The Battle Over Hydrogenation (1903-1920)
The Authors: Gary List and M.A. Jackson
The Journal of Industrial Engineering Chemistry(Volume 13, February 1921) carried, in part, the following note: “The United States Supreme Court, on December 6, 1920, handed down an important decision in favor of the defendant, in the ‘Hydrogenated Oil Case’ of Procter & Gamble vs. The Brown Company (formerly Berlin Mills Co.) reversing the Court of Appeals and holding with the District Court that claims 1 and 2 of the Burchenal patent no. 1,135,351, assigned to Procter & Gamble Company, are invalid.” Ninety years have passed since this landmark decision, and this article reviews the chain of events leading up to it.
“Crisco vs. Kream Krisp”
Hydrogenation of organic substances in the gas phase was discovered by the Frenchman Paul Sabatier in the latter part of the 19th century, and applications to the liquid phase had been patented by Wilhelm Normann, the German chemist, both in Britain and Germany in 1903. The British firm of Joseph Crosfield & Sons, who acquired the rights to the Normann patent, began an extensive campaign to interest European soap manufacturers in the technology, hoping to sell the patent rights for profit. According to Joseph Wilson (The History of Unilever, in two volumes), the struggles over the Normann patents in Europe were equally intense as the litigation in America that followed.
By 1907, Crosfield & Sons had carried out sufficient research to bring their hydrogenation technology to the United States. By November 1907, Edwin Kayser, a Crosfield chemist, took up residence in Cincinnati, Ohio, and shortly thereafter, contacted John Burchenal, business manager of Procter & Gamble. By early January 1908, a business arrangement had been made. Kayser then filed two U.S. patents applications (U.S. 1,004,034, Process for Making Metallic Catalyzers, granted September 11, 1911, and U.S. 1,004,035, Method for Saturating Fatty Acids or Their Glycerids with Hydrogen, granted September 26, 1911). Both patents were assigned to Procter & Gamble Co., Cincinnati, Ohio. There can be little doubt that, at this point, that the goal was to completely harden oils for the express purpose of producing raw materials for soap making. The work was kept secret except for Burchenal, Kayser, and Wallace Macaw, whose company (Georgia Mills and Elevator Company) had been acquired through a merger with Procter & Gamble. By late 1910, John Burchenal had filed two U.S. patent applications (Food Products 1,135,351 and 1,135,951, both granted April 13, 1915). Note that about five years elapsed from the time that the patents were applied for until their issuance. The former is based on partial hydrogenation of cottonseed, while the latter is based on blending completely hydrogenated cottonseed oil with liquid cottonseed oil. Originally, the product was named “Krispo” followed by “Cryst,” but settled for "Crisco" (an acronym for crystallized cottonseed oil) because of trademark and for obvious religious reasons. “Krispo” was similar to a trademark registered to Schwarzchild and Sulzberger, no. 40,317; Cooking Fats and Lard.
By June 1911, Crisco Shortening appeared on grocery store shelves and was the object of a massive advertising promotion. Housewives of that era cooked with lard and butter. Thus, Procter & Gamble faced the challenge of convincing them that the new product was superior. It was advertised as a healthier alternative to animal fats and more economical than butter. In addition, a cookbook containing 615 Crisco recipes was distributed free of charge to American housewives. Advertisements in the Ladies’ Home Journal touted Crisco as better than butter for cooking, as well as giving lighter and flakier fried foods. In 1911, $180,000 was spent on advertising and about $400,000 per year in the 1912-1915 time frame. Some 3,500 circulars and products were furnished to jobbers and, by the winter of 1911, Procter & Gamble had shipped cases of 6-1½ lb containers to about 15,000 retailers in America. Crisco became popular almost immediately. Sales in 1912 amounted to 2,600,000 lb (1,180,000 kg) and had reached 60,000,000 lb by 1916.
The plaintiff, the Berlin Mills Company, was located in northeast New Hampshire in the town of the same name. By the mid-19th century, Berlin Mills had become a thriving business town because of timber, water, and rail facilities. The company produced paper from local forests and produced hydrogen as a by-product. Hugh Moore had become Chief Chemist at Berlin Mills Company in 1903 and, by the fall of 1913, began work to use hydrogen for the production of edible products. Within a year, Moore filed two patent applications (Process and Apparatus for Hydrogenating Unsaturated Compounds and Methods and Means for Hydrogenating Unsaturated Compounds). Ultimately, the patents were issued; 1,121,860 on December 22, 1914, and 1,184,480 on May 23, 1916.
By September 1914, Berlin Mills had begun marketing a shortening product later trademarked (105,415, July 20, 1915) as “Kream-Krisp.” Initial production began at about 2,000 lb/d, but later had reached 30,000 lb/d, as will be discussed in detail later. Crisco was produced in a batch hydrogenation mode much like today, whereas Kream-Krisp was produced by a continuous process.
Kream Krisp was marketed as a wonderful new vegetable shortening. “Not only is Kream Krisp a pure food, wholesome, appetizing, better than butter for all good cooking, but is also a great money saver. It will cut your lard and butter bill squarely in two.” A loaf cake (all ingredients) baked with Kream Krisp cost about 15¢ in 1915 compared with about 24¢ using butter. The cost of Kream Krisp alone was 16¢/lb compared with 44¢/lb for best dairy butter. Kream Krisp had the appearance of a high-grade unsalted butter. Scraped surface heat exchangers for crystallizing and plasticizing shortenings were not developed until the late 1930s. The technology of this era consisted of chill rollers heretofore used for plasticizing lard and lard/vegetable oil compound shortenings.
A newspaper clipping of June 1915 described foods including doughnuts, cake, muffins, and pretzels made with Kream Krisp as delightful and the secret was attributed to the shortening. A Kream-Krisp booth, set up at the Maine State Exposition in Portland, Maine, in June 1915, distributed nearly 12,000 doughnuts over a week’s time.
By February 1915, Procter & Gamble was aware that Crisco Shortening had a strong competitor. John Burchenal drafted a polite letter to Berlin Mills Company to the effect that Kream-Krisp infringed on the Procter & Gamble patents, under which Crisco was being manufactured, and suggested that the two parties meet to discuss the matter. Despite numerous attempts on Procter & Gamble’s part to meet with Berlin Mills during the period February–September 1915, the law firm of Allen and Allen, representing Procter & Gamble, began a suit claiming infringement of the Burchenal patent 1,135,351.
Procter & Gamble Sues Berlin Mills In US District Court
Litigation of the Burchenal patent US 1,135,351 began on September 23, 1916, in the US District Court, southern district of New York. Procter & Gamble, as plaintiff, was represented by the firm Allen and Allen, while Barrow and May represented as defendent Berlin Mills Co. From the outset it was clear that Berlin Mills would show that if, indeed, anything was patentable, it was the invention of Edwin Kayser, not John Burchenal.
The patent in question had seven claims, of which only one and two were contested. Claim one: “A homogenous lard-like product consisting of incompletely hydrogenised [sic] vegetable oil” and claim two: “A homogenous lard-like product consisting of incompletely hydrogenised [sic] cottonseed oil.”
The 55-year-old John Burchenal, who had been business manager at Procter & Gamble since 1886, was the first defense witness for Berlin Mills. Mays very carefully attempted to show that Burchenal was not a chemist, nor did he have any technical knowledge of the hydrogenation process or the relationships between the physical and functional properties of hydrogenated oil. Marcus Allen repeatedly objected not only to this line of questioning but also to references to the Kayser patents as being immaterial.
Mays showed that Procter & Gamble’s hydrogenation process was a batch system, whereas Berlin Mills used a continuous hydrogenation process and, although the products, Crisco and Kream Krisp, had about the same melting point, their compositions were different. Analysis of Crisco showed that it was formulated by the hydrogenation of cottonseed oil to an iodine value (i.v.) in the 65–82 range, whereas Kream Krisp was hydrogenated to a lower i.v. and then blended with additional liquid oil. Crisco showed the following composition: saturates 20–25%, olein [sic] 65–75%, linolein [sic] 5–10%, melting point 33–40°C; Kream Krisp: saturates 28%, olein [sic] 34.3%, linolein [sic] 37.7%, melting point 35.1°C, i.v. about 95.
The trial lasted about a year, and the opinion was delivered October 3, 1917; it ordered (i) the Burchenal patent suit US 1,135,351 was void for lack of invention, (ii) claims one and two of said patent were not infringed by the defendant, and (iii) the bill of complaint be and same were dismissed, with court costs to the defendant to be taxed by the clerk of the court.
Procter & Gamble Appeals US District Court Decision
Procter & Gamble appealed to the US Circuit Court of Appeals for the Second Circuit on the basis that Burchenal’s patent was valid. The appeal was heard on October 10, 1918, with Judges Henry Galbraith Ward, Henry Wade Rogers, and Charles Merrill Hough presiding. In writing for the majority, Judge Hough said:
“The patent declares this (Crisco) is a description of a visible, tangible thing which, for some time, has been manufactured and sold by plaintiff, as to which, there is no evidence that anybody else ever made it before, or that, if this product is entitled to patent protection, there is clearly related prior art.
“Invention [i.e., the Burchenal patent] is denied, on the ground once taken by an examiner in the office, namely that ‘ If the problem of simulating lard were presented to an oil chemist, an incomplete hydrogenation of the cottonseed oil would, at once, suggest itself to him as a solution to the problem.’ That is (a), the matter is so obvious as not to rise to the dignity of invention.
“Another objection is that the hydrogenation of vegetable oils is not new. Prior to Burchenal’s effective date, it is admitted that the catalytic hydrogenation of an oil had been practiced, at least, in well-known laboratories and a hard stock produced, solid at ordinary temperatures and showing, on analysis, a very large percentage of stearic or palmitic acid. It is obvious that, if one starts with cottonseed oil, which is liquid at room temperature, because it has too little solid fat in it, and, by chemical means, so changes the molecular composition or arrangement of the substances as to increase the ratio of solid fat, there must have been a time during development of the process when the union of hydrogen had progressed far enough to convert the liquid to a semi-solid.
“Therefore, it is said (b) that no man is entitled to a patent upon the thing or product which has always been produced when the process of making another thing or product was half done.
“The third objection to invention is substantially this: The merit or value of what Burchenal claims and what the plaintiff makes and sells looks like lard, acts like lard, and can be used for the purposes of lard without offending the conservatism of chefs, housewives, and maid servants. But, before Burchenal, many inventions were made by mechanically mixing hard animal fats and cottonseed oil in varying proportion, and some of these mixtures show, upon analysis, substantially the same chemical characteristics as are shown by Burchenal’s chemically produced, homogenous semi-solid. Therefore, it is said that to make the same thing as had been made by earlier lard imitators, but in a different way, cannot warrant a patent upon the resulting thing or product, whatever may be produced, may be true, in respect of the process, by which that product is reached. This is as much to say that the Burchenal product, when completed and ready for use, must be old, because other men had earlier arrived at the same chemical result by other paths.”
Judge Hough continued:
“Objection (a) raises the question of fact encountered in a large proportion of patent causes and, concerning which discussion, is of small value if the record discloses no one ever tried to do the same thing in the same way. When novelty, in that sense, appears the question, really is one of measuring foresight by hindsight. The problem seems easy now, but when that object reached was desirable, useful and apt for commercial success, the bold fact that nobody ever did it before is persuasive, though, and not conclusive evidence of some invention. Burchenal’s imitation lard had these attributes and we consider it a sufficient answer to the statement that any oil chemist could have done the same thing to note that no oil chemist did do it during the more than score of years prior to Burchenal’s application when cottonseed oil (especially) as an abundant American product was endeavoring to supplant lard in the American market.
“The next objection to invention (b) really denies the possibility of invention ever residing in noting or discovering a use for something which, if naturally a by-product, may be termed a halfproduct or unfinished product of an existing method or procedure. Without resorting to the extreme doctrine of Potts vs. Creager (155 US 597), it seems to us that the question presented by this record depends on whether the thing produced by partial hydrogenation is a different thing from that which existed before hydrogenation began and that which would exist when it ended.
“The patent law does not speak in terms of science, though scientific evidence is necessary for the application of its rules. The chemical composition of steam, water, and ice is the same, but they are different things, and, in the same common sense, any oil, lard, and stearin are different things, although (with some latitude) the oil may be said ultimately to become stearine and to pass through the lard stage on the way. For substantially the same reasons we think there is nothing in the last (c) objection to invention. It may be assumed as true that by the mixture of cottonseed oil and animal stearine, a substance can be produced which, for practical purposes, is the same thing as Burchenal’s chemically changed cottonseed oil, but one is a mixture and the other is not, assuming the differences to be unimportant from the stand point of either chemist or cook, it is a vital difference from that of the law. We are therefore of opinion that there was invention in Burchenal’s disclosure.”
US Supreme Court Resolves The Issue
The case was argued before the US Supreme Court on November 15, 1920, and decided on December 6, 1920. The opinion of the court was delivered by Judge William B. Day. In essence, the court ruled with the district court and reversed the appellate court decision. The full text of Judge Day’s opinion can be found online at http://supreme.justia.com/us/254/156/case.html.
Judge Day was no stranger to controversial cases. By the early part of the 20th century, Congress began use of the commerce clause of the US Constitution to ameliorate social problems of which child labor was one. Others include interstate gambling, adulteration of food, and prostitution. Thus, in 1916, Congress enacted the Keating-Owen Child Labor Act, which used the commerce clause to bar goods made by children from interstate commerce. The law prohibited interstate commerce of any merchandise made by children under the age of 14 or that had been made in factories where children between 14 and 16 years worked for more than 8 hours, worked overnight, or worked for more than 6 days a week.
Roland Degenhart, who worked in a cotton mill in Charlotte, North Carolina, along with his two sons, sued, arguing that the Keating Act was unconstitutional. In a bitterly divided court (5–4), Judge Day, writing for the majority, stated that Congress has no power under the Constitution to force child labor laws on the states and to do so would destroy the federalism established under the 10th amendment.
Judge Day, 59th associate justice, authored 439 opinions while on the court (1903–1922) and only 18 dissents. His record clearly shows he distrusted large corporations and voted with the anti-trust majority in a number of landmark cases, including the breakup of Standard Oil under the Sherman Anti-Trust Act of 1890.
In 1920, Judge Day could not have possibly known what nearly 90 years of history has revealed regarding the edible oil industry. In the 1920s, soybean oil in the United States was nothing more than a laboratory curiosity, and it would take another 20 years before soybean oil became a replacement for cottonseed oil.
The decision by the US Supreme Court to uphold the Burchenal patent would have given Procter & Gamble a monopoly on the shortening industry and the use of hydrogenation to manufacture edible products for 20 years. Although Judge Day was not about to let that happen, history does vindicate his decision. It would take another 30 years or so, but hydrogenation would become the backbone of the edible oil industry for many years.
The patent law has four requirements. An invention must be (i) statutory, (ii) new, (iii) useful, and (iv) nonobvious or extending beyond the prior state of the art. The reading of the transcripts from the trial clearly show that the knowledge of hydrogenation in the 1916–1920 era was very meager indeed and very little prior art existed. Technical experts of the day viewed cottonseed oil as linoleate glyceride being converted to oleate, or directly to stearate. There is no evidence that the chemists of the day recognized that isomerization to elaidic acid was occurring or that mixtures of elaidic, stearic, and palmitic glycerides account for the functional properties of hydrogenated fats.
Hugh Moore, chief chemist at Berlin Mills, published the first paper on hydrogenation to be found in the open literature (Ind. Eng. Chem., 9, 451–462 (1917)). He concluded that little, if any, selectivity occurs during hydrogenation of cottonseed oil with nickel catalyst. By 1924, A.S. Richardson, Procter & Gamble, reported in the same journal that the analytical methods used by Moore were faulty and, indeed, selectivity occurs. That is, the conversion of linoleate to oleate occurs faster than oleate to stearate.
It should be noted that compound shortenings or lard substitutes were on the market some 25–30 years prior to the Crisco court case. Most notable were Cottolene and Cottosuet prepared by blending liquid cottonseed oil (70–80%) with beef stearines (20– 25%). Cottolene was first sold by N.K. Fairbanks in 1887, and Swift introduced Cottosuet in 1893. Both products remained popular well into the 1930s. Thus, by 1920, the issue of new (novelty) was no doubt influenced by the widespread popularity of these compound shortenings.
Acknowledgement: This document is based on a two-part article published originally in Inform, June 2007, Vol. 18 (6) 403-405 and June 2009, Vol. 20 (6) 395-397.