Fast Separations at Elevated Temperatures on Polybutadiene

Jonathan D. Thompson, James S. Brown, and Peter W. Carr ... John G. Dorsey and William T. Cooper , Barbara A. Siles , Joe P. Foley , Howard G. Barth...
0 downloads 0 Views 193KB Size
Anal. Chem. 1997, 69, 3884-3888

Fast Separations at Elevated Temperatures on Polybutadiene-Coated Zirconia Reversed-Phase Material Jianwei Li,† Yue Hu, and Peter W. Carr*

Department of Chemistry, University of Minnesota, Kolthoff and Smith Halls, 207 Pleasant Street, SE, Minneapolis, Minnesota 55455

This paper describes the results of thermal stability studies of polybutadiene-coated zirconia reversed phases and application to fast HPLC separations at elevated temperatures and high flow rates. The thermal stability of the material was evaluated at temperatures up to 200 °C, and the rapid analyses of polyaromatic hydrocarbons and typical reversed-phase test mixtures were carried out at 100 °C and a flow rate of 5 mL/min. We found that the material is completely stable at 200 °C for at least 1300 column volumes. Analysis time can be decreased about 18-fold at high temperatures and flow rates without any significant loss in resolution relative to that at conventional temperatures and normal flow rates. For the separation of a five-component reversed-phase test mixture, the analysis time was only 50 s. In a series of recent publications,1,2 we reported on the effect of temperature on the thermodynamic and dynamic properties of polybutadiene (PBD)-coated zirconia-based reversed-phase supports. We found that high column temperature can significantly decrease the analysis time and column back-pressure and can, in some instances, improve chromatographic selectivity.1,2 Moreover, elevated temperatures can increase column efficiency by as much as 30%, particularly at high flow rates.1 Improvements in selectivity at elevated temperatures have also been demonstrated in the work of Hancock et al.3,4 They optimized the separation of peptides and proteins by varying both mobile phase composition and temperature on the “sterically protected” C8 and C18 phases. Thus, it is becoming increasingly better recognized that temperature can be an effective separation variable in liquid chromatography, and Snyder et al. have published an interesting series of papers on the optimization of HPLC separations by varying both mobile phase composition and temperature.5-8 † Current address: 3M Pharmaceuticals, 3M Center, Building 270-4S-02, St. Paul, MN 55144-1000. (1) Li, J.; Carr, P. W. Anal. Chem. 1997, 69, 837-843. (2) Li, J.; Carr, P. W. Anal. Chem. 1997, 69, 2202-2206. (3) Hancock, W. S.; Chloupek, R. C.; Kirkland, J. J.; Snyder, L. R. J. Chromatogr. 1994, 686, 31-43. (4) Chloupek, R. C.; Hancock, W. S.; Marchylo, B. A.; Kirkland, J. J.; Boyes, B. E.; Snyder, L. R. J. Chromatogr. 1994, 686, 45-59. (5) Zhu, P. L.; Snyder, L. R.; Dolan, J. W.; Djordjevic, N. M.; Hill, D. W.; Sander, L. C.; Waeghe, J. T. J. Chromatogr. 1996, 756, 21-39. (6) Zhu, P. L.; Dolan, J. W.; Snyder, L. R. J. Chromatogr. 1996, 756, 41-50. (7) Zhu, P. L.; Dolan, J. W.; Snyder, L. R.; Hill, D. W.; Van Heukelem, L.; Waeghe, J. T. J. Chromatogr. 1996, 756, 51-62. (8) Zhu, P. L.; Dolan, J. W.; Snyder, L. R.; Djordjevic, N. M.; Hill, D. W.; Lin, J. T.; Sander, L. C.; Waeghe, J. T. J. Chromatogr. 1996, 756, 63-72.

3884 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

Fast separations at high temperatures on conventional columns (e.g., 100 mm × 4.6 mm i.d.) and particles (e.g., 5 µm) become feasible for two reasons. First, the logarithm of capacity factor in RPLC is proportional to the inverse of the column temperature; therefore, high temperature can significantly decrease analysis time, as observed in the majority of reversed-phase separations.9,10 Second, because elevated temperatures can significantly decrease column back-pressure and improve column performance, we can further decrease analysis time by operating at very high temperatures and flow rates. Horva´th et al.11-13 used pellicular polystyrene-divinylbenzene and ODS-silica stationary phases to separate proteins at temperatures up to 120 °C. Analysis time was decreased to less than 10 s. Accordingly, it appears that very fast separations can be achieved at elevated temperatures and high flow rates with conventional LC instrumentation (columns, detectors, etc.) without sacrificing the ruggedness of the separations. However, high-temperature separations on conventional bondedphase silica gels shorten their lifetimes considerably. The objective of the present study is to demonstrate the possibility of fast separations at high temperatures and flow rates on conventional HPLC equipment (e.g., pumps and detectors) using PBD-coated zirconia as the reversed-phase material. Such separations are practical due to the very high chemical and thermal stability of these materials. EXPERIMENTAL SECTION PBD-Coated Zirconia Particles and Columns. Two batches of zirconia particles (different in size) were used in this work as the substrate for the preparation of PBD-coated zirconia materials. The preparation, physical characterization, and packing of these particles were the same and have been detailed elsewhere.14,15 The physical characteristics of these materials were determined by BET measurements and carbon elemental analysis, as detailed elsewhere.14 The columns used were all 100 mm × 4.6 mm i.d. in all cases. Table 1 shows the physical characteristics of these materials. Reagents. All chemicals used were obtained from commercial sources and were reagent grade or better, unless noted below. The organic solvent was ChromAR HPLC grade acetonitrile (ACN, (9) Ooms, B. LC-GC 1996, 14, 306-323. (10) Keukelem, L. V.; Lewitus, A. J.; Kana, T.; Craft, N. E. Mar. Ecol.: Prog. Ser. 1994, 114 (3), 303-313. (11) Chen, H.; Horva´th, Cs. J. Chromatogr. 1995, 705, 3-20. (12) Chen, H.; Horva´th, Cs. Anal. Methods Instrum. 1993, 1 (4), 213-222. (13) Kalghatgi, K.; Horva´th, Cs. J. Chromatogr. 1988, 443, 343-353. (14) Li, J.; Carr, P. W. Anal. Chem. 1996, 68, 2857-2868. (15) Sun, L.; McCormick, A. V.; Carr, P. W. J. Chromatogr. 1994, 658, 465473. S0003-2700(97)00506-4 CCC: $14.00

© 1997 American Chemical Society

Table 1. Physical Characteristics of the Materials Used in This Studya

materialb

particle sizec (µm)

carbon loadd (wt %)

surface areae (m2/g)

pore volumef (mL/g)

pore diameterg (Å)

bare (PICA7)h bare (PICA10)h PBD/zirconiai PBD/zirconiaj PBD/zirconiak

2.5 4.5 2.5 2.5 4.5

0 0 5.6 6.1 2.1

34.3 28.4 21.4 15.0 23.0

0.17 0.13 0.09 0.10 0.09

220 190 256 266 161

a BET nitrogen porosimetry and carbon elemental analysis results for the materials used in this study. b Bare and PBD-coated zirconia. c Average particle size. d Carbon load as percent weight. e Total surface area per gram. f Total pore volume per gram. g Average pore diameter. h Bare zirconia used to prepare these phases. i PBD-zirconia (PICA7) used to carry out the thermal stability study. j PBD-zirconia (PICA7) used to carry out the separation of PAHs. k PBD-zirconia (PICA10) used to carry out the separation of a reversed-phase test mixture.

Mallinckrodt Chemical Co., Paris, KY). DI water was filtered through a 0.45 µm filter (Gelman Sciences Inc., Ann Arbor, MI) and then boiled to remove carbon dioxide before use. All solvents were filtered a second time with a 0.45 µm filtration disk. Solutes used in this study were benzene, toluene, ethylbenzene, npropylbenzene, n-butylbenzene, uracil, p-nitroaniline, methyl benzoate, and phenetole (Aldrich Chemical Co., Milwaukee, WI). Standard HPLC samples of polyaromatic hydrocarbons (16 solutes) were obtained from Supelco (Supelco, Inc., Bellefonte, PA). Chromatographic Apparatus. All chromatographic experiments were carried out on a fully automated Hewlett Packard 1090 liquid chromatograph with an autosampler, temperature controller, UV detector, and Chemstation (Hewlett Packard S.A., Wilmington, DE). The mobile phase composition was controlled by mixing two channels. Channel A was usually 50% ACN and 50% water (premixed and degassed), while channel B was either 100% water or 100% ACN. This approach allowed us to reduce the air bubbles generated by mixing ACN and water, particularly at high temperatures. It should be noted that the only extra piece of equipment needed for this experiment is a pressure regulator, and it was always attached to the detector outlet to provide back-pressure to eliminate the formation of bubbles in the mobile phase and thereby stabilize the baseline. Because the pressure regulator was attached behind the detector, it would not cause extracolumn broadening. The temperatures used to evaluate the column stability were greater than 100 °C. To achieve them, we used an apparatus provided by Systec (Systec, Inc., Minneapolis, MN) because the chromatograph oven only provides a maximum temperature of 100 °C. The mobile phase was preheated to the same temperature as the column oven temperature before it reached the column. The hot mobile phase exiting the column was immediately cooled by ice water before it reached the detector. We emphasize that, except for the use of the pressure regulator to stabilize the baseline, we are using conventional LC instrumentation (pumps, detectors, etc.) to do the fast separations. Chromatographic Conditions. Column Stability. A 5.6% (w/ w) carbon column was used to evaluate the stability of PBD-coated zirconia by chromatography of a homologous series of alkylbenzenes (benzene to n-butylbenzene) at a mobile phase composition of 40% (v/v) ACN. The temperatures used were 40, 150, and 200 °C, respectively. Solute concentrations were around 2 mg/mL,

Figure 1. Effect of mobile phase volumes on the thermal stability of PBD-coated zirconia. See Experimental Section for conditions. The vertical dashed line indicates the time the mobile phase reservoir was refilled. Solutes/symbols: 9, benzene; b, toluene; 2, ethylbenzene; 1, n-propylbenzene; [, n-butylbenzene.

the injection volumes were usually 1 µL, the flow rate was 1 mL/ min, and the detection wavelength was set at 254 nm. The initial temperature was set to 40 °C. The temperature was then raised to 150 °C, and the column was operated for more than 1350 column volumes. The temperature was then returned to 40 °C to see if the initial retention of solutes was recovered. The column temperature was increased to 200 °C, and the column was operated for more than 1300 column volumes and then cooled again to 40 °C. Separation of PAH Mixture. A 6.0% (w/w) column was used to separate the PAH mixture. The mobile phase composition was 50% ACN (premixed and degassed); the solute concentrations were varied from 0.1 to 2 mg/mL. The injected volume was 2 µL, and the flow rate was 1.0 mL/min at 30 °C and 5 mL/min at 100 °C. The detection wavelength was set to 254 nm. Separation of a Reversed-Phase Test Mixture. A 2.1% (w/w) column was used to separate a typical reversed-phase test mixture (uracil, p-nitroaniline, methyl benzoate, phenetole, and toluene). The mobile phase composition was 20% ACN (premixed and degassed); solute concentrations were varied from 0.1 to 2 mg/ mL. The injected volume was 2 µL, and the flow rate was 1.0 mL/min at 30 °C and 5 mL/min at 100 °C. The detection wavelength was set to 254 nm. RESULTS AND DISCUSSION Thermal Stability of PBD-Coated Zirconia. We demonstrated previously that PBD-coated zirconia is thermally stable at 100 °C.1 Because elevated temperatures can significantly reduce analysis time, decrease column back-pressure, and improve column efficiency (particularly at high flow rates),1 the use of high column temperatures may enable fast separations on conventional equipment by the use of high flow rates. The higher the temperature that can be achieved, the faster the separation can be obtained. Therefore, we wanted to re-evaluate the thermal stability of PBD-coated zirconia at temperatures above 100 °C (up to 200 °C). Figure 1 shows the column stability at 200 °C. It is evident that the material is thermally stable for at least 1300 column volumes at 200 °C. The vertical dashed line represents the time at which the mobile phase reservoirs were refilled. It is also obvious that the retention of the probe solutes increases slightly with time. This is probably due to the decrease in the amount of Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

3885

ACN in the ACN/water mixture because the mobile phase was purged with helium during the entire experiment. Furthermore, the temperature control apparatus was not very precise, and a slight temperature variation is very possible. We intentionally stopped the test at 1300 column volumes, but we believe that PBDcoated zirconia is fully stable at 200 °C toward loss of the active stationary phase. Effect of High Temperatures on the Recovery of Column Efficiency. The column used for the thermal stability study initially had a plate count of about 9000 (or about 90 000 plates/ m) when freshly packed. It was used for many experiments. The performance of the column deteriorated as it was being used due to gradual contamination. We washed the column with pure ACN to recover the efficiency of the column, but no significant improvement was seen. However, with some surprise, we noted considerable improvement in plate count upon exposure to elevated temperature. Table 1 shows the results of the column efficiency and retention reproducibility at 40 °C and at high temperatures. Before the column temperature was raised to 150 °C, the average column efficiency was about 3000 plates; however, after the column was operated at 150 °C for more than 1350 column volumes, the column efficiency at 40 °C became, on average, about 6000 plates. There is no decrease in retention, but rather a slight increase was observed. Moreover, after the column was heated at 200 °C for more than 1300 column volumes, the retention again increased slightly, and the efficiency at 40 °C jumped to about 9000 plates. Although we do not fully understand why retention increases, the data indicate that there is no loss in the stationary phase at temperatures up to 200 °C. Most importantly, the column efficiency improved significantly compared to the value (3000 plates) prior to being subjected to the thermal stress, and the initial efficiency was fully recovered. Currently, we believe that the improvement in the column efficiency could be due to the removal of strongly adsorbing materials. Effect of Flow Rate on Column Efficiency at High Temperatures. Although we have shown previously that the use of high temperatures can improve column efficiency by as much as 30%, this improvement takes place mainly at a high linear velocity region (a decrease in the C term of the Knox equation).1 This result is confirmed by the data in Table 2. Table 2 indicates that the average column efficiency at 200 °C is only about 2730 plates; when the column was cooled to 40 °C, its efficiency jumped to about 8680. All flow rates were 1 mL/min. This is because, as demonstrated in a previous study,1 the longitudinal molecular diffusion (the B term of the Knox equation) is more pronounced at high temperatures (the B term is 8 at 100 °C and 2 at 25 °C1) and low flow rates due to the interaction of molecules with the column walls, resulting in low efficiency at high temperatures relative to that at low temperatures. Similar results were also reported by other authors.16,17 However, at a high linear velocity region (where the C term is dominant), this contribution (the B term) is no longer significant. Thus, it will be advantageous to operate the columns at high temperatures and flow rates; this will not only improve the speed of separations but also increase the column efficiency due to minimization of the contribution of longitudinal diffusion. (16) Welsch, T.; Schmid, M.; Kutter, J.; Ka´lma´n, A. J. Chromatogr. 1996, 728, 299-306. (17) Warren, F. V.; Bidlingmeyer, B. A. Anal. Chem. 1988, 60, 2821-2824.

3886 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

Table 2. Effect of Temperature on Column Performancea solute ethylpropylbutylbenzene toluene benzene benzene benzene average initialb after 150 °Cc after 200 °Cd

2.82 2.89 3.02

Retention Time (min) 4.17 6.42 10.79 4.38 6.89 11.89 4.64 7.37 12.80

18.96 21.37 23.20

nah nah nah

initialb after 150 °Cc after 200 °Cd at 150 °Ce at 200 °Cf

2820 5710 7770 4370 3280

Column Efficiencyg 3170 3360 3610 6230 5880 5830 8790 8710 9460 5160 4770 4150 3520 2920 2250

3810 5890 na 3470 1690

3350 5910 8680 4380 2730

a Flow rate was 1 mL/min in all cases. b Initial operation of column at 40 °C. c After column was operated at 150 °C for >1350 column volumes. d After column was operated at 200 °C for >1300 column volumes. e Column efficiency at 150 °C and 1 mL/min. f Column efficiency at 200 °C and 1 mL/min. g Computed by 2π(tRH/A)2 (tR is the retention time, H is the peak height, and A is the peak area). h Does not apply.

Figure 2. Chromatograms of polyaromatic hydrocarbons. See Experimental Section for conditions. Plot A is the chromatogram at 30 °C and 1 mL/min, and plot B is the chromatogram at 100 °C and 5 mL/min. The inserted figures in plots A and B are “blow-ups” for the corresponding boxed regions. Solutes: 1, naphthalene; 2, acenaphthylene; 3, acenaphthene; 4, fluorene; 5, phenanthrene; 6, anthracene; 7, fluoranthene; 8, pyrene; 9, benzo[a]anthracene; 10, chrysene; 11, benzo[b]fluoranthene; 12, benzo[k]fluoranthene; 13, benzo[a]pyrene; 14, dibenzo[a,h]anthracene; 15, benzo[ghi]perylene; 16, indeno[1,2,3-cd]pyrene.

Fast Separations on PBD-Coated Zirconia. Figure 2 shows the separation of a mixture of 16 polyaromatic hydrocarbons (PAHs). These solutes are all nonpolar (hydrophobic), nonionizable species; thus, the sole difference in terms of retention between these solutes is their relative hydrophobicity. Figure 2A shows the chromatogram at 30 °C. The analysis time is 70 min, and the peaks are asymmetric, particularly the late-eluted peaks. We note that peaks 14-16 are very asymmetric and difficult to

Figure 3. Structures of two pairs of PAH structural isomers. First pair: benzo[a]anthracene (9) and chrysene (10). Second pair: benzo[b]fluoranthene (11) and benzo[k]fluoranthene (12).

detect. Two pairs of structural isomers (9/10 and 11/12, see Figure 3) are not resolved, and this is in agreement with our previous results that PBD-coated zirconia does not have the solute shape selectivity14 exhibited by conventional polymeric alkyl silane bonded phases. All other 14 peaks are fully resolved. As the temperature is raised to 100 °C and the flow rate is increased from 1 to 5 mL/min, the analysis time decreases from 70 to about 4 min (Figure 2B). This is an 18-fold decrease in analysis time. There is essentially no loss in resolution, except for the 3/4 pair. The detector response at 100 °C relative to that at 30 °C is improved significantly, especially for the late-eluted bands, due to the reduced retention times and sharper peaks. This is evidently important in trace analysis. It is noted that the shoulders on peaks 5 and 6 could be due to the air injected or bubbles in the mobile phase because it was not cooled prior to reaching the detector.18 We also point out (see Figure 2) that the peak symmetry becomes significantly better at higher temperatures. To make this point clearer, we replotted peak 13 from Figure 2 in Figure 4. It can be seen in this figure that the peak asymmetry is 3.2 at 30 °C but only 1.5 at 100 °C, even though the flow rate is 5 mL/ min. The improvement in peak symmetry is due not only to the decrease in the retention but, more likely, to the fast diffusion of solute in the stationary phase zone.19 It is also evident that the noise level is high at the higher temperature because we did not cool the mobile phase before it reached the detector. The baseline obtained at 200 °C in the thermal stability study was excellent because we cooled the effluent before it entered the detector. Figure 5 shows the separation of a typical reversed-phase test mixture. At 30 °C and 1 mL/min (Figure 5A), all peaks are fully resolved, and the analysis time is about 11 min. However, when the separation conditions were changed to a temperature of 100 °C and a flow rate of 5 mL/min (Figure 5B), the analysis time was decreased to about 50 s. The resolution at 100 °C is still very acceptable. The p-nitroaniline (solute 2) peak moves close to uracil (solute 1) because it is nearly unretained. This example clearly demonstrates that rapid separations can be practically achieved on conventional LC instrumentation based on the use of high temperatures and high linear velocities using a thermally stable stationary phase. Although we used the two samples to show the possibility of high-speed separation, there was no significant decrease in resolution. For other samples, high temperatures might either increase or decrease the selectivity. However, we point out that (18) Dolan, J. W.; Marchand, D. H.; Cahill, S. A. LC-GC, 1997, 15, 328-331. (19) Li, J.; Carr, P. W. Anal. Chem. 1997, 69, 2193-2201.

Figure 4. Comparison of peak asymmetry of benzo[a]pyrene at different temperatures. Benzo[a]pyrene is solute 13 in Figure 2. Plot A is the peak at 30 °C and 1 mL/min, and plot B is the peak at 100 °C and 5 mL/min. The peak asymmetry is measured at 10% peak height.

Figure 5. Chromatograms of a reversed-phase test mixture. See Experimental Section for conditions. Plot A is the chromatogram at 30 °C and 1 mL/min, and plot B is the chromatogram at 100 °C and 5 mL/min. Solutes: 1, uracil; 2, p-nitroaniline; 3, methyl benzoate; 4, phenetole; 5, toluene.

temperature is being used increasingly as a separation variable, and elevated temperatures can significantly improve selectivity in some situations.5-8 Thus, we conclude that high-speed analysis at high temperatures and flow rates does not necessarily entail a diminution in selectivity, but rather it is, in some cases, possible to simultaneously improve both selectivity and speed of analysis. Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

3887

CONCLUSIONS We have demonstrated in this study that PBD-zirconia is fully stable up to a temperature of 200 °C for at least 1300 column volumes. Rapid separations can be achieved at elevated temperatures and flow rates on this material. Analysis time can be improved by 18-fold at the high temperatures and flow rates with only minor losses in resolution relative to that at the lower temperatures and normal flow rates. Analysis time is less than 1 min (50 s) for the separation of a typical reversed-phase test mixture.

3888

Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

ACKNOWLEDGMENT The authors acknowledge the financial support by Grant GM 45988-05 from the National Institutes of Health.

Received for review May 16, 1997. Accepted July 18, 1997.X AC9705069 X

Abstract published in Advance ACS Abstracts, September 1, 1997.