CATALYTIC DEHYDROGENATION OF 2-PROPANOL TO ACETONE

S . S . L O K R A S , P . K . D E S H P A N D E , N. R . K U L O O R
A N D
Department of Chemical Engineering, Indian Institute of Science, Bangalore 12, India

Catalytic dehydrogenation of 2-propanol over Cu-Si02 catalyst was investigated.

The undesired side reaction of dehydration can be controlled by a selective
catalyst and choice of proper operating conditions. The kinetics of the heterogene-
ous catalytic reaction can be adequately expressed by a forward first-order
and reverse second-order mechanism. The rate-controlling step with chemically
pure 2-propanol is single-site surface reaction, while for the technical grade
alcohol the adsorption of alcohol is rate-controlling. The static bed data are
compared with the fluidized bed data.

ACETONE
is one of the important industrial solvents and
a raw material for synthesizing diacetone alcohol, iso-
phorone, methyl acrylate, bisphenol A, ketone, etc.
Dehydrogenation of 2-propanol displaced the fermentation
process in commercial manufacture of acetone in the mid-
dle 1920’s. With the advent of petrochemical industries,
2-propanol is now available in abundance a t competitive
prices and hence is a more suitable raw material.
Acetone can be obtained from 2-propanol by either
catalytic dehydrogenation or catalytic oxidation. The reac-
tions are
CHiCHOHCH3- CHlCOCHi + H? (1)
CHiCHOHCHi + ;1. O?+ CH7COCHl + H20 ( 2 )
Reaction 1 is endothermic, while Reaction 2 is exothermic
and requires more critical temperature control. Therefore,
industrial practice generally favors the dehydrogenation
route. During catalytic dehydrogenation, dehydration of
an alcohol occurs simultaneously, according to the reaction,
C3Hs + H20
CH?CHOHCH,--+ (3)
However, Reaction 3 can be controlled by selective
catalysts and proper operating conditions.
literature Survey
Extensive work has been carried put on the dehydro-
genation of 2-propanol. Balandin and coworkers (Balandin,
1958) studied the various catalysts in detail, to explain
the phenomena of catalysis in the light of the multiplet
theory and indicated the fundamental nature of bond
energy compared to the energy of activation. A large
number of patents deal with catalyst specifications and
optimum operating parameters. Kinetic work is com-
paratively less. Ciborowski (1959) and Balandin and Tet-
eni (1960) collected the kinetic data over copper catalysts,
while Germain et al. (1961, 1965), Garcia de la Banda
and Orlandini (1958, 1963), Garcia de la Banda (1960),
Sheely (1963), Krylov (1965), and Zhabrova et al. (1965)
worked with zinc oxide catalysts. Kinetics with chromium
catalysts were investigated by Garcia de la Banda (1960,
19621, Garcia de la Banda et al. (1966), and Orlandini
(1959). Liquid-phase reaction kinetics with Raney nickel
catalysts were studied by Kipperman and Nikolaeva
(1961) and Kipperman et al. (1962, 1963a,b). Rate-
controlling steps and reaction mechanism by the Hougen-
Watson approach were also investigated by Garcia de
la Banda et al.
Experimental
The apparatus (Figure 1) consists of 2-propanol feeding
and metering units, vaporizer with preheater, reactor, pro-
duct condensers, and receivers, and gas collecting drum.
The vaporizer consists of a 4-foot-long, 78-inch-i.d. 18-8
stainless steel tube. The reactor is a 2-inch-i.d., 18-inch-
long, 18-8 stainless steel tube having a conical bottom
with a plug cock, a 4-inch-diameter, 4-inch-high disen-
gaging section a t the top, and two pressure taps a t the
top and bottom. The heating is done electrically and
the temperatures are controlled by Sunvic energy regula-
tors and autotransformers. Reactor temperature is
measured by a Chromel-Alumel thermocouple and con-
trolled within i l ° C .
Analysis. Hydrogen and propylene in the collected gas
are determined in an Orsat apparatus and acetone in
the condensed product by Morasco’s Oxime method
(Mitchell et al., 1955). The liquid product was collected
at -20°C. by using a salt-ice mixture. As the vapor pres-
sure of acetone at -20°C. is 20 mm. of Hg, the carry-
over of acetone by the product gases was taken into
account. Since it was observed that moles of acetone
formed are equal to the moles of H1 collected, the only
side reaction being that of dehydration, only hydrogen
and propylene were estimated for routine analysis.
Catalyst Preparation. Catalysts, Cu-Si02 (1 t o 3, 1 to
4, 1 to 9, 1 to 19, and 1 to 99), Cu-pumice (1 to 4),
ZnO-SiOi (1 to 4), and Ni-Si02 (1 to 4) were prepared
by an impregnation technique. The impregnated nitrates
of copper and nickel and zinc acetate were oxidized a t
400°C. in a muffle furnace. The copper and nickel oxides
were reduced with hydrogen at 3OO0C. in the reactor
before being used. The size of the catalyst used in most
of the runs was -60 +lo0 Tyler mesh.

Ind. Eng. Chem. Process Des. Develop., Vol. 9, No. 2, 1970 293
 Feed bottla i
I ,
Flowmatar
Needla
valve
m "!
-
supor healer
Figure 1 . Flow diagram of setup for dehydrogenation of isopropyl alcohol to acetone
Catalyst Characteristics. The characteristics of Cu-Sios
(1 to 4, -60 +lo0 Tyler mesh) catalyst, which was found
to be the best among the lot, are surface area, 310.1
sq. cm. per gram, apparent density, 2.4818 grams per
cc., and bulk density, 0.8750 gram per cc.
Selection of Catalyst and Carrier Material. Pumice-
supported catalyst had low activity and short life and
hence silica gel was selected as a carrier. The performance
of catalysts Cu-Sios, Ni-SiO:!, and ZnO-SiOn, all in the
ratio of 1 to 4 by weight and having a size of -60 +lo0
Tyler mesh, was compared. The experiments were con-
ducted with a 4-inch bed height in the temperature range
of 275" to 325°C. and WIF range of 16 to 22. The copper
catalyst gave a much better performance than nickel and
zinc oxide. Conversions with nickel were rather low, and
with zinc oxide the dehydration reaction was more prom-
inent. Catalyst activity studies show that the activity
of Cu-Si02 (1 to 4) remains constant for 24 hours. All
the experiments were carried out under constant activity
conditions. Xoncatalytic conversion was not observed.
Some typical results a t 3OO0C. are presented in Table
I.
The catalyst performance can be partly explained in
the light of the energy correspondence principle, which
forms a part of the multiplet theory of catalyst activity
of Balandin (1958). Based on the bond energies data
of Cotrell (1958), Balandin et al. (1963), and Balandin
and Tolstopyatova (1965) for three catalysts, the activa-
tion energies for the dehydrogenation and dehydration
Table 1. Comparison of Cu-Sios, Ni-SiOs,
and ZnO-Si02 Catalysts
% Conversion at 300" C.
Catalyst Dehydro-
(1 :4) W F genation Dehydration
Ni-SiO, 22.2 20.4 2.0
ZnO-SiOP 24.2 23.6 58.8
Cu-SiO.! 24.0 88.4 3.2
294 Ind. Eng. Chem. Process Des. Develop., Vol. 9, No. 2, 1970
T h e r m coude
I
8 %
Reac tor
I
Condensers
Pi
and
receivers hx
liquid product
Gas collection system
of 2-propanol over Ni, Cu, and ZnO were computed (Table
11).
The activation energy for dehydrogenation is least for
copper, which indicates that it is a better dehydrogenation
catalyst. For nickel dehydration is less prominent than
dehydrogenation, while for zinc oxide the two reactions
seem to be probable to an equal extent. For copper
catalyst, the activation energy for dehydration is lower
than for dehydrogenation. However, experiments with cop-
per show lesser dehydration with increasing copper con-
tent. On the basis of those results, the copper catalyst
was selected for detailed experimentation.
Copper-Catalyst Compositions (1 to 3, 1 to 4, 1 to 9,
1 to 19, and 1 to 99) were investigated. Optimum values
of dehydrogenation and dehydration were obtained with
1 to 3 and 1 to 4 catalyst composition and hence 1
to 4 catalyst was chosen for further studies.
Results and Discussion
The experimental runs with Cu-Si02 (1 to 4, -60 +lo0
Tyler mesh) catalysts were carried out in the static and
fluidized beds. Chemically pure 2-propanol was used.
Technical grade 2-propanol was also used in the fluidized
beds.
Static Bed Reactor. The experimental data are presented
in Table 111. The range of temperature and time factor
( W I F ) employed was 250" to 325°C. and 16 to 30, respec-
tively. The extent of dehydration reaction was rather
low, the maximum a t 325°C. The conversion to acetone
increased with increase in time factor and the curves
of conversion us. time factor tend toward asymptotic values
Table II. Computed Values of Activation Energies for
Dehydrogenation and Dehydration of 2-Propanol
E . Kcal. G. Mole
Reaction Ni ZnO Cu
Dehydrogenation 33.6 20.5 15.0
Dehydration 51.3 20.8 12.2
 Table 111. Effect of Time Factor and
300.0 -
Temperature on Conversion
% Concwsion 7; Conwrsion
Dehydro- Dehydra- Dehydro- Dehydra-
No. W: F genation tion No. W:F genation tion
(At 250" C.) (At 275°C.) 200.0
1 16.6 60.0 1.5 1 15.9 67.4 2.0
2
3
18.6
19.5
73.3
75.0
1.1
1.6
2
3
18.9
22.0
84.0
87.0
1.6
2.2
m
--
4 21.8 73.1 1.1 4 25.9 88.1 2.8 a
5 24.4 77.7 1.2 6 27.8 91.7 1.8 :I'
6 24.7 81.5 2.2 6 29.0 93.6 2.4
7 30.0 85.9 1.6 100.0
(At 300" C.) (At 325O C.)
1 16.7 77.5 3.3 1 16.8 88.5 4.1
2 18.8 83.6 2.5 2 18.5 95.1 4.6
3 21.2 87.3 2.8 3 20.8 91.1 3.7
4 22.3 89.0 3.4 4 23.9 94.0 5.0
5 27.3 96.2 3.5 5 25.0 94.6 5.4
6 28.5 93.0 7 .O
0.0 5.0 IO 0 15.0 20 0
WIF

Figure 2. Plot of - R T / P [2 In ( 1 - x ) + x ] vs. W/F

at higher W / F . As the dehydrogenation reaction is endo-
thermic, increase in temperature increased the conversion.
However, the dehydration reaction increases at higher
temperatures.
Kinetics. The curves of conversions to acetone ( x ) LIS.
W / F were correlated by a second-degree polynomial, x
= a ( W / F ) + b ( W / F ) 2and constants a and b were cal-
culated by the least squares method. The rate of reaction
at any W I F value can be obtained by differentiating
the above polynomial with reference to W I F . Since the
extent of dehydration was rather small, it was not consid-
ered in the kinetic analysis. First-order irreversible and
forward first-order with reverse second-order mechanism
were considered and the following relationships were
obtained.
FIRST-ORDER IRREVERSIBLE
REVERSIBLE
(a)' ' + x(b)' ' } - 1 2 l n I--
x2b
) - $1
(5)
In (a)'Z- x ( b ) ' ' ( a
where a = K R T ; p and b = 1 + a.
For establishing the linearity between W / F and the
right-hand side of Equations 4 and 5 , the smoothed values
of x-Le., the values of conversion x calculated from the
fitted polynomial a t selected values of W / F (5, 10, 12,
16, 18, 20, and 22)-were used. The linearity was found
to be statistically significant. Figure 2 shows typical plots
for Equation 4. The values of the reaction velocity con-
stants obtained from the slopes of the curves are pre-
sented in Table IV. By the Arrhenius plots, activation
energies were calculated. The reaction velocity constants
can thus be expressed as
h = 11760 e-4?40'' (4)
and
h = 10,080 (5)
Standard deviations between experimental and cal-
culated conversion values from Equations 4 and 5 are
7.1 and 4.2%, respectively. Kolb and Burwell (1945)
experimentally determined the equilibrium constants for
this reaction, which is reversible in the temperature range
employed here. Hence it is preferable to employ Equation
5 in preference to pseudo-first-order irreversible Equation
4.
Hougen-Watson Reaction Mechanism. T o evaluate the
extent of internal and external diffusion in the reactor
system, experiments were carried out with -100 +150
Tyler mesh catalyst size and 4-inch and 2-inch high
catalysts beds. The conversion values with -60 +lo0 and
-100 +150-mesh size were similar under identical operating
conditions, indicating negligible internal diffusion.
Similarly with 2-inch and 4-inch high beds, the same
conversions were obtained at equal WIF values, showing
the absence of an external diffusion-controlling step.
When various mechanisms and rate-controlling steps
were tried, single-site surface reaction was found to be
the controlling step a t all the temperatures studied. The
values of the apparent reaction velocity constant, h,, and
the adsorption velocity constants, KA and K H (or K K ) ,
are presented in Table V. By the usual Arrhenius plot,
activation energies were obtained. The various constants
can thus be expressed as
The following rate equation can then be used.

Ind. Eng. Chem. Process Des. Develop., Vol. 9, No. 2, 1970 295
 230.4907 x e-3606,6'TCa - C H C ~ / K
r= (6)
1 + (0.1688 x 10-zez685i'r)CA+ (9.9536 x 10-12e123666.irT
CH

The standard deviation between the experimental and

Table VI. Effect of Time Factor and
calculated rates by Equation 6 is 49.8%. This figure is Temperature on Conversion
rather high. However, the temperature dependence of k,,
K A , and K H was as theoretically expected. The scatter 7c Conuersion 7% Conwrsion
in the Arrhenius-type plots caused this large error. Dehydm- Dehydm- Dehydm- Dehydra-
Fluidized Bed Reactor (Chemically Pure 2-Propanol). No. WIF genntion twn No. W / F genation tion
The reactor used for static bed studies was used for (At 250" C.) (At 2750 C.)
fluidized beds. A 4-inch high bed of Cu-SiO, (1 t o 4, 1 13.3 49.7 1.1 1 15.7 72.1 1.8
-60 +lo0 Tyler mesh) was used throughout. The incipient 2 17.4 57.0 1.8 2 16.9 73.3 2.6
fluidization velocity a t 300" C., calculated by using Leva's 3 19.1 71.0 2.0 3 17.0 72.8 2.0
(1959) correlating equation, was 16.50 pounds per hour 4 22.1 82.0 2.1 4 18.9 77.4 2.6
5 24.3 83.5 3.2 5 25.3 84.1 2.8
per square foot. For 140 grams of catalyst needed for 6 30.9 84.0 2.9 6 28.7 87.8 3.1
a 4-inch bed height ( W I F ) m l nis 51.54. The range of 7 29.0 94.2 5.5
W / F used was 14 to 30, and hence the bed was in the (At 30P C.) (At 325" C.)
fluidized form. The visual observation of constant pressure
1 17.0 78.1 3.8 1 16.9 87.6 3.5
drop a t all the flow rates used supplements the above 2 18.2 84.4 3.5 2 20.0 94.5 4.5
assumption. The range of temperature used was 250" t o 3 18.5 84.3 3.5 3 22.4 95.6 4.3
325'C. The experimental data are given in Table VI. 4 20.4 87.2 3.3 4 25.0 92.2 6.8
Comparison with static-bed data (Table 111) shows that 5 24.0 88.4 3.2 5 28.5 96.0 3.8
6 28.9 90.6 5.7 6 30.3 91.0 7.0
the over-all performance of a fluidized bed reactor is similar
to that of a static bed. The effect of temperature and
the time factor is similar to that obtained earlier and
the conversions are almost equal under similar operating as similar to a plug-flow reactor. Kinetics therefore can
conditions. These observations are contrary to the be adequately represented by Equations 4, 5, and 6.
expected low conversions in fluidized beds due to axial Fluidized Bed Reactor (Technical Grade 2-Propanol). The
backmixing and bubble phase. available technical grade 2-propanol was of a denatured
T o analyze these observations, residence-time distribu- variety. I t was purified by fractional distillation and the
tion studies were carried out with the same catalyst in heart-cut boiling a t 80.5"C. (680 mm. of Hg) was used
a glass tube having the same diameter as that of the for experimentation. This fraction compared favorably
reactor and a catalyst bed of 4 inches. Air was used with pure isopropyl alcohol in 2-propanol content (above
for fluidization and acetylene was used as a pulse tracer 99%) and refractive index (nz = 1.3820). The results
input. The exit gas samples were collected with the con- with Cu-Si02 (1 to 4, -60 +lo0 Tyler mesh) catalyst
tinuous rotating disk gas-sampling device of Kumar (1965) under fluidized bed conditions are presented in Table VII.
for time intervals of about 2 seconds. Acetylene in the The range of temperature and time factor ( W , F ) used
gas samples was estimated and the C curve was plotted. is 250" to 325" C. and 18 to 36, respectively.
From the relationship between the variance of the C curve The trend of conversion us. W I F curves is similar to
and reactor dispersion number, Dl uL, the experimental that obtained with C.P. 2-propanol and the effect of tem-
value of DluL was obtained as 0.01561. The tracer studies perature is the same. However, the conversions to
were calculated a t W,'F 18.0. The small value of D / u L dehydrogenation and dehydration obtained with technical
obtained partly explains the similar performance of static 2-propanol are considerably lower under otherwise identical
and fluidized beds, since the flow pattern can be taken conditions. The lower conversions cannot be attributed
to the lowering of the concentration of alcohol by the
denaturing agent, as the latter can be present in traces
only. The activity of the catalyst was not destroyed per-
manently. Thus the catalyst was not poisoned and the
Table IV. Values of Reaction Velocity
Constant at Various Temperatures relatively low conversions are probably due to the preferen-
tial adsorption of the denaturing agent on the catalyst
k , Liters surface, reducing the availability of active centers to
No. Temp., C. Hr , G Cat
alcohol adsorption.
1 250 3.882 Kinetics. The collected data were tested for the irreversi-
2 275 5.231
3 300 6.500
ble pseudo-first-order reaction mechanism given by Equa-
4 325 10.500 tion 4; satisfactory correlation was obtained. The
Arrhenius plot of calculated reaction velocity constants
Table V. Values of k,, K A , and KH at Various Temperatures and l / T yielded a straight line, from which the activation
energy was calculated. The reaction velocity can be
Temp.,
expressed as
No. c. k, KA KH
'
1 250 0.2563 2.9377 1.8641 k = 1.312 x 10' x
2 275 0.2559 1.5850 1.5923 ( E q u a t i o n 4, technical 2-propanol)
3 300 0.4677 1.9021 0.4402
4 325 0.5205 1.3880 0.3470 The energy of activation is higher with technical than
with C.P. 2-propanol. The standard deviation between

296 Ind. Eng. Chem. Process Des. Develop., Vol. 9, No. 2, 1970
 p = total pressure, atm.
Table VII. Effect of Time Factor and r = reaction rate
Temperature on Conversion R = gas constant
5; Conwsion X Conversion 7’ = absolute temperature, OK.

Dehydro- Dehydra- Dehydro- lleh-ydra-

W = weight of catalyst, grams
No. W )I.’ genation tion No. W IF genation tion
(At 2500C.) (At 275” C.) Literature Cited
1 18.4 34.2 0.1 1 18.5 47.6 0.3 Balandin, A. A., Advan. (’atal.ysis 10, 96--129 (1958).
2 22.6 32.2 0.3 2 24.0 54.4 0.9 Balandin, A. A., Dutzik, Z., Tolstopyatova, A. A . , 12~‘.
3 25.4 37.7 0.3 3 28.0 60.4 1.7 Akad. Nauk, SSSR Otd. Khim. Nauk 1964, 969--Y.
4 27.8 44.5 0.7 4 28.2 58.6 1.6 Balandin, A. A., Konenko, I. E., Tolstopyatova, A . A . .
5 35.2 47.5 1.9 5 34.6 75.2 2.3 Kinetics Kat& 2, 900-6 (1961).
(At 200” C.) (At 3250C.) Balandin, A. A,, Teteni, P., Problem3 Kinetica i Katuliza,
1 18.5 61.1 0.9 1 18.6 73.1 1.8 Akad. Nauk SSSR, No. 10, 339--42 (1960).
2 20.7 64.0 0.8 2 21.8 77.2 2.2 Balandin, A. A,, Tolstopyatova, A . A., ”Proceedings of
3 22.5 66.3 1.3 3 22.9 76.3 2.2 Third International Congress on Catalysis,” Vol. I, pp.
4 23.6 70.6 1.4 4 23.7 80.7 2.3 533-46, Korth-Holland Publishing Co., Amsterdam,
5 27.8 74.0 2.0 5 28.7 83.9 2.6 1965.
6 36.9 83.1 2.5 6 35.1 87.1 3.1 Ciborowski, S., Roczniki C’hern. 33, 145-55 (1959).
Cotrell, T . L., “Strength of Chemical Bonds,” 2nd ed.,
Butterworths, London, 1958.
Garcia de la Banda, J. F., J . Catalysis 1, 136 ,50 (1962).
Garcia de la Banda, J. F., U. S. Dept. of Commerce,
Office of Technical Services, P. B. Rept. 155,135, 94
experimental and calculated W F values for various con- (1960).
version values is 5.2 ‘-c. Garcia de la Baqda, J. F,, Orlandini, G. K., Anales Real
Hougen-Watson Reaction Mechanism. I n the case of Soc. Espaii. Fis. y Quim. (Madrid) 54B, 85-96 (1958).
technical 2-propanol, single-site adsorption of %-propanol Garcia de la Banda, J. F., Orlandini, G. K., Anales Real
Soc. Espaii. Fis. 3: Quirn. (Madrid). Ser. B, 59 (121,
was found to be the rate-controlling step over the entire
723-42 (1963).
temperature range employed. The calculated values of
apparent reaction velocity constants, k,, and the adsorp-
Garcia de la Handa, J. F., Orlandini, G. F.,
Lopez, A:
A., Anales Real SOC.Espaii. F&. J’ Quirn. (Madrid),
tion equilibrium constants, K H , followed Arrhenius-type Ser. B, 62 (2), 91-102 (1966).
relationship with reference t o temperature and following Germain, J. F., Gras, B., Beaufils, J . P., Compt. Rend.,
rate equation was developed. 261 (22) (Groupe 7), 4735-8 (1965).
Germain, <J. R., Gras, B., Beaufils, J . P.. Punsolle, L.,
Bull. Soc. Chim. France 1961, 1172--8.
Kipperman, S. L., Nikolaeva, N. V., Kinetica i Kataliz
2, 936-9 (1961).
Kipperman, S. I,., Nikolaeva, iX. V., Davydova, J. R.,
The temperature dependence of k , and K H was that Kinetica i Kataliz 4 (4), 605--13 (1963a).
theoretically expected. The calculated rates differ from Kipperman, S. L., Nikolaeva, N. V., Davydova, I. I t . ,
the experimental ones with a standard deviation of 28.541. Kinetica i Kataliz 4 ( 5 ) , 723-35 (196313).
Thus the most probable reason for lower conversions with Kipperman, S. L., Nikolaeva, N. V., Davydova. I . R.,
technical grade 2-propanol seems to be the preferential Tr. Inst. Khim. Nauk Akad. Nauk. Kaz. SSK 8, 3-20
adsorption of the denaturing agent, reducing the number (1962).
of active centers for alcohol adsorption. Hence it can Kolb, H. J., Burwell, R. L., J . A m . Chern. Soc. 67, 1084--
be concluded that, with this catalyst, denatured 2-propanol 8 (1945).
should not be used for acetone manufacture. Krylov, 0. V., Z h . Fiz. Khim. 39 (12), 2911-17 (1965).
Kumar, R., Brit. J . Chern. Eng. 10 (9), 640 (1965).
Nemencla ture Leva, M., “Fluidization,” p. 64, McGraw-Hill, New York,
a constant, K R T I P 1959.
Mitchell, J., Jr., Koltholf, I. M., Proskauer, E. S.,
b constant, 1 + K R T i P
Weissberger, A., eds., “Organic Analysis,” Vol. 1 , p.
C A molar concentration of adsorbed alcohol
259, Interscience, New York, 1955.
CH molar concentration of adsorbed hydrogen
Orlandini, G. K., Anales Real. Soc. Espaii F k . J ’ Qulm
Ck molar concentration of adsorbed ketone
E energy of activation, kcal./g. mole
(Madrid) 55B,465-76 (1959).
F feed rate, moles/ hour Sheely, C. Q., Jr., “Catalytic Dehydrogenation of
k reaction velocity constant Isopropanol,” €’h.D. thesis, University of Illinois, 1963.
apparent reaction velocity constant Zhabrova, G. M., Koginskii, S. Z., Shibanova, M . D.,
k,
K equilibrium constant Kinetica i Kataliz6 (6), 1018-24 (1965).
KA adsorption equilibrium constant for alcohol
KH adsorption equilibrium constant for hydrogen RECEIVED
for review June 10, 196Y
KK adsorption equilibrium constant for acetone ACCEPTED October 30, 1969

Ind. Eng. Chem. Process Des. Develop., Vol. 9,No. 2, 1970 297