Chapter 21: Metallic Structures V: Amorphous Metals

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Okay, let's unpack this.

We've dedicated many hours on this show navigating the beautiful, predictable,

and highly functional world of crystalline materials.

Right, structures defined by that long -range periodic order.

Exactly.

Today, we're flipping the script entirely.

We are diving deep into the messy, exciting,

and highly functional world of materials defined by their lack of structure, amorphous metals.

And what's so fascinating here is that while we call them amorphous, which sort of implies randomness, they are, well, they're far from it.

They're not just a jumble of atoms.

Not at all.

This very lack of crystal defects gives them these powerful, unique properties like exceptional corrosion resistance,

superior strength, and critically isotropic magnetic behavior.

Isotropic.

So the same in all directions.

Precisely.

These are not just scientific curiosities.

They are foundational materials in, say, high efficiency power transformation and specialized manufacturing.

Exactly.

Our mission for this deep dive is extremely focused.

We are taking on a single complex source, Chapter 21,

metallic structures via amorphous metals, and distilling every single major idea.

Every definition, every structure, every synthesis concept.

Yes, providing a detailed step -by -step masterclass, making this dense material science accessible and easy to visualize, even if this is your very first time encountering like DRPHS or TRG.

And we'll be navigating this material sequentially just as the chapter lays it out.

We'll start with the fundamentals, you know, defining what disorder even means on an atomic scale.

Okay.

Then we'll explore the theoretical models.

I mean, how do you describe the structure of something without a unit cell?

A huge challenge.

It is.

And finally, we'll cover the essential synthesis techniques, the thermodynamics that govern glass formation, and the cutting edge characterization tools that, you know, prove these structures exist.

Let's jump right in.

So when we use the term amorphous, we're using the Greek definition without shape or structure.

In the context of a solid, especially a metallic solid, what is the hard scientific definition?

How do we precisely contrast it with a crystal?

Well, the defining difference, it rests entirely on order.

Crystalline solids possess what we call long -range periodic order, or LRO.

LRO.

And this means if you pick an atom and measure its position relative to its neighbors, that exact arrangement, the crystal lattice, it repeats reliably across macroscopic distances.

Tens of micrometers or more, right?

Exactly.

Amorphous solids lack that LRO entirely.

So if they lack long -range order, are they just random puddles of atoms then?

Absolutely not.

And this is the crucial point.

They still possess short -range order, or SRO.

Okay, SRO.

So if you zoom in on a single atom and look at its immediate neighbors, the atoms in the local atomic arrangement is consistent, it's predictable.

So it's still packed densely.

Very densely, just like in a crystal.

But if you try to follow that structure out to, say, the fifth or tenth neighbor,

that local pattern just rate,

it breaks down entirely.

The order vanishes over larger distances, maybe beyond 15 angstroms.

And that structural configuration, local consistency, global disorder, that's what drives their unique functional properties, right?

Precisely.

Since the structure is structurally and chemically heterogeneous, there are no crystal planes, no grain boundaries, no preferred axes, the material's properties become isotropic.

Uniform in all directions.

Yes.

This uniformity is vital, especially for soft magnetic materials, where you really don't want the magnetization to prefer one direction over another.

And what's more, the absence of crystal defects dramatically changes their mechanical response.

Which leads to attractive deformation properties.

It does.

But amorphous metals are not easy to produce.

The text really stresses that we can't just melt, say, pure iron and let it cool naturally.

It'll just crystallize every time.

It'll always revert to its lowest energy crystalline state.

We have to suppress crystallization.

And that's the key synthesis hurdle.

It is.

Crystallization is thermodynamically preferred at ambient temperatures.

So to form an amorphous solid, we have to kinetically bypass that process.

Which means cooling it incredibly fast.

So fast that atoms do not have sufficient time to relax, diffuse, and rearrange themselves into that ordered lower energy crystalline state.

This requires immense undercooling, keeping the liquid stable far below its normal freezing point.

And for traditional amorphous metals, that means cooling rates of what?

We're talking 10 to the 6 Kelvin per second, or even greater, a million degrees a second.

Wow.

Okay.

Before we move on, we have to address the most common comparison.

Window glass.

Silicon dioxide.

Glass is also an amorphous solid.

So what distinguishes an amorphous metal from, say, a pane of window glass?

It's a fundamental difference in bonding.

Silicate glasses rely on strong directional covalent bonding.

They form these vast random networks, but those bonds impose severe geometric constraints.

Right.

They have to point in certain directions.

Yeah, exactly.

Metallic amorphous structures, though, are dominated by metallic bonding, which is non -directional.

So the atoms can pack more freely.

Much more freely.

Therefore, metallic glasses are typically characterized as highly dense, close -packed structures.

They're defined not by a random network, but by these things called coordination polyhedra, striving for the most efficient local packing possible.

Much like you see in crystalline metals, just

pattern.

That's the perfect way to put it.

So the source material provides a formal classification of order based on length scales.

If we were to categorize materials based on the range of SRO and LRO, how does that table of organization work?

It's really a spectrum.

The classification system organizes materials based on how far that atomic order extends.

At the very top, we have truly crystalline materials.

Where LRO extends for.

How far?

Beyond 10 micrometers.

Macroscopic distances.

Then if we shrink the size of those ordered regions, we enter the realm of microcrystalline materials.

Order between 100 nanometers and 10 micrometers.

Correct.

And then even smaller is nanocrystalline, where the order is just between 1 and 100 nanometers.

So in those nanocrystalline materials, the building blocks are still little crystals, but they're so small and separated by so many grain boundaries that they lose that overall LRO.

You've got it.

They have SRO in what we might call intermediate range order, but they lack true long -range periodicity across the bulk.

In the final categories.

The truly amorphous the third states.

Here, SRO exists, but LRO is completely totally nonexistent.

So the atomic correlations only exist out to the first few neighbors.

Give you 1 .5 to 2 nanometers and then you hit randomness.

The most immediate and, I guess, tangible proof of this difference comes from the measurement technique we rely on most

x -ray diffraction or XRD.

If we visualize the resulting patterns, how does a perfect crystal pattern transition as we move through nanocrystalline to fully amorphous?

It's a very clear progression.

A perfect crystal gives you these incredibly sharp, narrow peaks in the XRD pattern.

Because every single unit cell is contributing to that constructive interference at precise angles.

Exactly.

But when you move to nanocrystalline materials, you start seeing the

This is the classic fine particle broadening.

So the underlying crystal structure is still there.

You can still see the peaks, but they're just fatter.

Fatter is a good word for it.

And then when we lose LRO entirely,

what does the amorphous signature look like?

I'm guessing it's dramatically different.

Oh, completely.

Yeah.

The sharp peaks vanish.

You see only one or maybe two very broad, diffuse peaks or humps.

And those broad peaks correspond to what?

They correspond to the persistent short -range atomic correlations, primarily the distance to the first coordination shell.

The rest of the scattered intensity is just a relatively smooth background.

Confirming the absence of any repeating long -range structure.

It's the visual fingerprint of disorder.

This phenomenon of peak broadening is so important, it has its own mathematical descriptor, and we really need to understand the physics behind it.

We observe that peak breadth, which we call B, is inversely related to the crystal size T.

Why does making the crystal smaller make the diffraction peak fuzzier?

We have to go back to how diffraction works.

It occurs when X -rays scattering from different atomic planes travel a path length difference that is an exact integer multiple of the X -ray wavelength.

The Bragg condition.

The Bragg condition, exactly.

And that gives you that precise angle, theta B.

And that precise matching works best when you have enormous, essentially infinite crystal structure.

Right.

If you have an infinite stack of planes,

even a tiny deviation from the Bragg angle quickly causes destructive interference, which results in a perfectly sharp peak.

But if the crystal is tiny.

Right.

The crystal size, Te, is tiny.

Let's say only 10 unit cells thick.

The requirement for perfect constructive interference relaxes considerably.

There just aren't enough repeating units for small deviations from theta B to completely cancel out the signal.

I like to think of it like a choir.

If you have a thousand singers all holding the same note, that sound wave is perfectly sharp and focused.

Good analogy.

But if you only have three singers, the sound is fuzzier, spread across a wider range of audible frequencies, even if they're trying to sing the same note.

And that's exactly what's happening.

The signal is spread across a broader angular range and that measurable spread is the peak broadening, B.

Which brings us to the famous Scherrer formula.

It does.

It quantifies this relationship.

T equals 0 .9 lambda over B times cosine of theta B.

Okay.

Let's quickly break that down.

T is the crystal size we want to find.

Lambda is the X -ray wavelength.

We know that theta B is the angle of the peak.

And B is the measured width of that peak, usually taken at half its maximum height.

They have WHM and importantly measured in radians.

So the relationship is clear.

B is inversely proportional to T.

A larger breadth B means a smaller crystal size T.

But a huge challenge in applying this formula in material science is that we have to isolate the true size broadening from what we call instrumental broadening.

You mean even a huge perfect crystal sample when measured by a real world instrument will show some inherent peak width.

Precisely.

The apparatus itself, the X -ray source, the detector, they introduce a minimal amount of broadening.

So you have to correct for that.

You do.

To use show accurately, we have to measure a reference sample, typically a polycrystalline silicon sample with massive known grains, to quantify this instrumental component.

We then subtract that from the measured broadening of our unknown sample.

To isolate the true broadening caused solely by the small grain size.

Yes.

That careful correction is absolutely essential for deriving meaningful results.

Okay.

We've established that amorphous materials lack a unit cell, which means our traditional crystallographic techniques.

They just fail to describe the structure globally.

So if we can't use periodicity, how do we mathematically and geometrically describe the structure of disorder?

We have to shift our thinking.

We use statistical mechanics.

We go from describing where every atom is to describing the probability of finding atoms relative to one another.

And this is done through?

The pair correlation function, GR, and its close relative, the radial distribution function, RDFR.

Okay.

Walk us through the definition of GR.

So GR is the ratio of the actual atomic density at some distance R from a central atom divided by the average bulk density of the material.

Okay.

So if GR equals one.

If GR is one, the density at that distance is purely random.

There's no correlation.

If GR is greater than one, it means atoms are clustering at that specific distance.

The radial distribution function, RDR, is defined as four pi R squared times rho of R.

Physically R represents the number of atoms located within a spherical shell of thickness DR at a distance R from a central atom.

So it gives us a direct quantifiable count of the neighbors.

The count of the neighbors in the local structure, yes.

If we visualize the GR curves for different states of matter, we can really grasp this transition from total randomness to perfect order.

What would the curve for a gas look like?

Well, aside from the hard sphere exclusion radius, you can't find an atom directly on top of another.

The GR curve for a gas is just flat.

It's exactly a GR equals one.

Total randomness, no preferred distances.

None at all.

Now contrast that with a perfect crystal.

Oh, a perfect crystal is a series of infinitesimally sharp, intense spikes.

Each spike sits at a fixed distance, corresponding exactly to the location of the first neighbor shell, the second, the third, and so on, repeating indefinitely.

The spikes never decay.

So the amorphous structure is the middle ground, where we see evidence of close packing, but that signal quickly fades.

Precisely.

The amorphous GR curve shows these intense but broad peaks for the first and second coordination shells.

This shows that local packing is tight and highly correlated.

But the peaks die out fast.

Very fast.

They rapidly broaden and decay back down toward GR equals one.

This confirms that strong short -range order, but it quickly gives way to the average bulk density, signaling the absence of LRO.

And this is the crucial information we get from x -ray scattering experiments.

It is.

It's what validates all our models of amorphous structure.

In a crystal, we use the Wigner site cell to perfectly partition space based on nearest neighbors.

Since an amorphous structure lacks that underlying Brevet lattice, we need a non -periodic analog to describe atomic volume.

And that brings us to the Voronoi construction.

Yes, the Voronoi polyhedra are the amorphous analog to the Wigner site cell.

They're constructed using the exact same rule.

Partitioning space defined by the planes that bisect the vectors connecting neighboring atoms.

Exactly.

But the result is fundamentally different, and that difference reflects the disorder.

So how is it different?

In a crystal, the Wigner site cells are identical in shape and size, and they tile 3D space perfectly and periodically.

Like a perfect honeycomb.

A 3D honeycomb, yes.

In the amorphous state, the Voronoi construction results in irregular, non -identical polyhedra for every single atom.

So they still partition space perfectly, but they all have different shapes and sizes.

Different numbers of faces, different volumes, different shapes, all reflecting the local variations in the backing environment.

I find this fascinating.

It's a geometrical description of disorder.

It shows that even without a lattice, we can still precisely define the local, non -uniform volume occupied by every single atom in the structure.

And it allows us to analyze the local geometry.

We should also mention the Delaunay network, which connects the centers of these Voronoi polyhedra.

Forming a network of what?

A network of tetrahedral packing, which helps us map out the connectivity of the disordered system.

Okay, so the theoretical groundwork for understanding this geometry is built on the random packing of hard spheres or DRPHS model, pioneered by Bernal.

What is the foundational concept of DRPHS and why is it so applicable to metallic glasses?

It works so well because metallic bonding is non -directional, which means metallic atoms behave a lot like hard spheres.

So you can just model them as little balls.

Essentially, yes.

The DRPHS model assumes that the amorphous state is the result of trying to pack these spheres as densely as possible while maintaining randomness.

You're essentially freezing the random arrangement of the liquid state.

And the key result of that dense random packing is?

High coordination numbers, typically ranging from 8 to 12 neighbors for each atom.

Let's quantify that dense packing and compare it to ordered crystals because that difference is highly instructive.

Right.

If we look at the comparison in the source material, the cost of disorder becomes really clear.

The most efficient

Exactly.

BCC is a bit less, 68 % packed with a CN of 8.

The DRPHS model, despite being random, achieves a highly respectable packing fraction of 63 .8%.

So we lose about 10 percentage points of efficiency compared to a perfect crystal, but in return we gain those isotropic properties.

That's the trade -off.

And the average coordination number in DRPHS models is around 12 .0, which shows that locally the packing is nearly as tight as a packed crystal, but the geometry is inherently non -periodic.

So that's 63 .8 % figure.

That represents the maximum packing density you can get without things starting to line up and form a crystal.

It's the highest possible density for a truly random arrangement of same -sized spheres.

When we talk about how atoms locally organize themselves, we inevitably arrive at the concept of icosahedral short -range order, or ISRO.

This structure is foundational to understanding the

metals.

What makes the icosahedron so special for local packing?

The icosahedron is a non -crystallographic structure that is incredibly stable.

For a cluster of 13 identical atoms, one central atom and 12 neighbors, the icosahedral arrangement minimizes the total potential energy.

It's the most energy -efficient way to pack 13 spheres together.

It is.

Calculations using potentials like the Lennard -Jones potential consistently confirm that the energy of an ISRO cluster is significantly lower than that of clusters mimicking close -packed crystal structures like HGP or FCC.

So atoms fundamentally prefer this arrangement locally, but why is it called non -crystallographic?

Because the icosahedron has five -fold rotational symmetry.

Ah, and that's a crystallographic impossibility.

Exactly.

It's mathematically impossible to sustain five -fold symmetry over long distances in three -dimensional space without leaving gaps or requiring defects.

The structure cannot tile space periodically.

This is a perfect conflict, then.

Local stability versus global geometric impossibility.

And this conflict is key to why amorphous materials form.

Absolutely.

When metals are quenched from the liquid state, these energetically favorite ISRO clusters form rabidly in the supercooled liquid.

But they can't link up.

They can't link up to form a global crystal lattice without incurring high strain energy.

They create what we call geometric frustration.

I like that term.

This frustration imposes a vast energy barrier to crystallization.

The liquid essentially gets trapped in this metastable, highly stable, disordered state because the atoms just can't figure out how to arrange themselves periodically.

And we have evidence for this.

Oh yes.

Molecular dynamic simulations and experimental data both confirm that these ISRO clusters are the dominant local building blocks in many rapidly quenched amorphous metals.

Now that we know what they are and why they're geometrically stable, the logical next step is how do we actually produce them?

The overriding objective is simple.

Suppress crystallization.

What is the classic high -speed technique?

The foundational technique is melt spinning.

If you can visualize the process,

you have a molten alloy that's ejected through a small nozzle onto the surface of a rapidly rotating polished copper wheel.

Copper because it's highly conductive to pull the heat out as fast as possible.

Exactly.

Maximizing heat transfer.

The alloy hits the wheel, it spreads into an extremely thin layer, and it is instantly quenched.

And what kind of cooling rates are we talking about here?

I mean, we mentioned it before, but it's worth repeating.

Massive rates.

The contact time is minimal, leading to cooling rates that typically range from 5 to 10 to the 8 Kelvin per second.

Unbelievable.

This phenomenal rate of heat extraction freezes the disordered atomic arrangement of the liquid before the atoms can nucleate and grow the crystalline phase.

And what you're left with are these thin, continuous, amorphous ribbons.

Usually only about 20 micrometers thick, yes.

Besides melt spinning, the source material mentions other high -speed methods.

We have variations like splat quenching, melt atomization, laser melting.

They all rely on similar principles of extremely fast heat extraction over a minimal distance.

But the real game changer in the last few decades has been the development of bulk amorphous alloys, or BAAs.

That's right.

BAAs are special because they're thick.

We're talking centimeter -scale pieces, but they're still fully amorphous.

That suggests the need for those extreme cooling rates has been dramatically reduced.

That's the breakthrough.

BAAs are synthesized using specially engineered alloy systems that resist crystallization so profoundly that they can be produced with cooling rates as slow as 1 Kelvin per second.

Think about that.

Dropping from needing to cool a million degrees per second to just one degree per second.

It allowed for the manufacturing of complex thick components, opening up massive industrial potential.

And briefly, we should also mention the non -thermal methods as well.

Yes, you can achieve what's called solid state amorphization.

You can do this using mechanical alloying, which involves high -energy ball milling to disrupt the crystalline structure through repeated impacts.

Or through irradiation with energetic particles, which physically knocks atoms out of position, destabilizing the long -range order.

To understand why certain alloys like those used for BAAs exhibit such extreme resistance to crystallization, we have to look at the thermodynamics.

And this is quantified by the glass forming ability, or GFA.

GFA is the material's innate stability against crystallization.

The key thermodynamic metric is the reduced glass forming temperature, PRG, defined as the ratio of the glass transition temperature, TR, to the liquidus temperature, TL.

So TRG equals TG over TL.

Let's just clarify what TG is.

TG is the glass transition temperature.

It's the point where the supercooled liquid's viscosity increases so dramatically that atomic mobility effectively stops.

It's the point where we freeze in the amorphous structure.

Exactly.

And TL is just the normal freezing point of liquidus temperature.

So why does maximizing TRG maximize the GFA?

Well, if TG is close to TL, which means TRG is close to 1, the material spends less time in that critical temperature window, where it is supercooled but still has enough kinetic mobility to reorganize into a crystal.

You quickly pass through the danger zone into the rigid glass state.

That's it.

The text uses the conceptual time temperature transformation diagram, the TTT diagram, to illustrate this, particularly the role of deep eutectics.

This is often where students get lost, so let's walk through it slowly.

Okay, so imagine the TTT curve.

It plots the time required for a certain fraction of the material to crystallize at different temperatures.

It typically forms a C shape.

With a nose where crystallization is fastest?

Yes.

And to form a glass, you must cool so fast that your cooling curve completely bypasses that nose.

You miss it entirely.

No, introduce the eutectic composition.

Okay, eutectic points are compositions where the liquidus temperature is minimized.

For an alloy with a deep eutectic, the crystallization pathways for the component crystalline phases require what's called chemical partitioning.

Meaning, the liquid has to split into component phases with different chemical compositions.

It does.

And why does requiring that chemical partitioning help the GFA?

I'm guessing it's a kinetic problem.

It's slow.

It's very slow.

Diffusion is required for that partitioning to occur.

And diffusion is a slow kinetic process, especially in a cooling liquid.

By pushing the T knot curves where the liquid and solid free energies are equal far below the liquidus TL,

we ensure that crystallization requires massive kinetic effort.

Which gives us more time to cool the material and bypass the nose of the TTT curve.

Greatly reducing the required critical cooling rate, which we call RC.

Beyond the thermodynamics, the chapter outlines several kinetic and structural criteria essential for successful glass formation.

First, the balance between driving force and kinetic barriers.

Right.

You need a substantial thermodynamic driving force.

Meaning, the liquid must be significantly supercooled below T knot to even want to crystallize.

But this driving force must be resisted by strong kinetic barriers.

And the strongest kinetic barrier is that need for partitioning

Right, that huge diffusion hurdle.

Secondly,

even if the atoms want to rearrange, they have to be able to move heat away quickly enough.

Rapid heat transport is crucial during solidification to remove the latent heat of crystallization.

Which is why techniques like melt spinning work so well.

Exactly.

And the minimum cooling rate to bypass the nose of the TTT curve is that critical cooling rate RC.

Finally, we revisit the structural criteria, the atomic size rule, which is what makes the bulk amorphous alloys so robust.

And this is the steric hindrance argument.

Glass formation is highly favored when there are large differences in the atomic radii of the constituent elements.

And those elements have a strong negative heat of mixing.

Right, the delta H of mixing.

The large size disparity makes it geometrically impossible to form a dense periodic lattice without high strain.

The system is just structurally frustrated.

Hindering the diffusion and crystal growth necessary for the ordered phase to form.

You've got it.

Let's categorize the successful alloy systems that embody these rules.

Starting with the classic transition metal metalloid systems, or TLM.

These were the original high -speed ribbon alloys.

Right, these combined transition metals, or TL, like iron, nickel, or cobalt, with small metalloids, M, like boron, phosphorus, silicon, or carbon.

Systems like iron, boron, or cobalt, phosphorus.

Those are the classic examples.

They inherently possess the required deep eutectic compositions for good GFA.

How does studying the crystalline phase of these alloys help us understand the short -range order in the resulting amorphous material?

It provides the blueprint for the local geometry.

When these alloys do crystallize, they often form complex intermetallic compounds, like FA3C cementite.

Okay.

These crystalline structures feature specific coordination polyhedra, like octahedra or tetrahedra, where the small metalloid atoms sit interstitially within the packing of the larger transition metal atoms.

So we can infer that the amorphous phase maintains those local building blocks, the metalloid atoms occupying similar interstitial sites surrounded by the transition metals.

But those blocks are just randomly oriented and connected.

That's the idea.

For instance, in high iron amorphous alloys, the FA atoms locally form polyhedra, similar to those found in BCC iron.

But the smaller metalloid atoms ensure that the structure can't achieve global periodicity because they disrupt the regular stacking sequence.

Moving to the second major class, the early transition metal, late transition metal systems, ETM -LTM,

like iron zirconium, cobalt hafnium, or nickel niobium.

These systems are the technological foundation for the bulk amorphous alloys.

They really are.

These systems meet the glass -forming criteria perfectly.

The ETMs, like zirconium or hafnium, are significantly larger atoms.

And the LTMs, like nickel or iron, are typically smaller.

Which guarantees the necessary large size difference.

Couple that with the fact that these elements often have strong negative heats of mixing, delta H -mix, and they stabilize the liquid and impose massive steric hindrance upon cooling.

The source material uses the complex crystalline structure of phetuzir to help us visualize the chaos.

Why is that useful?

Because the complex crystalline compounds that would form, like phetuzir, already contain highly distinct coordination environments.

You get a mixture of large zirconium polyhedra and smaller iron cuboctahedra.

So even the crystal is geometrically complex.

It is.

This complexity shows that even the crystalline phase struggles to accommodate the size mismatch efficiently.

So when you quench the system, that size disparity following the atomic size rule makes it nearly impossible for the atoms to sort themselves out into any long -range periodic structure.

Which favors the amorphous state.

The structure itself seems to be actively fighting against order.

It is a profound level of self -frustration, and that's what ensures that the critical cooling rate plummets, allowing us to cast thick ingots.

The success of the ETMLTM systems led to the development of multi -component systems, the true BAAs which obey the famous empirical rules developed by Inui.

What are the three defining rules for forming a BAA?

Inui established that for BAAs, like the widely studied zirconium titanium nickel copper beryllium systems, you need three things.

Okay, what are they?

One,

a minimum of three, but often four or more component elements.

Two, significant differences in atomic size, generally greater than a 12 % difference between the constituents.

And three, negative heats of mixing between all components.

So this cocktail, it maximizes geometric complexity and chemical stability, yielding a highly viscous liquid that resists crystallization even at slow cooling rates.

That's the recipe.

Let's look at the incredible applications that opened up once BAAs were established, specifically in magnetics.

Ah, the magnetic properties of amorphous alloys, particularly iron -based materials like finite and nanoperm, are revolutionary.

Why?

Because these materials lack crystal anisotropy.

There are no preferred crystallographic directions, so their magnetic behavior is isotropic.

And they lack grain boundaries.

Which are defects that tend to pin magnetic domains.

So if there are no pinning sites, the magnetic domains can flip easily and quickly.

Exactly.

This results in very soft magnetic properties, high magnetic permeability, and extremely low coercivity.

Meaning minimal energy is needed to magnetize and demagnetize the material.

Yes.

And this is critical for high frequency applications, like power transformers and inductors, where energy loss due to hysteresis magnetic friction is minimized, leading to huge gains in energy efficiency.

And often, the peak performance is actually achieved not in a purely amorphous material, but in a metal amorphous nanocomposite.

Right.

These are hybrid materials.

You have a microstructure consisting of extremely fine nanocrystals, maybe 10 nanometers in diameter or less, embedded within a remaining amorphous matrix.

Fine met is a famous example.

It is.

The nanocrystalline grains are so small that the magnetic exchange coupling still operates across the grain boundaries and through the amorphous matrix.

This allows the material to behave macroscopically as one large, soft magnetic unit.

And materials like hip perms.

They combine this soft magnetic behavior with high saturation magnetization and thermal stability, which is crucial for applications that have to operate at high temperatures.

Okay, so we need to dive into the mathematical tools that allow us to prove the existence and describe the structure of these disordered systems.

The starting point remains X -ray scattering, but the theory changes because we can no longer assume fixed periodic atomic positions.

Since atoms are randomly oriented, we have to perform an intensity averaging over all orientations.

The chapter introduces this mathematically via the W scattering equation.

Conceptually, what does the W equation predict?

The W equation, in simplified terms, predicts the total scattered intensity chi of k for materials composed of small, randomly oriented clusters of atoms.

So it's not assuming periodicity.

No, it accounts for the interference of waves scattered by every pair of atoms in that cluster without that assumption.

The text shows how, as the number of atoms in the cluster n increases, the resulting scattering curve changes shape.

It does.

For a very small cluster, the intensity curve just oscillates slightly.

But as you increase n, those oscillations start to become sharper and more defined, mimicking the eventual sharp brag peaks of an infinite periodic crystal.

So the W equation shows the transition from the continuous oscillating scattering of a small, random group of atoms to the discrete sharp peaks of infinite periodicity.

It's the bridge between the two worlds.

The ultimate goal of all this analysis is to get from the measured scattering intensity i of k, which exists in reciprocal space, or k -space.

Where the peaks are.

Back to the real space atomic arrangement, row of r.

This transformation relies on Fourier analysis.

That is the crucial mathematical bridge.

The reduced intensity function phi of k, which we derive from the measured x -ray intensity, is directly related to the Fourier transform of the real space radial distribution function, the RDF of r.

So by performing this mathematical transformation, the inverse Fourier transform, we take the scattering data collected in reciprocal space and convert it directly into the RDF.

And that gives us the tangible real space probability map of atomic neighbor distances.

This confirms the physical structure predicted by models like DRPHS.

While bulk x -ray scattering gives us the average structure, the bulk RDF multi -component BAAs contain several different types of atoms, each with a unique local environment.

We need an element -specific technique.

And that's where Extended X -ray Absorption Fine Structure, or XAAFS, comes in.

The XAAFS is incredibly powerful because it is tunable.

We can tune the incident x -ray energy specifically to the absorption edge of, say, the zirconium atoms in the alloy.

So you can probe the local structure only around the zirconium atoms, ignoring the nickel or beryllium atoms for a moment.

Exactly.

Let's visualize the physical process that generates the signal.

Okay, it starts with X -ray absorption.

An incident x -ray photon with energy above a core electron's binding energy, say, from the k -shell ejects that core electron.

Okay, electron is ejected.

This electron immediately travels outward as a spherical photoelectron wave.

So we have an outgoing wave, and then what happens?

The photoelectron wave encounters the neighboring atoms.

These neighbors act like scattering centers, causing a portion of the wave to backscatter toward the absorbing atom.

And this backscattered wave interferes with the outgoing photoelectron wave.

It does.

If they interfere constructively, the probability of absorption is enhanced.

If they interfere destructively, it is reduced.

And this interference modulates the x -ray absorption coefficient beyond the core electron's absorption edge, creating an oscillating pattern.

That oscillatory fine structure is the EXAFS signal.

Since this interference depends entirely on the distance, number, and type of immediate neighbors, this provides precise site -specific data.

Exactly.

Unlike bulk x -ray scattering, which gives an average structure, EXAFS gives us the precise local coordination numbers and bond distances for a specific atomic species.

Which is essential for validating structural models of complex multi -component glasses.

The process of analysis involves measuring the absorption coefficient, U of K, subtracting the non -oscillatory background to get the oscillating AXAFS function, J of K, and then Fourier transforming, CH of K, to get a site -specific radial distribution function.

Our final advanced characterization tool is Muss -Bauer spectroscopy, a technique that sounds like something out of a physics textbook, relying on the recoil -free resonance absorption of gamma rays.

Muss -Bauer spectroscopy is unique because it probes the nuclear environment, relies on the ability of specific isotopes, primarily iron -57, to absorb gamma rays without recoiling.

A process only possible when the gamma ray energy perfectly matches the difference between the nuclear excited and ground states.

Right.

And the key is that this energy difference is exquisitely sensitive to the local environment of that iron -57 nucleus.

The nuclear energy levels are perturbed by the local electronic density and magnetic field.

That's it.

And since the energy differences are tiny, we measure them by modulating the gamma ray energy using the Doppler shift moving the source relative to the sample to create a velocity spectrum.

And the resulting spectrum is a fingerprint of the local chemical and magnetic structure.

What do the three main splitting patterns that we analyze?

The isomer shift, the quadrupole splitting, and the hyperfine field splitting actually tell a scientist.

Each one is a distinct piece of information.

The isomer shift tells us about the local chemical density specifically, the electron density crowding the nucleus.

A large shift means more electron density.

And the quadrupole splitting.

That occurs when the electric field around the nucleus is non -cubic or asymmetric.

It tells us about the symmetry of the local atomic arrangement.

So if the atoms were perfectly arranged in a highly symmetric cubic lattice, you wouldn't see quadrupole splitting.

But in an amorphous material, you'd expect a messy, broadened signal.

Precisely.

And finally, the hyperfine field splitting is perhaps the most famous application.

It directly measures the local magnetic environment, the effective magnetic field acting on the nucleus.

So this is critical for studying the magnetic properties of amorphous metals.

It is.

And also for tracking the extremely subtle initial stages of crystallization, where the environment of the iron atoms changes long before we can detect full LRO via X -ray diffraction.

Here's where it gets really interesting.

We started this deep dive exploring a realm defined by the absence of structure.

But what we found is that metallic disorder is highly structured at the atomic level, governed by statistical density and high stability.

The amorphous state is not random chaos.

It's a metastable ordered prison governed by kinetic frustration.

A great way to put it.

To summarize the essential knowledge gained from Chapter 21, let's just consolidate the four core concepts for you.

Let's do it.

One, structural definition.

Amorphous metals possess short -range order SRO -efficient local packing defined by coordination shells.

But they completely lack long -range order LRO, leading to those essential isotropic properties.

Okay, one.

Two, modeling and frustration.

Their structure is quantitatively described by the radial distribution function, the RDF, and modeled by the dense random packing of hard spheres, DRPHS, with that 63 .8 % efficiency.

And local stability is dominated by ISRO.

Yes, the non -crystallographic icosahedral short -range order, which geometrically frustrates crystallization.

Got it.

Three.

Three, synthesis criteria.

Excellent glass forming ability GFA requires high kinetic barriers, which we achieve through deep eutectic compositions.

Strong negative heats of mixing and significant atomic size differences, often greater than 12%.

And GFA is optimized when the reduced glass forming temperature TRG is high.

Exactly.

And finally,

four, structural proof.

We analyze these systems by linking reciprocal space data from techniques like x -ray scattering and element -specific EXAFS back to the real space atomic arrangements, the RDF, using the Fourier transform.

And we use nuclear techniques like Musbauer spectroscopy to probe local magnetic and chemical environments.

That's the whole picture.

That brings us to our final provocative thought for you to carry forward.

We know that icosahedral clusters are the most energetically favorable local arrangement for atoms, yet they fundamentally cannot tile 3D space.

When chemists design bulk amorphous alloys, BAAs, that are centimeters thick, how precisely do they engineer the remaining atomic structure, the matrix between these icosahedra, to balance that desperate local need for icosahedral stability with the global connectivity required for a solid piece of material?

Understanding how that global connectivity is maintained without sacrificing local preference.

Well, that is the cutting edge of amorphous materials research.

Thank you for joining us on this in -depth exploration of amorphous metals.

We hope this provided you with a thorough understanding of the material, enabling you to master this complex chapter.

We'll catch you next time for the next deep dive.