Citation
Surface enhanced raman spectroscopy on metallic colloids for the purpose of trace analysis

Material Information

Title:
Surface enhanced raman spectroscopy on metallic colloids for the purpose of trace analysis
Creator:
Angebranndt, Martin John, 1964-
Publication Date:
Language:
English
Physical Description:
ix, 138 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Chemicals ( jstor )
Colloids ( jstor )
Electrodes ( jstor )
Lasers ( jstor )
Molecules ( jstor )
Pyridines ( jstor )
Raman scattering ( jstor )
Signals ( jstor )
Silver ( jstor )
Spheroids ( jstor )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 132-137).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Martin John Angebranndt.

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:
026492261 ( ALEPH )
AJA5453 ( NOTIS )
25222112 ( OCLC )

Downloads

This item has the following downloads:


Full Text
















SURFACE ENHANCED RAMAN SPECTROSCOPY ON METALLIC COLLOIDS
FOR THE PURPOSE OF TRACE ANALYSIS














BY

MARTIN JOHN ANGEBRANNDT


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


1991














ACKNOWLEDGEMENTS

I would first like to thank Dr. James D. Winefordner for

his guidance and support. I am especially grateful for the

considerable freedom he allowed me, while keeping me from

straying too far off course. I am also grateful to my

colleagues in Jim's laboratory and throughout the department

for all of the good and bad times we shared. I hope

Washington D.C. is not too far out of the way. Another thank

you must go to Tony Antonazzi, who operated the photon

correlation spectrometer for the experiments on the copper

colloids.

Considerable recognition must also go to Frangois Lang,

Bob Wallace, Ward Campbell, Dr. Robert Ho, and John

Glendenning, my Aikido Senseis, as well as the other Aikidoka.

They managed to keep my physical and spiritual development

near the mental and technical levels of my academic pursuits.

Finally, I am grateful to my family. My parents and

sister, who helped and supported me when I needed it most and

my new wife, who helped me keep a smile during one of the most

trying periods in my life.
















TABLE OF CONTENTS


Pace


ACKNOWLEDGEMENTS .

LIST OF TABLES .

LIST OF FIGURES .


ABSTRACT

Chapter


INTRODUCTION .


OVERVIEW OF PREVIOUS WORK .

SERS on Electrochemical Surfaces
Analytical Application .
Physical Studies .
SERS on Colloidal Metals .
Analytical Applications .
Physical Studies .
SERS on Chromatographic Surfaces

THEORETICAL FOUNDATIONS .


. 20


The Raman Scattering Process .
The SERS Enhancement Process .
Electromagnetic Enhancement Theory .
Chemical Enhancement Theories .

COLLOIDS AND AGGREGATION .

SILVER COLLOIDS OF SULFA DRUG ANALYSIS

Introduction .
Experimental .
Apparatus .
Chemicals and Procedure .
Results and Discussion .


iii


rrii


. vi


viii


42

50









6 A STUDY OF COPPER COLLOIDS 68

Introduction 68
Experimental 70
Apparatus .. 70
Chemicals and Procedure .. 73
Results and Discussion .. 73

7 COPPER COLLOIDS IN A FLOWING STREAM 95

Introduction ... 95
Experimental ... 97
Apparatus .. 97
Results and Discussion .. 100

8 CONCLUSIONS AND SPECULATIONS ON THE FUTURE 111

Conclusions . 111
Speculations 113
Experiments on Filters ... 113
Novel Colloidal Surfaces .. 115
Methodology .. 115

APPENDIX ELECTROMAGNETIC DERIVATIONS .. 117

Molecule Near a Hemispheroid Protruding
From a Surface .. 117
Full Prolate Spheroid .... 126
The Enhancement Factor ... 128

REFERENCES . 132

BIOGRAPHICAL SKETCH . 138
















LIST OF TABLES


Page

5.1 Intensities and Positions of Raman Bands
for the Sulfa Drugs as KBr Pellets 57

5.2 Intensities and Positions of SERS Bands
of Sulfa Drugs on Silver Colloids ...... 60

5.3 Analytical Figures of Merit for Sulfa
Drugs on Silver Colloidal System. ... 66

6.1 Analytical Figures of Merit for Analytes on Static
Copper Colloid .. 93

6.2 Comparison of Static p-aminobenzoic acid Results
with Literature .. 94

7.1 Effect of Acidity on SERS Signals and Colloidal
Aggregation .. 102

7.2 Comparison of p-aminobenzoic acid FIA Data with
Literature .. 109

A.1 List of Symbols used in the Appendix 131
















LIST OF FIGURES


Page

3.1 A Schematic Diagram showing the Stokes and
Anti-Stokes Shifted Raman Scattering Relative to
the Rayleigh Scatter and their Origin 22

3.2 The Real and Imaginary Components of the Dielectric
Functions of Copper and Silver. .. 29

3.3 A Schematic Diagram Illustrating the Difference
between the Normal Raman and the SERS Processes 31

3.4 A Plot Showing the Relative SERS Enhancement for
Silver Spheroids as a Function of Polar Angle 34

3.5 A Plot Showing the Effect of Separation Distance on
the SERS Enhancement for Silver Spheroids 36

4.1 Calculated Extinction Profiles for 20 nm spheres of
Copper and Silver in Water .. 44

5.1 Normal Raman Spectra of the Sulfa Drugs as
KBr Pellets . .55

5.2 SERS Spectra of the Sulfa Drugs on
Silver Colloids .. 58

5.3 The Structures of the Sulfa Drugs .. 62

5.4 The SERS Spectrum of Sulfamethazine at 1, 10,
and 100 ppb Levels 64

6.1 A Schematic Diagram of the Experimental Layout of
the Optical Components Used in the Copper Colloid
Studies . ... .71

6.2 Normal Raman Scattering for Pyridine, SERS Spectrum
of 10 ppm Pyridine on a Copper Colloid, and the
Background of the Copper Colloid ... .75

6.3 Relative SERS Intensity of the Most Intense Band of
the Analyte as a Function of pH .. 77








6.4 Absorption Scans of the Copper Colloid as it
Aggregates over Time with PABA 80

6.5 A Photon Correlation Spectrum of the
Copper Colloid ....... .82

6.6 Intensities of the Two Most Intense Bands of PABA as
a Function of Time .............. .84

6.7 SERS Spectrum of 50 ppm PABA on a Copper Colloid
and the Background of the Copper Colloid ... 86

6.8 SERS Spectrum of 50 ppm PNBA on a Copper Colloid
and the Background of the Copper Colloid .. 88

6.9 SERS Signal as a Function of Analyte Concentration
for Pyridine, PABA, and PNBA .. 90

7.1 Schematic Diagram of the FIA Setup used to Produce
the Copper Colloids. 98

7.2 The SERS Spectrum of PABA of a FIA Produced Copper
Colloid . 103

7.3 The SERS-FIA Signals at 1397 cm'1 from a Series of
200 microgram injections of PABA into the System 105

7.4 SERS Signal as a Function of Amount of
PABA Injected .. .. 107

A.1 a). A Cross-section of a Prolate Spheroid
Illustrating the Elliptical Coordinate System.
b). An Illustration of the Model used for the
Derivation in the Appendix ... 119


vii


















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

SURFACE ENHANCED RAMAN SPECTROSCOPY ON METALLIC COLLOIDS
FOR THE PURPOSE OF TRACE ANALYSIS


BY

MARTIN JOHN ANGEBRANNDT

August, 1991



Chairman: Dr. James D. Winefordner
Major Department: Chemistry



The use of Surface Enhanced Raman Spectroscopy (SERS) on

silver and copper colloidal surfaces is discussed.

Qualitative and quantitative results in the determination of

three sulfapyrimidine drugs on silver surfaces are presented.

It is demonstrated that it is possible to differentiate

between these drugs at trace levels. Limits of detection in

the low ppb range were obtained with a linear dynamic range

extending for approximately two orders of magnitude above this

for the 1114 cm1 band.

The analytical figures of merit of copper colloidal

surfaces are evaluated for the first time. These copper


viii









colloids are produced without the addition of protecting

agents. The only previous preparation used citrate as an

antiaggregant. These colloids are less stable than silver

colloids, but will have fewer interference from fluorescent

solution species. Analytical results for p-aminobenzoic acid,

p-nitrobenzoic acid, and pyridine adsorbed onto a batch

produced colloid are presented. The detection limits are in

the low ppm range. This is comparable with previous results

for silver colloidal surfaces.

A Flow Injection Analysis (FIA) system was constructed to

evaluate the possibility of adapting copper colloid based-SERS

to on-line analysis. It was possible to continuously produce

the colloid and analyze a sample injected into the flowing

stream. Para-aminobenzoic acid was used as the test analyte.

The detection limit was found to be 150 gg. The major

obstacle in this analysis is the hydrogen gas produced as a

byproduct of the reduction. The gas forms bubbles in the line

and must be removed prior to the spectroscopic measurement.

The method used to remove the hydrogen gas is just adequate,

and an improvement in this aspect of the analysis would

increase the reproducibility significantly.














CHAPTER 1

INTRODUCTION

Raman Spectroscopy provides information which cannot be

obtained easily by other means [1]. It allows the researcher

to probe the symmetric vibrations of a molecule. This

information is complementary to that gleaned using infrared

spectroscopy. The primary reason that Raman spectroscopy can

be such a powerful tool for analytical analysis is that all

molecules possess some vibrations which are Raman active.

Therefore, this method can be applied when the various

luminescence spectroscopies cannot. The Raman spectrum also

contains more information than a luminescence spectrum, enough

to identify the molecule.

The major disadvantage of Raman spectroscopy is the low

quantum efficiency of the process [2]. Since only a very few,

approximately 1 in 10+6, of the photons are Raman scattered,

an intense excitation source is required to observe a signal

within a reasonable period of time. The signal observed is

located close to the excitation wavelength requiring the

excitation source to be also spectrally narrow, as well as

intense. A laser fits both of these qualifications and is

responsible for the increased interest in Raman Spectroscopy

since the invention of the laser in 1964. A double or triple









2

monochromator or a notch filter is needed to reject the

intense laser radiation, so that small Raman signals near the

laser line can be observed.

Because a Raman signal can be generated by any laser

wavelength, the researcher can choose the spectral region he

wishes to work in [3]. The ability to monitor the signal in

the visible or ultraviolet regions allows vibrational

spectroscopy to be carried out in aqueous solution, which is

difficult with infrared absorption. It also allows one to

choose a portion of the spectrum which is free from other

spectral interference, such as fluorescence from solution

species [4]. Equipment already existing in the laboratory for

other types of analysis, such as luminescence, can be adapted

to monitor Raman signals. It is also relatively easy to probe

vibrations below 600 cm1, which is difficult with IR

absorption spectroscopy because of the poor sources, optics,

and detectors available for the far infrared region.

One of the methods for increasing the quantum efficiency

of the Raman process is to adsorb the molecule of interest

onto a surface [5]. Surface Enhanced Raman Spectroscopy

(SERS) involves both electromagnetic enhancements in the local

field and the chemical enhancements in the molecular

polarizability to increase the intensity of the Raman scatter.

Several surfaces, most notably gold, silver and copper,

provide enhancements of up to six orders of magnitude. This

enhancement over the Raman scatter from the free molecule










opens the door for trace analysis using Raman spectroscopy.

The nature of the enhancement mechanism is dependent on the

orientation of the molecule relative to the surface to be

probed. The enhancement and the orientational information

combine to yield a probe of startling spectral and spatial

resolution. A limitation of this method is the necessity of

close interaction between the molecule and the surface. This

limits the analytical application of this technique to

molecules with functional groups containing nitrogen, sulfur,

oxygen and phosphorus.














CHAPTER 2

OVERVIEW OF PREVIOUS WORK

SERS on Electrochemical Surfaces

SERS was first observed on a silver electrode by

Fleishmann et al. [6] who were interested in using Raman

spectroscopy to study species adsorbed at the electrode

surface. Intense Raman scattering was observed from pyridine

adsorbed at the roughened silver electrode surface. They

attributed the high intensity of the Raman scatter from

pyridine to an increase in the number of molecules able to be

adsorbed onto the roughened surface.

Later Fleishmann et al. [7] investigated pyridine

adsorbed onto a copper electrode, but failed to see an

increase in the Raman scatter because they used 514.5 nm for

the excitation when red wavelengths were required. They did,

however, observe an increase in the adsorption of pyridine to

the electrode surface with increased chloride concentration,

the subject of many later studies.

SERS was later observed on copper and gold electrode

surfaces by Wenning et al. [8,9]. They reported that SERS was

only observed on gold and copper surfaces when irradiated by

"red" wavelengths. The observed enhancement in the intensity

of Raman scatter from pyridine was five orders of magnitude









5

for a copper electrode surface and four orders for a gold

electrode surface. These studies included measuring the

angular dependence of the SERS signals on electrode surfaces.

The optimum collection angle for SERS measurements was found

to be 550 from the normal of the surface.

Jeanmarie and Van Duyne [10] and Albrecht and Creighton

[11] independently investigated the silver electrode-pyridine

system. Interpreting the results within the framework of

experiments by Barradas and Conway [12] on the adsorption of

pyridine onto metal surfaces, they concluded that the increase

in the intensity of the Raman scatter was not due to increased

adsorption. Both groups based this on the fact that a single

roughening cycle of the electrode yielded a greater

enhancement than multiple roughening cycles. The increase in

the Raman scattering was correctly attributed to an increase

in the Raman cross-section due to the adjacent rough metal

surface, rather than an increase in number of molecules

absorbed onto the surface because of the increased surface

area of the electrode.

Analytical Applications

Carrabba et al. [13] evaluated the potential of SERS on

silver electrodes for determinations of organic water

contaminants. All of the experiments in this study were

performed in low ionic strength solutions to simulate in situ

groundwater conditions. No interference was observed from

humic acid materials, which would be present under field









6
conditions. The silver electrode potential controlled the

adsorption of pyridine and quinoline. Variation of the

potential allowed discrimination between these two compounds.

The detection limit for pyridine was estimated to be 8.5 pg.

Force [14] reported the first analytical application of

SERS at an electrode surface to a flowing system. The silver

electrode underwent a short Oxidation-Reduction Cycle (ORC)

prior to sample introduction. The electrode potential was

then adjusted so that the pyridine would adsorb onto the

surface. After the spectrum was obtained, the potential was

adjusted to desorb the pyridine and another short ORC

performed. This approach obviously can only be used for

analytes which adsorb reversibly onto the surface. A detection

limit of 250 nmol was obtained with linearity extending for

three orders of magnitude above the detection limit. Pothier

and Force [15] have also used this system for the analysis of

adenine, thymine and cytosine. They obtained detection limits

of 175, 233, and 211 pg, respectively, for these compounds,

using a multichannel analyzer. The linear dynamic range

extended four orders of magnitude above these values.

Cotton et al. [16] combined SERS with a resonance

enhancement performing SERRS on a silver electrode surface.

This technique was used to detect four nitrophenol compounds

previously separated by HPLC. The eluent was collected by a

fraction collector, and the spectra measured off-line using a









7

silver electrode surface. Detection limits of approximately

20 ppb were reported.

Physical Studies

The SERS spectrum of pyridine on copper, silver, and gold

electrode surfaces has been the subject of many

investigations. The excitation profile of the pyridine

breathing mode for each of these surfaces was reported and

discussed in terms of a chemical enhancement mechanism [17].

Changes in the relative intensities, frequency shifts and

increases in the half-width were indicative of metal-adsorbate

complexation. Increases in the intensity of the scattering

were observed when chloride ions were present, indicating that

chloride ions mediated the surface complex formation. They

also suggested that most of the surface features are smaller

than 5 nm, since a stronger oxidation-reduction cycle (ORC)

would be required for the formation of larger surface

features.

The pH dependence of the SERS signal of pyridine adsorbed

onto a copper electrode was measured by Cooney and Mernagh

[18]. They reported that an ORC was not necessary for the

observation of SERS from a copper electrode, as it was with

silver and gold surfaces. They attributed this to the strong

coordination of pyridine with the surface. The optimum pH for

the observation of SERS from the pyridine copper electrode

system was found to be = 7 Evidence of laser damage to the

electrode surface by the krypton ion laser also was discussed.








8

Crookell et al. [19] used an FT-IR spectrometer to

measure SERS in the near infrared region. The 1064 nm line of

a Nd-YAG laser was used to irradiate copper, silver and gold

electrode surfaces, with two narrow bandpass filters being

used to excluded the laser scatter. SERS signals were

observed from pyridine adsorbed on these surfaces. The

potential dependence of the SERS signal for the pyridine-

silver electrode system from -0.1 to -0.9 V vs. SCE was

reported. It was noted that this might be an ideal region,

since fluorescence was not often a problem, and it was still

possible to obtain high quality Raman spectra from aqueous

systems.

The SERS spectra of the adenosine and two of its

derivatives on a silver electrode surface were reported by

Koglin et al. [20]. They found indications that the w bonds

of the ring systems was able to interact with the surface,

giving rise to SERS. It was noted that the presence of (PO3)2

groups in a molecule significantly increased the Raman

intensity, due to their strong association with the surface.

The SERS spectra of adenine, guanine, thymine, and

cytosine on silver electrodes were reported and compared by

the same group [21]. Differentiation between these DNA bases

is possible from the frequencies of the breathing modes alone.

A partial assignment of the bands was also presented. A more

complete assignation of the bands arising from these DNA bases









9
and a discussion of the orientation of these molecules on the

surface was given by Otto et al. [22].

SERS on Colloidal Metals

SERS on colloidal metal surfaces was first observed in

1979. Creighton et al. [23] reported intense Raman scattering

from pyridine on colloidal silver and gold surfaces. The

excitation profiles for pyridine on both of these surfaces was

reported. A correlation between the extinction profiles of

the hydrosols, and the excitation profiles of pyridine on

these surface was shown. This was the first paper to assert

that the SERS process involves the excitation of surface

plasmons. Both of these colloids were prepared using a

borohydride reduction.

Creighton et al. [24] were also the first to observe SERS

with copper colloids. A borohydride reduction was again

used, but citrate was added to stabilize the colloid. The

SERS spectrum of pyridine, diphenyl sulfide, thiophenol and

4,4'-bipyridyl on this surface were reported. In a mixture of

pyridine and another of the analytes, competitive adsorption

was observed at the surface. A Raman excitation profile

similar to that of gold was obtained for pyridine. The

maximum enhancement observed was calculated to be 1.5 X 105,

50% of that of that for a similar gold hydrosol.











Analytical Applications

Torres [25] was first to use silver colloids for

analytical trace analysis. SERS spectra for p-aminobenzoic

acid, phenytoin, 2-aminofluorene, as well as uracil and some

of its derivatives, were reported. The effects of different

preparations on the analytical figures of merit of 2-

aminofluorene, and p-aminobenzoic acid were also studied. All

of the colloids were produced in batches (static conditions).

A later study by Morris et al. [26] used sedimentation to

fractionate silver colloids prior to the analysis. Crystal

violet was used as the analyte to evaluate the SERS activity

of these fractions. The linear dynamic range for the fraction

yielding the greatest enhancement was between 1 pg and 1 ng.

This fraction was only partially aggregated. The use of this

fractionation technique increases the amount of SERS active

surface available for the analyte to adsorb onto relative to

the total surface area of the colloid.

The use of photoreduced silver as a SERS substrate was

evaluated by Ahern et al. [27]. The silver metal was produced

via a photoreduction of silver nitrate. The spectra were good

relative to chemically produced colloids, exhibiting better

signal to noise characteristics than the silver hydrosols

produced in that study. It was not possible to observe SERS

from species which are highly soluble in water. There was

some problem with preparing a narrow distribution of particle

sizes.









11

Silver colloids were used in the trace analysis of sulfa

drugs by Sutherland et al. [28]. Limits of detection were in

the low ppb range, with the added ability to qualitatively

differentiate between sulfadiazine, sulfamerazine, and

sulfamethazine.

Laserna et al. [29] first applied colloidal SERS to Flow

Injection Analysis (FIA). For para-aminobenzoic acid (PABA),

the limit of detection was in the ppb, while a 5% RSD was

obtained. Their research demonstrated that the variance in

colloidal SERS intensities was due mainly to variation in

colloid preparation. FIA allowed the aggregation of the

hydrosol, the mixing of the analyte and the colloid, and the

time of observation to be precisely controlled. The effect of

pH on the SERS signals of PABA was also investigated. The

SERS spectrum of 9-aminoacridine [30] was later evaluated to

determine the applicability of this technique to other

systems. The effects of the addition of various ions was also

investigated.

Pararosaniline was used as the analyte by Freeman et al.

[31] to test an adaptation of colloidal SERS to HPLC. In this

paper, a previously prepared silver hydrosol was mixed with

the eluent from an HPLC. A linear range between 0.1 and 50

ppm was reported for this dye.

FIA was used by Taylor et al. [32] to evaluate combined

resonance and surface enhancement in amplifying the Raman

signal (SERRS) for application to HPLC detection. Their study









12

used micro-Raman optics and multichannel detection, allowing

the collection of sufficient information for both qualitative

and quantitative analysis in an on-line situation. The

detection limit for a dye, crystal violet, was determined to

be on the order of 600 molecules.

Ni et al. [33] have also used an FIA system to test the

feasibility of using colloidal-SERS for the analysis of HPLC

eluents. In this study, they observed an increase in the SERS

signal from uracil with increased temperature. They also

measured the effect of pH on the SERS signals from uracil,

cytosine, adenine and guanine. The scattering from these

bases was highest under acidic conditions. Detection limits

in the 100 nmol range were reported.

Cotton's group [34] recently used a multichannel analyzer

to detect the SERS signals from four purine bases which had

been separated by HPLC. Adenine, guanine, xanthine and

hypoxanthine were separated using reverse-phase HPLC.

Injecting increased amounts of analyte increased the width of

the peak, but not the height. Detection limits in the sub-

nanomolar range were reported.

Physical Studies

Silver hydrosols have also been used, by Heard et al.

[35], to study competitive adsorption onto a surface.

Competition was observed between pyridine and citrate for the

silver surface. It was found that chloride assisted the

pyridine in displacing the more strongly adsorbed citrate from









13

the surface. The observation of this competition using SERS

added credence to the chemical enhancement theories, since it

indicated that surface complex formation was important. The

addition of a stabilizer to the colloid before adding the

pyridine reduced the effect of the chloride.

Joo et al. [36] showed that it was possible to follow

surface induced reactions using SERS. Aromatic mercaptans,

which were chemisorbed onto the colloidal silver surface,

underwent dissociation. This reaction appeared to be a

surface induced photoreaction resulting in a scission at the

S-C bond nearest the aliphatic group. The location of the

scission is opposite the result produced via pyrolysis of the

same molecule.

Colloidal silver has also been used, by Garrell et al.

[37], as a SERS substrate in non-aqueous systems. Colloids

were prepared in a number of organic solvents including

acetonitrile and dimethylformamide. These two solvents proved

to be the most suitable for SERS analysis. The SERS spectrum

of t-butylamine on an acetonitrile silver sol was presented as

the first SERS spectrum from an anhydrous colloidal system.

Colloidal systems, such as these, may have use in non-aqueous

systems where electrochemical surfaces cannot always be used.

Several papers since the landmark paper of Creighton et

al. [23] have described the SERS signal from pyridine adsorbed

onto silver surfaces. In one study by Dawei et al. [38] a

silver chloride surface was used. The SERS activity of the









14

silver chloride colloid was seen to decrease with increased

exposure to light. The treatment of silver chloride colloids

with thiosulfate, ferricyanide, or hydrogen peroxide reduced

the SERS intensity of pyridine. Similar results were observed

for silver bromide colloids by the same group [39].

The SERS spectra of halide ions associated with the

surface were measured on a silver colloidal surface by Wetzel

et al. [40]. Chloride was found to enhance the signal of

pyridine on both gold and silver colloidal surfaces. The

addition of pyridine did not increase the intensity of the

SERS from the halide ions. The increase in the pyridine SERS

signal was attributed to the formation of a surface complex by

the pyridine which was mediated by the halide ions.

The SERS spectra of cyclohexene and pyridine on silver

colloids were measured by Yamada et al. [41]. The SERS

intensity from pyridine and cyclohexene was compared with the

intensity of the SERS active bands of water. The strong SERS

enhancement for these chemisorbed species, relative to water,

along with shifts in both the frequencies and the relative

intensities of the bands, were seen as supporting the chemical

enhancement mechanism.

Moskovits at. al. [42] observed that an increased

concentration of molecules at the surface could produce a

change in the orientation of the molecule relative to the

surface. At low coverages, 2-naphthoic acid seemed to lie

flat on silver colloidal surfaces, interacting through both









15
the w bonds of the ring system and the carboxylate group. At

higher concentrations, 2-naphthoic acid was oriented

perpendicular to the surface, adsorbed exclusively via the

carboxylate group.

The effect of pH on the adsorption of 2,2'-bipyridine

onto silver colloids has also been studied. Kim et al. [43]

observed that decreasing the pH using HBr was found to

increase the SERS signal from 2,2'-bipyridine. An opposite

trend was observed when HI and HC1 were used. Any change in

the pH caused a shift in the SERS band positions.

The SERS spectra of several amino acids and nucleotide

bases adsorbed onto colloidal silver surfaces were studied by

Suh et al. [44] to determine their orientation on the surface.

The majority of these molecules were found to lie flat on the

surface. An exception was noted for m-aminobenzoic acid,

which at low pH adopted a geometry of either standing up, or

tilting away, from the surface with adsorption occurring

exclusively via the carboxylate group.

Adenine and eight of its derivatives were investigated by

Kim et al. [45] using silver hydrosols. The effects of pH

and analyte concentration on the resulting spectra were

reported. Changes in the pH were affected by the use of

phosphoric acid. Acidic conditions were necessary for SERS to

be observed from these DNA bases, except when pyridine had

been added. This demonstrated that the addition of acid, or

some other adsorbate, might be necessary to induce aggregation









16

of the silver colloid. Of this series, only adenine was

observed to undergo a coverage dependant change in

orientation. A similar study was undertaken by the same group

for uracil and three of its derivatives [46]. Like the

adenine derivatives, these compounds preferred to lie flat on

the surface under the conditions studied. Uracil adsorbs end

on in the deprotonated form, but assumed a flat orientation

when additional sodium borohydride was added. The change in

orientation was attributed to the decrease in the surface

potential caused by the addition of the sodium borohydride.

The SERS signals from citrate adsorbed onto a slightly

aggregated silver hydrosol and a highly aggregated silver

hydrosol were compared by Blatchford et al. [47]. Both of

these silver hydrosols were protected using citrate to prevent

further aggregation. When additional sodium borohydride was

added, the SERS signal observed from the citrate decreased.

The addition of the sodium borohydride served to decrease the

potential of the surface.

The concentration and temperature dependence of the SERS

spectrum of PABA on a silver colloidal surface was studied by

Suh et al. [48]. The intensity of the 1452 cm'1 band was shown

to be linearly dependent on the PABA concentration. The

system was studied at pH 7, making the anion the predominant

species. The molecule was determined to be adsorbed flat on

the surface. A similar conclusion was reached for PABA

adsorbed onto a silver electrode surface by Park et al. [49].










SERS on ChromatoQraphic Substrates

In 1984, Tran [50] reported the first SERS spectra from

a chromatographic surface. He reported the SERS spectra of

four dyes adsorbed onto a cellulose filter paper. Limits of

detection for these dyes ranged from 0.5 ng for crystal violet

to 240 ng for methyl red. The analytes first were mixed with

silver hydrosols and then introduced onto the surface for

analysis using a syringe. A helium-neon laser was used as the

excitation source. For the case of crystal violet, the

surface enhancement was combined with a resonance Raman

enhancement (SERRS). This additional enhancement helped to

account for its much lower detection limit.

Later that year, Tran [51] performed a similar study

after first separating three dyes using paper chromatography.

These chromatograms were dried and then sprayed with a silver

hydrosol using a nebulizer. The SERRS spectrum of crystal

violet, malachite green, and basic fuchsine were evaluated on

two commercially available chromatography papers, yielding

detection limits of approximately 2 ng for these dyes.

A study by Berthod et al. [52] followed, with an

analysis of the analytical application of silver colloids to

coated filter papers. The SERS spectra of PABA and several

other nitro and amino containing heterocycles were obtained,

but no quantitative results were given. Electron micrographs

indicated that the coated filter paper surface is not

conducive to theoretical modeling, because of the entrapment








18

of silver colloids within the filter paper, as well as upon

the surface.

Several nitro containing heterocycles were detected by

Laserna et al. [53] on a silver coated filter paper, both

alone and in mixtures. This paper also demonstrated that the

components of a mixture could be spectroscopically resolved

using SERS. Detection limits in the low ppm range were

achieved with an RSD of 15%.

The first SERS spectrum from a silica gel plate was

obtained by Sequaris and Koglin [54]. They reported the SERS

spectrum of 9-methylguanine using 514.5 nm excitation. The

analyte first was mixed with the colloid and then applied to

the surface. The detection limit was determined to be 120 pg.

Sequaris and Koglin [55] later separated ten purine

derivatives using HPTLC; the plates were dried and then

sprayed with colloidal silver. Detection limits in this

latter study for these biological compounds were estimated to

be less then 5 ng/spot. Koglin [56] used a micro-Raman

instrument in a similar study to improve the spatial

resolution of this technique; detection limits were in the ng

range for purine, benzoic acid, and 1-nitropyrene, but the

spatial resolution was improved to 1 Am.

In addition to the work mentioned above, there have been

several reviews. A book on SERS, edited by Chang and Furtak

[57], was published in 1982 and offers the best introduction

to the field. A review by Seki [58] cites all references to









19
the SERS effect before 1985. Seki's review provides a good

place to start a search from, but lacks text describing the

observed results. Chang [59] has also published a report on

the status of SERS research of electrode interfaces. Cotton

[60] has recently published a review of SERS investigations

of compounds of biological interest.















CHAPTER 3

THEORETICAL FOUNDATIONS

The Raman Scattering Process

Raman scattering arises from the inelastic scattering of

photons by molecular vibrations [61]. Raman scatter is

dependent upon a change in the molecular polarizability during

a vibration, whereas a change in dipole moment is required for

infrared absorption. The Raman spectra reflect vibrational

and rotational motions, which cause changes in the molecular

polarizability. The experimental data within this

dissertation only involves vibrational transitions, therefore

only these will be discussed in this chapter and the appendix.



For a light of a frequency, vL (s '), the electric field,

E, (V/m) of the wave can be expressed as


E = Ecos(2vvLt) (3.1)



where E0, (V/m) is the amplitude of the wave. For a vibration

at frequency, vV (s ), the polarizability at any time, t, will

be



a(t) = a0(l+cos(27vvt)) (3.2)









21
where a (C2.m /J), is the molecular polarizability. Since the

electric field induces a dipole moment, A (C.m), in the

molecule through the molecular polarizability


A = -aE (3.3)


the time dependence of the induced dipole moment is given by

the supposition of the equations 3.1 and 3.2


p = aoEoCOS(2 vLt)+ (1/2)aoEoCOS(2w(viLvv)t) (3.4)


The first term corresponds to the light scattered elastically

by the molecule (Rayleigh Scatter), the second contains the

anti-Stokes and Stokes Raman (inelastic) scattering terms.

The anti-Stokes term is represented by the summation of the

frequency of the laser and the frequency of the vibration.

The Stokes term is shown as the difference between these

frequencies. The physical result is the appearance of bands

on either side separated by AvV from the exciting frequency

(vL) as shown in figure 3.1.
The oscillating electric field of the incident radiation

may, because of molecular symmetry, induce a dipole moment in

a molecule whose axis is not parallel to the applied field.































Figure 3.1. a) A schematic Raman spectrum showing the stokes
and anti-stokes Raman scattering shifted Av,, for the first
vibrational transition and, Av2, for the second vibrational
transition, relative to the Rayleigh scatter at vL. b) A
schematic diagram illustrating the origin of the Raman
scattered frequencies.



















LASER
LINE


ANTI-STOKES SHIFTED


STOKES SHIFTED


VIRTUAL LEVELS


STOKES SHIFTED


LASER LASER
LINE LINE




2
1


ANTI-STOKES SHIFTED


MOLECULAR
VIBRATIONAL LEVELS


a)


LASER
LINE









24
The relationship between the induced dipole and the applied

field is best shown by the matrix form of the polarizability

tensor



x = axx axy exz Ex

y = a a"w ayz Ey (3.5)

z. = azx a C (3.5)E


For spontaneous (normal) Raman scattering, the polarizability

matrix is real and symmetric, so the square matrix can be

diagonalized by choosing a principal axis. This attribute

allows one to calculate the magnitude and orientation of the

induced dipole.

The intensities of the Raman bands can be calculated

using the formula for the intensity of radiation, I, emitted

by an oscillating dipole.



I = (164fv /3c ')2 (3.6)


where c, is the speed of light. If the Stokes shifted lines

originate from the ground vibrational level, then the anti-

Stokes lines must originate from higher vibrational levels.

Since the populations of these vibrational levels follows a

Boltzmann distribution, the ratio of the relative intensities

of the Stokes and anti-Stokes bands can be expressed by

Is/Ias = ((VL+Vv)4/(VL-'v)4) exp(-hv,/kT) (3.7)








25

The selection rules for the harmonic oscillator allows only

Av =1, although often overtones (2vv) and combination bands

(vV, v2) are found due to anharmonicity in the molecular

vibration [62]. The presence of a surface directly adjacent

to the molecule may lower its symmetry, resulting in the

appearance of bands that are Raman forbidden in the free

molecule [61,2]. The selection rules then become "propensity

rules," with the z2 terms being the most intense. The z axis

extends outward into the surrounding medium, perpendicular to

the surface, and the x and y axes are parallel to the surface.
2 2
The xz and yz terms would be next most intense, and the x y ,

and xy terms would be the weakest.

The SERS Enhancement Process

The forces acting on a molecule located at a distance, r,

from a surface may have a variety of sources [63]. The most

important of these is the electric field of the laser. The

image field, which accounts for the polarization of the metal

by an induced dipole located near the surface, and the field

caused by the presence of induced dipoles located near the

molecule, may effect the absorption and resonance Raman

spectra of the molecule. The normal, non-resonant, Raman

intensity is only affected by the electric field of the laser.


(3.8)


Ar = ar'EL(r,vL) "exp[-i(vL+v)t]









26
where Ar (C-m), is the induced dipole, ar (C2m 2/J), is the

polarizability, EL(r,vL) (V/m), represents the field generated

by the laser at r, vL (s1 ), the incident laser frequency and,

v, (s ), the vibrational frequency of the mode.

From this equation, one can see the beginnings of the two

classifications of theories concerning the evolution of the

SERS effect. The first enhancement mechanism is termed

electromagnetic. This mechanism attributes the large

enhancements to local field enhancements due to either

resonances within the surface or points of high curvature.

This would result in a larger field term, EL(r,vL). The second

enhancement mechanism involves chemical effects; chemical

effects are due to direct interactions between the molecule

and the surface, which cause changes in the polarizability of

the molecule, ar

Electromagnetic Enhancement Theory

For the calculation of purely electromagnetic

enhancements there are three requirements. First, one must be

able to represent adequately the polarization of the molecule,

usually as a point-dipole. This condition is a "far field"

approximation, where the distance between the separation of

charge is negligible when compared with the distance between

the molecule and the surface. Second, the surface must have

a relatively simple shape, so that it may be easily modeled.

Third, Maxwell's equations must hold even at distances as

small as 1 A. The last is needed because the field felt by









27
the dipole as it nears an irregular surface becomes

increasingly inhomogeneous, causing the model to break down.

On smooth surfaces, the enhancement is observed to be

approximately 10-102 [64]. The enhancement from the smooth

surface should decay rapidly on an atomic scale as the

molecule is moved away from the surface. Spherical surfaces

are easy to model and able to couple directly with optical

fields [65].

The local field, ELc, caused by a surface located at the

center of the sphere is given by the equations



Etoc o [(E L)-Eo)/ (vL)+2c0] a3 (3.9)


where a is the diameter of the sphere, e(v) is the dielectric

function of the metal, and Eo is the dielectric function of

the surrounding medium. Each of these terms is composed of a

real, e,, and imaginary, e2, component as shown in equation

below



e = e1 + ie2 (3.10)



If the frequency of the laser is near the resonance frequency

of a sphere, the enhancement is inversely proportional to

square of the imaginary component of the dielectric function

of the metal surface, [e,] 2, at that frequency.









28
It is important to have a small imaginary component for

the dielectric constant when the resonance condition,

[l,(v)] = 2co is satisfied. Figure 3.2 shows the real and

imaginary components of the dielectric functions of bulk

copper and silver in a vacuum. Figure 3.2 can be used to

determine at which frequencies the surface plasmons of the

metal can couple with the incident radiation. The surface

states of a sphere can be in resonance with not only the

incident radiation, but also the scattered radiation. This

resonance condition allows the Raman scatter of the molecule

to be amplified [4]. Figure 3.3 schematically compares the

spontaneous (normal) Raman and SERS processes.

Prolate spheroids can be used to theoretically model

colloids which have undergone agglomeration. The

electromagnetic equations are similar to those for a sphere,

but illustrate the difference between of the local field at

the end and the sides of the spheroid. For the resonance

condition (E(VL) = -2c0), the local field at the tip of the

spheroid is enhanced by, G, the same amount as a sphere



G =I -1/e212 (3.11)


The enhancement on the side of the spheroid will be reduced

from this by a factor of E2. The tip also allows the

efficient coupling of the scattered Raman radiation with the































Figure 3.2. The real, el, (solid symbols) and imaginary, E2,
(open symbols) components of the dielectric functions of
copper (circles) and silver (triangles) [71].












































2.90


3.60


energy (eV)


S.=50
1.50


2.20


4.30


5.00













P4H 4
H r- &

$ 4
0.c i.,
00.
MS4

.g0
400
4)






4) 4)tr

0
4))

0>O4
*H ;
St
1-4 *4)




4 44m






44pC
= Q)




10 M

0 0
r- 4)
.H l 1

4 E-H14



E4 *

o wg1o




< 0- 0
-M 4





Vt 0



a 0Cl4
-U 4 ) U



*H U., 0
4-)
0 V) P
to~f








32
LU











LU
+










<^

w
LU L







Il v^



VV W^JJ^^J









33

surface, so that if vR vL then the scattered radiation

undergoes additional amplification.

If this occurs, a molecule interacting with the spheroid tip

will experience an enhancement of



G = le-1/214 (3.12)



with this being reduced by a factor of E4 for the sides of the

spheroid. This enhancement arises from the small radius of

curvature of the tip of the spheroid, not the aspect ratio of

the spheroid [66]. The basis for the increased field at the

tip of the spheroid is that in the maintenance of an

equipotential surface, the image charge would need to be close

to a point of high curvature. This translates to a more

intense field outside that portion of the surface. Figure 3.4

illustrates the variation in the enhancement factor for a

prolate spheroid as a function of polar angle.

The electromagnetic model accounts for the long range

enhancement observed by Murray et al. [67]. The results from

calculations of the distance dependence of the enhancement

from a prolate spheroid are shown in figure 3.5.

If the enhancement was purely electromagnetic, then the

Raman excitation spectrum would be dependent only on the

nature of the metal and the vibrational frequency of interest

[63]. This type of behavior has been observed on gold and













oV o
CDO0


$4 0



"0 0


I.-I

* 0 *


gont
. -
*C






4)
-,-IJ mo




-,.4
) *0
cu*2^







4'r 0 ONI
SC0
g,4



0 44
gog
g 0







040
*.4 q 0




4 000
r0


u V

0 41
.4 44 %




Np V 0 0-













0













L0)



L (D

C)




mC2L


(si!un -qJe) JoCoej .ueueouequ '*O












4; .94
O 0

Qg
.5 .

4)C
4a4


0

.1 >1


0
MO




(4-
Ck


Sto

E0 4
-0)
UCe

U)0 *






0 0






U00



404.

EaOto
0


0 .4.1 S

>9 ,
40


0 01
o 0






:5 1
Cr Q)












































(sl!un *qje) JO01oe-j luweueoueU 'D











silver surfaces for water, but not other adsorbates [59].

The observed correlation between the extinction profiles and

elastic scattering excitation profiles of non-aggregated

colloids can be accounted for by the electromagnetic theory.

The electromagnetic theory cannot account for differences in

the Raman excitation profiles for various molecules, since it

does not handle chemical effects [68]. Contrary to this is

the success of Weitz el. al. [69] in calculating the SERS

excitation spectrum of nitrobenzoate on silver island films

using the electromagnetic model.

Critics of the electromagnetic theory consistently point

out that it does not account for changes in the selection

rules and over-estimates the enhancement [70]. This tendency

to over-estimate the enhancement may be due to the use of the

bulk phase dielectric constants [65]. The use of the bulk

phase dielectric constants has been shown to be valid only for

cases where the length or diameter of the particle is less

than 1/15 of the exciting wavelength [4]. This breakdown been

observed in the measurement of the dielectric functions of

thin metal films [71].

Chemical Enhancement Theories

Chemical enhancement mechanisms require strong

interactions between the molecule and the surface. These

interactions result in an alteration of the polarizability

matrix elements of the molecule. This alteration is

attributed to either charge transfer [72,73], excitations of









39

electron hole pairs requiring "adatom" type surface defects,

or chemisorption induced resonance [74]. All of these models

are short range mechanisms, predicting no enhancement past the

first one or two monolayers adsorbed onto the surface.

This appears to be the case for an experiment involving

cyanide and pyridine [75]. The SERS bands from these

adsorbates were enhanced by up to six orders of magnitude.

The spectra were highly potential dependent, exhibiting SERS

only for the molecules that were strongly adsorbed to the

electrode surface. Competitive adsorption, similar to this,

has been observed on silver [35] and copper [24] colloids as

well.

Of these mechanisms, only the "adatom" model of Otto et

al. [76] has stood the test of time well. The charge transfer

and chemisorption induced resonance models do not specify the

necessity of active sites [4]. The enhancement from smooth

surfaces is small and can be accounted for by electromagnetic

mechanisms without the need to invoke the chemical models

[70]. The small enhancement observed on smooth surfaces is

commonly used to discount the effects of chemical enhancement

mechanisms.

The adatom model employs surface defects to break the

translational symmetry parallel to the surface. These

variations in surface morphology mediate the coupling between

photons and plasmons. This reduction in the symmetry of the

surface is also used in the electromagnetic model. The major









40

difference between the two theories on this particular point

is how far away the molecule can be away from the surface and

still feel a difference between a surface defect and the

planar surface. This is related to the relative degree of

roughness present in the surface. For the adatom model the

surface features are composed of several atoms or a defect in

the crystal structure, while the electromagnetic model uses

surface features on the order of 10-50 nm in the calculations.

Since surface defects mediate the photon plasmon

interactions, molecules adsorbed onto the surface at or near

these defects will be more easily excited. Coupling between

the electronic states of the molecule and these surface states

would account for both the breakdown of the selection rules

and the observed continuum background of inelastic scatter

[63].

Experiments in UHV conditions have examined the effect of

annealing surfaces on the SERS activity. In these

experiments, [77,78] Raman active surfaces were heated to

anneal them. Annealing the surface resulted in a loss in

Raman scattered intensity. In another experiment [59],

pyridine was adsorbed onto silver surfaces under UHV

conditions. For three low index faces, only one band

(992 cm ), attributed to physisorbed pyridine, was detected.

For a silver 540 face, the band at 992 cm1 was shifted to 1003

cm-1 due to chemisorption. In all cases this band was
cm due to chemisorption. In all cases this band was









41
unenhanced, which casts doubt on the importance of atomic

sized defects proposed in the adatom model.

There is also some data on colloids which is in conflict

with the adatom model's use of small surface defects. Cotton

et al. [34] have recently been using elevated temperatures (up

to 900 C) to increase the Raman scattering from molecules

adsorbed onto silver colloids. According to the adatom model

this heating should result in a decrease in the Raman scatter,

since much of the finer surface morphology would be smoothed

by the heating of the colloids, as annealing does to surfaces

in UHV conditions.

There is currently no theory which describes or explains

all of the SERS results. The previous section was compiled to

give a general overview of the competing theories. Under some

conditions, one theory may appear to adequately describe the

observed results. For aggregated colloidal systems, the

electromagnetic theory seems to best account for the observed

results. There are, however, indications that chemical

mechanisms do play some part. The electromagnetic model can

not account for shifts in the positions of the bands, or the

breakdown of the selection rules.














CHAPTER 4

COLLOIDS AND AGGREGATION

Metal colloids have been the focus of much attention.

The spectral properties of colloidal metals were first

quantitatively examined by Mie in 1908 [79]. Mie investigated

the scattering of light by gold particles of sizes

approximately the same size as the wavelength of light used

for excitation. His work formed the basis for further

experimental and theoretical work on what came to be known as

the Mie scattering phenomenon [80]. Absorption spectrometry

and light scattering techniques are still used to determine

the particle size and degree of polydispersity for colloidal

solutions.

The optical properties of colloidal spheres suspended in

aqueous solution can be described by the following parameters



Qext = (2/a')E (2n+l){Real(an+bn)} (4.1)

Qs = (2/a2)E (2n+l){lan2 + Ibn2} (4.2)

Qab = Qext Qsc (4.3)


where Qext is the extinction efficiency, Qsc, is the scattering

efficiency, Qb, is the absorption efficiency, a, is a size

parameter equal to (2wa/A), and an and bnrepresent a series of








43

coefficients dependent on the size of the particle and the

dielectric constant of the surface relative to that of the

medium. For very small particles, with radii less than 1/15

of the incident wavelength, these equations become



Qca = (8a4)/{31(e-1)/(e+2) 2} (4.4)
Qa = -4a{Im{(e-1)/(e+2)}} (4.5)


where Im represents the imaginary components of the dielectric

function of the metal surface relative to the surrounding

medium, and Qext is the summation of equations 4.4 and 4.5.

These simplifications are only valid for the Rayleigh limit,

when contribution from multipoles higher than the dipole can

be neglected.

The above equations have been used to calculate the

extinction spectra of spherical particles of silver and copper

in water [81,24] shown in figure 4.1. Peaks in the extinction

spectra enable one to follow the aggregation of a colloidal

system using conventional absorption spectrometry. It may be

possible to observe several peaks which correspond to several

different particle sizes [23], or only a single peak for one

particle size [24].

Another method of particle sizing is based on scattering.

Photon Correlation Spectroscopy (PCS) uses a diffusion size

model to determine particle size. Corrections can easily be

made for the refractive index of the particle at the













0

0
0
u









4,
r-l



C
1-I

0










N
0
m



*4-





0
4)
0










41
U
+>
0


4-)
U





4-,



o.
Hr














0
+-4
c4






















F-4 L-i
(0;
u *
rc)l
**p






E *-




















































O
LO 0 0
I,,- LOcJ
d d d


eoueqJosqV eA!~jeley


0
0
cO


0
LO
N


0
0
1--


0

LO
oC









o
0

0 .
0 c



0





Iq,
0
0
0



LO




m
14n
n3









46
scattering wavelength. Photon correlation spectroscopy yields

the distribution of the particle sizes, as well as, an average

particle size. One of the limitations of this method is that

the calculations usually assume a single mode distribution.

Electron microscopy can also be used to determine the range of

particle sizes, but care must be taken to avoid causing

further aggregation when preparing the sample for analysis

[82].

The aggregation process is controlled by the surface

charge of the particle and the thickness of the diffuse layer

surrounding the particle [83]. For colloidal particles, the

attraction is due mainly to van der Waals interactions. The

strength of these interactions vary as d6 with the

interparticle separation distance. The attractive forces, Ua,

(J/mol) for a spherical particle of radius a, is given by

A 8a2(d+a) d(d+4a)
U(d) --[ + 21n 1 (4.6)
12 d(d+a) (d+2a)2

where d is the distance between the surfaces, a is the radius

of the particle, and A is an empirically determined constant.

A is usually on the order of 10-19 J. From this equation, one

can expect uncharged particles to associate and eventually

precipitate from solution.

Since the majority of colloids have charged surfaces,

there are also repulsive forces to consider. The charge on

the surface is usually due to the adsorption of ions onto the

surface. There is also a diffuse layer, which shields the









47
surface charges from each other. This controls how near

similarly charged particles can approach each other. The

extension of this layer into the surrounding solution is given

by

*(d) 40(a/d) exp(a-d/X,) (4.7)

where 0 (V) is the potential of the surface and Xa (m) is a

distance parameter. The thickness of the diffuse layer is

related to the ionic strength of the solution through the

expression


Xa eeR (4.8)
F21

where e is the dielectric constant, eo is the permittivity of

free space, F (C/mol) is Faraday's constant, and I (mol/L) is

the ionic strength of the surrounding medium. From

expressions 4.7 and 4.8, we see that increasing the ionic

strength of the solution will compress the double layer,

resulting in increased aggregation. This accounts for the

"salting out" effect.

The repulsive forces, Ur, between two particles with ion

atmospheres surrounding them is difficult to calculate. The

trends observed follow the form


Ur(d) 27a3 2 ln(l-exp-dx') (4.9)
(l+a/Xa)2 ee,









48
The total potential energy of the system is the sum of

equations 4.6 and 4.9.

There are several ways to compress the diffuse layer and

induce aggregation [84]. The first is to increase the ionic

strength of the solution, so the extension of the diffuse

layer into the surrounding solution is reduced, making

collisions more probable. Another method is to adsorb neutral

compounds onto the surface, replacing the ions adsorbed onto

the surface and thereby reducing the surface charge of the

particle.

Cotton et al. [34] recently used temperature increases to

induce aggregation. This technique increases the velocities

of the particles in solution, so that the diffuse layer is not

as effective at preventing collisions between particles. More

quantitative explanations of the effect of increasing the

ionic strength of the solution, of raising the temperature and

of decreasing the surface charge of the particle on the

thickness of the double layer surrounding the particle can be

found in a treatment using the Gouy-Chapman theory [85].

The kinetic aspects of working with colloids also need to

be considered. A paper by Weitz et al. [84] describes the

aggregation process in fractal terms. The basis for the

application of fractal analysis is the self-similarity of the

colloids on various scales. The fractal dimension of a gold

colloid was shown to be 1.75. The use of fractal scaling of

the relationship between time and the mean radius of the









49

particles, yielded a logarithmic plot with a slope of 0.57 or

1/1.75. The fractal approach to quantifying colloidal

aggregation seems promising since it is accurate and provides

results which are easily visualized.















CHAPTER 5

SILVER COLLOIDS FOR SULFA DRUG ANALYSIS

Introduction

The development of drugs for therapeutic use both in

humans and animals has increased dramatically over the last

century. Over 10+8 kg of one particular class, sulfa drugs,

were administered in 1985. Of the 5000 sulfa drugs tested for

antibacterial activity, less than 30 are routinely used [86].

Sulfapyrimidines are an important subclass of these

sulfonamides (sulfa drugs). Sulfamethazine, sulfamerazine and

sulfadiazine are members of this group [87]. These drugs are

able to block the formation of dihydrofolic acid from p-

aminobenzoic acid, thereby inhibiting the biosynthesis of

folate cofactors in bacteria [88].

Sulfa drugs are not commonly prescribed to humans

anymore, because of the development of other antimicrobial

agents and the increased resistance of bacteria to these

drugs. Sulfa drugs are still widely used in veterinary

medicine, since they are easily administered. Often two or

more of these drugs are used in combination to produce the

necessary dosage. Crystalluria, the formation of crystals in

the renal tubules, can be caused by administering drugs which

have limited solubility. Since drugs of the same class do not









51

affect the solubility of each other strongly, the risk of

crystalluria is reduced.

The use of these drugs to treat livestock diseases has

resulted in the deposition of sulfa drug residues in the

tissue of swine. In 1973, the U.S. Food and Drug

Administration set a tolerance of 100 ppb for sulfonamides in

edible tissue. The current method for testing for this class

of compounds does not yield information about the identities

of the compounds found [86].

Most spectroscopic methods for determining sulfa drugs

involve diazotinization. The standard method of testing for

sulfa drugs is the Bratton and Marshall method [89], which

yields colored products by diazotinization; the colored

products can then be detected using absorption spectroscopy.

Later this procedure was modified by Iskander et al. [90] to

take advantage of the kinetics of the diazotinization to

differentiate between these compounds. This has some success

allowing differentiation between different classes of sulfa

drugs.

Diazotinization was used by Sato et al. [91] to allow

resonance Raman spectroscopy of these compounds to be measured

with excitation frequencies in the visible region. Detection

limits were estimated to be in the 10.8 M range.

Unfortunately, it was not possible to differentiate between

the various sulfa drugs spectrally.









52
Membrane electrodes were prepared by Yao et al. [92] and

used for the determination of Sulfa drugs. These membranes

were specific for sulfa drugs with detection limits in the pM

range but, were not able to differentiate between molecules of

the same class.

SERS on colloidal silver substrates will be shown to

provide qualitative and quantitative information for

sulfadiazine, sulfamethazine and sulfamerazine at levels in

the low ppb range.

Experimental

Apparatus

All of the spectra were obtained by irradiating the

samples with 100 mW of 514.5 nm line of an argon ion laser

(Spectra Physics, series 2000). The scattered radiation was

collected at 90 degrees and resolved using a 0.85 m double

monochromator (Spex Industries, model 1403) with the bandpass

set at 10 cm The detector was a cooled photomultiplier tube

(RCA, model C31034) coupled to photon counting electronics.

The acquisition, storage, and processing of the data was

carried out on an IBM compatible PC. All spectra presented

are single scans without any spectral smoothing.

Chemicals and Procedure

All chemicals were reagent grade or better and used

without further purification. Demineralized water was used

throughout the analyses. The sulfa drugs analyzed,

sulfadiazine, sulfamethazine, and sulfamerazine, were









53

purchased from Sigma Chemicals. The 100 ppm stock aqueous

solutions were prepared with 3% methanol to aid in

dissolution.

The silver colloids were prepared as follows: 100 ml of

an ice cold 1.75 X 103 M solution of AgNO3 was added dropwise,

with stirring, to 300 ml of an ice cold 7.2 X 10"4 M NaBH4

solution. An absorption spectrum of the resulting hydrosol

showed a narrow band with a maxima at 400 nm. Such a maxima

is indicative of a relatively monodisperse hydrosol containing

silver particle between 1 and 50 nm in diameter. The particle

sizes were also measured by filtration through polycarbonate

track-etched membranes, followed by SEM analysis. Each sample

analyzed was prepared by adding 0.1 ml of the analyte solution

to 0.9 ml of the room temperature colloidal solution. After

mixing, 0.5 ml of this solution was transferred to the sample

cell and analyzed.

Results and Discussion

The normal Raman and SERS spectra of sulfadiazine,

sulfamerazine and sulfamethazine are shown in figures 5.1 and

5.2 respectively. The shifts of the peaks are given in tables

5.1 and 5.2 in units of reciprocal centimeters. There are

several peaks in the normal Raman spectra, which are present

for all three of the compounds. The bands located near 1596,

1506 and 1094 cm1 in the these spectra are due to stretches

in the ring systems. The strong bands near 1150 cm1 are due

to symmetric vibration of the SO, group; these bands are









54
-1
accompanied by weaker vibrations near 1300 and 543 cm ,

attributed to the same group. The shoulders found in these

spectra at about 1635 cm'1 are commonly attributed to the amine

group. A similar feature found at 745 cm'1 is also attributed

to the amide.

Comparing these frequencies and relative intensities with

those found in the SERS spectra leads one to think that the

molecules are lying flat on the surface. The SO2 bands are

shifted from 1150 cm1 to 1114 cm"I, and are still among the

most intense in the spectrum. The other two modes appear near

550 and 1308 cm'1 for sulfamerazine and sulfamethazine, but not

for sulfadiazine. The bands near 1635 cm-' found in

sulfamethazine and sulfamerazine are shifted to 1630 cm .

This band appears in the SERS spectrum of sulfadiazine as

well at 1622 cm-1. The ring vibrations 1596, 1506, and 1094

cm are reduced in relative intensity, but not shifted

significantly. The large shifts in the vibrational

frequencies, as well as, the relative intensities of the bands

from the SO2 and amine groups indicate that they interact

strongly with the surface. From this information, the molecule

appears to lie nearly flat on the surface. This has

previously been suggested for molecules, such as DNA bases,

which also have more than one functional group.












4o0
* o


Cw

a4
14
$ 4

0
Ia

O,
#a













m u


8t.
010





to 0-4
4) I.
Me
044
' o*
01I
O N0




a)
0) (0

0 0

Id *to
aN
*-4 0)



E04






rni
0 C








q 0 r
SE C
.1d


SrCI
. P c
inu 10
0 .-
(U e Q
c r) k4
& fl) "










56







o
o



o
-.
0














I 0-'-
o








0
S4-(
o





o
i 1 o
0
C;








0
oD










0c





(0
n I) l u oo u
^ < ^ -
< \ 1 2

S S ^

^ ^ 1 r
f \ / o Q
\ I M Bo3












Table 5.1.
Raman Bands of Sulfa Drugs (in Solid Phase)
Raman Shift (cm )


sulfadiazine sulfamethazine sulfamerazine

229, sh 237, w 248,vw


286,
319,
339,


373,
402,
455,


544, w


578,
636,
661,

706,
750,
800,


vw
w
vw

vw
vw
w


273,
282,
300,
328,
342,
365,
380,

456,
504,
547,

585,
634,
659,
681,
715,
743,


vw
dbl,w
dbl,w
w
vw
w
w

w
w
m

w
w
vw
w
vw
m


290,
290,
330,


386, vw

439, w
529,vw
538, w
567, vw
579, w
637, w

685, w
716, vw


827, dbl,m 822, dbl,m 829,
848, dbl,m 840, dbl,m 842,
892, w 873,
939, vw 963, vw
995, m 999, m 999,
1098, s 1094, s 1092,
1150, vs 1153, vs 1144,
1191, w
1259, w 1241, vw
1312, vw 1298, w 1306,
1338, w 1333, w 1345,
1371, vw 1385,
1407, vw 1413, w 1416,
1438, vw 1440, vw
1466, vw
1478, vw
1505, w 1506, m 1509,
1581, sh 1560, w 1559,
1598, vs 1596, vs 1596,
1629, w 1640,
bb = broad band, sh = shoulder, m =
w = weak, v = very weak


dbl,m
dbl,m
vw

m
w
vs


vw
vw
vw
vw



w
vw
vs
w
medium, s = strong,










0) 0)
4JE
*9
W .4
. N




ua
0V

0,
,t






4*e
r*-4


.r 0,
4 )o
044
S





o i>
-H 4

S0V

to
0 *H 4
.N









0
4-4)
40
Ir4 k
|gc










r N

S43-r










* H -*
Naa
.r-4
















OH (
-40 0


0 E
1^a




a2
p^~
*

r OQ
a > i





























o
-o
{ i
i 's


o





cc
o



0
OC
o





0
o
o
o




0


O



0 0 0 0 0 0 0 0 0 00


(s,!un *qJe) Al!suelul peJalleeoS ueuwel












Table 5.2.
Surface Enhanced Raman Bands of Sulfa Drugs on Colloidal
Silver
Raman Shift (cm )



sulfadiazine sulfamerazine sulfamethazine



230, sh 230, sh 238, sh


574,
634,
684,

818,
974,


1068, sh
1114, vs
1186,vw


1340, m

1410, w


1502,
1594,
1630,

3050,
3258,
3378,


290,
342,
410,
462,
508,
544,
582,
626,
674,
762,
830,
976,
1002,
1066,
1114,
1174,
1252,
1312,
1356,
1390,

1454,
1502,
1594,
1630,
2930,
3068,
3258,
3396,


sh
vw
vw
vw
vw
vw
m
vw
w
s
m
vw
vw
sh
vs
vw
vw
m
sh
m


396, vw
458, vw
516, vw
554, w
590, m
638,vw
674, w
738, m
834, m
968, bb

1054, w
1112, vs
1186,vw
1260, vw
1302, vw

1394, m


1506,
1594,
1622,
2922,
3074,
3242,
3380,


bb = broad band, sh = shoulder, m = medium, s = strong,
w = weak, v = very


294, sh
364, vw
416,vw
458, vw
506, vw









61

From the chemical structures shown in figure 5.3, the

similarity and differences between these compounds is evident.

The SERS spectra reflect these similarities and differences as

well. The 2930 cm1 band, which is present with different

intensities for both sulfamethazine and sulfamerazine, is

absent in the sulfadiazine spectrum. The 2930 cm"' band is due

to the methyl groups on the pyrimidine ring. The locations of

the stronger peaks also differ slightly. The symmetric

stretch of the SO2 group is located at 1114 cm'I for

sulfadiazine and sulfamerazine, but is found at 1112 cm"' for

sulfamethazine. Similarly, a shoulder is seen at 1622 cm1 for

sulfamethazine, but is shifted to 1630 cm' in the other

spectra. There are also a number of peaks which appear in one

of the spectra, but not the others.

Figure 5.4 illustrates the ability of SERS to provide

quantitative information at ppb levels. The bands at 685 cm-'

and 1092 cm1 are attributed to vibrations of the benzene ring.

Variations in the intensities of these bands may be due to

slight oreintational changes of the molecule relative to the

surface. This presents a true test of the fingerprinting

capability of SERS. The intensity of the peak at 1114 cm"1 in

the SERS spectrum of sulfamethazine is useful for determining

concentrations of analyte between 10 and 500 ppb (see Table

5.3). This band was used for all of the analytical

determinations because it was linear with concentration over

several orders of magnitude. These results help to ensure


































Figure 5.3. Structures of the three sulfa drugs studied:
sulfadiazine, sulfamerazine, and sulfamethazine.











O
II
u SNH-
-- / I I K i-_I I

Sulfadiazine


O
11
SNH-
II
0


Sulfamerazine


II
u


Sulfamethazine


H\
N
Id!


N- /
H/


NM
H/

















4O
44





4J
cu



















O
,.4










0







0
0
M,Q













0
04
0











U4
*H




a4
(0

'o
0
0d










0
4iu
m












0o
0) A
C f
*i- Q





4-1



3 &










0r
































































0 0 0 0
-t C0 -


(S!iun "qje) Alisueaul peJealeoS ueweu



























Table 5.3.
Analytical Figures of Merite for Sulfa Drugs on the Silver
Colloidal System.
Compounds LOD's Sensitivity R
(ppb) (ppm )

sulfadiazine 1.0 66.74 0.997
sulfamerazine 1.0 143.82 0.999
sulfamethazine 10.0 16.62 0.999
a LOD is the limit of detection using three standard
deviations above the noise, the sensitivity is the slope of
the calibration curve, and R is the correlation coefficient of
the log-log plot.









67

that SERS will eventually provide high quality qualitative and

quantitative information as a method of routine analysis.














CHAPTER 6

A STUDY OF COPPER COLLOIDS

Introduction

The Raman scattering cross section can be increased up to

six orders of magnitude by bringing the molecule close to a

surface [6,10]. Copper, gold and silver surfaces have

generated the largest enhancements [58,69]. Because of this

enhancement, Raman spectroscopy can compete more effectively

as a trace analysis technique with other molecular

spectroscopies, such as fluorescence and phosphorescence. The

advantage of SERS over these techniques is two-fold. First,

not all molecules fluoresce or phosphoresce, whereas all

molecules have vibrational modes accessible to SERS. Second,

SERS spectra provide a wealth of information, allowing one to

identify the molecule and/or differentiate it from a similar

molecule adsorbed on the surface. These two reasons provide

the impetus for the development of SERS as a trace analytical

technique. Reviews by Garrell [1] and Campion and Woodruff

[5] have focused on the application of Raman techniques to

analytical problems.

Silver and gold colloids were first used as SERS

substrates in 1979 by Creighton et al. [23]. Much of the

research on SERS has utilized silver colloids as the









69

substrate. Silver colloids are easily produced, provide a

fresh surface for the adsorption of analyte molecules, and can

be excited by an argon ion laser. Silver hydrosols have been

used for the quantitative analysis of drugs [28,70,71] and

nitrogen containing heterocycles [72,53].

Silver hydrosols have been adapted for use with flow

injection analysis [30,31,33,29] and HPLC [34,52] systems. On

line measurements increase the reproducibility of colloid

preparation, thus lowering limits of detection. SERS has also

been adapted to both paper [52,54] and thin layer

chromatography [55] in efforts to further increase the

detection power and identification power of these separation

techniques.

Copper colloids were first evaluated for SERS in 1983 by

Creighton et al. [24]; in this study, the SERS enhancement of

pyridine by a citrate stabilized copper hydrosol was

investigated. A later study by Heard et al. [35] found there

was competitive adsorption between pyridine and citrate onto

the silver colloid surface. To minimize this interference, an

unprotected copper colloid was used in the present study. The

investigation of a copper colloid for use as an analytical

SERS substrate is reported for the first time. The colloidal

surface is prepared by a simple borohydride reduction of

cupric nitrate. These results illustrate the bright future of

SERS on copper colloids for analytical use.











Experimental

Apparatus

In Figure 6.1, a schematic diagram of the experimental

setup is shown. A Spectra Physics 2040 argon ion laser was

used to pump a dye laser (Spectra-Physics, model 375). DCM

(Exciton) was used to produce the necessary wavelengths. All

measurements were made relative to the 100 mW of 660 nm

exciting radiation. The stability of the output power and

lasing frequency were continuously monitored using both a

power meter and a 0.3 m monochrometer (McPhereson 218) with a

R928 Hamamatsu photomultiplier tube. The laser beam was

directed through the sample twice and the scattered radiation

collected at 90 degrees. A 0.22 m double monochrometer

(Spex 1680B) with slits set for 6 cm"1 resolution was used at

a scan rate of 3.0 nm/min for wavelength discrimination. The

Raman scatter was detected using a thermoelectrically cooled

(R636 Hamamatsu) photomultiplier tube. The photoelectron

pulses were amplified and counted using a (Stanford, model

440) photon counter with a 1 s acquisition time. The data was

collected on an IBM PC. The Raman bands used for quantitative

measurements were 1010, 1397, and 1387 cm1 for pyridine, p-

aminobenzoic acid, and p-nitrobenzoic acid, respectively.











W



0
C


0






0
U

r-I
u





4-1)

40



a)









a)

*-4



00
r0






*H
00


8.- ,

I0
o00






C
10
rz4







P4):





























Cfr
LUi




z
0



Z

0

Kc


) 0U
z
-7
I-
0
0


dS ---------^ OC
cr cr
(I Cc










Chemicals and Procedure

A 30 ml volume of a 2 X 10.3 M cupric nitrate solution was

added dropwise to 70 ml of a 1.0 X 10.2 M sodium borohydride

solution at room temperature with stirring. The analyte was

added as a 1.00 ml aliquot to the resulting solution. The pH

of the solution was then adjusted as needed using a small

amount of nitric acid. A sample of this pH adjusted solution

was then placed in a standard cuvette. Measurements began two

min. after the first two reagents were completely combined.

All glassware had been cleaned thoroughly with concentrated

nitric acid and the rinsed repeatedly with Barnstead water.

Pyridine, p-nitrobenzoic acid (PNBA), methanol and sodium

borohydride were obtained from Kodak (Rochester, NY). Cupric

nitrate and p-aminobenzoic acid (PABA) were obtained from

Fisher Scientific. The solutions of PABA and PNBA were

prepared in 20% (v/v) methanol. All chemicals were analytical

reagent grade or equivalent and used without further

purification.

Results and Discussion

Pyridine was chosen as the test compound for the initial

optimization because it is soluble in water, has a large Raman

cross section, adsorbs strongly to colloidal surfaces, and

SERS has been observed previously on copper colloids (figure

6.2). The enhancement factor for the pyridine 1010 cm'1 band,

corrected for prefilter and post filter effects was calculated

to be 1.8 X 105 which is in good agreement with previous









74

results [24]. Pyridine (Pk,=5.19) proved a good choice,

because of the relatively small effect that pH had upon its

SERS signal compared to those of the other analytes (figure

6.3). The optimal pH for the observation of SERS from

pyridine differs from that previously observed using a copper

electrode [18] and a different copper colloid [24]. The

optimal pH for the copper electrode and the copper colloid

used in this study differ by a 1.5 units, which may be due to

differences in the surface morphology or potential. The pH

difference between the copper colloids in the present study

and those used by Creighton et al. [24] is about 2.5 units.

This is probably due to their use of citrate to stabilize the

colloid. The adsorbed citrate will affect both the surface

potential of the colloid and the adsorption of the pyridine

onto the surface. The difference in the pH dependence of the

SERS intensities of PNBA (Pk,=3.45) and PABA (Pk,=4.65, 9.30)

are due to the differences between their dissociation

constants. The acidic form of the amino group on PABA will

adsorb more strongly than the basic form. At slightly basic

pH's, PABA is bound to the surface via both functional groups.

The effect of pH on PNBA is probably due mainly to the surface

charge of the colloid. The narrow pH ranges within which PABA

and PNBA exhibit SERS presents an additional method for differ

entiating between similar compounds.

The absorption spectrum of the colloid over time (figure

6.3) shows the growth of a peak at 570 nm. It has been












3^ .
* 0 V
C ftlP
*. 4 o
v 0 0
4 14


0 U
00



0
to-t



00



SO
t *
H0
olo


1t4



r)



05


O H





40 0
o 0
UM









rc.
a)







.o4
C
S0








04 4
30




|i4 ) (









76




o
0


0














U)

E





(um
C















(s:!un "qje) Al!suelul peielleoS uewel
% < } ^

-j ? S
-===^- -===3 < c/











W 04)

>4


( U
(0 'e
id 0
9)44




Og
(U0


Oo

0 0 O'






0
-00



10 0
944
0 >t
(a r-4 40







0






oogo
0


ot




0
c o
c1 C

















90 0 OD

0 > 0
u 4.






(U -*l4


-.-I to >4
44 tV C0'




























< 4
L










E--


ALISN3INI


Z
o 0
T- -

_J
0
(J)




0.
O_





co






N-


3AIV-1.D3









79
previously calculated that there is a correlation between the

570 nm peak and copper spheres having a diameter of z 20 nm.

The particle size distribution of the colloids was analyzed

with a Malvern 4700 photon correlation spectrometer. The data

analysis was performed using the optical constants for bulk

copper [71]. During the time span of interest, the average

particle diameter was found to be approximately 100 nm (see

figure 6.4).

The optical density in figure 6.3 reaches a maximum 780

seconds after the reagents were combined. If the 20 nm

particles are the precursors of the most SERS active particle

size, the SERS signals should increase while the 20 nm

particles are present and then decrease. This correlates well

with the SERS signals shown in figure 6.5., which is largest

for the 1397 cm'1 band at 800 s and decreases thereafter. The

time dependance of the 1397 cm1 SERS signal makes it difficult

to acquire a good spectrum using a scanning instrument

(Figures 6.7, 6.8).

The copper colloid system has a much lower background

than silver colloidal systems. Unfortunately, the SERS

signals are also smaller, so that the signal to noise ratio is

not improved. This may be due to lower adsorption onto the

copper surface as compared to silver surfaces. An additional

problem is that the copper colloid cannot be prepared more

than 30 min. ahead of time, where the silver colloid can be

made stable for future use [23,33,34].












. 0
0Q) I
*g







t o





0 .
'e -





-4J
r II
H ,





04)
4g *
0P
QrJo




U Q





?A .1 0

0

294 '
4) Q

0m0*


g o


M) 00) 0

$4 *4 )-4 0
0 4 N

4, 0 r. 0 0Q
rN -l -H I
or rl-r U 9






















































































eoueqjosqv


SDC:


Ln N d d d dC
.4 Ci d d 5


o 0c
a
D











* 0


FA
ton
a -d


tw
m *








0
*C 0


4)
Ow'r
000


Q'*


t4 O *4
V 4)4
0U1
ol 9



044





$40
.63
0O 0 g<
.0 4
U 4ffl
















$4-
0
04'1 4j






OU

0 .0











VA 44




$4 0
&'~ Q) ^
*rlX (^ ^ &I












83

















Ln
+ + + + +



+ + + + +




U-)


u-
+ + + + + "
0



+ + + + j-
+






+ + + + +



+ + + + +
\o u




+ + + + +







+ + + + +





U ) J -(\


uoilsodwodW luaejed






























Figure 6.6. Intensities of the 1606 cm" (filled triangle) and
1397 cm (filled circle) SERS bands of PABA as a function of
time. The .conditions were pH=9.0, [Cu ] = 6.0 X 10 [BH4 ]
= 1.9 X 10 and 50 ppm para-aminobenzoic acid.




















1397 cm-1


1606 cm-1


400
400


800


1200


1600


Time (s)


1 -




0-
0
O










*1


,-1 *
0oo






H*


o
0.


- 0







0 .


0o
1x4
A *



0 ,.
0
44 il






0


a,
uo



0
0
aV





4J
r/l
Wa
Wa










10
S-U
10 W




4a 0



0
rq
*ii ^
fcXUr-




























































o 0 0 0 0 o$
S00 o0 0
(sj!un "qje) Al!suehul pejeileoS ueweH=











-s

00
oa
r-l4

0
,4


$4
O
0




*d
0
o
Ca



*-,4







Ou



00


to
O




















ri8
4J.4






c 0












4m
0
I4O












So
o a
U* 4
4- 0
00




Q)
00








Co
0U


00
^^




9-3
b\g
H a
















































(SI!un "qje) A !suelul peJealeoS uejuey

































Figure 6.9. Calibration curves for pyridine, p-aminobenzoic
acid, and p-nitrobenzoic acid on the copper colloid.









































20 40 60


Concentration (ppm)


80


100




Full Text
SURFACE ENHANCED RAMAN SPECTROSCOPY ON METALLIC COLLOIDS
FOR THE PURPOSE OF TRACE ANALYSIS
BY
MARTIN JOHN ANGEBRANNDT
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
1991

ACKNOWLEDGEMENTS
I would first like to thank Dr. James D. Winefordner for
his guidance and support. I am especially grateful for the
considerable freedom he allowed me, while keeping me from
straying too far off course. I am also grateful to my
colleagues in Jim's laboratory and throughout the department
for all of the good and bad times we shared. I hope
Washington D.C. is not too far out of the way. Another thank
you must go to Tony Antonazzi, who operated the photon
correlation spectrometer for the experiments on the copper
colloids.
Considerable recognition must also go to Frangois Lang,
Bob Wallace, Ward Campbell, Dr. Robert Ho, and John
Glendenning, my Aikido Senseis, as well as the other Aikidoka.
They managed to keep my physical and spiritual development
near the mental and technical levels of my academic pursuits.
Finally, I am grateful to my family. My parents and
sister, who helped and supported me when I needed it most and
my new wife, who helped me keep a smile during one of the most
trying periods in my life.
ii

TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ii
LIST OF TABLES V
LIST OF FIGURES vi
ABSTRACT viii
Chapter
1 INTRODUCTION 1
2 OVERVIEW OF PREVIOUS WORK 4
SERS on Electrochemical Surfaces 4
Analytical Application 5
Physical Studies 7
SERS on Colloidal Metals 9
Analytical Applications 10
Physical Studies 12
SERS on Chromatographic Surfaces 17
3 THEORETICAL FOUNDATIONS 20
The Raman Scattering Process 20
The SERS Enhancement Process 25
Electromagnetic Enhancement Theory . . 26
Chemical Enhancement Theories .... 38
4 COLLOIDS AND AGGREGATION 42
5 SILVER COLLOIDS OF SULFA DRUG ANALYSIS . . 50
Introduction 50
Experimental 52
Apparatus 52
Chemicals and Procedure 52
Results and Discussion 53
iii

6 A STUDY OF COPPER COLLOIDS 68
Introduction 68
Experimental 70
Apparatus 70
Chemicals and Procedure 73
Results and Discussion 73
7 COPPER COLLOIDS IN A FLOWING STREAM .... 95
Introduction 95
Experimental 97
Apparatus 97
Results and Discussion 100
8 CONCLUSIONS AND SPECULATIONS ON THE FUTURE 111
Conclusions Ill
Speculations 113
Experiments on Filters 113
Novel Colloidal Surfaces 115
Methodology 115
APPENDIX ELECTROMAGNETIC DERIVATIONS 117
Molecule Near a Hemispheroid Protruding
From a Surface 117
Full Prolate Spheroid 126
The Enhancement Factor 128
REFERENCES 132
BIOGRAPHICAL SKETCH 138
iv

LIST OF TABLES
Page
5.1 Intensities and Positions of Raman Bands
for the Sulfa Drugs as KBr Pellets 57
5.2 Intensities and Positions of SERS Bands
of Sulfa Drugs on Silver Colloids 60
5.3 Analytical Figures of Merit for Sulfa
Drugs on Silver Colloidal System 66
6.1 Analytical Figures of Merit for Analytes on Static
Copper Colloid 93
6.2 Comparison of Static p-aminobenzoic acid Results
with Literature 94
7.1 Effect of Acidity on SERS Signals and Colloidal
Aggregation 102
7.2 Comparison of p-aminobenzoic acid FIA Data with
Literature 109
A. 1 List of Symbols used in the Appendix 131
v

LIST OF FIGURES
Page
3.1 A Schematic Diagram showing the Stokes and
Anti-Stokes Shifted Raman Scattering Relative to
the Rayleigh Scatter and their Origin 22
3.2 The Real and Imaginary Components of the Dielectric
Functions of Copper and Silver 29
3.3 A Schematic Diagram Illustrating the Difference
between the Normal Raman and the SERS Processes 31
3.4 A Plot Showing the Relative SERS Enhancement for
Silver Spheroids as a Function of Polar Angle . 34
3.5 A Plot Showing the Effect of Separation Distance on
the SERS Enhancement for Silver Spheroids ... 36
4.1 Calculated Extinction Profiles for 20 nm spheres of
Copper and Silver in Water 44
5.1 Normal Raman Spectra of the Sulfa Drugs as
KBr Pellets 55
5.2 SERS Spectra of the Sulfa Drugs on
Silver Colloids 58
5.3 The Structures of the Sulfa Drugs 62
5.4 The SERS Spectrum of Sulfamethazine at 1, 10,
and 100 ppb Levels 64
6.1 A Schematic Diagram of the Experimental Layout of
the Optical Components Used in the Copper Colloid
Studies 71
6.2 Normal Raman Scattering for Pyridine, SERS Spectrum
of 10 ppm Pyridine on a Copper Colloid, and the
Background of the Copper Colloid 75
6.3 Relative SERS Intensity of the Most Intense Band of
the Analyte as a Function of pH 77
vi

6.4 Absorption Scans of the Copper Colloid as it
Aggregates over Time with PABA 80
6.5 A Photon Correlation Spectrum of the
Copper Colloid 82
6.6 Intensities of the Two Most Intense Bands of PABA as
a Function of Time 84
6.7 SERS Spectrum of 50 ppm PABA on a Copper Colloid
and the Background of the Copper Colloid .... 86
6.8 SERS Spectrum of 50 ppm PNBA on a Copper Colloid
and the Background of the Copper Colloid .... 88
6.9 SERS Signal as a Function of Analyte Concentration
for Pyridine, PABA, and PNBA 90
7.1 Schematic Diagram of the FIA Setup used to Produce
the Copper Colloids 98
7.2 The SERS Spectrum of PABA of a FIA Produced Copper
Colloid 103
7.3 The SERS-FIA Signals at 1397 cm1 from a Series of
200 microgram injections of PABA into the System 105
7.4 SERS Signal as a Function of Amount of
PABA Injected 107
A.1 a). A Cross-section of a Prolate Spheroid
Illustrating the Elliptical Coordinate System,
b). An Illustration of the Model used for the
Derivation in the Appendix 119
vii

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
SURFACE ENHANCED RAMAN SPECTROSCOPY ON METALLIC COLLOIDS
FOR THE PURPOSE OF TRACE ANALYSIS
BY
MARTIN JOHN ANGEBRANNDT
August, 1991
Chairman: Dr. James D. Winefordner
Major Department: Chemistry
The use of Surface Enhanced Raman Spectroscopy (SERS) on
silver and copper colloidal surfaces is discussed.
Qualitative and quantitative results in the determination of
three sulfapyrimidine drugs on silver surfaces are presented.
It is demonstrated that it is possible to differentiate
between these drugs at trace levels. Limits of detection in
the low ppb range were obtained with a linear dynamic range
extending for approximately two orders of magnitude above this
for the 1114 cm 1 band.
The analytical figures of merit of copper colloidal
surfaces are evaluated for the first time. These copper
viii

colloids are produced without the addition of protecting
agents. The only previous preparation used citrate as an
antiaggregant. These colloids are less stable than silver
colloids, but will have fewer interferences from fluorescent
solution species. Analytical results for p-aminobenzoic acid,
p-nitrobenzoic acid, and pyridine adsorbed onto a batch
produced colloid are presented. The detection limits are in
the low ppm range. This is comparable with previous results
for silver colloidal surfaces.
A Flow Injection Analysis (FIA) system was constructed to
evaluate the possibility of adapting copper colloid based-SERS
to on-line analysis. It was possible to continuously produce
the colloid and analyze a sample injected into the flowing
stream. Para-aminobenzoic acid was used as the test analyte.
The detection limit was found to be 150 jug. The major
obstacle in this analysis is the hydrogen gas produced as a
byproduct of the reduction. The gas forms bubbles in the line
and must be removed prior to the spectroscopic measurement.
The method used to remove the hydrogen gas is just adeguate,
and an improvement in this aspect of the analysis would
increase the reproducibility significantly.
ix

CHAPTER 1
INTRODUCTION
Raman Spectroscopy provides information which cannot be
obtained easily by other means [1]. It allows the researcher
to probe the symmetric vibrations of a molecule. This
information is complementary to that gleaned using infrared
spectroscopy. The primary reason that Raman spectroscopy can
be such a powerful tool for analytical analysis is that all
molecules possess some vibrations which are Raman active.
Therefore, this method can be applied when the various
luminescence spectroscopies cannot. The Raman spectrum also
contains more information than a luminescence spectrum, enough
to identify the molecule.
The major disadvantage of Raman spectroscopy is the low
quantum efficiency of the process [2], Since only a very few,
approximately 1 in 10+6, of the photons are Raman scattered,
an intense excitation source is required to observe a signal
within a reasonable period of time. The signal observed is
located close to the excitation wavelength requiring the
excitation source to be also spectrally narrow, as well as
intense. A laser fits both of these qualifications and is
responsible for the increased interest in Raman Spectroscopy
since the invention of the laser in 1964. A double or triple
1

2
monochromator or a notch filter is needed to reject the
intense laser radiation, so that small Raman signals near the
laser line can be observed.
Because a Raman signal can be generated by any laser
wavelength, the researcher can choose the spectral region he
wishes to work in [3]. The ability to monitor the signal in
the visible or ultraviolet regions allows vibrational
spectroscopy to be carried out in aqueous solution, which is
difficult with infrared absorption. It also allows one to
choose a portion of the spectrum which is free from other
spectral interferences, such as fluorescence from solution
species [4]. Equipment already existing in the laboratory for
other types of analysis, such as luminescence, can be adapted
to monitor Raman signals. It is also relatively easy to probe
vibrations below 600 cm'1, which is difficult with IR
absorption spectroscopy because of the poor sources, optics,
and detectors available for the far infrared region.
One of the methods for increasing the quantum efficiency
of the Raman process is to adsorb the molecule of interest
onto a surface [5], Surface Enhanced Raman Spectroscopy
(SERS) involves both electromagnetic enhancements in the local
field and the chemical enhancements in the molecular
polarizability to increase the intensity of the Raman scatter.
Several surfaces, most notably gold, silver and copper,
provide enhancements of up to six orders of magnitude. This
enhancement over the Raman scatter from the free molecule

3
opens the door for trace analysis using Raman spectroscopy.
The nature of the enhancement mechanism is dependent on the
orientation of the molecule relative to the surface to be
probed. The enhancement and the orientational information
combine to yield a probe of startling spectral and spatial
resolution. A limitation of this method is the necessity of
close interaction between the molecule and the surface. This
limits the analytical application of this technique to
molecules with functional groups containing nitrogen, sulfur,
oxygen and phosphorus.

CHAPTER 2
OVERVIEW OF PREVIOUS WORK
SERS on Electrochemical Surfaces
SERS was first observed on a silver electrode by
Fleishmann et al. [6] who were interested in using Raman
spectroscopy to study species adsorbed at the electrode
surface. Intense Raman scattering was observed from pyridine
adsorbed at the roughened silver electrode surface. They
attributed the high intensity of the Raman scatter from
pyridine to an increase in the number of molecules able to be
adsorbed onto the roughened surface.
Later Fleishmann et al. [7] investigated pyridine
adsorbed onto a copper electrode, but failed to see an
increase in the Raman scatter because they used 514.5 nm for
the excitation when red wavelengths were required. They did,
however, observe an increase in the adsorption of pyridine to
the electrode surface with increased chloride concentration,
the subject of many later studies.
SERS was later observed on copper and gold electrode
surfaces by Wenning et al. [8,9]. They reported that SERS was
only observed on gold and copper surfaces when irradiated by
"red" wavelengths. The observed enhancement in the intensity
of Raman scatter from pyridine was five orders of magnitude
4

5
for a copper electrode surface and four orders for a gold
electrode surface. These studies included measuring the
angular dependence of the SERS signals on electrode surfaces.
The optimum collection angle for SERS measurements was found
to be 55° from the normal of the surface.
Jeanmarie and Van Duyne [10] and Albrecht and Creighton
[11] independently investigated the silver electrode-pyridine
system. Interpreting the results within the framework of
experiments by Barradas and Conway [12] on the adsorption of
pyridine onto metal surfaces, they concluded that the increase
in the intensity of the Raman scatter was not due to increased
adsorption. Both groups based this on the fact that a single
roughening cycle of the electrode yielded a greater
enhancement than multiple roughening cycles. The increase in
the Raman scattering was correctly attributed to an increase
in the Raman cross-section due to the adjacent rough metal
surface, rather than an increase in number of molecules
absorbed onto the surface because of the increased surface
area of the electrode.
Analytical Applications
Carrabba et al. [13] evaluated the potential of SERS on
silver electrodes for determinations of organic water
contaminants. All of the experiments in this study were
performed in low ionic strength solutions to simulate in situ
groundwater conditions. No interference was observed from
humic acid materials, which would be present under field

6
conditions. The silver electrode potential controlled the
adsorption of pyridine and quinoline. Variation of the
potential allowed discrimination between these two compounds.
The detection limit for pyridine was estimated to be 8.5 pg.
Forcé [14] reported the first analytical application of
SERS at an electrode surface to a flowing system. The silver
electrode underwent a short Oxidation-Reduction Cycle (ORC)
prior to sample introduction. The electrode potential was
then adjusted so that the pyridine would adsorb onto the
surface. After the spectrum was obtained, the potential was
adjusted to desorb the pyridine and another short ORC
performed. This approach obviously can only be used for
analytes which adsorb reversibly onto the surface. A detection
limit of 250 nmol was obtained with linearity extending for
three orders of magnitude above the detection limit. Pothier
and Forcé [15] have also used this system for the analysis of
adenine, thymine and cytosine. They obtained detection limits
of 175, 233, and 211 pg, respectively, for these compounds,
using a multichannel analyzer. The linear dynamic range
extended four orders of magnitude above these values.
Cotton et al. [16] combined SERS with a resonance
enhancement performing SERRS on a silver electrode surface.
This technique was used to detect four nitrophenol compounds
previously separated by HPLC. The eluent was collected by a
fraction collector, and the spectra measured off-line using a

7
silver electrode surface. Detection limits of approximately
20 ppb were reported.
Physical Studies
The SERS spectrum of pyridine on copper, silver, and gold
electrode surfaces has been the subject of many
investigations. The excitation profile of the pyridine
breathing mode for each of these surfaces was reported and
discussed in terms of a chemical enhancement mechanism [17].
Changes in the relative intensities, frequency shifts and
increases in the half-width were indicative of metal-adsorbate
complexation. Increases in the intensity of the scattering
were observed when chloride ions were present, indicating that
chloride ions mediated the surface complex formation. They
also suggested that most of the surface features are smaller
than 5 nm, since a stronger oxidation-reduction cycle (ORC)
would be required for the formation of larger surface
features.
The pH dependence of the SERS signal of pyridine adsorbed
onto a copper electrode was measured by Cooney and Mernagh
[18], They reported that an ORC was not necessary for the
observation of SERS from a copper electrode, as it was with
silver and gold surfaces. They attributed this to the strong
coordination of pyridine with the surface. The optimum pH for
the observation of SERS from the pyridine copper electrode
system was found to be « 7 . Evidence of laser damage to the
electrode surface by the krypton ion laser also was discussed.

8
Crookell et al. [19] used an FT-IR spectrometer to
measure SERS in the near infrared region. The 1064 nm line of
a Nd-YAG laser was used to irradiate copper, silver and gold
electrode surfaces, with two narrow bandpass filters being
used to excluded the laser scatter. SERS signals were
observed from pyridine adsorbed on these surfaces. The
potential dependence of the SERS signal for the pyridine-
silver electrode system from -0.1 to -0.9 V vs. SCE was
reported. It was noted that this might be an ideal region,
since fluorescence was not often a problem, and it was still
possible to obtain high quality Raman spectra from aqueous
systems.
The SERS spectra of the adenosine and two of its
derivatives on a silver electrode surface were reported by
Koglin et al. [20]. They found indications that the n bonds
of the ring systems was able to interact with the surface,
giving rise to SERS. It was noted that the presence of (P03)2
groups in a molecule significantly increased the Raman
intensity, due to their strong association with the surface.
The SERS spectra of adenine, guanine, thymine, and
cytosine on silver electrodes were reported and compared by
the same group [21]. Differentiation between these DNA bases
is possible from the frequencies of the breathing modes alone.
A partial assignment of the bands was also presented. A more
complete assignation of the bands arising from these DNA bases

9
and a discussion of the orientation of these molecules on the
surface was given by Otto et al. [22].
SERS on Colloidal Metals
SERS on colloidal metal surfaces was first observed in
1979. Creighton et al. [23] reported intense Raman scattering
from pyridine on colloidal silver and gold surfaces. The
excitation profiles for pyridine on both of these surfaces was
reported. A correlation between the extinction profiles of
the hydrosols, and the excitation profiles of pyridine on
these surface was shown. This was the first paper to assert
that the SERS process involves the excitation of surface
plasmons. Both of these colloids were prepared using a
borohydride reduction.
Creighton et al. [24] were also the first to observe SERS
with copper colloids. A borohydride reduction was again
used, but citrate was added to stabilize the colloid. The
SERS spectrum of pyridine, diphenyl sulfide, thiophenol and
4,4'-bipyridyl on this surface were reported. In a mixture of
pyridine and another of the analytes, competitive adsorption
was observed at the surface. A Raman excitation profile
similar to that of gold was obtained for pyridine. The
maximum enhancement observed was calculated to be 1.5 X 10+5,
50% of that of that for a similar gold hydrosol.

10
Analytical Applications
Torres [25] was first to use silver colloids for
analytical trace analysis. SERS spectra for p-aminobenzoic
acid, phenytoin, 2-aminofluorene, as well as uracil and some
of its derivatives, were reported. The effects of different
preparations on the analytical figures of merit of 2-
aminofluorene, and p-aminobenzoic acid were also studied. All
of the colloids were produced in batches (static conditions).
A later study by Morris et al. [26] used sedimentation to
fractionate silver colloids prior to the analysis. Crystal
violet was used as the analyte to evaluate the SERS activity
of these fractions. The linear dynamic range for the fraction
yielding the greatest enhancement was between 1 pg and 1 ng.
This fraction was only partially aggregated. The use of this
fractionation technique increases the amount of SERS active
surface available for the analyte to adsorb onto relative to
the total surface area of the colloid.
The use of photoreduced silver as a SERS substrate was
evaluated by Ahern et al. [27]. The silver metal was produced
via a photoreduction of silver nitrate. The spectra were good
relative to chemically produced colloids, exhibiting better
signal to noise characteristics than the silver hydrosols
produced in that study. It was not possible to observe SERS
from species which are highly soluble in water. There was
some problem with preparing a narrow distribution of particle
sizes.

11
Silver colloids were used in the trace analysis of sulfa
drugs by Sutherland et al. [28]. Limits of detection were in
the low ppb range, with the added ability to qualitatively
differentiate between sulfadiazine, sulfamerazine, and
sulfamethazine.
Laserna et al. [29] first applied colloidal SERS to Flow
Injection Analysis (FIA). For para-aminobenzoic acid (PABA),
the limit of detection was in the ppb, while a 5% RSD was
obtained. Their research demonstrated that the variance in
colloidal SERS intensities was due mainly to variation in
colloid preparation. FIA allowed the aggregation of the
hydrosol, the mixing of the analyte and the colloid, and the
time of observation to be precisely controlled. The effect of
pH on the SERS signals of PABA was also investigated. The
SERS spectrum of 9-aminoacridine [30] was later evaluated to
determine the applicability of this technique to other
systems. The effects of the addition of various ions was also
investigated.
Pararosaniline was used as the analyte by Freeman et al.
[31] to test an adaptation of colloidal SERS to HPLC. In this
paper, a previously prepared silver hydrosol was mixed with
the eluent from an HPLC. A linear range between 0.1 and 50
ppm was reported for this dye.
FIA was used by Taylor et al. [32] to evaluate combined
resonance and surface enhancement in amplifying the Raman
signal (SERRS) for application to HPLC detection. Their study

12
used micro-Raman optics and multichannel detection, allowing
the collection of sufficient information for both qualitative
and quantitative analysis in an on-line situation. The
detection limit for a dye, crystal violet, was determined to
be on the order of 600 molecules.
Ni et al. [33] have also used an FIA system to test the
feasibility of using colloidal-SERS for the analysis of HPLC
eluents. In this study, they observed an increase in the SERS
signal from uracil with increased temperature. They also
measured the effect of pH on the SERS signals from uracil,
cytosine, adenine and guanine. The scattering from these
bases was highest under acidic conditions. Detection limits
in the 100 nmol range were reported.
Cotton's group [34] recently used a multichannel analyzer
to detect the SERS signals from four purine bases which had
been separated by HPLC. Adenine, guanine, xanthine and
hypoxanthine were separated using reverse-phase HPLC.
Injecting increased amounts of analyte increased the width of
the peak, but not the height. Detection limits in the sub¬
nanomolar range were reported.
Physical Studies
Silver hydrosols have also been used, by Heard et al.
[35], to study competitive adsorption onto a surface.
Competition was observed between pyridine and citrate for the
silver surface. It was found that chloride assisted the
pyridine in displacing the more strongly adsorbed citrate from

13
the surface. The observation of this competition using SERS
added credence to the chemical enhancement theories, since it
indicated that surface complex formation was important. The
addition of a stabilizer to the colloid before adding the
pyridine reduced the effect of the chloride.
Joo et al. [36] showed that it was possible to follow
surface induced reactions using SERS. Aromatic mercaptans,
which were chemisorbed onto the colloidal silver surface,
underwent dissociation. This reaction appeared to be a
surface induced photoreaction resulting in a scission at the
S-C bond nearest the aliphatic group. The location of the
scission is opposite the result produced via pyrolysis of the
same molecule.
Colloidal silver has also been used, by Garrell et al.
[37], as a SERS substrate in non-aqueous systems. Colloids
were prepared in a number of organic solvents including
acetonitrile and dimethylformamide. These two solvents proved
to be the most suitable for SERS analysis. The SERS spectrum
of t-butylamine on an acetonitrile silver sol was presented as
the first SERS spectrum from an anhydrous colloidal system.
Colloidal systems, such as these, may have use in non-aqueous
systems where electrochemical surfaces cannot always be used.
Several papers since the landmark paper of Creighton et
al. [23] have described the SERS signal from pyridine adsorbed
onto silver surfaces. In one study by Dawei et al. [38] a
silver chloride surface was used. The SERS activity of the

14
silver chloride colloid was seen to decrease with increased
exposure to light. The treatment of silver chloride colloids
with thiosulfate, ferricyanide, or hydrogen peroxide reduced
the SERS intensity of pyridine. Similar results were observed
for silver bromide colloids by the same group [39],
The SERS spectra of halide ions associated with the
surface were measured on a silver colloidal surface by Wetzel
et al. [40]. Chloride was found to enhance the signal of
pyridine on both gold and silver colloidal surfaces. The
addition of pyridine did not increase the intensity of the
SERS from the halide ions. The increase in the pyridine SERS
signal was attributed to the formation of a surface complex by
the pyridine which was mediated by the halide ions.
The SERS spectra of cyclohexene and pyridine on silver
colloids were measured by Yamada et al. [41]. The SERS
intensity from pyridine and cyclohexene was compared with the
intensity of the SERS active bands of water. The strong SERS
enhancement for these chemisorbed species, relative to water,
along with shifts in both the frequencies and the relative
intensities of the bands, were seen as supporting the chemical
enhancement mechanism.
Moskovits at. al. [42] observed that an increased
concentration of molecules at the surface could produce a
change in the orientation of the molecule relative to the
surface. At low coverages, 2-naphthoic acid seemed to lie
flat on silver colloidal surfaces, interacting through both

15
the n bonds of the ring system and the carboxylate group. At
higher concentrations, 2-naphthoic acid was oriented
perpendicular to the surface, adsorbed exclusively via the
carboxylate group.
The effect of pH on the adsorption of 2,2'-bipyridine
onto silver colloids has also been studied. Kim et al. [43]
observed that decreasing the pH using HBr was found to
increase the SERS signal from 2,2'-bipyridine. An opposite
trend was observed when HI and HC1 were used. Any change in
the pH caused a shift in the SERS band positions.
The SERS spectra of several amino acids and nucleotide
bases adsorbed onto colloidal silver surfaces were studied by
Suh et al. [44] to determine their orientation on the surface.
The majority of these molecules were found to lie flat on the
surface. An exception was noted for m-aminobenzoic acid,
which at low pH adopted a geometry of either standing up, or
tilting away, from the surface with adsorption occurring
exclusively via the carboxylate group.
Adenine and eight of its derivatives were investigated by
Kim et al. [45] using silver hydrosols. The effects of pH
and analyte concentration on the resulting spectra were
reported. Changes in the pH were affected by the use of
phosphoric acid. Acidic conditions were necessary for SERS to
be observed from these DNA bases, except when pyridine had
been added. This demonstrated that the addition of acid, or
some other adsorbate, might be necessary to induce aggregation

16
of the silver colloid. Of this series, only adenine was
observed to undergo a coverage dependant change in
orientation. A similar study was undertaken by the same group
for uracil and three of its derivatives [46]. Like the
adenine derivatives, these compounds preferred to lie flat on
the surface under the conditions studied. Uracil adsorbs end
on in the deprotonated form, but assumed a flat orientation
when additional sodium borohydride was added. The change in
orientation was attributed to the decrease in the surface
potential caused by the addition of the sodium borohydride.
The SERS signals from citrate adsorbed onto a slightly
aggregated silver hydrosol and a highly aggregated silver
hydrosol were compared by Blatchford et al. [47]. Both of
these silver hydrosols were protected using citrate to prevent
further aggregation. When additional sodium borohydride was
added, the SERS signal observed from the citrate decreased.
The addition of the sodium borohydride served to decrease the
potential of the surface.
The concentration and temperature dependence of the SERS
spectrum of PABA on a silver colloidal surface was studied by
Suh et al. [48]. The intensity of the 1452 cm 1 band was shown
to be linearly dependent on the PABA concentration. The
system was studied at pH 7, making the anion the predominant
species. The molecule was determined to be adsorbed flat on
the surface. A similar conclusion was reached for PABA
adsorbed onto a silver electrode surface by Park et al. [49].

17
SERS on Chromatographic Substrates
In 1984, Tran [50] reported the first SERS spectra from
a chromatographic surface. He reported the SERS spectra of
four dyes adsorbed onto a cellulose filter paper. Limits of
detection for these dyes ranged from 0.5 ng for crystal violet
to 240 ng for methyl red. The analytes first were mixed with
silver hydrosols and then introduced onto the surface for
analysis using a syringe. A helium-neon laser was used as the
excitation source. For the case of crystal violet, the
surface enhancement was combined with a resonance Raman
enhancement (SERRS). This additional enhancement helped to
account for its much lower detection limit.
Later that year, Tran [51] performed a similar study
after first separating three dyes using paper chromatography.
These chromatograms were dried and then sprayed with a silver
hydrosol using a nebulizer. The SERRS spectrum of crystal
violet, malachite green, and basic fuchsine were evaluated on
two commercially available chromatography papers, yielding
detection limits of approximately 2 ng for these dyes.
A study by Berthod et al. [52] followed, with an
analysis of the analytical application of silver colloids to
coated filter papers. The SERS spectra of PABA and several
other nitro and amino containing heterocycles were obtained,
but no quantitative results were given. Electron micrographs
indicated that the coated filter paper surface is not
conducive to theoretical modeling, because of the entrapment

18
of silver colloids within the filter paper, as well as upon
the surface.
Several nitro containing heterocycles were detected by
Laserna et al. [53] on a silver coated filter paper, both
alone and in mixtures. This paper also demonstrated that the
components of a mixture could be spectroscopically resolved
using SERS. Detection limits in the low ppm range were
achieved with an RSD of 15%.
The first SERS spectrum from a silica gel plate was
obtained by Séquaris and Koglin [54]. They reported the SERS
spectrum of 9-methylguanine using 514.5 nm excitation. The
analyte first was mixed with the colloid and then applied to
the surface. The detection limit was determined to be 120 pg.
Séquaris and Koglin [55] later separated ten purine
derivatives using HPTLC; the plates were dried and then
sprayed with colloidal silver. Detection limits in this
latter study for these biological compounds were estimated to
be less then 5 ng/spot. Koglin [56] used a micro-Raman
instrument in a similar study to improve the spatial
resolution of this technique; detection limits were in the ng
range for purine, benzoic acid, and 1-nitropyrene, but the
spatial resolution was improved to 1 /um.
In addition to the work mentioned above, there have been
several reviews. A book on SERS, edited by Chang and Furtak
[57], was published in 1982 and offers the best introduction
to the field. A review by Seki [58] cites all references to

19
the SERS effect before 1985. Seki1s review provides a good
place to start a search from, but lacks text describing the
observed results. Chang [59] has also published a report on
the status of SERS research of electrode interfaces. Cotton
[60] has recently published a review of SERS investigations
of compounds of biological interest.

CHAPTER 3
THEORETICAL FOUNDATIONS
The Raman Scattering Process
Raman scattering arises from the inelastic scattering of
photons by molecular vibrations [61]. Raman scatter is
dependent upon a change in the molecular polarizability during
a vibration, whereas a change in dipole moment is required for
infrared absorption. The Raman spectra reflect vibrational
and rotational motions, which cause changes in the molecular
polarizability. The experimental data within this
dissertation only involves vibrational transitions, therefore
only these will be discussed in this chapter and the appendix.
For a light of a frequency, i;L (s 1) , the electric field,
E, (V/m) of the wave can be expressed as
E = E0cos (277vLt) (3.1)
where E0, (V/m) is the amplitude of the wave. For a vibration
at frequency, (s’1) , the polarizability at any time, t, will
be
a(t) = a0(l+cos(27n/vt))
(3.2)
20

21
where a (C2*m2/J) , is the molecular polarizability. Since the
electric field induces a dipole moment, n (C*m), in the
molecule through the molecular polarizability
M = a-E (3.3)
the time dependence of the induced dipole moment is given by
the supposition of the equations 3.1 and 3.2
M = a0E0cos(27Ti/Lt)+ (1/2) a0E0cos (2tt (i/l±j/v) t) (3.4)
The first term corresponds to the light scattered elastically
by the molecule (Rayleigh Scatter), the second contains the
anti-Stokes and Stokes Raman (inelastic) scattering terms.
The anti-Stokes term is represented by the summation of the
frequency of the laser and the frequency of the vibration.
The Stokes term is shown as the difference between these
frequencies. The physical result is the appearance of bands
on either side separated by Ai/y from the exciting frequency
(j/l) as shown in figure 3.1.
The oscillating electric field of the incident radiation
may, because of molecular symmetry, induce a dipole moment in
a molecule whose axis is not parallel to the applied field.

Figure 3.1. a) A schematic Raman spectrum showing the stokes
and anti-stokes Raman scattering shifted for the first
vibrational transition and, A^2, for the second vibrational
transition, relative to the Rayleigh scatter at vL. b) A
schematic diagram illustrating the origin of the Raman
scattered frequencies.

23
a)
LASER
LINE
ANTI-STOKES SHIFTED STOKES SHIFTED
b)
VIRTUAL LEVELS
LASER
LINE
LASER
LINE
MOLECULAR
VIBRATIONAL LEVELS
STOKES SHIFTED
2
1
0
ANTI-STOKES SHIFTED

24
The relationship between the induced dipole and the applied
field is best shown by the matrix form of the polarizability
tensor
X
1
=
a a
xx xy
^xz
E
X
My
=
a a
yx yy
ayz
Ey
Mz
=
OL QL
z x zy
azz_
3..
For spontaneous (normal) Raman scattering, the polarizability
matrix is real and symmetric, so the square matrix can be
diagonalized by choosing a principal axis. This attribute
allows one to calculate the magnitude and orientation of the
induced dipole.
The intensities of the Raman bands can be calculated
using the formula for the intensity of radiation, I, emitted
by an oscillating dipole.
I = (167tV/3c2)m2 (3.6)
where c, is the speed of light. If the Stokes shifted lines
originate from the ground vibrational level, then the anti-
Stokes lines must originate from higher vibrational levels.
Since the populations of these vibrational levels follows a
Boltzmann distribution, the ratio of the relative intensities
of the Stokes and anti-Stokes bands can be expressed by
Is/Ias = ((,/L+l/v)4/ (‘"l'^v)4) exp(-h*v/kT)
(3.7)

25
The selection rules for the harmonic oscillator allows only
A»/ =±1, although often overtones (2i/y) and combination bands
(i/y1 ± vw2) are found due to anharmonicity in the molecular
vibration [62]. The presence of a surface directly adjacent
to the molecule may lower its symmetry, resulting in the
appearance of bands that are Raman forbidden in the free
molecule [61,2], The selection rules then become "propensity
rules," with the z2 terms being the most intense. The z axis
extends outward into the surrounding medium, perpendicular to
the surface, and the x and y axes are parallel to the surface.
The xz and yz terms would be next most intense, and the x2, y2,
and xy terms would be the weakest.
The SERS Enhancement Process
The forces acting on a molecule located at a distance, r,
from a surface may have a variety of sources [63]. The most
important of these is the electric field of the laser. The
image field, which accounts for the polarization of the metal
by an induced dipole located near the surface, and the field
caused by the presence of induced dipoles located near the
molecule, may effect the absorption and resonance Raman
spectra of the molecule. The normal, non-resonant, Raman
intensity is only affected by the electric field of the laser.
Mr = (3.8)

26
where ¡ir (C• m) , is the induced dipole, ar (C2*m2/J) , is the
polarizability, EL(r,i/L) (V/m) , represents the field generated
by the laser at r, i/L (s 1) , the incident laser frequency and,
i/v (s 1) , the vibrational frequency of the mode.
From this equation, one can see the beginnings of the two
classifications of theories concerning the evolution of the
SERS effect. The first enhancement mechanism is termed
electromagnetic. This mechanism attributes the large
enhancements to local field enhancements due to either
resonances within the surface or points of high curvature.
This would result in a larger field term, EL(r,i/L) . The second
enhancement mechanism involves chemical effects; chemical
effects are due to direct interactions between the molecule
and the surface, which cause changes in the polarizability of
the molecule, ar.
Electromagnetic Enhancement Theory
For the calculation of purely electromagnetic
enhancements there are three requirements. First, one must be
able to represent adequately the polarization of the molecule,
usually as a point-dipole. This condition is a "far field"
approximation, where the distance between the separation of
charge is negligible when compared with the distance between
the molecule and the surface. Second, the surface must have
a relatively simple shape, so that it may be easily modeled.
Third, Maxwell's equations must hold even at distances as
small as 1 k. The last is needed because the field felt by

27
the dipole as it nears an irregular surface becomes
increasingly inhomogeneous, causing the model to break down.
On smooth surfaces, the enhancement is observed to be
approximately 10-102 [64]. The enhancement from the smooth
surface should decay rapidly on an atomic scale as the
molecule is moved away from the surface. Spherical surfaces
are easy to model and able to couple directly with optical
fields [65].
The local field, Eloc, caused by a surface located at the
center of the sphere is given by the equations
Eioc “ [(eK)-eo)/eK)+2eo] a3 (3.9)
where a is the diameter of the sphere, e(i/) is the dielectric
function of the metal, and e0 is the dielectric function of
the surrounding medium. Each of these terms is composed of a
real, e1f and imaginary, e2, component as shown in equation
below
e = + ie2 (3.10)
If the frequency of the laser is near the resonance frequency
of a sphere, the enhancement is inversely proportional to
square of the imaginary component of the dielectric function
of the metal surface, [e2m]+2'
at that frequency.

28
It is important to have a small imaginary component for
the dielectric constant when the resonance condition,
[ei(^)] = 2e0 is satisfied. Figure 3.2 shows the real and
imaginary components of the dielectric functions of bulk
copper and silver in a vacuum. Figure 3.2 can be used to
determine at which frequencies the surface plasmons of the
metal can couple with the incident radiation. The surface
states of a sphere can be in resonance with not only the
incident radiation, but also the scattered radiation. This
resonance condition allows the Raman scatter of the molecule
to be amplified [4]. Figure 3.3 schematically compares the
spontaneous (normal) Raman and SERS processes.
Prolate spheroids can be used to theoretically model
colloids which have undergone agglomeration. The
electromagnetic equations are similar to those for a sphere,
but illustrate the difference between of the local field at
the end and the sides of the spheroid. For the resonance
condition (e(i/L) = -2e0) , the local field at the tip of the
spheroid is enhanced by, G, the same amount as a sphere
G =|£-1/e2 | 2 (3.11)
The enhancement on the side of the spheroid will be reduced
. 2 •
from this by a factor of e . The tip also allows the
efficient coupling of the scattered Raman radiation with the

Figure 3.2. The real, e,, (solid symbols) and imaginary, e2,
(open symbols) components of the dielectric functions of
copper (circles) and silver (triangles) [71],

energy (eV)
Magnitude of the Dielectric Function
_,0 ->â–  K) W -F cn O' vi
O

Figure 3.3. A schematic diagram illustrating the difference between the normal Raman
scattering process and the SERS process. The incident frequency is denoted by v, while
the Raman shifted frequency is represented by v1. The A terms represent the
amplification of the incident field, E, by the spheroid.

u>
to

33
surface, so that if i/R » wL, then the scattered radiation
undergoes additional amplification.
If this occurs, a molecule interacting with the spheroid tip
will experience an enhancement of
G = |e-l/e21 4 (3.12)
... 4
with this being reduced by a factor of e for the sides of the
spheroid. This enhancement arises from the small radius of
curvature of the tip of the spheroid, not the aspect ratio of
the spheroid [66], The basis for the increased field at the
tip of the spheroid is that in the maintenance of an
equipotential surface, the image charge would need to be close
to a point of high curvature. This translates to a more
intense field outside that portion of the surface. Figure 3.4
illustrates the variation in the enhancement factor for a
prolate spheroid as a function of polar angle.
The electomagnetic model accounts for the long range
enhancement observed by Murray et al. [67]. The results from
calculations of the distance dependence of the enhancement
from a prolate spheroid are shown in figure 3.5.
If the enhancement was purely electromagnetic, then the
Raman excitation spectrum would be dependent only on the
nature of the metal and the vibrational frequency of interest
[63], This type of behavior has been observed on gold and

Figure 3.4. A plot showing the calculated enhancement for SERS on silver spheroids as
a function of polar angle from the center of the silver spheroid. 0 degrees is the tip
of the prolate spheroid and 90 degrees is the waist. 1, 2, and 3 correspond to the
eccentricity of the spheroid (a/b), where the minor axis, b, is held constant at 20 nm
[65].

Polar Angle (degrees)
G. Enhancement Factor (arb. units)
iQ
O
1000

Figure 3.5. A plot showing the calculated effect of separation distance on the SERS
enhancement for silver spheroids. 1,2, and 3 denote the eccentricity of the spheroid
(a/b), where the minor axis, b, is held constant at 20 nm [65].

Separation Distance (Á)
G. Enhancement Factor (arb. units)
O ro -F cr> co
-F
LZ

38
silver surfaces for water, but not other adsorbates [59].
The observed correlation between the extinction profiles and
elastic scattering excitation profiles of non-aggregated
colloids can be accounted for by the electromagnetic theory.
The electromagnetic theory cannot account for differences in
the Raman excitation profiles for various molecules, since it
does not handle chemical effects [68]. Contrary to this is
the success of Weitz el. al. [69] in calculating the SERS
excitation spectrum of nitrobenzoate on silver island films
using the electomagnetic model.
Critics of the electromagnetic theory consistently point
out that it does not account for changes in the selection
rules and over-estimates the enhancement [70]. This tendency
to over-estimate the enhancement may be due to the use of the
bulk phase dielectric constants [65]. The use of the bulk
phase dielectric constants has been shown to be valid only for
cases where the length or diameter of the particle is less
than 1/15 of the exciting wavelength [4], This breakdown been
observed in the measurement of the dielectric functions of
thin metal films [71],
Chemical Enhancement Theories
Chemical enhancement mechanisms require strong
interactions between the molecule and the surface. These
interactions result in an alteration of the polarizability
matrix elements of the molecule. This alteration is
attributed to either charge transfer [72,73], excitations of

39
electron hole pairs requiring "adatom" type surface defects,
or chemisorption induced resonance [74]. All of these models
are short range mechanisms, predicting no enhancement past the
first one or two monolayers adsorbed onto the surface.
This appears to be the case for an experiment involving
cyanide and pyridine [75] . The SERS bands from these
adsorbates were enhanced by up to six orders of magnitude.
The spectra were highly potential dependent, exhibiting SERS
only for the molecules that were strongly adsorbed to the
electrode surface. Competitive adsorption, similar to this,
has been observed on silver [35] and copper [24] colloids as
well.
Of these mechanisms, only the "adatom" model of Otto et
al. [76] has stood the test of time well. The charge transfer
and chemisorption induced resonance models do not specify the
necessity of active sites [4]. The enhancement from smooth
surfaces is small and can be accounted for by electromagnetic
mechanisms without the need to invoke the chemical models
[70]. The small enhancement observed on smooth surfaces is
commonly used to discount the effects of chemical enhancement
mechanisms.
The adatom model employs surface defects to break the
translational symmetry parallel to the surface. These
variations in surface morphology mediate the coupling between
photons and plasmons. This reduction in the symmetry of the
surface is also used in the electromagnetic model. The major

40
difference between the two theories on this particular point
is how far away the molecule can be away from the surface and
still feel a difference between a surface defect and the
planar surface. This is related to the relative degree of
roughness present in the surface. For the adatom model the
surface features are composed of several atoms or a defect in
the crystal structure, while the electromagnetic model uses
surface features on the order of 10-50 nm in the calculations.
Since surface defects mediate the photon plasmon
interactions, molecules adsorbed onto the surface at or near
these defects will be more easily excited. Coupling between
the electronic states of the molecule and these surface states
would account for both the breakdown of the selection rules
and the observed continuum background of inelastic scatter
[63].
Experiments in UHV conditions have examined the effect of
annealing surfaces on the SERS activity. In these
experiments, [77,78] Raman active surfaces were heated to
anneal them. Annealing the surface resulted in a loss in
Raman scattered intensity. In another experiment [59],
pyridine was adsorbed onto silver surfaces under UHV
conditions. For three low index faces, only one band
(992 cm 1) , attributed to physisorbed pyridine, was detected.
For a silver 540 face, the band at 992 cm 1 was shifted to 1003
cm 1 due to chemisorption.
In all cases this band was

41
unenhanced, which casts doubt on the importance of atomic
sized defects proposed in the adatom model.
There is also some data on colloids which is in conflict
with the adatom model's use of small surface defects. Cotton
et al. [34] have recently been using elevated temperatures (up
to 90° C) to increase the Raman scattering from molecules
adsorbed onto silver colloids. According to the adatom model
this heating should result in a decrease in the Raman scatter,
since much of the finer surface morphology would be smoothed
by the heating of the colloids, as annealing does to surfaces
in UHV conditions.
There is currently no theory which describes or explains
all of the SERS results. The previous section was compiled to
give a general overview of the competing theories. Under some
conditions, one theory may appear to adequately describe the
observed results. For aggregated colloidal systems, the
electromagnetic theory seems to best account for the observed
results. There are, however, indications that chemical
mechanisms do play some part. The electromagnetic model can
not account for shifts in the positions of the bands, or the
breakdown of the selection rules.

CHAPTER 4
COLLOIDS AND AGGREGATION
Metal colloids have been the focus of much attention.
The spectral properties of colloidal metals were first
quantitatively examined by Mie in 1908 [79]. Mie investigated
the scattering of light by gold particles of sizes
approximately the same size as the wavelength of light used
for excitation. His work formed the basis for further
experimental and theoretical work on what came to be known as
the Mie scattering phenomenon [80]. Absorption spectrometry
and light scattering techniques are still used to determine
the particle size and degree of polydispersity for colloidal
solutions.
The optical properties of colloidal spheres suspended in
aqueous solution can be described by the following parameters
Qext
= (2/a2) Z (2n+l) {Real (an+bn) }
(4.1)
Qsc
= (2/a2) Z (2n+l) { | an | 2 + |bn|2}
(4.2)
Qab
— Qext — Qsc
(4.3)
where Qext is the extinction efficiency, Qsc, is the scattering
efficiency, Qab, is the absorption efficiency, a, is a size
parameter equal to (2?ra/A) , and an and bn represent a series of
42

43
coefficients dependent on the size of the particle and the
dielectric constant of the surface relative to that of the
medium. For very small particles, with radii less than 1/15
of the incident wavelength, these equations become
Qsca = (8a4)/{3| (e-l)/(e+2) |2} (4.4)
Qab = -4a{Im{(e-l) / (e+2) }} (4.5)
where Im represents the imaginary components of the dielectric
function of the metal surface relative to the surrounding
medium, and Qext is the summation of equations 4.4 and 4.5.
These simplifications are only valid for the Rayleigh limit,
when contribution from multipoles higher than the dipole can
be neglected.
The above equations have been used to calculate the
extinction spectra of spherical particles of silver and copper
in water [81,24] shown in figure 4.1. Peaks in the extinction
spectra enable one to follow the aggregation of a colloidal
system using conventional absorption spectrometry. It may be
possible to observe several peaks which correspond to several
different particle sizes [23], or only a single peak for one
particle size [24].
Another method of particle sizing is based on scattering.
Photon Correlation Spectroscopy (PCS) uses a diffusion size
model to determine particle size. Corrections can easily be
made for the refractive index of the particle at the

Figure 4.1. Calculated extinction spectra of 20 nm spheres of silver [81] and copper
[24] in water.

Relative Absorbance
ui

46
scattering wavelength. Photon correlation spectroscopy yields
the distribution of the particle sizes, as well as, an average
particle size. One of the limitations of this method is that
the calculations usually assume a single mode distribution.
Electron microscopy can also be used to determine the range of
particle sizes, but care must be taken to avoid causing
further aggregation when preparing the sample for analysis
[82] .
The aggregation process is controlled by the surface
charge of the particle and the thickness of the diffuse layer
surrounding the particle [83], For colloidal particles, the
attraction is due mainly to van der Waals interactions. The
strength of these interactions vary as d 6 with the
interparticle separation distance. The attractive forces, Ua,
(J/mol) for a spherical particle of radius a, is given by
Ua(d) - --4- [ 8^ld+f° ♦ 21n ]
12 d{d+a) (cf+2a)2
(4.6)
where d is the distance between the surfaces, a is the radius
of the particle, and A is an empirically determined constant.
. -19 • •
A is usually on the order of 10 J. From this equation, one
can expect uncharged particles to associate and eventually
precipitate from solution.
Since the majority of colloids have charged surfaces,
there are also repulsive forces to consider. The charge on
the surface is usually due to the adsorption of ions onto the
surface. There is also a diffuse layer, which shields the

47
surface charges from each other. This controls how near
similarly charged particles can approach each other. The
extension of this layer into the surrounding solution is given
by
4>(d) - 4>Q(a/d) exp(a-d/Xa) (4.7)
where $ (V) is the potential of the surface and Xa (m) is a
distance parameter. The thickness of the diffuse layer is
related to the ionic strength of the solution through the
expression
a F21
(4.8)
where e is the dielectric constant, e0 is the permittivity of
free space, F (C/mol) is Faraday's constant, and I (mol/L) is
the ionic strength of the surrounding medium. From
expressions 4.7 and 4.8, we see that increasing the ionic
strength of the solution will compress the double layer,
resulting in increased aggregation. This accounts for the
"salting out" effect.
The repulsive forces, Ur, between two particles with ion
atmospheres surrounding them is difficult to calculate. The
trends observed follow the form
2na3 o2
Uz(d)
(1+a/xa)2 ee0
In (1-exp d/x*)
(4.9)

48
The total potential energy of the system is the sum of
equations 4.6 and 4.9.
There are several ways to compress the diffuse layer and
induce aggregation [84]. The first is to increase the ionic
strength of the solution, so the extension of the diffuse
layer into the surrounding solution is reduced, making
collisions more probable. Another method is to adsorb neutral
compounds onto the surface, replacing the ions adsorbed onto
the surface and thereby reducing the surface charge of the
particle.
Cotton et al. [34] recently used temperature increases to
induce aggregation. This technique increases the velocities
of the particles in solution, so that the diffuse layer is not
as effective at preventing collisions between particles. More
quantitative explanations of the effect of increasing the
ionic strength of the solution, of raising the temperature and
of decreasing the surface charge of the particle on the
thickness of the double layer surrounding the particle can be
found in a treatment using the Gouy-Chapman theory [85].
The kinetic aspects of working with colloids also need to
be considered. A paper by Weitz et al. [84] describes the
aggregation process in fractal terms. The basis for the
application of fractal analysis is the self-similarity of the
colloids on various scales. The fractal dimension of a gold
colloid was shown to be 1.75. The use of fractal scaling of
the relationship between time and the mean radius of the

49
particles, yielded a logarithmic plot with a slope of 0.57 or
1/1.75. The fractal approach to quantifying colloidal
aggregation seems promising since it is accurate and provides
results which are easily visualized.

CHAPTER 5
SILVER COLLOIDS FOR SULFA DRUG ANALYSIS
Introduction
The development of drugs for therapeutic use both in
humans and animals has increased dramatically over the last
+8
century. Over 10 kg of one particular class, sulfa drugs,
were administered in 1985. Of the 5000 sulfa drugs tested for
antibacterial activity, less than 30 are routinely used [86].
Sulfapyrimidines are an important subclass of these
sulfonamides (sulfa drugs). Sulfamethazine, sulfamerazine and
sulfadiazine are members of this group [87]. These drugs are
able to block the formation of dihydrofolic acid from p-
aminobenzoic acid, thereby inhibiting the biosynthesis of
folate cofactors in bacteria [88].
Sulfa drugs are not commonly prescribed to humans
anymore, because of the development of other antimicrobial
agents and the increased resistance of bacteria to these
drugs. Sulfa drugs are still widely used in veterinary
medicine, since they are easily administered. Often two or
more of these drugs are used in combination to produce the
necessary dosage. Crystalluria, the formation of crystals in
the renal tubules, can be caused by administering drugs which
have limited solubility. Since drugs of the same class do not
50

51
affect the solubility of each other strongly, the risk of
crystalluria is reduced.
The use of these drugs to treat livestock diseases has
resulted in the deposition of sulfa drug residues in the
tissue of swine. In 1973, the U.S. Food and Drug
Administration set a tolerance of 100 ppb for sulfonamides in
edible tissue. The current method for testing for this class
of compounds does not yield information about the identities
of the compounds found [86].
Most spectroscopic methods for determining sulfa drugs
involve diazotinization. The standard method of testing for
sulfa drugs is the Bratton and Marshall method [89], which
yields colored products by diazotinization; the colored
products can then be detected using absorption spectroscopy.
Later this procedure was modified by Iskander et al. [90] to
take advantage of the kinetics of the diazotinization to
differentiate between these compounds. This has some success
allowing differentiation between different classes of sulfa
drugs.
Diazotinization was used by Sato et al. [91] to allow
resonance Raman spectroscopy of these compounds to be measured
with excitation freguencies in the visible region. Detection
_ s
limits were estimated to be in the 10 M range.
Unfortunately, it was not possible to differentiate between
the various sulfa drugs spectrally.

52
Membrane electrodes were prepared by Yao et al. [92] and
used for the determination of Sulfa drugs. These membranes
were specific for sulfa drugs with detection limits in the /¿M
range but, were not able to differentiate between molecules of
the same class.
SERS on colloidal silver substrates will be shown to
provide qualitative and quantitative information for
sulfadiazine, sulfamethazine and sulfamerazine at levels in
the low ppb range.
Experimental
Apparatus
All of the spectra were obtained by irradiating the
samples with 100 mW of 514.5 nm line of an argon ion laser
(Spectra Physics, series 2000). The scattered radiation was
collected at 90 degrees and resolved using a 0.85 m double
monochromator (Spex Industries, model 1403) with the bandpass
set at 10 cm 1. The detector was a cooled photomultiplier tube
(RCA, model C31034) coupled to photon counting electronics.
The acquisition, storage, and processing of the data was
carried out on an IBM compatible PC. All spectra presented
are single scans without any spectral smoothing.
Chemicals and Procedure
All chemicals were reagent grade or better and used
without further purification. Demineralized water was used
throughout the analyses. The sulfa drugs analyzed,
sulfadiazine, sulfamethazine, and sulfamerazine,
were

53
purchased from Sigma Chemicals. The 100 ppm stock aqueous
solutions were prepared with 3% methanol to aid in
dissolution.
The silver colloids were prepared as follows: 100 ml of
an ice cold 1.75 X 10 3 M solution of AgN03 was added dropwise,
with stirring, to 300 ml of an ice cold 7.2 X 10 4 M NaBH4
solution. An absorption spectrum of the resulting hydrosol
showed a narrow band with a maxima at 400 nm. Such a maxima
is indicative of a relatively monodisperse hydrosol containing
silver particle between 1 and 50 nm in diameter. The particle
sizes were also measured by filtration through polycarbonate
track-etched membranes, followed by SEM analysis. Each sample
analyzed was prepared by adding 0.1 ml of the analyte solution
to 0.9 ml of the room temperature colloidal solution. After
mixing, 0.5 ml of this solution was transferred to the sample
cell and analyzed.
Results and Discussion
The normal Raman and SERS spectra of sulfadiazine,
sulfamerazine and sulfamethazine are shown in figures 5.1 and
5.2 respectively. The shifts of the peaks are given in tables
5.1 and 5.2 in units of reciprocal centimeters. There are
several peaks in the normal Raman spectra, which are present
for all three of the compounds. The bands located near 1596,
1506 and 1094 cm 1 in the these spectra are due to stretches
in the ring systems. The strong bands near 1150 cm 1 are due
to symmetric vibration of the S02 group; these bands are

54
accompanied by weaker vibrations near 1300 and 543 cm1,
attributed to the same group. The shoulders found in these
spectra at about 1635 cm 1 are commonly attributed to the amine
group. A similar feature found at 745 cm 1 is also attributed
to the amide.
Comparing these frequencies and relative intensities with
those found in the SERS spectra leads one to think that the
molecules are lying flat on the surface. The S02 bands are
shifted from 1150 cm1 to 1114 cm1, and are still among the
most intense in the spectrum. The other two modes appear near
550 and 1308 cm 1 for sulfamerazine and sulfamethazine, but not
-1 .
for sulfadiazine. The bands near 1635 cm found in
sulfamethazine and sulfamerazine are shifted to 1630 cm1.
This band appears in the SERS spectrum of sulfadiazine as
well at 1622 cm"1. The ring vibrations 1596, 1506, and 1094
cm 1 are reduced in relative intensity, but not shifted
significantly. The large shifts in the vibrational
frequencies, as well as, the relative intensities of the bands
from the S02 and amine groups indicate that they interact
strongly with the surface. From this information, the molecule
appears to lie nearly flat on the surface. This has
previously been suggested for molecules, such as DNA bases,
which also have more than one functional group.

Figure 5.1. The normal Raman spectra of the three sulfa drugs prepared as KBr pellets.
The spectrum a) sulfadiazine; b) sulfamerazine; c) sulfamethazine. The peak positions
and relative intensities are given in table 5.1.

200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0 1600.0 1800.0
Raman Shift (cm—1)
Raman Scattered Intensity (arb. units)
99

57
Table 5.1.
Raman Bands of ^ulfa Drugs (in Solid Phase)
Raman Shift (cm )
sulfadiazine
sulfamethazine
sulfamerazine
229, sh
237, w
248,vw
286, w
273, vw
282, dbl,w
290, vw
319, vw
300, dbl,w
290, vw
339, vw
328, w
330, vw
373, vw
342, vw
365, w
380, w
386, vw
402, vw
455, vw
456, w
439, w
504, w
529,vw
544, w
547, m
538, w
578, vw
585, w
567, vw
579, w
636, w
634, w
637, w
661, vw
659, vw
681, w
685, w
706, vw
715, vw
716, vw
750, vw
743, m
800, w
827, dbl,m
822, dbl,m
829, dbl,m
848, dbl,m
840, dbl,m
842, dbl,m
892, w
873, vw
939, vw
963, vw
995, m
999, m
999, m
1098, s
1094, S
1092, w
1150, vs
1153, VS
1144, vs
1259, W
1191, w
1241, vw
1312, vw
1298, w
1306, vw
1338, w
1333, w
1345, vw
1371, vw
1385, vw
1407, vw
1413, w
1416, vw
1438, vw
1440, vw
1505, w
1466, vw
1478, vw
1506, m
1509, w
1581, sh
1560, w
1559, vw
1598, vs
1596, vs
1596, vs
1629, w
1640, w
bb = broad band, sh = shoulder
, m = medium, s = strong,
w = weak, v =
very weak

Figure 5.2. The SERS spectra of the three sulfa drugs at 100 ppb concentrations in the
colloidal silver solution. a) sulfadiazine; b) sulfamerazine; c) sulfamethazine. The
peak positions and relative intensities are given in table 5.2 .

Raman Shift (cm- 1)
Raman Scattered Intensity (arb. units)
65

60
Table 5.2.
Surface Enhanced Raman Bands of Sulfa Drugs on Colloidal
Silver
Raman Shift (cm )
sulfadiazine
sulfamerazine
sulfamethazine
230,
sh
230,
sh
238,
sh
294,
sh
290,
sh
364,
vw
342,
vw
416,
vw
410,
vw
396,
vw
458,
vw
462,
vw
458,
vw
506,
vw
508,
vw
516,
vw
544,
vw
554,
w
574,
m
582,
m
590,
m
634,
w
626,
vw
638,
vw
684,
w
674,
w
674,
w
762,
s
738,
m
818,
s
830,
m
834,
m
974,
vw
976,
vw
968,
bb
1002,
vw
1068
, sh
1066,
sh
1054
/ w
1114
, vs
1114,
vs
1112
, vs
1186
, vw
1174,
vw
1186
, vw
1252,
vw
1260
, vw
1312,
m
1302
, vw
1340
, m
1356,
sh
1390,
m
1394
, m
1410
, w
1454,
w
1502
/ W
1502,
w
1506
/ w
1594
, S
1594,
s
1594
, s
1630
, sh
1630,
sh
1622
, sh
2930,
m
2922
, m
3050
, sh
3068,
sh
3074
, sh
3258
, bb
3258,
bb
3242
, bb
3378
, bb
3396,
bb
3380
, bb
bb =
broad band, sh
= shoulder
, m =
= medium, s = strong,
w = '
weak, v =
very

61
From the chemical structures shown in figure 5.3, the
similarity and differences between these compounds is evident.
The SERS spectra reflect these similarities and differences as
well. The 2930 cm 1 band, which is present with different
intensities for both sulfamethazine and sulfamerazine, is
absent in the sulfadiazine spectrum. The 2930 cm’1 band is due
to the methyl groups on the pyrimidine ring. The locations of
the stronger peaks also differ slightly. The symmetric
stretch of the S02 group is located at 1114 cm 1 for
sulfadiazine and sulfamerazine, but is found at 1112 cm 1 for
• • • • -1
sulfamethazine. Similarly, a shoulder is seen at 1622 cm for
sulfamethazine, but is shifted to 1630 cm 1 in the other
spectra. There are also a number of peaks which appear in one
of the spectra, but not the others.
Figure 5.4 illustrates the ability of SERS to provide
quantitative information at ppb levels. The bands at 685 cm’1
and 1092 cm 1 are attributed to vibrations of the benzene ring.
Variations in the intensities of these bands may be due to
slight oreintational changes of the molecule relative to the
surface. This presents a true test of the fingerprinting
capability of SERS. The intensity of the peak at 1114 cm'1 in
the SERS spectrum of sulfamethazine is useful for determining
concentrations of analyte between 10 and 500 ppb (see Table
5.3). This band was used for all of the analytical
determinations because it was linear with concentration over
several orders of magnitude. These results help to ensure

Figure 5.3. Structures of the three sulfa drugs studied
sulfadiazine, sulfamerazine, and sulfamethazine.

63
Sulfadiazine
Sulfamerazme
Sulfamethazine

Figure 5.4. The SERS spectrum of sulfamethazine on colloidal silver at three different
concentrations, a) 100 ppb; b) 10 ppb; c) 1 ppb

Raman Shift (cm—1)
Raman Scattered Intensity (arb. units)
S9

66
Table 5.3.
Analytical Figures
Colloidal System.
of Merit8 for
Sulfa Drugs on the
Silver
Compounds
LOD's
Sensitivity
R
(PPb)
(ppm )
sulfadiazine
1.0
66.74
0.997
sulfamerazine
1.0
143.82
0.999
sulfamethazine
10.0
16.62
0.999
a LOD is the limit of detection using three standard
deviations above the noise, the sensitivity is the slope of
the calibration curve, and R is the correlation coefficient of
the log-log plot.

67
that SERS will eventually provide high quality qualitative and
quantitative information as a method of routine analysis.

CHAPTER 6
A STUDY OF COPPER COLLOIDS
Introduction
The Raman scattering cross section can be increased up to
six orders of magnitude by bringing the molecule close to a
surface [6,10]. Copper, gold and silver surfaces have
generated the largest enhancements [58,69]. Because of this
enhancement, Raman spectroscopy can compete more effectively
as a trace analysis technigue with other molecular
spectroscopies, such as fluorescence and phosphorescence. The
advantage of SERS over these techniques is two-fold. First,
not all molecules fluoresce or phosphoresce, whereas all
molecules have vibrational modes accessible to SERS. Second,
SERS spectra provide a wealth of information, allowing one to
identify the molecule and/or differentiate it from a similar
molecule adsorbed on the surface. These two reasons provide
the impetus for the development of SERS as a trace analytical
technique. Reviews by Garrell [1] and Campion and Woodruff
[5] have focussed on the application of Raman techniques to
analytical problems.
Silver and gold colloids were first used as SERS
substrates in 1979 by Creighton et al. [23]. Much of the
research on SERS has utilized silver colloids as the
68

69
substrate. Silver colloids are easily produced, provide a
fresh surface for the adsorption of analyte molecules, and can
be excited by an argon ion laser. Silver hydrosols have been
used for the quantitative analysis of drugs [28,70,71] and
nitrogen containing heterocycles [72,53].
Silver hydrosols have been adapted for use with flow
injection analysis [30,31,33,29] and HPLC [34,52] systems. On
line measurements increase the reproducibility of colloid
preparation, thus lowering limits of detection. SERS has also
been adapted to both paper [52,54] and thin layer
chromatography [55] in efforts to further increase the
detection power and identification power of these separation
techniques.
Copper colloids were first evaluated for SERS in 1983 by
Creighton et al. [24]; in this study, the SERS enhancement of
pyridine by a citrate stabilized copper hydrosol was
investigated. A later study by Heard et al. [35] found there
was competitive adsorption between pyridine and citrate onto
the silver colloid surface. To minimize this interference, an
unprotected copper colloid was used in the present study. The
investigation of a copper colloid for use as an analytical
SERS substrate is reported for the first time. The colloidal
surface is prepared by a simple borohydride reduction of
cupric nitrate. These results illustrate the bright future of
SERS on copper colloids for analytical use.

70
Experimental
Apparatus
In Figure 6.1, a schematic diagram of the experimental
setup is shown. A Spectra Physics 2040 argon ion laser was
used to pump a dye laser (Spectra-Physics, model 375). DCM
(Exciton) was used to produce the necessary wavelengths. All
measurements were made relative to the 100 mW of 660 nm
exciting radiation. The stability of the output power and
lasing frequency were continuously monitored using both a
power meter and a 0.3 m monochrometer (McPhereson 218) with a
R928 Hamamatsu photomultiplier tube. The laser beam was
directed through the sample twice and the scattered radiation
collected at 90 degrees. A 0.22 m double monochrometer
(Spex 1680B) with slits set for 6 cm 1 resolution was used at
a scan rate of 3.0 nm/min for wavelength discrimination. The
Raman scatter was detected using a thermoelectrically cooled
(R63 6 Hamamatsu) photomultiplier tube. The photoelectron
pulses were amplified and counted using a (Stanford, model
440) photon counter with a 1 s acquisition time. The data was
collected on an IBM PC. The Raman bands used for quantitative
measurements were 1010, 1397, and 1387 cm1 for pyridine, p-
aminobenzoic acid, and p-nitrobenzoic acid, respectively.

Figure 6.1. A schematic diagram of the experimental layout of the optical components
used in the copper colloid studies.

MONOCHROMATOR
MIRROR
/
\
MIRROR
ARGON ION LASER
DYE LASER
METER
BEAM SPLITTER
SAMPLE CHAMBER
- RECORDER
MIRROR
PHOTON COUNTER
MONOCHROMATOR
(s-Miu-lj
COMPUTER
to

73
Chemicals and Procedure
A 30 ml volume of a 2 X 10 3 M cupric nitrate solution was
added dropwise to 70 ml of a 1.0 X 102 M sodium borohydride
solution at room temperature with stirring. The analyte was
added as a 1.00 ml aliquot to the resulting solution. The pH
of the solution was then adjusted as needed using a small
amount of nitric acid. A sample of this pH adjusted solution
was then placed in a standard cuvette. Measurements began two
min. after the first two reagents were completely combined.
All glassware had been cleaned thoroughly with concentrated
nitric acid and the rinsed repeatedly with Barnstead water.
Pyridine, p-nitrobenzoic acid (PNBA), methanol and sodium
borohydride were obtained from Kodak (Rochester, NY). Cupric
nitrate and p-aminobenzoic acid (PABA) were obtained from
Fisher Scientific. The solutions of PABA and PNBA were
prepared in 20% (v/v) methanol. All chemicals were analytical
reagent grade or equivalent and used without further
purification.
Results and Discussion
Pyridine was chosen as the test compound for the initial
optimization because it is soluble in water, has a large Raman
cross section, adsorbs strongly to colloidal surfaces, and
SERS has been observed previously on copper colloids (figure
6.2). The enhancement factor for the pyridine 1010 cm1 band,
corrected for prefilter and post filter effects was calculated
to be 1.8 X 10+5 , which is in good agreement with previous

74
results [24]. Pyridine (Pka=5.19) proved a good choice,
because of the relatively small effect that pH had upon its
SERS signal compared to those of the other analytes (figure
6.3). The optimal pH for the observation of SERS from
pyridine differs from that previously observed using a copper
electrode [18] and a different copper colloid [24]. The
optimal pH for the copper electrode and the copper colloid
used in this study differ by « 1.5 units, which may be due to
differences in the surface morphology or potential. The pH
difference between the copper colloids in the present study
and those used by Creighton et al. [24] is about 2.5 units.
This is probably due to their use of citrate to stabilize the
colloid. The adsorbed citrate will affect both the surface
potential of the colloid and the adsorption of the pyridine
onto the surface. The difference in the pH dependence of the
SERS intensities of PNBA (Pka=3.45) and PABA (Pka=4.65, 9.30)
are due to the differences between their dissociation
constants. The acidic form of the amino group on PABA will
adsorb more strongly than the basic form. At slightly basic
pH's, PABA is bound to the surface via both functional groups.
The effect of pH on PNBA is probably due mainly to the surface
charge of the colloid. The narrow pH ranges within which PABA
and PNBA exhibit SERS presents an additional method for differ
entiating between similar compounds.
The absorption spectrum of the colloid over time (figure
6.3) shows the growth of a peak at 570 nm. It has been

Figure 6.2. The Raman spectra for pyridine, a) a 10% aqueous solution of pyridine; b)
a 10 ppm concentration of pyridine in the copper colloidal solution; c) a background
spectrum of the copper colloid. Excitation was at 660 nm, with 150 mW laser output.

Raman Shift (cm-1)
o
o
a
Raman Scattered Intensity (arb. units)
N3 ot
9 L
750 H

Figure 6.3. Relative copper colloid-SERS intensity of most intense band of the analytes
as a function of pH. A 10 ppm solution and the 1010 cm 1 band were usedi for the
pyridine curve. A 50 ppm solution was used for the evaluation of the 1397 cm band of
PABA and the 1387 cm 1 band of PNBA.

RELATIVE INTENSITY
10
pH OF SOLUTION

79
previously calculated that there is a correlation between the
570 nm peak and copper spheres having a diameter of « 20 nm.
The particle size distribution of the colloids was analyzed
with a Malvern 4700 photon correlation spectrometer. The data
analysis was performed using the optical constants for bulk
copper [71]. During the time span of interest, the average
particle diameter was found to be approximately 100 nm (see
figure 6.4).
The optical density in figure 6.3 reaches a maximum 780
seconds after the reagents were combined. If the 20 nm
particles are the precursors of the most SERS active particle
size, the SERS signals should increase while the 20 nm
particles are present and then decrease. This correlates well
with the SERS signals shown in figure 6.5., which is largest
for the 1397 cm 1 band at 800 s and decreases thereafter. The
time dependance of the 1397 cm 1 SERS signal makes it difficult
to acquire a good spectrum using a scanning instrument
(Figures 6.7, 6.8).
The copper colloid system has a much lower background
than silver colloidal systems. Unfortunately, the SERS
signals are also smaller, so that the signal to noise ratio is
not improved. This may be due to lower adsorption onto the
copper surface as compared to silver surfaces. An additional
problem is that the copper colloid cannot be prepared more
than 30 min. ahead of time, where the silver colloid can be
made stable for future use [23,33,34].

Figure 6.4. Absorption scans of the copper colloid as it aggregates over time. The
lower scan was taken 180 s after mixing the reagents. The remaining scans show the
increase in absorbance occurring between 360 and 780 s at 30 s intervals. The
conditions were pH=9.0, [Cu] = 6.0 X 10 , [BH4] = 1.9 X 10 , and 50 ppm para-amino
benzoic acid.

3Ae^
m
o
Atosortance
X8

Figure 6.5. A photon correlation spectrum of the copper colloids used in these studies.
The data was corrected for the dielectric function of copper. The ordinate is the
percentage of the total composition that the fraction makes up.

03
OJ

Figure 6.6. Intensities of the 1606 cm (filled
1397 cm (filled circle) SERS bands of+^ABA as
time. The conditions were pH=9.0, [Cu¿] = 6.0
= 1.9 X 10 , and 50 ppm para-aminobenzoic acid.
triangle) and
a function of
X 10, [BH4‘]

Time (s)
Relative SERS Intensity
O -»■ N> U) Oi
O
00
(J1

Figure 6.7. a) the SER spectrum of 50 ppm of p-aminobenzoic acid on a copper colloid;
b) the background spectrum of the copper colloid. Excitation was at 660 nm, with 150 mW
laser output.

Raman Shift (cm—1)
Raman Scattered Intensity (arb. units)
M OJ 00
L 8
1000

Figure 6.8. a) the SERS spectrum of 50 ppm of p-nitrobenzoic acid on a copper colloid;
b) the background spectrum of the copper colloid. Excitation was at 660 nm, with 150 mW
laser output.

Raman Shift (cm—1)
Raman Scattered Intensity (arb. units)
68
1000

curves for pyridine, p-aminobenzoic
acid on the copper colloid.
Figure 6.9. Calibration
acid, and p-nitrobenzoic

o
9
8
7
e
5
4
3
2
1
O
20 40 60 80 100
Conceniration (ppm)

92
The colloidal copper surface has a linear dynamic range
of 1 to 1.5 orders of magnitude (figure 6.9). Limited dynamic
range is common with colloidal systems, since the signal is
dependent on the adsorption isotherm. Inability to control
directly adsorption onto the surface, as can be done with
electrode surfaces, is the greatest limitation of colloidal
based SERS. The limits of detection for pyridine, PABA, and
PNBA are estimated to be 1.5, 2.5, and 13 ppm, respectively
(Table 6.1, 6.2).

93
Table 6.1.
Analytical figures
Colloid.
of Merit
for Analytes
on Static
Copper
Compounds
LOD a
Sens.b
Slope c
Corr .d
(ppm)
log/log
Coeff.
Pyridine
1.5
0.597
1.04
0.993
p-aminobenzoic
acid
2.4
0.270
0.978
0.999
p-nitrobenzoic
acid
12.9
0.051
1.04
0.999
where, a, is the limit of detection for the compound, using
three standard deviations al^ovec the noise, , is the
sensitivity of the ^/nethod (ppm ') , c, is the slope of the
log-log plot, and , is the correlation coefficient for the
log-log plot.

94
Table 6.2.
Comparison of Static Copper Colloid p-aminobenzoic acid
Results with Literature Results on Colloidal Silver.
Author
LOD's (ppm)a
Slope b
Corr. Coeff.c
(Log/Log)
E. L. Torres
0.5
1.05
0.997
This Work
2.5
0.978
0.999
where, a, is the limit of detection for th^ compound, using
three standard deviations above the noise, , is the slope of
the log-log plot, and c, is the correlation coefficient for
the log-log plot.

CHAPTER 7
COPPER COLLOIDS IN A FLOWING STREAM
Introduction
The technique of Flow Injection Analysis (FIA) utilizes
control over the dispersion of the sample in a flowing stream
to maintain the integrity of the sample [93]. In FIA, the
sample is injected into the stream using a chromatographic
injection valve for analysis downstream. The injections need
to be spaced far enough apart so that dispersion of adjacent
samples do not cause them to mix.
Injection directly into the flowing stream has several
advantages over the previous on-line method. The integrity of
the sample was maintained previously by interspacing the
sample with bubbles. The bubbles kept the samples distinctly
separate, but needed to be removed before the analysis of the
samples took place. The compressibility of the added air also
caused pulsation in the streams. This meant that the movement
of the sample stream could not be exactly controlled. Since
there are fewer difficulties in controlling an entirely liquid
flowing stream, the equipment required for flow injection
analysis is relatively simple.
FIA has been used to evaluate the feasibility of adapting
methods to post-column HPLC analysis, since the analyte is
95

96
injected into a flowing stream and can be manipulated further
down stream as required. Changes in the system are seen
quickly, allowing rapid optimization of parameters. The time
required for the separation limits the pace of optimization in
HPLC.
Laserna et al. first demonstrated the application of SERS
to FIA [29,30]. For the determination of p-aminobenzoic acid
the % RSD was reduced from 10-15% to 3.2%. The detection
limit was also reduced by an order of magnitude. Ni et al.
[33] detected SERS from RNS bases in an FIA system. The
effects of pH and temperature on the SERS signals was also
investigated.
Combined resonance and surface enhanced Raman
spectroscopy (SERRS) has also been adapted to FIA [32].
Crystal violet was used as the analyte. The detection system
used an optical multichannel analyzer to analyze the scatter
collected via a Raman micro-probe. The detection limit was
. -12
found to be approximately 10 M.
Freeman et al. [31] were the first to adapt SERS to HPLC.
After first optimizing the system using FIA, SERS measurements
of pararosaniline hydrochloride were made after passing the
analyte through an HPLC column. The detection limits for this
system were in the ppb range. Sheng et al. [34] separated
four purine bases and analyzed the eluent with SERS active
silver colloids. The data was collected using an optical
multichannel analyzer. The detection limits were in the

97
nanomolar range. These results illustrate the progress of
SERS towards finally becoming a routine method for analysis.
Further progress toward this goal is covered in the next
section, where adaptation of copper colloids to FIA is
discussed.
Experimental
Apparatus
The instrumental system used in these experiments is the
same as that of the previous chapter (see Figure 6.1). Figure
7.1 is a schematic diagram of the setup for producing the
copper colloids on-line. A peristaltic pump (Rainin) was used
-2 .
to pump a solution 1.8 X 10 M of sodium borohydnde at 1.02
ml/min. A similar pump was used to pump solution at 0.48
ml/min from a reservoir of 2.0 X 10 3 M cupric nitrate prepared
in 1.36 X 101 M nitric acid. These were initially mixed by
flowing them together into a chromatographic "T" joint. The
colloidal solution flowed through a 6 m length of tubing to an
injection valve (Altex Scientific Inc, model 905) with a 20 /¿I
sample loop. From the injection valve the solution flowed
directly into the mixing chamber, where the plug of PABA was
thoroughly mixed with the colloidal copper. The mixing
chamber was open to the atmosphere to remove the hydrogen gas
that evolved in the formation of the colloid. The volume of
the solution in the mixing cell ranged from approximately 1.0
to 1.5 ml. The solution was pumped from the mixing chamber

Figure 7.1. Schematic diagram of the flow injection analysis
setup used to produce the copper colloids.

99
PUMP PUMP
PUMP

100
through the spectroscopic cell by a third peristaltic pump
metering at 0.95 ml/min. The spectroscopic cell had a length
of 3.0 cm with a cross-section of 0.2 cm by 0.2 cm and was
masked to reduce interference from laser scatter off the cell
walls. Approximately 60 s after the injection, the
concentration of the analyte was at its highest in the
spectroscopic cell.
Results and Discussion
Several 2.0 X 10 3 M solutions of cupric nitrate in
various concentrations of acid were prepared. These solutions
were used in the on-line preparation of copper colloids.
Since the time after the injection of the p-aminobenzoic acid
at which the concentration is at its highest is known to be
«60 s, this was used to evaluate the ability of the acid
present to induce aggregation of the copper colloid. This was
accomplished by injecting the analyte and turning all of the
pumps off 60 s after the injection. The SERS signal from this
volume of solution was then measured. The effect of
increasing the acidity of the copper solution on the
aggregation time is shown in Table 7.1.
A scan of the 1262 to 1818 cm 1 region under stopped
flow conditions is shown in figure 7.2. The SERS band at 1697
cm 1 is attributed to the symmetric stretching mode of the
carboxylate group. Stretching modes of the benzene ring give
rise to the bands at 1606 cm’1 and 1520 cm"1. This
demonstrates the ability of copper colloid based-SERS to

101
provide qualitative information about the structure of the
adsorbed species.
A series of nine 200 jug injections of p-aminobenzoic acid
are shown in figure 7.3. The shapes of these peaks seem to
follow the dilution of the analyte by the mixing chamber, but
are lacking in reproducibility. The variable shapes of the
peaks are a result of the hydrogen gas evolved in the lines.
The hydrogen bubbles are difficult to get through the small
diameter of the injection loop, causing irreproducibility in
the flow of the colloidal solution through that point. Since
the size and distribution of the bubbles is random and the
cell is open to the atmosphere, the irreproducibility in the
flow introduces variation in the volume of liquid in the cell.
These changes in the volume cause variation in the dilution of
the analyte by the hydrosol. The relationship between the
amount of p-aminobenzoic acid injected and the area of the
SERS-FIA peaks is shown in figure 7.4. The detection limit
for this system is approximately 150 jug. A comparison of this
with previous results on silver colloids is shown in Table
7.2.
The difference in the concentration of sodium borohydride
used in the present study and a previous study by Laserna et
al. [29] is an order of magnitude. This higher concentration
of sodium borohydride increased the amount of hydrogen evolved
and the difficulty in removing it.

102
Table 7.1.
The Effect of Increased Acidity on the Aggregation of the
Copper Colloid and the Resulting SERS Signal.
Nitric Acid Concentration (M) Time of Onset of SERS (s)
1.10
X
io'1
674
1.18
X
10 1
491
1.26
X
io'1
378
1.32
X
io'1
133
1.36
X
io'1
«60

Figure 7.2. The SERS spectrum of p-aminobenzoic acid on a FIA produced copper colloid,
between 1265 and 1815 cm .

Raman Shift (cm- 1)
Raman Scattered Intensity (arb. units)
ÍO 4*. o> oo
o o o o
*01
10000

Figure 7.3. The SERS signals, at 1397 cm1, from a series of 9 injections of 200 /Ltg of
p-aminobenzoic acid into the copper colloid-FIA system.

Time (s)
Raman Scattered Intensity (arb. units)
—• NJ
o o
o o
b
90T
3000

Figure 7.4 Calibration curve of the response of the SERS signals to various amounts of
p-aminobenzoic acid injected into the copper colloid-FIA system. The symbols of the two
lowest measurements are larger than the error in the measurement.

200
Amount of PABA injected (micrograms)
2500

109
Table 7.2
Comparison of p-aminobenzoic acid Copper Colloid-FIA Data with
Literature Rsults on Silver Colloids.
Authors
LOD's
Flow Rate
Cone.
of Metal
(Mg)
Laserna et al.
0.03
0.98 ml/min
1.2 X
10'3 M
This Work
150
2.22 ml/min
1.2 X
10‘2 M
where, a, is the limit of detection for the compound (ng) ,
using three standard deviations above the noise, , is the
flow rate through the spectroscopic cell (ml/min), and c, is
the concentration of the sodium borohydride in the final
solution.

110
This problem currently limits the analytical usefulness of
copper colloids prepared via FIA.

CHAPTER 8
CONCLUSIONS AND SPECULATIONS ON THE FUTURE
Conclusions
In the previous chapters, the results of SERS experiments
on both silver and copper colloidal surfaces were presented.
The concentration of three sulfapyrimidine drugs in the low
ppb range were measured using silver colloids as the SERS
active surface. These guantitative results are comparable
with other methods for determining the concentration of sulfa
drugs. Additionally, the ability to differentiate between the
three sulfa drugs at ppb levels was demonstrated. This was not
possible in previous analytical studies [89-92]. The
combination of the quantitative and qualitative merits of SERS
provides the impetus for continuing research until it becomes
a routine form of analysis.
One of the routes for making SERS more routine is
reducing the cost. Lower cost and the reduction of
interference from fluorescent species are the major reasons
for investigating copper colloids for analytical use. In
chapter 6, copper colloids prepared under batch (static)
conditions were evaluated. The detection limit for PABA was
comparable to those of Torres [25] on silver hydrosols.
Copper colloids are more active than those of silver, which
111

112
increases the difficulty of controlling the aggregation and
chemistry of the system. This increased chemical activity
might be exploited in the search for SERS active surfaces onto
which molecules adsorb strongly. Detection limits for
pyridine, p-aminobenzoic acid, and p-nitrobenzoic acid in the
low ppm range were presented.
One of the major difficulties with colloids is
irreproducible preparation. The production of copper colloids
on-line is a step toward reducing the variations found in
colloid preparation. The detection limit for PABA on a FIA
produced copper colloid is approximately 150 /xg, two orders of
magnitude above that for PABA on silver hydrosols produced
on-line. The major difference in the preparation of these two
hydrosols is the amount of sodium borohydride required to
produce the colloids and minimize dissolution. The
concentration of sodium borohydride used in the production of
the copper hydrosol is an order of magnitude higher than that
required for silver systems, and leads to the formation of
large amounts of hydrogen. Extensive efforts were required to
remove the bubbles in the copper hydrosol system, which
resulted in a much lower reproducibility than would be
expected from the static experiments. The future of FIA with
copper colloids is dependent upon the improvement of the
reproducibility and detection power.

113
Speculations on the Future
Experiments on Filters
The investigation of colloids on surfaces should aid in
discrimination between enhancements due to chemical and
electromagnetic enhancements. A series of experiments where
colloids of gold, silver, copper and possibly other metals are
filtered using surface capture filters could be informative.
Various dilutions of colloids could be used to provide
different surface coverages upon the filter. By comparing the
average distance of separation between the particles with the
SERS enhancement, it may be possible to determine the effects
of particle interactions.
The colloids could also be prepared with varying degrees
of aggregation and then filtered. This would reveal the
effects of particle size and shape on the SERS activity of the
surface. The surface capture filters can be used uncoated,
yielding information on the ability of the isolated spheroids
to enhance the Raman signals.
These filters could also be vapor deposited with various
metals. The purpose of this is to observe any differences
between the SERS intensities of metal colloids filtered onto
conductive metal surfaces which are able to be excited by the
incident radiation and those which are not. An example of
this would be a comparison between a small number of silver
colloids on a silver surface and a similar number on a copper
surface. These surfaces would be irradiated with an argon ion

114
laser. The filter surfaces would be electrically conducting,
so that the colloidal particles would be electrically
connected in both cases. For the case where both the surface
and the particles can interact with the incident radiation,
the electromagnetic theory would predict any additional
enhancement for the silver colloids on the silver surface over
that seen for silver colloids on the copper surface would be
equal to the enhancement from a flat surface. The adatom
model would predict a larger enhancement than the
electromagnetic theory for the silver on copper system due to
the chemical interactions between the copper surface and the
analyte. To resolve this effect adequately, a number of
analytes must be used. These analytes should either physisorb
or chemisorb onto the surface with minimal preference.
Perhaps the ultimate analytical application of SERS would
be to thin layer chromatography. A commercially produced SERS
active chromatographic surface would allow the user to
separate the compounds and then identify them, without
extensive sample preparation. Ideally, the initial assay
could determine where the spots are by using a UV lamp and
observing the fluorescence from the analytes, and then allow
further investigation as needed using SERS. The detection
limits could easily be in the ng or pg range, with the
advantage of identifying the molecule.

115
Experiments on Novel Colloidal Surfaces
Another area for experimentation is in the area of mixed
metal colloids. This might produce a colloidal surface which
is more stable against aggregation, yet shows increased
adsorption of molecules from the solution phase. One
possible combination might be a copper/gold hydrosol. Gold
hydrosols are extremely stable, but fairly expensive.
Alloying this with copper would decrease the cost and present
a surface which is more chemically active than a normal gold
surface. The copper/gold colloid would also probably be more
stable than a pure copper colloid. Additionally,
catalytically active colloidal surfaces [30] might be prepared
on which SERS can be used to follow the reaction process,
possibly discerning the mechanism of the reaction.
Methodology
The use of temperature to control the aggregation rate
will probably see more use. This technique should be more
reliable than using ionic strength, since it allows finer
control of the effect of the double layer on the aggregation.
In the future, it may be possible to prepare colloids of
copper and other active metals ahead of time and activate them
by simply heating them, as Cotton et al. [34] have recently
done with silver. This would simplify the adaptation of SERS
to on-line situations, such as HPLC.
The future of SERS will involve multichannel detectors,
such as photodiode arrays or charge coupled devices. These

116
currently allow the acquisition of unenhanced Raman spectra
from surfaces. This technology reduces the time required to
obtain spectra. The increased use of tunable solid state
lasers also will help to make SERS, and Raman spectroscopy in
general, a routine method of analysis. Solid state lasers,
such as the Ti-sapphire and diode lasers can have high power,
a wide tuning range, and are relatively simple to operate when
compared to dye lasers.

APPENDIX
This is an electromagnetic treatment of the SERS
phenomena as given by Gersten and Nitzen [65,94], who have
most explicitly treated the SERS phenomenon observed on
colloidal surfaces, and Adrian [95], who provides the link
between the derivation and the simplified equations shown in
chapter. The following derivation explains the origins of
"the lightening Rod Effect", the image enhancement mechanism,
and resonance effects.
Molecule Near a Hemispheroid Protruding from a Surface.
This derivation uses the model of a single conducting
hemispheroid protruding from a flat conducting surface. This
simplifies the initial derivation, allowing the contributions
from the conducting plane and the hemispheroid to be later
separated. One of the assumptions is that the potential of
the planar surface is zero. This allows the effect of the
high degree of curvature to be probed more easily. Additional
assumptions will be used and outlined as necessary in the body
of the text.
The following treatment is only rigorous for particles
which are much smaller than the incident wavelength, the
"Rayleigh limit". For this case, contributions from the
117

118
multipoles above the dipole can be neglected without
significant error. For a prolate spheroid surface, the
ellipitical coordinates can be defined as
x = f[($2-1) (1-rj2) ]1 /2cos0,
y = f[(£2-1) (l-rj2)]1/2sin0,
z = fir]
(A.l)
where tp in the azimuthal coordinate as measured from the xz
plane, and £ and rj are the orthogonal spherical coordinates.
The focal distance, f , of the prolate spheroid is related to
the major, a, and minor, b, semi-axes by the equation
f = (a2-b2)^.
(A.2)
Assume the case where the surface of the spheroid is defined
by the equation as £0=a/f and the position of the molecule
located a distance H directly above the protrusion is defined
by $1=(a+H)/f.
For the region outside the spheroid, with an applied
(laser) field, a polarized molecule nearby, and a field due to
the spheroid and the plane, the potential, $j, may be written
as

Figure A.l. a). A cross-section of a prolate spheroid
illustrating the elliptical coordinate system. b) . An
illustration of the model used for the derivation in the
appendix.

I
OZT
X)
*(o

121
01 = -Eofin + JL JLiía2 -1) (l-r?2) + ttri-Z 1)2]_1/2
f2 (A.3)
-[($2-l)(l-n2) + (Ir, + t i)]'1/2}+En cnpn(n)Qn(S) •
where Pn and Qn denote Legendre functions of the first and
second kind respectively. The first term is the contribution
to the potential of the surface due to an isolated spheroid an
a conducting plane. The second term contains contributions
from both the polarized molecule and its image a distance a+H
below the conducting plane. The final term represents the
potential associated with the polarized hemispheroid and its
image on the conducting plane. The potential is chosen so
that it vanishes at the boundary of the hemispheroid and the
conducting plane (rj=0) . A result of this is that only the odd
terms appear in the summation.
Inside the prolate spheroid, the general solution to the
Laplace equation can be applied.
*n = En bnPn(n)Pna)- (A.4)
The coefficients bn and cn in equations A. 3 and A. 4 must be
determined when the solutions for the surface and the inside
of the spheroid are matched at the boundary £=£0 for all cases
where CKrjci.

122
Since these are electrically conducting, *I# also needs to be
continuous across the surface.
#1 = -E0fir) + JL ^_{[(£2-l) (1-rj2) + (50»7-5i )2]"1/2
f2 d^
-1 («1-D (i-f)2) * (íoiHi)2r1/2) (A-5)
= rn C^n^ni^o) “ cnQn (£o) ]^n (*?)
If the normal components of the displacement vectors along the
surface are matched to avoid discontinuity in the transition
of the dielectric from that of the metal to that of the
surrounding medium.
*1 = -E0nn + JL dl {[($2-i) (i-r?2) + (50 »?-5i)2]"1/2
/2 d q 0d c1
- [ (i- i?2) * (S„ u+Ci)2]-1'2) = En [e Cvjbni’n <5o) - c„o'({„) ]P„ (r)) .
where e(v) is the dielectric function of the metal, relative
to the surrounding medium. The two above expressions, A.5 and
A.6, can be inverted to obtain the coefficients bn and cn by
employing the integral, A.7
jVn(rj)Pm(T7)drj = , (A.7)
which is valid when m and n are both odd, and A.8, for the
case where n is odd and ^>^>1.

123
Jo1 Pn(!7){l/[(Éo-l) dV) + (5o Hl)2]1/2-l/[(5o-l)
(1- rj2) + (£0 r?+$1)2]1/2)d r7 = 2Pn (£0) Qn ) ,
These yield expressions A.9 and A.10, respectively
(5 o) “ cnQn(£ o)
= -i^f£05n>1 + (4n+2)(/i/f2)Pn(£o)0n(So)/
e(v)bnp'n(Z0) -cno'(50)
= -V¿n), + (4n+2) (n/f2)P^ao)Q,na^)
Solving A.9 and A.10 for cn yields
+
c
n =
(4n+2)n
(e(v)-l)Eofl06nti
e(v)0i($o) “ loQ\^o)
(1 -c(v))p'(50)C?'(5i)fj<(5o)
e(v)Qntto)P'n(Zo) ~ Q'n(to)Pn(lo)
(A.11)
For the Rayleigh limit, the electromagnetic emission is due to
the dipole moment. If is evaluated for the case where £ is
large, we obtain
$1 r
as iQ~* oo
where,
Dl
(fi)2
~ Eq ftri
D = 2m + -ic^f2
(A.12)
(A.13)
The first term in the equation A. 13 accounts for both the
molecular dipole and its image in the plane. The second term

124
describes the dipole induced on the spheroid and its image in
plane. The expansion of these terms yields
D =
f3Eo
5o(«(v)-l)
+ 2n
1 + .
eívjCMÉo) - SoQÍ(Éo)
(l-e(v))5oOÍ(Ci)
e(v)Qi(50) - ^oQ\^o) j
(A.14)
Since the polarizability tensor of the molecule, a1, is
anisotropic, an orientation for the molecule relative to the
surface must be assumed. Assuming the orientation of the
principle axis perpendicular to the surface, the dipole can
then determined by the net electric field produced at the
molecules position by the nearby spheroid, and the externally
applied field.
M = «1
-tE7
cnQn(£l)
4(f£i)
(A.15)

125
combining equations A.14 and A.15 leads to the expressions
D =
+
2a-\E0
i-r
1 +
£o[e(v)-l]
eívJOÍí^o) -eoOÍ(^o)
[i-c(v)]50o{(5i)
e(v)0i(5o) -íoQÍí^o)^
£'
a1 2a1[e(v)-l]
r = ! + !
4(i$i)3 f3
(2n + l)Pntt0)p'na0) CQp C^1) ]2
e(v)P'nUo)Qntto) ~ Pn(Zo)Q'n(lo)
(A.16)
(A.17)
The first term of equation A. 16 is the contribution to the
dipole moment due an isolated hemispheroid on a conducting
plane. The denominator 1-r, may be termed the image
enhancement factor. From equation A.17, it can be seen that
r consists of two terms. The first term of equation A.17 is
due to the image of the molecular dipole in the plane, and the
second due to its image in the hemispheroid.
The contribution the the Raman signal from the nearby plane is
expressed by equation A.18
r- 2a.|/ [2 (a+H) ]3.
(A.18)

126
If the hemispheroid is a prefect conductor, then |e(v)|-+ °°
yielding
rm =
“1
2ai
4(f$i)
Yf (2n+l)
n
Pn($o) [Qp(^l)32
On (£o)
(A.19)
For a hemispheroid on a conducting plane, the dipole moment
given by equation A.16 reduces to
Deo
f3Eot o
3£?1 (ío)
+
20t-\Eg
(£1)
oi a0)
^2
(A.20)
Full Prolate Spheroid
If an insulated full spheroid (ie. not one embedded in a
conducting surface) is considered then, the image dipole term
associated with the planar surface (equation A.18) is
neglected, and the sum is then extended over all values of n.
Matching the potentials over the range, -1 < rj <1, yields
~ = f^Eo gp[6(v)-l]
3 e(v)Qi($0) -ScOÍ^o)
“l-E’o
i-r
1+
[l-e(v)]$o0(($i)
eívJOTÍ^oí-SoOÍíío)
where,
/
2
(A.21)

127
f = fl[e(v)-l] £
(2n+l)Pn(5o)p'[0' ($!) ]2
(A.22)
r° n
n e(v)P„ (50>On(£o> - Pn(Zo)Qn (So)
If we further assume the surface is a perfect conductor,
expressions A.21 and A.22 simplify to
( , \2
f3EoZ0 + cc^Eq 5oQi(5o)
(A.23)
30i (So) i-r. Q\ tto)
(A.24)
Equations A. 22 and A. 23 are similar to the results for the
hemispheroid on the conducting planar surface (equations A.19
and A.20).
These equations all contain terms in the denominator of the
form
A($0,v) =e(v)Qi(£0) -£o0i(£o)
(A.25)
This is the condition for the plasmon resonance upon the
surface. For a sphere this is
(A.26)

128
The conditions for the surface plasmon resonance on a sphere
is
e(v)+ (1+1)/1 = 0 (A.27)
where 1 is the quantum number for plasmon resonance on sphere.
For the dipole case (1=1), this reduces to e(v) = -2. As the
spheroid becomes more prolate (£0-* 1) , the equation becomes
e(v) = {(£0 -1) In[2/($Q -l)}*1. (A.28)
which means the real component of e(v) will be large and
negative. As the spheroid becomes more and more eccentric,
the surface plasmon frequency corresponding to the major axis
will decrease. The laser radiation will be in resonance with
only a fraction of the particles, but the enhancement from
these particles will be strong. It is necessary that the
imaginary component, Im(e(v)), be small at this resonance for
high efficiency in the coupling of the photons with the
surface plasmons.
The Enhancement Factor
The cross section of the Raman scattering from the isolated
molecule can be obtained in a semi-classical manner beginning
with
. _ da
a = a 0+AQ~— cosvvt.
oQ
(A.29)

129
where AQ, is a molecular mode coordinate and vv, is the
corresponding molecular vibrational frequency.
- £ $* <«>* (£)* (A. 30)
The Raman scattering from the molecule near the surface can
then be calculated using equation A.16 to be
a
rs “
8n I v \4
T \cj
1
[i-r]4
(l-ejSoQÍa!)
1 +
£Ql (So) - £qQ-\ (5o)
(A.31)
To obtain the enhancement due to the proximity of the surface,
the enhanced Raman scatter in equation A.31, can be divided by
the Raman scatter due to the isolated molecule (equation A. 30)
R =
l+(l-e(v))S0£>í(Si)/[e(v)QÍ($0) _ é0q¡(£0)]
I^f
(A.32)
For a perfect conductor, the dipole contribution is given by
equation A.23, yielding
(ed -
. (e(v) + xed)£>i(U
d-r)
(A.33)
Where e(v) is the dielectric function of the metal, ed is the
dielectric function of the surrounding medium, and x is

130
dependent upon the curvature of the surface. % is two for a
sphere.
For the case where the molecule is adsorbed on the surface
(Él = £0) and the photons couple efficiently with the surface
phonons, A.33 equation reduces to
e(v) - ed
e(v) + 2ed
(A.34)
Equation A.35 is an approximation of equation A.34 which is
modeled after a silver prolate spheroid 20 nm x 60 nm embedded
into a flat silver surface.
G = |e(v)-1/e(v)2|4 (A.35)
This equation best illustrates the importance of the imaginary
term of the dielectric function, e(v)2, of the metal.

131
Table A.l. List of Symbols used in the Appendix.
Symbol Meaning
o
$
f
l,r)
\o
^,*7=1
0
a
b
v
v '
M
P,Q
External Field Strength
Potential of the Surface
Scaling Parameter, Focal length of Spheroid
Elliptical Coordinates
Surface of Spheroid
Elliptical Coordinates of Molecule
Azimuthal Coordinate Measured from yz Plane
Semi-major Axis of Spheroid
Semi-minor Axis of Spheroid
Incedent Frequency
Raman Scattered Frequency
Induced Dipole
Polarizability of the Molecule
Polarizability of the Spheroid
Dielectric Function of the Surface
Dielectric Function of the Surrounding Medium
Height of Molecule Over Surface
Legendre Functions ( • indicates phase boundary)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
REFERENCES
Garrell, R.L. Anal. Chem. 1989, 61, 401A.
Campion, A. in Vibrational Spectroscopy of Molecules on
Surfaces; Yates, J.T.; Madely, T.G., Eds.; Plenum Press:
New York, 1987; Chapter 8.
Davidson, T.; Pons, B.S.;Bewick, A.; Schmidt, P.P. J.
Electroanal. Chem. 1981, 125, 237.
Otto, A. in Light Scattering in Solids III, Topics in
Applied Physics Series, Volume 54; Cordonna, M;
Guntherodt, G., Eds.; Springer-VerlagrNew York, 1984;
Chapter 6.
Campion, A.; Woodruff, W.H. Anal. Chem. 1987, 59, 1299A.
Fleishman, M. ; Hendra, P.J.; McQuillan, A.J. Chem. Phys.
Lett. 1974, 26, 163.
Paul, R.L.; McQuillan, A.J.; Hendra, P.J.; Fleishman, M.
J. Electroanl. Chem. 1975, 66, 248.
Pettinger, B.; Wenning, U.; Wetzel, H. Surf. Sci. 1980,
101, 409.
Wenning, U.; Pettinger, B.; Wetzel, H. Chem. Phys. Lett.
1980, 70, 49.
Jeanmarie, D.L.; Van Duyne, R.P. J. Electroanal. Chem.
1977, 84, 1.
Albrecht, M.G.; Creighton, J.A. J. Am. Chem. Soc. 1977,
99, 5215.
Barradas, R.G.; Conway, B.E. J. Electroanal. Chem. 1963,
6, 314.
Carrabba, M.M.; Edmonds, R.B.; Rauh, R.D. Anal. Chem.
1987, 59, 2559.
Forcé, R.K. Anal. Chem. 1988, 60, 1987.
Pothier, N.J.; Forcé, R.K. Anal. Chem. 1990, 62, 678.
132

16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
133
Ni, F.; Thomas, L.; Cotton, T.M. Anal. Chem. 1989,
61, 888.
Pettinger, B.; Wetzel, H. Ber. Bunsenges. Phys. Chem.
1981, 85, 473.
Mernagh, T.P.; Cooney, R.P. J. Raman Spectrosc. 1985, 16,
62.
Crookell, A.; Fleishman, M.; Hanniet, M.; Hendra, P.J.
Chem. Phys. Lett. 1988, 149, 123.
Koglin, E.; Séquaris, J.-M.; Valenta, P. J. Mol. Struct.
1980, 60, 421.
Koglin, E.; Séquaris, J.-M.; Valenta, P. J. Mol. Struct.
1982, 79, 185.
Otto, C.; van den Tweel, T.J.J.; de Muí, F.F.M.;
Greve, J. J. Raman Spectrosc. 1986, 17, 289.
Creighton, J.A.; Blatchford, C.G.; Albrecht, M.G.
J. Chem. Soc., Faraday Trans. 2. 1979, 75, 790.
Creighton, J.A.; Alvarez, M.S.; Weitz, D.A.; Garoff, S.;
Kim, M.W. J. Phys. Chem. 1983, 87, 4793.
Torres, E.L. Dissertation, University of Florida, 1986.
Sheng, R.-S.; Zhu, L. ; Morris, M.D. Anal. Chem. 1986,
58, 1116.
Ahern, A.; Garell, R.L. Anal. Chem. 1987, 59, 2813.
Sutherland, W.S.; Laserna, J.J.; Angebranndt, M.J.;
Winefordner, J.D. Anal. Chem. 1990, 62, 689.
Laserna, J.J.; Berthod, A.; Winefordner, J.D. Appl.
Spectrosc. 1987, 41, 1137.
Laserna, J.J.; Berthod, A.; Winefordner, J.D. Microchem.
J. 1988, 38, 125.
Freeman, R.D.; Hammaker, R.M.; McLean, C.E.; Fateley,
W.G. Appl. Spectrosc. 1988, 42, 456.
Taylor, G.T.; Sharma, S.K.; Mohanan, K. Appl. Spectrosc.
1990, 44, 635.
Ni, F; Sheng, R.; Cotton, T.M. Anal. Chem. 1990, 62,
1958.

34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
134
Sheng, R.; Ni, F.; Cotton, T.M. Anal. Chem. 1991, 63,
437.
Heard, S.M.; Greiser, F.; Barraclough, C.G.; Sanders,
J.V. J. Phys. Chem. 1985, 89, 389.
Joo, T.H.; Yim, Y.H.; Kim, K.; Kim, M.S. J. Phys. Chem.
1989, 93, 1422.
Garell, R.L.; Schultz, R.H. J. Colloid Inter. Sci. 1985,
105, 483.
Dawei, L.; Jian, W.; Houwen, X.; Xu, S.; Fan-Chen, L.
Spectrochim. Acta. 1987, 43A, 379.
Jian, W.; Dawei, L.; Houwen, X.; Xu, S.;
Fan-Chen, L. Spectrochim. Acta. 1987, 43A, 379.
Wetzel, H.; Gerischer, H. Chem. Phys. Lett. 1980, 76,
460.
Yamada, H.; Nagata, H.; Teranishi, H. J. Phys. Chem.
1986, 90, 2384.
Moskovits, M.; Suh, J.S. J. Phys. Chem. 1988, 92, 6327.
Kim, M.; Iton, K. J. Phys. Chem. 1987, 91, 126.
Suh, J.S.; Moskovits, M. J. Am. Chem. Soc. 1986, 108,
4711.
Kim, S.K.; Joo, T.H.; Suh, S.W.; Kim, M.S. J. Raman
Spectrosc. 1986, 17, 381.
Oh, W.S.; Suh, S.W. ; Kim, M.S. J. Raman Spectrosc. 1988,
19, 261.
Blatchford, C.G.; Siiman, O.; Kerker, M. J. Phys. Chem.
1983, 87, 2503.
Suh, J.S.; DiLella, D.P.; Moskovits, M. J. Phys. Chem.
1983, 87, 1540.
Park, H.; Lee, S.B.; Kim, K. ; Kim, M.S. J. Phys. Chem.
1990, 94, 7576.
Tran, C. Anal. Chem. 1984, 56, 824.
Tran, C. J. Chromatog. 1984, 292, 432.
Berthod, A.; Laserna, J.J.; Winefordner, J.D. J. Pharm.
Biomed. Anal. 1988, 6, 599.

135
53 Laserna, J.J.; Campiglia, A.D.; Winefordner, J.D. Anal.
Chem. 1989, 61, 1697.
54 Séquaris, J.-M.; Koglin, E. Fresenius Z. Anal. Chem.
1985, 321, 758.
55 Séquaris, J.-M.; Koglin, E. Anal. Chem. 1987, 59, 525.
56 Koglin, E. J. Mol. Struct. 1988, 173, 369.
57 Chang, R.K.; Furtak, T.K. Surface Enhanced Raman
Spectroscopy; Plenum PressrNew York, 1982.
58 Seki, H. J. Elec. Spectrosc. 1986, 39, 289.
59 Chang, R.K. Ber. Bunsenges. Phys. Chem. 1987, 91, 296.
60 Cotton, T.M. in Spectroscopy of Surfaces, Volume 16 of
Advances in Spectroscopy Series; Clark, R.J.H.; and
Hester, R.E., Eds.; Wiley:New York, 1988, Chapter 3.
61 Stetzel, J.M. in Raman Spectroscopy for Catalysis; Van
Nostrand ReinholdrNew York, 1990, Chapter 2.
62 Herzberg, G. Molecular Spectra and Molecular Structure;
Volume II, Kreiger: Malabar, FL (1991) Chapter 3.
63 Metiu, H. in Surface Enhanced Raman Spectroscopy; Chang,
R.K.; Furtak, T.K., Eds.; Plenum Press: New York, 1982.
64 Ford, G.W.; Weber, W.H. Phys. Rev. Lett. 1980, 44, 1774.
Lang, N.D.; Williams, A.R. Phys. Rev. B. 1978, 18, 616.
65 Gersten, J.; Nitzen, A. in Surface Enhanced Raman
Spectroscopy; Chang, R.K.; Furtak, T.K., Eds.;
Plenum Press: New York, 1982.
66 Murray, C. in Surface Enhanced Raman Spectroscopy; Chang,
R.K.; Furtak, T.K., Eds.; Plenum Press: New York, 1982.
67 Murray, C.A.; Aliara, D.L.; Rhinewine, M Phys. Rev. Lett.
1981,46, 57.
68 Mullins, D.R.; Campion, A. Chem. Phys. Lett. 1984, 110,
565.
69 Weitz, D.A.; Moskovits, M.; Creighton, J. in Chemistry
and Structural Interfaces: New Laser and Optical
Techniques; Hall, R.B.; Ellis, A.B., Eds.; VCH:New
York, 1983.
70 Loo, B.H.; Furtak, T.E. Chem. Phys. Lett. 1980, 71, 68.

136
71 Johnson, P.B.; Christy, R.W. Phys. Rev. B. 1972, 6, 4370.
72 Aussenegg, F.R.; Lippitch, M.E. Chem. Phys. Lett. 1978,
59, 214.
73 McCall, S.L.; Platzman, P.M. Phys. Rev. B. 1980, 22,
1660.
74 Effrima, S.; Metiu, H. J. Chem. Phys. 1979, 70, 2297.
75 Billman, J.; Otto, A. Appl. Surf. Sci. 1980, 6, 356.
76 Billman, J.; Kovacs, G. ; Otto, A. Surf. Sci. 1980, 92,
153.
77 Pockrand, I.; Otto, A. Solid State Comm. 1981, 38, 1159.
78 Wood, T.H.; Klein, M.V. J. Vac. Sci. Technol. 1979, 16,
459.
79 Mie, G. Ann. Phys., 1908, 25, 377.
80 Kerker, M. in Physical Chemistry and Colloids: The
Svedherg Symposium; Ránby, B. , Ed.; Blackwell Scientific
Publ.rOxford, 1987, Chapter 10.
81 Kerker, M. Pure and Appl. Chem. 1987, 56, 1429.
82 Blatchford, C.G.; Campbell, J.R.; Creighton, J.A. Surf.
Sci. 1982, 120, 435.
83 Rieger, P.H. in Electrochemistry; Prentice Hall,Inc.:
Englewood, NJ, 1987, p. 69-84.
84 Weitz, D.A.; Lin, M.Y.; Sandroff, C.J. Surf. Sci. 1985,
158, 147.
85 Devillez, C.; Sanfield, A.; Steiner, A. J. Colloid and
Interfac. Sci. 1978, 25, 295.
86 Bevill, R.F. in CRC Handbook Series in Zoonoses,
Volume 1, Section D: Antibiotics, Sulfamides, and Public
Health; CRC Press Inc.rBoca Raton, FL, 1984 p. 3,
30-31,and 355-365.
87 Sophian, L.H.; Piper, D.L.; Schneller, G.H. The
Sulfapyrimides; Press of A. ColishrNew York, 1952, P. 10.
88 Sigel, C.W.; Woolley, J.L. in Densitometry in Thin Layer
Chromatography; Touchstone, J.C.; Sherma, J., Eds;
Wiley:New York, 1979.

137
89 Bratton, A.C.; Marshall, E.K., jr. J. Biol. Chem. 1939,
128, 537.
90 Iskander, M.L.; Medien, H.A.A.; Nashed, S. Microchem. J.
1989, 39, 43.
91 Sato, S; Higuchi, S.; Tanaka, S. Anal. Chim. Acta 1980,
120, 209.
92 Yao, S.-Z.; Shiao, J.; Nie, L.-H. Talanta 1987, 34, 977.
93 Ruzika, J.; Hansen, E.H. Flow Injection Analysis, Volume
62 in the Chemical Analysis Series; Wiley: New York,
1981, p. 1-20.
94 Gersten, J.; Nitzan, A. J. Chem. Phys. 1980, 73, 3023.
95 Adrian, F. Chem. Phys. Lett. 1981, 78, 45.

BIOGRAPHICAL SKETCH
Martin John Angebranndt was born on July, 3 1964, in
Richmond, Va. to Martin William and Mary Angebranndt, just in
time for the fireworks. He has one sister, Melanie, three
years his junior. He holds the rank of Eagle Scout, and in
1982, graduated from Bishop Guertin H.S. in Nashua, N.H.
He graduated from the Rochester Institute of Technology,
in 1987, receiving a Bachelor of Science degree in chemistry.
Since then, he has attended the University of Florida, where
he received a Ph.D. degree in analytical chemistry in the
summer of 1991, working under the supervision of Dr. James D.
Winefordner. He was recently married to Chellyn Reid
Rinehart, and has taken a position in the U.S. Patent and
Trademark Office in Washington, D.C.
138

I certify that I have read this study and that in my
opinion it comforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
imes D. Winefordner, Chairman
Sraduate Research Professor of
Chemistry
I certify that I have read this study and that in my
opinion it comforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Vanecia Y. Young
Associate Professor £>f
I certify that I have read this study and that in my
opinion it comforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Jh
Anna Brajtjer-Toth
Associate iProfessor of Chemistry
I certify that I have read this study and that in my
opinion it comforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of ppctor of Philosophy.
Philip
Assista
ofessor of Chemistry
I certify that I have read this study and that in my
opinion it comforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
-
Richard Newman-Wolfe
Assistant Professor of Computer and
Information Sciences

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.
August, 1991
Dean, Graduate School

UNIVERSITY OF FLORIDA
262
08285 44
63




PAGE 1

685)$&( (1+$1&(' 5$0$1 63(&7526&23< 21 0(7$//,& &2//2,'6 )25 7+( 385326( 2) 75$&( $1$/<6,6 %< 0$57,1 -2+1 $1*(%5$11'7 $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

$&.12:/('*(0(176 ZRXOG ILUVW OLNH WR WKDQN 'U -DPHV :LQHIRUGQHU IRU KLV JXLGDQFH DQG VXSSRUW DP HVSHFLDOO\ JUDWHIXO IRU WKH FRQVLGHUDEOH IUHHGRP KH DOORZHG PH ZKLOH NHHSLQJ PH IURP VWUD\LQJ WRR IDU RII FRXUVH DP DOVR JUDWHIXO WR P\ FROOHDJXHV LQ -LPn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

PAGE 3

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

PAGE 4

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

PAGE 5

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

PAGE 6

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

PAGE 7

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f $ &URVVVHFWLRQ RI D 3URODWH 6SKHURLG ,OOXVWUDWLQJ WKH (OOLSWLFDO &RRUGLQDWH 6\VWHP Ef $Q ,OOXVWUDWLRQ RI WKH 0RGHO XVHG IRU WKH 'HULYDWLRQ LQ WKH $SSHQGL[ YLL

PAGE 8

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 685)$&( (1+$1&(' 5$0$1 63(&7526&23< 21 0(7$//,& &2//2,'6 )25 7+( 385326( 2) 75$&( $1$/<6,6 %< 0$57,1 -2+1 $1*(%5$11'7 $XJXVW &KDLUPDQ 'U -DPHV :LQHIRUGQHU 0DMRU 'HSDUWPHQW &KHPLVWU\ 7KH XVH RI 6XUIDFH (QKDQFHG 5DPDQ 6SHFWURVFRS\ 6(56f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

PAGE 9

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f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

PAGE 10

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

PAGE 11

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n ZKLFK LV GLIILFXOW ZLWK ,5 DEVRUSWLRQ VSHFWURVFRS\ EHFDXVH RI WKH SRRU VRXUFHV RSWLFV DQG GHWHFWRUV DYDLODEOH IRU WKH IDU LQIUDUHG UHJLRQ 2QH RI WKH PHWKRGV IRU LQFUHDVLQJ WKH TXDQWXP HIILFLHQF\ RI WKH 5DPDQ SURFHVV LV WR DGVRUE WKH PROHFXOH RI LQWHUHVW RQWR D VXUIDFH >@ 6XUIDFH (QKDQFHG 5DPDQ 6SHFWURVFRS\ 6(56f LQYROYHV ERWK HOHFWURPDJQHWLF HQKDQFHPHQWV LQ WKH ORFDO ILHOG DQG WKH FKHPLFDO HQKDQFHPHQWV LQ WKH PROHFXODU SRODUL]DELOLW\ WR LQFUHDVH WKH LQWHQVLW\ RI WKH 5DPDQ VFDWWHU 6HYHUDO VXUIDFHV PRVW QRWDEO\ JROG VLOYHU DQG FRSSHU SURYLGH HQKDQFHPHQWV RI XS WR VL[ RUGHUV RI PDJQLWXGH 7KLV HQKDQFHPHQW RYHU WKH 5DPDQ VFDWWHU IURP WKH IUHH PROHFXOH

PAGE 12

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

PAGE 13

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

PAGE 14

IRU D FRSSHU HOHFWURGH VXUIDFH DQG IRXU RUGHUV IRU D JROG HOHFWURGH VXUIDFH 7KHVH VWXGLHV LQFOXGHG PHDVXULQJ WKH DQJXODU GHSHQGHQFH RI WKH 6(56 VLJQDOV RQ HOHFWURGH VXUIDFHV 7KH RSWLPXP FROOHFWLRQ DQJOH IRU 6(56 PHDVXUHPHQWV ZDV IRXQG WR EH r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

PAGE 15

FRQGLWLRQV 7KH VLOYHU HOHFWURGH SRWHQWLDO FRQWUROOHG WKH DGVRUSWLRQ RI S\ULGLQH DQG TXLQROLQH 9DULDWLRQ RI WKH SRWHQWLDO DOORZHG GLVFULPLQDWLRQ EHWZHHQ WKHVH WZR FRPSRXQGV 7KH GHWHFWLRQ OLPLW IRU S\ULGLQH ZDV HVWLPDWHG WR EH SJ )RUF >@ UHSRUWHG WKH ILUVW DQDO\WLFDO DSSOLFDWLRQ RI 6(56 DW DQ HOHFWURGH VXUIDFH WR D IORZLQJ V\VWHP 7KH VLOYHU HOHFWURGH XQGHUZHQW D VKRUW 2[LGDWLRQ5HGXFWLRQ &\FOH 25&f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

PAGE 16

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f ZRXOG EH UHTXLUHG IRU WKH IRUPDWLRQ RI ODUJHU VXUIDFH IHDWXUHV 7KH S+ GHSHQGHQFH RI WKH 6(56 VLJQDO RI S\ULGLQH DGVRUEHG RQWR D FRSSHU HOHFWURGH ZDV PHDVXUHG E\ &RRQH\ DQG 0HUQDJK >@ 7KH\ UHSRUWHG WKDW DQ 25& ZDV QRW QHFHVVDU\ IRU WKH REVHUYDWLRQ RI 6(56 IURP D FRSSHU HOHFWURGH DV LW ZDV ZLWK VLOYHU DQG JROG VXUIDFHV 7KH\ DWWULEXWHG WKLV WR WKH VWURQJ FRRUGLQDWLRQ RI S\ULGLQH ZLWK WKH VXUIDFH 7KH RSWLPXP S+ IRU WKH REVHUYDWLRQ RI 6(56 IURP WKH S\ULGLQH FRSSHU HOHFWURGH V\VWHP ZDV IRXQG WR EH m (YLGHQFH RI ODVHU GDPDJH WR WKH HOHFWURGH VXUIDFH E\ WKH NU\SWRQ LRQ ODVHU DOVR ZDV GLVFXVVHG

PAGE 17

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f JURXSV LQ D PROHFXOH VLJQLILFDQWO\ LQFUHDVHG WKH 5DPDQ LQWHQVLW\ GXH WR WKHLU VWURQJ DVVRFLDWLRQ ZLWK WKH VXUIDFH 7KH 6(56 VSHFWUD RI DGHQLQH JXDQLQH WK\PLQH DQG F\WRVLQH RQ VLOYHU HOHFWURGHV ZHUH UHSRUWHG DQG FRPSDUHG E\ WKH VDPH JURXS >@ 'LIIHUHQWLDWLRQ EHWZHHQ WKHVH '1$ EDVHV LV SRVVLEOH IURP WKH IUHTXHQFLHV RI WKH EUHDWKLQJ PRGHV DORQH $ SDUWLDO DVVLJQPHQW RI WKH EDQGV ZDV DOVR SUHVHQWHG $ PRUH FRPSOHWH DVVLJQDWLRQ RI WKH EDQGV DULVLQJ IURP WKHVH '1$ EDVHV

PAGE 18

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nELS\ULG\O RQ WKLV VXUIDFH ZHUH UHSRUWHG ,Q D PL[WXUH RI S\ULGLQH DQG DQRWKHU RI WKH DQDO\WHV FRPSHWLWLYH DGVRUSWLRQ ZDV REVHUYHG DW WKH VXUIDFH $ 5DPDQ H[FLWDWLRQ SURILOH VLPLODU WR WKDW RI JROG ZDV REWDLQHG IRU S\ULGLQH 7KH PD[LPXP HQKDQFHPHQW REVHUYHG ZDV FDOFXODWHG WR EH ; b RI WKDW RI WKDW IRU D VLPLODU JROG K\GURVRO

PAGE 19

$QDO\WLFDO $SSOLFDWLRQV 7RUUHV >@ ZDV ILUVW WR XVH VLOYHU FROORLGV IRU DQDO\WLFDO WUDFH DQDO\VLV 6(56 VSHFWUD IRU SDPLQREHQ]RLF DFLG SKHQ\WRLQ DPLQRIOXRUHQH DV ZHOO DV XUDFLO DQG VRPH RI LWV GHULYDWLYHV ZHUH UHSRUWHG 7KH HIIHFWV RI GLIIHUHQW SUHSDUDWLRQV RQ WKH DQDO\WLFDO ILJXUHV RI PHULW RI DPLQRIOXRUHQH DQG SDPLQREHQ]RLF DFLG ZHUH DOVR VWXGLHG $OO RI WKH FROORLGV ZHUH SURGXFHG LQ EDWFKHV VWDWLF FRQGLWLRQVf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

PAGE 20

6LOYHU FROORLGV ZHUH XVHG LQ WKH WUDFH DQDO\VLV RI VXOID GUXJV E\ 6XWKHUODQG HW DO >@ /LPLWV RI GHWHFWLRQ ZHUH LQ WKH ORZ SSE UDQJH ZLWK WKH DGGHG DELOLW\ WR TXDOLWDWLYHO\ GLIIHUHQWLDWH EHWZHHQ VXOIDGLD]LQH VXOIDPHUD]LQH DQG VXOIDPHWKD]LQH /DVHUQD HW DO >@ ILUVW DSSOLHG FROORLGDO 6(56 WR )ORZ ,QMHFWLRQ $QDO\VLV ),$f )RU SDUDDPLQREHQ]RLF DFLG 3$%$f WKH OLPLW RI GHWHFWLRQ ZDV LQ WKH SSE ZKLOH D b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f IRU DSSOLFDWLRQ WR +3/& GHWHFWLRQ 7KHLU VWXG\

PAGE 21

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nV JURXS >@ UHFHQWO\ XVHG D PXOWLFKDQQHO DQDO\]HU WR GHWHFW WKH 6(56 VLJQDOV IURP IRXU SXULQH EDVHV ZKLFK KDG EHHQ VHSDUDWHG E\ +3/& $GHQLQH JXDQLQH [DQWKLQH DQG K\SR[DQWKLQH ZHUH VHSDUDWHG XVLQJ UHYHUVHSKDVH +3/& ,QMHFWLQJ LQFUHDVHG DPRXQWV RI DQDO\WH LQFUHDVHG WKH ZLGWK RI WKH SHDN EXW QRW WKH KHLJKW 'HWHFWLRQ OLPLWV LQ WKH VXEn QDQRPRODU UDQJH ZHUH UHSRUWHG 3K\VLFDO 6WXGLHV 6LOYHU K\GURVROV KDYH DOVR EHHQ XVHG E\ +HDUG HW DO >@ WR VWXG\ FRPSHWLWLYH DGVRUSWLRQ RQWR D VXUIDFH &RPSHWLWLRQ ZDV REVHUYHG EHWZHHQ S\ULGLQH DQG FLWUDWH IRU WKH VLOYHU VXUIDFH ,W ZDV IRXQG WKDW FKORULGH DVVLVWHG WKH S\ULGLQH LQ GLVSODFLQJ WKH PRUH VWURQJO\ DGVRUEHG FLWUDWH IURP

PAGE 22

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

PAGE 23

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

PAGE 24

WKH U ERQGV RI WKH ULQJ V\VWHP DQG WKH FDUER[\ODWH JURXS $W KLJKHU FRQFHQWUDWLRQV QDSKWKRLF DFLG ZDV RULHQWHG SHUSHQGLFXODU WR WKH VXUIDFH DGVRUEHG H[FOXVLYHO\ YLD WKH FDUER[\ODWH JURXS 7KH HIIHFW RI S+ RQ WKH DGVRUSWLRQ RI ELS\ULGLQH RQWR VLOYHU FROORLGV KDV DOVR EHHQ VWXGLHG .LP HW DO >@ REVHUYHG WKDW GHFUHDVLQJ WKH S+ XVLQJ +%U ZDV IRXQG WR LQFUHDVH WKH 6(56 VLJQDO IURP n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

PAGE 25

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f EDQG ZDV VKRZQ WR EH OLQHDUO\ GHSHQGHQW RQ WKH 3$%$ FRQFHQWUDWLRQ 7KH V\VWHP ZDV VWXGLHG DW S+ PDNLQJ WKH DQLRQ WKH SUHGRPLQDQW VSHFLHV 7KH PROHFXOH ZDV GHWHUPLQHG WR EH DGVRUEHG IODW RQ WKH VXUIDFH $ VLPLODU FRQFOXVLRQ ZDV UHDFKHG IRU 3$%$ DGVRUEHG RQWR D VLOYHU HOHFWURGH VXUIDFH E\ 3DUN HW DO >@

PAGE 26

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f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

PAGE 27

RI VLOYHU FROORLGV ZLWKLQ WKH ILOWHU SDSHU DV ZHOO DV XSRQ WKH VXUIDFH 6HYHUDO QLWUR FRQWDLQLQJ KHWHURF\FOHV ZHUH GHWHFWHG E\ /DVHUQD HW DO >@ RQ D VLOYHU FRDWHG ILOWHU SDSHU ERWK DORQH DQG LQ PL[WXUHV 7KLV SDSHU DOVR GHPRQVWUDWHG WKDW WKH FRPSRQHQWV RI D PL[WXUH FRXOG EH VSHFWURVFRSLFDOO\ UHVROYHG XVLQJ 6(56 'HWHFWLRQ OLPLWV LQ WKH ORZ SSP UDQJH ZHUH DFKLHYHG ZLWK DQ 56' RI b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

PAGE 28

WKH 6(56 HIIHFW EHIRUH 6HNLV UHYLHZ SURYLGHV D JRRG SODFH WR VWDUW D VHDUFK IURP EXW ODFNV WH[W GHVFULELQJ WKH REVHUYHG UHVXOWV &KDQJ >@ KDV DOVR SXEOLVKHG D UHSRUW RQ WKH VWDWXV RI 6(56 UHVHDUFK RI HOHFWURGH LQWHUIDFHV &RWWRQ >@ KDV UHFHQWO\ SXEOLVKHG D UHYLHZ RI 6(56 LQYHVWLJDWLRQV RI FRPSRXQGV RI ELRORJLFDO LQWHUHVW

PAGE 29

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f WKH HOHFWULF ILHOG ( 9Pf RI WKH ZDYH FDQ EH H[SUHVVHG DV ( (FRV Y/Wf f ZKHUH ( 9Pf LV WKH DPSOLWXGH RI WKH ZDYH )RU D YLEUDWLRQ DW IUHTXHQF\ Vff WKH SRODUL]DELOLW\ DW DQ\ WLPH W ZLOO EH DWf DOFRVQYWff f

PAGE 30

ZKHUH D &rP-f LV WKH PROHFXODU SRODUL]DELOLW\ 6LQFH WKH HOHFWULF ILHOG LQGXFHV D GLSROH PRPHQW Q &rPf LQ WKH PROHFXOH WKURXJK WKH PROHFXODU SRODUL]DELOLW\ 0 D( f WKH WLPH GHSHQGHQFH RI WKH LQGXFHG GLSROH PRPHQW LV JLYHQ E\ WKH VXSSRVLWLRQ RI WKH HTXDWLRQV DQG 0 D(FRV7L/Wf f D(FRV WW LOsMYf Wf f 7KH ILUVW WHUP FRUUHVSRQGV WR WKH OLJKW VFDWWHUHG HODVWLFDOO\ E\ WKH PROHFXOH 5D\OHLJK 6FDWWHUf WKH VHFRQG FRQWDLQV WKH DQWL6WRNHV DQG 6WRNHV 5DPDQ LQHODVWLFf VFDWWHULQJ WHUPV 7KH DQWL6WRNHV WHUP LV UHSUHVHQWHG E\ WKH VXPPDWLRQ RI WKH IUHTXHQF\ RI WKH ODVHU DQG WKH IUHTXHQF\ RI WKH YLEUDWLRQ 7KH 6WRNHV WHUP LV VKRZQ DV WKH GLIIHUHQFH EHWZHHQ WKHVH IUHTXHQFLHV 7KH SK\VLFDO UHVXOW LV WKH DSSHDUDQFH RI EDQGV RQ HLWKHU VLGH VHSDUDWHG E\ $L\ IURP WKH H[FLWLQJ IUHTXHQF\ L/f DV VKRZQ LQ ILJXUH 7KH RVFLOODWLQJ HOHFWULF ILHOG RI WKH LQFLGHQW UDGLDWLRQ PD\ EHFDXVH RI PROHFXODU V\PPHWU\ LQGXFH D GLSROH PRPHQW LQ D PROHFXOH ZKRVH D[LV LV QRW SDUDOOHO WR WKH DSSOLHG ILHOG

PAGE 31

)LJXUH Df $ VFKHPDWLF 5DPDQ VSHFWUXP VKRZLQJ WKH VWRNHV DQG DQWLVWRNHV 5DPDQ VFDWWHULQJ VKLIWHG IRU WKH ILUVW YLEUDWLRQDO WUDQVLWLRQ DQG $A IRU WKH VHFRQG YLEUDWLRQDO WUDQVLWLRQ UHODWLYH WR WKH 5D\OHLJK VFDWWHU DW L/ Ef $ VFKHPDWLF GLDJUDP LOOXVWUDWLQJ WKH RULJLQ RI WKH 5DPDQ VFDWWHUHG IUHTXHQFLHV

PAGE 32

Df /$6(5 /,1( $17,672.(6 6+,)7(' 672.(6 6+,)7(' Ef 9,578$/ /(9(/6 /$6(5 /,1( /$6(5 /,1( 02/(&8/$5 9,%5$7,21$/ /(9(/6 672.(6 6+,)7(' $17,672.(6 6+,)7('

PAGE 33

7KH UHODWLRQVKLS EHWZHHQ WKH LQGXFHG GLSROH DQG WKH DSSOLHG ILHOG LV EHVW VKRZQ E\ WKH PDWUL[ IRUP RI WKH SRODUL]DELOLW\ WHQVRU ; D D [[ [\ A[] ( ; 0\ D D \[ \\ D\] (\ 0] 2/ 2/ ] [ ]\ D]]B )RU VSRQWDQHRXV QRUPDOf 5DPDQ VFDWWHULQJ WKH SRODUL]DELOLW\ PDWUL[ LV UHDO DQG V\PPHWULF VR WKH VTXDUH PDWUL[ FDQ EH GLDJRQDOL]HG E\ FKRRVLQJ D SULQFLSDO D[LV 7KLV DWWULEXWH DOORZV RQH WR FDOFXODWH WKH PDJQLWXGH DQG RULHQWDWLRQ RI WKH LQGXFHG GLSROH 7KH LQWHQVLWLHV RI WKH 5DPDQ EDQGV FDQ EH FDOFXODWHG XVLQJ WKH IRUPXOD IRU WKH LQWHQVLW\ RI UDGLDWLRQ HPLWWHG E\ DQ RVFLOODWLQJ GLSROH U9Ff0 f ZKHUH F LV WKH VSHHG RI OLJKW ,I WKH 6WRNHV VKLIWHG OLQHV RULJLQDWH IURP WKH JURXQG YLEUDWLRQDO OHYHO WKHQ WKH DQWL 6WRNHV OLQHV PXVW RULJLQDWH IURP KLJKHU YLEUDWLRQDO OHYHOV 6LQFH WKH SRSXODWLRQV RI WKHVH YLEUDWLRQDO OHYHOV IROORZV D %ROW]PDQQ GLVWULEXWLRQ WKH UDWLR RI WKH UHODWLYH LQWHQVLWLHV RI WKH 6WRNHV DQG DQWL6WRNHV EDQGV FDQ EH H[SUHVVHG E\ ,V,DV /OYf fOnAYff H[SKrYN7f f

PAGE 34

7KH VHOHFWLRQ UXOHV IRU WKH KDUPRQLF RVFLOODWRU DOORZV RQO\ $} s DOWKRXJK RIWHQ RYHUWRQHV L\f DQG FRPELQDWLRQ EDQGV L\ s YZf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nO/f rH[S>LL/sLYfW@ f

PAGE 35

ZKHUH cLU &f Pf LV WKH LQGXFHG GLSROH DU &rP-f LV WKH SRODUL]DELOLW\ (/UL/f 9Pf UHSUHVHQWV WKH ILHOG JHQHUDWHG E\ WKH ODVHU DW U L/ V f WKH LQFLGHQW ODVHU IUHTXHQF\ DQG LY V f WKH YLEUDWLRQDO IUHTXHQF\ RI WKH PRGH )URP WKLV HTXDWLRQ RQH FDQ VHH WKH EHJLQQLQJV RI WKH WZR FODVVLILFDWLRQV RI WKHRULHV FRQFHUQLQJ WKH HYROXWLRQ RI WKH 6(56 HIIHFW 7KH ILUVW HQKDQFHPHQW PHFKDQLVP LV WHUPHG HOHFWURPDJQHWLF 7KLV PHFKDQLVP DWWULEXWHV WKH ODUJH HQKDQFHPHQWV WR ORFDO ILHOG HQKDQFHPHQWV GXH WR HLWKHU UHVRQDQFHV ZLWKLQ WKH VXUIDFH RU SRLQWV RI KLJK FXUYDWXUH 7KLV ZRXOG UHVXOW LQ D ODUJHU ILHOG WHUP (/UL/f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nV HTXDWLRQV PXVW KROG HYHQ DW GLVWDQFHV DV VPDOO DV N 7KH ODVW LV QHHGHG EHFDXVH WKH ILHOG IHOW E\

PAGE 36

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fHRfHA/fHR@ D f f ZKHUH D LV WKH GLDPHWHU RI WKH VSKHUH HLf LV WKH GLHOHFWULF IXQFWLRQ RI WKH PHWDO DQG H LV WKH GLHOHFWULF IXQFWLRQ RI WKH VXUURXQGLQJ PHGLXP (DFK RI WKHVH WHUPV LV FRPSRVHG RI D UHDO HX DQG LPDJLQDU\ H FRPSRQHQW DV VKRZQ LQ HTXDWLRQ EHORZ LH f ,I WKH IUHTXHQF\ RI WKH ODVHU LV QHDU WKH UHVRQDQFH IUHTXHQF\ RI D VSKHUH WKH HQKDQFHPHQW LV LQYHUVHO\ SURSRUWLRQDO WR VTXDUH RI WKH LPDJLQDU\ FRPSRQHQW RI WKH GLHOHFWULF IXQFWLRQ RI WKH PHWDO VXUIDFH >H=P@r DW WKDW IUHTXHQF\

PAGE 37

,W LV LPSRUWDQW WR KDYH D VPDOO LPDJLQDU\ FRPSRQHQW IRU WKH GLHOHFWULF FRQVWDQW ZKHQ WKH UHVRQDQFH FRQGLWLRQ >HLAf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f 5DPDQ DQG 6(56 SURFHVVHV 3URODWH VSKHURLGV FDQ EH XVHG WR WKHRUHWLFDOO\ PRGHO FROORLGV ZKLFK KDYH XQGHUJRQH DJJORPHUDWLRQ 7KH HOHFWURPDJQHWLF HTXDWLRQV DUH VLPLODU WR WKRVH IRU D VSKHUH EXW LOOXVWUDWH WKH GLIIHUHQFH EHWZHHQ RI WKH ORFDO ILHOG DW WKH HQG DQG WKH VLGHV RI WKH VSKHURLG )RU WKH UHVRQDQFH FRQGLWLRQ HL/f Hf WKH ORFDO ILHOG DW WKH WLS RI WKH VSKHURLG LV HQKDQFHG E\ WKH VDPH DPRXQW DV D VSKHUH _HOH_ f 7KH HQKDQFHPHQW RQ WKH VLGH RI WKH VSKHURLG ZLOO EH UHGXFHG f IURP WKLV E\ D IDFWRU RI H 7KH WLS DOVR DOORZV WKH HIILFLHQW FRXSOLQJ RI WKH VFDWWHUHG 5DPDQ UDGLDWLRQ ZLWK WKH

PAGE 38

)LJXUH 7KH UHDO H VROLG V\PEROVf DQG LPDJLQDU\ H RSHQ V\PEROVf FRPSRQHQWV RI WKH GLHOHFWULF IXQFWLRQV RI FRSSHU FLUFOHVf DQG VLOYHU WULDQJOHVf >@

PAGE 39

HQHUJ\ H9f 0DJQLWXGH RI WKH 'LHOHFWULF )XQFWLRQ B !‘ .f Z ) FQ 2, aQ2 8! R

PAGE 40

)LJXUH $ VFKHPDWLF GLDJUDP LOOXVWUDWLQJ WKH GLIIHUHQFH EHWZHHQ WKH QRUPDO 5DPDQ VFDWWHULQJ SURFHVV DQG WKH 6(56 SURFHVV 7KH LQFLGHQW IUHTXHQF\ LV GHQRWHG E\ Y ZKLOH WKH 5DPDQ VKLIWHG IUHTXHQF\ LV UHSUHVHQWHG E\ Y 7KH $ WHUPV UHSUHVHQW WKH DPSOLILFDWLRQ RI WKH LQFLGHQW ILHOG ( E\ WKH VSKHURLG

PAGE 41

8f WR

PAGE 42

VXUIDFH VR WKDW LI L5 } Z/ WKHQ WKH VFDWWHUHG UDGLDWLRQ XQGHUJRHV DGGLWLRQDO DPSOLILFDWLRQ ,I WKLV RFFXUV D PROHFXOH LQWHUDFWLQJ ZLWK WKH VSKHURLG WLS ZLOO H[SHULHQFH DQ HQKDQFHPHQW RI _HOH f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

PAGE 43

)LJXUH $ SORW VKRZLQJ WKH FDOFXODWHG HQKDQFHPHQW IRU 6(56 RQ VLOYHU VSKHURLGV DV D IXQFWLRQ RI SRODU DQJOH IURP WKH FHQWHU RI WKH VLOYHU VSKHURLG GHJUHHV LV WKH WLS RI WKH SURODWH VSKHURLG DQG GHJUHHV LV WKH ZDLVW DQG FRUUHVSRQG WR WKH HFFHQWULFLW\ RI WKH VSKHURLG DEf ZKHUH WKH PLQRU D[LV E LV KHOG FRQVWDQW DW QP >@

PAGE 44

3RODU $QJOH GHJUHHVf (QKDQFHPHQW )DFWRU DUE XQLWVf

PAGE 45

)LJXUH $ SORW VKRZLQJ WKH FDOFXODWHG HIIHFW RI VHSDUDWLRQ GLVWDQFH RQ WKH 6(56 HQKDQFHPHQW IRU VLOYHU VSKHURLGV DQG GHQRWH WKH HFFHQWULFLW\ RI WKH VSKHURLG DEf ZKHUH WKH PLQRU D[LV E LV KHOG FRQVWDQW DW QP >@

PAGE 46

6HSDUDWLRQ 'LVWDQFH ƒf (QKDQFHPHQW )DFWRU DUE XQLWVf 2 UR ) FU! FR ) /=

PAGE 47

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

PAGE 48

HOHFWURQ KROH SDLUV UHTXLULQJ DGDWRP W\SH VXUIDFH GHIHFWV RU FKHPLVRUSWLRQ LQGXFHG UHVRQDQFH >@ $OO RI WKHVH PRGHOV DUH VKRUW UDQJH PHFKDQLVPV SUHGLFWLQJ QR HQKDQFHPHQW SDVW WKH ILUVW RQH RU WZR PRQROD\HUV DGVRUEHG RQWR WKH VXUIDFH 7KLV DSSHDUV WR EH WKH FDVH IRU DQ H[SHULPHQW LQYROYLQJ F\DQLGH DQG S\ULGLQH >@ 7KH 6(56 EDQGV IURP WKHVH DGVRUEDWHV ZHUH HQKDQFHG E\ XS WR VL[ RUGHUV RI PDJQLWXGH 7KH VSHFWUD ZHUH KLJKO\ SRWHQWLDO GHSHQGHQW H[KLELWLQJ 6(56 RQO\ IRU WKH PROHFXOHV WKDW ZHUH VWURQJO\ DGVRUEHG WR WKH HOHFWURGH VXUIDFH &RPSHWLWLYH DGVRUSWLRQ VLPLODU WR WKLV KDV EHHQ REVHUYHG RQ VLOYHU >@ DQG FRSSHU >@ FROORLGV DV ZHOO 2I WKHVH PHFKDQLVPV RQO\ WKH DGDWRP PRGHO RI 2WWR HW DO >@ KDV VWRRG WKH WHVW RI WLPH ZHOO 7KH FKDUJH WUDQVIHU DQG FKHPLVRUSWLRQ LQGXFHG UHVRQDQFH PRGHOV GR QRW VSHFLI\ WKH QHFHVVLW\ RI DFWLYH VLWHV >@ 7KH HQKDQFHPHQW IURP VPRRWK VXUIDFHV LV VPDOO DQG FDQ EH DFFRXQWHG IRU E\ HOHFWURPDJQHWLF PHFKDQLVPV ZLWKRXW WKH QHHG WR LQYRNH WKH FKHPLFDO PRGHOV >@ 7KH VPDOO HQKDQFHPHQW REVHUYHG RQ VPRRWK VXUIDFHV LV FRPPRQO\ XVHG WR GLVFRXQW WKH HIIHFWV RI FKHPLFDO HQKDQFHPHQW PHFKDQLVPV 7KH DGDWRP PRGHO HPSOR\V VXUIDFH GHIHFWV WR EUHDN WKH WUDQVODWLRQDO V\PPHWU\ SDUDOOHO WR WKH VXUIDFH 7KHVH YDULDWLRQV LQ VXUIDFH PRUSKRORJ\ PHGLDWH WKH FRXSOLQJ EHWZHHQ SKRWRQV DQG SODVPRQV 7KLV UHGXFWLRQ LQ WKH V\PPHWU\ RI WKH VXUIDFH LV DOVR XVHG LQ WKH HOHFWURPDJQHWLF PRGHO 7KH PDMRU

PAGE 49

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f DWWULEXWHG WR SK\VLVRUEHG S\ULGLQH ZDV GHWHFWHG )RU D VLOYHU IDFH WKH EDQG DW FP ZDV VKLIWHG WR FP GXH WR FKHPLVRUSWLRQ ,Q DOO FDVHV WKLV EDQG ZDV

PAGE 50

XQHQKDQFHG ZKLFK FDVWV GRXEW RQ WKH LPSRUWDQFH RI DWRPLF VL]HG GHIHFWV SURSRVHG LQ WKH DGDWRP PRGHO 7KHUH LV DOVR VRPH GDWD RQ FROORLGV ZKLFK LV LQ FRQIOLFW ZLWK WKH DGDWRP PRGHOnV XVH RI VPDOO VXUIDFH GHIHFWV &RWWRQ HW DO >@ KDYH UHFHQWO\ EHHQ XVLQJ HOHYDWHG WHPSHUDWXUHV XS WR r &f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

PAGE 51

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f = QOf ^5HDO DQEQf ` f 4VF Df ( QOf ^ DQ _EQ_` f 4DE f§ 4H[W f§ 4VF f ZKHUH 4H[W LV WKH H[WLQFWLRQ HIILFLHQF\ 4VF LV WKH VFDWWHULQJ HIILFLHQF\ 4DE LV WKH DEVRUSWLRQ HIILFLHQF\ D LV D VL]H SDUDPHWHU HTXDO WR "UD$f DQG DQ DQG EQ UHSUHVHQW D VHULHV RI

PAGE 52

FRHIILFLHQWV GHSHQGHQW RQ WKH VL]H RI WKH SDUWLFOH DQG WKH GLHOHFWULF FRQVWDQW RI WKH VXUIDFH UHODWLYH WR WKDW RI WKH PHGLXP )RU YHU\ VPDOO SDUWLFOHV ZLWK UDGLL OHVV WKDQ RI WKH LQFLGHQW ZDYHOHQJWK WKHVH HTXDWLRQV EHFRPH 4VFD 6DFf^ HOfHf _` f 4DE D^,P^HOfHf `` f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f XVHV D GLIIXVLRQ VL]H PRGHO WR GHWHUPLQH SDUWLFOH VL]H &RUUHFWLRQV FDQ HDVLO\ EH PDGH IRU WKH UHIUDFWLYH LQGH[ RI WKH SDUWLFOH DW WKH

PAGE 53

)LJXUH &DOFXODWHG H[WLQFWLRQ VSHFWUD RI QP VSKHUHV RI VLOYHU >@ DQG FRSSHU >@ LQ ZDWHU

PAGE 54

5HODWLYH $EVRUEDQFH XL

PAGE 55

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f IRU D VSKHULFDO SDUWLFOH RI UDGLXV D LV JLYHQ E\ 8DGf -O > rOGr@ Q J"D! @ G^GDf F"Df f ZKHUH G LV WKH GLVWDQFH EHWZHHQ WKH VXUIDFHV D LV WKH UDGLXV RI WKH SDUWLFOH DQG $ LV DQ HPSLULFDOO\ GHWHUPLQHG FRQVWDQW f f $ LV XVXDOO\ RQ WKH RUGHU RI )URP WKLV HTXDWLRQ RQH FDQ H[SHFW XQFKDUJHG SDUWLFOHV WR DVVRFLDWH DQG HYHQWXDOO\ SUHFLSLWDWH IURP VROXWLRQ 6LQFH WKH PDMRULW\ RI FROORLGV KDYH FKDUJHG VXUIDFHV WKHUH DUH DOVR UHSXOVLYH IRUFHV WR FRQVLGHU 7KH FKDUJH RQ WKH VXUIDFH LV XVXDOO\ GXH WR WKH DGVRUSWLRQ RI LRQV RQWR WKH VXUIDFH 7KHUH LV DOVR D GLIIXVH OD\HU ZKLFK VKLHOGV WKH

PAGE 56

VXUIDFH FKDUJHV IURP HDFK RWKHU 7KLV FRQWUROV KRZ QHDU VLPLODUO\ FKDUJHG SDUWLFOHV FDQ DSSURDFK HDFK RWKHU 7KH H[WHQVLRQ RI WKLV OD\HU LQWR WKH VXUURXQGLQJ VROXWLRQ LV JLYHQ E\ !Gf !4DGf H[SDG;Df f ZKHUH 9f LV WKH SRWHQWLDO RI WKH VXUIDFH DQG ;D Pf LV D GLVWDQFH SDUDPHWHU 7KH WKLFNQHVV RI WKH GLIIXVH OD\HU LV UHODWHG WR WKH LRQLF VWUHQJWK RI WKH VROXWLRQ WKURXJK WKH H[SUHVVLRQ D ) f ZKHUH H LV WKH GLHOHFWULF FRQVWDQW H LV WKH SHUPLWWLYLW\ RI IUHH VSDFH ) &PROf LV )DUDGD\nV FRQVWDQW DQG PRO/f LV WKH LRQLF VWUHQJWK RI WKH VXUURXQGLQJ PHGLXP )URP H[SUHVVLRQV DQG ZH VHH WKDW LQFUHDVLQJ WKH LRQLF VWUHQJWK RI WKH VROXWLRQ ZLOO FRPSUHVV WKH GRXEOH OD\HU UHVXOWLQJ LQ LQFUHDVHG DJJUHJDWLRQ 7KLV DFFRXQWV IRU WKH VDOWLQJ RXW HIIHFW 7KH UHSXOVLYH IRUFHV 8U EHWZHHQ WZR SDUWLFOHV ZLWK LRQ DWPRVSKHUHV VXUURXQGLQJ WKHP LV GLIILFXOW WR FDOFXODWH 7KH WUHQGV REVHUYHG IROORZ WKH IRUP QD R 8]Gf D[Df HH ,Q H[S G[rf f

PAGE 57

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

PAGE 58

SDUWLFOHV \LHOGHG D ORJDULWKPLF SORW ZLWK D VORSH RI RU 7KH IUDFWDO DSSURDFK WR TXDQWLI\LQJ FROORLGDO DJJUHJDWLRQ VHHPV SURPLVLQJ VLQFH LW LV DFFXUDWH DQG SURYLGHV UHVXOWV ZKLFK DUH HDVLO\ YLVXDOL]HG

PAGE 59

&+$37(5 6,/9(5 &2//2,'6 )25 68/)$ '58* $1$/<6,6 ,QWURGXFWLRQ 7KH GHYHORSPHQW RI GUXJV IRU WKHUDSHXWLF XVH ERWK LQ KXPDQV DQG DQLPDOV KDV LQFUHDVHG GUDPDWLFDOO\ RYHU WKH ODVW FHQWXU\ 2YHU NJ RI RQH SDUWLFXODU FODVV VXOID GUXJV ZHUH DGPLQLVWHUHG LQ 2I WKH VXOID GUXJV WHVWHG IRU DQWLEDFWHULDO DFWLYLW\ OHVV WKDQ DUH URXWLQHO\ XVHG >@ 6XOIDS\ULPLGLQHV DUH DQ LPSRUWDQW VXEFODVV RI WKHVH VXOIRQDPLGHV VXOID GUXJVf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

PAGE 60

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

PAGE 61

0HPEUDQH HOHFWURGHV ZHUH SUHSDUHG E\ @ DQG XVHG IRU WKH GHWHUPLQDWLRQ RI 6XOID GUXJV 7KHVH PHPEUDQHV ZHUH VSHFLILF IRU VXOID GUXJV ZLWK GHWHFWLRQ OLPLWV LQ WKH 0 UDQJH EXW ZHUH QRW DEOH WR GLIIHUHQWLDWH EHWZHHQ PROHFXOHV RI WKH VDPH FODVV 6(56 RQ FROORLGDO VLOYHU VXEVWUDWHV ZLOO EH VKRZQ WR SURYLGH TXDOLWDWLYH DQG TXDQWLWDWLYH LQIRUPDWLRQ IRU VXOIDGLD]LQH VXOIDPHWKD]LQH DQG VXOIDPHUD]LQH DW OHYHOV LQ WKH ORZ SSE UDQJH ([SHULPHQWDO $SSDUDWXV $OO RI WKH VSHFWUD ZHUH REWDLQHG E\ LUUDGLDWLQJ WKH VDPSOHV ZLWK P: RI QP OLQH RI DQ DUJRQ LRQ ODVHU 6SHFWUD 3K\VLFV VHULHV f 7KH VFDWWHUHG UDGLDWLRQ ZDV FROOHFWHG DW GHJUHHV DQG UHVROYHG XVLQJ D P GRXEOH PRQRFKURPDWRU 6SH[ ,QGXVWULHV PRGHO f ZLWK WKH EDQGSDVV VHW DW FP 7KH GHWHFWRU ZDV D FRROHG SKRWRPXOWLSOLHU WXEH 5&$ PRGHO &f FRXSOHG WR SKRWRQ FRXQWLQJ HOHFWURQLFV 7KH DFTXLVLWLRQ VWRUDJH DQG SURFHVVLQJ RI WKH GDWD ZDV FDUULHG RXW RQ DQ ,%0 FRPSDWLEOH 3& $OO VSHFWUD SUHVHQWHG DUH VLQJOH VFDQV ZLWKRXW DQ\ VSHFWUDO VPRRWKLQJ &KHPLFDOV DQG 3URFHGXUH $OO FKHPLFDOV ZHUH UHDJHQW JUDGH RU EHWWHU DQG XVHG ZLWKRXW IXUWKHU SXULILFDWLRQ 'HPLQHUDOL]HG ZDWHU ZDV XVHG WKURXJKRXW WKH DQDO\VHV 7KH VXOID GUXJV DQDO\]HG VXOIDGLD]LQH VXOIDPHWKD]LQH DQG VXOIDPHUD]LQH ZHUH

PAGE 62

SXUFKDVHG IURP 6LJPD &KHPLFDOV 7KH SSP VWRFN DTXHRXV VROXWLRQV ZHUH SUHSDUHG ZLWK b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

PAGE 63

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

PAGE 64

)LJXUH 7KH QRUPDO 5DPDQ VSHFWUD RI WKH WKUHH VXOID GUXJV SUHSDUHG DV .%U SHOOHWV 7KH VSHFWUXP Df VXOIDGLD]LQH Ef VXOIDPHUD]LQH Ff VXOIDPHWKD]LQH 7KH SHDN SRVLWLRQV DQG UHODWLYH LQWHQVLWLHV DUH JLYHQ LQ WDEOH

PAGE 65

5DPDQ 6KLIW FPf§f 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf

PAGE 66

7DEOH 5DPDQ %DQGV RI AXOID 'UXJV LQ 6ROLG 3KDVHf 5DPDQ 6KLIW FP f VXOIDGLD]LQH VXOIDPHWKD]LQH VXOIDPHUD]LQH VK Z YZ Z YZ GEOZ YZ YZ GEOZ YZ YZ Z YZ YZ YZ Z Z YZ YZ YZ Z Z Z YZ Z P Z YZ Z YZ Z Z Z Z YZ YZ Z Z YZ YZ YZ YZ P Z GEOP GEOP GEOP GEOP GEOP GEOP Z YZ YZ YZ P P P V 6 Z YV 96 YV : Z YZ YZ Z YZ Z Z YZ YZ YZ YZ Z YZ YZ YZ Z YZ YZ P Z VK Z YZ YV YV YV Z Z EE EURDG EDQG VK VKRXOGHU P PHGLXP V VWURQJ Z ZHDN Y YHU\ ZHDN

PAGE 67

)LJXUH 7KH 6(56 VSHFWUD RI WKH WKUHH VXOID GUXJV DW SSE FRQFHQWUDWLRQV LQ WKH FROORLGDO VLOYHU VROXWLRQ Df VXOIDGLD]LQH Ef VXOIDPHUD]LQH Ff VXOIDPHWKD]LQH 7KH SHDN SRVLWLRQV DQG UHODWLYH LQWHQVLWLHV DUH JLYHQ LQ WDEOH

PAGE 68

5DPDQ 6KLIW FP f 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf

PAGE 69

7DEOH 6XUIDFH (QKDQFHG 5DPDQ %DQGV RI 6XOID 'UXJV RQ &ROORLGDO 6LOYHU 5DPDQ 6KLIW FP f VXOIDGLD]LQH VXOIDPHUD]LQH VXOIDPHWKD]LQH VK VK VK VK VK YZ YZ YZ YZ YZ YZ YZ YZ YZ YZ YZ YZ Z P P P Z YZ YZ Z Z Z V P V P P YZ YZ EE YZ VK VK Z YV YV YV YZ YZ YZ YZ YZ P YZ P VK P P Z Z : Z Z V V V VK VK VK P P VK VK VK EE EE EE EE EE EE EE EURDG EDQG VK VKRXOGHU P PHGLXP V VWURQJ Z n ZHDN Y YHU\

PAGE 70

)URP WKH FKHPLFDO VWUXFWXUHV VKRZQ LQ ILJXUH WKH VLPLODULW\ DQG GLIIHUHQFHV EHWZHHQ WKHVH FRPSRXQGV LV HYLGHQW 7KH 6(56 VSHFWUD UHIOHFW WKHVH VLPLODULWLHV DQG GLIIHUHQFHV DV ZHOO 7KH FP EDQG ZKLFK LV SUHVHQW ZLWK GLIIHUHQW LQWHQVLWLHV IRU ERWK VXOIDPHWKD]LQH DQG VXOIDPHUD]LQH LV DEVHQW LQ WKH VXOIDGLD]LQH VSHFWUXP 7KH FPf EDQG LV GXH WR WKH PHWK\O JURXSV RQ WKH S\ULPLGLQH ULQJ 7KH ORFDWLRQV RI WKH VWURQJHU SHDNV DOVR GLIIHU VOLJKWO\ 7KH V\PPHWULF VWUHWFK RI WKH 6 JURXS LV ORFDWHG DW FP IRU VXOIDGLD]LQH DQG VXOIDPHUD]LQH EXW LV IRXQG DW FP IRU f VXOIDPHWKD]LQH 6LPLODUO\ D VKRXOGHU LV VHHQ DW FP IRU VXOIDPHWKD]LQH EXW LV VKLIWHG WR FP LQ WKH RWKHU VSHFWUD 7KHUH DUH DOVR D QXPEHU RI SHDNV ZKLFK DSSHDU LQ RQH RI WKH VSHFWUD EXW QRW WKH RWKHUV )LJXUH LOOXVWUDWHV WKH DELOLW\ RI 6(56 WR SURYLGH TXDQWLWDWLYH LQIRUPDWLRQ DW SSE OHYHOV 7KH EDQGV DW FPf DQG FP DUH DWWULEXWHG WR YLEUDWLRQV RI WKH EHQ]HQH ULQJ 9DULDWLRQV LQ WKH LQWHQVLWLHV RI WKHVH EDQGV PD\ EH GXH WR VOLJKW RUHLQWDWLRQDO FKDQJHV RI WKH PROHFXOH UHODWLYH WR WKH VXUIDFH 7KLV SUHVHQWV D WUXH WHVW RI WKH ILQJHUSULQWLQJ FDSDELOLW\ RI 6(56 7KH LQWHQVLW\ RI WKH SHDN DW FPn LQ WKH 6(56 VSHFWUXP RI VXOIDPHWKD]LQH LV XVHIXO IRU GHWHUPLQLQJ FRQFHQWUDWLRQV RI DQDO\WH EHWZHHQ DQG SSE VHH 7DEOH f 7KLV EDQG ZDV XVHG IRU DOO RI WKH DQDO\WLFDO GHWHUPLQDWLRQV EHFDXVH LW ZDV OLQHDU ZLWK FRQFHQWUDWLRQ RYHU VHYHUDO RUGHUV RI PDJQLWXGH 7KHVH UHVXOWV KHOS WR HQVXUH

PAGE 71

)LJXUH 6WUXFWXUHV RI WKH WKUHH VXOID GUXJV VWXGLHG VXOIDGLD]LQH VXOIDPHUD]LQH DQG VXOIDPHWKD]LQH

PAGE 72

6XOIDGLD]LQH 6XOIDPHUD]PH 6XOIDPHWKD]LQH

PAGE 73

)LJXUH 7KH 6(56 VSHFWUXP RI VXOIDPHWKD]LQH RQ FROORLGDO VLOYHU DW WKUHH GLIIHUHQW FRQFHQWUDWLRQV Df SSE Ef SSE Ff SSE

PAGE 74

5DPDQ 6KLIW FPf§f 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf 6

PAGE 75

7DEOH $QDO\WLFDO )LJXUHV &ROORLGDO 6\VWHP RI 0HULW IRU 6XOID 'UXJV RQ WKH 6LOYHU &RPSRXQGV /2'nV 6HQVLWLYLW\ 5 33Ef SSP f VXOIDGLD]LQH VXOIDPHUD]LQH VXOIDPHWKD]LQH D /2' LV WKH OLPLW RI GHWHFWLRQ XVLQJ WKUHH VWDQGDUG GHYLDWLRQV DERYH WKH QRLVH WKH VHQVLWLYLW\ LV WKH VORSH RI WKH FDOLEUDWLRQ FXUYH DQG 5 LV WKH FRUUHODWLRQ FRHIILFLHQW RI WKH ORJORJ SORW

PAGE 76

WKDW 6(56 ZLOO HYHQWXDOO\ SURYLGH KLJK TXDOLW\ TXDOLWDWLYH DQG TXDQWLWDWLYH LQIRUPDWLRQ DV D PHWKRG RI URXWLQH DQDO\VLV

PAGE 77

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

PAGE 78

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

PAGE 79

([SHULPHQWDO $SSDUDWXV ,Q )LJXUH D VFKHPDWLF GLDJUDP RI WKH H[SHULPHQWDO VHWXS LV VKRZQ $ 6SHFWUD 3K\VLFV DUJRQ LRQ ODVHU ZDV XVHG WR SXPS D G\H ODVHU 6SHFWUD3K\VLFV PRGHO f '&0 ([FLWRQf ZDV XVHG WR SURGXFH WKH QHFHVVDU\ ZDYHOHQJWKV $OO PHDVXUHPHQWV ZHUH PDGH UHODWLYH WR WKH P: RI QP H[FLWLQJ UDGLDWLRQ 7KH VWDELOLW\ RI WKH RXWSXW SRZHU DQG ODVLQJ IUHTXHQF\ ZHUH FRQWLQXRXVO\ PRQLWRUHG XVLQJ ERWK D SRZHU PHWHU DQG D P PRQRFKURPHWHU 0F3KHUHVRQ f ZLWK D 5 +DPDPDWVX SKRWRPXOWLSOLHU WXEH 7KH ODVHU EHDP ZDV GLUHFWHG WKURXJK WKH VDPSOH WZLFH DQG WKH VFDWWHUHG UDGLDWLRQ FROOHFWHG DW GHJUHHV $ P GRXEOH PRQRFKURPHWHU 6SH[ %f ZLWK VOLWV VHW IRU FP UHVROXWLRQ ZDV XVHG DW D VFDQ UDWH RI QPPLQ IRU ZDYHOHQJWK GLVFULPLQDWLRQ 7KH 5DPDQ VFDWWHU ZDV GHWHFWHG XVLQJ D WKHUPRHOHFWULFDOO\ FRROHG 5 +DPDPDWVXf SKRWRPXOWLSOLHU WXEH 7KH SKRWRHOHFWURQ SXOVHV ZHUH DPSOLILHG DQG FRXQWHG XVLQJ D 6WDQIRUG PRGHO f SKRWRQ FRXQWHU ZLWK D V DFTXLVLWLRQ WLPH 7KH GDWD ZDV FROOHFWHG RQ DQ ,%0 3& 7KH 5DPDQ EDQGV XVHG IRU TXDQWLWDWLYH PHDVXUHPHQWV ZHUH DQG FP IRU S\ULGLQH S DPLQREHQ]RLF DFLG DQG SQLWUREHQ]RLF DFLG UHVSHFWLYHO\

PAGE 80

)LJXUH $ VFKHPDWLF GLDJUDP RI WKH H[SHULPHQWDO OD\RXW RI WKH RSWLFDO FRPSRQHQWV XVHG LQ WKH FRSSHU FROORLG VWXGLHV

PAGE 81

0212&+520$725 0,5525 ? 0,5525 $5*21 ,21 /$6(5 '<( /$6(5 0(7(5 %($0 63/,77(5 6$03/( &+$0%(5 5(&25'(5 0,5525 3+2721 &2817(5 0212&+520$725 >V0LXOM SXQ? &20387(5 WR

PAGE 82

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f PHWKDQRO DQG VRGLXP ERURK\GULGH ZHUH REWDLQHG IURP .RGDN 5RFKHVWHU 1
PAGE 83

UHVXOWV >@ 3\ULGLQH 3ND f SURYHG D JRRG FKRLFH EHFDXVH RI WKH UHODWLYHO\ VPDOO HIIHFW WKDW S+ KDG XSRQ LWV 6(56 VLJQDO FRPSDUHG WR WKRVH RI WKH RWKHU DQDO\WHV ILJXUH f 7KH RSWLPDO S+ IRU WKH REVHUYDWLRQ RI 6(56 IURP S\ULGLQH GLIIHUV IURP WKDW SUHYLRXVO\ REVHUYHG XVLQJ D FRSSHU HOHFWURGH >@ DQG D GLIIHUHQW FRSSHU FROORLG >@ 7KH RSWLPDO S+ IRU WKH FRSSHU HOHFWURGH DQG WKH FRSSHU FROORLG XVHG LQ WKLV VWXG\ GLIIHU E\ m XQLWV ZKLFK PD\ EH GXH WR GLIIHUHQFHV LQ WKH VXUIDFH PRUSKRORJ\ RU SRWHQWLDO 7KH S+ GLIIHUHQFH EHWZHHQ WKH FRSSHU FROORLGV LQ WKH SUHVHQW VWXG\ DQG WKRVH XVHG E\ &UHLJKWRQ HW DO >@ LV DERXW XQLWV 7KLV LV SUREDEO\ GXH WR WKHLU XVH RI FLWUDWH WR VWDELOL]H WKH FROORLG 7KH DGVRUEHG FLWUDWH ZLOO DIIHFW ERWK WKH VXUIDFH SRWHQWLDO RI WKH FROORLG DQG WKH DGVRUSWLRQ RI WKH S\ULGLQH RQWR WKH VXUIDFH 7KH GLIIHUHQFH LQ WKH S+ GHSHQGHQFH RI WKH 6(56 LQWHQVLWLHV RI 31%$ 3ND f DQG 3$%$ 3ND f DUH GXH WR WKH GLIIHUHQFHV EHWZHHQ WKHLU GLVVRFLDWLRQ FRQVWDQWV 7KH DFLGLF IRUP RI WKH DPLQR JURXS RQ 3$%$ ZLOO DGVRUE PRUH VWURQJO\ WKDQ WKH EDVLF IRUP $W VOLJKWO\ EDVLF S+nV 3$%$ LV ERXQG WR WKH VXUIDFH YLD ERWK IXQFWLRQDO JURXSV 7KH HIIHFW RI S+ RQ 31%$ LV SUREDEO\ GXH PDLQO\ WR WKH VXUIDFH FKDUJH RI WKH FROORLG 7KH QDUURZ S+ UDQJHV ZLWKLQ ZKLFK 3$%$ DQG 31%$ H[KLELW 6(56 SUHVHQWV DQ DGGLWLRQDO PHWKRG IRU GLIIHU HQWLDWLQJ EHWZHHQ VLPLODU FRPSRXQGV 7KH DEVRUSWLRQ VSHFWUXP RI WKH FROORLG RYHU WLPH ILJXUH f VKRZV WKH JURZWK RI D SHDN DW QP ,W KDV EHHQ

PAGE 84

)LJXUH 7KH 5DPDQ VSHFWUD IRU S\ULGLQH Df D b DTXHRXV VROXWLRQ RI S\ULGLQH Ef D SSP FRQFHQWUDWLRQ RI S\ULGLQH LQ WKH FRSSHU FROORLGDO VROXWLRQ Ff D EDFNJURXQG VSHFWUXP RI WKH FRSSHU FROORLG ([FLWDWLRQ ZDV DW QP ZLWK P: ODVHU RXWSXW

PAGE 85

5DPDQ 6KLIW FPf R R D 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf 1 RW / +

PAGE 86

)LJXUH 5HODWLYH FRSSHU FROORLG6(56 LQWHQVLW\ RI PRVW LQWHQVH EDQG RI WKH DQDO\WHV DV D IXQFWLRQ RI S+ $ SSP VROXWLRQ DQG WKH FP EDQG ZHUH XVHGL IRU WKH S\ULGLQH FXUYH $ SSP VROXWLRQ ZDV XVHG IRU WKH HYDOXDWLRQ RI WKH FP EDQG RI 3$%$ DQG WKH FP EDQG RI 31%$

PAGE 87

5(/$7,9( ,17(16,7< S+ 2) 62/87,21

PAGE 88

SUHYLRXVO\ FDOFXODWHG WKDW WKHUH LV D FRUUHODWLRQ EHWZHHQ WKH QP SHDN DQG FRSSHU VSKHUHV KDYLQJ D GLDPHWHU RI m QP 7KH SDUWLFOH VL]H GLVWULEXWLRQ RI WKH FROORLGV ZDV DQDO\]HG ZLWK D 0DOYHUQ SKRWRQ FRUUHODWLRQ VSHFWURPHWHU 7KH GDWD DQDO\VLV ZDV SHUIRUPHG XVLQJ WKH RSWLFDO FRQVWDQWV IRU EXON FRSSHU >@ 'XULQJ WKH WLPH VSDQ RI LQWHUHVW WKH DYHUDJH SDUWLFOH GLDPHWHU ZDV IRXQG WR EH DSSUR[LPDWHO\ QP VHH ILJXUH f 7KH RSWLFDO GHQVLW\ LQ ILJXUH UHDFKHV D PD[LPXP VHFRQGV DIWHU WKH UHDJHQWV ZHUH FRPELQHG ,I WKH QP SDUWLFOHV DUH WKH SUHFXUVRUV RI WKH PRVW 6(56 DFWLYH SDUWLFOH VL]H WKH 6(56 VLJQDOV VKRXOG LQFUHDVH ZKLOH WKH QP SDUWLFOHV DUH SUHVHQW DQG WKHQ GHFUHDVH 7KLV FRUUHODWHV ZHOO ZLWK WKH 6(56 VLJQDOV VKRZQ LQ ILJXUH ZKLFK LV ODUJHVW IRU WKH FP EDQG DW V DQG GHFUHDVHV WKHUHDIWHU 7KH WLPH GHSHQGDQFH RI WKH FP 6(56 VLJQDO PDNHV LW GLIILFXOW WR DFTXLUH D JRRG VSHFWUXP XVLQJ D VFDQQLQJ LQVWUXPHQW )LJXUHV f 7KH FRSSHU FROORLG V\VWHP KDV D PXFK ORZHU EDFNJURXQG WKDQ VLOYHU FROORLGDO V\VWHPV 8QIRUWXQDWHO\ WKH 6(56 VLJQDOV DUH DOVR VPDOOHU VR WKDW WKH VLJQDO WR QRLVH UDWLR LV QRW LPSURYHG 7KLV PD\ EH GXH WR ORZHU DGVRUSWLRQ RQWR WKH FRSSHU VXUIDFH DV FRPSDUHG WR VLOYHU VXUIDFHV $Q DGGLWLRQDO SUREOHP LV WKDW WKH FRSSHU FROORLG FDQQRW EH SUHSDUHG PRUH WKDQ PLQ DKHDG RI WLPH ZKHUH WKH VLOYHU FROORLG FDQ EH PDGH VWDEOH IRU IXWXUH XVH >@

PAGE 89

)LJXUH $EVRUSWLRQ VFDQV RI WKH FRSSHU FROORLG DV LW DJJUHJDWHV RYHU WLPH 7KH ORZHU VFDQ ZDV WDNHQ V DIWHU PL[LQJ WKH UHDJHQWV 7KH UHPDLQLQJ VFDQV VKRZ WKH LQFUHDVH LQ DEVRUEDQFH RFFXUULQJ EHWZHHS DQG V DW V LQWHUYDOV 7KH FRQGLWLRQV ZHUH S+ >&X@ ; >%+@ ; DQG SSP SDUDDPLQR EHQ]RLF DFLG

PAGE 90

D&A $WRVRUWDQFH ;

PAGE 91

)LJXUH $ SKRWRQ FRUUHODWLRQ VSHFWUXP RI WKH FRSSHU FROORLGV XVHG LQ WKHVH VWXGLHV 7KH GDWD ZDV FRUUHFWHG IRU WKH GLHOHFWULF IXQFWLRQ RI FRSSHU 7KH RUGLQDWH LV WKH SHUFHQWDJH RI WKH WRWDO FRPSRVLWLRQ WKDW WKH IUDFWLRQ PDNHV XS

PAGE 92

2-

PAGE 93

)LJXUH ,QWHQVLWLHV RI WKH FP ILOOHG FP ILOOHG FLUFOHf 6(56 EDQGV RIA$%$ DV WLPH 7KH FRQGLWLRQV ZHUH S+ >&X@ ; DQG SSP SDUDDPLQREHQ]RLF DFLG WULDQJOHf DQG D IXQFWLRQ RI ; >%+f@

PAGE 94

7LPH Vf 5HODWLYH 6(56 ,QWHQVLW\ 2 }‘ 1! 8f 2L 2 -

PAGE 95

)LJXUH Df WKH 6(5 VSHFWUXP RI SSP RI SDPLQREHQ]RLF DFLG RQ D FRSSHU FROORLG Ef WKH EDFNJURXQG VSHFWUXP RI WKH FRSSHU FROORLG ([FLWDWLRQ ZDV DW QP ZLWK P: ODVHU RXWSXW

PAGE 96

5DPDQ 6KLIW FPf§f 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf 0 2/

PAGE 97

)LJXUH Df WKH 6(56 VSHFWUXP RI SSP RI SQLWUREHQ]RLF DFLG RQ D FRSSHU FROORLG Ef WKH EDFNJURXQG VSHFWUXP RI WKH FRSSHU FROORLG ([FLWDWLRQ ZDV DW QP ZLWK P: ODVHU RXWSXW

PAGE 98

5DPDQ 6KLIW FPf 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf +

PAGE 99

FXUYHV IRU S\ULGLQH SDPLQREHQ]RLF DFLG RQ WKH FRSSHU FROORLG )LJXUH &DOLEUDWLRQ DFLG DQG SQLWUREHQ]RLF

PAGE 100

R H 2 &RQFHQLUDWLRQ SSPf

PAGE 101

7KH FROORLGDO FRSSHU VXUIDFH KDV D OLQHDU G\QDPLF UDQJH RI WR RUGHUV RI PDJQLWXGH ILJXUH f /LPLWHG G\QDPLF UDQJH LV FRPPRQ ZLWK FROORLGDO V\VWHPV VLQFH WKH VLJQDO LV GHSHQGHQW RQ WKH DGVRUSWLRQ LVRWKHUP ,QDELOLW\ WR FRQWURO GLUHFWO\ DGVRUSWLRQ RQWR WKH VXUIDFH DV FDQ EH GRQH ZLWK HOHFWURGH VXUIDFHV LV WKH JUHDWHVW OLPLWDWLRQ RI FROORLGDO EDVHG 6(56 7KH OLPLWV RI GHWHFWLRQ IRU S\ULGLQH 3$%$ DQG 31%$ DUH HVWLPDWHG WR EH DQG SSP UHVSHFWLYHO\ 7DEOH f

PAGE 102

7DEOH $QDO\WLFDO ILJXUHV &ROORLG RI 0HULW IRU $QDO\WHV RQ 6WDWLF &RSSHU &RPSRXQGV /2' D 6HQVE 6ORSH F &RUU G SSPf ORJORJ &RHII 3\ULGLQH SDPLQREHQ]RLF DFLG SQLWUREHQ]RLF DFLG ZKHUH D LV WKH OLPLW RI GHWHFWLRQ IRU WKH FRPSRXQG XVLQJ WKUHH VWDQGDUG GHYLDWLRQV DOARYHF WKH QRLVH LV WKH VHQVLWLYLW\ RI WKH AQHWKRG SSP nf F LV WKH VORSH RI WKH ORJORJ SORW DQG LV WKH FRUUHODWLRQ FRHIILFLHQW IRU WKH ORJORJ SORW

PAGE 103

7DEOH &RPSDULVRQ RI 6WDWLF &RSSHU &ROORLG SDPLQREHQ]RLF DFLG 5HVXOWV ZLWK /LWHUDWXUH 5HVXOWV RQ &ROORLGDO 6LOYHU $XWKRU /2'nV SSPfD 6ORSH E &RUU &RHIIF /RJ/RJf ( / 7RUUHV 7KLV :RUN ZKHUH D LV WKH OLPLW RI GHWHFWLRQ IRU WKA FRPSRXQG XVLQJ WKUHH VWDQGDUG GHYLDWLRQV DERYH WKH QRLVH LV WKH VORSH RI WKH ORJORJ SORW DQG F LV WKH FRUUHODWLRQ FRHIILFLHQW IRU WKH ORJORJ SORW

PAGE 104

&+$37(5 &233(5 &2//2,'6 ,1 $ )/2:,1* 675($0 ,QWURGXFWLRQ 7KH WHFKQLTXH RI )ORZ ,QMHFWLRQ $QDO\VLV ),$f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

PAGE 105

LQMHFWHG LQWR D IORZLQJ VWUHDP DQG FDQ EH PDQLSXODWHG IXUWKHU GRZQ VWUHDP DV UHTXLUHG &KDQJHV LQ WKH V\VWHP DUH VHHQ TXLFNO\ DOORZLQJ UDSLG RSWLPL]DWLRQ RI SDUDPHWHUV 7KH WLPH UHTXLUHG IRU WKH VHSDUDWLRQ OLPLWV WKH SDFH RI RSWLPL]DWLRQ LQ +3/& /DVHUQD HW DO ILUVW GHPRQVWUDWHG WKH DSSOLFDWLRQ RI 6(56 WR ),$ >@ )RU WKH GHWHUPLQDWLRQ RI SDPLQREHQ]RLF DFLG WKH b 56' ZDV UHGXFHG IURP b WR b 7KH GHWHFWLRQ OLPLW ZDV DOVR UHGXFHG E\ DQ RUGHU RI PDJQLWXGH 1L HW DO >@ GHWHFWHG 6(56 IURP 516 EDVHV LQ DQ ),$ V\VWHP 7KH HIIHFWV RI S+ DQG WHPSHUDWXUH RQ WKH 6(56 VLJQDOV ZDV DOVR LQYHVWLJDWHG &RPELQHG UHVRQDQFH DQG VXUIDFH HQKDQFHG 5DPDQ VSHFWURVFRS\ 6(556f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

PAGE 106

QDQRPRODU UDQJH 7KHVH UHVXOWV LOOXVWUDWH WKH SURJUHVV RI 6(56 WRZDUGV ILQDOO\ EHFRPLQJ D URXWLQH PHWKRG IRU DQDO\VLV )XUWKHU SURJUHVV WRZDUG WKLV JRDO LV FRYHUHG LQ WKH QH[W VHFWLRQ ZKHUH DGDSWDWLRQ RI FRSSHU FROORLGV WR ),$ LV GLVFXVVHG ([SHULPHQWDO $SSDUDWXV 7KH LQVWUXPHQWDO V\VWHP XVHG LQ WKHVH H[SHULPHQWV LV WKH VDPH DV WKDW RI WKH SUHYLRXV FKDSWHU VHH )LJXUH f )LJXUH LV D VFKHPDWLF GLDJUDP RI WKH VHWXS IRU SURGXFLQJ WKH FRSSHU FROORLGV RQOLQH $ SHULVWDOWLF SXPS 5DLQLQf ZDV XVHG WR SXPS D VROXWLRQ ; 0 RI VRGLXP ERURK\GQGH DW POPLQ $ VLPLODU SXPS ZDV XVHG WR SXPS VROXWLRQ DW POPLQ IURP D UHVHUYRLU RI ; 0 FXSULF QLWUDWH SUHSDUHG LQ ; 0 QLWULF DFLG 7KHVH ZHUH LQLWLDOO\ PL[HG E\ IORZLQJ WKHP WRJHWKHU LQWR D FKURPDWRJUDSKLF 7 MRLQW 7KH FROORLGDO VROXWLRQ IORZHG WKURXJK D P OHQJWK RI WXELQJ WR DQ LQMHFWLRQ YDOYH $OWH[ 6FLHQWLILF ,QF PRGHO f ZLWK D , VDPSOH ORRS )URP WKH LQMHFWLRQ YDOYH WKH VROXWLRQ IORZHG GLUHFWO\ LQWR WKH PL[LQJ FKDPEHU ZKHUH WKH SOXJ RI 3$%$ ZDV WKRURXJKO\ PL[HG ZLWK WKH FROORLGDO FRSSHU 7KH PL[LQJ FKDPEHU ZDV RSHQ WR WKH DWPRVSKHUH WR UHPRYH WKH K\GURJHQ JDV WKDW HYROYHG LQ WKH IRUPDWLRQ RI WKH FROORLG 7KH YROXPH RI WKH VROXWLRQ LQ WKH PL[LQJ FHOO UDQJHG IURP DSSUR[LPDWHO\ WR PO 7KH VROXWLRQ ZDV SXPSHG IURP WKH PL[LQJ FKDPEHU

PAGE 107

)LJXUH 6FKHPDWLF GLDJUDP RI WKH IORZ LQMHFWLRQ DQDO\VLV VHWXS XVHG WR SURGXFH WKH FRSSHU FROORLGV

PAGE 108

3803 3803 3803

PAGE 109

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m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f DQG FP 7KLV GHPRQVWUDWHV WKH DELOLW\ RI FRSSHU FROORLG EDVHG6(56 WR

PAGE 110

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

PAGE 111

7DEOH 7KH (IIHFW RI ,QFUHDVHG $FLGLW\ RQ WKH $JJUHJDWLRQ RI WKH &RSSHU &ROORLG DQG WKH 5HVXOWLQJ 6(56 6LJQDO 1LWULF $FLG &RQFHQWUDWLRQ 0f 7LPH RI 2QVHW RI 6(56 Vf ; LRn ; ; LRn ; LRn ; LRn m

PAGE 112

)LJXUH 7KH 6(56 VSHFWUXP RI SDPLQREHQ]RLF DFLG RQ D ),$ SURGXFHG FRSSHU FROORLG EHWZHHQ DQG FP

PAGE 113

5DPDQ 6KLIW FP f 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf 2 r R! RR R R R R r

PAGE 114

)LJXUH 7KH 6(56 VLJQDOV DW FP IURP D VHULHV RI LQMHFWLRQV RI /WJ RI SDPLQREHQ]RLF DFLG LQWR WKH FRSSHU FROORLG),$ V\VWHP

PAGE 115

7LPH Vf 5DPDQ 6FDWWHUHG ,QWHQVLW\ DUE XQLWVf f§f 12 2 R R E

PAGE 116

)LJXUH &DOLEUDWLRQ FXUYH RI WKH UHVSRQVH RI WKH 6(56 VLJQDOV WR YDULRXV DPRXQWV RI SDPLQREHQ]RLF DFLG LQMHFWHG LQWR WKH FRSSHU FROORLG),$ V\VWHP 7KH V\PEROV RI WKH WZR ORZHVW PHDVXUHPHQWV DUH ODUJHU WKDQ WKH HUURU LQ WKH PHDVXUHPHQW

PAGE 117

$PRXQW RI 3$%$ LQMHFWHG PLFURJUDPVf

PAGE 118

7DEOH &RPSDULVRQ RI SDPLQREHQ]RLF DFLG &RSSHU &ROORLG),$ 'DWD ZLWK /LWHUDWXUH 5VXOWV RQ 6LOYHU &ROORLGV $XWKRUV /2'nV )ORZ 5DWH &RQH RI 0HWDO 0Jf /DVHUQD HW DO POPLQ ; n 0 7KLV :RUN POPLQ ; f 0 ZKHUH D LV WKH OLPLW RI GHWHFWLRQ IRU WKH FRPSRXQG QJf XVLQJ WKUHH VWDQGDUG GHYLDWLRQV DERYH WKH QRLVH LV WKH IORZ UDWH WKURXJK WKH VSHFWURVFRSLF FHOO POPLQf DQG F LV WKH FRQFHQWUDWLRQ RI WKH VRGLXP ERURK\GULGH LQ WKH ILQDO VROXWLRQ

PAGE 119

7KLV SUREOHP FXUUHQWO\ OLPLWV WKH DQDO\WLFDO XVHIXOQHVV RI FRSSHU FROORLGV SUHSDUHG YLD ),$

PAGE 120

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f FRQGLWLRQV ZHUH HYDOXDWHG 7KH GHWHFWLRQ OLPLW IRU 3$%$ ZDV FRPSDUDEOH WR WKRVH RI 7RUUHV >@ RQ VLOYHU K\GURVROV &RSSHU FROORLGV DUH PRUH DFWLYH WKDQ WKRVH RI VLOYHU ZKLFK

PAGE 121

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

PAGE 122

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

PAGE 123

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

PAGE 124

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

PAGE 125

FXUUHQWO\ DOORZ WKH DFTXLVLWLRQ RI XQHQKDQFHG 5DPDQ VSHFWUD IURP VXUIDFHV 7KLV WHFKQRORJ\ UHGXFHV WKH WLPH UHTXLUHG WR REWDLQ VSHFWUD 7KH LQFUHDVHG XVH RI WXQDEOH VROLG VWDWH ODVHUV DOVR ZLOO KHOS WR PDNH 6(56 DQG 5DPDQ VSHFWURVFRS\ LQ JHQHUDO D URXWLQH PHWKRG RI DQDO\VLV 6ROLG VWDWH ODVHUV VXFK DV WKH 7LVDSSKLUH DQG GLRGH ODVHUV FDQ KDYH KLJK SRZHU D ZLGH WXQLQJ UDQJH DQG DUH UHODWLYHO\ VLPSOH WR RSHUDWH ZKHQ FRPSDUHG WR G\H ODVHUV

PAGE 126

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

PAGE 127

PXOWLSROHV DERYH WKH GLSROH FDQ EH QHJOHFWHG ZLWKRXW VLJQLILFDQW HUURU )RU D SURODWH VSKHURLG VXUIDFH WKH HOOLSLWLFDO FRRUGLQDWHV FDQ EH GHILQHG DV [ I>f UMf @ FRV \ I>ef OUMf@VLQ ] ILU@ $Of ZKHUH WS LQ WKH D]LPXWKDO FRRUGLQDWH DV PHDVXUHG IURP WKH [] SODQH DQG e DQG UM DUH WKH RUWKRJRQDO VSKHULFDO FRRUGLQDWHV 7KH IRFDO GLVWDQFH I RI WKH SURODWH VSKHURLG LV UHODWHG WR WKH PDMRU D DQG PLQRU E VHPLD[HV E\ WKH HTXDWLRQ I DEfA $f $VVXPH WKH FDVH ZKHUH WKH VXUIDFH RI WKH VSKHURLG LV GHILQHG E\ WKH HTXDWLRQ DV e DI DQG WKH SRVLWLRQ RI WKH PROHFXOH ORFDWHG D GLVWDQFH + GLUHFWO\ DERYH WKH SURWUXVLRQ LV GHILQHG E\ D+fI )RU WKH UHJLRQ RXWVLGH WKH VSKHURLG ZLWK DQ DSSOLHG ODVHUf ILHOG D SRODUL]HG PROHFXOH QHDUE\ DQG D ILHOG GXH WR WKH VSKHURLG DQG WKH SODQH WKH SRWHQWLDO M PD\ EH ZULWWHQ DV

PAGE 128

)LJXUH $O Df $ FURVVVHFWLRQ RI D SURODWH VSKHURLG LOOXVWUDWLQJ WKH HOOLSWLFDO FRRUGLQDWH V\VWHP Ef $Q LOOXVWUDWLRQ RI WKH PRGHO XVHG IRU WKH GHULYDWLRQ LQ WKH DSSHQGL[

PAGE 129

, 2=7 rR

PAGE 130

3M (Y+Q -/ f LQf QLf@n Dr $f >Of OUMf DQ Lf@n `6Q FQSQQfQ2 ZKHUH 3Q DQG 4Q GHQRWH /HJHQGUH IXQFWLRQV RI WKH ILUVW DQG VHFRQG NLQG UHVSHFWLYHO\ 7KH ILUVW WHUP LV WKH FRQWULEXWLRQ WR WKH SRWHQWLDO RI WKH VXUIDFH GXH WR DQ LVRODWHG VSKHURLG DQ D FRQGXFWLQJ SODQH 7KH VHFRQG WHUP FRQWDLQV FRQWULEXWLRQV IURP ERWK WKH SRODUL]HG PROHFXOH DQG LWV LPDJH D GLVWDQFH D+ EHORZ WKH FRQGXFWLQJ SODQH 7KH ILQDO WHUP UHSUHVHQWV WKH SRWHQWLDO DVVRFLDWHG ZLWK WKH SRODUL]HG KHPLVSKHURLG DQG LWV LPDJH RQ WKH FRQGXFWLQJ SODQH 7KH SRWHQWLDO LV FKRVHQ VR WKDW LW YDQLVKHV DW WKH ERXQGDU\ RI WKH KHPLVSKHURLG DQG WKH FRQGXFWLQJ SODQH UM f $ UHVXOW RI WKLV LV WKDW RQO\ WKH RGG WHUPV DSSHDU LQ WKH VXPPDWLRQ ,QVLGH WKH SURODWH VSKHURLG WKH JHQHUDO VROXWLRQ WR WKH /DSODFH HTXDWLRQ FDQ EH DSSOLHG r (Q EQ3QQf3QDf $f 7KH FRHIILFLHQWV EQ DQG FQ LQ HTXDWLRQV $ DQG $ PXVW EH GHWHUPLQHG ZKHQ WKH VROXWLRQV IRU WKH VXUIDFH DQG WKH LQVLGH RI WKH VSKHURLG DUH PDWFKHG DW WKH ERXQGDU\ e e IRU DOO FDVHV ZKHUH &.UMFL

PAGE 131

6LQFH WKHVH DUH HOHFWULFDOO\ FRQGXFWLQJ r, DOVR QHHGV WR EH FRQWLQXRXV DFURVV WKH VXUIDFH a( IWY $V ARBf B7"f }L f@n I > m' LIff r (RO+Lf@fnf $f UQ &AQAQLARf f FQ4Q eRf @AQ r"f ,I WKH QRUPDO FRPSRQHQWV RI WKH GLVSODFHPHQW YHFWRUV DORQJ WKH VXUIDFH DUH PDWFKHG WR DYRLG GLVFRQWLQXLW\ LQ WKH WUDQVLWLRQ RI WKH GLHOHFWULF IURP WKDW RI WKH PHWDO WR WKDW RI WKH VXUURXQGLQJ PHGLXP r (I=Q $ ^>6R'GU"f R+Lf@n I R A G >L L"f WR LHR@n r $ff (Q >H YMEQLfQ Rf FfRn(ff @3f Uff ZKHUH HYf LV WKH GLHOHFWULF IXQFWLRQ RI WKH PHWDO UHODWLYH WR WKH VXUURXQGLQJ PHGLXP 7KH WZR DERYH H[SUHVVLRQV $ DQG $ FDQ EH LQYHUWHG WR REWDLQ WKH FRHIILFLHQWV EQ DQG FQ E\ HPSOR\LQJ WKH LQWHJUDO $ I43QULf3P9fGQ A $f ZKLFK LV YDOLG ZKHQ P DQG Q DUH ERWK RGG DQG $ IRU WKH FDVH ZKHUH Q LV RGG DQG e_!e !

PAGE 132

-R 3Q7f^O>AOf OUMf 6R U"eOf@O>eR' UMf 6 U"f@fG U 3Q 6f 4Q f 7KHVH \LHOG H[SUHVVLRQV $ DQG $ UHVSHFWLYHO\ e Rf f FQ4QeRf LAIeQ! Qf LIf 3Q 6f 2S 6Rf HYfQAQARf FQRn6f 9Qf Qf QIfSA 6RfQ 6_f 6ROYLQJ $ DQG $ IRU FQ \LHOGV F Q QfQ HYfOf(RIOQWL HYfL6Rf f 6R4ARf H\!ff3Q8f4QDLf3+Df HYf4QDRf3Q=Rf a 4nQ=Rf3QWRf $f )RU WKH 5D\OHLJK OLPLW WKH HOHFWURPDJQHWLF HPLVVLRQ LV GXH WR WKH GLSROH PRPHQW ,I LV HYDOXDWHG IRU WKH FDVH ZKHUH e LV ODUJH ZH REWDLQ U DV L4ar RR ZKHUH 'O ILf P LFL-I $f $f 7KH ILUVW WHUP LQ WKH HTXDWLRQ $ DFFRXQWV IRU ERWK WKH PROHFXODU GLSROH DQG LWV LPDJH LQ WKH SODQH 7KH VHFRQG WHUP

PAGE 133

GHVFULEHV WKH GLSROH LQGXFHG RQ WKH VSKHURLG DQG LWV LPDJH LQ SODQH 7KH H[SDQVLRQ RI WKHVH WHUPV \LHOGV I(R RmYfOf Q HYf4Lff R2Rf OHY--A2ALf HYf4Lf 6R4Ln‹Rf $f 6LQFH WKH SRODUL]DELOLW\ WHQVRU RI WKH PROHFXOH D LV DQLVRWURSLF DQ RULHQWDWLRQ IRU WKH PROHFXOH UHODWLYH WR WKH VXUIDFH PXVW EH DVVXPHG $VVXPLQJ WKH RULHQWDWLRQ RI WKH SULQFLSOH D[LV SHUSHQGLFXODU WR WKH VXUIDFH WKH GLSROH FDQ WKHQ GHWHUPLQHG E\ WKH QHW HOHFWULF ILHOG SURGXFHG DW WKH PROHFXOHV SRVLWLRQ E\ WKH QHDUE\ VSKHURLG DQG WKH H[WHUQDOO\ DSSOLHG ILHOG 0 m W( &Q4QeOf IeLf $f

PAGE 134

FRPELQLQJ HTXDWLRQV $ DQG $ OHDGV WR WKH H[SUHVVLRQV D?( LU eR>HYfO@ HY-2ARf RLARf >OHYf@‹R^‹Of HYfLRf AR4ARf en D D>HYfO@ U Lf I Q Of3QWWfSnQDf &4S Af HYf3nQ8Rf4QWWRf a 3Q=Rf4nQORf $f $f 7KH ILUVW WHUP RI HTXDWLRQ $ LV WKH FRQWULEXWLRQ WR WKH GLSROH PRPHQW GXH DQ LVRODWHG KHPLVSKHURLG RQ D FRQGXFWLQJ SODQH 7KH GHQRPLQDWRU U PD\ EH WHUPHG WKH LPDJH HQKDQFHPHQW IDFWRU )URP HTXDWLRQ $ LW FDQ EH VHHQ WKDW U FRQVLVWV RI WZR WHUPV 7KH ILUVW WHUP RI HTXDWLRQ $ LV GXH WR WKH LPDJH RI WKH PROHFXODU GLSROH LQ WKH SODQH DQG WKH VHFRQG GXH WR LWV LPDJH LQ WKH KHPLVSKHURLG 7KH FRQWULEXWLRQ WKH WKH 5DPDQ VLJQDO IURP WKH QHDUE\ SODQH LV H[SUHVVHG E\ HTXDWLRQ $ U D_ > D+f @ $f

PAGE 135

,I WKH KHPLVSKHURLG LV D SUHIHFW FRQGXFWRU WKHQ _HYf_ rr \LHOGLQJ UP r DL ILf 4SAOf 2Q eRf $f )RU D KHPLVSKHURLG RQ D FRQGXFWLQJ SODQH WKH GLSROH PRPHQW JLYHQ E\ HTXDWLRQ $ UHGXFHV WR 'HR I(RW R e" Rf W?(J AR4? e f L Df A $f )XOO 3URODWH 6SKHURLG ,I DQ LQVXODWHG IXOO VSKHURLG LH QRW RQH HPEHGGHG LQ D FRQGXFWLQJ VXUIDFHf LV FRQVLGHUHG WKHQ WKH LPDJH GLSROH WHUP DVVRFLDWHG ZLWK WKH SODQDU VXUIDFH HTXDWLRQ $f LV QHJOHFWHG DQG WKH VXP LV WKHQ H[WHQGHG RYHU DOO YDOXHV RI Q 0DWFKLQJ WKH SRWHQWLDOV RYHU WKH UDQJH UM \LHOGV a IA(R J>YfO@ fYfLRf 6R2‹Rf rrR LU L >OHYf@Tnf HWYA8R-6R4DRf ZKHUH $f

PAGE 136

I IO>HYfO@ e QOf3QRfSn>n f @ $f Q Q HYfSn RfQRf 3Qef4Q ARf ,I ZH IXUWKHU DVVXPH WKH VXUIDFH LV D SHUIHFW FRQGXFWRU H[SUHVVLRQV $ DQG $ VLPSOLI\ WR ? I(R= FFA(T $f &"L 6Rf LU L mff $f (TXDWLRQV $ DQG $ DUH VLPLODU WR WKH UHVXOWV IRU WKH KHPLVSKHURLG RQ WKH FRQGXFWLQJ SODQDU VXUIDFH HTXDWLRQV $ DQG $f 7KHVH HTXDWLRQV DOO FRQWDLQ WHUPV LQ WKH GHQRPLQDWRU RI WKH IRUP $Yf HYf4Lef eRLeRf $f 7KLV LV WKH FRQGLWLRQ IRU WKH SODVPRQ UHVRQDQFH XSRQ WKH VXUIDFH )RU D VSKHUH WKLV LV $f

PAGE 137

7KH FRQGLWLRQV IRU WKH VXUIDFH SODVPRQ UHVRQDQFH RQ D VSKHUH LV HYf f $f ZKHUH LV WKH TXDQWXP QXPEHU IRU SODVPRQ UHVRQDQFH RQ VSKHUH )RU WKH GLSROH FDVH f WKLV UHGXFHV WR HYf $V WKH VSKHURLG EHFRPHV PRUH SURODWH er f WKH HTXDWLRQ EHFRPHV HYf ^e f OQ> Of`n $f ZKLFK PHDQV WKH UHDO FRPSRQHQW RI HYf ZLOO EH ODUJH DQG QHJDWLYH $V WKH VSKHURLG EHFRPHV PRUH DQG PRUH HFFHQWULF WKH VXUIDFH SODVPRQ IUHTXHQF\ FRUUHVSRQGLQJ WR WKH PDMRU D[LV ZLOO GHFUHDVH 7KH ODVHU UDGLDWLRQ ZLOO EH LQ UHVRQDQFH ZLWK RQO\ D IUDFWLRQ RI WKH SDUWLFOHV EXW WKH HQKDQFHPHQW IURP WKHVH SDUWLFOHV ZLOO EH VWURQJ ,W LV QHFHVVDU\ WKDW WKH LPDJLQDU\ FRPSRQHQW ,PHYff EH VPDOO DW WKLV UHVRQDQFH IRU KLJK HIILFLHQF\ LQ WKH FRXSOLQJ RI WKH SKRWRQV ZLWK WKH VXUIDFH SODVPRQV 7KH (QKDQFHPHQW )DFWRU 7KH FURVV VHFWLRQ RI WKH 5DPDQ VFDWWHULQJ IURP WKH LVRODWHG PROHFXOH FDQ EH REWDLQHG LQ D VHPLFODVVLFDO PDQQHU EHJLQQLQJ ZLWK D D$4A FRVYYW R4 $f

PAGE 138

ZKHUH $4 LV D PROHFXODU PRGH FRRUGLQDWH DQG YY LV WKH FRUUHVSRQGLQJ PROHFXODU YLEUDWLRQDO IUHTXHQF\ "I r!_if f§! 7KH 5DPDQ VFDWWHULQJ IURP WKH PROHFXOH QHDU WKH VXUIDFH FDQ WKHQ EH FDOFXODWHG XVLQJ HTXDWLRQ $ WR EH D UV f 7 Y ? 7 ?FM >LU@ OHfe4DLf e4L f eT4? Tf $f 7R REWDLQ WKH HQKDQFHPHQW GXH WR WKH SUR[LPLW\ RI WKH VXUIDFH WKH HQKDQFHG 5DPDQ VFDWWHU LQ HTXDWLRQ $ FDQ EH GLYLGHG E\ WKH 5DPDQ VFDWWHU GXH WR WKH LVRODWHG PROHFXOH HTXDWLRQ $ f 5 OOHYPR4DAWHY-4DRf T? Rf @ $f )RU D SHUIHFW FRQGXFWRU WKH GLSROH FRQWULEXWLRQ LV JLYHQ E\ HTXDWLRQ $ \LHOGLQJ HG HYf [HGfe!L8 GUf $f :KHUH HYf LV WKH GLHOHFWULF IXQFWLRQ RI WKH PHWDO HG LV WKH GLHOHFWULF IXQFWLRQ RI WKH VXUURXQGLQJ PHGLXP DQG [ LV

PAGE 139

GHSHQGHQW XSRQ WKH FXUYDWXUH RI WKH VXUIDFH b LV WZR IRU D VSKHUH )RU WKH FDVH ZKHUH WKH PROHFXOH LV DGVRUEHG RQ WKH VXUIDFH A ef DQG WKH SKRWRQV FRXSOH HIILFLHQWO\ ZLWK WKH VXUIDFH SKRQRQV $ HTXDWLRQ UHGXFHV WR HYf HG HYf HG $f (TXDWLRQ $ LV DQ DSSUR[LPDWLRQ RI HTXDWLRQ $ ZKLFK LV PRGHOHG DIWHU D VLOYHU SURODWH VSKHURLG QP [ QP HPEHGGHG LQWR D IODW VLOYHU VXUIDFH _HYfOHYf_ $f 7KLV HTXDWLRQ EHVW LOOXVWUDWHV WKH LPSRUWDQFH RI WKH LPDJLQDU\ WHUP RI WKH GLHOHFWULF IXQFWLRQ HYf RI WKH PHWDO

PAGE 140

7DEOH $O /LVW RI 6\PEROV XVHG LQ WKH $SSHQGL[ 6\PERO 0HDQLQJ R I ?R D E Y Y n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f LQGLFDWHV SKDVH ERXQGDU\f

PAGE 141

5()(5(1&(6 *DUUHOO 5/ $QDO &KHP $ &DPSLRQ $ LQ 9LEUDWLRQDO 6SHFWURVFRS\ RI 0ROHFXOHV RQ 6XUIDFHV
PAGE 142

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

PAGE 143

6KHQJ 5 1L ) &RWWRQ 70 $QDO &KHP +HDUG 60 *UHLVHU ) %DUUDFORXJK &* 6DQGHUV -9 3K\V &KHP -RR 7+
PAGE 144

/DVHUQD -&DPSLJOLD $' :LQHIRUGQHU -' $QDO &KHP 6TXDULV -0 .RJOLQ ( )UHVHQLXV = $QDO &KHP 6TXDULV -0 .RJOLQ ( $QDO &KHP .RJOLQ ( 0RO 6WUXFW &KDQJ 5. )XUWDN 7. 6XUIDFH (QKDQFHG 5DPDQ 6SHFWURVFRS\ 3OHQXP 3UHVVU1HZ
PAGE 145

-RKQVRQ 3% &KULVW\ 5: 3K\V 5HY % $XVVHQHJJ )5 /LSSLWFK 0( &KHP 3K\V /HWW 0F&DOO 6/ 3ODW]PDQ 30 3K\V 5HY % (IIULPD 6 0HWLX + &KHP 3K\V %LOOPDQ 2WWR $ $SSO 6XUI 6FL %LOOPDQ .RYDFV 2WWR $ 6XUI 6FL 3RFNUDQG 2WWR $ 6ROLG 6WDWH &RPP :RRG 7+ .OHLQ 09 9DF 6FL 7HFKQRO 0LH $QQ 3K\V .HUNHU 0 LQ 3K\VLFDO &KHPLVWU\ DQG &ROORLGV 7KH 6YHGKHUJ 6\PSRVLXP 5£QE\ % (G %ODFNZHOO 6FLHQWLILF 3XEOU2[IRUG &KDSWHU .HUNHU 0 3XUH DQG $SSO &KHP %ODWFKIRUG &* &DPSEHOO -5 &UHLJKWRQ -$ 6XUI 6FL 5LHJHU 3+ LQ (OHFWURFKHPLVWU\ 3UHQWLFH +DOO,QF (QJOHZRRG 1S :HLW] '$ /LQ 0< 6DQGURII &6XUI 6FL 'HYLOOH] & 6DQILHOG $ 6WHLQHU $ &ROORLG DQG ,QWHUIDF 6FL %HYLOO 5) LQ &5& +DQGERRN 6HULHV LQ =RRQRVHV 9ROXPH 6HFWLRQ $QWLELRWLFV 6XOIDPLGHV DQG 3XEOLF +HDOWK &5& 3UHVV ,QFU%RFD 5DWRQ )/ S DQG 6RSKLDQ /+ 3LSHU '/ 6FKQHOOHU *+ 7KH 6XOIDS\ULPLGHV 3UHVV RI $ &ROLVKU1HZ
PAGE 146

%UDWWRQ $& 0DUVKDOO (. MU %LRO &KHP ,VNDQGHU 0/ 0HGLHQ +$$ 1DVKHG 6 0LFURFKHP 6DWR 6 +LJXFKL 6 7DQDND 6 $QDO &KLP $FWD
PAGE 147

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

PAGE 148

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
PAGE 149

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\ $XJXVW 'HDQ *UDGXDWH 6FKRRO

PAGE 150

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 EMBEJDOND_92SAJZ INGEST_TIME 2017-07-13T22:04:18Z PACKAGE AA00003711_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES