The emergence and success of the US chemical industry has its roots in chemical research, but over the last 40 years, these programmes have been radically reduced, leading to ‘the death of chemical research’. We will look briefly at the history of US chemical research, in an effort to understand the forces that drove it forward and then caused it to shrink away. We will then conclude with a short discussion of the options available for the chemical industry to support the development of new chemical technologies in the future. It is useful, at this point, to define our terms:
• Chemical research: Laboratory efforts that focus on identification and early development of new and innovative chemical technologies.
• Chemical development: Laboratory efforts that focus on incremental improvements in the performance of or the processes for existing chemical technologies.
Research in the chemical industry
The roots of chemical research go back to the early 1800s. The earliest chemical technologies emerged from the efforts of individual entrepreneurs, who saw commercial potential in specific chemical technologies. Charles Goodyear, who in 1836, became fascinated with the potential for rubber, and after hundreds of failed experiments and several failed attempts at commercialisation, accidently discovered the ‘vulcanisation of rubber’; the key to making rubber a commercially viable product. However, due to poor business skills and the loss of European patent rights, he never personally profited from his invention, and died deeply in debt.
Other early chemical developments included the accidental discovery of nitrocellulose by Schonbein in 1846 and the extension of that work to the discovery of nitroglycerin by Sobrero in 1847, and the accidental discovery of the first synthetic dye - Mauve, a deep purple dye - by William Perkin, as he was trying to carry out a university assignment directed at the synthesis of quinine.
By the late 1800s, chemical technologies started to emerge from systematic research and development efforts: 1890s smokeless powder based on nitrocellulose; Rayon (viscose process) from the work of F. S. Cross in the UK; new dye technologies from Germany; Aspirin, the new miracle drug from Bayer in Germany; new photographic technologies from Eastman Kodak; and the synthesis of ammonia/nitric acid/nitrates via the Bosch/Haber process.
In the US, there existed a successful gunpowder company that was on the verge of becoming a major chemical/fibre company that for years ran the most successful chemical research programmes in America. E I du Pont had come to the US in 1802, escaping from France in the midst of the French Revolution. On the advice of Thomas Jefferson, he abandoned his idea of investing in land speculation and after observing the primitive state of gunpowder production in the US, he started a company to manufacture gunpowder using recently developed French equipment and manufacturing techniques. By the end of the 19th century, DuPont moved the technology forward and became the dominant explosives producer in the US.
In 1804, DuPont opened its first gunpowder plant and then expanded production by plant expansion and acquisition in the years that followed.
In the 1857, Lammot du Pont made the first major change in the production of gunpowder in 200 years, reducing the cost and increasing its explosive power. In the 1880s, Lammot du Pont established the Repauno Chemical Company after failing to convince the du Pont family that dynamite was the explosive of the future. DuPont would go on to become the world’s largest producer of dynamite. In the 1890s, DuPont became a major producer of nitrocellulose and smokeless gun powder.
In the early 1900s, DuPont moved into plastic films, molded plastics, varnishes and paints, all based on nitrocellulose. It was at this time that DuPont began to investigate the production of synthetic fibers, initially based on modified forms of cellulose.
At the turn of the century, a broader American chemical industry was starting to develop. Technologies that emerged at this time included Eastman’s pioneering work on photographic and X-ray film technology; Baekeland’s development of thermoset phenolic resins; and Bayer developed North American production facilities for dyes and Aspirin.
In 1903, recognising the expanding opportunities in chemistry and materials, DuPont established the Research Station at the site of an old gunpowder plant on the Brandywine River. Development groups were organised by business area and focused on specific market needs and process improvements.
World War I brought about major changes at DuPont. Along with the rising demand for explosives, the government requested that they develop dyes to replace those no longer available from Germany. While the project was largely a failure, it strengthened DuPont’s foundation in organic chemistry. In the early 1920s, after the war came to an end, DuPont began the transition from an explosives company to a major chemical company.
The company successfully moved into the manufacture of dyes after German dye and chemical technology was made public through the BIOS reports and related programsme. This area established what became the chemical department at DuPont. The company expanded its commitment to coatings, particularly for automotive applications; initially products were based on nitrocellulose followed by the development of the first alkyd resin based coating.
Based on what now was excess capacity for nitrocellulose, DuPont purchased several smaller companies and expanded into the area of moulded plastics and films based on nitrocellulose. In 1920, based on its experience with cellulose-based technologies, DuPont licensed European developed rayon technology and expanded into cellulose based fibres and films; opening its first rayon plant in 1921 and eventually having eight plants in operation. This experience in fibre production would help drive forward the commercialisation of Nylon in the late 1930s.
In 1924, Charles Stine became the head of research at DuPont, and organised a fundamental change in its approach to research. He went beyond the existing model of market driven development and established what we will call ‘The DuPont Model’ for research: ‘Hire smart chemists, provide them with laboratory and technical support, and they will invent materials that lead to commercially successful technologies.’ He established this research group within the chemical department and in 1927 sought out chemists to staff the effort. For the organic chemistry area, after considerable effort, he convinced Wallace Carothers to leave Harvard and come to work at DuPont. They agreed that Carothers would focus on the emerging area of synthetic polymers. His work, along with the support of other researchers at DuPont, would change the world.
In 1930, based on initial work by Fr. Julius Nieuwland at Notre Dame, Carothers and Charles Bolton pioneered the development of addition polymers, leading to the commercialisation of Neoprene and Chloroprene.
In 1931, based on work by the high pressure technology group, DuPont developed methyl methacrylate-based polymers that became Lucite and the polyvinyl butyral polymers, Butacite, that replaced nitrocellulose in automobile safety glass. In 1932, Carothers invented condensation polymers with the synthesis of the first polyesters, but this work was set aside in early 1935 with the invention of nylon 6,6.
Carothers continued his work on polymers, with a particular focus on polyamides, filing over 50 patents during his 9+ years at DuPont. He died on 29 April 1937 in a Philadelphia hotel.
The 1930s were a very productive period for DuPont research. In addition to Carothers’s work, DuPont moved into several other significant areas. In the late 1920s, a chemist at General Motors (GM) had discovered that low molecular weight chlorofluoro chemicals could be used to replace ammonia or sulfur dioxide as a low toxicity refrigerant. GM asked DuPont to produce these products commercially and this led DuPont into fluorine based chemical technologies. In 1938, while carrying out synthesis in this area, Roy Plunkett discovered polytetrafluoroethylene, Teflon. In addition, in the 1930s, DuPont became the major producer of titanium dioxide pigments and commercialised an early agriculture insecticide, phenothiazine.
The coming of World War II again brought major changes at DuPont along with the whole US chemical industry. As part of ‘total war’, the industry focused on a set of key chemical projects: the Manhattan Project with DuPont building and running the Hannaford, WA site; the development and production of synthetic rubber, led by Goodyear and others; and the large scale commercial production of Fleming’s 1928 antibiotic discovery, penicillin.
DuPont’s polymer technologies from the 1930s become significant technical contributors to the war effort: Nylon replaced silk in parachutes; Teflon in the Manhattan project and artillery shells; Lucite in air plane canopies; and Butacite in various forms of military safety glass. There were other, lesser known chemical developments that also occurred during the war, including significant advances in detergent chemistry and in the UK, Whinfield and Dickerson, extending Carothers early work on polyesters, invented polyethylene terephthalate, PET.
With the end of the war, DuPont found itself is a very desirable position. The European chemical industry had been decimated by the war; however, DuPont had profited from its explosive and materials groups, and was in a strong position to build on its existing technology. This resulted in the 1950s and 1960s being an extremely productive period in the company’s history.
In 1949, DuPont combined its plastic technologies into a single group and challenged them to bring out new, value added plastics. Technical developments included vinyl resins (Elvax) and polyamide resins (Zytel).
Polyester (PET) technology was purchased from ICI, which was in no position to exploit the technology. In the early 1950s, PET technology moved into commercial markets as boPET film (Mylar) and PET fibres (Dacron) were both commercialised. Both products proved to be immediate commercial successes.
Acrylic fibre, researched during the 1940s, was commercialised in the early 1950s as Orlon. Elastane fibre, discovered in 1958, was commercialised in 1962 as Lycra/Spandex, and Tyvek was discovered in 1955, but not commercialised until 1967. Corian, developed in 1968, was introduced to the public in 1971, while Nomex, discovered in 1963, was initially produced in 1967 and became available commercially in 1972. Kevlar, discovered by Stephanie Kwolek in 1965, was commercialised in 1971.
During this period, DuPont was very profitable. In the late 1960s, the Experimental Station changed its name to become DuPont Central Rese arch and a whole range of DuPont products had become household names. But by the late 1960s, DuPont technical management noted that the pipeline of high technology, high value products was drying up. In response, DuPont doubled down on its research efforts.
Budgets expanded and researchers were added under the ‘New Ventures program’, but by the late 1970s, the lack of significant results, the increasing cost of research and the economic downturn led to significant changes in DuPont research. Projects were canceled, personal levels were reduced and program leadership was restructured to provide ‘greater executive control’. Perhaps, this was the beginning of the end for Stine’s DuPont Model. This was also the end of big DuPont technical-commercial successes, with all the recognisable trade names; Lucite, Nylon, Mylar, Dacron, Orlon, Lycra, Tyvek, Corian, Nomex and Kevlar, now in the past.
This also led to a new period in the vision of how DuPont would generate profits in the future. While the technologies of the 1950s and 1960s were, in many cases, still very profitable; there was the growing threat of competition and price/profitability erosion on the horizon. DuPont entered a stage where focus for the future was not on growth based on new technologies coming out of Central Research, but on acquisition and diversification.
In the late 1970s, in an effort to get into the growing electronic components market, DuPont purchased Berg Electronics, which was then sold in 1993. In 1981, cash rich DuPont became a white knight in an attempt to protect Conoco from a take-over by Seagram. The three corporations became entwined in complex ways, with the relationships being finally dissolved between the mid-1990s to the early 2000s. In 1981, in an attempt to enter into the medical equipment business, DuPont purchased New England Nuclear and then sold it off in parts in the 1990s. In 1986, DuPont purchased Tau laboratories, as an effort to get into chemicals for the production of electronics, only to spin the company off in 1999.
DuPont, over the years, had developed candidate materials that it believed had potential as pharmaceutical actives. It did not pursue these materials due to lack of expertise in the testing and commercialisation of drug actives. In 1969, Endo Laboratories was purchased in hopes that they would provide the expertise that DuPont lacked. The relationship proved to be disappointing and in 1991, Endo Labs was combined with parts on New England Nuclear and these became part of a 1991 DuPont-Merck joint venture.
In the mid-1990s, DuPont announced a move into seeds, food and ‘natural fibers’. In 1999, it completed the purchase of the Pioneer Seed Company, the leading producer of hybrid seed corn. In 2000, DuPont announced the development of a fermentation route to 1,3 propane diol as the basis for Sorona, a ‘natural’ polyester. In 2011, it purchased Danisco Products with its food ingredient and enzyme based technologies.
In 1996, DuPont announced a joint venture with Dow in the area of elastomers, and in 1998, DuPont purchased Herberts, Europe’s largest producer of automobile coatings, and merged it group with its own coatings group.
Then came the selling off of DuPont. By the late 1990s, it became clear that the DuPont-Merck joint venture had failed to live up to expectations and, in 2001, it was sold. In 2004, DuPont sold Invista Fibers (Nylon, PET, Lycra and the Stain Master line), and in 2013, sold its entire coatings business. In 2015, DuPont spun off Chemours, which included chemicals, titanium dioxide and the fluorine products business.
In 2015, at the opening of the new corporate business center for the New DuPont, the company stated that ‘DuPont’s goals continue to be high growth, high value, global science and innovation’. In 2016, it announced the merger with Dow and, a few days later, that Central Research would be closed.
Should we cry or say ‘It was its time?’
The closing of Central Research marks the ‘death’ of the DuPont Model and with it, large chemical company commitments to significant research programmes. Was this ‘death of chemical research’ a case of murder or did it just die of natural causes?
Those that suggest murder could point to the takeover/divesture boom of the late 1970s and 1980s. Chemical companies were bought, overheads and costs were controlled, divisions were split up and the assets were sold off. Research, an expendable overhead, was quickly eliminated to reduce fixed costs and to improve the company’s bottom line. The rise of the MBA culture in the management of chemical companies, and the resulting focus on short-term returns, is inherently incompatible with long-term chemical research programs. Finally, there was a failure by research managers to recognise that the old model (the DuPont Model) was dying and to seek-out new research models that that could have provided a higher potential for generating a satisfactory return on investment.
Those that suggest chemical research died of natural causes would point to changing markets. In the 1930s, 1940s and 1950s, there were huge unmet needs in the areas of fibres and materials. By the early 1970s, these opportunities were well on their way to being met.
From that point on, new technologies, rather than walking into markets having large unmet needs, had to displace the current technologies and that was a much more challenging task. Early core technologies were accepted, in spite of high costs and processing issues. Equipment suppliers rapidly responded to the new needs and large technical service groups, supported by the high margins that the products generated, work with the emerging customer base.
Once these early market demands were met, new technologies needed to displace the existing technologies based on some combination of lower price, more efficient processing or significant technical improvements. This made market entry much more difficult and as time passed, patents expired, early core technologies became commoditised and processing equipment became increasingly refined, market entry by new technologies became even more challenging. New materials, successfully penetrating large volume current markets, became increasingly rare and as such, traditional research programmes, focus on identifying major market opportunities, became less and less productive.
The old DuPont Model died in the late 1960s, and while some people noticed, there was not another model available to justify large corporate research groups, so research directors just continued in the paths laid out by their predecessors. The cost of chemical research continued to rise, making large research groups harder and harder to justify. Finally, the perception of chemistry changed and with that change came the rise in regulatory requirements; limiting innovation and product development in many chemical markets.
While the takeover boom, failures in technical management and changes in the business environment may have accelerated the death of research in the chemical industry, the primary driver appears to be fundamental changes in nature of market opportunities.
What do we do now?
If this is the state of chemical research, what options are available for the chemical industry to identify develop and commercialise new technologies?
Rebuilding or restructuring internal research capabilities: This approach has a very low probability of success. It requires too much investment, has too much uncertainty and requires locating the right people to oversee and carry out the work. The probability of an acceptable return on investment over an intermediate time frame is very low. Attempts are being made to overcome these issues by setting up chemical research laboratories in India or China, but in all likelihood, these facilities will soon be experiencing the same problems as chemical research in the West.
Create an outside innovation office: This is a popular current option. Spin off a person and charge them with seeking out technology that has been developed outside of the company. Large consumer product companies, like Proctor & Gamble and Kimberly Clark, have abandoned their historical ‘Everything invented here’ approach to research and have sought to leverage the work of their internal groups with outside technology. The problems start with the basic assumption: ‘That there are people outside the company that can solve technical issues that internal people have not been able so solve and that these people will work very cheaply’.
In some cases, there may be existing ‘crossover’ technologies from other markets that have utility, but in most cases, ideas or concepts from individual entrepreneurs or small companies just disappear into the internal corporate research structures. To accept, confirm and then pay for outside technologies would reflect negatively on the abilities of the highly paid internal technology groups.
Outside innovation networks: This takes the most negative aspects of ‘the outside innovation office’ and puts them on steroids. Innovation networks are promoted as a low cost way of generating ideas on technical issues that internal technical staff have not been able to address. They assume that well developed technologies exist outside of the company, but that your internal people are not talented enough to find them. They often come with almost impossible to meet requirements, and this allows the requestors to ‘milk responders’ for new approaches to addressing the technical issue without actually engaging them: ‘Tell me all about your idea for this application, but don’t tell me anything proprietary’. Real chemical innovators do not work with innovation networks.
Utilise university-based research: Generally, university-based technologies are demonstration projects with no interest in actually applying the technology to the real world or in maintaining intellectual property rights. The interests of the professors often change quickly, based on current vogues in research and funding. Given the vast scale of academic publications today, finding interesting technology is itself difficult and if one finds an interesting technology, it is usually still in the ‘research’ stage. In addition, if the university has sought a patent on the technology - for US universities, usually only a US patent - you then face the issues of dealing with the university intellectual property department, which brings a whole new set of problems to the table.
In addition, recent trends in academic chemical research are not positive. They focus on the efficient generation of publications, which has moved chemical research into unproductive areas such as data generation collection, the re-analysis of old data and modeling. While such approaches generate relatively ‘easy’ publications, they do not generally lead to anything productive in terms of the development of new, commercially interesting technology.
Work with small, focused research companies: There is a general trend in mature industries: internally, they focus on the step-improvement of existing processes and technologies; while looking outside, for companies that have developed interesting new technologies, and they create joint ventures or they buy them. Smaller, innovative companies have the ability to pursue interesting academic technologies, to efficiently carry out chemical research and early stage development projects, to pursue promising ‘serendipitous’ observations and to seek outside funding to support early stage development work. Establishing a working relationship with such companies can establish the basis for licensing identified, commercially viable technologies, with relatively low up-front costs. The critical issues are the identification of such companies and establishing a mutually advantageous working agreement.
Develop an ‘internal-external’ research group: In the absence of being able to find small, research focused companies that are working in areas that are compatible with your technology, there is the option of identifying promising researchers (internal or external) and creating such a small start-up company.
With early stage funding, such a small company can be challenged to utilise academic research, outside funding opportunities and internal developments to identify and carry out research and early stage development work on new product concepts and technologies.
In return, the supporting company would receive some type of option on research results, such as a right of first refusal on new technologies.
The challenge here is to identify and use a creative model for addressing issues of early stage financial support, technology ownership and technology transfer. But if these can be overcome, such an approach can provide a company with a cost effective route to the early development of new technology, without the long-term commitments and management issues associated with an internal research group.
Whatever path you take forward in research, remember:
- Pace yourself and don’t oversell the technology. Really big, market hits will be increasingly rare.
- In the early stages of development, market needs and entry level paths to market need to be identified.
- Choose your targets carefully, in that regulatory or other barriers to change can greatly reduce the probability of success in some markets.
- Find the right people!
- Structure the programme to encourage productivity and reward success!
- Select a model and commit to it for a period of time. Utilise an early escape clause only if things are really going nowhere.
Be smart and future chemical technologies will reward both you and your company.