Jeffrey Perkel has been a scientific writer and editor since 2000. He holds a PhD in Cell and Molecular Biology from the University of Pennsylvania, and did postdoctoral work at the University of Pennsylvania and at Harvard Medical School.
Microfluidics isn’t a tool so much as a technology. Yet its potential increasingly is being felt throughout the life sciences, from single-cell manipulation and cell-motility studies to DNA sequencing and diagnostics.
Microfluidics, or the science of moving tiny liquid volumes through microscale channels, offers benefits in reagent use and waste generation, reproducibility, automation and ease of use. Yet that doesn’t make it easy. Microfluidics requires a certain level of expertise, explains Mark Burns, chemical engineering chair at the University of Michigan, who designs microfluidic tools.
“It’s the real divider for people in the field,” Burns says. “Either you know how to make the systems, or you don’t.”
Experts like Burns, of course, have the resources and know-how to design and build their own microfluidic circuits. Non-experts can buy off-the-shelf tools from companies such as Fluidigm and Revvity, assuming their application has been commercialized.
For bespoke chips, however—say, to improve on a design they’ve seen in the literature or simply to determine if microfluidics can solve a particular laboratory problem—non-experts are at a distinct disadvantage. Fortunately, there are tools and resources available to help.
The problem of microfluidics
In principle, microfluidics isn’t all that complicated. Simply design a circuit to solve a problem, fabricate a chip, and go. In practice, there’s a lot of skill and knowledge involved in carrying out those steps.
“Say you decided you wanted to do something very, very simple: Take two liquids, flow them through a channel, mix them and then measure or observe something from that event. That’s a very simple system,” says Burns.
First though, you need to answer several key questions. First, what material(s) should you use to make it? A rubbery material called PDMS is a popular and convenient choice, but it is incompatible with certain organic solvents and allows for gas exchange and evaporation. Glass and plastic provide greater rigidity and optical clarity, but they also are costlier to produce.
Next, you must design your device. Brian Johnson, a research engineer at the University of Michigan who works with Burns, says this process typically involves drawing a circuit diagram in a 2D CAD program, though any drawing software would probably work as long as it can render features with the appropriate precision and at the correct scale. (Johnson uses L-Edit from Tanner Research, but DraftSight® is a free alternative.)
That design is then rendered as a “transparency mask”—like a negative of the desired circuitry—which costs about $150, he says. This is used to build the chip features in the mold, which requires coating a silicon wafer with a photolithographic substrate such as SU-8 “photoresist,” and possibly the use of a cleanroom.
Finally, you’re ready to cast the PDMS onto the mold, let it cure and peel off the completed chip. Final assembly may include coring holes to access channels and bonding to another substrate or other PDMS piece to complete the structure, which may or may not work as desired in any event. “Like anything in science, it is iterative and may take multiple designs to make it work,” says Shell Ip, new market and technology consultant at FlowJEM, a custom microfluidics manufacturer.
All told, says Burns, “it could cost tens or hundreds of thousands of dollars plus weeks to months to get that first device made.” (Johnson explains the process of PDMS chip fabrication in an online webinar, which is available here.)
Microfluidics starter sets
To mitigate that cost, Burns’ lab has developed LEGO block-like PDMS building blocks researchers can use to prototype microfluidic designs and see if they make sense for their particular problem.
The set includes channels and intersections, reactors and valves plus channels of varying length. “We made the blocks to have essentially all the components to make what you would want in microfluidics research,” Burns explains.
Though the blocks are not commercialized, Burns says he’s “happy to work with anybody who contacts me”—if he has the bandwidth. “We do not have a large supply [of building blocks], and we aren’t making them, but if there was demand that would give me incentive to do that.”
Trianja Technologies has commercialized a similar tool made out of glass. The company’s “discovery fluidics” line includes more than 50 prefabricated single-function chips that perform such tasks as electrophoresis, mixing and separation. “We designed these with the idea to lower the barrier for scientists and provide a bunch of things they can test drive,” explains Michelle Lyles, president of commercial operations at Trianja.
Researchers who want to prototype a design can purchase chips that perform the various functions required—each costs $110, Lyles says—and test them one by one. “Once they do that, then we can go in and fine-tune the chip for their exact application.”
The difficulty new users encounter in the microfluidics space, Lyles says, is there are few if any standards—no established designs for key building blocks. As a result, almost every project requires the microfluidic equivalent of reinventing the wheel. “Almost every application that’s being attempted, requires some sort of prototyping, which in turn requires some iterative learning process,” she says.
Getting started
By all accounts, the best place to kick off a microfluidics project is with a literature search. “If you come up with an idea, chances are, somebody has already done it,” says Ip. Plus, “the literature can at least tell you what a device that does some specific task will look like.”
Ip notes, for instance, that neophytes often have trouble with the fundamentals of chip design, such as the importance of aspect ratio (the ratio of channel width to height), feature density and feature size.
Next, talk to your colleagues. Microfluidics is so common these days it’s a good bet your university or institute has a microfluidics expert on staff, but if not, you should be able to find an extramural expert with little difficulty. Between the literature and colleagues, Johnson says, “you’re going to find out 80% to 90% of what you need to know.”
Finally, try the Royal Society of Chemistry’s “Chips and Tips” blog. According to Ip, the blog (published by the same company that puts out the peer-reviewed journal Lab on a Chip) “is a place to publish stuff that maybe wouldn’t be publishable as a scientific discovery in a journal, but there are tips you can pick up that people want to share with the community”—different ways to plumb a device, for instance.
Such tools and resources may deprive newbies of some hard-won knowledge, Burns says, but should help them answer their most fundamental question: Whether microfluidics will advance their research in a way they otherwise couldn’t.