Water was purified with an Elix 3 UV system (Millipore Co., Billerica, MA). All reagents used were commercially available and of analytical grade. 1-Naphthol, 2,6-naphthalenedisulfonic acid, acetonitrile, and ethyl acetate were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Fused-silica capillary tubes with an inner diameter of 50 μm were purchased from Yasaka Industries, Inc. (Tokyo, Japan).

A schematic diagram of the present TRDC system comprised of a fused-silica capillary tube (120 cm total length and 100 cm effective length), micro-syringe pump (MF- 9090; Bioanalytical Systems, Inc., West Lafayette, IN), and absorption detector (modified SPD-10AV spectrophotometric detector; Shimadzu Co., Kyoto, Japan) is shown in Figure 1. The tube temperature was controlled by dipping the capillary tube (ca. 60 cm) in water maintained at a definite temperature (0˚C or 20˚C) in a beaker with stirring. The ternary mixed solvents of water-hydrophilic/hydrophobic organic solvent mixture (water-acetonitrile-ethyl acetate) solutions with various volume ratios were used as carrier solutions. Analyte solutions containing 1-naphthol and 2,6-naphthalenedisulfonic acid (1 mM each) as models were prepared with the carrier solutions.

The analyte solution was introduced directly into the capillary inlet side by the gravity method (20 cm height for 30 s). After analyte injection, the capillary inlet was connected through a joint to a micro-syringe. The syringe was placed on the micro-syringe pump. The carrier solution was fed into the capillary tube at 0.2 μL∙min^–1 flow rate under laminar flow conditions. On-capillary absorption detection (254 nm) was performed with the detector.

The phase diagram for the ternary mixed solvents of water-acetonitrile (hydrophilic organic solvent)-ethyl acetate (hydrophobic organic solvent) in a vessel at a temperature of 20˚C was constructed. The obtained phase diagram is shown in Figure 2. The dotted curve in the

diagram indicates the boundary between homogeneous and heterogeneous phases. The phase diagram showed that each component ratio of the solvents formed a homogeneous (one homogeneous phase) or heterogeneous (two homogeneous phases) solution. The solutions with component ratios of a-t were used as carrier solutions in the TRDC system in the following sections. The component ratios of the solvents or the carrier solutions a-t are plotted in the diagram based on separation performance of analytes described below, where circles (○), triangles (Δ), and crosses (×) indicated baseline-separation, split-separation (not baseline-separation), and non-separation for the analytes on the chromatograms, respectively.

To a solution with a constant water-ethyl acetate volume ratio of 5:7 was added acetonitrile in a vessel at a temperature of 20˚C to prepare carrier solutions (organic solvent-rich carrier solutions) with the following compositions: a) water-acetonitrile-ethyl acetate 5:13:7 volume ratio; b) 5:15:7 volume ratio; c) 5:17:7 volume ratio; d) 5:19:7 volume ratio; and e) 5:20:7 volume ratio. The five carrier solutions a)-e) were all homogeneous solutions. The analyte mixtures of 1-naphthol and 2,6-naphthalenedisulfonic acid were examined with the present TRDC system using carrier solutions a)-e).

The obtained chromatograms are shown in Figure 3. Carrier solutions a)-c) could separate the analytes in the mixture. 1-Naphthol (hydrophobic) was eluted first with near the average linear velocity in the capillary tube under laminar flow conditions and 2,6-naphthalenedisulfonic acid (hydrophilic) was eluted second with lower velocity than the average linear velocity. The elution order was reasonable in the TRDC system with the organic solvent-rich carrier solution as mentioned in the Introduction. The other homogeneous carrier solutions d) and e) of water-acetonitrile-ethyl acetate showed split-separation and non-separation of the analytes, respectively, under the present analytical conditions. The separation performances obtained with carrier solutions a)-e) are plotted with the following symbols: circles (○), triangles (Δ), and crosses (×) in the phase diagram in Figure 2. The separation performance decreased from carrier solution a) to e) as shown in Figure 3. However, even carrier solutions b) and c) in which the compositions were a little far from the boundary in the diagram also showed baseline-separation for the analytes. TRDP must occur with such allowance for the solvents compositions around the boundary in this area of the diagram.

The chromatograms of the analytes, the mixture of 1- naphthol and 2,6-naphthalenedisulfonic acid, were examined with the homogeneous carrier solutions f)-t). The compositions for the solutions f)-t) were positioned near the boundary in the phase diagram; f) water-acetonitrile-ethyl acetate of 0.25:0:10 volume ratio, g) 0.4:1:9 volume ratio, h) 0.55:2:8 volume ratio, i) 0.8:3:7 volume ratio, j) 1.1:4:6 volume ratio, k) 1.5:5:5 volume ratio, l) 2.05:6:4 volume ratio, m) 14:30:15 volume ratio, n) 4:6:2.4 volume ratio, o) 5:5:1.75 volume ratio, p) 6:4:1.3 volume ratio, q) 7:3:1.1 volume ratio, r) 8:2:0.95 volume ratio, s) 9:1:0.95 volume ratio, and t) 10:0:0.9 volume ratio. The carrier solutions f)-i) and p)-t) showed no separation behavior and presented just single peaks on the chromatograms with near the average linear velocity (data not shown). The carrier solutions j)-o) showed separation behavior as indicated in Figure 4. The component ratios of the solvents for solutions f)-t) were plotted in the phase diagram in Figure 2 with the symbols based on the chromatographic data.

As shown in Figure 2, the TRDC system was used with several homogeneous carrier solutions with component ratios of the solvents around the boundary curve in the phase diagram. Such specific carrier solutions caused tube radial distribution of the carrier solvents in the capillary tube under laminar flow conditions, i.e., the TRDP. The pressure imposed on the carrier solution in the capillary tube under laminar flow conditions might alter the carrier solution from homogeneous in the batch vessel to heterogeneous in the tube, affecting the tube radial distribution of the solvents in the capillary tube.

As shown in Figure 4, high resolutions of the analytes in the TRDC system were observed around the carrier solutions of l) and m) in the phase diagram. The chromategrams were examined in more detail with carrier solutions u)-w) that had compositions between solutions l) and m) along the boundary curve in the phase diagram (Figure 2). They were u) water-acetonitrile-ethyl acetate with 5:14.5:8.5 volume ratio, v) 5:13:7 volume ratio, and w) 6:13.5:6.5 volume ratio. The obtained chromatograms are shown in Figure 5 together with their resolutions (Rs). Carrier solutions v) and w) showed the best resolutions, ca. 14, under the present analytical conditions.

As characteristic separation behavior of the TRDC system using the PTFE capillary tube, we reported previously that a water-rich carrier solution elutes the more hydrophilic analyte first and an organic solvent-rich carrier solution elutes the more hydrophobic analyte first [7]. However, the reverse elution behavior of the analytes was not observed in Figure 4 under these analytical conditions. The analytical conditions, such as tube temperature, tube length, and flow rate, were significant and critical for separation performance in the TRDC system. Tentatively, the chromatograms were examined by the TRDC system using carrier solutions f)-t) where the capillary tube was dipped in the beaker controlled at 0˚C instead of 20˚C. The separation performance with solutions f)-t) were plotted with the symbols in the diagram in Figure 6. The chromatograms obtained with carrier solutions l)-q) that show separation are presented in Figure 7. Water-rich carrier solutions p) and q) gave the reverse elution order, 2,6-naphthalenedisulfonic acid first and then 1-naphthol second, although they showed splitseparation.

Furthermore, the analyte solutions were subjected to the TRDC system with the water-rich carrier solution of water-acetonitrile-ethyl acetate in a 15:3:2 volume ratio (water-rich carrier solution) under the analytical conditions of tube temperature 0˚C instead of 20˚C, tube effective length of 200 cm instead of 100 cm, and flow rate of 0.8 μL∙min^–1 instead of 0.2 μL∙min^–1. The obtained chromatogram is shown in Figure 8. 2,6-Naphthalenedisulfonic acid and 1-naphthol were eluted in this order with baseline-separation. The individual peaks were identified with the chromatogram of each analyte.

Almost the same elution behavior or elution order in the TRDC system with the fused-silica capillary tube was observed as that with the PTFE capillary tube reported in

our previous work [7]. However, there were also differences in separation behavior between them. In the fusedsilica capillary tube, the resolutions obtained with the organic solvent-rich carrier solution were better than those of the water-rich carrier solution. On the other hand, in the PTFE capillary tube, the resolutions obtained with the water-rich carrier solution were better than those of the organic solvent-rich carrier solutions as reported previously [7]. The interaction between the outer phase of water-rich and fused-silica inner wall as well as the outer phase of organic solvent-rich and PTFE (polymer material) inner wall may give prior conditions for making the outer phase (pseudo-stationary phase) in the capillary tube, while the interaction between the outer phase of water-rich and polymer material inner wall as well as the outer phase of organic solvent-rich and fused-silica inner wall may result in conditions inconvenient for forming the outer phase in the capillary tube.