So it's a pleasure to introduce today
see her also right here in this

house her masters and working in this.

City.

Spent many many years decades and
Bell Labs are in.

Labs where she rose to the to
the right of fellow Bell Labs and

director of the cereals research
we're seeing a lot of the.

Informal research that the graphic resists
she came to Georgia Tech in two thousand

and eight was currently fresher in
the school chemical like the engineering

points in chemistry and
biochemistry to real science.

In Nice ninety five like
the national engineering.

She is was the President of you know and
society in two thousand and three and.

The wars from the US She was also a C.S.
fellow as well as a fellow

of the AAA the Royal Society so
your question.

David thank you and it's really
a pleasure for me to be here and

I can you remember being giving

a seminar almost precisely two years ago.

Where And we had a snowstorm
that was just beginning so

we missed it this time but
the timing is kind of interesting so

hopefully it won't start stowing out
there and I think it is a little but

it's warm enough that it should I so
can everybody in back hear me all right.

Yes Good all right so what I'd like
to do is talk a little bit about

research that's going on in my group but
also collaboration

with a number of other groups related
to kind of developing structure

process property relationships for
the solution process.

Semiconductors and
the motivation of this if I can work

this properly is that yes
silicon technology and

basically in our Gannett materials
device technology has been around for

a while right so starting with
the first germanium transistors

that was developed at Bell Labs in
the one nine hundred forty S. timeframe

things have certainly advanced in terms
of initially micro fabrication now and

then a fabrication where silicon
devices now have features that are.

Approaching ten fifteen meters in size.

Just to put that in perspective when
I started working in the area of

foetal is a graphic materials.

It was a really big deal to have half
micron or quarter Micron images and

with a relation to let's say you
call a bacterium cell that's

about of my crime by five microns across
those features seem pretty small and

then a human hair is roughly one
hundred microns in diameter so

silicon devices today.

Are really the in the feature
size is a really tiny.

It's really costly to fabricate

those devices considering that it's
all subtractive processing for

the most part on very large
single crystal silicon substrates

Yeah there's a lot these devices can
do that every application doesn't

need the power of silicon or
other in our gadget materials so

that led to interest in Watt.

Typically it's been called
plastic electronics or fly.

Tronics can we now do something with
devices made out of flexible organic or

polymer materials use them
on flexible substrates and

they could be effective for things that
don't need silicon technology maybe for

a display that's going to be conformable
a display that could be rolled up

and lightweight new just
bring it with you very easily

solar cells that could be
rolled up in laid out for

a power source at some point Medical
by medical devices integrated sensors

that could be stretchable flexible and
biocompatible All right so

that led to a lot of
interest in developing.

Organic.

Semiconductors.

And there are a lot of things that we
need to think about in the development of

organic materials certainly
I don't want to just sorry.

We need to think about the material
something move over here.

How do you design the material to have
effective charge transport characteristics

of the devices are the actual
configurations of devices and

the device designs going to
be effectively the same for

organic materials versus
in organic materials and

then the processing how do you
actually control these large

organic floppy molecules
in solution as solvency

evaporating to provide the morphology
that we're going to need for

effective charge transport I and
initially.

There's a lot ton of
work in designing new me.

Tiriel says in looking at
the device this except.

Much less work associated
with the process.

Because most of the people doing this war.

Largely were applied physicists.

From the chemistry perspective the
materials perspective most of the people

doing the work wore organic chemists
who like to synthesize things and

there'd be a lot of beautiful synthetic
procedures lots of neat molecules being

prepared but in very small quantities and

results from one lab to another lab
just were not reproducible I and

that's largely because
the processes work troll.

And so I came to looking at our getting
some of conductors more from the process

standpoint based on the experience I
had with lithographic materials and

processes and
developing systems that could be used in

a manufacturing environment where
if you think about how silicon

devices are fabricated there
has to be a lot of control and

precision associated with those
processes and so can we now

develop the solution based
process sees required for

organic materials to really behave

reproducibly and.

Hopefully be developed into
a manufacturable system.

All right so I mentioned before that there
are a lot of possible opportunities for

polymers in electronics and
photonics where the biomedical sensors.

Possible displays stretchable
electronics we could use.

Relatively accessible inexpensive
roll to roll processing

techniques that were
cation techniques that.

This particular.

Piece of equipment is a groove your
printer the way newsprint is being printed

every day right so it's inexpensive.

Low cost very high throughput and
you could use very large areas

something that you can't do with silicon.

The issue though is as you can see
in the say a feminine edge sort of

what the morphology looks like and
there really isn't any firm basis yet

for knowing exactly what
morphology we really do

want to have to optimize
performance in a reproducible way

right so polymer materials are flexible
devices we wanted to develop

the fundamental structure process
property relationships so

that those relationships could God
the design of robust materials and

processes and
we also wanted to identify and

utilize sort of the fundamental mechanisms
that are associated with thin film

morphology evolution from a solution.

So thinking about that.

You know first thing is
to think about order and

disorder where if you look at these A.F.N.
images each one of them

shows a then from off ology surface before
all A-G. that's slightly different.

I and

if you had different materials they're
OK they're different so they're likely

to behave differently in this particular
case they're all taken there are.

Done using the exact same poll over and

not only just the exact same poem
are from different back Chiz the exact

same poem are from one
batch of material taken

from the same bottle of material
just process differently

a couple of them may have been just
processed on different days right so.

This says right up front that
the process characteristics and

how you process the material.

Makes a really big difference in terms
of what you get in that ten film and

that makes a big difference in what the
performance is going to be all right so

with this solution process
material what is the mechanism

of conducting channel formation I had
do we control the microstructure.

What is the role of crystal on a T.
do we want a perfectly crystal material or

do we want a partial partly
course full of material and

then in terms of conjugation for these
polymer chains you know we're thinking

about Inter chain conjugation and

in try to chain effect so
which is more important and

how do we balance all of
these different factors.

So there is.

So Palmer the semiconductor polymer
properties very much depend

on that final thin film.

That morphology is process dependent.

And we don't really know how
that morphology develops.

As the thin film is forming.

So microstructure having an impact
on semiconductor properties

in Palmers has been studied for a while.

Right so searing House talked about radio
regularity or how the monomer units

in this poem are arranged having
an impact on performance.

Klein talked about molecular weight
effects where low molecular weight

material was more crystalline but
didn't behave as well as

higher molecular weight material and
Kim and his coworkers looked at

functional izing dielectric
surface of these transistors and

that making an impact in the organization
of the Palm R. and thus performance.

Art So
what is the role of microstructure and

then how can that my constructor
be tuned reproducibly so

that we have a manufacturable process.

This doesn't want to cooperate with me.

So if we look at conducting
channel formation we have

a thin film transistor architecture
where are we're using and

silicon as the gate electrode.

Silicon oxide is electric We have
gold source drain electrodes and

we're depositing in our semiconductor
we start with a solution

it undergoes a number of
phase transitions and

we have our device Ideally the palm
of chains Roll be stacked up nice and

neat along this channel and
we have perfect transport

characteristics that's
the ideal case not the real.

So we started looking at
the process in solution and

we were taking drain current measurements
as solvent was about parading and Wi-Fi.

I think that the drain current
fluctuates as that film is forming.

So the Palmer chains are rearranging
is a function of time

they're percolation affects
their bulk effect so

they're taking place in the film their
interface is facts with the substrate and

all of that is going to influence how that
final microstructure evolves right so

looking at that system
simultaneously from the electric or

perspective coupled with
spectroscopic interrogation and

working with monster to sorrow
from material science here

we developed a way of being able to

interrogate the center of
a device at the same time that

we're looking doing Raman
spectroscopy measurements.

To try to correlate these changes
in the semiconducting performance

spectroscopic Les was what's going
on in the Palmer system in solution.

And use poly three hundred
five being the solvent was

Tricor benzene not a very good solvent for
manufacturing environment but

it's good for
studying this particular system and

the reason we did that is it's
a high boiling point solvent so

we don't have to worry about any dynamic
effects of the solvent evaporation

from that droplet right and
the time was set so

that we could actually
accumulate the Raman spectrum.

So what we find from the conducting

channel measurements is that
you know it takes about twelve.

Fourteen hours for this droplet to for
the solvent to evaporate and

have a solidified drop what.

About ten hours in we
see a very steep rise

in drain current and
you know there are distinct areas

in this rise where the behavior
is a little bit different and

if we look at the robin spectra
at the same time we can do.

Polarized Roman spectroscopy and
based on those

measurements we can see that there
appears to be a liquid crystal in phase

that the system is going through at some
point prior to having a solidified film

we also deceive distinct differences
between the isotropic solution and

ultimately the crystal in phase
when solvent is about parade I'd so

we do have a light tropic liquid
crystal unphased that's being formed.

There is long range order
with in that L C phase and

so if we can actually control and take
advantage of the little crystal energy in

the system Are there
positive consequences for

macroscopic charge transport.

Right so
we started looking at the liquid crystal

characteristics of poly so
we have five in solution and

then this this continued to be in
collaboration with Mohan service our.

So.

Yes we see that during solving
about peroration we do formalize

Crysler in phase we can take
a solution of P three H.T.

approximate the concentration that we yes.

Anticipate seeing this L.C. phase.

And just let it sit around.

Over time.

The solution changes.

We can see that there are lower
energy bands being formed

over in this case a three day period and
these absorption bands

are indicative of age aggregates
being formed so that we're seeing.

P two H.T. aggregating into structures.

Polarized offical my cross copy shows
that the age P three she solutions

have long range order and
monitor main characteristics right so

the aggregate those
aggregates are elongated and

they order in solution.

That part is partially good from a sort
of device fabrication perspective

it's actually not totally good because
a solution that changes over time

isn't going to be particularly user
friendly in a production environment

so what else is going on and
how can we now control our IT So

this just demonstrates again that
over a period of days in this case

thirty six days we do see
sort of aging induced

by our friend Jim's showing
mounted to mainline character.

By our friend has evolved over time and
it becomes strongest.

Around about a month into this
aging process all right and

we can do dichroic measurements
again the come for that

these are ordered elongated structures.

All right so the.

That's an issue that going forward is
going to need to be dealt with but

we do have aggregated
structures that can be ordered

Can we now develop process techniques
to influence that ordering and

maintain ordering on a map or
scopic scale so

that we can manipulate influence and
control.

The charged transport characteristics for
a device right so there are number of ways

that we've been looking at in the group
to sort of aggregate poly three X. five.

These include On occasion I think
the batteries going on the point

are ultrasonic ation using coast solvents
we've shown that we can use low dose U.V.

radiation to induce
ordering in aggregation

we can combine these methodology and
we've also explored

the use of Hansen solubility
parameters to try to optimize.

The solvent with
the semiconductor poem are to

have a system that maybe will more
often really form these aggregates

in an interconnected way for
macroscopic transport Alright this shows

Krista lenity can be manipulated
by changing the Salva and

this shows the kind of analysis you can
get using the Hansen solubility parameters

all right and
you can get a nicely organized.

Very high fiber content
thin film morphology.

By manipulating the material in solution.

So we looked a little bit more closely.

At P three H.T. aggregation and the impact
of a couple of techniques particularly ad.

Poor solvent.

Coupling that with the use of
ultrasonic ation to induce aggregation

art and
by changing the actual process conditions

you can still get changes in
that overall morphology and

some of those morphologies are better
some are not better for charge transport.

And that allowed us to start thinking
about the aggregation process

and the evolution process for

the ultimate thin film in
terms of Chrysalis ation.

We could go back to Chrysalis
ation of polyethylene where

it's pretty well understood
how polyethylene crystallizes.

We really need to think about
sort of the concentration

of the polymer in the solution.

Whether it's in the under
saturated regime and

on stable zone where everything
just falls out of solution or

in a medicine Abel's alone which is
where we most likely want to operate

to be able to control that
aggregation step right so

that led us to propose
a relatively straightforward

two step crystallization mechanism for
peace three H. T.

during the solvent
evolution process which is.

A process that's been investigated for
many poem or

Crystal in Palmer materials in the past.

Then we could take build
on that system and

this is work of a student who
was visiting from China up until

last October where we could
now think about that Krista.

Ization process.

I think about the temperature of how
we're treating that solution and

how we further.

Manipulate the system and
he developed a microfluidic approach

where we use a cooling bath
to induce nucleation and

then Using U.V. light we can grow

those nuclei nucleation sites into
larger fiber or structures and

then also we have a shear induced fly That

also optimizes growth of those spiders and

with that process we were able
to develop solutions that had

a very long integrated polymer fibers

and we could optimize
the process in order to get

very high mobility as in
politics we have five thing which

averaged about point one six
centimeters square propulsive.

Not great if you compare it to silicon but
for

many applications like
displays where you don't need.

That high immobility
it's perfectly adequate.

If we do sort of U.V. visit analysis and

you accidental surface we
can see that we definitely

are forming very well ordered
aggregates in the solution.

Those aggregates from the analysis
are also very highly crystal on.

And if we look at the sort of the this
is U.V. analysis the Exxon band where it

goes down which is an indicator that
there is a much longer conjugation line.

In the systems.

In the pipe by stacking distance of the
polymer backbones decreases again good for

charge transport and
the Herman's orientation factor.

Approaches almost perfect
perpendicular stacking of the poem or

backbones to the device
substrate all of which says

that we should have
a system that has better

charge transport characteristics
in Transistor device and

the results point to yeah
that's what we've got.

Right.

That process is also versatile.

This was an optimized
it also was successful

in promoting the church transport
characteristics of an electron

transport material not just
the whole transport material in so.

I think what we are thinking about
in terms of the crystal is ation

mechanism in terms of how.

The fibers grow and
Aryan in solution is applicable

not just a poly three X.
five but is also applicable to

alternative polymer systems that
may have better performance and

also may be more durable than P three H.
tedious.

I would go on further and

show that with that
microfluidic approach zero.

We do end up with materials
that are ordered ordered

based on polarized optical
microscope you results.

And we get that long range Orian Taishan
underflow conditions and then.

If we start to look at
the fiber structure and

do a little bit of image analysis
on the AI famine midges we see that

we have what could be called
bundles of fiber in solution and

that made us think about
what's known in the poem or

chrysalis ation literature as a shish
kabob mechanism of fiber growth where

with these fibers there are other parts

pulp to other parts of the polymer
chain so I have an elongated section of

the polymer chain that additional
fibers are growing from and

these so-called tie chains could
be leading to this bundling effect

i N. this is something which we're
continuing to investigate further

largely through a collaboration
with Martha Grove are also in

the chemical engineering department.

So this is all been fun.

Weekend manipulate through processing
we can get better performance of

P three inch T.V. We think we can do
this on other polymer systems but

there's still a problem
with these approaches that.

The solvents.

Are typically chloroform Clora
benzene tri Clora benzene.

And nobody in their right mind is
going to want to use those solvent

in a large scale high throughput
fabrication environment

so what else could we do.

So in a.

In a recent project with
Paul Russo in M.S.C..

We started looking at.

Proteins.

Which can be very nicely
dispersed in a quick environment.

And can those proteins and
now assist in the organization.

Of a conjugated polymer like P three.

So that alternately we may
have some sort of a quiz they

are almost latex like paint
processing approach where

the protein helps with
the assembly of the polymer.

It's all nicely dispersed in water and

then you can deposit
from effectively water.

It so we started looking at a fungal
protein called Serato Allman.

That effectively is
a Janice like particle and

it forms very strong biofilms
when dispersed in water it's but

no net it will also in capsule
a oils solvents accept or

so if we take POLLOCK We
have seen in a some solvent

dispersant in this.

Aqueous based protein mixture.

P. three H.T. gets in capsulated over
time those capsule sort of merge and

formed and Riddick structures we can
get crystalline structures forming.

And.

It sort of points towards the future of
one with yeah we Nabi able to develop by.

Compatible systems we may be able
to develop systems that can be.

Deposited from and aqueous medium.

These materials could also be handled
with an additive manufacturing technique

where we could specifically
deposit the active material.

On a subsidiary that's moving so
there are a lot of different things that

we can now start thinking
about given that we can form

very ordered crystalline P three each T.

Frew the protein assisted method.

We can also do some other things related
to blending the poem or as with.

Alternative materials our
One example was if we take

polyps we excell five scene.

Have a highly crystalline where
we do see enhanced mobility

The problem is that it's not very
mechanically stable so yeah you're

not going to be able to use this for any
kind of flexible stretchable device Well

you take one percent of that
pretreated poly three X.

will fly if being blend it
with metal Psylocke saying

during that curing process
we end up with a composite

system that is very stretchable
right the interesting

thing about that system
also is that in that bland.

The performance of.

P three.

Is substantially higher.

So the P.D.S. mass which is a non solvent.

Doesn't like to interact with P three.

H.T..

Forces peace creates T.

in to more order to line
structures that exhibit

very high mobility right if you just take.

P three H T that hasn't been
treated mix that with P.M.S..

You get no mobility.

And if we look at that from air famine
edges of both the air interface and

the substrate interface this is
straight piece that's been processed

no other additives and
the air and substory interfaces.

Are pretty similar to.

The P.M.S. system though the air
interface is almost entirely P.M.S.

while the substrate interface
shows this dendritic fiber says

network of peace three H.T. So the P.D.

A mass is influencing
the organization of P three P and

fortuitously it's also
enabling that organization in

a much better connected
ordered structure than.

The.

One hundred percent material so the steps
that we have to take here now are to show

that to integrate this with
an actual stretchable device and

show that upon stretching We also
maintain the performance characteristics.

And then in some additional work.

We also looked at blending
radio regular P three eighty

which is P three T.V.
that does have church transport

characteristics with radio random
piece reached the which the.

Doesn't really transport charges are and

in developing a a process of being able to

use this system on a flexible

polyethylene tariff that a light substrate
basically mylar that used to be used for

overhead transparencies were able
to develop a flexible device.

On plastic that maintains its mobility

up to about eighty eighty
percent of the radio

random material versus
the charts transformative

aerial and the device characteristics.

Are better.

Than straight P three H.T..

In addition to that.

We can have a fairly tight bending
radius and be able to maintain.

Integrity and device characteristics and
these are the transfer curves.

I also have to say that I had a very
patient student who did this work because

he did go up to about a thousand bands and
as I understand

it this was not a mechanical device set
up to do it by itself he very patiently.

Bent the system himself so and

the results are pretty impressive
that by again by taking advantage of.

Sort of the solubility characteristics
of one polymer in another polymer that.

We can manipulate.

The organization of the charge
transport material within that matrix.

And they have a positive impact
on the mechanical properties and

the device substrates that we can use it
so from the standpoint of flexibility.

On a variety of different substrates
that can be easily addressed

by appropriate choice of
a blended polymer system

I think as we go forward what we need to
do is look at alternative solvents so

that we have systems that.

Aren't going to be hazardous

to somebodies health of thirty using
it in the fabrication environment and

that's going to require both chemistry
in terms of the design of the Palmers so

that they will have the appropriate
solubility characteristics but

it also needs to be coupled
with sort of the process.

Science and engineering approach of
optimizing how we actually develop

robust processes for using these
materials in fabrication Vironment and

then we also have to look
at how we actually prepare

these samples and are the fabrication
methods also scalable so

can we really develop world
through a process can we develop

additive manufacturing processes I and so

there are a lot of exciting opportunities
really involving not only the science but

also the engineering of
conjugated systems for devices.

So really it's a combination of the
molecular structure with the processing

that is going to influence the electronic
properties of these materials

I think we do need to work
in a concerted way where.

It's not just one piece
of this triangle I think

it's thinking about all aspects they're
involved in device fabrication.

And we also need to understand.

But more importantly control all of
the interfaces in these materials

whether we're talking about interfaces
between the conjugated Palmers of selves

interfaces with the dielectric
substrate interfaces with conductors

we need to think about
all of those interfaces.

And then finally certainly.

I did not do the work myself there
are a lot of people I need to thank who

were involved in the work that I presented
today a small number of whom are presented

here and I need thought
the funding agencies and you for

listening and
I'm happy to answer questions.

If you have any

facts.

OK.

Right so.

Right.

So we actually have started
to look at that and.

A lot of the work in that regard is is
now in a collaboration between myself and

Martha Grover and
we have a couple of joint students who

are really looking and sort of the from
the process perspective being able to.

Model the actual process and
evolution of the thin film morphology

and from the knowledge that we
get there we also want to explore

can we now transfer and use those models

to start modeling the process and
the process characteristics for

let's say blade coding or
slot dye coding and

I think in Jet printing is
another opportunity so that's.

In

process.

Thanks thanks.