Gradient elution, improved separations and analytical figures of merit in high performance liquid chromatography using e...

MISSING IMAGE

Material Information

Title:
Gradient elution, improved separations and analytical figures of merit in high performance liquid chromatography using electrochemical detection
Physical Description:
xiii, 112 leaves : ill. ; 28 cm.
Language:
English
Creator:
Hadjmohammadi, Mohammad Reza, 1947-
Publication Date:

Subjects

Subjects / Keywords:
Liquid chromatography   ( lcsh )
Voltammetry   ( lcsh )
Electrochemical analysis   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1983.
Bibliography:
Includes bibliographical references (leaves 107-111).
Statement of Responsibility:
by Mohammad Reza Hadjmohammadi.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000504008
notis - ACS4022
oclc - 12177759
System ID:
AA00003432:00001

Full Text










GRADIENT ELUTION, IMPROVED SEPARATIONS
AND ANALYTICAL FIGURES OF MERIT IN
HIGH PERFORMANCE LIQUID CHROMATOGRAPHY
USING ELECTROCHEMICAL DETECTION







BY

MOHAMMAD REZA HADJMOHAMMADI


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

UNIVERSITY OF FLORIDA


1983































To my parents, brother and sisters














ACKNOWLEDGEMENTS


I would like to express my deepest gratitude to

Dr. John G. Dorsey for his acceptance, guidance, advice,

patience, concern and support throughout my work.

My sincere thanks are also due to Dr. J. D. Winefordner,

Dr. R. A. Yost, Dr. A. Brajter-Toth, Dr. C. M. Riley,

Dr. R. G. Bates, and Dr. J. L. Ward for their encouragement,

advice and support. I would also like to thank my colleagues

and friends in Dr. Dorsey's research group, all of whom have

contributed much to my progress and have made my course of

study pleasant and fruitful.

Acknowledgement is also due to the donors of the

Petroleum Research Fund, administrated by the American Chemical

Society, for support of this research and to Nelson H. C. Cooke,

Altex Scientific, for a gift of columns.

Sincere thanks are also due to Laura Griggs for her

patience, helpful hints, and expert typing.

Finally, mere thanks are not enough to my parents,

brother, and sisters, whose sacrifices for me can never be

repaid.


iii














TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS................. ....... ... ..... .... 111

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

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

KEY TO ABBREVIATIONS... .... ................... ....... ix

ABSTRACT.... ........................................... xi

CHAPTER

ONE INTRODUCTION................................ 1

TWO THEORY AND BACKGROUND....................... 10
Basic Parameters in Chromatography..... 10
Capacity Factor (k')............... 10
Peak Resolution (Rs) .............. 11
Efficiency (E) .................... 11
Control of Separation in Liquid
Chromatography.................. 12
Gradient Elution (GE) ................. 13
Limit of Detection..................... 14
Principles of Electrochemical Detection 15
Ion-pair Chromatography............... 16
Micellar Chromatography................ 18

THREE EXPERIMENTAL................................ 21
Cyclic Voltammetry Systems ............. 21
Liquid Chromatography System........... 21
Electrochemical Detector............... 22
Pretreatment of Glassy Carbon Electrode 24
Reagents..................... ........... 26







Page


FOUR SIGNAL CHANGE AND BASELINE SHIFT USING
ELECTROCHEMICAL METHODS..................... 27
Change of Residual Current with DC, DP
and NP Voltammetry in Isocratic Elu-
tion HPLC ............................ 29
Effect of Background Electrolytes and
Gradient Elution on Baseline Shift
Using Amperometric Detection......... 37
Effect of the Electrode Material on
Baseline Shift Using Amperometric and
NRDP Voltammetry Methods ............. 43

FIVE A COMPARISON OF MICELLAR AND HYDROORGANIC
MOBILE PHASES USING AMPEROMETRIC DETECTOR... 46
Hydrodynamic Voltammogram in Micellar
and Hydroorganic Mobile Phases........ 47
Analytical Figures of Merit Comparison
between Micellar and Hydroorganic
Mobile Phases......................... 56
Gradient Elution and Selectivity in
Micellar Mobile Phase................ 69

SIX RAPID SEPARATION AND DETERMINATION OF THYRO-
MIMETIC IODOAMINO ACIDS BY GRADIENT ELUTION
REVERSE PHASE LIQUID CHROMATOGRAPHY WITH
ELECTROCHEMICAL DETECTION................... 85
Standard Solutions..................... 87
Preparation of T4 Tablet Solution and
Injectable T4 Sample................. 87
Gradient Elution LC/EC................. 87
Isocratic Separations.................. 91
Assay of T4 Preparations............... 98

SEVEN CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK. 103
Conclusions............................. 103
Suggestions for Future Work............ 104

REFERENCES... ...................... ................... 107
BIOGRAPHICAL SKETCH................................... 112














LIST OF TABLES


Page

1 Residual Current Change, AI, and Decay Current,
decay, in DC, DP and NP Voltammetry in Isocratic
Elution.. ....................................... 35

2 Normalized Baseline Shift (nA/mm2) during
Gradient Elution with No Sample Injection........ 44

3 Analytical Figures of Merit for Phenol............ 57

4 Analytical Figures of Merit for B-6 Vitamins..... 59

5 Analytical Figures of Merit for Polyaromatic
Hydrocarbons .................................... 61

6 Analytical Figures of Merit for T2, T3 and T4.... 100














LIST OF FIGURES


Page

1 Electrochemical Cell with Glassy Carbon Electrode 23

2 Residual Current Change and Decay Current with 5%
CH3CN............................................ 30

3 Residual Current Change and Decay Current with
50% CH3CN... .................................... 31

4 Residual Current Change and Decay Current with
95% CH3CN......................................... 32

5 Residual Current Change by NP Voltammetry........ 33

6 Specific Conductance vs. Percentage of Solvent B. 39

7 Baseline Shift during Gradient Elution with No
Sample Injection, Using an Amperometric Detector. 42

8 Chemical Structures of Compounds Used in Chapter
Five............................................. 48

9 Hydrodynamic Voltammogram for Phenol............. 50

10 Hydrodynamic Voltammogram for B-6 Vitamins........ 52

11 Hydrodynamic Voltammogram for Polyaromatic Hydro-
carbons.. ....................................... 54

12 Analytical Curves for Phenol..................... 63

13 Analytical Curves for B-6 Vitamins............... 65

14 Analytical Curves for Polyaromatic Hydrocarbons.. 67

15 Gradient Micellar Chromatogram for Separation of
Vitamin B-6...................................... 71

16 Gradient Micellar Chromatogram for Separation of
Phenolic Compounds...... ......................... 73

17 Gradient Micellar Chromatogram for Separation of
Phenolic Compounds and Gradient with No Injection 75

vii







Page

18 Effect of SDS Concentration on k'............... 77

19 Isocratic Micellar Chromatogram for Separation of
Vitamin B-6 ................... .................. 80

20 Isocratic Micellar Chromatogram for Separation of
Vitamin B-6..................... .................... 82

21 Isocratic Micellar Chromatogram for Separation of
Vitamin B-6...................................... 84

22 Baseline during Gradient Program with Blank
Injection.. ..................................... 90

23 Separation of Seven Thyromimetic Iodoamino Acids. 93

24 Thyromimetic lodoamino Acids Used in This Study.. 94

25 Cyclic Voltammogram.............. ............. 95

26 Hydrodynamic Voltammogram for T2, T3 and T4...... 97

27 Analytical Curves for T2, T and'T ....99
S............
28 Isocratic Separation of T2, T3 and T 1............ 01


viii













KEY TO ABBREVIATIONS


BPC Bonded-phase chromatography

CMC Critical micelle concentration

CV Cyclic voltammetry

DC Direct current

DP Differential pulse

EC Electrochemical detection or electrochemical
detector

GC Gas chromatography

GCE Glassy carbon electrode

GE Gradient elution

HDV Hydrodynamic voltammogram

HPLC High performance liquid chromatography

IP Ion-pair

LC Liquid chromatography

LC/EC Liquid chromatography with electrochemical
detection

LC/MS Liquid chromatography mass spectrometry

LDR Linear dynamic range

LLC Liquid-liquid chromatography

LOD Limit of detection

MSRTP Micelle-stabilized room-temperature phosphorescence

NMR Nuclear magnetic resonance

NP Normal pulse

NRDP Non-ramping differential pulse







PAH

PAR

RP

RP-HPLC


RP-LC

RP-LC/EC


SDS

WE


Polyaromatic hydrocarbon

Princeton applied research

Reversed-phase or reverse-phase

Reversed-phase high performance liquid chromatog-
raphy

Reversed-phase liquid chromatography

Reversed-phase liquid chromatography with electro-
chemical detection

Sodium dodecyl sulfate

Working electrode













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


GRADIENT ELUTION, IMPROVED SEPARATIONS
AND ANALYTICAL FIGURES OF MERIT IN
HIGH PERFORMANCE LIQUID CHROMATOGRAPHY
USING ELECTROCHEMICAL DETECTION

By

Mohammad Reza Hadjmohammadi

December 1983


Chairman: Dr. John G. Dorsey
Major Department: Chemistry



The goal of this work was an improvement and better

understanding of electroanalytical methods used as liquid

chromatographic detection techniques. Particularly, methods

which would allow the use of chromatographic gradient elu-

tion with electrochemical detectors were investigated. Both

amperometric and pulse techniques were investigated with

traditional hydroorganic mobile phases. The use of gradient

elution with micellar mobile phases was shown to allow com-

patibility with electrochemical detectors. A further com-

parison of analytical figures of merit was made between

hydroorganic and micellar mobile phases using electro:1emical

detection

A comparison of baseline shift during gradient elution

with amperometric and nonramping differential pulse (NRDP)







methods was performed using both glassy carbon and gold

electrodes. The composition of the mobile phase is virtu-

ally constant during one pulse with gradient elution, but

high residual current changes preclude routine use of this

technique. A more stable baseline was achieved with an

amperometric detector and a glassy carbon electrode, than

it was otherwise with NRDP method. A higher concentration

of phosphoric acid in the organic modifier as opposed to

equal concentrations in both modifier and water produced a

more stable baseline during gradient elution using a glassy

carbon electrode and an amperometric detector.

The hydrodynamic voltammograms and analytical figures

of merit for phenol, two B-6 vitamins, and polyaromatic

hydrocarbons were compared in micellar and hydroorganic

mobile phases. The limit of detection in both mobile phases

was comparable, whereas the upper limit of linear dynamic

range was greater in micellar mobile phases.

Gradient chromatograms for separation of phenolic com-

pounds and B-6 vitamins with an anionic surfactant and

phenolic compounds with a nonionic surfactant using ampero-

metric detection are shown. Selectivity of micellar mobile

phases toward B-6 vitamins changes with surfactant concen-

trations.

A rapid separation of thyroxine and related thyroid

hormones is shown using gradient elution and electrochemical

detection. A five minute isocratic separation of thyroxine

and three related hormones is also reported. Limits of


xii







detection are in the sub-nanogram range with an upper limit

of linear dynamic range of 500 to 1000 nanograms for these

compounds. Analysis of levothyroxine sodium tablets and

injectable intravenous samples is described.


xiii













CHAPTER ONE

INTRODUCTION



Chromatographic methods can be classified according to

a number of schemes. The major one is based on the state of

the mobile phase. If the mobile phase is a gas, the method

is called gas chromatography (GC). If the mobile phase is a

liquid, the method is named liquid chromatography (LC).

Liquid and gas chromatography are each divided according to

the nature of the stationary phase. When the solid station-

ary phase has adsorption properties, the process is called

adsorption chromatography, and when the stationary phase is

a liquid supported by an inert matrix, the process is called

partition chromatography.

Special classifications have also been introduced. For

example, chromatography can be classified according to

whether the stationary phase is present as a thin-layer, a

paper, or a column. Chromatography can alternatively be

classified with respect to the flow of the mobile phase;

this classification includes one-way, two-dimensional, and

radial chromatography. According to the mechanism of the

retention, i.e., the interaction between solutes and the

stationary phase, chromatographic methods are classified as

adsorption, partition, ion-exchange, and gel permeation.








Finally, the chromatographic technique may be classified

according to the kind of sample introduction onto the sta-

tionary phase and migration through the system. This gives

rise to development, elution, displacement and frontal anal-

ysis chromatography. The first two techniques are the most

common, while the last two methods are relatively special-

ized and, therefore, of limited value and use.

In 1941, Martin and Synge [1] were led to contrive a

scheme of liquid-liquid chromatography for separation of the

amino acids of a wool hydrolyzate with a countercurrent

extractor [2]. To improve the efficiency of the counter-

current method, Martin and Synge considered a means of

immobilizing one phase while the second phase flowed over

it in such a manner as to maximize the contact of the im-

miscible liquids at the interface. Martin believed this

would facilitate the rapid distribution, or partitioning, of

the solutes between the two phases. The efficiency of their

partitioning column was almost 104 times greater than that

of the countercurrent method [3].

The difference between modern liquid chromatography and

traditional column chromatography (whether adsorption, par-

tition or ion-exchange) involves improvements in equipment,

materials, technique, and the application of theory. Modern

liquid chromatography provides more convenience, better

accuracy, higher speed, and the ability to carry out diffi-

cult separations. At the beginning of the 1970's, the

modern form of liquid chromatography was named high pressure









liquid chromatography (HPLC). Later, the P for "pressure"

was replaced by P for "performance." The reason was the

appearance of microparticles that allowed researchers to

perform, at a lower pressure drop, the same efficient and

rapid analyses done with other supports.

In conventional liquid-liquid chromatography (LLC), the

stationary phase is a bulk liquid, mechanically held to the

support by adsorption. In recent years, organic phases are

chemically bonded to the support, leading to a separate LC

method called bonded-phase chromatography (BPC). Bonded-

phase chromatography is the most widely used in modern LC,

and many laboratories use BPC columns for their LC separa-

tion. In contrast to LLC, bonded-phase chromatography pack-

ings are quite stable because the stationary phases are

chemically bound to the support and cannot be easily removed

or lost during use. The availability of a wide variety of

functional groups in BPC packing allows for both normal-

and reversed-phase chromatography.

Polar BPC packing are used for normal-phase separa-

tions. Samples of moderate to strong polarity are usually

well separated by normal phase-chromatography. The mobile

phase in normal phase chromatography is a hydrocarbon sol-

vent such as hexane, heptane, or isooctane, plus small

amounts of a more polar solvent. The mobile phase strength

can be varied for a given application by varying the concen-

tration of the more polar solvent component. Reverse phase







BPC normally involves a relatively nonpolar stationary phase

(e.g., C8 or C18 hydrocarbon) used in conjunction with polar

(e.g., aqueous) mobile phases to separate a wide variety of

less polar solutes.

The general reaction for preparing bonded-phase pack-

ings on silica-based supports to produce siloxanes is:


ClSi(Me)2R
Si-OH + or -- Si-O-Si(Me)2R + HC1
R'OSi(Me)2R


where R is the desired organic moiety and, for reversed-

phase (RP) packing, is an n-alkyl chain (n = 2,8,18). This

reaction is based on silanol groups on the surface of the

siliceous support. Fully hydrolyzed silica contains about

8 pmol of silanol groups per m2. Because of steric hin-

drance, a maximum of about 4.5 pmol of silanol groups per m2

can be reacted at best [4], and an end-capping process

usually can be done to cover most of the residual silanols

by a reaction similar to the one above using chlorotrimethyl-

silane. The concentration of the organic moiety per m2 of

BPC packing depends upon the surface area of the packing

particles (e.g., pellicular or porous support). Siloxane

bonded-phase packing are available with pellicular or

totally porous supports. Bonded-phases of this type are

hydrolytically stable throughout the pH range 2-8.5. Due to

the different methods for the preparation of BPC packing

and shielding the residual silanol groups by end-capping,

the surface coverage and overall volume of the organic




5

stationary phase tends to show large differences from manu-

facturer to manufacturer, and even from lot to lot. Without

end-capping of the residual silanol groups, a mixed-retention

mechanism can result and lead to asymmetric peaks.

Increasing the alkyl chain length on BPC results in an

increase in selectivity and retention times [5,6] (when the

columns were compared) using the same mobile phase. The

study of chain length (C8-C22) of bonded organic phases

showed that the selectivity [7] depended on the chain length

of the bonded phase and on the molecular structure of the

solute. The same study showed that the utilization of long

chain phases made it possible to reduce the water content

of the water:methanol mobile phase, which increases the

efficiency and loading capacity. Hemetsberger et al. [8,9]

studied the behavior and the effect of structure of bonded

phases. Kikta and Grushka [10] studied the retention behav-

ior on alkyl bonded phases as a function of chain length,

surface coverage, solute type, mobile phase composition, and

temperature. Colin and Guiochon [11] compared the resolu-

tion of RPC packing (C6-C22). As a result, the shorter

length alkyl chain columns generally gave the worst resolu-

tion and efficiency. A comparative study on the separation,

efficiency under optimum mobile phase conditions with three

different mobile phases, and three groups of solutes on

three commercially available alkyl bonded phases (C2, C8 and

C18) was done by Haleem [12].

In high-performance liquid chromatography (HPLC), as in

all analytical methods, the trend is to do it faster and







cheaper, make it more selective and sensitive, and combine

it with other methods. Selection of column type in LC has

been more restricted because of the viscosities and solute

diffusivities in the liquid phase which are orders of mag-

nitude different from the values in GC. The most efficient

columns presently used in LC are those packed with totally

porous small particles (with particle size down to a few

micrometers). Although substantial reductions of the plate

height are achieved while decreasing the particle size,

there are some practical limits to this procedure. As

pointed out by Halasz [13], there are difficulties in the

uniform packing of very small particles, as well as the

problem with evolved heat of friction.

Microbore HPLC as now commercially available uses 1-

to 2-mm diameter columns. The next generation of columns,

studied only in research laboratories for the past few

years, may be inner-coated microtubular or packed microcap-

illary columns, just 50 pm in diameter. The advantage of

these would be hundreds of thousands of theoretical plates

in a very long length at low cost. Eluents from these col-

umns also could be fed into a mass spectrometer with low

interference from the solvent. These small diameter columns

often require new instruments or modifications of existing

instruments to meet new needs in injection, pumping, or

detection.

The rapid progress in HPLC places great demands on

detection techniques. Unfortunately, the highly developed







GC detectors are mostly useless at this time in LC because

of principal differences between LC and GC. An ideal detector

should be universal; however, the great diversity of systems

to be analyzed makes the construction of a sensitive univer-

sal detector impossible, and thus detectors monitoring vari-

ous physicochemical properties of substances are employed,

the optimal detection conditions being determined specific-

ally for each system. It is possible to measure either bulk

properties, which depend on the variation of the composition

of the system (e.g., refractive index, electrical conduc-

tance, etc.), or properties that selectively characterize

certain components in the mobile phase (such as the absor-

bance at a certain wavelength, fluorescence, electric current

at a certain electrode potential, etc.). The bulk property

detectors are universal detectors, and they require that

the properties of the solutes be substantially different

from those of the mobile phase to attain sufficient signal

changes during detection. The bulk property detectors are

usually less sensitive and subject to a higher noise than

the measurement of specific properties of the solutes, and

they are rarely compatible with gradient elution. On the

other hand, the measurement of specific properties requires

that the mobile phase yields the lowest possible signal

under the given conditions.

Currently available commercial LC detectors are based

on a number of detection principles, including absorbance,

fluorescence, refractive index, electrochemical reaction,





8

and mass spectrometry. A number of these detectors are

likely to be improved in the next few years. For instance,

new types of LC/MS interfaces will probably appear, and

currently available interfaces should be further refined.

Nuclear magnetic Resonance (NMR) detectors should material-

ize, and GC detectors are being seriously considered for

their applicability to LC detection. The development of

laser based detectors [14] for chromatography is in progress

and in many cases offer better sensitivity and selectivity

than conventional LC detectors. The complexity and expenses

of the laser based detectors have delayed the acceptance of

these detectors in the laboratory.

Attempts to use electrochemical detection of molecules)

in effluents from chromatographic columns were made long

before the advent of HPLC. The first papers dealing with

polarographic detectors were those of Drake [15] and

Kemula [16]. Present electrochemistry offers a large group

of methods that can be used for continuous detection of

substances [17]. The field of electrochemical HPLC has been

reviewed several times [17-19]. A survey of scientific

papers on LC detector usage during the 1980-81 period showed

that 4.3% of LC analyses were based on electrochemical

detectors [20].

The suitability of electrochemical detection to a given

problem ultimately depends on voltammetric characteristics

of the components) of interest in a suitable mobile phase

and a suitable working electrode surface. Electrochemical








detection is more limited with respect to the mobile phase

composition than other LC detection methods because of the

fact that a complex surface reaction which depends on the

medium is involved.

Direct electrochemical detection is not likely to be

useful in normal-phase chromatography since nonpolar organic

mobile phases are not well suited to many electrochemical

reactions. The HPLC stationary phases of choice clearly

include all ion-exchange and reverse-phase (RP) materials

since these are compatible with polar mobile phases contain-

ing some dissolved ions. The ionic strength, pH, electro-

chemical reactivity of the mobile phase and background elec-

trolyte, and presence of electroactive impurities (dissolved

oxygen, halides, trace metal ions) are all important con-

siderations.

The choice of electrode material is one of the important

considerations, because of the ruggedness, potential range,

residual current, and long-term stability requirements.

Electrode materials such as platinum, glassy carbon, gold,

and mercury films may work well in some cases but may be

disastrous in others. These electrodes are subject to com-

plicated surface renewal problems but are not mechanically

awkward devices such as the dropping mercury electrode.

With all limitations mentioned, liquid chromatography with

electrochemical detection (LC/EC) has three distinct

advantages for applicable systems, namely, selectivity,

sensitivity and economy.














CHAPTER TWO

THEORY AND BACKGROUND



Basic Parameters in Chromatography


Capacity Factor (k')

The capacity factor is equal to n /n where n and n
s m s m
are the number of moles of solute in the stationary and

mobile phase, respectively. Therefore, k' can be written

according to the following:


n [X] V KV t -t0
k' = -- = (1)
n [X] V V t
m mm m 0


where


[X] = concentration of solute X in the sta-
tionary phase

[XI = concentration of solute X in the mobile
m
phase

Vs = volume of the stationary phase

V = the total volume of the mobile phase within
the column

K = distribution constant

tR = retention time of solute X

t = time for mobile phase or other unretained
molecules to pass through column







Peak Resolution (Rs)

By convention, peak resolution, Rs, is defined as the

ratio of the distance between the two peak maxima (At) to

the mean value of the peak width at base (Wb).


S At 2At
R = (2)
s Wbl + Wb2 Wbl + Wb2
2


For two closely spaced peaks, one can assume that the two

peak widths are the same, and Wb2 can be used instead of the

mean.



Efficiency (E)

Efficiency of chromatography, E, is defined as peak

retention time divided by peak width at base [211.


tR
E = (3)
Wb


The theoretical plate number, N, contains the same informa-

tion as E.



N = = 2 16[ = 16E2 (4)

or


N = 5.54 (5)
W h


where a and Wh are the standard deviation and peak width at

half height of the peak, respectively. An equation derived





12

by Foley and Dorsey [22] can be used for the calculation of

the number of theoretical plates of skewed peaks in a chro-

matographic system (Nys).
sys

41.7(tR/W 0.1)2
N = (6)
sys B/A + 1.25


where tR, W0.1 and B/A are retention time, peak width at

10% of peak height, and asymmetry factor, respectively.



Control of Separation in Liquid Chromatography

The key to separating components of a mixture is to

control resolution.


k'
R = (a-1)/ 2- (7)



An increase of separation factor, a, which is the ratio of

two capacity factors, k'2/k'1, results in a displacement of

one band center relative to the other and a rapid increase

in Rs. Increasing the number of theoretical plates narrows

the bands while increasing the peak height. For early

eluting peaks, an increase in k' can provide a significant

increase in resolution, however, with increasing k', band

height decreases and separation time increases.

The available options for increasing, a, in order of

decreasing utility are: change of stationary phase, change

of temperature, and special chemical effects. Increasing

the number of theoretical plates can be done by increasing

column length and decreasing flow rate for a given column.





13

Capacity factor can be increased by increasing the volume

of stationary phase, decreasing the strength of mobile

phase, and decreasing temperature.



Gradient Elution (GE)


The most convenient way to separate a complex mixture

of solutes is to use gradient elution. Gradient elution in

LC is similar to temperature programming in GC, except that

the composition of the mobile phase is changing during

separation time. To do GE, one needs at least two different

solvents (binary gradient). One of the solvents has higher

eluent strength, and usually its percentage increases during

the gradient. Ternary gradients using three solvents are

sometimes used in LC. Multisolvent gradients are rarely

required in LC, and because of the complexity, one usually

avoids the use of multisolvent gradients.

The purpose of gradient elution is to resolve early

eluting bands and decrease the retention time of strongly

retained compounds in comparison to isocratic elution. To

achieve this, the gradient must start with a weak mobile

phase, and the strength of the mobile phase increases during

the chromatographic run. Because of decreasing retention

times for late eluting bands, these peaks are greatly

sharpened in gradient elution when compared to isocratic

elution, and sensitivity for these bands is therefore much

improved. Gradient elution increases the peak capacity for

a mixture that contains a large number of individual








components and improves the peak shapes for bands that would

tail in isocratic elution. The gradient shape for binary

solvents can be linear, concave, convex, or any other shape.

The appropriate gradient shape is dependent on LC methods,

sample, and solvent composition. The steepness of the gra-

dient is the mobile phase strength change with time.



Limit of Detection


The limit of detection represents the ability of an

analytical method for quantification of a chemical component

in terms of concentration or absolute amount. In most ana-

lytical methods, the limit of detection, CL, is defined

according to the following equation.


KSb
C K (8)
L m


where Sb and m are the standard deviation of the blank and

slope of the calibration curve, respectively. On a statis-

tical basis, K = 3 is the most appropriate number for cal-

culation of the detection limit. For more information about

detection limits, one is referred to articles by Kaiser [23]

and Winefordner [24].

To calculate the detection limit in LC, the peak to

peak noise is measured while mobile phase passes through the

column. In a normal distribution, peak to peak or random

noise in LC is considered to be 5 times the standard devia-

tion of the blank, as in equation (8). Detection limit in







LC is usually defined as 3 times the peak noise or 3/5 of

the peak to peak noise divided by the slope of the calibra-

tion curve. The latter was used for reporting detection

limits in this dissertation.



Principles of Electrochemical Detection


The most common detector in LC/EC is the amperometric

detector which measures the current at constant potential.

The amperometric detector is more sensitive and less complex

than the coulometric detector. In a coulometric detector,

the amount of electricity for complete electrochemical con-

version of the analyte is measured at constant potential.

The lower sensitivity of the coulometric detector is due to

geometrical requirements necessary for complete electrochem-

ical conversion in a flowing stream. The requirement of

larger surface area of working electrode causes higher back-

ground current and noise which reduces the signal to noise

ratio in comparison to amperometric detector.

The technique of electrochemical detection is based

on electroactivity of components eluted from the column.

Sometimes it is possible to detect nonelectroactive compo-

nents by pre- or post-column derivitization. Selecting the

applied working electrode potential is the primary require-

ment in amperometric detection. The applied potential

should be held at the minimum value at which the current

reaches the limiting-current plateau of the analyte

(Eplateau), however, in most of the cases, Eplateau is







different for different components of a mixture to be ana-

lyzed. In this case the analyst should choose the optimum

applied potential to analyze all components of interest in

the mixture. The applied or analytical potential for an

analyte can be determined by a hydrodynamic voltammogram

(HDV). In an HDV, current is measured versus applied poten-

tial for analyte injected into an LC amperometric detection

system. The Eplateau can be precisely identified by HDV

measurement under analytically useful LC conditions, but it

is a time-consuming process, due to the time required for

the baseline to stabilize after each change of electrode

potential. The time required for stabilization of the base-

line is dependent upon the mobile phase composition and flow

rate. In the case of a glassy carbon electrode, for a change

of 0.1 V in applied potential, it takes 15 to 30 min to get

a stable baseline. Cyclic voltammetry (CV) is a much faster

method for determining Eplateau, which usually has a higher

magnitude than the CV peak potential (E ) under typical

measurement conditions for slow electron transfer reaction.

Data from CV and Eplateau can be related via a simple

empirical equation [25].



Ion-pair Chromatography


The extraction of ionized solutes into organic phases

has been well known for a number of decades. To extract

ionized species from aqueous solution, an ion-pairing reagent

of opposite electrical charge is added to the aqueous phase







resulting in ion-pairing between the solute ion and pairing

ion. The resultant complex which has a net low electrical

charge or polarity can easily be extracted by an organic

phase. To separate ionic species by reversed-phase HPLC,

an ion-pairing reagent can be added to the mobile phase.

Ion-pairing reagents can also be used as a probe to detect

and quantify compounds which cannot be directly detected

[26,27]. The following scheme shows overall phase transfer

of ion pairs.


A+ +B I (A,B)
aqueous organic
phase phase


In the case of reversed-phase chromatography, the organic

and aqueous phases on the above scheme are considered to be

stationary and mobile phases, respectively. The most popular

ion-pair reagents for cationic solutes are long-chain alkyl

sulfonate ions which are usually added to the mobile phase

to enhance separation of oppositely-charged sample ions.

The exact ion-pairing mechanism for the separation of

ionic samples is still uncertain. Three popular hypotheses

are: (1) the ion-pair model, (2) the dynamic ion-exchange

model, and (3) the ion-interaction model. The ion-pairing

model stipulates that the formation of an ion-pair occurs in

the aqueous mobile phase which is in agreement with solvo-

phobic theory [28], while the dynamic ion-exchange model

States that unpaired lipophilic alkyl ions adsorb onto the

nonpolar stationary phase, causing the column to behave as







an ion exchanger [28]. The ion-interaction model is based

upon conductance measurements. It proposes that neither the

ion-pairing nor the ion-exchange model can explain the ex-

perimental data in a consistent way [28]. The ion-interac-

tion model assumes that a primary layer of lipophilic ion

covers the surface of the stationary phase which is in

dynamic equilibrium with the bulk eluent. In the vicinity

of this primary layer exists a secondary layer of opposite

charge creating an electrical double layer on the surface.

The retention of the ionic components is due to the electro-

static force between these ions and the primary layer, as

well as an additional sorptionn) effect onto the nonpolar

stationary phase.



Micellar Chromatography


It is well known that surfactants, detergents, or

surface active agents are amphiphilic molecules (i.e.,

molecules in which a hydrophobic tail is joined to a hydro-

philic head-group). Surfactants can be anionic, cationic,

nonionic, and zwitterionic. Above a certain concentration,

surfactant molecules associate in aqueous solution to form

large molecular aggregates of colloidal dimensions termed

micelles. The concentration threshold at which a surfac-

tant starts to form micelles is called the critical micelle

concentration (CMC), and the number of surfactant molecules

in a micelle is called the aggregation number. The aggrega-

tion number and the CMC differ from one surfactant to







another, and even for the same surfactant in different media.

At concentrations greater than the CMC, a dynamic equilibrium

exists between the surfactant molecules and micelles. The

general size and shape of the particular micelle depend on

the aggregation number.

The term normal micelles is used for surfactant aggrega-

tion in aqueous media. The hydrophilic head groups are

directed toward and in contact with aqueous solution to

form a polar surface, while the hydrophobic tails are

directed away from the water to form a central nonpolar core.

In nonpolar solvents, the surfactant aggregates are termed

reversed or inverted micelles. In these micelles, polar

head groups are concentrated in the interior of the aggre-

gates and hence form a central hydrophilic core, while the

hydrophobic tail moieties extend into and are in contact

with the bulk nonpolar solvent.

The solubilizing power of micellar systems is one of

its most important aspects. This refers to the ability of

micelles to solubilize a wide variety of solutes that are

insoluble or only very slightly soluble in the bulk solvent

alone. The solubilization of solutes in micellar systems is

a dynamic process and depends upon such factors as the tem-

perature, the nature of solutes, the surfactant concentra-

tion, and the type of micelles. The amount of solute

solubilized is usually directly proportional to the concen-

tration of micelles. The solubilization of a solute at a

micelle site is dependent upon the type of solute and the







nature of the micelle. In a normal micelle, a nonpolar

solute is thought to be located near the center of the

hydrophobic core,while an ionic solute is adsorbed on the

polar micellar surface.

According to the properties of micellar systems men-

tioned above, micellar solutions can be used as mobile

phases in HPLC. The normal micellar solution can be used

as the mobile phase in reversed-phase HPLC, while reversed

micellar solutions are compatible with normal-phase HPLC.

Equations for partitioning behavior of solutes with micellar

mobile phases in LC have been derived by Armstrong and

Nome [29]. From these equations,one can calculate the

partition coefficients of solutes between water and

micelles, between the stationary phase and water, and be-

tween micelles and the stationary phase. One of the draw-

backs of micellar mobile phases in RP-HPLC is its poor

efficiency in comparison to hydroorganic mobile phase,

however, Dorsey et al. [30] showed that by the addition of

3% of propanol and a temperature of about 400C, micellar

mobile phases can approach efficiencies of hydroorganic

mobile phases. Possible advantages of micellar mobile

phases over hydroorganic mobile phase are: (1) the unique

selectivity of micellar mobile phases toward different

types of solutes, (2) the economy when compared to hydro-

organic mobile phases, and (3) the simplicity of purifica-

tion of the crystalline surfactants compared to organic

solvents.













CHAPTER THREE

EXPERIMENTAL



Cyclic Voltammetry Systems


To find an approximate analytical potential for solutes

of interest, a CV-1A cyclic voltammetry instrument and an

electrochemical cell made by Bio Analytical Systems, Inc.

(West Lafayette, Indiana) were used. The working and ref-

erence electrodes were glassy carbon and Ag/AgCl, respec-

tively. Before running the CV experiments, the sample solu-

tions were purged for 20 min with helium. Cyclic voltammetry

was then carried out in an inert helium atmosphere. A

Plotamatic MFE-715 (MFE, Salem, New Hampshire) X-Y recorder

and digital voltmeter were used to record the cyclic voltam-

mograms.



Liquid Chromatography System


The solvent delivery unit used during the chromato-

graphic run was a Waters 6000 A (Waters Associates, Milford,

Massachusetts), an Altex model 322 gradient liquid chro-

matograph with two model 100 A pumps (Altex Scientific,

Berkeley, California), or a Spectra-Physics SP 8700 solvent

delivery system (Spectra-Physics, Santa Clara, California).







The injection valve was either an Altex 210 or Rheodyne 7125

(Rheodyne, Cotati, California) with 5, 10 and 20 pL loops.

Various columns--an Altex Ultrasphere octyl, 250 x 4.6 mm;

an Altex Ultrasphere ODS, 150x 4.6 mm; and Rainin Microsorb

octyl column, 150 x 4.6 mm--were employed.



Electrochemical Detector


The electronic controller was an LC-4 amperometric

controller from Bio Analytical Systems, Inc. or Princeton

Applied Research (PAR, Princeton, New Jersey) model 174

polarographic analyzer. The electrochemical cell used was

either from Bio Analytical Systems, Inc., with glassy

carbon as working electrode, or a porous membrane separator

with gold as the working electrode, a generous gift of

K. A. Rubinson. The latter working electrode was used only

in work of Chapter Four of this dissertation to compare

electrode material, while the former was used from

Chapter Four through Chapter Six. The reference electrode

for both cells was Ag/AgCl from Bio Analytical Systems,

Inc.

The electrochemical cell from Bio Analytical Systems,

Inc. [31], is a thin-layer cell, as shown in Figure 1. The

thin-layer cell, reference electrode compartment, clamp,

and waste and connecting tubes are preassembled as one unit.

Addition of a reference electrode to this unit completes

the detector cell. All of the detector cell components

have been machined to accept standard plastic tube end















Reference
Electrode








Auxiliary
Electrode ,


To Waste-


















Electrode Positioned
Downstream From Inlet
Port


Figure 1. Electrochemical Cell with Glassy Carbon Electrode


From LC
Column







fittings (-28 thread) as supplied by BAS, Altex, LDC,

Omnifit, and others; therefore, the detector cell is directly

compatible with commercial HPLC systems.

The porous membrane separator contained gold as the

working electrode [32]. The 0.5 mm gold wire of 5 cm length

was covered with porous polymer tubing which can be attached

to the outlet of the column in an LC system. The gold elec-

trode covered with porous membrane was cemented into a small

reservoir which contained an external electrolyte solution

(1 or 2 M KCl), auxiliary (Pt), and reference electrodes.

To prevent any damage to the porous membrane gold electrode,

it was soaked in 1 M KCl for several hours before running

mobile phase through it from an LC system.



Pretreatment of Glassy Carbon Electrode


It is well known that the sensitivity of glassy carbon

electrodes decreases with use. Decreasing activity at the

glassy carbon electrode surface can be due to adsorption of

analytes, mobile phase components, and byproducts of redox

reactions onto the electrode surface as well as formation

of electroactive species, such as carbonyl and hydroxy

groups, from components in the electrode material. The

adsorption of these components at the electrode surface may

form a polymeric film which would decrease electrode re-

sponse; however, whatever the source of deactivation might

be, it is necessary to clean and reactivate the electrode

surface by mechanical, chemical, electrochemical, or by a






combination of these methods. More details on pretreatment

and the study of the glassy carbon electrode surface are given

elsewhere [33-35].

A deactivated electrode surface produces a large resid-

ual current as well as a noisy baseline which affects the

accuracy of analytical application and increases the detec-

tion limits. For reactivation or pretreatment of a glassy

carbon electrode surface the following procedure has given

satisfactory results. This procedure would be used if

analytical potential is positive or anodic.


(1) The cell was dismounted and electrode surface was
washed with methanol.

(2) A few drops of a slurry of alumina [0.1 im, Gamal,
Grade B, Fisher Scientific Company (Fairlawn,
New Jersey)] in water were placed at the surface
of the electrode. The surface was then carefully
polished with Buehler LTD polishing paper (Evan-
ston, Illinois) for 2-3 min.

(3) The electrode surface was rinsed with methanol and
polished with the same polishing paper which was
soaked in methanol, and then the cell was assem-
bled.

(4) An anodic potential (0.1-0.2 V higher than that
of analytical potential) was applied for 3-4 min
while mobile phase passed through the electrochem-
ical cell.

(5) A cathodic potential (-0.8 V vs. Ag/AgCl) was
applied for 2-3 min while mobile phase passed
through the electrochemical cell.

(6) The steps of (4) and (5) were repeated, and then
the potential was set at the analytical potential.
The potential switch was turned off before chang-
ing from the anodic-applied potential to the
cathodic-applied potential or vice versa.







Reagents


Methanol and acetonitrile, HPLC grade (Fisher Scientific

Company, Fairlawn, New Jersey) were employed as organic

mobile phase components. Methanol, HPLC grade, was used as

the solvent for test solutes. Deionized water, another

mobile phase component, was purified by a Barnstead Nanopure

system (Sybron Corp.) The purified water then was irradi-

ated for 24 hr by ultraviolet light in a model 816 HPLC

reservoir (Photronix Corp., Medway, Massachusetts). The

other chemicals comprising the mobile phases were: (1) rea-

gent grade phosphoric acid (Scientific Products, McGaw Park,

Illinois), (2) sodium perchlorate and Brij-35 (Aldrich

Chemical Co., Milwaukee, Wisconsin), and (3) sodium dodecyl

sulfate (SDS; Fisher Scientific Co.). The reagents mentioned

above mostly were used as a component of the mobile phase

or dissolved in the mobile phase. Other specific chemicals

(e.g., sample solutes) will be mentioned in chapters related

to them.













CHAPTER FOUR

SIGNAL CHANGE AND BASELINE SHIFT
USING ELECTROCHEMICAL METHODS



In this chapter the following will be considered:

(1) change of residual current with ordinary DC voltammetry,

normal pulse (NP) voltammetry, and differential pulse (DP)

voltammetry in isocratic elution HPLC; (2) effect of the

background electrolyte and gradient elution on baseline

shift using amperometric detection; and (3) effect of the

electrode material on baseline shift using amperometric and

non-ramping differential pulse (NRDP) voltammetry methods.

Pulse techniques such as DP and NP can be used as LC

detection schemes if a constant pulse amplitude superimposes

over an initial potential. This means that there is no

potential ramp, and it is better to be called non-ramping

pulse techniques. The term "pulse techniques" has been

used instead of the term "non-ramping pulse techniques" in

the literature, however. Fleet and Little [36] were among

the first authors to be concerned with application of

pulse techniques as an HPLC monitoring method using a mer-

cury drop electrode (MDE). The idea was examined in prac-

tice by Swartzfager [37], using carbon paste as a working

electrode. The advantages of pulse systems in monitoring

have been discussed by Kissinger [19], who noted the







possibility of separating the monitored electrochemical sys-

tem from systems with more positive or negative half-wave

potentials and prolonging electrode service life by electro-

chemical cleaning. Electrochemical cleaning during the

pulse operation is the subject of a patent by Fleet [38]. A

versatile voltammetric detector with double polarization

pulse and semi-differential scanning for LC is reported by

Stastny et al. [39]. Pulse techniques also offer the advan-

tage of decreased dependence of the measured current on

flow-rate in comparison to amperometric techniques [37],

because the short potential pulse duration minimizes the

development of the diffusion layer of the electrode and

dependence of the thickness of this layer on flow-rate.

With all the advantages of pulse techniques (more selectiv-

ity, increased electrode stability, and less current depen-

dence on flow-rate) over amperometric techniques, these

techniques are limited by poor sensitivity and complex in-

strumentation. The residual current at a solid electrode is

reportedly high [17]. A study of the change in magnitude of

the residual current with different electrochemical methods

using isocratic elution and with amperometric method using

gradient elution was performed and will be reviewed. The

residual current change during gradient elution with ampero-

metric detection is a baseline shift due to the change in

the composition of the mobile phase.

The electrochemical equipment used was a model 174

polarographic analyzer, an LC-4 amperometric controller, and







an electrochemical cell either with a glassy carbon elec-

trode (GCE) or a porous membrane gold (Au) electrode.



Change of Residual Current with DC, DP and
NP Voltammetry in Isocratic Elution HPLC


In differential pulse (DP) voltammetry and normal pulse

(NP) voltammetry, the current is measured during a time in-

terval of the pulse when the ratio of faradaic current to

charging current is a maximum. The electrochemical cell

can be considered as an RC-circuit, and thus the charging

current decays exponentially. The purpose of these experi-

ments was a comparison of the residual current change versus

applied electrode potential with different electrochemical

methods using isocratic elution.

The residual current change and decay current on a

glassy carbon electrode with different percentage of acetoni-

trile as an organic modifier in water are shown in Figures 2

through 5. The data from these figures are reported in

Table 1. All mobile phases contained 0.04 M NaClO4 as a

background electrolyte. The potential was scanned from

0.55 V (E ) to 1.30 V (E ) versus Ag/AgCl reference elec-

trode.

The residual current change, AI = If Ii, is due to

the difference between the current at the final potential

(Ef) or the current at the end of the ramp, If, and the cur-

rent at the initial potential, I.. The residual current

change, AI, can be considered to be due to the residual

current difference at two different potentials. Decay













"--------

o\
uL


In z
- u


U




.J




C,'
uJ







< 0A


ra 0
) c







0 -

1 4
7n I





4 r4

tr
U 0 0 r
u 00





O O

r- O



a 0
In a)

a Q Oc


( e

n LU














O C.,
*0 *0I


Z



U

o\O
0





*-P

















)-1
c
u












LLU



-P




4.)
-1
U








W (U
C,



















UL

4-1


in










V.-


















in)

0
























Z
LU







U



a0

c4-








U
I,-I










i i0









r=r r4

4. 0 0 )

CO4 -
eC






C5 rq
F EE O
0
Cm
C)





SOP
0 QQ
< E-





(Q g g Q
^! f0 n3l-
U -i i- (


















I----i


>1







0
--

o


z

>1

<
t Cdn



C'



> M

LU U


:3

cn
G)
P:







current is the amount of current depleted at the end of the

ramp (Ef) within a short time. This current depletion might

be considered as the charging current and decreasing current

due to impurities. Since the charging current decays quick-

ly (10-4 to 10-3 sec), it is impossible to find out the

origin of decay current with a stripchart recorder.

The data in Table 1 show that on the DP mode the resid-

ual current change, AI, and decay current, idecay are in

the limit of measurement error (measurement device was a

ruler). This means that the currents sampled shortly before

the end of the pulse and shortly before the rise of the

pulse are almost the same, and no difference in charging

current is sampled. On the DC mode, there is some current

change during the scan (potential applied to the electrode

changes with time) and some decay current at the end of the

scan. Increasing the scan rate will increase both AI, and

idecay in the DC mode because the magnitude of charging

current increases as the scan rate increases. In the NP

mode, AI is much larger than that of the DC mode, while the

decay current is in the limit of measurement error. The

current varies linearly with the applied potential on the

working electrode (WE) in the NP mode. The larger magnitude

of AI in the NP mode as compared to the DC mode might be due

to the surface clean up of the working electrode which

occurs with pulsing. This cleanup increases the electrode

response.








Table 1.


Residual Current Change, AI, and Decay Current,
decay, in DC, DP and NP Voltammetry in Isocratic
Elution. Flow rate: 1 mL/min; column: Altex Ultra-
sphere ODS 150 x 4.6 mm; Ei = 0.55 V, Ef = 1.30 V
vs. Ag/AgCl; working electrode: glassy carbon; low
pass filter = 0.3 sec; pulse amplitude = 100 mV.


Solvent A =
Solvent B =


0.04 M NaC104 in water.
0.04 M NaC104 in acetonitrile.


Scan
Rate
mV/sec


Electrochemical
Mode

DC
DP
DC
DC
DP
DC
DP
NP


Figure


Mobile
Phase


2-a
2-b
3-a
3-b
3-c
4-a
4-b
5


5% B
5% B
50% B
50% B
50% B
95% B
95% B
95% B








Table 1 Extended


Drop Knocker
Time
(sec)

1
1
1
1
1
0.5
0.5
0.5


AI Measure-
ment Error
(iA)

0.75 0.02
0.02
0.65 0.01
0.82 0.01
0.01
1.41 0.02
0.02
7.36 0.08


decay measure-
ment Error
(pA)


0.16


0.12
0.34


0.10

+


0.02
0.02
0.01
0.01
0.01
0.02
0.02
0.08








Effect of Background Electrolytes and Gradient
Elution on Baseline Shift Using Amperometric Detection


In isocratic elution, the composition of the mobile

phase during the course of separation is constant; however,

isocratic elution sometimes is incapable of separating com-

plex samples which have a wide range of retention times or

k' values. One of the solutions to this problem is gra-

dient elution (GE). The necessity (of the presence) of a

background electrolyte and changing composition of the

mobile phase causes changes in viscosity, dielectric constant,

and conductivity of the mobile phase. The dependence of

the diffusion coefficient on viscosity, variation of the

conductivity, and changing condition of the working elec-

trode surface during gradient elution will cause a change in

the residual current.

Walden's rule [40] states "the product of the equiva-

lent conductivity (X) and viscosity of the solvent (n) for

a particular electrolyte at a given temperature should be

a constant."


S* n = constant (9)


In a binary gradient (solvents A and B), the percentage of

organic modifier (solvent B) increases in hydroorganic mo-

bile phase during the gradient course. Decreasing the po-

larity of the mobile phase during the gradient course

decreases the conductivity of the mobile phase if the con-

centration of background electrolyte in both solvents is





38

the same. If the concentration of background electrolyte

in solvent B is higher than that of solvent A, the total

concentration of background electrolyte in mobile phase

increases during the gradient course. An increase in the

concentration of background electrolyte during the gradient

course may compensate for decreasing the polarity and chang-

ing the viscosity of the mobile phase, which are important

factors in baseline stability. As a result of this discus-

sion, a more stable baseline should be achieved with higher

concentration of background electrolyte in solvent B as

compared to equal concentration of background electrolyte

in both solvents.

If one assumes a binary gradient in which the percen-

tage of solvent B increases during the gradient course, one

can measure the conductivity or specific conductivity vs.

the percent of solvent B in A+B. The specific conductivity

vs. the percentage of organic modifier (solvent B) is shown

in Figure 6 for H3PO4 and NaClO4 as background electrolytes.

The conductance measurement was done with a conductance

bridge of the Janz-McIntyre type [41]. The specific conduc-

tivity, K, was calculated from using equation (10).


0.126
K = specific conductivity = R (10)


The cell constant is 0.126 cm-1, and R can be calculated

from equation (11).


Rmeasured 10000
R 10000 R (11)
measured

















40 1E-4


* d ~.. __


,
..
4
.ni.i
F- -1


Figure 6.


Specific Conductance vs. Percentage of Solvent B


(a) Solvent A = 0.15% H3PO4 in H20

Solvent B = 0.20% H PO4 in CH OH


(b) Solvent A = 0.05 M NaCIO4 in H20

Solvent B = 0.05 M NaCIO4 in CH30H


c -4







In the case of H3PO4 as a background electrolyte, the spe-

cific conductance decreases with the increasing percentage

of the organic modifier, while in the case of NaClO4 it

passes through a minimum at about 65% of organic modifier

(solvent B).

Figure 7 shows the baseline shift during gradient elu-

tion (with no sample injection) using an amperometric detec-

tor and glassy carbon electrode, with H PO4 and NaClO4 as

the background electrolytes. The column is equilibrated

with solvent A, and gradient starts with holding solvent A

for 3 min, then linear ramp to 20% A and 80% B over 30 min.

The baseline shift, Ai, is the difference in current before

starting the gradient program and at the end of the gradient.

Figure 7-a is due to 0.15% H3PO4 in H20 and 0.20% H3PO4 in

CH3OH, with Ai = -9 n A, while Figure 7-b is due to 0.20%

H3PO4 in both H20 and CH3OH with Ai = -14 n A. The differ-

ence in Ai for Figures 7-a and 7-b is due to the concentra-

tion difference of H3PO4 in H20. The lower baseline shift

in Figure 7-a as compared to Figure 7-b is in agreement with

Walden's rule as discussed above. Figure 7-c is due to

0.05 M NaClO4 in both H20 and CH3 OH, with Ai = +6 n A.

Sodium perchlorate produces a smaller baseline shift because

it is completely ionized in both solvents. The peaks in

Figure 7 might be due to impurities in H20 or background

electrolytes which collect at the top of the column during

the early part of the solvent program and equilibration.

More examples of gradient elution with hydroorganic and
























04




Z

QC
0
z

-4-)



0 Q)
4J)

-4



Wa)
rOO


00-'
c u

0 0)



C -H
C Q








.)l



M .
(U) 0
**-I 4-
r-l U
m) 0
cn r-


I
a)
a-,









0 >
o>
0


a).
4** o





) 0

4J

En

e m

M,-
o o





(13 l4
4 4


0
0
CN
3 U
C C


0 0
r-H i-i
U U
z z
x Z


Ln Ln
o o

0 0
0o

II II







0 0
C1n cl


0
o















II
m
C,


0
Cl,





o






0
c














c

































C\j
O0

-0
o











































m








-in








micellar mobile phases are given in Chapters Five and

Six.

The shift in baseline during gradient elution at a con-

stant applied potential might be due to different concentra-

tions of electroactive impurities, changing diffusion

coefficient of impurities with a change in the viscosity,

exhibition of different half-wave potential for specific

electroactive components, and effect of organic modifier on

the surface of the working electrode.



Effect of the Electrode Material on Baseline Shift
Using Amperometric and NRDP Voltammetry Methods


The baseline shift during gradient elution is evaluated

for glassy carbon and porous membrane gold electrodes. The

electrochemical methods used were amperometric and non-

ramping differential pulse (NRDP) voltammetry. In NRDP

method, the difference in current sampled at the end of the

pulse and shortly before the rise of the pulse is recorded.

Since the pulse time is very short, there is virtually no

change in the mobile phase composition with gradient elution

during one pulse. The normalized baseline shift (current

difference at the end and the start of gradient program

divided by the surface area of the working electrode) for

amperometric and NRDP voltammetry with their chromatographic

conditions are reported in Table 2.

Data in Table 2 show that the absolute amount of the

baseline shift in the amperometric method is much higher

for gold electrode as compared to glassy carbon electrode.








Table 2.


Normalized Baseline Shift (nA/mm2) during Gradient
Elution with No Sample Injection. Column: Altex
150x4.6 mm Ultrasphere ODS; E = 0.8 V vs. Ag/AgCl;
pulse amplitude = 100 mV; low pass filter = 3 sec.
Gradient program: initially 95% A, then linear
ramp to 5% A over 20 min, and finally holding for
5 min at 5% A.


Solvent A = 0.20% H PO4 in H20.
Solvent B = 0.20% H3PO4 in CH3OH.


Electrochemical
Method

Amperometric

NRDP

Amperometric


NRDP


Working
Electrode


Normalized Base-
line Shift
nA/mm2


Glassy Carbon

Glassy Carbon


Gold

Gold


-1.5


-116.0

-356.7

-152.8





45

This indicates that the residual current change with gradient

elution on the gold electrode surface is larger as compared

to glassy carbon electrode, which shows the gold is not an

appropriate working electrode under these conditions.

Although the absolute amount of the baseline shift in the

case of NRDP is also larger for the gold electrode as com-

pared to glassy carbon electrode, this gap is not as large

as in the case of amperometric method. This might be due to

cleanup of the electrode surface with pulsing.












CHAPTER FIVE

A COMPARISON OF MICELLAR AND HYDROORGANIC
MOBILE PHASES USING AMPEROMETRIC DETECTOR



In this chapter, the following will be considered:

(1) the comparison of the hydrodynamic voltammograms (HDV)

in micellar and hydroorganic mobile phases, (2) the compar-

ison of analytical figures of merit between micellar and

hydroorganic mobile phases, and (3) gradient elution and

selctivity with micellar mobile phases.

Ionic sufractants have been extensively used as ion-

pairing reagents in ion-pair chromatography in the past.

More recently, ionic surfactants and nonionic surfactants in

aqueous solution have been employed as the mobile phase in

RP-LC. The surfactant concentrations in the mobile phase

above the CMC have been shown to have properties similar to

conventional mobile phases for RP-LC [42]. Micellar solu-

tions have been used as media or matrices for room tempera-

ture phosphorescence [43], and the usefulness of micelle-

stabilized room-temperature phosphorescence (MSRTP) for

detection and quantification of aromatic molecules in HPLC

has been reported by Weinberger et al. [44]. Micellar solu-

tions have been used in electrochemistry to produce well-

defined redox waves for compounds that show only slight

shoulders or no waves in aqueous solutions [45]. Since the








redox voltammetry waves for organic compounds in micellar

solutions can be different from those of hydroorganic solu-

tions, three different classes of compounds were chosen in

this study. The structures of these three classes which

include various forms of B-6 vitamins (PLP,PL, PN, PNP and

PMP), phenolic compounds (phenol, hydroquinone, resorcinol,

catechol and o-cresol), and polyaromatic hydrocarbons

(anthracene and pyrene) are shown in Figure 8. The phos-

phate group in three of the B-6 vitamins can be partially

ionized. The extent of ionization is dependent upon the pH

of the mobile phase, however, in the low pH region, all

forms of B-6 vitamins will be protonated to produce cations.

A phosphoric acid buffer solution of pH 2.20 was used as the

background electrolyte in all mobile phases and in addition

generated an ionic site for ion-pairing of the nitrogenous

vitamins. With cationic solutes, sodium dodecyl sulfate

(SDS) in the mobile phase can act as an ion-pairing reagent

both below and above the CMC.



Hydrodynamic Voltammoqram in Micellar
and Hydroorganic Mobile Phases


Hydrodynamic voltammograms for phenol, two vitamins

(PLP and PL), and polyaromatic hydrocarbons (anthracene and

pyrene) with their experimental conditions are given in

Figures 9 through 11, respectively.

The signal achieved for the same amount of injected

phenol in the micellar mobile phase is greater than that of

the hydroorganic mobile phase when the applied potential is










HC=O


HO H CH: 0


H.C N H

Pyridcxal-5- phosphcae

PLP


pyridoxol

PL


CHo CH


CH2CH
CHtCH HO CH, O?-CH
Ii
YH 2 OH

H C N H

Pyridoxin.-5- phosp.c.a

PNP


U CHO -CP-OH
'I !OH

N N H H, C N H

Pyridoxina
Pyridoxcmine-5- phcsphc',a

PN PMP

Structures of B-6 Vitamins


on


01
OH


Hydroquinone

Structures


OH



o OR


OH


N^ ^


Resorcinol catcchol

of Phenolic Compounds


~8 ~
.7 2
6
N.5 NU ~


7 a


Anthracene Pyree
Pyrene
Structures of Polyaromatic Hydrocarbons



Figure 8. Chemical Structures of Compounds Used in Chapter
Five


0:I






Phenol


OH


o-Cresol
























Figure 9. Hydrodynamic Voltammogram for Phenol


Column: Altex Ultrasphere ODS, 150 x 4.6 mm
Flow Rate: 1 mL/min
Injection Volume: 20 pL (5 ppm)
Working Electrode: Glassy carbon electrode
Temperature: 300C


(o) Micellar mobile phase: 0.1 M SDS in
(3:97) (l-propanol:H20) + 5 x 10-2 M
NaClO4
(*) Hydroorganic mobile phase: (60:40) (H20:
CH OH) + 5 x 10-2 M NaCIO4
































































0,8 0,9


1.0 1, 1.2
E(v)


250 r


200 -


i(nA)


150 I


100 I


50 L
0.


7


I I I






















Figure 10. Hydrodynamic Voltammogram for B-6 Vitaminsa


Micellar mobile phase: 0.1 M SDS in (3:97)
(l-propanol:H20), pH = 2.20 (used
H3P4)

(A) PLP (250 ppm)
(0) PL (250 ppm)


Ion-pair mobile phase: 5 x 10-3 M SDS in
(12:88) (l-propanol:H20), pH = 2.20
(used H3PO4

(o) PLP (2000 ppm)
(*) PL (2000 ppm)

a All B-6 vitamins dissolved in 0.1 M HC1. Other con-
ditions as in Figure 9.

















1000 -







750 -





/
i (nA) 500 -

"I /





250 -







1,0 1.1 I.2 1.3 1.4
E (v)



















Figure 11. Hydrodynamic Voltammogram for Polyaromatic
Hydrocarbons


Column: Rainin microsorb octyl
Flow Rate: 2.5 mL/min


Micellar mobile phase: 0.1 M SDS in (3:97)
(l-propanol:H20), pH = 2.2 (used H PO4

(o) anthracene (50 ppm)
(*) pyrene (100 ppm)


Hydroorganic mobile phases:
(0) anthracene (50 ppm): (46.5:53.5)
(H20:CH3OH), pH = 2.20 (used H3PO4)

(A) pyrene (100 ppm): (43.2:56.8) (H20:
CH3OH), pH = 2.20 (used H3PO4)


a Anthracene and pyrene were dissolved in CH3OH. Other
conditions as in Figure 9.


































































0.8 0,9


iO I1I 1,2 1.3 1.4
E v)


500


400









300


i(nA)


200 -









00









0-
0,7







below 1.0 V, whereas above 1.0 V it is slightly greater in

the case of the hydroorganic mobile phase. According to

HDV, the best analytical potential for phenol is 1.1 V in

both mobile phases.

In the case of B-6 vitamins, the signal increases with

increasing potential in the same manner for both the micel-

lar mobile phase and the hydroorganic mobile phase which

contains small amounts of SDS as an ion-pairing reagent. The

signal for PLP and PL are greater in the micellar mobile

phase as compared with the ion-pairing mobile phase. At

first glance, the graph in Figure 10 shows that the signal

for PLP in the ion-pairing mobile phase is greater than that

in the micellar mobile phase at the same potential, however,

this is not the case because the absolute amount of injec-

tion of PL and PLP in the ion-pairing mobile phase is 8 times

greater than that in the micellar mobile phase. The poten-

tial 1.4 V might be chosen as an analytical potential for

vitamins.

Figure 11 shows the HDV for anthracene and pyrene in

both mobile phases. At potentials higher than 1.2 V, the

signal due to pyrene increases sharply, whereas the signal

due to anthracene decreases smoothly. One of the most im-

portant aspects of Figure 11 in comparison to Figures 9 and

10 is the unique selectivity response of polyaromatic hydro-

carbons with applied potential in both micellar and hydro-

organic mobile phases. This means that anthracene can be

measured in the presence of pyrene using a proper applied

potential. In order to compare analytical figures of merit







in micellar and hydroorganic mobile phases, the best possible

analytical potential to use for anthracene and pyrene would

be 1.1 V and 1.2 V, respectively.



Analytical Figures of Merit Comparison between
Micellar and Hydroorganic Mobile Phases


Analytical figures of merit such as limit of detection

(LOD), upper limit of linear dynamic range (LDR), sensitiv-

ity, correlation coefficient, and log-log slope, as well as

the experimental conditions for phenol, B-6 vitamins, and

polyaromatic hydrocarbons (PAH) in both micellar and hydro-

organic mobile phases, are reported in Tables 3 through 5,

respectively. Data in Tables 3 through 5 are derived from

Figures 12 through 14 (analytical or calibration curves),

respectively.

In order to compare analytical figures of merit for a

specific electroactive component in two different mobile

phase compositions using an amperometric detector, two con-

straints must be met. First, the flow rate should be nearly

the same for both mobile phases because the response of any

electrochemical detector is dependent on the rate of mass

transfer to the electrode surface [37]. Secondly, the

retention time of the electroactive component in both mobile

phases should be nearly the same because of the peak height

measurement. To get nearly the same retention time in both

mobile phases, the percentage of organic modifier in the

hydroorganic mobile phase was varied until the retention







Table 3. Analytical Figures of Merit for Phenol. Flow
rate: 1 mL/min; column: Altex Ultrasphere ODS,
150 x 4.6 mm; E = 1.1 V; working electrode:
glassy carbon electrode; injection volume: 20 lL.


Micellar mobile phase: 0.1 M SDS in 3% 1-propanol
+ 5 x 10-2 M NaCIO4.
Hydroorganic mobile phase: (H20:CH3OH)(60:40) +
5 x 10-2 M NaClO4.


Limit of Detection Upper limit of LDR
Mobile Phase (ppm) (ng) (ppm) (ng)

Micellar 0.0080 0.16 20a 400

Hydroorganic 0.0065 0.13 10 200


a 20 ppm was the most concentrated solution, so the upper
limit of LDR might be more than 20 ppm.







Table 3 Extended


Sensitivity
(nA/ppm) (nA/ng)

41.40 2.03

39.74 1.99


Correlation
Coefficient

0.9971

0.9973


Log-log
Slope

0.96

1.04


tR
(min)

5.0

3.8








Table 4.


Compound


Analytical Figures of Merit for B-6 Vitamins.
Column: Altex Ultrasphere ODS, 150 x 4.6 mm;
E = 1.4 V; working electrode: glassy carbon;
injection volume: 20 pL.


Micellar mobile phase: 0.1 M SDS in 3% 1-propanol
with pH = 2.20 (used H3P04).
Ion-pair mobile phase: 5 x 10-3 M SDS in 12%
1-propanol with pH = 2.20 (used H3PO4).


Mobile
Phase


Limit of Detection Upper Limit of LDR
(ppm) (ng) (ppm) (nq)


Micellar

Micellar


Ion-pair


PL Ion-pair


PLP

PL


PLP


0.316

0.236


0.0726


6.32

4.72


1.45


300

300


100


6000

6000


2000


0.0852 1.70


100 2000








Table 4 Extended


Sensitivity
(nA/ppm) (nA/ng)

3.80 0.170

5.09 0.255


4.96

4.22


0.248

0.211


Correlation
Coefficient

0.9999

0.9995


0.9999

0.9999


Log-log Flow Rate
Slope (mL/min)

0.99 1.0

1.00 1.0


0.95

0.95


1.1

2.0


tR
(min)

2.8

7.8


2.8

7.8





61


Table 5. Analytical Figures of Merit for Polyaromatic Hydro-
carbons. Flow rate: 2.5 mL/min; column: Rainin
microsorb octyl, 150 x 4.6 mm; injection volume:
20 jiL.


Micellar mobile phase: 0.1 M SDS in (3:97) (1-pro-
panol-H20), pH = 220 (used H3PO4)


Limit of
Detection


Compound


Anthracene


Pyrene


Anthracene


Pyrene


Mobile Phase


Micellar

Micellar


Hydroorganica

Hydroorganic


(ppm) (ng)

0.0933 1.87

1.77 35.4


0.0846 1.69


Upper Limit
of LDR
(ppm) (ng)


250

200


5000

4000


50 1000


a Hydroorganic mobile phase (used only for anthracene):
(46.5:53.5) (H20:CH30H); pH = 2.20 (used H3PO4); ionic
strength = 3.59 x 10-2 M.

b Hydroorganic mobile phase (used only for pyrene):
(43.20:56.80) (H20:CH3OH), pH = 2.20 (used H3PO4).





62

Table 5 Extended


Sensitivity


(nA/ppm)


0.707

0.667


3.545


(nA/ng)

0.0354

0.0338


0.177


Correlation
Coefficient

0.9996

0.9990


0.9906


1.2 9.3


Log-log
Slope

0.97

0.97


0.96


E
(V)

1.1

1.2


1.1


tR
(min)

13.7

9.3


13.7
















1000









750
/







500









250









4 8 12 15 20

C(ppm)

Figure 12. Analytical Curves for Phenol

(o) in micellar mobile phase
(*) in hydroorganic mobile phase

Experimental conditions are the same as Table 3.


























Figure 13. Analytical Curves for B-6 Vitamins

Micellar mobile phase:
(A) PLP
(0) PL

Ion-pair mobile phase:
(o) PLP
(*) PL

Experimental Conditions are the same as
Table 4.






















2000 r


1500 H


i(rnA)


1000 I


500 -


200

C(ppm)


300


400


vI ii






















Figure 14. Analytical Curves for Polyaromatic Hydrocar-
bons


Micellar mobile phase:
(o) anthracene
(*) pyrene


Hydroorganic mobile phase:
(0) anthracene


Hydroorganic mobile phase:
(A) pyrene


Experimental conditions are the same as
Table 5.



















400 r


300 1


i(nA)


200


100 1


C(ppm)


200


300







time of the electroactive component nearly matched that of

the micellar mobile phase.

In the case of phenol, the LOD is approximately the

same in both mobile phases, whereas the upper limit of the

LDR in the micellar mobile phase is at least twice that of

the hydroorganic mobile phase. For B-6 vitamins there is

not much difference in the LOD in both mobile phases, while

the upper limit of LDR in micellar mobile phase is 3 times

that of the hydroorganic mobile phase. The LDR for anthra-

cene is 5 times greater in the micellar mobile phase as

compared to the hydroorganic mobile phases, while the LOD

is approximately the same in both mobile phases. It is hard

to determine the precise analytical figures of merit for

pyrene in the hydroorganic mobile phase because, as

Figure 14 shows, the analytical curve is practically non-

linear in that concentration range. Pyrene, however, should

be similar to anthracene, and, as shown in Figure 14, the

upper limit of the LDR for the micellar mobile phase is much

greater when compared to the hydroorganic mobile phase.

The greater upper limit of LDR in the micellar mobile

phase as compared to the hydroorganic mobile phase may be

due to the higher ionic strength (resulting from the pre-

sence of ionic surfactants as well as background electro-

lyte), in the micellar mobile phase [19,46] or due to the

nature of surfactants. Increasing the ionic strength of the

mobile phase decreases the IR gradient across the electrode

face and between the electrode and the bulk solution. Sur-

factants or surface active agents can change the interfacial







properties. Hence, the interfacial properties of the micel-

lar mobile phase with working electrode will be quite dif-

ferent from those of hydroorganic mobile phase.



Gradient Elution and Selectivity in Micellar Mobile Phase


Gradient elution with the micellar mobile phase is

primarily used for the same purpose as it is with the hydro-

organic mobile phase. Gradient elution may also be used to

effect a change in selectivity because micellar mobile

phases have unique selectivities at different concentrations.

Armstrong and Henry [42] have used gradient flow rates with

a UV detector. The gradient micellar chromatogram, obtained

with an amperometric detector using an anionic surfactant

(SDS) and a nonionic surfactant (Brij-35) are shown in

Figures 15 through 17. Figure 15 shows the separation of

B-6 vitamins with SDS, and Figures 16 and 17 show separation

of phenolic compounds with SDS and Brij-35, respectively.

Micellar mobile phases have been shown [47] to offer

control over selectivity in liquid chromatography as the

concentration of surfactant changes in the mobile phase.

The selectivity of the micellar mobile phase toward B-6

vitamins is shown in Figure 18. The elution order of PL and

PMP changes above and below a certain concentration range

of SDS, whereas they coelute in this range. The void volume

of the column, determined by water injection, slightly

decreased as the concentration of SDS increased in the

mobile phase. For an Altex Ultrasphere ODS 25 x 4.6 mm, a












Figure 15. Gradient Micellar Chromatogram for Separa-
tion of Vitamin B-6


Solvent A: (3:97) (1-propanol:H20),
pH = 2.20
Solvent B: 0.2 M SDS in (3:97)(1-propanol:
H20), pH = 2.20 (used H3PO4)


Gradient program: started with 25% solvent
B and held at 25% B for 8 min, then
ramped to 100% B over 8 min and contin-
ued with 100% B


Flow rate: 1 mL/min; temperature: 300C;
working electrode: glassy carbon elec-
trode; E = 1.3 V


Column: Altex Ultrasphere ODS, 250 x 4.6 mm;
precolumn: 15 x 4.6 mm, packed manually
with 25-40 pm silica gel


Injection volume: 10 iL


Peaks are as follows: (1) PLP (200 ppm),
(2) PL (200 ppm), (3) PMP (400 ppm),
(4) PNP (400 ppm), (5) PN (400 ppm)
































200 nA


24 12

t(min)














Figure 16. Gradient Micellar Chromatogram for Separa-
tion of Phenolic Compounds


Gradient program: started with 25% solvent
B and held at 25% B for 4 min, then
ramped to 100% B over 8 min and con-
tinued with 100% B


Injection volume: 13 pL; E = 1.1 V


Peaks are as follows: (1) hydroquinone
(10 ppm), (2) resorcinol (8 ppm),
(3) catechol (20 ppm), (4) phenol
(18 ppm), (5) o-cresol (arbitrary con-
centration)


Other conditions as Figure 15.














































N)


I


8 12 6
t(min)












Figure 17. Gradient Micellar Chromatogram for Separa-
tion of Phenolic Compounds and Gradient
with No Injection


Solvent A: (3:97)(1-propanol:H20),
pH = 2.20 (used H3PO4)
Solvent B: 0.1 M Brij-35 in (3:97)(1-pro-
panol:H20), pH = 2.20 (used H3PO4)


Gradient program: started with 20% sol-
vent B and held at 20% B for 4 min,
then ramped linearly to 100% B over
6 min and continued with 100% B


Flow rate: 1 mL/min; temperature: 300C;
working electrode: glassy carbon
electrode


Column: Altex Ultrasphere ODS, 150 x
4.6 mm


Injection volume: (a) 20 iL; (b) none


Peaks are as follows: (1) hydroquinone
(10 ppm), (2) resorcinol (8 ppm),
(3) catechol (20 ppm), (4) phenol
(18 ppm), (5) o-cresol (arbitrary
concentration); these are for (a)






























4 2 loo nA









5



(a)





(b)

I |i
16 8 0
t(min)




























Figure 18. Effect of SDS Concentration on k'


Mobile phase: (3:97)(1-propanol:H20) with
different concentrations of SDS,
pH = 2.20 (used H PO4)


Flow rate: 1 mL/min; temperature: 300C;
working electrode: glassy carbon;
E = 1.3 V


Column: Altex Ultrasphere ODS, 250 x
4.6 mm


(Q) PLP
(o) PL
(x) PMP
(A) PN and PNP















1,20





1.00 x

0



0.80





0.60





0,40





0.20





0.00 O





-0,20 '''
-1.40 -1.20 -1.00 -0.80 -0.60
log [SDS] (mM)







change in concentration of SDS from 0.05 M to 0.15 M or more

in the mobile phase caused the void volume to change from

2.20 mL to 2.00 mL. The adsorption isotherm of surfactants

on ODS-hypersil is reported by Knox and Hartwick [48]. The

surfactant concentrations chosen were below the CMC and one

SDS concentration supposedly above the CMC was also used.

The adsorption isotherms shown by these researchers are in

methanol:water (20:80) solution which results in a different

CMC from that of pure water. It is doubtful they ever

reached the CMC for their system. There is no literature

adsorption isotherm available for SDS above the CMC.

Figures 19 through 21 show the isocratic micellar chromato-

grams for B-6 vitamins and their selectivity with three dif-

ferent concentrations of SDS in the mobile phase.

















Figure 19. Isocratic Micellar Chromatogram for Separa-
tion of Vitamin B-6


Mobile phase: 0.05 M SDS in (3:97)(1-pro-
panol:H20), pH = 2.20 (used H3PO4)


Flow rate: 1 mL/min; temperature: 300C;
working electrode: glassy carbon elec-
trode; E = 1.3 V


Column: Altex Ultrasphere ODS, 250 x 4.6
mm; precolumn: 15 x 4.6 mm, packed
manually with 25-40 pm silica gel


Injection volume: 10 pL


Peaks are as follows: (1) PLP (200 ppm),
(2) PL (200 ppm), (3) PMP (400 ppm),
(4) PNP (400 ppm), (5) PN (400 ppm)





















S


20
t(min)


4 ;5
1


200oo nA


40


Ji


w-


U


























Figure 20. Isocratic Micellar Chromatogram for Separa-
tion of Vitamin B-6


Mobile phase: 0.1 M SDS in (3:97) (1-pro-
panol:H20), pH = 2.20 (used H3PO4)


Other conditions are as in Figure 19.










4;5


2;3


2oo nA


20 10 0
t(min)




























Figure 21. Isocratic Micellar Chromatogram for Separa-
tion of Vitamin B-6


Mobile phase: 0.2 M SDS in (3:97)(1-pro-
panol:H20), pH = 2.20 (used H3PO4)


Other conditions are as in Figure 19.























L


2oonA


t (min)


4;5


1


i


J












CHAPTER SIX

RAPID SEPARATION AND DETERMINATION OF THYROMIMETIC
IODOAMINO ACIDS BY GRADIENT ELUTION REVERSE PHASE LIQUID
CHROMATOGRAPHY WITH ELECTROCHEMICAL DETECTION



The mechanism of action of the thyroid hormones, T3 and

T4, is of considerable interest in part because of the

amazing diversity of thyroid hormone effects. These agents

influence the metabolism of almost every class of foodstuff.

They exert profound effects on many enzymes and on almost

all organ systems, and they play an integral role in the

complex biological processes involved in growth and differ-

entiation [49]. The time-consuming procedures in USP mono-

graphs have been used for analysis of levothyroxine sodium

tablets [50] and liothyronine sodium tablets [511. The as-

say of the major thyroid hormones T3 and T4 also is done by

wet analysis [52-54], radioimmunoassay [55-58], chemical

derivatization followed by gas chromatography with electron

capture detection [59,60], thin-layer chromatography [61-63],

paper chromatography [64,65], electrophoresis [66], and gas

chromatography/mass spectrometry [67]. Thin-layer and paper

chromatography as well as the electrophoretic procedures do

not have good limits of detection for this kind of analysis.

Although the gas chromatographic procedures are sensitive,

they require the isolation of the iodoamino acids in a pure

form, which must then be converted to a volatile derivative








for chromatographic analysis. The radioimmunoassay proce-

dures are impractical for a small number of samples and have

the added problem of disposal of the radioactive wastes.

Liquid chromatography would normally be the method of

choice for these analyses because of the poor volatility of

the compounds. In fact, HPLC methods for separation of pure

iodoamino acids have appeared in the literature [68-72].

Recently, T3 and T4 tablets have been analyzed by reverse

phase LC using UV detection [73], and a gradient elution

separation of sixteen thyromimetic iodoamino acids has been

reported [74]; however, the low molar absorptivity of these

compounds at 254 nm precludes their determination at trace

levels. Detection at 220 nm improves this situation some-

what [74], and a clever catalytic post-column detection

scheme has also been shown [75]. Most recently, application

of amperometric electrochemical detection to these compounds

has been shown to give excellent limits of detection [76],

as has dansyl derivatization and subsequent fluorescence

detection [77].

The purpose of this study was to demonstrate the use-

fulness of gradient elution techniques with electrochemical

detection for the separation of seven thyromimetic iodoamino

acids. A rapid isocratic separation of TO, T2, T3 and T4,

as well as analysis of T4 both in tablets and injectable

intravenous samples is presented.







Standard Solutions


The compounds Tyr, MIT, DIT, TO, T2, T3 and T4 were

purchased from Sigma Chemical (St. Louis, Missouri) and

were stored in a freezer. Standard solutions were prepared

by dissolving appropriate amounts of each compound in meth-

anol containing 1% ammonium hydroxide and were stored in a

refrigerator.



Preparation of T4 Tablet Solution
and Injectable T4 Sample


Twelve tablets (1.5676 g) containing levothyroxine

sodium were dissolved in 20 mL of 0.01 M sodium hydroxide

using an ultrasonic bath. The sample solution was heated at

600C for 3 min, shaken for 3 min, and then filtered through

F2406-9 (S/P) filter paper. Before chromatographic injec-

tion, this solution was again filtered with a Rainin

(Rainin Instruments, Woburn, Massachusetts) HPLC sample

filter syringe using a 0.45 pm nylon-66 membrane filter.

The injectable sample was present as a powder and was pre-

pared by dissolving in 5 mL 0.9% sodium chloride solution.

This resulted in a clear solution which was then filtered

with the sample filter syringe.



Gradient Elution LC/EC


Amperometric electrochemical detectors are generally

considered incompatible with gradient elution