Advanced materials everyone can afford including Nature

[Music]

my name is marcus buehler i am the

mcafee professor of engineering at mit

and my goal is to create bio-inspired

sustainability

we want to be able to create almost any

function out of almost any material by

using

nature’s hierarchical design approach

that reassembles molecules to create

functional diversity out of universal

building blocks

this is a powerful paradigm that i

believe will change the way we construct

as engineers the way we

design as engineers and the way we can

actually address the sustainability

challenges that we’re facing today

the world in 30 years is going to look

very different than the world today

we’ll have a changing climate

and additional constraints on the

environment we have higher population

densities we have less resources likely

and we’re going to have more waste and

this waste is going to include more

chemical diversity and

maybe even nanostructures we also have

different climate conditions

we have changes in temperature and

humidity that we’ll have to deal with

on the other hand we have a lot of hope

and opportunities ahead of us

we expect major breakthroughs in

nanotechnology

we can assemble materials atom by atom

we can discern transformations using the

existing building blocks in waste or in

existing materials to designer materials

that are made

atom by atom and constructed molecule by

molecule with tailored

desirable functions we’ll also be able

to engineer living organisms

biomaterials and look to nature for

inspiration we can work with nature

instead of against nature and create new

solutions for humanity

biomass waste and advanced materials are

inextricably linked and provide major

opportunities for the future

biomass for instance is a renewable

material it’s a carbon sink

and it’s easily available today already

at 1 billion tons per year in the united

states

we can then use these biomass based

materials to create carbon materials

nanomaterials composites

or even energy materials such as

electrodes and batteries

the paradigm we’re using is called

meteoromics the design material is atom

by atom

just like nature constructs materials

with advanced function from virtually

any resource

we can use the meteorological approach

to build materials molecule by molecule

with advanced function

to do this we look at materials at

multiple levels from the nano scale all

the way to the macro scale

at the nano nanoscale we’re dealing with

chemistry the assembly of molecules

how molecules form micro nanostructures

and mesoscale structures

which then in turn assemble into

hierarchical levels

all the way up to the macro level that

we can see with our eyes

this multiscale structuring is really

the hallmark by which

biological living organisms are

constructed and their exemplary

manifestations of

biorefineries nature is able to

refine materials reconstruct reorganize

materials

from virtually any source we think about

spiders for instance spiders are amazing

species they build spider webs

in two dimensions and three dimensions

they create

very complex material constructions for

various purposes such as protecting

their prey

protecting their young their offspring

and these silk materials are all made

from what we call proteins or amino

acids

another kinds of protein materials are

those found in the ocean such as the

glues in marine materials like muscles

they create incredible glue materials

that work underwater in sea water under

virtually any conditions

we also find biological materials made

from proteins in the human body

such as in our cells in our nerve cells

in our skin cells to organ cells

virtually anywhere we are made from

proteins proteins are

nature’s choice to build materials and

proteins are really constructed from

very simple chemical building blocks

called amino acids these amino acids are

the same building blocks

wherever you look in nature however they

create an astonishing

array of diversity and properties out of

universality comes diversity

that’s the hallmark of nature can we

mimic these processes and construct

superior materials from universal or

simple building blocks

this might help us solve the

sustainability crisis we’re facing today

in fact nature uses this hierarchical

patterning approach to construct

materials nano to macro

and we can create multi-functionality

even by reassembling nature’s building

blocks

along the way as we need them that way

we can make materials that aren’t static

anymore these materials can be adjusted

modified

as the need changes we call this the

universality diversity paradigm which is

one of the most

foundational aspects of how materials

are used in nature

and could provide an important clue an

important solution to the climate crisis

we’re facing

think about a tree growing from a seed

forming the first leaves

these leaves grow provide photosynthesis

great biomass

the leaves fall off in the fall they rot

and they create new soil

and the cycle repeats these kind of

biological mechanisms where

creation of structures is repeated and

recycled

is foundational in nature we look at

biomass

like old leaves or wood or many other

kinds of tailings that we find in nature

they have a very rich set of chemical

foundations within them

these chemical building blocks of

molecules can be utilized if we manage

to reassemble them to create almost any

material function we might need

that’s what nature does when you think

about a spider a spider will eat a fly

and break down the amino acids the

proteins in the fly’s body

to reassemble them to make silk which

one of the which is one of the strongest

materials known

it’s an amazing polymer silk is not made

from petroleum

silk is made from a renewable resource

silk is made from biomass from waste

we cannot yet do that we cannot yet

mimic these processes fully in the

laboratory but provides an

amazing opportunity for future engineers

for scientists to create this future

economy in which we can recycle or

reassess

material combinations from the nano to

the macro level

the paradigm that nature uses is out of

simplicity

emerges complexity and structure and

superiority function and form

we use a multi-skill modeling approach

to address this issue and to solve this

pressing engineering challenge we

simulate materials out and by atom

through the scales from molecular

dynamics to coarse grain simulations all

the way to the continuum level where we

simulate materials

as a macroscopic object that has no

internal structure

by integrating multiple simulation

paradigms we can provide powerful

solutions to how materials work and

function

and we can design them for instance

think about a spider web we can look

inside the spider web

and see that the spider web has internal

structures there are atoms inside

molecules proteins inside and these

proteins are assembled in certain

architectures

if we understand how these proteins are

assembled we can mimic this we can make

our own protein materials

we don’t necessarily have to make a

spider web we can make materials that we

need for

for engineering for instance filtration

devices batteries

structural materials for construction

and the list goes on

this bottom bottom-up approach is very

powerful and can provide a direct

connection between the genetic sequence

in a material

all the way to the functional level we

can either engineer the dna

to design how proteins fold or we can

make materials in the laboratory

synthetically and assemble them atom by

atom

thereby we can take advantage of

foundational processes like size effects

and materials where

the small scale the small length scales

that can be controlled by these

nanotechnological approaches

can provide superior function in other

words materials can become resilient

if we make the nanostructures small

enough so they can prevent

from fracturing and being fragile that’s

one paradigm in which nature uses

defects

to the contrary defects are used to

create strength

out of weakness comes strength and the

key to this is architecture all the way

down to the nano scale by creating

confinement effects where geometry

plays a key role the hierarchical design

paradigm is a very powerful way

used by nature to generate function out

of weak building blocks

for example you can look at bone bone is

made from two components protein like

jello which is very wobbly

and chalk or minerals which are very

fragile and brittle

by combining jello and chalk into a

material

across hierarchical structuring bone is

created and bone is one of the toughest

materials we know

similar materials are seashells like

conch shells which have very very high

toughness values

we’re using this approach to design with

nature instead of against nature

provide solutions that are sustainable

and work with nature

we’ve utilized this approach in a

variety of ways for instance we’ve

utilized

silk made by silkworms and created

cocoons and then re-engineered the

materials in this cocoons to 3d print it

into

various kind of textiles into very

strong materials into tunable materials

or even medical devices or even engines

or motors we can use for tunability and

functional materials i’ve also made

filtration devices out of silk

a simple process where we reassemble and

microstructures the nanostructures that

nature is creating in the in the

formation of silk fibrils

we can create manual porosity mimicking

the structure of silk cocoons

but scaling them down all the way down

to the nano levels so we can filter

out molecules this might be another

powerful way to address pollution

silicon inspired materials and devices

are very interesting because silk is a

biomaterial

that’s compatible with the human body

and other environmental systems

it’s not a synthetic polymer it’s

something that nature creates you can

eat it

and you can work with it and it’s

inexpensive so creating filtration

devices out of silk

or advanced electronics is a powerful

way of working with nature instead of

against nature

the filtration devices are very

effective and can filter out very small

molecules such as heavy metals

metal particles and other kind of

organic substances

a second approach we’ve been exploring

in my laboratory is to use waste

and rearrange the molecules inside the

waste to create

future patterns of molecules that

resemble those

found in nature in other materials to

create superior function

thereby we can mimic what the spider

does the spider will eat flies

have offspring create new silk and these

silks are powerful and very effective

in creating advanced function we can

mimic this by using a process called

hydrothermal processing or hdp in which

we use

pressure and temperature to create new

materials

by using high temperature and high

pressure we can use water as a solvent

instead of relying on petrochemical

other aggressive chemical substances

which are toxic using water in a

supercritical state

allows us to create a reactor condition

in which we can transform

biomass or waste into three main

components a

solid which is a carbon-rich material a

biofuel which provides a foundation of

creating alternative petrochemical

sources

as well as a liquid phase which can be

utilized as an adhesive

we’ve also mixed these biomass sources

that we’ve extracted from hydrothermal

processing

with silk thereby we can mimic what

nature does

combining an existing material like silk

which is amazing

which has amazing properties with the

kinds of materials that nature has

created

you through waste using this

hydrothermal process by combining the

processes created using hydrothermal

processing

with other biological materials like

silk we can create composite materials

and take advantage of the best

properties of both components

and engineer new properties into this we

can develop for instance conductive and

flexible biomaterials

that can reach strength and

cytocombability by using silk

we can also reach environmental friendly

less expensive processing conditions and

packaging conditions for food for

instance

to enhance materials further by using

grapheno carbon nanotubes

we can overcome the suitability issue of

conventional grapheno carbon nanotube

materials

by using kite in the wood or activated

carbon in a recent research study we’ve

used

shrimp waste to create chitin-rich

materials that we’ve then

processed in the hydrothermal processing

plant to create

materials that we’ve used to make

electrodes for flow batteries

this could provide a solution to both

the waste problem

as well as providing new battery

technologies that can be very powerful

in storing alternative sources like

solar or wind

we’ve also used waste from chemical

treatment plants or sewage treatment

plants

for example from the deer island

facility near boston and

use sewage sludge to process it using

hydrothermal processing to create

a biocode mimicking oil as biobinders as

an adhesive for use in adhesive

construction industry for instance

in all this work we’re paying close

attention to the techno-economic

analysis in other words

how economically feasible these

processes are and it turns out they are

feasible

if we can scale it up we’re actually

able to utilize waste

by using this nanotechnology

nanoengineering approach to transform

the ingredients to transform the waste

into functional and usable materials we

can also use these products to create

new types of adhesives

for wood-based products for example

particle wood

or plywood where we need a glue or

adhesive to con construct

materials that have strong binding

between wood particles and the

surrounding faces

using biomass provides a way of

utilizing waste

either from sewage plants or perhaps

waste from wood facilities

for example sawdust to combine these

together with existing wood technologies

to create new types of paneling that are

formaldehyde-free

and work on a low energy sustainable

economic cycle

another exciting direction of this work

is to utilize these biomass-based

waste streams and transformations in the

creation of 3d printing materials or

inks

once we have transformed a biomass or

waste into an ink material we can

construct

any geometry any architecture we can

begin to assemble materials atom by atom

from the nano to the micro to the miser

to the macro level all from these waste

materials

the function of materials derives from

the hierarchical patterns across

different scales

nature has taught us how effective this

paradigm is the concept is now

can we learn from computational methods

how to actually construct these

materials

you can imagine if you’re thinking about

hierarchical patterning the design space

is gigantic it’s very big

and it’s very difficult to find

solutions for where to put materials

what to print or what kind of materials

to actually make in the first place

that’s why we use machine learning and

artificial intelligence to solve this

problem

in machine learning we have methods that

can very accurately capture

hierarchical patterns think about image

recognition or face recognition

these algorithms for example implemented

in convolutional neural networks

use deep layers within the neural

network to detect features across

length scales and time scales these

features are used to decide whether an

image or photo

is a car tree or cat or dog in our case

we’re using these convolutional neural

networks to determine

what function the material has by

looking at its microstructure across

length scales from the protein length

scale which is basically the code of dna

the chemical composition the micro scale

the mesoscale all the way to the macro

scale shape and form of the material

deep learning is a powerful way of

capturing structured process function

relationships that are otherwise very

difficult to understand

the many other machine learning methods

that can be used in addition to

convolutional neural networks

such as deep learning based methods

based on game theory or

gans generative adversarial neural

networks which basically implement a

game theoretic approach where multiple

neural networks

multiple ais play games with each other

to find solutions to physics problems

these solutions can not only be

the solution to a physics problem like

the equilibrium forces but they can also

be

the solution to a design problem where

the algorithm will determine the optimal

design given a set of constraints

this is extremely exciting because it

enhances human creativity

in fact we can complement or supplement

human creativity by creating

augmented reality or virtual reality

environments where we can interact with

the ai system

with our senses and we can make things

seem that we cannot yet see

like forces in a material by going into

an augmented reality environment

we can see forces for instance we can

see magnetic fields

we can see things that our own senses

our own eyes cannot yet recognize

but the ei algorithm can make it visible

to our senses

and we can then use this in the design

process in other words we can see

the immediate impact of how changing the

process the nano structure

the meso scale structure the shape of a

material affects certain properties

properties we can usually not see which

are however very important for

engineering applications

like the forces in the material are

critical to prevent materials from

breaking

now we can see them using ai and

algorithmics algorithmic development is

like virtual reality

that way we can push the frontier to the

next level in which we understand how

hierarchical structures can result in

defect halloween behavior

for instance we can make materials from

waste that are superior

that are superior in strength and

resilience in other words they’re very

hard to break

think about glass glass breaks very

easily however

we don’t want to build with glass

because we don’t want to build a

structure a house a car a train an

airplane out of blast it’s very fragile

using nature’s design paradigm we can

transform brittle elements like minerals

or glass particles

into structures that are very tough as a

whole by creating these architectural

features

from the nano to the macro scale and ai

methods can teach us

how to assemble these patterns these

building blocks

this is very similar to the kinds of

problems people have already solved

using ai methods like how to play chess

or go

which can be very effective and have

proven to be a very powerful way of

solving

complex game theoretic approaches now

we’re using ai machine learning to solve

similar problems

but instead of solving how to play go we

can solve the problem the puzzling

problem

of how to design the best possible

material out of waste

that is an exciting future that allows

us to mimic nature

and yet train the problem develop a

problem solution that mimics it

such that we provide solutions for our

own problems for own challenges that

we’re facing today

such as creating high volume tough

resilient materials

green materials carbon sinks or to

create electrodes for batteries or

filtration devices or

robots or materials that have actuation

properties that are smart that can

interact with the environment

these are all challenging problems of

materials design that need

a revolutionary approach such as ai and

machine learning

now to make these materials we can use

3d printing we talked earlier about 3d

printing as a powerful way

of assembling materials atom by atom

micro by micro

all the way to the macro level in fact

we’re using these multi-scale

additive manufacturing techniques to

then assemble these waste stream derived

inks

into materials that can then be created

and applied in various industrial

settings

we can make materials with tailored

properties we can dial in we can decide

exactly

what kind of strength the material

should have kind of elasticity

when it should break what kind of

tunability properties it should have

or how it would interact with the human

body in cells or animals or other types

of environmental systems

and all of these materials can be made

with nature they can be made out of

polymers or chemicals

that are actually found in nature so

think back about the material like silk

silk is found in nature proteins are

found in nature all of us are living

examples of how nature uses proteins to

build life

now we can create new material solutions

new technologies new electronics the

future of computing perhaps

future computing architectures out of

these silk based

or protein based or amino acid based

materials which are exciting solutions

genetic algorithms allow us to provide a

simulation of evolutionary processes

by combining it with ai and machine

learning which provides

very quick very rapid computational

solutions to complex physics problems to

understand how

design changes affect changes in the

fitness or performance

by combining these ai methods with

genetic algorithms we can essentially

simulate evolution

and within a couple of hours or days in

a computer simulation design

optimal proteins these are all solutions

that can be made today

and that we’re working very hard at mit

in my lab

to create future possibilities for

future generations

to thinking the work with nature instead

of against nature and providing

a platform to use waste as a way to

create the future of materials

the experimental testing that can be

carried out based on the materials we

have created

can be fed back right into the ai model

and therefore create a reinforcement

learning approach where

the performance measured in the

laboratory can

improve the model itself and thereby

improve the design experience overall

now the human input the

human creative input the input of the

engineer comes in

through these augmented reality virtual

reality setups where

we can interact with the computational

models in different ways

we can see things we cannot yet see like

internal forces electricity

magnetic forces or other things and we

can augment the picture the

images that we can see with our eyes

through these ai methods very

effectively

in the augmented reality of virtual

reality setup

through these methods we’re hoping to

mimic nature we’re trying to work

with nature instead of against nature if

we want to address the climate challenge

if you want to address the

challenges that we are facing in future

generations in terms of sustainability

creating more food more resilience more

resources for growing population in the

world

we have to look at nature and we have a

great opportunity now with

nanotechnology

emerging as one of the most exciting

trends in science and engineering

and a platform technology combined with

computational modeling like ai and

machine learning

we can put these things together and

create solutions for future generations

that mimic nature and they build on

nature and work with nature

the future of engineering lies in

thinking about how

natural materials are designed how

they’re created how they interact

the future of materials also features

living materials materials that aren’t

static today’s engineers create

materials that are made in the factory

after they’ve been designed by an

engineer and then shipped off to the

consumer or applied

in a product and they have a lifetime

and they fail and have to be repaired

now that’s a very different paradigm

than the kind of paradigm that nature

uses

uh think about our own body our bones

they grow and if you hurt your bones you

hurt your skin

our body will try to repair these and in

many cases our body is very effective in

repairing injuries or diseases

and we are trying to get to the point

where future materials future

technologies work

just like this where we can actually

mimic the paradigm instead of creating

materials

one time and then repairing them we’re

trying to mimic this paradigm where

instead of creating materials

and shipping them off to the consumer

until they fail we

built in a repair feature we built in

the ability of a material to be

living to be more like us the material

to be more like

human beings more like insects more like

the living world around us

and being able to sense damage to

respond to the environment

and to create entirely new solutions for

engineering solutions in that way

where of course the human need the human

demand for civilization is at the center

of this

and we can work with sustainable

solutions that are carbon sinks

that use waste materials that use the

immense complexity the richness of all

the chemical waste all the chemical

tailings we have today

in the construction of materials the key

to making this happen

are three one nanotechnology to use the

ability to construct materials atom by

atom molecule molecule

second computation to use computation as

a way of understanding

what to build what to design and three

to be able to measure and sense the

environment

either using augmented or virtual

reality or advanced experimental imaging

measurements where data can be generated

in large amounts

which in turn can improve the way we

model using the artificial intelligence

method for instance

to use a reinforcement learning approach

to improve models

as we go along finally this is changing

the way engineers work

the way the physical world is modeled

goes back today

pretty much to newton’s laws where we

solve equations like differential

equations

that have been written down on a piece

of paper the future

might rely on an alternative approach or

complementary approach where in addition

to solving equations like newton’s laws

schrodinger’s equation and others like

this we can also

generate data and behavior and

understand the behavior of physical

systems

directly from observations so instead of

having newton observe how an apple falls

from the tree and then deriving a model

for this

computers can do similar things

computers can observe

how physical systems or living systems

act interact behave

how they work and then from that derive

a

deep learning or neural network based

architecture

one of the biggest threats to climate

change is the use of resources

and that’s what this research really is

about the approach used by nature is

quite distinct

when nature uses the same chemical

ingredients so instead of using

limestone like for cement

or petroleum or other types of

ingredients nature uses

the same chemical building blocks amino

acids to create virtually any function

these protein-based materials have

functions as diverse as acting as a glue

as a sensor as a robot material a

robotic material that has

activation properties that can sense the

environment i can act as a signaling

material like a cable for nerve cells

for instance

it act as a very strong material like

seen in silk being one of the strongest

materials known stronger than steel

and the list goes on these materials are

exciting they’re powerful

and yet they’re made from the same

chemical building blocks so to address

the climate challenge

we believe we need to go to that mode of

operation we want to be able to be in a

position

where we can actually create almost any

function out of almost

any feature