Citation
Application of luminol chemiluminescence to the analysis of the beta lactam antibiotics

Material Information

Title:
Application of luminol chemiluminescence to the analysis of the beta lactam antibiotics
Creator:
Miyawa, John H
Publication Date:
Language:
English
Physical Description:
ix, 101 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Antibiotics ( jstor )
Chemiluminescence ( jstor )
Hydrogen ( jstor )
Ions ( jstor )
Lactams ( jstor )
Penicillin ( jstor )
Peroxides ( jstor )
Reagents ( jstor )
Signals ( jstor )
Sodium ( jstor )
Antibiotics, Lactam -- chemistry ( mesh )
Antibiotics, Lactam -- isolation & purification ( mesh )
Chemiluminescence ( mesh )
Chemiluminescence -- diagnostic use ( mesh )
Chromatography, High Pressure Liquid -- methods ( mesh )
Luminol -- chemistry ( mesh )
Spectrophotometry, Ultraviolet -- methods ( mesh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 97-100).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by John H. Miyawa.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Permission granted to the University of Florida to digitize, archive and distributed this item for non-profit and educational purposes only. Any reuse of this item in excess of fair use requires permission of the copyright holder.
Resource Identifier:
50083617 ( OCLC )
ocm50083617

Downloads

This item has the following downloads:


Full Text








APPLICATION OF LUMINOL CHEMILUMINESCENCE
TO THE ANALYSIS OF THE BETA LACTAM ANTIBIOTICS















By

JOHN H. MIYAWA


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
1995












ACKNOWLEDGMENTS


I would like to express my sincere gratitude to Dr. John Perrin for serving

as the chairman of my supervisory committee as well as providing guidance and

advice over the last four years.

I would also like to thank my supervisory committee members, Dr. Stephen

Schulman, Dr. Margaret James, Dr. Guenther Hochhaus and Dr. Vaneica Young

for their support and criticism of the material presented.

Further, I wish to thank the Department of Medicinal Chemistry for providing

the opportunity, Dr. K. Sloan for providing advice, my fellow students for tolerating

me.

Finally I would like to thank my whole family for their support and

encouragement.













TABLE OF CONTENTS


ACKNOWLEDGMENTS .........

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

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

ABSTRACT ..................

CHAPTERS

1. INTRODUCTION ....


Reaction mechanism .............................. 3
Application of luminol chemiluminescence to analysis in
solution ................................... 4
Optimization of analytical response ................... 13
The beta lactam antibiotics ......................... 16

2. OBJECTIVES ................................... 19


3. EXPERIMENTAL ................................ 24

Reagents ...................................... 24
Preparation of luminol solutions ................. 25
Preparation of Hydrogen Peroxide solutions ........ 25
Preparation of stock Sodium Dihydrogen Phosphate
solution ............................. 25
Preparation of Mobile Phase ................... 25
Preparation of Calibration series ................ 26
Ultraviolet spectrophotometric studies ................. 26
Static studies ................................... 27
Conductivity measurements ................... 27
Investigation of possible Peroxylate based mechanism ..... 29
Protocol for comparing different Fluorescers ........ 29


. . . . . . . . . . . . . . . ii

........=..................... V







Protocol for comparing different antibiotics .........
Flow studies ...................................
Flow Injection Analysis .......................
Chromatographic studies ..........................
Protocol for Calibration Curve determinations .......
Equipm ent .....................................
Software .................................
Source of water ............................

4. RESULTS AND DISCUSSION .......................

Mechanistic Studies ..............................
Static Studies ...................................
Flow Studies ...................................
Flow Injection analysis .......................
Liquid Chromatography ......................

5. CONCLUSIONS .................................


APPENDIX ....................

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

BIOGRAPHICAL SKETCH .........


. . . . . . . . . . . . . 10 1












LIST OF TABLES


Table page



1. Analysis in which luminecence has been applied .............. 13

2. Antibiotic compounds screened for enhancement.............. 37

3. Fluorescent compounds examined for potential use as reagents ... 47

4. Kinetic data obtained from Intensity-time profiles .............. 58

5. Analytical parameters for selected penicillins obtained from flow
injection analyses .............................. 78

6. Influence of selected excipients on chemiluminescent signal ...... 80

7. Chromatographic retention times for selected betalactam compounds
employing post-column chemiluminescence detection ...... 84

8. Analytical parameters for selected compounds ................ 89

9. Results from resolution measurements ..................... 90












LIST OF FIGURES



Figure pge

1. Chemical structure 2 aminophthalate ........................ 2

2. Schematic outline of chemiluminescent reaction pathway ......... 5

3. Typical luminescence profile .............................. 8

4. Various penicillins ..................................... 22

5. Schematic diagram of setup for static chemiluminescent
measurements .................................. 28

6. Three channel flow system .............................. 31

7. Two channel flow system ............................... 32

8. Chromatographic setup employed ......................... 33

9. Reaction scheme after White et al ......................... 40

10. Reaction scheme after Sattar et al ......................... 42

11. Ultraviolet visible spectrum of dicloxacillin on addition of varying
concentrations of 1102 ................................. 44

12. Intensity-time profiles for dicloxacillin system in presence of selected
fluorescers .................................... 48








13. Intensity-time profiles for selected penicillins in the presence of
fluorescein ..................................... 49

14. Postulated mechanism for scission of p- lactam ring ............ 50

15. Comparison of intensity-time profiles for dicloxacillin-luminol on
employing different oxidants ........................ 52

16. Reaction mechanism involving hydrazine generation ............ 54

17. Luminesence-time profiles for selected compounds in luminol
solutions ...................................... 57

18. Reaction model for HICHEM simulation ..................... 60

19. Input file for simulations on HICHEM ....................... 61

20. Concentration-time profiles for HICHEM simulation reactants and
intermediates ................................... 62

21. Simulated light emission profile ........................... 63

22. Reaction model for ACUCHEM simulation ................... 64

23. Input file for simulations on ACUCHEM ..................... 65

24. Simluated concentration-time profiles for 2-aminophthalate generation
for different values for the rate constant k .............. 67

25. Simulated light emission profiles for different values of rate constant
k . . . . . . . . . . . . . . . . . . . . . 68

26. Intensity time profile for luminol reaction at different methanol
concentrations .................................. 70

27. FIA chromatogram for enhancement by selected compounds ..... 75

28. Histogram for enhancement by selected compounds ........... 76

29. FIA calibration plots for selected compounds ................. 77

30. Scatter diagram for Infrared stretch frequencies for selected
compounds versus relative enhancement of luminol reaction 79







31. Calibration plot for dicloxacillin ........................... 85

32. Calibration plot for penicillin V ............................ 86

33. Calibration chart for penicillin G ........................... 87

34. Representative chromatogram for calibration measurements ..... 88












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

APPLICATION OF LUMINOL CHEMILUMINESCENCE
TO THE ANALYSIS OF THE BETA LACTAM ANTIBIOTICS


By

John H. Miyawa

December 1995

Chairman: John H. Perrin, PhD
Major Department: Medicinal Chemistry

The enhancement of luminol chemiluminescence by selected betalactam

antibiotics is examined and some observations and suggestions made as to the

mechanisms) of enhancement and the analytical utility of the enhancement.

The results obtained suggest the possibility of a peroxylate-like enhancement

mechanism in which cleavage of the beta-lactam ring is responsible for the energy

release manifesting as an enhancement of chemiluminescence.

The enhancement is applied to liquid chromatography (HPLC) as a post-column

chemiluminescence detection method. The utility of the post-column detection

technique is compared to ultraviolet detection (254nm) in terms of two peak

separation criteria.













CHAPTER 1
INTRODUCTION





The term chemiluminescence has generally referred to the luminescent

phenomena associated with a variety of chemical reactions. Chemiluminescence

can be more specifically described as the electromagnetic emission that arises

from the exothermic oxidation of an organic compound. Generally the exothermic

oxidation of the organic compound yields an energy rich product that is

luminescent because the molecule is either rigid or so small, that it is unable to

quickly dissipate internally the energy of the exothermic reaction.

Chemiluminescence is distinguished from the more efficient bioluminescence by

the fact that in bioluminescence visible light is produced from an enzymically

controlled reaction involving the chemical components of a living system. For a

molecule to exhibit chemiluminescence it must be able to form an electronically

excited species through a chemical reaction at ordinary temperatures. Other

conditions described by White and Roswell [1] are that the necessary energy must

be made available in a single reaction step and the molecule receiving this energy

must have a limited number of accessible vibrational energy states which would

otherwise act as an energy sink. If the excited state is emissive, it can








2
chemiluminesce directly; however, it can also transfer the energy to another

molecule, which following excitation then emits the energy as light. The

luminescence exhibited by luminol was one of the earliest cases of

chemiluminescence to be studied. Luminol (5-amino-2,3-dihydrophthalazine- 1,4-

dione) is an aminophthalic hydrazide (Fig 1) that is able to exist in several

tautomeric forms.


Figure 1: Luminol (3-aminophthalic acid hydrazide)





For luminol, the light emitting species is the excited 3-aminophthalate ion, which

emits a blue emission in water and a yellow green emission in dimethyl sulfoxide







3
[2]. This light emitting species is generated by an oxidative system that is usually

composed of either sodium or potassium hydroxide, hydrogen peroxide as the

oxidizing agent and an activator. The activator is commonly a transition metal ion,

metal-ion complex, hypochlorite, ferricyanide or perchlorate. The solvent system

has frequently been aqueous. However, lower alcohols as well as other water

miscible organic solvents such as dimethyl sulfoxide or dimethylformamide have

also been used in place of water [3]. The activating agent is not absolutely

necessary as in its place sonic waves have been used and in some aprotic media

no activating agent is necessary, only oxygen and a strong base.

In aprotic media the products from luminol oxidation have been isolated.

However, in aqueous systems the corresponding diacidic anion, despite being

quantitatively produced from the hydrazides in basic reaction media, has been

difficult to detect due to apparent further oxidation of the initially formed product.




Reaction mechanism


Basic solutions of hydrogen peroxide are common reagents in hydrazide

chemiluminescence. In these solutions the hydroperoxide ion which stems from the

initial reduction of oxygen is believed to be a critical reactant. The hydroperoxide

ion formed then reacts with the azaquinone form of luminol, by attacking one of

the carbonyl groups, leading to the formation of III. Compound III is then believed

to decompose by one of a variety of postulated routes to the phthaloyl peroxide








4

derivative VI, which then undergoes ring opening to yield the excited 3-amino-

phthalate ion as illustrated in Fig 2. The mechanism of decomposition of the

azaquinone has been the subject of a number of postulations, which have been

reviewed by White and Roswell [1].


Application of Luminol Chemiluminescence to analysis in solution

The intensity of chemiluminescence arising from the chemiluminescent

reaction can be described by the following equation.



I dC
ICL =+CL -a


where CL is the efficiency of the chemiluminescence and dC/dt the rate of the

chemical reaction. The efficiency of chemiluminescence depends on how

efficiently excited states are generated from the molecular reaction and on how

efficiently the excited states luminesce; i.e.,

ICL =4c*0 401.


where oxc is the excitation efficiency and 0,m the luminescence efficiency. Both

the excitation and luminescence efficiencies can be influenced by a variety of

reaction conditions such as solvents employed, concentration, pH and purity of

reagents.














pH= 10
go-


N-
I
.NH-


OH H0
0HH20


7 -
"OOH


N2


NH2


Figure 2: Schematic outline of chemiluminescent reaction pathway.







6
The intensity of luminescence can be used as the basis for determination

of any species whose concentration influences the rate or efficiency of the

chemiluminescent reaction. In applying chemiluminescence to analysis of the

species the reaction conditions should be adjusted so that the analyte of interest

is the limiting reagent in the system and all other reactants are in excess. To obtain

precise measurements the chemiluminescence reaction should be initiated in a

controlled and reproducible manner, largely because the emission intensity varies

with time as the reactants are consumed. Chemiluminescent analyses are reported

to have several advantages which include good sensitivity, a wide linear dynamic

range, low detection limits (femtomole to attomole range) and the requirement for

simple instrumentation [4].

The chemiluminescent signal is transient; hence, the measurement of

emitted light intensity is time dependent. The signal is therefore either recorded at

a specific time after mixing, or by integrating the light emission plot during the

entire time period or during a specific fraction of time when light is emitted.

Commonly the reactants are rapidly mixed and the emission intensity measured

as a function of time after mixing. This yields the plot illustrated below (Fig 3). The

initial part of the curve is influenced by the method of mixing employed, while the

general shape of the curve depends upon the kinetics of the reaction as well as

upon any changes in quantum yield with time.

Most chemiluminescence reactions are reported to have low efficiencies, less

than 10% [5], which has restricted their usefulness to analyses. The duration of the








7

reactions is influenced by the reaction conditions and may occur rapidly within one

sec, or last longer than 24 hours. In the development of chemiluminescent assay

methods the two basic factors that influence the intensity of chemiluminescence

(i.e. efficiency and rate) should be considered. The efficiency of the reaction

influences both analytical sensitivity and detection limits, while the reaction kinetics

determine both the precision and sample throughput.

Any substance able to quantitatively influence the light output can be

determined by chemiluminescence. In fact the luminol reaction has been used to

determine a number of compounds that are able to interact with the oxidant

initiating the chemiluminescent reaction. The substances so quantitated have not

necessarily been the analyte of interest but rather related to the analyte. These

substances have been divided into three categories by Grayeski [4].

Firstly those measured species that form one of the reagents that is consumed

in the course of the reaction. This case is exemplified by the determination of

hydrogen peroxide using the luminol system. This approach was applied to the

determination of hydrogen peroxide in irradiated water as early as 1955 [6].

The second category of reactions includes the analysis of compounds that are

able to generate one of the chemiluminescent reactants. An example here is the

indirect determination of glucose by treatment with the enzyme glucose oxidase.

The enzyme oxidizes the glucose to gluconic acid and hydrogen peroxide, the





























0.08


S0.06
C


0 500 1000 1500 2000 2500 3000 3500 4000
time (sec)


Figure 3: Typical luminescence profile








9
hydrogen peroxide so formed then being assayed by the luminol reaction.



Glucose Gluconic acid + 1102O

luminol + 1102 # aminophthalate + hv



More recently Zhou et al [7] reported a method for the determination of Vitamin

B12 by means of the luminol-hydrogen peroxide system. In their experimental

setup the bound cobalt in vitamin B12 is released by acidification of the vitamin. The

cobalt so released is then permitted to quantitatively catalyze the oxidation of

luminol by hydrogen peroxide.

The coupling of reactions is not always possible as there is often the problem

of incompatibility of conditions for all the reactions. This has however not impeded

use of the technique for the analysis of a range of biochemicals.

The third category consists of those substances able to modify the primary

chemiluminescent reaction. The analysis of a wide range of substances falls into

this category. This includes the analyses of metal ions such as Arsenic(As3.),

Cobalt(Cc* ), and Nickel(Ni2), nonmetallic inorganics such as the gases oxygen,

and nitrogen dioxide, the halide ions and a number of nitro-, amino-, or hydroxy-

group containing organic. These substances are reported to have either an

excitatory or inhibitory influence on the chemiluminescence exhibited by luminol.

More recently a number of drug substances have been reported to enhance a

variety of otherwise chemiluminescent reactions. For example a method has been







10
reported for determining quinine from the ability of quinine to enhance the

chemiluminescence exhibited by the oxidation of sulphite by cerium(IV) [8].

Similarly the determination of acetaldehyde by monitoring the chemiluminescence

emission from the luminol hexacyanoferrate (III) reaction in the presence of

xanthine oxidase has been described [9]. The steroidal hormones hydrocortisone

and betamethasone as well as the antihistamine promethazine have also been

determined from their ability to enhance the chemiluminescence of the cerium(IV)-

sulphite system [10,11]. The chemiluminescent oxidation of the antimycobacterial

isoniazid by N-bromosuccinimide has also been applied to the determination of

isoniazid [12]. A number of penicillins and cephalosporins have also been reported

to enhance the chemiluminescence exhibited by luminol [13,14,15]. Schulman et

al [14] Chen etal [15] have reported the enhancement luminol chemiluminescence

exhibited by several penicillins, and the cephalosporin cephalothin.

The instrumentation employed for chemiluminescence measurements

basically consists of a mixing device and a detection system. Three approaches

have been used to measure the intensity of emitted light.

The first approach involves the use of a static measurement system in

which the mixing of reagents is performed in a vessel held in front of the detector.

The chemiluminescent reagent is added to the analyte in a cuvette held in a dark

enclosure, and the intensity of the light emitted is measured through an adjacent

photomultiplier tube. In this static system the mixing is induced by the force of the

injection. This simple reagent addition system without a mixing device is of limited







11
usefulness in measuring, with precision, fast chemiluminescent reactions. The

procedure is also rather cumbersome requiring separate cuvettes for each

measurement as well as repeated opening of the light tight apparatus, which

necessitates special precautions to protect the photomultiplier tube.

The second approach is the two phase measurement system. In this system

the chemiluminescent reagents are immobilized on a solid support such as filter

paper, and the analyte is permitted to interact with the immobilized reagent by

diffusion or convection. The light emitted is measured using a microliter plate

reader or by contact printing with photographic detection [4]. The

chemiluminescent intensity of these systems is influenced by both the kinetics of

the reaction and the efficiency of the mass transfer processes bringing the

reactants together. The principal advantages of this system lie in the conservation

of reagents and the convenience of measurement.

The third approach involves the use of flow measurement systems. The

flow injection approach has been described as the most successful of the methods

[5]. It involves injection of the analyte into a stream of appropriate pH, remote from

the detector, and the chemiluminescent reagent flows in another stream. The two

streams meet at a T junction inside a light tight enclosure, then flow through a flat

coil placed immediately in front of a photomultiplier tube. This compact assembly

provides more rapid and reproducible mixing resulting in reproducible emission

intensities and permitting rapid sample throughput. The design of the mixing device

and the means of retaining the emitting solution in view of the detector are







12
important considerations. Mixing is reported as being most effective at a T piece

or Y junction; however, some workers have used the conventional FIA system in

which sample is injected into the surrounding flowing reagent to achieve mixing.

This approach is reproducible but reported as not producing rapid mixing [5]. In

flow measurement systems the chemiluminescent signal has to be measured

during the mixing, as a result only a section as opposed to the whole-intensity-time

curve is measured. An additional feature of these systems is that the shape of the

curve is now dependent on both the kinetics of the chemiluminescent reaction and

the parameters of the flow system. Chemical variables such as reagent

concentrations, pH, etc and physical parameters such as flow rate, reaction coil

length, sample size and the limitations of experimental apparatus all affect the

performance of flow injection procedures. The effects of these variables on the

observed analytical signal are not necessarily independent, as interactions can and

indeed do occur [16]. Most of the chemiluminescence determinations reported

have involved the determination of compounds able to quench the luminescence

of luminol or other chemiluminescent systems. A smaller number have been

reported for substances enhancing the chemiluminescent signal. A variety of

determinations fall into the latter group of compounds able to sensitize the

chemiluminescent reactions. The chemiluminescent systems have not necessarily

involved the luminol-peroxide system alone, but have included among others the

Cerium(IV)-sulfite system in which the luminescence arises from the oxidation of

sulphite by cerium(IV), hexacyanoferrate (111), the peroxodisulphate system and the









13

bromine based oxidative systems [17]. Examples of these analyses and some of

the analytical parameters reported are listed in Table 1.









Table 1: Analyses in which luminescence has been applied.



Analyte Reagent linear range Detection Ref
(ug/ml) limit (ug/ml)


Oxytetracycline hexacyanoferrate 1-10 0.5 17
Quinine Ce(IV) sulfite 5-500 0.64 8
Penicillin G *lum-perox-Co(ll) 0.0001 -0.01 6 exp(-5) 14
Promethazine Ce(IV) sulfite 10.3-51.3 11
Hydrocortisone Ce(IV) sulfite 0.1 -1.0 0.016 10
Betamethasone Ce(IV) sulfite 0.5-5.0 0.3 10
Isoniazid bromosuxinimide 0.05-20.0 0.024 12
Acetaldehyde #lum-XO-Fe(CN) exp(-7) exp(-3) 4 exp(-7) 9
Tetracycline hexacyanoferrate 0.1 -1.2 0.04 17
Doxycycline hexacyanoferrate 0.1 -10.0 0.2 17
Chlortetracycline hexacyanoferrate 1.0-10 0.1 17


* static measurement system
# lum-XO-Fe(CN) luminol-xanthine oxidase-hexacyanoferrate












Optimization of the analytical response

Most workers have arrived at optimal experimental conditions by examining

a single factor at a time, determining the optimal value for that factor before

proceeding with the next which was then optimized at the optimal level of the

preceding factorss.

Owing to the large number of variables (physical, chemical and

physicochemical) that are able to influence the signal obtained, there has been a

trend towards the application of optimization methods to analytical flow streams.

The application of chemometric principles to the determination of the optimal

combination of parameters that would afford the best analytical sensitivity,

precision and analytical robustness is relatively recent. Chemometric principles

have in the main been applied to optimizing chromatographic systems; however,

they can and have been applied to the optimization of analytical flow systems.

Iterative designs as opposed to grid search methods have proved to be the

more useful in optimizing analytical experimental conditions, the simplex approach

is the most commonly employed. The basic simplex approach as introduced by

Spendley et al [18], however, has the disadvantage of taking a rather long time to

locate the maximum, and the attendant risk of determining a secondary as

opposed to the primary maximum. Hence a number of variations to the basic

approach, variously described as modified simplex designs, have been developed

to facilitate determination of the true as opposed to secondary maximum and to








15

accelerate determination of the optimal conditions [19 23]. Indeed automated

approaches have been developed to reduce the search time involved and develop

more comprehensive response surfaces. In automated methods, pattern search

methods have been reported [24] as preferential, over simplex based procedures

where up to three experimental variables are being considered. In the literature a

number of applications of a modified simplex method to the optimization of

analytical flow streams have been described. Calatayud and Sancho [11] applied

a modified simplex optimization to the oxidative determination of promethazine by

flow injection analysis, similarly Chen [25] applied a more complex modified

simplex method to select the optimum reaction conditions for the 1,10-

phenanthroline / 1-12 /-cetyl-trimethylammonium bromide / Cu2 system.

The choice of an appropriate response function is probably the most critical

step in optimization. The response function selected should ideally provide a

means of optimizing all the factors able to influence analytical characteristics as

sensitivity, selectivity, precision, etc. In most analyses the response function

selected is usually the peak height at the peak maximum, though other peak

characteristics such as peak area and peak width have been employed. Indeed a

number of response functions that may be useful have been described by

Betteridge et al [26].








16
The Betalactam antibiotics

The penicillins and the cephalosporins are established groups of antibiotics

characterized by the presence of a substituted p-lactam fused to a sulfur

containing ring. For the penicillins the S-containing ring is a thiazolidine ring

whereas for the cephalosporins it is a dihydothiazine ring.

The penicillin and cephalosporin antibiotics both exert their antibiotic activity

by inhibiting the synthesis of bacterial cell walls. Thep lactam ring is reactive and

is believed to be inactivated following acylation by the transpeptidase enzyme

which normally crosslinks peptidoglycan strands during cell wall synthesis. The

four membered p- lactam ring is strained not only as a result of ring size but also

as a result of fusion to the second ring. The immediate consequence of this is non-

planarity of the p -lactam ring compromising amide resonance within the /-lactam

ring. In the cephalosporins this ring fusion effect on electron delocalization is

amplified by the enamine resonance outside the lactam ring. Reports in the

literature have suggested that the mechanism of action is related to both the liability

of the p-lactam amide bond and conformation of the antibiotic in the region of the

p- lactam ring [27]. The effect of ring size on ring-fused p-lactam reactivity as well

as substituent effects on base hydrolysis of the compounds has also been studied.

Generally the biological activity of the penicillins and cephalosporins has been

correlated to

degree of non-planarity in the p- lactam nitrogen

the lactam C=0 stretching frequency







17
as well as to the ease with which they undergo base hydrolysis.

A wide range of techniques have been used to develop methods for the

analysis of penicillins and /or cephalosporins.

Microbiological methods have found limited usefulness in the quantitative

determination of penicillins. Their main use has beer in qualitative determinations,

confirming microbial sensitivity to the compounds and in determining

therapeutically effective levels.

Titrimetric methods have been the mainstay of p- lactam analyses, these

have typically involved degrading the penicillin, by use of acid and/or base

solutions, to penicillamine or some sulfhydryl containing molecule which is then

analyzed by a mercurimetric or iodometric titration. The methods are limited by the

fact that they determine "total penicillins" necessitating a blank to correct for the

presence of degradation products) [28,29].

Colorimetric methods have similarly involved degrading the intactfl- lactam

to generate a chromophore with characterizable uv absorption spectrum. This has

formed the basis for the hydroxamate, methylene blue, dinitrobenzoic acid and the

more recent mercurimetric assay methods [30,31,32,33].

Optical rotatory dispersion (ORD) and circular dichroism (CD) techniques

have found limited application to the analysis of penicillins. They have been mainly

applied to kinetic determinations of the compounds.

Infrared techniques have not found use in the quantitative analysis of p-

lactams but rather have found use as diagnostic tools in qualitative determinations.








18

Of interest here is the p- lactam carbonyl stretch frequency which has been useful

in providing information on the structural integrity of the rings, the state of oxidation

of sulfur and the relative conformations of the #- lactam protons. In this respect an

increase in the #- lactam stretching frequency has been associated with an

increase in biological activity, a decrease in the lactam N planarity and increasing

ease of lactam amide bond hydrolysis by base.

Of the chromatographic methods gas chromatography has found limited

utility in the analysis of penicillins largely as a result of the non-volatility of the

penicillins and cephalosporins. Uquid chromatography has been used extensively

in both the development and application of the p lactam antibiotics. The antibiotics

have been separated on reverse phase columns typically octadecylsilane (ODS)

columns employing methanol, acetonitrile, aqueous buffers or combinations thereof

as the mobile phase. The detection method has typically been by uv at 254 nm.

The penicillins have been reported to enhance the luminescence arising

from the oxidation of luminol in alkaline media [13,14,15]. This enhancement has

been shown to be quantitative for penicillin V, penicillin G and cephalothin. It is

hypothesized that this enhancement is characteristic of all p- lactam antibiotics.








19
CHAPTER 2

OBJECTIVES




This study was essentially an extension of the work carried out by Schulman

et al [14] and Chen et al [15] on the enhancement of luminol chemiluminescence

by selected beta lactam antibiotics. Chen and coworkers carried out their analyses

in a static system, reporting "good" reproducibility with benzyl penicillin, piperacillin

and phenoxymethyl penicillin (penicillin V) and the cephalosporin cephalothin.

In the current study an attempt is made to answer the following questions.

1). Do all p-lactam antibiotics whether penicillin or cephalosporin enhance the

chemiluminescence signal?

2). To what extent do the compounds examined enhance the

chemiluminescence?

3). What structural features are necessary for the enhancement to be

measurable?

4). What is the mechanism of enhancement?

5). Can the chemiluminescence be applied to post-column detection in HPLC?

If so, what set of conditions would achieve maximal sensitivity, the best limit

of quantitation and best precision?



Chen et al were able to demonstrate that penicillin G, penicillin V, piperacillin and

cephalothin enhanced the chemiluminescence to varying extents. Preliminary








20
experiments since then have confirmed this, but have also demonstrated that a

number of penicillins do not enhance the chemiluminescence under the given

experimental conditions.

A systematic survey of the penicillins on the basis of chemical structure

should enable the prediction of which p-lactams enhance chemiluminescence. For

this purpose the penicillins can be divided into the following chemical groups, from

which an available antibiotic can be examined for the ability to enhance the

chemiluminescence.

penicillin G

phenoxyalkyl penicillins

isoxazolyl penicillins

ampicillins and related compounds

N-acylated ampicillins

cephalosporins

cephamycins

The extent(s) of enhancement will be compared in terms of the analytical

parameters such as the linear dynamic range, sensitivity, limit of quantitation,

precision, etc. Due to the poor reproducibility associated with variations in mixing

a flowing analytical stream was employed.

Finally an attempt is made to correlate the results obtained to the

reactivities/stability of the respective beta-lactam antibiotics and the application of

the method for determination of the -lactam antibiotics in dosage forms, indicating








21

accuracy, precision, repeatability and reproducibility of the methodss.














azetidinone
H3C(HO)HC

O/ SCH2CH2NH3


COOH

thienamycin
6- aminopenicillanic
acid .,,,


clavulanic acid
sulbactam


x


HH3
N S COOH
Y 0H3
Y& 0)):] 0C


Benzylpenicillin

ampicillin

amoxicillin


X=H, Y = H

X=NHI, Y= H

X=NH,, Y = OH


OCH3

H
c N S OH3
0
OCH3 O CH3
COOH
COOH


Figure 4: Chemicals structures of compounds examined.


NH


0


methicillin























Isoxazolyl penicillins;

cloxacillin

dicloxacillin

flucloxacillin


Piperacillin














penicillin V X = I


phenethicillin X = C113


Figure 4--continued.













CHAPTER 3
EXPERIMENTAL





Reagents

Fluorescein sodium and Rubrene were purchased from Fluka Chemical Co

(Buchs, Switzerland), Aminoacridine from Sigma Chemical Co (St Louis, MO), and

Rhodamine S from K & K Laboratories (Plainview, NY). All other reagents were

supplied by Fischer Scientific Co (Fairlawn, NJ) The antibiotics were obtained from

a variety of sources. Azetidinone from Aldrich Chemical Co (Milwaukee, Wl),

Ampicillin (ACS DOBFAR) from Interchem Corporation (Paramus, NJ), Hetacillin

was provided by Beecham Research Laboratories (Syracruse, NY), N- formimidoyl

thienamycin (Imipenem) was a gift from Merck & Co (Rahway, NJ), Methicillin and

Phenethicillin from Bristol Laboratories (Syracuse, NY), Lithium clavulinate a gift

from Smith Kline Beecham Pharmaceuticals (Philadelphia, PA), Sulbactam from

Pfizer Cephalothin from Eli Lilly Labs (Indianapolis, IN). The other antibiotics were

either purchased from Sigma Chemical Co (St Louis, MO) or Aldrich Chemical Co

(Milwaukee, Wl).

All chemicals and reagents were used as provided without further

purification.









Preparation of Luminol solutions.

Solutions of 0.001 M luminol were prepared by dissolving 88.6 mg 3-

Aminophthalhydrazide into a mixture of about 25 ml water and 2 ml of 2M sodium

hydroxide with stirring, then adjusting to volume in a 500 ml volumetric flask. The

solutions were employed for 48 hours prior to discarding.



Preparation of Hydrogen Peroxide solutions.

The hydrogen peroxide solutions were prepared by pipetting 2.0 mls of a

30% hydrogen peroxide solution into a 500 ml volumetric flask containing 2.0 ml

of 2M sodium hydroxide in about 25 ml water, then made to volume with water.



Preparation of Stock Sodium Dihvdroaen Phosphate solution.

A Stock solution of monobasic sodium phosphate (NaH-PO4) 0.01 M was

prepared by dissolving 690 mg of the anhydrous salt into 500 ml water. From the

stock solution 50 ml aliquots were diluted to 500 ml to obtain the 0.001 M

solutions.



Preparation of Mobile phase.

To 35.0 ml methanol in a measuring cylinder was added the 0.001 M

sodium dihydrogen phosphate solution to obtain a final volume of 100 ml. The

solution were filtered through a 0.45/ m membrane filter and degassed by stirring

the mobile phase under vacuum/suction.











Preparation of Calibration series.

Stock solutions of the p- lactam compounds were prepared in deionized

water prior to each analyses. Appropriate volumes of the solutions were then

diluted with water to obtain the desired concentration ranges. 20 /j I aliquots were

injected into the chromatographic system. Mobile phase methanol in 0.001 M

sodium phosphate (35 % MeOH) pH 6.3.






Ultraviolet Spectrophotometric studies

About 15 mg dicloxacillin was accurately weighed out into a 10 ml volumetric

flask and dissolved in a small quantity of 0.001 M sodium hydroxide before being

made to volume. A 2 ml aliquot of this stock solution was then diluted to 10 ml in

a volumetric flask to given the experimental solution. A 2.0 ml aliquot of the

experimental solution was transferred to a cuvette and used for the

spectrophotometric studies.

Employing a Shimadzu UV 160U instrument the ultraviolet spectrum in the

wavelength range 250 360 nm for the dicloxacillin sample was obtained against

0.001 M sodium hydroxide as reference and thereafter following addition of 10, 50,

80 and 100 ul of a 30 % 1202 solution. A spectrum was also obtained for a 100 A I

aliquot of -202 diluted in 2 ml of the 0.001 M sodium hydroxide solution.










Static studies.

The experimental setup for the studies is schematically illustrated in Figure

5. The reagent solution was prepared by diluting a 10 Il aliquot of 30 % hydrogen

peroxide to 10 ml in a volumetric flask using a 104 molar stock solution of 2-

aminophthalhydrazide in 0.001 M sodium hydroxide. For experimental runs, 201 I

of the dicloxacillin was added to 2 ml of the reagent solution in the cuvette.

Earlier studies involved the use of cobalt and copper ion solutions at a

concentration 104 M. They were however found unnecessary in preliminary

investigations. The metal ions employed were cobalt (II) and copper (II).

An alternate procedure adopted was to inject a specified 11 volume of 30%

1202 into the cuvette containing the luminol solution, prior to injection of the

antibiotic solution. Following injection of the antibiotic solution, the Intensity-time

profiles obtained were recorded on a disk over a specified time period.

Conductivity measurements

To a beaker containing 25 ml of a 4 X 1C04 M stock solution of luminol. was

added 25 ,1 of the 30 % H202 The changes in conductivity with time were then

measured using a conductivity electrode (Markkson Instruments).















syringe


light-proof enclosure


L-photomultiplier tube


1


cuvette


Figure 5: Schematic diagram of setup for static chemiluminescent
measurements.


.--------------------



L-----------------.--








29
Investigation of possible Peroxylate Based Mechanism.

Protocol for comparing different fluorescers

Approximately 1C4 M solutions were prepared by dissolving the appropriate

quantity of fluorescer (Fluorescein, Rhodamine S, aminoacridine or Rubrene) into

a 103 M sodium hydroxide solution. A 2 ml aliquot was then transferred into a 3

ml cuvette placed into the static setup illustrated in figure 4 and to this was added

a 100 I1 mixture of 1-102 (30%) and dicloxacillin (approx 1 mg/ml) (50:50). The

emission obtained was recorded on disk over a 1000 s time period using the A-D

board and Spectracalc software.



Protocol for comparing different antibiotics

Approximately 104 M solution of fluorescein was prepared by dissolving

about 38 mg of fluorescein sodium (anhydrous M.W 376.3) accurately weighed

into a 10U3 M sodium hydroxide solution. A 2 ml aliquot was then pipetted into a 3

ml cuvette, placed into the static setup (Figure 4) and a 100 pl mixture of H-202

(30%) and the appropriate p- lactam antibiotic (approx 1 mg/ml) (50:50) was

added.

The emission obtained was recorded on disk over a 1000 s time period using the

A-D board and Spectracalc software.








30
Flow studies

Flow Injection Analyses.

The set up employed is illustrated in Figures 6 and 7. A similar set up was

employed for chromatographic studies. In the two channel system the flowing

streams were alkaline peroxide into which the analyte was injected and alkaline

luminol.

For the three channel system flowing streams were alkaline peroxide, alkaline

luminol and water/phosphate/water methanol mixture. In both cases samples were

injected through a Rheodyne loop (Cotati California) injector fixed with a 20;1 loop.



Chromatoaraohic studies



The mobile phase primarily consisted of methanol-water mixtures of varying

proportions, with or without modifiers. The modifiers used were either phosphate

or imidazole solutions. In the screening experiments varying proportions of water

and other water miscible solvents e.g tetrahydrofuran, acetonitrile, were employed.

The flow rates were nominally 1 ml.min1, for the analytical stream and 0.5

ml.minrf1 for the reagent streams. When a flow rate change was necessary an effort

was made to retain this relative ratio to minimize band broadening arising from

post-column mixing.

A schematic representation of the setup is as illustrated in Figure 8.













analyte


peroxide (|
stream

pump photocell


metal ion )
stream 0 C

pump 2

luminol o__________
stream

pump 3 waste


Figure 6: Three channel flow system











analyte


stream -\
pump 1










luminol AC )
stream M n


photocell


pump 2


waste


Figure 7: Two channel flow system








column


mobile phase.


to waste


reagent I.


reagent II.


/
data acquisition
device.


mobile phase ; primarily water miscible mixtures
reagent streams ; an alkaline luminol stream and an alkaline peroxide stream

column; silica, octylsilane or octadecylsilane


Figure 8: Chromatographic setup employed.








34
Chromatographic columns employed in the studies include

Exsil R ODS Keystone scientific Inc

Dimensions 150 X 4.6 mm
Particle size 7Mm
Pore size 100 A

Nucleosil R C18 Keystone scientific Inc

Dimensions 150 X 4.6 mm
particle size 5 m
pore size 100 A

Microsorb Si Rainin Instrumant Company Inc

Dimensions 250 X 4.6 mm
particle size 5 jm
pore size 100 A


For the most part the mobile phase consisted of

methanol: 0.001 M sodium/potassium phosphate NaH2PO4 (35:65)
Flow rate 1.0 ml.minr'1.

Detection system

post-column chemiluminescence.
0.001 M luminol in 0.008 M NaOH 0.5 ml.mirf'n
-202 in 0.008 M NaOH 0.5 ml.minr1

ultraviolet at 254 nm

NB; All metal (stainless steel) tubing was of 0.010 inches internal diameter.











Protocol for Calibration Curve determinations

A 1 X 104 M solution of luminol made alkaline in 0.008 M sodium hydroxide

was employed as one of the reagent streams. The alkaline hydrogen peroxide

stream was prepared by diluting 1 ml of a 30% v/v stock solution to 500 ml in a

flask using 1 X 1U3 M sodium hydroxide. The two solutions were employed as

reagent streams in a two channel flow injection system as illustrated in Figure 6.

A 20 AI aliquot of the analyte solution in the respective calibration series was

injected into a stream of the mobile phase.



Equipment

1). FL-750 HPLC plus spectrofluorescence detector

McPherson Instrument Acton, Massachusetts.

2). Pharmacia LKB HPLC pump 2150

LKB-Produkter Bromma, Sweden.

3). LDC Analytical Constametric metering pump

LDC Analytical Riviera beach, Florida.

4). Servogor 120 recorder

Norma Goerz Instruments Osterreich, Austria

5). IBM PC XT

6). DT 2811 Analog and Digital Input/Output board

Data Translation Inc Malbora, Massachusetts.










7). Selectro Mark Analyzer

Markson Science Inc DelMar, Colorado.







Software

1). Spectracalc

Galactica Industries Corporation Salem New Hampshire.

2). Acuchem series 4.0

40 species version

NIST, Gaithersburg Maryland.

Source of Water

All water employed in the experiments was purified by filtration through a

millipore Milli-Q 50 water filtration apparatus Millipore Corporation (Bedford, MA)

prior to use.













Table 2: Antibiotic compounds screened for enhancement


Compound


Source


2- azetidinone
6- aminopenicillanic acid
amoxicillin
ampicillin
benzylpenicillin
cefotaxine
cephalosporin C
cephalothin Na
clavulinic acid
d- benzylpenicillenic acid
d- penicillamine
dicloxacillin
hetacillin
imipenem
methicillin
phenethicillin
phenoxymethylpenicillin
piperacilin
sulbactam
imipenem


Aid rich
Sigma
Sigma
ACS Dobfar
Sigma
Sigma
Sigma
Eli Lilly Labs
Bristol
Sigma
Sigma
Sigma
Beecham Research Labs
Merck Sharp & Dohme
Bristol Labs
Sigma
Sigma
Pfizer
Fluka
Merck & Co













CHAPTER 4
RESULTS AND DISCUSSION




Mechanistic studies

Mechanistic studies were carried out with the objective of being able to

propose a reaction model that would satisfactorily simulate the light emission

profile obtained on addition of the p lactams to the reagents. The simulations were

performed using the chemical kinetic software HICHEM [34] and the more recent

improvement ACUCHEM [34,35]. Much of the work in elucidating the mechanism

of luminol chemiluminescence was pioneered by White et al [36] and expanded

upon by Lind, Merbnyi and Eriksen [37,38,39]. In most of the reaction

mechanisms radicals, ions and radical ions are proposed as being the reactive

species in the various steps preceding formation of the luminescent species.

White et al [36] proposed two consecutive loses of protons to generate a

di-anion which is then able to undergo oxidative decomposition generating the

excited aminophthalate ion, which chemiluminesces as it returns to the ground

state (Figure 9). To this series of consecutive reactions, White et al applied the

steady state approximation. Making among others the assumption that k.1 (H20)

> > k,2(2) White et al derived the rate expression for photon emission.








39
rate = k [III] [OH] [02] (1)

where III is the monoanionic luminol base and

kk
k=. kjk2
k1 [H20]



A pseudo-first order rate constant of 2.5 X 10-3 s1 at 35 C was reported for the

monoanionic luminol base.

Merenyi and coworkers [37,38,39] in a series of publications also outlined

a possible mechanism for the chemiluminescent reaction of luminol. In the earliest

of the three papers [37], they describe the hydroperoxide as being the first critical

intermediate in the chemiluminescent pathway of luminol and suggest that the

hydroperoxide decomposes to the chemiluminescent aminophthalate ion at pH

values above its pK, (pK, = 10.4 0.1), whereas at pH values below its pK, the

hydroperoxide decomposes back to oxygen and its parent hydrazide. The paper

did not propose a mechanism for the formation of the hydroperoxide or the nature

of the oxidation stage.

In a later publication Merenyi et al [38] proposed a pathway for the

oxidation of the one-electron-oxidized luminol by molecular oxygen reporting a

one-electron reduction potential of 0.240 0.02 V vs NHE for 5-aminophthalazine-

1,4-dione. However little indication is provided as to how the one-electron-oxidized

luminol is generated as well as an indication of the levels of molecular oxygen

and/or 02-. in aqueous alkaline solutions.


















NH2 0


OH


H20 NH2 0


+ 02


+ N2


Figure 9: Reaction scheme after White et al [36].


pK, =6


-OH


H,O0








41
More recently Sattar and Epstein [40] outlined an overall reaction scheme

for the reaction of luminol with basic hydrogen peroxide from pulse radiolysis

studies carried out by Merbnyi and coworkers [37,38,39] (Figure 10).

The precise role of a number of compounds that sensitize chemiluminescent

reactions has not been fully elucidated. Various suggestions as to the role of the

sensitizing agents have ranged from being direct catalytic activators to acting via

complexation with the catalytic metal ions (usually Cor, CLu?) or some reaction

intermediates [41]. Some workers in the field [15,41] have suggested that the

enhancement mechanism involves the formation of an enhancer-peroxide adduct

which is then able to more rapidly generate the endoperoxide from luminol. The

endoperoxide then decomposes into nitrogen and the excited aminopthalate ion

which then serves as the luminescent species.

The concept of enhanced oxidative capability has been cited in a more

recent study of the enhancement of luminol chemiluminescence by nitric oxide.

Kikuchi et al [42] report that enhancement is due to the formation of peroxynitrite

from NO and H202. They demonstrated the formation of the peroxynitrite (ONOO')

anion, a strong oxidizing species, by demonstrating the existence of its specific

maximum at 302 nm in the uv spectrum. As regards the possible role of the

betalactam antibiotics in enhancing the chemiluminescence, Chen et al [15]

proposed the formation of an adduct between penicillin and the superoxide anion.

Employing difference spectrophotometry, they demonstrated the existence of a

chromophore with a maximum between 250-260 nm. This adduct was presumed





OH


02"

k,


NH2 0


OH


.OOH

"N
II


H-,0


k2OH
OH


+ N, + H20


dark product


Figure 10: Reaction scheme after Sattar and Epstein [40].


+ H'








43
to lengthen the lifetime of the oxidant. The penicillins however have been reported

by Martinez et al [43] to form metal ion complexes with cobalt whose spectra may

differ under various conditions or circumstances. Furthermore given that the

luminescence is known to be enhanced without the metal ion the postulated

adduct is not necessarily responsible for the enhancement alone. The ultraviolet

visible spectrum obtained on addition of I-I0O to dicloxacillin solutions, in the

absence of a metal ion, did not demonstrate the existence of a unique

chromophore (Figure 11). The demonstrated existence of an adduct alone does

not suggest that the adduct is responsible for the enhancement, it is also important

to show that the adduct does indeed enhance the luminescence. In the case of

nitric oxide, the peroxynitrite anion is an established oxidant, and demonstrating

a stronger oxidative potential for the adduct would lend credence to the argument.

Furthermore, it was not possible to demonstrate the existence of a uv light

absorbing adduct on the addition of the oxidizing agent sodium hypochlorite.

The oxidant activity could be demonstrated by redox potential measurementss,

possibly using a platinum electrode against a silver-silver chloride or other suitable

reference. Suitable equipment was not available to conduct the necessary

experiments.

A second possible mode of luminescence enhancement was that the

enhancer was simply able to donate the energy arising from its oxidation to 2-

aminophthalate which then re-emits the energy as light, similar to the

chemiluminescence exhibited by the peroxyoxalates. The possibility that the





































wavelength (nm)


--- 0% -*-+ 0.75% --ct1.2%
-E- 1.5 % --X<-- 1.5 % (control)


Figure 11: Ultraviolet spectrum of dicloxacillin on addition of varying concentrations
of H 02.








45
luminescence arose from a peroxyoxalate-like process associated with the opening

up of the p- lactam ring in the penicillin was also examined. The rationale being

that the opening up of the strained #- lactam ring is the primary energy releasing

reaction that simply requires an appropriate fluorescer, which is able to absorb the

energy and release it as light. The strained lactam ring would open subsequent

to nucleophilic attack of the electron deficient keto group by the peroxy anion. In

opening the p- lactam ring the energy released excites a receptor molecule which

then emits the energy as light. In the case of luminol the receptor molecule could

possibly be luminol per se or the aminophthalate ion. Several experiments were

carried out to determine whether this was indeed the case. Initial experiments

suggested that 3- aminophthalate and fluorescein were both able to support the

p- lactam enhanced chemiluminescence in the presence of hydrogen peroxide.

However the enhancement from mixtures of aminophthalate and hydrogen

peroxide, was found to diminish significantly with time on standing. This could have

been due to oxidation of the fluorescer. Mohan [44] has indicated the importance

of taking into account the fluorescer stability to hydrogen peroxide and photo-

oxidation. A possible variation of the mechanism is that the observed emission

could arise from the excitation of the fluorescer by the light energy emitted by

luminol which was then reemitted at a longer wavelength, by a mechanism akin to

that of fluors in liquid scintillation counters. Several fluorescent compounds were

tested (Table 3) and where possible intensity-time profiles were obtained.

Comparison of the chemiluminescence spectrum of the reaction mixture to that








46
of the pure luminol, fluorescein or any other fluorescent enhancer would help

ascertain this aspect of the mechanism. It was also shown that the more efficient

fluorescers rhodamine and fluorescein generated more intense light output profiles

than the aminophthalate ion (Figure 12). The intensity time emission profiles for the

oxidation of selected penicillins in the presence of fluorescein were also obtained

(Figure 13).

These observations led to the development of a peroxylate type

chemiluminescence mechanism, in which the opening of the p- lactam ring is the

primary energy releasing reaction. Subsequent events are as outlined in Figure 14.

Though opening of the p-lactam ring was the more obvious source of energy, it

is recognized that oxidation at other sites on the antibiotics molecules could serve

as energy sites. The suggested mechanism is supported in part by the recent

observations by Chen et al [46,47] that selected xanthone dyes exhibited

chemiluminescence on oxidation by hydrogen peroxide in alkaline conditions.

Trace Cobalt (II) is also reported to catalyse the chemiluminescent reaction in the

presence of the cationic surfactant cetyl-trimethyl ammonium bromide. It should

be noted however, that attempts to consistently reproduce the emission were not

productive.



















Table 3: Fluorescent compounds examined for potential use as reagents


Compound


Observations made


acridine
9- aminoacridine
2-aminophthalate
Fluorescein sodium
1- naphthalene carboxylic acid
Rhodamine S
Rubrene


no measurable signal
no measurable signal
signal obtained
signal obtained
no measurable signal
signal obtained
poorly soluble





































0 100 200 300 400 500 600 700
time (s)




Figure 12: Intensity-time profiles for dicloxacillin in the presence of selected
fluorescers.































0.3
4-
*5


0 100 200 300 400 500 600 700 800 900 1000
time (s)





Figure 13: Intensity-time profiles for selected# lactam compounds in the presence
of fluorescein.









H
R c N S CH3


N COH-
000


-00H


H
A N
0


fluorescer


H
R N.ICN S OH3
0
NN CH3
OOH c-
OOH \


OH


T -


"-OOC


fluorescer'


fluorescer*


fluorescer +


Figure 14: Postulated mechanism for scission of b-lactam ring.








51
A kinetic examination of the p-lactam hydrolysis reaction in the presence

and absence of luminol could provide insight on whether the chemiluminescence

enhancement is associated with beta-lactam ring opening or some other event.

A recent report by Nakashima et al [45] suggests that the I-2O2 stream is not

necessary for the chemiluminescence. The implication being that the penicillin-

peroxide adduct was not essential to the chemiluminescence enhancement. In an

attempt to reproduce their conditions, it was not possible to obtain a

chemiluminescence signal in the absence of the hydrogen peroxide stream. It is

however, possible that the formation of a penicillin hydrogen peroxide adduct may

not necessarily be the species responsible for the enhancement.

Determining whether the penicillins are able to enhance the luminescence

in the presence of differing oxidants may afford insight as to the nature of the

enhancement step. The alternative oxidants in this case would be the non-peroxide

oxidants such as C6, Br2 or sources of the same, namely the chlorates, bromates

or iodates as well as N-chloro- and N-bromo- succinimides. In this respect the

oxidant sodium hypochlorite was employed as a non peroxide oxidant. It was

shown that dicloxacillin did not increase the maximum intensity of emission, but an

enhancement that develops at a kinetically slower pace that was associated with

greater light output (Figure 15). As mentioned earlier there was no

spectrophotometric evidence of the formation of an adduct between the oxidant

and dicloxacillin.



























0.30 IUIyuIIa" pUiUAiUw

"V 0.25- \sodium hypochlorite

N 0.20-

0.15-

0.10-

0.05- ~~

0.00- -- -------------- ---
0 1o00 200 300 400 500 600 700 800 900 1000
time (s)








Figure 15: Intensity-time profile for dicloxacillin-luminol on employing different
oxidants.








53
An observation also made was that the mechanisms proposed by Mereinyi

and Sattar did not address themselves to the first of the two electron loss

proposed for luminol on oxidation. However Di az and Garci a [48] account for the

loss of the second hydrogen and its electron to a metal ion in a single electron

oxidation by the transition metal ion. In the absence of metal ion it is difficult to

account for single electron redox transfers to organic species. Arising from this

omission a reaction mechanism that accounts for the oxidative changes is

proposed (Figure 16). The mechanism proposes the transient formation of the

powerful reductant diimide. The diimide formed then undergoes either self

reduction or reduces the "enhancers". The generation of hydrazine may be

facilitated by the penicillins and other enhancers. The generation of hydrazine in

oxidative reactions been demonstrated by entrapping the hydrazine using stilbene,

or by use of anthracene [49].

Presumably the/p-lactam ring is reductively opened to yield a hypothetical

aldehyde and secondary amine. This mechanistic outline is developed from the

mechanism proposed by Corey et al for the thermal decomposition of sulfonyl

hydrazides [49,50,51].

The transient existence of diimide has been recognized in synthetic organic

chemistry where it is generated in situ. Diimide has been reported to reduce the

carbonyl function of aromatic aldehydes and ketones [51]. House [52] proposes

that the reduction occurs by a mechanism in which diimide serves as a nucleophile

that reversibly adds to the carbonyl group to form the hydroxy diazene.

















NH i NH
I -- + II
INH I NH
CO0-

NH2

NH
II --- N, NH2NH,
NH


Figure 16: Reaction mechanism involving hydrazine generation.








55
The diazene intermediate decomposes with loss of nitrogen to form the alcohol.

The mechanisms lack the equilibria that would account for the emission profile.

Furthermore there is the additional possible role for the mode of action of

enhancers which would possibly include their serving as substrate for diimide

reduction, and driving the reaction to completion by consumption of diimide. It is

assumed that light emission from the luminescent species occurs at a faster rate

than the chemical reactions generating the luminescent species.

The enhancement may also arise from an increase in the efficiency of

chemiluminescence from the excited species. Direct measurement of the efficiency

of chemiluminescence was not possible using available resources, as measured

chemiluminescence is determined by both the yield of the excited aminophthalate

species as well as the competing processes able to deactivate the excited

aminophthalate ion. However from examination of the emission profile and/or the

luminescent decay profile it would be possible to postulate whether the

luminescent species was the free aminophthalate ion per se or an adduct formed

between the penicillins and the anion. The decay profile should remain consistent

if the chemiluminescent species remains unaltered. Changes in total luminous flux

would reflect changes in the yield of the chemiluminescent aminophthalate ion. On

the other hand changes in the decay profile would suggest an altered luminescent

species. The nature of alterations would probably have to be characterized by

obtaining the absorption and/or the emission spectrum of the luminescing

species. The terminal phase of the decay profile obtained in a static system, was








56
characterized in terms of the best exponential fit and compared to that obtained

in the presence of the other enhancers.

The assumption made here was that, the light emission rate is indicative of

a critical rate limiting step in the chemiluminescent reaction, the terminal phase of

the light emission profile was regressed by fitting a first order exponential decay

equation. It is assumed that the terminal decay profile is free from the modifications

arising from continuous mixing in the static system. The exponential fitting was

carried out on profiles obtained for amoxicillin, benzylpenicillin, dicloxacillin and

penicillin V and compared to a control (Fig 17 & Table 4).

The results obtained compare well with the overall rate constant determined

following application of the steady state approximation by White et al [36] of k'=

2.4 X 103 s1 at 35WC. However a single sample t-test on the rate constants did not

reveal a significant difference between the rate constants, this would suggest a

similar if not identical final reaction step and would be consistent with the

observation made by other workers that the catalytic stage was not rate limiting

Sand not the terminal stage.

From examination of the various mechanistic possibilities a number of

reaction schemes were drawn up and tested for suitability. The object being to

develop a model able to at least qualitatively, simulate the chemiluminescent light

emission profile obtained in the presence and absence of the beta-lactam

antibiotics. For this purpose two related reaction simulation programs were

employed.








57









0.20

0.18-
0.16- i ..
0.16- dicloxacillin

0.14- penicillin G
0.12- \ penicillin V
2- phenethicillin
as 0.10-
C control
c. 0.08- amoxicillin
0.06-

0.04.

0.02"

0 500 1000 1500 2000 2500 3000 3500 4000
time (s)





Figure 17: Intensity-time luminescence profiles for selected compounds in luminol
solutions.















Table 4: Kinetic data obtained from Intensity-time profiles


Compound


k (Is)


R squared


amoxicillin 1.1 exp (-3) 0.908
dicloxacillin 0.9 exp (-3) 0.989
benzylpenicillin 0.8 exp (-3) 0.980
phenoxymethyl penicilli 0.8 exp (-3) 0.980
control 0.4 exp (-3) 0.964


All readings were obtained at an ambient temperature of 24.5 C







59
In the initial instance the reaction mechanism and associated rate constants,

described by Sattar and Epstein [40] (Fig 10) was applied to the HICHEM program

found to simulate at least qualitatively the observed results (Figures 18-21).

A major problem with the Sattar & Epstein outline is that the relatively high

rate constants preclude the possibility of simulating an enhancement that would

match the protracted chemiluminescence seen in vitro. It was concluded that the

relatively low rate constants were probably a result of the nature of the pulse

radiolytic techniques. The studies involved generating the radicals by in vitro pulse

radiolysis, where considerations of diffusion limited mixing processes do not apply.

Applying the inverse of the rate constants did permit simulation of light output

within a 3 microsecond time scale. The adjustment did however indicate a critical

feature necessary for any simulation. This is that the oxidation step be the slowest

step in any simulation. Secondly, their model reactions involve radicals, which is

not necessarily the case in the in vitro studies. Note the effect of adding radical

quenchers to the chemiluminescence reaction was not tested. A simpler reaction

mechanism model along the lines of the scheme proposed by White et al was also

drawn and tested on ACUCHEM software. The reaction model involved a divergent

reaction sequence (Fig 22), which was fitted into an ACUCHEM input file (Figure

23).

By varying the exponent term of the rate constant for the initial oxidation

reaction for luminol it was possible to demonstrate an enhancement of not only the

postulated light output but also to obtain profile similar to those seen in vitro.







00- H20
NN
k N -OH
NH2 2 NH2 0

(B) (E)


k3 -OH


COO" dark
N2 + H20 + [ |products
COO (DP)
(D)


Figure 18: Reaction model for HICHEM simulation
























RKB SIMUL OF LUMINOL RXN
1110
1, A+OO=B, 2.3E-08
2, B=E+OH, 2.5E-05
3, B+OH=D+N+HOH, 1.8E-05
4, E+OH=B, 1.0E-09
5, E=DP, 2.0E-03
END
A, 1 E-05
00, 1 E-05
B,0
END
1.OE-30,1.OE-32,1,0.00567
1,2,3,4,5,6,10,
11,12,13,14,15,16,17,18,19,20,
21,22,23,24,25,26,27,28,29,30

INPUT 0.0001
MULTIPLIER 0


Figure 19: Input file for simulations on HICHEM.








62











0.9-
0.8-
0.7-
5 W 0.6-
0
Ss 05-.5
W, 0.4-
S 0.3-

0.2-
0.1-
1E 5 .E-2 .d___ .. -
0 0.0005 0.001 0.0015 0.002 0.0025
time



-- A -*- B -*- -OH
--- D -p- E -A DP





Figure 20: Concentration time profiles for reactants and intermediates








63










9


8-

7-

? i 6-

0




1, o o/ bs oco
F r 2 5" -

2 E0
*c 4-

3-

2-

1: 1
0 0.0005 0.001








Figure 21: Simulated light emission profile


time








0

N


NH20
(A)


--OOH


ki


0 00
NH

NH
NH2)
(C)


Dark products


Ik2


Coo

Scoo-
NH;


+ H20 + N2


Figure 22 : Reaction model for ACUCHEM simulation


k3
No

















;INPUT FILE FOR LUMINOL SIMULATION
BASIC

1111

1, A+B=C, 5E+07
2, C=E+H, 1.8E+05
3, C=D, 2.0E+03
END
A, 1E-04
B,1E-04
END
0.001
3600


Figure 23: Input file for simulations on ACUCHEM.







66
Changes in the rate constant for the second step in the reaction model did not

yield changes in the light emission profile consistent with that observed

experimentally.

It appears that the proposals put forward in the literature [15,41] justifiably

recognized the initial oxidative step as being critical to the chemiluminescent light

output. They suggest that the various enhancers of chemiluminescence facilitate

the oxidation of luminol via formation of enhancer peroxides. The peroxy species

generated in the alkaline medium presumably undergoes nucleophilic attack of the

carbonyl group opening up the strained beta-lactam ring to a highly reactive

peroxy intermediate which is then able to oxidatively transfer the peroxy species

to luminol.








67









9.
8-












0 500 1000 1500 2000 2500 3000 3500 4000
time (s)
7-















... .. k=0.5 -+-- k=5 ^ k=50
6-







0k=0 -- k=200 0 k=500













Figure 24: Simulated concentration-time profiles for 2-aminophthalate generation
for different values for the rate constant I<1
C-2-

1 E2
0 500 1000 1500 2000 2500 3000 3500 46000

time (s)


M--- k=0.5 -+-k=5 -*K- k=50
-6 k=100 -M- k=200 -A, k=500





Figure 24: Simulated concentration-time profiles for 2-aminophthalate generation
for different values for the rate constant k1














1.4-
1.2-
1"

S0.6
.- o.si-^


0.4
0.2
0|11 T r 1^^ ^ei*--.
0 500 1000 1500 2000 2500 3000
time (s)

.. k=0.5 --- k=5 -- k=50
--- k=100 -- k=200 -A- k=500


Figure 25: Simulated light emission profiles for different values of rate constant k1.








69
Static Studies

The possibility of employing uv-visible spectrophotometry as an investigative

tool in the enhancement of luminol chemiluminescence was also examined. Except

for the appearance of a relatively weak shoulder in the region 400-460 nm, it was

not possible to demonstrate a significant change in the ultraviolet-visible spectrum

in the region 196-600 nm on addition of hydrogen peroxide to a luminol containing

solution (Fig 13). The presence of a chromophore perturbed by an asymmetric

center in the penicillins indicated the possibility of employing optical rotatory

dispersion or circular dichroic (ORD/CD) techniques as means of examining the

fate of the betalactam antibiotics during the reaction was also tried. This approach

has been applied by Rasmussen and Higuchi [53], Mitscher et al [54] and others,

in stability studies on the penicillins. However satisfactory experimental results were

difficult to obtain due to the very slow response of the variable wavelength

polarimeter employed. No available equipment is on hand for the current studies.

A problem that arose in applying the chiroptical method is the fact that in

addition to the optical activity exhibited by the penicillins, the hydroperoxide could

also be able to rotate plane-polarized light. This could complicate any interpretation

of optical activity vs time profiles.

From the static studies it was possible to demonstrate that the methanol

content of the analytical streams influences the extent of enhancement. Maximum

enhancement was found to lie between 10-20 % v/v methanol in water (Fig 26).





















0.02 10%
0.00- .--25%
5%
0.01


0.01-


0 500 1000 1500 2000 2500 3000
time (s)




Figure 26: Intensity-time profiles for luminol reaction at differing methanol
concentrations.








71
The possibility that degradation products of penicillin hydrolysis could in fact

be responsible for the enhancement of chemiluminescence was also examined, to

this end both penicillamine and penicillenic acid were found not to enhance luminol

chemiluminescence.

Flow Studies

Flow infection analysis

Preliminary work had indicated that the enhancement has better analytical

potential when applied to flowing streams, because the more consistent mixing

outweighs any loss of chemiluminescence due to the time of transfer to the

measuring cell. Quantitative signals could be obtained using either luminol or

isoluminol solutions at concentrations as low as 1CT5 to 10"6 M, metal ion solutions

in the same concentration range and hydrogen peroxide solutions of concentration

9 X 10U3 M. The optimum pH was obtained using a 104 M solution of sodium

hydroxide. This can be compared to the pH 11.7 employed by Yan [55]. However

due to convenience of preparation, a 13 M sodium hydroxide solution was

employed as solvent for all the analytes (i.e. 1102O, luminol and cobalt/copper). The

choice of metal ion was noted to qualitatively influence the shape of the peak.

Copper ions were noted to give rise to less baseline noise than cobalt ions, at

concentrations in the region of 10 M. In the literature different concentrations of

copper (CuP+) ions have also been reported to yield differing intensity-time profiles

[6]. Yan [55] selected copper as opposed to cobalt, citing as their grounds the

chromatographic characteristics obtained and the better signal to noise ratio as a








72
result of the lower background noise. Where the FIA setup involved more than two

streams, the order of mixing was found to influence the quality of the signal

obtained, better results were obtained for streams mixed in the order alkaline

luminol solution with alkaline hydrogen peroxide prior to merging with the metal ion

stream (Fig 10). This order was consistent with the reasoning that the metal ion

simply catalyzes the emission arising from the reaction of luminol in alkaline

peroxide. The beta-lactam analyte was then injected into the luminol stream prior

to the stream merging with the metal ion stream. Optimal flow rates for the setup

were determined to be approximately 1.25 ml.minrf'1 per channel, giving a total

output of 4.5 ml.mirfn1 flowing through the detector. A 24 ul1 cell permitted

measurements in the millivolt range which was difficult to achieve with the 12 1 I

cell. This was probably due to longer residence time of the chemiluminescent

stream within the larger volume detector.

It was also noted that the metal ion stream was not essential for

chemiluminescence enhancement, and could in fact be excluded altogether, the

effect of which was a reduction in the intensity of light emission associated with

removal of the catalytic influence of the metal ion. The reduction however does not

significantly compromise the enhancement brought about by the beta-lactams. As

a result all subsequent flow injection analyses excluded the metal ion stream. FIA

studies revealed that solvents such as acetonitrile, tetrahydrofuran and butanol

alone or in mixtures with water did not support the chemiluminescence. It was

observed that borate buffers did not support the chemiluminescence whereas








73
phosphate buffers were able to support post-column chemiluminescence detection.

This was found to be in agreement with reports by Nakashima et al [45] who

reported that borate as well as the imidazole buffer did support post-column

chemiluminescence detection. Attempts to employ the imidazole buffer pH 7.3 as

described in the BP 1988 [56] in the mobile phase did not yield a post column

detection system able to support chemiluminescence. However, it is perhaps worth

noting that Nakashima et al did not define the imidazole buffer employed in their

work.

Using a three channel setup consisting of a MeOH:0.001M KHFPO,, a

luminol stream and an alkaline peroxide (I-202) stream (Figure 6) a number of

penicillins, cephalosporins and related compounds were examined. The

chromatogram and subsequent histogram (Figure 27, 28) indicate that on a molar

basis dicloxacillin exhibited the best enhancement of chemiluminescence within the

group examined, clavulanic acid also exhibited a comparable enhancement of

chemiluminescence, then to a lower extent benzylpenicillin, penicillin V,

phenethicillin, 6-aminopenicillanic acid, cephalothin, and piperacillin. A number of

compounds exhibited weak enhancement while others did not afford a measurable

enhancement of chemiluminescence. Of significance was the fact that penicillamine

and the p-lactam ring azetidinone did not enhance the chemiluminescence

significantly.

Analytical parameters for selected penicillins were also obtained using the

flow injection analysis setup (Figure 29 and Table 5).







74

In an attempt to correlate the degree of enhancement to a measure of ring

strain, no correlation was found between the enhancement and the literature infra

red stretching frequencies for the beta-lactam carbonyl group (Figure 30).

The influence of selected dosage form excipients on the chemiluminescent signal

was also examined. The sugars galactose and sorbitol, p hydroxy cyclodextrin and

sodium carboxymethyl cellulose all compromise the chemiluminescence

enhancement and would have to be removed from the analyte solutions for good

results (Table 6).




















0.70


0.60-


0.50-


0.40-


0.30-


0.20-


0.10-


0.00-
0


500 1000 1500 2000
time (s)


200 3000
2500 3000


SFigure 27: FIA chromatogram for enhancement by selected compounds


j 4 U L.I a....J UJ 1..J U


3500

















6 aminopenicillanicacid
ampicillin
azeldinone
cefotaxidne
cephalosporin C
cephalothin




12.00-



10.00-



8.00-


.5,
s 6.00-
S
i-


7 clavulanic aMid
8 didoxacillin
9 penicillamine
10 hetacllin
11 imrnipenem
12 meihicillin
13 penicillin G


Figure 28: Histogram for enhancement by selected compounds.


14 penicillin V
15 phenethicillin
16 piperacillin
17 sulbactam














0.18 -

0.16 A

0.14 -

0.12 -

0.1 -

0.08 -
SA
0.06 -

0.04 -

0.02 -__----
0 -,- --,

0.OOE+00 2.00E-04 4.00E-04 6.00E-04 8.00E-04 1.00E-03 1.20E-03 1.40E-03 1.60E-03

concentration (m/I)


Sphenethicillin



x penV



* cephalothin



A ampicillin



piperacillin


A


penG


Figure 29: FIA calibration plots for selected compounds.












Table 5: Analytical parameters for selected compounds from FIA analysis


compound sensitivity L.O.Q R squared Precision
(V.I/m) (m/I) (s.d)


benzylpenicillin 168.15 8.5E-06 0.986 11.61 (n=5)
piperacillin 15.582 7.6E-05 0.98 3.59 (n=5)
phenethicillin 14.479 8.8E-05 0.986 6.14 (n=5) -,
penicillin V 7.643 0.00011 0.998 4.96 (n=4)
ampicillin 7.098 0.00039 0.998 2.21 (n=4)
cephalolthin 4.725 0.00082 0.97 15.84 (n=4)
hetacillin ns -
cephalosporin C ns -
methicillin ns -






































1640
1620


1680


1720


1760


1660 1700 1740
IR Stretching frequency (/cm)


S1800
1780


Figure 30: Scatter diagram for Infrared stretch frequencies for selected compounds
versus relative enhancement of luminol reaction.


I .rl


I lJW

90-

80-


" &i ,-


1600


dicloxacillin


penicillamine

i


x
x
X %e-4















Table 6: Influence of selected excipients on chemiluminescent
signal enhancement by benzylpenicillin.


Excipient


relative
concentration


% blank
signal


Galactose X 10 80.2
Sorbitol x 10 81.4
sodium carboxymethyl cellulose x 10 80
beta- hydroxy cyclodextrin x 20 0







81
Liquid Chromatoaraphy

The experimental objective was to develop a chromatographic system that would

permit separation and quantitation of a mixture of at least two beta-lactam

antibiotics employing a post-column chemiluminescence detection system.

Mobile phase development was based on the results from both static studies and

the flow injection analyses. Possible application of the enhancement as a post-

column detection technique in the chromatographic analysis of penicillins was also

examined. Chromatographic separation of penicillins is well documented [57,58]

in most cases uv detection has been employed for visualization of the column

eluate. Initial experiments indicated that separation with chemiluminescent detection

is possible using silica columns. However more recently Nakashima et al [47]

applied reverse phase separation on octadecylsilane columns with post-column

chemiluminescence detection. From the separation systems reported in the

literature it was decided to confine the mobile phase choices to mainly aqueous

based mobile phases, simply to preclude compromising the chemiluminescent

reaction by nonaqueous media or heats of solution arising from the mixing of two

or more different streams. For example whereas acetonitrile and tetrahydrofuran

were found to quench the luminescence, methanol generated a significant heat of

solution on mixing with aqueous streams resulting in a significant baseline drift.

Employing a 0.001 M KH2PO4 mobile phase useful chromatograms had been

obtained from which analytical sensitivity was measured and determination of limits

of detection was made for a number of penicillins (Table 5). From the initial







82
experiments silica column based separations with postcolumn chemiluminescence

detection, were found to be feasible, however emphasis was placed on developing

a reverse phase separation system due to the attendant advantage of lower

operating costs. The criteria applied to selecting the appropriate starting mobile

phase were that, the phase be partly or entirely aqueous, permit reverse phase

separation, not contain known quenchers of the chemiluminescence and be easily

adjusted to pH 10 by mixing with post column alkaline streams. A number of the

chromatographic systems described in the literature for the separation of penicillins

were not likely to support chemiluminescence, largely due to the organic modifier

or buffer employed in the chromatographic mobile phase, presumably to ensure

that only one ionized form of the penicillin exists in the system.

Attempts to employ a methanol-water-acetonitrile combination without the modifier

failed to achieve separation of the penicillins with all eluting at the solvent front.

Subsequent experimental runs were based on the separation reported by White

and Zarembo [59] these were initially carried out on an octadecylsilane (ODS)

column using a methanol in 0.01 M sodium dihydrogen phosphate NaN-PO4

(35:65) mobile phase with a uv detector set at 254 nm. This system was able to

separate dicloxacillin from penicillin G but did not support post column

chemiluminescence detection. Lowering the buffer concentration to 0.001 M

permitted post-column chemiluminescence detection, but significantly reduced the

solute retention times. The reduced retention times) compromised the usefulness

of post-column chemiluminescence detection as reasonable difference in retention








83
times is necessary to counter the effects of band broadening arising from the post-

column detection method. The use of 0.001 M potassium dihydrogen phosphate

(KF6PO4) alone as mobile phase for the elution of penicillin G gave rise to broad

peaks on a noisy baseline. The addition of 10% methanol to the 0.001 M KH2PO4

significantly reduced the noise, but did not yield useful peaks. The use of a 35 %

Methanol in 0.001M K2PO4 mobile phase eluted the penicillins but required the

introduction of a delay coil.The delay coil served to improve the extent of mixing

and allow the reaction to develop improving the signal to noise ratio. Using a 7.1

cm mixing loop the chemiluminescence enhancement was not detectable, whereas

lengthening the mixing loop to 47.0 cm afforded a measurable chemiluminescence

enhancement signal. Employing 35 % methanol in 0.001M KH2PO4 as mobile

phase on a reverse phase octadecylsilane (Exsil R ODS) column at 1.0 ml mirfn1,

with reagent streams of alkaline luminol and alkaline hydrogen peroxide each at

0.5 ml.minr1, the retention times for selected betalactam compounds were obtained

(Table 7) and calibration curves obtained for dicloxacillin, penicillin V and penicillin

G (Fig 31-34). Employing the three compounds the precision reported as a

standard deviation was determined (Table 8).

An estimate of the loss of resolution associated with the post-column

chemiluminescence detection was also determined employing penicillin V and

dicloxacillin as model compounds (Table 9). The results exhibit a loss of resolution

of 19.3 % for the system employed.














Table 7: Chromatographic retention times for selected compound


Compound


6-aminopenicillanic acid
amoxicillin
ampicillin
azetidinone
cefotaxine
cephalosporin C
cephalothin
clavulinic acid
d-penicillamine
dicloxacillin
hetacillin
imipenem
methicillin
penicillin G
penicillin V
phenethicillin
piperacillin
sulbactam


retention time
(min)
1.35
ns
ns
ns
1.8
1.2
1.9
1.2
ns
4.9
ns
1.2
2
3
3.1
3.6
1.2
ns


ns : no measurable signal obtained

Chromatographic system


Column
Mobile phase
Reagent streams

Detection system


Exsil R ODS
0.001 M K H2P04 :MeOH (65:35) 1.0 ml/min
alkaline luminol 0.5 ml/min
alkaline hydrogen peroxide 0.5 ml/min
Chemiluminescence

















120 -



100 -


80 -



60 -



40 -



20 -



n -


0 0.0005 0.001


0.0015 0.002
concentration (m/I)


0.0025 0.003 0.0035


Figure 31: Calibration plot for dicloxacillin.


i


0








86







120 x
x


100



80 -

4--
I /x
LM 60-o

a)

Q 40 -



20 X



0 I I -I
0 0.001 0.002 0.003 0.004 0.005
concentration (m/I)


Figure 32: Calibration plot for penicillin V.








87







200 -

180 -

160 -

140 -

E
E 120 -

._ 100 -
80
U 80 -A
a)
CL
0.
60 -

40 -

20 /
20



0 0.001 0.002 0.003 0.004 0.005
concentration (m/I)


SFigure 33: Calibration plot for penicillin G.













195 5970 EOO


I Ii -- -- g





... t__ i ... ; { __ __...



..- ....................... ... ____ ____ -
> 9 ---- __ ^ - ----- ------ --- --- ----- __ __ __ ___ ------- ^ __ ---- ^ - - 0 ----0^
: ] s' p ; :,-- ^ :-^ -.^ ,B_; .. -.r =
:5I.0 _^ ^ ^ ri^j[a:__=


.- *--4-*-


sa- -9'-


- b- -- =
Iwb


Figure 34: Typical chromatogram obtained.


Sjl&J


















Table 8: Analytical parameters for selected compounds.


Compound


X-coefficient R squared
mm.l/m


Precision replicates


Penicillin G 4.3 exp(4) 0.99 5.11 6
Penicillin V 2.5 exp(4) 0.98 24.44 6
Dicloxacillin 4.9 exp(4) 0.98 7.6 6













Table 9: Results obtained from resolution measurements


uv spectrophotometric detection

Compound retention
time (min)

penicillin V 3

dicloxacillin 5.55


Calculated resolution

chemiluminescence detection

Compound r
ti


peak
width (min)

0.8

0.9


Rs = 3.0


detention
me (min)


peak width
min


penicillin V 3.1 1.15

dicloxacillin 5.7 1.0


Calculated resolution


Rs = 2.4186












CHAPTER 5
CONCLUSIONS






In addition to the B-lactam antibiotics penicillin G, penicillin V, piperacillin and

cephalothin, a broader range of B- lactam ring containing compounds are able to

enhance the chemiluminescence exhibited by luminol. It appears that a "strained"

B-lactam ring is essential for the enhancement. However the presence of a p-

lactam ring alone does not insure enhancement of luminescence as other

structural features in the molecule appear to be able to modulate this enhancement

of luminol chemiluminescence.

The different penicillins enhance the chemiluminescence to varying degrees

exhibiting differing intensity profiles. Of the p- lactam compounds examined

dicloxacillin and clavulinate exhibit the most profound enhancement of

luminescence, more so than penicillin G and the phenoxyalkyl penicillins penicillin

V and phenethicillin. Cephalothin, methicillin, aminopenicillanic acid, piperacillin and

sulbactam exhibit much less enhancement. Ampicillin and its derivative hetacillin

on the other hand do not enhance the chemiluminescence as does the isolated fl-

lactam ring azetidinone. These differences are presumed to arise from differences

in the accessibility of the nucleophilic oxidant to the electron deficient carbon of the




Full Text
53
An observation also made was that the mechanisms proposed by Mereinyi
and Sattar did not address themselves to the first of the two electron loss
proposed for luminol on oxidation. However Diaz and Garca [48] account for the
loss of the second hydrogen and its electron to a metal ion in a single electron
oxidation by the transition metal ion. In the absence of metal ion it is difficult to
account for single electron redox transfers to organic species. Arising from this
omission a reaction mechanism that accounts for the oxidative changes is
proposed (Figure 16). The mechanism proposes the transient formation of the
powerful reductant diimide. The diimide formed then undergoes either self
reduction or reduces the "enhancers". The generation of hydrazine may be
facilitated by the penicillins and other enhancers. The generation of hydrazine in
oxidative reactions been demonstrated by entrapping the hydrazine using stilbene,
or by use of anthracene [49].
Presumably the 0-lactam ring is reductively opened to yield a hypothetical
aldehyde and secondary amine. This mechanistic outline is developed from the
mechanism proposed by Corey et al for the thermal decomposition of sulfonyl
hydrazides [49,50,51].
The transient existence of diimide has been recognized in synthetic organic
chemistry where it is generated in situ. Diimide has been reported to reduce the
carbonyl function of aromatic aldehydes and ketones [51]. House [52] proposes
that the reduction occurs by a mechanism in which diimide serves as a nucleophile
that reversibly adds to the carbonyl group to form the hydroxy diazene.


Table 9: Results obtained from resolution measurements
uv spectrophotometric detection
Compound retention peak
time (min) width (min)
penicillin V 3 0.8
dicloxacillin 5.55 0.9
Calculated resolution Rs = 3.0
chemiluminescence detection
Compound retention peak width
time (min)min
penicillin V 3.1 1.15
dicloxacillin 5.7 1.0
Calculated resolution
Rs = 2.4186


CHAPTER 4
RESULTS AND DISCUSSION
Mechanistic studies
Mechanistic studies were carried out with the objective of being able to
propose a reaction model that would satisfactorily simulate the light emission
profile obtained on addition of the p- lactams to the reagents. The simulations were
performed using the chemical kinetic software HICHEM [34] and the more recent
improvement ACUCHEM [34,35]. Much of the work in elucidating the mechanism
of luminol chemiluminescence was pioneered by White et al [36] and expanded
upon by Lind, Mernyi and Eriksen [37,38,39]. In most of the reaction
mechanisms radicals, ions and radical ions are proposed as being the reactive
species in the various steps preceding formation of the luminescent species.
White et al [36] proposed two consecutive loses of protons to generate a
di-anion which is then able to undergo oxidative decomposition generating the
excited aminophthalate ion, which chemiluminesces as it returns to the ground
state (Figure 9). To this series of consecutive reactions, White et al applied the
steady state approximation. Making among others the assumption that k, (HgO)
> > kg (02) White et al derived the rate expression for photon emission.
38


Intensity (V)
48
Figure 12: Intensity-time profiles for dicloxacillin in the prescence of selected
fluorescers.


Protocol for comparing different antibiotics 29
Flow studies 30
Flow Injection Analysis 30
Chromatographic studies 30
Protocol for Calibration Curve determinations 35
Equipment 35
Software 36
Source of water 36
4. RESULTS AND DISCUSSION 38
Mechanistic Studies 38
Static Studies 69
Flow Studies 71
Flow Injection analysis 71
Liquid Chromatography 81
5. CONCLUSIONS 91
APPENDIX 94
REFERENCE LIST 97
BIOGRAPHICAL SKETCH 101
iv


13
bromine based oxidative systems [17]. Examples of these analyses and some of
the analytical parameters reported are listed in Table 1.
Table 1: Analyses in which luminescence has been applied.
Analyte
Reagent
linear range
(ug/ml)
Detection
limit (ug/ml)
Ref
Oxytetracycline
hexacyanoferrate
1-10
0.5
17
Quinine
Ce(IV) sulfite
5-500
0.64
8
Penicillin G
lum-perox-Co(ll)
0.0001 -0.01
6 exp(-5)
14
Promethazine
Ce(IV) sulfite
10.3-51.3
-
11
Hydrocortisone
Ce(IV) sulfite
0.1 -1.0
0.016
10
Betamethasone
Ce(IV) sulfite
O
CJ1
CJ1
b
0.3
10
Isoniazid
bromosuxinimide
0.05 20.0
0.024
12
Acetaldehyde
#lum-XO-Fe(CN)
exp(-7) exp(-3)
4 exp(-7)
9
Tetracycline
hexacyanoferrate
0.1 -1.2
0.04
17
Doxycycline
hexacyanoferrate
0.1 -10.0
0.2
17
Chlortetracycline
hexacyanoferrate
1.0-10
0.1
17
* static measurement system
# lum-XO-Fe(CN) luminol-xanthine oxidase-hexacyanoferrate


26
Preparation of Calibration series.
Stock solutions of the p- lactam compounds were prepared in deionized
water prior to each analyses. Appropriate volumes of the solutions were then
diluted with water to obtain the desired concentration ranges. 20 /x I aliquots were
injected into the chromatographic system. Mobile phase methanol in 0.001 M
sodium phosphate (35 % MeOH) pH 6.3.
Ultraviolet Soectroohotometric studies
About 15 mg dicloxacillin was accurately weighed out into a 10 ml volumetric
flask and dissolved in a small quantity of 0.001 M sodium hydroxide before being
made to volume. A 2 ml aliquot of this stock solution was then diluted to 10 ml in
a volumetric flask to given the experimental solution. A 2.0 ml aliquot of the
experimental solution was transferred to a cuvette and used for the
spectrophotometric studies.
Employing a Shimadzu UV 160U instrument the ultraviolet spectrum in the
wavelength range 250 360 nm for the dicloxacillin sample was obtained against
0.001 M sodium hydroxide as reference and thereafter following addition of 10,50,
80 and 100 pI of a 30 % HgOj solution. A spectrum was also obtained for a 100 /I
aliquot of HjOj diluted in 2 ml of the 0.001 M sodium hydroxide solution.


Table 7: Chromatographic retention times for selected compound
Compound retention time
(min)
6-aminopenicillanic acid
1.35
amoxicillin
ns
ampicillin
ns
azetidinone
ns
cefotaxine
1.8
cephalosporin C
1.2
cephalothin
1.9
clavulinic acid
1.2
d-penicillamine
ns
dicloxacillin
4.9
hetacillin
ns
imipenem
1.2
methicillin
2
penicillin G
3
penicillin V
3.1
phenethicillin
3.6
piperacillin
1.2
sulbactam
ns
ns : no measurable signal obtained
Chromatographic system
Column Exsil R ODS
Mobile phase 0.001 M K H2P04 :MeOH (65:35)
Reagent streams alkaline luminol
alkaline hydrogen peroxide
Detection system Chemiluminescence
1.0 ml/min
0.5 ml/min
0.5 ml/min


61
RKB SIMUL OF LUMINOL RXN
1110
1, A+00 = B, 2.3E-08
2, B = E+OH, 2.5E-05
3, B+OH = D+N + HOH, 1.8E-05
4, E+OH = B, 1.0E-09
5, E = DP, 2.0E-03
END
A.1E-05
OO.1E-05
B,0
END
1.0E-30.1.0E-32,1,0.00567
I,2,3,4,5,6,10,
II,12,13,14,15,16,17,18,19,20,
21,22,23,24,25,26,27,28,29,30
INPUT 0.0001
MULTIPLIER 0
Figure 19: Input file for simulations on HICHEM.


58
Table 4: Kinetic data obtained from Intensity-time profiles
Compound k (Is) R squared
amoxicillin
1.1 exp (-3)
0.908
dicloxacillin
0.9 exp (-3)
0.989
benzylpenicillin
0.8 exp (-3)
0.980
phenoxymethyl penicilli 0.8 exp (-3)
0.980
control
0.4 exp (-3)
0.964
All readings were obtained at an ambient temperature of 24.5 C


81
Liquid Chromatography
The experimental objective was to develop a chromatographic system that would
permit separation and quantitation of a mixture of at least two beta-lactam
antibiotics employing a post-column chemiluminescence detection system.
Mobile phase development was based on the results from both static studies and
the flow injection analyses. Possible application of the enhancement as a post
column detection technique in the chromatographic analysis of penicillins was also
examined. Chromatographic separation of penicillins is well documented [57,58]
in most cases uv detection has been employed for visualization of the column
eluate. Initial experiments indicated that separation with chemiluminescent detection
is possible using silica columns. However more recently Nakashima et al [47]
applied reverse phase separation on octadecylsilane columns with post-column
chemiluminescence detection. From the separation systems reported in the
literature it was decided to confine the mobile phase choices to mainly aqueous
based mobile phases, simply to preclude compromising the chemiluminescent
reaction by nonaqueous media or heats of solution arising from the mixing of two
or more different streams. For example whereas acetonitrile and tetrahydrofuran
were found to quench the luminescence, methanol generated a significant heat of
solution on mixing with aqueous streams resulting in a significant baseline drift.
Employing a 0.001 M Kh^PO* mobile phase useful chromatograms had been
obtained from which analytical sensitivity was measured and determination of limits
of detection was made for a number of penicillins (Table 5). From the initial


65
¡INPUT FILE FOR LUMINOL SIMULATION
BASIC
1111
1, A+B=C, 5E+07
2, C = E+H, 1.8E+05
3, C = D, 2.0E+03
END
A, 1E-04
B.1E-04
END
0.001
3600
Figure 23: Input file for simulations on ACUCHEM.


82
experiments silica column based separations with postcolumn chemiluminescence
detection, were found to be feasible, however emphasis was placed on developing
a reverse phase seperation system due to the attendant advantage of lower
operating costs. The criteria applied to selecting the appropriate starting mobile
phase were that, the phase be partly or entirely aqueous, permit reverse phase
separation, not contain known quenchers of the chemiluminescence and be easily
adjusted to pH 10 by mixing with post column alkaline streams. A number of the
chromatographic systems described in the literature for the separation of penicillins
were not likely to support chemiluminescence, largely due to the organic modifier
or buffer employed in the chromatographic mobile phase, presumably to ensure
that only one ionized form of the penicillin exists in the system.
Attempts to employ a methanol-water-acetonitrile combination without the modifier
failed to achieve separation of the penicillins with all eluting at the solvent front.
Subsequent experimental runs were based on the separation reported by White
and Zarembo [59] these were initially carried out on an octadecylsilane (ODS)
column using a methanol in 0.01 M sodium dihydrogen phosphate NaHgPQ,
(35:65) mobile phase with a uv detector set at 254 nm. This sytem was able to
seperate dicloxacillin from penicillin G but did not support post column
chemiluminescence detection. Lowering the buffer concentration to 0.001 M
permitted post-column chemiluminescence detection, but significantly reduced the
solute retention times. The reduced retention time(s) compromised the usefulness
of post-column chemiluminescence detection as reasonable difference in retention


41
More recently Sattar and Epstein [40] outlined an overall reaction scheme
for the reaction of luminol with basic hydrogen peroxide from pulse radiolysis
studies carried out by Mernyi and coworkers [37,38,39] (Figure 10).
The precise role of a number of compounds that sensitize chemiluminescent
reactions has not been fully elucidated. Various suggestions as to the role of the
sensitizing agents have ranged from being direct catalytic activators to acting via
complexation with the catalytic metal ions (usually Co2*, Cu2+) or some reaction
intermediates [41]. Some workers in the field [15,41] have suggested that the
enhancement mechanism involves the formation of an enhancer-peroxide adduct
which is then able to more rapidly generate the endoperoxide from luminol. The
endoperoxide then decomposes into nitrogen and the excited aminopthalate ion
which then serves as the luminescent species.
The concept of enhanced oxidative capability has been cited in a more
recent study of the enhancement of luminol chemiluminescence by nitric oxide.
Kikuchi et al [42] report that enhancement is due to the formation of peroxynitrite
from NO and HgOj. They demonstrated the formation of the peroxynitrite (ONOO)
anion, a strong oxidizing species, by demonstrating the existence of its specific
maximum at 302 nm in the uv spectrum. As regards the possible role of the
betalactam antibiotics in enhancing the chemiluminescence, Chen et al [15]
proposed the formation of an adduct between penicillin and the superoxide anion.
Employing difference spectrophotometry, they demonstrated the existence of a
chromophore with a maximum between 250-260 nm. This adduct was presumed


46
of the pure luminol, fluorescein or any other fluorescent enhancer would help
ascertain this aspect of the mechanism. It was also shown that the more efficient
fluorescers rhodamine and fluorescein generated more intense light output profiles
than the aminophthalate ion (Figure 12). The intensity time emission profiles for the
oxidation of selected penicillins in the prescence of fluorescein were also obtained
(Figure 13).
These observations led to the development of a peroxylate type
chemiluminescence mechanism, in which the opening of the p- lactam ring is the
primary energy releasing reaction. Subsequent events are as outlined in Figure 14.
Though opening of the ^-lactam ring was the more obvious source of energy, it
is recognised that oxidation at other sites on the antibiotics molecules could serve
as energy sites. The suggested mechanism is supported in part by the recent
observations by Chen et al [46,47] that selected xanthone dyes exhibited
chemiluminescence on oxidation by hydrogen peroxide in alkaline conditions.
Trace Cobalt (II) is also reported to catalyse the chemiluminescent reaction in the
prescence of the cationic surfactant cetyl-trimethyl ammonium bromide. It should
be noted however, that attempts to consistently reproduce the emission were not
productive.


17
as well as to the ease with which they undergo base hydrolysis.
A wide range of techniques have been used to develop methods for the
analysis of penicillins and /or cephalosporins.
Microbiological methods have found limited usefulness in the quantitative
determination of penicillins. Their main use has been in qualitative determinations,
confirming microbial sensitivity to the compounds and in determining
therapeutically effective levels.
Titrimetric methods have been the mainstay of p- lactam analyses, these
have typically involved degrading the penicillin, by use of acid and/or base
solutions, to penicillamine or some sulfhydryl containing molecule which is then
analyzed by a mercurimetric or iodometric titration. The methods are limited by the
fact that they determine "total penicillins" necessitating a blank to correct for the
presence of degradation product(s) [28,29].
Colorimetric methods have similarly involved degrading the intact 0- lactam
to generate a chromophore with characterizable uv absorption spectrum. This has
formed the basis for the hydroxamate, methylene blue, dinitrobenzoic acid and the
more recent mercurimetric assay methods [30,31,32,33].
Optical rotatory dispersion (ORD) and circular dichroism (CD) techniques
have found limited application to the analysis of penicillins. They have been mainly
applied to kinetic determinations of the compounds.
Infrared techniques have not found use in the quantitative analysis of p-
lactams but rather have found use as diagnostic tools in qualitative determinations.


55
The diazene intermediate decomposes with loss of nitrogen to form the alcohol.
The mechanisms lack the equilibria that would account for the emission profile.
Furthermore there is the additional possible role for the mode of action of
enhancers which would possibly include their serving as substrate for diimide
reduction, and driving the reaction to completion by consumption of diimide. It is
assumed that light emission from the luminescent species occurs at a faster rate
than the chemical reactions generating the luminescent species.
The enhancement may also arise from an increase in the efficiency of
chemiluminescence from the excited species. Direct measurement of the efficiency
of chemiluminescence was not possible using available resources, as measured
chemiluminescence is determined by both the yield of the excited aminophthalate
species as well as the competing processes able to deactivate the excited
aminophthalate ion. However from examination of the emission profile and/or the
luminescent decay profile it would be possible to postulate whether the
luminescent species was the free aminophthalate ion per se or an adduct formed
between the penicillins and the anion. The decay profile should remain consistent
if the chemiluminescent species remains unaltered. Changes in total luminous flux
would reflect changes in the yield of the chemiluminescent aminophthalate ion. On
the other hand changes in the decay profile would suggest an altered luminescent
species. The nature of alterations would probably have to be characterized by
obtaining the absorption and/or the emission spectrum of the luminescing
species. The terminal phase of the decay profile obtained in a static system, was


79
to
c
0)
>
j
100
90-
80-
70-
60-
50
40-
30-
20-
10-
0
penicillamine
I
dicloxacillin
x
X X wX*,
1600
~r
1640 1680 1720 1760
1620 1660 1700 1740 1780
IR Stretching frequency (/cm)
1800
\
Figure 30: Scatter diagram for Infrared stretch frequencies for selected compounds
versus relative enhancement of luminol reaction.


21
accuracy, precision, repeatability and reproducibility of the method (s).


89
Table 8: Analytical parameters for selected compounds.
Compound X-coefficient R squared Precision replicates
mm.l/m
Penicillin G
4.3 exp(4)
0.99
5.11
6
Penicillin V
2.5 exp(4)
0.98
24.44
6
Dicloxacillin
4.9 exp(4)
0.98
7.6
6
\


87
concentration (m/I)
Figure 33: Calibration plot for penicillin G.


Intensity (a.u)
(Times 10E-5)
62
A
B
-OH
B D
E
-A- DP
Figure 20: Concentration time profiles for reactants and intermediates


Intensity (V)
phenethicillin
concentration (m/I)
X
pen V
cephalothin
ampicillin
piperacillin
pen G
~vj
>i
Figure 29: FIA calibration plots for selected compounds.


O'
O'
NH
I
NH
O
Figure 16: Reaction mechanism involving hydrazine generation.


Table 6: Influence of selected excipients on chemiluminescent
signal enhancement by benzylpenicillin.
Excipient relative % blank
concentration signal
Galactose X 10 80.2
Sorbitol x 10 81.4
sodium carboxymethyl cellulose x 10 80
beta- hydroxy cyclodextrin x 20 0


99
35). W. Braun, J.T. Herron and D.K. Kahaner, Int J. Chem. Kin. 1988, 20, 51.
36). E.H. White, O. Zafiriou, H.H. Kgi and J.H.M. Hill, J. Am Chem Soc 1964.
86, 940.
37). J. Lind, G. Mernyi and T. Eriksen, J. Am. Chem. Soc. 1983, 105, 7655.
38). G. Mernyi, J. Lind and T. E. Eriksen, J. Phys. Chem. 1984, 88, 2230.
39). G. Mernyi and J.S. Lind, J. Am. Chem. Soc. 1980, 102, 5830.
40). S. Sattar and I.R. Epstein, J. Phys. Chem. 1990, 94, 275.
41). T.G. Burdo and W.R. Seitz, Anal. Chem. 1975, 47, 1639.
42). K. Kikuchi, T Nagano, H. Hayakawa, Y. Hirata and M. Hirobe, Anal. Chem.
1993, 65, 1794.
43). J.H. Martinez, P.J. Martinez, P.Gutierrez and L.l. Martinez, Talanta 1992,
32, 637.
44). A.G. Mohan
Peroxylate chemiluminscence
in Chemi- and Bioluminescence (Ed J.G. Burr)
Marcel Dekker Inc, New York, 1985
45). K. Nakashima, S. Kawaguchi, S. Akiyama and S.G. Schulman, Biomed.
Chromat. 1993, 7, 217.
46). G.N.Chen J.P.Duan and Q.F. Hu, Anal. Chim. Acta 1994, 292, 159.
47). G.N. Chen J.P. Duan and Q.F. Hu, Mikrochim. Acta 1994, 116, 227.
48). A.N. Diaz and J. A. G. Garcia, Anal. Chem. 1994, 66, 988.
49). E.J. Corey, W.L. Mock and D.J. Pasto, Tet. Lett. 1961, 11, 353.
50). E.J. Corey, W.L. Mock and D.J. Pasto, J. Am. Chem. Soc. 1961, 83, 2957.
51). E.J. Corey, W.L. Mock and D.J. Pasto, J. Am. Chem. Soc. 1962, 84, 685.


37
Table 2: Antibiotic compounds screened for enhancement
Compound
Source
2- azetidinone
6- aminopeniciilanic acid
amoxicillin
ampicillin
benzylpenicillin
cefotaxine
cephalosporin C
cephalothin Na
clavulinic acid
d- benzylpenicillenic acid
d- penicillamine
dicloxacillin
hetacillin
imipenem
methicillin
phenethicillin
phenoxymethylpenicillin
piperacilin
sulbactam
imipenem
Aldrich
Sigma
Sigma
ACS Dobfar
Sigma
Sigma
Sigma
Eli Lilly Labs
Bristol
Sigma
Sigma
Sigma
Beecham Research Labs
Merck Sharp & Dohme
Bristol Labs
Sigma
Sigma
Pfizer
Fluka
Merck & Co


27
Static studies.
The experimental setup for the studies is schematically illustrated in Figure
5. The reagent solution was prepared by diluting a 10 pi aliquot of 30 % hydrogen
peroxide to 10 ml in a volumetric flask using a 1CT4 molar stock solution of 2-
aminophthalhydrazide in 0.001 M sodium hydroxide. For experimental runs, 20 n I
of the dicloxacillin was added to 2 ml of the reagent solution in the cuvette.
Earlier studies involved the use of cobalt and copper ion solutions at a
concentration 104 M. They were however found unnecessary in preliminary
investigations. The metal ions employed were cobalt (II) and copper (II).
An alternate procedure adopted was to inject a specified n I volume of 30%
H,02 into the cuvette containing the luminol solution, prior to injection of the
antibiotic solution. Following injection of the antibiotic solution, the Intensity-time
profiles obtained were recorded on a disk over a specified time period.
Conductivity measurements
To a beaker containing 25 ml of a 4 X 14 M stock solution of luminol. was
added 25 p I of the 30 % The changes in conductivity with time were then
measured using a conductivity electrode (Markkson Instruments).


10
reported for determining quinine from the ability of quinine to enhance the
chemiluminescence exhibited by the oxidation of sulphite by cerium(IV) [8].
Similarly the determination of acetaldehyde by monitoring the chemiluminescence
emission from the luminol hexacyanoferrate (III) reaction in the presence of
xanthine oxidase has been described [9]. The steroidal hormones hydrocortisone
and betamethasone as well as the antihistamine promethazine have also been
determined from their ability to enhance the chemiluminescence of the cerium (IV)-
sulphite system [10,11]. The chemiluminescent oxidation of the antimycobacterial
isoniazid by N-bromosuccinimide has also been applied to the determination of
isoniazid [12]. A number of penicillins and cephalosporins have also been reported
to enhance the chemiluminescence exhibited by luminol [13,14,15]. Schulman et
al [14] Chen etal [15] have reported the enhancement luminol chemiluminescence
exhibited by several penicillins, and the cephalosporin cephalothin.
The instrumentation employed for chemiluminescence measurements
basically consists of a mixing device and a detection system. Three approaches
have been used to measure the intensity of emitted light.
The first approach involves the use of a static measurement system in
which the mixing of reagents is performed in a vessel held in front of the detector.
The chemiluminescent reagent is added to the analyte in a cuvette held in a dark
enclosure, and the intensity of the light emitted is measured through an adjacent
photomultiplier tube. In this static system the mixing is induced by the force of the
injection. This simple reagent addition system without a mixing device is of limited


35
Protocol for Calibration Curve determinations
A 1 X104 M solution of luminol made alkaline in 0.008 M sodium hydroxide
was employed as one of the reagent streams. The alkaline hydrogen peroxide
stream was prepared by diluting 1 ml of a 30% v/v stock solution to 500 ml in a
flask using 1 X 103 M sodium hydroxide. The two solutions were employed as
reagent streams in a two channel flow injection system as illustrated in Figure 6.
A 20 n I aliquot of the analyte solution in the respective calibration series was
injected into a stream of the mobile phase.
Equipment
1). FL-750 HPLC plus spectrofluorescence detector
McPherson Instrument Acton, Massachusetts.
2). Pharmacia LKB HPLC pump 2150
LKB-Produkter Bromma, Sweden.
3). LDC Analytical Constametric metering pump
LDC Analytical Riviera beach, Florida.
4). Servogor 120 recorder
Norma Goerz Instruments Osterreich, Austria
5). IBM PC XT
6). DT 2811 Analog and Digital Input/Output board
Data Translation Inc Malbora, Massachusetts.


3
[2], This light emitting species is generated by an oxidative system that is usually
composed of either sodium or potassium hydroxide, hydrogen peroxide as the
oxidizing agent and an activator. The activator is commonly a transition metal ion,
metal-ion complex, hypochlorite, ferricyanide or perchlorate. The solvent system
has frequently been aqueous. However, lower alcohols as well as other water
miscible organic solvents such as dimethyl sulfoxide or dimethylformamide have
also been used in place of water [3]. The activating agent is not absolutely
necessary as in its place sonic waves have been used and in some aprotic media
no activating agent is necessary, only oxygen and a strong base.
In aprotic media the products from luminol oxidation have been isolated.
However, in aqueous systems the corresponding diacidic anion, despite being
quantitatively produced from the hydrazides in basic reaction media, has been
difficult to detect due to apparent further oxidation of the initially formed product.
Reaction mechanism
Basic solutions of hydrogen peroxide are common reagents in hydrazide
chemiluminescence. In these solutions the hydroperoxide ion which stems from the
initial reduction of oxygen is believed to be a critical reactant. The hydroperoxide
ion formed then reacts with the azaquinone form of luminol, by attacking one of
the carbonyl groups, leading to the formation of III. Compound III is then believed
to decompose by one of a variety of postulated routes to the phthaloyl peroxide


66
Changes in the rate constant for the second step in the reaction model did not
yield changes in the light emission profile consistent with that observed
experimentally.
It appears that the proposals put forward in the literature [15,41] justifiably
recognized the initial oxidative step as being critical to the chemiluminescent light
output. They suggest that the various enhancers of chemiluminescence facilitate
the oxidation of luminol via formation of enhancer peroxides. The peroxy species
generated in the alkaline medium presumably undergoes nucleophilic attack of the
carbonyl group opening up the strained beta-lactam ring to a highly reactive
peroxy intermediate which is then able to oxidatively transfer the peroxy species
to luminol.


peroxide
stream
pump 1
metal ion
stream
pump 2
luminol
stream
pump 3
Figure 6: Three channel flow system
analyte
photocell
waste


*
peroxide(
stream
pump 1
luminol
stream
pump 2
Figure 7: Two channel flow system
analyte
photocell
waste


o
o
o
Figure 9: Reaction scheme after White etal [36].


83
times is necessary to counter the effects of band broadening arising from the post
column detection method. The use of 0.001 M potassium dihydrogen phosphate
(KHgPQ,) alone as mobile phase for the elution of penicillin G gave rise to broad
peaks on a noisy baseline. The addition of 10% methanol to the 0.001 M Kf-^PC^
significantly reduced the noise, but did not yield useful peaks. The use of a 35 %
Methanol in 0.001 M Kh^P04 mobile phase eluted the penicillins but required the
introduction of a delay coil.The delay coil served to improve the extent of mixing
and allow the reaction to develop improving the signal to noise ratio. Using a 7.1
cm mixing loop the chemiluminescence enhancement was not detectable, whereas
lengthening the mixing loop to 47.0 cm afforded a measurable chemiluminescence
enhancement signal. Employing 35 % methanol in 0.001 M KI-^PQ, as mobile
phase on a reverse phase octadecylsilane (Exsil R ODS) column at 1.0 ml min1,
with reagent streams of alkaline luminol and alkaline hydrogen peroxide each at
0.5 ml.min1, the retention times for selected betalactam compounds were obtained
(Table 7) and calibration curves obtained for dicloxacillin, penicillin V and penicillin
G (Fig 31-34). Employing the three compounds the precision reported as a
standard deviation was determined (Table 8).
An estimate of the loss of resolution associated with the post-column
chemiluminescence detection was also detemined employing penicillin V and
dicloxacillin as model compounds (Table 9). The results exhibit a loss of resolution
of 19.3 % for the system employed.


76
1
6 aminopenicillanic acid
7 clavulanic acid
14
penicillin V
2
ampicillin
8 dicloxacillin
15
phenethicillin
3
azetidinone
9 penicillamine
16
piperacillin
4
cefotaxine
10 hetacillin
17
sulbactam
5
cephalosporin C
11 imipenem
6
cephalothin
12 methicillin
13 penicillin G
Figure 28: Histogram for enhancement by selected compounds.


59
In the initial instance the reaction mechanism and associated rate constants,
described by Sattar and Epstein [40] (Fig 10) was applied to the HICHEM program
found to simulate at least qualitatively the observed results (Figures 18-21).
A major problem with the Sattar & Epstein outline is that the relatively high
rate constants preclude the possibility of simulating an enhancement that would
match the protracted chemiluminescence seen in vitro. It was concluded that the
relatively low rate constants were probably a result of the nature of the pulse
radiolytic techniques. The studies involved generating the radicals by in vitro pulse
radiolysis, where considerations of diffusion limited mixing processes do not apply.
Applying the inverse of the rate constants did permit simulation of light output
within a 3 microsecond time scale. The adjustment did however indicate a critical
feature necessary for any simulation. This is that the oxidation step be the slowest
step in any simulation. Secondly, their model reactions involve radicals, which is
not necessarily the case in the in vitro studies. Note the effect of adding radical
quenchers to the chemiluminescence reaction was not tested. A simpler reaction
mechanism model along the lines of the scheme proposed by White et al was also
drawn and tested on ACUCHEM software. The reaction model involved a divergent
reaction sequence (Fig 22), which was fitted into an ACUCHEM input file (Figure
23).
By varying the exponent term of the rate constant for the initial oxidation
reaction for luminol it was possible to demonstrate an enhancement of not only the
postulated light output but also to obtain profile similar to those seen in vitro.


93
The enhancement can be applied to the flow injection analyses of the 6-
lactam antibiotic or as a post column detection technique in liquid chromatographic
analyses of penicillin mixtures whose retention times differ by more than 1 min.
Largely as a consequence of the cheaper and often more snesitive methods
available the technique has limited utility in the analysis of the compounds as raw
materials or in dosage forms. However, the technique has potential utility as a
means of directly quantitating the intact fused beta-lactam ring. This is particularly
so in the analysis of biological fluids where some of the other methods may be
compromised by the prescence of interferents such as amino acids that may
resemble some of the degradation products of ring opening.


23
Figure 4--continued.


36
7). Selectro Mark Analyzer
Markson Science Inc DelMar, Colorado.
Software
1). Spectracalc
Galctica Industries Corporation Salem New Hampshire.
2). Acuchem series 4.0
40 species version
NIST, Gaithersburg Maryland.
Source of Water
All water employed in the experiments was purified by filtration through a
millipore Milli-Q 50 water filtration apparatus Millipore Corporation (Bedford, MA)
prior to use.


CHAPTER 1
INTRODUCTION
The term chemiluminescence has generally referred to the luminescent
phenomena associated with a variety of chemical reactions. Chemiluminescence
can be more specifically described as the electromagnetic emission that arises
from the exothermic oxidation of an organic compound. Generally the exothermic
oxidation of the organic compound yields an energy rich product that is
luminescent because the molecule is either rigid or so small, that it is unable to
quickly dissipate internally the energy of the exothermic reaction.
Chemiluminescence is distinguished from the more efficient bioluminescence by
the fact that in bioluminescence visible light is produced from an enzymically
controlled reaction involving the chemical components of a living system. For a
molecule to exhibit chemiluminescence it must be able to form an electronically
excited species through a chemical reaction at ordinary temperatures. Other
conditions described by White and Roswell [1] are that the necessary energy must
be made available in a single reaction step and the molecule receiving this energy
must have a limited number of accessible vibrational energy states which would
otherwise act as an energy sink. If the excited state is emissive, it can
l


ACKNOWLEDGMENTS
I would like to express my sincere gratitude to Dr. John Perrin for serving
as the chairman of my supervisory commitee as well as providing guidance and
advice over the last four years.
I would also like to thank my supervisory commitee members, Dr. Stephen
Schulman, Dr. Margaret James, Dr. Guenther Hochhaus and Dr. Vaneica Young
for their support and criticism of the material presented.
Further, I wish to thank the Department of Medicinal Chemistry for providing
the opportunity, Dr. K. Sloan for providing advice, my fellow students for tolerating
me.
Finally I would like to thank my whole family for their support and
encouragement.
ii


92
p- lactam ring. No correlation was found between chemiluminescent enhancement
and ring strain as measured by the position of the p- lactam carbonyl IR stretching
frequency. The mechanism of enhancement does not necessarily involve formation
of a peroxy intermediate III, evidence for this obtained from the observation that
non- peroxide oxidant hypochlorite is also able to support the 6- lactam
enhancement of luminol chemiluminescence.
The enhancement may involve the oxidation of the 6-lactam ring, with
transfer of the energy generated by the hypothetical ring scission, to a suitable
receptor that is able to re-emit the energy as light. Under more drastic conditions
a weak emission is demonstrable from fluorescein. However there is need to obtain
emission spectra in order to characterize the emitting species.
It was also noted that the metal ion stream was not essential for
chemiluminescence enhancement, and could in fact could be excluded altogether.
The advantage of here being the potential reduction of analytical costs.
In application of the technique to LC analyses the mobile phase
combinations are limited to alcohol-water mixtures (primarily methanol-water) of not
more than 50% alcohol. The enhancement is not supported by acetonitrile or
tetrahydrofuran and is compromised by high (>0.01 M) concentrations of
phosphate or borate buffer.
The loss in resolution associated with use of the technique as a post-column
detection technique is of the order of 20 %.


LIST OF TABLES
Table page
1. Analysis in which luminecence has been applied 13
2. Antibiotic compounds screened for enhancement 37
3. Fluorescent compounds examined for potential use as reagents ... 47
4. Kinetic data obtained from Intensity-time profiles 58
5. Analytical parameters for selected penicillins obtained from flow
injection analyses 78
6. Influence of selected excipients on chemiluminescent signal 80
7. Chromatographic retention times for selected betalactam compounds
employing post-column chemiluminescence detection 84
8. Analytical parameters for selected compounds 89
9. Results from resolution measurements 90
v


34
Chromatographic columns employed in the studies include
Exsil R ODS Keystone scientific Inc
Dimensions
150 X 4.6 mm
Particle size
Pore size
7nm
100 A
Nucleosil R C18 Keystone scientific Inc
Dimensions
150 X 4.6 mm
particle size
pore size
5/iin
100 A
Microsorb Si Rainin Instrumant Company Inc
Dimensions
250 X 4.6 mm
particle size
pore size
5 |im
100 A
For the most part the mobile phase consisted of
methanol: 0.001 M sodium/potassium phosphate NaH,P04 (35:65)
Flow rate 1.0 ml.min1.
Detection system
post-column chemiluminescence.
- 0.001 M luminol in 0.008 M NaOH 0.5 ml.min1
- HgOg in 0.008 M NaOH 0.5 ml.min1
ultraviolet at 254 nm
NB; All metal (stainless steel) tubing was of 0.010 inches internal diameter.


70
Figure 26: Intensity-time
concentrations.
Profiles for luminol
reaction
* differin9 methanol


71
The possibility that degradation products of penicillin hydrolysis could in fact
be responsible for the enhancement of chemiluminescence was also examined, to
this end both penicillamine and penicillenic acid were found not to enhance luminol
chemiluminescence.
Flow Studies
Flow injection analysis
Preliminary work had indicated that the enhancement has better analytical
potential when applied to flowing streams, because the more consistent mixing
outweighs any loss of chemiluminescence due to the time of transfer to the
measuring cell. Quantitative signals could be obtained using either luminol or
isoluminol solutions at concentrations as low as 1CT5 to 10"6 M, metal ion solutions
in the same concentration range and hydrogen peroxide solutions of concentration
9 X 1CJ3 M. The optimum pH was obtained using a 1CT4 M solution of sodium
hydroxide. This can be compared to the pH 11.7 employed by Yan [55]. However
due to convenience of preparation, a Id3 M sodium hydroxide solution was
employed as solvent for all the analytes (i.e. H^, luminol and cobalt/copper). The
choice of metal ion was noted to qualitatively influence the shape of the peak.
Copper ions were noted to give rise to less baseline noise than cobalt ions, at
concentrations in the region of 16 M. In the literature different concentrations of
copper (CU2+) ions have also been reported to yield differing intensity-time profiles
[6], Yan [55] selected copper as opposed to cobalt, citing as their grounds the
chromatographic characteristics obtained and the better signal to noise ratio as a


29
Investigation of possible Peroxvlate Based Mechanism-
Protocol for comparing different fluorescers
Approximately 1G4 M solutions were prepared by dissolving the appropriate
quantity of fluorescer (Fluorescein, Rhodamine S, aminoacridine or Rubrene) into
a 1CT3 M sodium hydroxide solution. A 2 ml aliquot was then transferred into a 3
ml cuvette placed into the static setup illustrated in figure 4 and to this was added
a 100 ai I mixture of 1-1,0, (30%) and dicloxacillin (approx 1 mg/ml) (50:50). The
emission obtained was recorded on disk over a 1000 s time period using the A-D
board and Spectracalc software.
Protocol for comparing different antibiotics
Approximately 104 M solution of fluorescein was prepared by dissolving
about 38 mg of fluorescein sodium (anhydrous M.W 376.3) accurately weighed
into a 103 M sodium hydroxide solution. A 2 ml aliquot was then pipetted into a 3
ml cuvette, placed into the static setup (Figure 4) and a 100 mI mixture of H,02
(30%) and the appropriate p- lactam antibiotic (approx 1 mg/ml) (50:50) was
added.
The emission obtained was recorded on disk over a 1000 s time period using the
A-D board and Spectracalc software.


100
52). H.O. House
Modern Synthetic Reactions
S?d Edition
The Benjamin/Cummings Publishing Company, San Francisco, 1972.
53). C.E. Rasmussen and T. Higuchi, J. Pharm. Sci. 1971, 60, 1608.
54). L.A. Mitscher, P.W. Howison and T. Sokoloski, J. Antibiot. 1974, 27, 215.
55). B. Yan, Anal. Chim. Acta. 1991, 250, 145.
56). British Pharmacopea
H.M.S.O London, U.K., 1988.
57). Analytical Profiles of Drug Substances (Ed K. Florey)
Vols 15 & 17
Academic Press New York 1986, 1988.
58). J.J. Kirschbaum and A. Aszalos
Chapter 7
Modern Analysis of Antibiotics (Ed A. Aszalos)
Marcel Dekker Inc, New York, 1986.
59). E.R. White and J.E. Zarembo, J. Antibiot. 1981, 34, 836.


25
Preparation of Luminol solutions.
Solutions of 0.001 M luminol were prepared by dissolving 88.6 mg 3-
Aminophthalhydrazide into a mixture of about 25 ml water and 2 ml of 2M sodium
hydroxide with stirring, then adjusting to volume in a 500 ml volumetric flask. The
solutions were employed for 48 hours prior to discarding.
Preparation of Hvdrooen Peroxide solutions.
The hydrogen peroxide solutions were prepared by pipetting 2.0 mis of a
30% hydrogen peroxide solution into a 500 ml volumetric flask containing 2.0 ml
of 2M sodium hydroxide in about 25 ml water, then made to volume with water.
Preparation of Stock Sodium Dihvdroaen Phosphate solution.
A Stock solution of monobasic sodium phosphate (Nal-^P04) 0.01 M was
prepared by dissolving 690 mg of the anhydrous salt into 500 ml water. From the
stock solution 50 ml aliquots were diluted to 500 ml to obtain the 0.001 M
solutions.
Preparation of Mobile phase.
To 35.0 ml methanol in a measuring cylinder was added the 0.001 M
sodium dihydrogen phosphate solution to obtain a final volume of 100 ml. The
solution were filtered through a 0.45 /xm membrane filter and degassed by stirring
the mobile phase under vacuum/suction.


13. Intensity-time profiles for selected penicillins in the presence of
fluorescein 49
14. Postulated mechanism for scission of p- lactam ring 50
15. Comparison of intensity-time profiles for dicloxacillin-luminol on
employing different oxidants 52
16. Reaction mechanism involving hydrazine generation 54
17. Luminesence-time profiles for selected compounds in luminol
solutions 57
18. Reaction model for HICHEM simulation 60
19. Input file for simulations on HICHEM 61
20. Concentration-time profiles for HICHEM simulation reactants and
intermediates 62
21. Simulated light emission profile 63
22. Reaction model for ACUCHEM simulation 64
23. Input file for simulations on ACUCHEM 65
24. Simluated concentration-time profiles for 2-aminophthalate generation
for different values for the rate constant k, 67
25. Simulated light emission profiles for different values of rate constant
k, 68
26. Intensity time profile for luminol reaction at different methanol
concentrations 70
27. FIA chromatogram for enhancement by selected compounds 75
28. Histogram for enhancement by selected compounds 76
29. FIA calibration plots for selected compounds 77
30. Scatter diagram for Infrared stretch frequencies for selected
compounds versus relative enhancement of luminol reaction 79
vii


Intensity (V)
52
Figure 15: Intensity-time profile for dicloxacillin-luminol on employing different
oxidants.


96
Given that the concentration of A at any time t is given by the pseudo first order
relation
Assuming that at t=0, [C] = 0. Integrating with respect to time equation two would
yield the equation,
ic'-Jq
hence the rate of production of D which in turn reflects the light flux is given by
the equation
&D= le-k^-e-k't)
t k'-k,
NB A quantum yield of unity is assumed.


30
Flow studies
Flow Injection Analyses.
The set up employed is illustrated in Figures 6 and 7. A similar set up was
employed for chromatographic studies. In the two channnel system the flowing
streams were alkaline peroxide into which the analyte was injected and alkaline
luminol.
For the three channel system flowing streams were alkaline peroxide, alkaline
luminol and water/phosphate/water methanol mixture. In both cases samples were
injected through a Rheodyne loop (Cotati California) injector fixed with a 20/* I loop.
Chromatographic studies
The mobile phase primarily consisted of methanol-water mixtures of varying
proportions, with or without modifiers. The modifiers used were either phosphate
or imidazole solutions. In the screening experiments varying proportions of water
and other water miscible solvents e.g tetrahydrofuran, acetonitrile, were employed.
The flow rates were nominally 1 ml.min1, for the analytical stream and 0.5
ml.min1 for the reagent streams. When a flow rate change was necessary an effort
was made to retain this relative ratio to minimize band broadening arising from
post-column mixing.
A schematic representation of the setup is as illustrated in Figure 8.


47
Table 3: Fluorescent compounds examined for potential use as reagents
Compound
Observations made
acridine
9- aminoacridine
2-aminophthalate
Fluorescein sodium
1- naphthalene carboxylic acid
Rhodamine S
Rubrene
no measurable signal
no measurable signal
signal obtained
signal obtained
no measurable signal
signal obtained
poorly soluble


12
important considerations. Mixing is reported as being most effective at a T piece
or Y junction; however, some workers have used the conventional FIA system in
which sample is injected into the surrounding flowing reagent to achieve mixing.
This approach is reproducible but reported as not producing rapid mixing [5]. In
flow measurement systems the chemiluminescent signal has to be measured
during the mixing, as a result only a section as opposed to the whole-intensity-time
curve is measured. An additional feature of these systems is that the shape of the
curve is now dependent on both the kinetics of the chemiluminescent reaction and
the parameters of the flow system. Chemical variables such as reagent
concentrations, pH, etc and physical parameters such as flow rate, reaction coil
length, sample size and the limitations of experimental apparatus all affect the
performance of flow injection procedures. The effects of these variables on the
observed analytical signal are not necessarily independent, as interactions can and
indeed do occur [16]. Most of the chemiluminescence determinations reported
have involved the determination of compounds able to quench the luminescence
of luminol or other chemiluminescent systems. A smaller number have been
reported for substances enhancing the chemiluminescent signal. A variety of
determinations fall into the latter group of compounds able to sensitize the
chemiluminescent reactions. The chemiluminescent systems have not necessarily
involved the luminol-peroxide system alone, but have included among others the
Cerium(IV)-sulfite system in which the luminescence arises from the oxidation of
sulphite by cerium(IV), hexacyanoferrate (III), the peroxodisulphate system and the


CHAPTER 3
EXPERIMENTAL
Reagents
Fluorescein sodium and Rubrene were purchased from Fluka Chemical Co
(Buchs, Switzerland), Aminoacridine from Sigma Chemical Co (St Louis, MO), and
Rhodamine S from K & K Laboratories (Plainview, NY). All other reagents were
supplied by Fischer Scientific Co (Fairlawn, NJ) The antibiotics were obtained from
a variety of sources. Azetidinone from Aldrich Chemical Co (Milwaukee, Wl),
Ampicillin (ACS DOBFAR) from Interchem Corporation (Paramus, NJ), Hetacillin
was provided by Beecham Research Laboratories (Syracruse, NY), N- formimidoyl
thienamycin (Imipenem) was a gift from Merck & Co (Rahway, NJ), Methicillin and
Phenethicillin from Bristol Laboratories (Syracuse, NY), Lithium clavulinate a gift
from Smith Kline Beecham Pharmaceuticals (Philadelphia, PA), Sulbactam from
Pfizer Cephalothin from Eli Lilly Labs (Indianapolis, IN). The other antibiotics were
either purchased from Sigma Chemical Co (St Louis, MO) or Aldrich Chemical Co
(Milwaukee, Wl).
All chemicals and reagents were used as provided without further
purification.
24


63
Figure 21: Simulated light emission profile


Intensity (V)
57
O 500 1000 1500 2000 2500 3000 3500
time (s)
4000
Figure 17: Intensity-time luminescence profiles for selected compounds in luminol
solutions.


43
to lengthen the lifetime of the oxidant. The penicillins however have been reported
by Martinez ef a/ [43] to form metal ion complexes with cobalt whose spectra may
differ under various conditions or circumstances. Furthermore given that the
luminescence is known to be enhanced without the metal ion the postulated
adduct is not necessarily responsible for the enhancement alone. The ultraviolet
visible spectrum obtained on addition of l-^02 to dicloxacillin solutions, in the
absence of a metal ion, did not demonstrate the existence of a unique
chromophore (Figure 11). The demonstrated existence of an adduct alone does
not suggest that the adduct is responsible for the enhancement, it is also important
to show that the adduct does indeed enhance the luminescence. In the case of
nitric oxide, the peroxynitrite anion is an established oxidant, and demonstrating
a stronger oxidative potential for the adduct would lend credence to the argument.
Furthermore, it was not possible to demonstrate the existence of a uv light
absorbing adduct on the addition of the oxidizing agent sodium hypochlorite.
The oxidant activity could be demonstrated by redox potential measurement(s),
possibly using a platinum electrode against a silver-silver chloride or other suitable
reference. Suitable equipment was not available to conduct the necessary
experiments.
A second possible mode of luminescence enhancement was that the
enhancer was simply able to donate the energy arising from its oxidation to 2-
aminophthalate which then re-emits the energy as light, similar to the
chemiluminescence exhibited by the peroxyoxalates. The possibility that the


98
18). W.Spendley, G.B. Hext and F.R. Himsworth, Technometrics 1962, 4, 441.
19). J.A. Nelder and R. Mead, Comput. J. 1964, 7, 308.
20). P.B. Ryan, R. L. Barr and H. D. Todd, Anal. Chem. 1980, 52, 1460.
21). E.R. Aberg and A. G. T. Gustavsson, Anal Chim Acta 1982, 144, 39.
22). A. Gustawson and J. E. Sundkvist, Anal. Chim. Acta 1985, 167, 1.
23). P. Hedlund and A. Gustavsson, Anal. Chim. Acta 1992, 259, 243.
24) A.P. Wade, P.M. Shiundu and P.D. Wentzell, Anal. Chim. Acta. 1990, 237,
361.
25). G. N. Chen, Anal Chim Acta 1990, 236, 495.
26). D. Betteridge, T. J. Sly, A.P. Wade and D.G. Porter, Anal. Chem. 1986, 58,
2258.
27). R.M. Sweet and L.F. Dahl, J. Am. Chem. Soc. 1970, 90, 549.
28). D.H. Sieh
Analytical Profiles of Drug substances (Ed K. Florey)
Vol 17, p677
Academic Press Inc, New York 1988.
29). K.Kalberg and U. Forsman, Anal. Chim. Acta 1976, 83, 309.
30). Practical Pharmaceutical Chemistry (Eds A.H. Beckett J.B. Stenlake)
4th Edition, Part 1, p293
Athlone Press London 1984
31). Y.A. Beltagy, S.M. Rida and A. Issa, Pharmazie 1974, 29, 1974.
Through CA 87:189506
32). P. Vermeij, Pharm. Weekbl. Sci. Ed. 1979, 1, 217.
Through CA 90: 197304
33). H. Buundgaard and K. liner, J. Pharm. Pharmac. 1972, 24, 790.
34). R.L. Brown, NBSIR 81-2281
National Bureau of Standards, Gaithersburg, MD, 1981.


Table 5: Analytical parameters for selected compounds from FIA analysis
compound
sensitivity
(V.l/m)
L.O.Q
(m/I)
R squared
Precision
(s.d)
benzylpenicillin
168.15
8.5E-06
0.986
11.61 (n=5)
piperacillin
15.582
7.6E-05
0.98
3.59 (n=5)
phenethicillin
14.479
8.8E-05
0.986
6.14 (n=5)
penicillin V
7.643
0.00011
0.998
4.96 (n=4)
ampicillin
7.098
0.00039
0.998
2.21 (n=4)
cephalolthin
4.725
0.00082
0.97
15.84 (n=4)
hetacillin
ns
-
-
-
cephalosporin C
ns
-
-
-
methicillin
ns


-


I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
J
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
HA
Margaret O. James
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Guenther Hochhaus
Associate Professor of Pharmaceutics
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
f.A 1^4
Vaneica Y. Young /f
Associate Professorof Che^


20
experiments since then have confirmed this, but have also demonstrated that a
number of penicillins do not enhance the chemiluminescence under the given
experimental conditions.
A systematic survey of the penicillins on the basis of chemical structure
should enable the prediction of which p-lactams enhance chemiluminescence. For
this purpose the penicillins can be divided into the following chemical groups, from
which an available antibiotic can be examined for the ability to enhance the
chemiluminescence.
penicillin G
phenoxyalkyl penicillins
isoxazolyl penicillins
ampicillins and related compounds
N-acylated ampicillins
cephalosporins
cephamycins
The extent(s) of enhancement will be compared in terms of the analytical
parameters such as the linear dynamic range, sensitivity, limit of quantitation,
precision, etc. Due to the poor reproducibility associated with variations in mixing
a flowing analytical stream was employed.
Finally an attempt is made to correlate the results obtained to the
reactivities/stability of the respective beta-lactam antibiotics and the application of
the method for determination of the p-lactam antibiotics in dosage forms, indicating


k2
(D)
Figure 18 : Reaction model for HICHEM simulation


Intensity (V)
8
Figure 3: Typical luminescence profile


86
concentration (m/I)
Figure 32: Calibration plot for penicillin V.


39
rate = k [III] [OH] [02] (1)
where III is the monoanionic luminol base and
Jc- klk2
k.x [H20}
A pseudo-first order rate constant of 2.5 X 10'3 s'1 at 35 C was reported for the
monoanionic luminol base.
Mernyi and coworkers [37,38,39] in a series of publications also outlined
a possible mechanism for the chemiluminescent reaction of luminol. In the earliest
of the three papers [37], they describe the hydroperoxide as being the first critical
intermediate in the chemiluminescent pathway of luminol and suggest that the
hydroperoxide decomposes to the chemiluminescent aminophthalate ion at pH
values above its pKj (pKg = 10.4 0.1), whereas at pH values below its pKa the
hydroperoxide decomposes back to oxygen and its parent hydrazide. The paper
did not propose a mechanism for the formation of the hydroperoxide or the nature
of the oxidation stage.
In a later publication Mernyi et al [38] proposed a pathway for the
oxidation of the one-electron-oxidized luminol by molecular oxygen reporting a
one-electron reduction potential of 0.240 0.02 V vs NHE for 5-aminophthalazine-
1,4-dione. However little indication is provided as to how the one-electron-oxidized
luminol is generated as well as an indication of the levels of molecular oxygen
and/or 02~* in aqueous alkaline solutions.


Intensity
75
Figure 27: FIA chromatogram for enhancement by selected compounds


azetidinone
6- aminopcnicillanic
acid
sulhactam
thienamycin
clavulanic acid
X
Benzyl penicillin
ampicillin
amoxicillin
X=H Y = H
X=NH;,V=H
X=NH,. v = oh
methicillin
Figure 4: Chemicals structures of compounds examined.


''
k2
O)
cc
nh2
coo
coo'
(D)
+ h2o + n2
Figure 22 : Reaction model for ACUCHEM simulation



PAGE 1

$33/,&$7,21 2) /80,12/ &+(0,/80,1(6&(1&( 72 7+( $1$/<6,6 2) 7+( %(7$ /$&7$0 $17,%,27,&6 %\ -2+1 + 0,<$:$ $ ',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 OLNH WR H[SUHVV P\ VLQFHUH JUDWLWXGH WR 'U -RKQ 3HUULQ IRU VHUYLQJ DV WKH FKDLUPDQ RI P\ VXSHUYLVRU\ FRPPLWHH DV ZHOO DV SURYLGLQJ JXLGDQFH DQG DGYLFH RYHU WKH ODVW IRXU \HDUV ZRXOG DOVR OLNH WR WKDQN P\ VXSHUYLVRU\ FRPPLWHH PHPEHUV 'U 6WHSKHQ 6FKXOPDQ 'U 0DUJDUHW -DPHV 'U *XHQWKHU +RFKKDXV DQG 'U 9DQHLFD
PAGE 3

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

PAGE 4

3URWRFRO IRU FRPSDULQJ GLIIHUHQW DQWLELRWLFV )ORZ VWXGLHV )ORZ ,QMHFWLRQ $QDO\VLV &KURPDWRJUDSKLF VWXGLHV 3URWRFRO IRU &DOLEUDWLRQ &XUYH GHWHUPLQDWLRQV (TXLSPHQW 6RIWZDUH 6RXUFH RI ZDWHU 5(68/76 $1' ',6&866,21 0HFKDQLVWLF 6WXGLHV 6WDWLF 6WXGLHV )ORZ 6WXGLHV )ORZ ,QMHFWLRQ DQDO\VLV /LTXLG &KURPDWRJUDSK\ &21&/86,216 $33(1',; 5()(5(1&( /,67 %,2*5$3+,&$/ 6.(7&+ LY

PAGE 5

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

PAGE 6

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

PAGE 7

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

PAGE 8

&DOLEUDWLRQ SORW IRU GLFOR[DFLOOLQ &DOLEUDWLRQ SORW IRU SHQLFLOOLQ 9 &DOLEUDWLRQ FKDUW IRU SHQLFLOOLQ 5HSUHVHQWDWLYH FKURPDWRJUDP IRU FDOLEUDWLRQ PHDVXUHPHQWV YLLL

PAGE 9

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f RI HQKDQFHPHQW DQG WKH DQDO\WLFDO XWLOLW\ RI WKH HQKDQFHPHQW 7KH UHVXOWV REWDLQHG VXJJHVW WKH SRVVLELOLW\ RI D SHUR[\ODWHOLNH HQKDQFHPHQW PHFKDQLVP LQ ZKLFK FOHDYDJH RI WKH EHWDODFWDP ULQJ LV UHVSRQVLEOH IRU WKH HQHUJ\ UHOHDVH PDQLIHVWLQJ DV DQ HQKDQFHPHQW RI FKHPLOXPLQHVFHQFH 7KH HQKDQFHPHQW LV DSSOLHG WR OLTXLG FKURPDWRJUDSK\ +3/&f DV D SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ PHWKRG 7KH XWLOLW\ RI WKH SRVWFROXPQ GHWHFWLRQ WHFKQLTXH LV FRPSDUHG WR XOWUDYLROHW GHWHFWLRQ QPf LQ WHUPV RI WZR SHDN VHSDUDWLRQ FULWHULD ,;

PAGE 10

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

PAGE 11

FKHPLOXPLQHVFH GLUHFWO\ KRZHYHU LW FDQ DOVR WUDQVIHU WKH HQHUJ\ WR DQRWKHU PROHFXOH ZKLFK IROORZLQJ H[FLWDWLRQ WKHQ HPLWV WKH HQHUJ\ DV OLJKW 7KH OXPLQHVFHQFH H[KLELWHG E\ OXPLQRO ZDV RQH RI WKH HDUOLHVW FDVHV RI FKHPLOXPLQHVFHQFH WR EH VWXGLHG /XPLQRO DPLQRGLK\GURSKWKDOD]LQH GLRQHf LV DQ DPLQRSKWKDOLF K\GUD]LGH )LJ f WKDW LV DEOH WR H[LVW LQ VHYHUDO WDXWRPHULF IRUPV )LJXUH /XPLQRO DPLQRSKWKDOLF DFLG K\GUD]LGHf )RU OXPLQRO WKH OLJKW HPLWWLQJ VSHFLHV LV WKH H[FLWHG DPLQRSKWKDODWH LRQ ZKLFK HPLWV D EOXH HPLVVLRQ LQ ZDWHU DQG D \HOORZ JUHHQ HPLVVLRQ LQ GLPHWK\O VXOIR[LGH

PAGE 12

>@ 7KLV OLJKW HPLWWLQJ VSHFLHV LV JHQHUDWHG E\ DQ R[LGDWLYH V\VWHP WKDW LV XVXDOO\ FRPSRVHG RI HLWKHU VRGLXP RU SRWDVVLXP K\GUR[LGH K\GURJHQ SHUR[LGH DV WKH R[LGL]LQJ DJHQW DQG DQ DFWLYDWRU 7KH DFWLYDWRU LV FRPPRQO\ D WUDQVLWLRQ PHWDO LRQ PHWDOLRQ FRPSOH[ K\SRFKORULWH IHUULF\DQLGH RU SHUFKORUDWH 7KH VROYHQW V\VWHP KDV IUHTXHQWO\ EHHQ DTXHRXV +RZHYHU ORZHU DOFRKROV DV ZHOO DV RWKHU ZDWHU PLVFLEOH RUJDQLF VROYHQWV VXFK DV GLPHWK\O VXOIR[LGH RU GLPHWK\OIRUPDPLGH KDYH DOVR EHHQ XVHG LQ SODFH RI ZDWHU >@ 7KH DFWLYDWLQJ DJHQW LV QRW DEVROXWHO\ QHFHVVDU\ DV LQ LWV SODFH VRQLF ZDYHV KDYH EHHQ XVHG DQG LQ VRPH DSURWLF PHGLD QR DFWLYDWLQJ DJHQW LV QHFHVVDU\ RQO\ R[\JHQ DQG D VWURQJ EDVH ,Q DSURWLF PHGLD WKH SURGXFWV IURP OXPLQRO R[LGDWLRQ KDYH EHHQ LVRODWHG +RZHYHU LQ DTXHRXV V\VWHPV WKH FRUUHVSRQGLQJ GLDFLGLF DQLRQ GHVSLWH EHLQJ TXDQWLWDWLYHO\ SURGXFHG IURP WKH K\GUD]LGHV LQ EDVLF UHDFWLRQ PHGLD KDV EHHQ GLIILFXOW WR GHWHFW GXH WR DSSDUHQW IXUWKHU R[LGDWLRQ RI WKH LQLWLDOO\ IRUPHG SURGXFW 5HDFWLRQ PHFKDQLVP %DVLF VROXWLRQV RI K\GURJHQ SHUR[LGH DUH FRPPRQ UHDJHQWV LQ K\GUD]LGH FKHPLOXPLQHVFHQFH ,Q WKHVH VROXWLRQV WKH K\GURSHUR[LGH LRQ ZKLFK VWHPV IURP WKH LQLWLDO UHGXFWLRQ RI R[\JHQ LV EHOLHYHG WR EH D FULWLFDO UHDFWDQW 7KH K\GURSHUR[LGH LRQ IRUPHG WKHQ UHDFWV ZLWK WKH D]DTXLQRQH IRUP RI OXPLQRO E\ DWWDFNLQJ RQH RI WKH FDUERQ\O JURXSV OHDGLQJ WR WKH IRUPDWLRQ RI ,,, &RPSRXQG ,,, LV WKHQ EHOLHYHG WR GHFRPSRVH E\ RQH RI D YDULHW\ RI SRVWXODWHG URXWHV WR WKH SKWKDOR\O SHUR[LGH

PAGE 13

GHULYDWLYH 9, ZKLFK WKHQ XQGHUJRHV ULQJ RSHQLQJ WR \LHOG WKH H[FLWHG DPLQR SKWKDODWH LRQ DV LOOXVWUDWHG LQ )LJ 7KH PHFKDQLVP RI GHFRPSRVLWLRQ RI WKH D]DTXLQRQH KDV EHHQ WKH VXEMHFW RI D QXPEHU RI SRVWXODWLRQV ZKLFK KDYH EHHQ UHYLHZHG E\ :KLWH DQG 5RVZHOO >@ $SSOLFDWLRQ RI /XPLQRO &KHPLOXPLQHVFHQFH WR DQDO\VLV LQ VROXWLRQ 7KH LQWHQVLW\ RI FKHPLOXPLQHVFHQFH DULVLQJ IURP WKH FKHPLOXPLQHVFHQW UHDFWLRQ FDQ EH GHVFULEHG E\ WKH IROORZLQJ HTXDWLRQ A&/ EF/ ‘ G& GW ZKHUH r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

PAGE 14

)LJXUH 6FKHPDWLF RXWOLQH RI FKHPLOXPLQHVFHQW UHDFWLRQ SDWKZD\

PAGE 15

7KH LQWHQVLW\ RI OXPLQHVFHQFH FDQ EH XVHG DV WKH EDVLV IRU GHWHUPLQDWLRQ RI DQ\ VSHFLHV ZKRVH FRQFHQWUDWLRQ LQIOXHQFHV WKH UDWH RU HIILFLHQF\ RI WKH FKHPLOXPLQHVFHQW UHDFWLRQ ,Q DSSO\LQJ FKHPLOXPLQHVFHQFH WR DQDO\VLV RI WKH VSHFLHV WKH UHDFWLRQ FRQGLWLRQV VKRXOG EH DGMXVWHG VR WKDW WKH DQDO\WH RI LQWHUHVW LV WKH OLPLWLQJ UHDJHQW LQ WKH V\VWHP DQG DOO RWKHU UHDFWDQWV DUH LQ H[FHVV 7R REWDLQ SUHFLVH PHDVXUHPHQWV WKH FKHPLOXPLQHVFHQFH UHDFWLRQ VKRXOG EH LQLWLDWHG LQ D FRQWUROOHG DQG UHSURGXFLEOH PDQQHU ODUJHO\ EHFDXVH WKH HPLVVLRQ LQWHQVLW\ YDULHV ZLWK WLPH DV WKH UHDFWDQWV DUH FRQVXPHG &KHPLOXPLQHVFHQW DQDO\VHV DUH UHSRUWHG WR KDYH VHYHUDO DGYDQWDJHV ZKLFK LQFOXGH JRRG VHQVLWLYLW\ D ZLGH OLQHDU G\QDPLF UDQJH ORZ GHWHFWLRQ OLPLWV IHPWRPROH WR DWWRPROH UDQJHf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f 7KH LQLWLDO SDUW RI WKH FXUYH LV LQIOXHQFHG E\ WKH PHWKRG RI PL[LQJ HPSOR\HG ZKLOH WKH JHQHUDO VKDSH RI WKH FXUYH GHSHQGV XSRQ WKH NLQHWLFV RI WKH UHDFWLRQ DV ZHOO DV XSRQ DQ\ FKDQJHV LQ TXDQWXP \LHOG ZLWK WLPH 0RVW FKHPLOXPLQHVFHQFH UHDFWLRQV DUH UHSRUWHG WR KDYH ORZ HIILFLHQFLHV OHVV WKDQ b >@ ZKLFK KDV UHVWULFWHG WKHLU XVHIXOQHVV WR DQDO\VHV 7KH GXUDWLRQ RI WKH

PAGE 16

UHDFWLRQV LV LQIOXHQFHG E\ WKH UHDFWLRQ FRQGLWLRQV DQG PD\ RFFXU UDSLGO\ ZLWKLQ RQH VHF RU ODVW ORQJHU WKDQ KRXUV ,Q WKH GHYHORSPHQW RI FKHPLOXPLQHVFHQW DVVD\ PHWKRGV WKH WZR EDVLF IDFWRUV WKDW LQIOXHQFH WKH LQWHQVLW\ RI FKHPLOXPLQHVFHQFH LH HIILFLHQF\ DQG UDWHf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

PAGE 17

,QWHQVLW\ 9f )LJXUH 7\SLFDO OXPLQHVFHQFH SURILOH

PAGE 18

K\GURJHQ SHUR[LGH VR IRUPHG WKHQ EHLQJ DVVD\HG E\ WKH OXPLQRO UHDFWLRQ *OXFRVH r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f &REDOW&FIf DQG 1LFNHO1"f QRQPHWDOOLF LQRUJDQLFV VXFK DV WKH JDVHV R[\JHQ DQG QLWURJHQ GLR[LGH WKH KDOLGH LRQV DQG D QXPEHU RI QLWUR DPLQR RU K\GUR[\ JURXS FRQWDLQLQJ RUJDQLFV 7KHVH VXEVWDQFHV DUH UHSRUWHG WR KDYH HLWKHU DQ H[FLWDWRU\ RU LQKLELWRU\ LQIOXHQFH RQ WKH FKHPLOXPLQHVFHQFH H[KLELWHG E\ OXPLQRO 0RUH UHFHQWO\ D QXPEHU RI GUXJ VXEVWDQFHV KDYH EHHQ UHSRUWHG WR HQKDQFH D YDULHW\ RI RWKHUZLVH FKHPLOXPLQHVFHQW UHDFWLRQV )RU H[DPSOH D PHWKRG KDV EHHQ

PAGE 19

UHSRUWHG IRU GHWHUPLQLQJ TXLQLQH IURP WKH DELOLW\ RI TXLQLQH WR HQKDQFH WKH FKHPLOXPLQHVFHQFH H[KLELWHG E\ WKH R[LGDWLRQ RI VXOSKLWH E\ FHULXP,9f >@ 6LPLODUO\ WKH GHWHUPLQDWLRQ RI DFHWDOGHK\GH E\ PRQLWRULQJ WKH FKHPLOXPLQHVFHQFH HPLVVLRQ IURP WKH OXPLQRO KH[DF\DQRIHUUDWH ,,,f UHDFWLRQ LQ WKH SUHVHQFH RI [DQWKLQH R[LGDVH KDV EHHQ GHVFULEHG >@ 7KH VWHURLGDO KRUPRQHV K\GURFRUWLVRQH DQG EHWDPHWKDVRQH DV ZHOO DV WKH DQWLKLVWDPLQH SURPHWKD]LQH KDYH DOVR EHHQ GHWHUPLQHG IURP WKHLU DELOLW\ WR HQKDQFH WKH FKHPLOXPLQHVFHQFH RI WKH FHULXP ,9f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

PAGE 20

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

PAGE 21

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fVXOILWH V\VWHP LQ ZKLFK WKH OXPLQHVFHQFH DULVHV IURP WKH R[LGDWLRQ RI VXOSKLWH E\ FHULXP,9f KH[DF\DQRIHUUDWH ,,,f WKH SHUR[RGLVXOSKDWH V\VWHP DQG WKH

PAGE 22

EURPLQH EDVHG R[LGDWLYH V\VWHPV >@ ([DPSOHV RI WKHVH DQDO\VHV DQG VRPH RI WKH DQDO\WLFDO SDUDPHWHUV UHSRUWHG DUH OLVWHG LQ 7DEOH 7DEOH $QDO\VHV LQ ZKLFK OXPLQHVFHQFH KDV EHHQ DSSOLHG $QDO\WH 5HDJHQW OLQHDU UDQJH XJPOf 'HWHFWLRQ OLPLW XJPOf 5HI 2[\WHWUDF\FOLQH KH[DF\DQRIHUUDWH 4XLQLQH &H,9f VXOILWH 3HQLFLOOLQ fOXPSHUR[&ROOf H[Sf 3URPHWKD]LQH &H,9f VXOILWH +\GURFRUWLVRQH &H,9f VXOILWH %HWDPHWKDVRQH &H,9f VXOILWH 2 &&E ,VRQLD]LG EURPRVX[LQLPLGH $FHWDOGHK\GH OXP;2)H&1f H[Sf H[Sf H[Sf 7HWUDF\FOLQH KH[DF\DQRIHUUDWH 'R[\F\FOLQH KH[DF\DQRIHUUDWH &KORUWHWUDF\FOLQH KH[DF\DQRIHUUDWH r VWDWLF PHDVXUHPHQW V\VWHP OXP;2)H&1f OXPLQRO[DQWKLQH R[LGDVHKH[DF\DQRIHUUDWH

PAGE 23

2SWLPL]DWLRQ RI WKH DQDO\WLFDO UHVSRQVH 0RVW ZRUNHUV KDYH DUULYHG DW RSWLPDO H[SHULPHQWDO FRQGLWLRQV E\ H[DPLQLQJ D VLQJOH IDFWRU DW D WLPH GHWHUPLQLQJ WKH RSWLPDO YDOXH IRU WKDW IDFWRU EHIRUH SURFHHGLQJ ZLWK WKH QH[W ZKLFK ZDV WKHQ RSWLPL]HG DW WKH RSWLPDO OHYHO RI WKH SUHFHGLQJ IDFWRUVf 2ZLQJ WR WKH ODUJH QXPEHU RI YDULDEOHV SK\VLFDO FKHPLFDO DQG SK\VLFRFKHPLFDOf WKDW DUH DEOH WR LQIOXHQFH WKH VLJQDO REWDLQHG WKHUH KDV EHHQ D WUHQG WRZDUGV WKH DSSOLFDWLRQ RI RSWLPL]DWLRQ PHWKRGV WR DQDO\WLFDO IORZ VWUHDPV 7KH DSSOLFDWLRQ RI FKHPRPHWULF SULQFLSOHV WR WKH GHWHUPLQDWLRQ RI WKH RSWLPDO FRPELQDWLRQ RI SDUDPHWHUV WKDW ZRXOG DIIRUG WKH EHVW DQDO\WLFDO VHQVLWLYLW\ SUHFLVLRQ DQG DQDO\WLFDO UREXVWQHVV LV UHODWLYHO\ UHFHQW &KHPRPHWULF SULQFLSOHV KDYH LQ WKH PDLQ EHHQ DSSOLHG WR RSWLPL]LQJ FKURPDWRJUDSKLF V\VWHPV KRZHYHU WKH\ FDQ DQG KDYH EHHQ DSSOLHG WR WKH RSWLPL]DWLRQ RI DQDO\WLFDO IORZ V\VWHPV ,WHUDWLYH GHVLJQV DV RSSRVHG WR JULG VHDUFK PHWKRGV KDYH SURYHG WR EH WKH PRUH XVHIXO LQ RSWLPL]LQJ DQDO\WLFDO H[SHULPHQWDO FRQGLWLRQV WKH VLPSOH[ DSSURDFK LV WKH PRVW FRPPRQO\ HPSOR\HG 7KH EDVLF VLPSOH[ DSSURDFK DV LQWURGXFHG E\ 6SHQGOH\ HI DO >@ KRZHYHU KDV WKH GLVDGYDQWDJH RI WDNLQJ D UDWKHU ORQJ WLPH WR ORFDWH WKH PD[LPXP DQG WKH DWWHQGDQW ULVN RI GHWHUPLQLQJ D VHFRQGDU\ DV RSSRVHG WR WKH SULPDU\ PD[LPXP +HQFH D QXPEHU RI YDULDWLRQV WR WKH EDVLF DSSURDFK YDULRXVO\ GHVFULEHG DV PRGLILHG VLPSOH[ GHVLJQV KDYH EHHQ GHYHORSHG WR IDFLOLWDWH GHWHUPLQDWLRQ RI WKH WUXH DV RSSRVHG WR VHFRQGDU\ PD[LPXP DQG WR

PAGE 24

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

PAGE 25

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n SODQDULW\ RI WKH SODFWDP ULQJ FRPSURPLVLQJ DPLGH UHVRQDQFH ZLWKLQ WKH S ODFWDP ULQJ ,Q WKH FHSKDORVSRULQV WKLV ULQJ IXVLRQ HIIHFW RQ HOHFWURQ GHORFDOL]DWLRQ LV DPSOLILHG E\ WKH HQDPLQH UHVRQDQFH RXWVLGH WKH ODFWDP ULQJ 5HSRUWV LQ WKH OLWHUDWXUH KDYH VXJJHVWHG WKDW WKH PHFKDQLVP RI DFWLRQ LV UHODWHG WR ERWK WKH ODELOLW\ RI WKH SODFWDP DPLGH ERQG DQG FRQIRUPDWLRQ RI WKH DQWLELRWLF LQ WKH UHJLRQ RI WKH S ODFWDP ULQJ >@ 7KH HIIHFW RI ULQJ VL]H RQ ULQJIXVHG ODFWDP UHDFWLYLW\ DV ZHOO DV VXEVWLWXHQW HIIHFWV RQ EDVH K\GURO\VLV RI WKH FRPSRXQGV KDV DOVR EHHQ VWXGLHG *HQHUDOO\ WKH ELRORJLFDO DFWLYLW\ RI WKH SHQLFLOOLQV DQG FHSKDORVSRULQV KDV EHHQ FRUUHODWHG WR GHJUHH RI QRQSODQDULW\ LQ WKH S ODFWDP QLWURJHQ WKH ODFWDP & VWUHWFKLQJ IUHTXHQF\

PAGE 26

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f >@ &RORULPHWULF PHWKRGV KDYH VLPLODUO\ LQYROYHG GHJUDGLQJ WKH LQWDFW ODFWDP WR JHQHUDWH D FKURPRSKRUH ZLWK FKDUDFWHUL]DEOH XY DEVRUSWLRQ VSHFWUXP 7KLV KDV IRUPHG WKH EDVLV IRU WKH K\GUR[DPDWH PHWK\OHQH EOXH GLQLWUREHQ]RLF DFLG DQG WKH PRUH UHFHQW PHUFXULPHWULF DVVD\ PHWKRGV >@ 2SWLFDO URWDWRU\ GLVSHUVLRQ 25'f DQG FLUFXODU GLFKURLVP &'f WHFKQLTXHV KDYH IRXQG OLPLWHG DSSOLFDWLRQ WR WKH DQDO\VLV RI SHQLFLOOLQV 7KH\ KDYH EHHQ PDLQO\ DSSOLHG WR NLQHWLF GHWHUPLQDWLRQV RI WKH FRPSRXQGV ,QIUDUHG WHFKQLTXHV KDYH QRW IRXQG XVH LQ WKH TXDQWLWDWLYH DQDO\VLV RI S ODFWDPV EXW UDWKHU KDYH IRXQG XVH DV GLDJQRVWLF WRROV LQ TXDOLWDWLYH GHWHUPLQDWLRQV

PAGE 27

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f FROXPQV HPSOR\LQJ PHWKDQRO DFHWRQLWULOH DTXHRXV EXIIHUV RU FRPELQDWLRQV WKHUHRI DV WKH PRELOH SKDVH 7KH GHWHFWLRQ PHWKRG KDV W\SLFDOO\ EHHQ E\ XY DW QP 7KH SHQLFLOOLQV KDYH EHHQ UHSRUWHG WR HQKDQFH WKH OXPLQHVFHQFH DULVLQJ IURP WKH R[LGDWLRQ RI OXPLQRO LQ DONDOLQH PHGLD >@ 7KLV HQKDQFHPHQW KDV EHHQ VKRZQ WR EH TXDQWLWDWLYH IRU SHQLFLOOLQ 9 SHQLFLOOLQ DQG FHSKDORWKLQ ,W LV K\SRWKHVL]HG WKDW WKLV HQKDQFHPHQW LV FKDUDFWHULVWLF RI DOO S ODFWDP DQWLELRWLFV

PAGE 28

&+$37(5 2%-(&7,9(6 7KLV VWXG\ ZDV HVVHQWLDOO\ DQ H[WHQVLRQ RI WKH ZRUN FDUULHG RXW E\ 6FKXOPDQ HW DO >@ DQG &KHQ HW DO >@ RQ WKH HQKDQFHPHQW RI OXPLQRO FKHPLOXPLQHVFHQFH E\ VHOHFWHG EHWD ODFWDP DQWLELRWLFV &KHQ DQG FRZRUNHUV FDUULHG RXW WKHLU DQDO\VHV LQ D VWDWLF V\VWHP UHSRUWLQJ JRRG UHSURGXFLELOLW\ ZLWK EHQ]\O SHQLFLOOLQ SLSHUDFLOOLQ DQG SKHQR[\PHWK\O SHQLFLOOLQ SHQLFLOOLQ 9f DQG WKH FHSKDORVSRULQ FHSKDORWKLQ ,Q WKH FXUUHQW VWXG\ DQ DWWHPSW LV PDGH WR DQVZHU WKH IROORZLQJ TXHVWLRQV f 'R DOO "ODFWDP DQWLELRWLFV ZKHWKHU SHQLFLOOLQ RU FHSKDORVSRULQ HQKDQFH WKH FKHPLOXPLQHVFHQFH VLJQDO" f 7R ZKDW H[WHQW GR WKH FRPSRXQGV H[DPLQHG HQKDQFH WKH FKHPLOXPLQHVFHQFH" f :KDW VWUXFWXUDO IHDWXUHV DUH QHFHVVDU\ IRU WKH HQKDQFHPHQW WR EH PHDVXUDEOH" f :KDW LV WKH PHFKDQLVP RI HQKDQFHPHQW" f &DQ WKH FKHPLOXPLQHVFHQFH EH DSSOLHG WR SRVWFROXPQ GHWHFWLRQ LQ +3/&" ,I VR ZKDW VHW RI FRQGLWLRQV ZRXOG DFKLHYH PD[LPDO VHQVLWLYLW\ WKH EHVW OLPLW RI TXDQWLWDWLRQ DQG EHVW SUHFLVLRQ" &KHQ HW DO ZHUH DEOH WR GHPRQVWUDWH WKDW SHQLFLOOLQ SHQLFLOOLQ 9 SLSHUDFLOOLQ DQG FHSKDORWKLQ HQKDQFHG WKH FKHPLOXPLQHVFHQFH WR YDU\LQJ H[WHQWV 3UHOLPLQDU\

PAGE 29

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f RI HQKDQFHPHQW ZLOO EH FRPSDUHG LQ WHUPV RI WKH DQDO\WLFDO SDUDPHWHUV VXFK DV WKH OLQHDU G\QDPLF UDQJH VHQVLWLYLW\ OLPLW RI TXDQWLWDWLRQ SUHFLVLRQ HWF 'XH WR WKH SRRU UHSURGXFLELOLW\ DVVRFLDWHG ZLWK YDULDWLRQV LQ PL[LQJ D IORZLQJ DQDO\WLFDO VWUHDP ZDV HPSOR\HG )LQDOO\ DQ DWWHPSW LV PDGH WR FRUUHODWH WKH UHVXOWV REWDLQHG WR WKH UHDFWLYLWLHVVWDELOLW\ RI WKH UHVSHFWLYH EHWDODFWDP DQWLELRWLFV DQG WKH DSSOLFDWLRQ RI WKH PHWKRG IRU GHWHUPLQDWLRQ RI WKH SODFWDP DQWLELRWLFV LQ GRVDJH IRUPV LQGLFDWLQJ

PAGE 30

DFFXUDF\ SUHFLVLRQ UHSHDWDELOLW\ DQG UHSURGXFLELOLW\ RI WKH PHWKRG Vf

PAGE 31

D]HWLGLQRQH DPLQRSFQLFLOODQLF DFLG VXOKDFWDP WKLHQDP\FLQ FODYXODQLF DFLG ; %HQ]\O SHQLFLOOLQ DPSLFLOOLQ DPR[LFLOOLQ ; + < + ; 1+9 + ; 1+ Y RK PHWKLFLOOLQ )LJXUH &KHPLFDOV VWUXFWXUHV RI FRPSRXQGV H[DPLQHG

PAGE 32

)LJXUH FRQWLQXHG

PAGE 33

&+$37(5 (;3(5,0(17$/ 5HDJHQWV )OXRUHVFHLQ VRGLXP DQG 5XEUHQH ZHUH SXUFKDVHG IURP )OXND &KHPLFDO &R %XFKV 6ZLW]HUODQGf $PLQRDFULGLQH IURP 6LJPD &KHPLFDO &R 6W /RXLV 02f DQG 5KRGDPLQH 6 IURP t /DERUDWRULHV 3ODLQYLHZ 1
PAGE 34

3UHSDUDWLRQ RI /XPLQRO VROXWLRQV 6ROXWLRQV RI 0 OXPLQRO ZHUH SUHSDUHG E\ GLVVROYLQJ PJ $PLQRSKWKDOK\GUD]LGH LQWR D PL[WXUH RI DERXW PO ZDWHU DQG PO RI 0 VRGLXP K\GUR[LGH ZLWK VWLUULQJ WKHQ DGMXVWLQJ WR YROXPH LQ D PO YROXPHWULF IODVN 7KH VROXWLRQV ZHUH HPSOR\HG IRU KRXUV SULRU WR GLVFDUGLQJ 3UHSDUDWLRQ RI +YGURRHQ 3HUR[LGH VROXWLRQV 7KH K\GURJHQ SHUR[LGH VROXWLRQV ZHUH SUHSDUHG E\ SLSHWWLQJ PLV RI D b K\GURJHQ SHUR[LGH VROXWLRQ LQWR D PO YROXPHWULF IODVN FRQWDLQLQJ PO RI 0 VRGLXP K\GUR[LGH LQ DERXW PO ZDWHU WKHQ PDGH WR YROXPH ZLWK ZDWHU 3UHSDUDWLRQ RI 6WRFN 6RGLXP 'LKYGURDHQ 3KRVSKDWH VROXWLRQ $ 6WRFN VROXWLRQ RI PRQREDVLF VRGLXP SKRVSKDWH 1DOA3f 0 ZDV SUHSDUHG E\ GLVVROYLQJ PJ RI WKH DQK\GURXV VDOW LQWR PO ZDWHU )URP WKH VWRFN VROXWLRQ PO DOLTXRWV ZHUH GLOXWHG WR PO WR REWDLQ WKH 0 VROXWLRQV 3UHSDUDWLRQ RI 0RELOH SKDVH 7R PO PHWKDQRO LQ D PHDVXULQJ F\OLQGHU ZDV DGGHG WKH 0 VRGLXP GLK\GURJHQ SKRVSKDWH VROXWLRQ WR REWDLQ D ILQDO YROXPH RI PO 7KH VROXWLRQ ZHUH ILOWHUHG WKURXJK D [P PHPEUDQH ILOWHU DQG GHJDVVHG E\ VWLUULQJ WKH PRELOH SKDVH XQGHU YDFXXPVXFWLRQ

PAGE 35

3UHSDUDWLRQ RI &DOLEUDWLRQ VHULHV 6WRFN VROXWLRQV RI WKH S ODFWDP FRPSRXQGV ZHUH SUHSDUHG LQ GHLRQL]HG ZDWHU SULRU WR HDFK DQDO\VHV $SSURSULDWH YROXPHV RI WKH VROXWLRQV ZHUH WKHQ GLOXWHG ZLWK ZDWHU WR REWDLQ WKH GHVLUHG FRQFHQWUDWLRQ UDQJHV [ DOLTXRWV ZHUH LQMHFWHG LQWR WKH FKURPDWRJUDSKLF V\VWHP 0RELOH SKDVH PHWKDQRO LQ 0 VRGLXP SKRVSKDWH b 0H2+f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b +J2M VROXWLRQ $ VSHFWUXP ZDV DOVR REWDLQHG IRU D , DOLTXRW RI +M2M GLOXWHG LQ PO RI WKH 0 VRGLXP K\GUR[LGH VROXWLRQ

PAGE 36

6WDWLF VWXGLHV 7KH H[SHULPHQWDO VHWXS IRU WKH VWXGLHV LV VFKHPDWLFDOO\ LOOXVWUDWHG LQ )LJXUH 7KH UHDJHQW VROXWLRQ ZDV SUHSDUHG E\ GLOXWLQJ D SL DOLTXRW RI b K\GURJHQ SHUR[LGH WR PO LQ D YROXPHWULF IODVN XVLQJ D &7 PRODU VWRFN VROXWLRQ RI DPLQRSKWKDOK\GUD]LGH LQ 0 VRGLXP K\GUR[LGH )RU H[SHULPHQWDO UXQV Q RI WKH GLFOR[DFLOOLQ ZDV DGGHG WR PO RI WKH UHDJHQW VROXWLRQ LQ WKH FXYHWWH (DUOLHU VWXGLHV LQYROYHG WKH XVH RI FREDOW DQG FRSSHU LRQ VROXWLRQV DW D FRQFHQWUDWLRQ 0 7KH\ ZHUH KRZHYHU IRXQG XQQHFHVVDU\ LQ SUHOLPLQDU\ LQYHVWLJDWLRQV 7KH PHWDO LRQV HPSOR\HG ZHUH FREDOW ,,f DQG FRSSHU ,,f $Q DOWHUQDWH SURFHGXUH DGRSWHG ZDV WR LQMHFW D VSHFLILHG Q YROXPH RI b + LQWR WKH FXYHWWH FRQWDLQLQJ WKH OXPLQRO VROXWLRQ SULRU WR LQMHFWLRQ RI WKH DQWLELRWLF VROXWLRQ )ROORZLQJ LQMHFWLRQ RI WKH DQWLELRWLF VROXWLRQ WKH ,QWHQVLW\WLPH SURILOHV REWDLQHG ZHUH UHFRUGHG RQ D GLVN RYHU D VSHFLILHG WLPH SHULRG &RQGXFWLYLW\ PHDVXUHPHQWV 7R D EHDNHU FRQWDLQLQJ PO RI D ; 0 VWRFN VROXWLRQ RI OXPLQRO ZDV DGGHG S RI WKH b 7KH FKDQJHV LQ FRQGXFWLYLW\ ZLWK WLPH ZHUH WKHQ PHDVXUHG XVLQJ D FRQGXFWLYLW\ HOHFWURGH 0DUNNVRQ ,QVWUXPHQWVf

PAGE 37

N M V\ULQJH OLJKWSURRI HQFORVXUH SKRWRPXOWLSOLHU WXEH FXYHWWH )LJXUH 6FKHPDWLF GLDJUDP RI VHWXS IRU VWDWLF FKHPLOXPLQHVFHQW PHDVXUHPHQWV

PAGE 38

,QYHVWLJDWLRQ RI SRVVLEOH 3HUR[YODWH %DVHG 0HFKDQLVP 3URWRFRO IRU FRPSDULQJ GLIIHUHQW IOXRUHVFHUV $SSUR[LPDWHO\ 0 VROXWLRQV ZHUH SUHSDUHG E\ GLVVROYLQJ WKH DSSURSULDWH TXDQWLW\ RI IOXRUHVFHU )OXRUHVFHLQ 5KRGDPLQH 6 DPLQRDFULGLQH RU 5XEUHQHf LQWR D &7 0 VRGLXP K\GUR[LGH VROXWLRQ $ PO DOLTXRW ZDV WKHQ WUDQVIHUUHG LQWR D PO FXYHWWH SODFHG LQWR WKH VWDWLF VHWXS LOOXVWUDWHG LQ ILJXUH DQG WR WKLV ZDV DGGHG D DL PL[WXUH RI bf DQG GLFOR[DFLOOLQ DSSUR[ PJPOf f 7KH HPLVVLRQ REWDLQHG ZDV UHFRUGHG RQ GLVN RYHU D V WLPH SHULRG XVLQJ WKH $' ERDUG DQG 6SHFWUDFDOF VRIWZDUH 3URWRFRO IRU FRPSDULQJ GLIIHUHQW DQWLELRWLFV $SSUR[LPDWHO\ 0 VROXWLRQ RI IOXRUHVFHLQ ZDV SUHSDUHG E\ GLVVROYLQJ DERXW PJ RI IOXRUHVFHLQ VRGLXP DQK\GURXV 0: f DFFXUDWHO\ ZHLJKHG LQWR D 0 VRGLXP K\GUR[LGH VROXWLRQ $ PO DOLTXRW ZDV WKHQ SLSHWWHG LQWR D PO FXYHWWH SODFHG LQWR WKH VWDWLF VHWXS )LJXUH f DQG D P, PL[WXUH RI + bf DQG WKH DSSURSULDWH S ODFWDP DQWLELRWLF DSSUR[ PJPOf f ZDV DGGHG 7KH HPLVVLRQ REWDLQHG ZDV UHFRUGHG RQ GLVN RYHU D V WLPH SHULRG XVLQJ WKH $' ERDUG DQG 6SHFWUDFDOF VRIWZDUH

PAGE 39

)ORZ VWXGLHV )ORZ ,QMHFWLRQ $QDO\VHV 7KH VHW XS HPSOR\HG LV LOOXVWUDWHG LQ )LJXUHV DQG $ VLPLODU VHW XS ZDV HPSOR\HG IRU FKURPDWRJUDSKLF VWXGLHV ,Q WKH WZR FKDQQQHO V\VWHP WKH IORZLQJ VWUHDPV ZHUH DONDOLQH SHUR[LGH LQWR ZKLFK WKH DQDO\WH ZDV LQMHFWHG DQG DONDOLQH OXPLQRO )RU WKH WKUHH FKDQQHO V\VWHP IORZLQJ VWUHDPV ZHUH DONDOLQH SHUR[LGH DONDOLQH OXPLQRO DQG ZDWHUSKRVSKDWHZDWHU PHWKDQRO PL[WXUH ,Q ERWK FDVHV VDPSOHV ZHUH LQMHFWHG WKURXJK D 5KHRG\QH ORRS &RWDWL &DOLIRUQLDf LQMHFWRU IL[HG ZLWK D r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

PAGE 40

SHUR[LGH VWUHDP SXPS PHWDO LRQ VWUHDP SXPS OXPLQRO VWUHDP SXPS )LJXUH 7KUHH FKDQQHO IORZ V\VWHP DQDO\WH SKRWRFHOO ZDVWH

PAGE 41

r SHUR[LGH VWUHDP SXPS OXPLQRO VWUHDP SXPS )LJXUH 7ZR FKDQQHO IORZ V\VWHP DQDO\WH SKRWRFHOO ZDVWH

PAGE 42

PRELOH SKDVH UHDJHQW UHDJHQW ,, GDWD DFTXLVLWLRQ GHYLFH PRELOH SKDVH SULPDULO\ ZDWHU PLVFLEOH PL[WXUHV UHDJHQW VWUHDPV DQ DONDOLQH OXPLQRO VWUHDP DQG DQ DONDOLQH SHUR[LGH VWUHDP FROXPQ VLOLFD RFW\OVLODQH RU RFWDGHF\OVLODQH )LJXUH &KURPDWRJUDSKLF VHWXS HPSOR\HG

PAGE 43

&KURPDWRJUDSKLF FROXPQV HPSOR\HG LQ WKH VWXGLHV LQFOXGH ([VLO 5 2'6 .H\VWRQH VFLHQWLILF ,QF 'LPHQVLRQV ; PP 3DUWLFOH VL]H 3RUH VL]H QP $ 1XFOHRVLO 5 & .H\VWRQH VFLHQWLILF ,QF 'LPHQVLRQV ; PP SDUWLFOH VL]H SRUH VL]H LLQ $ 0LFURVRUE 6L 5DLQLQ ,QVWUXPDQW &RPSDQ\ ,QF 'LPHQVLRQV ; PP SDUWLFOH VL]H SRUH VL]H _LP $ )RU WKH PRVW SDUW WKH PRELOH SKDVH FRQVLVWHG RI PHWKDQRO 0 VRGLXPSRWDVVLXP SKRVSKDWH 1D+3 f )ORZ UDWH POPLQ 'HWHFWLRQ V\VWHP SRVWFROXPQ FKHPLOXPLQHVFHQFH 0 OXPLQRO LQ 0 1D2+ POPLQ +J2J LQ 0 1D2+ POPLQ XOWUDYLROHW DW QP 1% $OO PHWDO VWDLQOHVV VWHHOf WXELQJ ZDV RI LQFKHV LQWHUQDO GLDPHWHU

PAGE 44

3URWRFRO IRU &DOLEUDWLRQ &XUYH GHWHUPLQDWLRQV $ ; 0 VROXWLRQ RI OXPLQRO PDGH DONDOLQH LQ 0 VRGLXP K\GUR[LGH ZDV HPSOR\HG DV RQH RI WKH UHDJHQW VWUHDPV 7KH DONDOLQH K\GURJHQ SHUR[LGH VWUHDP ZDV SUHSDUHG E\ GLOXWLQJ PO RI D b YY VWRFN VROXWLRQ WR PO LQ D IODVN XVLQJ ; 0 VRGLXP K\GUR[LGH 7KH WZR VROXWLRQV ZHUH HPSOR\HG DV UHDJHQW VWUHDPV LQ D WZR FKDQQHO IORZ LQMHFWLRQ V\VWHP DV LOOXVWUDWHG LQ )LJXUH $ Q DOLTXRW RI WKH DQDO\WH VROXWLRQ LQ WKH UHVSHFWLYH FDOLEUDWLRQ VHULHV ZDV LQMHFWHG LQWR D VWUHDP RI WKH PRELOH SKDVH (TXLSPHQW f )/ +3/& SOXV VSHFWURIOXRUHVFHQFH GHWHFWRU 0F3KHUVRQ ,QVWUXPHQW $FWRQ 0DVVDFKXVHWWV f 3KDUPDFLD /.% +3/& SXPS /.%3URGXNWHU %URPPD 6ZHGHQ f /'& $QDO\WLFDO &RQVWDPHWULF PHWHULQJ SXPS /'& $QDO\WLFDO 5LYLHUD EHDFK )ORULGD f 6HUYRJRU UHFRUGHU 1RUPD *RHU] ,QVWUXPHQWV 2VWHUUHLFK $XVWULD f ,%0 3& ;7 f '7 $QDORJ DQG 'LJLWDO ,QSXW2XWSXW ERDUG 'DWD 7UDQVODWLRQ ,QF 0DOERUD 0DVVDFKXVHWWV

PAGE 45

f 6HOHFWUR 0DUN $QDO\]HU 0DUNVRQ 6FLHQFH ,QF 'HO0DU &RORUDGR 6RIWZDUH f 6SHFWUDFDOF *DO£FWLFD ,QGXVWULHV &RUSRUDWLRQ 6DOHP 1HZ +DPSVKLUH f $FXFKHP VHULHV VSHFLHV YHUVLRQ 1,67 *DLWKHUVEXUJ 0DU\ODQG 6RXUFH RI :DWHU $OO ZDWHU HPSOR\HG LQ WKH H[SHULPHQWV ZDV SXULILHG E\ ILOWUDWLRQ WKURXJK D PLOOLSRUH 0LOOL4 ZDWHU ILOWUDWLRQ DSSDUDWXV 0LOOLSRUH &RUSRUDWLRQ %HGIRUG 0$f SULRU WR XVH

PAGE 46

7DEOH $QWLELRWLF FRPSRXQGV VFUHHQHG IRU HQKDQFHPHQW &RPSRXQG 6RXUFH D]HWLGLQRQH DPLQRSHQLFLLODQLF DFLG DPR[LFLOOLQ DPSLFLOOLQ EHQ]\OSHQLFLOOLQ FHIRWD[LQH FHSKDORVSRULQ & FHSKDORWKLQ 1D FODYXOLQLF DFLG G EHQ]\OSHQLFLOOHQLF DFLG G SHQLFLOODPLQH GLFOR[DFLOOLQ KHWDFLOOLQ LPLSHQHP PHWKLFLOOLQ SKHQHWKLFLOOLQ SKHQR[\PHWK\OSHQLFLOOLQ SLSHUDFLOLQ VXOEDFWDP LPLSHQHP $OGULFK 6LJPD 6LJPD $&6 'REIDU 6LJPD 6LJPD 6LJPD (OL /LOO\ /DEV %ULVWRO 6LJPD 6LJPD 6LJPD %HHFKDP 5HVHDUFK /DEV 0HUFN 6KDUS t 'RKPH %ULVWRO /DEV 6LJPD 6LJPD 3IL]HU )OXND 0HUFN t &R

PAGE 47

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f 7R WKLV VHULHV RI FRQVHFXWLYH UHDFWLRQV :KLWH HW DO DSSOLHG WKH VWHDG\ VWDWH DSSUR[LPDWLRQ 0DNLQJ DPRQJ RWKHUV WKH DVVXPSWLRQ WKDW N +J2f ! NJ f :KLWH HW DO GHULYHG WKH UDWH H[SUHVVLRQ IRU SKRWRQ HPLVVLRQ

PAGE 48

UDWH N >,,,@ >2+@ >@ f ZKHUH ,,, LV WKH PRQRDQLRQLF OXPLQRO EDVH DQG -F NON N[ >+` $ SVHXGRILUVW RUGHU UDWH FRQVWDQW RI ; n Vn DW r& ZDV UHSRUWHG IRU WKH PRQRDQLRQLF OXPLQRO EDVH 0HUQ\L DQG FRZRUNHUV >@ LQ D VHULHV RI SXEOLFDWLRQV DOVR RXWOLQHG D SRVVLEOH PHFKDQLVP IRU WKH FKHPLOXPLQHVFHQW UHDFWLRQ RI OXPLQRO ,Q WKH HDUOLHVW RI WKH WKUHH SDSHUV >@ WKH\ GHVFULEH WKH K\GURSHUR[LGH DV EHLQJ WKH ILUVW FULWLFDO LQWHUPHGLDWH LQ WKH FKHPLOXPLQHVFHQW SDWKZD\ RI OXPLQRO DQG VXJJHVW WKDW WKH K\GURSHUR[LGH GHFRPSRVHV WR WKH FKHPLOXPLQHVFHQW DPLQRSKWKDODWH LRQ DW S+ YDOXHV DERYH LWV S.M S.J sf ZKHUHDV DW S+ YDOXHV EHORZ LWV S.D WKH K\GURSHUR[LGH GHFRPSRVHV EDFN WR R[\JHQ DQG LWV SDUHQW K\GUD]LGH 7KH SDSHU GLG QRW SURSRVH D PHFKDQLVP IRU WKH IRUPDWLRQ RI WKH K\GURSHUR[LGH RU WKH QDWXUH RI WKH R[LGDWLRQ VWDJH ,Q D ODWHU SXEOLFDWLRQ 0HUQ\L HW DO >@ SURSRVHG D SDWKZD\ IRU WKH R[LGDWLRQ RI WKH RQHHOHFWURQR[LGL]HG OXPLQRO E\ PROHFXODU R[\JHQ UHSRUWLQJ D RQHHOHFWURQ UHGXFWLRQ SRWHQWLDO RI s 9 YV 1+( IRU DPLQRSKWKDOD]LQH GLRQH +RZHYHU OLWWOH LQGLFDWLRQ LV SURYLGHG DV WR KRZ WKH RQHHOHFWURQR[LGL]HG OXPLQRO LV JHQHUDWHG DV ZHOO DV DQ LQGLFDWLRQ RI WKH OHYHOV RI PROHFXODU R[\JHQ DQGRU ar LQ DTXHRXV DONDOLQH VROXWLRQV

PAGE 49

R R R )LJXUH 5HDFWLRQ VFKHPH DIWHU :KLWH HWDO >@

PAGE 50

0RUH UHFHQWO\ 6DWWDU DQG (SVWHLQ >@ RXWOLQHG DQ RYHUDOO UHDFWLRQ VFKHPH IRU WKH UHDFWLRQ RI OXPLQRO ZLWK EDVLF K\GURJHQ SHUR[LGH IURP SXOVH UDGLRO\VLV VWXGLHV FDUULHG RXW E\ 0HUQ\L DQG FRZRUNHUV >@ )LJXUH f 7KH SUHFLVH UROH RI D QXPEHU RI FRPSRXQGV WKDW VHQVLWL]H FKHPLOXPLQHVFHQW UHDFWLRQV KDV QRW EHHQ IXOO\ HOXFLGDWHG 9DULRXV VXJJHVWLRQV DV WR WKH UROH RI WKH VHQVLWL]LQJ DJHQWV KDYH UDQJHG IURP EHLQJ GLUHFW FDWDO\WLF DFWLYDWRUV WR DFWLQJ YLD FRPSOH[DWLRQ ZLWK WKH FDWDO\WLF PHWDO LRQV XVXDOO\ &Rr &Xf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f DQLRQ D VWURQJ R[LGL]LQJ VSHFLHV E\ GHPRQVWUDWLQJ WKH H[LVWHQFH RI LWV VSHFLILF PD[LPXP DW QP LQ WKH XY VSHFWUXP $V UHJDUGV WKH SRVVLEOH UROH RI WKH EHWDODFWDP DQWLELRWLFV LQ HQKDQFLQJ WKH FKHPLOXPLQHVFHQFH &KHQ HW DO >@ SURSRVHG WKH IRUPDWLRQ RI DQ DGGXFW EHWZHHQ SHQLFLOOLQ DQG WKH VXSHUR[LGH DQLRQ (PSOR\LQJ GLIIHUHQFH VSHFWURSKRWRPHWU\ WKH\ GHPRQVWUDWHG WKH H[LVWHQFH RI D FKURPRSKRUH ZLWK D PD[LPXP EHWZHHQ QP 7KLV DGGXFW ZDV SUHVXPHG

PAGE 51

)LJXUH 5HDFWLRQ VFKHPH DIWHU 6DWWDU DQG (SVWHLQ >@ GDUN SURGXFW

PAGE 52

WR OHQJWKHQ WKH OLIHWLPH RI WKH R[LGDQW 7KH SHQLFLOOLQV KRZHYHU KDYH EHHQ UHSRUWHG E\ 0DUWLQH] HI D >@ WR IRUP PHWDO LRQ FRPSOH[HV ZLWK FREDOW ZKRVH VSHFWUD PD\ GLIIHU XQGHU YDULRXV FRQGLWLRQV RU FLUFXPVWDQFHV )XUWKHUPRUH JLYHQ WKDW WKH OXPLQHVFHQFH LV NQRZQ WR EH HQKDQFHG ZLWKRXW WKH PHWDO LRQ WKH SRVWXODWHG DGGXFW LV QRW QHFHVVDULO\ UHVSRQVLEOH IRU WKH HQKDQFHPHQW DORQH 7KH XOWUDYLROHW YLVLEOH VSHFWUXP REWDLQHG RQ DGGLWLRQ RI OA WR GLFOR[DFLOOLQ VROXWLRQV LQ WKH DEVHQFH RI D PHWDO LRQ GLG QRW GHPRQVWUDWH WKH H[LVWHQFH RI D XQLTXH FKURPRSKRUH )LJXUH f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f SRVVLEO\ XVLQJ D SODWLQXP HOHFWURGH DJDLQVW D VLOYHUVLOYHU FKORULGH RU RWKHU VXLWDEOH UHIHUHQFH 6XLWDEOH HTXLSPHQW ZDV QRW DYDLODEOH WR FRQGXFW WKH QHFHVVDU\ H[SHULPHQWV $ VHFRQG SRVVLEOH PRGH RI OXPLQHVFHQFH HQKDQFHPHQW ZDV WKDW WKH HQKDQFHU ZDV VLPSO\ DEOH WR GRQDWH WKH HQHUJ\ DULVLQJ IURP LWV R[LGDWLRQ WR DPLQRSKWKDODWH ZKLFK WKHQ UHHPLWV WKH HQHUJ\ DV OLJKW VLPLODU WR WKH FKHPLOXPLQHVFHQFH H[KLELWHG E\ WKH SHUR[\R[DODWHV 7KH SRVVLELOLW\ WKDW WKH

PAGE 53

$EVRUEDQFH b b b f§ b ; b FRQWUROf )LJXUH 8OWUDYLROHW VSHFWUXP RI GLFOR[DFLOOLQ RQ DGGLWLRQ RI YDU\LQJ FRQFHQWUDWLRQV RI +J2M

PAGE 54

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n R[LGDWLRQ $ SRVVLEOH YDULDWLRQ RI WKH PHFKDQLVP LV WKDW WKH REVHUYHG HPLVVLRQ FRXOG DULVH IURP WKH H[FLWDWLRQ RI WKH IOXRUHVFHU E\ WKH OLJKW HQHUJ\ HPLWWHG E\ OXPLQRO ZKLFK ZDV WKHQ UHHPLWWHG DW D ORQJHU ZDYHOHQJWK E\ D PHFKDQLVP DNLQ WR WKDW RI IOXRUV LQ OLTXLG VFLQWLOODWLRQ FRXQWHUV 6HYHUDO IOXRUHVFHQW FRPSRXQGV ZHUH WHVWHG 7DEOH f DQG ZKHUH SRVVLEOH LQWHQVLW\WLPH SURILOHV ZHUH REWDLQHG &RPSDULVRQ RI WKH FKHPLOXPLQHVFHQFH VSHFWUXP RI WKH UHDFWLRQ PL[WXUH WR WKDW

PAGE 55

RI WKH SXUH OXPLQRO IOXRUHVFHLQ RU DQ\ RWKHU IOXRUHVFHQW HQKDQFHU ZRXOG KHOS DVFHUWDLQ WKLV DVSHFW RI WKH PHFKDQLVP ,W ZDV DOVR VKRZQ WKDW WKH PRUH HIILFLHQW IOXRUHVFHUV UKRGDPLQH DQG IOXRUHVFHLQ JHQHUDWHG PRUH LQWHQVH OLJKW RXWSXW SURILOHV WKDQ WKH DPLQRSKWKDODWH LRQ )LJXUH f 7KH LQWHQVLW\ WLPH HPLVVLRQ SURILOHV IRU WKH R[LGDWLRQ RI VHOHFWHG SHQLFLOOLQV LQ WKH SUHVFHQFH RI IOXRUHVFHLQ ZHUH DOVR REWDLQHG )LJXUH f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f LV DOVR UHSRUWHG WR FDWDO\VH WKH FKHPLOXPLQHVFHQW UHDFWLRQ LQ WKH SUHVFHQFH RI WKH FDWLRQLF VXUIDFWDQW FHW\OWULPHWK\O DPPRQLXP EURPLGH ,W VKRXOG EH QRWHG KRZHYHU WKDW DWWHPSWV WR FRQVLVWHQWO\ UHSURGXFH WKH HPLVVLRQ ZHUH QRW SURGXFWLYH

PAGE 56

7DEOH )OXRUHVFHQW FRPSRXQGV H[DPLQHG IRU SRWHQWLDO XVH DV UHDJHQWV &RPSRXQG 2EVHUYDWLRQV PDGH DFULGLQH DPLQRDFULGLQH DPLQRSKWKDODWH )OXRUHVFHLQ VRGLXP QDSKWKDOHQH FDUER[\OLF DFLG 5KRGDPLQH 6 5XEUHQH QR PHDVXUDEOH VLJQDO QR PHDVXUDEOH VLJQDO VLJQDO REWDLQHG VLJQDO REWDLQHG QR PHDVXUDEOH VLJQDO VLJQDO REWDLQHG SRRUO\ VROXEOH

PAGE 57

,QWHQVLW\ 9f )LJXUH ,QWHQVLW\WLPH SURILOHV IRU GLFOR[DFLOOLQ LQ WKH SUHVFHQFH RI VHOHFWHG IOXRUHVFHUV

PAGE 58

,QWHQVLW\ WLPH Vf )LJXUH ,QWHQVLW\WLPH SURILOHV IRU VHOHFWHG S ODFWDP FRPSRXQGV LQ WKH SUHVFHQFH RI IOXRUHVFHLQ

PAGE 59

IOXRUHVFHU r IOXRUHVFHU KY )LJXUH 3RVWXODWHG PHFKDQLVP IRU VFLVVLRQ RI EODFWDP ULQJ

PAGE 60

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f $V PHQWLRQHG HDUOLHU WKHUH ZDV QR VSHFWURSKRWRPHWULF HYLGHQFH RI WKH IRUPDWLRQ RI DQ DGGXFW EHWZHHQ WKH R[LGDQW DQG GLFOR[DFLOOLQ

PAGE 61

,QWHQVLW\ 9f )LJXUH ,QWHQVLW\WLPH SURILOH IRU GLFOR[DFLOOLQOXPLQRO RQ HPSOR\LQJ GLIIHUHQW R[LGDQWV

PAGE 62

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f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

PAGE 63

2n 2n 1+ 1+ 2 )LJXUH 5HDFWLRQ PHFKDQLVP LQYROYLQJ K\GUD]LQH JHQHUDWLRQ

PAGE 64

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

PAGE 65

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t 7DEOH f 7KH UHVXOWV REWDLQHG FRPSDUH ZHOO ZLWK WKH RYHUDOO UDWH FRQVWDQW GHWHUPLQHG IROORZLQJ DSSOLFDWLRQ RI WKH VWHDG\ VWDWH DSSUR[LPDWLRQ E\ :KLWH HW DO >@ RI Nf ;&Vn DW A& +RZHYHU D VLQJOH VDPSOH WWHVW RQ WKH UDWH FRQVWDQWV GLG QRW UHYHDO D VLJQLILFDQW GLIIHUHQFH EHWZHHQ WKH UDWH FRQVWDQWV WKLV ZRXOG VXJJHVW D VLPLODU LI QRW LGHQWLFDO ILQDO UHDFWLRQ VWHS DQG ZRXOG EH FRQVLVWHQW ZLWK WKH REVHUYDWLRQ PDGH E\ RWKHU ZRUNHUV WKDW WKH FDWDO\WLF VWDJH ZDV QRW UDWH OLPLWLQJ DQG QRW WKH WHUPLQDO VWDJH )URP H[DPLQDWLRQ RI WKH YDULRXV PHFKDQLVWLF SRVVLELOLWLHV D QXPEHU RI UHDFWLRQ VFKHPHV ZHUH GUDZQ XS DQG WHVWHG IRU VXLWDELOLW\ 7KH REMHFW EHLQJ WR GHYHORS D PRGHO DEOH WR DW OHDVW TXDOLWDWLYHO\ VLPXODWH WKH FKHPLOXPLQHVFHQW OLJKW HPLVVLRQ SURILOH REWDLQHG LQ WKH SUHVHQFH DQG DEVHQFH RI WKH EHWDODFWDP DQWLELRWLFV )RU WKLV SXUSRVH WZR UHODWHG UHDFWLRQ VLPXODWLRQ SURJUDPV ZHUH HPSOR\HG

PAGE 66

,QWHQVLW\ 9f 2 WLPH Vf )LJXUH ,QWHQVLW\WLPH OXPLQHVFHQFH SURILOHV IRU VHOHFWHG FRPSRXQGV LQ OXPLQRO VROXWLRQV

PAGE 67

7DEOH .LQHWLF GDWD REWDLQHG IURP ,QWHQVLW\WLPH SURILOHV &RPSRXQG N ,Vf 5 VTXDUHG DPR[LFLOOLQ H[S f GLFOR[DFLOOLQ H[S f EHQ]\OSHQLFLOOLQ H[S f SKHQR[\PHWK\O SHQLFLOOL H[S f FRQWURO H[S f $OO UHDGLQJV ZHUH REWDLQHG DW DQ DPELHQW WHPSHUDWXUH RI &

PAGE 68

,Q WKH LQLWLDO LQVWDQFH WKH UHDFWLRQ PHFKDQLVP DQG DVVRFLDWHG UDWH FRQVWDQWV GHVFULEHG E\ 6DWWDU DQG (SVWHLQ >@ )LJ f ZDV DSSOLHG WR WKH +,&+(0 SURJUDP IRXQG WR VLPXODWH DW OHDVW TXDOLWDWLYHO\ WKH REVHUYHG UHVXOWV )LJXUHV f $ PDMRU SUREOHP ZLWK WKH 6DWWDU t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f ZKLFK ZDV ILWWHG LQWR DQ $&8&+(0 LQSXW ILOH )LJXUH f %\ YDU\LQJ WKH H[SRQHQW WHUP RI WKH UDWH FRQVWDQW IRU WKH LQLWLDO R[LGDWLRQ UHDFWLRQ IRU OXPLQRO LW ZDV SRVVLEOH WR GHPRQVWUDWH DQ HQKDQFHPHQW RI QRW RQO\ WKH SRVWXODWHG OLJKW RXWSXW EXW DOVR WR REWDLQ SURILOH VLPLODU WR WKRVH VHHQ LQ YLWUR

PAGE 69

N 'f )LJXUH 5HDFWLRQ PRGHO IRU +,&+(0 VLPXODWLRQ

PAGE 70

5.% 6,08/ 2) /80,12/ 5;1 $ % ( % (2+ ( %2+ '1 +2+ ( (2+ % ( ( '3 ( (1' $( 22( % (1' (( ,, ,1387 08/7,3/,(5 )LJXUH ,QSXW ILOH IRU VLPXODWLRQV RQ +,&+(0

PAGE 71

,QWHQVLW\ DXf 7LPHV (f $ f§% 2+ f§%f§ ( $ '3 )LJXUH &RQFHQWUDWLRQ WLPH SURILOHV IRU UHDFWDQWV DQG LQWHUPHGLDWHV

PAGE 72

)LJXUH 6LPXODWHG OLJKW HPLVVLRQ SURILOH

PAGE 73

nn N 2f FF QK FRR FRRn 'f KR Q )LJXUH 5HDFWLRQ PRGHO IRU $&8&+(0 VLPXODWLRQ

PAGE 74

c,1387 ),/( )25 /80,12/ 6,08/$7,21 %$6,& $% & ( & (+ ( & ( (1' $ ( %( (1' )LJXUH ,QSXW ILOH IRU VLPXODWLRQV RQ $&8&+(0

PAGE 75

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

PAGE 76

&RQFHQWUDWLRQ 0,f 7LPHV (f WLPH Vf ‘ N f§N N f§Kf§ N N s N )LJXUH 6LPXODWHG FRQFHQWUDWLRQWLPH SURILOHV IRU DPLQRSKWKDODWH JHQHUDWLRQ IRU GLIIHUHQW YDOXHV IRU WKH UDWH FRQVWDQW

PAGE 77

,QWHQVLW\ DXf 7LPHV (f N f§ff§ N [ N % N ; N s N )LJXUH 6LPXODWHG OLJKW HPLVVLRQ SURILOHV IRU GLIIHUHQW YDOXHV RI UDWH FRQVWDQW N

PAGE 78

6WDWLF 6WXGLHV 7KH SRVVLELOLW\ RI HPSOR\LQJ XYYLVLEOH VSHFWURSKRWRPHWU\ DV DQ LQYHVWLJDWLYH WRRO LQ WKH HQKDQFHPHQW RI OXPLQRO FKHPLOXPLQHVFHQFH ZDV DOVR H[DPLQHG ([FHSW IRU WKH DSSHDUDQFH RI D UHODWLYHO\ ZHDN VKRXOGHU LQ WKH UHJLRQ QP LW ZDV QRW SRVVLEOH WR GHPRQVWUDWH D VLJQLILFDQW FKDQJH LQ WKH XOWUDYLROHWYLVLEOH VSHFWUXP LQ WKH UHJLRQ QP RQ DGGLWLRQ RI K\GURJHQ SHUR[LGH WR D OXPLQRO FRQWDLQLQJ VROXWLRQ )LJ f 7KH SUHVHQFH RI D FKURPRSKRUH SHUWXUEHG E\ DQ DV\PPHWULF FHQWHU LQ WKH SHQLFLOOLQV LQGLFDWHG WKH SRVVLELOLW\ RI HPSOR\LQJ RSWLFDO URWDWRU\ GLVSHUVLRQ RU FLUFXODU GLFKURLF 25'&'f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b YY PHWKDQRO LQ ZDWHU )LJ f

PAGE 79

)LJXUH ,QWHQVLW\WLPH FRQFHQWUDWLRQV 3URILOHV IRU OXPLQRO UHDFWLRQ pr GLIIHULQ PHWKDQRO

PAGE 80

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f 7KH FKRLFH RI PHWDO LRQ ZDV QRWHG WR TXDOLWDWLYHO\ LQIOXHQFH WKH VKDSH RI WKH SHDN &RSSHU LRQV ZHUH QRWHG WR JLYH ULVH WR OHVV EDVHOLQH QRLVH WKDQ FREDOW LRQV DW FRQFHQWUDWLRQV LQ WKH UHJLRQ RI 0 ,Q WKH OLWHUDWXUH GLIIHUHQW FRQFHQWUDWLRQV RI FRSSHU &8f LRQV KDYH DOVR EHHQ UHSRUWHG WR \LHOG GLIIHULQJ LQWHQVLW\WLPH SURILOHV >@ @ VHOHFWHG FRSSHU DV RSSRVHG WR FREDOW FLWLQJ DV WKHLU JURXQGV WKH FKURPDWRJUDSKLF FKDUDFWHULVWLFV REWDLQHG DQG WKH EHWWHU VLJQDO WR QRLVH UDWLR DV D

PAGE 81

UHVXOW RI WKH ORZHU EDFNJURXQG QRLVH :KHUH WKH ),$ VHWXS LQYROYHG PRUH WKDQ WZR VWUHDPV WKH RUGHU RI PL[LQJ ZDV IRXQG WR LQIOXHQFH WKH TXDOLW\ RI WKH VLJQDO REWDLQHG EHWWHU UHVXOWV ZHUH REWDLQHG IRU VWUHDPV PL[HG LQ WKH RUGHU DONDOLQH OXPLQRO VROXWLRQ ZLWK DONDOLQH K\GURJHQ SHUR[LGH SULRU WR PHUJLQJ ZLWK WKH PHWDO LRQ VWUHDP )LJ f 7KLV RUGHU ZDV FRQVLVWHQW ZLWK WKH UHDVRQLQJ WKDW WKH PHWDO LRQ VLPSO\ FDWDO\]HV WKH HPLVVLRQ DULVLQJ IURP WKH UHDFWLRQ RI OXPLQRO LQ DONDOLQH SHUR[LGH 7KH EHWDODFWDP DQDO\WH ZDV WKHQ LQMHFWHG LQWR WKH OXPLQRO VWUHDP SULRU WR WKH VWUHDP PHUJLQJ ZLWK WKH PHWDO LRQ VWUHDP 2SWLPDO IORZ UDWHV IRU WKH VHWXS ZHUH GHWHUPLQHG WR EH DSSUR[LPDWHO\ POPLQ SHU FKDQQHO JLYLQJ D WRWDO RXWSXW RI POPLQ IORZLQJ WKURXJK WKH GHWHFWRU $ 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

PAGE 82

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f VWUHDP )LJXUH f D QXPEHU RI SHQLFLOOLQV FHSKDORVSRULQV DQG UHODWHG FRPSRXQGV ZHUH H[DPLQHG 7KH FKURPDWRJUDP DQG VXEVHTXHQW KLVWRJUDP )LJXUH f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f

PAGE 83

,Q DQ DWWHPSW WR FRUUHODWH WKH GHJUHH RI HQKDQFHPHQW WR D PHDVXUH RI ULQJ VWUDLQ QR FRUUHODWLRQ ZDV IRXQG EHWZHHQ WKH HQKDQFHPHQW DQG WKH OLWHUDWXUH LQIUD UHG VWUHWFKLQJ IUHTXHQFLHV IRU WKH EHWDODFWDP FDUERQ\O JURXS )LJXUH f 7KH LQIOXHQFH RI VHOHFWHG GRVDJH IRUP H[FLSLHQWV RQ WKH FKHPLOXPLQHVFHQW VLJDQDO ZDV DOVR H[DPLQHG 7KH VXJDUV JDODFWRVH DQG VRUELWRO S K\GUR[\ F\FORGH[WULQ DQG VRGLXP FDUER[\PHWK\O FHOOXORVH DOO FRPSURPLVH WKH FKHPLOXPLQHVFHQFH HQKDQFHPHQW DQG ZRXOG KDYH WR EH UHPRYHG IURP WKH DQDO\WH VROXWLRQV IRU JRRG UHVXOWV 7DEOH f

PAGE 84

,QWHQVLW\ )LJXUH ),$ FKURPDWRJUDP IRU HQKDQFHPHQW E\ VHOHFWHG FRPSRXQGV

PAGE 85

DPLQRSHQLFLOODQLF DFLG FODYXODQLF DFLG SHQLFLOOLQ 9 DPSLFLOOLQ GLFOR[DFLOOLQ SKHQHWKLFLOOLQ D]HWLGLQRQH SHQLFLOODPLQH SLSHUDFLOOLQ FHIRWD[LQH KHWDFLOOLQ VXOEDFWDP FHSKDORVSRULQ & LPLSHQHP FHSKDORWKLQ PHWKLFLOOLQ SHQLFLOOLQ )LJXUH +LVWRJUDP IRU HQKDQFHPHQW E\ VHOHFWHG FRPSRXQGV

PAGE 86

,QWHQVLW\ 9f SKHQHWKLFLOOLQ FRQFHQWUDWLRQ P,f ; SHQ 9 FHSKDORWKLQ DPSLFLOOLQ SLSHUDFLOOLQ SHQ aYM ‘!L )LJXUH ),$ FDOLEUDWLRQ SORWV IRU VHOHFWHG FRPSRXQGV

PAGE 87

7DEOH $QDO\WLFDO SDUDPHWHUV IRU VHOHFWHG FRPSRXQGV IURP ),$ DQDO\VLV FRPSRXQG VHQVLWLYLW\ 9OPf /24 P,f 5 VTXDUHG 3UHFLVLRQ VGf EHQ]\OSHQLFLOOLQ ( Q f SLSHUDFLOOLQ ( Q f SKHQHWKLFLOOLQ ( Q f SHQLFLOOLQ 9 Q f DPSLFLOOLQ Q f FHSKDOROWKLQ Q f KHWDFLOOLQ QV FHSKDORVSRULQ & QV PHWKLFLOOLQ QV f f

PAGE 88

WR F f mM SHQLFLOODPLQH GLFOR[DFLOOLQ [ ; ; Z;}r aU ,5 6WUHWFKLQJ IUHTXHQF\ FPf ? )LJXUH 6FDWWHU GLDJUDP IRU ,QIUDUHG VWUHWFK IUHTXHQFLHV IRU VHOHFWHG FRPSRXQGV YHUVXV UHODWLYH HQKDQFHPHQW RI OXPLQRO UHDFWLRQ

PAGE 89

7DEOH ,QIOXHQFH RI VHOHFWHG H[FLSLHQWV RQ FKHPLOXPLQHVFHQW VLJQDO HQKDQFHPHQW E\ EHQ]\OSHQLFLOOLQ ([FLSLHQW UHODWLYH b EODQN FRQFHQWUDWLRQ VLJQDO *DODFWRVH ; 6RUELWRO [ VRGLXP FDUER[\PHWK\O FHOOXORVH [ EHWD K\GUR[\ F\FORGH[WULQ [

PAGE 90

/LTXLG &KURPDWRJUDSK\ 7KH H[SHULPHQWDO REMHFWLYH ZDV WR GHYHORS D FKURPDWRJUDSKLF V\VWHP WKDW ZRXOG SHUPLW VHSDUDWLRQ DQG TXDQWLWDWLRQ RI D PL[WXUH RI DW OHDVW WZR EHWDODFWDP DQWLELRWLFV HPSOR\LQJ D SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ V\VWHP 0RELOH SKDVH GHYHORSPHQW ZDV EDVHG RQ WKH UHVXOWV IURP ERWK VWDWLF VWXGLHV DQG WKH IORZ LQMHFWLRQ DQDO\VHV 3RVVLEOH DSSOLFDWLRQ RI WKH HQKDQFHPHQW DV D SRVWn FROXPQ GHWHFWLRQ WHFKQLTXH LQ WKH FKURPDWRJUDSKLF DQDO\VLV RI SHQLFLOOLQV ZDV DOVR H[DPLQHG &KURPDWRJUDSKLF VHSDUDWLRQ RI SHQLFLOOLQV LV ZHOO GRFXPHQWHG >@ LQ PRVW FDVHV XY GHWHFWLRQ KDV EHHQ HPSOR\HG IRU YLVXDOL]DWLRQ RI WKH FROXPQ HOXDWH ,QLWLDO H[SHULPHQWV LQGLFDWHG WKDW VHSDUDWLRQ ZLWK FKHPLOXPLQHVFHQW GHWHFWLRQ LV SRVVLEOH XVLQJ VLOLFD FROXPQV +RZHYHU PRUH UHFHQWO\ 1DNDVKLPD HW DO >@ DSSOLHG UHYHUVH SKDVH VHSDUDWLRQ RQ RFWDGHF\OVLODQH FROXPQV ZLWK SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ )URP WKH VHSDUDWLRQ V\VWHPV UHSRUWHG LQ WKH OLWHUDWXUH LW ZDV GHFLGHG WR FRQILQH WKH PRELOH SKDVH FKRLFHV WR PDLQO\ DTXHRXV EDVHG PRELOH SKDVHV VLPSO\ WR SUHFOXGH FRPSURPLVLQJ WKH FKHPLOXPLQHVFHQW UHDFWLRQ E\ QRQDTXHRXV PHGLD RU KHDWV RI VROXWLRQ DULVLQJ IURP WKH PL[LQJ RI WZR RU PRUH GLIIHUHQW VWUHDPV )RU H[DPSOH ZKHUHDV DFHWRQLWULOH DQG WHWUDK\GURIXUDQ ZHUH IRXQG WR TXHQFK WKH OXPLQHVFHQFH PHWKDQRO JHQHUDWHG D VLJQLILFDQW KHDW RI VROXWLRQ RQ PL[LQJ ZLWK DTXHRXV VWUHDPV UHVXOWLQJ LQ D VLJQLILFDQW EDVHOLQH GULIW (PSOR\LQJ D 0 .KA32r PRELOH SKDVH XVHIXO FKURPDWRJUDPV KDG EHHQ REWDLQHG IURP ZKLFK DQDO\WLFDO VHQVLWLYLW\ ZDV PHDVXUHG DQG GHWHUPLQDWLRQ RI OLPLWV RI GHWHFWLRQ ZDV PDGH IRU D QXPEHU RI SHQLFLOOLQV 7DEOH f )URP WKH LQLWLDO

PAGE 91

H[SHULPHQWV VLOLFD FROXPQ EDVHG VHSDUDWLRQV ZLWK SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ ZHUH IRXQG WR EH IHDVLEOH KRZHYHU HPSKDVLV ZDV SODFHG RQ GHYHORSLQJ D UHYHUVH SKDVH VHSHUDWLRQ V\VWHP GXH WR WKH DWWHQGDQW DGYDQWDJH RI ORZHU RSHUDWLQJ FRVWV 7KH FULWHULD DSSOLHG WR VHOHFWLQJ WKH DSSURSULDWH VWDUWLQJ PRELOH SKDVH ZHUH WKDW WKH SKDVH EH SDUWO\ RU HQWLUHO\ DTXHRXV SHUPLW UHYHUVH SKDVH VHSDUDWLRQ QRW FRQWDLQ NQRZQ TXHQFKHUV RI WKH FKHPLOXPLQHVFHQFH DQG EH HDVLO\ DGMXVWHG WR S+ E\ PL[LQJ ZLWK SRVW FROXPQ DONDOLQH VWUHDPV $ QXPEHU RI WKH FKURPDWRJUDSKLF V\VWHPV GHVFULEHG LQ WKH OLWHUDWXUH IRU WKH VHSDUDWLRQ RI SHQLFLOOLQV ZHUH QRW OLNHO\ WR VXSSRUW FKHPLOXPLQHVFHQFH ODUJHO\ GXH WR WKH RUJDQLF PRGLILHU RU EXIIHU HPSOR\HG LQ WKH FKURPDWRJUDSKLF PRELOH SKDVH SUHVXPDEO\ WR HQVXUH WKDW RQO\ RQH LRQL]HG IRUP RI WKH SHQLFLOOLQ H[LVWV LQ WKH V\VWHP $WWHPSWV WR HPSOR\ D PHWKDQROZDWHUDFHWRQLWULOH FRPELQDWLRQ ZLWKRXW WKH PRGLILHU IDLOHG WR DFKLHYH VHSDUDWLRQ RI WKH SHQLFLOOLQV ZLWK DOO HOXWLQJ DW WKH VROYHQW IURQW 6XEVHTXHQW H[SHULPHQWDO UXQV ZHUH EDVHG RQ WKH VHSDUDWLRQ UHSRUWHG E\ :KLWH DQG =DUHPER >@ WKHVH ZHUH LQLWLDOO\ FDUULHG RXW RQ DQ RFWDGHF\OVLODQH 2'6f FROXPQ XVLQJ D PHWKDQRO LQ 0 VRGLXP GLK\GURJHQ SKRVSKDWH 1D+J34 f PRELOH SKDVH ZLWK D XY GHWHFWRU VHW DW QP 7KLV V\WHP ZDV DEOH WR VHSHUDWH GLFOR[DFLOOLQ IURP SHQLFLOOLQ EXW GLG QRW VXSSRUW SRVW FROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ /RZHULQJ WKH EXIIHU FRQFHQWUDWLRQ WR 0 SHUPLWWHG SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ EXW VLJQLILFDQWO\ UHGXFHG WKH VROXWH UHWHQWLRQ WLPHV 7KH UHGXFHG UHWHQWLRQ WLPHVf FRPSURPLVHG WKH XVHIXOQHVV RI SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ DV UHDVRQDEOH GLIIHUHQFH LQ UHWHQWLRQ

PAGE 92

WLPHV LV QHFHVVDU\ WR FRXQWHU WKH HIIHFWV RI EDQG EURDGHQLQJ DULVLQJ IURP WKH SRVWn FROXPQ GHWHFWLRQ PHWKRG 7KH XVH RI 0 SRWDVVLXP GLK\GURJHQ SKRVSKDWH .+J34f DORQH DV PRELOH SKDVH IRU WKH HOXWLRQ RI SHQLFLOOLQ JDYH ULVH WR EURDG SHDNV RQ D QRLV\ EDVHOLQH 7KH DGGLWLRQ RI b PHWKDQRO WR WKH 0 .IA3&A VLJQLILFDQWO\ UHGXFHG WKH QRLVH EXW GLG QRW \LHOG XVHIXO SHDNV 7KH XVH RI D b 0HWKDQRO LQ 0 .KA3 PRELOH SKDVH HOXWHG WKH SHQLFLOOLQV EXW UHTXLUHG WKH LQWURGXFWLRQ RI D GHOD\ FRLO7KH GHOD\ FRLO VHUYHG WR LPSURYH WKH H[WHQW RI PL[LQJ DQG DOORZ WKH UHDFWLRQ WR GHYHORS LPSURYLQJ WKH VLJQDO WR QRLVH UDWLR 8VLQJ D FP PL[LQJ ORRS WKH FKHPLOXPLQHVFHQFH HQKDQFHPHQW ZDV QRW GHWHFWDEOH ZKHUHDV OHQJWKHQLQJ WKH PL[LQJ ORRS WR FP DIIRUGHG D PHDVXUDEOH FKHPLOXPLQHVFHQFH HQKDQFHPHQW VLJQDO (PSOR\LQJ b PHWKDQRO LQ 0 .,A34 DV PRELOH SKDVH RQ D UHYHUVH SKDVH RFWDGHF\OVLODQH ([VLO 5 2'6f FROXPQ DW PO PLQ ZLWK UHDJHQW VWUHDPV RI DONDOLQH OXPLQRO DQG DONDOLQH K\GURJHQ SHUR[LGH HDFK DW POPLQ WKH UHWHQWLRQ WLPHV IRU VHOHFWHG EHWDODFWDP FRPSRXQGV ZHUH REWDLQHG 7DEOH f DQG FDOLEUDWLRQ FXUYHV REWDLQHG IRU GLFOR[DFLOOLQ SHQLFLOOLQ 9 DQG SHQLFLOOLQ )LJ f (PSOR\LQJ WKH WKUHH FRPSRXQGV WKH SUHFLVLRQ UHSRUWHG DV D VWDQGDUG GHYLDWLRQ ZDV GHWHUPLQHG 7DEOH f $Q HVWLPDWH RI WKH ORVV RI UHVROXWLRQ DVVRFLDWHG ZLWK WKH SRVWFROXPQ FKHPLOXPLQHVFHQFH GHWHFWLRQ ZDV DOVR GHWHPLQHG HPSOR\LQJ SHQLFLOOLQ 9 DQG GLFOR[DFLOOLQ DV PRGHO FRPSRXQGV 7DEOH f 7KH UHVXOWV H[KLELW D ORVV RI UHVROXWLRQ RI b IRU WKH V\VWHP HPSOR\HG

PAGE 93

7DEOH &KURPDWRJUDSKLF UHWHQWLRQ WLPHV IRU VHOHFWHG FRPSRXQG &RPSRXQG UHWHQWLRQ WLPH PLQf DPLQRSHQLFLOODQLF DFLG DPR[LFLOOLQ QV DPSLFLOOLQ QV D]HWLGLQRQH QV FHIRWD[LQH FHSKDORVSRULQ & FHSKDORWKLQ FODYXOLQLF DFLG GSHQLFLOODPLQH QV GLFOR[DFLOOLQ KHWDFLOOLQ QV LPLSHQHP PHWKLFLOOLQ SHQLFLOOLQ SHQLFLOOLQ 9 SKHQHWKLFLOOLQ SLSHUDFLOOLQ VXOEDFWDP QV QV QR PHDVXUDEOH VLJQDO REWDLQHG &KURPDWRJUDSKLF V\VWHP &ROXPQ ([VLO 5 2'6 0RELOH SKDVH 0 +3 0H2+ f 5HDJHQW VWUHDPV DONDOLQH OXPLQRO DONDOLQH K\GURJHQ SHUR[LGH 'HWHFWLRQ V\VWHP &KHPLOXPLQHVFHQFH POPLQ POPLQ POPLQ

PAGE 94

( ( F 4! ' 4 , , FRQFHQWUDWLRQ P,f )LJXUH &DOLEUDWLRQ SORW IRU GLFOR[DFLOOLQ

PAGE 95

FRQFHQWUDWLRQ P,f )LJXUH &DOLEUDWLRQ SORW IRU SHQLFLOOLQ 9

PAGE 96

FRQFHQWUDWLRQ P,f )LJXUH &DOLEUDWLRQ SORW IRU SHQLFLOOLQ *

PAGE 97

( )LJXUH 7\SLFDO FKURPDWRJUDP REWDLQHG

PAGE 98

7DEOH $QDO\WLFDO SDUDPHWHUV IRU VHOHFWHG FRPSRXQGV &RPSRXQG ;FRHIILFLHQW 5 VTXDUHG 3UHFLVLRQ UHSOLFDWHV PPOP 3HQLFLOOLQ H[Sf 3HQLFLOOLQ 9 H[Sf 'LFOR[DFLOOLQ H[Sf ?

PAGE 99

7DEOH 5HVXOWV REWDLQHG IURP UHVROXWLRQ PHDVXUHPHQWV XY VSHFWURSKRWRPHWULF GHWHFWLRQ &RPSRXQG UHWHQWLRQ SHDN WLPH PLQf ZLGWK PLQf SHQLFLOOLQ 9 GLFOR[DFLOOLQ &DOFXODWHG UHVROXWLRQ 5V FKHPLOXPLQHVFHQFH GHWHFWLRQ &RPSRXQG UHWHQWLRQ SHDN ZLGWK WLPH PLQfPLQ SHQLFLOOLQ 9 GLFOR[DFLOOLQ &DOFXODWHG UHVROXWLRQ 5V

PAGE 100

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

PAGE 101

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f RI QRW PRUH WKDQ b DOFRKRO 7KH HQKDQFHPHQW LV QRW VXSSRUWHG E\ DFHWRQLWULOH RU WHWUDK\GURIXUDQ DQG LV FRPSURPLVHG E\ KLJK 0f FRQFHQWUDWLRQV RI SKRVSKDWH RU ERUDWH EXIIHU 7KH ORVV LQ UHVROXWLRQ DVVRFLDWHG ZLWK XVH RI WKH WHFKQLTXH DV D SRVWFROXPQ GHWHFWLRQ WHFKQLTXH LV RI WKH RUGHU RI b

PAGE 102

7KH HQKDQFHPHQW FDQ EH DSSOLHG WR WKH IORZ LQMHFWLRQ DQDO\VHV RI WKH ODFWDP DQWLELRWLF RU DV D SRVW FROXPQ GHWHFWLRQ WHFKQLTXH LQ OLTXLG FKURPDWRJUDSKLF DQDO\VHV RI SHQLFLOOLQ PL[WXUHV ZKRVH UHWHQWLRQ WLPHV GLIIHU E\ PRUH WKDQ m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

PAGE 103

$33(1',; 'HULYDWLRQ RI UDWH HTXDWLRQ IRU FKHPLOXPLQHVFHQW OLJKW SURGXFWLRQ 5HDFWLRQ VHTXHQFH NL IFO $f ID 1+ 2 $VVXPSWLRQV PDGH LQFOXGH WKH UHDFWLRQV DUH ILUVW RUGHU RU DW OHDVW SVHXGR ILUVW RUGHU ZLWK UHVSHFW WR WKH SULPDU\ UHDFWDQWV $ DQG & ‘‘N">&@

PAGE 104

$SSO\LQJ WKH VWHDG\ VWDWH K\SRWKHVLV WR WKH LQWHUPHGLDWH & LH UDWH RI IRUPDWLRQ RI & HTXDOV UDWH RI FRQVXPSWLRQ WKHQ WKH UDWH RI IRUPDWLRQ RI & LV JLYHQ E\ WKH HTXDWLRQ KHQFH N[>$@ Nn>&@ ZKHUH Nf N N NBf DQG >F@ N N NB[ >$@ >%@ WKHUHIRUH WKH UDWH RI SURGXFW IRUPDWLRQ LV JLYHQ E\ W NNO N N NO >$@ >%@ ,W ZRXOG WKHUHIRUH DSSHDU WKDW HQKDQFHPHQW RI HLWKHU WKH UDWH RI SHUR[\ VSHFLHV IRUPDWLRQ RU FRQVXPSWLRQ ZRXOG LPSURYH \LHOGV RI SURGXFW

PAGE 105

*LYHQ WKDW WKH FRQFHQWUDWLRQ RI $ DW DQ\ WLPH W LV JLYHQ E\ WKH SVHXGR ILUVW RUGHU UHODWLRQ $VVXPLQJ WKDW DW W >&@ ,QWHJUDWLQJ ZLWK UHVSHFW WR WLPH HTXDWLRQ WZR ZRXOG \LHOG WKH HTXDWLRQ LFn-T KHQFH WKH UDWH RI SURGXFWLRQ RI ZKLFK LQ WXUQ UHIOHFWV WKH OLJKW IOX[ LV JLYHQ E\ WKH HTXDWLRQ t' OHNAHNnWf mW NnN 1% $ TXDQWXP \LHOG RI XQLW\ LV DVVXPHG

PAGE 106

5()(5(1&( /,67 f (+ :KLWH DQG ') 5RVZHOO &KDSWHU &KHPL DQG %LROXPLQHVFHQFH (G -* %XUUf 0DUFHO 'HNNHU ,QF 1HZ
PAGE 107

f :6SHQGOH\ *% +H[W DQG )5 +LPVZRUWK 7HFKQRPHWULFV f -$ 1HOGHU DQG 5 0HDG &RPSXW f 3% 5\DQ 5 / %DUU DQG + 7RGG $QDO &KHP f (5 $EHUJ DQG $ 7 *XVWDYVVRQ $QDO &KLP $FWD f $ *XVWDZVRQ DQG ( 6XQGNYLVW $QDO &KLP $FWD f 3 +HGOXQG DQG $ *XVWDYVVRQ $QDO &KLP $FWD f $3 :DGH 30 6KLXQGX DQG 3' :HQW]HOO $QDO &KLP $FWD f 1 &KHQ $QDO &KLP $FWD f %HWWHULGJH 7 6O\ $3 :DGH DQG '* 3RUWHU $QDO &KHP f 50 6ZHHW DQG /) 'DKO $P &KHP 6RF f '+ 6LHK $QDO\WLFDO 3URILOHV RI 'UXJ VXEVWDQFHV (G )ORUH\f 9RO S $FDGHPLF 3UHVV ,QF 1HZ
PAGE 108

f : %UDXQ -7 +HUURQ DQG '. .DKDQHU ,QW &KHP .LQ f (+ :KLWH 2 =DILULRX ++ .£JL DQG -+0 +LOO $P &KHP 6RF f /LQG 0HUQ\L DQG 7 (ULNVHQ $P &KHP 6RF f 0HUQ\L /LQG DQG 7 ( (ULNVHQ 3K\V &KHP f 0HUQ\L DQG -6 /LQG $P &KHP 6RF f 6 6DWWDU DQG ,5 (SVWHLQ 3K\V &KHP f 7* %XUGR DQG :5 6HLW] $QDO &KHP f .LNXFKL 7 1DJDQR + +D\DNDZD < +LUDWD DQG 0 +LUREH $QDO &KHP f -+ 0DUWLQH] 30DUWLQH] 3*XWLHUUH] DQG /O 0DUWLQH] 7DODQWD f $* 0RKDQ 3HUR[\ODWH FKHPLOXPLQVFHQFH LQ &KHPL DQG %LROXPLQHVFHQFH (G -* %XUUf 0DUFHO 'HNNHU ,QF 1HZ
PAGE 109

f +2 +RXVH 0RGHUQ 6\QWKHWLF 5HDFWLRQV 6"G (GLWLRQ 7KH %HQMDPLQ&XPPLQJV 3XEOLVKLQJ &RPSDQ\ 6DQ )UDQFLVFR f &( 5DVPXVVHQ DQG 7 +LJXFKL 3KDUP 6FL f /$ 0LWVFKHU 3: +RZLVRQ DQG 7 6RNRORVNL $QWLELRW f %
PAGE 110

%,2*5$3+,&$/ 6.(7&+ -RKQ 0L\DZD JUDGXDWHG IURP WKH 8QLYHUVLW\ RI 1DLUREL ZLWK D %DFKHORUnV GHJUHH LQ SKDUPDF\ % 3KDUPf LQ REWDLQHG KLV 0DVWHU RI 6FLHQFH 06Ff GHJUHH IURP WKH 8QLYHUVLW\ RI 6WUDWKFO\GH LQ DQG FDPH WR WKH 8QLYHUVLW\ RI )ORULGD LQ REWDLQLQJ KLV GRFWRUDO GHJUHH LQ 'HFHPEHU

PAGE 111

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

Absorbance
44
0%
1.5%
0.75 % 1.2 %
--X-- 1.5 % (control)
Figure 11: Ultraviolet spectrum of dicloxacillin on addition of varying concentrations
of HgOj.


73
phosphate buffers were able to support post-column chemiluminescence detection.
This was found to be in agreement with reports by Nakashima et al [45] who
reported that borate as well as the imidazole buffer did support post-column
chemiluminescence detection. Attempts to employ the imidazole buffer pH 7.3 as
described in the BP 1988 [56] in the mobile phase did not yield a post column
detection system able to support chemiluminescence. However, it is perhaps worth
noting that Nakashima et al did not define the imidazole buffer employed in their
work.
Using a three channel setup consisting of a MeOH:0.001M Kl-^P04, a
luminol stream and an alkaline peroxide (h^02) stream (Figure 6) a number of
penicillins, cephalosporins and related compounds were examined. The
chromatogram and subsequent histogram (Figure 27, 28) indicate that on a molar
basis dicloxacillin exhibited the best enhancement of chemiluminescence within the
group examined, clavulanic acid also exhibited a comparable enhancement of
chemiluminescence, then to a lower extent benzylpenicillin, penicillin V,
phenethicillin, 6-aminopenicillanic acid, cephalothin, and piperacillin. A number of
compounds exhibited weak enhancement while others did not afford a measurable
enhancment of chemiluminescence. Of significance was the fact that penicillamine
and the p -lactam ring azetidinone did not enhance the chemiluminescence
signficantly.
Analytical parameters for selected penicillins were also obtained using the
flow injection analysis setup (Figure 29 and Table 5).


6
The intensity of luminescence can be used as the basis for determination
of any species whose concentration influences the rate or efficiency of the
chemiluminescent reaction. In applying chemiluminescence to analysis of the
species the reaction conditions should be adjusted so that the analyte of interest
is the limiting reagent in the system and all other reactants are in excess. To obtain
precise measurements the chemiluminescence reaction should be initiated in a
controlled and reproducible manner, largely because the emission intensity varies
with time as the reactants are consumed. Chemiluminescent analyses are reported
to have several advantages which include good sensitivity, a wide linear dynamic
range, low detection limits (femtomole to attomole range) and the requirement for
simple instrumentation [4].
The chemiluminescent signal is transient; hence, the measurement of
emitted light intensity is time dependent. The signal is therefore either recorded at
a specific time after mixing, or by integrating the light emission plot during the
entire time period or during a specific fraction of time when light is emitted.
Commonly the reactants are rapidly mixed and the emission intensity measured
as a function of time after mixing. This yields the plot illustrated below (Fig 3). The
initial part of the curve is influenced by the method of mixing employed, while the
general shape of the curve depends upon the kinetics of the reaction as well as
upon any changes in quantum yield with time.
Most chemiluminescence reactions are reported to have low efficiencies, less
than 10% [5], which has restricted their usefulness to analyses. The duration of the


mobile phase.
reagent I.
reagent II.
data acquisition
device.
mobile phase ; primarily water miscible mixtures
reagent streams ; an alkaline luminol stream and an alkaline peroxide stream
column; silica, octylsilane or octadecylsilane
Figure 8: Chromatographic setup employed.


2
chemiluminesce directly; however, it can also transfer the energy to another
molecule, which following excitation then emits the energy as light. The
luminescence exhibited by luminol was one of the earliest cases of
chemiluminescence to be studied. Luminol (5-amino-2,3-dihydrophthalazine-1,4-
dione) is an aminophthalic hydrazide (Fig 1) that is able to exist in several
tautomeric forms.
Figure 1: Luminol (3-aminophthalic acid hydrazide)
For luminol, the light emitting species is the excited 3-aminophthalate ion, which
emits a blue emission in water and a yellow green emission in dimethyl sulfoxide


LIST OF FIGURES
Figure page
1. Chemical structure 2 amincphthalate 2
2. Schematic cutline cf chemiluminescent reacticn pathway 5
3. Typical luminescence prefile 8
4. Varicus penicillins 22
5. Schematic diagram cf setup fer static chemiluminescent
measurements 28
6. Three channel flew system 31
7. Twc channel flew system 32
8. Chromatographic setup employed 33
9. Reaction scheme after White et al 40
10. Reaction scheme after Sattar et al 42
11. Ultraviolet visible spectrum of dicloxacillin on addition of varying
concentrations of \\02 44
12. Intensity-time profiles for dicloxacillin system in presence of selected
fluorescers 48
vi


28
k j
syringe
light-proof enclosure
photomultiplier tube
cuvette
Figure 5: Schematic diagram of setup for static chemiluminescent
measurements.


Figure 10: Reaction scheme after Sattar and Epstein [40].
dark product


11
usefulness in measuring, with precision, fast chemiluminescent reactions. The
procedure is also rather cumbersome requiring separate cuvettes for each
measurement as well as repeated opening of the light tight apparatus, which
necessitates special precautions to protect the photomultiplier tube.
The second approach is the two phase measurement system. In this system
the chemiluminescent reagents are immobilized on a solid support such as filter
paper, and the analyte is permitted to interact with the immobilized reagent by
diffusion or convection. The light emitted is measured using a microliter plate
reader or by contact printing with photographic detection [4]. The
chemiluminescent intensity of these systems is influenced by both the kinetics of
the reaction and the efficiency of the mass transfer processes bringing the
reactants together. The principal advantages of this system lie in the conservation
of reagents and the convenience of measurement.
The third approach involves the use of flow measurement systems. The
flow injection approach has been described as the most successful of the methods
[5]. It involves injection of the analyte into a stream of appropriate pH, remote from
the detector, and the chemiluminescent reagent flows in another stream. The two
streams meet at a T junction inside a light tight enclosure, then flow through a flat
coil placed immediately in front of a photomultiplier tube. This compact assembly
provides more rapid and reproducible mixing resulting in reproducible emission
intensities and permitting rapid sample throughput. The design of the mixing device
and the means of retaining the emitting solution in view of the detector are


31. Calibration plot for dicloxacillin 85
32. Calibration plot for penicillin V 86
33. Calibration chart for penicillin G 87
34. Representative chromatogram for calibration measurements 88
viii


85
120
100
E
E
80
.c
Q>
(0
Q.
60
40
20
0
I I I I |
0 0.0005 0.001 0.0015 0.002 0.0025 0.003
concentration (m/I)
1
0.0035
Figure 31: Calibration plot for dicloxacillin.


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
APPLICATION OF LUMINOL CHEMILUMINESCENCE
TO THE ANALYSIS OF THE BETA LACTAM ANTIBIOTICS
By
John H. Miyawa
December 1995
Chairman: John H. Perrin, PhD
Major Department: Medicinal Chemistry
The enhancement of luminol chemiluminescence by selected betalactam
antibiotics is examined and some observations and suggestions made as to the
mechanism (s) of enhancement and the analytical utility of the enhancement.
The results obtained suggest the possibility of a peroxylate-like enhancement
mechanism in which cleavage of the beta-lactam ring is responsible for the energy
release manifesting as an enhancement of chemiluminescence.
The enhancement is applied to liquid chromatography (HPLC) as a post-column
chemiluminescence detection method. The utility of the post-column detection
technique is compared to ultraviolet detection (254nm) in terms of two peak
separation criteria.
IX


19
CHAPTER 2
OBJECTIVES
This study was essentially an extension of the work carried out by Schulman
et al [14] and Chen et al [15] on the enhancement of luminol chemiluminescence
by selected beta lactam antibiotics. Chen and coworkers carried out their analyses
in a static system, reporting "good" reproducibility with benzyl penicillin, piperacillin
and phenoxymethyl penicillin (penicillin V) and the cephalosporin cephalothin.
In the current study an attempt is made to answer the following questions.
1). Do all /?-lactam antibiotics whether penicillin or cephalosporin enhance the
chemiluminescence signal?
2). To what extent do the compounds examined enhance the
chemiluminescence?
3). What structural features are necessary for the enhancement to be
measurable?
4). What is the mechanism of enhancement?
5). Can the chemiluminescence be applied to post-column detection in HPLC?
If so, what set of conditions would achieve maximal sensitivity, the best limit
of quantitation and best precision?
Chen et al were able to demonstrate that penicillin G, penicillin V, piperacillin and
cephalothin enhanced the chemiluminescence to varying extents. Preliminary


9
hydrogen peroxide so formed then being assayed by the luminol reaction.
Glucose Gluconic acid +
luminol + aminophthalate + hv
More recently Zhou et al [7] reported a method for the determination of Vitamin
B12 by means of the luminol-hydrogen peroxide system. In their experimental
setup the bound cobalt in vitamin B12 is released by acidification of the vitamin. The
cobalt so released is then permitted to quantitatively catalyze the oxidation of
luminol by hydrogen peroxide.
The coupling of reactions is not always possible as there is often the problem
of incompatibility of conditions for all the reactions. This has however not impeded
use of the technique for the analysis of a range of biochemicals.
The third category consists of those substances able to modify the primary
chemiluminescent reaction. The analysis of a wide range of substances falls into
this category. This includes the analyses of metal ions such as Arsenic(As3+),
Cobalt(Ccf+), and Nickel(N?+), nonmetallic inorganics such as the gases oxygen,
and nitrogen dioxide, the halide ions and a number of nitro-, amino-, or hydroxy-
group containing organics. These substances are reported to have either an
excitatory or inhibitory influence on the chemiluminescence exhibited by luminol.
More recently a number of drug substances have been reported to enhance a
variety of otherwise chemiluminescent reactions. For example a method has been


BIOGRAPHICAL SKETCH
John Miyawa graduated from the University of Nairobi with a Bachelor's
degree in pharmacy (B. Pharm.) in 1984, obtained his Master of Science (M.Sc.)
degree from the University of Strathclyde in 1988 and came to the University of
Florida in 1992 obtaining his doctoral degree in December 1995.
101


14
Optimization of the analytical response
Most workers have arrived at optimal experimental conditions by examining
a single factor at a time, determining the optimal value for that factor before
proceeding with the next which was then optimized at the optimal level of the
preceding factor(s).
Owing to the large number of variables (physical, chemical and
physicochemical) that are able to influence the signal obtained, there has been a
trend towards the application of optimization methods to analytical flow streams.
The application of chemometric principles to the determination of the optimal
combination of parameters that would afford the best analytical sensitivity,
precision and analytical robustness is relatively recent. Chemometric principles
have in the main been applied to optimizing chromatographic systems; however,
they can and have been applied to the optimization of analytical flow systems.
Iterative designs as opposed to grid search methods have proved to be the
more useful in optimizing analytical experimental conditions, the simplex approach
is the most commonly employed. The basic simplex approach as introduced by
Spendley ef al [18], however, has the disadvantage of taking a rather long time to
locate the maximum, and the attendant risk of determining a secondary as
opposed to the primary maximum. Hence a number of variations to the basic
approach, variously described as modified simplex designs, have been developed
to facilitate determination of the true as opposed to secondary maximum and to


51
A kinetic examination of the p-lactam hydrolysis reaction in the presence
and absence of luminol could provide insight on whether the chemiluminescence
enhancement is associated with beta-lactam ring opening or some other event.
A recent report by Nakashima et al [45] suggests that the H,02 stream is not
necessary for the chemiluminescence. The implication being that the penicillin-
peroxide adduct was not essential to the chemiluminescence enhancement. In an
attempt to reproduce their conditions, it was not possible to obtain a
chemiluminescence signal in the absence of the hydrogen peroxide stream. It is
however, possible that the formation of a penicillin hydrogen peroxide adduct may
not necessarily be the species responsible for the enhancement.
Determining whether the penicillins are able to enhance the luminescence
in the presence of differing oxidants may afford insight as to the nature of the
enhancement step. The alternative oxidants in this case would be the non-peroxide
oxidants such as C^, Br2 or sources of the same, namely the chlorates, bromates
or iodates as well as N-chloro- and N-bromo- succinimides. In this respect the
oxidant sodium hypochlorite was employed as a non peroxide oxidant. It was
shown that dicloxacillin did not increase the maximum intensity of emission, but an
enhancement that develops at a kinetically slower pace that was associated with
greater light output (Figure 15). As mentioned earlier there was no
spectrophotometric evidence of the formation of an adduct between the oxidant
and dicloxacillin.


72
result of the lower background noise. Where the FIA setup involved more than two
streams, the order of mixing was found to influence the quality of the signal
obtained, better results were obtained for streams mixed in the order alkaline
luminol solution with alkaline hydrogen peroxide prior to merging with the metal ion
stream (Fig 10). This order was consistent with the reasoning that the metal ion
simply catalyzes the emission arising from the reaction of luminol in alkaline
peroxide. The beta-lactam analyte was then injected into the luminol stream prior
to the stream merging with the metal ion stream. Optimal flow rates for the setup
were determined to be approximately 1.25 ml.min1 per channel, giving a total
output of 4.5 ml.min1 flowing through the detector. A 24 /I cell permitted
measurements in the millivolt range which was difficult to achieve with the 12 n\
cell. This was probably due to longer residence time of the chemiluminescent
stream within the larger volume detector.
It was also noted that the metal ion stream was not essential for
chemiluminescence enhancement, and could in fact be excluded altogether, the
effect of which was a reduction in the intensity of light emission associated with
removal of the catalytic influence of the metal ion. The reduction however does not
significantly compromise the enhancement brought about by the beta-lactams. As
a result all subsequent flow injection analyses excluded the metal ion stream. FIA
studies revealed that solvents such as acetonitrile, tetrahydrofuran and butanol
alone or in mixtures with water did not support the chemiluminescence. It was
observed that borate buffers did not support the chemiluminescence whereas


56
characterized in terms of the best exponential fit and compared to that obtained
in the presence of the other enhancers.
The assumption made here was that, the light emission rate is indicative of
a critical rate limiting step in the chemiluminescent reaction, the terminal phase of
the light emission profile was regressed by fitting a first order exponential decay
equation. It is assumed that the terminal decay profile is free from the modifications
arising from continuous mixing in the static system. The exponential fitting was
carried out on profiles obtained for amoxicillin, benzylpenicillin, dicloxacillin and
penicillin V and compared to a control (Fig 17 & Table 4).
The results obtained compare well with the overall rate constant determined
following application of the steady state approximation by White et al [36] of k=
2.4 X1CJ3 s'1 at 3^C. However a single sample t-test on the rate constants did not
reveal a significant difference between the rate constants, this would suggest a
similar if not identical final reaction step and would be consistent with the
observation made by other workers that the catalytic stage was not rate limiting
and not the terminal stage.
From examination of the various mechanistic possibilities a number of
reaction schemes were drawn up and tested for suitability. The object being to
develop a model able to at least qualitatively, simulate the chemiluminescent light
emission profile obtained in the presence and absence of the beta-lactam
antibiotics. For this purpose two related reaction simulation programs were
employed.


TABLE OF CONTENTS
ACKNOWLEDGMENTS ii
LIST OF TABLES v
LIST OF FIGURES vi
ABSTRACT ix
CHAPTERS
1. INTRODUCTION 1
Reaction mechanism 3
Application of luminol chemiluminescence to analysis in
solution 4
Optimization of analytical response 13
The beta lactam antibiotics 16
2. OBJECTIVES 19
3. EXPERIMENTAL 24
Reagents 24
Preparation of luminol solutions 25
Preparation of Hydrogen Peroxide solutions 25
Preparation of stock Sodium Dihydrogen Phosphate
solution 25
Preparation of Mobile Phase 25
Preparation of Calibration series 26
Ultraviolet spectrophotometric studies 26
Static studies 27
Conductivity measurements 27
Investigation of possible Peroxylate based mechanism 29
Protocol for comparing different Fluorescers 29
hi


15
accelerate determination of the optimal conditions [19 23]. Indeed automated
approaches have been developed to reduce the search time involved and develop
more comprehensive response surfaces. In automated methods, pattern search
methods have been reported [24] as preferential, over simplex based procedures
where up to three experimental variables are being considered. In the literature a
number of applications of a modified simplex method to the optimization of
analytical flow streams have been described. Calatayud and Sancho [11] applied
a modified simplex optimization to the oxidative determination of promethazine by
flow injection analysis, similarly Chen [25] applied a more complex modified
simplex method to select the optimum reaction conditions for the 1,10-
phenanthroline / l-^02 /-cetyl-trimethylammonium bromide / Ci/+ system.
The choice of an appropriate response function is probably the most critical
step in optimization. The response function selected should ideally provide a
means of optimizing all the factors able to influence analytical characteristics as
sensitivity, selectivity, precision, etc. In most analyses the response function
selected is usually the peak height at the peak maximum, though other peak
characteristics such as peak area and peak width have been employed. Indeed a
number of response functions that may be useful have been described by
Betteridge et al [26].


CHAPTER 5
CONCLUSIONS
In addition to the B-lactam antibiotics penicillin G, penicillin V, piperacillin and
cephalothin, a broader range of B- lactam ring containing compounds are able to
enhance the chemiluminescence exhibited by luminol. It appears that a "strained"
B-lactam ring is essential for the enhancement. However the presence of a p-
lactam ring alone does not insure enhancement of luminescence as other
structural features in the molecule appear to be able to modulate this enhancement
of luminol chemiluminescence.
The different penicillins enhance the chemiluminescence to varying degrees
exhibiting differing intensity profiles. Of the p- lactam compounds examined
dicloxacillin and clavulinate exhibit the most profound enhancement of
luminescence, more so than penicillin G and the phenoxyalkyl penicillins penicillin
V and phenethicillin. Cephalothin, methicillin, aminopenicillanic acid, piperacillin and
sulbactam exhibit much less enhancement. Ampicillin and its derivative hetacillin
on the other hand do not enhance the chemiluminescence as does the isolated p-
lactam ring azetidinone. These differences are presumed to arise from differences
in the accessibility of the nucleophilic oxidant to the electron deficient carbon of the
91


fluorescer
*
fluorescer + hv
Figure 14: Postulated mechanism for scission of b-lactam ring.


Intensity
49
0.6
0 100 200 300 400 500 600 700 800 900
time (s)
1000
Figure 13: Intensity-time profiles for selected p lactam compounds in the prescence
of fluorescein.


APPENDIX
Derivation of rate equation for chemiluminescent light production
Reaction sequence
ki
fcl
(A)
fa
NH2 O
+
Assumptions made include, the reactions are first order or at least pseudo- first
order with respect to the primary reactants A and C.
5 D.
51
k?[C]
94


7
reactions is influenced by the reaction conditions and may occur rapidly within one
sec, or last longer than 24 hours. In the development of chemiluminescent assay
methods the two basic factors that influence the intensity of chemiluminescence
(i.e. efficiency and rate) should be considered. The efficiency of the reaction
influences both analytical sensitivity and detection limits, while the reaction kinetics
determine both the precision and sample throughput.
Any substance able to quantitatively influence the light output can be
determined by chemiluminescence. In fact the luminol reaction has been used to
determine a number of compounds that are able to interact with the oxidant
initiating the chemiluminescent reaction. The substances so quantitated have not
necessarily been the analyte of interest but rather related to the analyte. These
substances have been divided into three categories by Grayeski [4].
Firstly those measured species that form one of the reagents that is consumed
in the course of the reaction. This case is exemplified by the determination of
hydrogen peroxide using the luminol system. This approach was applied to the
determination of hydrogen peroxide in irradiated water as early as 1955 [6].
The second category of reactions includes the analysis of compounds that are
able to generate one of the chemiluminescent reactants. An example here is the
indirect determination of glucose by treatment with the enzyme glucose oxidase.
The enzyme oxidizes the glucose to gluconic acid and hydrogen peroxide, the


16
The Betalactam antibiotics
The penicillins and the cephalosporins are established groups of antibiotics
characterized by the presence of a substituted 0-lactam fused to a sulfur
containing ring. For the penicillins the S-containing ring is a thiazolidine ring
whereas for the cephalosporins it is a dihydothiazine ring.
The penicillin and cephalosporin antibiotics both exert their antibiotic activity
by inhibiting the synthesis of bacterial cell walls. The/9- lactam ring is reactive and
is believed to be inactivated following acylation by the transpeptidase enzyme
which normally crosslinks peptidoglycan strands during cell wall synthesis. The
four membered p- lactam ring is strained not only as a result of ring size but also
as a result of fusion to the second ring. The immediate consequence of this is non
planarity of the p-lactam ring compromising amide resonance within the p -lactam
ring. In the cephalosporins this ring fusion effect on electron delocalization is
amplified by the enamine resonance outside the lactam ring. Reports in the
literature have suggested that the mechanism of action is related to both the lability
of the p-lactam amide bond and conformation of the antibiotic in the region of the
p- lactam ring [27]. The effect of ring size on ring-fused 0-lactam reactivity as well
as substituent effects on base hydrolysis of the compounds has also been studied.
Generally the biological activity of the penicillins and cephalosporins has been
correlated to
- degree of non-planarity in the p- lactam nitrogen
- the lactam C=0 stretching frequency


74
In an attempt to correlate the degree of enhancement to a measure of ring
strain, no correlation was found between the enhancement and the literature infra
red stretching frequencies for the beta-lactam carbonyl group (Figure 30).
The influence of selected dosage form excipients on the chemiluminescent siganal
was also examined. The sugars galactose and sorbitol, p hydroxy cyclodextrin and
sodium carboxymethyl cellulose all compromise the chemiluminescence
enhancement and would have to be removed from the analyte solutions for good
results (Table 6).


APPLICATION OF LUMINOL CHEMILUMINESCENCE
TO THE ANALYSIS OF THE BETA LACTAM ANTIBIOTICS
By
JOHN H. MIYAWA
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
1995


Concentration (M/I)
(Times 10E-7)
67
time (s)
k=0.5 k=5 k=50
h k= 100 k=200 -- k=500
Figure 24: Simulated concentration-time profiles for 2-aminophthalate generation
for different values for the rate constant


Intensity (a.u)
(Times 10E-9)
68
k=0.5
k=5
-x- k=50
-B- k=100
-X- k=200
-- k=500
Figure 25: Simulated light emission profiles for different values of rate constant k,.


4
derivative VI, which then undergoes ring opening to yield the excited 3-amino-
phthalate ion as illustrated in Fig 2. The mechanism of decomposition of the
azaquinone has been the subject of a number of postulations, which have been
reviewed by White and Roswell [1].
Application of Luminol Chemiluminescence to analysis in solution
The intensity of chemiluminescence arising from the chemiluminescent
reaction can be described by the following equation.
^CL= dC
dt
where <*>CL is the efficiency of the chemiluminescence and dC/dt the rate of the
chemical reaction. The efficiency of chemiluminescence depends on how
efficiently excited states are generated from the molecular reaction and on how
efficiently the excited states luminesce; i.e.,
I CL ^ium
where 0exc is the excitation efficiency and 0|Um the luminescence efficiency. Both
the excitation and luminescence efficiencies can be influenced by a variety of
reaction conditions such as solvents employed, concentration, pH and purity of
reagents.


88
195 5970 E00
Figure 34: Typical chromatogram obtained


95
Applying the steady state hypothesis to the intermediate C.
i.e. rate of formation of C equals rate of consumption, then the rate of formation
of C is given by the equation
hence
kx[A] =k'[C]
where k = (k2 + k3 + k_1)
and
[c]
K
k2 +k3 +k_x
[A] [B]
therefore the rate of product formation is given by
5 D
6 t
k2kl
k2 +k3 +k-l
[A] [.B]
It would therefore appear that enhancement of either the rate of peroxy species
formation or consumption would improve yields of product.


18
Of interest here is the ^9- lactam carbonyl stretch frequency which has been useful
in providing information on the structural integrity of the rings, the state of oxidation
of sulfur and the relative conformations of the^- lactam protons. In this respect an
increase in the p- lactam stretching frequency has been associated with an
increase in biological activity, a decrease in the lactam N planarity and increasing
ease of lactam amide bond hydrolysis by base.
Of the chromatographic methods gas chromatography has found limited
utility in the analysis of penicillins largely as a result of the non-volatility of the
penicillins and cephalosporins. Liquid chromatography has been used extensively
in both the development and application of the p- lactam antibiotics. The antibiotics
have been separated on reverse phase columns typically octadecylsilane (ODS)
columns employing methanol, acetonitrile, aqueous buffers or combinations thereof
as the mobile phase. The detection method has typically been by uv at 254 nm.
The penicillins have been reported to enhance the luminescence arising
from the oxidation of luminol in alkaline media [13,14,15]. This enhancement has
been shown to be quantitative for penicillin V, penicillin G and cephalothin. It is
hypothesized that this enhancement is characteristic of all p- lactam antibiotics.


Figure 2: Schematic outline of chemiluminescent reaction pathway.


45
luminescence arose from a peroxyoxalate-like process associated with the opening
up of the p- lactam ring in the penicillin was also examined. The rationale being
that the opening up of the strained p- lactam ring is the primary energy releasing
reaction that simply requires an appropriate fluorescer, which is able to absorb the
energy and release it as light. The strained p- lactam ring would open subsequent
to nucleophilic attack of the electron deficient keto group by the peroxy anion. In
opening the^- lactam ring the energy released excites a receptor molecule which
then emits the energy as light. In the case of luminol the receptor molecule could
possibly be luminol per se or the aminophthalate ion. Several experiments were
carried out to determine whether this was indeed the case. Initial experiments
suggested that 3- aminophthalate and fluorescein were both able to support the
p- lactam enhanced chemiluminescence in the presence of hydrogen peroxide.
However the enhancement from mixtures of aminophthalate and hydrogen
peroxide, was found to diminish significantly with time on standing. This could have
been due to oxidation of the fluorescer. Mohan [44] has indicated the importance
of taking into account the fluorescer stability to hydrogen peroxide and photo
oxidation. A possible variation of the mechanism is that the observed emission
could arise from the excitation of the fluorescer by the light energy emitted by
luminol which was then reemitted at a longer wavelength, by a mechanism akin to
that of fluors in liquid scintillation counters. Several fluorescent compounds were
tested (Table 3) and where possible intensity-time profiles were obtained.
Comparison of the chemiluminescence spectrum of the reaction mixture to that


69
Static Studies
The possibility of employing uv-visible spectrophotometry as an investigative
tool in the enhancement of luminol chemiluminescence was also examined. Except
for the appearance of a relatively weak shoulder in the region 400-460 nm, it was
not possible to demonstrate a significant change in the ultraviolet-visible spectrum
in the region 196-600 nm on addition of hydrogen peroxide to a luminol containing
solution (Fig 13). The presence of a chromophore perturbed by an asymmetric
center in the penicillins indicated the possibility of employing optical rotatory
dispersion or circular dichroic (ORD/CD) techniques as means of examining the
fate of the betalactam antibiotics during the reaction was also tried. This approach
has been applied by Rasmussen and Higuchi [53], Mitscher etal [54] and others,
in stability studies on the penicillins. However satisfactory experimental results were
difficult to obtain due to the very slow response of the variable wavelength
polarimeter employed. No available equipment is on hand for the current studies.
A problem that arose in applying the chiroptical method is the fact that in
addition to the optical activity exhibited by the penicillins, the hydroperoxide could
also be able to rotate plane-polarized light. This could complicate any interpretation
of optical activity vs time profiles.
From the static studies it was possible to demonstrate that the methanol
content of the analytical streams influences the extent of enhancement. Maximum
enhancement was found to lie between 10-20 % v/v methanol in water (Fig 26).


97
REFERENCE LIST
1) E.H. White and D.F. Roswell
Chapter 4
Chemi- and Bioluminescence (Ed J.G. Burr)
Marcel Dekker Inc, New York, 1985.
2). G. Mernyi and J. S. Lind, J. Am. Chem. Soc. 1980, 102, 5330.
3). E.H. White, J. Chem. Educ. 1957, 34, 5.
4). M.L. Grayeski, Anal. Chem. 1987, 59, 1244A.
5). A.Townshend, Analyst 1990, 115, 495.
6). W.V. Maynoerd, W. Anderson, H.D. Evans and D. Rosen, Radiat. Res.
1955, 3, 379.
7). Y.K. Zhou, H. U, Y Uu and G.Y. Uang, Anal. Chim. Acta 1991, 243, 127.
8). 1.1. Koukli and A.C. Calokerinos, Anal. Chim. Acta 1990, 236, 463.
9). K.Sato and S. Tanaka, Anal. Chim. Acta 1990, 236, 459.
10). 1.1. Koukli and A.C. Calokerinos, Analyst 1990, 115, 1553.
11). J.M. Calatayud and T.G. Sancho, J. Pharm. Biomed. Anal. 1992, 10, 37.
12). S.A.Halvatzis, M.M. Timotheou-Potamia and A.C. Calokerinos
Analyst 1990, 115, 1229.
13). D.S. Milbrath, Proc. of l\/h Inter. Biolum. & Chem. Symp 1986 515
14). S.G. Schulman, J.H. Perrin, G.F. Yan and S.X. Chen, Anal. Chim. Acta.
1991, 255, 383.
15). S. Chen, G. Yan, M.A. Schwartz, J.H. Perrin and S.G. Schulman.
J. Pharm. Sci. 1991, 80, 1017.
16). D.Betteridge, T.J. Sly, A.P. Wade and J.E.W. Tillman, Anal. Chem. 1983, 55,
1292.
17). A.B. Syropoulos and A.C. Calokerinos, Anal. Chim. Acta 1991, 255, 403.