OPTIMIZATION
OF THE GRADIENT HPLC SEPARATION OF
SELECTED PHTHALATES
USING THE OVERLAPPING RESOLUTION
MAPPING TECHNIQUE Y. J. YAO, M. R. KHAN, C. P. ONG, H. K. LEE, and S. F. Y. LI* Department of Chemistry, National University of Singapore Kent Ridge, Republic of Singapore 0511
(Received March 1991) Abstract. Optimization procedures for the reversed-phase separations of six phthalates using the isoselective
multisolvent gradient elution (IMGE) system are described. A systematic experimental design has been employed to gather retention data on the compounds in a mixture. The data were then fitted into a second-order polynomial equation and an overlapping resolution mapping (ORM) technique of data analysis was subsequently used to establish the optimum solvent mixture for the highest resolution of all adjacent peaks in the chromatogram.
Introduction
Phthalate esters are used extensively as plasticizers in the formulation of polymers. Their migration into the surroundings has caused concern over their presence in the environment. Phthalates have been found as contaminants in m a n y samples of biological origin and foods such as milk (Cerbulis et al., 1967), h u m a n blood (Jaeger et al., 1972), egg (Ishida et al., 1981), fish (Williams, 1973), and water (Mayer et al., 1972). Phthalates were also found to migrate into chemicals, solvents and laboratory equipment (Ishida et al., 1980). Because of the widespread contamination and probable toxicity of the phthalates, there is a need to analyze for them. In our investigation, H P L C is used to study six phthalates (dimethylphthalate, diethylphthalate, dibutylphthalate, benzyl-n-butylphthalate, bis (2-ethylhexyl)phthalate and diallylphthalate). Five of them (excluding the last c o m p o u n d ) are on the priority pollutant list of the United States Environmental Protection Agency (USEPA). In this paper, the optimization of the gradient elution reversed-phase H P L C separation of the six phthalates is investigated. A mixture design statistical approach, combined with overlapping resolution mapping (ORM) is employed. This method requires as few as seven preliminary experiments based on the Snyder solvent selectivity triangle (Snyder, 1974) to predict the mobile phase compositions for optimal separation. A quaternary mobile phase system is used to provide the full range o f solvent strength and selectivity effects. The solvent system adopted is that of isoselective multisolvent gradient elution (IMGE). This system involves changing the ratio o f the carrier solvent (e.g. water in reversed-phase systems) to organic modifiers to achieve separation. During a typical run,
* Author to whom correspondence should be addressed. Environmental Monitoring and Assessment 19: 83-91, 1991. 9 1991Kluwer Academic Publishers. Printed in the Netherlands.
84
Y.L. YAO ET AL.
the solvent strength changes. However, separation selectivity remains constant as the ratio of the modifiers is constant.
Experimental INSTRUMENTATION All separations were carried out on a Perkin-Elmer Series 4 gradient pump equipped with a Micro UVIS20 (Carlo Erba, Italy) UV spectrophotometric detector and a Hewlett Packard 3390A integrator (California, U.S.A.). Columns used in this study included: (1) a 15 cm X 6.0 mm Shimadzu Shimpack CLC-ODS column (Kyoto, Japan); (2) a 25 cm X 4.6 mm Whatman Partisil-50DS 3 column (New Jersey, U.S.A.); and (3) a 1lcm X 4.7 mm Whatman Partisphere-5 Cl8 column (New Jersey, U.S.A.). Samples were injected with a Rheodyne 7125 injector. For the detection of the phthalates, the detector wavelength was set at 224 nm. A mobile phase flow rate of 1.0 ml/min was used, except in the case of isopropanol-water binary mixture (0.6 ml/min). All chromatographic runs were duplicated with reproducibility within +_ 2%. The void volume was obtained for all mobile phases with methanol as the unretained component. CHEMICALS AND REAGENTS
Dimethylphthalate, diethylphthalate, dibutylphthalate, benzyl-n-butylphthalate, diallylphthalate and bis(2-ethylhexyl)phthalate (at least 97% purity) were purchased from Fluka Chemie AG (Switzerland). Standard solutions were prepared by dissolving appropriate amounds of the phthalates in methanol. HPLC-grade methanol (J. T. Baker, U.S.A.) and acetonitrile (Ajax, Australia) and AR-grade isopropanol (Ajax) were used as the mobile phase modifiers. Milli-Q-treated water (Millipore, U.S.A.) was the solvent carrier. All solvents were filtered and thoroughly degassed with helium before use.
Results and Discussion In our previous investigation (Khan et al., 1990), the ORM technique under isocratic elution reversed-phase HPLC conditions was successfully employed to optimize the separation of five of the phthalates (excluding bis(2-ethylhexyl)phthalate). However, with the addition of bis(2-ethylhexyl)phthalate, the solvent strength of the mobile phase had to be increased to elute this compound. This resulted in the first five peaks eluting out too quickly. Satisfactory separation of all the six peaks was not achieved. To overcome this difficulty, gradient elution using the IMGE solvent system was considered in the present work. The first step in the optimization scheme was to define the three verticles of the solvent selectivity triangle (Snyder, 1974), which correspond to binary solvent mixtures of methanol/water, acetonitrile/water and isopropanol/water. All points in the triangle have the same solvent strength or eluting power. In IMGE, as the solvent strength is continuously increased during the run, a solvent selectivity prism can be adopted (Kirkland et al., 1983). The schematic in Figure 1 represents the solvent selectivity prism used in this study. The solvent compositions for the seven experimental runs are shown in Table 1.
OPTIMIZATION OF GRADIENT HPLC SEPARATION
D
85
A
H B Fig. 1.
F
G
E x p e r i m e n t a l design for seven gradient elution runs to obtain basic data for o p t i m i z a t i o n calculation.
Solvent compositions (A-G) are given in Table I.
The first vertex was established with a methanol/water mixture by using a linear gradient of 15 minutes, and a further 15 minutes of constant final methanol/water composition (mobile phase A in Table I). In this work, a total analysis time of 30 minutes was selected as a time constraint. The composition of the other two binary mixtures (mobile phases B and C) having the same initial solvent strength can be calculated using Equation (1) (Berridge, 1985) S, = ~,4',
(1)
where q5i is the volume fraction of the ith solvent and Si is the solvent strength, being 3.0 for methanol, 3.1 for acetonitrile, 4.2 for isopropanol and 0 for water (Snyder et al., 1979). Four other runs were conducted with mixtures of compositions corresponding to the midpoints of the sides and the center of the solvent selectivity triangle. The initial and final volume percentages of all mobile phase mixtures used are listed in Table I, and the respective retention times for the phthalates are listed in Table II. Having completed the experiments using the seven mobile phase systems, the resolution between adjacent peak pairs in the chromatograms obtained were calculated using Equation (2). R = 2 (t2- tl) w I ~- W 2
(2)
86
Y.L. YAO ET AL.
TABLE I Mobile phase systems used in IMGE scheme. Solvent (V/V %) Mobile phase A
Initial Final Initial Final Initial Final Initial Final Initial Final Initial Final Initial Final
B C D E F G
water
MeOH a
30.0 0.0 32.3 3.2 45.2 23.8 31.1 1.6 37.6 13.5 38.7 13.5 35.8 9.0
70.0 100.0
ACN b
67.7 96.8
35.0 50.0 35.0 50.0
54.8* 76.2*
33.9 48.4 33.9 48.4 22.6 32.3
23.3 33.3
IPA c
27.4 38.1 27.4 38.1 18.3 25.4
* Minor adjustment was made to this composition so that the last compound could be eluted out within the set time constraint. a MeOH = methanol. b A C N ----acetonitrile. c IPA = isopropanol. w h e r e fi, t2 a r e r e t e n t i o n t i m e s , a n d w~ a n d w2 a r e p e a k w i d t h s a t t h e b a s e l i n e o f t w o adjacent peaks. The R values obtained polynomial
were used to estimate
equation (Equation
the coefficients of a second-order
(3)) which describes the resolution surface. TABLE II
Retention times (in minutes) of the six phthalates in each of the eluent mixtures listed in Table I. compounds* mobile phases
DMP
DEP
DAP
BBP
DBP
DOP
A B C D E F G
4.165 4.360 5.550 4.330 3.380 3.570 3.760
6.125 5.960 6.550 6.410 4.390 4.600 5.080
7.840 6.990 7.540 7.820 5.450 5.300 6.010
13.645 11.930 11.950 13.330 10.880 9.660 11.230
14.025 12.970 12.130 14.010 11.050 10.770 11.960
22.095 27.600 26.030 23.940 23.870 24.840 24.330
DMP--dimethylphthalate; DEP=diethylphthalate; DAP=diallylphthalate; BBP=benzyl-n-butylphthalate; DBP=dibutylphthalate; DOP=bis(2-ethylhexyl)pht halate. *
OPTIMIZATION
OF
GRADIENT
HPLC
87
SEPARATION
R = a,xl + a2x2 Jr- a3x3 Jr- alexlx2 + al3xlx3 -}- ae3x2x3 q- a,e3xIx2x3
(3)
a~ are coefficients a n d x~are volume fractions of mobile phase mixtures A, B a n d C. In this study, a m i n i m u m resolution of unity for the various peak pairs was specified. With the aid of a BASIC p r o g r a m (Berridge, 1985), resolution c o n t o u r maps were generated for every adjacent peak pair. Subsequent overlapping of these resolution maps reveals an area in the solvent triangle in which the desired resolution can be achieved for all components, as illustrated by the region denoted by ~ in Figure 2. F r o m that region (where//: taken corresponds to 100% A C N : water binary mixture), a gradient r u n with initial mobile phase composition o f ( A C N : H 2 0 = 68.0 : 32.0) a n d final composition of ( A C N : H20 = 97.0 : 3.0) was chosen to confirm the success of the O R M scheme. Figure 3 shows the gradient elution c h r o m a t o g r a m of the test mixture carried out on C o l u m n 1 with this o p t i m u m mobile phase. Capacity factors of the six phthalates are listed in Table Ill. A p p a r e n t resolution for each peak pair in the mixture was greater than 1.0. As can be seen, all the six phthalates are satisfactorily separated within 27 minutes. The advantages of the I M G E separation over isocratic conditions for improved selectivity are thus clearly illustrated. To further demonstrate the usefulness of the O R M scheme, the optimal I M G E conditions o b t a i n e d for the six phthalates were extended to the use of two other C,~
0 5
.
I0
i00 95
.
15
90
20
. . . .
25 30
55
.
95 i00
-
. . . . .
-
.
.
.
.
.
.
.
. .
.
. .
.
.
-
.
.
-
.
.
.
9 .
50 45
.
+
+
+ + *
.
"
+
.
40 35
+
~ + +
~
~
*
*
+
*
-
~
+
30 +
+
+
~ *
e 9
~
25
-
* ~
20 +
15
. . . . . . . . . .
.
.
.
.
.
.
.
.
.
+
+ 9
IPA/wdter
60 55
. . . . . r.
+
.
.
65
+ ......
. . . . . .
.
70 -
. . . .
. . . . . . .
.
90
. . . . . . . . . . .
.
75 9 85
-
. . . . . .
70
80
. . . . .
75
-
-
. . . . . .
60
65
-
. . . . .
40 45 50
80
. . . . .
35
MeOH/water
85
. . . . .
o
.
.
.
.
.
.
o
~
.
* # # * . 5
*
+
+
~
~
#
#
0
i00
0
ACN/wate[
Fig. 2. Overlappingresolution mapping diagram for all fivepeak paris. - .: R < 0.7; - - : 0.7 -< R < 0.8; +-t- : 0.8--< R < 0.9; ** :0.9-< R < 1.0;~ ~ : R > _ 1.0.
88
Y.L. YAO ET AL.
2 3 4
[
[
I
0
2 6 min Time
Fig. 3. Chromatogram for optimum IMGE run on Column 1 obtained by using mobile phase of initial composition (ACN : H20 = 68.0 : 32.0), and final composition (ACN : H20 = 97.0 : 3.0). Peaks (1) dimethylphthalate; (2) diethylphthalate; (3) diallylphthalate; (4) benzyl-n-butylphthalate; (5) dibutyl-phthalate; (6) bis (2-ethylbexyl)phthalate.
columns (Columns 2 and 3). The chromatograms obtained (Figure 4 and 5) show well-resolved peaks, comparable or even better than the chrornatogram in Figure 3. With the three columns, all compounds were eluted out within reasonable analysis times. These results suggest that the optimum mobile phase composition once established for one C~s column could be easily extended for use with other C~8 columns. The advantages arising from this are substantial. F o r example, the preliminary analyses could be performed on a shorter column posessing the appropriate separation efficiency, followed by subsequent
OPTIMIZATION OU GRADIENT HPLC SEPARATION
89
TABLE III Capacity factors of the six phthalates with optimum mobile phase compositions using various columns. Compounds** Columns*
DMP
DEP
DAP
BBP
DBP
DOP
1 2 3
1.571 0.200 0.226
2.439 0.500 0.524
2.994 0.694 0.704
5.901 1.603 2.112
6.496 1.732 2.499
15.431 4.452 11,355
* Column details are given in the text. ** Identities of compounds are given in Table II. s e p a r a t i o n s b e i n g c a r r i e d o u t o n l o n g e r c o l u m n s w i t h e q u i v a l e n t o r b e t t e r efficiencies f o r improved resolution. Moreover, the small number of experiments necessary to obtain the o p t i m u m m o b i l e p h a s e c o m p o s i t i o n s results in l o w e r s o l v e n t c o n s u m p t i o n as o p p o s e d to
!
113 4
Is
6
l 20
o
min
Time
Fig. 4.
Chromatogram for optimum IMGE run on Column 2. HPLC conditions and peak designations are the same as Figure 3.
90
Y.L.
YAO
ET At..
3
15
G
[o
[15min Time
Fig. 5. Chromatogram for optimum IMGE run on Column 3. HPLC conditions and peak designations are the same as Figure 3.
conventional trial-and-error approaches. The simplicity of the I M G E approach is an attractive feature of this systematic optimization scheme, and the way in which optimum conditions are easily achieved fully demonstrates its potential as a method development tool. Acknowledgements The authors gratefully acknowledge the financial support of the National University of Singapore, and the technical assistance of Dr. K h o o Soo Beng, Ms Tan G e o k Kheng and Ms Frances Lim.
OPTIMIZATION OF GRADIENT HPLC SEPARATION
91
References Berridge, J. C.: 1985, Techniquesfor the Automated Optimization of HPLC Separations', Wiley, Chichester, pp. 70-94. Cerbulis, J., Ard, J. S.: 1967, 'Isolation and detection of dioctyl phthalate from milk lipids', J. Assoc. Offic. Anal Chem., 50, 646-650. Ishida, M., Suyama, K. and Ard, J.S.: 1980, 'Background contamination by phthalates commonly encountered in the chromatographic analysis of lipid samples', J. Chromatogr., 189, 421-424. Ishida, M., Suyama, K. and Adachi, S.: 1981, 'Occurrence of dibutyl and di (2-ethylhexyl)phthalate in chicken eggs', J. Agric. Food Chem., 29, 72-74. Jaeger, R. J. and Rubin, R. J.: 1972, Migration ofaphthalate esterplasticizerfrompolyvinylchloride bloodbags into stored human blood and its localization in human tissues', N. Engl. J. Med., 287, 1114-1118. Kirkland, J. J. and Glajch, J. L.: 1983, 'Optimization of mobile phases for multisolvent gradient elution liquid chromatography', J. Chromatogr., 255, 27-39. Khan, M. R., Ong, C. P., Li, S. F. Y. and Lee, H. K.: 1990, 'Optimization of the isocratic HPLC separation of selected phthalates using the overlapping resolution mapping technique', J. Chromatogr., 513, 360-366. Mayer, F. L., Jr., Stalling, D. L. and Johnson, J. L.: 1972, 'Phthalate esters as environmental contaminants', Nature (London), 238, 411-413. Snyder, L. R.: 1974, 'Classification of the solvent properties of common liquids', J. Chromatogr., 92, 223-230. Snyder, L.R., Dolan, J.W. and Grant, J.R.: 1979, 'Gradient elution in high performance liquid chromatography. I. Theoretical basis for reversed phase system', J. Chromatogr., 165, 3-30. Williams, D.T.: 1973, 'Dibutyl- and di(2-ethylhexyl)phthalate in fish', J. Agric. Food Chem., 21, 1128-1129.
OF THE GRADIENT HPLC SEPARATION OF
SELECTED PHTHALATES
USING THE OVERLAPPING RESOLUTION
MAPPING TECHNIQUE Y. J. YAO, M. R. KHAN, C. P. ONG, H. K. LEE, and S. F. Y. LI* Department of Chemistry, National University of Singapore Kent Ridge, Republic of Singapore 0511
(Received March 1991) Abstract. Optimization procedures for the reversed-phase separations of six phthalates using the isoselective
multisolvent gradient elution (IMGE) system are described. A systematic experimental design has been employed to gather retention data on the compounds in a mixture. The data were then fitted into a second-order polynomial equation and an overlapping resolution mapping (ORM) technique of data analysis was subsequently used to establish the optimum solvent mixture for the highest resolution of all adjacent peaks in the chromatogram.
Introduction
Phthalate esters are used extensively as plasticizers in the formulation of polymers. Their migration into the surroundings has caused concern over their presence in the environment. Phthalates have been found as contaminants in m a n y samples of biological origin and foods such as milk (Cerbulis et al., 1967), h u m a n blood (Jaeger et al., 1972), egg (Ishida et al., 1981), fish (Williams, 1973), and water (Mayer et al., 1972). Phthalates were also found to migrate into chemicals, solvents and laboratory equipment (Ishida et al., 1980). Because of the widespread contamination and probable toxicity of the phthalates, there is a need to analyze for them. In our investigation, H P L C is used to study six phthalates (dimethylphthalate, diethylphthalate, dibutylphthalate, benzyl-n-butylphthalate, bis (2-ethylhexyl)phthalate and diallylphthalate). Five of them (excluding the last c o m p o u n d ) are on the priority pollutant list of the United States Environmental Protection Agency (USEPA). In this paper, the optimization of the gradient elution reversed-phase H P L C separation of the six phthalates is investigated. A mixture design statistical approach, combined with overlapping resolution mapping (ORM) is employed. This method requires as few as seven preliminary experiments based on the Snyder solvent selectivity triangle (Snyder, 1974) to predict the mobile phase compositions for optimal separation. A quaternary mobile phase system is used to provide the full range o f solvent strength and selectivity effects. The solvent system adopted is that of isoselective multisolvent gradient elution (IMGE). This system involves changing the ratio o f the carrier solvent (e.g. water in reversed-phase systems) to organic modifiers to achieve separation. During a typical run,
* Author to whom correspondence should be addressed. Environmental Monitoring and Assessment 19: 83-91, 1991. 9 1991Kluwer Academic Publishers. Printed in the Netherlands.
84
Y.L. YAO ET AL.
the solvent strength changes. However, separation selectivity remains constant as the ratio of the modifiers is constant.
Experimental INSTRUMENTATION All separations were carried out on a Perkin-Elmer Series 4 gradient pump equipped with a Micro UVIS20 (Carlo Erba, Italy) UV spectrophotometric detector and a Hewlett Packard 3390A integrator (California, U.S.A.). Columns used in this study included: (1) a 15 cm X 6.0 mm Shimadzu Shimpack CLC-ODS column (Kyoto, Japan); (2) a 25 cm X 4.6 mm Whatman Partisil-50DS 3 column (New Jersey, U.S.A.); and (3) a 1lcm X 4.7 mm Whatman Partisphere-5 Cl8 column (New Jersey, U.S.A.). Samples were injected with a Rheodyne 7125 injector. For the detection of the phthalates, the detector wavelength was set at 224 nm. A mobile phase flow rate of 1.0 ml/min was used, except in the case of isopropanol-water binary mixture (0.6 ml/min). All chromatographic runs were duplicated with reproducibility within +_ 2%. The void volume was obtained for all mobile phases with methanol as the unretained component. CHEMICALS AND REAGENTS
Dimethylphthalate, diethylphthalate, dibutylphthalate, benzyl-n-butylphthalate, diallylphthalate and bis(2-ethylhexyl)phthalate (at least 97% purity) were purchased from Fluka Chemie AG (Switzerland). Standard solutions were prepared by dissolving appropriate amounds of the phthalates in methanol. HPLC-grade methanol (J. T. Baker, U.S.A.) and acetonitrile (Ajax, Australia) and AR-grade isopropanol (Ajax) were used as the mobile phase modifiers. Milli-Q-treated water (Millipore, U.S.A.) was the solvent carrier. All solvents were filtered and thoroughly degassed with helium before use.
Results and Discussion In our previous investigation (Khan et al., 1990), the ORM technique under isocratic elution reversed-phase HPLC conditions was successfully employed to optimize the separation of five of the phthalates (excluding bis(2-ethylhexyl)phthalate). However, with the addition of bis(2-ethylhexyl)phthalate, the solvent strength of the mobile phase had to be increased to elute this compound. This resulted in the first five peaks eluting out too quickly. Satisfactory separation of all the six peaks was not achieved. To overcome this difficulty, gradient elution using the IMGE solvent system was considered in the present work. The first step in the optimization scheme was to define the three verticles of the solvent selectivity triangle (Snyder, 1974), which correspond to binary solvent mixtures of methanol/water, acetonitrile/water and isopropanol/water. All points in the triangle have the same solvent strength or eluting power. In IMGE, as the solvent strength is continuously increased during the run, a solvent selectivity prism can be adopted (Kirkland et al., 1983). The schematic in Figure 1 represents the solvent selectivity prism used in this study. The solvent compositions for the seven experimental runs are shown in Table 1.
OPTIMIZATION OF GRADIENT HPLC SEPARATION
D
85
A
H B Fig. 1.
F
G
E x p e r i m e n t a l design for seven gradient elution runs to obtain basic data for o p t i m i z a t i o n calculation.
Solvent compositions (A-G) are given in Table I.
The first vertex was established with a methanol/water mixture by using a linear gradient of 15 minutes, and a further 15 minutes of constant final methanol/water composition (mobile phase A in Table I). In this work, a total analysis time of 30 minutes was selected as a time constraint. The composition of the other two binary mixtures (mobile phases B and C) having the same initial solvent strength can be calculated using Equation (1) (Berridge, 1985) S, = ~,4',
(1)
where q5i is the volume fraction of the ith solvent and Si is the solvent strength, being 3.0 for methanol, 3.1 for acetonitrile, 4.2 for isopropanol and 0 for water (Snyder et al., 1979). Four other runs were conducted with mixtures of compositions corresponding to the midpoints of the sides and the center of the solvent selectivity triangle. The initial and final volume percentages of all mobile phase mixtures used are listed in Table I, and the respective retention times for the phthalates are listed in Table II. Having completed the experiments using the seven mobile phase systems, the resolution between adjacent peak pairs in the chromatograms obtained were calculated using Equation (2). R = 2 (t2- tl) w I ~- W 2
(2)
86
Y.L. YAO ET AL.
TABLE I Mobile phase systems used in IMGE scheme. Solvent (V/V %) Mobile phase A
Initial Final Initial Final Initial Final Initial Final Initial Final Initial Final Initial Final
B C D E F G
water
MeOH a
30.0 0.0 32.3 3.2 45.2 23.8 31.1 1.6 37.6 13.5 38.7 13.5 35.8 9.0
70.0 100.0
ACN b
67.7 96.8
35.0 50.0 35.0 50.0
54.8* 76.2*
33.9 48.4 33.9 48.4 22.6 32.3
23.3 33.3
IPA c
27.4 38.1 27.4 38.1 18.3 25.4
* Minor adjustment was made to this composition so that the last compound could be eluted out within the set time constraint. a MeOH = methanol. b A C N ----acetonitrile. c IPA = isopropanol. w h e r e fi, t2 a r e r e t e n t i o n t i m e s , a n d w~ a n d w2 a r e p e a k w i d t h s a t t h e b a s e l i n e o f t w o adjacent peaks. The R values obtained polynomial
were used to estimate
equation (Equation
the coefficients of a second-order
(3)) which describes the resolution surface. TABLE II
Retention times (in minutes) of the six phthalates in each of the eluent mixtures listed in Table I. compounds* mobile phases
DMP
DEP
DAP
BBP
DBP
DOP
A B C D E F G
4.165 4.360 5.550 4.330 3.380 3.570 3.760
6.125 5.960 6.550 6.410 4.390 4.600 5.080
7.840 6.990 7.540 7.820 5.450 5.300 6.010
13.645 11.930 11.950 13.330 10.880 9.660 11.230
14.025 12.970 12.130 14.010 11.050 10.770 11.960
22.095 27.600 26.030 23.940 23.870 24.840 24.330
DMP--dimethylphthalate; DEP=diethylphthalate; DAP=diallylphthalate; BBP=benzyl-n-butylphthalate; DBP=dibutylphthalate; DOP=bis(2-ethylhexyl)pht halate. *
OPTIMIZATION
OF
GRADIENT
HPLC
87
SEPARATION
R = a,xl + a2x2 Jr- a3x3 Jr- alexlx2 + al3xlx3 -}- ae3x2x3 q- a,e3xIx2x3
(3)
a~ are coefficients a n d x~are volume fractions of mobile phase mixtures A, B a n d C. In this study, a m i n i m u m resolution of unity for the various peak pairs was specified. With the aid of a BASIC p r o g r a m (Berridge, 1985), resolution c o n t o u r maps were generated for every adjacent peak pair. Subsequent overlapping of these resolution maps reveals an area in the solvent triangle in which the desired resolution can be achieved for all components, as illustrated by the region denoted by ~ in Figure 2. F r o m that region (where//: taken corresponds to 100% A C N : water binary mixture), a gradient r u n with initial mobile phase composition o f ( A C N : H 2 0 = 68.0 : 32.0) a n d final composition of ( A C N : H20 = 97.0 : 3.0) was chosen to confirm the success of the O R M scheme. Figure 3 shows the gradient elution c h r o m a t o g r a m of the test mixture carried out on C o l u m n 1 with this o p t i m u m mobile phase. Capacity factors of the six phthalates are listed in Table Ill. A p p a r e n t resolution for each peak pair in the mixture was greater than 1.0. As can be seen, all the six phthalates are satisfactorily separated within 27 minutes. The advantages of the I M G E separation over isocratic conditions for improved selectivity are thus clearly illustrated. To further demonstrate the usefulness of the O R M scheme, the optimal I M G E conditions o b t a i n e d for the six phthalates were extended to the use of two other C,~
0 5
.
I0
i00 95
.
15
90
20
. . . .
25 30
55
.
95 i00
-
. . . . .
-
.
.
.
.
.
.
.
. .
.
. .
.
.
-
.
.
-
.
.
.
9 .
50 45
.
+
+
+ + *
.
"
+
.
40 35
+
~ + +
~
~
*
*
+
*
-
~
+
30 +
+
+
~ *
e 9
~
25
-
* ~
20 +
15
. . . . . . . . . .
.
.
.
.
.
.
.
.
.
+
+ 9
IPA/wdter
60 55
. . . . . r.
+
.
.
65
+ ......
. . . . . .
.
70 -
. . . .
. . . . . . .
.
90
. . . . . . . . . . .
.
75 9 85
-
. . . . . .
70
80
. . . . .
75
-
-
. . . . . .
60
65
-
. . . . .
40 45 50
80
. . . . .
35
MeOH/water
85
. . . . .
o
.
.
.
.
.
.
o
~
.
* # # * . 5
*
+
+
~
~
#
#
0
i00
0
ACN/wate[
Fig. 2. Overlappingresolution mapping diagram for all fivepeak paris. - .: R < 0.7; - - : 0.7 -< R < 0.8; +-t- : 0.8--< R < 0.9; ** :0.9-< R < 1.0;~ ~ : R > _ 1.0.
88
Y.L. YAO ET AL.
2 3 4
[
[
I
0
2 6 min Time
Fig. 3. Chromatogram for optimum IMGE run on Column 1 obtained by using mobile phase of initial composition (ACN : H20 = 68.0 : 32.0), and final composition (ACN : H20 = 97.0 : 3.0). Peaks (1) dimethylphthalate; (2) diethylphthalate; (3) diallylphthalate; (4) benzyl-n-butylphthalate; (5) dibutyl-phthalate; (6) bis (2-ethylbexyl)phthalate.
columns (Columns 2 and 3). The chromatograms obtained (Figure 4 and 5) show well-resolved peaks, comparable or even better than the chrornatogram in Figure 3. With the three columns, all compounds were eluted out within reasonable analysis times. These results suggest that the optimum mobile phase composition once established for one C~s column could be easily extended for use with other C~8 columns. The advantages arising from this are substantial. F o r example, the preliminary analyses could be performed on a shorter column posessing the appropriate separation efficiency, followed by subsequent
OPTIMIZATION OU GRADIENT HPLC SEPARATION
89
TABLE III Capacity factors of the six phthalates with optimum mobile phase compositions using various columns. Compounds** Columns*
DMP
DEP
DAP
BBP
DBP
DOP
1 2 3
1.571 0.200 0.226
2.439 0.500 0.524
2.994 0.694 0.704
5.901 1.603 2.112
6.496 1.732 2.499
15.431 4.452 11,355
* Column details are given in the text. ** Identities of compounds are given in Table II. s e p a r a t i o n s b e i n g c a r r i e d o u t o n l o n g e r c o l u m n s w i t h e q u i v a l e n t o r b e t t e r efficiencies f o r improved resolution. Moreover, the small number of experiments necessary to obtain the o p t i m u m m o b i l e p h a s e c o m p o s i t i o n s results in l o w e r s o l v e n t c o n s u m p t i o n as o p p o s e d to
!
113 4
Is
6
l 20
o
min
Time
Fig. 4.
Chromatogram for optimum IMGE run on Column 2. HPLC conditions and peak designations are the same as Figure 3.
90
Y.L.
YAO
ET At..
3
15
G
[o
[15min Time
Fig. 5. Chromatogram for optimum IMGE run on Column 3. HPLC conditions and peak designations are the same as Figure 3.
conventional trial-and-error approaches. The simplicity of the I M G E approach is an attractive feature of this systematic optimization scheme, and the way in which optimum conditions are easily achieved fully demonstrates its potential as a method development tool. Acknowledgements The authors gratefully acknowledge the financial support of the National University of Singapore, and the technical assistance of Dr. K h o o Soo Beng, Ms Tan G e o k Kheng and Ms Frances Lim.
OPTIMIZATION OF GRADIENT HPLC SEPARATION
91
References Berridge, J. C.: 1985, Techniquesfor the Automated Optimization of HPLC Separations', Wiley, Chichester, pp. 70-94. Cerbulis, J., Ard, J. S.: 1967, 'Isolation and detection of dioctyl phthalate from milk lipids', J. Assoc. Offic. Anal Chem., 50, 646-650. Ishida, M., Suyama, K. and Ard, J.S.: 1980, 'Background contamination by phthalates commonly encountered in the chromatographic analysis of lipid samples', J. Chromatogr., 189, 421-424. Ishida, M., Suyama, K. and Adachi, S.: 1981, 'Occurrence of dibutyl and di (2-ethylhexyl)phthalate in chicken eggs', J. Agric. Food Chem., 29, 72-74. Jaeger, R. J. and Rubin, R. J.: 1972, Migration ofaphthalate esterplasticizerfrompolyvinylchloride bloodbags into stored human blood and its localization in human tissues', N. Engl. J. Med., 287, 1114-1118. Kirkland, J. J. and Glajch, J. L.: 1983, 'Optimization of mobile phases for multisolvent gradient elution liquid chromatography', J. Chromatogr., 255, 27-39. Khan, M. R., Ong, C. P., Li, S. F. Y. and Lee, H. K.: 1990, 'Optimization of the isocratic HPLC separation of selected phthalates using the overlapping resolution mapping technique', J. Chromatogr., 513, 360-366. Mayer, F. L., Jr., Stalling, D. L. and Johnson, J. L.: 1972, 'Phthalate esters as environmental contaminants', Nature (London), 238, 411-413. Snyder, L. R.: 1974, 'Classification of the solvent properties of common liquids', J. Chromatogr., 92, 223-230. Snyder, L.R., Dolan, J.W. and Grant, J.R.: 1979, 'Gradient elution in high performance liquid chromatography. I. Theoretical basis for reversed phase system', J. Chromatogr., 165, 3-30. Williams, D.T.: 1973, 'Dibutyl- and di(2-ethylhexyl)phthalate in fish', J. Agric. Food Chem., 21, 1128-1129.
