Electrochromatographic separations of multi-component metal complexes on a microfluidic paper-based device with a simplified photolithography

Abstract

Electrochromatographic separations of three metal complexes were achieved through observing the visualization evolution process of moving chelation boundaries (MCBs) on microfluidic paper-based analytical devices (μPADs). A simplified photolithography was developed to pattern the μPADs within 5–10 min with three steps only, manually gelatinize, expose and develop, through thinning SU-8 2150 with trichloromethane. With a cutting method, many bubbles were clearly observed in the single-component MCB system on the μPADs, but were not found with this photolithography method under the same voltages. It was demonstrated that this photolithography for patterning narrow paper channels rejected Joule heating more effectively than the cutting paper method did. It was observed simultaneously that the apparent distances of electrophoretic and electroosmotic migrations of the blue [Cu-EDTA]²⁻ chelation boundary increased with increasing electrical field strength and width of the paper channels, respectively. The higher the Cu²⁺ concentrations were, the slower the electrophoretic migrations of the blue boundary were, and the faster the electroosmotic migrations of the blue boundary were on the μPADs with this photolithography. The separations and concentrations of the coloured zones, pink [Co–EDTA]²⁻, blue [Cu–EDTA]²⁻ and yellow [Fe–EDTA]²⁻, were clearly observed in the applied electric fields on the μPADs. The studies benefit extending the application of μPADs as low-cost, disposable analysis tools and deepen understanding of the comprehensive mechanisms concerning MCBs, partition adsorption and electroosmotic flow on μPADs.

---

1. Introduction

Microfluidic paper-based analytical devices (μPADs)^1–8 offer new opportunities for simple, low cost and portable analytical platforms that are as easy to use as paper strip tests. However, they are more complicated lateral-flow-type devices.^1–3 Whitesides’ group⁴ first introduced a photolithography method to fabricate μPADs by using a hydrophobic photoresist, SU-8 polymer. To date there are more than 10 methods^1–3 to pattern microfluidic paper channels. The processes of this photolithography⁴ included soak, spin, prebake, exposure, postbake, develop and oxygen plasma treatment prior to use. Its main disadvantages were the cumbersome manufacturing process and high cost. The group continually developed faster and lower cost techniques with home-made versions of the SU-8 photoresist to produce μPADs using an inkjet printer, a UV lamp and a hot plate with 7 steps in about 30 min.⁵ An alternative approach, wax printing, was independently presented by Lu et al.⁶ and Carrilho et al.⁷ with rapid fabrication in 5–10 min. Fenton et al. patterned μPADs by cutting a paper substrate with a computer-controlled x–y knife plotter, eliminating chemical treatments.⁸ However, lower cost, simpler and faster fabrication techniques for μPADs will need to be developed in the future for a variety of applications including health diagnostics, environmental monitoring and food quality testing.

The integration of advanced unit operations, such as separation of analytes and fluid manipulation, are still challenging on μPADs.⁹ Carvalhal et al.¹⁰ demonstrated chromatographic separation of uric and ascorbic acid by altering the solution pH to reduce the solubility of uric acid on a paper-based device. Yang et al.¹¹ separated plasma from whole blood by spotting antibodies onto paper to agglutinate red blood cells. Osborn et al.¹² separated analytes of different molecular weights using a laminar flow technique. However, it is not easy to implement laminar flow techniques on a paper-based device because they depend on capillary action for the different analytes. Abbas et al.¹³ exploited surface chemical gradients to separate complex samples by coating polyelectrolyte onto the paper. All these techniques are highly specific and do not easily extend to other systems. To the best of our knowledge, electrochromatographic separations of metal complexes have not been integrated onto μPADs.

In this paper, electrochromatographic separations of three metal complexes were performed in a μPAD with a simplified photolithography. Through thinning SU-8 2150 with trichloromethane (volume ratio 1 : 4) the simplified photolithography was demonstrated to fabricate μPADs in only three steps (manually gelatinize, expose and develop). It just needs exposure facilities (e.g. a UV lamp⁵). It is also a lower cost, simpler and faster (about 5–10 min) fabrication process compared with the photolithography proposed by Whitesides’ group. The visualization of electrophoresis and electroosmosis migrations of the coloured boundaries of metal complexes were investigated under different voltages, sizes of paper channels and analyte concentrations in the electrochromatographic devices. The studies will potentially extend the applications of μPADs through integrating pre-separation by paper electrochromatography or paper electrophoresis.

2. Materials and methods

2.1. Materials

A negative photoresist SU-8 2150 was purchased from MicroChem Corp. (Newton, MA). Whatman chromatography paper #1 (20.0 × 20.0 cm) was obtained from GE Healthcare Worldwide (Pudong Shanghai, China). Water with 18.25 MΩ cm⁻¹ resistance obtained from an Ulupure purification system was used throughout the experiments. Ethylenediaminetetraacetic acid disodium salt (Na₂EDTA, AR), copper sulphate (CuSO₄·5H₂O, AR), copper chloride (CuCl₂·2H₂O), cobalt chloride (CoCl₂·6H₂O), iron chloride (FeCl₃·6H₂O), sodium hydroxide (NaOH, AR), acetone and trichloromethane (CHCl₃, AR) were from Xi'an hairmer biotechnology Co., Ltd (Xi’an, China).

2.2. μPADs fabrication with a simplified photolithography and cutting method

The photolithography used in this paper for fabricating μPADs from Whatman #1 chromatography paper differed from the one reported by Martinez et al.⁴ First, SU-8 2150 negative photoresist was thinned with trichloromethane at the volume ratio (SU-8 2150 : trichloromethane) 1 : 4. Then about 1 mL of the thinned photoresist was coated onto 200 mm × 40 mm paper strips by hand in less than 30 s. The treated chromatography paper dried in less than 1 min due to the high volatility of trichloromethane. Second, the photomask was placed onto the coated paper, and exposed to ultraviolet (UV) light for 30 s (AB-M, Inc.). Lastly, after the exposure the unpolymerized photoresist was removed by washing the paper in acetone for about 3–5 min. The hydrophilicity of the paper channel and sampling areas must be ensured to obtain satisfactory flow rates of water solutions. Hence, the wick rate of pure water in the paper channel and sampling areas was checked by dripping water during development. Acetone washing was continually done until satisfactory hydrophilicity of the paper channel and sampling areas was obtained. With this simplified photolithography a total of 1.0 ml SU-8 photoresist and 5–10 min were spent to pattern a 200 mm × 40 mm paper strip. It would be estimated that our consumption of the photoresist is about 10% of the amount spent with the photolithography method of ref. 4 for the same size of paper strip. Martinez et al.⁵ reported a thinned SU-8 method that took about 30 min and used three pieces of equipment, an inkjet printer, a UV lamp and a hot plate. We inferred that the main diversities between Martinez's method and this simplified photolithography focused on the different thinning reagents and thinning ratios. In a word, this simplified photolithography method saves further the amount of SU-8 and fabrication time for patterning μPADs. Thin sheets of 200 mm × 200 mm Whatman #1 chromatography paper were cut into two different strips by hand using an ordinary small knife (Ningbo, China) and a ruler. The cutting method used here is similar to that used by Fenton et al.⁸

2.3. The separation principle of metal complexes on the electrochromatographic μPADs

Moving chelation boundary (MCB) electrophoresis,¹⁴ partition chromatography and electroosmotic flow (EOF) commonly occur in the electrochromatographic μPADs that we developed. Fig. 1 shows the structure, principles and pictures of the separation of three metal complexes on electrochromatographic μPADs.

Fig. 1 The structure and the principle of the electrochromatographic μPADs. (A) The electrochromatographic μPADs are composed of a μPAD, two centrifuge tubes and a rack. (B) The top side of the μPAD includes a sampling area, a hydrophilic channel, a UV cured photoresist, leading wires and tape. (C) The lower side of the μPAD includes double-sided tape and paper strips as two legs. The two dotted lines indicate the positions of the two legs replicated along the arrow direction. (D) The principle of the MCB in the hydrophilic paper channel. The symbols “+” and “−” indicate the anode and cathode, respectively. The top image of (D) shows the state for initially forming CBs among EDTA, Co²⁺, Cu²⁺ and Fe³⁺. The red dotted lines indicate the meeting position of [EDTAH₂]²⁻ and metal ions by capillary action only. The black arrows indicate the movement directions of [EDTAH₂]²⁻ and Co²⁺, Cu²⁺, or Fe³⁺, respectively. The red arrows indicate the direction of EOF in the paper channel. The middle image of (D) indicates the blue, pink and yellow zones and their positions. The bottom image of (D) is the principle of MCB formation with EDTA and three metal ions. (E) A picture of the MCB among [EDTAH₂]²⁻, Co²⁺, Cu²⁺ and Fe³⁺ in a 60 mm × 4 mm paper channel and 20 mm × 20 mm sampling area at 170 V formed by 0.20 M EDTA vs. 0.10 M Co²⁺, Cu²⁺ and Fe³⁺ at 15 min after their meeting.

The electrochromatographic μPAD is shown in Fig. 1(A). EDTA solution as the catholyte was in the left centrifuge tube, and the anolyte, Co²⁺, Cu²⁺ and Fe³⁺ mixed solution, was in the right one. Their solution pH was 5.0 in the two tubes. More than 95% of the EDTA existed as the subspecies [EDTAH₂]²⁻ under the given conditions.¹⁵ Two legs sucking the catholyte and the anolyte solutions were put into the centrifuge tubes, and a voltage was applied by electrophoresis apparatus (Xi'an TianLong Science and Technology Co., Ltd). As shown in Fig. 1(B), the leading wires were attached onto the top side of the two sampling areas of the patterned paper with transparent tape. The sampling areas and paper channel were covered by the transparent tape. As shown in Fig. 1(C), the two legs, paper strips with the same lengths and widths, were stuck onto the lower side of the two sampling areas of the patterned paper with double-sided tape. The sampling areas and paper channel of the lower side of the patterned paper were covered by the double-sided tape also. The top transparent tape, the middle patterned paper and the lower double-sided tape together constituted a closed unit to prevent solution evaporation. The other side of the double-sided tape was stuck to the centrifuge tube rack to fix the closed unit.

As shown in the top image of Fig. 1(D), the boundaries of [EDTAH₂]²⁻ and the three metal ions in the mixture solution moved in opposite directions through capillary action. Namely, [EDTAH₂]²⁻ moved towards the anode, and Co²⁺, Cu²⁺, and Fe³⁺ moved towards the cathode. Before the meeting between the boundaries of the flows of EDTA and the mixture solution, the left and the right of the paper channel were colorless and jasmine coloured, respectively. In this period, the solutions in the paper channel did not conduct electric current, although an electric field was applied. Actually, the process belonged to the frontal analysis¹⁶ of ion exchange chromatography based on partition between the stationary phase and mobile phase by capillary action. Generally, the process can be used as a pre-separation step for complicated mixtures. The electric current emerged when the boundaries of EDTA and the mixture solution met. At this moment the initial chelation boundary (CB) formed. This time was chosen as the zero point of recorded time for the CB migration. The electrokinetic phenomenona, electrophoresis and EOF, occurred in the paper channel, and chromatography separation occurred also. Thus, the process was called electrochromatography on the μPADs. As shown in the middle image of Fig. 1(D), the metal ions moved towards the cathode because of their positive charge, and [EDTAH₂]²⁻ migrated towards the anode due to its negative charge. After the EDTA and the mixture solution met, their boundaries intersected each other, and the initial CB migrated towards both the cathode and the anode. The direction of CB migration mainly depended on the directions and magnitudes of the EOF rate of the solution and the electrophoresis migration rates of the ions in the paper channels. At this stage the chelation reactions took place to generate the zones of the pink [Co–EDTA]²⁻, blue [Cu–EDTA]²⁻ and yellow [Fe–EDTA]²⁻. In the bottom image of Fig. 1(D), the MCB of the three metal complexes formed in the paper channel is visualised. Fig. 1(E) shows the experimental image of the MCB 15 min after the meeting of EDTA, Co²⁺, Cu²⁺ and Fe³⁺ in the paper channel. It must be emphasized that electrokinetic phenomenona (e.g. electrophoresis and EOF) played a leading role, and the chromatography action was less important after the meeting of EDTA and the metal ions in this period.

All experiments were carried out at room temperature (about 23 °C). A series of images were acquired with a 1 min time interval with a Sony (DSC-T900) digital camera. The positions of the front of the CB in the center region of the channel were measured with the ruler of the software Photoshop.

3. Results and discussion

In order to understand the separation process in the μPADs the key effects concerning electrochromatography were investigated for the paper size, voltage and Cu²⁺ concentration, etc. The photolithography and cutting methods were compared for the electric migrations of CBs in the patterned paper channels. Electrochromatographic separations of the three metal complexes were achieved on the μPADs when electric field and current were applied. However, the two components Co²⁺ and Cu²⁺ were not completely separated under the same conditions except with no applied electric field.

3.1. Comparing photolithography with the cutting method for electrokinetic effects

As shown in Fig. 2, the photolithography and cutting methods were compared through observing EOF, bubbles caused by Joule heating and distinct degrees of MCB on strips or in paper channels under the same conditions. The principle of MCB formation with EDTA and Cu²⁺ on paper strips or in paper channels is shown in Fig. S1† (for detailed procedures see ESI, Fig. S1†). Fig. 2(A) shows a serious broadening of the CB boundaries for the MCB on a 140 × 30 mm paper strip. In Fig. 2(A), ‘〈 ’ and ‘〉 ’ styles of fluid boundary appeared on the paper strip marked 1′. This means that the wick rates at the edges of the paper strip are faster than the ones in the central region. It also means that the edges of the flows of EDTA and Cu²⁺ meet prior to the meeting of the central flow region. This was verified since the color density at the edges of the initial CB was greater than that at the central region on the paper strip marked 0 min. It took about 3 min to form the initial blue thin CB by capillary action. The meeting position was about in the middle of the paper strip, at 7.05 cm from left to right. Therefore, the local flow rates were significantly different at the different local radial positions and randomly distributed paper-fibres contributed to the non-uniform or broadening fluid boundaries. These discussions are just for the stages (marked 1′ and 0 min) with no electric current before the meeting of the EDTA and Cu²⁺ boundaries. Only capillary action affects the fluid transport at this stage. After complete meeting of the EDTA and Cu²⁺ boundaries, the all blue chelation zones of the complex were located on the left side of the initial CB from 0–25 min. [Cu–EDTA]²⁻ with two negative charges should be located on the right side of the initial CB if there is only electrophoretic migration. EOF towards the cathode in the paper channel originated from the migration of hydrated cations, which dissociate from the carboxy groups of paper fibres as a stationary phase with pH 5.0. Therefore, we can reason that the EOF rate towards the left cathode is obviously larger than the rate of [Cu–EDTA]²⁻ electrophoretic migration. In addition, the blue zone gradually evolved, accompanying more irregular broadening from 0 to 25 min. Many bubbles were clearly observed near the cathode when Fig. 2(A) was enlarged to 300% in Word software. In Fig. 2(B) a paper strip with a 100 mm × 4 mm channel prepared by the cutting method was used under the same electrophoretic conditions. Many bubbles were also clearly observed in the blue zones when Fig. 2(B) was enlarged to 500% in Word software. This means that too much Joule heating was generated in both the 30 mm and 4 mm wide paper strips under the applied voltage of 150 V. In addition, the apparent EOF in the 4 mm wide paper strips was obviously smaller than in the 30 mm wide one, since the blue zone in Fig. 2(B) moved towards both the cathode and the anode, but the blue zone in Fig. 2(A) moved towards the cathode only. The larger the width of the paper strip was, the smaller the electrical resistance was, the more Joule heating there was, the smaller the solution viscosity was, and thus, the larger the EOF was. It was found that the irregular boundaries located at both sides of the blue zones directly result in blue band broadening. In Fig. 2(C) the MCB was achieved in a 4 mm wide paper channel patterned by photolithography. The bubbles and irregular boundaries were not found with the naked eye in the electrophoresis, even if Fig. 2(C) was enlarged to 500% in Word software. In fact, the bubbles were obtained in the electrophoretic processes of Fig. 2(A) and (B). Thus, Joule heating is a critical problem in traditional paper electrochromatography.^18,19 Although the cutting technology can be used to easily cut chromatography paper into various patterns without the need for chemical reagent treatment, the problems concerning Joule heating can be weakened by employing SU-8 treated paper channels as shown in Fig. 2(C) (e.g. 4 mm wide channels or narrower ones). As shown in Fig. 3–6 the following experiments for the formation of a MCB on electrochromatographic μPADs supported this viewpoint.

Fig. 2 CB migrations in paper strips from the cutting method and in a paper channel prepared by photolithography. (A) Cut paper strips: strip size 140 mm × 30 mm. 1′ denotes 1 min before the meeting of the EDTA and Cu²⁺ boundaries. (B) Cut paper strips: channel size 100 mm × 4 mm, sampling area 20 mm × 20 mm. (C) Photolithographic paper strip with SU-8 photoresist: channel size 100 mm × 4 mm, sampling area 20 mm × 20 mm. The effective length (120 mm) between the two electrodes was the same in (B) and (C). Experimental conditions: voltage 150 V; the cathode (−) on the left, and the anode (+) on the right; sample 0.20 M EDTA (−) and 0.20 M Cu²⁺ (+). The shoot times are listed on the right side of the images. Sampling mode: simultaneous sampling.

3.2. Effect of voltage

Fig. 3 shows the effects of voltage (e.g. 0, 30 and, 100 V) on the CB migration in the paper channel. Without an electric field, an evolution process of the CB is exhibited, as shown in Fig. 3(A), during 0–60 min. The shallower and indistinct blue CB was initially formed at 0 min. The meeting boundaries of [EDTAH₂]²⁻ and Cu²⁺ were driven by capillary action only before 0 min. During 0–60 min the centre of the blue zone inclined to the left side of the initial CB slightly. This could have resulted from the difference in hydrostatic pressure due to the slight difference in the level of electrolyte solution in the two centrifuge tubes. With a 2 mm length increase, the slight broadening of the blue zone during the 60 min is due to diffusion from a high concentration to a low one in the paper channel. It indicated that the broadening rate of the blue zone was 0.03 mm min⁻¹ in the paper channel. The blue chelation zone in Fig. 3(A) was very light compared to that in Fig. 3(B and C) and Fig. 2(C). In Fig. 3(A) the evolution of the blue chelation zone was driven by free diffusion and the difference of hydrostatic pressure, without an electric field. In Fig. 3(B and C) and Fig. 2(C) the evolution of the blue chelation zones was driven mainly by electrophoretic mobility and EOF in the electric field. Of course, free diffusion and the difference of hydrostatic pressure still had an effect, but this is slight. Therefore, the migration and broadening of the blue chelation zones in Fig. 3(B and C) and Fig. 2(C) are completely different from that shown in Fig. 3(A). In Fig. 2(C), the blue chelation zone on the left side of the black arrows was shallower than that on the other side during the 60 min. It is reasoned that the boundaries marked by the black arrows are the front boundaries of the Cu²⁺ phase. The chelation reaction occurred on the right side of the black arrows. The slight blue zone on the left side of the black arrows comes from blue [Cu–EDTA]²⁻, which was carried from the right side of the black arrows by EOF.

Fig. 3 Effect of voltage on CB migration in 100 mm × 4 mm paper channels prepared by photolithography. (A) 0 V; (B) 30 V; (C) 100 V; (D) curves of distance (x) against time (t) of MCB migration at different voltages. The red dotted line indicates the position of the initially formed CB. The size of the sampling area is 20 mm × 20 mm. Sample: 0.20 M EDTA and 0.20 M Cu²⁺. Sampling mode: simultaneous sampling. The MCB under 150 V is shown in Fig. 2(C). The detailed definitions of x_ep and x_eof are referred to in Fig. S1.†

For understanding, the red dotted line in Fig. 3(A–C) indicates the position of the CB initially formed in the paper channel. This was chosen as zero distance in CB migration. The distance, x_ep, from the front boundaries of the blue zones on the right side of the zero distance (red dotted line) was positive, and the negative distance, x_eof, was from the front boundaries on the left side to the zero distance in the paper channel. The changes of distance of the CB migrations with time, x_ep ∼ t and x_eof ∼ t, are shown in Fig. 3(D). The curves, x_ep ∼ t and x_eof ∼ t, indicated the non-linear relationship between CB migration and time. This is significantly different from the linear relationship or the constant velocity of CB migration in agarose gel-filled large tubes reported by Cao et al.¹⁴ EOF is a very complex electrokinetic phenomenon in paper electrochromatography. It depends upon many variable conditions, such as the paper material, its treatment, the electric field, the running time and hydrostatic pressure, etc.^17–19 In Fig. 3(D), x_ep and x_eof increase with the increasing electrical field strength within the same time (t). The effect of the length of the channel on the CB migration was similar to Fig. 3 (for detailed procedure see ESI, Fig. S2†).

Further, in the paper channel with no electric current the boundaries of EDTA and Cu²⁺ moved in opposite directions through capillary action only before 0 min. The errors among their meeting positions and times can reflect the quality of the paper channels patterned by the simplified photolithography. Their meeting times were 17.0, 17.0, 15.5 and 16.4 min, respectively. They were not shown in Fig. 3(A–C) and Fig. 2(C). Their meeting positions, indicated by the red dotted lines in Fig. 3(A–C) and Fig. 2(C) were 4.85, 5.02, 5.00 and 5.09 cm, respectively. The standard deviation (SD) of their meeting positions is less than 0.1010. This indicated that the simplified photolithography used in fabricating the μPAD is feasible.

3.3. Effect of the width of the paper channel

Fig. 4(A–E) show the effects of five widths of paper channel on CB migration. The non-linear curves, x_ep ∼ t and x_eof ∼ t, in Fig. 4(F) come from the measurements of the CB migration distances in Fig. 4(A–E) with different channel widths. From the non-linear curves, x_ep ∼ t, of Fig. 4(F) from bottom to top, the wider the paper channel (1, 2, 3, 4, 5 mm) is, the larger x_ep is. In particular, when the channel width changed from 3 mm to 2 mm, the influences of the hydrophobic walls of the paper channels formed by the photoresist SU-8 on CB migrations were remarkably reflected in the brilliant green curve and the red curve in Fig. 4(F). This means that the remarkable influence on the flow rates was observed due to the change of the hydrophobic specific surface from 3 mm to 2 mm. The fluid network of the paper channel was analogous to electrical resistance;²⁰ the channel width varies inversely with the electrical resistance. From the non-linear curves, x_eof ∼ t, from bottom to top, the narrower the paper channel width is, the slower the CB moves towards the left cathode. Before 0 min, the meeting of the boundaries, driven by capillary action, took 14.1, 11.0, 8.3, 7.0 and 7.1 min, which correspond to the widths 1, 2, 3, 4 and 5 mm, respectively, from Fig. 4(A) to Fig. 4(E). This supported the conclusion that the narrower the paper width is, the slower the flow rate of hydrophilic solution is in the hydrophilic paper channel with the hydrophobic wall. These results disagree with the points of Fu et al.²¹ that the flow rate of a fluid front is greater in a narrow paper strip than in a wider strip for a given time interval, without SU-8 treatment of the paper strips.

Fig. 4 Effects of the width of the paper channel on CB migration. Paper channel: (A) 60 mm × 1 mm; (B) 60 mm × 2 mm; (C) 60 mm × 3 mm; (D) 60 mm × 4 mm; (E) 60 mm × 5 mm. (F) Curves of distance (x) against time (t) of the MCB with different channel widths. The red dotted line indicates the position of the initial formed CB. The size of the sampling area is 16 mm × 16 mm. Voltage: 100 V; sample: 0.20 M EDTA and 0.20 M Cu²⁺; sampling mode: simultaneous sampling.

3.4. Effect of Cu²⁺ concentration

Fig. 5 shows that the CB migrations of the chelation zones depend on the Cu²⁺ concentration (varied from 0.050 to 1.0 M). Time-difference sampling was adopted in this experiment. Specifically, by capillary action EDTA solution was sucked by the left leg when the Cu²⁺ boundary reached the center (5 cm) of the 10 cm paper channel. The results of the time-difference sampling are shown in the images marked by 0′ in Fig. 5(A–C). In Fig. 5(D), from the non-linear curves, x_ep ∼ t, from top to bottom, the higher the Cu²⁺ concentration, the slower the migration of the front CB on the right of the red dotted lines. The CBs were formed by the reaction of equivalent moles of EDTA and Cu²⁺. Thus, the higher the concentration of Cu²⁺ phase in the reaction boundary, the more EDTA is consumed in moving the EDTA phase towards the right anode. Thus, the blue CB moved to the right slowly. These experimental results agreed with Fan et al.¹⁴ From the non-linear curves, x_eof ∼ t, from bottom to top, the higher the Cu²⁺ concentration, the faster the slight blue zones on the left side of the red dotted line move towards the left cathode. This meant that EOF quickly moved towards the left in the paper channels with high Cu²⁺ concentration. Further, the theoretical studies for these experimental results are currently being done by the authors.

Fig. 5 Effect of Cu²⁺ concentration on CB migration using time-difference sampling; (A) 0.050 M Cu²⁺; (B) 0.20 M Cu²⁺; (C) 1.0 M Cu²⁺. (D) Curves of distance (x) against time (t) of the MCB with different Cu²⁺ concentrations. The size of the sampling area is 20 mm × 20 mm. Paper channel: 100 mm × 4 mm; voltage: 150 V; sample: 0.20 M EDTA.

3.5. Separations of multi-components on the electrochromatographic μPADs

Realizing separations and concentrations of multi-components in paper-based microchannels is challenging for microfluidic applications.⁹ In this study MCB electrophoresis was achieved on μPADs to separate and concentrate multi-component metal complexes. This will enrich dramatically the applications of μPADs, and also deepen the understanding of the compositive mechanisms concerning MCBs, partition adsorption and EOF on μPADs. In this section the separations by MCB electrophoresis of two- and three-component metal complexes were exhibited, and the effects of the applied electric field were discussed.

Without an electric field, Co²⁺ and Cu²⁺ complexes were not separated when the μPADs were used for analogous paper chromatography. This was validated by observing a dim zone in the whole paper channel during 60 min, as shown in Fig. 6(A). Under the conditions employed here, the CB migration mainly depended on the hydrostatic pressure¹⁸ and the streaming potential²² on the μPADs. The width of the zone was less than 2 mm, as seen in Fig. 6(A), since both the streaming potential and hydrostatic pressure were very small.

As shown in Fig. 6(B), the clear separation and enrichment processes of two-component zones, pink [Co–EDTA]²⁻ and blue [Cu–EDTA]²⁻, were exhibited during 0–60 min. The clear boundary between the two zones of [Co–EDTA]²⁻ and [Cu–EDTA]²⁻, namely, the moving substitution boundary,²³ remained for a very long time (e.g. 60 min). This hints that the two-component zones would never mix, although the component amount in the zones increased with time continuously. This is just a concentration process. It also differed evidently from paper electrophoresis or paper electrochromatography without a chelation reaction between metal ions and EDTA. It is clear that this moving chelation reaction boundary in the channels with SU-8 treatment can effectively resist the dispersion of zones, although the zone dispersion caused by rough paper fibres is much greater than that in agarose gel-filled tubes.^14,23 Perhaps, the channels with SU-8 treatment can reduce the EOF in paper substrates, as shown in Fig. 2. In agarose gel-filled tubes the separation between [Co–EDTA]²⁻ and [Cu–EDTA]²⁻ was considered as an isotachophoresis mode.²³ However, on the electrochromatographic μPADs, the separation principle of the two components was more complicated in this study. Variable EOF existed in this electrochromatographic μPAD, as shown in Fig. 2–5, but it can be ignored in the agarose gel-filled tube.²³Ref. 23 revealed that the left and the right boundaries beside the pink zone of [Co–EDTA]²⁻ moved at the same velocity in the agarose gel-filled tube. This was called isotachophoresis. However, Fig. 6(B) shows the different apparent velocities of the two boundaries beside the pink zone of [Co–EDTA]²⁻ due to variable EOF on this μPAD. The separation between the two zones, [Co–EDTA]²⁻ and [Cu–EDTA]²⁻, should belong to the category of isotachophoresis, although chromatography partitioning existed in the separation process on the μPADs.

Fig. 6 Separations of multi-component metal complexes with or without applied electric field. (A) Paper channel 100 mm × 4 mm without electric field. Two component sample: 0.10 M Cu²⁺, 0.10 M Co²⁺ and 0.20 M EDTA. (B) Voltage: 150 V. The other conditions are the same as in (A). (C) Voltage: 170 V. Three component sample: 0.10 M Cu²⁺, 0.10 M Co²⁺, 0.10 M Fe³⁺ and 0.20 M EDTA. Paper channel: 60 mm × 4 mm.

For the three-component system, Co²⁺, Cu²⁺ and Fe³⁺ were chosen as the model ions of separation in the paper-based channel herein. A novel separation and enrichment of the three metal complexes was achieved on a μPAD for the first time, even if in agarose gel-filled tubes or other separation materials separation of the multi-component system based on MCBs was not found.^{14, 23–25} As shown in Fig. 6(C), the four clear boundaries, the left of the pink [Co–EDTA]²⁻ zone, the pink [Co–EDTA]²⁻–blue [Cu–EDTA]²⁻ boundary, the blue [Cu–EDTA]²⁻–yellow [Fe–EDTA]²⁻ boundary and the right of the yellow [Fe–EDTA]²⁻ zone, were formed and kept from 3 min to 18 min.

The boundary on the left of the pink [Co–EDTA]²⁻ zone in Fig. 6(C) reflected the EOF of the paper channel, similar to the boundary on the left of the [Cu–EDTA]²⁻ zone in the single-component system in Fig. 2(C). The two middle boundaries, the pink [Co–EDTA]²⁻–blue [Cu–EDTA]²⁻ and the blue [Cu–EDTA]²⁻–yellow [Fe–EDTA]²⁻ boundary, were named as moving substitution boundaries, and are related to the relative magnitude of the stablity constants of the metal complexes in the boundary.²⁶ The boundary on the right of the yellow [Fe–EDTA]²⁻ zone was related to the apparent mobility of the metal complex, similar to the boundary on the right of the [Cu–EDTA]²⁻ zone in the single-component system in Fig. 2(C). Theoretically, most metal complexes should be separated by the MCB mechanism on the μPADs. However, it is difficult to observe clear boundaries with the naked eye since many metal complexes are colorless.

The concentrations of metal ions employed were relatively high in this study. This disadvantage of the μPADs could be overcome by integrating a special concentration unit (e.g. a bipolar electrode²⁷) onto μPADs or by using sensitive detection apparatus. We are currently working to overcome these challenges and the results will be reported further.

4. Conclusions

(1) A simplified photolithography including the three steps, manually gelatinize, expose and develop, was performed to pattern μPADs in about 5–10 min through thinning the SU-8 2150 with trichloromethane at the volume ratio 1 : 4. With this simplified photolithography, the time and cost of μPAD fabrication were saved through omitting the steps, spin, prebake, postbake and oxygen plasma treatment.

(2) Electrochromatographic μPADs were invented by integrating the patterned paper strip, leading wires as electrodes attached onto the sampling areas, two fixed centrifuge tubes and a direct current power. Four clear boundaries with different colors were observed in the visualization evolution process of the separation and enrichment for three metal complexes on the μPADs. This will enrich the applications of μPADs.

(3) It was understood that, as two positive factors, paper-based channels with SU-8 treatment and moving substitution boundaries of multi-component metal complexes resisted the dispersion of component boundaries on rough paper fibres through reducing EOF or Joule heating effectively. This deepened the understanding of the comprehensive mechanisms concerning MCBs, partition adsorption and EOF on μPADs.

Acknowledgements

The authors are thankful for financial support from the National Science & Technology Pillar Program of China (no. 2009BAK59B02-04). The authors are also thankful for financial support from the National Natural Science Foundation of China (no.21377102). The authors have declared no conflict of interest.

References

1. A. W. Martinez, S. T. Phillips, G. M. Whitesides and E. Carrilho, Anal. Chem., 2010, 82, 3–10.
2. X. Li, D. R. Ballerini and W. Shen, Biomicrofluidics, 2012, 6, 011301.
3. D. R. Ballerini, X. Li and W. Shen, Microfluid. Nanofluid., 2012, 13, 769–787.
4. A. W. Martinez, S. T. Phillips, M. J. Butte and G. M. Whitesides, Angew. Chem., Int. Ed., 2007, 46, 1318–1320.
5. A. W. Martinez, S. T. Phillips, B. J. Wiley, M. Gupta and G. M. Whitesides, Lab Chip, 2008, 8, 2146–2150.
6. Y. Lu, W. W. Shi, L. Jiang, J. H. Qin and B. C. Lin, Electrophoresis, 2009, 30, 1497–1500.
7. E. Carrilho, A. W. Martinez and G. M. Whitesides, Anal. Chem., 2009, 81, 7091–7095.
8. E. M. Fenton, M. R. Mascarenas, G. P. Lopez and S. S. Sibbett, ACS Appl. Mater. Interfaces, 2009, 1, 124–129.
9. K. N. Ren, J. H. Zhou and H. K. Wu, Acc. Chem. Res., 2013, 46, 2396.
10. R. F. Carvalhal, M. S. Kfouri, M. H. D. O. Piazetta, A. L. Gobbi and L. T. Kubota, Anal. Chem., 2010, 82, 1162–1165.
11. X. Yang, O. Forouzan, T. P. Brown and S. S. Shevkoplyas, Lab Chip, 2012, 12, 274–280.
12. J. L. Osborn, B. Lutz, E. Fu, P. Kauffman, D. Y. Stevens and P. Yager, Lab Chip, 2010, 10, 2659–2665.
13. A. Abbas, A. Brimer, J. M. Slocik, L. Tian, R. R. Naik and S. Singamaneni, Anal. Chem., 2013, 85, 3977–3983.
14. L. Y. Fan, C. J. Li, W. Zhang, C. X. Cao, P. Zhou and Z. X. Deng, Electrophoresis, 2008, 29, 3989–3998.
15. A. Ringbom, Complexation in Analytical Chemistry, Interscience, 4th edn, 1963.
16. Q. Liu, L. F. OuYang, H. Liang, N. Li and X. D. Geng, J. Sep. Sci., 2012, 35, 1411–1423.
17. J. L. Engelke, H. H. Strain and S. E. Wood, Anal. Chem., 1954, 26, 1864–1868.
18. S. E. Wood and H. H. Strain, Anal. Chem., 1954, 26, 1869–1872.
19. M. A. Hubbe, Bioresources, 2006, 1, 116–149.
20. L. A. Godwin, K. S. Deal, L. D. Hoepfner, L. A. Jackson and C. J. Easley, Anal. Chim. Acta, 2013, 758, 101–107.
21. E. Fu, B. Lutz, P. Kauffman and P. Yager, Lab Chip, 2010, 10, 918–920.
22. V. Leung, A.-A. M. Shehata, C. D. M. Filipe and R. Pelton, Colloids Surf., A, 2010, 364, 16–18.
23. W. Zhang, J. F. Chen, L. Y. Fan, C. X. Cao, J. C. Ren, S. Li and J. Shao, Analyst, 2010, 135, 140–148.
24. C. X. Cao, L. Y. Fan and W. Zhang, Analyst, 2008, 133, 1139–1157.
25. J. Jin, J. Shao, S. Li, W. Zhang, L. Y. Fan and C. X. Cao, J. Chromatogr., A, 2009, 1216, 4913–4922.
26. F. Basolo and R. C. Johnson, Coordination Chemistry: The Chemistry of Metal Complexes, W. A. Benjamin, Inc., New York, 1964, pp. 54–62.
27. J. Dai, T. Ito, L. Sun and R. M. Crooks, J. Am. Chem. Soc., 2003, 125, 13026–13027.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra43625j

---