BRENNEN

Industries / Mining and materials / Chemicals and petrochemicals / Specialty chemicals / Adhesives and sealants

Chemistries

Chemistries are the families of molecules that do the sticking, from animal and plant glues to synthetic resins, and the reactions that turn a liquid or paste into a solid bond.

5

Milestones achieved, from 1938 to 2007.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1938

    Epoxy from a dental lab

    An epoxy is a resin whose molecules carry small, strained rings that a hardener opens and links into a solid network. Pierre Castan of Zurich claimed Swiss priority on 23 August 1938 for resins made from a phenol and epichlorohydrin and then heated with an acid anhydride until they became "infusible and insoluble"; his US patent, granted in 1943, notes that the products "adhere very well to glass." Novartis, the successor of Ciba, says the invention came from a dental laboratory that offered the patent to Ciba in 1943 because it had no use for it, and that Ciba sold it in Switzerland "shortly after the war" as a do-it-yourself glue; the brand's present owner, Huntsman, dates the Araldite launch to 1946.

    patents.google.comlive.novartis.comgo-araldite.com

    Credit

    Pierre Castan; Ciba

  2. 1954

    Glue that sets without air

    Anaerobic adhesives stay liquid while air reaches them and harden once they are shut off from it, as between the threads of a nut and bolt. General Electric developed one, Anaerobic Permafil, in the early 1950s, but it "tended to cure unexpectedly and prematurely"; Dave Dunn recalls that Bob Krieble, then a GE chemist, delivered samples with a portable air pump in his car trunk to keep them liquid. Bob's father, Vernon Krieble, a chemistry professor at Trinity College in Hartford, Connecticut, found that half-filled low-density polyethylene bottles let enough oxygen into the liquid to keep it stable on the shelf. He set up American Sealants Company in 1953, made the first commercial sales in 1954 and trademarked the name Loctite in 1956; the company became Loctite Corporation in 1963, and Henkel bought it in 1997. Dunn calls anaerobic liquids "the world's first 100%-reactive single-component adhesives."

    LessonA one-part glue is only useful if it stays liquid until it is used, so the bottle and the stabilizer are part of the chemistry.

    encyclopedia.trincoll.eduadhesivesmag.com

    Credit

    Vernon K. Krieble and Robert H. Krieble; General Electric

  3. 1968

    A weak glue's second life

    Spencer Silver, a 3M scientist, was trying to make a strong adhesive in 1968 and made instead one that held lightly and peeled off cleanly, so that it could be stuck down again. He spent five years looking for someone at 3M who could see a use for it, until in 1973 his colleague Art Fry wanted bookmarks that would not fall out of his church hymnal. Test sales in several large cities in 1977, under the name Press 'n Peel, had "mixed success" and 3M almost dropped the product; free samples handed out in Boise, Idaho, in 1978 led about 90% of those who tried it to say they would buy it, and Post-it Notes went on sale across the United States in 1980, by the Minnesota Historical Society's account.

    LessonAn adhesive that fails the job it was made for can suit another, and buyers may need to handle a new kind of product before they want it.

    mnhs.org

    Credit

    Spencer Silver and Art Fry, 3M

  4. 2006

    Soy glue replaces formaldehyde

    Most hardwood plywood was glued with urea-formaldehyde resin, which gives off formaldehyde, a probable human carcinogen. Kaichang Li of Oregon State University, "inspired by the superior properties of the protein that mussels use to adhere to rocks," modified some of the amino acids in soy protein to resemble that protein; his patent cures it with a formaldehyde-free resin, polyamidoamine-epichlorohydrin (PAE), and reports bonds that survived three cycles of 48 hours' soaking and 48 hours' drying without coming apart. Columbia Forest Products switched all its plywood plants to the soy adhesive, sold as PureBond, replacing more than 47 million pounds (about 21,000 tonnes) of urea-formaldehyde resin in 2006, with Hercules supplying the curing agent. The US Environmental Protection Agency gave the three a Presidential Green Chemistry Challenge Award in 2007 and reported that each plant's emissions of hazardous air pollutants fell by 50 to 90%.

    19january2025snapshot.epa.govpatents.google.com

    Credit

    Kaichang Li (Oregon State University); Columbia Forest Products; Hercules Incorporated, now Ashland

  5. 2007

    A primer copied from mussels

    Mussels hold to wet rock with proteins rich in a building block related to dopamine, a small molecule. Haeshin Lee, Shara Dellatore, William Miller and Phillip Messersmith of Northwestern University reported in Science in October 2007 that dipping an object in a water solution of dopamine leaves a thin, firmly attached film, polydopamine, on "noble metals, oxides, polymers, semiconductors, and ceramics," on which further layers can then be built. A coating that sticks to almost anything gave other laboratories a common first layer for bonding and coating; Crossref counted more than 10,000 later works citing the paper in November 2026.

    pmc.ncbi.nlm.nih.govdoi.org

    Credit

    Haeshin Lee, Shara Dellatore, William Miller and Phillip Messersmith, Northwestern University

Watch for

  1. Glues designed from protein data

    A team at Hokkaido University's Institute for Chemical Reaction Design and Discovery (WPI-ICReDD), led by Jian Ping Gong with Hongguang Liao as first author, mined about 25,000 adhesive protein sequences from the US National Center for Biotechnology Information's database, made 180 hydrogels (water-swollen polymer gels) whose building blocks follow patterns in those sequences, and used machine learning to design stronger ones. They report underwater adhesive strengths above 1 MPa (megapascal, about 10 kg on each square centimeter), which the university calls the highest reported; plates of ceramic, glass or titanium bonded in a salt-water tank that held a 1 kg load for more than a year; and a gel that sealed a leak from a 20 mm hole in a water pipe. A 2026 commentary by Chinese Academy of Sciences researchers notes that implant tests lasted about two weeks in mice, and Gong told Scientific American that the team still wants formulations "that can work on metal, for example."

    Achieved whena laboratory outside Hokkaido University remakes the published gels and measures underwater adhesion above 1 MPa, or the gels hold in a real use, such as a pipe repair or an underwater fixing, for months.

    global.hokudai.ac.jpscientificamerican.comijbs.com

    Claimed by

    Hokkaido University, in Nature, August 2025

Open

  1. 01

    Test the mined glues elsewhere

    An independent remake of the strongest hydrogel adhesives in the Hokkaido study, made from the paper's published recipes in another laboratory, with underwater lap-shear strength (the force to slide two bonded plates apart, per unit of bonded area) measured on glass, steel, concrete and plastic, and the same bonds retested after 30 and 90 days in natural seawater.

    The brief
    Credit · unclaimed
  2. 02

    Forecast sealant cure on site

    A free tool that predicts how long a moisture-curing sealant takes to skin over and to cure through a joint of given depth, from the local forecast of temperature and humidity. Silicone, polyurethane and MS polymer sealants all harden by reacting with water vapor from the air, so dry or cold weather slows them. The model is fitted to measurements of named products across a grid of climates.

    The brief
    Credit · unclaimed

Open 01 · the brief

Test the mined glues elsewhere

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An independent remake of the strongest hydrogel adhesives in the Hokkaido study, made from the paper's published recipes in another laboratory, with underwater lap-shear strength (the force to slide two bonded plates apart, per unit of bonded area) measured on glass, steel, concrete and plastic, and the same bonds retested after 30 and 90 days in natural seawater.

Why it is still open

Every strength figure so far comes from the authors' own laboratory. The one outside check found is computational: Ruchit Ambilduke, an undergraduate at the Indian Institute of Science, reran the authors' published code and data in August 2026 and matched eight of their nine model scores, but wrote that everything after a proposed recipe "requires synthesis and mechanical testing, so it cannot be reproduced computationally", and found two nearly identical recipes whose measured adhesion differed by 17.3 kPa. The 2026 commentary in the International Journal of Biological Sciences adds no experiments.

github.comijbs.comscientificamerican.com

Built with what exists

The paper's open-access methods and supplementary recipe tables (DOI 10.1038/s41586-025-09269-4) and the authors' code and data on GitHub (sheng-hu/hydrogels); the monomers and initiator it lists, bought from a chemical supplier; a universal testing machine for lap-shear tests in water; seawater tanks; an AI model such as Claude to pull recipes out of the supplementary files, each checked by hand; Python with scikit-learn to test whether the paper's model ranks the remade gels in the same order as the measurements.

  1. First step

    Download the paper's supplementary data, list the ingredients and amounts for the top three gels, and make one batch of the strongest to check that it sets as described.

  2. Done looks like

    At least three of the paper's top gels remade, with ten specimens per gel and surface, and the mean and spread of underwater strength set beside the paper's values, plus strength after 30 and 90 days in seawater, with all raw data posted.

Open 02 · the brief

Forecast sealant cure on site

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A free tool that predicts how long a moisture-curing sealant takes to skin over and to cure through a joint of given depth, from the local forecast of temperature and humidity. Silicone, polyurethane and MS polymer sealants all harden by reacting with water vapor from the air, so dry or cold weather slows them. The model is fitted to measurements of named products across a grid of climates.

Why it is still open

Makers publish cure figures for laboratory conditions. Incure, an adhesive maker, says moisture-cure silicones typically cure 2 to 3 mm deep in 24 hours at 25 °C and 50% relative humidity, and that published cure charts are made "under laboratory conditions that rarely match a specific production floor." GE's sealant guidance says cure time depends on sealant type, joint size, temperature, humidity and weather, and that dry air "can significantly slow curing." Searches of the web and GitHub in November 2026 found guidance of this kind but no public tool that forecasts cure from the weather.

incurelab.comgesealants.com

Built with what exists

Sealed boxes held at set humidities with saturated salt solutions, or a small climate chamber; cheap temperature and humidity loggers; a knife and calipers to cut beads open and measure the cured depth; Python with SciPy to fit cure depth against time, temperature and humidity; a free weather forecast service such as Open-Meteo; a phone web page.

  1. First step

    Lay beads of one silicone and one MS polymer sealant in two boxes held at different humidities, cut one bead from each box every day for a week, and plot cured depth against time.

  2. Done looks like

    For at least five sealants of three chemistries, measured cure depth at 24, 48 and 72 hours under at least nine combinations of temperature and humidity; at three outdoor sites not used for fitting, the tool predicts measured depths within 20%, with all data and code public.