Light/Matter Interactions in Biology (Prashant K Jain)
1 Video View·Sep 5, 2023
11/29/2016
Light/Matter Interactions at the Nano-Bio Interface Workshop
Professor Jain received his B.Tech. from the Institute of Chemical Technology in Mumbai, India in 2003 and his Ph.D. in Physical Chemistry from Georgia Tech in 2008. During 2008, he was a postdoctoral fellow at Harvard and from 2009-2011 a Miller Fellow at UC Berkeley. He joined the University of Illinois faculty as an Assistant Professor in Fall 2011. He has affiliations with the Materials Research Lab, the Department of Physics and the Beckman Institute. His research interests are in nano-optics and molecular imaging with the goal of understanding and controlling energy transport, light-matter interactions, and chemical transformations on nanometer length scales.
Well, Professor Bashir gave a really beautiful examples of sort of the cutting edge and also
sort of laid out the big challenges, which there are many, which is a great time to be,
which says it's a great time to be in this field.
What I'm going to do, my job is to really focus down on one subfield within the area
of nanobiomedicine, which is the use of light-matter interactions to really either understand molecular
biology or to develop new biotechnologies.
In particular, what I'm going to try and do is to share with you some of the core concepts
and physical principles in light-matter interactions.
And I'll do that by giving you classic examples in the area of nanoscience and plasmonics.
And it will sort of lay the groundwork for a lot of the tutorials and hands-on modules
that you will go through in the afternoon.
So all of the physics that you need for these afternoon sessions can be found in this talk.
Now you know, I'll be going fast, right?
I'll be trying to do that in 40 minutes.
But if you have questions, those can come up during the tutorials, right?
So this is mainly to introduce you to those main concepts.
Now my group works in a few different areas of nanoscience.
In particular, the use of light, plasmonics, and nanoparticles for understanding catalysis.
We are really interested in high spatial resolution imaging of catalysts so that we can understand
them and then use that understanding to make better catalysts, photocatalysts.
These are solid-state materials.
A big area is plasmonics, and particularly the use of light absorption by metal nanoparticles
for sensing as well as for solar-to-fuel conversion.
And then another area is understanding of phase transitions.
And all of these areas actually have required sort of understanding and advances in light-matter
interactions.
And since you're here on campus, we are happy to give you a tour of our lab if you have
some time and share with you the research.
But really something that we are very proud of is not just the research, but this outreach
effort that is going to be the highlight of the workshop, especially day one today.
What we've tried to do is give the community their own set of tools, a toolkit, to design,
understand, and model light-matter interactions in a wide range of systems, systems that involve
intricate nanostructures and their combinations with other materials or biomolecules.
And sort of a little bit of the history.
As you notice, none of my group's research is in the area of biomedicine.
But over the years, you know, I've been in this field since I was a grad student.
Over the last 10 or 12 years, I've always, you know, received emails from other people
or even initiated collaborations from other biomedical researchers who are doing really
neat stuff, cutting-edge stuff, but where they struggle is really the understanding
of the physical light-matter interactions.
And that's where me and my group has come in, helped them out by either giving them
a physical understanding or actually simulating and modeling some of their experimental phenomena.
And then at some point, we realized we can't do this for everyone.
And we really want to equip the community to do this on their own so that light-matter
interactions in biomedicine and biology can be advanced at a much faster pace.
And in particular, this has been achieved through this collaboration, this NanoBioNode.
And one of the highlights of the NanoBioNode, as you will see, are these set of tools.
These are called Nano DD-SCAT, which is sort of the primitive version.
And then we've developed this into a much more advanced version, Nano DD-SCAT Plus,
which can handle much more complex nanostructures.
And finally, the much more recent version is the combination of Nano DD-SCAT with something
known as a functionalization workbench, which has come together in a super tool which can
make these NanoBioHybrids.
And you'll see all of that throughout the tutorials today and tomorrow.
We have about 1,000 users, and already 10-plus publications have used some of these simulations.
These are real research-oriented, serious tools that you can use on your own.
We already have 1,000 users, but really what we're interested in doing with this workshop
is to give you a much more closer look at how these tools work, what are the physical
principles behind them, and also to make you part of our community so that we can
also benefit from your feedback.
We've gotten feedback at various steps which has taken us through this process, but we'd
like to improve, right?
Tool development, science never stops, right?
You can keep improving these tools.
These are open source.
Source code is available.
The workflow is fully available.
If you had programming experience, you can add your own modules and so on and so forth.
So really, we are excited to share this with you and make you part of our community and
maybe early on initiate collaborations to get you started.
But the real goal is for you to be independent in terms of designing and modeling your own
light matter interactions wherever you may need.
This is a large collaboration, in particular driven by students.
Students in my group, in particular two alumni, Sarah and Jeremy, who've now graduated.
They led the initial development, and now Praga is really leading this effort, and she'll
be conducting most of the hands-on modules this afternoon, and you'll get a lot of insight
and training from her.
And an undergrad researcher, Bara, who's been with us since high school.
He was a senior here at uni high, and he stayed with us and helped us through the development
of nanodigiscat, and he'll be coming in tomorrow at the very end and share with you a demonstration
of how he's using these tools for developing plasmon rulers.
Just to give you a sense of the entire development process, and NSF for funding this big effort,
and in particular these tools, both through the NanoBioNode and a career award where outreach
is a major component.
So let's start with why light matter interactions?
They're everywhere, right?
When you read data from a CD or an optical tomogram, OCT imaging, or even a wide range
of biosensors, microscopy, it's all light matter interactions, right?
It's a pervasive technology.
These concepts are everywhere.
But something changed in the last 25 or 30 years.
Light matter interactions became dramatically enhanced by this idea of sub-diffraction limited
focusing.
So over the last 150 years since we understood light, it was thought that you can't really
focus down light to below the diffraction limit, okay?
You can't focus down light below its wavelength, right, to a size scale smaller than its wavelength.
And that really limited the kind of light matter interactions you can have.
So even though we have a wide range of light matter interactions, you know, spectroscopy
and imaging, those were quite limited.
But recently, through the invention of nanoscience and nanoparticles, we've been able to focus
light down to the nanoscale.
And that has dramatically opened up new areas.
Firstly, because what it has done is that it has dramatically enhanced the cross-section
of interaction between light, which has the length scale of wavelengths, and biomolecules
or molecular materials, which are much smaller in their characteristic length scales, right?
And nanoparticles have helped bridge that gap, that disparity in length scales.
So they've enhanced the cross-sections of light matter interactions, but what they've
also been able to do by focusing this light down to the nanoscale, we've now been able
as a community to achieve a higher resolution and higher sensitivity of probing, of probing
matter in general and biomatter in particular.
And lastly, what it has opened up is it has made light matter interactions much more specific
and able to be engineered in many different ways, and I'll give you some examples.
And so really, this idea of sub-diffraction focusing of light has been made possible by
this idea of metal nanoparticles or plasmon resonances, okay?
When you take a metal nanoparticle, like a nanoparticle of gold, and you shine light
on it, metal, this being a metal, has a very high density of free electrons.
And in response to the electric field of the light, these electrons oscillate back and forth.
Okay, this is a dipolar oscillation, and it has a natural frequency, which is known as
a plasmon frequency, which depends on the size and the shape and the type of metal,
okay?
In reality, what this is is a confined form of light, right?
In the far field, light has the characteristic length scale of wavelength, hundreds of nanometers.
But when focused down to this nanoparticle in the form of this oscillation, this plasmon,
it has a much smaller length scale, which is dictated by the size of the nanoparticle.
Okay, that's really what plasmons are.
When you hear the word plasmon, think of them as nanoscale confined photons that can then
be engineered, okay?
And they can be engineered because you can change this natural frequency of this oscillation.
Turns out for nanoparticles of gold, silver, and copper, this frequency lies in the visible
region, okay?
So it really opens up visible light-matter interactions.
Now turns out this isn't a new field.
This has existed for a long, long time, except the people who were using it didn't really
know that they were using plasmons.
These are some artworks from back in the day, right?
This is the famous Lycurgus Cup from the 4th century.
And if you notice, the coating on this cup is made of gold-copper nanoparticles.
They actually took a TEM quite recently, and they found this is an alloy of gold and copper.
And it's pretty neat.
If you put a light bulb inside, it looks red in transmission.
And if you put a light bulb outside and you light this cup up, it looks green because
of the scattering.
It's a classic example of nanoparticles being used as pigments.
And the neat thing is they've lasted for thousands of years.
I actually went to Troyes.
It's a place in France, which is sort of thought to be the stained glass capital of the world.
It's full of stained glass artworks made of metal nanoparticles.
It's amazing.
So they've existed.
Michael Faraday actually made gold nanoparticles.
But really the field opened up much more recently in the last few years because of the ability
of chemists to use colloidal chemistry to make nanoparticles with size and shape control.
This is our analog of 3D printing.
What 3D printing did for engineers, colloidal chemistry did for nanoscientists and chemists
where you can make a wide range of shapes.
This is simply a very small catalog, but now you can make all kinds of octopods with extremely
good control.
And it turns out with shape and size control comes the ability to control these plasmon
resonances, their frequencies, and therefore their colors, which I'll discuss in more detail.
Not only can humans make gold nanoparticles like Michael Faraday did, it turns out bacteria
can also make their gold nanoparticles.
This is an example from my own graduate school days where this is SPD5 bacteria, not a particularly
good bacteria to come in contact with.
But when you expose them to gold salts, they make their own nanoparticles and sometimes
even shaped nanoparticles because of the enzymatic action of this bacteria where these enzymes
work as reducing agents that reduce that gold salt down to nanoparticles.
I'm not going to be talking any more about the detailed idea of synthesis, but really
that's what opened up the field.
And the second reason for advances in the field is the development of electron and optical
microscopy.
So even though the nanoparticles existed, the ability to see them with high enough spatial
resolution has really made them special.
And here's one image I'll share.
This is a simple light microscope being used to look at the scattering from individual
nanoparticles.
And it's amazing.
You can look at these individual silver dimers.
And each one of those dimers is like a polarized emitter.
When you rotate a polarizer, you can see the intensity go in and out in a sinusoidal fashion.
And each one of those is oriented along a different direction.
And so therefore, each polarizer goes in and out at a different angle of the polarizer.
So it gives you a neat example of the high sensitivity with which you can detect and
probe these nanoobjects.
And that really opened up the field.
In fact, if you look at, I'm going to skip that slide.
In fact, if you look at numbers and metrics of the interaction of light with these nanoparticles,
the absorption, their scattering cross-sections, and this is a busy slide, there are a lot
of numbers there, but it's sort of important.
This shows the role of simulations.
Back when it was hard to measure these cross-sections, these optical cross-sections, theory was a
ble to help us actually calculate their optical cross-sections.
And it was found very early on, this was 10 or so years ago, right back in 2006, it was
found that each nanoparticle, each gold nanoparticle, has an absorption cross-section that's typically
1,000 to 10,000 times larger than a dye molecule.
And these dyes were being used all over biology for imaging.
In fact, when you buy a confocal microscope, you can also pick one or a few sets of fluorophore
labels, right?
There are these imaging kits that you obtain.
And it turns out these particles can be much more sensitive because they have much larger
absorption cross-sections and much larger scattering cross-sections, which is the analog
of fluorescence from dye molecules.
But the biggest advantage is, unlike dye molecules, they don't bleach out.
That is why our ancestors were using these nanoparticles as pigments in artworks, which
have lasted until this very day without needing much care.
And that's really what's special about the use of these metal nanoparticles as non-photobleachable
labels, both for imaging as well as other applications, which I'll tell you a little
bit about.
Okay, so some examples, some very classic examples at this point.
This isn't very new, but over the last 10 or 15 years, I want to give you some examples
that will motivate some of your afternoon sessions.
One big example is the use of these nanoparticles in cell imaging.
And that's been achieved through a combination of the high scattering cross-sections.
Like I said, you can actually detect individual nanoparticles because of their scattering,
along with the ability to actually specifically target them to specific receptors and proteins
on the surface of a cell.
This is one classic example where a gold nanoparticle was tagged with EGFR, anti-EGFR
molecules.
EGFR is a growth factor receptor that's overexpressed on cancer cells.
And so when you expose cancer cells to these nanoparticles with anti-EGFR, you get very
specific binding.
In fact, the entire surface of the cell right here is coated with gold nanoparticles.
And this is a simple image from a dark field microscope that's present in every molecular
biology lab.
And that's all you need to detect this cell.
And when you do the same experiment with non-cancer cells, you see lack of any specific binding.
You just see simply the nanoparticles scattered over non-specifically.
And that became a very simple way to detect cancer cells from non-cancer ones.
Sort of gives you the idea of the use of imaging with a, you know, makes it a much simpler
technology.
Now, you know, when you excite a plasmonic nanoparticle, I showed you that cartoon.
You initiate this dipolar resonance, this oscillation.
These electrons now that are oscillating now eventually relax.
They relax in two ways.
They either scatter that light back out, which is known as radiative decay.
And that's what gives rise to that strong light scattering.
And the rest of that, that oscillation is dissipated in the form of heat.
These electrons collide with phonons, with the lattice vibrational modes.
And that's what leads to heat generation.
And that's the piece of the oscillation that's converted into absorption.
So these are two complementary processes.
Light scattering and absorption.
So light scattering has applications in imaging.
Absorption on the other hand is utilized right now for photothermal therapy, right?
You can actually use a laser beam.
And that's what's great about light matter interactions.
You can do things much more remotely than with an electrical excitation or
electrical probe,
where you actually have to get in physical contact with your probe, right?
So the idea here is that, I guess this slide is a little bit harder to see, but
that's not really important.
What I'm trying to show you here is the ability to deliver light
specifically to nanoparticles that have been targeted to cancer cells or
cancer tissue.
And right here, basically what it allows you to do is use these as
contrast agents to deliver that heat locally to cancer cells and
destroy cancer cells specifically by using a laser without hurting neighboring cells
because of your ability to target these anti-EGFR-coated nanoparticles
specifically to cancer cells, okay?
So that's that example right there.
Non-cancer cells don't really get affected as much,
whereas the cancerous cells are destroyed by delivering the laser light
to the nanoparticles, which is then converted into heat.
And this technology initiated in Al-Sayyad's group back in 2006.
And right now, it's actually,
these are a little bit uncomfortable images to look at.
But right now, it's actually being used for treating canine cancer in Egypt.
Rashid mentioned this thing about FDA approvals and
how much time it takes for an application to go into the clinic.
Turns out things in Egypt have moved much faster.
And so the Al-Sayyad group has moved this operation over there.
And they've actually been able to test several, dozens of cases of canine cancer
where the nanoparticles are directly injected into the tumor or
deliver intravenously.
And these nanoparticles, they're peg-coated.
They circulate in the blood cells, in the bloodstream.
And they eventually collect in the tumor cells.
And then you can excite the tumor cells with laser light.
And that allows you to basically laser surgically destroy
the tumor very specifically while being non-invasive to the neighboring tissue.
Okay, so it's a technology that's in progress.
Everyone in the community has been watching it.
But the key idea is what you're using here is the strong light absorption
of these metal nanoparticles, which allows you to deliver heat locally
as opposed to heating up the entire tissue or the entire organ, okay?
And then one example that is sort of brings you much closer
to chemistry is the ability of these nanoparticles to do chemical detection,
chemically specific detection.
So I said scattering and absorption are two complementary processes.
But in addition to those two processes,
this oscillating electron cloud generates an electric field
around a nanoparticle.
And these electric fields are orders of magnitude stronger
than those in a beam of light, okay?
The nanoparticle is able to concentrate that light in a very small region.
And so these electric fields are extremely high magnitudes.
And in these large electric fields,
what ends up happening is vibrational signatures,
vibrational scattering from molecules is usually enhanced, okay?
The vibrational scattering, Raman scattering,
all molecules have vibrational modes due to their functional groups that
have these bond vibrations.
And those give out specific scattering signatures known as Raman scattering.
And that's a pretty useful technology for
probing specific chemical groups.
And turns out in the vicinity of a metal nanoparticle,
the electric field is enhanced to the fourth power.
And the enhancement factors can be as large as six or seven orders of magnitude.
And so right next to a nanoparticle, it turns out you're able to detect
even single molecules and obtain surge spectra from single molecules.
And this is, at least on the laboratory scale,
this is being used for detection of DNA oligomers.
In a chemically specific manner.
And so that's really changed the field of imaging.
Where in addition to actually detecting and localizing your label,
you're able to get chemical information from around it.
So here are some examples where some researchers have been able to collect
surge spectra from different locations within a cell.
Within the nucleus, the cytoplasm, at the membrane.
And they're able to obtain chemical signatures.
They're helping them identify sort of the local chemical environment
at each one of these locations in real time while the cell
is responding to a drug that's been delivered to the cell.
It's becoming a really important technology in many, many ways.
And, but the core physics behind it is this electromagnetic field enhancement.
Okay?
And then one final example is this idea of biosensing, or sensing in general.
These plasmon resonances, like I said, the plasmon frequency is a natural
frequency associated with that electron density within that, those nanoparticles.
Turns out this frequency responds to changes in the medium
around the nanoparticle, right?
So when I take a nanoparticle and I change the refractive index of the medium
around it, the plasmon frequency shifts.
As you go to higher refractive indices, the plasmon frequency decreases, or
it shifts, the spectrum shifts to the red.
And this is one classic example where bare nanoparticles,
when coated with thiol molecules, there's an increase in the refractive index,
and that leads to a shift in the LSPR frequency.
Now, this is a very interesting example where if you excite this nanoparticle
with a laser beam, you can generate localized heat, and you can actually
desorb the self-assembled monolayer of thiol molecules, and
the plasmon resonance shifts back.
And turns out this is now becoming a laser-based drug delivery method,
where drug molecules are bound to the nanoparticle through this sulfur-gore
interaction, and then you can, by delivering a laser beam locally,
you can release those drug molecules.
This is actually becoming a technology.
But the key idea I want to share with you here is that the nanoparticle,
the LSPR of the nanoparticle, serves as a probe of the medium around it.
And this is being used in glucose sensing, where in a specific
recognition elements for glucose are bound to the gold nanoparticle surface.
And when glucose binds, you see a shift, and
that shift is being used to sensitively detect glucose.
Again, these are things that you'll be able to model using the tools.
Okay, very quickly, I want to tell you that the other big reason why
these nanoparticles are popular, and plasmonics is popular,
is the ability to engineer their colors, the strength of the interaction,
the degree of radiative, which is scattering versus non-radiative,
absorption efficiencies of these nanoparticles.
And the physics, as it turns out, for
individual nanoparticles and solution, is fully understood.
Okay, it's not new, it's based on Maxwell's equations.
And Maxwell's equations has applied to
boundary conditions associated with a small nanoparticle, right?
So there's no new physics here, it's all well understood.
And it's basically, if you look at something known as a polarizability
of the nanoparticle, you basically see how the permittivity,
the dielectric constant, as well as the shape,
can lead to modifications in the plasmon frequency.
Okay, so there's a polarizability associated with a given nanoparticle,
which tells you about the ease of interaction of light
with that electron cloud within the metal nanoparticle.
And in particular,
the denominator of this polarizability expression is quite important.
When this becomes close to zero,
that's when you get a maximum in your spectrum.
And that's where the resonance really is.
And that minimum, that resonance condition,
depends on the medium dielectric constant and
the shape through something known as a shape factor.
I'm not going to go into details of this, but
especially because you're going to be dealing with much more complex
nanostructures, where it turns out this physics is no longer useful.
This only works for individual nanoparticles, well-shaped,
spherical, or rod-shaped nanoparticles.
For more complex structures, you need simulation tools,
which is what you're gonna look at.
Okay, so here's one example of the role of theory, one early example.
So as I said, two complementary ways in which that plasmon relaxes,
scattering and absorption.
Scattering's great for imaging, absorption's great for
photothermal therapy.
So if you were a researcher who wanted to pick a nanoparticle, or
you wanted to design a nanoparticle, shape, size, medium, and
you wanted it for imaging, you'd want a high enough scattering cross-section.
Okay, because they compete with each other.
And so this is where theory becomes important,
because you'd want to know, how does the scattering to absorption ratio,
how do their contributions change with all these engineering variables?
And one easy engineering variable, as it turns out, is size.
And here I'm showing you, as you increase the size of the nanoparticle.
So at small sizes, all of the spectrum comes from absorption.
There's really no scattering.
Small particles don't scatter very well.
As you increase the size of the particle,
the contributions from scattering increases.
And in fact, at a size of about 80 nanometers, absorption for
Gore nanoparticles, scattering becomes almost as strong as absorption.
And beyond that size, scattering actually begins to dominate.
And this was calculating using something known as Mie theory,
which has existed for a century now for dealing with spherical nanoparticles.
Okay, where Mie theory solves Maxwell's equations for
spherical objects, all kinds of spherical objects.
But in particular, when applied to nanoparticles,
you can calculate spectra of this sort.
And the key physics here is, as you make the nanoparticle larger,
the characteristic length scale of the plasma in oscillation becomes longer and
begins to match the wavelength of the free space light, okay, the far field light.
When you match these two wavelengths,
the scattering becomes much more efficient.
And the rate at which the plasmon can radiate or
interact with this outgoing field becomes much, much stronger.
And therefore, the scattering cross-sections increase.
And as the scattering cross-sections increase,
in return, the absorption cross-sections decrease.
Because if the radiation is taking place at a much faster rate,
there isn't as much of a contribution or
dissipation into the electron phonon channel of absorption.
So that's the core physics, but you can understand that physics
when you actually perform a simulation and you can look at the spectra and
compare the cross-sections as a function of size, okay?
And turns out, similarly, another tuning variable is shape, okay?
Spherical particles are very well understood.
But when you make interesting shaped particles, in particular anisotropic
particles, it opens up a whole new range of properties and applications.
And I just want to give you one example here.
When you take a gold nanosphere and you turn it into a rod, okay?
I guess, yeah, I'll just use this slide here.
A gold nanosphere has a single plasmon resonance mode.
It's isotropic, okay?
It's a single dipolar plasmon resonance mode right around 500 nanometers.
But when you make a nanorod,
you now have two ways in which the plasmon can oscillate.
One is along the long axis, and the other is along the short axis, okay?
The short axis mode turns out as a very similar frequency as the gold
nanosphere, whereas the long axis mode is at much longer wavelengths.
And in fact, you can tune that long axis mode by changing the aspect ratio.
As you make the aspect ratio of the rod larger,
meaning you make the nanorod longer, you can shift the plasmon resonance,
the long axis mode, all the way into the near infrared.
And this is quite important, as I'll come to that in a second.
And you can treat these spherical particles theoretically.
You can design them, simulate them, model them, model all of their properties,
scattering, absorption, fields, using the Mie theory.
But when you go to nanorod-shaped particles, you can't really use Mie theory.
Okay, Mie theory is very specific to spherical objects, and
that's where the need for high-end simulations comes in, okay?
So why is this good?
What new do these nanorod resonances bring?
As it turns out, I gave you this example of photothermal therapy.
And the initial examples were with gold nanoparticles, gold nanospheres.
And they were being used for skin cancer, surface lesions, okay?
But if you actually want to do in vivo excitation, right?
Deep tissue, a few centimeters deep.
If the tissue is a few centimeters deep,
you need light that can penetrate through tissue.
And it turns out there's a biological water window, right,
between 700 and 900 nanometers.
That's the range of the spectrum where hemoglobin and
water do not absorb as much.
And you get a few centimeters, well, not a few, but a centimeter or
two of transmission through this tissue in that particular region.
And so you would want a nanoparticle that absorbs light right in that region.
And that's where the nanorod comes in.
If you make a nanorod, if I remember,
an aspect ratio of four gives you a resonance wavelength of 800 nanometers.
Okay, that relationship is very well understood because, again,
because of our ability to simulate these properties with very high precision.
And you can make aspect ratio four nanorod and
use it as a contrast agent with a 800 nanometer laser,
which is again a very common laser wavelength.
And turns out that's the nanorod technology developed by
Kathy Murphy's lab here at Illinois and perfected by the group at Georgia Tech
is now being used for the examples I showed you with canine cancer.
Because those tissues are a little bit deeper beyond the skin, okay?
So that's one example, and I'm not gonna talk about the synthesis.
I'm gonna skip that.
And very quickly, I guess,
a couple of you mentioned that you were interested in quantum dots.
So I added this piece last minute while Rashid was talking,
because I really wanna share this with you.
Because you also wanna see what the new directions are.
Michael Faraday knew about metal nanoparticles.
He knew about their special electromagnetic properties.
In particular, they didn't call them plasmons back then, but
he had this classic paper in 1857, okay?
And for 150 years, it's been taught,
whenever you say plasmon resonance, you think of a metal nanoparticle.
Okay, those go hand in hand.
But that has changed because of work that's happened in the community,
in particular in my lab.
What we found is plasmon resonances aren't limited to metal nanoparticles.
You can also induce artificial plasmon resonances
in semiconductor nanocrystals, quantum dots, as well as insulating nanoparticles,
oxide nanoparticles, by simply doping them with charge carriers.
All you need for a plasmon resonance is a bunch of charge carriers that can
oscillate back and forth in response to light.
And so if you take a semiconductor nanoparticle and you dope it,
which you can do with charge carriers,
you can induce an artificial plasmon resonance.
And the first example was in this material called copper sulfide,
which is a classic earth-abundant material for making solar cells.
And a collaborator of mine made these particles of different sizes.
And he was making solar cells, and his solar cells never worked.
And he found that the reason they never worked was these copper sulfide
nanoparticles were behaving sort of like metals.
They were shunting the electrical connection between the two electrodes.
So they weren't working like classical semiconductors.
And so I decided to look at their optical properties back when I was a postdoc.
And it turns out they behave like quantum dots.
Those of you who know about quantum dots know that their band gaps are size
dependent.
As you decrease the size, electron in a box idea,
their band gap increases, and they behaved exactly like that.
As the size was decreased, the band gap shifted to the blue, okay?
And that's really what a quantum dot does.
But it turns out when you look in this unexplored near-infrared region,
which I'll show you in the next slide, they show these extremely strong resonances,
which kind of look like the spectra that gold nanoparticles give you,
except these are in the near-infrared region.
I've turned around the scale here, okay?
That's the band gap, and that's the plasmon resonance.
And turns out these plasmon resonances come from holes,
positively charged carriers, in the copper sulfide lattice,
which are left behind because of copper vacancies.
Copper vacancies is where you have a copper atom missing.
Copper loves to leave its lattice and become copper oxide.
And so when copper leaves, it leaves a hole behind.
And each one of those holes contributes to this charge density
that can then give rise to a plasmon resonance.
Now the hole density within a copper sulfide nanoparticle,
the charge density, carrier density,
is much lower than in a metal nanoparticle like gold.
And that's why these resonances are in the near-infrared, okay?
Through the Drude relationship, the frequency, so
if you can't remember too many of the equations that we share with you,
this is one of the ones that you should try and remember.
The frequency of the localized surface plasmon resonance
goes as the square root of the carrier density, okay?
And in a semiconductor nanoparticle, the charge density,
the carrier density is much lower, and therefore those plasmon resonances
are much lower energy plasmons, and
typically in the near-infrared region of the spectrum.
But what's special is, you can actually,
now that you can induce these artificial plasmon resonances
by artificially introducing charge carriers to a nanoparticle,
you can actually tune them.
You can choose your own doping density.
You can choose your own carrier density, and
can engineer where in the spectrum is your plasmon resonance going to lie.
At very high doping densities, you can basically reach the visible region.
At low carrier densities, you can be in the terahertz region.
Turns out graphene is plasmonic in the terahertz region, okay?
And this has become a big deal.
It started with copper sulfide and copper selenide, but
turns out they're popping up everywhere.
People have found pretty much everything's plasmonic if you try and
dope it, and this is the list of materials, and this isn't even complete.
This was as of last year.
These are all the semiconductors and oxides which have shown plasmon resonances,
because they're either defective or
through engineering one has been able to dope them with charge carriers.
There's a whole new field, and because they're near-infrared, and
one thing that's special, you would ask, metal nanoparticles are great.
They have tunable resonances.
Why would you want semiconducting nanoparticles?
Sure, they have near-infrared resonances, but
gold nanorods can have near-infrared resonances as well,
if you make them long enough.
It turns out these resonances are special, because they can be turned on and off.
Okay, you probably know this, it's a very basic concept in physics,
and semiconductor physics in particular, you can make wires out of metals.
But if you want transistors that make our computer chips,
you can't use a metal, because a metal has electronic conductivity, but
one that is fixed, okay, whereas a semiconductor,
you can tune its electronic conductivity, you can turn it on and off,
which is what you need for a switch, a transistor, a digital circuit element.
And so we think these plasmonic quantum dots, or
semiconductors, can become building blocks for photonic computers.
Okay, just like electronic computers use semiconductor transistors,
photonic computers may be able to use plasmonic semiconductor nanocrystals
that turn their light absorption property on and
off by simply application of an electrical voltage, okay?
This idea has been around where it was thought that someday plasmonic
nanoparticles will become building blocks for
building waveguides for guiding light from one end to the other.
And there were some classic experiments at Caltech,
where they were able to excite a silver chain of nanoparticles at one end,
and they were able to detect light at the other end.
But they eventually realized this technology didn't go anywhere,
because you can't turn this switch on and off.
It's not a waveguide, it's a wire, okay?
Once you pick a nanoparticle,
its plasma and resonance properties are fixed, that's it.
You can tune it while you're in the lab and you can make different shapes, but
once you put it on the chip, that's it.
This is something you can change with a semiconductor nanoparticle,
because those can be tuned on and off, and
you can actually build photonic transistors out of them.
And that's where one of the big exciting things that we are doing in the lab.
I basically said all of that, but that's what's shown in the slide.
Essentially, they can serve as photonic switches and
amplifiers, these plasmonic quantum dots.
And the reason is they are actually tunable, and this is one example.
Right here is a nanoparticle of copper sulfide,
a quantum dot with really no or low charge carrier density.
Fairly pure, not doped.
And then when you inject charge carriers, okay, this is on a film.
You either inject charge carriers electrochemically or
by applying a potential, you can turn on the plasmon resonance.
And the plasmon resonance shifts as you increase the carrier density.
Okay, the Drude model,
the plasmon frequency goes as the square root of the charge carrier density.
So you can turn them on in this example, and you can reverse it back.
You can turn it off by applying the opposite voltage.
You can extract the charge carriers back.
You can extract those holes out, and there you go.
The plasmon resonance turns off.
You can turn these on and off,
something you can't do with metal nanoparticles.
And that's opening up very new fields and new areas.
And one example where they're already being used is because
they're near-infrared resonant,
they're being used as replacements for gold nanoparticles.
These are earth-abundant materials.
They're not known to be toxic, but their studies are still pending.
And they've been actually been able to use these in mice models and
show photothermal therapy with these nanoparticles.
But what I think is most exciting is the combination of their light absorption
with their electrical tunability or switchability,
where we think they could be used as optogenetic probes,
where by shining light you could basically induce voltage activity
because of the interaction between those dopants and light, okay?
So I'm gonna run out of time soon, so I'll give you one example.
One example of a complex nanostructure where really you can't do much with
basic Maxwell's theory and Mies theory, and you really need high-end simulations,
electrodynamic simulations, to really make sense of the problem and
to learn some physical principles that then you can then carry forward
to develop into new applications, okay?
So one great experiment back when I was a graduate student was done by
Chad Merkin here at Northwestern, who developed these amazing sensors,
sensors of DNA, and they call them DNA barcodes, okay?
And the idea was pretty simple.
A single metal nanoparticle, these were gold nanoparticles,
have a particular absorption spectrum with a peak maximum around 520 nanometers.
But if you bring these nanoparticles close together,
such that they begin to interact with each other,
these dipolar oscillations couple with each other.
They talk to each other, and they couple with each other favorably.
And that decreases their plasmon frequency.
And that's seen as a shift in the plasmon frequency, okay?
Chad Merkin was able to use this idea to make a visual sensor.
Works amazingly well.
So he was trying to detect DNA.
So in this example, he had two sets of particles, okay?
One set of particles with a particular sequence of DNA and
another set of particles with another strand of DNA.
And then when you mix these two solutions and you add a Velcro strand,
a strand with complementary sequences to both types of nanoparticles,
it Velcros those nanoparticles together into a three-dimensional network.
And as a result of that, the particles come close to each other,
the plasmons couple, and you see a shift in the frequency through a color change.
The color changes from red,
which is the classic wine red color of gold nanoparticles.
And you make an assembly, they turn into blue or purple, okay?
So you have a visual sensor of this strand.
If you add the wrong strand,
with even a single base mismatch, you don't get this assembly, okay?
So he was able to detect with single sequence specificity a DNA strand.
So great example, okay?
And it turns out you can calculate the spectra of these assemblies
now using these tools that we have available for you.
Now, there were similar ideas that use nanorods to sort of build this sensor
into a much more complex object where you could actually detect orientations, right?
You could imagine now bringing together anisotropic particles like nanorods.
And there, the degree of coupling would depend on the orientation of the two
nanorods.
And there were examples where in literature where the Kamatz group was
able to link nanorods end to end, and
they saw a similar shift in the long axis plasmon resonance mode.
And it makes sense, right?
The plasmons couple, there's a favorable interaction,
the plasmon energy decreases, it all makes sense.
Now, I did something that was very confusing.
I also assembled nanorods.
And I thought I assembled them side to side,
because the chemical method I was using sort of assembled them side to side by
sort of neutralizing the linkers that are present on the sides of the nanorods.
And through TEM, I was able to see that they linked side to side.
But the funny thing that happened in my experiment was my plasmon resonances,
as the particles were assembling, shifted to the blue.
This was quite striking, didn't make sense.
If particles couple and plasmons couple,
why would their plasmon resonance energy increase instead of decreasing?
That's not coupling, that's anti-favorable.
That's anti-coupling or anti-bonding.
Turns out that was true, but we had to test this out.
And this is what brought the need for theory and simulations.
And that's where I got started with these simulations.
And I realized, me theory is no good.
It's a complex structure, okay, a nanorod assembly.
And so, luckily for us, there was already this method available known as
discrete dipole approximation by these two astrophysicists, Drain and Flatau.
They weren't interested in nanoparticles.
They were interested in meteorology,
where they were looking at how dust particles scatter light.
And they had developed this wide range of tools for
looking at light scattering by dust particles.
Turns out that same theory was useful for looking at nanoparticles.
It's, after all, electrodynamics.
And the method's pretty simple.
It takes any, and George Schatz at Northwestern was responsible for
bringing this from their world to chemistry and to nanoscience, okay?
And the idea's pretty simple.
You take a nanostructure and you break it down into small virtual dipoles, okay?
And then you solve Maxwell's equations for
the interaction of each of those dipoles with the electromagnetic wave.
And then you take into account inter-dipole interactions, okay?
That's what all those terms are.
Those are the inter-dipole interactions.
And that's the interaction of each dipole with the electric field of light.
And when you do that, and you solve this set of equations consistently,
self-consistently, you eventually get a solution,
which gives you the polarizability of the nanoparticle.
And every possible optical property that you can imagine can be extracted
from that polarizability, the absorption spectrum, scattering spectrum,
resonances, how they shift with media, electric fields,
all properties associated with the scattering properties of those nanoparticles.
And you can use arbitrary objects,
because all you need to do is break this object down into virtual dipoles, okay?
I'll make a quick note about the dipole density, okay?
There are other methods available.
You've heard of COMSOL and finite difference time domain.
And those methods work, okay?
Those are methods to solve differential equations in general.
And they work for Maxwell's equations as well,
because Maxwell's equations are differential equations.
But this method is very specific to light matter interactions, okay?
The way it's set up, the way the physics is set up makes it fast and accurate, okay?
Much more specific to light matter interactions.
And that's what the really big winning advantage of DDA was.
And that's really the physics behind all the tools that we built,
discrete dipole approximation.
And you can look at arbitrary objects, as I'll show you.
So I'm running out of time, so I'm gonna, I guess, share one simulation.
And then I'm gonna sort of give you a quick overview of what you'll
see over the next two days, and in particular this afternoon.
Let me see if this video works.
This is a simulation or a set of simulations I ran back when I was a graduate student.
So this idea of nanorods being assembled,
I really wanted to understand what happens to the spectra as they come closer.
So in this particular example, the nanorods are coming in sideways or
end-to-end, and I'm looking at the spectrum.
And I made this pretty crude movie back then by stringing together my simulations.
And as the nanorods move closer, the plasmon resonance red shifts, okay?
And when you bring them side to side, and this is what I was able to do with
discrete dipole approximation.
And when I bring them side by side, the plasmon resonance blue shifts.
This confirmed my experiments.
And later we were able to explain this using the idea of bonding and
anti-bonding resonances that I'm not gonna go into.
But really what I'm trying to share with you is that using simulations,
we were able to not only model our experiments, but
also understand the physics behind that interaction,
behind that plasmonic coupling,
which allowed us to develop a whole set of physical principles.
And these physical principles now form the basis of something known as
a plasmon ruler, okay?
This idea of plasmonic coupling is being used in sensing all over for
measuring distances in biology.
Because that plasmon frequency shift
depends on the distance between the two nanoparticles.
I don't have time to go into that, but you'll see examples tomorrow, okay?
Now, really very quickly, I'm gonna skip all of this and
really tell you those simulations that I showed you with the nanorods were great.
I was able to understand something, but it took me months.
I had to go learn this method that was developed by astrophysicists.
And I had to learn coding, and I had to learn some Unix.
And I really had to be good, and I was good at programming, and that was fine.
And this is how it was done, okay?
This is how the nanorods are represented.
Very non-intuitive, a bunch of numbers, okay?
This is the location of each dipole.
And you have to construct these by hand, okay?
So I used a graphing software to put this together.
And it was fun, but it was painstaking, okay?
I could do only one simulation over a week or something.
And that's how all the inputs look like.
That's polarization right there, wavelengths, refractive indices,
very non-intuitive.
And that's what we wanted to change.
There's a bunch of files that you use, and different files are put together.
And then you look at the outputs, and
you get all the textual format that you convert into spectra.
Quite painstaking, and not very intuitive for
someone who's not in the field of actually doing simulations and modeling.
And that's what we decided to change by introducing NanoDDSCAD.
By giving you this very GUI base, it's like using CAD, okay?
Where you can design your own structure of any complexity using this open
source tool called Blender, can make your own structure.
And then we have a module that converts it into dipoles, and
then runs the simulation, and gives you spectra, which are plotted, okay?
Initially, these were separate pieces, but
now with the newest version, NanoDDSCAD+, these are integrated.
The workflow's been integrated into one tool.
You start, and you can plot spectra in real time.
And Progra is gonna show you a lot of that.
You can generate spectra, you can generate fields.
You can quickly look at them, go back, change your structure, okay?
The only bottleneck is the simulation time,
which for a larger structure can take time.
In particular, the dipole density is what matters.
If you increase the dipole density, you get a higher accuracy.
But also, the computational cost goes up polynomially, okay?
So that's something to keep in mind.
But these are fairly fast, as you can see, for complex structures.
One example where we were able to collaborate with a group and
use this tool, where a student in my group and
their experimental group, they were able to use this tool to design sensors.
These are, and I'm not gonna go into details,
these are sensors for measuring binding and measuring binding constants.
And in this particular case, there was a little nano plate
to which a little sphere is attached to a linker molecule.
It wasn't very clear what's the most sensitive location for
attaching that sphere.
But through these simulations, we were able to determine that it's the,
in the vertices, they're the most sensitive location for
putting that sphere to get the maximum plasma resonance shift.
And that cut down the number of experiments by an order of magnitude
that the experimentalists had to do.
Okay, and then the final version of this tool, where it's going,
is it's pretty powerful.
You can construct pretty complex nanostructures, and
you can simulate them.
But now what we've done is introduce this idea of multi-scale design and
simulation.
We've taken two very different tools together.
One is DDA, DDSCAD, to design the nanostructure, treat its plasmonics.
And the other one, which is the functionalization workbench,
which deals with molecular dynamics and structures, biomolecular structures.
You've heard of VMD, maybe some of you have heard of visual molecular dynamics.
We can draw these biomolecular structures.
We've integrated the two so that you can combine sort of a structural rendition
of a biomolecule or a biological structure and
combine it with a nanostructure, and then simulate the properties.
And you'll see a lot of that tomorrow, right?
So the best way to get, and so I'm not gonna go over the summary.
It's a little bit hard to see with the lights and the low contrast here.
But really the program today is really to focus on basic modules
in plasmonics design.
What's the effect of size?
What's the effect of shape?
What's the effect of the medium?
What's the effect of bringing two particles closer together?
How do you simulate the spectra?
How do you simulate fields?
How do you use Blender, design your own structures?
That's what you'll learn in today afternoon session.
And then tomorrow's session is focused more on the molecular dynamics aspects,
molecular structure, biological applications.
So morning lectures will focus on that.
And then afternoon what we'll do is we'll combine the two.
And we'll show you examples where both tools have been combined using this super
tool where you'll be able to use and construct very complex bio-nano
structures or nano bio-structures, whichever way you would like to refer to them.
And finally, Bara, the undergraduate in my lab, is gonna give you a demo
of construction of a plasmon ruler for studying RNA splicing,
which will show you how these two can be integrated.
The true sort of multi-scaling approach to design and simulations, okay?
So I've tried to cover a lot and I've gone really fast, but
we'll be here to answer your questions throughout the day.
And really, the best way to understand this is to do your own simulations,
which you'll be doing this afternoon.
So try and be engaged and ask questions.
Thank you.
If there are specific questions about the lecture, I can address them now.
It's not about the nanodisk, but you said that you are very interested in catalysis.
Yeah.
I'm a chemist, and I learned that plasma particles are generally cool.
That's what we thought for the longest time.
And there were these reports in the community that we never believed, or
we didn't believe for two years, where when you shine light on a metal
nanoparticle, turns out you can induce all kinds of chemical reactions.
One in particular was the generation of hydrogen from water through production.
And you're right, I mean,
a metal nanoparticle is not known to be very good at charge separation, right?
That's one of the critical pieces in photocatalysis.
You can excite with a very high cross section.
You can generate excited charge carriers, but they relax back very quickly,
because it's this electron-phonon coupling and
electron-electron interactions that I talked about.
It turns out in some systems where the actual acceptor, like titania or
say a small molecule, where it's able to sort of adsorb to the surface and
accept that electron with a high enough rate,
you can actually compete against that dissipation channel.
And you can transfer that electron, separate the charge, and
use those charges for doing photochemistry.
So right now, as we speak in my lab, Yang Tzu, who's here,
he's actually developed a method where he's able to count
how many electrons are extracted from metal nanoparticle by shining light.
He's been able to measure corner meters.
And you're right, they're terrible.
They're much worse than semiconductors.
But the great thing is, they don't degrade.
You can shine light on them for hours, they don't degrade.
And then, even though the corner meter is low, the cross sections are high.
So you benefit in terms of the overall yield of products you get,
even though not every photon is converted into a productive electron.
And then another postdoc in my lab is actually using that to do CO2 reduction.
Because what you're doing is,
you're taking advantage of the catalytic properties of the metal,
where the molecules are actually able to adsorb and form activated intermediates.
You combine that with the light absorption and
charge separation abilities, and achieve photocatalysis.
We don't know where this field is going in terms of making it efficient.
For some systems, it magically works, and
we're trying to understand the physics and chemistry behind that magic, okay?
Thank you, go ahead.
Just a quick one, so this kind of tool, does it take into consideration
these electrons transfer from nanoparticles to?
Great question, and the answer is no, right?
The tool really works based on classical electrodynamics,
where electron transfer effects aren't included,
quantum mechanical effects aren't included.
But that's something that's already being worked out in the community,
where, again, you need this idea of multi-scaling.
You take a method such as maybe some of you, or maybe you are familiar with
time-dependent density functional theory, from which you can actually calculate
dielectric functions associated with either certain quantum modes or
electron transfer events, and then use that dielectric function as the input,
which you'll see in this afternoon,
the dielectric function becomes an input in these classical electrodynamic simulations.
So by intermeshing those two together,
you can treat some of the quantum mechanical effects, but not all of them.
And again, that integrated tool is still not available.
Right now, you have to do those separately and
do your own homework to put them together.
But that's definitely something we are thinking about,
in particular with these plasmonic semiconductor quantum dots.
Turns out they're plasmons, but
they're different from metal nanoparticle plasmons in one more way.
They're quantum mechanical, okay?
They're the individual electronic interactions that quantum confinement
actually plays a big role in determining their plasmon resonances.
And so without introducing that quantum mechanical dielectric function,
you may not be able to get their plasmon resonances right,
especially at low carrier densities.
And so that's why we're really hoping to put these two together by combining
TDDFT and DDA.
This is where you all come in, right?
We've developed the tool, but if you have ideas or even have ability to
do your own coding and bring tools together, this is all open source.
You'll be adding a lot to the community, and
we'll be really appreciative.
But step one is to familiarize yourself with the basic technology of this toolkit.
Thank you.
If there are no other questions, I think I'll yield the floor over to Proga.
Ten minutes late, but I think we'll try to get you back in time.
You may be a little bit late for lunch, but
I guess there's a break right before you start.
Okay, so here's what we'll do.
The break was supposed to go until 10, 10.15, but
why don't we give people ten minutes.
So see you back here at 10.22, very high minute resolution, but
let's try and do that, 10.22, okay?
Great, thank you.
Thank you.

