Research, Science & environment

A clue to what lies beneath the bland surfaces of Uranus and Neptune

Layers of water and hydrocarbons that, like oil and water, don't mix can explain planets' unusual magnetic fields

a planet against a blue background is exploded in layered hemispheres resembling clouds, an ocean, an amber layer and a core
An exploded view of an ice giant planet such as Uranus or Neptune. A new theory proposes that below the dense atmosphere lies a water-rich layer (blue) that has separated from a deeper layer of hot, high-pressure carbon, nitrogen and hydrogen (amber). The pressure squeezes hydrogen out of methane and ammonia molecules, creating stratified hydrocarbon layers that cannot mix with the water layer, which prevents the convection that creates a dipolar magnetic field.

Courtesy of Quanta magazine

Diamond rain? Super-ionic water?

These are just two proposals that planetary scientists have come up with for what lies beneath the thick, bluish, hydrogen-and-helium atmospheres of Uranus and Neptune, our solar system’s unique, but superficially bland, ice giants.

A planetary scientist at the University of California, Berkeley, now proposes an alternative theory — that the interiors of both these planets are layered, and that the two layers, like oil and water, don’t mix. That configuration neatly explains the planets’ unusual magnetic fields and implies that earlier theories of the interiors are unlikely to be true.

In a paper appearing this week in the journal Proceedings of the National Academy of Sciences, Burkhard Militzer argues that a deep ocean of water lies just below the cloud layers and, below that, a highly compressed fluid of carbon, nitrogen and hydrogen. Computer simulations show that under the temperatures and pressures of the planets’ interiors, a combination of water (H2O), methane (CH3) and ammonia (NH3) would naturally separate into two layers, primarily because hydrogen would be squeezed out of the methane and ammonia that comprise much of the deep interior.

These immiscible layers would explain why neither Uranus nor Neptune has a magnetic field like Earth’s. That was one of the surprising discoveries about our solar system’s ice giants made by the Voyager 2 mission in the late 1980s.

“We now have, I would say, a good theory why Uranus and Neptune have really different fields, and it’s very different from Earth, Jupiter and Saturn,” said Militzer, a UC Berkeley professor of earth and planetary science. “We didn’t know this before. It’s like oil and water, except the oil goes below because hydrogen is lost.”

If other star systems have similar compositions to ours, Militzer said, ice giants around those stars could well have similar internal structures. Planets about the size of Uranus and Neptune — so-called sub-Neptune planets — are among the most common exoplanets discovered to date.

Convection leads to magnetic fields

As a planet cools from its surface downward, cold and denser material sinks, while blobs of hotter fluid rise like boiling water — a process called convection. If the interior is electrically conducting, a thick layer of convecting material will generate a dipole magnetic field similar to that of a bar magnet. Earth’s dipole field, created by its liquid outer iron core, produces a magnetic field that loops from the North Pole to the South Pole and is the reason compasses point toward the poles.

a blue ball encircled by a ring, with a wedge removed to show interior layers in blue, red and yellow
A diagram of the interior structure of Uranus showing four distinct layers: hydrogen (light blue); water (dark blue); hydrocarbons (red); and a rocky core (yellow). Uranus has a disordered magnetic field that originates from its water layer. The image also shows the extreme tilt of Uranus’s axis relative to its orbit and one of the faint rings of material that encircle the planet.

Burkhard Militzer, UC Berkeley, and NASA

But Voyager 2 discovered that neither of the two ice giants has such a dipole field, only disorganized magnetic fields. This implies that there’s no convective movement of material in a thick layer in the planets’ deep interiors.

To explain these observations, two separate research groups proposed more than 20 years ago that the planets must have layers that can’t mix, thus preventing large-scale convection and a global dipolar magnetic field. Convection in one of the layers could produce a disorganized magnetic field, however. But neither group could explain what these non-mixing layers were made of.

Ten years ago, Militzer tried repeatedly to solve the problem, using computer simulations of about 100 atoms with the proportions of carbon, oxygen, nitrogen and hydrogen reflecting the known composition of elements in the early solar system. At the pressures and temperatures predicted for the planets’ interiors — 3.4 million times Earth’s atmospheric pressure and 4,750 Kelvin (8,000°F), respectively — he could not find a way for layers to form.

Last year, however, with the help of machine learning, he was able to run a computer model simulating the behavior of 540 atoms and, to his surprise, found that layers naturally form as the atoms are heated and compressed.

“One day, I looked at the model, and the water had separated from the carbon and nitrogen. What I couldn’t do 10 years ago was now happening,” he said. “I thought, ‘Wow! Now I know why the layers form: One is water-rich and the other is carbon-rich, and in Uranus and Neptune, it’s the carbon-rich system that is below. The heavy part stays in the bottom, and the lighter part stays on top and it cannot do any convecting.’”

“I couldn’t discover this without having a large system of atoms, and the large system I couldn’t simulate 10 years ago,” he added.

The amount of hydrogen squeezed out increases with pressure and depth, forming a stably stratified carbon-nitrogen-hydrogen layer, almost like a plastic polymer, he said. While the upper, water-rich layer likely convects to produce the observed disorganized magnetic field, the deeper, stratified hydrocarbon-rich layer cannot.

When he modeled the gravity produced by a layered Uranus and Neptune, the gravity fields matched those measured by Voyager 2 nearly 40 years ago.

“If you ask my colleagues, ‘What do you think explains the fields of Uranus and Neptune?’ they may say, ‘Well, maybe it’s this diamond rain, but maybe it’s this water property which we call superionic,’” he said. “From my perspective, this is not plausible. But if we have this separation into two separate layers, that should explain it.”

2 wedge cross sections of planets and the layered interiors colored blue, orange and yellow
Models for the interior structures of the ice-giant planets Uranus and Neptune have two distinct, intermediate layers: an upper, water-rich convecting layer where disorganized magnetic fields are generated, and a lower, non-convecting hydrocarbon-rich layer. New computer simulations show that icy materials naturally separate at high pressure and temperature into these two layers.

Burkhard Militzer, UC Berkeley

Militzer predicts that below Uranus’ 3,000-mile-thick atmosphere lies a water-rich layer about 5,000 miles thick and below that a hydrocarbon-rich layer also about 5,000 miles thick. Its rocky core is about the size of the planet Mercury. Though Neptune is more massive than Uranus, it is smaller in diameter, with a thinner atmosphere, but similarly thick water-rich and hydrocarbon rich layers. Its rocky core is slightly larger than that of Uranus, approximately the size of Mars.

He hopes to work with colleagues who can test with laboratory experiments under extremely high temperatures and pressures whether layers form in fluids with the proportions of elements found in the protosolar system. A proposed NASA mission to Uranus could also provide confirmation, if the spacecraft has on board a Doppler imager to measure the planet’s vibrations. A layered planet would vibrate at different frequencies than a convecting planet, Militzere said. His next project is to use his computational model to calculate how the planetary vibrations would differ.

The research was supported by the National Science Foundation (PHY-2020249) as part of the Center for Matter at Atomic Pressures.

RELATED INFORMATION