Skip to Content
Dan k
Micro Fluidics Design
Explorations Dan Chen / January 23, 2016

Printing and design considerations

Using the FormLab 3D printer, we were able to consistently reproduce 1mm channel.

Making the channels easier to clean

  1. When using 1mm channels, try not to have one continuous long channels, if you really need a long channel try to vary the length, angle and patterns to have different pressure flow points. Making the channel slightly larger also helps.
  2. When you have a very long channel, consider placing a relief valve in the middle of the channel. You can design the cap or cover for the relief valve or close the valve using a fire sealed tube.
  3. It is easier to clean the long channel if you have the chip thickness around 3mm, anything above 4mm might be harder to clean
  4. chamfered edges on the 2 ends of the chip would make compressed air cleaning easier. The chip won’t fly off your hand as easily.
  5. Using the compressed air to dry the chip first, so it’s easier to grasp, then you can direct the compressed air right into the inlet or outlet or push some of the resin out.
  6. Alternating compressed air and isopropyl would help speed up the cleaning process. if one doesn’t work try another
  7. Flush the channels with isopropyl five or more times would help the micros grow in the chip. Also flow compress air through the chip for 30 or more seconds would help the resin dry the oven also make sure the isopropyl is completely   evaporated.
  8. Using the UV oven would increase the curing speed. We think 5 to 10 minutes is sufficient. Turn the chip over half way, because the UV light is only coming from the top of the oven.
  9. When laying out the print job, try to position the outlet or inlet facing away from the build base to allow gravity to pull the resin downward. This also helps to prevent the resin from trapping from the inside.

Avoiding printing errors:

  1. Try not to have a large flat surface at the top; this will cause the half-cured resin surface to collapse.
  2. Always wipe the build plate before printing, and check for particle such as hair or cured parts in the resin.

A Speedy way to get things off the build plate

Forget about placing the built plate on the formlab cleaning station. Place the built plate perpendicularly on a paper towel, then place the putty knife parallelly to the plate then scrape the parts off. This is much SAFER because we are resting the plate on the table and the knife is facing towards the table, giving you more leverage. Since we are not placing the built plate upside down, this prevents the resin from dripping towards the upper handle, which is a lot CLEANER then the formlab tray method.

Formlab also generates small “skirts” around the base, this makes it easier to take the parts off if you place the putty knife right under them.





Droplet Generator:

By altering the two inputs, air and liquid, we can create small liquid chambers with air gap as spacers. This design is great for locating certain bacteria in a large sample, such as sewage water.


Screen Shot 2015-01-21 at 8.17.41 PM




X-Mixer is the most effective mixer that we tested. We found that the most effective mixing is to flow the liquid right into each other by placing the channels facing each other. After the first mixing intersection, we split and dilute the solution in half then mix them again. We can probably reduce the mixing-intersections-blocks in half because it was already well mixed quarter the way through.

Screen Shot 2015-01-22 at 12.52.21 AM

Tips & tricks for printing on FormLab for Micro Fluidics


  1. Have the inlet or outlet facing up (facing away from the platform) The gravity would help the resin to drain.
  2. Try not to have a big flat top surface, this might cause the top to collapse.
  3. The print is able to produce 1mm channels 90% of the time

NOTE: These are my experiences from only 8 print sessions, that does not mean it will work every time or highly repeatable.


Other Designs



Gradient mixer:

Gradient mixer allows you to observe micros’ growth in different concentration of medium. You can also use this device to mix and incubate within the central disk structure, then flush out from the channel to the side.

Screen Shot 2015-01-21 at 6.57.11 PM

Screen Shot 2015-01-21 at 7.38.36 PM

IMG_3739Screen Shot 2015-01-21 at 7.37.48 PM

3D Mixer:

The 3D mixer has 4 inlets that works with gravity to pull the liquid slowly into the disk incubator. Once the mixture is in the disk incubator, we can use the 4 channels above the bio reacto as reservoir to store nutrients.

Screen Shot 2015-01-21 at 7.12.07 PM



Splitter dilutes the sample into smaller quantity for isolated incubation.

Screen Shot 2015-01-21 at 7.31.23 PM  Screen Shot 2015-01-21 at 6.58.24 PM

Screen Shot 2015-01-21 at 7.25.48 PMIMG_3882


Microbiology Experiments:

Initial Growth Tests:

We first tested the compatibility of E. coli RFP with ampicillin resistance on our resolution test chip.

We filled our chip with liquid media containing the E. coli, but we saw growth only on the tubing:

We then plated out the contents of the tubing and the contents of our chip on LB with ampicillin:

It was difficult to tell whether the growth from the chip was just a consequence of being in direct contact with the tubing.

We also tried growing E. coli with RFP in a cylindrical device:

We saw a white film in the channels and did not have much success when we plated out the contents of the chip:



We were interested in testing the compatibility of microbes on- and off-chip. Since we only had a few days to design and conduct our experiments, we chose to co-culture microbes with easily observable phenotypic differences. We chose to grow Providencia alcalifaciens, Providencia stuartii, Staphylococcus epidermidis, and the E. coli with RFP.

The stock plates display the phenotypic differences:

The Providencia strains were a deep orange with a rich, cheesy smell.

The Staph epidermidis formed smaller, whiter, dotty colonies.

We do not have a photo of the stock plates of the E. coli with RFP, but this plate was collected from one of our devices and shows the characteristic, bright red color of the colonies.

We set up combinations of bacteria to grow in different chips, testing whether we could grow microbes, whether certain microbes grow better on-chip, and whether growing them together vs. alone makes a difference.


Negative Control: BHI Media

BHI Media in a mixing device chip.

There appears to be one colony-forming unit, but otherwise is a successful negative control.


Positive Control: E. coli RFP

We used the droplet generator design to attempt to culture the E. coli RFP strain. We did not have much luck:


Positive Control: Providencia alcalifaciens:

We cultured the Providencia alcalifaciens on its own in a splitting chip.

The Providencia alcalifaciens do not appear to be very successful on their own.


Positive Control: Staphylococcus epidermidis:

We tested the independent growth on-chip in a splitter chip.

They grow relatively well on-chip.


Positive control: Providencia stuartii:

The Providencia stuartii do not look incredibly healthy when grown alone on-chip. But they do seem to grow.

We tried to co-culture the strains in a T-junction chip.



The Providencia a. seems to have out-competed the E. coli. But we have growth from the chip!


Staphylococcus epidermidis/Providencia stuartii:

We tried to co-culture the strains in a smaller T-junction chip.

The Staph appears to have out-competed the Providencia s.


Staphylococcus epidermidis/Providencia alcalifaciens:

We co-cultured the strains in a mixer designed to combine four different solutions. However, we pre-mixed the strains instead of mixing them in-chip.

There seems to be co-culture of the two species. The plate on the left shows the two species growing quite nicely together, with easily observable phenotypic differences.



The E. coli with RFP did not seem to be very robust for in-chip growth. They seemed to mutate away their RFP gene (or perhaps promoter). The Providencia alcalifaciens and Staphylococcus epidermidis grew spectacularly well together in-chip. The Providencia alcalifaciens, in fact, grew only when paired with the Staph. Further research may give us more insight into why this happened. At this point, we can only hypothesize that the species may have supplied one another with complimentary metabolites that supported each others’ growth.


Future Work:

We would love to investigate the co-culture further and conduct experiments testing the optimal chips for different microbes. We could also test which part of our procedure helped us to achieve growth, whether it was the microbes themselves or the treatment of the parts.



DAN K CHEN © 2024