The Catalytic Effectiveness Factor Under Nonisothermal Conditions

J. J. CARBERRY
E. 1. du Pont de Ne mours ond Compony., Wilmington, Delowore
The effectiveness factor for a porous- The work reported in this communi- The effectiveness factor is defined by
solid-catalyzed reaction is a measure of cation embraces both first-and second- D dC/dx df / d z
the extent to which species transport order irreversible catalytic reactions. - --
by Knudsen diffusion within the par- Pellet geometry is riot specified, since
9=
L k, Con +""
ticle retards the over-all rate process. the model is written for a single pore. Equations ( 2 ) and ( 3 ) were solved
The literature reveals correlations based Aris (1) has shown that a proper defini- simultaneously with an IBM-650 com-
upon analytical solution of the appro- tion of particle size ( L = l/a) renders puter.
priate equations when isothermal con- the isothermal effectiveness factor vir- The results of computation are shown
ditions prevail throughout the catalyst tually independent of particle shape, in Figures 1 and 2 in terms of the
particle. Following Thiele's original while Prater's derivation ( 2 ) proves classical reaction-diffusion modulus, +o,
work ( 5 ) ,Wheeler (8) and Weisz (7) that AT is also independent of geome- over a range of reaction exo and endo-
have made significant contributions to try. Thus it is assumed here that the thermicity (2a ) , where
the discussion of the isothermal effec- single-pore treatment (equivalent to
tiveness factor. flat-plate geometry) is of general ap-
The problem of reaction heat release plicability.
or abstraction within a catalyst particle The following dimensionless equa- Figures 1 and 2 are based upon fluid
has been recognized as a potentially tions are assumed to apply for kinetics film coefficients of heat and mass equal
serious one with respect to the influence of order n, where, in the case of second-to infinity relative to the equivalent
of the resulting temperature gradient order reaction, equality of species dif- transport coefficients within the porous
upon the effectiveness factor for both fusivities is assumed. For mass pellet. By definition of AH positive Val-
reactant and product. However analyti- ues of 01 signify exothermicity. The re-
cal solution of the governing differential lation shown for a = 0 corresponds to
equations presents marked difficulties. the classical isothermal effectiveness
Wheeler (8) and more recently Prater factor (1, 5 ) .
(2) have shown that for a simple first- An effectiveness greater than unity
order irreversible reaction the maximum for positive values of a is a consequence
temperature difference between the ex- of significant rate enhancement, due to
ternal and interim surface of the pellet a temperature rise within the pellet, in
is given by a region where the isothermal effective-
ness is close to unity. In other words
J d f / d z becomes greater than + .: This
where result emphasizes the inadequacy of the
C , AHD traditional notion of catalytic effective-
Valuable as the above relation is for p = I - __-. , k = k, exp
purposes of estimating the maximum To K ness in which 7 is understood to be the
possible temperature gradient, the pre-
cise influence of the temperature gradi-
ent upon effectiveness remained, until and
[
-E/RT, (t
- l)]
J
ratio of actual surface area utilization to
(4) the total area of the catalyst.
The parameter 01 does not uniquely
recently, undefined. govern the nonisothermal relationship
Schilson and Amundson ( 3 ) were the between 7 and $., for $" < 2, as may
first to undertake the task of solving the be seen in Table 1, where computed
nonisothermal problem analytically. The The boundary conditions, in terms of values of are shown over a range of
heat-generation functions for a single the bulk phase Nusselt numbers for heat 4H and E at constant OL and for values
reaction and for two simultaneous re- and mass transport, (N,,,) and ( N N % ) of $" of 0.85 and 0.7. Obviously a more
1,,,

actions were approximated by single respectively, are complex parameter exists which would
linear functions and also by two linear assemble the computed data more con-
relationships. Analytical solutions were cisely; however in view of the limited
then found by Schilson and Amundson accuracy with which the reaction and
z=1 j diffusion coefficients are generally
in the form of an integral equation ex- I 1
pressing temperature as a function of known the indicated manner of corre-
pellet radius. Specific cases were solved lation is deemed adequate. For > +.
in terms of the effectiveness factor, the z = 0 d f / d z = dy/dz = 0
details of which can be found in their TABLE2. INFLUENCE OF BULKHEAT AND

forthcoming publication ( 4 ) . M A ~ TRANSFER

S UPON q
TABLE1. INFLUENCE OF AH AND E UPON
Tinkler and Pigford (6) applied per- q AT CONSTANT VALUEOF 01 a=2 = 0.83 *
turbation techniques to obtain an ap- ( N N , ) ~=, 00 ( N X ~=
~ )00~ , ~
-AH, E,
proximate analytical expression for the kcal./ kcal./ AT,
(Na,),, 9 (N,t,)r, 9
nonisothermal effectiveness factor in mole mole "C. rl O0 1.17 x 1.17
terms of the heat-generation function .I. = 0.85 15 20 10.3 0.924 500 1.16 100 1.2
and the Nusselt number for heat trans- a = 1.0 30 10 20.5 0.922 50 1.13 10 1.4
fer at the external surface. Their at- 60 5 41 0.918 5 0.7 5 1.8
tention was fociised upon a first-order ' = 0.85
q 180 5 197 1.39 1 10
irreversible reaction catalyzed by a O1 = 3 18 50 28 1.87 ( Na,, ) h ( N\u ) "6 r
spherical pellet. 9J = 0.7 30 40 53 3 10 50 1.33
- a = 4 60 20 92 2.6 10 5 0.9
J. J. Carhrrry is at the Univrnity of Notre T , = 300°C. 1 5 2.0
Dame. Notre Dame, Indiana.

Page 350 A.1.Ch.E. Journal June, 1961

 in. IU, I I 1

?I

0.I

".".0.I
n ni
I 10 I00
I
I( 30
0.I I 100 1000

- I

Fig. 2. Nonisothermal catalytic effectiveness factor (second-

Fig. 1. Nonisothermal catalytic effectiveness factor. order reaction).

2.5 the effectiveness factor for a first- A suggested criterion for determin- AH reaction enthalpy change
=
order exothermic reaction is given by ing whether bulk transport can be con- k, mass transfer coefficient, cm./
=
sidered negligible relative to surface- sec.
reaction kinetics is k = sate constant, sec:' (moles/
Tk cc. ) 47L-U
-5 0.1 (7) k = over-all rate coefficient, Equa-
Of significant importance is the k, a tion (6)
influence upon 7 of finite heat and A number of consequences may be K = thermal conductivity of par-
mass transfer resistances at the ex- inferred from the results presented. ticle, cal./"C. cm. sec.
ternal surface of the particle. In Table While it might appear that a high ef- L = characteristic particle radius,
2 a few results are shown which illus- fectiveness factor is desirable (v > l ) , cm .
trate the influence of both bulk mass it must be realized that high values of n = reaction order
and heat transfer upon the noniso- 7 resulting from large values of -AH ( N N u ) h= Nusselt number for heat
thermal effectiveness factor. The Nus- are associated with considerable tem- transfer, h L / R
selt numbers are defined: perature gradients. When the reaction ( N N , , ) , = Nusselt number for mass
hL product is heat sensitive, degradation is transfer, k , L / D
)I = -
for heat, ( NNu) enhanced under such circumstances. In R = gas constant, cal./mole, O K .
K addition a high effectiveness factor due T = temperature, O K .
to exothermicity may ultimately render x = distance variable, cm.
conductivity of fluid film the catalyst inactive by virtue of sinter- y = dimensionless temperature,
conductivity of solid ing or site deactivation via phase and/or T/T.
chemical transformation. z = dimensionless distance, x / L
k, L
=-
for mass, (NN,Jm An appreciation of the influence of (Y = thermal parameter, Equation
D intraparticle exothermicity upon yield (5)
may be gathered from Table 3, where ii? = heat generation term, Equa-
diffusivity in fluid 6lm the yield of product B is shown as a tion (4)
diffusivity in solid function of (Y for the reaction sequence 7 = catalyst effectiveness factor
A-*B+C. $ = diffusion, reaction modulus,
An assessment of the bulk mass With respect to measurement of ac- Equation (4)
transport contribution is readily made, tivation energy it becomes apparent Subseriptr
in the case of first-order kinetics, by that, whereas diffusion retardation leads o = bulk-phase conditions
equating the surface rate to that of bulk to an observed activation energy of one- LITERATURE CITED
mass transport. There results an over- half the true value under isothermal 1. Aris, Rutherford, Chem. Eng. Sci., 6,
all rate coefficient given by conditions ( 8 ) , the opposite effect be- 262 ( 1957).
comes manifest for 7 >1; that is the 2. Prater, C. D., Chem. Eng. Sci., 8, 284
apparent measured activation energy ( 1958).
will be greater than the true value. 3. Schilson, R. E., and N. R. Amundson,
Paper presented at A.1.Ch.E. Meeting,
NOTATION St. Paul, Minn. (Sept., 1959).
a = external surface to volume ratio 4. Schilson, R. E., and N. R. Amundson,
TABLE
3. INFLUENCEOF HEATRELEASE
Chem. Eng. Sci., to be published.
of .a particle, cm.-l 5. Thiele, E. W., Znd. Eng. Chem., 31,
UPON YIELD C = species concentration, moles/ 916 (1939).
cc. 6 . Tinkler, John, and R. L. Pigford, Chem.
k, k, D = Knudsen diffusivity, sq. cm./
A-+B-+C Eng. Sci., to be published.
*1= 1.0 = 0.5
'I.? ( A / B ) e = 10 sec. 7. Weisz, P. B., Z. Physik. Chem., Neue
a1 a2 Yield of B% E = activation energy, cal./mole Folge, 11, 1 (1957).
0 0 93 f = dimensionless concentration, 8. Wheeler, Ahlborn, "Catalysis," Vol. 11,
c/c. Reinhold, New York ( 1955).
0.5 0.5 92
1.o 1.o 89 h = heat transfer coefficient, bulk Pnnrr nreseited at Gordon Conference on
Catalysis, New London, New Hampshire, August,
2.0 2.0 70 phase, cal./sq. cm. sec., "C. 1960.

Vol. 7, No. 2 A.1.Ch.E. Journal Page 351