Universities are often criticized for
for being ivory towers.

And the implication of being
an ivory tower is that you

know what you're talking about that
you're just dreaming impossible dreams.

Right.
I mean you don't know what reality is but

but my feeling always has been that as
the descendants of the monasteries really

universities are the monasteries in
modern dress that there's a great merit

in in having dreams that
dreams are not bad and

somebody should think about what's
the best you could possibly hope for

and and in that sense universities
are very are very positive.

Aren't they.
I mean there are places where

you can you can imagine not what is but
what could be.

And so that's what in the department here
with the strong component of membranes for

for chemical separation step.

That's what I thought
I'd talk about today.

I talk about what kind of dreams
What's the best you could hope for.

If if the membrane really work.

What would it do and if it worked
that well could it work better.

I mean how how far could you
push this particular idea and

it's a vehicle for this.

I want to talk specifically
about block copal ved.

And block O. Palmer streamlet curious
because they provide self assembled faces

on a net a meter level that it's hard
to imagine achieving in any other way.

And you can see this if you look at
that colored thing across at the top

that's sort of a series of little cubes.

And if you look at the one
at the extreme left you have

a sea of red with little drops of blue.

Well that's what happens when
you start out making a polymer

molecule that had seen it once
more all segment of blue.

Polymers generally don't mix because
of the large entropic penalty for

doing so and yet if you mix them together.

What will happen.

They're chemically covalent
Li bound together.

Now what will happen is the little
blue pieces the smallest

parts of one polymer will form
will form a different face.

You can think of it a bit like for
those of you who are women.

You can think of a small number
of women in a singles bar right.

And if you look at that the women
are not well distributed at all.

They from very small little
self protected groups.

Right.

You know what everyone goes back to OK So.

So that's the sort of phase separation and
I want you to imagine.

Now if you increase the amount of

the two Palmers again remember that
they are code they only bound together.

If you increase the amount the fraction
that is one polymer in this

case you're thinking of poly polystyrene
poly ice to preen mix together.

Then what will happen when those
two become closer to being equal in

volume fraction is that you will
automatically self assemble into Lamell

you'll get a layered type of structure and

under other circumstances one will
obtain cylinders these things

are often shown as face diagrams and
that's what's shown down here and

in this particular way of
showing the face diagrams the Y.

axis is proportional to
the inverse of temperature.

So when you go up.

That's getting cooler and this is the
fraction of for example the red material.

So I guess actually it's the blue material
here so this is this is pure red and

this is pure blue and as you cool the
system remember cooling is going upwards.

You will get a face separation just
the way you would with two liquids you.

I have at the bottom a consol you point
of a different form of a critical point.

And where you come in this region will
govern what type of self-assembly

you will get you can get self-assembly
as Will Mellor structure or

self-assembly as little spheres or

self-assembly as some by continuous
structure the so-called gyro

you can get a whole different groups of
self assembled structures and I want to

use these structures the size of which
normally is on the order of nanometers.

I want the youth to use these structures
to make particular self assembled

structures and membranes and that's what
I'm going to talk about three possible

self assembled structures first structures
that are barriers that is where

you want nothing to go through second
structures that form pores where for

example you have a way to make
completely modern dispersed Poor's and

ultra filtration membrane and

third ways in which you can get chemical
selective A-T. and in each case.

My goal is to improve over existing
technology by at least a factor of ten.

In other words this is a dream.

I'm not talking about practicality.

I want at least a ten fold increase.

Can you see what the ideas.

So basically the original
idea is to reduce permeable

it with flakes what's the easiest way
to have a less permeable film and

the easiest way to have
a less permeable film.

A perfect example is the have basically
the white stuff in the Oreo right.

You want that to be a rubber and
then these two parts to be a glass and

what you'd like is the permeable ity of
a glass a good barrier keeps your car

from rusting and the flexibility of a
rubber as in any cars as in any composite.

You want to have your cookie and
eat it too right.

You want both of those
properties at the same time.

And this is not hard to do in
these are examples of this.

This is this is a film
made with Mica Flake and

I always think of these flakes as
being especially appropriate for

this time of year because they remind
me of dead leaves matted together.

And if you imagine a gas molecule
diffusing through sector composite

the gas molecule must undergo
a very tortured type of path.

And as a result the Permian Pilkey of this
composite should be dramatically lower and

in fact it is the Permian Bill A T P
which is plotted here in this

kind of a confusing graph I apologize for
this what's shown here is

the reciprocal of the permeable of
the relative to some value right.

So so when you think about it.

This down here in the corner is a high
permeable of the unused as you go up

because I'm plotting it's versus
the reciprocal it's a low permeable and

this over here is basically the shape
squared times the volume fraction squared.

And you can see that you get a good
correlation and typically you can have

a permeable change on the order of
a factor of ten without any problem.

And a factor of fifty with some work.

Now what's good about this.

Well what's good about
this is the absolutely

still stunning fact that it's
independent of two things.

The result The reason this result is
interesting is not because Kermie

ability depends on the amount of
flakes that are there but because

the Keynsham Kermie ability is independent
of the polymer that you start with.

And it's also independent
of the size of the flakes.

I mean that's what's remarkable
that neither science nor

original polymer appears in the SO

for example I can take any polymer
introduce flakes into it if I'm expecting.

Factor of ten reduction and one are get
it not only in that Palmer but also in

other polymers and that's what's so
interesting about this particular result.

What's wrong with this result.

What's bad about this result.

What makes me want to
dream about the result

is that these flakes are utter
hell to a long line.

I mean think about it for a minute
if the flakes are aligned like this.

Then I do have the tortured pair but
if the flakes are aligned like this.

I don't have that same.

So if the frets are aligned
perpendicular to the film.

I have a perm ability of change or
perhaps ten percent if I take

the same ball Yuma flakes and
align them with the film.

I can get a factor of fifty five thousand
percent down instead of ten percent.

So I really need ways to
get better alignment.

Now I can get better alignment
with the mica flakes by using

a lacquer by having a big solvent but
nobody wants to reduce a lot of solvent.

What would be better is
to go to the drawing

at the right hand side of this chart
to make flakes that ran for ever.

By means of block O.
Palmers if I take a palm or

that self assembles into these structures
and the flakes run for great distances.

Now instead of going around the flex
I'll have to go through and

now if I have glassy flakes that
are imperfect Now I will really have

an effective barrier an effective paint
and effective way to keep anything out

using this thin film and remember it
doesn't matter which polymer I use.

So we look to start with at
Mellor structures that is

structures that consisted
basically of a red polymer and

a blue polymer and that consisted of those
things in the middle in that sort of

stacked up sandwich type of structure and
we looked at those results.

It's to see if we could get
changes in per me building now.

I confess I did not carefully choose
the polymers that were involved.

I basically took whatever was already
available on the shelf because I

want to make sure that it works
before I really start out and

sure enough if you mix up a polymer
like this in this case it was a poly

lactic acid the poly I supreme
in another poly lactic acid.

If I mix that up and cast it.

Sure enough it forms Lamell it forms
Lamell a spontaneously on whatever I spray

it on to hear the limb Ella and you can
see now the dramatic difference in size.

These are around fifteen nanometers
across where as the flakes

I was talking about before were in the
order of ten microns across in other words

I've gone down a factor of a thousand in
the size of my Flex which is good because

now if I'm going to paint some elaborate
surface with these like an airplane

wing I'm going it basically have a paint
that wraps around all of the corners and

still makes these flakes intact and now

this is I'm often accused by not of not
telling people how I made the experiments.

So the next few slides are the description
of this and if you find this talks

you know talks of four o'clock in the
afternoon are quite frankly rather deadly.

Have you ever noticed the time about

when students get jobs versus
the time of day of the interview.

If you ever noticed this.

No one gets jobs with the first
from the first interview so

if you're signing up for interviews
don't take that first slot because

the interviewer at that point
expects Christ coming again.

And is totally unrealistic
on aspirations and

as he goes on through the morning he
begins to feel a little more desperate but

frankly And
then before lunch before lunch.

You can be remembered.

That's quite true but
after lunch he's asleep so.

So there is there is no merit in that and
the best time day to interview actually is

towards the end of the day he's tired but
he hasn't found anybody that looks that

good yet and and he knows he has
to justify his trip in some way.

And so that if you look if you plot you
get exactly this you get all right well

this is the experimental stuff.

This is the experimental stuff
to show what do we really do.

And look this is really easy.

All you do is take a piece of paper.

And put it between two vessels right.

And you put in our case high
pressure gas on one side you measure

the pressure on the other side.

No no no big deal.

OK.

And here are the results.

Now.

This is this is kind of hard to understand
but it's probably worth the effort on this

side you're measuring the concentration
of the gas that is its pressure plotted

versus in this case the gases helium and
you're plotting that versus time.

And what you see this is not a big
difference between the rubber in the glass

between the ice cream.

These are the measurements for the pure I
supreme and these of the measurements for

the pure P L a poly lactic acid and
they're the data and

that solid line is what you would
expect from two resistances in series.

So the theory works.

Interesting.

Lee We studied roughly
a dozen of these systems and

in every case the theory the theoretical
prediction was that the Permian Billa T.

was a little faster then what we
observe and if you think about that.

It's an interesting result
that it should be so

general because it suggests that
the idea of resistances and

theories is not quite enough and
probably what you have is reduced chained

ability when you're going from one
of the polymer blocks to the next.

But that's one of these things
that's perfectly sensible to say and

no reviewer will ever complain.

But I certainly could
not prove in any way.

OK so it's reasonable but
completely on proven.

Now what's good about this.

Well why didn't I suddenly take off and
take other.

I mean one can imagine a liquid crystal
in block attached with the ice supreme

rubber and then all of a sudden I'd have
the opportunity for a real gain not just

factors of five or whatever but factors of
one hundred what why didn't I do that and

the answer is because the properties
of this composite are terrible.

Now why are they.

Terrible.
How are they terrible.

Well the answer is the composite
behaves very much like a glass.

Well why is that bad.

Well cuz it isn't flexible
remember what I want I want

the cookie I want to get
the inside flexibility and

the Permian built the low permeable and I
didn't get it and the reason I got a very

very big permeable of the change is
because my flakes are really big now.

So it's very hard to go around the end
of the flights because they're really

infinite.

OK but the physical properties of
the composite are terrible because

when you pull this apart.

What happens is the glass breaks and you
don't want that when you pull it apart.

Here give me your hand and I'll show you
what I mean when if we're really attached

and we pull I mean only it's going
to happen not to break your arm and

you're not a break.

OK but now give me your fingers.

If the fingers are just in or
locked and and you see the bonding

between the fingers is then it will pull
loose as that shown in the bottom here.

That's what you want.

You want it to pull not now you'll have in
the vertical direction a low per me Bill.

But you will also have that flexibility
caused by that slight that slight.

This entanglement if you like that
that slight just pulling a cork out

of a bottle and that's what you see.

OK now.

Why didn't we pursue that.

Then the reason is with the block.

PALMER The whole idea that I started
out with that's a terrible idea because

in this case the block zero Palmer with
the red in the blue blocks goes in and

out of the flake region.

You see.

So the only way I can get the mechanism
at the bottom there is to

break the covalent bonds that
make the molecule assemble.

So this dream is a disaster.

It didn't work.

OK.

It.

I can make barriers out of the mela but
they're very brittle.

They are not a success.

It's hard for
me to imagine how they could be a success.

Now what about making
ports now in making porous

on looking at what's there in the back and
now in Minneapolis a fairly good sized

city every drop of water in the city of
Minneapolis is ultra filtered before it

goes into the to the city system and
that is the system.

For for the city of Minneapolis
you do not understand the scale

the scale is the building
containing these hollow fiber

modules is one block it
has a thirty foot ceiling.

And those modules each
are eight hundred feet long.

So this whole it's just a Mensa isn't it.

So this whole thing is built up and

all the water after water treatment
is going through the ultrafilter and

the main reason is now there is
no virus in the Permian OK So

this water really is
very high quality water.

The result there's all kinds
of back flashing plush and

protocols and so forth but
maybe maybe I could take.

Membranes like this which
characteristically have one hundred

gallons per square foot per
day kind of kind of flux.

Maybe I can take that up to
about twenty thousand and

the answer to that is I probably can.

And if I could do that I could cut
the size the number of membranes

needed by a factor of between twenty and
fifty.

OK and hence cut the cost of
that enormous unit by that OK

now how would I do this now I'm going
to go to the second of those cubes.

I'm not going to go to the entire for
the entire Lamell or kind of structure.

I'm going to have perhaps twenty
five percent of one of the blocks.

OK in a sea of other blocks.

So now I have these
cylinders in this system and

now what I'm going to do is X.
out the cylinders.

Here are some of the cylinders of those
of you with an interest in in solid state

with material science this to me looks
like Crystal patterns the each of these

cylinders is about fifteen and a meters
in diameter and they can be long as as

as a centimeters in length and
apparently they're their

length the diameter ratio is constant
across the whole the whole business.

How much can we make we can routinely
make areas ten centimeters Square.

But Howard these made how are these
cylinders made because when I started out.

I'm talking about cores now right.

And when I started out I wasn't
talking about poor so much hoops that.

Well I'll get back to that and then
there's a second idea let me go back to

this the big truck problem here is
how do you get the cylinders aligned.

Remember that was
a problem with the flood.

How do you get and
the answer turns out to be

you spread out a film cast it
like you would from a from any.

Another membrane from a solvent and
then you a vapid rate use a very volatile

solvent and you evaporate fast and
what happens is

you nucleate the cylinders at the surface
of the film and the cylinders grow

kind of like imagine imagine bamboo Coming
up inside of this polymer solution.

The cylinders grow up through the solution

to reach the other side it's absolute
remarkable to me that they do this and

you know there's no theoretical reason
you're trapping them and actually

the Thurmond manically preferred region is
to have the cylinders lying in the fill.

But if you.

And if you do it with a slow evaporation
that's what happens to the cylinders but

if you do it fast.

They're vertical and then after you do it
you just drop the whole film into dilute

base dissolve one of the pallbearers and
now you've got the pores.

Now a second way to make the same
thing is don't go to the cylinders

instead go to the phase they're labeled G.

Can you see that and if you look at
that face you realize that the best

example of what this face is like for
me is a basket of apples.

If if you are a worm and you have
a basket of apples you can go across

the basket of apples eating
your way through Apple.

Right.
The whole way apples are continuous from

one side of the basket to the other and
if you were a fly you could

do the same thing through a whole
different set of channels.

Can't you both the air and
the apples are continuous phases.

We can make a block O. Palmers set that
both phases are continuous and then again.

Leach out one of the phases to
wind up with this by continuous

structure and
here are details of making the jar void in

this case you're talking about adding
cross-linking capability to the starter.

Nor boreen that gives you with the grubs
catalyst that gives you a polymer which

you showed in the upper right hand corner
and when you solution cast this it appears

to be an exactly homogeneous film but
again if you treat this for

a couple minutes in tenth molar
sodium hydroxide you dissolve out

one of the faces and you're left with that
kind of Swiss cheese kind of structure and

the advantage to that is it's isotropic
it's the same in all dimensions right.

So you don't have an alignment problem.

What's the disadvantage to that
the poor size is not exactly define

whereas those ones before had
much better defined poor sizes.

Now to give you more detail about what
the poorest were like in this case.

Look first at the small angle X.

ray patterns and you can see
This Is How You Can a dent the fi.

The base face when you start out
the patterns that you get for Lamell or

structures or four cylinder
cylindrical structures or for

gyro and structures are very different.

And by looking at the peaks
on these patterns.

You can then infer the spacing
of the self assembled

thermodynamically stable structure
the alignment is not there.

Meant in them equally stable but
the structure itself is do you see this

seems to me to be awfully hairy
if you haven't seen it before or

are you getting this you understand.

I mean it's sort of.

What are you trying to make you're
trying to make little tubes that parts

easy to understand isn't it.

But how about this Swiss cheese
business do you understand that that's

harder to get.

But that's why I think the basket
of apples as a better analogy.

Reminds me when I was a kid my grandmother
always used to used to ask me if

I'm going to go to a to a street market
and buy apples to make apple sauce.

Should I buy a big apples or
little apples which is a better buy.

OK And it's not altogether clear our
show has claimed it was better for

the by little apples and I think that
makes sense if you don't peel them and

Core them but I'm not in there that.

OK.

Now now start looking
across the top of this.

And here is the block O. polymer data at
the top here in this top graph labeled

block zero Palmer with the with
the poor diameter thirteen nanometers.

And you can see here are Permian says
that are given at better given for

these membranes and then a Permian
calculated without adjustable parameters

from the diameter found from
the small angle X. ray.

OK So
the first column is the measured Permian.

And the second column is the Permian
calculated for notes and

diffusion the close agreement
demonstrates that the mechanism is almost

certainly notes and diffusion.

The second thing that's
interesting is that the newts and

diffusion if it is correct should give
flux is proportional to the inverse

square root of the molecular way.

And the selectivity is
here are almost exactly

the inverse square root of the molecular
way the data at the bottom for

track etched membranes are interesting.

Only because the membrane it
has twice as big Poor's and

yet the flux is one hundred fifty
times smaller and the reason is

these poor are much higher Voigt fraction
there is just a hell of a lot more Poor's.

Now this is for the conventional pours for

the gyro structure you get very
similar kinds of pores in this case

the poorest do have a torch Wasi
remember these are not right.

Circular cylinders remember what that
would look like kind of like over cook

spaghetti with They've been chopped up.

OK or ramen noodles or
something like that.

That's always a graduate student
favorite dinner is not so.

So here for.

Diameter if you measure if you measure
flow of water through these things and

gas diffusion then you can calculate both
a size fourteen nanometers within and

then a meter of what we had before.

And the torture wasabi and

the torch you watch city is very close
to what Dan roster calculates for

random collections of spears So it really
looks like we know what these things are.

Moreover.

They show a sharper
molecular weight cut off

than the best existing
ultra filtration membranes.

At this point.

What we did was went to G.E.'s Monex and
said look can you help us

measure the properties of these new films
and they said sure we'll compare them with

the best membrane we have and
their best membrane is the Green Line and

our two membranes are the other two
colored lines and after they gave

us these results the manager involved
decided he wouldn't help us any more.

And he also would not tell us
which membrane he had given us

to that they had used as their standard.

So I can only infer given the number of
patent applications they've now put out.

I can only infer that these
are quite good performances and

by changing the length of the block.

Remember I was just talking before about
the percent of the system twenty percent

for example that was
are the blue Brocken and

eighty percent of the red block by
changing the length of those blocks but

keeping the percentage the same we can
adjust how sharp that that cut off is.

We have not pursued that and
I'm anxious to do so.

So what does that the other
thing we can think about and

these are not these are data from
Tom Russell's lab at the University of

Massachusetts is if we
get a much better flux.

What will happen is under the say.

Same pressure difference.

We're going to fell faster.

And so if we fell faster
we haven't gained anything.

Right.

I mean all we've done is just switched
from one problem to another but

in this case apparently because of
the very flat surface of of the poor

membranes we get considerably less fouling
than we would have another circumstances.

The data fit.

These are data for a membrane as I said
out of Tom Russell Scroope the data for

for our membranes appear
to be at least as good.

They are being taken by which renewal
Wickramasinghe in Colorado State and

I have not seen the details.

I only know that running
under typical process

conditions we're getting
no falling in a week.

Whether that is sustained or

not I don't know I mean what's
wrong with this picture.

I don't know I don't know but
at the moment they look pretty good.

OK so.

The barriers was a failure.

The net a porous will be
a big success if the fouling

turns out to be as minor
as it now appears.

But I don't know that it's
going to be that mine.

Now the next the third topic is
a membrane based on block O.

polymers that is selected for ammonia and
I've mentioned this to some of you today.

My my objective is to have an ammonia
plant that I can pull behind a tractor.

And I'm going to fuel this
ammonia plant with nitrogen

which I make from from
the membrane separation of air and

with hydrogen which I make
from a windmill on the farm.

So I want to take to start
to manufacture chemicals.

Not in a central facility but
once you don't have an oil tanker and

your energy source is
going to be dispersed.

I want to begin to manufacture
chemicals dispersal as well.

Now this really is a dream isn't it.

This is really not easy and yet
if you think about how you could do this.

You can imagine a number of
membranes already which have

these characteristics and one of them.

You're familiar with here.

It's the basically the self unaided
Taff one called Nathan on this

sort of mainstay of fuel cell shank you
and I talked about this then we saw.

OK so here is the structure.

And these structures these
structures have these

these membranes have the characteristic
that they are highly more permeable

to ammonia than to hydrogen and nitrogen
so when you have a hotter process.

What you do is you take hydrogen nitrogen
and push them together under pressure and

you make ammonia but

under reactor conditions you only
get twenty percent conversion.

So you would like to
take the ammonia out so

you could drive the process to one
hundred percent conversion and this this

stop has this spectacular selectivity
at least two hundred to one.

Interestingly the properties of ammonia.

This particular property of ammonia to
react with with salts like this was

suggested first by Edward
divers who was a apparently

very unpleasant English doctor
who fought with everyone and

so when he was sort of had fought with
everyone and they threw him out of

everywhere he went and
set up the study of chemistry in Japan.

Divers divers was the first chemist
employed by the Emperor in Japan and

in the University of Tokyo there is a bust
in the central core covered with trash and

old newspapers.

There's a bronze bust of this
rather stuffy looking Englishman

sitting there and
says that the base doctor divers and

this sort of technology
was very well understood.

Because if you remember refrigeration
originally depended heavily on ammonia and

so before the invention of freons this
chemistry of ammonia was extremely

well understood and
then it was all lost once freon appeared.

Now the trouble with this structure is
that this is fine for single gases.

But just as we were discussing for
carbon dioxide methane this particular

selectively completely disappears almost
completely disappears with mixed gases.

So just as you would have with carbon
dioxide methane separations which looks so

promising when you run a single gas.

Once you run gas mixtures
you lose that selectivity.

So the idea was again.

Could you go back you
can see what's coming.

If if you go back to a gyro structure and
now you make one of the structures.

Some sort of ionic group some
sort of a sulfa Nate and

the other structure just
a standard polymer.

Could you get a self assembled
structure which somehow

because of this local
architecture resisted swelling.

Could you get that kind of structure
that would keep it selectively and

the details of the structure
are shown on this slide.

Now I understand they're not there.

And that is because I was
physically restrained on my way to

the airport by two patent attorneys
from the University of Minnesota so

this was the slide they deleted.

OK And I apologize.

So I will tell you in words
what it was in there.

Damn those two people.

What was what was in there was
essentially styrene in a self

unaided styrene with coupling with nor
boreen and

the real difficulty is that you're
not sure which catalyst is used.

But let me assure you the.

There is one.

OK Now when you do this you get
this self assembled structure and

now the selectivity here is
greater than one hundred.

OK.

Nitrogen the red points are penetrating
rapidly hydrogen and night.

I'm sorry ammonia the red points
are penetrating rapidly and hydrogen and

nitrogen the green and

blue points penetrate about
two hundred times more slowly.

Now you have retained the selective
eighty with the mixed gases

that graph is both for mixed gases and for

pure gases that's both the actual
selectively in the ideal selective and

the points completely superimpose
Now there are other issues.

I mean this is potentially a major
breakthrough right because if

you could do.

C O two methane.

That's why I was asking you for

those articles that's clearly
the place to go if you could do.

C O two methane.

Then you would have a major way
to upgrade natural gas and so

that's why this is interesting but
C O two methane is more interesting.

Will I try and build my ammonia
plant behind a tractor Yes sure.

I think that's a that's a hoot.

I think somebody ought to build
one of those I think of the fun.

And so I will try to build one of
those but I recognize that the major

practical input of that may not be in
this system and then you can ask me

about other details of this there's major
parts of this that I don't understand.

So I just admit that going in.

OK Now let's back off and
I'll talk about the dreams.

OK I said I had three dreams.

The first one was a disaster.

That didn't work the systems
did self assemble.

I didn't even choose
a very good block Palmer.

But I already know that that's a failure
and the reason it's a failure.

Because the properties of
the composite are terrible.

Now.

Couldn't I make Lamell or.

Structures which had finite Lamell and

then could show that unlocking
of the fingers mechanism.

Maybe but I don't know how I don't
know how and so that one as far as I'm

concerned I've abandoned the second
one the monitor Spurstow pores.

When I made this slide a month ago.

I thought that the the fowling
of those Poor's

was was going to turn out to be
significant and so far it hasn't.

So so far I can be very optimistic but

there's no reason to work more
on this particular problem until

we have the following data in
the following data hold up the the.

The Permian bill of the of this
membrane is somewhat in excess of three

thousand square feet.

Three thousand gallons
per square foot per day.

It's big.

OK.

And then the third one.

Yeah OK it's ammonia selective.

But the operating conditions for
the ammonia reaction

are four hundred degrees and
one hundred eighty atmospheres.

And I'm running this separation at room
temperature at about five atmospheres.

So I got a long way to go before
I'm sure that this one will work.

What is my judgment my own judgement is
that it may find a niche application but

I don't think it's going to be compelling.

I think what you want is something
that actually will remove ammonia

reactor temperatures and there are ways
I think one can do that and we can

actually Martha you and I discuss those we
can discuss those maybe in the questions.

So what do I do next.

Well I think I keep dreaming.

OK.

I can think of lots of other things you
can do with this structure for example.

I could use this for preparative
chromatography because the inside of some

of these fours is filled with
carb oxalic acid groups or

acid chloride groups to which
I can attach specific Legan's.

And so

I potentially have a way to do preparative
come chromatography on a very large scale.

But do they have value.

Yet I don't know I think I'll
still stay with my original goal.

I like the idea of just dreaming people
will pick this up if it has value.

Maybe you will.

Thank you for listening.

Thanks.

That's

right.
The question basically is OK.

You gotta start but there's a long way
before things are practical and and

the answer is.

That's exactly true.

So let's just discuss what you have to do.

First you have to be able to
to put them into a membrane

geometry that is attractive the obvious
geometry to use like the one

used in the Minneapolis water
system is hollow fibers.

Could you could you equally
well do it with spiral wild.

Yeah I think you probably could but

I think if you can do it on
fibers it's probably better.

And we've done that.

No.
Do I anticipate problems in doing it.

Not that many.

I think that that's entirely
feasible I think that

it would be a tricky development project.

Then once I have those poorest are those

the are how long does
it take to etch them.

Remember I said I'm starting
with two polymers and

I've got to remove one of those
pollers before they become effective.

And and
the answer there is the etching data so

far suggests these things
are diffusion control.

So the data.

I showed you there actually take about
ten minutes to etch which is too

long to be practical.

But most of those data were for films that
were almost a millimeter in infecting us.

And if I go smaller remember it
the scale down is going to go

with the square of the size so.

So my feeling is I can get down under.

Let's say twenty seconds
without a lot of trouble.

Now if it should become a reaction
controlled then I would be in trouble but

there's no evidence so far that
these hydrolysis are anything more.

I think the really big issue
is the one of fouling and

I think the issue of being
able to start to think about

what sort of protocol you'd have for
backflush things like that.

I think that's an issue.

Could you use styrene is certainly
not the polymer of choice could you

use other polymers Yes.

There's a whole catalog of other polymers.

The thing that I find remarkable
is that the flow of data are set

member I did water flow and
and newts and diffusion in

order to get two parameters one of them
was the diameter of the tube and the other

was the torture wasa if I've got any
poorest that don't go all the way through.

Then I wouldn't get that result.

I wouldn't get a result so close and so

I think these structures
truly are remarkable.

OK.

What was the other thing
I want I remember.

The other thing that you should
ask is how much can I to

the diameter of the poor's And
the answer to that is so easy but

not as much as I'd like I can go
down to about four now animators.

OK but what I'd really like to
do is go up to about eighty.

And I probably can't get past
forty with the current chemistry

that that's what might
be the biggest problem.

I don't know let me just repeat so
that you heard

the question the question was could you
make it small enough to do desalination.

I think the answer by
conventional mechanism no.

But if you on the other hand went
to soften a groups on the inside

rather than carb oxalic acid groups then
I think you have a good chance to build

an oral membrane that would operate
by od on an exclusion rather than and

that would be a very but it's not clear at
all what the floor would be in that case.

So we've thought about that but we haven't
done any of that but that's a good idea.

You want to go to a small enough.

Small enough diameter so
so that your double layer

start to overlap within the pool and
then then you've got all kinds of tricks.

Having done a good idea.

I think.

Others.

Yes please go ahead.

I was struck by your
first three describing.

This problem skin.

So it has these multi structures.

That's correct.

But there's another thing that.

Cells.

Have additional Iowans that
chemical properties for

example might thought softer because
there's a reservoir for want to.

There you can.

I don't know the the issue on what he's
basically asking about is the thing

that we started with a nice saying could
I somehow change the the structure to

a greater extent he's drawing an analog
between between those structures and skin.

It's quite true that analog
is very good for skin and

Richard guy who is in pharmacy
you see F L F Now is it bad.

Actually took the series in data
that we developed there and

showed that they do apply for skin.

So so that type of argument that you're
saying does have application to scan

in terms of further modification in
order to keep the barrier properties but

regain the flexibility.

I think there's a reasonable
a quite a good prospect of that

if I build these systems
using clay flakes.

But I think it's a bad bet if I build
these films using block zero Palmers and.

I don't see colored pens but but
basically imagine I have red and

blue the problem is when they form a mela
the the molecules are going up and

down kind of like that terrible
an edible Christmas candy.

I know it's people already written on the
screens too I compliment them on that we

had a one professor who is famous for this
getting lost in the middle of Ph D. exams

and writing on the screen the projection
screen when he got excited about what.

OK.

But but basically that that is
a possibility if you go back to flakes.

But I'm going to work with BLOCK Palmers
because the molecule goes back and

forth it.

The molecules don't stretch out
like this they go like this.

They fold.

OK And if that happens or
they go up through one the metal and

then you're in one region
they go out the other side.

OK if you have that kind of structure.

The only way you're going to get
the right mechanical properties of the.

I'm.

OK Yes please.

The question is is fouling here going to
be worse than it would be conventionally

and the answer is no if I run at the same
low fluxes that are used conventionally.

But the whole reason to to run these is
to try and get bigger fluxes Now I could

argue that I what would be I'd settle for
the a smaller flocks and have

a lower pumping cost because I wouldn't
have to run as high pressure difference.

That's true but
I don't think that's a real knockout

advantage to these things I
think if you're looking for

a knockout advantage you want
more more water last membrane.

OK.

I don't know the question is Why don't
they fell don't know the the cliche.

That's used in the ultra filtration
literature is that fouling is

closely associated with the flatness
of the membrane surface.

And and in the gyro ones we made we don't
have an especially flat surface but

in those cylinders that surface
is flat within animators.

And it may be that simply we can we don't
get as many as much stuff caught in

you know just just imagine all
your teeth were polished and

you wouldn't have as much
stuff stuck in them persona.

If I'm if if I'm going to open
pores the maximum ports sizes

is somewhere around fifteen animators

which is smaller then than I would like
to be I'd like to have a little bigger.

No what happens is then the chains.

I mean the way to make bigger pores
you think is make bigger blocks right.

I mean you at the moment have bigger
butt but these things don't line up.

I mean the pictures in this way are
misleading the these things are like sort

of over cooked pasta.

With with different kinds of
pasta hitched up and then.

OK So what happens when you try and
go to bigger pores is it tends to just

form the same sorts of scale of structures
but with more bending of the molecules.

But that's a nice thing to think about
could you for example use use polymers in

which the blocks had stronger
bonds within the blocks so

that they didn't bend as much could you
then get the bigger poorest don't know.

Nice idea.

OK.

Yes.

Yeah.

The key to this is the reactor becomes
very small because the productivity

now is enormous and the capital cost it
drops by on the order of eighty percent.

No recycles and you're going to waste
some hydrogen sure but vent it.

You know I like you was trained to regard.

I was trained to regard hydrogen is as
close to walking around with dynamite and

things in my pockets and I was very
interested you member those crazy

civil engineers that didn't know
the difference in gas constant

We got civil engineers who have lots
of permission to put hydrogen in dirt.

You know about this.

They're doing it all over the country.

If.
You put hydrogen directly into dirt

the microorganisms that are there will
consume chlorinated car hydrocarbons

that are it and I'm absolutely astonished
these civil engineering graduate students

are taking cylinders of hydrogen with
a bit of copper tube in the end I'm not

making this up jamming it into
the dirt and opening the valve.

You know the farmer sitting there
smoking to talk and I mean.

OK so sure now.

Now should should the farmers do that.

I think we can I think we can do that.

No no I know I know my own
suspicion is that well.

I wouldn't put a membrane in that system
I can think of other ways to do it.

We can talk about those Africa's
they're off the subject of this talk.

Yes please.

OK he's talking about remember that the
big deal on the ammonia separator was to

run after it has an ideal selective
these are very good in other words

hydrogen doesn't go through in a mode it
does go through but then when you put both

gases There they go through at about the
same rate but but with a block O. polymer.

They don't go through about the same
rate in either case the ammonia

always goes through faster.

It's question is why is this and
the answer is I don't know.

I'm sorry but I don't I mean
you want an explanation that

it's because it's because
the structure of of the gyro Deezer

gyrates it's because the structure of the
gyro it is sufficiently rigid to prevent.

Prevent plastic station
by by the ammonia and so

the free volume doesn't exist and there's.

On a nano scale no swelling and
that's total bullshit I have no I have

no It may sound good but but I have no
reason to know that that's what happens.

The question is could could we tell
this by looking at small angles.

You know that's a that's a good idea.

We haven't done that what we did do
was change the size of the blocks.

And what I would have guessed would happen
is if we if we had the same percentage but

smaller blocks then we would have
smaller structures in the membrane.

And the effect should be still
better with smaller structures.

OK we did that experiment and the
structures with the smaller blocks were

slightly worse so that was the end of my
guessing how these things would turn out.

OK.

Yeah yeah yeah.

And this is what.

Yes You know there's enormous interest
in try blocks because the kind of.

Doctors that you can get now.

Are truly very elaborate and and the sort
of thing that you immediately begin to

think of is is could you build
a membrane reactor where you used two

of the other blocks to feed into regions
and you carried it and you can just.

And and where there are what that I
show roughly half a dozen structures for

the guy blocks.

There are more than thirty
already identified for

try blocks whether those have value or

not I mean quite frankly there's very
little value for DI box at the moment.

Yeah I mean what you're looking for
is is a tool box isn't it.

I mean what you want is is a whole bunch
of rules about how you can assemble this.

And that's what's so
fascinating about them is that they

are the thermodynamically stable faces
that's what's really interesting.

I mean the way you get them sometimes
you make them and they'll form.

You know you just smash them out in
a film and then you heat him up and

sure enough they they
rearranged to go into those.

Those faces.

Could that cause trouble in in
making them practical I don't know

we haven't had that when Neil and
in those poor systems.

Nobody asked the question that I thought
you'd ask me because I was sort of hoping

since I don't know
the answer one of you would.

So can I ask you a question.

Would that be.

Just OK I think I and
this is the end after

my question when Ok I have to
figure out how to get out of here.

I want to go back to one of the slides.

So here we are.

This is the one I want.

OK.

How do I do this.

Here it is.

Look at these poor.

Now what's fascinating to me
the first thing you look you think

wow that's a lot of poor.

And the second thing you think.

Yeah they do look about the same size but
I want you to look now at a different

sort of thing you see how
they kind of form groups.

OK What why is that and and I'll tell you
the explanation that people give because

I would like you to either tell
me if this is good or bad.

They say aha that's because you
have grain boundaries between

I forget what they say many
two dimensional crystals.

What do they mean.

I mean what causes this pattern.

Go ahead.

Yeah could be could be yet.

It looks it looks exactly
the same before and after action.

Qualitatively So what happens is you

get nuclei that forms on grow in these
things and then what happens is they grow.

Yeah that's probably right.

But but it's groups of them.

Charles that's what's interesting.

It's not just one.

I mean I can imagine if you had one and
then you sort of had for example like

really instability you get those nice
hexagons like I can understand but

this seems seems both less
ordered than that and

more ordered than random and
I don't understand why.

Yeah go ahead.

Are you.

Yeah see that's what Charles Assange.

Yeah I know that that's I think
that's probably the right idea but

I don't know how could I prove that.

Slow evaporation the cylinders
will whip down into the.

That's to me that that may be true but.

But that's spoken like a professor
who's not going to do the experiment.

No no no you just you just
take take a regular film and

or regular play and and
you can ask that out but

you may do it in a room with the
temperature turned up and you do it with

a volatile solid and you may do in
a laminar flow hood and things like that.

I don't know.

Can one of you give me a reference.

Good thank you.

Look.

Can I change the solvent systems.

Yeah.
If I change the solvent system I

get the same thing again.

But it looks a little different.

That's all.

But remember I'm always trying to use
fast evaporation if I don't use fast

evaporation.

Then the cylinders form then then I
look at this to be absolutely plain.

Because all of cylinders would be buried
down within the plane of the fill.

You must have fast evaporation in
order to nucleate on the surface and

get them to grow into the surface and.

I don't really know how to
make that quantitative either.

I mean you were saying that there were
several questions that have reflected on.

Yes can you change the temperature you can
change the temperature of the plate for

example you could change campus or
the casting fell.

Yeah that's a good idea.

Yeah that's good.

Man this is this is good you want to
go through all the slides like this

is just far.

Thank

you.

A little thank all

of.