A process analysis for microchannel deformation and bonding strength by in-mold bonding of microfluidic chips

1 Introduction

The microfluidic chip, also known as a microfluidic lab on a chip, is a novel tool for basic biological and chemical analysis rapidly, by controlling fluids in microchannel networks on a chip of a few square centimeters in size, with which many key laboratory unit operations such as sample injection, separation, metering, mixing, reaction and biochemical detection can be performed [1, 2]. With the development of life sciences and analytical chemistry, most applications demand devices which can be used as disposables, in order to eliminate risks of sample contamination. However, expensive glass, quartz or silicon chips cannot meet this demand. Hence, it is of high importance to develop low-cost fabrication strategies for the mass production of microfluidic chips [3]. Polymer chips and their fabrication methods have been proven to be the solution to this challenge. This trend towards the use of polymers as chip material is furthermore supported by the fact that many standard laboratory items focus on the microfabrication of polymer chips [4].

The fabrication process of a polymer microfluidic chip consists mainly of molding, cleaning, alignment and bonding; bonding is the most critical and time-consuming process [5]. To date, a series of techniques have been developed for bonding microfluidic chips, including thermal bonding, solvent bonding, glue bonding, microwave bonding, etc. Thermal bonding is the most commonly used approach because it allows the formation of microchannels with uniform surfaces composed entirely of the same polymeric materials. Research on microfluidic chip bonding technology has been focused on thermal bonding. One major challenge of thermal bonding is channel deformation caused by un-optimized temperature, pressure and time. Thus, properly controlling temperature, pressure and time is critical in achieving high bond strength while limiting deformation of the microchannels due to bulk polymer flow.

Thermal bonding can be analyzed using the diffusion theory, where macromolecular chain segments may move across the interface and establish entanglements on the opposite surface [6]. The bonding strength depends on the depth of interdiffusion of polymer molecules across the interface, which is a function of the healing temperature, pressure and time. The penetration depth is proportional to the healing strength. Bonding temperature has a significant influence on the mobility of the macromolecules and, therefore, on the interdiffusion process and bonding strength. Higher pressure enables better physical contact between the two mating surfaces, which enhances interdiffusion across the mating interface, resulting in higher chain entanglement and bond strength. Enough time is a necessary condition for polymer interdiffusion across the mating interface.

On methods of thermal bonding, many researchers [7–9] used an electric resistance furnace to heat the chip, and employed a crab-shaped clip, weight or spring press to apply the bonding pressure during thermal bonding. However, the bonding pressure cannot be well controlled. Some researchers [10–14] used a hot press machine or homemade experimental apparatus to bond the substrate and cover sheet. By these methods, bonding pressure can be controlled, but the bonding time increases owing to the slow preheating process. Overall, thermal bonding is isolated from other fabrication processes of polymer microfluidic chips, and makes the production cycle time longer. With the continuous development of life science technology, small batch microfluidic chip production in the laboratory cannot satisfy the growing market. The fabrication technology of the microfluidic chip is developing towards the high efficiency, low cost and large scale direction. Therefore, it is imperative to seek a highly integrated fabrication technology to realize the shift from laboratory to market in microfluidic chip application.

In this paper, an in-mold bonding technology, which integrated injection molding, alignment and bonding of chips, was developed to reach the goal of short cycle, low-cost and mass production of polymer microfluidic chips, based on technologies of thermal bonding and in-mold assembly. In-mold bonding experiments of the microfluidic chip were carried out to decrease the microchannels deformation, raise the bonding strength and reduce the production cycle of the chip.

2 Materials and methods

2.1 Setup of in-mold bonding device

2.1.1 Design of the microfluidic chip

A typical cross-channel electrophoresis chip was designed for the study of microfluidic chips with high efficiency, low cost and mass production. To remove the process of alignment and reduce cycle time, a design scheme which integrates both the microchannels and sample pools in the substrate was adopted to decrease alignment accuracy requirements between the substrate and cover sheet. The dimensions of microchannel cross-section were designed as 100×40 μm². The structure and dimensions of the substrate and the cover sheet are given in Figure 1.

Figure 1

The structure and dimensions of the cross-channel microfluidics chip (unit, mm): (A) the substrate (0.8 mm in thickness); (B) the cover sheet (0.6 mm in thickness).

2.1.2 Equipment of microfluidic chip in-mold bonding

The traditional fabrication method of microfluidic chip includes molding of the chip substrate and cover sheet, cooling, cleaning, alignment, heating, thermal bonding, and cooling etc., sometimes also including process of surface modification [15, 16]. The repetitive heating and cooling of the chip leads to, not only a waste of energy, but also an increase in production cycle and impair performance of chips.

Based on the points above, a novel injection molding tool was designed which should integrate injection molding, alignment and thermal bonding of polymer microfluidic chips, which could reduce the production cycle. The specific motion requirements of mold are given in Figure 2.

Figure 2

The specific mold operations.

In accordance with the motion requirements given in Figure 2, an in-mold bonding device was designed and fabricated, as shown in Figure 3A. Firstly, the substrate and cover sheet are molded simultaneously in the first mold close. The mold is a family mold, as shown in Figure 3B. At first mold open, the gate will be cut automatically to remove sprue by the ejection unit of the injection molding machine, so as not to interfere with the second mold close. The substrate and cover sheet are reserved in the respective cavities. Then, the plate “B” with the cover sheet is drawn for the alignment between the substrate and cover sheet by part 1. After the second mold close, the mold is heated for 2 min to allow the plates to reach the preset bonding temperature. The bonding pressure is provided by part 2. In-mold bonding was achieved at the required pressure for a fixed duration, as shown in Figure 3C. Finally, the bonded chip is demolded by part 3. Parts 1–3 are three hydraulic cylinders as shown in Figure 3B. The workflow of in-mold bonding is given in Figure 4.

Figure 3

The construction of in-mold bonding device: (A) the in-mold bonding device; (B) injection molding the substrate and cover sheet; and (C) in-mold bonding the substrate and cover sheet.

Figure 4

The workflow of in-mold bonding.

2.2 Experimental setup

2.2.1 Experimental material and equipment

In this study, the thermoplastic material polymethylmethacrylate (PMMA) (CM-205) was produced by Chimei Chemicals (Taiwan). The experiments were carried out on a horizontal injection molding machine (Arburg 370s500–100), which was used to perform microfluidic chip injection molding and in-mold bonding. The mold insert was made from electroplated nickel, shown in Figure 5. The surface morphology of the mold insert was measured using a Wyko NT9100 optical profilometer (Veeco Instruments, Inc.), as shown in Figure 6. The surface roughness average, Ra, is about 37 nm, which could have little influence on the quality of bonded chips.

Figure 5

The structure of the nickel mold insert.

Figure 6

The surface roughness of the mold insert.

2.2.2 In-mold bonding of microfluidic chips

Substrates and cover sheets were injection molded under a mold temperature of 90°C, injection rate of 297 mm/s, melt temperature of 250°C, packing time of 1.8 s, packing pressure of 100 MPa and cooling time of 20 s. Table 1 lists the processing conditions of in-mold bonding used in this study according to preliminary results of the simulation analysis [17–19], which researched the influence of bonding parameters on microchannels deformation, and showed that the small deformation could be obtained with a low pressure and short time when the bonding temperature was in the glass transition temperature vicinity of PMMA. When one processing parameter was investigated and varied, the other main processing parameters were kept at the default values. The default values are as follows: a bonding temperature of 100°C, a bonding pressure of 2 MPa and a bonding time of 240 s. In the process of bonding, the bonding temperature was supplied by the mold. However, the temperature at the bonding interface could not be measured. In this paper, the bonding temperature was approximately equal to the mold temperature, which was measured by a thermocouple embedded in the mold. The location for measuring temperature is shown in Figure 3A. The bonding pressure is the hydraulic pressure that is controlled by the injection molding machine (Arburg 370s500–100). The microchannel morphology was observed and measured on a Rational VMS-1510A optical image measuring instrument. The bonding strength of microfluidic chips was measured on SANS CMT4204 electronic universal testing machine, as shown in Figure 7. The experimental data were analyzed using least squares as the fitting procedure. The fitted curves are used as an aid for data visualization, to summarize the relationships among variables.

Table 1

Processing conditions used in this study.

Processing conditions	Units	Set no.
Factors under investigation	–	1–5	6–10	11–15
Bonding temperature	°C	90, 95, 100, 105, 110	100	100
Bonding pressure	MPa	2	1, 1.5, 2, 2.5, 3	2
Bonding time	s	240	240	120, 180, 240, 300, 360

Figure 7

Bonding strength test.

3 Results and discussion

Microchannel morphology of the microfluidic chip is shown in Figure 8. A microchannel with top width of 91 μm, bottom width of 40 μm and height of 39 μm can be injection molded. The bonded microchannel under default processing parameters is shown in Figure 9, with top width of 89 μm, bottom width of 39 μm and height of 34 μm. The in-mold bonding technology was successfully adapted to fabricate microfluidic chips with satisfactory quality. The deformations in height and in top width are two quality indexes to judge the deformation of microchannels in this study.

Figure 8

Topography of a substrate microchannel by injection molding.

Figure 9

Microchannel morphology of bonded chip (100°C, 2 MPa, 240 s).

3.1 Influence of bonding process on the microchannel deformation

The influence of bonding temperature on the microchannel deformation is shown in Figure 10. The height and top width deformation of microchannels would increase with increase of the bonding temperature. The deformation of microchannels was very small with the temperature below 100°C. However, it increased when the temperature exceeded 105°C. When the bonding temperature reached 110°C, the deformation of height was 11 μm, with a top width deformation of 13 μm. Therefore, the bonding temperature should not be higher than 105°C, which is the glass transition temperature of PMMA. The bonding temperature should not be too low, otherwise it would be difficult for substrates and cover sheets to successfully bond.

Figure 10

Effect of bonding temperature on microchannels deformation (2 MPa, 240 s): (A) decrease in height; (B) decrease in top width.

The influence of bonding pressure on the microchannel deformation is presented in Figure 11. It was observed that the height and top width deformation of microchannels increased with increase of the bonding pressure, and that the deformation in height was larger. There was an approximately linear relationship between the deformation of microchannels and the bonding pressure. The deformation in height was 9 μm, and the top width deformation was 7 μm. With regards to the required bonding strength, it is better to choose a smaller bonding pressure.

Figure 11

Effect of bonding pressure on microchannels deformation (100°C, 240 s): (A) decrease in height; (B) decrease in top width.

The effect of bonding time on the microchannels deformation is described in Figure 12. The microchannels deformation in height and top width increased with the increasing bonding time. The deformation of microchannels was approximately proportional to the bonding time. When the bonding time increased from 120 s to 360 s, the increment of height deformation was 6 μm, and the increment of top width deformation was 5 μm. It is better to choose a smaller bonding time if the bonding strength of chips meets operation requirements.

Figure 12

Effect of bonding time on microchannels deformation (100°C, 2 MPa): (A) decrease in height; (B) decrease in top width.

3.2 Influence of bonding process on the bonding strength

The influence of bonding temperature on bonding strength is presented in Figure 13. It was observed that the bond strength increased with increase of bonding temperature. When the temperature was below 100°C, most PMMA molecules could not break through an energy barrier for the diffusion process to proceed, so the bonding strength was very weak with a value of only 35 kPa. When the temperature was higher than 100°C, the interdiffusion ability of PMMA molecules greatly increased, which improved the bonding strength of chips. The bonding strength improved with increase in the bonding temperature, and was 825 kPa when the bonding temperature increased to 110°C.

Figure 13

Effect of bonding temperature on the bonding strength (2 MPa, 240 s).

The effect of bonding pressure on bonding strength is shown in Figure 14. When the pressure was below 2 MPa, the bonding strength was 68 kPa. It appeared some no-bonding areas, which could cause leakage of microchannel. When the pressure was 2 MPa, the bonding strength increased sharply to 434 kPa. When the bonding pressure was low, substrates and cover sheets were not closely contacted, which suppressed interdiffusion across the mating interface. This may due to warpage of the substrate and cover sheet, brought from the injection molding process. With the increase of bonding pressure, the substrates and cover sheets could be fully contacted, resulting in a high bonding strength.

Figure 14

Effect of bonding pressure on the bonding strength (100°C, 240 s).

Figure 15 shows the influence of bonding time on bonding strength. The bonding strength improved with increase of the bonding time. When the bonding time was 120 s, the bonding strength reached 340 kPa. The bonding strength increased to 614 kPa with a bonding time of 360 s. Under a bonding temperature of 100°C and a bonding pressure of 2 MPa, if the bonding time was lower than 240 s, there would be a no-bonding region in the chip. This region would greatly affect the performance of the final chips.

Figure 15

Effect of bonding time on the bonding strength (100°C, 2 MPa).

4 Process parameter setting

With the goal of smaller deformation, higher bonding strength and better sealing, the process of in-mold bonding is reasonably selected, according to the research above. It is reported that the chips with a bonding strength higher than 65 kPa can reach their functional targets [20]. The goal of bonding strength is set to be higher than 65 kPa. From experiments, the settings of the process parameters are: a bonding temperature of 102°C, a bonding pressure of 1.8 MPa and a bonding time of 240 s. With these parameters, the bonding strength of chip reached 350 kPa. The cover sheet and substrate were completely bonded that the chip microchannels height was 36 μm, the top width was 85 μm, and the bottom width was 38 μm. The maximum deformation of bonded microchannels did not exceed 10% of the microchannel sectional area before bonding, as shown in Figure 16.

Figure 16

Microchannel cross section morphology (102°C, 1.8 MPa, 240 s).

5 Conclusion

In this paper, in-mold bonding technology was advocated and developed for the efficient manufacture of microfluidic chips. This technology integrates thermal bonding into an injection mold to reduce the production cycle time and to simplify the manufacturing process. The influences of the bonding process parameters on the deformation and chip bonding strength of chip microchannels were studied with an in-mold bonding device. Some conclusions according to our experimental results can be drawn as follows:

1. The microchannels deformation would be increased with increase of bonding temperature, bonding pressure and bonding time. The microchannels deformation is most sensitive to the temperature. When the bonding temperature is below 100°C, the microchannels deformation of chips is very small. The deformation becomes large when the bonding temperature exceeds 105°C. The bonding pressure and bonding time have a large impact on the height deformation and a relatively small impact on the deformation of width.

2. The bonding temperature and bonding pressure have big impacts on bonding strength. However, only reaching a certain critical value, the bonding strength can increase sharply, up to 400 KPa. The bonding time has a small impact on bonding strength. However, the bonding time should exceed 240 s to guarantee excellent sealing of the chip.

3. Considering the microchannels deformation and bonding strength, the bonding process parameters are set as: a bonding temperature of 102°C, a bonding pressure of 1.8 MPa and a bonding time of 240 s. With these parameters, the microchannels deformation is <10% and the chip bonding strength can reach 350 kPa. The production cycle of chips was reduced to <5 min.