Micro Molding Advances
Every year, microfluidic molders push the envelope to meet the requirements of new analytic technologies. Much of this activity is hidden behind CDAs between the molder and the OEM. Slowly as the products are commercialized, we get to see some of the incredible advancements being made. One area that has pushed the boundaries is cell analysis by electrophysiology. A single cell is captured and analyzed. This requires very small orifices and flow channels. DNA sequencing and all the emerging technology surrounding it will become the biggest driver of microfluidics over the next 10 years. This will drive requirements to mold thin walls to allow thermal cycling to occur, the need for optical clarity and low autofluorescence to scan for test results, and dimensional precision between multiple channels for uniform results.
Injection molding continues to provide the most economical manufacturing process for disposable microfluidic cartridges. The tradeoff has been the limitation of what features can be molded. This is limited by the construction of the mold and processability of the thermoplastic. Classic molding relies on molds fabricated by machining processes and conventional injection molding. The CD, DVD, and Blu-ray products led to the commercialization of two new processes. On the mold making side, laser etched glass masters and electroplated/electroformed nickel inserts allowed much smaller and accurate features to be produced. Coupled with new injection/compression molding machines that produced flat stress free parts, new opportunities were created. Both approaches have a place in today’s market. Let’s look at what each approach offers and what limitations are presented.
Injection Compression Molding
The limitation of conventional injection molding is the inability to keep uniform pressure in all areas of the part as the polymer solidifies. This leads to a balancing act between pressure for replication and resulting flatness or lack thereof. Poor replication shows up in the edges of the fluidic channel (see Figure 1). Upon lamination, this corner radius can lead to air entrapment and poor performance.
Injection compression, sometimes called injection coining, solves this by reducing the cavity volume as the polymer solidifies. This is achieved by having the entire recessed surface of the core move slightly towards the parting line at the completion of the mold fill. This compensates for the reduction in volume during the solidification. This works best with thin, flat parts. The most successful use of this technology has been in the digital media disc. The best results are obtained when the part being molded has flat geometry with low aspect ratio for features that are perpendicular to the nominal plane of the parts surface. This keeps the pressure and volumetric change uniform throughout the part. The result is the creation of the pressure necessary to get excellent replication without inducing stress into the part.
Because the relationship between the core and cavity changes during compression, through holes can be problematic. A piloted hole on one side of the mold is necessary to receive the core pin as it telescopes during compression. If you have an SBS format-sized part with 100-200 through holes, the challenge is real (see Figure 2). One way to overcome this problem is to design the through holes in the mating lamination. Design the complicated fluidic geometry in one laminate and through holes for porting in the mating laminate.
Master/Replication of Mold Inserts
While this technology has given us digital-quality sound and high-definition video, it also allows molds to be produced with incredible detail for microfluidics. Features in the 5-10 um range are practical and there is little limitation on feature density. Depth of features can now be produced in the 2 mm range. Additionally, some innovative advances are combining replication inserts with traditional mold construction.
Machine Cut Molds
Conventional mold fabrication is limited by the quality and precision of the machine tools and cutters used in them. Advances in micromachining, end mill geometry, and end mill coatings allow the fabrication of features that were unheard of a couple of years ago (see Figure 4). End mills with a diameter of .02mm are commercially available and, with the right machining center, can produce surface finishes in the 1-2 um range. Surface treatments and coatings increase cutter life and allow reliable prediction of wear. Cavities with 10,000 mm2 of projected area and highly populated with fluidic channels can be machined flat to 5 um with features smaller than 50 µm. As cut, the surface finish is 2 µm (see Figure 6). If polishing is feasible, the surface finish can be improved to 80 nanometers or less (see Figure 7). These scans were produced on a white light interferometer (see Figure 5).
This method of mold fabrication has an advantage of flexibility for gate design, hybrid ejector systems and mechanical side actions in more complicated parts. Additionally, much deeper features can be fabricated. This allows hybrid designs of microfluidic incorporated into micro titre plates and other more complex cartridge designs.
Conventional Injection Molding
While conventional injection molding may not have the dazzle of injection/compression molding, it does provide an excellent process for many, if not most applications. Despite the replication/flatness balancing act, features 15-20 mm tall are easily accommodated. This allows the engineer more flexibility in the design of the product to incorporate wells, reservoirs, and ports. Because the core and cavity maintain the same relationship during the molding process, simple butt shutoffs can be used for cored holes (see figure 3). This helps reduce mold maintenance and associated cost for small through holes.
So, how can we deal with the replication/flatness issue? We need to understand the rheology that the plastic will undergo. Wall thicknesses of 1.00-1.200 mm and flow length from 25 to 75 mm are common. With the correct input values, it is possible to accurately model the molding process using CAE software. Combining this with empirical data, we can reliably predict the shrinkage at different points of the fill/pack and create a relatively low stress pack profile with optimal feature replication. This does requires that the mold be designed with variable shrink, primarily in the cross flow direction. All of this results in a flat part with good feature replication.
Conventional injection molding works best when features are large (100-200 µm), but is effective with features down to 50 µm; and nominal wall thicknesses 1 mm and greater.
Conclusion
The molding of microfluidic circuits is an exciting area of manufacturing today. Demands are ever increasing. This is pushing the capabilities of the manufacturers and they are responding. If you considered injection molding your fluidics years ago, but did not like the limitations you were presented with, look again.