by Dr Holger Becker
New, groundbreaking technological developments  are often initially expected to immediately revolutionise our way of living but frequently the revolution fails to appear. Many people argue that this is what happened with the much hyped innovative technology of microfluidics (or Lab-on-a-Chip).

A closer look reveals, however, that significant progress has been made in the microfluidics world. This article describes the current status and highlights directions for future developments. In 1990, the first “miniaturised total chemical analysis system” (µ-TAS) or “lab-on-a-chip” was introduced  to great acclaim [1]. There were several reasons for the excitement of the researchers who used the new system and other subsequent lab-on-a-chip systems:
• Reaction and analysis times were greatly reduced because of physical scaling effects
• The reduced size of the devices led to significant savings in reagent consumption and consequent reduction in waste
• It was realised that chip-based systems  were suitable to support a massive parallelisation of reactions, thus increasing the throughput of the system. Alternatively, a multitude of components or parameters could be measured simultaneously.
• The reduced footprint meant that smaller, even portable systems could be envisaged, e.g. for applications in point-of-care diagnostics (POC), in-line process control or on-site analysis.
• The concept allowed for sample-in  — answer-out systems with a high degree of user-friendliness and ease of operation,
• The systems can be used by relatively unskilled personnel or, e.g. in the case of POC diagnostics even by totally inexperienced users. Such advantages were all immediately recognised and are even more valid today.

 The universal appeal
An indicator of the progress in this area of research is that there is still exponentially growing academic activity. Although it could be argued that the field is attractive to people in academia simply because the cost of developing microfluidic chips and of performing experiments on them is comparatively low (certainly when compared e.g. to the capital and operating cost of a LC-MS system), this does not fully explain the large number of research and academic publications in the field. On the contrary, one of the drivers behind the continuing development activities is an intrinsic property of microfluidics, namely its applicability to an enormous number of scientific phenomena and its almost universal usefulness in a broad number of subjects.

 
Lab-on-a chip applications
Application examples range from ab initio computational studies of molecular dynamics through the whole field of analytical and bioanalytical chemistry to clinical diagnostics, from the synthesis of fine chemicals to the culturing of single cells, cellular matrices, tissue or even small organisms such as C. elegans. Particularly interesting applications are those that involve traditional scientific disciplines that are not normally associated with laboratory analyses. Examples are  the use of microfluidics in the geosciences to model sedimentation processes, or the coupling of neurons with an active microelectronic chip in a microfluidic manifold.

This breadth of applications of microfluidics can also be found on the commercial side of the field — it is this large variety of applications that currently defines business models in the microfluidics industry. Instead of a single application carried out at very high volume (lovingly dubbed the “killer-application”) and thus capable of generating large revenues and attracting for example large multinational companies, the field is dominated by small and medium sized enterprises (SMEs). Such SMEs typically have the flexibility to adapt quickly to varying demands for microfluidic functionality. Specific designs can be produced rapidly and the SMEs are able to manufacture relatively small production batches (in the order of thousands of chips instead of millions) in a comparatively short amount of time. This business model is in stark contrast to the one predominating in the microelectronics world. It should be noted that, at the beginning of the development of microfluidics,  the methods of production that were used to manufacture chips were mainly microfabrication methods and materials borrowed from the microelectronics industry. For this reason the microfluidics industry is frequently, and erroneously, compared to the microelectronics industry, ignoring several significant differences between the two. The most striking of these differences is the much wider range of basic elements used to generate functionality in a microfluidic system. While even after 40 years, an integrated circuit in microelectronics still consists basically of transistors and capacitors, the list of basic elements used in microfluidics contains channels, reservoirs, mixers, filters (to name just a few). In addition, for each of these, changes in individual geometrical dimensions can massively influence the functionality.

The comparative simplicity of componentry in microelectronics leads to a high level of standardisation, which is also apparent at a design level (e.g. functional component libraries). This is something that is almost totally missing in microfluidics.
The second difference lies in the material used. While microelectronics uses almost exclusively silicon, microfluidics can take place in glass, quartz, silicon, printed circuit boards and polymers. In total this can represent hundreds of usable materials, each with its own specific material properties. The number of materials means that the number of different fabrication technologies is correspondingly broad. Nevertheless two main technologies dominate the scene: silicone rubber casting (used mostly in academic settings) and injection molding of thermoplastic polymers (used mostly in industrial settings) [2].


 System development
The typical workflow involved in the development of an industrial microfluidic system using injection molding as the manufacturing method can be seen in Figure 1. The diagram shows in parallel not only the development phases of an instrument (normally a conventional development) but also those involved in the development of a microfluidic device which performs the sample manipulation and analysis. After defining the requirement specifications, the chip design phase starts. In this phase, the multidisciplinary nature of the development process is all-important since various aspects all have to be considered, such as the properties of the raw material, the basic physics of microfluidics and a detailed know-how of the application not to mention manufacturability considerations.

Recently, modelling and simulation have become important tools in this process, allowing early-phase statements on chip performance and possible limitations [Figure 2]. Upon completion of the design, a replication master, which is the geometrical inverse of the desired end-structure, has to be made using a process selected from a large range of microfabrication technologies. This master structure is then replicated in a polymer by means of injection molding. To have a final functional system, back-end processes such as the closing of the microchannels with a cover lid, the functionalisation of the channel surface, the immobilisation of biomolecules, the storage of reagents or the inclusion of electrodes then have to be carried out. It is noteworthy that a significant proportion of the fabrication cost of a microfluidic device is determined by these back-end processes so it is really worthwhile to consider them carefully at the earliest opportunity during the device design process.

 
Examples of commercially successful chips
Many commercially available examples of such polymer-based microfluidic structures have been developed. One example of this is a diagnostic platform chip for agglutination assays which has an interface fitted with a standard Luer port. [Figure 3] The sample is then divided into three different streams that react with different immobilised reagents stored on-chip in the reaction chambers. The results can be read by the naked eye. A more complex example is a chip for the dielectrophoretic assembly of cell co-cultures for cytotoxicity experiments in drug discovery. An AC-voltage at the electrodes generates a force gradient which guides the different cell types to the cellular matrix assembly area, where their reactions to exposure to different drugs can be analysed. Bioident Technologies Inc. have developed a nanotitre plate with integrated photodetectors made from organic semiconductor materials. This composition allows the read-out of fluorescent or luminescent reactions which take place in the wells using electronic instead of optical means, thus greatly simplifying the instrumental effort.

The above examples illustrate how microfluidics technology has matured. While much ground has already been covered, more work remains to be carried out both in academia and in industry before microfluidics becomes the routine technology that was initially expected of it. It is encouraging, however, to see the organisational developments in the industry, with good development and manufacturing infrastructure becoming available together with an increasing set of validated tools. Although development might have taken longer than many people anticipated at the beginning (as is usual for many disruptive technologies), microfluidics has certainly become a critical success factor in the development of novel analytical, bioanalytical or diagnostic systems.

References
[1] Manz A, Graber N,  Widmer H. Miniaturized total chemical analysis systems: a novel concept for chemical sensing. Sens. Actuators 1990; B1: 244-248.
[2] Becker H,  Gärtner C. Polymer Microfabrication Technologies for Microfluidic Systems,.Anal. Bioanal. Chem. 2008; 390: 89-111.

 The author
Dr. Holger Becker
microfluidic ChipShop GmbH
Carl-Zeiss-Promenade 10
D-07745 Jena
Germany
e-mail: hb@microfluidic-chipshop.com