Non-plasma Bonding of PDMS for Inexpensive Fabrication of Microfluidic Devices



In this video we demonstrate how to use the neuron microfluidic device without plasma bonding.

Cite this Article

Copy Citation | Download Citations

Harris, J., Lee, H., Vahidi, B., Tu, C., Cribbs, D., Cotman, C., Jeon, N. L. Non-plasma Bonding of PDMS for Inexpensive Fabrication of Microfluidic Devices. J. Vis. Exp. (9), e410, doi:10.3791/410 (2007).


In this video, we demonstrate how to use the neuron microfluidic device without plasma bonding. In some cases it may be desirable to reversibly bond devices to the Corning No. 1 cover glass. This could be due, perhaps, to a plasma cleaner not being available. In other instances, it may be desirable to remove the device from the glass after the culturing of neurons for certain types of microscopy or for immunostaining, though it is not necessary to remove the device for immunostaining since the neurons can be stained in the device. Some researchers, however, still prefer to remove the device. In this case, reversible bonding of the device to the cover glass makes that possible. There are some disadvantages to non-plasma bonding of the devices in that not as tight of a seal is formed. In some cases axons may grow under the grooves rather than through them. Also, because the glass and PDMS are hydrophobic, liquids do not readily enter the device making it necessary at times to force media and other reagents into the device. Liquids will enter the device via capillary action, but it takes significantly longer as compared to devices that have been plasma bonded. The plasma cleaner creates temporary hydrophilic charges on the glass and device that facilitate the flow of liquids through the device after bonding within seconds. For non-plasma bound devices, liquid flow through the devices takes several minutes. It is also important to note that the devices to be used with non-plasma bonding need to be sterilized first, whereas plasma treated devices do not need to be sterilized prior to use because the plasma cleaner will sterilize them.


Preparing Microfluidic Devices For Neuron Cell Culture

1) Clean Corning No.1 24 mm X 40 mm glass slides

  • Sonicate the glass slides in a water bath sonicator for 30 minutes
  • Rinse the glass slides with 70% EtOH
  • Wash slides three times with dH2O
  • Dry slides in a tissue culture hood for several hours, preferably overnight

Note: We have special metal trays that we load with the slides. The slides are separated individually and placed in slots in this metal rack. We can then place these metal loading trays in a glass dish that we fill with water, place in the water bath sonicator, and sonicate for 30 minutes. Subsequent washes are carried out in this dish.

2) Preparing the PDMS devices

  • Cut out the PDMS mold from the silicon wafer, punch holes in the PDSM cast and quarter it
  • Clean the PDMS devices by first blowing inert gas (Argon or Nitrogen) across them. Then use 3M Scotch Brand 471 tape to lift off any remaining debris

Note: It is important to keep the devices face up. Initially, when we cut the PDMS away from the silicon wafer, the side in contact with the wafer is the clean side: it contains the microfluidic channels. We try to be as careful as we can as to not damage the device, and keep the clean side face up until we are ready for plasma bonding.

  • Autoclave devices in plastic containers
  • After autoclaving, clean No. 1 Corning glass slides and plasma-treat devices using a Harrick brand plasma cleaner,
  • Place the devices clean side down on the glass slide.

The device is now plasma bonded to the glass slide.

3) Coating The Devices With Poly L-Lysine (PLL)

  • PLL is added to a well on each side of the device and allowed to flow through into the next well

=> 150 µl of PLL for the large wells and 30 µl PLL for the smaller wells. It does not have to be exact as long as the wells are full.

Note: If you look at a device you will see 4 wells: each two are connected. You want to put the PLL in one well so that it will flow through the device into the next well.

  • Allow the PLL to flow through the device for about 10 minutes and then add more PLL to the wells to fill them up
  • Place the devices containing PLL in an incubator at 37°C for a minimum of 4 hours (overnight is preferable)
  • Vacuum out the excess PLL after treatment, but be careful not to suck all the liquid out of the device

Note: You do not want to introduce air bubbles to the channels or chamber. Just vacuum out the excess PLL in the wells.

  • Add autoclaved dH2O to each well on either side of the device, and allow to flow to the other well by capillary action

Figure 1 is a diagram of a microfluidic device. Notice how the Blue (Soma side) is connected and the Red (Axonal side) are connected. When we say put the PLL, or water or media, on one well on each side of the device, what we mean, for example, is: Place 150 ul of autoclaved dH2O in one Blue Well then place 150 µl of autoclaved dH20 into one Red Well. The water (or PLL, or Media, or Cells) will flow through the device to the other corresponding well.

Figure 1

  Figure 1

Note: Please note that, from now on, when I say "place media, cells, water or PLL in the TOP wells", I am referring to the diagram above where the Soma would be on the left and the Axons will grow to the right.

  • Aspirate off the water (again being careful not to fully remove all the liquid from the device), then add another 150 µl of autoclaved dH2O to one of each connected well and allow it to flow through to the corresponding well.
  • Place the device containing dH2O in an incubator at 37°C for 1 hour, then repeat the wash step again. Wash the devices three times with dH2O for one hour each, which is the total.

Note: Due to the possibility that borate from the PLL can absorb into the PDMS, some in the Jeon Lab recommend incubating the devices overnight with autoclaved dH2O after a couple of initial quick rinses. This will ensure that all free borate will leach out of the PDMS prior to the loading of cells. If this is not done, there may be a release of borate into the media over time, which could lead to cytotoxicity.

  • Add Neural Basal Media (NBM) with the necessary factors (glutamax, B27, PenStrep) to the top wells

Note: The devices can be incubated overnight and plated with cells the next day, or incubated a minimum of 3 hours with NMB + factors before plating the cells.

Preparing The Neurons For Loading Into The Devices

We use 18 day old fetal rat cortex that we either buy from a company called Brain Bits or that is prepared here on campus for us. We prefer to get the tissue fresh.

  1. Obtain 1 fetal brain cortex (2 pieces of tissue) stored in Hibernate E
  2. Take 3 long glass pipettes and fire them each into successively smaller sizes using an alcohol lamp
  3. remove tissues from Hibernate E with a glass pipette and place in 2 ml of NMB + Factors in a sterile 15 ml tube.
  4. Trituration: Using a bulb and glass pipette the cortex is passed up and down about 5 to 10 times. Then, the next smaller pipette is used to pipette up and down the tissue. Finally, the smallest pipette is used to pipette up and down the tissue. We like to make sure that there are no visible chunks.

    Note: Recently, the Jeon Lab began comparing cells from tissue stored in Hibernate E to cells that were prepared from Tissue initially treated with 0.125% Trypsin in 50% Ca++ free and Mg++ free dissection buffer. The consensus is that tissue prepared using trypsin yield more viable cells than tissue placed initially in Hibernate E. If you are NOT going to use the tissue right away, then you need to store the tissue in Hibernate E.

    If you are going to use the tissue right away and would like increased viability, then treat the tissue with Trypsin prior to mechanical trituration as following: 1) dilute 1 ml 0.25% Trypsin (Gibco, Invitrogen) with 1 ml of Ca++ free Mg++ free dissection buffer (final Trypsin = 0.125%) into a 15 ml tube (keep ice cold); 2) place tissue in buffer and incubate immediately at 37°C for 8 minutes; 3) add 10 ml DMEM/10% FBS to stop the Trypsin reaction and continue protocol below.

  5. Centrifuge cells at 1100 rpm for 1 minute
  6. Aspirated media carefully off the cell pellet
  7. Add 1 ml of NMB + factors to the pellet and re-suspended it by pipetting up and down
  8. Pass the re-suspended cells by gravity flow through a filter to remove any clumps (BD Falcon cell strainer 40 um nylon)
  9. Count and plate cells:
    • Into a microfuge tube, add 60 µl NMB + factors, 20 µl of cell suspension, and 20 µl of trypan blue then mix
    • Add about 10 µl of this solution to a hemocytometer and count the live cells (not blue)
    • Adjust concentration to 2.5 - 4.5 x106 cells/ml

      Note: We generally obtain a concentration of about 2.5 - 4.5 x 106 cells/ml. depending on the volume the cells are resuspended in (normally 1 ml NBM + B27/Glutamax/PenStrep percortex). If the tissue was prepared with trypsin, the yield will likely be increased.

Plating The Devices

  1. Load cells (primary neurons) into one well of the device
  2. Plate 5 µl per small device and 20 µl per big device

    Note: You can load 10 µl of cells on the top of the device and 10 µl on the bottom of the device (half total volume), or directly all 20 µl of cells in the top somal well (5 µl in the top somal well for the small devices).

  3. Incubate for 10 minutes in a 37°C incubator so cells can start to attach
  4. Add approximately 150/30 µl of NMB + factors to each well depending on device size

    Note: Volumes may vary depending on the thickness of the PDMS. All that matters is that the wells are full.

Supplemental On Using The Devices Without Plasma Bonding

Without plasma treating, Christina Tu, a technician, who does it successfully every week without plasma, says to just load the cells right away.

Remember to remove the media from both wells on the soma side (axon side won't matter if you leave the media in). It's the fluid flow that allows the cells to enter the main channel. If you have media in the wells, you won't get fluid flow.


Here we demonstrated how to use the neuron microfluidic device without needing a plasma cleaner. We refer to this as non-plasma bonding.


The authors have nothing to disclose.


Name Type Company Catalog Number Comments
PDMS Reagent Dow Corning Sylgard 184 with curing agent Please consult Dow Corning to find a vendor near you
Cornig No. 1 Cover Glass Cover Glass Corning Corning No. X2935 244 Available through Fisher Scientific, Fish Catalog number 12-531D
B27 Reagent Invitrogen 17504-044 B27 is a proprietary supplement available through Invitrogen under their Gibco line of cell culture reagents.
Glutamax Reagent Invitrogen 35050-061 Glutamax is available through Invitrogen under their Gibco line of celll culture reagents.
60 mm Petri Dish Tool Fisher Scientific 08-757-13A 60 mm polystyrene sterile petri dishes are used to house the device bound to glass.
Poly-L-Lysine Reagent Sigma-Aldrich P-1274
BD Falcon 50 ml Tube Tool BD Biosciences Falcon No. 352098 Available through Fisher Scientific catalog number 14-959-49A
BD Falcon 15 ml tube Tool BD Biosciences Falcon No. 352097 Available through Fisher Scientific catalog number 14-959-70C



  1. Park, J. W., Vahidi, B., Taylor, A. M., Rhee, S. W., Jeon, N. L. Microfluidic culture platform for neuroscience research. Nat Protoc. 1, (4), 2128-2136 (2006).
  2. Taylor, A. M., Blurton-Jones, M., Rhee, S. W., Cribbs, D. H., Cotman, C. W., Jeon, N. L. A microfluidic culture platform for CNS axonal injury, regeneration and transport. Nat Methods. 2, (8), 599-605 (2005).



  1. can't view videos?? firefox latest plugins???

    Posted by: Anonymous
    November 4, 2007 - 10:33 PM
  2. Hi Ron. I am sorry you are having trouble. To view videos, please make sure that you have the latest version of Flash installed. If you still have problems, please send us an email to with operating system and browser information.

    Posted by: Anonymous
    November 5, 2007 - 12:48 AM
  3. What protocol do you use when making the PLL solution? 5mg of PLL to 50ml water?

    Posted by: Anonymous
    April 7, 2008 - 2:47 PM
  4. Hi Lee T
    We dissolve the PLL in borate buffer
    1.²4 g of boric acid and 1.9 g of sodium tetraborate is added to 300 ml of nanopure water (final volume will be 400 ml) and stirred to disolve for about 30 minutes.  Next the pH is adjusted to pH 8.5.  ²00 mg of PLL is added and stirred until disolved.  The final volume is adjusted to 400 ml.  The PLL solution is then sterile filtered and aliquoted into sterile 50 ml tubes in a tissure cutlure hood.  Aliquots are stored at -²0C until use.
    Joseph Harris

    Posted by: Anonymous
    May 7, 2008 - 5:57 PM
  5. Just wanted to let everyone know a company is forming to make, sell, and distribute the devices. Email for more information.

    Posted by: Anonymous
    July 4, 2008 - 1:01 PM
  6. Hi there, it it possible to get a protocol for fixation and staining of neurons in microfluidic chambers? Which method is better for this purpose (fixation and staining), plasma- or non-plasma bonding? Is it possible to remove the microfluidic device without damaging the neurons or dŒs it have to stay on during the procedure? By the way: A really impressive method, fantastic! Looking forward to your answer. Best, Harry

    Posted by: Anonymous
    November 6, 2008 - 2:32 PM
  7. Hello Harry Probably the best thing would be for you to check out our company website at We have a protocol there available for download.  It gŒs over staining in the device. Also, if you would like to discuss staining in more detail or have any other questions feel free to email us

    Posted by: Anonymous
    November 10, 2008 - 4:46 AM
  8. Hi Harry just to follow up.. The device dŒs not have to be removed for staining. However, if you are interested in visualizing in the mid-region of the grooves then it might not stain very well with the device on.  If you only want to visualize neurons in the main channels or axons just entering or leaving the grooves then staining with the device on is just fine. If you want to get a good stain of the middle of the microgroove region then it is best to remove the device before staining.  This means non-plasma bonding will have to be done as plasma bonding is irreversible. Yes, removing the device can and dŒs damage the neurons.  However, done with care and after some practice it is possible to remove the device and keep most of the sample intact. Some researchers like to fix the cells in the device, place the device on ice for 5 minutes then carefully remove it.  Others just remove the device as carefully as possible prior to staining. Please visit our website for our detailed protocol. Also feel free to email us

    Posted by: Anonymous
    November 10, 2008 - 4:55 AM
  9. Hello JŒ, thank you very much for your reply and support. The protocol on which you are refering to is indeed very detailed and helpful. I am now curious to see how my experiments work out. :-) All the best and thanks again, Harry

    Posted by: Anonymous
    November 11, 2008 - 5:59 PM
  10. This probably dŒsn't work for non-plasma PDMS-PDMS bonding... or, is there a way to treat one of the PDMS layers as glass?

    Posted by: Bryn G.
    September 23, 2009 - 8:05 PM
  11. Hello,

    This was a very good video indeed. Could you please let me know how did you make the wells on the PDMS. They are not micronsized wells that we used to make by replicating form a potolithographed Silicon wafer. Did you use some kind of belt puncher? What is the exact product number and brand of the puncher? I use one to make mm sized well on the PDMS slam, however the edges are not good. Or did you use other technique to make such nicely sheped wells? If I want to make recangular or square shaped wells (² mm length/width and .5-1.0 mm height) on PDMS what could I do? I will appreciate your kind advices.

    Posted by: Anonymous
    March 31, 2012 - 6:39 PM
  12. Hello Taifur

    The wells are 8 mm in diameter. We have special punches machined to punch the wells. However, you can use 8 mm biopsy punches made by Sklar. VWR or Fisher Sci should have them.

    Posted by: Joseph H.
    April 6, 2012 - 12:48 PM
  13. For the cells you receive in Hibernate E from Brain Bits, do you digest them with trypsin or even digest them at all prior to trituration. Why or why not?


    Posted by: Angela D.
    July 4, 2012 - 10:08 PM
  14. For the cells you receive in Hibernate E from Brain Bits, do you digest them with trypsin or even digest them at all prior to trituration. Why or why not?


    Posted by: Angela D.
    July 4, 2012 - 10:10 PM

Post a Question / Comment / Request

You must be signed in to post a comment. Please or create an account.

Usage Statistics