Primary Thermal Dissociation - ACS Publications

The writers are indebted to The Pfaudler Company of. Rochester, N. Y., for placing ... (2) Carpenter and Walsh, N. Y. State Agr. Expt. Sta., Tech. Bul...
0 downloads 0 Views 540KB Size
454

INDUSTRIAL AND ENGINEERING CHEMISTRY (3) (4) (5) (6) (7)

ACKNOWLEDQMENT

The writers are indebted to The Pfaudler Company of Rochester, K.Y., for placing a t their disposal the fruit juice concentrator with attached ester impregnating unit used in preparing the concentrates herein described, and to W. D. Pheteplace, Jr., for valuable assistance in carrying out the experiments.

(8) (9) (10) (11) (12) (13)

LITERATURE CITED (1) Carpenter, Pederson, and Walsh, IND. ENO.CHIM.,24, 1218-23 (1932). (2) Carpenter and Walsh, N. Y. State 4 g r . Expt. Sta., Tech. Bull. 202 (1932).

Vol. 26, No. 4

Eoff, J . IND.ENO.CHEM.,9, 587-8 (1917). Gore, Dept. Agr. Yearbook, pp. 227-44 (1914). Kleber, P h a m . Rev., 22, 94 (1904). McNair, J . Phus. Chem., 20, 633-9 (1916). Olsson, 2. physik. Chem., 133, 233-52 (1928). Power and Chestnut, J . Am. Chem. Soc., 42, 1509-26 (1920) Schonrock, table in A.O.A.C. Methods of Analysis, p. 510 (1930). Spencer, J . Phys. Chem., 33, 1987. 2012 (1929); 34, 410 (1930) T a r r and Baker, Univ. Del. Agr. Expt. Sta., Bull. 136 (1924). Todd (to Pfaudler Corp.), U.S. P a t e n t 1,856,979 ( M a y 3, 1932). Wilson, J . Assoc. Official Agr. Chem., 15, 635-42 (1932).

RECEIVEDOctober 5 , 1933. Approved by the director of the N. Y. State Agricultural Experiment Station for publication as Journal Paper 11.

Primary Thermal Dissociation Velocity Constants for Propane, n-Butane, and Isobutane R. E. PAUL^

AND

L. F. MAREK

Research Laboratory of Applied Chemistry, Massachusetts I n s t i t u t e of Technology, Cambridge, Mass.

N A PREVIOUS paper (5) depends upon the exchsion of Propane, n-butane, and isobutane are subjected an apparatus for studying secondary thermal reactions into limited thermal dissociation in copper and the thermal d i s s o c i a t i o n volving the products of the prifused-silica tubes under conditions which permit rates of the lower paraffin hydromary dissociation of the parent accurate evaluation of the time of contact and carbons was described. This hydrocarbon. I n the present temperature f o r the purpose of determining the apparatus permitted a c c u r a t e work, secondary reactions were control of the cracking temperaexcluded or limited to insignifivelocity constants of the primary thermal dissociature and the time of contact cant amounts by the restriction. The extent of dissociation is measured by through the use of a special t i o n of total dissociation of determination of the olefin content of the product electrically heated copper preparent hydrocarbon to a low gases by a modijied bromide-bromate method. heater section and a copper-coil v a l u e . I n only two runs did I n the temperature range investigated, the folr e a c t i o n c h a m b e r of known the dissociation exceed 9 per volume immersed in a lead bath. cent, and in most of the runs lowing equations represent the variation of I n the work with ethane and d i s s o c i a t i o n was below 5 per velocity constant with temperature: p r o p a n e , previously reported, cent. 74,850 analysis of the cracked gases Propane: loglo k = 16.60 - 2.3 RT was accomplished with a APPARATUSAND PROCEDURE Burrell gas analysis apparatus in (temp. range 550-650" C.) The general arrangement of the case of ethane and by low73,900 the experimental apparatus and temperature fractionation foln-Butane: loglok = 17.05 - 2.3RT the method of introducing the lowed by gas analysis in the case (temp. range 5 3 0 4 2 5 " C.) hydrocarbon gas were practiof propane. cally the same as in the earlier Because of the increasing com66,040 Isobutane: h g ] Qk = 14.89 - work (5). Two forms of reacplexity of the primary dissocia2.3 RT tion chambers were used in the tion products resulting from the (temp. range 5 5 0 4 1 0 " C.) p r e s e n t work, however. The thermal decomposition of parafChange of reactor surface to volume ratio, dilufirst (Figure 1) consisted of a fin h y d r o c a r b o n s higher in c o p p e r coil, 14.8 feet (4.51 molecular weight than ethane, tion with nitrogen, and change of surface from it was decided to adopt a difcopper to fused silica are found to have no effect meters) long, with an internal volume of 38 cc., an internal ferent analytical m e t h o d f o r under the experimental conditions. s u r f a c e a r e a of 487 sq. cm., t h e r e d e t e r m i n a t i o n of the and a surface to volume ratio n* r o *n a n e dissociation and for the determination of the n-butane and isobutane dissocia- of 12.8. Two-thirds of the runs reported here were made tion rates. The method depended upon the fact that, when in this copper coil apparatus. The first seven runs were p a r a f i hydrocarbons of the molecular weight under con- made with a similar coil having a volume of 30 cc. and sideration break owing to thermal instability, the pri- about the same surface to volume ratio. The reaction coil mary products consist of an olefin plus either hydrogen or a was heated by immersion in an electrically heated lead bath, paraffin of lower molecular weight. A determination of the the temperature of which was measured by a thermocouple. olefin content of the cracked product was, hence, depended The second type of reaction chamber consisted of a fused upon as a measure of dissociation. Obviously, the method silica tube bent in the form of a U, having one long and one short leg. The short leg of this tube terminated a t the upper 1 Present address, Humble Oil & Refining Company, Baytown, Texas;

April, 1934

INDUSTRIAL AND ENGINEERING

surface of the electrically heated lead bath in a thickwalled, fused-silica capillary which served as the exit C line. The long leg of this tube extended above the lead bath and served as the preheater section. It was fitted w i t h an e l e c t r i c a l heater wound on the outside, a fused-silica thermocouple well which extended to a point below t h e h e a t e r m i n d i n g a n d a b o v e thle upper lead b:Lth surface, and a fused-silica capillary preheater exit line which was sealed in a t a point opposite the tip of the thermocouple well, just above the lead bath surface. Both capillary exit lines were fitted with water jackets to insure rapid cooling of the gases A - Tncnwocoumc T-COPPER PREHLATLR and to permit connection to 8 - WATLR Jl\Cl(LT C-PRLIIEATLR E X I T U S *-LEAD 81," C-GAS INLLT rubber tubing. During the D- ELLCTUICAL HLATINC I -REACTION COIL UNIT J - R E A n l O N COlL C I I T CAS course of the work two of L - ASBCSTOS these fused-silica reactors FIGURE1. EXPERIMENTAL APwere used. One of these PARATUS FOR HYDROCARBON had an internal volume apDISSOCIATION p r o x i m a t i n g t h a t of the copper-coil reaction chamber a n d the other had an internal volume only two-thirds as large. Both had surface to volume ratios of nearly 4.5. During a run the temperature of the preheater was regulated so that the temperature of the gas leaving the preheater and entering the reaction chamber was the same as that of the lead bath. The hydrocarbon gas was stored in a reservoir, forced out by air-free water, dried with calcium chloride, metered with a calibrated flowmeter, and passed into the preheater and then through the reaction chamber (Figure 2). Samples of the exit gas from the preheater and from the reaction coil were collected in aspirator bottles a t atmospheric pressure. Saturated zinc sulfate solution was used as the confining liquid. Gas inlet rates were measured by water displacement and an orifice meter, and were corrected for pressure, temperature, and water vapor in the gas for subsequent calculation of times of contact. The gas samples mere analyzed for olefin content by an adaptation of the bromide-bromate method developed by Francis ( 1 ) . The method used m-as as follow: Exactly 25 cc. of approximately 0.02 iV potassium bromidebromate solution (restandardized nearly every run) were introduced into a 300-cc. glass-stoppered bottle fitted with a stopcock opening through the stopper and a rubber device for holding the stopper firmly in place. The bottle was evacuated of air by means of a vacuum pump. A sample of the gas t o be analyzed, measuring 100 cc. a t room conditions, was then introduced from a gas buret with care to prevent the entrance of air. Then 25 cc. of dilute sulfuric acid (27 cc. concentrated sulfuric acid in 2 liters of water) mere added. The top of the bottle, including all joints, was then thoroughly coated with melted paraffin wax, and the bottle shaken for 75 minutes in a motor-driven shaker. Sufficient half-saturated potassium iodide solution was added through the stopcock to give an excess over that required to react with the excess bromine. The stopper was then removed from the bottle and the iodine titrated with approximately 0.01 N sodium thiosulfate solution (frequently restandardized) with starch indicator. Usual precautions for restandardization of solutions and correction for blanks were taken. Two samples from the preheater and two samples from the reactor were analyzed for each run.

CHEMISTRY

455

The propane was prepared by fractionating Pyrofax in a low-temperature still. The first and last fractions were rejected and the middle cut was used after scrubbing with concentrated sulfuric acid. The n-butane and the isobutane were obtained from the Ohio Chemical and Manufacturing Company. This material was purified prior to use by low-temperature fractionation, and rejection of the first and last fractions.

RESULTS The results of the present experimental work are summarized in Tables I to I11 and presented graphically in Figures 3 to 5 . Velocity constants based on the rate of dis'appearance of the parent hydrocarbon only are plotted. Because of the fact that there is an increase in volume a t constant pressure of the gas during dissociation, the parent hydrocarbon drops in concentration a t a rate faster than that due solely to disappearance in the dissociation reaction Consequently, velocity constants based on the change in concentration of the parent hydrocarbon and on the proportion of the original hydrocarbon which disappeared a t any given time are not the same. Both of these constants are tabulated.

COOLlNC COlL

FIGURE2. FLOW DIAGRAM OF EXPERIXENTAL APPARATUS FOR HYDROCARBON DISSOCIATION

The general velocity equation for unidirectional, first order, homogeneous gas reactions may be written in the integrated form as follows: 12 where k

=

t

a

= =

b

=

=

2303

1 -a log,, 1-b t

velocity constant based on rate of change of the parent hydrocarbon, reciprocal seconds time of contact, seconds fraction of parent hydrocarbon dissociated in preheater fraction of parent hydrocarbon dissociated in preheater and reactor together

The velocity constant, k , expressed in this way and evaluated by the present data, Fives the rate of disappearance of the original hydrocarbon in an isothermal reaction under conditions where the pressure of the parent hydrocarbon remains practically Constant. The velocity constants shown by the tables and plots were calculated from the above equation without correction for the effect of reverse reactions. The extent of dissociation was sufficiently remote from equilibrium conditions ( 3 ) to make correction for the reverse reactions of little importance. The velocity constant, k', based on change in concentration

INDUSTRIAL AND ENGINEERING

456

1.0 0.8 0.4

0.4

0.4

0.4

0.2

0.2

0. I 0.08 0.06

0.I 0.08

0.04

0.06 0.04

0.02

tt

CHEMISTRY

0.02

0.01 0.008

K

0.006

coo4

0.01 0.000

0.006 0.004

0.002

0.002

0.001

-ONE dTUBE -ONE 0 -COPPER T U B E P G ' I ",::b: G:'; 0 - COPPER TUBE

0.0008

0.oooe

ATUS. ATMS.

5ILIC.A

0.0004

+ -rREY

-ONL

ATM

0.001 0.0008

V -MAW *M) kCLUER

A M HEW

0-COPPER TUBE

0.0004

0.0002

0.0004

0.0002

0.0001 IA60

Id20

llb0

Il:O

Ib40

llb0

0.0001

1420

525

8

1240

550 I220

575 1180

6 2 L I 450

f75

1100

IO40

IYO

-y

which were c a l c u l a t e d from the known rate of gas feed to the preheater by correcting to reaction temperature and allowing for v o l u m e i n c r e a s e due to dissociation at these two points. Inlet and exit gas rates did n o t h a v e to be corrected for temperature differences at the inlet and exit points of the reactor since the gas was a t the same temperature a t both. Thevelocity constant for the thermal dissociat i o n of p r o p a n e i n t h e temperature range 550" to 650" C. is accommodated by the equation:

I020

+$

FIGURE 3. THERMAL DISSOCIATION OF PROPANE

Vol. 26, No. 4

log,, k = 16.60 - 74 -L-.-850

FIGURE4. THERMAL DISSOCIATION

2.3 RT

OF

BUTANE

of the parent hydrocarbon for the same reaction, may be determined from the following relation, stoichiometrically derived from the preceding expression for k:

For the dissociation of nbutane in the temperature range 530" to 625" C. the equation takes the form: logla k = 17.05 - 73,900 ~

k'

2 303

= - log,,

t

1--a

(

l

2.3 RT

l + b

d ( I T U )

where k' = velocity constant based on coilcentration change, reciprocal seconds

log,, k = 14.89

TABLEI. THERMAL DISSOCIATION OF PROPANE RUN

TEMP.

a

b

c.

TIMEOF k, BASISOF CONTACT .MOLES

E', BAEIB OF CONCN.

Seconda 1 ATMOSPHERE

COPPER TUBE, PREBBURE

30 31 14 16 17" 18' 19b 20 21 22 23 24 25 28 29 26 27

552 552 573 573 573 573 573 585 588 6 10 6 10 6 10 610 631 631 652 652

0.0115 0.0138 0.0138

70 71 72 73 74 75

580 580 610 610 643 643

0.0052 0.0037 0.0059 0.0046 0.0145 0.0121

0.0054

~~~

0.0606 0.0729 0.0738

12.13 8.7 3.58 4.95 8.05 5.82 9.33 1.87 1.234 1.265 2.75 1.27 0.916 0.83 1.156 0.798 0.853

SILICA T U B E . PRBBSURE

0.0145 0.0101 0.0162 0.0116 0.0422 0.0341

p

0.000534 0.000596 0.00172 0.00164 0.00226 0.00178 0.00138 0.00312 0.00528 0.01380 0.01125 0.0121 0.0138 0.0401 0.0440 0,0772 0.0770

0.0010 0.0012 0.00339 0.00324 0.00445 0.00344 0.00270 0.00620 0.00868 0.0272 0.0220 0.0239 0.0273 0.0784 0.0852 0.1480 0.1446

I ATMOBPHERE

3.1 2.42 0.784 0.543 0.600 0.562

0.00296 0.00270 0.01331 0.01104 0.0477 0.0396

0.00607 0.00620 0.0253 0.0275 0.0955 0.0731

COPPER T U B E , PROPANE P R E S S U R E = 0.216, NITROQEN PRESSURE ATMOSPHERE

45 47 48 46 49 a b

599 608 610 632 639

0.0069 0.0173 0.0154 0.0115 0.0154

0.0302 0.0633 0.0583 0.0433 0.0700

2.54 4.94 3.61 1.217 0.97

0.00930 0.0097 0.0117 0.0270 0.0586

-

For the case of isobutane dissociation in the teniperature range of 5500 to 6 1 C.~the~equation becomes:

0.786

0.01128 0.01218 0.01355 0.0328 0.0706

Temperature peak 8 ' C. above 573' C. occurred during run. Gas narnple stood 60 hours before analysis.

The values of u and b were determined from the analysis of the exit gases from the preheater and the reactor chamber. Time of contact was calculated from the known volume of the reaction chamber and the average gas rate through the reactor. Average gas rate through the reactor was taken as the arithmetic average of the reactor inlet and exit rates

66,040 - 2.3 RT

The data of Marek and McCluer (5) for ethane recalculated to the present basis for velocity constant evaluation but not corrected for the effect of reverse reactions, and plotted to give least weight to the runs with high dissociation, give the following relations for k: logio

k

= 16.06

77,700 - 2.3 RT

The slopes of the equations for the velocity constants, e x p r e s s e d as above, are almost i d e n t i c a l 0.1 0.08 for the t h r e e hydro0.06 carbons-ethane, pro0.04 pane, and n-butane. 0.02 The heats of activation for t h e s e t h r e e 0.01 0.008 hydrocarbons lie be0.004 t w e e n approximately 0.004 74,000 and 78,000 on n the a b o v e b a s i s b u t show a slight progres0.001 0.0008 sive decrease in value e0006 with increase in molecua0004 lar weight of the hydrocarbon. The other constant of the equa0.0001 525 550 575 800 8 2 5 ' 650%. 1 1 tion, however, shows a I260 I220 1180 1140 1100 progressive i n c r e a s e 10' T 'K with increase in molecular weight of the hy- FIGURE 5. THERMAL DISSOCIATION OF ISOBUTANE drocarbon, expressing

IN D US TR Ih L AN D

April, 1934

E N G I N E E R I N G C H E hl I S T R Y

the greater ease of dissociation of the normal paraffins of higher molecular weight. TABLE11. THERMAL DISSOCIATION OF WBIJTANE RUN

TEMP.

r i

b

c.

TIMEO F k, Baals O F k', BASISOF CONTACT .MOLES COXCN. Seconds

COPPER T U B E , P R E S S U R E

51 52 53 .32 33 34 35 54 36 37 38 40 41 42 43

531 53 1 554 555 555 577 677 577 60 1 601 601 606 605 625 625

0.0021 0.0028 0.0028 0.0062 0.0047 0.0081 0.0072 0.0041 0.0087 0.0077 0.0087 0,0099 0.0089 0.0173 0.0163

8i 86 85 84 82 83 88 89

544 556 578 581 598 599 615 623

0.0078 0.0074 0.0085 0.0078 0.0062 0.0080 0.0069 0.0095

0.0078 0.0148 0.0125 0.0318 0.0217 0.0446 0.0366 0.0187 0.0462 0.0434 0.0472 0.0494 0.0470 0.0930 0.0928

SILICA 1'UBE, P R E S S U R l i

0.0294 0.0274 0.0447 0.0418 0.0305 0.0345 0.0400 0.0551

3

I ATMOSPHERE

7.7 14.9 2.59 7.58 4.35 4.24 3.07 1.04 0.975 0.806 1.26 1.08 0.877 1.005 0.803 3

0.000743 0.000791 0.00377 0.00345 0.00387 0.00666 0.00963 0.01435 0.0387 0.0442 0.0311 0.0367 0.0450 0.0801 0.1004

0.00148 0.00157 0.00750 0.00695 0.00768 0.0171 0.01881 0.02820 0.0780 0.0862 0.0611 0.0715 0.0875 0.1509 0,1900

0.00330 0.00536 0.01066 0.01715 0.0230 0.0296 0.0441 0.0664

0.00838 0.01054 0.0206 0.0334 0.0456 0.0579 0.0861 0.1286

COPPBR T U B E , B U T A N E PRESSURE 3 0.186, NITROQEN PRESSURE ATMOSPHERE

61 60 56 57 58 59

550 553 581 581 596 596

0.O03Sa 0.0030'' 0.0100 0.0080 0.0129 0,0120

0.0207 0.0168 0.0418 0.0363 0.0557 0.0513

12.4 5.91 4.41 3.22 1.796 1.21

0.00138 0.00336 0.00739 0.00896 0.02380 0.03352

=a

0815

0.00163 0.00278 0.0087 0.01069 0.0282 0 0397

Estimated, too low to measure with accuracy.

TABLE111. THERMAL DISSOCIATION O F ISOBUTASE RUN

TEMP.

c.

o

TIMEOF b

k, BASISOF k', BASIS OF CONTACT MOLES CONCN. Seconds

COPPER T U B E , PRESSURE

68 69 64 65 A 2 .~ 63

68

67

556 563 572 572 578 578 608 608

0.0036 0.0031 0.0089 0.0078 0.0134 0.0067 0.0112 0.0099

0.0135 0.0131 0.0366 0.0313 0.0420 0.0272 0.0460 0.0438

3

and 760 mm. mercury pressure, the results of Frey and Sniith (4) showed a progressive decrease in the calculated values for the velocity constant as the extent of dissociation was allowed to increase from 18.1 to 49.9 per cent. Pease and Morton ( 7 ) have also found in bomb experiments that the apparent dissociation rate dropped off as the reaction proceeded under isothermal conditions in the case of hydrocarbons higher in molecular weight than those under consideration here. Unless i t is possible to extrapolate the values of the velocity constants obtained a t these high amounts of dissociation to values at zero or low conversions, they should be used with caution in formulating values for heats of activation, general velocity equations, etc. The results of Pease and Morton ( 7 ) obtained from the dissociation of n-heptane in bomb experiments led t o the following relationship for the velocity constant:

1 ATMOSPHERE

6.64 3.71 3.420 2.035 1.08 0.915 0.767 0.710

1 ATMOSPHERE

3.51 2.30 4.45 3.10 3.90 2.39 1.15 1.07

0.00284 0.00433 0.00636 0.00775 0.00753 0.00872 0.0313 0.0326

0.00564 0.0086 0.01242 0.01517 0.01494 0.0171 0.0610 0.0634

The equation for propane given by Marek and RlaCluer ( 5 ) shows a slope different from the present equation but was based on a much smaller number of experimental determinations, Of the five runs reported in this previous work, only one was a t a total dissociation of less than 5 per cent, and it is noteworthy that a t the higher temperaturei. e., above 600" C.-the extent of dissociation in this former work was allowed to increase with the temperature, and that the corresponding values for the velocity constant deviated more widely from the present values as the temperature increased, The results of Pease and Durgan (6) obtained with propane and the butanes yield lower velocity-constant values and a lower heat of activation (65,000 calories) for the dissociations. The results of these workers are shown in Figures 3 to 5 and were obtained in packed and unpacked tubes, and in some instances with a 50 per cent nitrogen dilution. However, decomposition was carried to a much greater extent than in the present case, ranging from 14.7 to 43.8 per cent and being 20 per cent or greater in over half of the runs, As the dissociation of these hydrocarbons was allowed to increase in extent, the velocity of dissociation decreased so that, in the range of dissociation used by these workers, the calculated velocity was considerably lower than in the case where i t may be limited to a few per cent. For the case of the dissociation of propane in a silica tube a t 575" C.

437

46 500 log,, k = 9.85 - A 2.3 RT

The value of 46,500 for the activation energy for nheptane is out of keeping with the values presented here for propane and n-butane, but it should be noted that the equation is based on dissociation of 25 per cent and covers a rather narrow temperature range of 30" C . 4 . e., from 530' to 560" C. Aside from the extent of dissociation used, one of the major difficulties in determining velocity constants from much of the previous work has been in assigning values to the time of contact and to the temperature of the gas undergoing dissociation. Data have been reported from runs made under nonisothermal conditions, and the practice has generally been to average in some way the inlet and exit gas rates for the purpose of assigning a value to the time of reaction, and to use reactor wall temperature as the temperature of the gas. The present results with isobutane show a lower heat of activation (66,000 calories) than was found for the normal hydrocarbons. This lower value for the is0 compound is in keeping with the value of 59,000 calories obtained by Frey and Hepp (2) in an apparatus similar in effect to that used in the present work for the heat of activation in the case of isopentane dissociation in the range 400" to 575" C. Values for the velocity constants of the dissociation of propane and n-butane obtained from the silica reaction chamber are, in general, within the range of deviation of the values obtained from the copper coil reactor, although the surface to volume ratios varied from 4.5 in the case of the silica tubes to 12.8 in the case of the copper coils. This indicates that the effect of surface is practically negligible in this range of surface to volume ratios, and with these two materials. The effect of nitrogen dilution on the values of the velocity constants is also indeterminable in the concentrations used. Mixtures by volume of 1 propane to 3.65 nitrogen, and 1 n-butane to 4.4 nitrogen gave values of IC corresponding, respectively, to those obtained with the pure propane and n-butane.

LITERATURE CITED Francis, 1x1).ENG.CHEM.,18, 821 (1926). Frey and Hepp, Ibid., 25, 441 (1933). Frey and Huppke, Ibid., 25, 54 (1933). Frey and Smith, Ibid., 20, 950 (1928). (5) Marek and McCluer, Ibid., 23, 878 (1931). (6) Pease and Durgan, J . Am. Chem. Soc., 52,1262 (1930). (7) Pease and Morton, l b i d , 55, 3190 (1933).

(1) (2) (3) (4)

RECEIVED August 31, 1933.