The Concrete That Gets Stronger With Age — And Modern Engineers Still Can't Fully Copy It

The Paradox: A 1,900-Year-Old Dome vs. a 50-Year-Old Parking Garage

Rome's Pantheon was dedicated in 128 CE. Its dome is still, to this day, the largest unreinforced concrete dome ever built — no steel, no rebar, nothing but concrete holding up its own weight over a 43-metre span for nearly two thousand years. A short walk away, sections of Rome's ancient aqueducts still carry water. Along the Italian coast, Roman harbor piers that have spent two millennia getting pounded by waves are, by some measures, structurally sounder now than when they were poured.

Meanwhile, a typical modern reinforced concrete structure is engineered for a design life of roughly 50 to 100 years — and plenty fail well before that. Bridges, parking structures, and seawalls poured within the last century are already being torn down and rebuilt, cracked and spalling, while a Roman fish market from the reign of Trajan is still standing nearby. For most of the 20th century, engineers filed this under "ancient mystery, not our problem." It took two separate scientific investigations — one in 2017, one in 2023 — to prove that it very much is our problem, because the Romans weren't lucky. They were, in a very real sense, doing better materials science than we are, using tools that amounted to little more than fire, ash, and seawater.

The Detail Everyone Dismissed as Sloppy Work

Buried inside Roman concrete are small, white, gravel-like flecks called lime clasts. For decades, the standard explanation among archaeologists and materials scientists was almost insulting: the Romans, working with primitive tools, simply hadn't mixed their materials thoroughly enough. The clasts were treated as evidence of poor quality control — leftover, unreacted lime that never got properly blended in.

MIT civil engineering professor Admir Masic found that explanation hard to believe. Roman concrete recipes had been refined and passed down over centuries, following detailed formulas recorded by architects and builders of the era. It seemed strange that a civilization capable of engineering the Pantheon's dome would then get sloppy on the single most basic step — mixing the batch evenly. So in 2023, Masic's team, working with Harvard and labs in Italy and Switzerland, took a much closer look at what those clasts actually were, and how they got there.

The answer was hiding in how the concrete was mixed in the first place. Roman builders used what's called "hot mixing" — combining quicklime (rather than the pre-slaked, gentler lime modern preservationists assumed they used) directly with volcanic ash and water. That reaction produces intense heat, and it's this heat that traps small, unreacted fragments of lime inside the cooling concrete as those distinctive white clasts. Far from a mixing failure, it turned out to be functionally identical to a mechanism modern engineers spend millions of dollars trying to replicate in laboratory-designed "self-healing concrete": when a crack eventually forms and works its way to a clast, water dissolves the reactive lime, and the resulting solution recrystallizes as calcium carbonate, sealing the crack shut before it can spread. In lab tests, Masic's team ran water through cracked samples of both hot-mixed and conventionally mixed concrete — the hot-mixed version healed itself within two weeks, while the conventional sample kept leaking.

The Even Stranger Discovery: Concrete That Feeds on the Ocean

The lime clast finding was remarkable on its own. But it wasn't even the more surprising discovery — that distinction belongs to a separate line of research into Roman harbor concrete, led by University of Utah geologist Marie Jackson, using drill cores pulled from ancient piers by an international research project called ROMACONS between 2002 and 2009.

Modern marine concrete has a well-known enemy: seawater. Chloride ions in seawater seep through concrete, reach the embedded steel rebar, and corrode it. Corroding steel expands, and that expansion cracks the surrounding concrete from the inside out — a failure mode called spalling that can take down a modern seawall or pier within a few decades. Roman concrete has no steel rebar at all, which removes that failure mode entirely. But Jackson's team found something odder than mere absence of a weakness — they found active strengthening.

Jackson's team identified a mineral called aluminous tobermorite growing throughout the Roman marine mortar — a mineral so difficult to produce that, as Jackson put it, no laboratory had managed to grow it at room temperature. The Romans had. Her explanation: seawater slowly percolating through the concrete over centuries dissolved components of the volcanic ash, and the resulting highly alkaline fluid grew new interlocking crystals directly inside the material's pores and micro-cracks — a process closer to how a mineral forms inside a rock than how a manufactured material behaves. Rather than seawater corroding the structure, as it does to nearly every modern marine structure on Earth, it was building the Roman pier from the inside, one wave at a time.

Two Concretes, Two Philosophies

It's worth being precise about what this comparison does and doesn't mean. Roman concrete wasn't stronger in every sense — it typically had lower compressive strength than modern high-grade concrete, and it couldn't have supported the loads of a modern high-rise, which is exactly why steel-reinforced concrete was such a breakthrough when Portland cement was patented by Joseph Aspdin in 1824. Steel-reinforced concrete lets engineers build taller, span further, and pour faster, at a fraction of the cost per unit of strength. The trade Rome made — extraordinary longevity in exchange for lower load capacity and far slower construction — was the right trade for aqueducts and harbor walls. It is very much the wrong trade for a modern skyscraper. The real story isn't "Rome was better." It's that Rome solved a durability problem we still haven't solved, using a mechanism we only fully understood in the last few years.

Why This Isn't Just Ancient History

Cement production is responsible for roughly 8% of global CO₂ emissions — mostly from the intense heat required to produce Portland cement clinker. If the cement industry were a country, it would rank among the largest emitters on Earth. That fact, combined with the durability gap this research exposed, has turned Roman concrete from an archaeological curiosity into an active area of engineering research. Masic's lab and others are now working on modern self-healing concrete formulations that borrow the hot-mixed lime clast principle, aiming to extend the working life of structures and cut down on the emissions-heavy business of tearing down and repouring concrete every few decades. Separately, materials scientists are studying lower-carbon pozzolanic cement blends — using volcanic ash or industrial byproducts like fly ash in place of some clinker — that echo the Roman approach far more closely than standard modern concrete does.

None of this means your driveway needs to survive the fall of an empire. But it's a useful reminder that "concrete" isn't one uniform material — it's a mix design, and mix design is a real lever on both cost and lifespan, not just an afterthought a contractor picks off a shelf. Cheaper mixes cure faster and cost less today; better-specified mixes, proper curing time, and adequate reinforcement cover cost more upfront and buy decades of avoided repair. That trade-off exists at every scale, from a Roman harbor wall to a residential foundation footing.