Roman structures are still standing more than 1500 years after the last emperor was buried. And now the Romans’ secret of durable marine concrete has finally been cracked.

The Roman recipe – a mix of volcanic ash, lime (calcium oxide), seawater and lumps of volcanic rock – held together piers, breakwaters and harbours. Ad, believe it or not, the ancient water-based structures became stronger over time.

Scientists say this is the result of seawater reacting with the volcanic material in the cement and creating new minerals that reinforced the concrete. 

They spent a tremendous amount of work on developing this – they were very, very clever people,  said Marie Jackson, a geologist at the University of Utah. Scientists began their search with an ancient recipe for mortar, laid down by Roman engineer Marcus Vitruvius in 30 B.C.  It called for a concoction of volcanic ash, lime, and seawater, mixed together with volcanic rocks and spread into wooden molds that were then immersed in more sea water. History contains many references to the durability of Roman concrete, including this cryptic note by Pliny the Elder in 79 B.C who, in his Natural History, described concrete exposed to seawater as: “a single stone mass, impregnable to the waves and everyday stronger.”.

 What did it mean? To find out, the researchers studied drilled cores of a Roman harbor from Pozzuoli Bay near Naples, Italy. When they analyzed it, they found that the seawater had dissolved components of the volcanic ash, allowing new binding minerals to grow. Within a decade, a very rare hydrothermal mineral called aluminum tobermorite (Al-tobermorite) had formed in the concrete. Al-tobermorite, long known to give Roman concrete its strength, can be made in the lab, but it’s very difficult to incorporate it in concrete.

Now, they say, they’ve worked out why. Writing in the journal American Mineralogist, Ms. Jackson and her colleagues describe how they analyzed concrete cores from Roman piers, breakwaters and harbors.

Previous work had revealed lime particles within the cores contained the mineral aluminous tobermorite – a rare substance that is hard to make.  The mineral, said Ms. Jackson, formed early in the history of the concrete, as the lime, seawater and volcanic ash of the mortar reacted together in a way that generated heat.

But now Ms. Jackson and her team have made another discovery. “I went back to the concrete and found abundant tobermorite growing through the fabric of the concrete, often in association with phillipsite (another mineral),” she said.

She said this revealed another process that was also operating. Over time, seawater that seeped through the concrete dissolved the volcanic crystals, with aluminous tobermorite and phillipsite crystallising in their place.

These minerals, say the authors, helped to reinforce the concrete, preventing cracks from growing, with structures becoming stronger over time as the minerals grew.

By contrast, modern concrete, based on Portland cement, is not supposed to change after it hardens – meaning any reactions with the material cause damage.

Ms. Jackson said: “I think this opens up a completely new perspective for how concrete can be made – that what we consider corrosion processes can actually produce extremely beneficial mineral cement and lead to continued resilience, in fact, enhanced perhaps resilience over time.”

The findings offer clues for a concrete recipe that does not rely on the high temperatures and carbon dioxide production of modern cement, but also providing a blueprint for a durable construction material for use in marine environments. Ms. Jackson has previously argued Roman concrete should be used to build modern  seawalls.

So will you be seeing stronger piers and breakwaters anytime soon? Because both minerals take centuries to strengthen concrete, modern scientists are still working on recreating a modern version of Roman cement.

“There’s many applications but further work is needed to create those mixes. We’ve started but there is a lot of fine-tuning that needs to happen,” said Ms. Jackson. “The challenge is to develop methods that use common volcanic products – and that is actually what we are doing right now.”