Factors affecting selectivity of covalent chromatography.

Item Type text; Dissertation-Reproduction (electronic)

Authors Crampton, Mary Catherine.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.

Download date 16/07/2018 13:11:50

Link to Item http://hdl.handle.net/10150/184356

INFORMATION TO USERS

The most advanced technology has been used to photo­graph and reproduce this manuscript from the microfilm master. UMI films the original text directly from the copy submitted. Thus, some dissertation copies are in typewriter face, while others may be from a computer printer.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyrighted material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are re­produced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each oversize page is available as one exposure on a standard 35 mm slide or as a 17" x 23" black and white photographic print for an additional charge.

Photographs included in the original manuscript have been reproduced xerographically in this copy. 35 mm slides or 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

,."""".I,U 'I! ,

I" I'li II I, ail I , "1 lii~L I

300 North Zeeb Road, Ann Arbor, M148106· 1346 USA

Order Number 8814226

Factors affecting selectivity of covalent chromatography

Crampton, Mary Catherine, Ph.D.

The University of Arizona, 1988

V·M·I 300 N. Zccb Rd. Ann Arbor, MI 48106

PLEASE NOTE:

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark_./_.

1. Glossy photographs or pages __

2. Colored illustrations, paper or print __ _

3. Photographs with dark background __

4. Illustrations are poor copy __ _

5. Pages with black marks, not original copy V 6. Print shows through as there is t~xt on both sides of page __ _

7. Indistinct, broken or small print on several pages __ _

8. Print exceeds margin requirements __

9. Tightly bound copy with print lost in spine __ _

10. Computer printout pages with indistinct print __ _

11. Page(s) lacking when material received, and not available from school or author.

12. Page(s) seem to be missing in numbering only as text follows.

13. Two pages numbered . Text follows.

14. Curling and wrinkled pages __

15. Dissertation contains pages with print at a slant, filmed as received /

16. Other _________________________________ __

U·M-I

FACTORS AFFECTING SELECTIVITY OF COVALENT CHROMATOGRAPHY

by

Mary Catherine Crampton

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF CHEMISTRY

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

198 8

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have read

the dissertation prepared by Mary Catherine Crampton

entitled Factors Affecting Selectivit~ of Covalent Chromatography

and recommend that it be accepted as fulfilling the dissertation requirement

for the Degree of Doctor of Philosophy

August 20, 1987 Date

August 20, 1987 Date

August 20, 1987 Date

August 20, 1987

Date

Dr. Quintus Fernando ;;.f_

VJ :L~/JilI/k1-< Dr. W. Ronald Salzman

August 20, 1987

Date Dr.~:ft.~d~~ Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

:if" /", !T'/5:" ~ August 20, 1987 Dissertation Director Date

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

Dedication:

To my parents.

iii

ACKNOWLEDGMENTS

The author would like to thank Analytichem International for

their support and all the people there who impacted on this work,

especially

Dr. Mark Stolowi·tz.

The author would like to thank Dr. M. F. Burke for his

support,

encouragement and guidance throughout the entire course of study.

The author would like to thank all my committee members,

group members and the rest of the members of the staff, faculty and

students at the University of Arizona who offered encouragement and

guidance.

iv

TABLE OF CONTENTS

LIST OF TABLES . . . .

LIST OF ILLUSTRATIONS.

ABSTRACT ..

1. INTRODUCTION

Principles of Chromatography General Elution Chromatography. Solid Phase Extraction. Types of Interactions

Bonded Functiona1ities . Pheny1boronic Acid. Thio1 Phases.

The Solid Support.

2. BACKGROUND AND PROCEDURES.

Chemically Modified Surfaces Supports ....... .

Organic Supports . Inorganic Supports

Immobilized Functiona1ities Modification. . . . .

Procedures . . . . . . . . . . . Modification Methods ....

Reactions with bare silica Aminopropy1 based bonded silicas Activation of mercaptopropy1 bonded silica Propy1urea based bonded silicas.

Characterization of Bonded Silicas.

Summary.

Elemental Analysis . . . . Solid Phase Extraction . . Diagnostic Chromatography.

3. EFFECTS OF MODIFICATION.

Control of Non-specific Interactions Mobile Phase .. . Stationary Phase ....... .

v

Page

viii

ix

xiv

1

5 5 8 9

11 11 13 17

19

19 20 20 21 22 24 30 30 30 31 31 37 40 40 45 46 49

51

51 51 52

TABLE OF CONTENTS--Continued

Experimental . . . . . . . . Results and Discussion . . .

Dispersive Interactions Lipophilicity of the Bonded Phases Effects of Solvent Strength.

Electrostatic Interactions. Anion Exchange Sites . . . Cation Exchange Sites ... Effects of Ionic Strength.

Hydrogen Bonding. . . . . . . . Conclusions. . . . . . . . . . . . .

Controlling Liophilic Interactions. Controlling Electronic Interactions

4. PHENYLBORONIC ACID

Introduction . Acid base chemistry Compounds Expected to Interact with PBA

Diols ..... Polyhydroxyls. Hydroxyacids . Hydroxyamines. Other Functiona1ities.

Isolation of Compounds. Experimental . . . . . . . . .

Solid Phase Extraction. . Diagnostic Chromatography

Results and Discussion . . . . Retention Dependence on pH.

Diol ..... . Hydroxyamines .. . Mercaptoa1cohols . Salicylic Acid and Related Compounds Diketone . . . . . . . . . Keto-carboxylic Acid . . .

Effects of Modification on Retention. Secondary Interactions . . . Shift in pKa of PBA. . . . .

Demonstration of the Isolation of Compounds Diols .... Hydroxyacids

Conclusion. . . .

vi

Page

53 55 55 55 57 57 57 61 64 68 70 70 71

73

73 73 75 77 78 78 78 79 79 80 80 80 82 82 82 89 92 92

.100

.100

.104

.104

.104

.105

.106

.106

.106

TABLE OF CONTENTS--Continued

5. COVALENT CHROMATOGRAPHY OF THIOL CONTAINING COMPOUNDS.

Traditional Thiol Chemistry .. Separation of Thiol Containing Solid Phase Thiol Chemistry.

Mixed disulfide . . Phenylmercurial . . Phenylarsonous acid

Experimental . . . . Results & Discussion

Compounds

Disulfide Phase Organomercurial Phenylarsonous Acid Extraction of Reducing Agent.

Conclusions ......... .

6. CONCLUSIONS & REFERENCE WORKS.

REFERENCES . . . . . . . . . .

vii

Page

.111

.111

.114

.115

.115

.117

.121

.124

.125

.125

.133

.138

.144

.146

.148

.154

LIST OF TABLES

Table Page

2.1 Elemental analysis of modified silicas ... .43

2.2 Capacity factors of several probe molecules on propy1aminoPBA with methanol as the organic modifier .. 47

4.1 Molecules that interact with PBA. . .74

4.2 Buffers for solid phase extraction. .81

5.1 Dissociation Constants for Complexes with Methy1mercuric Acid. . . . . .................... 120

viii

Figure

1.1

1.2

1.3

1.4

LIST OF ILLUSTRATIONS

Page

A. Ionization of phenylboronic acid to the tetrahedral form

B. Interaction with diols. . . . . . . 12

Retention, detection and elution of thiols using a mixed disulfide phase. 14

Retention and elution of thiols on an immobilized organomercurial. 15

A. B.

Hydration of immobilized arsine oxide to arsonous acid Interaction of immobilized arsonous acid with dithio1

C. compounds to form a ring structure Interaction with monothio1s .16

2.1 Polymeric C18 bonded silica. . . .25

2.2 Propylaminophenylboronic acid andthe second interactions occurring on this phase. .... ..... .27

2.3 Propylureaphenylboronic acid immobilized on silica .29

2.4 Reaction of triethoxyaminopropyl silane with silica. .32

2.5 Reaction of triethoxychloropropyl silane with silica .33

2.6 Reaction of triethoxymercaptopropyl silane with silica .34

2.7 Reaction of chloropropyl bonded silica with m-aminophenylboronic acid bonded silica . . . . . 35

2.8 Reaction of 2,2-dithiodipyridine with mercaptopropyl bonded silica to form the mixed disulfide phase ......... 36

2.9 Reaction of propylamino bonded silica with 1,1'­carbodiimidizole followed by m-aminophenylboronic acid to form propylureaphenylboronic acid bonded silica.. . .38

2.10 Reaction of a CDI activated aminopropyl phase with p-aminophenyl-mercurial hydroxide . . . . . . 39

ix

LIST OF ILLUSTRATIONS--Continued

Figure Page

2.11 Reaction of a COl activated aminopropy1 phase with p-aminopheny1-arseneoxide . . . . . . . . . . . . . . 41

2.12 Reaction of a COl activated aminopropy1 phase with ammonia to form a propy1urea phase . . . . . . . . . .. . ... 42

3.1 Comparison of ionization of the two modification methods. 54

3.2 Retention of benzene and toluene on three modified silicas as a function of pH . . . . . . . . . . . . . . . . 56

3.3 Effects of acetonitrile concentration on the retention of benzene, aniline, phenol, and nitrate on propy1aminopheny1boronic acid bonded silica.

3.4 Retention of benzoic acid and phenylacetic acid on three

.58

bonded silicas as a function of pH . . . . . . . . .59

3.5 Effects of pH on the retention of benzene sulfonic acid and pheny1trimethy1ammonium on three modified silicas . 60

3.6 Retention of aniline and benzy1amine on three modified silicas as a function of pH . . . . . . . . . . . . 62

3.7 Effects of ionic strength on theretention of benzy1amine and benzoic acid on propy1urea (a), propy1ureapheny1boronic acid (b), propy1aminopheny1boronic acid (c) and aminopropyl bonded silica (d). . . . .. .............65

3.8 Effects of ionic strength on the retention of phenylacetic acid and benzenesulfonic acid on propylurea (a), propylureaphenylboronic acid (b), propylaminophenylboronic acid (c) and aminopropyl bonded silica (d) ...... 66

3.9 Ionic strength dependence on the retention of phenyltrimethylammonium on propylureaphenylboronic acid (a) and propylaminophenylboronic acid (b) . . . . . . . 67

3.10 Retention of phenol and benzyl alcohol on three bonded silicas as a function of pH ................ 69

4.1 Equilibration of PBA to its tetrahedral form. Reaction with diols and thereverse reaction when the pH is dropped ... 76

x

LIST OF ILLUSTRATIONS--Continued

Figure Page

4.2 Two PBA modified silica phases, a 1iophi1ic phase (a) and a hydrophilic phase (b). . . . . . . . . . .. .. 83

4.3 Retention of catechol on three modified silicas as a function of pH . . . . . . . . . . .. .. 84

4.4 Retention of 2-hydroxybenzy1 alcohol as a function of pH. . . . . . . . . . . . . ......... 86

4.5 Retention of aminothio1pheno1 (a), 2-hydroxybenzy1 alcohol (b) and 2-aminobenzy1 alcohol (c) on commercially available PBA columns by solid phase extraction. 88

4.6 Retention of aminopheno1 as a function of pH 90

4.7 Retention of 2-aminobenzy1 alcohol as a function of pH 91

4.8 Retention of 2-mercaptobenzy1 alcohol as a function of pH. 93

4.9 Retention of salicylic acid as a function of pH. 94

4.10 Salicylic acid (a) and salicylamide (b) retained on PBA as a function of pH . . . . . . . . .96

4.11 Retention of salicylamide as a function of pH on pheny1boronic acid . . . . . 98

4.12 Retention of thiosa1icy1ic acid as a function of pH on pheny1boronic acid ., .......... 99

4.13 Retention of benzi1 as a function of pH on phenyl~oronic acid. . . . . . .. ..... . . . .. . 101

4.14 Retention of benzi1 (a) and benzillic acid (b) on PBA as a function of pH .102

4.15 Retention of benzi11ic acid as a function of pH on pheny1boronic acid .103

4.16 Demonstration of the isolation of 1,3-dio1s on PBA .107

4.17 Demonstration of the retention of hydroxyacids on PBA. .108

xi

LIST OF ILLUSTRATIONS--Continued

Figure Page

5.1 Reduction of disulfide bonds by dithiothreitol followed by reaction with 2,2-dithiodipyridine . . . . . .. 113

5.2 Retention, detection and elution of thiols using a mixed disulfide phase. .. ........ . . .116

5.3 Retention and elution of thiols on an immobilized

5.4

organomercurial ..... . .. 118

A. B.

c.

Hydration of arsine oxide to arsonous acid Interaction of immobilized arsonous acid with dithiol compounds to form a ring structure Interaction with monothiols . . . . .. 122

5.5 Use of te mixed disulfide phase for the quantitation of cystiene (a) and dithithrietol (b). 126

5.6 Secondary interactions on the mixed disulfide phase as a function of % methanol (a) and -log ionic strength (b). 128

5.7 Retention of aniline (a) and benzoic acid (b) as a function of % methanol on the mixed disulfide phase at pH 4 and 6 . . . . . .................... 130

5.8 Reaction of 2-mercaptoetthanol with themixed disulfide phase .. ............ . . . . . 131

5.9 Retention as a function of % methanol (a) and -log ionic strength (b) on an exhausted mixed disulfide phase ... 132

5.10 Non-specific interaction of organomercurial bonded silica using 5% acetonitrile and 95% ammonium acetate buffer with an ionic strength of 0.1 M ................ 134

5.11 Oxytocin retained and eluted on an immobilized organomercurial . ..... . . . . . . . . . .137

5.12 Automated system for the retention of thiols on organomercurial. 139

5.13 Non-specific interactions on phenylarsonous acid bonded silica 140

xii

LIST OF ILLUSTRATIONS--Continued

Figure Page

5.14 Retention of cystiene on immobilized arsonous acid as a function of solvent strength. . . . . . . . . . . .. .142

5.15 Use of arsonous acid bonded silica for the separation of dithiols (DTT) and monothiols (cystiene) detected by the mixed disulfide phase. A. Without arsonous acid column. B. With arsonous acid collunn .............. 143

xiii

ABSTRACT

Of the interactions utilized in the separation of chemical

species, the reversible covalent bond is the strongest and most

selective. In order to exploit the selectivity of this interaction,

an understanding of the effects of several factors on the formation of

the covalent bond have been studied.

Covalent chromatography is most useful in solid phase

extraction. The strength of the covalent bond will allow for high

distribution coefficients needed to quantitatively retain chemical

species. The selectivity of the covalent interaction allows for the

retention of a specific compound or class of compounds. Under

conditions where the covalent bond is no longer formed or will break,

a low distribution coefficient is possible, and the compounds may be

eluted from the sorbent in a small volume allowing for

preconcentration.

The sorbent consists of three parts, the active functionality

capable of forming reversible covalent bonds, the solid support and

the spacer arm that tethers the active functionality to the support.

Silica supports demonstrate several advantages over organic supports.

However, silica supports have been limited by the activity of the

residual silanols and the use of hydrophobic spacer arms. This

research describes the preparation and characterization of a

modification method for silica with a hydrophilic spacer arm that

restricts the sample from the residual silanols.

xiv

Immobilized phenylboronic acid (PBA) interacts with compounds

containing polar functionalities in the correct configuration, but

this interaction is dependent on the local environment. Controlling

the local environment allows for the control of the interaction

provided the effects are understood. Diagnostic chromatography was

used to determine the effects of the solvent strength, ionic strength,

pH and composition of the mobile phase and the effects of the spacer

arm on the interactions of PBA with several compounds.

Three phases selective for thiols were also characterized.

Thiopyridone attached to mercaptopropyl bonded silica through a

disulfide linkage, is used for isolation and detection. Immobilized

phenylmercury was utilized for the extraction of thiol containing

species while immobilized phenylarsonous acid selectively extracts

dithiol compounds. A unique and powerful means of separation of

monothiol from dithiol compounds has been demonstrated.

xv

CHAPTER 1

INTRODUCTION

Solid phase biochemistry can be defined as any biochemical

process where one component is immobilized on a solid support(l).

This approach is popular due to the ease of sample handling and ease

of automation, as well as the fact that the immobilization of a

chemical species often allows for more efficient and more selective

interactions or reactions. This approach to chemistry is not limited

to biochemistry and can be separated into three types: synthesis,

sensors or detection systems, and separation systems. Solid phase

synthesis has been shown to be both simpler and more efficient than

solution synthesis especially for large molecules such as DNA or

proteins (1,2,3,4). Immobilized enzymes (5) as well as immobilized

redox reagents (6) are often used as catalysts in organic synthesis.

Immobilized enzymes are not only useful in synthesis, but are also

used as chemical sensors such as immobilized enzyme electrodes (7)

which provide selective and sensitive methods for quantitative analy­

sis. In addition, solid phase reactors coupled to traditional UV

liquid chromatography detectors have been employed in liquid chromato­

graphy as sensitive, selective and versatile detection systems (8).

Separation methods in solid phase chemistry have traditionally relied

on bioselective adsorption (affinity chromatography) or chemiselective

adsorption (e.g. covalent chromatography)(l).

1

2

In each of these types of chemistry, it is important to

understand the factors that affect the interactions of chemical

species with the surface. In general, the unique selectivity, which

is the goal of any solid phase chemistry, relies on the formation of a

reversible linkage between the species of interest and the solid

phase. This linkage usually consists of at least one covalent bond,

although other high energy interactions such as ion exchange or charge

transfer interactions, have also been used. The exploitation of solid

phase chemistry to its full potential relies on understanding the

conditions which allow one to form and break this linkage in a rapid

and efficient manner.

The covalent bond is not the only interaction possible between

the surface and the spec~es of interest. In some cases, the molecule

adsorbs to the surface through non-specific interactions. Understand­

ing these non-specific or secondary interactions and the factors which

affect them allows for the utilization of these interactions to

enhance the formation of the selective interaction or the minimization

of these interactions when adsorption is undesirable.

Chromatography has not only proven itself as a separations

technique, but also as a method of studying solid surfaces, (9,lO) to

gain insight into the types of interaction and the relative strengths

of the interactions available at a solid surface. Diagnostic chro­

matography is the use of the principles of chromatography to determine

a wide variety of information about a solid surface.

3

While covalent chromatography has proven itself for many

separations, its use has been limited by a lack of understanding of

the factors which affect the separations. Several factors affect the

formation of the covalent bond including pH, solvent strength, ionic

strength, solvent composition of the mobile phase and environment of

the solvated stationary phase. The optimum conditions for the cova­

lent reaction depend on the analyte and the matrix, understanding the

effects will allow for the determination of mobile and stationary

phases for an efficient separation.

In addition to the covalent interaction several other types of

interaction may be occurring on these phases. These secondary

interactions include electrostatic, H-bonding and dispersive inter­

actions. These interactions allow for the compound of interest to be

solvated by the stationary phase. The sample must spend some time in

the stationary phase to allow for the covalent reaction to occur.

These secondary interactions are also responsible for non-specific

retention. Molecules that are unable to form a covalent bond may be

retained on the surface through non-specific interactions. Recog­

nizing these interactions allow for the removal of contaminants that

might otherwise be left on the surface or eluted with the compound of

interest.

The physical stability and the efficiency of silica are not

matched by any of the soft gels commonly used as supports in covalent

chromatography. The limitation of silica as a solid support has been

due to a lack of creativity in the modification of silica, in that

4

traditional modification results in a lipophilic surface. Since a

hydrophilic support is desired for work with biomolecules(22 ,23) , a

hydrophilicly modified silica support has been developed and charac­

terized. This modification method allows for a broader range of

mobile phases due to a decrease in dispersive and anionic interactions

and will expand the use of silica as a support in all areas of solid

phase chemistry.

Four different functionalities have been immobilized on silica

using two different modification methods. These have been character­

ized through elemental analysis, solid phase extraction and diagnostic

chromatography. The effects of pH, ionic strength, solvent strength

and composition on the retention of several probe molecules have been

determined. The probe molecules were chosen because of their ability

to probe different interactions occurring on the phases. Thus the

effects of the changes in mobile phase and stationary phase on spe­

cific interactions are determined.

Once the factors affecting the selectivity of these phases is

understood the versatility can be increased by coupling several

columns together. The selectivity of the covalent bond allows for the

isolation of several different classes of compounds, each on a differ­

ent column-. Compounds that interfere with the immobilization of the

compound of interest may be removed prior to the immobilization

provided the interferants contain a distinctive functional group.

However choosing optimum conditions for this type of separation will

depend on an understanding of the selectivity of each column.

Principles of Chromatography

General Elution Chromatography

5

In liquid chromatography, the solute molecules are swept by

the flowing stream of the mobile phase past a second phase, the

stationary phase, which is most commonly a chemically modified solid

substrate. The general elution equation for chromatographic processes

can account for a partition mechanism, an adsorption mechanism and an

exclusion mechanism as shown in equation 1.1. (11)

Eq 1.1

where

Vr1 - the retention volume of the solute

Vm the volume of the mobile phase in the column

K the thermodynamic equilibrium constant (Eq l. 3)

Vs the volume of the stationary phase

K' the adsorption thermodynamic equilibrium constant

As the surface area of the support

K" the exclusion thermodynamic equilibrium constant

Vi the interstitial volume

The retention of a probe to partitioning is defined as KVs . This

mechanism "involves the intercalation of a molecule into the stationary

phase, so as to change the solvation of the ana1yte. The adsorption

mechanism defined as K'As involves the adsorption of a molecule on the

surface of the solid support. Retention due to exclusion is defined

as K"Vi' Smaller molecules are able to diffuse into small pores while

large molecules are excluded, and therefore, K" represents the frac­

tion of the pore volume available to a given analyte. Since the

volume of the stationary phase available to smaller molecules is

larger, the retention volume is also larger. In some cases, all of

these mechanisms may be occurring at the same time; however, the

separation is more easily controlled when one mechanism predominates.

6

If partitioning is the predominate mechanism for retention the

elution equation can be given as equation 1.2.

Eq 1.2

The thermodynamic equilibrium constant is defined by:

K [Als / [Alm Eq 1.3

where [Als is the concentration of the analyte in the stationary phase

and [Alm is the concentration in the mobile phase. If the analyte has

a strong interaction with the stationary phase, the distribution

coefficient will be large and the molecule will spend a significant

amount of time associated with the stationary phase. The molecules

which interact strongly with the surface will be retained longer than

those that interact weakly with the surface. Thus, molecules are

separated because of the different strengths of interaction with the

surface.

The distribution coefficient is difficult to determine

experimentally, because the volume of the stationary phase is unknown

and depends on the amount of mobile phase imbibed in the stationary

phase. The capacity factor (k') is defined as:

7

Eq. 1.4

where ns is the number of analyte molecules in the stationary phase,

nm is the number of analyte molecules in the mobile phase, K, Vs and

Vm are defined in equation 1.1 and Vm/Vs is the phase ratio. The

capacity factor is measured as

Eq. 1.5

where Vr is the retention volume.

In differential migration chromatography, a k' of less than 2

does not allow for the resolution of compounds. Above a k' of 20 the

increase in resolution is minimal and the increase in analysis time

limits the number of analyses possible. For this reason, a k' between

2 and 20 is considered optimal (12). Assuming a phase ratio between

.1 and 10 -means optimal distribution coefficients are between 0.2 and

200. In order to control the retention of a molecule the distribution

coefficient may be altered through changes in the mobile or stationary

phases or the altering the phase ratio.

8

Solid Phase Extraction

The general elution equation (equation 1.1) is incomplete, in

that it does not take into account the capability of a phase to form a

covalent or an ionic bond with an analyte. Including this type of

retention, the equation is expanded to

Eq 1.6

where K'IIN accounts for retention due to bonds occurring with specific

sites. N is defined as the number of sites and K'II is the bonding

coefficient.

The strong interactions that are described by the last term in

the general elution equation are most useful in solid phase extraction

(SPE). This extraction involves the same principles as the more

traditional differential elution separations. In SPE an analyte is

quantitatively retained and isolated on a solid sorbent. In order to

quantitatively retain the analyte while eluting any contaminants a

distribution coefficient of 1000 or greater is needed. The strength

of the covalent bond allows for this high distribution coefficient

while the selectivity allows for the retention of only compounds with

specific functional groups. Once the analyte is isolated on the

sorbent, the conditions are changed to allow for a distribution

coefficient to be 0.001 or less. The analyte is eluted in the small­

est volume possible to allow for preconcentration. This has also been

called digital chromatography, since in one set of conditions, the

solute is retained on the column and a second elutes the solute off

the column. SPE usually isolates a class of compounds rather than

single components. This type of separation is useful to clean up a

sample prior to HPLC or GC separations and for analyses that do not

require a pure compound but are affected by a complex matrix.

Types of Interactions

9

The most commonly used interaction in liquid chromatography

today is the dispersive or induced dipole- induced dipole interactions

of reverse phase chromatography. The dispersion interactions, as

with all of the Van der Waals forces, are relatively weak and involve

about 2 kcal/mol. These interactions result from the instantaneous

dipoles formed in molecules due to the movement of electrons and the

formation of corresponding dipoles in nearby molecules. While the

universal nature of these interactions make them useful for many

separations, it also limits selectivity.

Normal phase chromatography is older than reverse phase, but a

lack of reproducibility due to experimental parameters limits it's

use. Normal phase chromatography relies on hydrogen bonding inter­

actions. These interactions are stronger than dispersion forces and

average about 10 kcal/mol.

The electrostatic interactions of ion chromatography, at about

100 kcal/mol, are stronger than those of normal or reverse phase

chromatography. They are also more selective, occurring only with

ionized solutes. The strength of electrostatic interactions allows

10

for quantitative retention of compounds and early uses of electro­

static interactions for separations employed in this ion exchange

chromatography. More recently with the work of H. Small electrostatic

interactions have found wide use in elution chromatography (13).

The strongest and most selective interaction available in

chromatography today is the covalent bond. Reversible covalent bonds

that form between functional groups bound to a solid support and

selective solutes have been used in both general elution chromato­

graphy and SPE. One type of chromatography that has utilized covalent

bonds for elution chromatography has been ligand exchange chromato­

graphy. In this type of chromatography the mobile phase contains a

compound which is capable of competing with the species of interest

for the active site. The real advantages of this interaction are

realized in SPE. The strength of the covalent bond from 50-200

kcal/mol allows for large distribution coefficients under conditions

where the solute reacts. These large K values needed for quantitative

retention in SPE are achieved. Under conditions where the solute does

not react or the bond will break, very small distribution coefficients

exist allowing for a quantitative elution in small volumes. Changes

in the mobile phase allow for the retention or elution of the solute.

The chemical nature of the functionality on the surface dictates the

factors that control the interaction. The physical nature of the

solid support and the spacer arm that links the support and the

functionality will also affect the selectivity of a functionality.

11

Bonded Functionalities

The ideal functionality in covalent chromatography would be

able to form reversible stable bonds with specific functional groups.

For efficient separations, the bond formation must be more rapid than

the diffusion which controls band spreading on the column. In addi­

tion, the bond would be easily and controllably reversed in order

to elute the molecule from the packed bed. Once again the reverse

reaction must be fast to allow for elution in the smallest possible

volume.

Several different functionalities have been bound to a silica

surface through two bonding methods. The effects of bonding methods

as well as solvent strength and composition, ionic strength and pH on

the retention of molecules on these phases has been studied.

Phenylboronic Acid

Phenylboronic acid (PBA) (figure 1.1) reacts readily with a

wide variety of polar functionalities if they exist with the appro­

priate geometric configuration. The most commonly studied molecules

are the 1,2- and l,3-diols (14). However, other combinations of

amines, mercaptans, carboxylic acids, ketones or hydroxy groups will

also react with PBA (24).

PBA is a Lewis acid with a pKa near 8.3. It ionizes through

acceptance of a hydroxy group at high pH. It is the acid-base chem­

istry that controls the formation of covalent complexes. At high pH,

it is tetrahedral and ionized, at low pH it is trigonal planar and

A

+2H,O~

Figure 1.1 A. Ionization of phenylboronic acid to the tetrahedral form. B. Interaction with diols.

12

neutral. These characteristics can be used to explain the pH depen­

dence of retention for several functiona1ities.

Thio1 Phases

13

The use of a mixed disulfide to covalently link thiol contain­

ing compounds to surfaces has been used for the immobilization of

compounds (15,16,17,18). A thiol bonded surface activated by

reaction with 2,2'-dithiodipyridine (figure 1.2) makes it possible to

separate thiol containing peptides from non-thiol containing peptides

or to separate enzymes with active thiol sites from denatured enzymes

or to protect or store enzymes with active thiol sites. In addition

the thiopyridone is displaced when a solute interacts with the sur­

face. The chromophore released may allow for an increase in the

sensitivity of detection for thiol containing compounds (15). The

ability to use this phase for isolation and quantification makes this

a very useful phase. Unfortunately, this phase has a few limitations

due to the ease of oxidation of the free thio1s, which will be dis­

cussed in Chapter 5.

A second phase selective for thio1 containing compounds is

shown in figure 1.3. This phase is an organomercurial bonded phase.

The mercu~y is very stable and forms strong interactions with thio1s

(19,20) under a wide range of conditions. This phase is useful for

the isolation of thio1s, but the displacement of a chromophore has not

been demonstrated.

Immobilized phenylarsonous acid (figure 1.4) also interacts

with free thio1s. This phase is able to interact at two sites. When

z

}- S-S-R 0 }-S-S-o + i: III. R-SH ". 11& 0

'" Q c

S=O ,N

H z

}-S-S-R }-S-S-R' 0

-= + It'-Slot '\ ::) -.. R-SH

Figure 1.2 Retention, detection and elution of thiols using a mixed disulfide phase.

14

Z i }-H9-OH + R-SH c(

0_ Z}-~ Hg-S-R + R'-SH "\ • }-H9-S-R'

R-SH

15

Figure 1.3 Retention and elution of thio1s on an immobilized organomercurial.

A

JO _all I O-AS"'OH HS-C-

B + I

HS-C-HN I

~

JO A ,all C 0- s'OH

+ 2 R-SH HN

~

~ ~

rA-r .... OH M AS'OII

HN

~

I

,..5-<;-)O-AS'S-T-

HN

~

JOT ,S-R . a AS"' S_ R ~ ~ HN

~

+2 1120

+2 H2O

Figure 1.4 A. Hydration of immobilized arsine oxide to arsonous acid. B. Interaction of immobilized arsonous acid with dithiol compounds to form a ring structure. C. Interaction with monothiols.

16

interacting with a dithiol, a ring structure is formed. This ring

structure is more stable than the structure formed with monothiols,

and one can immobilize dithiol compounds without retaining the mono­

thiol containing compounds (21).

17

These phases used separately or in series will allow for easy

methods of separating and detecting thiol containing compounds.

The Solid Support

All areas of solid phase chemistry, including chromatography,

rely on properly modified solid supports. While no support to date is

considered ideal, several different types have been used. The ideal

support would be physically stable. It would be rigid so as to

withstand high pressure drops necessary to facilitate fast flow rates.

In addition, it would have a large surface area to allow for a large

number of active sites in a small volume. The active sites need to be

accessible to the solute requiring a pore size to be of a known

uniform size. The support must be insoluble under a wide range of

conditions. It must not be easily degraded through microbial attack.

In addition to the physical characteristics of the support, it is also

important to covalently bind the active functionality to the support.

Once chemically modified, the ideal support would be com­

pletely reacted to contain the highest concentration of active sites

possible. The modified support would not diminish the ability of the

functionality to interact and would not interact directly with the

solute.

18

Of the solid supports available, silica gels are physically

strong, easily reacted and available in specific and consistent sizes

and pore volumes. Silica gels have found limited use in covalent

chromatography due to the surface activity of silica and the modifi­

cation methods used. To expand the use of silica gel in covalent

chromatography and all areas of solid phase chemistry, the silica must

be properly modified to interact with chemical species in a controlled

manner. Proper modification relies on understanding the silica

surface and the types of active groups present.

CHAPTER 2

BACKGROUND AND PROCEDURES

Chemically modified silica surfaces are routinely used for

separations, however, the mechanism of most separations are not well

understood. The general elution equation (Eq. 1.6) defines retention

of molecules as due to several different mechanisms. In addition,

each mechanism may rely on several different parameters; for example,

partition chromatography may rely on more than one interaction for the

retention of molecules. Usually in cases with multiple interactions,

one interaction is considered to predominate and experimental condi­

tions are set to assure this. However, the separation depends on all

the interactions occurring at the surface, and the predominant or

primary interaction is affected by the secondary or nonspecific

interactions. If a molecule is excluded from the surface due to ionic

interactions, it will not form a covalent bond even when the condi­

tions are such that the covalent bond should form. In addition, if a

molecule is well solvated by the stationary phase, a covalent bond

will be stabilized. Understanding all the interactions occurring at

the surface allows for better control of the separation.

Chemically Modified Surfaces

A chemically modified surface consists of three parts, the

solid support, the immobilized functionality and the spacer arm that

attaches the functionality to the surface.

19

20 supports

There are many different types of supports available today. A

complete review of these supports can be found elsewhere (12,22,23).

The important properties of solid supports are physical and chemical

stability, porosity, particle shape and particle size. Solid supports

may be classified as inorganic or organic, with further classification

depending on chemical or physical properties.

Organic Supports. Organic supports include natural macro­

molecules, polymers and synthetic polymers. The organic supports may

be hydrophilic or lipophilic. The lipophilic supports have not been

widely used in biochemical applications due to the aqueous nature of

the samples. Of these supports, the most commonly used are the

hydrophilic natural polysaccharides.

Polysaccharides include cellulose, dextran, agarose and

starch. These supports are hydrophilic, and the surface hydroxyl

groups are easily reacted to bind surface moieties. Dextran and

agarose gels are the most widely used of the polysaccharides and have

been utilized as macroporous supports in several areas of chromato­

graphy. Starch has found little use because it is degraded easily

through microbial attack. Cellulose has found limited use, because

until recently it was not available in rigid macroporous particles.

Of dextran and agarose supports, the agarose gels are mechanically

more stable and more resistant to microbial attack; however, they

suffer from several disadvantages. They cannot be heat sterilized;

they disintegrate in alkaline solutions and organic solvents; they

shrink irreversibly on drying and are more expensive than other

21

supports. The main disadvantage of both dextran and agarose is that

they are soft gels compared to silica. They lack pressure stability

and are not useful in high performance systems. Even in medium

pressure systems, one must be concerned with the compressibility of

these gels.

Commonly used synthetic polymers include polystyrene, poly­

acrylate types, polyacrylamides, and many others. A wide range of

synthetic polymers are easily prepared with different porosities due

to the various polymers and crosslinkers available. Unfortunately

many of these supports shrink and swell with changes in ionic or

solvent strength. This shrinking and swelling will cause concern in

many types of separations, especially in SPE where the change in

mobile phases is extreme.

Inorganic Supports. The inorganic supports have high mechan­

ical strength, thermal stability, resistance to organic solvents and

microbial attack, easy handling, long shelf lives and do not change

structure over wide ranges of pH, pressure, or temperature. Highly

porous inorganic supports with high surface areas are available and

allow for a high concentration of binding sites in a small volume.

The mechanical stability of these supports allows for small particle

sizes, since they are able to withstand the pressure caused by small

particles. Inorganic supports offer more efficient separations than

any of the organic supports.

Porous inorganic supports inclUde several natural products;

however, these have a wide range of pore sizes. Controlled pore glass

has been used for some applications, but is expensive and has proven

to be less stable than silica or alumina.

Silica is available in a wide range of sizes and porosities

to fit any application. Furthermore, the chemical modification of

silica has been widely studied, and a wide range of alkylsilanes is

commercially available.

22

The availability of silica in small uniform particles offers

an advantage in reducing band broadening. Decreasing band broadening

will allow for more efficient separations, and in SPE, eluting the

compound of interest in a smaller volume. Band broadening in chro­

matography is commonly expressed as the theoretical plate height. The

five major contributors to band broadening are eddy diffusion, mobile

phase transfer, longitudinal diffusion, stagnant mobile phase transfer

and stationary phase mass transfer, but longitudinal diffusion is

negligible in liquid chromatography. Since smaller particles decrease

eddy diffusion, mobile phase transfer and stagnant mobile phase

transfer, smaller particles will have a significant effect on the

reduction of band broadening. If all other parameters are optimal,

the minimum plate height is proportional to the particle size.

Immobilized Functionalities

In covalent chromatography, the efficiency of the separation

also depends on the kinetics of the covalent interaction. Retention

of molecules depends on two steps,much like the kinetics of ion

chromatography (11). The first step is the partitioning of the

molecule into the stationary phase. The second step is the reaction

23

of the molecule with the immobilized functionality. The first step is

governed by the non-specific interactions, and an understanding and

optimization of these interactions are necessary prior to the the

study of the covalent interaction. Exclusion of the compound of

interest will inhibit the formation of the covalent bond, and reten­

tion through non-specific interactions will stabilize the covalent

interaction. At the same time it is important that non-specific

interactions are not so strong that contaminants are not removed from

the phase.

The amount of time an unretained molecule spends in contact

with the packed bed is dependent on flow rate and void volume of the

column as well as the nonspecific retention. A small packed bed with

a void volume of 50 u1 using a flow rate of 2 m1/minute has a resi­

dence time of 1.5 seconds for an unretained molecule. Fo~ a k' of 1,

the compound spends 1.5 seconds in the stationary phase. For a k' of

5 there is a residence time of 7.5 seconds. Using the theoretical

plate hypothesis (10) and assuming 1000 plates means that each encoun­

ter with the stationary phase lasts, on the average, 7.5 milliseconds

and the reaction will occur in this time. If the molecule does not

react in this time, the flow rate may be slowed or stopped to allow

more time for the reaction. Alternatively, the time of reaction could

be decreased by raising the temperature.

In order to assure that an adsorbent is not overloaded, the

amount of the sample should be about 1% of the stationary phase. In

these cases, the reaction kinetics are pseudo first order. For 99.9%

24

of the compound of interest to react in 7.5 milliseconds, a first

order rate constant of 1000 seconds- 1 is needed. Reactions with rate

constants of this magnitude or higher can easily be utilized in

covalent chromatography.

Modification

The most commonly used surface for chromatography is an

octadecy1si1ane modified si1ica(C18). Extensive research into an

understanding of the nature of this surface has been carried out

(9,10). Figure 2.1 is a polymerically bound octadecy1si1ane surface

equilibrated with a methanol/water mobile phase. On the molecular

level this surface is a heterogeneous phase. Associated with the

residual si1ano1s on the surface is about two layers of water mole­

cules. Moving away from the surface, the hydrocarbon chain solvated

with the organic component of the mobile phase forms a nonpolar

region. Moving to the end of the chain, the polarity will gradually

increase to the polarity of the mobile phase. Thus, the surface shows

a polarity gradient (9) with different molecules finding their most

favorable interactions at different areas of the phase, and therefore

intercalating to the area of its most favorable interaction. Separa­

tions on C18 phases rely on several types of interactions. The

residual si1ano1s of the surface ionize above a pH of 6. Thus, at

neutral to basic pH, the si1ano1s are ionized and are capable of

electrostatic interactions. Unionized si1ano1s and the associated

water are capable of hydrogen bonding. The C18 chain is capable of

dispersive interactions. Incorporated solvent molecules will also

H H H H If HCH HCH HCH HCH HCH I \ \',

HCH HCH HCH HCH HCH \ / / I I HCH HCH HCH HCH I1CH I \ \ \ \

HCH HCH HCH HCH HCH ,I 1// HCH HCH HCH HCH HCH

HC' JcH ~H ~H ~H \ / / / / HCH HCH HCH HCH HCH I \ \ \ \

HCH HCH HCH HCH HCH \ / I 1 / HCH HCH HCH HCH HCH I \ \' \

HCH HCH HCH HCH HCH \. / I / / HCH HCH HCH HCH HCH I \ \' \ HC~;eH 1H ;CH H~

HCH HCH HCH HCH HCH / \ \ \ /

HCH HCH HCH HCH HCH \ 1 I 1 I HCH HCH HCH HCH HCH I \ \' \

HCH HCH HCH HCH HCH \ I / I I HCH HCH HCH HCH HCH I '\ \ \ \

HCH HCH HCH HCH HCH \ I '\ I /

-51-Q-5i-O-H H H-O- 5i-0-5i-0-5i-0-H I '\ , 1 \ '\

o 0 0 0 0 '\ " \ , \

51 5i 51 51 51 51 Si 51 Si /\/,/\.1'1'1"'1\1

o 0 0 0 0 0 0 0 0

Figure 2.1 Polymeric C18 bonded silica.

25

26

interact with the sample. There is a great difference in the energet­

ics of these interactions; however, all of these interactions are

important in the retention.

While alkyl bonded silicas have proven themselves for reverse

phase chromatography, the hydrocarbon nature of these phases has been

a limitation in covalent chromatography. Covalent chromatography is

most applicable to biochemical separations. Most of these samples are

aqueous and often the compounds of interest are denatured in organic

solutions. If a lipophilic phase is used, organic solutions are

needed to assure that non-specific adsorption does not occur and that

sufficient solvation of the sorbent is occurring.

The residual silanols which may be important in the separa­

tions of reverse phase chromatography (9) result in irreversible

retention of biopolyrners. If a silica support is used in covalent

chromatography, care must be taken to assure that the residual sila­

nols are unable to interact with the sample.

Silica based stationary phases used in covalent chromatography

have traditionally relied on alkyl silane backbones. N-phenylboronic

acid-N-propylsilylamino silica shown in figure 2.2 is a commercially

available phenyl boronic acid phase. Applications of this phase are

discussed in more detail in chapter 4. The non-specific or secondary

interactions available to the sample are demonstrated in figure 2.2b.

Molecules may form electrostatic interactions at the ionized sites.

They may interact through hydrogen bonding or dispersive forces.

These non-specific interactions may stabilize a covalent bond, but

they may also cause non-specific retention of unwanted molecules.

c;BCQ0

2 d CQf)2 ~

Cl HH Cl lIf Cl / I 1/1

HCH HCH HCH HCH HCH \ / I / I HCH HCH HCH HCH HCH / \ \ \ \

HCH HCH HCH HCH HCH \ / \ / / Si-O-Si-O-H H H-O- 5i-0-Si-0-5i-0-H I, \ /'

0, 0, 0 0 0 \ \ 1

Si 51 5i 5i Si 5i 5i 5i 5i

t \/\/ \1 \/\1\1\/\/ o 000 0 000

HVOROPHOBIC. w""

+ rQI .ARGE'TRANSFER

....... N~B,.OH H I ~ HVOROGEN + OH'" BONOING

. ,1~ ..... OH

IONIC -0- e~""OH OM

Figure 2.2 Prc:.pylaminophenylboronic acid and the secondary interactions occurring on this phase.

27

28

Non-specific retention may be eliminated by changing the

mobile phase to reduce the strength of the non-specific interactions.

Increasing the solvent strength decreases retention due to dispersion

forces but may increase retention of ionic compounds. Increasing

ionic strength will decrease retention due to electrostatic interac­

tions but may increase retention due to dispersion forces. Changes in

the pH will affect the ionization of bQ~h the stationary phase and the

sample, thus affecting the electrostatic interactions. Non-specific

retention of molecules on residual silanols may be reduced by the

addition of an amine to the mobile phase. The amines will interact

strongly with the silanols, competing with the sample for this inter­

action. Since the mobile phase additive is in a much higher concen­

tration than the sample, it will effectively eliminate retention due

to silanol interaction. The silanols are often reacted with tri­

methylchlorosilane. This endcapping reaction removes most of the

available residual silanols but increases the lipophilicity of the

surface.

N-phenylboronic acid-N'-propylsilylurea silica is shown in

figure 2.3. The difference in this phase is the hydrophilicity of the

surface. On this surface, the urea forms a hydrogen bonding barrier

that restricts molecules from intercalating to the silica surface. In

addition to the hydrogen bonding of the urea group the residual

silanols are also capable of hydrogen bonding. There are two to three

layers of water molecules associated with the surface silanols (10).

The water associated with the surface silanols is highly structured

29

g-(OO) 2

U. HO,PH NIl 0 N /!\,H \ II I HO C=O ----- HN- C ,O=C

/ H \, \ HNH HN NH NH / / \ \

HCH HCH HCH HCH \ \ / / HCH HCH HCH HCH / / / \ \

HCH HCH HCH HCH HCH \ / '\ / /

HO-Si-O-Si-OH H0-5i-O-Si- O-Si-OH I \ H / \ ,

o a '0 0 0

\ " " '\ '\ Si Si Si Si Si Si Si Si Si / \/ \/ \/ \1 \/\/ \/ \/ 000000 000

Figure 2.3 Propylureaphenylboronic acii immobilized on silica.

30

through hydrogen bonding, and this structure will extend past the

propyl linkage to merge with the water that is hydrogen bonded to the

urea group. In order to interact with the surface silanols, the

molecule would have to break the hydrogen bonds and disrupt the

structure of the water molecule. In addition, the residual amine

groups that remain unreacted are capable of interaction with the

silanols and will compete with the sample for the interaction.

Procedures

Several silica based adsorbents were prepared using several

active functionalities and two different types of modification

methods. Elemental analysis, diagnostic chromatography and solid

phase extraction were employed to characterize the phases.

Modification Methods

Modification of the silica was done in two steps. The silica

was first modified with a reactive organic moiety, either an amino­

propyl, thiopropyl or a chloropropyl silane, followed by reaction with

the covalently reactive species, the disulfide, phenylmercurial,

phenylarsonous or phenyl boronic acid group.

Reactions With Bare Silica. To prepare aminopropy1 bonded

silica, forty micron irregular silica gel with 60 A average porosity

was acid washed, rinsed with deionized water and dried overnight at

80°C. The silica gel was suspended in dried 80% carbon tetrachloride

20% methanol. For each gram of silica, 1.2 mi11iequiva1ents of

triethoxyaminopropy1 silane (Petrach) was diluted by a factor of five

31

with dried carbon tetrachloride. This was added slowly to the silica

gel suspension. The reaction mixture was covered and left on the

shaker table for 6 hours. The reaction mixture was filtered and the

silica was washed thoroughly with SO% carbon tetrachloride 20% meth­

anol with a final rinse with methanol. The silica was dried with

suction, then placed in an SO°C oven overnight. This reaction is

shown in figure 2.4.

Mercaptopropyl and chloropropyl bonded silicas were prepared

in the same manner as the aminopropyl bonded silica using triethoxy­

mercaptopropyl silane and triethoxychloropropyl silane (Petrach) as

the silating reagents. These reactions are shown in figure 2.5 and

2.6.

Aminopropyl based bonded silicas. To prepare N-phenylboronic

acid-N-propylsilylamine silica (figure 2.7), chloropropyl silica was

reacted with m-aminopheny1boronic acid. For each gram of silica, 1

milliequivalent of m-aminophenylboronic acid was used. The reactants

are slurried in an aqueous solution with triethylamine as an acid

scavenger and left on a shaker table for 6 hours.

Activation of the mercaptopropy1 bonded silica. Mercapto­

propyl bonded silica oxidizes easily, so care was taken to insure this

phase is kept in a non-oxidizing atmosphere. This phase was reacted

with 2,2'-dithiodipyridine (Aldrich), figure 2.S. 2,2-dithiodipyri­

dine was dissolved in deaerated methanol. Thiopropyl modified silica

was added to the solution, and the reaction mixture was stirred

overnight. The phase was washed with methanol and dried overnight at

SO°C.

+ H H H H H

I I 1// 0, \ 0\ 0\ 0,

Si Si Si Si Si Si Si Si Si /\/\/\/\/\1\/\/\/ 00000 0 000

HNH HNH HNH HNH HNH I / I / /

HCH HCH HCH HCH HCH \ / / I / HCH HCH HCH Ht:H HCH / \ \ \ \

HCH HCH HCH HCH HCH

_ \Si-O-S:-'O-H H H-O - 'Si.-o-Sr- 0 -Si:O-H I , \ / \ 00000 \ \ \ , \

5i 5i 5i 5i 51 5i 51 5i 5i / \/ ,/\1 \/ \/ \1 \1\ /

o 0 0 0 0 0 0 0 0

32

Figure 2.4 Reaction of triethoxyaminopropyl silane with silica.

(CH3CH20)351CH2CH2CH2CI

+ H H H H H

I I I I / 0 0 0 0 0

\ '\ '\ \ '\ 5i 51 51 51 51 51 5i 5i 51

1\/ \./,\/\/ \ / \ / \ I \/ 0 0 0 0 0 0 0 0 0

~ Cl Cl CI Cl CI

/ / / / / HCH HCH HCH HCH HCH

\ / I / I HCH HCH HCH HCH HCH / \ \ \ \

HCH HCH HCH HCH HCH \ / \ / / 5i-O-Sl-0-H H H-O-51-0-Si-O-5i-O-H

I '\ \ 1 '\ o 0 0 0 0 \ \ \ \ /

51 51 51 5i 51 51 5i 51 51 / \/\/ \/\/\/\/\1\/ 600000000

33

Figure 2.5 Reaction of triethoxychloropropyl silane with silica.

H H I /

0 0

(CH3CH20)3SiC~CH2CH2SH

+

H H / /

0 0

H /

0

'SI , \ \ \

5i Si Si SI Si Si 51 5i I \ / \ / \ / ,I \ I \ I " / \ /

0 0 0 0 0 0 0 0 . 0

SH)H 5H 5H 5H I / / I

HCH HCH HCH HCH HCH \ I / I / j'H HC~ HC~ HC~ H~

HeH HCH HCH HCH HCH \ / \ / /

Si 0 SI-0-H H H-O-Si-O-Si-O-Si-OooH / \ \ I \

o 0 0, O. 0 \ \ \ \

5i 51 Si 51 51 51 5i SI 51 . / \ / \ / \. / '\ / ,/ \ /., / \ I 000000000

Figure 2.6 Reaction of triethoxymercaptopropyl silane with silica.

34

CI CI CI CI CI / / / / /

HeH HCH JH HeH HCH \ 1 1 / HeH HeH HCH HCH HeH / \ \ \ \

~ ItCH HeH HCH HeH HCH \ / , / /

iJ(oH)') SI-O-SI- O-H H H- 0- SI-O-SI-O-SI-O-H / \ \ 1 ,

0 0 0 0 , 0, \ , /

SI SI 51 51 SI 51 51 51 51 ('/\/\/ \1\1\1\1\/

000 0 0 0 0 0

d(af)2 Cl HH Cl NH CI

/ / / / / HCH HCH HCH HCH HCH

\ 1 / / / HeH HCH HeH HCH HCH / \ \ \ \.

HCH HCH HeH HeH HCH \ / ,/ / SI-O-SI-O-H H H-0-SI-0-51-0-SI-0-H

/ \ \ 1 \ o 0 0 0 .J , \ \ , /

51 51 51 SI 51 51 51 SI SI / ,/\/ \/\/\/\/\1\/ o 0 000 0 0 0 0

Figure 2.7 Reaction of chloropropyl bonded silica with m­aminophenylboronic acid to form propylaminophenylboronic acid bonded silica.

35

36

SH JH SH SH SH I / / /

HOt HCH HCH HeH HeH \ / / / /

@-s~~ ItCH HCH HeH HCH H~ I \ \ \ HeH HCH HCH HeH HCH

) \ / , / / - 51-0-51-0- H H H-0-Sl-0-51-0 -51-0-tf I \ \ I \ "-

0 0 0, 0 0 \ \ \ \

51 51 51 51 51 51 51 51 51 \ / \ / \ 1 \/ \ /, / ,/, /' I "-

o 0 000 0 0 0 0

5H S/S@ S/S-@ 5H SH

I / / / I HeH HCH . HCH HeH HeH \ / / / I

/HCH HCH HeH HeH HCH

\ \ \. \ HeH HeH HeH HeH HCH

\ / '/ / -51-0-51-0-H H H-0-51-0-SI-0-SI-Oif I \ , I \ , o 0 0, 0 0 \ \ \ \

51 51 51 51 51 51 51 51 51 \/\/\1\/\1'\./\/\/\ I' o 0 000 0 0 0 0

Figure 2.8 Reaction of 2,2-dithiodipyrirline with mercaptopropyl bonded silica to form the mixed disulfide phase.

37

Propylurea based bonded silicas. To prepare N-Phenylboronic

acid-N'-propylsilylurea silica (figure 2.9), aminopropyl bonded silica

was dried for 3 hours at sooe and cooled to room temperature in a

desiccator. The aminopropyl bonded silica was activated by reaction

with N,N-carbonyldiimidizole in methylene chloride. Triethylamine was

added as an acid scavenger. This reaction is left for 3 hours at room

temperature. The.activated silica is removed from the solution by

filtration and washed with methylene chloride and dimethylsulfoxide

(DMSO). The silica is added to a solution of m-aminophenylboronic

acid hemisulfate in DMSO. This reaction is heated to 40 0 e and left

for 24 hours. The silica is filtered and then placed in a solution of

ammonia saturated DMSO and reacted for three hours. The ammonia will

react with any residual carbo imide sites. The final product is

filtered and washed with 50% DMSO, water, 0.02 M hydrochloric acid, 1

M sodium chloride and again with water. The product is dried at room

temperature.

To prepare N-phenylmercury-N'-propylsilylurea silica, (figure

2.10), aminopropyl silica is activated in the same manner as for

N-phenylboronic acid-N'-propylsilylurea silica (figure 2.9). The

activated silica gel is added to a solution of p-aminophenylmercuric

acetate in DMSO. The reaction is heated to 40 0 e for 24 hours. The

silica is filtered and placed in a solution of ammonia saturated DMSO

and stirred for 3 hours at room temperature. The finished product is

rinsed and dried just as the previous reaction.

l1li HIIH HIIH-

/ / I ltC' ~HeH ? ,H /

HeH HeH HCH / \ \ \ ()

HCH HeH HCH HeH ~O \ / ,I'

_ Si-0.51-G-H H H-O- SI-0-51-0' ,'N

/ \ \ / \ 00000

~,;

\ \ \ \ \ SI SI SI 51 51 51 51 51 SI 1\/\/\1\1'\/\/\/\1 000 0 0 0 0 0 0

Q 0 QQ ~~

Nt. HIIIt "" HHH 1 / I /

HCH' HeH HCH ItCH ,/ I / HCH HCH HCH HCH

1 \ " ItCH HCH HCH HCH \ / \ /

-SI-0.51-0-H" H-O- 51-0·SI-0-/ \ , /"

o 0 0 0 0 , '\. '\. \ \ ~ ~ ~ g ~ ~ ~ ~ ~ /\ /,/\1,1 \/ \1 \/ \/

d. 0 0 0 0 0 0 0 0

d(QIl'1. 4~ ~ 1:.'/ e.o ~'Q /

HH HIlI HH HHH / / / 1

HCH HCH HCH He"

HCH ~ ~ ~ / \ \ \

HCN HCH HCH HCH \ / \ / I

-51-0.51-0-H" H-O- 51-0·SI- 0-51· / \ \ / \ I

o 0 0 0 0 " " \ , " 51 SI 51 51 51 51 51 51 SI 1\/\/\/\/\/\1'\/\1 000000000

>

>

Figure 2.9 Reaction of propylamino bonded silica with 1,1'­carbodiimidizole followed by m-aminophenylboronic acid to form propylureaphenylboronic acid bonded silica.

38

Q () I ro "0 I

NH HHH NH HNH / / / /

HeH HCH HCH HeH \ / / I HCH HCH HCH HeH / \ \ \

HeH HeH HCH HeH \ / \ / I

H2N-@-~ -5i-0-51-0- H H H- 0 - 5i· 0-5i- O-SI-/ '\ \ / \ ,

0 0 0 0 0 \ \ \ \ \

Si Si 5i Si Si Si 5i Si Si /\/\/\/'\/'/'\1'\1\1

o 0 0 0 0 0 I) a 0

Q '-0 I '-0 I NH HUH NH HNH

/ / / / HeH HeH HCH HeH

\ / 1 / HeH HCH HeH HeH / \ \ \

HeH HeH . HCH HeH \ / \ / J

-5i-0-5i-0- H H H- 0 - 5i· 0-5i- 0-5i-/, \ / \ I

o 0 0 0 0 \ \ \ \ \

5i 5i 5i 5i Si 5i 5i Si 5i 1\/\/\/,\/,/,\/,/\/ 000000000

~

Figure 2.10 Reaction of a cor activated aminopropy1 phase with p-aminopheny1mercurial hydroxide.

39

40

To prepare N-phenylarseneoxide-N'propylsilylurea silica

(Figure 2.11), aminopropyl bonded silica is activated with N,N'­

carbonyldiimidazole as in the previous reaction. The activated

silica is then reacted with p-aminophenylarsine oxide in DMSO for

three hours at room temperature. The silica is filtered and placed in

a solution of ammonia saturated DMSO and stirred for 3 hours at room

temperature. The finished product is rinsed and drie~ just as the

previous reactions.

A ureidopropyl phase (figure 2.12) was prepared by activating

aminopropyl bonded silica with N,N'-carbonyldiimidazole as in the

previous reaction. The silica is filtered and placed in a solution of

ammonia saturated DMSO and stirred for 3 hours at room temperature.

The finished product is rinsed and dried as described previously.

In order to assure that these phases are consistent, the same

batch of activated aminopropyl silica was used in all three reactions.

Characterization of Bonded Silicas

Elemental Analysis. The surfaces were characterized at the

through elemental analysis both at the University of Arizona and at

Analytichem (Harbor City, CA). The results are found in table 2.1.

This information is complemented with the ICP emission data for boron,

mercury and arsenic. These data allow for the determination of the

the different types of groups bound to the surface.

In the preparation of the propylaminoPBA phase (figure 2.7),

it is possible for it to contain 2 bonded moieties, the chloropropyl

group and the propylaminoPBA moiety. From the elemental analysis

41

Q [) Coro c·o I I

NH HNH NH HNH I / / /

IICH HCH HeH HCH \ / I / HCH HeH HeH HCH

'7~""IIIO / \ \ ,

HCH HeH HeH HCH \ / \ / I )

-Si-O-Si-O-H It H-O- Si· O-Si- O-Si-I " \ / \ I

0 0 0 0 0 \ \ \ \ " Si Si' Si Si Si Si Si Si Si /\1\1\1,,1'1'1'1"1

0 0 0 a 0 0 0 0 0

Q ~. C·O CIO I I

NH HNH NH HNH I / I /

HOt HCH HeH HCH \ / I / HCH HeH HCH HCH / \ \'

HCH HCH HeH HeH \ / \ / I

-Si-O-Si-O-H It H-O- Si· O-Si- O-Si-/ \ \ / \ I

o 0 0 0 a \ \ \ \ \

Si Si Si Si Si Si Si Si Si 1\/\1\1,,1'1'1'1,,1 00 a 000 000

Figure 2.11 Reaction of a CDr activated aminopropy1 phase with p-aminopheny1arseneoxide.

Q f) C-o C·o I I

NH HHH NH HNH / / / I

HCH HCH HCH HCH \ / / / HCH HCH HCH HCH / \ \ \

HCH HCH HCH HeH \ / I \ /

-SI-O-SI-O- H H H- 0 - SI· O-SI- O-SI-/ '\ \ / \ I

0 0 0 0 0 \ \ \ \ \

SI SI SI SI SI Si SI Si Si /\/\/\/\/,/'\/'\/\/

0 0 0 0 0 0 0 0 0

KV" H~H e-o e.o I I

NH HNH NH HNH / / / I

HCH HeH HCH HCH \ / / / HCH HCH HCH HCH / \ \ \

HCH HCH HCM HCM \ / \ / I

-SI-O-SI-O-H H H-O- SI· O-Si- O-SI-/ '\ \ / \ I

o 0 0 0 0 \ \ \ \ \

51 SI Si SI SI 51 Si SI 5i /\/\/\/,\/'/'\/'\/\/ 000000000

42

NH3

;,.

Figure 2.12 Reaction of a CDr activated aminopropyl phase with ammonia to form a propylurea phase.

43

Table 2.1 Elemental analysis of modified silicas.

Bonded Silica %C %H %N %M

Aminopropyl 5.44 1. 52 1.77

Mercaptopropyl 3.16 0.74 0.00

Mixed Disulfide 6.80 0.66 0.13

PropylaminoPBA 5.95 0.97 0.47

PropylureaPBA 9.26 1.71 2.40 0.53!!

PropylureaHgOH 7.50 1. 52 2.19 6.43h

PropylureaAsO 7.76 1.52 2.37 1.22£

Propy1urea 6.73 1.59 2.99

a %Boron, b %Mercury, c %Arsenic

44

(table 2.1), the concentration of both can be determined. The chloro­

propyl phase initially bound with 0.98 millimoles/gram. Approximately

35% reacted with m-aminophenylboronic acid to give 0.34 millimoles/

gram PBA and 0.64 millimoles/gram unreacted chloropropyl. The cova­

lent interactions in this phase are occurring on a lipophilic hydro­

carbon backbone.

The preparation of the hydrophilic propylureaPBA phase is

more complicated, and there are four possibilities of surface

moieties: the PBA phase, an unactivated aminopropyl, an activated

aminopropyl and the urea propyl phase. Assuming that the ammonia in

DMSO was able to react with all the activated sites, we can assume

that no activated aminopropyl sites remain. The aminopropyl bonded

silica contains 1.42 millimoles/gram. During the reaction of this

phase to the ureapropyl phenylboronic acid phase (figure 2.9), it is

unlikely that any of the bonded groups were removed, so the total

number of bonded moieties is the same for both phases. From the boron

analysis, the concentration of the PBA functionality is 0.49

millimoles/gram, or 35% of the bonded functionalities. Using the

carbon data and solving for the remaining variables indicates that

0.32 millimoles/gram or 23% remain unreacted as the aminopropyl and

0.61 millimoles/gram or 43% are ureapropyl. The unreacted aminopropyl

groups are capable of ionizing, and the hydrophilic urea groups will

provide a completely different environment than in the hydrocarbon

phase.

The ureapropylsilane silica shown in figure 2.12 may have

three functionalities on the surface: the unreacted aminopropyl, the

45

activated aminopropyl and the ureapropyl functionality. Once again,

the activated aminopropyl has been fully reacted with the ammonia in

DMSO. Using the carbon data and the assumption that 1.42 millimoles/

gram is the total number of functionalities on the surface, it can be

determined that 0.34 millimoles/gram or 24% remain as the aminopropyl

and 1.08 millimoles/gram or 76% are reacted to the ureapropyl.

Using the same assumptions, the data can be used to determine

the concentrations of the different functionalities on the propyl­

ureaorganometallic phases. For the phenylmercurial phase, there are

0.32 milliequivalents/gram or 20% of the propylureamercurial function­

ality, 0.48 milliequivalents/gram or 34% of the ureidopropyl function­

ality and 0.62 milliequivalents/gram or 46% aminopropyl groups. The

arsonous acid phase has 0.16 milliequivalents/gram or 10.6% propyl­

ureaphenylarsonous acid, 0.79 milliequivalents/gram or 52%

uriedopropyl groups and 0.56 milliequivalents/gram or 37% aminopropyl

groups.

Solid Phase Extraction. The capacity of the phases was

determined using solid phase extraction. Plastic syringe barrels were

filled with 100 milligrams of aminopropylPBA bonded silica. The

silica was washed first with methanol or acetonitrile, then with water

and finally the mobile phase. Approximately two micromoles of sample

dissolved in the mobile phase was applied to the column. The effluent

from the column was captured in a 5 ml volumetric flask and the

concentration of sample was determined through HPLC with a UV detec­

tor. HPLC was used to separate the compound of interest from the

solvent. During the aspiration of the sample into the volumetric

flask, some of the organic solvent would evaporate. This change in

the solvent made simple UV experiments to determine concentration

difficult. UV data was used to confirm the identification of eluted

components.

46

Solid phase extraction experiments were used to determine the

capacity of the phases and the changes in capacity when the mobile

phase was changed.

Diagnostic Chromatography. The chromatographic results

indicate relative strengths of interactions and the trends in inter­

actions as the ionic strength, solvent strength, solvent composition

and pH are varied. In all cases, only one variable is changed while

the rest are held constant.

Careful consideration of mobile phases is needed to assure

that the mobile phase is not competing with the sample for the active

sites on the surface. Alcohols are expected to bind to PBA (14),

therefore acetonitrile was chosen as the organic modifier. Data in

table 2.2 confirms that as the methanol concentration increases, the

retention of benzene does not decrease. This indicates that the

lipophilic nature of the stationary phase is increasing with the

increase in methanol concentration. The increase in lipophilicity of

the phase is caused by the covalent interaction between the methanol

and the phenylboronic acid. The interaction of PBA with monofunc­

tional compounds is weak and is only noticed at high concentrations

(greater than 10%).

Table 2.2 Capacity factors of several probe molecules on

propylaminoPBA with methanol as the organic modifier

Solvent

Concentration

5

10

15

5

10

25

5

10

25

Probe

Benzene

Benzene

Benzene

Phenol

Phenol

Phenol

Nitrate

Nitrate

Nitrate

Capacity

Factor

27.6

40.5

4l.5

7.3

6.3

6.8

0.0

0.0

0.0

47

48

The choice of buffer composition is based on the buffer

capacity, range of pH and a limited interaction with the stationary

phase. Phosphate buffers bind with PBA and are not acceptable buffers

(14). Ammonium acetate is a commonly used buffer for liquid chromato­

graphy. While both ammonia and carboxylic acids show some interaction

with PBA (chapter 4), these interactions are weaker than that of

phosphate since each contains only a single interaction site while

phosphate has four. The ammonium acetate also demonstrates inter­

action with the mercury phase (20). However a great number of mole­

cules will act as ligands to mercury but the interaction with thiols

is so strong it easily displaces the buffer even with the difference

in concentrations.

The modified silica was packed into a 2 cm long 2 mm inner

diameter column. A second column containing octadecylsilyl silica was

placed in system just prior to the injection valve. In the case of

high pH mobile phases, the CIS column would dissolve into the mobile

phase so that the mobile phase would be equilibrated with silanes and

would not dissolve the phase of interest.

Mobile phases were prepared using doubly distilled filtered

water and filtered reagent grade acetonitrile or methanol. All

mobile phases are a volume/volume percent and were measured using

graduated cylinders. Stock solutions of the ammonium acetate and KCl

were prepared by weighing the salts and dissolving them in doubly

distilled filtered water. These solutions were then added to mobile

phases to assure that the buffer concentration and the the ionic

strength were consistent.

49

Chromatographic experiments were done using an IBM 9533

chromatograph, with a 9522 detector. This is a 254nm fixed wavelength

ultraviolet detector. The data collection was an IBM 9000 lab com­

puter or a strip chart recorder.

Retention of the molecules is reported as capacity factor

(eq. 1.5). The void volume was measured as the volume needed to

elute nitrate from the column. This volume was assumed to be the

same for all pH ranges but was measured at a pH of 9.

Several probe molecules were chosen to determine the relative

strengths of interactions occurring at the surface. These probe

molecules include compounds expected to interact with the covalent

functionality as well as molecules which should not interact. The

retention of these molecules on different phases and when using

different mobile phases will allow for determination of the factors

which affect selectivity of the surface and which factors must be

controlled when using or preparing the silica surface.

Summary

Diagnostic chromatography has been used to characterize two

different classes of modified silica surfaces. Careful choice of

probe molecules allows for the determination of the relative

strengths of interactions and the effects of different mobile phases

and modifications on the versatility of silica adsorbents.

Of the commonly used solid sorbents, silica is physically

strong and easily modified. The reactivity of silica may be a

50

limitation. Unless the surface is completely reacted the unreacted

sites of the silica can interact with many compounds causing irrever­

sible adsorption. Controlling the modification of the surface is

important when using silica.

CHAPTER 3

EFFECTS OF MODIFICATION

While the covalent interaction is the primary interaction and

the control of this interaction will be most important in covalent

separations, the control of the non-specific or secondary interactions

will also be important. Secondary interactions are most commonly

dispersive, hydrogen bonding or electrostatic interactions that may

occur between the compound of interest and the sorbent, the matrix and

the sorbent or between the compound of interest and the matrix. Non­

specific interactions may be necessary for the formation of the

covalent bond, or they may cause retention of contaminants. Optimiz­

ing non-specific interactions is necessary for efficient separations.

Control of the Non-specific Interactions

Mobile Phase

Traditionally, non-specific interactions are controlled by

altering the mobile phase. Dispersive interactions can be reduced by

increasing the concentration of the organic modifier. Ionic inter­

actions can be reduced by increasing the ionic strength. Electro­

static repulsion is minimized by an increase in ionic strength.

Changing the pH to suppress ionization of the sample or the stationary

phase may also eliminate electrostatic interactions. Interactions

51

52

with residual silanols has been reduced by the addition of an amine to

compete with the sample for these sites.

In some cases, altering the mobile phase may not be possible.

In many biochemical applications, increasing the organic modifier may

not be an option. The compound of interest is often denatured ir­

reversibly at concentrations of 20% acetonitrile or less. Increasing

the ionic strength above 0.1 molar is difficult when working with

liquid chromatographic pumping systems because of the salt deposits

that block lines or ruin seals. In automated systems the ionic

strength must be kept to a minimum to increase the lifetime of the

pumps.

Stationary Phase

In addition to altering the mobile phase, the stationary phase

may also be altered. Residual silanols have been reacted with tri­

methylchlorosilane. This molecule is smaller than other silanes and

is believed to react with silanols that are sterically not available

to the larger molecules. Even the most carefully reacted silica

surfaces will have residual silanols. These silanols may irreversibly

adsorb samples so that any residual silanols must be sterically unable

to interact with the compound of interest.

A more versatile modification of silica is needed in order to

overcome the present limitations while still allowing for the effi­

ciency of the small rigid particles. A new modification method with a

hydrophilic surface that restricts samples from the residual silanols

has been developed. This propylurea phase (figure 2.3) is described

in chapter two.

Experimental

Diagnostic chromatography as described in chapter two was

employed to compare the differences in a traditionally modified

53

silica and the new modification. The effects of solvent strength,

ionic strength and pH on dispersive, hydrogen bonding and electro­

static interactions were determined using molecules that probe each of

these interactions.

The modified silicas characterized in this work all contain

some ionizable species. These are shown in figure 3.1. In addition.

each contains residual silanols that ionize at pH 6 and above. PBA

will ionize through the acceptance of a hydroxyl group with a pKa of

9.2 (24). The meta amino functionality which binds the PBA to the

surface in the lipophilic phase will accept a proton with a pKa

similar to that of N-ethylaniline which is 5.12 (25). The urea or

hydrophilic phases have two types of nitrogen atoms, the aminopropyl,

and the urea nitrogens. The pKa of propyl amine is 10.7 (25) and the

pKa of urea is 0.10. Due to the low pKa of urea, the urea type

nitrogens of the phase are not ionizing. Conversely, due to the high

pKa of propylamine, the amine groups on the surface are ionized

throughout the experiments.

As the pH of the solution is altered, the different bonded

moieties will ionize. A study of retention vs pH will allow for the

54

S(al)3"

Q "hydrophilic" "lipophilic" d(01I) 3 NH ~ f ~ I C-O C-O

;m; I , NH NH Cl NH Cl I I I I I

~ fIlz ~Z ~ 1HZ 9Hz ~ ~ ~ ~ C~ 9~ - - - , -f2 0

~ 0 C~ ~

0 CHZ 0 C~ I I I I I I I

Q«(JI)2 Basic pH

d(01I) 2 NH ~ f ~ I C-O C-O

~ I , .

NH NH Cl NH Cl I I I I I

~ fH2 ~Z ~ ~ 9~ ~ - F2 - ~ ~ ~ <fiz - -~ 0

~ 0

~ ~ 0 f2 0

~ I I I I QB«(JI) 2 Neutral pH

~ J 0(011)2 NH F2 f ~ I C-O Cm()

~ I I NH NH Cl NHz Cl r , I

~ I

~ ~. ~ fHz 7HZ

~ H ~ ~ ~z ~2 H 9H2 ~ ~z CHZ 0 f2 0

~ ~ 6

~ 0 CHZ I I I I I I

Acidic pH Figure 3.1 ~o\Uparison of ionization of the two modification

methods.

determination of the types surface charges on the phase. The pH of

the solution was varied between 3.4 and 9.2, and the retention of

several molecules able to probe different interactions was measured.

For these data, the ionic strength was held at 0.1 molar.

55

In order to assure that the bonded silica is solvated, it is

necessary that the mobile phase contain some organic modifier. In

this work 5% acetonitrile was chosen. Below 5%, it is difficult to

keep the organic strength constant throughout the experiments. Above

5% acetonitrile, the capability of the mobile phase to dissolve salts

decreases, and the compatibility of the mobile phase with biopo1ymers

is diminished. Therefore, 5% was chosen as a good trade off between

reproducibility, compatibility and practicality.

Results and Discussion

Dispersive Interactions

Lipophi1icity of the Bonded Phases. The difference in

lipophilicity in the two phases is demonstrated by the retention of

benzene and toluene (figure 3.2). Toluene has a capacity factor of 10

on aminopropylPBA and 2 on propylureaPBA. Benzene has a capacity

factor of 5 on aminopropylPBA and 1 on propy1ureaPBA. Both molecules

show a lower capacity factor on propylureaPBA with a difference of a

factor of five. AminopropylPBA is more lipophilic and the dispersive

interactions are much stronger on this phase causing the higher

capacity factor. The variations in capacity factor with a change in

pH may be due to slight differences in the concentration of the

BENZENE 7

.,

5

4

:.. 3

2

: ; :~ 0

3 5 7 9

Cl proPvNrea ~ar prop~A

TOLUENE 13

IZ

n

~ 10

9

IS

:.. 7

., 5

4

3 ==---: 2 ... -CJ Ei EI -r--

3 5 7 9

Cl propylur .. propylonaPar 0 propytP8A

Figure 3.2 Retention of benzene and toluene on three modified silicas as a function of pH.

56

57

organic modifier in the mobile phase which would cause noticeable

differences in k', or due to the difference in surface charge as the

pH changes. A highly charged surface will exclude hydrocarbon

samples and decrease retention. Electrostatic interactions are a

better measure of charged surface sites than lipophilic interactions.

Effects of Solvent Strength. The effect of increasing the

concentration of acetonitrile on the lipophilic surface is shown in

figure 3.3. Above 10%, the addition of more organic modifier results

in only small decreases in retention for these small probe molecules.

The effects of increasing ionic strength would be greater for larger

molecules that are well retained on the phase at 5%.

Electrostatic Interactions

Anion Exchange Sites. As the pH of the mobile phase changes,

the charge on the surface will change. Ionizable sites on the

surface or the immobilized functionality will be charged or neutral

due to the pH of the surrounding solvent. While the pH at the surface

will not always be the same as the mobile phase, an increase in mobile

phase pH will increase the pH at the surface. At low pH, amine

functionalities on the surface will be positively charged. As the pH

of the surface increases, the amines may neutralize or the the ioniza­

tion of surface silanols and the PBA functionality will produce a

negatively charged surface. Retention of anions is expected to be

high at low pH and decrease as the surface becomes negative. Benzoic

acid (figure 3.4), phenylacetic acid (figure 3.4), and benzenesulfonic

acid (figure 3.5) probe the anion exchange sites on the surface.

SOL VENT STRENGTH DEPENDENCE

6,---------------------------------------------------~

5

4

:.. 3

2

5 7 9 II 15 f7 19 21 23 2S

" ACETOHITRLE Q BENZENE o ~E A NITRATE

Figure 3.3 Effects of acetonitrile concentration on the retention of benzene, aniline, phenol, and nitrate on propylaminophenylboronic acid bonded silica.

58

BENZOIC ACID

7,----------------------------------------------------, 6

5

4

2

0 ~ 5 7 9

0 PfO\>yIon. ptOfJ."....,el' prupytP8A

PHENYLACETIC ACID 7

6

5

4

:.. ~

2

" "

O+-------r------,-------r------.-------~----~~----~ 5 7 9

Figure 3.4 Retention of benzoic acid and phenylacetic a:id on three bonded silicas as a function of pH.

59

60

BENZENESULFONIC ACID

8~--------~-----------------------------r

.~

..

2

0 ---i :J 5 7 9

0 propytura. _yturaJ _yIP9A

PHENYL TRIMETHYLAMMONIUM 4S

40

:IS

:JO

2S

:.. 20

IS

10

S

0 :J 5 7 9

0 propyka'aa _ytura""ett 0 propylP9A

Figure 3.5 Effects of pH on the retention of benzenesulfonic acid and phenyltrimethylammonium on three modified silicas.

61

Benzenesulfonic acid retention shows a continuous decrease in reten­

tion as the pH is increased. Higher retention of this probe on the

hydrophilic surface at low pH indicates that this phase has more

positively charged sites than the lipophilic surface up to a pH of 7.

The increase in anion exchange capacity is due to the residual amino­

propyl groups on the silica. The decrease in retention at

neutral pH is not due to the suppressed ionization of these sites but

to the ionization of the PBA to its anionic form and the ionization of

the surface silanols. These sites compete with the sample for inter­

action with the cations. Above a pH of 7, all the phases are nega­

tively charged and the anions are not retained on the surface.

Benzoic acid and phenylacetic acid show the same trend as benzene­

sulfonic acid. The difference is that the carboxylic acids with pKa

near 4.0 are not ionized at the low pH, so they are not retained

through electrostatic interactions. As the pH increases to the pKa of

these probes, the retention increases.

Cation Exchange Sites. Phenyltrimethylammonium (figure 3.5),

aniline (figure 3.6) and benzyl amine (figure 3.6) will probe the

cation exchange characteristics of the surface. The cation exchange

sites will occur with the ionization of residual silanols or the

immobilized PBA. Elemental analysis shows that the hydrophilic phase

has 0.49 millimoles/gram of PBA and the lipophilic phase contains 0.34

millimoles/gram. If the PBA is the only important site, we would

expect more cationic exchange sites on the lipophilic phase. The

amines on the surface are capable of interaction with the PBA to

:..

0

:..

=

ANILINE 2..8

2.4

2.2

2

L8

L8

L4

1.2

\ 0.0

0.0

0.4

Q.2

\ 0

3 S 7 9

PIOCIvlllrea prc,PVtVINfi'9r 0 propvtP9A

BENZYLAMINE 7

8

S ~~ ..

2

G-- -----.. 9 9

~ 3 S 7 9

propykn_ PIOCIvtureePal" 0 PIOCIyiPtlA

Figure 3.6 Retention of aniline and benzylamine on three modified silicas as a function of pH ..

62

63

decrease the pKa (26) or to neutralize the the charge. The only

amines on the lipophilic phase is the meta amine which does not

interact with the PBA groups and is not protonated at the pH where

the PBA ionizes. The hydrophilic phase contains a large number of

amine functionalities that are protonated at high pH and are able to

interact with the PBA. The affects of this amine will be to neutral­

ize the surface.

The quaternary amine, phenyltrimethylammonium, shows little

retention on any of the phases at low pH. As the pH increases, the

retention increases, but the retention on the lipophilic phase is much

higher. Comparing these data to the data of catechol (figure 4.3),

the catechol is retained much better on the hydrophilic phase.

Catechol interacts only with the ionized PBA indicating that the

hydrophilic phase contains more ionized PBA functionality. The

residual silanols and the propyl amine functionalities are important

sites for electrostatic interactions. The residual silanols are

available to the sample on the lipophilic phase, but the organized

water of the hydrophilic phase block these sites.

Aniline with a pka of 4.6 is not retained at low pH (less than

5), since the surface and the probe are positively charged. At

neutral pH (between 5 and 8), aniline interacts with the silica

surface on the lipophilic phase or with PBA on either phase. At high

pH, the hydroxyl groups compete for the PBA site and the retention

drops.

Benzylamine with a pKa of 9.33 (25) remains charged throughout

the pH region studied and there is no change in retention on the

64

lipophilic phase. The drop in retention at pH 9 for the hydrophilic

phase is due to competition for the boronic acid from hydroxyl groups.

This interaction will be discussed later.

Effects of Ionic Strength. Changing the ionic strength of

the mobile phase will allow further probing of the electrostatic

interactions. The retention of benzene, toluene, phenol, benzyl

alcohol and aniline is not changed as the ionic strength is increased.

These molecules are not charged and electrostatic interactions are not

expected. Benzylamine (figure 3.7) is a cation at pH of 7. There is

little change in the retention as the ionic strength is increased.

Benzyl amine is not retained through electrostatic interactions. The

anions, benzoic acid (figure 3.7) phenylacetic acid (figure 3.8) and

benzenesulfonic acid (figure 3.8), demonstrate ionic strength depend­

ence on the urea phases. These molecules are retained on the amino­

propyl groups of the urea phases. This interaction is weak, however,

because of the competition for these sites from the residual silanols

and the PBA functionalities. At pH 7, the lipophilic phase has little

cationic character and the anions show little retention and no ionic

strength dependence.

The retention of phenyltrimethylammonium (figure 3.9) on the

lipophilic phase decreases as the ionic strength increases, indicating

that the cation is retained through electrostatic interactions. On

the urea phase, there is no ionic strength dependence indicating that

there is no electrostatic interaction. The negative charges on the

surface are not likely due to ionized PBA groups because these

Figure 3.8 Effects of ionic strength on the retention of phenylacetic acid and benzenesulfonic acid on propylurea (a), propylureaphenylboronic acid (b), propylaminophenylboronic acid (c) and aminopropyl bonded silica (d).

1% pH 7

111

a

k 6

4 a e e

&= z

a 12.' 31.5 63

ion ie stzoongth

pH 9

18

38

~ 28

18

8--------~--------------~--------------~ 8 68

ionic strength 115

67

Fi~ure 3.9 Ionic strength dependence on the retention of phenyltrimethylammonium on propylureaphenylboronic acid (a) and propylaminophenylboronic acid (b).

68

groups, would not be significantly charged at this pH. This retention

is therefore due to the availability of ionized residual si1ano1s on

the surface.

If the pH is increased to nine, the retention of the cation

shows ionic strength dependence for both surfaces (figure 3.9). At

this pH, the PBA functionality is charged on both phases, but the k'

for the cation is higher on the hydrophobic phase indicating an

increase in ion exchange sites. The increase in the number of cation

exchange sites is due to the cationic aminopropy1 groups on the

hydrophilic surface which interact with the anions on the surface, and

to the availability of the si1ano1s to the sample on the lipophilic

surface. The residual si1ano1s are not available to the sample on the

lipophilic surface due to the hydrogen bonding of the urea groups and

the organization of water which forms a barrier to the molecule.

Hydrogen Bonding

The hydrogen bonding capability of the surface was probed with

phenol (figure 3.10) and benzyl alcohol (figure 3.10). In 95% water

the hydrogen bonding characteristics are expected to be insignificant.

The retention of the molecules is similar to retention of benzylamine

(figure 3.6). On the lipophilic phase the retention is not changed as

the pH changes. On this surface these molecules are retained at the

silica surface. The polar end of the molecule interacts with the

surface silanols or with the associated water while the lipophilic

phenyl ring interacts with the hydrocarbon chains (9). This inter­

action is sufficiently removed from the PBA that the ionization of the

69

PHENOL 2.1 2

1.9

1.8 L7

US

L5 L4

L3

:.. L2 LI

o.s o..a 0.7 o..a o.s 0.4

Q.!I :) 5 7 9

0 propyluru PIOf'y/AIrotIi'att 0 propyfP8A

BENZYL ALCOHOL 2.:)

2.2 2.1 2

L9 1.8 L7 US L5 L4 :.. L3 L2 LI I

0.9 0 o..a 0.7 o..a o.s 0.4

:) 5 7 9

0 propykn- prop.,....J 0 PIOf'yIPfIA

Fig~re 3.10 Retention of phenol and benzyl alcohol on three bonded silicas as a function of pH.

70

PBA does not affect the retention of these molecules. On the hydro­

philic phase, a drop in the retention at pH 9.2 for both the amines

and alcohols indicates that these molecules are retained at the PBA

and the ionization of this site has removed the site of interaction

for these probes. These molecules are retained through interaction

with the boron. The k' values are smaller for the hydrophilic than

the lipophilic phase indicating that the interaction with the PBA

molecule is weaker than the interaction at the surface. On the hydro­

philic phase the molecules are unable to intercalate past the urea

functionalities to the silica surface.

Conclusions

The retention of these probe molecules containing only one

functionality can be used to understand the complex interactions of

larger molecules with more than one functional group.

Controlling Lipophilic Interactions

Lipophilic interactions were much stronger on the propyl­

aminoPBA phase. These interactions are due to the availability of the

propyl linkage for interaction with the sample. For biochemical

interactions, increasing the organic modifier is not an acceptable

way to eliminate these interactions. While the addition of a small

amount of organic modifier was able to minimize these interactions for

the small molecule, the much larger biopolymers may have stronger

interactions with the surface. The advantage of the more hydrophilic

71

propylureaPBA phase is the minimization of these interactions without

excluding lipophilic molecules from the surface.

Controlling Electrostatic Interactions

The residual aminopropyl groups on the propylureaPBA surface

help control the anionic nature of the surface. The cationic amine

sites are capable of neutralizing the surface most likely with direct

interaction with the residual silanols or with the ionized PBA sites.

The ability of the residual amines on the lipophilic surface

also helps to assure that anions are not excluded from the surface.

While it is important that the retention of molecules be minimized to

restrict irreversible adsorption or nonspecific retention of mole­

cules, molecules excluded from the surface will not be able to

covalently react with the functionality. The presence of the amines

keeps exclusion to a minimum.

Interaction of the molecule with the residual silanols is

always unwanted. The interactions with these sites is often very

strong and often irreversible. Since it is not possible to complete­

ly react all of the residual silanols, and since most endcapping

reagents increase the lipophilicity of the resulting surface, develop­

ing a surface that will restrict the molecule from the surface is

important. The hydrogen bonding of the urea group as well as the

highly structured water molecules imbibed in the lipophilic phase

restrict sample from the residual silanols and the lipophilic inter­

actions of the propyl functionality.

72

Controlling the retention of molecules on the surface depends

on the ability to reproduce a surface exactly. Since there are fewer

steps and variables in the preparation of the propylaminoPBA, the

preparation of this phase is more consistent from batch to batch. In

addition, since the unreacted amine functionalities on the propylurea

phases interact with other functionalities on the phase, an increase

in the unreacted sites will have an even greater effect on the inter­

actions, both primary and secondary.

CHAPTER 4

PHENYLBORONIC ACID

Introduction

Boronic acid and phenylboronic acid (PBA) readily and reversi­

bly react with molecules containing polar functionalities. The

interaction of borate with sugars was first reported by Biot in 1842

(14). While a boric acid with four reactive sites is able to interact

with two sugar molecules, organoboronic acids, with three sites, can

only interact with a single sugar molecule (27). The organoboronic

acids are often used because of the simplicity of the one to one

interaction.

While diols and polyols are the most common ligands to PBA,

there are several types of polar functionalities capable of interac­

tion with the boronates. 1,2-hydroxyacids (28,29) and 1,2-hydroxy­

amines (30) as well as several other functionalities are expected to

bind to PBA. Examples of substances expected to react with PBA are

found in table 4.1.

Acid Base Chemistry

The interaction of the PBA with polar functionalities is

intrinsically dependent on the acid base chemistry of the PBA as well

as the acid base characteristics of the sample and its matrix.

73

Table 4.1 Molecules that interact with PBA.

Compound Class

Polyhydroxy

1 ,2 dihydroxy

1,3 hydroxy

Hydroxy acids

Diketo

Hydroxyamines

Mercaptoalcohols

Examples

Mannitol, Carbohydrates

Catechol, Tannis

Tris Pyridoxine

Lactate, Benzillic, Salicylic acids

Benzil

Aminophenol

2-mercapto 'benzylalcohol

74

75

Therefore, the pH of the solution can be used to control the selectiv­

ity of the PBA, provided the effects of pH on the formation of the

covalent complex are understood.

Phenylboronic acid is a Lewis acid, accepting a hydroxyl group

as shown in figure 4.1 with a pKa of 8.86 (27,31) in an aqueous

solution. The pKa is alte-ed by substituent groups on the ring thus

the pKa of immobilized PBA will depend upon the immobilization proce­

dure. In general, PBA is bound through an amine functionality in the

meta position. Maestas et al (24) report the pKa of PBA bound in this

way as 9.2. The increase in pKa is attributed to the meta amine

functionality. The addition of a nitro group on the benzene ring in

the ortho position will decrease the pKa of the boronate substantial­

ly. Transfer RNA has been immobilized on ortho-nitro substituted

meta-aminophenylboronic acid at neutral pH (32). Residual amine

groups on a surface reduce the apparent pKa of immobilized PBA through

a surface buffering effect or direct interaction with a PBA (26).

Compounds Expected to React with PBA

Monofunctionalities such as amines and alcohols are reported

to interact with PBA(l). However, monofunctional interactions are

weaker than those that involve di- or trifunctionalities, because the

di or tri functionality will interact with two or three sites to

stabilize the interaction.

Difunctionalities, such as diols, must have the correct

configuration. They must be cis, coaxial or able to rotate into the

same plane in order to physically reach the PBA. In addition the

EQUIL1UIVU lON

~B .... OH+2H'O I

~3:'OH+ • "'OH

OH OH

REACTION

lQl R,@ O H '""H.,.... 0 e OH + +-- e 0 B:: H~"'H .B.# 'l._ R i "OH R HO\\~ A .t<L

. OM 0 \""1' I'

RH ELUTION

76

R

+ HO-f"", H H01""H

R

Figure 4.1 Equilibration of PBA to its tetrahedral form. Reaction with diols and the reverse reaction when the pH is dropped.

77

geometry of the PBA will determine the ability to bind different

molecules. At low pH, the PBA is in the trigonal planar neutral form

and will form charge transfer interactions. At high pH, the PBA is a

tetrahedral anion, and it is capable of undergoing condensation

reactions with dio1s and po1yols. Since the pH governs the conforma­

tion and the conformation governs the reactivity, the pH of the mobile

phase will control the reactions occurring on an immobilized PBA.

The charge on the compound is also important. Molecules that

bind to PBA in the anion form may be excluded from PBA sorbents if the

compound of interest is also an anion. Cationic functionalities on

the compound of interest may stabilize the covalent interactions with

the anionic form through electrostatic interactions.

Dio1s. A great deal of attention has been paid to the

interaction of PBA with cis or l,2-dio1s, especially catecholamines,

nuc1eotides and t-RNA. These compounds form a covalent complex with

PBA under basic conditions, when the PBA is in the ionized or tetrahe­

dral form. Under acidic conditions, the l,2-diols do not interact

with the PBA. Under these conditions, the PBA is in the trigonal form

with a 120 0 bond angle and the ring strain is believed to be too great

to form the complex (14).

l,3-dio1s interact with PBA in the same way as the l,2-diols;

however, the l,3-diol has more freedom and spends less time with its

hydroxy groups in the same plane. The decrease in the number of

molecules with the correct configuration weakens the interaction.

78

Po1yhydroxy1s. Po1yhydroxyl containing compounds interact

with PBA similarly to vicinal diols except that the possibility of a

tridentate interaction is possible for compounds with the correct

configuration. (14,27) This extra interaction further stabilizes the

complex. This type of interaction explains the high formation con­

stants reported for molecules capable of tridentate interactions (14).

Molecules able to form tridentate complexes have higher formation

complexes than the bidentate compounds.

Hydroxyacids. The interactions of alpha-hydroxy aliphatic

acids or ortho-hydroxy aromatic acids with PBA have been studied by

several researchers (28,29). These interact at low pH to form a

complex but are excluded at high pH due to the electrostatic repul­

sions. Recognition of these electrostatic repulsions is important in

the primary interactions of PBA. Lactic acid has also been shown to

interact with PBA in solutions of high ionic strength and pH of 2 and

6 (28).

Hydroxyamines. Amines are able to donate an electron pair to

the boron atom in its trigonal state (30). The bond formed between

nitrogen and boron is weaker than the bond between oxygen and boron.

The interaction of the nitrogen with boronic acid is further stabi­

lized by the interaction of the hydroxy group with the PBA. At low

pH, the protons compete with the boron for the nitrogen, and at high

pH, the hydroxide competes with the nitrogen for the boron, causing

the strongest interactions at neutral pH. The interaction of amines

with boron is further stabilized by a vicinal hydroxy group that will

react with the boron in the same fashion as the diols.

79

Other Functiona1ities. The interaction of PBA with dike to

and triketo compounds with PBA has been reported (24). Understanding

the effects of pH on the retention of these molecules and other

related compounds will allow for a better understanding of the effects

of pH on biologically significant molecules.

Isolation of Compounds

The reaction of PBA with compounds containing polar function­

a1ities has been studied using pH titration (28) and pH depression

(33) techniques, as well as, infrared (29) and Raman (34) spectro­

scopic methods in solution. Complexation with cell walls has been

determined using fluorescent boronic acids (35). Pheny1boronic acid

solid supports have been used to isolate catecho1amines (36, 37,38),

nuc1eosides (39,40,41), tRNA (32,36,42,43), carbohydrates (41), and

enzymes (44). Automated systems using PBA have been developed (36,

40,45) for the isolation of compounds. A great deal of work on PBA

has been done to demonstrate its ability to isolate compounds, but few

studies to determine the factors controlling the interactions have

been carried out.

The association constants for PBA and several compounds have

been measured by spectroscopic methods in unbuffered solutions. As

mentioned, the pH dependence of complexation with PBA is important in

controlling the isolation and subsequent elution of compounds. A

study using UV spectroscopy to determine the effects of pH on the

complexation of PBA with different dihydroxy substituted naphthalenes

80

(46), shows that the retention of these molecules will reach a maximum

at a pH of 9-10 and that this maximum depends on the molecule.

While spectroscopic data on PBA in solution can yield a

great deal of information, the best method for determining conditions

for efficient separations is the use of diagnostic chromatography.

Determining the factors which affect retention of molecules on immobi­

lized PBA using solvents suitable for chromatography will be more

informative for developing chromatographic methods.

Experimental

The strength of interactions of immobilized PBA was investi­

gated in two different experiments. Solid phase extraction experi­

ments and diagnostic chromatography were used to determine the rela­

tive strengths of interactions of several probe molecules from a pH of

3 to a pH of 9.

Solid Phase Extraction

SPE experiments as described in chapter two were carried out

using application buffers covering a wide range of pH. These buffers

are listed in table 4.2. The amount of sample eluted in 3 milliliters

(k'-30) was measured using HPLC. The HPLC analysis separated the

sample from the application buffer allowing for better reproducibility

and accuracy.

Diagnostic Chromatography

The retention of several probe molecules over a range of pH

was measured by liquid chromatography as described in chapter two.

Table 4.2

pH

2.0

4.1

6.1

9.0

Buffers for solid phase extraction

Components

30 mM Dich1oroacetic Acid, 17 mM NaOH

8.7 mM Acetic Acid, 1.8 mM Sodium Acetate

4.35 mM Acetic Acid, 94.4 mM Sodium Acetate

10 mM Ammonium Acetate, 9.9 mM NaOH

81

82

The different modification methods for PBA immobilized on silica

(figure 4.2) described in chapter two were used, and the effects of

modification as well as the effects of pH on the covalent interaction

were determined.

Results and Discussion

Retention Dependence on pH

Diol. A comparison of the retention of catechol (figure 4.3)

on a surface containing no PBA with surfaces with immobilized PBA

demonstrates the retention of catechol through covalent interactions.

At pH above 7, the retention of catechol is much less on the phase

containing no PBA, indicating that a covalent bond between catechol

and PBA is occurring on the PBA phases. The difference in the reten­

tion is too large to be explained by retention through non-specific

interactions. The increase in retention as the pH increases is

consistent with the spectroscopic data (4.10). The increase in pH

increases number of PBA functionalities in the ionized or tetrahedral

form; therefore, the catechol is interacting with the ionized PBA, but

not with the trigonal planer molecule. The lack of interaction with

the trigonal planar form is attributed to the ring strain from the

120 0 bond angle of this form.

The interaction of PBA with catechol depends on the modifica­

tion methods. Catechol interaction is stronger on the hydrophilic

urea phase (figure 4.2b) than on the lipophilic propyl phase (figure

4.2a) in the pH range from 4.4 to 7. Equation 1.5 demonstrates that

the retention of molecules due to covalent bonds is proportional to

the number of reactive sites. The number of pheny1boronic sites on

;==<B«(Jd) 2 cfat

)2 '-(

f r ,H ~ HCH HCH tiCH HCH

\ \ I / HCH HCH HCH HCH / / / \ \

HCH HCH HCH HCH HCH \ / \ I /

HO-5i-0-5i-Cii HQ-Si-0-5i- Q-SI-OH I \ H / \

o '0 0 0 '\ " " \ " 5i 5i 5i 5i 5i 5i 5i 5i 5i /\/\/\/\/\/\1\/\/

o 0 0 0 0 0 000

,-((Jd) 2

\..(NH 0 H ':JL,.,pH \ 1/ r Hoe"oH C=O ----- HN- C .O.C

1 H \ \ HNH HIl NH ~IH

HC/H / \ \ HCH HCH. HOI

\ \ I / HCH HCH HCH HCH / / \ \ /

HCH HCH HCH HCH HCH \ 1 \ I /

HO-5i-0-5i-(- HQ-Si-0-51- Q-5i-OH / \ '\. / \ . ~ 0 I.i 0 0

\ " " " " 5i /51 5i 5i 51 5i 5i 5i ~l /\ \/\/\/\/\1\/\/-COOO 00 Q 00

Figure 4.2 Two PBA modified silica phases, a lipophilic phase (a) and a hydrophilic phase (b).

83

CATECHOL 100

90

eo

70

60

:.. 50

40

30

20

10

0 3 5 7 9

a propyl .... ll""PylunaPBlH

o propylPBA

-~-fc5\ . HO-fc5\

~ "OH ._)::::Y "B HO

HO

-N-\Q/ H 0JQJ / Be

HO/ '0 0

Figure 4.3 Retention of catechol on three modified silicas as a function of pH.

84

85

the surface is determined from the elemental analysis (chapter two);

the lipophilic phase contains 0.34 milliequivalents/gram and the

hydrophilic phase contains 0.48 milliequivalents/gram of PBA. The

capacity of the hydrophilic phase is 1.4 times that of the lipophilic

phase, but the capacity factor of the hydrophilic phase at pH 7 is 3

times the capacity factor of the lipophilic phase. The difference in

retention is not solely due to the difference in the number of PBA

groups bound to the surface.

Catechol interacts only with the PBA that is ionized. There­

fore, the number of active sites is not the number of PBA function­

alities on the surface, but the number of these sites that are ion­

ized. The acid base equilibrium of any acid depends on its environ­

ment and the non-specific interactions (chapter three) have shown that

the there is a significant difference in the environment of the PBA on

the differently modified surfaces. In addition, the presence of amine

functionalities on the surface have been shown to cause a decrease in

the pKa of PBA through direct interaction or through a surface buffer­

ing effect (26). A significant number of amines are left unreacted on

the hydrophilic phase, while there are no amines in the lipophilic

phase that are capable of interaction with the PBA. The residual

amines on the surface are causing a decrease in the pKa of the PBA so

that more of the PBA on the hydrophilic surface are ionized at a

given mobile phase pH.

The retention of the 1,3 diol, 2-hydroxybenzyl alcohol (figure

4.4) shows the same trends as the retention of catechol. This mole­

cule interacts with the tetrahedral form of PBA, and the

100

90

80

70

00

:.. 50

40

30

20

10

0 3

Cl propyl ....

-~-«S\ . ~,OH

8 , OH

2-HYDROXYBENZYL ALCOHOL

5

PfOPyl ... aPBr

HO-@ HOCH z

7 9

o propylPBA

-N-!c3\ ~ H ~I /0 0 Be

/ "-HO OCHz

Figure 4.4 Retention of 2-hydroxybenzyl alcohol as a function of pH.

86

difference in the ionization of PBA due to the modification of the

silica is apparent.

87

Comparing the capacity factor for catechol, a cis diol, with

the 1,3-diol it is apparent that the catechol interaction is occurs at

lower pH. At pH of 7 on the hydrophilic phase, the capacity factor of

the catechol is 3.7 times the capacity factor of 2-hydroxybenzyl

alcohol. The capacity factor of the 1,3-diol at pH of 8 is the same

as the capacity factor for catechol at a pH of 7. For the 1,3-diol to

have the same retention volume as the 1,2-diol, the pH, and therefore,

the number of ionized PBA sites is increased. Thus, the reaction

constant, K (equation 1.5), for the 1,3-diol is weaker under these

conditions. This weakness is due to the ability of the 1,3-diol to

rotate the hydroxy groups into more than one configuration. The cis

diol retains its hydroxyl groups in the correct geometry for reaction

with the PBA. The hydroxyl groups of the 1,3 diol are capable of

rotating into the correct geometry, but not all of the 1,3 diols will

have the correct configuration at anyone time.

The trends in the chromatographic data fit with the trends in

the solid phase extraction data for 2-hydroxybenzyl alcohol (figure

4.5), except that at a pH of 9, the diol should have been strongly

retained. The difference may be in the equilibration of the station­

ary phase. In order to assure that the phase was equilibrated, it was

washed with 2 ml of O.02M NaOH. The continuous flow of the chromato­

graphic system allows for easy equilibration of the phase, but with

SPE, care must be taken to assure that the phase is equilibrated.

Figure 4.5

i ~

~ ~

SdLWPHASEEXTRAcnON roo

00

00

70

60 b 50

40

30

W

~

0 2 4 6 8

~H

Figure 4.5 Retention of aminothiolphenol (a), 2-hydroxybenzyl alcohol (b) and 2-aminobenzyl alcohol (c) on commercially available PBA columns by solid phase extraction.

88

89

Hydroxyamines. Amines have been reported to interact with the

unfilled orbital on boronic acid in the trigonal planar form (figure

4.6 and 4.7). This interaction is stabilized by a hydroxy group that

is capable of interaction with the PBA to form a five or six membered

ring. Competition between protons and the PBA for the amine site and

competition of hydroxide ions and the amine for the boron cause this

interaction to be strongest at neutral pH. The retention data for

aminophenol (figure 4.6) and 2-aminobenzyl alcohol (figure 4.7)

demonstrate an increase in retention at neutral pH. However, this

retention is very low and these compounds could not be isolated by

covalent chromatography under these condition. In fact, the retention

of these molecules is lower than the retention of the monofunction­

ali ties (phenol (figure 3.10) and aniline (figure 3.6». The SPE data

(figure 4.5) is consistent with the chromatography data, but the

retention of these probes at pH 6 is higher. This increase is due to

the buffer used in SPE at this pH which does not contain a competing

amine. The retention is still too low to be useful for covalent

chromatography.

The conditions chosen for the SPE and the chromatographic

experiments were not optimized for the retention of amines. The

ammonia in the buffer used for covalent chromatography is competing

with the amines for the boron interaction. Carboxylic acids also

complex with PBA. The ammonium acetate buffer may be competing for

the PBA sites.

The SPE retention of aminothiolphenol (figure 4.5) is much

greater than that of the hydroxyamines. This increase in retention is

90

AMINOPHENOl 1.8

L7

us L5

L4

L3

L2

U

:.. I

0.9

0..8

0.7

0,8

o.s 0.4

o.:l

Q.2

:J 5 7 9

c PfOIIvturea propwturoaPett Q 1W000vtPBA

+

-@,oH -@OH ©rN~ OH Sr B' ~ HO' 'OM

Figure 4.6 Retention of aminophenol as a function of pH.

91

2-AMINOBENZYL ALCOHOL. L9

La

L7

La

L5

L4

L3

:.. 1.2

L1

0.9

0.8

0.7

0.8

o.s 3 5 7 9

c:J propyluree o propyIPBA

Figure 4.7 Retention of 2-aminobenzyl alcohol as a function of pH.

92

due in part to the lipophilic nature of this molecule. This molecule

is retained through lipophilic interaction with the propyl chains.

Once this molecule is solvated by the propyl chains, the covalent

interaction occurs. The ability of this molecule to be well solvated

by the phase allows for the stabilization of the covalent bond. The

data demonstrate that this interaction predominates at neutral pH for

the same reasons as the aminophenol.

Mercaptoalcohols. The retention of 2-mercaptobenzyl alcohol

(figure 4.8) is different on all the phases so each phase will be

discussed separately. On the ureaPBA phase, the retention is similar

to the retention of l,3-diol at low pH (figure 4.4), but at a pH of

above 5.6, the diol is well retained and the mercaptoalcohol retention

decreases. This decrease is due to the ionization of the thiol and

the PBA to anions. The electrostatic repulsions cause the molecule to

be excluded from the stationary phase and no covalent bonds may form.

On the lipophilic phase, at this solvent strength the probe is

well retained through the non-specific interactions due to the

lipophilicity of the probe. The dispersion forces are not the only

interaction responsible for the retention on this phase, but these

interactions stabilize the formation of a complex as shown in figure

4.11. At high pH where the PBA is ionized as well as the thiol, the

probe is excluded from the phase due to electrostatic repulsion.

Salicyclic Acid and Related Compounds. The effect of pH on

the retention salicylic acid (figure 4.9) is the opposite of the

effect on diol compounds. For the salicylic acid, retention decreases

as the pH is increased in this case the covalent complex forms with

:..

100

go

00

70

00

50

40

:xJ

20

10

0 3

CJ propylur ••

2-MERCAPTOBENZYL ALCOHOL

5

propyUuPol"

7

-<9!~OH Be

Ha' 'OH

9

o propyIPBA

Figure 4.8 Retention of 2-mercaptobenzyl alcohol as a function of pH.

93

94

SALICYLIC ACID 100

90

eo

70

60

:.. 50

40

:JO

20

10

0 3 5 7 9

CJ propylun. PfOfJvUeel'afi 0 propylP9A

-~-©>/o~ -~-©> HO~ "OH

/B~ B e 0

0_

C HO O=C HO" 'OH II

I 0 OH

Figure 4.9 Retention of salicylic acid as a function of pH.

the trigonal planar form of PBA not with the ionized form. This is

consistent with the SPE data (figure 4.10) and with reported separa­

tions of alpha-hydroxy aliphatic acids (28,29). Some salicylic acid

with a pKa of 3.0 is ionized throughout the entire pH range studied.

At low pH, the salicylic acid complexes with the PBA as shown in

figure 4.9. As the pH increases, the PBA ionizes and the this mole­

cule is excluded from the surface due to electrostatic repulsion.

95

Comparing the retention of salicylic acid on the different

phases, a shift in the pKa is apparent. The PBA on the lipophilic

phase is ionizing at a higher pH than the hydrophilic phase. In

addition, the decrease in retention of salicylic acid occurs over a

wide pH range for the hydrophilic phase but is very sharp on the

lipophilic phase. The retention data for the hydrophilic phase

implies that the PBA ionizes over a wide range. This indicates that a

full range of PBA groups exist with the extremes being a PBA group

surrounded by arnines or interacting directly with an amine and a PBA

group completely removed from all amine functionalities and their

buffering effects. The decrease in retention on the lipophilic phase

is much sharper and occurs between 5.6 and 7. This pH is consistent

with the pH range for the ionization of silanols (47) which are shown

by the non-specific interactions (Chapter 3) to be available to the

sample. The ionization of the PBA occurs at a higher pH (24). The

electrostatic repulsions are occurring between the ionized surface

silanols and the salicylic acid and are excluding the probe from the

surface. Salicylic acid is unable to probe the ionization of the PBA

SOLID PHASE EXTRACTION

wo~-------------------------------------------------~

90

00 a 70

60

50

20

\0

o+-----.-----.-----~~~~-===~==:~~----~ 2 4 6 8

Cl SAUCYUC ACID pH

SAUCYLAMDE'

Figure 4.10 Salicylic acid (a) and salicylamide (b) retained on PBA as a function of pH.

96

on the hydrophobic phase, due to the exclusion of the probe from the

phase at the pH where PBA ionizes.

97

The interaction of salicylamide (figure 4.11) with PBA is much

weaker than the interaction of salicylic acid, but it follows the same

trend except that the amide is capable of interacting with the boron

through the oxygen or the nitrogen (48), as shown in figure 4.11. At

low pH, the interaction occurs through the keto group. As the pH

increases, the keto interaction decreases and the amine interaction

increases. The dip at pH 4.4 occurs because of the two types of

interactions. At high pH, hydroxide ions compete with both interac­

tions and the molecule is not retained. The SPE retention demon­

strates the same trends.

The pH dependence on the retention of thiosalicylic acid

(figure 4.12) shows different behavior on the different PBA phases.

The difference in retention is due to the differences in non-specific

interactions. Thiosalicylic interacts with the PBA as shown in figure

4.12. This interaction is weaker than the interaction of salicylic

acid. The propylPBA phase is lipophilic and the thiosalicylic acid

will intercalate into this phase at pH 3, above this the thiosalicylic

acid ionizes (49) and the lipophilic interactions decrease. At a pH of

5.6, the ionization of the silanols excludes the probe from the phase.

The hydrophilic phase excludes the lipophilic probe molecule

at low pH where the molecule is unionized. When the thiosalicylic

acid ionizes, the positively charged amine groups on the surface allow

for solvation and covalent interaction. As the pH increases, the PBA

ionizes and the thiosalicylic acid is no longer retained.

Cl

98

SALICYLAMIDE

7~-------------------------------------------

8

5

4

3

2

0 3 5 7 9

propyl .... propyl ... ePB1H

0 propylPBA

-~-©; o-@ -~-@/o-@ -~-@ S/ 0 ' "OH

/B", B0 / '" Hci 'OH

HO O::~ HO HzN-C

NHz 8

Figure 4.11 Retention of salicylamide as a function of pH on phenylboronic acid.

100

90

60

70

60

:.. 50

40

:lO

20

10

0 :l

0 propy" •••

THIOSALlCYLlC ACID

5 7 9

prOpy .... ePBIH

0 propylPBA

-~~/o-fc5\ /,\ );::Y

HO S-C

" o

:0-PJ s-c

" o

Figure 4.12 Retention of thiosa1icy1ic acid as a function of pH on pheny1boronic acid.

99

100

Diketone. Benzil interaction with PBA (Figure 4.13) is

dependent on the non-specific interactions. Benzil is a lipophilic

molecule, and the dispersive interactions available to this probe on

the lipophilic phase stabilize the interactions with the PBA (figure

4.13). At pH 9, the hydroxide ions compete for the site on the boron

and benzil is no longer retained. The lipophilic molecule is excluded

from the charged surface. On the hydrophilic phase at a very low pH

the phase has enough of a positive nature to exclude benzil. As the

pH increases the benzil is retained as shown in figure 4.20, but above

a pH of 5, the surface amines or hydroxide ions compete with the probe

for the Boronic acid and it is no longer retained.

The SPE retention of benzil (figure 4.14) is consistent with

the chromatography data. For these data, 20% acetonitrile was used

and the dispersion forces were weakened causing a decrease in the

retention.

Keto-carboxylic acid. The pH dependence of the retention of

benzillic acid on the lipophilic phase (figure 4.15) is consistent

with the SPE data and with salicylic acid. It is retained at low pH

but elutes when the residual silanols ionize due to electrostatic

repulsion. On the hydrophilic phase, the retention data are similar

to benzil. Benzillic acid is unretained at a low pH due to poor

solubility in the phase. At a pH of 4.4, the carboxylic acid is

ionized but the PBA is not. For this small pH range, the molecule is

well solvated by the surface and the covalent interaction occurs. At

higher pH, about 5.6, the PBA begins to ionize and the covalent

:..

Cl

100

90

80

70

80

50

40

:10

20

10

0 3 5

propyl .... +

BENZIL

7

propylunoaPBIH

0

-<Q(,OH Be

Ho' 'OH

9

Pl'opylPBA

Figure 4.13 Retention of benzi1 as a function of pH on pheny1boronic acid.

101

i ~ ~

SOLID PHASE EXTRACTION 100

90

80 a 70

80

50

40

:lO b 20

10

0 2 4 6 8

CJ BENlD. ~Zl.UCACID

Figure 4.14 Retention of benzil (a) and benzillic acid (b) on PBA as a function of pH.

102

:..

0

103

BENZILLIC ACID 100

90

60

70

60

50

40

30

20

10

0 3 5 7 9

propylur •• propylureaPoiH

0 propyiPBA

-~-<Qf./~-@ -Z-<Qf ~ ,OH Ho-~~ B0 / '\ 4'c H<1 'OH HO 0 bH C o~ be

Figure 4.15 Retention of benzi11ic acid as a function of pH on pheny1boronic acid.

interaction cannot occur. At higher pH, 7 and above, the electro­

static repulsion excludes the molecule from the phase.

Effects of Modification on Retention

104

In comparison, the two PBA phases show a great deal of simi­

larity in the capability of forming covalent interactions with

compounds containing polar functionalities. The differences in the

local environment do cause disparities in the retention values. These

differences are due to the effects of the non-specific interactions on

the covalent bond and the apparent shift in the pKa of PBA.

Secondary Interactions

The secondary interactions discussed in detail in chapter

three may support or resist the formation of the covalent bond. The

formation of the covalent bond requires the diffusion of the molecule

of interest to the surface and the solvation of the molecule of

interest into the stationary phase prior to the reaction. The solva­

tion of a species of interest is more important for the reaction of

the ketones and thiol compounds than the other polar functionalities.

An understanding of the effects of the secondary interactions will

allow for the optimization of solvents for efficient separations.

Shift in pKa of PBA. The retention data for several probe

molecules indicate a shift in the pKa of PBA on the urea phase. While

the chromatographic data does not give an absolute pKa for either

phase, the shift in pKa is apparent. This shift is due to electronic

and solvation effects. The amine functionality in the meta position

of the PBA increase its pKa of the boronic acid from 8.6 to 9.2. The

105

urea functionality in the meta position would also increase the pKa of

PBA but not to the same extent as the amine functionality. The

solvation effects of the different modification methods will also

affect the pKa . It has been reported (26) that amine functionalities

on the surface will lower the pKa of immobilized PBA. These amines

either interact with the PBA directly or cause a surface buffering

effect. These data do not distinguish between the two models.

A related effect of amines on the PBA is the increase in the

heterogeneity of the PBA groups. While a decrease in the retention of

salicylic acid or the increase in retention of the diols is very sharp

on the lipophilic phase, it is not sharp on the hydrophilic phase.

This difference is due to the availability of the amines, to different

PBA groups. While some groups may be isolated from amines others are

very close to one or more causing a disparity in the interaction

energies of PBA.

Demonstration of Isolation of Compounds

Once the effects of pH and modification on retention are

understood, efficient methods for the isolation of specific compounds

or classes of compounds can be developed. The sample matrix and the

wash solvent should be optimized to assure strong interaction of the

compound of interest with the stationary phase, while removing contam­

inants. The elution solvent should be chosen to assure that minimal

interaction is occurring.

To demonstrate the use of PBA sorbents, a 2 centimeter long

2mm id column was dry packed with 40 micron propylureaPBA bonded

106

silica and coupled to a 10 centimeter 10 micron octadecylsilyl modi­

fied silica column in a traditional liquid chromatographic system.

The system was equilibrated with a retaining solvent and the sample

was injected. Fifty column volumes of the retaining solvent was

washed through the column. The PBA column was then removed from the

system. The system without the PBA column was equilibrated with an

elution solvent and the PBA column was replaced in the system.

Diols. Figure 4.16 shows the isolation of diols. Trace A is

is the 2-hydroxybenzyl alcohol injected through the system at pH 4.4.

The retention time is 1.5 minutes. Trace B is the same compound

injected at a pH of 9.2. At time 1, the PBA column is removed and

the system equilibrated to pH of 4.4. At time 2, the PBA is returned

to the system and the peak at 1.5 minutes was saved. The UV spectra

of this peak compared to a standard (figure 4.17) confirms that this

peak is 2-hydroxybenzyl alcohol.

Hydroxyacid. A similar experiment with salicylic acid (figure

4.17) demonstrates the retention using a pH of 4.4 and elution with a

pH of 9.2. The tailing in the elution peak demonstrates slow kinetics

of the equilibration of PBA to the ionized form.

Conclusions

While PBA has been used for the isolation of compounds, the

effects of pH on the retention and elution of compounds was not well

understood (50). Understanding the effects of pH on the retention of

molecules will allow for more efficient use of this phase for the

immobilization of compounds. Recognizing that many types of polar

107

e

Time (min)

Figure 4.16 Demonstration of the retention of 1,3-dio1s on PBA.

PBA coupled to a C18 column using a mobile phase of 5% acetonitrile and .1M ionic strength and a flow rate of 1 m1/min. A) Mobile phase pH 4.4 and a sample of 2-hydroxybenzy1 alcohol. B) Injection of 2-hydroxybenzy1 alcohol at pH 9.2, after 2.5 minutes the PBA column is removed from the system and the mobile phase pH is dropped to pH 4.4 at 5.5 minutes the PBA column is returned to the system.

Time (min)

Figure 4.17 Demonstration of the retention of hydroxyacids on PBA.

108

PBA coupled to C1S column with a mobile phase of 5% acetonitrile and an ionic strength 0.1 M. and a flow rate of 1 m1/min. A. Injection of salicylic acid using a mobile phase of pH 9.2. B. Injection of salicylic acid with a pH of 4.4. After 2.5 minutes the PBA column is removed and the system is equilibrated with pH 9.2 and the PBA column is returned to the system after 5.5 minutes.

109

functionalities are capable of binding to PBA will allow for the use

of this phase to isolate these molecules. This recognition will also

help one to realize what types of contaminants will be retained on the

surface.

tVhile the pH controls the retention of compounds on the

surface, the modification of the surface can alter the pH range where

different molecules are retained.

The compounds studied interacted with the immobilized PBA in a

manner consistent with the solution data. Diols interact through a

condensation reaction with two hydroxy groups on the boron. This

interaction only occurs when the hydroxy groups on the diol are cis or

coplanar and when the boronic acid is in a tetrahedral form because of

the bond angle of the hydroxy groups. Since the diols only interact

with this form, the interaction increases as the pH increases.

Amines interact with PBA as Lewis bases donating electrons to

the empty orbital of the boron. This interaction may be stabilized

through the reaction of a vicinal diol, but under the conditions used,

the interaction was weak. The amines interact only with the trigonal

planar form of the PBA and protons compete with the boron for the

amine causing this interaction to be strongest at neutral pH.

The interaction of 2-mercaptobenzyl alcohol with PBA on the

hydrophilic phase seemed to follow the interaction of alcohols. The

ionization of the thiol group, however, eliminated the interaction due

to the electrostatic repulsion. On the lipophilic phase, the lipo­

philic compound was well solvated. The thiol interacts with the PBA

as a lewis base, and the alcohol can react with the hydroxy group

of the boron. This is a weak interaction, but it is stabilized

through the lipophilic interactions.

110

Salicylic acid interacts with the PBA through a carboxylic

acid and a hydroxy group. The carboxylic acid donates electrons to

the boronic acid in the trigonal planar form. The ionized PBA and the

carboxylic acid do not interact because of electrostatic repulsion.

As the pH increases and the PBA ionizes, the retention decreases.

Salicylamide interacts through a hydroxy group and the amide.

The amide has two possibilities of interaction, the keto group or the

amine group. The keto interaction predominates at pH 3.6, and the

amine interaction predominates at pH 5.6. At high pH, the hydroxide

ions compete with the amide for the boron site and the amide is not

retained.

Thiosalicylic acid, benzil and benzillic acid all interact

with PBA as Lewis bases, and none of these molecules contain a hydroxy

groups. The interaction of these molecules with PBA are dependent on

the modification method. Each of these compounds shows different

retention behavior for the two different modification methods. The

dispersive interactions are important in the solvation of these

molecules to stabilize the otherwise weak interactions with PBA.

CHAPTER 5

COVALENT CHROMATOGRAPHY OF THIOL CONTAINING COMPOUNDS

Sulfide and disulfide groups play an important role in the

structure and activity of biochemical compounds (51). The tertiary

structure of proteins depends in part on the formation of disulfide

bonds between cysteine residues, and the active sites in enzymes often

contain free thiols. In order to prepare synthetic biomolecules or to

understand their functions, an understanding of the thiol groups or

the disulfide linkage is needed (52). Several different reagents have

been developed to work with thiol chemistry (8,15,52,53,54,55). These

reagents all utilize the redox potentials of sulfide disulfide exc­

hanges to study thiol chemistry. In general, there are three types of

reagents traditionally used with thiols in solution chemistry, the

reducing agent, assay compounds and blocking compounds. Each of

these types of solution chemistry could be simplified with the

development of separation methods.

Traditional Thiol Chemistry

Several disulfide compounds have been employed to assay thiols

by colorimetric or spectroscopic methods (15,52,53,56). The number of

thiol groups may be utilized to determine the concentration of a thiol

containing compound or the extent of denaturation of a biomolecule

III

112

which in turn gives insight into the structure of the molecule. Free

thiols in solution may react with the disulfide forming a mixed

disulfide and a molecule that contains a chromophore or electrophore.

For example, 2,2-dithiodipyridine may be reacted with a free thiol to

form a mixed disulfide and 2-thiopyridone as shown in figure 5.1 (15).

Thiopyridone is almost exclusively in the tautomeric thio form with

the hydrogen on the nitrogen. This causes a large difference in the

UV spectra of the thiopyridone and its corresponding disulfide,

allowing one to spectrometrically follow the reaction of dithiodipy­

ridine with free thiols (18).

Much of the work involving thiols requires the use of a

reducing agent to reduce any disulfide bonds to thiols and to assure

that no reoxidation of the thiols occurs. The most commonly used

reducing agents are dithiothreitol (OTT), dithioerythritol (DTE) and

2-mercaptoethanol (ME) (57). The interaction of OTT with a disulfide

is shown in figure 5.1. The reducing agent is added in excess and

contains free thiols. The free thiols of the reducing agent will

interfere with the assay of the free thiols on the compound of inter­

est unless they are removed or blocked. Arsenite has been used to

block the thiols of OTT and DTE in solution by binding to the dithiols

and not the monothiols (54). The rates of formation and breakdown of

the complexes are needed to determine the number of uncomplexed thiols

at equilibrium. This assay could be simplified by the use of immobi­

lized arsonous acid to remove the dithiol from solution.

R-S-S-R • U - R-SH •

HO OH

J-\ . H R-SH

HO OH

113

D~-HO OH

@-S-S-@ 'ZR-SH - --'N~ 2 S"==/_" • R-S-S-R'

Figure 5.1 Reduction of disulfide bonds by dithiothreitol followed by reaction with 2,2-dithiodipyridine.

114

Due to the reactivity of thiols, these groups are often

blocked during reactions or storage. Reagents that reversibly react

with thiols are often used for this purpose.

Separations of Thiol Containing Compounds

It is often advantageous to separate the thiol containing

compounds from the non-thiol containing compounds. When working with

a mixture of peptides, the free thiols may need to be blocked. Prior

to the blocking reaction, it may be desirable to separate the thiol

containing peptides from the non-thiol containing peptides. Because

the amount of active enzyme is important, the separation of the active

from the denatured enzyme is of interest. When the active site of the

enzyme is a thiol, the active enzyme will contain an available and

active free thiol and the denatured enzyme will not. A separation of

thiol containing compounds from nonthiol containing compounds will

allow for the separation of the isolation of the active enzyme for

accurate measurements. When working with a reducing agent, some of

this reagent may have oxidized to the disulfide, separation of the

oxidized and reduced forms of DTT, DTE or ME will allow for accurate

measurements of the active reducing reagent (57). The use of a solid

adsorbent that reversibly reacts with the thiol containing compounds

to isolate these compounds on the surface will be demonstrated as a

quick, efficient method of separation.

In addition to these separations, the separation of a compound

of interest and the reducing reagent is often necessary or the excess

115

reducing agent will compete with the compound of interest for blocking

or assay reagents. When the reducing agent is extracted either by

solid phase chemistry or solution chemistry, care must be taken to

assure that the free thiol of the compound of interest is not

oxidized to the disulfide bond prior to the assay. Disulfides form

readily and the free thiols must be kept in a non-oxidizing

atmosphere. The environment of the sample is easily controlled when

using solid phase adsorbents by directly coupling the extraction

column to a second column or a reaction vessel.

Solid Phase Thiol Chemistry

The use of solid phase chemistry rather than solution chemis­

try provides simpler sample handling, especially with smaller sample

size. In addition, covalent chromatography has been shown to be an

effective method for the separation of thiol containing compounds from

non-thiol containing compounds (58). Several types of phases have

been employed for the immobilization of thiol containing compounds

(17,57,58,59). Thiols can then be stored on these phases to prevent

oxidation or the thiol may be eluted for further analysis.

Mixed disulfide. The reversibility of sulfide/disulfide

reactions is utilized to reversibly immobilize compounds of interest

on a solid support (15,16,17,18,60). A mercaptopropyl derivatized

surface is activated by reaction with 2,2'-dithiodipyridine (figure

2.8). A thiol containing compound introduced to the phase will

displace the thiopyridone and be immobilized on the column (figure

5.2). A competing thiol may then be used to elute the compound. This

z 0

~5-S-o + }- S-S-R ~ ..

R-SH • '"~ 0

'" a c sD

,N H

z

~S-S-R }-S-S-R' 0 .: + R~SH ~ '\ .. ..

R-SH

Figure 5.2 Retention, detection and elution of thiols using a mixed disulfide phase.

116

phase is useful in the isolation of compounds through SPE or the

storage of compounds with active thiol groups.

117

While the isolation of compounds is an important area of

separations research, it is also important to develop selective and

inexpensive methods of detection. The displacement of the thiopy­

ridone can also be used to improve the sensitivity of detection of

thiol containing compounds with a simple fixed wavelength ultraviolet

flow through detector. Extra column volume and the kinetics of the

displacement reaction cause wider peaks, but the trade off between

wider peaks and an increase in sensitivity is often acceptable.

Phenylmercurial. A second phase which is commonly used for

the separation of thiol and non-thiol containing compounds is a

phenylmercurial phase (figure 2.10). Mercurials are among the most

reactive and selective of the thiol reagents. Immobilized organomer­

curials interact with thiol containing compounds, immobilizing the

thiol compound on the surface (figure 5.3) allowing for an effective

method to separate thiol containing compounds from non-thiol contain­

ing compounds (61). A competing thiol or mercuric chloride can be

used to elute the compound of interest. An excellent review of

mercurial chemistry is available (20).

Organomercurials interact with thiols similar to mercuric

chloride (HgC12) or inorganic mercury through a bond with covalent and

ionic character. The interaction is generally somewhat weaker for the

organomercurials but may be stabilized through dispersive interac­

tions. Inorganic mercury interacts with four ligands, but only two

are bound tightly. The other two are bound weakly due possibly to a

I }-H9-S-R + R'-SH '\' }-H9-S-R'

R-SH

ll8

Figure 5.3 Retention and elution of thiols on an immobilized organomercurial.

119

change in bond configuration (20). Organomercurials are capable of

binding to three ligands but only one is bound tightly. Under the

conditions of covalent chromatography were there is an excess of

mercury sites, only one thiol will bind with the surface. The ligands

bound to the immobilized organomercurial during its use will depend on

the pH, the buffer components, the organic modifier and any other

molecules present in the mobile phase and their concentration. The

association constants for methyl mercuric ion with several ions is

given in table 5.1. The thiol compound is capable of displacing any

ligand bound to the mercury so the buffer choice and composition will

not affect the retention of thiols but it will affect nonspecific

interactions.

Organomercurials interact with a thiol functionality and are

selective for this group, not for the compound; however, the chemical

nature of the compound as well as the immobilized reagent will affect

the optimal conditions for the interaction (20). At high pH, the

interaction of mercury with thiols is expected to improve due to the

electrostatic forces resulting from the ionization of the thiol. The

electrostatic attraction between the anionic thiol group and the

cationic mercury would stabilize the formation of a covalent bond.

This is not always the case because of competition from hydroxide ions

and the increase in pH will affect the conformation of a protein

making the thiols less accessible. The effects of pH will be differ­

ent for every compound and the optimization of the phase for individ­

ual compounds may be necessary. However, optimum interaction usually

occurs in the pH range between 4 and 8.

Table 5.1

Ligand

C1

OH

Acetate

Ammonia

Cysteine

Dissociation Constants for Complexes with Methy1mercuric Acid

pK

5.4

9.5

3.6

8.4

15.7

120

121

Immobilized organomercurials have been used in the study of DNA

replication through the separation of newly synthesized DNA and RNA

from the parental strands (19,6~,63,64,65) and separations of syn­

thetic peptides (58) and proteins (66,67). It has also been used for

the isolation of compounds from natural products (68). The separation

of active enzymes from inactive enzymes (5.24) is possible on this

phase if the inactivity is due to an oxidized thiol from the active

site.

Pheny1arsonous Acid. Phenylarsonous acid (PAA) is the hydrat­

ed form of phenylarseneoxide also called phenylarsine oxide or

arsenosobenzene (figure 5.4). In aqueous solutions, the hydrated form

is dominant (19). Organoarsonous acids are capable of interacting

with thiols to form an arsenic thiol complex and water. These mole­

cules are capable of interaction with both mono and dithiol compounds.

Several studies have seen that the interaction of monothiols with PAA

are weaker than the interactions of dithiols (69). This is due to the

ring formed by the dithio1 complex. The ring structure stabilizes the

complex. The stability of the ring depends on the size of the ring,

with the most stable rings having 5 or 6 atoms and the least stable

having 7. With rings larger than 7, the interaction is stronger

because the disruption of the intramolecular interactions of the

dithiol with rings larger than 7, the disruption of the intramolecular

interactions decreases as the ring size increases (70). While the 7

membered ring may be the least stable under certain conditions, this

ring structure is still stronger than the interaction with monothiol

compounds.

A

)OJ A _all I 0- s .... OH HS-C-

B + I

HS-C-HN I

~

)OJ A -DH C 0- s"'OH + 2 R-SH

HN

~

~ ~

rAtA ,..OH N s,{)11

HN

~

I )OJ fl-C-o -AS,S_~_ HN

~

,0- ,5-" o AS'S_R f--~ HN

~

122

+2 li20

+2 H 0 2

Figure 5.4 A. Hydration of arsine oxide to arsonous acid. B. Interaction of immobilized arsonous acid with dithiol compounds to form a ring structure. C. Interaction with monothiols.

123

A study of arsonous acid immobilized on an agarose support (21)

demonstrated the selectivity of immobilized arsine oxide for dithiols

over monothiols. In this study, the monothiols, cysteine (cys) and

2-mercaptoethanol were eluted at neutral or acid pH but were retained

with O.OIM NaOH. Dihydrolipoic acid (DHLA) , a dithiol, was retained

using 0.01 M NaOH but eluted at 0.2 M NaOH. This study showed weaker

interactions of DTT and DTE than DHLA, because the DHLA forms a 5

membered ring while DTT and DTE form 7 membered rings. In order to

use this phase for the extraction of DTT and DTE the optimization of

conditions for retention will be important.

This phase will allow for the extraction of a dithiol from a

solution. With the appropriate conditions, this phase will extract

dithiol reducing reagents from a solution containing monothiols with

minimal retention of monothiols. When the reducing agent is extract­

ed, either by solid phase chemistry or solution chemistry, care must

be taken to assure that the free thiol of the compound of interest is

not oxidized to the disulfide bond prior to the assay. Disulfides

form readily, and the free thiols must be kept in a non-oxidizing

atmosphere. The environment of the sample is easily controlled when

using solid phase adsorbents in a column that is directly coupled to a

second column which may contain a sorbent for the immobilization of

the compound through the free thiol. Coupling the immobilized

arsonous acid column to a mixed disulfide column allows for the

extraction of the dithiol reducing agent followed by the

124

quantification of free thiols in the sample through the detection of

the displaced thiopyridone. In addition, the thiol containing com­

pounds may be isolated on this phase and eluted with a competing

thiol. This reduction in sample handling can increase the ease and

efficiency of the quantification of thiols.

The immobilization of thiol reagents will allow for the

isolation, detection or storage of thiol containing compounds with an

ease of sample handling, small sample volumes and speed. In order to

fully exploit this chemistry, the effects of immobilization on selec­

tivity and reactivity must be understood. In addition, the efficiency

and selectivity of the immobilized reagents may be controlled through

the mobile phase. Diagnostic chromatography was employed to determine

the selectivity and the ability to control the selectivity of the

immobilized thiol reagents discussed in this chapter.

Experimental

The capacity of the phases as measured by elemental analysis

and described in chapter 2 is not necessarily the capacity available

to the sample, because it may be sterically restricted from some or

even all of the active sites. However, it is useful to compare the

number of sites bound to the surface and the number available to the

sample. The capacity of the mixed disulfide phase available to

cysteine was measured using frontal elution chromatography. A mobile

phase containing cysteine was used to completely wash the thiopyridone

from the phase. The effluent was captured in a volumetric flask and

the concentration of thiopyridone was measured using UV spectroscopy.

125

The capacity of the phenyl mercurial phase was measured using repeated

injections of cysteine. When the cysteine is eluted, the column was

considered overloaded and the capacity of the phase was determined

from the number of injections.

Diagnostic chromatography as described in chapter two was

performed to measure the retention of several probe molecules on the

column to determine the extent of non-specific interactions and their

contributions to the retention of thiol compounds.

The detection of the thiol compounds was either through direct

measurement of absorption at 254 nm or through the use of the mixed

disulfide phase as a reactor bed. The use of this phase allows for an

increase in sensitivity and therefore the use of lower concentrations.

Results & Discussion

Disulfide Phase

Studies of the mixed disulfide phase (figure 5.2) involving

the determination of peak height vs concentration show that the

amount of thiopyridone eluted from the surface is proportional to the

amount of cysteine or DTT immobilized over a small concentration range

(figure 5.5). This phase may be used to quantify the amount of thiols

injected into the column provided it has been carefully standardized.

The capacity of this phase for immobilizing cysteine was

measured by frontal elution chromatography. A solution of cysteine in

50% methanol/ 50% water was washed through a column which contained 71

milligrams of the mixed disulfide phase. UV spectroscopy was used to

126

CONCENTRATION OF CYSTIENE VERSUS PEAK AREA ON THE YXED 0ISUI.I'1De: PtfASE

28

ae a 24

22

20

• I •

M ~ 12 ~

10

8

a .. 2

J 0 0 2 3 ..

CONCENTRA 110H ImMI

CONCENTRA TION VERSUS PEAK HEIGHT AND AREA FOR on ON IIIXED DISlAJ'lOE: PtfASE

24

22 b

20

IG

IG 1/1 .... ; M ... ~ 12 e: I 10

/ /

: ~ / /

/ / .. /

/ 2

0 0 ..

Figure 5.5 Use of the mixed disulfide phase for the quantification of cysteine (a) and dithiothreitol (b).

127

determine the concentration of thiopyridone in the effluent. It was

determined that 38 micromoles were eluted from the system, which

corresponds to 0.5 milliequivalents per gram. The initial concentra­

tion of propylmercaptan was 0.9 milliequivalents/gram. About

60% of the original mercaptopropyl groups were accessible to the

cysteine. This is consistent with the elemental analysis. Cysteine

is not expected to be excluded from the active sites, because it is

smaller than the activation molecule. The available capacity will be

more important as larger molecules are used. The larger molecules may

be sterically or otherwise restricted from the active site and there­

fore unable to react. In those cases, the modification of the silica

would need to include a longer spacer arm that could reach to the free

thiol on the compound of interest.

Determination of the non-specific interactions on this phase

are needed in order to develop effective isolation and detection

procedures. The retention of benzene, phenol and benzoic acid at a pH

of 4 show typical reverse phase retention (figure 5.6). As the

methanol concentration is increased, the retention of these probes

decreases. These molecules are most likely interacting with the

propyl linkages and the mobile phase associated with these chains (9).

As the ionic strength changes (figure 5.6), retention for these

molecules is unaffected, indicating that these molecules are not

retained through electrostatic interactions. The retention of aniline

at a pH of 4 demonstrates electrostatic interactions. The retention

is unaffected by a change in methanol (figure 5.6) but decreases as

FORI·1A TE BUFF ER pH 4

A

2.5

1 • 5

. 5

5 25 4S

B

~ ANILINE

2.5 1!l8ENZENE

~PHENOL

2.0

..:.: 1.5

1.0

:z!s:

2." 3." Figu~e 5.6 Secondary interactions on the mixed disulfide phase

as a functio~ of % methanol (A) and -log ionic strength (B).

128

129

the ionic strength is increased (figure 5.6). At pH 6, aniline

retention decreases as the methanol concentration is increased (figure

5.7) which is consistent, because at pH 6, the aniline is not charged

and electrostatic interactions are not responsible for its retention.

At pH 4, benzoic acid is not charged and demonstrates typical reverse

phase interactions. When the pH is increased to pH 6, the benzoic

acid is charged and elutes near the void volume and the retention is

unaffected by a change in methanol. These data indicate a strong

cation exchange site on the surface. Since the charge exists at a pH

of 4, it is not likely residual silanols, since these ionize at pH 6

and above (47). The negative charge on this surface is attributed to

the oxidation of the free thiols to a sulfonic acid.

In order to demonstrate how the secondary interactions might

change as the surface is exhausted, the thiopyridone was washed from

the surface using 2-mercaptoethanol (figure 5.8). The lipophilic

interactions are not significantly affected as demonstrated by the

retention of benzene and phenol, but the retention of aniline increas­

es (figure 5.9). The retention of aniline is dependent on ionic

strength but not on methanol concentration, demonstrating electrosta­

tic interactions between aniline and negative surface charge. The

increase in retention is due to an increase in cation exchange capac­

ity. 2-mercaptoethanol reduces the disulfide on the surface to leave

a free thiol as shown in figure 5.10. When the disulfide bond is

cleaved, the free thiols then oxidized to sulfonate groups, increasing

the cation excha.nge capability of the surface.

Figure 5.7

a

4.9

3.5

3.0 .:.::

2.5

2.0

5

3." b

2.5

2."

1.5 .:.::

1.9

.5

5

Retention of aniline

FORI·1A TE BUFF ER

ANILINE

<!> pH = 4

I!l = 6

15 25 35 45

BENZOIC ACID

<!> pH = 4

[!] p II 6

[!I

15 25 35 45

(a) and benzoic acid (b) as a function of % methanol on the mixed disulfide phase at pH 4 and 6.

130

-s-@ + 2 HOCH 2CH 2 SH

+

+

Figure 5.8 Reaction of 2-mercaptoethanol with the mixed disulfide phase.

131

FORMATE BUFFER pH = 4 MERCAPTOPYRIDIHE REMOVED FROM SURFACE

a

7.0

4.111

1.111

b

8.

5.5

3.0

.s

5 15

~ ANILINE

(!l8ENZENE

~PHENOL

25 35 45.13

,.W 2.5 3.0 3.5 4.111 Figure 5.9 Retention as a function of % methanol (a) and - log

ionic strength (b) on an exhausted mixed disulfide phase.

132

133

When using the mixed disulfide phase, care must be taken to

assure that all mobile phases are carefully deaerated to assure that

all oxygen is removed. Secondly, it is important to keep this phase

in a non oxidizing atmosphere during storage and preparation. The

oxidation of a thiol to a disulfide is reversible but the oxidation a

sulfonate is not. Once the thiols have oxidized to sulfonates,

regeneration of the phase is impossible.

Organomercurial

The secondary interactions of the organomercurial phase were

studied at pH 5 and 7 with a 4lmM ammonium acetate buffer and 5%

acetonitrile. The capacity factors of several probe molecules are

shown in figure 5.10. The ligand strongly attached to the mercury may

be the acetate or the ammonia from the buffer or hydroxide ions.

Assuming that the dissociation constants for methylmercury (table 5.1)

are similar to those for phenylmercury (20), the extent of complex­

ation can be determined. Only the deprotonated form of the ammonia is

capable of complexation with the mercury, but the complexation is

strong, and even at pH 5 when only a negligible amount of ammonia is

deprotonated, the ammonia complex exists in the same concentration

range as the acetate complex. The concentration of unprotonated

ammonium is about 4 orders of magnitude lower than the acetate but the

dissociation constant is four orders of magnitude lower. When the pH

is increased to 7, the ammonia concentration increases by a factor of

100 and this is the dominant ligand complexed to mercury.

134

Mercurial

~ ~------------------------------------------------------~

9

o

7

:.. S

4

3

2

PAA BSA

IZZI pHS

Figure 5.10 Non-specific interactions of organomercurial bonded silica using 5% acntonitrile and 95 % ammonium acetate buffer with c:'.n ionic strength of 0.1 M.

135

The acetate complex with mercury is neutral. The negative

acetate neutralizes the positive mercury. The ammonia is a neutral

compound and unable to neutralize the cation. At pH 5, some of the

mercury sites are complexed with acetate, but as the pH is increased

to 7, the mercury complexes with ammonia and the acetate sites are

negligible. Therefore, the anion exchange capability of the surface

increases.

The increase in anion exchange capacity of the phase with an

increase in pH is demonstrated by the increase in retention of

benzene sulfonic acid at pH 7 (figure 5.10). One expects the anion

exchange capability of the surface to decrease as the pH increases,

and this is the case for the phases discussed in chapter 3. The

retention of the rest of the probes does not change as drastically;

however, the decrease in retention of benzene and toluene at pH 7 is

consistent with an increase in positive surface charges.

The retention of these probe molecules under these conditions

are, in general, within the optimum range for covalent chromatography.

The molecules are not excluded from the phase and they are not strong­

ly retained. The exceptions are the ionic species. The cation,

phenyltrimethylammonium, is excluded from the column and the anion,

benzenesulfonic acid, at pH 7 is retained for almost 10 column vol­

umes. An increase in the ionic strength will reduce the electrostatic

interactions and eliminate the excessive retention of the anion and

the exclusion of the cation.

136

The positive nature of the surface is an advantage in retain­

ing thiol containing compounds. The positive charge on the surface

will stabilize the retention of ionized thiols. Cysteine was well

retained on the mercury phase in a wide variety of conditions.

Cysteine was retained at a k' of over 100 on this phase with aceto­

nitrile concentrations up to 100% and with ionic strengths from O.lM

to lmM. A great advantage of this phase is the versatility of the

mobile phase, allowing for a wide range of sample matrixes without a

loss of efficiency.

The use of a phenylmercuria1 column for the retention, isola­

tion and subsequent elution of several thiol containing compounds has

been demonstrated. The strong retention of cysteine on this column

demonstrates its possibility for this use. Oxytocin (figure 5.11) was

dissolved in acetonitrile with a ten fold excess of DTT (figure 5.1).

Using a octadecy1si1ane 10 centimeter 2 millimeter inner diameter

column in a typical chromatographic experiment the oxytocin was eluted

in 1 minute (figure 5.11, Trace A). When an organomercurial column is

placed in front of the column the octadecylsilane column (figure 5.11

Trace B), the oxytocin is not eluted after 5 minutes. At 5 minutes,

the mercurial column is removed and the system is equilibrated with a

mobile phase of 100% acetonitrile and 0.02 M cysteine. The mercurial

column was replaced in the system, and a peak is eluted 1.2 minutes.

When using a mobile phase with no cysteine, the oxytocin interacts

with the mercury forming a covalent bond with at least one of the free

thiols from the cysteine residues. When a mobile phase with cysteine

137

OXYTOCIN

H_Cys_Tyr_Ile_Gln_ASn_Cys-Pro-Leu-Gly-NH2

B

A ---

t t t 1 2 3

Figure 5.11 Oxytocin retained and eluted on an immobilized organomercurial.

A. Chromatogram of oxytocin on CIS with an acetonitrile mobile phase and a flow rate of 1 ml/min. B. Chromatogram with the phenylmercurial column and the same conditions. At 2 the mercurial column was removed and the system was equilibrated with .02 M cysteine in acetonitrile. At 3 the mercurial column is replaced in the system.

is used the cysteine displaces the oxytocin through mass action,

eluting it from the column.

138

The DTT used to reduce the disulfide bridges does not appear

in the chromatogram because the absorption of DTT at 254 nm is too low

to detect it. In most cases, it is necessary to remove the DTT prior

to isolation on the organomercurial column. Removal of the DTT would

assure that the reducing agent does not overload the mercurial column

A second demonstration of the usefulness of an immobilized

organomercurial column for the isolation of thio1 containing compounds

is shown in figure 5.12. The same experimental procedure was used

except the sample was 2,2-dithiodipyridine reduced with DTT and the

mobile phase was 5% acetonitrile and 95% 20 mM ammonium acetate

buffer. The elution mobile phase contains 0.02 M cysteine. Trace A

is an unreduced dithiodipyridine which is not retained on the mercury

column. This compound elutes in the void volume. Trace B is the

reduced sample which is retained on the column and then eluted by the

cysteine mobile phase.

Pheny1arsonous Acid

The non-specific interactions of the pheny1arsonous acid phase

were probed at pH 5 and 7 (figure 5.13) with a mobile phase of 41mM

ammonium acetate buffer and 5% acetonitrile. The effects of pH on

retention of the probe molecules is consistent with the PBA phases

(chapter 3). As the pH increases, the retention of the cations

increase and the retention of anions decrease, demonstrating fewer

A

1

B

10

2 3

20 Time (min)

Figure 5.12 Automated system to using organomercurial column coupled to a CIS column with a mobile phase of 5% acetonitrile 95% 20 roM ammonium acetate and a flow rate of lml/min.

139

At 1 the sample is injected. At 2 the mercurial column is removed and the system is equilibrated with the a .02 M solution of cysteine in the original mobile phase. At 3 the mercurial column is replaced in the system. Trace A is unreduced dithiodipyridine and trace B is dithiodipyridine in an excess of OTT.

3.2

3

2.8

2.8

2..4

2.2

2

La

1.6

L4

L2

o.a 0.8

0.4

a.z 0

11;(11

Figure 5.13 Nonspecific interactions on phenylarsono\ls acid bonded silica.

140

141

positive sites or an increase in negative sites on the surface as the

pH increases.

The non-specific interactions are within the optimum range for

covalent chromatography. However, an increase in ionic strength to

reduce the exclusion of the cation at pH 5 may be necessary to retain

cations.

The retention of cysteine as a function of acetonitrile

concentration was measured (figure 5.14). As the acetonitrile concen­

tration increases, the zwitterionic cysteine is better solvated

by the hydrophilic urea phase and the retention increases. To mini­

mize the interaction of cysteine with this phase, 5% acetonitrile was

chosen as the concentration of organic modifier.

In order to demonstrate the ability for this phase to separate

mono and dithiols, a sample of DTT and cysteine was used. The concen­

tration of DTT was twice the concentration of cysteine. The results

of chromatographic separations using an arsonous acid column coupled

to a C1S column is shown in figure 5.15. Trace A is the sample using

a propylurea column coupled to an octadecylsilane column. The first

peak at 2.1 minutes is the cysteine the second peak at 4.5 minutes is

the DTT. In trace B, the urea column is replaced with a phenylar­

sonous acid column. The cysteine peak is eluted at 3.5 minutes and

the DTT peak is not eluted after 20 minutes. The arsonous acid column

is effective at separating mono and dithiols.

The poor peak shape is partly due to the use of a mixed

disulfide column as a means of detection. A small packed bed was

j j c . -:

142

Retention versus Organic Modifier

4.8

4.8

4.4

4.2

4

3.8

3.8

3.4

3.2

3

2.8

2.8

2.4

2.2 5 10 15

" Acstonitil

Figure 5.14 Retention of cysteine on immobilized arsonous acid as a function of solvent strength.

OTT

ffi COLUMN I Ii CIS

A COLUMN I BLANK B COLUMN I PHENYLARSENEOX IDE

TimeCmin)

143

~ DETECTION I

Figure 5.15 Use of arsonous acid bonded silica for the separation of dithiols (OTT) and monothiols (cysteine) detected by the mixed disulfide phase. A. Without arsonous acid column. B. With arsonous acid column.

144

placed in front of the 254 fixed wavelength detector to allow for

greater sensitivity in detecting these molecules. This detection does

cause some band broadening. This column, the urea column and the

arsonous acid column are all dry packed beds of 40 micron particles

that also add to band broadening.

Extraction of the Reducing Agent

Phenylarsonous acid has demonstrated its ability to separate

dithiol containing compounds from monothiol containing compounds.

This is an effective way to isolate the compound of interest from the

reducing agent provided that one is a dithiol and the other is a

monothiol. In the case of DTT, the arsonous acid only interacts with

the reduced form. The cyclized DTT will not be retained. The cycliz­

ed form will not interact with the mercuric or the mixed disulfide

phases, so in most cases the presence of this form of DTT will not

matter.

If the reducing agent has another functionality that is

capable of forming a reversible covalent bond, then the extraction of

the agent may be accomplished through this this functionality. DTT

contains a diol which is capable of reaction with PBA in either its

oxidized or reduced form. SPE studies to immobilize DTT on PBA

however proved difficult. A 100 milligram Bond Elut (Analytichem Int)

column with a capacity of 0.49 millieqivalents/gram was conditioned 2

milliliters of acetonitrile, 2 milliliters of water and 2 milliliters

of pH 12 buffer, followed by 2 milliliters of 20% acetonitrile 80% pH

145

9 ammonium acetate buffer (84mM). Two micromoles of DTT were washed

through the column with two milliliters of the final solution. The

effluent was saved and measured by liquid chromatography with a solid

phase detection using a mixed disulfide bed coupled to a 254 fixed

wavelength detector. Only 50% of the sample or less was retained on

the column. Only when the ionic strength of the wash buffer is raised

to 2 M NaCl is 97% of the sample retained.

The difficulty in retaining DTT on the PBA phase is due to the

exclusion of DTT from the surface. PBA only interacts with diols when

it is in a tetrahedral ionized form. DTT with pKa values of 7.8 and

8.4 for the two thiol groups, is ionized at the pH where it would form

a covalent complex with PBA. The electrostatic repulsion decreases

the ability of these molecules to react. At high ionic strength, the

DTT is no longer excluded from the column and the retention increases.

Unfortunately, 2 M NaCl is not an acceptable mobile phase for a

liquid chromatograph due to experimental limitations. In order to

automate a system for this extraction, it is important that the mobile

phase be compatible with pumping systems.

In comparison, the use of immobilized PBA to extract DTT will

not be as rugged nor as versatile as the use of immobilized arsonous

acid. The difficulty in using PBA is attributed to electrostatic

repulsion between the PBA and the DTT. The non-specific interactions

occurring on a phase are very important in the development of solid

phase extractions as well as any solid phase chemistry.

146 Conclusions

There thiol reagents were immobilized on a solid support with

no apparent change in the selectivity or reactivity from solution

chemistry. The advantage of solid phase chemistry is the ability to

immobilize these molecules on a solid surface for ease in separation

and sample handling and storage. The effects of the mobile phase

composition on the selectivity of the immobilized reagents were

determined for more efficient use of these phases.

The mixed disulfide phase may be used for the detection of

thiols as in solution chemistry (15) or for isolation or storage of

thiol containing compounds. One limitation of this phase is the care

needed to assure that the thiols on the surface do not oxidize to

sulfonate groups. Unlike the oxidation to disulfide bonds, the

oxidation to sulfonates is not reversible and the phase cannot be

regenerated. The non-specific retention of cations at the sulfonate

sites can be minimized through the use of high ionic strength mobile

phases. While the mobile phase controls the non-specific interaction,

the covalent interaction forms under a wide variety of conditions

from 50% methanol to 5% acetonitrile, and within the pH range of 4 to

9. This phase is compatible with a wide variety of sample matrixes

and will be useful for detection in a variety of applications.

The organomercurial phase is useful for the isolation and

storage of thiol containing compounds. This phase demonstrates

compatibility with a wide variety of mobile phases and limited

secondary interactions, making this a very versatile and easy to use

147

phase. The ability for mercury to complex with so many different

ligands will not affect is reaction with thio1s since these are

strongly retained but will affect the secondary interactions. Changes

in the mobile phase may cause unexpected changes in the non-specific

interactions.

The immobilized arsine oxide is useful for the extraction of

dithiol compounds from a solution of monothiol compounds. Tnis

extraction will simplify much of thiol chemistry since the extraction

may be performed just prior to the assay, immobilization, or storage

of the monothiol in controlled conditions. This simplifies sample

handling and eliminates the reoxidation of the thiols to a disulfide.

CHAPTER SIX

CONCLUSIONS AND FUTURE WORKS

Reversible covalent bonds between an immobilized reagent and

a compound of interest may be exploited for a wide variety of uses,

including the isolation of compounds through covalent or affinity

chromatography, solid phase synthesis or solid phase reactor beds. In

all areas of solid phase chemistry, the efficiency may be improved

through the use of rigid inert supports and an understanding of the

factors which affect the selectivity and reactivity of the immobilized

reagent.

The physical stability and the efficiency of chemically

modified silica adsorbents is not matched by any of the soft gels

commonly used as supports in covalent chromatography. One limitation

of silica has been the use of lipophilic modification methods which

are not compatible with the aqueous samples of biochemical applica­

tions. In order to assure that the lipophilic phase is solvated, an

organic solvent must be added to the mobile phases. The amount of

solvent needed to solvate the traditionally bonded silicas was detri­

mental to the stability of some biochemical compounds. A hydrophilic

phase is desirable for most biochemical applications.

A second limitation to silica as a solid support is due to

the reactivity of the support. Residual si1ano1s are capable of

148

149

strongly interacting with some compounds causing irreversible adsorp­

tion. Traditionally, these sites were reacted with trimethylchloro­

silane; however, this increases the lipophilicity of the phase.

In order to retain the advantages of silica, a new modifica­

tion method that would exclude the sample from residual silanols and

provide a homoenerget~c hydrophilic surface was developed and charac­

terized. This phase is a propylurea bonded silica. The urea groups

will hydrogen bond directly to other urea groups or thorough water

molecules to form a hydrogen bonded barrier that will restrict com­

pounds from intercalating past the urea into the surface silanols.

Chemically modified silica surfaces have been shown to have 2-3 layers

of highly structured water at the surface. Between the hydrogen

bonding nature of the silica surface and the urea groups, this entire

phase is a highly structured hydrogen bonded hydrophilic surface.

Molecules do not intercalate to the surface because of the energy

required to break the hydrogen bonds and displace a water molecule.

As demonstrated by electrostatic and hydrogen bonding inter­

actions, the residual silanols that are available to the sample on the

traditionally modified silica sorbents are not available to the sample

on the urea modified phases. The elimination of the non-specific

retention at these sites will expand the use of silica in covalent

chromatography, and all areas of chemistry utilizing immobilized

reagents by allowing for compatibility with a wider range of samples

and mobile phases.

150

Comparisons of the data of immobilized functionalities to the

solution work shows no apparent change on reactivity or selectivity.

Diagnostic chromatography and solid phase extraction were used to

determine the effects of pH on the reaction of several molecules with

PBA. The effects of pH on the reaction with immobilized PBA was

consistent with solution chemistry.

Diagnostic chromatography of several immobilized thiol rea­

gents showed that immobilization did not affect the selectivity of

these phases. Thiopyridone immobilized on a support through a disul­

fide bond will be displaced by thiol containing compounds. The

thiopyridine may be detected as it elutes from the system to quantify

the compound of interest or the compound of interest may be immobi­

lized, isolated and eluted with a competing thiol. Mercurial com­

pounds immobilized on solid supports will strongly interact with thiol

containing compounds under a wide variety of conditions. Arsonous

acid immobilized on a surface will interact with thiols but the

interaction with dithiols is much stronger than the interaction with

monothiols. The interactions of immobilized organomercurial and

arsonous acid are consistent with the interactions of these molecules

in solution.

The covalent reaction between the compound of interest and the

immobilized reagent occurs in two steps. The first step is the

diffusion of the molecule to the immobilized reagent and the second is

the formation of the covalent bond. The nonspecific interactions can

improve the first step by solvating the molecule in the stationary

151

phase. For example, electrostatic forces between opposite charges

would attract the molecule to the surface improving the first step,

while the repulsion of like charges would cause exclusion of the

molecule from the surface and the covalent bond will not form. While

non-specific repulsion may be minimized by changing the mobile phase,

a more effective method is to modify the surface so the local envi­

ronment of the reagent is compatible with the sample.

The local environment may affect the formation of the covalent

bond. The pKa of PBA on the hydrophilic propylurea modified silica is

shifted to a lower value. The amine groups on this surface are

interacting directly with the PBA or causing a surface buffering

effect to decrease the pKa . The preparation of a lipophilic phase

with residual propyl amine functionalities will have the same affect

(27). The retention of molecules will be affected by the shift in

pKa . Catechol and other diols are retained at a lower pH on the

hydrophilic phase, because these molecules interact with the ionized

form of PBA. Compounds that interact with the neutral form of PBA

will require lower pH in order for quantitative retention.

Understanding the effects of the modification on the retention

of molecules is necessary in order to develop separation methods. In

addition, modification methods of silica may be altered to fit with

the requirements for a separation to improve selectivity. Once an

adsorbent has been designed, the mobile phase may be used to fine tune

the separation as long as the non-specific interactions are recognized

and their effects are understood.

152

The use of several phase coupled together would allow the

isolation of several classes of compounds from one sample or for the

removal of an interferant prior to the isolation of a compound of

interest. Successfully coupling columns together requires an under­

standing of the interactions of all the columns and the use of samples

and mobile phases that will be compatible with all the sorbents. In

figure 1.4, an immobilized arsonous acid was coupled to immobilized

mercurial column. The arsonous acid column allowed for the extraction

of the dithio1 reducing agent so this would not compete with the

thiopyridone for the active sites in the organomercurial column. The

mixed disulfide adsorbent was used with the PBA, the mercurial and the

arsonous acid column as a method to detect low concentrations of thio1

containing compounds. The ability to couple the columns together to

isolate several classes of compounds will allow for high speed

separations of a number of compounds from a small sample.

Extending the reported studies using different buffers and

organic solvents to measure the effects of these on the selectivity

could be carried out, as well as the use of other techniques to

characterize the surface. Titration of the surface would allow for

the determination of the pKa of immobilized PBA under a wide variety

of mobile phases. This would complement the retention data of mole­

cules that interact with only one form of PBA. Chromatographic

temperature studies and IR spectroscopy as described by Schunk (9)

could be used to determine the organization of the solvated stationary

phase and the retention mechanisms occurring on the phase. The

153

isolation of biologically significant molecules is simplified when the

retention of small probe molecules is understood. While the retention

of the small probes will allow for the estimation of the interactions

of larger and more complicated molecules, systematic investigations of

the complicated molecules is needed for the isolation of these com­

pounds, and will allow for further understanding of the interactions

of the surface.

One complication of larger molecules is the inability of

reagents immobilized using short spacer arms to reach into the mole­

cule to react with the active sites. The folding of proteins or

enzymes often causes active sites to exist within the folded chains.

Modification methods should be developed that would have spacer arms

long enough to reach into folded proteins, have a chemical nature

compatible with the protein and restrict interactions of the compound

of interest with residual silanols. Development of these spacer arms

requires an understanding of the effect of the spacer arm on retention

and the effects of the spacer arm on the compound of interest.

The isolation of compounds as well as the use of multidimen­

sional chromatography would be significantly enhanced through the

automation of system with column switching capability and mobile phase

control. The preparation of an automated system would require the

understanding of the sample and the matrix to assure complete and

efficient separation.

REFERENCES

1. Scouten, William H. Solid Phase Chemistry Wiley, NY (1983) 1-16.

2. Stewart, John M. "Solid Phase Pepti1e and Protein Synthesis" Solid Phase Biochemistry Wiley, NY (1983) 507-534.

3. Wallace, R. B., Keiichi Itakura "Solid Phase Synthesis and Biological Applications of Po1ydeoxyribonuc1eotides." Solid Phase Biochemistry Wiley, NY (1983) 631-664.

4. Merrifield, R. B. "Solid Phase Synthesis I. Synthesis of a Tetrapeptide" J. Amer. Chern. Soc 85 (1963) 2419-2430.

5. Tramper, J. "Organic Synthesis Using Immobilized Enzymes" Solid Phase Biochemistry Wiley, NY (1983) 393-478.

6. Bergo1d, Alan, William Scouten "Borate Chemistry" Solid Phase Biochemistry Wiley, NY (1983) 149-187.

7. Guiban1t, G. G. "Immobilized Enzyme Electrode Probes" Solid Phase Biochemistry Wiley, NY (1983) 479-503.

8. Keen, James, William Jakoby "A Convenient Assay Method for Disulfide Interchange and Other Thio1 Transfer Reactions" Biochem 90 (1978) 136-140.

9. Schunk, T. C. "Chemical Interactions at the Solid-Liquid Interface: Investigations Employing Diagnostic Separations" Ph.D. Dissertation, University of Arizona (1985).

10. Gorse, J., "Characterization of the Solid-Liquid Interface on Chemically Modified Particulate Surfaces" Ph.D. Dissertation, University of Arizona (1985).

11. Karger, Barry L., Lloyd R. Snyder, Casaba Horvath An Introduction to Separation Science, Wiley, NY (1973).

12. Snyder, L. R., J. J. Kirkland Introduction to Modern Liquid Chromatography 2nd Edition, Wiley, NY (1979).

13. Small, Hamish, Timothy S. Stevens, William C. Bauman "Novel Ion Exchange Chromatographic Method Using Condeuctimetric Detection" Anal Chern 17 (1975) 11 1801-1809.

14. Lewis, C., W. Scouten "Immobilized Protein Modification Reagents" Solid Phase Biochemistry Wiley, NY (1983) 671-675.

154

155

15. Grassetti, D. R., J. F. Murry, .Tr. "Determination of Sulfhydryl Groups with 2,2/- or 4,4/-Dithiodipyridine" Arch. Biochem. Biophys. 119 (1967) 41-49.

16. Carlsson, Jan, Rolf Axen, Keith Brocklehurst, Eric M. Crook "Immobilization of Urease by Thiol Disulfide Interchange with Concommitant Purification" Eur. J Biochem. 44 (1974) 189-194.

17. Brocklehurst, Keith, Jan Carlsson, Marek P.J. Kierstan, Eric M. Crook "Covalent Chromatography" Biochem. J. 133 (1973) 573-584.

18. Carlsson, Jan, Anders Svenson "Preparation of Bovine Mercaptoalbumin by Means of Covalent Chromatography" FEBS Letters 42 (1974) 2 183-186.

19. Yoshida, Shonen, Toshiteru Morita, Mineo Saneyoshi "Isolation of the Newly Synthesized DNA of HeLa cells Containing p-2/-Deoxy-6-Thioguanylate on an Organomercurial Agarose Column" Biochem Biophis Acta 607 (1980) 527-529.

20. Webb, J. Leyden "Mercurials" Enzvrne and Metobolic Inhibitors Vol. II Academic Press, NY (1966) 729-907.

21. Hannes tad , Ulf, Per Lunquist, Bo Sorbo "An Agarose Derivative Containing an Arsenical for Affinity Chromatography of Thiol Compounds" Anal Biochem. 126 (1982) 200-204.

22. Kennedy, J. F., J.M. Cabral "Immobilized Enzymes" Solid Phase Biochemistry Wiley, NY (1983) 253-292.

23. Unger, K. K., R. Jansen "Packings and Stationary Phases in Preparative Column Liquid Chromatography" J Chromatogr. 373 (1986) 227-264.

24. Maestas, R. R., J. R. Preito, G. D. Juehn, J. H. Hageman "Polyacrylamide-Boronate Beads Saturated with Biomolecules: A New General Support for Affinity Chromatography" J. Chromatogr. 189 (1980) 225-231.

25. Dean, John A., editor Langes Handbook of Chemistry 11th ed., McGraw Hill, NY (1973).

26. Lochmuller, C. H., Walter B. Hill, P. M. Gross "Manipulation of Stationary Phase Acid-Base Properties by a Surface Buffering Effect: Boric Acid-Saccharide Complexation" Chromatography and Separations Chemistry, Advances and Developments (1986) 210-225.

27. Yabroff, David L., G. E.K. Branch, H. J. Almquist "The Interactions of the Boron Sextet with Adjacent Groups" J. Amer. Chern. Soc. 55 (1933) 2935-2941.

156

28. Freidman, Samuel, Benjamin Pace, Richard Pizer "Complexation of Phenylboronic Acid with Lactic Acid. Stability Constant and Reaction Kinetics" J. Amer. Chern Soc. 56 (1934) 5381-5384.

29. Larsson, Ragnar, Gennaro Nunziata "An Infrared Spectroscopic Investigation of the Complexes Formed Between Boric Acid and Lactic Acid in Aqueous Solutions" Acta Chern. Scand 24 (1970) 2156-2168.

30. Torssell, Kurt, John H. McClendon, G. Fred Somers "Chemistry of Arylboric Acids VIII. The Relationship Between Physico-Chemical Properties and Activity in Plants" Acta Chern. Scand 12 (1958) 7 1373-1385.

31. Branch, G.E.K., David L. Yabroff, Bernard Bettman "The Dissociation Constants of the Chlorophenyl and Phenethyl Boric Acids" J. Amer. Chern. Soc. 56 (1934) 937-941.

32. Johnson, Barbara J. B. "Synthesis of a Nitrobenzene Acid Substituted Polyacrylamide and Its Use in Purifying Isoaccepting Transfer Ribonucleic Acids" Biochemistry 20 (1981) 21 6103-6108.

33. Lorand, John P., John O. Edwards "Polyol Complexes and Structure of the Benceneboronate Ion" J Org Chern. 24 (1959) 769.-774.

34. Oertel, Richard P. "Raman Study of Aqueous Monoborate-Polyol Complexes. Equilibria in the Monoborate-l,2 Ethanediol System" Inorg Chern. 11 (1972) 3 544-549.

35. Burnett, Thomas J. Henry C. Peebles, James M. Hageman "Synthesis of a Fluorescent Boronic Acid Which Reversibly Binds to Cell Walls and a Diboronic Acid Which Agglutinates Erythorocytes" Biochem. Biophys. Res. Comm. 96 (1980) 1 157-162.

36. Kemper, K., E. Hagemier, D. Ahrens, K-S. Boos, E. Schlimme "Group Selective Prefractionation and Analysis of Urinary Catecholamines by On-Line HPLAC-HPLC Chromatography" Chromatographica 19 (1984) 288-291.

37. Kemper, K., E. Hagemier, K-S. Boos, E. Schlimme "Group Selective Prefractionation and Analysis of Nucleosicles and Catecholamines by High Performacne Liquid Chromatographic Techniques" Fresenius Z. Anal. Chern. 317 (1984) 680-681.

38. Higa, Sadayoshi, Tomokazu Suzuki, Akira Hayashi, Isao Tsuge, Yuichi Yamamura "Isolation of Catecholamines in Biological Fluids by Boric Acid Gels" Anal Biochem. 77 (1977) 18-42.

39. Gehrke, Charles W., Kenneth C. Kuo, George E. Davis, Robert D. Suits, T. Phillip Waalkes, Ernest J. Borek" Quantitative High Performance Liquid Chromatography of Nucleosides in Biological Materials" J. Chromatogr. 150 (1978) 455-476.

157

40. Hagemier, E., K-S. Boos, E. Schlimme, K. Lechtenborger, A. Kettrup "Synthesis and Application of a Boronic Acid-Substituted Silica for High-Performance Affinity Chromatography" J. Chromatogr. 208 (1983) 291-295.

41. Glad, Mangus, Sten Olsen, Lennart Hansson, Mats-olle Mansson, Klaus Mosbach "High Performance Liquid Affinity Chromatography of Nucleosides, Nucleotides and Carbohydrates with Boronic Acid­Substituted Microparticulate Silica" J Chromatogr. 200 (1980) 254-260.

42. Singhal, Ram P., Raui K. Baja, Charles M. Buess, David B. Smoll, Vikram N. Vakharia "Reversed Phase Boronate Chromatography for the Separation of O-Methylribose Nucleosides and Aminoacyl­tRNA's" Anal. Biochem. 109 (1980) 1-11.

43. Hagemier, E., K. Kemper, K-S. Boos, E. Schlimme "Development of a Chromatographic Method for the Quantitative Determination of Minor Ribonucleosides in Physical Fluids" J. CUn. Chern. CUn. Biochem. 22 (1984) 175-184.

44. Bouriotis, V., I. J. Galpin, P.D.G. Dean "Applications of Immobilized Phenylboronic Acids as Supports for Group Selective Ligands in the Affinity Cl'\romatography of Enzymes" J. Chromatogr. 210 (1981) 267-278.

45. Hagemier, E., K. Kemper, K-S. Boos, E. Schlimme "On-Line High Performance Liquid Affinity Chromatography-High Performance Liquid Chromatography Analysis of Monomeric Ribonucleoside Compounds in Biological Fluid" J. Chromatogr. 208 (1983) 663-669.

46. Seinkiewicz, Paul A., David C. Roberts "Chemical Affinity Systems I. pH Dependence of Boronic Acid-Diol Affinity in Aqueous Solution" Inorg. Nucl Chern. 42 (1980) 1559-1575.

47. Iler, R. K. The Chemistry of Silica Wiley, NY (1979).

48. Homer, R. B., C. D. Johnson "Acid Base and Complexing Properties of Amides" The Chemistry of Amides Wiley, NY (1970).

49. Crampton, M. R. "Acidity and Hydrogen Bonding" The Chemistry of the Thiol Group, Wiley, NY (1974) 379-415.

50. Sotowitz, Mark L. "Covalent Chromatography: Immobilized Phenyl boronic Acid for Sample Preparation" Sample Preparation and Isolation Using Bonded Silicas Analytichem Int. Harbon City, CA (1985) 41-44.

51. Anfinsen, Christian B. "Principles That Govern the Folding of Protein Chains" Science 181 (1973) 4096 223-230.

52. Ellman, George L. "A Colorimetric Method for Determining Low Concentrations of Mercaptans" Arch Biochem. Biphys. 74 (1958) 443-450.

53. Ellman, George L. "Tissue Sulfhydryl Groups" Arch Biochem. Biophys. 82 (1959) 70-77.

158

54. Zahler, W. L., Cleland, W. W. "A Specific and Sensitive Assay for Disulfides" J. BioI. Chern. 243 (1968) 4 716-719.

55. Cleland, W.W. "Dithiothrieto1, A New Protective Reagent for SH Groups" Biochem 3 (1964) 480-482.

56. Anderson, W. L., D. B. Wet1aufer "A New Method for Disulfide Analysis of Peptides" Anal. Biochem. 67 (1975) 493-502.

57. Ingebretsen, Ole Christian, Mikae1 Farstad "Separation of the Oxidized and Reduced Forms of Dithiothrieto1 and 2-Mercapto­ethanol by reversed-phase high-performance liquid chromatography. Application of the Method to Biological Extracts and the Determination of Disu1fides" J. Chromatogr. 210 (1981) 522-526.

58. Krieger, David E., Bruce W. Ericson, R. B. 11errifie1d "Affinity Purification of Synthetic Pepddes" Proc Nat1. Acad. Sci. 73 (1976) 9 3160-3164.

59. Yu, Mu-quing, Gui-qui Liu, Qinhan Jin "Determination of Trace Arsenic, Antimony, Selinium and Tellurium in Various Oxidation States in Water by Hydride Generation and Atomic Absorption Spectrophotometry After Enrichment and Separation with Thio1 Cotton" Ta1anta 30 (1983) 4 265-270.

60. Bishop, C.A., T. M. Kittson, D. R. K. Harding, W. S. Hancock "Investigation of the isolation of a mixed disulfide, diethy1thiocarbamoy1 2-pyridy1 disulfide by analytical and preparative high-performance liquid chromatography" :L. Chromatogr. 208 (1981) 141-147.

61. Liener, Irvin E. "Binding of Thio1s by a Water Insoluble Organo Mercurial Copolymer of Ethylene and Maleic Acid" Arch. Biochem. Biophys. 121 (1967) 67-72.

62. Ono, Masao, Masya Kawakami "Separation of Newly Synthesized RNA by Organomercurial Agarsoe Affinity Chromatography" J. Biochem. 81 (1977) 1247-1252.

63. Smith, Michael M. Anthony Reeve, Ru Chih C. Huang "Transcription of Bateriophase ADNA in Vitro Using Purine Nucleoside 5'-(~­S]Triphosphates as Affinity Probes for RNA Chain Initiation" Biochem 17 (1978) 3 493-500.

159

64. Mar, Eng-Chun, Eng Shang Huang "Comparative Studies of Herpes Group Virus Induced DNA Polymeraces" Intervirology 12 (1979) 73-83.

65. Melvin, William T., Hamish M. Keir "Affinity Chromatography of Thiol Containing Purines and Ribo Nucleic Acid" Biochem J. 168 (1977) 595-597.

66. Sun, Irene Yi Chi, Edward M. Johnson, Vincent G. Allfrey "Affinity Purification of Newly Phosphorylated Protein Molecules" J Biochem. 255 (1980) 2 742-747.

67. Shami, Y. S. Ship, A. Rothstein "Rapid Quantitative Separation of the Major Glycoproteins (PAS 1, 2 and 3) From Other Red Cell Membrane Proteins in a Nondenaturing Medium by Affinity Chromatography" Anal. Biochem. 80 (1977) 438-445.

68. Carroll, Anthony R., Robert R. Wagner "Adenosin-5'-O- (3-thiotriphosphate) as an Affinity Probe for Studying Leader RNA's Transcribed by Vesicular Stomatitis Viris" J. Biol. Chern. 254 (1979) 19 9339-9341.

69. Webb, J. Leyden "Arsenicals" Enzvrne and Metobolic Inhibitors Vol. III Academic Press, NY (1966) 595-793.

70. Whittaker, V. P. "An Experimental Investigation of the Ring Hypothesis of Arsenical Toxicity" Biochem. J. 41 (1947) 56-61.