1. The furnace used by the submitters and checkers was an 1870-watt hinged type manufactured by the Hevi-Duty Electric Company, Milwaukee, Wisconsin; length 18 in.; inside diameter 2 3/8; in.; catalog No. M-2018. With the packing described, the total surface area in the packed tube is about 2000 cm.² In addition to this setup the checkers used a similar tube (20 mm. inside diameter) packed with 1/8;-in. I.D. single-turn Pyrex^® glass helices and inserted in a 550-watt furnace manufactured by the Hoskins Mfg. Co., Detroit, Michigan; length 12 in.; inside diameter 1 ½ in.; catalog No. FD303A.

2. To prevent heat loss from the ends of the furnace opening, the ends are packed with Pyrex^® glass wool or Fiberglas^® insulation PF-105. In addition, the checkers capped each end of the furnace with a plate constructed from ¼-in. Transite^® sheet in which a hole just large enough for a loose sliding fit on the tube had been bored. The plate at the bottom of the furnace was held in place with a rubber stopper also bored to fit the tube, and the plate at the top of the furnace was held in place with an iron ring attached to the ring stand supporting the funnel.

3. The lower part of the barrel of the dropping funnel should be bent in such a way as to offset it from the path of the thermocouple to permit adjustment of the latter. The checkers used a 8-mm. glass tube bent for offset and fitted at the upper end with a nitrogen inlet and a standard-taper ground joint to permit attachment of funnels of various sizes.

4. Attachment of dropping funnel and thermocouple well to the pyrolysis tube may be made with a rubber stopper suitably bored.

5. The first two traps may be packed loosely with glass wool to prevent mechanically entrained impurities (or aerosol) from passing through into the final trap. Omission of the glass wool may allow as much as 0.5–1.0 g. of colored material to be collected in the product. However, this colored impurity is easily removed by simple distillation. The checkers cooled the two traps in ice-ethanol.

6. The checkers passed the nitrogen through a gas absorption bottle filled to a depth of 10 cm. with concentrated sulfuric acid which had been calibrated roughly for flow rate. The rate of flow used was one bubble per 7 seconds (ca. 145 ml. per hr.) for the larger furnace, and one bubble per 10 seconds (ca. 100 ml. per hr.) for the smaller furnace.

7. Suitable diketene can be obtained from the Aldrich Chemical Company, Milwaukee, Wisconsin. If at all colored, this material should be distilled or sublimed before pyrolysis, for use of colored material may lead to a colored product. Distillation is easily carried out through use of the apparatus illustrated in Fig. 1. The impure diketene is placed in flask A and is cooled in ice and stirred with a magnetic stirrer. The Dewar trap is filled with dry ice and ethanol. Evacuation of the system with a vacuum pump is begun very carefully with appropriate upward adjustment of the pressure if necessary to prevent bumping and splashing. After the initial degassing and removal of low-boiling materials, the diketene distils smoothly (at 0.1–1.0 mm. pressure), collecting as a white solid on the cold surface of the Dewar trap. After the distillation is completed, the apparatus is partially disassembled and the dry ice and ethanol are poured out. As the diketene melts, it flows into flask B. The diketene should be stored under nitrogen in tightly stoppered brown bottles in a refrigerator or, better, below its freezing point of −6.5°, as otherwise it may slowly decompose.

8. Equally good results have been obtained with charges one-fourth to twice that of the present procedure.

9. The ketene obtained by simple distillation from the final trap is essentially pure. The purity can be determined by passing a weighed sample into 1N sodium hydroxide solution and titrating for excess base. The checkers found it very difficult to avoid entirely the collection of minute traces of high-boiling colored impurities with the ketene . However, these impurities were easily removed by simple distillation of the ketene , just before use, from the original trap into a clean trap by warming the former in ice water and cooling the latter in dry ice and ethanol.

10. The checkers' yields were 54–66% for the larger furnace and 58–78% for the smaller furnace. The yield was found to be somewhat dependent on the rate of addition, with the rate specified in the procedure giving a good yield in a reasonable length of time. The checkers used an addition rate of 0.25–0.30 ml. per min. for the smaller furnace to obtain the yields cited.

11. Major impurities that collect on standing are all high-boiling, e.g., diketene, b.p. 127°, and dehydroacetic acid, b.p. 270°. If the ketene has been stored long enough to allow a considerable portion of higher-boiling materials to accumulate, it is desirable to insert a trap cooled in an ice bath between the ketene-containing trap and the reaction vessel in order to minimize mechanical entrainment of the impurities. Repeated warming of the container to remove portions of the ketene naturally hastens transformation of the ketene to high-boiling materials.

12. The carbonaceous material that is deposited inside the pyrolysis tube is easily removed by passing a stream of oxygen through at about 550° after a thorough flushing with nitrogen.

13. The submitters no longer consider this method to be the most convenient one for preparing laboratory quantities of ketene . It has been supplanted at the Du Pont Experimental Station by pyrolysis of acetic anhydride according to the procedure of Fisher, MacLean, and Schnizer.³ It is recommended that the wires of the preheater be glass-covered, since polymerization of the ketene at −78° seemed to occur frequently when bare wires were used, possibly because of catalysis by trace amounts of metal. Using this procedure 55 g. (1.31 moles) of ketene per hour could be produced (private communication, J. B. Sieja and H. D. Carson).