WEBVTT

00:00.160 --> 00:05.040
First we discovered Earth's liquid core.&nbsp;
Then we discovered a solid core within

00:05.040 --> 00:09.096
that liquid core. Then we discovered that&nbsp;
the solid core has bizarre properties that

00:09.096 --> 00:15.909
are more liquid like. So is the inner&nbsp;
core solid or liquid? The answer is yes.

00:12.667 --> 00:18.107
Links in the description. Now onto the&nbsp;
episode. And no. It may be that Earth's

00:18.107 --> 00:25.147
inner core is in an exotic state of matter&nbsp;
that is simultaneously crystalline and fluid.

00:25.147 --> 00:31.067
The super ionic state the mysteries of&nbsp;
our own planet's interior have in many

00:31.067 --> 00:35.947
ways been harder to crack than&nbsp;
those of the rest of the cosmos.

00:35.947 --> 00:41.067
I mean, we can send probes to the edge of the&nbsp;
solar system and the 46 billion light years to

00:41.067 --> 00:47.067
the cosmic horizon, a largely transparent.&nbsp;
And we can see the most distant galaxies.

00:47.067 --> 00:51.627
But the 6400 km to Earth's center are both

00:51.627 --> 00:55.867
opaque to light and far beyond the&nbsp;
reach of any conventional drill.

00:55.867 --> 01:02.107
The best we can do is to listen to the&nbsp;
faint rumblings of distant earthquakes,

01:02.107 --> 01:07.627
and then try to piece together how those seismic&nbsp;
waves bounce around Earth's interior. To give

01:07.627 --> 01:14.027
you an idea of how seismology has lagged&nbsp;
cosmology and the rest of physics,

01:14.027 --> 01:20.587
it wasn't until 1909, after Einstein, special&nbsp;
relativity, that anyone realized that seismic

01:20.587 --> 01:26.987
waves could teach us about Earth's structure. It&nbsp;
was Andrija Mohorovičić that noticed that some

01:26.987 --> 01:35.547
of the P waves that's primary or pressure waves,&nbsp;
found a shortcut from the earthquake's epicenter

01:35.547 --> 01:42.267
to seismic stations hundreds of kilometers away.&nbsp;
He brilliantly inferred that the waves traveled

01:42.267 --> 01:49.147
through a denser and hence higher velocity region&nbsp;
deeper in the Earth, thereby discovering the crust

01:49.147 --> 01:56.027
mantle boundary. That opened the floodgates.&nbsp;
In 1914, just a year before general relativity

01:56.027 --> 02:01.787
and the following prediction of black holes,&nbsp;
we discovered Earth's molten core. That was

02:01.787 --> 02:08.507
by Benno Gutenberg, who found that S-waves,&nbsp;
for secondary or shear, were blocked across

02:08.507 --> 02:13.147
a broad shadow region on the opposite&nbsp;
side of the planet to their origin while

02:13.147 --> 02:19.307
the P-waves could make it straight through.&nbsp;
Shear waves only propagate through solids.

02:19.307 --> 02:25.227
So he realized that this meant there&nbsp;
must be a liquid interior to the Earth.

02:25.227 --> 02:32.267
But even those P waves play funny around the&nbsp;
molten core, refracting to create smaller shadow

02:32.267 --> 02:40.027
zones. Taking this game one level deeper. It was&nbsp;
in 1936, long after we discovered the expansion

02:40.027 --> 02:46.347
of the universe, and inferred the Big Bang&nbsp;
that Inge Lehmann discovered the inner solid

02:46.347 --> 02:52.987
core. She found that some P waves did enter&nbsp;
their shadow zones, and brilliantly figured

02:52.987 --> 02:59.627
out that these must be reflecting off a solid&nbsp;
region even deeper than the molten interior.

02:59.627 --> 03:05.787
Well, better late than never. These ideas&nbsp;
have been refined and verified over the years,

03:05.787 --> 03:11.787
leading to our modern, multi-layered model&nbsp;
of Earth's interior. Our planet is an

03:11.787 --> 03:18.347
oblate spheroid with a low density crust that&nbsp;
carries slow P-waves and even slower S-waves,

03:18.347 --> 03:23.307
followed by a denser, higher&nbsp;
wave speed rocky mantle that

03:23.307 --> 03:28.907
slowly flows even though it's still solid,&nbsp;
allowing it to support both wave types.

03:28.907 --> 03:33.067
Then we have the liquid outer&nbsp;
core breathing a solid inner core,

03:33.067 --> 03:39.467
both mostly of iron. It all hangs together&nbsp;
neatly, explaining our dynamic planet,

03:39.467 --> 03:45.947
from its volcanism to its tectonics to its&nbsp;
magnetic field. For all intents and purposes,

03:45.947 --> 03:52.907
it's right. So what do we do when new,&nbsp;
even more refined seismic data disagrees

03:52.907 --> 03:58.267
with this model? As our seismic monitoring&nbsp;
became more sensitive and more widespread,

03:58.267 --> 04:03.947
we started to probe the inner structure in more&nbsp;
detail, and we started to see some glitches.

04:03.947 --> 04:09.147
For example, P-waves pass through&nbsp;
the inner core faster in the polar

04:09.147 --> 04:14.187
direction than the equatorial direction.&nbsp;
This is typically explained by the fact

04:14.187 --> 04:20.907
that sound speed in a crystal can depend on the&nbsp;
orientation relative to the crystal's lattice.

04:20.907 --> 04:27.067
If Earth's inner core is crystalline iron&nbsp;
and the lattice orientation correlates

04:27.067 --> 04:32.907
with planetary spin, then the polar&nbsp;
equatorial difference can be explained.

04:32.907 --> 04:38.107
There even seems to be an east west&nbsp;
hemispheric asymmetry in wave speed,

04:38.107 --> 04:43.387
and that's taken as evidence of large scale&nbsp;
lumpiness and or melt regions in the inner core.

04:43.387 --> 04:50.427
One of the hardest things to explain is really&nbsp;
deep in the weeds of the data, but it may reveal

04:50.427 --> 04:56.587
something really startling about the heart&nbsp;
of our planet. S-waves, which are completely

04:56.587 --> 05:03.307
blocked by the molten outer core, do arise in&nbsp;
the inner core through conversion of P-waves,

05:03.307 --> 05:10.667
and it's actually wild that we can even identify&nbsp;
surface seismic waves that have gone through a

05:10.667 --> 05:18.507
P-S, then S-P transition through Earth's core,&nbsp;
then back to P-waves to return to the surface.

05:18.507 --> 05:24.187
But we do find that these core shear&nbsp;
waves travel way too slowly in that

05:24.187 --> 05:28.347
core compared to what we expect,&nbsp;
given the core's composition.

05:28.347 --> 05:35.147
S-waves in the core are also losing energy&nbsp;
much more quickly as they travel than is

05:35.147 --> 05:41.787
expected for a stiff material like crystalline&nbsp;
iron. Now, this inner core S-wave thing is

05:41.787 --> 05:47.147
a real current mystery, and to figure it&nbsp;
out, we're going to need some geophysics.

05:47.147 --> 05:53.867
There are two main ways to deform a solid&nbsp;
compression/expansion. So changing the volume. And

05:53.867 --> 06:00.747
that's what P-waves do. Or shearing changing the&nbsp;
shape. And that's our S-waves. The more resistant

06:00.747 --> 06:07.067
a material is to either of these, the faster that&nbsp;
material will propagate waves in that mode. Think

06:07.067 --> 06:13.227
about it as a more easily deformable material,&nbsp;
being sort of soft and sludgy in that mode.

06:13.227 --> 06:20.907
Crystalline iron is very stiff to both compression&nbsp;
and shear, and so should have fast P and S waves.

06:20.907 --> 06:26.827
And Earth's core does transmit P waves as fast&nbsp;
as we expect. But the S waves are too slow.

06:27.387 --> 06:32.027
To put numbers on this. The relative&nbsp;
shear-ability to compressibility of a

06:32.027 --> 06:39.467
material is encapsulated by the Poisson's&nbsp;
ratio, which is usually around 0.2 or 0.3

06:39.467 --> 06:45.467
for a typical solid. But Earth's&nbsp;
core seems to be at around 0.45,

06:45.467 --> 06:50.587
which is close to the Poisson's ratio&nbsp;
for rubber. So not very compressible,

06:50.587 --> 06:57.787
but very shareable. Let's call such a&nbsp;
material squidgy. That's not a technical term,

06:57.787 --> 07:06.587
so don't use it on any geophysicists. So what&nbsp;
sort of core material could be squidgy like this?

07:06.587 --> 07:12.267
Well we're going to stick with iron in general&nbsp;
because we know that this matches the density.

07:12.267 --> 07:18.267
And because we see that asteroids do develop&nbsp;
these iron cores as the heavy metal drips to

07:18.267 --> 07:20.667
the center during cooling. But there are various

07:20.667 --> 07:24.267
things that could increase&nbsp;
the cores's squidginess.

07:24.267 --> 07:28.827
The first idea people had was to&nbsp;
just add other stuff to the crystal

07:28.827 --> 07:33.467
lattice. Alloying iron crystal&nbsp;
with hydrogen, carbon, oxygen,

07:33.467 --> 07:40.187
silicon sulfur will significantly increase&nbsp;
the poisson's ratio and so lower S-wave speed.

07:40.187 --> 07:45.067
But it's now generally believed that all by&nbsp;
itself, just alloying with any reasonable

07:45.067 --> 07:50.747
amount of these light elements could not&nbsp;
reduce S-wave speed by enough. Another

07:50.747 --> 07:57.307
possibility is to make the core grainy. Rather&nbsp;
than one giant spherical crystal, the lattice

07:57.307 --> 08:03.307
of the core could be fragmented, potentially on&nbsp;
microscopic scales. When molten metal solidifies,

08:03.307 --> 08:10.027
its crystal lattice grows from multiple nucleation&nbsp;
points, resulting in many misaligned sectors.

08:10.027 --> 08:16.667
Over time, these grains can merge and align to&nbsp;
produce larger lattices or even fragment further,

08:16.667 --> 08:21.067
and it's not clear what we should&nbsp;
expect to have happened in Earth's core.

08:21.067 --> 08:25.867
Because the boundaries between grains&nbsp;
lacked the strength of the pure lattice.

08:25.867 --> 08:30.827
The more able to slide against each&nbsp;
other, and this is greatly amplified

08:30.827 --> 08:37.787
if microscopic membranes of molten or heat&nbsp;
softened metal formed between those grains.

08:37.787 --> 08:43.627
This granular ization and melts or pre-built&nbsp;
can potentially explain the high Poisson's

08:43.627 --> 08:48.827
ratio in the core. But the challenge with&nbsp;
this explanation is that there's only a

08:48.827 --> 08:54.187
narrow window of grain size, plus melt or&nbsp;
pre-melt that gives us the right core properties

08:54.187 --> 09:01.147
to explain S-wave propagation, push it too&nbsp;
far and the core goes from squidgy to goopy,

09:01.147 --> 09:06.907
meaning that's waves lose way too much&nbsp;
energy compared to what we observe.

09:06.907 --> 09:13.627
There's also the issue that fine grained crystal&nbsp;
structures have a harder time organizing globally.

09:13.627 --> 09:19.067
And remember, we need some level of global&nbsp;
lattice alignment to explain the polar versus

09:19.067 --> 09:24.107
equatorial speed difference. Most likely&nbsp;
there is some of these granular ization,

09:24.107 --> 09:29.547
etc. going on, but there are strong arguments&nbsp;
that they aren't the whole explanation. We're

09:29.547 --> 09:35.547
looking for a way to retain the general rigidity&nbsp;
of the core, potentially allowing a global crystal

09:35.547 --> 09:42.427
lattice alignment, but to still lubricate&nbsp;
things a bit to slow down them S-waves.

09:42.427 --> 09:47.147
What we need is a new state of matter, a&nbsp;
state of matter that allows the core to be

09:47.147 --> 09:55.067
simultaneously solid and liquid. And so let me&nbsp;
introduce the superionic state. In this state,

09:55.067 --> 10:00.827
we have a rigid crystal lattice of some element&nbsp;
to a molecule with some other atom moving

10:00.827 --> 10:06.987
freely within this structure. There are lots of&nbsp;
examples of superionic materials with various

10:06.987 --> 10:11.707
technological applications, from batteries&nbsp;
and fuel cells to various types of sensor.

10:11.707 --> 10:18.107
We've even observed superionic ice in which&nbsp;
hydrogen moves freely within an oxygen lattice.

10:18.107 --> 10:23.387
In the case of Earth's core, the primary&nbsp;
lattice would be the hexagonal lattice,

10:23.387 --> 10:28.427
formed by iron alloyed with nickel.&nbsp;
Lighter elements like hydrogen, oxygen,

10:28.427 --> 10:34.907
and most notably carbon, could then live&nbsp;
in the interstitial spaces of that lattice,

10:34.907 --> 10:37.467
moving with a freedom that is liquid like.

10:37.467 --> 10:41.227
Now, iron doesn't do this in normal&nbsp;
circumstances. At low temperature,

10:41.227 --> 10:46.187
impurities like carbon tend to stay&nbsp;
at fixed positions in the lattice,

10:46.187 --> 10:52.107
but molecular dynamic simulations have&nbsp;
shown that as temperatures increase,

10:52.107 --> 10:58.107
carbon atoms begin to move between interstitial&nbsp;
sites and above a certain temperature, they

10:58.107 --> 11:04.507
exhibit liquid like behavior. According to these&nbsp;
simulations, the high pressure in a core of the

11:04.507 --> 11:12.907
Earth creates the perfect conditions for the iron&nbsp;
carbon lattice to enter this superionic state.

11:12.907 --> 11:20.187
Even better, in these simulations, the alloy&nbsp;
shows a lower shear velocity than pure iron,

11:20.187 --> 11:25.867
and a Poisson's ratio of around 0.43, very&nbsp;
close to the observed seismic properties

11:25.867 --> 11:31.227
of the real inner core. Simulations are&nbsp;
great because they can suggest and sharpen

11:31.227 --> 11:35.947
hypotheses that we can then go and test in&nbsp;
the real world. Now it's going to be a long,

11:35.947 --> 11:40.347
long time before we can access&nbsp;
Earth's core to do direct tests.

11:40.347 --> 11:45.307
So in the meantime, we need to recreate&nbsp;
the conditions of the core in our labs.

11:45.307 --> 11:51.467
And this is where the new study by Huang,&nbsp;
He, Zhang et al comes in. This team created

11:51.467 --> 11:55.867
a hexagonal close packed iron lattice&nbsp;
with a small amount of carbon dissolved

11:55.867 --> 12:01.227
interstitially. Then they smacked it with&nbsp;
particles traveling at high speed to create

12:01.227 --> 12:07.147
shock pressures and corresponding temperatures&nbsp;
high enough to unlock the super ionic state.

12:07.147 --> 12:13.387
Now, the theme of this episode seems to be sound&nbsp;
speed in different materials. So one side note

12:13.387 --> 12:18.507
the projectile in this experiment&nbsp;
was accelerated by a light gas gun.

12:18.507 --> 12:23.707
The projectile velocity in a firearm is&nbsp;
constrained by the speed of sound. Just

12:23.707 --> 12:27.867
because a bullet is pushed out&nbsp;
of the nozzle by expanding gas,

12:27.867 --> 12:34.987
and that expansion is limited by sound speed,&nbsp;
but sound speed increases with molecular mess.

12:34.987 --> 12:40.747
For air, that's basically dinitrogen. If&nbsp;
instead you fill the chamber with hydrogen or

12:40.747 --> 12:47.307
helium, then a higher sound speed behind the&nbsp;
bullet results in a faster muzzle velocity,

12:47.307 --> 12:52.667
where normal firearms can achieve&nbsp;
muzzle velocities of Mach a few. A

12:52.667 --> 12:59.947
light gas projectile can hit Mach 20 plus or&nbsp;
actual orbital velocities. Because of that,

12:59.947 --> 13:06.267
they're often used to study the impacts of,&nbsp;
say, micro meteoroids on space based hardware.

13:06.267 --> 13:11.707
But in this experiment, the light gas gun is&nbsp;
giving us a sort of particle accelerator for

13:11.707 --> 13:17.867
geology. Anyway, ends tangent mode. And let's&nbsp;
get back to the experiment. Once the supersonic

13:17.867 --> 13:22.427
state was created in the iron lattice, its&nbsp;
properties were studied similar to how we

13:22.427 --> 13:27.787
study Earth's interior by looking at vibrations&nbsp;
on the surface. In this case, with photon Doppler

13:27.787 --> 13:33.547
velocimetry, which is basically pointing&nbsp;
a super precise speed gun at the surface,

13:33.547 --> 13:40.987
now those vibrations reveal the structure of the&nbsp;
interior of the sample, just as with seismology.

13:40.987 --> 13:46.107
In this way, the team could infer various&nbsp;
properties, including the shear velocity and

13:46.107 --> 13:52.507
the Poisson's ratio, and these values were&nbsp;
consistent with what we expect from super

13:52.507 --> 13:58.347
ionic iron. As predicted by simulations. It's&nbsp;
worth mentioning that this experiment didn't

13:58.347 --> 14:03.387
achieve the full pressure and temperature of&nbsp;
Earth's inner core, but at the very least,

14:03.387 --> 14:08.907
we seem to have produced this strange&nbsp;
phase of matter in an iron carbon alloy,

14:08.907 --> 14:16.507
which is an experimental validation of a mechanism&nbsp;
that had previously existed only in simulation,

14:16.507 --> 14:24.347
and the supersonic state produced exhibits strong&nbsp;
shear softening consistent with seismic data that

14:24.347 --> 14:30.187
strengthens its viability as a candidate for&nbsp;
the unusual "squidginess" of Earth's inner core.

14:30.187 --> 14:35.627
If the superironic hypothesis turned out to&nbsp;
be true, it may help explain other things.

14:35.627 --> 14:39.707
For example, preferential flow of&nbsp;
the interstitial carbon along Earth's

14:39.707 --> 14:44.747
rotational axis could partially account&nbsp;
for the polar equatorial speed difference,

14:44.747 --> 14:50.347
which loosens the constraints on global&nbsp;
lattice alignment, and the flow of this

14:50.347 --> 14:55.307
carbon may even participate in the geo-dynamo&nbsp;
effect that powers Earth's magnetic field.

14:55.307 --> 15:00.027
And we should also consider this a victory&nbsp;
for the philosophy at the heart of all of

15:00.027 --> 15:06.747
physics that the secrets of nature are best&nbsp;
revealed by smashing things to other things

15:06.747 --> 15:13.147
really hard. Earth's interior below us&nbsp;
is in many ways harder to explore than

15:13.147 --> 15:18.267
the vast cosmos above. Fortunately,&nbsp;
the secrets of both are encoded in

15:18.267 --> 15:23.547
vibrations that reach the narrow sliver&nbsp;
of our human world from both directions.

15:23.547 --> 15:48.459
And so we build our maps of both&nbsp;
astrophysical and geophysical space time.

15:23.547 --> 15:28.827
vibrations that reach the narrow sliver&nbsp;
of our human world from both directions.

15:23.547 --> 15:48.459
And so we build our maps of both&nbsp;
astrophysical and geophysical space time.
