Use of calixarenes to modify selectivities in capillary electrophoresis

Sensitive determination of inorganic anions at trace levels in samples of snow water from sierra nevada (Granada, Spain) by capillary ion electrophore...
1 downloads 0 Views 373KB Size
Correspondence Anal. Chem. 1994,66,747-750

Use of Calixarenes To Modify Selectivities in Capillary Electrophoresis Dvlra Shohat and Eli Grushka’ Department of Inorganic and Analytical Chemistry, The Hebrew University, Jerusalem, Israel

Calixarenes are macrocyclic oligomers having the shape of a conical vase. Its inner cavity can accommodate various guest molecules, in similarity to other macrocyclic molecules such as crown ethers and cyclodextrins. Thus, calixarenes can be used to manipulate selectivities in chromatography and in capillary electrophoresis. We report here initial results that demonstrate the value of calixarenes as selectivity modifiers in capillary electrophoresis. Retention behaviors of chlorinated phenols, benzenediols, and toluidines in the presence of psulfonate calix[6]arene was investigated. The separationwas carried out under conditions where the solutes are uncharged and largely unresolved. The addition of the calixarene, which forms complexes with the solutes, facilitated the separation. The influence of the calix(6Jarene concentration was studied in various buffer conditions. The use of additives to alter the selectivity of a separation method is well-known. In chromatography such additives, due to their interactions with the solutes to be separated, change the distribution of the solutes between the stationary and the mobile phases. In electrophoresis, the interactions with the additives change the effective mobilities of the solutes. In both cases, these changes affect the selectivity of the separation system. Typical examples of additives, used both in chromatography and in electrophoresis, are crown ethers1 and cyclodextrins.2 Both these classes of additives are macrocyclic molecules having cavities which are available to host-guest type interactions with the solutes. In the present work, we report on another class of macrocyclic additives, the calixa r e n e ~which , ~ are also capable of host-guest interactions with appropriate solute molecules. While calixarenes can be used to change selectivities in several separation techniques, we will discuss in this work their use only in capillary electrophoresis. Subsequent publications will be devoted to the use of calixarenes in chromatography. . . The are macrocyclic molecules made up Of phenolic units meta linked by methylene bridges. They possess basket-shaped cavities. They have been named “calixarenes” by Gutsche3v4because of the resemblance of the four-membered ring to a (in Greek’ For specifying the

so3 Figure 1. pSulfonic calix[b]arene.

macrocyclic ring size, a number, representing the number of repeating phenolic units, is inserted between the “x” and “a” in calixarene. The calixarenes are particularly stable substances. The common calixarenes, having phenolic derivatives such as p-methyl, p-tert-butyl, and p-phenyl as the building blocks, are insoluble in water and in most of the common organic solvents. However, substitution of the R group with a sulfonate group gives a water-soluble calixarene; see Figure 1. Much of the interest in the calixarenes derives from their promise as useful selective complexation agents. This selectivity is due to the presence of the cavity, as well as to the outer functional groups. The complexation ability516of calixarenes made from phenols in which the para R groups are ethers, esters: amines, or carboxyls or resorcinol-deri~ed~*~ calixarenes was examined in nonaqueous solvents. The synthesis of the sulfonatocalixarenes by Shinkai’O enabled complexation of ions11-14 as well as organic molecules in aqueous solutions. To study the potentialof calixarenes as selectivity modifiers in capillary electrophoresis, we have examined the electro-

~

(1) Kuhn, R.; Stoeklin, F.; Erni F. Chromatographfa 1992, 33, 32. (2) Terabe, S.;Ozaki, H.;Otsuka, K.; Ando, T. J . Chromatogr. 1985,332,211. (3) Gutsche, C. D. CaNxarenes; R. Society of Chemistry: London, 1989. (4) Gutsche, C. D.; Bauer, L. J. J. Am. Chem. Soc. 1985, 107, 6052.

QQQ3-270Q194/Q36&Q747$04.5Q/ 0 0 1994 Amerlcan Chemical Society

J.; Gutsche, C. D. J. Am. Chem. Soc. 1985, 107, 6063. (6) Olmstead, M. M.; Sigel, G.; Hope,H.;Xu, X.; Power, P. P. J. Am. Chem. Soc. 1985, 107, 8087. ( 7 ) Gutsche, C . D.; Bauer, L. J. J. Am. Chem. Soc. 1985, 107, 6059. (8) Schneider, H. J.: Kramer, R.: Simova. S.: Schneider. U. J. Am. Chem. Soc. 1988, 110, 6442. (9) Aoyama, Y.;Tanaka, Y.;Toi, H.;Ogoshi, H.J. Am. Chem. Soc. 1988,110, (5) Bauer, L.

63. (10) Shinkai, S.; Mori, S.; Koreishi, H.; Tsubaki, T.; Manabe, 0.J . Am. Chem. SOC.1986, 108, 2409. (11) Shinkai,S.;Mori,S.;Arimura,T.;Manabe,O.J.Chem.Soc.,Chem.Commun. 1987, 238.

Analytical Chemistry, Vol. 66,No. 5, March 1, 1994 747

II

1

I

I

0.0002 AU

WVVVVV%

OOOlAU 3

3

35

4

45

5

55

6

65

7

75

6

Migralion time (mm)

Flgure 2. Electropherogram of a mlxture Containing p, m, and pbenzenedlols and p, m, whlorophenols. Phosphate buffer; pH 7 (0.05 M) without calixarene: (A) 1,&benzenedbl and pchlorophenol; (B) 1,9benzenedlol and 1,2-benzenedlol; (C) whlorophenol and mchlorophenol.

phoretic behavior of several classes of disubstituted benezenes, some of which were uncharged at the conditions of the experiments. Separation of phenolic compounds have been widely studied because they are priority environmental pollutants. For example, Otsuka and TerabelsJ6reported on the separation of phenols in micellar electrokinetic capillary chromatography (MECC) using SDS micellar solution in borate-phosphate buffer. While the separation in MECC is achieved by liquid-liquid partition using the micelles as a pseudostationary phase, the separation described here is enabled by the interaction of the phenolic compounds with the calixarene. In the present research we have investigated the effects of adding 4-sulfonic calix [6]arene to the running buffer on the selectivity in capillary electrophoresis.

EXPERIMENTAL SECTION Instrumentation. The work was performed on a homemade CE unit with a Glassman series EH power supply and a Linear UVIS 200 detector operated at 220 nm. A BarSpec (Rishon-Letzion) Chrom-A-set was used to control the power supply, the injection process, and data collection and analysis. The quartz capillary, purchased from Polymicro Technologies was 50 pm i.d. by 80 cm long with an effective length of 40 cm . Reagents. Electrophoresis was carried out in a phosphate buffer using NaH2P04 (Baker) or Na2HP04 (Merck) adjusted to the desired pH with NaOH, H3PO4, and/or borate (Frotarum). Four different buffer solutions, all having pH 7, were prepared: (a) 0.05 M sodium dihydrogen phosphate adjusted with NaOH, (b) 0.05 M disodium hydrogen phosphate adjusted with phosphoric acid, (c) 0.015 M disodium hydrogen phosphate adjusted with phosphoric acid, and (d) 0.05 M disodium hydrogen phosphate adjusted with 0.1 M borate. Another buffer, at pH 8, was prepared using 0.015 M disodium hydrogen phosphate adjusted with NaOH. The ionic strength was adjusted as required by the experiment. (12) Shinkai, S.; Mori, S.; Araku, K.; Manabe, 0. Bull. Chem. Soc. Jpn. 1987, 60, 3679. (13) Shinkai, S.;Koreishi, H.; Ucda, K.; Arimura, T.; Manabe, 0.J. Chem. SOC.. Chem. Commun. 1986, 233. (14) Shinkai, S.; Koreishi, H.; Ueda, K.; Arimura, T.; Manabe, 0.J. Am. Chem. SOC.1987, 109, 6371. ( I S ) Otsuka, K.; Terabe, S.; Ando, T. J. Chromatogr. 1985, 318,39. (16) Terabe, S.; Otsuka, K.; Ichikawa, K.; Tsuchiya, A.; Ando, T. Anal. Chem 1984, 56, 113.

740

Ana&ti&Chemistry,

Voi. 66, No. 5, March 1, 1994

32

I,

34

36

38

114

42

44

46

/

M i p t i o n lime (min)

Flgure 3. Electropherogram of a mixture contalnlng p, m, and ebenzenedlol and p, m, and o-chlorophenols. Phosphate buffer; pH 7 (0.05 M) withpsulfonlc callx[6]arene0.004 M: (A) l,&benrenedlol; (B) pchlorophenol; (C) 1,bbenzenedbl; (D) 1,2-benzenedid; (E) mchlorophenol, (F) o-chlorophenol. The shoulders apparent on some of the peaks are due to detector signal noise.

For test solutes we used m-chlorophenol (Aldrich) 0- and p-chlorophenols (Fluka), 1,4-, 1,3-, and 1,2-benzenediols (Aldrich), and p-, m-, and o-toluidines (Merck). In the separation of the toluidines, 0.01 M triethylamine (TEA) (Fluka) was added to the buffer to minimize wall adsorption and to improve resolution. 4-Sulfonic calix [6]arene was purchased from Janssen Chemica.

RESULTS AND DISCUSSION The chlorophenols and the benzenediols are largely uncharged at the pH range of the experiments. Therefore, their separation by conventional CE should be difficult. Indeed, in the absence of the calixarene, test mixtures containing either all chlorophenols or all benzenediols gave two broad peaks as the para and meta isomers coeluted in all cases. Figure 2 shows an electropherogram containing a mixture of all six chlorophenols and benzenediols run at pH 7. Without the calixarene, this mixture could only be resolved partially. The toluidines, under the same conditions, could not be resolved at all. It should be noted that, at pHs 9 and above, the chlorophenols as well as the benzenediols are completely or partially ionized and thus they can be separated by conventional electroph~resis.~~ Similarly, at acidic pHs, the toluidines can be easily separated. However, we intentionally chose to work in a pH range where little or no separations took place since we wanted to compare migration behaviors in the absence and presence of the calixarene. The rational for choosing the above solutes as test compounds is threefold: (a) All the solutes, being aromatic, have good molar absoptivities for easy detection above the background of the calixarene. (b) The isomers of each class of compounds allow examination of the effects of molecular shape. (c) The electrophoretic behavior of the solutes are well-known, in the absence of a calixarene, over a limited range of pH. The addition of p-sulfonic calix[6]arene to the buffer solutions led to full separations of all three isomers for each class of compounds studied. In fact, when the phosphate concentration in the buffer was 0.05 M, at pH 7, we could separate completely all six solutes in a mixture containing all chlorophenols and benzenediols; see, for example, Figure 3. (17) Kaneta, T.; Tanaka, S.; Yoshida, H. J . Chromatogr. 1991, 538, 385.

-0.21 0.00Et00

5.00E-04

1.00E-03

1.50E-03 Calixarene Conc.

2.00E-03

2.50E-03

3.00E-03

(M)

Flgure 4. Peak-to-valley ratio versus psuHonlc calix[6]arene concentration: (A) 1,4-benzenediol and pchlorophenol; (6)pchlorophenol and 1,3-benzenediol; (C) 1,9benzenedIol and 1,2-benzenediol; (D) 1,2-benrenedlol and mchlorophenol; (E) mchlorophenol and echlorophenol.

No attempt was made to optimize the efficiency by better matching the buffer ions mobilities to those of the solutes. With lower phosphate concentration only five components could be completely separated since m- and o-benzenediols comigrated. It should be noted that under the same conditions we could separate all three benzenediols when they were injected in the absence of the chlorophenols. The presence of chlorophenols in the mixture affects the migration behavior of the benzenediols, perhaps due to mutual competition for the host molecules. We are now investigating further the mutual effects of one solute on the migration of another. The separation of the above six compounds was not effected by the presence of the borate ions in the buffer. Since borate ions are known to form complexes with phenols,17 it was thought that their addition to the buffer would influence the selectivity of the system. However, such was not the case, perhaps due to stronger complexation with the calixarene. The separation of thep-, 0-, and m-toluidines was carried out in a pH 7 buffer made of 0.05 M sodium dihydrogen phosphate-NaOH and 0.01 M TEA. A full separation of all three isomers could be obtained with a p-sulfonic calix[6]arene concentration of 4 mM and higher. The addition of TEA was necessary to the separation process. Without the addition of TEA, only two wide peaks were observed even at high calixarene concentration. The presence of the amine diminishs wall adsorption of the toluidines and stabilizes the complex formation between the additive and the solute. In all three classes of compounds studied here, the migration order is p < m < o isomers. At the pH of the analysis the p-sulfonic calix[6]arene is negatively charged. Thus, its electrophoretic mobility is in the direction opposite to the electroosmotic flow. There is ample evidence in the literature indicating that elongated solutes can penetrate the calixarene cavity.1c13 Our results are yet inconclusive as to whether the solutes used here enter the calixarene cavity or whether they interact with outer functional groups of the calixarene. However, the migration time order indicates that the complexes formed with the para isomers are less negative than with the meta complexes, which are even less negative than the ortho complexes. To the best of our knowledge, no previous investigations were done on the nature of the interactions

between chlorophenols, dihydroxybenzenes, and toluidines and calixarenes. Furthermore, much to our surprise, there is little or no work relating steric'effects in positional isomers on the complexing abilities of calixarenes. We are now continuing this study with the aim of elucidating the exact mechanism of complex formation in geometric isomers. The effect of the calixarene concentration on the separation of the chlorophenols and the benzenediols is shown in Figure 4, where the peak-to-valley ratiois plottedversus concentration. The peak-to-valley ratio is calculated from the following expression:l8

p="

c

h,+f

where P is the ratio, hv is the height of the valley between two neighboring peaks, and f is the height from the valley to the line connecting the maxima of neighboring peaks. At complete overlap, the peak-to-valley ratio is zero. At baseline separation, that is, when the resolution is 1.5 or greater, the peak-tovalley ratio is unity. We used peak-to-valley ratio rather than the conventional resolution so that we could inject mixtures of the solutes rather than individual components and, thus, eliminate the mutual interaction effects mentioned before. Figure 4 shows that without the calixarene in the buffer there is only a partial separation between the 1,2-benzenediol and the o-chlorophenol and between the p-chlorophenol and the 1,3-benzenediol. Figure 2, which is the electropherogram of a mixture of all chlorophenols and benzenediols, without calixarene in the buffer, can be explained in terms of Figure 4. As the calixarene concentration increases, the trend is an increase in the resolution. At concentrations higher than 3 mM p-sulfonic calix[6]arene, all neighboring solutes are baseline resolved. The reason for the decrease in the peakto-valley ratio, at 1.5 mM calixarene, for m- and o-chlorophenols and for p-chlorophenol and 1,3-benzenediol is not clear to us; we think it is due to an experimental artifact. Figure 5 gives the peak-to-valley plots for the toluidines. In the presence of TEA in the buffer, the resolution of the pair (1 8) Schoenmakers, P. J. Optimization of Chromatographic Selectivity; Elsevier; Holland, 1986; p 119.

Analyt/calChemistry, Vol. 66, No. 5, Merch 1, 1994

749

1.oo

0.80

.0

CI

ep

-

0.60

P)

g b

'?

Y

0.40

2 0.20

0.00 0.00E+00

1.00E-03

2.00E-03

3.00E-03

4.00E-03

5.00E-03

6.00E-03

Calixarene Conc. (MI

Flgure 5. Peak-to-valley ratio versus psulfonic calix[8]arene concentration with toluMines as solutes: (A) p and mtoluidines; (B) m and +toluMines.

p- and m-toluidine increases very rapidly with the increase of calixarene concentration. At concentrations higher than 0.005 M, all three isomers are well resolved. CONCLUSIONS Calixarenes can be used as additives in CE to adjust selectivities. Successful separations of chlorophenols, benzenediols, and toluidines are observed with the addition of 3-4 mM p-sulfonic calix[6]arene. The examples discussed above were obtained with buffers at pH 7. At pH 8, only 2 mM calixarene is sufficient for complete separation of a mixture containing all the chlorophenols and benzenediols. The addition of the calix[6]arene to the buffer is limited by its UV absorption. The baseline in all cases where separation occurred was noisy due to UV absorption by the additive. At the highest additive concentration, the signalto-noise ratio was about 10. However, as the additive

750

Analytical Chemistty, Vol. 66,No. 5, March 1, 1994

concentration was raised, higher concentrationsof test mixtures could be injected, up to their solubility limit. While the high UV absorbance of the calixarenes can be a drawback, it can also be a blessing for indirect detection. The possibility of using calixarenes for selectivity manipulation in an indirect mode of detection is now under investigation.

ACKNOWLEDGMENT This research was supported by Grant 88-00021 from the United States-Israel Binational Science Foundation (BSF), Jerusalem, Israel. Received for review June 10, 1993. Accepted December 8, 1993.' Abstract published in Advance ACS Absrrocrs, January 15, 1994.