Reverse micellar mobile phases for normal phase chromatography

REVERSE MICELLAR MOBILE PHASES FOR NORMAL PHASE CHROMATOGRAPHY


By

MARIA A. HERNANDEZ TORRES























A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


1983














ACKNOWLEDGEMENTS

I would like to express my gratitude and appreciation to my

research director, Dr. John G. Dorsey, for his invaluable assis-

tance, guidance and encouragement throughout the course of this

project.

Also, my special thanks are extended to Mr. John S. Landy, for

his help in the initial setting of the instrument and his continuous

assistance.

I would like to thank my colleagues in Dr. Dorsey's research

group for their friendship and daily encouragement.

I would also like to thank my family and friends for their pa-

tience and continuous support.

Finally, I would like to mention Patty Hickerson for her prompt-

ness and accuracy in typing the final draft of this thesis.














TABLE OF CONTENTS

Page
ACKNOWLEDGEMENTS. . . .. ii

ABSTRACT . . . iv

CHAPTER

I INTRODUCTION . . 1

II THEORY . . . 9

III EXPERIMENTAL . . 12

Instrumentation. . . 12

Reagents . . . 12

Procedure and Calculations . 13

IV RESULTS AND DISCUSSION . . 15

Retention Changes with Surfactant Concentration. .. 15

Efficiency . . 23

Influence of Water Content on Retention. .. 26

V CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK. 31

REFERENCES . . . 33

BIOGRAPHICAL SKETCH . . 35














Abstract of Thesis Presented to the
Graduate School of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Master of Science

REVERSE MICELLAR MOBILE PHASES FOR
NORMAL PHASE CHROMATOGRAPHY

By

Marfa A. Hernandez Torres

December 1983
Chairman: John G. Dorsey
Department: Chemistry

Reverse micelles, formed in hexane, can be used in place of polar

modifier to control the strength and selectivity of normal phase mo-

bile phases for liquid chromatography. The effect of surfactant con-

centration (Aerosol OT) on retention of various solutes in a normal

phase liquid chromatographic system was investigated. Surfactant con-

centration above the critical micelle concentration range was used to

observe the change in capacity factor, k', as the water content in the

mobile phase was varied. Small changes in k' values were obtained at

low water content in the micellar mobile phase. Finally, the column

efficiencies for dry hexane, 5/95 isopropanol/hexane and 5 x 10-2 M

AOT in hexane mobile phases were compared for Ultrasil-NH2 and Ultra-

sphere-Si columns.














CHAPTER I

INTRODUCTION

A great deal of research has been put into the improvement of

both the stationary and mobile phases of high performance liquid

chromatography (HPLC) to achieve better separations.1-3

This study is limited to normal phase HPLC, for which the reten-

tion of the normal phase separation increases with increasing solute

polarity and decreases with increasing mobile phase polarity. The

solute retention mechanism is dominated by interaction of polar sta-

tionary phase sites with polar solutes.

The advantages of normal phase chromatography are summarized by

Abbot4 as follows:

1) ability to separate highly hydrophilic species which cannot

be retained on reverse phase

2) use of predominantly organic solvent-based mobile phases,

avoiding silica dissolution problems experienced in reversed

phase

3) ability to differentiate solutes based on differences in hy-

drophilic structure rather than hydrophobic structure

4) compatibility of organic phase with molecules having either

low stability or aggregation problems in aqueous phases

5) lower viscosity of organic mobile phases, resulting in lower

operating pressures








6) mobile phase volatility, allowing simpler, more efficient

concentration and transfer steps in off-line fraction collection

and structural characterization

7) ability to separate isomers which is of great significance

in natural product and pharmaceutical chemistry

8) ability to obtain class separations which cannot be obtained

in reversed phase

9) greater compatibility of organic mobile phase with on-line

coupling

10) availability of wide range of organic solvents to capital-

ize on special solvent effects

11) availability of a wide range of stationary phase selectivi-

ties, which can be utilized to obtain high separation performance

in multi-stage chromatography.

Normal phase chromatography typically encompasses adsorption

chromatography on silica and partition chromatography on cyano and

amino bonded phases.

Adsorption may be defined as the concentration of solute mole-

cules at the interface of two immiscible phases. In liquid-solid ad-

sorption chromatography, the mobile phase is a liquid, while the sta-

tionary phase is a finely divided, usually porous, solid. The atoms

in the bulk of the solid are subjected to equal forces in all direc-

tions, whereas the surface atoms experience unbalanced forces which

can attract molecules from the surrounding solution to restore the

balance.








In a multicomponent system, selective adsorption occurs due to

competition between the solutes and the mobile phase for the surface.

It is governed by the differences in the strengths of the adsorption

forces between the adsorbent and the adsorbates. In general, polar

compounds are more strongly adsorbed by polar solids than are non-

polar compounds. Also, adsorption of a polar compound is enhanced in

a nonpolar medium, but reduced in a polar medium due to increased com-

petition of the mobile phase for the surface.

The partition process of a solute occurs between a liquid mobile

phase and a liquid absorbed on, or chemically bonded onto, a porous

support. Classical liquid-liquid chromatography involves the parti-

tion of the sample components between a liquid stationary phase and

the liquid mobile phase. The separation occurs due to differences in

the solubilities of the sample components in the two liquid phases.

Since liquids which are practically immiscible will usually still have

at least ppm solubility in each other, the mobile phase must be pre-

saturated with the polar phase to prevent gradual column stripping and

loss of separating power. As a result, column packing have been

developed in which the stationary phase is permanently bonded to the

support by chemical bonding. Bonded phase packing are prepared by

the reaction of the surface Si-OH groups of the support particles

with various reagents5 (Figure 1). These siliceous supports are pre-

pared with a variety of functional groups, ranging from very polar to

nonpolar, resulting in widely diverse selectivities. The amino(-NH2)

and cyano(-CN) functional groups are polar in nature, and they are

commonly used with low polarity mobile phases for normal phase sepa-

rations.










































crw
Z


z



=!



AJ
50

f




0 w b

0 -i-




o: o
-/- 4--








z<-


41I
nI o
--T 3
*n In --v ---

t 0 I.t- I








It has been recognized that retention on active stationary phases

such as silica and alumina with nonpolar and moderately polar eluents

is strongly influenced by the water content of the eluent.6-9 The

surface of the silica gel is covered with Si-OH and Si-O-Si groups

which can interact with the sample molecules. There are two kinds of

hydroxyl groups covering the silica gel surface: free (OH) and reac-

tive (HOH) hydroxyl groups. The reactive hydroxyl groups are very

strong bonding agents which may permanently adsorb polar sample and

water molecules onto the gel, giving nonreproducible results. Since

the water content of the absorbent markedly affects relative capacity

ratio (k') values and band migration rates, water content must be held

constant for repeatable separation.

Surfactants are amphiphatic molecules which have distinct hydro-

phobic and hydrophilic regions. Over a narrow concentration range,

defined as the critical micelle concentration (CMC), surfactants dy-

namically associate to form large molecular aggregates. In nonpolar

solvents, these have been termed reversed or inverted micelles, since

their polar groups are concentrated in the interior of the aggregate

while their hydrophobic groups extend into and are surrounded by the

bulk apolar solvent (Figure 2). In aqueous solution, the micelle is

a compact roughly spherical body with a liquid-like hydrocarbon core.

The polar head groups are at the surface (Figure 3). Micellar sys-

tems have been introduced in reverse phase HPLC, and an improvement

in the separations has been reported.2,3,10,'11 When using micellar

mobile phases, solutes do not partition to the bulk solvent but rather

to the discrete aggregates creating a unique separation mechanism:











































Reverse micelle representation. The ionic head group (0)
points toward the center of the micelle, and the hydro-
carbon chain (No/) surrounds the central ionic core
(from 12).


Figure 2.











































Dill-Flory's representation of a normal micelle. The
ionic head groups are indicated by the circles, and the
hydrocarbon chains are pointing toward the center of
the micelle (from12).


Figure 3.








pseudo phase liquid chromatography. Some of the advantages of the
13
micelle system in RPHPLC are:

1) many hydrophobic, amphiphilic, and even hydrophilic molecules

interact differently with the micelles, allowing for a wide range

of applications

2) low cost

3) nontoxic

4) non-UV absorbing

5) simultaneous separations of hydrophobic and hydrophilic

solutes are accomplished without aqueous organic gradients

6) the addition of solutes to the mobile phase to control ionic

strength, pH, and buffer capacity will not produce solubility

problems

7) a relatively small change in the amount of surfactant will

essentially give a different mobile phase.

Our studies are directed mainly to observe the behavior of sol-

utes in reverse micellar systems. Solute's retentions as a function

of surfactant concentration are examined. Also, this study attempts

to find whether the water in the chromatographic system will be en-

trapped so that the retention of the solutes in the normal phase HPLC

will not be changed. The best column efficiency is sought among all

the conditions studied.













CHAPTER II

THEORY

Due to the new findings and applications of reversed or inverted

micelles in the course of the last few years, the scientific interest

with regard to nonpolar detergent solutions has extensively in-

creased.14

The obvious difference between micelles in polar and nonpolar

surfactant solutions is their mutual structural reversion. In non-

polar solvents, the inverted micelle is visualized to be built up by

a polar core covered by hydrocarbon tails videe supra). Inverted

micelles have moderate aggregation numbers which contrast the large

micellar aggregates in aqueous surfactant solutions, e.g., for 4.9 x

10-4 M of Aerosol OT in pentane at 250C the mean aggregation number

(n) is 15, while for 8.2 x 10-3 M of sodium n-dodecylsulfate (SDS)

in aqueous solution at 250C n is 64.

In the formation of a micelle in apolar media, there seems to be

a general agreement with regard to the existence of premicellar ag-

gregates. The monomer 4 n-mer type association is unlikely to rep-

resent the behavior of surfactants.16 The most universally used

treatment, the multiple equilibrium model, assumes the stepwise forma-
kI
tion of aggregates in an indefinite association: monomer dimer
k2 k3 k n
k2 trimer 3 n n-mer, where the distribution of the

different aggregates depends on the stoichiometric surfactant








concentration. The higher the concentration of the surfactant, the

larger the aggregates. Under this treatment, the CMC can be defined

as the concentration of aggregates within a narrow distribution cen-

tering around the average micellar size. On the other hand, it can

be assumed that there is no unique concentration at which the exist-

ence of micelles becomes detectable; therefore, it is unjustified to

speak of one concentration, or even a narrow range of concentration,

describing the CMC of surfactants in hydrocarbon solutions.)7

The driving force behind the process of aggregation and micelli-

zation of surfactants in hydrocarbon media is essentially due to

dipole-dipole interactions between polar heads of the amphiphilic

molecules.

One important property of the micellar system is the detergency

or the ability of surfactant molecules to take up (solubilize) polar

material, i.e., water in the polar core of the inverted micelles.

There are a large number of parameters which influence and even

determine the extent of the solubilization of the polar material.

These parameters include the structure of the surfactant, the physico-

chemical property of the solubilizate, temperature, pressure, electric

fields and cosurfactants.

Considerable amounts of water can be solubilized by reverse mi-

celles. This surfactant solubilized water is often called waterpool.

Initial water molecules are bound to the polar head groups, and their

motion is restricted. Additional water molecules occupy the core of

the micelle, and their properties resemble bulk water. A rapid ex-

change takes place between these two kinds of water molecules.








The effective polarity, acidity and microscopic viscosity of the

water pools are expected to be substantially different from those in

bulk water.

Polar substrates are expected to be localized in water pools;

therefore, controlled amounts of surfactant-entrapped water in non-

polar solvents provide a unique medium for interactions of polar sub-

strate.













CHAPTER III

EXPERIMENTAL

Instrumentation

A Spectra-Physics 8700 (Santa Clara, California) liquid chromato-

graph equipped with a model 7125 Rheodyne sample injection valve with

a 10 microliter loop, a model 153 Beckman UV detector (254 nm) with

an 8 microliter flow cell, and a model 1210 Linear strip chart re-

corder were used. The columns used were: an Altex (Berkeley, CA)

Ultrasphere-Si (15 cm x 4.6 mm ID) 5 pm particle size and an Altex

Ultrasil-NH2 (25 cm x 4.6 mm ID) 10 pm particle size. Both columns

were thermostatted at 30C 0.2 using a water jacket and a Haake D1

water circulator.

Reagents

HPLC grade n-hexane (Fisher Scientific) was used as the solvent;

it was dried using Linde Molecular Sieves (Union Carbide) type 3A.

Reagent grade 2-propanol (Fisher Scientific) was used as organic

modifier and to wash the columns.

Reagent grade Aerosol OT (dioctyl sodium sulfosuccinate, Fisher

Scientific) was used without further purification; it was dissolved in

hexane and then filtered with a Rainin solvent filtration apparatus

through a 0.45 pm membrane filter.

Water used in the mobile phase was purified using a Barnstead

Nanopure System (Symbron Corporation).








The solutes were used without further purification, and all the

solute solutions were prepared in n-hexane. The solutes were: 2,4-

dinitrotoluene, naphthol (Matheson Coleman & Bell) and phenol

(Mallinckrodt).

Reagent grade n-pentene (J.T. Baker) was used as an unretained

compound for the column's void volume calculation.

Procedure and Calculations

The appropriate weight of surfactant (Aerosol OT) was weighed

and dissolved in hexane. The concentration range was from 1 x 10-4 M

to 1 x 10-1 M with alternate increments of 5x and 2x, respectively.

The columns were thermostated at 300C. A silica gel precolumn was

used before the injector to prevent the silica in the column from

being dissolved by the surfactant. The columns were washed with 2-

propanol, and rinsed and stored in n-hexane.

A stock surfactant solution was prepared of 0.50M,and aliquots

of 50 ml were taken and diluted with n-hexane to 5.0 x 10-2 M for the

study of the influence of mobile phase water content on the retention

time of the solutes. In each of the six 5.0 x 10-2 M Aerosol OT solu-

tions, the volume percentage of water was varied (0%, 0.01%, 0.02%,

0.05%, 0.10%, 0.25%).

The capacity factor, k', was used to compare the retention data

of the solutes and was calculated using the following equation:

V V
k' = rV 0 (1)
o

where Vr = the retention volume of a given solute and is calculated

as the product of flow rate times retention time (ml), and Vo = the








column's void volume (ml). n-Pentene, assumed to be an unretained

solute, was used to calculate the void volume of the column.

The efficiencies of the columns were compared using the reduced

plate height, h, given by:

h N.dp (2)


where L = the column length, dp = the particle diameter, and N = the

plate number. The plate number, N, was calculated as follows:18
t 2
41.7 ( )
N = 0.1 (3)
B/A + 1.25

where tR = the retention time of the compound (cm), w0.1 = the peak

width at 10% of the peak height (cm), and B/A = the asymmetry factor.














CHAPTER IV

RESULTS AND DISCUSSION

Retention Changes with Surfactant Concentration

Figures 4 and 5 show the behavior of phenol and naphthol in

hexane/Aerosol OT mobile phases. These graphs show two linear compo-

nents, one above and one below the CMC. Since the formation of mi-

celles in nonpolar solvents occurs in sequential steps, it is diffi-

cult to assign the CMC to one point. The intersection of the two

linear components was found to be 3.5 x 10-3MAOT in hexane. There-

fore, the concentration range from 1.0 x 10-3 M to 5.0 x 10-3 M AOT

in hexane was assigned as the CMC, in agreement with the formation

of reverse micelles as the multiple equilibrium model postulates. In

both columns, Ultrasphere-Si and Ultrasil-NH2, the first large change

in k' was observed in the same range of concentration, the CMC range.

In other words, the same value of CMC range for AOT in hexane was ob-

tained for the two different columns and two different solutes. Table

I shows mean aggregation numbers and critical micelle concentrations

of Aerosol OT in different hydrocarbon solvents.

At concentrations above the CMC range, the slope of the four

curves are slightly negative; with the Ultrasphere-Si column having a

greater negative slope. This can be attributed to the fact that the

silica surface has silanol groups that interact with polar compounds























150

1.40
o 1.30

1.20

o 1.10

1.00

-0.90

0.80
0.70 0
-0S0
0.60

050

-0.40

-0.30

0.20

0.10

-50 -4.0 -3.0 -2.0 0

LOG [AOT -.







Figure 4. Effect of surfactant concentration on retention
using Ultrasphere-Si column. (0) naphthol, (*)
phenol.




J I


1.80

1.70

,- 1.60

-o- 1.50

1- .40

*- 1.30

1- .20
1.10

1.00

0.90 -

0.80 O
J
0.70

0.60

050

-0.40

030

0.20

0.10
I I II I
-5.0 -4.0 -3.0 -2.0 -1.0


LOG [AOT]


Figure 5. Effect of surfactant concentration on retention using
Ultrasil-NH2 column. (1) naphthol, (0) phenol.








Table I

Mean Aggregation Numbers (n) and CMCs of 14
Aerosol OT in Different Hydrocarbon Solvents (from )

Temperature CMC
Solvent (0C) Technique (M) n

Cyclohexane 37 VPO 3.9 x 10-4 17

28 4 LS 1.3 x 10-3 56

Benzene 37 VPO 3.5 x 10-4 13

RT PA 2.2 x 10-3

28 4 LS 2.7 x 10-3 23.6

20 TCNQ 2.0 x 10-3
CCI4 25 VPO 1.6 x 10-4 17

37 VPO 4.0 x 10-4 17

20 TCNQ 6.0 x 10-4
Pentane 25 LS 4.9 x 10-4 15

a VPO: vapor pressure osmometry
LS: light scattering
PA: positron annihilation
TCNQ: solubilization of 7,7,8,8-tetracyanoquinodimethane








in a very strong way. Also, the polar heads of the surfactant can

interact with these active sites, and retention of polar compounds

will decrease as the number of active sites available is reduced by

the presence of free surfactant. In the Ultrasil-NH2 column, the

interaction of the polar surfactant head groups with the amino sta-

tionary phase groups is less strong; therefore, the concentration of

free surfactant in the mobile phase for this column affects the re-

tention of polar compounds to a lesser extent. Recently, Deming and

Tang proposed that addition of surfactant to the eluent reduces the

interfacial tension and thus decreases the retention time of the in-

jected samples.19

Looking at Table II, one can observe a big change in k' value

going from dry hexane to 5 x 10-4 M AOT in hexane mobile phases. Two

possible reasons for this change are (1) before the CMC range, the

surfactant is in a monomeric form, and the polar heads compete for

active sites in the stationary phase, and (2) the formation of small

aggregates at very low concentration of surfactant that can interact

with solutes. For both points, there is a decrease in the number of

active sites available at the stationary phase; therefore, a decrease

in the retention time of solutes is observed.

Above the CMC range, the k' of naphthol and phenol decrease very

rapidly as the formation of aggregates increases. The reason for this

behavior is that in a reverse micellar system a polar solute will par-

tition to the polar core of the micelle spending less time interacting

with the stationary phase. This separation mechanism with reverse mi-

celles is similar to the one discussed by Armstrong and Nome20 for




L9


Table II

k' Value for Phenol and Naphthol in Dry Hexane and
5 x10-4M AOT in Hexane for Ultrasphere-Si and Ultrasil-NH2 Columns


Column Mobile Phase k' phenol k' naphthol


Ultrasphere-Si dry hexane 72.53 47.50

5 x 10-4 M AOT in hexane 27.89 20.30

Ultrasil-NH2 dry hexane 47.52 53.75

5 x 10-4 M AOT in hexane 29.44 40.87








micellar systems in reversed phase HPLC, or pseudo phase liquid

chromatography. In normal phase chromatography, the solvent strength

is varied in order to optimize k'. Increasing the strength of the

mobile phase by increasing the solvent polarity, a decrease in sample

k' value is observed. Preferred solvents used in normal phase to

increase the strength of the mobile phase are acetonitrile, ethanol,

ethylacetate, CHCu3, tetrahydrofuran and isopropyl ether.

Comparing micellar and organic mobile phases for normal phase

HPLC, one cansay that (1) Organic mobile phases have somewhat lower

viscosity than micellar mobile phases, resulting in lower operating

pressures. For example, at a flow rate of 1 ml/min., 30C, the pres-

sure for 5/95 isopropanol/hexane mobile phase was 302 psig vs. 350

psig for 5 x 10-4 M AOT in hexane mobile phase for Ultrasphere-Si

column. Table III shows the viscosity values for most common solvents

used in normal phase HPLC. (2) Surfactants are cheaper than the HPLC

grade organic solvents; an approximate cost per gallon of organic

solvent is thirty dollars vs. twenty dollars for 500 g of Aerosol OT

(Table III). (3) Surfactants are non-toxic vs. some of the most

commonly used organic modifiers, for example tetrahydrofuran. (4)

Micellar mobile phases offer a different selectivity for separations

of compounds, and (5) For both mobile phases, a UV detector can be

used. The molar absorptivity coefficient for Aerosol OT was calcu-

lated to be 0.9986 L/mole cm at a wavelength equal to 254 nanometers

(Table III).








Table III

Viscosity, UV Cutoff and Price per Gallon for 21,22
Most Commonly Used Solvents in Normal Phase HPLC From
UV Cutoff Viscosity Price per
Solvent (nm) (cP, 250C) Gallon

Hexane 195 0.30 28.65

i-Propyl ether 220 0.38 27.85

Tetrahydrofuran 212 0.46 47.15

i-Propanol 205 1.9 24.80

Chloroform 245 0.53 29.85

Acetonitrile 190 0.34 55.85

Ethanol 210 1.08 26.75

Ethylacetate 256 0.43 26.05








Efficiency

In looking for an optimum separation, one which gives adequate

sample resolution with a minimum of time and effort, one of the para-

meters to be considered is column efficiency. The number of theo-

retical plates, N, was calculated in order to compare the performance

of the system under the different experimental conditions studied.

Foley and Dorsey's equation was used to calculate the number of theo-

retical plates.18 videe supra) This equation was preferred over the

most commonly used,

N = 5.54 (tr/W.5 )2 (4)

because the former corrects for the asymmetry of skewed peaks giving

more accurate values for N. The reduced plate height was calculated

by

h = H/dp (5)

where H is the plate height, and dp is the particle diameter.

Table IV shows the theoretical plate numbers, plate heights, re-

duced plate height and asymmetry ratios for phenol in 5 x 10-2 M AOT

in hexane, 5/95 isopropanol/hexane and dry hexane mobile phases for

Ultrasphere-Si and Ultrasil-NH2 columns.

Better efficiencies were observed for 5/95 isopropanol/hexane

mobile phase for both columns. The reason for smaller values of N in
-2
5 x 102 M AOT is the higher viscosity of the micellar solution, hav-

ing smaller diffusion coefficients, therefore, slower mass transfer.

This slow mass transfer results from the slower rate of movement of a

polar sample between the stationary phase and the interior of the








Table IV


Theoretical Plate Numbers (N), Plate Heights (H, mm), Reduced
Plate Heights (h) and Asymmetry Ratios (B/A) for Phenol in
5 x 10-2 M AOT in Hexane, 5/95 Isopropanol/Hexane and Dry
Hexane Mobile Phases for Ultrasphere-Si and Ultrasil-NH2 Columns


Ultrasphere-Si


5 x 10-2 M AOT
in Hexane

3301

0.0454

9.08

1.50


5/95
Isopropanol/Hexane

8174

0.0184

3.68

1.14


Dry Hexane

3987

0.0376

7.40

2.02


Ultrasil-NH2


N

H (mm)

h


0.277


27.7


B/A 2.18


H (mm)


B/A


2409

0.104

10.4

1.22








micelle. Also, higher viscosity in the interior of the micelle in

comparison of the bulk solution contributes to slow the adsorption-

desorption interchange.

Greater values for asymmetry ratio in micellar system were ob-

tained in accordance with the N values. Optimization of the micellar

system has to be done in order to be a feasible medium for separa-

tion.

For Ultrasphere-Si columns, higher values of N were calculated

for the three systems studied. This can be explained in terms of a

smaller particle diameter, 5 pm vs. 10 pm for Ultrasil-NH2. It seems

that the fact of smaller particle diameter dominates over the longer

column for Ultrasil-NH2 and stronger active sites in Ultrasphere-Si.

Lower efficiency was observed for phenol in Ultrasphere-Si in

dry hexane mobile phase. This was expected since it is generally

accepted that with a dry mobile phase, at low water contents, much

lower efficiencies will be obtained because of the strong active

sites of the adsorbent surface, leading to slow adsorption-desorp-

tion kinetics.6 The theoretical plate number for phenol in Ultrasil-

NH2 could not be calculated because of the long retention (more than

three hours).

In Szepezy, Combellas, Claude and Rosset's paper,6 very low and

rapidly changing values of efficiency at low water contents were

pointed out, indicating the necessity for selection and control of

the water content in the mobile phase. For this reason, the next part

of this experiment was performed.




"-


Influence of Water Content on Retention

One of the disadvantages in using normal phase chromatography

is the changes in the chromatographic characteristics (capacity

ratio, selectivities, efficiencies) with increasing water content

of the mobile phase.

In order to observe the influence of the water content on reten-

tion, plots of k' vs. volume percent of H20 in 5 x 10-2 M AOT/hexane

were drawn for both columns. The surfactant concentration in the

mobile phase was chosen above the CMC range to assure the existence

of micelles.

Figures 6 and 7 show the behavior of phenol, naphthol, and 2,4-

dinitrotoluene in Ultrasil-NH2 and Ultrasphere-Si columns. A smooth

change in k' was observed at low water concentrations. For the amino

column, only a slight change was noticed; this can be attributed to

the fact that the silica surface is more active. videe supra)

Retention values are quite sensitive to small changes in mobile

phase with low water concentration. In actual practice, it is diffi-

cult to avoid small changes in mobile phase water content, for several

reasons:23

-The water contents of the starting (supposedly dry) solvents

that comprise the mobile phase vary somewhat.

-The water is readily picked up or lost to the atmosphere, de-

pending on its humidity and water content of the solvent.

-Further changes in mobile phase water content can occur as a

result of contact with the walls of intermediate containers, the

solvent reservoir, and so on.






SLI


0 O~ a) N CD it) q- r') OI -


*r*
4--

to
S- 0O



C 0
"o
04-'<
0

.0) .
C (-'


- *-
-0
Cr I


CL













14-.
.) --
5-

CC
0 to

C r-
00
*r- E























4-,--
(0 .C
*r- Q.
*-



C a








4-- 0
C .U
Q.
5-0 1


O
.4-


oE

4- 0
L. U























E
r-
0
0



*-




40)




C
0






C.

r- 0






C 0.

4- C
C -








0 1
S*r-
C4













0
4-)
O C

.4-
0-1-









+- ---


o 0 0 o
,t ) Nj -




'- -,


The importance of controlling the water content at low water

concentration in order to obtain reproducibility of k' values has

been pointed out by many authors.6'7'23 Szepezy, Combellas, Claude

and Rosset demonstrated in their paper6 that for silica packing, a

small increase in water content corresponds to a large decrease in

k' values. Their k' values change up to one order of magnitude for

dibutylphthalate varying from 100 to 600 ppm of water.

The solubility of water in hexane at 20C, 1 atm, is 0.0111 g

water per 100 g of hexane, or 0.0073 percent V/V.24

Using reversed micellar system, a maximum change of 38 percent

in k' was observed at low water compositions (Table V). This small

change in k' values as the water content is increased can be ex-

plained by the fact that reverse micelles tend to solubilize a large

amount of water in their interior. The first molecules of water are

attached to the counterion (sodium) which connects, via two hydrogen

bridges to two sulfonate molecules, forming a trimer. Additional

water molecules occupy the core of the micelle, and polar substrates

are expected to be localized in this water pool. Small changes in

the amount of water do not affect the retention, since almost all the

water in the system is restricted to the interior of the micelle.

Instead, it has been concluded that small amounts of water are a pre-

requisite to the formation of closed, concentration independent sur-

factant aggregates.25 Large amounts of water cause the micelle to

swell, increase the mean aggregation number and to assume a different

shape; therefore, it will be interesting to find out the amount of

water that the micelle can hold without drastic changes in the k'

value.








Table V

Percent of k' Values Changes as the Water Content
Varies from 0.0 to 0.25 Percent for Ultrasphere-Si
and from 0.0 to 1.0 Percent for Ultrasil-NH2 Columns
in 5 x 10-2 M AOT in Hexane Mobile Phase


% Change % Change % Change
Column Phenol Naphthol 2,4-Dinitrotoluene

Ultrasphere-Si 11.28 10.14 32.53

Ultrasil-NH2 24.03 19.81 38.27














CHAPTER V

CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK

Reverse micellar systems are a promising mobile phase for normal

phase HPLC, as normal micellar systems have been shown to be in re-

versed phase HPLC. As the concentration of surfactant in the mobile

phase is increased, the retention times of polar compounds decrease.

Drastic reductions in k' values for surfactant concentrations above

the critical micellar concentration range were observed. Only small

changes in capacity ratio values for polar solutes were obtained at

low water contents in the mobile phase. Due to the fact that the

amount of water in the mobile phase is critical in order to obtain

reproducible retention values in normal phase chromatography, an

addition of a small amount of water in micellar mobile phases may be

a feasible means to obtain reproducible k' values. Further studies

should be done in this area. It will be interesting to look for the

maximum amount of water than can be held by the micellar mobile phase

without an abrupt change in k' value.

Lower efficiencies were obtained for micellar mobile phases be-

cause of the higher viscosity of these mobile phases, leading to

slower mass transfer. Raising the temperature or adding small amounts

of polar organic modifiers3 may be a good approach to improving effi-

ciencies for reverse micellar mobile phases, and should be investi-

gated.








The selectivity (a), or relative retention of solutes, has a

considerable effect on resolution. The value of a should be as large

as possible, other factors being equal. The use of a micellar mobile

phase and changing temperature are two options that can be studied in

the future in order to increase a, thereby improving resolution.

A similar study can be performed using a cyano column and more

polar solutes in order to clarify the retention mechanism of polar

compounds in reverse micellar systems.

Aerosol OT is an anionic surfactant. A study of different anionic

surfactants, cationic surfactants and nonionic surfactants will give

different types of aggregation systems from which the separation mecha-

nism can be observed.














REFERENCES


1. R.E. Majors. Recent advances in HPLC packing and columns. J.
Chromatogr. Sci. 18, 488-511 (1980).

2. D.W. Armstrong and S.J. Henry. Use of an aqueous micellar mo-
bile phase for separation of phenols and polynuclear aromatic
hydrocarbons via HPLC. J. Liq. Chromatogr. 3, 657-662 (1980).

3. J.G. Dorsey, M.T. DeEchegaray and J.S. Landy. Efficiency en-
hancement in micellar liquid chromatography. Anal. Chem. 55,
924-928 (1983).

4. S.R. Abbot. Practical aspects of normal-phase chromatography.
J. Chromatogr. Sci. 18, 540-550 (1980).

5. L.R. Snyder and J.J. Kirkland. "Introduction to Modern Liquid
Chromatography," 2nd Ed., John Wiley and Sons, New York, 1979,
pp. 272-277.

6. L. Szepezy, C. Combellas, M. Claude and R. Rosset. Influence
of water content of the mobile phase on chromatographic perform-
ance in adsorption chromatography. J. Chromatogr. 237, 65-78
(1982).

7. L.R. Snyder. "Principles of Adsorption Chromatography," Marcel
Dekker, Inc., New York, 1968, pp. 157-159.

8. J.J. Kirkland. "Modern Practice of Liquid Chromatography,"
John Wiley and Sons, New York, 1971, pp. 215-218.

9. R.W. Yost, L.S. Ettre and R.D. Conlon. "Practical Liquid Chromo-
tography," Perkin Elmer Corporation, New York, 1980, pp. 60-62.

10. D.W. Armstrong. Pseudophase liquid chromatography: applications
to TLC. J. Liq. Chromatogr. 3, 895-900 (1980).

11. D.W. Armstrong. Application of pseudophase liquid chromato-
graphy (PLC). Amer. Lab. 13, 14-20 (1981).

12. K.A. Dill and P.J. Flory. Proc. Natl. Acad. Sci. USA 78, 676-
680 (1981).

13. M.T. DeEchegaray. Effect on temperature on the elution of non-
ionic solutes in micelles containing mobile phases for reverse
phase HPLC, Master's Thesis, University of Florida, Gainesville,
1982, p. 8.




.J-T


14. H.F. Eicke. Surfactants in nonpolar solvents aggregation and
micellization. Topics in Current Chemistry 87, 1-83 (1980).

15. B. Lindman and H. Wennerstrom. Micelles amphiphilic aggrega-
tion in aqueous solution. Topics in Current Chemistry 87, 85-
145 (1980).

16. J.H. Fendler. Interaction and reactions in reversed micellar
system. Acc. Chem. Res. 9, 153-161 (1976).

17. K.L. Mittal. "Micellization, Solubilization and Microemul-
sions," 1, Plenum, New York, 1977, pp. 429-454.

18. J.P. Foley and J.G. Dorsey. Equations for calculation of
chromatographic figures of merit for ideal and skewed peaks.
Anal. Chem. 55, 730-737 (1983).

19. M. Tang and S.N. Deming. Interfacial tension effects of non-
ionic surfactants in reversed-phase liquid chromatography.
Anal. Chem. 55, 425-428 (1983).

20. D.W. Armstrong and F. Nome. Partitioning behavior of solutes
eluted with micellar mobile phase in liquid chromatography.
Anal. Chem. 53, 1662-1666 (1981).

21. L.R. Snyder and J.J. Kirland. "Introduction to Modern Liquid
Chromatography," 2nd Ed., John Wiley and Sons, New York, 1979,
pp. 248-250.

22. Rainin Catalog, 1983, pp. 124-125,

23. L.R. Snyder and J.J. Kirland, "Introduction to Modern Liquid
Chromatography," 2nd Ed., John Wiley and Sons, New York, 1979,
pp. 374-376.

24. C. Black, G.G. Joris and H.S. Taylor. The solubility of water
in hydrocarbons. J. Chem. Phys. 16, 53 (1948).


25. H.F. Eicke and H. Chisten. Is water critical to the formation
of micelles in apolar media? Helv. Chim. Acta. 61, 2258-2263
(1978).













BIOGRAPHICAL SKETCH

Maria A. Hernandez Torres was born in Mayaguez, Puerto Rico,

on May 29th, 1958. She received her elementary and high school edu-

cation at the Colegio de La Milagrosa, MayagUez, Puerto Rico. In

1980, she completed her Bachelor of Science degree in chemistry at

the University of Puerto Rico Mayaguez Campus. She is now pursuing

a Master of Science degree in analytical chemistry at the University

of Florida.








I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate in scope and quality, as a thesis for the degree
of Master of Science.




John]G. Dorsey, Chairman \
Ass itant Professor of Chemistry^


I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate in scope and quality, as a thesis for the degree
of Master of Science.



ames D. Winefordfer
graduate Research Professor
of Chemistry

This thesis was submitted to the Graduate Faculty of the Depart-
ment of Chemistry in the College of Liberal Arts and Sciences and to
the Graduate School, and was accepted for partial fulfillment of the
requirements of the degree of Master of Science.




Dean for Graduate Studies and
December 1983 Research










































UNIVERSITY OF FLORIDA
11 1 11111111 IIIIlllllllllllllllllllllllll1111111 111
3 1262 08556 7534