chapter

3
Introduction to Mass Transfer

OVERVIEW
As indicated in Chapter 1, understanding combustion requires a combined knowledge
of thermodynamics (Chapter 2), heat and mass transfer, and chemical reaction rate
theory, or chemical kinetics (Chapter 4). Since most readers of this book are unlikely
to have had much, if any, exposure to the subject of mass transfer, we present in
this chapter a brief introduction to this topic. Mass transfer, a fundamental topic
in chemical engineering, is quite complex, much more so than is suggested by
the following discussion. We provide here only a rudimentary treatment of the
fundamental rate laws and conservation principles governing mass transfer, leaving a
more comprehensive treatment of mass transfer until Chapter 7 and to other textbooks
[1– 4]. To develop some physical insight into mass transfer, we briefly examine the
process from a molecular point of view. This has the added advantage of showing
the fundamental similarity of mass transfer and heat conduction in gases. Lastly, we
illustrate the application of mass transfer concepts to the mathematical descriptions
of the evaporation of a liquid layer and a droplet.

RUDIMENTS OF MASS TRANSFER

Imagine opening a bottle of perfume and placing the opened bottle in the center of a
room. Using your nose as a sensor, the presence of perfume molecules in the immedi-
ate vicinity of the bottle can be detected shortly after the bottle is opened. At a later
time, you will find the perfume odor everywhere in the room. The processes whereby
the perfume molecules are transported from a region of high concentration, near the
bottle, to a region of low concentration, far from the bottle, are the subject of mass
transfer. Like heat transfer and momentum transfer, mass may be transported by

79

tur80199_ch03_079-[Link] 79 1/4/11 11:58:32 AM

 80 C H A P T E R 3 • Introduction to Mass Transfer

molecular processes (e.g., collisions in an ideal gas) and /or turbulent processes. The
molecular processes are relatively slow and operate on small spatial scales, whereas
turbulent transport depends upon the velocity and size of an eddy carrying the trans-
ported material. Our focus here is on molecular transport, while Chapters 11, 12, and
13 deal with the turbulent processes.

Mass Transfer Rate Laws

Fick’s Law of Diffusion Let us consider a nonreacting gas mixture comprising just
two molecular species: species A and B. Fick’s law describes the rate at which one
species diffuses through the other. For the case of one-dimensional binary diffu-
sion, Fick’s law on a mass basis is

dYA
m A′′ = YA (m A′′ + m B′′ ) − ρDAB , (3.1)
dx
Mass flow of Mass flow of speciees A Mass flow of species A
species A per unit area associated with bulk associated with molecular
flow per unit area difffusion per unit area

where m A′′ is the mass flux of species A and YA is the mass fraction. In this book, the
mass flux is defined as the mass flowrate of species A per unit area perpendicular to
the flow:
m A′′ = m A /A. (3.2)

The units of m A′′ are kg /s-m2. The idea of a “flux” should be familiar to you since the
“heat flux’’ is the rate at which energy is transported per unit area; i.e., Q ′′ = Q /A with
units of J/s-m2 or W/m2. The binary diffusivity, DAB, is a property of the mixture and
has units of m2/s. Values for some binary diffusion coefficients, i.e., diffusivities, at
1 atm are provided in Appendix D.
Equation 3.1 states that species A is transported by two means: the first term on
the right-hand side representing the transport of A resulting from the bulk motion of
the fluid, and the second term representing the diffusion of A superimposed on the
bulk flow. In the absence of diffusion, we obtain the obvious result that

m A′′ = YA (m A′′ + m B′′ ) = YA m ′′ ≡ Bulk flux of species A, (3.3a)

where m ′′ is the mixture mass flux. The diffusional flux adds an additional compo-
nent to the flux of A:

dYA
− ρDAB ≡ Diffusional flux of species A, m A,diff
′′ . (3.3b)
dx

This expression says that the diffusional flux of A, m A,diff

′′ , is proportional to the
gradient of the mass fraction, where the constant of proportionality is − ρDAB . Thus,
we see that species A preferentially moves from a region of high concentration to

tur80199_ch03_079-[Link] 80 1/4/11 11:58:35 AM

 Rudiments of Mass Transfer 81

a region of low concentration, analogous to energy traveling in the direction from

high temperature to low. Note that the negative sign causes the flux to be positive in
the x-direction when the concentration gradient is negative. An analogy between the
diffusion of mass and the diffusion of heat (conduction) can be drawn by comparing
Fourier’s law of conduction,

dT
Q x′′ = − k , (3.4)
dx

with Fick’s law of diffusion in the absence of bulk flow, Eqn. 3.3b. Both expres-
′′ diff or Q x′′) being proportional to the gradient of a scalar
sions indicate a flux (m A,
quantity ((dYA /dx) or (dT/dx)). We will explore this analogy further when we discuss
the physical significance of the transport properties, rD and k, the proportionality
constants in Eqns. 3.3b and 3.4, respectively.
Equation 3.1 is the one-dimensional component of the more general expression

 A′′ = YA ( m
m  A′′ + m
 B′′ ) − ρDAB ⵱YA, (3.5)

where the bold symbols represent vector quantities. In many instances, the molar
form of Eqn. 3.5 is useful:

N A′′ = χ A ( N A′′ + N B′′ ) − cDAB⵱χ A, (3.6)

where N A′′ is the molar flux (kmol /s-m2) of species A, cA is the mole fraction, and c
is the mixture molar concentration (kmolmix /m3).
The meanings of bulk flow and diffusional flux become clearer if we express the
total mass flux for a binary mixture as the sum of the mass flux of species A and the
mass flux of species B:

m ′′ = m A′′ + m B′′ . (3.7)

Mixture Species A Species B
mass flux mass flux mass flux

The mixture mass flux on the left-hand side of Eqn. 3.7 is the total mixture flow-
rate m per unit of area perpendicular to the flow. This is the m that you are familiar
with from previous studies in thermodynamics and fluid mechanics. Assuming one-
dimensional flow for convenience, we now substitute the appropriate expressions for
the individual species mass fluxes (Eqn. 3.1) into Eqn. 3.7 and obtain

dYA dY
m ′′ = YA m ′′ − ρDAB + YB m ′′ − ρDBA B (3.8a)
dx dx
or
dYA dY
m ′′ = (YA + YB ) m ′′ − ρDAB − ρ D BA B . (3.8b)
dx dx

tur80199_ch03_079-[Link] 81 1/4/11 11:58:36 AM

 82 C H A P T E R 3 • Introduction to Mass Transfer

For a binary mixture, YA + YB = 1 (Eqn. 2.10); thus,

dYA dYB
− ρ DAB − ρDBA = 0, (3.9)
dx dx
Diffusional flux Diffusional flux
of species A of species B

i.e., the sum of the diffusional fluxes of the species is zero. In general, overall mass
conservation requires that Σ m i,′′ diff = 0.
It is important at this point to emphasize that we are assuming a binary gas and
that species diffusion is a result of concentration gradients only, which is termed
ordinary diffusion. Real mixtures of interest in combustion contain many compo-
nents, not just two. Our binary gas assumption, however, allows us to understand the
essential physics of many situations without the complications inherent in an analy-
sis of multicomponent diffusion. Also, gradients of temperature and pressure can
produce species diffusion, i.e., the thermal diffusion (Soret) and pressure diffusion
effects, respectively. In many systems of interest, these effects are usually small,
and, again, their neglect allows us to understand more clearly the essential physics
of a problem.

Molecular Basis of Diffusion To obtain insight into the molecular processes that
result in the macroscopic laws of mass diffusion (Fick’s law) and heat diffusion or
conduction (Fourier’s law), we apply some concepts from the kinetic theory of gases
(e.g., [5, 6]). Consider a stationary (no bulk flow) plane layer of a binary gas mixture
consisting of rigid, nonattracting molecules in which the molecular mass of each
species A and B is essentially equal. A concentration (mass-fraction) gradient ex-
ists in the gas layer in the x-direction and is sufficiently small that the mass-fraction
distribution can be considered linear over a distance of a few molecular mean free
paths, l , as illustrated in Fig. 3.1. With these assumptions, we can define the follow-
ing average molecular properties derived from kinetic theory [1, 5, 6]:

1/ 2
Mean speed of ⎛ 8kB T ⎞
v≡ = ⎜⎝ π m ⎟⎠ , (3.10a)
species A molecules A

Wall collision frequency 1 ⎛ nA ⎞

Z A′′ ≡ = (3.10b)
of A molecules per unitt area 4 ⎜⎝ V ⎟⎠
v,

1
λ ≡ Mean free path = ,
( )σ
ntot
(3.10c)
2π V
2

Average perpendicular distance

2
a ≡ from plane of last colllision to plane = λ , (3.10d)
where next collision occurs 3

tur80199_ch03_079-[Link] 82 1/4/11 11:58:36 AM

 Rudiments of Mass Transfer 83

Figure 3.1 Schematic diagram illustrating mass diffusion of species A molecules from a
region of high concentration to one of low concentration. Mass-fraction distribution is shown
at the top.

where kB is Boltzmann’s constant, mA the mass of a single A molecule, nA /V the

number of A molecules per unit volume, ntot / V the total number of molecules per unit
volume, and s the diameter of both A and B molecules.
Assuming no bulk flow for simplicity, the net flux of A molecules at the x-plane
is the difference between the flux of A molecules in the positive x-direction and the
flux of A molecules in the negative x-direction:

m A′′ = m A′′ , ( + ) x -dir − m A′′ , ( − ) x -dir , (3.11)

which, when expressed in terms of the collision frequency, becomes

m A′′ = ( Z A′′ ) x − a mA − ( Z A′′ ) x + a mA . (3.12)

⎛ Number of A ⎞ ⎛ Number of A ⎞
⎛ Net mass ⎞ ⎜ molecules crossing ⎟ ⎛ Mass of ⎞ ⎜ moleccules crossing ⎟ ⎛ Mass of ⎞
⎜ flux of ⎟ ⎜ plane at x originating ⎟ ⎜ single ⎟ ⎜ plane at x originating ⎟ ⎜ single ⎟
⎝ species A ⎠ ⎜ from plane at x − a ⎟ ⎝ molecule ⎠ ⎜ from plane at x + a ⎟ ⎝ molecule ⎠
⎜⎝ per unit time and area ⎟⎠ ⎜⎝ per unit time and area ⎟⎠

tur80199_ch03_079-[Link] 83 1/4/11 11:58:36 AM

 84 C H A P T E R 3 • Introduction to Mass Transfer

We can use the definition of density ( r ≡ mtot /Vtot ) to relate Z A′′ (Eqn. 3.10b) to the
mass fraction of A molecules:
1 nA m A 1
Z A′′ m A = ρ v = YA ρ v . (3.13)
4 mtot 4
Substituting Eqn. 3.13 into Eqn. 3.12 and treating the mixture density and mean
molecular speeds as constants yields
1
m A′′ = ρ v (YA, x − a − YA, x + a ). (3.14)
4
With our assumption of a linear concentration distribution,

dYA YA, x + a − YA, x − a YA, x + a − YA, x − a

= = . (3.15)
dx 2a 4λ / 3
Solving Eqn. 3.15 for the concentration difference and substituting into Eqn. 3.14,
we obtain our final result:
vλ dYA
m A′′ = − ρ . (3.16)
3 dx
Comparing Eqn. 3.16 with Eqn. 3.3b, we identify the binary diffusivity DAB as
vλ
DAB = . (3.17)
3
Using the definitions of the mean molecular speed (Eqn. 3.10a) and mean free path
(Eqn. 3.10c), together with the ideal-gas equation of state PV = nkBT, the temperature
and pressure dependence of DAB can easily be determined, i.e.,
1/ 2
2 ⎛ k3T ⎞ T
DAB = ⎜ 3B ⎟ (3.18a)
3 ⎝ π mA ⎠ σ 2P
or
DAB ∝ T 3/ 2 P −1 . (3.18b)

Thus, we see that the diffusivity depends strongly on temperature (to the 32 power)
and varies inversely with pressure. The mass flux of species A, however, depends on
the product rDAB, which has a square-root temperature dependence and is independ-
ent of pressure:
ρDAB ∝ T 1/ 2 P 0 = T 1/ 2 . (3.18c)

In many simplified analyses of combustion processes, the weak temperature depend-

ence is neglected and rD is treated as a constant.

Comparison with Heat Conduction To see clearly the relationship between mass
and heat transfer, we now apply kinetic theory to the transport of energy. We assume

tur80199_ch03_079-[Link] 84 1/4/11 11:58:37 AM

 Rudiments of Mass Transfer 85

a homogeneous gas consisting of rigid nonattracting molecules in which a tempera-

ture gradient exists. Again, the gradient is sufficiently small that the temperature dis-
tribution is essentially linear over several mean free paths, as illustrated in Fig. 3.2.
The mean molecular speed and mean free path have the same definitions as given in
Eqns. 3.10a and 3.10c, respectively; however, the molecular collision frequency of
interest is now based on the total number density of molecules, ntot /V, i.e.,

Average wall collision 1⎛n ⎞

Z ′′ ≡ = ⎜ tot ⎟ v . (3.19)
frequency per unit area 4 ⎝ V ⎠

In our no-interaction-at-a-distance hard-sphere model of the gas, the only energy

storage mode is molecular translational, i.e., kinetic, energy. We write an energy
balance at the x-plane (see Fig. 3.2) where the net energy flow (per unit area) in the
x-direction is the difference between the kinetic energy flux associated with mol-
ecules traveling from x − a to x and those traveling from x + a to x:

Q x′′ = Z ′′( ke) x − a − Z ′′( ke) x + a .

(3.20)

Figure 3.2 Schematic diagram illustrating energy transfer (heat conduction) associated
with molecular motion in a gas. The temperature distribution is shown at the top.

tur80199_ch03_079-[Link] 85 1/4/11 11:58:37 AM

 86 C H A P T E R 3 • Introduction to Mass Transfer

Because the mean kinetic energy of a molecule is given by [5]

1 2 3
ke = mv = kBT , (3.21)
2 2
the heat flux can be related to the temperature as

3
Q x′′ = kB Z ′′(Tx − a − Tx + a ). (3.22)
2
The temperature difference in Eqn. 3.22 relates to the temperature gradient following
the same form as Eqn. 3.15, i.e.,

dT Tx + a − Tx − a (3.23)
= .
dx 2a
Substituting Eqn. 3.23 into Eqn. 3.22, employing the definitions of Z ′′ and a, we
obtain our final result for the heat flux:

Q x′′ = − kB ⎛⎜ ⎞⎟ v λ
1 n dT (3.24)
.
2 ⎝ V ⎠ dx

Comparing the above with Fourier’s law of heat conduction (Eqn. 3.4), we can iden-
tify the thermal conductivity k as

1 ⎛ n⎞
k= k ⎜ ⎟ v λ. (3.25)
2 B⎝V⎠

Expressed in terms of T and molecular mass and size, the thermal conductivity is

1/ 2
⎛ k3 ⎞ (3.26)
k=⎜ 3 B 4⎟ T 1/ 2 .
⎝ π mσ ⎠

The thermal conductivity is thus proportional to the square root of temperature,

k ∝ T 1/ 2 , (3.27)

as is the rDAB product. For real gases, the true temperature dependence is greater.

Species Conservation
In this section, we employ the rate law of species transport (Fick’s law) to develop
a basic species mass conservation expression. Consider the one-dimensional con-
trol volume of Fig. 3.3, a plane layer Δ x thick. Species A flows into and out of
the control volume as a result of the combined action of bulk flow and diffusion.
Within the control volume, species A may be created or destroyed as a result of
chemical reaction.

tur80199_ch03_079-[Link] 86 1/4/11 11:58:37 AM

 Rudiments of Mass Transfer 87

Figure 3.3 Control volume for one-dimensional analysis of conservation of species A.

The net rate of increase in the mass of A within the control volume relates to the
mass fluxes and reaction rate as follows:
dmA, c v
= [m A′′ A]x − [ m A′′ A] x + Δx + m A′′′ V, (3.28)
dt
Rate of increase Mass flow of A Mass flow of A Mass production
of mass of A into the out of the rate of species A
within control volume control volume control volume by chemical reaaction

where the species mass flux m A′′ is given by Eqn. 3.1, and m A′′′ is the mass production
rate of species A per unit volume (kgA /m3-s). In Chapter 5, we specifically deal with
how to determine m A′′′. Recognizing that the mass of A within the control volume is
mA, cv = YA mcv = YA ρVcv and that the volume Vcv = AΔ x , Eqn. 3.28 can be rewritten:

∂( ρYA ) ⎡ ∂Y ⎤
AΔx = A ⎢YA m ′′ − ρDAB A ⎥
∂t ⎣ ∂x ⎦ x
(3.29)
⎡ ∂Y ⎤
− A ⎢YA m ′′ − ρDAB A ⎥ + m A′′′AΔx.
⎣ ∂x ⎦ x + Δx

Dividing through by AΔx and taking the limit as Δx → 0, Eqn. 3.29 becomes

∂( ρYA ) ∂ ⎡ ∂Y ⎤
= − ⎢YA m ′′ − ρDAB A ⎥ + m A′′′ (3.30)
∂t ∂x ⎣ ∂x ⎦
or, for the case of steady flow where ∂( ρYA ) / ∂t = 0,

d ⎡ dY ⎤
m A′′′ − ⎢YA m ′′ − ρDAB A ⎥ = 0. (3.31)
dx ⎣ dx ⎦

tur80199_ch03_079-[Link] 87 1/4/11 11:58:38 AM

 88 C H A P T E R 3 • Introduction to Mass Transfer

Equation 3.31 is the steady-flow, one-dimensional form of species conservation for a

binary gas mixture, assuming species diffusion occurs only as a result of concentra-
tion gradients; i.e., only ordinary diffusion is considered. For the multidimensional
case, Eqn. 3.31 can be generalized as

m A′′′ −  ′′A
⵱⋅m = 0. (3.32)
Net rate of production Net flow of speciess A
of species A by out of control volume,
chemical reaction, per unit volume
per unit volume

In Chapter 7, we will employ Eqns. 3.31 and 3.32 to develop the conservation of
energy principle for a reacting system. Chapter 7 also provides a more detailed treat-
ment of mass transfer, extending the present development to multicomponent (non-
binary) mixtures and including thermal diffusion.

SOME APPLICATIONS OF MASS TRANSFER

The Stefan Problem

Consider liquid A, maintained at a fixed height in a glass cylinder as illustrated
in Fig. 3.4. A mixture of gas A and gas B flow across the top of the cylinder. If
the concentration of A in the flowing gas is less than the concentration of A at the
liquid–vapor interface, a driving force for mass transfer exists and species A will
diffuse from the liquid–gas interface to the open end of the cylinder. If we assume
a steady state exists (i.e., the liquid is replenished at a rate to keep the liquid level
constant, or the interface regresses so slowly that its movement can be neglected)

Figure 3.4 Diffusion of vapor A through a stagnant column of gas B, i.e., the Stefan
problem.

tur80199_ch03_079-[Link] 88 1/4/11 11:58:38 AM

 Some Applications of Mass Transfer 89

and, furthermore, assume that B is insoluble in liquid A, then there will be no net
transport of B in the tube, producing a stagnant layer of B in the column.
Mathematically, the overall conservation of mass for this system can be
expressed as

m ′′( x ) = constant = m A′′ + m B′′. (3.33)

Since m B′′ = 0, then

m A′′ = m ′′( x ) = constant. (3.34)

Equation 3.1 now becomes

dYA
m A′′ = YA m A′′ − ρDAB . (3.35)
dx

Rearranging and separating variables, we obtain

m A′′ dYA
− dx = . (3.36)
ρDAB 1 − YA

Assuming the product rDAB to be constant, Eqn. 3.36 can be integrated to yield

m A′′ (3.37)
− x = − ln[1 − YA ] + C ,
ρDAB

where C is the constant of integration. With the boundary condition

YA ( x = 0) = YA, i , (3.38)

we eliminate C and obtain the following mass-fraction distribution after removing

the logarithm by exponentiation:

⎡ m ′′ x ⎤
YA ( x ) = 1 − (1 − YA, i ) exp ⎢ A ⎥ . (3.39)
⎣ ρDAB ⎦

The mass flux of A, m A′′ , can be found by letting YA(x = L) = YA, ∞ in Eqn. 3.39.
Thus,

ρDAB ⎡ 1 − YA , ∞ ⎤
m A′′ = ln ⎢ ⎥. (3.40)
L ⎢⎣ 1 − YA, i ⎥⎦

From Eqn. 3.40, we see that the mass flux is directly proportional to the product
of the density, r, and the mass diffusivity, DAB, and inversely proportional to the
length, L. Larger diffusivities thus produce larger mass fluxes.
To see the effects of the concentrations at the interface and at the top of the
tube, let the mass fraction of A in the freestream flow be zero, while arbitrarily

tur80199_ch03_079-[Link] 89 1/4/11 11:58:38 AM

 90 C H A P T E R 3 • Introduction to Mass Transfer

Table 3.1 Effect of interface mass fraction on mass flux

YA, i  A′′ / ( ρ AB / L )

m

0 0
0.05 0.0513
0.10 0.1054
0.20 0.2231
0.50 0.6931
0.90 2.303
0.999 6.908

varying YA, i , the interface mass fraction, from zero to unity. Physically, this could
correspond to an experiment in which dry nitrogen is blown across the tube outlet
and the interface mass fraction is controlled by the partial pressure of the liquid,
which, in turn, is varied by changing the temperature. Table 3.1 shows that at small
values of YA, i, the dimensionless mass flux is essentially proportional to YA, i . For YA, i
greater than about 0.5, the mass flux increases very rapidly.

Liquid–Vapor Interface Boundary Conditions

In the example above, we treated the gas-phase mass fraction of the diffusing species
at the liquid–vapor interface, YA, i , as a known quantity. Unless this mass fraction is
measured, which is unlikely, some means must be found to calculate or estimate its
value. This can be done by assuming equilibrium exists between the liquid and vapor
phases of species A. With this equilibrium assumption, and the assumption of ideal
gases, the partial pressure of species A on the gas side of the interface must equal the
saturation pressure associated with the temperature of the liquid, i.e.,

PA, i = Psat (Tliq, i ). (3.41)

The partial pressure, PA, i, can be related to the mole fraction of species A,
χ A, i = Psat /P, and to the mass fraction:

Psat (Tliq, i ) MWA

YA, i = , (3.42)
P MWmix, i

where the molecular weight of the mixture also depends on χ A, i , and, hence, on Psat.
The above analysis has transformed the problem of finding the vapor mass frac-
tion at the interface to the problem of finding the temperature at the interface. In
some cases, the interface temperature may be given or known, but, in general, the
interface temperature must be found by writing energy balances for the liquid and
gas phases and solving them with appropriate boundary conditions, including that at
the interface. In the following, we will establish this interface boundary condition,
but leave the gas-phase and liquid-phase energy balances for later.

tur80199_ch03_079-[Link] 90 1/4/11 11:58:39 AM

 Some Applications of Mass Transfer 91

Figure 3.5 Energy balance at the surface of evaporating fluid.

In crossing the liquid–vapor boundary, we maintain continuity of temperature, i.e.,

Tliq, i ( x = 0 − ) = Tvap, i ( x = 0 + ) = T (0), (3.43)

and energy is conserved at the interface as illustrated in Fig. 3.5. Heat is transferred
from the gas to the liquid surface, Qg−i . Some of this energy goes into heating the
liquid, Qi −l , while the remainder causes the phase change. This energy balance is
expressed as

Q g−i − Q i −l = m (hvap − hliq ) = mh

 fg (3.44)

or

Q net = mh
 fg . (3.45)

Equation 3.45 can be used to calculate the net heat transferred to the interface if the
 is known. Conversely, if Q net is known, the evaporation rate can
evaporation rate, m,
be determined.

Liquid benzene (C6H6), at 298 K, is contained in a 1-cm-diameter glass tube and maintained Example 3.1
at a level 10 cm below the top of the tube, which is open to the atmosphere. The following
properties of benzene are given:

Tboil = 353 K@1 atm,

h fg = 393 kJ/kg @ Tboil ,
MW = 78.108 kg /kmol ,
ρl = 879 kg /m 3 ,
DC H = 0.88 ⋅ 10 −5 m 2 /s@ 298 K.
6 6 − air

A. Determine the mass evaporation rate (kg /s) of the benzene.

B. How long does it take to evaporate 1 cm3 of benzene?
C. Compare the evaporation rate of benzene with that of water. Assume
DH O−air = 2.6 ⋅ 10 −5 m 2 /s.
2

tur80199_ch03_079-[Link] 91 1/4/11 11:58:39 AM

 92 C H A P T E R 3 • Introduction to Mass Transfer

Solution
A. Find m C H .
6 6

Since the configuration given represents the Stefan problem, we can apply Eqn. 3.40:

ρ DC6H6 −air ⎡ 1 − YC H , ∞ ⎤
m C′′ H = ln ⎢ 6 6
⎥.
6 6 L ⎢⎣ 1 − YC6H6 , i ⎥⎦

In the above, D, L, and YC H , ∞ (= 0) are all known; however, we need to evaluate the benzene
6 6
mass fraction at the interface, YC H , i , and an appropriate mean density, ρ , before proceeding.
6 6
From Eqn. 3.42, we know that

MWC H
YC , H = χC H , i 6 6
6 6, i 6 6 MWmix, i

where
Psat (Tliq, i )
χ C6 H 6 , i = .
P
To evaluate Psat /P, we integrate the Clausius–Clapeyron equation, Eqn. 2.19,

dP h fg dT
= ,
P Ru / MWC H T 2
6 6

from the reference state (P = 1 atm, T = Tboil = 353 K) to the state at 298 K, i.e.,

Psat ⎡ h fg ⎛1 1 ⎞⎤
= exp ⎢ − − ⎥.
P (= 1 atm ) ( Ru / MWC H ) ⎝ T Tboil ⎟⎠ ⎥
⎜
⎣⎢ 6 6 ⎦
Evaluating the above,

Psat ⎡ −393, 000 ⎛ 1 1 ⎞⎤

= exp ⎢ ⎜⎝ − ⎟⎠ ⎥
P (= 1 atm ) ⎣ 8315 / 78 . 108) 298 353 ⎦
= exp(−1.93) = 0.145.

So, Psat = 0.145 atm and χ C6H6 = 0.145 . The interface mixture molecular weight is then

MWmix , i = 0.145 ( 78.108 ) + (1 − 0.145)28.85

= 35.99 kg/ kmol,
where the simplified composition of air is assumed. The interface benzene mass fraction is then

78.108
YC H , i = 0.145 = 0.3147.
6 6 35.99
For isothermal, isobaric conditions, we can estimate the mean gas density in the tube by using
the ideal-gas law and the mean mixture molecular weight as follows:

P
ρ= ,
( Ru / MW )T

tur80199_ch03_079-[Link] 92 1/4/11 11:58:39 AM

 Some Applications of Mass Transfer 93

where
1
MW = ( MWmix, i + MWmix, ∞ )
2
1
= (35.99 + 28.85) = 32.42.
2

Thus,
101, 325
ρ= = 1.326 kg /m 3 .
⎛ 8315 ⎞ 298
⎜⎝ ⎟⎠
32.42

We can now evaluate the benzene mass flux (Eqn. 3.40):

1.326(0.88 ⋅ 10 −5 ) ⎡ 1 − 0 ⎤
m C′′ H = ln ⎢
6 6 0.1 ⎣ 1 − 0.3147 ⎦⎥
= 1.167 ⋅ 10 ln(1.459)
-4

= 1.167 ⋅ 10 -4 (0.378) = 4.409 ⋅ 10 −5 kg/s-m 2

and

π D2 π (0.01)2
m C H = m C′′ H = 4.4 09 ⋅ 10 −5 = 3.46 ⋅ 10 −9 kg/s
6 6 6 6 4 4

B. Find the time needed to evaporate 1 cm3 of benzene.

Since the liquid level is maintained, the mass flux is constant during the time to evaporate
the 1 cm3; thus,
mevap ρliqV
t= = ,
m C H m C H
6 6 6 6

879(kg / m 3 )1 ⋅ 10 −6 (m 3 )
t = = 2.54 ⋅ 10 5 s or 70.6 hr
3.46 ⋅ 10 −9 ( kg /s)

C. Find m C H / m H O .
6 6 2
Finding m H O follows the above; however, the problem is simplified since we can use the
2
steam tables (e.g., [7]), to find Psat at 298 K rather than invoking the Clausius–Clapeyron
approximation.
From the steam tables,
Psat (298 K) = 3.169 kPa

so
3169
χ H2O, i = = 0.03128
101,325

and

MWmix, i = 0.03128(18.016) + (1 − 0.03128)28.85 = 28.51.

tur80199_ch03_079-[Link] 93 1/4/11 11:58:40 AM

 94 C H A P T E R 3 • Introduction to Mass Transfer

Thus,

MWH 18.016
2O
YH = χH = 0.03128 = 0.001977.
2 O, i 2 O, i MWmix, i 28.51

The mean molecular weight and mean density in the tube are

1
MW = (28.51 + 28.85) = 28.68
2
101, 325
ρ= = 1.173 kg/m 3,
⎛ 8315 ⎞ 298
⎜⎝ ⎟
28.68 ⎠

where we assume the air outside of the tube is dry.

The evaporation flux is then

1.173(2.6 ⋅ 10 −5 ) ⎡ 1 − 0 ⎤
m H′′ = ln ⎢
2O 0.1 ⎣ 1 − 0.01977 ⎥⎦
= 3.050 ⋅ 10 −4 ln(1.020)
= 3.050 ⋅ 10 −4 (0.01997) = 6.09 ⋅ 10 −6 kg/s-m 2

so,

m C H 4.409 ⋅ 10 −5
6 6
= = 7.2
m H 6.09 ⋅ 10 −6
2O

Comment
Comparing the details of the calculations in parts A and C, we see that the higher vapor pres-
sure of the benzene dominates over the higher diffusivity of the water, thus causing the ben-
zene to evaporate more than seven times faster than the water.

Droplet Evaporation
The problem of the evaporation of a single liquid droplet in a quiescent environment is
just the Stefan problem for a spherically symmetric coordinate system. Our treatment of
droplet evaporation illustrates the application of mass-transfer concepts to a problem of
practical interest. In Chapter 10, we will deal with the droplet evaporation /combustion
problem in greater detail; however, we foreshadow the Chapter 10 development by
introducing the concept of an evaporation constant and droplet lifetimes.
Figure 3.6 defines the spherically symmetric coordinate system. The radius r
is the only coordinate variable. It has its origin at the center of the droplet, and the
droplet radius at the liquid–vapor interface is denoted rs. Very far from the droplet
surface (r → ∞), the mass fraction of droplet vapor is YF, ∞.
Physically, heat from the ambient environment supplies the energy necessary
to vaporize the liquid, and the vapor then diffuses from the droplet surface into the
ambient gas. The mass loss causes the droplet radius to shrink with time until the

tur80199_ch03_079-[Link] 94 1/4/11 11:58:40 AM

 Some Applications of Mass Transfer 95

Figure 3.6 Evaporation of a liquid droplet in a quiescent environment.

droplet is completely evaporated (rs = 0). The problem that we wish to solve is the
determination of the mass flowrate of the vapor from the surface at any instant in
time. Knowledge of this will then enable us to calculate the droplet radius as a func-
tion of time and the droplet lifetime.
To mathematically describe this process completely, the following conservation
laws are required:
Droplet: mass conservation, energy conservation.
Droplet vapor /ambient gas mixture (rs < r < ∞): overall mass conservation,
droplet vapor (species) conservation, and energy conservation.
Thus, we see that for this complete description we need at least five equations.
These equations, in general, take the form of either ordinary or partial differential
equations, depending on the simplifying assumptions applied.

Assumptions For our brief treatment here, we can greatly reduce the number of
unknowns and, hence, the number of equations, by invoking the same assumptions
that we employed in the one-dimensional cartesian problem.
1. The evaporation process is quasi-steady. This means that, at any instant in
time, the process can be described as if it were in steady state. This assumption
eliminates the need to deal with partial differential equations.

tur80199_ch03_079-[Link] 95 1/4/11 11:58:40 AM

 96 C H A P T E R 3 • Introduction to Mass Transfer

2. The droplet temperature is uniform and, furthermore, the temperature is

assumed to be some fixed value below the boiling point of the liquid. In many
problems, the transient heating of the liquid does not greatly affect the droplet
lifetime. Determination of the temperature at the droplet surface depends
on the heat-transfer rate to the droplet. Thus, our assumption of a specified
temperature eliminates the need to apply conservation of energy to the gas
phase surrounding the liquid droplet and to the droplet itself. As will be shown
in Chapter 10, heat-transfer considerations frequently dominate the droplet
evaporation problem.
3. The mass fraction of vapor at the droplet surface is determined by liquid–vapor
equilibrium at the droplet temperature.
4. We also assume that all thermophysical properties—specifically, the rD product—
are constant. Although properties may vary greatly as we move through the gas
phase from the droplet surface to the faraway surroundings, constant properties
allow a simple closed-form solution.

Evaporation Rate With the above assumptions, we can find the mass evaporation
 and the droplet radius history, rs(t), by writing a droplet vapor species con-
rate, m,
servation equation, and a droplet liquid mass conservation equation. From species
conservation, we determine the evaporation rate m,  and thus, knowing m (t ), we can
easily find the drop size as a function of time.
As in the cartesian-coordinate Stefan problem, the species originally in the liq-
uid phase is the species transported, while the ambient fluid is stagnant. Thus, our
previous analysis (Eqns. 3.33–3.35) needs only to be modified to take into account
the change in coordinate system. Overall mass conservation is expressed as

m (r ) = constant = 4π r 2 m ′′, (3.46)

where m ′′ = m A′′ + m B′′ = m A′′, since m B = 0. Note that it is the mass flowrate, not the
mass flux, that is constant. Species conservation for the droplet vapor (Eqn. 3.5)
becomes
dYA
m A′′ = YA m A′′ − ρDAB . (3.47)
dr
Substituting Eqn. 3.46 into Eqn. 3.47, and rearranging to solve for the evaporation
rate m (= m A ), yields

ρDAB dYA
m = − 4π r 2 . (3.48)
1 − YA dr

Integrating Eqn. 3.48 and applying the boundary condition that at the droplet
surface the vapor mass fraction is YA, s, i.e.,

YA (r = rs ) = YA , s , (3.49)

tur80199_ch03_079-[Link] 96 1/4/11 11:58:41 AM

 Some Applications of Mass Transfer 97

yields

(1 − YA , s ) exp[− m / (4πρDABr )] (3.50)

YA (r ) = 1 − .
exp[− m / (4πρDABrs )]

The evaporation rate can be determined from Eqn. 3.50 by letting YA = YA, ∞ for

r → ∞ and solving for m:

⎡ (1 − YA, ∞ ) ⎤
m = 4π rs ρDAB ln ⎢ ⎥. (3.51)
⎢⎣ (1 − YA, s ) ⎥⎦

This result (Eqn. 3.51) is analogous to Eqn. 3.40 for the cartesian problem.
To see more conveniently how the vapor mass fractions at the droplet surface,
YA, s, and far from the surface, YA, ∞, affect the evaporation rate, the argument of the
logarithm in Eqn. 3.51 is used to define the dimensionless transfer number, BY :

1 − YA, ∞ (3.52a)
1 + BY ≡
1 − YA, s
or
YA, s − YA, ∞
BY = . (3.52b)
1 − YA, s

Using the transfer number, BY , the evaporation rate is expressed as

m = 4π rs ρDAB ln(1 + BY ). (3.53)

From this result, we see that when the transfer number is zero, the evaporation rate
is zero; and, correspondingly, as the transfer number increases, so does the evapora-
tion rate. This makes physical sense in that, from the appearance of the mass-fraction
difference YA, s − YA, ∞ in its definition, BY can be interpreted as a “driving potential”
for mass transfer.

Droplet Mass Conservation We obtain the droplet radius (or diameter) history
by writing a mass balance that states that the rate at which the mass of the droplet
decreases is equal to the rate at which the liquid is vaporized; i.e.,

d md (3.54)
= − m ,
dt
where the droplet mass, md , is given by

md = ρl V = ρl π D 3/6, (3.55)

and V and D (= 2rs) are the droplet volume and diameter, respectively.

tur80199_ch03_079-[Link] 97 1/4/11 11:58:41 AM

 98 C H A P T E R 3 • Introduction to Mass Transfer

Substituting Eqns. 3.55 and 3.53 into Eqn. 3.54 and performing the differentia-
tion yields

dD 4ρDAB
=− ln(1 + BY ). (3.56)
dt ρl D

In the combustion literature, however, Eqn. 3.56 is more commonly expressed in

terms of D2 rather than D. This form is
dD2 8 ρDAB
=− ln(1 + BY ). (3.57)
dt ρl

Equation 3.57 tells us that the time derivative of the square of the droplet diameter
is constant; hence, D2 varies linearly with t with the slope −(8ρDAB /ρl ) ln(1 + BY ), as
illustrated in Fig. 3.7a. This slope is defined to be the evaporation constant K:

8ρDAB (3.58)
K= ln(1 + BY ).
ρl

Figure 3.7 The D2 law for droplet evaporation. (a) Simplified analysis. (b) Experimental
data from Ref. [8] for water droplets with T∞ = 620°C.
Reprinted by permission of The Combustion Institute.

tur80199_ch03_079-[Link] 98 1/4/11 11:58:42 AM

 Some Applications of Mass Transfer 99

We can use Eqn. 3.57 (or 3.56) to find the time it takes a droplet of given initial size
to evaporate completely; i.e., the droplet lifetime, td. Thus,
0 td

∫ d D2 = − ∫ K dt, (3.59)
D02 0

which yields

td = D02 /K . (3.60)

We can change the upper limits in Eqn. 3.59 to provide a general relationship
expressing the variation of D with time (t):

D 2 (t ) = D02 − Kt. (3.61)

Equation 3.61 is referred to as the D2 law for droplet evaporation. Experimental data
show that the D2 law holds after an initial transient period as shown in Fig. 3.7b.
The D2 law is also used to describe burning of fuel droplets, which is discussed in
Chapter 10.

In mass-diffusion-controlled evaporation of a fuel droplet, the droplet surface temperature is Example 3.2
an important parameter. Estimate the droplet lifetime of a 100-µm-diameter n-dodecane drop-
let evaporating in dry nitrogen at 1 atm if the droplet temperature is 10 K below the dodecane
boiling point. Repeat the calculation for a temperature 20 K below the boiling point, and
compare the results. For simplicity, assume that, in both cases, the mean gas density is that of
nitrogen at a mean temperature of 800 K. Use this same temperature to estimate the fuel vapor
diffusivity. The density of liquid dodecane is 749 kg /m3.
Given: n-dodecane droplet

D = 100 μm ,
P = 1 atm ,
ρl = 749 kg /m 3 ,
Ts = Tboil − 10 (or 20),
ρ = ρN2 @T = 800 K.

Find: droplet lifetime, td .

Solution
We can estimate the droplet lifetimes by employing Eqn. 3.60, after calculating the evaporation
rate constants, K, from Eqn. 3.58. Evaluation of properties will be important in our solution.
Properties required:

Tboil = 216.3°C + 273.15 = 489.5 K (Appendix Table B.1),

h fg = 256 kJ / kg (Appendix Table B.1),
MWA = 170.337 kg /kmol,
DAB = 8.1 ⋅ 10 −6 m 2 /s @ 399 K (Appendix Table D.1).

tur80199_ch03_079-[Link] 99 1/4/11 11:58:42 AM

 100 C H A P T E R 3 • Introduction to Mass Transfer

We start by calculating BY , which requires knowledge of the fuel mass fraction at the sur-
face. As in Example 3.1, we integrate the Clausius–Clapeyron equation to find the saturation
pressure at the given droplet surface temperature. For T = Tboil − 10 (= 479.5 K),

Psat ⎡ −256,000 ⎛ 1 1 ⎞⎤
= exp ⎢ ⎜⎝ − ⎟ ⎥ = 0.7998
P (= 1 atm) ⎣ (8315/170.337) 479.5 489.5 ⎠ ⎦

so Psat = 0.7998 and χ A (= χ dodecane ) = 0.7998. We employ Eqn. 2.11 to calculate the fuel mass
fraction at the surface:
170.337
YA, s = 0.7998 = 0.9605.
0.7998(170.337) + (1 − 0.7998)28.014

We now evaluate the transfer number BY (Eqn. 3.52b):

YA , s − YA, ∞ 0.9605 − 0
BY = = = 24.32.
1 − YA, s 1 − 0.9605

To evaluate the evaporation constant, we need to estimate rDAB, which we treat as

ρ N DAB (T = 800 K). Extrapolating the tabulated value to 800 K using Eqn. 3.18b gives
2

800 ⎞ 3/ 2
DAB (T ) = 8.1 ⋅ 10 −6 ⎛⎜ = 23.0 ⋅ 10 −6 m 2 /s,
⎝ 399 ⎟⎠

and using the ideal-gas law to evaluate ρ N :

2

101,325
ρ N2 = = 0.4267 kg /m 3 .
(8315 / 28.014)800

Thus,
8 ρ DAB
K= ln (1 + BY )
ρl
8(0.4267)23.0 ⋅ 10 −6
= ln(1 + 24.32)
749
= 3.39 ⋅ 10 −7 m 2 /s,
and the droplet lifetime is

(100 ⋅ 10 −6 )2
td = D 2/ K =
3.39 ⋅ 10 −7
td = 0.030 s

Repeating the calculations for T = Tboil − 20 = 469.5 K, we can compare the various parameters:

ΔT T Psat K td
(K) (K) (atm) Ys BY (m2/s) (s)

10 479.5 0.7998 0.9605 24.32 3.39 · 10−7 0.030

20 469.5 0.6336 0.9132 10.52 2.56 · 10−7 0.039

tur80199_ch03_079-[Link] 100 1/4/11 11:58:42 AM

 Nomenclature 101

From the above table, we see that about a 2 percent change in the droplet surface temperature
results in a 30 percent increase in the droplet lifetime. This large effect of temperature mani-
fests itself in the BY term, where the denominator, 1 − YA, s , is clearly sensitive to temperature
when YA, s is large.
Comments
Evaporation of fuel droplets at elevated temperatures is important for many practical devices,
particularly gas-turbine engines and diesel engines. In these devices, evaporation occurs when
droplets are injected into hot compressed air or into zones where combustion is already in
progress. In spark-ignition engines employing fuel injection, temperatures in the intake system
are typically much closer to ambient levels and pressures are subatmospheric. In most cases,
forced-convection effects are important in the evaporation process. These effects will be dis-
cussed in Chapter 10, which focuses on droplet combustion.

SUMMARY
In this chapter, you were introduced to the concept of mass diffusion or mass trans-
fer, and Fick’s law governing the rate of species mass transfer in a binary mixture
was presented. A new transport property, the mass diffusivity, provides the propor-
tionality between the species diffusional flux and the species concentration gradi-
ent, analogous to the kinematic viscosity and thermal diffusivity in momentum and
heat transfer, respectively. You should be familiar with the physical interpretation of
Fick’s law and how to apply it to simple binary systems. As an example of simple
binary diffusion, the Stefan problem was developed. From this example, you saw
how overall mass and species mass conservation equations apply to a single species
diffusing through a stagnant layer of a second species. With application of appropri-
ate boundary conditions, we were able to solve for the liquid evaporation rate. A
similar analysis was performed for the spherically symmetric problem of droplet
evaporation. You should have some physical appreciation for this problem, as well as
being familiar with how to calculate droplet evaporation rates and droplet lifetimes.

NOMENCLATURE
a Defined in Eqn. 3.10d
A Area (m2)
BY Dimensionless transfer number or Spalding number, Eqn. 3.52
c Molar concentration (kmol / m3)
D Diameter (m)
DAB Binary diffusivity or diffusion coefficient (m2/s)
h Enthalpy (J / kg)
hfg Heat of vaporization (J / kg)
k Thermal conductivity (W/ m-K)
kB Boltzmann constant (J/ K)
ke Kinetic energy (J)
K Evaporation constant, Eqn. 3.58 (m2/s)

tur80199_ch03_079-[Link] 101 1/4/11 11:58:43 AM

 102 C H A P T E R 3 • Introduction to Mass Transfer

L Tube height (m)

m Mass (kg)
m Mass flowrate (kg /s)
m ′′ Mass flux (kg / s-m2)
MW Molecular weight (kg / kmol)
n Number of molecules
N Number of moles
N Molar flowrate (kmol / s)
N ′′ Molar flux (kmol / s-m2)
P Pressure (Pa)
Q Heat-transfer rate (W)
Q ′′ Heat flux (W/ m2)
r Radius (m)
Ru Universal gas constant (J / kmol-K)
t Time (s)
T Temperature (K)
v Average molecular speed (m / s)
V Volume (m3)
x Cartesian coordinate (m)
y Cartesian coordinate (m)
Y Mass fraction (kg / kg)
Z ′′ Molecular collision frequency per unit area (no. /m2-s)
Greek Symbols
l Molecular mean free path (m)
r Density (kg / m3)
s Molecular diameter (m)
c Mole fraction (kmol / kmol)
Subscripts
A Species A
B Species B
boil Boiling point
cv Control volume
d Droplet
diff Diffusion
evap Evaporation
g Gas
i Interface
l, liq Liquid
mix Mixture
s Surface
sat Saturation
tot Total
vap Vapor
∞ Freestream or far removed from surface

tur80199_ch03_079-[Link] 102 1/4/11 11:58:43 AM

 Review Questions 103

REFERENCES
1. Bird, R. B., Stewart, W. E., and Lightfoot, E. N., Transport Phenomena, John Wiley &
Sons, New York, 1960.
2. Thomas, L. C., Heat Transfer–Mass Transfer Supplement, Prentice-Hall, Englewood
Cliffs, NJ, 1991.
3. Williams, F. A., Combustion Theory, 2nd Ed., Addison-Wesley, Redwood City, CA,
1985.
4. Kuo, K. K., Principles of Combustion, 2nd Ed., John Wiley & Sons, Hoboken, NJ, 2005.
5. Pierce, F. J., Microscopic Thermodynamics, International, Scranton, PA, 1968.
6. Daniels, F., and Alberty, R. A., Physical Chemistry, 4th Ed., John Wiley & Sons, New
York, 1975.
7. Irvine, T. F., Jr., and Hartnett, J. P. (eds.), Steam and Air Tables in SI Units, Hemisphere,
Washington, 1976.
8. Nishiwaki, N., “Kinetics of Liquid Combustion Processes: Evaporation and Ignition Lag
of Fuel Droplets,” Fifth Symposium (International) on Combustion , Reinhold, New York,
pp. 148–158, 1955.

REVIEW QUESTIONS
1. Make a list of all of the boldfaced words in Chapter 3. Make sure you
understand the meaning of each.
2. Assuming constant properties, recast Eqn. 3.4 into a form where the
constant of proportionality in Fourier’s law is the thermal diffusivity,
a = k / rcp, rather than k. The gradient (or spatial derivative) of what prop-
erty now appears?
3. Starting with the sketch in Fig. 3.3, derive the species conservation equa-
tion (Eqn. 3.30) without reference to the text derivation.
4. Using words only, explain what is happening in the Stefan problem.
5. How is the evaporating droplet problem similar to the Stefan problem?
6. If no heat is supplied to the liquid A (Fig. 3.4), what happens to the tem-
perature of the liquid? Can you write a simple energy conservation equa-
tion to justify your answer?
7. Define a transfer number, BY , that can be used with Eqn. 3.40. Further-
more, assume the flow area is A, and write an expression for m in terms
of fluid properties, geometrical parameters, and BY . Compare your result
with Eqn. 3.53 and discuss.
8. Why, and in what way, does the presence of species A in the free stream
(Fig. 3.4) affect the evaporation rate?

tur80199_ch03_079-[Link] 103 1/4/11 11:58:43 AM

 104 C H A P T E R 3 • Introduction to Mass Transfer

9. How does the mass average velocity vary with distance from the surface of
an evaporating droplet at any instant in time?
10. Explain what is meant by “quasi-steady” flow.
11. Starting with Eqns. 3.53 and 3.54, derive the “D 2 law’’ for droplet evaporation.

PROBLEMS
3.1 Consider an equimolar mixture of oxygen (O2) and nitrogen (N2) at 400 K
and 1 atm. Calculate the mixture density r and the mixture molar concen-
tration c.
3.2 Calculate the mass fractions of O2 and N2 in the mixture given in
problem 3.1.
3.3 Estimate the value of the binary diffusivity of n-octane in air at a
temperature of 400 K and pressure of 3.5 atm utilizing the value given in
Appendix D. Also, compare the ratio Dref / D (T = 400 K, P = 3.5 atm) with
the ratio of the rD products, i.e., (rD)ref /(rD) (T = 400 K, P = 3.5 atm).
3.4 Equation 3.18a was derived assuming that molecules A and B are essen-
tially the same mass and size. Reference [1] indicates that Eqn. 3.18a can
be generalized for the case where mA ≠ mB and sA ≠ sB by using

mA mB 1
m= and σ = (σ A + σ B ).
(mA + mB ) / 2 2

With this information, estimate the binary diffusivity for O2 in N2 at 273 K

for σ O = 3.467 and σ N 2 = 3.798 Å. Compare your simple estimate with the
2
handbook value of 1.8 · 10−5 m2/s. Should you expect good agreement? Why
not? Note: You will need to use Avogadro’s number 6.022 · 1026 molecules/
kmol to calculate the mass of a molecule.
3.5 Consider liquid n-hexane in a 50-mm-diameter graduated cylinder. The
distance from the liquid–gas interface to the top of the cylinder is 20 cm.
The steady-state n-hexane evaporation rate is 8.2 · 10−8 kg /s and the
n-hexane mass fraction at the liquid–air interface is 0.482. The diffusivity
of n-hexane in air is 8.0 · 10−4 m2/s.
A. Determine the mass flux of n-hexane vapor. Give units.
B. Determine the bulk flux of n-hexane vapor, i.e., that portion associ-
ated with the bulk flow, at the liquid–gas interface.
C. Determine the diffusional f lux of n-hexane vapor at the liquid–gas
interface.
3.6 Consider water in a 25-mm-diameter test tube evaporating into dry air
at 1 atm. The distance from the water–air interface to the top of the

tur80199_ch03_079-[Link] 104 1/4/11 11:58:44 AM

 Problems 105

tube is L = 15 cm. The mass fraction of the water vapor at the water–air
interface is 0.0235, and the binary diffusivity for water vapor in air is
2.6 · 10−5 m2/s.
A. Determine the mass evaporation rate of the water.
B. Determine the water vapor mass fraction at x = L / 2.
C. Determine the fraction of the water mass flow that is contributed by
bulk flow and the fraction contributed by diffusion at x = L / 2.
D. Repeat part C for x = 0 and x = L. Plot your results. Discuss.
3.7 Consider the physical situation described in problem 3.6, except that the
interface vapor mass fraction is unknown. Find the mass evaporation
rate of the water when the liquid water is at 21°C. Assume equilibrium at
the interface, i.e., PH 2O (x = 0) = Psat (T ). The air outside the tube is dry.
3.8 Repeat problem 3.6 when the air outside the tube is at 21°C and has a
relative humidity of 50 percent. Also, determine the rate of heat transfer
required to maintain the liquid water at 21°C.
3.9 Consider liquid n-hexane in a 50-mm-diameter graduated cylinder. Air
blows across the top of the cylinder. The distance from the liquid–air
interface to the open end of the cylinder is 20 cm. Assume the diffusivity
of n-hexane is 8.8 · 10−6 m2/s. The liquid n-hexane is at 25 °C. Estimate the
evaporation rate of the n-hexane. (Hint: Review the Clausius–Clapeyron
relation as applied in Example 3.1.)
3.10 Calculate the evaporation rate constant for a 1-mm-diameter water droplet
at 75 °C evaporating into dry, hot air at 500 K and 1 atm.
3.11 Determine the influence of the ambient water vapor mole fraction on the
lifetimes of 50-µm-diameter water droplets. The droplets are evaporating
in air at 1 atm. Assume the droplet temperature is 75 °C and the mean air
temperature is 200 °C. Use values of χ H O, ∞ = 0.1, 0.2, and 0.3.
2

3.12 Consider the two general forms of Fick’s law presented in this chapter:

 A′′ = YA (m
m  A′′ + m
 B′′ ) − ρDAB ⵱YA (3.5)

N A′′ = χ A ( N A′′ + N B′′ ) − cDAB ⵱χ A, (3.6)

where c is the molar concentration (= r /MWmix).

The first expression (Eqn. 3.5) represents the mass flux of species A
(kgA/s-m2) relative to a stationary laboratory reference frame, and the
second is an equivalent expression for the molar flux of species A
(kmolA/s-m2), also relative to a stationary laboratory reference frame.
The first term on the right-hand-side of Eqns. 3.5 and 3.6, respectively,
is the bulk transport of species A at the mass-average velocity, V (for
Eqn. 3.5) and the molar-average velocity, V* (for Eqn. 3.6). You are used
to working with the mass-average velocity as the velocity of choice in

tur80199_ch03_079-[Link] 105 1/4/11 11:58:44 AM

 106 C H A P T E R 3 • Introduction to Mass Transfer

fluid mechanics. Both of these velocities are relative to our fixed labora-
tory reference frame. Useful relationships associated with these two ve-
locities are the following [1]:

 A′′ + m
m  B′′ = ρV (I)

or
ρYA vA + ρYBvB = ρV (II)

and
N A′′ + N B′′ = cV * (III)

or

cχ A vA + cχ BvB = cV * , (IV)

where vA and vB are the species velocities relative to the fixed frame.
The second term on the right-hand side of Eqns. 3.5 and 3.6, respectively,
expresses the diffusion flux of species A relative to the mass-average
velocity V (for Eqn. 3.5) and relative to the molar-average velocity V * (for
Eqn. 3.6).
Use the above relations (Eqns. I–IV), and others that you may need,
to transform the 1-D planar form of Eqn. 3.6 to the 1-D planar form of
Eqn. 3.5. Hint: There are a few elegant ways to do this, as well as tedious
brute-force approaches.
3.13 Rederive the simple Stefan problem result (Eqn. 3.40) using the molar
flux form of Fick’s law (Eqn. 3.6) with boundary conditions expressed in
mole fractions. Note that the molar concentration c (= P/RuT ) is constant
for an ideal gas, unlike the density r, which depends on the mixture mo-
lecular weight. Leave the final result in terms of the mole fractions.
3.14 Use the result from problem 3.13 to solve part A of Example 3.1 in the
text. How does this result for m C H compare with the text’s result? Which
6 6
is more accurate? Why?

tur80199_ch03_079-[Link] 106 1/4/11 11:58:44 AM