This is an example of multiple zone-plates placed over individual microfluidic channels. Credit: Laboratory of Ken Crozier, Harvard School of Engineering and Applied Sciences.
With a silicone rubber “stick-on” sheet containing dozens of miniature, powerful lenses, engineers at Harvard are one step closer to putting the capacity of a large laboratory into a micro-sized package.
Microfluidics, the ability to manipulate tiny volumes of liquid, is at the heart of many lab-on-a-chip devices. Such platforms can automatically mix and filter chemicals, making them ideal for disease detection and environmental sensing.
The performance of these devices, however, is typically inferior to larger scale laboratory equipment. While lab-on-a-chip systems can deliver and manipulate millions of liquid drops, there is not an equally scalable and efficient way to detect the activity, such as biological reactions, within the drops.
The Harvard team’s zone-plate array optical detection system, described in an article appearing in Lab on a Chip (Issue 5, 2010), may offer a solution. The array, which integrates directly into a massively parallel microfluidic device, can analyze nearly 200,000 droplets per second; is scalable and reusable; and can be readily customized
Unlike a typical optical detection system that uses a microscope objective lens to scan a single laser spot over a microfluidic channel, the team’s zone-plate array is designed to detect light from multiple channels simultaneously. In their demonstration, a 62 element zone-plate array measured a fluorescence signal from drops traveling down 62 channels of a highly parallel microfluidic device.
The device works by creating a focused excitation spot inside each channel in the array and then collects the resulting fluorescence emission from water drops traveling through the channels, literally taking stop-motion pictures of the drops as they pass.
“Water drops flow through each channel of the device at a rate of several thousand per second,” explains lead author Ethan Schonbrun, a graduate student at SEAS. “Each channel is monitored by a single zone plate that both excites and collects fluorescence from the high speed drops. By using large arrays of microfluidic channels and zone plate lenses, we can speed up microfluidic measurements.”
The series of images are then recorded by a digital semiconductor (CMOS) camera, allowing high speed observation of all the channels simultaneously. Moreover, the array is designed so that each zone plate collects fluorescence from a well-defined region of the channel, thereby avoiding cross talk between adjacent channels. The end result is a movie of the droplets dancing through the channels.
“Our approach allows us to make measurements over a comparatively large area over the chip. Most microscopes have a relatively limited view and cannot see how the whole system is working. With our device, we can place lenses wherever we want to make a measurement,” adds Crozier.
The system can detect nearly 200,000 drops per second, or about four times the existing state-of-the-art detection systems. Further, the lens array is scalable, without any loss in efficiency, and can be peeled on-and-off like a reusable sticker. Ultimately, the integrated design offers the sensitivity of a larger confocal microscope and the ability to measure over a larger area, all in a much smaller, cheaper package.
“Because we have this massively parallel approach—effectively like 62 microscopes—we can get very high measurement or data rates,” says Crozier. “This device has shown we can measure up to 200,000 drops per second, but I think we can push it even further.”
Nanophotonics experts Schonbrun and Crozier originally developed the zone-plate technology to enhance optical tweezers so they could grab particles in a liquid using light. Using the high numerical aperture that makes efficient optical tweezers, they realized that arrays of zone plates could also be used to implement an efficient and scalable optical detection platform.
The researchers, who have filed a patent on their invention, are optimistic that with further research and development, the device could enhance a range of microfluidic and microfluidic-based lab-on-chip devices and speed the advance of using them for applications such as in-the-field biological assays.
The chips work like coin sorters, only they are much, much smaller. Liquids flow until they hit a wall where big particles get stuck and small particles pass through a super-thin slot at the bottom. Each chip’s slot is set a little smaller than the size of the particle to be detected. After the particles get trapped against the wall, they form a line visible with a special camera.
“One of the goals in the ‘lab on a chip’ community is to try to measure down to single particles flowing through a tube or a channel,” said Hawkins, who is also writing a book about aspects of lab-on-a-chip development.
Capturing single particles has important applications besides simply knowing if a particular virus or protein is present.
“One of the things I hope to see is for these chips to become a tool for virus purification,” said David Belnap, an assistant professor of chemistry and co-author on the paper.
He explained that a tool like the BYU chip would advance the pace of his research, allowing him and other researchers to consistently obtain pure samples essential for close inspection of viruses.
Overcoming obstacles to make the chips
A huge barrier to making chips that can detect viruses is $100 million – that’s the cost of machinery precise enough to make chips with nano-sized parts necessary for medical and biological applications.
The BYU group developed an innovative solution. First they used a simpler machine to form two dimensions in micrometers — 1,000 times larger than a nanometer. They formed the third dimension by placing a 50 nanometer-thin layer of metal onto the chip, then topping that with glass deposited by gasses. Finally they used an acid to wash away the thin metal, leaving the narrow gap in the glass as a virus trap.
So far, the chips have one slot size. Hawkins says his team will make chips soon with progressively smaller slots, allowing a single channel to screen for particles of multiple sizes. Someone “reading” such a chip would easily be able to determine which proteins or viruses are present based on which walls have particles stacked against them.
After perfecting the chips’ capabilities, the next step, Hawkins says, is to engineer an easy-to-use way for a lab technician to introduce the test sample into the chip