Episode #82 | Advancing Microfluidic Technology Through 3D Printing (Virtual Event Recording) - The Lattice (Official 3DHEALS Podcast)

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 2 (00:02):
Good morning everybody.
I'm going to let people totrickle in.
Thanks for joining us.
I know it's very early in themorning on the West Coast and
sometimes in the middle of thenight across the globe, so thank
you very much for joining us.
We have quite a large audiencetoday.
Hopefully, we will deliver theknowledge that you're looking
for and also a very engaging andexciting conversation among the

(00:24):
speakers and you.
My name is Jenny Chen, founderand CEO of 3D Heels.
We're a small media company andwe have three missions.
One is to educate everybody.
10 years ago, when I startedgot interested in 3D printing, I
realized my understanding of 3Dprinting was very shallow and I

(00:45):
wanted to learn more, and thenrealized everybody else's
knowledge is equally shallow.
So it's time to learn aboutthis technology, and it is
constantly evolving.
We have new things coming up,new founders, new technologies
invented every day, and I thinkit's a great platform here for
us to learn about them.
Number two is networking.
We used to host a lot ofin-person events.

(01:07):
In fact, in 2018, I justrealized we hosted 38 in-person
events all over the world withthe help of community managers,
but that was the glorious daybefore COVID.
And after COVID we're almostentirely virtual, although
occasionally we do organizein-person events as well.
However, the way that you cannetwork here on the virtual

(01:30):
platform is to input your sociallink in the chat and obviously
you can tell people who you are,your name, where you're from,
and typically we have 30 to 40countries around the world who
are either watching this live oron demand, so it's quite a
vibrant community every time.
So do that.

(01:50):
And number three is we have aprogram called Pitch3D.
We help early stage startup,which means up to series A, with
fundraising, and this programis entirely free to you.
I spend my free time to do this, so if you're in that category,
feel free to reach out to me,and we have I can send share a

(02:12):
link for you to to understandwhat that program is about.
But without further ado, ourtopic today is quite popular and
we host this once a year, everyyear.
I think that's a redundantdescription, but anyhow, you
know, I know about microfluidicsat the same time as I knew, or

(02:35):
started to learn about 3Dprinting about a decade ago, and
I learned.
I actually went to Dr AliKudhasani's lab I don't know if
you know him and he was showingme how he makes these
microfluidics chips himself orhis team in the lab, and it's
quite an elaborate process, alot of arts involved, and he's

(02:57):
telling me that 3D printing,while very ideal in theory to
make a chip, it's reallychallenging and we were just not
there.
And after all these years, itis really pleasant to see that
we now have more options to dothis with 3D printing, and
everybody is recognizing thepower.
So the first speaker today isalso Hamdeep Patel, and he's

(03:23):
also the founder CEO ofCatWorks3D, who is also the
sponsor for this event.
So thank you very much for thesupport.
Without you guys, thesupporters, we could not host
this on a regular basis.
So thank you very much, butwithout further ado, please
share with us what you'reworking on.

Speaker 1 (03:42):
Yeah, thank you very much, Jenny, for that
introduction and let me justlaunch my presentation here.
I am, my name is Hemdy Patel.
As Jenny said, I am the cofounder and CEO of Cadwork City.
We are a 3d printing companyand solution provider for

(04:04):
microfluidics.
Let me, I guess I can touch onwhat Jenny had just said.
In terms of the availabletechnology for microfluidics,
there have been quite a numberof them and all of them had a
very unique set of deliverableswhich was very, very important
for researchers around the world.
There was the soft lithographythat Jenny had spoken about, the

(04:28):
CNC milling and polyjet.
There's injection molding.
All of them had a really strongset of deliverables for
microfluidics at the microscale.
And then, around eight to 10years ago, 3d printing made its
way into microfluidics and atthat time what the technology
was was that printers wereproviding solutions for macro

(04:51):
scale printing and then theywere trying to adopt or modify
that solution in order to reachthe microfluidic or micro scale
that is needed for microfluidicsNow, right across the platform.
All of these technologies havetheir barriers or at least their
pain points.
In terms of the first threethat were traditionally used,

(05:13):
pain points usually meant messyor very difficult technology to
work with.
It was expensive or the lengthof time.
When it came to 3D printing atleast the first and some of the
more commercial availableprinting you're noticing that
there is pixelation, there isn'ta clear-cut device strategy,

(05:35):
there are issues in terms ofsurface roughness, and those are
the things that we're trying toaddress with our platform.
To address with our platform,3d printing, I think as everyone
now has become aware is thefact that it does allow users,
researchers and engineers to gofrom concept to

(05:56):
commercialization a lot morefaster.
This iteration and design cyclegets a lot more quicker.
In most cases, you're able tofind iteration cycles are
speeding up this design processby nearly 80% or more.
In terms of our company, and thefocus of what we're doing right
now is that we've sort oflooked at microfolios in these

(06:19):
two unique areas.
One is the need to be able tocast PDMS, and we've got a
master mold material that usersare able to print and it's as
simple as pour, print and castIn this case in terms of the
cleaning stage.

(06:40):
Really, these molds are readyto use right off the printer,
where you don't need any releaseagents or any sort of
post-processing of the actualdevice itself other than being
able to cast Eupedemus or anylike a silicone or agar-gar

(07:01):
right on the actual mold itself.
The second batch of users thatwe have is those that are
looking at printing outmonolithic devices.
Again, the process is quitesimple using our technology,
pour, print and clean, and thiscleaning process is as simple as
it's noted.

(07:21):
You clean it in IPA compressedair.
You don't need to polish yoursurfaces, you don't need to coat
it.
It is optically clear directlyout of the printer itself.
We are located in Canada and ourclient base is worldwide.
We've got research teams aroundthe world that are using our

(07:45):
platform, along with Big Pharmaand government agencies and
really the thread that connectsall of these entities together.
In terms of what they werelooking for, they were looking
for a complete 3D printingsystem that was optimized for
microfluidics, materials thatwere made specifically for

(08:07):
microfluidics itself, and also amethodology that was made
specifically for microfluidicsand delivering the needs of
microfluidic users.
We were aiming to take them outof the clean room, take them
out of expensive and long drawnout processes and then do

(08:28):
everything at bench side.
The one key fact and this isthe overarching mandate and the
mission of our company is toactually focus on deliverables
undebeatable deliverables thatour users expect when they're
using traditional methods.
They are wanting to be able toduplicate that on a 3D platform,

(08:52):
and that's what we are aimingto do with our platform.
So if we were to do just a verysimple cross-comparison between
what you design and what you'regetting, if you take a simple
Y-channel design here, thisY-channel has a series of
herringbone patterns down thecenter and if you were to print

(09:15):
it on a commercially availableprinter using high-resolution,
commercially available materialand then compare that with our
platform that has been optimizedfor microfluidics.
This is what you get, sovisually, when you expect,
inspect it, you don't see muchdifferent.
It seemed to be very similar towhat your CAD file it's.
It's when you sort of starttaking a much more closer

(09:38):
examination of your printed fileand take a look at what those
herringbone patterns do looklike Now off of a commercial
printer.
As you can see, this is whereit starts to.
The issues are very prevalent.
You're starting to seepixelation, this sort of boxing
pattern, as it's trying to makethe actual herringbone, whereas

(10:02):
our device is now reallyoutlining the intended design of
the researcher.
If we take a further look atanother feature and this is the
Y junction, that happens again,that same pixelization that you
can see on the channel walls,and this also impacts everything

(10:22):
down the line.
So if you're using this as yourmaster mold and you're going to
cast PDMS, this defect nowtranslates onto PDMS and so,
effectively, you end up with aPDMS that doesn't do or function
the way that you would expect.
Another type of file that wouldbe expected by most users is
this simple straight line withserpentine.

(10:43):
Again, if we do just a verysimple comparison, visual
comparison, they look fine.
It's when we start scanningunder a digital microscope and
that's where the differencesbecome very apparent.
At 50 microns you are startingto test the limits of the
material and the machine itself,whereas on our platform you can

(11:06):
still see a very clean straightedge on this raised channel and
really the big comparisonhappens in the serpentine.
These are features that arevery difficult to reproduce on a
printer.
If you look at the commerciallyavailable printer with their
material set, you can startnoticing that pixelization, or

(11:30):
the noise that you're going toadd to your design, is now very
much prominent there, whereas ifyou look at our platform, it
has a very distinct channelshape and the rounding of the
corners are very distinct aswell.
If we take a look at our clearmaterial, this has become one of

(11:52):
our strongest materials outthere because there are a lot of
users who are trying to moveaway from PDMS onto a clear
material and we use a verysimple channel structure 80
microns wide, with three portscommercially available printer.
There's always a distinctyellowing of the device and
there's a frosted finish on thesurface, whereas using our

(12:16):
platform with the clearmicrofluidic material, you've
got a very clear, distinct,transparent material and then,
if we take a real closer look atthe actual channel, you're off
of the commercially available 80microns.
It was not, it didn't come outat all, whereas on our platform,

(12:39):
you're able to form thischannel.
To form this channel.
Now the question that I alwaysget at this point of any
conversation is what's thedifference between commercially
available printers and what isit that you have you done on
your side that allows you to doand deliver these type of
features and products?
And that is we knew right offthe get-go is the fact that if

(13:05):
you're going to be printingmicrofluidic parts or devices or
features, you had to reallydeconstruct your printing
functionality down to this mostbasic part, and that was
actually each layer that wasprinted.
So you're now needing tooptimize each layer that's
printed on your device and thenthat is the foundation of the

(13:26):
next layer that's actuallyprinted.
So if we take a look at what acommercially available printer
with high resolution materials,this is what the surface
actually looks like.
This is a typical pixelationsort of checkerboard pattern
that you see.
If you compare it to ours, yousee a much more smoother finish

(13:47):
and just on the visuals you seethat there is a significant
difference between the two andthis is more of a qualitative
analysis.
But if we're going to do a muchmore quantitative analysis and
do a surface roughness analysisbetween the two, this is what
you get off of a commercial 3Dprinter.

(14:08):
The pattern that you see, theseries of troughs and peaks,
actually align with the actualpixelization of the checkerboard
pattern that you see there.
The same analysis.
When you do it off of ourplatform, you see a much more
smoother finish.
Off of our platform, you see amuch more smoother finish and if
you were going to look atnumbers, you would be looking at

(14:28):
a roughness and RA value ofanywhere from 1.5 to 11.5
microns for a commerciallyavailable 3D printer, whereas on
our platform we've been able toreduce that down to 0.18
microns.
So that's an improvement ofnearly 85%.

(14:51):
So, when it comes to our techand what we want to deliver,
there are some very simplethings that we want to deliver.
We want to help researchers getout of the clean room and bring
out most of their research, orat least their iteration cycle
at BentSight.
We want to ensure that thesetup is very easy.
We want to bring down theirfabrication time from weeks and

(15:14):
months down to a single day.
Since we manufacture our ownmaterials and reformulated
ourselves, we're able to monitorthe type of raw materials and
raw products that we're startingwith, so to ensure that for
enclosed spaces where we're notgoing to be, none of our
materials have odors in it.
The fact that you're using a 3Dprinter allows you to fully

(15:39):
customize your devices in alldifferent areas and, lastly, the
machine is fully open source,so it is open to any third-party
material without the need forany licensing.
We are more than happy to helpresearchers or any teams you
know set up profiles forthird-party materials as well,

(16:03):
and that sort of sets up whatthe machine and our setup has
been up until today.
In fact, we're proud to launchthat, and this is some of the
articles that were written.
I think I just jumped aheadthere.
Let me just jump ahead rightnow.
This is a launch of a newmaterial that we're going to be

(16:26):
bringing to our users on May the8th.
It is a material that's been inthe works for quite a while.
Take a look, thank you.

(17:20):
So some of the key features ofthis new material is that we
have had it validated over abroad spectrum of cell lines of
over 90% cell viability.
It is optically clear, very,very optically clear, to the
point where you can actually domicroscopic analysis directly on
chip itself.
And it is inert under UV light,so you can start doing your

(17:43):
analysis under fluorescent lightwithout any issues.
And this, as I said, has been amaterial in the works for quite
a while.
And here's some links that youguys can take a look at if you
can get updates, white paper onthis material.

(18:05):
If you want to book a meetingwith us and get more information
, you can scan this, and theselinks will be available after
the show as well, thank you.
Oh, before I go any further,the optical transparency is very
important to us.
The optical transparency isvery important to us, and so,
for this actual presentation,you've been viewing this right

(18:28):
through that chip that we haddesigned and that was on the
video.
So that is how opticallytransparent the actual material
is, and this is unpolished.
It hasn't been processed otherthan using IPA and a compressed
air.

Speaker 2 (18:45):
That is more clear than my glasses lens, to be
honest.
Thank you for showing us agreat, great presentation, md.
We have one question from theaudience, but I highly encourage
people who have questions toput it in the QA box.
I can't follow the chat.
Chat is for social and if youhave questions, please put them
in the QA box, and yourquestions are way better than

(19:07):
mine, trust me.
So we have one question fromLuciano.
I don't know, I feel like youalready kind of addressed it.
It says pixelated edges anduneven surfaces.
In that context, which approachor process would you recommend
when working with chips thatinvolves optics?

Speaker 1 (19:31):
That involves optics Like I can only base it on our
knowledge.
If you're using domed shapes, Iguess they would need to answer
is it domed shapes or is itflat structures, pixelations
you're going to find rightacross the board.
On our platform, we have beenable to minimize all of that, so

(19:52):
that would be like you'd haveto get our system in order to
minimize that type of stuff.
Cause this our system is set upwith, where we've optimized the
actual printer platform, thematerial and the methodology in
order to deliver what you'reable to see.

Speaker 2 (20:10):
Great, we have another question from Yan Liu.
He said I don't think you hadtalked about it.
He said what wavelength is theprinter?
Is it 405 or 385?

Speaker 1 (20:24):
Oh, I guess 3D printing in general is available
in a wide range of wavelengths.
Ours is set at 385.
We find that the availablephotopolymers are I guess our
CTO, who manufactures all of ourphotopolymers, loves 385.
And so we built our platform on385 nanometers.

(20:46):
Got it?

Speaker 2 (20:48):
Okay.
Julie asks were neural lineagecells viable on this material
Very specific?

Speaker 1 (20:55):
Very specific.
I think it might be somethingyou might want to sign up to our
white paper.
I don't think we did any neurallineage cells.
There were.
We did a variety of adherentcells and suspension cells.
I can't remember the exactdetails of which ones they were,
but while we're conversing I'llsee if I can pull up that

(21:20):
information for you and I canshare it.

Speaker 2 (21:22):
I mean.
Another quick question, sincethis is a more general question,
is how do you work with youguys if they're not sure if
their cells are going to workwith you?
I mean, how do you work withresearchers who are not sure if
your system is going to work forthem?

Speaker 1 (21:34):
Well, I guess the first thing that we always do is
that we ask that we first,during our first conversation,
we try to figure out exactlywhat they're doing and what
we're doing and if it isactually a match.
The next stage of that would betrying to find out if they're
doing cell lines, any particularcell lines and they want to
evaluate the material.
We're more than happy to printout a sample for them so that

(21:58):
they can evaluate it on theirside, and then we can proceed
and have a conversation furtherat that point, Great.

Speaker 2 (22:05):
Okay, let's take a look.
We have more and more questions.
This is like explosion ofquestions.
I'm not sure I can address allof them, guys, so hold on one
second, let me just take a look.
Okay, what is the smallestchannel, height and chamber you
can achieve?

Speaker 1 (22:22):
So right now we have.
So this trip here, we've gotthat one there, I believe is 114
wide by 150 tall.
We have done 60 microns wide at, I believe, 300 microns tall,
like.
The one thing that I guesspeople fail to understand is

(22:46):
that as you get smaller andsmaller, it isn't.
You're now trying to.
You're challenged by optics ofeverything.
So there's your light.
Your UV light is passingthrough a series of membranes
and glass and surfaces, whichcauses defects and diffractions

(23:07):
that will cause deviations onyour size.
You're also trying to optimizeyour materials in order to print
at that scale as well.
So it is a work in progress.
Right now, the newest materialthat we have, we know that we
can go down to a width of 60microns and if we're able to

(23:28):
share, in fact I do have thatdocument.
Yesterday I just got an emailfrom my….

Speaker 2 (23:34):
You can upload it as well, if you can't, I can upload
it as well, because we canupload PDFs for everybody, all
the speakers, if you have stuff.
So we're going to have to moveon, mainly because of time
concerns.
However, hemdeep, you canactually write your answer in
the Q&A box.
Just type answers for thesepeople.
So sorry guys, we have to moveon because we have only 90

(23:57):
minutes here.
All right, so we're going tomove on to our next speaker.
Thank you so much, hemdeep.
Excellent presentation, Ireally learned a lot.
Our next speaker Thank you somuch, hamdi.
Excellent presentation, Ireally learned a lot.
Our next speaker is PaulMarshall, ceo and, I guess,
founder or co-founder ofRapidFlex.
Hi, paul.
Hi Jenny From across the pondright.
Are you still in the UK?

Speaker 4 (24:19):
Yeah, we are based in Newcastle, northeast England.
Now, if I can share mypresentation, let's try that.
Does that come up okay?

Speaker 2 (24:35):
Yep.

Speaker 4 (24:36):
Fantastic.
Right, let me find the otherwindow, which is there.
It's not.
It's this one.
Bear with me, I'm on adifferent screen.
So, yeah, I'm Paul Marshall, I'mCEO and one of the founders of
Rapid Fluidics, based here inNewcastle, northeast England.
Unlike Cadworks, we are a userof 3D printers, so we provide

(24:58):
rapid prototypes ofmicrofluidics using a range of
3D printing systems, includingCadworks, and what I'm going to
do is explain the services thatwe offer, what we're trying to
do and how we do it.
So let's go through.
So, yeah, we providemicrofluidic prototypes.
The basis of the company camefrom my experience in a

(25:22):
diagnostics company developing apoint of care molecular
diagnostic system usingdisposable microfluidic
cartridge which, at volume, isgoing to be injection molded,
but for prototyping it wasexpensive, it was time consuming
, it was poor quality, whatwe're trying to do.
So we worked with the localuniversity here in Newcastle to
come up with a process for 3Dprinting microfluidic channels

(25:45):
suitable for PCR amplification.
Once that all was completed, Isaw the opportunity to take that
service to a wider audience,offering next-day delivery,
effectively, of bespokemicrofluidic components.
We also work with our customerson the design side of things.
We're not just a prototypingshop.
We have a team of very skilledengineers here who can work on

(26:08):
anything from the you know themicrofluidic side of things, the
mechanical design formanufacture up to the full
systems, electronics andsoftware as well, and then,
obviously, when it comes to thescale up, when you're taking
your device from prototypethrough to mass manufacturing,
you know there is a limit as towhere you can go to in terms of
quantities with additivemanufacturing.

(26:28):
So that's where we'll bring inpartners.
So we know a range of injectionmolding companies, for example,
around the world, and we'llmake those connections.
We can bring in the design formass manufacturing at an early
stage, at the prototyping stage,or we can partner up and share
the experience.
So our main capability asHemdeep just demonstrated there

(26:54):
is the ability to create fullyenclosed microfluidic devices.
Now we started off very much onthe more traditional, the more
traditional two and a half Dside of things, creating flat,
linear microfluidic systems.
We've expanded from that intothe more three-dimensional side
of things as well to really showthe benefits of 3D printing.

(27:15):
So that's got us into manifolds, anatomical geometry, things
like that.
We're now moving intoalternative manufacturing as
well.
We're looking at thermoplasticforming, so pressure forming and
hot embossing.
I'll show a few examples ofthose later on in the
presentation, partially as astepping stone towards the
injection molding so you cantransfer materials to

(27:36):
thermoplastics, and also as amore suitable batch production
side of things so we can turnaround higher quantities in a
quicker time than you can with3D printing.
And then also the embeddedelectronics.
This is something that's quiteneat that's come about over the
last year or so.
Adapting the process of printpause print to actually embed
different components into the 3Dprinted part Materials is

(28:03):
always an interesting one.
Yeah, absolutely love thatlittle reveal.
You just did hem deep.
That was wicked um.
With 3d printing in resins, um,it's not thermoplastics.
There's limitations.
You can't weld it, for instance.
It's a thermoset.
Most of it is methylacrylatebased resins, um, but we have
got a range, we've got access toan angel we all start with.
You know the basic modelingresins that you can buy on

(28:24):
amazon for 20 quid a bottle.
Um, and that works forsomething.
You can actually, if youcorrectly wash and cure these,
these materials, correctlythey're moderately by putting
these out through micro fluidchannels in standard resins.
You don't have to go for thebiocompatible materials, but you
know they'll glow like crazy onthe uv.
They're not totallybiocompatible.

(28:45):
There's definitely limitationsthere.
So we've got a range.
We can go from the the cheapresins up to the more expensive
materials.
We can go opaque, we can go totransparent um high temperature
resistances, for even the basicmaterials will cope with um pcr
type temperatures but we can goup to 170 or so with the
materials we've got in stock atthe moment.
Um.
We've been doing some prettycool stuff with flexible

(29:05):
materials.
Recently We've been workingclosely with Applied Molecules
over there in Boston with someof their new materials and we've
got some really nicesilicon-based materials we've
been developing with them.
At the moment, as I say,biocompatible, which is a bit of
a misnomer when it comes tothis particular application for
microfluidics, especially lessso for medical devices, the

(29:26):
biocompatible certification isrelated to human contact, so
wearables, implants and so onnot necessarily biocompatible at
a cellular level, but itprobably is and we can work with
, you know, with third partiesto verify that the materials are
suitable for particularapplications.
And then, obviously, wementioned integrated electronics

(29:46):
.
In the same way, we canintegrate flexible membrane, we
can integrate optical components.
Again, it's even this printableor print technique that we
sometimes use.
There we go.

Speaker 1 (29:59):
Capabilities.

Speaker 4 (30:00):
So this is a little out of date and not totally
accurate, if I'm totally honestwith you, but it's the question
everyone always asks how smallcan you go?
It depends on the material, itdepends on the printer.
As I say, we have a range ofprinters.
We have the CADworks, we have acouple of the Sega printers.
We've got the commerciallyavailable printers, as Henry
phrased it, then the SLA LEDprinters which allow us to go to

(30:23):
large format printing as well.
The DLPs, ultimately, are moreaccurate.
You can get feature sizes downto maybe two pixel sizes With
the LED machines.
You're looking four or fivepixels to get the decent
definition.
And it varies on the materials.
The low viscosity materials youcan get better definition.
The biocompatible materialstend to be higher viscosity, so
it's harder to get that size.

(30:45):
The footprint footprint thatwe've mentioned there, the 51 by
29, that's a bit of a limitingfactor.
That's our seagull max 27 umwhich has this footprint of 51
by 29.
That will go smaller.
So, as we say there, you knowwe can create with a
non-biocompatible material.
We can create a channel 100microns feature size at 70

(31:05):
microns we can actually gosmaller.
We've created channels 50 by 50microns, fully enclosed.
So 50 by 50 microns squaresection channel using the ASEGA
MAX27.
If it's an open channel, so ifwe're not enclosing it, we've
managed to create channels 50 by10 microns, so 10 microns deep,
50 microns wide.
Again, moving to thebiocompassable materials, you

(31:28):
have 200 or 300 microns minimumand that's really the more
comfortable side.
So we can go down to 50, butit's a hell of a challenge and
it depends on the complexity ofthe geometry.
We prefer to be slightly larger, but then we can go really
quite large.
We've recently done somemanifolds, created some
manifolds that were threemillimeters, with an array of
channels that were in the regionof one millimeter.

(31:51):
So much more on themillifluidic side of things.
But as an alternative to TNTmachining and bonding, it showed
the opportunities for ourparticular customer who needed
these large manifolds.
We were able to do them in amatter of days, whereas the lead
time on the machine passed forsix or eight weeks, so much
quicker than when we can do that.

(32:15):
One of the really cool thingswith 3D printing is you can
integrate other geometry.
So if you want external fluidconnections, so lower locks,
mini lowers you can integratestainless steel, tubing, barbs,
threads.
We can go down to M1.6, rathernot Again, m3 is a nice size,
m2.5 maybe yeah, m1.6 we havedone.

(32:38):
But the more standard M6 orquarter 28 fertilizers
absolutely fine can integratethat channel tube directly into
the part without any moreminimal secondary
post-processing.
We have what we call themodular microfluidic system.
This is a system that's on ourworkshop, which has nothing to
do with 3D printing really,other than the fact that we do

(32:59):
3D print the components thatform it, and it's an over-centre
locking clamp system to createa breadboard style system for
developing an instrument.
The intention is to startadding components heaters,
coolers, magnet actuators and soon, light sources and the like,
so you can build up aninstrument.
That's work in progress, but itthat allows us to connect to a

(33:22):
flat and open port progress, butit allows us to connect to a
flat and open port.
And then we get on to manifolds.
This is this isn't is quite aninteresting application of
microfluidics came from anexisting customer who asked us
if we could make a valvemanifold.
We'd never even thought of itbefore but we were able to, they

(33:44):
say, with the other to machineand bond a manifold.
You're often looking six weeksor so lead time.
We were able to make theseparts in a day or two and that's
opened up some hugeopportunities for us within the
fluid handling lab automationsector.
We've been doing a lot of workwith a variety of valve
companies.
The top picture there you cansee see is an optimized model

(34:08):
where we've basically taken allthe excess material out of the
manifold.
We were presented with a designwhich was to be originally
three layers of acrylic to bemachined, drilled, bonded
together, and, as I say, that'dbe six weeks or so, we were able
to print that in 11 hours.
I think it was to print that.
But then we took the wholedesign, optimised it, so we took

(34:28):
out all the excess material,made the most of 3D printing,
got rid of all the right anglecorners for the fluid tracks to
improve fluid flow, improve themanufacturability, and we got
that print time down from 11hours to two and a half hours.
And typically the quantitiesthat manifolds like this are
required in is a few hundred, isa few thousand.

(34:49):
So actually the batchproduction of 3d printing works
really really well here.
So when we looked at the costingof that, you know we could 3d
print the equivalent for prettymuch the same price as a
machined part but massivelyquicker, optimized it.
We reduced the cost.
So with the material we gotdown, um, we got 70% of the
material out of there.
We knocked 80% of the cost upto that, so over a quantity of

(35:11):
200 of those.
It works out a saving of£36,000 to 3D print them rather
than machining, which blew meaway when we worked that out,
and again a lot of interest inthat from the market there.
So yeah, minimize volumes,reduce weight, save on shipping
costs.
There's absolute winners.
It's a fantastic opportunitythat we're pushing.
All right, is that video goingto run?

(35:37):
Let's see.
All right, struggling to playthe video, never mind, it's
obviously too large.
So, embedded electronics, thisis an interesting one.
So the basis of our originalconcept when we set the company
up five years ago and the PhDthat predated that was using
this process, calledprint-pause-print, which is

(35:59):
fairly common knowledge and onlycertain printers allow you to
do this where you printeffectively, the open channels,
take the part out of the printer, you take it out, you then use
a secondary material, asacrificial material, to fill
the channel so you can put itback in, print the lid directly
on, and that's how you get yourfully sealed micro-threaded and
the knowledge of how to use thatmaterial.

(36:19):
That's our specialty.
That's what we can do reallyreally well.
We know what material it is, weknow how to apply it, we know
how to remove it.
We don't always use it, though.
As printers have got better,it's less applicable.
It certainly doesn't work withcircular cross-sections and
complex geometry, but thatpausing process allows us to do
other things.
So, as I say, we can put glasscomponents in there, we can put
flexible membranes in, wemembranes in, we can put

(36:41):
electronics in.
Knowing that electronics arereally useful for, especially
for electro-biosensing,electrochemical biosensing, we
thought, well, let's try and puta PCB in that.
And because we've got somelarger format printers, we
thought let's put it into a wellplate.
So if this video played and Ithink it's on our website it

(37:03):
demonstrates a very customizedwell plate.
So the left-hand side, there iswells of a 96 well size.
We've got 16 and 32 well sizesin there as well.
We've got micro channeledbetween the wells.
The pcb is wired up to acontrol system that detects
fluid.
It's got fluid detection.
It's got heaters.
It's got thermistors as well tomonitor the temperature.
You've got light sources andyou've got an open one in there
as well.
So we're're trying to promote.

(37:24):
You don't have to be restrictedto the layout of a well plate,
but you can still use a wellplate reader to run your
particular experiments.
Obviously, as we are embeddingthe electronics in that, we can
have the fluid directly incontact with the electronics or
we can insulate it and embed itslightly within the resin itself
.
So let's say you've gotelectrical insulation and then

(37:50):
we get onto my favorite stuffthe anatomical models.
These things are amazing.
This came out if those of youwho follow me on LinkedIn you'll
see I posted about thisrecently because quite a lot
recently.
It came originally fromcreating a leaf model.
That came from the idea of canwe avoid using straight lines?
Can we take a sketch, turn itinto a 3D STL and 3D print that?

(38:12):
And we started off with takingan X-ray of a leaf to generate
the microfluidic channels of theleaf venation.
That got the attention ofpeople involved in vascular
systems and that opened upfurther opportunities.
So the model at the top thereis actually a prostate.
It's taken directly from anX-ray.
We've converted the X-ray tomonochrome image, black and

(38:33):
white image.
We've converted the X-ray tomonochrome image, black and
white image and then convertedthat into a three-dimensional
circular sectioned STL file thatwe 3D printed, the one at the
bottom.
It's called, if I can get theLatin right, metamorablus.
It's a vasculature network, Ithink, in the brain, which is
for research, r&d purposes, ofresearch chemicals going in

(38:54):
through the vasculature and atthe moment you can only use on a
live brain, which means usinglab animals.
So in the move to reduce labanimal testing, by 3D printing
these shapes which come directlyfrom the CT scan, we can create
vasculature networks forresearch.
The channels that go across themiddle, if you imagine that
sort of H-shaped the channelstask as your networks for

(39:14):
research.
The channels they go across themiddle.
If you imagine that sort of Hshaped, the map channels that go
across the central bar.
They're 250 microns in diameter, fully circular and all over
the place.
These things are brilliant,absolutely love them.
We've all.
We've done kidneys, we've donelivers, we work in the heart
model at the moment and itreally shows what you can do
with 3d, 3d printing that yousimply cannot make any other way
.
So what's next?

(39:36):
Is that video going to play?
Rouch, the video's not playing.
If it played, there's a reading.
Oh no, it's trying to.

Speaker 2 (39:45):
I think it's going to work.
Yeah, the video's going to work.

Speaker 4 (39:51):
Yeah, there we go.
Yeah, we're live.
So that's a time lapse ofmaking key chains, which you may
ask.
Why do we make key chains For aswag to give away at trade
shows?
But it demonstrates the batchproduction capability.
So we made these chains there.
They're not the size of abullet.
Perhaps You've got a spiralmicrofluidic channel going there
.
Obviously we've got the logoembossed into it and we make

(40:12):
those 150 at a time.
You've got three rows of 50, so5 by 10, I think it is
components all in one go.
It's, I don't know, aneight-hour print or something
that's.
One printer can make 150 partsin eight hours.
Two printers can make 300 parts.
You can make 600 parts in a day.
You can really see the benefit.
If you design for additivemanufacturing, batch production,

(40:33):
larger scale production startsto become very, very feasible,
and that's what I'm keen topromote, as 3D printing has a
lot of advantages.
It's very early days really inthe manufacturing side and I
think it's additivemanufacturing.
The world over, every sector islooking at how to scale up and
it's slowly becoming much morenormal.
The example at the bottom ispressure formed thermoplastic.

(40:55):
I haven't got a photo of thehot embossed parts, but we've
been 3D printing tools to do hotembossing so we can turn around
the tool in a matter of hoursand then hot emboss it so you
can then translate intothermoplastic.
So if you need to use COC, cop,polystyrene, whatever the idea
is, we can actually do a genuinerapid prototype in COC or
polystyrene and so on and getthat out within a day or two.

(41:19):
Yes, the tool isn't gonna lastvery long.
You might only get 10 partsfrom it.
You might only get two, dependson the complexity, but you can
do it.
That's the key thing here and Ithink there we go.
That is the last slide.
Hopefully this one plays.
Please play, Yay there we go.

Speaker 1 (41:39):
So that's the leaf that I mentioned.

Speaker 4 (41:40):
It's just super cool.
I love that one.
As I say, we're based in NorthEast England, New Caledonia.
There's a growing biotechsector here, but we work with
customers all across the UK,across Europe, across North
America, a coupled in Australiaas well.
We're slowly getting over thereas well.
We make small parts.
It's very easy to ship themglobally.
Tariffs or no, it's not amassive problem and hopefully

(42:04):
that's kind of coveredeverything we do.

Speaker 2 (42:08):
So there we go.
Thank you, paul.
Your presentation always wowedthe audience with the incredible
visuals and videos.
So people who have questionsplease put them in the Q&A box,
because I can't keep track and Icannot handle explosion of
questions, 10 at a time, thelast minute of the session.
So my question is how do youwork with typically work with

(42:30):
customer uh, who wants to let'ssay, a large customer who wants
to mass produce something?
How do you walk them throughyour design process to your
manufacturing process?

Speaker 4 (42:42):
yes, the application depends what they're trying to
get out of it if if somebodycomes along to me.
So say, hypothetically nottotally hypothetically
scientific startups spins out ofuniversity.
They've got a novel assay, anovel biosensor.
They're trying to turn it intoa platform.
They they want to work with usto do the rapid prototyping, but
their five-year plan involvestransitioning to to injection

(43:03):
molding.
Depending on where they'relocated in the world, we'll
basically hook up with a localinjection molding company who's
got the specialist skills andwe'll get them involved at a
really early stage so we canconsider the design for
manufacturing.
So you know exactly what'sneeded to actually, you know, to
get them involved, we'll gothrough the process.
We'll help them on the on theiterations of design.

(43:26):
If they don't have anyengineering support, if they are
genuine, just a smallscientific company, we'll help
them on the design side.
If it's getting more complex,we've got a team of microfluidic
experts we can bring in when weneed to.
We'll help them on themanufacturing side of things.
We'll help them on theinstrumentation side and then,
when it involves scaling up, asI say, we'll start that
conversation at the very earlystage so that we know what we

(43:49):
need to do, and this is whywe're moving into the
thermoplastic forming as well.
So, as I say we've got thatstepping stone so we can
actually transition, becauseultimately the resins that
you're 3d printing in do behavedifferently, um to the
thermoplastic.
So we understand them and weknow what what the limitations
are.
We can do everything possibleto modify contact angle with,

(44:09):
with coatings for instance, tomake them the correct
hydrophobicity to represent whatthey're going to need in the
future.
But there is that transitionand we can kind of cover that,
depending on what they need todo.

Speaker 2 (44:21):
Okay, we have a question from the audience that
says how to integrate microvalveinto these 3D printed
microphotics.

Speaker 4 (44:30):
That's a really good question and something we enjoy
doing.
So the easiest way is simplyputting a flexible membrane over
the top.
Um, so again, with the printprint print technique you can
change material.
So you, you, you print the bulkof the part, change the
material and then print aflexible material to create a
flat membrane that can beexternally actuated either by an

(44:51):
external pneumatic system ormechanical actuator.
So that's the simplest way ofdoing that.
We've been looking at ways ofintegrating that more into the
part, if we can the other way isobviously you can actually use
an external, you know flangemounted valve, micro valve and
the likes of stygo or spc oranyone great.

Speaker 2 (45:10):
I have another question from actually several
questions from luano.
He says how did you overcomethe clocking issue?
Are the channels fully enclosedor do you use a welded hemi
layers?
Hemilayers, sorry, Hemilayers.

Speaker 4 (45:27):
I don't know how do we ensure that?

Speaker 2 (45:32):
Yeah.

Speaker 4 (45:33):
I'm going to say that this is our secret sauce.
This is our knowledge of how weensure that these channels are
completely clear.
I'd like to say it's magic.
It isn't.
I've just got some very cleverpeople.

Speaker 2 (45:45):
Yeah, no proprietary information.
It's fine.

Speaker 4 (45:47):
Don't tell us secrets , you can tell us you know, a
lot of it comes down toexperience.
We've been doing this for fiveyears, so we know how to well,
the team knows how to drive theprinters to use the correct
print settings for the correctmaterials, so we can actually
ensure that these processcorrectly well, at least we know
it's achievable by human, yeah,okay.

Speaker 2 (46:10):
So, uh, let's see what is acceptance rate in
fabrication for your works Idon't know, what that means yeah
no, no yeahsometimes yeah, sometimes I was
like is this like a scientificterm that I don't know, because
it's very possible.
Okay, it sounds like somebodyfrom uk is asking a question.

(46:32):
Lots of love from there.
It says a few stack questionshere.
Okay, for embedded electronics,are they being fully embedded
into the resin print?
Does this, yeah, okay, or isthis voided being created?
And so the answer is yes, allright.
All right, let's see.
Are you working with injectionmolder on the scale up?

(46:55):
I think you just said ittheoretically.

Speaker 4 (46:58):
Oh, I see, yeah, yeah , yeah.
I see that questionTheoretically going after their
work.
We're not going after that work.
We're never going to.
For the people who want amillion parts, it's just not,
it's not within our ballpark.
I have no intention of doingthat.
I want to work with theinjection molding companies and
we can do the, the, the one off,the two off, the 10 off, the
iterations.
We can do the hundreds off,maybe the thousands off for the

(47:19):
batch, batch productions, forthe data generation.
But when it gets to actuallymaking a commercial product, the
target costs that people wanttheir disposable cartridges to
come to, we can never match thatprice with 3d printing.
You know, in theory and I havepriced up making a million parts
a year for a job and it wasn'tinsane, it wasn't as cheap as 3d

(47:40):
printing, but actually itbasically meant buying a whole
room full of 3d printers,employing a team of people to
operate them.
But when you broke it down intothe unit cost per part, it
wasn't stupid.
It was an order of magnitude ortwo higher than you can get
with 3d printing.
So it's never going to goanywhere.
So that's, that's how we workand that's why I get on well

(48:01):
with the injection moldingcompanies well, if I've learned
anything from doing thesewebinars is never say never so
oh, yeah, absolutely yeah, allright, cool, we have quite a few
questions still, like I, I said, you know, the earlier you
submit your question, earlieryou can get in line.

Speaker 2 (48:17):
But we are running out of time and I want to move
on.
Paul, excellent presentationagain.
I think I need to rewatch thisto really fully grasp
everybody's content.
So we're going to move on andthen we're going to come back to
you later for a paneldiscussion.
Okay, yeah, awesome.
Okay, our next speaker isprofessor Christopher Marais,

(48:42):
who is a professor at McGilluniversities or lots of
Canadians today.
Let's see, let's see, let's see.
Okay, there we go.
Now we're in business.

Speaker 1 (48:57):
If you can share your screen.

Speaker 2 (48:59):
that would be great, I can there we go.
Is that?

Speaker 5 (49:03):
working.
Yep, it's working now.
All right, wonderful, okay.
Thank you, jenny, for theinvitation and thank you
everyone for being here.
I think I present a slightlydifferent viewpoint on this.
I am a very, very end user,right?
So we certainly don't make the3D printers.
We don't work with other peopleto design, you know, different

(49:25):
usages of that printer.
We instead use the 3D printerdirectly in my research lab to
solve a number of differentproblems.
So my work is in microscaletissue engineering.
I'm currently at McGillUniversity in the departments of
chemical and biomedicalengineering.
My work is in building tinyversions of tissues so that we
can understand how they function, and I'm particularly excited

(49:47):
and interested in the mechanicsof those tissues and how
mechanical forces change diseasefunction.
Here's a sort of big pictureoverview of what my lab does.
We build organs on a chip, sothese are again tiny versions of
tissues that can be fluidicallyconnected to each other.
We look at various diseaseprocesses and we look at various
developmental processes, ourgeneral workflow.

(50:08):
The thing that has emerged overyears in the lab is we build a
microscale version of thattissue.
We use the fact that we haveconstructed it from the ground
up to manipulate some feature ofthe microenvironment.
This could be mechanics, thiscould be oxygen availability,
this could be nutrientavailability, this could be a
number of different parametersthat are external to the cell

(50:29):
but still play a reallyimportant role in getting that
cell to function in a realisticway.
Because we've built the tissueourselves, because we've
manipulated various parametersourselves, it allows us to ask
questions and probe the tissuein ways that are quite unusual,
right Like quite impossible todo in an animal model or in a

(50:50):
human being.
This leads to some prettyunusual insights.
My group works withmicrofluidics and fabrication.
My PhD was spent largely in ahuman being.
This leads to some prettyunusual insights.
My group works withmicrofluidics and fabrication.
My PhD was spent largely in aclean room building these kinds
of devices where we canprecisely control fluidic flow.
We also spend a lot of timedesigning our own biomaterials
and we use them for variousorganoid applications.
So organoids are stemcell-derived culture systems

(51:13):
which can really capture a lotof the complexity of in vivo
tissues.
We merge those ideasmicrofluidics and fabrication
with biomaterials and organoidsthrough computational models to
really get a biophysicalunderstanding of what is
happening inside of thosetissues.
And then we often have to dealwith custom imaging problems as
well.
So we build our own microscopesand custom design our

(51:35):
microscopes so that we can imageall of these together to get
the kind of data that we want.
I have lots of differentinterests, but we've applied
this sort of workflow to anumber of different biological
systems.
We have active projects in thepancreas, in the lung, in breast
cancer, understanding theplacenta, understanding brain
development and very recently inthe gut and also now in spinal

(51:57):
cord.
So my interests are reallyquite broad in the biology side.
But I think that 3D printing isemerging as one of these tools
that can really facilitate thiscycle, this ability to rapidly
probe and understand biologicalsystems, this ability to rapidly
probe and understand biologicalsystems.
For today I'm going to ignoreall of the cool stuff that we do

(52:18):
in a biological system, but I'mgoing to talk a little bit
about our microfluidics andfabrication and then how we
integrate that with biomaterialsand organoids.
I'm anticipating that if youhave questions about it, I'd be
happy to take them.
Questions about it?
I'd be happy to take them.
So here's how we traditionallydid this.

(52:39):
I learned how to do this in aclean room, so this was 15 years
ago.
Training in the clean room tookabout three weeks to a month.
This is the fabrication process, which you've already seen
slides off, so I'll skip it.
We'd end up with our devices,and every time we wanted to
change one of these devices, itwas roughly two to three weeks
once you got good at it.
So after four years in my PhD,I got good at this.
I was able to produce a newstyle of device every two weeks

(53:02):
For someone who's brand new.
This whole process takes a lotlonger.
There's a lot of skill that'sinvolved in making this happen,
but you can anticipate a two tothree week turnaround time and
any changes in the device thatyou wanted to make.
When I was doing my postdoc atthe University of Michigan, I
was first exposed to this ideaof 3D printing, so this was back
in 2012.

(53:23):
Scott Hollister and Glenn Greenthey produced a really
remarkable workflow which justgot me extraordinarily excited
about this field.
What they did was they wereworking with patients, in this
case a baby who had a trachealproblem, and the trachea would
collapse frequently.
This is a life-threateningillness, of course, and Scott

(53:45):
Hollister and his team came upwith a strategy to build a
bioresorbable tracheal stent.
It would fit in, hold theairway open and then slowly get
absorbed into the tissue.
This was incredibly exciting.
I had nothing to do with thiswork, right, but I was in the
building at the time.
So this is the Carl Gerstackerbuilding on the University of
Michigan campus, and the feelingthat you got was just so

(54:08):
exciting that this was apossibility.
Right, this is something thatyou could do, and for me, this
was touted as the first timethat 3D printing had saved a
life.
I got very, very enthusiasticabout it and, just like with all
technologies that have comethrough my lab, I went through
various cycles of believing anddisbelieving in this technology.
So this is the Gartner hypecycle.

(54:29):
I'm sure you're all aware of it.
For me, this is my personalview of 3D printing over the
last little while.
For me, this finding being inthe building at the time that Dr
Hollister was doing this work,that was the technology trigger.
Everyone got really, reallyexcited about it, and then we
started to hit this trough ofdisillusionment.
I feel as though we'recurrently in this stage where

(54:54):
we're slowly figuring out whatthis is good for and what we can
use it for, and then, ideally,we'd eventually get to plateaus
of productivity afterwards.
So when we first saw thishappening, we said, okay, what
else can we use 3D printing for?
We tried a number of differentthings and we got disillusioned
fairly quickly.
So I tried to encapsulate thosedifficulties as best as I could

(55:15):
.
First is there are limited bydefinition.
If you 3D print something, youhave to limit your materials
right.
The material has to becompatible with the whole
process of 3D printing, and thatlimits the number of the things
that you can use.
This is particularly importantfor biological applications.
You've heard a lot about thebiological compatibility of

(55:37):
resins that are coming outHemdeep.
I'm super excited to try thenew CytoClear resin.
90% is way better than we'veseen before, but for many of the
applications in my lab, 90% isjust not enough.
When you start with a stem celland you grow it up to an
organoid, the cell might survive, but it just doesn't behave the
way that it's supposed to orthat we expect it to.

(55:57):
So cytocompatibility is reallya touchy subject for many people
and we found it to beparticularly painful when trying
to work with the limited rangeof materials that are available
for 3D printing.
I'll also say that theresolutions that we've got from
3D printing.
I'll also say that theresolutions that we've got from
3D printing.
They don't compare to the toolsI've already used, so I've

(56:18):
learned how to.
I grew up in the clean room.
I know how to make devices thatgo down to five microns one
micron resolution.
The 3D printed approaches thatare generally available to us
don't really compare to thatkind of resolution.
So for me, I had to make a bigswitch in terms of the design
practices that I use to try andavoid these sort of situations

(56:42):
where I need really highresolution prints.
I think another key issue isthat there's limited
availability of theinfrastructure but, more
importantly, limitedavailability of maintenance
capabilities and skills to keepthese machines going.
Right now, this feels like aproblem that is getting more and
more tractable, right Likewhere the instruments are

(57:03):
becoming cheaper and cheaper.
More and more expertise isbecoming available, but we're
also interested in makingdevices in really
resource-limited regions of theworld, and if you want to make
devices where they need to be,you need an entire
infrastructure of people to havethat happen, and so this was
one of those big concerns for us.
But honestly, with the rate oftechnology development and the

(57:32):
rate that things are changing,I'm hoping that this will be a
thing of the past soon.
So what do we do with the 3Dprinter in the lab?
Well, we have the filamentprinters.
I won't talk about the detailsof this, of course, but I'll say
some things about what this isuseful for.
It's easy to swap out resins.
Biopla, we found, is one of themost biocompatible materials
that we can get and it doesn'tmess with our cells at all.

(57:53):
Printing with PETG is great.
From building solid devices, wecan print flexible devices.
Resolutions here go down toabout one millimeter in XY, 0.3
millimeters in the Z direction.
Here's an example of the kindsof things we do.
We rapidly turn aroundmechanical devices.
This is an example of a systemthat we built that will allow us

(58:16):
to rapidly produce droplets,microfluidic droplets, from one
of these pores.
The 3D printed parts allow usto really improve the uniformity
of droplets that come out of it.
There's a ton of sample holders, sample forms, cases and
supports All of the stuff thatis needed for a lab to run but
which we don't often spend a lotof time thinking about how much

(58:38):
time and effort it takes us toput together, and my students
have loved the ability torapidly iterate on these devices
and produce things very, veryquickly.
Now, of course, there'sdownsides to those filament
printing.
They're not really good formicrofluidics.
Resin printing would be.
We've had a range of resinprinters in the lab.
High-resolution printing can godown to less than 100 microns

(58:59):
in X and Y, 20 microns in Z, but, as has already been discussed,
photopolymerization chemistriesreally change your capabilities
, especially with reallyadvanced biological systems.
What was really cool to us,though, when we first started in
this, this was our firstprinter back in 2016.

(59:23):
When we first started in this,we realized that the 3D printers
allow us to build designs thatyou cannot manufacture in the
conventional cleanroom, and wewanted to try and capitalize on
that to come up withapplications where we could use
this capacity to build genuinelyweird structures that are not
machinable any other way in inthe lab.
So there's a few way, differentways.
This is gone.
This is one application.

(59:43):
We work with the company herein montreal, nanofacile.
They are making a reallyinteresting system to produce
lipid nanoparticles that containdifferent system to produce
lipid nanoparticles that containdifferent cargoes.
The lipid nanoparticleproduction system is now a
3D-printed toroidal micromixer,so we print this directly into a
resin, much like CytoClear thatCadillacs was mentioning, and

(01:00:07):
we can use these to producelipid nanoparticles with really
really good consistency andreally good reliability in the
device, right?
So simple application, but wewere able to tune these
parameters on demand, with youknow, with roughly two hours
worth of turnaround time, whichis really quite amazing.
We then set ourselves to try tounderstand how do we use these

(01:00:30):
tools to work with more advancedand complex biology, and here's
an example of something that wedid.
We 3D printed spheres using aresin mold.
So this is on CadWorks' blackresin system.
You can print these largedevices.
This looks like.
Imagine a golf ball sitting onthe lawn in front of you, right,

(01:00:51):
that's what these devices looklike.
This device in itself has notgot great biocompatibility,
particularly for stem cellapplications.
What we did instead was toreplica mold this device into a
polyacrylamide gel.
Polyacrylamide isextraordinarily good at being
protein resistant and also beingcompatible with most biological

(01:01:11):
processes.
This is a hydrogel now, andbecause of the shape of the
hydrogel that we get from makingthese overhanging structures,
single cells can enter thelittle vessels, enter the little
chambers there, they canaggregate together and form
spheroids or, more recently,organoids, and because of the
size of them, they can't comeout again.
Because they can't come outagain.

(01:01:32):
This allows us to monitor inreal time, on an organoid by
organoid basis, each one ofthese tissues.
We can treat them directly onthe chip, which means that we
can wash out the liquid andreplace the liquid with the
treatment agent.
We can label them, we cananalyze them all while sitting
on the chip and if you want totake them out, you simply flip
the chip over and, because it'sa flexible material, you

(01:01:54):
centrifuge them and all of theorganoids pop out the bottom.
So this is what that 3D printedmold looks like.
Here's this closeup view of the3D printed part.
This would not have beenpossible even five years ago,
but you really get very cleanresolutions on the surfaces and
you're able to produce thesereally fine scale, high
resolution parts with excellentsurface finishes.

(01:02:17):
This allows us to make thesehighly uniform devices.
Each one of these little unitsis a pancreatic organoid that
we're taking from stem cells toislets.
The idea here is we're makingpancreatic islets eventually for
replacement therapies.
It was really interesting usingthis device because the
organoids all maintained exactlythe same size over 27 days of

(01:02:39):
culture in the devices.
In conventional models, whereyou've got a stirred tank,
bioreactor or something likethat, these organoids all fuse
together and they create weirdstructures that are difficult to
model and difficult to workwith.
So there's big advantages incombining 3D printed tools with
bioengineered and biocompatiblematerials in these ways right,

(01:03:00):
and that's been a common themein the lab.
We're going to continuepursuing that for many different
applications.
I'd like to talk a little bit,though.
I've talked, about replacingthe material, using the material
as a mold and then replacingthe material for the stuff that
comes in contact with thebiology.
I think that there's somepossibilities to use the
materials directly with biology,even with the more advanced

(01:03:21):
kind of stuff that we do.
So the reason we wanted to goto 3D printed devices directly
for organs on a chip is becausewhenever collaborators come to
us to talk to us about thingsthat they want to do, they're
excited now, right, or they'reexcited yesterday when they read
the paper that gave them thisidea.
Chances are pretty good that amonth from now they're going to

(01:03:42):
be less excited and lessenthusiastic, right, and then
people have trouble funnelingresources towards new projects.
The other issue that came up isthat my students can't always
keep up with the demand, right.
Some of these experiments takethree months to do.
We make a device for them andthen it goes away for three
months and then three monthslater we hear back.
My students have graduated andmoved on to something else.

(01:04:04):
Right Like it's very, verydifficult for them to keep up
with the demand that happens forthese longer-term developmental
projects.
I'll also say thatfine-resolution device
fabrication it's still finicky,still takes some skill, it
doesn't transfer easily betweenstudents and you'll always need
these iterations.
So we've been spending a lot oftime over the last year or so
wondering how do we designdevices so that they're as fast

(01:04:26):
and as cheap as possible to make, so that my collaborators can
turn this around in their lab.
Recently, 3d printers havestarted coming out at $500 price
points.
At this price point I canafford to put one in my
collaborators lab.
They can press the button andwe can work on the design
components of it.
This allows us to collaboratewith people who are not just in

(01:04:48):
Montreal but all across thecountry and all across the world
as needed, and there's now alab in New Zealand that's using
some of our devices with thissystem.
I won't show you the 3Dprinting process, of course you
all know that, but the conceptis that we can press the button
and have these devices turn up.
However, the difficulties stillexist.
Right Like materials,compatibility is a problem.

(01:05:09):
3d live bioprinting is still achallenge that I don't know how
to resolve really well foradvanced biology.
And then there's resolutionlimitations.
So I'll point out two thingsthat have come up recently.
David Yonko, a colleague hereat McGill, recently published
this paper where they looked atusing, where they came up with a
new kind of ink, and the newkind of ink can be used in the

(01:05:30):
really inexpensive printers.
You don't get the resolutionsthat you get with some of our
more advanced printers,especially like the ones you've
seen today, but you do get goodenough resolution for some of
these organ-on-a-chipapplications.
And his group came up with anink.
When I saw the biocompatibilitydata for that ink, that's when
I got really excited and startedusing these for our more

(01:05:51):
advanced biological models.
It's custom formulated, so wedo have to make this from
scratch in the lab.
We're taking that ink and thenwe're combining it with some
clever integration strategiesthat I think will reduce the
resolution limitations.
So maybe we don't need betterand better 3D printers to go to
finer and finer resolutions.
Perhaps we can get away byintegrating them with other

(01:06:12):
components.
You've already seen some of theworld-to-chip interfaces.
I completely agree with Paul.
This is surprisinglygame-changing for us.
Every time a microfluidicdevice fails, it's because of
this world-to-chip interface.
And the ability to put incompletely monolithic
world-to-chip interfacestructures, such as lure locks,
really hugely improves thereproducibility and robustness

(01:06:35):
of our devices, particularlywhen we put them in other
people's labs.
We've also taken to integratingnanoporous membranes into the
devices.
So here's an example where youhave a nanoporous membrane that
is fitted into this space, whilethe rest of the device is 3D
printed.
This allows us to flow insolutions, and those solutions

(01:06:56):
are maintained stably underneaththe nanoporous membrane.
They can flow into the channelsand move through the fine
networks later on.
The last thing this allows usto do the idea of integrating
different parts is we canintegrate bioengineered parts
into these devices.
So here's an example of ourbioengineered spheroid making
system.
We flip that over and stick itinto the device, and then we can

(01:07:19):
take a plug and seal the wholedevice off, so that you now have
a well-engineered system thatcan be compatible with the tiny
pore sizes that are availablevia the metaporous membrane.
That can be built in a coupleof hours in someone else's lab.
So that is very exciting for us.
Here's a super simpledemonstration where we just try

(01:07:41):
and solve a very simple problemthat has big applications and
big ramifications for others.
Culturing cells under flow andnot high shear stresses, but
just replacement of fluid, issomething that is really
promising in terms of manyorganoid development platforms.
Here's a fully 3D printeddevice where we just fill up the
tank and allow the tank todrain very slowly through a high

(01:08:03):
resistance channel and ourcells are cultured in that
region there.
We could print it this way, butthis is slow and expensive.
We've also figured out ways tointegrate our devices with
existing biological standardequipment that's in the lab.
This is the 50-mil Falcon tubethat my students cut up and
screwed into the 3D-printeddevice and that works without
leaks or anything.

(01:08:23):
It's quite amazing.
In these devices we've shown wecan get flow rates that are less
than half a microliter perminute.
That's maintained over a verylong time.
So again, these capabilities.
They don't come easily inmicrofluidics but we're able to
produce them with a 3D printedpart.
I'll stop there.
This is my lab group who reallydoes all of the work.
They're a fun group of peopleto work with and they're quite

(01:08:45):
willing to mess with bothbiology and engineering and it
takes a special kind of studentto do that.
So I appreciate them very, verymuch and our funding sources,
and that if you have anyquestions, I'd be happy to take
them.

Speaker 2 (01:08:57):
Okay, Thank you so much, Professor.
Let me see if there are somequestions for you.
Yes, there is.
Julie is asking are therecoding innovations that can
assist in integration of thevarious materials in the
printing flow that you'reworking with?

Speaker 5 (01:09:15):
That's a great question.
I can imagine several ideaswhere you know as a product is
being printed, a monitoringsystem looks at it and adjusts
the exposure time or adjusts theflow rate or adjusts the
removal of a liquid in responseto that right.

(01:09:36):
I think that means doing morethan just coding.
I think that's a hardware andsoftware integration that will
allow for real-time monitoringof this process.
The process.
It's getting way, way betterthan it used to be.
Right Like.
It's gotten to the point wheremy students brand new, they show
up in my lab two days laterthey're producing printed parts.

(01:09:56):
Right Like.
That is pretty wonderful, butit's still finicky in the sense
that if we have an off day inMontreal and the humidity is
really high, that changes theoutcome of the device.
So I can certainly seeapplications that would put a
process flow control loopdirectly into these systems, but
I don't know if that's just asoftware solution.

Speaker 2 (01:10:16):
In other words, how can AI help your work?
Just want to bring somekeywords.
I mean, have you guys tried?
I mean, when I do a reallyquick news search for the last
week, you know, whatever thatcomes out, it's always there's
like at least 20% of them wouldhave machine learning and AI
embedded in the title.

(01:10:37):
So just curious, have you guyslike try to explore some of the
ways that can optimize your work?
Oh for, us.

Speaker 5 (01:10:44):
Yeah, I don't currently have anyone in the lab
doing that kind of doing thatkind of project.
I can certainly see theapplications, though I do a
little bit of work with breastcancer diagnosis especially, and
in that field there is a wholelot of interest.
Just like with every AI system,you need an excellent training
data set, which means that youneed a lot of well-characterized

(01:11:06):
, well-labeled data, and I thinkthat's where that is going to
be the next big push in all ofthese different areas.
The AI models have proven theycan work.
Now we need enough data to feedthem.

Speaker 2 (01:11:17):
Yeah, data is the key .
Okay, we have a question fromPeter.
How does post-processing affectbiocompatibility?

Speaker 5 (01:11:26):
Yeah, For example, soaking chips longer in IPA.
Oh, okay, so the idea is thatthere are leachates that are
within the photopolymerizedresins, right, Right, those
leachates have to come out.
Generally, if you've got aphotoactivated molecule or a
photoinitiator, presence of thatphotoinitiator produces free

(01:11:48):
radicals, which are great forthe polymerization reaction,
which are terrible for thebiology, right, yeah, so
removing them is a no-brainer.
That has to happen.
There's many different tools todo that.
You can do an IPA wash forseveral weeks.
Eventually you'll get to thepoint.
But that sort of takes awayfrom the objective of printing a
device really fast and thenhaving it turn around.

(01:12:08):
Currently we print our devicesand we do five-day washes in IPA
and in PBS.
It's not ideal, but it isnecessary to remove as much of
it as we can.
We remove as much as we can.
We don't get it all and we knowthat what's left does make a
difference to the biology.
As you get to more and moresensitive assays, the viability

(01:12:30):
stuff cells will survive anawful lot, right?
The differentiation from amesenchymal stem cell down all
the way through a neuralprogenitor lineage to a neural
cell that's much morecomplicated and that tends to be
much more sensitive to anythingthat's left in the material.

Speaker 2 (01:12:49):
Maybe there is an idea for automation and
accelerate this washing andpost-processing process.

Speaker 5 (01:12:55):
Yeah, yeah, Okay cool , many of Paul's innovations
being very useful even for that.

Speaker 2 (01:13:02):
Thank you so much for your presentation.
We're running out of time so weneed to move on to our last
speaker so we can have a paneldiscussion.
Thank you so much, professor,and you can type in the answer
for some of the questions, andpeople can continue to put
questions for you, but I'm goingto introduce our last but not
the least speaker so that we canstay on time, jeff.

(01:13:23):
So our next speaker is JeffSchultz, the CEO and co-founder
or yeah, I guess are you the CEO?
I'm not really sure, actually.
Or yeah, I guess are you theCEO I'm not really sure,
actually Of Phase a brand new 3Dprinting microphotics company.

Speaker 3 (01:13:41):
All right, well, thank you, jenny.
Thanks for everybody for the 3DHeels for hosting today.
I'll try to get to the point.
Phase is an early stage startup.
We're actually still intechnology development, but I
have a long career in additivemanufacturing, 3d printing, and
I want to sort of share somethings that I've learned along

(01:14:03):
the way in the aerospace anddefense adoption of 3D printing
and how we, as a collectivegroup of people interested in
microfluidics and 3D printing,should you know, think about how
to move forward and show you alittle bit of what we're doing
at Faze.
So, as Jenny mentioned, I'm JeffSchultz.

(01:14:24):
I'm one of the co-founders ofFaze.
The other co-founder is ZekeBarlow.
My background is sort of thematerial science behind 3D
printing and also developingbusiness around unique 3D
printing and additivemanufacturing technology.
Fase is currently fundedthrough NIH research contracts

(01:14:47):
to develop an additivemanufacturing process to 3D
print, pdms.
We also have funding from theNorth Carolina Biotechnology
Center and also the state ofNorth Carolina and we're located
just north of Charlotte, northCarolina.
In our current researchprograms we collaborate with
Georgia Tech, master General,harvard Medical and Virginia

(01:15:12):
Tech.
On this slide I'll take amoment to pause and sort of tell
you about the genesis of wherewe are and why we went down the
road of trying to develop ourtechnology.
Zeke and I are friends,colleagues with Rafael Davalos,

(01:15:34):
who was at Virginia Tech, now atGeorgia Tech, and we were
discussing with him.
Actually sorry.

Speaker 2 (01:15:45):
Jenny is my video off .

Speaker 3 (01:15:46):
I didn't mean to leave my video off.

Speaker 2 (01:15:52):
No, your video is not on.

Speaker 3 (01:15:56):
It says it's disabled by the host, so maybe you can
turn it on.
So we were discussing withRaphael, who's a longtime user
of both 3D printing andmicrofluidics, and we were
actually developing a technologythat used liquid
dielectrophoresis to shapephotocurable resins and then we

(01:16:25):
would 3D print from there.
And we were discussing thetechnology with him and he said
you know, it's a great idea.
He's like I just want to beable to 3D print SILGARD 184.
It's a great idea.
He's like I just want to beable to 3D print Silgard 184.
He's like that's the material.
We know, for better or worse,it has its limitations and
complications, but we understandit and we would just like to be

(01:16:48):
able to print devices.
As Chris said, the process ofphotolithography is laborious
and slow and it takes many weeks.
And he said you know,eventually we run out of funding
.
So we always end up with kindof a microfluidic device that
has kind of a B minus likeusability, like if we just had,
you know, one or two more repson the molds, you know we would

(01:17:12):
get something good.
But if I could just 3D printPDMS, you know that would be
SILGARD 184 specifically, thatwould.
That would really just, youknow, make our device designs
easier.
So that that was the genesis ofPhase and and why we, you know,
sort of headed down the road oflooking at how do we 3D print
SILGARD 184 as it is, withoutany 3D print SILGARD 184 as it

(01:17:39):
is, without any, you know,additions of photo initiators or
photo catalysts, just printthermally curable PDMS?
I want to quickly also kind ofI mentioned I have you know,
hopefully a perspective that issomewhat unique amongst the
crowd.
It is somewhat unique amongstthe crowd.

(01:18:00):
My early career started outdeveloping a PhD was funded by a
company that's now part of 3DSystems and really understanding
what the properties, both therheologic and thermodynamic
properties of a material arethat actually make them 3D
printable.
You know this was probablybefore most people, a lot of
people on this call, were evenborn.
This was in the late 90s.
And then I moved into metaladditive manufacturing where we

(01:18:21):
developed a technology where Iwas the first inventor on all
the patents to actually printreally large scale metal
structures which is now beingapplied to printing tank holes
and very large componentsrelatively quickly out of
advanced materials.
And then I moved on to reallydeveloping sort of production

(01:18:47):
scale additive manufacturing.
I was the GM for a large Swissindustrial conglomerate's
additive manufacturing business.
And this was production additivemanufacturing.
We operated one of the largestor at the time the largest
called Service Bureau sort of aproduction of metal additive

(01:19:09):
manufacturing parts and thensort of moved on to left that
and really wanted to get backinto technology.
Development identifiedmicrofluidics as a place where
3D printing should absolutely bethe norm but just wasn't being
adopted.
Is why isn't, as people who aretrying to develop additive

(01:19:37):
manufacturing, 3d printing formicrofluidics is?
You know why isn't additivemanufacturing being adopted at
the rate that we think it should?
And you know how do we enable3D printing in the microfluidics
world?
And I like to think about thissort of from an equation
standpoint where you know theadoption risks have to be much,
much less than the rewards.

(01:19:59):
And you know a lot of thiscomes from my history with, you
know, going through the adoptionside bioimplants for
orthopedics as an end user, youknow there's introducing a new

(01:20:24):
material, you know there'salways unexpected interactions.
So that's, you know, asignificant risk when you're
thinking about a large researchprogram or assay development
where you know the end useroften sees the microfluidic
device as the packaging and theydon't want to risk their

(01:20:44):
project on packaging, eventhough I think most of the
people on the call appreciatethat microfluidics can be a lot
more than just the packaging.
There's also a newmanufacturing process,
especially on the morebiologically heavy end users not
familiar with CAD or operating3D printers, much more familiar

(01:21:06):
with pipettes and syringes andmicroscopes.
Often going to 3D printingintroduces a completely new
design.
There's also the associatedcapital costs, although you know
, we have seen a dramaticreduction in the cost of
printers, you know.
And then there's also, you know, the learning curve risk for
every user.
And then on the reward side,which I think is what we as a

(01:21:29):
community sort of need to sellmore and empower is, you know.
Sell more and empower is youknow.
I think we've seen it heretoday with a lot of the, you
know, chip to world interfacethings that we can do that are
inherently hard when you'redoing soft lithography.
So how do we not just make amicrofluidic chip that is has
better performance, but how dowe also make it much more usable

(01:21:51):
in the lab?
By putting lure lock interfaceson or integrating the tube
connectors, putting in valves,and those things can actually
lead to overall lower systemcost and improved usability if
you think about the time a gradstudent might spend in the lab

(01:22:12):
trying to make those devices andthen like developing chips that
do ultimately lead to improvedprotocols and assays.
You know, if you're in aresearch environment or even a
small biotech, you know, and ifyou're trying to introduce a
product, you know this.
You know an advanced design canpotentially lead towards you
know a market-leading positionin your assay and ultimately

(01:22:36):
lead to customer growth in newdesign applications.
So I want to walk through acouple of sort of historical
things where I think that it caninform us about the adoption of
3D printing and microfluidicsthat come from the aerospace,
defense and biotech side.
So prior to 2015,.

(01:23:03):
Fuel nozzles for sort of jetaero engines were basically this
is a conventional aero enginefuel nozzle it was about 20
parts, 20 welds.
Because you have so many partsand so many welds, you have a
high potential for failure.
Assembly costs were high.

(01:23:23):
And then GE introduced thisLEAP fuel nozzle, which is one
part 25% lighter, improveddurability.
Lighter improved durability,the ability to add more
complicated flow channels endedup in overall improved
performance and this really waswhat launched a good part of the

(01:23:51):
3D print things and put them insafety applications that
require real safety criticalqualifications and the things
that I think we should take awayfrom this as thinking about
microfluidics.
This was like an extremely highreward situation where it

(01:24:12):
basically allowed GE to makefuel nozzles that were better,
cheaper, faster, stronger.
The other thing that I think isreally interesting about this is
the material they used forthese fuel nozzles was because
here again we have a new design,a new manufacturing process.
They leveraged a cobalt chromealloy that was actually a
relatively well-known alloy.

(01:24:33):
So they knew how to qualify thepart, inspect the part, because
they knew what the fatigue life, what the corrosion performance
, what the basic mechanicalproperties needed to be to show
that they manufactured this partcorrectly.
You know, and I've kind of losttrack, but I know there have

(01:24:56):
been, you know, over 200,000 ofthese printed and you know most
of us, I'm sure, have flown on aplane that had these in them.
So when people talk about, youknow, is additive manufacturing,
3d printing, a commerciallyviable process?
You know it absolutely is.
It's all about, you know, aswe've heard before, designing
the part for the process.

(01:25:17):
Another interesting applicationspinal implants.
This is a conventional partTi6-4.
It was machined.
And then, if you look at the 3Dprinted version, same material,
ti6-4, but it has much betterosseointegration.
Looks much more like you wouldthink the inside of a bone
should look.
Again, significantly improvedperformance.

(01:25:40):
3d printing allows you tochange, have multiple sizes, but
again, this was the samematerial.

Speaker 5 (01:25:46):
There wasn't a material change.

Speaker 3 (01:25:47):
It was just a process change and design change here.
Finally, you know the area oforthodontia.
You know we've had braces, youknow, since we were all little
kids.
But you know, in the pastdecade we've seen dental liners,
you know, conventionally calledInvisalign.
I think most people didn't evenrealize for a long time that

(01:26:07):
these were actually 3D printeddesign.
3d printed, you know, butthat's been, you know, a game
changer in terms of changing theway people approach orthodontia
, especially as applyingorthodontia to adults.
And again, this was a it wasn'ta known material, but it was a
known-ish material in thatmarket because dentists have

(01:26:29):
long used UV curable resins, sothere was always a familiarity
with it.
So it wasn't exactlyintroducing, you know, a brand
new material.
So what are the sort oftakeaways that you know?
I think we should all thinkabout?
I think one of the things it'sbeen alluded to.

(01:26:49):
But I think for 3D printing toreally shine, we need to not do
like-for-like substitutions.
You're never going to 3D printa part that's currently
injection molded or a verysimple 2D microfluidic design,
and be competitive.
I mean you need to enhance theperformance.

(01:27:13):
Again, there's alwaysprototyping scenarios where you
might make some very simplemicrofluidics, but we need to
help educate our end users abouthow to incorporate multiple
elements.
You know, maybe you havemultiple mixing elements valves,
membranes, all in one deviceand those things you know are

(01:27:34):
enabled by 3d printing.
So, again, not like for, likeyou know, the you know
performance improvements in thedevices have to be, you know,
significantly better than theswitching costs when you're
trying to convince someone thatthey don't need to use, you know
, soft lithography to to maketheir devices um.
It also opens, um, you know,new alternative approaches, like

(01:27:55):
in the Invisalign example, andthe risk reward really has to be
highly favorable, especiallyfor the first mover.
On PDMS is that we employindustry standard materials.

(01:28:17):
That you know.
If there is an ISO spec, anASME spec, an ASTM spec, it
really helps bridge a gap forthe users in terms of adopting a
new technology.
And again, don't 3D print 2Ddevices.
If you look back, you know thisis one of Whiteside's devices
and it's from a you know famouspaper.
And you know this is one ofWhiteside's devices and it's
from a you know famous paper andyou know who wants to connect

(01:28:37):
all those tubes.
But if you look, take anexample, this is actually a
pretty complex 3D printedmanifold that Moog has used as
an exemplar.
Really, you know you have a lotof tubing connections now and
you just have to bolt up fourflanges and now you have a
relatively complicated manifold.
So I think that's how we needto think about selling our

(01:29:01):
customers on 3D printing.
So now on to phase specificallythe device.
You see on the screen.
There is a 300-layer sort ofbasic 3D mixer design with 2D
mixing elements.
You know, one of the thingsthat you know we focus on is one

(01:29:23):
Sylvard 184.
That's the material we alwaysuse with known biocompatibility.
We actually I'll actually focusa little bit back on the slide
there for some key aspects wealways print on a microscope
slide so our devices fit intosort of standard workflows.

(01:29:43):
The user gets a device that is3D printed right on a microscope
slide so it's ready to fit intheir conventional microscopy
workflows conventionalmicroscopy workflows.
One thing that we do that sortof eliminates some of the
problems that are conventionallyfound when you bond a soft

(01:30:05):
lithography PDMS part onto amicroscope slide is there's
often leaking at the interface.
We actually print down a solidum pdms substrate right on or
not print down a solid pdms baseon top of the microscope slide
before we start printing anychannels so our devices can
actually be completely removedfrom the device to um without

(01:30:29):
leaking.
Again, we have the designfreedom of 3d printing.
Uh, our goal is to replicatethe resolution of the in vitro
environment.
I'll talk about our size scalein a bit Rapid design iterations
.
That's key to what we do andour success.
Typically takes us about 30minutes to print a part and we

(01:30:53):
will go do three or four designiterations a day usually and
that's very important for us togetting the right microfluidic
design with the right material.
And again, we focus on scalableproduction.
Our process is fast, ourhardware allows us to produce

(01:31:13):
things at volumes and pricepoints where we could actually
go into production and we also,using some similar techniques
that we've heard about before,we integrate electrodes and
membranes into our devices.
I'll move quickly so we havesome more time for the panel

(01:31:34):
discussion.
On the left you can see anormally closed vacuum actually.

All Episodes

Episode #82 | Advancing Microfluidic Technology Through 3D Printing (Virtual Event Recording)

Episode Transcript

Popular Podcasts

Stuff You Should Know

New Heights with Jason & Travis Kelce

24/7 News: The Latest

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Episode #82 | Advancing Microfluidic Technology Through 3D Printing (Virtual Event Recording)