Prof at Duke Chemistry
Durham, NC
Joined March 2013
- Tweets2.6K
- Following712
- Followers843
- Likes5.7K
In 1888 Fritz Blau, a chemist, heated a dry copper salt and distilled a new compound, a plain white solid that turned intense red on contact with a trace of iron. He named it dipyridyl; its current name is 2,2'-bipyridine. Blau needed ten more years to explain the red color. Three bipyridine molecules bind a single iron atom to form a tightly bound group, or complex. Chemists found this binding so useful that a century later they called bipyridine the most widely used metal-binding molecule in their field.
Clinical kits still use Blau's red reaction every day to measure iron in blood samples. When ruthenium replaces iron in the same three-molecule complex, it gives off light under an electric current, and hospital analyzers use that light to read HIV and hepatitis tests. Farmers also spray diquat, a weedkiller with a molecular structure based directly on bipyridine. The US imports about 40,000 pounds of bipyridine a year and reports no production of its own.
A long-acting schizophrenia injection sells for $1,040 per gram, 11 times the price of gold.
Haloperidol decanoate is the long-acting form of haloperidol. After injection, the body releases the medicine slowly, letting patients stay covered between visits instead of relying on daily pills. That steady release helps keep schizophrenia symptoms from returning.
BLD Pharmatech Ltd. in China lists haloperidol decanoate at $260 for 250 milligrams. Aaron Chemicals LLC in Hong Kong lists it at $106 for 100 milligrams.
This is one of 838 molecules selling today for more than gold per gram. Here's the list:
In 1966 Leland Clark, a biochemist, submerged a mouse in a clear liquid and watched it keep breathing. The liquid was perfluoro-2-butyltetrahydrofuran, a fluorocarbon that holds enough dissolved oxygen for an animal to breathe the liquid. Clark's mice drew it in and out of their lungs for hours before returning to air and living for weeks. Clark's report in Science led researchers to study liquid breathing and artificial blood for decades.
The fluid later came into regular use in electronics. 3M sold it as Fluorinert FC-75 to cool circuits that air cannot cool, since it carries heat without conducting electricity. After making it for decades, 3M shut down the production lines when it left fluorochemical manufacturing entirely at the end of 2025. The US now has no domestic production. For 2019, US companies reported to the EPA that they imported about 56,000 pounds of it.
A green dye used to color fabric sells for $140 per gram, 1.5 times the price of gold.
The dye gives textiles a durable green shade. Manufacturers care because fabric color has to survive washing, light, and repeat production without drifting from one batch to the next.
TargetMol Chemicals in the United States sells C.I. Vat Green 1 for $70 for 500 milligrams. MedChemExpress MCE in Sweden sells it for $145.20 for 1 gram.
This is one of 838 molecules selling today for more than gold per gram. Here's the list:
In 1806 two French chemists boiled down asparagus juice and isolated white crystals of an amino acid, the first anyone had extracted in pure form from a living thing. They named the newly discovered compound asparagine, after the vegetable.
Eighty years later Arnaldo Piutti, an Italian chemist, extracted 100 grams of asparagine's mirror-image form from 6,500 kilograms of vetch seeds and tasted it. Natural asparagine is tasteless, but the mirror-image form was intensely sweet. This difference gave the first evidence that taste receptors can distinguish a molecule from its mirror-image form.
Labs today buy the two forms as a synthetic 50/50 blend called DL-asparagine, mostly to feed bacteria in culture media. The US imports about 570,000 pounds a year and reports no production of its own.
A chemical used in antidotes and glaucoma medicine sells for $453 per gram, 5 times the price of gold.
Physostigmine blocks one cleanup enzyme, so nerve signals last longer. That can reverse certain drug poisonings in emergency care, and in the eye it helps fluid drain to lower pressure.
Labotest KG in Germany lists physostigmine at $453 for 1000 milligrams. GlpBio in Montclair, US, sells 25 milligrams for $250.
This is one of 838 molecules selling today for more than gold per gram. Here's the list:
DuPont researchers tested cheap cotton for celluloid, a plastic, at the Experimental Station in 1915. They were trying to replace expensive tissue paper with short-staple cotton as a raw material. DuPont was weighing whether to develop its own manufacturing operation or buy an existing producer.
Initially, DuPont had wanted to sell nitrocellulose, long cellulose chains bearing nitrate groups, to celluloid makers. Celluloid combined it with a plasticizer, an ingredient that made the material more flexible. Manufacturers used expensive camphor, a molecule with two linked carbon rings in the ketone family. But American producers would not buy DuPont's nitrocellulose, and the company lacked the manufacturing information to make celluloid itself. A cheaper plasticizer might give DuPont an advantage if it entered the business.
By 1913, researchers had made good transparent sheets using acetaldol, an aldehyde with an alcohol group, as the plasticizer. Those sheets helped DuPont gain access to German manufacturing knowledge that American producers had withheld. Walter Carpenter studied the German business and presented entry options in 1914. There was still a risk with the substitute: DuPont feared it might make the plastic brittle in winter and unusable.
By the fall of 1914, researchers had developed excellent celluloid with a plasticizer of 50 percent camphor and 50 percent acetaldol. DuPont's Executive Committee funded a semiworks, a trial manufacturing plant. But after a company reorganization replaced the committee's members, the new committee turned toward buying a producer before the trial plant had issued a single operating report. The cotton experiment was another attempt to lower manufacturing costs.
After nine months of research, that experiment had failed. In September 1915, DuPont bought the Arlington Company of Arlington, New Jersey, the largest U.S. celluloid manufacturer. DuPont later opened the Delta Laboratory there and built a small plant to manufacture synthetic camphor using a purchased patent. Rising camphor prices during World War I made that plant successful for a time.
DuPont paid $260 million in 2026 dollars for Arlington; its return on investment averaged only 4.3 percent from 1916 through 1920, far below the company's goal of 15 percent. DuPont's research contributed little to the celluloid business. The camphor plant became a commercial disaster when the war ended and the wartime prices it depended on disappeared.
When you snap a glow stick, you start a reaction that a chemist discovered by accident in a janitor's closet. In 1962 Edwin Chandross, a young Bell Labs chemist, added hydrogen peroxide to a flask of leftover oxalyl chloride and a trace of dye, and the liquid glowed in the dark closet he used as his lab's darkroom. His department's patent attorney looked at the glowing liquid and declined to file a patent.
Chemists at American Cyanamid read his paper and decided to develop the reaction. In Stamford, Connecticut, they spent years replacing his short-lived chemical with a stable oxalate ester, with funding from the US military. By 1971 Michael Rauhut had developed bis(2,4,5-trichloro-6-carbopentoxyphenyl) oxalate, the compound sold as Cyalume and still used in glow sticks today. The white powder lights a life vest at sea and lets a soldier mark a position where a flame would be a hazard. The US now imports about 22,000 pounds of it a year, with no reported US production.
A bladder infection antibiotic sells for $1,050 per gram, 11 times the price of gold.
Amdinocillin blocks a protein bacteria use to build their cell wall. The cells lose their shape and die instead of multiplying. That matters because bladder infections are common, and resistant bacteria keep narrowing the useful antibiotic options.
MedKoo Biosciences in the US sells it as a research chemical for $1,050 for 1 gram. Cymit Quimica in Barcelona, Spain sells it for laboratory use at €580 ($678.72) for 1 gram.
This is one of 838 molecules selling today for more than gold per gram. Here's the list:
I read “Science and Corporate strategy: DuPont R&D, 1902–1980” by Hounshell and Smith, it’s such a great documentation of the history of Dupont product development and commercialization. It has so many great lessons and interesting history that few people will read. So I thought I would try to make a series of videos about each product developed over this 80 year period as documented in the book. Kind of history channel style.
DuPont kept backing a machine to make black powder, an explosive mixture, at its Experimental Station during the financial panic of 1907. Executives believed that a safer, cheaper way to make it justified years of experiments.
The work began with Alfred du Pont, who in 1902 requested $110,000 in 2026 dollars for experiments in wet mixing. The black powder combined sulfur, a chemical element, charcoal, a carbon-rich solid, and sodium nitrate, a salt made of sodium and nitrate ions. Wet mixing meant combining those ingredients in a slurry, a mixture of solids and liquid, then drying it. Alfred wanted safer, faster production, but earlier inventors had struggled to keep sodium nitrate from crystallizing on the powder. DuPont expanded his idea into a continuous process, with connected operations intended to lower the cost of each unit of powder.
DuPont worked with machinery inventor J. N. Wingett, but development moved slowly from 1903 through 1907. By October of that last year, the Executive Committee had authorized $3.9 million for the project, excluding Wingett's salary. The panic brought instructions to finish the machinery for no more than another $110,000. Yet experiments continued: for the remainder of 1908 and throughout 1909, the committee funded the work at $74,000 per month, excluding the salaries of Wingett and Francis du Pont.
Material caked inside the machinery, and the powder itself had low density and performed poorly when fired. Under a microscope, researchers could still see crystals on the grain surfaces. Despite these problems, the process seemed developed enough for DuPont to begin patent work in mid-1909. Attorneys found almost all 200 of Wingett's claims worthless and unpatentable, so DuPont ended his contract.
The Station's full-time staff kept working through the first quarter of 1911. Further work nearly eliminated the mechanical faults, but the powder remained unsatisfactory. DuPont narrowed the effort to wet mixing followed by its established batch operations, which handled separate quantities of powder: pressing it, forming grains, smoothing their surfaces by glazing, and drying them.
In January 1912, DuPont ordered the continuous-process machinery scrapped, abandoning a venture that had consumed about $6.9 million in development spending, excluding salaries. The venture lost money because the technical defects left it without a commercial advantage; the limited wet-mix approach offered no savings either, and the Station ended that work too. DuPont had hired Wingett before developing the idea in house, Francis du Pont had managed him poorly, and the Station had taken too long to judge the process.
Fin Sparre inspected the Fabrikoid artificial-leather factory in Newburgh, New York, in 1910, after DuPont had already signed the purchase contract. He spent two days examining its processes and found inefficient, often dangerous operations. DuPont had committed itself before finding out how the factory worked.
Congressional pressure against the "Powder Trust" had pushed DuPont to find uses beyond powder for its nitrocellulose capacity. Nitrocellulose is cellulose, the polymer in plant fibers, chemically modified with nitrate ester groups. Fabrikoid consisted of a nitrocellulose coating on woven cloth. Irénée du Pont, head of DuPont's Development Department, saw an opening for a premium artificial leather that the company's chemical knowledge could help it make.
In September 1909, DuPont funded an experimental artificial-leather plant at its closing Oakland site. Within six months, chemists and engineers produced the premium grade Irénée wanted. Their key technique was solvent recovery, a process for collecting solvents for reuse. DuPont then assumed that the Fabrikoid Company's use of solvent recovery meant the artificial-leather maker's entire process was advanced. That assumption lay behind the purchase that brought Sparre to Newburgh.
After his inspection, DuPont spent two years repairing the factory and its market reputation. The material still stretched too much diagonally across its woven backing, and its coating wore away at folds and corners. Poor weather resistance threatened its use for automobile tops and upholstery. DuPont also tested shoe uppers, the material above the soles, on about 40 Wilmington mail carriers and sold 150 pairs locally. The shoes looked good and resisted water, but they lacked ventilation, softened, and cracked with repeated bending.
DuPont pursued four years of research to improve the material and try to match genuine leather. Researchers made definite improvements, while the company aimed Fabrikoid at automobile tops and upholstery. As the automobile industry grew in the 1910s, Fabrikoid's business grew with it.
DuPont put $43 million in 2026 dollars into the company purchase, and the business's return on investment rose from 0.5 percent in 1911 to 20 percent in 1919. It made money, with profits averaging about 15 percent over that period, because chemical knowledge and solvent recovery improved production, sustained research improved the product, and automobile growth expanded demand. But the material's remaining physical defects defeated the shoe application: DuPont never produced satisfactory artificial leather for boots and shoes.
Chemists usually handle catalysts carefully because the classic ones stop working on contact with air, so they work in sealed gloveboxes under argon. In 2000 Amir Hoveyda's lab at Boston College reported a ruthenium catalyst that chemists could store in a jar on a shelf and recover from a flask for reuse. His group combined its design with a catalyst from Robert Grubbs at Caltech to create the Hoveyda-Grubbs catalyst.
The catalyst enables olefin metathesis, a reaction that breaks carbon double bonds and swaps their halves to build new molecules. The method earned Grubbs a share of the 2005 Nobel Prize in Chemistry. Today the Hoveyda-Grubbs catalyst is among the most widely used in the field. For example, drug chemists use it to close the oversized molecular rings in drug candidates, and materials labs blend it into epoxy resins that repair their own cracks. US import records list it as a catalyst for making synthetic rubber.
Chemists invented the catalyst in Massachusetts and named it for two American professors, Hoveyda and Grubbs. The US imports about 11,000 pounds of it a year and reports no domestic production.