total process capital and operating cost is lower when a tubular reactor is employed in the design because less reactant A must be separated and recycled. There is no advantage in a unit in which there is partial conversion of B. T h e cost saving resulting from a smaller reactor volume is more than offset by the increased capital and operating costs for a B-removal unit. Acknowledgment
The authors express their gratitude to the National Science Foundation and the University of Delaware Research Foundation for financial support. Nomenclature
C A = molar concentration of reactant A, moles/volume CB = molar concentration of reactant B, moles/volume Ci = molar concentration of product PI, moles/volume C,, = heat capacity of reactant A, energy/degree-mole C p B = heat capacity of reactant B, energy/degree-mole -.CPi = heat capacity of product Pi, energy/degree-mole C, = heat capacity of mixture, energy/degree-mole Ei = activation energy of reaction i, energy/mole AHi = heat of reaction i, energy/mole ki = reaction rate constant of reaction i ki, = frequency factor of reaction i, energy/mole q = volumetric flow rate, volume/time Q = heat removed or added to reactor, energy/time-volume rA = rate of depletion of reactant A, moles/volume-time r g = rate of depletion of reactant B, moles/volume-time
ri
R T
V xi
= rate of formation of product Pi, moles/volume-time = gas constant = temperature of reactor = reactor volume = pounds of product Pi
SUBSCRIPTS = inlet condition e = outlet condition o
GREEKLETTERS = rate constant ratio of reaction i to reaction 1 ai = conversion of reactant A, CA/CA, y = conversion of reactant B, C,/CA, 7 = density of product P, pi p = density of mixture e = holding or space time of reactor literature Cited
Buzzelli, D. T., M.Ch.E. thesis, University of Delaware, Newark, Del., 1966. Eastham, A. M., Latremouille, G. A., Can. J . Chem. 30, 169-76 (1952). \ . _ _ _
Ferrero,’P., Beebe, F., Flamme, L. R., Bull. Sod. Chin. Belg. 56, 349-68 (1947). Friedman, M. H., White, R. R., A.I.Ch.E. J . 8, 581-6 (November .n,n\ IYU.6).
MacMullin, R. B., Chem. Eng. Progr. 44, 183 (1948). Maget, J. R., J . Polymer Sci. A2, 1281-90 (1964). Parker, W. A., Prados, J. W., Chern. Eng. Progr. 60,74 (1964). U. S. Tariff Commission, Washington, D. C., “U. S. Production and Sales of Synthetic Organic Chemicals,” 1965. RECEIVED for review February 12, 1968 ACCEPTEDJune 24, 1968
DEHYDROGENATION OF BUTYL ALCOHOL IN FIXED CATALYST BEDS U . R. R A O , R A J I N D E R K U M A R , A N D N . R . K U L O O R Department of Chemical Engineering, Indian Institute of Science, Bangalore 12, South India
The vapor-phase dehydrogenation of 1 -butanol to butyraldehyde was studied in a fixed bed of catalyst from 2 5 0 ” to 360” C. Of all the catalysts studied during preliminary investigation, the one containing 90% copper, 8% chromia, and 2% carbon supported on pumice was best, with high activity and selectivity. The data are expressed in the form of a first-order irreversible reaction rate equation. Single-site surface reaction (hydrogen adsorbed) is the rate-controlling mechanism at all the temperatures studied. The rate data obtained in the entire range of experimental conditions fit the rate equation based on this mechanism with a standard deviation of *22.8%.
of dehydrogenation for obtaining butyraldehyde Tfromuse1-butanol is of interest because it not only permits HE
no side reactions but also yields pure hydrogen as a by-product. T h e stoichiometric equation for the dehydrogenation of I-butanol is 1-C4H40H + C3H7CHO Hz
+
T h e following reactions are also possible.
+
DEHYDRATION.1-C4H90H -+ I-CdHs HzO CONDENSATION. 2(CdHgOH) CaH&H=CCzH&HO -+
+ HzO
TISCHENKO’S REACTION. Z(C4HgOH) + 2(C8H&HO) 2(CsH&2HO)
-+
+ 2Hz
C3H&OOCJIo
Dehydration is predominant at elevated temperatures (>400° C.), whereas condensation and Tischenko’s reactions are favored by higher pressures. Thus, dehydrogenation can be carried out best at low temperatures and atmospheric pressure. T h e equilibrium constants for dehydrogenation were calcuVOL. 8
NO.
1
JANUARY 1969
9
A . COMPRESSED AIR 6 . FIXED BED REACTOR
I GLASS WOOL
2 ANHYDROUS CALCIUM CHLORIDE
C a
3 ANIMAL CHARCOAL
WATER COOLED CONDENSER
4 CYLINDER CONTAINING NITROGEN OR HYDROGEN
Y
WATER
IN
OUT
WAY STOPCOCK
d W
PORCELAIN PACKINQ
GONlAl N I N 0 ICE SALT MIXTURE
'ER
Figure 1 ,
WATER
Experimental setup
lated at various temperatures, and the equilibrium conversions were evaluated by using the equation
achieved and maintained for half an hour. Both the liquid and gaseous products were collected and analyzed. In certain runs, electrolytic hydrogen was passed along with the reactant. Choice of Catalyst
At temperatures higher t h a n 523.1' K., the equilibrium conversion is greater than 99.9%. Therefore in the analysis that follows, the reaction is considered irreversible. The equilibrium constants for other reactions were also evaluated, but are not reported here, because during the experiments none of these reactions proceeded to an appreciable extent. As dehydration can become important at higher temperatures, the range selected for the present study was from 250' to 360' C. Experimental Setup
A flow diagram of the equipment is given in Figure 1. The reactor is a stainless steel tube 5.0 cm. in i.d., heated by a radiation heater, the input to which is controlled by Dimmerstats. The reactor is partially packed with porcelain beads to raise the preheated reactants to the reaction temperature. Two thermowells measure the temperature of the catalyst bed. The preheater and the superheater are of 1.25- and 0.625-cm. diameter, respectively, and are heated by radiation heaters. Above the catalyst a perforated stainless steel plate prevents the fluidization and carryover of the catalyst particles. Arrangements are provided for the entry of the reactant and withdrawal of products (Figure 1). The entry of alcohol is brought about by using the acid-egg principle. The gaseous product from the reaction is sampled without disturbing the flow conditions of the reactor by using the gassampling device of Kumar and Kuloor (1964), Experimental Procedure
1-Butanol, analytical reagent grade, was metered by a flowmeter and passed through the preheater to the reactor. T h e run was taken only after steady-state conditions were 10
l&EC PROCESS D E S I G N A N D DEVELOPMENT
A number of catalysts are reported in the patent literature, most of them based on copper. The use of alkali metal oxides like KzO as promoters with reduced copper has been reported to give high conversions, few side reactions, and pure hydrogen (German Patent, 1963). Zinc oxide catalyst has been reported to give high conversions to dehydration and hence it cannot be employed in the present case. I t was observed during preliminary runs that the conversion increased with the time of usage of the catalyst. O n examination of the catalyst after reaction, a certain amount of carbon deposit was found. These results confirm the finding of Kotelkov (1954) that the presence of a carbon deposit improves the performance of platinized Xichrome as a dehydrogenation catalyst. Therefore, the authors decided to base the catalysts on copper, chromia, and carbon for preliminary runs. Many catalysts of varying composition and support material were prepared for comparison (Table I). The catalysts in Table I were compared for activity and selectivity. Activity is indicated by over-all conversion, whereas selectivity is indicated by conversion to the desired product. The comparison for conversion was made for a predetermined value (38 f 0.5) of W / F a t various temperatures, chosen such that diffusion effects are negligible and the reaction takes place in the chemically controlled regime. The catalysts were compared at various temperatures, because a catalyst which is not good at lower temperatures may be the best at higher temperatures because of the favorable values of frequency factor and activation energy. The conversions
Table 1.
Catalyst No.
Composition of Catalysts Employed in Preliminary Investigation
Components
Ratio of Components
...
I
... ... ... ...
I1 111
IV V
Copper-chromia 70% Cu f 3Oy0 CrzO3, calculated as Cr 90% Cu f 10% Ci-203, calculated as Cr Copper-chromia Copper-chromia VI11 99% Cu f 1% CrtO,, calculated as Cr Copper-chromia-carbon 90% Cu f 8% CrzO3 f 2% C, calculated as Cr IX Copper-chromia-carbon X 84% Cu f 7% Crz03 9% C, calculated as Cr For catalysts VI to X,20% active material and 80% of support by weight was employed. VI. VI1
+
obtained, with the conditions of the experiments, are presented in Figure 2. All the catalysts studied were highly selective and none permitted more than 3y0conversion into undesirable products. Figure 2 shows that copper on asbestos catalyst gives low conversions compared with some of the others. The addition of 10% chromia, with pumice as a support material, improves the performance of the catalyst to a small extent. T h e addition of large percentages of carbon a t the cost of copper does not greatly improve catalyst activity. I n fact, the catalyst is worse than the one containing no carbon and 10% chromia in the active material. However, the incorporation of a small percentage of carbon (2% of the active material) gives highly encouraging results. This catalyst gives conversions much higher than those given by other catalysts throughout the temperature range of 275' to 350' C . Hence this catalyst (90% copper, 8% chromia calculated as chromium, and 2y0 carbon: 20% active material and 80% pumice) is chosen for complete kinetic analysis. Catalyst Characteristics. Size, -28 4-48 Tyler mesh. Surface area by BET method, 2.277 sq. meters per gram. Influence of Variables
The main variables influencing the conversions obtained are the time factor, composition of the feed, and temperature. All were studied during this investigation. Whereas a limited number of runs were conducted to study the influence of feed composition, a complete factorial design was resorted to for the study of time factor and temperature. For each temperature, a fresh batch of catalyst was em-
support Asbestos (99%) Asbestos (9870) Asbestos (9770) Asbestos (9570) Asbestos (90%) Pumice Pumice Pumice Pumice Pumice
ployed. Before discarding a catalyst batch after experimentation, its activity was tested through a standard run. I t was invariably observed that, during the collection of data, the catalyst underwent no change in activity. Effect of Time Factor. Normally, for catalytic reactions of this nature, it is customary to express the time factor as W/F. However, this can be done only when the over-all surface area of the catalyst is very large and virtually independent of particle size. I n the present case the surface area varies considerably when the particle size is changed. Thus, the authors decided to express the time factor as S / F instead of W/F. S is the total surface area of the catalyst employed and is related to W a s
S = WA1 (2) where A1 is the surface area of the unit mass of the catalyst. The conversions obtained for various time factors have been plotted in Figure 3 with temperature as parameter. I n the absence of side reactions, the conversion increases with S / F , as expected. As the data were obtained in an integral reactor, the curves of the conversion time factor have to be differentiated to obtain the reaction rates. T o eliminate personal bias during the numerical differentiation of the data, the authors decided to fit the data to polynomials and then evaluate the rates by analytical differentiation. Regression through the polynomials of second and third degree was tried. The third-degree polynomials gave higher variance than the second-degree ones a t all temperatures. Thus the second-degree polynomial of the form x = a(S/F) b(S/F)2 (3)
+
C A T A L Y S T S GIVEN IN TABLE ?
CATALYST XC-
U
Figure 2. Performance of various catalysts
em
275
300 TEMPERATURE 'C
325
350
-.-C
VOL. 8
NO. 1
JANUARY 1 9 6 9
11
!$
B
0.1
E 2 2 2
0.4
LA Figure 3.
Effect of time factor on conversion
% g
gg
0.3
= ::
0.2
*
0.1
2
'2 U
2
0
0
was used. The zero-power term of S / F was excluded from the polynomial to satisfy the boundary condition: x = 0 a t S / F = 0. All the polynomials are presented as solid lines in Figure 3. T h e values of their constants are given in Table 11. Effect of Temperature. Figure 3 shows that the conversion for any S / F value increases with increasing temperature, This behavior is expected from the Arrhenius equation, which gives the effect of temperature on homogeneous reactions. For catalytic reactions the final apparent effect of temperature is due to the effect of both adsorption constants and the specific reaction rate constant. I n the present case, as the change in conversion with temperature is high, the effect of adsorption constants appears to be less. No decrease in the activity of the catalyst was observed with rise in temperature within the range studied. Higher temperatures could not be employed, however, because of the greater chances of dehydration. Influence of Addition of Products. Hydrogen, in the molal ratio of 1 to 1, was passed along with 1-butanol a t various temperatures. T h e same conversion-(S/F) curve is reproduced as for pure feed. This is probably due to the activating influence of hydrogen on the catalyst. Butyraldehyde was also added to the feed (87% 1-butanol and 13% butyraldehyde) and the experiments were conducted at three temperatures covering the same range (Figures 4, 5,
20
60
40
T,HE FACrOR
=
F
=
BO I00 la0 140 160 SURFACE AREA OF THE CATALYST SO. METERS FEED R A T E QRAH H O L E S / HOUR
180
ZOO
and 6 ) . The addition of butyraldehyde results in a drop in conversion. The data are, again, expressed by polynomials of second degree as for pure feed. The constants of the polynomials for various temperatures are given in Table 111. Lattice parameters of the catalyst, before and after use, were evaluated by an x-ray diffraction technique. Use of the catalyst resulted in a compression of crystal lattice. Rate Equation. Two methods were employed for the quantitative expression of rate. The first is based on the concept of order of reaction, and the second on the mechanism of reaction approach of Hougen and Yang (1950). O r d e r of Reaction. The design equation for the catalytic flow reactor can be put for the present case in the form
(4)
Jo r
r is a function of concentration of alcohol and is expressed as: (5) The value of n can be evaluated by a suitable plot of r us. C A or by assuming an order, solving Equation 4, and then testing whether the value of k remains constant. In the present case, the differential method was not employed because the poly-
PURE F E E D
Table II. Constants of Equation 3 at Various Temperatures Comtants pmp., a b No. c. Polynomial a ( S / F ) 4- b(S/FY = x 1 2 3 4 5
Table 111.
No. 1
2 3
250 300 325 345 360
0.000971 0.001831 0.003293 0.003538 0.004321
-0.000000708 -0.000002293 -0.000007275 -0.000007391 -0.000010313
Effect of Butyraldehyde in Feed on Conversion Molal Constants Concn. of a b Polynomial Temp., Alcoholx = a(S/F) b(S/F)2 ' C. in Feed
+
250 325 360
0.87 0.87 0.87
0.000370 0.001600 0,003591
-0.0000008000 -0.0000001579 -0.0000118400
o
Figure 4.
250' C. 12
l & E C PROCESS D E S I G N A N D DEVELOPMENT
40
20 FPCTOR
4-
-
40
eo
100
SURFPltE P R E P SO. METERS FEE0 R A T E GR4MHOLES/UOUR
Effect of
-
120
1m
160
zoo
butyraldehyde on conversion at
I
I
k Values Calculated by Integral Method
Table IV.
Temp., C. 250 300 325 345 360
No. 1 2 3 4 5
A
=
E =
Integral Method 0.04247 0.08804 0.16350 0.18630 0.23675
kcalcdt
x loa 10.277 x l o 3 cal. per gram mole 1.903
O n substituting the pertinent values in Equation 7, we get the following relationship among x , S / F , and T: FACiOR
::
A = SURFACE
Figure 5. 3 2 5 " C.
AREA 9O.METERS
FEED RATE GRAM MOLES/HOUR
F
-x
--c
Reliability of Rate Equation
T o verify if the derived equation represents the d a t a adequately, S / F was calculated at various values of conversion (Table V). The equation represents the data well. T h e over-all standard deviation works out as =t5.4'%. Apart from this, the conversions for different S / F values were calculated for different temperatures (broken lines in Figure 3); Equation 10 predicts the data well. Rate Equation Based on Reaction Mechanism. Before trying to express the data in terms of various rate-controlling mechanisms, it is necessary to check whether physical steps like external and internal diffusion play a n important part. T h e role of external diffusion was checked by conducting experiments, a t 360" C., by doubling the bed height from 1.87 to 3.74 cm. T h e highest temperature was purposely chosen, because here the reaction rate is highest and there is maximum chance that diffusion will play a n important part. T h e results a t the two bed heights are not significantly different from each other (Figure 9). The role of internal diffusion is normally studied by varying the size of the particle but maintaining the same weight. Such a plot is given in Figure 10A, where two distinct lines are obtained. However, this does not indicate that internal diffusion is significant in the present case, because the gross area is considerably changed on reducing the particle size. This is taken into account by employing different catalyst
r = kCA (6) C A is expressed in terms of conversion and introduced in Equation 4, which then becomes dx
RT
- 2 In
(1
- x)
(7)
Equation 7 is presented graphically in Figure 7 as the plot of RHS-(S/F), where straight lines passing through the origin are obtained. Experimentally obtained values have been employed in Figure 7. The values of k a t various temperatures are found from the slopes (Table IV). Effect of Temperature on Reaction Velocity Constant. T h e reaction velocity constant is related to the reaction temperature by the Arrhenius equation as follows: k =
Ae-E/RT
(8)
or i n k = In A
- E/RT
(10)
Equation 10 is the integrated rate equation. I t can be used for obtaining conversion for any set of conditions for pure feed. For mixed feed, Equation 10 can be modified by taking a suitable expression for C Ain terms of x in Equation 7.
nomial used is of rather low degree. T h e rates found by it cannot fit a first-order expression. Therefore, the integral method was used. I t was possible to fit the present d a t a in the first-order rate expression.
= -x
=
1903 e-10*277/1.987T (1/0.082 7') ( S / F )
Effect of butyraldehyde on conversion at
-S/F =
- 2 ln(1 - x )
(9)
T o verify the applicability of the above relationship, In k is plotted against 1 / T in Figure 8 , where In k varies linearly with 1 / T with a negative slope. From this plot, the values of E and A are found to be
M PURE F E E D MIXED FEED BASED ON PURE COMPONENT FEED RATE 0
-
x
0.2
Figure 6. Effect of butyraldehyde on conversion at 3 6 0 " C.
-
0 F
A
C
=~ 9 . ~ -~ SURFACE P R L P SO.YETERS F - FEED RATE ORAM MOLESlWOUR VOL 8
NO. 1
JANUARY 1 9 6 9
13
0.8
3.s'C
/ 0.6
A
t
9C
0.5
0.4
0.3
0.2
0.1
0 llME FAtTOR&
F
Figure 7.
SURFACE AREA OF THE CATALYST FEED RATE GRM. MOLESlHOLM
Plot of - x
-2
In ( 1
-
SQ METERS.
sizes but maintaining constant S. The resulting plot is given in Figure 10B where the two sets of data fall on a single curve, showing that internal diffusion does not play a significant role in the present case. The various chemical mechanisms possible for the reaction under investigation are given in the Appendix, along with the actual and simplified equations. For pure feed, the CH and CK terms are equal, and the data cannot be used for the simultaneous evaluation of the coefficients of these terms. To evaluate the coefficients when C H and C, terms appear in the same equation, the mixed-feed data have been used. The resulting equations, however,
Table V.
Temp. 250' C . (S/F)calod
(S/F)exptl
22.77 45.54 68.31 91.08 113.85
22.70 45.20 68.98 91.97 114.96
22.77 45.54 68.31 91 .OS 113.85
0
b d
Temp. 345' C .
Temp. 325' C .
(S/%lod
28.02 50.50 68.58 91.63 114.72
Table VI.
c
have been checked for all the data, including the pure feed data, at all temperatures. The values of the coefficients for all 10 equations, evaluated by regression analysis techniques, are given in Table VI. The rates used in the evaluation of the adsorption constants, etc., were obtained by differentiating the appropriate polynomial. All the coefficients are positive for the single-site surface reaction mechanism, hydrogen adsorbed, a t all temperatures. Hence in all probability, this can be considered as the rate-controlling mechanism. From the coefficients, the values of apparent specific velocity constant (i/a) and adsorption equilibrium constants of alcohol and aldehyde ( b / a and
(s/F)ex~tl
(S/F)mlcd
(s/F)exptl
36.94 40.08 46.68 70.04 168.28
36.92 39.78 46.55 70.04 150.57
33.66 36.66 40.08 47.65 72.64 168.50
....
(s/%Iod
Temp. 360' C. (S/%ptl
(s/F)op.hd
36 72
i o . 04 43.81 52.61 81.92 171.42
33.28 47.54 72.48 109.28 167.22
33.26 47.66 75.09 114.46 155.09
Catalyst 90% copper, 8% chromia, and 2y0carbon on pumice.
Constunts a
- 1/T
Experimental and Calculated S / F Valuesa
Temp. 300' C .
(S/F)exptl
Arrhenius plot of In k
Figure 8.
x) vs. time factor
7
2
3
2061.46 -4433.68
1120.00 336.00 5109.00
-0,0044 -0.0035 2372.22
., ,
...
...
...
Coefficients of Reaction Mechanism Equations Mechanism 4 5 6 7 8
2421.67 -0.1645 -2745.13 -2048.00
.. .. .. .,
2660.26 -2311.78 2372.22 -4972.04
2729.83 -5102.05
...
...
-4136.56 4095.99 8192.00
...
9 0.00 0.00 2372.22
...
70 0.12 x 10'6 -0.14 X 1016 -2372.22 0.27 X 1016
Temperature 325' C. a
b c
d
463.18 -488.06
...
...
667.70 106.70 2066.00
...
-202.62 -900.62 704.32
...
-1053.00 -620.53 1696.56 520.00
, .
.. ..
..
1973.69 -2791.83 704.32 -4067.53
536.75 -561.64
...
...
510.20 23.08 -512.00
0.00 549.26
61.07 54.74 0.5 X
0.00 -470.00 122.29
...
0.00
...
- 0 . 1 1 x 10'2 -0.12 1011 549.26 0.24 X 10"
x
Temperature 360" C. a
b C
d
14
108.58 -102.31
...
...
140.80 15.50 387.30
...
167.07 70.85 88.10
...
118.05 186.49 -45.40 -256.00
.. .. , .
..
I & E C PROCESS D E S I G N A N D D E V E L O P M E N T
751.82 -437.30 88.11 -1092.90
124.00 -117.73
...
...
...
-0.27 X 10" 0.31 X 10" 122.29 0.59 X 10"
t \.
1 I
A.
=7\
t
l
i
x-
I
I
k.m
\
X
I 0
a
I
I
I
00
I
FACTOR
= 9 F
Figure 9.
-
-
w(
-
I
I
I 80
60
40
I
I
I
100
I
I
I00
I
I
I 160
140
SURFACE AREA SQ.METERS FEED RATE GRAM MOLES / HOUR
Effect of bed height a t
AVERAQE PARTICLE S I Z E 0*2516 AVERAQE PARTICLE S I L E 0.442
360" C.
mm mm
A Tml
d
l
0.0015 "0.0016
I
Figure 11.
20
0
40
W WEIQHT OF CATALYST F ' n - a u i n ALCOHOL FEEO RATE
*-
0.0017
12-
I
I
I
I
00018
Plot of log
I
OMllP
I
0.0020
KA, KK, k vs. 1 / T
c / a ) can be calculated (Table VII). The coefficients were found to be significantly different from zero through the application of Student's t test. T h e dependence of k, K A , and K K on temperature provides a further check on the validity of the mechanism found. Whereas K A and K , should decrease with increase of temperature, k must increase. Also each should follow an Arrhenius type equation. This is evident from Figure 11, where k, K,, and K , are plotted against l / T . All show linear relationships; whereas In k - l / T shows a negative slope, the other two show positive slopes, as expected.
60 QRAMS ORAN M L E S / H O U R
-1
Table VII. Specific Reaction Velocity Constant and Adsorption Equilibrium Constants of Alcohol and Aldehyde at Various Temperatures Temp., c. k KA KK 6 F '
Figure 10.
SURFACE AREA OF THE CATALYST n--BuTvL
ALCOHOL FEEO RATE
250 325 360
SP. M E T E R S QRAMMOLESIHOUR
0.000853 0,001499 0.007104
4.562 3.098 2.750
0.3000 0.1590 0.1101
Effect of particle size on conversion a t 250" C.
Table VIII.
Temb. 250' C . rexDtl
1 2 3 4 5 (1
7
8
9 10
roslcd
0.000809
0.000609 0.000553 0.000506 0.000463 0.000431
0.000370 0.000369 0.000368 0.000367 0.000367
0.000403 0.000394 0.000385 0.000376 0.000359
0.000954 0.000939 0.000922 0.000842
Experimental and Calculated Reaction Rates Temp. 300' C. Temp. 325' C. Temp. 345' C. rem rcalcd rexptl rcaled rexptl Toaiod PURE 0.001723 0.001582 0.002961 0.002440 0.003201 0.003445 0.001619 0.001351 0.002630 0.002075 0.002864 0.002690 0.001515 0.001171 0.002299 0.001652 0.002528 0.002191 0.001410 0.001042 0.001967 0.001366 0.002191 0.001845 0.001310 0.000933 0.001646 0.001200 0.001855 0.001598
MIXED 0.001621 0.001606 0.001592 0.001576 0.001563
0.001952 0.001784 0.001663 0.001542 0.001436
VOL. 8
NO. 1
Temp. 360" C. rexptl
rdcd
0.003852 0.003382 0.002912 0.002442 0.002208
0.004090 0.003650 0.002450 0.002030 0.001497
0.003052 0.002513 0.001972 0.001436
0.002810 0.002203 0.001690 0.001440
JANUARY 1 9 6 9
15
From these plots, the values of E and A have been calculated. After substituting these values in the equation for the proposed mechanism, we obtain
Appendix.
This equation can be used for calculating the rates at all the temperatures studied. Equation 11 has been verified on the basis of the data obtained for pure as well as mixed feed at all temperatures. The actual and calculated values of r are given in Table V I I I . The values of r, obtained from Equation 11, when compared with experimental values (Table V I I I ) show a standard deviation of ~ t 2 2 . 8 7 ~ The . comparison in Table VI11 is made over the S / F range of 10 to about 150. Though the accuracy of Equation 11 is less than that of Equation 10, it can be a p plied with more confidence if extrapolation of data to a wider temperature range is desired.
Various Possible Mechanisms with Their Actual and Simplified Equations
Reaction type A + K Mechanism
+H
Actual Equation
Simplijed Equation
1. Adsorption of alcohol controlling single site r = ~[ CA (CKCH/K)] ( K adsorbed) [I f K H C H ]
2.
yl = a
+ bxl
Surface reaction controlling single site (H adsorbed)
3. Desorption of hydrogen controlling (single site) 4.
Adsorption of alcohol controlling (dual site)
5.
Surface reaction controlling (dual site)
6 . Desorption of hydrogen controlling (dual site)
7 . Adsorption of alcohol controlling (hydrogen adsorbed, single site) 8.
Surface reaction controlling single site ( K adsorbed)
9. Desorption of K controlling (single site) IO.
Desorption of K controlling (dual site)
Acknowledgment
W/F
=
One of the authors (U.R.R.) thanks the authorities of the University Grants Commission, New Delhi, for the award of a Senior Research Fellowship to him.
a, b, c, d
k
= =
r
=
X
=
Nomenclature
A AI
= frequency factor = surface area of unit mass of catalyst, sq. meters/
gram C K = concentration of alcohol, hydrogen, and aldehyde, respectively E = energy of activation, cal./gram mole K = thermodynamic equilibrium constant KA, K H , K K = adsorption equilibrium constants of alcohol, hydrogen, and aldehyde, respectively = total pressure of system, atm. PT S = total surface area of catalyst, sq. cm. = time factor based on surface area of catalyst, S/F total surface area of catalystjmolal feed rate of reactant, gram moles/hour T = temperature, K. = mass of catalyst, gram CA, CH,
w
16
l & E C P R O C E S S DESIGN AND DEVELOPMENT
XE XI,
xz,
XI
= =
time factor based on mass of catalyst, weight of catalyst, grams/molal feed rate of reactant, gram moles/hour constants in rate equations reaction velocity constant, liters/hour, unit surface area of catalyst rate of reaction, gram moles converted/hour, unit surface area of catalyst conversion, gram moles of n-butyraldehyde produced/gram moles of n-butyl alcohol passed equilibrium conversion concentration of hydrogen, aldehyde, and alcohol, respectively, in reaction mechanism equations
literature Cited
Ger. Patent 1.147.933 (Mav 1963). Hougen, 0.A.: Y k g , K.H.,Chem. Eng. Progr. 46, 146 (1950). Kotelkov, N.Z . , J . Appl. Chem. USSR27,961 (1954)(Engl. trans.). Kurnar, R., Kuloor, N. R., Brit. Chem. Eng. 9,400 (1964). RECEIVEDfor review April 12, 1967 ACCEPTED July 29, 1968