Re: DIY protein purification on a microfluidic CD for CDROM drives
Bryan Bishop
> On Wed, Jul 15, 2009 at 12:55 AM, Bryan Bishop wrote:
>> "This work demonstrates the development of microfluidic compact discs
>> (CDs) for protein purification and fractionation integrating a series
>> of microfluidic features, such as microreservoirs, microchannels, and
>> microfluidic fractionators.
>
> ... so, what, you laser etch a CD, stick it in a drive and control the
> speed of spin to purify proteins with microfluidics? Shit, I'm sold.
> There are so many busted CD/DVD drives with working motors out there
> that I can't buy them, people have to pay *me* to take them away.
In this case, no, they casted their own PDMS design by using an SU-8
master mold after using some sort of photolithography method (with the
SU-8). Then you spill or spin PDMS over SU-8, polymerize it, then peel
it off. But the SU-8 step can be tricky if you can't do
photolithography. There's probably a way to make it work with laser
cutting, however. Laser cutting is probably overkill. In the paper
that I linked to, you have to do this extra step where you polymerize
MPNS (or something) in a microwave oven for three minutes, which makes
up a part that you run your solutions through. You also have to hook
it up to a pump so that you can make sure things pass through it.
Anyway, be sure to read the paper.
>> The CDs were fabricated with polydimethylsiloxane (PDMS)
>
> Is a CD fabricated with PDMS a commodity item? Alternately, how are
PDMS is considered the easy way to do microfluidics in the literature,
all labs should have this sort of capacity, yeah.
> they made? From some cursory googling it sounds like the PDMS is
> actually a photomask layer; would it work to etch directly into the
I recommend you read through some of the papers here:
http://adl.serveftp.org/papers/microfluidics/
> polycarbonate of a CD, or is there something special about the PDMS
Etching a CD would work, yes, but then consider how you are going to
make layers or something, without liquid leaking, etc.
> that makes it extra-specially useful for microfluidics?
What?
> Semi-relatedly, anybody know off-hand how narrow a channel TechShop's
> laser can cut? How shallow?
Are they using the fablab epilog laser?
Bryan Bishop
> On Tue, Jul 14, 2009 at 6:34 PM, Meredith L. Patterson wrote:
>> actually a photomask layer; would it work to etch directly into the
>
> I recommend you read through some of the papers here:
>
> http://adl.serveftp.org/papers/microfluidics/
>
>> polycarbonate of a CD, or is there something special about the PDMS
>
> What?
Oh, right. So, yeah, there are some properties of PDMS that make it
great to use. For one, it's not a polymer that polymerizes in the
presence of water, so routing water through PDMS microfluidic circuits
is okay. I forget whether or not there are commonly problems with
biological agents running through it- I suspect not, and that's why
people like it so much. Things kept nice and clean, not clogged, etc.
But this deserves some checking in the literature.
Bryan Bishop
> On Wed, Jul 15, 2009 at 1:45 AM, Bryan Bishop wrote:
>> In this case, no, they casted their own PDMS design by using an SU-8
>> master mold after using some sort of photolithography method (with the
>> SU-8).
>
That's easy to say when you have a laser cutter laying around. :-) But
congrats on the access to the laser. Blow me up some planets in my
name or something, please.
> Besides, who do you know with a cleanroom? I don't see one listed on
I know a few guys. Get rid of the carpet.
> the TechShop assets page. A laser cutter is ready to hand (it's an
I wonder if any fablabs have a cleanroom. I'll ask around.
> Epilog Helix 45, I don't know the specs on that, I know a guy who's
> used the one at TechShop and Marnia knows a guy who teaches the class
> there), I for one would like to see if laser cutting a plain ol' CD
> will do the job.
The Austin Robot Group had a fellow (Paul) who bought a Quantum
FireBall 9000 (or something), a desktop CNC machine and got
linuxcnc/emc running on it. At the Maker Faire last year, he demoed it
by cutting patterns that people walked by and drew in some app, and
then the desktop router cut it out. I'm still confused as to how
etching patterns would wrok, since you wouldn't have a top.
I think a good experiment would be to see how the hell saran wrap
stretched over a CD fares. Is that an effective covering? Someone on
the list could try by cutting some saran wrap out, stretching it over
a CD, taping it at the edges, and then seeing if it still spins in a
drive, or if it gets locked because of thickness issues, and possible
ways to work around that.
>> I recommend you read through some of the papers here:
>
> Bryan, honey, I think you forget the part where I'm a corporate CTO
> restarting her PhD while working on six different open-source projects
You also haven't ever told me, so that's probably why I don't know.
What company are you being a CTO at? I thought you quit IDT.
> at once and trying to deal with terrible RSI. I don't have a laser
Well, you've got me there, I don't have RSI.
> cutter, the folks in SF do (or will soon) and I don't know much about
That's why I am skeptical as to how "diy" a laser cutter is. I mean,
it's great to have one laying around, use it if you have it, but the
majority of us aren't going to
just-so-happen-have-a-laser-cutter-in-the-garage-out-back.
> microfluidics; I'm picking stuff up as I go along but I do not have
> time to *sleep* much less shift everything I'm working on down a few
> days so I can digest enough about microfluidics to understand what's
> going on in those papers.
>
> Shit happens. Not everyone is you. Fucking learn to deal with it already.
Where the hell does that attitude come from? Read them if you want, or
don't. What's the problem?
>>> polycarbonate of a CD, or is there something special about the PDMS
>>
>> Etching a CD would work, yes, but then consider how you are going to
>> make layers or something, without liquid leaking, etc.
>
> Okay, is single-layer microfluidics useful for anything?
The electrowetting-on-dielectrics that Jonathan has been playing
around with is single-layer in some cases. It requires some fairly
high voltage power supplies. I'm having trouble recalling other
single-layer or atmosphere-exposed applications.
>>> that makes it extra-specially useful for microfluidics?
>>
>> What?
>
> I dunno, y'all were talking about hydrogen donors and acceptors in
> some microfluidics thread earlier, I don't really have the chemistry
That was about nylon fibers and columns for chromatography, btw. Not
necessarily microfluidics.
> to grasp it from first principles. I really do know nothing whatsoever
> about microfluidics, which is why this is the first time I've spoken
> up on a thread about it.
Okay.
> --mlp, evidently punchy tonight
Meh.
Bryan Bishop
> On Wed, Jul 15, 2009 at 2:22 AM, Bryan Bishop wrote:
>> I wonder if any fablabs have a cleanroom. I'll ask around.
>
> else. Len's been wanting to get into the COSIC cleanroom to recover a
> crashed disk for I dunno how long now.
I thought hard drives were also under a vacuum, not just ultra-clean.
>>> Shit happens. Not everyone is you. Fucking learn to deal with it already.
>>
>> Where the hell does that attitude come from? Read them if you want, or
>> don't. What's the problem?
>
> There are a lot of people on this list who have been frustrated for a
> long time about the fact that your response to basically every request
> for information is "read the archive, read the archive". It's gotten
Because it sounds like nobody bothers to remember anything at all..
it's like we go through the same thing over and over again.
> to the point where some folks are just reading digest, or not reading
> at all anymore, because productive discussion has fallen off -- in
> large part because a huge proportion of the list's content is either
> stuff you're forwarding or arguments stemming from you being upset
> about people not doing things your way.
I'm upset that they don't even bother to try open source tools for an
open source project- so shoot me. What a crime!
> it. People are here because they learn through the process of
> dialogue. I certainly learn better from talking shop with people who
Top-posting really inhibits dialog, if you haven't noticed.
> already know their way around a lab than I do from an academic paper
> which presupposes that I already have a lot of background knowledge
Same here.
> which I frankly *don't* have. I wish I had time to soak it all up. But
> I don't.
You already know about the tools that I talk about anyway, I don't
think you need to play catch up.
> As an alternative to just posting a big list of links every time
> someone has a question, have you considered including a quick summary,
> in your own words, of what is useful about a particular paper? I think
Yep, I used to do that quite often. But then everybody was
complaining, and didn't even bother reading anything, etc. So, look,
you guys need to make up your minds.
> that would go a long way toward promoting discussion among people who
> are too overworked, or too busy with projects of their own, or too
> unfamiliar with the territory, to shift gears and tackle a
diybio is *hard*. We're here to make it easier. There's no way around this.
> three-inch-high stack of papers every time they have a question they
They should at least have the decent courtesy to remember the last
time the topic came up. What's the point of discussing if we all
maintain the memory retention rate of a fish?
> want to explore. I already have a foot-high stack of papers that I
> need to read for work, and it grows every day. "Here, take six hours
> out of your day and you'll find the answer" is a hell of a turn-off.
I don't think it should take six hours a day to do something.
> "The basic answer to your question is $(foo), it's explained in more
> detail in $(bar)" works much, much better.
Yep.
> Anyway, this is why people have been snarking at you for the last
> three months. Snark annoys me. I'd rather have it out and get it the
> hell over with. You people wouldn't like me when I'm annoyed.
It would be much easier if people sat down and *learned* something.
I'm sorry that it's so hard to open up google and look around for
tutorials, but diybio is hard (right now)- it takes learning, and it's
not going to be something that you digest in a single email. I am
simply not that awesome of a teacher.