Reversed-Phase Electrochromatography with a Monolithic

Reversed-Phase Electrochromatography with a Monolithic Microcolumn Prepared in a 2.2-mm-Inner Diameter Fused-Silica Tube. Ning Deng, You-zhao He*, Lei...
0 downloads 0 Views 222KB Size
Anal. Chem. 2005, 77, 5622-5627

Reversed-Phase Electrochromatography with a Monolithic Microcolumn Prepared in a 2.2-mm-Inner Diameter Fused-Silica Tube Ning Deng, You-zhao He,* Lei Wang, Xiao-kui Wang, and Qing-de Su

Department of Chemistry, University of Science and Technology of China,Hefei, Anhui, 230026, P. R. China

Capillary electrochromatography possesses advantages of high separation efficiency and velocity, but it also has its disadvantages due to its small inner diameter, such as poor detection sensitivity, low sample capacity, and some trouble in its column preparation. To overcome these shortcomings, a monolithic microcolumn with a surface area larger than 200 m2/g was prepared by sol-gel polycondensation of tetraethoxysilane-hydrochloric acidpoly(ethylene glycol) and filling with fine quartz sand in a 2.2-mm-i.d. fused-silica tube. The prepared microcolumn was used in the separation of aromatic compounds by reversed-phase electrochromatography. Some factors that affected electroosmotic flow were explored, such as electric field strength, buffer concentration, and buffer pH. Acetonitrile concentration in the mobile phase was investigated for phenol, benzene, and naphthalene separation. The separation results were satisfying with the electrochromatographic microcolumns. The detection limits of phenol, benzene, and naphthalene were 0.07, 0.26, and 0.04 mg/L, respectively. Capillary electrochromatography (CEC) is a novel electrokinetic separation technique that combines the high separation efficiency of capillary electrophoresis with the fine analyte adaptability of high-performance liquid chromatography (HPLC). Pretorius et al. were the influential pioneers of CEC, who demonstrated the advantages of electroosmosis as a pumping mechanism for chromatographic separation.1 Jorgenson and Lukacs published the CEC separation of 9-methylanthracene and perylene with an ODS-packed capillary column.2 Tsuda demonstrated the possibility of CEC separation driven by electroosmosis and pressure flow simultaneously.3 So the flat pluglike electroosmotic flow (EOF) or EOF combined with pressurized flow has been adopted to drive the mobile phase in CEC. Moreover, Knox and Grant made another significant contribution of heat effect theory for the development of this technique.4 Packed5,6 and open7 capillary column are two major types used in current CEC practice. In CEC, most of the separations are * To whom correspondence should be addressed. E-mail: yzhe@ustc.edu.cn. Phone: (86)-551-3607072. Fax: (86)-551-3603388. (1) Pretorius, V.; Hopkins, B. J.; Schieke, J. D. J. Chromatogr. 1974, 99, 2330. (2) Jorgenson, J. W.; Lukacs, K. D. J. Chromatogr. 1981, 218, 209-216. (3) Tsuda, T. Anal. Chem. 1987, 59, 521-523. (4) Knox, J. H.; Grant, I. H. Chromatorgraphia 1987, 24, 135-143.

5622 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

performed by using capillary columns packed with HPLC stationary phases. However, there is some difficulty in the preparation of packed capillaries. The preparation qualities of stationary phase and frit are of key importance to achieve satisfying separation capability of CEC. A uniform stationary phase of packed capillaries is required to obtain high separation efficiency and fine reproducibility with many small particles. Moreover, the greatest challenge is the preparation of highly permeable and mechanically strong end frits. The nonuniformity of the packing column bed and the active sites in end frits can lead to the formation of bubbles in capillary columns, which cause baseline noise or even current breakdown. To avoid the trouble associated with the packed capillaries, open capillary columns have been used in CEC. The stationary phases are chemically bonded onto the inner wall of open capillary columns directly, so the capillary columns have a very low phase ratio that leads to low retention time and low sample capacity. The monolithic column8,9 holds great potential for CEC separation because it can effectively overcome both the preparation difficulties associated with the conventionally packed capillary columns and the disadvantages resulting from the open capillary columns. However, this type of CEC column also has shortcomings from its small diameter, including limited sample loadability and restricted concentration detectability. The inner diameter of the capillary columns is usually less than 100 µm in order to eliminate Joule heating in electroosmosis-driven conditions, because the temperature difference between the center and inner wall of a capillary column is proportional to the square of column diameter.4 Yan et al. showed that 320-µm-i.d. packed columns could be used in CEC separation, but a slightly positive deviation from the linearity of Ω plots was observed.10 Zare et al. also reported that a 550-µm-i.d. pack column was used in CEC separation and the separation efficiency of CEC achieved 104 plates/m.11 To reduce the effect of Joule heating in large-diameter monolithic columns, an additional approach should be proposed (5) Yan, C.; Dadoo, R.; Zhao, H.; Zare, R. N.; Rakestraw, D. J. Anal. Chem. 1995, 67, 2026-2029. (6) Dittmann, M. M.; Rozing, G. P. J. Chromatogr., A 1996, 744, 63-74. (7) Tan, Z. X. J.; Remcho, V. T. Anal. Chem. 1997, 69, 581-586. (8) Tang, Q. L.; Xin, B. M.; Lee, M. L. J. Chromatogr., A 1999, 837, 35-50. (9) Ishizuka, N.; Minakuchi, H.; Nakanishi, K.; Soga, N.; Nagayama, H.; Hosoya, K.; Tanaka, N. Anal. Chem. 2000, 72, 1275-1280. (10) Yan, C.; Schaufelberger, D.; Erni, F. J. Chromatogr., A 1994, 670, 15-23. (11) Chen, J.-R.; Zare, R. N.; Peters, E. C.; Svec, F.; Frecht, J. J. Anal. Chem. 2001, 73, 1987-1992. 10.1021/ac050589q CCC: $30.25

© 2005 American Chemical Society Published on Web 08/04/2005

to prepare large-diameter electrochromatography (EC) columns that can offer high qualities of separation efficiency, reproducibility, loadability, and detectability. In our previous work,12 a 2.7mm-i.d. monolithic microcolumn bed was prepared by sol-gel polycondensation of potassium silicate-potassium hydroxideformamide and filling with fine quartz sand. However, the surface area of this microcolumn was too low to separate the analytes on baseline. In this paper, a new microcolumn fabrication method by sol-gel polycondensation of tetraethoxysilane-hydrochloric acid-poly(ethylene glycol) and filling with fine quartz sand was developed and its surface area could be higher than 200 m2/g. Its electrokinetic properties were investigated. Phenol, benzene, and naphthalene were separated with the microcolumn bonded with C8 by reversed-phase EC and detected at 254 nm by an online spectrophotometer. The detection limit of phenol, benzene, and naphthalene was 0.07, 0.26, and 0.04 mg/L, respectively. Kuban et al. showed that the detection limit of phenol was 1.86 mg/L with a 70 cm × 50 µm i.d. capillary and an on-column spectrophotometer detecting at 210 nm by capillary zone electrophoresis.13 EXPERIMENTAL SECTION Chemicals and Materials. Fused-silica tubes were purchased from Glass Instruments Co. (Shanghai, China). Tridistilled water obtained from a distilled water system (SZ-3 Huxi Analytical Instruments Factory, Shanghai, China) was used to prepare the solutions in this work. Acetonitrile was of chromatographic grade and purchased from Tedia Co. (Fairfield, OH). Octyltriethoxysilane (C8-TEOS) was of analytical grade and purchased from the Aldrich Chemical Co. (Milwaukee, WI), Other materials and reagents of 100-125-µm fine quartz sand, tetraethoxysilane (TEOS), poly(ethylene glycol) (PEG; Mw 10 000), tris(hydroxymethyl)aminomethane (Tris), hydrochloric acid, sodium hydroxide, ethanol, methanol, acetone, thiourea, phenol, benzene, and naphthalene were of analytical grade and purchased from the Chemical Reagent Co. (Shanghai, China). A stock buffer solution was prepared by adjusting 50 mmol/L Tris solution with 5 mol/L HCl to pH 8.5. The mobile phase was 45% (v/v) acetonitrile/water solution containing 6 mmol/L Tris buffer solution (pH 8.5) prepared by mixing acetonitrile and water with an appropriate amount of the stock buffer solution. The mobile phase was degassed for 15 min by an S 2200 ultrasonic cleaner (120 W, 35 kHz, J & L Ltd., Shanghai, China) before use. Three analyte solutions of 1 mmol/L phenol, 2 mmol/L benzene, and 0.2 mmol/L naphthalene were prepared by dissolving the analyte reagents with the mobile-phase solution. The 1 mmol/L thiourea dissolved in the mobile-phase solution was used as EOF marker. Instrumentation. A microcolumn EC system consisted of a DYY-12C electrophoretic power supply (20-5000 V, Liuyi Instrument Factory, Beijing, China), a UV-9100 spectrophotometer (Ruili Anal. Instrument Ltd., Beijing, China) measuring at 254 nm, and an IFIS-C flow injection analysis system comprising two peristaltic pumps and one eight-way injection valve (Ruimai Electr. Tech. Ltd., Xi’an, China). A PEEK influent unit, a PEEK flow cell, and a (12) Qu, Q. S.; He, Y. Z.; Gan, W. E.; Deng, N.; Lin, X. Q. J. Chromatogr., A 2003 983, 255-262. (13) Kuban, P.; Berg, M.; Garcia, C.; Karlberg, B. J. Chromatogr., A 2001, 912, 163-170.

Figure 1. Schematic diagram of microcolumn electrochromatography system: M, mobile phase; S, sample solution; FI, flow injection system; I, influent unit; MC, EC microcolumn; FC, flow cell; LP, light path; V, valve; W and W1, waste; E, power supply; A, amperemeter; D, detector; CP, computer.

reversed-phase EC microcolumn were homemade. A schematic diagram of the microcolumn EC system is shown in Figure 1. As shown in Figure 1, the main conduit of the influent unit (I) looks like T shape, and a vertical electrode cell locates at the end of the conduit. The outlet of the electrode cell is higher than the inlet, of which the position is propitious to expel the gas formed on the anode in EC separation. A platinum electrode is fixed on the top of the electrode cell. The size of inlet conduit, electrode cell, and outlet path is 12 mm × 1 mm i.d., 7 mm × 3 mm i.d., and 7 mm × 3 mm i.d., respectively. To minimize the dispersion of analyte zone, both the branch conduits connected to the microcolumn in the influent unit and flow cell are 1 mm × 1 mm i.d. The microcolumn, the flow conduits and the electrode are sealed and connected by silicone rings and screw fittings. A X-650 scanning electron microscope (Hitachi) and An ASAP 2000 accelerated surface area and porosimeter (Micromeritics) was employed to measure the scanning electron micrographs and surface area of the microcolumns, respectively. Preparation of Silica Monolithic Microcolumn. To compare the heating effect of two types of EC microcolumns, both monolithic microcolumns without and filled with fine quartz sand were prepared in this work. Four milliliters TEOS was added into the solution of 0.30 g of PEG dissolved with 3.0 mL of 0.9 mol/L HCl. The mixture solution was stirred for 20 min and degassed by the ultrasonic cleaner for 10 min. Then the solution was introduced into a 3.2-mm-i.d. PTFE mold sealed with plastic film at its bottom. The mold filled with the mixture solution was sealed again on its top and placed vertically in a drying oven at 60 °C overnight. After polycondensation, the inner diameter of the prepared silica monolithic microcolumn would shrink to ∼2.2-mm i.d.. The prepared monolithic microcolumn was washed with water and ethanol in order and then dried at 58 °C in the drying oven. Preparation of Silica Monolithic Microcolumn Filled with Fine Quartz Sand.14 A 2.2-mm-i.d. fused-silica tube was sealed at its bottom with plastic film and filled with the mixture solution prepared by the method mentioned above, and 100-125-µm fine (14) He, Y. Z.; Wang, X. K.; Deng, N.; Wang, L.; Han, F. Chinese Patent Applied No. 200510038677.5, 2005.

Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

5623

Figure 2. Scanning electron micrographs of silica microcolumns. The micrograph of the microcolumn without fine quartz sand is observed under a magnification of 2000 in Figure 2A, and those of the columns filled with fine quartz sand are obtained under a magnification of 100 in (B) and 5000 in (C).

quartz sand was added into the tube slowly to a required height by vibration. Then the fused-silica tube was sealed again on its top with plastic film and placed vertically in the drying oven at 60 °C overnight. The prepared microcolumn was flushed with water and ethanol in order and dried at 58 °C in the drying oven. EC Microcolumn Bonded with C8 Stationary Phase. After the solvent was evaporated, the microcolumn was bonded by C8 stationary phase, referring to the method described by Guo and Colo´n.15 The reaction solution was prepared by mixing 0.67 mL of TEOS, 0.93 mL of C8-TEOS, 1.0 mL of ethanol, 0.38 mL of water, and 35 µL of 0.1 mol/L HCl, stirring at room temperature overnight, introducing it into the microcolumn, keeping inside the column for 10 min, and removing the reacted solution by pressure. Then the microcolumn was dried at 100 °C for 10 h in the drying oven. Last, the column was flushed with acetone and methanol and then equilibrated with the mobile phase. Separation Procedure of Microcolumn EC. During analyte loading, the injection valve was turned to its loading position and the analyte solution was introduced into a sample loop (8 cm × 0.5 mm i.d.) of the valve by one peristaltic pump. Then the valve was turned to the injection position and the electrophoretic power supply (E) switched on at 1550 V. Another peristaltic pump propelled the mobile phase and the analyte zone toward the EC microcolumn at a velocity of 0.53 mm/s detected with a 1.0-mL pipet. When the analyte zone passed into the influent unit (I), the analyte solution was electrokinetically injected into the microcolumn (C) by EOF velocity of 0.12 mm/s. The introduced volume of the analyte solution can be determined by the length of the sample loop, injection voltage, and pump flow rate. The programmed time to turn off the waste valve and reduce the pumping velocity was 30 s after the eight-way injection valve turned to the injection position, and the electrochromatogram was recorded. A multiple flow of 0.25 mm/s was adopted to carry out the electrochromatographic separation by EOF and the peristaltic pump. Three analytes were separated by the microcolumn and passed through the flow cell (FC). The absorbance was measured by the spectrophotometer (D) at 254 nm and transferred to a computer (CP). An amperemeter (A) was used to measure the electric current during the EC separation. Last, the waste valve (V) was turned on, and the mobile phase was introduced into the (15) Guo, Y.; Colo´n, L. A. Anal. Chem. 1995, 67, 2511-2516.

5624

Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

influent unit again and expelled the gas formed on the anode in the EC separation. In addition, EOF velocity was measured independently by EC separation of thiourea without the pressurized flow. In this situation, the waste valve (V) was kept open during the detection of EOF velocity. RESULTS AND DISCUSSION Characteristic of Microcolumn. Nakanishi and Soga prepared the silica monolithic columns by hydrolysis and polycondensation of tetramethoxysilane.16,17 In this paper, TEOS was hydrolyzed followed by polycondensation under acidic condition in a PTFE mold or a fused-silica tube to form a silica monolithic microcolumn with satisfactory mechanical strength and high permeability. The structural morphology of the silica monolithic microcolumn was observed on its cross section by the scanning electron microscopy (SEM). Figure 2A shows that the silica morphology is an entire cross network formed by smooth cylinder silica. The surface area of the silica monolithic microcolumn was 200-360 m2/g measured by a nitrogen adsorption and desorption method performed by the accelerated surface area and porosimeter. The porosity of the silica monolithic microcolumn was ∼60% as evaluated by Rathore and Horva´th’s method,18 which implies that this microcolumn possesses high permeability. If a silica microcolumn were prepared in fused-silica tube directly, the gap between the silica microcolumn and the inner wall of fused-silica tube would form because of column shrinkage during polycondensation. However, The shrinkage associated with the polymerization can be solved practically by packing fine quartz sand into the sol-gel solution in the fused-silica tube. Panels B and C in Figure 2 show the SEM photographs of the microcolumn prepared by sol-gel polycondensation and filling with fine quartz sand in the fused-silica tube. It can be seen in Figure 2B that the shrinkage gap between the microcolumn and the fused-silica tube has been avoided and the fine quartz sand has solidified tightly by the silica polymer. The porosity of the microcolumn decreased to 51% by packing fine quartz sand into the fused-silica tube. Curves 1 and 2 in Figure 3 display the effect of electric field strength on electric current in the silica microcolumn filled without (16) Nakanishi, K.; Soga, N. J. Non-Cryst. Solids 1992, 139, 1-13, 14-24. (17) Nakanishi, K.; Soga, N. J. Non-Cryst. Solids 1992, 142, 36-44, 45-54. (18) Rarthore, A. S.; Horva´th, Cs. Anal. Chem. 1998, 70, 3069-3077.

Figure 3. Effect of electric field strength on electric current in monolithic microcolumns. Curves 1 and 2 display the effect in a normal microcolumn and quartz sand microcolumn, respectively. The mobile phase is 45% (v/v) acetonitrile mobile phase containing pH 8.5, 6 mmol Tris buffer solution. Both sizes of the normal microcolumn and quartz sand microcolumn are 10 cm × 2.2 mm i.d.

and with fine quartz sand, respectively. All the reported data are expressed with the average values of three individual detections throughout this work except for an additional declaration. Hereinafter, the microcolumn filled without and with fine quartz sand can be named a normal microcolumn and quartz sand microcolumn, respectively. The electric current in the normal microcolumn is much higher than that of the quartz sand microcolumn under the same electric field strength, and the electric current of the former increases faster than that in the latter. The slope of curve 1 increases obviously with an electric field strength higher than 120 V/cm, which implies that the excessive Joule heating occurs in the normal microcolumn. On the contrary, the slop of curve 2 keeps constant with the electric field strength enhanced to 300 V/cm, which indicates that the Joule heating has been limited by packing fine quartz sand. Influence Factors of EOF in Quartz Sand Microcolumn. EOF is generated by solvated pair ions in the diffusion layer of the electrical double layer under an external electric field, and its velocity can be expressed with eq 1, where veo is EOF velocity,

veo )

LM orζ E ) teo η

(1)

LM is column length from its inlet to detection window, teo is migration time of EOF marker, o is vacuum permittivity, r is relative dielectric constant of mobile phase, ζ is zeta potential, η is viscosity of mobile phase, and E is external electric field strength. Equation 1 indicates that the EOF velocity depends on the factors of applied electric field strength, zeta potential, and mobile-phase properties. Effect of Electric Field Strength on EOF. Figure 4 shows the effect of applied electric field strength on EOF velocity and electric current. It can be seen that the EOF and electric current increase approximately linearly with the increase of applied electric field strength. If the excessive Joule heating occurs, the plots of EOF and electric current versus applied electric field strength will

Figure 4. Effect of electric field strength on electric current and electroosmotic flow velocity in a quartz sand microcolumn. The size of the quartz sand microcolumn is 10 cm × 2.2 mm i.d. A 1 mmol/L solution of thiourea is used as EOF marker. The detection wavelength is 254 nm. Other conditions are the same as in Figure 3.

show positive deviation from linearity since the viscosity and molar conductivity of the buffer solution have been affected by temperature. Figure 4 indicates that Joule heating can be ignored in the quartz sand microcolumn from 60 to 300 V/cm, for the EOF velocity and current increase linearly with the increase of applied electric field strength. Effect of Buffer Concentration. The zeta potential ζ on the electrical double layer can be expressed as eq 2, where δ is the

ζ ) δσ/or

(2)

thickness of the electrical double layer and σ is the total excess charges per unit area of the Stern plane. The thickness of the electrical double layer is a function of ionic strength and is expressed with eq 3, where R is the gas

δ ) (orRT/2F2C)1/2

(3)

constant (8.314 J/K), T is absolute temperature, F is the Faraday constant (96 500 C/mol), and C is molar concentration of electrolyte. Based on eqs 1-3, the thickness of electrical double layer increases with the decrease of buffer concentration. It leads to the increase of zeta potential and the enhancement of EOF velocity. Figure 5 shows a plot of EOF velocity versus the inverse of the square root of buffer concentration. The buffer concentration was adjusted by mixing an appropriate amount of 50 mmol/L buffer stock solution with acetonitrile water solution. As illustrated in the figure, the EOF velocity increases with the decrease of Tris concentration when buffer concentration is lower than 7.5 mmol/L (horizontal axis value >0.365). However, when the buffer concentration is 10 mmol/L (horizontal axis value 0.316), the EOF velocity increases due to the decrease of mobile-phase viscosity caused by excessive Joule heating. The linear relationship between the EOF velocity and the inverse of the square root of the buffer concentration can be observed in a buffer concentration range lower than ∼7.5 mmol/L. Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

5625

Table 1. Resolution of Phenol and Benzene with Different Acetonitrile Concentrations in the Mobile Phasea acetonitrile concentration (%, v/v)

resolution

65

60

55

50

45

40

0.56

0.81

1.43

1.64

2.04

2.37

a Mobile phase contains pH 8.5, 6 mmol/L Tris buffer solution, and other conditions are the same as in Figure 4.

Figure 5. Effect of buffer concentration on EOF velocity in a quartz sand microcolumn. The mobile phase is 45% acetonitrile aqueous mobile phase containing pH 8.5 Tris buffer solution. The electrical field strength is 155 V/cm. Other conditions are the same as in Figure 4.

Figure 7. Electrochromatograms of three aromatic compounds: phenol (1), benzene (2), and naphthalene (3). The electrical field strength is 155 V/cm, and other conditions are the same as in Figure 4.

Figure 6. Effect of buffer pH on EOF velocity in a quartz sand microcolumn. The mobile phase is 45% acetonitrile aqueous mobile phase containing 6 mmol Tris buffer solution. The electrical field strength is 155 V/cm. Other conditions are the same as in Figure 4.

Effect of Buffer pH on EOF. The effect of buffer pH on EOF velocity was investigated with 45% (v/v) acetonitrile/water solution containing 6 mmol/L Tris buffer solution at different pH value. As shown in Figure 6, the EOF velocity increases with the increase of buffer pH. The ionization of a silanol group on the surface of fine quartz sand and fused-silica tube is enhanced with the increase of buffer pH that results in the increase of zeta potential and EOF velocity. Reversed-Phase EC Separation with Quartz Sand Microcolumn. The concentration of organic solvent in the mobile phase is the most important parameter that affects the resolution of analyte separation, so the acetonitrile concentration in the mobile phase for reversed-phase EC separation of phenol, benzene, and naphthalene was investigated. The acetonitrile concentration was adjusted by mixing appropriate volumes of acetonitrile and water, with the buffer concentration kept constant. Table 1 shows the resolution between phenol and benzene under the mobile phase 5626 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

containing pH 8.5, 6 mmol/L Tris buffer solution with different acetonitrile concentrations. As shown in the table, the resolution of phenol and benzene increases with the decrease of acetonitrile concentration. Phenol and benzene can be separated on baseline when the concentration of acetonitrile is lower than 55%, but the optimal separation efficiency is achieved when concentration of acetonitrile is 45%. Figure 7 shows the EC separation of phenol, benzene, and naphthalene under the mobile phase of 45% (v/v) acetonitrile aqueous mobile phase containing pH 8.5, 6 mmol Tris buffer solution. It can be seen that three compounds were separated on baseline and the peaks of these compounds were symmetrical. The separation efficiency obtained for phenol was 3.0 × 104 plates/m. The detection limits (S/N ) 3) of phenol, benzene, and naphthalene were 0.07, 0.26, and 0.04 mg/L, respectively. Separation Reproducibility with Quartz Sand Microcolumn. Run-to-run reproducibilities of retention time, viz. relative standard deviation (RSD), for phenol and benzene were 4.2 and 4.3%, respectively, which were obtained by six individual runs with the same microcolumn and the same separation conditions. The retention times of phenol and benzene were 14.1 ( 0.6 and 16.2 ( 0.7 min, respectively. Column-to-column reproducibilities (RSD) of retention time for phenol and benzene were 4.2 and 4.8%, respectively, which were measured with four pieces of microcolumns under the same separation conditions. The retention times of phenol and benzene were 14.2 ( 0.6 and 16.4 ( 0.8 min, respectively. It indicates that the reproducibility and reliability of

the microcolumns are satisfied. There were no significant changes for the aromatic compound separation in two months with the same microcolumn. In addition, a commercial detector can be employed in the proposed separation technique, such as a UVvisible spectrophotometer. CONCLUSIONS A silica monolithic microcolumn was prepared by sol-gel polycondensation of tetraethoxysilane-hydrochloric acid-poly(ethylene glycol) and filling with fine quartz sand in a 2.2-mm-i.d. fused-silica tube and was used in reversed-phase EC separation successfully. Some parameters of the microcolumn, such as electric current, EOF velocity, and reproducibility, were investigated by using a laboratory-made electrochromatography system. The proposed EC separation with the quartz sand microcolumn was proved to be an effective separation technique that can overcome the shortcomings of CEC. The microcolumn separation can be detected with a commercial spectrophotometer detector. The laboratory-made electrochromatography system can be operated with electroosmotic, pressurized, and electroosmotic/pres-

surized flow. The CE separation technique can be a complementary method for CEC to improve the sample loading capacity and concentration detectability and can be developed to a fully automated EC system. However, low flow rate of mobile phase can lead to the restricted separation efficiency because of the limited working pressure of the peristaltic pumps. Large detection volume in the flow cell for matching the light path of the spectrophotometer can cause the broadening of analyte peaks and the reduction of separation efficiency. The surface area and distribution of macro- and mesopore have to be measured. These problems should be improved in due course. ACKNOWLEDGMENT This work is supported by the Natural Science Foundation of China, 20275035 and 29975026.

Received for review April 7, 2005. Accepted July 3, 2005. AC050589Q

Analytical Chemistry, Vol. 77, No. 17, September 1, 2005

5627