.

 

 

 

Background

 

 

 

 

 

 

 

.

 

 

 

 

 

 

 

 

 

Thermodynamics of VLE

Importance of VLE | Thermo. of VLE | Lever Rule & VLE | Molecular Interactions | Molecular Simulation

Classical thermodynamics allows us to obtain a relationship between the equilibrium mole fraction in the liquid phase, x, and the vapor phase, y.  Equilibrium between the liquid and vapor phases requires that the temperatures, pressures, and partial fugacities of each component be equal in the two coexisting phases.  By equating the partial fugacity of component 1 in the liquid phase to its partial fugacity in the vapor phase, we can write

for the more volatile component.  (We use x for the liquid phase mole fraction of the more volatile component and 1 - x for the mole fraction of the less volatile component.  We use a similar procedure for y, but we use subscript 1 for the remaining properties of the more volatile component and subscript 2 for those of the less volatile component.)  This equation, and a similar equation for the less volatile component is exact, but difficult to use because it contains the partial fugacity coefficient at the system pressure, the pure component fugacity coefficient at the corresponding pure-component vapor pressure (sat), and the pointing correction factor F.  For most systems, these terms represent only small corrections, and for our purposes the VLE equation can be simplified to

A similar equation can be written for the less volatile component.  Notice that there are mainly two things that affect the relationship of y to x: the relative vapor pressures of the two pure components, P1* and P2*, and the nonidealities of the mixture as manifest in deviations of the activity coefficient (gamma) from unity.  This can perhaps better be seen by adding the above equation to the similar equation for component 2 which gives

If the activity coefficients are unity, then the system is ideal and is said to obey Raoult's Law.  Activity coefficients greater that unity give pressures that are larger than the sum of the pure vapor pressures and are said to have positive deviations from Raoult's Law.  Notice that from the above equation that positive deviations mean that the partial pressure in the vapor phase (left-hand side of the equation) is larger than it would be for an ideal system; i.e., the component "doesn't like" to be in the liquid phase.  We shall see that this corresponds to smaller attractions between molecules in the mixture than in the pure component.

Shown at the right are isotherms of pressure versus composition for two different binary systems.  The less volatile component is the same in both cases, but the more volatile component is different.  Notice the interplay of the two factors that affect the separability of the components or the ration of y to x.  In the case of the red line, the vapor pressures of the two components are very close together and very small positive deviations from ideality can cause there to be a maximum in the total pressure.  This maximum is a minimum-boiling azeotrope.  (Recall that higher volatility corresponds to a lower boiling temperature, so if a constant pressure plot were made instead of a constant temperatre plot, the maximum in pressure would correspond to a minimum in the boiling temperatures of the possible mixtures).  On the other hand, if the boiling points (or relative volatilities) of the two pure components are quite different, then it takes very large positive deviations from ideality to produce a minimum-boiling azeotrope as shown by the blue line.  The black lines show the ideal pressures for the system and the difference between the colored lines and the black lines are due to the nonidealities of the liquid mixtures as expressed by the activity coefficients.

 

A common way to show the relationship of y to x is on a x-y plot.  The same systems shown above on a pressure plot are shown on an x-y plot at the right by using the equation shown above.  Here the magenta line represents the pressure that would be observed for the system with disparate vapor pressures if the activity coefficient were unity.  Note that even if the system is ideal that the large difference in vapor pressures produces a large difference between y and x.  The system containing components of approximately the same volatility shows very little difference between the x and y values.  Mixtures of these two components would be very difficult to separate by distillation; the hypothetical magenta mixture could be separated by distillation with only a few trays.  However, the nonidealities also have an affect upon the mixture separability, in this case producing an azeotrope in both the red and blue system.  The azeotrope point is the point where x = y and the y-x plot crosses the 45-degree (black) line.  A liquid mixture of this composition when boiled gives off a vapor of exactly the same composition as the liquid and no separation takes place.

 

Mixtures may also exhibit negative deviations from Raoult's Law.  We shall see that this is produced when the attractions between the unlike molecules are greater than those between identical molecules in the pure components.  This added attraction is described by classical thermodynamics with activity coefficients that are less than unity.  As can be seen from the equations above, activity coefficients that are less than one lower the partial pressure of the component below the ideal value for the given composition; the molecules "prefer" to stay in the liquid phase. A Pxy plot for a system exhibiting negative deviations is shown at the right.  This particular system does not have an azeotrope, but if the negative deviations were large enough in comparison to the relative volatilities of the two pure components, then a minimum pressure would occur.  This would correspond to a maximum-boiling azeotrope.

 

When plotted for constant pressure conditions, the VLE data better illustrate the reason for the terminology "maximum- and minimum-boiling" azeotropes. Thus, a minimum on the Txy plot, as in the figure at the right, corresponds to a minimum-boiling azeotrope, or the minimum possible boiling temperature over the whole range of compositions for this mixture. This minimum temperature would of course be a maximum in system pressure at constant temperature on a Pxy plot.

The plots generated in this module's simulations are all at constant pressure like the one on the upper right. The corresponding yx diagram is also generated by the simulator. The mixture at the right is ethanol-toluene at 756 mm Hg.  The corresponding yx plot is also shown in the lower right. At this point you should be able to say whether this system exhibits positive or negative deviations from Raoult's Law. Can you also say from a molecular view point why it exhibits the deviations that it does?

 

In addition to Txy and  yx plots, the module simulator also generates activity coefficient and Gibbs energy plots.  Several mixtures will be seen to have large enough nonidealities relative to the separation in pure-component vapor pressures that they exhibit azeotropy. Of particular importance in viewing the simulations is for you to identify the types of molecular interactions (e.g., strong cross interactions).

 

The Gibbs excess energy is related to the individual activity coefficients by

The second equation shows that the change in Gibbs energy upon mixing two components consists of the ideal mixing term and the excess term.  Even if all of the interactions between molecules are identical, there is an increase in entropy, consequently a decrease in Gibbs energy, just due to the ability to randomize the molecules in the mixture relative to two separate regions of all one kind of molecule.  This is the ideal change in Gibbs energy represented by the first term.  The excess term then results from non-ideal behavior; i.e., differences in intermolecular interactions due to the unique electron density structure of the individual molecules and correlational effects between the those electron populations.  These molecular effects result in nonunity activity coeffients and either positive or negative excess Gibbs functions.

 

A plot of the excess Gibbs energy and the change in Gibbs energy upon mixing, similar to those generated by the module simulator, is shown at the right for the case of ethanol + toluene (positive deviations from Raoult's Law).  Remember that positive deviations occur because the molecules don't like each other in the mixture as much as within the pure components.  This produces activity coefficients greater than one and gE > 0.  What does this mean about the liquid mixture.  You might recall from your thermodynamics class that favorable processes at constant T and P proceed with a decrease in the Gibbs energy.  Note from the above equations that the ideal mixing term is always negative since mole fractions are always less than one.  Thus two components that form an ideal mixture are always completely soluble because the free energy goes down with the mixing process.  However, gE can be positive and large enough that the mixing Gibbs energy actually bows up.  In such a case, the Gibbs energy of the mixture can be minimized by splitting into two phases rather than following the upward trend of the single-phase Gibbs energy.  The plot at the right is approaching the limit of liquid-liquid partial miscibility, but the curve does not yet have a region that bows up.  Negative deviations, gE < 0, can never lead to liquid-liquid insolubility because the Gibbs energy of mixing is even more negative than the ideal case.  This is because the molecules "like each other" in the mixture. 

 

 

A sample of an activity coefficient plot similar to those generated by the module's simulator is shown at the right.  Often activity coefficients are compared on semilog plots as they are here.  Activity coefficients must go to unity as the pure component limit is approached for each of the two components. This is because they physically indicate deviations from ideal behavior for the component relative to the pure component standard state.  That is, just adding a molecule or two of a second component doesn't change the behavior of the liquid much from the pure-component case, so the activity coefficient will be very close to one.  But, when you have mostly component 2, then the activity coefficient of component 1 may be quite different than pure component 1 because molecule 1 is almost completely surrounded by molecules of component 2 and the 1-2 interactions will dominate the behavior of the activity coefficient of component 1.  Here note that the infinite dilution values of the activity coefficients are close are both close to 1.6.

The module simulator also generates plots of the relative volatility (usually given the symbol alpha).  The distribution coefficient, K = y/x, shows the volatility of an individual component; i.e., the composition in the vapor phase relative to the composition in the liquid phase.  The relative volatility, on the other hand, compares the volatility of one component to the other. It is defined by

for a two-component system (where as usual, we use x and y for the more volatile component mole fractions).  A relative volatility of 1 indicates that both components are equally volatile and no separation takes place.  The plot at the right shows a y-x plot for different values of the relative volatility.  As can be seen, the higher the relative volatility, the more separable are the two components.  This means fewer stages in a distillation column in order to effect the same separation between the overhead and bottoms streams.

 

______________________

[Back]

More Background: Importance of VLE | Thermo. of VLE | Lever Rule & VLE | Molecular Interactions | Molecular Simulation