Citation
Approaching single molecule detection by laser-induced fluorescence of flowing dye solutions in a capillary

Material Information

Title:
Approaching single molecule detection by laser-induced fluorescence of flowing dye solutions in a capillary
Creator:
Lehotay, Steven John, 1965-
Publication Date:
Language:
English
Physical Description:
viii, 165 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Dyes ( jstor )
Fluorescence ( jstor )
Laser power ( jstor )
Lasers ( jstor )
Microscopes ( jstor )
Molecules ( jstor )
Photons ( jstor )
Signals ( jstor )
Solvents ( jstor )
Wavelengths ( jstor )
City of Orlando ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 159-164).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Steven John Lehotay.

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:
025224618 ( ALEPH )
AJG6068 ( NOTIS )
26576147 ( OCLC )

Downloads

This item has the following downloads:


Full Text









APPROACHING SINGLE MOLECULE DETECTION
BY LASER-INDUCED FLUORESCENCE OF
FLOWING DYE SOLUTIONS IN A CAPILLARY














By

STEVEN JOHN LEHOTAY


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

1992


UNIERmSITYO n a 1W L'!'~ U'' "






























This dissertation is dedicated to my father, Andrew L. Lehotay. Though he

died, he still lives in me.













ACKNOWLEDGEMENTS


I think Albert Einstein once said something like, "If I have seen farther, it is

because I have been standing on the shoulders of giants." I may not be able to see

very far, but it is more of a problem with "visual acuity" than elevation. The "giants"

in my case are Chris Stevenson, Ramee Indralingam, and Tye Barber whose research

efforts have had direct bearing on this dissertation. I would like to thank them for

their work, otherwise, this thesis would not have been possible.

Most of all, I would like to thank Dr. James D. Winefordner for his guidance,

insight, and knowledge. I am grateful for the opportunity to have been one of his

students and will forever be amazed and inspired by his intelligence, diligence, and

personality. He brought this project out of the clutches of despair with the simple

placement of a black piece of construction paper between the metal vapor filter and

the monochromator.

I also sincerely thank Dr. Benjamin W. Smith and Dr. Giuseppe A. Petrucci

for their help, patience, and friendship through the months of research leading to this

dissertation. I cannot count the numerous times I turned to Ben for advice, and

Giuseppe spent much time with me aligning the Ti:sapphire laser. He also modeled

the focusing aspects in a capillary presented in the dissertation.








I am also grateful for the work of several others in the group: Mike Wensing

wrote the program that calculates the Voigt profiles for the metal vapor cells; Nancy

Petrucci took the Raman spectra of the solvents; Yuan-Hsiang Lee helped collect

some of the data; and Wellington Masamba and Dennis Hueber helped with the

diode array software and use of the HR1000. Furthermore, I should thank the entire

JDW research group for their input, companionship, and spirit. They have all helped

make life in graduate school as stimulating, rewarding, and fun as it has been.

I very much appreciate the monetary support granted me by the state of Florida

and Texaco during my graduate school years.

Finally, I would like to thank my wife, Joann, for being loving, supportive, and

understanding during the stressful times, and all other times, leading to this

dissertation. I am a lucky soul to have her with me.













TABLE OF CONTENTS



ACKNOWLEDGEMENTS ...................................... iii

ABSTRACT ................................................ vii

CHAPTER 1 INTRODUCTION AND THEORY ..................... 1

Introduction ............................................. 1
Applications of Single Molecule Detection ................. 2
Choice of Analytical Technique for Single Molecule
Detection .................................... 6
Theory of Single Molecule Detection .......................... 8
Definitions ........................................ 9
Statistics of Data in Single Molecule Detection ............. 14
Theory of Laser-Induced Fluorescence ......................... 16
Sources of Noise in LIF and Means of Noise Reduction ........... 25
Laser Scatter ....................................... 25
Raman Scatter ..................................... 30
Background Fluorescence ............................. 30
History of Single Molecule Detection .......................... 32


CHAPTER 2 OPTIMIZATION OF INSTRUMENTATION AND
PARAMETERS ......................................... 41

The Metal Vapor Filter .................................... 41
Theory of the Metal Vapor Filter ....................... 41
Choice of Metal for the Filter .......................... 43
Calculation of Spectral Linewidths and Absorbances for Rb .... 45
The Laser .............................................. 55
Criteria of the Laser fo Single Molecule Detection ........... 56
The Ti:Sapphire Laser .............................. 58
The Ti:Sapphire Laser/Rb Metal Vapor Filter Combination ... 62
The Sample ............................................. 65
Choice of Analyte .................................. 65
Choice of Solvent ................................... 75








Sample Containment ............
Optical Considerations Regarding the
Focusing the Laser ........
Collection of the Fluorescence
Detection ....................
Choice of the Detector .....
Photon Counting ..........
Control of the Sample Flow .......
Experimental .................


..........
Capillary .

. .

. .

. .
.. l.. .. .


CHAPTER 3 RESULTS AND DISCUSSION ........

Studies of the Metal Vapor Filter ............
Absorption Properties ................
Transmittance Properties ..............
Additional Spectral Filtering ................
Spectral Filters .....................
Polarization ........................
R results ................................
Limits of Detection ..................
Noises of the System .................
Discussion of Limits of Detection .............


CHAPTER 4 CONCLUSIONS AND FUTURE WORK.

Conclusions .............................
Future W ork ............................
Elimination of Scatter ................
Other Future Improvements ...........


REFERENCE LIST............................

BIOGRAPHICAL SKETCH .....................


S83
S89
92
S94
100
100
103
108
112


119
119
130
131
131
134
136
136
141
145













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

APPROACHING SINGLE MOLECULE DETECTION
BY LASER-INDUCED FLUORESCENCE OF
FLOWING DYE SOLUTIONS IN A CAPILLARY

By

Steven John Lehotay

May 1992

Chairperson: James D. Winefordner
Major Department: Chemistry

The ability to detect single molecules in a solution has long been the ultimate

goal in ultratrace chemical analysis. In the strictest sense, single molecule detection

is defined as the efficient detection of a monomeric chemical species with near-100%

statistical certainty. Of current analytical techniques, laser-induced fluorescence has

the best chance of achieving single molecule detection due to its high sampling

efficiency, nondestructive probing, and high sensitivity. The limitations of laser-

induced fluorescence stem from noises associated with specular scatter from the

laser, Rayleigh and Raman scattering from the solvent, and background fluorescence

from sample contaminants. This dissertation concerns a new approach to eliminate

or greatly reduce these sources of noise for the purpose of single molecule detection.

In this approach, laser specular scatter is completely absorbed by a metal vapor filter,








which is simply a glass cell containing a metal element under reduced pressure.

When heated, the metal enters the vapor state and specifically absorbs the laser

scatter provided the laser spectral linewidth is narrower than the absorption band of

the metal vapor. Raman scatter from the solvent is reduced by containing the

sample in the small volume of a narrow capillary, and virtual elimination of

background fluorescence is accomplished by using near-infrared excitation with a

titanium:sapphire laser which specifically induces the fluorescence of a polymethine

dye. Currently, the Ti:sapphire laser is the only source with the power, tunability,

and narrow spectral linewidth that can be used in conjunction with a metal vapor

filter for single molecule detection. Rubidium was the element chosen for the metal

vapor filter due to its low melting point and ground state absorption transitions at

780.02 and 794.76 nm. In this dissertation, the theory of single molecule detection

and a review of previous approaches are presented, and the development and results

of this new approach are discussed. The lowest detection limit attained was 800

molecules of the dye, IR 140, in methanol flowing through a 140 pL probe volume.

Although single molecule detection was not achieved, the theoretically possible

detection limit of a single molecule could be attained through future experimental

modifications.













CHAPTER 1
INTRODUCTION AND THEORY


Introduction


The singular purpose of this research is to detect a single molecule in a

chemical solution. If one defines single molecule detection (SMD) as being able to

determine individual molecules in a sample containing billions of undetected

molecules of that species, the goal of this research has already been accomplished

in many instances by others. However, in order to fulfill the requirements for

practical application of SMD, the definition must be expanded to require the

individual detection of nearly every molecule of the analyte contained in the sample

solution. SMD remains elusive by this definition, and if SMD is to be truly realized,

this definition must be met.

Obviously, the achievement of SMD is no simple task. It has taken many

years of experimentation and instrumental refinements for other researchers to reach

near-SMD detection limits. In this project, it is hoped that with a thorough

understanding of the theory and careful design and implementation of experimental

components, the approach described in this dissertation will be able to achieve SMD

in its initial attempts.








2
The purpose of this dissertation is to 1) present the theoretical criteria for

SMD; 2) review previous research on the subject; 3) discuss the development of a

novel approach to SMD; 4) show the results of experiments designed to attain SMD;

and 5) discuss these results and future research possibilities.


Applications of Single Molecule Detection


For the analytical chemist, being able to detect individual atoms and

molecules in a sample is a worthwhile goal in itself, and indeed, it constitutes the

goal of this project. However, analytical chemists are not usually involved in the type

of research that could be labeled as "purely academic." As scientists, analytical

chemists often take pride in the practical nature of their work. Therefore, other than

the fundamental, inherently worthwhile aspect of pursuing SMD, there are several

important potential applications of this research.

DNA sequencing. In consideration of the highly publicized project to decode

the human DNA sequence, Jett et atL at Los Alamos National Laboratory have

proposed a strategy for rapid DNA sequencing relying on the detection of individual

nucleotides. This strategy involves specifically tagging the four different DNA (or

RNA) nucleotides with four different highly fluorescent dye molecules, then detecting

each of these molecules as they pass through a volume probed by a tightly focused

argon ion laser. If successful, the scheme is projected to sequence DNA 1000 times

faster than current techniques.' Although it remains to be seen if this application can

be accomplished as proposed, research concerning the detection aspects of the








3

project have come very close to SMD.2" This research at Los Alamos will be

discussed in more detail later in this chapter.

Immunoassays. Like DNA sequencing, the immunological assay has become

a very prominent analytical technique in biotechnology.7 Immunoassay is a general

type of technique that relies on antigen:antibody interaction for specific analyses of

many types of biological species. Currently, the technique requires considerable time

for an organism to produce sufficient antibody for an assay. If these type of analyses

could be performed on a molecular scale, much time and trouble could be saved, and

potentially more information of biological systems could be obtained. SMD in an

immunological system would enable the study of fundamental biological processes

at the molecular level.

Flow cytometry. Flow cytometric methods are another large area of research

in the biological sciences.8 In fact, the DNA sequencing strategy mentioned above

relies on a sheath flow cuvette, the main component of a flow cytometer, to contain

the sample.1 Currently, the ability to rapidly detect and sort single cells through flow

cytometric methods.8 In being able to detect single molecules as opposed to single

cells, flow cytometry would possibly allow separation of individual molecules.

Obviously, this would enable some very interesting studies to be performed.

Tracer studies. It is common practice in the determination of environmental

flow patterns, fates of chemicals in the environment, and metabolic pathways in

organisms to add a tracer compound into the system and follow its path. Due to

dilution effects, decomposition, and other losses, the analysis at some removed time








4

and place from the injection point may require the detection of the tracer in very

small quantities, potentially single molecules. These kinds of studies could be carried

out in a system on a global scale (such as the study of ocean currents or wind

patterns) or on a microscale (such as cell transport).

Fundamental research. For many physicists and physical chemists, it would

be interesting to determine whether the characteristics of an individual molecule

match its bulk properties. In the case of fluorescence, questions could be answered

concerning fluorescence lifetimes, quantum yields, quenching, and environmental

effects. Experiments involving SMD in solution may uncover important

physicochemical effects on a submolecular scale.

Detection in capillary electrophoresis and microcolumn chromatography. In

the 1980s, Jorgenson's group made a major advance in separation science with the

development of capillary zone electrophoresis (CZE).9 CZE is capable of achieving

separation efficiences as large as 106 theoretical plates and allows for the separation

of several types of chemicals in aqueous solution, particularly biological species, that

were not easily separated previously.10 Similarly, it has been known for some time

that decreasing column diameter (or particle size) in liquid chromatography increases

separation efficiency." With the recently developed commercial ability to coat very

narrow capillaries with stationary phase material, microcolumn high-performance

liquid chromatography (HPLC) has become a viable method for separation analysis.

However, due to the small quantities involved in microcolumn HPLC and CZE,

detection is often the limiting factor in applications of these techniques. By their








5
nature, separation methods dilute the original analyte concentration, and for

complicated separations requiring a great deal of time, the analyte concentration at

the detector may be a small fraction of the original value. For many sample limited

applications, very small amounts of a particular analyte may be separated by

electrophoretic or chromatographic methods, but cannot be measured due to

detection problems. In essence, for separation techniques utilizing capillaries, SMD

may be required for some separations; however, very low detection limits are

required for all of these separations.

Due to the great potential of capillary separation techniques and their need

for sensitive detectors, a great deal of research has been undertaken concerning

detection of solutions flowing in capillaries. The most promising detection method

has been laser-induced fluorescence utilizing fluorescent probe molecules tagged to

the analyte.2 Based on the use of capillaries in both microcolumn chromatography

and capillary electrophoresis and the great applicability and needs of these methods,

a capillary was chosen as the sample container for this project. This point will

become more prominent later in this dissertation.

Unforeseen applications. In general, applicability of any technique depends

on whether the technique suits the specific needs of the application. In most cases,

the need for a type of analysis is the impetus for its development; this is only

partially true in this instance. As mentioned previously, the purpose of this project

is to achieve SMD. The experiment is designed to be useful for detection in capillary

separation techniques, and in the future could be used as such, but for the present,








6
SMD is an end in itself. If this project is successful, the announcement of SMD to

scientists in various disciplines would possibly cause the design of experiments to suit

the needs of the detector. It is difficult to predict what outcomes would arise by

accomplishing SMD, but the possibilities are very exciting.


Choice of Analytical Technique for Single Molecule Detection


Of all analytical methods currently in use, laser-induced fluorescence (LIF)

has the best chance of success for SMD in practice. Other techniques have been

considered, as discussed in the following paragraphs, but all were dismissed based on

theoretical or practical grounds.

All electrochemical methods in solution are incapable of rapid single molecule

detection due to the insufficient detection sensitivities to overcome the inherent

noises and response times associated with electrochemical measurements. Also,

molecular absorption spectrometry is eliminated on the fundamental basis that the

absorption cross section of a molecule is much smaller than the cross section of any

known light source beam. Similarly, Raman spectrometry also lacks sensitivity based

upon the small scattering cross-section." Several other types of spectroscopic

detection methods such as magnetic resonance, refractive index, thermal lensing,

photoacoustic spectroscopy, and all forms of thermally excited emission spectrometry

are also incapable of single molecule detection with current technology. Additionally,

radiochemical techniques possess insufficient sensitivity to detect a single molecule

on a practical basis.14








7

Mass spectrometry. On the other hand, many techniques utilizing mass

spectrometric detection"1 have sufficient signal-to-noise ratios to detect single

molecules and are routinely capable of detecting individual molecular ions striking

the electron multiplier. In this respect, mass spectrometry is capable of single

molecule detection, but as mentioned earlier, practical SMD must occur with near

100% sampling efficiency. Such is not the case with mass spectrometry because of

unavoidable sample losses during introduction into the ionizing chamber and during

transport and separation of ions by the mass spectrometer. Mass spectrometry would

be a viable, practical approach to SMD, with the added benefit of molecular

identification, if these sample losses can be eradicated. Unfortunately, these are very

difficult problems to solve and it is not expected that mass spectrometry will be

useful for true SMD in the near future.

Scanning tunneling microscopy. A technique capable of true SMD, but not

yet useful for practical application, is scanning tunneling microscopy (STM). STM

has been demonstrated to allow viewing the atomic structure of molecules such as

benzene," and this capability has been suggested as an approach to DNA

sequencing.1 Scientists have developed STM techniques to the point that they are

able to view and manipulate single atoms contained on a metal surface.'

Despite these remarkable advances, STM is impractical for SMD of large

samples and in solutions. The analytical procedure for treatment of a sample for

STM involves coating the solid with an electrically conductive layer (commonly gold).

This process is as much art as science and can be very time consuming and expensive.








8
Once the sample is coated, the analysis can also be time consuming in trying to view

single molecules on a large surface. Furthermore, data interpretation is often a

subjective process often criticized for the imposition of imaginative viewing by the

researchers.

These problems are being addressed by current research, and if they can be

resolved, the potential for STM is enormous.19 Even so, STM has become a valuable

tool for the surface analysis of metals and many other solid materials, but it is not

currently practical as a general technique for SMD.20

Laser-induced fluorescence. As in mass spectrometry, lack of 100% sampling

efficiency plagues many atomic spectrochemical methods in flames and furnaces, such

as atomic fluorescence, laser-enhanced ionization, and resonance ionization.2122

However, the direct molecular fluorescence analysis in a solution is theoretically and

practically capable of detecting an individual highly fluorescent molecule. Laser-

induced fluorescence (LIF) is commonly the most sensitive fluorescence technique,

and for this and other reasons to be discussed in a later section in this chapter, LIF

is the chosen method in the attempt to attain SMD.


Theory of Single Molecule Detection


In recent years, since technology has been developed that is capable of

extremely low limits of detection, much discussion has appeared in the literature

concerning the requirements and statistics concerning detection of single species.21'

A useful summary of most of the concepts of SMD,28 as well as application of the








9
theory for the counting of atoms and molecules, has recently been given,' and

much of the following discussion is based on these writings.


Definitions


Due to the many variables involved in any given analytical procedure, and the

many differences in those parameters when compared with other techniques,

acceptable terminology for SMD must be defined. Otherwise, the claims to SMD for

one type of analysis may not actually meet the requirements for another.

Single molecule. In general chemistry textbooks, a molecule, or compound,

is defined as, "a substance composed of more than one element, chemically

combined."3 By this definition, single molecule detection is accomplished by looking

at DNA under a microscope or touching a piece of plastic. This is unsuitable for the

ego of analytical chemists who perform ultratrace analysis, so for the purposes of this

dissertation, the word "molecule" has been modified to signify a monomeric

compound of reasonable size. For tagging purposes, "reasonable size" depends on

the application, but in general, a molecular weight of less than 1000 g/mol is

considered reasonable.

Single molecule detection. In classic papers, Alkemade23' stated the criteria

for the detection of individual species. These works mostly concerned single atom

detection (SAD), but apply also to SMD. He mentioned that SAD involves two basic

requirements: 1) an efficiency of detection of unity and 2) attainment of the intrinsic

noise limit. These two factors are defined below.








10
As in the case of defining a "single molecule," this definition must be modified

as well. In his papers, Alkemade indicates that SAD refers only to the spatially and

temporally probed region and does not account for sampling efficiency. By his

definition, achieving SAD by focusing a pulsed laser to a very small region of a flame

remains possible, despite that in this system, for each atom that is detected, hundreds

of atoms are not aspirated into the flame, thousands do not pass through the focus

volume, and thousands more pass unprobed during the time between laser pulses.

For the purposes of practical use of SMD in solution by LIF, more stringent

requirements for SMD are necessary. With this in mind, the third criterion for SMD

is that the sampling efficiency, E,, must be nearly 100%. By this definition, it is fair

to say that SAD/SMD has not yet been achieved.

Efficiency of detection. The first criterion of Alkemade for SAD is that the

efficiency of detection must be unity, or -d = 1. This means that each time that a

single species appears in the probe volume during the probe time, it must be

detected. In many cases, it is possible that single molecules can be detected, but only

a certain percentage of the times that a molecule is present is it detected. In these

cases, the researchers cannot make a valid claim to SMD.

Extrinsic and intrinsic noise. According to Alkemade's second condition for

SAD, extrinsic noise must be eliminated at which time intrinsic noise becomes

prevalent. Extrinsic noise is the background produced by external factors such as

stray light, thermal fluctuations, and electric and magnetic fields. These factors can








11

be virtually eliminated experimentally with the techniques to be described later in

this chapter.

Intrinsic noise is the result of fluctuations of the signal itself. Sources of

intrinsic noise include the noise of the detector and power fluctuations of the source.

These are inherent features of the detection process that can be greatly reduced, but

will prevail as the limiting source of noise in the absence of extrinsic noise.

Limit of detection. One of the major figures of merit for any analytical

technique is limit of detection (LOD). LOD is defined as the concentration at which

the signal is 3 times larger than the standard deviation of the blank (O), or

LOD = 3q,/sensitivity, (1-1)

where sensitivity is the linear slope of the analytical calibration curve of the detection

system. This definition was developed at a meeting of the International Union of

Pure and Applied Chemists (IUPAC) in 1976 to settle differences in the subjective

way in which detection limits were previously determined.3

In the signal domain, the measure of LOD is given the symbol X, which is

defined as,

Xd = bl + 3bb (1-2)

where Cbi is the mean signal of the blank. When using this expression, a calibration

curve is not necessary; the analyte concentration at which this criterion is met is the

LOD.

Limit of guaranteed detection. Despite the IUPAC definition, LOD does not

always correspond to the best measure of lowest level of analyte detectability for a








12
system. In many circumstances, it is very difficult to actually observe a difference in

the signal at the limit of detection. Kaiser35 was aware of these problems and

introduced a term known as the limit of guaranteed detection (LOGD). Based on

statistical concepts, LOGD is set at twice the standard deviation requirement chosen

for LOD, which means

LOGD = 60b,/sensitivity, (1-3)

or in the signal domain,

X, = ibl + 6abl, (1-4)

where X, is the signal produced at the LOGD.

False positives and negatives. Another way of looking at detection in a

chemical analysis is the occurrence of false positives and false negatives. For

example, at the LOGD, the probability of the occurrence of a false positive is

essentially zero. A false positive, or type I error, occurs when the data exceed the

criteria for the detection of the analyte, but in actuality, noise, not signal, has been

the cause of the occurrence. False positives occur at a probability a. False

negatives, or type II errors, transpire when the analyte is present at a sufficient

concentration to be detected, but the signal does not exceed the detection level.

Type II errors occur with probability B.

In approaching the definition of detection limits based on the consideration

of types I and II errors, Xd (LOD in the concentration domain) corresponds to the

signal level that exceeds the background with a confidence of 1-a, and Xg (LOGD

in the concentration domain) is the signal level that gives a confidence of 1-f that








13

the analyte is actually being detected. Based on the definitions of LOD and LOGD,

the minimum confidence level is 99.86% which means that a and f must be 0.0014

or less.

Destructive and nondestructive probing. As mentioned previously, LIF has

a greater chance of success in realizing true SMD than other analytical methods.

The reason for this given earlier is the potential for high sampling efficiency by LIF.

However, an equally important factor is that LIF is a nondestructive probing method.

This means that the molecule is not destroyed in the detection process and more

than one detected event can occur per molecule. In fact, LIF is capable of producing

106 photons per fluorophore in a timespan of a few milliseconds which permits the

possibility of a higher noise level for an experiment that is still able to attain SMD.6

In destructive methods of detection, only one detection event can result from

each molecule. Destructive methods of detection include mass spectrometry,

radiometric methods, laser-enhanced ionization, and resonance ionization. In these

methods, the noise level must be extremely low, or signal of that one event must be

very high, to quantifiably detect a single molecule.

Symbols. Based on the definition of SMD, the laser must be continuous-wave

or very high repetition rate and probe the entire sample flow region. The focused

region of the laser is termed the probe volume, Vp, and the transit time an analyte

molecule in the Vp is the residence time, t,. During this interaction time, the number

of individual molecules) in the Vp, symbolized by Np, may give rise to a number of








14

detected events, N, (photoelectrons in the case of a photomultiplier tube). The mean

background level during t, is symbolized by jbI*


Statistics of Data in Single Molecule Detection


Poisson distribution. At low concentrations, Np follows a Poisson distribution,

and when photon counting is used for data collection at low levels, N, and ^bl also

follow a Poisson distribution. The Poisson probability distribution is given by'


P(X) = () ,(,), (1-6)
X!

where P(X) is the probability of X events occurring (with X being Np, N0 or noise),

and t is the mean value of X. In Poisson distributions, the variance equals the mean

(o2 = I, where a refers to the standard deviation)" which applies at higher means

when Gaussian and Poisson distributions have a large overlap.28

In typical analytical measurements, the occurrence of noise, signal, and

numbers of analyte species in the detection region follow Gaussian probability

distributions and the expressions for LOD and LOGD were designed for these types

of analyses. However, in the case of SMD, a problem with the determination of

LOD and LOGD through the calibration curve method is that the slope of the

calibration curve at concentrations much higher than the detection limit, which are

Gaussian in nature, may lead to errors at near-SMD levels, which follow a Poisson

distribution.28








15
Criteria of signal and noise for SMD. Based on a Poisson probability

distribution,37 the values for Xd and X for a low-level counting experiment at various

background levels are given in Table 1. The table was constructed as follows:2

1) The Poisson probability distribution with a mean equal to the chosen

background level was found in reference 37. The detection limit, Xd, was

determined as the number of counts (minus background) at which the sum of

the remaining probabilities of the distribution beyond Xd did not exceed

0.0014 (confidence = 1-a or 99.86%).

2) A distribution was then found such that the sum of the probabilities greater

than Xd exceeded 0.9986 (1-3f). The mean of this distribution is X.


Table 1. Signal levels required to achieve single molecule detection with 99.86%
confidence at given mean blank levels.

Mean Blank Level Detection Limit Guaranteed Limit
(Ibb, in counts) (X, in counts) (XY, in counts)
a < 0.0014 1 f 0.0014

0.001 1 6.6

0.25 4 12.7

1 6 16

5 14 28

10 22 39

25 42 64

50 73 102

100 132 169

All parameters defined in text.








16

Based on these confidence levels, Table 1 gives the signal level required for

the realization of SMD at a given background level. Because of the differences in

practical aspects and statistics of data at near-SMD levels, Curie8" has advocated that

the confidence level be set to 99.5% (1.65a) rather than the IUPAC level. This

would significantly lower the values for X, and X, reported in Table 1.

The only remaining theoretical topic is whether the proposed LIF system is

able to meet the signal and noise levels presented in Table 1. These considerations

of LIF and its sources of noise are discussed in the next section.


Theory of Laser-Induced Fluorescence


Fluorescence is a physical phenomenon involving the absorption of a photon

of light by a molecule, causing an electron to climb to a vibronic level of an excited

singlet state, followed by emission of a photon of typically lower energy as the

electron returns to the ground vibronic level. Not all molecules undergo

fluorescence, and the ones that do often are highly conjugated and contain aromatic

functional groups. Fluorescence is useful as an analytical procedure because the

intensity of the emission is dependent upon concentration, and the excitation and

emission wavelengths of the light give some information as to the identity of the

fluorescent species. Due to factors to be discussed below, fluorescence is often a

very sensitive type of analysis.

As the name implies, laser-induced fluorescence (LIF) simply uses a laser as

the excitation source for fluorometric analysis. Lasers are able to produce higher








17
spectral irradiance (W/cm2) than common broad-band sources such as the xenon arc

lamp. Thus, LIF typically gives lower LODs for the analysis of fluorescent

compounds. Also, the narrow emission bandwidth of the laser is useful in many

situations that require selective excitation and detection of a fluorescent probe

species added to a system.

The practical formula that gives the average number of detected events, N.,

in the type of IUF system used for SMD is3


No = ( )a,Y,( -)?Tt, (1-7)
hv *S 4r

where ~L is the laser power (W), hvL gives the energy per laser photon (J) with h

being Planck's constant (6.636 x 10"' J-s) and VL being frequency of the light (Hz),

SL is the cross-sectional area of the focused laser beam (cm2), oa is the cross section

of absorption for the molecule (cm2), YF is the fluorescence quantum efficiency, Op

is the solid angle of collection of the fluorescence (sr), t is the cathodic efficiency of

the detector dimensionlesss), T is the transmittance of the optical components

dimensionlesss), and tr is the residence time of each molecule in the probe volume

(s). These parameters are given by the manufacturer (as in the case of q), easily

measured (tL, T, and S), referenced in the literature (YF), or calculated from other

known parameters (A, OF, and t,). The following paragraphs discuss the

determination of these parameters for the system to be described in this thesis.

The cross section of absorption. The absorption cross section, OA, is not as

much of an actual "size" of a molecule as it is a statistical quantity. The units of cm2








18
arise from the fact that irradiance (W/cm2) is used in the expression to determine

the availability of light. The value for oa is the probability that the light available in

a certain area is absorbed.

The simplest way to determine o^ for a molecule is to measure the absorbance

of a known concentration in solution. Assuming the concentration falls within the

region of linear response, the molar absorptivity, EA (M1cmH), at the chosen

wavelength can be found from Beer's law,

A = ^AC, (1-8)

where A is absorbance dimensionlesss), f is path length (cm), and C is concentration

(M). By knowing ^A, A can be found from,

aA = 1000eA/N, (1-9)

where N is Avogrado's number (6.02 x 103). The absorption cross-section can also

be determined from fundamental parameters,9 but this method was used in this

project based on its simplicity and its use of an experimental measurement. The

absorption coefficients for the dyes to be tested in this project (given in Chapter 2)

are 200,000 M-cm7': therefore, the values for oA are approximately 3 x 1016 cm2.

Fluorescence quantum yield. The quantum efficiency of fluorescence, or

quantum yield, Y,, is the probability of emission of a fluorescence photon once a

photon has been absorbed. In mathematical form, Y, is given by39











Y = k (1-10)
kp+k.

where kF is the rate of fluorescence (s'), and k, is the rate of nonradiative

deactivation (s-) of the excited singlet state (Si). Nonradiative decay of S, occurs

through the processes of external conversion (collisional deactivation), internal

conversion (nonfluorescent de-excitation), and intersystem crossing (S, to triplet,

T).39 These factors are difficult to quantify, and depend strongly upon the molecule

itself and environmental factors such as temperature, pressure, solvent, and presence

of other species. Therefore, the Y, for a particular system must be measured.

Basically three different methods are used to quantify Y,, the simplest and

most common of which is the Parker-Rees method.39 This involves the comparison

of the fluorescence emission of an unknown fluorophore, Y,,, with a fluorophore of

known quantum efficiency, YF,k. Assuming constant power of excitation, this method

uses the equation,


YF = YF,k EAk (1-11)
EA,u'u

where the differences in the absorption coefficients and detection efficiencies (a

function of wavelength) for the known and unknown fluorophores must be taken into

account. The most commonly used reference fluorophore is quinine which has a

quantum yield of 0.59 in an acidic aqueous solution.








20

The Parker-Rees method is not always simple to use, and due to other

correction factors not included above, is not always accurate. An easier way to

determine YF for a particular system is to search the physical chemistry literature for

an accurate determination of Y, in the solvent to be used. THrough an intensive

literature search, it was found that that the dye has Yp 142

Solid angle of collection. A microscope objective was to be utilized for the

collection of fluorescence in this approach. The solid angle of collection, OF, can be

calculated from the stated specifications of the manufacturer and the distance of the

point source fluorescence emission to the collection optics. It should be stressed,

however, that erroneously large values for 0, result if one does not use the proper

equation for the calculation. The general expression normally used to calculate 0 for

a lens is given as9


,F = rtan20, (1-12)

where 0 is the angle defined by a line extended from the point source to the center

of the lens and a second line from the point source to the edge of the lens. The

problem with this equation is that it breaks down at large solid angles encountered

with microscope objectives. The calculated value of OF can exceed the true value by

a factor of 15% for 0 = 30* and by a factor of 300% for a 0 = 60*. The correct

expression for Ol is given by4


S= 27(1-cos0) = 4rsin 2(). (1-13)

With this equation, 0, is less susceptible to error than with the previous equation.








21
When using a microscope objective, 0 is found from the stated numerical

aperature, N.A. of the objective, in that


N.A. = n sine = -, (1-14)
2f

where n is the refractive index dimensionlesss) of the medium between the object

and objective, 4 is the aperture (cm) and f is the focal length (cm) of the objective.

The calculated OF for the microscope objective to be used in this experiment is 1.5

sr which, when corrected for the 4r sr of a sphere, corresponds to a collection

efficiency of 11.9% of a point source. The magnification (40X) and N.A. (0.65) for

the microscope objective used in this project, as well as other considerations of the

collection optics, are discussed in Chapter 2.

Calculation of the expected LIF signal. Now that the parameters of IF have

been discussed, it is possible to estimate N, for the conditions of this experiment.

Table 2 contains the values of the the relevant parameters determined by the

methods described above and in Chapter 2. By incorporating these values into

Equation 1-7 above, the theoretical signal level of a single molecule in this project

is 81 counts per 2 ms measurement period of the photon counter.

Although the accuracy of these parameters is thought to be very good, and the

Equation 1-7 is theoretically sound, the determination of N, may not be truly valid

by this method for two reasons: 1) the potential for photodecomposition of the dye

before t, and 2) optical saturation of fluorescence. According to Equation 1-7,

simply increasing laser power, decreasing focus size, or increasing t, allows one to







22

obtain as large a signal as desired. However, physical limitations to negate this

possibility are described below.


Table 2. Parameters of the system to be used in the attempt of laser-induced
fluorescence detection of single molecules and the expected signal level
calculated from Equation 1-7.
Parameter Value
Laser Power, 'L 200 mW
Laser Frequency, vL 3.77 x 1014 Hz
Laser Focus Area, SL 3.5 x 10-5 cm2
Absorption Cross-Section, aA 3 x 10-16 cm2
Dye Quantum Yield, Y, 1
Collection Efficiency, QF/4T 0.119
Detector Efficiency, q 0.1
Optical Transmittance, T 0.5
Residence Time, t, 0.002 s
Expected Signal Level, N, 81 photoelectrons


Optical saturation. Optical saturation occurs when the fluorescence signal

becomes independent of laser power. This effect happens when the rate of

fluorescence is limited by the time it takes the molecule to cycle through the

excitation/de-excitation process before it becomes available to go through another

cycle. Optical saturation" is characterized by the relationship,


ELoAYF 2 (g )A,21 (1-15)
hvL g +g2

where EL' is the laser irradiance at optical saturation (W/cm2), g, and g2 are the

statistical weights dimensionlesss) of the ground state and excited singlet state, Si,








23
respectively, and A,2 is the Einstein coefficient of spontaneous emission of the

fluorescence (s"'). This rate of spontaneous emission is simply the inverse of the

fluorescence lifetime, rp (s). To determine EL', which is the laser irradiance when

the slope of a log-log plot of signal versus irradiance becomes 0.5, the relevant

equation is41


g2 4,rhc 2A,
EL = ( )(h ), (1-16)
gl+g, YX.s

where c is the velocity of light (3 x 1010 cm/s), AXF is the full width at half maximum

of the fluorescence excitation and emission bands (cm), and X is the excitation

wavelength of the laser (cm). The value of A2, for the dye to be used in this system

is 1.25 x 109 s"' (rp = 800 ps);42 the bandwidths of the excitation/emission spectra are

50 nm; the laser excitation wavelength to be used is 794.76 nm. The values for g,

and gz are assumed to be equal (for So and S, vibronic levels, this is usually a valid

assumption). When these values are put into Equation 1-16, the resulting saturating

irradiance is 5900 W/cm2, which corresponds to 207 mW laser power in the 3.5 x 105

cm2 focus size. By this account, the stated laser power above does not saturate the

transition and the calculated 81 photoelectrons per counting period is theoretically

obtainable with the stated parameters.

If laser irradiance, EL, is increased above EL', N, is no longer given by

Equation 1-7, but by











N, = ( )A21( F )Tt. (1-17)
g +g' 4,r

It is not desirable to require the use of this equation, because in an actual analysis,

the laser power should be kept just below saturating conditions. For EL > EL', the

signal changes by less than a factor of 2, but the noise continues to increase with EL.

Photodecomposition. The second pitfall of Equation 1-7 is degradation of the

the dye before it emits as many photons as the theory predicts in the 2 ms sampling

time. Other researchers performing SMD have encountered this problem4"4 and

developed a procedure to determine the optimum parameters to reduce dye

degradation. In this optimization technique, the end result is that t, should roughly

correspond to the time it takes the molecule to decompose under the laser irradiance

of the experiment. For the dyes in these experiments, the molecule typically

undergoes 106 fluorescence cycles before it degrades.43 If this value holds true for

the dye to be used in this project, the time it would take for the molecule to

photodecompose at the conditions listed in Table 2 is 146 ms, nearly 100 times

longer than the measurement time of this experiment. Based on this estimate, it is

hoped that photodecomposition will not become a factor in this attempt at SMD.

However, if optical saturation occurs to the point that the molecule undergoes

fluorescence at the theoretical limiting rate (Az2 = 1.25 x 109 for the fluorophore42

of Table 2), it would only take 0.8 ms to go through 1 million cycles.

Noise level required for SMD. Based on the theory of LIF presented here,

the average signal level should consist of 81 counts above background per counting







25

interval. Using the statistical theory applied to SMD presented earlier, the maximum

mean noise level, ^pb, permitted during the 2 ms interval to attain an LOD of 1

molecule with 99.86% confidence is 56 counts. For LOG = 1 with 99.86%

confidence to be achieved with 81 counts above background, IbI would have to be 35

counts or less. Remember in the calculation of these values it is assumed that the

data follow a Poisson probability distribution with the variance equal to the mean.

There is some question as to whether this is true when actual data are collected.3

A key to the success of this project is the reduction of noise to this required level.

The next section is a discussion of sources of noise in LIF and how they can be

eliminated or reduced.


Sources of Noise in LIF and Means of Noise Reduction


The three limiting extrinsic sources of noise for typical LIF experiments

consist of scattered light, Raman scatter, and background fluorescence. Other less

severe sources of noise exist, but by far, these three are the largest sources.


Laser Scatter


Laser scatter can be divided into two categories, specular scatter arising from

reflections from optics and other surfaces, and Rayleigh scattering from the solvent.

Laser specular scatter. In nearly every analysis utilizing LIF, scattered light

from the laser constitutes the most severe source of noise. This feature is not

surprising when one considers the amount of light that is delivered by the laser. For








26
example, the number of photons focused on the sample by the 200 mW laser at

794.76 nm to be used in this project is 8 x 1017 s' or 1.6 x 1015 photons during the 2

ms counting period. Of these, only 1 photon in 117 billion (a~SL) is absorbed by a

molecule in this time, which essentially still leaves some 1.6 x 10is nonabsorbed

photons. Realizing that 0,/4T is nearly 12%, qi is 10%, and T is 50% for the system,

the detector would produce = 9.6 x 1012 photoelectrons per 2 ms assuming that the

laser light is scattered isotropically from a point source. However, this assumption

is not true; the vast majority of the laser light continues unhindered through the

sample container and is not scattered within the 1.5 sr region collected by the

microscope objective (which is why fluorescence is collected 90 from the angle of

excitation). For the sake of argument, assume only 0.1% of the photons are

scattered. This still corresponds to nearly 10 billion photoelectrons. The 81

photoelectrons emitted by a single fluorophore in the same time period pales in

comparison. If the above assumptions are correct, an absorbance of > 10 is required

to reduce the scatter to a level compatible for SMD with this system.

Monochromators. As can be surmised based on this analysis, laser scatter is

an enormous problem experimentally. Even with the spectral selectivity of a

monochromator, stray light rejection is typically on the order of 10s which still leaves

some 50,000 photoelectrons produced under the conditions stated above (the 0.1%

isotropic point source of laser scatter at the focus of the collection optics is not valid

except in the case of capillaries as will be discussed later). Furthermore, with

monochromators, the spectral bandpass of the emission process is also greatly








27
reduced which lowers the signal as well as the noise. Therefore, the use of a

monochromator in the SMD approach is inadequate to the task.

Spectral filters. The use of spectral filters is a common approach to reduce

laser scatter in IF experiments. Typical rejection of interference filters and long

pass spectral filters is 10i, which is poorer than the rejection obtained with most

monochromators, but the filters generally have a much greater optical throughput of

the fluorescence signal. In most cases, two or more filters are used together or in

conjunction with a monochromator. Again, the problem of this approach is that as

the filters reject more laser scatter, the signal is also reduced. This type of filtering

is explored in more detail in Chapter 3.

Polarization. Lasers are typically highly polarized sources of emission.

Through the use of polarized filters, the spectroscopist can take advantage of this

trait to reduce laser scatter because the large majority of scattered light retains its

polarization. Conversely, fluorescence emission is nonpolarized. Optimally,

polarized filters are capable of 109 rejection of light polarized in the same direction

as the filter and are still able to pass light of the opposite polarization. This means

that approximately half of the nonpolarized fluorescence should pass through the

filter.

In practice, polarized light rejection does not work as well as expected from

theory. Problems arise with the purity of the laser emission polarity and maintaining

polarity when scattered from complex materials. Experimentation with stray light

rejection by polarization is discussed in Chapter 3.








28

Spatial filtering. As mentioned above, the focused laser light does not scatter

equally in all directions. The pattern of laser scatter produced at the sample greatly

depends on focusing and the shape of the sample container (this aspect becomes very

important later in this dissertation when considering the capillary container used in

this project). Spatial filtering exploits the nonisotropic feature of the scatter through

the placement of a small slit or pinhole between the collection optics and the

detector. With careful positioning and focusing, the aperture can collect a large

percentage of the fluorescence while blocking much of the laser scatter arising from

the edges of the container. The implementation of this simple concept by Dovichi

et a"45 (along with the use of 3 spectral filters and sheath flow cuvette sample

container) was a great break-through in initial studies on single molecule detection.

The problem with the spatial filter in SMD, however, is that it can limit the

probe volume of the analysis. By collecting emission from only a portion of the

focused region, the sampling efficiency is reduced which circumvents one of the

conditions for SMD. It is also very difficult to position the optics and spatial filter

for optimum effect.

Rayleigh scatter. Another drawback with spatial filters is that they do nothing

to limit the amount of Rayleigh scattering reaching the detector. Unlike specular

scatter, which arises at interfaces between media of different refractive indices,

Rayleigh scatter occurs in the medium itself due to light interaction at the molecular

level. Furthermore, the fraction of light that is scattered specularly (measured as a

percentage) is very high compared to Rayleigh scatter which typically has a cross-








29
section on the order of 1028 cm2. This corresponds to the production of 1

photoelectron at an average of every 2 ms for methanol in the 140 nL Vp of the

conditions presented in Table 2. Although this is very small, at the light levels

involved in this project, every additional noise photoelectron could become

significant.

The metal vapor filter. A possible way to eliminate the serious problem of

laser scatter is through the use of a metal vapor filter (MVF). This device is central

to the success of this project and its theory is presented in Chapter 2 and

experimental results in Chapter 3. The theory is too extensive to be presented now,

but according to theory, the MVF is capable of essentially totally absorbing the laser

scatter (or the entire laser emission for that matter) provided the laser emission

bandwidth is narrower than the absorption band of the metal vapor. Moreover, the

MVF is completely specific to the laser wavelength and does nothing the hinder the

transmittance of nearly the entire fluorescence emission band of the fluorophore.

Based on the calculated 10 billion laser scatter photoelectrons, it must suffice to say

that the required rejection of the MVF must be on the order of 1010 or higher to

achieve SMD, which is theoretically possible as shown in Chapter 2.

Even though the MVF is capable of eliminating the problems with laser

specular scatter and Rayleigh scatter under conditions of this experiment, other

extrinsic sources of noise exist that would thwart SMD. Both Raman scatter and

background fluorescence from the solvent and optics occur at removed wavelengths

from the laser, which allow their passage through the MVF to the detector.










Raman Scatter


Unlike Rayleigh scattering, Raman scatter is an inelastic process that occurs

with even lower probability, typically with a cross-section of 103 cm2 per molecule.39

This hardly appears significant until one realizes that in a large volume, the number

of molecules in a material is so great that the value for N, becomes significant. As

LIF researchers have learned, noise due to Raman scatter from the solvent becomes

the limiting source of noise when the laser scatter is reduced. The way around this

problem is the reduce Vp which limits the number of atoms and molecules in the

container that scatter light. This is one of the reasons why all LIF approaches to

single molecule detection utilize a small Vp. In this project, with the parameters

listed in Table 2, the Vp of 140 pL containing methanol would give rise to about 1

photoelectron per every 100 counting intervals. This is an acceptable level for SMD,

but a more difficult to quantify amount of Raman scatter arises from the quartz of

the capillary. This and other aspects of Raman scatter are discussed in Chapter 2.


Background Fluorescence


As discussed in the theory, fluorescence has a rather high absorption cross-

section (oA = 3 x 1016 cm2 for the fluorophore presented in Table 2) and can be

measured very sensitively. It is a selective technique as well, but it is often unable

to specifically detect one fluorophore at low concentration in the presence of

another. Furthermore, no solvent is absolutely pure, and even the presence of an

ultratrace concentration of fluorescent interferents can negate the possibility of SMD.








31

There are two main methods to avoid this problem. In the first case, the

purest available solvent should be used, and secondly, the detection should be

designed to be as specific to the analyte as possible. The former method is not trivial

in even the best available solvents,4 and attempts at SMD are consigned to basically

working with standard solutions of the purest solvent. However in the application

of these techniques to a real sample, interfering species become a severe problem

in the production of background fluorescence. In a real analysis, it is not realistic to

assume that the analyte will be the only detectable fluorescent species in the sample.

Excitation at long wavelength. The best way to avoid background fluorescence

is to specifically analyze the analyte. Very few species fluoresce at far-red/near-

infrared wavelengths, and the only known dyes to do so at the laser excitation

wavelength of this project are given in Chapter 2. Therefore, the method of

detection for this project is very selective to the molecule of interest.

An additional benefit of using laser light at 794.76 nm, as opposed to the

more commonly used 325 nm emission from a HeCd laser or the 514.5 nm line from

an Ar+ laser, is that Rayleigh and Raman scattering processes are reduced by a

factor of 1/X4 (and specular scatter is also reduced to a large extent).39 This is a

substantial reduction in noise when pursuing SMD.









History of Single Molecule Detection


Now that the pertinent concepts of SMD and LIF have been introduced, the

previous accomplishments of LIF analysis nearing SMD can be reviewed without

having to define terms or explain the rationale behind the design of the experiments.

Single atom detection. In the past, there have been several instances of SAD

most notably through the research of Letokhov"'47 and Hurst.48 Their separate work

concerns the use of resonance ionization spectrometry to detect atoms in an ion trap

or in an atomic beam. As discussed earlier, these experiments meet the

requirements of SAD as stated by Alkemade,"'2 but do not conform to the practical

definition of SMD of this dissertation.

LIF of solids. In the case of molecules, single molecule detection has been

accomplished under somewhat artificial circumstances. Hirschfeld49 implemented LIF

with a microscope to detect a single protein molecule (MW 20,000) tagged with

80-100 fluorescein molecules on a solid substrate. Kirsch et aL,5 in a similar type of

procedure, were able to detect 8000 rhodamine 6G molecules. More recently,

Moerner5'53 has detected single pentacene molecules in a solid matrix at low

temperature using laser-excited fluorescence.

Keller's approach to SMD. In the analysis of flowing solutions, Keller's group

performed several experiments leading to the claim of single molecule detection

(although none of the reports satisfies even Alkemade's definition of SMD).2' 45'"

The origin of Keller's project at Los Alamos National Laboratory, which is designed

for the application of SMD for DNA sequencing,' began with research by Dovichi








33
et aL45 who bested the previous lowest LOD by nearly two orders of magnitude in

obtaining an LOD of 35,000 rhodamine 6G molecules. The subsequent experiments

at Los Alamos concerned refinements of the basic set-up developed by Dovichi.45

As shown in Figure 1, this basic set-up utilizes a tightly focused argon ion

laser to excite a highly fluorescent dye flowing in a sheath flow cuvette. A

microscope objective collects the fluorescence, spatial and spectral filtering reduces

scattered light, and a cooled photomultiplier tube coupled with photon counting

electronics measures the signal.

Table 3 is a list of the parameters and detection limits of published results

reported by the Los Alamos group. From the table, it is apparent that the

researchers have been slowly lowering the detection limit with difficulty. In Ref. 4

and Ref. 6, in which single molecule detection was claimed, the sampling efficiency

(e,) was very poor, water was not used as the solvent, and the sheath flow cuvette was

abandoned. Also, the traditional method of determining the LOD was not used;

instead autocorrelation analysis of a single sample was performed. In Ref. 4, the

research team used a laser with 70 ps pulses and a microchannel plate detector with

sophisticated signal collection to help discriminate the fluorescence from the scatter.

The data presented in Ref. 4 do appear to be single molecule events, but the signal

to noise ratios are not reported so a statistical treatment to determine LOD cannot

be performed. The researchers base their claim of Ed = 70% (of those molecules

passing through the center of the Vp) on computer simulations that appear similar

to the actual data. Weighted quadratic sum plots of the data presented showed























1/2 Wave
Plate C
Polarizing
Prism


Spectral
Filter


Mirror


Figure 1. The instrumental approach to single molecule detection used by
Keller's group at Los Alamos National Laboratory. Redrawn from
references 2, 3, and 5.









Table 3.


Comparison of the reported parameters and results of the laser-
induced fluorescence experiments of Keller's group at Los Alamos
National Laboratory.


Parameter Ref. 2 Ref. 3 Ref. 4 Ref. 5 Ref.6
(1984) (1987) (1990) (1991) (1991)
Analyte R6G R6G R6G R6G R6G
Solvent H20 H20 H20/C2HsOH H20 C2FHOH
Cell SFC SFC flowcell SFC flowcell
Laser, Ar+ Ar+ Nd:YAG Ar+ Ar+
XL (nm) 514.5' 514.5 532b 514.5 514.5
SL (cm2) 3.8x10 1.1x106 4.4x10-7 2.1x10- 2.1xl05
EL (kW/cm2) 130 700 6.8 40 23.4
Stream Size 30 42 4000 44 250
(/Mm)
F/4'r 0.06 0.045 --- --- -
E, 0.6 0.06 1.9x10-7 0.1 0.05
Vp (pL) 11 0.6 0.44 11 10.7
Flow Rate 25 0.012 5760 1.0 0.18
(AL/min)
Flow Velocity 60 14.2 0.075 5.4 4.85
(cm/s)
t, (ms) 0.037 0.085 10 1.8 2
rc (s) 1 1 0.004 1 0.0004
LOD (M) 1.3x10-13 2.2x10-13 -- 9x10-1 ---
LOD (#/rc) 33,000 1,200 1" 33 "1d
R6G = Rhodamine 6G; SFC = Sheath Flow Cuvette; e, = Sampling Efficiency;
7, = Time Constant of Measurement;
LOD (#/7,) = number of molecules passing through Vp during 7, at the LOD (M).
'Pulsed at 10 kHz, 50% duty cycle; bPulsed at 82 MHz, 0.57% duty cycle.
cTime discrimination method able to detect passage of single molecules with reported
70% detection efficiency (Ed).
dReported detection limit based on autocorrelation analysis; Ed not given.








36
photoelectron bursts arising from passage of single molecules which were not

presented in any of the other references. The same sort of basis was used in the

claim to single molecule detection made in Ref. 6, but in this case, passage of

individual molecules in the V, was not noticeable. The claim to single molecule

detection was based on the use of a 20 point sliding sum distribution which agreed

with theoretical results. No efficiency of detection can be calculated from such a

determination because the single events could not be counted.

In references 3 through 6, the spatial filter viewed only a small portion (5 Jtm)

of the laser focus volume to reduce the background sources of noise discussed

earlier. This technique limited the solid angle of collection of the microscope

objective which is why p0/4r is not reported in those cases (except Ref. 2). In Ref.

5, which reported an LOD of 9 x 10-' M based on conventional methods to

determine the LOD, the 4bI of the PMT was 212,764 +/- 461 counts/s. This high

noise level is a result of the large background sources of noise discussed earlier. At

this noise level, SMD is not possible with the approach presented in this thesis. I

a separate study utilizing B-phycoerythrin as the analyte, Nguyen et aL5 in Keller's

group claimed the first instance of single molecule detection in solution. It is

noteworthy that B-phycoerythrin is a very large (MW 250,000 g/mol) protein

possessing the equivalent fluorescence of 25 rhodamine 6G molecules."

Mathies' research. At the University of California at Berkeley, the research

group of Mathies contested this initial claim to SMD made by Keller's group; they

repeated the LIF study with B-phycoerythrin.ss Through a more rigorous statistical








37

approach, their research effort showed that the previous work did not obtain SMD,

and their results demonstrated the detection of 15% of the passing single molecules.55

Based on the definition of SMD presented in this thesis, B-phycoerythrin, with

a MW 250,000 g/mol, does not qualify as a "single molecule," and the sampling

efficiency of the experiment is much less than unity. Keller's group2' and other

researchers43 are aware of these shortcomings and have worked to lower the LOD

for smaller fluorophores. In their proposal,' Keller's group' mentioned that sampling

efficiency must be increased to perform the desired DNA sequencing application, but

the reduction of noise through the use of the spatial filter was integral to the

detection limits they have achieved.

Also, the sheath flow cuvette as the sample container is inadequate to achieve

SMD with e, = Ed = 1. Further addressed in Chapter 2, sampling efficiency and

detection efficiency are diametrically opposed relationships with the sheath flow

cuvette. To increase sampling efficiency, the laser beam focus must be increased, but

to increase detection efficiency, the focus must be kept small. Furthermore, with

small probe volumes, the sample flow rate must be lowered to maintain the residence

time, but with a sheath flow cuvette, the flow stream becomes broader with

decreasing flow rate thus requiring a larger beam focus. This has been a difficult

problem with the sheath flow cuvette, and researchers using this device have resorted

to finding an optimum trade-off." In this respect, it is unlikely that research with a

sheath flow cuvette will ever achieve true SMD. In fact, the time discrimination

approach of the Los Alamos group with the frequency doubled Nd:YAG laser








38
appears more promising for true SMD than the approach exhibited in Figure 1.

Indeed, two of the recent papers by Keller do not use the sheath flow cuvette.4'6

Winefordner's approach. To avoid the problem of this trade-off with a sheath

flow cuvette and its high cost, Winefordner's group has decided to probe the entire

sample stream in their attempts at SMD. Furthermore, the diode laser was chosen

as the excitation source for purposes of greater analyte selectivity, lower noise from

Raman scatter, lower cost, simplicity, and the many other advantages of diode lasers

to be discussed in Chapter 2. In two separate studies (with different lasers), LODs

of 40,000 and 3,000 molecules flowing in the probe volume of the near-infrared dye,

IR 140, have been measured in a liquid jet (flow stream emanating from a

capillary).5657

Despite the many advantages of working with diode lasers and the low

detection limits attained with their use, LIF with diode lasers was found to lack the

sensitivity to achieve SMD in a liquid jet. Furthermore, the liquid jet only operates

at high flow rates which gives unsuitable residence times for SMD. This flow

condition is one of the reasons why the method described in this thesis employs a

capillary for sample containment.

Ramsey's approach. Ramsey's group at Oak Ridge National Laboratory has

designed the first instrumental set-up capable of truly realizing SMD in solution as

defined in this dissertation.4 The design of this approach appears in Figure 2. The

key to the experiment is the use of the electrodynamic trap to contain droplets of the

sample instead of a flowing stream. In this way, the entire sample is probed (one

























I 0
2E


04.
E =
e=
S f
"^


00

coiL.
a b..


0

z
4)
*O



C

0
o





4)
t3


I













rn
(0

4)






O
U,

a









4)
a,
'4-



o
4)





'o



I9

5 o
60
s ,D


*3


ab
I..
I-








40
drop at a time) and the interaction time can be on the order of days if necessary.

The noise rejection is not as high as in other experiments and the dye tends to

photodecompose before SMD can be quantitatively attained. The LOD currently

stands at an average of 25 molecules of rhodamine 6G per droplet.

The approach to SMD of this dissertation. It is hoped that the SMD approach

described in Chapter 2 will be the first method to truly achieve SMD in a solution.

Like Ramsey's approach, it is designed to sample the entire solution, and has been

shown to theoretically perform the desired assignment. More of the theory, design

and development considerations, and some results will be presented in the following

chapter.













CHAPTER 2
CHOICE OF INSTRUMENTATION AND
OPTIMIZATION OF PARAMETERS


The Metal Vapor Filter


Because of its role to remove the laser specular and Rayleigh scatter from the

collected light, the metal vapor filter (MVF) is the most important component of the

experimental set-up to attain single molecule detection (SMD). With few exceptions,

laser scatter is the limiting source of noise in ultratrace analyses using laser-induced

fluorescence (LIF). Therefore, satisfactory performance of the metal vapor filter is

crucial to the success of the experiment.


Theory of the Metal Vapor Filter


The MVF typically is a 2-3 inch long glass cell that contains a surplus amount

of a metal in its elemental form enclosed in a nitrogen environment. Figure 3 is a

simple drawing of the MVF. Upon gentle to moderate heating of the cell, a portion

of the solid or liquid metal enters the vapor state depending on temperature,

pressure, and thermodynamic properties of the element. The absorption properties

of the MVF depend on the number density, or concentration, of the metal in the

vapor state, the length of the cell, and the pressure in the cell.











Pyrex


Cylinder


Metal Element


Figure 3. The metal vapor filter.

Absorption. Increases in the metal vapor number density and absorption path

length lead to directly proportional increases in absorbance assuming the linewidth

of the source is more narrow than the absorption bandwidth. This relationship is

exhibited by Beer's Law,39

A = 0.434Ao^n, (2-1)

where A is the absorbance dimensionlesss), A^ is the absorption cross-section (cm2),

f is the absorption path length (cm), and n is the number density of the absorber

(cm"). In the case of the MVF, the equation to determine the absorbance of an

atomic transition can be calculated from,58

e If.S,vnil
A = 0.434e (2-2)
4E,m,c


where e is the charge of an electron (1.6 x 1019 C2), f is the oscillator strength for

the electronic transition from level i to level j (unitless empirical value for each








43
transition), S,v is the Voigt shape function of the absorption bandwidth (Hz-'), nA is

the number density of the atom in level i (m-3), e, is the permittivity of free space

(8.854 x 10-2 Ns'/C2), m, is the mass of an electron (9.11 x 10"3 kg), and c is the

velocity of light (2.998 x 108 m/s).


Choice of Metal for the Filter


For the MVF to be useful in practice, the electronic absorption transition

must start from a ground state due to the small fraction of the element existing in

an excited state as described by the Boltzmann distribution. Therefore, level i in

Equation 2-2 must be the ground state, and nA must be the number density in the

ground state. Another important consideration for the practical use of a MVF is the

thermodynamics of the chosen element. In order to attain a large enough number

density, n, to satisfactorily absorb the large light intensities associated with a laser,

the metal must have a low melting point and high vapor pressure. Table 4 is a list

of elements that could be of practical use in a MVF along with their melting points,

ground state transition wavelengths, and approximate o^ of the overall transition. Of

these elements, rubidium is an excellent choice for use in a metal vapor filter due to

its low melting point and strong absorption lines at 780.023 nm and 794.760 nm. For

these reasons and more (based on characteristics of lasers and fluorescence dyes

available, which will be discussed later in this chapter), rubidium has been chosen as

the element to include in the MVF for use in the SMD system.










Thermodynamic characteristics of several elements for possible use in
the metal vapor filter.


Element
Al
Ba
Ca

Cd
Cs


Hg
In
K


Li

Mg
Na

Pb
Rb


Sn
Sr
TI
Zn


M.P. ("C)'
660
714
838

321
29


-38
156
64


181

650
98

327
39


232
768
303
420


X (nm)a
308.216
553.548
657.278
422.673
228.802
852.110
894.350
455.530
253.652
410.476
766.491
769.898
404.410
670.784
323.260
285.213
588.995
330.232
283.306
780.023
794.760
420.180
286.333
460.733
377.572
307.590
213.856


gA (cm)b
9 x 10-13
2.5 x 10-"
8 x 10-1
2.1 x 10-"
1.3 x 10"1
9.5 x 10"1
4 x 10-"
8 x 1012
8.7 x 10-12
4 x 10-12
2 x 10-"
1 x 1011
3 x 10-13
6 x 1012
3.5 x 10-13
1.3 x 10-1
1 x 10-11
7 x 10-4
4 x 10-12
1.3 x 10-"
6 x 10-"
1.1 x 10-1
1.3 x 10-12
5 x 10-12
5 x 10-12
2.3 x 10-14
1.2 x 10-"


Table 4.


"Values from Lange's Handbook of Chemistry, 13th Ed., JA. Dean, Ed., McGraw-Hill,
New York, 1985.
bValues calculated by Ramee Indralingam from Equation 2-2.









Calculation of Spectral Linewidths and Absorbances for Rb


The two most important features for use of the MVF are the absorption

coefficients (or absorption cross sections) and spectral profiles for the atomic

transitions at 780.023 and 794.760 nm. These properties give an idea of the required

specifications of the laser to be used in conjunction with the MVF in SMD. In order

to quantify these two parameters of interest, some of the theory of line broadening

will be given.

Atomic spectral profiles. The emission or absorption spectral profile of a

single, stationary atom contained in a vacuum absent of electric and magnetic fields

would have a spectral linewidth determined by the lifetime of the electronic

transition as stated by the Heisenberg uncertainty principle." Typically, atomic

transition lifetimes are on the order of 10' s which corresponds to a natural, or

fundamental, linewidth of 10 MHz or 0.021 pm at 800 nm. However, in a real system

there are several factors, such as motion of the atom, atomic collisions, and presence

of electric fields, which act to broaden the spectral profile for a given transition.

Furthermore, fine and hyperfine structure exist due to quantum splitting of electronic

transitions and existence of isotopes for a particular element. These components

each have an individual, quantifiable effect, and after they have been factored

together, the overall peak shape, termed the Voigt spectral profile, can be calculated.

Rubidium has two main isotopes, Rb8 and Rb87, which exist naturally in the

ratio 2.59:1 for RbS:Rb7. The fine structure for transitions of these isotopes and

their fi values are given in Table 5 at 780.023 and 794.760 nm.59









5. Physical parameters of the rubidium hyperfine structure required for
the calculation of the Voigt spectral profile of the metal vapor filter.

For the Absorption Band at 780.023 nm (vo = 12.820 cm-'):


for Rb7 for Rb"8

V-p0 fij V-VQ f
(cm-l) (cm-')

0.1323 0.0417 0.0553 0.0833
0.1347 0.104 0.0563 0.108
0.1401 0.104 0.0585 0.0864
-0.0933 0.0208 -0.0451 0.0309
-0.879 0.104 -0.0430 0.108
-0.790 0.292 -0.0389 0.250


For the Absorption Band at 794.760 nm (vp = 12.582 cm^):

for Rb87 for Rb85

P-VP fij V-Po f4
(cm') (cm-')

0.1255 0.0208 0.0521 0.0309
0.1527 0.104 0.0642 0.108
-0.1025 0.104 -0.0493 0.108
-0.0753 0.104 -0.0372 0.0864


Each of these lines is broadened by the same extent as described below and

their normalized absorbances are added together at individual wavelengths to form


the Voigt profile.


Table








47
Lorentzian broadening. There are two types of broadening taking place in the

MVF which factor into the Voigt profile, namely, Doppler broadening and collisional

(or pressure) broadening. The different types of collisional broadening can be

grouped together to produce the total Lorentzian profile. In the case of the MVF,

it is realistic to assume that all collisions of the atoms in the vapor state occur with

diatomic nitrogen (N2) and are adiabatic in nature (the atom remains in the same

electronic state during the collision). Thus, the Lorentzian linewidth, AvL (Hz), can

be calculated from the equation to determine the profile due to adiabatic collisions,

which is"

1/2
2a,An 2RT (2-3)
2L-j (2-3)


where R is the gas constant (8.314 J/K mol), T is the temperature (K), u is reduced

mass of Rb and N2 (kg), a, is the collisional cross-section of Rb (m2), and n, is the

number density of nitrogen in the metal vapor cell (mn3; in the case of nitrogen, n,

= 9.74 x 10u P/T, with pressure, P, in Torr and temperature, T, in K). For Rb in

a nitrogen environment, a, has been measured to be 2.49 x 10"9 m2 at 780.023 nm

and 2.30 x 10-19 m2 at 794.76 nm.58

Doppler broadening. Doppler broadening, APD (Hz), a result of the atoms

moving at different velocities upon absorption of light, can be determined from39










2p, 2(ln2)RT] (2-4)
c M

where R is the gas constant (8.314 J/K), M is the formal weight of Rb (kg/mol), and

vm is the center frequency of the overall transition (Hz).

The a-parameter. To determine the proportion of each broadening effect in

the overall linewidth, the "a-parameter" is used, where, a = 0.83(AvL/APD). In the

case of the Rb metal vapor cell at pressures > 75 Torr, the effect of collisional

broadening is greater than that of Doppler broadening, which is expressed by a > 2.

The Voigt integral value, b(a,0), at the line centers of both Rb resonance lines can

be calculated (within 10%) to be 6(a,0) m 0.56/a (for a > 2).39

Voigt linewidth. The multiplication of 5(a,0) by the Doppler shape function,

S,D (s), results in the Voigt shape function, S,v (s). For Doppler broadening, SD at

frequency v (Hz) can be determined from the relationship,39


2 1/2
S,D = 2 l e -4(n2)(,-,^/(A, (2-5)


and the Voigt linewidth, Avv (Hz), is found from



APV [.2_+'2 ] 1/2 (2-6)
1/2



All linewidths denoted by the subscript v correspond to the full width of the

peak at half of the maximum intensity. To convert any of these linewidths from

frequency, Av (Hz), to wavelength, AX (nm), the value is multiplied by wavelength








49
squared and divided by the velocity of light, c (3 x 107 nm/s). For example, AXv at

800 nm is found from the relationship, Avv(800 nm)2/c.39

Rb number density. At this point, all of the parameters, except nA and e, have

been defined that are necessary to determine the absorbance, as described by

Equation 2-2, and the Voigt spectral shape function, as described by Equation 2-5.

For the metal vapor cell I is fixed and n, is dependent on temperature and pressure.

For the evaporation of liquid Rb, nA follows the expression,"

log(n.) = -A/T (B+ 1)log(T) + C + DT +18.985, (2-7)

where A = 4529.6, B = 2.991, C = 15.8825, and D = 0.00059 with T in K. There

have been many different measurements of the values of these constants by physicists

with variability in nA as great as 25%, but the above values were chosen from

reference 60 because of their close agreement with other referenced values61 and

high precision.

Voigt spectral profile. The overall absorption line shape of Rb vapor was

calculated with the help of a computer spreadsheet and a computer program (written

in basic language by Michael Wensing). The program summed the absorbances of

the individual components of the profile versus wavelength based on values

calculated from the above equations in spreadsheet format. Plots of the Voigt

profiles at different temperatures and pressures were then generated (with I of 4.4

cm). These plots are exhibited in Figures 4 and 5. As the figures show, increasing

cell temperature has a large effect on the absorbance, which essentially mirrors the

increasing Rb vapor number density, and increasing pressure increases linewidth at








50

the expense of absorbance. The presence of more than one peak at low pressure

exhibits the fine structure of the overall transition.

For this work, two different Rb metal vapor cells were available for use in the

project. Both cells contained 500 mg of 99.99% pure rubidium metal in nitrogen

(this is more than enough to produce a large vapor number density without

expending all of the condensed metal). One cell (to be referred as cell #1),

manufactured by Rudy Strohschein in the University of Florida glass shop, has an

absorption path length of 4.4 cm and a pressure of 200 Torr at room temperature.

The other cell (cell #2, manufactured by Opthos, Rockville, MD) has a 4.7 cm path

length and a room temperature pressure of 500 Torr. These parameters were

entered into the computer programs to calculate the Voigt profiles for these cells at

100C for both the 780.023 nm and 794.760 nm lines. These profiles appear in

Figures 6 and 7. The Voigt linewidths are 21 pm (at 780 nm) and 20 pm (at 794.76

nm) for cell #1, and 36 pm (at 780 nm) and 36 pm (at 794.76 nm) for cell #2. Cell

#1 is capable of a larger absorbance at constant temperature than cell #2, which

makes cell #1 preferential for use in the experiment provided that the laser emission

linewidth is less than 21 pm.

Tye Barber41 experimentally verified the Voigt profiles generated from

calculations using a single-mode diode laser with cell #1. He found the absorption

bandwidth to be 21 pm at 780 nm which is in excellent agreement with the calculated

bandwidth. The shape and intensity of the absorption peak also closely agreed with

the computer generated plot at the same temperature.41








51



o -
SLO


o



"- a






o 0.
o 4
/ I '-"


0 0









o -
o



o T
\o










I I I I I I LO
o0 0 000 O' -
eo-- oq \-
CUJ 0 cO (0 t1 C J












o






,,1..-
4)






oo
F--



0 0
0 0

o \
0-- 8'
E


C,

Sco

0 ou -q-o, \ r





ca



0 o







O o noL m o m o LO O1
It I- Co C, C'j CJ '-

eoueqjosqv 4
IllIIII C













-Lo


0
4 )


_4)






a) 0
o









r-4
oO







a,s
-- ^----------oa a- O


0)
C











-S
C)
-a l

\\ i'





O
CD LO 9t CO C\J C0
eoueq N




*E


7-







54


o


o*
SLO


o 4
z 0



O
Lo 0 o


CM s

o .
0M -0





a)


L0
o 0
eoueq

O









o 0
I I I \ I ,- ,


eoueqJosqv a
\\ o O
\ \ "" l
\ \ ^il










The Laser


Instrumentation for fluorometric analysis. To perform molecular fluorescence

analysis, five basic components are necessary: 1) a source of light, 2) a fluorescent

sample, 3) a container for the sample, 4) a spectral filter, and 5) a detector. In the

SMD project, the MVF, which is able to remove the source light while still allowing

the passage of the fluorescence to the detector, serves as the fourth component listed.

The excitation source is the subject of this section, and each of the remaining

components will be discussed individually in the following sections.

Light sources. For selective analysis, the excitation source should have a

narrow emission band, and for improved sensitivity, the source should have high

emission intensity. Arc lamp sources are capable of very large radiances, but in

order to obtain narrow bandwidth, the light must be passed through a

monochromator which filters the excitation source into less intense bands. With the

invention of the laser, a source became available that was capable of high power in

a narrow spectral bandwidth. However, a problem with the laser as an excitation

source is its lack of tunability. Only those analytes with an excitation band at the

laser emission wavelength can be determined by LIF. Of course, there are many

different lasers with many different laser lines, but the cost and practicality of using

so many different sources makes LIF an analytical technique for unique

circumstances.









Criteria of the Laser for Single Molecule Detection


The choice of laser to be used in the detection of single molecules must be

made based on several criteria. These criteria include:

(1) the laser must be capable of sufficient irradiance (W/cm2) at the excitation

wavelength of a given fluorophore to attain SMD;

(2) the laser emission wavelength must fall at or be able to be tuned to the

absorption wavelength of the metal vapor cell and remain at this wavelength

over time;

(3) the emission profile of the laser must be Gaussian-shaped with a linewidth

less than the absorption bandwidth of the metal vapor;

(4) the source must be continuous or have a pulse rate of at least m 10 kHz in

order to efficiently probe a flowing sample.

Other desirable traits of the source include low cost, simple maintenance, and easy

operation.

Choosing the laser. Most lasers meet criterion (1) if the beam can be focused

to a very small area. Because no non-tunable lasers happen to emit light at an

absorption band of an element for practical use in a MVF, criterion (2) establishes

that a tunable laser must be used. Only dye lasers, diode lasers, Raman-shifted

lasers, and the Ti:sapphire laser are tunable. Requirement (3) negates Raman-

shifted lasers and all pulsed dye lasers except the copper vapor-dye laser system.

Criterion (4) precludes the use of a pulsed laser unless it is capable of a repetition

rate greater than = 10 kHz. The upper limit of the repetition rate of the Cu vapor-








57
dye laser system is 10 kHz, thus single-mode diode lasers and Ti:sapphire laser have

the best characteristics for SMD.

Diode lasers. Diode lasers are solid state electronic emitters of

electromagnetic radiation much like light-emitting diodes except the diode laser

possesses a cavity to induce lasing.2 The lasing material is usually gallium doped

arsenide which has an energy gap that corresponds to wavelengths in the red to near-

infrared wavelengths. Diode lasers have several advantages over conventional lasers

that make them generally accepted among spectroscopists as the "laser of the future."

The favorable characteristics62 of diode lasers include: (1) inexpensive, (2) easy to

operate, (3) maintenance free, (4) long-lived, (5) robust, (6) small, (7) tunable, (8)

powerful, (9) efficient, (10) versatile, (11) high power stability, and (12) narrow

spectral linewidth.

Despite these advantages, diode lasers will remain the "laser of the future"

unless the following problems can be corrected. Firstly, diode lasers are tunable, but

not continuously tunable allowing coverage of all wavelengths in the tuning range.

Diode lasers commonly exhibit mode hopping which is the tendency of the laser to

jump instantaneously from one wavelength to another and exhibit lack of tunability

in the region between these modes. Secondly, since diode lasers have a small laser

cavity, the laser beam has a large beam divergence which makes focusing difficult.

Lastly, diode lasers are available over a limited wavelength range of 650-1300 nm

with each laser tunable over a 20-30 nm range. If the laser had a low divergence, it

would not be difficult to frequency double the emission to wavelengths half of the








58

fundamental output utilizing nonlinear optical properties of certain crystals. This

would enable diode lasers to be capable of emission from 325-1300 nm, but the poor

doubling efficiency currently obtained results in laser powers too low for general

applications.6

However, it is fair to say that the current intensive research in the area of

diode lasers in the brief time since their invention has resulted in more powerful

lasers, wider range of lasing wavelengths, more efficient single-mode operation, and

lower cost which will soon result in a laser useful for SMD. Already, diode lasers

can be used in conjunction with a MVF.41 Special electronic control of the diode

laser current with active feedback virtually eliminates mode hopping, but the low

power and focusing difficulties of these lasers does not yet meet all of the criteria

listed earlier. Undoubtedly, diode laser excitation would be the preferential

approach for SMD due to the many advantages of diode lasers, but until the above

problems are solved, another laser must be used.


The Titanium:Sapphire Laser


Of all possible sources currently available, only the titanium:sapphire

(Ti:A1203) laser meets all of the outlined criteria. Table 6 lists the specifications of

the manufacturer of the Ti:sapphire laser purchased for use in this project. As the

table shows, the laser is capable of two lasing arrangements, the standing-wave and

ring cavity configurations. The specifications for both configurations are the same

except that the spectral linewidth is narrower in the ring configuration.








59
Figure 8 is a drawing of the design of the Ti:sapphire laser. The diagram

shows the laser in the ring cavity arrangement with the beam passing through the

optical diode. In the standing-wave configuration, the two flat mirrors are tilted to

return the beam to the curved mirrors on either side of the crystal instead of through

the optical diode.


Table 6. Specifications of the titanium:sapphire laser used in the single
molecule detection project.

Parameter Specified Value
Maximum Power 750 mW
(at 800 nm) (10% of pump power up to 7.5 W)
Tuning Range 700-1000 nm
Spectral Linewidth <2 GHz standing wave
(at 5 W Pump Power) <40 MHz ring configuration
Spectral Profile Gaussian
Spatial Profile TEMoo
Polarization Horizontal
Beam Size 1 mm at exit


The Ti:sapphire laser is a passive device with no moving or electronic parts.

An important aspect of the Ti:sapphire laser is that it requires the services of a pump

laser to initiate the lasing action. The source of the lasing is a solid state AO13

crystal doped with titanium. The crystal has broad excitation and emission bands

with maxima = 530 nm and 800 nm, respectively." Due to peak absorption in the

green part of the spectrum, argon ion, Nd-YAG, and copper vapor lasers are used

to pump the Ti:sapphire laser. The Ar+ laser is used most often to take advantage









60


















93
U
t4
.0'




4)
1

.9
4)
4)





E4
0
4)


4)
O











0


I
60



li





00
S

*M








61

of its continuous laser action, but pulsed operation can be applicable in many

situations. The output power of the Ti:sapphire is proportional to the power of the

pump laser and, according to the manufacturer specifications, is capable of 10%

efficiency of conversion, but in the SMD experiments, nearly 20% efficiency was

attained with careful alignment.

Comparison of the Ti:sapphire laser with dye lasers. The Ti:sapphire laser

was invented in the 1980s and is poised to replace the dye laser for many

applications. In the early 1970s, dye lasers were hailed as the solution to the lack of

laser tunability. There are several problems with dye lasers, however, concerning

both scientific and practical aspects in their use. Each dye has a limited tuning range

based on concentration, solvent, pH, and pump laser. Power is strongly dependent

on wavelength and is greatly changed when a dye laser is scanned; efficiency of

conversion is often less than 1% for dye lasers. For these reasons, dye lasers are

seldom used to scan over a wavelength range, and the more common application is

to tune a dye laser to a fixed wavelength. In practice, dye lasers are noted for the

problems: dye decomposition, solvent flammability, chemical toxicity, waste disposal,

chemical spills, and extended down time (often due to solvent pumps). For these

reasons, spectroscopists have been anxious for a laser to replace the dye laser.

The Ti:sapphire laser has many advantages over the dye laser, but is not the

panacea for the laser spectroscopist. Since the Ti:sapphire is a passive, solid state

laser, it has none of the practical problems associated with using laser dyes. No

moving parts translates to no down time for mechanical reasons, and no dyes means








62

no messes, chemical hazards, or waste. Also, the Ti:sapphire laser is capable of

continuous tuning with moderate change in power over a continuous range of 120 nm

(fundamental) or 60 nm (frequency doubled) before requiring a change of optics.

Few individual dyes can match this feat; however, the overall wavelength coverage

of 350-500 nm and 700-1000 nm is still inferior to a network of dyes that allows for

coverage of the spectrum from the ultraviolet to the near-infrared. Therefore, the

Ti:sapphire laser is not able to replace the dye laser in all applications.


The Ti:Sapphire Laser/Rb Metal Vapor Filter Combination


The stated linewidth specifications of < 2 GHz in the Ti: sapphire standing-

wave configuration and < 40 MHz in the ring cavity configuration corresponds to <

4.1 pm and < 0.081 pm, respectively, at the 780 nm Rb transition. For the 794.76

nm transition, these linewidths are < 4.2 pm and < 0.084 pm, respectively. Due to

the Gaussian nature of the spectral profile for the laser as opposed to the Lorentzian

wings of the Voigt profile, the laser beam for either laser configuration should be

virtually totally absorbed by the Rb metal vapor cell at temperatures > 100*C. This

aspect is shown in Figure 9 in which the emission profiles of the Ti:sapphire in both

configurations are superimposed onto the Voigt profile of cell #1 at 794.76 nm. As

shown, the specified laser linewidths are much narrower than the profile of the

absorption line. This condition is especially notable in the ring cavity profile where

the laser line is a thin line on the same scale as the other profiles.













A4!SuGIuI Jesei


eoueqJosqv


CI-



08



1-

ifT









8-









4')
40
















C) e-


1-
E



C,
T

-0 r-

C
-0)


T- 0








64

Absorbance is the log of the inverse of transmittance, T, (A = log(l/T))

which means that the absorbance of 10 for the Rb metal vapor filter at 100*C

corresponds to a transmittance of 10-10, or 99.99999999% of the light is absorbed.

Therefore, with a metal vapor absorbance of 10, the calculated maximum percentage

of light scattered by the capillary is 5.9% (if it is an isotropic point source at the

focus of the microscope objective) that would still allow for an LOD = 1 molecule

based on the parameters listed in Table 2 of Chapter 1. In other words, the laser

scatter collected in the 1.5 sr solid angle of collection (0,/47r) must be limited to less

than 0.7% (5.9% x 11.9%) to achieve a background count rate of 56 counts per 2 ms

increment (maximum theoretical AbM possible in order to attain SMD for Xd = 81

counts. These calculations are based on the discussion given in Chapter 1.

It should be pointed out that the Ti:sapphire laser/Rb MVF combination has

been used previously to perform Raman spectroscopy.6', As in LIF, the limiting

source of noise in Raman spectroscopy is typically laser specular scatter. With a

similar laser/MVF set up to be used here, Raman spectra were obtained that showed

no evidence of the laser despite viewing at the laser wavelength."'" This is a major

accomplishment in Raman spectroscopic analysis because it is now possible to

measure Raman bands very close to the excitation wavelength. Furthermore, the

conditions for optimum performance of Raman spectroscopy are very similar to those

of the SMD; since the Ti:sapphire laser and Rb MVF combination was successful in

that application, there was little reason to believe it would not be successful in

detecting single molecules by LIF.








65
Ti:sapphire laser configuration. Because the linewidth in either the standing-

wave configuration or ring cavity configuration is much narrower than the Voigt

profile of the Rb vapor cell, it does not matter which laser configuration is used for

single molecule detection. Both configurations were tested for use in the project with

no significant difference in results.


The Sample


After the excitation source, the sample constitutes the second of the five basic

components of a fluorometer. In the case presented so far, the analyte must absorb

strongly at 780 or 795 nm (Rb ground state electronic transitions) and fluoresce with

high quantum efficiency at removed wavelengths. Furthermore, as discussed in

Chapter 1, the analyte should not be a polymer or other chemical species with a

molecular weight greater than = 1000 g/mol. Another consideration is the solubilty

of the molecular species because this analysis is to be attempted in solution phase.


Choice of Analyte


Polymethine dyes. Very few chemicals fluoresce beyond 700 nm. In fact, the

only known molecules to fluoresce strongly in the far red part of the spectrum are

the cyanine dyes known as polymethines.67 Table 7 is a list of polymethine dyes that

includes their common names, chemical formulas, excitation and emission maxima,

and fluorescence quantum yields of those available in the literature. As expected for

fluorescent molecules, these dyes have highly conjugated, semi-symmetrical structures











Fluorescent dyes for possible use in the single molecule detection project.


Dye Molecular x X, C eA" Yp
Formula (nm) (nm) (MN'cm'1)

Rhodamine 800 C26H26N30 Cl 682" 700' 89,500 0.39'
Methylene Blue CiXH{N3S Cl 668' 683" 66,60 --
Nile Blue C2oH2N30 Cl 640" 672' 77,500"
Oxazine 750 C24HN3O Cl 673 691 82,500 ---
IR 125 C43H47N2S, Na 780 806 150,000 0.13c
IR 132 Cs2H48N304S2 C104 810 846 210,000
IR 140 C38H34C12N3S2 C104 800 833 180,000 1.0d
IR 144 CoH8sN408S2 NEt3 698b 708b 127,000 ---
DTTC CH25HN2S2 I 746b 777b --- 0.38c
DTDC C2H23N2S2 I 647 668b --- 0.73
DOTC CsHN202 678b 703b --- 0.63'
HITC C28H33N2 I 736b 764b 240,000 0.28c
HDITC C36H37N, C104 771b 805b
DDTC C32HN2 765" 855 -- 0.16'
DQDC CH27N2 765' 835C -- .001'
DQTC C29H2N2 825' 865' -- .035c

'Values in water from T. Imasaka, A. Tsukamoto and N. Ishibashi, Anal Chem., 61. 2285
(1989).
bValues in methanol from D. Andrews-Wilberforce and G. Patonay, AppL Spectrosc., 43
1450 (1989).
"Values in dimethylsulfoxide from R.C. Benson and H.A. Kues, J. Chem. Eng. Data, 22, 379
(1977).
dValue in ethanol from D.J.S. Birch, G. Hungerfold, B. Nadolsi, R.E. Imhof and A.D.
Dutch, J. Phys. E: Sci Instrum., 21, 857 (1988).
All other values determined experimentally in methanol.


Table 7:








67

containing several aromatic rings with the length of the structures correlating to

excitation and emission wavelengths. In fact, Benson and Kues6 have worked out

an empirical relationship between the structure of polymethine dyes and their

fluorescent characteristics.

Of the dyes listed in Table 7, only three, IR 125, IR 140, and IR 132, are

potentially useful for single molecule detection with excitation at 780 or 794.76 nm.

Figure 10 gives the structures of these polymethine dyes. It is rather fortuitous that

X," of IR 125 falls at the 780 nm Rb line, and that IR 140 has k" at 800 which

is very near the 794.76 nm transition. Meanwhile, IR 132 is the least favorable of

these dyes in this respect with X,", of 810 nnm.

Another important parameter for SMD is the absorption cross-section, Ao

(cm2), of the analyte which is proportional to the molar absorptivity, cA, of the dye

in bulk solution as shown in Chapter 1. In a simple experiment using Hewlett-

Packard 8450A and Varian 634 (for wavelengths greater than 800 nm)

spectrophotometers, the absorption spectra of known concentrations of the dyes in

methanol were measured. The wavelengths of maximum absorption corresponded

to the fluorescence excitation maxima for the dyes, and the measured EAm for IR

125, IR 140, and IR 132 were 1.5 x 10', 2.1 x 105, and 1.5 x 10' M-Ncm', respectively.

These figures are in agreement with values reported in the literature.69

The emission spectra of the dyes in methanol solution are shown in Figure 11

with excitation at 780 and 795 nm. For this experiment, the Ti:sapphire laser was

used as the excitation source of the flowing dye solutions contained in a 1 cm path











N j CH=CH)3 -CH Jg
(CH2)4 IR 125
80- 803 Na


CIJ IH=CH -r N O H-
j(X--^r--O
CI H5 IR 140 i
C2 C2H5

8 N 8
Q/[OH=-CH j CH-H hh I
N 0 0
o4 (CH)3OC IR 1 32 o
4 (CH2 OCCH (CH2)30CCH3


Figure 10. Chemical structures of the polymethine dyes to be tested for use in
the single molecule detection project.
length cuvette. A Spex 1680 double monochromator with 1 mm slits was used to
collect the spectra and a cooled Hamamatsu R636 photomultiplier tube served as the
detector. These spectra have been normalized to a laser power of 145 mW and dye
concentration of 4.6 x 107 M. The X maxima for IR 125, IR 140, and IR 132 with








69

A



-0 0
o 1.
co c
co






a)










00
coo
6( 0












o
I I 0 co



00000









(tyu) luejjno I|/4d .
2
0 00








70
this system were 806, 833, and 843 nm, respectively. Each of these dyes was tested

for application to the single molecule detection project as presented in Chapter 3.

More detailed characteristics of each dye will be discussed separately below.

IR 125. The formal chemical name for IR 125 is anhydro-l,1-dimethyl-2-[7-

[1,1-dimethyl-3-(4-sulfobutyl0-2-(1H)-benz(e)indolinylidene]-1,3,5-heptatrienyl]-3-(4-

sulfobutyl)-1H-benz(e)indolinium hydroxide sodium salt; its Chemical Abstracts

Services (CAS) number is [3599-32-4]. The compound is more commonly known by

many chemists as indocyanine green, or ICG, which is associated with its use as an

indicator. Laser spectroscopists, on the other hand, are more familiar with the name

IR 125 which is associated with its use as a laser dye. IR 125 is the only water

soluble dye of the three species, and is also readily soluble in most organic solvents.

In water, IR 125 has a pKl of 3.27.70 Like the other polymethine dyes, IR 125 is a

zwitterionic salt due to the presence of an aromatic heterocyclic ring containing

nitrogen.

Applications. IR 125 is the most widely used of the mentioned polymethine

dyes, mainly due to its solubility in water. By far, the most common uses of IR 125

are as a laser dye" and as a clinical indicator dye for testing of in vivo blood flows

and hepatic functions in animals and humans.7 It is useful in clinical applications

due to its large molar absorptivity at long wavelengths where blood does not absorb

strongly. It has also been used in angiography73 and many other studies in blood.2

A very interesting aspect of IR 125 for future applications is that it has been bound

to surfactants74 and proteins75 for analytical purposes. Patonay at Georgia State








71
University is actively pursuing the use of IR 125 and other polymethine dyes for

tagging purposes.67 If his research is successful, this SMD technique could become

very important in many tagging applications.

Indirect fluorometric detection. Another interesting use of IR 125 concerns

its chromatographic properties. It has been analyzed by high-performance liquid

chromatography (HPLC) for the clinical applications previously mentioned,7678 and

based on these studies, it was thought to be an excellent choice as a visualization

agent for indirect fluorometric detection in HPLC.79 Indirect detection works on the

principle that by monitoring the concentration of a continuously present indicator

species, termed the visualization agent, the presence of other species can be

determined by fluctuations in the concentration of the visualization agent.8 In this

manner, it is a universal method of detection with detection limits based on three

factors: 1) the size of the effect of the analyte on the signal of the visualization

agent (known as the transfer ratio); 2) the ability to measure these signal fluctuations

(or dynamic reserve); and 3) the concentration of the visualization agent.7 In HPLC,

ion exchange chromatography and capillary electrophoresis, the magnitude of the

transfer ratio can be very large based on separation properties, and IR 125 is a good

choice for this purpose because it can be detected very selectively and sensitively due

to its fluorescence at long wavelengths. A diode laser is suitable as the excitation

source due to its very high power stability which greatly reduces noise on a large,

constant signal. Indirect fluorometric detection with diode laser excitation of the IR

125 visualization agent was used in the detection of alcohols using reversed-phase








72
HPLC,79 and current work is underway to use indirect fluorometric detection in

capillary electrophoresis with the system. If successful, it will be possible to obtain

a LOD of 10 pM for an analyte with a transfer ratio near unity.: Indirect

fluorometric detection is of interest to separation scientists because it has the

potential of being a sensitive and universal detector which is a rare combination.

IR 140. IR 140 is formally known as 5,5'-dichloro-11-diphenylamino-10,12-

ethylenethiatricarbocyanine perchlorate; its CAS number is [53655-17-7]. The major

use of IR 140 is as a laser dye," but it has also been used as the analyte in several

LIF experiments with diode laser excitation.56,s7'",81 The goal of these projects was

very much similar as this one, which was to attain the smallest possible limit of

detection of the dye in a flowing stream. Based on these earlier experiments with

this dye and the very high fluorescence quantum efficiency (YF = 1) and a high

spontaneous emission (A21 = 1.26 x 109 s~1 based on 791 ps fluorescence lifetime,

rT),42 IR 140 is emphasized in the experimental studies.

IR 132. Like IR 140, the only known use of IR 132 is as a laser dye.71 It's

formal name is 3,3'-di(3-acetoxypropyl)-11-diphenylamino-10,12-ethylene-5,6,5',6'-

dibenzothiatricarbocyanine perchlorate (CAS # [62669-62-9]).

With the advent of diode lasers, these dyes have become very important as

probe species. Several analytical chemists have used these dyes in an attempt to

apply diode laser source to analytical techniques." Since diode lasers are only useful

at wavelengths longer than 650 nm, and polymethine dyes are one of the few

molecules fluorescent at these wavelengths, they have been thrust into several








73

applications.81 In the SMD project, the optimum dye is to be chosen based on the

conditions of the excitation wavelength and filtering to be discussed later in this

chapter and the next.

Stability of the Dyes. Due to the highly conjugated structures of these dyes,

it was anticipated that these dyes decompose readily. There have been numerous

studies involving the decomposition of IR 125 in blood plasma, water, and electrolytic

solutions, and indeed, IR 125 degrades in a matter of hours in these solutions.2

However, the decomposition rate is vastly reduced in organic solvents. A study was

performed to determine whether decomposition of the dyes would be a problem for

the single molecule detection project. Figure 12 shows the fluorescence of the dyes

in methanol over a period of 25 days. These measurements were made with a Spex

Fluorolog 2 spectrofluorometer with 1 mm slits and a cooled Hamamatsu R928

photomultiplier tube detector. The Xx was set to 764 nm because the 500 W Xe arc

lamp used as the excitation source has a greater emission intensity (and produces a

larger signal) at this wavelength than at the wavelengths of ECA of the dyes. Stock

dye solutions of m 1 x 10' M were kept at room temperature in the dark. Each time

a spectrum was taken, a 0.1 mL aliquot of solution was pipetted into a 1 cm path

length cuvette and diluted with 3 mL of solvent. As the figure shows, there was quite

some day-to-day variation in the procedure, but the overall fluorescence did not

decrease significantly. Therefore, the dyes do not readily decompose in organic

solvent (assuming that the degradation products do not fluoresce at the same

wavelengths as the dyes).













CM


0

I + 2
0*o



a"
N OO CO






,o *
0 0





S+ oJo
co


C-o .a
7*- 2








0
>o






-T z:, *
0 0 0







C\l
8- 3 0-r------ 0 Sn 4








75
Effect of degassing the solutions. In a similar study utilizing the same

instrument, the effect of bubbling different gases on the photodecomposition of the

dye solutions in methanol was examined. In this study, = 0.1 jiM dye solutions were

prepared, split into 3 volumes, and continuously sparged with nitrogen, helium, and

air. Their fluorescence emission spectra were measured after m 1 minute. These

solutions were then placed in a chamber radiated by light of wavelengths greater than

700 nm. For this purpose, a 500 W Eimac arc lamp was placed in the chamber, and

its white light emission was filtered by a series of long pass spectral filters. The

emission spectra of 3 mL aliquots of these solutions were measured periodically over

a period of 2 hours. Figure 13 shows the effect of degassing on the

photodecomposition rate of IR 140 in methanol. This figure clearly shows that air,

or more specifically oxygen, causes increased dye degradation as opposed to an

oxygen free solution. Figure 13 further shows that nitrogen and helium greatly

reduced the photodecomposition with respect to air, therefore, all solutions used in

the single molecule detection project should be sparged with one of these two gases.

Based on cost and ease of access, N2 was chosen for this purpose.


The Choice of Solvent


Since the goal of this project was to achieve single molecule detection with no

particular sample type in mind other than a solution, the choice of solvent was based

mainly on signal to noise of the dye fluorescence in that solvent. Of course, water

would be the ideal solvent due to its great applicability to many types of analyses, but

































































Iubj!S


a
k






rn
4)





0
0
4)



0
'U
4u
a
C0





r
O

1-1
4)


o



4)




0
()





U
*4)
s0







4)
*-4
-.9




y-4
05
CU
4)2

0
".f .


E

4--
cm
UC
a)
>








77
since IR 140 and IR 132 are water insoluble, and IR 125 decomposes rapidly in

water,82 H20 is not one of the choices as solvent for this experiment. Instead, several

organic solvents covering a wide range of properties, were tested.

Table 8 is a list of six solvents, and their pertinent characteristics. Of the

enormous number of organic solvents, these six were singled out due to their general

availability and history of use in fluorescence analyses. Polarity tends to be the most

important factor in the fluorescence intensity and maxima emission and excitation

wavelengths. For most fluorescent solutes, nonpolar solvents, such as hexane, tend

to shift the spectra to longer wavelengths and reduce intensity. Nonpolar solvents

are useful when a red-shift is desirable, or when the analyte is soluble only in

nonpolar solutions.


Table 8.

Solvent
Acetone
Acetonitrile
Dimethylsulft
Ethanol
n-Hexane
Methanol
Water


Physical properties of

Formula
CH3COCH3
CH3CN
)xide CH3SOCH3
CzHsOH
CH3(CH2)4CH3
CH3OH
H20


the solvents tested for use in

Density n'
0.7908 1.3588
0.840 1.3420
1.100 1.4783
0.7894 1.3614
0.6594 1.3749
0.7913 1.3284
1.0000 1.3330


this project.

eb
20.7
37.5
46.6
24.55
1.89
32.70
80.10


"n is refractive index at 589 nm; be is dielectric constant
All values for 200C from Lange's Handbook of Chemistry, 13th Ed., J.A. Dean, Ed.,
McGraw-Hill, New York, 1985.


I








78
The emission spectra of the IR dyes were taken with the laser tuned to 780

and 795 nm using the same instrumental system and normalized to the same

conditions as described in Figure 11 for methanol. The results of these spectra with

,, = 795 nm are compiled in Figure 14 which shows the normalized peak emission

intensities of IR 125, IR 140, and IR 132. Acetone and acetonitrile consistently yield

the largest fluorescence signal for the dyes whereas the nonpolar solvents, hexane

and dimethylsulfoxide produce less intense fluorescence. This is a typical behavior

for fluorescent compounds.

Raman spectra of the solvents. Based on Figure 14, acetone or acetonitrile

would be the best choices for use in the SMD project. However, signal is not the

only criterion on which the selection of solvent is based. Signal to noise ratio is the

most important parameter in any analysis to minimize limits of detection. As

discussed in Chapter 1, Raman scatter is the second most severe source of noise after

laser specular scatter, and if the MVF removes the laser scatter as it should, Raman

scatter from the solvent will then be the limiting source of noise. Since Raman

scatter is shifted in wavelength from that of the laser, it is transmitted through the

MVF to the detector. Therefore, it must be reduced at the source, by the reduction

of V, and choice of solvent, or filtered by a spectral filter other than the MVF.

Figure 15 contains the Raman spectra of the 4 solvents that gave the highest

fluorescence signals in Figure 14. These spectra were taken under the same

conditions with the Spex Raman microprobe instrument utilizing an Ar+ laser source

and cooled RCA C31034 detector. In Figure 15, the Raman shift in cm"1 from the















*I V I I


1 f I


0
-





CO


Ct
10
a:


( *II








51





CMd





15


SEEM


leu6!s Need p9z!lweJON


_ __


L


m


























E



00)
3I:
ooc'


leuB!S








81

514.5 nm Ar' laser line has been converted to wavelength as if the excitation

occurred at 794.76 nm. By referring to the fluorescence of the dyes in Figure 11, one

can see that the Raman spectra of the solvents overlap with the fluorescence signal

of the dyes. Only methanol is devoid of Raman peaks at wavelengths less than 860

nm. Assuming Raman scatter is the limiting source of noise, methanol is then the

best choice of solvent, despite the slightly lower intensity of the IR dyes in methanol

than in acetonitrile and acetone. These solvents give several very intense Raman

peaks around 820 nm.

To ensure that the extension of the Ar' excited Raman spectra to the

Ti:sapphire was valid, the spectra of methanol and acetonitrile were measured in a

1 cm square flow cell with Ti:sapphire laser excitation and the Spex 1680 double

monochromator/Hamamatsu R636 PMT combination. Figure 16 shows the resulting

spectra from this analysis with excitation at 795 nm. The wavelengths of the peaks

correlates with the wavelengths shown in Figure 15. It is unknown why the

background for the methanol spectrum is higher than it is for the acetonitrile in this

case. In the case of the argon ion laser excitation, the high background for methanol

is believed to be fluorescence of an impurity.

Dimerism. Another reason to choose methanol over acetonitrile as the

solvent is that the polymethine dyes dimerize in acetonitrile and not in methanol.

Dimerization is the tendency of solute molecules to associate with each other in pairs

instead of exist individually in solution. In the literature, other researchers have
















c O
0a
Oo N

6 co .


o-'
o
00
a) C




0
oi
-0 4)

<1 o a P



So




co
U I,
C)


cO

oo 0
4a





(vd) .UJJn4 1 1
0cc 0

0 0 0 02
+ + + +
w w w w








83
noted this problem with polymethine dyes in certain solvents.67'983 Dimerism is

exposed in calibration curves of signal versus concentration. In acetonitrile, the

calibration curve was nonlinear, and when converted to log-log scale, the slope of the

line was 2. Figure 17 is a calibration curve of IR 125 in acetonitrile taken with

Ti:sapphire laser excitation at 795 nm; based on the log-log slope of 1.99,

dimerization is clearly shown in this figure. This situation is much like the analysis

of sulfur with a flame photometric detector in which signal is proportional to

(concentration)2 due to the production of S2 in the flame. This is an unacceptable

situation when calculating limits of detection based on the sensitivity of an analysis.

Methanol, on the other hand, produces linear calibration curves and log-log plots

with a slope near unity. This indicates that dimerism of the dyes does not occur in

methanol solutions, and that LOD calculations based on sensitivity can be made of

this dye/solvent system.


Sample Containment


In an ideal system, the method of sample containment should not perturb the

sample in any way or introduce noise into the measurement. Despite this seemingly

innocuous task, sample contact with the sample cell may cause problems such as

introduction of reaction surfaces and matrix interference. In spectroscopy, light

transitions from one medium to another decreases the amount of light reaching the

sample due to scatter, fluorescence, and absorption by the containing medium.








84




ccJ
-W








0- |
\- 8 -






U) 00
\ en q,





S~- |S




\0 0
( U I1- 'I d



I I e5
a-


\- 0






\4 a-


o

LO '+ 0+J 0
+ + + + + + 0 '

V) PUPS I*9
Lx ^

*s 1^
o i 3








85
Furthermore, laser specular scatter from the container constitutes the major source

of noise in most IF analyses.

Liquid jets. Some of the problems with sample containers can be eliminated

by using a liquid jet, which is merely the term used for a flowing stream emanating

from a nozzle. The optimal probe region occurs at a point just after the orifice

where the flow stream narrows before it spreads again. Previous studies in this

group'657 used a liquid jet in the analysis of IR 140 in methanol with diode laser

excitation. In these studies, noise due to laser scatter was greatly reduced by exciting

the stream outside of the capillary, and the sensitivity of the measurement was

increased because more of the laser light was reaching the sample. However, the

major problem with a liquid jet is that it only operates at high flow rates. At lower

flow rates, the sample drips from the orifice one drop at a time which also causes

measurement difficulties. Another problem is that residence time of the analyte in

the probe volume is < 1 ms at flow velocities of liquid jets. For SMD, the desired

residence time (t,) is in the range of 1-10 ms, depending on laser power and dye

decomposition.

Levitation. Ideally, the analytical chemist would like to ensure that the

sample is isolated from external factors and positioned in the probe region for a

length of time sufficient for complete analysis. The most obvious way to accomplish

this goal is through sample levitation. Means of sample levitation involve the use of

physical forces to counteract gravitational attraction, or as in the case of performing

analyses in outer space, the reduction of gravitational effects. Of course, since space-








86
based research in analytical chemistry is overly expensive and generally impractical,

the former means of levitation are far more common than the latter. Practical types

of sample levitation possible make use of aerodynamic, acoustic," photophoretic,"

electrodynamic,6 and magnetic forces.7 As discussed in Chapter 1, Ramsey's

approach to SMD makes use of the advantages of electrodynamic levitation.4

Although sample introduction into the trap is complicated and time consuming, it

does remedy some problems related to a flowing sample contained in quartz. The

way in which Ramsey contains the sample is reviewed in the following paragraph in

order to give solid grounds for comparison with the other methods.

Electrodvnamic levitation. In electrodynamic levitation, a charged species

(atom, molecule, particle, or droplet) is introduced into a chamber containing a ring

electrode, to which is applied an ac (radiofrequency) potential, and top and bottom

electrodes to which are applied dc potentials." A field applied to the ring electrode

controls the lateral position of the charged species, and the dc potentials applied to

the top and bottom electrodes control the vertical motion. Figure 2 in Chapter 1

shows an electrodynamic trap in the drawing of Ramsey's set-up. With samples

having high mass/charge ratios (> = 10 kg/C), such as the charged, micron-sized

droplets introduced into the chamber by Ramsey, the droplets can be stably levitated

at the center of the trap. Sample introduction is accomplished with a piezoelectric

droplet generator at the top electrode of the chamber; the electrode voltages are

then altered until the droplet is trapped.43 The size of each droplet must be

determined from the complicated analysis of scattering patterns of a HeNe laser








87
beam focused onto the droplet.43 Furthermore, the analyte must be contained in a

solution partially consisting of glycerin to reduce solvent evaporation under the laser.

After a few minutes required for introducing, trapping, and determining droplet

volume, the fluorescence analysis begins. Based on these aspects, one can see that

practical analysis of large volume samples would be tedious by this method and rapid

measurements in a flowing stream would be impossible.

Sheath flow cuvette. The approach to single molecule detection of Keller's

group makes use of a sheath flow cuvette for sample containment.2" The sheath flow

cuvette is a type of flow cell in which an external solvent stream, or sheath, is used

to compress an internal sample stream.8 Figure 1 in Chapter 1 contains a partial

drawing of a sheath flow cuvette used in Keller's set-up. The narrowing quartz walls

of the flow cell cause the sheath flow, originating from outside the internal capillary,

to compress the sample stream as it effuses from the internal capillary. The solvent

in the sheath is usually the same solvent as the sample in order to avoid changes in

refractive index, but use of different solvents has certain advantages in some

instances.

In the set-up at Los Alamos, the laser probes a region in the cuvette where

the quartz tube is square (flat surfaces do not scatter as much light as round

surfaces). In a situation much like the liquid jet, but not as severe, the flow rates

required to induce the desired narrowing of the sample stream, greatly decrease t,

When the flow rate is lowered in an attempt to increase t, diffusion of the analyte

into the sheath occurs and the irradiance of the laser required to cover the entire








88
sample stream is not sufficient for SMD. To deal with this problem, Keller's group

decided to avoid it altogether by lowering flow rate to a point where t, is adequate

for the detection of the single molecules as they flow through Vp,2'3'5 but the laser

probes only 6% of the sample flow region (and the spatial filter only views a

portion of that region). This does not meet the 100% sampling efficiency

requirement of SMD expressed in Chapter 1.

The capillary. Unlike the other sample containment methods discussed, the

capillary is simply a glass or quartz tube requiring no instrumental adjustments.

Capillaries of various sizes are commonly used in several types of analytical methods.

In separation techniques using capillaries (gas chromatography, supercritical fluid

chromatography, microcolumn HPLC, capillary electrophoresis), the inner diameter

(i.d.) of the capillary is an important characteristic because it effects column volume,

flow, pressure, heat dissipation, and separation factors.9" In gas chromatography,

capillary i.d. is on the order of hundreds of micrometers, whereas in capillary

electrophoresis, the capillaries range from 10-100 /im i.d. It is for this latter

technique that this experiment is designed, so the sample capillary is on the order of

those used for capillary electrophoresis. Furthermore, narrower capillaries have

smaller probe volumes which is desirable in the reduction of Raman scatter from the

solvent. The key in choice of capillary i.d. is the laser focus size and t, at obtainable

flow rates which will discussed in later paragraphs. As it turns out, a 50 Jm i.d.

capillary is nearly ideal for the system.








89
Capillaries commercially available for capillary electrophoresis are made of

fused silica, or quartz, with an outer diameter (o.d.) of 150 or 360 Gim. For this

application, the thinner walled tubing was chosen because: 1) thinner walls absorb

less heat and better dissipate heat than thicker walls; 2) thinner walls have a smaller

interaction volume with the laser beam (less absorption, less Raman scatter); 3)

focusing with narrower capillaries is easier; 4) the microscope can be positioned

closer to the source of the fluorescence emission, if necessary, with thinner walls; and

5) the greater lensing curvature of the narrower capillaries focus the laser to a

greater extent at the sample than thicker capillaries. Some of these points will be

discussed in more detail in the section below.

Capillaries used for electrophoresis are commercially available and are coated

with polyimide which endows the brittle quartz with flexibility. At the probe region,

the polyimide coating must be removed, which is usually accomplished through

stripping or burning it off. Burning is advantageous because it does not create

grooves or scratches on the capillary walls as stripping may do. However, burning

does leave a sooty layer on the capillary, but this artifact is easily removed by wiping

with a wet tissue. Once the polyimide coating is removed, the capillary must be

handled delicately because the quartz tube snaps easily upon slight bending.


Optical Considerations Regarding the Capillary


Focusing in optical systems is paramount in attaining the optimum conditions

in a spectroscopic analysis. The rounded surfaces of a narrow capillary create several








90
considerations involving focusing of the laser onto and collection of emission from

the sample stream not normally encountered. Previously, researchers performing LIF

for detection in capillary electrophoresis have resorted to special means to reduce

the problems associated with round surfaces.12 In many cases, the sample cell is

made to be rectangular,88 or a separate cell is used altogether such as a sheath flow

cuvette.9 An interesting approach is to manufacture an immersible cell of a desired

shape containing a fluid of the same refractive index of the capillary.9' However,

in this SMD approach, confidence has been placed on the ability of the MVF to

remove the laser scatter, so the capillary is used without alterations. The subject of

ways to reduce laser scatter by the capillary will be discussed in more detail in

Chapter 3.

Laser specular scatter. Specular scatter from a rounded surface occurs in all

directions arising from both the outer and inner walls of the capillary. The modeling

of this system is very difficult due to the shapes of the focus beam, capillary,

collection optics, and the difficulty of predicting where rays will end up from a

capillary. As mentioned in Chapter 2, up to 0.7% laser scatter of the total laser

intensity can be tolerated in the direction of the collection optics according to

calculations presented in Chapter 1. Through elaborate experimentation and optical

ray tracing, Bruno et aL9 have shown that the least amount of scatter from a round

capillary occurs in the direction perpendicular to the excitation beam. This is

because the increased angle of incidence of the incoming light at the edges of the








91
capillary causes closely spaced rays to disperse more widely at 900 to the laser beam.

Fluorescence is to be collected from this direction (90*) in this approach.

Rayleigh scatter. Laser specular scatter produced at interfaces of different

refractive indices is by far the most intense form of scatter produced from the

capillary, but Rayleigh and Raman scatter also arise from within the quartz to a

lesser extent. Rayleigh scatter is very weak and occurs at the same wavelength as the

laser emission. These forms of scatter have slightly broader spectral profiles than the

source, but due to the small Rayleigh cross-section, small capillary volume

illuminated, and broad absorption profile of the metal vapor filter (with respect to

the laser), this form of scatter is expected to be negligible as shown from theory in

Chapter 1.

Raman scatter from the capillary. Raman scatter from the capillary may

prove to be a more formidable problem than Rayleigh scatter. Quartz has a Raman

emission band" at 464 cm'n which corresponds to 825 nm with excitation at 795 nm

and 809 nm for 780 nm excitation. The Raman spectrum overlaps the fluorescence

emission spectra for the polymethine dyes, and there is little to be done to filter this

radiation unless a portion of the fluorescence is to be filtered as well. The possibility

of using a capillary made of a different type of glass does exist, but commercially

available capillaries of this sort are made of fused silica, and there is no guarantee

that other glasses will not have the same problem. The cross section for Raman

scatter is typically on the order or 10-30 cm2 per molecule,3 and the volume of fused

silica illuminated is about 100 nL. Without knowing the MW of quartz, the number








92
of expected photons arising from the quartz cannot be calculated. It is estimated

based on the parameters presented in Table 2 in Chapter 1 that the maximum signal

from the quartz of the capillary is on the order of 10 photoelectrons per 2 ms

counting interval.


Focusing the Laser


Laser focus onto the capillary. Figure 18 is a diagram, generated by a

commercially available optical ray tracing program (Beam 4, Stellar Software,

Berkeley, CA), that represents the laser focusing onto the 50 im i.d., 150 Jim o.d.

quartz capillary to be used in this system. The incoming light is collimated and was

assumed to be of 786 nm wavelength. The quartz has a refractive index, n, of

1.45356 at this wavelength4 and the inner tube was said to contain methanol (n =

1.327 at 589 nm). The focusing of the laser beam by the capillary wall is an


Figure 18.


Focusing aspects of the 0.05 mm i.d., 0.15 mm o.d. quartz capillary
containing methanol with a 0.075 mm beam of collimated light at
786 nm entering from the left. Diagram generated by an optical
ray tracing program.




Full Text
APPROACHING SINGLE MOLECULE DETECTION
BY LASER-INDUCED FLUORESCENCE OF
FLOWING DYE SOLUTIONS IN A CAPILLARY
By
STEVEN JOHN LEHOTAY
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
1992
UNIVERSITY 5

This dissertation is dedicated to my father, Andrew L. Lehotay. Though he
died, he still lives in me.

ACKNOWLEDGEMENTS
I think Albert Einstein once said something like, "If I have seen farther, it is
because I have been standing on the shoulders of giants." I may not be able to see
very far, but it is more of a problem with "visual acuity" than elevation. The "giants"
in my case are Chris Stevenson, Ramee Indralingam, and Tye Barber whose research
efforts have had direct bearing on this dissertation. I would like to thank them for
their work, otherwise, this thesis would not have been possible.
Most of all, I would like to thank Dr. James D. Winefordner for his guidance,
insight, and knowledge. I am grateful for the opportunity to have been one of his
students and will forever be amazed and inspired by his intelligence, diligence, and
personality. He brought this project out of the clutches of despair with the simple
placement of a black piece of construction paper between the metal vapor filter and
the monochromator.
I also sincerely thank Dr. Benjamin W. Smith and Dr. Giuseppe A. Petrucci
for their help, patience, and friendship through the months of research leading to this
dissertation. I cannot count the numerous times I turned to Ben for advice, and
Giuseppe spent much time with me aligning the Tksapphire laser. He also modeled
the focusing aspects in a capillary presented in the dissertation.
in

I am also grateful for the work of several others in the group: Mike Wensing
wrote the program that calculates the Voigt profiles for the metal vapor cells; Nancy
Petrucci took the Raman spectra of the solvents; Yuan-Hsiang Lee helped collect
some of the data; and Wellington Masamba and Dennis Hueber helped with the
diode array software and use of the HR1000. Furthermore, I should thank the entire
JDW research group for their input, companionship, and spirit. They have all helped
make life in graduate school as stimulating, rewarding, and fun as it has been.
I very much appreciate the monetary support granted me by the state of Florida
and Texaco during my graduate school years.
Finally, I would like to thank my wife, Joann, for being loving, supportive, and
understanding during the stressful times, and all other times, leading to this
dissertation. I am a lucky soul to have her with me.
IV

TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
ABSTRACT vii
CHAPTER 1 INTRODUCTION AND THEORY 1
Introduction 1
Applications of Single Molecule Detection 2
Choice of Analytical Technique for Single Molecule
Detection 6
Theory of Single Molecule Detection 8
Definitions 9
Statistics of Data in Single Molecule Detection 14
Theory of Laser-Induced Fluorescence 16
Sources of Noise in LIF and Means of Noise Reduction 25
Laser Scatter 25
Raman Scatter 30
Background Fluorescence 30
History of Single Molecule Detection 32
CHAPTER 2 OPTIMIZATION OF INSTRUMENTATION AND
PARAMETERS 41
The Metal Vapor Filter 41
Theory of the Metal Vapor Filter 41
Choice of Metal for the Filter 43
Calculation of Spectral Linewidths and Absorbances for Rb .... 45
The Laser 55
Criteria of the Laser fo Single Molecule Detection 56
The Ti:Sapphire Laser 58
The Ti:Sapphire Laser/Rb Metal Vapor Filter Combination ... 62
The Sample 65
Choice of Analyte 65
Choice of Solvent 75
v

Sample Containment 83
Optical Considerations Regarding the Capillary 89
Focusing the Laser 92
Collection of the Fluorescence 94
Detection 100
Choice of the Detector 100
Photon Counting 103
Control of the Sample Flow 108
Experimental 112
CHAPTER 3 RESULTS AND DISCUSSION 119
Studies of the Metal Vapor Filter 119
Absorption Properties 119
Transmittance Properties 130
Additional Spectral Filtering 131
Spectral Filters 131
Polarization 134
Results 136
Limits of Detection 136
Noises of the System 141
Discussion of Limits of Detection 145
CHAPTER 4 CONCLUSIONS AND FUTURE WORK 147
Conclusions 147
Future Work 150
Elimination of Scatter 150
Other Future Improvements 155
REFERENCE LIST 159
BIOGRAPHICAL SKETCH 165
vi

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
APPROACHING SINGLE MOLECULE DETECTION
BY LASER-INDUCED FLUORESCENCE OF
FLOWING DYE SOLUTIONS IN A CAPILLARY
By
Steven John Lehotay
May 1992
Chairperson: James D. Winefordner
Major Department: Chemistry
The ability to detect single molecules in a solution has long been the ultimate
goal in ultratrace chemical analysis. In the strictest sense, single molecule detection
is defined as the efficient detection of a monomeric chemical species with near-100%
statistical certainty. Of current analytical techniques, laser-induced fluorescence has
the best chance of achieving single molecule detection due to its high sampling
efficiency, nondestructive probing, and high sensitivity. The limitations of laser-
induced fluorescence stem from noises associated with specular scatter from the
laser, Rayleigh and Raman scattering from the solvent, and background fluorescence
from sample contaminants. This dissertation concerns a new approach to eliminate
or greatly reduce these sources of noise for the purpose of single molecule detection.
In this approach, laser specular scatter is completely absorbed by a metal vapor filter,
Vll

which is simply a glass cell containing a metal element under reduced pressure.
When heated, the metal enters the vapor state and specifically absorbs the laser
scatter provided the laser spectral linewidth is narrower than the absorption band of
the metal vapor. Raman scatter from the solvent is reduced by containing the
sample in the small volume of a narrow capillary, and virtual elimination of
background fluorescence is accomplished by using near-infrared excitation with a
titanium:sapphire laser which specifically induces the fluorescence of a polymethine
dye. Currently, the Ti:sapphire laser is the only source with the power, tunability,
and narrow spectral linewidth that can be used in conjunction with a metal vapor
filter for single molecule detection. Rubidium was the element chosen for the metal
vapor filter due to its low melting point and ground state absorption transitions at
780.02 and 794.76 nm. In this dissertation, the theory of single molecule detection
and a review of previous approaches are presented, and the development and results
of this new approach are discussed. The lowest detection limit attained was 800
molecules of the dye, IR 140, in methanol flowing through a 140 pL probe volume.
Although single molecule detection was not achieved, the theoretically possible
detection limit of a single molecule could be attained through future experimental
modifications.
vm

CHAPTER 1
INTRODUCTION AND THEORY
Introduction
The singular purpose of this research is to detect a single molecule in a
chemical solution. If one defines single molecule detection (SMD) as being able to
determine individual molecules in a sample containing billions of undetected
molecules of that species, the goal of this research has already been accomplished
in many instances by others. However, in order to fulfill the requirements for
practical application of SMD, the definition must be expanded to require the
individual detection of nearly every molecule of the analyte contained in the sample
solution. SMD remains elusive by this definition, and if SMD is to be truly realized,
this definition must be met.
Obviously, the achievement of SMD is no simple task. It has taken many
years of experimentation and instrumental refinements for other researchers to reach
near-SMD detection limits. In this project, it is hoped that with a thorough
understanding of the theory and careful design and implementation of experimental
components, the approach described in this dissertation will be able to achieve SMD
in its initial attempts.
1

2
The purpose of this dissertation is to 1) present the theoretical criteria for
SMD; 2) review previous research on the subject; 3) discuss the development of a
novel approach to SMD; 4) show the results of experiments designed to attain SMD;
and 5) discuss these results and future research possibilities.
Applications of Single Molecule Detection
For the analytical chemist, being able to detect individual atoms and
molecules in a sample is a worthwhile goal in itself, and indeed, it constitutes the
goal of this project. However, analytical chemists are not usually involved in the type
of research that could be labeled as "purely academic." As scientists, analytical
chemists often take pride in the practical nature of their work. Therefore, other than
the fundamental, inherently worthwhile aspect of pursuing SMD, there are several
important potential applications of this research.
DNA sequencing. In consideration of the highly publicized project to decode
the human DNA sequence, Jett et al.x at Los Alamos National Laboratory have
proposed a strategy for rapid DNA sequencing relying on the detection of individual
nucleotides. This strategy involves specifically tagging the four different DNA (or
RNA) nucleotides with four different highly fluorescent dye molecules, then detecting
each of these molecules as they pass through a volume probed by a tightly focused
argon ion laser. If successful, the scheme is projected to sequence DNA 1000 times
faster than current techniques.1 Although it remains to be seen if this application can
be accomplished as proposed, research concerning the detection aspects of the

3
project have come very close to SMD.2'6 This research at Los Alamos will be
discussed in more detail later in this chapter.
Immunoassays. Like DNA sequencing, the immunological assay has become
a very prominent analytical technique in biotechnology.7 Immunoassay is a general
type of technique that relies on antigemantibody interaction for specific analyses of
many types of biological species. Currently, the technique requires considerable time
for an organism to produce sufficient antibody for an assay. If these type of analyses
could be performed on a molecular scale, much time and trouble could be saved, and
potentially more information of biological systems could be obtained. SMD in an
immunological system would enable the study of fundamental biological processes
at the molecular level.
Flow cytometry. Flow cytometric methods are another large area of research
in the biological sciences.8 In fact, the DNA sequencing strategy mentioned above
relies on a sheath flow cuvette, the main component of a flow cytometer, to contain
the sample.1 Currently, the ability to rapidly detect and sort single cells through flow
cytometric methods.8 In being able to detect single molecules as opposed to single
cells, flow cytometry would possibly allow separation of individual molecules.
Obviously, this would enable some very interesting studies to be performed.
Tracer studies. It is common practice in the determination of environmental
flow patterns, fates of chemicals in the environment, and metabolic pathways in
organisms to add a tracer compound into the system and follow its path. Due to
dilution effects, decomposition, and other losses, the analysis at some removed time

4
and place from the injection point may require the detection of the tracer in very
small quantities, potentially single molecules. These kinds of studies could be carried
out in a system on a global scale (such as the study of ocean currents or wind
patterns) or on a microscale (such as cell transport).
Fundamental research. For many physicists and physical chemists, it would
be interesting to determine whether the characteristics of an individual molecule
match its bulk properties. In the case of fluorescence, questions could be answered
concerning fluorescence lifetimes, quantum yields, quenching, and environmental
effects. Experiments involving SMD in solution may uncover important
physicochemical effects on a submolecular scale.
Detection in capillary electrophoresis and microcolumn chromatography. In
the 1980s, Jorgenson’s group made a major advance in separation science with the
development of capillary zone electrophoresis (CZE).9 CZE is capable of achieving
separation efficiences as large as 106 theoretical plates and allows for the separation
of several types of chemicals in aqueous solution, particularly biological species, that
were not easily separated previously.10 Similarly, it has been known for some time
that decreasing column diameter (or particle size) in liquid chromatography increases
separation efficiency.11 With the recently developed commercial ability to coat very
narrow capillaries with stationary phase material, microcolumn high-performance
liquid chromatography (HPLC) has become a viable method for separation analysis.
However, due to the small quantities involved in microcolumn HPLC and CZE,
detection is often the limiting factor in applications of these techniques. By their

5
nature, separation methods dilute the original analyte concentration, and for
complicated separations requiring a great deal of time, the analyte concentration at
the detector may be a small fraction of the original value. For many sample limited
applications, very small amounts of a particular analyte may be separated by
electrophoretic or chromatographic methods, but cannot be measured due to
detection problems. In essence, for separation techniques utilizing capillaries, SMD
may be required for some separations; however, very low detection limits are
required for all of these separations.
Due to the great potential of capillary separation techniques and their need
for sensitive detectors, a great deal of research has been undertaken concerning
detection of solutions flowing in capillaries. The most promising detection method
has been laser-induced fluorescence utilizing fluorescent probe molecules tagged to
the analyte.12 Based on the use of capillaries in both microcolumn chromatography
and capillary electrophoresis and the great applicability and needs of these methods,
a capillary was chosen as the sample container for this project. This point will
become more prominent later in this dissertation.
Unforeseen applications. In general, applicability of any technique depends
on whether the technique suits the specific needs of the application. In most cases,
the need for a type of analysis is the impetus for its development; this is only
partially true in this instance. As mentioned previously, the purpose of this project
is to achieve SMD. The experiment is designed to be useful for detection in capillary
separation techniques, and in the future could be used as such, but for the present,

6
SMD is an end in itself. If this project is successful, the announcement of SMD to
scientists in various disciplines would possibly cause the design of experiments to suit
the needs of the detector. It is difficult to predict what outcomes would arise by
accomplishing SMD, but the possibilities are very exciting.
Choice of Analytical Technique for Single Molecule Detection
Of all analytical methods currently in use, laser-induced fluorescence (LIF)
has the best chance of success for SMD in practice. Other techniques have been
considered, as discussed in the following paragraphs, but all were dismissed based on
theoretical or practical grounds.
All electrochemical methods in solution are incapable of rapid single molecule
detection due to the insufficient detection sensitivities to overcome the inherent
noises and response times associated with electrochemical measurements. Also,
molecular absorption spectrometry is eliminated on the fundamental basis that the
absorption cross section of a molecule is much smaller than the cross section of any
known light source beam. Similarly, Raman spectrometry also lacks sensitivity based
upon the small scattering cross-section.13 Several other types of spectroscopic
detection methods such as magnetic resonance, refractive index, thermal lensing,
photoacoustic spectroscopy, and all forms of thermally excited emission spectrometry
are also incapable of single molecule detection with current technology. Additionally,
radiochemical techniques possess insufficient sensitivity to detect a single molecule
on a practical basis.14

7
Mass spectrometry. On the other hand, many techniques utilizing mass
spectrometric detection15 have sufficient signal-to-noise ratios to detect single
molecules and are routinely capable of detecting individual molecular ions striking
the electron multiplier. In this respect, mass spectrometry is capable of single
molecule detection, but as mentioned earlier, practical SMD must occur with near
100% sampling efficiency. Such is not the case with mass spectrometry because of
unavoidable sample losses during introduction into the ionizing chamber and during
transport and separation of ions by the mass spectrometer. Mass spectrometry would
be a viable, practical approach to SMD, with the added benefit of molecular
identification, if these sample losses can be eradicated. Unfortunately, these are very
difficult problems to solve and it is not expected that mass spectrometry will be
useful for true SMD in the near future.
Scanning tunneling microscopy. A technique capable of true SMD, but not
yet useful for practical application, is scanning tunneling microscopy (STM). STM
has been demonstrated to allow viewing the atomic structure of molecules such as
benzene,16 and this capability has been suggested as an approach to DNA
sequencing.17 Scientists have developed STM techniques to the point that they are
able to view and manipulate single atoms contained on a metal surface.18
Despite these remarkable advances, STM is impractical for SMD of large
samples and in solutions. The analytical procedure for treatment of a sample for
STM involves coating the solid with an electrically conductive layer (commonly gold).
This process is as much art as science and can be very time consuming and expensive.

8
Once the sample is coated, the analysis can also be time consuming in trying to view
single molecules on a large surface. Furthermore, data interpretation is often a
subjective process often criticized for the imposition of imaginative viewing by the
researchers.
These problems are being addressed by current research, and if they can be
resolved, the potential for STM is enormous.19 Even so, STM has become a valuable
tool for the surface analysis of metals and many other solid materials, but it is not
currently practical as a general technique for SMD.20
Laser-induced fluorescence. As in mass spectrometry, lack of 100% sampling
efficiency plagues many atomic spectrochemical methods in flames and furnaces, such
as atomic fluorescence, laser-enhanced ionization, and resonance ionization.21,22
However, the direct molecular fluorescence analysis in a solution is theoretically and
practically capable of detecting an individual highly fluorescent molecule. Laser-
induced fluorescence (LIF) is commonly the most sensitive fluorescence technique,
and for this and other reasons to be discussed in a later section in this chapter, LIF
is the chosen method in the attempt to attain SMD.
Theory of Single Molecule Detection
In recent years, since technology has been developed that is capable of
extremely low limits of detection, much discussion has appeared in the literature
concerning the requirements and statistics concerning detection of single species.21'27
A useful summary of most of the concepts of SMD,28 as well as application of the

9
theory for the counting of atoms and molecules, has recently been given,29'32 and
much of the following discussion is based on these writings.
Definitions
Due to the many variables involved in any given analytical procedure, and the
many differences in those parameters when compared with other techniques,
acceptable terminology for SMD must be defined. Otherwise, the claims to SMD for
one type of analysis may not actually meet the requirements for another.
Single molecule. In general chemistry textbooks, a molecule, or compound,
is defined as, "a substance composed of more than one element, chemically
combined."33 By this definition, single molecule detection is accomplished by looking
at DNA under a microscope or touching a piece of plastic. This is unsuitable for the
ego of analytical chemists who perform ultratrace analysis, so for the purposes of this
dissertation, the word "molecule" has been modified to signify a monomeric
compound of reasonable size. For tagging purposes, "reasonable size" depends on
the application, but in general, a molecular weight of less than 1000 g/mol is
considered reasonable.
Single molecule detection. In classic papers, Alkemade23,24 stated the criteria
for the detection of individual species. These works mostly concerned single atom
detection (SAD), but apply also to SMD. He mentioned that SAD involves two basic
requirements: 1) an efficiency of detection of unity and 2) attainment of the intrinsic
noise limit. These two factors are defined below.

10
As in the case of defining a "single molecule," this definition must be modified
as well. In his papers, Alkemade indicates that SAD refers only to the spatially and
temporally probed region and does not account for sampling efficiency. By his
definition, achieving SAD by focusing a pulsed laser to a very small region of a flame
remains possible, despite that in this system, for each atom that is detected, hundreds
of atoms are not aspirated into the flame, thousands do not pass through the focus
volume, and thousands more pass unprobed during the time between laser pulses.
For the purposes of practical use of SMD in solution by LIF, more stringent
requirements for SMD are necessary. With this in mind, the third criterion for SMD
is that the sampling efficiency, e„ must be nearly 100%. By this definition, it is fair
to say that SAD/SMD has not yet been achieved.
Efficiency of detection. The first criterion of Alkemade for SAD is that the
efficiency of detection must be unity, or ed = 1. This means that each time that a
single species appears in the probe volume during the probe time, it must be
detected. In many cases, it is possible that single molecules can be detected, but only
a certain percentage of the times that a molecule is present is it detected. In these
cases, the researchers cannot make a valid claim to SMD.
Extrinsic and intrinsic noise. According to Alkemade’s second condition for
SAD, extrinsic noise must be eliminated at which time intrinsic noise becomes
prevalent. Extrinsic noise is the background produced by external factors such as
stray light, thermal fluctuations, and electric and magnetic fields. These factors can

11
be virtually eliminated experimentally with the techniques to be described later in
this chapter.
Intrinsic noise is the result of fluctuations of the signal itself. Sources of
intrinsic noise include the noise of the detector and power fluctuations of the source.
These are inherent features of the detection process that can be greatly reduced, but
will prevail as the limiting source of noise in the absence of extrinsic noise.
Limit of detection. One of the major figures of merit for any analytical
technique is limit of detection (LOD). LOD is defined as the concentration at which
the signal is 3 times larger than the standard deviation of the blank (abl), or
LOD = 3 where sensitivity is the linear slope of the analytical calibration curve of the detection
system. This definition was developed at a meeting of the International Union of
Pure and Applied Chemists (IUPAC) in 1976 to settle differences in the subjective
way in which detection limits were previously determined.34
In the signal domain, the measure of LOD is given the symbol Xj, which is
defined as,
Xj = Mw + 3ob„ (1*2)
where ^bl is the mean signal of the blank. When using this expression, a calibration
curve is not necessary; the analyte concentration at which this criterion is met is the
LOD.
Limit of guaranteed detection. Despite the IUPAC definition, LOD does not
always correspond to the best measure of lowest level of analyte detectability for a

12
system. In many circumstances, it is very difficult to actually observe a difference in
the signal at the limit of detection. Kaiser35 was aware of these problems and
introduced a term known as the limit of guaranteed detection (LOGD). Based on
statistical concepts, LOGD is set at twice the standard deviation requirement chosen
for LOD, which means
LOGD = 6abl/sensitivity,
(1-3)
or in the signal domain,
Xg = Mbi + 6abl,
(1-4)
where Xg is the signal produced at the LOGD.
False positives and negatives. Another way of looking at detection in a
chemical analysis is the occurrence of false positives and false negatives. For
example, at the LOGD, the probability of the occurrence of a false positive is
essentially zero. A false positive, or type I error, occurs when the data exceed the
criteria for the detection of the analyte, but in actuality, noise, not signal, has been
the cause of the occurrence. False positives occur at a probability a. False
negatives, or type II errors, transpire when the analyte is present at a sufficient
concentration to be detected, but the signal does not exceed the detection level.
Type II errors occur with probability /3.
In approaching the definition of detection limits based on the consideration
of types I and II errors, Xd (LOD in the concentration domain) corresponds to the
signal level that exceeds the background with a confidence of 1-a, and Xg (LOGD
in the concentration domain) is the signal level that gives a confidence of 1-/8 that

13
the analyte is actually being detected. Based on the definitions of LOD and LOGD,
the minimum confidence level is 99.86% which means that a and (5 must be 0.0014
or less.
Destructive and nondestructive probing. As mentioned previously, LIF has
a greater chance of success in realizing true SMD than other analytical methods.
The reason for this given earlier is the potential for high sampling efficiency by LIF.
However, an equally important factor is that LIF is a nondestructive probing method.
This means that the molecule is not destroyed in the detection process and more
than one detected event can occur per molecule. In fact, LIF is capable of producing
106 photons per fluorophore in a timespan of a few milliseconds which permits the
possibility of a higher noise level for an experiment that is still able to attain SMD.6
In destructive methods of detection, only one detection event can result from
each molecule. Destructive methods of detection include mass spectrometry,
radiometric methods, laser-enhanced ionization, and resonance ionization. In these
methods, the noise level must be extremely low, or signal of that one event must be
very high, to quantifiably detect a single molecule.
Symbols. Based on the definition of SMD, the laser must be continuous-wave
or very high repetition rate and probe the entire sample flow region. The focused
region of the laser is termed the probe volume, Vp, and the transit time an analyte
molecule in the Vp is the residence time, tr. During this interaction time, the number
of individual molecule(s) in the Vp, symbolized by Np, may give rise to a number of

14
detected events, Ne (photoelectrons in the case of a photomultiplier tube). The mean
background level during tr is symbolized by /xbl.
Statistics of Data in Single Molecule Detection
Poisson distribution. At low concentrations, Np follows a Poisson distribution,
and when photon counting is used for data collection at low levels, Ne and /¿b, also
follow a Poisson distribution. The Poisson probability distribution is given by36
P(X) = (1-6)
where P(X) is the probability of X events occurring (with X being Np, Ne, or noise),
and n is the mean value of X. In Poisson distributions, the variance equals the mean
(a2 = n, where a refers to the standard deviation)36 which applies at higher means
when Gaussian and Poisson distributions have a large overlap.28
In typical analytical measurements, the occurrence of noise, signal, and
numbers of analyte species in the detection region follow Gaussian probability
distributions and the expressions for LOD and LOGD were designed for these types
of analyses. However, in the case of SMD, a problem with the determination of
LOD and LOGD through the calibration curve method is that the slope of the
calibration curve at concentrations much higher than the detection limit, which are
Gaussian in nature, may lead to errors at near-SMD levels, which follow a Poisson
distribution.28

15
Criteria of signal and noise for SMD. Based on a Poisson probability
distribution,37 the values for Xd and Xg for a low-level counting experiment at various
background levels are given in Table 1. The table was constructed as follows:28
1) The Poisson probability distribution with a mean equal to the chosen
background level was found in reference 37. The detection limit, X¿, was
determined as the number of counts (minus background) at which the sum of
the remaining probabilities of the distribution beyond Xd did not exceed
0.0014 (confidence = 1-a or 99.86%).
2) A distribution was then found such that the sum of the probabilities greater
than Xd exceeded 0.9986 (1-/3). The mean of this distribution is Xg.
Table 1. Signal levels required to achieve single molecule detection with 99.86%
confidence at given mean blank levels.
Mean Blank Level
(/xbl, in counts)
Detection Limit
(Xd, in counts)
a < 0.0014
Guaranteed Limit
(Xg, in counts)
/3 = 0.0014
0.001
1
6.6
0.25
4
12.7
1
6
16
5
14
28
10
22
39
25
42
64
50
73
102
100
132
169
All parameters defined in text.

16
Based on these confidence levels, Table 1 gives the signal level required for
the realization of SMD at a given background level. Because of the differences in
practical aspects and statistics of data at near-SMD levels, Curie38 has advocated that
the confidence level be set to 99.5% (1.65a) rather than the IUPAC level. This
would significantly lower the values for X,, and Xg reported in Table 1.
The only remaining theoretical topic is whether the proposed LIF system is
able to meet the signal and noise levels presented in Table 1. These considerations
of LIF and its sources of noise are discussed in the next section.
Theory of Laser-Induced Fluorescence
Fluorescence is a physical phenomenon involving the absorption of a photon
of light by a molecule, causing an electron to climb to a vibronic level of an excited
singlet state, followed by emission of a photon of typically lower energy as the
electron returns to the ground vibronic level. Not all molecules undergo
fluorescence, and the ones that do often are highly conjugated and contain aromatic
functional groups. Fluorescence is useful as an analytical procedure because the
intensity of the emission is dependent upon concentration, and the excitation and
emission wavelengths of the light give some information as to the identity of the
fluorescent species. Due to factors to be discussed below, fluorescence is often a
very sensitive type of analysis.
As the name implies, laser-induced fluorescence (LIF) simply uses a laser as
the excitation source for fluorometric analysis. Lasers are able to produce higher

spectral irradiance (W/cm2) than common broad-band sources such as the xenon arc
lamp. Thus, LIF typically gives lower LODs for the analysis of fluorescent
compounds. Also, the narrow emission bandwidth of the laser is useful in many
situations that require selective excitation and detection of a fluorescent probe
species added to a system.
The practical formula that gives the average number of detected events, Ne,
in the type of LIF system used for SMD is13
N. = <-^-KYPÁ»Ttr> (1-7)
h*VSL 4ir
where L is the laser power (W), hi>L gives the energy per laser photon (J) with h
being Planck’s constant (6.636 x 10'34 J-s) and vL being frequency of the light (Hz),
SL is the cross-sectional area of the focused laser beam (cm2), aA is the cross section
of absorption for the molecule (cm2), YF is the fluorescence quantum efficiency, 0F
is the solid angle of collection of the fluorescence (sr), t? is the cathodic efficiency of
the detector (dimensionless), T is the transmittance of the optical components
(dimensionless), and tr is the residence time of each molecule in the probe volume
(s). These parameters are given by the manufacturer (as in the case of tj), easily
measured (L, T, and SL), referenced in the literature (YP), or calculated from other
known parameters (aA, flF, and tr). The following paragraphs discuss the
determination of these parameters for the system to be described in this thesis.
The cross section of absorption. The absorption cross section, aA, is not as
much of an actual "size" of a molecule as it is a statistical quantity. The units of cm2

18
arise from the fact that irradiance (W/cm2) is used in the expression to determine
the availability of light. The value for aA is the probability that the light available in
a certain area is absorbed.
The simplest way to determine of a known concentration in solution. Assuming the concentration falls within the
region of linear response, the molar absorptivity, eA (M 'cm1), at the chosen
wavelength can be found from Beer’s law,
A = ejC, (1-8)
where A is absorbance (dimensionless), ( is path length (cm), and C is concentration
(M). By knowing eA, aA can be found from,
where N is Avogrado’s number (6.02 x 1023). The absorption cross-section can also
be determined from fundamental parameters,39 but this method was used in this
project based on its simplicity and its use of an experimental measurement. The
absorption coefficients for the dyes to be tested in this project (given in Chapter 2)
are ** 200,000 M ’cm'1: therefore, the values for aK are approximately 3 x 10"16 cm2.
Fluorescence quantum yield. The quantum efficiency of fluorescence, or
quantum yield, YF, is the probability of emission of a fluorescence photon once a
photon has been absorbed. In mathematical form, YF is given by39

19
Y
F
kF
kF+knr
(1-10)
where kF is the rate of fluorescence (s'1), and k„ is the rate of nonradiative
deactivation (s1) of the excited singlet state (S,). Nonradiative decay of occurs
through the processes of external conversion (collisional deactivation), internal
conversion (nonfluorescent de-excitation), and intersystem crossing (St to triplet,
Tj).39 These factors are difficult to quantify, and depend strongly upon the molecule
itself and environmental factors such as temperature, pressure, solvent, and presence
of other species. Therefore, the YF for a particular system must be measured.
Basically three different methods are used to quantify YF, the simplest and
most common of which is the Parker-Rees method.39 This involves the comparison
of the fluorescence emission of an unknown fluorophore, YF u, with a fluorophore of
known quantum efficiency, YF k. Assuming constant power of excitation, this method
uses the equation,
V V €A,k^k
1 F,u 1 F,k ’
«A.u’Ju
(1-11)
where the differences in the absorption coefficients and detection efficiencies (a
function of wavelength) for the known and unknown fluorophores must be taken into
account. The most commonly used reference fluorophore is quinine which has a
quantum yield of 0.59 in an acidic aqueous solution.

20
The Parker-Rees method is not always simple to use, and due to other
correction factors not included above, is not always accurate. An easier way to
determine YP for a particular system is to search the physical chemistry literature for
an accurate determination of YF in the solvent to be used. THrough an intensive
literature search, it was found that that the dye has YP * l.42
Solid angle of collection. A microscope objective was to be utilized for the
collection of fluorescence in this approach. The solid angle of collection, flp, can be
calculated from the stated specifications of the manufacturer and the distance of the
point source fluorescence emission to the collection optics. It should be stressed,
however, that erroneously large values for flp result if one does not use the proper
equation for the calculation. The general expression normally used to calculate Q for
a lens is given as39
iIF = xtan20, (1-12)
where 0 is the angle defined by a line extended from the point source to the center
of the lens and a second line from the point source to the edge of the lens. The
problem with this equation is that it breaks down at large solid angles encountered
with microscope objectives. The calculated value of flP can exceed the true value by
a factor of 15% for 0 = 30° and by a factor of 300% for a 6 = 60°. The correct
expression for fiP is given by40
ilF = 2t(1-cos0) = 4irsin2(^). (1-13)
With this equation, fiF is less susceptible to error than with the previous equation.

21
When using a microscope objective, 6 is found from the stated numerical
aperature, N.A. of the objective, in that
N.A. = n sin0 = —, (1-14)
2f
where n is the refractive index (dimensionless) of the medium between the object
and objective, <£ is the aperture (cm) and f is the focal length (cm) of the objective.
The calculated i)F for the microscope objective to be used in this experiment is 1.5
sr which, when corrected for the \r sr of a sphere, corresponds to a collection
efficiency of 11.9% of a point source. The magnification (40X) and N.A. (0.65) for
the microscope objective used in this project, as well as other considerations of the
collection optics, are discussed in Chapter 2.
Calculation of the expected LIF signal. Now that the parameters of LIF have
been discussed, it is possible to estimate Ne for the conditions of this experiment.
Table 2 contains the values of the the relevant parameters determined by the
methods described above and in Chapter 2. By incorporating these values into
Equation 1-7 above, the theoretical signal level of a single molecule in this project
is 81 counts per 2 ms measurement period of the photon counter.
Although the accuracy of these parameters is thought to be very good, and the
Equation 1-7 is theoretically sound, the determination of Nc may not be truly valid
by this method for two reasons: 1) the potential for photodecomposition of the dye
before tn and 2) optical saturation of fluorescence. According to Equation 1-7,
simply increasing laser power, decreasing focus size, or increasing tr allows one to

22
obtain as large a signal as desired. However, physical limitations to negate this
possibility are described below.
Table 2. Parameters of the system to be used in the attempt of laser-induced
fluorescence detection of single molecules and the expected signal level
calculated from Equation 1-7.
Parameter Value
Laser Power, $L 200 mW
Laser Frequency, vL 3.77 x 1014 Hz
Laser Focus Area, SL 3.5 x 10"5 cm2
Absorption Cross-Section, aA 3 x 10"16 cm2
Dye Quantum Yield, YF 1
Collection Efficiency, Qf/47t 0.119
Detector Efficiency, 17 0.1
Optical Transmittance, T 0.5
Residence Time, tr 0.002 s
Expected Signal Level, Ne
81 photoelectrons
Optical saturation. Optical saturation occurs when the fluorescence signal
becomes independent of laser power. This effect happens when the rate of
fluorescence is limited by the time it takes the molecule to cycle through the
excitation/de-excitation process before it becomes available to go through another
cycle. Optical saturation13 is characterized by the relationship,
(1-15)
K §i+g2
where EL* is the laser irradiance at optical saturation (W/cm2), gt and g2 are the
statistical weights (dimensionless) of the ground state and excited singlet state, Slf

23
respectively, and A2, is the Einstein coefficient of spontaneous emission of the
fluorescence (s'1). This rate of spontaneous emission is simply the inverse of the
fluorescence lifetime, tf (s). To determine EL8, which is the laser irradiance when
the slope of a log-log plot of signal versus irradiance becomes 0.5, the relevant
equation is41
E, = (
gi+g2 YX
(1-16)
where c is the velocity of light (3 x 1010 cm/s), AXF is the full width at half maximum
of the fluorescence excitation and emission bands (cm), and X0 is the excitation
wavelength of the laser (cm). The value of A21 for the dye to be used in this system
is 1.25 x 109 s'1 (rF = 800 ps);42 the bandwidths of the excitation/emission spectra are
50 nm; the laser excitation wavelength to be used is 794.76 nm. The values for g,
and g2 are assumed to be equal (for S0 and S, vibronic levels, this is usually a valid
assumption). When these values are put into Equation 1-16, the resulting saturating
irradiance is 5900 W/cm2, which corresponds to 207 mW laser power in the 3.5 x 10‘5
cm2 focus size. By this account, the stated laser power above does not saturate the
transition and the calculated 81 photoelectrons per counting period is theoretically
obtainable with the stated parameters.
If laser irradiance, EL, is increased above EL8, Ne is no longer given by
Equation 1-7, but by

24
N = (_^)A21(iíi),TtI. (1-17)
gj+g2 4ir
It is not desirable to require the use of this equation, because in an actual analysis,
the laser power should be kept just below saturating conditions. For EL > ELS, the
signal changes by less than a factor of 2, but the noise continues to increase with EL.
Photodecomposition. The second pitfall of Equation 1-7 is degradation of the
the dye before it emits as many photons as the theory predicts in the 2 ms sampling
time. Other researchers performing SMD have encountered this problem4344 and
developed a procedure to determine the optimum parameters to reduce dye
degradation. In this optimization technique, the end result is that tr should roughly
correspond to the time it takes the molecule to decompose under the laser irradiance
of the experiment. For the dyes in these experiments, the molecule typically
undergoes « 106 fluorescence cycles before it degrades.43 If this value holds true for
the dye to be used in this project, the time it would take for the molecule to
photodecompose at the conditions listed in Table 2 is 146 ms, nearly 100 times
longer than the measurement time of this experiment. Based on this estimate, it is
hoped that photodecomposition will not become a factor in this attempt at SMD.
However, if optical saturation occurs to the point that the molecule undergoes
fluorescence at the theoretical limiting rate (A21 = 1.25 x 109 for the fluorophore42
of Table 2), it would only take 0.8 ms to go through 1 million cycles.
Noise level required for SMD. Based on the theory of LIF presented here,
the average signal level should consist of 81 counts above background per counting

25
interval. Using the statistical theory applied to SMD presented earlier, the maximum
mean noise level, /xbl, permitted during the 2 ms interval to attain an LOD of 1
molecule with 99.86% confidence is 56 counts. For LOG = 1 with 99.86%
confidence to be achieved with 81 counts above background, /xb, would have to be 35
counts or less. Remember in the calculation of these values it is assumed that the
data follow a Poisson probability distribution with the variance equal to the mean.
There is some question as to whether this is true when actual data are collected.32
A key to the success of this project is the reduction of noise to this required level.
The next section is a discussion of sources of noise in LIF and how they can be
eliminated or reduced.
Sources of Noise in LIF and Means of Noise Reduction
The three limiting extrinsic sources of noise for typical LIF experiments
consist of scattered light, Raman scatter, and background fluorescence. Other less
severe sources of noise exist, but by far, these three are the largest sources.
Laser Scatter
Laser scatter can be divided into two categories, specular scatter arising from
reflections from optics and other surfaces, and Rayleigh scattering from the solvent.
Laser specular scatter. In nearly every analysis utilizing LIF, scattered light
from the laser constitutes the most severe source of noise. This feature is not
surprising when one considers the amount of light that is delivered by the laser. For

26
example, the number of photons focused on the sample by the 200 mW laser at
794.76 nm to be used in this project is 8 x 1017 s'1 or 1.6 x 1015 photons during the 2
ms counting period. Of these, only 1 photon in 117 billion ( molecule in this time, which essentially still leaves some 1.6 x 1015 nonabsorbed
photons. Realizing that fiF/4ir is nearly 12%, 77 is 10%, and T is 50% for the system,
the detector would produce * 9.6 x 1012 photoelectrons per 2 ms assuming that the
laser light is scattered isotropically from a point source. However, this assumption
is not true; the vast majority of the laser light continues unhindered through the
sample container and is not scattered within the 1.5 sr region collected by the
microscope objective (which is why fluorescence is collected 90° from the angle of
excitation). For the sake of argument, assume only 0.1% of the photons are
scattered. This still corresponds to nearly 10 billion photoelectrons. The 81
photoelectrons emitted by a single fluorophore in the same time period pales in
comparison. If the above assumptions are correct, an absorbance of > 10 is required
to reduce the scatter to a level compatible for SMD with this system.
Monochromators. As can be surmised based on this analysis, laser scatter is
an enormous problem experimentally. Even with the spectral selectivity of a
monochromator, stray light rejection is typically on the order of 105 which still leaves
some 50,000 photoelectrons produced under the conditions stated above (the 0.1%
isotropic point source of laser scatter at the focus of the collection optics is not valid
except in the case of capillaries as will be discussed later). Furthermore, with
monochromators, the spectral bandpass of the emission process is also greatly

27
reduced which lowers the signal as well as the noise. Therefore, the use of a
monochromator in the SMD approach is inadequate to the task.
Spectral filters. The use of spectral filters is a common approach to reduce
laser scatter in LIF experiments. Typical rejection of interference filters and long
pass spectral filters is 103, which is poorer than the rejection obtained with most
monochromators, but the filters generally have a much greater optical throughput of
the fluorescence signal. In most cases, two or more filters are used together or in
conjunction with a monochromator. Again, the problem of this approach is that as
the filters reject more laser scatter, the signal is also reduced. This type of filtering
is explored in more detail in Chapter 3.
Polarization. Lasers are typically highly polarized sources of emission.
Through the use of polarized filters, the spectroscopist can take advantage of this
trait to reduce laser scatter because the large majority of scattered light retains its
polarization. Conversely, fluorescence emission is nonpolarized. Optimally,
polarized filters are capable of 109 rejection of light polarized in the same direction
as the filter and are still able to pass light of the opposite polarization. This means
that approximately half of the nonpolarized fluorescence should pass through the
filter.
In practice, polarized light rejection does not work as well as expected from
theory. Problems arise with the purity of the laser emission polarity and maintaining
polarity when scattered from complex materials. Experimentation with stray light
rejection by polarization is discussed in Chapter 3.

28
Spatial filtering. As mentioned above, the focused laser light does not scatter
equally in all directions. The pattern of laser scatter produced at the sample greatly
depends on focusing and the shape of the sample container (this aspect becomes very
important later in this dissertation when considering the capillary container used in
this project). Spatial filtering exploits the nonisotropic feature of the scatter through
the placement of a small slit or pinhole between the collection optics and the
detector. With careful positioning and focusing, the aperture can collect a large
percentage of the fluorescence while blocking much of the laser scatter arising from
the edges of the container. The implementation of this simple concept by Dovichi
et aL4S (along with the use of 3 spectral filters and sheath flow cuvette sample
container) was a great break-through in initial studies on single molecule detection.
The problem with the spatial filter in SMD, however, is that it can limit the
probe volume of the analysis. By collecting emission from only a portion of the
focused region, the sampling efficiency is reduced which circumvents one of the
conditions for SMD. It is also very difficult to position the optics and spatial filter
for optimum effect.
Rayleigh scatter. Another drawback with spatial filters is that they do nothing
to limit the amount of Rayleigh scattering reaching the detector. Unlike specular
scatter, which arises at interfaces between media of different refractive indices,
Rayleigh scatter occurs in the medium itself due to light interaction at the molecular
level. Furthermore, the fraction of light that is scattered specularly (measured as a
percentage) is very high compared to Rayleigh scatter which typically has a cross-

29
section on the order of 10"28 cm2. This corresponds to the production of 1
photoelectron at an average of every 2 ms for methanol in the 140 nL Vp of the
conditions presented in Table 2. Although this is very small, at the light levels
involved in this project, every additional noise photoelectron could become
significant.
The metal vapor filter. A possible way to eliminate the serious problem of
laser scatter is through the use of a metal vapor filter (MVF). This device is central
to the success of this project and its theory is presented in Chapter 2 and
experimental results in Chapter 3. The theory is too extensive to be presented now,
but according to theory, the MVF is capable of essentially totally absorbing the laser
scatter (or the entire laser emission for that matter) provided the laser emission
bandwidth is narrower than the absorption band of the metal vapor. Moreover, the
MVF is completely specific to the laser wavelength and does nothing the hinder the
transmittance of nearly the entire fluorescence emission band of the fluorophore.
Based on the calculated 10 billion laser scatter photoelectrons, it must suffice to say
that the required rejection of the MVF must be on the order of 1010 or higher to
achieve SMD, which is theoretically possible as shown in Chapter 2.
Even though the MVF is capable of eliminating the problems with laser
specular scatter and Rayleigh scatter under conditions of this experiment, other
extrinsic sources of noise exist that would thwart SMD. Both Raman scatter and
background fluorescence from the solvent and optics occur at removed wavelengths
from the laser, which allow their passage through the MVF to the detector.

30
Raman Scatter
Unlike Rayleigh scattering, Raman scatter is an inelastic process that occurs
with even lower probability, typically with a cross-section of 10'30 cm2 per molecule.39
This hardly appears significant until one realizes that in a large volume, the number
of molecules in a material is so great that the value for Ne becomes significant. As
LIF researchers have learned, noise due to Raman scatter from the solvent becomes
the limiting source of noise when the laser scatter is reduced. The way around this
problem is the reduce Vp which limits the number of atoms and molecules in the
container that scatter light. This is one of the reasons why all LIF approaches to
single molecule detection utilize a small Vp. In this project, with the parameters
listed in Table 2, the Vp of 140 pL containing methanol would give rise to about 1
photoelectron per every 100 counting intervals. This is an acceptable level for SMD,
but a more difficult to quantify amount of Raman scatter arises from the quartz of
the capillary. This and other aspects of Raman scatter are discussed in Chapter 2.
Background Fluorescence
As discussed in the theory, fluorescence has a rather high absorption cross-
section (aA = 3 x 1016 cm2 for the fluorophore presented in Table 2) and can be
measured very sensitively. It is a selective technique as well, but it is often unable
to specifically detect one fluorophore at low concentration in the presence of
another. Furthermore, no solvent is absolutely pure, and even the presence of an
ultratrace concentration of fluorescent interferents can negate the possibility of SMD.

31
There are two main methods to avoid this problem. In the first case, the
purest available solvent should be used, and secondly, the detection should be
designed to be as specific to the analyte as possible. The former method is not trivial
in even the best available solvents,4 and attempts at SMD are consigned to basically
working with standard solutions of the purest solvent. However in the application
of these techniques to a real sample, interfering species become a severe problem
in the production of background fluorescence. In a real analysis, it is not realistic to
assume that the analyte will be the only detectable fluorescent species in the sample.
Excitation at long wavelength. The best way to avoid background fluorescence
is to specifically analyze the analyte. Very few species fluoresce at far-red/near-
infrared wavelengths, and the only known dyes to do so at the laser excitation
wavelength of this project are given in Chapter 2. Therefore, the method of
detection for this project is very selective to the molecule of interest.
An additional benefit of using laser light at 794.76 nm, as opposed to the
more commonly used 325 nm emission from a HeCd laser or the 514.5 nm line from
an Ar+ laser, is that Rayleigh and Raman scattering processes are reduced by a
factor of l/\04 (and specular scatter is also reduced to a large extent).39 This is a
substantial reduction in noise when pursuing SMD.

32
History of Single Molecule Detection
Now that the pertinent concepts of SMD and LIF have been introduced, the
previous accomplishments of LIF analysis nearing SMD can be reviewed without
having to define terms or explain the rationale behind the design of the experiments.
Single atom detection. In the past, there have been several instances of SAD
most notably through the research of Letokhov46,47 and Hurst.48 Their separate work
concerns the use of resonance ionization spectrometry to detect atoms in an ion trap
or in an atomic beam. As discussed earlier, these experiments meet the
requirements of SAD as stated by Alkemade,23,24 but do not conform to the practical
definition of SMD of this dissertation.
LIF of solids. In the case of molecules, single molecule detection has been
accomplished under somewhat artificial circumstances. Hirschfeld49 implemented LIF
with a microscope to detect a single protein molecule (MW * 20,000) tagged with
80-100 fluorescein molecules on a solid substrate. Kirsch et al.,50 in a similar type of
procedure, were able to detect 8000 rhodamine 6G molecules. More recently,
Moerner51'53 has detected single pentacene molecules in a solid matrix at low
temperature using laser-excited fluorescence.
Keller’s approach to SMD. In the analysis of flowing solutions, Keller’s group
performed several experiments leading to the claim of single molecule detection
(although none of the reports satisfies even Alkemade’s definition of SMD).2"6,45,54
The origin of Keller’s project at Los Alamos National Laboratory, which is designed
for the application of SMD for DNA sequencing,1 began with research by Dovichi

33
et al.*5 who bested the previous lowest LOD by nearly two orders of magnitude in
obtaining an LOD of 35,000 rhodamine 6G molecules. The subsequent experiments
at Los Alamos concerned refinements of the basic set-up developed by Dovichi.45
As shown in Figure 1, this basic set-up utilizes a tightly focused argon ion
laser to excite a highly fluorescent dye flowing in a sheath flow cuvette. A
microscope objective collects the fluorescence, spatial and spectral filtering reduces
scattered light, and a cooled photomultiplier tube coupled with photon counting
electronics measures the signal.
Table 3 is a list of the parameters and detection limits of published results
reported by the Los Alamos group. From the table, it is apparent that the
researchers have been slowly lowering the detection limit with difficulty. In Ref. 4
and Ref. 6, in which single molecule detection was claimed, the sampling efficiency
(e,) was very poor, water was not used as the solvent, and the sheath flow cuvette was
abandoned. Also, the traditional method of determining the LOD was not used;
instead autocorrelation analysis of a single sample was performed. In Ref. 4, the
research team used a laser with 70 ps pulses and a microchannel plate detector with
sophisticated signal collection to help discriminate the fluorescence from the scatter.
The data presented in Ref. 4 do appear to be single molecule events, but the signal
to noise ratios are not reported so a statistical treatment to determine LOD cannot
be performed. The researchers base their claim of ed = 70% (of those molecules
passing through the center of the Vp) on computer simulations that appear similar
to the actual data. Weighted quadratic sum plots of the data presented showed

34
Ar+
Laser
1/2 Wave
Plate [
Polarizing
Prism
Mirror
Amplifier/
Discriminator
Counter
Spectral
Filter
Spatial
Filter
iObjective
Lens
Lens
Figure 1. The instrumental approach to single molecule detection used by
Keller’s group at Los Alamos National Laboratory. Redrawn from
references 2, 3, and 5.

35
Table 3. Comparison of the reported parameters and results of the laser-
induced fluorescence experiments of Keller’s group at Los Alamos
National Laboratory.
Parameter
Ref. 2
(1984)
Ref. 3
(1987)
Ref. 4
(1990)
Ref. 5
(1991)
Ref.6
(1991)
Analyte
R6G
R6G
R6G
R6G
R6G
Solvent
h2o
h2o
HjO/QIIjOH
h2o
C,H5OH
Cell
SFC
SFC
flowcell
SFC
flowcell
Laser,
XL (nm)
Ar+
514.5“
Ar+
514.5
Nd:YAG
532b
Ar+
514.5
Ar+
514.5
SL (cm2)
3.8xl0"6
l.lxlO6
4.4xl0'7
2.1xl0"5
2.1xl0‘5
El (kW/cm2)
130
700
6.8
40
23.4
Stream Size
(fim)
30
42
4000
44
250
nF/4ir
0.06
0.045
—
—
—
e,
0.6
0.06
1.9xl0"7
0.1
0.05
Vp (PL)
11
0.6
0.44
11
10.7
Flow Rate
(/xL/min)
25
0.012
5760
1.0
0.18
Flow Velocity
(cm/s)
60
14.2
0.075
5.4
4.85
tr (ms)
0.037
0.085
10
1.8
2
Tc (s)
1
1
0.004
. 1
0.0004
LOD (M)
1.3xl0'13
2.2xl0"13
—
9xl015
—
LOD (#/rc)
33,000
1,200
~lc
33
»ld
R6G = Rhodamine 6G; SFC = Sheath Flow Cuvette; e, = Sampling Efficiency;
7C = Time Constant of Measurement;
LOD (#/tc) = number of molecules passing through Vp during tc at the LOD (M).
“Pulsed at 10 kHz, 50% duty cycle; bPulsed at 82 MHz, 0.57% duty cycle.
cTime discrimination method able to detect passage of single molecules with reported
70% detection efficiency (ed).
dReported detection limit based on autocorrelation analysis; ed not given.

36
photoelectron bursts arising from passage of single molecules which were not
presented in any of the other references. The same sort of basis was used in the
claim to single molecule detection made in Ref. 6, but in this case, passage of
individual molecules in the Vp was not noticeable. The claim to single molecule
detection was based on the use of a 20 point sliding sum distribution which agreed
with theoretical results. No efficiency of detection can be calculated from such a
determination because the single events could not be counted.
In references 3 through 6, the spatial filter viewed only a small portion (5 /xm)
of the laser focus volume to reduce the background sources of noise discussed
earlier. This technique limited the solid angle of collection of the microscope
objective which is why flF/4ir is not reported in those cases (except Ref. 2). In Ref.
5, which reported an LOD of 9 x 10"15 M based on conventional methods to
determine the LOD, the /¿bl of the PMT was 212,764 + /- 461 counts/s. This high
noise level is a result of the large background sources of noise discussed earlier. At
this noise level, SMD is not possible with the approach presented in this thesis, ñ
a separate study utilizing B-phycoerythrin as the analyte, Nguyen et al54 in Keller’s
group claimed the first instance of single molecule detection in solution. It is
noteworthy that B-phycoerythrin is a very large (MW * 250,000 g/mol) protein
possessing the equivalent fluorescence of 25 rhodamine 6G molecules.54
Mathies’ research. At the University of California at Berkeley, the research
group of Mathies contested this initial claim to SMD made by Keller’s group; they
repeated the LIF study with B-phycoerythrin.55 Through a more rigorus statistical

37
approach, their research effort showed that the previous work did not obtain SMD,
and their results demonstrated the detection of 15% of the passing single molecules.55
Based on the definition of SMD presented in this thesis, B-phycoerythrin, with
a MW « 250,000 g/mol, does not qualify as a "single molecule," and the sampling
efficiency of the experiment is much less than unity. Keller’s group2"6 and other
researchers43 are aware of these shortcomings and have worked to lower the LOD
for smaller fluorophores. In their proposal,1 Keller’s group1 mentioned that sampling
efficiency must be increased to perform the desired DNA sequencing application, but
the reduction of noise through the use of the spatial filter was integral to the
detection limits they have achieved.
Also, the sheath flow cuvette as the sample container is inadequate to achieve
SMD with e, = ed = 1. Further addressed in Chapter 2, sampling efficiency and
detection efficiency are diametrically opposed relationships with the sheath flow
cuvette. To increase sampling efficiency, the laser beam focus must be increased, but
to increase detection efficiency, the focus must be kept small. Furthermore, with
small probe volumes, the sample flow rate must be lowered to maintain the residence
time, but with a sheath flow cuvette, the flow stream becomes broader with
decreasing flow rate thus requiring a larger beam focus. This has been a difficult
problem with the sheath flow cuvette, and researchers using this device have resorted
to finding an optimum trade-off.44 In this respect, it is unlikely that research with a
sheath flow cuvette will ever achieve true SMD. In fact, the time discrimination
approach of the Los Alamos group with the frequency doubled Nd:YAG laser

38
appears more promising for true SMD than the approach exhibited in Figure 1.
Indeed, two of the recent papers by Keller do not use the sheath flow cuvette.4,6
Winefordner’s approach. To avoid the problem of this trade-off with a sheath
flow cuvette and its high cost, Winefordner’s group has decided to probe the entire
sample stream in their attempts at SMD. Furthermore, the diode laser was chosen
as the excitation source for purposes of greater analyte selectivity, lower noise from
Raman scatter, lower cost, simplicity, and the many other advantages of diode lasers
to be discussed in Chapter 2. In two separate studies (with different lasers), LODs
of 40,000 and 3,000 molecules flowing in the probe volume of the near-infrared dye,
IR 140, have been measured in a liquid jet (flow stream emanating from a
capillary).56,57
Despite the many advantages of working with diode lasers and the low
detection limits attained with their use, LIF with diode lasers was found to lack the
sensitivity to achieve SMD in a liquid jet. Furthermore, the liquid jet only operates
at high flow rates which gives unsuitable residence times for SMD. This flow
condition is one of the reasons why the method described in this thesis employs a
capillary for sample containment.
Ramsey’s approach. Ramsey’s group at Oak Ridge National Laboratory has
designed the first instrumental set-up capable of truly realizing SMD in solution as
defined in this dissertation.43 The design of this approach appears in Figure 2. The
key to the experiment is the use of the electrodynamic trap to contain droplets of the
sample instead of a flowing stream. In this way, the entire sample is probed (one

Figure 2.
The instrumental approach to single molecule detection used by Ramsey’s group at Oak Ridge National
Laboratory. Redrawn from reference 43.
u>
VO

40
drop at a time) and the interaction time can be on the order of days if necessary.
The noise rejection is not as high as in other experiments and the dye tends to
photodecompose before SMD can be quantitatively attained. The LOD currently
stands at an average of 25 molecules of rhodamine 6G per droplet.
The approach to SMD of this dissertation. It is hoped that the SMD approach
described in Chapter 2 will be the first method to truly achieve SMD in a solution.
Like Ramsey’s approach, it is designed to sample the entire solution, and has been
shown to theoretically perform the desired assignment. More of the theory, design
and development considerations, and some results will be presented in the following
chapter.

CHAPTER 2
CHOICE OF INSTRUMENTATION AND
OPTIMIZATION OF PARAMETERS
The Metal Vapor Filter
Because of its role to remove the laser specular and Rayleigh scatter from the
collected light, the metal vapor filter (MVF) is the most important component of the
experimental set-up to attain single molecule detection (SMD). With few exceptions,
laser scatter is the limiting source of noise in ultratrace analyses using laser-induced
fluorescence (LIF). Therefore, satisfactory performance of the metal vapor filter is
crucial to the success of the experiment.
Theory of the Metal Vapor Filter
The MVF typically is a 2-3 inch long glass cell that contains a surplus amount
of a metal in its elemental form enclosed in a nitrogen environment. Figure 3 is a
simple drawing of the MVF. Upon gentle to moderate heating of the cell, a portion
of the solid or liquid metal enters the vapor state depending on temperature,
pressure, and thermodynamic properties of the element. The absorption properties
of the MVF depend on the number density, or concentration, of the metal in the
vapor state, the length of the cell, and the pressure in the cell.
41

42
Pyrex k Cylinder
Nitrogen
Metal Element
Figure 3. The metal vapor filter.
Absorption. Increases in the metal vapor number density and absorption path
length lead to directly proportional increases in absorbance assuming the linewidth
of the source is more narrow than the absorption bandwidth. This relationship is
exhibited by Beer’s Law,39
A = 0.434aAfn, (2-1)
where A is the absorbance (dimensionless), i is the absorption path length (cm), and n is the number density of the absorber
(cm'3). In the case of the MVF, the equation to determine the absorbance of an
atomic transition can be calculated from,58
e 2f..S vn 1
A = 0.434—L, (2-2)
4e m c
o e
where e is the charge of an electron (1.6 x 10'19 C2), fy is the oscillator strength for
the electronic transition from level i to level j (unitless empirical value for each

43
transition), S,v is the Voigt shape function of the absorption bandwidth (Hz1), n¡ is
the number density of the atom in level i (m'3), eD is the permittivity of free space
(8.854 x 10'12 Ns2/C2), me is the mass of an electron (9.11 x 10'31 kg), and c is the
velocity of light (2.998 x 108 m/s).
Choice of Metal for the Filter
For the MVF to be useful in practice, the electronic absorption transition
must start from a ground state due to the small fraction of the element existing in
an excited state as described by the Boltzmann distribution. Therefore, level i in
Equation 2-2 must be the ground state, and n¡ must be the number density in the
ground state. Another important consideration for the practical use of a MVF is the
thermodynamics of the chosen element. In order to attain a large enough number
density, n¡, to satisfactorily absorb the large light intensities associated with a laser,
the metal must have a low melting point and high vapor pressure. Table 4 is a list
of elements that could be of practical use in a MVF along with their melting points,
ground state transition wavelengths, and approximate aA of the overall transition. Of
these elements, rubidium is an excellent choice for use in a metal vapor filter due to
its low melting point and strong absorption lines at 780.023 nm and 794.760 nm. For
these reasons and more (based on characteristics of lasers and fluorescence dyes
available, which will be discussed later in this chapter), rubidium has been chosen as
the element to include in the MVF for use in the SMD system.

44
Table 4. Thermodynamic characteristics of several elements for possible use in
the metal vapor filter.
Element
m.p. r°cv
X (nml*
gA (cm2)b
A1
660
308.216
9 x 1013
Ba
714
553.548
2.5 x 10 "
Ca
838
657.278
422.673
8 x 1016
2.1 x 10u
Cd
321
228.802
1.3 x 10 "
Cs
29
852.110
894.350
455.530
9.5 x 10 “
4 x 1011
8 x 1012
Hg
-38
253.652
8.7 x 1012
In
156
410.476
4 x 1012
K
64
766.491
769.898
404.410
2 x 10“
1 x 1011
3 x 1013
Li
181
670.784
323.260
6 x 1012
3.5 x 1013
Mg
650
285.213
1.3 x 10 "
Na
98
588.995
330.232
1 x 10“
7 x 1014
Pb
327
283.306
4 x 1012
Rb
39
780.023
794.760
420.180
1.3 x 1010
6 x 10"
1.1 x 1011
Sn
232
286.333
1.3 x 1012
Sr
768
460.733
5 x 1012
T1
303
377.572
5 x 1012
Zn
420
307.590
213.856
2.3 x 1014
1.2 x 1011
‘Values from Lange’s Handbook of Chemistry, 13th Ed., J.A. Dean, Ed., McGraw-Hill,
New York, 1985.
bValues calculated by Ramee Indralingam from Equation 2-2.

45
Calculation of Spectral Linewidths and Absorbances for Rb
The two most important features for use of the MVF are the absorption
coefficients (or absorption cross sections) and spectral profiles for the atomic
transitions at 780.023 and 794.760 nm. These properties give an idea of the required
specifications of the laser to be used in conjunction with the MVF in SMD. In order
to quantify these two parameters of interest, some of the theory of line broadening
will be given.
Atomic spectral profiles. The emission or absorption spectral profile of a
single, stationary atom contained in a vacuum absent of electric and magnetic fields
would have a spectral linewidth determined by the lifetime of the electronic
transition as stated by the Heisenberg uncertainty principle.39 Typically, atomic
transition lifetimes are on the order of 10'8 s which corresponds to a natural, or
fundamental, linewidth of 10 MHz or 0.021 pm at 800 nm. However, in a real system
there are several factors, such as motion of the atom, atomic collisions, and presence
of electric fields, which act to broaden the spectral profile for a given transition.
Furthermore, fine and hyperfine structure exist due to quantum splitting of electronic
transitions and existence of isotopes for a particular element. These components
each have an individual, quantifiable effect, and after they have been factored
together, the overall peak shape, termed the Voigt spectral profile, can be calculated.
Rubidium has two main isotopes, Rb85 and Rb87, which exist naturally in the
ratio 2.59:1 for Rb85:Rb87. The fine structure for transitions of these isotopes and
their f¡j values are given in Table 5 at 780.023 and 794.760 nm.59

46
Table 5. Physical parameters of the rubidium hyperfine structure required for
the calculation of the Voigt spectral profile of the metal vapor filter.
For the Absorption Band at 780.023 nm (yn = 12.820 cm'h
for Rb87
for Rb85
V-Pq
(cm1)
fij
P-Po
(cm1)
u
0.1323
0.0417
0.0553
0.0833
0.1347
0.104
0.0563
0.108
0.1401
0.104
0.0585
0.0864
-0.0933
0.0208
-0.0451
0.0309
-0.879
0.104
-0.0430
0.108
-0.790
0.292
-0.0389
0.250
For the Absorption Band at 794.760 nm (yn = 12.582 cm'1!:
for Rb87
for Rb85
p-Po
(cm1)
u
p-po
(cm1)
fij
0.1255
0.0208
0.0521
0.0309
0.1527
0.104
0.0642
0.108
-0.1025
0.104
-0.0493
0.108
-0.0753
0.104
-0.0372
0.0864
Each of these lines is broadened by the same extent as described below and
their normalized absorbances are added together at individual wavelengths to form
the Voigt profile.

47
Lorentzian broadening. There are two types of broadening taking place in the
MVF which factor into the Voigt profile, namely, Doppler broadening and collisional
(or pressure) broadening. The different types of collisional broadening can be
grouped together to produce the total Lorentzian profile. In the case of the MVF,
it is realistic to assume that all collisions of the atoms in the vapor state occur with
diatomic nitrogen (N2) and are adiabatic in nature (the atom remains in the same
electronic state during the collision). Thus, the Lorentzian linewidth, AvL (Hz), can
be calculated from the equation to determine the profile due to adiabatic collisions,
which is58
1/2
(2-3)
where R is the gas constant (8.314 J/K mol), T is the temperature (K), u is reduced
mass of Rb and N2 (kg), ac is the collisional cross-section of Rb (m2), and nx is the
number density of nitrogen in the metal vapor cell (m'3; in the case of nitrogen, nx
= 9.74 x 1024 P/T, with pressure, P, in Torr and temperature, T, in K). For Rb in
a nitrogen environment, ac has been measured to be 2.49 x 10"19 m2 at 780.023 nm
and 2.30 x 10'19 m2 at 794.76 nm.58
Doppler broadening. Doppler broadening, Apd (Hz), a result of the atoms
moving at different velocities upon absorption of light, can be determined from39
2a n
A C X
Apl =
2RT
TTU

48
1/2
(2-4)
y
where R is the gas constant (8.314 J/K), M is the formal weight of Rb (kg/mol), and
vm is the center frequency of the overall transition (Hz).
The a-parameter. To determine the proportion of each broadening effect in
the overall linewidth, the "a-parameter" is used, where, a = 0.83(A^L/A>/D). In the
case of the Rb metal vapor cell at pressures > 75 Torr, the effect of collisional
broadening is greater than that of Doppler broadening, which is expressed by a > 2.
The Voigt integral value, 5(a,0), at the line centers of both Rb resonance lines can
be calculated (within 10%) to be 5(a,0) « 0.56/a (for a > 2).39
Voigt linewidth. The multiplication of 5(a,0) by the Doppler shape function,
S,D (s), results in the Voigt shape function, S,v (s). For Doppler broadening, S,D at
frequency v (Hz) can be determined from the relationship,39
*1 1/2
]n2 (2-5)
ir
and the Voigt linewidth, Avy (Hz), is found from
A,d =
2v_
2(ln2)RT
M
AXy
1/2
(
Av
~2
-)2+Av
2
D
(2-6)
All linewidths denoted by the subscript v correspond to the full width of the
peak at half of the maximum intensity. To convert any of these linewidths from
frequency, Av (Hz), to wavelength, AX (nm), the value is multiplied by wavelength

49
squared and divided by the velocity of light, c (3 x 1017 nm/s). For example, A\v at
800 nm is found from the relationship, Aj/v(800 nm)2/c.39
Rb number density. At this point, all of the parameters, except n¡ and (, have
been defined that are necessary to determine the absorbance, as described by
Equation 2-2, and the Voigt spectral shape function, as described by Equation 2-5.
For the metal vapor cell Í is fixed and n¡ is dependent on temperature and pressure.
For the evaporation of liquid Rb, n, follows the expression,60
log(n¡) = -A/T - (B+ l)log(T) + C + DT +18.985, (2-7)
where A = 4529.6, B = 2.991, C = 15.8825, and D = 0.00059 with T in K. There
have been many different measurements of the values of these constants by physicists
with variability in n¡ as great as 25%, but the above values were chosen from
reference 60 because of their close agreement with other referenced values61 and
high precision.
Voigt spectral profile. The overall absorption line shape of Rb vapor was
calculated with the help of a computer spreadsheet and a computer program (written
in basic language by Michael Wensing). The program summed the absorbances of
the individual components of the profile versus wavelength based on values
calculated from the above equations in spreadsheet format. Plots of the Voigt
profiles at different temperatures and pressures were then generated (with Í of 4.4
cm). These plots are exhibited in Figures 4 and 5. As the figures show, increasing
cell temperature has a large effect on the absorbance, which essentially mirrors the
increasing Rb vapor number density, and increasing pressure increases linewidth at

50
the expense of absorbance. The presence of more than one peak at low pressure
exhibits the fine structure of the overall transition.
For this work, two different Rb metal vapor cells were available for use in the
project. Both cells contained 500 mg of 99.99% pure rubidium metal in nitrogen
(this is more than enough to produce a large vapor number density without
expending all of the condensed metal). One cell (to be referred as cell #1),
manufactured by Rudy Strohschein in the University of Florida glass shop, has an
absorption path length of 4.4 cm and a pressure of 200 Torr at room temperature.
The other cell (cell #2, manufactured by Opthos, Rockville, MD) has a 4.7 cm path
length and a room temperature pressure of 500 Torr. These parameters were
entered into the computer programs to calculate the Voigt profiles for these cells at
100°C for both the 780.023 nm and 794.760 nm lines. These profiles appear in
Figures 6 and 7. The Voigt linewidths are 21 pm (at 780 nm) and 20 pm (at 794.76
nm) for cell #1, and 36 pm (at 780 nm) and 36 pm (at 794.76 nm) for cell #2. Cell
#1 is capable of a larger absorbance at constant temperature than cell #2, which
makes cell #1 preferential for use in the experiment provided that the laser emission
linewidth is less than 21 pm.
Tye Barber41 experimentally verified the Voigt profiles generated from
calculations using a single-mode diode laser with cell #1. He found the absorption
bandwidth to be 21 pm at 780 nm which is in excellent agreement with the calculated
bandwidth. The shape and intensity of the absorption peak also closely agreed with
the computer generated plot at the same temperature.41

200
180
160
140
120
100
80
60
40
20
0
Ca
4.4
¡0 -40 -30 -20 -10 0
Wavelength Shift (pm)
mlated Voigt absorption profiles of the 794.76 nm Rb line at different temperatures. Path length is
cm and N2 pressure is 200 Torr at 25°C.

45-
40-
35-
30
25-
20
15
10
5
0;
ilculated Voigt absorption profiles of the 794.76 nm
and path length is 4.7 cm.
Rb line at different pressures. Temperature is 373
L/l
K)

Figure 6. Calculated Voigt absorption profiles at 780.023 nm and 100°C for the Rb cells to be used in the SMD
project. Linewidths for cells #1 and #2 are 21 pm and 36 pm, respectively.
Ln
OJ

Absorbance
Figure 7. Calculated Voigt absorption profiles at 794.76 nm and 100°C for the Rb cells to be used in the SMD
project. Linewidths for cells #1 and #2 are 20 pm and 35 pm, respectively.

55
The Laser
Instrumentation for fluorometric analysis. To perform molecular fluorescence
analysis, five basic components are necessary: 1) a source of light, 2) a fluorescent
sample, 3) a container for the sample, 4) a spectral filter, and 5) a detector. In the
SMD project, the MVF, which is able to remove the source light while still allowing
the passage of the fluorescence to the detector, serves as the fourth component listed.
The excitation source is the subject of this section, and each of the remaining
components will be discussed individually in the following sections.
Light sources. For selective analysis, the excitation source should have a
narrow emission band, and for improved sensitivity, the source should have high
emission intensity. Arc lamp sources are capable of very large radiances, but in
order to obtain narrow bandwidth, the light must be passed through a
monochromator which filters the excitation source into less intense bands. With the
invention of the laser, a source became available that was capable of high power in
a narrow spectral bandwidth. However, a problem with the laser as an excitation
source is its lack of tunability. Only those analytes with an excitation band at the
laser emission wavelength can be determined by LIF. Of course, there are many
different lasers with many different laser lines, but the cost and practicality of using
so many different sources makes LIF an analytical technique for unique
circumstances.

56
Criteria of the Laser for Single Molecule Detection
The choice of laser to be used in the detection of single molecules must be
made based on several criteria. These criteria include:
(1) the laser must be capable of sufficient irradiance (W/cm2) at the excitation
wavelength of a given fluorophore to attain SMD;
(2) the laser emission wavelength must fall at or be able to be tuned to the
absorption wavelength of the metal vapor cell and remain at this wavelength
over time;
(3) the emission profile of the laser must be Gaussian-shaped with a linewidth
less than the absorption bandwidth of the metal vapor;
(4) the source must be continuous or have a pulse rate of at least * 10 kHz in
order to efficiently probe a flowing sample.
Other desirable traits of the source include low cost, simple maintenance, and easy
operation.
Choosing the laser. Most lasers meet criterion (1) if the beam can be focused
to a very small area. Because no non-tunable lasers happen to emit light at an
absorption band of an element for practical use in a MVF, criterion (2) establishes
that a tunable laser must be used. Only dye lasers, diode lasers, Raman-shifted
lasers, and the Tirsapphire laser are tunable. Requirement (3) negates Raman-
shifted lasers and all pulsed dye lasers except the copper vapor-dye laser system.
Criterion (4) precludes the use of a pulsed laser unless it is capable of a repetition
rate greater than * 10 kHz. The upper limit of the repetition rate of the Cu vapor-

57
dye laser system is 10 kHz, thus single-mode diode lasers and Tirsapphire laser have
the best characteristics for SMD.
Diode lasers. Diode lasers are solid state electronic emitters of
electromagnetic radiation much like light-emitting diodes except the diode laser
possesses a cavity to induce lasing.62 The lasing material is usually gallium doped
arsenide which has an energy gap that corresponds to wavelengths in the red to near-
infrared wavelengths. Diode lasers have several advantages over conventional lasers
that make them generally accepted among spectroscopists as the "laser of the future."
The favorable characteristics62 of diode lasers include: (1) inexpensive, (2) easy to
operate, (3) maintenance free, (4) long-lived, (5) robust, (6) small, (7) tunable, (8)
powerful, (9) efficient, (10) versatile, (11) high power stability, and (12) narrow
spectral linewidth.
Despite these advantages, diode lasers will remain the "laser of the future"
unless the following problems can be corrected. Firstly, diode lasers are tunable, but
not continuously tunable allowing coverage of all wavelengths in the tuning range.
Diode lasers commonly exhibit mode hopping which is the tendency of the laser to
jump instantaneously from one wavelength to another and exhibit lack of tunability
in the region between these modes. Secondly, since diode lasers have a small laser
cavity, the laser beam has a large beam divergence which makes focusing difficult.
Lastly, diode lasers are available over a limited wavelength range of 650-1300 nm
with each laser tunable over a 20-30 nm range. If the laser had a low divergence, it
would not be difficult to frequency double the emission to wavelengths half of the

58
fundamental output utilizing nonlinear optical properties of certain crystals. This
would enable diode lasers to be capable of emission from 325-1300 nm, but the poor
doubling efficiency currently obtained results in laser powers too low for general
applications.63
However, it is fair to say that the current intensive research in the area of
diode lasers in the brief time since their invention has resulted in more powerful
lasers, wider range of lasing wavelengths, more efficient single-mode operation, and
lower cost which will soon result in a laser useful for SMD. Already, diode lasers
can be used in conjunction with a MVF.41 Special electronic control of the diode
laser current with active feedback virtually eliminates mode hopping, but the low
power and focusing difficulties of these lasers does not yet meet all of the criteria
listed earlier. Undoubtedly, diode laser excitation would be the preferential
approach for SMD due to the many advantages of diode lasers, but until the above
problems are solved, another laser must be used.
The Titanium:Sapphire Laser
Of all possible sources currently available, only the titanium:sapphire
(Ti:Al203) laser meets all of the outlined criteria. Table 6 lists the specifications of
the manufacturer of the Ti:sapphire laser purchased for use in this project. As the
table shows, the laser is capable of two lasing arrangements, the standing-wave and
ring cavity configurations. The specifications for both configurations are the same
except that the spectral linewidth is narrower in the ring configuration.

59
Figure 8 is a drawing of the design of the Tirsapphire laser. The diagram
shows the laser in the ring cavity arrangement with the beam passing through the
optical diode. In the standing-wave configuration, the two flat mirrors are tilted to
return the beam to the curved mirrors on either side of the crystal instead of through
the optical diode.
Specifications of the titaniumrsapphire laser used in the single
molecule detection project.
Table 6.
Parameter
Maximum Power
(at 800 nm)
Tuning Range
Spectral Linewidth
(at 5 W Pump Power)
Spectral Profile
Spatial Profile
Polarization
Beam Size
Specified Value
750 mW
(10% of pump power up to 7.5 W)
700-1000 nm
< 2 GHz standing wave
<40 MHz ring configuration
Gaussian
TEMoo
Horizontal
1 mm at exit
The Ti:sapphire laser is a passive device with no moving or electronic parts.
An important aspect of the Tirsapphire laser is that it requires the services of a pump
laser to initiate the lasing action. The source of the lasing is a solid state A1203
crystal doped with titanium. The crystal has broad excitation and emission bands
with maxima * 530 nm and 800 nm, respectively.64 Due to peak absorption in the
green part of the spectrum, argon ion, Nd-YAG, and copper vapor lasers are used
to pump the Tirsapphire laser. The Ar+ laser is used most often to take advantage

Figure 8. The optical arrangement of the Ti:sapphire laser used in this project.

61
of its continuous laser action, but pulsed operation can be applicable in many
situations. The output power of the Ti:sapphire is proportional to the power of the
pump laser and, according to the manufacturer specifications, is capable of 10%
efficiency of conversion, but in the SMD experiments, nearly 20% efficiency was
attained with careful alignment.
Comparison of the Thsapphire laser with dye lasers. The Ti:sapphire laser
was invented in the 1980s and is poised to replace the dye laser for many
applications. In the early 1970s, dye lasers were hailed as the solution to the lack of
laser tunability. There are several problems with dye lasers, however, concerning
both scientific and practical aspects in their use. Each dye has a limited tuning range
based on concentration, solvent, pH, and pump laser. Power is strongly dependent
on wavelength and is greatly changed when a dye laser is scanned; efficiency of
conversion is often less than 1% for dye lasers. For these reasons, dye lasers are
seldom used to scan over a wavelength range, and the more common application is
to tune a dye laser to a fixed wavelength. In practice, dye lasers are noted for the
problems: dye decomposition, solvent flammability, chemical toxicity, waste disposal,
chemical spills, and extended down time (often due to solvent pumps). For these
reasons, spectroscopists have been anxious for a laser to replace the dye laser.
The Ti:sapphire laser has many advantages over the dye laser, but is not the
panacea for the laser spectroscopist. Since the Ti:sapphire is a passive, solid state
laser, it has none of the practical problems associated with using laser dyes. No
moving parts translates to no down time for mechanical reasons, and no dyes means

62
no messes, chemical hazards, or waste. Also, the Ti:sapphire laser is capable of
continuous tuning with moderate change in power over a continuous range of 120 nm
(fundamental) or 60 nm (frequency doubled) before requiring a change of optics.
Few individual dyes can match this feat; however, the overall wavelength coverage
of 350-500 nm and 700-1000 nm is still inferior to a network of dyes that allows for
coverage of the spectrum from the ultraviolet to the near-infrared. Therefore, the
Tksapphire laser is not able to replace the dye laser in all applications.
The TkSapphire Laser/Rb Metal Vapor Filter Combination
The stated linewidth specifications of < 2 GHz in the Ti: sapphire standing-
wave configuration and < 40 MHz in the ring cavity configuration corresponds to <
4.1 pm and < 0.081 pm, respectively, at the 780 nm Rb transition. For the 794.76
nm transition, these linewidths are < 4.2 pm and < 0.084 pm, respectively. Due to
the Gaussian nature of the spectral profile for the laser as opposed to the Lorentzian
wings of the Voigt profile, the laser beam for either laser configuration should be
virtually totally absorbed by the Rb metal vapor cell at temperatures > 100°C. This
aspect is shown in Figure 9 in which the emission profiles of the Tksapphire in both
configurations are superimposed onto the Voigt profile of cell #1 at 794.76 nm. As
shown, the specified laser linewidths are much narrower than the profile of the
absorption line. This condition is especially notable in the ring cavity profile where
the laser line is a thin line on the same scale as the other profiles.

Figure 9. Calculated spectral line profiles of the Ti:sapphire laser in the standing-wave and ring configurations at
794.76 nm in comparison with the Rb Voigt profile of cell #1 at 100°C.
Laser Intensity

64
Absorbance is the log of the inverse of transmittance, T, (A = log(l/T))
which means that the absorbance of 10 for the Rb metal vapor filter at 100°C
corresponds to a transmittance of 1010, or 99.99999999% of the light is absorbed.
Therefore, with a metal vapor absorbance of 10, the calculated maximum percentage
of light scattered by the capillary is 5.9% (if it is an isotropic point source at the
focus of the microscope objective) that would still allow for an LOD = 1 molecule
based on the parameters listed in Table 2 of Chapter 1. In other words, the laser
scatter collected in the 1.5 sr solid angle of collection (fiF/4x) must be limited to less
than 0.7% (5.9% x 11.9%) to achieve a background count rate of 56 counts per 2 ms
increment (maximum theoretical ¿ibl possible in order to attain SMD for Xd = 81
counts. These calculations are based on the discussion given in Chapter 1.
It should be pointed out that the Tirsapphire laser/Rb MVF combination has
been used previously to perform Raman spectroscopy.65,66 As in LIF, the limiting
source of noise in Raman spectroscopy is typically laser specular scatter. With a
similar laser/MVF set up to be used here, Raman spectra were obtained that showed
no evidence of the laser despite viewing at the laser wavelength.65,66 This is a major
accomplishment in Raman spectroscopic analysis because it is now possible to
measure Raman bands very close to the excitation wavelength. Furthermore, the
conditions for optimum performance of Raman spectroscopy are very similar to those
of the SMD; since the Ti:sapphire laser and Rb MVF combination was successful in
that application, there was little reason to believe it would not be successful in
detecting single molecules by LIF.

65
Tksapphire laser configuration. Because the linewidth in either the standing-
wave configuration or ring cavity configuration is much narrower than the Voigt
profile of the Rb vapor cell, it does not matter which laser configuration is used for
single molecule detection. Both configurations were tested for use in the project with
no significant difference in results.
The Sample
After the excitation source, the sample constitutes the second of the five basic
components of a fluorometer. In the case presented so far, the analyte must absorb
strongly at 780 or 795 nm (Rb ground state electronic transitions) and fluoresce with
high quantum efficiency at removed wavelengths. Furthermore, as discussed in
Chapter 1, the analyte should not be a polymer or other chemical species with a
molecular weight greater than « 1000 g/mol. Another consideration is the solubilty
of the molecular species because this analysis is to be attempted in solution phase.
Choice of Analyte
Polvmethine dves. Very few chemicals fluoresce beyond 700 nm. In fact, the
only known molecules to fluoresce strongly in the far red part of the spectrum are
the cyanine dyes known as polymethines.67 Table 7 is a list of polymethine dyes that
includes their common names, chemical formulas, excitation and emission maxima,
and fluorescence quantum yields of those available in the literature. As expected for
fluorescent molecules, these dyes have highly conjugated, semi-symmetrical structures

66
Table 7: Fluorescent dyes for possible use in the single molecule detection project.
Dye
Molecular
Formula
K
(run)
^cm
(nm)
max
eA
(M^cm1)
yf
Rhodamine 800
q6h26n3o Cl
682*
700*
89,500*
0.39*
Methylene Blue
C16H18N3S Cl
668*
683*
66,600*
—
Nile Blue
QoH^O Cl
640*
672*
77,500*
—
Oxazine 750
C^H^NjO Cl
673*
691*
82,500*
—
IR 125
C43H47N2S2 Na
780
806
150,000
0.13‘
IR 132
c52h48n3o4s2 C104
810
846
210,000
—
IR 140
C38H34C12N3S2 cio4
800
833
180,000
1.0d
IR 144
CjoHj8N4OgS2 NEt3
698b
oo
o
127,000
—
DTTC
C25H25N2S21
746b
777b
—
0.38‘
DTDC
C23H23N2S21
647b
668b
—
0.73‘
DOTC
C25H25N202 1
678b
703b
—
0.63‘
HITC
I
736b
764b
240,000
0.28‘
HDITC
^36^37N2 C104
77 lb
805b
—
—
DDTC
C32H29N2
765‘
855‘
—
0.16°
DQDC
^27^27^
765‘
835°
—
.001°
DQTC
^29^29^2
825‘
865‘
—
.035‘
"Values in water from T. Imasaka, A. Tsukamoto and N. Ishibashi, Anal. Chem., 61, 2285
(1989).
bValues in methanol from D. Andrews-Wilberforce and G. Patonay, Appl. Spectrosc., 43,
1450 (1989).
‘Values in dimethylsulfoxide from R.C. Benson and H.A. Kues, /. Chem. Eng. Data, 22,379
(1977).
dValue in ethanol from D.J.S. Birch, G. Hungerfold, B. Nadolsi, R.E. Imhof and A.D.
Dutch, J. Phys. E: Sci. Instrum., 21, 857 (1988).
All other values determined experimentally in methanol.

67
containing several aromatic rings with the length of the structures correlating to
excitation and emission wavelengths. In fact, Benson and Kues68 have worked out
an empirical relationship between the structure of polymethine dyes and their
fluorescent characteristics.
Of the dyes listed in Table 7, only three, IR 125, IR 140, and IR 132, are
potentially useful for single molecule detection with excitation at 780 or 794.76 nm.
Figure 10 gives the structures of these polymethine dyes. It is rather fortuitous that
Xexn“x of IR 125 falls at the 780 nm Rb line, and that IR 140 has X^”4* at 800 which
is very near the 794.76 nm transition. Meanwhile, IR 132 is the least favorable of
these dyes in this respect with Xex““x of 810 nm.
Another important parameter for SMD is the absorption cross-section, aA
(cm2), of the analyte which is proportional to the molar absorptivity, eA, of the dye
in bulk solution as shown in Chapter 1. In a simple experiment using Hewlett-
Packard 8450A and Varían 634 (for wavelengths greater than 800 nm)
spectrophotometers, the absorption spectra of known concentrations of the dyes in
methanol were measured. The wavelengths of maximum absorption corresponded
to the fluorescence excitation maxima for the dyes, and the measured ea”** for IR
125, IR 140, and IR 132 were 1.5 x 105, 2.1 x 10s, and 1.5 x 105 M'W'1, respectively.
These figures are in agreement with values reported in the literature.69
The emission spectra of the dyes in methanol solution are shown in Figure 11
with excitation at 780 and 795 nm. For this experiment, the Ti:sapphire laser was
used as the excitation source of the flowing dye solutions contained in a 1 cm path

68
Figure 10. Chemical structures of the polymethine dyes to be tested for use in
the single molecule detection project.
length cuvette. A Spex 1680 double monochromator with 1 mm slits was used to
collect the spectra and a cooled Hamamatsu R636 photomultiplier tube served as the
detector. These spectra have been normalized to a laser power of 145 mW and dye
concentration of 4.6 x 10'7 M. The Xem maxima for IR 125, IR 140, and IR 132 with

600
500
400
300
200
100
0
fluorescence emission spectra of the polymethine dyes in methanol with excitation at the Rb lines,
spectra have been normalized to 145 mW laser power and 4.6 x 10 7 M dye concentration.
O-N

70
this system were 806, 833, and 843 nm, respectively. Each of these dyes was tested
for application to the single molecule detection project as presented in Chapter 3.
More detailed characteristics of each dye will be discussed separately below.
IR 125. The formal chemical name for IR 125 is anhydro-l,l-dimethyl-2-[7-
[l,l-dimethyl-3-(4-sulfobutyl0-2-(lH)-benz(e)indolinylidene]-l,3,5-heptatrienyl]-3-(4-
sulfobutyl)-lH-benz(e)indolinium hydroxide sodium salt; its Chemical Abstracts
Services (CAS) number is [3599-32-4]. The compound is more commonly known by
many chemists as indocyanine green, or ICG, which is associated with its use as an
indicator. Laser spectroscopists, on the other hand, are more familiar with the name
IR 125 which is associated with its use as a laser dye. IR 125 is the only water
soluble dye of the three species, and is also readily soluble in most organic solvents.
In water, IR 125 has a pK, of 3.27.70 Like the other polymethine dyes, IR 125 is a
zwitterionic salt due to the presence of an aromatic heterocyclic ring containing
nitrogen.
Applications. IR 125 is the most widely used of the mentioned polymethine
dyes, mainly due to its solubility in water. By far, the most common uses of IR 125
are as a laser dye71 and as a clinical indicator dye for testing of in vivo blood flows
and hepatic functions in animals and humans.72 It is useful in clinical applications
due to its large molar absorptivity at long wavelengths where blood does not absorb
strongly. It has also been used in angiography73 and many other studies in blood.72
A very interesting aspect of IR 125 for future applications is that it has been bound
to surfactants74 and proteins75 for analytical purposes. Patonay at Georgia State

71
University is actively pursuing the use of IR 125 and other polymethine dyes for
tagging purposes.67 If his research is successful, this SMD technique could become
very important in many tagging applications.
Indirect fluorometric detection. Another interesting use of IR 125 concerns
its chromatographic properties. It has been analyzed by high-performance liquid
chromatography (HPLC) for the clinical applications previously mentioned,76"78 and
based on these studies, it was thought to be an excellent choice as a visualization
agent for indirect fluorometric detection in HPLC79 Indirect detection works on the
principle that by monitoring the concentration of a continuously present indicator
species, termed the visualization agent, the presence of other species can be
determined by fluctuations in the concentration of the visualization agent.80 In this
manner, it is a universal method of detection with detection limits based on three
factors: 1) the size of the effect of the analyte on the signal of the visualization
agent (known as the transfer ratio); 2) the ability to measure these signal fluctuations
(or dynamic reserve); and 3) the concentration of the visualization agent.79 In HPLC,
ion exchange chromatography and capillary electrophoresis, the magnitude of the
transfer ratio can be very large based on separation properties, and IR 125 is a good
choice for this purpose because it can be detected very selectively and sensitively due
to its fluorescence at long wavelengths. A diode laser is suitable as the excitation
source due to its very high power stability which greatly reduces noise on a large,
constant signal. Indirect fluorometric detection with diode laser excitation of the IR
125 visualization agent was used in the detection of alcohols using reversed-phase

72
HPLC,79 and current work is underway to use indirect fluorometric detection in
capillary electrophoresis with the system. If successful, it will be possible to obtain
a LOD of 10 pM for an analyte with a transfer ratio near unity.79 Indirect
fluorometric detection is of interest to separation scientists because it has the
potential of being a sensitive and universal detector which is a rare combination.
IR 140. IR 140 is formally known as 5,5’-dichloro-ll-diphenylamino-10,12-
ethylenethiatricarbocyanine perchlorate; its CAS number is [53655-17-7]. The major
use of IR 140 is as a laser dye,71 but it has also been used as the analyte in several
LIF experiments with diode laser excitation.56,57,63'81 The goal of these projects was
very much similar as this one, which was to attain the smallest possible limit of
detection of the dye in a flowing stream. Based on these earlier experiments with
this dye and the very high fluorescence quantum efficiency (YF = 1) and a high
spontaneous emission (A21 = 1.26 x 109 s'1 based on 791 ps fluorescence lifetime,
rF),42 IR 140 is emphasized in the experimental studies.
IR 132. Like IR 140, the only known use of IR 132 is as a laser dye.71 It’s
formal name is 3,3’-di(3-acetoxypropyl)-ll-diphenylamino-10,12-ethylene-5,6,5’,6’-
dibenzothiatricarbocyanine perchlorate (CAS # [62669-62-9]).
With the advent of diode lasers, these dyes have become very important as
probe species. Several analytical chemists have used these dyes in an attempt to
apply diode laser source to analytical techniques.81 Since diode lasers are only useful
at wavelengths longer than 650 nm, and polymethine dyes are one of the few
molecules fluorescent at these wavelengths, they have been thrust into several

73
applications.81 In the SMD project, the optimum dye is to be chosen based on the
conditions of the excitation wavelength and filtering to be discussed later in this
chapter and the next.
Stability of the Dyes. Due to the highly conjugated structures of these dyes,
it was anticipated that these dyes decompose readily. There have been numerous
studies involving the decomposition of IR 125 in blood plasma, water, and electrolytic
solutions, and indeed, IR 125 degrades in a matter of hours in these solutions.82
However, the decomposition rate is vastly reduced in organic solvents. A study was
performed to determine whether decomposition of the dyes would be a problem for
the single molecule detection project. Figure 12 shows the fluorescence of the dyes
in methanol over a period of 25 days. These measurements were made with a Spex
Fluorolog 2 spectrofluorometer with 1 mm slits and a cooled Hamamatsu R928
photomultiplier tube detector. The Xcx was set to 764 nm because the 500 W Xe arc
lamp used as the excitation source has a greater emission intensity (and produces a
larger signal) at this wavelength than at the wavelengths of eAmax of the dyes. Stock
dye solutions of * 1 x 10 5 M were kept at room temperature in the dark. Each time
a spectrum was taken, a 0.1 mL aliquot of solution was pipetted into a 1 cm path
length cuvette and diluted with 3 mL of solvent. As the figure shows, there was quite
some day-to-day variation in the procedure, but the overall fluorescence did not
decrease significantly. Therefore, the dyes do not readily decompose in organic
solvent (assuming that the degradation products do not fluoresce at the same
wavelengths as the dyes).

Figure 12. Stability of the polymethine dye stock solutions in methanol kept in the dark over a period of 25 days.
The variations in the signal is due to daily instrumental and pipetting deviations.
-o

75
Effect of degassing the solutions. In a similar study utilizing the same
instrument, the effect of bubbling different gases on the photodecomposition of the
dye solutions in methanol was examined. In this study, «0.1 /¿M dye solutions were
prepared, split into 3 volumes, and continuously sparged with nitrogen, helium, and
air. Their fluorescence emission spectra were measured after » 1 minute. These
solutions were then placed in a chamber radiated by light of wavelengths greater than
700 nm. For this purpose, a 500 W Eimac arc lamp was placed in the chamber, and
its white light emission was filtered by a series of long pass spectral filters. The
emission spectra of 3 mL aliquots of these solutions were measured periodically over
a period of 2 hours. Figure 13 shows the effect of degassing on the
photodecomposition rate of IR 140 in methanol. This figure clearly shows that air,
or more specifically oxygen, causes increased dye degradation as opposed to an
oxygen free solution. Figure 13 further shows that nitrogen and helium greatly
reduced the photodecomposition with respect to air, therefore, all solutions used in
the single molecule detection project should be sparged with one of these two gases.
Based on cost and ease of access, N2 was chosen for this purpose.
The Choice of Solvent
Since the goal of this project was to achieve single molecule detection with no
particular sample type in mind other than a solution, the choice of solvent was based
mainly on signal to noise of the dye fluorescence in that solvent. Of course, water
would be the ideal solvent due to its great applicability to many types of analyses, but

Signal
Wavelength (nm)
Figure 13. Stability of IR 140 in methanol in the presence of different gases over a two hour period under near¬
infrarad light.
as

77
since IR 140 and IR 132 are water insoluble, and IR 125 decomposes rapidly in
water,82 H20 is not one of the choices as solvent for this experiment. Instead, several
organic solvents covering a wide range of properties, were tested.
Table 8 is a list of six solvents, and their pertinent characteristics. Of the
enormous number of organic solvents, these six were singled out due to their general
availability and history of use in fluorescence analyses. Polarity tends to be the most
important factor in the fluorescence intensity and maxima emission and excitation
wavelengths. For most fluorescent solutes, nonpolar solvents, such as hexane, tend
to shift the spectra to longer wavelengths and reduce intensity. Nonpolar solvents
are useful when a red-shift is desirable, or when the analyte is soluble only in
nonpolar solutions.
Table 8. Physical properties of the solvents tested for use in this project.
Solvent
Formula
Density
n*
eb
Acetone
CH3COCH3
0.7908
1.3588
20.7
Acetonitrile
CH3CN
0.840
1.3420
37.5
Dimethylsulfoxide
CH3SOCH3
1.100
1.4783
46.6
Ethanol
QHsOH
0.7894
1.3614
24.55
«-Hexane
CH3(CH2)4CH3
0.6594
1.3749
1.89
Methanol
CH3OH
0.7913
1.3284
32.70
Water
h2o
1.0000
1.3330
80.10
an is refractive index at 589 nm; be is dielectric constant
All values for 20°C from Lange’s Handbook of Chemistry, 13th Ed., J.A. Dean, Ed.,
McGraw-Hill, New York, 1985.

78
The emission spectra of the IR dyes were taken with the laser tuned to 780
and 795 nm using the same instrumental system and normalized to the same
conditions as described in Figure 11 for methanol. The results of these spectra with
Xex = 795 nm are compiled in Figure 14 which shows the normalized peak emission
intensities of IR 125, IR 140, and IR 132. Acetone and acetonitrile consistently yield
the largest fluorescence signal for the dyes whereas the nonpolar solvents, hexane
and dimethylsulfoxide produce less intense fluorescence. This is a typical behavior
for fluorescent compounds.
Raman spectra of the solvents. Based on Figure 14, acetone or acetonitrile
would be the best choices for use in the SMD project. However, signal is not the
only criterion on which the selection of solvent is based. Signal to noise ratio is the
most important parameter in any analysis to minimize limits of detection. As
discussed in Chapter 1, Raman scatter is the second most severe source of noise after
laser specular scatter, and if the MVF removes the laser scatter as it should, Raman
scatter from the solvent will then be the limiting source of noise. Since Raman
scatter is shifted in wavelength from that of the laser, it is transmitted through the
MVF to the detector. Therefore, it must be reduced at the source, by the reduction
of Vp and choice of solvent, or filtered by a spectral filter other than the MVF.
Figure 15 contains the Raman spectra of the 4 solvents that gave the highest
fluorescence signals in Figure 14. These spectra were taken under the same
conditions with the Spex Raman microprobe instrument utilizing an Ar+ laser source
and cooled RCA C31034 detector. In Figure 15, the Raman shift in cm'1 from the

Normalized Peak Signal
Figure 14. The effect of different solvents on the normalized maximum fluorescence intensity of the polymethine dyes
with Ti:sapphire laser excitation at 794.76 nm (ACN = acetonitrile; MeOH = methanol; EtOH = ethanol;
DMSO = dimethylsulfoxide).

Signal
Wavelength (nm)
Figure 15. Raman spectra of four solvents converted to excitation at 794.76 nm.
00
©

81
514.5 nm Ar+ laser line has been converted to wavelength as if the excitation
occurred at 794.76 nm. By referring to the fluorescence of the dyes in Figure 11, one
can see that the Raman spectra of the solvents overlap with the fluorescence signal
of the dyes. Only methanol is devoid of Raman peaks at wavelengths less than 860
nm. Assuming Raman scatter is the limiting source of noise, methanol is then the
best choice of solvent, despite the slightly lower intensity of the IR dyes in methanol
than in acetonitrile and acetone. These solvents give several very intense Raman
peaks around 820 nm.
To ensure that the extension of the Ar+ excited Raman spectra to the
Ti:sapphire was valid, the spectra of methanol and acetonitrile were measured in a
1 cm square flow cell with Tirsapphire laser excitation and the Spex 1680 double
monochromator/Hamamatsu R636 PMT combination. Figure 16 shows the resulting
spectra from this analysis with excitation at 795 nm. The wavelengths of the peaks
correlates with the wavelengths shown in Figure 15. It is unknown why the
background for the methanol spectrum is higher than it is for the acetonitrile in this
case. In the case of the argon ion laser excitation, the high background for methanol
is believed to be fluorescence of an impurity.
Dimerism. Another reason to choose methanol over acetonitrile as the
solvent is that the polymethine dyes dimerize in acetonitrile and not in methanol.
Dimerization is the tendency of solute molecules to associate with each other in pairs
instead of exist individually in solution. In the literature, other researchers have

PMT Current (pA)
Wavelength (nm)
Figure 16. Raman spectra of methanol and acetonitrile with Ti:sapphire laser excitation at 794.76 nm.

83
noted this problem with polymethine dyes in certain solvents.67,69,83 Dimerism is
exposed in calibration curves of signal versus concentration. In acetonitrile, the
calibration curve was nonlinear, and when converted to log-log scale, the slope of the
line was 2. Figure 17 is a calibration curve of IR 125 in acetonitrile taken with
Ti:sapphire laser excitation at 795 nm; based on the log-log slope of 1.99,
dimerization is clearly shown in this figure. This situation is much like the analysis
of sulfur with a flame photometric detector in which signal is proportional to
(concentration)2 due to the production of S2 in the flame. This is an unacceptable
situation when calculating limits of detection based on the sensitivity of an analysis.
Methanol, on the other hand, produces linear calibration curves and log-log plots
with a slope near unity. This indicates that dimerism of the dyes does not occur in
methanol solutions, and that LOD calculations based on sensitivity can be made of
this dye/solvent system.
Sample Containment
In an ideal system, the method of sample containment should not perturb the
sample in any way or introduce noise into the measurement. Despite this seemingly
innocuous task, sample contact with the sample cell may cause problems such as
introduction of reaction surfaces and matrix interferences. In spectroscopy, light
transitions from one medium to another decreases the amount of light reaching the
sample due to scatter, fluorescence, and absorption by the containing medium.

PMT Signal (nA)
Figure 17. Log-log calibration curve of IR 125 in acetonitrile with 350 mW Tirsapphire laser excitation at 780 nm.
Flowing sample in flowcell using Spex 1680 monochromator at 806 nm with 5 mm slits (9 nm bandwidth)
and cooled R636 detection. Slope of 2 is a sign that dimerism is occurring.

85
Furthermore, laser specular scatter from the container constitutes the major source
of noise in most LIF analyses.
Liquid jets. Some of the problems with sample containers can be eliminated
by using a liquid jet, which is merely the term used for a flowing stream emanating
from a nozzle. The optimal probe region occurs at a point just after the orifice
where the flow stream narrows before it spreads again. Previous studies in this
group56,57 used a liquid jet in the analysis of IR 140 in methanol with diode laser
excitation. In these studies, noise due to laser scatter was greatly reduced by exciting
the stream outside of the capillary, and the sensitivity of the measurement was
increased because more of the laser light was reaching the sample. However, the
major problem with a liquid jet is that it only operates at high flow rates. At lower
flow rates, the sample drips from the orifice one drop at a time which also causes
measurement difficulties. Another problem is that residence time of the analyte in
the probe volume is < 1 ms at flow velocities of liquid jets. For SMD, the desired
residence time (tr) is in the range of 1-10 ms, depending on laser power and dye
decomposition.
Levitation. Ideally, the analytical chemist would like to ensure that the
sample is isolated from external factors and positioned in the probe region for a
length of time sufficient for complete analysis. The most obvious way to accomplish
this goal is through sample levitation. Means of sample levitation involve the use of
physical forces to counteract gravitational attraction, or as in the case of performing
analyses in outer space, the reduction of gravitational effects. Of course, since space-

86
based research in analytical chemistry is overly expensive and generally impractical,
the former means of levitation are far more common than the latter. Practical types
of sample levitation possible make use of aerodynamic, acoustic,84 photophoretic,85
electrodynamic,86 and magnetic forces.87 As discussed in Chapter 1, Ramsey’s
approach to SMD makes use of the advantages of electrodynamic levitation.43
Although sample introduction into the trap is complicated and time consuming, it
does remedy some problems related to a flowing sample contained in quartz. The
way in which Ramsey contains the sample is reviewed in the following paragraph in
order to give solid grounds for comparison with the other methods.
Electrodvnamic levitation. In electrodynamic levitation, a charged species
(atom, molecule, particle, or droplet) is introduced into a chamber containing a ring
electrode, to which is applied an ac (radiofrequency) potential, and top and bottom
electrodes to which are applied dc potentials.86 A field applied to the ring electrode
controls the lateral position of the charged species, and the dc potentials applied to
the top and bottom electrodes control the vertical motion. Figure 2 in Chapter 1
shows an electrodynamic trap in the drawing of Ramsey’s set-up. With samples
having high mass/charge ratios (> * 10s kg/C), such as the charged, micron-sized
droplets introduced into the chamber by Ramsey, the droplets can be stably levitated
at the center of the trap. Sample introduction is accomplished with a piezoelectric
droplet generator at the top electrode of the chamber; the electrode voltages are
then altered until the droplet is trapped.43 The size of each droplet must be
determined from the complicated analysis of scattering patterns of a HeNe laser

87
beam focused onto the droplet.43 Furthermore, the analyte must be contained in a
solution partially consisting of glycerin to reduce solvent evaporation under the laser.
After a few minutes required for introducing, trapping, and determining droplet
volume, the fluorescence analysis begins. Based on these aspects, one can see that
practical analysis of large volume samples would be tedious by this method and rapid
measurements in a flowing stream would be impossible.
Sheath flow cuvette. The approach to single molecule detection of Keller’s
group makes use of a sheath flow cuvette for sample containment.2'6 The sheath flow
cuvette is a type of flow cell in which an external solvent stream, or sheath, is used
to compress an internal sample stream.8 Figure 1 in Chapter 1 contains a partial
drawing of a sheath flow cuvette used in Keller’s set-up. The narrowing quartz walls
of the flow cell cause the sheath flow, originating from outside the internal capillary,
to compress the sample stream as it effuses from the internal capillary. The solvent
in the sheath is usually the same solvent as the sample in order to avoid changes in
refractive index, but use of different solvents has certain advantages in some
instances.
In the set-up at Los Alamos, the laser probes a region in the cuvette where
the quartz tube is square (flat surfaces do not scatter as much light as round
surfaces). In a situation much like the liquid jet, but not as severe, the flow rates
required to induce the desired narrowing of the sample stream, greatly decrease tr.
When the flow rate is lowered in an attempt to increase tn diffusion of the analyte
into the sheath occurs and the irradiance of the laser required to cover the entire

88
sample stream is not sufficient for SMD. To deal with this problem, Keller’s group
decided to avoid it altogether by lowering flow rate to a point where tr is adequate
for the detection of the single molecules as they flow through Vp,2,3,5 but the laser
probes only « 6% of the sample flow region (and the spatial filter only views a
portion of that region). This does not meet the 100% sampling efficiency
requirement of SMD expressed in Chapter 1.
The capillary. Unlike the other sample containment methods discussed, the
capillary is simply a glass or quartz tube requiring no instrumental adjustments.
Capillaries of various sizes are commonly used in several types of analytical methods.
In separation techniques using capillaries (gas chromatography, supercritical fluid
chromatography, microcolumn HPLC, capillary electrophoresis), the inner diameter
(i.d.) of the capillary is an important characteristic because it effects column volume,
flow, pressure, heat dissipation, and separation factors.9'11 In gas chromatography,
capillary i.d. is on the order of hundreds of micrometers, whereas in capillary
electrophoresis, the capillaries range from 10-100 /an i.d. It is for this latter
technique that this experiment is designed, so the sample capillary is on the order of
those used for capillary electrophoresis. Furthermore, narrower capillaries have
smaller probe volumes which is desirable in the reduction of Raman scatter from the
solvent. The key in choice of capillary i.d. is the laser focus size and tr at obtainable
flow rates which will discussed in later paragraphs. As it turns out, a 50 /an i.d.
capillary is nearly ideal for the system.

89
Capillaries commercially available for capillary electrophoresis are made of
fused silica, or quartz, with an outer diameter (o.d.) of 150 or 360 f*m. For this
application, the thinner walled tubing was chosen because: 1) thinner walls absorb
less heat and better dissipate heat than thicker walls; 2) thinner walls have a smaller
interaction volume with the laser beam (less absorption, less Raman scatter); 3)
focusing with narrower capillaries is easier; 4) the microscope can be positioned
closer to the source of the fluorescence emission, if necessary, with thinner walls; and
5) the greater lensing curvature of the narrower capillaries focus the laser to a
greater extent at the sample than thicker capillaries. Some of these points will be
discussed in more detail in the section below.
Capillaries used for electrophoresis are commercially available and are coated
with polyimide which endows the brittle quartz with flexibility. At the probe region,
the polyimide coating must be removed, which is usually accomplished through
stripping or burning it off. Burning is advantageous because it does not create
grooves or scratches on the capillary walls as stripping may do. However, burning
does leave a sooty layer on the capillary, but this artifact is easily removed by wiping
with a wet tissue. Once the polyimide coating is removed, the capillary must be
handled delicately because the quartz tube snaps easily upon slight bending.
Optical Considerations Regarding the Capillary
Focusing in optical systems is paramount in attaining the optimum conditions
in a spectroscopic analysis. The rounded surfaces of a narrow capillary create several

90
considerations involving focusing of the laser onto and collection of emission from
the sample stream not normally encountered. Previously, researchers performing LIF
for detection in capillary electrophoresis have resorted to special means to reduce
the problems associated with round surfaces.12 In many cases, the sample cell is
made to be rectangular,88 or a separate cell is used altogether such as a sheath flow
cuvette.89 An interesting approach is to manufacture an immersible cell of a desired
shape containing a fluid of the same refractive index of the capillary.90-92 However,
in this SMD approach, confidence has been placed on the ability of the MVF to
remove the laser scatter, so the capillary is used without alterations. The subject of
ways to reduce laser scatter by the capillary will be discussed in more detail in
Chapter 3.
Laser specular scatter. Specular scatter from a rounded surface occurs in all
directions arising from both the outer and inner walls of the capillary. The modeling
of this system is very difficult due to the shapes of the focus beam, capillary,
collection optics, and the difficulty of predicting where rays will end up from a
capillary. As mentioned in Chapter 2, up to 0.7% laser scatter of the total laser
intensity can be tolerated in the direction of the collection optics according to
calculations presented in Chapter 1. Through elaborate experimentation and optical
ray tracing, Bruno et al.9i have shown that the least amount of scatter from a round
capillary occurs in the direction perpendicular to the excitation beam. This is
because the increased angle of incidence of the incoming light at the edges of the

91
capillary causes closely spaced rays to disperse more widely at 90° to the laser beam.
Fluorescence is to be collected from this direction (90°) in this approach.
Rayleigh scatter. Laser specular scatter produced at interfaces of different
refractive indices is by far the most intense form of scatter produced from the
capillary, but Rayleigh and Raman scatter also arise from within the quartz to a
lesser extent. Rayleigh scatter is very weak and occurs at the same wavelength as the
laser emission. These forms of scatter have slightly broader spectral profiles than the
source, but due to the small Rayleigh cross-section, small capillary volume
illuminated, and broad absorption profile of the metal vapor filter (with respect to
the laser), this form of scatter is expected to be negligible as shown from theory in
Chapter 1.
Raman scatter from the capillary. Raman scatter from the capillary may
prove to be a more formidable problem than Rayleigh scatter. Quartz has a Raman
emission band94 at 464 cm'1 which corresponds to 825 nm with excitation at 795 nm
and 809 nm for 780 nm excitation. The Raman spectrum overlaps the fluorescence
emission spectra for the polymethine dyes, and there is little to be done to filter this
radiation unless a portion of the fluorescence is to be filtered as well. The possibility
of using a capillary made of a different type of glass does exist, but commercially
available capillaries of this sort are made of fused silica, and there is no guarantee
that other glasses will not have the same problem. The cross section for Raman
scatter is typically on the order or 10'30 cm2 per molecule,39 and the volume of fused
silica illuminated is about 100 nL. Without knowing the MW of quartz, the number

92
of expected photons arising from the quartz cannot be calculated. It is estimated
based on the parameters presented in Table 2 in Chapter 1 that the maximum signal
from the quartz of the capillary is on the order of 10 photoelectrons per 2 ms
counting interval.
Focusing the Laser
Laser focus onto the capillary. Figure 18 is a diagram, generated by a
commercially available optical ray tracing program (Beam 4, Stellar Software,
Berkeley, CA), that represents the laser focusing onto the 50 /¿m i.d., 150 ¿on o.d.
quartz capillary to be used in this system. The incoming light is collimated and was
assumed to be of 786 nm wavelength. The quartz has a refractive index, n, of
1.45356 at this wavelength40 and the inner tube was said to contain methanol (n =
1.327 at 589 nm). The focusing of the laser beam by the capillary wall is an
Figure 18. Focusing aspects of the 0.05 mm i.d., 0.15 mm o.d. quartz capillary
containing methanol with a 0.075 mm beam of collimated light at
786 nm entering from the left. Diagram generated by an optical
ray tracing program.

93
interesting feature of this system. The point of this diagram is that in order to
achieve 100% sampling efficiency, the diameter of the focused laser beam at the
outer wall of the capillary should be slightly greater than the i.d. of the capillary to
illuminate the entire sample. The calculated minimum laser focus diameter for the
50 fim i.d., 150 ¿an o.d. capillary to illuminate the entire sample region was « 70 ¿un
at the outer wall. When a 50 ¿un i.d., 360 ¿un o.d. capillary was modeled under the
same conditions the light rays were not focused to the same extent as in the narrower
tube and rays at the edge of the laser did not probe the sample. This feature refers
to point 5 earlier in why the narrower o.d. capillary was chosen for the SMD set-up.
The laser focusing lens. Since the laser beam was * 5 mm in diameter at the
capillary without focusing, the beam had to be focused to attain the calculated
optimum spot size. Focal length, diameter, beam divergence, laser wavelength, and
lens-type were important parameters to evaluate when making the choice of lens to
focus the laser.95 For this application, a longer focal length lens was chosen over a
lens with a shorter focal length for several reasons. With shorter focal lengths, the
lens produces a tightly focused spot of small volume; the beams converge and diverge
from this volume at steep angles and pose difficulties in optimizing alignment. Also,
if the focus volume is smaller than the i.d. of the capillary, molecules would be able
to pass the probe region without being excited which defeats the purpose of this
experiment. Conversely, the focus volume of a long focal length lens is larger than
the volume produced by a lens with shorter focal length, but the focusing angle is less

94
steep, the beam at the waist is more or less coherent, and the waist persists over a
longer distance.
For this experiment, an achromat lens having a 12 inch focal length and 1.2
inch diameter was chosen. An achromat reduces the spherical and chromatic
aberrations of a regular lens (the laser bandwidth is so narrow that chromatic
aberrations do not develop), and the antireflective coating on the lens reduces laser
scatter. Figure 19 is the result of a study to determine the optimum focus distance
of the lens with the laser. In this study, a 25 ¿an slit was translated (using calipered
optical micropositioners driven by a stepping motor) through the laser beam at set
distances from the lens. A photodiode behind the slit measured laser intensity, and
the laser beam diameter was determined from when intensity fell to a value 1/e2
(13.5%) of the peak. The minimum beam waist was found to be 70 ¿un in diameter
and occurred over a 3 mm region 13.3 inches from the lens. A large focal region is
the desired trait for focusing the laser beam in this experiment.
Collection of the Fluorescence
Collection of emission from the capillary. Due to the isotropic nature of
fluorescence (equal emission in all directions), optical considerations involving the
capillary and the collection of fluorescence are rather straightforward. The same
computer software was used to generate representative diagrams of the effect of the
capillary on the fluorescence emission from within the capillary. Figure 20 shows the
effect of the capillary on point source emission at different sites within the capillary.

Focused Beam Diameter (mm)
Distance from Lens (mm)
Figure 19. Focus size of the 794.76 nm Tirsapphire laser beam versus distance from the 12 inch focal length
achromatic lens as measured by translated a 25 /¿m slit across the beam.
Ln

96
Figure 20. Point source fluorescence emission at 830 nm from points within the
capillary containing methanol. Diagram generated by an optical ray
tracing program.
In this case, the light was said to be 830 nm (nqiuu1z = 1.45282)40 and methanol
was again the solvent. From the center of the capillary, the emission is not altered
on the same plane by the round walls, but the emission from points elsewhere in the
capillary are differentially bent as the diagram shows.

97
The microscope objective. A microscope objective was chosen as the means
to collect the fluorescence emission based on its large solid angle of collection and
high magnification. Fiber optics were also considered, but the lower solid angle of
collection, 0F (sr), and smaller optical transmittance of optical fibers promptly
nullified their use for this purpose. The important parameters of interest for a
microscope objective are its magnification and numerical aperture, N.A. Numerical
aperture and the calculation of fiF are defined in Chapter 1. Basically, the larger the
N.A., the more light is collected for the objective. The collection optics chosen for
this experiment is a 40X (magnification) objective with a N.A. of 0.65. It meets the
standard specifications for most microscopes with working distance of 0.42 mm and
6 mm aperture.
Field of depth and field of view. In a study similar to the one performed to
determine optical focusing of the laser, the collection properties of the microscope
objective were measured. To determine the optimal distance to position the emission
source, a 25 /¿m slit emitting the diffused emission from a tungsten light bulb was
used as a point source. Beginning at the point where the slit and outer lens were
touching, the microscope objective was translated away from the slit. A photodiode
positioned against the butt of the objective measured the amount of light transmitted
by the objective. Figure 21 shows a plot of photodiode signal versus distance from
the slit (the objective was moved horizontally at each point to find the maximum
signal). This plot gives the field of view of the microscope objective, and according
to this study, the optimum distance for collection of the fluorescence is 760 /¿m.

50
45-
40
35
30:
25
20
15
10
5
0
0
Peak Intensity occurs with Slit 0.76 mm
from Microscope Objective
1.00
Z00 3JOO 4^00 Z00 ZOO
Distance of Slit from Objective (mm)
7.00
8.00
ield of depth of the 40X, 0.65 N.A. microscope objecive used in this project determined by moving a 25
n slit emanating a point source of light gradually away from the objective. The photodiode was placed
rectly behind the microscope objective.
oo

50
45
40
35
30
25
20
15
10
5
0
r
>
Field of view of the 40X, 0.65 N.A. microscope objective used in this project determined by translating
a 25 /xm slit emanating a point source of light in front of the objective at 708 /im away. The photodiode
was placed directly behind the microscope objective.
SO
Vo

100
Figure 22 is the photodiode signal in the same experiment as the microscope
objective is translated horizontally across the slit at a point 708 /¿m away. This is
known as the field of view of the objective, which in this case is approximately 2 mm
in diameter. This means that light can enter the capillary from regions well outside
of the capillary, but this aspect of the lens is also what allows for the 1.5 sr solid
angle of collection for the fluorescence.
Detection
Choice of the Detector
The detector is the last of the five basic components of a fluorescence
instrument to be discussed. For this project, the criteria of the detector include:
(1) high detection efficiency, r¡, at the fluorescence wavelengths for the dyes (780-
850 nm);
(2) faster response time than the residence time of a molecule in the probe
volume;
(3) ability to detect near single photon events; and
(4) the area of detection should be of reasonable size.
Several photodetectors are currently available that meet some of these requirements,
but no detector is able to outperform others in every category. Therefore, the chosen
detector was the one that was able to perform generally well overall.
Table 9 is a list of possible photodetectors for use in this project.

101
Table 9. Possible detectors for use in this project.
Photodetector
V (%)
@800 nm
Tb
Area
(mm2)
S/N
Cost
($)
Hamamatsu R636
(PMT)
10
2x1 O'3
36
+ + +
662
Hamamatsu HC210
(Photodiode)
60
2.5x10s
6.72
+
180
RCA SPM-100
(Avalanche
Photodiode)
25
0.4
0.0025
+ +
2,500
Spex CCD
50
>1000
106
+ +
30,000
= Efficiency of Detection; br = Response Time
The charge-coupled detector, or CCD, is attractive in its high detection efficiency, tj,
in the near-infrared and large detection area. However, its prohibitive cost, and
millisecond minimum measurement time eliminates it from further consideration.
The Hamamatsu HC210 silicon photodiode, which is packaged with a pre-amplifier,
is inexpensive, easy to use, and possesses a high efficiency of detection, but this
detector is too slow for this application. Furthermore, the photodiode has a small
detection area which would require careful focusing. An interesting detector is the
RCA SPCM-100 which is termed an avalanche photodiode for photon counting
because it emits single TTL pulses when a photon is detected. The detection
efficiency is high in the NIR and response time is adequate for this application, but
like the photodiode, its detector area is very small. The small area is not as big of
a problem as the $2,500 cost of the detector. Perhaps in the future, the RCA SPCM-
100 detector or CCD will be tested.

102
The photomultiplier tube. Through the process of elimination, the
photomultiplier tube (PMT) is currently the best choice as the detector for this
project. The PMT is the most commonly used photodetector, and generally has the
lowest noise equivalent power. However, PMTs are not generally as sensitive in the
red part of the spectrum as they are at shorter wavelengths. This is because the
energy of photons is inversely proportional to wavelength, and the probability of
higher energy blue photons to release electrons at the photocathode, via the
photoelectric effect, is greater than that of lower energy red photons. This is a
disadvantage of working in the near-infrared part of the spectrum, but the advantages
of reduced noise and greater selectivity should overcome this drawback, especially
when one considers that there are red-sensitive PMTs to choose from that have very
low noise and high cathodic efficiency.
Hamamatsu R636. Since the fluorescence emission was focused by the
collection optics in a circular shape, a head-on PMT with a circular photocathode is
best suited for this project. However, a lack of a suitable head-on PMT housing in
the lab forced the selection of a side-on tube with rectangular photocathode shape
(designed for the detection of light arising from the slit of a monochromator). The
best side-on PMT available for this application was the Hamamatsu R636. This tube
has a 10% cathodic efficiency at 830 nm, 2 ns rise time, gain of 1.8 x 10s and 36 mm2
cathodic area. The most important feature is that it has very low noise
characteristics with a dark current of 20 pA at -20°C operating at -1500 V. The R636

103
is also useful for many other applications based on its flat sensitivity of response
across the UV/visible/NIR spectrum.
Photon Counting
Analog signal collection. Two general ways are available to collect the signal
from the PMT. In the analog mode, the current is sent to an electrometer or
current-to-voltage amplifier and then sent to a recorder. The response time of this
type of measurement is usually 10-1000 ms which makes it impractical for the
detection of single molecules with short residence times. Another analog approach
is to use a fast oscilloscope to discern the individual bursts of current when photons
strike the photocathode. These bursts would be counted in incremental time periods
to give the detected flux of photons. In essence this is what photon counting does.
However, before delving into photon counting, the next section contains a brief
review of the operation of photomultiplier tubes.
How a PMT operates. The photocathode of a PMT is a metallic material that
readily undergoes the photoelectric effect. The material used is based on the
application, and in the case of red-sensitive tubes, the material is GaAs(Cs). The
tube is placed under high voltage so that any electron released at the cathode is
drawn at high velocity toward the anode. However, the charge of a single electron
is minute, so a series of several (typically 9) dynodes is used to amplify the signal.
A dynode chain is basically an eletron multiplier; when an electron strikes a dynode
surface, two or more electrons are released to continue from dynode to dynode thus

104
amplifying the signal exponentially. By the time the signal reaches the anode, it can
consist of millions of electrons which generate a burst in the signal current.
The biggest source of noise with PMTs are the thermal release of electrons
from the dynodes (or photocathode). The probability of these releases decreases
with decreasing temperature which is why many analyses call for cooled PMTs. The
noise pulses from a PMT most often arise from a dynode and not the cathode which
means that these pulses will be of smaller size than photopulses generated at the
cathode. The key feature of photon counting is that a discrimination level is set to
differentiate between the smaller noise pulses from the larger signal pulses.97
Pulse height distribution. To obtain maximal signal to noise ratios using
photon counting, the optimum discrimination level must be set. To help in this task,
a pulse height distribution plot can be taken which gives an indication of the size and
occurrence of the signal and noise pulses for the PMT. In practice, this is
accomplished by setting a window of upper and lower levels that a pulse must occur
within and scanning that window across the signal level (a pre-amplifier is used in
conjunction with photon counting so discrimination level is in terms of voltage).
Figure 23 is a series of pulse height distribution plots for the R666 detector, which
is the high gain version of the R636). The pulse height distribution is dependent on
voltage to the PMT, and as shown in Figure 23, the distribution of pulses shifts to
larger signal levels as the voltage is increased. For this reason, the photon counter
is better able to differentiate the signal pulses from the noise pulses, and the PMT

105
is typically operated at the specified maximum voltage, which is -1200 V in the case
of the R666, and -1500 V for the R636.
Setting the discrimination level. Once the PMT has been characterized by a
pulse height distribution plot, the next step is to determine the optimum
discrimination level. Figure 24 is a plot of signal to noise ratio versus discrimination
level for the cooled R666 PMT operated at -1200 V. Based on this plot, it is
apparent that the discrimination level should be set at -50 mV. Similar data were
collected with the R636 which showed the optimum discrimination level was -10 mV
(the pre-amplifier was increased from 5X to 25X in this case).
Despite the low noise characteristic of the R636, photon counting requires
high gain for best results. The R636 has a gain of 1.8 x 105 which is on the threshold
of use in photon counting. It was thought that the high gain (2 x 106) of the R666
may improve upon the signal/noise of the R636 when photon counting is utilized.
A comparison of these two tubes and analytical results are given in Chapter 3. It
turns out that both tubes gave the same signal to noise ratios when photon counting
was used as when analog signal collection was used.
Nonetheless, photon counting is the better approach to signal collection over
analog measurements based on its capability to distinguish short residence times in
single molecule detection. In a flowing stream, a major consideration is that, due to
the incremental nature of molecules flowing through the probe volume, the collected
data must consist of increments equal to or shorter than the residence time of the
molecules. Photon counting is able to meet this requirement. In addition to

Figure 23. Effect of increasing voltage on the signal of the R666 cooled PMT versus discrimination level. Graphed
data is the 25 point smoothed average output of 100 scans with a 2 mV discrimination window. Room
light was used as the light source for the signal.
o
as

Signal/Noise Ratio
Discrimination Level (mV)
Figure 24. Signal to noise ratio of the cooled R666 PMT versus discrimination level with room light serving as the
signal source and standard deviation of the dark current being the noise. Average of 100 scans of 2000
points in 2 ms counting intervals.
o
-â– j

108
fast response time, photon counting enables computerized control of detection
parameters, computerized data collection, and data processing. When working with
a large amount of raw data, computerized analysis is a very important feature.
Control of the Sample Flow
Although flowing sample through a capillary is not as complicated as the
sheath flow cuvette, it is not trivial at the flow rates necessary for SMD. In order to
achieve the desired 10 ms residence time of a flowing molecule in a 140 pL volume,
the flow rate must be 0.84 /xL/min. To achieve such a low flow rate reproducibly
requires the use of a special pump, especially at the high pressure encountered when
pumping methanol through a 17 inch long, 50 /xm i.d. capillary.
Capillary electrophoresis. Due to the laminar nature of flow, the flow velocity
in the center of the tube is greater than that nearer the tube walls. Only capillary
electrophoresis, in which the impetus for flow originates at the wall itself, is able to
nearly attain plug flow (when the flow velocity is equal across the tube diameter).
Conducting capillary electrophoresis on the sample can achieve flow rates as low as
0.84 /xL/min, and plug flow is certainly desirable, but setting up a capillary
electrophoresis system would be beyond the scope of the research project at this
stage.
Syringe pump. In an attempt to meet the flow conditions needed for SMD,
two types of pumps in the lab were tested. A syringe pump (Sage (Model 355),
Cambridge, MA) had the advantage of continuous delivery of a low flow rate, but it

109
had the disadvantages: 1) of not being able to operate at high pressure, 2) requiring
frequent syringe changes, and 3) irreproducible flow because of sticking of the
syringe. These problems, and the additional problems of leaks and noncontinuous
flow rate, made use of the syringe pump in our lab intolerable in this application.
HPLC pump. An HPLC pump (Autochrom (Model M500), Berlin, Germany)
applicable in microcolumn chromatography was the second pump tested. This pump
had the advantages of better flow reproducibility, operation at high pressure, and no
need for syringe changes, but it had impediments concerning a higher minimum flow
rate, pulsed flow, and excessive maintenance. Despite these problems, it was deemed
superior to the syringe pump in practice. The ideal pumping method, other than
capillary electrophoresis, would be a syringe pump designed for HPLC.
Figure 25 is a calibration of the flow rate through the capillary using the
HPLC pump. In this case, flow rate was measured by weighing the amount of
methanol accumulated into a 10 mL graduated cylinder over a timed period of
several hours (at low flow). A second cylinder containing a measured volume of
methanol was kept beside the first, and the change in weight was used to account for
evaporation losses. The weight of methanol was converted to volume from density
and the flow rate was calculated. The flow rate at the lowest pump setting was 5
¿iL/min which corresponds to a flow velocity of 4.4 cm/s and tr = 1.6 ms. Based on
the number of photons necessary for SMD, and the minimum counting period of 2
ms for the photon counter, a slower flow rate is needed. As shown in Chapter 3, it
was not necessary

Flow Rate (mL/min)
Figure 25. Calibration of the HPLC pump flowing methanol through the capillary. Flow rate at lowest pump setting
is 5 ¿iL/min.
o

Ill
to lower the flow rate at this stage and the lowest pump flow setting was used, but
in the future, some way must be developed to lower the flow rate.
To accomplish this task, the flow can be split into two tubes of different sizes
and lengths. Initial experiments of a split flow system were performed to determine
the feasibility of the method. A long piece of narrow-bore stainless steel tubing was
teed to the tubing before the capillary, and the flow rate was measured by the same
method as before. The results of the study show that the flow rate going through the
capillary at the lowest setting was less than 0.84 /¿L/min, and this method could be
used in future experiments. Besides reducing the flow into the capillary, splitting the
flow stream presents a way to monitor the average flow in the capillary. Since the
summation of flows in both tubes is always the same, the flow in the capillary can be
deduced by measuring the flow in the split tubing. This is an easy way to estimate
tr during an analysis.
However, the accuracy of this determination is questionable due to the
laminar nature of flow and bulk measurement of flow rate. Another way to
determine flow rate involves placing microspheres into the flow stream and
measuring how many counting periods one takes to traverse the Vp.44 This method
could not be done using this system because it would be difficult to introduce the
spheres into the capillary and the tr was expected to be shorter than a single counting
period. This microsphere method will be attempted in future experiments when the
flow rate has been lowered.

112
Experimental
The set-up. Now that the choice of instrumentation and optimization of
parameters has been discussed, the set-up as an integrated unit can be presented.
Figure 26 is the basic instrumental arrangement for the approach to SMD of this
dissertation. Some individual experiments may have used additional components or
minor modifications of the set-up shown here, but these instances will be mentioned
with the results of those experiments in Chapter 3. Otherwise all experiments used
the set-up described here.
The lasers. The basic approach involved pumping the Schwartz Electro-Optics
(Orlando, FL) Titan CWBB Tirsapphire laser in either the ring or standing wave
configuration with the Spectra-Physics (Mountain View, CA) Model 2040 argon ion
laser. During experiments the Ar+ laser was operated in the constant power mode,
and aperture setting #8 was found to give the best spatial profile in the all-lines
mode (the Tirsapphire required a TEM^ pumping laser beam). Power of the
Tirsapphire laser was measured with a Coherent (Palo Alto, CA) Model 210 power
meter.
The dark box. The output of the Tirsapphire laser was sent by two mirrors
and a prism into the dark box (formerly a computer printer box) which was
approximately 2x2x2 feet. Inside the box, black foam was attached to the inner
walls, black felt was spread on the optical table, and black metal baffles were placed
to separate the detector from laser scatter (other than the amount collected by the
microscope objective). The laser beam entered the box through a 6 inch long x 0.5

Figure 26. Basic instrumental arrangement for the approach to single molecule detection described in
dissertation.

114
inch diameter tube with an iris placed at the entrance to keep room light out of the
box. A Corion (Holliston, MA) LG-697 long-pass filter was positioned at the other
end of the tube in the box to remove the small amount of green Ar+ emission that
was contained within the Ti:sapphire beam. In the box, room light was excluded to
the extent that the PMT dark current was able to be reached when the box was
closed and the PMT shutter opened (while the laser was blocked before the
capillary).
Focusing of the laser. A 12" focal length achromatic lens with an anti-
reflective coating focuses the beam onto the 50 /¿m i.d., 150 /¿m o.d. capillary
(Polymicro Technologies, Phoenix, AZ) inside the box. The lens was mounted on an
XY micropositioner (Newport, Fountain Valley, CA) and height was adjusted with
the post holder. The capillary was mounted with similar positioning capabilities
(except height adjustment) and was held in the vertical position by two plastic prongs
about 3 cm apart. The polyimide coating of the capillary was burned off with a
butane lighter in the region of the laser focus. The positioning of the laser beam and
capillary at the optimum location in front of the microscope objective was
accomplished by flowing a concentrated dye solution through the capillary, and the
optimum position was assumed to occur where the signal was maximal. In the
procedure, finding the the correct focus height was done first by adjusting the height
of the lens by turning the post-holder adjustment screw. The optimum capillary
position in the field of view of the objective was found by translating the capillary
across the plane in front of the objective. To determine the optimum distance from

115
the objective, both the lens and capillary were translated starting from the front
surface of the objective to farther away until the maximum PMT signal was reached.
The focus size was the final adjustment which was accomplished by moving the lens
closer to or farther away from the capillary. The optimum distance of the lens from
the capillary was * 13.3 inches, and the distance of the capillary from the
microscope objective was * 550 ¿tm.
Fluorescence collection. The fluorescence collection optics used in these
experiments was a 40X, 0.65 N.A. Nikon (Japan) microscope objective. The
objective was screwed into the horizontal microscope tube leading to the detector.
After studying the focusing aspects of the microscope objective, it was determined
that the light cone would be 6 mm in diameter at a point just after the MVF. A
round metal aperture of 6 mm diameter was placed at this point, and a glass
planoconvex lens with 9.5 cm focal length was placed just after the aperture to focus
the collected emission onto the 3 mm wide photocathode. The distance between the
the PMT and microscope objective was 27 cm; both components were mounted to
the optical table in fixed positions.
The metal vapor filter. The metal vapor filter was wrapped in heating tape
and placed in the microscope tube. Two MVFs were available for use: cell #1,
made by Rudy Strohschein (Gainesville, FL), contained 500 mg Rb in 200 Ton-
nitrogen with a 4.4 cm long internal path length; cell #2, made by Opthos (Rockville,
MD), contained 500 mg Rb in 500 Torr N2 with a 4.7 cm internal path length. The
cells were 1 inch in diameter pyrex cylinders. Temperature of the heating tape was

116
controlled by adjusting the voltage with a varible autotransformer and monitoring the
temperature with a K-type thermocouple and digital readout (Omega, Stamford, CT).
Once thermal equilibrium was reached, temperature of the MVF varied less than 1°C
over a period of hours. Typical operating temperature was 100°C.
The end of the microscope tube was butted against a Uniblitz (Rochester,
NY) Model SD-122B electronically operated shutter which was attached to the PMT
housing. The shutter was round with a diameter of 1 inch. To minimize stray light
reaching the PMT, black Apiezon grease was caulked around the outside of the
shutter, 6 mm aperture, and around the connection between the shutter and
microscope shaft. A Corion RG-795 long pass spectral filter (or other filter from
Corion to be discussed in the next chapter) was positioned behind the shutter to
absorb light of wavelengths < 750 nm.
Detection. The PMT was either a Hamamatsu (Japan) R636 or R666 and was
contained in a cooled (* -20°C) housing (Products for Research Model TE177RF,
Danvers, MA). The PMT was biased with a Fluke (Seattle, WA) high voltage power
supply at -1500 V for the R636 and -1200 V for the R666. This power supply was
found to give the most stable voltage supply in the lab.
Signal collection. The PMT signal was collected with either analog or photon
counting methods. In the analog set-up, the PMT current was measured by a
Keithley (Cleveland, OH) Model 614 electrometer, and recorded by a Fisher (Fair
Lawn, NJ) Recordall strip chart recorder. In this method, a 1 Hz low pass filter was
connected between the electrometer and recorder to reduce noise. In photon

117
counting, the signal was sent by a 14" cable to a pre-amplifier (Stanford Research
Systems (Sunnyvale, CA) Model SR445). An adjustable * 50 Q snubber was teed to
the other end of the cable to reduce ringing effects of the photoelectron pulses. The
R666 required 5X amplification to produce an optimum discrimination level of -50
mV, and the R636 signal was amplified 25X to give a -10 mV discrimination level.
A computer (Datatech-1230C) controlled the Stanford Research Systems Model
SR400 photon counter/discriminator and was used for data collection. The SR400
scans were typically 2000 points with 2 ms intervals and 2 ms dwell times which
means there was only a 50% sampling efficency with this instrument. For application
of SMD, a different photon counter able to scan continuously would have to be used.
Pumping. An Autochrom (Berlin, Germany) Model M500 HPLC pump
typically set at its lowest flow setting (0.01) was used to generate solvent flow. The
flow setting corresponded to an average flow rate of 5 /xL/min and residence time
of 1.6 ms in the 140 pL probe volume. At this flow rate, pump pressure was < 0.2
MPa. Sample introduction simply involved changing the beaker of solution for the
pump, but a 5 /xL 6-port HPLC injector is available for future applications. The
HPLC tubing leading to the capillary contained a 0.2 /tm in-line frit, and the 17 inch
long capillary is connected after the frit with the use of a graphite reducing ferrule.
Reagents. Fisher Optima methanol was used for the solvent in all calibration
curves, but other tested solvents included ethanol (Florida Distillers, Lake Alfred,
FL), acetone (Fisher), HPLC grade actetonitrile (Fisher), hexane (Baxter/Burdick
& Jackson, Muskegon, MI), and A.C.S. grade dimethlysulfoxide (Fisher). The dyes

118
used in the project were laser-grade IR 125, IR 140, and IR 132, all from Exciton
(Dayton, OH). Solutions were continuously bubbled with nitrogen gas when being
pumped into the capillary. Since the dyes decompose in oxygen under light, the
blank solvent was degassed several hours by bubbling with dry air under fluorescent
lighting to reduce the chance of contamination of the blank with polymethine dyes.
The samples were filtered with 0.45 pirn pore size Acrodisc 3 membrane filters before
being pumped through the system.
Procedures and other parameters will be given with the description and results
of each experiment in Chapter 3.

CHAPTER 3
RESULTS AND DISCUSSION
Studies of the Metal Vapor Filter
Absorption Properties
As shown in Chapter 2, the rubidium metal vapor filters (MVFs) at
temperatures greater than = 100°C should have absorbances greater than can be
measured. Several experiments were undertaken to determine whether the
MVF/Ti:sapphire laser combination were actually able to meet theoretical
expectations.
Tuning the laser. Before performing these experiments, however, the tuning
range of the Ti:sapphire laser had to be calibrated in order to tune to the 780 and
794.76 nm Rb lines. The caliper adjustment for the birefringent filter was calibrated
to the corresponding wavelength by sending the attenuated laser output into the Spex
(Edison, NJ) Model 1680 double spectrometer, and the caliper setting was adjusted
to maximize the R636 PMT response at known wavelength of the monochromator.
The caliper settings for 780 and 794.8 nm were specially noted.
To tune the laser to a Rb transition, the MVF was heated to * 100°C, and
the laser was passed through the MVF onto a photodiode. The caliper was slowly
adjusted near the position determined from the calibration experiment. The
119

120
minimum signal was found, then the Fabry Perot etalon in the laser was adjusted to
further minimize the signal. This became a trivial procedure once the general
settings were known. It is noteworthy that once the laser was tuned to the absorption
line, adjustments were seldom necessary. Also, the maximum absorbance was more
like a plateau than a peak because the etalon and birefringent tuner could be
adjusted significantly before the laser was tuned off-wavelength and the signal
increased sharply.
The first absorbance study performed involved cell #1 (4.4 cm internal path
length, 200 Torr). The cell was placed in a Lindberg (Watertown, WI) Model 55035
tube furnace with temperature control, and the laser beam was sent directly through
the MVF onto a photodiode; the signal was collected by a strip chart recorder.
When the cell was sufficiently heated, the laser was tuned to the minimum
transmittance at 780 nm as described above, then the furnace was allowed to
gradually cool. A few hours later, when the cell reached room temperature, it was
removed, and the photodiode signal was calibrated with the use of neutral density
filters of known absorbances. The measured transmittance of each data point was
converted to absorbance, and a plot was made of the results.
Figure 27 is a comparison of the results of this initial study with the
theoretical absorbance values for cell #1 vs. temperature calculated from theory
presented in Chapter 2. The comparison shows that the general shape of the two
plots are similar until * 90 °C where the experimental absorbance tops out and the
theoretically derived absorbance continues upward.

o
9
8
7
6
5
4
3
2
1
O
lomparison of the theoretical and experimental results for absorbance of the Rb metal vapor cell #1
srsus temperature at 780 nm. The Ti:sapphire laser was in the standing-wave configuration with power
* 200 mW before the cell.
to

122
The plot exhibited in Figure 27 can be divided into three regions: 1) the
differences in the absorbances from 25-50°C is most likely due to light losses at the
windows of the MVF not accounted for in the theory; 2) from 60-90°C, the
differences in values are possibly due to the differences between the external tube
furnace temperature of the plot and the actual temperature inside the MVF; and 3)
the reason that the plot reaches a maximum absorbance at temperature > 90 °C is
due to stray light. Undoubtedly, stray light was interfering with the measurements
at all temperatures, but it became the limiting factor in the absorbance
measurements at temperatures > 90 °C because the amount of laser light passing
through the MVF was less than the amount of light reaching the photodiode that did
not pass through the MVF.
Stray light is often the limiting factor in absorbance measurements. Due to
this aspect and limitations of the detector, most spectrophotometers cannot measure
absorbances greater than 3 or 4. The maximum absorbance of 4.8 obtained in this
experiment was approximately the expected upper limit for the measurement before
being limited by stray light and detector dark noise. Based on the theory and
appearance of the results at temperatures below 90 °C, it was believed that the actual
absorbance continued upward as predicted by theory, but that these absorbances
were not measurable with this system. A similar experiment carried out at the 794.76
nm Rb line gave a maximum absorbance of 3.9, again at temperatures > 90°C.
For purposes of comparison, the absorbance of cell #2 vs. temperature was
determined under similiar conditions. Figure 28 is a semi-log plot of the results of

Absorbance
Temperature (°C)
Figure 28. Semi-log plot of the theoretical and experimental results for absorbance of the Rb metal vapor cell #2
versus temperature at 794.76 nm. The Ti:sapphire laser was in the standing-wave configuration with power
« 200 mW before the cell. w

124
this study and the theoretical expectations. Again, the absorbance reached a
maximum and remained at this value as the temperature increased. In the case of
cell #2, the temperature was higher (« 110°C) as was the absorbance (A = 5.2)
when the maximum was reached. These results agree with theory that the cell under
higher pressure should require more heating to achieve the same absorbance as the
cell under lower pressure. Also, scatter was less severe in this case exhibited by the
higher absorbance attainable.
Stray light. After these studies, it was believed that stray light was responsible
for the limiting absorbances encountered at high temperatures, and that when actual
experiments were performed in the dark box with the PMT, that much greater
absorbances would be achieved. However, when cell #1 was placed in the set-up
shown in the experimental section, the absorbance again reached a maximum at a
lower temperature than expected. Figure 29 shows how the maximum absorbance
obtained was 5.4 even with the use of a Corion S10-830 interference band pass filter
which had an absorbance of 3.34 at 794.76 nm. These results were disheartening
because it was believed that stray light was not able to reach the PMT due to the
presence of the aperture after the MVF. Based on the results of this experiment,
other possible explanations as well as stray light were also considered as the source
of the problem, as enumerated below.
Atomic fluorescence. When a Rb atom absorbs light at 780.023 or 794.760
nm, the excited electron may emit light at the same wavelength or at another
transition wavelength. This process is known as atomic fluorescence, resonance

14-
13
12
11
10
9
8
7
6
5
4
3
2
1
0
/
c
v
irison of the theoretical and experimental results for absorbance of the Rb metal vapor cell #1
temperature at 794.76 nm when the laser is reflected by the capillary. The S10-830 bandpass
rence filter was placed before the cooled R636 PMT which is why the plot begins with A = 3.34.

126
fluorescence being the term for emission occurring at the same wavelength as the
absorption. Atomic fluorescence processes were expected to be nearly entirely
quenched by collisional deactivation with the concentrated nitrogen gas in the cell,
and the little fluorescence that does take place was expected to be totally absorbed
by the other Rb atoms in the cell (since the fluorescence occurs at the same
wavelengths as the absorption). It was thought that atomic fluorescence would not
be a factor in the use of the MVF for this application, but nothing was done to rule
out this possibility up to this point in the project.
Raman scatter. A second unlikely possibility to be the additional source of
light in the experiment is Raman scatter from the solvent and optics. As calculated
from theory in Chapter 1, the Raman scatter from the solvent and the capillary under
experimental conditions was expected to be very small. The S10-830 ruled out
Raman scatter from the methanol solvent, but the quartz band at 464 cm'1 from the
capillary or pyrex band from the MVF front window could pass through the filter and
reach the PMT. Again, the intensity of the light reaching the detector was thought
too great to be Raman scatter, but since the detection system gives no information
concerning wavelength, this possibility cannot be eliminated.
Laser side-lobes. A fourth explanation for the lower absorbance is that the
laser linewidth was not as narrow as stated by the manufacturer, or that the laser
output contained broad-band background emission on top of which rests the narrow
laser line. Also, based on data collected by Ramee Indralingam,65 it was postulated
that the laser emitted side-lobes at high and low wavelengths from the center line.

127
MVF absorption study with the Spex 1680. All of the potential reasons for
the presence of extra light could be explained by doing the absorption studies of the
MVF using a monochromator to give a wavelength dependence on the absorption
band. In this way, provided the wavelength resolution was high enough, the
absorption band could be monitored as temperature was increased. For this purpose,
the Spex 1680 double monochromator was used as in the case before when the laser
wavelength was calibrated. This time, MVF cell #2 was placed in front of the
entrance aperture where the laser entered the monochromator. The slits were set
to 0.05 mm which gave a 0.1 nm bandwidth reaching the R636 detector. At high cell
temperature, the light reaching the detector all occurred within the resolution of the
monochromator at 794.8 nm. Thus, after this experiment, stray laser light, resonance
fluorescence, and broader than stated laser linewidth were still possibilities, but the
broad-band background and Raman scatter possibilities were eliminated.
Interestingly, the source of the side-lobes noted previously65 was believed to
be determined by this experiment. As the cell cooled and more laser power began
passing into the monochromator, two small bands on both sides of the central peak
began to appear. The size of these bands was a function of power entering the
monochromator, so these bands are related to the spectral separation abilities of the
gratings in the monochromator. It is believed that these lobes are a result of
diffraction by the slits or focusing aberations by the gratings or mirrors in the
monochromator.

128
MVF absorption study with the HR 1000. Due to the inadequate spectral
resolution of the Spex 1680 for this study, the experiment was repeated with the use
of the HR 1000 1 m monochromator (ISA/Jobin-Yvon, Metuchen, NJ) fitted with a
cooled, intensified linear diode array detector (Princeton Instruments Model IRY-
1024G, Princeton, NJ). The entrance slit to the monochromator was set to 10 ¿un
which gave a bandwidth of » 50 pm (still not enough resolution to resolve the
transition line, but it was the smallest obtainable bandwidth in the lab at this
wavelength). This experiment gave the same results as before, even with the
improved resolution and when the cell temperature was increased to 166°C.
However, in this instance, a black card was placed between the metal vapor filter and
the slit, but the signal did not decrease. This means that stray laser light can be the
only explanation for the anomalous results obtained in the previous experiments
because the card would have blocked resonance fluorescence and Raman scatter
from the cell.
The stray light was reduced as much as possible in the experiment using the
HR 1000, but the scatter could not be totally eliminated without redesigning the
optical set-up. Figure 30 illustrates the resulting signals of the diode array with the
MVF blocked and unblocked and the difference between the two scans. These
spectra were collected from an accumulation of 100 scans of 0.33 s each. The reason
that the scan performed with the MVF blocked had a higher background than the
scan taken with the MVF unbocked is not known, but it is believed to be due to
differences in scattered room light at the time of the scans. This made no difference

Intensity
200
Difference
-100+-
786
788 790 792 794 796 798
Wavelength (nm)
800
802
Figure 30. Data from the linear photodiode array detector on the HR1000 monochromator (bandwidth = 50 pm)
that shows the occurrence of scattered light even when the 166°C MVF cell #2 was blocked. The
conclusion of these data is that the MVF absorbs the laser, but that scatter is still prevalent.

130
in the signal at the Ti:sapphire emission wavelength as evident in the spectrum in
which the scans were subtracted from one another. Based on the noise of the
difference between these two scans and the use of neutral density filters as
calibration standards, the absorbance of the MVF cell #2 at 166°C was > 7.
Transmittance Properties
Rb metal deposition on the MVF windows. For the MVF to be useful in the
SMD project, not only must it absorb the laser scatter, but also it must transmit the
fluorescence. The great advantage of the MVF is that it specifically removes laser
scatter noise and maintains a high optical throughput at nearly all other wavelengths.
A potential drawback with the MVF is that the windows of the cell can become
coated with metallic Rb which can physically block the light at all wavelengths. This
can occur as the cell cools, and Rb vapor tends to condense at the surfaces of lowest
temperature. If the cell is heated from the walls, the windows become the coolest
surfaces, and indeed, Rb metal was observed to accumulate on the windows during
the cooling process.
A simple study was performed to quantify the transmittance of light using
the Hewlett-Packard 8450A spectrophotometer. The transmittance of cell #2 before
its first use was 85% from 400-800 nm (and assumed to remain at this level at longer
wavelengths). After the initial use of the cell in the experiment manifested in Figure
28, the transmittance fell to 80%. After a second heating period, the cell was cooled

131
more rapidly, and the %T dropped to 51%. If this trend continued, the cell would
be useful for only a few experiments.
Cleaning the windows. However, a method to clean the the cell windows was
developed. At temperatures > 39 °C, exists as a liquid, and much like liquid water,
cohesive forces are greater than the adhesive forces of the metal with the glass. In
other words, Rb liquid prefers to form one large glob rather than several smaller
globs in the cell. By simply raising the cell temperature > 40 °C with a heat gun or
in an oven and rotating the cell to gather all the globules condensed on the windows
into one mass, the windows can be cleaned. The transmission spectrum of the cell
after this procedure returned to 80%.
In order to avoid the problem altogether, Norma Ayala has developed a way
to halt the metal deposition on windows of metal vapor cells during the cooling
process.97 This is accomplished by placing a cold tip at the cell wall during cooling
which causes the metal to condense at the wall rather than on the windows.
Additional Filtering of Light
Spectral Filtering
Based on the results of this study, it was realized that despite the ability of the
MVF to absorb the laser scatter to an undetectable level, laser scatter was a more
formidable problem than expected. Therefore, additional spectral filtering was
required to help reduce the scattered light that passes around the MVF.

132
It must be pointed out that SMD is not likely to be realized when spectral
filters are used. In the calculations presented in Chapter 1, the optical transmittance
used was 0.5 (based on the 90%T for the quartz capillary, = 80%T for the
microscope objective, » 80%T for the MVF (85 %T when new), 90%T for the
focusing lens and 95%T for the PMT window). The overall transmittance factors of
the collection optics are important to being able to achieve SMD because the signal
is in direct proportion to them.
Figure 31 gives the transmittance spectra of the five filters (all from Corion)
tested for possible use in this experiment as measured by the Varían 634
spectrophotometer. As can be inferred from Figure 31, the LG and RG filter
designations refer to long pass spectral filters with the stated number corresponding
to the wavelength at which the %T is half of the maximum value. The other two
filters are interference band pass filters of 10 nm full width at half maximum
bandwith with peak transmittance centered at the designated wavelength. The
wavelength ranges of the transmittance can be compared with the fluorescence
emission spectra of the dyes appearing in Figure 11 in Chapter 2.
Table 10 lists the %transmittance of the total emission passed by each
dye/filter combination which were determined by multiplying PMT signals at each
wavelength of each dye, as shown in Figure 11, by the transmittances of the filters,
as shown in Figure 31, and then dividing the sums of the resulting values for the
spectra with the filters by the overall PMT signal for each dye obtained without
accounting for the filters. Table 10 also lists the absorbances of the filters at 780 and

so¬
so
70
60
50
40
30
20
10
O
S10-820
RG-830
30 790 800 810 820 830 840 850 860
Wavelength (nm)
870 880
890
spectra of spectral filters for possible use in experiments.
u>
OJ

134
795 nm which were measured in the same method previously described in the
experiments used to measure absorbances of the MVFs with the photodiode.
Table 10. Transmittances of the possible spectral filters to be used in combination
with the metal vapor filter.
Filter
Absorbance
%Transmittance
780.02 nm
794.76 nm
IR 125
IR 140
IR 132
RG-830
2.79
2.33
20.88
38.48
49.66
LG-840
2.89
2.14
16.61
38.48
35.77
RG-850
3.2
3.21
14.51
30.06
42.33
S10-820
3.3
3.24
18.86
14.95
10.01
S10-830
3.3
3.34
12.46
15.43
11.86
Based on the values given in Table 10, Figure 32 shows the calculated relative
signal to background ratios with 794.76 nm excitation of the polymethine dyes in
methanol with the different filters. As the figure shows, IR 132 with the RG-850
spectral filter is expected to give the largest signal/background ratio.
Polarization
Another way to reduce scattered light as presented in Chapter 1 is through
polarization filters. This possibility was explored in this project by placing a calcite
polarizer in the Tirsapphire laser beam path before the dark box. The polarizer was
rotated to obtain maximum power as measured by the power meter; it was found that
80% of the power passed through the calcite polarizer (and a minimum of 11% could
be obtained). A second polarizer (Melles Griot, Irvine, CA) was placed after

Signal/Background
Filter
IR 125
IR 132
IR 140
Figure 32. Calculated signal/background ratios of the polymethine dyes in methanol at normalized concentration and
power with the different spectral filters (excitation at 795 nm). ^
L/l

136
the first before the power meter and rotated to minimize the power. This simple
experiment showed that this system was capable of 105 light rejection of the laser
light when it is passed directly through the filters.
However, when the polarized laser was scattered from the capillary, and the
second polarizer was placed inside the microscope tube after the focusing lens, a
factor of only 3 x 103 light rejection was obtained. The difference between the
rejection values was due to the slight changes in polarization of the laser light as it
was scattered from the many surfaces between the capillary and second polarizer.
Furthermore, the PMT signal was very sensitive to very slight adjustments in the
rotation of the second polarized filter, even in the direct laser beam, so it is
understandable that the rejection factor was less in the set-up than in the direct path
of the polarized laser. Still, taking advantage of polarization is a good way to reduce
polarized laser scatter and still transmit much of the nonpolarized fluorescence.
With the Hewlett-Packard 8450A, the second polarizer was found to have a
transmittance of 35% at 800 nm. Reducing the noise by a factor 3 x 103 and losing
65% of the signal is normally considered a very good trade-off.
Results
Limits of Detection
Once the system was set up and characterized, there remained only one other
task: to measure dye standards of known concentrations and determine the lowest
detectable concentration. Several different experiments were carried out, and slight

137
modifications in the instrumentation were made in each case in an attempt to obtain
the lowest possible LOD. Of these, four experiments were chosen to be
representative of the detection ability of the system. Table 11 contains these results
and a description of each experiment follows.
Experiment #1 The first experiment did not use the system showed in Figure
26, but instead utilized the Spex 1680 monochromator in a more traditional analytical
LIF procedure with detection in a 1 cm square, 3.75 mL flowcell. Solvent was flowed
through the cell using a syringe pump set at 1 mL/min. This experiment was
performed in order to give a comparison in the detection with a monochromator
system as opposed to the approach using a capillary and the MVF.
When using the monochromator, the laser could be tuned to the wavelength
of maximum excitation instead of to a Rb line, and the emission could be collected
at the wavelength range of maximum intensity. The versatility of choosing these
parameters is an advantage of the monochromator system, but in the case of the
SMD approach, the analysis was designed for a specific analyte; therefore this
versatility is not needed.
Before performing the experiment to determine the LOD for the dye, a study
to optimize monochromator slit widths and laser power was performed. It was found
that 5 mm entrance and exit slits (9 nm bandwidth) gave the largest signal to noise
ratio at fixed laser power, and signal to noise increased linearly with power up to ~
300 mW (the maximum operable laser power at the cell due to optical losses). A
calibration curve for IR 140 was taken at these conditions (listed in Table 11), and

138
Table 11. Results of the experiments to determine limits of detection for the
approach.
Parameter
Experiment
#1
#2
#3
#4
Dye
IR 140
IR 132
IR 140
IR 140
Solvent
Methanol
Methanol
Methanol
Methanol
Sample
Container
1 cm
Flowcell
50¿un i.d.
Capillary
50/xm i.d.
Capillary
50/im i.d.
Capillary
Filtering
“Spex 1680
MVF #2
RG-850
MVF #1
S10-830
MVF #1
S10-830
Detector
R636
R636
R666
R636
Xcx (nm)
800
794.76
794.76
794.76
$l (mW)
280
65
240
240
SL (cm2)
*3x 104
3.5 x 10'5
3.5 x 10'5
3.5 x 10 s
El (W/cm2)
*930
2,300
6,900
6,900
Flow Rate
(¿íL/min)
1000
20
5
5
Flow Velocity
(cm/s)
17
17
4.4
4.4
Vp (nL)
«160
0.14
0.14
0.14
tr (ms)
*1
0.4
1.6
1.6
Tc (S)
0.3
0.002
0.032
1
Mb!
140 pA
b4900
b4.1
450 pA
Sbl
7.2 pA
b150
b0.52
4 pA
Sensitivity
150 A/M
c7.7 x 10"
c7.7 x 109
1.27 A/M
LOD (M)
3.0 x 1013
6.0 x 1010
2.0 x 1010
9.5 x 1012
LOD (#/Vp)
« 29,000
51,000
17,000
800
LOD (#/rc)
8,700,000
260,000
340,000
500,000
sMonochromator set at 833 nm with 9 nm bandwidth; bCounts per 2 ms; cCounts/M

139
an LOD of 3 x 1013 M was obtained. This corresponds to an average number of
» 29,000 molecules in Vp at the LOD, or approximately 8.7 million molecules
passing through Vp during the minimum measurement period at the LOD.
Experiment #2 Experiment #2 utilized the approach shown in Figure 26.
The analyte chosen for this analysis was IR 132 which, as presented earlier, was
expected to give the largest signal with the RG-850 filter/MVF combination at
794.76 nm excitation. A study to optimize flow rate and laser power showed that
higher flow gave a larger signal/noise ratio, and that 65 mW power (EL = 2.3
kW/cm2) at the capillary yielded optimum power at all flow rates.
The low optimum power and increasing signal versus flow rate suggested that
the dye was decomposing during its residence time. This was not noticed in previous
studies with IR 132 because the system with the Spex 1680 involved a lower laser
irradiance (« 1 kW/cm2). Apparently, with irradiance of 2.3 kW/cm2,
photodecomposition of IR 132 takes place in under 0.4 ms. However, this was not
verified with subsequent studies.
In any event, IR 132 analysis in methanol solvent flowing through a 50 /¿m i.d.
capillary at 0.02 mL/min with R636 PMT detection in photon counting intervals of
2 ms and MVF/RG-850 spectral filtering was able to obtain an LOD of 6.0 x 10'10
M (51,000 molecules on average in the 140 pL Vp). This does not seem to be an
especially low LOD, but with the detection criterion used by Keller’s group,2’3'5,54 this
corresponds an LOD of 260,000 molecules during the 2 ms counting interval.

140
Experiment #3 Due to the decomposition problem with IR 132, it was
decided to perform all future analyses in the capillary with IR 140 (IR 125 fluoresced
at wavelengths too short for optimum use with the filters). Also, photon counting
detection with the R636 proved difficult (mentioned in Chapter 4); therefore,
experiment #3 used the R666 for this application. Furthermore, in the event that
Raman scatter from the solvent (not taken into account in Figure 32) was a more
significant source of noise than anticipated, the S10-830 bandpass filter replaced the
RG-850 long pass filter.
In the case of IR 140, the lowest possible flow rate could be set with little
change in signal, and the saturating laser power closely corresponded to the 207 mW
value calculated in Chapter 1. Therefore, to achieve the maximum signal to noise
laser power was set to 240 mW (just above saturation) at the capillary (EL = 6.9
kW/cm2), and to achieve the longest tr the flow rate was set to 5 /xL/min. The LOD
under these conditions was 2.0 x 10'10 M (17,000 IR 140 molecules in 140 pL). With
the R666, detector noise was found to be greater than with the R636; therefore, an
average of 16 scans of 2 ms intervals was used to collect the data which gave a 32 ms
measurement time. In this regard, the detection limit was 360,000 molecules per
measurement interval.
Experiment #4 The final experiment presented in Table 11 was essentially
the same as experiment #3 except the R636 was used for detection and the analog
mode was used for data collection. Again, signal was found to be independent of
flow rate, and laser power of 240 mW was used in the analysis. Figure 33 is a plot

141
of the dependence of the signal to noise ratio on laser power (at 3.5 x 10'5 cm2 focus
area) of flowing 4.62 nM IR 140 solution in methanol. As exhibited, the longer time
constant (tc) of 1 s (set by a 1 Hz low pass filter) gives a higher signal to noise ratio
than the 300 ms time constant of the chart recorder. For this reason, the longer time
constant was used in the analysis. As shown in Table 11, the concentrational limit
of detection was 9.5 x 10'12 M or 800 molecules in the Vp. After accounting for the
tr of 1.6 ms and rc of 1 s, the LOD becomes 500,000 IR 140 molecules. The log-log
calibration curve for this experiment appears in Figure 34.
Noises of the System
Table 12 lists the noise parameters for the analytical experiments discussed
above. As mentioned previously, laser scatter was by far the limiting source of noise
in these experiments. Due to the overwhelming magnitude of the laser scatter, the
effect of Raman scatter from the solvent and capillary could not be quantitated.
Background fluorescence was not expected to occur, but again, this aspect could not
be quantitated based on the limitations caused by laser scatter.
Scatter bv the capillary. In an attempt to quantify the amount of light
scattered by the capillary, the power of the laser before and after striking the
capillary was measured. At a point between the focusing lens and capillary, laser
power was 205 mW, and the power was 105 mW in the same path on the other side
of the capillary. This means that = 50% of the laser light was scattered overall. By
observing the scatter with a near-infrared viewer, it appeared that the scatter was

600
550
500
450
400
350
300
250
200
150
100
50
0
i. ]
1 s time constant
50 100 150 200 250 300 350 400 450
Laser Power (mW)
'feet of laser power (focused to 3.5 x 10'5 cm2 area) on the signal/noise ratio of a 4.62 nM IR 140
lution in methanol flowing in the 50 ¿un i.d. capillary at 5 ¿iL/min at rc = 1 and 0.3 s. This study was
¡rformed in the power/flow optimization for experiment 4.
-e-
N>

Signal (nA)
Figure 34. Calibration curve for experiment #4. Conditions presented in Table 11.
u>

144
Table 12. Instrumental noises of the approach to single molecule detection for
the experiments listed in Table 11.
Parameter
Experiment
#1
#2
#3
#4
Detector
R636
R636
R666
R636
Signal
Collection
Analog
Photon
Counting
Photon
Counting
Analog
Pre-amplifier
< 0.05 pA
1.1 mV
0.5 mV
< 0.05 pA
PMT in Dark
18
± 4pA
“0.126
± 0.014
‘0.41
± 0.028
12.6
+ 1.0 pA
Blank
140
± 7.2 pA
“4900
+ 150
*4.0
± 0.52
450
± 4 pA
Laser Power (mW)
280 ± 3
80 ± 1
240 ± 2.5
240 ± 2.5
“Counts per 2 ms.
most intense near the transmitted beam, but the light was still very intense in a 360°
ring around the capillary. Assuming that the light is scattered uniformly (not a valid
assumption), the scatter collected by the microscope objective would be 476
photelectrons based on the 11.9% solid angle of collection of the 100 mW scattered
by the capillary. Judging by the large amount of scatter produced by the capillary,
it is not difficult to realize why scatter was such a formidable problem in the real
situation. Based on experimental evidence of the MVF presented earlier, the scatter
must have passed around the MVF and reached the detector. Ways to avoid this
problem are presented in Chapter 4.

145
Discussion of Limits of Detection
Method of LOD calculation. To determine the LODs for the experiments
listed in Table 11, the IUPAC approach given in Chapter 1 was followed. The noise
was determined by taking the average and standard deviation of 16 separate
measurements (photon counting) or by taking 1/5 of the peak-to-peak noise of the
blank over a period of several minutes. The statistical method also described in
Chapter 1 involving Poission distributions was not necessary based on the great many
counts per time interval and inability to attain near-SMD concentrations. If one
wishes to know the LOGD for any of these analyses, the stated LODs can be
doubled to give LOGD.
Once the LOD has been determined in a flowing sample, interpreting the
results and asserting which approach is best is subject for debate. As Table 11 shows,
the best concentration LOD (the lowest) tends to result in poor LOD during the time
constant. This is because longer time constants result in lower noise, but more
molecules traverse the detection region in the measurement period. For SMD, the
measurement interval must be equal to or less than the residence time, but how long
should the measurement time be when SMD is not achieved? This problem can lead
to imaginative mathematical manipulations to determine the best way to present the
LOD, but essentially, the fact remains that stray light was affecting the experiments
with the capillary to generally the same extent. The differences in the values
presented are more or less variations in the way in which the data was collected.

146
Comparison of LODs In an actual analysis, the method of data collection
depends on the type of analysis performed. The LOD determined with the
monochromator (with an LOD of 300 fM in the 3.75 mL flowcell) would be the best
approach if a sample of low concentration and large volume required analysis for one
of the polymethine dyes. However, this method would not be useful as a method of
detection for transient signals in analytical separations or analysis of limited sample
volumes. In these cases, the best method of approach would be to perform the
analysis in the capillary (or other small volume flowcell). In SMD experiments of a
static sample, longer measurement times are advantageous provided the analyte does
not decompose, but for SMD in a flowing solution, longer measurement times are
a luxury that cannot be afforded. Therefore, despite the lower concentrational LOD
obtained with longer rc, the best approach for future analyses is to keep the fast
measurement time and work to lower the LOD through reduction of actual noise,
and not through the electronic reduction of noise after data collection.

CHAPTER 4
CONCLUSIONS AND FUTURE WORK
Conclusions
This project began confidently as a novel method to detect single molecules
in a flowing solution contained in a capillary. It was strongly believed that SMD
would be possible using the system described in this dissertation based on the reports
of near-SMD by other researchers using approaches that do not reach the intrinsic
noise level.2-6,43^5,54,55 Most notably, the ongoing work at Los Alamos,2'6,54'55 as close
as it is to detecting single molecules in the probe volume, is far from being able to
achieve such results with 100% sampling efficiency. Based on these limitations,
Keller’s approach cannot be used in the proposed DNA sequencing method because
if the identification of one nucleotide is missed, the entire DNA chain from that
point is out of sequence.
For this reason, the approach described here is designed to resist any
reduction in sampling efficiency for the sake of lowering LOD. Furthermore,
capillary electrophoresis has become the accepted method for DNA sequencing. It
is possible to transfer the flow from a capillary into a sheath flow cuvette, but this
is a complicated process that leads to a poorer resolution of separation.89,90 It would
be advantageous if detection could be performed in the capillary itself, despite its
147

148
problem with greatly increased scatter. Therefore, it was decided that using a sample
container other than the capillary is not an option for this project. It was realized
when these two decisions were made (100% sampling efficiency and detection in a
capillary) that it would be much more difficult to reach the intrinsic noise limit and
a detection efficiency near unity, but it was felt that the metal vapor filter would
overcome these added difficulties.
Based on the theoretical treatment of the parameters in this approach, it was
thought that the metal vapor filter would be able to perform as predicted and SMD
would be achieved. With the best measurement that could be made with the
available instrumention in the lab, the MVF did perform as predicted, but the
practical aspects of blocking scattered laser light not passing through the MVF was
not given as much attention as necessary. The limiting source of noise in
experimental work in this dissertation was a result of scattered light.
In this respect, this project has been both satisfying and disappointing. It is
satisfying that the Ti:sapphire laser is a narrow enough source to fall within the
absorption band of an atomic transition, and that the Rb MVF is able to absorb the
direct full power of the Tirsapphire laser output beyond the point of measurability,
but it is disappointing to still be limited by laser scatter. However, there is little
doubt that with the appropriate refinements, discussed above, this project should be
able to eliminate laser scatter as a source of noise.
An LOD of 800 molecules continuously flowing through a 140 pL volume is
the lowest detection limit ever attained in this laboratory, but it is still much higher

149
than the detection limits obtained by other researchers attempting SMD. Of course,
they were not performing detection in a capillary, and only two other known reported
LODs in a capillary are lower than this.90,98
Comparison with Keller’s work. In the project presented here, a relatively
short period of time was spent with one graduate student doing the great majority
of the work. During eight years of effort at Los Alamos National Laboratory,
Keller’s team of scientists have improved their detection limit from 33,000 molecules
with 60% sampling efficiency2,54 to 33 molecules with 10% sampling efficiency5 (Ref.
4 involved a different approach, and as in Ref. 6, the concentration LOD was not
reported). With this in mind, it is believed that the continuous-wave Ar+ laser
approach of Keller has little chance of achieving true SMD. On the other hand, in
the initial stages with the technique reported in this dissertation, a detection limit of
260,000 IR 140 molecules (based on the same criteria as above) with 100% sampling
efficiency was obtained. Again, it must be stressed that the limiting noise of the
experiment, laser scatter, is orders of magnitude higher from a capillary than from
the sheath flow cuvette. With the modifications to this approach presented below,
barring unforeseen impediments, it should be possible to achieve single molecule
detection in a flowing sample contained in a capillary.

150
Future Work
Elimination of Scatter
Based on the results of the experiments designed to reach the single molecule
detection limit, there is still much room for improvement in the approach. Currently,
the most pressing concern with the system is light scattered around the MVF
reaching the detector, and the single most important improvement which can be
made is the total blockage of light not passing through the MVF. In the design of
the system, it was believed that this scatter would be totally blocked by the aperture
before the focusing lens, but due to the many reflective surfaces (heating tape, front
and back faces and inner walls of the MVF and microscope tube), the scatter was a
formidable problem.
To solve this problem, the way in which light is passed to the detector must
be redesigned. The best way to accomplish this is to surround the PMT with metal
vapor, but the requirement of heated vapor positioned near a cooled tube makes this
a difficult concept to implement. An easier way to block scatter is to include a series
of apertures or baffles down the microscope tube. Figure 35 is a drawing of this
proposed method of collecting the fluorescence. The size of the optimum apertures
before and after the MVF can be determined by using adjustable irises. In this
approach, it is important that the apertures still pass nearly all of the collected
fluorescence from the laser probe volume (unlike Keller’s approach) or else sampling
efficiency is compromised. Furthermore, the MVF would have to be fitted in the

Figure 35. Suggested design of the light passage between the microscope objective and photomultiplier tube for
in future experiments to achieve single molecule detection.

152
tube so that no light would pass betwen it and the tube wall. In this design, the
heating tape (or small tube furnace) would be placed outside of the tube.
Longer MVF path length. The reason that the MVF cell path length was 2
inches in this approach is that the spacing of the microscope tube and PMT housing
could accommodate only this length. Since the total distance between the
microscope objective and PMT is 10.6 inches, there is certainly enough space for a
longer cell, and the new design would allow for this change. Since absorbance is
directly proportional to path length, and number density is independent of this
parameter, doubling the cell length from 2 inches to 4 inches would double the
absorbance of from 10 to 20 at 100°C (theoretical absorbance of cell #1). In this
way, light rejection capability of the MVF can be greatly magnified with little trouble.
Furthermore, a longer cell would also require less heating to achieve the same
absorbance and put less thermal stress on the microscope objective.
Reduction of the collection of scattered light. The elimination of the
extraneous scattered light alone should allow achievement of the intrinsic noise limit
for the system. However, if scattered light is still detectable after the above proposed
changes have been made in the set-up, other techniques are available that can greatly
reduce the amount of laser scatter collected by the microscope objective. As stressed
earlier in the dissertation, capillaries generate much more scatter than other methods
of containing the sample. As described, this approach relied upon the MVF to
totally remove the scatter (the laser could be directed straight through and not be

153
detected) so nothing was done to limit the collection of scatter by the microscope
objective.
Tilting the capillary. The simplest way to reduce the amount of light scattered
into the collection optics is to tilt the angle of the capillary along the viewing axis of
the microscope objective. Since the laser scatter occurs predominantly
perpendicularly to the axis of the capillary, the capillary can be tilted such that the
bulk of the scatter passes above (or below depending on the tilt direction of the
capillary) the microscope objective. Concurrently, the fluorescence from the flowing
solution occurs isotropically; therefore the fluorescence is still viewed, albeit at a
slightly reduced collection efficiency. This is a common technique in fluorescence
measurements in capillaries.90,98
In the experiments presented in this dissertation, the reason that the capillary
was kept vertical to the laser and microscope objective is that a reduction in the solid
angle of collection could not be tolerated and still be able to achieve SMD. Based
on the optimum distance of the capillary from the microscope objective and the
aperture of the objective, the capillary could not be tilted to the degree necessary to
send the scatter out of the field of view of the microscope objective and still maintain
the same solid angle of collection. There simply was not space enough for the
capillary to be tilted to the extent required with the microscope objective currently
used in the system.
Spatial filter. A second possibility for the reduction of scatter collected by the
microscope objective is to use a spatial filter. As shown in Figure 22, the field of

154
view of the microscope objective is approximately 2 mm at the focus; considering that
the size of the signal in the capillary is a rectangle 0.05 mm wide and 0.07 mm long,
most of the collected region is not useful information. A spatial filter (simply a
rectangle with the dimensions proportional to the Vp) could be placed between the
the capillary and microscope objective. In this way, the scatter coming from the
outer walls of the capillary and other external sources in the box would be blocked
before entering the microscope objective. Other SMD approaches use the spatial
filter,2'6,43^5’54,55 which typically is placed after the microscope objective and defines
the size of the Vp. To maintain 100% sampling efficiency, the spatial filter must
define a region equal to or larger than the width of the capillary. The
implementation of the spatial filter would require careful alignment of the collection
optics, but this is not a great difficulty.
Immersion cell. If scatter persists after attempting the previous methods of
reduction, the capillary could be contained in an immersion cell. An immersion cell
is a specially designed cell containing a fluid of the same refractive index as the
quartz of the capillary.91,92 The cell envelops the capillary, and since specular scatter
only arises when light encounters a change in refractive index, the scatter from the
outer walls of the capillary is reduced.
One difficulty of applying this concept is finding a fluid that matches the
refractive index of quartz at the laser wavelength. The fluid (most likely a mixture
of substances) must not absorb light at the laser wavelength, nor at wavelengths of
the fluorescence emission. Additionally, the immersion cell is difficult to design.

155
The front surface of the cell should be flat to minimize scatter of the laser entering
the cell, and due to the space requirements of the collection optics, the immersion
cell would have to be designed to incorporate the lens of the microscope objective
inside the fluid. It is not uncommon for objectives to be designed to interface with
a fluid. In fact, this situation is preferable because two changes in refractive index
are eliminated if the light does not traverse air between the object and objective.
However, due to the intracies of the design and manufacture of such an immersion
cell, it would be used only as a last resort in this project.
Other Future Improvements
Other future improvements of the experiment are trivial in comparison to the
light scattering problem, but they should be addressed anyway. These may become
more important if extraneous laser scatter can be eliminated and the intrinsic noise
level is reached.
Photon counting. First of all, the photon counting instrument used in this
approach is inadequate to achieve true SMD because of the lack of continuous
collection of the detector signal. The counter is able to count for periods as small
as 100 ns, but the 2 ms minimum dwell time (interval between counting times) would
only allow 1 point per molecule at the 1.6 ms maximum residence time of a single
molecule in the probe volume at the minimum pump setting (increasing tr is another
option discussed below). Furthermore, the noise of the preamplifier used in
conjunction with the photon counter is not ideal for use with the R636 PMT. The

156
R636 was found to be the best photomultiplier tube for the project based on its noise
equivalent power of 0.2 fW (versus 1.4 fW for the R666), but due to the low gain of
the R636, the noise of the pre-amplifier became limiting. Due to the drifting voltage
level of the pre-amplifier, the optimum discrimination level would have to be re-set
every few hours. The R666 has sufficient gain to be used with the photon counter,
but since detector noise is an intrinsic source of noise, the R666 may limit the ability
to detect single molecules. A head-on, red-sensitive PMT, such as the Hamamatsu
R943-02 suitable for photon counting would be best.
Control of flow. Another option would be to increase tr (lower flow rate) to
allow for the production of more fluorescence and longer counting times, but due to
decomposition, the residence time should not be typically more than 10 ms at near¬
saturation laser irradiance. Syringe pumps or capillary electrophoresis are the best
methods of solvent flow able to achieve the flow rates necessary for a tr = 10 ms in
a narrow capillary. In this case, 2.5 data points on average would be collected per
molecule with the photon counter set at 2 ms counting periods (and 2 ms minimum
dwell time). Ideally, it would be beneficial to apply one of the two methods of
pumping and use a better photon counting system.
Laser focusing. The manner of focusing in this set-up, though functional, was
not user friendly. The several directions of movement necessary to position the laser
and capillary when the microscope objective and detector were fixed make optimum
focusing arduous. Conversely, if the the laser was fixed in space and the capillary
placed at the optimum focus, it would be easier to position the microscope objective

157
and detector (spaced apart so the focused light collected by the objective matches the
size of the photocathode) to maximize signal/noise. In this way, only one set of
micropositioners and fewer manipulations would be necessary. The reason this was
not done in previous experiments is that the base of the microscope and detector
housing were too bulky for the micropositioners. Recently, a large micropositioner
was found that may be able to manipulate the microscope objective/PMT
combination.
Holding the capillary. Also, a more rigid means of holding the capillary in
place could be developed. Due to the small size of the laser focus, any variation in
the position of the capillary caused significant change in the scatter and signal. By
relieving the capillary of adjustable movement, as discussed above, positioning would
be more stable. Furthermore, the holding prongs for the capillary could be spaced
closer together (and still leave room for the microscope objective) to minimize
oscillations and the associated noise.
Ar+ laser power fluctuation. A final source of noise that should be minimized
if possible is a long-term power drift in the Ar+ laser, the origin of which is unknown.
The frequency of the cycling power fluctuations is 8 - 10 minutes, and the operation
of possible sources of this drift related to the instrumentation at the set-up (heat
exchangers) was found not to be correlated with the laser fluctuations. This noise
is possibly caused by the air handling system of the building. It is conceivable that
the thermal expansion and contraction of the optics and table defocused the laser
slightly and produced power drift. Operating the laser in the constant power mode

158
should have corrected this problem, but the fluctuation was still as great as 1%.
Although the drift was not cause for great concern in the experiments presented in
this dissertation, laser power stability is a source of intrinsic noise that may become
very important if the problem with laser scatter is solved.
Black body radiation. A final factor that was discerned as a potential problem
in experiments when the MVF was heated to high temperatures (> « 130 °C) is
black body radiation. The red-sensitive PMT was able to detect the near-infrared
emission of matter when long pass filter was used. This was not significant at MVF
temperatures < 120°C or when an interference filter, which absorbed radiation >
840 nm, was placed before the PMT. In future experiments, black body radiation
would not be a significant factor with the use of a bandpass filter (which would also
block the methanol Raman band at 867 nm with 795 nm excitation) and a longer
MVF discussed previously that requires lower temperature to achieve the necessary
absorbance to achieve SMD.

REFERENCE LIST
1. J.H. Jett, R.A. Keller, J.C. Martin, B.L. Marrone, R.K. Moyzis, R.L. Ratliff,
N.K. Seitzinger, E.B. Shera and C.C. Stewart,/. Biomol. Struct. Dynam.,2,301
(1989).
2. N.J. Dovichi, J.C. Martin, J.H. Jett, M. Trkula and R.A. Keller, Anal. Chem.,
56, 348 (1984).
3. D.C. Nguyen, R.A. Keller and M. Trkula, /. Opt. Soc. Am B, 4, 138 (1987).
4. E.B. Shera, N.K. Seitzinger, L.M. Davis, R.A. Keller and S.A. Soper, Chem.
Phvs. Lett.. 174. 553 (1990V
5. J.H. Hahn, S.A. Soper, H.L. Nutter, J.C. Martin, J.H. Jett and R.A. Keller,
Appl. Spectrosc.. 45. 743 (1991).
6. S.A. Soper, E.B. Shera, J.C. Martin, J.H. Jett, J.H. Hahn, H. L. Nutter and
R.A. Keller, Anal. Chem., 63, 432 (1991).
7. J.M. van Emon, Ed., Immunochemical Assays for Environmental Analysis,
American Chemical Society, Washington, DC, 1990.
8. H.W. Shapiro, Practical Flow Cytometry, Less, New York, 1988.
9. J.W. Jorgenson and K.D. Lukács, Anal. Chem., 53, 1298 (1981).
10. J.W. Jorgenson, Anal. Chem., 58, 743A (1986).
11. P. Kucera, Ed., Microcolumn High-Performance Liquid Chromatography,
Elsevier, New York, 1984.
12. N.J. Dovichi, F. Zarrin, T.G. Nolan and D.J. Bornhop, Spectrochim. Acta. 43B.
639 (1988).
13. N. Omenetto and J.D. Winefordner, CRC Crit. Rev. Anal. Chem., J3, 59
(1981).
159

160
14. P.J. Elving, V. Krivan and I.M. Kolthoff, Eds., "Nuclear Activation and
Radioisotopic Methods of Analysis," Vol. 14, Part I, Treatise on Analytical
Chemistry, 2nd Ed., John Wiley, New York, 1986.
15. J.T. Watson, Introduction to Mass Spectrometry, 2nd Ed., Rasen, New York,
1985.
16. H. Otani, R.J. Wilson, S. Chiang and C.M. Mate, Phys. Chem. Lett., 60, 2398
(1988).
17. S.M. Lindsay, talk presented in the Division of Analytical Chemistry at the
200th National Meeting of the American Chemical Society, Washington, DC,
August, 1990.
18. R.M. Metzger and C.A. Panetta, New J. Chem.. 15. 209 (1991).
19. Science. 254. 1707 (1991).
20. C. Joachim, New J. Chem., JL5, 223 (1991).
21. N. Omenetto, B.W. Smith and J.D. Winefordner, Spectrochim. Acta, 43B, 1111
(1988).
22. J.D. Winefordner, B.W. Smith and N. Omenetto, Spectrochim. Acta. 44B. 1397
(1989).
23. C.Th.J. Alkemade, "Detection of Small Numbers of Atoms and Molecules" in
Analytical Applications of Lasers, E.H. Piepmeier, Ed., John Wiley, New York,
1986.
24. C.Th.J. Alkemade, Appl. Spectrosc.. 35. 1 (1981).
25. G.S. Hurst, J. Chem. Ed., 59, 895 (1982).
26. G.M. Hieftje, J. Chem. Ed., 59, 900 (1982).
27. H. Falk, Prog. Anal. Atom. Spectrosc., 3, 181 (1980).
28. C.L. Stevenson and J.D. Winefordner, Chemtracts - Anal Phys. Inorg. Chem.,
2, 217 (1990).
29. C.L. Stevenson, Special Considerations in Estimating Detection Limits, Ph.D.
Dissertation, University of Florida, 1991.

161
30. C.L. Stevenson and J.D. Winefordner, Appl. Spectrosc., 45, 4161 (1991).
31. C.L. Stevenson and J.D. Winefordner, Appl Spectrosc., in press (1992).
32. C.L. Stevenson and J.D. Winefordner, Appl. Spectrosc., in press (1992).
33. D.D. Ebbing, General Chemistry, 2nd Ed., Houghton Mifflin, Boston, 1987, p.
31.
34. IUPAC, Pure Appl. Chem., 45, 101 (1976).
35. H. Kaiser, Z. Anal. Chem., 209, 1 (1965). Translated by A.C. Menzies in The
Limit of Detection of a Complete Analytical Procedure, Hofner, New York,
1969.
36. W.J. Dixon and F.J. Massey, Introduction to Statistical Analysis, 4th Ed.,
McGraw-Hill, New York, 1983.
37. E.C. Molina, Poisson’s Exponential Binomial Limit, 4th Ed., Van Nostrand,
New York, 1947.
38. L.A. Curie in Detection in Analytical Chemistry: Importance and Practice, L.A.
Curie, Ed., American Chemical Society, Washington, DC, 1988, ch. 1.
39. J.D. Ingle, Jr. and S.R. Crouch, Spectrochemical Analysis, Prentice Hall,
Englewood Cliffs, NJ, 1988.
40. Melles Griot, Optics Guide 5, Melles Griot, Irvine, CA, 1990.
41. T.E. Barber, Diode Laser Atomic Spectrometry, Ph.D. Dissertation, University
of Florida, 1992.
42. D.J.S. Birch, G. Hungerfold, B. Nadolsi, R.E. Imhof and A.D. Dutch, J. Phys.
E: Scl Instrum.. 21. 857 (1988).
43. W.B. Whitten, J.M. Ramsey, S.A. Arnold and B.V. Bronk, Anal. Chem., 63,
1027 (1991).
44. R.A. Mathies, K. Peck and L. Stryer, Anal. Chem., 62, 1786 (1990).
45. N.J. Dovichi, J.C. Martin, J.H. Jett and R.A. Keller, Science, 219. 845 (1983).
46.
V.S. Letokhov, Laser Photoionization Spectroscopy, Academic Press, Orlando,
1987.

162
47. G.I. Bekov, V.S. Letokhov, O.I. Matveev and V.I. Mishin, Sov. Phys. JETP, 48,
1052 (1978).
48. G.S. Hurst, M.H. Nayfeh and J.P. Young, Appl. Phys. Lett., 30, 229 (1977).
49. T. Hirschfeld, Appl Opt., 15, 2965 (1976).
50. B. Kirsch, E. Voigtman and J.D. Winefordner, Anal. Chem., 51, 2007 (1985).
51. W.E. Moerner and L. Kador, Phys. Rev. Lett., 62, 2535 (1989).
52. W.P. Ambrose and W.E. Moerner, Nature, 349. 225 (1991).
53. W.E. Moerner, New J. Chem.. 15. 199 (1991).
54. D.C. Nguyen, R.A. Keller, J.H. Jett and J.C. Martin, Anal. Chem., 59 2158
(1987).
55. K. Peck, L. Stryer, A.N. Glazer and R.A. Mathies, Proc. Natl. Acad. Set USA,
86, 4087 (1989).
56. P.A. Johnson, T.E. Barber, B.W. Smith and J.D. Winefordner, Anal. Chem.,
61, 861 (1989).
57. S.J. Lehotay, P.A. Johnson, T.E. Barber and J.D. Winefordner, Appl
Spectrosc.. 44. 1577 (1990).
58. C. Th. J. Alkemade, Tj. Hollander, W. Snelleman and P.J. Th. Zeegers, Metal
Vapors in Flames, Pergamon Press, Elmsford, NY, 1982.
59. H.M. Gibbs and R.J. Hill, Phys. Rev., 153, 132 (1967).
60. H. Kostkowski and A. Bass, J. Opt. Soc. Am., 46, 1060 (1956).
61. A. Gallagher and E.L. Lewis, J. Opt. Soc. Am., 63, 864 (1973).
62. O. Svelto, Principles of Lasers, Plenum Press, New York, 1989.
63. P.A. Johnson, The Semiconductor Laser Diode as an Excitation Source for
Analytical Atomic and Molecular Fluorescence Spectroscopy, Ph.D. Dissertation,
University of Florida, 1989.
64. W.R. Rapoport and C.P. Khattak, Appl. Opt.. 27. 2677 (1988).

163
65. R. Indralingam, Evaluation of Narrow Line Atomic Filters and the Development
of a Metal Vapor Filter for Raman Spectroscopy, Ph.D. Dissertation, University
of Florida, 1991.
66. R. Indralingam, J.B. Simeonsson, G.A. Petrucci, B.W. Smith and J.D.
Winefordner, Anal. Chem., in press (1992).
67. G. Patonay and M.D. Antoine, Anal. Chem., 63, 321A (1991).
68. R.C. Benson and H.A. Kues, /. Chem. Eng. Data, 22, 379 (1977).
69. D. Andrews-Wilberforce and G. Patonay, Appl. Spectrosc., 43, 1450 (1989).
70. O.G. Bjornsson, R.Murphy and V.S. Chadwick, Experientia, 38, 1441 (1982).
71. J.P. Webb, F.G. Webster and B.E. Plourde, IEEE J. Quantum Electron.. 11.
114 (1975).
72. M.R. Tripp, C.R. Swayze and I J. Fox, in Dye Curves, D.A. Bloomfield, Ed.,
University Park Press, Baltimore, 1974, Ch. 18.
73. R.C. Benson and H.A. Kues, Phys. Med. Biol, 23, 159 (1978).
74. M.A. Roberson, D. Andrews-Wilberforce, D.C. Norris and G. Patonay, Anal.
Lett, 23, 719 (1990).
75. D. Andrews-Wilberforce and G. Patonay, Spectrochim. Acta. 46A. 1153 (1990).
76. P.L. Rappoport and J.J. Thiessen, /. Pharm. Sci.,11, 157 (1982).
77. B. Hollins, B. Noe and J.M. Henderson, Clin. Chem., 33, 765 (1987).
78. M.B. Dorr and G.M. Pollack, J. Pharm. ScL, 78, 328 (1989).
79. S.J. Lehotay, A.M. Pless And J.D. Winefordner, Anal. ScL, in press (1992).
80. E.S. Yeung and W.G. Kuhr, Anal. Chem., 63, 275A (1991).
81. T. Imasaka and N. Ishibashi, Anal. Chem., 62, 363A (1990).
82. J. Gathje, R.R. Steuer and K.R.K. Nicholes, J. Appl. Physiol., 29, 181 (1970).
83. W. West and S. Pearce, J. Phys. Chem., 69, 1894 (1965).

164
84. E.H. Trinh and C.J. Hsu, J. Acoust. Soc. Am., 80, 1757 (1986).
85. A. Ashkin, Phys. Rev. Lett.. 24. 156 (1970).
86. S. Arnold and N. Hessel, Rev. Sci. Instrum., 56, 2066 (1985).
87. R. Williams and J.R. Matey, Appl. Phys. Lett., 52, 751 (1988).
88. J.W. Lyons and L.R. Faulkner, Anal. Chem., 54, 1960 (1982).
89. Y.F. Cheng, S. Wu, D.Y. Chen and N.J. Dovichi, Anal Chem., 62, 496 (1990).
90. H. Swerdlow, J.Z. Zhang, D.Y. Chen, H.R. Harke, R. Grey, S. Wu, N.J.
Dovichi and C. Fuller, Anal. Chem.. 63. 2835 (1991).
91. Y. Kurosu, T. Sasaki and M. Saito, /. High Res. Chromatogr., 14, 186 (1991).
92. A.E. Bruno, B. Kraittiger, F. Maystre and H.M. Widmer, Anal. Chem., 63,
2689 (1991).
93. A.E. Bruno, E. Gassmann, N. Pericles and K. Anton, Anal. Chem., 61, 876
(1989).
94. D.R. Sage and S.O. Farwell, Anal. Chem.. 53. 2123 (1981).
95. D.C. O’Shea, Elements of Modem Optical Design, John Wiley & Sons, New
York, 1985, Ch. 7.
96. R.W. Engstrom, Photomultiplier Handbook, RCA, Lancaster, PA, 1980.
97. N. Ayala, Resonance Line Laser as Excitation Source for Atomic Spectrometry,
Ph.D. Dissertation, University of Florida, 1992.
98. L. Hernandez, J. Escalona, N. Joshi and N. Guzman,/. Chromatoer.. 559. 183
(1991).

BIOGRAPHICAL SKETCH
Steven John Lehotay was born in Orlando, Florida, on February 4, 1965. In
June, 1983, he graduated from Maynard Evans High School in Orlando, and
afterwards, enrolled at the University of Florida where he earned a B.S. degree in
chemistry in May of 1987. During his undergraduate years, Steve met Joann Basile,
and they were married on July 9, 1988, in Gainesville. From August 1987 to January
1992, he attended graduate school at the University of Florida and studied analytical
chemistry in the research group of Dr. James D. Winefordner. In May of 1992, he
received a Ph.D. in chemistry with a minor degree in environmental engineering
sciences. He began his postdoctoral career at the United States Department of
Agriculture Research Laboratory in Beltsville, Maryland.
165

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
¡/James D. Winéfordner, Chair
Graduate Research Professor of Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
j&JL-AA-
Richard A. Yost
Professor of Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
\)C* ^A.CA - l ^ -
Vaneica Y. Youb{| /] j
Associate Professor ofChemistr/
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Eric R. Allen
Professor of Environmental Engineering
Sciences
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
David P. Chynovveth
Professor of Agricultural Engineering

This dissertation was submitted to the Graduate Faculty of the Department
of Chemistry in the College of Liberal Arts and Sciences and to the Graduate School
and was accepted as partial fulfillment of the requirements for the degree of Doctor
of Philosophy.
May, 1992
Dean, Graduate School

UNIVERSITY OF FLORIDA
262 08285 410




PAGE 1

$3352$&+,1* 6,1*/( 02/(&8/( '(7(&7,21 %< /$6(5,1'8&(' )/825(6&(1&( 2) )/2:,1* '<( 62/87,216 ,1 $ &$3,//$5< %\ 67(9(1 -2+1 /(+27$< $ ',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,'$ 81,9(56,7< 2) ),25(n"

PAGE 2

7KLV GLVVHUWDWLRQ LV GHGLFDWHG WR P\ IDWKHU $QGUHZ / /HKRWD\ 7KRXJK KH GLHG KH VWLOO OLYHV LQ PH

PAGE 3

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

PAGE 4

, DP DOVR JUDWHIXO IRU WKH ZRUN RI VHYHUDO RWKHUV LQ WKH JURXS 0LNH :HQVLQJ ZURWH WKH SURJUDP WKDW FDOFXODWHV WKH 9RLJW SURILOHV IRU WKH PHWDO YDSRU FHOOV 1DQF\ 3HWUXFFL WRRN WKH 5DPDQ VSHFWUD RI WKH VROYHQWV
PAGE 5

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

PAGE 6

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

PAGE 7

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b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

PAGE 8

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

PAGE 9

&+$37(5 ,1752'8&7,21 $1' 7+(25< ,QWURGXFWLRQ 7KH VLQJXODU SXUSRVH RI WKLV UHVHDUFK LV WR GHWHFW D VLQJOH PROHFXOH LQ D FKHPLFDO VROXWLRQ ,I RQH GHILQHV VLQJOH PROHFXOH GHWHFWLRQ 60'f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

PAGE 10

7KH SXUSRVH RI WKLV GLVVHUWDWLRQ LV WR f SUHVHQW WKH WKHRUHWLFDO FULWHULD IRU 60' f UHYLHZ SUHYLRXV UHVHDUFK RQ WKH VXEMHFW f GLVFXVV WKH GHYHORSPHQW RI D QRYHO DSSURDFK WR 60' f VKRZ WKH UHVXOWV RI H[SHULPHQWV GHVLJQHG WR DWWDLQ 60' DQG f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f QXFOHRWLGHV ZLWK IRXU GLIIHUHQW KLJKO\ IOXRUHVFHQW G\H PROHFXOHV WKHQ GHWHFWLQJ HDFK RI WKHVH PROHFXOHV DV WKH\ SDVV WKURXJK D YROXPH SUREHG E\ D WLJKWO\ IRFXVHG DUJRQ LRQ ODVHU ,I VXFFHVVIXO WKH VFKHPH LV SURMHFWHG WR VHTXHQFH '1$ WLPHV IDVWHU WKDQ FXUUHQW WHFKQLTXHV $OWKRXJK LW UHPDLQV WR EH VHHQ LI WKLV DSSOLFDWLRQ FDQ EH DFFRPSOLVKHG DV SURSRVHG UHVHDUFK FRQFHUQLQJ WKH GHWHFWLRQ DVSHFWV RI WKH

PAGE 11

SURMHFW KDYH FRPH YHU\ FORVH WR 60'n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

PAGE 12

DQG SODFH IURP WKH LQMHFWLRQ SRLQW PD\ UHTXLUH WKH GHWHFWLRQ RI WKH WUDFHU LQ YHU\ VPDOO TXDQWLWLHV SRWHQWLDOO\ VLQJOH PROHFXOHV 7KHVH NLQGV RI VWXGLHV FRXOG EH FDUULHG RXW LQ D V\VWHP RQ D JOREDO VFDOH VXFK DV WKH VWXG\ RI RFHDQ FXUUHQWV RU ZLQG SDWWHUQVf RU RQ D PLFURVFDOH VXFK DV FHOO WUDQVSRUWf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fV JURXS PDGH D PDMRU DGYDQFH LQ VHSDUDWLRQ VFLHQFH ZLWK WKH GHYHORSPHQW RI FDSLOODU\ ]RQH HOHFWURSKRUHVLV &=(f &=( LV FDSDEOH RI DFKLHYLQJ VHSDUDWLRQ HIILFLHQFHV DV ODUJH DV WKHRUHWLFDO SODWHV DQG DOORZV IRU WKH VHSDUDWLRQ RI VHYHUDO W\SHV RI FKHPLFDOV LQ DTXHRXV VROXWLRQ SDUWLFXODUO\ ELRORJLFDO VSHFLHV WKDW ZHUH QRW HDVLO\ VHSDUDWHG SUHYLRXVO\ 6LPLODUO\ LW KDV EHHQ NQRZQ IRU VRPH WLPH WKDW GHFUHDVLQJ FROXPQ GLDPHWHU RU SDUWLFOH VL]Hf LQ OLTXLG FKURPDWRJUDSK\ LQFUHDVHV VHSDUDWLRQ HIILFLHQF\ :LWK WKH UHFHQWO\ GHYHORSHG FRPPHUFLDO DELOLW\ WR FRDW YHU\ QDUURZ FDSLOODULHV ZLWK VWDWLRQDU\ SKDVH PDWHULDO PLFURFROXPQ KLJKSHUIRUPDQFH OLTXLG FKURPDWRJUDSK\ +3/&f KDV EHFRPH D YLDEOH PHWKRG IRU VHSDUDWLRQ DQDO\VLV +RZHYHU GXH WR WKH VPDOO TXDQWLWLHV LQYROYHG LQ PLFURFROXPQ +3/& DQG &=( GHWHFWLRQ LV RIWHQ WKH OLPLWLQJ IDFWRU LQ DSSOLFDWLRQV RI WKHVH WHFKQLTXHV %\ WKHLU

PAGE 13

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

PAGE 14

60' LV DQ HQG LQ LWVHOI ,I WKLV SURMHFW LV VXFFHVVIXO WKH DQQRXQFHPHQW RI 60' WR VFLHQWLVWV LQ YDULRXV GLVFLSOLQHV ZRXOG SRVVLEO\ FDXVH WKH GHVLJQ RI H[SHULPHQWV WR VXLW WKH QHHGV RI WKH GHWHFWRU ,W LV GLIILFXOW WR SUHGLFW ZKDW RXWFRPHV ZRXOG DULVH E\ DFFRPSOLVKLQJ 60' EXW WKH SRVVLELOLWLHV DUH YHU\ H[FLWLQJ &KRLFH RI $QDO\WLFDO 7HFKQLTXH IRU 6LQJOH 0ROHFXOH 'HWHFWLRQ 2I DOO DQDO\WLFDO PHWKRGV FXUUHQWO\ LQ XVH ODVHULQGXFHG IOXRUHVFHQFH /,)f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

PAGE 15

0DVV VSHFWURPHWU\ 2Q WKH RWKHU KDQG PDQ\ WHFKQLTXHV XWLOL]LQJ PDVV VSHFWURPHWULF GHWHFWLRQ KDYH VXIILFLHQW VLJQDOWRQRLVH UDWLRV WR GHWHFW VLQJOH PROHFXOHV DQG DUH URXWLQHO\ FDSDEOH RI GHWHFWLQJ LQGLYLGXDO PROHFXODU LRQV VWULNLQJ WKH HOHFWURQ PXOWLSOLHU ,Q WKLV UHVSHFW PDVV VSHFWURPHWU\ LV FDSDEOH RI VLQJOH PROHFXOH GHWHFWLRQ EXW DV PHQWLRQHG HDUOLHU SUDFWLFDO 60' PXVW RFFXU ZLWK QHDU b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f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f 7KLV SURFHVV LV DV PXFK DUW DV VFLHQFH DQG FDQ EH YHU\ WLPH FRQVXPLQJ DQG H[SHQVLYH

PAGE 16

2QFH WKH VDPSOH LV FRDWHG WKH DQDO\VLV FDQ DOVR EH WLPH FRQVXPLQJ LQ WU\LQJ WR YLHZ VLQJOH PROHFXOHV RQ D ODUJH VXUIDFH )XUWKHUPRUH GDWD LQWHUSUHWDWLRQ LV RIWHQ D VXEMHFWLYH SURFHVV RIWHQ FULWLFL]HG IRU WKH LPSRVLWLRQ RI LPDJLQDWLYH YLHZLQJ E\ WKH UHVHDUFKHUV 7KHVH SUREOHPV DUH EHLQJ DGGUHVVHG E\ FXUUHQW UHVHDUFK DQG LI WKH\ FDQ EH UHVROYHG WKH SRWHQWLDO IRU 670 LV HQRUPRXV (YHQ VR 670 KDV EHFRPH D YDOXDEOH WRRO IRU WKH VXUIDFH DQDO\VLV RI PHWDOV DQG PDQ\ RWKHU VROLG PDWHULDOV EXW LW LV QRW FXUUHQWO\ SUDFWLFDO DV D JHQHUDO WHFKQLTXH IRU 60' /DVHULQGXFHG IOXRUHVFHQFH $V LQ PDVV VSHFWURPHWU\ ODFN RI b VDPSOLQJ HIILFLHQF\ SODJXHV PDQ\ DWRPLF VSHFWURFKHPLFDO PHWKRGV LQ IODPHV DQG IXUQDFHV VXFK DV DWRPLF IOXRUHVFHQFH ODVHUHQKDQFHG LRQL]DWLRQ DQG UHVRQDQFH LRQL]DWLRQ +RZHYHU WKH GLUHFW PROHFXODU IOXRUHVFHQFH DQDO\VLV LQ D VROXWLRQ LV WKHRUHWLFDOO\ DQG SUDFWLFDOO\ FDSDEOH RI GHWHFWLQJ DQ LQGLYLGXDO KLJKO\ IOXRUHVFHQW PROHFXOH /DVHU LQGXFHG IOXRUHVFHQFH /,)f LV FRPPRQO\ WKH PRVW VHQVLWLYH IOXRUHVFHQFH WHFKQLTXH DQG IRU WKLV DQG RWKHU UHDVRQV WR EH GLVFXVVHG LQ D ODWHU VHFWLRQ LQ WKLV FKDSWHU /,) LV WKH FKRVHQ PHWKRG LQ WKH DWWHPSW WR DWWDLQ 60' 7KHRU\ RI 6LQJOH 0ROHFXOH 'HWHFWLRQ ,Q UHFHQW \HDUV VLQFH WHFKQRORJ\ KDV EHHQ GHYHORSHG WKDW LV FDSDEOH RI H[WUHPHO\ ORZ OLPLWV RI GHWHFWLRQ PXFK GLVFXVVLRQ KDV DSSHDUHG LQ WKH OLWHUDWXUH FRQFHUQLQJ WKH UHTXLUHPHQWV DQG VWDWLVWLFV FRQFHUQLQJ GHWHFWLRQ RI VLQJOH VSHFLHVn $ XVHIXO VXPPDU\ RI PRVW RI WKH FRQFHSWV RI 60' DV ZHOO DV DSSOLFDWLRQ RI WKH

PAGE 17

WKHRU\ IRU WKH FRXQWLQJ RI DWRPV DQG PROHFXOHV KDV UHFHQWO\ EHHQ JLYHQn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f EXW DSSO\ DOVR WR 60' +H PHQWLRQHG WKDW 6$' LQYROYHV WZR EDVLF UHTXLUHPHQWV f DQ HIILFLHQF\ RI GHWHFWLRQ RI XQLW\ DQG f DWWDLQPHQW RI WKH LQWULQVLF QRLVH OLPLW 7KHVH WZR IDFWRUV DUH GHILQHG EHORZ

PAGE 18

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f PXVW EH QHDUO\ b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fV VHFRQG FRQGLWLRQ IRU 6$' H[WULQVLF QRLVH PXVW EH HOLPLQDWHG DW ZKLFK WLPH LQWULQVLF QRLVH EHFRPHV SUHYDOHQW ([WULQVLF QRLVH LV WKH EDFNJURXQG SURGXFHG E\ H[WHUQDO IDFWRUV VXFK DV VWUD\ OLJKW WKHUPDO IOXFWXDWLRQV DQG HOHFWULF DQG PDJQHWLF ILHOGV 7KHVH IDFWRUV FDQ

PAGE 19

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f /2' LV GHILQHG DV WKH FRQFHQWUDWLRQ DW ZKLFK WKH VLJQDO LV WLPHV ODUJHU WKDQ WKH VWDQGDUG GHYLDWLRQ RI WKH EODQN DEOf RU /2' UEOVHQVLWLYLW\ f ZKHUH VHQVLWLYLW\ LV WKH OLQHDU VORSH RI WKH DQDO\WLFDO FDOLEUDWLRQ FXUYH RI WKH GHWHFWLRQ V\VWHP 7KLV GHILQLWLRQ ZDV GHYHORSHG DW D PHHWLQJ RI WKH ,QWHUQDWLRQDO 8QLRQ RI 3XUH DQG $SSOLHG &KHPLVWV ,83$&f LQ WR VHWWOH GLIIHUHQFHV LQ WKH VXEMHFWLYH ZD\ LQ ZKLFK GHWHFWLRQ OLPLWV ZHUH SUHYLRXVO\ GHWHUPLQHG ,Q WKH VLJQDO GRPDLQ WKH PHDVXUH RI /2' LV JLYHQ WKH V\PERO ;M ZKLFK LV GHILQHG DV ;M 0Z DEO rf ZKHUH AEO LV WKH PHDQ VLJQDO RI WKH EODQN :KHQ XVLQJ WKLV H[SUHVVLRQ D FDOLEUDWLRQ FXUYH LV QRW QHFHVVDU\ WKH DQDO\WH FRQFHQWUDWLRQ DW ZKLFK WKLV FULWHULRQ LV PHW LV WKH /2' /LPLW RI JXDUDQWHHG GHWHFWLRQ 'HVSLWH WKH ,83$& GHILQLWLRQ /2' GRHV QRW DOZD\V FRUUHVSRQG WR WKH EHVW PHDVXUH RI ORZHVW OHYHO RI DQDO\WH GHWHFWDELOLW\ IRU D

PAGE 20

V\VWHP ,Q PDQ\ FLUFXPVWDQFHV LW LV YHU\ GLIILFXOW WR DFWXDOO\ REVHUYH D GLIIHUHQFH LQ WKH VLJQDO DW WKH OLPLW RI GHWHFWLRQ .DLVHU ZDV DZDUH RI WKHVH SUREOHPV DQG LQWURGXFHG D WHUP NQRZQ DV WKH OLPLW RI JXDUDQWHHG GHWHFWLRQ /2*'f %DVHG RQ VWDWLVWLFDO FRQFHSWV /2*' LV VHW DW WZLFH WKH VWDQGDUG GHYLDWLRQ UHTXLUHPHQW FKRVHQ IRU /2' ZKLFK PHDQV /2*' DEOVHQVLWLYLW\ f RU LQ WKH VLJQDO GRPDLQ ;J 0EO -Ef f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f FRUUHVSRQGV WR WKH VLJQDO OHYHO WKDW H[FHHGV WKH EDFNJURXQG ZLWK D FRQILGHQFH RI D DQG ;J /2*' LQ WKH FRQFHQWUDWLRQ GRPDLQf LV WKH VLJQDO OHYHO WKDW JLYHV D FRQILGHQFH RI WKDW

PAGE 21

WKH DQDO\WH LV DFWXDOO\ EHLQJ GHWHFWHG %DVHG RQ WKH GHILQLWLRQV RI /2' DQG /2*' WKH PLQLPXP FRQILGHQFH OHYHO LV b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f LQ WKH 9S V\PEROL]HG E\ 1S PD\ JLYH ULVH WR D QXPEHU RI

PAGE 22

GHWHFWHG HYHQWV 1H SKRWRHOHFWURQV LQ WKH FDVH RI D SKRWRPXOWLSOLHU WXEHf 7KH PHDQ EDFNJURXQG OHYHO GXULQJ WU LV V\PEROL]HG E\ [EO 6WDWLVWLFV RI 'DWD LQ 6LQJOH 0ROHFXOH 'HWHFWLRQ 3RLVVRQ GLVWULEXWLRQ $W ORZ FRQFHQWUDWLRQV 1S IROORZV D 3RLVVRQ GLVWULEXWLRQ DQG ZKHQ SKRWRQ FRXQWLQJ LV XVHG IRU GDWD FROOHFWLRQ DW ORZ OHYHOV 1H DQG E DOVR IROORZ D 3RLVVRQ GLVWULEXWLRQ 7KH 3RLVVRQ SUREDELOLW\ GLVWULEXWLRQ LV JLYHQ E\ 3;f f ZKHUH 3;f LV WKH SUREDELOLW\ RI ; HYHQWV RFFXUULQJ ZLWK ; EHLQJ 1S 1H RU QRLVHf DQG Q LV WKH PHDQ YDOXH RI ; ,Q 3RLVVRQ GLVWULEXWLRQV WKH YDULDQFH HTXDOV WKH PHDQ D Q ZKHUH D UHIHUV WR WKH VWDQGDUG GHYLDWLRQf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

PAGE 23

&ULWHULD RI VLJQDO DQG QRLVH IRU 60' %DVHG RQ D 3RLVVRQ SUREDELOLW\ GLVWULEXWLRQ WKH YDOXHV IRU ;G DQG ;J IRU D ORZOHYHO FRXQWLQJ H[SHULPHQW DW YDULRXV EDFNJURXQG OHYHOV DUH JLYHQ LQ 7DEOH 7KH WDEOH ZDV FRQVWUXFWHG DV IROORZV f 7KH 3RLVVRQ SUREDELOLW\ GLVWULEXWLRQ ZLWK D PHDQ HTXDO WR WKH FKRVHQ EDFNJURXQG OHYHO ZDV IRXQG LQ UHIHUHQFH 7KH GHWHFWLRQ OLPLW ;r ZDV GHWHUPLQHG DV WKH QXPEHU RI FRXQWV PLQXV EDFNJURXQGf DW ZKLFK WKH VXP RI WKH UHPDLQLQJ SUREDELOLWLHV RI WKH GLVWULEXWLRQ EH\RQG ;G GLG QRW H[FHHG FRQILGHQFH D RU bf f $ GLVWULEXWLRQ ZDV WKHQ IRXQG VXFK WKDW WKH VXP RI WKH SUREDELOLWLHV JUHDWHU WKDQ ;G H[FHHGHG f 7KH PHDQ RI WKLV GLVWULEXWLRQ LV ;J 7DEOH 6LJQDO OHYHOV UHTXLUHG WR DFKLHYH VLQJOH PROHFXOH GHWHFWLRQ ZLWK b FRQILGHQFH DW JLYHQ PHDQ EODQN OHYHOV 0HDQ %ODQN /HYHO LEO LQ FRXQWVf 'HWHFWLRQ /LPLW ;G LQ FRXQWVf D *XDUDQWHHG /LPLW ;J LQ FRXQWVf $OO SDUDPHWHUV GHILQHG LQ WH[W

PAGE 24

%DVHG RQ WKHVH FRQILGHQFH OHYHOV 7DEOH JLYHV WKH VLJQDO OHYHO UHTXLUHG IRU WKH UHDOL]DWLRQ RI 60' DW D JLYHQ EDFNJURXQG OHYHO %HFDXVH RI WKH GLIIHUHQFHV LQ SUDFWLFDO DVSHFWV DQG VWDWLVWLFV RI GDWD DW QHDU60' OHYHOV &XULH KDV DGYRFDWHG WKDW WKH FRQILGHQFH OHYHO EH VHW WR b Df UDWKHU WKDQ WKH ,83$& OHYHO 7KLV ZRXOG VLJQLILFDQWO\ ORZHU WKH YDOXHV IRU ; DQG ;J UHSRUWHG LQ 7DEOH 7KH RQO\ UHPDLQLQJ WKHRUHWLFDO WRSLF LV ZKHWKHU WKH SURSRVHG /,) V\VWHP LV DEOH WR PHHW WKH VLJQDO DQG QRLVH OHYHOV SUHVHQWHG LQ 7DEOH 7KHVH FRQVLGHUDWLRQV RI /,) DQG LWV VRXUFHV RI QRLVH DUH GLVFXVVHG LQ WKH QH[W VHFWLRQ 7KHRU\ RI /DVHU,QGXFHG )OXRUHVFHQFH )OXRUHVFHQFH LV D SK\VLFDO SKHQRPHQRQ LQYROYLQJ WKH DEVRUSWLRQ RI D SKRWRQ RI OLJKW E\ D PROHFXOH FDXVLQJ DQ HOHFWURQ WR FOLPE WR D YLEURQLF OHYHO RI DQ H[FLWHG VLQJOHW VWDWH IROORZHG E\ HPLVVLRQ RI D SKRWRQ RI W\SLFDOO\ ORZHU HQHUJ\ DV WKH HOHFWURQ UHWXUQV WR WKH JURXQG YLEURQLF OHYHO 1RW DOO PROHFXOHV XQGHUJR IOXRUHVFHQFH DQG WKH RQHV WKDW GR RIWHQ DUH KLJKO\ FRQMXJDWHG DQG FRQWDLQ DURPDWLF IXQFWLRQDO JURXSV )OXRUHVFHQFH LV XVHIXO DV DQ DQDO\WLFDO SURFHGXUH EHFDXVH WKH LQWHQVLW\ RI WKH HPLVVLRQ LV GHSHQGHQW XSRQ FRQFHQWUDWLRQ DQG WKH H[FLWDWLRQ DQG HPLVVLRQ ZDYHOHQJWKV RI WKH OLJKW JLYH VRPH LQIRUPDWLRQ DV WR WKH LGHQWLW\ RI WKH IOXRUHVFHQW VSHFLHV 'XH WR IDFWRUV WR EH GLVFXVVHG EHORZ IOXRUHVFHQFH LV RIWHQ D YHU\ VHQVLWLYH W\SH RI DQDO\VLV $V WKH QDPH LPSOLHV ODVHULQGXFHG IOXRUHVFHQFH /,)f VLPSO\ XVHV D ODVHU DV WKH H[FLWDWLRQ VRXUFH IRU IOXRURPHWULF DQDO\VLV /DVHUV DUH DEOH WR SURGXFH KLJKHU

PAGE 25

VSHFWUDO LUUDGLDQFH :FPf WKDQ FRPPRQ EURDGEDQG VRXUFHV VXFK DV WKH [HQRQ DUF ODPS 7KXV /,) W\SLFDOO\ JLYHV ORZHU /2'V IRU WKH DQDO\VLV RI IOXRUHVFHQW FRPSRXQGV $OVR WKH QDUURZ HPLVVLRQ EDQGZLGWK RI WKH ODVHU LV XVHIXO LQ PDQ\ VLWXDWLRQV WKDW UHTXLUH VHOHFWLYH H[FLWDWLRQ DQG GHWHFWLRQ RI D IOXRUHVFHQW SUREH VSHFLHV DGGHG WR D V\VWHP 7KH SUDFWLFDO IRUPXOD WKDW JLYHV WKH DYHUDJH QXPEHU RI GHWHFWHG HYHQWV 1H LQ WKH W\SH RI /,) V\VWHP XVHG IRU 60' LV 1 A.<3ƒ}7WU! f Kr96/ LU ZKHUH L!/ LV WKH ODVHU SRZHU :f KL!/ JLYHV WKH HQHUJ\ SHU ODVHU SKRWRQ -f ZLWK K EHLQJ 3ODQFNfV FRQVWDQW [ n -Vf DQG Y/ EHLQJ IUHTXHQF\ RI WKH OLJKW +]f 6/ LV WKH FURVVVHFWLRQDO DUHD RI WKH IRFXVHG ODVHU EHDP FPf D$ LV WKH FURVV VHFWLRQ RI DEVRUSWLRQ IRU WKH PROHFXOH FPf <) LV WKH IOXRUHVFHQFH TXDQWXP HIILFLHQF\ ) LV WKH VROLG DQJOH RI FROOHFWLRQ RI WKH IOXRUHVFHQFH VUf W" LV WKH FDWKRGLF HIILFLHQF\ RI WKH GHWHFWRU GLPHQVLRQOHVVf 7 LV WKH WUDQVPLWWDQFH RI WKH RSWLFDO FRPSRQHQWV GLPHQVLRQOHVVf DQG WU LV WKH UHVLGHQFH WLPH RI HDFK PROHFXOH LQ WKH SUREH YROXPH Vf 7KHVH SDUDPHWHUV DUH JLYHQ E\ WKH PDQXIDFWXUHU DV LQ WKH FDVH RI WMf HDVLO\ PHDVXUHG !/ 7 DQG 6/f UHIHUHQFHG LQ WKH OLWHUDWXUH <3f RU FDOFXODWHG IURP RWKHU NQRZQ SDUDPHWHUV D$ IO) DQG WUf 7KH IROORZLQJ SDUDJUDSKV GLVFXVV WKH GHWHUPLQDWLRQ RI WKHVH SDUDPHWHUV IRU WKH V\VWHP WR EH GHVFULEHG LQ WKLV WKHVLV 7KH FURVV VHFWLRQ RI DEVRUSWLRQ 7KH DEVRUSWLRQ FURVV VHFWLRQ D. LV QRW DV PXFK RI DQ DFWXDO VL]H RI D PROHFXOH DV LW LV D VWDWLVWLFDO TXDQWLW\ 7KH XQLWV RI FP

PAGE 26

DULVH IURP WKH IDFW WKDW LUUDGLDQFH :FPf LV XVHG LQ WKH H[SUHVVLRQ WR GHWHUPLQH WKH DYDLODELOLW\ RI OLJKW 7KH YDOXH IRU D$ LV WKH SUREDELOLW\ WKDW WKH OLJKW DYDLODEOH LQ D FHUWDLQ DUHD LV DEVRUEHG 7KH VLPSOHVW ZD\ WR GHWHUPLQH U$ IRU D PROHFXOH LV WR PHDVXUH WKH DEVRUEDQFH RI D NQRZQ FRQFHQWUDWLRQ LQ VROXWLRQ $VVXPLQJ WKH FRQFHQWUDWLRQ IDOOV ZLWKLQ WKH UHJLRQ RI OLQHDU UHVSRQVH WKH PRODU DEVRUSWLYLW\ H$ 0 nFPf DW WKH FKRVHQ ZDYHOHQJWK FDQ EH IRXQG IURP %HHUfV ODZ $ HM& f ZKHUH $ LV DEVRUEDQFH GLPHQVLRQOHVVf LV SDWK OHQJWK FPf DQG & LV FRQFHQWUDWLRQ 0f %\ NQRZLQJ H$ D$ FDQ EH IRXQG IURP D$ H$1 f ZKHUH 1 LV $YRJUDGRfV QXPEHU [ f 7KH DEVRUSWLRQ FURVVVHFWLRQ FDQ DOVR EH GHWHUPLQHG IURP IXQGDPHQWDO SDUDPHWHUV EXW WKLV PHWKRG ZDV XVHG LQ WKLV SURMHFW EDVHG RQ LWV VLPSOLFLW\ DQG LWV XVH RI DQ H[SHULPHQWDO PHDVXUHPHQW 7KH DEVRUSWLRQ FRHIILFLHQWV IRU WKH G\HV WR EH WHVWHG LQ WKLV SURMHFW JLYHQ LQ &KDSWHU f DUH rr 0nFPn WKHUHIRUH WKH YDOXHV IRU D. DUH DSSUR[LPDWHO\ [ FP )OXRUHVFHQFH TXDQWXP \LHOG 7KH TXDQWXP HIILFLHQF\ RI IOXRUHVFHQFH RU TXDQWXP \LHOG <) LV WKH SUREDELOLW\ RI HPLVVLRQ RI D IOXRUHVFHQFH SKRWRQ RQFH D SKRWRQ KDV EHHQ DEVRUEHG ,Q PDWKHPDWLFDO IRUP <) LV JLYHQ E\

PAGE 27

< ) N) N)NQU f ZKHUH N) LV WKH UDWH RI IOXRUHVFHQFH Vnf DQG LV WKH UDWH RI QRQUDGLDWLYH GHDFWLYDWLRQ Vf RI WKH H[FLWHG VLQJOHW VWDWH 6f 1RQUDGLDWLYH GHFD\ RI 6L RFFXUV WKURXJK WKH SURFHVVHV RI H[WHUQDO FRQYHUVLRQ FROOLVLRQDO GHDFWLYDWLRQf LQWHUQDO FRQYHUVLRQ QRQIOXRUHVFHQW GHH[FLWDWLRQf DQG LQWHUV\VWHP FURVVLQJ 6W WR WULSOHW 7Mf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f H$X9X f ZKHUH WKH GLIIHUHQFHV LQ WKH DEVRUSWLRQ FRHIILFLHQWV DQG GHWHFWLRQ HIILFLHQFLHV D IXQFWLRQ RI ZDYHOHQJWKf IRU WKH NQRZQ DQG XQNQRZQ IOXRURSKRUHV PXVW EH WDNHQ LQWR DFFRXQW 7KH PRVW FRPPRQO\ XVHG UHIHUHQFH IOXRURSKRUH LV TXLQLQH ZKLFK KDV D TXDQWXP \LHOG RI LQ DQ DFLGLF DTXHRXV VROXWLRQ

PAGE 28

7KH 3DUNHU5HHV PHWKRG LV QRW DOZD\V VLPSOH WR XVH DQG GXH WR RWKHU FRUUHFWLRQ IDFWRUV QRW LQFOXGHG DERYH LV QRW DOZD\V DFFXUDWH $Q HDVLHU ZD\ WR GHWHUPLQH <3 IRU D SDUWLFXODU V\VWHP LV WR VHDUFK WKH SK\VLFDO FKHPLVWU\ OLWHUDWXUH IRU DQ DFFXUDWH GHWHUPLQDWLRQ RI <) LQ WKH VROYHQW WR EH XVHG 7+URXJK DQ LQWHQVLYH OLWHUDWXUH VHDUFK LW ZDV IRXQG WKDW WKDW WKH G\H KDV <3 r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f ZKHUH LV WKH DQJOH GHILQHG E\ D OLQH H[WHQGHG IURP WKH SRLQW VRXUFH WR WKH FHQWHU RI WKH OHQV DQG D VHFRQG OLQH IURP WKH SRLQW VRXUFH WR WKH HGJH RI WKH OHQV 7KH SUREOHP ZLWK WKLV HTXDWLRQ LV WKDW LW EUHDNV GRZQ DW ODUJH VROLG DQJOHV HQFRXQWHUHG ZLWK PLFURVFRSH REMHFWLYHV 7KH FDOFXODWHG YDOXH RI IO3 FDQ H[FHHG WKH WUXH YDOXH E\ D IDFWRU RI b IRU r DQG E\ D IDFWRU RI b IRU D r 7KH FRUUHFW H[SUHVVLRQ IRU IL3 LV JLYHQ E\ LO) WFRVf LUVLQAf f :LWK WKLV HTXDWLRQ IL) LV OHVV VXVFHSWLEOH WR HUURU WKDQ ZLWK WKH SUHYLRXV HTXDWLRQ

PAGE 29

:KHQ XVLQJ D PLFURVFRSH REMHFWLYH LV IRXQG IURP WKH VWDWHG QXPHULFDO DSHUDWXUH 1$ RI WKH REMHFWLYH LQ WKDW 1$ Q VLQ f§ f I ZKHUH Q LV WKH UHIUDFWLYH LQGH[ GLPHQVLRQOHVVf RI WKH PHGLXP EHWZHHQ WKH REMHFW DQG REMHFWLYH e LV WKH DSHUWXUH FPf DQG I LV WKH IRFDO OHQJWK FPf RI WKH REMHFWLYH 7KH FDOFXODWHG Lf) IRU WKH PLFURVFRSH REMHFWLYH WR EH XVHG LQ WKLV H[SHULPHQW LV VU ZKLFK ZKHQ FRUUHFWHG IRU WKH ?U VU RI D VSKHUH FRUUHVSRQGV WR D FROOHFWLRQ HIILFLHQF\ RI b RI D SRLQW VRXUFH 7KH PDJQLILFDWLRQ ;f DQG 1$ f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f WKH SRWHQWLDO IRU SKRWRGHFRPSRVLWLRQ RI WKH G\H EHIRUH WQ DQG f RSWLFDO VDWXUDWLRQ RI IOXRUHVFHQFH $FFRUGLQJ WR (TXDWLRQ VLPSO\ LQFUHDVLQJ ODVHU SRZHU GHFUHDVLQJ IRFXV VL]H RU LQFUHDVLQJ WU DOORZV RQH WR

PAGE 30

REWDLQ DV ODUJH D VLJQDO DV GHVLUHG +RZHYHU SK\VLFDO OLPLWDWLRQV WR QHJDWH WKLV SRVVLELOLW\ DUH GHVFULEHG EHORZ 7DEOH 3DUDPHWHUV RI WKH V\VWHP WR EH XVHG LQ WKH DWWHPSW RI ODVHULQGXFHG IOXRUHVFHQFH GHWHFWLRQ RI VLQJOH PROHFXOHV DQG WKH H[SHFWHG VLJQDO OHYHO FDOFXODWHG IURP (TXDWLRQ 3DUDPHWHU 9DOXH /DVHU 3RZHU / P: /DVHU )UHTXHQF\ Y/ [ +] /DVHU )RFXV $UHD 6/ [ FP $EVRUSWLRQ &URVV6HFWLRQ D$ [ FP '\H 4XDQWXP
PAGE 31

UHVSHFWLYHO\ DQG $ LV WKH (LQVWHLQ FRHIILFLHQW RI VSRQWDQHRXV HPLVVLRQ RI WKH IOXRUHVFHQFH Vnf 7KLV UDWH RI VSRQWDQHRXV HPLVVLRQ LV VLPSO\ WKH LQYHUVH RI WKH IOXRUHVFHQFH OLIHWLPH WI Vf 7R GHWHUPLQH (/ ZKLFK LV WKH ODVHU LUUDGLDQFH ZKHQ WKH VORSH RI D ORJORJ SORW RI VLJQDO YHUVXV LUUDGLDQFH EHFRPHV WKH UHOHYDQW HTXDWLRQ LV JLJ
PAGE 32

1 /f$ $}7W f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m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f LW ZRXOG RQO\ WDNH PV WR JR WKURXJK PLOOLRQ F\FOHV 1RLVH OHYHO UHTXLUHG IRU 60' %DVHG RQ WKH WKHRU\ RI /,) SUHVHQWHG KHUH WKH DYHUDJH VLJQDO OHYHO VKRXOG FRQVLVW RI FRXQWV DERYH EDFNJURXQG SHU FRXQWLQJ

PAGE 33

LQWHUYDO 8VLQJ WKH VWDWLVWLFDO WKHRU\ DSSOLHG WR 60' SUHVHQWHG HDUOLHU WKH PD[LPXP PHDQ QRLVH OHYHO [EO SHUPLWWHG GXULQJ WKH PV LQWHUYDO WR DWWDLQ DQ /2' RI PROHFXOH ZLWK b FRQILGHQFH LV FRXQWV )RU /2* ZLWK b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

PAGE 34

H[DPSOH WKH QXPEHU RI SKRWRQV IRFXVHG RQ WKH VDPSOH E\ WKH P: ODVHU DW QP WR EH XVHG LQ WKLV SURMHFW LV [ Vn RU [ SKRWRQV GXULQJ WKH PV FRXQWLQJ SHULRG 2I WKHVH RQO\ SKRWRQ LQ ELOOLRQ U$6/f LV DEVRUEHG E\ D PROHFXOH LQ WKLV WLPH ZKLFK HVVHQWLDOO\ VWLOO OHDYHV VRPH [ QRQDEVRUEHG SKRWRQV 5HDOL]LQJ WKDW IL)LU LV QHDUO\ b LV b DQG 7 LV b IRU WKH V\VWHP WKH GHWHFWRU ZRXOG SURGXFH r [ SKRWRHOHFWURQV SHU PV DVVXPLQJ WKDW WKH ODVHU OLJKW LV VFDWWHUHG LVRWURSLFDOO\ IURP D SRLQW VRXUFH +RZHYHU WKLV DVVXPSWLRQ LV QRW WUXH WKH YDVW PDMRULW\ RI WKH ODVHU OLJKW FRQWLQXHV XQKLQGHUHG WKURXJK WKH VDPSOH FRQWDLQHU DQG LV QRW VFDWWHUHG ZLWKLQ WKH VU UHJLRQ FROOHFWHG E\ WKH PLFURVFRSH REMHFWLYH ZKLFK LV ZK\ IOXRUHVFHQFH LV FROOHFWHG r IURP WKH DQJOH RI H[FLWDWLRQf )RU WKH VDNH RI DUJXPHQW DVVXPH RQO\ b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b LVRWURSLF SRLQW VRXUFH RI ODVHU VFDWWHU DW WKH IRFXV RI WKH FROOHFWLRQ RSWLFV LV QRW YDOLG H[FHSW LQ WKH FDVH RI FDSLOODULHV DV ZLOO EH GLVFXVVHG ODWHUf )XUWKHUPRUH ZLWK PRQRFKURPDWRUV WKH VSHFWUDO EDQGSDVV RI WKH HPLVVLRQ SURFHVV LV DOVR JUHDWO\

PAGE 35

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

PAGE 36

6SDWLDO ILOWHULQJ $V PHQWLRQHG DERYH WKH IRFXVHG ODVHU OLJKW GRHV QRW VFDWWHU HTXDOO\ LQ DOO GLUHFWLRQV 7KH SDWWHUQ RI ODVHU VFDWWHU SURGXFHG DW WKH VDPSOH JUHDWO\ GHSHQGV RQ IRFXVLQJ DQG WKH VKDSH RI WKH VDPSOH FRQWDLQHU WKLV DVSHFW EHFRPHV YHU\ LPSRUWDQW ODWHU LQ WKLV GLVVHUWDWLRQ ZKHQ FRQVLGHULQJ WKH FDSLOODU\ FRQWDLQHU XVHG LQ WKLV SURMHFWf 6SDWLDO ILOWHULQJ H[SORLWV WKH QRQLVRWURSLF IHDWXUH RI WKH VFDWWHU WKURXJK WKH SODFHPHQW RI D VPDOO VOLW RU SLQKROH EHWZHHQ WKH FROOHFWLRQ RSWLFV DQG WKH GHWHFWRU :LWK FDUHIXO SRVLWLRQLQJ DQG IRFXVLQJ WKH DSHUWXUH FDQ FROOHFW D ODUJH SHUFHQWDJH RI WKH IOXRUHVFHQFH ZKLOH EORFNLQJ PXFK RI WKH ODVHU VFDWWHU DULVLQJ IURP WKH HGJHV RI WKH FRQWDLQHU 7KH LPSOHPHQWDWLRQ RI WKLV VLPSOH FRQFHSW E\ 'RYLFKL HW D/6 DORQJ ZLWK WKH XVH RI VSHFWUDO ILOWHUV DQG VKHDWK IORZ FXYHWWH VDPSOH FRQWDLQHUf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f LV YHU\ KLJK FRPSDUHG WR 5D\OHLJK VFDWWHU ZKLFK W\SLFDOO\ KDV D FURVV

PAGE 37

VHFWLRQ RQ WKH RUGHU RI FP 7KLV FRUUHVSRQGV WR WKH SURGXFWLRQ RI SKRWRHOHFWURQ DW DQ DYHUDJH RI HYHU\ PV IRU PHWKDQRO LQ WKH Q/ 9S RI WKH FRQGLWLRQV SUHVHQWHG LQ 7DEOH $OWKRXJK WKLV LV YHU\ VPDOO DW WKH OLJKW OHYHOV LQYROYHG LQ WKLV SURMHFW HYHU\ DGGLWLRQDO QRLVH SKRWRHOHFWURQ FRXOG EHFRPH VLJQLILFDQW 7KH PHWDO YDSRU ILOWHU $ SRVVLEOH ZD\ WR HOLPLQDWH WKH VHULRXV SUREOHP RI ODVHU VFDWWHU LV WKURXJK WKH XVH RI D PHWDO YDSRU ILOWHU 09)f 7KLV GHYLFH LV FHQWUDO WR WKH VXFFHVV RI WKLV SURMHFW DQG LWV WKHRU\ LV SUHVHQWHG LQ &KDSWHU DQG H[SHULPHQWDO UHVXOWV LQ &KDSWHU 7KH WKHRU\ LV WRR H[WHQVLYH WR EH SUHVHQWHG QRZ EXW DFFRUGLQJ WR WKHRU\ WKH 09) LV FDSDEOH RI HVVHQWLDOO\ WRWDOO\ DEVRUELQJ WKH ODVHU VFDWWHU RU WKH HQWLUH ODVHU HPLVVLRQ IRU WKDW PDWWHUf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

PAGE 38

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n FP IRU WKH IOXRURSKRUH SUHVHQWHG LQ 7DEOH f DQG FDQ EH PHDVXUHG YHU\ VHQVLWLYHO\ ,W LV D VHOHFWLYH WHFKQLTXH DV ZHOO EXW LW LV RIWHQ XQDEOH WR VSHFLILFDOO\ GHWHFW RQH IOXRURSKRUH DW ORZ FRQFHQWUDWLRQ LQ WKH SUHVHQFH RI DQRWKHU )XUWKHUPRUH QR VROYHQW LV DEVROXWHO\ SXUH DQG HYHQ WKH SUHVHQFH RI DQ XOWUDWUDFH FRQFHQWUDWLRQ RI IOXRUHVFHQW LQWHUIHUHQWV FDQ QHJDWH WKH SRVVLELOLW\ RI 60'

PAGE 39

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f 7KLV LV D VXEVWDQWLDO UHGXFWLRQ LQ QRLVH ZKHQ SXUVXLQJ 60'

PAGE 40

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r f WDJJHG ZLWK IOXRUHVFHLQ PROHFXOHV RQ D VROLG VXEVWUDWH .LUVFK HW DO LQ D VLPLODU W\SH RI SURFHGXUH ZHUH DEOH WR GHWHFW UKRGDPLQH PROHFXOHV 0RUH UHFHQWO\ 0RHUQHUn KDV GHWHFWHG VLQJOH SHQWDFHQH PROHFXOHV LQ D VROLG PDWUL[ DW ORZ WHPSHUDWXUH XVLQJ ODVHUH[FLWHG IOXRUHVFHQFH .HOOHUfV DSSURDFK WR 60' ,Q WKH DQDO\VLV RI IORZLQJ VROXWLRQV .HOOHUfV JURXS SHUIRUPHG VHYHUDO H[SHULPHQWV OHDGLQJ WR WKH FODLP RI VLQJOH PROHFXOH GHWHFWLRQ DOWKRXJK QRQH RI WKH UHSRUWV VDWLVILHV HYHQ $ONHPDGHfV GHILQLWLRQ RI 60'f 7KH RULJLQ RI .HOOHUfV SURMHFW DW /RV $ODPRV 1DWLRQDO /DERUDWRU\ ZKLFK LV GHVLJQHG IRU WKH DSSOLFDWLRQ RI 60' IRU '1$ VHTXHQFLQJ EHJDQ ZLWK UHVHDUFK E\ 'RYLFKL

PAGE 41

HW DOr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f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b RI WKRVH PROHFXOHV SDVVLQJ WKURXJK WKH FHQWHU RI WKH 9Sf RQ FRPSXWHU VLPXODWLRQV WKDW DSSHDU VLPLODU WR WKH DFWXDO GDWD :HLJKWHG TXDGUDWLF VXP SORWV RI WKH GDWD SUHVHQWHG VKRZHG

PAGE 42

:DYH 3ODWH > 3RODUL]LQJ 3ULVP 0LUURU $PSOLILHU 'LVFULPLQDWRU &RXQWHU 6SDWLDO )LOWHU 6SHFWUDO )LOWHU 2EMHFWLYH /HQV )LJXUH 7KH LQVWUXPHQWDO DSSURDFK WR VLQJOH PROHFXOH GHWHFWLRQ XVHG E\ .HOOHUfV JURXS DW /RV $ODPRV 1DWLRQDO /DERUDWRU\ 5HGUDZQ IURP UHIHUHQFHV DQG

PAGE 43

7DEOH &RPSDULVRQ RI WKH UHSRUWHG SDUDPHWHUV DQG UHVXOWV RI WKH ODVHU LQGXFHG IOXRUHVFHQFH H[SHULPHQWV RI .HOOHUfV JURXS DW /RV $ODPRV 1DWLRQDO /DERUDWRU\ 3DUDPHWHU 5HI f 5HI f 5HI f 5HI f 5HI f $QDO\WH 5* 5* 5* 5* 5* 6ROYHQW KR KR ,,M24,,M2+ KR &+2+ &HOO 6)& 6)& IORZFHOO 6)& IORZFHOO /DVHU ;/ QPf $U f $U 1G<$* E $U $U 6/ FPf [O OO[O2 [On [O [Of (O N:FPf 6WUHDP 6L]H ILPf Q)LU f§ f§ f§ H [O 9S 3/f )ORZ 5DWH [/PLQf )ORZ 9HORFLW\ FPVf WU PVf 7F Vf /2' 0f [On [O f§ [O f§ /2' UFf aOF }OG 5* 5KRGDPLQH 6)& 6KHDWK )ORZ &XYHWWH H 6DPSOLQJ (IILFLHQF\ & 7LPH &RQVWDQW RI 0HDVXUHPHQW /2' WFf QXPEHU RI PROHFXOHV SDVVLQJ WKURXJK 9S GXULQJ WF DW WKH /2' 0f f3XOVHG DW N+] b GXW\ F\FOH E3XOVHG DW 0+] b GXW\ F\FOH F7LPH GLVFULPLQDWLRQ PHWKRG DEOH WR GHWHFW SDVVDJH RI VLQJOH PROHFXOHV ZLWK UHSRUWHG b GHWHFWLRQ HIILFLHQF\ HGf G5HSRUWHG GHWHFWLRQ OLPLW EDVHG RQ DXWRFRUUHODWLRQ DQDO\VLV HG QRW JLYHQ

PAGE 44

SKRWRHOHFWURQ EXUVWV DULVLQJ IURP SDVVDJH RI VLQJOH PROHFXOHV ZKLFK ZHUH QRW SUHVHQWHG LQ DQ\ RI WKH RWKHU UHIHUHQFHV 7KH VDPH VRUW RI EDVLV ZDV XVHG LQ WKH FODLP WR VLQJOH PROHFXOH GHWHFWLRQ PDGH LQ 5HI EXW LQ WKLV FDVH SDVVDJH RI LQGLYLGXDO PROHFXOHV LQ WKH 9S ZDV QRW QRWLFHDEOH 7KH FODLP WR VLQJOH PROHFXOH GHWHFWLRQ ZDV EDVHG RQ WKH XVH RI D SRLQW VOLGLQJ VXP GLVWULEXWLRQ ZKLFK DJUHHG ZLWK WKHRUHWLFDO UHVXOWV 1R HIILFLHQF\ RI GHWHFWLRQ FDQ EH FDOFXODWHG IURP VXFK D GHWHUPLQDWLRQ EHFDXVH WKH VLQJOH HYHQWV FRXOG QRW EH FRXQWHG ,Q UHIHUHQFHV WKURXJK WKH VSDWLDO ILOWHU YLHZHG RQO\ D VPDOO SRUWLRQ APf RI WKH ODVHU IRFXV YROXPH WR UHGXFH WKH EDFNJURXQG VRXUFHV RI QRLVH GLVFXVVHG HDUOLHU 7KLV WHFKQLTXH OLPLWHG WKH VROLG DQJOH RI FROOHFWLRQ RI WKH PLFURVFRSH REMHFWLYH ZKLFK LV ZK\ IO)LU LV QRW UHSRUWHG LQ WKRVH FDVHV H[FHSW 5HI f ,Q 5HI ZKLFK UHSRUWHG DQ /2' RI [ 0 EDVHG RQ FRQYHQWLRQDO PHWKRGV WR GHWHUPLQH WKH /2' WKH EO RI WKH 307 ZDV FRXQWVV 7KLV KLJK QRLVH OHYHO LV D UHVXOW RI WKH ODUJH EDFNJURXQG VRXUFHV RI QRLVH GLVFXVVHG HDUOLHU $W WKLV QRLVH OHYHO 60' LV QRW SRVVLEOH ZLWK WKH DSSURDFK SUHVHQWHG LQ WKLV WKHVLV D VHSDUDWH VWXG\ XWLOL]LQJ %SK\FRHU\WKULQ DV WKH DQDO\WH 1JX\HQ HW DO LQ .HOOHUfV JURXS FODLPHG WKH ILUVW LQVWDQFH RI VLQJOH PROHFXOH GHWHFWLRQ LQ VROXWLRQ ,W LV QRWHZRUWK\ WKDW %SK\FRHU\WKULQ LV D YHU\ ODUJH 0: r JPROf SURWHLQ SRVVHVVLQJ WKH HTXLYDOHQW IOXRUHVFHQFH RI UKRGDPLQH PROHFXOHV 0DWKLHVf UHVHDUFK $W WKH 8QLYHUVLW\ RI &DOLIRUQLD DW %HUNHOH\ WKH UHVHDUFK JURXS RI 0DWKLHV FRQWHVWHG WKLV LQLWLDO FODLP WR 60' PDGH E\ .HOOHUfV JURXS WKH\ UHSHDWHG WKH /,) VWXG\ ZLWK %SK\FRHU\WKULQ 7KURXJK D PRUH ULJRUXV VWDWLVWLFDO

PAGE 45

DSSURDFK WKHLU UHVHDUFK HIIRUW VKRZHG WKDW WKH SUHYLRXV ZRUN GLG QRW REWDLQ 60' DQG WKHLU UHVXOWV GHPRQVWUDWHG WKH GHWHFWLRQ RI b RI WKH SDVVLQJ VLQJOH PROHFXOHV %DVHG RQ WKH GHILQLWLRQ RI 60' SUHVHQWHG LQ WKLV WKHVLV %SK\FRHU\WKULQ ZLWK D 0: m JPRO GRHV QRW TXDOLI\ DV D VLQJOH PROHFXOH DQG WKH VDPSOLQJ HIILFLHQF\ RI WKH H[SHULPHQW LV PXFK OHVV WKDQ XQLW\ .HOOHUfV JURXS DQG RWKHU UHVHDUFKHUV DUH DZDUH RI WKHVH VKRUWFRPLQJV DQG KDYH ZRUNHG WR ORZHU WKH /2' IRU VPDOOHU IOXRURSKRUHV ,Q WKHLU SURSRVDO .HOOHUf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

PAGE 46

DSSHDUV PRUH SURPLVLQJ IRU WUXH 60' WKDQ WKH DSSURDFK H[KLELWHG LQ )LJXUH ,QGHHG WZR RI WKH UHFHQW SDSHUV E\ .HOOHU GR QRW XVH WKH VKHDWK IORZ FXYHWWH :LQHIRUGQHUfV DSSURDFK 7R DYRLG WKH SUREOHP RI WKLV WUDGHRII ZLWK D VKHDWK IORZ FXYHWWH DQG LWV KLJK FRVW :LQHIRUGQHUfV JURXS KDV GHFLGHG WR SUREH WKH HQWLUH VDPSOH VWUHDP LQ WKHLU DWWHPSWV DW 60' )XUWKHUPRUH WKH GLRGH ODVHU ZDV FKRVHQ DV WKH H[FLWDWLRQ VRXUFH IRU SXUSRVHV RI JUHDWHU DQDO\WH VHOHFWLYLW\ ORZHU QRLVH IURP 5DPDQ VFDWWHU ORZHU FRVW VLPSOLFLW\ DQG WKH PDQ\ RWKHU DGYDQWDJHV RI GLRGH ODVHUV WR EH GLVFXVVHG LQ &KDSWHU ,Q WZR VHSDUDWH VWXGLHV ZLWK GLIIHUHQW ODVHUVf /2'V RI DQG PROHFXOHV IORZLQJ LQ WKH SUREH YROXPH RI WKH QHDULQIUDUHG G\H ,5 KDYH EHHQ PHDVXUHG LQ D OLTXLG MHW IORZ VWUHDP HPDQDWLQJ IURP D FDSLOODU\f 'HVSLWH WKH PDQ\ DGYDQWDJHV RI ZRUNLQJ ZLWK GLRGH ODVHUV DQG WKH ORZ GHWHFWLRQ OLPLWV DWWDLQHG ZLWK WKHLU XVH /,) ZLWK GLRGH ODVHUV ZDV IRXQG WR ODFN WKH VHQVLWLYLW\ WR DFKLHYH 60' LQ D OLTXLG MHW )XUWKHUPRUH WKH OLTXLG MHW RQO\ RSHUDWHV DW KLJK IORZ UDWHV ZKLFK JLYHV XQVXLWDEOH UHVLGHQFH WLPHV IRU 60' 7KLV IORZ FRQGLWLRQ LV RQH RI WKH UHDVRQV ZK\ WKH PHWKRG GHVFULEHG LQ WKLV WKHVLV HPSOR\V D FDSLOODU\ IRU VDPSOH FRQWDLQPHQW 5DPVH\fV DSSURDFK 5DPVH\fV JURXS DW 2DN 5LGJH 1DWLRQDO /DERUDWRU\ KDV GHVLJQHG WKH ILUVW LQVWUXPHQWDO VHWXS FDSDEOH RI WUXO\ UHDOL]LQJ 60' LQ VROXWLRQ DV GHILQHG LQ WKLV GLVVHUWDWLRQ 7KH GHVLJQ RI WKLV DSSURDFK DSSHDUV LQ )LJXUH 7KH NH\ WR WKH H[SHULPHQW LV WKH XVH RI WKH HOHFWURG\QDPLF WUDS WR FRQWDLQ GURSOHWV RI WKH VDPSOH LQVWHDG RI D IORZLQJ VWUHDP ,Q WKLV ZD\ WKH HQWLUH VDPSOH LV SUREHG RQH

PAGE 47

)LJXUH 7KH LQVWUXPHQWDO DSSURDFK WR VLQJOH PROHFXOH GHWHFWLRQ XVHG E\ 5DPVH\fV JURXS DW 2DN 5LGJH 1DWLRQDO /DERUDWRU\ 5HGUDZQ IURP UHIHUHQFH X! 92

PAGE 48

GURS DW D WLPHf DQG WKH LQWHUDFWLRQ WLPH FDQ EH RQ WKH RUGHU RI GD\V LI QHFHVVDU\ 7KH QRLVH UHMHFWLRQ LV QRW DV KLJK DV LQ RWKHU H[SHULPHQWV DQG WKH G\H WHQGV WR SKRWRGHFRPSRVH EHIRUH 60' FDQ EH TXDQWLWDWLYHO\ DWWDLQHG 7KH /2' FXUUHQWO\ VWDQGV DW DQ DYHUDJH RI PROHFXOHV RI UKRGDPLQH SHU GURSOHW 7KH DSSURDFK WR 60' RI WKLV GLVVHUWDWLRQ ,W LV KRSHG WKDW WKH 60' DSSURDFK GHVFULEHG LQ &KDSWHU ZLOO EH WKH ILUVW PHWKRG WR WUXO\ DFKLHYH 60' LQ D VROXWLRQ /LNH 5DPVH\fV DSSURDFK LW LV GHVLJQHG WR VDPSOH WKH HQWLUH VROXWLRQ DQG KDV EHHQ VKRZQ WR WKHRUHWLFDOO\ SHUIRUP WKH GHVLUHG DVVLJQPHQW 0RUH RI WKH WKHRU\ GHVLJQ DQG GHYHORSPHQW FRQVLGHUDWLRQV DQG VRPH UHVXOWV ZLOO EH SUHVHQWHG LQ WKH IROORZLQJ FKDSWHU

PAGE 49

&+$37(5 &+2,&( 2) ,167580(17$7,21 $1' 237,0,=$7,21 2) 3$5$0(7(56 7KH 0HWDO 9DSRU )LOWHU %HFDXVH RI LWV UROH WR UHPRYH WKH ODVHU VSHFXODU DQG 5D\OHLJK VFDWWHU IURP WKH FROOHFWHG OLJKW WKH PHWDO YDSRU ILOWHU 09)f LV WKH PRVW LPSRUWDQW FRPSRQHQW RI WKH H[SHULPHQWDO VHWXS WR DWWDLQ VLQJOH PROHFXOH GHWHFWLRQ 60'f :LWK IHZ H[FHSWLRQV ODVHU VFDWWHU LV WKH OLPLWLQJ VRXUFH RI QRLVH LQ XOWUDWUDFH DQDO\VHV XVLQJ ODVHULQGXFHG IOXRUHVFHQFH /,)f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

PAGE 50

3\UH[ N &\OLQGHU 1LWURJHQ 7 0HWDO (OHPHQW )LJXUH 7KH PHWDO YDSRU ILOWHU $EVRUSWLRQ ,QFUHDVHV LQ WKH PHWDO YDSRU QXPEHU GHQVLW\ DQG DEVRUSWLRQ SDWK OHQJWK OHDG WR GLUHFWO\ SURSRUWLRQDO LQFUHDVHV LQ DEVRUEDQFH DVVXPLQJ WKH OLQHZLGWK RI WKH VRXUFH LV PRUH QDUURZ WKDQ WKH DEVRUSWLRQ EDQGZLGWK 7KLV UHODWLRQVKLS LV H[KLELWHG E\ %HHUfV /DZ $ D$IQ f ZKHUH $ LV WKH DEVRUEDQFH GLPHQVLRQOHVVf U$ LV WKH DEVRUSWLRQ FURVVVHFWLRQ FPf LV WKH DEVRUSWLRQ SDWK OHQJWK FPf DQG Q LV WKH QXPEHU GHQVLW\ RI WKH DEVRUEHU FPnf ,Q WKH FDVH RI WKH 09) WKH HTXDWLRQ WR GHWHUPLQH WKH DEVRUEDQFH RI DQ DWRPLF WUDQVLWLRQ FDQ EH FDOFXODWHG IURP H I6 YQ $ f§/YLf f H P F R H ZKHUH H LV WKH FKDUJH RI DQ HOHFWURQ [ n &f I\ LV WKH RVFLOODWRU VWUHQJWK IRU WKH HOHFWURQLF WUDQVLWLRQ IURP OHYHO L WR OHYHO M XQLWOHVV HPSLULFDO YDOXH IRU HDFK

PAGE 51

WUDQVLWLRQf 6Y LV WKH 9RLJW VKDSH IXQFWLRQ RI WKH DEVRUSWLRQ EDQGZLGWK +]f LLM LV WKH QXPEHU GHQVLW\ RI WKH DWRP LQ OHYHO L Pnf H' LV WKH SHUPLWWLYLW\ RI IUHH VSDFH [ n 1V&f PH LV WKH PDVV RI DQ HOHFWURQ [ n NJf DQG F LV WKH YHORFLW\ RI OLJKW [ PVf &KRLFH RI 0HWDO IRU WKH )LOWHU )RU WKH 09) WR EH XVHIXO LQ SUDFWLFH WKH HOHFWURQLF DEVRUSWLRQ WUDQVLWLRQ PXVW VWDUW IURP D JURXQG VWDWH GXH WR WKH VPDOO IUDFWLRQ RI WKH HOHPHQW H[LVWLQJ LQ DQ H[FLWHG VWDWH DV GHVFULEHG E\ WKH %ROW]PDQQ GLVWULEXWLRQ 7KHUHIRUH OHYHO L LQ (TXDWLRQ PXVW EH WKH JURXQG VWDWH DQG Qc PXVW EH WKH QXPEHU GHQVLW\ LQ WKH JURXQG VWDWH $QRWKHU LPSRUWDQW FRQVLGHUDWLRQ IRU WKH SUDFWLFDO XVH RI D 09) LV WKH WKHUPRG\QDPLFV RI WKH FKRVHQ HOHPHQW ,Q RUGHU WR DWWDLQ D ODUJH HQRXJK QXPEHU GHQVLW\ Qc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f UXELGLXP KDV EHHQ FKRVHQ DV WKH HOHPHQW WR LQFOXGH LQ WKH 09) IRU XVH LQ WKH 60' V\VWHP

PAGE 52

7DEOH 7KHUPRG\QDPLF FKDUDFWHULVWLFV RI VHYHUDO HOHPHQWV IRU SRVVLEOH XVH LQ WKH PHWDO YDSRU ILOWHU (OHPHQW PS UrFY ; QPOr J$ FPfE $ [ %D [ &D [ [ X &G [ &V [ f [ [ +J [ ,Q [ [ f [ [ /L [ [ 0J [ 1D [ [ 3E [ 5E [ [ [ 6Q [ 6U [ 7 [ =Q [ [ f9DOXHV IURP /DQJHfV +DQGERRN RI &KHPLVWU\ WK (G -$ 'HDQ (G 0F*UDZ+LOO 1HZ
PAGE 53

&DOFXODWLRQ RI 6SHFWUDO /LQHZLGWKV DQG $EVRUEDQFHV IRU 5E 7KH WZR PRVW LPSRUWDQW IHDWXUHV IRU XVH RI WKH 09) DUH WKH DEVRUSWLRQ FRHIILFLHQWV RU DEVRUSWLRQ FURVV VHFWLRQVf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n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cM YDOXHV DUH JLYHQ LQ 7DEOH DW DQG QP

PAGE 54

7DEOH 3K\VLFDO SDUDPHWHUV RI WKH UXELGLXP K\SHUILQH VWUXFWXUH UHTXLUHG IRU WKH FDOFXODWLRQ RI WKH 9RLJW VSHFWUDO SURILOH RI WKH PHWDO YDSRU ILOWHU )RU WKH $EVRUSWLRQ %DQG DW QP \Q FPnK IRU 5E IRU 5E 93T FPf ILM 33R FPf X )RU WKH $EVRUSWLRQ %DQG DW QP \Q FPn IRU 5E IRU 5E 33 FPf X SSR FPf ILM (DFK RI WKHVH OLQHV LV EURDGHQHG E\ WKH VDPH H[WHQW DV GHVFULEHG EHORZ DQG WKHLU QRUPDOL]HG DEVRUEDQFHV DUH DGGHG WRJHWKHU DW LQGLYLGXDO ZDYHOHQJWKV WR IRUP WKH 9RLJW SURILOH

PAGE 55

/RUHQW]LDQ EURDGHQLQJ 7KHUH DUH WZR W\SHV RI EURDGHQLQJ WDNLQJ SODFH LQ WKH 09) ZKLFK IDFWRU LQWR WKH 9RLJW SURILOH QDPHO\ 'RSSOHU EURDGHQLQJ DQG FROOLVLRQDO RU SUHVVXUHf EURDGHQLQJ 7KH GLIIHUHQW W\SHV RI FROOLVLRQDO EURDGHQLQJ FDQ EH JURXSHG WRJHWKHU WR SURGXFH WKH WRWDO /RUHQW]LDQ SURILOH ,Q WKH FDVH RI WKH 09) LW LV UHDOLVWLF WR DVVXPH WKDW DOO FROOLVLRQV RI WKH DWRPV LQ WKH YDSRU VWDWH RFFXU ZLWK GLDWRPLF QLWURJHQ 1f DQG DUH DGLDEDWLF LQ QDWXUH WKH DWRP UHPDLQV LQ WKH VDPH HOHFWURQLF VWDWH GXULQJ WKH FROOLVLRQf 7KXV WKH /RUHQW]LDQ OLQHZLGWK $Y/ +]f FDQ EH FDOFXODWHG IURP WKH HTXDWLRQ WR GHWHUPLQH WKH SURILOH GXH WR DGLDEDWLF FROOLVLRQV ZKLFK LV f ZKHUH 5 LV WKH JDV FRQVWDQW -. PROf 7 LV WKH WHPSHUDWXUH .f X LV UHGXFHG PDVV RI 5E DQG 1 NJf DF LV WKH FROOLVLRQDO FURVVVHFWLRQ RI 5E Pf DQG Q[ LV WKH QXPEHU GHQVLW\ RI QLWURJHQ LQ WKH PHWDO YDSRU FHOO Pn LQ WKH FDVH RI QLWURJHQ Q[ [ 37 ZLWK SUHVVXUH 3 LQ 7RUU DQG WHPSHUDWXUH 7 LQ .f )RU 5E LQ D QLWURJHQ HQYLURQPHQW DF KDV EHHQ PHDVXUHG WR EH [ P DW QP DQG [ n P DW QP 'RSSOHU EURDGHQLQJ 'RSSOHU EURDGHQLQJ $SG +]f D UHVXOW RI WKH DWRPV PRYLQJ DW GLIIHUHQW YHORFLWLHV XSRQ DEVRUSWLRQ RI OLJKW FDQ EH GHWHUPLQHG IURP D Q $ & ; $SO 57 78

PAGE 56

f \ ZKHUH 5 LV WKH JDV FRQVWDQW -.f 0 LV WKH IRUPDO ZHLJKW RI 5E NJPROf DQG YP LV WKH FHQWHU IUHTXHQF\ RI WKH RYHUDOO WUDQVLWLRQ +]f 7KH DSDUDPHWHU 7R GHWHUPLQH WKH SURSRUWLRQ RI HDFK EURDGHQLQJ HIIHFW LQ WKH RYHUDOO OLQHZLGWK WKH DSDUDPHWHU LV XVHG ZKHUH D $\/$\'f ,Q WKH FDVH RI WKH 5E PHWDO YDSRU FHOO DW SUHVVXUHV 7RUU WKH HIIHFW RI FROOLVLRQDO EURDGHQLQJ LV JUHDWHU WKDQ WKDW RI 'RSSOHU EURDGHQLQJ ZKLFK LV H[SUHVVHG E\ D 7KH 9RLJW LQWHJUDO YDOXH Df DW WKH OLQH FHQWHUV RI ERWK 5E UHVRQDQFH OLQHV FDQ EH FDOFXODWHG ZLWKLQ bf WR EH Df m D IRU D f 9RLJW OLQHZLGWK 7KH PXOWLSOLFDWLRQ RI Df E\ WKH 'RSSOHU VKDSH IXQFWLRQ 6' Vf UHVXOWV LQ WKH 9RLJW VKDSH IXQFWLRQ 6Y Vf )RU 'RSSOHU EURDGHQLQJ 6' DW IUHTXHQF\ Y +]f FDQ EH GHWHUPLQHG IURP WKH UHODWLRQVKLS r @Q f LU DQG WKH 9RLJW OLQHZLGWK $\Y +]f LV IRXQG IURP $G YB OQf57 0 $AY $Y a f$L! f $OO OLQHZLGWKV GHQRWHG E\ WKH VXEVFULSW Y FRUUHVSRQG WR WKH IXOO ZLGWK RI WKH SHDN DW KDOI RI WKH PD[LPXP LQWHQVLW\ 7R FRQYHUW DQ\ RI WKHVH OLQHZLGWKV IURP IUHTXHQF\ $\ +]f WR ZDYHOHQJWK $; QPf WKH YDOXH LV PXOWLSOLHG E\ ZDYHOHQJWK

PAGE 57

VTXDUHG DQG GLYLGHG E\ WKH YHORFLW\ RI OLJKW F [ QPVf )RU H[DPSOH $?Y DW QP LV IRXQG IURP WKH UHODWLRQVKLS $MY QPfF 5E QXPEHU GHQVLW\ $W WKLV SRLQW DOO RI WKH SDUDPHWHUV H[FHSW Qc DQG W KDYH EHHQ GHILQHG WKDW DUH QHFHVVDU\ WR GHWHUPLQH WKH DEVRUEDQFH DV GHVFULEHG E\ (TXDWLRQ DQG WKH 9RLJW VSHFWUDO VKDSH IXQFWLRQ DV GHVFULEHG E\ (TXDWLRQ )RU WKH PHWDO YDSRU FHOO  LV IL[HG DQG Qc LV GHSHQGHQW RQ WHPSHUDWXUH DQG SUHVVXUH )RU WKH HYDSRUDWLRQ RI OLTXLG 5E Q IROORZV WKH H[SUHVVLRQ ORJQcf $7 % OfORJ7f & '7 f ZKHUH $ % & DQG ZLWK 7 LQ 7KHUH KDYH EHHQ PDQ\ GLIIHUHQW PHDVXUHPHQWV RI WKH YDOXHV RI WKHVH FRQVWDQWV E\ SK\VLFLVWV ZLWK YDULDELOLW\ LQ Qc DV JUHDW DV b EXW WKH DERYH YDOXHV ZHUH FKRVHQ IURP UHIHUHQFH EHFDXVH RI WKHLU FORVH DJUHHPHQW ZLWK RWKHU UHIHUHQFHG YDOXHV DQG KLJK SUHFLVLRQ 9RLJW VSHFWUDO SURILOH 7KH RYHUDOO DEVRUSWLRQ OLQH VKDSH RI 5E YDSRU ZDV FDOFXODWHG ZLWK WKH KHOS RI D FRPSXWHU VSUHDGVKHHW DQG D FRPSXWHU SURJUDP ZULWWHQ LQ EDVLF ODQJXDJH E\ 0LFKDHO :HQVLQJf 7KH SURJUDP VXPPHG WKH DEVRUEDQFHV RI WKH LQGLYLGXDO FRPSRQHQWV RI WKH SURILOH YHUVXV ZDYHOHQJWK EDVHG RQ YDOXHV FDOFXODWHG IURP WKH DERYH HTXDWLRQV LQ VSUHDGVKHHW IRUPDW 3ORWV RI WKH 9RLJW SURILOHV DW GLIIHUHQW WHPSHUDWXUHV DQG SUHVVXUHV ZHUH WKHQ JHQHUDWHG ZLWK  RI FPf 7KHVH SORWV DUH H[KLELWHG LQ )LJXUHV DQG $V WKH ILJXUHV VKRZ LQFUHDVLQJ FHOO WHPSHUDWXUH KDV D ODUJH HIIHFW RQ WKH DEVRUEDQFH ZKLFK HVVHQWLDOO\ PLUURUV WKH LQFUHDVLQJ 5E YDSRU QXPEHU GHQVLW\ DQG LQFUHDVLQJ SUHVVXUH LQFUHDVHV OLQHZLGWK DW

PAGE 58

WKH H[SHQVH RI DEVRUEDQFH 7KH SUHVHQFH RI PRUH WKDQ RQH SHDN DW ORZ SUHVVXUH H[KLELWV WKH ILQH VWUXFWXUH RI WKH RYHUDOO WUDQVLWLRQ )RU WKLV ZRUN WZR GLIIHUHQW 5E PHWDO YDSRU FHOOV ZHUH DYDLODEOH IRU XVH LQ WKH SURMHFW %RWK FHOOV FRQWDLQHG PJ RI b SXUH UXELGLXP PHWDO LQ QLWURJHQ WKLV LV PRUH WKDQ HQRXJK WR SURGXFH D ODUJH YDSRU QXPEHU GHQVLW\ ZLWKRXW H[SHQGLQJ DOO RI WKH FRQGHQVHG PHWDOf 2QH FHOO WR EH UHIHUUHG DV FHOO f PDQXIDFWXUHG E\ 5XG\ 6WURKVFKHLQ LQ WKH 8QLYHUVLW\ RI )ORULGD JODVV VKRS KDV DQ DEVRUSWLRQ SDWK OHQJWK RI FP DQG D SUHVVXUH RI 7RUU DW URRP WHPSHUDWXUH 7KH RWKHU FHOO FHOO PDQXIDFWXUHG E\ 2SWKRV 5RFNYLOOH 0'f KDV D FP SDWK OHQJWK DQG D URRP WHPSHUDWXUH SUHVVXUH RI 7RUU 7KHVH SDUDPHWHUV ZHUH HQWHUHG LQWR WKH FRPSXWHU SURJUDPV WR FDOFXODWH WKH 9RLJW SURILOHV IRU WKHVH FHOOV DW r& IRU ERWK WKH QP DQG QP OLQHV 7KHVH SURILOHV DSSHDU LQ )LJXUHV DQG 7KH 9RLJW OLQHZLGWKV DUH SP DW QPf DQG SP DW QPf IRU FHOO DQG SP DW QPf DQG SP DW QPf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

PAGE 59

&D c :DYHOHQJWK 6KLIW SPf PODWHG 9RLJW DEVRUSWLRQ SURILOHV RI WKH QP 5E OLQH DW GLIIHUHQW WHPSHUDWXUHV 3DWK OHQJWK LV FP DQG 1 SUHVVXUH LV 7RUU DW r&

PAGE 60

LOFXODWHG 9RLJW DEVRUSWLRQ SURILOHV RI WKH QP DQG SDWK OHQJWK LV FP 5E OLQH DW GLIIHUHQW SUHVVXUHV 7HPSHUDWXUH LV /7? 1!

PAGE 61

)LJXUH &DOFXODWHG 9RLJW DEVRUSWLRQ SURILOHV DW QP DQG r& IRU WKH 5E FHOOV WR EH XVHG LQ WKH 60' SURMHFW /LQHZLGWKV IRU FHOOV DQG DUH SP DQG SP UHVSHFWLYHO\ /Q 2-

PAGE 62

$EVRUEDQFH )LJXUH &DOFXODWHG 9RLJW DEVRUSWLRQ SURILOHV DW QP DQG r& IRU WKH 5E FHOOV WR EH XVHG LQ WKH 60' SURMHFW /LQHZLGWKV IRU FHOOV DQG DUH SP DQG SP UHVSHFWLYHO\

PAGE 63

7KH /DVHU ,QVWUXPHQWDWLRQ IRU IOXRURPHWULF DQDO\VLV 7R SHUIRUP PROHFXODU IOXRUHVFHQFH DQDO\VLV ILYH EDVLF FRPSRQHQWV DUH QHFHVVDU\ f D VRXUFH RI OLJKW f D IOXRUHVFHQW VDPSOH f D FRQWDLQHU IRU WKH VDPSOH f D VSHFWUDO ILOWHU DQG f D GHWHFWRU ,Q WKH 60' SURMHFW WKH 09) ZKLFK LV DEOH WR UHPRYH WKH VRXUFH OLJKW ZKLOH VWLOO DOORZLQJ WKH SDVVDJH RI WKH IOXRUHVFHQFH WR WKH GHWHFWRU VHUYHV DV WKH IRXUWK FRPSRQHQW OLVWHG 7KH H[FLWDWLRQ VRXUFH LV WKH VXEMHFW RI WKLV VHFWLRQ DQG HDFK RI WKH UHPDLQLQJ FRPSRQHQWV ZLOO EH GLVFXVVHG LQGLYLGXDOO\ LQ WKH IROORZLQJ VHFWLRQV /LJKW VRXUFHV )RU VHOHFWLYH DQDO\VLV WKH H[FLWDWLRQ VRXUFH VKRXOG KDYH D QDUURZ HPLVVLRQ EDQG DQG IRU LPSURYHG VHQVLWLYLW\ WKH VRXUFH VKRXOG KDYH KLJK HPLVVLRQ LQWHQVLW\ $UF ODPS VRXUFHV DUH FDSDEOH RI YHU\ ODUJH UDGLDQFHV EXW LQ RUGHU WR REWDLQ QDUURZ EDQGZLGWK WKH OLJKW PXVW EH SDVVHG WKURXJK D PRQRFKURPDWRU ZKLFK ILOWHUV WKH H[FLWDWLRQ VRXUFH LQWR OHVV LQWHQVH EDQGV :LWK WKH LQYHQWLRQ RI WKH ODVHU D VRXUFH EHFDPH DYDLODEOH WKDW ZDV FDSDEOH RI KLJK SRZHU LQ D QDUURZ VSHFWUDO EDQGZLGWK +RZHYHU D SUREOHP ZLWK WKH ODVHU DV DQ H[FLWDWLRQ VRXUFH LV LWV ODFN RI WXQDELOLW\ 2QO\ WKRVH DQDO\WHV ZLWK DQ H[FLWDWLRQ EDQG DW WKH ODVHU HPLVVLRQ ZDYHOHQJWK FDQ EH GHWHUPLQHG E\ /,) 2I FRXUVH WKHUH DUH PDQ\ GLIIHUHQW ODVHUV ZLWK PDQ\ GLIIHUHQW ODVHU OLQHV EXW WKH FRVW DQG SUDFWLFDOLW\ RI XVLQJ VR PDQ\ GLIIHUHQW VRXUFHV PDNHV /,) DQ DQDO\WLFDO WHFKQLTXH IRU XQLTXH FLUFXPVWDQFHV

PAGE 64

&ULWHULD RI WKH /DVHU IRU 6LQJOH 0ROHFXOH 'HWHFWLRQ 7KH FKRLFH RI ODVHU WR EH XVHG LQ WKH GHWHFWLRQ RI VLQJOH PROHFXOHV PXVW EH PDGH EDVHG RQ VHYHUDO FULWHULD 7KHVH FULWHULD LQFOXGH f WKH ODVHU PXVW EH FDSDEOH RI VXIILFLHQW LUUDGLDQFH :FPf DW WKH H[FLWDWLRQ ZDYHOHQJWK RI D JLYHQ IOXRURSKRUH WR DWWDLQ 60' f WKH ODVHU HPLVVLRQ ZDYHOHQJWK PXVW IDOO DW RU EH DEOH WR EH WXQHG WR WKH DEVRUSWLRQ ZDYHOHQJWK RI WKH PHWDO YDSRU FHOO DQG UHPDLQ DW WKLV ZDYHOHQJWK RYHU WLPH f WKH HPLVVLRQ SURILOH RI WKH ODVHU PXVW EH *DXVVLDQVKDSHG ZLWK D OLQHZLGWK OHVV WKDQ WKH DEVRUSWLRQ EDQGZLGWK RI WKH PHWDO YDSRU f WKH VRXUFH PXVW EH FRQWLQXRXV RU KDYH D SXOVH UDWH RI DW OHDVW m N+] LQ RUGHU WR HIILFLHQWO\ SUREH D IORZLQJ VDPSOH 2WKHU GHVLUDEOH WUDLWV RI WKH VRXUFH LQFOXGH ORZ FRVW VLPSOH PDLQWHQDQFH DQG HDV\ RSHUDWLRQ &KRRVLQJ WKH ODVHU 0RVW ODVHUV PHHW FULWHULRQ f LI WKH EHDP FDQ EH IRFXVHG WR D YHU\ VPDOO DUHD %HFDXVH QR QRQWXQDEOH ODVHUV KDSSHQ WR HPLW OLJKW DW DQ DEVRUSWLRQ EDQG RI DQ HOHPHQW IRU SUDFWLFDO XVH LQ D 09) FULWHULRQ f HVWDEOLVKHV WKDW D WXQDEOH ODVHU PXVW EH XVHG 2QO\ G\H ODVHUV GLRGH ODVHUV 5DPDQVKLIWHG ODVHUV DQG WKH 7LUVDSSKLUH ODVHU DUH WXQDEOH 5HTXLUHPHQW f QHJDWHV 5DPDQ VKLIWHG ODVHUV DQG DOO SXOVHG G\H ODVHUV H[FHSW WKH FRSSHU YDSRUG\H ODVHU V\VWHP &ULWHULRQ f SUHFOXGHV WKH XVH RI D SXOVHG ODVHU XQOHVV LW LV FDSDEOH RI D UHSHWLWLRQ UDWH JUHDWHU WKDQ r N+] 7KH XSSHU OLPLW RI WKH UHSHWLWLRQ UDWH RI WKH &X YDSRU

PAGE 65

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f LQH[SHQVLYH f HDV\ WR RSHUDWH f PDLQWHQDQFH IUHH f ORQJOLYHG f UREXVW f VPDOO f WXQDEOH f SRZHUIXO f HIILFLHQW f YHUVDWLOH f KLJK SRZHU VWDELOLW\ DQG f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

PAGE 66

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f ODVHU PHHWV DOO RI WKH RXWOLQHG FULWHULD 7DEOH OLVWV WKH VSHFLILFDWLRQV RI WKH PDQXIDFWXUHU RI WKH 7LVDSSKLUH ODVHU SXUFKDVHG IRU XVH LQ WKLV SURMHFW $V WKH WDEOH VKRZV WKH ODVHU LV FDSDEOH RI WZR ODVLQJ DUUDQJHPHQWV WKH VWDQGLQJZDYH DQG ULQJ FDYLW\ FRQILJXUDWLRQV 7KH VSHFLILFDWLRQV IRU ERWK FRQILJXUDWLRQV DUH WKH VDPH H[FHSW WKDW WKH VSHFWUDO OLQHZLGWK LV QDUURZHU LQ WKH ULQJ FRQILJXUDWLRQ

PAGE 67

)LJXUH LV D GUDZLQJ RI WKH GHVLJQ RI WKH 7LUVDSSKLUH ODVHU 7KH GLDJUDP VKRZV WKH ODVHU LQ WKH ULQJ FDYLW\ DUUDQJHPHQW ZLWK WKH EHDP SDVVLQJ WKURXJK WKH RSWLFDO GLRGH ,Q WKH VWDQGLQJZDYH FRQILJXUDWLRQ WKH WZR IODW PLUURUV DUH WLOWHG WR UHWXUQ WKH EHDP WR WKH FXUYHG PLUURUV RQ HLWKHU VLGH RI WKH FU\VWDO LQVWHDG RI WKURXJK WKH RSWLFDO GLRGH 6SHFLILFDWLRQV RI WKH WLWDQLXPUVDSSKLUH ODVHU XVHG LQ WKH VLQJOH PROHFXOH GHWHFWLRQ SURMHFW 7DEOH 3DUDPHWHU 0D[LPXP 3RZHU DW QPf 7XQLQJ 5DQJH 6SHFWUDO /LQHZLGWK DW : 3XPS 3RZHUf 6SHFWUDO 3URILOH 6SDWLDO 3URILOH 3RODUL]DWLRQ %HDP 6L]H 6SHFLILHG 9DOXH P: b RI SXPS SRZHU XS WR :f QP *+] VWDQGLQJ ZDYH 0+] ULQJ FRQILJXUDWLRQ *DXVVLDQ 7(0RR +RUL]RQWDO PP DW H[LW 7KH 7LVDSSKLUH ODVHU LV D SDVVLYH GHYLFH ZLWK QR PRYLQJ RU HOHFWURQLF SDUWV $Q LPSRUWDQW DVSHFW RI WKH 7LUVDSSKLUH ODVHU LV WKDW LW UHTXLUHV WKH VHUYLFHV RI D SXPS ODVHU WR LQLWLDWH WKH ODVLQJ DFWLRQ 7KH VRXUFH RI WKH ODVLQJ LV D VROLG VWDWH $ FU\VWDO GRSHG ZLWK WLWDQLXP 7KH FU\VWDO KDV EURDG H[FLWDWLRQ DQG HPLVVLRQ EDQGV ZLWK PD[LPD r QP DQG QP UHVSHFWLYHO\ 'XH WR SHDN DEVRUSWLRQ LQ WKH JUHHQ SDUW RI WKH VSHFWUXP DUJRQ LRQ 1G<$* DQG FRSSHU YDSRU ODVHUV DUH XVHG WR SXPS WKH 7LUVDSSKLUH ODVHU 7KH $U ODVHU LV XVHG PRVW RIWHQ WR WDNH DGYDQWDJH

PAGE 68

)LJXUH 7KH RSWLFDO DUUDQJHPHQW RI WKH 7LVDSSKLUH ODVHU XVHG LQ WKLV SURMHFW

PAGE 69

RI LWV FRQWLQXRXV ODVHU DFWLRQ EXW SXOVHG RSHUDWLRQ FDQ EH DSSOLFDEOH LQ PDQ\ VLWXDWLRQV 7KH RXWSXW SRZHU RI WKH 7LVDSSKLUH LV SURSRUWLRQDO WR WKH SRZHU RI WKH SXPS ODVHU DQG DFFRUGLQJ WR WKH PDQXIDFWXUHU VSHFLILFDWLRQV LV FDSDEOH RI b HIILFLHQF\ RI FRQYHUVLRQ EXW LQ WKH 60' H[SHULPHQWV QHDUO\ b HIILFLHQF\ ZDV DWWDLQHG ZLWK FDUHIXO DOLJQPHQW &RPSDULVRQ RI WKH 7KVDSSKLUH ODVHU ZLWK G\H ODVHUV 7KH 7LVDSSKLUH ODVHU ZDV LQYHQWHG LQ WKH V DQG LV SRLVHG WR UHSODFH WKH G\H ODVHU IRU PDQ\ DSSOLFDWLRQV ,Q WKH HDUO\ V G\H ODVHUV ZHUH KDLOHG DV WKH VROXWLRQ WR WKH ODFN RI ODVHU WXQDELOLW\ 7KHUH DUH VHYHUDO SUREOHPV ZLWK G\H ODVHUV KRZHYHU FRQFHUQLQJ ERWK VFLHQWLILF DQG SUDFWLFDO DVSHFWV LQ WKHLU XVH (DFK G\H KDV D OLPLWHG WXQLQJ UDQJH EDVHG RQ FRQFHQWUDWLRQ VROYHQW S+ DQG SXPS ODVHU 3RZHU LV VWURQJO\ GHSHQGHQW RQ ZDYHOHQJWK DQG LV JUHDWO\ FKDQJHG ZKHQ D G\H ODVHU LV VFDQQHG HIILFLHQF\ RI FRQYHUVLRQ LV RIWHQ OHVV WKDQ b IRU G\H ODVHUV )RU WKHVH UHDVRQV G\H ODVHUV DUH VHOGRP XVHG WR VFDQ RYHU D ZDYHOHQJWK UDQJH DQG WKH PRUH FRPPRQ DSSOLFDWLRQ LV WR WXQH D G\H ODVHU WR D IL[HG ZDYHOHQJWK ,Q SUDFWLFH G\H ODVHUV DUH QRWHG IRU WKH SUREOHPV G\H GHFRPSRVLWLRQ VROYHQW IODPPDELOLW\ FKHPLFDO WR[LFLW\ ZDVWH GLVSRVDO FKHPLFDO VSLOOV DQG H[WHQGHG GRZQ WLPH RIWHQ GXH WR VROYHQW SXPSVf )RU WKHVH UHDVRQV VSHFWURVFRSLVWV KDYH EHHQ DQ[LRXV IRU D ODVHU WR UHSODFH WKH G\H ODVHU 7KH 7LVDSSKLUH ODVHU KDV PDQ\ DGYDQWDJHV RYHU WKH G\H ODVHU EXW LV QRW WKH SDQDFHD IRU WKH ODVHU VSHFWURVFRSLVW 6LQFH WKH 7LVDSSKLUH LV D SDVVLYH VROLG VWDWH ODVHU LW KDV QRQH RI WKH SUDFWLFDO SUREOHPV DVVRFLDWHG ZLWK XVLQJ ODVHU G\HV 1R PRYLQJ SDUWV WUDQVODWHV WR QR GRZQ WLPH IRU PHFKDQLFDO UHDVRQV DQG QR G\HV PHDQV

PAGE 70

QR PHVVHV FKHPLFDO KD]DUGV RU ZDVWH $OVR WKH 7LVDSSKLUH ODVHU LV FDSDEOH RI FRQWLQXRXV WXQLQJ ZLWK PRGHUDWH FKDQJH LQ SRZHU RYHU D FRQWLQXRXV UDQJH RI QP IXQGDPHQWDOf RU QP IUHTXHQF\ GRXEOHGf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r& 7KLV DVSHFW LV VKRZQ LQ )LJXUH LQ ZKLFK WKH HPLVVLRQ SURILOHV RI WKH 7NVDSSKLUH LQ ERWK FRQILJXUDWLRQV DUH VXSHULPSRVHG RQWR WKH 9RLJW SURILOH RI FHOO DW QP $V VKRZQ WKH VSHFLILHG ODVHU OLQHZLGWKV DUH PXFK QDUURZHU WKDQ WKH SURILOH RI WKH DEVRUSWLRQ OLQH 7KLV FRQGLWLRQ LV HVSHFLDOO\ QRWDEOH LQ WKH ULQJ FDYLW\ SURILOH ZKHUH WKH ODVHU OLQH LV D WKLQ OLQH RQ WKH VDPH VFDOH DV WKH RWKHU SURILOHV

PAGE 71

)LJXUH &DOFXODWHG VSHFWUDO OLQH SURILOHV RI WKH 7LVDSSKLUH ODVHU LQ WKH VWDQGLQJZDYH DQG ULQJ FRQILJXUDWLRQV DW QP LQ FRPSDULVRQ ZLWK WKH 5E 9RLJW SURILOH RI FHOO DW r& /DVHU ,QWHQVLW\

PAGE 72

$EVRUEDQFH LV WKH ORJ RI WKH LQYHUVH RI WUDQVPLWWDQFH 7 $ ORJO7ff ZKLFK PHDQV WKDW WKH DEVRUEDQFH RI IRU WKH 5E PHWDO YDSRU ILOWHU DW r& FRUUHVSRQGV WR D WUDQVPLWWDQFH RI RU b RI WKH OLJKW LV DEVRUEHG 7KHUHIRUH ZLWK D PHWDO YDSRU DEVRUEDQFH RI WKH FDOFXODWHG PD[LPXP SHUFHQWDJH RI OLJKW VFDWWHUHG E\ WKH FDSLOODU\ LV b LI LW LV DQ LVRWURSLF SRLQW VRXUFH DW WKH IRFXV RI WKH PLFURVFRSH REMHFWLYHf WKDW ZRXOG VWLOO DOORZ IRU DQ /2' PROHFXOH EDVHG RQ WKH SDUDPHWHUV OLVWHG LQ 7DEOH RI &KDSWHU ,Q RWKHU ZRUGV WKH ODVHU VFDWWHU FROOHFWHG LQ WKH VU VROLG DQJOH RI FROOHFWLRQ IWf PXVW EH OLPLWHG WR OHVV WKDQ b b [ bf WR DFKLHYH D EDFNJURXQG FRXQW UDWH RI FRXQWV SHU PV LQFUHPHQW PD[LPXP WKHRUHWLFDO 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

PAGE 73

7NVDSSKLUH ODVHU FRQILJXUDWLRQ %HFDXVH WKH OLQHZLGWK LQ HLWKHU WKH VWDQGLQJ ZDYH FRQILJXUDWLRQ RU ULQJ FDYLW\ FRQILJXUDWLRQ LV PXFK QDUURZHU WKDQ WKH 9RLJW SURILOH RI WKH 5E YDSRU FHOO LW GRHV QRW PDWWHU ZKLFK ODVHU FRQILJXUDWLRQ LV XVHG IRU VLQJOH PROHFXOH GHWHFWLRQ %RWK FRQILJXUDWLRQV ZHUH WHVWHG IRU XVH LQ WKH SURMHFW ZLWK QR VLJQLILFDQW GLIIHUHQFH LQ UHVXOWV 7KH 6DPSOH $IWHU WKH H[FLWDWLRQ VRXUFH WKH VDPSOH FRQVWLWXWHV WKH VHFRQG RI WKH ILYH EDVLF FRPSRQHQWV RI D IOXRURPHWHU ,Q WKH FDVH SUHVHQWHG VR IDU WKH DQDO\WH PXVW DEVRUE VWURQJO\ DW RU QP 5E JURXQG VWDWH HOHFWURQLF WUDQVLWLRQVf DQG IOXRUHVFH ZLWK KLJK TXDQWXP HIILFLHQF\ DW UHPRYHG ZDYHOHQJWKV )XUWKHUPRUH DV GLVFXVVHG LQ &KDSWHU WKH DQDO\WH VKRXOG QRW EH D SRO\PHU RU RWKHU FKHPLFDO VSHFLHV ZLWK D PROHFXODU ZHLJKW JUHDWHU WKDQ m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

PAGE 74

7DEOH )OXRUHVFHQW G\HV IRU SRVVLEOH XVH LQ WKH VLQJOH PROHFXOH GHWHFWLRQ SURMHFW '\H 0ROHFXODU )RUPXOD UXQf AFP QPf PD[ H$ 0AFPf \I 5KRGDPLQH TKQR &O r r r r 0HWK\OHQH %OXH &+16 &O r r r f§ 1LOH %OXH 4R+A2 &O r r r f§ 2[D]LQH &A+A1M2 &O r r r f§ ,5 &+16 1D f ,5 FKQRV & f§ ,5 &+&16 FLR G ,5 &MR+M12J6 1(W E [! 22 2 I f§ '77& &+16 E E f§ f '7'& &+16 E E f§ f '27& &+1 E E f§ f +,7& 40, E E f +',7& AA1 & OE E f§ f§ ''7& FKQ f f f§ r '4'& AAA f r f§ r '47& &+1 f f f§ f 9DOXHV LQ ZDWHU IURP 7 ,PDVDND $ 7VXNDPRWR DQG 1 ,VKLEDVKL $QDO &KHP f E9DOXHV LQ PHWKDQRO IURP $QGUHZV:LOEHUIRUFH DQG 3DWRQD\ $SSO 6SHFWURVF f f9DOXHV LQ GLPHWK\OVXOIR[LGH IURP 5& %HQVRQ DQG +$ .XHV &KHP (QJ 'DWD f G9DOXH LQ HWKDQRO IURP '-6 %LUFK +XQJHUIROG % 1DGROVL 5( ,PKRI DQG $' 'XWFK 3K\V ( 6FL ,QVWUXP f $OO RWKHU YDOXHV GHWHUPLQHG H[SHULPHQWDOO\ LQ PHWKDQRO

PAGE 75

FRQWDLQLQJ VHYHUDO DURPDWLF ULQJV ZLWK WKH OHQJWK RI WKH VWUXFWXUHV FRUUHODWLQJ WR H[FLWDWLRQ DQG HPLVVLRQ ZDYHOHQJWKV ,Q IDFW %HQVRQ DQG .XHV KDYH ZRUNHG RXW DQ HPSLULFDO UHODWLRQVKLS EHWZHHQ WKH VWUXFWXUH RI SRO\PHWKLQH G\HV DQG WKHLU IOXRUHVFHQW FKDUDFWHULVWLFV 2I WKH G\HV OLVWHG LQ 7DEOH RQO\ WKUHH ,5 ,5 DQG ,5 DUH SRWHQWLDOO\ XVHIXO IRU VLQJOH PROHFXOH GHWHFWLRQ ZLWK H[FLWDWLRQ DW RU QP )LJXUH JLYHV WKH VWUXFWXUHV RI WKHVH SRO\PHWKLQH G\HV ,W LV UDWKHU IRUWXLWRXV WKDW ;H[Qf[ RI ,5 IDOOV DW WKH QP 5E OLQH DQG WKDW ,5 KDV ;H[Pr[ DW ZKLFK LV YHU\ QHDU WKH QP WUDQVLWLRQ 0HDQZKLOH ,5 LV WKH OHDVW IDYRUDEOH RI WKHVH G\HV LQ WKLV UHVSHFW ZLWK ;F[PD[ RI QP $QRWKHU LPSRUWDQW SDUDPHWHU IRU 60' LV WKH DEVRUSWLRQ FURVVVHFWLRQ D$ FPf RI WKH DQDO\WH ZKLFK LV SURSRUWLRQDO WR WKH PRODU DEVRUSWLYLW\ H$ RI WKH G\H LQ EXON VROXWLRQ DV VKRZQ LQ &KDSWHU ,Q D VLPSOH H[SHULPHQW XVLQJ +HZOHWW 3DFNDUG $ DQG 9DUDQ IRU ZDYHOHQJWKV JUHDWHU WKDQ QPf VSHFWURSKRWRPHWHUV WKH DEVRUSWLRQ VSHFWUD RI NQRZQ FRQFHQWUDWLRQV RI WKH G\HV LQ PHWKDQRO ZHUH PHDVXUHG 7KH ZDYHOHQJWKV RI PD[LPXP DEVRUSWLRQ FRUUHVSRQGHG WR WKH IOXRUHVFHQFH H[FLWDWLRQ PD[LPD IRU WKH G\HV DQG WKH PHDVXUHG IRU ,5 ,5 DQG ,5 ZHUH [ [ V DQG [ 0n:n UHVSHFWLYHO\ 7KHVH ILJXUHV DUH LQ DJUHHPHQW ZLWK YDOXHV UHSRUWHG LQ WKH OLWHUDWXUH 7KH HPLVVLRQ VSHFWUD RI WKH G\HV LQ PHWKDQRO VROXWLRQ DUH VKRZQ LQ )LJXUH ZLWK H[FLWDWLRQ DW DQG QP )RU WKLV H[SHULPHQW WKH 7LVDSSKLUH ODVHU ZDV XVHG DV WKH H[FLWDWLRQ VRXUFH RI WKH IORZLQJ G\H VROXWLRQV FRQWDLQHG LQ D FP SDWK

PAGE 76

)LJXUH &KHPLFDO VWUXFWXUHV RI WKH SRO\PHWKLQH G\HV WR EH WHVWHG IRU XVH LQ WKH VLQJOH PROHFXOH GHWHFWLRQ SURMHFW OHQJWK FXYHWWH $ 6SH[ GRXEOH PRQRFKURPDWRU ZLWK PP VOLWV ZDV XVHG WR FROOHFW WKH VSHFWUD DQG D FRROHG +DPDPDWVX 5 SKRWRPXOWLSOLHU WXEH VHUYHG DV WKH GHWHFWRU 7KHVH VSHFWUD KDYH EHHQ QRUPDOL]HG WR D ODVHU SRZHU RI P: DQG G\H FRQFHQWUDWLRQ RI [ n 0 7KH ;HP PD[LPD IRU ,5 ,5 DQG ,5 ZLWK

PAGE 77

IOXRUHVFHQFH HPLVVLRQ VSHFWUD RI WKH SRO\PHWKLQH G\HV LQ PHWKDQRO ZLWK H[FLWDWLRQ DW WKH 5E OLQHV VSHFWUD KDYH EHHQ QRUPDOL]HG WR P: ODVHU SRZHU DQG [ 0 G\H FRQFHQWUDWLRQ 21

PAGE 78

WKLV V\VWHP ZHUH DQG QP UHVSHFWLYHO\ (DFK RI WKHVH G\HV ZDV WHVWHG IRU DSSOLFDWLRQ WR WKH VLQJOH PROHFXOH GHWHFWLRQ SURMHFW DV SUHVHQWHG LQ &KDSWHU 0RUH GHWDLOHG FKDUDFWHULVWLFV RI HDFK G\H ZLOO EH GLVFXVVHG VHSDUDWHO\ EHORZ ,5 7KH IRUPDO FKHPLFDO QDPH IRU ,5 LV DQK\GUROOGLPHWK\O> >OOGLPHWK\OVXOIREXW\OO+fEHQ]HfLQGROLQ\OLGHQH@OKHSWDWULHQ\O@ VXOIREXW\OfO+EHQ]HfLQGROLQLXP K\GUR[LGH VRGLXP VDOW LWV &KHPLFDO $EVWUDFWV 6HUYLFHV &$6f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

PAGE 79

8QLYHUVLW\ LV DFWLYHO\ SXUVXLQJ WKH XVH RI ,5 DQG RWKHU SRO\PHWKLQH G\HV IRU WDJJLQJ SXUSRVHV ,I KLV UHVHDUFK LV VXFFHVVIXO WKLV 60' WHFKQLTXH FRXOG EHFRPH YHU\ LPSRUWDQW LQ PDQ\ WDJJLQJ DSSOLFDWLRQV ,QGLUHFW IOXRURPHWULF GHWHFWLRQ $QRWKHU LQWHUHVWLQJ XVH RI ,5 FRQFHUQV LWV FKURPDWRJUDSKLF SURSHUWLHV ,W KDV EHHQ DQDO\]HG E\ KLJKSHUIRUPDQFH OLTXLG FKURPDWRJUDSK\ +3/&f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f WKH VL]H RI WKH HIIHFW RI WKH DQDO\WH RQ WKH VLJQDO RI WKH YLVXDOL]DWLRQ DJHQW NQRZQ DV WKH WUDQVIHU UDWLRf f WKH DELOLW\ WR PHDVXUH WKHVH VLJQDO IOXFWXDWLRQV RU G\QDPLF UHVHUYHf DQG f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

PAGE 80

+3/& DQG FXUUHQW ZRUN LV XQGHUZD\ WR XVH LQGLUHFW IOXRURPHWULF GHWHFWLRQ LQ FDSLOODU\ HOHFWURSKRUHVLV ZLWK WKH V\VWHP ,I VXFFHVVIXO LW ZLOO EH SRVVLEOH WR REWDLQ D /2' RI S0 IRU DQ DQDO\WH ZLWK D WUDQVIHU UDWLR QHDU XQLW\ ,QGLUHFW IOXRURPHWULF GHWHFWLRQ LV RI LQWHUHVW WR VHSDUDWLRQ VFLHQWLVWV EHFDXVH LW KDV WKH SRWHQWLDO RI EHLQJ D VHQVLWLYH DQG XQLYHUVDO GHWHFWRU ZKLFK LV D UDUH FRPELQDWLRQ ,5 ,5 LV IRUPDOO\ NQRZQ DV fGLFKORUROOGLSKHQ\ODPLQR HWK\OHQHWKLDWULFDUERF\DQLQH SHUFKORUDWH LWV &$6 QXPEHU LV >@ 7KH PDMRU XVH RI ,5 LV DV D ODVHU G\H EXW LW KDV DOVR EHHQ XVHG DV WKH DQDO\WH LQ VHYHUDO /,) H[SHULPHQWV ZLWK GLRGH ODVHU H[FLWDWLRQn 7KH JRDO RI WKHVH SURMHFWV ZDV YHU\ PXFK VLPLODU DV WKLV RQH ZKLFK ZDV WR DWWDLQ WKH VPDOOHVW SRVVLEOH OLPLW RI GHWHFWLRQ RI WKH G\H LQ D IORZLQJ VWUHDP %DVHG RQ WKHVH HDUOLHU H[SHULPHQWV ZLWK WKLV G\H DQG WKH YHU\ KLJK IOXRUHVFHQFH TXDQWXP HIILFLHQF\ <) f DQG D KLJK VSRQWDQHRXV HPLVVLRQ $ [ Vn EDVHG RQ SV IOXRUHVFHQFH OLIHWLPH U)f ,5 LV HPSKDVL]HG LQ WKH H[SHULPHQWDO VWXGLHV ,5 /LNH ,5 WKH RQO\ NQRZQ XVH RI ,5 LV DV D ODVHU G\H ,WfV IRUPDO QDPH LV fGLDFHWR[\SURS\OfOOGLSKHQ\ODPLQRHWK\OHQHff GLEHQ]RWKLDWULFDUERF\DQLQH SHUFKORUDWH &$6 >@f :LWK WKH DGYHQW RI GLRGH ODVHUV WKHVH G\HV KDYH EHFRPH YHU\ LPSRUWDQW DV SUREH VSHFLHV 6HYHUDO DQDO\WLFDO FKHPLVWV KDYH XVHG WKHVH G\HV LQ DQ DWWHPSW WR DSSO\ GLRGH ODVHU VRXUFH WR DQDO\WLFDO WHFKQLTXHV 6LQFH GLRGH ODVHUV DUH RQO\ XVHIXO DW ZDYHOHQJWKV ORQJHU WKDQ QP DQG SRO\PHWKLQH G\HV DUH RQH RI WKH IHZ PROHFXOHV IOXRUHVFHQW DW WKHVH ZDYHOHQJWKV WKH\ KDYH EHHQ WKUXVW LQWR VHYHUDO

PAGE 81

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f DW WKLV ZDYHOHQJWK WKDQ DW WKH ZDYHOHQJWKV RI H$PD[ RI WKH G\HV 6WRFN G\H VROXWLRQV RI r [ 0 ZHUH NHSW DW URRP WHPSHUDWXUH LQ WKH GDUN (DFK WLPH D VSHFWUXP ZDV WDNHQ D P/ DOLTXRW RI VROXWLRQ ZDV SLSHWWHG LQWR D FP SDWK OHQJWK FXYHWWH DQG GLOXWHG ZLWK P/ RI VROYHQW $V WKH ILJXUH VKRZV WKHUH ZDV TXLWH VRPH GD\WRGD\ YDULDWLRQ LQ WKH SURFHGXUH EXW WKH RYHUDOO IOXRUHVFHQFH GLG QRW GHFUHDVH VLJQLILFDQWO\ 7KHUHIRUH WKH G\HV GR QRW UHDGLO\ GHFRPSRVH LQ RUJDQLF VROYHQW DVVXPLQJ WKDW WKH GHJUDGDWLRQ SURGXFWV GR QRW IOXRUHVFH DW WKH VDPH ZDYHOHQJWKV DV WKH G\HVf

PAGE 82

&2 4B F F HQ f FH 4B [ [ ; [ 0 ,5 ; ; ; [ ; ; ; [ f 0 ,5 f§ A ; ; r .f§ ; ; [ 0,5 ; a ; ; f R f§ ‘ 'D\V )LJXUH 6WDELOLW\ RI WKH SRO\PHWKLQH G\H VWRFN VROXWLRQV LQ PHWKDQRO NHSW LQ WKH GDUN RYHU D SHULRG RI GD\V 7KH YDULDWLRQV LQ WKH VLJQDO LV GXH WR GDLO\ LQVWUXPHQWDO DQG SLSHWWLQJ GHYLDWLRQV

PAGE 83

(IIHFW RI GHJDVVLQJ WKH VROXWLRQV ,Q D VLPLODU VWXG\ XWLOL]LQJ WKH VDPH LQVWUXPHQW WKH HIIHFW RI EXEEOLQJ GLIIHUHQW JDVHV RQ WKH SKRWRGHFRPSRVLWLRQ RI WKH G\H VROXWLRQV LQ PHWKDQRO ZDV H[DPLQHG ,Q WKLV VWXG\ m 0 G\H VROXWLRQV ZHUH SUHSDUHG VSOLW LQWR YROXPHV DQG FRQWLQXRXVO\ VSDUJHG ZLWK QLWURJHQ KHOLXP DQG DLU 7KHLU IOXRUHVFHQFH HPLVVLRQ VSHFWUD ZHUH PHDVXUHG DIWHU } PLQXWH 7KHVH VROXWLRQV ZHUH WKHQ SODFHG LQ D FKDPEHU UDGLDWHG E\ OLJKW RI ZDYHOHQJWKV JUHDWHU WKDQ QP )RU WKLV SXUSRVH D : (LPDF DUF ODPS ZDV SODFHG LQ WKH FKDPEHU DQG LWV ZKLWH OLJKW HPLVVLRQ ZDV ILOWHUHG E\ D VHULHV RI ORQJ SDVV VSHFWUDO ILOWHUV 7KH HPLVVLRQ VSHFWUD RI P/ DOLTXRWV RI WKHVH VROXWLRQV ZHUH PHDVXUHG SHULRGLFDOO\ RYHU D SHULRG RI KRXUV )LJXUH VKRZV WKH HIIHFW RI GHJDVVLQJ RQ WKH SKRWRGHFRPSRVLWLRQ UDWH RI ,5 LQ PHWKDQRO 7KLV ILJXUH FOHDUO\ VKRZV WKDW DLU RU PRUH VSHFLILFDOO\ R[\JHQ FDXVHV LQFUHDVHG G\H GHJUDGDWLRQ DV RSSRVHG WR DQ R[\JHQ IUHH VROXWLRQ )LJXUH IXUWKHU VKRZV WKDW QLWURJHQ DQG KHOLXP JUHDWO\ UHGXFHG WKH SKRWRGHFRPSRVLWLRQ ZLWK UHVSHFW WR DLU WKHUHIRUH DOO VROXWLRQV XVHG LQ WKH VLQJOH PROHFXOH GHWHFWLRQ SURMHFW VKRXOG EH VSDUJHG ZLWK RQH RI WKHVH WZR JDVHV %DVHG RQ FRVW DQG HDVH RI DFFHVV 1 ZDV FKRVHQ IRU WKLV SXUSRVH 7KH &KRLFH RI 6ROYHQW 6LQFH WKH JRDO RI WKLV SURMHFW ZDV WR DFKLHYH VLQJOH PROHFXOH GHWHFWLRQ ZLWK QR SDUWLFXODU VDPSOH W\SH LQ PLQG RWKHU WKDQ D VROXWLRQ WKH FKRLFH RI VROYHQW ZDV EDVHG PDLQO\ RQ VLJQDO WR QRLVH RI WKH G\H IOXRUHVFHQFH LQ WKDW VROYHQW 2I FRXUVH ZDWHU ZRXOG EH WKH LGHDO VROYHQW GXH WR LWV JUHDW DSSOLFDELOLW\ WR PDQ\ W\SHV RI DQDO\VHV EXW

PAGE 84

6LJQDO :DYHOHQJWK QPf )LJXUH 6WDELOLW\ RI ,5 LQ PHWKDQRO LQ WKH SUHVHQFH RI GLIIHUHQW JDVHV RYHU D WZR KRXU SHULRG XQGHU QHDUn LQIUDUDG OLJKW DV

PAGE 85

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r HE $FHWRQH &+&2&+ $FHWRQLWULOH &+&1 'LPHWK\OVXOIR[LGH &+62&+ (WKDQRO 4+V2+ m+H[DQH &+&+f&+ 0HWKDQRO &+2+ :DWHU KR DQ LV UHIUDFWLYH LQGH[ DW QP EH LV GLHOHFWULF FRQVWDQW $OO YDOXHV IRU r& IURP /DQJHfV +DQGERRN RI &KHPLVWU\ WK (G -$ 'HDQ (G 0F*UDZ+LOO 1HZ
PAGE 86

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n IURP WKH

PAGE 87

1RUPDOL]HG 3HDN 6LJQDO 6ROYHQW )LJXUH 7KH HIIHFW RI GLIIHUHQW VROYHQWV RQ WKH QRUPDOL]HG PD[LPXP IOXRUHVFHQFH LQWHQVLW\ RI WKH SRO\PHWKLQH G\HV ZLWK 7LVDSSKLUH ODVHU H[FLWDWLRQ DW QP $&1 DFHWRQLWULOH 0H2+ PHWKDQRO (W2+ HWKDQRO '062 GLPHWK\OVXOIR[LGHf

PAGE 88

6LJQDO :DYHOHQJWK QPf )LJXUH 5DPDQ VSHFWUD RI IRXU VROYHQWV FRQYHUWHG WR H[FLWDWLRQ DW QP k

PAGE 89

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

PAGE 90

307 &XUUHQW S$f :DYHOHQJWK QPf )LJXUH 5DPDQ VSHFWUD RI PHWKDQRO DQG DFHWRQLWULOH ZLWK 7LVDSSKLUH ODVHU H[FLWDWLRQ DW QP

PAGE 91

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f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

PAGE 92

307 6LJQDO Q$f )LJXUH /RJORJ FDOLEUDWLRQ FXUYH RI ,5 LQ DFHWRQLWULOH ZLWK P: 7LUVDSSKLUH ODVHU H[FLWDWLRQ DW QP )ORZLQJ VDPSOH LQ IORZFHOO XVLQJ 6SH[ PRQRFKURPDWRU DW QP ZLWK PP VOLWV QP EDQGZLGWKf DQG FRROHG 5 GHWHFWLRQ 6ORSH RI LV D VLJQ WKDW GLPHULVP LV RFFXUULQJ

PAGE 93

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f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

PAGE 94

EDVHG UHVHDUFK LQ DQDO\WLFDO FKHPLVWU\ LV RYHUO\ H[SHQVLYH DQG JHQHUDOO\ LPSUDFWLFDO WKH IRUPHU PHDQV RI OHYLWDWLRQ DUH IDU PRUH FRPPRQ WKDQ WKH ODWWHU 3UDFWLFDO W\SHV RI VDPSOH OHYLWDWLRQ SRVVLEOH PDNH XVH RI DHURG\QDPLF DFRXVWLF SKRWRSKRUHWLF HOHFWURG\QDPLF DQG PDJQHWLF IRUFHV $V GLVFXVVHG LQ &KDSWHU 5DPVH\fV DSSURDFK WR 60' PDNHV XVH RI WKH DGYDQWDJHV RI HOHFWURG\QDPLF OHYLWDWLRQ $OWKRXJK VDPSOH LQWURGXFWLRQ LQWR WKH WUDS LV FRPSOLFDWHG DQG WLPH FRQVXPLQJ LW GRHV UHPHG\ VRPH SUREOHPV UHODWHG WR D IORZLQJ VDPSOH FRQWDLQHG LQ TXDUW] 7KH ZD\ LQ ZKLFK 5DPVH\ FRQWDLQV WKH VDPSOH LV UHYLHZHG LQ WKH IROORZLQJ SDUDJUDSK LQ RUGHU WR JLYH VROLG JURXQGV IRU FRPSDULVRQ ZLWK WKH RWKHU PHWKRGV (OHFWURGYQDPLF OHYLWDWLRQ ,Q HOHFWURG\QDPLF OHYLWDWLRQ D FKDUJHG VSHFLHV DWRP PROHFXOH SDUWLFOH RU GURSOHWf LV LQWURGXFHG LQWR D FKDPEHU FRQWDLQLQJ D ULQJ HOHFWURGH WR ZKLFK LV DSSOLHG DQ DF UDGLRIUHTXHQF\f SRWHQWLDO DQG WRS DQG ERWWRP HOHFWURGHV WR ZKLFK DUH DSSOLHG GF SRWHQWLDOV $ ILHOG DSSOLHG WR WKH ULQJ HOHFWURGH FRQWUROV WKH ODWHUDO SRVLWLRQ RI WKH FKDUJHG VSHFLHV DQG WKH GF SRWHQWLDOV DSSOLHG WR WKH WRS DQG ERWWRP HOHFWURGHV FRQWURO WKH YHUWLFDO PRWLRQ )LJXUH LQ &KDSWHU VKRZV DQ HOHFWURG\QDPLF WUDS LQ WKH GUDZLQJ RI 5DPVH\fV VHWXS :LWK VDPSOHV KDYLQJ KLJK PDVVFKDUJH UDWLRV r NJ&f VXFK DV WKH FKDUJHG PLFURQVL]HG GURSOHWV LQWURGXFHG LQWR WKH FKDPEHU E\ 5DPVH\ WKH GURSOHWV FDQ EH VWDEO\ OHYLWDWHG DW WKH FHQWHU RI WKH WUDS 6DPSOH LQWURGXFWLRQ LV DFFRPSOLVKHG ZLWK D SLH]RHOHFWULF GURSOHW JHQHUDWRU DW WKH WRS HOHFWURGH RI WKH FKDPEHU WKH HOHFWURGH YROWDJHV DUH WKHQ DOWHUHG XQWLO WKH GURSOHW LV WUDSSHG 7KH VL]H RI HDFK GURSOHW PXVW EH GHWHUPLQHG IURP WKH FRPSOLFDWHG DQDO\VLV RI VFDWWHULQJ SDWWHUQV RI D +H1H ODVHU

PAGE 95

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fV JURXS PDNHV XVH RI D VKHDWK IORZ FXYHWWH IRU VDPSOH FRQWDLQPHQW 7KH VKHDWK IORZ FXYHWWH LV D W\SH RI IORZ FHOO LQ ZKLFK DQ H[WHUQDO VROYHQW VWUHDP RU VKHDWK LV XVHG WR FRPSUHVV DQ LQWHUQDO VDPSOH VWUHDP )LJXUH LQ &KDSWHU FRQWDLQV D SDUWLDO GUDZLQJ RI D VKHDWK IORZ FXYHWWH XVHG LQ .HOOHUf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f ,Q D VLWXDWLRQ PXFK OLNH WKH OLTXLG MHW EXW QRW DV VHYHUH WKH IORZ UDWHV UHTXLUHG WR LQGXFH WKH GHVLUHG QDUURZLQJ RI WKH VDPSOH VWUHDP JUHDWO\ GHFUHDVH WU :KHQ WKH IORZ UDWH LV ORZHUHG LQ DQ DWWHPSW WR LQFUHDVH WQ GLIIXVLRQ RI WKH DQDO\WH LQWR WKH VKHDWK RFFXUV DQG WKH LUUDGLDQFH RI WKH ODVHU UHTXLUHG WR FRYHU WKH HQWLUH

PAGE 96

VDPSOH VWUHDP LV QRW VXIILFLHQW IRU 60' 7R GHDO ZLWK WKLV SUREOHP .HOOHUfV JURXS GHFLGHG WR DYRLG LW DOWRJHWKHU E\ ORZHULQJ IORZ UDWH WR D SRLQW ZKHUH WU LV DGHTXDWH IRU WKH GHWHFWLRQ RI WKH VLQJOH PROHFXOHV DV WKH\ IORZ WKURXJK 9S EXW WKH ODVHU SUREHV RQO\ r b RI WKH VDPSOH IORZ UHJLRQ DQG WKH VSDWLDO ILOWHU RQO\ YLHZV D SRUWLRQ RI WKDW UHJLRQf 7KLV GRHV QRW PHHW WKH b VDPSOLQJ HIILFLHQF\ UHTXLUHPHQW RI 60' H[SUHVVHG LQ &KDSWHU 7KH FDSLOODU\ 8QOLNH WKH RWKHU VDPSOH FRQWDLQPHQW PHWKRGV GLVFXVVHG WKH FDSLOODU\ LV VLPSO\ D JODVV RU TXDUW] WXEH UHTXLULQJ QR LQVWUXPHQWDO DGMXVWPHQWV &DSLOODULHV RI YDULRXV VL]HV DUH FRPPRQO\ XVHG LQ VHYHUDO W\SHV RI DQDO\WLFDO PHWKRGV ,Q VHSDUDWLRQ WHFKQLTXHV XVLQJ FDSLOODULHV JDV FKURPDWRJUDSK\ VXSHUFULWLFDO IOXLG FKURPDWRJUDSK\ PLFURFROXPQ +3/& FDSLOODU\ HOHFWURSKRUHVLVf WKH LQQHU GLDPHWHU LGf RI WKH FDSLOODU\ LV DQ LPSRUWDQW FKDUDFWHULVWLF EHFDXVH LW HIIHFWV FROXPQ YROXPH IORZ SUHVVXUH KHDW GLVVLSDWLRQ DQG VHSDUDWLRQ IDFWRUVn ,Q JDV FKURPDWRJUDSK\ FDSLOODU\ LG LV RQ WKH RUGHU RI KXQGUHGV RI PLFURPHWHUV ZKHUHDV LQ FDSLOODU\ HOHFWURSKRUHVLV WKH FDSLOODULHV UDQJH IURP XQ LG ,W LV IRU WKLV ODWWHU WHFKQLTXH WKDW WKLV H[SHULPHQW LV GHVLJQHG VR WKH VDPSOH FDSLOODU\ LV RQ WKH RUGHU RI WKRVH XVHG IRU FDSLOODU\ HOHFWURSKRUHVLV )XUWKHUPRUH QDUURZHU FDSLOODULHV KDYH VPDOOHU SUREH YROXPHV ZKLFK LV GHVLUDEOH LQ WKH UHGXFWLRQ RI 5DPDQ VFDWWHU IURP WKH VROYHQW 7KH NH\ LQ FKRLFH RI FDSLOODU\ LG LV WKH ODVHU IRFXV VL]H DQG WU DW REWDLQDEOH IORZ UDWHV ZKLFK ZLOO GLVFXVVHG LQ ODWHU SDUDJUDSKV $V LW WXUQV RXW D WP LG FDSLOODU\ LV QHDUO\ LGHDO IRU WKH V\VWHP

PAGE 97

&DSLOODULHV FRPPHUFLDOO\ DYDLODEOH IRU FDSLOODU\ HOHFWURSKRUHVLV DUH PDGH RI IXVHG VLOLFD RU TXDUW] ZLWK DQ RXWHU GLDPHWHU RGf RI RU IrP )RU WKLV DSSOLFDWLRQ WKH WKLQQHU ZDOOHG WXELQJ ZDV FKRVHQ EHFDXVH f WKLQQHU ZDOOV DEVRUE OHVV KHDW DQG EHWWHU GLVVLSDWH KHDW WKDQ WKLFNHU ZDOOV f WKLQQHU ZDOOV KDYH D VPDOOHU LQWHUDFWLRQ YROXPH ZLWK WKH ODVHU EHDP OHVV DEVRUSWLRQ OHVV 5DPDQ VFDWWHUf f IRFXVLQJ ZLWK QDUURZHU FDSLOODULHV LV HDVLHU f WKH PLFURVFRSH FDQ EH SRVLWLRQHG FORVHU WR WKH VRXUFH RI WKH IOXRUHVFHQFH HPLVVLRQ LI QHFHVVDU\ ZLWK WKLQQHU ZDOOV DQG f WKH JUHDWHU OHQVLQJ FXUYDWXUH RI WKH QDUURZHU FDSLOODULHV IRFXV WKH ODVHU WR D JUHDWHU H[WHQW DW WKH VDPSOH WKDQ WKLFNHU FDSLOODULHV 6RPH RI WKHVH SRLQWV ZLOO EH GLVFXVVHG LQ PRUH GHWDLO LQ WKH VHFWLRQ EHORZ &DSLOODULHV XVHG IRU HOHFWURSKRUHVLV DUH FRPPHUFLDOO\ DYDLODEOH DQG DUH FRDWHG ZLWK SRO\LPLGH ZKLFK HQGRZV WKH EULWWOH TXDUW] ZLWK IOH[LELOLW\ $W WKH SUREH UHJLRQ WKH SRO\LPLGH FRDWLQJ PXVW EH UHPRYHG ZKLFK LV XVXDOO\ DFFRPSOLVKHG WKURXJK VWULSSLQJ RU EXUQLQJ LW RII %XUQLQJ LV DGYDQWDJHRXV EHFDXVH LW GRHV QRW FUHDWH JURRYHV RU VFUDWFKHV RQ WKH FDSLOODU\ ZDOOV DV VWULSSLQJ PD\ GR +RZHYHU EXUQLQJ GRHV OHDYH D VRRW\ OD\HU RQ WKH FDSLOODU\ EXW WKLV DUWLIDFW LV HDVLO\ UHPRYHG E\ ZLSLQJ ZLWK D ZHW WLVVXH 2QFH WKH SRO\LPLGH FRDWLQJ LV UHPRYHG WKH FDSLOODU\ PXVW EH KDQGOHG GHOLFDWHO\ EHFDXVH WKH TXDUW] WXEH VQDSV HDVLO\ XSRQ VOLJKW EHQGLQJ 2SWLFDO &RQVLGHUDWLRQV 5HJDUGLQJ WKH &DSLOODU\ )RFXVLQJ LQ RSWLFDO V\VWHPV LV SDUDPRXQW LQ DWWDLQLQJ WKH RSWLPXP FRQGLWLRQV LQ D VSHFWURVFRSLF DQDO\VLV 7KH URXQGHG VXUIDFHV RI D QDUURZ FDSLOODU\ FUHDWH VHYHUDO

PAGE 98

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b ODVHU VFDWWHU RI WKH WRWDO ODVHU LQWHQVLW\ FDQ EH WROHUDWHG LQ WKH GLUHFWLRQ RI WKH FROOHFWLRQ RSWLFV DFFRUGLQJ WR FDOFXODWLRQV SUHVHQWHG LQ &KDSWHU 7KURXJK HODERUDWH H[SHULPHQWDWLRQ DQG RSWLFDO UD\ WUDFLQJ %UXQR HW DO KDYH VKRZQ WKDW WKH OHDVW DPRXQW RI VFDWWHU IURP D URXQG FDSLOODU\ RFFXUV LQ WKH GLUHFWLRQ SHUSHQGLFXODU WR WKH H[FLWDWLRQ EHDP 7KLV LV EHFDXVH WKH LQFUHDVHG DQJOH RI LQFLGHQFH RI WKH LQFRPLQJ OLJKW DW WKH HGJHV RI WKH

PAGE 99

FDSLOODU\ FDXVHV FORVHO\ VSDFHG UD\V WR GLVSHUVH PRUH ZLGHO\ DW r WR WKH ODVHU EHDP )OXRUHVFHQFH LV WR EH FROOHFWHG IURP WKLV GLUHFWLRQ rf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f WKLV IRUP RI VFDWWHU LV H[SHFWHG WR EH QHJOLJLEOH DV VKRZQ IURP WKHRU\ LQ &KDSWHU 5DPDQ VFDWWHU IURP WKH FDSLOODU\ 5DPDQ VFDWWHU IURP WKH FDSLOODU\ PD\ SURYH WR EH D PRUH IRUPLGDEOH SUREOHP WKDQ 5D\OHLJK VFDWWHU 4XDUW] KDV D 5DPDQ HPLVVLRQ EDQG DW FPn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n FP SHU PROHFXOH DQG WKH YROXPH RI IXVHG VLOLFD LOOXPLQDWHG LV DERXW Q/ :LWKRXW NQRZLQJ WKH 0: RI TXDUW] WKH QXPEHU

PAGE 100

RI H[SHFWHG SKRWRQV DULVLQJ IURP WKH TXDUW] FDQQRW EH FDOFXODWHG ,W LV HVWLPDWHG EDVHG RQ WKH SDUDPHWHUV SUHVHQWHG LQ 7DEOH LQ &KDSWHU WKDW WKH PD[LPXP VLJQDO IURP WKH TXDUW] RI WKH FDSLOODU\ LV RQ WKH RUGHU RI SKRWRHOHFWURQV SHU PV FRXQWLQJ LQWHUYDO )RFXVLQJ WKH /DVHU /DVHU IRFXV RQWR WKH FDSLOODU\ )LJXUH LV D GLDJUDP JHQHUDWHG E\ D FRPPHUFLDOO\ DYDLODEOH RSWLFDO UD\ WUDFLQJ SURJUDP %HDP 6WHOODU 6RIWZDUH %HUNHOH\ &$f WKDW UHSUHVHQWV WKH ODVHU IRFXVLQJ RQWR WKH P LG RQ RG TXDUW] FDSLOODU\ WR EH XVHG LQ WKLV V\VWHP 7KH LQFRPLQJ OLJKW LV FROOLPDWHG DQG ZDV DVVXPHG WR EH RI QP ZDYHOHQJWK 7KH TXDUW] KDV D UHIUDFWLYH LQGH[ Q RI DW WKLV ZDYHOHQJWK DQG WKH LQQHU WXEH ZDV VDLG WR FRQWDLQ PHWKDQRO Q DW QPf 7KH IRFXVLQJ RI WKH ODVHU EHDP E\ WKH FDSLOODU\ ZDOO LV DQ )LJXUH )RFXVLQJ DVSHFWV RI WKH PP LG PP RG TXDUW] FDSLOODU\ FRQWDLQLQJ PHWKDQRO ZLWK D PP EHDP RI FROOLPDWHG OLJKW DW QP HQWHULQJ IURP WKH OHIW 'LDJUDP JHQHUDWHG E\ DQ RSWLFDO UD\ WUDFLQJ SURJUDP

PAGE 101

LQWHUHVWLQJ IHDWXUH RI WKLV V\VWHP 7KH SRLQW RI WKLV GLDJUDP LV WKDW LQ RUGHU WR DFKLHYH b VDPSOLQJ HIILFLHQF\ WKH GLDPHWHU RI WKH IRFXVHG ODVHU EHDP DW WKH RXWHU ZDOO RI WKH FDSLOODU\ VKRXOG EH VOLJKWO\ JUHDWHU WKDQ WKH LG RI WKH FDSLOODU\ WR LOOXPLQDWH WKH HQWLUH VDPSOH 7KH FDOFXODWHG PLQLPXP ODVHU IRFXV GLDPHWHU IRU WKH LP LG XQ RG FDSLOODU\ WR LOOXPLQDWH WKH HQWLUH VDPSOH UHJLRQ ZDV m XQ DW WKH RXWHU ZDOO :KHQ D XQ LG XQ RG FDSLOODU\ ZDV PRGHOHG XQGHU WKH VDPH FRQGLWLRQV WKH OLJKW UD\V ZHUH QRW IRFXVHG WR WKH VDPH H[WHQW DV LQ WKH QDUURZHU WXEH DQG UD\V DW WKH HGJH RI WKH ODVHU GLG QRW SUREH WKH VDPSOH 7KLV IHDWXUH UHIHUV WR SRLQW HDUOLHU LQ ZK\ WKH QDUURZHU RG FDSLOODU\ ZDV FKRVHQ IRU WKH 60' VHWXS 7KH ODVHU IRFXVLQJ OHQV 6LQFH WKH ODVHU EHDP ZDV a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

PAGE 102

VWHHS WKH EHDP DW WKH ZDLVW LV PRUH RU OHVV FRKHUHQW DQG WKH ZDLVW SHUVLVWV RYHU D ORQJHU GLVWDQFH )RU WKLV H[SHULPHQW DQ DFKURPDW OHQV KDYLQJ D LQFK IRFDO OHQJWK DQG LQFK GLDPHWHU ZDV FKRVHQ $Q DFKURPDW UHGXFHV WKH VSKHULFDO DQG FKURPDWLF DEHUUDWLRQV RI D UHJXODU OHQV WKH ODVHU EDQGZLGWK LV VR QDUURZ WKDW FKURPDWLF DEHUUDWLRQV GR QRW GHYHORSf DQG WKH DQWLUHIOHFWLYH FRDWLQJ RQ WKH OHQV UHGXFHV ODVHU VFDWWHU )LJXUH LV WKH UHVXOW RI D VWXG\ WR GHWHUPLQH WKH RSWLPXP IRFXV GLVWDQFH RI WKH OHQV ZLWK WKH ODVHU ,Q WKLV VWXG\ D DQ VOLW ZDV WUDQVODWHG XVLQJ FDOLSHUHG RSWLFDO PLFURSRVLWLRQHUV GULYHQ E\ D VWHSSLQJ PRWRUf WKURXJK WKH ODVHU EHDP DW VHW GLVWDQFHV IURP WKH OHQV $ SKRWRGLRGH EHKLQG WKH VOLW PHDVXUHG ODVHU LQWHQVLW\ DQG WKH ODVHU EHDP GLDPHWHU ZDV GHWHUPLQHG IURP ZKHQ LQWHQVLW\ IHOO WR D YDOXH H bf RI WKH SHDN 7KH PLQLPXP EHDP ZDLVW ZDV IRXQG WR EH XQ LQ GLDPHWHU DQG RFFXUUHG RYHU D PP UHJLRQ LQFKHV IURP WKH OHQV $ ODUJH IRFDO UHJLRQ LV WKH GHVLUHG WUDLW IRU IRFXVLQJ WKH ODVHU EHDP LQ WKLV H[SHULPHQW &ROOHFWLRQ RI WKH )OXRUHVFHQFH &ROOHFWLRQ RI HPLVVLRQ IURP WKH FDSLOODU\ 'XH WR WKH LVRWURSLF QDWXUH RI IOXRUHVFHQFH HTXDO HPLVVLRQ LQ DOO GLUHFWLRQVf RSWLFDO FRQVLGHUDWLRQV LQYROYLQJ WKH FDSLOODU\ DQG WKH FROOHFWLRQ RI IOXRUHVFHQFH DUH UDWKHU VWUDLJKWIRUZDUG 7KH VDPH FRPSXWHU VRIWZDUH ZDV XVHG WR JHQHUDWH UHSUHVHQWDWLYH GLDJUDPV RI WKH HIIHFW RI WKH FDSLOODU\ RQ WKH IOXRUHVFHQFH HPLVVLRQ IURP ZLWKLQ WKH FDSLOODU\ )LJXUH VKRZV WKH HIIHFW RI WKH FDSLOODU\ RQ SRLQW VRXUFH HPLVVLRQ DW GLIIHUHQW VLWHV ZLWKLQ WKH FDSLOODU\

PAGE 103

)RFXVHG %HDP 'LDPHWHU PPf 'LVWDQFH IURP /HQV PPf )LJXUH )RFXV VL]H RI WKH QP 7LUVDSSKLUH ODVHU EHDP YHUVXV GLVWDQFH IURP WKH LQFK IRFDO OHQJWK DFKURPDWLF OHQV DV PHDVXUHG E\ WUDQVODWHG D P VOLW DFURVV WKH EHDP

PAGE 104

)LJXUH 3RLQW VRXUFH IOXRUHVFHQFH HPLVVLRQ DW QP IURP SRLQWV ZLWKLQ WKH FDSLOODU\ FRQWDLQLQJ PHWKDQRO 'LDJUDP JHQHUDWHG E\ DQ RSWLFDO UD\ WUDFLQJ SURJUDP ,Q WKLV FDVH WKH OLJKW ZDV VDLG WR EH QP QTLXX] f DQG PHWKDQRO ZDV DJDLQ WKH VROYHQW )URP WKH FHQWHU RI WKH FDSLOODU\ WKH HPLVVLRQ LV QRW DOWHUHG RQ WKH VDPH SODQH E\ WKH URXQG ZDOOV EXW WKH HPLVVLRQ IURP SRLQWV HOVHZKHUH LQ WKH FDSLOODU\ DUH GLIIHUHQWLDOO\ EHQW DV WKH GLDJUDP VKRZV

PAGE 105

7KH PLFURVFRSH REMHFWLYH $ PLFURVFRSH REMHFWLYH ZDV FKRVHQ DV WKH PHDQV WR FROOHFW WKH IOXRUHVFHQFH HPLVVLRQ EDVHG RQ LWV ODUJH VROLG DQJOH RI FROOHFWLRQ DQG KLJK PDJQLILFDWLRQ )LEHU RSWLFV ZHUH DOVR FRQVLGHUHG EXW WKH ORZHU VROLG DQJOH RI FROOHFWLRQ ) VUf DQG VPDOOHU RSWLFDO WUDQVPLWWDQFH RI RSWLFDO ILEHUV SURPSWO\ QXOOLILHG WKHLU XVH IRU WKLV SXUSRVH 7KH LPSRUWDQW SDUDPHWHUV RI LQWHUHVW IRU D PLFURVFRSH REMHFWLYH DUH LWV PDJQLILFDWLRQ DQG QXPHULFDO DSHUWXUH 1$ 1XPHULFDO DSHUWXUH DQG WKH FDOFXODWLRQ RI IL) DUH GHILQHG LQ &KDSWHU %DVLFDOO\ WKH ODUJHU WKH 1$ WKH PRUH OLJKW LV FROOHFWHG IRU WKH REMHFWLYH 7KH FROOHFWLRQ RSWLFV FKRVHQ IRU WKLV H[SHULPHQW LV D ; PDJQLILFDWLRQf REMHFWLYH ZLWK D 1$ RI ,W PHHWV WKH VWDQGDUG VSHFLILFDWLRQV IRU PRVW PLFURVFRSHV ZLWK ZRUNLQJ GLVWDQFH RI PP DQG PP DSHUWXUH )LHOG RI GHSWK DQG ILHOG RI YLHZ ,Q D VWXG\ VLPLODU WR WKH RQH SHUIRUPHG WR GHWHUPLQH RSWLFDO IRFXVLQJ RI WKH ODVHU WKH FROOHFWLRQ SURSHUWLHV RI WKH PLFURVFRSH REMHFWLYH ZHUH PHDVXUHG 7R GHWHUPLQH WKH RSWLPDO GLVWDQFH WR SRVLWLRQ WKH HPLVVLRQ VRXUFH D P VOLW HPLWWLQJ WKH GLIIXVHG HPLVVLRQ IURP D WXQJVWHQ OLJKW EXOE ZDV XVHG DV D SRLQW VRXUFH %HJLQQLQJ DW WKH SRLQW ZKHUH WKH VOLW DQG RXWHU OHQV ZHUH WRXFKLQJ WKH PLFURVFRSH REMHFWLYH ZDV WUDQVODWHG DZD\ IURP WKH VOLW $ SKRWRGLRGH SRVLWLRQHG DJDLQVW WKH EXWW RI WKH REMHFWLYH PHDVXUHG WKH DPRXQW RI OLJKW WUDQVPLWWHG E\ WKH REMHFWLYH )LJXUH VKRZV D SORW RI SKRWRGLRGH VLJQDO YHUVXV GLVWDQFH IURP WKH VOLW WKH REMHFWLYH ZDV PRYHG KRUL]RQWDOO\ DW HDFK SRLQW WR ILQG WKH PD[LPXP VLJQDOf 7KLV SORW JLYHV WKH ILHOG RI YLHZ RI WKH PLFURVFRSH REMHFWLYH DQG DFFRUGLQJ WR WKLV VWXG\ WKH RSWLPXP GLVWDQFH IRU FROOHFWLRQ RI WKH IOXRUHVFHQFH LV P

PAGE 106

3HDN ,QWHQVLW\ RFFXUV ZLWK 6OLW PP IURP 0LFURVFRSH 2EMHFWLYH =22 =22 = =22 'LVWDQFH RI 6OLW IURP 2EMHFWLYH PPf LHOG RI GHSWK RI WKH ; 1$ PLFURVFRSH REMHFLYH XVHG LQ WKLV SURMHFW GHWHUPLQHG E\ PRYLQJ D Q VOLW HPDQDWLQJ D SRLQW VRXUFH RI OLJKW JUDGXDOO\ DZD\ IURP WKH REMHFWLYH 7KH SKRWRGLRGH ZDV SODFHG UHFWO\ EHKLQG WKH PLFURVFRSH REMHFWLYH RR

PAGE 107

! )LHOG RI YLHZ RI WKH ; 1$ PLFURVFRSH REMHFWLYH XVHG LQ WKLV SURMHFW GHWHUPLQHG E\ WUDQVODWLQJ D [P VOLW HPDQDWLQJ D SRLQW VRXUFH RI OLJKW LQ IURQW RI WKH REMHFWLYH DW P DZD\ 7KH SKRWRGLRGH ZDV SODFHG GLUHFWO\ EHKLQG WKH PLFURVFRSH REMHFWLYH 62

PAGE 108

)LJXUH LV WKH SKRWRGLRGH VLJQDO LQ WKH VDPH H[SHULPHQW DV WKH PLFURVFRSH REMHFWLYH LV WUDQVODWHG KRUL]RQWDOO\ DFURVV WKH VOLW DW D SRLQW P DZD\ 7KLV LV NQRZQ DV WKH ILHOG RI YLHZ RI WKH REMHFWLYH ZKLFK LQ WKLV FDVH LV DSSUR[LPDWHO\ PP LQ GLDPHWHU 7KLV PHDQV WKDW OLJKW FDQ HQWHU WKH FDSLOODU\ IURP UHJLRQV ZHOO RXWVLGH RI WKH FDSLOODU\ EXW WKLV DVSHFW RI WKH OHQV LV DOVR ZKDW DOORZV IRU WKH VU VROLG DQJOH RI FROOHFWLRQ IRU WKH IOXRUHVFHQFH 'HWHFWLRQ &KRLFH RI WKH 'HWHFWRU 7KH GHWHFWRU LV WKH ODVW RI WKH ILYH EDVLF FRPSRQHQWV RI D IOXRUHVFHQFH LQVWUXPHQW WR EH GLVFXVVHG )RU WKLV SURMHFW WKH FULWHULD RI WKH GHWHFWRU LQFOXGH f KLJK GHWHFWLRQ HIILFLHQF\ Uc DW WKH IOXRUHVFHQFH ZDYHOHQJWKV IRU WKH G\HV QPf f IDVWHU UHVSRQVH WLPH WKDQ WKH UHVLGHQFH WLPH RI D PROHFXOH LQ WKH SUREH YROXPH f DELOLW\ WR GHWHFW QHDU VLQJOH SKRWRQ HYHQWV DQG f WKH DUHD RI GHWHFWLRQ VKRXOG EH RI UHDVRQDEOH VL]H 6HYHUDO SKRWRGHWHFWRUV DUH FXUUHQWO\ DYDLODEOH WKDW PHHW VRPH RI WKHVH UHTXLUHPHQWV EXW QR GHWHFWRU LV DEOH WR RXWSHUIRUP RWKHUV LQ HYHU\ FDWHJRU\ 7KHUHIRUH WKH FKRVHQ GHWHFWRU ZDV WKH RQH WKDW ZDV DEOH WR SHUIRUP JHQHUDOO\ ZHOO RYHUDOO 7DEOH LV D OLVW RI SRVVLEOH SKRWRGHWHFWRUV IRU XVH LQ WKLV SURMHFW

PAGE 109

7DEOH 3RVVLEOH GHWHFWRUV IRU XVH LQ WKLV SURMHFW 3KRWRGHWHFWRU 9 bf # QP 7E *rf $UHD PPf 61 &RVW f +DPDPDWVX 5 307f [ 2n +DPDPDWVX +& 3KRWRGLRGHf [V 5&$ 630 $YDODQFKH 3KRWRGLRGHf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

PAGE 110

7KH SKRWRPXOWLSOLHU WXEH 7KURXJK WKH SURFHVV RI HOLPLQDWLRQ WKH SKRWRPXOWLSOLHU WXEH 307f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f 7KH EHVW VLGHRQ 307 DYDLODEOH IRU WKLV DSSOLFDWLRQ ZDV WKH +DPDPDWVX 5 7KLV WXEH KDV D b FDWKRGLF HIILFLHQF\ DW QP QV ULVH WLPH JDLQ RI [ V DQG PP FDWKRGLF DUHD 7KH PRVW LPSRUWDQW IHDWXUH LV WKDW LW KDV YHU\ ORZ QRLVH FKDUDFWHULVWLFV ZLWK D GDUN FXUUHQW RI S$ DW r& RSHUDWLQJ DW 9 7KH 5

PAGE 111

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f 7KH WXEH LV SODFHG XQGHU KLJK YROWDJH VR WKDW DQ\ HOHFWURQ UHOHDVHG DW WKH FDWKRGH LV GUDZQ DW KLJK YHORFLW\ WRZDUG WKH DQRGH +RZHYHU WKH FKDUJH RI D VLQJOH HOHFWURQ LV PLQXWH VR D VHULHV RI VHYHUDO W\SLFDOO\ f G\QRGHV LV XVHG WR DPSOLI\ WKH VLJQDO $ G\QRGH FKDLQ LV EDVLFDOO\ DQ HOHWURQ PXOWLSOLHU ZKHQ DQ HOHFWURQ VWULNHV D G\QRGH VXUIDFH WZR RU PRUH HOHFWURQV DUH UHOHDVHG WR FRQWLQXH IURP G\QRGH WR G\QRGH WKXV

PAGE 112

DPSOLI\LQJ WKH VLJQDO H[SRQHQWLDOO\ %\ WKH WLPH WKH VLJQDO UHDFKHV WKH DQRGH LW FDQ FRQVLVW RI PLOOLRQV RI HOHFWURQV ZKLFK JHQHUDWH D EXUVW LQ WKH VLJQDO FXUUHQW 7KH ELJJHVW VRXUFH RI QRLVH ZLWK 307V DUH WKH WKHUPDO UHOHDVH RI HOHFWURQV IURP WKH G\QRGHV RU SKRWRFDWKRGHf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f )LJXUH LV D VHULHV RI SXOVH KHLJKW GLVWULEXWLRQ SORWV IRU WKH 5 GHWHFWRU ZKLFK LV WKH KLJK JDLQ YHUVLRQ RI WKH 5f 7KH SXOVH KHLJKW GLVWULEXWLRQ LV GHSHQGHQW RQ YROWDJH WR WKH 307 DQG DV VKRZQ LQ )LJXUH WKH GLVWULEXWLRQ RI SXOVHV VKLIWV WR ODUJHU VLJQDO OHYHOV DV WKH YROWDJH LV LQFUHDVHG )RU WKLV UHDVRQ WKH SKRWRQ FRXQWHU LV EHWWHU DEOH WR GLIIHUHQWLDWH WKH VLJQDO SXOVHV IURP WKH QRLVH SXOVHV DQG WKH 307

PAGE 113

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f 'HVSLWH WKH ORZ QRLVH FKDUDFWHULVWLF RI WKH 5 SKRWRQ FRXQWLQJ UHTXLUHV KLJK JDLQ IRU EHVW UHVXOWV 7KH 5 KDV D JDLQ RI [ ZKLFK LV RQ WKH WKUHVKROG RI XVH LQ SKRWRQ FRXQWLQJ ,W ZDV WKRXJKW WKDW WKH KLJK JDLQ [ f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

PAGE 114

$YJ &RXQWV LQ PV )LJXUH (IIHFW RI LQFUHDVLQJ YROWDJH RQ WKH VLJQDO RI WKH 5 FRROHG 307 YHUVXV GLVFULPLQDWLRQ OHYHO *UDSKHG GDWD LV WKH SRLQW VPRRWKHG DYHUDJH RXWSXW RI VFDQV ZLWK D P9 GLVFULPLQDWLRQ ZLQGRZ 5RRP OLJKW ZDV XVHG DV WKH OLJKW VRXUFH IRU WKH VLJQDO R DV

PAGE 115

6LJQDO1RLVH 5DWLR 'LVFULPLQDWLRQ /HYHO P9f )LJXUH 6LJQDO WR QRLVH UDWLR RI WKH FRROHG 5 307 YHUVXV GLVFULPLQDWLRQ OHYHO ZLWK URRP OLJKW VHUYLQJ DV WKH VLJQDO VRXUFH DQG VWDQGDUG GHYLDWLRQ RI WKH GDUN FXUUHQW EHLQJ WKH QRLVH $YHUDJH RI VFDQV RI SRLQWV LQ PV FRXQWLQJ LQWHUYDOV R ‘M

PAGE 116

IDVW UHVSRQVH WLPH SKRWRQ FRXQWLQJ HQDEOHV FRPSXWHUL]HG FRQWURO RI GHWHFWLRQ SDUDPHWHUV FRPSXWHUL]HG GDWD FROOHFWLRQ DQG GDWD SURFHVVLQJ :KHQ ZRUNLQJ ZLWK D ODUJH DPRXQW RI UDZ GDWD FRPSXWHUL]HG DQDO\VLV LV D YHU\ LPSRUWDQW IHDWXUH &RQWURO RI WKH 6DPSOH )ORZ $OWKRXJK IORZLQJ VDPSOH WKURXJK D FDSLOODU\ LV QRW DV FRPSOLFDWHG DV WKH VKHDWK IORZ FXYHWWH LW LV QRW WULYLDO DW WKH IORZ UDWHV QHFHVVDU\ IRU 60' ,Q RUGHU WR DFKLHYH WKH GHVLUHG PV UHVLGHQFH WLPH RI D IORZLQJ PROHFXOH LQ D S/ YROXPH WKH IORZ UDWH PXVW EH [/PLQ 7R DFKLHYH VXFK D ORZ IORZ UDWH UHSURGXFLEO\ UHTXLUHV WKH XVH RI D VSHFLDO SXPS HVSHFLDOO\ DW WKH KLJK SUHVVXUH HQFRXQWHUHG ZKHQ SXPSLQJ PHWKDQRO WKURXJK D LQFK ORQJ [P LG FDSLOODU\ &DSLOODU\ HOHFWURSKRUHVLV 'XH WR WKH ODPLQDU QDWXUH RI IORZ WKH IORZ YHORFLW\ LQ WKH FHQWHU RI WKH WXEH LV JUHDWHU WKDQ WKDW QHDUHU WKH WXEH ZDOOV 2QO\ FDSLOODU\ HOHFWURSKRUHVLV LQ ZKLFK WKH LPSHWXV IRU IORZ RULJLQDWHV DW WKH ZDOO LWVHOI LV DEOH WR QHDUO\ DWWDLQ SOXJ IORZ ZKHQ WKH IORZ YHORFLW\ LV HTXDO DFURVV WKH WXEH GLDPHWHUf &RQGXFWLQJ FDSLOODU\ HOHFWURSKRUHVLV RQ WKH VDPSOH FDQ DFKLHYH IORZ UDWHV DV ORZ DV [/PLQ DQG SOXJ IORZ LV FHUWDLQO\ GHVLUDEOH EXW VHWWLQJ XS D FDSLOODU\ HOHFWURSKRUHVLV V\VWHP ZRXOG EH EH\RQG WKH VFRSH RI WKH UHVHDUFK SURMHFW DW WKLV VWDJH 6\ULQJH SXPS ,Q DQ DWWHPSW WR PHHW WKH IORZ FRQGLWLRQV QHHGHG IRU 60' WZR W\SHV RI SXPSV LQ WKH ODE ZHUH WHVWHG $ V\ULQJH SXPS 6DJH 0RGHO f &DPEULGJH 0$f KDG WKH DGYDQWDJH RI FRQWLQXRXV GHOLYHU\ RI D ORZ IORZ UDWH EXW LW

PAGE 117

KDG WKH GLVDGYDQWDJHV f RI QRW EHLQJ DEOH WR RSHUDWH DW KLJK SUHVVXUH f UHTXLULQJ IUHTXHQW V\ULQJH FKDQJHV DQG f LUUHSURGXFLEOH IORZ EHFDXVH RI VWLFNLQJ RI WKH V\ULQJH 7KHVH SUREOHPV DQG WKH DGGLWLRQDO SUREOHPV RI OHDNV DQG QRQFRQWLQXRXV IORZ UDWH PDGH XVH RI WKH V\ULQJH SXPS LQ RXU ODE LQWROHUDEOH LQ WKLV DSSOLFDWLRQ +3/& SXPS $Q +3/& SXPS $XWRFKURP 0RGHO 0f %HUOLQ *HUPDQ\f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f $ VHFRQG F\OLQGHU FRQWDLQLQJ D PHDVXUHG YROXPH RI PHWKDQRO ZDV NHSW EHVLGH WKH ILUVW DQG WKH FKDQJH LQ ZHLJKW ZDV XVHG WR DFFRXQW IRU HYDSRUDWLRQ ORVVHV 7KH ZHLJKW RI PHWKDQRO ZDV FRQYHUWHG WR YROXPH IURP GHQVLW\ DQG WKH IORZ UDWH ZDV FDOFXODWHG 7KH IORZ UDWH DW WKH ORZHVW SXPS VHWWLQJ ZDV L/PLQ ZKLFK FRUUHVSRQGV WR D IORZ YHORFLW\ RI FPV DQG WU PV %DVHG RQ WKH QXPEHU RI SKRWRQV QHFHVVDU\ IRU 60' DQG WKH PLQLPXP FRXQWLQJ SHULRG RI PV IRU WKH SKRWRQ FRXQWHU D VORZHU IORZ UDWH LV QHHGHG $V VKRZQ LQ &KDSWHU LW ZDV QRW QHFHVVDU\

PAGE 118

)ORZ 5DWH P/PLQf )LJXUH &DOLEUDWLRQ RI WKH +3/& SXPS IORZLQJ PHWKDQRO WKURXJK WKH FDSLOODU\ )ORZ UDWH DW ORZHVW SXPS VHWWLQJ LV L/PLQ R

PAGE 119

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

PAGE 120

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f 7LWDQ &:%% 7LUVDSSKLUH ODVHU LQ HLWKHU WKH ULQJ RU VWDQGLQJ ZDYH FRQILJXUDWLRQ ZLWK WKH 6SHFWUD3K\VLFV 0RXQWDLQ 9LHZ &$f 0RGHO DUJRQ LRQ ODVHU 'XULQJ H[SHULPHQWV WKH $U ODVHU ZDV RSHUDWHG LQ WKH FRQVWDQW SRZHU PRGH DQG DSHUWXUH VHWWLQJ ZDV IRXQG WR JLYH WKH EHVW VSDWLDO SURILOH LQ WKH DOOOLQHV PRGH WKH 7LUVDSSKLUH UHTXLUHG D 7(0A SXPSLQJ ODVHU EHDPf 3RZHU RI WKH 7LUVDSSKLUH ODVHU ZDV PHDVXUHG ZLWK D &RKHUHQW 3DOR $OWR &$f 0RGHO SRZHU PHWHU 7KH GDUN ER[ 7KH RXWSXW RI WKH 7LUVDSSKLUH ODVHU ZDV VHQW E\ WZR PLUURUV DQG D SULVP LQWR WKH GDUN ER[ IRUPHUO\ D FRPSXWHU SULQWHU ER[f ZKLFK ZDV DSSUR[LPDWHO\ [[ IHHW ,QVLGH WKH ER[ EODFN IRDP ZDV DWWDFKHG WR WKH LQQHU ZDOOV EODFN IHOW ZDV VSUHDG RQ WKH RSWLFDO WDEOH DQG EODFN PHWDO EDIIOHV ZHUH SODFHG WR VHSDUDWH WKH GHWHFWRU IURP ODVHU VFDWWHU RWKHU WKDQ WKH DPRXQW FROOHFWHG E\ WKH PLFURVFRSH REMHFWLYHf 7KH ODVHU EHDP HQWHUHG WKH ER[ WKURXJK D LQFK ORQJ [

PAGE 121

)LJXUH %DVLF LQVWUXPHQWDO DUUDQJHPHQW IRU WKH DSSURDFK WR VLQJOH PROHFXOH GHWHFWLRQ GHVFULEHG LQ GLVVHUWDWLRQ

PAGE 122

LQFK GLDPHWHU WXEH ZLWK DQ LULV SODFHG DW WKH HQWUDQFH WR NHHS URRP OLJKW RXW RI WKH ER[ $ &RULRQ +ROOLVWRQ 0$f /* ORQJSDVV ILOWHU ZDV SRVLWLRQHG DW WKH RWKHU HQG RI WKH WXEH LQ WKH ER[ WR UHPRYH WKH VPDOO DPRXQW RI JUHHQ $U HPLVVLRQ WKDW ZDV FRQWDLQHG ZLWKLQ WKH 7LVDSSKLUH EHDP ,Q WKH ER[ URRP OLJKW ZDV H[FOXGHG WR WKH H[WHQW WKDW WKH 307 GDUN FXUUHQW ZDV DEOH WR EH UHDFKHG ZKHQ WKH ER[ ZDV FORVHG DQG WKH 307 VKXWWHU RSHQHG ZKLOH WKH ODVHU ZDV EORFNHG EHIRUH WKH FDSLOODU\f )RFXVLQJ RI WKH ODVHU $ IRFDO OHQJWK DFKURPDWLF OHQV ZLWK DQ DQWL UHIOHFWLYH FRDWLQJ IRFXVHV WKH EHDP RQWR WKH XQ LG P RG FDSLOODU\ 3RO\PLFUR 7HFKQRORJLHV 3KRHQL[ $=f LQVLGH WKH ER[ 7KH OHQV ZDV PRXQWHG RQ DQ ;< PLFURSRVLWLRQHU 1HZSRUW )RXQWDLQ 9DOOH\ &$f DQG KHLJKW ZDV DGMXVWHG ZLWK WKH SRVW KROGHU 7KH FDSLOODU\ ZDV PRXQWHG ZLWK VLPLODU SRVLWLRQLQJ FDSDELOLWLHV H[FHSW KHLJKW DGMXVWPHQWf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

PAGE 123

WKH REMHFWLYH ERWK WKH OHQV DQG FDSLOODU\ ZHUH WUDQVODWHG VWDUWLQJ IURP WKH IURQW VXUIDFH RI WKH REMHFWLYH WR IDUWKHU DZD\ XQWLO WKH PD[LPXP 307 VLJQDO ZDV UHDFKHG 7KH IRFXV VL]H ZDV WKH ILQDO DGMXVWPHQW ZKLFK ZDV DFFRPSOLVKHG E\ PRYLQJ WKH OHQV FORVHU WR RU IDUWKHU DZD\ IURP WKH FDSLOODU\ 7KH RSWLPXP GLVWDQFH RI WKH OHQV IURP WKH FDSLOODU\ ZDV r LQFKHV DQG WKH GLVWDQFH RI WKH FDSLOODU\ IURP WKH PLFURVFRSH REMHFWLYH ZDV r WP )OXRUHVFHQFH FROOHFWLRQ 7KH IOXRUHVFHQFH FROOHFWLRQ RSWLFV XVHG LQ WKHVH H[SHULPHQWV ZDV D ; 1$ 1LNRQ -DSDQf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f FRQWDLQHG PJ 5E LQ 7RQ QLWURJHQ ZLWK D FP ORQJ LQWHUQDO SDWK OHQJWK FHOO PDGH E\ 2SWKRV 5RFNYLOOH 0'f FRQWDLQHG PJ 5E LQ 7RUU 1 ZLWK D FP LQWHUQDO SDWK OHQJWK 7KH FHOOV ZHUH LQFK LQ GLDPHWHU S\UH[ F\OLQGHUV 7HPSHUDWXUH RI WKH KHDWLQJ WDSH ZDV

PAGE 124

FRQWUROOHG E\ DGMXVWLQJ WKH YROWDJH ZLWK D YDULEOH DXWRWUDQVIRUPHU DQG PRQLWRULQJ WKH WHPSHUDWXUH ZLWK D .W\SH WKHUPRFRXSOH DQG GLJLWDO UHDGRXW 2PHJD 6WDPIRUG &7f 2QFH WKHUPDO HTXLOLEULXP ZDV UHDFKHG WHPSHUDWXUH RI WKH 09) YDULHG OHVV WKDQ r& RYHU D SHULRG RI KRXUV 7\SLFDO RSHUDWLQJ WHPSHUDWXUH ZDV r& 7KH HQG RI WKH PLFURVFRSH WXEH ZDV EXWWHG DJDLQVW D 8QLEOLW] 5RFKHVWHU 1
PAGE 125

FRXQWLQJ WKH VLJQDO ZDV VHQW E\ D FDEOH WR D SUHDPSOLILHU 6WDQIRUG 5HVHDUFK 6\VWHPV 6XQQ\YDOH &$f 0RGHO 65f $Q DGMXVWDEOH r 4 VQXEEHU ZDV WHHG WR WKH RWKHU HQG RI WKH FDEOH WR UHGXFH ULQJLQJ HIIHFWV RI WKH SKRWRHOHFWURQ SXOVHV 7KH 5 UHTXLUHG ; DPSOLILFDWLRQ WR SURGXFH DQ RSWLPXP GLVFULPLQDWLRQ OHYHO RI P9 DQG WKH 5 VLJQDO ZDV DPSOLILHG ; WR JLYH D P9 GLVFULPLQDWLRQ OHYHO $ FRPSXWHU 'DWDWHFK&f FRQWUROOHG WKH 6WDQIRUG 5HVHDUFK 6\VWHPV 0RGHO 65 SKRWRQ FRXQWHUGLVFULPLQDWRU DQG ZDV XVHG IRU GDWD FROOHFWLRQ 7KH 65 VFDQV ZHUH W\SLFDOO\ SRLQWV ZLWK PV LQWHUYDOV DQG PV GZHOO WLPHV ZKLFK PHDQV WKHUH ZDV RQO\ D b VDPSOLQJ HIILFHQF\ ZLWK WKLV LQVWUXPHQW )RU DSSOLFDWLRQ RI 60' D GLIIHUHQW SKRWRQ FRXQWHU DEOH WR VFDQ FRQWLQXRXVO\ ZRXOG KDYH WR EH XVHG 3XPSLQJ $Q $XWRFKURP %HUOLQ *HUPDQ\f 0RGHO 0 +3/& SXPS W\SLFDOO\ VHW DW LWV ORZHVW IORZ VHWWLQJ f ZDV XVHG WR JHQHUDWH VROYHQW IORZ 7KH IORZ VHWWLQJ FRUUHVSRQGHG WR DQ DYHUDJH IORZ UDWH RI [/PLQ DQG UHVLGHQFH WLPH RI PV LQ WKH S/ SUREH YROXPH $W WKLV IORZ UDWH SXPS SUHVVXUH ZDV 03D 6DPSOH LQWURGXFWLRQ VLPSO\ LQYROYHG FKDQJLQJ WKH EHDNHU RI VROXWLRQ IRU WKH SXPS EXW D [/ SRUW +3/& LQMHFWRU LV DYDLODEOH IRU IXWXUH DSSOLFDWLRQV 7KH +3/& WXELQJ OHDGLQJ WR WKH FDSLOODU\ FRQWDLQHG D WP LQOLQH IULW DQG WKH LQFK ORQJ FDSLOODU\ LV FRQQHFWHG DIWHU WKH IULW ZLWK WKH XVH RI D JUDSKLWH UHGXFLQJ IHUUXOH 5HDJHQWV )LVKHU 2SWLPD PHWKDQRO ZDV XVHG IRU WKH VROYHQW LQ DOO FDOLEUDWLRQ FXUYHV EXW RWKHU WHVWHG VROYHQWV LQFOXGHG HWKDQRO )ORULGD 'LVWLOOHUV /DNH $OIUHG )/f DFHWRQH )LVKHUf +3/& JUDGH DFWHWRQLWULOH )LVKHUf KH[DQH %D[WHU%XUGLFN t -DFNVRQ 0XVNHJRQ 0,f DQG $&6 JUDGH GLPHWKO\VXOIR[LGH )LVKHUf 7KH G\HV

PAGE 126

XVHG LQ WKH SURMHFW ZHUH ODVHUJUDGH ,5 ,5 DQG ,5 DOO IURP ([FLWRQ 'D\WRQ 2+f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

PAGE 127

&+$37(5 5(68/76 $1' ',6&866,21 6WXGLHV RI WKH 0HWDO 9DSRU )LOWHU $EVRUSWLRQ 3URSHUWLHV $V VKRZQ LQ &KDSWHU WKH UXELGLXP PHWDO YDSRU ILOWHUV 09)Vf DW WHPSHUDWXUHV JUHDWHU WKDQ r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f 0RGHO GRXEOH VSHFWURPHWHU DQG WKH FDOLSHU VHWWLQJ ZDV DGMXVWHG WR PD[LPL]H WKH 5 307 UHVSRQVH DW NQRZQ ZDYHOHQJWK RI WKH PRQRFKURPDWRU 7KH FDOLSHU VHWWLQJV IRU DQG QP ZHUH VSHFLDOO\ QRWHG 7R WXQH WKH ODVHU WR D 5E WUDQVLWLRQ WKH 09) ZDV KHDWHG WR m r& DQG WKH ODVHU ZDV SDVVHG WKURXJK WKH 09) RQWR D SKRWRGLRGH 7KH FDOLSHU ZDV VORZO\ DGMXVWHG QHDU WKH SRVLWLRQ GHWHUPLQHG IURP WKH FDOLEUDWLRQ H[SHULPHQW 7KH

PAGE 128

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f 7KH FHOO ZDV SODFHG LQ D /LQGEHUJ :DWHUWRZQ :,f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r r& ZKHUH WKH H[SHULPHQWDO DEVRUEDQFH WRSV RXW DQG WKH WKHRUHWLFDOO\ GHULYHG DEVRUEDQFH FRQWLQXHV XSZDUG

PAGE 129

R 2 ORPSDULVRQ RI WKH WKHRUHWLFDO DQG H[SHULPHQWDO UHVXOWV IRU DEVRUEDQFH RI WKH 5E PHWDO YDSRU FHOO VUVXV WHPSHUDWXUH DW QP 7KH 7LVDSSKLUH ODVHU ZDV LQ WKH VWDQGLQJZDYH FRQILJXUDWLRQ ZLWK SRZHU r P: EHIRUH WKH FHOO WR

PAGE 130

7KH SORW H[KLELWHG LQ )LJXUH FDQ EH GLYLGHG LQWR WKUHH UHJLRQV f WKH GLIIHUHQFHV LQ WKH DEVRUEDQFHV IURP r& LV PRVW OLNHO\ GXH WR OLJKW ORVVHV DW WKH ZLQGRZV RI WKH 09) QRW DFFRXQWHG IRU LQ WKH WKHRU\ f IURP r& WKH GLIIHUHQFHV LQ YDOXHV DUH SRVVLEO\ GXH WR WKH GLIIHUHQFHV EHWZHHQ WKH H[WHUQDO WXEH IXUQDFH WHPSHUDWXUH RI WKH SORW DQG WKH DFWXDO WHPSHUDWXUH LQVLGH WKH 09) DQG f WKH UHDVRQ WKDW WKH SORW UHDFKHV D PD[LPXP DEVRUEDQFH DW WHPSHUDWXUH r& LV GXH WR VWUD\ OLJKW 8QGRXEWHGO\ VWUD\ OLJKW ZDV LQWHUIHULQJ ZLWK WKH PHDVXUHPHQWV DW DOO WHPSHUDWXUHV EXW LW EHFDPH WKH OLPLWLQJ IDFWRU LQ WKH DEVRUEDQFH PHDVXUHPHQWV DW WHPSHUDWXUHV r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r& LW ZDV EHOLHYHG WKDW WKH DFWXDO DEVRUEDQFH FRQWLQXHG XSZDUG DV SUHGLFWHG E\ WKHRU\ EXW WKDW WKHVH DEVRUEDQFHV ZHUH QRW PHDVXUDEOH ZLWK WKLV V\VWHP $ VLPLODU H[SHULPHQW FDUULHG RXW DW WKH QP 5E OLQH JDYH D PD[LPXP DEVRUEDQFH RI DJDLQ DW WHPSHUDWXUHV r& )RU SXUSRVHV RI FRPSDULVRQ WKH DEVRUEDQFH RI FHOO YV WHPSHUDWXUH ZDV GHWHUPLQHG XQGHU VLPLOLDU FRQGLWLRQV )LJXUH LV D VHPLORJ SORW RI WKH UHVXOWV RI

PAGE 131

$EVRUEDQFH 7HPSHUDWXUH r&f )LJXUH 6HPLORJ SORW RI WKH WKHRUHWLFDO DQG H[SHULPHQWDO UHVXOWV IRU DEVRUEDQFH RI WKH 5E PHWDO YDSRU FHOO YHUVXV WHPSHUDWXUH DW QP 7KH 7LVDSSKLUH ODVHU ZDV LQ WKH VWDQGLQJZDYH FRQILJXUDWLRQ ZLWK SRZHU m P: EHIRUH WKH FHOO Z

PAGE 132

WKLV VWXG\ DQG WKH WKHRUHWLFDO H[SHFWDWLRQV $JDLQ WKH DEVRUEDQFH UHDFKHG D PD[LPXP DQG UHPDLQHG DW WKLV YDOXH DV WKH WHPSHUDWXUH LQFUHDVHG ,Q WKH FDVH RI FHOO WKH WHPSHUDWXUH ZDV KLJKHU m r&f DV ZDV WKH DEVRUEDQFH $ f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

PAGE 133

b 9 RPSDULVRQ RI WKH WKHRUHWLFDO DQG H[SHULPHQWDO UHVXOWV IRU DEVRUEDQFH RI WKH 5E PHWDO YDSRU FHOO UVXV WHPSHUDWXUH DW QP ZKHQ WKH ODVHU LV UHIOHFWHG E\ WKH FDSLOODU\ 7KH 6 EDQGSDVV WHUIHUHQFH ILOWHU ZDV SODFHG EHIRUH WKH FRROHG 5 307 ZKLFK LV ZK\ WKH SORW EHJLQV ZLWK $ 1! /U?

PAGE 134

IOXRUHVFHQFH EHLQJ WKH WHUP IRU HPLVVLRQ RFFXUULQJ DW WKH VDPH ZDYHOHQJWK DV WKH DEVRUSWLRQ $WRPLF IOXRUHVFHQFH SURFHVVHV ZHUH H[SHFWHG WR EH QHDUO\ HQWLUHO\ TXHQFKHG E\ FROOLVLRQDO GHDFWLYDWLRQ ZLWK WKH FRQFHQWUDWHG QLWURJHQ JDV LQ WKH FHOO DQG WKH OLWWOH IOXRUHVFHQFH WKDW GRHV WDNH SODFH ZDV H[SHFWHG WR EH WRWDOO\ DEVRUEHG E\ WKH RWKHU 5E DWRPV LQ WKH FHOO VLQFH WKH IOXRUHVFHQFH RFFXUV DW WKH VDPH ZDYHOHQJWKV DV WKH DEVRUSWLRQf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n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

PAGE 135

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

PAGE 136

09) DEVRUSWLRQ VWXG\ ZLWK WKH +5 'XH WR WKH LQDGHTXDWH VSHFWUDO UHVROXWLRQ RI WKH 6SH[ IRU WKLV VWXG\ WKH H[SHULPHQW ZDV UHSHDWHG ZLWK WKH XVH RI WKH +5 P PRQRFKURPDWRU ,6$-RELQ
PAGE 137

,QWHQVLW\ 'LIIHUHQFH f§ :DYHOHQJWK QPf )LJXUH 'DWD IURP WKH OLQHDU SKRWRGLRGH DUUD\ GHWHFWRU RQ WKH +5 PRQRFKURPDWRU EDQGZLGWK r SPf WKDW VKRZV WKH RFFXUUHQFH RI VFDWWHUHG OLJKW HYHQ ZKHQ WKH r& 09) FHOO ZDV EORFNHG 7KH FRQFOXVLRQ RI WKHVH GDWD LV WKDW WKH 09) DEVRUEV WKH ODVHU EXW WKDW VFDWWHU LV VWLOO SUHYDOHQW

PAGE 138

LQ WKH VLJQDO DW WKH 7LVDSSKLUH HPLVVLRQ ZDYHOHQJWK DV HYLGHQW LQ WKH VSHFWUXP LQ ZKLFK WKH VFDQV ZHUH VXEWUDFWHG IURP RQH DQRWKHU %DVHG RQ WKH QRLVH RI WKH GLIIHUHQFH EHWZHHQ WKHVH WZR VFDQV DQG WKH XVH RI QHXWUDO GHQVLW\ ILOWHUV DV FDOLEUDWLRQ VWDQGDUGV WKH DEVRUEDQFH RI WKH 09) FHOO DW r& ZDV 7UDQVPLWWDQFH 3URSHUWLHV 5E PHWDO GHSRVLWLRQ RQ WKH 09) ZLQGRZV )RU WKH 09) WR EH XVHIXO LQ WKH 60' SURMHFW QRW RQO\ PXVW LW DEVRUE WKH ODVHU VFDWWHU EXW DOVR LW PXVW WUDQVPLW WKH IOXRUHVFHQFH 7KH JUHDW DGYDQWDJH RI WKH 09) LV WKDW LW VSHFLILFDOO\ UHPRYHV ODVHU VFDWWHU QRLVH DQG PDLQWDLQV D KLJK RSWLFDO WKURXJKSXW DW QHDUO\ DOO RWKHU ZDYHOHQJWKV $ SRWHQWLDO GUDZEDFN ZLWK WKH 09) LV WKDW WKH ZLQGRZV RI WKH FHOO FDQ EHFRPH FRDWHG ZLWK PHWDOOLF 5E ZKLFK FDQ SK\VLFDOO\ EORFN WKH OLJKW DW DOO ZDYHOHQJWKV 7KLV FDQ RFFXU DV WKH FHOO FRROV DQG 5E YDSRU WHQGV WR FRQGHQVH DW WKH VXUIDFHV RI ORZHVW WHPSHUDWXUH ,I WKH FHOO LV KHDWHG IURP WKH ZDOOV WKH ZLQGRZV EHFRPH WKH FRROHVW VXUIDFHV DQG LQGHHG 5E PHWDO ZDV REVHUYHG WR DFFXPXODWH RQ WKH ZLQGRZV GXULQJ WKH FRROLQJ SURFHVV $ VLPSOH VWXG\ ZDV SHUIRUPHG WR TXDQWLI\ WKH WUDQVPLWWDQFH RI OLJKW XVLQJ WKH +HZOHWW3DFNDUG $ VSHFWURSKRWRPHWHU 7KH WUDQVPLWWDQFH RI FHOO EHIRUH LWV ILUVW XVH ZDV b IURP QP DQG DVVXPHG WR UHPDLQ DW WKLV OHYHO DW ORQJHU ZDYHOHQJWKVf $IWHU WKH LQLWLDO XVH RI WKH FHOO LQ WKH H[SHULPHQW PDQLIHVWHG LQ )LJXUH WKH WUDQVPLWWDQFH IHOO WR b $IWHU D VHFRQG KHDWLQJ SHULRG WKH FHOO ZDV FRROHG

PAGE 139

PRUH UDSLGO\ DQG WKH b7 GURSSHG WR b ,I WKLV WUHQG FRQWLQXHG WKH FHOO ZRXOG EH XVHIXO IRU RQO\ D IHZ H[SHULPHQWV &OHDQLQJ WKH ZLQGRZV +RZHYHU D PHWKRG WR FOHDQ WKH WKH FHOO ZLQGRZV ZDV GHYHORSHG $W WHPSHUDWXUHV r& H[LVWV DV D OLTXLG DQG PXFK OLNH OLTXLG ZDWHU FRKHVLYH IRUFHV DUH JUHDWHU WKDQ WKH DGKHVLYH IRUFHV RI WKH PHWDO ZLWK WKH JODVV ,Q RWKHU ZRUGV 5E OLTXLG SUHIHUV WR IRUP RQH ODUJH JORE UDWKHU WKDQ VHYHUDO VPDOOHU JOREV LQ WKH FHOO %\ VLPSO\ UDLVLQJ WKH FHOO WHPSHUDWXUH r& ZLWK D KHDW JXQ RU LQ DQ RYHQ DQG URWDWLQJ WKH FHOO WR JDWKHU DOO WKH JOREXOHV FRQGHQVHG RQ WKH ZLQGRZV LQWR RQH PDVV WKH ZLQGRZV FDQ EH FOHDQHG 7KH WUDQVPLVVLRQ VSHFWUXP RI WKH FHOO DIWHU WKLV SURFHGXUH UHWXUQHG WR b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

PAGE 140

,W PXVW EH SRLQWHG RXW WKDW 60' LV QRW OLNHO\ WR EH UHDOL]HG ZKHQ VSHFWUDO ILOWHUV DUH XVHG ,Q WKH FDOFXODWLRQV SUHVHQWHG LQ &KDSWHU WKH RSWLFDO WUDQVPLWWDQFH XVHG ZDV EDVHG RQ WKH b7 IRU WKH TXDUW] FDSLOODU\ b7 IRU WKH PLFURVFRSH REMHFWLYH } b7 IRU WKH 09) b7 ZKHQ QHZf b7 IRU WKH IRFXVLQJ OHQV DQG b7 IRU WKH 307 ZLQGRZf 7KH RYHUDOO WUDQVPLWWDQFH IDFWRUV RI WKH FROOHFWLRQ RSWLFV DUH LPSRUWDQW WR EHLQJ DEOH WR DFKLHYH 60' EHFDXVH WKH VLJQDO LV LQ GLUHFW SURSRUWLRQ WR WKHP )LJXUH JLYHV WKH WUDQVPLWWDQFH VSHFWUD RI WKH ILYH ILOWHUV DOO IURP &RULRQf WHVWHG IRU SRVVLEOH XVH LQ WKLV H[SHULPHQW DV PHDVXUHG E\ WKH 9DUDQ VSHFWURSKRWRPHWHU $V FDQ EH LQIHUUHG IURP )LJXUH WKH /* DQG 5* ILOWHU GHVLJQDWLRQV UHIHU WR ORQJ SDVV VSHFWUDO ILOWHUV ZLWK WKH VWDWHG QXPEHU FRUUHVSRQGLQJ WR WKH ZDYHOHQJWK DW ZKLFK WKH b7 LV KDOI RI WKH PD[LPXP YDOXH 7KH RWKHU WZR ILOWHUV DUH LQWHUIHUHQFH EDQG SDVV ILOWHUV RI QP IXOO ZLGWK DW KDOI PD[LPXP EDQGZLWK ZLWK SHDN WUDQVPLWWDQFH FHQWHUHG DW WKH GHVLJQDWHG ZDYHOHQJWK 7KH ZDYHOHQJWK UDQJHV RI WKH WUDQVPLWWDQFH FDQ EH FRPSDUHG ZLWK WKH IOXRUHVFHQFH HPLVVLRQ VSHFWUD RI WKH G\HV DSSHDULQJ LQ )LJXUH LQ &KDSWHU 7DEOH OLVWV WKH bWUDQVPLWWDQFH RI WKH WRWDO HPLVVLRQ SDVVHG E\ HDFK G\HILOWHU FRPELQDWLRQ ZKLFK ZHUH GHWHUPLQHG E\ PXOWLSO\LQJ 307 VLJQDOV DW HDFK ZDYHOHQJWK RI HDFK G\H DV VKRZQ LQ )LJXUH E\ WKH WUDQVPLWWDQFHV RI WKH ILOWHUV DV VKRZQ LQ )LJXUH DQG WKHQ GLYLGLQJ WKH VXPV RI WKH UHVXOWLQJ YDOXHV IRU WKH VSHFWUD ZLWK WKH ILOWHUV E\ WKH RYHUDOO 307 VLJQDO IRU HDFK G\H REWDLQHG ZLWKRXW DFFRXQWLQJ IRU WKH ILOWHUV 7DEOH DOVR OLVWV WKH DEVRUEDQFHV RI WKH ILOWHUV DW DQG

PAGE 141

VRn VR 2 6 5* :DYHOHQJWK QPf VSHFWUD RI VSHFWUDO ILOWHUV IRU SRVVLEOH XVH LQ H[SHULPHQWV X! 2-

PAGE 142

QP ZKLFK ZHUH PHDVXUHG LQ WKH VDPH PHWKRG SUHYLRXVO\ GHVFULEHG LQ WKH H[SHULPHQWV XVHG WR PHDVXUH DEVRUEDQFHV RI WKH 09)V ZLWK WKH SKRWRGLRGH 7DEOH 7UDQVPLWWDQFHV RI WKH SRVVLEOH VSHFWUDO ILOWHUV WR EH XVHG LQ FRPELQDWLRQ ZLWK WKH PHWDO YDSRU ILOWHU )LOWHU $EVRUEDQFH b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b RI WKH SRZHU SDVVHG WKURXJK WKH FDOFLWH SRODUL]HU DQG D PLQLPXP RI b FRXOG EH REWDLQHGf $ VHFRQG SRODUL]HU 0HOOHV *ULRW ,UYLQH &$f ZDV SODFHG DIWHU

PAGE 143

6LJQDO%DFNJURXQG )LOWHU ,5 ,5 ,5 )LJXUH &DOFXODWHG VLJQDOEDFNJURXQG UDWLRV RI WKH SRO\PHWKLQH G\HV LQ PHWKDQRO DW QRUPDOL]HG FRQFHQWUDWLRQ DQG SRZHU ZLWK WKH GLIIHUHQW VSHFWUDO ILOWHUV H[FLWDWLRQ DW QPf A /O

PAGE 144

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b DW QP 5HGXFLQJ WKH QRLVH E\ D IDFWRU [ DQG ORVLQJ b RI WKH VLJQDO LV QRUPDOO\ FRQVLGHUHG D YHU\ JRRG WUDGHRII 5HVXOWV /LPLWV RI 'HWHFWLRQ 2QFH WKH V\VWHP ZDV VHW XS DQG FKDUDFWHUL]HG WKHUH UHPDLQHG RQO\ RQH RWKHU WDVN WR PHDVXUH G\H VWDQGDUGV RI NQRZQ FRQFHQWUDWLRQV DQG GHWHUPLQH WKH ORZHVW GHWHFWDEOH FRQFHQWUDWLRQ 6HYHUDO GLIIHUHQW H[SHULPHQWV ZHUH FDUULHG RXW DQG VOLJKW

PAGE 145

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f JDYH WKH ODUJHVW VLJQDO WR QRLVH UDWLR DW IL[HG ODVHU SRZHU DQG VLJQDO WR QRLVH LQFUHDVHG OLQHDUO\ ZLWK SRZHU XS WR a P: WKH PD[LPXP RSHUDEOH ODVHU SRZHU DW WKH FHOO GXH WR RSWLFDO ORVVHVf $ FDOLEUDWLRQ FXUYH IRU ,5 ZDV WDNHQ DW WKHVH FRQGLWLRQV OLVWHG LQ 7DEOH f DQG

PAGE 146

7DEOH 5HVXOWV RI WKH H[SHULPHQWV WR GHWHUPLQH OLPLWV RI GHWHFWLRQ IRU WKH DSSURDFK 3DUDPHWHU ([SHULPHQW '\H ,5 ,5 ,5 ,5 6ROYHQW 0HWKDQRO 0HWKDQRO 0HWKDQRO 0HWKDQRO 6DPSOH &RQWDLQHU FP )ORZFHOO XQ LG &DSLOODU\ [P LG &DSLOODU\ LP LG &DSLOODU\ )LOWHULQJ f6SH[ 09) 5* 09) 6 09) 6 'HWHFWRU 5 5 5 5 ;F[ QPf O P:f 6/ FPf r[ [ n [ n [ V (O :FPf r )ORZ 5DWH /PLQf )ORZ 9HORFLW\ FPVf 9S Q/f m WU PVf r UH Vf 0EL S$ E E S$ 6EO S$ E E S$ 6HQVLWLYLW\ $0 F [ F [ $0 /2' 0f [ [ [ [ /2' 9Sf m /2' UFf V0RQRFKURPDWRU VHW DW QP ZLWK QP EDQGZLGWK E&RXQWV SHU PV F&RXQWV0

PAGE 147

DQ /2' RI [ 0 ZDV REWDLQHG 7KLV FRUUHVSRQGV WR DQ DYHUDJH QXPEHU RI } PROHFXOHV LQ 9S DW WKH /2' RU DSSUR[LPDWHO\ PLOOLRQ PROHFXOHV SDVVLQJ WKURXJK 9S GXULQJ WKH PLQLPXP PHDVXUHPHQW SHULRG DW WKH /2' ([SHULPHQW ([SHULPHQW XWLOL]HG WKH DSSURDFK VKRZQ LQ )LJXUH 7KH DQDO\WH FKRVHQ IRU WKLV DQDO\VLV ZDV ,5 ZKLFK DV SUHVHQWHG HDUOLHU ZDV H[SHFWHG WR JLYH WKH ODUJHVW VLJQDO ZLWK WKH 5* ILOWHU09) FRPELQDWLRQ DW QP H[FLWDWLRQ $ VWXG\ WR RSWLPL]H IORZ UDWH DQG ODVHU SRZHU VKRZHG WKDW KLJKHU IORZ JDYH D ODUJHU VLJQDOQRLVH UDWLR DQG WKDW P: SRZHU (/ N:FPf DW WKH FDSLOODU\ \LHOGHG RSWLPXP SRZHU DW DOO IORZ UDWHV 7KH ORZ RSWLPXP SRZHU DQG LQFUHDVLQJ VLJQDO YHUVXV IORZ UDWH VXJJHVWHG WKDW WKH G\H ZDV GHFRPSRVLQJ GXULQJ LWV UHVLGHQFH WLPH 7KLV ZDV QRW QRWLFHG LQ SUHYLRXV VWXGLHV ZLWK ,5 EHFDXVH WKH V\VWHP ZLWK WKH 6SH[ LQYROYHG D ORZHU ODVHU LUUDGLDQFH m N:FPf $SSDUHQWO\ ZLWK LUUDGLDQFH RI N:FP SKRWRGHFRPSRVLWLRQ RI ,5 WDNHV SODFH LQ XQGHU PV +RZHYHU WKLV ZDV QRW YHULILHG ZLWK VXEVHTXHQW VWXGLHV ,Q DQ\ HYHQW ,5 DQDO\VLV LQ PHWKDQRO VROYHQW IORZLQJ WKURXJK D P LG FDSLOODU\ DW P/PLQ ZLWK 5 307 GHWHFWLRQ LQ SKRWRQ FRXQWLQJ LQWHUYDOV RI PV DQG 09)5* VSHFWUDO ILOWHULQJ ZDV DEOH WR REWDLQ DQ /2' RI [ n 0 PROHFXOHV RQ DYHUDJH LQ WKH S/ 9Sf 7KLV GRHV QRW VHHP WR EH DQ HVSHFLDOO\ ORZ /2' EXW ZLWK WKH GHWHFWLRQ FULWHULRQ XVHG E\ .HOOHUfV JURXSfn WKLV FRUUHVSRQGV DQ /2' RI PROHFXOHV GXULQJ WKH PV FRXQWLQJ LQWHUYDO

PAGE 148

([SHULPHQW 'XH WR WKH GHFRPSRVLWLRQ SUREOHP ZLWK ,5 LW ZDV GHFLGHG WR SHUIRUP DOO IXWXUH DQDO\VHV LQ WKH FDSLOODU\ ZLWK ,5 ,5 IOXRUHVFHG DW ZDYHOHQJWKV WRR VKRUW IRU RSWLPXP XVH ZLWK WKH ILOWHUVf $OVR SKRWRQ FRXQWLQJ GHWHFWLRQ ZLWK WKH 5 SURYHG GLIILFXOW PHQWLRQHG LQ &KDSWHU f WKHUHIRUH H[SHULPHQW XVHG WKH 5 IRU WKLV DSSOLFDWLRQ )XUWKHUPRUH LQ WKH HYHQW WKDW 5DPDQ VFDWWHU IURP WKH VROYHQW QRW WDNHQ LQWR DFFRXQW LQ )LJXUH f ZDV D PRUH VLJQLILFDQW VRXUFH RI QRLVH WKDQ DQWLFLSDWHG WKH 6 EDQGSDVV ILOWHU UHSODFHG WKH 5* ORQJ SDVV ILOWHU ,Q WKH FDVH RI ,5 WKH ORZHVW SRVVLEOH IORZ UDWH FRXOG EH VHW ZLWK OLWWOH FKDQJH LQ VLJQDO DQG WKH VDWXUDWLQJ ODVHU SRZHU FORVHO\ FRUUHVSRQGHG WR WKH P: YDOXH FDOFXODWHG LQ &KDSWHU 7KHUHIRUH WR DFKLHYH WKH PD[LPXP VLJQDO WR QRLVH ODVHU SRZHU ZDV VHW WR P: MXVW DERYH VDWXUDWLRQf DW WKH FDSLOODU\ (/ N:FPf DQG WR DFKLHYH WKH ORQJHVW WU WKH IORZ UDWH ZDV VHW WR [/PLQ 7KH /2' XQGHU WKHVH FRQGLWLRQV ZDV [ n 0 ,5 PROHFXOHV LQ S/f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

PAGE 149

RI WKH GHSHQGHQFH RI WKH VLJQDO WR QRLVH UDWLR RQ ODVHU SRZHU DW [ n FP IRFXV DUHDf RI IORZLQJ Q0 ,5 VROXWLRQ LQ PHWKDQRO $V H[KLELWHG WKH ORQJHU WLPH FRQVWDQW WFf RI V VHW E\ D +] ORZ SDVV ILOWHUf JLYHV D KLJKHU VLJQDO WR QRLVH UDWLR WKDQ WKH PV WLPH FRQVWDQW RI WKH FKDUW UHFRUGHU )RU WKLV UHDVRQ WKH ORQJHU WLPH FRQVWDQW ZDV XVHG LQ WKH DQDO\VLV $V VKRZQ LQ 7DEOH WKH FRQFHQWUDWLRQDO OLPLW RI GHWHFWLRQ ZDV [ n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b RI WKH ODVHU OLJKW ZDV VFDWWHUHG RYHUDOO %\ REVHUYLQJ WKH VFDWWHU ZLWK D QHDULQIUDUHG YLHZHU LW DSSHDUHG WKDW WKH VFDWWHU ZDV

PAGE 150

L @ V WLPH FRQVWDQW /DVHU 3RZHU P:f nIHHW RI ODVHU SRZHU IRFXVHG WR [ n FP DUHDf RQ WKH VLJQDOQRLVH UDWLR RI D Q0 ,5 OXWLRQ LQ PHWKDQRO IORZLQJ LQ WKH XQ LG FDSLOODU\ DW L/PLQ DW UF DQG V 7KLV VWXG\ ZDV cUIRUPHG LQ WKH SRZHUIORZ RSWLPL]DWLRQ IRU H[SHULPHQW r 1!

PAGE 151

6LJQDO Q$f )LJXUH &DOLEUDWLRQ FXUYH IRU H[SHULPHQW &RQGLWLRQV SUHVHQWHG LQ 7DEOH X!

PAGE 152

7DEOH ,QVWUXPHQWDO QRLVHV RI WKH DSSURDFK WR VLQJOH PROHFXOH GHWHFWLRQ IRU WKH H[SHULPHQWV OLVWHG LQ 7DEOH 3DUDPHWHU ([SHULPHQW 'HWHFWRU 5 5 5 5 6LJQDO &ROOHFWLRQ $QDORJ 3KRWRQ &RXQWLQJ 3KRWRQ &RXQWLQJ $QDORJ 3UHDPSOLILHU S$ P9 P9 S$ 307 LQ 'DUN s S$ f s f s S$ %ODQN S$ f s r s s S$ /DVHU 3RZHU P:f s s s s f&RXQWV SHU PV PRVW LQWHQVH QHDU WKH WUDQVPLWWHG EHDP EXW WKH OLJKW ZDV VWLOO YHU\ LQWHQVH LQ D r ULQJ DURXQG WKH FDSLOODU\ $VVXPLQJ WKDW WKH OLJKW LV VFDWWHUHG XQLIRUPO\ QRW D YDOLG DVVXPSWLRQf WKH VFDWWHU FROOHFWHG E\ WKH PLFURVFRSH REMHFWLYH ZRXOG EH SKRWHOHFWURQV EDVHG RQ WKH b VROLG DQJOH RI FROOHFWLRQ RI WKH P: VFDWWHUHG E\ WKH FDSLOODU\ -XGJLQJ E\ WKH ODUJH DPRXQW RI VFDWWHU SURGXFHG E\ WKH FDSLOODU\ LW LV QRW GLIILFXOW WR UHDOL]H ZK\ VFDWWHU ZDV VXFK D IRUPLGDEOH SUREOHP LQ WKH UHDO VLWXDWLRQ %DVHG RQ H[SHULPHQWDO HYLGHQFH RI WKH 09) SUHVHQWHG HDUOLHU WKH VFDWWHU PXVW KDYH SDVVHG DURXQG WKH 09) DQG UHDFKHG WKH GHWHFWRU :D\V WR DYRLG WKLV SUREOHP DUH SUHVHQWHG LQ &KDSWHU

PAGE 153

'LVFXVVLRQ RI /LPLWV RI 'HWHFWLRQ 0HWKRG RI /2' FDOFXODWLRQ 7R GHWHUPLQH WKH /2'V IRU WKH H[SHULPHQWV OLVWHG LQ 7DEOH WKH ,83$& DSSURDFK JLYHQ LQ &KDSWHU ZDV IROORZHG 7KH QRLVH ZDV GHWHUPLQHG E\ WDNLQJ WKH DYHUDJH DQG VWDQGDUG GHYLDWLRQ RI VHSDUDWH PHDVXUHPHQWV SKRWRQ FRXQWLQJf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f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

PAGE 154

&RPSDULVRQ RI /2'V ,Q DQ DFWXDO DQDO\VLV WKH PHWKRG RI GDWD FROOHFWLRQ GHSHQGV RQ WKH W\SH RI DQDO\VLV SHUIRUPHG 7KH /2' GHWHUPLQHG ZLWK WKH PRQRFKURPDWRU ZLWK DQ /2' RI I0 LQ WKH P/ IORZFHOOf ZRXOG EH WKH EHVW DSSURDFK LI D VDPSOH RI ORZ FRQFHQWUDWLRQ DQG ODUJH YROXPH UHTXLUHG DQDO\VLV IRU RQH RI WKH SRO\PHWKLQH G\HV +RZHYHU WKLV PHWKRG ZRXOG QRW EH XVHIXO DV D PHWKRG RI GHWHFWLRQ IRU WUDQVLHQW VLJQDOV LQ DQDO\WLFDO VHSDUDWLRQV RU DQDO\VLV RI OLPLWHG VDPSOH YROXPHV ,Q WKHVH FDVHV WKH EHVW PHWKRG RI DSSURDFK ZRXOG EH WR SHUIRUP WKH DQDO\VLV LQ WKH FDSLOODU\ RU RWKHU VPDOO YROXPH IORZFHOOf ,Q 60' H[SHULPHQWV RI D VWDWLF VDPSOH ORQJHU PHDVXUHPHQW WLPHV DUH DGYDQWDJHRXV SURYLGHG WKH DQDO\WH GRHV QRW GHFRPSRVH EXW IRU 60' LQ D IORZLQJ VROXWLRQ ORQJHU PHDVXUHPHQW WLPHV DUH D OX[XU\ WKDW FDQQRW EH DIIRUGHG 7KHUHIRUH GHVSLWH WKH ORZHU FRQFHQWUDWLRQDO /2' REWDLQHG ZLWK ORQJHU UF WKH EHVW DSSURDFK IRU IXWXUH DQDO\VHV LV WR NHHS WKH IDVW PHDVXUHPHQW WLPH DQG ZRUN WR ORZHU WKH /2' WKURXJK UHGXFWLRQ RI DFWXDO QRLVH DQG QRW WKURXJK WKH HOHFWURQLF UHGXFWLRQ RI QRLVH DIWHU GDWD FROOHFWLRQ

PAGE 155

&+$37(5 &21&/86,216 $1' )8785( :25. &RQFOXVLRQV 7KLV SURMHFW EHJDQ FRQILGHQWO\ DV D QRYHO PHWKRG WR GHWHFW VLQJOH PROHFXOHV LQ D IORZLQJ VROXWLRQ FRQWDLQHG LQ D FDSLOODU\ ,W ZDV VWURQJO\ EHOLHYHG WKDW 60' ZRXOG EH SRVVLEOH XVLQJ WKH V\VWHP GHVFULEHG LQ WKLV GLVVHUWDWLRQ EDVHG RQ WKH UHSRUWV RI QHDU60' E\ RWKHU UHVHDUFKHUV XVLQJ DSSURDFKHV WKDW GR QRW UHDFK WKH LQWULQVLF QRLVH OHYHOnA 0RVW QRWDEO\ WKH RQJRLQJ ZRUN DW /RV $ODPRV DV FORVH DV LW LV WR GHWHFWLQJ VLQJOH PROHFXOHV LQ WKH SUREH YROXPH LV IDU IURP EHLQJ DEOH WR DFKLHYH VXFK UHVXOWV ZLWK b VDPSOLQJ HIILFLHQF\ %DVHG RQ WKHVH OLPLWDWLRQV .HOOHUf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

PAGE 156

SUREOHP ZLWK JUHDWO\ LQFUHDVHG VFDWWHU 7KHUHIRUH LW ZDV GHFLGHG WKDW XVLQJ D VDPSOH FRQWDLQHU RWKHU WKDQ WKH FDSLOODU\ LV QRW DQ RSWLRQ IRU WKLV SURMHFW ,W ZDV UHDOL]HG ZKHQ WKHVH WZR GHFLVLRQV ZHUH PDGH b VDPSOLQJ HIILFLHQF\ DQG GHWHFWLRQ LQ D FDSLOODU\f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

PAGE 157

WKDQ WKH GHWHFWLRQ OLPLWV REWDLQHG E\ RWKHU UHVHDUFKHUV DWWHPSWLQJ 60' 2I FRXUVH WKH\ ZHUH QRW SHUIRUPLQJ GHWHFWLRQ LQ D FDSLOODU\ DQG RQO\ WZR RWKHU NQRZQ UHSRUWHG /2'V LQ D FDSLOODU\ DUH ORZHU WKDQ WKLV &RPSDULVRQ ZLWK .HOOHUfV ZRUN ,Q WKH SURMHFW SUHVHQWHG KHUH D UHODWLYHO\ VKRUW SHULRG RI WLPH ZDV VSHQW ZLWK RQH JUDGXDWH VWXGHQW GRLQJ WKH JUHDW PDMRULW\ RI WKH ZRUN 'XULQJ HLJKW \HDUV RI HIIRUW DW /RV $ODPRV 1DWLRQDO /DERUDWRU\ .HOOHUfV WHDP RI VFLHQWLVWV KDYH LPSURYHG WKHLU GHWHFWLRQ OLPLW IURP PROHFXOHV ZLWK b VDPSOLQJ HIILFLHQF\ WR PROHFXOHV ZLWK b VDPSOLQJ HIILFLHQF\ 5HI LQYROYHG D GLIIHUHQW DSSURDFK DQG DV LQ 5HI WKH FRQFHQWUDWLRQ /2' ZDV QRW UHSRUWHGf :LWK WKLV LQ PLQG LW LV EHOLHYHG WKDW WKH FRQWLQXRXVZDYH $U ODVHU DSSURDFK RI .HOOHU KDV OLWWOH FKDQFH RI DFKLHYLQJ WUXH 60' 2Q WKH RWKHU KDQG LQ WKH LQLWLDO VWDJHV ZLWK WKH WHFKQLTXH UHSRUWHG LQ WKLV GLVVHUWDWLRQ D GHWHFWLRQ OLPLW RI ,5 PROHFXOHV EDVHG RQ WKH VDPH FULWHULD DV DERYHf ZLWK b VDPSOLQJ HIILFLHQF\ ZDV REWDLQHG $JDLQ LW PXVW EH VWUHVVHG WKDW WKH OLPLWLQJ QRLVH RI WKH H[SHULPHQW ODVHU VFDWWHU LV RUGHUV RI PDJQLWXGH KLJKHU IURP D FDSLOODU\ WKDQ IURP WKH VKHDWK IORZ FXYHWWH :LWK WKH PRGLILFDWLRQV WR WKLV DSSURDFK SUHVHQWHG EHORZ EDUULQJ XQIRUHVHHQ LPSHGLPHQWV LW VKRXOG EH SRVVLEOH WR DFKLHYH VLQJOH PROHFXOH GHWHFWLRQ LQ D IORZLQJ VDPSOH FRQWDLQHG LQ D FDSLOODU\

PAGE 158

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f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fV DSSURDFKf RU HOVH VDPSOLQJ HIILFLHQF\ LV FRPSURPLVHG )XUWKHUPRUH WKH 09) ZRXOG KDYH WR EH ILWWHG LQ WKH

PAGE 159

)LJXUH 6XJJHVWHG GHVLJQ RI WKH OLJKW SDVVDJH EHWZHHQ WKH PLFURVFRSH REMHFWLYH DQG SKRWRPXOWLSOLHU WXEH IRU LQ IXWXUH H[SHULPHQWV WR DFKLHYH VLQJOH PROHFXOH GHWHFWLRQ

PAGE 160

WXEH VR WKDW QR OLJKW ZRXOG SDVV EHWZHQ LW DQG WKH WXEH ZDOO ,Q WKLV GHVLJQ WKH KHDWLQJ WDSH RU VPDOO WXEH IXUQDFHf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r& WKHRUHWLFDO DEVRUEDQFH RI FHOO f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

PAGE 161

GHWHFWHGf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f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

PAGE 162

YLHZ RI WKH PLFURVFRSH REMHFWLYH LV DSSUR[LPDWHO\ PP DW WKH IRFXV FRQVLGHULQJ WKDW WKH VL]H RI WKH VLJQDO LQ WKH FDSLOODU\ LV D UHFWDQJOH PP ZLGH DQG PP ORQJ PRVW RI WKH FROOHFWHG UHJLRQ LV QRW XVHIXO LQIRUPDWLRQ $ VSDWLDO ILOWHU VLPSO\ D UHFWDQJOH ZLWK WKH GLPHQVLRQV SURSRUWLRQDO WR WKH 9Sf FRXOG EH SODFHG EHWZHHQ WKH WKH FDSLOODU\ DQG PLFURVFRSH REMHFWLYH ,Q WKLV ZD\ WKH VFDWWHU FRPLQJ IURP WKH RXWHU ZDOOV RI WKH FDSLOODU\ DQG RWKHU H[WHUQDO VRXUFHV LQ WKH ER[ ZRXOG EH EORFNHG EHIRUH HQWHULQJ WKH PLFURVFRSH REMHFWLYH 2WKHU 60' DSSURDFKHV XVH WKH VSDWLDO ILOWHUnAf ZKLFK W\SLFDOO\ LV SODFHG DIWHU WKH PLFURVFRSH REMHFWLYH DQG GHILQHV WKH VL]H RI WKH 9S 7R PDLQWDLQ b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f PXVW QRW DEVRUE OLJKW DW WKH ODVHU ZDYHOHQJWK QRU DW ZDYHOHQJWKV RI WKH IOXRUHVFHQFH HPLVVLRQ $GGLWLRQDOO\ WKH LPPHUVLRQ FHOO LV GLIILFXOW WR GHVLJQ

PAGE 163

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f ZRXOG RQO\ DOORZ SRLQW SHU PROHFXOH DW WKH PV PD[LPXP UHVLGHQFH WLPH RI D VLQJOH PROHFXOH LQ WKH SUREH YROXPH DW WKH PLQLPXP SXPS VHWWLQJ LQFUHDVLQJ WU LV DQRWKHU RSWLRQ GLVFXVVHG EHORZf )XUWKHUPRUH WKH QRLVH RI WKH SUHDPSOLILHU XVHG LQ FRQMXQFWLRQ ZLWK WKH SKRWRQ FRXQWHU LV QRW LGHDO IRU XVH ZLWK WKH 5 307 7KH

PAGE 164

5 ZDV IRXQG WR EH WKH EHVW SKRWRPXOWLSOLHU WXEH IRU WKH SURMHFW EDVHG RQ LWV QRLVH HTXLYDOHQW SRZHU RI I: YHUVXV I: IRU WKH 5f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f WR DOORZ IRU WKH SURGXFWLRQ RI PRUH IOXRUHVFHQFH DQG ORQJHU FRXQWLQJ WLPHV EXW GXH WR GHFRPSRVLWLRQ WKH UHVLGHQFH WLPH VKRXOG QRW EH W\SLFDOO\ PRUH WKDQ PV DW QHDUn VDWXUDWLRQ ODVHU LUUDGLDQFH 6\ULQJH SXPSV RU FDSLOODU\ HOHFWURSKRUHVLV DUH WKH EHVW PHWKRGV RI VROYHQW IORZ DEOH WR DFKLHYH WKH IORZ UDWHV QHFHVVDU\ IRU D WU PV LQ D QDUURZ FDSLOODU\ ,Q WKLV FDVH GDWD SRLQWV RQ DYHUDJH ZRXOG EH FROOHFWHG SHU PROHFXOH ZLWK WKH SKRWRQ FRXQWHU VHW DW PV FRXQWLQJ SHULRGV DQG PV PLQLPXP GZHOO WLPHf ,GHDOO\ LW ZRXOG EH EHQHILFLDO WR DSSO\ RQH RI WKH WZR PHWKRGV RI SXPSLQJ DQG XVH D EHWWHU SKRWRQ FRXQWLQJ V\VWHP /DVHU IRFXVLQJ 7KH PDQQHU RI IRFXVLQJ LQ WKLV VHWXS WKRXJK IXQFWLRQDO ZDV QRW XVHU IULHQGO\ 7KH VHYHUDO GLUHFWLRQV RI PRYHPHQW QHFHVVDU\ WR SRVLWLRQ WKH ODVHU DQG FDSLOODU\ ZKHQ WKH PLFURVFRSH REMHFWLYH DQG GHWHFWRU ZHUH IL[HG PDNH RSWLPXP IRFXVLQJ DUGXRXV &RQYHUVHO\ LI WKH WKH ODVHU ZDV IL[HG LQ VSDFH DQG WKH FDSLOODU\ SODFHG DW WKH RSWLPXP IRFXV LW ZRXOG EH HDVLHU WR SRVLWLRQ WKH PLFURVFRSH REMHFWLYH

PAGE 165

DQG GHWHFWRU VSDFHG DSDUW VR WKH IRFXVHG OLJKW FROOHFWHG E\ WKH REMHFWLYH PDWFKHV WKH VL]H RI WKH SKRWRFDWKRGHf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f WR PLQLPL]H RVFLOODWLRQV DQG WKH DVVRFLDWHG QRLVH $U ODVHU SRZHU IOXFWXDWLRQ $ ILQDO VRXUFH RI QRLVH WKDW VKRXOG EH PLQLPL]HG LI SRVVLEOH LV D ORQJWHUP SRZHU GULIW LQ WKH $U ODVHU WKH RULJLQ RI ZKLFK LV XQNQRZQ 7KH IUHTXHQF\ RI WKH F\FOLQJ SRZHU IOXFWXDWLRQV LV PLQXWHV DQG WKH RSHUDWLRQ RI SRVVLEOH VRXUFHV RI WKLV GULIW UHODWHG WR WKH LQVWUXPHQWDWLRQ DW WKH VHWXS KHDW H[FKDQJHUVf ZDV IRXQG QRW WR EH FRUUHODWHG ZLWK WKH ODVHU IOXFWXDWLRQV 7KLV QRLVH LV SRVVLEO\ FDXVHG E\ WKH DLU KDQGOLQJ V\VWHP RI WKH EXLOGLQJ ,W LV FRQFHLYDEOH WKDW WKH WKHUPDO H[SDQVLRQ DQG FRQWUDFWLRQ RI WKH RSWLFV DQG WDEOH GHIRFXVHG WKH ODVHU VOLJKWO\ DQG SURGXFHG SRZHU GULIW 2SHUDWLQJ WKH ODVHU LQ WKH FRQVWDQW SRZHU PRGH

PAGE 166

VKRXOG KDYH FRUUHFWHG WKLV SUREOHP EXW WKH IOXFWXDWLRQ ZDV VWLOO DV JUHDW DV b $OWKRXJK WKH GULIW ZDV QRW FDXVH IRU JUHDW FRQFHUQ LQ WKH H[SHULPHQWV SUHVHQWHG LQ WKLV GLVVHUWDWLRQ ODVHU SRZHU VWDELOLW\ LV D VRXUFH RI LQWULQVLF QRLVH WKDW PD\ EHFRPH YHU\ LPSRUWDQW LI WKH SUREOHP ZLWK ODVHU VFDWWHU LV VROYHG %ODFN ERG\ UDGLDWLRQ $ ILQDO IDFWRU WKDW ZDV GLVFHUQHG DV D SRWHQWLDO SUREOHP LQ H[SHULPHQWV ZKHQ WKH 09) ZDV KHDWHG WR KLJK WHPSHUDWXUHV m r&f LV EODFN ERG\ UDGLDWLRQ 7KH UHGVHQVLWLYH 307 ZDV DEOH WR GHWHFW WKH QHDULQIUDUHG HPLVVLRQ RI PDWWHU ZKHQ ORQJ SDVV ILOWHU ZDV XVHG 7KLV ZDV QRW VLJQLILFDQW DW 09) WHPSHUDWXUHV r& RU ZKHQ DQ LQWHUIHUHQFH ILOWHU ZKLFK DEVRUEHG UDGLDWLRQ QP ZDV SODFHG EHIRUH WKH 307 ,Q IXWXUH H[SHULPHQWV EODFN ERG\ UDGLDWLRQ ZRXOG QRW EH D VLJQLILFDQW IDFWRU ZLWK WKH XVH RI D EDQGSDVV ILOWHU ZKLFK ZRXOG DOVR EORFN WKH PHWKDQRO 5DPDQ EDQG DW QP ZLWK QP H[FLWDWLRQf DQG D ORQJHU 09) GLVFXVVHG SUHYLRXVO\ WKDW UHTXLUHV ORZHU WHPSHUDWXUH WR DFKLHYH WKH QHFHVVDU\ DEVRUEDQFH WR DFKLHYH 60'

PAGE 167

5()(5(1&( /,67 -+ -HWW 5$ .HOOHU -& 0DUWLQ %/ 0DUURQH 5. 0R\]LV 5/ 5DWOLII 1. 6HLW]LQJHU (% 6KHUD DQG && 6WHZDUW %LRPRO 6WUXFW '\QDP f 1'RYLFKL -& 0DUWLQ -+ -HWW 0 7UNXOD DQG 5$ .HOOHU $QDO &KHP f '& 1JX\HQ 5$ .HOOHU DQG 0 7UNXOD 2SW 6RF $P % f (% 6KHUD 1. 6HLW]LQJHU /0 'DYLV 5$ .HOOHU DQG 6$ 6RSHU &KHP 3K\V /HWW f -+ +DKQ 6$ 6RSHU +/ 1XWWHU -& 0DUWLQ -+ -HWW DQG 5$ .HOOHU $SSO 6SHFWURVF f 6$ 6RSHU (% 6KHUD -& 0DUWLQ -+ -HWW -+ +DKQ + / 1XWWHU DQG 5$ .HOOHU $QDO &KHP f -0 YDQ (PRQ (G ,PPXQRFKHPLFDO $VVD\V IRU (QYLURQPHQWDO $QDO\VLV $PHULFDQ &KHPLFDO 6RFLHW\ :DVKLQJWRQ '& +: 6KDSLUR 3UDFWLFDO )ORZ &\WRPHWU\ /HVV 1HZ
PAGE 168

3(OYLQJ 9 .ULYDQ DQG ,0 .ROWKRII (GV 1XFOHDU $FWLYDWLRQ DQG 5DGLRLVRWRSLF 0HWKRGV RI $QDO\VLV 9RO 3DUW 7UHDWLVH RQ $QDO\WLFDO &KHPLVWU\ QG (G -RKQ :LOH\ 1HZ
PAGE 169

&/ 6WHYHQVRQ DQG -' :LQHIRUGQHU $SSO 6SHFWURVF f &/ 6WHYHQVRQ DQG -' :LQHIRUGQHU $SSO 6SHFWURVF LQ SUHVV f &/ 6WHYHQVRQ DQG -' :LQHIRUGQHU $SSO 6SHFWURVF LQ SUHVV f '' (EELQJ *HQHUDO &KHPLVWU\ QG (G +RXJKWRQ 0LIIOLQ %RVWRQ S ,83$& 3XUH $SSO &KHP f + .DLVHU = $QDO &KHP f 7UDQVODWHG E\ $& 0HQ]LHV LQ 7KH /LPLW RI 'HWHFWLRQ RI D &RPSOHWH $QDO\WLFDO 3URFHGXUH +RIQHU 1HZ
PAGE 170

*, %HNRY 96 /HWRNKRY 2, 0DWYHHY DQG 9, 0LVKLQ 6RY 3K\V -(73 f *6 +XUVW 0+ 1D\IHK DQG -3
PAGE 171

5 ,QGUDOLQJDP (YDOXDWLRQ RI 1DUURZ /LQH $WRPLF )LOWHUV DQG WKH 'HYHORSPHQW RI D 0HWDO 9DSRU )LOWHU IRU 5DPDQ 6SHFWURVFRS\ 3K' 'LVVHUWDWLRQ 8QLYHUVLW\ RI )ORULGD 5 ,QGUDOLQJDP -% 6LPHRQVVRQ *$ 3HWUXFFL %: 6PLWK DQG -' :LQHIRUGQHU $QDO &KHP LQ SUHVV f 3DWRQD\ DQG 0' $QWRLQH $QDO &KHP $ f 5& %HQVRQ DQG +$ .XHV &KHP (QJ 'DWD f $QGUHZV:LOEHUIRUFH DQG 3DWRQD\ $SSO 6SHFWURVF f 2* %MRUQVVRQ 50XUSK\ DQG 96 &KDGZLFN ([SHULHQWLD f -3 :HEE )* :HEVWHU DQG %( 3ORXUGH ,((( 4XDQWXP (OHFWURQ f 05 7ULSS &5 6ZD\]H DQG )R[ LQ '\H &XUYHV '$ %ORRPILHOG (G 8QLYHUVLW\ 3DUN 3UHVV %DOWLPRUH &K 5& %HQVRQ DQG +$ .XHV 3K\V 0HG %LRO f 0$ 5REHUVRQ $QGUHZV:LOEHUIRUFH '& 1RUULV DQG 3DWRQD\ $QDO /HWW f $QGUHZV:LOEHUIRUFH DQG 3DWRQD\ 6SHFWURFKLP $FWD $ f 3/ 5DSSRSRUW DQG -7KLHVVHQ 3KDUP 6FL f % +ROOLQV % 1RH DQG -0 +HQGHUVRQ &OLQ &KHP f 0% 'RUU DQG *0 3ROODFN 3KDUP 6F/ f 6/HKRWD\ $0 3OHVV $QG -' :LQHIRUGQHU $QDO 6F/ LQ SUHVV f (6
PAGE 172

(+ 7ULQK DQG &+VX $FRXVW 6RF $P f $ $VKNLQ 3K\V 5HY /HWW f 6 $UQROG DQG 1 +HVVHO 5HY 6FL ,QVWUXP f 5 :LOOLDPV DQG -5 0DWH\ $SSO 3K\V /HWW f -: /\RQV DQG /5 )DXONQHU $QDO &KHP f <) &KHQJ 6 :X '< &KHQ DQG 1'RYLFKL $QDO &KHP f + 6ZHUGORZ -= =KDQJ '< &KHQ +5 +DUNH 5 *UH\ 6 :X 1'RYLFKL DQG & )XOOHU $QDO &KHP f < .XURVX 7 6DVDNL DQG 0 6DLWR +LJK 5HV &KURPDWRJU f $( %UXQR % .UDLWWLJHU ) 0D\VWUH DQG +0 :LGPHU $QDO &KHP f $( %UXQR ( *DVVPDQQ 1 3HULFOHV DQG $QWRQ $QDO &KHP f '5 6DJH DQG 62 )DUZHOO $QDO &KHP f '& 2f6KHD (OHPHQWV RI 0RGHP 2SWLFDO 'HVLJQ -RKQ :LOH\ t 6RQV 1HZ
PAGE 173

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

PAGE 174

, 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\ c-DPHV :LQIRUGQHU &KDLU *UDGXDWH 5HVHDUFK 3URIHVVRU RI &KHPLVWU\ 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\ Mt-/$$ 5LFKDUG $
PAGE 175

7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH 'HSDUWPHQW RI &KHPLVWU\ LQ WKH &ROOHJH RI /LEHUDO $UWV DQG 6FLHQFHV DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 0D\ 'HDQ *UDGXDWH 6FKRRO

PAGE 176

81,9(56,7< 2) )/25,'$


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 EPXIAJ7EP_JWXDO0 INGEST_TIME 2011-07-29T20:01:13Z PACKAGE AA00003275_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES