I
H. J. HEPP and L. E. DREHMAN Research and Development Department, Phillips Petroleum Co., Bartlesville, Okla.
Isomerization of n-Heptane Aluminum Chloride-Hydrocarbon
C o m p l e x Catalyst
This process may lead to the development of more efficient processes for isomerizing heavier paraffins
I
s o M m I m m o N Is AN increasingly important means of upgrading straight chain or slightly branched paraffins to highly branched, high octane number products. When n-pentane or higher-boiling paraffins are isomerized in the presence of aluminum chloride or boron fluoride catalysts, disproportionation reactions take place which yield lower and higher boiling products and cause catalyst degradation ( 7 7, 72). Hydrogen, naphthenes, and aromatics inhibit the disproportionation reaction, but their effectiveness falls off rapidly as the molecular weight of the paraffin increases. Isobutane suppresses disproportionation for n-pentane, but has little effect with heavier paraffins (9,72, 73). I n the work described here, the effectiveness of isobutane in suppressing disproportionation of n-heptane in the presence of an active aluminum chloridehydrocarbon complex catalyst was investigated. Such catalysts have been employed in alkylation ( 7 , 5) and isomerization (4, 5, 77). Unlike solid anhydrous aluminum chloride, which was used in most isomerization studies, activity of the complex can be varied by changing its aluminum chloride content. Hydrogen chloride is a vigorous promoter, and thus affords another means for controlling activity. Side reactions were not completely eliminated, but when the mole ratio of isobutane to n-heptane was increased to 5, the fraction of n-heptane converted to isoheptanes was increased to a respectable 707,. Presumably, a further in-
crease in isobutane would decrease disproportionation still more. Apparatus and Materials
All experiments were of the batch type. Catalyst, hydrogen chloride, and feed were charged to a reactor held at the desired temperature by a water bath. stirred for the desired time, drained, and the phases were separated and analyzed. The reactor was a 580-cc. Hastelloy B autoclave, 3l/2 inches inside diameter and 4 inches deep, equipped with two 4-bladed arrowhead impellers rotated a t 1750 r.p.m. A conventional packing gland with Teflon packing was employed. Suitable openings for introducing reactants and draining products were provided. Auxiliary equipment included a diaphragm pump and an 18mm. glass debutanizing column packed with 8 inches of l/S-inch glass helices. The catalyst used in much of the work was an aluminum chloride-hydrocarbon complex formed in the alkylation of isobutane and ethylene. I t was a fluid, red-brown oil containing 56 to 60 weight per cent aluminum chloride. Its viscosity was approximately 300 cs. at 100' F. Upon hydrolysis it yielded a highboiling unsaturated oil having a density of 0.916 and a viscosity of 258 Saybolt universal seconds a t 100' F. Carbon and hydrogen content was 87.4 and 11.6 weight per cent, respectively. Infrared examination indicated the presence of branched parafins, branched internal olefins, and poly-substituted aromatics.
A similar complex made from a hexane cut of natural gasoline contained 64 to 657, aluminum chloride. While more active, product composition was unaffected. The hydrocarbons were Phillips Petroleum Co. pure grade isobutane and reference fuel grade n-heptane. Anhydrous hydrogen chloride was purchased from the Harshaw Chemical Co. and was used without further purification. Procedure
A measured volume of anhydrous hydrogen chloride was introduced into the reactor. Catalyst was then charged to the reactor from a calibrated-gage glass reservoir by pressuring in with nitrogen. A new sample of catalyst was used for each experiment. The stirrer was then operated for a few minutes to dissolve hydrogen chloride in the complex. The premixed isobutane-heptane feed was then pumped into the reactor from a container set on a balance. When the reactor was filled the pump was shut off and the stirrer started. After the desired time had elapsed the stirrer was stopped. Reaction time was taken as the period of stirrer operation. The catalyst was allowed to settle for a few minutes, after which the reactor contents were drained through a bottom fitting into a tared 1-liter, two-necked flask immersed in adry ice-acetone bath. T h e flask was connected to a side entry of the debutanizing column whose reflux condenser was cooled with dry ice-acetone. This permitted hydrogen VOL. 52, NO. 3
MARCH 1960
207
chloride to pass overhead into a water scrubber where it was collected. When the reactor pressure was reduced to atmospheric, it was disconnected from the flask. The hydrocarbon phase was decanted into the debutanizer through the side entry. Nearly complete separation from the catalyst was obtained since the catalyst was very viscous a t this low temperature. T o recover remaining hydrocarbon, the flask was warmed to 150" F., then disconnected from the column and weighed. Residual hydrogen chloride was stripped from the hydrocarbon product and isobutane taken overhead to a final temperature of 50' F. Distillate was collected a t dry-ice temperature and transferred cold to a tared sample container. T h e debutanized product was weighed, water washed, and dried. The isobutane cut was analyzed by mass spectrometer or gas chromato-
graphic methods. The heavier products were analyzed by fractionation in a l/2 X 36 inch Hypercal column, or by chromatography. Hydrogen chloride was determined by titrating the aqueous solution in the scrubber. The catalyst was analyzed for aluminum chloride, aluminum hydroxide, hydrogen chloride, and ferrous chloride by the method of Hale (7). After draining the reactor, some 11 to 13 cc. of catalyst remained on the walls. This was not removed prior to the next experiment, since a solvent would be required, and it would be difficult to ensure that all solvent had been removed. Thus 7 to 870 of old catalyst was present in most experiments. This had no observable effect on catalyst activity although it complicated material balance calculations. The amount of hydrocarbon reacting with aluminum chloride to form complex
was calculated from changes in aluminum chloride, aluminum hydroxide, ferrous chloride, and hydrogen chloride content of the new and used catalyst. The material balance was corrected to secure an aluminum chloride balance for the starting and used catalyst. Hydrocarbon so complexed was added to the hydrocarbon weight to determine total product composition. Results
Operating conditions were held reasonably constant for the series shown in Table I, except that the aluminum chloride content of the catalyst varied from about 56 to 64%. Catalyst activity thus varied somewhat, and this affected conversion. However, at the same conversion, disproportionation was not much affected. ~
Table 1. Experiment No.
At the Same Conversion, Disproportionation Was Not Substantially Affected
1
2
3
4
5
6
7
8
9
Temp., F. Pressure, p.s.i.g. Time, min.
108 90 15
108 90 30
112 335 15
113 525 30
113 430 15
106 280
112 445 30
106 315 30
113 320 15
Charges, g. Isobutane n-Heptane HCl AlCb-complex
0
0
239.0 19.7 199.0
240.0 21.6 199.0
41.4 52.1 189.8 180.5 19.7 19.7 286.1 230.0
52.1 182.6 19.7 250.0
83.0 144.2 19.7 241.0
98.5 143.5 20.0 201.3
Total 457.7 Mole ratio iCa/nC-i 0 Products, g. Hydrocarbon HCl AlCh-complex
Loss Total AlCL in complex, Wt. % Initial Final Composition products,
10
11
12
13
14
112 300 15
110 500 15
116 320 15
117 270 15
110 130 15
104.0 117.5 144.0 101.2 20.0 19.7 201.6 250.0
163.0 83.2 20.0 224.0
158.2 73.6 20.0 225.6
160.4 78.8 20.0 258.4
171.0 60.0 20.0 227.0
131.0 46.0 20.0 227.0
460.6 0
537.0 482.3 504.4 487.9 463.3 469.6 488.4 0.38 0.50 0.49 0.99 1.18 1.24 2.0
490.2 3.4
477.4 3.7
517.6 3.5
478.0 4.9
424.0 4.9
213.3 7.0 220.0 17.4
209.8 8.8 232.8 9.2
214.8 12.8 302.8 6.6
212.1 11.5 256.7 2.0
214.8 12.8 264.4 12.4
224.7 12.6 251.8
...
231.5 7.7 204.0 19.1
224.4 9.7 208.4 27.1
219.9 15.9 245.5 7.1
235.7 11.1 228.8 14.6
211.6 7.9 234.0 23.9
222.9 15.1 258.2 21.4
217.9 10.0 233.0 17.0
174.7 7.8 236.4
457.7
460.6
537.0
482.3
504.4
489.1
463.3
469.6
488.4
490.2
477.4
517.6
478.0
424.0
57.9 51.6
57.9 50.8
62.4 59.3
62.4 60.3
62.4 60.3
62.4 61.8
55.7 55.2
55.7 55.2
62.4 63.2
56.2 57.9
56.2 57.3
63.8 64.7
56.2 58.0
56.2 ' 58.5
0.22 22.55 1.20 17.30 2.32 2.42
0.29 24.53 1.36 18.95 2.62 3.06
0.40 29.95 2.42 18.75 3.61 6.15
0.40 33.00 3.00 16.50 2.32 7.25
0.46 31.10 1.69 19.60 3.84 6.08
0.44 34.00 2.96 18.70 4.46 8.48
0.30 33.53 1.77 17.20 3.54 5.19
0.17 34.60 1.97 17.06 3.93 6.30
0.55 45.95 1.87 13.50 2.10 3.00
0.14 61.89 1.45 5.91 0.61 0.78
0.06 59.31 1.08 6.90 0.70 1.33
0.20 59.90 1.68 9.65 1.52 2.21
0.13 69.32 1.40 5.68 0.65 0.75
0.13 69.00 1.19 6.03 0.77 0.98
7.92 2.77 1.50 0.57 4.79 (6*30 2.33 2.84 1.63 5.16 1.23 12.70 10.61 8.14 5.37
8.40 2.72 1.52
8.55 2.88 1.65 0.73 4.98 2.38 1.51 1.55 11.20 1.87
7.50 2.38 1.46 0.66 5.38 2.21 1.81 1.11 7.64 0.80
9.25 3.24 1.89 0.77 6.36 J5.00
8.61 3.94 1.68 0.64 5.84
1.91 0.67 0.21 0.96 3.25'
2.35 0.80 0.28 1.14
1.16 9.76 0
2.41 0.77 0.25 0.80 3.62O 5.94 5.87 9.98 1.65 (-0.67)
2.77 0.92 0.36 1.02
1.36 10.60 0
2.23 0.68 0.24 1.22 3.944 7.22 5.97 7.12 1.82 1.55 4.10 0 (-0.95)
3.66 3.53 (-0.26)
4.80 1.19 (-1.37)
2.99 1.34 (-1.62)
100.00
100.00
100.00
100.00
100.00
4.3 79.0
11.6 68.4
8.2 89.0
4.7 81.4
5.4 88.4
15
5.1
of wt.
%
Propane Isobutane n-Butane Isopentane n-Pentane Neohexane Z-Methylpentane diisopropyl 3-Methylpentane %-Hexane Triotane Dimethylpentanes 2-Methylhexane 3-Methylhexane n-Heptane Heavier6 Complexedb
+
Total Reacted hydrocarbon, yo Isobutane %-Heptane
8.38 2.81 1.23 1.14 9.44 8.95 15.58 6.40
8.70 3.15 1.36 0.84
0.58
4.78 3.08 1.57 1.25 8.40 5.23
\
5.19 1.77 0.73 1.55 8.50
100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
91.1
94.8
98.5
98.4
2,BDimethyl plus 2,4-diniethylpentane only.
208
~~
INDUSTRIAL AND ENGINEERING CHEMISTRY
98.0
98.2
13.3 97.5
12.7 97.9
Calculated from AlCla balance.
11.0 95.7
{E;
{I;:
n- H E P T A N E IS 0 MER IZ AT I O N The pentanes recovered were essentially an equilibrium mixture of iso- and n-pentane. As pointed out by Schneider (72), the hexanes formed contain a considerable percentage of neohexane: although it falls short of the equilibrium amount. Diisopropyl was also present, but was not determined. The heptanes cut contained triptane and both lowboiling and high-boiling dimethylpentanes. These highly branched products probably result from secondary reactions. The products boiling above n-heptane decreased with increasing isobutane concentration because of decreased disproportionation. The hydrocarbon reacting with aluminum chloride to form complex also decreased with increasing isobutane concentration. Above an isobutane-heptane mole ratio of 1.2 complexing of the feed became quite small. In Table I1 the effect of reaction time, temperature, and hydrogen chloride content on the reaction a t a 1 to 1 mole ratio of hydrocarbon feed is shown. It is remarkable that the reaction appears to go to completion in 15 minutes or less. Increasing the time to 120 minutes had little effect on the products. I t is unlikely that catalyst activity was decreased seriously over this period, suggesting that some sort of over-all equilibrium or steady state condition among the products was attained. Experiments 4, 5, and 6 were made to explore the effect of temperature on the reaction. A rather steep temperature coefficient of reaction rate is indicated by comparing the runs made a t 70" F. and 106" F. However, bringing the catalyst and hydrocarbon phases into contact may not have been so effective at the lower temperature, because catalyst viscosity increased to about 1000 cs. at 70" F. At 106' F., 98y0 of the n-heptane reacted; thus little change was apparent with further increase in temperature. As expected, hydrogen chloride is a vigorous promoter, and increasing its concentration increased reaction rate. There is some evidence in the runs reported that increased hydrogen chloride increases disproportionation.
mole ratio was further increased there was a net disappearance of isobutane which leveled off a t about 25 moles isobutane reacted per 100 moles of nheptane reacted. The total moles of products slightly exceeded the 125 moles of reactants a t the higher isobutane levels, indicating that products of the initial reaction were further disproportionated. Isoheptanes yield increased rapidly with increasing isobutane to a value of about 70 moles per 100 moles of nheptane reacted. Efficiencies probably could be driven higher by further increases in isobutane, but this was not pursued. The composition of the disproportionation products alone, and the ratio of the reactants yielding disproportionation ~~
Table 11.
~
The effect of increasing isobutane concentration is shown in Figure 1, where the data of Table I are plotted as cumulative moles of products per 100 moles of n-heptane which reacted us. mole ratio of isobutane to n-heptane charged to the reactor. As the mole ratio of isobutane to nheptane charged was increased from zero, isobutane yield fell off rapidly, until a t a ratio of about 1 there was no net production of isobutane. As. the
~~~~
~~
~~
That the Reaction Goes to Completion in 15 Minutes or Less Is Remarkable (Effect of time, temperature, and HC1 concentration)
Experiment No.
1
2
3
4
5
6
7
8
9
Temp., O F. Pressure, p.s.i.g. Time, min.
106 290 5
111 250 30
111 250 120
70 250 15
106 260 15
139 250 15
112 250 15
112 330 15
114 440 15
Charges, g. Isobutane n-Heptane HCl AlCls-complex Total
84.0 141.5 19.7 243.0 488.2
83.0 126.5 19.7 237.0 466.2
83.0 144.1 19.7 239.0 485.8
83.2 143.8 19.7 249.4 496.1
83.0 144.1 19.7 241.0 487.8
83.9 140.5 19.7 236.8 480.9
84.3 143.3 9.1 236.8 473.5
83.0 131.5 29.5 236.8 480.8
158.1 77.1 25.0 257.0 517.2
Products, g. Hydrocarbon HCI AlCl3-complex Loss Total
228.0 15.4 248.5 (3.7) 488.2
194.6 11.8 261.3 (1.5) 466.2
215.5 12.8 241.1 16.4 485.8
228.7 22.3 252.3 (7.2) 496.1
224.7 12.6 251.8 (1.3) 487.8
210.5 12.0 247.5 10.9 480.9
221.1 7.6 234.6 10.2 473.5
207.0 26.4 242.8 4.6 480.8
214.5 27.3 263.5 11.9 517.2
62.4 63.3
62.4 61.7
62.4 59.8
56.1 57.3
62.4 61.8
62.4 61.0
63.0 63.4
63.0 61.8
63.8 63.2
0.23 39.09 0.63 0.63 0.18 0.77
0.56 37.69 3.81 18.59 4.77 9.09
0.58 35.31 6.01 17.86 4.80 10.10
0.06 38.05 0.75 6.05 0.67 1.73
0.44 34.04 2.98 18.78 4.44 8.50
0.76 36.51 4.06 21.91 4.63 6.38
0.25 36.12 0.82 1.89 0.19 1.13
0.71 36.92 2.79 19.76 4.11 7.38
0.33 56.35 1.84 16.42 2.79 4.79
6.38 2.41 1.32 0.43 3.26 2.65 2.18 1.04 6.53 2.13
3.94 1.47 0.69 0.52 2.99 2.43 1.54 2.98 2.07 -1.15
AlC13 in complex, wt. % Initial Final Composition of products, wt. $70 Propane Isobutane n-Butane Isopentane n-Pentane Neohexane 2-Methylpentane diisopropyl 3-Methylpentane n-Hexane Triptane Dimethvloentanes" 2-Methylhexane 3-Methglhexane n-Heptane Heaviers ComplexedC Total
+
Discussion
products were calculated from the data of Table I and are plotted on a cumulative basis in Figures 2 and 3. The moles of isobutane reacting with 100 moles of n-heptane to produce disproportionation products increased until a t an isobutane to n-heptane charge ratio of about 5 , equimolar parts of isobutane and n-heptane were reacting. The ratio of the moles of pentanes to hexanes produced increased but slowly from 1.62 to 1.93 over the range of 1 to 5 moles isobutane per mole of n-heptane charged. This suggests that the reactions are cocurrent, and proceed through similar, if not identical intermediates. The ratio of the moles of total disproportionation products to moles of reactants is substantially constant a t 1.I to 1 for isobutane-n-heptane charge
0.50 6.50 6.73 2.17 7.49 6.05 0.18 2.54 1.93 0.69 2.41 2.46 0.04 1.57 0.90 0.20 1.46 1.61 0.141 0.82 0.70 0.52 3.20 1.72 f 5.80 3.42 2.55 4.68' 2.39 2.06 8.60 3.55 4.50 1.32 1.53 8.37 2.41 49.67 0.30 0.40 25.85 1.07 0.66 0.18 6.50 5.20 2.07 7.68 6.66 -3.14 1.17 3.32 -1.86 0.63 0.57 100.00 100.00 100.00 100.00 100.00 100.00
I;:
0.81 0.32 0.06 1.51 6.315~ 22.54 28.12 0.94 -1.06
100.00 100.00 100.00
Reacted hydrocarbon,
% n-Heptane 2,2balance. a
20.7
99.5
+ 2,4-dimethylpentane only.
99.4
5.5
98.2
99.0
Total dimethylgentanes.
VOL. 52, NO. 3
55.4
98.4
91.1
Calculated from -41C1,
MARCH 1960
209
0
w
225 r
'
'I40
I
I
I
W
7-
PENTAN E S
I
e
I
,
/
200 -
I
I
/
e
c5
ISOHEPTANES
-OO
0
I
,HEAVIER,
I
2
Ow-
-
I
"
I
3
4 5 MOLE RATIO OF ISOBUTANE TO N-HEPTANE Figure 1. lsoheptanes yield increases rapidly as concentration of isobutane in the feed is increased 0
Hexanes
c0
OO
+
I 2 3 4 5 MOLE R A T I O OF ISOBUTANE TO N-HEPTANE
Figure 3. Composition of disproportionation products is but little affected b y increase in isobutane concentration above a mole ratio of 1 to 1
rc
0
z v)
W.
References
xloe
=g
ki=5 0 -
-
on
u o w a a n
MOLE RATIO OF ISOBUTANE TO N-HEPTANE
w
v)
-I
0
ZS
Figure 2. As the concentration of isobutane in the feed i s increased relatively more isobutane takes part in the disproportionation reaction
ratios above 2; this indicates that heavier products tend to disproportionate. A carbonium ion mechanism for the disproportionation of heavy paraffins catalyzed by aluminum chloride was proposed by Pines and others (7 7) and Schneider (72). T h e first step in this mechanism is the conversion of n-heptane to tert-heptyl ion by transfer of a hydride ion to another carbonium ion, followed by methyl and proton shifts. T h e tertheptyl ion may then either accept a hydride ion from a neighboring paraffin to form isoheptane and a new carbonium ion ( 2 ) , or it may lose a proton to form the corresponding olefin. Interaction
210
of the olefinic species with a tert-butyl or other carbonium ion leads to disproportionation products. T h e hydride exchange reaction of the bcrt-heptyl ion thus leads to the formation of heptane isomers, while proton loss leads to disproportionation. Isobutane seems to decrease disproportionation by increasing the probability that the tertheptyl ion will undergo hydride exchange rather than proton loss by supplying a large concentration of exchangeable hydrogens in the neighborhood of the ion. T h e relative amounts of pentanes and hexanes in the disproportionation products probably reflect the relative stabilities to proton loss of these two ions.
INDUSTRIAL AND ENGINEERING CHEMISTRY
(1) Alden, R. C., Frey, F. E., McReynolds, L. A., Hepp, H. J., Oil Gas J. 44, 70 (1946). (2) Black, H. S., Pines, H., Schmerling, L., J . -4m. Chem. SOC.6 8 , 153 (1946). (3) Egloff, G., Hulla, G., Komarewsky, V. I., "Isomerization of Pure Hydrocarbons," Reinhold, New York, 1942. (4) Evering, B. L., Fragen, Weems, Chcm. Eng. News22, 1898 (1944). (5) Evering, B. L., d'ouville, E. L., J . Am. Chem. SOC. 71, 440 (1949). (6) .Evering, B. L., d'ouville, E. L., Lien, A. P., Waugh, R. C., 111th Meeting, ACS, Atlantic City, N. J., 1947. (7) Hale, M. N., IND. END.CHEM.,ANAL. ED. 18, 568 (1946). (8) Ipatieff, V. N., Schmerling, L., IND. ENC.CHEM.40, 2354 (1948). (9) Mavity, J. M., Pines, H., Wackher, R. C., Brooks, J. A., Ibid., 40, 2374 (1948). (10) Montgomery, C. W., McAtees, J. H., Franke, N. W., J . Am. Chcm. SOC.59, 1768 (1937). (11) Pines, H., Kvetinskas, B., Kassel, L. S., Ipatieff, V. N., Ibid., 67, 631 (1945). (12) Schneider, A., Ibid., 74, 2553 (1952); 73, 5021, 5024 (1951). (13) Sensel, E. E., Goldsby, A. R., IND. E N D . CHEM.44, 271 6 ( 1 952). (14) Thompson, R. B., Chenicek, J. A., Ibid., 40, 1265 (1948). RECEIVED for review September 13, 1959 ACCEPTED December 7, 1959 Division of Petroleum Chemistry, 135th Meeting, ACS, Boston, Mass., April 5-10, 1959.