chemical vignettes ROBERT C. BRASTED Unwenity of Mmnerota Minneapolm, 55455
Reaction Yield
= f(3hape It's In)
Suggested hy: Dr. Alan P. Benz, Stamford, Conn.
The column welcomes suggestions and material from industry since in the opinion of the column editor there are still real deficiencies in academic training for the chemist who plans on an industrial career. Any small contribution that this column can make in bridging this unsolved academic-industry gap might enhance its value. The inevitable role of the industrial chemist whether he he fresh out of his formal schooling (graduate BS, doctorate, or post doctorate) or a veteran in the trade is to produce something salable in the best possible yield. The classical (?) training in mechanisms and rates of reactions as well as the overall research approach may not be adequate for the task. Dr. Benz makes this quite clear in the following. The scale up of a reaction to production quantities may depend upon such a variable as the shape of a reaction vessel. Not many chapters in texts are of help here. A reaction carried out in a round bottom flask in the research laboratory might give entirely different yields in a beaker-shaped kettle in the plant. A few of the variables encountered are; degree of agitation, wall effects, and the ratio of the surface to the total volume of the container. A case in point is described. The plant yields in a vitamin BG synthesis were consistently lower than those in the laboratory even though every effort was made to reproduce conditions defined by the research laboratory. The final step in this process prior to isolation of product was a diazotization of a nitro compound to an amino compound. Due to the elevated temperature used to decompose the diazo compound and the air flow across the surface to the vent, there was significant volume decrease. This change actually was desirable since the reaction was carrried out in a diluted state and later evaporation was necessary. However, when the reaction was returned to the laboratory for further testing by duplicating the plant conditions in shape of container and evaporation the yield dropped to the identical value found in the plant in a large scale operation. The solution to the yield problem lay in the continual dilution to make up the evaporation loss in order to keep the total container surface constant. What appeared to he a money saving evaporation during the reaction was actually impractical. It is not suggested that every degree holder in chemistry he trained as a producing engineer or control 762
/
Journal of Chemicol Education
chemist, but it is suggested that those responsible for the academic instruction in synthesis as well as theory he aware of the fact that there may be non-text hook "happenings."
Chemical Induction, A Potential Plant Hazard Suggested by: Dr. Ahn P. Benz, Stamfwd, Conn.
An earlier column pointed out a need for looking for the unexpected or a t least the non-textbook approach in many industrial scale-up reactions. In some industrial processes the scaling-up operation may lead to unexpected results of frightening dimensions even though following acceptable "rules of the thumb" when applied to laboratory procedures. It is quite possible that a concentration build up of ominous proportions may occur during an induction period when hundreds of pounds of reactants are used in place of a few moles (or grams) typical of the laboratory reaction. Two instances are reported by Dr. Benz. In one of the vitamin production steps there is a nitration reaction using fuming nitric acid dispersed in acetic anhydride. The temperature must be slightly elevated to decompose coproduced tetranitromethane as fast as it is formed. Availability of brine cooling of the reaction vessel would seem logical since we all think of the "old wives tale" of doubling a reaction rate with a 10" rise in temperature. The rule should operate reasonably well if the temperature is lowered. Nitrations are famous for their exothermicity with this vitamin step being no exception. The production process problem lies in the overheating during the induction period which cannot be adequately controlled by brine cooling, necessitating a cut off of the HNOa(CH3C0)20 mixture which step in turn is followed by cooling. The time-temperature log of the system takes the appearance of a razor-back hog's posterior. The system alternately fumed and sputtered then would lie dormant. This is no way to "run the store" in a plant. A simple solution was found in controlling the initiation period for the reaction by further diluting the nitric acid (already in acetic anhydride) with glacial acetic acid. I n the end more useable product is available from a lower concentration of reactants. Another nitration step in vitamin synthesis provides further proof of the danger in automatically assuming that cooling will slow d o h a large scale reaction. In deference to the known exothermicity of nitrations the plant engineers provided a reaction vessel of great durability and with a top secured by some 36 bolts.
