So from a more biological and

chemistry related talk I want to go
back to a more engineering related talk.

My main faculty member in
electrical engineering and

in my research group we
developed sensor platforms.

We have really used the facilities here
of the micro trying to research center

to do so and I want to introduce you
to one of those platforms a mass

sensitive platform that we actually use
currently mainly for chemical sensing but

where we are also looking into
applications in biology and medicine.

So just because I wasn't sure what the
background of this group is I think it's

very very diverse I first want to give
a very very brief introduction into

biochemical micro sensors and focus on
again the the platform the mass and such.

If not from that we are working on
a little bit of a conclusion at the end.

So just a couple of nice pictures to
begin with biochemical sensors and

some of them you might have seen some of
them you might have use already yourself

on the far on the far more left there

is there is more an example of a of
a chemical sensor This is one of these

weather stations that you can
actually buy now everywhere right and

they have miniaturized sensors in
them pressure sensor is one of those

temperature sensor is another one but
the simplest chemical sensor to do you can

actually think of all these weather
stations have you made it to sensors in

it and all of them are or
let's say most of them are actually today.

So they can based there is a company
a small company in Switzerland called

Sun Siri on who markets you might have two
sensors based on silicon technology and

knowledge supply.

To a lot of the companies like
Oregon Scientific who basically sell these

weather stay.

It's the end product.

An end to obligations here for

micro sensors in the in
the in the medical world.

One of them pretty large market
right glucose sensors and

this is just one example that I
show of a of a fairly small device

from Lifescan which is a Johnson and
Johnson company.

So these devices what they have basically
it is an electro chemical cell that is

a sensor which measures the to
glucose concentration in the blobs and

then a more sophisticated device for point
of care analysis from Abbott the ice.

Stat analyzer which is a plot analyzer
where you basically buy the base unit and

then depending on on what you want to
measure you by different cartridges you

insert the blood sample into the car
hundred the cartridge into this handheld

analyzer and within a couple of
minutes you have the paramita that

you do want to measure and again there is
a whole number of different cartridges

available just go on the Web site and
you see a lot of information on that.

So if you look at all these sensors.

If you look at them from a very
very basic point of view.

They're very very similar.

At the input.

You have some chemical or
biological signal.

And if the output you want to have some
sort of electrical signal right in the way

you have the transduction process is
also always very very similar and

this is now your show more for a chemical
sensor system but for a bio chemical

sensor system basically the same way you
have some sort of functionally zation.

Which translates the chemical or
biological information into

a physical information and then you
have an underlying physical transducer.

Which measures a physical property change
a mass of dielectric constant change.

Changing optical absorption or or or and
then translates into an electrical signal.

That's a transducer and then we can have
very very sophisticated signal processing

at the end of the device right until
you get the desired output form and.

So we are doing and my group is certainly
not the only one there are a lot of groups

here at Georgia Tech across
the US across the world

is trying to integrate all that
part here into a small system.

So all the way from this recognition
membrane from the sensitive layer to

the physical transducer to the to
just circuitry all that put it.

Try to put it into a small system.

Now my particular expertise is on that
transducer side I have some expertise

in the sensitive layer but I often
collaborate with colleagues here mainly or

Georgia Tech in terms of
this recognition membrane.

So when you look at the transducer itself.

Again you have different possibilities.

Right.

And whoever speaks here will say well
the one that I'm using is the only

one right and it's the right one to
use and so on and so on and so on.

Right.
But in reality.

Each one of those has advantages and
has disadvantages.

So we talked about the electro
chemical cells that

are in the in the glucose monitor but
they are also in this iced out instrument.

This is probably the oldest
form of devise and

it comes in different flavors right there
are a lot of devices on the market which

are based on electrochemical
sensing different kinds of right.

And again there are experts on the electro
chemical sensors here at Georgia Tech to

our genitalia was such just sitting here
a couple of minutes ago still use he's in

one example with his group then these
are the mass sensitive sensors which could

be based on acoustic wave devices and you
might remember hopefully remember to talk.

By Bill hunt this morning.

Right.
Who was talking about acoustic wave

sensors and then I'm more used
to silicon based devices like

cantilevers as mass and as they have
sensors and what we measure is in the end

we have a we have a small mass scale right
we can detect masses absolute mass or

mass changes very very small mass
changes so when an a light binds to

a surf A's we detect that by
the uptake in mass basically.

Then you can have thermal sensor so
devices that generate heat based on and

based on an interaction color metric
devices are examples there Pallister

devices that are vitally widely used them
for chemical sensing and then last but

not least there is a whole large
group of optical sensors right and

we heard in the previous talk about
fluorescents detection right which

is very very common in the biological
world which in my in my

distribution here would fall on the does
protect the rear of optical sensors.

All right so
I focus really on these mass sensitive

devices and one of the advantages of these
devices will show you a couple of others.

One of the advantages of these devices
is that we believe you can do label for

your analysis right a lot of the optical
techniques you require labeling right

you can get around that with
the with the massive and

with some of the other devices.

However specific city specific binding and

things like that are always something
that you do have to consider.

Of course.

Sensitive layer since this is out of.

A book chapter a good friend of mine and
various Hillerman used

a faculty member in Switzerland and
Zurich wrote about chemical sensors and

this kind of gives you an example of what
possibilities you have in these Indies.

Recognition men.

Brains and I I really don't want to
go into detail into that just saying

you have a lot of possibilities
there are different.

There are many many different interfaces
between the chemical biological war and

the physical world right and some of
them are basically categorized here

down there at the bottom right is forward
for biological sensors you can work with

enzymes protein cells and
you have various possibilities there

in our group we also work a lot with
polymers are sensitive layers for

chemical sensing applications and
typically those are used to detect bola

got a volatile organics in the air so we
use them for environmental monitoring for

example benzene concentrations to
give you one to give you one example.

But you could also think
about using that for

breath analysis right where you have all
the tile organics in your breath and

you could use these type of sensors
to detect some of these substances.

Now coming to those mass sensitive or
resonant devices that we are using.

This is an example off of
work that I was involved

them back at my time in Switzerland this
is my this is my background basically this

isn't a radio for can deliver sensors and
these cantilevers have a correct the wrist

to go resonance frequency so they vibrate
at the correct the wrist and frequency as

such the Golden Gate Bridge has also
a natural frequency at which it vibrates

only those frequencies are a lot higher
because the structures are a lot smaller.

Right.
So what we do now is very very simple we

coat those resonators with our sensitive
layer are their biological entity or

a polymer film for chemical sensing and

then analyze it is absorbed into
these films or odds or by a small.

Onto the surface are bound onto
the surface of these recognition membranes

This makes the cantilever heavier and

you can measure that as a change in
its natural resonance frequency.

That's basically the way these
devices work you can use them for

Chemical Safety application surveillance
monitoring environmental monitoring but

also I hope in areas such
as medical diagnostics.

And the one of the important thing is when
you miniature rise the sensors you cannot

only make one you can make
many on the same substrate

the idea space to functionalize
each one of them individually.

So that you can not only
detect one on a light but

you can detect multiple an alliance.

Right.
So that's the basic concept behind it.

Little bit different.

The way it works for chemicals
because they're this is physical sort

of the chemicals into the sensitive layer.

So you're sensitive layer will not only
absorb benzene but also told you in

an exciting for example would you use
different sensitive layers now and

each one of them absorbs the different
compounds in a little bit different way

and then you can use feature extraction
algorithms at the end to really find out

how much Menzies how much value and how
much you had into your in the air sample

in terms of biological entities the
recognition membrane the interaction is

usually a lot more specific right to use
for example antigen antibody interactions.

So dead you can selectively
detect with one particular

sensor a certain by a molecule.

Now.

After show at least one equation right.

So the physics I'm a physicist by
training I guess that's why I'm

to physics behind these structures is
however incredibly simple right in the end

you have a mass spring dashboard
system that you might have my.

Remember even from your
high school years right.

And there the natural frequency of such
a mastering dashboard system is basically

proportional to the square root of
the spring constant in the mass

we change that mass and with that
we change the frequency right and

this is what we measure so we need and
we detect small mass changes.

Now interesting is the limit of
detection of these devices right and

the limit of fifty action.

Depends on the properties of
your resonator of your library

ding structure and

the smallest change of frequency that you
can resolve this is this Dell Tasmin here.

It also depends on the sensitivity of
your transducer notes mainly a function

of the recognition membrane that you have
on top of it and we come back to that

in a nutshell in order to make
that limit of detection low.

Would you have to be able is detect small
changes in the resonance frequency and

have a high sensitivity
of your transducer.

And one way to to detect small changes
in the in the frequency used to

develop transducers that have a high I.Q.
factor.

At least engineers among you probably have
heard about the Q. factor and a lot of

the work that we are doing is in welding
optimizing to devices so that we get high.

Q. factor operation in air and

in a liquid environment right now
I'm not sure what I bore you here.

So I don't want to go
too much into detail but

from traditional cantilever devices
that have been used by many many groups.

We have we have changed to two devices
that are shown here which is kind

of a small disconnect small disk
rotates constantly back and

forth and it has a characteristic
frequency at which it is rotates right and

we measure that frequency once again
in a sensing application we coat

the surf Ace's of this disc with our
recognition membrane and then the.

Ask uptake again changes to resonance
frequency we can detect that and

have a measure.

All of the mass are taken off the analyzed
concentration in the environment.

Yes You know you could well.

You could more signal if you covered
the whole surface you don't have to but

you get more signal.

So we usually because our word to
position techniques are not yet

a sophisticated we cover everything but
you could only do parts.

So that's the device again.

Right then.

Again fabricated here in the M.R.C.
facilities has then in them in the middle.

It has certain structures so that you can
excite this vibration and that you can

also detect despite abrasion on chip
right that see that's the important part.

So what do you see their typical
frequencies are the order of four hundred

to six hundred kilo herds this is how
fast that thing also laid back and

forth four hundred to six hundred
thousand times per second and

that gives you an idea of the deed to
mass sensitivities So with that one.

Pico gram so ten to the minus
twelve gram mass change results

in a frequency change of about one hertz
out of that four hundred killer ads.

And we have to detect that we have
to be able to detect that and

we can actually quite easily detect that.

Now the important thing if
you can do one you can do for

this is I would normally chip has
four of these devices now on it but

you can do ten you can do you can do
one hundred if you want right just

to give you an idea of the mentions so
to die a metre of this little this

here is two hundred microns so it's of
the order of the diameter of a hair.

Maybe not the diameter of
a hair two times a day.

I mean over here.

All right just to give
you an idea of the size.

So well we can we can.

Excited and detected I don't
want to say too much about that

you can look at the frequency spectrum.

Again this is more interesting forty
engineers you can scan the frequency and

look at the vibration amplitude of the
oscillation and we really see here nicely

one peak this is this resonant mode that
we are interested in very high Q. in

air we get about fifty eight hundred does
this really for an Arabic ation a very

high value for can deliver so you get to
values below with Thousand typically.

And that translates directly into
into resolution of the device and

I skip over it out as well I guess and

start now becoming more towards the
chemical and the biochemical sensing so

for a sensing application what we do
is we incorporate this little disc

into a feedback loop so we build a circuit
around it and it continuously oscillating

always at this natural frequency
always at that particular frequency.

And then you can measure
the stability of the frequency right.

And it's a little bit different.

It's better in a error because the BAM
ping is less than erudite it is actually

in water but you get a frequency
stability in air of the order

of ten to the minus eight that means when
we are at four hundred kilo hurts our

frequency stability is four hundred
kilo Hertz times ten to the minus eight.

So is of the order of a couple
million Hertz right over time.

And in water.

It's a little bit.

D. It's about two orders of magnitude.

Did we lose in water from that we can
calculate what mass change we could

actually theoretically result
with such a structure right and

you get an error you get down
to mass resolutions of the order

off fourteenth I'm told Gram right so
point zero one four

Pico gram or often people like
to have that more translate.

It into a mass loading per
area because your film that

you put down is really an area film you
have there a mass loading of the order

of twenty five Pico gram
per cent a metre square of.

Recognition film area and

these are values which are at
least comparable if not if not

better then what you get
with quartz resonators So

some of you might have heard of
these quartz crystal oscillator.

For bio sensing applications they have

the best ones of them as far as I
know have massive resolutions master

area resolutions in that area so
we can get there with a much

much smaller wise device that can be
much much more easily a rate right.

So many devices in peril.

Now.

Time.

How do I have to fifteen or so.

Right.
When we switch over to that as well and

count out to the actual measurements
in the chemical sensing and

in the biological sensing so.

What we do for the chemical sensing
is we take these resonators and

then we coat them with the polymer and

polymers that we are using are listed
here as a beauty Hydra and

there are many many more I told you so
I look saying to you

are things our typical film that
enrich all the tile organics as well.

So you.

We use currently a spray coating
technique to coat these discs so

we go to some flax which is very close
to here by a little pain brush and

basically dilute our polymer in a soul and

spray it over to describe the sounds
very very crude but it works well.

So.

We want to evolve there and in the.

And what I envision is that we have kind
of a dispensing technique right that we

have a small inkjet system
which can deposit controlled

the mound of of polymer onto
those disk resonators right

if we have that then we can deposit
different systems in a different way and

I think one of the one of the instruments
that is on the wish list for

the clean room is one of
these injecting tools.

David is saying yes in the background.

Maybe already bald.

So you see.

That's how it works here it's wonderful.

So this is how we coat these devices and
then we expose it to different and

allied concentrations.

And one example measurement is shown
here for again for the chemical world.

Just because I have a lot more
data own on the chemical side than

on the biological side so what we have
here is now a sensor coated with P.-I B.

with poorly I said beautifully and
I think the layer thickness is for

Micron here it is for Micron spray
coder and I suppose P.R.B. and

P.C.H. there's two sensors and we expose
them to different concentrations of.

One of the Syleena is Amir's And
would you see is a soon as you expose it

to the unallowed concentration in a this
is done in a chemical measurement set up

the frequency goes down because an Allied
is absorbed in the polymer film Ride to

this system gets heavier and the resonance
frequency goes down and you see that these

frequency changes nicely scale with
the with the and light concentration and

the important thing is now that
the sensitivities are different for

different coatings right so we can do kind
of a fingerprint analysis in order to find

out what an ally and what concentration
we really have in the environment.

Now important thing is to
delimit off detection and

this is this is really only at.

The moment a theoretical
limit off detection.

If we look at the sensitivity
that we have and

if we look at our frequences to billet.

And we can call quite late
a theoretical limit of detection

kind of three times the noise right.

And then there we typically get
values in the low ppm range

for four volatile organics and for

some of the four some of the poly
membranes that we're using.

We're we're pretty confident that we can
we can push that down probably another

order of magnitude or so if you want to
get the low we even that if you want to

get into the low P.P.B. range right
at the moment I believe you have

to do pretty concentration there is there
is at the moment I see no way around.

If you ask me again in five years I might
answer that question differently but

at the moment I would go
with pre-crime situation.

So this is in the air.

You can also do that in the liquid phase
and this is actually one of the main

motivations behind doing this
work because there's very little

micro sensors available for
the liquid phase operation and

we again this is a coded
sensor now we expose it to

silent concentrations in water and
again you see a signal that is more or

less proportional to the unallowed
concentration in the water.

What we still see at the moment is
much much longer time constants and

in the air which bothers me a lot but

we're working on trying to
reduce those time constants.

I think a lot of it has to do with our
poor design of the measurement chamber

at this point of time.

So this is forty this is forward.

This is for D.D.
chemical sensor saw it in the last five to

ten minutes I really want to spend on the
the really preliminary measurements and

we have done with bio
functionalize ation right.

So we.

We have actually functionalized an enzyme
on top of these on top of these resonators

and we have used a covalent bond
ing scheme that we have found in

the literature right here is basically
the reference where we surface activates

the our surf ace which is
a silicon nitride layer and

then we silent in the surf a sweet
Closs across link with the hide and

then we covalent bond our and
ends I am and we used at that time I

was days on two to serve
phase of the sensors.

I'll turn to find.

We are working more now on that is using
gold sort of phases of trial chemistry.

All right so some of you might have heard
bad the advantage should we see there is

that we can nicely do localized
functional use ation right only where our

gold surfaces are which
is more difficult for D.

for the covalent binding that
we have done in the first place.

And then we take our sensors

we take the sensor stride we mount them

on a circuit board this looks all very
very engineering like at the moment but

this can all be put in a nice package and
you don't see the circuitry anymore.

We have to palm someone palm
with an antibody solution

on the days and

then another palm with own Lee with
only a buffer solution in it and

say water there and then we basically run

it over the sensor and then again in a
very very crude fashion at the moment but

I hope we can refine it in
the future we kind of per yard

it clearly injects with antibody solution

point three six nano Graham in
this particular case three times.

Times into the measurements so right and
would you see typical are signals like

that that the frequency is fairly stable
and then a while after the first injection

because the liquid has to travel from
the T. towards the measurement chamber.

We do see we do see a frequency drop and
then again

after the second injection after a while
we see a next frequency drop and

we don't see anything anymore after
the third injection because we think all

the binding sides on the on
the sensor surf Ace are you already

with successful binding experiments but we
need to do it is really very preliminary.

And we need to do a lot more work
in that in that direction but

this is typically what do you see you
have to be very careful with temperature

this divisor D.C. viruses have
to be temperature stabilized but

that's another nice thing about
using a raise of your eyes.

You can use another resonator which
is on code to write do a differential

measurement in order to cancel out
temperature and if effects right.

You could also try to use differential
take Meek's to eliminate unspecific

by ending things like damage right so
the possibility of having multiple devices

not only one sensor really I think open
not opens opens up a number of avenues.

Now.

Told me that all this
year is about cancer so

I have to show at least one
slide about about cancer.

This is not what coming from for me to
is not we're coming from Georgia Tech.

So I want to say that immediately this
is we're coming out of Berkeley and

while ago what they did is they use
cantilever structures that bend

under changes of surface stress
functionalized these these surf ace.

So that they can bind the prostate
specific antigen to the surface.

And then detect the deflection of the
cantilever which is basically proportional

to the P.S.A.
concentration in the sample and

with that with that experiment in the in
the Nature Biotechnology article they

could show that they could detect P.S.A.
and clinically relevant concentrations.

Now this device here is really old
a static device and even has an optical

Reed Right so you need quite a big
system around it in order to

to read out the signal and we believe with
our with our dynamic approach we have one

on one side much much less sensitive
to environmental noise and

on the other side we have a completely
integrated read out scheme right and

we don't need to have
a large system around it.

So does this basically
one of the motivations.

Behind the work that we are doing.

And the springs me to the summary and
hopefully gives you some time for

questions so we have developed a sensor
platform which achieves quite high.

Q. factors and

because of that quite high frequency
of resolutions both in air and water.

We have tested it for
detection of volatile organics in the in

the gas face but also in the liquid
phase we have done first crude

bio sensing experiments we
have to do a lot more dare and

topics we are currently
working on is improving

our packaging scheme improving
the functional is ation the uniformity of

the function is ation reducing on
specific binding local deposition and

then addressing also signs of sensitivity
and sensor drift and on our last slides.

I really want to say that it's
not me who's doing the work but

these are the some of the guys
who are doing the work.

So.
He meantime unfortunately graduated

Sometimes you have to let him go.

He's now in California.

And Stuart they are still around.

This is porousness I Kofi is a chemist and
I work together with them for

the surface functional ization and Ph D.

student who is working on
the surface function is ation.

And then here is some of the names.

Some of the collaboration within
Georgia Tech with chemistry but

also with Professor Jim Spain
in where is he.

Or what department is he is not in
chemistry forgot the department

of us here at Georgia Tech and
then with a group of months group when and

of course you need some money to do
the work we need a lot more from that and

Currently this is this is mainly funded
from the National Science Foundation.

Thanks so much thank.