The Pantheon in Rome has stood for nearly 2,000 years, its enormous concrete dome still spanning an interior that continues to astonish architects. Ancient Roman harbours, aqueducts and monumental walls also survive long after many newer concrete structures have required extensive repairs or replacement.
Their longevity raises an intriguing question: Did Roman builders discover a way to make concrete repair itself?
Scientists are now finding evidence that the answer may lie in a carefully chosen combination of volcanic materials, lime-rich fragments and chemical reactions that continue long after concrete has hardened. Research published in leading scientific journals has begun to explain how some Roman concrete could resist damage and seal small cracks.
But there is an important distinction between understanding an ancient material and proving that it can outperform modern concrete in every situation. Roman structures were built using different designs, served different purposes and experienced different environmental conditions. Their survival is not evidence that all ancient concrete was superior.
The real discovery is more useful: researchers are identifying specific mechanisms that could help engineers develop longer-lasting construction materials today.
The volcanic ingredient that changed construction
Roman concrete was not made using exactly the same recipe as modern Portland cement concrete. Depending on the location and intended use, builders combined lime, water, stone fragments and volcanic materials such as ash.
One particularly important ingredient was pozzolana, a reactive volcanic material associated with the region around Pozzuoli, near Naples. When suitable volcanic ash reacts with lime in the presence of water, it forms cementitious compounds that bind the mixture together.
This process, known as the pozzolanic reaction, helped Roman concrete harden into a durable material. Unlike ordinary lime mortar, which relies heavily on carbon dioxide from the air to harden, hydraulic lime-and-pozzolan mixtures can develop binding compounds even in damp conditions.
That mattered enormously for ports, seawalls, baths and other structures exposed to water.
Research published in the Proceedings of the National Academy of Sciences has examined the mineral structure of Roman architectural mortar and found that reactions involving volcanic materials helped strengthen the binding matrix. The formation of minerals such as strätlingite can reinforce the interfaces between aggregate particles and help obstruct the spread of cracks.
Read the original research: Mechanical resilience and cementitious processes in Imperial Roman architectural mortar — PNAS.
The volcanic ash was not a magical ingredient, however. Its chemical composition, particle size, proportion and interaction with lime all mattered. Different Roman structures used different mixtures, and a recipe that performed well in one setting would not automatically be ideal for another.
The white fragments that may hold the secret
For years, researchers examining Roman concrete noticed small, pale inclusions scattered throughout the material. These fragments, called lime clasts, were once sometimes treated as evidence of imperfect mixing.
A 2023 study led by researchers at the Massachusetts Institute of Technology offered a different interpretation.
Published in Science Advances, the research proposed that Roman builders used a technique called hot mixing. In this process, quicklime—calcium oxide—was mixed with other dry ingredients before water was added. The reaction between quicklime and water releases heat and produces calcium hydroxide, creating conditions that can leave distinctive lime-rich fragments in the hardened material.
The researchers argued that these clasts may act as reservoirs of reactive calcium within the concrete. When cracks form and water enters, calcium can dissolve from the lime-rich regions and move into the damaged area. Under suitable chemical conditions, it can contribute to the formation of calcium carbonate or additional cementitious compounds that fill pores and narrow cracks.
The important point is that the researchers did not rely solely on an attractive historical theory. They examined ancient material using chemical and microscopic techniques, then developed and tested modern cementitious mixtures containing lime clasts. Those experiments demonstrated that the approach can produce self-healing behaviour under laboratory conditions.
Read the peer-reviewed paper: Hot mixing: Mechanistic insights into the durability of ancient Roman concrete — Science Advances.
The findings do not mean that every Roman builder deliberately designed concrete to repair itself, or that every crack in an ancient wall will close. They do, however, provide experimental support for the idea that lime-rich inclusions can contribute to the long-term resilience of some Roman-style mixtures.
How can concrete repair itself?
Concrete may look like a solid, inert block, but it contains pores and a complex network of minerals. Water entering a small crack can trigger chemical processes that change the material around it.
In a lime-clast mixture, one possible sequence is relatively straightforward:
- A small crack develops and allows water to enter.
- Water interacts with reactive calcium-rich material inside the concrete.
- Dissolved calcium moves through the crack or nearby pores.
- Calcium carbonate or other mineral products precipitate under suitable conditions.
- These deposits partially fill the opening, potentially reducing water movement and slowing further deterioration.
This is known as autogenous self-healing: the material uses components already present in it to repair limited damage, without an external repair crew adding a new substance.
The process has limits. It depends on the crack’s width, water availability, temperature, the remaining reactive minerals and the concrete’s chemistry. Mineral deposits may seal a pathway without restoring the original strength of the damaged section. A large structural fracture cannot be assumed to heal safely simply because a smaller crack becomes less visible.
Self-healing also is not unique to Roman concrete. Modern researchers are investigating several approaches, including mineral-producing bacteria, encapsulated healing agents and engineered cement mixtures that encourage mineral precipitation. The Roman approach is one promising option among several.
What has science actually demonstrated?
The strongest evidence comes from combining observations of ancient samples with controlled experiments.
First, the materials have been identified. Researchers have used microscopy, chemical mapping and mineral analysis to examine lime clasts and the surrounding cementitious matrix in Roman mortar. These observations help reconstruct how the material was made and how its components interacted over time.
Second, hot mixing has gained archaeological support. In 2025, research on an unfinished construction site in Pompeii provided unusually direct evidence of Roman building practices. Analysis of walls and raw materials supported the conclusion that builders mixed quicklime with dry pozzolanic ingredients before adding water. The findings strengthened the case that hot mixing was a real technique, rather than merely a modern explanation for the composition of surviving concrete.
Read the study: An unfinished Pompeian construction site reveals ancient Roman building technology — Nature Communications.
Third, modern mixtures have shown self-healing potential. The 2023 Science Advances study tested cementitious formulations incorporating lime clasts and demonstrated that these inclusions can contribute to crack-filling reactions. That is meaningful experimental evidence, although it is not the same as proving that a modern bridge made with the technique will remain maintenance-free for centuries.
Fourth, scientists are investigating more than one durability mechanism. Volcanic materials can continue reacting, while calcium-bearing phases may carbonate and fill pores or fractures. The relative importance of each mechanism can vary with the mixture and its exposure conditions.
The evidence is therefore stronger than the claim that Roman concrete simply lasted because ancient builders possessed a lost secret. Scientists have identified plausible mechanisms, observed them in ancient materials and tested key parts of the explanation in modern formulations. What remains uncertain is how reliably these effects can be reproduced across large structures and decades of real-world exposure.
Could modern buildings use Roman concrete technology?
Potentially, yes. The goal is not to replace all modern concrete with an ancient recipe, but to adapt useful principles to contemporary engineering.
Engineers could incorporate carefully controlled lime-rich inclusions into selected cement mixtures, use suitable supplementary cementitious materials, or design binders that promote mineral formation when cracks develop. If these approaches reduce water penetration and delay corrosion or other forms of deterioration, they could extend the service life of certain structures.
That possibility matters because concrete production has a substantial environmental footprint. The International Energy Agency identifies cement production as a significant source of industrial carbon dioxide emissions. Making concrete last longer could reduce the need for repeated repairs and replacement, cutting material demand over a structure’s lifetime.
See the International Energy Agency’s overview of cement for context on the sector’s emissions and decarbonisation challenges.
Still, the path from laboratory sample to infrastructure is demanding. Modern buildings often rely on steel reinforcement, standardised performance requirements, rapid construction schedules and tightly controlled material specifications. A mixture that performs well in a small test specimen must also be assessed for strength, shrinkage, setting time, corrosion risk, compatibility with reinforcement, freeze-thaw resistance and long-term durability.
Researchers and construction companies must establish whether the benefits persist under realistic conditions, how much the technology costs, and whether the required ingredients can be sourced consistently. Independent testing, standards and long-term field trials are essential before widespread adoption.
Roman concrete is a lesson, not a perfect blueprint
It is tempting to frame the story as a contest between ancient wisdom and modern engineering. The reality is more interesting.
Roman builders developed materials suited to the resources, techniques and structures of their time. Modern engineers have created reinforced concrete systems capable of supporting skyscrapers, long-span bridges and complex infrastructure that the Romans could not have built in the same way. Many modern structures also face demanding loads, environmental exposure and design constraints that differ substantially from those of ancient monuments.
The survival of Roman buildings reflects not just chemistry, but also structural design, maintenance history, material selection and the conditions each structure has endured. Some ancient concrete has deteriorated, and some modern concrete is designed to last for many decades or longer.
The enduring achievement of Roman concrete is that it offers scientists a real-world record of material behaviour over exceptionally long periods. Its volcanic ingredients, lime clasts and mineral-forming reactions are now providing ideas that can be tested with modern methods.
The mystery is no longer simply why a Roman dome has survived. It is how engineers can translate the most promising parts of that ancient chemistry into concrete that uses fewer resources, resists damage more effectively and lasts longer.
The answer may not be a lost Roman formula. It may be a better understanding of what concrete can do when its chemistry is designed to keep working long after construction ends.
