If you're going to build your

chemical engineering undergraduate degree.

There

are some great out

west.

That's your case has only recently and.

Here in

his research for the story.

Thank you very much Ron.

And it's a pleasure to be here
today in spite of the cold weather.

I thought I was done with all
this when I left Cleveland but

I guess not the step to do and
I'm going to talk about two ongoing

projects in my research group that
deal with membranes for fuel cells.

And how we are trying to look
at nano morphology as a way to

optimize the performance of these
membranes during fuel cell operation so

to begin with we start with a sly
fuel cells we want to one where we.

I show you here a hydrogen oxygen and
a direct methanol fuel cell

the basic components of the pen
proton exchange membrane fuel cells

are a polymer Erik membrane
material that separate

an anodyne cathode at the end of
hydrogen this oxidized the protons and

electrons the ECT electrons move
through an external circuit and

the protons move across this hydrated
polymeric membrane protons and

electrons mix with oxygen
from air to produce water so

the overall reaction in hydrogen plus
oxygen goes through each two zero.

In the direct methanol fuel cell rather
than using a gaseous fuel we're using

an equally a solution of methanol where
it's oxidized to C O two protons and

electrons and again the electrons pass
through the external circuit you can

extract useful work from those electrons
and the protons move across the proton

exchange membrane where they again
react in the same way at the cathode

the air cathode with oxygen and
so here is the overall reaction.

So that there is considerable interest
in these kind of fuel cells for

portable power or for perhaps electric
vehicles stationary power and

my research is focused on the proton
exchange membrane components.

Now these are schematic cartoons
of the of the fuel cell

operation in reality what we have
is we have a very thin proton

exchange membrane in on the opposing
surfaces we have a catalytic powder.

And that's where the electrode
reactions occur.

There are these gas
diffusion layers these are.

Carbon cloth or carbon paper
that a laugh out of the influx

of gases then they remove all of
water product from the cathode and

then their backing plate with flow
channels through which the hydrogen and

oxygen are met them all in an air flow.

So again this is the real picture of
what we're looking at a very thin

membrane that has to have
certain properties and so

what does this membrane do the remembering
separates the anode and cathode So

there's not a no physical short
circuit between those two electrodes

it prevents mixed scene
of the fuel in Occident.

So there is no chemical short circuit and.

It.

The membrane provide say
conduct the pathway for

Proton transports of protons have
to go through this membrane quickly

because the protons are carrying
the current through the polymer.

So that's the purpose so one of
the requirements of course we need a high.

I on a conductivity and zero electronic
contacted you know electron movement.

We have to have long term chemical and

mechanical stability these
fuel cells operate that.

Sixty eighty degrees and
some applications you like to have them go

up to one hundred twenty degrees C.
and they're going to go

through thermal cycling you turn
the fuel cell on you turn it off.

It's been used in Minnesota in your
automobile in the middle of winter.

It's all it's going to freeze perhaps or
be exposed to low temperatures.

So it's going to go through
a series of thermal cycling during

the fuel cell operation you
can produce peroxide and

hydroxyl radicals at the electrodes and
they can eat into polymer material.

So you worry about chemical stability.

They are going to swell and
shrank these are hydrated.

Homer so they swell and
shrink depending on temperature and

operation of the fuel cell and so

there's a number of mechanical and
chemical stability questions involved.

We want to be low Occident and fuel
crossover no pinholes has to be compatible

with the catalytic layers that are pressed
on to the surfaces of this membrane.

And of course low cost.

So here are all of the.

All of the requirements and if you go
back and look through the literature

the primary material the mid
the membrane material that seem the most

work over the last fifteen years or
so for fuel cells is

this material with a trade name
a fee on Buy made by Dupont

it is a perfectly fine a cassock polymer

backbone is polytech to floor ethylene
then there is the linkages and

it's side chain and
seen in a cell phonic acid group so

lead the key here is that if you
want to be able to transport protons

across this membrane put a lots a lot
of a Citigroup's in the membrane.

All right.

And these cell Fina Cassy groups when
hydrated will dissociate into an S.O.S.

three minus and an H. plus and so now you
have acidic groups that can now begin

to move across the membrane when there's
an electric field occur and current flow.

There are problems though with Matthew and
it has a lot of attractive properties say

it's has good chemical resistance
this Teflon based material.

This is so fine a gas a group is a super

acid with electron withdrawing
troops from the C.F. Tuesday.

So it's a very strong acid
lots of dissociation.

But there are problems and
the two main problems with may feel on

is that if you use it in
a methanol fuel cell.

There's lots of let methanol leakage.

So methanol goes from the end where
it's supposed to be oxidized and

gets into the membrane and
diffuses across the cathode and

when that happens a whole series
of deleterious effects occur.

You lose power in the fuel cell you have a
chemical short circuit some of the oxygen

that you'd like to use for reducing
water now is used to chemically oxidized

the methanol you're producing lots
of unnecessary warder at this

electrode which causes problems of
gas transport to catalytic surfaces.

So people want a fuel cell membrane
per D.M.F. C direct methanol fuel

cell operation that has a lower methanol
per me ability than what you get here.

The other problem with the Fianna is that
if you go to higher temperatures above

eighty degrees C. and you start lowering
the humid idiot of the air that's

passing into the system or lower
the humid of the of the hydrogen gas.

So if you're running under low humidity
conditions the membrane tends a dry out

and when the membrane dries out
the conduct of it he drops.

And so
can you find a way to improve the conduct

of it even a fee on under lower humidity
conditions and higher temperatures

temperatures maybe in one hundred
two hundred twenty degree C. range.

So what's been done if you
look chronologically at

what's been done in the fuel
cell membrane arena

people have known these limitations
on a few on for some time.

So the first thing was what I would
just switch out the polymer and

we'll find a polymer material that
a work better than the FIA and so

they investigated sulfa need to pull
the cellphone self unaided poly ether.

He tones poly emissions.

And the conclusion was no they couldn't
find anything that really worked as well

as an A.T.M. It really didn't solve
these two fundamental problems and

in addition only it was difficult
to get as high enough contacted me

as one might get in a few on the
mechanical properties and they were very

good the chemical stability issues of
some of these materials were suspect.

So then people said OK well
let's try something else.

Let's look at.

Polymer inorganic composites or

polymer polymer brands let's blend
a couple of polymers together and

maybe in that way we'll get the best of
both worlds one one polymer will promote.

Control of mechanical properties
another might control the proton

conduct to Vittie.

So people look at the addition
of zirconium phosphate silica

particles into polymers many of them

like these or they look at blends
of pole even eluding Flora.

Adding some floral polymer
to these materials and

again what people found is no you're
diluting the ion exchange polymer conduct

a beauty is dropping whatever benefits you
may get in terms of physical mechanical

stability you lose somewhere else and so
more recently people has been the same.

Well maybe we really need to do is
look more carefully at the membrane

morphology the Nano morphology let's
get down to molecular dimensions and

see if we can control or
tailor what's going on on the molecular

level to give us some improved
performance which is not going to be so

simple as to pick a polymer off the shelf
or mix a couple polymers together

we really got to get into the molecular
interactions which control pro.

Time transport which control
hydration of these polymers and

methanol Permian billeted and that's
what my group has been doing and we work

in four areas right now and I'm just
going to talk about two of them today.

I'll be talking about the fabrication
of nano fiber composite membranes and

I'll talk about what we call
stretch recasting a few on.

So in one case we're going to modify
the morphology of this particular

material by stretching.

And I know another case will be
looking at different kind of

artificially created structures.

We also look at rigid rod polymers and
block current polymers or

alkaline fuel cells but I'm going to
be focusing just on those two topics.

So let's begin by talking
about nano composite.

Nano fiber composite membranes.

So here's a schematic cartoon
of what we're trying to do.

We're trying to fabricate a fifty
separated polymer system.

Where we have proton conducting nanofibers
embedded in an inert polymer matrix.

So that the structure that
we get is not this similar

to a black hole polymer system but
in a block or

polymer system you have hydrophilic
in a hydrophobic block and

then those blocks self aggregate
into some molecular structure

could be lamb or layers or
it could be tubular type geometries and

what we want to do is we don't want to
leave it up to nature to decide what this

family structure looks like we want
to get in there and make our own.

So what we do is we will
elect to spin fibers.

We're a lecture spin fibers.

We will control the density
the fibers the number of channels so

we can control the size of those.

Birds and the number of fibers.

We will connect the fibers
together at the cross points and

we'll make a three D.
network so right there.

Will connect those fibers to one another
and create a three D. network of a proton

conducting polymer and then we will
fill the streets in between the polymer

within the nerve matrix that will give
this membrane mechanical strength and

be a very good battery here for
gas or other.

Transport type properties.

So what do we do in such
a system like this is that

we are going to decouple mechanical and
proton conducting functions.

One of the simplest things to
do in the fuel cell membrane

a RINO is to take a polymer and

add a lot of these some phonic acid
groups this S o three age group.

And as much as you possibly can but as you
increase the number of S O three groups

you increase the hydro felicity of
the polymer eventually it becomes water

soluble that's no good but you're
going to create lots of swelling and

if in the dry state those
materials tend to be brittle.

So you try to make a homogeneous
membrane out of a very high

polymer very high concentration of
acid groups and you run into problems.

OK In terms of the mechanical
integrity of the membrane.

Well we can use fiber is a very
high concentration as a groups.

And they're going to their swelling is
going to be controlled by the surrounding

inert matrix and the overall properties

of the membrane are going to be controlled
to a great extent by this inert matrix.

So we could also independently
control the size and

the number of these nanofibers.

And eventually and we've just started to
do this is that we can explore a certain

into a facial effects in particular
capillary condensation right now.

I'm going to show you some results where
we have a solid polymer nano fiber.

But in principle we can make a nano fiber
with a nano cool water that's empty.

All right.

And in stride.

That open core we can perhaps condense
water by capillary condensation so

this membrane will hold on to water
absorb water under low humidity

conditions at high temperatures and that's
something that people are looking at.

So that's the concept.

Here's what we've done our initial
work was to look at a sofa

made a poly cell phone and
here is the repeating murder structure.

Lots of aromatic grains here.

Ether group.

Soul foam group and
we synthesize this particular polymer and

we ed S o three hs those who
are Citigroup's to aromatic rings.

So this is the polymer that we're looking
at at two point five million mile per

gram think of that as
an equivalent as as a concentration

of about two point five molar because
the density of these materials are about

one gram per cubic centimeter So
this is this is a solid acid

polymer with a acid concentration
of about two point five molar.

We take this and we Electra spin it
the way in which you Electra spin.

You put it into a solution.

Dissolve this polymer push it.

Alice through needles and
put a large voltage difference between

the needle tip and a collector and
what we use in our studies is

a rotating drum collector that oscillates
laterally in order to give us a very

uniform what we call
Matt Electra spa nanofibers

and they're variety of
processing variable.

As that we have to look at applied
voltage the distance between

the needle tip and
the collector the flow rate of solution.

The polymer concentration the solution
viscosity lots of parameters here.

If you don't do it right you
are spray you make droplets

if you don't do it right
you get beads on fibers.

There are various types of geometries that
you'll get we want the Nano fibers and

so we use this particular apparatus.

And we make mats the mats look like
teflon tape and like they're white.

And.

Here's our initial mat.

And here are the conditions under
which we created that Mat fourteen

kilovolts that's a voltage drop in the
distance between our spinner and collector

our rotating drum speed the fiber mat
density is around twenty percent.

It's not very dense the and
here's a histogram.

Of the average of the fiber diameters.

And in this particular case and here is
that little bang if occasion high bang to

the cation S.C.M.
of the math surface and you can say that

the average fiber diameter is a little
bit over one hundred nanometers So

he's a very small fibers they're
twenty percent fiber density.

So what we want is we want a higher
density of fibers those fibers are going

to be the pathways through which
protons move at twenty percent and

not enough pathways.

We want to densify the mad and the way we
do that is through mechanical compaction.

If we compress the mat at
seven hundred P.S.I.I. for

three minutes that twenty percent fiber
volume fraction increases the thirty.

If we compress at a higher pressure
thirteen thousand P.S.I.I.

it increases to six.

The four percent.

So there's a correlation.

The more pressure you
use to compact the map.

The closer these fibers get so
now we have a fiber mat shown here

that has a much higher density of these
fibers the next thing we want to do

instead is to weld these
intersecting fibers.

So we have a three D. network.

Right that protons can move along one
fiber hidden intersection point then move

another way in another way.

OK Because in principle when you
Electra spin this initial math.

Might have been one continuous fiber.

All right.

Extremely long.

All right.

And what we want is this network
of interconnecting fibers and

we can create these into a fiber
welds by exposing the mat to vapor or

and in this case we use the M.F. paper.

And when you do that.

This is the structure you get and

you can see that we've created welds
some of the fibers have also been welded

together but generally speaking we still
end up with these very small welds.

So now we have this interconnected
fiber mat a very high polymer this

particular polymer that we use here if we
try to make a homogeneous film it would

swell like mad in water and if it dried
out it would break because it's brittle.

The next and final step is we
want to fill in the enter fiber

voids with an inner polymer And
so what we have use

is something called Norlin optical
it he sort of and away sixty three.

That's a photo curable your
thing based Pretty Polly Mary.

It's a little bit.

Tiriel.

There's no solvent.

So we impregnate imbibe this anyway.

Sixty three into the fiber mat.

Then we U.V. cure it.

And there's no solvent loss and so

there's complete filling of the fiber.

Spaces.

OK the space between the fibers.

If we use the polymer in a solvent.

We can impregnate that.

But then when you evaporate the solvent
you're going to get a volume loss and

you're going to get holes
between your fibers

here there are no holes everything
that we add Polemarchus ices.

So we use this.

And then we get a membrane.

And here's some data and
this is a a compacted

well bit impregnated mat Friess
fractured cross-section.

You can see all of the fibers in
between those fibers is this and

away sixty three and what we did is
we made a whole series of membranes

with the volume fraction of
fibers ranging from around thirty

percent up to eighty we've since GO DOWN
gone down to ten and ninety here.

So we made different fiber mats of
different fiber volume fraction and

we measured the proton content T.V. and

we measured water swelling how
much water gets in the membrane.

The straight line data here
between a homogeneous and

away film which has zero contact Tiffany
and a homogeneous poly cell phone

film which has this convict a body we see
a straight line convict a pretty effect.

Which tells us that if you have a fifty
fifty five or mat in dirt material.

You're going to have fifty percent of
the conduct of any that you would get in

a homogeneous membrane so
conduct levity scales linearly

with the amount of proton conducting
fiber material we have in or membrane.

We then look at the water uptake.

And that doesn't follow this linear
relationship where we see for

example at fifty percent fiber volume
fractioned is that we're getting something

less than the fifty percent swelling
that you might imagine for water up to.

And what's happening is the complaining
effect of this and away sixty three.

It's a hydrophobic material that material
surrounding each fiber in preventing

the fiber from swelling as it would if it
was a freestanding home or genius film.

So we see this dip in the water uptake.

When we get to very high
concentrations of fibers

now it appears that it scales linearly and
what's happening there is

that we no longer have a percolation
threshold for The anyway.

To anyway.

There's not enough any way to surround or
have a fibers so

now the fibers are can can
swell without restraint and

now the swelling is directly a function
of the fiber volume fraction.

So here are some data.

The problem right now is again this
is the conduct of medium water so.

We also wanted to look at other
properties of these membranes.

So we've looked at the oxygen
barrier properties of the films and

we look at mechanical properties.

So here's a homogeneous film made from
our self a native poly cellphone.

Here's the oxygen Kermie ability.

Here's the oxygen perm you build me up or
any way or not material.

Again it's very low.

And there's a composite membrane.

And he saw our competitor
commercial maffia on.

The same thing is that
it's about a fifty times.

Lower fifty the oxygen premie
ability decreases by a factor

of about fifty in our film as
compared to commercial nephew and

so these are very good
oxygen barrier materials.

We also look at mechanical
properties Young's modulus and

again you can see that in our
membrane it has a much higher.

Strength to associate.

Did with it as compared
to commercial Matthew.

So.

It's a good oxygen barrier has
very good mechanical properties

that it still doesn't have
the comic to be we want.

So what we have done.

This and have the conduct of
it to to match up with what

the Department of Energy which is
funding this project is looking for.

So what we did is we decided to stop.

Electra spinning Pollie cell phone and
start to spin.

Nafi on like materials and
that's not such an easy thing.

So the problem is that when
this polymer is not soluble

not truly soluble in solvent
it forms a mysel or solution.

And to do.

Electra spinning and you have to have
chain polymer Cheney long geisha in and

entanglement in order to make
a continuous fiber and so

people have tried to electricity in these
materials with very little success.

What they did was however they
course spin the Nafi on with a high

molecular weight doping and so
that's what we decided to do but where's

the others were looking at twenty twenty
five percent of the development or more.

We were looking at very low concentrations
we did not want to contaminate or

dilute this perf little
self on a cast of material.

So what we did is we've Electress Brun

this material this is a material
from three M. corporation.

That has a very high concentration
of the sole phonic acid groups.

So we Electra spin this with a dope.

To further increase the proton
conduct to Vittie we Ed.

The self unaided UK the phenol
polyhedral wholly a legal marriage.

So saying which I will from now
on say is self unaided poss.

All right.
This is molecular silica shown here.

It's got lots of sulphuric
acid groups on it.

So we're going to take this material blend
it with this put in the polymer doping and

start making nanofibers and
membranes these should

increase the number of acidic groups and
should also help with water retention.

So here's here's some lecture spin
the results here we come spin

with polyethylene oxide P.E.O. We've
also done work with poly acrylic acid.

This is the total weight just
to show you some different

structures you can get
with Electra spinning.

This is a material with.

A twenty.

This is a got about a one point four
million more per gram one point four

men or a Citigroup and per floor so
fine a cast of material.

And we could spin that with
the poly acrylic acid five percent.

And if we're at thirty percent total
polymer concentration when we do.

Electra spinny you can see that we
get these ribbons that form and

they are the with of these is
about six microns very large.

If we drop that total polymer
concentration to twenty five percent.

You could see that there
we get ribbons here and

there may be open channels at
the two ends we're not sure yet.

And this particular case we drop
down from twenty to fifteen.

Eventually we make nanofibers So
here's a case where we make nanofibers

we add self unaided POS to those and if we

do that we get some new membrane materials
and here's proton conduct to Vittie now.

As a function of relative humidity the
Department of Energy has set a target for

high temperature low
humidity conduct to Vittie.

Their target is point one semen percent
a meter at fifty percent humidity and

a hundred twenty degrees.

And in fact our materials have
met that Dia we target and

these measurements were done actually at
a third party independent laboratory.

And that the DIA we subcontracted so

that there would be no ambiguity as to
how people measure conduct to Vittie and

everyone in the particular research
program had to send their samples out.

Here's a commercial not Fiamma Terrio
you can see how it drops.

So we have this material per floor so
phonic acid with this S. pos

there's very nice job of contact to Vittie
and that's a very hard target to hit.

Now I'd like to switch gears
in the next few minutes and

talk about direct methane or fuel cells.

Again what we're trying to do
in a direct methanol fuel cell

is to have a high contact to Vittie.

And let the root of
the methanol permeable it.

The problem.

Historically has been whenever you
try to make the membrane a better

methanol barrier.

You're also going to
drop the conduct of it.

All right.

Conduct liberty depends on
the world of content and

the acidity the number of
acid sites in the membrane.

If you want to block bethen off.

Chances are you're also going to block
water getting into the membrane methanol

and mortar too close to one another
in terms of properties and so

if you dropped methanol transport
in absorption into our film you're

also going to drop water you drop
water proton conduct to the drops and

then you lose because you have to
maintain high proton conduct to Vittie so

I thought well as well let's uni
actually only one dimension.

Maybe will change the orientation of
the Nano structure of naif you may.

Be that all change the polymer
Crystal will be in some way and

maybe we can change the methanol permeable
it without changing contact evictee So

again here is that if you.

And I go ways to decrease crossover
without the loss of conduct to Vittie So

we want to push the power densities
the amount of power you get out of this

direction ethanol fuel cell
higher methanol crossover

hurts you in terms of power output.

So what happens.

Well first let's talk about
the evolution of structure.

In our Nafi on film when you
cast it from a solution.

Nappy and does not dissolve.

In solids it form some kind
of my cellar solution.

And I represent that by these rod
like particles in the solvent.

You tend to take this polymer
solvent solution cast it into

a film on a glass plate and you start
to allow the solvent to dissolve and

in this case we're using methyl
a set of might as the solving.

It starts that it starts to evaporate and
lose solvent and you create these

us Fredricka domains where the charge
remember there's still acidic charge here.

And so three H. that charges on
the outside of the spheres at some

point during this evaporation process.

There's a phase immersion.

Instead of being solid spheres with
charges on the outside surrounded by

celebrant it now becomes holes.

All right where the polymer now is
surrounding this solving here the solvent

we're surrounding the polymer here the
parliamentary now surrounds the solvent in

these what people think are spherical
domains what the charges now facing inward

when you finish drying this membrane
you have to go through one more.

Kneeling step.

The membrane itself after casting has very
poor or mechanical properties brittle.

All right.

So you have to raise the temperature
hundred fifty to two hundred degrees and

you carry out and dealing step during that
annealing step you create Crystal limited.

So if you're on a semi Crystal in polymer.

And so

it's the crystal limited I give it a lot
of its attractive mechanical properties.

So you have to go through this
a kneeling step when you do that.

You can no longer dissolve
this film in this solvent.

You have to go through a long complicated
high pressure high temperature

procedure to get that
polymer to re dissolved.

Also this particular polymer.

Now has the proper mechanical
properties of proper

conduct trim properties that you would
expect from a commercial Matthew on film.

So our idea here then is
to take this membrane and

to go through these steps and just prefer.

I merely we're going to stretch.

We're going to leave a little bit of
that D.M.A. see solvent in the membrane

that will plasticized the polymer and
allow us to stretch it.

And then after stretch keen
we're going to and nail it and

when we a nail it we're going to
have formed Crystal lights and

that's going to lock
the structure into our membrane.

So this is very different than
taking a commercial piece of Nafi

on in stretching it because if you do
that you get in the last the defamation.

It'll bounce back put it in hot water and
it will relax back it's to

its original position and
our case we are stretching it and

then a kneeling so
that the structure becomes permanent.

So that's what we are going to do.

And so here are some of the procedures.

Right.
We make a solution of may feel we

cast it into a Teflon dish.

We partially evaporate the solvent.

Then we heat the film to one hundred
twenty five degrees and we start to

stretch it to a desired drop ratio final
dimension relative to the initial length.

We keep that in the stretching frame
to fully evaporate the solvent.

Then we kneel that film and
then we can barely a lid to

soak it we could do whatever we
want the structure is permanent and

the structure is different than
in commercial masking out.

So the methodological
changes are no different.

Now what we did originally is we did
lots of electro chemical studies but

I'll cut to the chase.

All right.
We did lots of lots

of chemical studies and measurements and
then we looked at T.M..

But I'm going to give you the T.M.
picture right now here we are looking at

a cross section of the membrane
in the stretching direction.

These are T.M.
membrane samples that were died with lead

acetate the dark domains
are the hydrophilic

domains that I remember I showed you
those hydrophilic domains those fears.

So here is non stretch Nafi on that
this is twenty nine of meters.

And you can see these domains.

And as you stretch here were doubling
the size of our film draw a ratio of two.

All right.

Here we go from two to four to
seven this work was carried out.

At the fuel cell center
in Vancouver Canada and

what you see is that as you stretch
the size of the domain increases.

And the number of the domains increase.

So we have changed the Nano
morphology of may feel

when we stretch this
morphology is permanent.

It doesn't go away.

All right.

It doesn't relax back past
because there's a crystal in.

Palmer that's keeping that
that structure in place so

what does a suit two properties.

Well I hear we have.

Conduct heavy versus drag ratio
that's the final length compared to

the initial lengths.

And we also look at
methanol per me ability so

this is through playing conduct a pretty
measured in water twenty five degrees and

method are permeable it's sixty
degrees it's in the diffusion So

we see how much methanol
goes across the membrane.

What we see is there's almost no change
in the conduct of any when we stretch.

But there's a drop in permeability and

that drop plateaus out it looks like
around a draw ratio of four or five.

So what we've done here is we haven't
changed the proton conducting properties

of the membrane but we have changed
how methanol gets across and

that's exactly what people want.

Normally what you see in conduct to me
is that the conduct of any will power.

Well this permeable of the line.

So you save really nothing if it's
a better barrier for for methanol but

it also becomes a better it becomes
a barrier for protons also so

this is a very unique combination of
properties for fuel so membranes.

What else do we know where we thought
OK maybe it swells differently.

So we took a membrane and

we put it in water and we looked at
the swelling gravimetric swelling put

in water see how much water absorbs and
what we found is that then the membrane

swelling the weight of the wet minus dry
Oberle wet doesn't change when you swell.

So we're not really seen a difference
in solving a take by the membrane.

Whether when we look at
partitioning of ethanol.

We see this sort of thing for some that.

There's a decrease in methanol
partitioning scription in our membrane

when we stretch.

So this is the ratio of the membrane
phase methanol concentration as a as

compared to that in the external solution.

And we see this drop in this drop from
one point five or more smear it's

exactly the drop in the Permian billet
permeability is the product of diffusive

beauty and solubility So what we're saying
now is the reason why methanol doesn't get

across the membrane because it then
getting the membrane can't get in.

So then the question became
well why doesn't it get in

how all of these swell pours
somehow blocking methanol so

we decided to look what happens
to water in these membranes and

we did some low temperature
differential scanning calorimetry.

And that shown here.

Well temperature D.S.E.

we took a membrane soak it in water drop
the temperature to minus fifty degrees.

In a D.S. C. and we slowly raise
the temperature and we see it.

What temperature and how much heat
is absorbed to melt the water.

Here's typical Here's a net commercial
Nafi on sample right around zero degrees.

You see this endothermy here that really

represents the heat of melting
of the water in the membrane.

When you go to a dry ratio of two three
four seven the methane peak disappears.

Melting peak disappears means that
the rotor never melt and never froze so

good with water in our membranes
that are hydrating these so

phonic acid S o three eight sites in
our stretch Nafi on does not freeze.

All right.

And in fact if you look at the freezer and
non freezable where

order you could see that it was the most
of the order for a draw a ratio of five.

Does not freeze in or
membranes doesn't freeze and so

what's happening here is that you have
these large spherical clusters with so

phonic acid groups surround on the inside
of the spherical shells and you have

hydrated water that's water that forms
a hydration sheath that's oriented Roder.

That water doesn't behave like water.

All right.
It's interacting with the yes Citigroup's

and.

When you go to stretch material.

You make smaller clusters
smaller spherical clusters and

a higher fraction has a better absorbed
water remember the total warders the same

the total water here is the same this
is where dispersal phonic gas a group

the total rotors are the same but

the ratio of the bang to free where it
goes way down in our particular system.

Now how does this affect
methanol Well them and

I nam equally weirder is preferentially
going to hydrate a source phonic acid site

as compared to methanol So
how does Bethnal get in in this

configuration methanol gets
into this free water regime.

And stretch Nafi on it's in the spray or.

Course shell hydrated clusters right or

Bethnal will not replace this
water of hydration don't I.

Now McLay not going to happen and
therefore we're back.

We propose we created this barrier for
methanol absorption which then decreases.

Which then decreases
the overall permeability

we see as expected when we stretch.

We get more crystal
limit be in the polymer.

We also see a.

Increase in the mechanical properties and
we have to say we have not yet

done the experiments.

See if those mechanical properties
translate into a longer life

time of this membrane in a fuel cell.

So how do these membranes work in
an actual fuel cell here is typical

yourself performance data so
voltage versus current density.

For one method all five hundred
cubic centimeter person

minute air at one atmosphere.

We took three layers of our
films we put them together

we stacked them together we could also do
it is one thing membrane that stretch but

we took three steps films we put it
together and here is the performance data

the open diamonds here
a commercial Nafi on.

The open triangles area commercial
Nafi on with a different about.

The clothes diamonds here
are commercial Nafi on and

they just had different
amounts of catalyst.

And what you could see is that
these three layer systems shown

in red have much higher performance
then what you see in blue here.

And in fact if you look at the power
density of point four volts

you get about a fifty percent
boost in power density and

the reason for
that is twofold one is this membrane

is much thicker than a thinner
than a commercial Nafi a membrane.

And so there's less resistance to Proton
transport proton conduct a beauty and

we have a membrane that has
a little permeable would be so

we have a membrane here with a lower
resistance to Proton transport and

a lower priming ability to
methanol the net result

is you get this big jump
in conduct to fifty.

I mean in power density I'm sorry.

We look at thickness of facts.

Maybe make the nephew on thicker
maybe make the Nafi on thinner.

Maybe eventually it will be as good
as our stretch material not true.

Here is power density in our fuel cell
as a function of membrane thick this.

And here's commercial Navy on
samples they were all around forty

milliwatts perscribe centimeter
under these particular conditions

here are the stretch films and
these are different batches.

Some are single film some are multiple
films where what you see is a mass

maximum an optimum thickness of
about one hundred fifty microns.

If you make the membrane
even our membrane too thin.

Then do you get too much methanol going
across in the and the and the power drops.

If you make the membrane too thick.

Now you block methanol but resistance
to Proton transport goes up and.

The power density goes down again and
so you can see this quite nicely in our

system you see it in commercial maffia
on these things don't cross and

lift your way out here maybe and you can
see this big boost in power density.

So let me conclude.

First let's talk about the nanofibers
we have an entirely new membrane

fabrication scheme.

We're going to make these fibers first and

then we're going to fill in the space
between them with an earth polymer.

We decoupled although I haven't shown you
the results we had decouple the mechanical

properties from the transport
properties the transport properties of

our membranes are very close to
two three M. membranes right.

But the mechanical properties are a whole
lot better because of mechanical

properties are dependent on our inert
material and we have a very stiff tough.

Material that keeps the membrane.

Mechanical properties high.

It's a new platform for fuel cells.

We haven't looked at it for D.M.
AC or for alkaline fuel cells but

that's something that we plan to
look at this P.F.S. a sulf unaided.

They prefer also fine a cast and sofa need
to pos films has made a two thousand and

ten deal we target which for

us is important because we
get money from the deal we.

And the next steps here.

Is to continue to try and
further improve the conduct to Vittie at

relative humidities even
lower than fifty percent.

And so we are creating nano
prosody fibers remember

we believe these fibers from P.F.S.
say with polyethylene oxide

we can leach that out and
create prosody within our nanofibers

that prosody now can perhaps be used to
trap water be a capital airy condensation

we could also create core shell
fibers you can coast Spain

to polymers you could have a core of
one polymer and a shell of another.

So you have in

that type of configuration the core
could be rather sizable you remove it.

Now you have an empty core down your
nano fiber that can trap water or

the core can be impermeable
hydrophobic polymer and

which case now the hydrophilic
polymer only becomes an annular tube.

Alright rather than a complete fiber and

in between those things we
think we can do some have

some effect on the overall swelling in
the comic to Vittie that sort of thing.

For appreciation after yon we
get the excellent membrane.

I mean fuel cell performance
in the methanol fuel so

those that they did that I showed
you were some of the highest.

The M.S.C. data out there.

We've gone as high as two hundred fifty
milliwatts pre-schooler center centimeter

at eighty degrees and for those people who
know anything about direct methanol fuel

cell that's a pretty high number and
the reason

why you did that is that we have this
new morphology and our membrane has.

Allured.

I only drop a resistance and a Unless

methanol crossover then typical membranes
and that combination is very unusual.

Our next step is to look at low
temperature proton conduct to Vittie

remember I showed you in the D.S.C.
financial scanning calorimetry data

that a lot of the water in our
stretch an Afghan doesn't freeze.

If that's the case maybe we'll
have high contact to Vittie.

Way not the fuel cell is sitting
in an automobile in Minnesota.

All right.

Because the world is not going to
freeze even in minus fifty and

so we've done some criminal he studies and

yes we get about a forty to fifty percent
boost in the low temperature conduct to

Vittie of our membranes and that's because
the water doesn't freeze in there and

that's an important consideration
under for practical applications.

So we have to investigate that a little
bit more and then we have to it.

We know that we get improved accountable
properties where we don't know how that

translates to longer do ability of
our membrane membrane will survive

former cycling on and off thermal and

humidity cycling we have yet to see
that Nafi on fails after a certain time.

In this kind of cycling
maybe are stretched now if

you know on higher Crystal entity better
mechanical properties will do better.

So with that I want to thank.

U.S. Department of Energy who funded.

The nano fiber work and a whole group
of people in my lab and elsewhere.

Pat Mather who's now at Syracuse who
helped with the Nano fiber work.

Richard visit Gen Lincoln mentally.

John Choi.

Lewis root and a couple of undergraduate
and undergraduate and even a high

school student who worked on these
various aspects of the two projects now.

Now.
All right

some work on one some work on
the other and also Ken she.

At the N.R.C.

laboratory in Vancouver who did the work
on stretch Nafi And so with that.

Thank you.

And I be happy to answer questions.

Thank you thank you.

OK.

Right.

Good question.

I mean I put cost at the bottom of
that list because it is important but

at this point in time.

It's not a prime consideration but I will
say that there are companies right now

that spin that use electricity in to
create commercial products have a filters.

OK.

Support layers.

You can make they use the nanofibers
as the support material

for various kinds of membranes for

example there are industrial processes
where you have multiple fibers.

I mean multiple spinner rats
with multiple fibers and

they can make very large sheets of
this at relative Lee moderate cost.

OK.

So we haven't done that cost analysis but
we can point to websites and and

commercial entities that
are making consumer products and

other types of products using the lecture
spinning so we don't think it's

the worst thing in the world there
are a number of processing steps here.

We're concerned about
that when we do P.F.S.

say for example the cell
phonic acid when we compact.

It will be at the same time so
we compact and weld in one step.

So that saves some some of the cost for
example of the post spinning.

Methods.

So we're aware of cost.

We think it's going to be OK.

First we have to make the get the right
properties them will start our

own company.

And try it out.

Yes.

Good question.

We don't know yet.

And that's what I write when I
one of my students is working on

the goal the question is what is the
optimum diameter of the nanofibers for.

High proton convict to fifty.

So the question is do you want.

First of all we can get down to maybe
forty fifty nanometers right now.

I showed you one hundred nanometers for
poly cell phone with the P.F.S.

say they're more like three or

four hundred nanometers we're
working on shrinking those fibers.

But the real question is do
you have very small fibers and

lots of them so that the volume fraction
of the fibers stays at seventy or

seventy five percent or do you have
bigger fibers in fewer such that

the total polymer P.F.S. a polymer still
occupies seventy or seventy five percent.

So if you were big fibers
are many more smaller fibers.

We don't know we don't know from
a torturous to the point of view and

cross contact points you would like to
have a greater number of smaller fibers.

But we don't know if there's going to we
would like eventually to see if there's

an inherent change in the Proton conduct
pivoting if we get to fibers that are very

very small we'd like to see an enhancement
of proton conduct a wity maybe we'll

assess finding Kasa groups line up on
the exterior surface of our fibers.

Now we have a surface diffusion.

Effect maybe that will enhance proton
conduct to Vittie in some way or

there every point in word and

they'll be concentrated we
would like to see in effect.

We haven't looked yet so
but it's a good question.

We have this the question is to have
a happy look at the very end Taishan of

the fibers right now they're random.

And we do know that the conduct
liberty is isotropic.

So in the plane of the film and
also perpendicular to the plane.

We see the exact same proton
convict a body which tells us that

we have this random three D.
network of fibers.

Ideally of course we want to fibers
just a point straight through.

The film with a short is distances perhaps
as possible but when there is no way

that where where off of doing that in
the selector spinning apparatus this fiber

whips around when you're Electra
spinning and it has its own mind.

We have not yet
oriented it in any way we can change

the rotation speed of the drum in which
case all the fibers will line up.

Parallel to one another but

we don't want that because they have
the power where it running the parallel

to one another in the surface plane of
the membrane they don't cross over.

And so that doesn't do as any good.

So right now random orientation.

Was.

As you recall use that
question was how how how

then can you make the need
a membrane when you draw it out.

We get down to twenty thirty.

Microns.

For the for the membrane after stretching.

But as you notice everything kind
of dies out after a drop ratio for.

Right.
Everything plateaus the permeable of

the plateau as we see
the solubility decrease plateau.

We're not exactly sure what's happening.

I mean part of the problem is
as you stretch this membrane you

create Crystal limited
you don't have to anneal.

Stretching will create crystalline
that you line up polymer chains and

there will be a Crystal Crystal ice.

So we don't know if we're just
not stretching fast enough.

And that Crystal entity is setting in
by the time we get to a draw ratio for

and we're not seeing any further
decrease further change.

OK we do see a change in the Nano
morphology going from for a drop ratio of

four to seven but that doesn't seem to be
affecting the properties of the membrane.

So we can stretch it more but

we don't see any beneficial
effect at this point in time.

Even the mechanical properties
tend to plateau out.

Now we're trying to do stretch
a faster restructure faster maybe we

can do it before.

It crystallizes but stretching a faster.

You have to be careful because
any cracks are then and.