Integrated Microfluidic and Optical Sorting Device

From September 2008 to March 2009, I teamed up with two high school classmates (Kim Hoang Vuong and Terry Zhi Hao Gani) in a science research project to compete in the Singapore Science and Engineering Fair (SSEF).

SSEF is a national competition organised by Singapore’s Ministry of Education (MOE), Singapore’s Agency for Science, Technology and Research (A*STAR) and Science Center Singapore. It is affiliated to the prestigious Intel International Science and Engineering Fair (Intel ISEF), which is regarded as the Olympics of science competitions. Top winners in SSEF will represent Singapore to compete at the Intel ISEF.

Our project, entitled Integrated Microfluidic and Optical Sorting Device, won a Bronze Medal at SSEF 2009 and we co-authored a paper in 2010 with our supervisor in the journal Biomicrofluidics. The work was done at the Center for Ion Beam Application at the National University of Singapore.

Our work

Video Demonstration

Experiment setup

Our team recorded the video below to show how yeast cells of different sizes respond to an optical force. The fluid flow speed is fairly slow for ease of viewing. The videos shows a “lab-on-a-chip” system consisting of two coupled “channels” of fluid flows. The channels are perpendicular to each other but are at different vertical heights. The horizontal channel is above the vertical one.

The line laser is roughly horizontal and stretching across the middle of the “sorting box” (where the two channels intersect). However, this laser is not visible because its wavelength (1064 nm) is in the infrared range. The laser shines on the sorting box from below.

There are two directions of fluid flows in this video: one “upward” flow in the lower channel and one “leftward” flow in the upper channel. Yeast cells are introduced into the lower channel (outside the field of view).

Please note that because the two fluid flows are perfectly level with the ground, the cells flowing “upward” in the lower channel are NOT actually flowing against gravity.

The focus of the camera is on the lower channel. Because of the extremely close working distance of the imaging lens (about a few millimeters above the sorting box), the depth of focus is extremely “shallow.” That is, anything that is not inside the lower channel will appear out of focus.

You can observe that (mostly large) cells respond strongly to the optical force and are pushed to the upper channel (thus going out of focus) and they follow the flow to the left. Small cells tend to continue on the upward flow in the lower channel.

This video was recorded through a Nikon TE2000-U optical microscope.

Nanofabrication

Nanofabrication

Our team of three high school students fabricated most of these “lab-on-a-chip” systems by ourselves. We first obtain a master silicon wafer from our supervisor. These wafers are made by the technique of proton-beam writing using a beam line from the state-of-the-art 3.5-million volt accelerator at CIBA.

We then mix liquid PDMS and a curing agent together in vacuum to ensure there are no air bubbles inside the chip. We then heat this mixture (with a wafer as the mould) in an oven until hardening. This process creates two etched channels (whose width is on the order of tens of microns), one on each half of the chip.

The two halves are then “bonded” to each other to form a single chip. Before bonding occurs, however, these two halves must be surface-cleaned, first by sonication in ethanol and then by exposure to plasma inside a Harrick plasma cleaner. This bonding ensures that the two halves will not come apart under pressure from the pump that introduces the fluid flows.

Finally the chip is bonded to a glass slide for ease of handling and placement in a microscope.

We did extensive testing to determine the vacuum pressure and plasma exposure time for the strongest bonding. The thickness of the chip is also an important consideration. Too thin chips will not have sufficient mechanical strength; too thick chips will degrade the video recording quality (which we rely on for efficiency analysis) and hamper the working of the imaging lens (which is just a few millimeters from the surface of the chip).

Fluid Flow

We introduce yeast cells of different sizes, suspended in de-ionized water, into the chip by a kdScientific KDS250 automatic syringe pump. The pump injection speed must be adjusted to create the right fluid flow speed in the chip because flow speed is a very important determinant of the sorting efficiency, as pointed out in our paper. The fluid flow speed in the chip must be low enough be laminar but fast enough to ensure high throughput. Furthermore, an injection speed that is too high from the pump will cause very large pressure inside the chip and break the chip.

We had to choose the aqueous solution with care because the wrong solution can absorb too much laser energy and heat up the chip to the point of destruction. We settle on de-ionized water after experimenting with other common solvents, such as ethanol, isopropanol and acetone.

Data Analysis

We record videos of all the experiments and use the freeware VirtualDub to break the videos into static frames. We load these videos into IDL and manually trace the trajectory of every single cell as they travel through the sorting box. This analysis tells us what fraction of cells and of what size are deflected by the laser beam at a particular flow speed.

The image below shows typical trajectories of the cells. Figure (a) shows what happens in the absence of the laser. The vast majority of cells do not change direction. Figure (b) and (c) show the trajectories of small and large cells, respectively. As expected, fewer small cells than large cells are deflected by the line laser.

Particle trajectories

More Details

To learn even more details about our project, you can view the paper we submitted to the SSEF referees, our poster presentation or the published journal paper.

Media Coverage

Singapore Science and Engineering Fair Award acceptance

SSEF 2009 award ceremony. Fromt left: Mr. Heng Chee How (Singapore’s minister of state), Kim Hoang Vuong (team member), Huy Nguyen (team member) and Terry Zhi Hao Gani (team member).

Where They Are Now

It was my pleasure to work alongside very dedicated teammates in this project.

Kim graduated with Distinction from high school and graduated with a bachelor’s degree in quantitative economics and finance from Singapore Management University in 2014. He’s currently the Vietnam Business Manager for YOOSE, a Singapore-based advertising company.

Terry won a gold medal at the International Chemistry Olympiad 2009 (at the astounding young age of 15!) and graduated with High Distinction from high school. He received his bachelor’s degree in chemical engineering from the National University of Singapore in 2013 at the age of 19, the youngest ever from that department. He is now a PhD student in chemical engineering at MIT.