- Email: [email protected]
Separation of vitamins by supercritical fluid chromatography with water-modified carbon dioxide as the mobile phase
J. Biochem. Biophys. Methods 43 (2000) 113–123 www.elsevier.com / locate / jbbm
Separation of vitamins by supercritical fluid chromatography with water-modified carbon dioxide as the mobile phase Dongjin Pyo Department of Chemistry, Kangwon National University, Chuncheon 200 -701, South Korea Received 16 December 1999
Abstract Supercritical fluid chromatography (SFC) has become a technique for solving problems that are difficult to be monitored by other chromatographic methods. However, the most widely used fluid, is no more polar than hexane. Polar samples which are difficult to be analyzed with pure supercritical CO 2 because of their high polarity can be separated by adding polar modifiers to supercritical CO 2 . In this paper various vitamins were well separated using water-modified supercritical CO 2 fluid. The amount of water dissolved in supercritical CO 2 was measured using an amperometric microsensor made of a thin film of perfluorosulfonate ionomer (PFSI). 2000 Elsevier Science B.V. All rights reserved. Keywords: Supercritical fluid chromatography; Vitamins; Modifier; Mixing device; Water-modified carbon dioxide; Perfluorosulfonate ionomer
1. Introduction It is generally acknowledged that supercritical fluids are attractive as mobile phases in chromatography owing to their low viscosity and a high diffusivity relative to liquids. Varying the density is the main way of controlling retention in supercritical fluid chromatography (SFC), analogous to varying the temperature in gas chromatography (GC). Additional means of controlling retention in SFC include varying the temperature or the mobile phase composition. The density of the mobile phase in SFC is about 200–500 times that in gas chromatography. The effect of shorter intermolecular 0165-022X / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0165-022X( 00 )00051-8
114
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
distances and the resulting increase in molecular interactions is an enhanced solubilizing capability of the solvent towards various solutes. Compounds with much higher molecular weight than gas chromatography normally allows can therefore be chromatographed. However, the most commonly used mobile phases in SFC are all relatively non-polar fluids. Carbon dioxide, the most widely used fluid, is no more polar than hexane [1], even at high densities. Solute polarity should be between that of the stationary phase and the mobile phase in order to have a well-behaved separation. Few real samples contain only non-polar solutes, so a major objective of research in SFC has been directed toward increasing the range of solute polarity that can be handled by the technique. To bring the SFC technique into routine use, mobile phases that are more polar than the commonly used carbon dioxide are necessary. The solvent strength of supercritical CO 2 , even at high density, is not sufficient for the elution of polar solutes. Polar mobile phases such NH 3 [2] exhibit useful properties, but a more practical way to extend the range of compounds separable by SFC is to use a mixed mobile phase. The solubility of the solute in the supercritical phase can be influenced considerably by adding modifiers to the mobile phase. The use of modifier has been reported by Jentoft and Gouw [3] and by Novotny et al. [2]. The latter group showed that adding 0.1% 2-propanol to pentane as the coefficient (K) values for many polynuclear aromatic hydrocarbons by 20–30%. Thus, the addition of modifiers (generally organic solvents) to a supercritical mobile phase changes the polarity of the mobile phase and also leads to deactivation of the column packing material. In capillary SFC, most separations are made with pure CO 2 because of its compatibility with a flame ionization detection (FID); except for formic acid and water, the addition of any common modifier precludes the use of FID [4]. Modifiers are essential in packed-column SFC for the elution of polar compounds [5] and are extensively used. For the analysis of vitamins, capillary SFC using pure CO 2 as the mobile phase [15] was used. However, packed-column SFC with modified CO 2 has several advantages, such as short retention times and better peak shapes over capillary SFC with pure CO 2 . Several workers [5–7] have reported the influence of modifiers on peak shape, selectivity and retention time in capillary and packed-column SFC. One of the simple and effective ways for the addition of modifiers to supercritical fluid mobile phase reported in the literature is to use a saturator column [8,9], which is usually a silica column saturated with polar modifiers. In our laboratory [10], a m-Porasil column saturated with polar modifiers has been used successfully as a saturator column for a while. A serious disadvantage with this system is that the amount of modifiers dissolved in the mobile phase varies as the mobile phase passes through the saturator column, since the modifier holding capacity of the silica column is limited. Therefore, we attempted to use the porous material for the preparation of mixed fluids for SFC which can maintain the amount of polar modifiers dissolved in supercritical fluid constant for a longer time. In our previous papers [11,12], porous stainless steel filters have been studied and applied to the analysis of vitamins and fatty acids. However, a high porous ion chromatographic filter could maintain the amount of modified in CO 2 constant for a much longer time than any other materials. In this paper, vitamins were separated using water-modified carbon dioxide mobile phase which was generated by passing supercritical CO 2 through a high porous ion chromatographic filter.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
115
Fig. 1. Modifier sensor.
2. Experimental A CCS (Computer Chemical Systems, Avondale, PA, USA) Model 5000 supercritical fluid chromatograph was used with a 100-mm 3 2-mm i.d. packed column (Nucleosil ˚ and the diol). The particle size of the packings was 5 mm, the pore size was 100 A 2 surface area was 350 m / g. This system was equipped with a C14W loop injector (Valco Instruments, Houston, TX, USA) and a flame ionization detector. SFC-grade carbon dioxide (Scott Specialty Gases, Plumsteadville, PA, USA) was used as a basic mobile phase. Experimental conditions for SFC separations are as the follows: supercritical CO 2 at 1508C, pressure programmed from 27.58 to 34.47 MPa at 0.276 MPa min 21 , detector at 3008C and 10 ml min 21 restrictor flow-rate at 10.34 MPa. For the addition of modifiers to supercritical CO 2 , a teflon filter which is manufactured for ion chromatog-
Fig. 2. The arrangement of the water content measuring device: P, pump; V, solenoid valve; S, sensor; FM, flow meter; MP, magnesium perchlorate; R, recorder; M, multimeter; DC, 12-V power supply; PPC, programmable process controller; Ch, Charcoal; Si, silica gel; CS, current source.
116
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
raphy by Dionex was used. The ion chromatographic filter was saturated with modifiers using a 50-ml hypodermic syringe and placed between the pump and injector. To measure the amount of modifier dissolved in the supercritical fluid, an amperometric microsensor was designed and made of perfluorosulfonate ionomer (PFSI) [10]. A constant-current power supply (0.1 mA) (Sungeun, Seoul, South Korea) was used to measure the voltage drop across the sensor. The sensor output was recorded on a strip-chart recorder (Knauer, Berlin, Germany). Fig. 1 shows the cross-section of the modifier sensor used. Platinum wire was wrapped with PFSI thin film and another platinum wire was wound in a coil over the assembly. The sensor made in this way was placed in a plastic tube and the entire modifier measuring device was assembled as shown in Fig. 2.
3. Results and discussion When modifiers are used with supercritical CO 2 in order to chromatograph more polar substances, the binary mixture of eluents can contaminate the instrument. Especially,
Fig. 3. Mixing device.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
117
when water or formic acid is used as a modifier, the modifier remaining in a pump may cause a corrosion of the pump and when methanol is used as a modifier, methanol remaining in a pump can be eluted slowly during the next run. This may affect the time to achieve chemical equilibrium for the next separations, and also many modifiers can diffuse in the laboratory and contaminate the air in the laboratory. A good way to overcome these problems is to use a saturator column [8,9] to add polar modifiers to supercritical CO 2 . In our previous paper [10], water was used as a modifier and a m-Porasil column was used as a saturator column with a similar system design to that of Engelhardt et al. [8]. With this design, a polar modifier (water) can be added to pressurized carbon dioxide fluid after the pump, and thus no modifier remains in the pump. However, when dealing with the use of a saturator column, it should be mentioned that a serious problem always arises. The problem is the bad reproducibility of the amount of modifier dissolved in the supercritical carbon dioxide. When the same experiments were repeated several times, it was very difficult to obtain reproducible results in the chromatograms since the amount of modifier in the mobile phase does not stay constant with time [13]. For these reasons, a new mixing device (Fig. 3) was developed, in which a high porous ion chromatographic filter was used to hold a large amount of water. Two three-way valves on both sides of mixing chamber were used for adding modifier. Modifier was injected using a 50-ml hypodermic syringe. While in the saturator column [8–10] water is held on the stationary phase by hydrogen bonding, with this device water is held physically inside the small pores of the filters. After being saturated with water, the device is placed between a pump and an injector. With this design, supercritical CO 2 is delivered from
Fig. 4. Measurements of the water content in supercritical CO 2 as a function of time.
118
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
the pump to the device which is saturated with water. When supercritical CO 2 goes through the device, water held within the small pores of the filters can be dissolved in the pressurized supercritical fluids. Thus, nonpolar supercritical CO 2 can demonstrate the characteristics of a polar mobile phase because it can absorb polar solvent, H 2 O. Therefore, after passing through the mixing device, supercritical CO 2 is changed to a new mobile phase with a different polarity. To measure the amount of water dissolved in supercritical CO 2 after the mixing device, a modifier sensor [10] which consists of a polymer film [14], i.e., a film of a perfluorosulfonate ionomer (PFSI), was used. PFSI polymer has a high affinity for water. When the PFSI film was in contact with two electrodes and a constant current flowed through the film, the water that partitions into the film from the surrounding environment was electrolytically decomposed. The change of voltage across the two electrodes was used as a measure of the water content of the environment surrounding the sensor. The resistance of the PFSI film was changed according to the water content
Fig. 5. Chromatograms of 4-nitrophenol using the mixing device (A) for 2 h after saturation with water, (B) 2.5 h after saturation and (C) 3 h after saturation.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
119
Fig. 6. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) nicotinitic acid; (2) nicotinitic amide; and (3) ascorbic acid. Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
120
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
of the supercritical CO 2 fluid and the voltage difference between two platinum wires was recorded and measured. To evaluate the performance of our mixing device, the amount of water dissolved in the supercritical CO 2 fluid after passing through the mixing device saturated with water was monitored as a function of time (Fig. 2). Fig. 4 shows that the amount of water in the mobile phase remained constant for about 2 h. Fig. 4 demonstrates that the mixing device with a high porous ion chromatographic filter can be used successfully to generate water-modified carbon dioxide mobile phase. After going through this mixing device, supercritical CO 2 was changed to a more polar mobile phase, and it was possible to separate polar samples using this new mobile phase. As a control, 4-nitrophenol dissolved in dichloromethane was separated using our mixing device (Fig. 5). For 2 h after water was injected using a hypodermic syringe, 4-nitrophenol was eluted in 11.50
Fig. 7. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) vitamin K; (2) vitamin E; and (3) vitamin D. Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
121
Fig. 8. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) vitamin B 1 ; (2) vitamin B 6 ; and (3) vitamin B 2 . Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
122
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
Fig. 9. Standard calibration curves of vitamins B 1 , B 6 and B 2 .
min; however, after 2.5 h it was eluted in 12.81 min and after 3 h it was eluted in 13.27 min. These results demonstrate that the water content in CO 2 can be kept constant for 2 h; however, after 2 h, the water content decreases slowly. This is, of course, because only limited amounts of water were held in the pores of the ion chromatographic filter. Finally, experiments to separate various vitamins with this mixing device were performed. Figs. 6–8 are chromatograms for mixtures of vitamins obtained using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. In contrast to the chromatogram of Figs. 6A, 7A and 8A, excellent separations (see Figs. 6B, 7B and 8B) were obtained when water was added to supercritical CO 2 . This is because the solvent strength of pure supercritical CO 2 is not sufficient for the elution of polar compounds, such as, vitamins. The standard calibration curves of vitamin B 1 , B 6 , B 2 (Fig. 9) show a good linear relationship in the range of 50–200 mg / ml, and the reproducibility of retention times of these vitamins (Table 1) was very good under the water mixing conditions.
Acknowledgements This investigation was supported by a grant from KOSEF (97-2-03-0101-3).
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
123
Table 1 Supercritical fluid retention times of vitamins B 1 , B 6 and B 2
Vitamin B 1 Vitamin B 6 Vitamin B 2
Mean
R.S.D.
5.98 8.09 11.23
2.1 2.4 4.1
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
Yonker CR, Frye SL, Lalkwarf DR, Smith RD. J Phys Chem 1986;90:3022. Novotny M, Bertsch W, Zlatkis A. J Chromatogr 1971;61:17. Jentoft RE, Gouw TH. J Chromatogr Sci 1970;8:138. Wright BW, Smith RD. J Chromatogr 1986;355:367. Schmidt S, Blomberg LG, Campbell ER. Chromatographia 1988;25:775. Blilie AL, Greibrokk T. Anal Chem 1985;57:2239. Yonker CR, Smith RD. J Chromatogr 1986;361:25. Engelhardt H, Gross A, Mertens R, Petersen M. J Chromatogr 1989;477:169. Schwartz HE, Barthel PJ, Moring SE, Yates TL, Lauer HH. Fresenius’ Z Anal Chem 1988;330:204. Pyo D, Ju D. Analyst (London) 1993;118:253. Pyo D, Park D, Kim H, Lee H, Lee T. Anal Sci Technol 1997;10:131. Pyo D, Ju D. Anal Sci 1994;10:171. Pyo D, Hwang H. Anal Chim Acta 1993;280:103. Huang H, Dasgupta PK. Anal Chem 1990;62:1935. Shen Y, Bradshaw JS, Lee ML. Chromatographia 1996;43:53.
Separation of vitamins by supercritical fluid chromatography with water-modified carbon dioxide as the mobile phase Dongjin Pyo Department of Chemistry, Kangwon National University, Chuncheon 200 -701, South Korea Received 16 December 1999
Abstract Supercritical fluid chromatography (SFC) has become a technique for solving problems that are difficult to be monitored by other chromatographic methods. However, the most widely used fluid, is no more polar than hexane. Polar samples which are difficult to be analyzed with pure supercritical CO 2 because of their high polarity can be separated by adding polar modifiers to supercritical CO 2 . In this paper various vitamins were well separated using water-modified supercritical CO 2 fluid. The amount of water dissolved in supercritical CO 2 was measured using an amperometric microsensor made of a thin film of perfluorosulfonate ionomer (PFSI). 2000 Elsevier Science B.V. All rights reserved. Keywords: Supercritical fluid chromatography; Vitamins; Modifier; Mixing device; Water-modified carbon dioxide; Perfluorosulfonate ionomer
1. Introduction It is generally acknowledged that supercritical fluids are attractive as mobile phases in chromatography owing to their low viscosity and a high diffusivity relative to liquids. Varying the density is the main way of controlling retention in supercritical fluid chromatography (SFC), analogous to varying the temperature in gas chromatography (GC). Additional means of controlling retention in SFC include varying the temperature or the mobile phase composition. The density of the mobile phase in SFC is about 200–500 times that in gas chromatography. The effect of shorter intermolecular 0165-022X / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0165-022X( 00 )00051-8
114
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
distances and the resulting increase in molecular interactions is an enhanced solubilizing capability of the solvent towards various solutes. Compounds with much higher molecular weight than gas chromatography normally allows can therefore be chromatographed. However, the most commonly used mobile phases in SFC are all relatively non-polar fluids. Carbon dioxide, the most widely used fluid, is no more polar than hexane [1], even at high densities. Solute polarity should be between that of the stationary phase and the mobile phase in order to have a well-behaved separation. Few real samples contain only non-polar solutes, so a major objective of research in SFC has been directed toward increasing the range of solute polarity that can be handled by the technique. To bring the SFC technique into routine use, mobile phases that are more polar than the commonly used carbon dioxide are necessary. The solvent strength of supercritical CO 2 , even at high density, is not sufficient for the elution of polar solutes. Polar mobile phases such NH 3 [2] exhibit useful properties, but a more practical way to extend the range of compounds separable by SFC is to use a mixed mobile phase. The solubility of the solute in the supercritical phase can be influenced considerably by adding modifiers to the mobile phase. The use of modifier has been reported by Jentoft and Gouw [3] and by Novotny et al. [2]. The latter group showed that adding 0.1% 2-propanol to pentane as the coefficient (K) values for many polynuclear aromatic hydrocarbons by 20–30%. Thus, the addition of modifiers (generally organic solvents) to a supercritical mobile phase changes the polarity of the mobile phase and also leads to deactivation of the column packing material. In capillary SFC, most separations are made with pure CO 2 because of its compatibility with a flame ionization detection (FID); except for formic acid and water, the addition of any common modifier precludes the use of FID [4]. Modifiers are essential in packed-column SFC for the elution of polar compounds [5] and are extensively used. For the analysis of vitamins, capillary SFC using pure CO 2 as the mobile phase [15] was used. However, packed-column SFC with modified CO 2 has several advantages, such as short retention times and better peak shapes over capillary SFC with pure CO 2 . Several workers [5–7] have reported the influence of modifiers on peak shape, selectivity and retention time in capillary and packed-column SFC. One of the simple and effective ways for the addition of modifiers to supercritical fluid mobile phase reported in the literature is to use a saturator column [8,9], which is usually a silica column saturated with polar modifiers. In our laboratory [10], a m-Porasil column saturated with polar modifiers has been used successfully as a saturator column for a while. A serious disadvantage with this system is that the amount of modifiers dissolved in the mobile phase varies as the mobile phase passes through the saturator column, since the modifier holding capacity of the silica column is limited. Therefore, we attempted to use the porous material for the preparation of mixed fluids for SFC which can maintain the amount of polar modifiers dissolved in supercritical fluid constant for a longer time. In our previous papers [11,12], porous stainless steel filters have been studied and applied to the analysis of vitamins and fatty acids. However, a high porous ion chromatographic filter could maintain the amount of modified in CO 2 constant for a much longer time than any other materials. In this paper, vitamins were separated using water-modified carbon dioxide mobile phase which was generated by passing supercritical CO 2 through a high porous ion chromatographic filter.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
115
Fig. 1. Modifier sensor.
2. Experimental A CCS (Computer Chemical Systems, Avondale, PA, USA) Model 5000 supercritical fluid chromatograph was used with a 100-mm 3 2-mm i.d. packed column (Nucleosil ˚ and the diol). The particle size of the packings was 5 mm, the pore size was 100 A 2 surface area was 350 m / g. This system was equipped with a C14W loop injector (Valco Instruments, Houston, TX, USA) and a flame ionization detector. SFC-grade carbon dioxide (Scott Specialty Gases, Plumsteadville, PA, USA) was used as a basic mobile phase. Experimental conditions for SFC separations are as the follows: supercritical CO 2 at 1508C, pressure programmed from 27.58 to 34.47 MPa at 0.276 MPa min 21 , detector at 3008C and 10 ml min 21 restrictor flow-rate at 10.34 MPa. For the addition of modifiers to supercritical CO 2 , a teflon filter which is manufactured for ion chromatog-
Fig. 2. The arrangement of the water content measuring device: P, pump; V, solenoid valve; S, sensor; FM, flow meter; MP, magnesium perchlorate; R, recorder; M, multimeter; DC, 12-V power supply; PPC, programmable process controller; Ch, Charcoal; Si, silica gel; CS, current source.
116
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
raphy by Dionex was used. The ion chromatographic filter was saturated with modifiers using a 50-ml hypodermic syringe and placed between the pump and injector. To measure the amount of modifier dissolved in the supercritical fluid, an amperometric microsensor was designed and made of perfluorosulfonate ionomer (PFSI) [10]. A constant-current power supply (0.1 mA) (Sungeun, Seoul, South Korea) was used to measure the voltage drop across the sensor. The sensor output was recorded on a strip-chart recorder (Knauer, Berlin, Germany). Fig. 1 shows the cross-section of the modifier sensor used. Platinum wire was wrapped with PFSI thin film and another platinum wire was wound in a coil over the assembly. The sensor made in this way was placed in a plastic tube and the entire modifier measuring device was assembled as shown in Fig. 2.
3. Results and discussion When modifiers are used with supercritical CO 2 in order to chromatograph more polar substances, the binary mixture of eluents can contaminate the instrument. Especially,
Fig. 3. Mixing device.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
117
when water or formic acid is used as a modifier, the modifier remaining in a pump may cause a corrosion of the pump and when methanol is used as a modifier, methanol remaining in a pump can be eluted slowly during the next run. This may affect the time to achieve chemical equilibrium for the next separations, and also many modifiers can diffuse in the laboratory and contaminate the air in the laboratory. A good way to overcome these problems is to use a saturator column [8,9] to add polar modifiers to supercritical CO 2 . In our previous paper [10], water was used as a modifier and a m-Porasil column was used as a saturator column with a similar system design to that of Engelhardt et al. [8]. With this design, a polar modifier (water) can be added to pressurized carbon dioxide fluid after the pump, and thus no modifier remains in the pump. However, when dealing with the use of a saturator column, it should be mentioned that a serious problem always arises. The problem is the bad reproducibility of the amount of modifier dissolved in the supercritical carbon dioxide. When the same experiments were repeated several times, it was very difficult to obtain reproducible results in the chromatograms since the amount of modifier in the mobile phase does not stay constant with time [13]. For these reasons, a new mixing device (Fig. 3) was developed, in which a high porous ion chromatographic filter was used to hold a large amount of water. Two three-way valves on both sides of mixing chamber were used for adding modifier. Modifier was injected using a 50-ml hypodermic syringe. While in the saturator column [8–10] water is held on the stationary phase by hydrogen bonding, with this device water is held physically inside the small pores of the filters. After being saturated with water, the device is placed between a pump and an injector. With this design, supercritical CO 2 is delivered from
Fig. 4. Measurements of the water content in supercritical CO 2 as a function of time.
118
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
the pump to the device which is saturated with water. When supercritical CO 2 goes through the device, water held within the small pores of the filters can be dissolved in the pressurized supercritical fluids. Thus, nonpolar supercritical CO 2 can demonstrate the characteristics of a polar mobile phase because it can absorb polar solvent, H 2 O. Therefore, after passing through the mixing device, supercritical CO 2 is changed to a new mobile phase with a different polarity. To measure the amount of water dissolved in supercritical CO 2 after the mixing device, a modifier sensor [10] which consists of a polymer film [14], i.e., a film of a perfluorosulfonate ionomer (PFSI), was used. PFSI polymer has a high affinity for water. When the PFSI film was in contact with two electrodes and a constant current flowed through the film, the water that partitions into the film from the surrounding environment was electrolytically decomposed. The change of voltage across the two electrodes was used as a measure of the water content of the environment surrounding the sensor. The resistance of the PFSI film was changed according to the water content
Fig. 5. Chromatograms of 4-nitrophenol using the mixing device (A) for 2 h after saturation with water, (B) 2.5 h after saturation and (C) 3 h after saturation.
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
119
Fig. 6. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) nicotinitic acid; (2) nicotinitic amide; and (3) ascorbic acid. Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
120
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
of the supercritical CO 2 fluid and the voltage difference between two platinum wires was recorded and measured. To evaluate the performance of our mixing device, the amount of water dissolved in the supercritical CO 2 fluid after passing through the mixing device saturated with water was monitored as a function of time (Fig. 2). Fig. 4 shows that the amount of water in the mobile phase remained constant for about 2 h. Fig. 4 demonstrates that the mixing device with a high porous ion chromatographic filter can be used successfully to generate water-modified carbon dioxide mobile phase. After going through this mixing device, supercritical CO 2 was changed to a more polar mobile phase, and it was possible to separate polar samples using this new mobile phase. As a control, 4-nitrophenol dissolved in dichloromethane was separated using our mixing device (Fig. 5). For 2 h after water was injected using a hypodermic syringe, 4-nitrophenol was eluted in 11.50
Fig. 7. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) vitamin K; (2) vitamin E; and (3) vitamin D. Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
121
Fig. 8. Chromatograms of mixtures of vitamins using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. Peaks: (1) vitamin B 1 ; (2) vitamin B 6 ; and (3) vitamin B 2 . Separation conditions: 1508C; pressure programming, 27.58–34.47 MPa (rate, 0.276 MPa / min); column, 100-mm 3 2-mm i.d. packed column (Nucleosil diol).
122
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
Fig. 9. Standard calibration curves of vitamins B 1 , B 6 and B 2 .
min; however, after 2.5 h it was eluted in 12.81 min and after 3 h it was eluted in 13.27 min. These results demonstrate that the water content in CO 2 can be kept constant for 2 h; however, after 2 h, the water content decreases slowly. This is, of course, because only limited amounts of water were held in the pores of the ion chromatographic filter. Finally, experiments to separate various vitamins with this mixing device were performed. Figs. 6–8 are chromatograms for mixtures of vitamins obtained using (A) pure supercritical CO 2 mobile phase and (B) water-modified supercritical CO 2 mobile phase. In contrast to the chromatogram of Figs. 6A, 7A and 8A, excellent separations (see Figs. 6B, 7B and 8B) were obtained when water was added to supercritical CO 2 . This is because the solvent strength of pure supercritical CO 2 is not sufficient for the elution of polar compounds, such as, vitamins. The standard calibration curves of vitamin B 1 , B 6 , B 2 (Fig. 9) show a good linear relationship in the range of 50–200 mg / ml, and the reproducibility of retention times of these vitamins (Table 1) was very good under the water mixing conditions.
Acknowledgements This investigation was supported by a grant from KOSEF (97-2-03-0101-3).
D. Pyo / J. Biochem. Biophys. Methods 43 (2000) 113 – 123
123
Table 1 Supercritical fluid retention times of vitamins B 1 , B 6 and B 2
Vitamin B 1 Vitamin B 6 Vitamin B 2
Mean
R.S.D.
5.98 8.09 11.23
2.1 2.4 4.1
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
Yonker CR, Frye SL, Lalkwarf DR, Smith RD. J Phys Chem 1986;90:3022. Novotny M, Bertsch W, Zlatkis A. J Chromatogr 1971;61:17. Jentoft RE, Gouw TH. J Chromatogr Sci 1970;8:138. Wright BW, Smith RD. J Chromatogr 1986;355:367. Schmidt S, Blomberg LG, Campbell ER. Chromatographia 1988;25:775. Blilie AL, Greibrokk T. Anal Chem 1985;57:2239. Yonker CR, Smith RD. J Chromatogr 1986;361:25. Engelhardt H, Gross A, Mertens R, Petersen M. J Chromatogr 1989;477:169. Schwartz HE, Barthel PJ, Moring SE, Yates TL, Lauer HH. Fresenius’ Z Anal Chem 1988;330:204. Pyo D, Ju D. Analyst (London) 1993;118:253. Pyo D, Park D, Kim H, Lee H, Lee T. Anal Sci Technol 1997;10:131. Pyo D, Ju D. Anal Sci 1994;10:171. Pyo D, Hwang H. Anal Chim Acta 1993;280:103. Huang H, Dasgupta PK. Anal Chem 1990;62:1935. Shen Y, Bradshaw JS, Lee ML. Chromatographia 1996;43:53.