Ultrasonic microdroplet generators are useful devices with broad applications ranging from aerosolized drug delivery to three-dimensional (3D) printing-based additive manufacturing. One such technology comprises a microfabricated array of nozzles with droplet production driven by a piezoelectric transducer. The present study focuses on refining a critical fabrication step, anisotropic wet etching of pyramidal nozzles using a basic potassium hydroxide (KOH) solution. Given the integral role of nozzle geometry in device operation, high-precision techniques including Reactive Ion Etching (RIE), Deep Reactive Ion Etching (DRIE), and KOH wet etching were employed. A tapering geometry is preferred for acoustic wave focusing and efficient droplet generation, and KOH etching naturally yields pyramids due to preferential removal of the (100) plane versus the (111) plane of single-crystal silicon. Though wet etching is less precise than dry etching, it is difficult to form these 3D shapes using dry etching alone. Thus, this work focused on realizing the highest possible level of control over KOH etching. Challenges were encountered in using a conventional etching setup to achieve uniform vietching and good surface smoothness, which are crucial to definition of the pyramidal tip geometry. These aspects are also important for use of KOH etching to define microstructures in a range of microfluidic systems. Here, we introduce an ultrasonicassisted etching method to enhance the KOH etching process, addressing issues like nonuniform etching rates and surface roughness. This research not only provides insight into the microfabrication of ultrasonic microdroplet generators but also contributes to further improvements in microfluidic device manufacturing.
1.Introduction
Microfluidic devices with controllable liquid dispensing, specifically those that can produce uniform droplets, have great potential in many technological applications, such as traditional manufacturing, the chemical industry, biology/biomedicine , thermal management, and three-dimensional (3-D) printing . The controlled generation of droplets, in terms of size, size distribution, frequency, velocity, and composition, is foundational for applications ranging from high-throughput screening in drug discovery to advanced materials synthesis. Within this context, the acoustic microfluidic droplet generator stands out for its unique droplet formation mechanism.
2.Chip Fabrication
For basic control of droplet size and ejection mode (continuous jet or discrete droplet), we made nozzle microarrays with three different orifice diameters: 10, 20, and 40 µm. We began with a double-side polished silicon substrate [(100)-oriented, 100-mm diameter, 500- µm thick; University Wafer] precoated on both sides with a silicon nitride (𝑆𝑖3𝑁4) masking layer [~ 650-nm thick, low pressure chemical vapor deposited (LPCVD); Georgia Institute of Technology Institute for Electronics and Nanotechnology].
As illustrated in Figure1, the blue layer represents silicon nitride (𝑆𝑖3𝑁4), the gray layer signifies silicon, and the red layer denotes photoresist.
Figure 1. Illustration of fabrication procedure.
To summarize, the project used (i) standard photolithography to create the design patterns, (ii) reactive ion etching (RIE) to create square windows in the nitride film, and (iii) anisotropic potassium hydroxide (KOH) wet etching of the silicon to form pyramids. For the orifices at the tip of the nozzles, we relied on back side alignment and deep reactive ion etching (DRIE) of any remaining silicon at the nozzle tips to form circular orifices.
3.Ultrasonic Wet Etching
During the KOH etching with a magnetic bar, gas bubbles generated from the surface of features. The reaction between KOH and silicon produces silicate and hydrogen gas. The hydrogen gas forms bubbles which adhere to the surface of the silicon wafer. The adhesion of gas bubbles will lead to a varying etching rate at different areas of the silicon wafer.
According to the review by Pal et al., employing ultrasonic agitation during KOH etching of silicon notably enhances the etch rate while simultaneously improving the surface morphology. Moreover, the vibrations produced by the ultrasonic waves are beneficial in enhancing the quality of the etched surface. The use of ultrasonic energy facilitates the removal of adhered hydrogen gas bubbles, thus improving the contact between the etching surface and the KOH solution.
4.Test Wafer Design
The microfeatures were designed based on a 4-inch wafer, with each color in Figure 2 (a) denoting a distinct square array size. The size of squares in each nozzle arrays were range from 615μm to 665μm. The size of the features was chosen to examine the etch rate of the upper layer of the nozzle. The linear arrangement of square arrays shown in Figure 2(c) represents sizes that range from 590μm to 640μm, alignment features in the original design, with each successive array increasing by an increment of 5μm.
Figure 2(a) The CAD design of the testing features. (b) Dimension of microarrays from 615μm to 665μm. (c) Detailed view of side line squares from 590μm to 640μm with increments of 5μm.
During the fabrication process, the use of primer was found to affect the photoresist spin coated on the silicon nitride layer, inadvertently leading to overexposure during the laser writing. To solve this issue, a dose test was executed using the calibration matrix function on the Heidelberg laser writer operation program. The original settings for the photoresist S1827 were: Focus at 0%, Intensity at 100%, Laser Power at 70mW, and Filter at 100%. Post-dose testing, the microscope observations indicated that the best result was made with the following adjusted parameters: Focus at -25%, Intensity at 60%, while maintaining the Laser Power at 70mW and the Filter at 100%.
5.Wet Etching
KOH etching inherently produces pyramidal structures due to the preferential removal of the (100) plane over the (111) plane in single-crystal silicon. However, under varying concentrations, temperatures, and other conditions, the etching rates of the (100) and (111) planes and the structural changes induced by etching cannot be determined through simple calculations without performing experimental characterization of the etch process. To calibrate the KOH etching of silicon wafers coated with silicon nitride, three wafers mentioned in section 2.2 were used in this study. This was done to negate any potential effects of silicon thickness and the silicon nitride mask on the test results.
For each circular nozzle array corresponding to a nozzle size (i.e., discrete base sizes of 615μm, 625μm, 635μm, 645μm, 655μm, and 665μm), two of four arrays were selected for measurement, resulting in a total of 12 circular nozzle arrays being evaluated, as shown in Figure3(a) For each circular array, several fixed squares (nozzles) were selected for measurement to ensure the reliability of the data. In the side line, two of each size were also picked for measurement, as labeled in Figure3(b).
Figure 3 (a) Detailed view of picked square for each measured microarray. (b) Detailed view of picked square for measurement.
Figure 4provides a detailed view of the various parameters that require measurement. The 'Outer Length' (O) is size of square on the silicon nitride layer, while the 'Inner Length' (I) denotes the length of bottom area of the etched feature. The angle of the side wall from (100) surface to the (111) plane was calculated by equation.
Figure4. Illustration of measured parameters of the pyramid/ trapezoid during the KOH etching
6.Conclusions
To conclude this thesis, the exploration of ultrasonic KOH etching processes has yielded significant insights, particularly into the temperature-dependent nuances of the etching behavior. The research was primarily focused on understanding the etching dynamics at various temperatures, though the peculiarities observed at 65°C remain to be explained. Notably, the unique faceting and ripple-pattern etching issues observed at this temperature presented an intriguing anomaly, prompting a deeper investigation into the causes behind such irregularities.
Significantly, when devoid of the peculiar faceting and ripple-like features, the 65°C KOH etching process demonstrated superior performance. This finding is crucial, as it suggests that under optimal conditions, the 65°C etching process could potentially offer the most efficient and precise outcomes. This aspect of the research has substantial implications for the field of microfluidics and related disciplines, particular for our fabrication of nozzle microarrays, where the precision of the etching process is paramount. Understanding and harnessing the optimal conditions for KOH etching at this temperature could lead to advancements in fabrication techniques essential for developing high-precision microfluidic devices.
下一篇： 2D MXene 电化学晶体管
