Anal. Chem. 1999, 71, 5444-5446
Hadamard Transform Capillary Electrophoresis Takashi Kaneta, Yasuko Yamaguchi, and Totaro Imasaka*
Department of Chemical Systems and Engineering, Graduate School of Engineering, Kyushu University, Hakozaki Fukuoka, 812-8581 Japan
This paper reports the first demonstration of a multiplex sample injection technique in capillary electrophoresis. The sample was injected into a capillary (effective length, 4 cm) as a pseudorandam Hadamard sequence by a photodegradation technique using a high-power gating laser, and the fluorescence signal, which was measured using a probe excitation beam, was decoded by an inverse Hadamard transformation. The signal-to-noise ratio was improved by a factor of 8, which was in good agreement with the theoretically predicted value of 8.02. This approach is potentially useful for the enhancement of the sensitivity by 3 orders of magnitude in high-resolution capillary electrophoresis, combined with fluorescence detection. Capillary electrophoresis has attracted considerable attention because of its widespread applicability in fields such as in DNA sequencing and protein analysis. The major advantage of capillary electrophoresis lies in its capability for fast and high-resolution separations. However, the injection volume is limited to 1 nL, which constitutes a limiting factor in terms of detection limits. In practice, it is difficult to prepare and manipulate a 1-nL sample, and the injection of sample sizes greater than 1 µL would be highly advantageous. However, direct injection of a 1-µL sample severely degrades the separation resolution. For this reason, the analyte must be concentrated into a small volume before sample separation,1,2 requiring additional procedures and, hence, time for analysis. An alternative approach to this problem might be the use of a multiplexing technique. Fourier and Hadamard transform techniques are currently employed in nuclear magnetic resonance spectroscopy,3 infrared absorption spectroscopy,4 and ion cyclotron resonance mass spectroscopy. Recently, this technique has also been applied to time-of-flight mass spectroscopy.5 The advantage of the multiplexing technique is a high throughput of the signal, thus providing superior signal-to-noise ratios. This substantially improves sensitivity and decreases the detection limit. This type of approach has already been applied to separation techniques, such as gas chromatography, for which a cross-correlation technique6,7 has been employed. The advantage of the Hadamard transform technique in gas chromatography has also been suggested but only by means of computer simulation.8 To date, studies of (1) (2) (3) (4) (5) (6) (7)
Quirino, J. P.; Terabe, S. Science 1998, 282, 465-468. Chien, R.-L.; Burgi, D. S. Anal. Chem. 1992, 64, 489A-496A. Blu ¨ mich, B.; Ziessow, D. J. Magn. Reson. 1982, 46, 385-405. Sugimoto, N. Appl. Opt. 1986, 25, 863-865. Brock, A.; Rodriguez, N.; Zare, R. N. Anal. Chem. 1998, 70, 3735-3741. Annino, R.; Gonnord, M.-F.; Guichon, G. Anal. Chem. 1979, 51, 379-382. Phillips, J. B. Anal. Chem. 1980, 52, 468A-478A.
5444 Analytical Chemistry, Vol. 71, No. 23, December 1, 1999
Figure 1. (A) Schematic illustration of Hadamard transform capillary electrophoresis; (B) geometry of capillary.
the Hadamard transform technique have not been reported relative to applications to liquid-phase separation techniques. In this work, we report the first use of the Hadamard transform technique for application to capillary electrophoresis. We conclude that this approach would be of value in terms of improving sensitivity by signal multiplexing with minor modifications of the analytical system. EXPERIMENTAL SECTION Figure 1A illustrates the homemade experimental setup for Hadamard transform capillary electrophoresis. A high-voltage power supply, model HCZE-30PN0.25 (Matsusada Precision Devices, Shiga, Japan), is used for applying the voltage. A capillary (25 µm i.d., 375 µm o.d., total length 40 cm, GL Sciences Inc., (8) Kaljurand, M.; Ku ¨ llik, E. Chromatographia 1978, 11, 328-330. 10.1021/ac990625j CCC: $18.00
© 1999 American Chemical Society Published on Web 10/22/1999
Tokyo, Japan) is hold perpendicularly to the laser beam axis by a homemade holder which is constructed on a three-dimensional (XYZ) translation stage (Σ707-60PC, Sigma Koki, Saitama, Japan), as shown in Figure 1B. An argon ion laser emitting at 514.5 nm (GLG3200, Nippon Electric Co., Tokyo, Japan) is split into two parts for use as gating and probe beams, respectively. The gating beam is focused by an objective lens (Nikon, Tokyo, Japan, magnification ×5) onto the capillary. The beam is passed through or blocked by an optical shutter (F77-4, Suruga Seiki, Shizuoka, Japan) which is modulated by a controller (F77-6, Suruga Seiki), interfaced with a personal computer (PC9801 RX, Nippon Electric Co.). When the shutter is opened, the light-absorbing analyte is photodegradated by strong irradiation by a visible light. Thus, the sample is injected only when the shutter is closed. This technique has already been reported as a fast sample injection technique in capillary electrophoresis.9 The modulation was performed based on an S-matrix code which can be found in textbooks.10,11 Thus, a series of data obtained by pseudorandom sequence injection is given by
[η] ) [S] × [E]
Figure 2. Electropherogram for sodium fluorescein. Concentration of sodium fluorescein, 10 nM; buffer, 10 mM KHCO3 + NaOH (pH, 9.3); voltage, 15 kV; gating beam, 120 mW; probe beam, 9 mW.
(1)
where S is the n × n matrix consisting of “zero” or “one” element and E is a series of data representing an electropherogram. The probe beam is focused by a lens onto the capillary, positioned 4 cm from the focal point of the gating beam. Fluorescence is collected by an objective lens (Olympus, Tokyo, Japan, magnification ×50) and passed through notch (514.5 nm; bandwidth,
