Japan’s Hayabusa probe returned to Earth on 13 June 2010, breaking apart in a bright streak over the Woomera Test Range in South Australia. About three hours before that final reentry, it had released a 40-centimetre capsule, which parachuted down into the outback intact. Inside were the first samples ever returned from the surface of an asteroid. The mission had taken seven years and very nearly did not make it back at all.
The spacecraft, known as MUSES-C until it was renamed Hayabusa just before its launch on 9 May 2003, was built by what is now the Japan Aerospace Exploration Agency primarily as an engineering demonstration. Its target was 25143 Itokawa, a near-Earth asteroid roughly 550 metres long, which Hayabusa reached in September 2005. The headline that survives is the clean one: first asteroid sample return. The actual story is messier, and more interesting for it.
What went wrong, in order
The list of failures is long enough that the return reads less like a plan executed than a plan salvaged. As Space.com summarised at the time, the probe suffered a fuel leak, repeated communications breakdowns, and malfunctions in its ion engines, problems that together added about three years to the journey. A small hopping lander called MINERVA, meant to bounce across the surface taking pictures, was released during a descent and drifted off into space without ever reaching the asteroid.
Then there was the sampling system itself, which is the part most worth dwelling on.
The sampler that did not fire
Hayabusa was designed to collect material by touching a horn-shaped device to the surface and firing a five-gram tantalum projectile into the asteroid at about 300 metres per second. The fragments thrown up by that impact would be funnelled through the horn into a sealed canister. Tantalum was chosen deliberately, because it is rare in meteorites and so would not be confused with the asteroid material.
The projectiles never fired. JAXA’s own account notes that during the first touchdown the spacecraft detected a possible hazard and issued an abort, and the abort command also cancelled the firing. The paper reporting the touchdowns in Science, led by JAXA’s team, recorded that the pyrotechnic control device gave no indication the projectiles had completed firing. Hayabusa touched down twice on a smooth patch named the Muses Sea on 19 and 25 November 2005, but the mechanism that was supposed to generate the sample did not work as intended either time.
Why there were samples anyway
What saved the science was the asteroid’s almost negligible gravity. When the horn made contact with the surface, the touchdowns themselves disturbed the regolith, and a small amount of dust drifted up into the collection area and stayed there. According to NASA, examination of the capsule later revealed roughly 1,500 dust particles from Itokawa, presumed to have been kicked into the catcher during the contacts rather than collected by the intended firing.
The grains were tiny, mostly tens of micrometres across, smaller than the width of a human hair. Confirming they were asteroid material rather than terrestrial contamination took JAXA several months of analysis. In November 2010 the agency reported that at least 1,500 grains had been identified as rocky particles, most of extraterrestrial origin and traceable to Itokawa.
What the dust actually settled
The payoff was not the quantity but a specific scientific question the grains could answer. Researchers had long debated whether the common S-type asteroids, the stony bodies that dominate the inner asteroid belt, were the parent objects of the most common meteorites found on Earth, the ordinary chondrites. The spectra did not quite match, and the suspected reason was space weathering, the gradual alteration of an airless surface by solar wind and micrometeorite bombardment.
The Itokawa grains, described across a set of papers in the 26 August 2011 issue of Science, closed much of that gap. The mineralogy matched a class called LL chondrites, tying a specific meteorite type directly to a specific asteroid for the first time using returned material rather than inference. That is the result the seven-year ordeal delivered: not a large sample, but a clean link between the rocks that fall to Earth and the bodies they come from.
What it set up
Hayabusa’s value turned out to be as much procedural as scientific. The mission proved that a sample could be retrieved from an asteroid and brought home, and it gave JAXA a hard education in everything that could go wrong doing it. The successor, Hayabusa2, launched in 2014 to the asteroid Ryugu with a redesigned sampling system, collected surface material in February 2019 and subsurface material that July, and delivered its capsule, this time with the collection mechanism working as designed, to the same Woomera range on 6 December 2020.
The richest comparison now comes from NASA’s OSIRIS-REx, which delivered material from asteroid Bennu on 24 September 2023. Bennu is a carbonaceous asteroid rather than a stony one like Itokawa, and early analysis of its sample has reported amino acids, nucleobases, and evidence of ancient salty water, the kind of organic and aqueous chemistry the dry Itokawa grains were never going to carry. That work is ongoing, and the Itokawa and Ryugu samples are increasingly read alongside it as the catalogue of directly sampled asteroids grows.
