Here.

There yes the.

University.

Right after.

First.

Person.

There's no one.

Here and this is for

commercializing.

You know did more detail about.

That.

Like you.

Like.

First.

Question.

Thanks once again thank you it's
really great to be here I appreciate

the invitation you'll have a wonderful
school Department of Chemical Engineering.

Turn the microphone off thanks I
got going it's a quickly there but

yeah thanks again for
having me today it's really a pleasure and

honor to be here as Mark alluded to
the title My talk is going to focus on

what I call disassembling and
it wasn't perhaps a.

Organized disassembling we did it was
perhaps somewhat unexpected this is

something.

And how we can learn from that and
how we're able to maybe now reconstruct

what we found there and
find new materials for C O two capture so

to give a brief overview of where all this
year it's you capture stuff originated

is it really comes from
natural gas sweetening So

that's the removal of C O
two hydrogen sulfide other.

Things from methane thing and the natural
gas that comes out of the ground and

that's historically has historically where
the capture of C O two first started and

now as we'll see it's very different than
the capture of suits you were proposing to

do from power plants in the fact that
the conditions are so much different

in natural gas wells we see a wide
range of compositions and pressures and

in addition of C O two hydrogen sulfide we
find all sorts of species such as water

CARBONELL sulfide carbon carbon
die sulfide hydrocarbons be taxed

benzene talking in at the benzene xylene
all sorts of stuff can be present and in

these natural gas weaning right we've got
to remove the C O two the age to us and

all the other bad guys because we want to
get clean fuel that we can burn Also H two

S. is really toxic C O two and H two S.
are corrosive there's a lot of

reasons economically and industrial to be
doing this if we want to use natural gas

as a fuel and the common way to remove
these things acid gases is through

a Chris I mean solvents which
are efficient but how energy intensive and

that's going to play a role here and
what we're going to talk about next is

optimizing or minimizing the energy
intensive news of such a process and

so to point out this natural gas remain
committed current really large processing

facilities now this isn't just sweetening
but there's certainly some natural

gas weaning units involved here or it
could be really small portable units that

I drag out to a wellhead So
this would fit on the back of the truck.

Now from power plants this is what we hear
about today when we hear about the term

carbon capture Yes carbon capture Seaquest
ration and even carbon utilization So

this is really going to occur at coal
fired power plants predominantly And

here are the conditions are going to be
much more consistent than they were at

natural gas wells but they're also going
to be very different here we've got low

pressure gas almost always we've got
a one atmosphere pressure of the total

stream which is twelve to fifteen percent
C O two now the things we're going to find

in that stream are also fairly consistent
from power plants a power plant.

Nitrogen oxygen of course you have
to burn coal so we've got to get our

oxygen from somewhere of course and
make water there Sol first so

we could also to get things like mercury
ASH a whole bunch of nasty other things

that go along with it now most of these
sort of to Mercury actually is already

being scrubbed right so really then
we're dealing with Niger Noxon H two O.

some SOL for with that C O two now the
magnitude of this is much more significant

as we've got billions and billions of tons
of C O two that we're meeting annually so

when you hear about greenhouse gases and
global warming and

these kind of things well what this is one
of the major sources of those emissions

power plants now the part of Energy said
in order for this to be economical they

want to limit the impact of such a process
on C O E which is the cost of electricity

the thirty five percent which more or
less means that for every dollar I pay on

the electric bill today I can't pay more
than a dollar thirty five when I do this

process at least that's the goal and
our best processes those that have been

commercialized in the past for
a natural gas I mean technologies really

can only get in the fifty to eighty
percent range so that's a big impact on

the cost of electricity because it's not
just the cost of our electric bills that

propagates it's a cost of everything we
make and use in society right electricity

power drives the economy to an extent or
to a large extent so if we increase

the cost year then we're increasing
the cost of just about everything right so

here's a typical block flow diagram of
a power plant we're going to put that

suit you capture process all the way down
the line it's going to be one of a number

of units that are downstream of the boiler
which is where we're actually making

the steam and making the electricity right
so we've got to remove particulates sulfur

before we even remove the C O two and
the ultimate goal is to separate

it from the flue gas compress it and
store underground use it for

some beneficial purpose hopefully So
at this point I always like

to show this slide because this shows more
or less the process conditions that exist

in industry where we might remove
C O two from some sort of gas and.

On that Y. axis we've got partial
pressure of C O two in the feed right so

anything from one P S side to a thousand
P.S.-I is represented on this graph and

we've got partial pressure of
C O two in the product right so

basically this is a ratio then of what
coming what's going into what's coming out

right and so that roughly tells me
how much C O two I'm going to remove

in my process where it's at
high pressures we can use

what are known as physical solvents which
don't rely on chemical reaction and so

as we come down here as we get to lower
C O two partial pressures in the feed and

lower partial pressure of the C O two
in the product we get into more and

more reactive strongly binding solvent and
this is just for

solvent but there's membranes there's
other technologies that will talk

about a little bit later that would
have their own places on this chart but

this is strictly for solvent based process
and so really then as we come down

as I said natural gas would be all
up here even some down here but

see what you captured from bluegrass
rest firmly in this aqueous I mean

section of the graph and this is from Oil
Gas Journal article from one thousand nine

hundred seven but I think it still makes
the point when we're operating at really

low partial pressures coming in and we
have to remove ninety percent of C O two

they're operating at very very low partial
pressure to you it's you coming out so

we're very firmly in this regime where
we need a reactive solvent to accomplish

ninety percent remove We almost
have no other choice according to

that crowd now as chemical
engineering majors at one point or

another we've all seen this particular
flow chart right absorber stripper

absorber regenerator the C O two comes in
contact with the solvent we get the clean

gas of the product gas coming out of
the top we send this year to rich solvent

through heat exchanger heated up release
the C O two to events or a compressor

they don't mean solve and then gets
recycled and repeat the process over and

over again as Gary Rochelle points out and
his science article from a few years ago

this has been in operation for

about eighty years and in a lot of
cases this is the best we can do

it's actually a very efficient process if
you look at some of Dr shells article.

I mean this is it's been used for eighty
years for a reason because it gets the job

done but it's maybe not going to give us
the energy efficiency we need unless we

start to tweak some of the process or
maybe tweak the solve it so just for

some quick background on house you have
to capture works between the means

we've got the car bammy forming right so
we take to it means those are the primary

means secondary means they direct directly
with C O two so what happens then is I get

to mean molecules I pick up a molecule
C O two and I make this salt and

so then primary and secondary is behave
almost the same way tertiary means those

are hindered it means they behave a little
bit differently right so they can't react

to radically would suit you because they
don't have that Proton on the nitrogen so

it's got to react with water first to make
carbonic acid and the carbonic acid is

the neutralized of the tertiary I mean
to make a bicarbonate ammonium salt

Now each of these as a preferential
pressure where one might use them right so

here we see that there's a stronger
binding energy right anyway has a binding

energy about a twenty five B. to use
per pound of C O two that's going to be

most effective at low pressures right this
very eczema thermic favorable reaction so

if I don't have a lot of C O
two floating around well

then I need something that's going to grab
onto it pretty strongly D.E.A.'s somewhere

in the middle we've got seven thirty
five beats used per pound of C O two so

that's a moderate pressure
we might use D.N.A.

and then we can use that at high pressures
because it's going to drive this reaction

and so we're going to get a more
favorable boost in the reaction rate from

the high pressure so we can use a more
weakly binding I mean and get away with

the tertiary I mean which is going to
help us with our energy efficiency but

as I mentioned zero as you capture
that involves low partial pressures so

we kind of have to get something
that's going to bind quite strongly so

we're going to look at things
like EMEA a primary and

secondary means in order to accomplish
the C O two capture goal now some

general acquirements on what we want to
accomplish then in designing solvents or

or analyzing the impact of
the solvent on a process then of.

I want to minimize the impact
of cost of electricity that is

that thirty five percent target we'd
like to reduce the process capital and

operating expenses right if we can drive
down the cost of actually building and

running this process that's going to be a
great thing that's also going to play into

the cost of electricity for the sake of
scaling we'd like to keep these things as

small as possible so if we can have a
target of one hundred grams of C O two per

liter of solvent that's a rough
estimate of where we'd like to be and

of course we'd like to have things like
a low viscosity to get high mass and

he transfer rates where we simultaneously
want to minimize the amount of energy it

takes to regenerate that's all and
that's to kick the suits you back out and

then we'd also like it to be stable in
the presence of some really tough actors

oxygen US or two would set or we want
these things to last a long time and

not to grade so this is asking a lot but
this is what it's going to take really

to get to that thirty five percent
efficiency now I'm going to

take a step back away from a career
Solomons and give you a bit of history

where adequate quids came into play a lot
of us have heard about on a good grades I

know there's a lot of faculty members
here who are working on a good grades but

maybe some of the students haven't
been as familiar with these yet so

on a good grades are more insults
effectively they're usually organic

materials that are a liquid salt at room
temperature they're not sodium chloride

in water it's a pure organic materials
we'll see on the next slide or a mixture

of organics but they're charged and
they've drawn a lot of interest because

they have some interesting properties
that make them really useful for

a chemical engineering process sees
they're nonvolatile we like to think so

at least even up to three degrees Celsius
Now if they don't break down before three

hundred cells is effectively because
they're salts they can evaporate so

we have this negligible or
almost zero volatility

some of them are quite thermally
stable but that's going to depend on

a lot of factors including the composition
of the On a good itself and the and

in the environment that it is in again
they're also chemically stable but that's

going to be also dependent on factors in
your process and they have a load of my.

Viscosity and I'd say that with the caviar
because we see in a few slides that there

is an apparent lower limit on viscosity
but we don't see an upper limit

we can actually make them as viscous
as we want and that might be useful for

some applications but maybe not so much
for suits you capture now the big question

then was Could these be used that low
pressure what is the C O two capacity in

good at low pressure so this was a problem
that people have been studying for

a long time and continue to study can
you make these to have a higher C O two

capacity and then if so could they be used
for suits you capture from power plants as

an alternative solvent to those aqueous
the means that we're talking about

right because if they don't
evaporate we get some benefit

to our process in fact what I didn't say
was that in the process a lot of energy

can be wasted purely by evaporating the
water so if we have a nonvolatile solvent

the theory is then that we can save
some energy so a brief overview then of

what goes into making a lot of good some
chemistry here that we'll get into some

more and more chemistry as we go along
as I mentioned there molten salts really

then we've got a cat ion and an ion that
are big and bulky and we pair those up

they can't crystallize so we get
a nice liquid at room temperature and

that depends on a lot of factors not every
combination here every combination one

could think of as necessary going to
be a liquid at room temperature but

some of them are and
then basically the selection and

design of these ions is going to influence
the physical properties right density

viscosity hydrophobic city conductivity
it cetera et cetera et cetera

C O two site ability all of
these things are then functions

of these functional groups right so
we have a lot of ways to play with tuning

the physical properties of these systems
I highlighted the images only in have

to when paired because that's one that my
group has used a lot both in the past and

currently but
this doesn't necessarily mean that's

all that there is there's a whole
bunch of potential combinations and

even things I'm not showing on this slide.

So normally you know in the past what I
did was we looked at a number of different

a good speed.

And so really then what we're
most interested with was building

functionalities on to the cattle and the
cattle is relatively easy to modify and

ions are not so easy to modify necessarily
so could we build in functionalities

here these are groups that mimic things
that are already used in gas processing or

gas treating applications so the original
article which had this hydrocarbon side

chain but if you look into the literature
we normally see that hydrocarbons aren't

really the best way to separate
C O two out of our membranes and

materials and other solvents there
are very polar hydrocarbons and

non-polar group so we could be said
well what if we put in either group or

a number of groups on that side gene can
we improve upon the properties of these

hydrocarbon based salts and then things
like night trials for native groups so

Locsin some of that was purely out
of academic interest maybe but

nonetheless we can modify these
with a number of different polar or

different functional groups and figure
out how suits he was influenced by those

presence of those groups so for
comparison quickly then we're going to put

up a chart of organic solvents and
these numbers keep some of them in mind

as you look out on it with us we've got
to see tonight we've got a number of

solvents organic solvents and
water we've got the side body parameter.

Of C O two in terms of standard cubic
centimeters per volume of solvent perhaps

fear and then we've got the Henry's
constant Right which is the mole fraction

of C O two at a given pressure so
here then we see a c tonight Reza Sayah

ability of C O two at seven point
one C.C.'s Percy C perhaps fear and

that's the highest on this list we
have and I borrowed this list from

Benny Freeman's group at the University of
Texas when I published this paper and so

these are their numbers and we've got
C O two at seven point one But then

we see Henry's constant of sixty point
four So keep those numbers in mind and

then also we see Henry's constants that
span maybe from like forty five seventy

eight one hundred fifty water being
an outlier there are six hundred four.

And then in C O two receive anything from
seven point one C.C.'s to about two point

one and hexane of course water then is
also an outlier see it's you sideways

quite low in pure water so we made these
similar measurements on a number of

Ana goods and I want to go through
everything that's here but

you can see some groups where
we had hydrocarbon based.

Trials for a benzo will it cetera and
looking back at the numbers we're going to

start of the hundreds constants first
because this is what got people

excited this is the on it could have
a very low Henry's constant right so

the mole fraction of C O two in the out
of good is actually bigger than it is in

the organic solve Well I don't agree with
their big molecule So when I have a big

molecular weight right it's effectively
less moles in my solution so a little

bit of C O two looks like a bigger mole
fraction now when we revert that converted

back then to the volume of a saw ability
standard cubic centimeters per cubic

centimeter promise for we see numbers in
the range of like two point four two point

one one point nine We
don't see a huge increase

in Syria to cite ability compared to
organic common organic solvents right so

we wouldn't consider using a typical
organic solvent in a post combustion so

you had to capture a process that just
doesn't have the capacity for C O two so

the Henrys consonantal
could can be misleading so

we look at the volume based Cybill
there's actually not that much

C O two dissolving in these systems.

And when I plotted that and
we've got organic molecules on top

These are the empty diamonds we parted
some polymer data on the bottom

the hollow squares again this came
from Benny Freeman's group but U.T.

Austin when we superimposed on it
Good words over that we saw almost no

dependents right we saw a very
flat trend there's not much up or

down that we could tweak the C O two side
already in the Solomons and then yeah by

cooling it we got more right so blue
is cooler and green is forty degrees C.

So by cooling it you can get more but
that's to be expected and cooling you.

Really is there going to be an option
in many of these processes so

we weren't absorbing a whole lot of
C O two at forty degrees or even

twenty five degrees Celsius so we were
somewhere in between small molecules and

polymers which you might not and might
not be unexpected because Rwanda grids

are much bigger than these small
molecules in terms of molecular weights

they're not quite Palmers but
they're somewhere in between and

I think that result makes makes a bit of
sense when they get it in terms of volume.

Now I borrowed this from do you we think
it illustrates the point as to what we're

looking for in C O two capture for power
plants were needing a chemical solvent

write something that has a chemical
reaction with C O two because again

we're going to be operating at low partial
pressures and so the X. axis being partial

pressure the Y. axis being absorption
capacity here we're going to be operating

and this is all relative right we're going
to be operating at relatively low partial

pressure and so a chemical solvent
going to give us very high loading and

then after a while it's not going to
load any more as pressure increases

whereas a physical solvent has a linear
response roughly in pressure right so

a physical solvent going to better see
the high pressures because we can get high

loadings but a chemical solvent is going
to work better at low pressures So

again we recognise that on a good
your physical solvents Henries a lot

type behavior that linear loading
of C O two or other gases so we

needed to increase the C O two Cybill in
all go a little bit more quickly through

this part so looking back at that original
chart we said well maybe on a good can

work well up here right so maybe at high
pressures this is an application but

that wasn't really we're focusing on
we needed something that could work

down here at the See if you
capture from flu gas region.

So me and
Dr camp are here having coffee one day

we said well we have some amines why don't
we just add the means to go on a quick

right we know what means get high C O two
capacities on a good goods have some nice

properties maybe they don't
absorb a lot of C O two but

if we put some amines in there well we
should get something that's the best.

Both Worlds and what do you know it
shouldn't have been unexpected but

it was an exciting result anyway we just
looked at the guy on a good alone if we

had forty degree Celsius and two P.S.-I of
partial pressure we saw that we got about

two point six grams of C O two per liter
of solve it now when we added sixteen

volume percent and the sixteen percent was
probably based on a mass to mass ratio but

we converted this to volume so
sixteen wasn't arbitrary.

What we saw then was we got seventy grams
of C O two per liter of solvent under

the same conditions so that's one hundred
ten times increase in C O two and

I said well that's a pretty
great result we can combine

the nice properties of on a good goods
with the reactivity of a means and

we can have some super solvents right
we're combining the best of both worlds.

So we saw then that this
worked quite well and based on

choosing different means any A had
a higher loading than D.N.A. at a given

pressure this would be expected what we
saw was that the kinetics were fast but

it was all based on mass transfer it was
going to react as fast as we could stir it

and here viscosity became an important
consideration we also notice that in some

cases we can actually pursue petite the
amine carbonate so that reaction product

that forms typically doesn't precipitate
from an acreage solution it's a well.

Solved but not a good grids especially
ones with this particular and

iron it precipitated which is
a really interesting result but

it's an unconventional process so we don't
really pursue this further I know others

are looking at precipitation
a means maybe not.

So this is an interesting result and

people have pursued this avenue
of research but we haven't but

it does happen and that's what's shown in
that picture there now quickly then we

said we started mapping out what
the benefits of such a solution could be

well with there with the precipitation not
included right we can use these solvents

in a conventional process going back to
that process for a diagram that I showed

at the beginning there wouldn't
be much need to change that.

Process using this idea
of a combined hybrid and

it could mean solving AI we
did some calculations and

using some process simulations even we saw
that by removing the bulk of the water and

I underline bold because you can
never remove all the water going to

absorb some water from fluid gas
there's an energy savings and

that's goes back to what I said about not
evaporating water in the generator and

some of the modeling maybe the most
optimistic models that we could get down

to two point three Giga Joules
per metric ton of C O two

whereas conventional aqueous solvents
are in the range of three to four

good jewels per metric ton of C O two so
at a significant savings and maybe a more

realistic estimate conservative estimate
is something like two point seven But

that's still an improvement over
conventional processes now we said if

we're going to scale this down a good
you're using in the lab are going to work.

Primarily because they're too expensive
now we were set up to find a very simple

and quick to do otherwise the economics
of the cost of the solvent will make

the whole process an economical so we said
we want something that's thermally and

chemically stable relatively low
viscosity tolerance of water no cost and

commercially available so more or less a
super wishlist of the most ideal molecules

and basically then as I said we want to
consider the process right you always have

to consider the process necessarily more
than the molecule it's easier to tweak

the molecule to meet the process than
tweak the process to meet the molecule So

again we want to make sure we didn't have.

Super viscous solvent to get so
that we could get good gas liquid contact

we were concerned about the heat
capacity because it's going to change

the temperature profile in the absorber in
fact it's going to get hotter potentially

because the heat capacity of the bulk
solvent is lower than water then we

started to think about OK so what about
the manufacture in the maintenance here

we have to do new approaches to solvent
reclaiming because we can to still on

a good words they have no vapor pressure
so if they get dirty What are we

going to do with them we have to filter
them or or do something unique into.

Of getting our good solvent
back we then asked what

how long are these going to last and
if I want to manufacture these in Bowl and

gold is some serious ball in this paper
I did an analysis on what it would take

to fill a series you capture process
on one full size power plant and

it's something like a thousand
tons on day one and

then you have to continuously replace
some of your solvent as you go along so

if I'm going to do this every
power plant in the world

well this is this is a major consideration
how would you ever manufacture these.

Big scales so in two thousand and ten we
got a project funded by Do we need to yell

a number of companies University Alabama
included we had some engineering

firm chemical company every Electric Power
Research and just do we have a power plant

involved this is Excel Energy in
Boulder Colorado where I come from you for

a move to Alabama so we initiated this
project in two thousand and ten to do you

bring these to do a pilot in the field
Well the best plans that we have often

fall apart and we have to ask ourselves
what can we learn from this so relatively

quickly Ivana was able to provide us this
relatively low cost non foreign aid and

again I point out non foreign aid because
there's some toxicity concerns there when

these things are foreign aid like I showed
earlier so we have this one C two min

which means at the methyl and it is
only I'm E.T.F. So for Ethyl sulfate So

this is this is our lowest cost this
is your best hope of meeting these cost

estimates that you need for
your process and they were kind enough to

give us drums of this stuff as part of
the project well during the course of

testing we observe them that the A means
we're putting into this particular on

a good we're actually attacking this at
all sulfate which was believed in a lot

of cases to be relatively stable
that this group wouldn't pop off but

in fact it did and what it's we saw
then was enemy a secondary I mean

would come in the ball off and
we'd make a tertiary and.

In which then is not going to
be particularly reactive to

C O two hundred low pressures and
we make this so if you're a cassiterite So

the Hemi sulfuric acid or by sulphate and
I which of course then is a pretty

strong acid which can throw an eight
that caused corrosion etc etc etc So

really then this was a losing proposition
and the things that we said in the lab

that we're going to hold up aren't really
going to work in the field and do we said

well you know this is actually a useful
result please publish this we don't want

others going down that path and so earlier
this year we wrote up this result and

we said you know a lot of good
goods with the means maybe.

This is going to work out the way we
thought it was but in the process of

analyzing the degradation products
we saw something else come out.

And looking at it we saw that the cattle
in that image is only human right

the positive charge would revert back to
the one mouthful and one Ethel in middle

and that's not pretentious about the same
mechanism we saw earlier we're still not

sure what the mechanism is but this guy or
these guys were showing up and

we said well if it's a degradation product
of a not a good it must be a more stable

molecule on a good reason to go back to
this state then maybe we should be looking

at images and if the middle schools were
more stable can we use them as well for C.

or to capture applications and as we'll
get to in a minute I think yes and

that's really where that
project has gone and

it's more about now scaling images olds
for carbon capture than ever is and

a quick wits so a little bit of background
here middles olds are really interesting

molecules it's a five member heteros cycle
as you can see here tonight you're going

spaced out by one carbon on the bottom
two on top and when we get our two

sides of an equal reactivity right so
this nitrogen is effectively inert

it's not going to do much this is where
there's a lone payer it can be a proton or

supped your community file I mean as
older fellow than any natural products

they're also found as cores of many
pharmaceuticals so they're been around for

a while but as we'll see the baby.

Characterizations of the middle
schools haven't even been

carried out until just recently as we
know that these are building blocks for

the midazolam based Isles and we can
produce in number of different molecules

by varying this our group just like on
a good we have this inherent to an ability

of the structure based on functional is
ation here and as we'll see later we can

even introduce functionalities that any
of these other carbon ring positions so

it's actually potentially even more
versatile than adequate quits in some

respects so again as I've been
saying these are connected right and

we didn't even pay much attention
during the early days of Onyx liquid so

to say we're on a good base materials
because we're making emitters olds in

order to make new article good materials
and I'll show a few examples here that

include a good membrane so polarized on it
would you have to start with a middles ols

we made a liquid crystals out of the use
also starting with the middle as old

as with your poly Cateye and
that cetera et cetera et cetera and

they all began by synthesizing
in the middle but

we weren't focusing on that itself we're
only using it as a building block and

not as a useful material.

And I think that we it was an oversight
and you can in retrospect you can say

why don't I think of this five six years
ago but that's just the way science is

sometimes So again here's an example where
used in the middle of all to make liquid

crystals or really long chain molecules so
we get a hydrophilic

Quranic head that separates from
hydrocarbon tails we get some useful or

interesting nano scale behaviors in these
things we can make cat ions right so

we can string these neutral molecules
together the same way we build on a good

goods so we can make these big cat ions
which have some interesting properties and

gas separation membranes as
we first looked at them they

have actually very very low
probability used to C O two and

hydrogen which was somewhat unexpected
based on an include good chemistry.

So really then we said if we want to use
them it is always we actually can't buy

many of them so.

Then went out and try to find ways
to make these things that are very

efficient use commodity chemicals and
can give us high yields right so here then

the goal was then to make these things in
the lab on meaningful scales engineering

scales at least that we could test with it
with a good amount of solvent so we can do

is taking a middle course some metal
hydroxide sodium hydroxide is a good one

we can do it underneath conditions that
solvent for your we can add some C.H.F.

is a good solvent here and
we're going to do protein at the middle so

now we're seeing the other phase
right we saw the positive side

good the neutral molecule and
when it actually can undergo and transform

into a negative negatively charge salt
as well so when we take some molecule

whatever that might be Ethyl bromide
perhaps Our want to add that in C.H.F.

this group attaches here and we get our
middle with our functional group and

it's relatively easy work up and
as as we showed we can do this for

any number of species
here in this paper we did

twenty different species all sorts
of different functional groups and

this is just pure the synthesis we want to
show that this method is actually broadly

applicable and so as you can see here on
the right just about every Ewald was about

eighty percent or greater and we give them
pretty big scales anything from twenty

grams all the way up to about one hundred
grams so this synthesis is going to be

very efficient and that's going to be
useful to help us moving forward because

we're not necessarily making things
that are very difficult to deal with or

we can't make enough of them to do
meaningful tests now at this point we went

into the literature we said OK we made all
these compounds were not necessary doing

making molecules for the sake of making
them let's start characterizing them and

we found pretty quickly that there
was virtually almost no physical or

chemical property data from it as olds in
the literature I mean these are molecules

that are even simpler than a nickel grids
even though it's have been around for

nearly a decade and
they've been thoroughly characterized

no one has really looked
at the precursors and

so we started as well as this guy.

Gave a read can in Germany he's done a lot
of measurements on vapor pressures we've

been working with him since on doing some
further characterizations and we knew we

could simulate some of these properties
as well so even if we couldn't yet

make them all kewl you get a really good
estimate as to how they might behave

using this software package and what we
found pretty quickly though is that most

of these are low volatility liquids
not quite zero vapor pressure but

very very low less than one millimeter
of mercury in temperature and

they have boiling points
above two hundred cs so

in fact that's well above where a carbon
capture process is going to operate so

pretty quickly then we examine the
fundamental properties then city viscosity

We also looked at the relationships to
the change that it is all as well as

temperature and we want to compare these
tronic we said OK so they have benefits if

I've changed the structure from a charge
structure to a neutral structure what

happens to the thermodynamics and
the interactions with C O two.

So here is the profile of these
things I think the only thing that's

really important here to take home is that
most of these emitters ols exist as very

very low viscosity liquids we're seeing
about one to two percent of poise at most

temperature conditions except when
the chainline gets very long twelve C.

fourteen we see a big spike but
that's to be expected and

to put it in perspective right a typical
water for example is one set of poise

a typical water based solvent for
c it's you captures something like

a few five cent a poise perhaps So we're
in the right ballpark to discuss what is

low enough where this can be a useful
solvent potentially Now we compared those

viscosities Tanika grids we saw something
encouraging that we were one or two orders

of magnitude less viscous than on a good
so all those things we're concerned about

heat transfer mass transfer rates well
there's a big advantage there because

we're not a charge molecule anymore right
we only viscosity is only a function

to change the not the and ion as well so
we saw one to two order of magnitude

reduction in viscosity or so I think.

This is an interesting plot it tells us a
lot and these aren't arbitrary underneath

this are the actual data points I just
did the over way to make it look nice but

if you go into the paper there is real
data underneath both of these windows that

supports the fact that I could goods tend
to be one to two times more viscous than

their corresponding missiles now we
didn't see a big advantage in C O

two side which is interesting and
we won't go into why that is today but

again the same thing we were doing right
we're starting with a small molecule and

adding functional groups so again that
really wasn't the way to increase you know

to save you if nothing it was slightly
decreasing as we extended that al-Q.

chain but nonetheless it prove the point
that it is a little bit more sizable

C O two than in emitters old inadequate
words but it's all going to do is extend

that chain length on the molecule but it's
always going to hurt me I'm never going

to move up in terms of C O two as
I change that solubility program

only going to move down and
to the left and

there are some reasons for that we just
won't talk about given the time so

then the idea was OK so if it's not going
to absorb water on its own Can we also

drop in I mean into emitters all solution
and get some chemistry to happen right so

again just to review with and
not a career solution or

make restitution we've taken to mean
with C O two right we're going to get

this water on forming so that's our
intermediate and then again we're going to

get that one molecule C O two for every
two means making the carbon the ammonium

sold when we added a means to emitters
ols we saw something different happen

we saw that a one to one ratio of C O
two per I mean could be achieved and

here's what we think happened so
it's either through this water on

getting the approach made by
the midazolam proton transfer or

we're making these carbonate and midazolam
salts so in fact here you can see that

the bulk solvent the emitters all which
is about eighty percent of the volume

is also going to participate in
the reaction so we saw then we could form

the carbon emitters all in salt
by one of these mechanisms.

The fact we're absorbing a lot more C O
two than we thought we would based on

the tree that we saw in the previous side.

So again you know this is a very tunable
system as we've been talking about out of

goods and it means earlier we can.

We can control the properties here
by turning the emitters all and

the I mean to influence both
chemical properties and

physical properties rates
of C O two loading and

viscosity among other properties can be
controlled by varying the emitters OHL and

the nature of the amine So
we've got these excess loadings again but

we're also able to lower the viscosity
to about thirty sent up or

is which is approaching then what a.

Rich solution is which is about
twenty cent of who is at the most so

we're getting very close to what's
already done in industry and

we think you know some experiments are
just going go even lower maybe less than

ten thousand employees by changing
the image is all and I mean further.

So here is a plot we made that just
shows a number of different means with

beautiful images all in the system and
so in most cases we see that

THAT point five loading can be executed at
very low pressures right so the horizontal

dash line represents point five miles
of Super Bowl the mean right one for

every two the vertical dash line is
the partial pressure of C O two and

flue gas so here then we're getting even
these higher loadings at very very low

pressure or exceeding the stoichiometry
at very low pressures especially for

any A in any papers seen as a little
bit different story it actually

pursue potatoes are never dissolved so
this is actually a slurry reaction

I don't really talk much about that
because it's a two phase reaction that

would again be a different process so we
won't worry too much about papers seen but

what we saw there was this was compatible
with a number of different species so

we also then said Well as to as I
mentioned earlier answer to you can be

a problem in carbon capture a process
that it tends to degrade it

means pretty quickly so

what we wanted to see was to see
if it would degrade a Midazolam.

And what we saw then was no in fact we
got a reversible reaction and so to.

An image as all is can bind in this one
to one or two to one type of adequate and

it was can be reversible read it degrade
the images all the way would degrade and

I mean so this is again another
type of reversible and it could and

this paper here this that we
publish this in other is also

a nice review on the reverse of records
and they're out of the work that

has happened here at Georgia Tech we
talked about there and I really think this

is reversible on a good way approach is
the way to go with solvents I think it's

very efficient in the fact that I can turn
on a good words on and off by causing them

some neutral molecules
to undergo a reaction.

So I'll speed it up a little bit here but
also we looked at this stuff for removal

from natural gas we do see a preferential
absorption of C O two over methane and

that has a strong dependence on
the nature of the middle itself so

we see the activities tend to drop off
as the chain length gets longer but

then we see the selectivity the ratio of
C O two to methane suddenly increase above

C sixteen and there is a lot of reasons
for that we think that involves volume but

again we will go too deeply into that but
again these things can separate C O two

from methane so there may be some
application here and natural gas treating.

I won't spend too much time on this
there's a lot of numbers a lot of data

cells basically here that we did compare
the Met metal in that is all solvents to

other commercial solvents and we did find
that they have a number of properties that

appear to be competitive with things that
are E. out there and in some cases perhaps

even better properties than some
of the commercial solvents for

natural gas treating But what I want
to jump in to you now is the work and

we take this right so we've only focused
then on very simple molecules that

we've modified the side chain Well what
if we put in these electron donating even

electron withdrawing troops we're going
to modify the reactivity it is nitrogen

We're also going to control the volatility
of the viscosity by branching and

this is something we've just
started now it's the extension of

what we're doing earlier but
nonetheless I think this.

To be really the key to
controlling both physical and

chemical properties simultaneously and
we think then as the P.K.

increases with electron donating groups
we should see even better performances

in suits you capture because it's
going to be a more basic molecule and

that's precisely where we're going
as I said his image is all platform

that is pretty versatile and as we've
seen it can be a physical solvent for

C O two those are the measurements we make
perhaps not the best physical solvent for

C O two we've seen that
it's a promoter for

a means right an enhanced is the the story
that we can achieve with it within

the mean and see if you capture We also
know there can be a chemical solvent for

C O two in a queer solution and this data
doesn't just come from my group the others

that have published a little bit on this
because when C O two and water present we

can form that by carbonate and the meat
is all being basic enough we get some

equilibrium then we form
this bicarbonate salts.

And really then this is
going to be advantageous for

a number of things perhaps even some
facilitated transport down the road

it's also that we've seen that it's
a chemical solid for us or to H two S.

with or without water we also think we can
make this into membranes which we won't

have time to really get into today and
then finally talk about in that wrapping

up this talk a little bit is that these
images olds now we can go back and

perhaps make new on a good base and
cereals with a little bit more

directed design now that we know some
of the things that we can do here maybe

we can revisit the out of good grades and
look at them in slightly different.

State of matter if you will and
that means liquid to polymer

the transition there so as I've been
mentioning this not being able to

synthesize a missile is actually affords
us new degrees of control on attic liquids

that we haven't even explored before now
historically everyone has focused on A and

then the and I can very it's not
that it's for this particular And

I it's just the representation
here right so we see.

That there is five positions on the middle
going clockwise here one two three

four five everybody's focused primarily
on one now there's reasons for

that one method and
it is always very convenient

starting material you can buy it in four
gallon jugs from Aldrich is very easy

to work with now some of these you have
to go out and synthesize yourself but

there's been very little focus on how
these on a goods would behave there's been

a few reports of the one two three Try
functionalized but I think they've been

they haven't been thoroughly explored
to make any conclusions on them and

what we're really interested then is how
these substitution patterns are going to

affect properties like free volume
I think we need to then go back and

incorporate these into memory means and
take a new look as to how we might do Auld

new materials based on on three four
even five points of substitution so

even greater degrees of control so
a little bit of history then about

why we want to build polymers
from logically quids in fact

the chemistry monikered to so
versatile so modifiable that if we make

these things that former rival groups
we can actually make them into

polymer films that you can use as gas
separation membranes potentially and

that's an alternative technology
that we didn't talk about earlier

persecutes you capture and then the goal
there is to get high flux high throughput

too as well as high selectivity for
C O two in nitrogen.

So again there's a lot of control that
we can have here on C O two separation

in Poljana could membranes we
can change the Palmer side chain

the type of an ion the type of our group
and then we can make these composites and

so I don't have time to talk about
where these composites come from but

really it's a combination of A good
with some free on it could go so

non-polarized and I could sitting
in between the Palmer chains and

that's the way we really were able to
increase per me ability now this is some

old work that we haven't there was never
published but nonetheless we saw that for

the pure pulling our eyes on a good
where there was no freon and.

We saw pretty low.

But as we increase the amount
of good in that composite

we're able to actually increase the C O
two by an order of magnitude or

more without sacrificing
suits your nitrogen.

Right so it goes against this activity
tradeoff which is illustrated here

by that upper bound this is
what's known as a robust and

plot where it plots of membrane
performance flux versus

activity versus permeability and so really
when we want to move up into the right

this is going to be the region that's
most attractive for see it you capture so

it's nice that we can increase probably
ability without losing see the activity

but we'd really like to take things in
this direction up into the right So

really then I think with that with a more
rational view of how to control these

properties we can we can try to make these
things approach that desirable region and

so I won't have much to talk about
those that were just beginning

although I will say what we've seen
already is encouraging and for

those people say well there's
a metal group here or there and

it turns out that metal groups can have an
impact on properties it seems trivial but

in fact it's not only
compared to membrane with and

without a metal group but these two
position we saw a big increase in sea

activity without permeable suffering so

now we're moving up on that
road of Simplot so to say

I really think we need to understand
this in terms of free volume but

if we look at the numbers here we saw that
the membrane on the left which is a point

to rise version of this die methyl and
it is only and and the member on the right

which is a group there we saw that they
had roughly the same C O two permit

abilities about nine to ten barriers
which in that unit of permute go but for

back pain and

nitrogen we saw a very different seal
activities in fact the presence of this

metal group dropped the methane from the
ability by almost fifty percent maybe not

fifty forty percent such that
this you know activity here for

years you mother was forty six were on
the right it was only thirty two so we saw

a really big increase in C O two methane
selectively We saw a modest increase in.

Nitrogen selectively but nonetheless it's
still an improvement so we're thinking

then OK this can lead to a whole bunch of
other structure property relationships and

we really want to find out how
the substitution pattern on this ring

is really going to influence
selectivity and permeable.

And so to do that I think we need to take
an approach that looks at free volume

again this is something we're just
starting from both the middle schools and

the grids and it was a colleague were
employing molecular dynamics simulations

as well as Cosmos term to determine
the relationship between structure and

free volume and we're hoping that these
can tell us how to design better materials

for high performance year
to capture memory and so

on the left here we see the free volume
in images all solvent as a function of

the probe radio which is really just the
size of the particle that's dissolving or

diffusing through that matrix and
we see that they all had

relatively similar free volumes in
these emitters ols and this other

one on the right here is this other paper
here with my grad student we modeled for

a volume in grids using a cosmic them
approach and basically it says that as

we shrink the free volume the subjectivity
of the material should also shrink and

we're hoping then to find the combination
then going up the optical

excuse me optimal combination that
provides high suited to probability and

high ceilings to see the activity but the
first step then is the model of the data

that's already been published and
get a feel for

how this space really is distributed so
we're using both like you dynamics and

these brightly colored as there
are the cosmos surfaces of the cabin and

on and on and so I really think then
free volume could be the key to

improving materials and making better
membranes in this case so in summary then

the conventional on a good roads as we as
we talked about what we found then is that

there are really better suited for

high pressures physical solvent cations
that aren't going to be reactive

when the means were present we had some
serious issues with convention on a grid I

really think that the reversible on a good
grades including the work that's being.

Done In Georgia Tech is really one of the
better or best ways to go with this based

approach because we're seeing then that
we can use neutral molecules to react to

things like C o two S O two etc and really
take advantage of the out of good good

properties without ever having to make an
ionic good per se it's entirely reversible

in the process as we've seen that images
old perhaps hold a great number of

opportunities a solvents as well as memory
and also just purely from molecular design

I think there's there's a whole lot that
hasn't been explored in that realm and

then ultimately now our goals then are to
use a midazolam not only to design

new molecules but also design neurotically
grids for membrane approaches and

really focus on the role of free volume on
the transport properties so just quit your

knowledge to my group of collaborators and
could he turn to you a professor of

Redskin over in Germany Dr Carlisle
from an engineering Also J.P.

Macklemore I've told a few of you about
our chemical engineering i Pad apps and

i Phone apps he's been really
instrumental in helping

get those going if you can find
this on the App Store i Tunes We

also have a post several Ph D.
students and I'm a student and

a whole bunch of undergrads working
on these projects right now and

just to acknowledge funding sources
we have do we BOTH any S.T.R. and

engineering working closely with them and
S.F. A.C.S. P.R.F. the U.A.E.

research grants committee and
Dr Bell Moore who does most of my animal

are samples for mean we couldn't do
much without without Ken getting us and

Mars every day so
with that I thank you for

having me here and
I'd be God's answer questions.

To.

You sure.

That your want to.

Know.

How it will.

Yeah I mean with the methods we're using
in the lab it's easy to do fifty grams or

one hundred grams but
you know we'll industry we're going to use

sodium hydroxide to do these things and do
two phase reactions maybe not you'd like

to find some sort of gas phase reaction or
some solution reaction so

I think more time just needs to be
invested in determining whether or

not that's something that can be scaled
But you know what we're seeing in the labs

encouraging if we want to do a kilogram
but like I said we might need tons and

tons of this stuff and I think there
just has to be some R. and D. investment

in looking at the fundamental reactions
that make it is a loss in order to scale

that.

Yeah I think this chart tells us pretty
quickly where we can go with that so

the reason those polymers are so low
is because they're it's clearly part of

the home we haven't introduced any good
that's non-polarized into the matrix So

really then the Green Square here is
the permeability have got a good and just

the support so it's a supportive liquid
membrane so that's probably our program

it is somewhere in the neighborhood
of a thousand barrels right so

this is a polymer that's just the polymer
this is just the pure on a quick wit and

so we saw we are somewhere in between so
there's something a thousand barrels maybe

two thousand if we create the right
combinations it's somewhere in this regime

I don't think we're going to achieve
ten thousand barriers in the system so

I it's a few hundreds of maybe two
thousand optimally in that system.

Yeah and again I mean it's
going to also be dependent on

how thin you can make them
the performance is going to depend on and

getting them less than a none and
less than a micron.

Bill.

Or in the memory.

You know that was just a moderate
Profumo atmosphere is that we haven't

gone into a higher pressure situation but
you know as we were talking about earlier

they have some advantages that
they resist Plus this is Asian and

so yeah I think going to
high pressure once we

have a material we think can perform
is going to be really important for.

The I was just pure gas was proof of

concept.

Yeah I mean that's part of that paper as
well and that was one of the results that

we saw was that in some cases that means
didn't degrade as fast as in water they

were protected there without having a lot
of water around some of those mechanisms

didn't necessarily occur for
I mean degradation but

in other cases it was worse I mean so
I there's a trade off and

as I was saying I'm not sure that they've
got a good approach really is ever going

to be optimal because you might get
some advantages but then a new set of

disadvantages can pop up.

Yeah.

Right yeah.

Yeah yeah and that's I think that's
the term that I like the most is that

in this case this composite that's a not
it's really a nonvolatile plasticizer and

because the polymer backbone is chart.

And that freon equipment is charged
they have some interactions and so

if I squeeze on this it's not
a sponge I don't I don't liquid

isn't going to come pouring out
of the Matrix it's actually you.

Know there but
it's physically cross-linked or

on a crossing having to
think about it yeah.

OK.

So.

You start with.

This.

Yeah you know it's really
good question because.

People are going to use cosmos therm very
skeptically right I mean it's a black box

you make your molecule and
you get some data but what we saw was that

Turner Turner's numbers from these M.D.'s
actually wind up really well with what

cosmos there was predicting for
three volumes in these systems so

really that in some cases if we're just
interested in free via Cosmo might be

the way to go I mean it's quick and
it's easy and if we can confirm and

in some cases from like your dynamics and
we've probably got something

that's trustworthy what cars no doesn't
do well as is predicted like things like

C O two site ability I mean it it's known
that it gets in the right ballpark but

it doesn't get you the precise answer
that Cosmo doesn't line up well with.

You know C O two Cybill So there are some
things that I think we have to stick to my

core dynamics with but other things
that are purely based on the size and

shape of the molecule called
Moore does a pretty good job.

Asked.

Right.

Yeah yeah that's that's precisely
what we're going to use that for

because obviously like an IMAX isn't
cheap compared to Cosmo in terms of time.

Yeah you know for sure I think you
have to show advantages compared to

the long list of other solvents
that we had up there and.

Very I mean I there's any number
of metrics by which to evaluate.

And process for natural gas reading and
specifically H two S.

removal but
here you know we saw that in that.

First year to methane remit than what is
on on all these are pure sites you visit

these are single gas the activities we saw
pretty good still activity for metal and

that is all for
see if you methane and M.P.

is actually considered to be
a very good solvent for H two S.

removal right so we see a selectively
of ten point two So we are we have to

go make these measurements but
by analogy here we have a high selectivity

in metal emitters all we also have a react
reactive site for H two S. potentially So

I think there are some interesting
considerations for H two S. removal but

again when you when you design a process
you've got to evaluate it on a whole bunch

of metrics but yeah that's something we'd
certainly want to try it seems like a good

possibility.

Here.