2662
Anal. Chem. 1984, 56, 2662-2670
trans-l0,ll-dihydrodiol racemate, 92622-79-2; DMBA transl0,11-dihydro-(10R,11R)-diol, 92693-72-6; DMBA trans-l0,11dih~dro-(lOS,llS)-diol,92693-73-7; trans-benz[alanthracene 3,4-dihydro-(3S,4S)-diol, 67335-43-7; trans-7-methylbenz[a]anthracene 8,9-dihydro-(8R,gR)-diol, 88244-40-0.
LITERATURE CITED (1) Dippie, D. In “Chemical Carcinogens”; Searie, C. E., Ed.; American Chemical Society: Washington, DC, 1982 ACS Monograph No. 173 pp 245-314. (2) Yang, S. K.; Chou, M. W.; Roller, P. P. J . Am. Chem. SOC. 1979, 101, 237-239. (3) Chou, M. W.; Yang, S. K. J . Chromtogr. 1979, 785, 635-654. (4) Yang, S.K.; Fu, P. P. Biochem J . 1984, 233, 775-782. (5) Maiaveiile, C.; Bartsch, H.; Tierney, B.; Grover, P. L.; Sims, P. Blochem. Blophys. Res. Commun. 1977, 8 3 , 1468-1473. (6) Siaga, T. J.; Gleason, G. C.; DlGiovanni, J . D.; Sukumaran, K. B; Harvey, R. G. Cancer Res. 1979, 3 9 , 1934-1936. (7) Huberman, E.; Chou, M. W.; Yang, S. K. Proc. Natl. Acad. Sci. U . S.A 1979, 76, 862-866. (8) Wislocki, P. G.; Juliana, M. M.; MacDonald, J. S; Chou, M. W.; Yang, S. K.; Lu, A. Y. H. Carcinogenesis 1981, 2, 51 1-514. (9) Conney, A. H. Cancer Res. 1982, 42, 4875-4917. (10) Yang, S.K.; Geiboin, H. V.; Weber, J. D.; Sankaran, V.; Fisher, D. L.; Engel, J. F. Anal. Blochem. 1977, 7 8 , 520-526. (11) Thakker, D. R.; Levin, W.; Yagi, H.; Turujman, S.; Kapadia, D.; ConneY, A. H.; Jerina, D. M. Chem.-Bioi. Interact. 1979, 278 145-161. (12) Kim Y. H.; Tishbee, A.; GII-Av, E. J . Chem. SOC. Chem. Commun. 1981, 75-76. (13) Weems, H. B.; Yang, S. K. Anal. Biochem. 1982, 725, 156-161. (14) Yang, S. K.; Li, X. C. J . Chromatogr. 1984, 291, 265-273. (15) Fu, P. P.; Yang, S.K. Biochem. Blophys. Res. Commun. 1982, 709, 927-934.
.
(16) Fu, P. P.; Yang, S. K. Carclnogensls 1983, 4 , 979-984. (17) Pirkle, W. H.; House, D. W.; Finn, J. M. J . Chromatogr. 1980, 192, 143-1 56. (18) pirkle, w, H,; Finn, J. M. J . Org, Chem. 1981, 4 6 , 2935-2938, (19) Sukumaran, K. B.; Harvey, R. G. J . Am. Chem SOC. 1979, 101, 1353-1354. (20) Lehr, R. E.; Schaefer-Ridder, M.; Jerina, D. M. J . Org. Chem. 1977, 42, 736-744. (21) Zacharias, D. E.; Glusker, J. P.; Fu, P. P.; Harvey, R. G. Cancer Res. 1975, 37, 775-782. (22) Tierney, B.; Hewer, A.; MacNicoli, A. D.; Gervasi, P. G Rattle, H.; Walsh, C.; Grover, P. L.; Sims, P. Chem.-Blol. Interact. 1978, 2 3 , 243-257. (23) Tierney, B.; Abercrombie, B.; Walsh, C.; Hewer, A.; Grover, P. L.; Sims, P. Chem.-Bioi. Interact. 1978, 21 289-298. (24) Yang, S. K.; Fu, P. P. Chem.-Blol. Interact. 1984, 49, 71-88. (25) Yang, S. K. Drug Metab. Dispos. 1982, 10, 205-211.
RECEIVED for review June 6,1984. Accepted August 16,1984. This work was supported by U.S. Public Health Service Grant CA29133 and Uniformed Services University of the Health Sciences Protocol No. R07502. The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the Department of Defense or the Uniformed Services University of the He&, Sciences. The experiments reported herein were conducted according to the principles set forth in the “Guide for the Care and Use of Laboratory Animals”, Institute of
Resources,
Research
DHEW,
Publication No. (NIH) 78-23.
Separation Strategy of Multicomponent Mixtures by Liquid Chromatography with a Single Stationary Phase and a Limited Number of Mobile Phase Solvents Marina De Smet, Guido Hoogewijs, Marc Puttemans, and D. L. Massart* Farmaceutisch Instituut, Vrije Universiteit Brussel, Laarbeeklaan 103, B-1090 Brussel, Belgium
A methodology for the development of separatlon procedures for dlfferent groups of substances (mostly drugs) Is descrlbed. Chromatography Is performed on a cyanopropyl column wlth a llmlted number of moblle phase solvents to carry out normal-phase as well as reversed-phase HPLC. Inltlally, a gradlent elutlon with two solvents Is performed from which a starting lsocratlc mobile phase composltlon Is selected to elute the substances In a sultable capaclty factor range. Other binary and ternary solvent compositions, determined wlth an experlmental deslgn and having the same solvent strength but dlfferent selectivity, are tested to adJust the resolution of peak pairs and to Improve the separatlon between the solutes. Benramldes, local anesthetlcs, phenothlarlnes, trlcycllc antldepressants, benrodlareplnes, cortlcosterolds, sutfonamldes, barbiturates, and food preservatlves are separated elther In normal phase or In reversed phase or In both.
In the last few years, pioneering work on the optimization
of HPLC separations has been carried out by Glajch, Kirkland, Snyder, and co-workers (1-6). We will call in this paper their collection of works the GKS strategy. An overview of the resulting methodology can be found in an article by Lehrer
(7). In short, the philosophy of the GKS strategy is the following: A limited number of solvents (a total of eight) and stationary phases (two) are sufficient to achieve the great majority of all HPLC separations. The optimal constitution of the mobile phase can be found by using a formal optimization algorithm, which consists in determining first a suitable solvent strength and then in optimizing the solvent selectivity. In our laboratory we have worked along the same lines on more limited sets of substances. Using information theory, we found that nearly every separation of pharmaceutically important bases can be achieved with a single stationary phase (i.e., a cyanopropyl bonded phase) and one of two preferred mobile phases, namely, n-hexane-dichloromethane-acetonitrile-propylamine (50/50/25/0.1) and acetonitrile-waterpropylamine (90/ 10/0.01) (8). Optimization is then carried out with one of the two systems as a starting point and consists in fine tuning the volume ratio of the eluent components (9). Because of the success of our strategy for basic drugs with a single column, we thought that it could be possible to develop a general strategy along the same lines as the GKS strategy, but with the cyanopropyl bonded phase as single phase instead of the two different columns. The CN bonded column is of intermediate polarity and can be used in the reversed-phase mode as well as in the normal mode. The GKS methodology
0 1984 American Chemical Society 0003-2700/84/0358-2662~01.50/0
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
uses two stationary phase types, namely, one for normal-phase and one for reversed-phase chromatography. If it were possible to use a single stationary phase type, this would be an important advantage since the selection of one of two stationary phases and the replacement of one of these by the other are the only nonautomated step in the GKS methodology. In particular, since only one stationary phase is used, one requires knowledge about retention behavior only about that single phase. The elimination of the other stationary phases makes therefore the selection easier. The use of a single stationary phase opens the way for fully automated selection and optimization of HPLC separations. Another simplification compared with the original GKS strategy is that we use only six solvents (instead of eight) and consider only binary and ternary solvent mixtures and not quaternary eluent compositions. This simplification is based on the wish to work with a medium-priced HPLC apparatus using the customary three solvent reservoirs. Application of the complete GKS strategy requires a special apparatus with a four-solvent delivery system. The six solvents are water as the solvent strength adjusting solvent in the reversed-phase mode, nhexane as the solvent strength adjusting solvent in normalphase separations, and methanol, acetonitrile, tetrahydrofuran, and dichloromethane as solvent selectivity adjusting solvents. The GKS strategy consists of three steps. The first step is the selection of an acceptable solvent strength to elute the solutes in a suitable capacity factor range. The second step consists of carrying out an experimental design method called Simplex, The third step is a fine tuning of the eluent composition to obtain the best separation, using a method of data analysis called overlapping resolution mapping (ORM). It is the aim of the present paper to demonstrate that acceptable separations of drugs in general can be obtained with a single stationary phase and only six mobile phase solvents. We use only the two first steps, step 2 using the optimization triangle as described in ref 7 . If one wants to obtain optimal instead of merely acceptable separations, the third step is however necessary. We therefore undertook to develop separations using the simplified strategy for groups of basic substances, acidic substances, and neutral substances of different polarities so that normal- and reversed-phase separations would be needed. These groups consist of seven local anesthetics (basic substances), five benzamides (basic substances), six phenothiazines (basic substances), eight tricyclic antidepressants and benzodiazepines (basic substances), six corticosteroids (neutral substances), six sulfonamides (amphoteric substances considered here as acids), seven barbiturates (acidic substances), and seven food preservatives (acidic substances).
2663
Table I. Solutes concn of group
substances
stds, wg/mL
benzamides
alizapride tiapride sulpiride sultopride metoclopramide
10 10 10 10 10
local anesthetics
amylocaine benzocaine procaine tetracaine pramoxine mepivacaine lidocaine
20 2 10 10
corticosteroids
phenothiazines
sulfonamides
barbiturates
food preservatives
150 50 50
prednisone prednisolone betamethasone prednisonacetate prednisolonacetate betamethasonacetate
2 2
alimemazine triflupromazine levomepromazine chlorpromazine promethazine promazine
2
tricyclic antidepressants amitryptiline and benzodiazepines nortryptiline imipramine chlorimipramine desipramine diazepam medazepam nitrazepam
EXPERIMENTAL SECTION Apparatus. A Varian 5060 liquid chromatograph, equipped with a Valco loop injector (sample loop 100 1L) and a UV detector with fixed wavelength (254 nm), was used. Chromatograms were recorded on a Varian 9176 recorder and retention times were recorded with a Varian CDS 401 chromatographic data system. The columns were 250 X 4 mm i.d. stainless steel columns packed with Lichrosorb CN with a particle size of either 10 wm or 5 wm. The flow rate was 1 mL/min or 2 mL/min. The detector attenuation was 0.08 aufs. All experiments were carried out at ambient temperature. Reagents. All drugs were of pharmacopeia1 or equivalent purity. A survey of the drugs is given in Table I. The solutes were dissolved in a solvent mixture, containing equal volumes of the solvents used in the gradient elution run (see section on the simplified strategy), Le., in reversed-phase methanol-water (5050) and in the normal-phase n-hexane-dichloromethane (5050). n-Hexane, dichloromethane, tetrahydrofuran, acetonitrile, and methanol were all liquid chromatographic grade. Double distilled water which was further purified using a Water-I-system(Gelman Sciences) was used. n-Hexane, dichloromethane, tetrahydrofuran,
0
2 2
2 2 2
2 2 2 2
1 2 1 1
2 1 1 2
sulfadiazine methoxysulfadiazine sulfadimethoxine sulfadimidine sulfamerazine isosulfamerazine
2
secobarbital phenobarbital pentobarbital methylphenobarbital butobarbital barbital diphenylhydantoine
100
sorbic acid benzoic acid methyl p-hydroxybenzoate ethyl p-hydroxybenzoate propyl p-hydroxybenzoate 3-hydroxybenzoicacid 4-hydroxybenzoic acid
2 2
2 2
2 10 100 10 100 100
15
0.3 10 1 1 1
15 2.5 ~.
and propylamine were purchased from Fluka A.G. (Buchs, Switzerland). Acetonitrile, methanol, and acetic acid were obtained from E. Merck (Darmstadt, G.F.R.). The shifting from normal phase to reversed phase is done by carrying out a gradient elution starting from 100% dichloromethane to 100% methanol (or acetonitrile) in 30 min. This latter mobile phase is held for 15 min. Then a second gradient elution is carried out to the desired methanol-water composition,which is maintained for 1 h to obtain constant retention times.
RESULTS AND DISCUSSION The Simplified Strategy. Initial Selection of Conditions. Two criteria determine the initial chromatographic conditions, namely:
2664
ANALYTICAL CHEMISTRY, VOL. 58, NO. 14, DECEMBER 1984
The polarity: apolar substances are usually separated by normal phase chromatography and polar substances by reversed-phase chromatography. In some instances, it may be advantageous to try out both possibilities (see Applications, Phenothiazines). Acid-base characteristics: for acidic substances, one adds 1% acetic acid to all eluting agents. For basic substances, one adds propylamine. In normal-phase chromatography a concentration of 0.1% is applied, but in reversed phase chromatography such a quantity has deleterious effects on the column packing so that only 0.01% is added. Step 1. To determine the solvent strength at which convenient capacity factors are obtained, a gradient elution is carried out in normal phase starting with n-hexane and going to dichloromethane in 10 min and, in reversed phase starting with methanol-water (50:50) to methanol in 20 min. One determines the solvent composition ci at which each solute i leaves the column with the following equation: c, = initial concentration of solvent selectivity adjusting solvent (i.e., 50% in reversed phase, 0% in normal phase) [(time of elution - distance between gradient generator and column inlet in time units) X gradient rate in % of solvent selectivity adjusting solvent per minute]. Step 2. In this step one carries out a t the same solvent strength, elutions with mobile phases with different solvent selectivity as described in the GKS strategy. This requires the following: The selection of a first isocratic composition from the gradient elution: in our simplified version one determines the geometric mean of the solvent compositions c, (i = 1, ..., n) at which the solutes i leave the column.
+
As the value of c, calculated in eq 1resulted in most cases in a poor retention of the solutes, the solvent strength of the first isocratic composition was lowered by multiplication of c by a factor 3/4. This value was obtained experimentally co = (3/4)c
(2)
co is then the concentration in volume percent of the solvent selectivity adjusting solvent in the first isocratic composition. The selection of two additional binary mixtures with the same solvent strength. The first isocratic composition is a binary mixture of water and methanol (reversed phase) or n-hexane and dichloromethane (normal phase). One now selects two additional binary mixtures between the solvent strength determining solvents (water in reversed phase, nhexane in normal phase) and the two remaining solvent selectivity adjusting solvents, acetonitrile and tetrahydrofuran, with the same solvent strength as the first isocratic composition. This is determined by using the equation
s = CS&
(3)
where S is the solvent strength of the mobile phase, s, is the solvent strength weighing factor of solvents i, and +, is the volume fraction of solvent i. The values of s, for the different solvents of interest are given in ref 7. In the normal phase procedure the mobile phase composition with n-hexane and acetonitrile js only calculated theoretically. The elution with this mobile phase is not carried out since these two golvents are not miscible over the entire range of mixing ratios, but in the presence of tetrahydrofuran or dichloromethane complete miscibility is obtained with the theoretical n-hexaneacetonitrile mixture (see ternary mobile phase compositions). Selection of three ternary mixtures with the same solvent strength. These are obtained by mixing the three binary mixtures two by two in 1:l proportions.
This yields a total of six eluting agents for reversed-phase chromatography and five for normal-phase chromatography. Chromatograms are obtained with these solvent mixtures, and the best one is chosen. Comparison with the GKS Strategy. In the GKS strategy two stationary phase types (i.e., one for reversed phase and one for normal phase) and eight solvents are used to carry out the separation of multicomponent mixtures, while in the simplified strategy only the cyanopropyl column and six solvents are needed. In both strategies the selection of the solvents is founded on Snyder’s solvent classification scheme ( I O , l I ) , which classified the solvents into eight groups, according to different solvent properties such as polarity, proton donor, proton acceptor, and dipole interacter. The greatest difference in selectivity is obtained by choosing solvents belonging to different classes, i.e., with different properties. In the GKS strategy the solvents used in reversed phase are water, methanol, acetonitrile, and tetrahydrofuran, while in normal phase n-hexane, dichloromethane, chloroform, and diethyl ether are the mobile phase solvents. In the simplified strategy the solvents are water, methanol, acetonitrile, tetrahydrofuran, dichloromethane, and n-hexane. The first four mentioned are used in reversed phase, the last four in normal phase. The determination of the first isocratic mobile phase composition with a suitable solvent strength is different in both strategies. In the GKS strategy the mobile phase with suitable solvent strength is selected rather by trial and error. In the simplified strategy a linear gradient elution is carried out as was first proposed by Lehrer (7) and implemented in the DuPont Sentinel system. Equations 1 and 2 are then applied to determine the first isocratic mobile phase.
Applications of the Simplified Strategy for the Separation of Multicomponent Mixtures. Local Anesthetics. These basic substances are separated in the reversed-phase mode (in the presence of 0.01% propylamine) and the chromatograms resulting from the complete optimization procedure are shown in Figures 1to 7. From the gradient run the first isocratic mobile phase composition is determined. It consists of methanol-water (4555). The two other binary mobile phase compositions with the same solvent strength are acetonitrile-water (36:64) and tetrahydrofuran-water (28:72). The ternary mobile phase solvents are obtaiped by mixing two of the three binary mixtures in 1:l proportions. The three ternary eluting agents are methanol-acetonitrile-water (22:18:60), methanol-tetrahydrofuran-water (22:14:64),and acetonitrile-tetrahydrofuran-water (18:14:68). The reversed-phase gradient separates completely four of the drugs but pramoxine, lidocaine, and mepivacaine are not resolved. The first isocratic composition yields a somewhat better result since only lidocaine and mepivacaine remain completely unresolved. Binary and ternary solvent mixtures containing tetrahydrofuran permit resolution of this peak pair and yield, for instance, the very acceptable result of Figure 7. The results of the optimization scheme are given in Table 11. Benzamides. These basic substances are investigated with the reversed-phase system in the presence of propylamine. The gradient elution separates the five substances but tiapride and sulpiride are only partially resolved. The results obtained with the optimization scheme &re given in Table 111. The best result is obtained with acetonitrile-water-propylamine (49:51:0.01). Nearly complete resolution of all substances is achieved (Figure 8). Phenothiazines. The separation of these six substances is carried out in the reversed-phase as well as in the normalphase mode with addition of propylamine. The reversed-phase gradient yields only four peaks because triflupromazine and promethazine on the one hand and levomepromazine and
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
2665
Table 11. k’Values of Local Anesthetics
mobile phase compositions: % vol CHBOH CHBCN THF HzO 45
0
0 0 22
36
0 0
0
28 0
22
18 0
0
18
55 64 72 60 64 68
14 14
k’values of the solutesb 3 4 5
1
2
1.9 2.0 3.2 1.9
2.7 2.9 3.7 2.7 2.8 2.9
2.2
2.3
2.7 2.9 4.7 2.7 2.9 3.2
3.8 4.0 6.0 3.8 3.9 4.1
3.4 3.4 6.5 3.2 3.9 3.8
6
7
9.2 8.4
10.3 10.0 15.5 9.3 10.5 9.9
8.2
8.7 7.8 7.2
“Propylamine (0.01%) is added to all the mobile phases. bSolutes: (1) benzocaine, (2) mepivacaine, (3) lidocaine, (4) pramoxine, (5) amylocaine, (6) procaine, (7) tetracaine. 2.3
Table 111. k’Values of the Benzamides
mobile phase compositions,” % vol CH30H CH3CN THF HzO 60
0
0 0
49
30 30
25
0
25
40 51 66 45 53 58
0 0
34 0 17 17
0 0
1
k’ values of solutesb 2 3 4 5
0.2 6.9 7.9 0.2 6.3 7.5 0.1 5.1 5.3 0.1 5.8 6.8 0.2 4.8 5.3 0.1 4.8 5.3
9.5 12.1 9.4 12.8 6.3 12.5 8.2 10.9 6.4 9.7 6.6 10.8
“Propylamine (0.01%) is added to all the mobile phases. Solutes: (1)alizapride; (2) sulpiride; (3) tiapride; (4) sultopride; (5) metoclopramide.
15 rnin
10
5
0
Flgure 2. Chromatogram of the separatidn of local anesthetics on a Lichrosorb CN (10 Hm) column by isocratic elution: mobile phase, methanol-water-propylamlne (45:55:0.01); flow rate, 1 mL/min; detection sensitivity, 0.08 aufs. 7
Table IV. k’Values of Phenothiazines in the Reversed-Phase Mode
1
5
mobile phase compositions: k’ values of the solutesb % vol CH,OH CH&N THF HzO 1 2 3 4 5 6
6
70
0
0 0
57
0 0 41 0 21 21
0 28 0 28
35 35 0
30 43 59 37 44 51
4.9 5.3 5.5 4.6 4.5 4.5
5.4 5.7 5.4 5.1 4.5 4.6
5.5 6.4 6.1 5.6 5.5 5.1
5.4 7.4 10.1 7.9 10.1 12.5 8.6 9.2 11.3 6.2 7.9 9.8 6.6 8.0 10.2 6.8 7.9 9.7
‘Propylamine (0.01%) is added to all the mobile phases. Solutes: (1) promethazine; (2) levomepromazine;(3) alimemazine; (4) triflupromazine; (5) chlorpromazine: (6) promazine.
0
5
10
15 rnin
Flgure 1. Chromatogram showing the separation of a mixture of local anesthetic on a Lichrosorb CN (10 km) column by gradient elution, methanol-propylamine methanol-water-propylamine (50:50:0.01) (1OO:O.Ol) in 20 min: (1) benzocaine, (2) mepivacaine, (3) lidocaine, (4) pramoxine, (5) amylocaine, (6) tetracaine, (7) procaine; flow rate, 1 mL/min; detection sensitivity, 0.08 aufs. +
alimemazine on the other hand are not separated. The optimization procedure (Table IV) yields an acceptable separation with acetonitrile-water and methanol-acetonitrilewater mixtures. T h e latter is shown in Figure 9. The normal-phase gradient does separate all the solutes but not completely. The first isocratic composition with n-hexane
Table V. k’Values of the Phenothiazines in the Normal Phase Mode
mobile phase compositions: % vol CHzCl THF CH&N n-Hex 26
0
0
20 10
0 0 0
80
0 10
7 7
80
13 13 0
74 77 83
k’ values of the solutesb 1 2 3 4 5 6 1.0 1.3 1.6 0.5 0.7 0.7 0.6 0.8 0.9 0.9 0.9 0.9 0.7 0.8 0.9
2.3 3.2 4.3 1.0 1.5 1.9 1.2 1.8 2.3 0.9 1.0 1.2 1.1 1.3 1.5
“Propylamine (0.1%) is added to all the mobile phases. Solutes: (1) alimemazine; (2) levomepromazine; (3) triflupromazine; (4) chlorpromazine; (5) promethazine; (6) promazine.
2666
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
2.3
2.3
5
0 0
5
15 min
10
15 rncn
10
on a Lichrosorb CN (10 pm) column: mobile phase, methanol-acetonitrile-water-propylamine (22: 18:60:0.0 1); flow rate, 1 mL/min; detection sensitivity, 0.08 aufs. Flgure 5. Chromatogram of the separation of local anesthetics
Figure 3. Chromatogram of the separatlon of local anesthetics on a Lichrosorb CN (10 pm) column: mobile phase, acetonitrile-waterpropylamine (36:64:0.01); flow rate, 1 mL/min; detection sensitivity, 0.08 aufs.
I
7
1
T
-
0
- 5
10
~
15
20
25
min
Flgure 4. Chromatogram of the separation of local anesthetlcs on a
Lichrosorb CN (10 pm) column: mobile phase, tetrahydrofuranwater-propylamine (28:72:0.01); flow rate, 1 mL/tqin; detection sensltivity, 0.08 aufs.
and dichloromethane, however, yields an acceptable separation (Figure 10). The further optimization (Table V) does not improve on this. Corticosteroids. The gradient elution in the reversed-phase mode yields two weakly resolved peaks, Le., an extremely bad
0
5
-
10
~~
7
15
2 0 rnin
on a Lichrosorb CN (10 pm) column: mobile phase, methanol-tetrahydrofuran-water-propylamine (22:14:64:0.01); flow rate, 1 rnL/min; detection sensitivity, 0.08 aufs. Flgure 6. Chromatogram of the separation of local anesthetics
separation. Application of the optimization procedure (Table VI) yields a much better result for the mixture of acetonitrile-water (26:74) (Figure 11). Tricyclic Antidepressants and Benzodiazepines. The gradient elution is performed according to the normal-phase procedure after addition of 0.1% propylamine. The gradient
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
2667
1
J _3
0 10 20 MIN Flgure 8. Chromatogram of the separation of benzamldes on a Li-
’
chrosorb CN (IO pm) column: mobile phase, acetonitrile-waterpropylamine (49:51:0.01); (1) alizapride, (2) sulpiride, (3) tlaplde, (4) sultopride, (5) metoclopramide; flow rate, 1 mL/min; detection sensitivity, 0.08 aufs.
0 5 10 15 rnin Flgure 7. Chromatogram of the separation of local anesthetics on a Lichrosorb CN (IO pm) column: moblle phase, acetonitrile-tetrahydrofuran-water-propylamine (18:14:68:0.01); flow rate, 1 mL/min; detection sensitivity, 0.08 aufs.
L
Table VI. k’Values of Corticosteroids
mobile phase compositions, % vol CHBOH CH&N THF HzO 32
0
0 0 16
26
0 0
0
18
13
0
16
0
0
13
9 9
68 74 82 71 75 78
k’ values of the solutes0 1 2 3 4 5 6 2.7 3.1 2.2 2.5 3.9 4.1 2.6 2.9 2.6 2.8 2.6 2.8
3.4 2.9 6.0 3.8 3.4 3.5
3.9 3.4 7.5 3.2 4.1 4.3
4.8 4.9 4.1 4.6 8.6 11.7 4.5 4.7 4.8 5.5 5.0 6.0
Solutes: (1) prednisolone; (2) prednisone; (3) betamethasone; (4) prednisolonacetate; (5) prednisonacetate; (6) betamethasonacetate.
I
5 10 15 min Flgure 9. Chromatogram of the separation of phenothiazines in re-
0
elution and resulting isocratic f i s t composition do n o t separate a m i t r y p t i l i n e f r o m chlorimipramine n o r n o r t r y p t i l i n e f r o m nitrazepam. T h e optimization procedure (Table VII) permits achieving separation o f t h e latter t w o but n o t o f t h e former. T h i s is achieved w i t h dichloromethane-acetonitrile-n-hexane-propylamine (30:16:54:0.1) (Figure 12). T o achieve separation o f a m i t r y p t i l i n e f r o m chloripramine, one w o u l d have t o use a lower solvent strength, which would, however, also result in greater It’ values for a l l the solutes o f this group.
versed phase on a Lichrosorb CN (IO pm) column: mobile phase, methanol-acetonitrile-water-propylamine (35:28:37:0.01); (1) promethazine, (2)levomepromazine, (3) alimemazine, (4) trifluopromazine, (5) chlorpromazine, (6) promazine; flow rate, 1 mL/min; detection sensitivity, 0.08 aufs.
Sulfonamides. These amphoteric substances are separated by elution in the normal-phase mode after addition o f acetic acid. T h e gradient yields an acceptable separation but t h e
Table VII. k’Values of Tricyclic Antidepressants and Benzodiazepines
mobile phase compositions; % vol CHzClz THF CH3CN n-Hex 60
0
0
46
0 0
0
16
30 30 0
23 23
0 16
40 54 54 46 61
k’ values of the solutesb 1
2
3
4
5
6
7
8
0.3 0.3
0.7 0.6 0.4
0.9
0.9 0.4
1.3
6.6
0.4
0.6
1.0
0.6 0.6 0.5
0.5
0.7
0.5 0.5
1.6
0.8
1.1
0.5
1.0
6.8 3.9 4.5 4.9 3.1
10.3 6.6 6.9 7.9 5.1
0.3
0.3 0.5
0.5 0.6
Propylamine (0.1%) is added to all the mobile phases. *Solutes: (1) medazepam; (2) diazepam; (3) amitryptiline; (4) chloorimipramine; (5) imipramine; (6) nitrazepam; (7) nortryptiline; (8) desipramine.
2668
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
I
I
0
I
I
I
.
l
l
I
8 rnin
6
L
2
Flgure 10. Chromatogram of the Separation of phenothiazine in normal phase on a Lichrosorb CN (10 pm) column: mobile phase, n-hexane-dichloromethane-propylamine (74:26:0.1); (1) alimemazine, (2) levomepromazine, (3) triflupromazine, (4) chlorpromazine, (5) promethazine, (6) promazine; flow rate, 2 mL/min; detection sensitivity, 0.08 aufs.
0 2 1 6 a 10 12 min Flgure 12. Chromatogram of the separation of tricyclic antidepressants and benzodiazepines on a Lichrosorb CN (10 pm) column: mobile phase, n -hexane-dichloromethane-acetonitrile-propylamine (54:30:16:0.1); (1) medazepam, (2) diazepam, (3) amitryptiline, (4) chlorimipramine, (5) imipramlne, (6) nitrazepam, (7) nortryptlllne, (8) desipramine: flow rate, 2 mL/min; detection sensitivity, 0.08 aufs.
2 2
0
0
2
L
6
8
10 rnin
Flgure 11. Chromatogram of the separation of corticosteroids on a
Lichrosorb CN (10 pm) column: mobile phase, acetonitrile-water (26:74): (1) prednisolone, (2) prednlsone, (3) betamethasone, (4) prednisolonacetate, (5) prednisonacetate, (6) betamethasonacetate: flow rate, 1 mllmin; detection sensitivity, 0.08 aufs.
isocratic elution with the first isocratic composition is to be preferred (Figure 13). The other binary and ternary mobile phase compositions do not yield better separtions (Table VIII). Barbiturates. These substances are separated in the normal-phase mode and acetic acid is added to all the mobile phase compositions in view of their acidic properties. Both this gradient and the first isocratic composition permit good separation of these substances. The latter is given in Figure 14. The use of tetrahydrofuran and acetonitrile in the mobile phases does not permit improvement in the separation of these solutes (Table IX).
15 rnin
10
5
Figure 13. Chromatogram of the separation of sulfonamide on Lichrosorb CN (5 pm) column: mobile phase, n-hexane-dichloromethane-acetic acM (50:50:0.1); (1) sulfadimethoxine, (2) sulfadimidine, (3) sulfamerazlne, (4) Isosulfamerazine, (5) sulfadiazine, (6) methoxysulfadiazine: flow rate, 2 mLlmin: detection sensitivity, 0.08 aufs.
Table VIII. k’Values of Sulfonamides
mobile phase compositions,” % vol CH2CI, THF CH&N n-Hex 50
0
0
38 19
25 25 0
0
19
50
0 0 0
62
13 13
62 68
56
k’ values of the solutesb 2 3 4 5 6
1
3.9 5.8 6.6 6.9 1.5 2.5 2.9 3.2 1.5 2.1 2.4 2.5 1.5 1.6 2.1 2.2 2.1 2.3 2.7 2.7
7.4 8.2 3.2 3.5 2.7 2.8 2.6 2.6
3.0 3.2
“Acetic acid (1%) is added to all the mobile phases. bSolutes: (1)sulfadimethoxine; (2) sulfadimidine; (3) sulfamerazine; (4)isosulfamerazine; ( 5 ) sulfadiazine; (6)methoxysulfadiazine.
Food Preservatives. The food preservatives are acidic substances so that 1% acetic acid is added to all solvents. In
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984 2669 Table IX. k’Values of Barbiturates mobile phase compositions,“% vol CHzClz THF CH&N n-Hex 47
0
0
36 18
23 23
0
18
0
53 64 59 65 70
0 0 0 12 12
k’ values of solutesb 4 5
1
2
3
1.2
2.7 0.4
1.4 0.2 0.2 0.9
0.2 0.2 0.7 0.5
0.4 1.3 0.9
0.5
0.8
0.3 0.3 0.7 0.7
1.5 0.3 0.3 0.8 0.5
6
7
2.1
4.4 0.7
0.3 0.3
0.7
1.1
1.5
0.7
1.2
“Acetic acid (1%)is added to all the mobile phases. bSolutes: (1) secobarbital; (2) phenobarbital; (3) pentobarbital; (4) methylphenobarbital; (5) butobarbital; (6) barbital; (7) diphenylhydantoine. Table X. k’Values of Food Preservatives in the Reversed-Phase Mode mobile phase compositions”% vol CH30H CHaCN THF 34
0
0 0
28
0 0 20 0 10 10
0 14 0
17 17 0
14
k’values of solutesb 4 5
HZO
1
2
3
66
1.2 1.1
1.2 1.1
1.5 1.3
1.5 1.4
1.7
1.8 1.3 1.3 1.3
1.3 1.3 1.8
2.2
2.4
1.4
1.5
1.5
1.7 1.5
1.5 1.8 1.6
72 80
69 73 76
1.3 1.2 1.2
1.4
6
7
1.7 1.5 3.1
1.9 1.8
4.2
1.7
1.9
2.0 1.9
2.3
2.3
“Acetic acid (1%)is added to all the mobile phases. bSolutes: (1)4-hydroxybenzoic acid; (2) 3-hydroxybenzoic acid; (3) sorbic acid; (4) benzoic acid; (5) methyl p-hydroxybenzoate; (6) ethyl p-hydroxybenzoate; (7) propyl p-hydroxybenzoate, 67
1
1
0
2
2
6
8 rnin
Chromatogram of the separation of barbiturate on a Lichrosorb CN (5 pm) column: mobile phase, n-hexane-dichloromethane-acetic acid (53:47:1); (1) secobarbital,(2)phenobarbital, (3) pentobarbital, (4) methylphenobarbital, (5) butobarbital, (6) barbital, (7) diphenylhydantoine; flow rate, 2 mL/min; detectlon sensitivity, 0.08 aufs. Flgure 14.
the first instance, the separation of the solutes is studied in the reversed phase mode. The gradient elution yields only two poorly resolved peaks, i.e., an extremely bad separation (Table X). Isocratically the best separation is obtained with tetrahydrofuran-water (20:80) but sorbic acid, 3-hydroxybenzoic acid, and 4-hydroxybenzoic acid coelute as well as benzoic acid and methyl p-hydroxybenzoate. Since reversed phase does not yield an acceptable separation for this group of substances, the normal-phase system was tested. The result obtained with the gradient elution is shown in Figure 15. A quite acceptable result is obtained so that the procedure could have stopped here. If an isocratic elution is wanted, the first isocratic composition is calculated to be n-hexane-dichloromethane (6040). This is successful in separating all the solutes except sorbic acid and benzoic acid (Figure 16). The k’values of these substances are too small and since the addition of
0
5
10
15 rnin
Chromatogram of the separation of food preservatives on a Lichrosorb CN (5 pm) column by gradient elution: n-hexane-acetic acid (1OO:l) dichloromethane-acetic acid (1OO:l) in 10 min; (1) sorbic acid, (2)benzoic acid, (3) propyl p -hydroxybenzoate, (4) ethyl p-hydroxybenzoate, (5) methyl p -hydroxybenzoate, (6) 4-hydroxybenzoic acid, (7)3-hydroxybenzoicacid; flow rate, 2 mL/min; detection sensitivity, 0.08 aufs. Flgure 15.
-+
tetrahydrofuran and acetonitrile (Table XI) does not improve the separation, further optimization for this peak pair should be carried out a t a lower solvent strength.
CONCLUSIONS As a conclusion, we may state that the simplified strategy with a single stationary phase and a reduced set of eluting
2670 Table
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
XI.
k'Values
of Food Preservatives in the Normal-Phase M o d e
mobile phase compositions:
CHzClz
THF
40
0 30 15 0 15
0
20 20 0
% vol
k'
CHBCN
n-Hex
0
60 70 65 70 75
,O
0 10 10
1
2
0.3
0.3 0.1 0.1 0.3 0.3
0.1
0.1 0.3 0.3
3
values of solutes
4
5
6
7
3.2
3.4
0.2 0.2 0.6 0.5
3.8 0.2 0.2 0.7 0.5
5.5
0.1 0.1
0.4 0.4 1.2
6.3 0.3 0.3 1.3 0.7
0.5 0.4
0.8
Snyder (IZ),when applying normal phase chromatography. Snyder suggests a displacement mechanism for solute retention in normal-phase chromatography with amino columns (IZ),so that solvent strength does not vary linearly with the concentration of polar modifiers. This is also true for CN columns (as can be verified from Tables V, VII, VIII, IX, and XI) and we are now developing an approach to optimizing chromatography on CN columns, which is similar to that proposed by Snyder. Of course, we are aware that the complete GKS is more general and will probably yield solutions in some cases for which our simplified strategy will not. However, there always is a trade-off between, on the one hand, the complexity and therefore the cost of such a strategy and, on the other hand, its generality and its quality. If one takes this into account, the simplified strategy seems to offer a reasonable alternative to the complete version.
ACKNOWLEDGMENT The authors thank A. De Schrijver and M. De Vreese for technical assistance.
LITERATURE CITED
I
b
5
10 min
Flgure 16. Chromatogram of the separation of food preservatives on Lichrosorb CN (5 pm) column by isocratic elution: mobile phase, nhexane-dichloromethane-acetic acid ( 6 0 4 0 1); flow rate, 2 mL/min; detection sensitivity, 0.08 aufs.
a
agents permits the separation of a wide variety of related substances without needing special apparatus. Therefore, this simplified strategy seems a good way of attacking most separation problems of low molecular weight substances. The addition of a fine-tuning optimization step, which was not investigated in the present article, is under study now. It would also be valuable to take into account recent work of
(1) Snyder, L. R.; Kirkland, J. J. "Introduction to Modern Liquid Chromatography", 2nd ed.; Wiley-Interscience: New York, 1979; Chapter 6. (2) Glajch, J. L.; Kirkland, J. J.; Squire, K. M.; Minor, J. M. J . Chromatogr. 1080, 199,57-79. (3) Snyder, L. R.; Glajch, J. L. J . Chromatogr. 1081, 214, 1-19. (4) Glajch, J. L.; Snyder, L. R. J . Chrometogr. 1981, 214, 21-34. (5) Snyder, L. R.; Glajch, J. L.; Kirkland, J. J. J . Chromatogr. 1981, 218, 299-326. (6) Glajch, J. L.; Kirkland, J. J.; Snyder, L. R. J . Chromatogr. 1982, 238, 269-280. (7) Lehrer, Int. Lab. 1981 (Nov-Dec), 76-88. (8) Detaevekier, M. R.; Hoogewljs, G.; Massart, D. L. J. fharm. Biomed. Anal. 1083, 1 , 331-337. (9) Hoogewijs. G.; Massart, D. L. J . Liq. Chromatogr. 1083, 6 , 2521-2543. (10) Snyder, L. R. J . Chromatogr. 1074, 92, 223-230. (11) Snyder, L. R. J . Chromatogr. Sci. 1078, 16, 223-234. (12) Snyder, L. R.; Schunk, T. C. Anal. Chem. 1082, 5 4 , 1764-1772.
RECEIVED for review November 23, 1983. Resubmitted July 26, 1984. Accepted September 4, 1984.
