Choosing the Right Microfluidic Chip: Materials, Fabrication, and Cost
March 10, 2025
No Comments
Selecting the right material for your microfluidic chip is a critical decision. The material impacts everything from ease of fabrication and cost, to chemical compatibility, biocompatibility, and optical transparency. Early microfluidic devices were made from silicon and glass (borrowing techniques from microelectronics), but today a variety of polymers are used as well (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center). The most common materials can be grouped into three categories:
Each material has pros and cons. Table 1 provides a comparison of three widely used chip materials: PDMS, glass, and representative thermoplastics.
| Material | Advantages | Limitations | Typical Uses |
|---|---|---|---|
|
PDMS (silicone elastomer) |
– Low cost and easy to prototype (cast from molds) (Materials for microfluidic chips fabrication : a review 2017 – Elveflow) – Optically transparent (clear from UV to visible) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Biocompatible; gas permeable (allows cell respiration) (Materials for microfluidic chips fabrication : a review 2017 – Elveflow) – Soft, enabling built-in valves/pumps via membrane deformation (Materials for microfluidic chips fabrication : a review 2017 – Elveflow) |
– Absorbs/adsorbs small hydrophobic molecules (drug compounds, biomolecules) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Swells or degrades in many organic solvents; limited chemical compatibility (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) – Hydrophobic surface can cause bubbles and needs treatment for aqueous flows (Materials for microfluidic chips fabrication : a review 2017 – Elveflow) – Not ideal for large-scale production (manual casting process) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) |
– Rapid prototyping in research labs (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Cell culture devices (need gas exchange)- Low-cost disposable bioassays (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) |
|
Glass (borosilicate, etc.) |
– Excellent optical transparency (glass is clear and has low autofluorescence) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) – Highly chemically resistant (can handle organic solvents, acids/bases) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Biocompatible and non-toxic; no leaching- Rigid and thermally stable (suitable for high-pressure or high-temp processes) |
– Fabrication is complex and expensive (requires micromachining or etching) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Bonding glass layers needs high temperatures/pressures or specialized surface treatments (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Not gas-permeable (not suitable for long-term cell culture that needs oxygen exchange) (Materials for microfluidic chips fabrication : a review 2017 – Elveflow) – Brittle (risk of chip breakage if dropped or stressed) |
– Analytical chemistry (e.g., capillary electrophoresis chips) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Chemical reaction platforms (where solvent compatibility is crucial)- Detection cells for optics (fluorescence, absorbance) |
|
Thermoplastics (e.g., PMMA, COC, PC) |
– Low material cost, disposable if needed (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Scalable mass production by injection molding or embossing (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Good optical properties (many are as transparent as glass; COC has very high clarity) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) – Hard plastic provides mechanical stability; easier integration with fittings and instrumentation (no flexing) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) |
– Some plastics have limited solvent resistance (e.g., PMMA can craze with certain solvents) (Considerations When Switching from PDMS to Thermoplastic Microfluidics) – Surface often hydrophobic; may need coating or plasma treatment for wetting or to reduce adsorption (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) – Molding of microfeatures requires precision (molds can be expensive upfront)- Not gas-permeable (if cell culture requires O₂/CO₂ exchange, needs special design) |
– High-volume manufacturing of cartridges for diagnostics- Point-of-care devices (disposable test chips) – Microfluidic devices for which PDMS prototyping is translated to plastic for scale-up |
Table 1: Comparison of common microfluidic chip materials, highlighting their key properties, drawbacks, and example applications. PDMS = polydimethylsiloxane; PMMA = poly(methyl methacrylate); COC = cyclic olefin copolymer; PC = polycarbonate.
As shown in the table, PDMS (Polydimethylsiloxane) is by far the most popular material for prototyping in academic labs. It became popular because of its rapid prototyping capability – you can pour PDMS onto a mold, cure it, and peel off a microstructured chip within hours (Materials for microfluidic chips fabrication : a review 2017 – Elveflow). It’s also optically clear and fairly inert in aqueous environments. However, PDMS is not a cure-all; it has notable weaknesses in solvent compatibility (it swells in many organics) (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center) and can adsorb proteins and small molecules, which can be problematic for analytical accuracy (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center). PDMS devices are great for experiments and small-scale use, but if you need hundreds of devices or have to run organic solvents, you’ll likely consider other materials for scale-up.
Glass remains an excellent choice when chemical inertness and optical clarity are paramount. You can treat a glass chip much like a microscope slide – apply harsh solvents, image through it with high-end optics, etc., without the material affecting your assay. Glass chips can be made via etching or ultrasonic machining and then bonded, but this process is not trivial in a typical lab. It often requires cleanroom facilities and specialized equipment, which is why glass is less common at the prototyping stage (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center). Companies and foundries can produce glass microfluidic devices for you, and they are used in some commercial analytical instruments (for example, capillary electrophoresis chips and DNA sequencers). Glass is also used in hybrid chips – for instance, a PDMS layer bonded to glass to combine the best of both (PDMS microchannels with a rigid, transparent glass cover).
Thermoplastics like acrylic (PMMA), polycarbonate, and cyclic olefin copolymer (COC) are increasingly popular, especially for translating a prototype into a product. These plastics can be injection molded or hot-embossed to replicate microchannel patterns at large scale, which makes them suitable for commercialization. Each thermoplastic has its own chemical compatibility profile – e.g., COC and COP have excellent solvent resistance (even better than some glass, in that they don’t shatter and can handle organic solvents) (Considerations When Switching from PDMS to Thermoplastic Microfluidics). PMMA is less solvent-resistant (it might crack or craze in strong solvents), but is easy to machine. Most thermoplastics are biocompatible and have been used in FDA-approved devices. A downside is that many are hydrophobic; thus surface treatments (plasma, polymer coatings) are used to make channels wettable or to prevent cells/proteins from sticking (How to choose the best chip material for microfluidic fabrication? – Microfluidics Innovation Center). Unlike PDMS, once fabricated, plastic chips are not easily modified – but they are much more robust (you can drill holes for tubing connectors, they won’t deform under pressure, etc.). For these reasons, it’s common to prototype a device in PDMS, then transition to a thermoplastic for a more rugged version suitable for higher-throughput use or distribution to other labs.
Material selection should be driven by your application’s needs: If you need fast turnaround and microvalves, PDMS is great. If you need organic solvent compatibility or long-term cell culture, consider glass or specialty polymers. If you aim for mass production (making thousands of chips), thermoplastics with injection molding will be most cost-effective in the long run.
Just as important as the material is the method used to fabricate the microfluidic structures. Different fabrication techniques are suited to different materials and production scales. Here we overview the most common methods:
Soft lithography is the workhorse method for academia to make PDMS chips. It involves creating a master mold (usually by photolithography on silicon wafers using SU-8 photoresist to build up channel patterns), and then casting liquid PDMS pre-polymer over this mold (Materials for microfluidic chips fabrication : a review 2017 – Elveflow). After curing, the flexible PDMS slab with embossed channels is peeled off. It can then be bonded to glass or another PDMS layer (often by oxygen plasma treatment) to seal the channels. Soft lithography is quick and relatively inexpensive for prototyping. A single mold can be reused dozens of times to cast chips, and making the mold is the main upfront effort.
Key benefits of soft lithography:
Limitations:
Despite these limitations, soft lithography has revolutionized microfluidics by dramatically lowering the barrier to fabricating custom designs. Most concepts are first proven in PDMS chips via soft lithography before moving to other methods.
For prototyping in plastics (or making molds), computer numerical control (CNC) micromilling is a useful technique. A precision milling machine can directly carve microchannels into materials like PMMA, COC, or even metals. Milling can achieve channel widths on the order of tens of microns (though typically a bit larger than photolithography resolution) with very smooth surfaces if done well (Considerations When Switching from PDMS to Thermoplastic Microfluidics). It’s particularly good for rapid prototyping of thermoplastic chips one at a time, or for creating molds (e.g., milling a brass mold that will later be used for injection molding). Similarly, laser ablation or laser engraving can cut channels into plastics by vaporizing material along a designed path (Considerations When Switching from PDMS to Thermoplastic Microfluidics).
Benefits:
Drawbacks:
In practice, milling is great for making quick prototypes in hard plastic to test form-factor or fluid connections, and for situations where PDMS isn’t suitable (e.g., testing with a solvent that would destroy PDMS). Some startup companies offer rapid micro-milling services for microfluidics. Laser micromachining is similarly used for fast turnaround. These methods can also complement lithography (for example, drilling access holes in glass chips, or cutting out devices).
For large-scale production of polymer chips, replica molding techniques are standard. Injection molding involves injecting molten thermoplastic into a precision-machined mold cavity that defines the microchannel geometry. Hot embossing involves pressing a mold (often called a stamp) into a heated thermoplastic substrate to imprint the pattern. Both require an initial high-precision mold (typically made of steel, nickel, or silicon) but once you have the mold, you can churn out numerous copies quickly.
Both injection molding and embossing create rigid polymer chips with the exact design of the mold. Features like vertical sidewalls and smooth surfaces are achievable. These methods are considered when you have a finalized design that you want to manufacture at scale. It’s not uncommon for an academic group to prototype in PDMS, and if the application shows promise (say a diagnostic device), partner with a company to injection mold the design in plastic for a field trial.
Note: The transition from PDMS to thermoplastic isn’t always plug-and-play – sometimes designs need tweaking due to the material change (for example, PDMS’s flexibility allows valve structures that rigid plastics can’t replicate directly (Considerations When Switching from PDMS to Thermoplastic Microfluidics)). But many microfluidic geometries can be transferred successfully, and there are even foundries that specialize in microfluidic injection molding to help with this process.
3D printing (additive manufacturing) has emerged as an intriguing option for microfluidics because it can directly create complex 3D channel architectures that are difficult with planar lithography (Can 3D printing be used to manufacture microfluidics – uFluidix). There are two main uses of 3D printing in this field: (1) printing molds (which you then use for casting PDMS or molding plastic), and (2) printing the actual microfluidic device itself in one go.
Several 3D printing technologies are used:
The advantages of 3D printing are clear: design freedom (truly 3D channel networks, including features like curved channels, internal supports, etc.), and fast iteration – you can CAD a design and have a physical prototype the same day in some cases (Can 3D printing be used to manufacture microfluidics – uFluidix). This “print on demand” approach encourages rapid experimentation and can realize fluidic structures that are impossible to make with standard lithography (like a knot in a channel, or a smooth 3D branching network).
However, current 3D printing still has some challenges for microfluidics:
That said, 3D printing has already proven useful for things like rapid prototyping of microfluidic connectors and manifolds (for example, printing a custom adapter to interface a microfluidic chip with tubing) and for integrated devices like organ-on-chip platforms that require complex chamber geometries. It’s also democratizing fabrication – you don’t necessarily need a cleanroom, just a reasonably good 3D printer and some know-how.
In summary, researchers have an expanding toolbox of fabrication methods. Often the progression is: design concept → PDMS prototype (soft lithography) → refined design → thermoplastic or 3D printed version for more rigorous testing → injection molding for production. It’s not one-size-fits-all; the method depends on the material and the stage of your project.
(Conventional microfabrication methods summary: photolithography/etching for silicon/glass, casting for PDMS, molding/embossing for plastics, micromachining and 3D printing as newer alternatives (Can 3D printing be used to manufacture microfluidics – uFluidix).)
When choosing a material and fabrication method, it’s vital to consider what fluids and conditions your chip will be exposed to, as well as any requirements for imaging or detection. Different materials have very different compatibilities:
In summary, match your material to your chemistry and detection. If in doubt, perform simple tests: flow your solvent or sample through a sample piece of the material and observe for swelling, leaks, or analyte loss. Many researchers start with PDMS for convenience, but if they encounter issues like drug molecules disappearing (absorbed into PDMS) or channels deforming under certain pressures, they consider switching to a hard plastic or glass. Fortunately, there are now many options, and even hybrid approaches (e.g., glass-coated PDMS, or PDMS on PCB boards, etc.) if you need multiple characteristics in one device.
Cost is a major factor in choosing fabrication approaches. The economics of microfluidic chip production can be very different depending on whether you need 10 units, 1000 units, or 1 million units. Here’s a breakdown to guide decisions on prototyping versus scaling up:
In practical terms, use PDMS/3D-print prototypes to nail down the science, but engage in “design for manufacture” once you have a proven concept. The transition could involve making a thermoplastic version and testing it on a smaller scale (maybe 50–100 via a milled mold or an external service) to validate that it performs the same. Then invest in high-volume tooling. Remember, each step of scale-up will have its own costs and learning curve – budget both money and time for it.
To give a sense of scale: an academic lab might prototype a microfluidic device with a budget of $1000 or less (materials and some facility use). To get 100 units made in plastic might cost on the order of $5k–$10k via a service. To get 5000 units injection molded might cost $50k (including a durable steel mold) but then each additional unit is maybe $1. The exact numbers vary, but this gradient shows how cost per chip can drop by orders of magnitude when moving to industrial methods, at the expense of upfront investment (Prototyping Microfluidics vs. Manufacturing Them – uFluidix) (Prototyping Microfluidics vs. Manufacturing Them – uFluidix).
In summary, prototype cheaply, but plan for scale if needed. When the time comes to move beyond the lab, involve companies or engineers with manufacturing expertise. They can advise on material substitution (e.g., which polymer matches your PDMS chip’s performance) and on modifications to make fabrication reliable and cost-effective. Many microfluidic innovations fail not because the science is bad, but because the device couldn’t be made at scale or at a reasonable cost. By understanding the cost trade-offs and preparing for them, you can ensure your microfluidic chip is both scientifically useful and manufacturable.
Call to Action: Selecting materials and fabrication methods can be daunting, but making informed choices early will save you time and money down the road. If you’re unsure which material best suits your experiment or how to scale from a prototype to many chips, we’re here to help. Our team specializes in custom microfluidic chip development – from advising on material compatibility to arranging cost-effective fabrication. Whether you need a single prototype in PDMS or a batch of 1000 plastic cartridges, we can guide you to the optimal solution. Contact us to discuss your project’s requirements, and let us help you design a custom microfluidic chip that balances performance, manufacturability, and cost for your specific research application.
Shuhan He, MD is a dual-board certified physician with expertise in Emergency Medicine and Clinical Informatics. Dr. He works at the Laboratory of Computer Science, clinically in the Department of Emergency Medicine and Instructor of Medicine at Harvard Medical School. He serves as the Program Director of Healthcare Data Analytics at MGHIHP. Dr. He has interests at the intersection of acute care and computer science, utilizing algorithmic approaches to systems with a focus on large actionable data and Bayesian interpretation. Committed to making a positive impact in the field of healthcare through the use of cutting-edge technology and data analytics.
Monday – Friday
9 AM – 5 PM EST
DISCLAIMER: ConductScience and affiliate products are NOT designed for human consumption, testing, or clinical utilization. They are designed for pre-clinical utilization only. Customers purchasing apparatus for the purposes of scientific research or veterinary care affirm adherence to applicable regulatory bodies for the country in which their research or care is conducted.
