Nanoscience in the Roman Empire

My last post covered the early surface science developed by Benjamin Franklin to explain the calming of rough waters by oil, but prominent examples of use of nanotechnology have been around for thousands of years. The Romans have long been regarded as expert engineers, but it turns out they might have been some of the earliest nanoscience pioneers as well.

The Lycurgus Cup is a cage cup (so named because of the way the glass is cut back to leave a decorative “cage” surrounding the body of the cup) crafted in the 4th Century AD by Roman Empire craftsmen. The cup is named after the mythical King Lycurgus of Thrace, cut in relief and shown wrapped in vines as a punishment for attempting to kill a follower of the Greek god Dionysus. In addition to being a marvel of pre-modern glass working, the Lycurgus Cup is also a striking example of an early application of nanoscience.

The Lycurgus Cup (a) reflecting and (b) transmitting light

Of Glass and Gold

The most eye-catching feature of the Lycurgus Cup, aside from the level of skill apparent in the craftsmanship, is the unique way that the cup interacts with light. When the cup is illuminated from the front, it appears an opaque pale green, but when lit from behind, it becomes translucent and red instead. Known now as dichroic glass, a cup made from this material would have aroused much admiration in the time of the Roman Empire. A cup that changes color depending on the way it is lit? What’s the trick? As is the case with most topics I’ll cover, the unexpected macroscale phenomena (in this case the apparent change in color depending on lighting) is based in the nanoscale structure and composition of the system. In fact, the behavior of this cup was so unusual that researchers initially argued over whether or not it was even made of glass in the first place. This debate was settled in 1959 when the cup was determined to be made of glass by x-ray diffraction.[1]

In 1962, independent elemental analyses carried out by General Electric Company and The Corning Museum of Glass both confirmed that the cup was made of soda-lime-silica glass, typical for Roman glass of the period. More interestingly, both studies found trace amounts of gold (40 parts per million) and silver (300 parts per million) were added into the glass during production.[2] Simply adding trace amounts of gold and silver to glass does not result in these unique color changing properties, so researchers were left scratching their heads at the cause of the dichroism. However, similar behavior had been noticed previously in colloidal systems (dispersions of small solid or liquid particles in a liquid or gas phase) due to light scattering, leading scientists to hypothesize the presence of metallic nanoparticles within the glass.[3,4]

The details of light scattering by small particles are beyond the scope of this short post; it is enough to know that particles scatter light depending on a number of factors including their size and composition, and the wavelength of the light. As a rule, particles that are significantly smaller than the wavelength of light tend to scatter shorter wavelength light (blues/violet) more than longer wavelength light (orange/red) (see “Rayleigh Scattering”). This is the same behavior that provides us with stunning red sunsets, due to light scattering from tiny gas particles. However, larger particles on the same size scale as the wavelength of light scatter all visible light more or less equally (see “Mie Scattering”). This is why clouds, which contain water droplets that are larger than atmospheric gas particles, look white or gray as opposed to red.

Effective Electrum

Despite these theories that metallic nanoparticles might be contributing to the Cup’s dichroism, it wasn’t until the late 1980’s that their presence was experimentally determined. Using transmission electron microscopy (the particles were too small to be seen using traditional microscopes), researchers identified metallic nanoparticles approximately 50–100 nm in diameter, confirming earlier hypotheses. Elemental analyses of these particles showed them to be an alloy of silver and gold (known as electrum), with a small amount of copper as well. Incidentally, the dichroic behavior is a result of both the nanoparticles being metallic and of their size. Nanoparticles both absorb and scatter light, and at 50 nm in size do both at just the right proportion that incident light results in a mostly opaque surface, while still transmitting some light. The transmitted light appears red because most of the blue light is scattered away, as explained previously. If the nanoparticles were much larger, no light would be transmitted. Much smaller, and the glass would be completely transparent.

Electrum nanoparticle seen under the transmission electron microscope.

As it’s unlikely that the Romans were able to create nanoparticles on demand, these nanoparticles were probably formed during a heat treatment step after the addition of gold and silver to the glass. Still, the level of expertise required to create an object such as the Lycurgus Cup is indisputable. It is a testament to the ingenuity of previous generations that it took 20th century scientists over 30 years to unravel the mystery of this early example of nanotechnology. It’s also an exhibition in the difficulty inherent in characterizing and truly understanding nanoscale systems. Dichroic glass is now routinely manufactured, and has even inspired record-breaking sensing devices! While nanoscience is often seen as the science of the future, the Lycurgus Cup is reminder that it’s actually been around for much longer than we tend to think.


  1. B. Harden & J.M.C. Toynbee, The Rothschild Lycurgus Cup, Archaeologia 97, 180 (1959).
  2. Freestone, N. Meeks, M. Sax & C. Higgit, A Roman Nanotechnology, Gold Bull 40, 270 (2007).
  3. H. Brill, The chemistry of the Lycurgus Cup, Proc 7th Internat. Cong. Glass, comptes rendus 2 223, 1 (1965).
  4. J. Barber & I.C. Freestone, An investigation of the origin of the colour of the Lycurgus Cup by analytical transmission electron microscopy, Archaeometry 32, 33 (1990).