A stellar source of lithium is caught in the act

Of all the elements of the periodic table, lithium has perhaps the most complicated and mysterious origins, in part because it can be created in so many ways. Its dominant isotope, ⁷Li, is one of the few species (along with deuterium, helium-3, and helium-4) to have been produced in Big Bang nucleosynthesis. Cosmic-ray spallation—nuclear fission initiated when energetic protons and other particles collide with interstellar carbon, nitrogen, and oxygen—is a significant source of Li, beryllium, and boron. And Li can be both produced and destroyed in stars, although questions remain about the nature of the stellar sources and how much Li they produce. In the effort to derive a clear picture of the chemical evolution of our galaxy by combining observations of elemental abundances with their known mechanisms of creation and destruction, the so-called “lithium problem” has been an especially tough one to crack.

Now Akito Tajitsu, of the National Astronomical Observatory of Japan, and his colleagues have found direct evidence of Li production in a stellar system—specifically, a classical nova: a thermonuclear explosion on a white dwarf that blows away the dwarf’s outer layers. ¹ The researchers used NAOJ’s Subaru Telescope, located on Mauna Kea in Hawaii, to monitor the absorption spectrum of the cast-off layers. They found a set of absorption lines that pointed to the presence of ⁷Be, a radioisotope and progenitor of ⁷Li. The short-lived ⁷Be must have been produced in the nova explosion.

“This is actually the first time that the formation of lithium has been caught ‘in the act,’” says Francesca D’Antona of the Rome Observatory. If the observation can be repeated for other novae, it will help to determine how much of today’s Li could have been produced in such explosions—a question that modelers of Li evolution would dearly like to see answered.

The lithium problem

The best data on Li abundances come from meteorites, which record the composition of our solar system just before the Sun started burning. Observations of stars of various ages provide information about other eras. Together, they suggest that Li abundance has been on the rise for most of our galaxy’s history; the source of the increase lacks a clear explanation. ² And of the solar system’s Li, at most 30% is accounted for by Big Bang nucleosynthesis and cosmic-ray spallation together. Core-collapse supernovae can explain at most another 20%, which leaves at least 50% for all other stellar sources. ³

The mechanism by which Li is synthesized in stars, shown in red in figure 1, seems almost contradictory. First, ³He and ⁴He combine to make ⁷Be, which decays with a half-life of about 50 days into ⁷Li. The He fusion step requires high temperatures—but if the nascent ⁷Be and ⁷Li remain at those temperatures, they’re quickly destroyed via the reactions shown in blue.

A stellar Li factory therefore needs a way to create ⁷Be and quickly cool it. Novae could easily qualify: The explosion itself is hot, and the cast-off layers are speedily ejected into interstellar space. A few classes of giant stars have also been identified as possible candidates. But models of each of them show that they collectively fall well short of producing enough Li to explain the solar system’s present abundance. ³

Blueshifted beryllium

Tajitsu and company have been using the Subaru Telescope and its High Dispersion Spectrograph (HDS) to study novae for several years. In mid-August 2013, when a nova appeared that was especially bright across a broad continuum of optical wavelengths, they hoped they could glean new information about the material expelled in the explosion. Light from the still-exploding star passes through the cast-off layers and thus captures their absorption spectra. Because the absorbing material is moving rapidly away from the star toward us, the spectrum is strongly blueshifted.

The researchers thought they might see absorption lines that they could attribute to Li itself, but they didn’t— perhaps because all the Li produced in the nova was ionized. Instead, they saw a group of curious near-UV features, labeled A, B, C, and D in figure 2. The features were identifiable for only a few days, between six and seven weeks after the explosion reached its peak: Before that, they were too saturated for their exact wavelengths to be determined, and after that, they had faded away to nothing.

Nearby known spectral lines, from hydrogen and calcium, showed that the spectrum actually had two distinct velocity components, one at 1103 km/s and one at 1268 km/s. Undoing that double blueshift revealed that the four unusual lines exactly matched a pair of resonances in Be⁺. Thanks to the HDS’s high resolution, the researchers could conclusively assign the absorption features to radioactive ⁷Be⁺ rather than stable ⁹Be⁺, even though their resonance wavelengths differ by less than 0.02 nm.

Comparing the Be⁺ lines to the analogous Ca⁺ lines allowed Tajitsu and colleagues to form a rough estimate of how much Be was expelled—and thus how much Li was created—in the explosion: perhaps 3 to 10 times as much as models predict. But that’s just for one nova, whose Li production might be unusually high or low. Many more observations will be needed to determine the average Li yield.

Though our galaxy hosts about 40 novae per year, only about 10 of them are observed, and just a fraction of those are bright enough for observers to have a chance of seeing the Be⁺ lines. But at least now the researchers know what to look for, and they know they have the right instruments. “The beryllium-7 lines are located in the near-UV range,” says Tajitsu. “To see them, we need a good location at high altitude, a large-aperture telescope, and a UV-sensitive spectrograph. The combination of the Subaru Telescope and the HDS is perfectly suitable.”