1. Basic tools for microfabrication
For relatively simple microfluidics and devices, for example moulded PDMS microchannels bonded to glass slides, investment in a laminar flow hood for the clean space and PDMS processing tools costing >£10K can be adequate.
This strategy will also increase background knowledge in microfabrication and become useful when the scale-up stage is reached.
The device performance can then be optimised in an accompanying bio-lab. And with the optimised prototypes and detailed Standard Operating Procedures (SOPs) in hand, companies can then approach a foundry that meets the requisite flexibility, quality, cost, volume and scale-up manufacturing timetable.
If, however, the (initial) market for the product remains small, then a basic in-house manufacturing set-up or its parallel replication should suffice. Typical examples of device type, equipment, materials and cost are shown below.
macro machined/plated masters used to mould PDMS and bond to glass, hot embossed or injection moulded thermo-plastics, paper-based lateral flow assay
laminar flow hood, degassing chamber, plasma chamber, bonder/embosser, surface profiler, microscope, dicer, 3D printer and oven
glass (slides), PDMS, thermo-plastics (PS, COC, COP, PMMA, PET and PEEK), ceramics, metals and paper (lateral flow assays)
2. Open access facilities
If more complex devices embody your technology, for example with smaller features, greater precision and parallel wafer-scale processing or even embedded electrical tracks, sensors, actuators or on-chip electronics, then more conventional microfabrication techniques can be the solution.
Bear in mind that your staff will need to be trained in semiconductor-type process tools and understand cleanroom working practices. Equipment down time, IP confidentiality, scheduling, cross-contamination, low yield and poor reproducibility can be issues to contend with when using this approach.
After a few design–fabrication–test cycles to fully optimise the design, scale-up manufacturing can be undertaken at the same cluster or a foundry.
as in (1), plus deep silicon/glass etched channels/vias bonded to silicon/glass/polymer, metal tracks, sensors, actuators/moveable structures (diaphragm, inter-digitated fingers), moulded thermo-plastics, printing/SLA and laser machined chips
as in (1) plus 4” - 6” wafer size photolithography, D-RIE glass/silicon, sputter/e-beam evaporator, CVD/PECVD/ALD, furnace, electroplating, laser machining, injection moulding and metrology (ellipsometers, surface profiler, SEM/EDX, probe station…)
as in (1) plus silicon, polymers liquid and dry film (photoresist, SU8, Ordyl, polyimide etc.) and piezoelectrics
~£1K per day (depends on the short-/long-term usage)
3. In-house cleanrooms
For either simple or more complex microfabricated devices, it may make sense to invest in your own in-house R&D or low-medium scale-up production facility.
The cost of a fully equipped cleanroom depends on the processes required and whether it is just for R&D or manufacturing. But, an investment of between £1M and £5M is typical, along with a further £1M to £2M for a bio-lab.
Hiring of qualified process engineers and biologists will be necessary, along with understanding of running costs, including; service, maintenance, validation/calibration, consumables and investment in Quality (ISO 14644, 9000 or 13485) and Health & Safety procedures.
The rewards of having your own hi-tech labs can be lower manufacturing cost, IP generation, faster time-to-market, better control over processes and income stream via your in-house manufacturing.
as in (2)
as in (2), but generally 4”- 6” for R&D and 6”- 8” for scale-up
as in (2) plus silicon, polymers (photoresist, SU8, dry film resists, PMMA etc.)
~£1M to £5M
4. Rental facilities
An alternative to in-house and potentially costly high-tech labs is to rent. Cleanroom rentals are rare, but bioclusters or incubators with wet lab rental spaces are becoming common. These empty sterile labs will need to be equipped with appropriate bio-lab equipment depending on the project or biological needs.
as in (1) and possibly (2)
as in (1) and possibly (2), micro-bio safety cabinets, cell counters, centrifuges, flow cytometry, incubators, thermo cyclers, shakers, freezers, bio printers and microplate readers
cells, beads, lab reagents/chemicals, buffers, hydrogels, perylene etc.
~£50K to £150K per year rental + £0.5M-£2M for above equipment
If there is no appetite for a costly in-house lab or there are space constraints, then direct collaboration or outsourcing to a foundry/manufacturer/cluster/specialist could be the answer. This approach can also be more beneficial if the scale-up and development is done at the same foundry. However, the development time and costs can quickly escalate if the device geometry or the process steps are particularly challenging or indeed if the design is not quickly optimised.
as in (1), (2) and (4)
as in (1), (2) and (4), but typically 6” - 8”
as in (1), (2) and (4), although PDMS is not always allowed in (2)
£50K - £250K “pre-production” cost + £1K-£5K per 6” - 8” wafer depending on device complexity (number of masks), i.e. the smaller the chip, the lower the individual chip cost
A “basic” in-house clean area with simple lab equipment may work well for many simple microfabricated or microfluidic devices and biology. But for more complex devices, the use of a nearby microfabrication cluster, hub or university with a well-equipped and staffed cleanroom would be preferable.
Alternatively, investing and running your own cleanroom, semiconductor-type equipment and bio-lab should generate more IP, provide better control over processes, reduce time-to-market, and generate longer-term revenue through your own manufacturing, but at a significant financial cost.
In our experience, many companies employ more than one of these strategies as their plans mature.
But, until these approaches become more commonplace, this blog should provide some useful insight into early stage decisions on cleanrooms and bio-labs to make your life science start-up or SME a success.