- Email: [email protected]
Single fiber-in-capillary annular column for gas chromatographic separation
Journal of Chromatography A, 1216 (2009) 3343–3348
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Single fiber-in-capillary annular column for gas chromatographic separation Pengzhi Li a , Zemin Xu a , Xiupei Yang a , Weiwen Bi a , Dan Xiao a,b,∗ , Martin M.F. Choi c,∗∗ a
College of Chemistry, Sichuan University, Chengdu 610064, China College of Chemical Engineering, Sichuan University, Chengdu 610065, China c Department of Chemistry, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China b
a r t i c l e
i n f o
Article history: Received 30 December 2008 Received in revised form 13 February 2009 Accepted 17 February 2009 Available online 21 February 2009 Keywords: Gas chromatography Annular column Tunable selectivity
a b s t r a c t A method for preparation of a stationary phase-adjustable column with in-column stationary phasecoated fused-silica fiber annular column was successfully developed. The surface of a 0.12 mm o.d. bare optical fiber was first coated with a stationary phase and then inserted into a fused-silica capillary (noncoated or coated) as an annular column for gas chromatographic study. The optical fiber and capillary were coated with polydimethylsiloxane (SE-30) and polyethylene glycol 20M (PEG-20M) as nonpolar and polar stationary phases, respectively. Among the investigated annular and open tubular columns, the PEG20M-coated fiber-in-PEG-20M-coated capillary annular column showed the highest column efficiency with a minimum plate height of 0.35 mm and an optimum gas velocity of 25 cm/s. When a SE-30/PEG20M-coated fiber-in-uncoated capillary annular column was applied to separate a 9-component complex mixture, the total analysis time was 5.3 min and the column length was 12 m. By contrast, when a SE-30coated fiber-in-PEG-20M-coated capillary annular column was used to separate the same 9-component mixture, the analysis time was reduced to 3.5 min and the column length was shortened by half to 6 m. Our results show that the stationary phase-coated fiber-in-stationary phase-coated capillary annular column is a better choice for gas chromatographic separation as it is more efficient and flexible. In addition, the proposed annular column design provides flexibility in using two or even more types of stationary phases to achieve optimal analytical separation. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Open tubular gas chromatography (GC) is the most common method for the analysis of volatile organic compound mixtures. However, there is always a growing need to develop columns or separation strategies with enhanced selectivity for the analysis of samples of complex mixtures. Although augmenting column length can enhance separation efficiency by generating larger number of theoretical plates, longer analysis time is resulted. A more convenient alternative is to apply narrower-bore and thinner-film column which can improve column efficiency and provide higher carrier gas velocity (). Unfortunately, this will reduce sample loading on the column resulting in narrower dynamic analyte working range and lower sensitivity of detection. Recently, considerable attention has been drawn to the use of two stationary phases in GC to enhance selectivity. Mixed packings each holding a single stationary phase [1,2], or a packing coated
∗ Corresponding author at: College of Chemical Engineering, Sichuan University, Chengdu 610065, China. Tel.: +86 28 8541 5029; fax: +86 28 8541 6029. ∗∗ Corresponding author. Tel.: +852 3411 7839; fax: +852 3411 7348. E-mail addresses: [email protected] (D. Xiao), [email protected] (M.M.F. Choi). 0021-9673/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2009.02.040
with mixed stationary phases (or mixed solvents) for GC columns have been reported [3,4]. However, there are very few studies of using mixed stationary phases for open tubular column (OTC) due to their low inter-solubility and poor film-formability especially when they have very different polarities [5,6]. To date a number of papers have reported the use of series-coupled ensembles [7–10] in two capillary columns with different stationary phases. Such column designs can eliminate the deficiencies associated with mixed stationary phases while preserving their unique advantages. Selectivity can also be adjusted more conveniently by controlling the carrier gas pressure at the junction of two series-coupled capillary columns [11–14]. However, the use of extra connectors increases slightly the dead-volume with a concomitant decrease in separation efficiency. Fused-silica fiber is widely applied in solid-phase microextraction [15,16] and other analytical applications [17–19]. The chemical inertness and thermostability of silica fiber make it a suitable support for coating extraction phase(s) including polydimethysiloxane, polyethylene glycol, etc. In the last few years, stationary phasecoated fibers in capillary were used in packed capillary GC [20–24]. A majority of their work involved packing large number of fibers (15 m o.d.) in a capillary column to improve the loading of stationary phase with the expense of separation efficiency and head column pressure.
3344
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
Similarly, other types of annular column design have been documented. For instances, annular column devices as suppressor in ion chromatography were employed to reduce dispersion and improve mass transfer [25–27] as compared to hollow fiber suppressor [28]. These devices have a substantially larger available membrane surface area per unit dispersion of an injected band and also exhibit superior performance in terms of better mass transfer and smaller band broadening. Other annular columns developed as chromatographic sensors were reported by Bruckner and Synovec [29]. Their emphasis was on the sensing aspect of the annular column but not on optimizing chromatographic separation. The polymer cladding (12-m thickness) of a polymer-clad optical fiber was employed as the stationary phase. However, this caused a large resistance to mass transfer in the stationary phase and resulted in considerable band broadening of the later eluting compounds. Their chromatographic sensors were also applied to liquid chromatographic separation [30,31]. In addition, fibers packed into a narrow-bore capillary for capillary electrochromatography (CEC) were studied and high separation efficiency could be achieved [32,33]. In this work, we propose a simple strategy of using a single stationary phase-coated fiber in a stationary phase-(un)coated capillary column for GC separation. The major attribute of our annular column design is that only one single fiber with larger o.d. (0.12 mm) was used to coat the stationary phase instead of many fibers of smaller o.d. (typically, 15 m) packed in the column. As such, this can simplify the annular column design and possibly produces longer annular column since our fiber is longer and more rigid. Secondly, fused-silica fiber is more chemically inert and thermostable than other polymer fibers. Thirdly, our annular column can accommodate a single fiber with various types of stationary phases or both the fiber and capillary coated with different stationary phases without mixing. Fourthly, no connector is required in contrast to series-coupled column; thus reducing the void volume. Finally, the selectivity of our annular column can be conveniently tuned by adjusting the length-ratio of the inserted series-coupled fibers, concentration-ratio of stationary phases and the fiber/capillary surfaces to be coated. To our knowledge, this is the first annular column based on a single stationary phase-coated fiber-in-stationary phase-coated capillary and has been successfully applied to GC separation of complex mixtures. 2. Experimental 2.1. Chemicals Fused-silica capillaries (0.20 mm i.d. and 0.25 mm i.d.) and optical fiber (0.12-mm core diameter) were obtained from Ruipu Chromatography Equipment (Hebei, China). Polyethylene glycol 20M (PEG-20M) was purchased from Guangzhou Chemicals (Guangdong, China). Trimethylchlorosilane was obtained from Kelong Chemicals (Sichuan, China). Polydimethylsiloxane (SE30) was from Sinopharm Chemicals Reagent (Shanghai, China). Acetone, dichloromethane and methanol were purchased from Chengdu Chemicals (Sichuan, China). All organic solvents and chemicals used were of analytical-reagent grade or above. 2.2. Apparatus Gas chromatography was performed on a Fuli model 9790 series GC equipped with a flame ionization detector (Zhejiang, China) and nitrogen as the carrier gas. Chromatograms were recorded on a Wuhao Information Technology WH-500 integrator (Shanghai, China). The injector temperature was set at 250 ◦ C and the detector temperature was 300 ◦ C.
2.3. Deactivation of fiber and capillary The fibers were soaked in dichloromethane for several minutes to remove the outer layer. (Noted that the fiber became fragile afterwards and had to be handled with care.) The fibers were cleansed sequentially with dichloromethane and deionized water, and then soaked in 1 M HCl for 30 min. Finally, the fibers were washed with deionized water again, dried under room temperature and were ready for further use. The fused-silica capillaries were sequentially rinsed with acetone and methanol, followed by a brief nitrogen purge for 10 min. The inner wall of the fused-silica capillary and the surface of the fiber were deactivated prior to coating of stationary phases. Firstly, the fiber was inserted into the capillary which was then filled with 25% (v/v) trimethylchlorosilane in pentane solution and both ends of the capillary were sealed. It was heated in a GC oven from 25 to 400 ◦ C at 5 ◦ C/min with a hold time of 300 min at 400 ◦ C. Afterwards, the column was cooled to room temperature and the sealed ends were cut open. The column was rinsed with dichloromethane for 5 min, followed by a brief nitrogen purge for 20 min. The fiber was withdrawn from the capillary and both the capillary and the fiber were ready for coating with stationary phases. A laboratory-made capillary filling device with a Fenghua model JY7134 vacuum pump (Shandong, China) was used to fill the capillary with stationary phase coating solution. 2.4. Preparation of single fiber-in-capillary annular column The deactivated fused-silica capillary was coated by static method with a dichloromethane solution of stationary phase. In this work, 2.0 and 4.0% (w/v) stationary phase solutions were used to prepare 1.0 and 2.0 m thick stationary phase films on capillary, respectively. The stationary phase was coated on the fiber by dipping and withdrawing it through a stationary phase solution (SE-30 or PEG-20M in dichloromethane) in a u-tube at a speed of 5 cm/s for five times. Then the stationary phase-coated fiber was dried under nitrogen overnight. A 2.5% (w/v) stationary phase solution would produce a 2-m stationary phase film on a fiber as determined by scanning electron microscopy (not shown). Similarly, a 1.25% (w/v) stationary phase solution could coat a 1-m stationary phase film on the fiber. The stationary phase-coated fiber was longitudinally inserted into an uncoated or stationary phase-coated capillary. Note that the fiber is fragile and must be handled carefully. In this way, a typical 1–10 m annular column can be prepared within 2 h depending on the length of the fiber required. When an annular column longer than 10 m is needed, a suction pump connected to the other end of the capillary will assist in inserting the fiber into the capillary. Slight rotation of the fiber helps the insertion process. Alternatively, two individual fibers can be inserted from each end of a capillary. Finally, all the annular columns were conditioned under a flow stream of nitrogen in a GC oven overnight before use. In this work, various schemes of single fiber-in-capillary annular columns were prepared and evaluated for their separation efficiency. The capillary could be uncoated or coated with stationary phase. The first scheme was a stationary phase-coated fiber positioned into an uncoated capillary. The second was two individual fibers, coated with two different stationary phases and then inserted into the opposite ends of an uncoated capillary. The third was a stationary phase-coated fiber loaded into a stationary phase-coated capillary. The fiber and capillary can be coated with the same or different stationary phases in order to tune the selectivity of separation as required. Fig. 1 displays the schematic constructions of these annular columns. SE-30 and PEG-20M were employed to coat the fiber or capillary. The following is some examples of our prepared annular columns: (a) SE-30-coated fiberin-uncoated capillary (SE-CF/C), (b) SE-30 and PEG-20M-coated
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
3345
Fig. 3. Van Deemter plots of n-undecane using (a) PEG-CF/C annular column (6 m × 0.20 mm, 2.0 m), (b) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), and (c) PEG-CF/PEG-CC annular column (6 m × 0.20 mm, 1.0 m). Oven temperature was 125 ◦ C. Fig. 1. Schematic diagrams of various types of single fiber-in-capillary annular columns prepared in this work. (a) Stationary phase-coated fiber-in-capillary, (b) dual stationary phase-coated fiber-in-capillary and (c) stationary phase-coated fiber-in-stationary phase-coated capillary annular columns. Stationary phases I and II can be of the same or different types.
fiber-in-uncoated capillary (SE/PEG-CF/C), (c) PEG-20M-coated fiber-in-PEG-20M-coated capillary (PEG-CF/PEG-CC), (d) SE-30coated fiber-in-PEG-20M-coated capillary (SE-CF/PEG-CC), and (e) PEG-20M-coated fiber-in-SE-30-coated capillary (PEG-CF/SE-CC). 3. Results and discussion 3.1. Effect of capillary column radius on separation In order to study the applicability of a stationary phase-coated annular column for GC separation, the SE-CF/C (1.0 m SE-30 coated on a 6 m × 0.12 mm o.d. fiber in a 0.25 mm i.d. uncoated capillary) was firstly applied to separate a mixture of n-alkanes (n-nonane, n-decane, n-undecane, and n-dodecane) and the chromatogram is depicted in Fig. 2a. All the n-alkanes are well separated from each other and the peak shape is good, demonstrating that fused-silica fiber is a good solid support for holding the stationary phase. Sil-
ica fiber possesses the advantages of chemical inertness and good thermostability as compared to polymer fibers. It can be coated with various types of stationary phases as required and positioned in a capillary column for GC separation. To evaluate the effect of capillary radius on separation, another SE-CF/C with 0.20 mm i.d. uncoated capillary was applied for the same separation and the chromatogram is shown in Fig. 2b. The separation between each n-alkane is further and better. All the n-alkanes show longer retention times, indicating that they have stronger interaction with the stationary phase coated on the fiber. The diffusion path of the analytes from the carrier gas to the stationary phase or vice versa is shorter with narrower capillary, allowing more efficient mass transfer and better separation, which is consistent with the chromatographic theory. Taking n-undecane as the probe compound, the column efficiency was determined by measuring the number of theoretical plates per meter. The 0.20 mm and 0.25 mm i.d. capillaries could obtain 3200 and 2900 m−1 , respectively. These results again demonstrate that a narrower annular space should enjoy lower dispersion, more efficient mass transfer and better separation efficiency which is consistent with other work in ion chromatography [25,28]. In essence, a stationary phasecoated fiber can be applied to GC separation and a narrower capillary will produce higher column efficiency. As such, 0.20 mm i.d. capillary was chosen for most of our annular column studies. 3.2. Separation efficiency of single fiber-in-capillary annular column
Fig. 2. Effect of column radius on separation performance. (a) SE-CF/C (6 m × 0.25 mm) and (b) SE-CF/C (6 m × 0.20 mm) annular columns. Sample is a mixture of n-alkanes. 1: n-nonane, 2: n-decane, 3: n-undecane, and 4: n-dodecane. Conditions: 100 ◦ C; linear carrier gas velocity 25 cm/s; and splitless injection.
In order to investigate the separation efficiency of the single fiber-in-capillary annular column, in this work, three types of GC annular column were evaluated and compared: (1) PEG-CF/C (6 m × 0.20 mm, 2.0 m), (2) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), and (3) PEG-CF/PEG-CC (6 m × 0.20 mm, 1.0 m). Van Deemter plots were performed to compare the H and of these columns. Fig. 3 shows the Van Deemter plots of plate height H against for the investigated columns using n-undecane as the probe compound. All the curves display typical Van Deemter behavior. The PEG-CF/C has larger H value than that of PEG-20M-coated OTC when the carrier gas velocity is higher than their optimum velocity of 20 cm/s. The mass transfer resistance term of the Van Deemter equation for PEG-CF/C is 0.014 mm s/cm and is slightly larger than that of PEG-20M-coated OTC with 0.010 mm s/cm. This is probably due to more solute diffusion in the moving mobile phase
3346
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
Fig. 4. Separation of a mixture of aromatic compounds on (a) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), (b) PEG-CF/C (6 m × 0.20 mm, 2.0 m), and (c) PEGCF/PEG-CC (6 m × 0.20 mm, 1.0 m) annular columns. Solutes 1: benzene, 2: toluene, 3: m-xylene, 4: p-xylene, and 5: o-xylene. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 70 to 90 ◦ C at 5 ◦ C/min; and splitless injection.
contributing to the plate height [34] with a fiber in an annular column. When both the fiber and capillary were coated with the stationary phase (i.e., PEG-CF/PEG-CC), the minimum plate height value is 0.35 mm which is smaller than 0.44 mm for PEG-CF/C and 0.41 mm for OTC; hence, the PEG-CF/PEG-CC annular column has the best column efficiency. This column design also has the highest optimum carrier gas velocity of 25 cm/s which can certainly speed up sample analysis. In addition, it is well known that thinner film and smaller void volume of OTC allow more efficient mass transfer and higher column efficiency but reduce sample loading. However, our PEG-CF/PEG-CC annular column can retain the same sample loading capacity even though thinner stationary phase films are coated on the fiber and the interior capillary surfaces. Furthermore, separation of aromatic compounds on our annular column is another example to demonstrate the advantages of PEG-CF/PEG-CC annular column. Columns coated with polyethylene glycol are commonly used for separating aromatic compounds, especially the position isomers of dimethylbenzene. The GC separation of benzene, toluene and isomers of dimethylbenzenes is depicted in Fig. 4. For both PEG-20M-coated OTC and PEG-CF/C annular column, solutes were only partially separated and the isomers of dimethylbenzene were co-eluted (Fig. 4a and b). When a PEG-CF/PEG-CC annular column was employed for the same separation, solutes turned to be better resolved as illustrated in Fig. 4c. Benzene, toluene and o-xylene were completely resolved while mxylene and p-xylene were partially separated. This result illustrates that the use of both stationary phase-coated fiber and capillary can improve the separation efficiency of a column, attributing to the thinner and larger surface of stationary phase and lower void volume which enhance the mass transfer of solute. Furthermore, the retention times of solutes on the PEG-CF/PEG-CC and PEG-CF/C annular columns were slightly longer than that of the OTC under the same separation conditions as the larger coating area allows stronger retention of solutes. 3.3. Dual stationary phase-coated fiber-in-capillary annular column Although the Van Deemter plots of H against (Fig. 2) and the separation of aromatic compounds have shown that the PEG-CF/C annular column is not as efficient as OTC, the stationary phasecoated fiber-in-capillary annular column is more advantageous
since two stationary phases can be used without pre-mixing and the selectivity of stationary phase can be tuned easily. As such, this allows flexibility of using various stationary phases simultaneously to improve separation. Fig. 5 displays the GC separation of a complex organic compound mixture on a SE-30-coated OTC, a PEG-20M-coated OTC, a SE/PEG-CF/C annular column, and a seriescoupled column of SE-30-coated OTC and PEG-20M-coated OTC. This complex mixture contained nine components of the three homologous series of n-alkanes, alcohols and esters. Only isoamylol, n-decane and n-undecane were baseline separated and other solutes were poorly resolved by the SE-30-coated OTC (Fig. 5a). A nearly complete separation of solutes was obtained by the PEG20M-coated OTC (Fig. 5b). When SE-30 and PEG-20M were coated on the fibers and inserted in an uncoated capillary (i.e., SE/PEG-CF/C annular column in Fig. 1b), they could act as dual stationary phase as depicted in the chromatogram of Fig. 5c. Most of the solutes were well resolved except 1-hexanol and n-undecane. It is easily observed from Fig. 5a–c that the SE/PEG-CF/C annular column possesses properties of both SE-30 and PEG-20M which results in a combined elution order based on SE-30 and PEG-20M. It is well understood that elution order is dependent on the type of stationary phase in the separation column. The elution order of the solutes on the SE-30-coated OTC is different from that on the PEG-20M-coated OTC (Fig. 5a and b). n-Propanol/isobutanol and isooctane/butyl acetate pairs were co-eluted on the SE-30-coated OTC (Fig. 5a) while n-propanol, butyl acetate and isobutanol were not completely resolved on the PEG-20M-coated OTC (Fig. 5b). By contrast, better separations of isooctane, n-propanol, isobutanol, and butyl acetate were achieved on the SE/PEG-CF/C annular column. In essence, the retention characteristics of these solutes on SE-30 and PEG-20M are maintained in the SE/PEG-CF/C annular column. Separation of the 9-component mixture was also applied to a SE-30 OTC (6 m × 0.20 mm, 1.0 m) series-coupled with a PEG-20M OTC (6 m × 0.20 mm, 1.0 m) as depicted in Fig. 5d. Good separation of the mixture was also achieved. The performance of our annular column is similar to the series-coupled column design [11–14]. However, our work demonstrates that selectivity and separation of solutes on a GC column can be tuned by placing fibers coated with
Fig. 5. Separation of a complex mixture of organic compounds on (a) SE-30-coated OTC, (b) PEG-20M-coated OTC, (c) SE/PEG-CF/C annular column (12 m × 0.20 mm, 1.0 m), and (d) SE-30-coated OTC (6 m × 0.20 mm, 1.0 m) series-coupled with PEG-20M-coated OTC (6 m × 0.20 mm, 1.0 m) with a pressure valve at the junction. Solutes 1: n-propanol, 2: isobutanol, 3: isoamylol, 4: isooctane, 5: butyl acetate, 6: amyl acetate, 7: 1-hexanol, 8: n-decane, and 9: n-undecane. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 65 to 110 ◦ C at 7 ◦ C/min; and splitless injection. For the series-coupled column, linear carrier gas velocity for SE-30-coated OTC and PEG-20M-coated OTC are 18 and 32 cm/s, respectively.
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
one or more types of stationary phases into a capillary. This allows more flexibility in using stationary phases; moreover, it is easier and more convenient to coat stationary phases on fibers than on OTCs. Finally, it is possible that the analytical separation of GC column can be greatly enhanced by merging our annular column design with the series-coupled column approach. 3.4. Stationary phase-coated fiber-in-stationary phase-coated capillary annular column Figs. 3 and 4 have shown that the stationary phase-coated fiberin-stationary phase-coated capillary annular column is a promising column design for efficient separation. Is it possible to reduce the column length but still maintain separation efficiency? To answer this question, two types of single fiber-in-capillary annular columns with dual stationary phase (i.e., PEG-CF/SE-CC and SE-CF/PEG-CC) were prepared and applied to the same 9-component mixture to evaluate their separation efficiency. The columns were also halved in length, i.e., 6 m. In theory, resolution is a function of the square root of column length. Decreasing the column length from 12 to 6 m will lower the resolution by a factor of 1.4. Fig. 6a displays the GC separation of the complex mixture on the PEG-CF/SE-CC annular column. The solutes were almost completely separated except the n-propanol/isooctane pair. This demonstrates that shorter column and dual stationary phase coated on fiber and capillary is a promising strategy to improve separation. Furthermore, the analysis time can also be shortened. Fig. 6b depicts the GC separation of the same complex mixture on the SE-CF/PEG-CC annular column. The solutes were well separated from each other. Again, this demonstrates the success of single fiber-in-capillary annular column with dual stationary phase. The selectivity changes with the phase-ratio of the two stationary phases. The chromatograms of the two annular columns (Fig. 6a and b) show that retention and selectivity of solutes changed when the stationary phase on the fiber and capillary are swop. For example, n-propanol/isobutanol/isooctane/butyl acetate solutes were better separated on the SE-CF/PEG-CC than on PEG-CF/SE-CC annular columns. In this work, the dual stationary phase-coated fiber in an uncoated capillary (stationary phase/stationary phase-CF/C) and stationary phase-coated fiber-in-stationary phase-coated capillary (stationary phase-CF/stationary phase-CC) were the two different column designs of single fiber-in-capillary annular column using
3347
dual stationary phase. The main difference between the two designs was the positioning of the stationary phases on the fiber and/or capillary. In the stationary phase/stationary phase-CF/C annular column, both stationary phases are series-coupled on a single fiber. The stationary phase-coated fiber and stationary phase-coated capillary are parallel in the stationary phase-CF/stationary phase-CC annular column. It is also interesting to compare the performances of the SE/PEG-CF/C with SE-CF/PEG-CC annular columns. As shown in Figs. 5c and 6b, the elution orders of the solutes are very similar. The SE-CF/PEG-CC is superior to SE/PEG-CF/C annular column in terms of separation efficiency and column design. First, the SECF/PEG-CC annular column has a shorter analysis time (3.5 min) than the SE/PEG-CF/C annular column (5.3 min). Second, 1-hexanol and n-undecane were much better resolved on the SE-CF/PEG-CC annular column. Third, the column length is shorter, i.e., 6-m long SE-CF/PEG-CC versus 12-m long SE/PEG-CF/C annular columns. In brief, shorter column has the advantages of lower pressure drop and shorter sample analysis time. Our results show that the stationary phase-CF/stationary phase-CC annular column is a better choice for GC separation as it is more efficient and flexible. 4. Conclusion In this work, a novel stationary phase-coated fiber-in-stationary phase-(un)coated capillary annular column was demonstrated and its performance was evaluated for the first time in GC. Only one single fiber with larger o.d. (0.12 mm) was used to coat the stationary phase as compared to others using many fibers of smaller o.d. (typically, 15 m) packed in the column. Our silica fiber was proved to be a good support material for stationary phase. Various types of stationary phases can be coated on a single fiber or both fiber and capillary without stationary phase mixing according to the analysis required. The inserted stationary phase-coated fiber can decrease void volume, facilitate faster mass transfer, and thus improve column efficiency. Selectivity can be adjusted by using various stationary phase-coated fiber-in-stationary phase-(un)coated capillary annular columns. The stationary phase-CF/stationary phase-CC design provides feasibility in tuning the selectivity of the separated solutes. Finally, the separation of any complex mixtures can be realized by optimizing the types, concentration and coating length of stationary phases, or even introducing a third stationary phase. This work is currently in progress in our laboratory. Lastly, more work can also be pursued by incorporating our annular column with the series-coupled column design to further enhance the separation power of a GC column. But it should be borne in mind that although longer annular columns can provide more theoretical plates, the total analysis time will be lengthened. Acknowledgements Financial support from National Natural Science foundation of China (20575042, 20775050) and Education Ministry of China (105141) are gratefully acknowledged. References [1] [2] [3] [4]
Fig. 6. Separation of a complex mixture of organic compounds on (a) PEG-CF/SECC and (b) SE-CF/PEG-CC (6 m × 0.20 mm, 1.0 m) annular columns. Solutes 1: npropanol, 2: isobutanol, 3: isoamylol, 4: isooctane, 5: butyl acetate, 6: amyl acetate, 7: 1-hexanol, 8: n-decane, and 9: n-undecane. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 65 to 90 ◦ C at 7 ◦ C/min; and splitless injection.
[5] [6] [7] [8] [9] [10]
C.F. Chien, R.J. Laub, M.M. Kopecni, Anal. Chem. 52 (1980) 1402. J.H. Purnell, M.H. Wattan, J. Chromatogr. 555 (1991) 173. R.J. Laub, J.H. Purnell, J. Chromatogr. 112 (1975) 71. J. Sanz, I. Martínez-Castro, G. Reglero, M.D. Cabezudo, Anal. Chim. Acta 194 (1987) 91. A. Ruderisch, J. Pfeiffer, V. Schurig, J. Chromatogr. A 994 (2003) 127. X.Y. Shi, Y.Q. Zhang, R.N. Fu, Anal. Chim. Acta 424 (2000) 271. J.H. Purnell, P.S. Williams, J. Chromatogr. 292 (1984) 197. E. Benická, J. Krupˇcík, D. Repka, P. Kuljovsky, R.E. Kaiser, Anal. Chem. 62 (1990) 985. A. Jayatilaka, S.K. Poole, C.F. Poole, T.M.P. Chichila, Anal. Chim. Acta 302 (1995) 147. J.M. Sanchez, R.D. Sacks, Anal. Chem. 75 (2003) 2231.
3348 [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
H. Smith, R.D. Sacks, Anal. Chem. 69 (1997) 5159. A.J. Grall, E.T. Zellers, R.D. Sacks, Environ. Sci. Technol. 35 (2001) 163. T. Veriotti, R.D. Sacks, Anal. Chem. 73 (2001) 3045. J. Whiting, R.D. Sacks, Anal. Chem. 74 (2002) 246. C.L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145. C. Dietz, J. Sanz, C. Cámara, J. Chromatogr. A 1103 (2006) 183. A.L. Lopes, F. Augusto, J. Chromatogr. A 1056 (2004) 13. M. Berg, J. Bolotin, T.B. Hofstetter, Anal. Chem. 79 (2007) 2386. F. Hernández, T. Portolés, E. Pitarch, F. López, J. Anal. Chem. 79 (2007) 9494. Y. Saito, M. Imaizumi, K. Nakata, T. Takeichi, K. Kotera, H. Wada, K. Jinno, J. Microcol. Sep. 13 (2001) 259. [21] Y. Saito, A. Tahara, M. Imaizumi, T. Takeichi, H. Wada, K. Jinno, Anal. Chem. 75 (2003) 5525. [22] Y. Saito, M. Ogawa, M. Imaizumi, K. Ban, A. Abe, T. Takeichi, H. Wada, K. Jinno, Anal. Bioanal. Chem. 382 (2005) 825. [23] Y. Saito, M. Ogawa, M. Imaizumi, K. Ban, A. Abe, T. Takeichi, H. Wada, K. Jinno, J. Chromatogr. Sci. 43 (2005) 536.
[24] M. Ogawa, Y. Saito, M. Imaizumi, H. Wada, K. Jinno, Chromatographia 63 (2006) 459. [25] P.K. Dasgupta, Anal. Chem. 56 (1984) 96. [26] P.K. Dasgupta, Anal. Chem. 56 (1984) 103. [27] P.K. Dasgupta, R.Q. Bligh, M.A. Mercurio, Anal. Chem. 57 (1985) 484. [28] T.S. Stevens, J.C. Davis, H. Small, Anal. Chem. 53 (1981) 1488. [29] C.A. Bruckner, R.E. Synovec, Talanta 43 (1996) 901. [30] R.E. Synovec, A.W. Sulya, L.W. Burgess, M.D. Foster, C.A. Bruckner, Anal. Chem. 67 (1995) 473. [31] M.D. Foster, R.E. Synovec, Anal. Chem. 68 (1996) 1456. [32] K. Jinno, H. Watanabe, Y. Saito, T. Takeichi, Electrophoresis 22 (2001) 3371. [33] Z.-G. Shi, W. Fang, Y.-Q. Feng, J. Liq. Chromatogr. Rel. Technol. 31 (2008) 3094. [34] C.F. Poole, The Essence of Chromatography, Elsevier, Amsterdam, 2003, p. 32.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Single fiber-in-capillary annular column for gas chromatographic separation Pengzhi Li a , Zemin Xu a , Xiupei Yang a , Weiwen Bi a , Dan Xiao a,b,∗ , Martin M.F. Choi c,∗∗ a
College of Chemistry, Sichuan University, Chengdu 610064, China College of Chemical Engineering, Sichuan University, Chengdu 610065, China c Department of Chemistry, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China b
a r t i c l e
i n f o
Article history: Received 30 December 2008 Received in revised form 13 February 2009 Accepted 17 February 2009 Available online 21 February 2009 Keywords: Gas chromatography Annular column Tunable selectivity
a b s t r a c t A method for preparation of a stationary phase-adjustable column with in-column stationary phasecoated fused-silica fiber annular column was successfully developed. The surface of a 0.12 mm o.d. bare optical fiber was first coated with a stationary phase and then inserted into a fused-silica capillary (noncoated or coated) as an annular column for gas chromatographic study. The optical fiber and capillary were coated with polydimethylsiloxane (SE-30) and polyethylene glycol 20M (PEG-20M) as nonpolar and polar stationary phases, respectively. Among the investigated annular and open tubular columns, the PEG20M-coated fiber-in-PEG-20M-coated capillary annular column showed the highest column efficiency with a minimum plate height of 0.35 mm and an optimum gas velocity of 25 cm/s. When a SE-30/PEG20M-coated fiber-in-uncoated capillary annular column was applied to separate a 9-component complex mixture, the total analysis time was 5.3 min and the column length was 12 m. By contrast, when a SE-30coated fiber-in-PEG-20M-coated capillary annular column was used to separate the same 9-component mixture, the analysis time was reduced to 3.5 min and the column length was shortened by half to 6 m. Our results show that the stationary phase-coated fiber-in-stationary phase-coated capillary annular column is a better choice for gas chromatographic separation as it is more efficient and flexible. In addition, the proposed annular column design provides flexibility in using two or even more types of stationary phases to achieve optimal analytical separation. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Open tubular gas chromatography (GC) is the most common method for the analysis of volatile organic compound mixtures. However, there is always a growing need to develop columns or separation strategies with enhanced selectivity for the analysis of samples of complex mixtures. Although augmenting column length can enhance separation efficiency by generating larger number of theoretical plates, longer analysis time is resulted. A more convenient alternative is to apply narrower-bore and thinner-film column which can improve column efficiency and provide higher carrier gas velocity (). Unfortunately, this will reduce sample loading on the column resulting in narrower dynamic analyte working range and lower sensitivity of detection. Recently, considerable attention has been drawn to the use of two stationary phases in GC to enhance selectivity. Mixed packings each holding a single stationary phase [1,2], or a packing coated
∗ Corresponding author at: College of Chemical Engineering, Sichuan University, Chengdu 610065, China. Tel.: +86 28 8541 5029; fax: +86 28 8541 6029. ∗∗ Corresponding author. Tel.: +852 3411 7839; fax: +852 3411 7348. E-mail addresses: [email protected] (D. Xiao), [email protected] (M.M.F. Choi). 0021-9673/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2009.02.040
with mixed stationary phases (or mixed solvents) for GC columns have been reported [3,4]. However, there are very few studies of using mixed stationary phases for open tubular column (OTC) due to their low inter-solubility and poor film-formability especially when they have very different polarities [5,6]. To date a number of papers have reported the use of series-coupled ensembles [7–10] in two capillary columns with different stationary phases. Such column designs can eliminate the deficiencies associated with mixed stationary phases while preserving their unique advantages. Selectivity can also be adjusted more conveniently by controlling the carrier gas pressure at the junction of two series-coupled capillary columns [11–14]. However, the use of extra connectors increases slightly the dead-volume with a concomitant decrease in separation efficiency. Fused-silica fiber is widely applied in solid-phase microextraction [15,16] and other analytical applications [17–19]. The chemical inertness and thermostability of silica fiber make it a suitable support for coating extraction phase(s) including polydimethysiloxane, polyethylene glycol, etc. In the last few years, stationary phasecoated fibers in capillary were used in packed capillary GC [20–24]. A majority of their work involved packing large number of fibers (15 m o.d.) in a capillary column to improve the loading of stationary phase with the expense of separation efficiency and head column pressure.
3344
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
Similarly, other types of annular column design have been documented. For instances, annular column devices as suppressor in ion chromatography were employed to reduce dispersion and improve mass transfer [25–27] as compared to hollow fiber suppressor [28]. These devices have a substantially larger available membrane surface area per unit dispersion of an injected band and also exhibit superior performance in terms of better mass transfer and smaller band broadening. Other annular columns developed as chromatographic sensors were reported by Bruckner and Synovec [29]. Their emphasis was on the sensing aspect of the annular column but not on optimizing chromatographic separation. The polymer cladding (12-m thickness) of a polymer-clad optical fiber was employed as the stationary phase. However, this caused a large resistance to mass transfer in the stationary phase and resulted in considerable band broadening of the later eluting compounds. Their chromatographic sensors were also applied to liquid chromatographic separation [30,31]. In addition, fibers packed into a narrow-bore capillary for capillary electrochromatography (CEC) were studied and high separation efficiency could be achieved [32,33]. In this work, we propose a simple strategy of using a single stationary phase-coated fiber in a stationary phase-(un)coated capillary column for GC separation. The major attribute of our annular column design is that only one single fiber with larger o.d. (0.12 mm) was used to coat the stationary phase instead of many fibers of smaller o.d. (typically, 15 m) packed in the column. As such, this can simplify the annular column design and possibly produces longer annular column since our fiber is longer and more rigid. Secondly, fused-silica fiber is more chemically inert and thermostable than other polymer fibers. Thirdly, our annular column can accommodate a single fiber with various types of stationary phases or both the fiber and capillary coated with different stationary phases without mixing. Fourthly, no connector is required in contrast to series-coupled column; thus reducing the void volume. Finally, the selectivity of our annular column can be conveniently tuned by adjusting the length-ratio of the inserted series-coupled fibers, concentration-ratio of stationary phases and the fiber/capillary surfaces to be coated. To our knowledge, this is the first annular column based on a single stationary phase-coated fiber-in-stationary phase-coated capillary and has been successfully applied to GC separation of complex mixtures. 2. Experimental 2.1. Chemicals Fused-silica capillaries (0.20 mm i.d. and 0.25 mm i.d.) and optical fiber (0.12-mm core diameter) were obtained from Ruipu Chromatography Equipment (Hebei, China). Polyethylene glycol 20M (PEG-20M) was purchased from Guangzhou Chemicals (Guangdong, China). Trimethylchlorosilane was obtained from Kelong Chemicals (Sichuan, China). Polydimethylsiloxane (SE30) was from Sinopharm Chemicals Reagent (Shanghai, China). Acetone, dichloromethane and methanol were purchased from Chengdu Chemicals (Sichuan, China). All organic solvents and chemicals used were of analytical-reagent grade or above. 2.2. Apparatus Gas chromatography was performed on a Fuli model 9790 series GC equipped with a flame ionization detector (Zhejiang, China) and nitrogen as the carrier gas. Chromatograms were recorded on a Wuhao Information Technology WH-500 integrator (Shanghai, China). The injector temperature was set at 250 ◦ C and the detector temperature was 300 ◦ C.
2.3. Deactivation of fiber and capillary The fibers were soaked in dichloromethane for several minutes to remove the outer layer. (Noted that the fiber became fragile afterwards and had to be handled with care.) The fibers were cleansed sequentially with dichloromethane and deionized water, and then soaked in 1 M HCl for 30 min. Finally, the fibers were washed with deionized water again, dried under room temperature and were ready for further use. The fused-silica capillaries were sequentially rinsed with acetone and methanol, followed by a brief nitrogen purge for 10 min. The inner wall of the fused-silica capillary and the surface of the fiber were deactivated prior to coating of stationary phases. Firstly, the fiber was inserted into the capillary which was then filled with 25% (v/v) trimethylchlorosilane in pentane solution and both ends of the capillary were sealed. It was heated in a GC oven from 25 to 400 ◦ C at 5 ◦ C/min with a hold time of 300 min at 400 ◦ C. Afterwards, the column was cooled to room temperature and the sealed ends were cut open. The column was rinsed with dichloromethane for 5 min, followed by a brief nitrogen purge for 20 min. The fiber was withdrawn from the capillary and both the capillary and the fiber were ready for coating with stationary phases. A laboratory-made capillary filling device with a Fenghua model JY7134 vacuum pump (Shandong, China) was used to fill the capillary with stationary phase coating solution. 2.4. Preparation of single fiber-in-capillary annular column The deactivated fused-silica capillary was coated by static method with a dichloromethane solution of stationary phase. In this work, 2.0 and 4.0% (w/v) stationary phase solutions were used to prepare 1.0 and 2.0 m thick stationary phase films on capillary, respectively. The stationary phase was coated on the fiber by dipping and withdrawing it through a stationary phase solution (SE-30 or PEG-20M in dichloromethane) in a u-tube at a speed of 5 cm/s for five times. Then the stationary phase-coated fiber was dried under nitrogen overnight. A 2.5% (w/v) stationary phase solution would produce a 2-m stationary phase film on a fiber as determined by scanning electron microscopy (not shown). Similarly, a 1.25% (w/v) stationary phase solution could coat a 1-m stationary phase film on the fiber. The stationary phase-coated fiber was longitudinally inserted into an uncoated or stationary phase-coated capillary. Note that the fiber is fragile and must be handled carefully. In this way, a typical 1–10 m annular column can be prepared within 2 h depending on the length of the fiber required. When an annular column longer than 10 m is needed, a suction pump connected to the other end of the capillary will assist in inserting the fiber into the capillary. Slight rotation of the fiber helps the insertion process. Alternatively, two individual fibers can be inserted from each end of a capillary. Finally, all the annular columns were conditioned under a flow stream of nitrogen in a GC oven overnight before use. In this work, various schemes of single fiber-in-capillary annular columns were prepared and evaluated for their separation efficiency. The capillary could be uncoated or coated with stationary phase. The first scheme was a stationary phase-coated fiber positioned into an uncoated capillary. The second was two individual fibers, coated with two different stationary phases and then inserted into the opposite ends of an uncoated capillary. The third was a stationary phase-coated fiber loaded into a stationary phase-coated capillary. The fiber and capillary can be coated with the same or different stationary phases in order to tune the selectivity of separation as required. Fig. 1 displays the schematic constructions of these annular columns. SE-30 and PEG-20M were employed to coat the fiber or capillary. The following is some examples of our prepared annular columns: (a) SE-30-coated fiberin-uncoated capillary (SE-CF/C), (b) SE-30 and PEG-20M-coated
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
3345
Fig. 3. Van Deemter plots of n-undecane using (a) PEG-CF/C annular column (6 m × 0.20 mm, 2.0 m), (b) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), and (c) PEG-CF/PEG-CC annular column (6 m × 0.20 mm, 1.0 m). Oven temperature was 125 ◦ C. Fig. 1. Schematic diagrams of various types of single fiber-in-capillary annular columns prepared in this work. (a) Stationary phase-coated fiber-in-capillary, (b) dual stationary phase-coated fiber-in-capillary and (c) stationary phase-coated fiber-in-stationary phase-coated capillary annular columns. Stationary phases I and II can be of the same or different types.
fiber-in-uncoated capillary (SE/PEG-CF/C), (c) PEG-20M-coated fiber-in-PEG-20M-coated capillary (PEG-CF/PEG-CC), (d) SE-30coated fiber-in-PEG-20M-coated capillary (SE-CF/PEG-CC), and (e) PEG-20M-coated fiber-in-SE-30-coated capillary (PEG-CF/SE-CC). 3. Results and discussion 3.1. Effect of capillary column radius on separation In order to study the applicability of a stationary phase-coated annular column for GC separation, the SE-CF/C (1.0 m SE-30 coated on a 6 m × 0.12 mm o.d. fiber in a 0.25 mm i.d. uncoated capillary) was firstly applied to separate a mixture of n-alkanes (n-nonane, n-decane, n-undecane, and n-dodecane) and the chromatogram is depicted in Fig. 2a. All the n-alkanes are well separated from each other and the peak shape is good, demonstrating that fused-silica fiber is a good solid support for holding the stationary phase. Sil-
ica fiber possesses the advantages of chemical inertness and good thermostability as compared to polymer fibers. It can be coated with various types of stationary phases as required and positioned in a capillary column for GC separation. To evaluate the effect of capillary radius on separation, another SE-CF/C with 0.20 mm i.d. uncoated capillary was applied for the same separation and the chromatogram is shown in Fig. 2b. The separation between each n-alkane is further and better. All the n-alkanes show longer retention times, indicating that they have stronger interaction with the stationary phase coated on the fiber. The diffusion path of the analytes from the carrier gas to the stationary phase or vice versa is shorter with narrower capillary, allowing more efficient mass transfer and better separation, which is consistent with the chromatographic theory. Taking n-undecane as the probe compound, the column efficiency was determined by measuring the number of theoretical plates per meter. The 0.20 mm and 0.25 mm i.d. capillaries could obtain 3200 and 2900 m−1 , respectively. These results again demonstrate that a narrower annular space should enjoy lower dispersion, more efficient mass transfer and better separation efficiency which is consistent with other work in ion chromatography [25,28]. In essence, a stationary phasecoated fiber can be applied to GC separation and a narrower capillary will produce higher column efficiency. As such, 0.20 mm i.d. capillary was chosen for most of our annular column studies. 3.2. Separation efficiency of single fiber-in-capillary annular column
Fig. 2. Effect of column radius on separation performance. (a) SE-CF/C (6 m × 0.25 mm) and (b) SE-CF/C (6 m × 0.20 mm) annular columns. Sample is a mixture of n-alkanes. 1: n-nonane, 2: n-decane, 3: n-undecane, and 4: n-dodecane. Conditions: 100 ◦ C; linear carrier gas velocity 25 cm/s; and splitless injection.
In order to investigate the separation efficiency of the single fiber-in-capillary annular column, in this work, three types of GC annular column were evaluated and compared: (1) PEG-CF/C (6 m × 0.20 mm, 2.0 m), (2) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), and (3) PEG-CF/PEG-CC (6 m × 0.20 mm, 1.0 m). Van Deemter plots were performed to compare the H and of these columns. Fig. 3 shows the Van Deemter plots of plate height H against for the investigated columns using n-undecane as the probe compound. All the curves display typical Van Deemter behavior. The PEG-CF/C has larger H value than that of PEG-20M-coated OTC when the carrier gas velocity is higher than their optimum velocity of 20 cm/s. The mass transfer resistance term of the Van Deemter equation for PEG-CF/C is 0.014 mm s/cm and is slightly larger than that of PEG-20M-coated OTC with 0.010 mm s/cm. This is probably due to more solute diffusion in the moving mobile phase
3346
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
Fig. 4. Separation of a mixture of aromatic compounds on (a) PEG-20M-coated OTC (6 m × 0.20 mm, 2.0 m), (b) PEG-CF/C (6 m × 0.20 mm, 2.0 m), and (c) PEGCF/PEG-CC (6 m × 0.20 mm, 1.0 m) annular columns. Solutes 1: benzene, 2: toluene, 3: m-xylene, 4: p-xylene, and 5: o-xylene. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 70 to 90 ◦ C at 5 ◦ C/min; and splitless injection.
contributing to the plate height [34] with a fiber in an annular column. When both the fiber and capillary were coated with the stationary phase (i.e., PEG-CF/PEG-CC), the minimum plate height value is 0.35 mm which is smaller than 0.44 mm for PEG-CF/C and 0.41 mm for OTC; hence, the PEG-CF/PEG-CC annular column has the best column efficiency. This column design also has the highest optimum carrier gas velocity of 25 cm/s which can certainly speed up sample analysis. In addition, it is well known that thinner film and smaller void volume of OTC allow more efficient mass transfer and higher column efficiency but reduce sample loading. However, our PEG-CF/PEG-CC annular column can retain the same sample loading capacity even though thinner stationary phase films are coated on the fiber and the interior capillary surfaces. Furthermore, separation of aromatic compounds on our annular column is another example to demonstrate the advantages of PEG-CF/PEG-CC annular column. Columns coated with polyethylene glycol are commonly used for separating aromatic compounds, especially the position isomers of dimethylbenzene. The GC separation of benzene, toluene and isomers of dimethylbenzenes is depicted in Fig. 4. For both PEG-20M-coated OTC and PEG-CF/C annular column, solutes were only partially separated and the isomers of dimethylbenzene were co-eluted (Fig. 4a and b). When a PEG-CF/PEG-CC annular column was employed for the same separation, solutes turned to be better resolved as illustrated in Fig. 4c. Benzene, toluene and o-xylene were completely resolved while mxylene and p-xylene were partially separated. This result illustrates that the use of both stationary phase-coated fiber and capillary can improve the separation efficiency of a column, attributing to the thinner and larger surface of stationary phase and lower void volume which enhance the mass transfer of solute. Furthermore, the retention times of solutes on the PEG-CF/PEG-CC and PEG-CF/C annular columns were slightly longer than that of the OTC under the same separation conditions as the larger coating area allows stronger retention of solutes. 3.3. Dual stationary phase-coated fiber-in-capillary annular column Although the Van Deemter plots of H against (Fig. 2) and the separation of aromatic compounds have shown that the PEG-CF/C annular column is not as efficient as OTC, the stationary phasecoated fiber-in-capillary annular column is more advantageous
since two stationary phases can be used without pre-mixing and the selectivity of stationary phase can be tuned easily. As such, this allows flexibility of using various stationary phases simultaneously to improve separation. Fig. 5 displays the GC separation of a complex organic compound mixture on a SE-30-coated OTC, a PEG-20M-coated OTC, a SE/PEG-CF/C annular column, and a seriescoupled column of SE-30-coated OTC and PEG-20M-coated OTC. This complex mixture contained nine components of the three homologous series of n-alkanes, alcohols and esters. Only isoamylol, n-decane and n-undecane were baseline separated and other solutes were poorly resolved by the SE-30-coated OTC (Fig. 5a). A nearly complete separation of solutes was obtained by the PEG20M-coated OTC (Fig. 5b). When SE-30 and PEG-20M were coated on the fibers and inserted in an uncoated capillary (i.e., SE/PEG-CF/C annular column in Fig. 1b), they could act as dual stationary phase as depicted in the chromatogram of Fig. 5c. Most of the solutes were well resolved except 1-hexanol and n-undecane. It is easily observed from Fig. 5a–c that the SE/PEG-CF/C annular column possesses properties of both SE-30 and PEG-20M which results in a combined elution order based on SE-30 and PEG-20M. It is well understood that elution order is dependent on the type of stationary phase in the separation column. The elution order of the solutes on the SE-30-coated OTC is different from that on the PEG-20M-coated OTC (Fig. 5a and b). n-Propanol/isobutanol and isooctane/butyl acetate pairs were co-eluted on the SE-30-coated OTC (Fig. 5a) while n-propanol, butyl acetate and isobutanol were not completely resolved on the PEG-20M-coated OTC (Fig. 5b). By contrast, better separations of isooctane, n-propanol, isobutanol, and butyl acetate were achieved on the SE/PEG-CF/C annular column. In essence, the retention characteristics of these solutes on SE-30 and PEG-20M are maintained in the SE/PEG-CF/C annular column. Separation of the 9-component mixture was also applied to a SE-30 OTC (6 m × 0.20 mm, 1.0 m) series-coupled with a PEG-20M OTC (6 m × 0.20 mm, 1.0 m) as depicted in Fig. 5d. Good separation of the mixture was also achieved. The performance of our annular column is similar to the series-coupled column design [11–14]. However, our work demonstrates that selectivity and separation of solutes on a GC column can be tuned by placing fibers coated with
Fig. 5. Separation of a complex mixture of organic compounds on (a) SE-30-coated OTC, (b) PEG-20M-coated OTC, (c) SE/PEG-CF/C annular column (12 m × 0.20 mm, 1.0 m), and (d) SE-30-coated OTC (6 m × 0.20 mm, 1.0 m) series-coupled with PEG-20M-coated OTC (6 m × 0.20 mm, 1.0 m) with a pressure valve at the junction. Solutes 1: n-propanol, 2: isobutanol, 3: isoamylol, 4: isooctane, 5: butyl acetate, 6: amyl acetate, 7: 1-hexanol, 8: n-decane, and 9: n-undecane. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 65 to 110 ◦ C at 7 ◦ C/min; and splitless injection. For the series-coupled column, linear carrier gas velocity for SE-30-coated OTC and PEG-20M-coated OTC are 18 and 32 cm/s, respectively.
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
one or more types of stationary phases into a capillary. This allows more flexibility in using stationary phases; moreover, it is easier and more convenient to coat stationary phases on fibers than on OTCs. Finally, it is possible that the analytical separation of GC column can be greatly enhanced by merging our annular column design with the series-coupled column approach. 3.4. Stationary phase-coated fiber-in-stationary phase-coated capillary annular column Figs. 3 and 4 have shown that the stationary phase-coated fiberin-stationary phase-coated capillary annular column is a promising column design for efficient separation. Is it possible to reduce the column length but still maintain separation efficiency? To answer this question, two types of single fiber-in-capillary annular columns with dual stationary phase (i.e., PEG-CF/SE-CC and SE-CF/PEG-CC) were prepared and applied to the same 9-component mixture to evaluate their separation efficiency. The columns were also halved in length, i.e., 6 m. In theory, resolution is a function of the square root of column length. Decreasing the column length from 12 to 6 m will lower the resolution by a factor of 1.4. Fig. 6a displays the GC separation of the complex mixture on the PEG-CF/SE-CC annular column. The solutes were almost completely separated except the n-propanol/isooctane pair. This demonstrates that shorter column and dual stationary phase coated on fiber and capillary is a promising strategy to improve separation. Furthermore, the analysis time can also be shortened. Fig. 6b depicts the GC separation of the same complex mixture on the SE-CF/PEG-CC annular column. The solutes were well separated from each other. Again, this demonstrates the success of single fiber-in-capillary annular column with dual stationary phase. The selectivity changes with the phase-ratio of the two stationary phases. The chromatograms of the two annular columns (Fig. 6a and b) show that retention and selectivity of solutes changed when the stationary phase on the fiber and capillary are swop. For example, n-propanol/isobutanol/isooctane/butyl acetate solutes were better separated on the SE-CF/PEG-CC than on PEG-CF/SE-CC annular columns. In this work, the dual stationary phase-coated fiber in an uncoated capillary (stationary phase/stationary phase-CF/C) and stationary phase-coated fiber-in-stationary phase-coated capillary (stationary phase-CF/stationary phase-CC) were the two different column designs of single fiber-in-capillary annular column using
3347
dual stationary phase. The main difference between the two designs was the positioning of the stationary phases on the fiber and/or capillary. In the stationary phase/stationary phase-CF/C annular column, both stationary phases are series-coupled on a single fiber. The stationary phase-coated fiber and stationary phase-coated capillary are parallel in the stationary phase-CF/stationary phase-CC annular column. It is also interesting to compare the performances of the SE/PEG-CF/C with SE-CF/PEG-CC annular columns. As shown in Figs. 5c and 6b, the elution orders of the solutes are very similar. The SE-CF/PEG-CC is superior to SE/PEG-CF/C annular column in terms of separation efficiency and column design. First, the SECF/PEG-CC annular column has a shorter analysis time (3.5 min) than the SE/PEG-CF/C annular column (5.3 min). Second, 1-hexanol and n-undecane were much better resolved on the SE-CF/PEG-CC annular column. Third, the column length is shorter, i.e., 6-m long SE-CF/PEG-CC versus 12-m long SE/PEG-CF/C annular columns. In brief, shorter column has the advantages of lower pressure drop and shorter sample analysis time. Our results show that the stationary phase-CF/stationary phase-CC annular column is a better choice for GC separation as it is more efficient and flexible. 4. Conclusion In this work, a novel stationary phase-coated fiber-in-stationary phase-(un)coated capillary annular column was demonstrated and its performance was evaluated for the first time in GC. Only one single fiber with larger o.d. (0.12 mm) was used to coat the stationary phase as compared to others using many fibers of smaller o.d. (typically, 15 m) packed in the column. Our silica fiber was proved to be a good support material for stationary phase. Various types of stationary phases can be coated on a single fiber or both fiber and capillary without stationary phase mixing according to the analysis required. The inserted stationary phase-coated fiber can decrease void volume, facilitate faster mass transfer, and thus improve column efficiency. Selectivity can be adjusted by using various stationary phase-coated fiber-in-stationary phase-(un)coated capillary annular columns. The stationary phase-CF/stationary phase-CC design provides feasibility in tuning the selectivity of the separated solutes. Finally, the separation of any complex mixtures can be realized by optimizing the types, concentration and coating length of stationary phases, or even introducing a third stationary phase. This work is currently in progress in our laboratory. Lastly, more work can also be pursued by incorporating our annular column with the series-coupled column design to further enhance the separation power of a GC column. But it should be borne in mind that although longer annular columns can provide more theoretical plates, the total analysis time will be lengthened. Acknowledgements Financial support from National Natural Science foundation of China (20575042, 20775050) and Education Ministry of China (105141) are gratefully acknowledged. References [1] [2] [3] [4]
Fig. 6. Separation of a complex mixture of organic compounds on (a) PEG-CF/SECC and (b) SE-CF/PEG-CC (6 m × 0.20 mm, 1.0 m) annular columns. Solutes 1: npropanol, 2: isobutanol, 3: isoamylol, 4: isooctane, 5: butyl acetate, 6: amyl acetate, 7: 1-hexanol, 8: n-decane, and 9: n-undecane. Conditions: linear carrier gas velocity 25 cm/s; temperature program from 65 to 90 ◦ C at 7 ◦ C/min; and splitless injection.
[5] [6] [7] [8] [9] [10]
C.F. Chien, R.J. Laub, M.M. Kopecni, Anal. Chem. 52 (1980) 1402. J.H. Purnell, M.H. Wattan, J. Chromatogr. 555 (1991) 173. R.J. Laub, J.H. Purnell, J. Chromatogr. 112 (1975) 71. J. Sanz, I. Martínez-Castro, G. Reglero, M.D. Cabezudo, Anal. Chim. Acta 194 (1987) 91. A. Ruderisch, J. Pfeiffer, V. Schurig, J. Chromatogr. A 994 (2003) 127. X.Y. Shi, Y.Q. Zhang, R.N. Fu, Anal. Chim. Acta 424 (2000) 271. J.H. Purnell, P.S. Williams, J. Chromatogr. 292 (1984) 197. E. Benická, J. Krupˇcík, D. Repka, P. Kuljovsky, R.E. Kaiser, Anal. Chem. 62 (1990) 985. A. Jayatilaka, S.K. Poole, C.F. Poole, T.M.P. Chichila, Anal. Chim. Acta 302 (1995) 147. J.M. Sanchez, R.D. Sacks, Anal. Chem. 75 (2003) 2231.
3348 [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
P. Li et al. / J. Chromatogr. A 1216 (2009) 3343–3348
H. Smith, R.D. Sacks, Anal. Chem. 69 (1997) 5159. A.J. Grall, E.T. Zellers, R.D. Sacks, Environ. Sci. Technol. 35 (2001) 163. T. Veriotti, R.D. Sacks, Anal. Chem. 73 (2001) 3045. J. Whiting, R.D. Sacks, Anal. Chem. 74 (2002) 246. C.L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145. C. Dietz, J. Sanz, C. Cámara, J. Chromatogr. A 1103 (2006) 183. A.L. Lopes, F. Augusto, J. Chromatogr. A 1056 (2004) 13. M. Berg, J. Bolotin, T.B. Hofstetter, Anal. Chem. 79 (2007) 2386. F. Hernández, T. Portolés, E. Pitarch, F. López, J. Anal. Chem. 79 (2007) 9494. Y. Saito, M. Imaizumi, K. Nakata, T. Takeichi, K. Kotera, H. Wada, K. Jinno, J. Microcol. Sep. 13 (2001) 259. [21] Y. Saito, A. Tahara, M. Imaizumi, T. Takeichi, H. Wada, K. Jinno, Anal. Chem. 75 (2003) 5525. [22] Y. Saito, M. Ogawa, M. Imaizumi, K. Ban, A. Abe, T. Takeichi, H. Wada, K. Jinno, Anal. Bioanal. Chem. 382 (2005) 825. [23] Y. Saito, M. Ogawa, M. Imaizumi, K. Ban, A. Abe, T. Takeichi, H. Wada, K. Jinno, J. Chromatogr. Sci. 43 (2005) 536.
[24] M. Ogawa, Y. Saito, M. Imaizumi, H. Wada, K. Jinno, Chromatographia 63 (2006) 459. [25] P.K. Dasgupta, Anal. Chem. 56 (1984) 96. [26] P.K. Dasgupta, Anal. Chem. 56 (1984) 103. [27] P.K. Dasgupta, R.Q. Bligh, M.A. Mercurio, Anal. Chem. 57 (1985) 484. [28] T.S. Stevens, J.C. Davis, H. Small, Anal. Chem. 53 (1981) 1488. [29] C.A. Bruckner, R.E. Synovec, Talanta 43 (1996) 901. [30] R.E. Synovec, A.W. Sulya, L.W. Burgess, M.D. Foster, C.A. Bruckner, Anal. Chem. 67 (1995) 473. [31] M.D. Foster, R.E. Synovec, Anal. Chem. 68 (1996) 1456. [32] K. Jinno, H. Watanabe, Y. Saito, T. Takeichi, Electrophoresis 22 (2001) 3371. [33] Z.-G. Shi, W. Fang, Y.-Q. Feng, J. Liq. Chromatogr. Rel. Technol. 31 (2008) 3094. [34] C.F. Poole, The Essence of Chromatography, Elsevier, Amsterdam, 2003, p. 32.