A novel method of manufacturing a microchannel with integrating three-dimensional microstructure arrays for mixing experiment

Basic operating units, such as micropumps, microchannels, and micromixing chambers, are fabricated in the microfluidic chip.¹ Due to its characteristics, such as small size, low cost, and fast analysis, the microfluidic chip realizing rapid analysis of liquids is mainly applied in biological, chemical, and medical analysis as well as other fields. In order to ensure the practicability of the microfluidic chip, efficiently processing microstructures in the microfluidic chip is necessary to meet the accuracy in accordance with the designrequirements.

Different microstructure processing technologies can process different materials. For example, the polydimethylsiloxane (PDMS) chip² can be obtained through the molding method;³ however, the microchannel formed by PDMS has the disadvantage of being easily collapsed and deformed. Fortunately, the PMMA material microstructure adopts micro-hot-embossing⁴ compensates for the deficiency of PDMS. Nonetheless, if the microstructure changes, the mold needs to be remanufactured, and hence it is not suitable for the development and production of prototype microfluidic chips. After the method of hydrogel molding⁵ is proposed, the difficulty of the mold is effectively solved. Unfortunately, the width of the processed microchannel becomes larger at the corner. Although paper-based microfluidic chips are processed via inkjet printing⁶ and screen printing⁷ to avoid the above problems, paper-based microfluidic chips⁸ can usually only be utilized once. The substrate directly produced via 3D-printing⁹ technology can make up for the shortcomings of 2D-printed paper-based microfluidic chips, whereas the transparent printing materials available for microfluidic chips are very limited. It is useful to note that the glass material has obvious advantages in many aspects, such as optical properties and mechanical properties.¹⁰ The traditional glass microstructure processing method is wet chemical etching;¹¹ nevertheless, the photolithography process steps (such as glue coating, glue baking, exposure, development, and post-baking) are complicated and difficult to promote. Compared with wet etching, the method of employing glass-etching cream¹² greatly simplifies the process flow. Howbeit, the processed microchannel has poor edge regularity.

Remarkably, it is discovered that the femtosecond laser¹³ can process microchannels with regular edges on the glass surface. However, despite all that, this method is difficult to deal with rectangular cross section channels. Noticeably, the micromilling technique¹⁴ can control the cross-sectional shape of the microchannel through the shape of the tool. Inadequate current research only processed some simple microchannels and holes.¹⁵ 

In this paper, the glass microcolumn array utilized for mixing microfluidic chips is processed via the method of micromilling. The contact angles of the three microstructures are compared, and the influence of the size parameters of the microcolumn array on the contact angle is further discussed. The glass chip with the microcolumn structure and the PDMS chip are satisfactorily bonded. The mixing experiment of demonized water and blue ink verifies the feasibility of the mixing chip with microcolumn arrays. After comparing with other process methods, the characteristics of this method are summarized.

The microfluidic chip is composed of a PDMS chip and a glass chip with the microstructures. The process steps are divided into three parts including glass chip preparation, PDMS chip preparation, and chip bonding in Fig. 1. The specific steps are shown in Table I.

In order to verify the practicability of the mixing microfluidic chip, a mixing experiment is implemented as shown in Fig. 1(c). A syringe pump (LSP02-2A, Longer Constant Flow Pump Co., Ltd., China) is employed to inject demonized water and blue ink into the microchannels simultaneously. Before the mixing experiment, set the flow rate (0.01, 0.02, and 0.03 ml/min) of the liquid first and ensure that the connection between the microfluidic chip and the syringe (1 ml) is reliable. When performing mixing experiments, a microscope (OLYMPUSIX73, Olympus Co., Tokyo, Japan) is utilized to record images of liquid mixing. After the recording is completed, the mixed pictures taken from the video will be processed with the ImageJ software. Finally, the mixing efficiency is measured by the gray value extracted from the image.

The schematic diagram and the chip photograph are illustrated in Figs. 2(a) and 2(b), respectively. Three different shapes including a microcolumn, a microcone, and a microsphere are fabricated, as shown in Figs. 2(c)–2(e), respectively.

Compared with the hydrophilic surface, a slip flow phenomenon will occur when the liquid flows on the hydrophobic surface. This will lead to low flow resistance and high flow rate.^16,17 According to the Reynolds equation,

Here, ρ, v, μ, and d are the density, velocity, dynamic viscosity, and the characteristic size of the microchannel, respectively.

It can be seen from the equation that the greater the speed is, the greater the Reynolds number is. Furthermore, the large Reynolds number is beneficial to the turbulent flow state of the fluid. Consequently, more sufficient the mixing is, the surface is more hydrophobic.

The liquid contact angle is a measuring standard of a surface hydrophilicity. In order to obtain a hydrophobic surface corresponding to a large contact angle, a contact angle experiment is implemented. All the measurement data obtained in the contact angle measurement experiment in this article are the average of six samples.

Figure 3 demonstrates the contact angles of three different microstructures. The bottom diameter and depth of the three structures are identical. The bottom diameter is 200 µm, and the array depth is 100 µm. It can be seen that the contact angle of the three microstructures of microcone (35.9 ± 2), microsphere (52.1 ± 1.8), and microcolumn (74.3 ± 2.4) gradually increases, which may be related to the surface shape of the microstructure. Compared with the plane of the microcolumn, the inclined surface of the microcone and the curved surface of the microsphere are not conducive to the retention of liquid droplets, which strengthens the tendency of the liquid to flow toward the bottom of the array. Compared with microspheres, the apex of the microcone further strengthens this flow tendency. Therefore, the contact angles of the microcone array, the microsphere array, and the microcolumn array increase in turn. Consequently, among the three structures, the hydrophobicity of the microcolumn array is the most accomplished.

To further investigate the relationship between the microstructure size and contact angle, the influence of three parameters including the depth, diameter, and spacing on the contact angle is investigated through taking the microcolumn array with better hydrophobicity as an example.

Specifically, the Wenzel model¹⁸ [Eq. (2)] is introduced to build the theoretical model. Meanwhile, Eq. (2) on the roughness factor R expressed through the depth H, the diameter D, and the spacing L (Fig. 4) is employed. Moreover, the model of microcolumn contact angle (4) is acquired through bringing Eq. (3) into Eq. (2),

Here, R, θ_w, S₁, S₂, and θ_c represent the solid surface roughness factor (the ratio of the actual contact area to the apparent contact area), the contact angle under the Wenzel model, the apparent contact angle on the smooth and flat surface, the actual contact area between liquid and solid, the projected area of the bottom of the water drop on the horizontal plane under the Wenzel model, and the contact angle of the microcolumn array, respectively.

Figure 5 exhibits the change in the contact angle with different microcolumn array depths. As illustrated in Fig. 5(a), for the given parameters including the diameter of 300 µm and the spacing of 200 µm, as the depth of the microcolumn increases from 50 to 200 µm, the contact angle decreases from 79.9° ± 2.4° to 37.9° ± 1.4°. According to Eq. (4), when the array depth H increases, cos θ_w increases due to the increase in the surface roughness factor R. Consequently, the contact angle θ_w decreases. It can be seen that the trend is consistent with the experimental results. Figures 5(b)–5(e) display the photographs of contact angles with different array depths. It can be concluded that the greater the depth of the microcolumn array is, the smaller the contact angle is.

Figure 6 reveals the change in the contact angle with different microcolumn array spacings. As shown in Fig. 6(a), for the given parameters including the diameter of 300 µm and the depth of 100 µm, as the spacing of the microcolumn increases from 250 to 400 µm, the contact angle increases from 48.2° ± 1.8° to 66.2° ± 2.4°. According to Eq. (4), when the array spacing L increases, cos θ_w decreases due to the decrease in the surface roughness factor R. Consequently, the contact angle θ_w increases. It can be seen that the trend is consistent with the experimental results. Figures 6(b)–6(e) expose the photographs of contact angles with different array spacings. It can be concluded that the greater the spacing of the microcolumn array is, the larger the contact angle is.

Figure 7 demonstrates the change in the contact angle with different microcolumn diameters. As indicated in Fig. 7(a), for the given parameters including the spacing of 200 µm and the depth of 100 µm, as the diameter of the microcolumn increases from 250 to 400 µm, the contact angle increases from 74.6° ± 1.3° to 83.9° ± 1.4°. According to Eq. (4), when the microcolumn diameter D increases, cos θ_w decreases due to the decrease in the surface roughness factor R. Consequently, the contact angle θ_w increases. It can be seen that the trend is consistent with the experimental results. Figures 7(b)–7(e) present the photographs of contact angles with different microcolumn diameters. It can be concluded that the greater the diameter of the microcolumn array is, the larger the contact angle is.

The mixing image is processed with the ImageJ software to obtain the gray value in Fig. 7(b). Then, the mixing phenomenon is quantitatively analyzed. Since the uniformity of the gray value of the pixels in the image can represent the mixing degree of the working fluid, the gray value of the standard deviation method is employed to calculate the mixing index,

In Eq. (5), the mixing index I_E¹⁹ (0 < I_E < 1) is defined. The larger the I_E value, the better the mixing effect is. C_i is the gray value of each pixel in the mixing image. N is the total number of pixels.

Three sampling areas are selected as shown in Fig. 8(b). The mixing index of P1, P2, and P3 with different flow rates is given in Fig. 8(a). It can be observed that the mixing index of P1, P2, and P3 gradually increases, and the index change from P1 to P2 is greater than the change from P2 to P3. Picking up the flow rate of 0.01 ml/min as an example, the difference in the mixing index between P1 and P2 is 0.027, and the difference in the mixing index between P2 and P3 is 0.009. From the perspective of the mixing index, the mixing effect of the microcolumn array is three times that of the ordinary microchannel. This is due to the secondary flow²⁰ when the fluid passes through the microcolumn array during the flow process, which increases the contact area between the fluids. Thus, the mixing effect is enhanced. Consequently, the mixing chip with glass microcolumn arrays is available. Moreover, it can be found from the figure that the flow rate also has an effect on the mixing efficiency. The larger the flow rate is, the better the mixing effect is. As the flow rate increases, the great convection intensity of the fluid under the action of the mixing element is beneficial to uniform mixing.

Table II lists the different microstructure processing methods. Considering comprehensively the process steps and processing time, wet chemical etching and molding methods are complicated and time-consuming. Although the micro-hot-embossing method makes up for the shortcomings of the above two methods, the obligatory metal male mold is hard to obtain. If the chip structure changes, the mold must be redesigned and manufactured, which is not appropriate for scientific research. The two methods of paper-based microfluidic chip and 3D-printing are flexible and do not require molds. Unfortunately, due to substantive restrictions, the choice of transparent printing consumables is very limited. Fortunately, laser etching and hydrogel molding can process materials with better optical properties; however, they are difficult to process rectangular microchannels. In contrast, the micromilling technique can control the channel shape via changing the tool size and shape. In summary, the microstructure processing method proposed in this article has the advantages of simple steps, short processing time, and more materials to choose from.

In order to manufacture glass microstructures, the micromilling technique is proposed. The successful bonding of the glass chip with the microcolumn array structure and the PDMS chip is realized. In the hydrophilic experiment, through comparing the contact angles of the three arrays of microcone, microsphere, and microcolumn, it is determined that the microcolumn array (with a contact angle of 74.3° ± 2.4°) is the optimal structure for making microfluidic mixing chips. After that, the influence of the spacing, diameter, and depth of the microcolumn array on the contact angle is further investigated. The experimental results demonstrate that the increase in the array depth will make the contact angle smaller, and the increase in the microcolumn diameter and array spacing will make the contact angle larger. After increasing the depth of the microcolumn array from 50 to 200 µm, correspondingly, the contact angle decreased from 79.9° ± 2.4° to 37.9° ± 1.4°; as the spacing of the microcolumn array increased from 250 to 400 µm, the contact angle increased from 48.2° ± 1.8° to 66.2° ± 2.4°; as the diameter of the microcolumn increased from 250 to 400 µm, the contact angle increased from 74.6° ± 1.3° to 83.9° ± 1.4°. In order to verify the feasibility of the mixing microcolumn chip made via this method, a mixing experiment is carried out. Experiments indicate that the mixing index increases by 0.027 after the solution flows through the microcolumn array. This proves that the mixing microcolumn chip is available. Finally, a comparative analysis with other process methods is performed. In summary, this method has the advantages of high efficiency and flexibility and can process three-dimensional microstructures on a variety of materials.

This project was supported by the National Natural Science Foundation of China (Grant No. 51505077), the Project Agreement for Science and Technology Development of Jilin Province (Grant No. JJKH20200105KJ), and the Science and Technology Innovation Development Project of Jilin City (Grant Nos. 201750230, 20166013, and 20166012).

The authors declare no conflict of interest.

The data that support the findings of this study are available within the article.