND

GRANULAR

MATERIAL

SPHERICITY RESEARCHERS

Sand 0.85 Ives, K. J.[39]

Sand 0.75-0.85 Tchobanoglous, G.

et.al [40]

3.5.1.2 Porosity of the bed

Porosity (n%) is defined as the ratio of the volume of voids to the total volume and

calculated from Eq. (3.2). The following parameters are required to calculate the porosity

of the granular material such as dry and wet weight of the media, depth of the media in the

filter column, and density of the media. As it can be seen from Table 3.4, several

researchers had determined the porosity of sand.

V

Vn v%

(3.2)

TABLE 3.4 POROSITY OF SAND

GRANULAR

MATERIAL

POROSITY

(n)

GRAIN

SHAPE RESEARCHERS

Sand 0.43 n/a Ives, K. J.[39]

Sand and grit 0.40-0.45 n/a Kézdi, Á.[41]

Sand 0.40-0.46 n/a Tchobanoglous, G.

et.al [40]


TIMEA GREGO 13

3.5.2 Mechanism of filtration

The removal mechanism of fine suspended flocs from water by filtration through a

granular media is a complex process and can be classified as transport, attachment and

detachment mechanisms. Filtration can be further classified as surface (cake)- or depth

filtration but in all cases of filtration, laminar flow regime predominates [42].

Darcy‟s Law defines laminar flow of a fluid through permeable media (Figure 3.6).

FIGURE 3.6 HYDRAULICS OF A FLUID THROUGH PERMEABLE MEDIA [44, 51]

General equation Eq. (3.3) of his experiment (laminar flow condition) [45]

s

m

L

hAKQ

3

(3.3)

Where

Q flow rate [m3/s]

A cross section area of the column [m2]

Δh head loss of the column [m]

L length of the column [m]

K coefficient of permeability

L

hI

hydraulic gradient [dimensionless]


TIMEA GREGO 14

“Permeability depends on the viscosity and the density of water, porosity, the size of the

pores, and the capillary forces between the medium and the liquid” [44].

Darcy’s Law

IkdL

dHk

A

Qqv ppa

(3.4)

Where

va approach- or filtration velocity [m/s]

kp permeability coefficient [m/s] [46]

I. The transport mechanisms during filtration

Straining or screening process takes place when the suspended particulates

percolating through a pore greater than the pore opening. This mechanism occurs

on the surface of the filter and is “independent of the filtration rate” [47].

Sedimentation: if the density of the particulates is greater than that of water they

will attached to the grain surface by gravitational force. Sedimentation is not

effective at removing particulates smaller than 1µm [47].

According to Camps, T. R. [48], settling behaviour of particles in suspension can be

classified as „free settling‟, „flocculent settling‟, „hindered (zone) settling‟ and

„compression settling‟ [48]. In accordance with the fine sand filtration experiment, it was

assumed that the settling is discrete and there is no interface among the sand particles [48].

During “the settling of discrete particles”, there are three flow regimes in a settling basin

e.g. laminar (10-4

<Re<1)-, transition (1<Re<104)- and turbulent flow (10

4<Re<10

6) (Figure

3.7) [48].


TIMEA GREGO 15

FIGURE 3.7 DRAG COEFFICIENT (CD) FOR SPHERES AS A FUNCTION OF RE OVER LAMINAR-,

TRANSITION- AND TURBULENT FLOW REGIMES. IT CAN BE SEEN FROM THE FIGURE, THAT RE

INCREASES WITH DECREASING CD [48].

However, the value of Reynolds‟ number for permeable material under laminar flow is

Re<5 and can be calculated using Eq. (3.5).

tS

e

vdR

(3.5)

Where

Re Reynolds number [dimensionless]

µ dynamic viscosity of fluid [kg/ms]

ρ density of fluid [kg/m3]

vt terminal velocity [m/s]

dS sieve diameter [m]

During settling, three types of forces (gravitational force, buoyant force, and drag force)

affect the motion of a particle and drag occurs when water contacts with the surface of the

particle (Figure 3.8) [49].


TIMEA GREGO 16

FIGURE 3.8 GRAVITATION FORCE (FG), BUOYANT FORCE (FB) AND DRAG FORCE (FD) AFFECT

THE MOTION OF A PARTICLE IN WATER DURING SEDIMENTATION [50]

Gravitation force is defined by Eq. (3.6)

gmFG ][N (3.6)

Where

g gravitational constant [m/s]

m mass of the particle [kg]

The buoyant force is given by Eq. (3.7)

S

OHB

gmF

2

(3.7)

Where

OH2 Density of water

3m

kg

S Density of particle

3m

kg

FD

FG

FB

PARTICLE


TIMEA GREGO 17

The drag force is given by Newton‟s Law

2

2

2 vACF OHD

D

][N

(3.8)

Where

FD Drag force on particle [N]

CD Drag coefficient [dimensionless]

A Cross section area of the particle [m2]

vs settling or terminal velocity of the particle [m/s]

General equation for settling velocity is defined by the following Eq. (3.9)

]/[

3

4

2

2 smC

dgv

OHD

OHss

(3.9)

Figure 3.7 shows the three different flow regimes such as laminar, transition and turbulent.

Value of drag coefficient depends on the flow regime thus Reynolds‟s number.

Drag coefficient and settling velocity for Laminar flow 10-4

<Re<1 but for permeable

materials Re<5

e

DR

C24

(3.10)

Settling velocity for laminar fluid flow is defined by Stockes’ Law.

]/[

18

2

smdg

v ss

(3.11)

Where

g gravitational force [m/s2]

ρH2O density of the water [kg/m3]

ρs density of solid [kg/m3]

µH2O viscosity of the water [kg/ms]


TIMEA GREGO 18

Camp‟s method for Transitional flow 1<Re<104 [49]

Drag coefficient form Eq. (3.9)

2

2

23

4

tOH

OHSD

v

dgC

(3.12)

Camp rearranged Eq. (3.12) to define

32

2

2

2

3

4

tOH

OHOHS

e v

g

R

Cd

(3.13)

and

OH

OHOHS

eD

dgRC

2

2

2

2

3

2

3

)(4

(3.14)

Camp‟s curves of eR

Cd and

2

eDRC versus Re can be found in Appendix B.

Inertial force: It can be seen from Figure 3.9 as flow lines approach the surface of

the filter grain the particle is forced to move out of the streamline [42]. “If the

particles have adequate inertia they maintain a trajectory” [42] therefore, they

deposit on the surface of the grain.

Diffusion (Brownian motion): natural source of water contains dissolved and

suspended solids. Suspended solids in water can be characterised as organic and

inorganic materials, bacteria, virus and other particulates. Particles with a particle

size greater than 1µm such as large colloids and soil can be removed during

filtration without additional chemical of physical treatment. Suspended particles

with a particle diameter less than 1µm are stable in the suspension and defined as

colloids.

Colloidal particulates can be further classified according to their affinity for water

such as hydrophobic and hydrophilic [37].


TIMEA GREGO 19

FIGURE 3.9 TRANSPORT MECHANISMS DURING FILTRATION

THE MECHANISMS ARE: INTERCEPTION, INERTIA, GRAVITY, DIFFUSION, AND HYDRODYNAMIC

BY WHICH GRANULAR FILTER MEDIA REMOVE IMPURITIES FROM WATER [42]


TIMEA GREGO 20

Three physical forces control the motion of colloidal particulates in water; namely,

electrostatic forces, van der Waals forces, and Brownian motion which is a random

movement of colloidal suspension [47].

Hydrodynamic action can be described in two ways. Firstly, when the shear field is

homogenous with a constant velocity gradient across the flow lines, spherical

particles will experience centrifugal force. As a consequence of this particles will

move across the shear field. Secondly, when the shear field is heterogeneous, as

on the filter pores, the particle will be diverted by the same force [32, 51].

II. Attachment mechanisms

1. Electrical double layer interaction: surface of a filter media, colloidal particles

and impurities are negatively (anionic) charged in water thus these materials

attract cationic ions, also known as “counter - ions” [37].

A stern layer around the electronegative particle is formed by counter ions such as

hydrogen. The second outer layer is the diffused layer where distribution of

cations and anions are presented but counter ions predominate. Combination of

the two outer layers is called the double layer [Figure 3.10]. As was mentioned

above, filter grain and impurities are negatively charged in water, thus the double

layer interaction will cause repulsion [37].

ELECTROSTATIC FIELD AROUND AN ANIONIC

PARTICLE

ILLUSTRATION OF ZETA POTENTIAL, SHEAR

PLANE, AND ELECTROSTATIC POTENTIAL

FIGURE 3.10 A, B ELECTRICAL DOUBLE LAYER [52]


TIMEA GREGO 21

2. Van der Waals forces: the attraction or repulsion between atoms and molecules.

The magnitude of the force depends on the mass and the distance of the atoms and

molecules [52].

III. Detachment mechanisms

During the filtration cycle, suspended particulates are entrapped in the filter bed causing

reduction in the porosity. As a result of this, the head loss, filtration velocity, and the shear

force of the influent water thorough the porous media will be increased. At this point, the

filter bed becomes clogged and there will be a reduction in the filtration rate [37]. To

increase the filter performance, the porous media needs to be backwashed. During the

backwashing process, the colloidal particulates, which were entrapped among the grains,

become detached from the grains‟ surface [51].

3.5.3 Types of granular media filters

Granular media filters utilized in water treatment can be classified according to the filter

media type, filtration rate, driving force, and direction of flow [48]. In theory, sand filters

remove micro- and macro substances, and materials within the size spectrum of 1-1000µm

(Figure 3.11).


TIMEA GREGO 22

FIGURE 3.11 DIFFERENT SEPARATION PROCESSES IN ACCORDANCE WITH THE SIZE DISTRIBUTION

OF POSSIBLE IMPURITIES PRESENT IN WATER [53]

GRANULAR MEDIA FILTRATION USED FOR REMOVING PARTICULATES OF SIZE RANGE BETWEEN

1.0µm-1000µm

Classification by filter media: There are three different filter media for the successful

operation of the filtration process such as single-, dual–, and multi-media (Figure 3.12).

Single medium filter consists of a layer of sand or anthracite whereas, dual medium filter

utilizes sand and crushed anthracite. The third filter type is the multi-media, which is a

triple layer of sand, crushed anthracite and garnet. The design parameters of different

media are shown in Table 3.5. A single medium filter removes macro particulates in the

range of 40-1000µm. Effective removal of micro particulates from water requires a dual-

or multi-media filter bed installation [54]. In theory, dual media filter (sand and anthracite)

are used in rapid filtration for micron particulate removal. For the filtration experiment fine

and coarse sand were used as dual media filter.


TIMEA GREGO 23

FIGURE 3.12 DIFFERENT SAND FILTER MEDIA SYSTEMS

A), SINGLE MEDIUM FILTER (B) DUAL MEDIA FILTER, (C) MULTI-MEDIA FILTER [37]

Classification by filtration rate: filters can be slow (10 m3/m

2d)- or rapid (120 m

3/m

2d)

sand filters.

Classification by driving force: gravity or pressure filters.

Classification by direction of flow: downward flow or upward flow filters [37].

TABLE 3.5 DESIGN PARAMETERS OF SINGLE-MEDIUM, DUAL-, AND MULTI-MEDIA FILTERS [37]


TIMEA GREGO 24

3.5.4 Hydraulics of sand filter

3.5.4.1 Head loss in clean filter

Throughout the operation of a clean filter under slow flow rate, suspended solids firstly

deposit in the top layers of the filter bed. Secondly, as the filtration time increases, the

suspended particles move deeper into the filter bed. As a result of this, head loss across the

filter will increase.

The Kozeny-Carman equation Eq. (3.15) describes the flow of clean water through a

permeable media when Re<10 [46]

2

3

26)1(

2

2

sOH

OHapproach

dg

vK

dL

dh

(3.15)

Where

Ψ Sphericity [dimensionless]

ε=n Porosity [%]

vapproach [m/s] can be calculated from the relation of volumetric low rate to cross

section area [46]

ds Diameter of the spheres [m]

K Permeability is a constant and range 4-5

OH2 viscosity of water at given temperature

sPa

sm

kg

OH2 density of water at given temperature

3m

kg

3.5.4.2 Filter backwashing

During continuous operation of uniform or stratified filter beds, suspended particles will be

entrapped among the pore spaces thus, reduction in porosity and increase in headloss

occurs. A clogged filter media state is formed and the system requires backwashing.

During backwashing, the upward fluid velocity expands the filter bed and causes

fluidisation.


TIMEA GREGO 25

Hydraulics of expanded beds [46]

Determination of the expanded bed porosity using Fair – Hatch equation

22.0

t

ev

v

(3.16)

Where

εe porosity of expanded bed [%]

vt settling velocity of the particulates [m/s]

v approach velocity [m/s]

Depth of expanded bed

][

1

)1(22.0

m

v

v

LL

t

e

(3.17)

Where

L depth of filter bed [m]

Le depth of expanded bed [m]

Determination of minimum fluidisation velocity

e

sesmf

dgv

1180

)( 23

[m/s2]

(3.18)

Where

εe Porosity of the expanded filter bed


TIMEA GREGO 26

3.5.5 Time-independent Non-Newtonian Fluids

Alginates have many industrial applications due to their chemical and physical properties

such as thickening and gelling. Notably, alginate solutions are time-independent,

pseudoplastic systems (non-Newtonian fluid) thus the viscosity of the solution depends on

concentration of the polymer, molecular weight, pressure, temperature, and rate of shear.

As stated, viscosity of such fluids depends on the concentration therefore, alginate

solutions can be further categorised as Newtonian or non-Newtonian fluids. In a published

article „Alginate-A Polysaccharide of Industrial Interest and Diverse Biological Functions‟,

Sabra, W et. al. [55] stated in accordance with other researchers, that rheologically a

medium viscosity, 0.5 % sodium alginate solution considered to be Newtonian fluid under

low (0.1-10 s-1

) and medium shear (10-1,000 s-1

) rates and time-independent under high

shear rate (1,000-100,000 s-1

). Whereas, rheologically a medium viscosity, 2.5 % aqueous

sodium alginate hydrocolloid is a non-Newtonian fluid at the end spectrum of low shear

rate range and high shear rate [55]. For cell encapsulation of living cells, viscosity of the

insoluble alginate gel is crucial because it affects the bead shape and resultant

microenvironment for cell growth and survival.

It is important to mention that non-Newtonian fluids can be subdivided into three groups

such as time-independent, time-dependent and viscoelastic fluids. In the next section

distinction between Newtonian and non-Newtonian fluids is presented.


TIMEA GREGO 27

3.5.5.1 Viscosity of Newtonian fluids

Fluids behave differently under the application of external force (shear stress) and such

force causes continuous deformation, which can be defined as consistency. Consistency of

low molecular weightv liquids and gases at steady pressure and temperature is constant and

it is called viscosity. Viscosity of Newtonian fluids is independent of the applied rate of

shear and can be explained by the following Eq. (3.19) [56].

FIGURE 3.13 VISCOSITY OF NEWTONIAN FLUIDS [57]

][ sPadr

dv

(3.19)

Where

τ shear stress [Pa]

γ shear rate [s-1

]

r thickness of the fluid layer [m]

Correlation between shear stress and shear rate can be presented by flow curve, which is

linear for Newtonian fluid (Figure 3.14).

X

Y

dr

dv

Shearing force


TIMEA GREGO 28

FIGURE 3.14 FLOW CURVE OF NEWTONIAN FLUIDS [57]

3.5.5.2 Viscosity of non-Newtonian fluids [57]

Solutions containing polymers of high molecular weighted are classified as non-Newtonian

fluids thus their “flow curve is non-linear” [57]. Four factors affect the viscosity of a non-

Newtonian (real) solution such as concentration, molecular weight, temperature, pressure

and shear rate. Further categorisation of real fluids:

“Time-independent non-Newtonian fluids” [57] (Bingham plastic, pseudoplastic and

dilatant fluids) (Figure 3.15)

FIGURE 3.15 FLOW CURVES OF VARIOUS TIME-INDEPENDENT NON-NEWTONIAN FLUIDS [57]

τ

γ

η=τ/γ

τ

γ

τB

Bingham (Plastic)

Dilatant

Newtonian

Pseudoplastic


TIMEA GREGO 29

a) Bingham plastic

As can be seen from Figure 3.15, flow curve of a Bingham plastic fluid is a

straight line and τB is given by the intersection on the shear stress axis [57].

Rheological properties of a Bingham plastic fluid is described by the

following Eq. (3.20)

][PaB (3.20)

Where


τB Yield stress [Pa]

γ shear rate [1/s]

b) Pseudoplastic fluids: viscosity of these liquids decrease with increasing

shear rate. Flow curve of these fluids is slightly convex (Figure 3.15).

Ostwald (power law) is used for describing the rheological properties of the

fluid [57].

][)( Pak n (3.21)

Where


γ shear rate [s-1

]

k consistency index, [55]

n flow behaviour index, for pseudoplastic fluids between 1 to 0 [55]

Important to note, “k depends on temperature and measures the consistency of the fluid.

The higher “k” the more viscous the fluid”, n measures the degree of non-Newtonian

behaviour [57].

c) Dilatant fluid: viscosity of the solution increases by increasing shear rate.

A 0.2%, alginate solution may be considered to be a Newtonian liquid as the low

concentration produces more water like properties [55], therefore further analysis will be

carried out to determine the Newtonian or non-Newtonian fluid dynamics behaviour of Na-

alginate solution having different concentration.


TIMEA GREGO 30

4 MATERIAL AND METHODS

The major objective of the present work was to investigate the use of fine sand filtration as

an alternative purification method for particulate removal from Na-alginate. This section

includes:

purification of Na-alginate by sand filtration, a method for large scale

production,

in vitro analytical methods analysis of purity of alginate,

description of the experimental work,

equipments and materials are presented (Figure 4.1),

fine sand filtration experiments.

The alginate used was research grade, medium viscosity sodium alginate (Sigma-Aldrich

UK) with the following specification: viscosity of 2% at 25°C approximately 3500cps and

molecular weight of alginic acid sodium salt range between 80,000-120,000. Granular

media (RH-70 and RH-110) were supplied by Sibelco UK Ltd.

FIGURE 4.1 LIFE CYCLE OF THE EXPERIMENTAL WORK

The experimental work was divided into four stages:

Stage I. Fine sand filtration with the aim of removing macro particulates from research

grade Na-alginate;

Stage II. Lyophilisation 0.2% filtered alginate solution is shell frozen in a dry ice acetone

bath in 1L round bottomed flasks, to be lyophilised to sublimate water, leaving dry alginic

acid salt.


TIMEA GREGO 31

Stage III. Reconstitution of alginate to final concentration of 1.0%, 1.25%, 1.5%, 1.75%,

1.875% and 2%, and autoclaving for 10 mins at 121°C.

Stage IV. HepG2 cell encapsulation preparation of 1.875% alginate beads containing a

liver cell line using Inotech microencapsulator.

4.1 Preparation of 2% HEPES buffered alginate solution

To make 2% sodium alginate solution, 8.76 g NaCl was dissolved in Milli-Q water under

constant stirring, than 15 ml 1M HEPES (Sigma-Aldrich UK), bringing to pH7.4 using

NaOH. The saline solution was placed in a foil covered bottle. Na-alginate powder (20g),

was gradually added to the saline solution allowing to dissolve and stirred overnight by a

mechanical stirrer [58]. For the filtration test, 2% aqueous Na-alginate solution was diluted

to 1:10 in Milli-Q water and filtered through sand.

Milli-Q water is produced by Millipore; Milli-Q(R)

Integral water purification system. The

four cartridges remove microorganism, ionic and organic contaminants and subsequently

the water is filtered through Millipak Express 0.22 µm membrane filter thus, the system

provides fine suspended material and bacterial free ultra pure water for laboratory use

(Table 4.1) [59].

TABLE 4.1 WATER QUALITY PARAMETERS OF MILLI-Q WATER [59]


TIMEA GREGO 32

4.2 Testing of filter media

Filtration media tests were conducted to achieve purification of sodium alginate solution.

Two different granular materials were used as filter media (RH-70 and RH-110).

4.2.1 Chemical analysis of filter media

The aim of the analysis was to determine the chemical composition of the granular filter

media, which are 99% silica sand. The examination was done by SIBELCO UK Ltd

(Appendix C).

4.2.2 Physical analysis of filter media

4.2.2.1 Grain size distribution analysis

The aim of the sieve analysis is to assess the particle distribution of sediments, to

determine the particle sizes and to evaluate the cumulative percent passing of the granular

material. Grain size distribution was done by sieve analysis. Results of the analysis are

presented in the particle distribution graph (Appendix D), which is a semilogarithmic

graph, where the X axis (logarithmic scaled) shows particle size and the Y axis (linear

scaled) % passing of the sediment. Another important parameter, the uniformity coefficient,

that is a ratio between 60%- and 10% of granular materials passing Eq. (4.1), is also

determined during the analysis.

Uniformity coefficient

10

60

d

dU

(4.1)

Where

d10 also know as effective grain size (deff)


TIMEA GREGO 33

4.2.2.2 Density of sands

The density of granular material is required for the calculation of settling velocity of the

particulates, and determination of headloss through the filter bed. In theory, the test is

carried out in a density bottle [39]. Density of filter medium was evaluated by SIBELCO

UK Ltd. hence density of RH-70 was 2,650 kg/m3 and assumption was made for the

density of RH-110. Assumption analysis is described in more details in Section 5.1.1.2.

4.2.3 Settling velocity of individual grains

There are two main reasons to determine the “settling velocity of individual grains of

uniform size in water of known temperature” [39]. Firstly, during backwashing in a “water

alone” system, it is establishes the “stratification behaviour” of the grains. Secondly, the

hydraulic diameter can be calculated and later the sphericity of the grains can be

determined [39]. In theory, settling velocity is determined based on the average time

required to settle 20 individual wet sediment grains (sand or anthracite) in a 1 m column at

a known water temperature.

4.2.4 Shape (sphericity) of filter media

The settling velocity can be used to determine the shape of the filter grains using the Eq.

(4.2) [39]. Sphericity of filter media was determined by SIBELCO Ltd thus grain shapes

are sub-angular and sphericity medium (Appendix D).

s

h

d

d

(4.2)

Where

Ψ sphericity [dimensionless],

Ψ=1 the grain is sphere,

Ψ<1 the grain is non-sphere [39].


TIMEA GREGO 34

4.2.5 Porosity of filter media

Porosity of the RH-110 filter bed was determined using Eq. (3.2) and an assumption was

made for RH-70 based on literature. Dry and wet weight of the medium and depth of the

medium in the filter column were measured, and density of the sands was calculated.

4.2.6 Determination of headloss in clean granular media filter

4.2.6.1 Flow of non-Newtonian alginate fluid in porous media

For the filtration experiment a 2% sodium-alginate solution was prepared and was diluted

to 1:10 ratio with Milli-Q water. As a result of this, the following assumptions were made

and based on Sabra, W. et al. [55] and Cancela, M. A. et al. [60]:

a 0.2%, medium viscosity Na-alginate solution is considered as a Newtonian fluid

between low and medium shear rate ranges,

the viscosity of the fluid depends on temperature and shear rate,

flow regime is laminar,

flow of the fluid through porous media is described by Darcy‟s Law.

Headloss development in the permeable media, flow rate, and filtration rate calculations

are based on the above stated assumptions and the results are presented in Chapter 5.


TIMEA GREGO 35

4.3 Systems preparation for filtration

4.3.1 Treatment of filter media before use

Acid treatment was used to eliminate the micronutrient impurities from the sand. The sands

used in the experiment were washed with 20% hydrochloric acid (HCl) and left overnight.

HCl was supplied by Fisher Scientific UK Ltd with the following specification: analytical

reagent grade 32% HCl, Code H/1100/PB17, Batch 0929448. After acid bath, the granular

materials were mechanically stirred and rinsed in Milli-Q water several times, gradually

increasing the pH to 7.

4.3.2 Treatment of the filtration systems before use

Glass tubes were autoclaved for 15 min at 121°C, and due to the construction material

(PVC) of the large scale system, the filter tube was washed by a 70% Methanol/H2O.

4.4 Filtration systems installation

Particular objectives of the research project were to determine the most effective granular

media filter and flow rate for micron particulate removal. Therefore, filtration tests (single

medium and dual media filters) were run and the process operation was subdivided into

five phases. Technical parameters of the small scale systems were the same in Phase I-III.

Phase I. Aim was to determine the most efficient granular filter medium size.

Technical parameters of the small scale systems:

• Two glass tube were used for the test (Figure 4.2),

• d(external)=0.018m of the tube,

• d(internal)=0.0143m of the tube,

• Height (H) = 0.40m of the tube.


TIMEA GREGO 36

FIGURE 4.2 PHASE I. FILTRATION: SMALL SCALE SYSTEM BEFORE THE FILTER MEDIA

INSTALLATION

Parameters of the small scale filtration systems:

Uniform filter medium RH-70 and RH-110,

Filter media depth in both columns 0.25m,

Variable flow rates were tested for the efficient operation of the filtration system.

Phase II. The aim of this phase was to establish the most suitable flow rate according to

the removal efficiency. Technical parameters of the small scale system and the filtration

system are the same as in Phase I.

Phase III. The aim of this phase was to specify the most efficient filter medium depth.

Parameter of the small scale filtration system:

Uniform filter bed of RH-70 and RH-110,

Filter media depth in both columns 0.30m,

Variable flow rates.

Sand

RH-110

Sand

RH-70


TIMEA GREGO 37

Phase IV. For the more efficient removal of macro particulates a dual-media filter was

introduced at the small scale system. A large scale (Figure 4.3) single media filter was also

set up with the aim of comparing the filtration results such as flow rate and particulate

removal efficiency to the performance of small scale systems.

Technical parameters of the small scale systems were the same as in Phase I-III.

Technical parameters of the large scale system (Appendix E):

• PVC built filter tube,

• d(external)=0.06m of the tube,

• d(internal)=0.0465m of the tube,

• Height (H) = 1.25m of the tube.

Parameters of the large scale filtration system:

• Mono (single) medium filter of RH-110,

• Filter medium depth 0.6m.

Parameters of the small scale filtration system (Figure 4.4):

• Dual media filter combination of RH-70 and RH-110,

• Total depth of the dual media filter was 0.30m (RH-70=0.13m and RH-110=0.17m)


TIMEA GREGO 38

FIGURE 4.3 PHASE IV. FILTRATION: LARGE SCALE, SINGLE MEDIA FILTRATION SYSTEM


TIMEA GREGO 39

FIGURE 4.4 PHASE IV. FILTRATION: DUAL MEDIA FILTRATION SYSTEM

Phase V. Due to an unforeseen technical problem, which contaminated the effluent during

the single media filter operation, both systems in Phase IV. were shut down on 9th

July

2009. Contamination caused turbidity in the effluent and gave a „yellowish‟ colour for the

filtered 0.2% sodium-alginate solution. Thus, Phase V. is repetition of Phase IV.

Problem statement

Cleaning technique of the single media was insufficient therefore, the filtered sodium

alginate was contaminated. The possible reasons for the contamination were

Filter media was sitting in acid bath overnight rather than continuously stirred,

Sand might be contaminated by microorganisms.

Dual-media filter (DMF)

combination of:

RH-70 (13cm)

RH-110 (17cm)


TIMEA GREGO 40

Problem Solving

As a consequence of the above, the following problem solving steps were applied. Firstly,

a new sterilisation technique of the single medium was conducted. Sand was washed in

20% HCl solution overnight to remove biological material. After acid bath, the granular

materials were mechanically agitated and rinsed in Milli-Q water several times, gradually

increasing the pH to 7. Finally, the sand was autoclaved for 15mins at 121°C.

Secondly, to make sure that the effluent purity level was adequate water samples were

taken and turbidity level was measured. Thirdly, the filter bed was backwashed because the

turbidity level of the first sample was too high (102.4 NTU). Milli-Q water (0.3 NTU) was

used as control sample for monitoring the effluent turbidity concentration. Following

backwashing the turbidity was reduced to 2.4 NTU, which is higher than Milli-Q water but

significantly lower than before. This may be a reason for potential of added material to the

alginate during filtration.

Parameters of the large scale filtration system:

• Mono (single) medium filter of RH-110,

• Filter medium depth 0.8m.

Parameters of the small scale filtration system (Figure 4.4):

• Dual media filter combination of RH-70 and RH-110,

• Determination of depth of the filter bed is explained in Chapter 5.

In brief, the total depth of the dual media filter was 0.30 m (RH-70=0.13m and

RH-110=0.17m)

4.5 Sand filtration experiment

Before each filtration test fresh 2% HEPES buffered sodium alginate solution was prepared

which was highly viscous, and therefore diluted to 1:10 ratio in Milli-Q water. The alginate

solution was prepared one day priori to treatment. Gravity was used as driving force for the

filtration and the volume of the filtration tank was 10L. Sand filtration was operating with

constant inflow.


TIMEA GREGO 41

4.5.1 Filtered aqueous sodium-alginate samples

Filtration experiments were run overnight and control samples were taken before filtration

for further analysis. Filtered aqueous Na-alginate samples were collected on the following

day and immediately lyophilised. Table 4.2 shows the collected filtered sodium alginate

samples during the fine sand filtration periods.


TIMEA GREGO 42

TABLE 4.2 FILTERED NA-ALGINATE SAMPLES DURING THE FINE SAND FILTRATION PERIODS

FILTRATION TEST

FILTRATION PERIOD FILTRATION

TIME (DAY)

VOLUME OF THE

FILTERED SAMPLE

(ml) STARTED FINISHED

PHASE I.-II. FILTRATION

SA

MP

LE

CO

DE

RH-70/I. 4.30pm

16th June 09

9am

17th June 09 0.90 332

RH-70/II. 9am

17th June 09

5.15pm

17th June 09

0.34 182

RH-70/III. 5.15pm

17th June 09

9.30am

18th June 09 0.68 275

RH-70/IV. 9.30 am

18/06/09

5.15pm

18/06/09 0.36 250

RH-70/V. 5.15pm

18/06/09

12.40 pm

19/06/09 0.83 442

RH-110/I. 4.30pm

16th June 09

9am

17th June 09 0.90 250

RH-110/II. 9am

17th June 09

5.15pm

17th June 09 0.34 136

RH-110/III. 5.15pm

17th June 09

9.30am

18th June 09 0.68 255

RH-110/IV. 9.30 am

18/06/09

5.15pm

18/06/09 0.36 244

RH-110/V. 5.15pm

18/06/09

12.40 pm

19/06/09

0.83 332

SA

MP

LE

CO

DE

RH-70/VI. 4.20pm

30/06/09

9.45am

01/07/09 0.73

275

RH-70/VII. 9am

01/07/09

9.30am

02/07/09 1.02

525

RH-110/VI. 4.20pm

30/06/09

9.45am

01/07/09 0.73

275

RH-110/VII. 9am

01/07/09

9.30am

02/07/09 1.02

525

PHASE IV. FILTRATION (THE SYSTEM WAS SHUT DOWN ON 09/07/09)

SA

MP

LE

CO

DE

DMF-I. 5pm

07/07/09

10am

08/07/09 0.71 500

DMF-II. 10am

08/07/09

12.40pm

08/07/09

0.11 230

MMF-I. 6pm

07/07/09

10am

08/07/09

0.67 500

MMF-II. 10am

08/07/09

12.40pm

08/07/09

0.11 500

PHASE V. FILTRATION

SA

MP

LE

CO

DE

DMF-III. 2.30pm

21/07/09

11.20am

22/07/09 0.45 285

DMF-IV. 11.20am

22/07/09 Filter bed was clogged. 176

MMF-III. 6pm

21/07/09

10am

22/07/09 0.33 490

MMF-IV. 6pm

22/07/09

9am

23/07/09 0.62 520


TIMEA GREGO 43

4.5.2 Sodium alginate viscosity measurement

Viscosity of the sodium alginate solutions was measured to verify the rheological

properties of the fluid. The dynamic viscosity of the alginate solutions having different

concentration, autoclaved and non-autoclaved, fresh and old samples were measured using

Bohline CVO 100 Analogue RSO Viscometer. Samples were dissolved in Milli-Q water

and viscosity measurements were made at 25°C.

Na-alginate powder, with medium viscosity of 2% at 25°C approximately 3500cps. The

Molecular weight of alginic acid Na-salt ranged between 80,000-120,000 (as specified by

the manufacturer). According to this information, viscosity of a 2% Na-alginate solution

could be calculated.

sPaPcps 32 10101 smkgsPa /][

Therefore viscosity of 2% sodium alginate solution at 25°C is smkgcps /5.3500,3 .

4.5.3 Turbidity

It was described previously, during the single medium filter operation the effluent water

was contaminated, hence filter backwashing was performed. This approach, which is part

of rapid filtration, was used to clean the filter bed and to remove much of the impurities,

which were entrapped among the pores. Filter backwashing had been running for 35 mins

and, as a result, turbidity level of the effluent was reduced from the initial concentration of

102.5 Nephelometric Turbidity Units (NTU) to 2.4 NTU. Turbidity was measured in HACH

Ratio/XR Turbidity meter, USA.

4.6 Lyophilisation of filtered sodium-alginate

Lyophilisation or freeze – drying, which is the second stage during the experiment life-

cycle (Figure 4.1), is a method by which filtered sodium-alginate solution is frozen and

then water is removed by sublimation. Freeze-drying process has become a standard

operation process in the medical field for providing “stable state” for liquid materials [61].


TIMEA GREGO 44

Lyophilisation of the solution involves the following steps:

a. Shell freezing of filtered sodium alginate solution: 1L round-bottomed flasks

were manually spun in an acetone bath, which was cooled below 0°C by dry ice.

During this process, a slim layer of frozen sodium alginate formed around the

internal wall of the flask (Figure 4.5). The samples were kept in a freezer to prevent

melting of alginate before they could be transferred to the vacuum.

FIGURE 4.5 SHELL FREEZING PROCESS OF FILTERED NA-ALGINATE SOLUTION

b. Drying of the frozen alginate under vacuum: the sample was placed under a

vacuum and placed in circuit with a freezer at -40°C, where water sublimed to the

freezer section and giving the end product of dry alginic acid (Figure 4.6).

FIGURE 4.6 DRIED ALGINIC ACID AFTER LYOPLISIATION


TIMEA GREGO 45

4.7 Reconstruction of sodium-alginate

At the third stage of the experimental work (Figure 4.1) dried sodium alginates were

rehydrated at ambient temperature in Milli-Q water to a final 1.0%, 1.25%, 1.5%, 1.75%,

1.875% and 2% solution under constant stirring. Therefore, the most appropriate

concentration could be chosen for immobilisation of HepG2 cells in alginate matrix. In

additional to this, to improve the accuracy of particulate analysis the samples were

centrifuged.

4.7.1 Analytical centrifugation

22 samples (Table 4.2) were prepared for two separate centrifuge runs. Dry Na-alginates

samples were dissolved in 1ml Milli-Q water to a final 2% solution under constant

agitation, and then centrifuged for 100mins at 16,000 relative centrifugal force (rcf). “x g”

is the unit of rcf and “g” presents the force of gravity. After the first centrifugation

supernatants were removed from pellet, and resuspended in 1ml Milli-Q water and the tube

was filled up to 1.5ml and centrifuged again as before. At the end of the second

centrifugation, supernatant was taken off and resuspended in 250µl Milli-Q water.

FIGURE 4.7 ANALYTICAL CENTRIFUGE ANALYSIS BY EPPENDORF CENTRIFUGE 5415R


TIMEA GREGO 46

4.7.2 Determination of the size distribution of the particulates

Particle size distribution of filtered, 2% Na-alginate solutions were determined by a

NIKON Eclipse TE200 light microscope. After centrifugation the samples were placed

onto glass microscope slides and to prevent the alginate droplets drying out the samples

were covered with a protective glass sheet and the mean particle size distributions were

calculated. Particle counting of filtered solutions were done by PAMAS SBSS Liquid

particle counting system with the aim of determining size ranges and quantities of the

particles by light obscuration.

4.7.3 Encapsulation of filtered sodium-alginate

Due to lack of experience in cells encapsulation, the empty alginate beads and

immobilisation of HepG2 cells in alginate matrix were done by Amir Gander and assisted

by Timea Grego.

4.7.3.1 Encapsulation of filtered sodium alginate

The aim of the empty beads encapsulation was to analyse the morphology of the beads

thus, filtered, final 1.0%, 1.25, 1.5%, 1.75%, 1.875% and 2% concentration of Na-alginate

solutions were prepared in Milli-Q water, and autoclaved at 121°C for 10 mins. A 50:50

ratio alginate was mixed with high glucose complete media solution, and applied into a

syringe, which was connected to Inotech encapsulation system, which is a droplet

generator utilising vibration [63]. Alginate and media mix passed through a 200µm nozzle

under constant electromagnetic vibration causing bead formation. Polymerisation of the

alginate matrix occurs when the droplets fall into the polymerisation buffer in a 204 mM

CaCl2, 150 mM NaCl and 15 mM HEPES buffer solution (pH 7.4) and had been stirred for

10mins. Then, the beads were washed in Alpha MEM media (Invitrogen) and resuspended

in PBS + Ca2+

(Lonza) [63].


TIMEA GREGO 47

4.8 Encapsulation of HepG2 cells

A 20 ml filtered sodium-alginate with a final concentration of 1.875% was prepared in

Milli-Q water. HepG2 cells were cultured in monolayer for four days, then harvested by

trypsinisation and added to a 50:50 alginate to media mix at 1x106/ml beads. Alginate and

cell-media mix were passed through the 200 µm nozzle of the Inotech encapsulator

creating alginate-HepG2 beads upon immersion into the polymerisation buffer containing

Ca2+

for 10 mins [63].

4.8.1.1 Morphology of empty beads and cell cultured alginate beads

Morphology of the beads was analysed by NIKON Eclipse TE200 microscope. Empty and

HepG2 containing alginate beads were placed on a microscope slide and photographs were

captured.

4.8.1.2 Determination of cells viability in beads

For the live and dead quantitative analysis, “alginate beads were mixed in medium and

transferred from the culture vessel to a 1.5ml microfuge tube. The beads were left to settle

and the volume of the beads was adjusted to 200µl. After this, a 200µl sample of HepG2

alginate microcapsules was washed twice and re-suspended in 500µl of PBSCa+Mg

”. To

identify the viability of the encapsulated cells, “20µl of 1mg/ml propidium iodide (PI) dead

dye” and for the non-viability analysis “10µl of flourescein diacetate (FDA) live dye was

added and incubated for 90seconds, than the tubes were mixed gently”. “The beads were

washed with PBSCa+Mg

again and finally resuspended in 0.5ml of PBSCa+Mg”

. The dyed

microcapsules were placed on a microscope slide for image analysis “using an excitation

filter of 510-560ηm and an emission filter of 590ηm for PI dyed cells, and for FDA dyed

cells, and an excitation filter of 465-495ηm and an emission filter of 515-55ηm. Digital

images were captured with NIKON Eclipse TE200 microscope and the built in DX1200

camera and utilising Lucia imaging software package” [62].


TIMEA GREGO 48

5 EXPERIMENTAL RESULTS AND DISCUSSION

Fine sand filtration was investigated as a possible purification method for purifying

research grade sodium-alginate on a large scale for clinical use. In this section, results of

the filter media test, viscosity measurement, empty beads and HepG2 cell encapsulation in

purified Na-alginate are presented and discussed.

5.1 Testing of filter media

Physical and chemical analysis of the filter media and the various filter tests were

undertaken to determine its suitability for fine sand filtration. In the following sections, the

results of the analysis are presented and discussed in more details.

5.1.1 Physical analysis of filter media

5.1.1.1 Grain size distribution analysis

The aim of the granulometric analysis, which was conducted by SIBELCO UK Ltd, was to

determine the effective grain size and uniformity coefficient of the filter media. Not just

the previously mentioned parameters, but also the fine and very fine particles size

distribution were determined. It is important to know the distribution of these fractions

because these granular materials have large specific surface area, and their surface charge

(zeta potential) is unequal, therefore during sorption these particles present large surface

activity.

The value of the particle diameter at which 60% and 10% of the sands is finer was read

from the graphs. Therefore, uniformity coefficient and effective size range were

determined. In accordance with the sieve analysis, RH-70 was considered as coarse sand

with U=1.6 and deff=135µm, whereas RH-110 was fine sand with U=1.7 and deff=82µm.

Table 5.1 shows the reduced data of grain size distribution analysis of coarse and fine sand

such as cumulative passing and retained fractions of the media and classification of

sediments. Figure 5.1 and 5.2 shows the grain size distribution graphs of coarse and fine

sand. As can be seen from the graphs, the more steep a graph, the more uniform the

particulate distribution.


TIMEA GREGO 49

TABLE 5.1 REDUCED DATA OF GRAIN SIZE DISTRIBUTION ANALYSIS OF

RH-70 (COARSE SAND) AND RH-110 (FINE SAND)

BS STANDARD SIEVE

NOMINAL APERTURES

(µm)

MESH

NUMBERS

PASSING

(%)

RETAINED

(%)

CUMULATIVE

RETAINED (%)

SEDIMENT

CATEGORIES

RH-70 (COARSE SAND)

710 22 100.00 0.00 0.20 Coarse sand

500 30 99.80 0.20

355 44 96.90 2.90 23.10 Medium sand

250 60 76.70 20.20

180 85 33.80 42.90 69.00 Fine sand

125 120 7.70 26.10

90 170 2.30 5.40 7.00

Very fine

sand 63 240 7.70 1.60

<63 240-440 0.00 0.70 0.70 Slit/Clay

RH-110 (FINE SAND)

355 44 100.00 0.00 0.20 Medium sand

250 60 99.80 0.20

180 85 89.10 10.70 59.00 Fine sand

125 120 40.80 48.30

90 170 15.20 25.60 37.10

Very fine

sand 63 240 3.70 11.50

<63 240-440 0.00 3.70 3.70 Slit/Clay


PARTICLE SIZES (µm) VERSUS CUMULATIVE PASSING OF THE GRAINS (%)


TIMEA GREGO 50


PARTICLE SIZE (µm) VERSUS CUMULATIVE PASSING OF THE PARTICLES

5.1.1.2 Density of sands

SIBLECO UK Ltd could only determine the density of RH-70 (Appendix D) and an

assumption was made for RH-110 which was based on literature [39-41, 58]

(Table 5.2). As a consequence of the above mentioned, density of granular media is

ρ=2,650kg/m3.

TABLE 5.2 DETERMINED DENSITY OF RH-70 FILTER MEDIUM COMPARED TO DATA AVAILABLE IN

LITERATURE

SEDIMENTS DENSITY (kg/m3) RESEARCHERS

Quartz sand 2,650 Ives, K. J. (1990) [39]

Sand and grit 2,400-2,600 Judd (2006) [58]

Sand and grit 2,650 Kézdi, Á. (1969) [41]

Sand 2,550-2,650 Tchobanoglous, G. et.al

[40]

RH-70 (Moist)

Sand

2,650 SIBELCO UK Ltd.


TIMEA GREGO 51

5.1.2 Sphericity (shape) of filter media

According to physical analysis of the filter media, grain shape of RH-70 and RH-110 was

sub-angular with medium sphericity (Appendix D). Moreover, two charts of sphericity and

roundness were provided by SIBELCO UK Ltd., and as can been seen from the charts the

more rounded a grain, the higher the sphericity (Ψ) value (Appendix F). In accordance

with the charts, the shape of filter media is non-spherical and the estimated value of

sphericity (0.78) was based on the provided charts and literature (Table 5.3).

TABLE 5.3 ESTIMATED SPHERICITY OF RH-70 AND RH-110 COMPARED TO DATA AVAILABLE IN

LITERATURE

GRANULOMETRIC

MATERIAL

SPHERICITY GRAIN SHAPE RESEARCHERS

Sand 0.85 n/a Ives, K. J. (1990)

[39]

Sand 0.75-0.85 n/a Tchobanoglous, G.

et.al [40]

RH-70 and

RH-110 (Moist)

Sands 0.78

(a) Sub-angular

SIBELCO UK Ltd.

and (a) Casey [46]

5.1.3 Settling velocity of individual grains

The settling velocity of 20 individual coarse and fine sand grains settling through 1 m

column at water temperature 20°C (Appendix G) was measured and the average settling

velocity was calculated (Table 5.4).

TABLE 5.4 AVERAGE SETTLING VELOCITY OF COARSE AND FINE SAND GRAINS

GRANULAR MATERIAL AVERAGE SETTLING VELOCITY

vav (m/s)

RH-70 (COARSE) 0.05

RH-110 (FINE) 0.03


TIMEA GREGO 52

5.1.4 Porosity of filter media

In this section, porosity calculation of RH-110 (fine sand) is presented. As can be seen

from Table 5.3 and Table 5.5, there is a correlation between porosity and sphericity of the

sediments thus, the higher sphericity value, the lower the porosity. According to the

physical analysis of the filter media, grains shape is sub-angular. Therefore, Table 5.5

shows the assumed porosity of filter media based on literature review.

TABLE 5.5 ESTIMATED POROSITY OF RH-70 AND RH-110

TYPE OF GRANULAR

MATERIALS

POROSITY

(n)

GRAIN

SHAPE

RESEARCHERS

Sand 0.43

n/a Ives, K. J. (1990, p

144.) [39]

Sand and grit 0.40-0.45

n/a Kézdi, Á. (1969, p 48.)

[41]

Sand 0.40-0.46

n/a Tchobanoglous, G.

et.al [40]

RH-70 and

RH-110 (Moist) Sand 0.43(a)

Sub-

angular

SIBELCO UK Ltd

(a) Casey [46]

research.

Porosity of RH-110

In this part, the porosity (n) and void ratio (e) of the single (RH-110) medium filter, water

content (w), specific volume (V) had been determined.

Mass of the wet sand (RH-110) Gw=2.80kg

Mass of the dry sand (RH-110) Go=2.08kg

Depth of the mono media filter (MMF) HMMF=0.8m

Internal diameter of the filter tube DMMF=0.0465m

Density of the sand (assumption) 3/2650 mkgs

Density of the sand was adopted from Kézdy, Á. and Ives, K. [39, 41]

Temperature during filter operation T=20°C

Single medium filter is a complex system and can be divided into three parts i.e. air, water

and solids (Figure 5.3).


TIMEA GREGO 53

FIGURE 5.3 MASS –VOLUME RELATIONSHIP OF A SEDIMENT SAMPLE [51]

Volume related symbols [64]

V=Vs+Vv Total volume of the sand

Va Volume of the air

Vv=Va+VW Volume of the voids

Vs Volume of the solids

Vw Volume of the water

Mass related symbols [64]

M=Mw+Ms Total mass of sand sample

Ma Mass of air (0)

Ms Mass of solids

Mw Mass of water

STEP 1 Determination the total volume of the sand

33222

1036.18.04

14.30465.0

4

14.3

4mxm

mH

Dh

dV MMF

MMF

(5.1)

STEP 2 Calculation of water content (w%)

%6.3410008.2

08.280.2100%

0

0

kg

kgkg

G

GGw n (5.2)


TIMEA GREGO 54

STEP 3 Volume of solids (Vs)

34

3

0 1085.7/2650

08.2mx

mkg

kgGV

s

s

(5.3)

STEP 4 Volume of voids (Vv)

343433 1075.51085.71036.1 mxmxmxVVV sv

(5.4)

STEP 5 Void ratio (e)

732.01085.7

1075.534

34

mx

mx

V

Ve

s

v (5.5)

STEP 6 Porosity (n%) of RH-110

%2.421001036.1

1075.5100

33

34

mx

mx

V

Vn v (5.6)

Comparing the calculated porosity value to the estimated porosity based on the literature

(Table 5.5), it can be see that porosity of RH-110 was a little bit over 42% and close to

43%. Thus, the calculated and assumed sphericity (Table 5.3) are valid.


TIMEA GREGO 55

5.2 Hydraulic of filtration

As was described previously, a 0.2%, medium viscosity sodium-alginate solution was

considered as a Newtonian fluid between low and medium shear rate ranges. Viscosity

measurements were conducted to determine the rheological properties of Na-alginate

solutions having different concentration and to validate the above stated assumption. As a

result of this, a 0.2%, medium viscosity Na-alginate solution is Newtonian fluid. In

section 5.3, the results of the viscosity measurements are discussed in more details. In

accordance with the above, head loss in clean water and during backwashing, volumetric

flow rates, and filtration rate calculations were determined.

Firstly, it was necessary to determine the effective size, and the uniformity coefficient of

the granular filter media. These values were obtained from the sieve analysis and presented

in Table 5.6. Secondly, the headloss in the clean filter bed was calculated. Detailed

calculation of hydraulic of filtration can be found (Appendix H).

TABLE 5.6 PHYSICAL PARAMETERS OF THE FILTER MEDIA

PHYSICAL PARAMETER SYMBOL UNIT RH-70 RH-110

SPHERICITY Ψ dimensionless 0.78 0.78

POROSITY n dimensionless 0.43 0.422

OPERATIONAL

TEMPERATURE

T °C 20 20

UNIFORMITY COEFFICIENT 10

60

d

dU

dimensionless 1.6 1.7

EFFECTIVE SIZE d10 µm 135 82

GEOMETRIC MEAN OF

AVERAGE SIEVES ds

m 1.77x10-4

1.04x10-4

AVERAGE SETTLING

VELOCITY vs

m/s 0.05 0.03

5.2.1 Determination flow regimes and drag coefficient

STEP 1. Determination of the value of Reynolds‟ number (Re) for coarse and fine sands

using Eq. (3.5)

RH-70

9

)/100087.1(

)/05.0()/23.998()/1077.1(3

34

2

2

mskg

smmkgsmvdR

OH

OHse


TIMEA GREGO 56

RH-110

As can be seen from the results, the flow region of coarse sand grains was transition and

fine sand grains was laminar.

STEP 2. Determination of the drag coefficient for laminar and transition flow using Eq.

(3.10) and Eq. (3.12)

Drag coefficient for laminar flow regime (RH-110)

824

e

DR

C

Drag coefficient for transitional flow regime (RH-70)

53.1)/05.0()/100087.1()3(

)1077.1()/23.998/650,2()/81.94(23

433

smmskg

mmkgmkgsmCD

STEP 3. Calculation of eR

Cd”velocity term” [49] and

2

eDRC “diameter term” [49] value for

transitional flow regime using Eq. (3.13) and Eq. (3.14)

RH-70

17.0)/05.0()/23.998()3(

)/100087.1()/77.1651()/81.94(3

3

smmskg

mskgmskgsm

R

Cd

e

54.117)/100087.1()3(

)1077.1()/23.998()/77.1651()/81.94(3

342

mskg

mmskgmskgsmRC eD

Also, the value of eR

Cd and

2

eDRC can be read from Camp‟s curves (Appendix B).

3)/100087.1(

)/03.0()/23.998()/1004.1(3

34

2

2

mskg

smmkgsmvdR

OH

OHse


TIMEA GREGO 57

STEP 4. Control calculation of drag coefficient using the yield value of eR

Cd and

2

eDRC

The drag coefficient for transitional flow was determined at Step 2 and gave CD=1.53.

Determination the control value of drag coefficient

From From

53.117.0 D

e

D CR

C 45.154.117

2 DeD CRC

As can be seen from the results, drag coefficient calculated from the “diameter term” was

slightly less than the CD value yield from the general equation of settling velocity (Step 2).

STEP 5. Control calculation of settling velocity using Eq. (3.9)

Settling velocity of 20 individual particles was measured and averaged velocity was

calculated thus, settling velocity of coarse particles was 0.05m/s and fine particles settled

out under 0.03m/s velocity.

Determination of the control value of settling velocity

For RH-70 For RH-110

]/[05.0

3

4

2

2 smC

dgv

OHD

OHss

]/[02.0

3

4

2

2 smC

dgv

OHD

OHss

Control results show that settling velocity of coarse sand was equal with the result obtained

from laboratory analysis. However, control settling velocity of fine sand was 0.01m/s less

than the measured value under laboratory condition. One possible reason of the difference

is human error, the settling velocity was measured in the laboratory using a stop watch.

5.2.2 Determination of Clean-Water Head loss in granular-media filter

During the five phases of filtration experiment, 22 filtered Na-alginate samples were

collected thus, the volumetric flow rates (Q), retention times (t), and filtration rates

(q=vapproach) were calculated. As can be seen from Table 4.2, majority of the fine sand

filtrations were run overnight thus, constant downward flow filtration was not performed.


TIMEA GREGO 58

As a consequence of this, dual media filter was clogged on 22/07/2009. Filter backwashing

was not operated because the filtration experiment had finished. Head loss through the

filter bed could not be measured because manometer was not installed at the system.

Therefore, the theoretical head loss was determined.

STEP 1. DETERMINATION OF EQUIVALENT HYDRAULIC DIAMETER

It is known, that sphericity is the ratio between the hydraulic diameter (dh) and the

geometric mean of the sieve sizes (ds) Eq. (4.2). Since the sphericity (Table 5.3) and the

average sieves sizes (ds) were known only, the equivalent hydraulic diameter had to be

calculated. Reduced data of sieve analysis and calculation of geometric mean of average

sieves can be found (Appendix I).

TABLE 5.7 ESTIMATED SPHERICITY VALUE AND CALCULATED EQUIVALENT HYDRAULIC DIAMETER

TYPE OF

GRANULAR

MATERIAL

SPHERICITY

(Ψ)

EQUIVALENT

HYDRAULIC

DIAMETER

sh dd [m]

DIAMETER

OF AVERAGE

SIEVES dS [m]

REYNOLDS‟

NUMBER

RH-70

(COARSE)

0.78 1.38x10-4

1.77X10-4

9

RH-110 (FINE) 0.78 8.12x10-5

1.04x10-4

3

Table 5.8, summaries the calculated head loss in a clean filter for small scale, (single

medium systems) and large scale, (single medium systems) using Eq. (3.15). Figure 5.4

(A-F) show volumetric flow rates (Q) versus sampling dates. As can be seen from the

graphs, filtrations were run overnight, thus flow rates were not constant. During Phase I-II.

filtration, fine filter medium (RH-110) purified 25% less Na-alginate solution compare to

RH-70. Downward flow was increased on 18th

June, to establish the most efficient flow

rate for successful filtration of Na-alginate solution. Filter media depth were increased by 5

cm at Phase III. but it had not significantly changed the volume of filtered solution. From

Figure 5.4 F can be seen, overnight operation of the mono media filter had caused 40%

decrease in the volumetric flow rate, which indicates that the filter was being clogged and

filter backwashing would be required.


TIMEA GREGO 59

TABLE 5.8 VOLUMETRIC FLOW RATES, APPROACH VELOCITIES AND HEADLOSSES DURING

FILTRATION

FILTRATION

TEST

FILTRATION PERIOD VOLUMETRIC

FLOW RATE

(Q) [m3/s]

APPROACH

VELOCITY

vapp=Q/A [m/s]

HEADLOSS IN

CLEAN FILTER

EQ. (3.15) [m]

FILTER MEDIUM

DEPTH

[m] STARTED FINISHED

PHASE I.-II. FILTRATION

SA

MP

LE

CO

DE

RH-70/I. 4.30pm

16th June 09

9am

17th June 09 4.33x10

-9 2.70x10

-5

2.68x10-2

0.25

RH-70/II. 9am

17th June 09

5.15pm

17th June 09

6.17x10-9

3.84x10

-5

3.82x10-2

0.25

RH-70/III. 5.15pm

17th June 09

9.30am

18th June 09 4.67x10

-9 2.91x10

-5

2.89x10-2

0.25

RH-70/IV. 9.30 am

18/06/09

5.15pm

18/06/09 8.00x10

-9 4.98x10

-5

4.95x10-2

0.25

RH-70/V. 5.15pm

18/06/09

12.40 pm

19/06/09 6.17x10

-9 3.84x10

-5

3.82x10-2

0.25

RH-110/I. 4.30pm

16th June 09

9am

17th June 09 3.17x10

-9 1.97x10

-5

6.16x10-2

0.25

RH-110/II. 9am

17th June 09

5.15pm

17th June 09 4.50x10

-9 2.80x10

-5

8.76x10-2

0.25

RH-110/III. 5.15pm

17th June 09

9.30am

18th June 09 4.33x10

-9 2.70x10

-5

8.43x10-2

0.25

RH-110/IV. 9.30 am

18/06/09

5.15pm

18/06/09 7.67x10

-9 4.77x10

-5

1.49x10-1

0.25

RH-110/V. 5.15pm

18/06/09

12.40 pm

19/06/09

4.67x10-9

2.91x10

-5

9.08x10-2

0.25

PHASE III. FILTRATION

SA

MP

LE

CO

DE

RH-70/VI. 4.20pm

30/06/09

9.45am

01/07/09 4.33x10

-9

2.70x10-5

3.50x10-2

0.3

RH-70/VII. 9am

01/07/09

9.30am

02/07/09 6.00x10

-9

3.74x10-5

4.85x10-2

0.3

RH-110/VI. 4.20pm

30/06/09

9.45am

01/07/09 4.33x10

-9

2.70x10-5

3.50x10-2

0.3

RH-110/VII. 9am

01/07/09

9.30am

02/07/09 6.00x10

-9

3.74x10-5

4.85x10-2

0.3

PHASE IV. FILTRATION

SA

MP

LE

CO

DE

DMF-I. 5pm

07/07/09

10am

08/07/09 8.17x10

-9 5.08x10

-5

0.3

DMF-II. 10am

08/07/09

12.40pm

08/07/09

2.40x10-8

1.49x10

-4

0.3

MMF-I. 6pm

07/07/09

10am

08/07/09

8.67x10-9

5.10x10

-6

3.83x10-2

0.6

MMF-II. 10am

08/07/09

12.40pm

08/07/09

5.22x10-8

3.07x10

-5

2.30x10-1

0.6

PHASE V. FILTRATION

SA

MP

LE

CO

DE

DMF-III. 2.30pm

21/07/09

11.20am

22/07/09 7.33x10

-9 4.57x10

-5

0.3

DMF-IV. 11.20am

22/07/09 FILTER BED WAS CLOGGED.

MMF-III. 6pm

21/07/09

10am

22/07/09 1.70x10

-8 1.00x10

-5

1.00x10-1

0.8

MMF-IV. 6pm

22/07/09

9am

23/07/09 9.67x10

-9 5.69x10

-6

5.69x10-2

0.8


TIMEA GREGO 60

(A) (B)

(C)

(D)

PHASE IV. FILTRATION WAS SHUT DOWN.

(E)

(F)

FIGURE 5.4 (A-F) VOLUMETRIC FLOW RATES VERSUS SAMPLING DATES DURING

PHASE I-III AND V. FILTRATION PERIODS


TIMEA GREGO 61

5.2.3 Turbidity

During the Phase IV filtration, the effluent became contaminated, consequently,

backwashing was performed and turbidity level was measured. The aim of the

backwashing was to remove the possible material and substances entrapped among the

pores thus, filter bed expansion and backwash flow rates were not measured during the

process operation. However, the theoretical head loss through the filter bed was

determined. Four samples were taken for turbidity analysis after backwashing. As can be

seen from Table 5.9, filter cleaning had been running with Milli-Q water for four days and

the turbidity level significantly decreased from the initial concentration of 102.5 NTU to

2.4 NTU. Comparing the final sample concentration to the control sample (Milli-Q water),

it can be clearly seen that the effluent was still turbid but due to lack of time for the

experiment the filter cleaning step was stopped. Therefore, there is a possibility that

filtered Na-alginate solutions physical properties were modified due to the contamination.

TABLE 5.9 TURBIDITY LEVEL AND CONCENTRATION (NTU) OF THE EFFLUENT SAMPLES AFTER

BACKWASHING

EFFLUENT SAMPLES

(AFTER BACKWASHING)

TURBIDITY

(NEPHELOMETRIC

TURBIDITY UNITS=NTU)

TURBIDITY

CONCENTRATION

(%)

16/07/2009 102.5 34,067%

17/07/2009 88.8 29,500%

20/07/2009 4.59 1,430%

21/07/2009 2.4 700%

Milli-Q water (Control

sample)

0.3 0

5.2.3.1 Calculation of the theoretical head loss in fluidised

Porosity

nRH-110=0.422

32650

m

kgS

Filter bed depth before backwashing

L=0.8m )20(2 CTOH

=998.23

3m

kg


TIMEA GREGO 62

mLnhOH

OHS 77.0)(

12

2

(5.1)

5.3 Evaluation of the viscosity measurements of sodium alginate solutions

The goals of the analysis were to determine the rheological properties of aqueous Na-

alginate under conventional and extreme conditions, and to evaluate the effects of the

filtration method on the solution viscosity. It is well known that molecular weight,

temperature and solution concentration affect the viscosity, which determines the size and

shape of the alginate beads. Table 5.10 summarises the prepared samples for the analysis.

Viscosity of the samples was measured at 25°C unless otherwise stated.

TABLE 5.10 PREPARED SAMPLES FOR VISCOSITY ANALYSIS

SAMPLE CODES SOLUTION CONCENTRATION

(%)

COMMENTS

CONTROL SAMPLES 0.25%, 0.5% 1%, 1.25%,

1.5%, 1.875%, 2%, 2.5%, 3%,

3.5%, 4%

Fresh NaAlg solution and

autoclaved *

FRESH ALGINATE 0.2%, 2% autoclaved

not lyophilised

2% was measured at 25°C

and 37°C

0.2%, 2% not autoclaved

not lyophilised

0.2%, 2% autoclaved

not lyophilised

UNFILTERED

NaAlg SOLUTION

0.2%, 2% autoclaved

lyophilised

OLD NAALG SOLUTION 0.2%, 2% 1 week old sample

RH-70 0.2%, 2%

autoclaved

lyophilised RH-110

DMF 0.2%, 2%

autoclaved

lyophilised MMF

(*) Sterilisation of the sample at 121°C for 10mins unless otherwise stated.

The results of the analysis were plotted on three different graphs i.e. dynamic viscosity as a

function of shear rate and shear stress, and the flow curves, which shows the Newtonian or

non-Newtonian behaviour of the fluids. Viscosity of a Newtonian fluid changing by stress

and temperature but independent of shear rate, whereas a non-Newtonian fluid viscosity

highly dependent on these three factors.


TIMEA GREGO 63

In the following sections, the flow curves are presented and the majority of the analysis can

be found (Appendix J).

5.3.1 Effect of different concentration on solution viscosity

Figure 5.5 A-D shows the shear stress as a function of shear rate (flow curve) of various

alginate solution concentrations.

(A)

(B)


TIMEA GREGO 64

(C)

(D)

FIGURE 5.5 A-D FLOW CURVES OF NEWTONIAN AND TIME-INDEPENDENT NON-NEWTONIAN

FLUIDS. FLOW CURVES SHOW THE RELATIONSHIP BETWEEN SHEAR STRESS (Pa) AND SHEAR

RATE (1/S) FOR VARIOUS ALGINATE CONCENTRATIONS.

As can be seen from Figure 5 (A), flow curves of 0.2% and 0.25% NaAlg solutions are

linear confirming these samples as exhibiting Newtonian fluid behaviour. As a result, the

assumption was made in Section 3.5.5 was verified. Figure 5 (B-D) show, the fluids

behaviour was changed by increasing the solutions concentration. As a result of this, the

assumption that alginate solutions below 0.5% shows Newtonian characteristic made in

Section 3.5.5 was verified.


TIMEA GREGO 65

Figure 5 B-D show, the fluids behaviour was changed by increasing the solutions

concentration. As a result, the viscosity of the solutions was gradually decreasing as shear

rates were increasing. These systems showed non-Newtonian (pseudoplastic) fluid

properties.

5.3.2 Effect of autoclaving on solution viscosity

Table 5.11 shows the effect of autoclaving on the 0.2% and 2% aqueous Na-alginate

solutions viscosity, measurement were made at 25°C. The effects of the sterilisation can be

seen on Figure 5.6-5.7. A 82% decrease in the viscosity of 2% NaAlg solution autoclaved

for 20mins was measured compared to a reduction of 40% for 2% alginate autoclaved for

10mins. Clearly it can be seen from Table 5.11 and Figure 5.6-5.7, that 20mins autoclaving

caused higher degree of depolymerisation of the uronic acid residues. Flow curve of a 2%

NaAlg solution after 20mins heat treatment almost showed Newtonian fluid characteristic.

Autoclaving altered the physical characteristic of the alginate solution therefore, the size

and sphericity of the beads could be affected which may result a decline in cell growth and

viability.

TABLE 5.11 EFFECT OF AUTOCLAVING AT 121°C FOR 10 AND 20MINS ON THE VISCOSITY OF NAALG

SOLUTIONS

NA-ALGIANTE

CONCENTRATION (%)

VISCOSITY OF THE SOLUTIONS (Pa s)

NON-AUTOCLAVED AUTOCLAVED

(10MINS)

AUTOCLAVED

(20MINS)

0.2 0.007 0.07

0.25 0.008

2 0.802 0.480 0.143


TIMEA GREGO 66

FIGURE 5.6 DYNAMIC VISCOSITY OF 0.2% NAALG SOLUTION AS A FUNCTION OF

SHEAR STRESS (Pa)

FIGURE 5.7 FLOW CURVES OF AUTOCLAVED NaAlg (CONTROL) SOLUTIONS

5.3.3 Effect of temperature on solution viscosity

The effects of temperature on the alginate solution show in Figure 5.8. The viscosity of the

solution decreased by almost 34% at 37°C compared to the viscosity at 25°C. Flow curves

demonstrated (Figure 5.9) that non-Newtonian behaviour of the fluid significantly

decreased by increased temperature. Moreover, a 2% NaAlg solution at 37°C showed

almost Newtonian behaviour. The temperature results demonstrated the sensitivity of

alginate to increased heat.


TIMEA GREGO 67

FIGURE 5.8 VISCOSITY OF 2% NAALG SOLUTION AT DIFFERENT TEMPERATURES

FIGURE 5.9 FLOW CURVES OF 2% NAALG SOLUTION AT DIFFERENT TEMPERATURES

5.3.4 Effect of filtration on the solution viscosity

Figure 5.10-5.11 represent the viscosity changes of 0.2% alginate solutions compared to

non-filtered aqueous Na-alginate after fine and coarse sand, mono medium and dual media

filtration. It can be seen from Figure 5.10 that non-filtered but freeze-dried, and autoclaved

0.2% sample was pseudoplastic compared to the purified sodium alginate solutions, which


TIMEA GREGO 68

exhibited Newtonian fluid characteristic. Clearly, viscosity of a pseudoplastic fluid

changes by increased shear rate, whereas viscosity of a Newtonian fluid is staying

constants by increasing shear rate. Viscosity has the greatest effect on alginate gel

properties and due to the dilution and filtration process alginate solutions lost its strength

and caused non-spherical beads morphology during encapsulation. This could be one of the

reasons for the cells viability and growth decline.

FIGURE 5.10 FLOW CURVES OF FINE SAND FILTERED 0.2% NA-ALGINATE SOLUTIONS

FIGURE 5.11 FLOW CURVES OF MONO MEDIUM AND DUAL MEDIA FILTERED 0.2% NA-ALGINATE

SOLUTIONS


TIMEA GREGO 69

5.4 Reconstruction of sodium-alginate

5.4.1 Evaluation of the effect of fine sand filtration on the size distribution of the

particulates

Figure 5.12 indicates, the broad spectrum of micron particulates presented in non-purified

2% sodium alginate solution. During the filtration period, the particulate size was lowered

in the solution compared to non-purified sample and Figure 5.13 A-B and Figure 5.14 A-B

show the effect of coarse and fine sand filtration on the particulate length distribution of

filtered Na-alginate. As can be seen from the graphs, fine sand filter removed larger micron

particulates compared to coarse sand. Increasing downward flow had slightly changed the

removal efficiency of the filtered thus, less micron particles were removed. Particles size

and shape has significant effect on the filtration efficiency therefore, it was difficult to

design an effective filtration system because the size distribution of particulates in alginate

suspension was not known. It is also important to know the surface characteristics such as

surface charge of the particles, which is predominant for micron size particles, in the

surrounding fluid [65].

During filtration suspended particles interacts with the liquid causing repulsion or

attraction. These factors would determine the settling behaviour of the particulates. If

particle size distribution of aqueous Na-alginate had known, a more effective filter media

system would have designed.

FIGURE 5.12 PARTICLE SIZE DISTRIBUTION OF NON-PURIFIED 2% SODIUM ALGINATE SOLUTION

MEAN PARTICLE SIZE

5µM±2µM


TIMEA GREGO 70

(A)

(B)

FIGURE 5.13 LENGTH DISTRIBUTION (µm) OF PARTICLES IN RH-70 SAND FILTERED NAALG

SOLUTION. (A) PARTICLES LENGTH UNDER SLOW FLOW RATE (B) PARTICLES LENGTH UNDER

INCREASED FLOW RATE

RH-70 (coarse sand),

increased flow rate

MEAN PARTICLE SIZE

4.70µm

RH-70 (coarse sand), slow

flow rate

MEAN PARTICLE SIZE

4.43µm

PARTICLE SIZE (µm)

PARTICLE SIZE (µm)


TIMEA GREGO 71

(A)

(B)

FIGURE 5.14 LENGTH DISTRIBUTION (µm) OF PARTICLES IN RH-110 SAND FILTERED NAALG

SOLUTION. (A) PARTICLES LENGTH UNDER SLOW FLOW RATE (B) PARTICLES LENGTH UNDER

INCREASED FLOW RATE

RH-110/III. (fine sand),

slow flow rate

MEAN PARTICLE SIZE

3.60µm

RH-110/I. (fine sand),

increased flow rate

MEAN PARTICLE SIZE

3.90µm

PARTICLE SIZE (µm)

PARTICLE SIZE (µm)


TIMEA GREGO 72

Figure 5.15 show the removal efficiency of Phase V filtration. Although sand depth was

0.80 m, particulates reduction had not changed significantly compared to small scale

filtration run. Large scale filtration system was operated less efficient due to the

contaminated filter medium. Although filter backwashing was performed, the turbidity

level of the effluent was not decreased to 0.3NTU. The smaller particulates could not

percolate through the filter bed because of the entrapped organic and probably inorganic

impurities in the filter medium.

FIGURE 5.15 PARTICLES LENGTH DISTRIBUTION (µm) OF LARGE SCALE FINE SAND FILTERED

NAALG SOLUTION

After Phase I-III filtration experiment, firstly, it could be determined that fine sand filter

retained larger micron particulates compared to coarse sand. Secondly, slow downward

flow rate was more efficient compared to increased flow. Therefore, the filter media were

operated at slow downward flow rate during Phase IV-V filtration.

PARTICLE SIZE (µm)

MMF-II. Large scale, fine

sand filtration.

Sand depth 0.8m

MEAN PARTICLE SIZE

3.97µm


TIMEA GREGO 73

5.4.2 Evaluation of particle size measurement of filtered sodium alginate solutions

Particle size measurements of filtered Na-alginate solutions were carried out at

Hepatology, Royal Free Hospital. Results obtained from the analysis were used to calculate

the particles size and number distribution of 1g Na-alginate powder, which can be seen

from Figure 5.16. Detailed calculation of particle size measurement can be found

Appendix K.

FIGURE 5.16 DISTRIBUTION OF PARTICLES SIZES (µm) AND NUMBERS OF NA-ALGINATE

POWDER

As can be seen from the results, the most contaminated sample was the unpurified and non-

autoclaved Na-alginate solution. During the filter run almost all large (10-20µm)

particulates were removed from the solutions. An 82% decrease in the number of 2µm

particulates of autoclaved solution was detected compared to raw Na-alginate solution.

Dual media filter had eliminated 5-10µm particulates and 94% of 2µm particulates

compared to unfiltered alginate solution. Mono medium filter (MMF) and fine sand filter

removal efficiency of 2µm were almost the same. The reason of the inadequate MMF

operation compared to small scale RH-110 filter was the contaminated filter bed.


TIMEA GREGO 74

5.4.3 Encapsulation of filtered sodium-alginate

As a consequence of the fine sand filtration experiment, viscosity of the sodium alginate

solution was changed. Since the aim of the encapsulation was to produce micron beads

with spherical shape, the Na-alginate concentrations -1.25%, 1.5%, 1.75%, 1.87%, and 2%

- were analysed and compared to 1.0% normally used.

5.4.3.1 Empty beads morphology

As can be seen from Figure 5.17 beads formed from non-purified Na-alginate solution had

more spherical geometry compared to microcapsules prepared from purified alginate

Figure 5.18. At the first encapsulation stage a 2% alginate solution was used for empty

microcapsulation. Stretch marks Figure 5.18 were found on the surface of the microbeads,

moreover the shape of the beads were non-spherical. Additional to this, decreasing the

solution concentration caused more deformed, non-spherical beads and irregular sizes.

As can be seen, the solution properties of alginate strongly dependent on the viscosity,

which could be affected by the molecular weight of the polymer, concentration and

possible residual impurities. Sodium alginate powder was research grade, thus information

regarding the chemical composition of the polymer was not available.

As the encapsulation process went further, it was found that, 1.875 % purified alginate

solution gave more homogeneous beads, which is another important parameter of alginate

beads in application for immobilising living cells. Therefore, HepG2 cells were

encapsulated in 1.875% DMF and MMF filtered solutions.

FIGURE 5.17 ALGINATE BEADS FROM NON-PURIFIED 2% SODIUM ALGINATE SOLUTION


TIMEA GREGO 75

A C

B D

FIGURE 5.18. EFFECT OF THE NA-ALGINATE SOLUTION CONCENTRATION ON THE MORPHOLOGY

AND SIZE OF EMPTY ALGINATE BEADS. BEADS WERE PREPARED USING 1.25% (A) AND 1.5% (B)

DMF FILTERED SOLUTION; C 1.75% MMF FILTERED SOLUTION AND D 2% FINE SAND FILTERED

AQUEOUS NA-ALGINATE.


TIMEA GREGO 76

5.5 Encapsulation of HepG2 cells

5.5.1 Morphology of cell cultured alginate beads

Alginate micro spheres from mono medium filtered sodium alginate were more spherical

(Figure 5.19) than beads from dual media filtered alginate at day 4 (Figure 5.20).

FIGURE 5.19 PHASE 4: ENCAPSULATED HEPG2 CELLS; MONO MEDIUM FILTERED 1.875% SODIUM

ALGINATE SOLUTION WAS USED FOR CELL IMMOBILISATION

FIGURE 5.20 PHASE 4: ENCAPSULATED HEPG2 CELLS; DUAL MEDIA FILTERED 1.875% SODIUM

ALGINATE SOLUTION WAS USED FOR CELL IMMOBILISATION


TIMEA GREGO 77

5.5.2 Protein production from encapsulated HepG2 cell

Curves are showing the protein synthesis as a function of time for three different purified

alginate culture conditions (Figure 5.24). As can be seen, protein production from dual

media filtered (DMF) alginate encapsulated HepG2 cells were lower than mono media

filtered at day 9 of culture. The largest production in albumin was at day 9 for mono media

filtered and non filtered alginate entrapped HepG2.

FIGURE 5.24 ALBUMIN (PROTEIN) PRODUCTION FROM ENCAPSULATED HEPG2 CELLS IN A NON

FILTERED, MONO MEDIUM, AND DUAL MEDIA FILTERED ALGINATE SOLUTION


TIMEA GREGO 78

5.5.3 Assessment of cell viability in alginate beads

Figure 5.21 shows the viability of the purified sodium alginate encapsulated HepG2 cells

compared to non filtered microcapsules. As can be seen from the graphs, at day 5 viability

of the cells reached a peak. HepG2 cells cultured in non filtered and mono media filtered

alginate beads had levelled out between days 5-15 compared to cells maintained in the non

filtered alginate matrix; viability of the cells were significantly lower. It is unclear what

caused this decrease, it may have been due to the lack of support provided by the beads, or

potentially an anomalous result. The experiment would need to be repeated to confirm this.

FIGURE 5.21 VIABILITY OF ENCAPSULATED HEPG2 CELLS IN A NON FILTERED AND MONO MEDIUM,

AND DUAL MEDIA FILTERED ALGINATE SOLUTION

Figure 5.22 demonstrates that cells number of mono media purified alginate encapsulated

HepG2 cells slightly declined compared to cells in non purified alginate beads. It also can

be seen, that the dual media filtered alginate solution had the greatest impact on alginate

solution properties, cells number dropped below 2x10-6

.


TIMEA GREGO 79

FIGURE 5.22 CELLS NUMBER OF PURIFIED AND NON PURIFIED ALGINATE ENCAPSULATED HEPG2

CELLS

Analysis of FDA and PI stained beads using confocal microscopy showed a low number of

dead cells in both mono media filtered and dual media filtered at day 4 (Figure 5.23).

Overall, there is a clear difference between HepG2 cells encapsulated in non filtered

alginate, DMF and MMF.


TIMEA GREGO 80

(A) DAY 3 LIVE HEPG2 CELLS; DMF (C)DAY 3 LIVE HEPG2 CELLS; MMF

(B) DAY 3 DIED HEPG2 CELLS; DMF (D) DAY 3 DIED HEPG2 CELLS; MMF

FIGURE 5.23 LIVE AND DEAD ENCAPSULATED HepG2 CELLS IN MONO MEDIUM AND DUAL MEDIA

FILTERED ALGINATE MICROCAPSULES AT DAY 3 STAINED WITH FDA AND PI.

(A) DMF FILTERED NaAlg ENCAPSULATED CELLS GROWTH AND (B) CELLS DECLINE.

(C) MMF FILTERED NaAlg ENCAPSULATED CELLS GROWTH AND (D) CELLS DECLINE.

GREEN SHOWS LIVE CELLS, AND READ REPRESENT DEAD CELLS.


TIMEA GREGO 81

6 CONCLUSIONS

The hypothesis of the dissertation was that sand filtration would remove micron particulates

from aqueous sodium alginate, thereby providing a biocompatible polymer matrix for cell

encapsulation. It was believed that sand filtration could act as a promising purification process

for large scale production of purified alginate without altering the physical composition of the

polymer. Therefore, the micron particulates removal efficiency of the sand filtration was

investigated by setting up various filtration systems. As was described in this dissertation, it

was revealed that filtering sodium alginate solution through a dual media filter could

eliminate micron particulates between 10 to 20µm, and reduced 94% of 2µm particulates

compared to non filtered Na-alginate solution. Filtration process showed a major impact on

the dynamic viscosity of the solution, even thought the micron particulates were removed to

some extent. As a consequence of this, the stability and the morphology of the micro beads

were hugely affected. Furthermore, it was showed that immobilisation of HepG2 cells in fine

sand filtered, aqueous sodium alginate had not provide a coherent gel matrix for the cells.

The following findings and conclusions of this work had contributed to understand the

disadvantages of solid-liquid based purification of sodium alginate, therefore providing an

idea for a unique and an alternative purification method, which is presented in the

recommendation chapter.

The results of the sand filtration experiment have shown that during the first stage of the

investigation; fine sand (RH-110) filter had greater removal efficiency for particulates

between 2µm and 5µm compared to coarse sand (RH-70) and to the particulate size

distribution of non filtered sodium alginate solution. Mean particle size removal of fine sand

was 3.60µm whereas coarse sand 4.43µm under slow flow rate. Varying the flow rate during

Phase I-II filtrations had resulted to purify more sodium alginate but there was a decline in the

particulates removal efficiency. Fine sand filter could purify 0.39ml more alginate solution for

the fourth sampling compared to the first sample. For the same period, coarse sand filter

purified 0.32ml more than at the first sampling.


TIMEA GREGO 82

Increasing, the inflow rate gave higher volumetric flow rate but provided less purified alginate

solution. Moreover, filtration time (t) was between 0.73 and 1.02 days for fine and coarse

sand filter. The results indicated that the filter was being clogged and the headloss was

increased. It is necessary to mentioned, Phase I-III filtration experiments were small scale

systems, filters were running overnight and that problem was accumulated more during Phase

V filtration when dual media filter become clogged.

This work has also showed that filter bed was contaminated due to insufficient cleaning

process, which was described in section 4.4. During Phase IV-V the filter media was not clean

before filtration; therefore it could not provide acceptable purity level for sodium alginate

solution. Although the turbidity level of the mono media filtered effluent was reduced to

2.4NTU from the initial 102.5NTU concentration, the filter was not purely clean. This

probably increased the impurity content of the filtered Na-alginate solution.

What has been revealed from the viscosity measurement enabled to understand the

rheological properties of the polymer solution and the analysis showed the complexity of the

aqueous sodium alginate. At the concentration spectrum of 1% to 4% the viscosity behaviour

of alginate solution was pseudoplastic. Any decrease in concentration altered the dynamic

viscosity of the solution, thus providing a thin fluid like water and it determined the effect on

the cell encapsulation. Several encapsulation tests were undertaken to identify the most

suitable concentration for spherical alginate beads production. As a consequence of this, it

was shown that 1.875% filtered sodium alginate solution gave more spherical beads but the

size distribution of the beads were multiform.

The work had highlighted that viscosity of sodium alginate decreased by not only

concentration, temperature, time but also autoclaving. Increasing temperature by 12°C a 2%

alginate solution showed almost Newtonian fluid characteristic. This finding suggests that

increasing temperature even higher would lead to a total loss in thickness of the fluid.

Autoclaving a 2% alginate solution for 20mins had drastically decreased the dynamic

viscosity of the fluid compared to a non autoclaved sample; viscosity dropped by 82%.

The solution property of alginates not only depends on the viscosity but also on the chemical

composition and impurities presented in the powder of sodium alginate. The study has

reinforced the impact of unknown detailed chemical and composition analysis of the polymer

since, it made difficult to be able to better predict the reasonable outcome of the filtrations.


TIMEA GREGO 83

This research has also shown that immobilisation of HepG2 cells within mono medium and

dual media filtered (DMF) alginate spheres had altered the growth of cells, although function

as measured by albumin was relatively unaffected compared to cell number and protein

synthesis of the cells compared to unfiltered spheroid cultures.

Growth function of HepG2 cells cultured in DMF had dropped below 2x10-6

between days 4

and 9 and gradually started to increase, and at day 14 had reached its peak (10x10-6

). Same

process was observed during viability, when after a sudden increase at day 9 the viability of

the cells dropped by 2%. Albumin (protein) production of DMF HepG2 microspheres was

flocculating. Protein content of cell at day 4 and 12 was on peak, reaching 7µg/million. Mono

media filtered (MMF) micro bead had least modified the growth, function, and viability of the

cells compared to DMF. Viability of HepG2 cells entrapped in the flexible matrix was

levelled out between day 5 and 10. Albumin production of the cells had fall to 6µg/million by

day 6 compared to 8µg/million of unfiltered beads by the same time.


TIMEA GREGO 84

7 RECOMMENDATIONS

The experimental works were evaluated not only by engineering and medical science aspects

but also the efficiency of the currently publish purification processes were put into

consideration [14, 18-19, 25]. Therefore, a key question was addressed:

“Is the liquid-solid based purification method the most appropriate for the purification of

“micron particulates free” alginate biopolymer suitable for cell encapsulation?”

For immobilisation of cells in alginate matrix, the most important property is the polymer‟s

viscosity. As could be seen, many published in-house purification protocols and the fine sand

filtration had utilised alginate solubility in water. Thus, liquid-solid separation processes have

been investigated for alginate purification.

In this section, an alternative approach for alginate purification, utilising dry alginate powder

in a cyclonic separation from gas stream and possible production of “micron particulate free”

alginate for cell encapsulation is discussed. Utilisation of cyclones are diverse in various

science and engineering technologies, for instance air pollution control, spraying of

pharmaceutical drugs, and powders classification on the basis of their physical and chemical

properties [66-67]. It is important to mention, that the proposed separation process would be a

unique solution for alginate purification and it might help to optimize the rheological

behaviour of the polymer.

As can be seen from, Figure 7.1 micron and macro particles (2µm-1000µm) can be removed

from gas streams by cyclones.


TIMEA GREGO 85

FIGURE 7.1 PARTICLE SIZE DISTRIBUTION OF SEVERAL MATERIALS ( ) IN A FUNCTION OF

SEPARATION TECHNIQUES ( ); CYCLONES SEPARATION PARTICLE SIZE 1µm-1,000µm [66].

7.1 Operation of the gas-solid cyclone

The operation of the commonly designed cyclone is based on reverse flow where particles are

suspended in a gas stream and separated by centrifugal force [66]. Figure 7.2 shows a

schematic diagram and process steps of a reverse flow cyclone.

FIGURE 2 REVERSE FLOW CYCLONE SEPARATOR. SUSPENDED PARTICLES IN GAS ENTER IN THE CYCLONE AND

CENTRIFUGAL FORCES (VORTEX) DRIVE THEM TO THE CENTRAL PART, WHERE THE PARTICLES WILL BE SEPARATED

[68.]

Separation processes

Particles size range of

different materials


TIMEA GREGO 86

Cyclones can be categorised by their size such as large- and small cyclones, which is more

efficient in particulates removal below 5µm [70]. Figure 7.3 shows the grade efficiency curve

of cyclones separator. As can be seen from the diagram, 50% of micron particles (~6µm) can

be removed from gases by cyclones.

FIGURE 7.3 GRADE EFFICIENCY CURVES OF GAS-SOLID SEPARATION DEVICES VERSUS PARTICLE SIZE (µm)

CYCLONES REMOVE 50% OF THE PARTICLES (6µm) [68]

Successful design and separation efficiency of micron particulates from gases depends on the

following factors [66]

1. Physical factors of the powder

Particle size: it is suggested to undertake a detailed particle size distribution

analysis of the polymer powder because it will determine the behaviour of the

particles under certain conditions. Analysis would involve a “direct-, indirect-

and a classification methods” [69]. Direct technique will estimate the size of

individual alginate particles, whereas indirect system will determine the

chemical characteristic of the particles. It is important to mention, there has not

been established a polymer texture classification standard for alginate powder.

Particle Shape has the greatest impact on the separation efficiency.

Density

2. Chemical composition of the powder

Determination of molecular weight (Mw): Mw has an influence on the viscosity

of alginate solution. Mw can be determined by measuring the intrinsic viscosity of the

polymer solution [71].


TIMEA GREGO 87

To summarise, the solid-gas purification process presented might be used as an alternative

purification method for micron particulates removal of alginate, and for the production of

large scale purified alginate for clinical use. The system utilises gas thus there is a possibility

that physical composition of Na-alginate will not be altered significantly. However, further

study needs to be undertaken to validate the above stated therefore, the following process

stages are suggested

Stage 1. Detailed physical and chemical characterisation of Na-alginate and a feasibility study

of the proposed purification method,

Stage 2. System Design in accordance with Stage 1,

Stage 3. A laboratory scale System Architecture based on Stage 1-2,

Stage 4. Cyclone separation of suspended Na-alginate powder in gas and particulates removal

between 1-6µm,

Stage 5. Sterilisation of purified alginate before cell encapsulation,

Stage 6. Preparation of purified Na-alginate solution for cells encapsulation and viscosity

measurement,

Stage 7. Viability analysis of HepG2 cells.

The length of the dissertation is: 15,885 words

(exc. of references, table of contents, index, acknowledgements and appendices)


TIMEA GREGO

REFERENCES

1. Wong, M. (2004) „Alginates in Tissue Engineering‟. Methods in Molecular

Biology. Biopolymer Methods in Tissue Engineering. vol. 238, pp 77-86. Human

Press

2. Dusseault, J., Tam, K. S., Ménard, M., Polizu, S., Jourdan, G., Yahia, L‟H., Hallé

J-P. „Evaluation of alginate purification methods: Effect on polyphenols, endotoxin,

and protein contamination‟ (2005) DOI: 10.1002/jbm.a.30541

3. http://www.statistics.gov.uk/downloads/theme_health/HSQ41.pdf

viewed on 20th

June 2009

4. Bronzino, J. D. „Chapter 133 Liver Support System‟ The Biomedical Engineering

Handbook 2nd Edition, CRC Press ISBN 0-8493-8346-3

5. Rozga, J. (2006). „Liver support technology – an update‟. Xenotransplantation’.

Vol 13, pp 380-389.

6. Damelin, L. H., Coward, S., Kirwan, M., Collins, P., Selden, C., and Hodgson, H.

J. F. (2007) „Fat-Loaded HepG2 Spheroids Exhibit Enhanced Protection From Pro-

Oxidant and Cytokine Induced Damage‟. Journal of Cellular Biochemmistry.

101:723-734.

7. Selden, C. (2009) A clinical scale bioartificial liver machine to treat patients with

liver failure. Handouts (unpublished)

8. Jones, J. The removal of toxic lactate levels using Immobilised lactate oxidase

within a BioArtificial Liver.

9. 21CFR§184.1724

http://law.justia.com/us/cfr/title21/21-3.0.1.1.14.2.1.169.html

http://www.statistics.gov.uk/downloads/theme_health/HSQ41.pdf

http://law.justia.com/us/cfr/title21/21-3.0.1.1.14.2.1.169.html


TIMEA GREGO

viewed on 26th

July 2009

10. Dornish, M., Kaplan, D., and Skaugrud, O. „Standard and Guidelines for

Biopolymers in Tissu-Engineered Medical Products.‟ Volume 944 Issue

BIOARTIFICIAL ORGANS III: TISSUE SOURCING, IMMUNOISOLATION,

AND CLINICAL TRIALS, Pages 388-397

11. „Chapter Stabilising and Suspending Agents‟ Martindale: The Complete Drug

Reference Book 33rd

edition. Pharmaceutical Press, May 2002. ISBN-10:

0853694990 and ISBN-13: 978-0853694991

12. „Chapter Sodium Alginate‟ European Pharmacopoeia - Supplement 2000, 3rd

edition. ISBN 92-871-3881-8

13. Sodium alginate

http://www.fao.org/docrep/W6355E/w6355e0x.htm

viewed on 26th

July 2009

14. van de Ven, W. J. C., van‟t Sant, K., Pünt, I. G. M., Zwinjnenburg, A.,

Kemperman, A. J. B., van der Meer, W. G. J., Wessling, M. „Hollow fiber dead-end

ultrafiltration: Influence of ionic environment on filtration of alginates. (2008)

Journal of Membrance Science pp 218-229

15. Compendium of Food Additive Specifications. Addendum 5. (FAO Food and

Nutrition Paper - 52 Add. 5) 49th session, Rome, 17-26 June 1997,WORLD

HEALTH ORGANIZATION,

http://www.fao.org/docrep/W6355E/w6355e0x.htm#sodium alginate

viewed on 26th

July 2009

16. Zeta Plus Maximizer Dual-Zone Depth Filters

http://www.tcproducts.pl/zeta%20plus%20maximizer.pdf

viewed on 3rd

August 2009

http://www3.interscience.wiley.com/journal/120754551/issue



http://www.fao.org/docrep/W6355E/w6355e0x.htm

http://www.tcproducts.pl/zeta%20plus%20maximizer.pdf


TIMEA GREGO

17. Prokop, A. and Wang, T. G. „Purification of polymers used for fabrication of an

Immunoisolation Barrier‟ Annals New York Academy of Sciences

18. Vos, P., De Haan, B. J., Wolters, G. H. J., Strubbe, J. H., Van Schilfgaarde, R.

„Improved biocompatibility but limited graft survival after purification of alginate

for microencapsulation of pancreatic islets‟ Diabetologia (1997) 40: 262-270

19. Zimmermann, U. Et al. Production of purified alginates suitable for use in

immunoisolated transplantation Applied Microbiology and Biotechnology (1994)

40: 638-643.

20. Sharma, A. and Gupta, M. N. „Three phase partitioning of carbonhydrate polymers:

separation and purification of alginates‟ (2001) Carbonhydrate polymers 48 (2002)

pp 391-395.

21. Prof Robinson, K., Prof Smith, A., James, A. And Whyndham, M. (2007) GS06

System Engineering Management handouts.

22. Draget, K. I. „Chapter 22 Alginates‟ Handbook of hydrocolloids‟ Woodhead

Publishing, 2000 ISBN 1855735016, 9781855735019

http://docencia.izt.uam.mx/epa/archivos/quimalim/alginato.pdf

Downloaded 15th

June 2009

23. Gacesa, P. „Alginates‟ (1988) Carbonhydrate Polymers 8 pp 161-182

24. http://www.sb-roscoff.fr/Phyto/Science_en_Fete_2002/Laminaria.jpg

25. Tam, S. K., Dusseault, J., Polizu, S., Menard, M., Halle, J-P., and Yahia, L‟H.

„Impact of residual contamination on the biofunctional properties of purified

alginates used for cell encapsulation‟ (2006) Biomaterials 27 pp 1296-1305.

26. Hollinger, G., An Introduction to biomaterials (2006) Taylor & Francis Group,

LLC USA

http://docencia.izt.uam.mx/epa/archivos/quimalim/alginato.pdf

http://www.sb-roscoff.fr/Phyto/Science_en_Fete_2002/Laminaria.jpg


TIMEA GREGO

27. Dr. Draget, K. I., Prof. Dr. Smidsrod, O., and Dr. Skjak-Braek, G. „Chapter 1

Alginate from Algae‟ Polysaccahrids and Polyamides in the Food Industry.

Properties, Production and Patents. ISBN 3-527-31345-1.

28. Smidsrod, O. And Skjak-Braek, G. Alginate as immobilization matrix for cells.

Trends in Biotechnology (March 1990) Vol. 8. pp. 71-78.

29. Khotimchenko, Yu. S., Kovalev V. V., Savchenko O. V., and Ziganshina, O. A.

„Physical–Chemical Properties, Physiological Activity, and Usage of Alginates, the

Polysaccharides of Brown Algae‟ Russian Journal of Marine Biology, Vol. 27,

Suppl. 1, 2001, pp. S53–S64.

30. Alginates by FMC BioPolymer

www.fmcbiopolymer.com/Portals/bio/Content/Docs/.../Alginates.pdf

viewed on 5th

July 2009

31. Dumitriu, S. Polymeric biomaterials ISBN -0-8247-0569-6

http://books.google.com/books?id=iCm1SJBDZwkC&pg=PA172&lpg=PA172&dq=impurities+are

+there+in+alginate&source=bl&ots=0YnXcMGrzB&sig=Emi1nskj4BKY0OnutCHDTraMIDo&hl=

en&ei=FtsuSvfGOJPLjAfZ08yVCw&sa=X&oi=book_result&ct=result&resnum=8

32. HASSAN, R. M. „Alginate polyelectrolyte ionotropic gels‟ Part II Kinetics and

mechanism of exchange of chelated nickel (11) by hydrogen ions in capillary

ionotropic nickel alginate polymembrane gel complex

Journal of Materials Science 28 (1993) 384-388

33. Finch, C. A. Industrial water soluble polymers, The Royal Society of Chemistry

1996.

http://www.fmcbiopolymer.com/Portals/bio/Content/Docs/.../Alginates.pdf

http://books.google.com/books?id=iCm1SJBDZwkC&pg=PA172&lpg=PA172&dq=impurities+are+there+in+alginate&source=bl&ots=0YnXcMGrzB&sig=Emi1nskj4BKY0OnutCHDTraMIDo&hl=en&ei=FtsuSvfGOJPLjAfZ08yVCw&sa=X&oi=book_result&ct=result&resnum




TIMEA GREGO

34. Hoek C., Mann D. G., and Jahns H. M. Algae: An introduction to phycology.

http://books.google.com/books?id=xuUoiFesSHMC&pg=PA275&lpg=PA275&dq

=Algae+By+Christiaan+Hoek,+D.+G.+Mann,+Hans+Martin+Jahns+on-

line&source=bl&ots=WTCBiTY0R7&sig=nzqoc0ln8QlyCze2sXqyvn-

4fwY&hl=en&ei=3KZ0Ss30McPOjAeJyuinBg&sa=X&oi=book_result&ct=result

&resnum=1#v=onepage&q=&f=false

35. http://www.lsbu.ac.uk/water/hyalg.html downloaded on 25th July 2009

36. Wang, W. (2000) „Review on Lyophilisation and development of solid protein

pharmaceuticals‟. International Journal of Pharmaceutics 203. Pp 1-60.

37. Qasim, S. R., Motley, E. M., Zhu, G. (2000) „Filtration‟ in Water Works

Engineering: Planning, Design and Operation. pp. 355-423. Upper Saddle River:

Prentice Hall PTR

38. Ives, K. J. And Sholji, I. (1965) „Research on variables affecting filtration‟. Journal

of the Sanitary Engineering Division Proceedings of the American Society of Civil

Engineering. Vol. 91, No. SA4, Part 1.

39. Ives, K. J. (1990) Testing of filter media. Journal of Water SRT-Aqua. Vol. 39, No.

3, pp. 144-151

40. Tchobanoglous, G., Franklin, L. B., and Stensel, H. D. (2003) „Wastewater

Engineering: Treatment and reuse‟. Metcalf & Eddy, Inc. - 4th

ed.

41. Kézdi, Á. (1969) Talajmechanika (Soil mechanics) I. Harmadik kiadás,

Tankönyvkiadó, Budapest

42. Ives, K. J. (1975) The Scientific Basis of Filtration. Noordhoff International

Publishing. ISBN 90 286 0523 1

http://books.google.com/books?id=xuUoiFesSHMC&pg=PA275&lpg=PA275&dq=Algae+By+Christiaan+Hoek,+D.+G.+Mann,+Hans+Martin+Jahns+on-line&source=bl&ots=WTCBiTY0R7&sig=nzqoc0ln8QlyCze2sXqyvn-4fwY&hl=en&ei=3KZ0Ss30McPOjAeJyuinBg&sa=X&oi=book_result&ct=result&resnum=1#v=onepage&q=&f=false





http://www.lsbu.ac.uk/water/hyalg.html


TIMEA GREGO

43. Picture of Darcy‟s Law

http://uwp.edu/~li/geol460-00/images2/fig1.gif

44. Grego, T. (2009). Laboratory report on Granular hydraulics

45. Nazaroff, W.W. and Alvarez-Cohen, L. Environmental Engineering Science,

U.S.A., John Wiley&Sons, Inc. (2001)

46. Campos, L. C. (2009) Lecture notes on Filtration. Water and wastewater treatment

process.

47. Huisman L, Wood W. E. (1974) Slow sand filtration. Geneva: World Health

Organisation.

www.who.int/entity/water_sanitation_health/publications/ssf3.pdf

downloaded on 5th

July 2009

48. Reyolds, T. D. and and Richards, P. A. „Unit Operations and Processes in

Environmental Engineering‟. Boston, PWS Publishing Company

49. Campos, L. C. (2009) Lecture notes on Sedimentation. Water and wastewater

treatment processes.

50. Prof. Gregory, J. (2009) Lecture notes on Coagulation and flocculation.

51. Ives, K. (1970) „Review Paper on Rapid Filtration‟. Water Research. Vol. 4, pp.

210-223.

52. Qasim, S. R., Motley, E. M., Zhu, G. (2000) „Coagulation, Flocculation, and

Precipitation‟ in Water Works Engineering: Planning, Design and Operation. pp.

229-300. Upper Saddle River: Prentice Hall PTR

53. Microfiltration and ultrafiltration by ACTEW Corporation Ltd (221 London Circuit

Canberra City ACT 2601)

http://www.actew.com.au/water2water/documents/MicrofiltrationandUltrafiltration.pdf

http://uwp.edu/~li/geol460-00/images2/fig1.gif


TIMEA GREGO

downloaded and viewed on 6th

August 2009

54. EVERFILT® - "The Filtration People”. Sand filters.

http://www.everfilt.com/sand_filters.htm

viewed on 27th

June 2009

55. Sabra, W. and Deckwer, W-D. (2004) „Alginate-A Polysaccharide of Industrial

Interest and Diverse Biological Functions‟ in Dumitriu, S. Polysaccharides:

Structural Diversity and Functional Versatility.pp 515-535. 2nd

Edition. CRC Press.

http://books.google.co.uk/books?id=kvUTPxPbkowC&pg=PA518&lpg=PA518&d

q=alginate+pseudoplastic&source=bl&ots=M3P2t59_ta&sig=bYDAUkYOkOUIT

_FS_5C9yEPra1Q&hl=en&ei=NkR3Su2nGozUjAf7jemnBg&sa=X&oi=book_resu

lt&ct=result&resnum=4#v=onepage&q=alginate%20pseudoplastic&f=false

56. Báder, I. (1998). „Nem–Newtoni folyadékok viszkózus sajátságainak

meghatározása rotációs viszkoziméterrel‟. Fizikai-Kémiai Laboratóriumi

gyakorlatok. Kohó- és Bányamérnök-Hallgatók részére. Miskolci Egyetemi Kiadó

57. Wilkinson, W. L. (1960) Non-Newtonian Fluids: Fluid mechanics, mixing and heat

transfer. Pergamon Press Ltd. London.

58. Standard Operating Procedure for making 2% HEPES buffered alginate solutions.

Version No. and date: CFH/08/S05/00, 20/08/2008. Centre for Hepatology at the

Royal Free University College London (UCL) - Royal Free Hospital Campus

59. Millipore. Milli-Q(R)

Integral System

http://www.millipore.com/publications.nsf/a73664f9f981af8c852569b9005b4eee/3

bea69d1793a3484852573ee0063b349/$FILE/pb1008en00.pdf

downloaded on 5th

June 2009

60. Cancela, M. A., Álvarez, E., Maceiras, R. (2003) „Polymers in alimentary industry:

Properties of the sodium alginate‟. Electronic Journal of Environmental,

Agricultural and Food Chemistry. Vol 2 (3), pp 380-387.


TIMEA GREGO

61. Franks, F. „Freeze-drying of Pharmaceuticals and Biopharmaceuticals‟.

http://books.google.co.uk/books?id=-

0gkSrMyyNAC&pg=PA53&lpg=PA53&dq=%E2%80%98Freeze-

drying+of+Pharmaceuticals+and+Biopharmaceuticals&source=bl&ots=i-

STZBdSgK&sig=dY1BHmISBJFcVXQ8JDz6jm6MN9Q&hl=en&ei=3mCcSoCW

A5HMjAeY7NjbDQ&sa=X&oi=book_result&ct=result&resnum=1#v=onepage&q

=&f=false

62. Standard Operating Procedure for vital dye staining of alginate-encapsulated cells

for Fluorescence microscopy. Version No. and date: CFH/08/S03/00, 20/02/2008.

Centre for Hepatology at the Royal Free University College London (UCL) - Royal

Free Hospital Campus (unpublished work)

63. Gander, A. (2009) Upgrade Report „Development of a Quality Management

System for a Bioartificial Liver, to Remove Cellular Contaminants for Regulatory

Approval‟. Centre for Hepatology at the Royal Free University College Medical

School, UCL-Royal Free Campus

64. Aysen, A. (2005) Soil Mechanics: Basic Concepts and Engineering Applications.

Taylor & Francis Ltd.

65. Wakeman, R. (2007) „The influence of particle properties on filtration‟. Separation

and Purification Technology 58, pp.234-241.

66. Hoffmann, A. C. and Stein, L. E. (2007) „Gas Cyclones and Swirl Tubes:

Principles, Design, and Operation‟. Springer.

http://books.google.co.uk/books?id=jq0dhRC6X6wC&printsec=frontcover&dq=

Gas+cyclones+and+swirl+tubes+By+Alex+C.+Hoffman,+Louis+E.+Stein#v=one

page&q=&f=false

67. Svarovsky, L. (2000) „Solid-Liquid Separation‟. 4th

edition, Butterworth-

Heinemann, Oxford.

http://books.google.co.uk/books?id=-0gkSrMyyNAC&pg=PA53&lpg=PA53&dq=%E2%80%98Freeze-drying+of+Pharmaceuticals+and+Biopharmaceuticals&source=bl&ots=i-STZBdSgK&sig=dY1BHmISBJFcVXQ8JDz6jm6MN9Q&hl=en&ei=3mCcSoCWA5HMjAeY7NjbDQ&sa=X&oi=book_result&ct=result&resnum=1#v=onepage&q=&f=false






http://books.google.co.uk/books?id=jq0dhRC6X6wC&printsec=frontcover&dq=Gas+cyclones+and+swirl+tubes+By+Alex+C.+Hoffman,+Louis+E.+Stein#v=onepage&q=&f=false




TIMEA GREGO

68. Rhodes, M. (2008) „Introduction to Particle Technology‟. John Wiley&Sons.

69. British Standard 8471:2007. „Guide to particle sizing methods.‟

70. Module 6: Air Pollutants and Control Techniques, Particulate Matter, Control

Techniques.

http://www.epa.gov/apti/bces/module6/matter/control/control.htm

view on 22nd

August 2009

71. Standard Guide for Characterization and Testing of Alginates as Starting Materials

Intended for Use in Biomedical and Tissue-Engineered Medical Products

Application (F 2064)

72. British Standard (BS) 410: 1986 Specification for Test Sieves

http://www.epa.gov/apti/bces/module6/matter/control/control.htm


TIMEA GREGO

APPENDICES


TIMEA GREGO

A. RECENTLY PUBLISHED IN HOUSE PURIFICATION METHODS


TIMEA GREGO

ALGINATE PURIFICATION METHODS

METHODS AUTHOR AND YEAR OF

PUBLICATION

REMOVED IMPURITIES ADVANTAGES DISADVANTAGES

ENDOTOXIN MITOGENIC

IMPURITIES

POLYPHENOL PROTEINS

CHEMICAL PURIFICATIONS

1. FREE FLOW

ELECTROPHORESIS

(FEE) [10]

(SEE FIGURE 3.7.)

Zimmermann U. et al.

(1992)

+ + D/M D/M

- Endotoxin level was

almost zero after

purification

- Alginates with

different ration of

mannuronic acid and

guluronic acid can be

treated

- Expensive

equipments are

required,

- Extremely long process,

- Not suitable for large

scale purified alginate

production.

- Viscosity of the

purified alginate was

changed

- Fibrotic overgrowth

around micro beads

2. ‘MULTI – STEP’

PURIFICATION

PROCESS [9]

Vos, P., De Haan, B.

J., Wolters, G. H. J.,

Strubbe, J. H., Van

Schilfgaarde, R.

(1997)

„Improved

biocompatibility but

+ D/M + +

- Before and after the

purification process,

there was no

mannuronic- and

guluronic acid content

changes observed

- Improved the

- Alginate beads had

fibrotic overgrowth

between 10-30%

- Limited graft function

as a consequence of

fibrotic tissues.

- Proteins content did


TIMEA GREGO

limited graft survival

after purification of

alginate for

microencapsulation of

pancreatic islets‟

Diabetologia (1997)

40: 262-270

biocompatibility of

alginate beads as a

result of endotoxin and

protein level reduction.

- Endotoxin content

was drastically lower

than in the non-

purified alginate. [3]

- Polyphenols content

was much lower

compare to the non-

purified alginate. [3]

not changed

significantly in two

purified alginates


TIMEA GREGO

METHODS AUTHOR AND YEAR OF

PUBLICATION

REMOVED IMPURITIES ADVANTAGES DISADVANTAGES

ENDOTOXIN

(PYROGENS)

[7]

MITOGENIC

IMPURITIES

POLYPHENOL PROTEINS

CHEMICAL PURIFICATIONS

3. SIMPLIFIED

‘MULTI – STEP’

PURIFICATION

PROCESS [7]

Prokop, A. and Wang,

T. G. „Purification of

polymers used for

fabrication of an

Immunoisolation

Barrier‟ Annals New

York Academy of

Sciences

+ + + [3] + [3]

- Endotoxin removal

by charcoal

- after purification,

the endotoxin levels

were still too high

(ASTM [1] guide stated

endotoxin level should

be below 5 EU/kg of

body weight)

4. THREE PHASE

PARTITIONING

METHOD (TPP)

Sharma, A. and Gupta,

M. N. „Three phase

partitioning of

carbonhydrate polymers:

separation and

purification of alginates‟

(2001)

D/M D/M D/M +

- „Elegant technique‟

[15] for elimination of

protein content of

alginate

- „easily scalable‟

[15]

- The test was carried

out on 2 ml alginate

solutions

- There was no

information regarding

the chemical components

of the alginate

- The viscosity of the

non-Newtonian polymer

was not measured. It is

understandable because


TIMEA GREGO

Symbols: (+) can be removed; (-) not suitable method for removal; (D/M) does not mention

of the prepared amount

of alginate.

- Further research is

needed whether this

method is suitable for

large scale alginate

purification.


TIMEA GREGO

B. CAMP’S CURVES OF

2

ed

e

d RandCR

C

VERSUS RE


TIMEA GREGO

C. CHEMICAL ANALYSIS OF RH-70 AND RH-110


TIMEA GREGO

D. GRAIN SIZE DISTRIBUTION ANALYSIS OF RH-70 AND RH-110


TIMEA GREGO

E. TECHNICAL DRAWING OF LARGE SCALE SYSTEM


TIMEA GREGO

F. SPHERICITY AND ANGULARITY GRAPHS


TIMEA GREGO

G. SETTLING VELOCITY OF INDIVIDUAL SAND GRAINS


TIMEA GREGO

H. DETAILED CALCULATION OF HYDRAULIC OF FILTRATION


TIMEA GREGO

I. REDUCED DATA FROM SIEVE ANALYSIS

AND BS 410:1986

SPECIFICATION FOR TEST SIEVES [52]


TIMEA GREGO

MESH NUMBERS


TIMEA GREGO

MESH

NUMBERS

INTERVAL

OF MESH

NUMBERS

BS STANDARD

SIEVE NOMINAL

APERTURES (µm)

[52]

BS

SIEVE

SIZE

[mm]

AVERAGE SIEVE SIZE

(dS)

RH-70 RH-110

22 22-30

710 0.710 0.60

25 600 0.600

30 30-44

500 0.50 0.43

36 425 0.425

44 44-60

355 0.355 0.30

52 300 0.300

60 60-85

250 0.25 0.21 0.21

72 212 0.212

85 85-120

180 0.18 0.15 0.15

100 150 0.15

120 120-170

125 0.125 0.11 0.11

150 106 0.106

170 170-240

90 0.090 0.08 0.08

200 75 0.075

240

240-440

63 0.063

0.05 0.05

300

<63

53 0.053

350 45 0.045

400 38 0.038

440 32 0.032

Geometric mean of average sieve size 0.177 0.04


TIMEA GREGO

J. RESULTS OF THE VISCOSITY ANALYSIS


TIMEA GREGO

K. PARTICLE SIZE MEASUREMENT OF

FILTERED SODIUM ALGINATE SOLUTION

MSc dissertation FINAL Timea Grego

Documents

clinical use timea grego

abstract alginate

dissertation timea grego

timea grego programme

clinical application

house purification methods

tissue engineering

polymer solution