A new imidazolium ionic polymer as a gas chromatography stationary phase for separation of high and wide temperature range complex samples

Abstract

A new imidazolium ionic liquid polymer was synthesized and prepared as GC stationary phase for separation of coal tar samples. The coated capillary GC columns exhibited good thermal stability from room temperature to 380 °C, high separation efficiencies for coal tar samples, phthalates, PAHs and a maximum programmable temperature up to 400 °C.

---

Ionic liquids (ILs) have attracted a great deal of attention due to their unique properties such as thermal stability, low vapour pressure, electrical conductivity, solubility and inflammability.^1–7 These properties can be tuned by changing the chemical composition of cations and anions, which allowed their applications in different areas (e.g. separations, catalyst, electrochemistry and others).^8–14

Recently, ILs have been synthesized as stationary phases for chromatography and made great success in both research and applications. Unfortunately, they are still not quite compared with commercially available columns. The high thermal stability and operating temperature are nowadays demanded in design of novel stationary phases, analyzing environmental and petrochemical samples. However, ILs can lose the viscosity, vaporize or decompose under high temperature, which implies the shorter column life and reproducibility. In order to overcome the limitations, a number of authors have shown polymeric ionic liquids (PILs) as great promises to increase their mechanical strength and thermal stability. More recently, Amstrong et al. prepared cross-linked IL stationary phases which the column stability was up to 280 °C.¹⁵ Armstrong et al. also demonstrated a series of dicationic based ionic liquids with high thermal stability, which the thermal stability of these ionic liquid columns achieved 350–400 °C.^16–18 Our previous study showed the column stability of PILs up to 350 °C.¹⁹ On the basis of our studies in preparation of PILs based stationary phases for gas chromatography (GC), we have proposed the new PIL as the stationary phase for GC, which possess good thermal stability at 380 °C, programmable operation temperature up to 400 °C and wide operating temperature (50–400 °C) without lose separation efficiencies to coal tar samples, phthalates, and PAHs.

In order to manufacture more thermal stable PILs, we used N-vinylimidazole group to cross-link new PILs. The PILs in this study were derived from our previous study and modified with N-vinylimidazole group.^19,20 The synthesis process was briefly described as following, one of PILs (PIL 1) was obtained by self-polymerization reaction, 3 g of 1-(3-chlorohexyl) imidazole (ImC₆Cl) was added in propanol at 80 °C for 48 hours, then 4.5 g of N-vinylimidazole was added and reacted for another 24 hours. The other PILs (PIL 2) was obtained as following, 5 g of ImC₆Cl was added in ethylene glycol at 95 °C for 8 hours, then 7.6 g of N-vinylimidazole was added and reacted for another 24 hours. Previous studies showed that the PILs GC stationary phases containing NTf₂ (bis(trifluoromethylsulfonyl)imide) have high thermal stability and better separation efficiencies;^19,20 therefore, metathesis reactions were carried out to replace the chloride to NTf²⁻ by using LiNTf₂ (lithium bis(trifluoromethylsulfonyl)imide) aqueous solution. After replacing the chloride to NTf²⁻, 0.0882 g of PIL 1 and 0.101 g of PIL 2 were dissolved in DMSO and further polymerized with 0.1% of azobisisobutyronitrile (AIBN) at 90 °C for 24 hours. The new polymeric ionic liquid was obtained after removing DMSO with vacuum. The relative molecular weight of new polymeric ionic liquids was about 90 000 g mol⁻¹, which was evaluated by gel permeation chromatography (GPC) (equipped with Jasco-880PU pump, Water Styragel HR4E (M_w: 50–100 000) and a UV detector). The chemical structure of the new polymeric ionic liquid was shown in Fig. 1.

Fig. 1 The proposed chemical structure of the new polymeric ionic liquid.

The PIL capillary columns were prepared by using the dynamic coating method. The new polymeric ionic liquid was dissolved in acetone at a concentration of 1% (w/v, 10 c.c.) as stationary phase solution. The nitrogen (10 psi) was used to inject the prepared stationary phase solution into untreated capillary fused silica tubing (10 m × 0.25 mm i.d.). After capillary fused silica tubing was filled with stationary phase solution, and nitrogen gas was used to push the solution out of the column at speed of 2 to 4 cm per second, then column was dried by nitrogen stream for overnight. The new polymeric ionic liquids coated capillary column was installed into the GC injector and flushed with nitrogen (0.7 mL min⁻¹) and held at 50 °C for 3 hours, then heated from 50 °C to 100 °C with 15 °C min⁻¹ heating rate. And then slowly heated again to 400 °C and held for 10 minutes. After conditioning, the column was ready for use. The film thicknesses of prepared PIL capillary column was estimated as 0.15 μm.^21–24 The new polymeric ionic liquid column dimension in this study is 10 m × 0.25 mm i.d. × 0.15 μm. Our previous study column dimension is 10 m × 0.25 mm i.d. × 0.20 μm m. DB-5 column dimension is 10 m × 0.25 mm i.d. × 0.25 μm. GC equipped with mass spectrometer (Agilent GC7820A 5975MSD, USA) was used for identification of the analytes and GC (Hewlett Packard 5890 Series II, USA) equipped with a flame ionization detector (FID) was used for the GC analyses. The systems were controlled using Chem. Station and Chromatography Data System (SISC, Taiwan). Nitrogen was used as the carrier gas for the GC and the dynamic coating procedure. A thermal gravimetric analyzer TGA (TGA-50, Shimadzu, Japan) was used to analyze the decomposition temperature of new polymeric ionic liquid.

Fig. 2 shows TGA results of new polymeric ionic liquid. TGA result showed significant weight loss temperatures of 450 °C was observed for new polymeric ionic liquid, which matched the goal of this study.

Fig. 2 TGA results of new polymeric ionic liquid, Temperature program: 100 to 500 °C at a heating rate of 10 °C min⁻¹; N₂ flow rate: 20 mL min⁻¹; platinum plate.

The thermal stability of stationary phase is very important to GC, which was evaluated by GC-FID baseline test. Fig. 3 shows a comparison of the baseline test of new polymeric ionic liquids column in this paper with our previous study and a commercially available 95% dimethyl, 5% diphenylpolysiloxane column (DB-5). In Fig. 3, line 1 to 3 describe the FID signal response of DB-5 (recommended min. bleeding temp: 330 °C, max. program temp: 360 °C), our previous study, and new polymeric ionic liquid column respectively. DB-5 (line 1) showed baseline was low at low temperature, but rose higher as oven temperature reached 330 °C and column bleeding occurred at 360 °C. Our previous study (line 2) showed FID response almost unchanged under 330 °C and demonstrated bleeding temperature around 380 °C. In this study (line 3), results showed the new polymeric ionic liquid column could be stable at 350 °C and demonstrated bleeding temperature around 400 °C. The maximum programmable temperature was investigated by slowly increasing oven temperature and the FID response was observed. The retention factor (k′) of a standard solution of naphthalene was routinely tested after repeatedly heating of column to the temperature of 400 °C, the RSD of k′ was 0.8% (n = 5).¹⁹ The mixtures of phthalates, PAHs, n-alkanes, and coal tar samples were tested for practical applications.

Fig. 3 GC-FID responses for (1) DB-5 column; (2) Our previous study;¹⁹ (3) This paper new polymeric ionic liquids column. Oven temperature program: 100 to 400 °C at 15 °C min⁻¹ heating rate, 1 mL min⁻¹ flow rate.

Fig. 4 shows a comparison of the chromatogram of a mixture of six phthalates of new polymeric ionic liquid column with a commercially available DB-5 column. New polymeric ionic liquid column performed acceptable separation efficiency, symmetrical and sharp peak shapes were obtained. Fig. 4 also shows the different retention orders (peak 5–7) in new polymeric ionic liquid column and DB-5 column. Fig. 4(a) shows that new polymeric ionic liquid column has retention order in DNOP, BBP, BEHP, but Fig. 4(b) shows that DB-5 column has retention order in BEHP, DNOP, BBP. The different retention order might due to the dipole–induced dipole interactions among different phthalates, imidazolium cations and NTf2 anions.

Fig. 4 Chromatogram for EPA phthalate mix each 25 μg mL⁻¹ in (a) new polymeric ionic liquids column; (b) DB-5 column, peak (1) solvent: dichloromethane; (2) dimethylphthalate, DMP; (3) diethylphthalate, DEP; (4) di-n-butylphthalate, DNBP; (5) di-n-octylphthalate, DNOP; (6) benzylbutylphthalate, BBP; (7) bis(2-ethylhexyl)phthalate, BEHP. Oven temperature program: 50 to 400 °C at 15 °C min⁻¹ heating rate, split ratio: 10 : 1, 0.7 mL min⁻¹ flow rate.

In order to emphasize the advantage of new polymeric ionic liquids column, PAHs with very low volatility and high boiling point such as indeno[1,2,3-cd]pyrene (b.p. 536 °C) and benzo[g,h,i]perylene (b.p. 536 °C) were employed as analytes for the test. There is a problem for GC columns to purge out these low volatility compounds at relatively high temperature, which caused GC column unsustainable. Typically, other separation method such as HPLC was recommended to separate these low volatility compounds. Fig. 5 shows a comparison of the chromatogram of a mixture of 18 PAHs of new polymeric ionic liquid column with a commercially available DB-5 column. Fig. 5 shows the same retention order of PAHs in new polymeric ionic liquid and DB-5 column. Compared with Fig. 5(b) 18 PAHs in DB-5 column, Fig. 5(a) shows symmetrical peaks and high efficiency obtained in new polymeric ionic liquid column except for peak 8, 9, 12–14, and 15–19. This might due to the similar volatilities (peak 8, 9, 12–14, 15 and 16), π–πinteractions (peak 18 and 19), ring strain of PAHs, and shorter tested column length (10 m). Better separation efficiency could be achieved by extending the length of the column.

Fig. 5 Chromatogram for PAHs mix each 25 μg mL⁻¹ in (a) new polymeric ionic liquid column (b) DB-5 column, peak (1) solvent: dichloromethane; (2) naphthalene; (3) 1-methyl naphthalene; (4) 2-methyl naphthalene; (5) acenaphthylene; (6) acenaphthene; (7) fluorene; (8) anthracene; (9) phenanthrene; (10) fluoranthene; (11) pyrene; (12) benz[a]anthrene; (13) chrysene; (14) benzo[k]fluoranthene; (15) benzo[b]fluoranthene; (16) benzo[a]pyrene; (17) indeno[1,2,3-cd]pyrene; (18) dibenz[a,h]anthracene; (19) benzo[g,h,i]perylene. Oven temperature program: 50 to 400 °C at 15 °C min⁻¹ heating rate, split ratio: 10 : 1, 0.5 mL min⁻¹ flow rate.

Fig. 6 shows a comparison of the chromatogram of a mixture of 15 n-alkanes of new polymeric ionic liquid column with a commercially available DB-5 column. Compared with Fig. 6(b) the chromatogram of a mixture of n-alkanes in DB-5 column, Fig. 6(a) shows new polymeric ionic liquids column has poor resolution and efficiency for n-alkanes. Even though Fig. 6(a) shows the retention of all tested n-alkanes were eluted out within 12 min and had more stable baseline, but front tailoring effects were observed. The result implies interactions between n-alkanes and new polymeric ionic liquids are relative small. This can be attributed to the large amount of imidazolium cations and NTf₂ anions were used in synthesis of new polymeric ionic liquid and preparation of column.

Fig. 6 Chromatogram for n-alkanes mix each 25 μg mL⁻¹ in (a) new polymeric ionic liquid column; (b) DB-5 column, peak (1) solvent: dichloromethane; (2) C₉; (3) C₁₀; (4) C₁₂; (5) C₁₄; (6) C₁₆; (7) C₁₈; (8) C₂₀; (9) C₂₂; (10) C₂₄; (11) C₂₆; (12) C₂₈; (13) C₃₀; (14) C₃₂; (15) C₃₄; (16) C₃₆. Oven temperature program: 50 to 400 °C at 10 °C min⁻¹ heating rate, split ratio: 10 : 1, 0.7 mL min⁻¹ flow rate.

It is an important issue to analyze environmental or petrochemical samples, which also show the ability of the high temperature GC column. In this study, coal tar samples obtained from Chinese steel cooperation were also used to show the benefit of new polymeric ionic liquid column. Fig. 7 shows a comparison of the chromatogram for coal tar samples in new polymeric ionic liquid and DB-5 columns. New polymeric ionic liquid column (Fig. 7(a)) showed more stable baseline separation than DB-5 column (Fig. 7(b)). Therefore, new polymeric ionic liquid column was coupled with GC-MS for further identifications (Fig. 8). Fig. 8 shows the total ion current (TIC) chromatogram of coal tar sample in new polymeric ionic liquid column. 7 major compounds were identified (showed in Fig. 8) according to the standards and database. No major analytes were found and investigated in coal tar samples after 215 °C.

Fig. 7 Chromatogram for coal tar samples in (a) new polymeric ionic liquid column; (b) DB-5column. Oven temperature program: 50 to 400 °C at 15 °C min⁻¹, split ratio: 10 : 1, 0.7 mL min⁻¹ flow rate.

Fig. 8 The total ion current (TIC) chromatogram of coal tar sample in new polymeric ionic liquid column. Peak (1) dibenzo[b,d]furan, (2) acenaphthylene, (3) 1H-phenalene, (4) anthracene, (5) fluoranthene, (6) 1,5-dihydropyrene, (7) 9H-carbazole. Oven temperature program: 50 (1 min) to 400 °C at 15 °C min⁻¹, split ratio: 10 : 1, 0.7 mL min⁻¹ flow rate.

It is worth to mention that the new polymeric ionic liquid stationary phases are applicable and stable from low to high temperature without losing separation efficiency. Most ultra high temperature-resistant stationary phases become relatively rigid and less flexible to solutes under low temperatures, which cause poor separation efficiency at low temperatures. However, the straight hexyl chains on new polymeric ionic liquid stationary phases are not long enough to separate n-alkanes. According to previous study,¹² the longer alkyl chain on imidazolium based polymeric ionic liquid could have higher dispersion force. Therefore, new polymeric ionic liquid stationary phases showed poor separation efficiencies for n-alkanes compounds.

New polymeric ionic liquid columns showed good reproducibility after routinely tests under the condition of 120 chromatographic runs each week for five months. This can be attributed to facile synthetic methods, simple and reliable coating procedures. After a few months of tests, new polymeric ionic liquid columns showed only a slight retention time shift was observed (RSD < 0.9%). In order to evaluate the column, we also calculate the resolution (Rs) and height equivalent to a theoretical plate (HETP in mm) as shown in Table 1. The results showed that new polymeric ionic liquid column has better Rs and stronger interaction but less efficiency for PAHs ((5) acenaphthylene and (6) acenaphthene) than DB-5 column; new polymeric ionic liquid column has less Rs but better efficiency for phthalate mix ((2) dimethylphthalate, DMP; (3) diethylphthalate, DEP) than DB-5 column; new polymeric ionic liquid column has less resolution and weak interaction for alkanes ((4) C₁₂; (5) C₁₄) than DB-5 column. Obviously, new polymeric ionic liquid column is not suitable for separating alkanes.

Table 1 Resolution and efficiency measurements of polymeric ionic liquid and DB-5 columns

Figures		EPA phthalate mix		PAHs mix		n-alkanes mix
Compound's number		2	3	5	6	4	5
Resolution	Polymer ionic liquid column	3.5		5		5.6
	DB-5 column	10.8		2.8		22.5
N	Polymer ionic liquid column	106 276	115 600	25 600	32 400	3091	6084
	DB-5 column	71 396	96 348	70 543	76 618	50 625	99 225
HETP (mm)	Polymer ionic liquid column	0.094	0.087	0.391	0.309	3.235	1.644
	DB-5 column	0.140	0.104	0.142	0.131	0.198	0.101

---

In conclusion, the proposed new polymeric ionic liquid stationary phase shows highly thermal stability and separation efficiency for GC. The results recommend that new polymeric ionic liquids column has excellent thermal stability and separation efficiency for low-volatility and high boiling point analytes. The new polymeric ionic liquid column also has wide operating temperature (50–400 °C), which can expend applications of the GC column. The further work is to design more thermal stable and wide applicable ionic polymers included their derivatives, which are investigated in progress.

Acknowledgements

The authors thank the National Science Council (Ministry of Science and Technology) for financial supports (NSC100-2113-M006-001-MY3).

Notes and references

1. J. L. Anderson, D. W. Armstrong and G. T. Wei, Anal. Chem., 2006, 78, 2893–2902.
2. M. Koel, Crit. Rev. Anal. Chem., 2005, 35, 177–192.
3. G. A. Baker, S. N. Baker, S. Pandey and F. V. Bright, Analyst, 2005, 130, 800–808.
4. S. Pandey, Anal. Chim. Acta, 2006, 556, 38–45.
5. J. F. Liu, J. A. Jonsson and G. B. Jiang, TrAC, Trends Anal. Chem., 2005, 24, 20–27.
6. K. J. Fraser, E. I. Izgorodina, M. Forsyth, J. L. Scott and D. R. MacFarlane, Chem. Commun., 2007, 3817.
7. T. Welton, Chem. Rev., 1999, 99, 2071–2083.
8. D. W. Armstrong, L. F. He and Y. S. Liu, Anal. Chem., 1999, 71, 3873–3876.
9. D. S. Silvester, Analyst, 2011, 136, 4871.
10. J. L. Anderson and D. W. Armstrong, Anal. Chem., 2003, 75, 4851–4858.
11. H. Qiu, A. K. Mallik, M. Takafuji, S. Jiang and H. Ihara, Analyst, 2012, 137, 2553.
12. Y. N. Hsieh, R. S. Horng, W. Y. Ho, P. C. Huang, C. Y. Hsu, T. J. Whang and C. H. Kuei, Chromatographia, 2008, 67, 413–420.
13. C. Zhanga and J. L. Anderson, J. Chromatogr. A, 2014, 1344, 15–22.
14. J. L. Anderson, R. F. Ding, A. Ellern and D. W. Armstrong, J. Am. Chem. Soc., 2005, 127, 593–604.
15. J. L. Anderson and D. W. Armstrong, Anal. Chem., 2005, 77, 6453–6462.
16. M. Qi and D. W. Armstrong, Anal. Bioanal. Chem., 2007, 388, 889–899.
17. Z. S. Breitbach and D. W. Armstrong, Anal. Bioanal. Chem., 2008, 390, 1605–1617.
18. T. Payagala, Y. Zhang, E. Wanigasekara, K. Huang, Z. S. Breitbach, P. S. Sharma, L. M. Sidisky and D. W. Armstrong, Anal. Chem., 2009, 81, 160–173.
19. W.-Y. Ho, Y.-N. Hsieh, W.-C. Lin, C. L. Kao, P.-C. Huang, C.-F. Yeh, C.-Y. Pan and C.-H. Kuei, Anal. Methods, 2010, 2, 455–45721.
20. Y.-N. Hsieh, C.-H. Kuei, Y.-K. Chou, C.-C. Liu, K.-L. Leu, T.-H. Yang, M.-Y. Wang and W.-Y. Ho, Tetrahedron Lett., 2010, 51, 3666–3669.
21. K. D. Bartle, Anal. Chem., 1973, 45, 1831.
22. F. Fairbrother and A. E. Stubbs, J. Chem. Soc., 1935, 527.
23. M. Novontny, L. Blomberg and K. D. Bartle, J. Chromatogr. Sci., 1970, 8, 309.
24. M. Novontny, K. D. Bartle and L. Blomberg, J. Chromatogr., 1969, 45, 469.

---