- Email: [email protected]
Separation performance of cucurbit[7]uril in ionic liquid-based sol-gel coating as stationary phase for capillary gas chromatography
Journal of Chromatography A, 1371 (2014) 237–243
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Separation performance of cucurbit[7]uril in ionic liquid-based sol-gel coating as stationary phase for capillary gas chromatography Xiaogang Wang, Meiling Qi ∗ , Ruonong Fu Key Laboratory of Cluster Science, Ministry of Education of China, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials and School of Chemistry, Beijing Institute of Technology, Beijing 100081, China
a r t i c l e
i n f o
Article history: Received 27 February 2014 Received in revised form 26 September 2014 Accepted 21 October 2014 Available online 30 October 2014 Keywords: Cucurbit[7]uril Stationary phase Ionic liquid Sol–gel coating Capillary gas chromatography
a b s t r a c t Here we report the separation performance of a new stationary phase of cucurbit[7]uril (CB7) incorporated into an ionic liquid-based sol-gel coating (CB7-SG) for capillary gas chromatography (GC). The CB7-SG stationary phase showed an average polarity of 455, suggesting its polar nature. Abraham system constants revealed that its major interactions with analytes include H-bond basicity (a), dipole–dipole (s) and dispersive (l) interactions. The CB7-SG stationary phase achieved baseline separation for a wide range of analytes with symmetrical peak shapes and showed advantages over the conventional polar stationary phase that failed to resolve some critical analytes. Also, it exhibited different retention behaviors from the conventional stationary phase in terms of retention times and elution order. Most interestingly, in contrast to the conventional polar phase, the CB7-SG stationary phase exhibited longer retentions for analytes of lower polarity but relatively comparable retentions for polar analytes such as alcohols and phenols. The high resolving ability and unique retention behaviors of the CB7-SG stationary phase may stem from the comprehensive interactions of the aforementioned interactions and shape selectivity. Moreover, the CB7-SG column showed good peak shapes for analytes prone to peak tailing, good thermal stability up to 280 ◦ C and separation repeatability with RSD values in the range of 0.01–0.11% for intraday, 0.04–0.41% for inter-day and 2.5–6.0% for column-to-column, respectively. As demonstrated, the proposed coating method can simultaneously address the solubility problem with CBs for the intended purpose and achieve outstanding GC separation performance. © 2014 Elsevier B.V. All rights reserved.
1. Introduction For capillary gas chromatography (GC), a selective stationary phase and an efficient column preparation method are the requisites for a GC column to achieve high-resolution separations. Cucurbit[n]urils (CBs, n = 5–8, 10) are a family of pumpkin-shaped macrocyclic compounds with a hydrophobic cavity and two identical carbonyl-laced portals [1–3]. The hydrophobic interior provides a potential interaction site for nonpolar to low polar analytes while the ureido carbonyl groups at the portals offer selective interactions with polar analytes via dipole and H-bond interactions. Also, CBs possess high thermal stability. The unique structures and novel physicochemical properties of CBs have attracted great attention in supramolecular chemistry [4–6], separation science [7–14] and many other fields [15,16]. The aforementioned features also offer
∗ Corresponding author. Tel.: +86 10 68912668; fax: +86 10 68914780. E-mail addresses: [email protected], [email protected] (M. Qi). http://dx.doi.org/10.1016/j.chroma.2014.10.066 0021-9673/© 2014 Elsevier B.V. All rights reserved.
CBs as ideal candidates for capillary GC stationary phases. However, their poor solubility in common organic solvents for column preparation greatly limits their use for the intended purpose. Up to now, only two reports on using CBs as stationary phases for capillary GC are available [13,14], showing their high selectivity and thermal stability as stationary phases for capillary GC separations. Over the past decades, macrocyclic GC stationary phases such as cyclodextrins and calixarenes were reported. Since they have good solubility in dichloromethane or methanol commonly used in capillary column fabrication, it is relatively easier to fabricate a capillary GC column of these stationary phases with high column efficiency by static coating method. Often, they were reported to be used in a mixture with polysiloxanes for improving their GC separation performance. However, CBs are difficult to get a high-efficiency column by the same way due to their insolubility in the solvents mentioned above. Most recently, we successfully used CB6 as the GC stationary phase by simultaneously employing a guanidinium-based ionic liquid (GBIL) and a sol-gel coating
238
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
method [13]. The GBIL was N , N - (6 , 6 -dihydroxy)dihexyl-N, N, N , N -tetramethyl guanidinium bis(trifluoromethylsulfonyl)imide (DHHTMG-NTf2 ) (denoted as GBIL in the following). This GBIL was chosen after our many trials with different ionic liquids including imidazolium-based ILs [20] and guanidinium ILs [21–23] previously investigated in our laboratory since there were no any related reports available. Among the ILs tested, the indicated GBIL was found to be efficient in enhancing the CBs solubility in the sol solution and made it achievable to prepare the capillary GC column of CBs by sol–gel coating method. The proposed method proved to be efficient and feasible for use of CBs as new alternatives of GC stationary phases. Solgel coating method is known as a one-step coating method for preparation of capillary GC columns [17–19]. This method usually uses a tetraalkoxysilane as the sol–gel precursor, trifluoroacetic acid (TFA) as the catalyst, dichloromethane as the solvent, poly (methylhydrosiloxane) (PMHS) as the deactivation reagent, and a given stationary phase. The sol–gel process may involve the reactions of catalytic hydrolysis of the alkoxide precursor, polycondensation of the hydrolyzed products and immobilization of the stationary phase, leading to a three-dimensional (3D) sol-gel coating network with high physicochemical stability. Among the family members, CB7 (Fig. 1) has a larger dimension than CB6 in terms of diameter and cavity, and easier availability than CB8 that has a quite low yield in its synthesis. The larger dimension offers CB7 interactions with a wider range of analytes and may also exhibit different selectivity and resolving ability from CB6 in capillary GC separations. Here we report the fabrication of CB7 sol–gel (SG) capillary column and its GC separation performance for analytes of different variety. Meanwhile, a well-recognized commercial column with comparable polarity was also used for comparison. McReynolds constants and Abraham system constants were determined to characterize the molecular interactions of the CB7-SG stationary phase with analytes. Separation performance was examined by esters, aldehydes, alcohols, the Grob mixture and a more complex mixture containing 19 analytes of different types. Moreover, separation repeatability (intra-day, inter-day and column-to-column) and thermal stability were investigated as well. 2. Experimental 2.1. Materials and equipment Tetraethoxysilane (TEOS) was purchased from Fluka Chemika Co. (Steinheim, Switzerland). Poly(methylhydrosiloxane) (PMHS) and trifluoroacetic acid (TFA, 95%) were purchased from Alfa Aesar (Heysham, UK). Dichloromethane and the rest of chemicals used in this work were purchased from Beijing Chemical Reagent Company (Beijing, China). All the chemicals were of analytical grade and dissolved in dichloromethane. Untreated fused-silica capillary tubing (0.25 mm, i.d.) was purchased from Yongnian Ruifeng Chromatogram Apparatus Co., Ltd. (Hebei, China). A commercial HP-INNOWAX capillary column (10 m × 0.25 mm i.d., 0.25 m film thickness, Agilent Technologies) was used for comparison. An Agilent 7890A gas chromatograph equipped with a split/splitless injector and a flame ionization detector (FID) was used for the GC separations. All the separations were performed under the following GC conditions: nitrogen of high purity (99.999%) as carrier gas at a flow rate of 1 mL/min, injection port at 250 ◦ C, split injection mode with a split ratio at 30:1, FID detector at 300 ◦ C. A Hitachi S4800 scanning electron microscope (SEM) (Hitachi, Japan) was also used for the observation of the column inner surface morphology.
2.2. Methods 2.2.1. Capillary column preparation Column preparation for the CB7-SG capillary column followed the method in the reference [13] with minor modification. CB7 and GBIL were synthesized following the same methods provided in our previous work [14] and [13], respectively. Briefly, CB7 was synthesized by the reaction of glycoluril and paraformaldehyde in a concentrated HCl solution at room temperature and purified by multiple precipitation and solvent extraction [14]. GBIL was obtained by reaction of tetramethyl guanidine, tetrabutylammonium bromide and 6-bromo-1-hexanol at 75 ◦ C for 48 h and followed by anion exchange and purification via solvent extraction [13]. Prior to coating, a capillary column (10 m × 0.25 mm, i. d.) was successively pretreated with 1 mol/L NaOH and 0.1 mol/L HCl, rinsed with water and dried overnight at 120 ◦ C under nitrogen. After this, 100 L TEOS, 125 mg GBIL and 10 L PMHS were added into a vial and mixed well. Then, 15 mg CB7, 100 L of TFA and 250 L CH2 Cl2 were added and sonicated for 3 min. After another 250 L CH2 Cl2 were added, the mixture was sonicated for 3 min. Next, the resulting sol solution was pumped into the pretreated capillary column and stayed for 30 min at room temperature. After the extra sol solution was expelled from the column, the column was conditioned from 40 ◦ C for 30 min and then to 190 ◦ C at a rate of 1 ◦ C/min and held at the final temperature for 6 h. Column efficiency of the prepared column was determined by naphthalene at 100 ◦ C.
2.2.2. McReynolds constants and Abraham system constants McReynolds constants are often used to characterize the polarity of a GC stationary phase and its molecular interactions with analytes. Five probe compounds, namely benzene, n-butanol, 2-pentanone, nitropropane and pyridine, are used for the determination and serve to measure –electron interaction (X ), H-bond donor and acceptor interaction (Y ), H-bond acceptor interaction (Z ), dipole interactions (U ) and strong H-bond acceptor interaction (S ), respectively. For a GC column, McReynolds constants are measured by determining the differences of the five probes in the retention indices on the given stationary phase and squalane. The average of the differences is used for the polarity estimation. A GC stationary phase with average polarity over 400 is identified as polar phase. In this work, McReynolds constants of the CB7-SG column were determined at 120 ◦ C. Abraham system constants for a given stationary phase is determined on the basis of Abraham’s linear solvation parameter model, which can be used to quantitatively evaluate the individual intermolecular interactions between a GC stationary phase and probe compounds [24–27] and is set out below in the form for GC with the logarithm of retention factor (log k) as the dependent variable.
log k = c + eE + sS + aA + bB + lL The capital letters (E, S, A, B, L) are solute descriptors that represent different interactions (Table S1). For probe compounds, their values of solute descriptors are available in the reference [25] and the probe compounds used in this work and their solute descriptors are listed in Table S2. The lowercase letters (e, s, a, b and l) are system constants representing the contributions of a stationary phase to retention mechanism by the defined molecular interactions (Table S1). System constants for a given stationary phase can be calculated on the retention factors of probe compounds using multiple linear regression analysis.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
239
3. Results and discussion 3.1. McReynolds constants and Abraham system constants Characteristic parameters of the as-prepared CB7-SG column regarding column efficiency, McReynolds constants and Abraham system constants were determined. The column efficiency measured by naphthalene at 100 ◦ C was 2078 plates/m for CB7-SG column and 3727 plates/m for HP-INNOWAX column. CB7-SG column had lower column efficiency than the commercial column but higher efficiency than that of the statically coated CB7 column (1400 plates/m) [14]. Table 1 lists the McReynolds constants of CB7SG and commercial columns for comparison. As shown, CB7-SG column had an average polarity of 455, suggesting its polar nature, which was quite close to that of the commercial column. The high polarity of CB7-SG stationary phase can mainly be attributed to the polar GBIL used to improve the solubility of CB7 in the sol solution. Though CB7-SG stationary phase exhibited almost the same average polarity with the commercial phase, their individual molecular interactions differed greatly. The former exhibited higher Z , U and S values but lower Y value than the latter, suggesting its stronger dipole–dipole and H-bond acceptor interactions but weaker Hbond donator interaction with analytes. These differences may be indicative of their different separation performance. Table 2 lists the Abraham system constants of CB7-SG stationary phase at three temperatures of 80 ◦ C, 100 ◦ C and 120 ◦ C and shows that the stationary phase exhibits strong intermolecular interactions with analytes via dipole–dipole interaction (s), H-bond basicity (a) and dispersive interactions (l) but weak H-bond acidity (b) and –/n– interactions (e). This result is in good agreement with that of McReynolds constants. Additionally, it can be observed that the s, a, b and l values decreased with the rise of temperature but the e value slightly increased. The interactions related with s, a and b depend on the orientation of a stationary phase and solutes and become weakened as temperature rises due to the translational and rotational energy increase of the molecules. Dispersive interactions (l) are strongly distance-dependent and can be lessened with rising temperature owing to the increased distance between a stationary phase and a solute. With respect to e value, the slightly rising value as temperature increases may be attributed to that electrons are less tightly bound to atoms at higher temperatures. Moreover, the inner surface morphology of the CB7-SG capillary column was also examined by scanning electron microscopy (SEM) and the result is shown in Fig. S1 in Supplementary material. As can be observed, the inner coating showed an uneven and roughened surface inherently originating from the sol-gel coating process, which was favorable for enhancing the coating surface area and interactions of the stationary phase with analytes for GC separations. The film thickness of the sol–gel coating cannot be accurately determined due to its typical uneven surface. In this work, the film thickness of the sol–gel coating was roughly estimated to be around 0.8 m by the SEM image. Thicker coating layer is another feature of sol–gel coating procedure in contrast to static coating method. 3.2. Separation performance Separation performance of CB7-SG column was evaluated by separations of different mixtures including esters, aldehydes, alcohols, the Grob mixture and a more complex mixture containing 19 analytes of different types. Fig. 2 shows the GC separations of esters, aldehydes and alcohols on the CB7-SG column. All the analytes were baseline resolved and eluted in the order of their boiling points except 1, 4-butyrolactone that eluted much later than expected with regard to its boiling point. From a structural perspective, 1, 4-butyrolactone (a cyclic ester) is quite different from other linear
Fig. 1. Chemical structure of cucurbit[7]uril (CB7).
fatty acid methyl esters. Its longer retention may result from its stronger interaction with CB7 possibly due to its better geometric match with the CB7 portals. This observation suggested the existence of shape selectivity of the stationary phase for GC separations. Aldehydes and alcohols are known to be prone to severe peak tailing in GC analysis, which harms their resolution with neighboring analytes and determination. However, as evident in Fig. 2, CB7-SG column exhibited good separation performance for these analytes with almost symmetrical peak shapes. Also, CB7SG column showed advantages over the CB6-GBIL column [13] for aldehydes and alcohols. Observably, the CB7-SG column produced relatively narrower peaks with better peak shapes than the CB6GBIL column. The good peak shapes of these analytes on CB7-SG column are of prime importance for baseline resolution of analytes in a complex mixture and their determination. The Grob test mixture containing 12 analytes is well recognized for comprehensive evaluation of separation performance of a GC column and a chromatographic system. Fig. 3 shows the GC separations of the mixture on the CB7-SG and commercial columns, suggesting some significant differences of CB7-SG column from the commercial one in terms of resolution, retention time and elution order of some analytes. CB7-SG column achieved a large resolution for the peak pair of 2,3-butanediol (peak 3) and methyl decanoate (peak 6), which coeluted on the commercial column. Interestingly, in contrast to the commercial column, the CB7-SG column extended the retentions for most of the analytes including alkanes, esters, amine and aniline but slightly decreased the retentions of 2,3-butanediol and 2-ethylhexanoic acid. Alkanes and esters of nonpolar or low polar nature were supposed to retain shorter on a polar GC column as they behaved on the commercial column by the general principle “like attracts like”. This finding suggests that the separation process on CB7-SG column is determined not only by polarity but also by other factors. Host-guest interaction may be one of them in this case since analytes with low polarity and a linear structure tend to stay in the hydrophobic cavity of a CB molecule. This explanation can also apply for the dramatically prolonged retention of dicyclohexylamine (peak 12) on CB7-SG column. The above findings demonstrate the great contribution of CB7 to the separation of different analytes via H-bond, dipole–dipole, dispersive and host–guest interactions. Moreover, the analytes of 2-ethylhexanoic acid and dicyclohexylamine often show distorted peak shapes when there are any acidic or basic active sites on the inner wall of a capillary column. As evident in Fig. 3, these analytes achieved symmetrical peak shapes on CB7SG column, suggesting the good column inertness. For example, dicyclohexylamine showed a symmetry factor of 0.95 on CB7-SG column but 1.30 on the commercial column. Additionally, comparison of CB7-SG column with CB7 and CB8 columns [14] revealed their selectivity difference. Take the Grob mixture for example. Noticeably, CB7-SG column well resolved all the 12 analytes with good peak shapes while the CB7 column coeluted two pairs of them, demonstrating its better separation
240
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
Table 1 McReynolds constants of the CB7-SG and commercial columns. Column
X
Y
Z
U
S
Average polarity
CB7-SG HP-INNOWAX
294 322
402 536
394 368
611 572
572 510
455 462
X : benzene, Y : 1-butanol, Z : 2-pentanone, U : nitropropane, S : pyridine. Temperature: 120 ◦ C
Table 2 Abraham system constants of the CB7-SG column. T (◦ C)
c
e
s
a
b
l
n
R2
F
80 100 120
−2.681(0.04) −2.719(0.03) −2.726(0.03)
0.079(0.03) 0.097(0.03) 0.111(0.03)
1.774(0.04) 1.694(0.04) 1.608(0.03)
1.796(0.04) 1.622(0.03) 1.442(0.03)
0.080(0.05) 0.048(0.05) 0.041(0.04)
0.522(0.01) 0.463(0.01) 0.407(0.01)
38 39 39
0.99 0.99 0.99
2757 3109 3221
n, R2 , and F stand for number of probe compounds used in multiple linear regression, coefficient of determination and Fisher coefficient, respectively. The values in parentheses are the standard deviation for the system constants.
performance than CB7 column. Moreover, elution order for some of the analytes also varied on the two columns. Generally, analytes of H-bonding nature, such as 2,3-butanediol, 2,6-dimethylphenol, 2,6-dimethylaniline and dicyclohexylamine, retained relatively longer on the CB7-SG column, which agrees well with its higher Z , U and S values in the McReynolds constants or higher s and a values in the Abraham system constants. Similar conclusion can also be reached as a result of comparison with CB8 column [14]. Furthermore, CB7-SG column achieved much higher resolution for nonanal and 1-octanol (peaks 4–5) (R = 4.2) in the Grob mixture than the CB6-GBIL column (R = 1.0) [13] and showed narrower peak shapes for the analytes, suggesting its better separation performance. In
brief, the CB7-SG column showed advantages over the reported CB7, CB8 and CB6-GBIL columns. On the basis of the above results, a more complex mixture comprising 19 analytes of different types, including n-alkanes, substituted benzenes, esters, ketones, aldehydes, alcohols, phenols, amide and anilines, was examined to further investigate the separation performance of CB7-SG stationary phase and the results are showed in Fig. 4. Though the CB7-SG column had a lower column efficiency, it achieved much better resolution for the analytes than the commercial column that failed to resolve three peak pairs of analytes, i.e., 1-octanol/benzonitrile (peaks 7/10), 2-nitrophenol/methyl dodecanoate (peaks 12/14) and
Fig. 2. GC separations of (a) esters, (b) aldehydes and (c) alcohols on CB7-SG column. Peaks for (a): (1) n-propyl acetate, (2) n-butyl acetate, (3) methyl hexanoate, (4) methyl heptanoate, (5) methyl nonanoate, (6) methyl decanoate, (7) methyl undecanoate, (8) methyl dodecanoate, (9) 1,4-butyrolactone. Peaks for (b): (1) butanal, (2) pentanal, (3) heptanal, (4) octanal, (5) nonanal, (6) benzaldehyde, (7) salicylaldehyde. Peaks for (c): (1)1-propanol, (2) 1-butanol, (3) 1-pentanol, (4) 1-hexanol, (5) 1-heptanol, (6) 1-octanol, (7) 1-nonanol, (8) 1-decanol, (9) 1-undecanol. Temperature program: 40 ◦ C (1 min) to 160 ◦ C at 5 ◦ C/min.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
241
Fig. 3. GC separations of the Grob test mixture on (a) CB7-SG column and (b) HP-INNOWAX column. Peaks: (1) n-decane, (2) n-undecane, (3) 2, 3-butanediol, (4) nonanal, (5) 1-octanol, (6) methyl decanoate, (7) methyl undecanoate, (8) 2-ethylhexanoic acid, (9) 2,6-dimethylphenol, (10) methyl dodecanoate, (11) 2,6-dimethylaniline, (12) dicyclohexylamine. Temperature program: 40 ◦ C (1 min) to 160 ◦ C at 5 ◦ C/min.
1,4-butanediol/2-chloroaniline (peaks 16/17). This demonstrated the high selectivity of the CB7-SG stationary phase. In addition, CB7-SG column exhibited different retention behaviors from the commercial column. Generally, the analytes exhibited longer retentions on CB7-SG column but their differences in retention times on the two columns (tR ) varied greatly. Noticeably, N,Ndimethylformamide achieved the largest tR value (>6 min), i.e., it prolonged its retention on CB7-SG column by more than 6 min. The analytes of a medium tR value (tR = 2–4 min) included n-alkanes, aldehydes, substituted benzenes, esters and anilines (2,6-dimethylaniline). Relatively, alcohols and phenols showed a tR value less than 2 min (cyclohexanol, 2-nitrophenol) or less than 1 min (1-octanol, 1,4-butanediol, phenol, p-cresol), suggesting their comparable retentions on the two columns. The above findings suggest that analytes such as alcohols and phenols, which can interact with a stationary phase mainly via dipole–dipole and H-bond interactions, exhibit similar retention behaviors on the two columns while other analytes such as n-alkanes, aldehydes, substituted benzenes, esters, anilines and amide exhibited stronger interactions with the CB7-SG stationary phase via multiple interactions that may also include shape selective interactions in addition to dipole–dipole, H-bond and dispersive interactions. These multiple interactions contributed to the unique separation performance of CB7-SG column, which differs greatly from the conventional polar stationary phase. The different
retention behaviors of analytes also affected their elution orders on CB7-SG column. Specifically, the longer retentions of benzonitrile (peak 10), methyl dodecanoate (peak 14) and 2-chloroaniline (peak 17) on CB7-SG column lead to the baseline resolution of the three peak pairs that coeluted on the commercial column. Additionally, cyclohexanol (peak 4) and ethoxybenzene (peak 5) reversed their elution order due to the extended retention of ethoxybenzene on CB7-SG column. Briefly, the above results demonstrate the unique separation performance of CB7-SG column and its advantages over the conventional polar column. 3.3. Separation repeatability and thermal stability Separation repeatability of CB7-SG column concerning intraday, inter-day and column-to-column repeatability was examined by separations of esters and aldehydes. Relative standard deviation (RSD%) in retention times of the analytes in the mixtures was used for the evaluation. The obtained results are shown in Table 3. RSD values of the analytes are in the range of 0.01–0.11% for intra-day, 0.04–0.41% for inter-day and 2.5–6.0% for columnto-column, respectively, demonstrating the good repeatability of CB7-SG columns in separation performance. Our previous work has demonstrated the good thermal stability of CBs columns either prepared by static coating or by sol-gel coating methods [13,14]. In this work, thermal stability of the
Fig. 4. GC separations of a mixture of 19 analytes on (a) CB7-SG column and (b) HP-INNOWAX column. Peaks: (1) n-undecane, (2) heptanal, (3) bromobenzene, (4) cyclohexanol, (5) ethoxybenzene, (6) 1,2–dichlorobenzene, (7) 1-octanol, (8) N,N-dimethylformamide, (9) benzaldehyde, (10) benzonitrile, (11) acetophenone, (12) 2-nitrophenol, (13) nitrobenzene, (14) methyl dodecanoate, (15) 2,6-dimethylaniline, (16) 1,4-butanediol, (17) 2-chloroaniline, (18) phenol, (19) p-cresol. Temperature program: 40 ◦ C to 160 ◦ C at 5 ◦ C/min.
242
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
Table 3 Separation repeatability of CB7-SG capillary column in retention times (tR , min) for separations of the indicated analytes. Analyte
n-Propyl acetate n-Butyl acetate Methyl hexanoate Methyl heptanoate Methyl nonanoate Methyl decanoate Methyl undecanoate Methyl dodecanoate 1,4-Butyrolactone Butanal Pentanal Heptanal Octanal Nonanal Benzaldehyde Salicylaldehyde
Intra-day
(n = 5)
Inter-day
(n = 5)
Column-to-column
(n = 3)
Mean
RSD%
Mean
RSD%
Mean
RSD%
3.51 5.17 7.31 9.43 13.62 15.59 17.46 19.24 20.76 2.55 4.00 7.96 10.17 12.32 14.57 16.42
0.09 0.05 0.03 0.01 0.01 0.02 0.04 0.04 0.11 0.06 0.07 0.07 0.04 0.03 0.01 0.01
3.51 5.16 7.30 9.41 13.60 15.56 17.43 19.20 20.72 2.56 4.02 7.97 10.18 12.33 14.57 16.41
0.35 0.34 0.25 0.21 0.19 0.19 0.21 0.21 0.36 0.40 0.41 0.20 0.15 0.09 0.04 0.04
3.59 5.28 7.61 9.93 14.26 16.40 18.22 20.46 20.62 2.55 4.05 8.35 10.66 13.06 14.97 16.82
5.8 5.4 4.9 3.9 3.2 3.1 2.9 2.6 3.3 5.4 6.0 4.0 3.5 3.7 3.0 2.5
CB7-SG column was further evaluated by GC separations of esters and aldehydes after the column was conditioned up to 220 ◦ C, 250 ◦ C and 280 ◦ C for 4 h, respectively. Table S3 shows that the RSD values of the analytes in retention times were in the range of 0.11–3.5%, suggesting the good thermal stability and separation repeatability of CB7-SG column after the column was subjected up to the indicated temperatures. Importantly, the column still exhibited excellent separation performance for the analytes without showing any appreciable baseline drift or any decrease in retention times and resolution, indicating that CB7-SG column can be used up to a higher temperature than the commercial polar column that is recommended to be used below 260/270 ◦ C. Good separation repeatability and thermal stability of CB7-SG column demonstrates its potential for practical use. 4. Conclusions This work demonstrates the unique GC separation performance of CB7-SG capillary column for a wide range of analytes and its good separation repeatability and thermal stability. The CB7-SG column differs from the conventional stationary phase of comparable polarity in retention behaviors and selectivity for some critical analytes such as diols, phenols, amines and anilines. Also, it shows advantages over the previously reported CB7 and CB6-GBIL columns for its high resolving ability and good peak shapes for the aforementioned analytes. The outstanding features of the CB7-SG stationary phase may inherently originate from the full integration of CB7, ionic liquid and sol-gel coating method. The established method in this work may advance the use of CBs in chromatographic separations and introduction of other promising candidates with similar problems to separation science. Acknowledgments The authors are grateful for the financial support by the National Natural Science Foundation of China (21075010) and the 111 Project B07012 in China. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2014.10.066.
References [1] J. Lagona, P. Mukhopadhyay, S. Chakrabarti, L. Isaacs, The cucurbit[n]uril family, Angew. Chem. Int. Ed. 44 (2005) 4844–4870. [2] D. Bardelang, K.A. Udachin, D.M. Leek, J.C. Margeson, G. Chan, C.I. Ratcliffe, J.A. Ripmeester, Cucurbit[n]urils (n=5-8): a comprehensive solid state study, Cryst. Growth Des. 11 (2011) 5598–5614. [3] E. Masson, X.X. Ling, R. Joseph, L. Kyeremeh-Mensah, X.Y. Lu, Cucurbituril chemistry: a tale of supramolecular success, RSC Adv. 2 (2012) 1213–1247. [4] J.W. Lee, S. Samal, N. Selvapalam, H.J. Kim, K. Kim, Cucurbituril homologues and derivatives: new opportunities in supramolecular chemistry, Acc. Chem. Res. 36 (2003) 621–630. [5] A.C. Bhasikuttan, H. Pal, J. Mohanty, Cucurbit[n]uril based supramolecular assemblies: tunable physico-chemical properties and their prospects, Chem. Commun. 47 (2011) 9959–9971. [6] M. Florea, W.M. Nau, Strong binding of hydrocarbons to cucurbituril probed by fluorescent dye displacement: a supramolecular gas-sensing ensemble, Angew. Chem. Int. Ed. 50 (2011) 9338–9342. [7] S.M. Liu, X. Li, C.T. Wu, Y.Q. Feng, Preparation and characterization of perhydroxyl-cucurbit[6] uril bonded silica stationary phase for hydrophilicinteraction chromatography, Talanta 64 (2004) 929–934. [8] L. Xu, S.M. Liu, C.T. Wu, Y.Q. Feng, Separation of positional isomers by cucurbit[7] uril-mediated capillary electrophoresis, Electrophoresis 25 (2004) 3300–3306. [9] F. Wei, Y.Q. Feng, Rapid determination of aristolochic acid I and II in medicinal plants with high sensitivity by cucurbit[7] uril-modifier capillary zone electrophoresis, Talanta 74 (2008) 619–624. [10] W.J. Cheong, J.H. Go, Y.S. Baik, S.S. Kim, E.R. Nagarajan, N. Selvapalam, Y.H. Ko, K. Kim, Preparation of cucurbituril anchored silica gel by cross polymerization and its chromatographic applications, Bull. Korean Chem. Soc. 29 (2008) 1941–1945. [11] L.S. Li, S.W. Wang, C. Liu, L.L. Xu, Preparation of glycoluril and perhydroxycucurbit[6] uril GC stationary phase and comparative studies on their separation performances, Acta. Chim. Sin. 65 (2007) 1855–1862. [12] L.S. Li, X.Y. He, H. Chen, Y.S. Fang, Preparation and property evaluation of cucurbit[7] uril coated celite as gas chromatography stationary phase, Chin. J. Appl. Chem. 29 (2012) 304–310. [13] L. Wang, X.G. Wang, M.L. Qi, R.N. Fu, Cucurbit[6] uril in combination with guanidinium ionic liquid as a new type of stationary phases for capillary gas chromatography, J. Chromatogr. A 1334 (2014) 112–117. [14] P. Zhang, S.J. Qin, M.L. Qi, R.N. Fu, Cucurbit[n]urils as a new class of stationary phases for gas chromatographic separations, J. Chromatogr. A 1334 (2014) 139–148. [15] K. Kim, N. Selvapalam, Y.H. Ko, K.M. Park, D. Kim, J. Kim, Functionalized cucurbiturils and their applications, Chem. Soc. Rev. 36 (2007) 267–279. [16] L. Isaacs, Cucurbit[n]urils: from mechanism to structure and function, Chem. Commun. 6 (2009) 619–629. [17] D.X. Wang, S.L. Chong, A. Malik, Sol-gel column technology for single-step deactivation, coating and stationary-phase immobilization in high-resolution capillary gas chromatography, Anal. Chem. 69 (1997) 4566–4576. [18] C. Shende, A. Kabir, E. Townsend, A. Malik, Sol-gel poly(ethylene glycol) stationary phase for high-resolution capillary gas chromatography, Anal. Chem. 75 (2003) 3518–3530. [19] M.M. Liang, M.L. Qi, C.B. Zhang, R.N. Fu, Peralkylated-beta-cyclodextrin used as gas chromatographic stationary phase prepared by sol–gel technology for capillary column, J. Chromatogr. A 1059 (2004) 111–119.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243 [20] K. Lu, W. Liu, M.L. Qi, R.N. Fu, Retention behaviors of novel ionic liquid stationary phases and their selectivity for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1475–1478. [21] L.Z. Qiao, K. Lu, M.L. Qi, R.N. Fu, Novel guanidinium-based ionic liquids as stationary phases for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1133–1136. [22] K. Lu, L.Z. Qiao, M.L. Qi, Y.H. Zhang, R.N. Fu, Selectivity of guanidinium ionic liquid for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1358–1360. [23] L.Z. Qiao, K. Lu, M.L. Qi, R.N. Fu, Separation performance of guanidinium-.based ionic liquids as stationary phases for gas chromatography, J. Chromatogr. A 1276 (2013) 112–119.
243
[24] M.H. Abraham, G.S. Whiting, R.M. Doherty, W.J. Shuely, Hydrogen bonding: XV. A new characterisation of the McReynolds 77-stationary phase set, J. Chromatogr. 518 (1990) 329–348. [25] M.H. Abraham, Scales of solute hydrogen-bonding: their construction and application to physicochemical and biochemical processes, Chem. Soc. Rev. 22 (1993) 73–83. [26] M.H. Abraham, C.F. Poole, S.K. Poole, Classification of stationary phases and other materials by gas chromatography, J. Chromatogr. A 842 (1999) 79–114. [27] C.F. Poole, S.K. Poole, Separation characteristics of wall-coated open-tubular columns for gas chromatography, J. Chromatogr. A 1184 (2008) 254–280.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Separation performance of cucurbit[7]uril in ionic liquid-based sol-gel coating as stationary phase for capillary gas chromatography Xiaogang Wang, Meiling Qi ∗ , Ruonong Fu Key Laboratory of Cluster Science, Ministry of Education of China, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials and School of Chemistry, Beijing Institute of Technology, Beijing 100081, China
a r t i c l e
i n f o
Article history: Received 27 February 2014 Received in revised form 26 September 2014 Accepted 21 October 2014 Available online 30 October 2014 Keywords: Cucurbit[7]uril Stationary phase Ionic liquid Sol–gel coating Capillary gas chromatography
a b s t r a c t Here we report the separation performance of a new stationary phase of cucurbit[7]uril (CB7) incorporated into an ionic liquid-based sol-gel coating (CB7-SG) for capillary gas chromatography (GC). The CB7-SG stationary phase showed an average polarity of 455, suggesting its polar nature. Abraham system constants revealed that its major interactions with analytes include H-bond basicity (a), dipole–dipole (s) and dispersive (l) interactions. The CB7-SG stationary phase achieved baseline separation for a wide range of analytes with symmetrical peak shapes and showed advantages over the conventional polar stationary phase that failed to resolve some critical analytes. Also, it exhibited different retention behaviors from the conventional stationary phase in terms of retention times and elution order. Most interestingly, in contrast to the conventional polar phase, the CB7-SG stationary phase exhibited longer retentions for analytes of lower polarity but relatively comparable retentions for polar analytes such as alcohols and phenols. The high resolving ability and unique retention behaviors of the CB7-SG stationary phase may stem from the comprehensive interactions of the aforementioned interactions and shape selectivity. Moreover, the CB7-SG column showed good peak shapes for analytes prone to peak tailing, good thermal stability up to 280 ◦ C and separation repeatability with RSD values in the range of 0.01–0.11% for intraday, 0.04–0.41% for inter-day and 2.5–6.0% for column-to-column, respectively. As demonstrated, the proposed coating method can simultaneously address the solubility problem with CBs for the intended purpose and achieve outstanding GC separation performance. © 2014 Elsevier B.V. All rights reserved.
1. Introduction For capillary gas chromatography (GC), a selective stationary phase and an efficient column preparation method are the requisites for a GC column to achieve high-resolution separations. Cucurbit[n]urils (CBs, n = 5–8, 10) are a family of pumpkin-shaped macrocyclic compounds with a hydrophobic cavity and two identical carbonyl-laced portals [1–3]. The hydrophobic interior provides a potential interaction site for nonpolar to low polar analytes while the ureido carbonyl groups at the portals offer selective interactions with polar analytes via dipole and H-bond interactions. Also, CBs possess high thermal stability. The unique structures and novel physicochemical properties of CBs have attracted great attention in supramolecular chemistry [4–6], separation science [7–14] and many other fields [15,16]. The aforementioned features also offer
∗ Corresponding author. Tel.: +86 10 68912668; fax: +86 10 68914780. E-mail addresses: [email protected], [email protected] (M. Qi). http://dx.doi.org/10.1016/j.chroma.2014.10.066 0021-9673/© 2014 Elsevier B.V. All rights reserved.
CBs as ideal candidates for capillary GC stationary phases. However, their poor solubility in common organic solvents for column preparation greatly limits their use for the intended purpose. Up to now, only two reports on using CBs as stationary phases for capillary GC are available [13,14], showing their high selectivity and thermal stability as stationary phases for capillary GC separations. Over the past decades, macrocyclic GC stationary phases such as cyclodextrins and calixarenes were reported. Since they have good solubility in dichloromethane or methanol commonly used in capillary column fabrication, it is relatively easier to fabricate a capillary GC column of these stationary phases with high column efficiency by static coating method. Often, they were reported to be used in a mixture with polysiloxanes for improving their GC separation performance. However, CBs are difficult to get a high-efficiency column by the same way due to their insolubility in the solvents mentioned above. Most recently, we successfully used CB6 as the GC stationary phase by simultaneously employing a guanidinium-based ionic liquid (GBIL) and a sol-gel coating
238
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
method [13]. The GBIL was N , N - (6 , 6 -dihydroxy)dihexyl-N, N, N , N -tetramethyl guanidinium bis(trifluoromethylsulfonyl)imide (DHHTMG-NTf2 ) (denoted as GBIL in the following). This GBIL was chosen after our many trials with different ionic liquids including imidazolium-based ILs [20] and guanidinium ILs [21–23] previously investigated in our laboratory since there were no any related reports available. Among the ILs tested, the indicated GBIL was found to be efficient in enhancing the CBs solubility in the sol solution and made it achievable to prepare the capillary GC column of CBs by sol–gel coating method. The proposed method proved to be efficient and feasible for use of CBs as new alternatives of GC stationary phases. Solgel coating method is known as a one-step coating method for preparation of capillary GC columns [17–19]. This method usually uses a tetraalkoxysilane as the sol–gel precursor, trifluoroacetic acid (TFA) as the catalyst, dichloromethane as the solvent, poly (methylhydrosiloxane) (PMHS) as the deactivation reagent, and a given stationary phase. The sol–gel process may involve the reactions of catalytic hydrolysis of the alkoxide precursor, polycondensation of the hydrolyzed products and immobilization of the stationary phase, leading to a three-dimensional (3D) sol-gel coating network with high physicochemical stability. Among the family members, CB7 (Fig. 1) has a larger dimension than CB6 in terms of diameter and cavity, and easier availability than CB8 that has a quite low yield in its synthesis. The larger dimension offers CB7 interactions with a wider range of analytes and may also exhibit different selectivity and resolving ability from CB6 in capillary GC separations. Here we report the fabrication of CB7 sol–gel (SG) capillary column and its GC separation performance for analytes of different variety. Meanwhile, a well-recognized commercial column with comparable polarity was also used for comparison. McReynolds constants and Abraham system constants were determined to characterize the molecular interactions of the CB7-SG stationary phase with analytes. Separation performance was examined by esters, aldehydes, alcohols, the Grob mixture and a more complex mixture containing 19 analytes of different types. Moreover, separation repeatability (intra-day, inter-day and column-to-column) and thermal stability were investigated as well. 2. Experimental 2.1. Materials and equipment Tetraethoxysilane (TEOS) was purchased from Fluka Chemika Co. (Steinheim, Switzerland). Poly(methylhydrosiloxane) (PMHS) and trifluoroacetic acid (TFA, 95%) were purchased from Alfa Aesar (Heysham, UK). Dichloromethane and the rest of chemicals used in this work were purchased from Beijing Chemical Reagent Company (Beijing, China). All the chemicals were of analytical grade and dissolved in dichloromethane. Untreated fused-silica capillary tubing (0.25 mm, i.d.) was purchased from Yongnian Ruifeng Chromatogram Apparatus Co., Ltd. (Hebei, China). A commercial HP-INNOWAX capillary column (10 m × 0.25 mm i.d., 0.25 m film thickness, Agilent Technologies) was used for comparison. An Agilent 7890A gas chromatograph equipped with a split/splitless injector and a flame ionization detector (FID) was used for the GC separations. All the separations were performed under the following GC conditions: nitrogen of high purity (99.999%) as carrier gas at a flow rate of 1 mL/min, injection port at 250 ◦ C, split injection mode with a split ratio at 30:1, FID detector at 300 ◦ C. A Hitachi S4800 scanning electron microscope (SEM) (Hitachi, Japan) was also used for the observation of the column inner surface morphology.
2.2. Methods 2.2.1. Capillary column preparation Column preparation for the CB7-SG capillary column followed the method in the reference [13] with minor modification. CB7 and GBIL were synthesized following the same methods provided in our previous work [14] and [13], respectively. Briefly, CB7 was synthesized by the reaction of glycoluril and paraformaldehyde in a concentrated HCl solution at room temperature and purified by multiple precipitation and solvent extraction [14]. GBIL was obtained by reaction of tetramethyl guanidine, tetrabutylammonium bromide and 6-bromo-1-hexanol at 75 ◦ C for 48 h and followed by anion exchange and purification via solvent extraction [13]. Prior to coating, a capillary column (10 m × 0.25 mm, i. d.) was successively pretreated with 1 mol/L NaOH and 0.1 mol/L HCl, rinsed with water and dried overnight at 120 ◦ C under nitrogen. After this, 100 L TEOS, 125 mg GBIL and 10 L PMHS were added into a vial and mixed well. Then, 15 mg CB7, 100 L of TFA and 250 L CH2 Cl2 were added and sonicated for 3 min. After another 250 L CH2 Cl2 were added, the mixture was sonicated for 3 min. Next, the resulting sol solution was pumped into the pretreated capillary column and stayed for 30 min at room temperature. After the extra sol solution was expelled from the column, the column was conditioned from 40 ◦ C for 30 min and then to 190 ◦ C at a rate of 1 ◦ C/min and held at the final temperature for 6 h. Column efficiency of the prepared column was determined by naphthalene at 100 ◦ C.
2.2.2. McReynolds constants and Abraham system constants McReynolds constants are often used to characterize the polarity of a GC stationary phase and its molecular interactions with analytes. Five probe compounds, namely benzene, n-butanol, 2-pentanone, nitropropane and pyridine, are used for the determination and serve to measure –electron interaction (X ), H-bond donor and acceptor interaction (Y ), H-bond acceptor interaction (Z ), dipole interactions (U ) and strong H-bond acceptor interaction (S ), respectively. For a GC column, McReynolds constants are measured by determining the differences of the five probes in the retention indices on the given stationary phase and squalane. The average of the differences is used for the polarity estimation. A GC stationary phase with average polarity over 400 is identified as polar phase. In this work, McReynolds constants of the CB7-SG column were determined at 120 ◦ C. Abraham system constants for a given stationary phase is determined on the basis of Abraham’s linear solvation parameter model, which can be used to quantitatively evaluate the individual intermolecular interactions between a GC stationary phase and probe compounds [24–27] and is set out below in the form for GC with the logarithm of retention factor (log k) as the dependent variable.
log k = c + eE + sS + aA + bB + lL The capital letters (E, S, A, B, L) are solute descriptors that represent different interactions (Table S1). For probe compounds, their values of solute descriptors are available in the reference [25] and the probe compounds used in this work and their solute descriptors are listed in Table S2. The lowercase letters (e, s, a, b and l) are system constants representing the contributions of a stationary phase to retention mechanism by the defined molecular interactions (Table S1). System constants for a given stationary phase can be calculated on the retention factors of probe compounds using multiple linear regression analysis.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
239
3. Results and discussion 3.1. McReynolds constants and Abraham system constants Characteristic parameters of the as-prepared CB7-SG column regarding column efficiency, McReynolds constants and Abraham system constants were determined. The column efficiency measured by naphthalene at 100 ◦ C was 2078 plates/m for CB7-SG column and 3727 plates/m for HP-INNOWAX column. CB7-SG column had lower column efficiency than the commercial column but higher efficiency than that of the statically coated CB7 column (1400 plates/m) [14]. Table 1 lists the McReynolds constants of CB7SG and commercial columns for comparison. As shown, CB7-SG column had an average polarity of 455, suggesting its polar nature, which was quite close to that of the commercial column. The high polarity of CB7-SG stationary phase can mainly be attributed to the polar GBIL used to improve the solubility of CB7 in the sol solution. Though CB7-SG stationary phase exhibited almost the same average polarity with the commercial phase, their individual molecular interactions differed greatly. The former exhibited higher Z , U and S values but lower Y value than the latter, suggesting its stronger dipole–dipole and H-bond acceptor interactions but weaker Hbond donator interaction with analytes. These differences may be indicative of their different separation performance. Table 2 lists the Abraham system constants of CB7-SG stationary phase at three temperatures of 80 ◦ C, 100 ◦ C and 120 ◦ C and shows that the stationary phase exhibits strong intermolecular interactions with analytes via dipole–dipole interaction (s), H-bond basicity (a) and dispersive interactions (l) but weak H-bond acidity (b) and –/n– interactions (e). This result is in good agreement with that of McReynolds constants. Additionally, it can be observed that the s, a, b and l values decreased with the rise of temperature but the e value slightly increased. The interactions related with s, a and b depend on the orientation of a stationary phase and solutes and become weakened as temperature rises due to the translational and rotational energy increase of the molecules. Dispersive interactions (l) are strongly distance-dependent and can be lessened with rising temperature owing to the increased distance between a stationary phase and a solute. With respect to e value, the slightly rising value as temperature increases may be attributed to that electrons are less tightly bound to atoms at higher temperatures. Moreover, the inner surface morphology of the CB7-SG capillary column was also examined by scanning electron microscopy (SEM) and the result is shown in Fig. S1 in Supplementary material. As can be observed, the inner coating showed an uneven and roughened surface inherently originating from the sol-gel coating process, which was favorable for enhancing the coating surface area and interactions of the stationary phase with analytes for GC separations. The film thickness of the sol–gel coating cannot be accurately determined due to its typical uneven surface. In this work, the film thickness of the sol–gel coating was roughly estimated to be around 0.8 m by the SEM image. Thicker coating layer is another feature of sol–gel coating procedure in contrast to static coating method. 3.2. Separation performance Separation performance of CB7-SG column was evaluated by separations of different mixtures including esters, aldehydes, alcohols, the Grob mixture and a more complex mixture containing 19 analytes of different types. Fig. 2 shows the GC separations of esters, aldehydes and alcohols on the CB7-SG column. All the analytes were baseline resolved and eluted in the order of their boiling points except 1, 4-butyrolactone that eluted much later than expected with regard to its boiling point. From a structural perspective, 1, 4-butyrolactone (a cyclic ester) is quite different from other linear
Fig. 1. Chemical structure of cucurbit[7]uril (CB7).
fatty acid methyl esters. Its longer retention may result from its stronger interaction with CB7 possibly due to its better geometric match with the CB7 portals. This observation suggested the existence of shape selectivity of the stationary phase for GC separations. Aldehydes and alcohols are known to be prone to severe peak tailing in GC analysis, which harms their resolution with neighboring analytes and determination. However, as evident in Fig. 2, CB7-SG column exhibited good separation performance for these analytes with almost symmetrical peak shapes. Also, CB7SG column showed advantages over the CB6-GBIL column [13] for aldehydes and alcohols. Observably, the CB7-SG column produced relatively narrower peaks with better peak shapes than the CB6GBIL column. The good peak shapes of these analytes on CB7-SG column are of prime importance for baseline resolution of analytes in a complex mixture and their determination. The Grob test mixture containing 12 analytes is well recognized for comprehensive evaluation of separation performance of a GC column and a chromatographic system. Fig. 3 shows the GC separations of the mixture on the CB7-SG and commercial columns, suggesting some significant differences of CB7-SG column from the commercial one in terms of resolution, retention time and elution order of some analytes. CB7-SG column achieved a large resolution for the peak pair of 2,3-butanediol (peak 3) and methyl decanoate (peak 6), which coeluted on the commercial column. Interestingly, in contrast to the commercial column, the CB7-SG column extended the retentions for most of the analytes including alkanes, esters, amine and aniline but slightly decreased the retentions of 2,3-butanediol and 2-ethylhexanoic acid. Alkanes and esters of nonpolar or low polar nature were supposed to retain shorter on a polar GC column as they behaved on the commercial column by the general principle “like attracts like”. This finding suggests that the separation process on CB7-SG column is determined not only by polarity but also by other factors. Host-guest interaction may be one of them in this case since analytes with low polarity and a linear structure tend to stay in the hydrophobic cavity of a CB molecule. This explanation can also apply for the dramatically prolonged retention of dicyclohexylamine (peak 12) on CB7-SG column. The above findings demonstrate the great contribution of CB7 to the separation of different analytes via H-bond, dipole–dipole, dispersive and host–guest interactions. Moreover, the analytes of 2-ethylhexanoic acid and dicyclohexylamine often show distorted peak shapes when there are any acidic or basic active sites on the inner wall of a capillary column. As evident in Fig. 3, these analytes achieved symmetrical peak shapes on CB7SG column, suggesting the good column inertness. For example, dicyclohexylamine showed a symmetry factor of 0.95 on CB7-SG column but 1.30 on the commercial column. Additionally, comparison of CB7-SG column with CB7 and CB8 columns [14] revealed their selectivity difference. Take the Grob mixture for example. Noticeably, CB7-SG column well resolved all the 12 analytes with good peak shapes while the CB7 column coeluted two pairs of them, demonstrating its better separation
240
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
Table 1 McReynolds constants of the CB7-SG and commercial columns. Column
X
Y
Z
U
S
Average polarity
CB7-SG HP-INNOWAX
294 322
402 536
394 368
611 572
572 510
455 462
X : benzene, Y : 1-butanol, Z : 2-pentanone, U : nitropropane, S : pyridine. Temperature: 120 ◦ C
Table 2 Abraham system constants of the CB7-SG column. T (◦ C)
c
e
s
a
b
l
n
R2
F
80 100 120
−2.681(0.04) −2.719(0.03) −2.726(0.03)
0.079(0.03) 0.097(0.03) 0.111(0.03)
1.774(0.04) 1.694(0.04) 1.608(0.03)
1.796(0.04) 1.622(0.03) 1.442(0.03)
0.080(0.05) 0.048(0.05) 0.041(0.04)
0.522(0.01) 0.463(0.01) 0.407(0.01)
38 39 39
0.99 0.99 0.99
2757 3109 3221
n, R2 , and F stand for number of probe compounds used in multiple linear regression, coefficient of determination and Fisher coefficient, respectively. The values in parentheses are the standard deviation for the system constants.
performance than CB7 column. Moreover, elution order for some of the analytes also varied on the two columns. Generally, analytes of H-bonding nature, such as 2,3-butanediol, 2,6-dimethylphenol, 2,6-dimethylaniline and dicyclohexylamine, retained relatively longer on the CB7-SG column, which agrees well with its higher Z , U and S values in the McReynolds constants or higher s and a values in the Abraham system constants. Similar conclusion can also be reached as a result of comparison with CB8 column [14]. Furthermore, CB7-SG column achieved much higher resolution for nonanal and 1-octanol (peaks 4–5) (R = 4.2) in the Grob mixture than the CB6-GBIL column (R = 1.0) [13] and showed narrower peak shapes for the analytes, suggesting its better separation performance. In
brief, the CB7-SG column showed advantages over the reported CB7, CB8 and CB6-GBIL columns. On the basis of the above results, a more complex mixture comprising 19 analytes of different types, including n-alkanes, substituted benzenes, esters, ketones, aldehydes, alcohols, phenols, amide and anilines, was examined to further investigate the separation performance of CB7-SG stationary phase and the results are showed in Fig. 4. Though the CB7-SG column had a lower column efficiency, it achieved much better resolution for the analytes than the commercial column that failed to resolve three peak pairs of analytes, i.e., 1-octanol/benzonitrile (peaks 7/10), 2-nitrophenol/methyl dodecanoate (peaks 12/14) and
Fig. 2. GC separations of (a) esters, (b) aldehydes and (c) alcohols on CB7-SG column. Peaks for (a): (1) n-propyl acetate, (2) n-butyl acetate, (3) methyl hexanoate, (4) methyl heptanoate, (5) methyl nonanoate, (6) methyl decanoate, (7) methyl undecanoate, (8) methyl dodecanoate, (9) 1,4-butyrolactone. Peaks for (b): (1) butanal, (2) pentanal, (3) heptanal, (4) octanal, (5) nonanal, (6) benzaldehyde, (7) salicylaldehyde. Peaks for (c): (1)1-propanol, (2) 1-butanol, (3) 1-pentanol, (4) 1-hexanol, (5) 1-heptanol, (6) 1-octanol, (7) 1-nonanol, (8) 1-decanol, (9) 1-undecanol. Temperature program: 40 ◦ C (1 min) to 160 ◦ C at 5 ◦ C/min.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
241
Fig. 3. GC separations of the Grob test mixture on (a) CB7-SG column and (b) HP-INNOWAX column. Peaks: (1) n-decane, (2) n-undecane, (3) 2, 3-butanediol, (4) nonanal, (5) 1-octanol, (6) methyl decanoate, (7) methyl undecanoate, (8) 2-ethylhexanoic acid, (9) 2,6-dimethylphenol, (10) methyl dodecanoate, (11) 2,6-dimethylaniline, (12) dicyclohexylamine. Temperature program: 40 ◦ C (1 min) to 160 ◦ C at 5 ◦ C/min.
1,4-butanediol/2-chloroaniline (peaks 16/17). This demonstrated the high selectivity of the CB7-SG stationary phase. In addition, CB7-SG column exhibited different retention behaviors from the commercial column. Generally, the analytes exhibited longer retentions on CB7-SG column but their differences in retention times on the two columns (tR ) varied greatly. Noticeably, N,Ndimethylformamide achieved the largest tR value (>6 min), i.e., it prolonged its retention on CB7-SG column by more than 6 min. The analytes of a medium tR value (tR = 2–4 min) included n-alkanes, aldehydes, substituted benzenes, esters and anilines (2,6-dimethylaniline). Relatively, alcohols and phenols showed a tR value less than 2 min (cyclohexanol, 2-nitrophenol) or less than 1 min (1-octanol, 1,4-butanediol, phenol, p-cresol), suggesting their comparable retentions on the two columns. The above findings suggest that analytes such as alcohols and phenols, which can interact with a stationary phase mainly via dipole–dipole and H-bond interactions, exhibit similar retention behaviors on the two columns while other analytes such as n-alkanes, aldehydes, substituted benzenes, esters, anilines and amide exhibited stronger interactions with the CB7-SG stationary phase via multiple interactions that may also include shape selective interactions in addition to dipole–dipole, H-bond and dispersive interactions. These multiple interactions contributed to the unique separation performance of CB7-SG column, which differs greatly from the conventional polar stationary phase. The different
retention behaviors of analytes also affected their elution orders on CB7-SG column. Specifically, the longer retentions of benzonitrile (peak 10), methyl dodecanoate (peak 14) and 2-chloroaniline (peak 17) on CB7-SG column lead to the baseline resolution of the three peak pairs that coeluted on the commercial column. Additionally, cyclohexanol (peak 4) and ethoxybenzene (peak 5) reversed their elution order due to the extended retention of ethoxybenzene on CB7-SG column. Briefly, the above results demonstrate the unique separation performance of CB7-SG column and its advantages over the conventional polar column. 3.3. Separation repeatability and thermal stability Separation repeatability of CB7-SG column concerning intraday, inter-day and column-to-column repeatability was examined by separations of esters and aldehydes. Relative standard deviation (RSD%) in retention times of the analytes in the mixtures was used for the evaluation. The obtained results are shown in Table 3. RSD values of the analytes are in the range of 0.01–0.11% for intra-day, 0.04–0.41% for inter-day and 2.5–6.0% for columnto-column, respectively, demonstrating the good repeatability of CB7-SG columns in separation performance. Our previous work has demonstrated the good thermal stability of CBs columns either prepared by static coating or by sol-gel coating methods [13,14]. In this work, thermal stability of the
Fig. 4. GC separations of a mixture of 19 analytes on (a) CB7-SG column and (b) HP-INNOWAX column. Peaks: (1) n-undecane, (2) heptanal, (3) bromobenzene, (4) cyclohexanol, (5) ethoxybenzene, (6) 1,2–dichlorobenzene, (7) 1-octanol, (8) N,N-dimethylformamide, (9) benzaldehyde, (10) benzonitrile, (11) acetophenone, (12) 2-nitrophenol, (13) nitrobenzene, (14) methyl dodecanoate, (15) 2,6-dimethylaniline, (16) 1,4-butanediol, (17) 2-chloroaniline, (18) phenol, (19) p-cresol. Temperature program: 40 ◦ C to 160 ◦ C at 5 ◦ C/min.
242
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243
Table 3 Separation repeatability of CB7-SG capillary column in retention times (tR , min) for separations of the indicated analytes. Analyte
n-Propyl acetate n-Butyl acetate Methyl hexanoate Methyl heptanoate Methyl nonanoate Methyl decanoate Methyl undecanoate Methyl dodecanoate 1,4-Butyrolactone Butanal Pentanal Heptanal Octanal Nonanal Benzaldehyde Salicylaldehyde
Intra-day
(n = 5)
Inter-day
(n = 5)
Column-to-column
(n = 3)
Mean
RSD%
Mean
RSD%
Mean
RSD%
3.51 5.17 7.31 9.43 13.62 15.59 17.46 19.24 20.76 2.55 4.00 7.96 10.17 12.32 14.57 16.42
0.09 0.05 0.03 0.01 0.01 0.02 0.04 0.04 0.11 0.06 0.07 0.07 0.04 0.03 0.01 0.01
3.51 5.16 7.30 9.41 13.60 15.56 17.43 19.20 20.72 2.56 4.02 7.97 10.18 12.33 14.57 16.41
0.35 0.34 0.25 0.21 0.19 0.19 0.21 0.21 0.36 0.40 0.41 0.20 0.15 0.09 0.04 0.04
3.59 5.28 7.61 9.93 14.26 16.40 18.22 20.46 20.62 2.55 4.05 8.35 10.66 13.06 14.97 16.82
5.8 5.4 4.9 3.9 3.2 3.1 2.9 2.6 3.3 5.4 6.0 4.0 3.5 3.7 3.0 2.5
CB7-SG column was further evaluated by GC separations of esters and aldehydes after the column was conditioned up to 220 ◦ C, 250 ◦ C and 280 ◦ C for 4 h, respectively. Table S3 shows that the RSD values of the analytes in retention times were in the range of 0.11–3.5%, suggesting the good thermal stability and separation repeatability of CB7-SG column after the column was subjected up to the indicated temperatures. Importantly, the column still exhibited excellent separation performance for the analytes without showing any appreciable baseline drift or any decrease in retention times and resolution, indicating that CB7-SG column can be used up to a higher temperature than the commercial polar column that is recommended to be used below 260/270 ◦ C. Good separation repeatability and thermal stability of CB7-SG column demonstrates its potential for practical use. 4. Conclusions This work demonstrates the unique GC separation performance of CB7-SG capillary column for a wide range of analytes and its good separation repeatability and thermal stability. The CB7-SG column differs from the conventional stationary phase of comparable polarity in retention behaviors and selectivity for some critical analytes such as diols, phenols, amines and anilines. Also, it shows advantages over the previously reported CB7 and CB6-GBIL columns for its high resolving ability and good peak shapes for the aforementioned analytes. The outstanding features of the CB7-SG stationary phase may inherently originate from the full integration of CB7, ionic liquid and sol-gel coating method. The established method in this work may advance the use of CBs in chromatographic separations and introduction of other promising candidates with similar problems to separation science. Acknowledgments The authors are grateful for the financial support by the National Natural Science Foundation of China (21075010) and the 111 Project B07012 in China. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2014.10.066.
References [1] J. Lagona, P. Mukhopadhyay, S. Chakrabarti, L. Isaacs, The cucurbit[n]uril family, Angew. Chem. Int. Ed. 44 (2005) 4844–4870. [2] D. Bardelang, K.A. Udachin, D.M. Leek, J.C. Margeson, G. Chan, C.I. Ratcliffe, J.A. Ripmeester, Cucurbit[n]urils (n=5-8): a comprehensive solid state study, Cryst. Growth Des. 11 (2011) 5598–5614. [3] E. Masson, X.X. Ling, R. Joseph, L. Kyeremeh-Mensah, X.Y. Lu, Cucurbituril chemistry: a tale of supramolecular success, RSC Adv. 2 (2012) 1213–1247. [4] J.W. Lee, S. Samal, N. Selvapalam, H.J. Kim, K. Kim, Cucurbituril homologues and derivatives: new opportunities in supramolecular chemistry, Acc. Chem. Res. 36 (2003) 621–630. [5] A.C. Bhasikuttan, H. Pal, J. Mohanty, Cucurbit[n]uril based supramolecular assemblies: tunable physico-chemical properties and their prospects, Chem. Commun. 47 (2011) 9959–9971. [6] M. Florea, W.M. Nau, Strong binding of hydrocarbons to cucurbituril probed by fluorescent dye displacement: a supramolecular gas-sensing ensemble, Angew. Chem. Int. Ed. 50 (2011) 9338–9342. [7] S.M. Liu, X. Li, C.T. Wu, Y.Q. Feng, Preparation and characterization of perhydroxyl-cucurbit[6] uril bonded silica stationary phase for hydrophilicinteraction chromatography, Talanta 64 (2004) 929–934. [8] L. Xu, S.M. Liu, C.T. Wu, Y.Q. Feng, Separation of positional isomers by cucurbit[7] uril-mediated capillary electrophoresis, Electrophoresis 25 (2004) 3300–3306. [9] F. Wei, Y.Q. Feng, Rapid determination of aristolochic acid I and II in medicinal plants with high sensitivity by cucurbit[7] uril-modifier capillary zone electrophoresis, Talanta 74 (2008) 619–624. [10] W.J. Cheong, J.H. Go, Y.S. Baik, S.S. Kim, E.R. Nagarajan, N. Selvapalam, Y.H. Ko, K. Kim, Preparation of cucurbituril anchored silica gel by cross polymerization and its chromatographic applications, Bull. Korean Chem. Soc. 29 (2008) 1941–1945. [11] L.S. Li, S.W. Wang, C. Liu, L.L. Xu, Preparation of glycoluril and perhydroxycucurbit[6] uril GC stationary phase and comparative studies on their separation performances, Acta. Chim. Sin. 65 (2007) 1855–1862. [12] L.S. Li, X.Y. He, H. Chen, Y.S. Fang, Preparation and property evaluation of cucurbit[7] uril coated celite as gas chromatography stationary phase, Chin. J. Appl. Chem. 29 (2012) 304–310. [13] L. Wang, X.G. Wang, M.L. Qi, R.N. Fu, Cucurbit[6] uril in combination with guanidinium ionic liquid as a new type of stationary phases for capillary gas chromatography, J. Chromatogr. A 1334 (2014) 112–117. [14] P. Zhang, S.J. Qin, M.L. Qi, R.N. Fu, Cucurbit[n]urils as a new class of stationary phases for gas chromatographic separations, J. Chromatogr. A 1334 (2014) 139–148. [15] K. Kim, N. Selvapalam, Y.H. Ko, K.M. Park, D. Kim, J. Kim, Functionalized cucurbiturils and their applications, Chem. Soc. Rev. 36 (2007) 267–279. [16] L. Isaacs, Cucurbit[n]urils: from mechanism to structure and function, Chem. Commun. 6 (2009) 619–629. [17] D.X. Wang, S.L. Chong, A. Malik, Sol-gel column technology for single-step deactivation, coating and stationary-phase immobilization in high-resolution capillary gas chromatography, Anal. Chem. 69 (1997) 4566–4576. [18] C. Shende, A. Kabir, E. Townsend, A. Malik, Sol-gel poly(ethylene glycol) stationary phase for high-resolution capillary gas chromatography, Anal. Chem. 75 (2003) 3518–3530. [19] M.M. Liang, M.L. Qi, C.B. Zhang, R.N. Fu, Peralkylated-beta-cyclodextrin used as gas chromatographic stationary phase prepared by sol–gel technology for capillary column, J. Chromatogr. A 1059 (2004) 111–119.
X. Wang et al. / J. Chromatogr. A 1371 (2014) 237–243 [20] K. Lu, W. Liu, M.L. Qi, R.N. Fu, Retention behaviors of novel ionic liquid stationary phases and their selectivity for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1475–1478. [21] L.Z. Qiao, K. Lu, M.L. Qi, R.N. Fu, Novel guanidinium-based ionic liquids as stationary phases for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1133–1136. [22] K. Lu, L.Z. Qiao, M.L. Qi, Y.H. Zhang, R.N. Fu, Selectivity of guanidinium ionic liquid for capillary gas chromatography, Chin. Chem. Lett. 21 (2010) 1358–1360. [23] L.Z. Qiao, K. Lu, M.L. Qi, R.N. Fu, Separation performance of guanidinium-.based ionic liquids as stationary phases for gas chromatography, J. Chromatogr. A 1276 (2013) 112–119.
243
[24] M.H. Abraham, G.S. Whiting, R.M. Doherty, W.J. Shuely, Hydrogen bonding: XV. A new characterisation of the McReynolds 77-stationary phase set, J. Chromatogr. 518 (1990) 329–348. [25] M.H. Abraham, Scales of solute hydrogen-bonding: their construction and application to physicochemical and biochemical processes, Chem. Soc. Rev. 22 (1993) 73–83. [26] M.H. Abraham, C.F. Poole, S.K. Poole, Classification of stationary phases and other materials by gas chromatography, J. Chromatogr. A 842 (1999) 79–114. [27] C.F. Poole, S.K. Poole, Separation characteristics of wall-coated open-tubular columns for gas chromatography, J. Chromatogr. A 1184 (2008) 254–280.