Anal. Chem. 2003, 75, 5525-5531
Polymer-Coated Fibrous Materials as the Stationary Phase in Packed Capillary Gas Chromatography Yoshihiro Saito,† Ai Tahara,† Motohiro Imaizumi,† Tsutomu Takeichi,† Hiroo Wada,‡ and Kiyokatsu Jinno*,†
School of Materials Science, Toyohashi University of Technology, Toyohashi 441-8580, Japan, and Shinwa Chemical Industries, Ltd., 50 Kagekatsu-cho, Fushimi-ku, Kyoto 612-8307, Japan
Synthetic polymer filaments have been introduced as the support material in packed capillary gas chromatography (GC). The filaments of the heat-resistant polymers, Zylon, Kevlar, Nomex, and Technora, were longitudinally packed into a short fused-silica capillary, followed by the conventional coating process for open-tubular GC columns. The separation of several test mixtures such as n-alkylbenzenes and n-alkanes was carried out with these polymercoated fiber-packed capillary columns. With the coating by various polymeric materials on the surface of these filaments, the retentivity was significantly improved over the parent fiber-packed column (without polymer coating) as well as a conventional open-tubular capillary of the same length. The results demonstrated a good combination of Zylon as the support and poly(dimethylsiloxane)based materials as the coating liquid-phase for the successful GC separation of n-alkanes and polycyclic aromatic hydrocarbons (PAHs), while successful applications for other separations such as poly(ethylene glycol) coating for the separation of alcohols were also obtained. From the results it has been suggested that the selectivity of the fiber-packed column could be tuned by selecting different coating materials, indicating the promising possibility for a novel usage of fine fibrous polymers as the support material that can be combined with newly synthesized coating materials specially designed for particular separations. Taking advantage of good thermal stability of the fibers, the column temperature could be elevated to higher than 350 °C with the combination of a short metallic capillary. Gas chromatography (GC) is one of the most versatile separation methods for volatile compounds, and a wide variety of polymeric materials as the stationary phases have been developed for various applications.1-3 In contrast to the successful applica* Corresponding author. Phone: +81-532-44-6805. Fax: +81-532-48-5833. E-mail:
[email protected]. † Toyohashi University of Technology. ‡ Shinwa Chemical Industries, Ltd. (1) Eiceman, G. A.; Gardea-Torresday, J.; Overton, E.; Carney, K.; Dorman, F. Anal. Chem. 2002, 74, 2771-2780. (2) Eiceman, G. A.; Hill, H. H.; Gardea-Torresday, J. Anal. Chem. 2000, 72, 137R-144R. 10.1021/ac030052h CCC: $25.00 Published on Web 09/09/2003
© 2003 American Chemical Society
tions and subsequent commercialization of these polymer-coated columns, such as wall-coated, support-coated, and pellicular-coated, the reports for the fibrous polymer-packed stationary phases, however, were somewhat limited except for characterizing the surface of the fibers in the inverse GC (IGC) technique.4-8 The IGC method is based on the observation of the specific interactions between the surface of the fibers and the standard solutes injected as the probes. Therefore, it is quite natural that a synthetic fiber can be employed as a stationary-phase material in GC separations, if the fiber possesses both the thermal stability and resistance to the gaseous chemical species throughout the column during the separation process. In terms of the employment of fibrous materials as the stationary phase in liquid-phase separation methods, Jinno et al. have already reported the use of fibrous cellulose acetate (CA) stationary phase for the separation of alcohols9 and polycyclic aromatic hydrocarbons (PAHs)10 in microcolumn LC as well as the application in capillary electrochromatography (CEC).11,12 Recently, the application of other synthetic fibers with a good solvent stability as the stationary phase was reported.13,14 Although the separation efficiency of those columns was not comparable with conventional particle-packed columns, for example, octadecylsilica (ODS), and there were several limitations such as on mobile-phase composition and on packing density, it has been (3) Abraham, M. H.; Poole, C. F.; Poole, S. K. J. Chromatogr. A 1999, 842, 79-114. (4) Inverse Gas Chromatography: Characterization of Polymers and Other Materials; Lloyd, D. R., Ward, T. C., Schreiber, H. P., Eds.; ACS Symposium Series 391; American Chemical Society: Washington, DC, 1989. (5) Munk, P. Polymer Characterization Using Inverse Gas Chromatography. In Modern Methods of Polymer Characterization; Barth, H. G., Mays, J. W., Eds.; John Wiley & Sons: New York, 1991. (6) Voelkel, A. Crit. Rev. Anal. Chem. 1991, 22, 411-439. (7) Al-Saigh, Z. Y. Trends Polym. Sci. 1997, 5, 97-102. (8) Al-Saigh, Z. Y. Int. J. Polym. Anal. Charact. 1997, 3, 249-291. (9) Kiso, Y.; Jinno, K.; Nagoshi, T. J. High Resolut. Chromatogr. Chromatogr. Commun. 1986, 9, 763-764. (10) Kiso, Y.; Takayama, K.; Jinno, K. J. High Resolut. Chromatogr. 1989, 12, 169-173. (11) Jinno, K.; Wu, J.; Sawada, H.; Kiso, Y. J. High Resolut. Chromatogr. 1998, 21, 617-619. (12) Jinno, K.; Watanabe, H.; Kiso, Y. J. Biochem. Biophys. Methods 2000, 48, 209-218. (13) Jinno, K.; Watanabe, H.; Saito, Y.; Takeichi, T. Electrophoresis 2001, 22, 3371-3376. (14) Saito, Y.; Kawazoe, M.; Imaizumi, M.; Morishima, Y.; Nakao, Y.; Hatano, K.; Hayashida, M.; Jinno, K. Anal. Sci. 2002, 18, 7-17.
Analytical Chemistry, Vol. 75, No. 20, October 15, 2003 5525
clearly confirmed that the fibrous polymer-packed capillary could play a role as the stationary phase in the chromatographic process. Miniaturized sample preparation with fibrous extraction medium has been also reported by Saito et al., where several hundreds of polymeric filaments were longitudinally packed into a short capillary to prepare the extraction cartridge.15-20 The method, fiber-in-tube solid-phase extraction (FIT-SPE), was successfully applied to the sample preparation of complex sample mixtures.19-24 With the FIT-SPE technique, a trace amount of analytes could be extracted onto the surface of the fine filaments in the extraction tube by simply passing the aqueous sample solution into the extraction tube. The extracted analytes are desorbed by passing a small amount of solvents in a similar manner and are simultaneously transferred to the injection loop of LC system for the subsequent chromatographic separation. Therefore, several advantageous features can be found with the FIT format. Comparing with conventional in-tube solid-phase extraction (in-tube SPME) developed by Pawliszyn et al.,25-27 in which a polymer-coated fused-silica capillary (open-tubular capillary GC column) was employed as the extraction tube, the surface area of the extraction medium was dramatically increased in FITSPE because a large number of fine filaments could be packed into the extraction tube, while maintaining a good permeability through the tube. The improved extraction power of FIT-SPE has a good agreement with the results in the previous investigations, where the extraction efficiency was improved by multiple draw/ eject procedure25-27 and by inserting a stainless wire into the extraction capillary.22,28 Based on the above successful applications of fibrous materials as the stationary phase and the extraction medium, fiber-packed capillaries have been introduced for the separation of volatile compounds in GC.14,29 The results also showed promise for fibrous material as the stationary phase in GC, especially as the packing material for a short column having a large sample loading capacity. In this work, polymer coating onto the packed filaments in a short capillary has been investigated using several conventional polymeric materials. Several synthetic fibers were employed taking into account heat resistance and solvent resistance as the require(15) Saito, Y.; Nakao, Y.; Imaizumi, M.; Takeichi, T.; Kiso, Y.; Jinno, K. Fresenius J. Anal. Chem. 2000, 368, 641-643. (16) Saito, Y.; Imaizumi, M.; Takeichi, T.; Jinno, K. Anal. Bioanal. Chem. 2002, 372, 164-168. (17) Morishima, Y.; Saito, Y.; Fujimoto, C.; Takeichi, T.; Jinno, K. Chromatographia 2002, 56, 585-590. (18) Saito, Y.; Jinno, K. Anal. Bioanal. Chem. 2002, 373, 325-331. (19) Saito, Y.; Jinno, K. J. Chromatogr., A 2003, 1000, 53-67. (20) Saito, Y.; Ohta, H.; Jinno, K. J. Sep. Sci. 2003, 26, 242-250. (21) Jinno, K.; Kawazoe, M.; Saito, Y.; Takeichi, T.; Hayashida, M. Electrophoresis 2001, 22, 3785-3790. (22) Saito, Y.; Nakao, Y.; Imaizumi, M.; Morishima, Y.; Kiso, Y.; Jinno, K, Anal. Bioanal. Chem. 2002, 373, 81-86. (23) Saito, Y.; Nojiri, M.; Imaizumi, M.; Nakao, Y.; Morishima, Y.; Kanehara, H.; Matsuura, H.; Kotera, K.; Wada, H.; Jinno, K. J. Chromatogr., A 2002, 975, 105-112. (24) Imaizumi, M.; Saito, Y.; Hayashida, M.; Takeichi, T.; Wada, H.; Jinno, K. J. Pharm. Biomed. Anal. 2003, 30, 1801-1808. (25) Eisert, R.; Pawliszyn, J. Anal. Chem. 1997, 69, 3140-3147. (26) Kataoka, H.; Narimatsu, S.; Lord, H. L.; Pawliszyn, J. Anal. Chem. 1999, 71, 4237-4244. (27) Applications of Solid-Phase Microextraction; Pawliszyn, J., Ed.; Royal Society of Chemistry: Cambridge, U.K., 1999. (28) Saito, Y.; Kawazoe, M.; Hayashida, M.; Jinno, K. Analyst 2000, 125, 807809. (29) Saito, Y.; Imaizumi, M.; Nakata, K.; Takeichi, T.; Kotera, K.; Wada, H.; Jinno, K. J. Microcolumn Sep. 2001, 13, 259-264.
5526 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
Table 1. Heat-Resistant Fibers Used
fiber
diameter (µm) [%RSD] (n ) 10)]
no. of packed filamentsa
temp (°C) for 5% weight lossb
temp (°C) for 10% weight lossb
Zylon Kevlar Nomex Technora
11.5 [2.1] 12.4 [2.0] 17.5 [1.8] 12.5 [2.0]
170 ( 2 146 ( 2 73 ( 2 144 ( 2
690 550 430 480
710 570 450 490
a For fused-silica capillary of 0.32 mm i.d. b RSDs were less than 2.0% for all measurements (n ) 3).
ments for GC stationary phase at elevated column temperatures. For the coating onto the packed filaments, typical commercially available polymeric materials were chosen, and the separation was compared with that of the parent fiber-packed capillary column. With the combination of the excellent thermal stability of the fiber and metal capillary, temperature-programmed separations of semivolatile compounds have been established with the maximum column temperature at higher than 350 °C. EXPERIMENAL SECTION Reagents. All solvents and sample solutes were of analytical grade and purchased either from Kishida Chemical, Osaka, Japan or Tokyo Chemical Industries, Tokyo, Japan. The standard sample mixtures were prepared with either n-hexane, n-heptane, or dichloromethane as the sample solvent, and typical concentrations of these mixtures were about 0.1-0.5 wt/vol % unless otherwise specified. Column Preparation. As shown in Table 1, four types of fibers were obtained from respective manufacture representatives in Japan as follows: Zylon (Toyobo Co., Ltd., Shiga, Japan); Kevlar (Du Pont-Toray Co., Ltd., Tokyo, Japan); Nomex (Du Pont, Tokyo, Japan); and Technora (Teijin, Ltd., Osaka, Japan). The chemical structures of these fibers are also illustrated in Figure 1. To prepare the fiber-packed columns, these fibers were longitudinally packed into fused-silica capillaries of either 0.53 mm i.d. or 0.32 mm i.d. (Shinwa Chemical Industries, Ltd., Kyoto, Japan) as described previously.29 The initial length of the columns was 1.0 m, and the number of Zylon filaments packed in these capillaries was about 170 and 330 for 0.32 mm i.d. and 0.53 mm i.d. columns, respectively, while the number of the packed filaments for other three fibers (Table 1) was determined to maintain the packing density about the same as Zylon fiber. The packing density was calculated as the proportion of the cross-section area of total packed filaments to that of the capillary. Figure 2 shows the typical cross-section photograph of the fiber-packed column taken by scanning electron microscope (SEM; Model JSM-5900LV, JEOL, Tokyo, Japan). For the preparation of coated fiber packings, five types of polymeric materials were employed: HR-1, 100%-methyl-polysiloxane; HR-52, 5%-phenyl-95%-methyl-polysiloxane; HR-1701, 7%phenyl-7%-cyanopropyl-86%-methyl-polysiloxane; HR-17, 50%-phenyl50%-methyl-polysiloxane and HR-20M, poly(ethylene glycol) (Shinwa Chemical Industries). The coating procedure was similar to the preliminary experiments23,29 as described below. First, a fiberpacked capillary of 1.0 m length was connected to the pressureproof vessel containing 10 mL of acetone and washed with the solvent pumped by N2 gas at the pressure of 500 kPa. The suitable
Figure 1. Chemical structures of fibrous polymers: (A) Zylon; (B) Kevlar; (C) Nomex; and (D) Technora.
Figure 2. Typical cross-section photograph of a polymer-coated fiber-packed column (0.53 mm i.d.).
number of filaments packed for the preparation of the coated fiberpacked capillary was optimized by the preliminary experiments. After the same volume of the following solvents, water, acetone, and chloroform, were pumped in a similar manner, the capillary was left to dry at room temperature for about 3 h using N2 flow at about 1.0 mL/min. Second, the capillary was subject to heating in a GC oven with the flow of N2 gas (at about 1.0 mL/min), while the outlet of the column was disconnected from the detector. The temperature was programmed from room temperature to 300 °C at 2 deg/min and then held for about 10 h. Next, a solution of the polymeric coating material in n-hexane/acetone (90/10) containing a cross-linking reagent was pumped through the packed capillary. The concentration of polymer in the coating solution was set at 5.0% unless otherwise specified. After the total volume of the polymer solution (0.5 mL) was pumped, the N2 flow (ca. 1.0 mL/min) was maintained for more than 5 h. For the crosslinking and chemical bonding reaction, the column was installed in the GC oven again without being connected to the detector and the programmed heating was carried out as follows: from 40 to 280 °C at 0.5 deg/min and then held about 48 h to be certain of a complete reaction. The successful fiber-packing and polymer-coating treatments were confirmed by the separation of a standard sample containing three n-alkanes, n-dodecane, n-tetradecane, and n-hexadecane, and the relative standard deviations (RSDs) for the retention factors
were less than about 2.5% on three columns separately prepared with the same coating, where the average values of the retention factors for three consecutive runs on each column were used for the calculation. The RSDs for multiple injection onto the same column were less than 1.0% (n ) 3) for all the columns studied. These results showed the reproducible preparation of polymercoated fiber-packed columns with a precise control of the several experimental conditions such as the solvent washing and coating of the filaments and the number of packed filaments in the capillary (Table 1). To prepare the columns of 2.0-m length, two of these columns were connected with capillary column connector (GL Sciences, Tokyo, Japan) after the coating process as described above. Thermal Analysis. For the thermogravimetric analysis (TGA) of these fibrous materials, Thermo Plus 2 TG-DTA TG8120 (Rigaku, Tokyo, Japan) was employed. GC Measurements. A HP 5890-II Gas Chromatograph (Yokogawa Analytical Systems, Musashino, Japan) with a split/ splitless injection port and a flame ionization detector (FID) was used for all GC measurements. The injector and detector temperature was set at 300 °C, and a typical injection volume was 1.0 µL unless otherwise specified. As the carrier gas, He or N2 was used, and the carrier gas and air were supplied from the respective gas cylinder through the cartridge packed with molecular sieve. The other separation conditions such as carrier gas flowrate, column head pressure, and temperature programs were determined by the results of preliminary experiments for each sample. All GC measurements were done at least three times, and the RSDs for the retention times were less than 1.0% as described above. The data collection was made with Borwin Chromatography Data Handling Software (Jasco, Tokyo, Japan) running on a personal computer. RESULTS AND DISCUSSION Evaluation of the Thermal Stability of Fibrous Materials. Prior to the preparation of fiber-packed columns, the TGA measurements of the fibers (Figure 1) were carried out using the temperature-programmed run from room temperature to 850 °C at the rate of 10 deg/min under Ar atmosphere. From the TGA measurements the temperatures for 5% and 10% weight loss were calculated since these have been normally employed as the indicator of the thermal stability of such polymeric materials. In general, relatively higher thermal stability can be accomplished with a polymer having a linear rigid-rod chemical structure if the functionalities included in these structures are similar. The results in Table 1 have a good agreement with this general trend; that is, Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
5527
Figure 3. Comparison of four fibers as the support material: (A) Zylon; (B) Kevlar; (C) Nomex; and (D) Technora. Sample, n-alkanes mixture (C11H24 to C20H42). Conditions: column size, 0.32 mm i.d. × 1.0 m; temperature program, 80 °C (1.5 min) to 250 °C at 10 deg/ min; splitless injection.
Zylon (most rigid structure) possesses the highest thermal stability, Kevlar shows better thermal stability than Nomex because of the linearity of the structure, and an intermediate stability between Kevlar and Nomex was obtained with Technora. In terms of the thermal stability for GC applications, Zylon is the most appropriate fiber; however, three other fibers might show a different selectivity for a particular class of compounds due to the different surface characteristics induced by the chemical structure.13 Comparison of Fibrous Materials. To evaluate the contribution of the fibrous support to the separation, four fibrous materials were coated with the same polymeric phase, and the chromatographic separations of an alkanes mixture have been compared, where the packing density of the filaments in these capillaries were maintained almost the same. From the chromatograms in Figure 3, different separation characteristics can be observed depending on the type of the fibers used as the support material. Good separation was obtained with Zylon and Kevlar, while tailing of the peaks was found for other two fibers, especially for Technora. Furthermore, the tailing of the solvent, dichloromethane, was observed for Nomex and Technora even in the
other separation conditions. In the preliminary experiments, Nomex and Technora showed significant peak tailing for the alkanes, if the fibers were employed as the stationary phase without any polymeric coating. Therefore, there is a slight contribution of the fibrous support to the separation, although there is a larger effect of the polymeric coating on the separation characteristics for the separation. It can be also assumed that the surface condition induced by the chemical structure has a certain effect on the wettability of filaments by the coating materials. The existence of polar functional groups on the fiber surface can be expected for polyamides with a low degree of crystallinity such as Nomex and Technora. Then, the combination of the fibrous support and the coating material should be systematically investigated more along with the optimization of the coating conditions for each set; it can be said that Zylon is the most appropriate support material for these polysiloxane-based coating materials. In the following experiments, we have mainly selected Zylon fiber as the support material. Effect of Polymer Coating on Retention. For the evaluation of the effect of polymer concentration for the coating process on the separation characteristics, polymer-coated fiber-packed capillary columns have been prepared with different concentrations of HR-1 material. As described above, the polymeric coating process was carried out after packing 170 filaments of Zylon into the fused-silica capillary by a similar procedure for the preparation of conventional open-tubular capillary. Figure 4 shows the typical chromatograms for the separation of an n-alkane mixture on these polymer-coated fiber-packed capillaries prepared with different polymer concentrations from 1.0% to 20%. For comparison, the chromatogram obtained with a noncoated column in the same condition is also illustrated in the figure. Without the polymer coating onto the packed filaments, only limited sample loading capacity and retentivity could be obtained as reported previously.29 On increasing the polymer concentration, the retentivities for these analytes were increased, and wider peak widths were observed. It was also confirmed that an acceptable retentivity for the sample on the column with 1.0% coating could be obtained at lower column temperature and/or low column inlet pressure,
Figure 4. Chromatograms for the separation of n-alkane mixture (from C11H24 to C20H42) on polymer-coated fiber-packed capillary columns prepared with different polymer concentration. Conditions: fiber, Zylon (170 filaments); polymer (HR-1) concentration, (A) 0% (without coating), (B) 1.0%, (C) 5.0%, (D) 10%, and (E) 20%; column size, 0.32 mm i.d. × 2.0 m; column inlet pressure, 20 kPa; temperature program, 110 (1.5 min) to 260 °C at 3 deg/min; splitless injection. 5528 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
Table 2. Retention Factors (k)a,b for Various Compounds on Four Different Polymer-Coated Fiber-Packed Columns polymer coating
Figure 5. Effect of the fibrous materials and the polymeric coating on the retentivity for three aromatic compounds. (A) Fused-silica capillary of 0.53 mm i.d. × 1.0 m only; (B) fiber-packed fused-silica without polymer coating; and (C) fiber-packed fused-silica with polymer coating. Peaks: (a) naphthalene; (b) n-tridecane; and (c) biphenyl. Column for (B) and (C): fused-silica (0.53 mm i.d. × 1.0 m) packed with Zylon (330 filaments), and additionally HR-1 (5.0%) coating was made for column (C). Chromatographic conditions: column inlet pressure, 2.5 kPa; column temperature, 120 °C (isothermal); split injection with the ratio of (10:1). Other conditions are in the text.
however, with significant peak leading and an extended analysis time. Furthermore, polymer concentrations of more than 10% seem not to be appropriate in this separation, because of the overlapping peaks in the chromatogram. These results clearly demonstrated the contribution of the polymeric coating to the retention and the sample loading capacity. Similar results were also observed during the retentivity comparison with noncoated filaments as depicted in Figure 5. Comparing the retentivity on a parent fiber-packed column to that on the polymer-coated one, the contribution of the polymeric coating can be clearly observed, although a partial separation was obtained on the noncoated column as reported earlier.29 The retention values of three n-alkanes (C11H24, C12H26, and C13H28) were measured on the polymer-coated fiber-packed column and several conventional open-tubular columns of the same coating, where the linear velocity was maintained as the same at isothermal condition, and the retention factors were determined with the retention time of the solvent as the dead-time. The calculated retention factors for these three compounds on the polymer-coated fiber-packed capillary were about 50 and 100 times larger than that on open-tubular capillaries of the same length of 0.25 and 0.32 mm i.d., respectively, having the film thickness of 0.25 µm. Compared with the open-tubular column of 1.0 µm film thickness (0.25 mm i.d.), the coated fiber-packed column showed more retentivity by a factor of about 3. The above results clearly indicated that a good retentivity could be obtained, even for a short column length, on the polymer-coated fiber-packed capillary. Selectivities on Various Coating Materials. Upon successful results on the polymer coating by poly(dimethylsiloxane) to the packed filaments, several other polymeric materials have been also introduced, because the separation characteristics of the fiberpacked columns could be changed with a different type of coating. The retention data have been summarized in Table 2, where four typical stationary phases were included as the coating material. Figure 6 also shows the typical chromatograms for a standard mixture on fiber-packed columns coated with three different
analyte
HR-1
HR-52
HR-1701
HR-17
1-octanol 2,6-dimethylphenol 2,6-dimethylaniline naphthalene methyl-n-nonanoate 1-decanol n-tridecane biphenyl dicyclohexylamine
3.90 [1.9] 5.10 [2.3] 6.90 [2.0] 8.27 [1.4] 9.55 [2.0] 12.2 [2.5] 15.0 [1.7] 21.4 [1.5] 27.1 [2.0]
6.63 [2.2] 8.96 [1.9] 12.9 [2.2] 14.7 [1.8] 15.7 [2.4] 20.8 [2.1] 24.0 [2.0] 39.8 [2.0] 47.6 [1.9]
8.96 [1.8] 16.4 [1.9] 20.5 [1.9] 18.9 [2.0] 17.2 [1.9] 27.8 [2.2] 17.8 [2.4] 51.2 [2.1] 51.3 [1.5]
4.85 [2.2] 11.0 [2.0] 18.0 [1.8] 19.3 [2.4] 11.4 [1.8] 15.8 [1.9] 9.27 [2.5] 57.0 [2.4] 40.4 [1.8]
a Chromatographic conditions are the same as in Figure 4. b In square brackets %RSDs (n ) 3) were calculated for three different columns with the same coating, where the average values of the retention factors for three consecutive runs on each column were used for the calculation. The RSDs for the multiple injections on the same column were all less than 1.0% (n ) 3).
Figure 6. Different selectivities obtained with different types of polymeric coatings. (A) HR-52; (B) HR-1701; and (C) HR-17. Peaks: (a) naphthalene; (b) n-tridecane; and (c) biphenyl. Other chromatographic conditions are the same as in Figure 4.
polymeric stationary phases. The elution order of these three analytes was changed mainly due to the relatively polar nature of aromatic compounds, as observed on conventional open-tubular columns with the same polymeric coatings. The elution order obtained with nonpolymer-coated fiber-packed column (bare Zylon packed) is the same as HR-52 coated one (chromatogram A) in Figure 6, although the retentivitiy was quite limited as described above. These data clearly demonstrate that the selectivity of the fiber-packed columns can be changed based on the type of the surface coating, even if the same fibrous material is used as the support. Applications of Polymer-Coated Fiber-Packed Columns. Figure 7 illustrates the separation of four standard mixtures with different polymer-coated fiber-packed columns. Before these separation conditions have been optimized, the comparison of the separation performance for each mixture was made with various polymer-coated columns prepared by the fibrous materials listed in Table 1. Among the fiber-packed columns coated with HR-1, Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
5529
Figure 7. Typical chromatograms on polymer-coated fiber-packed capillary columns. Samples: (A) alkylbenzenes, C6H5C2H5-C6H5C10H21; (B) PAHs; (C) alcohols, C4H9OH-C8H17OH; and (D) alcohols, C6H13OH-C15H31OH. Conditions for (A): column, Technora-coated with HR-1 (5%); temperature program and inlet pressure, 40 (1.5 min) to 200 °C at 10 deg/min and 10 kPa. Conditions for (B), (C) and (D) are as follows: (B) Zylon-coated with HR-1 (5%), 60 °C (3.0 min) to 300 °C at 12 deg/min, 30 kPa; (C) Zylon-coated with HR-20M (5%), 30 °C to 150 °C at 15 deg/min, 10 kPa; and (D) Zylon-coated with HR-20M (5%), 50 °C to 180 °C at 5 deg/min, 10 kPa. Column, fusedsilica of 0.32 mm i.d. × 1.0 m. Injection was made with splitless mode. Other conditions are found in the text.
Technora support showed the best separation for alkylbenzenes, suggesting that a small contribution of the support material on the separation characteristics along with a major contribution of the polymeric coating thereon. On the other hand, Zylon support was found suitable for the separation of PAHs mixtures containing naphthalene, fluorene, phenanthrene, pyrene, and triphenylene. The elution order of PAHs is the same as conventional opentubular capillary columns prepared with the same polymeric liquid phase, for example, HR-1. The separation of analogous alcohol mixtures was also studied on introducing poly(ethylene glycol) (HR-20M) phase as the coating material. The results demonstrated an appropriate employment of the coating for the separation of alcohols, as expected from the similar trend in open-tubular columns. With the coating, a short fiber-packed capillary of only 1.0-m length shows good retentivity, without peak tailing and/or distortion. Since significant peak tailing was observed for the separation of alcohols on the nonpolymer-coated Zylon column, the dominant contribution of the surface coating to the separation has been demonstrated. Comparing with conventional open-tubular capillary columns of the same length, however, the improved retentivitiy can be partially attributed to the function of the fibrous material that allows more surface area of polymeric phase than open-tubular ones. Taking advantage of the excellent thermal stability of the fibers and a metal capillary, the column temperature can be elevated to more than 350 °C, as shown in Figure 8 for the separation of the alkanes mixture. The metal column was prepared using a similar procedure for fused-silica capillaries, except for the aging process before the separation. The aging temperature was set at higher temperature than the final temperature of the scheduled run. With this aging procedure quite stable baselines could be obtained for the temperature-programmed separation over the upper limit of conventional fused-silica capillaries having polyimide coating, 5530 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
Figure 8. Typical chromatogram for the separation of alkanes. Sample: alkane mixture (containing about C19H40 to C44H90) taken from a commercially available candle. Conditions: column, stainless steel capillary (0.25 mm i.d. × 1.0 m; without deactivation treatment; Shinwa Chemical Industries) packed with HR-1 (5.0%) coated Zylon (110 filaments); inlet pressure, 20 kPa; column temperature, programmed from 180 to 360 °C at 5 deg/min; splitless injection (0.5 µL) of the alkane mixture (1.0 wt/vol %) in dichloromethane. Other conditions are as described in the text.
where no measurable decrease in the separation performance was observed during the aging process. Therefore, one can conclude that the fiber-packed metallic capillary should have a wide applicability to the high-temperature separation of other complex mixtures containing low-volatile analytes. Based on the larger retentivity, a shorter column could be also developed for the fast separation. In contrast to the conventional open-tubular capillary, the wall effect from the inner surface might be significantly reduced by introducing the polymer-coated fibrous materials, especially for a shorter column length. Study for more extensive applications of the polymer-coated fiber-packed capillary as the extraction medium for trace amounts of organics in a complex sample matrix23,24 is currently underway in our laboratory along with the systematic comparison of other heat- and solvent-resistant filaments as the stationary phase and/ or the support material in other chromatographic techniques. CONCLUSIONS A polymer-coated fiber-packed capillary has been successfully developed as a novel format of capillary GC columns. With the coating by various polymeric materials on the surface of the packed filaments, the retentivity was significantly improved over a conventional open-tubular capillary of the same length. The excellent thermal stability of the fiber allowed a good separation performance over the normal operational temperature range in fused-silica capillary columns. Several experimental parameters for column preparation such as polymer film thickness on the packed filaments, a suitable combination of polymeric coating and fiber material should be further studied; the results in the present work suggest the future possible use of polymer-coated fibrous
materials as stationary phases in the temperature-programmed separation in various GC applications. ACKNOWLEDGMENT A part of this research was financially supported by Grant-inAid for Scientific Research ((B) No. 14340233) from The Japan Society for the Promotion of Science (JSPS), Grant-in-Aids for Young Scientists ((B) No. 13740421 and No. 15750066) from The Ministry of Education, Culture, Sports, Science and Technology (MEXT), and a Research Grant for Young Faculties from Toyo-
hashi University of Technology. The authors thank Dr. Tarek Agag, School of Materials Science, Toyohashi University of Technology, for his valuable help during the TGA measurements. The authors also acknowledge the financial support from The Tatematsu Foundation and The Ogasawara Foundation for the Promotion of Science and Engineering. Received for review February 6, 2003. Accepted July 11, 2003. AC030052H
Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
5531
