Nanoparticles as Pseudostationary Phase in Capillary

Anal. Chem. 2002, 74, 4595-4601

Accelerated Articles

Nanoparticles as Pseudostationary Phase in Capillary Electrochromatography/ESI-MS Peter Viberg, Magnus Jornten-Karlsson,† Patrik Petersson,† Peter Spe´gel, and Staffan Nilsson*

Technical Analytical Chemistry, Center for Chemistry and Chemical Engineering, Lund University, P.O. Box 124, 221 00 Lund, Sweden

A novel technique that uses polymer nanoparticles as pseudostationary phase in capillary electrochromatography with electrospray ionization mass spectrometry detection is described. A continuous full filling technique in which the nanoparticles were suspended in the entire electrolyte volume as well as a conventional partial filling technique is presented. No nanoparticles entered the mass spectrometer, which was fitted with an orthogonal electrospray interface, despite the continuous flow of nanoparticles into the interface. Nanoparticles (average diameter 160 nm) were prepared from methacrylic acid, methyl methacrylate, and trimethylolpropane trimethacrylate by utilizing a precipitation polymerization technique. Salbutamol, nortriptyline, and diphenhydramine were used as analytes. The interaction between analytes and nanoparticles was found to be predominantly ionic.

The development of a novel capillary electrochromatography/ mass spectrometry (CEC/MS) method using polymer nanoparticles as pseudostationary phase is described. Amines were chosen as analytes in the experiments, and the focus was to introduce the new technique by proving interaction between the analytes and nanoparticles and to show that electrospray ionization (ESI)MS can be used for detection despite a continuous flow of nanoparticles into the CEC/ESI-MS interface. Capillary electrochromatography is a powerful separation technique, which combines the efficiency of capillary electrophoresis (CE) with the selectivity of liquid chromatography.1,2 * To whom correspondence should be addressed. E-mail: Staffan.nilsson@ teknlk.lth.se. Fax: +46 46 222 45 25. † Permanent address: AstraZeneca R&D Lund, Lund, Sweden. (1) Altria, K. D.; Smith, N. W.; Turnbull, C. H. Chromatographia 1997, 46, 664. (2) Cikalo, M. G.; Bartle, K. D.; Robson, M. M.; Myers, P.; Euerby, M. R. Analyst 1998, 123, 87R-102R. 10.1021/ac0204045 CCC: $22.00 Published on Web 08/10/2002

© 2002 American Chemical Society

Commonly used interaction phases in CEC are stationary phases consisting of packed particle beds3-5 or monolithic beds.6 Gold particles have been used in open-tubular CEC.7 The gold particles, dissolved in the run buffer, were adsorbed to the capillary wall and functioned as interaction phase. Another way of improving the selectivity in CE is to use micelles as interaction phase, i.e., micellar electrokinetic chromatography (MEKC).8,9 MEKC involves solubilization of surfactants above the critical micelle concentration (cmc) in the electrolyte to form micelles into which the analytes may migrate. Other modifiers that can interact with the analytes can also be added to the electrolyte to affect the separation of the analytes, such as cyclodextrine,10 proteins,11 polymers,12 and oil13 (as a microemulsion with water). A problem with stationary phases in chromatography is that they may be contaminated by substances present in the sample and electrolyte and by the analytes themselves. Contamination is of specific concern in the analysis of samples in complex matrixes, such as blood plasma or cell lysates. Packed particle beds require a retaining frit, often prepared by sintering of silica particles3,14,15 (3) Kennedy, R. T.; Jorgenson, W. Anal. Chem. 1989, 61, 1128-1135. (4) Jorgenson, J. W.; Lukacs, K. D. J. Chromatogr. 1981, 218, 209-216. (5) Colo´n, L. A.; Maloney, T. D.; Fermier, A. M. J. Chromatogr., A 2000, 887, 43-53. (6) Schweitz, L.; Andersson, L. I.; Nilsson, S. Anal. Chem. 1997, 69, 11791183. (7) Neiman, B.; Grushka, E.; Lev, O. Anal. Chem. 2001, 73, 5220-5227. (8) Terabe, S.; Otsuka, K.; Ichikawa, K.; Tsuchiya, A.; Ando, T. Anal. Chem. 1984, 56, 111-113. (9) Terabe, S.; Otsuka, K.; Ando, T. Anal. Chem. 1985, 57, 834-841. (10) Fanali, S. J. Chromatogr., A 2000, 875, 89-122. (11) Valtcheva, L.; Mohammad, J.; Pettersson, G.; Hjerte´n, S. J. Chromatogr. 1993, 638, 263-267. (12) Palmer, C. P. Electophoresis 2000, 21, 4054-4072. (13) Furumoto, T.; Fukomoto, T.; Sekiguchi, M.; Sugiyama, T.; Watarai, H. Electophoresis 2001, 22, 3438-3443. (14) Hilder, E. F.; Klampf, C. W.; Macka, M.; Haddad, P. R.; Myers, P. Analyst 2000, 125, 1-4. (15) Behnke, B.; Grom, E.; Bayer, E. J. Chromatogr., A 1995, 716, 207-213.

Analytical Chemistry, Vol. 74, No. 18, September 15, 2002 4595

or in situ polymerization.16,17 However, these frits have the disadvantage of introducing band-broadening effects.18-21 Fritless packed capillaries have been manufactured by trapping of the packing material in internally tapered capillaries,22,23 which may minimize band broadening. The use of additives in the buffer, such as surfactants, cyclodextrins, or proteins, suffers from problems with increased background noise that may introduce problems in trace analysis; for instance, sodium dodecyl sulfate (SDS), a commonly used surfactant in MEKC, is known to decrease the signal response in ESI-MS due to suppression of the analyte ions in the electrospray ionization and to contaminate the mass spectrometer.24 An attractive approach when buffer additives are used as interaction phases in CEC is to use the partial filling technique, which allows the analytes and buffer additives to enter the detector at different times and hence minimizes interference from the additives during detection of the analytes.11 Partial filling has been used with interaction phases such as proteins,11 cyclodextrins,25 and surfactants.26 Another approach is to apply the partial filling technique to particles in CEC. Molecularly imprinted microparticles have been used in enantiomer separations.27 However, avoiding coelution of interaction phase and analytes often requires time-consuming optimizations. Our aim was to develop a CEC/MS method that does not involve the disadvantages mentioned above; i.e., the system should work without retaining frits, the interaction phase should not be contaminated by previous analysis, and it should be tolerated by ESI-MS. In this approach, polymer nanoparticles were used as interaction phase in CEC separations of the three amines, nortriptyline, salbutamol, and diphenhydramine, without packing the particles in columns or using retaining frits. We present to our knowledge the first attempt to use particles as pseudostationary phase in CEC/MS with particles present in the entire electrolyte volume and hence a constant flow of particles into the CEC/ESI-MS interface (hereafter referred to as continuous full filling) and the use of a conventional partial filling technique. The technique, although not fully optimized for a specific separation, shows promising features for analysis of analytes in complex matrixes. EXPERIMENTAL SECTION Chemicals and Reagents. Nortriptyline, diphenhydramine, and 2,2′-azobis(isobutyronitrile) (AIBN) were obtained from Sigma (St. Louis, MO), and salbutamol was synthesized by AstraZeneca (Lund, Sweden). Ammonium carbonate and trimethylolpropane (16) Chen, J.-R.; Dulay, M. T.; Zare, R. N.; Svec, F.; Peters, E. Anal. Chem. 2000, 72, 1224-1227. (17) Haskins, W. E.; Wang, Z.; Watson, C. J.; Rostand, R. R.; Witowski, S. R.; Powell, D. H.; Kennedy, R. T. Anal. Chem. 2001, 73, 5005-5014. (18) Behnke, B.; Johansson, J.; Bayer, E.; Nilsson, S. Electrophoresis 2000, 21, 3102-3108. (19) Pyell, U. J. Chromatogr., A 2000, 892, 257-278. (20) Rebscher, H.; Pyell, U. Chromatographia 1994, 38, 737-743. (21) Rebscher, H.; Pyell, U. Chromatographia 1996, 42, 171-176. (22) Rapp, E.; Bayer, E. J. Chromatogr., A 2000, 887, 367-378. (23) Tallarek, U.; Rapp, E.; Scheenen, T.; Bayer, E.; Van As, H. Anal. Chem. 2000, 72, 2292-2301. (24) Lu, W.; Poon, G. K.; Carmichael, P. L.; Cole, R. B. Anal. Chem. 1996, 68, 668-674. (25) Rudaz, S.; Cherkaoui, S.; Gauvrit, J.-Y.; Lante´ri, P.; Veuthey, J.-L. Electrophoresis 2001, 22, 3316-3326. (26) Nelson, W. M.; Lee, C. S. Anal. Chem. 1996, 68, 3265-3269. (27) Schweitz, L.; Spe´gel, P.; Nilsson, S. Analyst 2000, 125, 1899-1901.

4596

Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

trimethacrylate (TRIM) were obtained from Aldrich (Gillingham, U.K.). Methanol was purchased from J.T. Baker (Phillipsburg, NJ), and acetonitrile, methacrylic acid (MAA) and methyl methacrylate (MMA) were purchased from Merck (Hohenbrunn, Germany). Formic acid was obtained from Riedel-de haen (Seelze, Switzerland), and water was purified by a MilliQ purification system (Millipore, Bedford, MA). Preparation of Polymer Nanoparticles. Nanoparticles were prepared by utilizing a previously described precipitation polymerization technique.27,28 The technique is based on the precipitation of polymeric spherical particles as they reach their solubility limit in the system due to their increasing molecular weight or as a result of cross-linking. A monomer containing a carboxylic acid group, MAA, 0.0545 mol L-1, a hydrophobic monomer, MMA, 0.0545 mol L-1, and a cross-linking monomer, TRIM, 0.109 mol L-1, were dissolved in acetonitrile in a screw-capped borosilicate glass tube. A radical initiator, AIBN, 0.0012 mol L-1, was added, and the prepolymerization solution was sonicated for 10 min followed by degassing by a flow of nitrogen gas for 6 min. Polymerization was initiated by UV irradiation and allowed to proceed overnight. The obtained particles were collected and washed by successive centrifugation (14 500 rpm for 15 min; Sigma 201M) and resuspended (sonication for 15 min) twice in acetonitrile/acetic acid (75:25 v/v) and once in acetonitrile. Finally, the particles were dried and stored at room temperature until use. The sizes of the nanoparticles were determined by transmission electron microscopy (TEM) on a JEOL JM 2000 FX instrument. Capillary Electrochromatography. CEC experiments were performed on a HP3DCE system (Agilent Technologies, Waldbronn, Germany), and Chem Station software was used for data processing. A 75-cm-long, 50-µm-i.d., and 375-µm-o.d. fused-silica capillary obtained from Polymicro Technologies (Phoenix, AZ) was used in all experiments. The electrolyte was a mixture of acetonitrile and an aqueous buffer (1:1 v/v). Prior to mixing with acetonitrile, the aqueous buffer, 50 mM ammonium carbonate, was adjusted to pH 8.2 with a 10% v/v ammonia/water solution. Sample solutions were prepared by dissolving nortriptyline, salbutamol, and diphenhydramine in the electrolyte to a concentration of 100, 25, and 10 µg mL-1. Figure 1 presents the structures of the sample molecules and a proposed structure of the nanoparticle polymer. Nanoparticles were suspended in the electrolyte to form stable slurries (10, 2.5, 0.44, 0.22, and 0.11 mg mL-1). Prior to use, all solutions were degassed by sonication. The capillary was rinsed prior to analysis with 0.1 mol L-1 sodium hydroxide (5 min at 1 bar), water (5 min at 1 bar), and electrolyte (10 min at 1 bar). Samples were injected into the capillary hydrodynamically (5 s at 50 mbar), and the separation voltage was 20 kV (267 V/cm). All experiments were performed with the capillary at ambient temperature. Partial Filling Experiments. Partial filling experiments were performed by introducing a plug of nanoparticle slurry into the capillary by hydrodynamic pumping at 50 mbar, after which the sample was introduced (Figure 2A). Different volumes of the slurry were evaluated at two different slurry concentrations (10 and 2.5 mg mL-1). The filling of slurry in the capillary, in percent by volume (% v/v), was determined by injecting sample (5 s at 50 mbar) into the electrolyte-filled capillary, followed by infusion of (28) Schweitz, L.; Spe´gel, P.; Nilsson, S. Electrophoresis 2001, 22, 4053-4063.

Figure 1. Structures of the ionized forms of nortriptyline, salbutamol, and diphenhydramine and a proposed structure of the polymer in the nanoparticles.

Figure 2. Schematic illustrations of a CEC-column during (A) the partial filling technique and (B) the continuous full filling technique. The capillary (black) is illustrated during three subsequent times (t0, t1, t2) for the two techniques. Samples are illustrated by dark gray bars, slurry is illustrated by light gray fields, and background electrolyte is illustrated by white fields (confined by the capillary). By the time the sample molecules reach the interface on the detector (not included in the illustration) they have passed the slurry plug in (A) while in (B) a mix of the samples and the slurry reach the interface.

the slurry at 50 mbar during continuous mass spectrometric scanning. The time measured for the sample to reach the mass spectrometer then corresponds to the time it takes to fill the capillary with slurry (100% v/v filling). A slurry injection at X% of that time was assumed to correspond to X% v/v filling of the capillary. Interaction between the sample molecules and the nanoparticles was studied by recording the changes in retention times of the sample molecules at different slurry volumes. Continuous Full Filling. Another approach was to perform experiments with a continuous flow of slurry through the CEC column. In these experiments, the sample was injected into a capillary filled with slurry and the separation was performed with slurry used as electrolyte. Figure 2B illustrates the continuous full filling technique and how it differs from the partial filling technique. Interaction between the sample molecules and the nanoparticles was studied by recording the changes in retention times of the sample molecules at different slurry concentrations. Mass Spectrometric Detection. Detection was performed on an Agilent Technologies LC/MSD ion trap SL mass spectrometer equipped with an orthogonal ESI interface operated in positive full-scan mode scanning between m/z 200 and 300 with a maximum ion accumulation time between 50 and 75 ms and a total ion current (TIC) target between 75 000 and 100 000. Sheath

Figure 3. Schematic illustration of the orthogonal electrospray interface between the CEC and MS. Positive ions are pulled out of the electrospray plume and accelerated in an electrical field toward the inlet to the mass spectrometer.

Figure 4. Transmission electron micrograph of polymer nanoparticles. The nanoparticles (black beads) have an average diameter of 160 nm (n ) 51). The gray network in the background is a carbon grid on which the nanoparticle slurry was deposited.

liquid flow consisting of methanol, water, and formic acid (1:1 v/v and 0.1% v/v) was pumped at 0.6 mL min-1 by an Agilent Technologies series 1100 quaternary pump and split 1:100 by a fixed splitter. The CE was coupled to the ESI interface using an Agilent Technologies triple tube coaxial nebulizer held at ground potential. The electrospray interface was orthogonal; i.e., the sheath liquid flow and the flow from the CE column (particles, sample, and electrolyte) was electrosprayed orthogonally to the inlet of the mass spectrometer, as illustrated in Figure 3. The orthogonal ESI interface is essential to avoid nanoparticles to enter the mass spectrometer in continuous full filling analysis. RESULTS AND DISCUSSION Preparation of Polymer Nanoparticles. The preparation of nanoparticles yielded spherical beads as seen in the TEM micrograph in Figure 4. The average diameter of the nanoparticles, as determined from the TEM micrograph, was 160 nm. The used precipitation polymerization technique allows nanoparticles to be prepared without the use of stabilizing surfactants, which is beneficial as surfactants decrease the analyte signal intensity in Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

4597

Figure 5. Eight electrochromatograms from partial filling separations of nortriptyline (peak 1), salbutamol (peak 2), and diphenhydramine (peak 3) at 10 µg mL-1. Each electrochromatogram displays the total ion chromatogram (top curve) and the reconstructed ion chromatogram for nortriptyline, monitoring m/z 264.4 ( 0.5 (bottom curve). Column A is from experiments using 10 mg mL-1 slurry concentration injected for 0, 10, 20, and 40 s (top to bottom). Column B is from experiments using 2.5 mg mL-1 slurry concentration injected for 0, 40, 80, and 160 s (top to bottom). Peak 4 is contaminants from the slurry.

ESI-MS detection. With polymerization methods that require surfactants, extensive washing would have been necessary. Partial Filling of Nanoparticles. The required interaction between the analyte molecules and the nanoparticles was confirmed in the partial filling experiments. Figure 5 shows electrochromatograms from partial filling separations of nortriptyline, salbutamol, and diphenhydramine at 10 mg mL-1 slurry concentration (A) and 2.5 mg mL-1 slurry concentration (B). The retention times for nortriptyline, salbutamol, and diphenhydramine at different slurry injection times are presented in Figure 6 (10 mg mL-1). To minimize effects from variations in the endoosmotic flow (EOF) between different runs, the retention times are normalized to the retention time of salbutamol in Figure 7. Normalization to salbutamol was chosen as Figure 6 shows that salbutamol has weak or no interactions with the nanoparticles. Normalized retention times will be used hereafter. Reconstructed 4598 Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

ion chromatograms (RIC) were used to determine which peak corresponds to which sample molecule (only RICs from nortriptyline shown here). Peaks assigned number 4 in Figure 5 are contaminants present in the slurry plug. There is a significant change in normalized retention time for nortriptyline and diphenhydramine, which indicates interaction between the nanoparticles and nortriptyline and diphenhydramine. Figure 1 shows the structure of nortriptyline, salbutamol, and diphenhydramine and a proposed structure of the nanoparticle’s surface at pH 8.2. If it is assumed that the interaction is based on ion-ion interaction between the positive charges on the sample molecules and the negative charges on the nanoparticles, the changes in retention times of the sample molecules can be explained. Salbutamol is a secondary amine with a bulky tert-butyl group. The tert-butyl group is likely to shield the protonated amine group and hence counteract interaction with the nanoparticles. Nortriptyline and

Figure 6. Retention times for nortriptyline, salbutamol, and diphenhydramine versus different slurry injection times. The concentration of nanoparticles in the slurry was 10 mg mL-1. A 1-s injection corresponds to 0.13% v/v filling of the capillary.

Figure 7. Normalized retention times for nortriptyline, salbutamol, and diphenhydramine versus different slurry injection times from partial filling experiments (10 mg mL-1 in (A) and 2.5 mg mL-1 in (B)). A 1-s injection corresponds to 0.13% v/v filling of the capillary.

diphenhydramine are secondary amines that lack the shielding tert-butyl group. This can explain their higher degree of interaction with the nanoparticles, compared to salbutamol. Initial experiments were performed using the same nanoparticles at acidic pH to investigate interactions based on hydrophobic effects (not presented). At acidic pH, the analytes are positively charged while the nanoparticles are neutral, which excludes ion-ion interactions. However, the neutral nanoparticles required high concentrations of organic solvent in the electrolyte to form stable slurries, which suppressed possible hydrophobic interactions. Filling the capillary with slurry required 780 s (13 min) of hydrodynamic pumping at 50 mbar. A one second injection then corresponds to filling 1/780 ) 0.13% v/v of the capillary. The total volume of the 75-cm-long and 50-µm-i.d. capillary is 1.47 µL, so

during 1 s of injection, 1.9 nL of slurry corresponding to 0.019 (10 mg mL-1 slurry) or 0.0047 µg (2.5 mg mL-1 slurry) of nanoparticles is introduced into the capillary. The amount of particles required for nortriptyline to coelute with salbutamol can be calculated from Figure 7. The point where the regression line for nortriptyline crosses the regression line for salbutamol gives the slurry injection time for coelution, which is 23 s for 10 mg mL-1 slurry and 102 s for 2.5 mg mL-1 slurry. Using the previously calculated slurry injection volumes gives that 0.44 (10 mg mL-1 slurry) and 0.48 µg (2.5 mg mL-1 slurry) of nanoparticles were required for coelution. The results indicate that the observed interaction appears to be affected mostly by the amount of particles injected into the CEC column and not as much by the slurry concentration; i.e., a 10-s injection of 10 mg mL-1 slurry gives retention times similar to a 40-s injection of 2.5 mg mL-1 slurry. Continuous Full Filling. In the continuous full filling experiments (Figure 2B), nanoparticles enter the capillary as a slurry (a mix of nanoparticles and electrolyte). The nanoparticles move through the capillary toward the mass spectrometer by EOF and their own electromobility and act as interaction phase for sample molecules (a pseudostationary phase). After passing through the capillary, the nanoparticles enter the orthogonal ESI interface, where the nanoparticles pass straight out of the capillary without entering the mass spectrometer while the sample molecules deviates 90° and enter the mass spectrometer. There is a continuous flow of nanoparticle slurry in to and out of the capillary throughout the continuous full filling analysis. Figure 3 is a schematic illustration of the orthogonal electrospray during continuous full filling experiments. The positively charged analytes are pulled out of the electrospray plume and accelerated toward the inlet to the mass spectrometer by electrical forces while the negatively charged and high molecular weight nanoparticles do not diverge from the electrospray plume and hence never enter the mass spectrometer. The mass spectrometer was used for continuous full filling experiments for several days without any noticeable change in response or any other effects from the nanoparticles, and the lenses of the mass spectrometer showed no deposits of particles. Electrochromatograms from continuous full filling separations of nortriptyline, salbutamol, and diphenhydramine at different slurry concentrations (0.11, 0.22, and 0.44 mg mL-1) can be seen in Figure 8A (100 µg mL-1 sample solution) and in Figure 8B (25 µg mL-1 sample solution). The retention times for nortriptyline, salbutamol, and diphenhydramine at different slurry concentrations are presented in Figure 9. At high sample concentrations (Figure 8A; 100 µg mL-1 sample solution), the peaks from nortriptyline, salbutamol, and diphenhydramine have similar signal intensities, while at low sample concentrations (Figure 8B; 25 µg mL-1 sample solution), nortriptyline and salbutamol have considerably lower signal intensities than diphenhydramine, which is not the case in the partial filling experiments (Figure 5). In Figure 10, the peak areas for nortriptyline, salbutamol, and diphenhydramine (100 µg mL-1) are plotted at different slurry concentrations. Nortriptyline’s peak area decreases dramatically (by 90%) when the slurry concentration is increased from 0.11 to 0.44 mg mL-1, but salbutamol’s area remains approximately constant. At this point, we can only speculate about the reasons for these Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

4599

Figure 8. Six electrochromatograms from continuous full filling separations of nortriptyline (peak 1), salbutamol (peak 2), and diphenhydramine (peak 3) at (A) 100 and (B) 25 µg mL-1 sample concentration. Each electrochromatogram displays the total ion chromatogram (top curve) and the reconstructed ion chromatogram for nortriptyline, monitoring m/z 264.4 ( 0.5 (bottom curve). The slurry concentrations in the experiments are 0, 0.22, and 0.44 mg mL-1 (top to bottom).

Figure 9. Normalized retention times for nortriptyline, salbutamol, and diphenhydramine versus different slurry concentrations from continuous full filling experiments. The sample concentration was 100 µg mL-1.

effects. One explanation may be that especially nortriptyline but also possibly salbutamol has greater affinity for the nanoparticles in the electrospray than diphenhydramine and hence enter the mass spectrometer to a smaller extent. Changing the sheath liquid flow so that the sample molecules desorb more easily from the particles could solve this problem, e.g., by increasing the organic content or changing the pH so that the analytes and particles no longer are oppositely charged. Another possible explanation for the lower signal intensity of nortriptyline and salbutamol 4600 Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

Figure 10. Peak areas for nortriptyline, salbutamol, and diphenhydramine at different slurry concentrations in continuous full filling experiments.

could be the contaminants, from the manufacturing of the nanoparticles, present in the slurry (peak 4 in Figure 5). Contaminants can suppress nortriptyline and salbutamol in the formation of gas-phase ions in the electrospray. This could be avoided by more thorough washing of the particles during their preparation. CONCLUSIONS It was shown that polymer nanoparticles suspended in electrolyte can be used as interaction phase in CEC using continuous

full filling and partial filling techniques. By utilizing an orthogonal electrospray interface on the mass spectrometer during continuous full filling, the nanoparticle suspension could be continuously infused into the mass spectrometer interface. At low sample concentrations, the signals from nortriptyline and salbutamol were lower than the signal from diphenhydramine compared to similar experiments performed with higher sample concentrations. More thorough washing of the nanoparticles to remove contaminants in the nanoparticle suspension in order to decrease suppression of the ionization of sample molecules in the electrospray ionization is yet to be examined. Optimizing the sheath liquid composition to improve desorption of sample molecules from the nanoparticles during the electrospray to improve the signals from sample molecules also needs to be investigated. A new way of thinking in separation technology is introduced with the continuous full filling technique with nanoparticles as interaction phase. The possibility to use different monomers in the synthesis of the nanoparticles opens up a new field of tailormade interaction phases (hydrophobic, hydrophilic, neutral, or charged) to suit different analytical problems, and there is no need for changing the column when varying the interaction phase as only the slurry has to be changed. Another speculative approach is to use the continuous full filling technique for mimicking column switching by introducing several different nanoparticle slurries (29) Tan, H.; Yeung, E. S. Anal. Chem. 1998, 70, 4044-4053. (30) Kang, S. H.; Gong, X.; Yeung, E. S. Anal. Chem. 2000, 72, 3014-3021.

(A-C) into the capillary or on a chip. Slurry A may be chosen to interact with the sample matrix but not with the sample molecules and hence function as a precolumn, while the following slurries B and C may be chosen to interact with the sample molecules in different ways, e.g., by ion-ion interaction and hydrophobic interaction. Multiplex CEC29,30 is another attractive application that may be useful for parallel continuous full filling separations using different nanoparticle slurries. The continuous full filling technique is easily optimized, does not require time-consuming particle packing or retaining frits, and has the advantage of an unused interaction phase for every new analysis. We believe that the continuous full filling technique will be of great value in future analysis of analytes in complex matrixes and in other analysis where contamination of the stationary phase should be avoided. ACKNOWLEDGMENT The authors thank Professor Karl-Gustav Wahlund at Lund University for valuable discussions on ion-ion interaction chromatography and TFR, NFR, Crafoordska stiftelsen, and Fysiografiska sa¨llskapet i Lund for financial support.

Received for review June 20, 2002. Accepted June 27, 2002. AC0204045

Analytical Chemistry, Vol. 74, No. 18, September 15, 2002

4601