Citation
Quantitative study of mechanisms of bacterial adhesion to surfaces

Material Information

Title:
Quantitative study of mechanisms of bacterial adhesion to surfaces
Creator:
Ruta, Alina Gabriela, 1971-
Publication Date:
Language:
English
Physical Description:
x, 98 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Bacterial adhesion ( jstor )
Biomaterials ( jstor )
Cells ( jstor )
Fibrosis ( jstor )
Infections ( jstor )
Kinetics ( jstor )
Particle interactions ( jstor )
Receptors ( jstor )
Trucks ( jstor )
Chemical Engineering thesis, Ph.D ( lcsh )
Dissertations, Academic -- Chemical Engineering -- UF ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 90-97).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Alina Gabriela Ruta.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
029546358 ( ALEPH )
40154072 ( OCLC )

Downloads

This item has the following downloads:


Full Text










QUANTITATIVE STUDY OF MECHANISMS OF
BACTERIAL ADHESION TO SURFACES











By

ALINA GABRIELA RUTA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998


























Copyright 1998

by

Alina Gabriela Ruta





























This Ph.D. dissertation is dedicated with all my heart to my parents, especially to my
mother who continually reminds me what determination and character are and who gives
me strength through her courage. Without my parents' unfailing love, sacrifice and
guidance I would not be writing this today.













ACKNOWLEDGMENTS


I would like to acknowledge first of all the support of my advisor, Dr. Richard

Dickinson, who has guided me throughout my Ph.D. work and from whom I've learned

what it really means to be a good research scientist. Through his genuine concern, high

motivation and expectation for quality results as well as the rational step approach to any

problem I have learned invaluable lessons.

I also owe much appreciation to our collaborator, Tim Foster from Trinity College

in Dublin, Ireland, for supplying the bacteria used in our experiments as well as for

helpful suggestions. Also I would like to acknowledge the help from my colleagues,

Aaron Clapp and Steve Truesdail, with the EWLS/3DOT technique.

The support of the ERC for Particle Science and Technology at the University of

Florida has been extremely valuable. The group discussions and numerous presentations

given there have been a motivating, inspiring and an invaluable learning experience.

My thanks also go to Mrs. Shirley Kelly, Nancy Krell, and Peggy-Jo Daugherty in

the Chemical Engineering Department at the University of Florida, for their genuine

concern and for putting a smile on my face when I needed it most. I would like to thank

as well the other supporting personnel in the Chemical Engineering Department who

make a lot of things possible, although all too often their help goes unnoticed.













TABLE OF CONTENTS
page

ACKNOW LEDGMENTS .............................................................................................. iv

LIST OF FIGURES....................................................................................................... vii

ABSTRACT................................................................................................................... ix

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

2 BACTERIAL ADHESION.......................................................................................... 4

2-1 Molecular Structure of Bacterial Cell Surfaces ...................................................... 4
2-2 Non-Specific (Colloidal) Mechanisms of Bacterial Adhesion ................................ 6
2-3 Specific M echanisms of Bacterial Adhesion.......................................................... 8
2-4 M ethods Used to M measure Bacterial Adhesion in Vitro.......................................... 9

3 STAPHYLOCOCCUSAUREUS................................................................................. 13

3-1 Structure/Characteristics...................................................................................... 13
3-2. S. aureus in Device-Centered Infections............................................................. 15

4 MEASUREMENT OF REAL-TIME BACTERIAL ATTACHMENT
KINETICS TO Glass and PROTEIN COATED SURFACES.................................. 16

4-1 Apparatus............................................................................................................ 16
4-1-1 Theory........................................................................................................ 17
4-1-2 Design........................................................................................................ 19
4-1-3 PPFC Used in Attachment Studies.............................................................. 21
4-2 Methodology for Measuring Particle Attachment Kinetics................................... 22
4-2-1 Data Acquisition......................................................................................... 22
4-2-2 Data Analysis............................................................................................. 22
4-3 Measurement of the Role of Receptor Length in Bacterial Attachment to
Protein Coated Surfaces....................................................................................... 26
4-3-1 Bacterial M utant Strains ............................................................................. 26
4-3-2 Experimental Protocol................................................................................ 28
4-3-3 Modified Experimental Protocol................................................................. 30
4-3-4 Results Using Initial Protocol in Section 4-3-2 ........................................... 31
4-3-5 Results Using the Modified Experimental Protocol in Section 4-3-3........... 34
4-4 Measurement of the Role of Electrolyte Concentration in Bacterial
Adhesion to Glass Surfaces.................................................................................. 39








5 MATHEMATICAL MODEL FOR PARTICLE TRANSPORT TO THE
BOTTOM ATTACHING SURFACE IN THE PARALLEL PLATE FLOW
CELL ...................................................................................................................... 42

5-1 M odeling Particle Attachm ent Kinetics ............................................................... 42
5-2 M odel Description............................................................................................... 44
5-3 Dim ensionless Equations..................................................................................... 48
5-4 M odel Solution.................................................................................................... 49

6 EVANESCENT WAVE LIGHT SCATTERING COUPLED WITH THREE
DIMENSIONAL OPTICAL TRAPPING................................................................ 56

6-1 Introduction......................................................................................................... 56
6-2 Apparatus / Principle of Operation....................................................................... 57
6-3 Force M easurem ent............................................................................................. 63
6-4 Data Analysis...................................................................................................... 66
6-4-1 Overview .................................................................................................... 66
6-4-2 Calculating Particle Position....................................................................... 66
6-4-3 Calibrating the Optical Trap ....................................................................... 69
6-4-4 Calculating Force and Potential.................................................................. 71
6-5 Prelim inary Results............................................................................................. 72
6-6 Discussion........................................................................................................... 74

7 SUM M ARY AND CON CLU SION S......................................................................... 77

IM AGE AN ALYSIS PROGRAM S............................................................................... 82

LIST OF REFEREN CES............................................................................................... 90

BIOGRAPHICAL SKETCH ......................................................................................... 98













LIST OF FIGURES


Figpag
Figure 2-1. Bacterial cell wall structures.......................................................................... 5

Figure 2-2. DLVO interaction potential ........................................................................... 8

Figure 2-3. A schematic ofreceptor-ligand interactions promoting adhesion................. 10

Figure 3-1. Scanning Electron Micrograph of Staphyloccocus aureus............................ 14

Figure 3-2. Structure of clumping factor........................................................................ 14

Figure 4-1. Schematic of parallel plate flow cell designed "in-house" ............................ 20

Figure 4-2. Schematic of parallel plate flow cell............................................................ 23

Figure 4-3. Schematic of the automated video microscopy system................................. 24

Figure 4-4. Quantification of attachment kinetics........................................................... 25

Figure 4-5. Procedure for calculating kf(x)................................................................... 26

Figure 4-6. S. aureus mutants with variable length of the repeat region used in PPFC
experiments.................................................................................................. 27

Figure 4-7. Effect of receptor length on bacterial attachment kinetics in the PPFC......... 33

Figure 4-8. Comparison of the mechanistic model predictions to measured values of
k ................................................................................................................ 35

Figure 4-9. Mechanistic model results-effect of energy barrier.................................... 36

Figure 4-10. Mechanistic model results-effect of the magnitude of non-specific repulsive
forces ......................................................................................................... 37
Figure 4-11. Plot of k+ vs. receptor length for S. aureus mutants on Fg and skim milk
control surfaces......................................................................................... 38

Figure 4-12. A plot of intrinsic attachment rate constant, k+, vs. electolyte
concentration............................................................................................. 41

Figure 5-1. Illustration of the convective-diffusive model for particle transport in
PPFC ........................................................................................................... 45









Figure 5-2. Plot of the effective attachment rate constant vs. dimensionless downstream
distance for various values of the dimensionless intrinsic attachment rate
constant, a = k+/v ........................................................................................ 52

Figure 5-3. Plot of the dimensionless attachment rate constant as a fimunction of
downstream distance for various values of the the dimensionless flow
param eter, fl ................................................................................................. 53

Figure 5-4. Plot of the dimensionless attachment rate constant as a function of
downstream distance for various values of the the dimensionless boundary
layer thickness, y.......................................................................................... 54

Figure 5-5. Plot of the dimensionless k+ and average kffas a function of shear rate for a S.
aureus mutant with a missing R region......................................................... 55

Figure 6-1. Structural schematic of key components of the EWLS/3DOT apparatus..... 57

Figure 6-2. Schematic of the principle of Evanescent Wave Light Scattering................. 59

Figure 6-3. Evanescent Wave Light Scattering of Staphylococcus aureus...................... 60

Figure 6-4. Schematic of the three-dimensional optical trap........................................... 61

Figure 6-5. Schematic of forces in the three dimensional optical trap............................. 62

Figure 6-6 Histograms for a trapped l/an polystyrene particle....................................... 68

Figure 6-7. Trap calibration and measurements using the EWLS/3DOT technique........ 71

Figure 6-8. Force and potential energy curves for a 1/Pm polystyrene particle interacting
with glass in water....................................................................................... 73













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
QUANTITATIVE STUDY OF MECHANISMS OF
BACTERIAL ADHESION TO SURFACES

By

Alina Gabriela Ruta

August 1998
Chairman: Richard B. Dickinson
Major Department: Chemical Engineering

Biomaterial-associated infections impede the long-term use of implanted and

intravascular devices. Bacterial adhesion to the protein-coated surfaces of implanted

biomaterials and subsequent catastrophic resistant infection are major barriers to

extended use of these devices.

Because of the size (approximately 1 tm in diameter), bacteria can be considered

colloidal particles. Therefore, studies of bacteria-surface interactions have been in terms of

classical colloidal theories and accounted only for the non-specific interactions such as

hydrophobic, Van der Waals and steric stabilization. However, due to bacteria's complex

surface structure and inherent metabolism, specific, receptor-ligand, type interactions are

important in bacterial adhesion. Hence, a comprehensive understanding of the adhesion

mechanisms of bacteria to protein coated surfaces can be obtained only if the dual nature

of the bacteria-surface interactions is considered.








This dissertation describes a novel method for measuring bacterial attachment

kinetics with the aid of automated video microscopy and image analysis in well-defined

laminar flow conditions using a parallel plate flow cell (PPFC). An intrinsic attachment

rate constant is extracted after the experimental data has been corrected for transport

effects in the PPFC via a transport model. This technique has been applied to measure

the role of receptor length in the rate of specific attachment of Staphylococcus aureus to

ligand-coated surfaces.

To further understand bacteria-surface interactions, a novel technique, called

Evanescent Wave Light Scattering coupled with Three Dimensional Optical Trapping

was developed to measure colloidal force interactions between protein coated surfaces

and individual bacteria.

The development of these methodologies for measuring bacterial attachment rate

constants and interaction forces provides new tools for quantitative studies of bacterial

adhesion. This work will serve to enhance fundamental understanding of the specific and

non-specific mechanisms of bacterial adhesion and may ultimately contribute to the

design of infection-resistant biomaterials and novel surface coatings for enhanced

filtration of microorganisms.














CHAPTER 1
INTRODUCTION

Infection of implanted and intravascular devices is potentially life threatening,

and impedes the long-term use of many biomedical devices (Dankert et al. 1986, Hench

1980, Sugarman & Young 1984, Bisno & Waldvogel 1989, Wadstrom & Eliasson 1990).

Approximately 45% of hospital-acquired infections are associated with implanted or

extra corporeal medical devices (Dankert et al. 1986, Stamm 1978). Device-centered

infections result in the impairment of the host defense mechanisms, extreme resistance to

anti-microbial therapies, and, as a result, high mortality rates (Gristina et al. 1987).

Despite this great clinical importance, the pathogenesis of these infections remains poorly

understood and as a consequence efficient means to eradicate biomaterial-centered

infections without removal or exchange of the implant, or prophylactic treatment are

unavailable. Recent reports from the Center for Disease Control indicate that Staph

bacteria cause 13% of the nation's two million infections each year, which kill 60,000 to

80,000 people (Center for Disease Control 1997). Also the same organization reports that

Staphylococcus aureus is becoming increasingly resistant to antibiotic treatment. This

suggests the urgency to progress in understanding of bacterial adhesion mechanisms in

order to devise new biomaterials or procedures that prevent bacterial colonization.

Bacterial adhesion to biomaterial surfaces plays a central role in the pathogenesis

of device-centered infections (Dankert et al. 1986, Gristina 1987, Gristina et al. 1993,

Bisno & Waldvogel 1989, Dickinson & Bisno 1993, Jansen & Peters 1993, Dougherty








1988, Vaudaux et aL 1990). The fundamental physical and molecular mechanisms

governing bacterial adhesion to these surfaces remain poorly understood and their

elucidation could ultimately provide a rational basis for the design of infection-resistant

biomaterials. Control of bacterial adhesion is currently implemented in the form of

surfactant or antibiotic coatings on the biomaterial surface, but these provide only partial,

short-term protection from bacterial adhesion, since these coatings are usually

biodegraded in the body. Both specific (i.e., receptor-mediated interactions--receptors are

molecules on the cell surface capable of binding other molecules on the surface of a

material or another cell, termed ligands) and non-specific mechanisms (colloidal

interactions) may be involved. Therefore, it is equally important to develop a tested

theoretical basis for interpretation of acquired adhesion data and to be able to predict the

probability of adhesion occurring under a given set of local conditions.

The goal of the work described in this thesis is to enhance fundamental

understanding of microbial adhesion mechanisms to surfaces via quantitative

measurements. This understanding may aid in designing infection-resistant biomaterials

or in developing techniques to prevent bacterial infection of implanted intravascular and

biomedical devices. It may also aid in designing surfaces for better microbe removal in

wastewater treatment.

This dissertation describes the development of novel quantitative experimental

techniques for studying microbial attachment to solid surfaces and the application of

these techniques to study the role of specific and non-specific interactions in bacterial

attachment. Chapter 2 provides a brief background on the current state of knowledge

with respect to biology and physics underlying bacterial adhesion mechanisms and the








common measurement techniques used to study microbial adhesion. Chapter 3 discusses

the characteristics of the species used in this study, Staphylococcus aureus. Chapter 4

discusses the methodology used to directly measure real-time bacterial attachment

kinetics under well-defined flow conditions, and the results of a study of the dependence

of bacterial attachment on the length of a fibrinogen-binding molecule, "clumping

factor". Also presented in Chapter 4 are measurements of attachment rate constant as a

function of electrolyte concentration. Chapter 5 presents the derivation of a mathematical

model for convective-diffusive bacteria transport to the collector surface of the parallel-

plate flow chamber, which was used to correct the experimental data for transport effects

and yield an intrinsic attachment rate constant. Chapter 6 discusses a novel technique that

is able to measure the interaction forces between a single bacterium and a surface.

Finally, Chapter 7 contains a summary of the key accomplishments and results,

concluding with recommendations for future work.













CHAPTER 2
BACTERIAL ADHESION

2-1 Molecular Structure of Bacterial Cell Surfaces


Bacteria are classified according to their reaction with the Gram stain; they can be

classified either as Gram negative or Gram positive (Dankert et al. 1986). A schematic of

cell walls of Gram negative and Gram positive bacteria can be seen in Figure 2-1 (a & b).

Gram negative bacteria have their plasma membranes surrounded by a peptidoglycan

layer. The outer membrane, a phospholipid bilayer is anchored into the peptidoglycan

layer by a lipoprotein (Figure 2-1(a)). Capsular polysaccharide and some outer-

membrane proteins can extend from the outer membrane. Flagella, fimbriae or capsules

can also extend from the outer membrane. To be noted is that the peptidoglycan layer is

embedded as a separate layer within the outer membrane of Gram negative bacteria, thus

preventing it from interacting directly with molecules on the surface or in its proximity.

As seen in Figure 2-1(b), Gram positive bacteria have a relatively thick outer wall

consisting of peptidoglycan, teichoic and teichuronic acids as well as other

polysaccharides or proteins. These are not entirely embedded in the cell wall; some can

also be exposed at the surface and can play an important role in generating adhesion to

surfaces.









Capailar Polysacchwide

SCapular Prtein
./V y^ Upopolysaochide
,(91Z -------


Capsule



Cell Wall


. Cytophasmic
Membrane


enzymes,& proteases


Figure 2-1. Bacterial cell wall structures
(a) Cell wall structure of Gram-negative bacteria.
positive bacteria. Source: Dankert et al. 1986.


(b) Cell wall structure of Gram-


Several bacterial species have macromolecules that can bind specifically to

ligands in the cell's environment, called 'adhesins' or 'receptors'. Both specific (i.e.

receptor-mediated) and non-specific (colloidal) interactions may be important in bacterial








adhesion to solid surfaces (Busscher & Weerkamp 1987, Busscher et al 1992). Bacterial

adhesion has traditionally been studied quantitatively in terms of theories developed for

adhesion of colloidal particles (Bell et al. 1984, Rutter & Vincent 1980). The specific,

receptor-mediated, interactions have been traditionally treated separately, largely

qualitatively, from a biochemical point of view (Sullam et al. 1990, Lindon 1986,

Vercelotti et al. 1985, Vaudaux et al. 1984(a), Ryden et al. 1983, Espersen &

Clemmensen 1982). Quantitative studies of bacterial adhesion accounting for both

specific and non-specific interactions have been lacking. Although progress has been

made in identifying molecular components responsible for the adhesion of some cell

types, the role of quantitative physical and molecular parameters in governing bacterial

adhesion to protein-coated surfaces, such as those of implanted biomaterials, remain to be

elucidated. The purpose of this chapter is to discuss both the non-specific and specific

bacterial adhesion mechanisms and some of the techniques that have been used for

measuring bacterial adhesion.



2-2 Non-Specific (Colloidal) Mechanisms of Bacterial Adhesion


Based on size (approximately 1 grm in diameter), bacteria can be considered

colloidal particles. Therefore, until now, physicochemical models for bacterial adhesion

have been primarily based on theories for colloid-surface interactions (Marshall 1985,

Marshall et al. 1971, Ruckenstein 1975, Pethica 1980, Rutter & Vincent 1980).

Classical DLVO theory (derived by Derjaguin, Landau, Verwey, and Overbeek) accounts

for electrostatic and Van der Waals forces between the cell and a surface to predict the

interaction potential, O(z), as a function of separation distance, z (Derajaguin & Landau








1941, Verway & Overbeek 1948). A hypothetical DLVO interaction profile, O(z), for a

particle and surface of like charge is shown in Figure 2-2. Several features of this plot are

noteworthy. At large distances (>50 nm) the particle is attracted toward the surface by

long-range van der Waals forces and repelled by an electrostatic force created by the

overlap of a double layers of diffuse counter ions near the surfaces. The so-called

secondary minimum (z = z2) is located where these long range electrostatic repulsion

forces balance van der Waals attraction forces, which corresponds to a transient adhesion

state where the particle can typically diffuse out of this region due to Brownian motion. If

the particle has sufficient energy it may be able to surmount the free energy barrier (z =

z.a,) and descend into the primary minimum (z =z,) where it will be considered fully

adherent at a separation distance where close range steric repulsion forces balance van

der Waals attraction forces. One hypothesis is that the rearrangement of macromolecular

surface structures such as binding molecules, fimbriae or extruded polymers ("slime")

occurs during adhesion in the secondary minimum, which promotes the transition over

the energy barrier, resulting in firm attachment in the primary minimum (Pethica 1980).

A limitation of viewing bacterial attachment as deposition of an inert colloidal

particle is that the macromolecular interactions are not explicitly taken into account.

However, work in our group resulted in a recent publication that presents a model for

bacterial attachment that accounts for both colloidal forces and the formation of

macromolecular bonds (Dickinson 1997). This model will be used to help interpret

experimental data in Chapter 4.















Energy Barrier
to Attachment


2 Minimum


125

100

75

p50

^25


-25

-50


300


400


1 Minimum


z [nm]


Figure 2-2. DLVO interaction potential.
A hypothetical plot of potential energy, $z), vs. separation distance, z, as predicted from
the DLVO theory. The primary minimum corresponds to the point of attachment. To
attach, a bacterium must move by Brownian motion from the secondary minimum, over
the energy barrier and into the primary minimum.




2-3 Specific Mechanisms of Bacterial Adhesion

A biomaterial exposed to a biological environment rapidly acquires a protein

conditioning film that can affect surface properties and cell adhesion (Gristina et al. 1993,

Baier et al. 1984). Attachment can result from specific binding of molecular receptors


200


........... Electrostatic
-- -. van der Waals


500








(a.k.a. adhesins) on the cell surface to complimentary ligands on the biomaterial surface.

As shown schematically in Figure 2-3, to bind to surface-adsorbed ligands, receptors

must extend past repulsive forces, including steric repulsion forces induced by the cell

capsule and cell walL Therefore, the effective length of receptor-ligand interaction is

expected to be a relevant determinant in the ability of the cell to attach to a surface. In

fact many bacteria have specific structures such as fimbriae that allow the cell to attach

by extending over any repulsive barrier to adhesion (Foster & McDevitt 1994).

Staphylococcus aureus is a common opportunistic pathogen involved in device-

centered infections (Dankert et al. 1986). It is believed to adhere to biomaterials via

specific binding of cell-surface receptors to adsorbed host proteins such as fibrinogen

(Fg) (Foster & McDevitt 1994, Ohtomo & Yoshida 1984, Boden & Flock 1989, Cheung

& Fischetti 1991) and fibronectin (Fn) (Kuusela et al. 1985, Toy et al. 1985) on the

biomaterial surface. This will further be described in the next chapter (Chapter 3).




2-4 Methods Used to Measure Bacterial Adhesion in Vitro


Two aspects of bacterial adhesion are typically studied: the number of adhered

bacteria deposited from suspension, and measurement of the force required to dislodge a

cell from a surface (Dankert et al. 1986). The first aspect is more relevant here because

the focus of this is primarily on the interactions involved in the initial attachment of cells,

rather than the strength of the resulting adhesion.












CWl V"d...


F m rL rLI







Figure 2-3. A schematic of receptor-ligand interactions promoting adhesion
Schematic of the region of interaction between a bacterium and a surface governed by
specific receptor-ligand interactions. Cell surface receptors extend through the cell wall
to specifically bind to adsorbed host proteins on the biomaterial surface.


The most prevalent technique for measuring the number of adhered cells is simply
incubating a collector in a bacterial suspension. Then, after an arbitrary incubation time,

the number of adhered bacteria is counted directly under a microscope or from

photomicrographs, or indirectly by measuring the chemoluminescence (Ewetz &
Strangert 1974), fluorescence (Donkersloot 1972), or radioactivity (Sjollema 1989).

However, these are often not sensitive enough, or, depend on unstable characteristics of
the cells. These techniques also generally suffer from ill-defined flow conditions which

can affect the transport rate of bacteria to the surface as well as influence the probability

of a cell attaching to the surface (Dickinson & Cooper 1995). Furthermore, 'blind'
assays in which cell attachment is not directly observed cannot account for secondary

effects such as cell-to-cell interactions or cell agglomeration in suspension or on the








surface. These limitations cloud the significance of the experimental results when the

primary interest is to quantify the interaction of the cell with the substratum.

To overcome these limitations, some recent studies have used parallel plate flow

chambers (PPFC's) to study cell attachment or detachment (Wilkinson et al. 1984,

Sjollema 1989, Lawrence et aL 1990, Jen & Lin 1991, van Kooten et aL 1992). Others

who studied shear dependent cell adhesion (Fowler & McKay 1980, Crouch et al. 1985,

Cozens-Roberts et al. 1990b, Dickinson & Cooper 1995) have used radial-flow chambers.

Flow chambers offer the advantages of well-defined flow conditions and the possibility

of real-time direct observation and enumeration of the number of attached cells. This

allows the measurement of the kinetics of cell attachment or detachment. However, in

previous PPFC studies, the actual cell concentration near the surface was not known, and

it could vary as a function of the downstream distance in the chamber (Dickinson &

Cooper 1995). This implies that the measured attachment rate depended on the position in

the chamber chosen for observation, as well as on the flow rate. For example, it was

observed (Dickinson & Cooper 1995) that rapid attachment of particles, can lead to

downstream depletion of cells near the surface, and conversely, for slow attachment and

sedimentation can lead to a downstream accumulation of cells near the surface.

Therefore, the measured effective attachment rate constant is dependent on the

downstream distance in the PPFC, therefore is not an intrinsic measure of the probability

of attachment which depends only on local conditions such as interaction between the

cell and the surface. To estimate an intrinsic rate constant of attachment, the rate of

attachment must be measured as a function of position in the PPFC, and the variable








transport of cells to the surface must be taken into account. A methodology and analysis

to accomplish this will further be discussed in Chapter 4.

Only recently have investigators focused on directly measuring the surface forces

between bacteria and solid substrata. The novel technique described in Chapter 6, can be

used to directly measure the long range interaction forces between bacteria and an

underlying surface, which are relevant as the cell approaches the surface in the process of

attachment. It is a considerable improvement over existent techniques and it will allow

the comparison with the bacterial adhesion measurements based on enumeration of

adherent bacteria that were performed by using a PPFC. This combination of techniques

used to study bacterial adhesion will enhance understanding of the bacterial adhesion

mechanisms occurring in vitro for conditions relevant to in vivo bacterial adhesion to

biomaterial surfaces.














CHAPTER 3
STAPHYLOCOCCUS A UREUS




3-1 Structure/Characteristics


Staphylococcus aureus is a spherical Gram-positive pathogen that expresses cell
surface binding molecules ('receptors') for plasma and extracellular matrix components.
It is a spherical microorganism with an approximate diameter of one micron. Important to
keep in mind for the attachment kinetics experiments that are performed (Chapter 4)
using this microorganism is that it retains its spherical morphology upon adhesion to a
surface, as it can be observed from Figure 3-1.
An important fibrinogen receptor on the surface of S. aureus, called "clumping

factor", has been extensively characterized (McDevitt et al. 1994). The important features

of its structure are shown schematically in Figure 3-2. "Clumping factor" is a 92 kDa

protein consisting of three distinct domains: a membrane and cell wall spanning domain,

a serine-aspartic acid dipeptide repeat domain (also called "the R region"), and the

fibrinogen-binding domain at the end of the molecule. The dipeptide repeat region is of

308 amino acid residues in length (McDevitt & Foster 1995). Assuming an a-helical

structure the length of this region is estimated to be 54 nm. It has been hypothesized that

the hydrophilic R-region extends the fibrinogen-binding domain to promote its ability to

interact with its environment. By genetically engineering mutants with R-regions of

reduced length it was recently shown that adhesion depends on the length of the R-region

(Hartford et al. 1997). As discussed in the next chapter, we have demonstrated under









well-defined flow conditions, that the length of the R-region promotes adhesion by

promoting an enhanced rate of attachment, which further suggests that the function of this

region allows the cell to "reach" across the energy barrier.


Figure 3-1. Scanning Electron Micrograph ofStaphyloccocus aureus




Fibrinogen-binding Protein
clfA "Clumping Factor"


Membrane and
Cell Wall-Spanning Ser-Asp Repeat Domain
Domains .


Fibrinogen-Binding
Domain


"Stalk Region"


Figure 3-2. Structure of clumping factor


IIIII!III












3-2. S. aureus in Device-Centered Infections


Staphylococcus aureus (S. aureus) was used in this study because it is a highly

pathogenic bacterial strain causing 45% of the hospital acquired infections and it is

extremely resistant to antibiotic treatment. S. aureus is a predominant pathogen in

generating antibiotic-resistant as well as device-centered infections (Waldvogel 1995).

S. aureus expresses several specific cell surface binding molecules that interact

with plasma proteins commonly found adsorbed to biomaterials, such as fibrinogen

(Cheung & Fischetti 1990, Herrmann et al. 1993, Herrmann et al. 1988, McDevitt et al.

1994), fibronectin (Herrmann et al.1988, Vaudaux et al. 1984(a), 1984(b), 1989, Vaudaux

1994), collagen, vitronectin, laminin, thrombospondin, bone sialoprotein, elastin, etc.

However, studies performed strongly indicate that fibrinogen is the most active plasma

component in promoting an initial in vivo adhesion of S. aureus (Francois et al. 1996).

As mentioned, S. aureus has been shown to attach to fibrinogen-coated surfaces via

"clumping factor" (McDevitt et al. 1994, Dickinson et al 1995). The prevention of

bacterial attachment and onset of infection are highly dependent on the initial, short-term

contact of the biomaterial in the body. Therefore, the focus of this study is on the

molecular interactions and parameters relevant at a short time scale after the

implantation. Specifically, our focus is on the role of specific and non-specific

interactions involved in the initial rate (or, equivalently, probability) of attachment to

surfaces.














CHAPTER 4
MEASUREMENT OF REAL-TIME BACTERIAL ATTACHMENT KINETICS TO
GLASS AND PROTEIN COATED SURFACES




4-1 Apparatus


The role of the cell surface properties, fluid and substrata in promoting adhesion

can be quantified by measuring bacterial attachment kinetics to different coatings on a

glass substrate under well-defined flow conditions. The principal tool involved in this

study is a parallel plate flow cell (PPFC) which provides laminar, well defined flow

conditions for measuring the number of bacteria attaching to surfaces. As it was

discussed in Section 2-3, the possibility of direct observation of attachment kinetics in the

PPFC offers advantages over other blind assays for the measurement of bacterial

adhesion.

The aim of this chapter is to present a detailed description of a PPFC, including

the principles involved in its design and testing, the experimental protocol for automated

enumeration of the adherent bacteria, and the procedure for data analysis. A

methodology is also described for the data analysis that offers an improvement over

traditional approaches used in interpreting the attachment kinetics results from PPFC

experiments. Finally, the experimental results are provided as a function of the length of

the clumping factor molecule. They are obtained from an application of this methodology








to the study ofS. aureus attachment to fibrinogen-coated surfaces. In a second study the

role of electrolyte on attachment kinetics of S. aureus to bare glass surfaces is also

described.




4-1-1 Theory


Flow systems are employed in adhesion studies because they) allow the control of

hydrodynamic conditions in terms of shear rate, flow velocity, and Reynolds number

(Sjollema 1989). Convection also affects transport of bacteria to the collector surface.

The advantage of using a PPFC is that it allows real-time direct observation of the surface

where particles attach under well-defined laminar flow conditions.

Flow in the PPFC must be steady, laminar and fully developed in order to

minimize or eliminate any entrance effects (van Wagenen & Andrade 1980, Bowen 1985,

van Kooten et al. 1992). The design of the PPFC is similar to previous designs (van

Kooten et al. 1992, Sjollema 1989), but modified to fit within the in-house automated

video microscopy system. The parameter that needs to be chosen appropriately to assure

minimal entrance effects is the characteristic entry length, Le, defined as follows in

equation 4-1.

Le = a- h. Re (4-1)
Where: a = proportionality constant

h = channel height (cm)

Re = Reynolds number








The constant, a, is reported to be between 0.013 (van Wagenen & Andrade 1980), and

0.044 (Bowen 1985). In designing the PPFC described in the next section, an average,

0.0285, of the two values was used.

The Reynolds number in the PPFC is defined as:

p.Q
Re h (4-2)
(w +h) /
where: p = fluid density (g/cm 3)

U = viscosity (g/cm-s)

w = channel width (cm)

Q = volumetric flow rate (cm3 I/s)

For a uniform flow profile throughout the entire plate length, Le must be small

compared to the length of the chamber. The criterion is Le < b -l1, where b is a

proportionality constant and I is the chamber length. Based on the above equation, Van

Wagenen and Andrade postulated that Le should be less than 10% of the chamber length

in order to have minimal entrance effects, i.e., b = 0.10 (van Wagenen & Andrade 1980).

From the fluid velocity profile (deferred to Chapter 5), the shear stress on the

surface of the PPFC is given by:

Shear stress = 6Qu / wh2 (dyn/cm2) (4-3)
For a typical experiment the desired shear stress is determined, and the appropriate flow

rate can be calculated from the known fluid and chamber parameters.








4-1-2 Design


Initially, a PPFC was designed and constructed in-house based on the theory

described above and on an existing design (van Kooten et al. 1992). However, some

additions and improvements were made over the designs described in the literature. This

design can be seen in Figure 4-1. The PPFC consists of two parts: The bottom part was

machined out of brass and was plated with solid nickel containing 10% phosphorus. The

coat thickness is between 0.001 0.0012 inches; this prevents corrosion of the bottom

part in the saline operating environment. This part has machined into it an 0-ring channel

that allows complete sealing of the chamber when a glass slide is superimposed on it and

closed with the top Plexiglas part which, in turn has another glass slide fitted in on an 0-

ring channel. The flow channel is formed when 2 stainless steel (0.2mm uniform

thickness) spacers are positioned parallel to each other along the length of the chamber,

and sandwiched between the top and bottom glass slides. The entire assembly is sealed to

become leak-proof by tightening of the screws.

In order to ensure a laminar flow field at the channel entrance, between the two

parallel plates, the depth and width of the of the inlet and outlet channels gradually

decrease and increase, respectively up to the final channel depth (0.02 cm) and width (5

cm).

For proper operation of this device, air was eliminated from the lines before

performing any experiments because entrapped air bubbles could cause stagnation of

flow or back flow in the chamber changing the known flow rate. Liquid was allowed to

flow prior to starting the actual experiment in order to fill up with liquid any remaining

gaps outside the flow area.







The temperature of the flow chamber was controlled using four silicone rubber

heaters attached to the bottom brass part of the chamber. These heaters could be

connected to a temperature controller with a relay to switch the load on and off on the

heaters (Watlow, Sebring, FL).



inlet Plexiglas 'outlet




parallel
plates spacers
(4x7.5 cm) 0 sparsn)
0- ring









Figure 4-1. Schematic of parallel plate flow cell designed "in-house"



The flow inside the PPFC was tested to be laminar by using a rheoscopic fluid

(Kalliroscope Corporation, Groton, Massachusetts). This fluid allows visualization of the

flow pattern in the PPFC. The rheoscopic fluid has been developed initially by an artist,

Paul Matisse, who used these fluids to develop artworks based on the display of

convection currents. The rheoscopic fluid is a suspension of microscopic crystalline

pellets. When in motion, the pellets align their longest dimension parallel to the planes of








shear. In the presence of incident light, areas of varying orientation reflect differing

intensities of light and their movement produces striking visual images of the currents

taking place. Unlike dye based fluid tracers which eventually become uniformly

dispersed within the medium, a rheoscopic fluid never looses its ability to make the flow

patterns visible and can be used many times over.


4-1-3 PPFC Used in Attachment Studies


The actual PPFC used to generate results presented in this thesis was

commercially available from CytoDyne (San Diego, CA), and it is shown in Figure 4-2.

The change to using this slightly different design was made because of decreased time

necessary for set-up as compared to the originally designed chamber. This PPFC consists

of a bottom polycarbonate part that can be readily cleaned between experiments. A

silicone rubber gasket is mounted on the channel machined in the bottom polycarbonate

part. The gasket serves two purposes: it provides a tight seal and leak-proof environment

and at the same time serves as the spacer that creates the flow channel. A 75 mm x 50

mm glass slide with an optically transparent coating can be mounted easily over the

gasket on the bottom part of the PPFC, and vacuum is applied as a means of sealing the

PPFC assembly. This allows test slides to be quickly and easily exchanged, and reduces

the likelihood of spills and leaks.












4-2 Methodology for Measuring Particle Attachment Kinetics



4-2-1 Data Acquisition



An automated video microscopy system (Figure 4-3) was programmed to

recursively scan a central rectangular region of the parallel-plate flow cell. Within this

region, four fields were recursively scanned at each of the ten different downstream

positions, as illustrated in Figure 4-4. Images from each field were captured using a CCD

video camera, and stored in a file on a Pentium 166 computer for further analysis. This

process was automated using an Optimas image analysis program entitled SCAN.MAC,

which is presented in the APPENDIX.



4-2-2 Data Analysis


A second Optimas program recursively opened and processed the images to count

the number of attached cells in each, and stored these counts to an EXCEL file for further

analysis and for calculation of an attachment rate constant. The image analysis program

used in this step is called BREPLAY.MAC and it also appears in the APPENDIX.

Data were analyzed in the following manner: first, the cell density in each field

was obtained by dividing the number of attached particles in each field by the area of the

field (automatically recorded by SCAN.MAC). Then, the particle density was plotted

versus time for each downstream position (Figure 4-5).









Glass slide
(75x 50 mm) .. ~


Spacergasket '
(200o m)

Vacuum
channel rFow direction

< > Plexiglass ~C \4
/ Pg Vacuum

Inlet Outlet




Figure 4-2. Schematic of parallel plate flow cell
This PPFC was used in the adhesion experiments whose results are presented in this
dissertation.


The slope of the resulting straight line was obtained by linear regression. Dividing this

slope by the initial concentration of particles, Co, yielded an effective attachment rate

constant, kej, for each field. The values from the four fields at each downstream position

were averaged to provide ke(x).
As discussed in more detail in Chapter 5, keff is generally a function of the
downstream distance in the PPFC, because rapid attachment can lead to a downstream
depletion of cells near the surface, and slow attachment can lead to downstream
accumulation due to sedimentation. In order to correct for this effect of particle transport

on the attachment in the PPFC and to determine an intrinsic attachment rate constant, k+,











Inverted
Microscope
Stage ._
trizd Controller ] -\

Flow Motorized \
Flow StageII J
Apparatus Sa A _j -

^ -\ f _!!Monitor I l1111l

Pentium 166 MHz



Figure 4-3. Schematic of the automated video microscopy system
The system is composed of an inverted microscope equipped with an x-y motorized stage
and z-focus control through the microscope objective. The PPFC is mounted on the
motorized stage and the computer is programmed to automatically move the stage at
certain positions and acquire images of particles adhering to the bottom plate in the
PPFC. These images are stored on the computer hard drive for further analysis.


that is independent of transport effects, a mathematical model was developed for

transport and deposition of colloidal particles in the PPFC. This model accounts for cell

diffusion, convection, and sedimentation to calculate the expected particle concentration

at the surface for a given flow rate and value of k+. The derivation and numerical solution

of the model is detailed in Chapter 5. By fitting the mathematical model predictions to the

experimentally obtained values of kejf(x) by nonlinear least-squares regression, an

estimate of k+ was obtained. The value of this procedure is that the intrinsic attachment

rate constant is independent of any global parameters in the system, such as the

downstream distance in the PPFC.








Parallel-Plate Flow Cell
x


/Outlet


Real-time Measurement
Automated Image
Analysis and Data
Acquisition


Figure 4-4. Quantification of attachment kinetics
The schematic illustrates the procedure for quantifying attachment kinetics in the PPFC.
The marked fields are scanned in a recursive fashion and images of particles adherent to
the surface at those particular fields are stored in the computer memory. From such a
procedure a cell density as a function of downstream position in the PPFC can be
determined.




k+ physically reflects the ability of the particle near the surface to overcome any
separating energy barrier and attach to the surface. In the interest of continuity of the
material presented here, the detailed model description is postponed until Chapter 5.


/
Inlet









Cell Density


0 500 1000 1500
Time, t (seconds)


Figure 4-5. Procedure for calculating ke(x)
This is a plot of cell density vs. time for two of the S. aureus mutants (R=0 amino acid
residues, and R=307 amino acid residues) at a particular field on the surface of the PPFC.
The slope of these lines were obtained by linear regression. Dividing the slope by the
initial particle concentration present in suspension, Co, the effective attachment rate
constant was obtained.




4-3 Measurement of the Role of Receptor Length in Bacterial Attachment to Protein


Coated Surfaces


4-3-1 Bacterial Mutant Strains


To study the role of the receptor length in bacterial adhesion, genetically

engineered mutants, procured from a collaborating group at Trinity College in Ireland,

were used in the experiments (Hartford et al. 1997). The S. aureus mutants with varying

length of the serine-aspartic acid (Ser-Asp) repeat domain of the fibrinogen receptor


vs. Time








(clumping factor -clfA) were genetically engineered by plasmid insertion of the modified

clfA gene into a fibrinogen deficient mutant of the bacterium and cloned (Hartford et al.

1997). The cloned fibrinogen receptor (clumping factor, clfA) was that of S. aureus

Newman strain.



Membrane and Fibrinogen-Binding
Cell Wall-Spanning Ser-Asp Repeat Domain Domain
Domains Ser-Asp Repeat Domain Doma
pCF54
'0 307

-""'C pCF78
o 157
-- pCF81
6 133
I i"-- pCF83
0 39

0_C pCF84
o 21
0- -- D pCF85


S pCF56
0
-m--- A2




Figure 4-6. S. aureus mutants with variable length of the repeat region used in PPFC
experiments



The mutants were stored in a -70C freezer and each individual vial was defrosted

only for the short period necessary for the transfer of an innoculumn to tryptic soy broth








for overnight growth. The S. aureus mutants used in experiments were grown overnight

in 8 ml of tryptic soy broth (TSB) to which 5 microliters of chloramphenicol were added.

Normally the ratio ofchloramphenicol to TSB was l l per ml of TSB. All mutant strains,

except the clumping factor negative mutant, A2, were grown in TSB with

chloramphenicol.


4-3-2 Experimental Protocol



The first step of an experiment was the preparation of the phosphate buffer saline

(PBS) with double distilled water and a commercial phosphate buffer dry mixture

(Dubelco's PBS). The buffered media was supplemented with 0.1mg each magnesium

and calcium chloride per ml of solution (at pH 7.35). Therefore, to one liter of double

distilled water there were added, under continuous stirring on a magnetic plate, two

envelopes of powdered Dubelco's PBS and 0.1 g each of magnesium and calcium

chloride. This buffer was the base for the albumin and fibrinogen solutions prepared for

coating of the glass slides (50 mm x 75 mm; Fisher, Pittsburgh, PA). The buffer was

filtered under vacuum through a Fisher Scientific 0.2-micron filter to ensure an initially

particle-free solution.

Prior to their respective coatings, the glass slides were cleaned overnight in a

chromic acid solution, Chromerge, (Fisher, Pittsburgh, PA), and subsequently rinsed with

water, and submerged for two hours in double distilled water.

The next step consisted of coating the pre-cleaned glass slides with the respective

coating solutions by swirling the submerged slides in a Petri dish on top of a Roto Mixer

in the following manner. Slides were incubated either in bovine serum albumin (BSA)








solution (0.5 % by weight) or in a solution of human fibrinogen (0.02mg fibrinogen /ml)

for one hour, then for 1 hour in the BSA solution to block any remaining surface sites

potentially not covered by the fibrinogen. The slides coated with BSA served as a non-

adhesive control surface. This last step ensured that nonspecific interactions between the

bacteria and the glass were eliminated. The albumin and fibrinogen solutions were

prepared using the pre-made buffer, which was filtered before use in the coating process.

Bacteria were grown overnight in tryptic soy broth and prior to each experiment

were first spun down to a pellet in a centrifuge in order to remove the nutrient media.

They were subsequently washed twice in PBS (by spinning down and re-suspending),

and prior to re-suspension in the final media they were sonicated lightly to dissociate

aggregates. Next, bacteria were suspended in 100 ml PBS at an approximate

concentration of 2x107 bacterial cells/ml. Such a low concentration was desirable to

eliminate particle-particle interactions during the deposition experiments. The bacterial

concentration could be accurately determined in an automated fashion by using the

automated video microscopy system coupled with image analysis to count the particles in

a certain region of a Petroff-Hauser counting chamber. The image analysis program used

in this step is called COUNT.MAC and appears in the APPENDIX. The particles were

counted at a magnification of 20X. The bacteria at the determined concentration were

further used in the attachment experiments.

The PPFC was assembled by placing the treated slide over the silicone rubber

gasket. The entire assembly was sealed by applying a vacuum provided by a vacuum

pump. The protein coated slide was placed such that the top-facing surface in the Petri

dish during the coating process was inverted and faced downward toward the flow in the








PPFC. This way the protein-coated surface was exposed to flowing bacteria during the

experiment. For the seal to hold, it was absolutely necessary to place the tube for the

effluent from the PPFC at a level much lower than the surface on which it was

assembled. Once this step was completed, the PPFC was mounted on the x-y motorized

stage of the inverted Nikon microscope and the experiment was initiated. The bottom

surface of the chamber was observed using a 40X objective. The tube for the effluent

from the PPFC was at a lower level than that of the microscope stage.

First, only PBS solution flowed at a constant rate of 0.94ml/min, corresponding

to a shear rate of 100 s-1, to insure a pulse free flow from the peristaltic pump, which was

used to pump the bacteria suspension in the chamber. The flow was stabilized by

manually adjusting a screw in the top part of the peristaltic pump. When this was

achieved, the PBS flow was stopped and the change was made to pumping bacterial

solution into the PPFC. Once bacteria were visualized on the computer monitor as

attaching to the protein coated surface the Optimas program that recursively scans fields

over the attaching surface was started (SCAN.MAC). From this point until the

completion of the experiment, the data were acquired automatically and the images of the

fields with attached particles were stored in the computer for further analysis.



4-3-3 Modified Experimental Protocol

The experimental protocol described in Section 4-3-2 was modified in the latest

experiments because of problems with the recent batches of bovine serum albumin (BSA)

purchased from Sigma Scientific. The new BSA batches were apparently not as pure as

previous batches. It is believed it might have been contaminated with fibrinogen (Fg) or








immunoglobulins that can enhance adhesion of a S. aureus by binding to protein A on the

cell surface. For this reason, BSA was no longer used for blocking non-specific adhesion

sites on the glass surface.

In literature it is well documented that milk casein, the major protein in milk,

blocks nonspecific adhesion. It has been used extensively in developing ELISA assays

for various applications with the common denominator of milk casein blocking

nonspecific adhesion sites on surfaces (Wu et al. 1998, Brown et al. 1997). As an

extension, dry fat-free milk has also been used as a blocking agent for nonspecific

adhesion (Zeng et al. 1996). Therefore, it was a logical extension to try liquid skim milk

instead of BSA to coat the glass slides in order to attain a non-adhesive control. Milk has

on average a composition of 87% water, 4.9%/ lactose, 3.7% fat, 1.3% other ingredients

and 3.1% milk protein, out of which 85-90% is casein (Ferguson 1998).

The protocol described in 4-3-2 was changed in terms of the coating used for

blocking non-specific adhesion to the surface, and the coating time for Fg. The slides

used in testing the effect of receptor length were coated first overnight in a 0.02 mg/ml of

human fibrinogen in PBS. The slides were rinsed with PBS to remove any unattached Fg,

and skim milk was added in the Petri dish. The slides were then swirled on the Roto

Mixer for one hour. The non-adhesive control slides were coated only with skim milk for

one hour. Protein coated slides were rinsed and refrigerated in PBS prior to experiment.



4-3-4 Results Using Initial Protocol in Section 4-3-2

A sequential strategy was used to attain the significant results presented in this

section. Series of experiments were run initially using 1 pm polystyrene beads










(Polysciences, Inc.) as model bacteria in varying electrolyte concentration in the

suspension media in order to establish the experimental protocol for use with the S.

aureus bacteria. This initial approach was used to eliminate any problems with the PPFC,

such a non-uniform laminar field or leaks in the system. It also helped to eliminate the

bugs in the Optimas program (macro) code used in the automatic acquisition of data for

the attachment experiments.

Once the protocol was perfected, genetically engineered S. aureus mutants with

varying length receptors were used to measure the effect of receptor length on

attachment. These mutants were obtained from our collaborators at Trinity College in

Ireland, and this system would be more similar to an in vivo condition. The pertinent

results are further described. The sequence of mutants with the corresponding receptor

lengths is illustrated in Figure 4-6.

A plot of intrinsic attachment rate constant, k+, vs. length of the repeat region of

clumping factor (ClfA) for S. aureus attaching to a fibrinogen coated surface appears in

Figure 4-7. The measured rate constant of attachment on fibrinogen-coated surfaces was

found to be significantly lower for the mutant strains with shorter stalk regions than the

Newman wildtype. It is clearly visible that an increase in k+ occurred as the length of the

receptors present on the S. aureus surface increased. There appeared to be a critical

receptor length for which there was a sharp increase in the attachment rate constant. Also

the baseline (at lower receptor lengths) indicates background binding to the fibrinogen-

coated surfaces, which may be due to the presence of other fibrinogen-binding proteins

on the bacterial surface. Based on comparison to the mutant strain that was completely








devoid of clumping factor, the shortened receptor was still effective in promoting

attachment to the fibrinogen-coated surface. However, the length of the repeat region in

ClfA seemed to play a critical role in promoting attachment, which is consistent with the

hypothesis that this region serves as a stalk to project the ligand-binding domain to

increase the probability of binding with ligands in the cell's environment.


H-


Wildtype


-60 0 50 100 150 200 250 300 360
Length of Repeat Region (residues)


Figure 4-7 Effect of receptor length on bacterial attachment kinetics in the PPFC



This was the first study to demonstrate that receptor is a critical molecular

parameter in determining the rate of cell attachment under well-defined flow conditions.

These results were also consistent with theoretical predictions that, in the presence of

repulsive non-specific forces, k+ should be a strong function of the length of bridging

molecules in the deposition of particles to a surface (Dickinson 1997). A qualitative


pCF56


Shear rate = 100 s-1








comparison between this theory and the experimental results is shown in Figure 4-8.

The mechanistic model predicts a sharp increase in k+ at a critical receptor length, then a

weakening dependence as the length continues to increase. As, illustrated in Figure 4-9,

this trend results from a lower effective energy barrier to attachment by moving the

characteristic distance of receptor-ligand binding farther from the surface. Eventually,

this distance is sufficiently large to negate the repulsive force, and the attachment

becomes rate limited only by the rate Brownian motion and the rate of the receptor-ligand

reaction. The position of the sharp increase on the abscissa depends on the magnitude of

the repulsive forces, as shown in Figure 4-10.



4-3-5 Results Using the Modified Experimental Protocol in Section 4-3-3

These experiments were performed using the same genetically engineered S.

aureus mutant strains of varying receptor lengths with the only difference being the use

of skim milk to block non-specific adhesion. The results are similar to the ones presented

in Section 4-3-4; however, they are more complete since they include all of the available

S. aureus mutant strains with varying receptor length, and they measure attachment of all

strains on 'non-adhesive' control surfaces.















6



., 4

v- 3

+ 2



0
0 10 20 30 40 50
Estimated Length of clfA Stalk Region (nm)




Figure 4-8. Comparison of the mechanistic model predictions to measured values of k+




Using this protocol, two types of experiments were performed as characterized by

the surface used for adhesion. One type of experiment studied the attachment of S. aureus

mutant strains onto fibrinogen-coated surfaces. This experiment revealed the importance

of receptor length in adhesion of bacteria to ligand-coated surfaces. The second type of

experiment was a control experiment in which adhesion of the same S. aureus mutant

strains used in the previous experiment was studied on glass surfaces coated with skim

milk. As described in Section 4-3-3, skim milk was used to block non-specific adhesion










40
^ 30

S20 ------------ Decreasing
Q UEnergy
: 10 I----------------- Barriers
0

S-10
"5 I A [ /Increasing
-20 \ Y Length
-30 %
-40
0 50 100 150

Separation Distance (nm)





Figure 4-9. Mechanistic model results-effect of energy barrier
This plot shows the results of the mechanistic model. It indicates that the observed
increase in attachment with increase in receptor length is due to the decrease in the
energy barriers to attachment.


onto the slide. This has proved an excellent control for S. aureus attachment to fibrinogen

surfaces. The results appear in Figure 4-11 below. An increase in the intrinsic

attachment rate constant was observed with increasing length of the repeat region for the

S. aureus fibrinogen receptor. These data are consistent with the hypothesis that receptor

length plays a critical role in the rate of bacterial attachment to surfaces. A sharp

increase and saturating effect was again observed.















6
3 Increasing
4 Non-specific
S 3 Repulsive
+ Force


0 20 40 60 80 100
Bond Length (nm)



Figure 4-10. Mechanistic model results--effect of the magnitude of non-specific repulsive
forces.



For the control experiments, the attachment rate onto the skim milk-coated glass

surfaces was lower than in attachment to fibrinogen substrate. There was however, some

background binding of S. aureus mutants to the control surfaces as well, which could

have resulted due to other proteins present on the bacterial surface. Another interesting

observation was that the attachment rate appeared to decrease with increasing length of

the receptor on the skim milk-coated surfaces. A possible explanation for this observation

is that in the absence of the fibrinogen ligand, the stalk-like clumping factor increased the

long-range steric repulsion forces by blocking any non-specific binding between the cells








and the surface. This repulsive effect would be expected to increase with increasing

receptor length, as observed.


rildtype on Fg


on Fg
on SM


Figure 4-11. Plot of k+ vs. receptor length for S. aureus mutants on Fg and skim milk
control surfaces
'CF-mutant' refers to the mutant strain devoid of clumping factor. Wildtype refers to the
native Newman strain. 'SM' indicates skim milk-coated control surface. 'Fg' indicates
fibrinogen-coated surface.




Comparison of the attachment kinetics results obtained with the non-specific

adhesion blockers, BSA and skim milk, respectively, showed a similar trend, i.e., an

increase in k+ with increase in the receptor length. However, the values of k+ are different

for the two sets of experiments. k+ values for the experiments using skim milk were lower

than the values obtained when BSA was used as the non-specific adhesion blocker. A


i 1.5
S1.25



b 0.75
i-^
x 0.5

S0.25

0


0 100 200 300 400
Length of Repeat Region (residues)








possible reason for this could be that the skim milk affected the nonspecific repulsive

interactions. An increase in the repulsive interactions would have caused the observed

decrease in k+. Also the position of sharp increase in k+ vs. R was different. For the

surfaces treated with BSA, the sharp increase in k+ occurred at a longer length of the R

region than for the surfaces treated with skim milk.

These experimental and theoretical results suggest that molecular properties other

than simply the affinmity of the binding sites between receptor and ligand govern the

function of the receptor in mediating cell attachment. This observation is relevant to

further understanding of the molecular and physical basis for bacterial attachment to

biomaterial surfaces.



4-4 Measurement of the Role of Electrolyte Concentration in Bacterial Adhesion to Glass

Surfaces

Bacterial adhesion is also crucial in water purification by filtration. In the design

of filters for separation of bacteria from wastewater streams, enhanced bacterial

attachment to filter surfaces is desirable. Filter surfaces can be modified to decrease the

potential energy barrier to attachment. This can be achieved by several means. The filter

surface can either be coated with a surfactant or in the particular case of wastewater

treatment with a metal hydroxide which will lower the surface energy. At the University

of Florida Engineering Research Center for Particle Science and Technology, a novel

method was investigated for bacteria separation from wastewater streams. It consisted in

the use of metal hydroxide coated sand in a packed bed through which bacteria

suspensions flowed. This coating is believed to enhance bacterial deposition by altering








the electrostatic interactions to lower the energy barrier to attachment (Truesdail et al.

1998).

As an initial step, to better understand the role of electrostatic forces on bacterial

deposition, the effect of electrolyte concentration in the media on k+ was measured using

the above methodology for measuring particle attachment kinetics under well-defined,

laminar flow conditions in the PPFC. The surfaces used in the study were glass slides

thoroughly cleaned in chromic acid solution overnight, then rinsed and submerged in

double distilled water prior to experiments. The bacterial strain used was the common

pathogen, S. aureus. It was suspended at a concentration of 2x10 7 cells/ml in sodium

chloride solutions of a different electrolyte concentration for each particular experiment.

Electrolyte concentrations used in the attachment kinetics experiments were lxlO"4,

lxlO-3, 102, and 1M.

Figure 4-12 is a plot of the negative log of the electrolyte concentration vs. k+.

The data showed a sharp increase in k+ with increasing electrolyte concentration. In terms

of DLVO theory, this observation is consistent with a decrease in the energy barrier to

attachment for bacteria as the characteristic length (Debye length) and repulsive

electrostatic forces decrease (Prieve & Ruckenstein 1974). Eventually, attractive van der

Waals forces predominate and the rate of attachment becomes limited by only the rate of

diffusion or sedimentation over the boundary layer near the surface.










2

1.5


X
+0.5

0-
0 1 2 3 4 5
-log(Molarity of NaCI)




Figure 4-12. A plot of intrinsic attachment rate constant, k+, vs. electolyte concentration
Attachment of S. aureus wildtype on glass was measured as a function of electrolyte
concentration.













CHAPTER 5
MATHEMATICAL MODEL FOR PARTICLE TRANSPORT TO THE BOTTOM
ATTACHING SURFACE IN THE PARALLEL PLATE FLOW CELL


5-1 Modeling Particle Attachment Kinetics


The purpose of this section is to give a background on the modeling of colloidal

particle attachment and on what has been published in the literature as well as highlight

the differences between the procedure described in this dissertation (Section 5-2) and its

advantages over other protocols. This model developed here is used in conjunction with

attachment kinetics experimental results of the effective attachment rate constants to

predict an intrinsic attachment rate constant independent of any global parameters in the

parallel plate flow cell.

Several experimental systems are known and used in the modeling of particle

deposition onto collectors. Each system comes with its inherent quantitative formulation

of the particle transport to the attaching surface in well-defined conditions. In each case

the particle deposition rate can be calculated from the particle transport equation in the

particular system. Systems used include the rotating disk, stagnation-point flow, parallel-

plate flow cell, spherical, and cylindrical collectors.

For example, Marmur et al. studied theoretically the kinetics of cell deposition

from stagnant solutions under conditions in which the cellular adhesion rate is determined

by both the escape over the potential energy barrier and by the cell transport rate to the

surface (Marmur et al. 1976). Other researchers have studied the kinetics of latex








particle deposition from flowing suspensions using a stagnation point flow created by a

rotating disk (Dabros et al. 1977) or spherical and cylindrical collectors usually applied in

filtration modeling (Adamczyk & Van de Ven 1981 & Adamczyk et al. 1983). The

rotating disk has been used in particle deposition studies due to its simple and well-

defined hydrodynamics and mass transfer. In stagnation-point flow the thickness of the

hydrodynamic and diffusion boundary layers are constant close to the stagnation point. If

a transparent collector surface is used, then, in stagnation point flow one can view under

microscope the particle deposition. Until 1982 the deposition rates for particles have been

evaluated from turbidity changes of a suspension before and after leaving a filter. In such

experiments the attached particle density was measured either in an indirect way or the

collector surface was removed from the experimental environment and particles counted

under microscope post-experimentally. Direct visualization under microscope cannot be

applied when a rotating disk is used due to its motion.

Particle deposition in a stagnation-point flow can provide a first approximation

for particle deposition on complex systems such as spherical and cylindrical collectors

(Elimelech et al. 1995). Other researchers have studied particle deposition in stagnation

point flow (Dabros & van de Ven 1983, 1987, Chari & Rajagopalan 1985a,b, Adamczyck

et al. 1986).

Parallel-plate flow cells have been used to study particle attachment kinetics onto

stationary surfaces. Bowen and Epstein (1979) studied latex particle deposition in a

parallel-plate flow cell using an indirect method for measuring the attached particle

density based on radioactive isotope activity measurements of the isotropically active

particles. Other literature counts exist of the use of a parallel-plate flow cell for








theoretical and experimental studies of particle deposition (Adamczyck & van de Ven

1981, Sjollema & Busscher 1989, 1990). However, none predict an intrinsic attachment

rate constant independent of any global parameters as the model described in Section 5-2.

Radial-flow cells have also been used to study shear-dependent bacterial adhesion

kinetics to biomaterial surfaces (Dickinson & Cooper 1995). Intrinsic attachment and

detachment rate constants are estimated by fitting mathematical models to resulting

experimental data. The model for cell attachment accounts for the global transport of the

cell in the chamber to estimate the cell concentration near the collector surface. The

model described in the next section is similar to this attachment model in the radial flow

cell. However, it is developed for a different geometry of the PPFC.



5-2 Model Description


As mentioned in Chapter 4, the purpose of a mathematical model was to correct

for variable transport effects in order to eliminate the dependence of the effective

attachment rate constant, keff, on downstream position in the PPFC. Specifically, the

model estimates the concentration of particles "near" the surface at a boundary layer

distance, F, and the resulting position-dependent value of kejl(x), for a given intrinsic

attachment rate constant, k+. The model prediction is then fit to the experimental data for

kej,(x) by nonlinear least-squares regression to provide the best estimate of k+. The

rationale is that k+ is a local parameter, which depends only on local surface properties

and flow conditions, not the global transport effects in the PPFC.


























Figure 5-1. Illustration of the convective-diffusive model for particle transport in PPFC
Schematic of a lateral view of the PPFC displaying a parabolic flow profile, and
illustrating the conditions of the convective-diffusive mathematical model used to correct
for transport effects to the surface of the PPFC.


A schematic of the flow cell and model assumptions appears in Figure 5-1. Upon

neglecting entry and exit effects (c.f. Section 4-1-1), the parabolic velocity profile, v,(z),

can be solved from the steady-state Navier-Stokes equation to yield


602 ( h)
z)= z (5-1)

where Q is the volumetric flow rate, w is the width of the chamber, and h is the gap

thickness. The shear rate, S, on the surface is therefore

6Q
S wh2 (5-2)

which allows Eq. (5-1) to be rewritten as


v,(z) = Sz ( l-) (5-3)








The concentration of cells in suspension, c(z,x), is a function of both vertical

position and downstream distance. Exact solution of c(z,x) requires knowledge of all

forces on the cells in suspension and how these forces depend on position. However, in

our case, the forces near the collector surface are generally not known, but can be

assumed to act on the particle over a short distance, F, which is on the order of a particle

diameter. At separation distances greater than F, only convection, Brownian forces, and

gravitational forces are assumed to act on the particle. It is also assumed that the cell

concentration is sufficiently low to neglect cell-to-cell interactions, a condition that is

experimentally imposed. Accounting for diffusion and sedimentation, the net particle

flux in the z-direction is

Jz = -vc(x,z) D ac(xz) (z> ) (5-4)
dz

where D is the particle diffusion coefficient and vs is the sedimentation velocity.

Assuming diffusion is negligible compared to convection in the x-direction, the steady-

state continuity equation in the region is

o = _VX(z) 0C + D d c +, 9C ( 55
0+ +v -- (z>f) (5-5)
8 X z2 Z 0 z

At separation distances less than F, the steady-state flux of particles over

boundary layer is assumed proportional to the concentration at z= Fr, such that

-Jz =k+c(F,z) (0
This flux is then equal to the rate of accumulation of cells on the surface, such that

dc, -=k+ c(r,z) (5-7)
dt








The attached cell density, cs(x,t), is assumed to remain sufficiently low over the duration

of the experiment to neglect any dependence of k+ on cs(x,t) (also experimentally

imposed). Recall from the previous chapter that the rate of accumulation is

experimentally measured to determine the effective attachment rate constant, kef(x),

which is defined in terms of the known inlet concentration, Co, i.e.

dc = k ff (x)co0 (5-8)
dt

Equating Eq. (5-7) and (5-8) yields

keff(x) _c(r,x) (57)
k+ Co

Therefore, determining k+ from keft(x) requires solution of the cell concentration at the

boundary layer, c(F,x), from Eq. (5-4).

Solution of Eq. (5-4) requires an initial condition for the x-domain and two

boundary conditions for the z-domain. The assumed initial condition is that c=co at x=O

and z > -. The appropriate lower boundary condition is obtained by matching the fluxes

at z = F, i.e.

k+c(x,F) = D c(xz) Ir +v,c(x,F) (5-8)
d z

For the upper boundary condition, it is assumed that the concentration is undisturbed

from co far from the surface, such that


c = Co (z ) >>


(5-9)








5-3 Dimensionless Equations


Before solving, the governing equations were made dimensionless by scaling with

respect to the characteristic dimensions. The characteristic downstream distance was the

length of the chamber, L= 75 mm. The characteristic vertical dimension was chosen to

be D/vs, which is the predicted decay length of concentration near the surface under the

special case of no flow and no attachment (i.e., the decay length of the exponential

solution to Eq. (5-4) under for S=O and k, = 0). The dimensionless spatial coordinates

are then

x
X (5-10)
L

Z Vs
Y- D (5-11)

Eq. (5-4) becomes

U 1(y-y2) I y( (5-12)


where u=c/co, and the characteristic dimensionless parameters are

D
D -v, (5-13)




iv( ) (5-14)

The dimensionless inlet and upper boundary conditions are u(-=O,y)=1 and u(,y=H)=l,

respectively, where H >> 1 is the chosen upper boundary distance. (In the results shown

here, H was chosen to be equal to 5, which was confirmed to be appropriately large by








checking that the solution yielded U =0). The dimensionless lower boundary
Y y=H

condition is

U (a-l)u(4,y) (5-15)
SY y=Fr

where y-= Fv/D, and a -= k+/v.



5-4 Model Solution

Eq. 5-12 was solved numerically using a centered finite difference approximation

of the y-domain (y < y < H) to convert Eq. (5-12) into a system ofn ordinary differential

equations of the form

u.= 1 [u,-+1-2-u, +u1,- Ui+l + U, (2 d 4 p (y, _7y,2) Ay2 + 2Ay
d U_ 1 u2-2u,+uo + u2 +uO
0 4 fi (y-r-y2) Ay2 2Ay (5-17)

where un = 1, corresponding the boundary condition at y = H, and from the finite

difference approximation of Eq. (5-15),

u0 = u2 -2(a )Ay (5-18)

which corresponds to the boundary condition at y = y.

For given values of parameters, y, f/, q7, and a, the system of ordinary differential

equations shown in Eqs. (5-16) and (5-17) were numerically integrated in the 4-direction

using a fourth order Runga-Kutta routine with step size control (Dormand & Prince

1980). In the fitting of the model solution to the data, f,, r and y were fixed parameters,








and a was a fitted parameter. To estimate the fixed parameters, the diffusion coefficient

was estimated using Stokes Einstein equation for one-micron diameter particle, to find D

= 3xl 07cm2/min. The sedimentation velocity for S. aureus was estimated previously at

vs=3.5xlO4cm/min (Dickinson & Cooper 1995). The boundary layer thickness was

chosen to be one cell diameter, or F=l xlO4cm. From these values, f8 was estimated to

equal to 1.7 for 5=-100 s', ywas estimated to be 0.12, and r1 was estimated to be 0.034.

Because the model was used to estimate the fitted parameter, a, the sensitivity of

the model solutions to uncertainties in the fixed parameters was relevant to the accuracy

of this estimate. The sensitivity of the dimensionless solution for kejVs vs. x/L to the

dimensionless parameter a, f/, and y are shown in Figures 5-2, 5-3, and 5-4, respectively.

(The solution was not significantly dependent on rq over a large range of values;

therefore, a corresponding plot for r is not shown). Figure 5-2 shows the effect of

dimensionless attachment rate constant, a, on the resulting effective rate constant, keff.

As shown in this plot, if k+ is larger than vs (a>]), the model predicts a decrease in keff

with downstream distance due to a downstream depletion of cells in suspension from

rapid attachment of cells near the surface (c.f. Eq. 5-7). Similarly, for k+ less than v,, the

model predicts an increase in keff with downstream position due to sedimentation and an

accumulation of cells near the surface. This plot clearly illustrates the importance of

correcting for variable transport to the surface as a function of position in the PPFC,

because the effective attachment rate constant can vary with measure position and can

deviate significantly from the intrinsic rate constant. As observed in Figures 5-3 and

5-4, the model solutions are somewhat insensitive to the values of 8 and y over a large








range of reasonable values. This implies that moderate uncertainty in these parameter

estimates will not result in large uncertainties in the model prediction of keffy.

The optimal value k, consistent with the experimental data for kej(x) was

estimated by fitting the model solution to the experimental data using nonlinear least-

squares regression. This required recursive solution of the model for iterative values of

k, to minimize the weighted sum of the squares of the residuals between the model

solution and the data points for ke,(x). Specifically, a regression program was written

which minimized the function

S(k+)= (k.,,,(x,)-k (x,))2 /cr, (5-19)
i

with respect to k+, where a, is the standard error in the measured value of keff at position,

x,. The uncertainty in k+, ark, was calculated by numerically calculating


2 S(k evaluated at the optimal value of k+.


The need to correct for transport effects in the PPFC, especially in studies of the

role of fluid shear rate on attachment, is further demonstrated in Figure 5-5. Here values

of position-averaged keff and k+ are plotted as a function of shear rate for the mutant strain

lacking the stalk region on fibrinogen-coated surfaces. The data for keff suggest a strong

dependence of attachment rate on shear stress. However, the comparison of these

uncorrected values to the estimated values of k+ suggests that the shear-dependence of keff

resulted from a greater accumulation of cells near the surface at lower flow rates, and not

due to any effect of fluid forces on the probability of cell attachment.









2

1.8
1.6
a =1.7
1.4 \
1.2 71



S0.8 a =0.9

0.6 -5

0.4
a =0.1
0.2

0 ---
0 0.2 0.4 0.6 0.8 1
Downstream Distance, x/L


Figure 5-2. Plot of the effective attachment rate constant vs. dimensionless downstream
distance for various values of the dimensionless intrinsic attachment rate constant
a = k+/v,







2 1 T----



1.5 .Increasing=2


; 1

ar=0.2

0.5
"*x Increasing f3


0 0.2 0.4 0.6 0.8 1
Downstream Distance, x/L



Figure 5-3. Plot of the dimensionless attachment rate constant as a function of
downstream distance for various values of the the dimensionless flow parameter, f8
The curves correspond to 8Jranging from 0.5 to 2.5. This plot illustrates the model
predictions are only weakly sensitive to fi for both large and small values of a.







2


Increasing y
1.5 c=2.0






0.5 a=0.2

JIncreasing y


0 0.2 0.4 0.6 0.8 1


Downstream Distance, x/L


Figure 5-4. Plot of the dimensionless attachment rate constant as a function of
downstream distance for various values of the the dimensionless boundary layer
thickness, y
The curves correspond to ranging from 0.06 to 0.3. This plot lustrates the model
predictions are only weakly sensitive to y for both large and small values of a.







0.3
0.25-t k+
.2 o.2
0E
0.2 ..eff avg
S 0.15

1 0.1
S 0.05 ---
0 i
0 100 200 300

Shear Rate (1/s)

Figure 5-5. Plot of the dimensionless k+ and average keff as a function of shear rate for a S.
aureus mutant with a missing R region.














CHAPTER 6
EVANESCENT WAVE LIGHT SCATTERING COUPLED WITH THREE
DIMENSIONAL OPTICAL TRAPPING'


6-1 Introduction




To gain full understanding of the role of surface interactions in bacterial

attachment, it is desirable to correlate the measure attachment rate constants with the

surface forces between the cell and substratum. However, no method was previously

available to measure the interaction forces between a single bacterium and a surface.

Evanescent Wave Light Scattering coupled with Three Dimensional Optical Trapping

(EWLS/3DOT) is a novel technique developed to directly measure the potential energy of

interaction between a particle and an (optically transparent) surface from an initial

measurement of the separation distance between the particle and the surface. By

integration of the potential curve one can also obtain interaction forces between a particle

and a surface.









Several members of my group have contributed to the development of the EWLS/3DOT. My role was in
initiating the design, assembly, and preliminary testing of the apparatus. Aaron Clapp has since been
primarily responsible for the analysis for obtaining force measurements from the raw signal. Sections of
this chapter closely follow our paper submitted to Rev. Sci. Instruments, co-authored by Clapp, myself, and
Dickinson.








6-2 Apparatus / Principle of Operation




The EWLS/3DOT apparatus consists of several parts which are integrated into an

existing inverted optical microscope (Diaphot 200, Nikon Corporation, Mellville, N.Y.)

equipped with an x-y motorized stage (LUDL Electronic Products, Hawthorne, NY) and

a z-focus control through the microscope objective (Figure 6-1). The key components

have the following important purposes: (a) generation of an evanescent wave for

measurement of particle-surface separation distance; (b) laser trapping of the particle

from suspension; (c) measurement of light scattering intensity from trapped particle; and

(d) visualization of the trapped particle. The other components will be described in the

order of their respective relation to the main components.


SHe-Ne laser




X-Y motorized stage


aperture filter


Figure 6-1. Structural schematic of key components of the EWLS/3DOT apparatus








The principle of operation of this novel technique is based on the generation of an

evanescent wave (Figure 6-2) at a glass-liquid interface upon the total internal reflectance

of a laser beam directed through a dove prism (BK7 glass, 45 degree angle; Optosigma,

Santa Ana, CA). The dove prism is optically coupled using an immersion oil of the same

index of refraction (n d =1.515) as the prism and the glass chamber holding the particle

suspension. The laser generating the evanescent wave is a variable 17mW randomly

polarized Helium-Neon laser of wavelength 632.8 nm (Melles Griot, Irvine, CA). The

laser beam is guided to the hypotenuse of a dove prism by a set of elliptical mirrors

mounted on a pole at the exit of the beam from the laser housing. The angle of the top

mirror can be manually adjusted using a thumbscrew. This adjusts the lateral rotation and

vertical pitch of the mirror mount in order to accurately guide the laser beam into the

prism at an angle, which provides total internal reflectance of the beam at the surface of

the glass chamber. This creates the evanescent wave at the glass-liquid interface (i.e.,

precisely controlling the penetration depth of the evanescent wave).

The intensity of the evanescent wave is known to decay exponentially with

increasing distance from the glass-liquid interface. Hence, a particle present in the

vicinity of this evanescent wave will scatter light proportional to the intensity of the laser

beam creating this evanescent wave, i.e., 632.8 nm, and it is expected that as the particle

is brought closer to the glass surface, its scattering intensity will increase exponentially.

This can be seen in Figure 6-3. The increase in scattering intensity as the particle is

brought incrementally toward the surface can be visualized on a computer screen.











Glass Slide Total Internally
as u Reflected Beam

-- Evanescent
-- Wave
1?53, (A= 633nm)



Scattered Intensity: 1(z) = I(O)e-A




Figure 6-2. Schematic of the principle of Evanescent Wave Light Scattering


The novelty of this technique stems from the integration of a three-dimensional

optical trap (Figures 6-4 and 6-5) with the total internal reflection microscopy developed

by Prieve (Prieve & Frej 1990). The three-dimensional optical trap is generated by

focusing a laser beam from a 100mW diode laser (Cell Robotics, Albuquerque, NM)

mounted underneath the nosepiece of the microscope objective, through a high numeric

aperture objective (Plan Fluor 100X, 1.3 numerical aperture, oil immersion; Nikon

Corporation, Japan) to a diffraction-limited spot (of diameter X, the wavelength of

light) at the focal point (Block 1992). This tightly focused beam produces intensity

gradients in three dimensions, giving rise to the lateral and axial trapping components. A

particle with a refractive index higher than that of the surrounding medium will

experience a trapping force directed toward the region of highest light intensity (Block

1992).















Su r b)

















Figure 6-3. Evanescent Wave Light Scattering of Staphylococcus aureus
Images in the left column (a) are of a S. aureus bacterium far from the surface. The top
image is obtained from a real experiment and it can clearly be seen that the bacterium is
scattering less light than in the case where the particle is closer to the surface (right
column images (b)). The bottom pictures in both columns are schematics illustrating the
light scattering magnitude.


A trapped particle can be manipulated laterally by moving the motorized stage in

the x and y directions and it can be positioned axially, by vertically positioning the

objective with a high-resolution stepping motor (Ludl Electronic Products, Hawthorne,

NY). The 3DOT can be used to bring a particle in the vicinity of the glass surface with

excellent resolution. It allows the trapped particle to sample positions which, otherwise








under free floating conditions, would not be attainable due to the potential energy barrier

between the particle and the surface.


S100X Objective
1.3 NA/Oil Immersion


Figure 6-4. Schematic of the three-dimensional optical trap
Lateral view of how the high numerical aperture objective tightly focuses the laser beam
to a diffraction limited spot. A micron-sized particle present at the focus of the laser beam
is trapped.


Essential to this technique is the automated video microscopy system, which

includes an inverted Nikon microscope and a monochrome CCD video camera (Thomas

Optical Measuring Systems) mounted into one of the eyepieces. The image of a trapped

particle within the glass chamber is captured and sent to a frame grabbing board (g Tech)

which processes the image and displays it on a video monitor. Therefore, the trapping and

manipulation of a particle can be directly observed on the computer monitor.














Zo



Force /

Distance, z
Focused Trapping Laser



Figure 6-5. Schematic of forces in the three dimensional optical trap
This figure illustrates the forces acting on a micron-sized particle at the laser beam focus.
The potential at the trap focus is harmonic.


Central to understanding the measurement of the particle scattering intensity is the

existence of the two separate light paths present in this system. One, consisting of the He-

Ne laser beam being totally internally reflected and in turn generating the evanescent

wave and a second one, generated by a halogen lamp, used to illuminate the chamber

from the top and thus allow visualization of the particle through the microscope

objective. A particle present in this evanescent wave will scatter light at the same

wavelength as the He-Ne laser. Hence, the scattering wavelength that needs to be picked

up by the photomultiplier tube (PMT) (Oriel Instruments, Stratford, CT) is in a very

narrow range around 632.8 nm, the wavelength of the He-Ne laser. This is done by

mounting on the PMT a band pass filter (BPF), (Nikon, Japan) which blocks all other

wavelengths except 633 +/- 2 nm. Since the scattering intensity of a trapped particle is








expected to be quite low, a very narrow aperture (1mm diameter) is also mounted ahead

of the BPF on the PMT in order to block any extraneous scattering from the system. The

highly sensitive PMT delivers a continuous voltage signal, which is linearly proportional

to the scattered intensity.

The voltage signal from scattering intensity of the particle is pre-amplified before

being sent to a data acquisition board (Keithley Instruments, Cleveland, OH) and stored

for further analysis and determination of the particle-surface separation distance. This

results in a real-time measurement of the particle position near the test surface.

The force and potential energy determination methodologies were derived in our

group and submitted for publication along with some preliminary results (Clapp et al.

1997).



6-3 Force Measurement


The EWLS/3DOT apparatus is capable of measuring interaction forces between a

micron-sized particle and a flat test surface. A variety of particles and surfaces can

potentially be measured provided the particle system is appropriate for measurement by

the optical techniques involved. In spite of these restrictions, there are still numerous

colloidal systems that can be explored using the technique, especially considering the

variety of coatings, particle types, and solution characteristics (e.g., ionic strength, pH,

surfactant concentration) that are available. However, for the purpose of validating the

technique, selection of a well-defined system is important. Our initial studies using the

technique have focused on aqueous suspensions of 1-pm polystyrene microspheres

(1.072 0.019 mm; Polysciences, Warrington, PA) interacting with glass substrata.








The suspensions prepared for experiments must have a relatively low particle

density such that the apparatus observes only the interaction of a single particle and the

test surface. Solutions were prepared using 1 pm polystyrene microsphere densities on

the order of 106 particles/cm3, adequately concentrated to allow location of a particle but

sufficiently dilute to avoid interactions with other particles during measurements. In

addition to the 1-/in particles a small amount of 22-/pm polystyrene microspheres

(22.011 3.031 Wm; Polysciences) were added to serve as spacers between the glass slide

and cover glass. Solutions were prepared using 22-/an polystyrene microsphere with

densities on the order of 104 particles/cm3 sufficient to maintain a fixed liquid gap.

Maintaining the gap is crucial to the experiments because the technique intends to study

the interaction of a single particle with one surface only. In addition, calibration of the

trap depends on a region of zero interaction force only attainable with larger gaps.

Slides were prepared for experiment by placing a small drop (40 pl) of particle

solution onto the center of the slide surface and overlaying a piece of cover glass. Excess

water was drawn into a tissue by capillary forces, thus establishing a fixed liquid gap as

the cover glass compressed the 22-/pm spacer microspheres. A small drop of immersion

oil was placed onto the objective lens and the slide was mounted onto the motorized stage

with the cover glass facing down toward the objective. The objective was raised until the

oil bead was compressed between the objective and cover glass. By translating the

motorized stage and adjusting the position of the objective with the fine focus control, an

isolated free particle was located in the field of view. The optical trap was activated, and

the free particle was positioned into the trap center by precise movements of the stage

and focus motors. Once trapped, the particle was fixed near the focal point of the








objective (trap center), and for any small translation of the stage or objective, the particle

remained fixed with respect to the focal point.

With a particle trapped in the liquid gap, a small amount of immersion oil was

placed upon the upper surface of the glass slide, centered over the objective lens. The

dove prism was placed over the oil, optically coupling the prism to the slide below. The

He-Ne laser was activated and its beam directed into the center of the hypotenuse face of

the dove prism by the adjustable mirror at an angle sufficient for total internal reflection

at the glass-water interface. The trapped particle was then moved axially (z) towards the

glass surface until it was pressed as far as possible into the surface by the trap. The

mirror was adjusted to maximize the scattered intensity of the particle. This laser

alignment step improved the ability of the PMT to sense changes in intensity due to small

fluctuations in the axial direction and the signal-to-noise ratio.

With the optics properly adjusted, the particle was withdrawn from the surface to

a point far from the surface (>5 /anm) where no scattering was measurable. A computer

program was written in Microsoft Visual Basic to collect data and precisely control the

trap position in the axial direction. For each trap position, scattered intensities from the

particle measured by the PMT were sampled by the data acquisition system for a given

time interval after which the position of the trap was moved a small distance (10 nm or

more) toward the surface. The procedure was repeated until the axial trapping force

imposed on the particle was insufficient to move the particle any closer to the surface

with subsequent moves of the trap center (outside of the linear trapping force regime).

The experimental data was saved at each trap position to a PC file and analyzed at a later

time.











6-4 Data Analysis


6-4-1 Overview



Data collection and axial trap position are both controlled via in-house software.

The digital sampling parameters of the PC data acquisition board are user-defined at run-

time along with the trap center position. The program records three quantities for

sampling at a given trap position, relative axial trap center position, sampled voltage

mean, and sampled voltage variance. As the software begins acquiring data, the program

records the first trap-center position as zero and refers to all future trap-center positions

relative to the initial position. This method records all the data that is necessary for

determining the mean relative separation distance, trap stiffness, and mean interaction

force.





6-4-2 Calculating Particle Position


As discussed elsewhere (Chew et al. 1979), the intensity of scattered light, I(z),

from a particle decays exponentially with normal separation distance, z, from the

reflecting surface:

I(z) = I(O)e-* (6-1)
where, 1(0) is the intensity of the particle contacting the surface, and where the

evanescent wave decay constant is determined by








4ir [n2 sin2 2 12 (6-2)

Calculating the mean particle position from the mean collected intensity, for a

large number of samples, involves deriving normally distributed statistics from a

lognormally-distributed data set. Because of the stiffness of the optical trap, fluctuations

of the particle position within the trap are generally small2 and closely obey a Gaussian

distribution with mean, (z), and variance, a2. This condition and equation (6-1) imply

that the fluctuations in I obey a lognormal distribution with mean, (I), and variance, al,

from which the first two statistical moments ofz can be calculated as




) 2 +1 (6-3)



where a, is the variance in I. Figure 6-6(a) is a histogram of relative separation distance

for a trapped 1-apm polystyrene particle. The solid line is a normal probability

distribution, based on the mean and variance of separation distance, which indicates that

the positions are normally distributed. Figure 6-6(b) is a histogram of sampled intensities

(Voltage units) for a trapped one-micron polystyrene particle. The solid line is a

lognormal probability distribution, which demonstrates that the sample intensities follow

a lognormal distribution. Equation (6-3) indicates that measurement of 1(0) allows

calculation of the absolute separation distance, (z). However, 1(0) is not necessary for



2 This assumption is violated when the negative curvature of the interaction potential, a2 /az2, is
comparable to the trap stifhfness, y, which occasionally occurs in regions of strong attraction and requires a
more complex analysis not discussed here.








calculation of the relative separation distance between different measured positions,

1
therefore, all positions can be measured to an undetermined constant, -Iln 1(0).
18


0.14...


0.1
0.08
0.06
0.04


a)








3 0.4 0.5 0.6 0.7 C


Intensity (V)


Relative Separation Distance (nm)



Figure 6-6 Histograms for a trapped 1/nm polystyrene particle
(a) Histogram of sampled intensities for a trapped 1/m polystyrene particle. The solid
line is a plot of a lognormal probability distribution using data statistics. The plot
indicates the lognormal nature of the sampled intensities as expected due to the
exponential decay of the intensity with separation distance, and the normally distributed
behavior of the particle position, (b) Histogram of relative separation distances for a
trapped 1/pm polystyrene particle. The solid line is a plot of a normal probability
distribution using statistical quantities from the sampled data. The plot indicates that the
particle positions are normally distributed as expected based on the Brown fluctuations of
the particle in the optical trap.










6-4-3 Calibrating the Optical Trap

The three-dimensional optical trap is characterized by its lateral and axial trapping

components. For the force measurements described here, only the axial component of the

trap needs to be calibrated. The two values that are required for force measurements are

the axial trap stiffness, y, and the trap center position, z0. Because trap stiffness is

sensitive to experimental conditions, the trap must calibrated each time an experiment is

conducted.

In order to properly calibrate the trap, a trapped Brownian particle must be

observed sufficiently far from the test surface such that the particle motion is influenced

only by random Brownian fluctuations, viscous drag, and the restoring force of the trap

toward the equilibrium position. However, the particle must also be near enough to the

test surface to scatter light from the penetrating evanescent wave. The trap imposes an

axial harmonic potential on the particle near the trap center:


S(z)= (Z Zo)2 (6-5)

With stiffness parameter, y, Boltzmann's distribution law implies

p(z) exp[- (z)/ATI, hence the parameters z0 and y can be obtained from the

stationary mean and variance of particle positions at a trap position far enough from the

surface to neglect surface forces. For this calibration position, zo = (z), and

kT
Y = --- (6-6)
z-








where, k is Boltzmann's constant, T is the temperature of the surrounding liquid, and ar2

is the variance of particle position. From equation (6-5), the force on the particle from the

optical trap is


F(z) = do (z) Y(Z- Zo) (6-7)
dz= -) (6-7)
This requires a value for the trap center position, zo. Far from the test surface, the particle

will fluctuate around the trap center, and thus have a mean position, (z), equal to the

position of the trap center, zo. The mean force in this case would be zero, and the trap

center position would be equal to the mean position.

Axial trapping efficiency depends on the distance of the particle above the cover

glass (Wright et al. 1994). The absolute effect of this distance dependence of trapping

efficiency can be reduced by maintaining a small gap, such that there is sufficient

trapping anywhere within the sample chamber; the 22-micron gap reported here should

be adequately small to ensure adequate trapping efficiency. The relative effects of the

trapping efficiency with distance must remain small over the range of calibration and

measurement near the surface. Because the total distance traveled by the particle in the

axial direction during sampling is less than about 300 nm, we can assume that the axial

trapping efficiency will remain constant over the small range of distances. An illustration

of the trap calibration measurements can be seen in Figure 6-7.






71






oo \Trap Potential
1 T(z) (z=-zO)
75 kT


dOT
d = ZoZ =zZ)
25



-25 Io z) Scan at 10 nm increments zo

100 200 300 400 500
z (nm)



Figure 6-7. Trap calibration and measurements using the EWLS/3DOT technique






6-4-4 Calculating Force and Potential


The value for the interaction force far from the test surface is zero, but as the

particle is stepped at known increments toward the surface, the surfaces forces eventually

begin to deflect the mean position, (z), from the trap position, zo. With the values for g

and zo known, the mean force of interaction between the particle and test surface can be

calculated from (z). At equilibrium, the surface force must balance the trapping force

at the position (z) ; i.e.








F((z)) = -F ((z))= y((z z) (6-8)
The potential energy of mean force at any position, z, can be estimated by

numerically integrating the mean force data:

00
O(z)= fF,((z))d(z) (6-9)

As the mean of the Brownian force approaches zero (i.e., for large sample sizes),

the potential of mean force will converge to the mean potential.



6-5 Preliminary Results



Preliminary results obtained in our lab are illustrated in Figure 6-8(a). It is a plot

of the interaction force between a 1-pm polystyrene sphere and a flat glass surface in

distilled water. The force data was numerically integrated using the trapezoid rule to

produce the corresponding potential energy curve in Figure 6-8(b). Each data point on

the force and potential energy plots was calculated by averaging the separation distance

of the particle over a period of five seconds at 1 ms intervals. It was found that the mean

position of the particle converged to a stationary value after about 2 seconds, thus the 5

second sampling periods showed no significant difference in the measured values. The

trap center was moved by 10 nm increments (smallest resolution of the stepping motor) in

the z direction (axial) such that the measured intensity varied as little as possible between

positions. For regions where the intensity varied little with subsequent movements toward

the surface (i.e., far away from or very close to the surface), the trap center was moved by

larger increments (20-200 nm) to speed up the measurement.










a)


1.8E-11 1
1.6E-11
1.4E-11
1.2E-11
1E-11
8E-12
6E-12
4E-12
2E-12
0
-2E-12


b)


150 200


250


Relative Separation Distance (nm)



Figure 6-8 Force and potential energy curves for a Ip/an polystyrene particle interacting
with glass in water
(a). Force curve for a 1,ma polystyrene particle interacting with glass in water. The
abscissa represents relative separation distance. Far from the surface there is zero force
between the particle and surface. As the particle nears the surface, there is an attraction
(negative force values) and then a strong repulsion between the particle and surface. (b).
Corresponding potential energy obtained by numerically integrated force data in Figure
6-9(a) to yield the interaction potential plot. There is a secondary minimum followed by a
strong repulsive energy barrier as the particle is moved toward the surface.


50 100 150 200 250

Relative Separation Distance (nm)


60
50
40
30o
20'

0t

-10-
-20-


*-!<

0








The accuracy of the separation distance measurements is strongly dependent on

the precision of the inverse penetration depth, 8/. From equation (2), ft is a strong function

of the incident angle of the laser beam, 0 1, just above the critical angle. However,

equation (2) also suggests that /f becomes somewhat insensitive to changes in the incident

angle for values much higher than the critical angle. The data presented in Figure 6-9

(a)&(b) is obtained at a measured incident angle of approximately 67.3 0, sufficiently

above the critical angle of 61.4 0. Therefore, 8 can be accurately estimated.

By observing the trends in the force and potential energy data (Figure 6-9(a)&(b))

it is obvious the existence of a small secondary minimum of approximately -13 kT and a

maximum measurable repulsion of 55 kT. These results demonstrate the ability of the

technique to resolve small longer-range forces, as well as relatively strong repulsion

forces not previously attainable with either AFM or TIRM alone. The addition of the

three dimensional optical trap proves to have a dual role: it allows for a sensitive force

measurement and at the same time provides a means for sampling forces over a larger

range of distances than allowed by a naturally diffusing particle.





6-6 Discussion



The technique described here is a novel combination of three-dimensional optical

trapping for force measurement and evanescent wave light scattering for position

measurement to provide a means for direct determination of the surface forces between a

small (-1 micron) particle and a test surface. The similarities between TIRM and








EWLS/3DOT are the determination of particle-surface separation distance from the

scattered light intensity of a particle in the vicinity of the evanescent wave and the use of

an optical trap for micromanipulation of the particle. The principal difference of this new

technique described herein is the use of a three-dimensional laser trap as a force

transducer which provides faster direct measurement of forces on smaller particles than

has been reported with TIRM. Furthermore, because the range of measurable separation

distances is determined by maximum measurable force, not by potential energy sampled

by a Brownian particle, a significantly large range of separation distances can be

measured. This technique's force measurement methodology is similar to the AFM in

that it acts as a force transducer with the particle acting as a probe, rather than deriving

potential energy from position histograms as with TIRM.

AFM is a technique, which successfully measures forces between a single particle

and a surface. However, EWLS/3DOT offers a novel alternative method for determining

forces impossible to calculate by AFM because of its sensitivity limitations. In many

ways, this new technique is complimentary to AFM measurements, and should provide a

more complete picture of the interaction forces especially for weaker longer-range forces.

Another benefit of EWLS/3DOT over other existent techniques is its ability to measure

forces on biological cells such as micron-sized bacteria. The ability of the technique to

resolve small forces along with its non-intrusive (i.e., sterile) manipulation of the particle

make it ideal for measuring biological systems. Certainly great care must be taken to

ensure that the optical trap does not damage the cells. Trapping lasers in the near infrared

wavelength have been shown to be ideal for such applications (Block 1992).








The other noteworthy feature of the EWLS/3DOT technique is its ease of

implementation. The complete system consists of only a few modifications to a standard

light microscope, and hence it adapts well to existing microscopy set-ups allowing for

multiple functionality.














CHAPTER 7
SUMMARY AND CONCLUSIONS




The general goal of this work has been to enhance the fundamental understanding

of the physicochemical mechanisms of bacterial adhesion to solid surfaces through the

development of novel methodologies and through quantitative measurement of the role of

cell surface properties on attachment. Better understanding of bacterial adhesion

mechanisms could ultimately result in the design of an infection resistant biomaterial, or

of a technique that would prevent pathogenic microorganism attachment to the surface of

implanted biomaterials.

As it was previously described, bacteria can adhere to surfaces via non-specific

colloidal forces and via specific binding of cell surface receptors to adsorbed proteins on

the biomaterial surface. It was shown previously that attachment of S. aureus to

fibrinogen coated surfaces in well-defined flow conditions is primarily governed by

specific binding of the cell surface protein, "clumping factor", to fibrinogen (Dickinson et

al. 1995). Clumping factor has a hydrophilic stalk-like region, which has been

hypothesized to enhance its function by allowing the molecule to reach over energy

barriers and allow attachment. In this study, the attachment kinetics of S. aureus to

surfaces was quantified as a function of the length of the clumping factor stalk region. In

this thesis, intrinsic attachment rate constants to fibrinogen coated surfaces were








compared for a series of site-directed S. aureus mutants with variable lengths of the stalk

region.

To quantify the ability of bacteria attach to surfaces, a methodology and analysis

was developed for the measurement of bacterial attachment kinetics under well-defined

flow conditions using automated image analysis. These measurements, combined with a

model for convective-diffusive cell transport in the flow cells, allowed estimation of an

intrinsic attachment rate constant. This methodology can be directly extended to

measuring the attachment kinetics of micron sized particles to various coated substrates.

By this means intrinsic attachment rate constants (k+) can be obtained for particles

attaching to surfaces coated with substrates of interest in filters designed to separate

bacteria from waste water streams. This parameter, k+, is necessary in existing filtration

models; until now these attachment rate constants were approximated from the DLVO

theory.

This methodology and analysis was applied to demonstrate the effect of

sequentially varying one molecular parameter, receptor length, on the attachment of

bacteria to surfaces under well defined, laminar flow conditions. The results of this study

using S. aureus mutants of varying receptor lengths are unique and they prove that

receptor length plays a significant role in adhesion, consistent with predictions of a

mechanistic mathematical model (Dickinson 1997). The first set of experiments reported

attachment of S. aureus on fibrinogen using bovine serum albumin (BSA) as the non-

specific adhesion blocker. An increase in the attachment rate constant with increase in the

length of the repeat region of the clumping factor was observed. The length of the

clumping factor repeat region plays an important role in increasing adhesion of S. aureus








to fibrinogen coated surfaces. This receptor may act as a stalk to project the ligand-

binding domain of S. aureus (the biologically active fibrinogen-binding region) away

from the cell surface to 'reach' the ligand-coated surface. Another set of results were

obtained for attachment of S. aureus to fibrinogen coated surfaces, however, instead of

using BSA as the non-specific adhesion blocker, skim milk was substituted. The results

show the same trend as the previous ones, i.e., an increase in k, with increase in the

receptor length, however, the values of k+ are different for the two sets of experiments. k+

values for the experiments using skim milk were smaller than the values obtained when

BSA was used as the non-specific adhesion blocker. A possible explanation for this

could be that the skim milk affects the nonspecific interactions, specifically the repulsive

interactions. An increase in the repulsive interactions would cause the observed decrease

in k+. Results from adhesion control experiments on skim milk varying the receptor

length show a decrease in k+ values with increase in the receptor length. This is an

interesting result since it shows that the steric repulsion forces may increase as the length

of the receptor increases. In the control experiments only background binding onto skim

milk takes place, but as the receptor length increases it experiences a higher repulsion

close to the surface, and consequently the attachment rate would decrease as observed.

Another study was undertaken in an effort to determine the effect on S. aureus

attachment to glass as the electrolyte concentration was changed. It was shown that the

increasing electrolyte concentration enhances attachment of S. aureus to glass surfaces,

qualitatively consistent with predictions based on DLVO theory.








EWLS/3DOT is a novel technique for measuring interaction forces between small

particles and surfaces, and has provided the first direct measurement of the forces

between a single bacterium and a solid substrate. This technique will allow future

correlation between interaction forces and attachment rates.

In conclusion, the results of the work presented here show that the receptor length

is a critical parameter in the attachment of bacteria to surfaces. In specific attachment

experiments the increase in receptor length increases attachment of bacteria to surfaces,

thus, the specific receptor-ligand interactions. In control experiments where the ligands

are absent from the surface, non-specific steric repulsive interactions come into play to

decrease attachment with increase in the receptor length. The mathematical model for

particle attachment to the study surface in the PPFC predicted a dependence of the

attachment rate on shear. In preliminary experiments, shear was observed to have an

effect on bacterial attachment in accordance with the model predictions. Future

experiments studying in more depth the effect of shear will shed more light on this. Also

in the future, once the EWLS/3DOT methodology will be well developed it will allow

measurement of adhesion forces between S. aureus mutants of varying lengths and

fibrinogen coated surfaces.

This work in this dissertation has provided new tools for measurement of bacteria-

surface interactions and enhanced the fundamental understanding of specific and non-

specific bacterial attachment mechanisms. A combination of measurements of bacterial

attachment kinetics and bacteria-surface interaction forces may lead to a more clear

understanding of the mechanisms of bacterial adhesion to surfaces and hence, may lead to

techniques for prevention of this process or its enhancement, depending upon the






81

application. This combined approach may aid in the design of infection-resistant

biomaterials or in the practical application to improve filter design for separation of

bacteria from wastewater streams. However, the techniques can be applied to enhance

fundamental understanding of bacterial adhesion relevant to other applications such as

tooth decay and biofouling.














APPENDIX
IMAGE ANALYSIS PROGRAMS

This appendix contains three image analysis programs: COUNT.MAC,

SCAN.MAC, and BREPLAY.MAC. The first one is used in obtaining the particle

concentration in the suspension used for the parallel plate flow cell (PPFC) experiments

as well as the error in measuring it. The second one is used to store to the hard drive in a

Microsoft EXCEL file successive images of particles attaching on the bottom surface of

the PPFC. After the experiments are completed, the stored images from the EXCEL files

are recalled and the number of particles in each image is counted using the third program,

BREPLAY.MAC.

Following are the three programs briefly described above. They are all written in

the Analytical Language for Images (ALI) used within the image analysis program,

Optimas.



// File name: COUNT.MAC

/*This macro counts cells (particles) in the Petroff-Hauser counting chamber, and returns
a cell (particle) concentration and the error in measuring it. This is used later in
determining an effective attachment rate constant for particles. */

// May 1998 last update

Acquire;
Brightness(150);
Contrast(150);
ans=Prompt("Has calibration been set? If not, click , set calibration from menu,
then restart macro.",2); if(!ans) {pauseQ;};









hLib = LoadMacroLibrary("surface.oml");
RunMacro ("C:/optimas/macros/flowcell/ninit.mac");

// Commands to clear the square in the middle of the screen

RunMacro("C:/optimas/macros/flowcell/bcomml.mac"); // Loads motors function
// library
mwindowo;

// End commands for clearing the square in the middle of the screen

SelectFullScreen (0);

INTEGER rconc[];
tier=-.01;
INTEGER nf=0;

ans=TRUE;
while (ans)
{
// SelectFullScreen (0);
Acquire();
show("Move to new field, click OK, then select ROI with mouse
(left button only!).");
SelectROI0;
AreaF=(ROI[0,1]-ROI[l,1])*(ROI[l,0]-ROI[0,0]); //Field area in active units
SetExport (mPtPoints, 1, TRUE);

if(nf--=0) {Freeze(); diff= SetThresh(); Acquireo;}
AutoExtract = TRUE;
Freeze();
nsh=CountCells(diff);
rconc=rconc:((1.*nsh)/(AreaF*(1 e-4)*(le-4))/tiefl e6);
stats=Moments(rconc);
conc=stats[0]; stdev=stats[l1]; unc=0.0;
if(nf>0) {stdev=stdev/sqrt(l1.*nf); unc=stdev* 100/conc;}
nf=-nf+ 1;
show(ToText(nf):" fields. ":ToText(conc,"%5.2f'):" +/-
:ToText(stdev,"%5.2f'):" xlO^6 cells/ml. ":ToText(unc,"%5.2f'):" % uncertainty");
ans=Prompt("Another Field?",2);
}


show(ToText(nf):" fields. ":ToText(conc,"%5.2f"):" +/- ":ToText(stdev,"%5.2f'):" xlO^6
cells/ml. ":ToText(unc,"%5.2f'):" % uncertainty");
// End of Program COUNT.MAC









// File name: SCAN.MAC

// This macro scans multiple fields, either user-selected or by a raster scan, then stores
//the images of the fields to hard disk. An excel file is prepared with the image
// and field information. The images must be analyzed and data extracted in a second
// program (e.g. BREPLAY.MAC).

// May 1998 last update

// Initialize
acquire;
OutputLUT ("Normal", 0.0);
Brightness(150);
Contrast( 150);
ans=Prompt("Has calibration been set? If not, click , set calibration from menu,
then restart macro.",2); if(!ans) pauseeo;;

RunMacro("C:/optimas/macros/flowcell/bcomml.mac"); // Loads motors function
library
hlib=LoadMacroLibrary("optdlg.oml");
mwindowo;

SelectFullScreen (0);
beep(40,32);

hLib = LoadMacroLibrary("surface.oml");
// RunMacro("C:/OPTIMAS/MACROS/flowcell/FOCPLANE.MAC"); / Loads
functions for calculating // focal plane
RunMacro("C:/OPTIMAs/MACROS/flowcell/ninit.mac"); // Loads various functions

// Open Excel
show("Open Excel then Click OK");
fw=ROI[l,0]-ROI[0,0]; fl=ROI[0,lj-ROI[l,l]; I/Field width and length

Inami=Prompt("Enter Image Prefix","CHAR","Img");
sheet="Sheet V"; hChan=DDEInitiate("Excel",sheet);
AreaF=fw*fl; // Field area in active units

GetDateTimeO;
(Rng=Range(1,1,1,1)); DDEPoke (hchan,Rng,DateTime);

Rng=Range(2,1,2,1); DDEPoke (hchan,Rng,"Field Area:");
Rng=Range(2,2,2,2); DDEPoke (hchan,Rng,AreaF);


Rng=Range(2,5,2,5); DDEPoke (hchanRng,"Bacteria Cone:");








cconc=Prompt("Enter Bacteria Concentration: ","REAL","1 .0e6");
Rng=Range(2,6,2,6); DDEPoke (hchan,Rng,cconc);

show("With joystick, move to center of scan area then click OK to focus with arrow
keys");
Here(OL,OL,OL); //Sets reference point position to 0,0,0

nt= prompt("Enter number of time steps ","INTEGER","30");
mode=Prompt("Select scan mode; (0=User Selected Fields, 1= Raster
Scan)","INTEGER"," 1");

LONG x[100],y[100],z[100],pos[3],posl [3],pos2[3],pos3[3],postmp[3];

if(mode--==0)
{
nff=prompt("Input number of fields","INTEGER");

// Set Raster and Focus

/* Input Field Coordinates */
i=0;
while (i {
beep(40,32);
text="Move to field ":ToText(i):" and set focus with joystick";
show(text);
pos=--findpos(); // Finds current position
x[i]=pos[0]; // Store current position of field i to arrays x,y,z
y[i]=pos[l];
z[i]=pos[2];
// Store position info to spreadsheet
Rng=Range(3,3+i,3,3+i); DDEPoke (hchan,Rng,x[i]);
Rng=Range(4,3+i,4,3+i); DDEPoke (hchan,Rng,y[i]);
Rng=Range(5,3+i,5,3+i); DDEPoke (hchan,Rng,i);
i=i+l;
}
}
if(mode==l) // Set raster scan area by defining opposite comers of rectangle
{

beep(40,32);
show("Move to top limit of scan area with joystick and focus");
posl--findposO;
movabs(OL,OL,OL); WaitO;
show("Move to inlet position with joystick and focus");
pos2--=findposO;








LONG xl=2L*posl [0]; //Length of scan area
LONG yl=2L*pos2[1]; // Breadth of scan area
LONG zll=2L*posl[2]; //Height at inlet area
LONG zl2=2L*pos2[2]; // Height at top area
movabs(OL,OL,OL); WaitQ; movrel(OL, OL, 1000L);
pos3=OL:OL:OL;
// Plane=CalcPlane(posl,pos3,pos2); // This functions calculates parameters for
calculating focal plane
nxx=Prompt("Input number of fields in the x-direction (downstream
direction)","INTEGER");
nyy=Prompt("Input number of fields in the y-direction integerGER");
nff=-nxx*nyy;
LONG dx=xl/(lL*(nxx-1));
LONG dy=yl/(lL*(nyy-1));
LONG dzl=zll/(1L*(nxx-1));
LONG dz2=zl2/(1L*(nyy- 1));
i=0;
while(i {
j=0;
while(j {
ifld=i*nyy+j; // Changes double index (ij) to single index, ifld
if(nxx>l) {x[ifld]=posl[0]+dx*(i-nxx/2); } // Calculates field positions
if(nxx=l) {x[ifld]=posl[0];}
if(nyy>l 1) {y[ifld]=posl [1 ]+dy*(j-nyy/2);}
if(nyy==l) {y[ifld]=posl[l ];}
if (nyy>l) {z[ifld]=posl [2]+dzl *(i-nxx/2)+dz2*(j-nyy/2);}
if(nyy==l) {z[ifld]=posl [2];}
j=j+l;
}
i=i+l;
}
}

ifld=0;
while(ifld {
xp=x[ifld]; yp=y[ifld]; zp=z[ifld]; //zp=CalcZ(xp,yp,Plane);
show(xp,yp,zp);
movabs(xp,yp,zp); Waito;
txt="Field ":ToText(ifld);
show("Correct focus for": txt);
postmp=findpos0;
z[ifld]=postmp[2]; // Corrects focus
// Store field coordinates to spreadsheet








Rng=Range(3,3+ifld,3,3+ifld); DDEPoke (hchan,Rng,x[ifld]);
Rng=Range(4,3+ifld,4,3+ifld); DDEPoke (hchan,Rng,y[ifld]);
Rng=Range(5,3+ifld,5,3+ifld); DDEPoke (hchan,Rng,ifld);
ifld=ifld+ 1;
}
}
// Store Images

INTEGER foc[nffJ=l; // foc[i]=l for delay after Acquire for focus check
/* Start Scanning Selected Fields */

it=0; StartTime=DOSTime; // Calculates initial time

while(it {
i=o;
Tm=DOSTimeo-StartTime; // Current time
Rng=Range(6+it,l,6+it,1); DDEPoke (hchan,Rng,it); //Store time
Rng=Range(6+it,2,6+it,2); DDEPoke (hchan,Rng,Tm);
while (i {
xp=x[i];
yp--y[i];
zp=z[i];
movabs(xp,yp,zp); Waito;
Acquire();
if(KeyHit()o=18 ) {beep(50,32); show("Pausing Click OK to Continue");} // Alt
Key
iw=0;
while ((foc[i]=l) && ((keyhitO!=Oxl 0) && (iw<15))) //Allow chance to check
focus if foc[i]=l
{DelayMS(100); iw=iw+1 ;}
if ((keyhit0==0xl0) II (foc[i]=2)) // key causes pause and refocus if not
in focus
{
Beep(50,32);
txt="Field ":ToText(ifld);
show("Correct focus for" : txt);
postmp = findposo;
z[i]=postmp[2]; // Reset z-position
ifoc=l;
foc[i]=l; // Keep at 1 if focusing was necessary








else
{foc[i]=0;}
Iname=Inami:ToText(i):"_":ToText(it); FreezeO;
IRoute="C:/OPTIMAs/MACROS/flowcell/images/":Iname:".tif';
Savelmage (Iroute, ROI, 8,,,"", "", "", TRUE, 0, FALSE); // Save image
Rng=Range(6+it,3+i,6+it,3+i); DDEPoke (hchanRng,Lname);
if (KeyHit(O==Oxl 0)
{beep(30,32); foc[i]=2;} // Key here allows this field to be focused next
time
if (foc[i]<2)
{i=i+l;}
}
it-it+l;
}

DDETerminate(hchan); // Close connection to excel

// End of Program SCAN.MAC



// File name: BREPLAY.MAC

// This macro replays the image sequence taken by SCAN.MAC to extract statistics for
// bacteria. It is written in the ALI (Analytical Language for Images) language of
// Optimas.

// May 1998 last update

// Start Initialization

ans=Prompt("Has calibration been set? If not, click , set calibration from menu,
then restart macro.",2); if(!ans) {pause();};
show("Open Excel Datafile then click OK");
RunMacro ("C:/optimas/macros/flowcell/ninit.mac"); // Contains some handy functions
show("After clicking OK, open a test image from the sequence");
Openlmage0; //Opens TestImage for Thresholding
diff--SetThresh(); // Sets thresholding parameters
AutoExtract = TRUE;

// End Initiation


nt= prompt("Enter number of time steps (rows) ","INTEGER","30");
nfst= prompt("Enter first field number (columns) ","INTEGER","0");
nlst= prompt("Enter last field number (columns) ","INTEGER","24");








CHAR Inst=prompt("Enter file prefix (e.g. Imgxx.tif, file prefix is
Img).","CHAR","Img");
it=0; // Set it corresponding to first used time step for data.
while(it {
i=nfst;
while (i<---nlst)
{
Rng=Range(6+it,3+i,6+it,3+i);
CHAR Iname=Inst:ToText(i):"_":ToText(it);
IRoute="C:/OPTIMAs/MACROS/FLOWCELL/IMAGES/":Iname:".tif;
Openlmage (IRoute, ROI,);
nb=CountCells(diff);
sheet="Sheetl "; hChan=DDEInitiate("Excel",sheet);
DDEPoke(hchan,Rng,nb);
DDETerminate(hchan);
i=i+l;


// End Data Extraction Routine
it=it+l;
}End of Program BREPLAYMAC
II End of Program BREPLAY.MAC













LIST OF REFERENCES


Adamczyk, Z., and T.G.M.,Van de Ven, "Deposition of Particles Under External Forces
in Laminar Flow through Parallel-Plate and Cylindrical Channels,"J Colloid
Interface Sci., 89, 497 (1981).

Adamczyk, Z., T., Dabros, J.,Czarnecki, and T.G.M., Van de Ven, "Particle Transport to
Solid Surfaces," Adv. Colloid Interface Sci., 19, 183 (1983).

Adamczyk, Z., M., Zembala, B., Siwek, and J. Czarnecki, "Kinetics of Latex Deposition
from flowing suspensions," J. Colloid Interface Sci., 110, 188 (1986).

Atchinson, J., The Lognormal Distribution, with Special References to Its Uses in
Economics, Academic Press, Cambridge (1957).

Baier, R., A.E. Meyer, J.R. Natiella, R.R. Natiella, J.M Carter, "Surface Properties
Determine Bioadhesive Outcomes: Methods and Results," J. Biomed Mater. Res.
18, 337 (1984).

Bell, G.I., M.Dembo, P.Bongrad. "Cell adhesion. Competition between Nonspecific
Repulsion and Specific Bonding," Biophys. J., 45, 1051 (1984).

Bisno, A.L. & F.A. Waldvogel Eds., "Infections Associated with Indwelling Medical
Devices," Am. Soc. Microbiol., Washington, D.C. (1989).

Block, S.M., "Making Light Work with Optical Tweezers," Nature, 360,493 (1992).

Boden, M.K., and J.I. Flock, "Fibrinogen-Binding Protein/Clumping Factor from
Staphylococcus aureus," Infect. Immun. 57,2358 (1989).

Bowen, B.D., "Streaming Potential in the Hydrodynamic Entrance Region of Cylindrical
and Rectangular Capillaries," J. Colloid Interface, 106(2), 367 (1985).

Bowen, B. D., and N., Epstein, "Fine Particle Deposition in Smooth Parallel-Plate
Channels," J. Colloid Interface Sci., 72, 81 (1979).

Brown, L., M. Westby, B.E. Souberbielle, P.W. Szawlowski, G. Kemp, P. Hay, A.G.
Dalgleish, "Optimization of a Petptide-Based Indirect ELISA for the Detection of
Antibody in the Serum of HIV-1 Seropositive Patients," J. Immunol. Methods,
200(1-2), 79 (January 15 1997).




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EBBGMZUBU_YOWQ67 INGEST_TIME 2014-05-23T23:29:57Z PACKAGE AA00020703_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

48$17,7$7,9( 678'< 2) 0(&+$1,606 2) %$&7(5,$/ $'+(6,21 72 685)$&(6 %\ $/,1$ *$%5,(/$ 587$ $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

&RS\ULJKW E\ $OLQD *DEULHOD 5XWD

PAGE 3

7KLV 3K' GLVVHUWDWLRQ LV GHGLFDWHG ZLWK DOO P\ KHDUW WR P\ SDUHQWV HVSHFLDOO\ WR P\ PRWKHU ZKR FRQWLQXDOO\ UHPLQGV PH ZKDW GHWHUPLQDWLRQ DQG FKDUDFWHU DUH DQG ZKR JLYHV PH VWUHQJWK WKURXJK KHU FRXUDJH :LWKRXW P\ SDUHQWVf XQIDLOLQJ ORYH VDFULILFH DQG JXLGDQFH ZRXOG QRW EH ZULWLQJ WKLV WRGD\

PAGE 4

$&.12:/('*0(176 ZRXOG OLNH WR DFNQRZOHGJH ILUVW RI DOO WKH VXSSRUW RI P\ DGYLVRU 'U 5LFKDUG 'LFNLQVRQ ZKR KDV JXLGHG PH WKURXJKRXW P\ 3K' ZRUN DQG IURP ZKRP ,f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

PAGE 5

7$%/( 2) &217(176 SDJH $&.12:/('*0(176 LY /,67 2) ),*85(6 YLL $%675$&7 L[ ,1752'8&7,21 %$&7(5,$/ $'+(6,21 0ROHFXODU 6WUXFWXUH RI %DFWHULDO &HOO 6XUIDFHV 1RQ6SHFLILF &ROORLGDOf 0HFKDQLVPV RI %DFWHULDO $GKHVLRQ 6SHFLILF 0HFKDQLVPV RI %DFWHULDO $GKHVLRQ 0HWKRGV 8VHG WR 0HDVXUH %DFWHULDO $GKHVLRQ LQ 9LWUR 67$3+
PAGE 6

0$7+(0$7,&$/ 02'(/ )25 3$57,&/( 75$163257 72 7+( %27720 $77$&+,1* 685)$&( ,1 7+( 3$5$//(/ 3/$7( )/2: &(// 0RGHOLQJ 3DUWLFOH $WWDFKPHQW .LQHWLFV 0RGHO 'HVFULSWLRQ 'LPHQVLRQOHVV (TXDWLRQV 0RGHO 6ROXWLRQ (9$1(6&(17 :$9( /,*+7 6&$77(5,1* &283/(' :,7+ 7+5(( ',0(16,21$/ 237,&$/ 75$33,1* ,QWURGXFWLRQ $SSDUDWXV 3ULQFLSOH RI 2SHUDWLRQ )RUFH 0HDVXUHPHQW 'DWD $QDO\VLV 2YHUYLHZ &DOFXODWLQJ 3DUWLFOH 3RVLWLRQ &DOLEUDWLQJ WKH 2SWLFDO 7UDS &DOFXODWLQJ )RUFH DQG 3RWHQWLDO 3UHOLPLQDU\ 5HVXOWV 'LVFXVVLRQ 6800$5< $1' &21&/86,216 ,0$*( $1$/<6,6 352*5$06 /,67 2) 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ YL

PAGE 7

/,67 2) ),*85(6 )LJXUH SDJJ )LJXUH %DFWHULDO FHOO ZDOO VWUXFWXUHV )LJXUH '/92 LQWHUDFWLRQ SRWHQWLDO )LJXUH $ VFKHPDWLF RI UHFHSWRUOLJDQG LQWHUDFWLRQV SURPRWLQJ DGKHVLRQ )LJXUH 6FDQQLQJ (OHFWURQ 0LFURJUDSK RI 6WDSK\ORFFRFXV DXUHXV )LJXUH 6WUXFWXUH RI FOXPSLQJ IDFWRU )LJXUH 6FKHPDWLF RI SDUDOOHO SODWH IORZ FHOO GHVLJQHG fLQKRXVHf )LJXUH 6FKHPDWLF RI SDUDOOHO SODWH IORZ FHOO )LJXUH 6FKHPDWLF RI WKH DXWRPDWHG YLGHR PLFURVFRS\ V\VWHP )LJXUH 4XDQWLILFDWLRQ RI DWWDFKPHQW NLQHWLFV )LJXUH 3URFHGXUH IRU FDOFXODWLQJ NH-J[f )LJXUH 6 DXUHXV PXWDQWV ZLWK YDULDEOH OHQJWK RI WKH UHSHDW UHJLRQ XVHG LQ 33)& H[SHULPHQWV )LJXUH (IIHFW RI UHFHSWRU OHQJWK RQ EDFWHULDO DWWDFKPHQW NLQHWLFV LQ WKH 33)& )LJXUH &RPSDULVRQ RI WKH PHFKDQLVWLF PRGHO SUHGLFWLRQV WR PHDVXUHG YDOXHV RI N )LJXUH 0HFKDQLVWLF PRGHO UHVXOWVf§HIIHFW RI HQHUJ\ EDUULHU )LJXUH 0HFKDQLVWLF PRGHO UHVXOWVf§HIIHFW RI WKH PDJQLWXGH RI QRQVSHFLILF UHSXOVLYH IRUFHV )LJXUH 3ORW RI N YV UHFHSWRU OHQJWK IRU 6 DXUHXV PXWDQWV RQ )J DQG VNLP PLON FRQWURO VXUIDFHV )LJXUH $ SORW RI LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW NW YV HOHFWRO\WH FRQFHQWUDWLRQ )LJXUH ,OOXVWUDWLRQ RI WKH FRQYHFWLYHGLIIXVLYH PRGHO IRU SDUWLFOH WUDQVSRUW LQ 33)& YLL

PAGE 8

)LJXUH 3ORW RI WKH HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW YV GLPHQVLRQOHVV GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH GLPHQVLRQOHVV LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW D V NY )LJXUH 3ORW RI WKH GLPHQVLRQOHVV DWWDFKPHQW UDWH FRQVWDQW DV D IXQFWLRQ RI GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH WKH GLPHQVLRQOHVV IORZ SDUDPHWHU )LJXUH 3ORW RI WKH GLPHQVLRQOHVV DWWDFKPHQW UDWH FRQVWDQW DV D IXQFWLRQ RI GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH WKH GLPHQVLRQOHVV ERXQGDU\ OD\HU WKLFNQHVV \ )LJXUH 3ORW RI WKH GLPHQVLRQOHVV N DQG DYHUDJH NHI DV D IXQFWLRQ RI VKHDU UDWH IRU D 6 DXUHXV PXWDQW ZLWK D PLVVLQJ 5 UHJLRQ )LJXUH 6WUXFWXUDO VFKHPDWLF RI NH\ FRPSRQHQWV RI WKH (:/6'27 DSSDUDWXV )LJXUH 6FKHPDWLF RI WKH SULQFLSOH RI (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ )LJXUH (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ RI 6WDSK\ORFRFFXV DXUHXV )LJXUH 6FKHPDWLF RI WKH WKUHHGLPHQVLRQDO RSWLFDO WUDS )LJXUH 6FKHPDWLF RI IRUFHV LQ WKH WKUHH GLPHQVLRQDO RSWLFDO WUDS )LJXUH +LVWRJUDPV IRU D WUDSSHG SP SRO\VW\UHQH SDUWLFOH )LJXUH 7UDS FDOLEUDWLRQ DQG PHDVXUHPHQWV XVLQJ WKH (:/6'27 WHFKQLTXH )LJXUH )RUFH DQG SRWHQWLDO HQHUJ\ FXUYHV IRU D SP SRO\VW\UHQH SDUWLFOH LQWHUDFWLQJ ZLWK JODVV LQ ZDWHU YP

PAGE 9

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 48$17,7$7,9( 678'< 2) 0(&+$1,606 2) %$&7(5,$/ $'+(6,21 72 685)$&(6 %\ $OLQD *DEULHOD 5XWD $XJXVW &KDLUPDQ 5LFKDUG % 'LFNLQVRQ 0DMRU 'HSDUWPHQW &KHPLFDO (QJLQHHULQJ %LRPDWHULDODVVRFLDWHG LQIHFWLRQV LPSHGH WKH ORQJWHUP XVH RI LPSODQWHG DQG LQWUDYDVFXODU GHYLFHV %DFWHULDO DGKHVLRQ WR WKH SURWHLQFRDWHG VXUIDFHV RI LPSODQWHG ELR PDWHULDOV DQG VXEVHTXHQW FDWDVWURSKLF UHVLVWDQW LQIHFWLRQ DUH PDMRU EDUULHUV WR H[WHQGHG XVH RI WKHVH GHYLFHV %HFDXVH RI WKH VL]H DSSUR[LPDWHO\ SP LQ GLDPHWHUf EDFWHULD FDQ EH FRQVLGHUHG FROORLGDO SDUWLFOHV 7KHUHIRUH VWXGLHV RI EDFWHULDVXUIDFH LQWHUDFWLRQV KDYH EHHQ LQ WHUPV RI FODVVLFDO FROORLGDO WKHRULHV DQG DFFRXQWHG RQO\ IRU WKH QRQVSHFLILF LQWHUDFWLRQV VXFK DV K\GURSKRELF 9DQ GHU :DDOV DQG VWHULF VWDELOL]DWLRQ +RZHYHU GXH WR EDFWHULDfV FRPSOH[ VXUIDFH VWUXFWXUH DQG LQKHUHQW PHWDEROLVP VSHFLILF UHFHSWRUOLJDQG W\SH LQWHUDFWLRQV DUH LPSRUWDQW LQ EDFWHULDO DGKHVLRQ +HQFH D FRPSUHKHQVLYH XQGHUVWDQGLQJ RI WKH DGKHVLRQ PHFKDQLVPV RI EDFWHULD WR SURWHLQ FRDWHG VXUIDFHV FDQ EH REWDLQHG RQO\ LI WKH GXDO QDWXUH RI WKH EDFWHULDVXUIDFH LQWHUDFWLRQV LV FRQVLGHUHG ,;

PAGE 10

7KLV GLVVHUWDWLRQ GHVFULEHV D QRYHO PHWKRG IRU PHDVXULQJ EDFWHULDO DWWDFKPHQW NLQHWLFV ZLWK WKH DLG RI DXWRPDWHG YLGHR PLFURVFRS\ DQG LPDJH DQDO\VLV LQ ZHOOGHILQHG ODPLQDU IORZ FRQGLWLRQV XVLQJ D SDUDOOHO SODWH IORZ FHOO 33)&f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

PAGE 11

&+$37(5 ,1752'8&7,21 ,QIHFWLRQ RI LPSODQWHG DQG LQWUDYDVFXODU GHYLFHV LV SRWHQWLDOO\ OLIH WKUHDWHQLQJ DQG LPSHGHV WKH ORQJWHUP XVH RI PDQ\ ELRPHGLFDO GHYLFHV 'DQNHUW HW DO +HQFK 6XJDUPDQ t
PAGE 12

9DXGDX[ HW DO f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f§UHFHSWRUV DUH PROHFXOHV RQ WKH FHOO VXUIDFH FDSDEOH RI ELQGLQJ RWKHU PROHFXOHV RQ WKH VXUIDFH RI D PDWHULDO RU DQRWKHU FHOO WHUPHG OLJDQGVf DQG QRQVSHFLILF PHFKDQLVPV FROORLGDO LQWHUDFWLRQVf PD\ EH LQYROYHG 7KHUHIRUH LW LV HTXDOO\ LPSRUWDQW WR GHYHORS D WHVWHG WKHRUHWLFDO EDVLV IRU LQWHUSUHWDWLRQ RI DFTXLUHG DGKHVLRQ GDWD DQG WR EH DEOH WR SUHGLFW WKH SUREDELOLW\ RI DGKHVLRQ RFFXUULQJ XQGHU D JLYHQ VHW RI ORFDO FRQGLWLRQV 7KH JRDO RI WKH ZRUN GHVFULEHG LQ WKLV WKHVLV LV WR HQKDQFH IXQGDPHQWDO XQGHUVWDQGLQJ RI PLFURELDO DGKHVLRQ PHFKDQLVPV WR VXUIDFHV YLD TXDQWLWDWLYH PHDVXUHPHQWV 7KLV XQGHUVWDQGLQJ PD\ DLG LQ GHVLJQLQJ LQIHFWLRQUHVLVWDQW ELRPDWHULDOV RU LQ GHYHORSLQJ WHFKQLTXHV WR SUHYHQW EDFWHULDO LQIHFWLRQ RI LPSODQWHG LQWUDYDVFXODU DQG ELRPHGLFDO GHYLFHV ,W PD\ DOVR DLG LQ GHVLJQLQJ VXUIDFHV IRU EHWWHU PLFUREH UHPRYDO LQ ZDVWHZDWHU WUHDWPHQW 7KLV GLVVHUWDWLRQ GHVFULEHV WKH GHYHORSPHQW RI QRYHO TXDQWLWDWLYH H[SHULPHQWDO WHFKQLTXHV IRU VWXG\LQJ PLFURELDO DWWDFKPHQW WR VROLG VXUIDFHV DQG WKH DSSOLFDWLRQ RI WKHVH WHFKQLTXHV WR VWXG\ WKH UROH RI VSHFLILF DQG QRQVSHFLILF LQWHUDFWLRQV LQ EDFWHULDO DWWDFKPHQW &KDSWHU SURYLGHV D EULHI EDFNJURXQG RQ WKH FXUUHQW VWDWH RI NQRZOHGJH ZLWK UHVSHFW WR ELRORJ\ DQG SK\VLFV XQGHUO\LQJ EDFWHULDO DGKHVLRQ PHFKDQLVPV DQG WKH

PAGE 13

FRPPRQ PHDVXUHPHQW WHFKQLTXHV XVHG WR VWXG\ PLFURELDO DGKHVLRQ &KDSWHU GLVFXVVHV WKH FKDUDFWHULVWLFV RI WKH VSHFLHV XVHG LQ WKLV VWXG\ 6WDSK\ORFRFFXV DXUHXV &KDSWHU GLVFXVVHV WKH PHWKRGRORJ\ XVHG WR GLUHFWO\ PHDVXUH UHDOWLPH EDFWHULDO DWWDFKPHQW NLQHWLFV XQGHU ZHOOGHILQHG IORZ FRQGLWLRQV DQG WKH UHVXOWV RI D VWXG\ RI WKH GHSHQGHQFH RI EDFWHULDO DWWDFKPHQW RQ WKH OHQJWK RI D ILEULQRJHQELQGLQJ PROHFXOH fFOXPSLQJ IDFWRUf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

PAGE 14

&+$37(5 %$&7(5,$/ $'+(6,21 0ROHFXODU 6WUXFWXUH RI %DFWHULDO &HOO 6XUIDFHV %DFWHULD DUH FODVVLILHG DFFRUGLQJ WR WKHLU UHDFWLRQ ZLWK WKH *UDP VWDLQ WKH\ FDQ EH FODVVLILHG HLWKHU DV *UDP QHJDWLYH RU *UDP SRVLWLYH 'DQNHUW HW DO f $ VFKHPDWLF RI FHOO ZDOOV RI *UDP QHJDWLYH DQG *UDP SRVLWLYH EDFWHULD FDQ EH VHHQ LQ )LJXUH D t Ef *UDP QHJDWLYH EDFWHULD KDYH WKHLU SODVPD PHPEUDQHV VXUURXQGHG E\ D SHSWLGRJO\FDQ OD\HU 7KH RXWHU PHPEUDQH D SKRVSKROLSLG ELOD\HU LV DQFKRUHG LQWR WKH SHSWLGRJO\FDQ OD\HU E\ D OLSRSURWHLQ )LJXUH Dff &DSVXODU SRO\VDFFKDULGH DQG VRPH RXWHU PHPEUDQH SURWHLQV FDQ H[WHQG IURP WKH RXWHU PHPEUDQH )ODJHOOD ILPEULDH RU FDSVXOHV FDQ DOVR H[WHQG IURP WKH RXWHU PHPEUDQH 7R EH QRWHG LV WKDW WKH SHSWLGRJO\FDQ OD\HU LV HPEHGGHG DV D VHSDUDWH OD\HU ZLWKLQ WKH RXWHU PHPEUDQH RI *UDP QHJDWLYH EDFWHULD WKXV SUHYHQWLQJ LW IURP LQWHUDFWLQJ GLUHFWO\ ZLWK PROHFXOHV RQ WKH VXUIDFH RU LQ LWV SUR[LPLW\ $V VHHQ LQ )LJXUH Ef *UDP SRVLWLYH EDFWHULD KDYH D UHODWLYHO\ WKLFN RXWHU ZDOO FRQVLVWLQJ RI SHSWLGRJO\FDQ WHLFKRLF DQG WHLFKXURQLF DFLGV DV ZHOO DV RWKHU SRO\VDFFKDULGHV RU SURWHLQV 7KHVH DUH QRW HQWLUHO\ HPEHGGHG LQ WKH FHOO ZDOO VRPH FDQ DOVR EH H[SRVHG DW WKH VXUIDFH DQG FDQ SOD\ DQ LPSRUWDQW UROH LQ JHQHUDWLQJ DGKHVLRQ WR VXUIDFHV

PAGE 15

, , &DSVXODU 3RO\VDFFKDULGH &DSVXODU 3URWHLQ /LSRSRO\VDFFKDULGH &DSVXOH &HOO :DOO 3KRVSKROLSLG ELOD\HU (Q]\PLF t VWUXFWXUDO SURWHLQV /LSRSURWHLQ 3HSWLGRJO\FDQ ,QQHU 3KRVSKROLSLG EL OD\HU U F\WRSODVPLF ZLWK PHPEUDQH SURWHLQV nPHPEUDQH DQG SURWHDVHV )LJXUH %DFWHULDO FHOO ZDOO VWUXFWXUHV Df &HOO ZDOO VWUXFWXUH RI *UDPQHJDWLYH EDFWHULD Ef &HOO ZDOO VWUXFWXUH RI *UDPn SRVLWLYH EDFWHULD 6RXUFH 'DQNHUW HW DO 6HYHUDO EDFWHULDO VSHFLHV KDYH PDFURPROHFXOHV WKDW FDQ ELQG VSHFLILFDOO\ WR OLJDQGV LQ WKH FHOOfV HQYLURQPHQW FDOOHG fDGKHVLQVf RU fUHFHSWRUVf %RWK VSHFLILF LH UHFHSWRUPHGLDWHGf DQG QRQVSHFLILF FROORLGDOf LQWHUDFWLRQV PD\ EH LPSRUWDQW LQ EDFWHULDO

PAGE 16

DGKHVLRQ WR VROLG VXUIDFHV %XVVFKHU t :HHUNDPS %XVVFKHU HW DO f %DFWHULDO DGKHVLRQ KDV WUDGLWLRQDOO\ EHHQ VWXGLHG TXDQWLWDWLYHO\ LQ WHUPV RI WKHRULHV GHYHORSHG IRU DGKHVLRQ RI FROORLGDO SDUWLFOHV %HOO HW DO 5XWWHU t 9LQFHQW f 7KH VSHFLILF UHFHSWRUPHGLDWHG LQWHUDFWLRQV KDYH EHHQ WUDGLWLRQDOO\ WUHDWHG VHSDUDWHO\ ODUJHO\ TXDOLWDWLYHO\ IURP D ELRFKHPLFDO SRLQW RI YLHZ 6XOODP HW DO /LQGRQ 9HUFHORWWL HW DO 9DXGDX[ HW DO Df 5\GHQ HW DO (VSHUVHQ t &OHPPHQVHQ f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f 0HFKDQLVPV RI %DFWHULDO $GKHVLRQ %DVHG RQ VL]H DSSUR[LPDWHO\ SP LQ GLDPHWHUf EDFWHULD FDQ EH FRQVLGHUHG FROORLGDO SDUWLFOHV 7KHUHIRUH XQWLO QRZ SK\VLFRFKHPLFDO PRGHOV IRU EDFWHULDO DGKHVLRQ KDYH EHHQ SULPDULO\ EDVHG RQ WKHRULHV IRU FROORLGVXUIDFH LQWHUDFWLRQV 0DUVKDOO 0DUVKDOO HW DO 5XFNHQVWHLQ 3HWKLFD 5XWWHU t 9LQFHQW f &ODVVLFDO '/92 WKHRU\ GHULYHG E\ 'HUMDJXLQ /DQGDX 9HUZH\ DQG 2YHUEHHNf DFFRXQWV IRU HOHFWURVWDWLF DQG 9DQ GHU :DDOV IRUFHV EHWZHHQ WKH FHOO DQG D VXUIDFH WR SUHGLFW WKH LQWHUDFWLRQ SRWHQWLDO I!]f DV D IXQFWLRQ RI VHSDUDWLRQ GLVWDQFH ] 'HUDMDJXLQ t /DQGDX

PAGE 17

9HUZD\ t 2YHUEHHN f $ K\SRWKHWLFDO '/92 LQWHUDFWLRQ SURILOH M!]f IRU D SDUWLFOH DQG VXUIDFH RI OLNH FKDUJH LV VKRZQ LQ )LJXUH 6HYHUDO IHDWXUHV RI WKLV SORW DUH QRWHZRUWK\ $W ODUJH GLVWDQFHV QPf WKH SDUWLFOH LV DWWUDFWHG WRZDUG WKH VXUIDFH E\ ORQJUDQJH YDQ GHU :DDOV IRUFHV DQG UHSHOOHG E\ DQ HOHFWURVWDWLF IRUFH FUHDWHG E\ WKH RYHUODS RI D GRXEOH OD\HUV RI GLIIXVH FRXQWHU LRQV QHDU WKH VXUIDFHV 7KH VRFDOOHG VHFRQGDU\ PLQLPXP ] ]f LV ORFDWHG ZKHUH WKHVH ORQJ UDQJH HOHFWURVWDWLF UHSXOVLRQ IRUFHV EDODQFH YDQ GHU :DDOV DWWUDFWLRQ IRUFHV ZKLFK FRUUHVSRQGV WR D WUDQVLHQW DGKHVLRQ VWDWH ZKHUH WKH SDUWLFOH FDQ W\SLFDOO\ GLIIXVH RXW RI WKLV UHJLRQ GXH WR %URZQLDQ PRWLRQ ,I WKH SDUWLFOH KDV VXIILFLHQW HQHUJ\ LW PD\ EH DEOH WR VXUPRXQW WKH IUHH HQHUJ\ EDUULHU ] =PD[f DQG GHVFHQG LQWR WKH SULPDU\ PLQLPXP ] ]Mrf ZKHUH LW ZLOO EH FRQVLGHUHG IXOO\ DGKHUHQW DW D VHSDUDWLRQ GLVWDQFH ZKHUH FORVH UDQJH VWHULF UHSXOVLRQ IRUFHV EDODQFH YDQ GHU :DDOV DWWUDFWLRQ IRUFHV 2QH K\SRWKHVLV LV WKDW WKH UHDUUDQJHPHQW RI PDFURPROHFXODU VXUIDFH VWUXFWXUHV VXFK DV ELQGLQJ PROHFXOHV ILPEULDH RU H[WUXGHG SRO\PHUV fVOLPHff RFFXUV GXULQJ DGKHVLRQ LQ WKH VHFRQGDU\ PLQLPXP ZKLFK SURPRWHV WKH WUDQVLWLRQ RYHU WKH HQHUJ\ EDUULHU UHVXOWLQJ LQ ILUP DWWDFKPHQW LQ WKH SULPDU\ PLQLPXP 3HWKLFD f $ OLPLWDWLRQ RI YLHZLQJ EDFWHULDO DWWDFKPHQW DV GHSRVLWLRQ RI DQ LQHUW FROORLGDO SDUWLFOH LV WKDW WKH PDFURPROHFXODU LQWHUDFWLRQV DUH QRW H[SOLFLWO\ WDNHQ LQWR DFFRXQW +RZHYHU ZRUN LQ RXU JURXS UHVXOWHG LQ D UHFHQW SXEOLFDWLRQ WKDW SUHVHQWV D PRGHO IRU EDFWHULDO DWWDFKPHQW WKDW DFFRXQWV IRU ERWK FROORLGDO IRUFHV DQG WKH IRUPDWLRQ RI PDFURPROHFXODU ERQGV 'LFNLQVRQ f 7KLV PRGHO ZLOO EH XVHG WR KHOS LQWHUSUHW H[SHULPHQWDO GDWD LQ &KDSWHU

PAGE 18

)LJXUH '/92 LQWHUDFWLRQ SRWHQWLDO $ K\SRWKHWLFDO SORW RI SRWHQWLDO HQHUJ\ IW]f YV VHSDUDWLRQ GLVWDQFH ] DV SUHGLFWHG IURP WKH '/92 WKHRU\ 7KH SULPDU\ PLQLPXP FRUUHVSRQGV WR WKH SRLQW RI DWWDFKPHQW 7R DWWDFK D EDFWHULXP PXVW PRYH E\ %URZQLDQ PRWLRQ IURP WKH VHFRQGDU\ PLQLPXP RYHU WKH HQHUJ\ EDUULHU DQG LQWR WKH SULPDU\ PLQLPXP 6SHFLILF 0HFKDQLVPV RI %DFWHULDO $GKHVLRQ $ ELRPDWHULDO H[SRVHG WR D ELRORJLFDO HQYLURQPHQW UDSLGO\ DFTXLUHV D SURWHLQ FRQGLWLRQLQJ ILOP WKDW FDQ DIIHFW VXUIDFH SURSHUWLHV DQG FHOO DGKHVLRQ *ULVWLQD HW DO %DLHU HW DO f $WWDFKPHQW FDQ UHVXOW IURP VSHFLILF ELQGLQJ RI PROHFXODU UHFHSWRUV

PAGE 19

DND DGKHVLQVf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t 0F'HYLWW f 6WDSK\ORFRFFXV DXUHXV LV D FRPPRQ RSSRUWXQLVWLF SDWKRJHQ LQYROYHG LQ GHYLFH FHQWHUHG LQIHFWLRQV 'DQNHUW HW DO f ,W LV EHOLHYHG WR DGKHUH WR ELRPDWHULDOV YLD VSHFLILF ELQGLQJ RI FHOOVXUIDFH UHFHSWRUV WR DGVRUEHG KRVW SURWHLQV VXFK DV ILEULQRJHQ )Jf )RVWHU t 0F'HYLWW 2KWRPR t
PAGE 20

&GO 0HPEUDQH &HOO :DOO &DSVXOH &RQGLWLRQLQJ )LOP 6 0DWHULDOr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t 6WUDQJHUW f IOXRUHVFHQFH 'RQNHUVORRW f RU UDGLRDFWLYLW\ 6MROOHPD f +RZHYHU WKHVH DUH RIWHQ QRW VHQVLWLYH HQRXJK RU GHSHQG RQ XQVWDEOH FKDUDFWHULVWLFV RI WKH FHOOV 7KHVH WHFKQLTXHV DOVR JHQHUDOO\ VXIIHU IURP LOOGHILQHG IORZ FRQGLWLRQV ZKLFK FDQ DIIHFW WKH WUDQVSRUW UDWH RI EDFWHULD WR WKH VXUIDFH DV ZHOO DV LQIOXHQFH WKH SUREDELOLW\ RI D FHOO DWWDFKLQJ WR WKH VXUIDFH 'LFNLQVRQ t &RRSHU f )XUWKHUPRUH fEOLQGf DVVD\V LQ ZKLFK FHOO DWWDFKPHQW LV QRW GLUHFWO\ REVHUYHG FDQQRW DFFRXQW IRU VHFRQGDU\ HIIHFWV VXFK DV FHOOWRFHOO LQWHUDFWLRQV RU FHOO DJJORPHUDWLRQ LQ VXVSHQVLRQ RU RQ WKH

PAGE 21

VXUIDFH 7KHVH OLPLWDWLRQV FORXG WKH VLJQLILFDQFH RI WKH H[SHULPHQWDO UHVXOWV ZKHQ WKH SULPDU\ LQWHUHVW LV WR TXDQWLI\ WKH LQWHUDFWLRQ RI WKH FHOO ZLWK WKH VXEVWUDWXP 7R RYHUFRPH WKHVH OLPLWDWLRQV VRPH UHFHQW VWXGLHV KDYH XVHG SDUDOOHO SODWH IORZ FKDPEHUV 33)&fVf WR VWXG\ FHOO DWWDFKPHQW RU GHWDFKPHQW :LONLQVRQ HW DO 6MROOHPD /DZUHQFH HW DO -HQ t /LQ YDQ .RRWHQ HW DO f 2WKHUV ZKR VWXGLHG VKHDU GHSHQGHQW FHOO DGKHVLRQ )RZOHU t 0F.D\ &URXFK HW DO &R]HQV5REHUWV HW DO E 'LFNLQVRQ t &RRSHU f KDYH XVHG UDGLDOIORZ FKDPEHUV )ORZ FKDPEHUV RIIHU WKH DGYDQWDJHV RI ZHOOGHILQHG IORZ FRQGLWLRQV DQG WKH SRVVLELOLW\ RI UHDOWLPH GLUHFW REVHUYDWLRQ DQG HQXPHUDWLRQ RI WKH QXPEHU RI DWWDFKHG FHOOV 7KLV DOORZV WKH PHDVXUHPHQW RI WKH NLQHWLFV RI FHOO DWWDFKPHQW RU GHWDFKPHQW +RZHYHU LQ SUHYLRXV 33)& VWXGLHV WKH DFWXDO FHOO FRQFHQWUDWLRQ QHDU WKH VXUIDFH ZDV QRW NQRZQ DQG LW FRXOG YDU\ DV D IXQFWLRQ RI WKH GRZQVWUHDP GLVWDQFH LQ WKH FKDPEHU 'LFNLQVRQ t &RRSHU f 7KLV LPSOLHV WKDW WKH PHDVXUHG DWWDFKPHQW UDWH GHSHQGHG RQ WKH SRVLWLRQ LQ WKH FKDPEHU FKRVHQ IRU REVHUYDWLRQ DV ZHOO DV RQ WKH IORZ UDWH )RU H[DPSOH LW ZDV REVHUYHG 'LFNLQVRQ t &RRSHU f WKDW UDSLG DWWDFKPHQW RI SDUWLFOHV FDQ OHDG WR GRZQVWUHDP GHSOHWLRQ RI FHOOV QHDU WKH VXUIDFH DQG FRQYHUVHO\ IRU VORZ DWWDFKPHQW DQG VHGLPHQWDWLRQ FDQ OHDG WR D GRZQVWUHDP DFFXPXODWLRQ RI FHOOV QHDU WKH VXUIDFH 7KHUHIRUH WKH PHDVXUHG HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW LV GHSHQGHQW RQ WKH GRZQVWUHDP GLVWDQFH LQ WKH 33)& WKHUHIRUH LV QRW DQ LQWULQVLF PHDVXUH RI WKH SUREDELOLW\ RI DWWDFKPHQW ZKLFK GHSHQGV RQO\ RQ ORFDO FRQGLWLRQV VXFK DV LQWHUDFWLRQ EHWZHHQ WKH FHOO DQG WKH VXUIDFH 7R HVWLPDWH DQ LQWULQVLF UDWH FRQVWDQW RI DWWDFKPHQW WKH UDWH RI DWWDFKPHQW PXVW EH PHDVXUHG DV D IXQFWLRQ RI SRVLWLRQ LQ WKH 33)& DQG WKH YDULDEOH

PAGE 22

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

PAGE 23

&+$37(5 67$3+
PAGE 24

ZHOOGHILQHG IORZ FRQGLWLRQV WKDW WKH OHQJWK RI WKH 5UHJLRQ SURPRWHV DGKHVLRQ E\ SURPRWLQJ DQ HQKDQFHG UDWH RI DWWDFKPHQW ZKLFK IXUWKHU VXJJHVWV WKDW WKH IXQFWLRQ RI WKLV UHJLRQ DOORZV WKH FHOO WR fUHDFKf DFURVV WKH HQHUJ\ EDUULHU )LJXUH 6FDQQLQJ (OHFWURQ 0LFURJUDSK RI 6WDSK\ORFFRFXV DXUHXV )LEULQRJHQELQGLQJ 3URWHLQ FOI$ f&OXPSLQJ )DFWRUf 0HPEUDQH DQG &HOO :DOO6SDQQLQJ 6HU$VS 5HSHDW 'RPDLQ 'RPDLQV ZPDPPPPPPPPPPPPPPPPPPPPPPPP f6WDON 5HJLRQf )LEULQRJHQ%LQGLQJ 'RPDLQ )LJXUH 6WUXFWXUH RI FOXPSLQJ IDFWRU

PAGE 25

6 DXUHXV LQ 'HYLFH&HQWHUHG ,QIHFWLRQV 6WDSK\ORFRFFXV DXUHXV 6 DXUHXVf ZDV XVHG LQ WKLV VWXG\ EHFDXVH LW LV D KLJKO\ SDWKRJHQLF EDFWHULDO VWUDLQ FDXVLQJ b RI WKH KRVSLWDO DFTXLUHG LQIHFWLRQV DQG LW LV H[WUHPHO\ UHVLVWDQW WR DQWLELRWLF WUHDWPHQW 6 DXUHXV LV D SUHGRPLQDQW SDWKRJHQ LQ JHQHUDWLQJ DQWLELRWLFUHVLVWDQW DV ZHOO DV GHYLFHFHQWHUHG LQIHFWLRQV :DOGYRJHO f 6 DXUHXV H[SUHVVHV VHYHUDO VSHFLILF FHOO VXUIDFH ELQGLQJ PROHFXOHV WKDW LQWHUDFW ZLWK SODVPD SURWHLQV FRPPRQO\ IRXQG DGVRUEHG WR ELRPDWHULDOV VXFK DV ILEULQRJHQ &KHXQJ t )LVFKHWWL +HUUPDQQ HW DO +HUUPDQQ HW DO 0F'HYLWW HW DO f ILEURQHFWLQ +HUUPDQQ HW DO 9DXGDX[ HW DO Df Ef 9DXGDX[ f FROODJHQ YLWURQHFWLQ ODPLQLQ WKURPERVSRQGLQ ERQH VLDORSURWHLQ HODVWLQ HWF +RZHYHU VWXGLHV SHUIRUPHG VWURQJO\ LQGLFDWH WKDW ILEULQRJHQ LV WKH PRVW DFWLYH SODVPD FRPSRQHQW LQ SURPRWLQJ DQ LQLWLDO LQ YLYR DGKHVLRQ RI 6 DXUHXV )UDQFRLV HW DO f $V PHQWLRQHG 6 DXUHXV KDV EHHQ VKRZQ WR DWWDFK WR ILEULQRJHQFRDWHG VXUIDFHV YLD fFOXPSLQJ IDFWRUf 0F'HYLWW HW DO 'LFNLQVRQ HW DO f 7KH SUHYHQWLRQ RI EDFWHULDO DWWDFKPHQW DQG RQVHW RI LQIHFWLRQ DUH KLJKO\ GHSHQGHQW RQ WKH LQLWLDO VKRUWWHUP FRQWDFW RI WKH ELRPDWHULDO LQ WKH ERG\ 7KHUHIRUH WKH IRFXV RI WKLV VWXG\ LV RQ WKH PROHFXODU LQWHUDFWLRQV DQG SDUDPHWHUV UHOHYDQW DW D VKRUW WLPH VFDOH DIWHU WKH LPSODQWDWLRQ 6SHFLILFDOO\ RXU IRFXV LV RQ WKH UROH RI VSHFLILF DQG QRQVSHFLILF LQWHUDFWLRQV LQYROYHG LQ WKH LQLWLDO UDWH RU HTXLYDOHQWO\ SUREDELOLW\f RI DWWDFKPHQW WR VXUIDFHV

PAGE 26

&+$37(5 0($685(0(17 2) 5($/7,0( %$&7(5,$/ $77$&+0(17 .,1(7,&6 72 */$66 $1' 3527(,1 &2$7(' 685)$&(6 $SSDUDWXV 7KH UROH RI WKH FHOO VXUIDFH SURSHUWLHV IOXLG DQG VXEVWUDWD LQ SURPRWLQJ DGKHVLRQ FDQ EH TXDQWLILHG E\ PHDVXULQJ EDFWHULDO DWWDFKPHQW NLQHWLFV WR GLIIHUHQW FRDWLQJV RQ D JODVV VXEVWUDWH XQGHU ZHOOGHILQHG IORZ FRQGLWLRQV 7KH SULQFLSDO WRRO LQYROYHG LQ WKLV VWXG\ LV D SDUDOOHO SODWH IORZ FHOO 33)&f ZKLFK SURYLGHV ODPLQDU ZHOO GHILQHG IORZ FRQGLWLRQV IRU PHDVXULQJ WKH QXPEHU RI EDFWHULD DWWDFKLQJ WR VXUIDFHV $V LW ZDV GLVFXVVHG LQ 6HFWLRQ WKH SRVVLELOLW\ RI GLUHFW REVHUYDWLRQ RI DWWDFKPHQW NLQHWLFV LQ WKH 33)& RIIHUV DGYDQWDJHV RYHU RWKHU EOLQG DVVD\V IRU WKH PHDVXUHPHQW RI EDFWHULDO DGKHVLRQ 7KH DLP RI WKLV FKDSWHU LV WR SUHVHQW D GHWDLOHG GHVFULSWLRQ RI D 33)& LQFOXGLQJ WKH SULQFLSOHV LQYROYHG LQ LWV GHVLJQ DQG WHVWLQJ WKH H[SHULPHQWDO SURWRFRO IRU DXWRPDWHG HQXPHUDWLRQ RI WKH DGKHUHQW EDFWHULD DQG WKH SURFHGXUH IRU GDWD DQDO\VLV $ PHWKRGRORJ\ LV DOVR GHVFULEHG IRU WKH GDWD DQDO\VLV WKDW RIIHUV DQ LPSURYHPHQW RYHU WUDGLWLRQDO DSSURDFKHV XVHG LQ LQWHUSUHWLQJ WKH DWWDFKPHQW NLQHWLFV UHVXOWV IURP 33)& H[SHULPHQWV )LQDOO\ WKH H[SHULPHQWDO UHVXOWV DUH SURYLGHG DV D IXQFWLRQ RI WKH OHQJWK RI WKH FOXPSLQJ IDFWRU PROHFXOH 7KH\ DUH REWDLQHG IURP DQ DSSOLFDWLRQ RI WKLV PHWKRGRORJ\

PAGE 27

WR WKH VWXG\ RI 6 DXUHXV DWWDFKPHQW WR ILEULQRJHQFRDWHG VXUIDFHV ,Q D VHFRQG VWXG\ WKH UROH RI HOHFWURO\WH RQ DWWDFKPHQW NLQHWLFV RI 6 DXUHXV WR EDUH JODVV VXUIDFHV LV DOVR GHVFULEHG 7KHRU\ )ORZ V\VWHPV DUH HPSOR\HG LQ DGKHVLRQ VWXGLHV EHFDXVH WKH\ DOORZ WKH FRQWURO RI K\GURG\QDPLF FRQGLWLRQV LQ WHUPV RI VKHDU UDWH IORZ YHORFLW\ DQG 5H\QROGV QXPEHU 6MROOHPD f &RQYHFWLRQ DOVR DIIHFWV WUDQVSRUW RI EDFWHULD WR WKH FROOHFWRU VXUIDFH 7KH DGYDQWDJH RI XVLQJ D 33)& LV WKDW LW DOORZV UHDOWLPH GLUHFW REVHUYDWLRQ RI WKH VXUIDFH ZKHUH SDUWLFOHV DWWDFK XQGHU ZHOOGHILQHG ODPLQDU IORZ FRQGLWLRQV )ORZ LQ WKH 33)& PXVW EH VWHDG\ ODPLQDU DQG IXOO\ GHYHORSHG LQ RUGHU WR PLQLPL]H RU HOLPLQDWH DQ\ HQWUDQFH HIIHFWV YDQ :DJHQHQ t $QGUDGH %RZHQ YDQ .RRWHQ HW DO f 7KH GHVLJQ RI WKH 33)& LV VLPLODU WR SUHYLRXV GHVLJQV YDQ .RRWHQ HW DO 6MROOHPD f EXW PRGLILHG WR ILW ZLWKLQ WKH LQKRXVH DXWRPDWHG YLGHR PLFURVFRS\ V\VWHP 7KH SDUDPHWHU WKDW QHHGV WR EH FKRVHQ DSSURSULDWHO\ WR DVVXUH PLQLPDO HQWUDQFH HIIHFWV LV WKH FKDUDFWHULVWLF HQWU\ OHQJWK /H GHILQHG DV IROORZV LQ HTXDWLRQ /H DK 5H f :KHUH D SURSRUWLRQDOLW\ FRQVWDQW K FKDQQHO KHLJKW FPf 5H 5H\QROGV QXPEHU

PAGE 28

7KH FRQVWDQW D LV UHSRUWHG WR EH EHWZHHQ YDQ :DJHQHQ t $QGUDGH f DQG %RZHQ f ,Q GHVLJQLQJ WKH 33)& GHVFULEHG LQ WKH QH[W VHFWLRQ DQ DYHUDJH RI WKH WZR YDOXHV ZDV XVHG 7KH 5H\QROGV QXPEHU LQ WKH 33)& LV GHILQHG DV 5H 34 ZKHUH Z Kf f IM S IOXLG GHQVLW\ JFP f f S YLVFRVLW\ JFPVf Z FKDQQHO ZLGWK FPf 4 YROXPHWULF IORZ UDWH FP Vf )RU D XQLIRUP IORZ SURILOH WKURXJKRXW WKH HQWLUH SODWH OHQJWK /H PXVW EH VPDOO FRPSDUHG WR WKH OHQJWK RI WKH FKDPEHU 7KH FULWHULRQ LV /H EO ZKHUH E LV D SURSRUWLRQDOLW\ FRQVWDQW DQG LV WKH FKDPEHU OHQJWK %DVHG RQ WKH DERYH HTXDWLRQ 9DQ :DJHQHQ DQG $QGUDGH SRVWXODWHG WKDW /H VKRXOG EH OHVV WKDQ b RI WKH FKDPEHU OHQJWK LQ RUGHU WR KDYH PLQLPDO HQWUDQFH HIIHFWV LH E YDQ :DJHQHQ t $QGUDGH f )URP WKH IOXLG YHORFLW\ SURILOH GHIHUUHG WR &KDSWHU f WKH VKHDU VWUHVV RQ WKH VXUIDFH RI WKH 33)& LV JLYHQ E\ 6KHDU VWUHVV 4SZK G\QFPf f )RU D W\SLFDO H[SHULPHQW WKH GHVLUHG VKHDU VWUHVV LV GHWHUPLQHG DQG WKH DSSURSULDWH IORZ UDWH FDQ EH FDOFXODWHG IURP WKH NQRZQ IOXLG DQG FKDPEHU SDUDPHWHUV

PAGE 29

'HVLJQ ,QLWLDOO\ D 33)& ZDV GHVLJQHG DQG FRQVWUXFWHG LQKRXVH EDVHG RQ WKH WKHRU\ GHVFULEHG DERYH DQG RQ DQ H[LVWLQJ GHVLJQ YDQ .RRWHQ HW DO f +RZHYHU VRPH DGGLWLRQV DQG LPSURYHPHQWV ZHUH PDGH RYHU WKH GHVLJQV GHVFULEHG LQ WKH OLWHUDWXUH 7KLV GHVLJQ FDQ EH VHHQ LQ )LJXUH 7KH 33)& FRQVLVWV RI WZR SDUWV 7KH ERWWRP SDUW ZDV PDFKLQHG RXW RI EUDVV DQG ZDV SODWHG ZLWK VROLG QLFNHO FRQWDLQLQJ b SKRVSKRUXV 7KH FRDW WKLFNQHVV LV EHWZHHQ LQFKHV WKLV SUHYHQWV FRUURVLRQ RI WKH ERWWRP SDUW LQ WKH VDOLQH RSHUDWLQJ HQYLURQPHQW 7KLV SDUW KDV PDFKLQHG LQWR LW DQ 2ULQJ FKDQQHO WKDW DOORZV FRPSOHWH VHDOLQJ RI WKH FKDPEHU ZKHQ D JODVV VOLGH LV VXSHULPSRVHG RQ LW DQG FORVHG ZLWK WKH WRS 3OH[LJODV SDUW ZKLFK LQ WXUQ KDV DQRWKHU JODVV VOLGH ILWWHG LQ RQ DQ 2 ULQJ FKDQQHO 7KH IORZ FKDQQHO LV IRUPHG ZKHQ VWDLQOHVV VWHHO PP XQLIRUP WKLFNQHVVf VSDFHUV DUH SRVLWLRQHG SDUDOOHO WR HDFK RWKHU DORQJ WKH OHQJWK RI WKH FKDPEHU DQG VDQGZLFKHG EHWZHHQ WKH WRS DQG ERWWRP JODVV VOLGHV 7KH HQWLUH DVVHPEO\ LV VHDOHG WR EHFRPH OHDNSURRI E\ WLJKWHQLQJ RI WKH VFUHZV ,Q RUGHU WR HQVXUH D ODPLQDU IORZ ILHOG DW WKH FKDQQHO HQWUDQFH EHWZHHQ WKH WZR SDUDOOHO SODWHV WKH GHSWK DQG ZLGWK RI WKH RI WKH LQOHW DQG RXWOHW FKDQQHOV JUDGXDOO\ GHFUHDVH DQG LQFUHDVH UHVSHFWLYHO\ XS WR WKH ILQDO FKDQQHO GHSWK FPf DQG ZLGWK FPf )RU SURSHU RSHUDWLRQ RI WKLV GHYLFH DLU ZDV HOLPLQDWHG IURP WKH OLQHV EHIRUH SHUIRUPLQJ DQ\ H[SHULPHQWV EHFDXVH HQWUDSSHG DLU EXEEOHV FRXOG FDXVH VWDJQDWLRQ RI IORZ RU EDFN IORZ LQ WKH FKDPEHU FKDQJLQJ WKH NQRZQ IORZ UDWH /LTXLG ZDV DOORZHG WR IORZ SULRU WR VWDUWLQJ WKH DFWXDO H[SHULPHQW LQ RUGHU WR ILOO XS ZLWK OLTXLG DQ\ UHPDLQLQJ JDSV RXWVLGH WKH IORZ DUHD

PAGE 30

7KH WHPSHUDWXUH RI WKH IORZ FKDPEHU ZDV FRQWUROOHG XVLQJ IRXU VLOLFRQH UXEEHU KHDWHUV DWWDFKHG WR WKH ERWWRP EUDVV SDUW RI WKH FKDPEHU 7KHVH KHDWHUV FRXOG EH FRQQHFWHG WR D WHPSHUDWXUH FRQWUROOHU ZLWK D UHOD\ WR VZLWFK WKH ORDG RQ DQG RII RQ WKH KHDWHUV :DWORZ 6HEULQJ )/f )LJXUH 6FKHPDWLF RI SDUDOOHO SODWH IORZ FHOO GHVLJQHG fLQKRXVHf 7KH IORZ LQVLGH WKH 33)& ZDV WHVWHG WR EH ODPLQDU E\ XVLQJ D UKHRVFRSLF IOXLG .DOOLURVFRSH &RUSRUDWLRQ *URWRQ 0DVVDFKXVHWWVf 7KLV IOXLG DOORZV YLVXDOL]DWLRQ RI WKH IORZ SDWWHUQ LQ WKH 33)& 7KH UKHRVFRSLF IOXLG KDV EHHQ GHYHORSHG LQLWLDOO\ E\ DQ DUWLVW 3DXO 0DWLVVH ZKR XVHG WKHVH IOXLGV WR GHYHORS DUWZRUNV EDVHG RQ WKH GLVSOD\ RI FRQYHFWLRQ FXUUHQWV 7KH UKHRVFRSLF IOXLG LV D VXVSHQVLRQ RI PLFURVFRSLF FU\VWDOOLQH SHOOHWV :KHQ LQ PRWLRQ WKH SHOOHWV DOLJQ WKHLU ORQJHVW GLPHQVLRQ SDUDOOHO WR WKH SODQHV RI

PAGE 31

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f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

PAGE 32

0HWKRGRORJ\ IRU 0HDVXULQJ 3DUWLFOH $WWDFKPHQW .LQHWLFV 'DWD $FTXLVLWLRQ $Q DXWRPDWHG YLGHR PLFURVFRS\ V\VWHP )LJXUH f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f 7KHQ WKH SDUWLFOH GHQVLW\ ZDV SORWWHG YHUVXV WLPH IRU HDFK GRZQVWUHDP SRVLWLRQ )LJXUH f

PAGE 33

)LJXUH 6FKHPDWLF RI SDUDOOHO SODWH IORZ FHOO 7KLV 33)& ZDV XVHG LQ WKH DGKHVLRQ H[SHULPHQWV ZKRVH UHVXOWV DUH SUHVHQWHG LQ WKLV GLVVHUWDWLRQ 7KH VORSH RI WKH UHVXOWLQJ VWUDLJKW OLQH ZDV REWDLQHG E\ OLQHDU UHJUHVVLRQ 'LYLGLQJ WKLV VORSH E\ WKH LQLWLDO FRQFHQWUDWLRQ RI SDUWLFOHV &R \LHOGHG DQ HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW NHII IRU HDFK ILHOG 7KH YDOXHV IURP WKH IRXU ILHOGV DW HDFK GRZQVWUHDP SRVLWLRQ ZHUH DYHUDJHG WR SURYLGH NHMM[f $V GLVFXVVHG LQ PRUH GHWDLO LQ &KDSWHU NHII LV JHQHUDOO\ D IXQFWLRQ RI WKH GRZQVWUHDP GLVWDQFH LQ WKH 33)& EHFDXVH UDSLG DWWDFKPHQW FDQ OHDG WR D GRZQVWUHDP GHSOHWLRQ RI FHOOV QHDU WKH VXUIDFH DQG VORZ DWWDFKPHQW FDQ OHDG WR GRZQVWUHDP DFFXPXODWLRQ GXH WR VHGLPHQWDWLRQ ,Q RUGHU WR FRUUHFW IRU WKLV HIIHFW RI SDUWLFOH WUDQVSRUW RQ WKH DWWDFKPHQW LQ WKH 33)& DQG WR GHWHUPLQH DQ LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW N

PAGE 34

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f E\ QRQOLQHDU OHDVWVTXDUHV UHJUHVVLRQ DQ HVWLPDWH RI N ZDV REWDLQHG 7KH YDOXH RI WKLV SURFHGXUH LV WKDW WKH LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW LV LQGHSHQGHQW RI DQ\ JOREDO SDUDPHWHUV LQ WKH V\VWHP VXFK DV WKH GRZQVWUHDP GLVWDQFH LQ WKH 33)&

PAGE 35

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

PAGE 36

&HOO 'HQVLW\ YV 7LPH )LJXUH 3URFHGXUH IRU FDOFXODWLQJ NH-J?f 7KLV LV D SORW RI FHOO GHQVLW\ YV WLPH IRU WZR RI WKH 6 DXUHXV PXWDQWV 5 DPLQR DFLG UHVLGXHV DQG =" DPLQR DFLG UHVLGXHVf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f 7KH 6n DXUHXV PXWDQWV ZLWK YDU\LQJ OHQJWK RI WKH VHULQHDVSDUWLF DFLG 6HU$VSf UHSHDW GRPDLQ RI WKH ILEULQRJHQ UHFHSWRU

PAGE 37

GXPSLQJ IDFWRU FOI$f ZHUH JHQHWLFDOO\ HQJLQHHUHG E\ SODVPLG LQVHUWLRQ RI WKH PRGLILHG FOI$ JHQH LQWR D ILEULQRJHQ GHILFLHQW PXWDQW RI WKH EDFWHULXP DQG FORQHG +DUWIRUG HW DO f 7KH FORQHG ILEULQRJHQ UHFHSWRU FOXPSLQJ IDFWRU FOI$f ZDV WKDW RI 6 DXUHXV 1HZPDQ VWUDLQ 0HPEUDQH DQG &HOO :DOO6SDQQLQJ 'RPDLQV 6HU$VS 5HSHDW 'RPDLQ f R )LEULQRJHQ%LQGLQJ 'RPDLQ S&) ‘r ‘ & S&) S&) S&) S&) S&) R & $ S&) )LJXUH 6 DXUHXV PXWDQWV ZLWK YDULDEOH OHQJWK RI WKH UHSHDW UHJLRQ XVHG LQ 33)& H[SHULPHQWV 7KH PXWDQWV ZHUH VWRUHG LQ D & IUHH]HU DQG HDFK LQGLYLGXDO YLDO ZDV GHIURVWHG RQO\ IRU WKH VKRUW SHULRG QHFHVVDU\ IRU WKH WUDQVIHU RI DQ LQQRFXOXP WR WU\SWLF VR\ EURWK

PAGE 38

IRU RYHUQLJKW JURZWK 7KH 6 DXUHXV PXWDQWV XVHG LQ H[SHULPHQWV ZHUH JURZQ RYHUQLJKW LQ PO RI WU\SWLF VR\ EURWK 76%f WR ZKLFK PLFUROLWHUV RI FKORUDPSKHQLFRO ZHUH DGGHG 1RUPDOO\ WKH UDWLR RI FKORUDPSKHQLFRO WR 76% ZDV OSO SHU PO RI 76% $OO PXWDQW VWUDLQV H[FHSW WKH FOXPSLQJ IDFWRU QHJDWLYH PXWDQW $ ZHUH JURZQ LQ 76% ZLWK FKORUDPSKHQLFRO ([SHULPHQWDO 3URWRFRO 7KH ILUVW VWHS RI DQ H[SHULPHQW ZDV WKH SUHSDUDWLRQ RI WKH SKRVSKDWH EXIIHU VDOLQH 3%6f ZLWK GRXEOH GLVWLOOHG ZDWHU DQG D FRPPHUFLDO SKRVSKDWH EXIIHU GU\ PL[WXUH 'XEHOFRfV 3%6f 7KH EXIIHUHG PHGLD ZDV VXSSOHPHQWHG ZLWK PJ HDFK PDJQHVLXP DQG FDOFLXP FKORULGH SHU PO RI VROXWLRQ DW S+ f 7KHUHIRUH WR RQH OLWHU RI GRXEOH GLVWLOOHG ZDWHU WKHUH ZHUH DGGHG XQGHU FRQWLQXRXV VWLUULQJ RQ D PDJQHWLF SODWH WZR HQYHORSHV RI SRZGHUHG 'XEHOFRfV 3%6 DQG J HDFK RI PDJQHVLXP DQG FDOFLXP FKORULGH 7KLV EXIIHU ZDV WKH EDVH IRU WKH DOEXPLQ DQG ILEULQRJHQ VROXWLRQV SUHSDUHG IRU FRDWLQJ RI WKH JODVV VOLGHV PP [ PP )LVKHU 3LWWVEXUJK 3$f 7KH EXIIHU ZDV ILOWHUHG XQGHU YDFXXP WKURXJK D )LVKHU 6FLHQWLILF PLFURQ ILOWHU WR HQVXUH DQ LQLWLDOO\ SDUWLFOHIUHH VROXWLRQ 3ULRU WR WKHLU UHVSHFWLYH FRDWLQJV WKH JODVV VOLGHV ZHUH FOHDQHG RYHUQLJKW LQ D FKURPLF DFLG VROXWLRQ &KURPHUJH )LVKHU 3LWWVEXUJK 3$f DQG VXEVHTXHQWO\ ULQVHG ZLWK ZDWHU DQG VXEPHUJHG IRU WZR KRXUV LQ GRXEOH GLVWLOOHG ZDWHU 7KH QH[W VWHS FRQVLVWHG RI FRDWLQJ WKH SUHFOHDQHG JODVV VOLGHV ZLWK WKH UHVSHFWLYH FRDWLQJ VROXWLRQV E\ VZLUOLQJ WKH VXEPHUJHG VOLGHV LQ D 3HWUL GLVK RQ WRS RI D 5RWR 0L[HU LQ WKH IROORZLQJ PDQQHU 6OLGHV ZHUH LQFXEDWHG HLWKHU LQ ERYLQH VHUXP DOEXPLQ %6$f

PAGE 39

VROXWLRQ b E\ ZHLJKWf RU LQ D VROXWLRQ RI KXPDQ ILEULQRJHQ PJ ILEULQRJHQ POf IRU RQH KRXU WKHQ IRU KRXU LQ WKH %6$ VROXWLRQ WR EORFN DQ\ UHPDLQLQJ VXUIDFH VLWHV SRWHQWLDOO\ QRW FRYHUHG E\ WKH ILEULQRJHQ 7KH VOLGHV FRDWHG ZLWK %6$ VHUYHG DV D QRQn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f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

PAGE 40

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n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f )URP WKLV SRLQW XQWLO WKH FRPSOHWLRQ RI WKH H[SHULPHQW WKH GDWD ZHUH DFTXLUHG DXWRPDWLFDOO\ DQG WKH LPDJHV RI WKH ILHOGV ZLWK DWWDFKHG SDUWLFOHV ZHUH VWRUHG LQ WKH FRPSXWHU IRU IXUWKHU DQDO\VLV 0RGLILHG ([SHULPHQWDO 3URWRFRO 7KH H[SHULPHQWDO SURWRFRO GHVFULEHG LQ 6HFWLRQ ZDV PRGLILHG LQ WKH ODWHVW H[SHULPHQWV EHFDXVH RI SUREOHPV ZLWK WKH UHFHQW EDWFKHV RI ERYLQH VHUXP DOEXPLQ %6$f SXUFKDVHG IURP 6LJPD 6FLHQWLILF 7KH QHZ %6$ EDWFKHV ZHUH DSSDUHQWO\ QRW DV SXUH DV SUHYLRXV EDWFKHV ,W LV EHOLHYHG LW PLJKW KDYH EHHQ FRQWDPLQDWHG ZLWK ILEULQRJHQ )Jf RU

PAGE 41

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f $V DQ H[WHQVLRQ GU\ IDWIUHH PLON KDV DOVR EHHQ XVHG DV D EORFNLQJ DJHQW IRU QRQVSHFLILF DGKHVLRQ =HQJ HW DO f 7KHUHIRUH LW ZDV D ORJLFDO H[WHQVLRQ WR WU\ OLTXLG VNLP PLON LQVWHDG RI %6$ WR FRDW WKH JODVV VOLGHV LQ RUGHU WR DWWDLQ D QRQDGKHVLYH FRQWURO 0LON KDV RQ DYHUDJH D FRPSRVLWLRQ RI b ZDWHU b ODFWRVH b IDW b RWKHU LQJUHGLHQWV DQG b PLON SURWHLQ RXW RI ZKLFK b LV FDVHLQ )HUJXVRQ f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

PAGE 42

3RO\VFLHQFHV ,QFf DV PRGHO EDFWHULD LQ YDU\LQJ HOHFWURO\WH FRQFHQWUDWLRQ LQ WKH VXVSHQVLRQ PHGLD LQ RUGHU WR HVWDEOLVK WKH H[SHULPHQWDO SURWRFRO IRU XVH ZLWK WKH 6 DXUHXV EDFWHULD 7KLV LQLWLDO DSSURDFK ZDV XVHG WR HOLPLQDWH DQ\ SUREOHPV ZLWK WKH 33)& VXFK D QRQXQLIRUP ODPLQDU ILHOG RU OHDNV LQ WKH V\VWHP ,W DOVR KHOSHG WR HOLPLQDWH WKH EXJV LQ WKH 2SWLPDV SURJUDP PDFURf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f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f LQGLFDWHV EDFNJURXQG ELQGLQJ WR WKH ILEULQRJHQ FRDWHG VXUIDFHV ZKLFK PD\ EH GXH WR WKH SUHVHQFH RI RWKHU ILEULQRJHQELQGLQJ SURWHLQV RQ WKH EDFWHULDO VXUIDFH %DVHG RQ FRPSDULVRQ WR WKH PXWDQW VWUDLQ WKDW ZDV FRPSOHWHO\

PAGE 43

GHYRLG RI FOXPSLQJ IDFWRU WKH VKRUWHQHG UHFHSWRU ZDV VWLOO HIIHFWLYH LQ SURPRWLQJ DWWDFKPHQW WR WKH ILEULQRJHQFRDWHG VXUIDFH +RZHYHU WKH OHQJWK RI WKH UHSHDW UHJLRQ LQ &OI$ VHHPHG WR SOD\ D FULWLFDO UROH LQ SURPRWLQJ DWWDFKPHQW ZKLFK LV FRQVLVWHQW ZLWK WKH K\SRWKHVLV WKDW WKLV UHJLRQ VHUYHV DV D VWDON WR SURMHFW WKH OLJDQGELQGLQJ GRPDLQ WR LQFUHDVH WKH SUREDELOLW\ RI ELQGLQJ ZLWK OLJDQGV LQ WKH FHOOfV HQYLURQPHQW )LJXUH (IIHFW RI UHFHSWRU OHQJWK RQ EDFWHULDO DWWDFKPHQW NLQHWLFV LQ WKH 33)& 7KLV ZDV WKH ILUVW VWXG\ WR GHPRQVWUDWH WKDW UHFHSWRU LV D FULWLFDO PROHFXODU SDUDPHWHU LQ GHWHUPLQLQJ WKH UDWH RI FHOO DWWDFKPHQW XQGHU ZHOOGHILQHG IORZ FRQGLWLRQV 7KHVH UHVXOWV ZHUH DOVR FRQVLVWHQW ZLWK WKHRUHWLFDO SUHGLFWLRQV WKDW LQ WKH SUHVHQFH RI UHSXOVLYH QRQVSHFLILF IRUFHV N VKRXOG EH D VWURQJ IXQFWLRQ RI WKH OHQJWK RI EULGJLQJ PROHFXOHV LQ WKH GHSRVLWLRQ RI SDUWLFOHV WR D VXUIDFH 'LFNLQVRQ f $ TXDOLWDWLYH

PAGE 44

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fQRQDGKHVLYHf FRQWURO VXUIDFHV

PAGE 45

(VWLPDWHG /HQJWK RI FOI$ 6WDON 5HJLRQ QPf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

PAGE 46

)LJXUH 0HFKDQLVWLF PRGHO UHVXOWVf§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

PAGE 47

)LJXUH 0HFKDQLVWLF PRGHO UHVXOWVf§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

PAGE 48

DQG WKH VXUIDFH 7KLV UHSXOVLYH HIIHFW ZRXOG EH H[SHFWHG WR LQFUHDVH ZLWK LQFUHDVLQJ UHFHSWRU OHQJWK DV REVHUYHG /HQJWK RI 5HSHDW 5HJLRQ UHVLGXHVf )LJXUH 3ORW RI N YV UHFHSWRU OHQJWK IRU 6 DXUHXV PXWDQWV RQ )J DQG VNLP PLON FRQWURO VXUIDFHV f&)PXWDQWf UHIHUV WR WKH PXWDQW VWUDLQ GHYRLG RI FOXPSLQJ IDFWRU :LOGW\SH UHIHUV WR WKH QDWLYH 1HZPDQ VWUDLQ f60f LQGLFDWHV VNLP PLONFRDWHG FRQWURO VXUIDFH f)Jf LQGLFDWHV ILEULQRJHQFRDWHG VXUIDFH &RPSDULVRQ RI WKH DWWDFKPHQW NLQHWLFV UHVXOWV REWDLQHG ZLWK WKH QRQVSHFLILF DGKHVLRQ EORFNHUV %6$ DQG VNLP PLON UHVSHFWLYHO\ VKRZHG D VLPLODU WUHQG LH DQ LQFUHDVH LQ N ZLWK LQFUHDVH LQ WKH UHFHSWRU OHQJWK +RZHYHU WKH YDOXHV RI N DUH GLIIHUHQW IRU WKH WZR VHWV RI H[SHULPHQWV N YDOXHV IRU WKH H[SHULPHQWV XVLQJ VNLP PLON ZHUH ORZHU WKDQ WKH YDOXHV REWDLQHG ZKHQ %6$ ZDV XVHG DV WKH QRQVSHFLILF DGKHVLRQ EORFNHU $

PAGE 49

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

PAGE 50

WKH HOHFWURVWDWLF LQWHUDFWLRQV WR ORZHU WKH HQHUJ\ EDUULHU WR DWWDFKPHQW 7UXHVGDLO HW DO f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n O[O 2n 2n DQG 0 )LJXUH LV D SORW RI WKH QHJDWLYH ORJ RI WKH HOHFWURO\WH FRQFHQWUDWLRQ YV N 7KH GDWD VKRZHG D VKDUS LQFUHDVH LQ N ZLWK LQFUHDVLQJ HOHFWURO\WH FRQFHQWUDWLRQ ,Q WHUPV RI '/92 WKHRU\ WKLV REVHUYDWLRQ LV FRQVLVWHQW ZLWK D GHFUHDVH LQ WKH HQHUJ\ EDUULHU WR DWWDFKPHQW IRU EDFWHULD DV WKH FKDUDFWHULVWLF OHQJWK 'HE\H OHQJWKf DQG UHSXOVLYH HOHFWURVWDWLF IRUFHV GHFUHDVH 3ULHYH t 5XFNHQVWHLQ f (YHQWXDOO\ DWWUDFWLYH YDQ GHU :DDOV IRUFHV SUHGRPLQDWH DQG WKH UDWH RI DWWDFKPHQW EHFRPHV OLPLWHG E\ RQO\ WKH UDWH RI GLIIXVLRQ RU VHGLPHQWDWLRQ RYHU WKH ERXQGDU\ OD\HU QHDU WKH VXUIDFH

PAGE 51

[O2 FPPLQf )LJXUH $ SORW RI LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW N YV HOHFWRO\WH FRQFHQWUDWLRQ $WWDFKPHQW RI 6 DXUHXV ZLOGW\SH RQ JODVV ZDV PHDVXUHG DV D IXQFWLRQ RI HOHFWURO\WH FRQFHQWUDWLRQ

PAGE 52

&+$37(5 0$7+(0$7,&$/ 02'(/ )25 3$57,&/( 75$163257 72 7+( %27720 $77$&+,1* 685)$&( ,1 7+( 3$5$//(/ 3/$7( )/2: &(// 0RGHOLQJ 3DUWLFOH $WWDFKPHQW .LQHWLFV 7KH SXUSRVH RI WKLV VHFWLRQ LV WR JLYH D EDFNJURXQG RQ WKH PRGHOLQJ RI FROORLGDO SDUWLFOH DWWDFKPHQW DQG RQ ZKDW KDV EHHQ SXEOLVKHG LQ WKH OLWHUDWXUH DV ZHOO DV KLJKOLJKW WKH GLIIHUHQFHV EHWZHHQ WKH SURFHGXUH GHVFULEHG LQ WKLV GLVVHUWDWLRQ 6HFWLRQ f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f 2WKHU UHVHDUFKHUV KDYH VWXGLHG WKH NLQHWLFV RI ODWH[

PAGE 53

SDUWLFOH GHSRVLWLRQ IURP IORZLQJ VXVSHQVLRQV XVLQJ D VWDJQDWLRQ SRLQW IORZ FUHDWHG E\ D URWDWLQJ GLVN 'DEURV HW DO f RU VSKHULFDO DQG F\OLQGULFDO FROOHFWRUV XVXDOO\ DSSOLHG LQ ILOWUDWLRQ PRGHOLQJ $GDPF]\N t 9DQ GH 9HQ t $GDPF]\N HW DO f 7KH URWDWLQJ GLVN KDV EHHQ XVHG LQ SDUWLFOH GHSRVLWLRQ VWXGLHV GXH WR LWV VLPSOH DQG ZHOO GHILQHG K\GURG\QDPLFV DQG PDVV WUDQVIHU ,Q VWDJQDWLRQSRLQW IORZ WKH WKLFNQHVV RI WKH K\GURG\QDPLF DQG GLIIXVLRQ ERXQGDU\ OD\HUV DUH FRQVWDQW FORVH WR WKH VWDJQDWLRQ SRLQW ,I D WUDQVSDUHQW FROOHFWRU VXUIDFH LV XVHG WKHQ LQ VWDJQDWLRQ SRLQW IORZ RQH FDQ YLHZ XQGHU PLFURVFRSH WKH SDUWLFOH GHSRVLWLRQ 8QWLO WKH GHSRVLWLRQ UDWHV IRU SDUWLFOHV KDYH EHHQ HYDOXDWHG IURP WXUELGLW\ FKDQJHV RI D VXVSHQVLRQ EHIRUH DQG DIWHU OHDYLQJ D ILOWHU ,Q VXFK H[SHULPHQWV WKH DWWDFKHG SDUWLFOH GHQVLW\ ZDV PHDVXUHG HLWKHU LQ DQ LQGLUHFW ZD\ RU WKH FROOHFWRU VXUIDFH ZDV UHPRYHG IURP WKH H[SHULPHQWDO HQYLURQPHQW DQG SDUWLFOHV FRXQWHG XQGHU PLFURVFRSH SRVWH[SHULPHQWDOO\ 'LUHFW YLVXDOL]DWLRQ XQGHU PLFURVFRSH FDQQRW EH DSSOLHG ZKHQ D URWDWLQJ GLVN LV XVHG GXH WR LWV PRWLRQ 3DUWLFOH GHSRVLWLRQ LQ D VWDJQDWLRQSRLQW IORZ FDQ SURYLGH D ILUVW DSSUR[LPDWLRQ IRU SDUWLFOH GHSRVLWLRQ RQ FRPSOH[ V\VWHPV VXFK DV VSKHULFDO DQG F\OLQGULFDO FROOHFWRUV (OLPHOHFK HW DO f 2WKHU UHVHDUFKHUV KDYH VWXGLHG SDUWLFOH GHSRVLWLRQ LQ VWDJQDWLRQ SRLQW IORZ 'DEURV t YDQ GH 9HQ &KDUL t 5DMDJRSDODQ DE $GDPF]\FN HW DO f 3DUDOOHOSODWH IORZ FHOOV KDYH EHHQ XVHG WR VWXG\ SDUWLFOH DWWDFKPHQW NLQHWLFV RQWR VWDWLRQDU\ VXUIDFHV %RZHQ DQG (SVWHLQ f VWXGLHG ODWH[ SDUWLFOH GHSRVLWLRQ LQ D SDUDOOHOSODWH IORZ FHOO XVLQJ DQ LQGLUHFW PHWKRG IRU PHDVXULQJ WKH DWWDFKHG SDUWLFOH GHQVLW\ EDVHG RQ UDGLRDFWLYH LVRWRSH DFWLYLW\ PHDVXUHPHQWV RI WKH LVRWURSLFDOO\ DFWLYH SDUWLFOHV 2WKHU OLWHUDWXUH FRXQWV H[LVW RI WKH XVH RI D SDUDOOHOSODWH IORZ FHOO IRU

PAGE 54

WKHRUHWLFDO DQG H[SHULPHQWDO VWXGLHV RI SDUWLFOH GHSRVLWLRQ $GDPF]\FN t YDQ GH 9HQ 6MROOHPD t %XVVFKHU f +RZHYHU QRQH SUHGLFW DQ LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW LQGHSHQGHQW RI DQ\ JOREDO SDUDPHWHUV DV WKH PRGHO GHVFULEHG LQ 6HFWLRQ 5DGLDOIORZ FHOOV KDYH DOVR EHHQ XVHG WR VWXG\ VKHDUGHSHQGHQW EDFWHULDO DGKHVLRQ NLQHWLFV WR ELRPDWHULDO VXUIDFHV 'LFNLQVRQ mIH &RRSHU f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fQHDUf WKH VXUIDFH DW D ERXQGDU\ OD\HU GLVWDQFH 7 DQG WKH UHVXOWLQJ SRVLWLRQGHSHQGHQW YDOXH RI NHJ[f IRU D JLYHQ LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW N 7KH PRGHO SUHGLFWLRQ LV WKHQ ILW WR WKH H[SHULPHQWDO GDWD IRU NHJ[f E\ QRQOLQHDU OHDVWVTXDUHV UHJUHVVLRQ WR SURYLGH WKH EHVW HVWLPDWH RI N 7KH UDWLRQDOH LV WKDW N LV D ORFDO SDUDPHWHU ZKLFK GHSHQGV RQO\ RQ ORFDO VXUIDFH SURSHUWLHV DQG IORZ FRQGLWLRQV QRW WKH JOREDO WUDQVSRUW HIIHFWV LQ WKH 33)&

PAGE 55

U R R R R R 2 )LJXUH ,OOXVWUDWLRQ RI WKH FRQYHFWLYHGLIIXVLYH PRGHO IRU SDUWLFOH WUDQVSRUW LQ 33)& 6FKHPDWLF RI D ODWHUDO YLHZ RI WKH 33)& GLVSOD\LQJ D SDUDEROLF IORZ SURILOH DQG LOOXVWUDWLQJ WKH FRQGLWLRQV RI WKH FRQYHFWLYHGLIIXVLYH PDWKHPDWLFDO PRGHO XVHG WR FRUUHFW IRU WUDQVSRUW HIIHFWV WR WKH VXUIDFH RI WKH 33)& $ VFKHPDWLF RI WKH IORZ FHOO DQG PRGHO DVVXPSWLRQV DSSHDUV LQ )LJXUH 8SRQ QHJOHFWLQJ HQWU\ DQG H[LW HIIHFWV FI 6HFWLRQ f WKH SDUDEROLF YHORFLW\ SURILOH Y[]f FDQ EH VROYHG IURP WKH VWHDG\VWDWH 1DYLHU6WRNHV HTXDWLRQ WR \LHOG f ZKHUH 4 LV WKH YROXPHWULF IORZ UDWH Z LV WKH ZLGWK RI WKH FKDPEHU DQG K LV WKH JDS WKLFNQHVV 7KH VKHDU UDWH 6 RQ WKH VXUIDFH LV WKHUHIRUH ZKLFK DOORZV (T f WR EH UHZULWWHQ DV f

PAGE 56

7KH FRQFHQWUDWLRQ RI FHOOV LQ VXVSHQVLRQ F][f LV D IXQFWLRQ RI ERWK YHUWLFDO SRVLWLRQ DQG GRZQVWUHDP GLVWDQFH ([DFW VROXWLRQ RI F][f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f =f&r]f ]!7f f R] ZKHUH LV WKH SDUWLFOH GLIIXVLRQ FRHIILFLHQW DQG LV WKH VHGLPHQWDWLRQ YHORFLW\ $VVXPLQJ GLIIXVLRQ LV QHJOLJLEOH FRPSDUHG WR FRQYHFWLRQ LQ WKH [GLUHFWLRQ WKH VWHDG\ VWDWH FRQWLQXLW\ HTXDWLRQ LQ WKH UHJLRQ LV A aY[]faa Y aa ]! f R [ G] G ] $W VHSDUDWLRQ GLVWDQFHV OHVV WKDQ 7 WKH VWHDG\VWDWH IOX[ RI SDUWLFOHV RYHU ERXQGDU\ OD\HU LV DVVXPHG SURSRUWLRQDO WR WKH FRQFHQWUDWLRQ DW ] ) VXFK WKDW NF7 ]f ] 7f f 7KLV IOX[ LV WKHQ HTXDO WR WKH UDWH RI DFFXPXODWLRQ RI FHOOV RQ WKH VXUIDFH VXFK WKDW AI NF7]f DW f

PAGE 57

7KH DWWDFKHG FHOO GHQVLW\ FV[Wf LV DVVXPHG WR UHPDLQ VXIILFLHQWO\ ORZ RYHU WKH GXUDWLRQ RI WKH H[SHULPHQW WR QHJOHFW DQ\ GHSHQGHQFH RI N RQ FV[Wf DOVR H[SHULPHQWDOO\ LPSRVHGf 5HFDOO IURP WKH SUHYLRXV FKDSWHU WKDW WKH UDWH RI DFFXPXODWLRQ LV H[SHULPHQWDOO\ PHDVXUHG WR GHWHUPLQH WKH HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW NHMM[f ZKLFK LV GHILQHG LQ WHUPV RI WKH NQRZQ LQOHW FRQFHQWUDWLRQ &R LH a NA[fF f DW (TXDWLQJ (T f DQG f \LHOGV 0rf FUVf f N F 7KHUHIRUH GHWHUPLQLQJ N IURP NHJ[f UHTXLUHV VROXWLRQ RI WKH FHOO FRQFHQWUDWLRQ DW WKH ERXQGDU\ OD\HU F7[f IURP (T f 6ROXWLRQ RI (T f UHTXLUHV DQ LQLWLDO FRQGLWLRQ IRU WKH [GRPDLQ DQG WZR ERXQGDU\ FRQGLWLRQV IRU WKH ]GRPDLQ 7KH DVVXPHG LQLWLDO FRQGLWLRQ LV WKDW F FR DW [ DQG ] U 7KH DSSURSULDWH ORZHU ERXQGDU\ FRQGLWLRQ LV REWDLQHG E\ PDWFKLQJ WKH IOX[HV DW ] ) LH NF[ 7f &;f=f _U YVF[ 7f f R ] )RU WKH XSSHU ERXQGDU\ FRQGLWLRQ LW LV DVVXPHG WKDW WKH FRQFHQWUDWLRQ LV XQGLVWXUEHG IURP FR IDU IURP WKH VXUIDFH VXFK WKDW F F ]}7f f

PAGE 58

'LPHQVLRQOHVV (TXDWLRQV %HIRUH VROYLQJ WKH JRYHUQLQJ HTXDWLRQV ZHUH PDGH GLPHQVLRQOHVV E\ VFDOLQJ ZLWK UHVSHFW WR WKH FKDUDFWHULVWLF GLPHQVLRQV 7KH FKDUDFWHULVWLF GRZQVWUHDP GLVWDQFH ZDV WKH OHQJWK RI WKH FKDPEHU / PP 7KH FKDUDFWHULVWLF YHUWLFDO GLPHQVLRQ ZDV FKRVHQ WR EH 'YV ZKLFK LV WKH SUHGLFWHG GHFD\ OHQJWK RI FRQFHQWUDWLRQ QHDU WKH VXUIDFH XQGHU WKH VSHFLDO FDVH RI QR IORZ DQG QR DWWDFKPHQW LH WKH GHFD\ OHQJWK RI WKH H[SRQHQWLDO VROXWLRQ WR (T f XQGHU IRU 6 DQG N f 7KH GLPHQVLRQOHVV VSDWLDO FRRUGLQDWHV DUH WKHQ (T f EHFRPHV f f G X B G X G mn e 3 \Y\f 8 \ G \! ZKHUH X FFR DQG WKH FKDUDFWHULVWLF GLPHQVLRQOHVV SDUDPHWHUV DUH f 7f f 7KH GLPHQVLRQOHVV LQOHW DQG XSSHU ERXQGDU\ FRQGLWLRQV DUH XA \f O DQG Xe\ +f O UHVSHFWLYHO\ ZKHUH + } LV WKH FKRVHQ XSSHU ERXQGDU\ GLVWDQFH ,Q WKH UHVXOWV VKRZQ KHUH + ZDV FKRVHQ WR EH HTXDO WR ZKLFK ZDV FRQILUPHG WR EH DSSURSULDWHO\ ODUJH E\

PAGE 59

FKHFNLQJ WKDW WKH VROXWLRQ \LHOGHG G X G \ f 7KH GLPHQVLRQOHVV ORZHU ERXQGDU\ \ + FRQGLWLRQ LV G X G \ D fX=\f f \ U ZKHUH \ Y' DQG D NYV 0RGHO 6ROXWLRQ (T ZDV VROYHG QXPHULFDOO\ XVLQJ D FHQWHUHG ILQLWH GLIIHUHQFH DSSUR[LPDWLRQ RI WKH AGRPDLQ \ \ +f WR FRQYHUW (T f LQWR D V\VWHP RI Q RUGLQDU\ GLIIHUHQWLDO HTXDWLRQV RI WKH IRUP XM G 3 WQ\ f mP m X0 XLBO $\ $\ MQOf f G m L 3 UALUf X Q X $\ X XR $\ f ZKHUH Xf FRUUHVSRQGLQJ WKH ERXQGDU\ FRQGLWLRQ DW \ + DQG IURP WKH ILQLWH GLIIHUHQFH DSSUR[LPDWLRQ RI (T f X X ^D Of$\ f ZKLFK FRUUHVSRQGV WR WKH ERXQGDU\ FRQGLWLRQ DW \ \ )RU JLYHQ YDOXHV RI SDUDPHWHUV \ WM DQG D WKH V\VWHP RI RUGLQDU\ GLIIHUHQWLDO HTXDWLRQV VKRZQ LQ (TV f DQG f ZHUH QXPHULFDOO\ LQWHJUDWHG LQ WKH AGLUHFWLRQ XVLQJ D IRXUWK RUGHU 5XQJD.XWWD URXWLQH ZLWK VWHS VL]H FRQWURO 'RUPDQG t 3ULQFH f ,Q WKH ILWWLQJ RI WKH PRGHO VROXWLRQ WR WKH GDWD c Uc DQG \ ZHUH IL[HG SDUDPHWHUV

PAGE 60

DQG D ZDV D ILWWHG SDUDPHWHU 7R HVWLPDWH WKH IL[HG SDUDPHWHUV WKH GLIIXVLRQ FRHIILFLHQW ZDV HVWLPDWHG XVLQJ 6WRNHV (LQVWHLQ HTXDWLRQ IRU RQHPLFURQ GLDPHWHU SDUWLFOH WR ILQG [OnFPPLQ 7KH VHGLPHQWDWLRQ YHORFLW\ IRU 6 DXUHXV ZDV HVWLPDWHG SUHYLRXVO\ DW YL [OnFPPLQ 'LFNLQVRQ t &RRSHU f 7KH ERXQGDU\ OD\HU WKLFNQHVV ZDV FKRVHQ WR EH RQH FHOO GLDPHWHU RU 7 O[OnFP )URP WKHVH YDOXHV IW ZDV HVWLPDWHG WR HTXDO WR IRU 6n O Vn ZDV HVWLPDWHG WR EH DQG Uc ZDV HVWLPDWHG WR EH %HFDXVH WKH PRGHO ZDV XVHG WR HVWLPDWH WKH ILWWHG SDUDPHWHU D WKH VHQVLWLYLW\ RI WKH PRGHO VROXWLRQV WR XQFHUWDLQWLHV LQ WKH IL[HG SDUDPHWHUV ZDV UHOHYDQW WR WKH DFFXUDF\ RI WKLV HVWLPDWH 7KH VHQVLWLYLW\ RI WKH GLPHQVLRQOHVV VROXWLRQ IRU NHJYV YV [/ WR WKH GLPHQVLRQOHVV SDUDPHWHU D DQG \ DUH VKRZQ LQ )LJXUHV DQG UHVSHFWLYHO\ 7KH VROXWLRQ ZDV QRW VLJQLILFDQWO\ GHSHQGHQW RQ Uc RYHU D ODUJH UDQJH RI YDOXHV WKHUHIRUH D FRUUHVSRQGLQJ SORW IRU LV QRW VKRZQf )LJXUH VKRZV WKH HIIHFW RI GLPHQVLRQOHVV DWWDFKPHQW UDWH FRQVWDQW D RQ WKH UHVXOWLQJ HIIHFWLYH UDWH FRQVWDQW NHJ $V VKRZQ LQ WKLV SORW LI N LV ODUJHU WKDQ YY D!Of WKH PRGHO SUHGLFWV D GHFUHDVH LQ NHMM ZLWK GRZQVWUHDP GLVWDQFH GXH WR D GRZQVWUHDP GHSOHWLRQ RI FHOOV LQ VXVSHQVLRQ IURP UDSLG DWWDFKPHQW RI FHOOV QHDU WKH VXUIDFH FI (T f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

PAGE 61

UDQJH RI UHDVRQDEOH YDOXHV 7KLV LPSOLHV WKDW PRGHUDWH XQFHUWDLQW\ LQ WKHVH SDUDPHWHU HVWLPDWHV ZLOO QRW UHVXOW LQ ODUJH XQFHUWDLQWLHV LQ WKH PRGHO SUHGLFWLRQ RI NHMI 7KH RSWLPDO YDOXH N FRQVLVWHQW ZLWK WKH H[SHULPHQWDO GDWD IRU NHIO[f ZDV HVWLPDWHG E\ ILWWLQJ WKH PRGHO VROXWLRQ WR WKH H[SHULPHQWDO GDWD XVLQJ QRQOLQHDU OHDVW VTXDUHV UHJUHVVLRQ 7KLV UHTXLUHG UHFXUVLYH VROXWLRQ RI WKH PRGHO IRU LWHUDWLYH YDOXHV RI N WR PLQLPL]H WKH ZHLJKWHG VXP RI WKH VTXDUHV RI WKH UHVLGXDOV EHWZHHQ WKH PRGHO VROXWLRQ DQG WKH GDWD SRLQWV IRU NHMM[f 6SHFLILFDOO\ D UHJUHVVLRQ SURJUDP ZDV ZULWWHQ ZKLFK PLQLPL]HG WKH IXQFWLRQ = [ f NHIIMDWD rffn r@ f ZLWK UHVSHFW WR N ZKHUH FU LV WKH VWDQGDUG HUURU LQ WKH PHDVXUHG YDOXH RI NHII DW SRVLWLRQ [c 7KH XQFHUWDLQW\ LQ N FUN ZDV FDOFXODWHG E\ QXPHULFDOO\ FDOFXODWLQJ G6Nf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e VXJJHVWV WKDW WKH VKHDUGHSHQGHQFH RI NHMM UHVXOWHG IURP D JUHDWHU DFFXPXODWLRQ RI FHOOV QHDU WKH VXUIDFH DW ORZHU IORZ UDWHV DQG QRW GXH WR DQ\ HIIHFW RI IOXLG IRUFHV RQ WKH SUREDELOLW\ RI FHOO DWWDFKPHQW

PAGE 62

)LJXUH 3ORW RI WKH HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW YV GLPHQVLRQOHVV GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH GLPHQVLRQOHVV LQWULQVLF DWWDFKPHQW UDWH FRQVWDQW D V NYV

PAGE 63

'RZQVWUHDP 'LVWDQFH [/ )LJXUH 3ORW RI WKH GLPHQVLRQOHVV DWWDFKPHQW UDWH FRQVWDQW DV D IXQFWLRQ RI GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH WKH GLPHQVLRQOHVV IORZ SDUDPHWHU 7KH FXUYHV FRUUHVSRQG WR SUDQJLQJ IURP WR 7KLV SORW OOXVWUDWHV WKH PRGHO SUHGLFWLRQV DUH RQO\ ZHDNO\ VHQVLWLYH WR IRU ERWK ODUJH DQG VPDOO YDOXHV RI D

PAGE 64

NHIYV )LJXUH 3ORW RI WKH GLPHQVLRQOHVV DWWDFKPHQW UDWH FRQVWDQW DV D IXQFWLRQ RI GRZQVWUHDP GLVWDQFH IRU YDULRXV YDOXHV RI WKH WKH GLPHQVLRQOHVV ERXQGDU\ OD\HU WKLFNQHVV \ 7KH FXUYHV FRUUHVSRQG WR UDQJLQJ IURP WR 7KLV SORW OOXVWUDWHV WKH PRGHO SUHGLFWLRQV DUH RQO\ ZHDNO\ VHQVLWLYH WR \ IRU ERWK ODUJH DQG VPDOO YDOXHV RI D

PAGE 65

GLPHQVLRQOHVV )LJXUH 3ORW RI WKH GLPHQVLRQOHVV N DQG DYHUDJH NHII DV D IXQFWLRQ RI VKHDU UDWH IRU D 6 DXUHXV PXWDQW ZLWK D PLVVLQJ 5 UHJLRQ

PAGE 66

&+$37(5 (9$1(6&(17 :$9( /,*+7 6&$77(5,1* &283/(' :,7+ 7+5(( ',0(16,21$/ 237,&$/ 75$33,1* ,QWURGXFWLRQ 7R JDLQ IXOO XQGHUVWDQGLQJ RI WKH UROH RI VXUIDFH LQWHUDFWLRQV LQ EDFWHULDO DWWDFKPHQW LW LV GHVLUDEOH WR FRUUHODWH WKH PHDVXUH DWWDFKPHQW UDWH FRQVWDQWV ZLWK WKH VXUIDFH IRUFHV EHWZHHQ WKH FHOO DQG VXEVWUDWXP +RZHYHU QR PHWKRG ZDV SUHYLRXVO\ DYDLODEOH WR PHDVXUH WKH LQWHUDFWLRQ IRUFHV EHWZHHQ D VLQJOH EDFWHULXP DQG D VXUIDFH (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ FRXSOHG ZLWK 7KUHH 'LPHQVLRQDO 2SWLFDO 7UDSSLQJ (:/6'27f LV D QRYHO WHFKQLTXH GHYHORSHG WR GLUHFWO\ PHDVXUH WKH SRWHQWLDO HQHUJ\ RI LQWHUDFWLRQ EHWZHHQ D SDUWLFOH DQG DQ RSWLFDOO\ WUDQVSDUHQWf VXUIDFH IURP DQ LQLWLDO PHDVXUHPHQW RI WKH VHSDUDWLRQ GLVWDQFH EHWZHHQ WKH SDUWLFOH DQG WKH VXUIDFH %\ LQWHJUDWLRQ RI WKH SRWHQWLDO FXUYH RQH FDQ DOVR REWDLQ LQWHUDFWLRQ IRUFHV EHWZHHQ D SDUWLFOH DQG D VXUIDFH 6HYHUDO PHPEHUV RI P\ JURXS KDYH FRQWULEXWHG WR WKH GHYHORSPHQW RI WKH (:/6'27 0\ UROH ZDV LQ LQLWLDWLQJ WKH GHVLJQ DVVHPEO\ DQG SUHOLPLQDU\ WHVWLQJ RI WKH DSSDUDWXV $DURQ &ODSS KDV VLQFH EHHQ SULPDULO\ UHVSRQVLEOH IRU WKH DQDO\VLV IRU REWDLQLQJ IRUFH PHDVXUHPHQWV IURP WKH UDZ VLJQDO 6HFWLRQV RI WKLV FKDSWHU FORVHO\ IROORZ RXU SDSHU VXEPLWWHG WR 5HY 6FL ,QVWUXPHQWV FRDXWKRUHG E\ &ODSS P\VHOI DQG 'LFNLQVRQ

PAGE 67

$SSDUDWXV 3ULQFLSOH RI 2SHUDWLRQ 7KH (:/6'27 DSSDUDWXV FRQVLVWV RI VHYHUDO SDUWV ZKLFK DUH LQWHJUDWHG LQWR DQ H[LVWLQJ LQYHUWHG RSWLFDO PLFURVFRSH 'LDSKRW 1LNRQ &RUSRUDWLRQ 0HOOYLOOH 1
PAGE 68

7KH SULQFLSOH RI RSHUDWLRQ RI WKLV QRYHO WHFKQLTXH LV EDVHG RQ WKH JHQHUDWLRQ RI DQ HYDQHVFHQW ZDYH )LJXUH f DW D JODVVOLTXLG LQWHUIDFH XSRQ WKH WRWDO LQWHUQDO UHIOHFWDQFH RI D ODVHU EHDP GLUHFWHG WKURXJK D GRYH SULVP %. JODVV GHJUHH DQJOH 2SWRVLJPD 6DQWD $QD &$f 7KH GRYH SULVP LV RSWLFDOO\ FRXSOHG XVLQJ DQ LPPHUVLRQ RLO RI WKH VDPH LQGH[ RI UHIUDFWLRQ Q G f DV WKH SULVP DQG WKH JODVV FKDPEHU KROGLQJ WKH SDUWLFOH VXVSHQVLRQ 7KH ODVHU JHQHUDWLQJ WKH HYDQHVFHQW ZDYH LV D YDULDEOH P: UDQGRPO\ SRODUL]HG +HOLXP1HRQ ODVHU RI ZDYHOHQJWK QP 0HOOHV *ULRW ,UYLQH &$f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f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

PAGE 69

6FDWWHUHG ,QWHQVLW\ ]f f§ f" A )LJXUH 6FKHPDWLF RI WKH SULQFLSOH RI (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ 7KH QRYHOW\ RI WKLV WHFKQLTXH VWHPV IURP WKH LQWHJUDWLRQ RI D WKUHHGLPHQVLRQDO RSWLFDO WUDS )LJXUHV DQG f ZLWK WKH WRWDO LQWHUQDO UHIOHFWLRQ PLFURVFRS\ GHYHORSHG E\ 3ULHYH 3ULHYH t )UHM f 7KH WKUHHGLPHQVLRQDO RSWLFDO WUDS LV JHQHUDWHG E\ IRFXVLQJ D ODVHU EHDP IURP D O22P: GLRGH ODVHU &HOO 5RERWLFV $OEXTXHUTXH 10f PRXQWHG XQGHUQHDWK WKH QRVHSLHFH RI WKH PLFURVFRSH REMHFWLYH WKURXJK D KLJK QXPHULF DSHUWXUH REMHFWLYH 3ODQ )OXRU ; QXPHULFDO DSHUWXUH RLO LPPHUVLRQ 1LNRQ &RUSRUDWLRQ -DSDQf WR D GLIIUDFWLRQOLPLWHG VSRW RI GLDPHWHU a ; WKH ZDYHOHQJWK RI OLJKWf DW WKH IRFDO SRLQW %ORFN f 7KLV WLJKWO\ IRFXVHG EHDP SURGXFHV LQWHQVLW\ JUDGLHQWV LQ WKUHH GLPHQVLRQV JLYLQJ ULVH WR WKH ODWHUDO DQG D[LDO WUDSSLQJ FRPSRQHQWV $ SDUWLFOH ZLWK D UHIUDFWLYH LQGH[ KLJKHU WKDQ WKDW RI WKH VXUURXQGLQJ PHGLXP ZLOO H[SHULHQFH D WUDSSLQJ IRUFH GLUHFWHG WRZDUG WKH UHJLRQ RI KLJKHVW OLJKW LQWHQVLW\ %ORFN f

PAGE 70

)LJXUH (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ RI 6WDSK\ORFRFFXV DXUHXV ,PDJHV LQ WKH OHIW FROXPQ Df DUH RI D 6 DXUHXV EDFWHULXP IDU IURP WKH VXUIDFH 7KH WRS LPDJH LV REWDLQHG IURP D UHDO H[SHULPHQW DQG LW FDQ FOHDUO\ EH VHHQ WKDW WKH EDFWHULXP LV VFDWWHULQJ OHVV OLJKW WKDQ LQ WKH FDVH ZKHUH WKH SDUWLFOH LV FORVHU WR WKH VXUIDFH ULJKW FROXPQ LPDJHV Eff 7KH ERWWRP SLFWXUHV LQ ERWK FROXPQV DUH VFKHPDWLFV LOOXVWUDWLQJ WKH OLJKW VFDWWHULQJ PDJQLWXGH $ WUDSSHG SDUWLFOH FDQ EH PDQLSXODWHG ODWHUDOO\ E\ PRYLQJ WKH PRWRUL]HG VWDJH LQ WKH [ DQG \ GLUHFWLRQV DQG LW FDQ EH SRVLWLRQHG D[LDOO\ E\ YHUWLFDOO\ SRVLWLRQLQJ WKH REMHFWLYH ZLWK D KLJKUHVROXWLRQ VWHSSLQJ PRWRU /XGO (OHFWURQLF 3URGXFWV +DZWKRUQH 1
PAGE 71

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f PRXQWHG LQWR RQH RI WKH H\HSLHFHV 7KH LPDJH RI D WUDSSHG SDUWLFOH ZLWKLQ WKH JODVV FKDPEHU LV FDSWXUHG DQG VHQW WR D IUDPH JUDEELQJ ERDUG S 7HFKf ZKLFK SURFHVVHV WKH LPDJH DQG GLVSOD\V LW RQ D YLGHR PRQLWRU 7KHUHIRUH WKH WUDSSLQJ DQG PDQLSXODWLRQ RI D SDUWLFOH FDQ EH GLUHFWO\ REVHUYHG RQ WKH FRPSXWHU PRQLWRU

PAGE 72

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f 2ULHO ,QVWUXPHQWV 6WUDWIRUG &7f LV LQ D YHU\ QDUURZ UDQJH DURXQG QP WKH ZDYHOHQJWK RI WKH +H1H ODVHU 7KLV LV GRQH E\ PRXQWLQJ RQ WKH 307 D EDQG SDVV ILOWHU %3)f 1LNRQ -DSDQf ZKLFK EORFNV DOO RWKHU ZDYHOHQJWKV H[FHSW QP 6LQFH WKH VFDWWHULQJ LQWHQVLW\ RI D WUDSSHG SDUWLFOH LV

PAGE 73

H[SHFWHG WR EH TXLWH ORZ D YHU\ QDUURZ DSHUWXUH PP GLDPHWHUf LV DOVR PRXQWHG DKHDG RI WKH %3) RQ WKH 307 LQ RUGHU WR EORFN DQ\ H[WUDQHRXV VFDWWHULQJ IURP WKH V\VWHP 7KH KLJKO\ VHQVLWLYH 307 GHOLYHUV D FRQWLQXRXV YROWDJH VLJQDO ZKLFK LV OLQHDUO\ SURSRUWLRQDO WR WKH VFDWWHUHG LQWHQVLW\ 7KH YROWDJH VLJQDO IURP VFDWWHULQJ LQWHQVLW\ RI WKH SDUWLFOH LV SUHDPSOLILHG EHIRUH EHLQJ VHQW WR D GDWD DFTXLVLWLRQ ERDUG .HLWKOH\ ,QVWUXPHQWV &OHYHODQG 2+f DQG VWRUHG IRU IXUWKHU DQDO\VLV DQG GHWHUPLQDWLRQ RI WKH SDUWLFOHVXUIDFH VHSDUDWLRQ GLVWDQFH 7KLV UHVXOWV LQ D UHDOWLPH PHDVXUHPHQW RI WKH SDUWLFOH SRVLWLRQ QHDU WKH WHVW VXUIDFH 7KH IRUFH DQG SRWHQWLDO HQHUJ\ GHWHUPLQDWLRQ PHWKRGRORJLHV ZHUH GHULYHG LQ RXU JURXS DQG VXEPLWWHG IRU SXEOLFDWLRQ DORQJ ZLWK VRPH SUHOLPLQDU\ UHVXOWV &ODSS HW DO f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f WKDW DUH DYDLODEOH +RZHYHU IRU WKH SXUSRVH RI YDOLGDWLQJ WKH WHFKQLTXH VHOHFWLRQ RI D ZHOOGHILQHG V\VWHP LV LPSRUWDQW 2XU LQLWLDO VWXGLHV XVLQJ WKH WHFKQLTXH KDYH IRFXVHG RQ DTXHRXV VXVSHQVLRQV RI ]P SRO\VW\UHQH PLFURVSKHUHV s PP 3RO\VFLHQFHV :DUULQJWRQ 3$f LQWHUDFWLQJ ZLWK JODVV VXEVWUDWD

PAGE 74

7KH VXVSHQVLRQV SUHSDUHG IRU H[SHULPHQWV PXVW KDYH D UHODWLYHO\ ORZ SDUWLFOH GHQVLW\ VXFK WKDW WKH DSSDUDWXV REVHUYHV RQO\ WKH LQWHUDFWLRQ RI D VLQJOH SDUWLFOH DQG WKH WHVW VXUIDFH 6ROXWLRQV ZHUH SUHSDUHG XVLQJ PL SRO\VW\UHQH PLFURVSKHUH GHQVLWLHV RQ WKH RUGHU RI r SDUWLFOHVFP DGHTXDWHO\ FRQFHQWUDWHG WR DOORZ ORFDWLRQ RI D SDUWLFOH EXW VXIILFLHQWO\ GLOXWH WR DYRLG LQWHUDFWLRQV ZLWK RWKHU SDUWLFOHV GXULQJ PHDVXUHPHQWV ,Q DGGLWLRQ WR WKH OPL SDUWLFOHV D VPDOO DPRXQW RI MXP SRO\VW\UHQH PLFURVSKHUHV s SP 3RO\VFLHQFHVf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f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

PAGE 75

REMHFWLYH WUDS FHQWHUf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f WRZDUGV WKH JODVV VXUIDFH XQWLO LW ZDV SUHVVHG DV IDU DV SRVVLEOH LQWR WKH VXUIDFH E\ WKH WUDS 7KH PLUURU ZDV DGMXVWHG WR PD[LPL]H WKH VFDWWHUHG LQWHQVLW\ RI WKH SDUWLFOH 7KLV ODVHU DOLJQPHQW VWHS LPSURYHG WKH DELOLW\ RI WKH 307 WR VHQVH FKDQJHV LQ LQWHQVLW\ GXH WR VPDOO IOXFWXDWLRQV LQ WKH D[LDO GLUHFWLRQ DQG WKH VLJQDOWRQRLVH UDWLR :LWK WKH RSWLFV SURSHUO\ DGMXVWHG WKH SDUWLFOH ZDV ZLWKGUDZQ IURP WKH VXUIDFH WR D SRLQW IDU IURP WKH VXUIDFH DQf ZKHUH QR VFDWWHULQJ ZDV PHDVXUDEOH $ FRPSXWHU SURJUDP ZDV ZULWWHQ LQ 0LFURVRIW 9LVXDO %DVLF WR FROOHFW GDWD DQG SUHFLVHO\ FRQWURO WKH WUDS SRVLWLRQ LQ WKH D[LDO GLUHFWLRQ )RU HDFK WUDS SRVLWLRQ VFDWWHUHG LQWHQVLWLHV IURP WKH SDUWLFOH PHDVXUHG E\ WKH 307 ZHUH VDPSOHG E\ WKH GDWD DFTXLVLWLRQ V\VWHP IRU D JLYHQ WLPH LQWHUYDO DIWHU ZKLFK WKH SRVLWLRQ RI WKH WUDS ZDV PRYHG D VPDOO GLVWDQFH QP RU PRUHf WRZDUG WKH VXUIDFH 7KH SURFHGXUH ZDV UHSHDWHG XQWLO WKH D[LDO WUDSSLQJ IRUFH LPSRVHG RQ WKH SDUWLFOH ZDV LQVXIILFLHQW WR PRYH WKH SDUWLFOH DQ\ FORVHU WR WKH VXUIDFH ZLWK VXEVHTXHQW PRYHV RI WKH WUDS FHQWHU RXWVLGH RI WKH OLQHDU WUDSSLQJ IRUFH UHJLPHf 7KH H[SHULPHQWDO GDWD ZDV VDYHG DW HDFK WUDS SRVLWLRQ WR D 3& ILOH DQG DQDO\]HG DW D ODWHU WLPH

PAGE 76

'DWD $QDO\VLV 2YHUYLHZ 'DWD FROOHFWLRQ DQG D[LDO WUDS SRVLWLRQ DUH ERWK FRQWUROOHG YLD LQKRXVH VRIWZDUH 7KH GLJLWDO VDPSOLQJ SDUDPHWHUV RI WKH 3& GDWD DFTXLVLWLRQ ERDUG DUH XVHUGHILQHG DW UXQn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f WKH LQWHQVLW\ RI VFDWWHUHG OLJKW ]f IURP D SDUWLFOH GHFD\V H[SRQHQWLDOO\ ZLWK QRUPDO VHSDUDWLRQ GLVWDQFH ] IURP WKH UHIOHFWLQJ VXUIDFH ]f fHr f ZKHUH f LV WKH LQWHQVLW\ RI WKH SDUWLFOH FRQWDFWLQJ WKH VXUIDFH DQG ZKHUH WKH HYDQHVFHQW ZDYH GHFD\ FRQVWDQW LV GHWHUPLQHG E\

PAGE 77

" VLQ m @n f &DOFXODWLQJ WKH PHDQ SDUWLFOH SRVLWLRQ IURP WKH PHDQ FROOHFWHG LQWHQVLW\ IRU D ODUJH QXPEHU RI VDPSOHV LQYROYHV GHULYLQJ QRUPDOO\ GLVWULEXWHG VWDWLVWLFV IURP D ORJQRUPDOO\GLVWULEXWHG GDWD VHW %HFDXVH RI WKH VWLIIQHVV RI WKH RSWLFDO WUDS IOXFWXDWLRQV RI WKH SDUWLFOH SRVLWLRQ ZLWKLQ WKH WUDS DUH JHQHUDOO\ VPDOO DQG FORVHO\ REH\ D *DXVVLDQ GLVWULEXWLRQ ZLWK PHDQ ]f DQG YDULDQFH FU 7KLV FRQGLWLRQ DQG HTXDWLRQ f LPSO\ WKDW WKH IOXFWXDWLRQV LQ REH\ D ORJQRUPDO GLVWULEXWLRQ ZLWK PHDQ f DQG YDULDQFH D@ IURP ZKLFK WKH ILUVW WZR VWDWLVWLFDO PRPHQWV RI ] FDQ EH FDOFXODWHG DV IH! AOQ 7n OQ f LOQ ,f IU f f ZKHUH R@ LV WKH YDULDQFH LQ )LJXUH Df LV D KLVWRJUDP RI UHODWLYH VHSDUDWLRQ GLVWDQFH IRU D WUDSSHG OP SRO\VW\UHQH SDUWLFOH 7KH VROLG OLQH LV D QRUPDO SUREDELOLW\ GLVWULEXWLRQ EDVHG RQ WKH PHDQ DQG YDULDQFH RI VHSDUDWLRQ GLVWDQFH ZKLFK LQGLFDWHV WKDW WKH SRVLWLRQV DUH QRUPDOO\ GLVWULEXWHG )LJXUH Ef LV D KLVWRJUDP RI VDPSOHG LQWHQVLWLHV 9ROWDJH XQLWVf IRU D WUDSSHG RQHPLFURQ SRO\VW\UHQH SDUWLFOH 7KH VROLG OLQH LV D ORJQRUPDO SUREDELOLW\ GLVWULEXWLRQ ZKLFK GHPRQVWUDWHV WKDW WKH VDPSOH LQWHQVLWLHV IROORZ D ORJQRUPDO GLVWULEXWLRQ (TXDWLRQ f LQGLFDWHV WKDW PHDVXUHPHQW RI f DOORZV FDOFXODWLRQ RI WKH DEVROXWH VHSDUDWLRQ GLVWDQFH ]f +RZHYHU f LV QRW QHFHVVDU\ IRU ‘ 7KLV DVVXPSWLRQ LV YLRODWHG ZKHQ WKH QHJDWLYH FXUYDWXUH RI WKH LQWHUDFWLRQ SRWHQWLDO G! G LV FRPSDUDEOH WR WKH WUDS VWLIIQHVV \ ZKLFK RFFDVLRQDOO\ RFFXUV LQ UHJLRQV RI VWURQJ DWWUDFWLRQ DQG UHTXLUHV D PRUH FRPSOH[ DQDO\VLV QRW GLVFXVVHG KHUH

PAGE 78

FDOFXODWLRQ RI WKH UHODWLYH VHSDUDWLRQ GLVWDQFH EHWZHHQ GLIIHUHQW PHDVXUHG SRVLWLRQV WKHUHIRUH DOO SRVLWLRQV FDQ EH PHDVXUHG WR DQ XQGHWHUPLQHG FRQVWDQW f§,Qf )LJXUH +LVWRJUDPV IRU D WUDSSHG DQ SRO\VW\UHQH SDUWLFOH Df +LVWRJUDP RI VDPSOHG LQWHQVLWLHV IRU D WUDSSHG DQ SRO\VW\UHQH SDUWLFOH 7KH VROLG OLQH LV D SORW RI D ORJQRUPDO SUREDELOLW\ GLVWULEXWLRQ XVLQJ GDWD VWDWLVWLFV 7KH SORW LQGLFDWHV WKH ORJQRUPDO QDWXUH RI WKH VDPSOHG LQWHQVLWLHV DV H[SHFWHG GXH WR WKH H[SRQHQWLDO GHFD\ RI WKH LQWHQVLW\ ZLWK VHSDUDWLRQ GLVWDQFH DQG WKH QRUPDOO\ GLVWULEXWHG EHKDYLRU RI WKH SDUWLFOH SRVLWLRQ Ef +LVWRJUDP RI UHODWLYH VHSDUDWLRQ GLVWDQFHV IRU D WUDSSHG DQ SRO\VW\UHQH SDUWLFOH 7KH VROLG OLQH LV D SORW RI D QRUPDO SUREDELOLW\ GLVWULEXWLRQ XVLQJ VWDWLVWLFDO TXDQWLWLHV IURP WKH VDPSOHG GDWD 7KH SORW LQGLFDWHV WKDW WKH SDUWLFOH SRVLWLRQV DUH QRUPDOO\ GLVWULEXWHG DV H[SHFWHG EDVHG RQ WKH %URZQ IOXFWXDWLRQV RI WKH SDUWLFOH LQ WKH RSWLFDO WUDS

PAGE 79

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f M]]Rf f :LWK VWLIIQHVV SDUDPHWHU \ %ROW]PDQQfV GLVWULEXWLRQ ODZ LPSOLHV KHQFH WKH SDUDPHWHUV ]T DQG \ FDQ EH REWDLQHG IURP WKH VWDWLRQDU\ PHDQ DQG YDULDQFH RI SDUWLFOH SRVLWLRQV DW D WUDS SRVLWLRQ IDU HQRXJK IURP WKH VXUIDFH WR QHJOHFW VXUIDFH IRUFHV )RU WKLV FDOLEUDWLRQ SRVLWLRQ ] ]f DQG N7 f

PAGE 80

ZKHUH N LV %ROW]PDQQfV FRQVWDQW 7 LV WKH WHPSHUDWXUH RI WKH VXUURXQGLQJ OLTXLG DQG D@ LV WKH YDULDQFH RI SDUWLFOH SRVLWLRQ )URP HTXDWLRQ f WKH IRUFH RQ WKH SDUWLFOH IURP WKH RSWLFDO WUDS LV )^]f \] a rf f 7KLV UHTXLUHV D YDOXH IRU WKH WUDS FHQWHU SRVLWLRQ ]R )DU IURP WKH WHVW VXUIDFH WKH SDUWLFOH ZLOO IOXFWXDWH DURXQG WKH WUDS FHQWHU DQG WKXV KDYH D PHDQ SRVLWLRQ ]f HTXDO WR WKH SRVLWLRQ RI WKH WUDS FHQWHU ]T 7KH PHDQ IRUFH LQ WKLV FDVH ZRXOG EH ]HUR DQG WKH WUDS FHQWHU SRVLWLRQ ZRXOG EH HTXDO WR WKH PHDQ SRVLWLRQ $[LDO WUDSSLQJ HIILFLHQF\ GHSHQGV RQ WKH GLVWDQFH RI WKH SDUWLFOH DERYH WKH FRYHU JODVV :ULJKW HW DO f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

PAGE 81

)LJXUH 7UDS FDOLEUDWLRQ DQG PHDVXUHPHQWV XVLQJ WKH (:/6'27 WHFKQLTXH &DOFXODWLQJ )RUFH DQG 3RWHQWLDO 7KH YDOXH IRU WKH LQWHUDFWLRQ IRUFH IDU IURP WKH WHVW VXUIDFH LV ]HUR EXW DV WKH SDUWLFOH LV VWHSSHG DW NQRZQ LQFUHPHQWV WRZDUG WKH VXUIDFH WKH VXUIDFHV IRUFHV HYHQWXDOO\ EHJLQ WR GHIOHFW WKH PHDQ SRVLWLRQ ]f IURP WKH WUDS SRVLWLRQ ]R :LWK WKH YDOXHV IRU J DQG ]R NQRZQ WKH PHDQ IRUFH RI LQWHUDFWLRQ EHWZHHQ WKH SDUWLFOH DQG WHVW VXUIDFH FDQ EH FDOFXODWHG IURP ]f $W HTXLOLEULXP WKH VXUIDFH IRUFH PXVW EDODQFH WKH WUDSSLQJ IRUFH DW WKH SRVLWLRQ ]f LH

PAGE 82

)ƒ]ff a) ]!f U]f ]Rf f 7KH SRWHQWLDO HQHUJ\ RI PHDQ IRUFH DW DQ\ SRVLWLRQ ] FDQ EH HVWLPDWHG E\ QXPHULFDOO\ LQWHJUDWLQJ WKH PHDQ IRUFH GDWD f = $V WKH PHDQ RI WKH %URZQLDQ IRUFH DSSURDFKHV ]HUR LH IRU ODUJH VDPSOH VL]HVf WKH SRWHQWLDO RI PHDQ IRUFH ZLOO FRQYHUJH WR WKH PHDQ SRWHQWLDO 3UHOLPLQDU\ 5HVXOWV 3UHOLPLQDU\ UHVXOWV REWDLQHG LQ RXU ODE DUH LOOXVWUDWHG LQ )LJXUH Df ,W LV D SORW RI WKH LQWHUDFWLRQ IRUFH EHWZHHQ D OP SRO\VW\UHQH VSKHUH DQG D IODW JODVV VXUIDFH LQ GLVWLOOHG ZDWHU 7KH IRUFH GDWD ZDV QXPHULFDOO\ LQWHJUDWHG XVLQJ WKH WUDSH]RLG UXOH WR SURGXFH WKH FRUUHVSRQGLQJ SRWHQWLDO HQHUJ\ FXUYH LQ )LJXUH Ef (DFK GDWD SRLQW RQ WKH IRUFH DQG SRWHQWLDO HQHUJ\ SORWV ZDV FDOFXODWHG E\ DYHUDJLQJ WKH VHSDUDWLRQ GLVWDQFH RI WKH SDUWLFOH RYHU D SHULRG RI ILYH VHFRQGV DW PV LQWHUYDOV ,W ZDV IRXQG WKDW WKH PHDQ SRVLWLRQ RI WKH SDUWLFOH FRQYHUJHG WR D VWDWLRQDU\ YDOXH DIWHU DERXW VHFRQGV WKXV WKH VHFRQG VDPSOLQJ SHULRGV VKRZHG QR VLJQLILFDQW GLIIHUHQFH LQ WKH PHDVXUHG YDOXHV 7KH WUDS FHQWHU ZDV PRYHG E\ QP LQFUHPHQWV VPDOOHVW UHVROXWLRQ RI WKH VWHSSLQJ PRWRUf LQ WKH ] GLUHFWLRQ D[LDOf VXFK WKDW WKH PHDVXUHG LQWHQVLW\ YDULHG DV OLWWOH DV SRVVLEOH EHWZHHQ SRVLWLRQV )RU UHJLRQV ZKHUH WKH LQWHQVLW\ YDULHG OLWWOH ZLWK VXEVHTXHQW PRYHPHQWV WRZDUG WKH VXUIDFH LH IDU DZD\ IURP RU YHU\ FORVH WR WKH VXUIDFHf WKH WUDS FHQWHU ZDV PRYHG E\ ODUJHU LQFUHPHQWV QPf WR VSHHG XS WKH PHDVXUHPHQW

PAGE 83

5HODWLYH 6HSDUDWLRQ 'LVWDQFH QPf 5HODWLYH 6HSDUDWLRQ 'LVWDQFH QPf )LJXUH )RUFH DQG SRWHQWLDO HQHUJ\ FXUYHV IRU D IMP SRO\VW\UHQH SDUWLFOH LQWHUDFWLQJ ZLWK JODVV LQ ZDWHU Df )RUFH FXUYH IRU D ?MP SRO\VW\UHQH SDUWLFOH LQWHUDFWLQJ ZLWK JODVV LQ ZDWHU 7KH DEVFLVVD UHSUHVHQWV UHODWLYH VHSDUDWLRQ GLVWDQFH )DU IURP WKH VXUIDFH WKHUH LV ]HUR IRUFH EHWZHHQ WKH SDUWLFOH DQG VXUIDFH $V WKH SDUWLFOH QHDUV WKH VXUIDFH WKHUH LV DQ DWWUDFWLRQ QHJDWLYH IRUFH YDOXHVf DQG WKHQ D VWURQJ UHSXOVLRQ EHWZHHQ WKH SDUWLFOH DQG VXUIDFH Ef &RUUHVSRQGLQJ SRWHQWLDO HQHUJ\ REWDLQHG E\ QXPHULFDOO\ LQWHJUDWHG IRUFH GDWD LQ )LJXUH Df WR \LHOG WKH LQWHUDFWLRQ SRWHQWLDO SORW 7KHUH LV D VHFRQGDU\ PLQLPXP IROORZHG E\ D VWURQJ UHSXOVLYH HQHUJ\ EDUULHU DV WKH SDUWLFOH LV PRYHG WRZDUG WKH VXUIDFH

PAGE 84

7KH DFFXUDF\ RI WKH VHSDUDWLRQ GLVWDQFH PHDVXUHPHQWV LV VWURQJO\ GHSHQGHQW RQ WKH SUHFLVLRQ RI WKH LQYHUVH SHQHWUDWLRQ GHSWK )URP HTXDWLRQ f LV D VWURQJ IXQFWLRQ RI WKH LQFLGHQW DQJOH RI WKH ODVHU EHDP L MXVW DERYH WKH FULWLFDO DQJOH +RZHYHU HTXDWLRQ f DOVR VXJJHVWV WKDW EHFRPHV VRPHZKDW LQVHQVLWLYH WR FKDQJHV LQ WKH LQFLGHQW DQJOH IRU YDOXHV PXFK KLJKHU WKDQ WKH FULWLFDO DQJOH 7KH GDWD SUHVHQWHG LQ )LJXUH DftEf LV REWDLQHG DW D PHDVXUHG LQFLGHQW DQJOH RI DSSUR[LPDWHO\ r VXIILFLHQWO\ DERYH WKH FULWLFDO DQJOH RI r 7KHUHIRUH FDQ EH DFFXUDWHO\ HVWLPDWHG %\ REVHUYLQJ WKH WUHQGV LQ WKH IRUFH DQG SRWHQWLDO HQHUJ\ GDWD )LJXUH DftEff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a PLFURQf SDUWLFOH DQG D WHVW VXUIDFH 7KH VLPLODULWLHV EHWZHHQ 7,50 DQG

PAGE 85

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f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f PDQLSXODWLRQ RI WKH SDUWLFOH PDNH LW LGHDO IRU PHDVXULQJ ELRORJLFDO V\VWHPV &HUWDLQO\ JUHDW FDUH PXVW EH WDNHQ WR HQVXUH WKDW WKH RSWLFDO WUDS GRHV QRW GDPDJH WKH FHOOV 7UDSSLQJ ODVHUV LQ WKH QHDU LQIUDUHG ZDYHOHQJWK KDYH EHHQ VKRZQ WR EH LGHDO IRU VXFK DSSOLFDWLRQV %ORFN f

PAGE 86

7KH RWKHU QRWHZRUWK\ IHDWXUH RI WKH (:/6'27 WHFKQLTXH LV LWV HDVH RI LPSOHPHQWDWLRQ 7KH FRPSOHWH V\VWHP FRQVLVWV RI RQO\ D IHZ PRGLILFDWLRQV WR D VWDQGDUG OLJKW PLFURVFRSH DQG KHQFH LW DGDSWV ZHOO WR H[LVWLQJ PLFURVFRS\ VHWXSV DOORZLQJ IRU PXOWLSOH IXQFWLRQDOLW\

PAGE 87

&+$37(5 6800$5< $1' &21&/86,216 7KH JHQHUDO JRDO RI WKLV ZRUN KDV EHHQ WR HQKDQFH WKH IXQGDPHQWDO XQGHUVWDQGLQJ RI WKH SK\VLFRFKHPLFDO PHFKDQLVPV RI EDFWHULDO DGKHVLRQ WR VROLG VXUIDFHV WKURXJK WKH GHYHORSPHQW RI QRYHO PHWKRGRORJLHV DQG WKURXJK TXDQWLWDWLYH PHDVXUHPHQW RI WKH UROH RI FHOO VXUIDFH SURSHUWLHV RQ DWWDFKPHQW %HWWHU XQGHUVWDQGLQJ RI EDFWHULDO DGKHVLRQ PHFKDQLVPV FRXOG XOWLPDWHO\ UHVXOW LQ WKH GHVLJQ RI DQ LQIHFWLRQ UHVLVWDQW ELRPDWHULDO RU RI D WHFKQLTXH WKDW ZRXOG SUHYHQW SDWKRJHQLF PLFURRUJDQLVP DWWDFKPHQW WR WKH VXUIDFH RI LPSODQWHG ELRPDWHULDOV $V LW ZDV SUHYLRXVO\ GHVFULEHG EDFWHULD FDQ DGKHUH WR VXUIDFHV YLD QRQVSHFLILF FROORLGDO IRUFHV DQG YLD VSHFLILF ELQGLQJ RI FHOO VXUIDFH UHFHSWRUV WR DGVRUEHG SURWHLQV RQ WKH ELRPDWHULDO VXUIDFH ,W ZDV VKRZQ SUHYLRXVO\ WKDW DWWDFKPHQW RI 6 DXUHXV WR ILEULQRJHQ FRDWHG VXUIDFHV LQ ZHOOGHILQHG IORZ FRQGLWLRQV LV SULPDULO\ JRYHUQHG E\ VSHFLILF ELQGLQJ RI WKH FHOO VXUIDFH SURWHLQ fFOXPSLQJ IDFWRUf WR ILEULQRJHQ 'LFNLQVRQ HW DO f &OXPSLQJ IDFWRU KDV D K\GURSKLOLF VWDONOLNH UHJLRQ ZKLFK KDV EHHQ K\SRWKHVL]HG WR HQKDQFH LWV IXQFWLRQ E\ DOORZLQJ WKH PROHFXOH WR UHDFK RYHU HQHUJ\ EDUULHUV DQG DOORZ DWWDFKPHQW ,Q WKLV VWXG\ WKH DWWDFKPHQW NLQHWLFV RI 6 DXUHXV WR VXUIDFHV ZDV TXDQWLILHG DV D IXQFWLRQ RI WKH OHQJWK RI WKH FOXPSLQJ IDFWRU VWDON UHJLRQ ,Q WKLV WKHVLV LQWULQVLF DWWDFKPHQW UDWH FRQVWDQWV WR ILEULQRJHQ FRDWHG VXUIDFHV ZHUH

PAGE 88

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f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f 7KH ILUVW VHW RI H[SHULPHQWV UHSRUWHG DWWDFKPHQW RI 6 DXUHXV RQ ILEULQRJHQ XVLQJ ERYLQH VHUXP DOEXPLQ %6$f DV WKH QRQn VSHFLILF DGKHVLRQ EORFNHU $Q LQFUHDVH LQ WKH DWWDFKPHQW UDWH FRQVWDQW ZLWK LQFUHDVH LQ WKH OHQJWK RI WKH UHSHDW UHJLRQ RI WKH FOXPSLQJ IDFWRU ZDV REVHUYHG 7KH OHQJWK RI WKH FOXPSLQJ IDFWRU UHSHDW UHJLRQ SOD\V DQ LPSRUWDQW UROH LQ LQFUHDVLQJ DGKHVLRQ RI 6 DXUHXV

PAGE 89

WR ILEULQRJHQ FRDWHG VXUIDFHV 7KLV UHFHSWRU PD\ DFW DV D VWDON WR SURMHFW WKH OLJDQGn ELQGLQJ GRPDLQ RI 6 DXUHXV WKH ELRORJLFDOO\ DFWLYH ILEULQRJHQELQGLQJ UHJLRQf DZD\ IURP WKH FHOO VXUIDFH WR fUHDFKf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

PAGE 90

(:/6'27 LV D QRYHO WHFKQLTXH IRU PHDVXULQJ LQWHUDFWLRQ IRUFHV EHWZHHQ VPDOO SDUWLFOHV DQG VXUIDFHV DQG KDV SURYLGHG WKH ILUVW GLUHFW PHDVXUHPHQW RI WKH IRUFHV EHWZHHQ D VLQJOH EDFWHULXP DQG D VROLG VXEVWUDWH 7KLV WHFKQLTXH ZLOO DOORZ IXWXUH FRUUHODWLRQ EHWZHHQ LQWHUDFWLRQ IRUFHV DQG DWWDFKPHQW UDWHV ,Q FRQFOXVLRQ WKH UHVXOWV RI WKH ZRUN SUHVHQWHG KHUH VKRZ WKDW WKH UHFHSWRU OHQJWK LV D FULWLFDO SDUDPHWHU LQ WKH DWWDFKPHQW RI EDFWHULD WR VXUIDFHV ,Q VSHFLILF DWWDFKPHQW H[SHULPHQWV WKH LQFUHDVH LQ UHFHSWRU OHQJWK LQFUHDVHV DWWDFKPHQW RI EDFWHULD WR VXUIDFHV WKXV WKH VSHFLILF UHFHSWRUOLJDQG LQWHUDFWLRQV ,Q FRQWURO H[SHULPHQWV ZKHUH WKH OLJDQGV DUH DEVHQW IURP WKH VXUIDFH QRQVSHFLILF VWHULF UHSXOVLYH LQWHUDFWLRQV FRPH LQWR SOD\ WR GHFUHDVH DWWDFKPHQW ZLWK LQFUHDVH LQ WKH UHFHSWRU OHQJWK 7KH PDWKHPDWLFDO PRGHO IRU SDUWLFOH DWWDFKPHQW WR WKH VWXG\ VXUIDFH LQ WKH 33)& SUHGLFWHG D GHSHQGHQFH RI WKH DWWDFKPHQW UDWH RQ VKHDU ,Q SUHOLPLQDU\ H[SHULPHQWV VKHDU ZDV REVHUYHG WR KDYH DQ HIIHFW RQ EDFWHULDO DWWDFKPHQW LQ DFFRUGDQFH ZLWK WKH PRGHO SUHGLFWLRQV )XWXUH H[SHULPHQWV VWXG\LQJ LQ PRUH GHSWK WKH HIIHFW RI VKHDU ZLOO VKHG PRUH OLJKW RQ WKLV $OVR LQ WKH IXWXUH RQFH WKH (:/6'27 PHWKRGRORJ\ ZLOO EH ZHOO GHYHORSHG LW ZLOO DOORZ PHDVXUHPHQW RI DGKHVLRQ IRUFHV EHWZHHQ 6 DXUHXV PXWDQWV RI YDU\LQJ OHQJWKV DQG ILEULQRJHQ FRDWHG VXUIDFHV 7KLV ZRUN LQ WKLV GLVVHUWDWLRQ KDV SURYLGHG QHZ WRROV IRU PHDVXUHPHQW RI EDFWHULD VXUIDFH LQWHUDFWLRQV DQG HQKDQFHG WKH IXQGDPHQWDO XQGHUVWDQGLQJ RI VSHFLILF DQG QRQn VSHFLILF EDFWHULDO DWWDFKPHQW PHFKDQLVPV $ FRPELQDWLRQ RI PHDVXUHPHQWV RI EDFWHULDO DWWDFKPHQW NLQHWLFV DQG EDFWHULDVXUIDFH LQWHUDFWLRQ IRUFHV PD\ OHDG WR D PRUH FOHDU XQGHUVWDQGLQJ RI WKH PHFKDQLVPV RI EDFWHULDO DGKHVLRQ WR VXUIDFHV DQG KHQFH PD\ OHDG WR WHFKQLTXHV IRU SUHYHQWLRQ RI WKLV SURFHVV RU LWV HQKDQFHPHQW GHSHQGLQJ XSRQ WKH

PAGE 91

DSSOLFDWLRQ 7KLV FRPELQHG DSSURDFK PD\ DLG LQ WKH GHVLJQ RI LQIHFWLRQUHVLVWDQW ELRPDWHULDOV RU LQ WKH SUDFWLFDO DSSOLFDWLRQ WR LPSURYH ILOWHU GHVLJQ IRU VHSDUDWLRQ RI EDFWHULD IURP ZDVWHZDWHU VWUHDPV +RZHYHU WKH WHFKQLTXHV FDQ EH DSSOLHG WR HQKDQFH IXQGDPHQWDO XQGHUVWDQGLQJ RI EDFWHULDO DGKHVLRQ UHOHYDQW WR RWKHU DSSOLFDWLRQV VXFK DV WRRWK GHFD\ DQG ELR IRXOLQJ

PAGE 92

$33(1',; ,0$*( $1$/<6,6 352*5$06 7KLV DSSHQGL[ FRQWDLQV WKUHH LPDJH DQDO\VLV SURJUDPV &28170$& 6&$10$& DQG %5(3/$<0$& 7KH ILUVW RQH LV XVHG LQ REWDLQLQJ WKH SDUWLFOH FRQFHQWUDWLRQ LQ WKH VXVSHQVLRQ XVHG IRU WKH SDUDOOHO SODWH IORZ FHOO 33)&f H[SHULPHQWV DV ZHOO DV WKH HUURU LQ PHDVXULQJ LW 7KH VHFRQG RQH LV XVHG WR VWRUH WR WKH KDUG GULYH LQ D 0LFURVRIW (;&(/ ILOH VXFFHVVLYH LPDJHV RI SDUWLFOHV DWWDFKLQJ RQ WKH ERWWRP VXUIDFH RI WKH 33)& $IWHU WKH H[SHULPHQWV DUH FRPSOHWHG WKH VWRUHG LPDJHV IURP WKH (;&(/ ILOHV DUH UHFDOOHG DQG WKH QXPEHU RI SDUWLFOHV LQ HDFK LPDJH LV FRXQWHG XVLQJ WKH WKLUG SURJUDP %5(3/$<0$& )ROORZLQJ DUH WKH WKUHH SURJUDPV EULHIO\ GHVFULEHG DERYH 7KH\ DUH DOO ZULWWHQ LQ WKH $QDO\WLFDO /DQJXDJH IRU ,PDJHV $/,f XVHG ZLWKLQ WKH LPDJH DQDO\VLV SURJUDP 2SWLPDV )LOH QDPH &28170$& r7KLV PDFUR FRXQWV FHOOV SDUWLFOHVf LQ WKH 3HWURII+DXVHU FRXQWLQJ FKDPEHU DQG UHWXUQV D FHOO SDUWLFOHf FRQFHQWUDWLRQ DQG WKH HUURU LQ PHDVXULQJ LW 7KLV LV XVHG ODWHU LQ GHWHUPLQLQJ DQ HIIHFWLYH DWWDFKPHQW UDWH FRQVWDQW IRU SDUWLFOHV r 0D\ ODVW XSGDWH $FTXLUHf %ULJKWQHVVf &RQWUDVW f DQV 3URPSW+DV FDOLEUDWLRQ EHHQ VHW" ,I QRW FOLFN 1R! VHW FDOLEUDWLRQ IURP PHQX WKHQ UHVWDUW PDFURf LIDQVf ^SDXVH4`

PAGE 93

K/LE /RDG0DFUR/LEUDU\&nVXUIDFHRPOf 5XQ0DFUR &RSWLPDVPDFURVIORZFHOOQLQLWPDFf &RPPDQGV WR FOHDU WKH VTXDUH LQ WKH PLGGOH RI WKH VFUHHQ 5XQ0DFUR&RSWLPDVPDFURVIORZFHOOEFRPPOPDFf /RDGV PRWRUV IXQFWLRQ OLEUDU\ PZLQGRZ4 (QG FRPPDQGV IRU FOHDULQJ WKH VTXDUH LQ WKH PLGGOH RI WKH VFUHHQ 6HOHFW)XOO6FUHHQ f ,17(*(5 UFRQF4 WLHI ,17(*(5 QI DQV 758( ZKLOH DQVf ^ n 6HOHFW)XOO6FUHHQ f $FTXLUHf VKRZ0RYH WR QHZ ILHOG FOLFN 2. WKHQ VHOHFW 52, ZLWK PRXVH OHIW EXWWRQ RQO\ff 6HOHFW52,f $UHD) 52,> @52,> @fr52,> @5>@f )LHOG DUHD LQ DFWLYH XQLWV 6HW([SRUW P3W3RLQWV 758(f LIQI f )UHH]Hf GLII 6HW7KUHVKf $FTXLUHf` $XWR([WUDFW 758( )UHH]Hf QVK &RXQW&HOOVGLILf UFRQF UFRQF rQVKf $UHD)r Hfr HffWLHIW Hf VWDWV 0RPHQWVUFRQFf FRQF VWDWV>@ VWGHY VWDWV>O@ XQF LIQI!f ^VWGHY VWGHYVTUW rQIf XQF VWGHYr FRQF` QI QIO VKRZ7R7H[WQIf ILHOGV 7R7H[WFRQFbInf0 7R7H[WVWGHYbInf [O2$ FHOOVPO 7R7H[WXQFbInf b XQFHUWDLQW\f DQV 3URPSW$QRWKHU )LHOG"f ` VKRZ7R7H[WQIf ILHOGV 7R7H[WFRQFbInf 7R7H[WVWGHYbInf [O2$ FHOOVPO 7R7H[WXQFbInf b XQFHUWDLQW\f (QG RI 3URJUDP &28170$&

PAGE 94

)LOH QDPH 6&$10$& 7KLV PDFUR VFDQV PXOWLSOH ILHOGV HLWKHU XVHUVHOHFWHG RU E\ D UDVWHU VFDQ WKHQ VWRUHV WKH LPDJHV RI WKH ILHOGV WR KDUG GLVN $Q H[FHO ILOH LV SUHSDUHG ZLWK WKH LPDJH DQG ILHOG LQIRUPDWLRQ 7KH LPDJHV PXVW EH DQDO\]HG DQG GDWD H[WUDFWHG LQ D VHFRQG SURJUDP HJ %5(3/$<0$&f 0D\ ODVW XSGDWH ,QLWLDOL]H DFTXLUHf 2XWSXW/87 1RUPDO f %ULJKWQHVVf &RQWUDVWO f DQV 3URPSW+DV FDOLEUDWLRQ EHHQ VHW" ,I QRW FOLFN 1R! VHW FDOLEUDWLRQ IURP PHQX WKHQ UHVWDUW PDFURf LIDQVf SDXVHf` 5XQ0DFUR&RSWLPDVPDFURVIORZFHOOEFRPPOPDFf /RDGV PRWRUV IXQFWLRQ OLEUDU\ KOLE /RDG0DFUR/LEUDU\RSWGOJRPOf PZLQGRZ4 6HOHFW)XOO6FUHHQ f EHHSf K/LE /RDG0DFUR/LEUDU\&nVXUIDFHRPOf 5XQ0DFUR&237,0$60$&526IORZFHOO)2&3/$1(0$&f /RDGV IXQFWLRQV IRU FDOFXODWLQJ IRFDO SODQH 5XQ0DFUR&237,0$V0$&526IORZFHOOQLQLWPDFf /RDGV YDULRXV IXQFWLRQV 2SHQ ([FHO VKRZ2SHQ ([FHO WKHQ &OLFN 2.f IZ 52,> @52,>@ IO 52,>O@52,>OO@ )LHOG ZLGWK DQG OHQJWK ,QDPL 3URPSW (QWHU ,PDJH 3UHIL[&+$5,PJf VKHHWn6KHHWO K&KDQ ''(,QLWLDWH([FHOVKHHWf $UHD) IZrIO )LHOG DUHD LQ DFWLYH XQLWV *HW'DWH7LPHf 5QJ 5DQJHOOOOff ''(3RNH KFKDQ5QJ'DWH7LPHf 5QJ 5DQJHf ''(3RNH KFKDQ5QJ)LHOG $UHDf 5QJ 5DQJHf ''(3RNH KFKDQ5QJ$UHD)f 5QJ 5DQJHf ''(3RNH KFKDQ5QJ%DFWHULD &RQHf

PAGE 95

FFRQF 3URPSW(QWHU %DFWHULD &RQFHQWUDWLRQ 5($/Hf 5QJ 5DQJHf ''(3RNH KFKDQ5QJFFRQFf VKRZ :LWK MR\VWLFN PRYH WR FHQWHU RI VFDQ DUHD WKHQ FOLFN 2. WR IRFXV ZLWK DUURZ NH\Vf +HUH///f 6HWV UHIHUHQFH SRLQW SRVLWLRQ WR QW SURPSW(QWHU QXPEHU RI WLPH VWHSV n,17(*(5f PRGH 3URPSW6HOHFW VFDQ PRGH 8VHU 6HOHFWHG )LHOGV 5DVWHU 6FDQf,17(*(5 f /21* [>@\>@]>@SRV>@SRVO>@SRV>@SRV>@SRVWPS>@ LI PRGH f ^ QII SURPSW,QSXW QXPEHU RI ILHOGV,17(*(5f 6HW 5DVWHU DQG )RFXV r ,QSXW )LHOG &RRUGLQDWHV r L ZKLOH LQIIf ^ EHHSf WH[Wn0RYH WR ILHOG 7R7H[WLf DQG VHW IRFXV ZLWK MR\VWLFN VKRZWH[Wf SRV ILQGSRVf )LQGV FXUUHQW SRVLWLRQ [>L@ SRV>@ 6WRUH FXUUHQW SRVLWLRQ RI ILHOG L WR DUUD\V [\] \>L@ SRV>O@ ]>L@ SRV>@ 6WRUH SRVLWLRQ LQIR WR VSUHDGVKHHW 5QJ 5DQJH L Lf ''(3RNH KFKDQ5QJ[>LMf 5QJ 5DQJHLLf ''(3RNH KFKDQ5QJ\>LMf 5QJ 5DQJHLLf ''(3RNH KFKDQ5QJLf +O ` ` LI PRGH Of 6HW UDVWHU VFDQ DUHD E\ GHILQLQJ RSSRVLWH FRPHUV RI UHFWDQJOH ^ EHHSf VKRZ0RYH WR WRS OLPLW RI VFDQ DUHD ZLWK MR\VWLFN DQG IRFXVf SRVO ILQGSRVf PRYDEV///f :DLWf VKRZ0RYH WR LQOHW SRVLWLRQ ZLWK MR\VWLFN DQG IRFXVf SRV ILQGSRVf

PAGE 96

/21* [O /rSRVO >@ /HQJWK RI VFDQ DUHD /21* \O /rSRV>O@ %UHDGWK RI VFDQ DUHD /21* ]OO /rSRVO >@ +HLJKW DW LQOHW DUHD /21* ]O /rSRV>@ +HLJKW DW WRS DUHD PRYDEV///f :DLWf PRYUHO/ / /f SRV /// 3ODQH &DOF3ODQHSRVOSRVSRVf 7KLV IXQFWLRQV FDOFXODWHV SDUDPHWHUV IRU FDOFXODWLQJ IRFDO SODQH Q[[ 3URPSW,QSXW QXPEHU RI ILHOGV LQ WKH [GLUHFWLRQ GRZQVWUHDP GLUHFWLRQf,17(*(5f Q\\ 3UrPSW,QSXW QXPEHU RI ILHOGV LQ WKH \GLUHFWLRQ f,17(*(5f QII Q[[rQ\\ /21* G[ [OO/rQ[[Off /21* G\ \OO/rQ\\Off /21* G]O ]OOO/rQ[[Off /21* G] ]OO/rQ\\Off L ZKLOHLQ[[f ^ M ZKLOHMQ\\f ^ LIOG LrQ\\M &KDQJHV GRXEOH LQGH[ LMf WR VLQJOH LQGH[ LIOG LI Q[[!Of [>LIOG@ SRVO>@G[rLQ[[f` &DOFXODWHV ILHOG SRVLWLRQV LIQ[[ Of ^[>LIOG@ SRVO>@` LI Q\\!Of ^\>LIOG@ SRVO>O@G\rMQ\\f` LIQ\\ Of ^\>LIOG@ SRVO> @` LI Q\\!Of ^]>LIOG@ SRVO>@G]OrLQ[[fG]rMQ\\f` LIQ\\ Of ^]>LIOG@ SRVO>@` M ML ` L LO ` LIOG ZKLOHLIOGQIIf ^ [S [>LIOG@ \S \>LIOG@ ]S ]>LIOG@ ]S &DOF=[S\S3ODQHf VKRZ[S\S]Sf PRYDEV[S\S]Sf :DLWf W[W )LHOG f7R7H[WLIOGf VKRZ&RUUHFW IRFXV IRU W[Wf SRVWPS IPGSRVf ]>LIOG@ SRVWPS>@ &RUUHFWV IRFXV 6WRUH ILHOG FRRUGLQDWHV WR VSUHDGVKHHW

PAGE 97

5QJ 5DQJHLIOGLIOGf ''(3RNH KFKDQ5QJ[>LIOG@f 5QJ 5DQJHLIOGLIOGf ''(3RNH KFKDQ5QJ\>LIOG@f 5QJ5DQJHLIOGLIOGf ''(3RNH KFKDQ5QJLIOGf LIOG LIOGO ` ` 6WRUH ,PDJHV ,17(*(5 IRFIQIIOAO IRF>L@ O IRU GHOD\ DIWHU $FTXLUH IRU IRFXV FKHFN r 6WDUW 6FDQQLQJ 6HOHFWHG )LHOGV r LW 6WDUW7LPH '267LPHf &DOFXODWHV LQLWLDO WLPH ZKLOHLWQWf ^ L 7P '267LPHf6WDUW7LPH &XUUHQW WLPH 5QJ 5DQJHLWOLWOf ''(3RNH KFKDQ5QJLWf 6WRUH WLPH 5QJ 5DQJHLWLWf ''(3RNH KFKDQ5QJ7Pf ZKLOH LQIIf 0RYH WR SRVLWLRQV ^ [S [>L@ \S \>L@ ]S]>L@ PRYDEV[S\S]Sf :DLWf $FTXLUHf LI .H\+LWf f EHHSf VKRZ3DXVLQJ &OLFN 2. WR &RQWLQXHf` $OW .H\ LZ ZKLOH IRF>L@ Of tt NH\KLW2OA2[O2f tt LZfff $OORZ FKDQFH WR FKHFN IRFXV LI IRF>L@ O 'HOD\06O22f LZ LZO` LI NH\KLWf 2[O2f __ IRF>L@ ff 6+,)7! NH\ FDXVHV SDXVH DQG UHIRFXV LI QRW LQ IRFXV ^ %HHSf W[W )LHOG 7R7H[WLIOGf VKRZ&RUUHFW IRFXV IRU W[Wf SRVWPS IPGSRVf ]>L@ SRVWPS>@ 5HVHW ]SRVLWLRQ LIRF O IRF>L@ O .HHS DW LI IRFXVLQJ ZDV QHFHVVDU\ `

PAGE 98

HOVH ^IRF>L@ ` ,QDPH 7QDPL7R7H[WLfB7R7H[WLWf )UHH]Hf ,5RXWH &237,0$V0$&526IORZFHOOLPDJHV,QDPHWLIn 6DYHOPDJH ,URXWH 52, 758( )$/6(f 6DYH LPDJH 5QJ 5DQJHLWLLWLf ''(3RNH KFKDQ5QJ,QDPHf LI .H\+LWf [Of EHHSf IRF>L@ ` 6+,)7! .H\ KHUH DOORZV WKLV ILHOG WR EH IRFXVHG QH[W WLPH LI IRF>L@f ^+O` ` LW KO ` ''(7HUPLQDWHKFKDQf &ORVH FRQQHFWLRQ WR H[FHO (QG RI 3URJUDP 6&$10$& )LOH QDPH %5(3/$<0$& 7KLV PDFUR UHSOD\V WKH LPDJH VHTXHQFH WDNHQ E\ 6&$10$& WR H[WUDFW VWDWLVWLFV IRU EDFWHULD ,W LV ZULWWHQ LQ WKH $/, $QDO\WLFDO /DQJXDJH IRU ,PDJHVf ODQJXDJH RI 2SWLPDV 0D\ ODVW XSGDWH 6WDUW ,QLWLDOL]DWLRQ DQV 3URPSW+DV FDOLEUDWLRQ EHHQ VHW" ,I QRW FOLFN 1R! VHW FDOLEUDWLRQ IURP PHQX WKHQ UHVWDUW PDFURf LIDQVf SDXVHf` VKRZ2SHQ ([FHO 'DWDILOH WKHQ FOLFN 2.f 5XQ0DFUR &RSWLPDVPDFURVIORZFHOOQLQLWPDFf &RQWDLQV VRPH KDQG\ IXQFWLRQV VKRZ$IWHU FOLFNLQJ 2. RSHQ D WHVW LPDJH IURP WKH VHTXHQFHf 2SHQ,PDJHf 2SHQV 7HVWOPDJH IRU 7KUHVKROGLQJ GLII 6HW7KUHVKf 6HWV WKUHVKROGLQJ SDUDPHWHUV $XWR([WUDFW 758( (QG ,QLWLDWLRQ QW SURPSW(QWHU QXPEHU RI WLPH VWHSV URZVf,17(*(5f QIVW SURPSW(QWHU ILUVW ILHOG QXPEHU FROXPQVf,17(*(5f QOVW SURPSW(QWHU ODVW ILHOG QXPEHU FROXPQVf,17(*(5f

PAGE 99

&+$5 ,QVW SURPSW(QWHU ILOH SUHIL[ HJ ,PJ[B[WLI ILOH SUHIL[ LV ,PJfnn&+$5,PJf LW 6HW LW FRUUHVSRQGLQJ WR ILUVW XVHG WLPH VWHS IRU GDWD ZKLOHLWQWf ^ L QIVW ZKLOH L QOVWf ^ 5QJ 5DQJHLWLLWLf &+$5 ,QDPH ,QVW7R7H[WLfB7R7H[WLWf ,5RXWH &237,0$V0$&526)/2:&(//,0$*(6,QDPHWLI 2SHQOPDJH ,5RXWH 52,f QE &RXQW&HOOVG LIIf VKHHW 6KHHW K&KDQ ''(,QLWLDWH([FHOVKHHWf ''(3RNHKFKDQ5QJQEf ''(7 HUPLQDWHKFKDQf L LO ` (QG 'DWD ([WUDFWLRQ 5RXWLQH LW LWO ` (QG RI 3URJUDP %5(3/$<0$&

PAGE 100

/,67 2) 5()(5(1&(6 $GDPF]\N = DQG 7*09DQ GH 9HQ f'HSRVLWLRQ RI 3DUWLFOHV 8QGHU ([WHUQDO )RUFHV LQ /DPLQDU )ORZ WKURXJK 3DUDOOHO3ODWH DQG &\OLQGULFDO &KDQQHOVf &ROORLG ,QWHUIDFH 6FL f $GDPF]\N = 7 'DEURV -&]DPHFNL DQG 7*0 9DQ GH 9HD f 3DUWLFOH 7UDQVSRUW WR 6ROLG 6XUIDFHVf $GY &ROORLG ,QWHUIDFH 6FL f $GDPF]\N = 0 =HPEDOD % 6LZHN DQG &]DUQHFNL f.LQHWLFV RI /DWH[ 'HSRVLWLRQ IURP IORZLQJ VXVSHQVLRQVf &ROORLG ,QWHUIDFH 6FL f $WFKLQVRD / 7KH /RJQRUPDO 'LVWULEXWLRQ ZLWK 6SHFLDO 5HIHUHQFHV WR ,WV 8VHV LQ (FRQRPLFV $FDGHPLF 3UHVV &DPEULGJH f %DLHU 5 $( 0H\HU -5 1DWLHOOD 55 1DWLHOOD -0 &DUWHU f6XUIDFH 3URSHUWLHV 'HWHUPLQH %LRDGKHVLYH 2XWFRPHV 0HWKRGV DQG 5HVXOWVf %LRPHG 0DWHU 5HV f %HOO *, 0'HPER 3%RQJUDG f&HOO DGKHVLRQ &RPSHWLWLRQ EHWZHHQ 1RQVSHFLILF 5HSXOVLRQ DQG 6SHFLILF %RQGLQJf %LRSK\V f %LVQR $/ t )$ :DOGYRJHO (GV f,QIHFWLRQV $VVRFLDWHG ZLWK ,QGZHOOLQJ 0HGLFDO 'HYLFHVf $P 6RF 0LFURELRO :DVKLQJWRD '& f %ORFN 60 f0DNLQJ /LJKW :RUN ZLWK 2SWLFDO 7ZHH]HUVf 1DWXUH f %RGHD 0. DQG -, )ORFN f)LEULQRJHQ%LQGLQJ 3URWHLQ&OXPSLQJ )DFWRU IURP 6WDSK\ORFRFFXV DXUHXVf ,QIHFW ,PPXQ f %RZHD %' f6WUHDPLQJ 3RWHQWLDO LQ WKH +\GURG\QDPLF (QWUDQFH 5HJLRQ RI &\OLQGULFDO DQG 5HFWDQJXODU &DSLOODULHVf &ROORLG ,QWHUIDFH f f %RZHD % DQG 1 (SVWHLD f)LQH 3DUWLFOH 'HSRVLWLRQ LQ 6PRRWK 3DUDOOHO3ODWH &KDQQHOVf &ROORLG ,QWHUIDFH 6FL f %URZD / 0 :HVWE\ %( 6RXEHUELHOOH 3: 6]DZORZVNL .HPS 3 +D\ $* 'DOJOHLVK f2SWLPL]DWLRQ RI D 3HWSWLGH%DVHG ,QGLUHFW (/,6$ IRU WKH 'HWHFWLRQ RI $QWLERG\ LQ WKH 6HUXP RI +,9 6HURSRVLWLYH 3DWLHQWVf ,PPXQRO 0HWKRGV f -DQXDU\ f

PAGE 101

%XVVFKHU +$+:HHUNDPS f6SHFLILF DQG 1RQVSHFLILF ,QWHUDFWLRQV LQ %DFWHULDO $GKHVLRQ WR 6ROLG 6XEVWUDWDf )(06 0LFURELRO 5HY f %XVVFKHU +00 &RZDQ DQG +& YDQ GHU 0HL f2Q WKH 5HODWLYH ,PSRUWDQFH RI 6SHFLILF DQG 1RQVSHFLILF $SSURDFKHV WR 2UDO 0LFURELDO $GKHVLRQ f )(06 0LFURELRO 5HY f &HQWHU IRU 'LVHDVH &RQWURO t 7KH $VVRFLDWHG 3UHVV KWWSZZZZOVOLEQ\XVRHU FGFKWPO DFFHVVHG $XJXVW f &KDUL DQG 5 5DMDJRSDODQ f'HSRVLWLRQ RI &ROORLGDO 3DUWLFOHV LQ 6WDJQDWLRQ3RLQW )ORZf &KHP 6RF )DUDGD\ 7UDQV Df &KDUL DQG 5 5DMDJRSDODQ f7UDQVSRUW RI &ROORLGDO 3DUWLFOHV 2YHU (QHUJ\ %DUULHUVf &ROORLG ,QWHUIDFH 6FL Ef &KHXQJ $/ DQG 9$ )LVFKHWWL f7KH 5ROH RI )LEULQRJHQ LQ 6WDSK\ORFRFFDO $GKHUHQFH WR &DWKHWHUV LQ 9LWURf ,QIHFW 'LV f &KHXQJ $/ DQG 9$ )LVFKHWWL f7KH 5ROH RI )LEULQRJHQ LQ 0HGLDWLQJ 6WDSK\ORFRFFDO $GKHUHQFH WR )LEHUVf9 6XUJ 5HV f &KHZ + '6 :DQJ DQG 0 .HUNHU f(ODVWLF 6FDWWHULQJ RI (YDQHVFHQW (OHFWURPDJQHWLF :DYHVf $SSO2SW f $XJXVW f &ODSS $5 $* 5XWD DQG 5% 'LFNLQVRQ f7KUHH'LPHQVLRQDO 2SWLFDO 7UDSSLQJ DQG (YDQHVFHQW :DYH /LJKW 6FDWWHULQJ IRU 'LUHFW 0HDVXUHPHQW RI /RQJ 5DQJH )RUFHV EHWZHHQ D &ROORLGDO 3DUWLFOH DQG D 6XUIDFHf 6XEPLWWHG WR WKH ([SHU 0HWKRGV 'HFHPEHU f &R]HQV5REHUWV & -$ 4XLQQ '$ /DXIIHQEXUJHU f5HFHSWRU 0HLGDWHG &HOO $WWDFKPHQW DQG 'HWDFKPHQW .LQHWLFV ,, ([SHULPHQWDO 0RGHO 6WXGLHV ZLWK WKH 5DGLDO)ORZ 'HWDFKPHQW $VVD\f %LRSK\Vf &URXFK &+: )RZOHU 5( 6SLHU f7KH $GKHVLRQ RI $QLPDO &HOOV WR 6XUIDFHVWKH 0HDVXUHPHQW RI &ULWLFDO 6KHDU 6WUHVV 3HUPLWWLQJ $WWDFKPHQW RU &DXVLQJ 'HWDFKPHQWf &KHP 7HFKQRO %LRWHFKQRO % f 'DEURV 7 DQG 7*0 9DQ GH 9HQ &ROORLG ,QWHUIDFH 6FL f 'DEURV 7 DQG 7*0 9DQ GH 9HQ f$ 'LUHFW 0HWKRG IRU 6WXG\LQJ 3DUWLFOH 'HSRVLWLRQ 2QWR 6ROLG 6XUIDFHVf &ROORLG t 3RO\PHU 6FL f 'DEURV 7 = $GDPF]\N DQG &]DPHFNL &ROORLG ,QWHUIDFH 6FL f

PAGE 102

'DQNHUW $+ +RJW DQG )HLMHQ f%LRPHGLFDO 3RO\PHUV %DFWHULDO $GKHVLRQ &RORQL]DWLRQ DQG ,QIHFWLRQf &5& &ULW 5HY %LRFRPSDW f 'HUMDJXLQ %9 DQG /'/DQGDX $FWD 3K\VLFD &KLPLFD 8665 f 'LFNLQVRQ *0 DQG $/ %LVQR f,QIHFWLRQV $VVRFLDWHG ZLWK 3URVWKHWLF 'HYLFHV &OLQLFDO &RQVLGHUDWLRQVf ,QW $UWLI 2UJ f 'LFNLQVRQ 5% f4XDQWLWDWLYH &RPSDULVRQ RI &OXPSLQJ )DFWRU DQG &RDJXODVH0HGLDWHG 6WDSK\ORFRFFXV DXUHXV $GKHVLRQ WR 6XUIDFH%RXQG )LEULQRJHQ 8QGHU )ORZf ,QIHFW OPPXQ f f 'LFNLQVRQ 5% f$ '\QDPLF 0RGHO IRU WKH $WWDFKPHQW RI D %URZQLDQ 3DUWLFOH 0HGLDWHG E\ 'LVFUHWH 0DFUR PROHFXODU %RQGVf RI &ROORLG t ,QWHUIDFH 6FL f 'LFNLQVRQ 5% DQG 6/ &RRSHU f$QDO\VLV RI 6KHDU 'HSHQGHQW %DFWHULDO $GKHVLRQ .LQHWLFV WR %LRPDWHULDO 6XUIDFHVf $,&K( f f 'RQNHUVORRW -$ f)OXRURPHWULF 'HWHUPLQDWLRQ RI 'HR[LULERQXFOHLF $FLG LQ %DFWHULD ZLWK (WKLGLXP %URPLGHf $SSO 0LFURELRO f 'RUPDQG -5 DQG 33ULQFH f$ )DPLO\ RI (PEHGGHG 5XQJH.XWWD )RUPXODHf &RPS $SSO 0DWK f 'RXJKHUW\ 6+ f3DWKRELRORJ\ RI ,QIHFWLRQ LQ 3URVWKHWLF 'HYLFHVf 5HY ,QIHF 'LV f (OLPHOHFK 0 *UHJRU\ ; -LD 5 :LOOLDPV f0RGHOLQJ RI 3DUWLFOH 'HSRVLWLRQ RQWR ,GHDO &ROOHFWRUVf 3DUWLFOH 'HSRVLWLRQ DQG $JJUHJDWLRQ f§ 0HDVXUHPHQW 0RGHOOLQJ DQG 6LPXODWLRQ %XWWHUZRUWK+HLQHPDQQ HG *UHDW %ULWDLQ f (VSHUVHQ ) &OHPPHQVHQ f,VRODWLRQ RI D )LEURQHFWLQ %LQGLQJ 3URWHLQ IURP 6WDSK\ORFRFFXV DXUHXVf ,QIHFW t ,PP f f (ZHW] / 6WUDQJHUW f$ 6LPSOH 0HWKRG IRU 'HWHFWLRQ RI %DFWHULD ZLWK DQ $XWRPDWHG &KHPLOXPLQHVFHQFH 7HFKQLTXHf IHWR 3DWK 0LFURELRO 6FDQG f )HUJXVRQ -' f0LON 3URWHLQf KWWSZZZGKLDSVXHGXIHUVSURW KWPO DFFHVVHG -XQH f )RVWHU 7DQG 0F'HYLWW f6XUIDFH$VVRFLDWHG 3URWHLQV RI 6WDSK\ORFRFFXV DXUHXV 7KHLU 3RVVLEOH 5ROHV LQ 9LUXOHQFHf )(06 0LFURELRO /HWW f

PAGE 103

)RZOHU +: $0F.D\ f7KH 0HDVXUHPHQW RI 0LFURELDO $GKHVLRQf HG %HUNOH\ 5&: -0 /\QFK 0HOOLQJ 35 5XWWHU % 9LQFHQW 0LFURELDO $GKHVLRQ WR 6XUIDFHV &KLHFKHVWHU (QJODQG (OOLV +RUZRRG /WG f )UDQFRLV 3 3 9DXGDX[ 7)RVWHU '3 /HZ f+RVW %DFWHULD ,QWHUDFWLRQV LQ )RUHLJQ %RG\ ,QIHFWLRQVf ,QIHFWLRQ &RQWURO DQG +RVSLWDO (SLGHPLRORJ\ f f *ULVWLQD $* f%LRPDWHULDO &HQWHUHG ,QIHFWLRQ 0LFURELDO $GKHVLRQ YHUVXV 7LVVXH ,QWHJUDWLRQf 6FLHQFH 6HSWHPEHU f *ULVWLQD $* &' +REJRRG /; :HEE t 41 0\UYLN f$GKHVLYH &RORQL]DWLRQ RI %LRPDWHULDOV DQG $QWLELRWLF 5HVLVWDQFHf %LRPDWHULDOV f *ULVWLQD $* *LULGKDU %/ *DEULHO 37 1D\ORU 41 0\UYLN f&HOO %LRORJ\ DQG 0ROHFXODU 0HFKDQLVPV LQ $UWLILFLDO 'HYLFH ,QIHFWLRQVf ,QW $UWLI 2UJ 9RO 1R f +DUWIRUG 3 )UDQFRLV 3 9DXGDX[ DQG 7)RVWHU f7KH 'LSHSWLGH 5HSHDW 5HJLRQ RI WKH )LEULQRJHQ%LQGLQJ 3URWHLQ &OXPSLQJ )DFWRUf LV 5HTXLUHG IRU )XQFWLRQDO ([SUHVVLRQ RI WKH )LEULQRJHQ%LQGLQJ 'RPDLQ RQ WKH 6WDSK\ORFRFFXV DXUHXV &HOO 6XUIDFHf 0RO 0LFURELRO f f +HQFK // f6SHFLDO 5HSRUW 7KH ,QWHUIDFLDO %HKDYLRU RI %LRPDWHULDOV f %LRPHG 0DWHU 5HV f +HUUPDQQ 0 3 9DXGDX[ 3LWWHW 5 $XFNHQWKDOU 3' /HZ ) 6FKXPDFKHU 3HUGUHDX 3HWHUV DQG )$ :DOGYRJHO f)LEURQHFWLQ )LEULQRJHQ DQG /DPLQLQ $FW DV 0HGLDWRUV RI $GKHUHQFH RI &OLQLFDO 6WDSK\ORFRFFDO ,VRODWHV WR )RUHLJQ 0DWHULDOf ,QIHFW 'LV f +HUUPDQQ 0 4/DL 50 $OEUHFKW ') 0RVKHU 5$ 3URFWRU f$GKHVLRQ RI 6WDSK\ORFRFFXV DXUHXV WR 6XUIDFH %RXQG 3ODWHOHWV 5ROH RI )LEULQRJHQ)LEULQ DQG 3ODWHOHW ,QWHJU£LVf ,QIHFW 'LV f -DQVHQ % DQG 3HWHUV f)RUHLJQ %RG\ $VVRFLDWHG ,QIHFWLRQVf $QWLPLFURE &KHPRWKHU f -HQ &-6 /LQ f'LUHFW 2EVHUYDWLRQ RI 3ODWHOHW $GKHVLRQ WR )LEULQRJHQ DQG )LEULQ &RDWHG 6XUIDFHVf $P 3K\VLRO f .XXVHOD 3 7 9DUWLR 0 9XHQWR (% 0\KUH f$WWDFKPHQW RI 6WDSK\ORFRFFL DQG 6WUHSWRFRFFL RQ )LEURQHFWLQ )LEURQHFWLQ )UDJPHQWV DQG )LEULQRJHQ %RXQG WR D 6ROLG 3KDVHf ,QIHF DQG ,PPXQ f

PAGE 104

/DXUHQFH 0% &: 6PLWK 6* (VNLQ /9 0FOQWLUH f(IIHFW RI 9HQRXV 6KHDU 6WUHVV RQ &' 0HGLDWHG 1HXWURSKLO $GKHVLRQ WR &XOWXUHG (QGRWKHOLXPf %ORRG f /LQGRQ -1 DQG *1 0F0DQDPDUD f'RHV WKH &RQIRUPDWLRQ RI $GVRUEHG )LEULQRJHQ 'LFWDWH 3ODWHOHW ,QWHUDFWLRQV ZLWK $UWLILFLDO 6XUIDFHV"f %ORRG f f 0DUPXU $ :1 *LOO DQG ( 5XFNHQVWHLQ f.LQHWLFV RI &HOO 'HSRVLWLRQ 8QGHU WKH $FWLRQ RI DQ ([WHUQDO )LHOGf %XOOHWLQ RI 0DWKHPDWLFDO %LRORJ\ f 0DUVKDOO .& f0HFKDQLVPV RI %DFWHULDO $GKHVLRQ RQ 6ROLG:DWHU ,QWHUIDFHVf %DFWHULDO $GKHVLRQ 0HFKDQLVPV DQG 3K\VLRORJLFDO 6LJQLILFDQFH '& 6DYDJH HGf 3OHQXP 3UHVV 1HZ $@ f 3HWKLFD %$ f7KH 0HDVXUHPHQW RI 0LFURELDO $GKHVLRQf 0LFURELDO $GKHVLRQ WR 6XUIDFHV 5&: %HUNOH\ HGf (OOLV +DUZRRG /LPLWHG &KLFKHVWHU (QJODQG f 3ULHYH '& DQG 1$ )UHM f7RWDO ,QWHUQDO 5HIOHFWLRQ 0LFURVFRS\ $ 4XDQWLWDWLYH 7RRO IRU WKH 0HDVXUHPHQW RI &ROORLGDO )RUFHVf /DQJPXLU f 5XFNHQVWHLQ ( f'\QDPLFV RI &HOO 'HSRVLWLRQ RQ 6XUIDFHVf 7KHRU %LRO f

PAGE 105

5XFNHQVWHLQ ( DQG 3ULHYH '& f5DWH RI 'HSRVLWLRQ RI %URZQLDQ 3DUWLFOHV XQGHU WKH $FWLRQ RI /RQGRQ DQG 'RXEOHOD\HU )RUFHVf &KHP 6RF )DUDGD\ ,, f 5XWWHU 35 DQG % 9LQFHQW f7KH $GKHVLRQ RI 0LFURRUJDQLVPV WR 6XUIDFHV 3K\VLFR &KHPLFDO $VSHFWVf 0LFURELDO $GKHVLRQ WR 6XUIDFHV 5&: %HUNOH\ HGf (OOLV +DUZRRG /LPLWHG &KLFKHVWHU (QJODQG f 5\GHQ & 5XELQ DQG 3 6SH]LDOH f)LEURQHFWLQ 5HFHSWRUV IURP 6WDSK\ORFRFFXV DXUHXVf %LRO &KHP f f 6MROOHPD f'LUHFW 2EVHUYDWLRQ RI 3DUWLFOH 'HSRVLWLRQ LQ D 3DUDOOHO 3ODWH )ORZ &HOOf 0LFURELRO 0HWKRGV f 6MROOHPD DQG +%XVVFKHU f 'HSRVLWLRQ RI 3RO\VW\UHQH /DWH[ 3DUWLFOHV 7RZDUG 3RO\PHKW\OPHWKDFU\ODWH LQ D 3DUDOOHO 3ODWH )ORZ &HOOf &ROORLG ,QWHUIDFH 6FL f 6MROOHPD DQG +%XVVFKHU f 'HSRVLWLRQ RI 3RO\VW\UHQH 3DUWLFOHV LQ D 3DUDOOHO 3ODWH )ORZ FHOO 7KH ,QIOXHQFH RI &ROOHFWRU 6XUIDFH 3URSHUWLHV RQ WKH ([SHULPHQWDO 'HSRVLWLRQ 5DWHf &ROORLGV 6XUIDFHV f 6SLHOPDQ /$ DQG )LW]3DWULFN f7KHRU\ RI 3DUWLFOH &ROOHFWLRQ 8QGHU /RQGRQ DQG 'RXEOH/D\HU )RUFHV IRU $SSOLFDWLRQ WR :DWHU )LOWUDWLRQf &ROORLG ,QWHUIDFH 6FL f 6WDPP :( f,QIHFWLRQV 5HODWHG WR 0HGLFDO 'HYLFHVf $QQ ,QWHUQ 0HG f 6XJDUPDQ % DQG (-
PAGE 106

9DQ .RRWHQ 7* -0 6FKDNHQUDDG +& 9DQ GHU 0HL +%XVVFKHU f'HYHORSPHQW DQG 8VH RI 3DUDOOHO3ODWH )ORZ &KDPEHU IRU 6WXG\LQJ &HOOXODU $GKHVLRQ WR 6ROLG 6XUIDFHVf %LRPHG 0DW 5HV f 9DQ:DJHQHQ 5$ DQG -' $QGUDGH f)ODW 3ODWH 6WUHDPLQJ 3RWHQWLDO ,QYHVWLJDWLRQV +\GURG\QDPLFV DQG (OHFWURNLQHWLF (TXLYDOHQFHf &ROORLG ,QWHUIDFH 6FL f f 9DXGDX[ 3 f+RVW )DFWRUV 3UHGVSRVLQJ WR DQG ,QIOXHQFLQJ 7KHUDS\ RI )RUHLJQ %RG\ ,QIHFWLRQVf ,QIHFWLRQV DVVRFLDWHG ZLWK ,QGZHOOLQJ 0HGLFDO 'HYLFHV %LVQR $/HGf :DVKLQJWRQ '& $PHULFDQ 6RFLHW\ IRU 0LFURELRORJ\ 3UHVV f 9DXGDX[ 3 3LWWHW $ +DHEHUOL ( +XJJOHU 8( 1\GHJJHU '3 /HZ )$ :DOGYRJHO f+RVW )DFWRUV 6HOHFWLYHO\ ,QFUHDVH 6WDSK\ORFRFFDO $GKHUHQFH RQ ,QVHUWHG &DWKHWHUV $ 5ROH IRU )LEURQHFWLQ DQG )LEULQRJHQ)LEULQf ,QIHFW 'LV f 9DXGDX[ 3 5 6X]XNL )$ :DOGYRJHO -0RUJHQWKDOHU 8( 1\GHJJHU f)RUHLJQ %RG\ ,QIHFWLRQ 5ROH RI )LEURQHFWLQ DV D /LJDQG IRU WKH $GKHUHQFH RI 6WDSK\ORFRFFXV DXUHXVf ,QIHFW 'LV f Dff 9DXGDX[ 3 )$ :DOGYRJHO -0RUJHQWKDOHU 8( 1\GHJJHU f$GVRUSWLRQ RI )LEURQHFWLQ RQWR 3RO\PHWK\OPHWKDFU\ODWH DQG 3URPRWLRQ RI 6WDSK\ORFRFFXV DXUHXV $GKHUHQFHf ,QIHFW ,PPXQ Eff 9DXGDX[ 3 +
PAGE 107

:HLVV / f6WXGLHV RQ &HOOXODU $GKHVLRQ LQ 7LVVXH &XOWXUH ,9 7KH $OWHUDWLRQ RI 6XEVWUDWD %\ &HOO 6XUIDFHVf ([S &HOO 5HV f :LONLQVRQ 3& -0 /DFNLH -9 )RUUHVWHU *$ 'XQQ f &KHPRNLQHWLF $FFXPXODWLRQ RI +XPDQ 1HXWURSKLOV RQ ,PPXQH &RPSOH[&RDWHG 6XEVWUDWD $QDO\VLV DW D %RXQGDU\f9 &HOO %LRO f :ULJKW :+ *6RQHN DQG 0: %HPV f3DUDPHWULF 6WXG\ RI WKH )RUFHV RQ 0LFURVSKHUHV +HOG E\ 2SWLFDO 7ZHH]HUVf $SSO 2SW f f :X -7 3 =KDQJ *+ /LX DQG / :LOVRQ f'HYHORSPHQW RI DQ ,PPXQRDVVD\ 6SHFLILF IRU WKH 36 $$&7 &RPSOH[ :LWKRXW WKH 3UREOHP RI +LJK %DFNJURXQGf &OLQ /DE $QDO f f =HQJ )< -$ 2ND DQG 3+ :HLJHO f5HQDWXUDWLRQ DQG /LJDQG %ORWWLQJ RI WKH 0DMRU 6XEXQLW RI WKH 5DW $VLDORJO\FRSURWHLQ 5HFHSWRU DIWHU 'HQDWXULQJ 3RO\DFU\ODPLGH *HO (OHFWURSKRUHVLVf *O\FRELRORJ\ f $SULO f

PAGE 108

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

PAGE 109

, FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 5LFKDUG % 'LFNLQVRQ &KDLUPDQ $VVLVWDQW 3URIHVVRU RI &KHPLFDO (QJLQHHULQJ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ V -WMSQ ;; 6S\URV $ 6YRURQRV 3URIHVVRU RI &KHPLFDO (QJLQHHULQJ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ L/2 'LQHVK 2 6KDK 3URIHVVRU RI &KHPLFDO (QJLQHHULQJ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\A A &KDQJ :AIUDUN $VVRFLDWH3URIHVVRU RI &KHPLFDO (QJLQHHULQJ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ YY &KULVWRSKHU %DWLFK 3URIHVVRU RI 0DWHULDOV 6FLHQFH DQG (QJLQHHULQJ

PAGE 110

7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH &ROOHJH RI (QJLQHHULQJ DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ .DUHQ $ +ROEURRN 'HDQ *UDGXDWH 6FKRRO

PAGE 111

/' ‘ 5: 81,9(56,7< 2) )/25,'$