The Flash of Genius, 4 Collected by Alfred B. Garrett
Ohio State University Columbus
I
Freon:
Albert 1. Henne
Thomas Midgley, a young mechanical engineer of the General Motors research laboratory of which Charles Kettering was director, had worked out the problem of developing s n antiknock for gasolinelead tetraethyl. Ten years later (in early 1930) he was a t work on another important problem.' Here is Midgley's account of the discovery of Freon, a. refrigerant now used in most modern refrigerator^:^^^
"I was in the laboratory one morning and called Kettering in Detroit about something of minor importance. After we had finished this discussion, he said: 'Midge, I was talking with Lester Keilholtz last
.
Thomas Midgley and
1 M~nr.r.nv. n o~ ~ n-s Jn.. .~ -~A~ N D ,HENNE.ALBERTL.. Ind. and ...-.---, T ---Eng. Cha.,22,542 (1930). Z M ~ ~THOMAS, o ~ ~JR., ~ I, d . and Eng. Chem., 29, 241-4 (1937). Thiais the Perkin Medaladdress. a K ~ ~D. ~H., ~"The ~ Genius ~ ~ of nIndustrial , Research,'' Reinhold Publishina Corp., 330 42nd St., New York, 1948, pp. 48-51.
night and we came to the conclusion that the refrigeration industry needs a new refrigerant if they ever expect to get anywhere. So I told Lester that I would call you and have you see him to talk it over. He is leaving for Dayton tonight.' L. S. Keilholtz was chief engineer of Frigidaire at that time, we spent the next morning in his office discussing this problem. What was wanted was obvious-a nontoxic, nonflammable refrigerant. On leaving Kleilholto, I expressed myself as very doubtful that we would be able to find a single substance suited to the task but that there might be some hope of greatly reducing the existing hazards by using mixtures where nonflammable but toxic materials would be mixed with nontoxic, flammable compounds to give a mixture substantially nonflammable and considerably less toxic than the refrigerants then commonly employed.
Volume
39, Number 7, July 1962
/
361
"In this frame of mind, I returned to our laboratory where I found A. L. Henne and Robert MacNary waiting t o have lunch with me. We discussed the problem during lunch and I have Henne's word that my skepticism of solving the problem with a single compound intrigued his interest so much that he gave up his free afternoon and he, MacNary, and I went to t,he library and started to work. "The desired combination of properties was a boiling point between 0 " and -40°C, stabilky, nontoxicity, and nonflammability. International Critical Tables gave us a partial summary of the volatile organic compounds. The now-proved mistake that carbon t,etrafluoride boiled a t - 15°C st,ruck us in the face and started us thinking about fluorine. KO one could doubt a t that time that it was terribly toxic, probably too toxic to use even wit.h isobutane. Perhaps we could add some chlorine compound with beneficial result,^. Henne suggested chlorofluorides as a class t,o be investigated further. And so the discussion ran. "Recognizing that t,he International Critical Table list was very incomplete, I decided to bring into play t,he periodic table. Perhaps volatility could be related to it in some way. It takes but a fraction of a second t,o see that this is true. The elements on the right-hand side (of the periodic table) are the only ones which make compounds sufficiently volatile for the purpose in hand. I n fact,, only a certain number of these need be considered. Volatile compounds of boron, silicon, phosphorus, arsenic, antimony, bismuth, tellurium, and iodine are all too unstable and toxic to consider. The inert gases are too low in boiling point. Now look over the remaining elements. Every refrigerant used has been made from combinations of t,hese elements. Flammability decreases from left to right. Toxicity (in general) decreases from the heavy elements a t the bottom t,o t,he lighter element,^ a t the top. These two desiderata focus on fluorine. I t was an exciting dcdnct,ioo. Seemingly no one previously had considered it possible that fluorine might be nontoxic in some of its compounds. This possibility had certainly been disregarded by t.he refrigeration engineers. If the problem before 11swere solvable by the use of a single compound, then that compound would certainly contain fluorine. The heats of formation between the halogens and carbon werc checked. They increase from iodine to fluorine, thus indicating a high degree of stability for fluorine-carbon compounds. Everything looked right except that old fear of hydrofluoric acid burns. As it turns out, hydrofluoric acid is less toxic than hydrochloric acid, but we didn't know it, that afternoon. Next came methods of preparation. Carbon tetrafluoride seemed rather hard to make. And then how
could dichlorodifluoromethane boil a t -2O0C and carbon tetrafluoride a t -15"C? I t just didn't make sense. Plottings of boiling points, hunting for data, corrections, slide rules, log paper, eraser dirt, pencil shavings, and all the rest of the paraphenalia that takes place of t,ea leaves and crystal spheres in the life of the scientific clairvoyant were brought into play. We decided that carbon tetrafluoride boiled a t about - 136°C or else it was a very special kind of substance. (Kot long after this a publication on the subject appeared. Carbon tetrafluoride boils a t -128' not -15°C.) Feeling pretty certain a t the time that, - 1 5 T was wrong and that it was a sizable research problem to make carbon t,etrafluoride, we selected dichloromonofluoromethane as the starting point for experimentation. I called one of the chemical supply houses by telephone and ordered five 1-02 bottles of antimony t,riflnoride. I believe this was all there was in the country a t the time. "The five 1-07.bott,les of antimony trifluoride arrived. One was taken a t random, and a few grams of dichloromonofluoromet,hane were prepared. A guinea pig was placed under a bell jar with it and, much to the surprise of the physician in charge, didn't suddenly gasp and die. I n fact, it wasn't even irritated. Our predictions were fulfilled. We then t,ook another hott:le and made a few more grams and tried it again. This time the animal did what the physician expected. We repeatrd again but this time we smelled the mat,erial first. The answer was phosgene; a simple caustic wash was all that was needed to make it perfectly safe. Then we examined the two remaining bot,tles of antimony trifluoride. They were not pure. I n fact, they were both badly contaminded wit,h a double salt containing water of crystallization. This makes phosgene in ample quantities as an impurity. Of five bot,tles marked 'antimony triflouride,' one had really contained good material. We had chosen that one by accident, for our first trial. Had we chosen any one of the other four, t,he animal would have died as expected by everyone else in the world except ourselves. I believe we would have given up what would then have seemed a 'bum hunch.' "And the moral of this last little story is simply this: You must be lucky as well as have good associates and assist,ants t o succeed in this world of applied chemistry snfficiently well to receive the Pcrkin Medal. "Albert L. Henne deserves fully as much credit as I for devcloping organic fluorides as refrigerants." This is an crcellent oxample of s di~covcrymade as a. result of the combination of guncral experience and seient,ific hunches based on scientific data.
NOTE:These papers are the third and fourth EDITOR'S in a continuing series of stories of discovery told in the ~ 287 words of the discoveren. See THE J O U R X A39, (1962) (June) for previous items.
362 /
lournal of Chemical Education
