Hello there! Welcome to our session on the thermodynamics of solutions. This is a fascinating and fundamental part of physical chemistry. We're moving beyond simple ideal gases to see what happens when we actually mix substances together. We'll start with the "perfect" scenario, the ideal solution, and then build up to the much more complex (and realistic) world of non-ideal solutions. Let's begin!
Tutor's Note: The goal here is to connect the macroscopic properties we can measure (like vapor pressure, boiling point, heat released) to the microscopic interactions between molecules. Thermodynamics gives us the exact mathematical language to do this.
An ideal solution is a theoretical benchmark, much like an ideal gas. It's a mixture where the intermolecular forces between unlike molecules (A-B) are exactly the same as the forces between like molecules (A-A and B-B). Think of mixing benzene and toluene—they are so similar chemically and structurally that they mix "perfectly."
The most precise way to define an ideal solution is through its chemical potential, $\mu_i$. For any component 'i' in an ideal solution, its chemical potential is given by:
$ \mu_i(P, T, x_i) = \mu_i^*(P, T) + RT \ln x_i $
Where:
If the vapor above the solution is ideal, this definition leads directly to Raoult's Law:
$ P_i = x_i P_i^* $
Where $P_i$ is the partial pressure of 'i' above the solution and $P_i^*$ is the vapor pressure of pure 'i'.
What happens when we mix two pure components (say, $n_A$ moles of A and $n_B$ moles of B) to form an ideal solution? We want to find $\Delta G_{mix}$, $\Delta S_{mix}$, $\Delta H_{mix}$, and $\Delta V_{mix}$.
1. Gibbs Free Energy of Mixing ($\Delta G_{mix}$)
The total Gibbs free energy before mixing is: $ G_{initial} = n_A \mu_A^* + n_B \mu_B^* $
The total Gibbs free energy after mixing is: $ G_{final} = n_A \mu_A + n_B \mu_B $
Using our definition $\mu_i = \mu_i^* + RT \ln x_i$, we get:
$$ G_{final} = n_A (\mu_A^* + RT \ln x_A) + n_B (\mu_B^* + RT \ln x_B) $$
$$ G_{final} = (n_A \mu_A^* + n_B \mu_B^*) + RT (n_A \ln x_A + n_B \ln x_B) $$
$$ G_{final} = G_{initial} + RT (n_A \ln x_A + n_B \ln x_B) $$
Therefore, $\Delta G_{mix} = G_{final} - G_{initial}$:
$$ \Delta G_{mix} = RT (n_A \ln x_A + n_B \ln x_B) $$
For 1 mole of total solution ($n_T = 1$, so $n_A = x_A$ and $n_B = x_B$), the molar Gibbs free energy of mixing is:
$$ \Delta G_{mix, m} = RT (x_A \ln x_A + x_B \ln x_B) $$
Since mole fractions are always less than 1, $\ln x_i$ is always negative. This means $\Delta G_{mix}$ is always negative. This proves that mixing of ideal solutions is a spontaneous process.
2. Entropy of Mixing ($\Delta S_{mix}$)
We use the fundamental relationship: $ S = -(\partial G / \partial T)_P $.
$$ \Delta S_{mix} = -\left(\frac{\partial \Delta G_{mix}}{\partial T}\right)_P = -\frac{\partial}{\partial T} [RT (n_A \ln x_A + n_B \ln x_B)] $$
Since $n_i$ and $x_i$ are independent of temperature:
$$ \Delta S_{mix} = -R (n_A \ln x_A + n_B \ln x_B) $$
For 1 mole of solution:
$$ \Delta S_{mix, m} = -R (x_A \ln x_A + x_B \ln x_B) $$
Since $\ln x_i$ is negative, $\Delta S_{mix}$ is always positive. This makes perfect sense! Mixing increases disorder. This positive $\Delta S_{mix}$ is the sole driving force for ideal mixing.
3. Enthalpy of Mixing ($\Delta H_{mix}$)
We use the Gibbs-Helmholtz equation: $\Delta G = \Delta H - T\Delta S$.
$$ \Delta H_{mix} = \Delta G_{mix} + T \Delta S_{mix} $$
$$ \Delta H_{mix} = [RT (n_A \ln x_A + n_B \ln x_B)] + T [-R (n_A \ln x_A + n_B \ln x_B)] $$
$$ \Delta H_{mix} = 0 $$
This is a key feature! No heat is absorbed or evolved when mixing ideal components. This is the direct thermodynamic consequence of our initial assumption: the A-B, A-A, and B-B interactions are identical. No bonds are strengthened or weakened, so no energy change occurs.
4. Volume of Mixing ($\Delta V_{mix}$)
We use the relationship: $ V = (\partial G / \partial P)_T $.
$$ \Delta V_{mix} = \left(\frac{\partial \Delta G_{mix}}{\partial P}\right)_T = \frac{\partial}{\partial P} [RT (n_A \ln x_A + n_B \ln x_B)] $$
Since R, T, n, and x are all independent of pressure (at least in this context):
$$ \Delta V_{mix} = 0 $$
This also makes sense. If the interactions are all the same, the molecules will pack together in the mixture just as they did when pure. 100 mL of A and 100 mL of B will give exactly 200 mL of solution.
Key Takeaway: For an Ideal Solution $\Delta G_{mix} < 0$ (spontaneous), $\Delta S_{mix} > 0$ (disorder-driven), and $\Delta H_{mix} = 0$, $\Delta V_{mix} = 0$.
In reality, A-B interactions are almost never the same as A-A and B-B interactions. This gives rise to non-ideal solutions, which are the norm. To handle this, we introduce two new concepts: activity and activity coefficient.
We loved our simple ideal equation: $\mu_i = \mu_i^* + RT \ln x_i$. It's so elegant! To keep this same form, we invent a new quantity called activity, $a_i$, which is like an "effective mole fraction."
The chemical potential for a component in a real solution is defined as:
$$ \mu_i = \mu_i^\circ + RT \ln a_i $$
Here, $\mu_i^\circ$ is the chemical potential in a standard state (which for a solvent is usually the pure liquid, $\mu_i^*$).
How does activity $a_i$ relate to the actual mole fraction $x_i$? We define the activity coefficient, $\gamma_i$ (gamma), as the "fudge factor" that connects them:
$$ a_i = \gamma_i x_i $$
Now we can write the chemical potential in its most useful form:
$$ \mu_i = \mu_i^\circ + RT \ln(\gamma_i x_i) = (\mu_i^\circ + RT \ln x_i) + RT \ln \gamma_i $$
This is brilliant! The first part, $(\mu_i^\circ + RT \ln x_i)$, is the chemical potential for an ideal solution. The second part, $RT \ln \gamma_i$, is the correction term that accounts for all the non-ideal behavior.
Tutor's Note: The activity coefficient $\gamma_i$ is our measure of non-ideality.
As $x_i \to 1$ (the component becomes pure), $\gamma_i \to 1$ and $a_i \to x_i$. (This is the Raoult's Law limit).
- If $\gamma_i = 1$, then $a_i = x_i$, and the solution is ideal (at least for that component at that concentration).
- If $\gamma_i > 1$, then $a_i > x_i$. The component has a higher "effective concentration" than its mole fraction suggests. It's "unhappy" in the solution and wants to escape. This leads to a positive deviation from Raoult's Law (vapor pressure is higher than ideal).
- If $\gamma_i < 1$, then $a_i < x_i$. The component is "too comfortable" in the solution. It has a lower escaping tendency. This leads to a negative deviation from Raoult's Law (vapor pressure is lower than ideal).
We have our ideal mixing functions ($\Delta G_{mix}^{ideal}$, etc.) and our real mixing functions ($\Delta G_{mix}^{real}$). The difference between them is the Excess Function.
An excess function, $M^E$, is defined as the difference between the actual thermodynamic property of mixing and the value it would have if the solution were ideal.
$$ M^E = \Delta M_{mix}^{real} - \Delta M_{mix}^{ideal} $$
1. Excess Gibbs Free Energy ($G^E$)
$$ G^E = \Delta G_{mix}^{real} - \Delta G_{mix}^{ideal} $$
Let's derive the most important relation for $G^E$. For a binary mixture:
$\Delta G_{mix}^{real} = n_A (\mu_A - \mu_A^*) + n_B (\mu_B - \mu_B^*) $
$ \Delta G_{mix}^{real} = n_A (RT \ln a_A) + n_B (RT \ln a_B) = RT(n_A \ln(\gamma_A x_A) + n_B \ln(\gamma_B x_B)) $
$ \Delta G_{mix}^{real} = RT(n_A \ln x_A + n_B \ln x_B) + RT(n_A \ln \gamma_A + n_B \ln \gamma_B) $
Look closely! The first term is just $\Delta G_{mix}^{ideal}$. The second term must be $G^E$.
$$ G^E = RT(n_A \ln \gamma_A + n_B \ln \gamma_B) $$
For 1 mole of solution:
$$ G_{m}^E = RT(x_A \ln \gamma_A + x_B \ln \gamma_B) $$
This is the central equation! $\gamma$ is our measure of non-ideality, and $G^E$ is the thermodynamic quantity that stores that information.
- For positive deviation, $\gamma_i > 1$, so $\ln \gamma_i > 0$, which means $G^E > 0$.
- For negative deviation, $\gamma_i < 1$, so $\ln \gamma_i < 0$, which means $G^E < 0$.
2. Excess Enthalpy ($H^E$)
$$ H^E = \Delta H_{mix}^{real} - \Delta H_{mix}^{ideal} $$
Since $\Delta H_{mix}^{ideal} = 0$, this one is simple:
$$ H^E = \Delta H_{mix}^{real} $$
The excess enthalpy is just the actual, measurable heat of mixing! This is what you would measure with a mixing calorimeter.
3. Excess Entropy ($S^E$)
$$ S^E = \Delta S_{mix}^{real} - \Delta S_{mix}^{ideal} $$
This tells us if the mixture is more or less ordered than an ideal solution. For example, in the acetone-chloroform case, the strong H-bonding creates a more ordered structure, so $S^E$ would be negative.
4. Excess Volume ($V^E$)
$$ V^E = \Delta V_{mix}^{real} - \Delta V_{mix}^{ideal} $$
Since $\Delta V_{mix}^{ideal} = 0$:
$$ V^E = \Delta V_{mix}^{real} $$
This is the real, measurable volume change upon mixing. You can find it by accurately measuring the densities of the pure components and the mixture.
The excess functions are thermodynamic properties, so they obey all the same rules. The most important one is the Gibbs-Helmholtz equation, which connects $G^E$, $H^E$, and $S^E$:
$$ G^E = H^E - T S^E $$
We can also get $H^E$ from the temperature dependence of $G^E$:
$$ H^E = -T^2 \left[ \frac{\partial (G^E/T)}{\partial T} \right]_P $$
And $S^E$ can be found from the temperature dependence of $G^E$:
$$ S^E = -\left(\frac{\partial G^E}{\partial T}\right)_P $$
This is incredibly powerful! It means if we can determine $G^E$ (e.g., from vapor pressure measurements) at different temperatures, we can calculate both $H^E$ and $S^E$ without ever using a calorimeter!
The Gibbs-Duhem equation is one of the most fundamental relationships in chemical thermodynamics. It states that the properties of the components in a mixture are not independent of each other. If you change one, the others must change in a predictable way.
We start with the fundamental equation for the total Gibbs free energy, G, of an open system:
$$ dG = VdP - SdT + \sum \mu_i dn_i $$
Now, if we integrate this equation at constant T, P, and composition (i.e., we just make the system bigger), G must be a linear function of the moles, $n_i$. This is because G is an extensive property. This integration gives:
$$ G = \sum \mu_i n_i $$
This makes perfect sense: the total G is the sum of the moles of each component times its contribution to G (the chemical potential).
Now, let's take the total differential of this new equation (G = $\sum \mu_i n_i$):
$$ dG = \sum (\mu_i dn_i + n_i d\mu_i) $$
We now have two equations for dG. Let's set them equal to each other:
$$ VdP - SdT + \sum \mu_i dn_i = \sum \mu_i dn_i + \sum n_i d\mu_i $$
The $\sum \mu_i dn_i$ terms cancel, leaving:
$$ VdP - SdT = \sum n_i d\mu_i $$
This is the general form of the Gibbs-Duhem equation. For the most common experimental conditions, we hold T and P constant (so dT=0 and dP=0). This gives the most famous form:
$$ \sum n_i d\mu_i = 0 \quad (\text{at constant T, P}) $$
For a binary system (A and B):
$$ n_A d\mu_A + n_B d\mu_B = 0 $$
If we divide by the total moles, $n_T$, we get the form in mole fractions:
$$ x_A d\mu_A + x_B d\mu_B = 0 $$
Tutor's Note: This equation tells you that $\mu_A$ and $\mu_B$ are not independent variables! In a binary mixture, if you change the composition (dx_A), the change in $\mu_A$ (d$\mu_A$) dictates the change in $\mu_B$ (d$\mu_B$). You can't change one without the other.
This is a very practical application of the Gibbs-Duhem equation. We just derived $x_A d\mu_A + x_B d\mu_B = 0$. Let's see what this means for our activity coefficients.
We know $\mu_i = \mu_i^\circ + RT \ln a_i = \mu_i^\circ + RT \ln(\gamma_i x_i)$.
Let's find $d\mu_i$ (at constant T, P):
$$ d\mu_i = RT \cdot d[\ln(\gamma_i x_i)] = RT \cdot d[\ln \gamma_i + \ln x_i] $$
$$ d\mu_i = RT (d\ln \gamma_i + d\ln x_i) = RT \left(d\ln \gamma_i + \frac{dx_i}{x_i}\right) $$
Now substitute this back into $x_A d\mu_A + x_B d\mu_B = 0$:
$$ x_A \left[ RT \left(d\ln \gamma_A + \frac{dx_A}{x_A}\right) \right] + x_B \left[ RT \left(d\ln \gamma_B + \frac{dx_B}{x_B}\right) \right] = 0 $$
Divide by RT:
$$ (x_A d\ln \gamma_A + dx_A) + (x_B d\ln \gamma_B + dx_B) = 0 $$
Group the terms:
$$ (x_A d\ln \gamma_A + x_B d\ln \gamma_B) + (dx_A + dx_B) = 0 $$
In a binary mixture, $x_A + x_B = 1$, so $dx_A + dx_B = 0$. The second parenthesis vanishes!
We are left with the Gibbs-Duhem-Margules equation:
$$ x_A d\ln \gamma_A + x_B d\ln \gamma_B = 0 $$
We can also write this in derivative form by dividing by $dx_A$:
$$ x_A \left( \frac{\partial \ln \gamma_A}{\partial x_A} \right)_P + x_B \left( \frac{\partial \ln \gamma_B}{\partial x_A} \right)_P = 0 $$
We know $G^E$ is the key, but $G_m^E = RT(x_A \ln \gamma_A + x_B \ln \gamma_B)$ isn't a practical equation on its own. We need a model that gives $G^E$ as a simple function of composition ($x$).
We can express $G_m^E$ as a power series in mole fraction. The simplest non-trivial model is the two-suffix Margules equation:
$$ G_m^E = A x_A x_B $$
Here, 'A' is an empirical parameter (which may depend on T and P) that is determined by fitting experimental data. This model is for simple solutions and assumes a symmetric $\gamma$ vs. $x$ plot.
Let's apply our thermodynamics! How can we get the activity coefficients $\gamma_A$ and $\gamma_B$ from this model? We need to use partial molar properties.
The chemical potential $\mu_i$ is the partial molar Gibbs free energy, $\bar{G}_i$.
We know $RT \ln \gamma_i$ is the excess partial molar Gibbs free energy, $\bar{G}_i^E$.
So, $RT \ln \gamma_i = \bar{G}_i^E = \left(\frac{\partial (n_T G_m^E)}{\partial n_i}\right)_{T,P,n_{j \neq i}}$
Let's use $G^E = n_T G_m^E = n_T (A x_A x_B) = n_T A \left(\frac{n_A}{n_T}\right) \left(\frac{n_B}{n_T}\right) = \frac{A n_A n_B}{n_A + n_B}$
Now we take the partial derivatives:
$$ RT \ln \gamma_A = \left(\frac{\partial G^E}{\partial n_A}\right)_{T,P,n_B} = \frac{\partial}{\partial n_A} \left[ \frac{A n_A n_B}{n_A + n_B} \right] $$
Using the quotient rule:
$$ RT \ln \gamma_A = A \left[ \frac{n_B(n_A+n_B) - n_A n_B(1)}{(n_A+n_B)^2} \right] = A \left[ \frac{n_A n_B + n_B^2 - n_A n_B}{(n_A+n_B)^2} \right] $$
$$ RT \ln \gamma_A = A \left( \frac{n_B}{n_A+n_B} \right)^2 = A x_B^2 $$
By symmetry, for component B:
$$ RT \ln \gamma_B = A x_A^2 $$
So, for the Margules model $G_m^E = A x_A x_B$:
$$ \ln \gamma_A = (A/RT) x_B^2 \quad \text{and} \quad \ln \gamma_B = (A/RT) x_A^2 $$
(You can check for yourself that these expressions satisfy the Gibbs-Duhem-Margules equation! It's a great exercise.)
Another common model is the Regular Solution Model, which assumes $S^E = 0$. This means $G^E = H^E$. It's a "one step up" from ideal solutions, where we allow a heat of mixing ($H^E \neq 0$) but assume the entropy of mixing is still purely random ($S^E = 0$).
And that's the complete story! We've gone from the simple ideal solution to a complete thermodynamic framework for describing and measuring the behavior of real, non-ideal mixtures. I hope this detailed breakdown helps you master these concepts!
Great! You've got the theory, now let's apply it by tackling those 15 questions one by one. I'll break down the answer for each, just like we discussed.
Step 1: Thermodynamic Definition
An ideal solution is defined thermodynamically by the chemical potential $\mu_i$ of each component 'i'. For a component in an ideal solution, its chemical potential is given by:
$$ \mu_i(P, T, x_i) = \mu_i^*(P, T) + RT \ln x_i $$
Where:
Step 2: Deriving Thermodynamic Functions of Mixing
We calculate the change in a thermodynamic property $M$ when we mix pure components to form 1 mole of solution. $\Delta M_{mix} = M_{final} - M_{initial}$.
Step 3: Molecular Interpretation
$\Delta H_{mix} = 0$ because, at a molecular level, an ideal solution assumes the intermolecular forces between unlike molecules (A-B) are identical to the average forces between like molecules (A-A and B-B). Therefore, no net energy is required to break old interactions, and no net energy is released when forming new ones.
$\Delta V_{mix} = 0$ for the same reason. Since the forces are identical, the molecules pack together in the mixture with the same efficiency and spacing as they did when pure. The volumes are perfectly additive (e.g., 50 mL A + 50 mL B = 100 mL solution).
A non-ideal solution is a mixture that does not obey Raoult's Law over the entire composition range. This is the reality for most solutions, as the intermolecular forces between unlike molecules (A-B) are different from those between like molecules (A-A and B-B).
1. Necessity and Definition:
The simple chemical potential equation for ideal solutions ($\mu_i = \mu_i^* + RT \ln x_i$) is mathematically simple but fails for real solutions. To preserve this elegant thermodynamic form, we introduce two new concepts:
$$ \mu_i = \mu_i^\circ + RT \ln a_i $$
where $\mu_i^\circ$ is the chemical potential in a defined standard state (for a solvent, this is typically the pure liquid, $\mu_i^*$).$$ a_i = \gamma_i x_i $$
The activity coefficient $\gamma_i$ is our numerical measure of non-ideality. For an ideal solution, $\gamma_i = 1$.2. Derivation of Usefulness:
By substituting the definition of $a_i$ into the chemical potential equation, we get:
$$ \mu_i = \mu_i^\circ + RT \ln(\gamma_i x_i) = (\mu_i^\circ + RT \ln x_i) + RT \ln \gamma_i $$
This brilliantly separates the chemical potential into two parts:
3. Relation to Limiting Laws (Standard States):
We follow the same derivation as for ideal solutions, but now we must use the full expression for chemical potential, including the activity coefficient.
Let's derive the expressions for mixing $n_A$ moles of A and $n_B$ moles of B.
$G_{initial} = n_A \mu_A^* + n_B \mu_B^*$.
$G_{final} = n_A \mu_A + n_B \mu_B = n_A(\mu_A^* + RT \ln a_A) + n_B(\mu_B^* + RT \ln a_B)$.
$G_{final} = (n_A \mu_A^* + n_B \mu_B^*) + RT(n_A \ln a_A + n_B \ln a_B)$.
1. Gibbs Free Energy of Mixing ($\Delta G_{mix}$):
$\Delta G_{mix} = G_{final} - G_{initial}$
$$ \Delta G_{mix} = RT(n_A \ln a_A + n_B \ln a_B) $$
Substituting $a_i = \gamma_i x_i$:
$$ \Delta G_{mix} = RT(n_A \ln(\gamma_A x_A) + n_B \ln(\gamma_B x_B)) $$
$$ \Delta G_{mix} = \underbrace{RT(n_A \ln x_A + n_B \ln x_B)}_{\Delta G_{mix}^{ideal}} + \underbrace{RT(n_A \ln \gamma_A + n_B \ln \gamma_B)}_{G^E} $$
So, $\Delta G_{mix} = \Delta G_{mix}^{ideal} + G^E$.
2. Entropy of Mixing ($\Delta S_{mix}$):
$\Delta S_{mix} = -(\partial \Delta G_{mix} / \partial T)_P$. Applying this derivative to the full expression gives:
$$ \Delta S_{mix} = \underbrace{-R(n_A \ln x_A + n_B \ln x_B)}_{\Delta S_{mix}^{ideal}} - \underbrace{\left[ RT \left( n_A \frac{\partial \ln \gamma_A}{\partial T} + n_B \frac{\partial \ln \gamma_B}{\partial T} \right) + R(n_A \ln \gamma_A + n_B \ln \gamma_B) \right]}_{S^E} $$
This is complicated! A much easier way is to use the relationship $S^E = -(\partial G^E / \partial T)_P$.
$$ \Delta S_{mix} = \Delta S_{mix}^{ideal} + S^E $$
$$ \Delta S_{mix} = -R(n_A \ln x_A + n_B \ln x_B) - \left(\frac{\partial G^E}{\partial T}\right)_P $$
3. Enthalpy of Mixing ($\Delta H_{mix}$):
We know $\Delta H_{mix} = \Delta G_{mix} + T \Delta S_{mix}$.$$ \Delta H_{mix} = (\Delta G_{mix}^{ideal} + G^E) + T(\Delta S_{mix}^{ideal} + S^E) $$
$$ \Delta H_{mix} = (\Delta G_{mix}^{ideal} + T\Delta S_{mix}^{ideal}) + (G^E + TS^E) $$
We know from the ideal solution derivation that $\Delta G_{mix}^{ideal} + T\Delta S_{mix}^{ideal} = \Delta H_{mix}^{ideal} = 0$.
From the definition of excess functions, $G^E + TS^E = H^E$.
Therefore:
$$ \Delta H_{mix} = H^E $$
This is a crucial result: the enthalpy of mixing for a real solution is simply equal to its excess enthalpy. It is the heat you would measure in a calorimeter.
1. Statement and Derivation:
The Gibbs-Duhem equation is a fundamental relationship that shows that the chemical potentials (or other partial molar properties) of components in a mixture are not independent of each other.
Derivation:
$$ dG = VdP - SdT + \sum_{i} \mu_i dn_i $$
$$ G = \sum_{i} \mu_i n_i $$
$$ dG = \sum_{i} (\mu_i dn_i + n_i d\mu_i) $$
$$ VdP - SdT + \sum_{i} \mu_i dn_i = \sum_{i} \mu_i dn_i + \sum_{i} n_i d\mu_i $$
$$ VdP - SdT = \sum_{i} n_i d\mu_i $$
This is the most general form of the Gibbs-Duhem equation.$$ \sum_{i} n_i d\mu_i = 0 $$
2. Significance:
For a binary (two-component) system at constant T and P, the equation is: $$ n_A d\mu_A + n_B d\mu_B = 0 $$
Dividing by the total moles $n_T$ gives: $$ x_A d\mu_A + x_B d\mu_B = 0 $$
This means the chemical potentials are coupled. If you change the composition, causing $\mu_A$ to change, $\mu_B$ must also change in a compensating way. For example, if you add more A, $\mu_A$ increases ($d\mu_A > 0$), which means $\mu_B$ must decrease ($d\mu_B < 0$). This is the "golden rule" that keeps the thermodynamics of the mixture consistent.
This is a direct application of the Gibbs-Duhem equation, but written in terms of activity coefficients instead of chemical potentials.
Derivation:
$$ x_A d\mu_A + x_B d\mu_B = 0 $$
$$ \mu_i = \mu_i^\circ + RT \ln a_i = \mu_i^\circ + RT \ln(\gamma_i x_i) $$
$$ d\mu_i = RT \cdot d[\ln(\gamma_i x_i)] = RT \cdot d[\ln \gamma_i + \ln x_i] $$
$$ d\mu_i = RT (d\ln \gamma_i + d\ln x_i) = RT \left(d\ln \gamma_i + \frac{dx_i}{x_i}\right) $$
$$ x_A \left[ RT \left(d\ln \gamma_A + \frac{dx_A}{x_A}\right) \right] + x_B \left[ RT \left(d\ln \gamma_B + \frac{dx_B}{x_B}\right) \right] = 0 $$
$$ (x_A d\ln \gamma_A + dx_A) + (x_B d\ln \gamma_B + dx_B) = 0 $$
$$ (x_A d\ln \gamma_A + x_B d\ln \gamma_B) + (dx_A + dx_B) = 0 $$
$$ x_A d\ln \gamma_A + x_B d\ln \gamma_B = 0 $$
Application (Consistency Test):
If you perform an experiment to measure vapor pressures, you can calculate $\gamma_A$ and $\gamma_B$ across the entire composition range. You can then plot $\ln \gamma_A$ and $\ln \gamma_B$ vs. mole fraction, $x_A$. The slopes of these two curves at any given composition must obey the G-D-M equation (written in derivative form):
$$ x_A \left( \frac{\partial \ln \gamma_A}{\partial x_A} \right) + x_B \left( \frac{\partial \ln \gamma_B}{\partial x_A} \right) = 0 $$
If your data does not satisfy this equation, your experimental data is thermodynamically inconsistent and therefore invalid.
Besides testing for data consistency (as mentioned in Q6), a key application is to calculate the activity coefficient of one component (solute) by measuring the activity coefficient of the other (solvent).
Tutor's Note: This is incredibly useful! Think about measuring the properties of sugar (a non-volatile solute) in water. You can't measure the vapor pressure of the sugar. But you can easily measure the vapor pressure of the water above the solution. From the water's properties, we can calculate the sugar's properties!
Step-by-Step Procedure:
$$ d\ln \gamma_B = -\frac{x_A}{x_B} d\ln \gamma_A $$
$$ \int_{\text{at } x_B=0}^{\text{at } x_B'} d\ln \gamma_B = - \int_{\ln \gamma_A \text{ at } x_B=0}^{\ln \gamma_A \text{ at } x_B'} \left(\frac{x_A}{x_B}\right) d\ln \gamma_A $$
$$ \ln \gamma_B(\text{at } x_B') - \ln \gamma_B(\text{at } x_B=0) = - \int_0^{\ln \gamma_A(\text{at } x_B')} \left(\frac{x_A}{x_B}\right) d\ln \gamma_A $$
(Note: The $\ln \gamma_B(\text{at } x_B=0)$ term is tricky, related to the standard state, but in practice this integration is done numerically).
This technique is also used to find solute activity from colligative properties like freezing point depression or osmotic pressure, as these measurements also give the activity of the solvent.
1. Definition and Physical Significance:
An Excess Thermodynamic Function, denoted $M^E$, is the difference between the actual thermodynamic property of a real mixture ($\Delta M_{mix}^{real}$) and the value it would have if it were an ideal solution ($\Delta M_{mix}^{ideal}$) at the same T, P, and composition.
$$ M^E = \Delta M_{mix}^{real} - \Delta M_{mix}^{ideal} $$
Physical Significance: Excess functions are our primary way to quantify non-ideality. They isolate only the contribution from the non-ideal interactions, stripping away the ideal (purely statistical) effects of mixing.
2. Derivation of $G^E = RT(x_1 \ln \gamma_1 + x_2 \ln \gamma_2)$:
$\Delta G_{mix}^{real} = RT(x_A \ln a_A + x_B \ln a_B) = RT(x_A \ln(\gamma_A x_A) + x_B \ln(\gamma_B x_B))$
$\Delta G_{mix}^{ideal} = RT(x_A \ln x_A + x_B \ln x_B)$
$$ G^E = [RT(x_A \ln(\gamma_A x_A) + x_B \ln(\gamma_B x_B))] - [RT(x_A \ln x_A + x_B \ln x_B)] $$
$$ G^E = RT[ (x_A \ln \gamma_A + x_A \ln x_A) + (x_B \ln \gamma_B + x_B \ln x_B) ] - RT[x_A \ln x_A + x_B \ln x_B] $$
$$ G^E = RT(x_A \ln \gamma_A + x_B \ln \gamma_B) $$
This is the fundamental equation linking the macroscopic excess energy $G^E$ to the microscopic-level correction factors $\gamma_i$.The Excess Gibbs Free Energy ($G^E$) is most commonly determined from vapor-liquid equilibrium (VLE) data, which involves measuring the vapor pressure and composition of the vapor above a liquid mixture.
Experimental Method (Static or Dynamic Vapor Pressure Measurement):
Calculations to find $G^E$:
$$ P_A = y_A P_{total} $$
$$ P_B = y_B P_{total} $$
$$ P_i = \gamma_i x_i P_i^* $$
We rearrange this to solve for the activity coefficient $\gamma_i$:$$ \gamma_A = \frac{P_A}{x_A P_A^*} = \frac{y_A P_{total}}{x_A P_A^*} $$
$$ \gamma_B = \frac{P_B}{x_B P_B^*} = \frac{y_B P_{total}}{x_B P_B^*} $$
$$ G^E = RT(x_A \ln \gamma_A + x_B \ln \gamma_B) $$
There are two primary ways to determine excess enthalpy ($H^E$) and excess entropy ($S^E$): one direct and one indirect.
The excess enthalpy $H^E$ is, by definition, the real enthalpy of mixing $\Delta H_{mix}$. This can be measured directly using an isothermal mixing calorimeter.
Experimental Setup:
$$ H^E = \frac{\Delta H_{mix}}{n_A + n_B} $$
If we don't have a calorimeter, we can find all the excess functions by measuring $G^E$ (from Q9) at several different temperatures.
We use the fundamental thermodynamic relationships that connect G, H, and S.
Key Relationships:
1. $G^E = H^E - T S^E$
2. $S^E = -\left(\frac{\partial G^E}{\partial T}\right)_P$ (from $S = -(\partial G / \partial T)_P$)
3. $H^E = G^E + T S^E = G^E - T \left(\frac{\partial G^E}{\partial T}\right)_P$ (This is the Gibbs-Helmholtz equation, also written as $H^E = -T^2 \left[ \frac{\partial (G^E/T)}{\partial T} \right]_P$)
Procedure:
$$ H^E = G^E + T S^E $$
This indirect method is extremely powerful, as it allows us to determine the enthalpic (heat) and entropic (disorder) contributions to non-ideality just from vapor pressure measurements.
1. Definition and Assumptions:
The Regular Solution Model, proposed by Hildebrand, is a simple model that is one step more complex than an ideal solution. It is defined by two key assumptions:
$$ \Delta S_{mix}^{real} = \Delta S_{mix}^{ideal} = -R(x_A \ln x_A + x_B \ln x_B) $$
2. Derivations of Excess Functions:
$$ H^E = \Delta H_{mix} = \beta x_A x_B $$
(Note: $\beta$ is related to the interaction energies: $\beta \propto [E_{AB} - \frac{1}{2}(E_{AA} + E_{BB})]$).$$ G^E = \beta x_A x_B $$
3. Relation to Deviations:
The behavior of the regular solution is entirely controlled by the sign of the interaction parameter $\beta$:
Limitation: The main limitation of the model is the assumption $S^E = 0$. In reality, strong interactions (negative deviation) often lead to ordering ($S^E < 0$), and weak interactions (positive deviation) can lead to non-random "clumping" (also $S^E < 0$). The model is a good first approximation but fails for complex systems.
1. The Model:
The Margules equation is a more flexible, empirical model. It doesn't make physical assumptions like the Regular Solution model. Instead, it just represents $G^E$ as a polynomial power series of mole fractions. It is a data-fitting equation.
The simplest form is the two-suffix (or one-parameter) Margules equation, which is mathematically identical to the Regular Solution model:
$$ G^E = A x_A x_B $$
Here, 'A' is an empirical parameter (which may depend on T and P) that is found by fitting experimental data. Unlike the Regular Solution's $\beta$, 'A' contains both enthalpic and entropic contributions ($A = H^E/x_A x_B - T(S^E/x_A x_B)$).
2. Deriving Activity Coefficients:
We need to find the activity coefficients $\ln \gamma_A$ and $\ln \gamma_B$. We use the fundamental relationship from Q3:
$$ RT \ln \gamma_i = \mu_i^E = \bar{G}_i^E \quad (\text{The excess partial molar Gibbs energy}) $$
And $\bar{G}_i^E$ is found by: $\bar{G}_i^E = \left(\frac{\partial (n_T G_m^E)}{\partial n_i}\right)_{T,P,n_{j \neq i}}$
Let's derive $\ln \gamma_A$:
$$ n_T G_m^E = (n_A + n_B) \cdot A \cdot \left(\frac{n_A}{n_A + n_B}\right) \left(\frac{n_B}{n_A + n_B}\right) = \frac{A n_A n_B}{n_A + n_B} $$
$$ RT \ln \gamma_A = \left(\frac{\partial (n_T G_m^E)}{\partial n_A}\right) = \frac{\partial}{\partial n_A} \left[ \frac{A n_A n_B}{n_A + n_B} \right] $$
$$ RT \ln \gamma_A = A n_B \left[ \frac{(1)(n_A + n_B) - (n_A)(1)}{(n_A + n_B)^2} \right] $$
$$ RT \ln \gamma_A = A n_B \left[ \frac{n_B}{(n_A + n_B)^2} \right] = A \left[ \frac{n_B}{n_A + n_B} \right]^2 $$
$$ RT \ln \gamma_A = A x_B^2 \quad \text{or} \quad \ln \gamma_A = (A/RT) x_B^2 $$
$$ RT \ln \gamma_B = A x_A^2 \quad \text{or} \quad \ln \gamma_B = (A/RT) x_A^2 $$
3. Checking with Gibbs-Duhem-Margules (G-D-M):
The G-D-M equation is: $ x_A d\ln \gamma_A + x_B d\ln \gamma_B = 0 $. Let's see if our derived expressions satisfy this. We'll use the derivative form: $ x_A \left( \frac{d\ln \gamma_A}{dx_A} \right) + x_B \left( \frac{d\ln \gamma_B}{dx_A} \right) = 0 $.
$$ \frac{d\ln \gamma_A}{dx_A} = (A/RT) \cdot 2(1 - x_A) \cdot (-1) = -2(A/RT)x_B $$
$$ \frac{d\ln \gamma_B}{dx_A} = (A/RT) \cdot 2x_A $$
$$ x_A \left[ -2(A/RT)x_B \right] + x_B \left[ 2(A/RT)x_A \right] $$
$$ = -2(A/RT)x_A x_B + 2(A/RT)x_A x_B = 0 $$
The equation equals zero. The expressions are thermodynamically consistent.
Here is a side-by-side comparison of the thermodynamic functions of mixing and the excess functions for the three main cases. (All plots are vs. mole fraction, $x$, from 0 to 1).
An azeotrope (from Greek, "to boil unchanged") is a liquid mixture of two or more components that has a constant boiling point and whose vapor has the same composition as the liquid. This means you cannot separate an azeotrope by simple distillation.
Azeotrope formation is a direct consequence of non-ideal behavior (i.e., $G^E \neq 0$).
1. Definition of Partial Molar Properties:
A partial molar property describes how the total value of an extensive property $M$ (like Volume, Enthalpy, or Gibbs Energy) of a mixture changes when you add one mole of a specific component 'i', while keeping T, P, and the moles of all other components constant.
The partial molar property of 'i' ($\bar{M}_i$) is defined mathematically as the partial derivative:
$$ \bar{M}_i = \left(\frac{\partial M}{\partial n_i}\right)_{T, P, n_{j \neq i}} $$
Example: Partial Molar Volume ($\bar{V}_i$)
This is not the volume of 1 mole of pure 'i'. It is the change in the total solution volume when 1 mole of 'i' is added to a very large amount of the solution. If you add 1 mole of water to a large pool of pure water, the volume increases by 18 mL ($\bar{V}_{water} = 18$ mL/mol). But if you add 1 mole of water to a large pool of pure ethanol, the volume only increases by ~14 mL! This is because the water molecules are fitting into a different structural environment. Thus, $\bar{V}_i$ depends on the composition.
The most famous partial molar property is the chemical potential, which is the partial molar Gibbs free energy: $\mu_i = \bar{G}_i$.
2. Method of Intercepts (Graphical Technique):
This is a clever graphical method to determine the partial molar properties $\bar{M}_A$ and $\bar{M}_B$ at any composition, if you have a plot of the molar property of the solution ($M_m$) vs. mole fraction.
Let's use molar volume ($V_m$) as our example.
3. Relation to Gibbs-Duhem Equation:
This graphical method is not magic; it is the direct geometric representation of the Gibbs-Duhem equation.
The Gibbs-Duhem equation for a molar property M at constant T, P is:
$$ x_A d\bar{M}_A + x_B d\bar{M}_B = 0 $$
The equation for the molar property $M_m$ is: $M_m = x_A \bar{M}_A + x_B \bar{M}_B$.
If you differentiate this: $dM_m = x_A d\bar{M}_A + \bar{M}_A dx_A + x_B d\bar{M}_B + \bar{M}_B dx_B$.
The Gibbs-Duhem part $(x_A d\bar{M}_A + x_B d\bar{M}_B)$ is zero! So:
$$ dM_m = \bar{M}_A dx_A + \bar{M}_B dx_B $$
Since $dx_A = -dx_B$: $dM_m = (\bar{M}_B - \bar{M}_A) dx_B$.
This means the slope of the $M_m$ vs. $x_B$ plot is $ \frac{dM_m}{dx_B} = (\bar{M}_B - \bar{M}_A) $.
The equation for the tangent line you just drew is $y = (\text{slope})x + (\text{intercept})$. It can be shown with algebra that a line with slope $(\bar{M}_B - \bar{M}_A)$ that passes through the point $(x_B, M_m)$ must have intercepts of $\bar{M}_A$ (at $x_B=0$) and $\bar{M}_B$ (at $x_B=1$). The method works precisely because the partial molar properties obey the Gibbs-Duhem equation.
Time Remaining: 900s