OUR Solar System was built from the dust of dead stars, It鈥檚 an often repeated fact. But if you ask how this dust actually started to form planets, you might get an embarrassed silence. Planets, it seems, grow too fast-no one knows why the dust clumps together so quickly.
Steinn Sigurdsson of Pennsylvania State University has an explanation, but it may be hard to swallow. He thinks that the dust might have been under the influence of forces from hyperspace. 鈥淧lanets may owe their existence to a space dimension that has gone unnoticed until now,鈥� says Sigurdsson.
So what鈥檚 the problem with the conventional view of how planets formed? The idea is that you start out with an interstellar cloud of gas, sprinkled with dust grains made mostly of iron, silicon and ice. The cloud begins to shrink under its own gravity, perhaps jolted by the shock wave from a nearby supernova, and soon settles into a thick rotating disc called a protoplanetary nebula.
Advertisement
The tiny dust grains-each just a few micrometres across-drift around slowly within this disc. There is a little turbulence within the cloud, so occasionally grains collide and stick. When the grains get to about a centimetre across, they begin to move independently, sweeping up dust at a much faster rate. Only when they reach 10 metres across does gravity come into its own, pulling in material more rapidly. These rocks merge into planetesimals, many kilometres across, which collide to form planets.
The problem is that this process can be desperately slow. The bottleneck, according to Sigurdsson, is with the dust: to grow into pebbles those tiny dust particles have to hit each other head on and stick. In the inner parts of discs, where planets such as the Earth formed, dust is dense and grains collide frequently. However, in the outer disc, where planets like Uranus and Neptune formed, the material is rarefied and cold. Collisions are rare, and even when two grains do collide, they probably rebound most of the time. According to simulations, the typical timescale for forming a planet in the outer disc is 300 million years.
But proto-planetary discs don鈥檛 last that long. Astronomers see dense proto-planetary discs around stars that are 2 to 3 million years old, but only thin and depleted discs around stars that are 20 to 30 million years old. 鈥淎n obvious inference is that the process of planet formation therefore takes at most a few tens of millions of years,鈥� says Sigurdsson. If it takes any longer, the material will all be gone-blown away by the star鈥檚 radiation and the stellar wind.
So planets like Uranus and Neptune should not exist at all. Jupiter and Saturn would also have had difficulty forming in the time available. That鈥檚 especially important to us, as Jupiter is responsible for sweeping up most of the Solar System鈥檚 rogue comets and protecting the Earth from catastrophic impacts. 鈥淪omething must intervene to speed things up,鈥� says Sigurdsson.
One suggestion is that the grains become charged and so get pulled together by electrostatic forces. Ultraviolet light from a nearby hot star could charge up the grains by knocking out electrons from their surface. Some of these electrons would stick to neutral grains, making them negatively charged and therefore attracted to the positive grains. However, there is some doubt about whether ultraviolet could penetrate far into the thick, choking dust of a proto-planetary disc. And, even if it did, most of the ejected electrons would float about among the grains or attach themselves to floating hydrogen atoms. 鈥淭he problem is that a grain would be more likely to attract one of these than another charged grain,鈥� says Sigurdsson.
Another way dust aggregation might be speeded up is if the proto-planetary disc is ultra-thin, like a CD. This would boost the density and so the chance of grain collisions. 鈥淭he problem here, however, is that thin discs are unstable,鈥� says Sigurdsson. 鈥淭heir own gravity causes them to buckle, or vertical turbulence puffs them out.鈥�
A final possibility is that grains are extremely sticky and readily cling together when they collide. But the surface chemistry of these substances in space isn鈥檛 well understood.
Sigurdsson wasn鈥檛 convinced by any of these explanations, and began to look at alternatives. In particular, he considered the controversial idea that, on a submillimetre scale, gravity might become much stronger than Newton鈥檚 law predicts.
In 1998, Nima Arkani-Hamed and Savas Dimopoulos at Stanford University and Gia Dvali of the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy, suggested that there could be an extra dimension of space in which gravity alone acts and which until now has gone unnoticed. If this is so, then gravity-which is weak over large distances-gets stronger at the tiny distances encompassed by the extra dimension (see Diagram).
鈥淎ctually, there have to be two gravity-only dimensions,鈥� says Sigurdsson. 鈥淲ith only one, gravity is significantly modified on the scale of the Solar System-something which we do not observe.鈥� At scales smaller than these two extra dimensions, gravity would extend into five space dimensions rather than the usual three. 鈥淭he two extra dimensions would cause gravity to drop off with an inverse fourth law rather than an inverse square law,鈥� says Sigurdsson. 鈥淚f, for instance, there was an extra dimension 100 micrometres in extent, gravity would be 100 times stronger than predicted by Newton on a scale of 10 micrometres.鈥�
This short-range gravity boost is just what astronomers need to speed up dust aggregation. Grains flying close to each other would feel this force yank them together.
It doesn鈥檛 yank too hard, though. Like many others who simulate this process, Sigurdsson assumed that interstellar grains grow into fluffy, snowflake-like structures, whose large surface areas provide large targets and help to speed up the process of grain aggregation. The risk is that this delicate fractal structure would be squashed or broken apart if the speed of impacts were too great. So the turbulence within the disc can鈥檛 be too strong, and the acceleration caused by Sigurdsson鈥檚 modified gravity can鈥檛 be too extreme. Assuming only a small degree of turbulence, he calculated that the impact speeds would be less than a millimetre per second-not fast enough to disrupt the aggregates. And a by-product of assuming the low turbulence is that gravity has a much better chance of attracting grains to each other.
Sticky situation
The net result is to remove the bottleneck in planet formation. 鈥淚f there was a gravity-only dimension of about 80 micrometres in extent, I calculate that planets like Uranus and Neptune would form in a few tens of millions of years,鈥� says Sigurdsson. So the giants of our Solar System would have had a chance to grow before all their raw material was lost. And with them to take the flak, Earth would have been protected from bombardment by life-destroying rogue comets.
Sigurdsson confesses that his work was done as a bit of fun. 鈥淚 publicised it simply to see if anyone could easily shoot it down,鈥� says Sigurdsson. 鈥淏ut no one has-yet.鈥�
鈥淭he idea does not strike me as crazy,鈥� says Stevenson. 鈥淏ut it is unlikely to be correct.鈥� Stevenson doesn鈥檛 ascribe to the need for low levels of turbulence. 鈥淭he problem as I see it lies in the assumption that encounter velocities of dust particles would be so low that modified gravity matters at this length scale. It is far more likely that grain stickiness is what matters.鈥�
If Sigurdsson鈥檚 mechanism for speeding up planet formation were merely a theorists鈥� fantasy, it wouldn鈥檛 need to be taken too seriously. But it isn鈥檛. His proposal also explains a deeply puzzling result obtained by an experiment carried out on a space shuttle in 1999.
A team led by Jurgen Blum of Jena University in Germany released micrometre-sized silicate spheres in microgravity to see how the grains would aggregate. What the physicists found was that the grains formed just the kind of fractal grains everyone had hoped for. However, the experiment also threw up a puzzle. To everyone鈥檚 surprise, all the grains grew at the same rate. Small grains stuck to small grains and then, when they were all used up, big grains stuck to big grains, and so on (Physical Review Letters, vol 85, p 2426). There was never a mixture of small and large grains.
The explanation turned out to be that the growing aggregates were much fluffier than expected-more like wriggly fractal strings, or 鈥渟eedlings鈥�, as the researchers call them. These have a very large cross section, so they are extremely efficient at sweeping up any small fry.
The question then is, why should they take on these stringy, open shapes? The only way for them to be so tenuous is if all collisions between aggregates are glancing blows, so that any new material sticks on somewhere near the extremities of the growing fluff. But you would expect particles to collide at random angles, leading to aggregates that are more densely packed.
According to Sigurdsson, modified gravity makes glancing collisions much more likely. Grains which by rights should have flown past each other without sticking would be snared by the powerful non-Newtonian force. Because it gets stronger as you get closer in, grains would spiral in towards each other. 鈥淭here are no stable orbits in an inverse fourth law force field so, once caught, a grain spirals in to make its glancing collision,鈥� says Sigurdsson. 鈥淭he result is a kind of gravitational focusing.鈥�
The great attraction of Sigurdsson鈥檚 idea is that it is easily testable. In fact, it is likely to be proved right or wrong within a year. Experiments at three American universities are currently probing Newton鈥檚 inverse square law of gravity on submillimetre scales. 鈥淭he latest data I鈥檝e seen, from the group at the University of Washington in Seattle, is that Newton鈥檚 law holds down to 218 micrometres,鈥� says Sigurdsson. 鈥淚鈥檒l be holding my breath when they get down to 80.鈥�
