Chapter 1: Thermodynamics & Phase Diagrams

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.

Have you ever looked closely at an old bronze statue?

Maybe one that's been outside for centuries and then, you know, compared it to a cheap tool that snaps after just a few uses.

Why do some materials seem to maintain their structure, their strength,

seemingly forever, while others rapidly degrade or change shape just because you've heated them up?

Well, it all comes down to control.

Control over the material's inner life, specifically it's about control over the microscopic changes happening between the atoms.

If you want to design an alloy that performs reliably, and I mean whether it's a turbine blade or the casing of a battery, you have to understand first and foremost the direction that nature is pushing that material.

The direction.

So before you can even think about how fast something changes.

Exactly.

Before you punch the controls for speed or rate of change, you need to know the destination.

And that direction,

that destination, that's defined by thermodynamics.

That's the one.

And today, we are giving you the foundational source code for material science.

This deep dive is, well, it's essentially chapter one of the Expert's Guide to Material Stability.

Phase transformations in metals and alloys.

Our mission is crystal clear.

We are unpacking the set of thermodynamic concepts that allow us to unequivocally predict the equilibrium state of an alloy at any given temperature, any given pressure.

This is the tool.

This is the fundamental physics.

It's what helps us understand why phase diagrams look the way they do and really how to read the hidden language of material stability.

This feels like a real prerequisite then.

We're going to define stability using what has to be the most powerful concept in the field.

Yeah.

Gibbs free energy.

Yep.

We'll show you geometrically how temperature and pressure force atoms to rearrange themselves and we'll break down the logic behind these complex binary and ternary phase diagrams using visual analogies.

If you want a complete shortcut to an expert level understanding of material behavior, this is where you have to start.

But we absolutely must start with the essential caveat, which defines the boundary of our discussion today.

And what's that?

Thermodynamics tells us if a change is possible.

It tells us whether the initial state is unstable relative to the final state.

It gives you the direction and the potential.

But it doesn't tell you everything.

It does not tell us how fast that transformation happens.

That crucial variable, the rate that belongs to the realm of kinetics, we'll touch upon it briefly at the end, but the deep exploration of that, that waits for our next discussion.

Okay, let's begin.

Let's establish the common language we need.

When a material scientist discusses change, they start with, what, three fundamental definitions.

First,

the system.

The system is just the chunk of material we are studying.

It's the alloy, the compound, the specific piece of metal you've isolated.

Simple enough.

And inside that system...

Inside that system, we have the components.

These are the chemically distinct elements or compounds that make it up.

Think iron and carbon or copper and zinc.

We measure the composition of the system based on the relative amounts of these components.

Okay, so system components.

And the third one is the most critical, the phase.

A phase is a portion of the system that is physically distinct and is completely homogenous.

And that's crucial.

It must be homogenous in both its properties and its composition.

Give me an example.

Think of ice cubes in a glass of water.

Right.

Same component, HO, but you have two distinct phases,

solid ice and liquid water.

They have different properties, different structures.

Or think of solid sugar at the bottom of a glass of tea.

Before it dissolves, you've got two phases, solid sugar, liquid tea.

And once it dissolves completely...

You've just got one homogenous phase,

sweet tea.

Exactly.

So if the system, the components, and the phases are all defined, how does the system decide which phases should even exist?

How does nature determine stability?

For almost every transformation that matters in engineering, I mean, those happening at constant temperature, T, and constant pressure, the ultimate measure of stability is the Gibbs free energy.

We just call it G.

The famous Gibbs free energy equation.

I remember this one.

Equation 1 .1 is, doula dollar is HTSV.

It's deceptively simple, but every material decision we discuss today is going to flow from it.

It really does.

So let's break down the components of that equation, starting with H.

H stands for enthalpy.

It is, generally speaking, a measure of the heat content of the system.

Formally, H is defined in equation 1 .2 as dolly H plus PV or.

E is the internal energy, the total kinetic and potential energy of the atoms.

And the PV term.

For condensed phases, so solids and liquids, that pressure volume term, PV, it's negligible.

It's tiny compared to the internal energy.

So for most practical applications in metals and alloys, we can basically say H is the same as E.

That's right.

For a gut check, just think of H as the energy locked up in the atomic bonds and the vibrations A transformation that forms stronger, more stable bonds will result in a lower enthalpy.

So if the system only cared about enthalpy, it would just seek out the strongest possible bonds, the most efficient packing?

It would.

But it can't ignore the other term.

The minus DS term.

What is S?

S is entropy.

It's the measure of randomness, of disorder in the system.

The drive toward chaos.

That's the one.

The higher the entropy, the greater the number of ways the atoms can be arranged or the thermal energy can be distributed.

It's why liquids inherently have much, much higher entropy than solids.

The atoms just have far more freedom of movement.

Which brings us back to the central concept of the whole chapter, then.

Stability is the state that achieves the lowest possible G.

This means the system has to find the optimal trade -off between these two competing drives.

It is a fundamental compromise.

On one side, you have the drive for low enthalpy H, strong ordered bonds, low internal energy.

But on the other side?

The drive for high entropy S, maximum disorder, maximum randomness.

And temperature T is the umpire that decides which side wins that compromise.

Precisely.

At low temperatures, the T term is small.

This means the whole minus TS term is small, and so low enthalpy dominates.

And that's why solids are stable when it's cold.

That's why solid ordered phases, where atoms sit in fixed, strongly bonded locations, are the stable states at low T.

But crank up the temperature.

And T becomes a huge factor.

It makes that negative minus TS term very large and very negative.

Now, the drive for high entropy takes over.

Which is why things melt or boil.

Exactly.

The liquid and gas phases offer so much more atomic freedom, they become the stable state because that entropic gain just overwhelms the energetic cost of breaking all those bonds.

Now that we know what drives stability, we need to understand that stability isn't just one single condition.

Section 1 .3, which is illustrated conceptually in Figure 1 .1, it forces us to distinguish between two types of equilibrium.

And this distinction is crucial for understanding why we can use the materials we do.

First, you have stable equilibrium, that's configuration A in the figure.

It is the absolute lowest possible free energy state.

The system is completely settled, it has zero desire to change.

But then we often encounter metastable equilibrium, which is configuration B.

Metastable equilibrium is a local minimum in G.

The system appears stable against small fluctuations, but it's not at the global minimum.

It's trapped, you could say.

Trapped by what?

By an energy barrier.

The hump, shown separating B from A in Figure 1 .1.

It's stuck in a little valley, but there's a much deeper valley it could get to if only it had enough energy to get over the hill.

Let's talk about the most famous example of this because it's so powerful.

Carbon.

Right.

At standard room temperature and atmospheric pressure.

The lowest possible free energy state for carbon.

The stable equilibrium, configuration A, is graphite.

A pencil lead.

A pencil lead.

Diamond, which we value so highly, is actually stuck in a metastable state.

It's in configuration B.

So the transformation from diamond to graphite is, it's actually thermodynamically possible.

It is.

The change in free energy, delta G, is negative.

So thermodynamically, every diamond on earth is just a clock ticking down until it eventually turns into graphite.

That is the physics of it, yes.

However, that energy barrier, that kinetic hump between the diamond state and the graphite state is immense.

Ah, so it's a huge hill to climb.

A massive one.

This barrier ensures that the rate of transformation is so infinitesimally slow that for all human purposes we perceive the diamond as permanent.

And that's the difference.

Thermodynamics tells you the direction, the possibility.

Kinetics tells you the rate.

That distinction brings us to the formal transformation criterion, equation 1 .4.

A transformation is only possible if the change in Gibbs free energy is negative.

So delta G, which is G final minus G initial, has to be less than zero.

And the key limitation is exactly that.

Thermodynamics determines the possibility based on the sign of delta G.

But kinetics, the size of that energy barrier in figure 1 .1, determines the rate.

For engineering purposes, a transformation that's thermodynamically possible but kinetically blocked, that's basically the same as being stable.

It is.

This interplays why material science is so complex and so interesting.

Okay, finally, just a quick note on terminology before we move on.

We need to categorize thermodynamic properties.

Right.

We have intensive properties, which are independent of the system's size.

Think temperature, T, and pressure, P.

And then the others.

Then we have extensive properties like energy E, enthalpy H, entropy S, and of course Gibbs free energy G.

These are proportional to the amount of material you have.

So when we use these in calculations, we almost always normalize them.

We do.

We express them as molar quantities energy per mole to make sure we're always comparing apples to apples.

Okay, we've got the language down.

Let's move on to the simplest case.

We begin our quantitative analysis with the simplest system.

A single component, like a pure element, studied at a fixed pressure.

To predict its phase changes, we have to track exactly how its Gibbs free energy, G, changes as a function of temperature, T.

And the fundamental link for tracking H and S, that comes from something we can actually measure in a lab, right?

Cp2, the specific heat.

Yes.

Cp2 is what we measure.

As you can see in figure 192A, Cp2 is just the heat required to raise the temperature of a specific amount of the substance by one Kelvin.

That's equation 1 .5.

And through integration, we can then build the H and S curves from that measured Cp2.

Exactly.

If you plot enthalpy H versus temperature T, the slope of that curve at any given point is just Cpw $.

That's figure 1 .2b.

You're simply summing up all the heat you've added to raise the temperature from absolute zero.

And for entropy S, it's a bit different.

For entropy, the relationship is slightly different, yes.

The variation of S with T is the integral of Cptd divided by T.

So the integral of Ctt do as in figure 1 .2c.

Why the division by T?

You need to account for the fact that adding a bit of heat at a very low temperature has a much greater impact on disorder than adding that same amount of heat at a very high temperature.

It's about the relative change.

That makes sense.

This careful tracking of H and S then leads us to what has to be the single most critical relationship for phase stability, how G changes with T.

It is.

The core thermodynamic relationship equation 1 .9 is 0G8L is SdT plus VdP2.

But since we're holding pressure constant for now, that simplifies really elegantly, doesn't it?

It does.

It simplifies to equation 1 .10.

The slope of the G versus T curve, dEvrT T2, is equal to minus S.

That's the mathematical connection that ties everything together.

Wait, let me just make sure I appreciate the elegance of that.

Since entropy S is always a positive value.

Always.

Then dGGTT2 must always be negative.

Absolutely right.

It means G must decrease as temperature increases.

This simple fact explains why materials are generally less stable and transition to new phases as they heat up.

The system is always seeking a lower G.

And it also explains perfectly why the liquid phase eventually takes over from the solid phase.

Yes.

Because we know the liquid has a higher entropy than the solid, right?

Solid is greater than a Jol.

More disordered.

Much more.

So the magnitude of the slope for the liquid's G curve, which is minus the liquid, has to be greater.

Therefore, the liquid's Gibbs free energy curve must drop faster with increasing temperature than the solid's G curve.

And figure 1 .4 illustrates this competition perfectly.

At low temperatures, the solid G curve is lowest, so it's stable.

But because the liquid G curve is dropping faster, the two curves absolutely have to intersect at some point.

And that intersection point is the very definition of equilibrium.

It's T dollars the equilibrium melting temperature, where the Gibbs free energy of the solid equals the Gibbs free energy of the liquid.

Above T dollars, the liquid phase achieves the lowest free energy and becomes the stable one.

The figure also highlights the idea of latent heat.

When you supply heat right at T dollars, the temperature doesn't immediately go up.

That's right.

The heat goes into changing the phase.

It's used for breaking the solid bonds and increasing the enthalpy by an amount delta H.

So what if a pure metal has allotropes, you know, different crystal structures,

like iron?

We just plot a separate G curve for each structure, take a pure iron.

Room temperature, it's BCC, body -centered cubic, which we call ferrite.

Then at 910 degrees Celsius, it transitions to FCC, phase -centered cubic, which we call austenite.

And that 910 degrees Celsius is precisely where the G curve for the BCC structure intersects the G curve for the FCC structure.

It's just another purely energetic and entropic compromise.

That's all it is.

Now moving to section 2 .2, let's explore the effect of pressure.

Remember the full equation,

dLdG OHA SdT plus VdPt.

Okay, so this time we hold T constant.

And we get dGdP equals V volume.

That's equation 1 .11.

The physics here feels intuitive, but it's crucial.

Applying pressure is going to favor the phase that takes up less space, the one with the smaller molar volume.

The system tries to accommodate the external pressure by becoming denser.

It's as simple as that.

And to quantify this, we use the Clausius -Claiboron equation.

Equation 1 .14.

It tells us how the equilibrium temperature, dT, shifts when we change the pressure, dP.

The relationship is dPdT equals delta H, T delta V.

Let's use that iron example again, shown in figure 1 .5.

We're looking at the alpha BCC to gamma FCC austenite transformation.

Right, so BCC is less closely packed than FCC, so when you transform from BCC alpha to FCC gamma, the volume change, delta V, is negative.

It gets smaller.

And since gamma forms at a higher temperature, the process needs heat, so delta H is positive.

A positive delta H divided by a negative delta V means the slope, dPdT, has to be negative.

So increasing the pressure of positive dP causes a decrease in the transition temperature and negative dT.

Exactly.

It expands the stability range of gamma phase, the denser phase.

It's consistent with the principle that pressure favors smaller volumes, and you see this continue with epsilon phi, the densest phase shown in figure 1 .5, which only becomes stable at incredibly high pressures.

Okay, finally, in this section, section 2 .3 looks at the driving force for solidification that seems important for kinetics later on.

It's very important.

When a liquid is cooled below its melting point, $2 we call that under -cooling by an amount delta T2, we want to know the magnitude of the free energy difference, delta G, that's pushing the liquid to transform into a solid.

Figure 1 .6 shows this driving force visually.

It's just the vertical distance between the liquid and solid G curves at that specific temperature T.

And since we know delta G is zero right at T dollars, we can write delta HTM, delta HESO.

This lets us define the entropy of fusion, delta H fusion Tm inside A.

Which led to Richard's rule, this observation that for most metals, the entropy of fusion is a nearly constant value.

Roughly equal to the gas constant R.

And that constancy allows for a really powerful and practical approximation,

especially for engineers, as long as the under -cooling, delta T, is pretty small.

That's the approximation.

By assuming the difference in specific heats between the liquid and solid is negligible, we get equation 1 .17.

Delta G is approximately equal to delta H fusion times delta T Paul.

So the driving force for solidification is just linearly proportional to how far we cool the liquid below its melting point.

The greater the under -cooling, the harder the system tries to solidify.

This linear approximation is absolutely essential for calculating nucleation rates the very beginning of the kinetic process, which we will cover in great depth in future material.

Okay, we've established the rules for pure elements.

Now we jump into the real complexity of alloys by adding composition as a variable.

We're focusing on a binary solution, so components A and B.

Right.

And the Gibbs free energy of the solution now depends not just on T and P, but also on the concentration, X.

Section 3 .1 describes this two -step thought process for calculating that G cell.

I'm looking at figures 1 .7 and 1 .8 here.

Step one.

Start with sexy moles of pure A and x -bally moles of pure B, but keep them separate.

Their combined free energy, which we call GD1, is just the simple weighted average.

G doll equals XAGA plus XBGB1.

And on the G versus X diagram, where the X axis is composition, that's just a straight line connecting the G values of pure A and pure B.

Exactly.

Now, step two.

Mix them together homogeneously to form one mole of solution.

The final free energy is that initial G doll, and plus the free energy change that results from the act of mixing itself.

We call that delta G mix.

So G doll one dollar plus delta G mix, that's equation one dollar 20.

And that delta G mix, of course, is that same compromise we talked about before.

Delta G mix, delta H mix, T delta ses.

Equation one dollar and 21.

So we have the heat of mixing, delta H mix, and the entropy of mixing, delta S mix.

Exactly.

Did the barns get stronger, making it exothermic with a negative delta H?

Or did they get weaker, making it endothermic with a positive delta H?

And then you have the change in entropy, which is the difference between that chaotic mixed state and the highly ordered unmixed state.

Let's start with the theoretical,

almost beautiful world of ideal solutions.

That's section 3 .2.

An ideal solution makes a big assumption.

It assumes that delta H of mixing is zero.

What does that mean physically?

It means the atoms of A and B don't particularly care whether they are neighbors or not.

The energy of an A -B bond is exactly the average of an A and a B -B bond.

There's no energetic tenality or benefit to mixing.

So in that case, mixing is driven only by entropy.

Only by entropy.

The dominant term here is the configurational entropy, delta S mix.

And that's just the entropy that comes from all the different ways you can arrange the A and B atoms on the lattice site.

That's it.

Since the unmixed state has zero configurational entropy, delta S mix is always positive.

Mixing always increases the overall disorder.

And the formula for that, equation 1 .25,

is delta S mix, thatch A, ln and xA plus xB ln and xB A would.

And since xB O and xB O on fractions less than one, their natural logarithms are negative, which makes that whole term positive.

Which gives us the ideal delta G mix curve shown in figure 1 .9.

It's parabolic and it's always negative.

Meaning mixing is always thermodynamically favorable in an ideal solution.

And it's always a good idea.

So when we combine this parabolic delta G mix curve with that straight G to one line from before,

we get the total G curve in figure 1 .10.

And I noticed something.

The G curve is deeply curved.

And it approaches the pure component axis almost vertically.

That deep curvature is the fingerprint of that entropic term.

It ensures that mixing even a tiny, tiny bit of component B into pure A or vice versa is always favorable.

The entropic benefit is so large that the G curve has to shoot downward near the pure components.

And that guarantees that solubility is never zero in this model.

Never zero.

The shape of this curve is absolutely key to understanding the next concept.

Which is section 3 .3, chemical potential, or MU.

I still like thinking of it as the desire of a component to leave the phase.

It's an excellent intuitive definition.

Formally, MU A is the instantaneous change in total free energy when you add one infinitesimal amount of A to the system.

It's an equation 1 .29.

It tells us the energetic cost or benefit of adding that component at a specific concentration.

And the most important takeaway for phase diagrams is the geometric construction.

That's figure 1 .11.

Absolutely.

If you pick any composition X on the G curve and you just draw a tangent line to that curve, the points where that tangent line intersects the pure component axis, the sex A1 and sex B1 and y axis, those points define the chemical potentials, MU A and MU B.

So if the tangent is steeper near component A, that component has a lower chemical potential.

It's more stable, less likely to leave.

Precisely.

And for an ideal solution, that geometric construction results in a very simple formula.

MU A plus RT LN XAO.

That free LN XA term is simply the free energy reduction that you get due to the entropy of mixing.

Okay.

So we're leaving the ideal solution and stepping into the real world now with regular solutions in section 3 .4.

Now we have to incorporate the heat of mixing, delta H mix.

Right.

And we model this by assuming delta H mix is governed by the energy difference between the AB bonds and the average of the AA and BB bonds.

But we're still making a big simplification.

We are.

We stick with the major simplification that the atoms are still randomly arranged.

And that is the definition of the regular solution approximation.

And that gives us a parabolic heat of mixing.

Delta H mix omega XAXBC equation 1 .36.

Right.

And omega, that's the effective bonding interaction term.

This term is vital because its sign, positive or negative, dictates the material's entire behavior.

Okay.

Let's take case one.

Omega is less than zero.

If omega is less than zero, delta H of mixing is negative.

It's exothermic.

This means atoms prefer unlike neighbors.

The AB bonds are stronger than the average.

So there's an attraction.

It's a strong attraction.

This force lowers the enthalpy, making mixing favorable at all temperatures.

You can see this in figures 1 .15A and 1 .15B.

This system is highly stable and it often leads to ordered structures.

Okay, now case two, omega is greater than zero.

Now, delta H of mixing is positive.

It's endothermic.

The atoms dislike each other.

A prefers A neighbors, B prefers B neighbors.

There's a repulsion that raises the enthalpy.

And here, temperature is the deciding factor.

It is.

At a high temperature, like in figure 1 .15C, the entropic drive, that large negative minus T delta S mixed term, it just overwhelms the positive delta H of mixing and the solution remains stable.

It stays mixed.

But at a low temperature.

At a low T, like in figure 1 .15B, that positive delta H mix dominates.

It causes the G curve to develop that telltale negative curvature in the middle.

And that negative curvature, that's the thermodynamic signature of instability, isn't it?

It is.

The system has realized that by separating into two distinct phases, it could actually lower its overall free energy.

It's unstable as a single phase.

So the total G curve in figure 1 .15C is the sum of all these different parts.

It is.

And the resulting chemical potential equation, equation 1 .40, are now more complex.

They combine that entropy term with the parabolic omega term.

To simplify the mathematics of these complex chemical potentials, section 3 .5 introduces activity, which we denote with an A.

Activity is just a normalized measure of chemical potential.

We define it so that would A equal GA plus RT LNAAR.

That's equation 1 .41.

You can see it's the same form as the ideal solution equation, but with activity A instead of mole fraction X.

So if a solution's ideal, activity A equals the mole fraction X.

Exactly.

But when solutions are non -ideal, we use something called the activity coefficient, gamma, where gamma A equals AA divided by XA.

Okay, let's look at figure 1 .1C.

Line 1 is the ideal case.

What about line 2?

If the heat of mixing is negative, so there's attraction between the atoms, line 2 shows the component is less active than ideal.

Why less active?

Because that strong attraction is holding the atoms firmly in the solution.

It reduces their tendency, their activity, to leave.

And conversely, if the heat of mixing is positive, so there's repulsion.

That's line 3.

The component is more active.

The atoms dislike their neighbors, so they have a higher tendency to escape the solution.

The physical meaning remains consistent,

then.

Activity measures the effect of concentration, or the willingness of an atom to transition to another phase, whether it's evaporating or moving into a precipitate.

Exactly.

And in very dilute solutions, these concepts simplify to Henry's Law for the solute and Rauwut's Law for the solvent.

So now we move to real solutions in section 3 .6, acknowledging that the regular solution model is still often too simple.

Why?

Because atoms won't always arrange randomly.

They won't.

They will actively organize themselves to minimize G.

So if there's a strong attraction, a negative in -atomic energy.

That drives ordering.

Atoms prefer unlike neighbors, as you see in figure 1 .1 AA.

And if there's repulsion?

That drives clustering.

Atoms seek out like neighbors, like in figure 1 .1 B.

Furthermore, the model ignores size effects.

What happens when atoms A and B are very different sizes?

That size difference introduces elastic strain energy into the system, which raises the delta H of mixing.

The system might accommodate this by forming interstitial solutions, like in figure 1 .0 OCC, or by forming entirely new crystal structures.

Within that idea of ordering, from section 3 .7, we distinguish between short -range order, SRO, and long -range order, LRO.

Right.

In SRO, atoms locally prefer unlike neighbors, but the overall structure is still random.

That's figure 1 .19.

LRO is different.

In LRO, the ordering becomes systemic.

It forms a superlattice, which is a new ordered phase.

The QO system is the textbook example, shown in figure 1 .20.

At high temperatures, it's just a random FCC lattice.

And at low temperatures.

Below a critical temperature, the atoms snap into ordered alternate layers, forming structures like QO.

And I'm guessing that as temperature increases, the drive for randomness for entropy eventually destroys that long -range order.

It does.

Above a critical temperature, T -critical, the LRO is gone.

And as you can see in figure 1 .21, that critical temperature peaks at the ideal stoichiometric compositions, which reflects where the bond strength, that omega term, is at its maximum.

Okay, finally, we have to consider intermediate phases.

This is section 3 .8.

These are phases that don't resemble the crystal structure of either pure component A or B.

They exist because, at specific compositions, they achieve the absolute lowest free energy.

You can see one represented in figure 1 .23.

These can be intermetallic compounds with very narrow stoichiometric stability ranges.

Like AB or AB.

Or they can tolerate wider composition fluctuations.

So what dictates the structure they adopt?

It's driven by three factors, moving beyond our simple thermodynamic models.

Factor 1 is atomic size.

If the sizes differ significantly, say by a ratio of 1 .1 to 1 .6, the system prioritizes filling space efficiently.

This leads to complex structures like lave's phases, like midjihiro, which you can see in figure 1 .24, or interstitial compounds, where smaller atoms just occupy the gaps between larger ones.

What's factor 2?

Valency.

This governs so -called electron phases, where stability is based on maintaining a critical electron -to -atom ratio.

Classical examples are the alpha and beta brasses.

And the third factor.

Electronegativity.

Large differences in electronegativity drive strong charge transfer, which leads to ionic bonds and stable stoichiometric compounds like Mg or Ocanis.

Okay, so we've covered single phases.

Now we apply these G versus X curve concepts to the real world, where we almost always have multiple phases coexisting a heterogeneous system.

Right.

We need to find the lowest possible free energy state when the material can choose between different crystal structures.

To begin, we have to plot the G curves for different structures.

Let's say FCC alpha and BCC beta on the same diagram.

But there's a problem, right?

Pure component B might only exist stably as BCC.

That's right.

As figure 1 .25 illustrates, we have to calculate the necessary fictional free energy cost to force pure B into the unstable FCC alpha structure.

That's the distance BC in the figure.

So we add that cost to GB before we incorporate the energy of mixing for the alpha phase.

And that allows us to draw realistic intersecting G curves for both the alpha and beta phases on the same plot.

Once we have two intersecting G curves, we need to understand the molar free energy of phase mixtures.

This is figure 1 .26.

If an alloy X is a mixture of alpha and beta phases, its total molar free energy, G, is not found on the G curves themselves.

No, it's not.

The total G of the mixture actually lies on the straight line that connects the G values of the two phases.

This is the geometric proof of the liberal.

So by physically separating into two phases, the system achieves a lower free energy than if it were forced into one single homogenous phase.

Exactly.

The distance B in figure 1 .26 is the G of that two phase mixture.

This leads us to what has to be the single most important geometrical construction in this whole field, the common tangent rule.

That's figure 1 .27.

When the alpha and beta G curves cross, a single homogenous phase in that middle region is unstable.

The system will minimize its free energy by separating into two distinct phases.

And how does it find the lowest energy?

You draw a straight line that is simultaneously tangent to the alpha curve at composition alpha E and tangent to the beta curve at composition beta E.

Ah, so it's like a tightrope stretched between the two curves.

A perfect analogy.

The total free energy, G E, which is represented by this tangent line segment, is the absolute lowest energy the system can possibly achieve.

So the system spontaneously changes its composition, it unmixes, until it finds this lowest energy envelope.

And the compositions alpha E and beta E are the equilibrium compositions.

This common tangent construction is the geometric definition of heterogeneous equilibrium.

More profoundly, it dictates the formal equilibrium condition, which is equation 1 .46.

Which says what?

It says that the slope of the G curve at alpha E must be identical to the slope of the G curve at beta E.

And since the slope is defined by the chemical potential.

Means the chemical potential, mu, of each component must be identical in both phases.

Mu A in alpha equals mu A in beta, and mu B in alpha equals mu B in beta.

And if, for just a moment, the chemical potential of component A was higher in alpha than in data, what would happen?

A atoms would spontaneously leave alpha and migrate to beta,

continuing the transformation until those potentials equalize.

Equality of chemical potential is the material's definition of true peace.

It's done changing.

Section 4 .4 states this using that alternative metric activity.

The activity condition, equation 1 .47, is functionally identical, right?

Identical.

The activity A of each component must be equal in both phases at equilibrium.

And figure 1 .28 visualizes this beautifully.

In that two -phase region defined by the common tangent, the activities of both components are just flat, horizontal lines.

They're constant across the entire two -phase field, which confirms the system is in equilibrium.

If the line wasn't flat, it would mean mu was unequal, and atoms would be on the move.

So now we can take this entire mechanism, the common tangent construction on the G -X diagram,

and use it to build the map of material behavior.

The temperature composition, or T -X phase diagram.

This is where it all comes together.

Section 5 .1 shows the procedure, in figure 1 .29, for a simple system where the solid and liquid phases are completely miscible.

We're just tracking the curves as we change temperature.

Right.

At a very high temperature, T, the G -curve for the liquid, gl, is lowest everywhere.

That's figure 1 .29a.

The entire system is liquid.

Okay.

Now we drop to an intermediate temperature, P.

The curves cross.

And because the entropy of the liquid is higher,

the liquid G -curve has dropped faster than the solid G -curve.

Now we apply the common tangent, as in figure 1 .29c.

The tangent points define the equilibrium compositions.

Point B is the solidest, the solid composition, and point C is the liquidest, the liquid composition.

And we just plot those two points on our T -X diagram at that temperature, T.

The region between B and C is the two -phase region, solid plus liquid.

We just repeat this for every temperature until we hit T, where the solid G -curve is lowest everywhere, and the system is 100 % solid.

So plotting all those tangent points, B and C, across the entire temperature range, that literally derives the smooth, liquidest, and solidest lines that define the phase diagram.

It proves that the complex phase diagram is simply a thermodynamic consequence of the relative stabilities of the two phases.

It's not magic.

Okay, let's look at a system with a miscibility gap.

That's section 5 .2, figure 1 .30.

This happens when we have a positive heat of mixing for the solid phase.

Meaning the A and B atoms repel each other.

And that repulsion leads to that negative curvature in the solid G -curve at low temperatures, which makes the single homogenous phase unstable.

Right.

The system minimizes its free energy by separating into two solid phases, alpha prime, which is A -rich, and alpha double prime, which is B -rich, and they're connected by a common tangent.

And as you increase the temperature.

The entropic term starts to fight that repulsion.

It flattens the G -curve, and the compositions of alpha prime and alpha double prime move closer together.

The result on the T -X diagram in figure 1 .30 is a miscibility gap, a dome -shaped two -phase region where two solid solutions coexist.

And this is also why these systems typically exhibit a minimum melting point.

Melting helps relieve that atomic repulsion, which makes the liquid phase relatively more stable.

Conversely, systems with strong attraction, a negative omega,

they often exhibit a maximum melting point in the phase diagram, right?

They do, corresponding to the composition with the highest attraction.

This often coincides with an ordered phase composition, as you can see in figure 1 .31.

This attraction can lead to eutectic systems only if the components have very different crystal structures.

Okay, section 5 .4 tackles intermediate phases.

These just introduce entirely new G -curves onto the diagram.

A eutectic phase diagram that involves an in a metallic compound, like in figure 1 .34, is just the result of applying the common tangent rule between the liquid curve and all the solid curves, alpha, beta, and the intermediate phase at every single temperature.

This leads to one of the most important and maybe expert -level insights in the whole chapter.

It's highlighted conceptually in figure 1 .35.

A phase's stability range on the phase diagram is not simply dictated by where its own G -curve is lowest.

That's right.

That feels counterintuitive.

I would assume the phase is stable, where its energy is at a global minimum.

But look closely at figure 1 .35.

The actual stable composition range of a phase is determined by the common tangent to its neighboring phases.

If a mixture of alpha and beta can achieve a lower free energy than a homogenous gamma phase, even at gamma's optimal composition, then the gamma phase is excluded.

It can't exist.

So a phase is only stable over the range where it forms part of that minimum free energy envelope, the one created by the common tangent construction.

That distinction is incredibly important for predicting the limits of solubility.

And it leads us directly to the Gibbs phase rule, section 5 .5.

This is the unifying principle for all phase equilibrium.

And the core condition is still the same.

It is.

The chemical potential, mu, of each component must be identical in every phase present.

So the phase rule quantifies the degrees of freedom f a system has.

It's f a Cp plus 2.

Components minus phases plus 2 variables, temperature and pressure.

But we can simplify this for material science by assuming constant pressure.

So it becomes f equals Cp plus 1.

That's equation 1 .50.

OK.

Let's apply that to a binary system.

So C equals 2.

If you have one phase, p equals 1.

Then f equals 2 minus 1 plus 1, which is 2.

You have two degrees of freedom.

You can independently change temperature and composition.

Now if you have two phases in equilibrium, p equals 2, like solid and liquid.

Then f equals 2 minus 2 plus 1, which is 1.

You have one degree of freedom.

If you fix the temperature, you've used your one degree of freedom.

The compositions of the two phases are now automatically fixed by the common tangent rule.

You have zero freedom left.

And if you have three phases, p equals 3, like at a eutectic point.

Then f equals 2 minus 3 plus 1, which is 0.

This is an invariant point.

The temperature and all three compositions are fixed by nature.

They cannot be changed without one of the phases disappearing.

OK.

Section 5 .6 looks at the effect of temperature on solid solubility.

I'm looking at figure 1 .36.

We want to know why a solvent, A, can only dissolve a maximum amount of solute B.

That maximum solubility, which we call XB alpha, is set by the common tangent between the alpha G curve and the adjacent phase's G curve, which is beta in this case.

This boundary condition allows us to derive the relationship in equation 1 .53.

The solubility, XB alpha, is approximately A times the exponential of minus Q over RT.

And Q is the enthalpy change required when a mole of B dissolves in A.

Right.

And since Q is usually positive dissolving, B usually requires some energy input.

The solubility increases rapidly with temperature.

But the key insight here goes back to G equals H minus TS.

It does.

Even if Q, the enthalpy cost H, is large and positive, the entropic gain, S, for mixing will always allow some amount of mixing to occur, as long as T is above absolute zero.

So solubility can never actually reach zero.

It can't.

If it did, the system would be violating the laws of thermodynamics.

The final concept in this section, 5 .7, is truly profound.

Thermodynamics even governs crystalline imperfections, like the equilibrium vacancy concentration.

A truly perfect crystal is in fact unstable.

That's right.

To create a vacancy, an empty lattice site, you have to break atomic bonds.

That costs enthalpy, so delta HV is positive.

But introducing those empty sites dramatically increases the configurational entropy, delta because you now have many, many more ways to arrange the atoms and the holes.

So the material seeks a compromise.

It creates just enough vacancies to maximize that entropy gain without paying too high a price in enthalpy.

Exactly.

The equilibrium concentration, XVE, is reached when G is minimized.

You can see this in figure 1 .37.

The resulting formula, equation 1 .57, is another exponential function of temperature.

XVE is approximately the exponential of minus delta GV over RT.

So since the free energy to create a vacancy is positive, the concentration increases exponentially with temperature.

It does.

It can reach significant levels, up to a tenth of a percent, right at the melting point.

Defects are thermodynamically inevitable.

Okay, we've discussed so much, but it's all assumed we're dealing with bulk, infinitely large phases.

A big assumption.

In reality, phase transformations involve interfaces.

And when those interfaces are curved, like the ones surrounding small precipitates, they introduce an extra energetic cost that fundamentally changes the equilibrium condition.

This is the capillarity, or Gibbs -Thompson effect, section 6 .1.

A curved interface surrounding a small particle of phase beta exerts an extra internal pressure, delta P, on that phase.

It's analogous to the surface tension of a soap bubble.

So for a spherical particle of radius R.

The excess pressure is approximately 2 times gamma over R.

That's equation 1 .58.

Gamma is the interfacial energy.

The smaller the particle, the greater the pressure inside it.

And since we know DG equals VDP, this extra pressure raises the total free energy of that small beta particle.

It does, by an amount delta G equals delta P times the molar volume.

If we plot this on the GX diagram, as in figure 1 .38b, the G curve for the small particle G beta shifts upward compared to the G curve for a bulk flat interface.

And this upward shift of the G curve has a massive practical consequence on solubility in particle size.

If we apply the common tangent rule to that now raised G beta curve,

the tangent point shifts.

It does.

And that shift means that the solubility of solute B in the matrix, alpha, is greater when the matrix is in equilibrium with small curved beta particles than it is with a big flat beta phase.

This is the thermodynamic proof that small particles are inherently less stable than large particles.

They require a higher concentration of solute in the surrounding matrix just to prevent them from dissolving.

And the size dependence in equation 1 .62 shows this can be a big effect.

It can.

For very small nanometer size precipitates, the equilibrium solubility can be 5 or 10 % higher.

This is the root cause of what engineers call Oswald ripening or coarsening, right?

Exactly.

This thermodynamic instability drives coarsening.

When you have a mixture of small and large precipitates, the small ones have a higher free energy and a higher equilibrium solubility around them.

So the matrix is locally supersaturated around the larger, more stable particles.

Right.

So solute atoms will diffuse away from the shrinking small particles and deposit onto the growing large particles.

The system is just trying to minimize its total surface area and its total free energy.

This effect is why fine microstructures are always trying to coarsen over time, especially at high temperatures.

Okay, now we take the final step up in complexity.

Three components.

A, B, and C.

This is essential because most commercial alloys are ternary or even higher.

Right.

And we can no longer use a simple 2D TX plot.

We need the Gibbs Triangle, shown in figure 1 .4, to represent composition.

The corners are 100 % pure A, B, and C.

Any point inside is a mixture.

And to represent free energy, we have to go to 3D.

We use a free energy surface, which is plotted vertically over the Gibbs Triangle, as in figure 1 .41A.

So when two phases, say liquid and solid, are in equilibrium,

what's the new rule?

The equilibrium condition requires a common tangential plane to touch both G surfaces simultaneously.

That's figure 1 .41B.

And in a 2D isothermal section,

a slice of the phase diagram at a fixed temperature,

the points where that plane touches the G surfaces,

those define the equilibrium compositions.

And they're connected by a tie line.

As that common tangential plane rolls across the two G surfaces, it generates the two -phase region on the diagram.

What happens when we have three -phase equilibrium?

That's figure 1 .42.

A tie triangle appears.

This happens when the common tangential plane touches the G surfaces of three phases simultaneously, like liquid plus alpha plus beta.

Let's check the Gibbs phase rule again.

F equals C, P plus 1.

So C is 3, P is 3.

So F equals 3 minus 3 plus 1, which is 1.

We have one degree of freedom, but since we fixed the temperature to get this isothermal section, we've used it.

So the compositions at the corners of that triangle are fixed.

Zero degrees of freedom left.

Right.

And dropping the temperature further, we might hit the ternary eutectic point.

E, where four phases are in equilibrium.

Now P is 4, so F is 0.

It's an invariant point where t and all compositions are fixed by nature.

So section 7 .4 outlines the solidification path.

We track this using a liquid as surface projection, like in figure 1 .43.

This is like a map showing only the temperatures at which freezing begins.

Exactly.

For an alloy X, as you show in figure 1 .44, as it cools, it first solidifies the primary alpha phase.

The liquid is therefore depleted of component A.

Which causes the remaining liquid composition to move away from the A corner.

The liquid composition then hits a eutectic valley, EE, where two phases, alpha and beta, solidify together.

The liquid composition moves down this valley, cooling as it goes, until it finally reaches the ternary eutectic point, E.

And at that point, the remaining liquid transforms isothermally into three solid phases, alpha plus beta plus gamma.

Exactly.

Finally, we need to issue a crucial warning on vertical sections, figure 1 .45.

Engineers often use these slices of the ternary diagram because they look like familiar binary diagrams.

They are profoundly misleading when it comes to composition.

In a binary diagram, the horizontal line cutting through the two -phase region is the tie line.

It tells you the exact compositions of the two phases in equilibrium.

But in a ternary system?

In a ternary system, a vertical section only shows what phases are present at a given temperature along that specific slice.

The tie lines in the two -phase regions almost never lie along the vertical plane of that section.

So you cannot use a vertical section to determine the compositions of the phases in equilibrium.

You cannot.

Which is the whole point of a phase diagram.

For composition information in ternary systems, you have to consult the full isothermal sections or the liquid -to -surface projection.

Ignoring this warning leads to massive errors in materials processing.

OK.

We circle back to where we started.

That hump separating the metastable state from the stable state.

Thermodynamics defines the possibility, delta G, but kinetics defines the rate at which we overcome that barrier.

Transformation relies on thermal activation, which you can see in figure 1 .47.

To move from the initial state G to the final state G, an atom has to temporarily achieve a higher free energy, the activated state, which has an energy of delta GA.

And the energy for that jump is supplied by random thermal motion.

Exactly.

The probability of an atom achieving this necessary activation energy is the key to the rate of transformation.

And that probability is proportional to?

An exponential term.

It's proportional to E to the power of minus delta GA over kT.

This gives us the foundation for the Arrhenius rate equation, equation 1 .72.

The transformation rate is proportional to the exponential of minus delta GA over RT.

This shows that exponential relationship.

The rate of transformation is incredibly sensitive to two things, the size of that activation energy barrier, delta GA, and critically, temperature.

A slight raise in temperature can increase the transformation rate dramatically because so many more atoms now have the necessary thermal energy to jump that kinetic hump.

And this is the fundamental equation for understanding how fast processes like diffusion occur, which in turn governs almost all phase transformations.

It is.

If we can calculate delta GA, we can predict the speed of change.

Okay.

We have completed our foundational deep dive into the thermodynamics of phase transformations.

To recap the most essential takeaways, the five concepts you absolutely must carry forward.

First, Gibbs free energy, G equals H minus TS, is the ultimate dictator of stability.

It's always balancing the drive for low energy H and high disorder S.

The second.

Phase diagrams are fundamentally constructed by finding the common tangent between the G versus X curves of competing phases at various temperatures.

Third takeaway.

Equilibrium requires that the chemical potential, mu, or activity, A, of every single component must be identical in all coexisting phases.

No exceptions.

Fourth.

The Gibbs phase rule, F equals C minus P plus one, controls the freedom of the system.

It guarantees that three phases coexisting in a binary alloy always results in an invariant point.

And the fifth and final takeaway.

The presence of small curved interfaces raises the free energy of a phase.

This is the capillarity effect.

It raises its equilibrium solubility and it drives the coarsening mechanism that we know as Oswald Wipening.

We started by establishing that thermodynamics dictates that solubility is never truly zero and that diamond is constantly, if imperceptibly, decaying into graphite.

Our materials are never truly static.

They're just stuck in a messistible state until time and temperature conspire to help them overcome that kinetic barrier.

So think about the metals around you, the steel in your phone, the aluminum in a plane wing.

Which of those materials is performing reliably only because it is successfully trapped in a state that nature actively wants to destroy?

Thank you for diving deep into the source material on decoding phase transformation thermodynamics with us.

We hope this was the perfect foundational course you needed.

We'll see you next time for the deep dive into the physics of kinetics and diffusion.