A Transdermal Measurement Platform Based on Microfluidics

The Franz diffusion cell is one of the most widely used devices to evaluate transdermal drug delivery. However, this static and nonflowing system has some limitations, such as a relatively large solution volume and skin area and the development of gas bubbles during sampling. To overcome these disadvantages, this study provides a proof of concept for miniaturizing models of transdermal delivery by using amicrofluidic chip combinedwith a diffusion cell.The proposed diffusionmicrochip system requires only 80 μL of sample solution and provides flow circulation. Twomodel compounds, Coomassie Brilliant BlueG-250 and potassium ferricyanide, were successfully tested for transdermal delivery experiments. The diffusion rate is high for a high sample concentration or a large membrane pore size. The developed diffusion microchip system, which is feasible, can be applied for transdermal measurement in the future.


Introduction
Transdermal drug delivery has been a topic of considerable research in the field of pharmaceuticals for decades [1].Current and future prospects of transdermal drug delivery have been reviewed thoroughly in many articles.[2][3][4].Transdermal delivery is an extremely attractive and innovative route through which drugs exert systemic effects [4].It has benefits such as avoiding first-pass metabolism, sustained therapeutic action, and greater patient compliance; thus, it is a potential noninvasive route for drug administration [5].
The Franz diffusion cell is one of the most accepted models for transdermal study [6][7][8][9].The Franz diffusion cell is a static and nonflowing system [10].To extend its applications, various improvements have been proposed.A flow-through cell with an automatic fraction collector saved time and labor compared with the conventional Franz cell [11].Sequential injection chromatography coupled with the Franz cell was developed for the separation and simultaneous determination of two substances in a topical pharmaceutical formulation [12].However, the Franz diffusion cell still has some limitations such as the requirement of a relatively large solution volume and a large skin area.During sampling, gas is often generated between the donor and receptor and interferes with the transdermal diffusion.
This paper proposes a transdermal delivery system with a microfluidic chip for simple, fast, and low-volume sample detection.Two model compounds, Coomassie Brilliant Blue G-250 and potassium ferricyanide, were tested in this platform.This diffusion microchip system can be applied for evaluating transdermal delivery in the future.

Development of the Diffusion Microchip.
The diffusion microchip (Figure 1) comprises six layers, of which layers (i), (ii), (iv), and (vi) are conventional polymethylmethacrylate substrates constructed by a CO 2 laser machine, and (iii) and (v) are polydimethylsiloxane (PDMS) films.The top layer, layer (i), has dimensions of 25.0 mm × 25.0 mm, a sample donor (diameter/depth: 8.0 mm/2.0 mm), and four screw orifices.Layers (ii) and (vi) are clamp layers with thicknesses of 2.0 mm and 4.0 mm, respectively.Layer (iv) is the main channel layer for microfluidics and is 2.0 mm thick.The receptor has the same size as the donor, and the diameters of the inlet and outlet are 2.0 mm.The length, width, and depth of the main microchannel are 60.0, 1.0, and 2.0 mm, respectively.The PDMS film, which contains 15 holes for the receptor, inlet/outlet, and 12 screw orifices, is applied to prevent leakage of liquid from the microchannel.Figure 2 shows the composition of the diffusion microchip device.First, layers (ii)-(vi) in Figure 1 are connected with eight screws (diameter/pitch: 2/8 mm) to construct the receptor part.After the membrane is inserted, the top layer providing the donor site and the receptor part are connected with another four screws to form the diffusion microchip device.

Setup of the Diffusion Microchip System. As shown in
Figure 3, the experimental system is comprised of a diffusion microchip, a solution reservoir, and a peristaltic pump (PS103, YOTEC, Hsinchu, Taiwan).A tube (Tygon E-3603 Tubing, Saint-Gobain, Aurora, USA) was installed at the inlet and outlet and then fed through the peristaltic pump to the solution reservoir (15 mL centrifuge tube).The solution was driven by the peristaltic pump, slowly entered the diffusion microchip through the inlet entrance (red line in Figure 3), and proceeded to the microchannel.The outflow exited through the outlet (blue line in Figure 3) and was conveyed through the peristaltic pump to the solution reservoir for further circulation.
The diffusion solution (80 L) was loaded into the donor cell and covered with Parafilm5 M (Sigma, St. Louis, USA) to prevent evaporation.At predetermined time points, 200 L of the solution was collected from the reservoir for detecting the sample concentration by using an absorbance reader (Sunrise6 ELISA, Tecan, Zürich, Switzerland).After measurement, the solution was returned to the reservoir to maintain a fixed flow volume.

Effectiveness of the Diffusion Microchip System.
To test the feasibility of this diffusion microchip system, Coomassie Brilliant Blue G-250 (Bionovas, Toronto, Canada) and potassium ferricyanide (Merck, Darmstadt, Germany) were selected, respectively, as model color compounds for observing molecular diffusion through the donor cell into the receptor and sequential flow in the microchannel.As shown in Figure 3, a membrane with a 5.00 m pore size (Nylon, Toson, Hsinchu, Taiwan) was placed between the receptor and donor and fixed with four screws.A combination of 80 L of 1% Coomassie Brilliant Blue G-250 and potassium ferricyanide was used as the diffusion solution.At predetermined time points, 1 mL of solution in the reservoir was collected for examination by using a UV-visible spectrophotometer (V-630, JASCO, Tokyo, Japan).After examination, the solution was returned to the reservoir to maintain a fixed solution volume.

Transdermal Drug Delivery
2.4.1.Effects of the Membrane Pore Size.Three membrane pore sizes, 0.10, 0.45, and 5.00 m, were tested on this platform; 80 L of 1.2% potassium ferricyanide in deionized water was used as the diffusion solution.At predetermined time points, 200 L of the solution was collected from the reservoir to detect the sample concentration by using an absorbance reader at 420 nm.The solution was returned to the reservoir after the assay.

Effects of the Sample Concentration.
Potassium ferricyanide solution (80 L) at three concentrations in deionized  water, 0.4%, 0.8%, and 1.2%, was used as the diffusion solution in this experiment.The membrane pore size was 5.00 m.
The other procedures were the same as those described in Section 2.4.1.

Characteristics of the Diffusion Microchip.
The Franz diffusion cell is one of the most commonly used models for in vitro transdermal studies.The Franz diffusion cell measures 11.0, 2.4, and 8.6 cm in length, width, and height, respectively (Table 1 and Figure 4(a)).The membrane used in this device is 25 mm in diameter.The diffusion microchip developed in this study is relatively small.The length, width, and height of the chip are 7.5, 2.4, and 1.2 cm, respectively (Table 1 and Figure 4(b)).The membrane in the chip has a diameter of only 10 mm.Furthermore, the required injection solution volume for diffusion is only 80 L.Compared with the Franz diffusion cell, the diffusion microchip considerably reduces the sample and skin material requirements.As shown in Figure 5, the diffusion microchip can circulate the sample by using the peristaltic pump; however, the sample is stirred using a magnet in the Franz diffusion cell.Figure 5 illustrates sample diffusion through a membrane and its circulation in the microchannel.This system can stimulate transdermal drug delivery with blood circulation in vivo.Sampling is performed in the solution reservoir outside the chip, and no bubbles occur during sampling.Thus, the source of error source in the Franz diffusion cell is eliminated.

Verification of the Diffusion Microchip
System.The diffusion of Coomassie Brilliant Blue G-250 and potassium ferricyanide was tested in the diffusion microchip system.In accordance with previous reports [31][32][33], Coomassie Brilliant Blue G-250 exhibited maximum absorption at approximately 595 nm, as shown in Figure 6.Furthermore, no signal was detected in the circulating deionized water at the initial time.As time passed, the sample concentration in the solution reservoir increased, indicating continuous diffusion of Coomassie Brilliant Blue G-250 through the membrane.
According to a previous study [34], potassium ferricyanide has an absorption peak at 424 nm, as shown in Figure 7.At the initial time, no absorbance was observed in the tested sample.The absorbance at approximately 424 nm in the solution increased with diffusion time.After 5 h, there was nearly no change in the absorbance curve.

Evaluation of Transdermal Drug Delivery
3.3.1.Effects of the Membrane Pore Size. Figure 8 shows the effects of the membrane pore size on transdermal delivery for 1.2% potassium ferricyanide solution.Regardless of the membrane pore size, the absorption value at 424 nm increased with diffusion time.When the absorbance reached a plateau, the diffusion of potassium ferricyanide was complete.The smaller the pore size of the membrane, the smaller the value of the absorbance increase and the longer the time required for the absorbance to reach the plateau.The total time for the diffusion process was approximately 4 h, 5.5 h, and 7 h for pore sizes of 5.00, 0.45, and 0.10 m, respectively.

Effects of the Sample Concentration.
Figure 9 shows the effects of the sample concentration on transdermal delivery.Three potassium ferricyanide concentrations, 0.4%, 0.8%, and 1.2%, were compared with the membrane pore size fixed at 5 m.The results indicate that the diffusion rate was higher at a higher concentration of potassium ferricyanide.The time to reach the concentration plateau was shorter at a lower concentration of potassium ferricyanide.The higher the concentration of potassium ferricyanide was, the higher the plateau concentration reached.

Conclusions
This study proposes a diffusion microchip for evaluating transdermal delivery.Compared with the widely applied Franz diffusion cell, the diffusion microchip system requires a lower volume of sample, and no gas bubbles formed during the experiments.Model compounds, Coomassie Brilliant Blue G-250 and potassium ferricyanide, were used to demonstrate the feasibility of this platform.A high sample concentration or a large membrane pore size resulted in a high diffusion rate.In the future, this diffusion microchip system can be used to determine molecular diffusivity in transdermal delivery.

Conflicts of Interest
The authors declare no conflicts of interest.

Figure 1 :
Figure 1: Expanded view of the diffusion microchip device.The top layer has a sample donor and the second layer contains the receptor and inlet/outlet.The main channel layer provides circulating flow that transports targets for measurement.

Figure 5 :
Figure 5: Illustration of diffusion microchip with the diffusion process.(a) No sample added; (b) sample added to donor; (c) sample diffusion through the membrane; (d) sample circulation in the microchannel.

Figure 6 :Figure 7 :
Figure 6: Absorbance curves of Coomassie Brilliant Blue G-250 measured in the solution at different diffusion times.Initially, 1% Coomassie Brilliant Blue G-250 was used as the diffusion solution in the donor cell.
m pore size membrane 0.45 m pore size membrane 0.10 m pore size membrane

Figure 8 :Figure 9 :
Figure 8: Effects of the membrane pore size on transdermal delivery of 1.2% potassium ferricyanide.

Table 1 :
Size difference between the traditional transdermal drug delivery and diffusion microchip system.