Ketene Production and Utilization

EXPERIMENTAL STUDY
G. H. MOREY
Commercial Solvents Corp., Terre Haute, Ind.

HE production of ketene could be put through the pyA study of the preparation of ketene and rolysis chamber was thus deby the pyrolysis of aceits utilization in the manufacture of ortone and its utilization as creased. Another entirely unganic esters was made. Some improvean acetylating agent and in variexpected result was found in ments in apparatus for the production and ous organic syntheses have rethe fact that butyl alcohol diutilization of ketene have been effected. ceived considerable study durluted with considerable butyl ing the past few years (W,Q,6, It was found that some of the statements acetate would not react with ketene at all, although tem8, 9). Hurd and Tallyn (6) apin the literature relative to the reactivity pear to have been the first to peratures ranging up to the of ketene are slightly misleading. For state that the yield of ketene boiling point of butanol were instance, ketene would not react with from the pyrolysis of acetone deused. butanol which was diluted with considerpends mainly on the fraction Apparatus of acetone decomposed per pass able butyl acetate unless a catalyst was through the reaction chamPYROLYSIS CHAMBER. The used. As the acetylated product accumuber. It was found that the pyrolysis chambers used in lates, the speed of reaction of ketene with smaller the fraction of acetone most of the work on ketene debutanol was reduced inordinately. This decomposed per pass, the larger scribed in the literature usually is probably true of other alcohols. the yield of ketene. This is consisted of glass or quartz equivalent to saying that a short tubes, since many metals cataSulfuric acid served as an excellent contact time in the pyrolysis lyze the decomposition of ketene. catalyst for the reaction between ketene of acetone is necessary to obAside from the fact that quartz and alcohols. Even tert-butanol, which tain good yields of ketene. The is expensive and fragile, there is has been reported as reacting with ketene contact time used in Hurd and also the, objection that it is a only to a very slight extent, was esterified Tallyn's work was considerpoor conductor of heat. I n a ably less than one second for pyrolysis where i t is necessary with ease when sulfuric acid was used as t e m p e r a t u r e s r a n g i n g from to bring the gases to reaction catalyst 695" to 735" C. ( 3 ) . Bydecomtemperature in as short a time posing less than 10 per cent as possible, good heat transfer of the-acetone per pass; Hurd and Dull (6) obtained yields of is important. It is known that cbpper (I, 7)is one of the few ketene in excess of 70 per cent of theory. Rice made a sysmetals that is noncatalytic toward ketene. tematic study of the factors controlling the yield of ketene and showed that the yield depends solely on the fraction of The pyrolysis chamber used in the resent work consisted of acetone decomposed. The yield of ketene was found to a copper tube, "18 inch i. d. and 8 " incx 0. d., which was driven approach 100 per cent as the fraction of acetone decomposed into a half-inch stainless steel pipe, 20 inches long (Figure 1). Both the copper tubing and the stainless steel ipe are standard approaches zero (8). Rice also explained the mechanism of commercial sizes. However, unless the stainyess steel pipe is the decomp6sition of acetone to produce ketene on the slightly oversized, its internal diameter must be increased a few basis of the formation and action of free radicals (8). thousandths of an inch before it wilI take the copper tube. This Before the present experimental study was undertaken, can be done conveniently by dissolving out some of the metal with hydrochloric acid. There is such a snug fit between the a search of the literature disclosed that many desirable imcopper tubing and the jacket that no air oxidation of the copper provements could be made in the design and construction of can result at the high temperatures used. The pyrolysis a laboratory apparatus for the production and efficient utilizachamber was placed in a vertical electric furnace whose heating tion of ketene. For this reason the apparatus finally decore was 1 inch in diameter and 20 inches long. At the top, inlet end of the pyrolysis chamber, a copper thermometer well, veloped is described here in some detail. As the work W , was silver-soldered into it. At the bottom of this well, small progressed, it was found that some of the statements in the projections extended to the chamber walls to keep it centered. literature regarding ketene are slightly misleading. For one A quarter-inch 0 . d. copper tube to act as inlet for the acetone thing, ketene is not absorbed as avidly by water and alcohols was also silver-soldered into the side of the pyrolysis chamber about an inch frdm the top end. A similar piece of copper tubing as one would suppose. It was found, for instance, that when was also silver-soldered to the bottom, exit end. the ketene-containing gases (freed from most of the acetone) GENERAL ASSEMBLY,The complete apparatus is illustrated were passed into water through a sintered glass bubbler, conin Figure 1. The reservoir from which acetone was fed into the siderable ketene escaped absorption. This was even the case system is the 2-liter graduated cylinder, G. From here the acetone flowed into the constant leveling device, L, which ensured when two such bubblers were used in series. The sintered a constant and uniform flow of acetone. The reservoir was glass disks were a n inch in diameter and furnished a very filled with acetone in the following manner: The rubber tubing fine spray of bubbles. Aside from the inefficiency of this was removed from the short piece of glass tubing, H, and a type of absorption apparatus, additional objection was found funnel was inserted. Pinch clamp PI was closed, and rubber in its great resistance to gas flow; the rate at which acetone nipple N was removed from the T tube and replaced by a length

1129

1130

INDUSTRIAL AND ENGINEERING CHEMISTRY

THE

VOL. 31, NO. 9

TABLEI. RESULTS SHOWING THE YIELDS OF BUTYL ACETATEOBTAIXABLE FROM

Run No.
14 15 16 17 19 20 21 22 24 25 26 27 30 31 32 33 34 35 36 37 40

CONVERSION OF ACETONE TO KETENE

--l-ButanolDown Down column column
C1

Temp. ' C.
670 680 720 715 670 720 720 720 715 715 715 715 715 665 670 670 715 700 650 710 716

Flow Rate of Acetone Cc./hr.

880 880 843 750 466 50 1 540 850 687 703 685 719 678 85 1 571 668 670 651 670 567 695

Acetone Used

cc.
1320 1320 1265 1125 852 877 945 1487 1374 1407 1475 1439 1357 1490 1001 1336 1454 1302 1340 1133 942

-Recovered AcetonFrom From From reflux column column condenser CI CZ cc. cc. cc. 1195 55 8 1195 57 8 1031 92 20 965 70 7 56 718 8 76 37 607 75 56 593 1082 81 76 98 28 1070 1106 108 23 1020 94 1144 110 zi 1024 104 41 1264 24 74 24 81 770 82 38 1048 88 15 1235 13 1165 60 14 1100 93 34 91 820 11 90 720

Fraction of Acetone Deoomposed

%
4.56 9.6 7.4 8.2 17.9 23.4 16.7 13.0 12.1 10.5 11.4 13.5 8.6 12.6 12.5 8.0 4.9 9.6 16.6 12.8
4 -.. 7 .

---Butyl Acetate-From From column column CI Cz -----Qrams----..

87.45 85.73 149.3 119.3 95.81 171.7 214.35 250.1 201.1 186.5 148.8 188.4 191.4 140.0 129.2 194.8 124.8 87.78 157.5 172.4 128.4 0.45 1.65 6.7 1.7 1.52 8.7 12.39 15.8 2.7 4.6 1.2 3.5 22.6 13.4 20.1 17.6 11.8 0.39 5.8 30.3 10.8

c z

-Cc./hr.560 660 560 560 560 560 560 610 550 640 640 580 580 580 520 470 500 510 670 480 830 300 300 300 300 300 300 300 420 370 430 430 440 470 470 430 410 470 460 490 410 390

Yield of Butyl Acetate Mole %

88 93 82 93 90.3 73.6 65.7 68.6 73.3 72.0 73.0 75.0 75.0 76.8 76.5 81.0 86.5 88.4 82.0 69.1 73.0

of rubber tubing attached t o a vacuum aspirator. Acetone was then poured through the funnel into the constant leveling device from which it was transferred into the graduated cylinder by suction. After the reservoir was filled in this manner, the acetone level was brought t o the point in the constant leveling device shown in the drawing, care being taken not to allow any acetone to rise in tube S . The volume of acetone in the graduated cylinder was then read and the temperature noted. About 5 minutes before a run was started, the substance with which the ketene was t o react (butanol, ethanol, water, etc.) was started flowing down absorption columns CI and Cz. 1-Butanol was used for most of this work. This was done by applying air pressure on the surface of the liquid in the storage flasks a t the top of the columns. A mercury blow-off tube and manometer (not shown in the drawing) were attached to each of these storage flasks. After the furnace was brought t o the proper temperature, the acetone flow was started by opening pinch clamp PI, closing pinch clamp PB,and applying pressure with nitrogen gas at A . The pressure was regulated by the mercury blow-off tube, B, and read on the manometer, M . The acetone was forced up the capillary tube, D , and into the pyrolysis chamber through orifice 0. A sintered glass filter, F , was sealed t o the bottom of tube D in order t o prevent the plugging of orifice 0 by small solid particles. A spray of water on the outside of the exit end of the reaction chamber cooled the hot pyrolysis products as they emerged from the furnace, whence they were introduced into the bottom of a reflux condenser. A 2-liter flask was attached t o the bottom of this reflux condenser by a ground-glass joint held firmly together by strong rubber bands. The flask was heated by a hot plate. The unconverted acetone in the pyrolysis products was condensed out and flowed into the flask, where it was ke t boiling by the hot plate. No ketene would dissolve in the ioiling acetone, and it passed out of the condenser along with methane and small amounts of other gases. These other gases have been shown (8) to consist of carbon monoxide, ethylene, and hydrogen. A small amount of acetone vapor also escaped from the reflux condenser, but it was always absorbed in the stream of 1-butanol flowing down columns C1 and CZ. These columns were inch i. d. and 30 inches long, and were packed with single-turn glass s irals (10). Since the spirals were originally wound on l/,-inc! steel rods, they were leached with hydrochloric acid to remove any iron with which they may have become contaminated. On entering absorption column CL, the gaseous ketene reacted rapidly with the 1-butanol t o form but 1 acetate. Any ketene which escaped from column C1 was caugzt in column Cz and the product collected in the 2-liter flask a t the bottom of this column. As a final precaution t o prevent the escape of even a trace of ketene, a coarse-mesh sintered-glass bubbler, 2, containing aniline in benzene, was attached after column CS. Usually only a negligible amount of ketene reached this bubbler. To stop a run, the volume of acetone in the reservoir was first noted; then pinch clamp PS was opened, and nitrogen allowed t o flow through the apparatus in order to flush out any ketene from the reaction chamber and reflux condenser. The flow of butanol down the absorption columns was continued for a t least 15 minutes longer, in order t o wash down all butyl acetate into the collecting flasks.

Temperature Distribution in Pyrolysis Chamber

The internal temperature and the wall temperature of the pyrolysis chamber varied considerably throughout its length; it was relatively cool at the inlet and where acetone was vaporhing and became hotter toward the interior, u p t o a point about 4 inches from the exit where the hottest zone was usually located. From this point on toward the exit, the temperature fell off rapidly. There was also a considerable difference between the wall temperature and that measured a t the center of the reactor. In order to measure the wall temperature accurately, a pyrolysis chamber was made like the one described, except that the stainless steel jacket was slit lengthwise on one side. The width of the slit was 1/8 inch, which gave sufficient room for a thermocouple which could be raised and lowered. This slit was covered with thick asbestos paper to shield the thermocouple from direct radiation from the furnace walls. The end of the thermocouple was bent so that a tension held i t against the wall of the copper pyrolysis chamber and thus ensured metallic contact. The temperature distribution for a typical run is as follows:
-At
Center of TubeI n . from inlet 8 10 11 12 13 14 15 16 17 1s

c.
520 533 635 657 670 670 663 645 616

Copper Wall Temp.

c.

...

615 640 677 702 720 727 723 706 675 623

From the nature of the temperature distribution in the pyrolysis chamber, it is evident that any calculation of reaction time a t a certain temperature level would be a very arbitrary procedure. Reaction time over a range of temperatures could be calculated, however. The temperature of a run was taken as that of the hottest zone measured in the thermocouple well down the center of the tube. Thermocouple T , located on the outside of this hot zone, led t o a photoelectric relay which regulated the temperature.

Results
The results for several runs are set forth in Table I. About 1000 t o 1500 cc. of liquid acetone were passed through the pyrolysis chamber for each run. The unreacted acetone

SEPTEMBER, 1939

INDUSTRIAL AND ENGINEERING CHEMISTRY

1131

which was recovered in the flask a t the bottom of the reflux condenser was measured in a graduate a t the temperature of measurement of the acetone in the reservoir. The acetone which escaped from the reflux condenser and was absorbed in columns C1and Czwas determined by Messenger's method. The ketene was absorbed in butyl alcohol to form butyl acetate. The amount of butyl acetate was determined by esterification with a measured amount of standard alkali and back titration with standard acid. From Table I it is evident that considerable ketene escaped absorption in column C1. (At first, only one absorption column was used. Bubbler 2 then quickly became clogged with crystals of acetanilide.) A great excess of butyl alcohol was always used, and the proportion of butyl acetate in the mixture collected from column CI was never greater than 25 per cent. Repeated analyses of the coiitents of bubbler 2 did not show any ketene or acetone escaping from column Cz, however. In Figure 2 the mole per cent yield of ketene is plotted against the fraction of acetone decomposed per pass through the pyrolysis chamber. The yields check those of Rice ( 8 ) ,although most of his work was done with a low partial pressure of acetone

in the pyrolysis chamber and nitrogen or hydrogen was employed as a carrier gas. When a small fraction of the acetone is decomposed per pass (say 10 per cent), the recovered acetone in the flask a t the bottom of the reflux condenser will be water-white but will possess a rather pungent odor. This is due to a small amount of uninvestigated impurity, presumably acetic acid and acetic anhydride. The pungency and the acid content increase inordinately when the condensate is put through the pyrolysis chamber a second and third time. One hundred cubic centimeters of the first condensate consumed 2.10 CC. of 0.99 N sodium hydroxide solution. With one and two recyclings, the values for 100 cc. of condensate increased to 8.45 and 16.47 cc. of 0.99 N sodium hydroxide, respectively. Apparently if the impurity is left in the recycled acetone, it acts in such a manner as to increase greatly the amount of impurity which forms on the next passage through the pyrolysis chamber. It is likely that after the acetone had been recycled eight or nine times, the amount of impurity would be considerable. This means that the recycle acetone must 'be fractionated before it is used again. With this method of

AND UTILIZATION OF KETENE

rw,-COPPER
REACTOR AIR PRESSURE

BUTANOL

WATER

S P R A Y (

D R A I N L -

1
\

1132

INDUSTRIAL AND ENGINEERING CHEMISTRY

THE

VOL. 31, NO. 9

TABLE 11. RESULTS OF RUNSON

Run No. Compn. of Mixt. Fed down Column 1 Butyl aoetate 1-Butanol

RATEOF ABSORPTION OF KETENE BY BUTANOL

Compn. of Product from Column 1 Butyl acetate 1-Butanol -Butyl AcetatFormed in Formed in column 1 oolumn 2
Crams Crams

IN

MIXTURES RICH IN BUTYL ACETATE

Temp. of Column 1

%
43 44 45 46 47 48 67.8 74.0 74.0 74.1 74.1 74.1

%
32.2 26.0 26.0 25.9 25.9 25.9

c.

%
71.1 74.5 76.4 75.6 74.9 82.8

%
28.9 25.5 23.6 24.4 25.1 17.2

Remarks" Column not heated or cooled externally Column heated externally Some butyl acetate distd. from hot oolumn 1 to column 2 Some butyl acetate distd. from hot oolumn 1 to oolumn 2 Column 1 cooled with water jacket 0.6% sulfurio acid present: some polymer formed water present in mixt. to oolumn 1 sodium acetate present Column not heated or cooled externally 0 24 sulfuric acid and 0 33 water present 0:242 sulfuric acid and 0133% water present 0.24% .. sulfurio acid and 0.33% .. water uresent

50 90 115 120 <20 50

104.7

7.7 ...

64.6 153.0

...
...

...

...
4.5
131.3 52 78 1 0 6 7

... 245

49 74.1 25.9 50 77.8 22.2 151 50 74.1 25.9 60-70 83.8 16.2 227 51 26.2 73.8 50 41.8 58.2 209 54 49.7 50.3 <20 58.4 41.6 188 55 49.7 50.3 <20 60.0 30.0 257 56 73.4 26.6 <20 82.2 17.8 278 0 The peroentages of various materials referred to are for the mixtures fed down column 1.

operating, only a negligible amount of acetone would be lost by conversion into the impurity.

Speed of Reaction of Ketene with Butanol

As the 1-butanol becomes diluted with butyl acetate formed by reaction with ketene, the latter reacts progressively less readily until with mixtures containing as much as 75 per cent butyl acetate, no reaction will occur a t all. The presence of even 30 per cent butyl acetate reduces the speed of absorption of ketene considerably. For the commercial production of butyl acetate from butanol and ketene, this behavior con-

PERC NT

1 0

CONVER~ION PER PAS$ 20 30

FIQURE 2. CONVERSION OF ACETONE us. YIELD OF KETENE

74 per cent butyl acetate and 26 per cent 1-butanol by weight was fed down column C1. I n no case would the ketene react with the 1-butanol in such a mixture. It was then thought that perhaps no reaction occurred owing to lack of physical solubility of the ketene at these higher temperatures, and accordingly a run was made in which the column was cooled with a water jacket. Still no reaction occurred between the ketene and the 1-butanol in the mixture. These results indicate that a catalyst is necessary to cause ketene to react with 1-butanol in mixtures rich in butyl acetate. Several materials were tested for such a purpose, among which were water, sodium acetate, and sulfuric acid. All of these materials exerted a catalytic effect, but the sulfuric acid was by far the most active catalyst. The column should be cooled with a water jacket when sulfuric acid is used as a catalyst. The results of several experiments are set forth in Table 11. I n all cases in which a catalyst was not present (as in runs 44 to 47), the butyl acetate content fed down column C1 did not increase appreciably. Pure 1-butanol was fed down column Cz to catch the ketene which escaped from column C1. However, with a few tenths per cent by weight of sulfuric acid present, very little ketene escaped absorption in column C1. Even tert-butanol was found to react readily with ketene to form tert-butyl acetate when a small amount of sulfuric acid was used as catalyst. Rice (8) found that this alcohol would not react readily with ketene, even when the temperature was raised to just below the boiling point of the alcohol. The diminution of the speed of reaction of ketene as the acetylated product accumulates in the mixture will probably occur with most substances possessing alcohol groups which can react with ketene. This may explain some of the conflicting statements in the literature regarding the activity of ketene as an acetylating agent.

stitutes a serious handicap, since butyl acetate and 1-butanol form a constant-boiling mixture which is 26.2 per cent butyl acetate and 73.8 per cent butyl alcohol by weight. It would therefore be necessary to get a mixture which contained considerably more than 26.2 per cent butyl acetate before any product could be economically fractionated from the constantboiling mixture. I n order to increase the speed of absorption of ketene by the 1-butanol in the mixture, column CI was heated with heating wire which was wrapped around it. Raising the temperature has been shown to speed up the reaction between ketene and a great many substances (8). Runs were made with column temperatures of 75", 90, and 115" C., and in one run the lower third of the column was heated to 120'C. (vi.., above the boiling point of 1-butanol but below that of butyl acetate). In most of these runs a mixture of

Literature Cited
(1) Clarke and Waring, U. 5. Patent 1,723,724 (1929). 45,3095 (1923). (2) Hurd, C. D., J . Am. Chem. SOC., (3) Hurd, C. D., private communication. (4) Hurd and Cochran, J. Am. Chem. SOC., 45,515 (1923). (5) Hurd and Dull, Ibid., 54, 3428 (1932). (6) Hurd and Tallyn, Ibid.,47, 1427 (1925); Tallyn, M.S. thesis, Univ. Ill., 1924. (7) Reid, G. H., U. S. Patent 1,975,663 (1934). (8) Rice, Greenberg, Waters, and Vollrath, J. Am. Chem. Soc., 56, 1760 (1934). (9) R i c e and Vollrath, Proc. Natl. Acad. Sci., 15,702 (1929). (10) Wilson, Parker, and McLaughlin, J . Am. Chem. SOC.,55, 2795 (1933).
P R ~ B E N Tbefore E D the Division of Organio Chemistry a t the 96th Meeting of t h e American Chemioal Sooiety, Milwaukee, Wis.