461

Appendix A
Residue Curve Maps for Reactive Mixtures

A.1
Coupling Reaction and Distillation

Useful insights into the thermodynamics of a multiphase reacting system can be

obtained by analyzing the situation when chemical and phase equilibrium (C&PE)
are achieved simultaneously. The Gibbs rule can be used to find the degrees of
freedom F:

F = c +2−P −R (A.1)

where P, c, and R are the number of phases, components and independent reac-
tions, respectively. If only vapor–liquid equilibrium exists then P = 2, R = 0 and F
= c, usually selected among (c − 1) molar fractions, plus one state variable, P or T,
but when one independent chemical reaction takes place then R = 1 and F has
one dimension less. Some examples below will illustrate this fundamental rule.
Insightful solution of a C&PE problem can be found by graphical representation,
particularly using ternary diagrams. Let us consider the general stoichiometric
reaction:
c c
νA A + νBB + .... ↔ νP + νRR + ... or ∑ν A = 0
i i with νt = ∑ νi (A.2)
i =1 i =1

The composition can be expressed with respect to a reference species k as follows

[1]:

x io (vk − vt x k ) + vi (x k − x k 0 )
xi = (A.3)
vk − vt x k 0

The relation (A.3) describes the so-called stoichiometric lines converging into a pole
π, whose location can be determined by the following relation:

vi vi
xi = = (A.4)
∑i t
v v
i

Chemical Process Design: Computer-Aided Case Studies. Alexandre C. Dimian and Costin Sorin Bildea
Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ISBN: 978-3-527-31403-4
462 Appendix A Residue Curve Maps for Reactive Mixtures

Figure A.1 Conventional and transformed composition variables in a ternary diagram.

Figure A.1 illustrates the graphical interpretation in a right-angle triangle. Pure

components are marked in the vertices A, B, C. The molar fractions xA and xB are
represented on the edges CA and CB, while xC is visualized by the height of the
point representing the ternary mixture with respect to AB. In Figure A.1 the
chemical equilibrium curve is displayed too drawn by means of the relation:

aPvP aRvR .... xPvP xRvR .... γ PvP γ RvR ....

K eq = = = ∏(x i γ i )vi = K xK γ (A.5)
a AvA aBvB .... x AvA xBvB .... γ vAA γ BvB .... i

The mass-balance problem can be solved graphically. The median connecting the
vertex C with the AB edge corresponds to the transformation of an equimolar AB
mixture into C. Extending this line with an equal segment gives the position of
the pole π of coordinates (0,1) and (1,0). From this point, stoichiometric lines can
be drawn for any initial composition of the reaction mixtures. When the reaction
preserves the number of moles (νt = 0) the stoichiometric lines are parallel.

A.1.1
Transformed Variables

For more species a change of variables appears useful in order to reduce the
dimensionality of a graphical representation. For simplicity, let us consider the
reaction A + B = C taking place in one phase containing initially only A and B.
Before the reaction the number of moles is N0 = NA + NB. After the reaction
the number of moles becomes N A* , N B* and N C*, the total number being
 A.2 Reactive Residue Curve Map Analysis 463

N t = N A* + N B* + N C* = N 0 − N C* . The actual molar fractions may be written as:

x A = N A*/N t xB = N B*/N t xC = N C*/N t with Nt = N0/(1 + xC). The composition can
be expressed only as function of A and B using C used as reference, since the
three species are linked by a stoichiometric equation. Since N A = N A* + N C* and
N B = N B* + N C* one gets the expressions:

x A + xC x B + xC
XA = and X B = (A.6)
1 + xC 1 + xC

XA and XB are new composition variables Xi that verify the relation ∑Xi = 1.
More generally, the composition of a reacting system characterized by c molar
fractions can be reduced to (c − 1) new composition variables by the following
transformation [1, 2]:

νk x i − νi x k
Xi = (A.7)
νk − νt x k

The reference k component should preferably be a product.

For example, the AB mixture expressed in Figure A.1 by XA and XB mole frac-
tions on the AB edge leads at equilibrium to a mixture (xA, xB, xC) obtained by
intersecting the equilibrium curve with the stoichiometric line passing through
the initial mixture. Conversely, a ternary mixture where a chemical reaction at
equilibrium takes place may be described only by two transformed composition
variables.

A.2
Reactive Residue Curve Map Analysis

Residue curve (RCM) and distillation curve (DCM) maps are today standard tools
for designing distillation systems dealing with nonideal mixtures involving azeo-
tropes. A residue curve characterizes the evolution of the liquid composition in a
vessel during a batchwise distillation experiment. The whole compositional space
may be spanned by residue curves considering different initial mixture composi-
tions. For nonreactive mixtures the RCM is obtained by solving the component
dynamic material balance expressed by the following differential equation:

dx i
= xi − yi (A.8)
dξ

in which ξ = H/V is a “warped-time” variable defined as the ratio of molar liquid

holdup H by the molar vapor rate V, while xi and yi are vapor and liquid com-
positions, respectively. Getting reliable RCMs, their classification and use in the
464 Appendix A Residue Curve Maps for Reactive Mixtures

Figure A.2 Construction of the distillation lines for nonreactive (left)

and reactive mixtures (right).

design of distillation systems is discussed thoroughly in the book of Doherty and

Malone [2].
A similar representation is based on distillation lines [1], which describe the
composition on successive trays of a distillation column with an infinite number
of stages at infinite reflux (∞/∞ analysis). In contrast with relation (A.8) the distil-
lation lines may be obtained much easier by algebraic computations involving a
series of bubble and dew points, as follows:

x i ,1 → y i*,1 = x i ,2 → y i*,2 = x i ,3 → y i ,3 .... (A.9)

Figure A.2 (left) shows the construction of a distillation for an ideal ternary system
in which A and C are the light (stable node) and the heavy (unstable node) boilers,
while B is an intermediate boiler (saddle). The initial point xi,1 produces the vapor
y i,*1 that by condensation gives a liquid with the same composition such that the
next point is x i ,2 = y i*,1, etc. Accordingly, the distillation line describes the evolution
of composition on the stages of a distillation column at equilibrium and total reflux
from the bottom to the top. The slope of a distillation line is a measure of the rela-
tive volatility of components. The analysis in RCM or DCM leads to the same
results.
Similarly, when a chemical reaction takes place the residue curves can be found
by the equation [2]:

dx i
= x i − y i + Da(vi − vT x i )R (A.10)
dξ
 A.2 Reactive Residue Curve Map Analysis 465

In the above Da denotes the Damköhler number as the ratio of the characteristic
process time H/V to the characteristic reaction time 1/ro. The reaction rate ro is a
reference value at the system pressure and an arbitrary reference temperature, as
the lowest or the highest boiling point. For catalytic reactions ro includes a refer-
ence value of the catalyst amount. R is the dimensionless reaction rate R = r/ro.
The kinetics of a homogeneous liquid-phase reaction is described in general as
function of activities:

r = k (∏ aivi )pr − (∏ a vjj )eq /K eq  (A.11)

Thus, the parameter Da is a measure of the reaction rate, but its absolute value
cannot be taken as a basis for comparing different systems.
For sufficient high values of the Da number the reaction can reach an equilib-
rium state. This situation, in which phase and chemical equilibrium are reached
simultaneously is very useful in analyzing real systems. In this case the RCM can
be computed by using the following equation [2]:

dX i
= X i − Yi (A.12)
dξ

Equations (A.8) and (A.12) are similar, but the last makes use of transformed
variables.
Analogous with the procedure presented before, reactive distillation lines can be
obtained by computing a series of dew and bubble points incorporating a chemical
equilibrium term, as follows:

x i ,1 → y i*,1 ↔ x ieq,2 → y i*,2 ↔ x ieq,3 → y i*,3 ..... (A.13)

Graphical construction of reactive distillation lines at equilibrium is shown in

Figure A.2 (right) for the reversible reaction A + B ↔ C. The initial point xi,1, at
chemical equilibrium produces a vapor y i,*1, which by condensation and equilib-
rium reaction gives the liquid with the composition xi,2. This is found by crossing
the stoichiometric line passing through y i,*1 with the chemical equilibrium curve.
Then, the liquid xi,2 produces the vapor y i,*2, etc. Similarly, the points 11, 12, 13,
etc. shows the situation in which the mixture becomes richer in B and poorer in
A and C.
Figure A.2 (right) emphasizes a particular position where phase equilibrium and
stoichiometric lines are collinear. In other words the liquid composition remains
unchanged because the resulting vapor, after condensation, is converted into the
original composition. This point is a potential reactive azeotrope, but when the
composition satisfies chemical equilibrium too it becomes a true reactive azeotrope.
Some examples of residue curve maps are presented below. Ideal mixtures are
used to illustrate the basic features, which may be applied to some important
industrial applications.
466 Appendix A Residue Curve Maps for Reactive Mixtures

A.2.1
Three Reactive Components Mixtures

Let us consider the reversible reaction A + B ↔ C for which the relative volatilities
are in the order 3/2/1 and the equilibrium constant Kx = 6.75. The physical and
reactive distillation lines may be obtained simply by computation in ExcelTM by
applying the relations (A.13), as illustrated in Figure A.3. Note that in this case
the starting point, liquid with composition (0.1, 0.l, 0.8), is not at chemical equi-
librium. The co-ordinates of the triangle are in normal mole fractions. It may be
observed that after a short straight path the reactive distillation line superposes
the chemical equilibrium curve. Starting from other points shows the same trend.
Inverting the volatilities to 3/1/2 does not change the situation. This result is
consistent with the Gibbs phase rule. Since c = 3, P = 2 and R = 1, we have F = 3
− 2 − 1 + 2 = 2. If one variable, pressure or temperature, is fixed in order to deter-
mine the phase state, it remains that only one composition variable is needed.
This may be XA or XB as transformed variables. Such situations designate one
degree-of-freedom systems.
Figure A.3 also illustrates graphically the formation of a reactive azeotrope as
the point where a particular stoichiometric line becomes tangential to the nonreac-
tive residue curve and intersects simultaneously the chemical equilibrium curve.
Let us now consider the reaction A + B ↔ C conducted in the presence of an
inert I, which is the lightest component. The number of degrees of freedom
becomes F = 4 − 2 − 1 + 2 = 3. One specification as P or T leaves two composition
variables. It is convenient to start with mixtures of C and I that enables us to
span the whole composition space. The representation of four components by
means of a tetrahedron is possible, although not comfortable. As can be seen from

Figure A.3 Residues curves for a mixture involving three components, reaction A + B ↔ C.
 A.2 Reactive Residue Curve Map Analysis 467

Figure A.4 Reactive residue curve maps for a ternary system

containing inert, reaction A + B + I ↔ C + I.

Figure A.4 (left), the equilibrium curve from the previous case becomes a surface
of conical shape limited by the inert-free equilibrium curve and the edges AI and
BI. It is to be noted that all the reactive residue curves start at the reactive azeotrope
but with different paths towards the vertex I.
Another possibility is the representation in a two-dimensional diagram, as in
Figure A.4 (right). The component C being chosen as the reference, the relation
(A.3) gives the transformed co-ordinates: XA = (xA + xC)/(1 + xC) and XB =
(xB + xC)/(1 + xC). The residue curves run from the reactive azeotrope to the vertex
of component I. This situation is denoted by “two degrees of freedom systems”.

A.2.2
Four Reactive Components Mixtures

The advantage of bidimensional representation is evident if four reactive compo-

nents are involved, as in the class of reversible reactions A + B ↔ C + D. This situ-
ation covers an important number of industrial applications, as the esterification
of acids with alcohols. Selecting C as the reference, the transformed variables are
XA = xA + xC, XB = xB + xC and XD = xD − xC since νt = 0. The transformed variables
sums to one, but only two are used as co-ordinates. Accordingly, the pure compo-
nents may be placed in the corner of a square diagram, reactants or products on
the same diagonal. Figure A.5 displays the reactive distillation map traced as before
for the relative volatilities 4/2/6/1 and the equilibrium constant Keq = 5.
The above behavior has been verified experimentally in a number of cases.
Figure A.6 illustrates the RCM for the reactive system methanol / acetic acid/
methyl acetate / water at physical and chemical equilibrium [3] in which experi-
468 Appendix A Residue Curve Maps for Reactive Mixtures

Figure A.5 Residue curve map for an ideal reactive mixture

with relative volatilities 4/6/2/1 and reversible reaction
A + B ↔ C + D with Kx = 5.

Figure A.6 Comparison of computed (solid line) and

experimentally measured equilibrium reactive RCM for the
methyl acetate system at atmospheric pressure [3].

mental and calculated data are compared. The liquid activities were computed by
using UNIQUAC. Note the existence of a minimum-boiling azeotrope of metha-
nol/methyl acetate, which is the unstable node. The trajectories run from the
azeotrope to the acetic acid vertex, which is the stable node. The other components,
 A.3 Influence of Reaction Kinetics on the Residue Curve Map 469

methyl acetate, methanol and water are saddles. The agreement is good, taking
into account the sensitivity of modelling to experimental errors: the equilibrium
constant Kx was in the range of 4.6–8.2, the correction for liquid activities
K = 3.4–4.5, leading to a true equilibrium constant Keq in the range 20–30.

A.3
Influence of Reaction Kinetics on the Residue Curve Map

The kinetics of a reaction rate has a substantial influence on residue curve maps.
Distillation boundaries and physical azeotropes can vanish, while other singular
points due to kinetic effects might appear. The influence of the kinetics on RCM
can be studied by integrating Eq. (A.10) for finite Da numbers. In addition, the
singular points satisfy the relation:

xi − yi
DaR = (A.14)
x i − νT x i

As an illustration, Figure A.7 presents the evolution of RCM for the mixture isobu-
tene/ methanol/MTBE at a pressure of 8 bar from pure physical equilibrium to
simultaneous chemical and phase equilibrium [4]. Note that the absolute values
of Da from this work cannot be compared against other papers because of different
scaling of variables. Plot (a) shows the physical phase equilibrium (Da = 0) in
which there are two physical azeotropes, MeOH/MTBE (x = 0.55) and MeOH/IB
(x = 0.08), the first being saddle and the second unstable node, both linked by a
distillation boundary. Consequently, two distillation fields appear above and below
the separatrix in which MeOH and MTBE are stable nodes. Plot (b) shows the situ-
ation at Da = 10−4 when the reaction is slow. The upper distillation region is not
affected by reaction, unlike near to the MTBE corner. The stable point moves from
MTBE to a new position containing about 76.5% MTBE, 20% MeOH and 3.5%
IB. Raising Da further to 2 × 10−4 (plot c) leads to a situation in which the saddle
point disappears and all trajectories point out to the methanol vertex; the reaction
rate becomes dominant. For Da = 1, when the reaction rate is high enough (plot
d), the trajectories are collected by the chemical equilibrium curve, the RCM being
similar to the Figure A.3.
Summing up, the influence of the kinetics of a chemical reaction on the
vapor–liquid equilibrium is very complex. Physical distillation boundaries
may disappear, while new kinetic stable and unstable nodes may appear. As
result, the residue curve map with chemical reaction could look very different
from the physical plots. As a consequence, evaluating the kinetic effects on
residue curve maps is of great importance for conceptual design of reactive
distillation systems. However, if the reaction rate is high enough such that the
chemical equilibrium is reached quickly, the RCM simplifies considerably. But
even in this case the analysis may be complicated by the occurrence of reactive
azeotropes.
470 Appendix A Residue Curve Maps for Reactive Mixtures

Figure A.7 Influence of reaction kinetics on the residue curve map

(a) no reaction Da = 0, (b) slow reaction Da = 10−4, (c) intermediate
reaction Da = 2 × 10−4, (d) fast reaction Da = 1 [4].

A.4
Influence of the Accuracy of Thermodynamic Data

The accuracy of the thermodynamic data has a significant effect on RCM computa-
tion. In the case of slow reactions both kinetics and phase equilibrium should be
modelled accurately. If the reaction is fast enough the chemical reaction prevails.
In many cases chemical equilibrium may be taken as the reference. Consequently,
accurate knowledge of the chemical equilibrium constant is needed. When reactive
azeotropes and/or phase splitting might occur accurate modelling of phase equi-
librium is also needed.
 A.5 Reactive Azeotropes 471

It should be kept in mind that the calculation of the equilibrium constant is

subject to high sensitivity to small errors in thermochemical data due to its expo-
nential dependency to the standard Gibbs free energy variation, as expressed by
the relation:

K eq = exp(− ∆G 0 (T , P )/RT ) = exp[ −(∑ vi ∆G 0f ,i (T , P )/RT ]

i

∆G0 is the standard Gibbs free energy change of reaction obtained from the stan-
dard energy of formation of pure components ∆G 0f , i at 25 °C, P = 1 atm and a
suitable aggregation state. Once having determined Keq in standard conditions, the
van’t Hoff equation may be used to calculate Keq at other temperatures, as
follows:

 d ln K eq  = ∆HR (T )
0

 dT  P RT 2

where ∆HR0 is the standard enthalpy change of reaction. Alternatively, ∆G0 may be
calculated from enthalpy and entropy changes as expressed by the relation ∆G0 =
∆H0 − T∆S0. Whichever method is used the accuracy of data should be very good
in order to avoid false predictions.
For example, for a liquid-phase reaction at 400 K the product RT has the value
0.792 cal/mol. For ∆G0 = −2 kcal/mol one gets Keq = 15. For an absolute error of
±1 kcal/mol the values are 44 and 3.5, respectively, obviously leading to very dif-
ferent results. An error of ±1 kJ/mol would be acceptable.
Therefore, experimental validation of the chemical equilibrium constant is rec-
ommended for accurate design. In addition, for nonideal mixtures the equilibrium
constant has to be expressed in term of activities. Accurate modelling of phase
equilibrium is required particularly for binaries with a strong influence on Kγ.

A.5
Reactive Azeotropes

From a mathematical viewpoint the necessary and sufficient condition for azeot-
ropy is:

y1 − x 1 yi − xi
= i = 2, ..... , c − 1 (A.15)
v1 − v t x 1 v i − v t x i

The mathematical solution of Eq. (A.15) is tedious. An elegant graphical solution

has been proposed by Stichlmair and Fair [1]. The occurrence of a reactive azeo-
trope is expressed geometrically by the necessary condition that the tangent to the
residue (distillation) curve be collinear with the stoichiometric line. Such points
form the locus of potential reactive azeotropes. In order to become a true reactive
azeotrope the intersection point must also belong to the chemical equilibrium
472 Appendix A Residue Curve Maps for Reactive Mixtures

Figure A.8 Graphical identification of reactive azeotropes for the reaction A + B ↔ C.

Figure A.9 Reactive azeotropes for the reaction A + B ↔ C

when A and B form a minimum azeotrope.

curve. Thus, reactive azeotropes can be identified by a graphical construction, as

illustrated in Figure A.8 for the reaction A + B ↔ C. Residues curves emerge from
the unstable node C, turn to the saddle B and terminate into the stable node C.
Stoichiometric lines tangent can be traced from the pole π, as previously explained.
The graphical construction is able to find points that satisfy the conditions of reac-
tive azeotropy, which form an arched curve around the edge BC. It may be
observed that when the residue curves are concave they cannot cross the stoichio-
 References 473

metric lines, and as a result there are no reactive azeotropes. This would happen
if B is the highest boiler and C and intermediate.
This simple graphical illustration allows the formulation of a practical rule:
reactive azeotropes may occur for ideal mixtures having “segregated volatilities”
(reactants either lighter or heavier with respect to products), but should not form
in the case of “mixed volatilities”.
A similar construction can be imagined when A and B form a minimum boiling
azeotrope, as illustrated in Figure A.9. This time a second reactive azeotrope curve
appears in the upper distillation region, between the AB-minimum azeotrope and
the reactant A. On the other hand, when the volatility order is changed, as for
example B becomes the heaviest and C the intermediate, no reactive azeotropes
can be found since the residue curves and the equilibrium curves are aligned in
the same direction.

References

1 Stichlmair, J. G., Fair, J. R., Distillation: 4 Thiel, C., Sundmacher, K., Hoffmann, U.,
Principles and Practice, Wiley-VCH, Residue curve maps for heterogeneously
Weinheim, Germany, 1998 reactive distillation of fuel ethers MTBE and
2 Doherty M. F., Malone M. F., Conceptual TAME, Chem. Eng. Sci., 52, no. 6, 993–1005,
design of distillation systems, McGraw- 1997
Hill, New York, USA, 2001
3 Song, W., Venimadhavan, G., Manning,
J., Malone, M. F., Doherty, M. F, Eng.
Chem. Res., 37, 1917–1928, 1998