PDMS: The Favorite Material of Microfluidics (for now)

Precise microfluidic channel in PDMS

Precise microfluidic channel in PDMS

Whether you’ve been learning about microfluidics here at Microfluidic Future or somewhere else, you’ve undoubtedly come across the elastomer poly(dimethylsiloxane) (PDMS). PDMS has radically changed the capabilities of microfluidics (and its price tag) since it was first brought into microfluidics by George Whitesides in 1998. PDMS has effectively replaced glass and silicon which were borrowed from existing micromachining industries. PDMS has great resolution and can contain sub-0.1 µm features. But how is PDMS used, and what makes it so great? Hopefully you’ll have these answers by the end of this post.

PDMS Fabrication

PDMS fabrication process, beginning with photolithography

PDMS fabrication process, beginning with photolithography

Soft lithography is a widely used technique to create PDMS structures. This process first requires a mold, or a master, generally produced by photolithography. The mold must represent the hollow space of the microfluidic chip, where the fluids will eventually flow. PDMS is formed by mixing one component which contains silicon hydride groups and another component which contains a vinyl group. This mixture is then poured on top of the master. PDMS cures to a solid after an hour at 70o C, and can then be peeled away from the master without damage to either part, allowing the master mold to be reused. The earlier steps can be completed within 24 hours, and replicas can be mass produced after designing and creating a master.

After removing the PDMS replica, we have an open space in the structure from the master, but our channels have no bottom! We next have to seal the PDMS to another surface, such as glass, silicon, or itself. More importantly, PDMS can be sealed reversibly or irreversibly.

Van der Waals forces allow reversible sealing to smooth surfaces. This seal is airtight, but can only withstand pressures around 5 psi. Adhesive silicone or cellophane tapes can also be used to create a reversibly seal that is stronger than PDMS alone. Reversible seals allow for reuse or added functionality. Remember that the PDMS used in the SIMBAS lab-on-a-chip was sealed reversibly to the glass, allowing reuse or future analysis.

In some cases, a lab-on-a-chip may require an irreversible seal, especially if operating under higher pressures. PDMS can be irreversibly sealed by exposing it (and possibly the other surface) to oxygen plasma. However, the two pieces of chip must be aligned within a minute or else the oxidized PDMS surface will reconstruct in the air.

Complex PDMS structure "sandwiching" a membrane between two layers

Complex PDMS structure "sandwiching" a membrane between two layers

Complex structures can be created by “sandwiching” multiple layers of PDMS. Furthermore, additional components like membranes can be included in this sandwich. This can be a bit of a challenge depending on the complexity of the features, and micro-stages have been created to ensure perfect alignment after oxidation. Alternatively, PDMS can be sealed with the aid of polar solvents. A thin film of a polar solvent is placed between PDMS layers and then heated, evaporating the solvent and sealing the layers.

Instead of using photolithography to create the mold, 3D printing can also be used. 3D printing prints small globules of material according to a CAD file to make 3D structures. 3D printing itself has been growing in capabilities due to its extremely quick prototyping speed. It can achieve a resolution between 50 and 100 µm, which is suitable for microfluidic devices. This mold would then be used in the same way to create PDMS replicas.

Microfluidic Components

Microfluidic chips can be very complex, featuring pumps, valves and mixers. These parts can easily be implemented in a PDMS structure. Mixers can be incorporated into the masters, transferring them to the floor of a channel. Other parts that are made separately, like some of the microvalves I discussed previously, can be placed in the PDMS while curing. This allows a single master to be very versatile while moving independent features about the device.

Key Features

Here are some key features of PDMS adapted from “Poly(dimethylsiloxane) as a Material for Fabricating Microfluidic Devices” by Whitesides.





transparent; UV cutoff, 240 nm

optical detection from 240 to 1100 nm


insulating; breakdown voltage,

2 x 107 V/m

allows embedded circuits; intentional breakdown to open connections


elastomeric; tunable Young's modulus, typical value of

~750 kPa

conforms to surfaces; allows actuation by reversible deformation; facilitates release from molds


insulating; thermal conductivity, 0.2 W/(m*K); coefficient of thermal expansion, 310 µm/(m*°C)

can be used to insulate heated solutions; does not allow dissipation of resistive heating from electrophoretic separation


low surface free energy

~20 erg/cm

replicas release easily from molds; can be reversibly sealed to materials


impermeable to liquid water; permeable to gases and nonpolar organic solvents

contains aqueous solutions in channels; allows gas transport through bulk material; incompatible with many organic solvents


inert; can be oxidized by exposure to plasma

unreactive toward most reagents; surface can be etched; can be modified to be hydrophilic and also reactive toward silanes; etching can alter topography of surfaces



can be implanted in vivo; supports mammalian cell growth

Some clear advantages and unique properties can be seen from this table. I would also like to point out some additional advantages and disadvantages.


  • Quicker production time than glass and silicon microfluidic chips
  • Cheaper than glass and silicon


  • Unless oxidized or treated, the surface of PDMS is hydrophobic which can encourage air bubbles to form in channels and protein adsorption
  • Elasticity restricts aspect ratio of structures as sagging/shrinking can occur.

 This is one part of my Microfluidics Beginer’s Guide. Check out the rest of it and keep learning!


ResearchBlogging.org McDonald, J., & Whitesides, G. (2002). Poly(dimethylsiloxane) as a Material for Fabricating Microfluidic Devices Accounts of Chemical Research, 35 (7), 491-499 DOI: 10.1021/ar010110q

The Microvalve: The Traffic Light of Microfluidics

You could say that valves in microfluidics (or microvalves) are like traffic lights that control flow along microfluidic channels. But I’d say that they’re more like police barricades, stopping anyone they want, wherever they want. The sole purpose of microvalves is to control flow within a microfluidics device, allowing them to become very complex and more automated. Without microvalves, all reactions and mixing must occur in the same space, unless they were premixed elsewhere, which might just eliminate the advantage of microfluidics.

Every microvalve that has ever existed seems to be covered in “A review of microvalves” by Kwang W Oh and Chong H Ahn from the Journal of Micromechanics and Microengineering. I’ll be covering three of the valves mentioned, but feel free to check out all the other microvalves that didn’t make the cut. A companion paper for this post, “Incorporation of prefabricated screw, pneumatic, and solenoid valves into microfluidic devices” by George M Whitesides et al. , featured in Lab on a Chip, describes a way to prefabricate these three valves so that they can easily be added to existing lab-on-a-chips.

Solenoid Microvalves

Prefabricated Solenoid Microvalve

Prefabricated Solenoid Microvalve

You guessed it, solenoid microvalves are based on…solenoids! A solenoid is a coil of wire wound in a helix, like a compressed spring. I won’t go into the physics behind it, but a magnetic field is created by passing current through the solenoid. If you put a metallic object within the coil and vary the current, you could move the object. This is the idea behind solenoid microvalves. A solenoid microvalve is simply a solenoid with an actuator inside of it. The actuator is situated above the channel of a microfluidic device and pushes down on the ceiling of the channel to collapse it and obstruct its flow. In this manner, the solenoid microvalve requires an elastomeric lab-on-a-chip, and is rather bulky.

Whitesides’ Prefab:

Whitesides’ prefabrication does not involve a dramatic redesign of solenoid microvalves. Instead, he places the same solenoid in PDMS housing, which can accommodate the solenoid’s large footprint.

Screw Microvalves

Screw Microvalve

Screw Microvalve

This is another one. You guessed it, microvalves are based on…screws! This is an extremely low tech microvalve that requires little more than a screw. The screw is incorporated into the microfluidic device and deflects the membrane of the channel by twisting it. A ball can be placed beneath the screw to prevent damage to the device. This design is great for lab-on-a-chips that are intended to be disposable or for a low-resource setting. It requires no power, has a small profile (although you need access to screw it manually) and is simple to use. Obviously, it’s not completely automated, but I suppose you could incorporate a small electric screwdriver somewhere.

Whitesides’ Prefab:

Whitesides’ prefabrication method doesn’t really differ too much from the conventional screw microvalve design. The screw is essentially embedded in a housing which can be joined above an existing microfluidic channel. One of the key advantages argued by Whitesides is that prefabrication regulates the screw microvalves. All the prefabricated screw valves are created in an identical manner, and would require the same degree of rotation to shut off flow. This is important because it may not be possible to see how well the flow is obstructed, resulting in leakage or damage to the channels. Screw valves created by different people at different times may be placed differently. A half-turn of one person’s valve might not properly close the channel, while a half-turn of someone else’s valve might go too far and damage the device.

Pneumatic Microvalves

Quake Microvalve ( source )

Quake Microvalve (source)

I know it says pneumatic, but these are commonly referred to as Quake Valves, named after Stephen Quake. Quake valves require additional channels, often perpendicular to the targeted microfluidic channels. The additional channels share a thin, common membrane with the targeted channels. When air at the right pressure is applied through the pneumatic channels, the shared membrane is deflected and obstructs the flow of fluid. This completely changed the field of microfluidics since its arrival in 2000, allowing such feats as 400 simultaneous PCR reactions.

As cool as these microvalves are, they certainly have their drawbacks. First, they require more planning because you not only need to incorporate an additional layer of pneumatic channels, but you also have to route all your channels so that they don’t overlap where you don’t want them to. A small change in the design of your microfluidic device could require a massive redesign. While the solenoid valve certainly has a large footprint over the lab-on-a-chip due to the large size of the solenoid, the Quake valves only require the inclusion of an additional layer, keeping the area directly around the device cleaner. But this setup requires a tank of pressurized gas near the device, so in reality its footprint isn’t so small. This obviously hinders a device from easily leaving the lab. The pneumatic valves can be controlled electronically, allowing a device consisting of multiple independent valves to become more automatic. For instance, Albert Folch of University of Washington created this video of a Microfluidic Ballet by controlling valves according to the frequency of the music by Dimitri Shostakovich.

Whitesides Prefab:

Prefabricated Quake Microvalve

Prefabricated Quake Microvalve

Whitesides’ prefabrication of Quake valves diverges from the most from the conventional out of all three microvalves shown here. Simply consisting of a chamber attached to an air supply placed over a channel, this method doesn’t require any additional layers of valves. A conventional Quake valve would require an entire redesign of the pneumatic channel layer, while a prefabricated Quake valve can simply be relocated over a new channel. When activated, the chamber presses down on the channel wall and obstructs the flow.

Final Thoughts

We’ve only examined three different microvalves here, and you can see how radically unique they are. As we develop new microfluidic applications and materials, we’ll continually develop novel ways to control the flow. A high concentration of valves indicates greater device design sophistication, but even so, less can be more, and our best solution might just be the simplest one.

This is one part of my Microfluidics Beginer’s Guide. Check out the rest of it and keep learning!





Oh, K., & Ahn, C. (2006). A review of microvalves Journal of Micromechanics and Microengineering, 16 (5) DOI: 10.1088/0960-1317/16/5/R01

Elizabeth Hulme, S., Shevkoplyas, S., & Whitesides, G. (2009). Incorporation of prefabricated screw, pneumatic, and solenoid valves into microfluidic devices Lab on a Chip, 9 (1) DOI: 10.1039/b809673b

Microfluidics? What's That? A Beginner's Guide

Author's note: This post was chosen as an Editor's Selection at ResearchBlogging.org. Thanks for the support!


Microfluidic chemostat used to study microbes

Microfluidic chemostat used to study microbes

I don’t quite have the resources to poll the United States and the rest of the world, but if I did, this is what I’d ask:

  1. Do you know what microfluidics is?
  2. Can you explain it to me?
  3. Do you currently use anything with this technology?

We may never know the results of the poll, but I think I'd hear "No" and "What is microfluidics?" Have no fear, because today you’re lucky enough to read my Beginner’s Guide to Microfluidics.

To start with, let’s look at the word itself. It is the word fluidics with the prefix micro- strapped onto the front. Thus, we’re looking at the manipulation of fluid at the micro scale, often flowing through channels. There isn't a strict definition, but some like to refer to flow with a channel dimension less than 1,000 µm as microfluidics. It's more complicated than that, so to guide us, I’ll be referencing heavily from a review article George M Whitesides wrote in 2006, “The origins and the future of microfluidics.” For those unfamiliar with Dr. Whitesides, he is a legend in many spheres including microfluidics, and if you are reading a paper on microfluidics, it most likely either cites him, or can be traced to something that does. In this particular article, he looks at four “parents” of microfluidics that have brought us to the technology we have today:

  1. Microanalytical methods
  2. Biodefense
  3. Molecular biology
  4. Microelectronics

Microfluidics’ Loving Parents


Microanalysis is the chemical analysis and identification of small amounts of matter, such as capillary electrophoresis. First developed in the 1960s, it is a method to separate ionic species using their size to charge ratio. Capillary electrophoresis is still used today, showing that microanalysis not only provided needs for microfluidics to meet, but it also provided some techniques.


Awareness of chemical and biological warfare escalated after the Cold War, fueling the United States’ Defense Advanced Research Projects Agency (DARPA) to heavily fund microfluidics research in the 1990s. Biodefense remains a major area or research in homeland defense as demonstrated by the recently unveiled microfluidics-based device that detects anthrax or other suspect DNA. The suitcase-sized system recovers cells from a sample and analyzes the DNA in less than an hour, while it currently takes 1 to 3 days in the lab. 

Molecular Biology

In the past few decades, genomics activity has progressed quickly and has demanded newer technology to grow even further. Microfluidics had the opportunity to process DNA faster, cheaper and more precisely. Stephen Quake, a professor at Stanford University, has been leading cutting-edge research in sequencing. The ability to sequence DNA in weeks rather than months has allowed him to start a company, Helicos, implementing microfluidic methods. There have been several microfluidic applications for sequencing as researchers and companies race to sequence a human genome for only $1,000.


While the three previous “parents” have provided some tools and needs for microfluidics, the final “parent” helped it find its form and substance. Microelectronics was also developing at the time (although it was further along), and people saw a way to manufacture micro-sized structures precisely. Microfluidic components were first made of silicon and glass with microelectronics techniques such as photolithography. Eventually, silicon was replaced by materials with more suitable properties. Silicon is not durable, cheap nor transparent to visible or ultraviolet light. This is important because light is often used to either determine the results of a test, or to or to provide excitation energy. The most popular substitute material has been PDMS (polydimethylsiloxane). PDMS is a soft, transparent elastic polymer that is permeable to carbon dioxide and oxygen, making it useful for housing cells. Other cheaper, alternative materials have also been developed, such as thermoplastics, paper and string.

Advantages of Microfluidics

These four parents have given birth to a field that focuses on biochemical and medical applications, and with good reason. While everyone loves small things (I’m sure that micro is sad that it never got its time in the sun before the world became obsessed with nano), there’s more to it than that. A microfluidic device does have a smaller footprint than a huge machine in a lab, but it also uses a lot less fluid too. Why would you make a whole pot of soup just to taste and then dump it down the drain? On the other hand, I would make as little as I could and still taste it. So we can see that microfluidics is beneficial because it uses less costly chemicals and reagents, but it's also attractive due to the physics happening at the micro scale. Now, we’re not talking about leaving the world of classical physics and entering quantum physics, but matter does behave differently, most evidently in the way liquids flow.

Turbulent flow vs Laminar flow

Turbulent flow vs Laminar flow

Laminar flow around air bubbles ( source )

Laminar flow around air bubbles (source)

In fluid mechanics, there are two main types of flow: laminar and turbulent. Turbulent flow is just how it sounds, violent and chaotic. Laminar flow is peaceful, and particles will basically maintain their course without mixing. The state of a flowing liquid is determined by its Reynolds number. This is proportional to the dimensions of the channel and the velocity of the liquid. Slower liquids travelling in narrower channels are more likely to be laminar. This not only makes pretty images like those from Albert Folch’s lab, it gives a new level of control to researchers. Two laminar fluids flowing side-by-side will only slowly mix by diffusion, rather than convection. This allows us to do some clever manipulations, like selectively inserting components to induce mixing in some of the streams in a multi-stream channel. Other techniques exist to create droplets, which allow for controlled mixing, and capillary action is utilized to eliminate the need for pumps.

Microfluidic Applications and Issues

Paul Yager  with a lab-on-a-chip that  detects malaria

Paul Yager with a lab-on-a-chip that detects malaria

Microfluidics can be incorporated into a lab-on-a-chip which is used as an assay or diagnostic test and replaces its large, costly full-size lab counterpart. This has three basic components, a sample, a process and a validation method. For example, you might input a fluid such as saliva or blood to be analyzed. Next, you’ll have to process this blood mechanically or chemically. This may mean mixing it or applying filters to isolate what you want, and then introducing a possible chemical reaction. Even so, what good is this reaction if you can’t tell if it happened? You may be able to see a change in color, or look at it under a fluorescent microscope. These three components can be fulfilled in dozens of ways to create whichever test you so desire.

You must be saying to yourself, “Wow! What a technology! So why haven’t I heard more about this if it can do so many magical things?” Well, space travel is cool too, but when was the last time that you were out in space? Microfluidics just hasn’t made its breakthrough to the rest of us consumers. We can create some pretty cool applications with it, but how often can those prototypes be easily mass-produced? And how easy will they be able to use and troubleshoot and fix? Both concerns are even more important for point-of-care applications. Point-of-care refers to a medical application that is being used right where the caregiver meets the patient. In the developed world, this means that your samples don’t have to be sent to a lab somewhere to be processed, and in the developing world, this gives access to life-saving diagnostics that are normally too expensive to purchase and operate. More importantly, the quick response time of the devices would provide results to people who have travelled to a far-away clinic and need to return home again. No matter how elegant or complex a microfluidic device is, it must be easy and cheap to make, simple to use and durable.

Now, I don’t want it to seem like there are no companies producing microfluidics-based technologies. Some relevant ones are Helicos, FluidigmClaros DiagnosticsMicroFluidic Systems, Affymetrix and Advanced Liquid Logic. If you want to see a more complete list, check out the one at FluidicMems. I hope that I’ve helped you to understand the basics of microfluidics, and you can continue to learn about the field on this site. Oh, and if you were wondering how long it will be until you can get your hands on some microfluidic tech, you probably already have: your inkjet printhead incorporates microfluidics!

This is one part of my Microfluidics Beginner’s Guide. Check out the rest of it and keep learning!


ResearchBlogging.org Whitesides, G. (2006). The origins and the future of microfluidics Nature, 442 (7101), 368-373 DOI: 10.1038/nature05058