Henne van Heeren, CEO, enablingMNT, Vania Silverio, R&D microfluidics scientist, Institute of Systems and Computer Engineering for Microsystems and Nanotechnologies (INESC MN), Christina Pecnik, development engineer in life science and diagnostics, IMT Microtechnologies, and Elsa Batista, head of volume and flow laboratory, Portuguese Institute of Quality (IPQ)
Microfluidics is a technology that is used to create a wide diversity of miniaturised products, ranging from microreactors for the chemical industry to medical diagnostic devices. The COVID-19 pandemic has highlighted the importance of fast, reliable and specific testing as well as opportunities for RNA-based vaccines. Efforts around the world are therefore increasing significantly to create miniaturised point-of-care (POC) devices for fast and precise detection of COVID infections, DNA sequencing devices for identifying virus variants and RNA synthesising devices for future vaccines. The microfluidics industry is working hard to develop standards that aim to decrease the transition time to market, decrease cost and improve the quality of these miniaturised devices and their manufacturing methods.
A group of organisations, which includes microfluidic supply chain leaders, initiated the foundation of the Microfluidics Association (MFA) to officially develop such standards. In 2016, a first step towards microfluidic standardisation was made with the development of IWA 23:2016¹. This document was created to facilitate the uptake of microfluidic devices by making them easier to use, reducing the cost for assembling and enabling plug-and-play functionality. Recently, a new standard, ISO 22916:2022,² was developed based on the information from IWA 23:2016. In this document, the position of microfluidic ports is defined (see figure 1).
The positioning of microfluidic ports for a standard connection, as defined in ISO 22916:2022.Courtesy of the Microfluidics Association (MFA).
However, although this new standard defined microfluidic classes as a first step toward metrology standardisation, it still lacks dedicated metrological specifications required for accurate and reproducible manufacturing. The MFA has therefore started to work closely with MFMET, the European Metrology Programme for Innovation and Research (EMPIR) initiative for establishing metrology standards in microfluidic devices, to define metrology challenges for microfluidics and to take actions to overcome these challenges. After a short overview of main microfluidic technologies, this article will give a short update of the progress of this endeavour.
Main fabrication processes
The two most used materials for microfluidics are glass and polymer. The processing of these materials is very different.
Glass wafers are processed in a similar way to semiconductor wafers, namely a structure is generated onto a glass wafer through lithographic processes. Follow-up processes such as wet etching and/or powder blasting result in microfluidic channels and other structures (e.g., access holes). The assembly of wafers to create fully closed microfluidic chips can be undertaken by fusion, thermal, adhesive or low-temperature bonding of a cover wafer. The separation of the wafer into single devices is performed by dicing. Connectors attached to the device allow the interaction of the glass chip with the world. The advantages of glass processing are the accuracy of structuring, the durability of the devices, easy integration of sensors and the transparency of the used material, which is important for optical sensing.
Polymeric microfluidic devices and parts are produced in a single step, taking advantage of the thermal deformation or melting of the polymer. The most used fabrication processes for polymeric microfluidics are injection moulding, embossing and stamping. In injection moulding, powder or granular polymer is heated as a moulding compound until it melts or becomes plastically formable. The polymeric mass is then injected and compacted by a rotating screw under high pressure into a closed mould. This mould determines the shape and surface structure of the finished part. It is seen as the most cost-efficient fabrication method for microfluidic parts in large quantities. The assembly of microfluidic parts into a fully closed component can be carried out at low temperatures without harming the structures and features. It also allows for the preservation of functionalised surfaces for sensing purposes.
Embossing and stamping rely on the deformation of the polymer when subjected to heat and pressure. In these processes, a solid plate or sheet of the polymer is forced into an open cavity containing the mould and is thermoformed (i.e. pressed against the mould at a given temperature). Embossing uses a higher temperature in the thermoforming process than stamping.
It must be kept in mind that whatever material and process is chosen, in the end, the microfluidic part must be combined with other parts to create a complete device. This assembly is in most cases an expensive part of the total production cost. After assembly, functional testing of the created device becomes possible; but in most cases, such testing is destructive.
All manufacturers are required to check the quality of their products and processes during processing. There is therefore an increasing interest in the development of testing methods for different key measurement quantities in a variety of materials and production schemes for microfluidics. For this, measurement accuracy and traceability are essential. In more established industries, standard measurement protocols form the mainstay of product and processing quality control, but it is not so in microfluidics yet.
Metrology challenges during the production of microfluidic devices
From a recent series of interviews conducted by one of the authors, it became clear that the lack of standard metrology tools and protocols is impeding fast industrial development. To quote one of the interviewed experts: “Our company has been in the business of high-tech microfabrication since the 1970s and has vast experience of measurement during fabrication. Now, since we have been involved in microfluidics for a few years, we realise that the measurement challenges for this part of production need completely new expertise; expertise that we are slowly, too slowly, building up, and we encounter the problem that we cannot (yet) rely on standard protocols and affordable off-the-shelf instrumentation.”
The lack of terminology, harmonisation and common understanding of microfluidic terms are hindering the community. Another issue is that several of the microfluidic tests used in academia and in industry do not have a link to traceable standards for quantities such as flow, length, volume, etc.
More specifically, the main challenges relate to the following factors.
- Flow resistivity/pressure decay inside the device
Deviation from the intended cross-section leads to an increase or decrease of the flow resistivity and thereby influences the performance of the device. These deviations can be caused by variations in etching in the case of glass products or deforming during bonding of polymeric products. Surface properties such as roughness or wettability will also influence flow resistivity and pressure decay.
The microfluidics industry therefore needs tools and protocols for the measurement of channel dimensions (in 3D). The preferred time for measurement is after etching in the case of glass and after bonding in the case of polymers. The ability to measure the entire area of the wafer or even sample area of it in a short amount of time is desirable and it would give an indication of the process homogeneity and stability over the wafer. It would also predict internal flow resistivity and pressure decay. Particular measurement challenges are the range of sizes, from nanometres to hundreds of microns or even a few millimetres, and the complex cross sections and patterns (see figure 2).
An example of a complex microfluidic channel structure. Courtesy of the Institute of Systems and Computer Engineering for Microsystems and Nanotechnologies (INESC MN).
2. Bonding quality
The bonding quality of the microfluidic part and its covering layer can be tested by tensile strength measurement, i.e., pulling both substrates from each other. This is however a time-consuming process and destructive. Alternatives are to pressurise the device with air or liquid and check for leakage or check the bonding interface for contamination and voids. Adhesively bonded substrates must be tested not only for their bond strength, but also inspected for contamination inside the channel as this can affect the wettability of the surface. The visibility of features inside the channel after bonding or the optical window for measurement can be impaired when surface roughness is high or when the optical transparency is limited.
3. Optical transmission of materials used
Most microfluidic-based diagnostic devices operate via optical detection of biochemical reactions inside the microfluidic channels. Measurement of the optical transmission of the used material is therefore essential. Optical transmission methods rely on an optical path, the light wavelength and the detector. The measurement result depends very much on the material properties and the fabrication process, and no standard method is yet agreed, making it difficult to qualify the materials used.
4. Testing leakage, burst pressure and maximal operating temperature
Due to the small amounts of medium leaking in or out of microfluidic devices, detection is more challenging than for leakage in devices that have larger internal volumes. To know if a device leaks or not is not enough. Knowledge of the amount of medium leaking from a microfluidic device over a certain period of time is essential for the statistical process control needed to safeguard product quality. There is a strong preference towards using air or dry nitrogen as a test medium as it allows the quantitative measurement to be undertaken in a fast and non-destructive way, leaving no contamination inside the microchannels. However, most microfluidic devices use liquids as a medium. The relation between leakages of gas and of liquid in such small amounts does not solely depend on the viscosity difference; other factors might also play a role.
The burst test is a destructive test performed to determine the pressure at which a given component will rupture or fail. It can be defined as the point at which a product fails while being internally pressurised, i.e., the expression of the maximum pressure that can be endured the product before it breaks. This will enable a sufficient safety factor to be applied when in use. The burst test provides the margin between the maximum working pressure in the field and the pressure when the product completely fails (the burst pressure).
The maximal operational pressure test is used to check if a device or system functions properly before it leaves the manufacturing plant or may be part of a final quality check after production or repair/routine maintenance. The device is pressurised for a period of time at a pressure lower than the burst pressure, generally at two-thirds of the burst pressure, and checked for pressure decay. If the burst pressure is not known, an alternative might be testing at 150 percent of the expected working pressure. This is also a method to check for early life failures, i.e., failures that are unexpected and occur at the beginning of the product’s life and may be caused by chemical or mechanical wear. These differ from failures occurring at the end of the product’s expected life.
There are no industry wide accepted protocols for these tests yet, although the MFA has set up a working group with the objective of proposing such protocols. A whitepaper created by MFA members and other experts is expected soon.
5. Measurement of fast changing flow rates
Specialised labs are able to measure microfluidic flows very precisely (see figure 3). One of the advantages of microfluidic based diagnostic devices is that they provide measurement results fast. In a short time, the flow of the medium to be measured must start and stabilise. Precise control of flow is therefore essential. Calibrated flow measurement tools and methods for measuring fast changing flows is still lacking.
Calibrated equipment for the measurement of microfluidic flows. Courtesy of the Czech Metrology Institute (CMI).
Different types of equipment are used to generate and control flows in microfluidic systems; however, volumetric flow rate accuracy has not yet reached a satisfactory level. It is necessary to characterise components within a system to correctly model and predict the system level behaviour to optimise component usability and interchangeability. Currently, there are no industry-wide accepted tests, qualification protocols or generic compliance certificates to objectively compare microfluidic flow control products from different suppliers.
In a concerted effort to tackle the increasing demand for passive flow, national metrology institutes (NMIs) have already developed protocols and calibrations services for small flow rates³,.⁴ Traceability to national standards has been available since 2012 down to 0.1 μl/min, but there is the need to transpose this information to microfluidics technology, especially for flow control specifications.
Conclusions and future plans
The microfluidics industry needs test protocols that facilitate:
- communication for mutual understanding between customer and supplier;
- standard protocols to normalise testing practices between companies; and
- manufacturing practices and commerce between different customers and companies around the world⁵.
There is therefore a need for affordable, fast and accurate instrumentation for standard measurements of dimensions of microfluidics structures and characterisation of the interface between material and medium to be able to predict:
- flow resistivity/pressure decay inside the device;
- quality of the bonding;
- optical transmission of the materials used;
- microfluidic leakage, burst pressure and maximal operating temperature;
- fast changing flow rates; and
- measurement of fast changing flow rates.
Fulfilling these microfluidics industry needs would facilitate a decrease in fabrication costs and production time. Protocols deriving from such systems would not only be important for the user but also process mapping (in)stabilities for the manufacturer. This is essential for further quality improvement and yield increase.
Standardisation of performance characteristics is needed for the different classes of microfluidics components, including test conditions, measurement protocols and guidelines; because, so far, each microfluidics product is tested according to its own protocol, making generalising the result and bringing that information back to the design and fabrication stages difficult. Due to the testing methods and procedures for microfluidics still being user/manufacturer- or device-specific, the development of industry-wide accepted tests, qualification protocols or generic compliance certificates for microfluidics is fundamental for addressing different key measurement quantities in a variety of materials and production schemes to ensure measurement accuracy and traceability.
Furthermore, there is work to be done to develop general standards and guidelines for interfaces and connectivity.
Acknowledgements
The authors acknowledge the help of MFA members and other microfluidic experts for their help in formulating the microfluidic challenges. Some of the authors are participating in the 20NRM02 MFMET project, which has received funding from the EMPIR programme co-financed by the participating States and from the European Union’s Horizon 2020 research and innovation programme.
enablingMNT
INESC MN
IMT Microtechnologies
Portuguese Institute for Quality (IPQ)
¹International Organization for Standardization (ISO) (2016). IWA 23:2016 Interoperability of microfluidic devices—Guidelines for pitch spacing dimensions and initial device classification.
Available at: https://bit.ly/3HXHNfB
²International Organization for Standardization (ISO) (2022). ISO 22916:2022 Microfluidic devices—Interoperability requirements for dimensions, connections and initial device classification.
Available at: https://bit.ly/37hFKGv
³National Metrology Institute of the Netherlands (VSL), Technical Centre for Aeraulic and Thermal Industries (CETIAT), Danish Technological Institute (DTI), Federal Institute of Metrology (METAS), Portuguese Institute of Quality (IPQ) and Scientific and Technological Research Institution of Turkey (TUBITAK) (2016). JRP HLT07 Metrology for Drug Delivery (MeDD). [online] Braunschweig: EURAMET.
Available at: https://bit.ly/3hPiZvy
⁴Batista, E., Furtado, A., Pereira, J., Ferreira, M., Bissig, H., Graham, E., Niemann, A., Timmerman, A., Alves e Sousa, J., Ogheard, F. and Boudaoud, A. (2020). New EMPIR project—Metrology for Drug Delivery. Flow Measurement and Instrumentation, volume 72, p.101716.
Available at: https://bit.ly/3pS3RSx
⁵Reyes, D. and van Heeren, H. (2019). Proceedings of the First Workshop on Standards for Microfluidics. Journal of Research of the National Institute of Standards and Technology, volume 124, article no: 124001.
Available at: https://bit.ly/3sUWLhZ