The GALAH Survey: Searching for the Universe's missing lithium

The amount of lithium in the universe has changed since the Big Bang, can we measure the original amount?

Can studying the Milky Way help us confirm the amount of lithium (Li) created in the Big Bang? (Adapted from jpl.nasa.gov and the GALAH Survey website)

By Dirk Goës

We hear a lot about the element lithium these days, as it is one of the key elements in the clean energy revolution.  It was also one of the first elements created in the Big Bang.  Lithium’s ability to store electricity in a lightweight high-density form has made it ideal for building rechargeable batteries for everything from smartphones to electrical trucks. 

Mathematical models of the Big Bang have predicted how much lithium was created at the beginning of the Universe.  But are we able to confirm this amount by measuring the amount of lithium in our own Milky Way galaxy?   

Primordial Lithium

Immediately after the Universe was created in the Big Bang it consisted of a hot dense environment made up of atomic particles, such as protons, neutrons and electrons.  These particles began to fuse together into atoms, primarily hydrogen and helium with trace amounts of other elements including lithium.  This process is known as Big Bang Nucleosynthesis (BBN). 

At this time the Universe was expanding rapidly and within twenty minutes of the Big Bang it was no longer hot and dense enough for nuclear fusion to occur.  At his point the Universe was made up of the elements of hydrogen (75%) and helium (25%) with tiny amounts of other elements.  These trace amounts are so small that they hardly take anything away from the 75% hydrogen, 25% helium split.  The trace elements produced consisted primarily of deuterium, helium-3 and lithium-7.  Beryllium-7 was also produced but it decayed into lithium-7. 

The elements described above are all considered to be “light elements”, that is elements with a small number of protons.  Models of BBN calculate that lithium (hitherto lithium-7), the subject of this article, was produced at a ratio of approximately 1 lithium atom for every 1.8 billion hydrogen atoms. 

Table 1: The light elements created in the first 20 minutes of the Universe. 

Stellar Nucleosynthesis

As the Universe evolved the first stars and galaxies formed.  In the cores of these first stars, the intense pressure and high temperatures caused nuclear fusion to commence, forming heavier elements such as carbon (6 protons), nitrogen (7 protons) and oxygen (8 protons).  When these first stars exploded as supernova even heavier elements such as iron (26 protons) were produced.  This started the process of stellar recycling where the material distributed from one generation of stars became the raw material for forming the next generation of stars. 

As galaxies matured and became more complex many different types of stellar events contributed to fusing atoms together into heavier elements.  Events such as exploding massive stars (supernovae), exploding white dwarfs (supernovae type-IA), dying low-mass stars (like what will happen to our Sun) and merging neutron stars.  For example, merging neutron stars are thought to be one of the main sources of gold (79 protons) production. 

All the processes described above come under the banner of stellar nucleosynthesis.  Over the 13.8-billion-year history of the Universe these processes have changed the mix of elements to approximately 74% hydrogen, 24% helium and 2% heavier elements.  This includes a change in the amount of lithium in the Universe making it difficult for modern astronomers to confirm the primordial amount predicted by BBN calculations. 

For example, lithium can be destroyed in the interior of stars, but it is also created by red giant stars and novae.  Red giant stars produce beryllium in their cores which decays into lithium.  This lithium is then released into the interstellar medium (the space between stars) when the red giant experiences a dredge-up phase.  A nova is a thermonuclear explosion on the surface of a white dwarf star that occurs as the result of the build-up of hydrogen gas pulled off a companion star via gravity.  This explosion also produces beryllium which decays into lithium and is also released into the interstellar medium. 

Figure 1: The periodic table of elements showing by what cosmic mechanisms each element is created (Credit: GALAH Survey website). 

Searching for Lithium

In the last 13.8 billion years the Universe has evolved from containing only trace amounts of elements heavier than hydrogen and helium to containing about two percent heavier elements.  Astronomers call all the heavier elements ‘Metals’.  Therefore, hydrogen and helium are non-metals and everything else from lithium & beryllium to silicon & calcium to iron & gold are all considered metals.  While this may seem confusing at first, it is an easy way to delineate between the elements that where primarily created in the Big Bang and those that there created subsequently.  The ratio of metals in a star is called its metallicity. 

If you want to verify the amount of lithium that was created in the Big Bang, then you need to take into account that the mix of elements in the Universe has changed over time.  This means you need to observe and measure celestial objects where you think conditions would not have significantly altered the amount of lithium over time. 

This exercise was famously undertaken by the husband-and-wife team of François and Monique Spite at the Paris Observatory.  They studied the low-mass stars orbiting the outer reaches of our Milky Way Galaxy, known as halo stars.  These stars have been measured to be older than 11 billion years and to contain a very small amount of heavier elements or metals.  Therefore, they make good candidates for measuring primordial lithium.

François and Monique Spite found that these halo stars contain lithium at a ratio of approximately 1 lithium atom for every 6.3 billion hydrogen atoms.  This is a much lower ratio than the 1 lithium atom for every 1.8 billion hydrogen atoms predicted by models of the Big Bang.  This discrepancy came to be known as the ‘cosmological lithium problem’ and the measurement made by the François and Monique Spite is known as the ‘Spite Plateau’. 

The question then becomes, what happened to the lithium?  Low-mass stars such as our Sun have what is called a convective envelope.  This is the zone below the stars surface where hot gases rise and cool gas sinks back towards the interior in a constant cycle.  Below that is the radiative zone where energy is being pushed out from the nuclear core. 

Lithium can be destroyed at the relatively low temperature of approximately 2.5 million degrees K (Kelvin).  If the lithium present at the surface of a star is being pushed down into the hotter interior by convection it can easily be destroyed.  This is one possible explanation of what reduced the amount of lithium in the low-mass halo stars. 

Stellar models (the physics of how stars work) have shown that if you look at a range of low-mass stars the cooler stars tend to have strong convective envelopes (more likely to destroy lithium) whereas the warmer stars tend to have weaker convective envelopes (less likely to destroy lithium).  This indicates that studying these warmer stars is good place to measure the amount of lithium left over from the Big Bang.  Previous studies of Milky Way stars have shown that this may be correct but the number of stars in those studies is small, making it difficult to identify a definite trend. 

Figure 2: The convection zone in a cool and warm low mass stars. Not to scale. (Diagram: Dirk Goës). 

Enter the GALAH survey

The ongoing GALAH survey is one of the largest optical surveys of Milky Way Stars ever undertaken.  GALAH stands for Galactic Archaeology with HERMES.  HERMES is the spectrograph attached to the Anglo Australian Telescope (AAT).  GALAH has captured the spectra of one million Milky Way Stars and identified up to 30 elements per star including lithium. 

In a 2020 research study led by Xudong Gao of the Max-Planck-Institute for Astronomy and Karin Lind of Stockholm University (Gao et al. (2020)) used the dataset from the GALAH survey to examine a large set of low mass stars from cool to warm temperatures to see if they could identify a set of stars that contain a primordial amount of lithium. 

From 650,000 field stars (stars that are not part of a cluster) in the GALAH survey they selected low mass stars in the surface temperature range 5900 K to 7000 K (Kelvin).  For comparison the surface temperature of our Sun is 5800 K. 

The researchers also selected stars within a certain range of metallicity.  The number of heavier elements or metals in a star is often expressed on a logarithmic scale as the ratio between hydrgrogen (H) and iron (Fe) atoms as compared to our Sun.  A ratio of zero or [Fe/H] = 0 denotes that a star has the same ratio of metals as our Sun.  For this study they selected stars with a range of metallicities from [Fe/H] = -3.0 (low metallicity) to [Fe/H] = +0.5 (high metallicity).  This resulted in a set of over 100,000 stars. 

The results of the analysis revealed three distinct groups of stars based on their metallicity.  At the low metallicity end only cool old stars with a lithium ratio equal to the Spite Plateau were identified.  At the high metallicity end both cool and warm stars with lithium ratios higher than the Big Bang value were identified.  However, at the mid-range of metallicity between -1.0 ≤ [Fe/H] ≤ -0.5 a group of 117 warm stars were identified with a lithium ratio of approximately 1 lithium atom for every 2 billion hydrogen atoms which is consistent with the BBN value of 1 lithium atom for every 1.8 billion hydrogen atoms. 

Table 2: The results from Gao et al. (2020) expressed as the ratio of lithium to hydrogen atoms.  In the research paper the results are expressed as lithium abundance A(Li), please see ‘Note 1’ for an explanation.

In conclusion they found that the cool group of stars is more depleted in lithium than the warm group of stars by a factor of three.  And that the lithium abundance of the warm group of stars is very close to the BBN value.  Figure 3 from Gao et al. (2020) shows the three groups by metallicity and lithium abundance. 

Figure 3: Shows the lithium abundance A(Li) of the three groups against metallicity [Fe/H].  The lithium abundance of the BBN Group (SBBN Pitrou+) and the Warm Group (Warm plateau) are very close whereas the Cool Group (Cool Plateau) has a much lower lithium abundance.  The right of the chart shows that above a metallicity value of [Fe/H] ~ -0.5 lithium abundance increases rapidly as shown by the black squares and black stars (Credit: Gao et al. (2020)). 

Among other effects, in the cool group of stars lithium depletion has been driven by their deeper convective envelopes over the long lifetimes of these stars (> 11 billion years).  In the warm group of stars with their shallower convective envelopes, lithium has not been depleted, neither has significant lithium been produced.  At higher metallicities (> -0.5 [Fe/H]) where most of the stars in the study reside, lithium abundance increases significantly.  These stars represent the more recent population of stars in our galaxy where red giant stars and events such as novae are producing lithium. 

While earlier studies of the Small Magellanic Cloud (SMC) and the open star cluster (NGC 2243) have indicated similar results, the GALAH survey and its high-resolution spectroscopy have made it possible to determine these results on a much larger set of more than 100,000 Milky Way stars. 

Note 1

Thank you to Dr Sven Buder from the Australian National University (ANU) for answering my questions regarding this article.  Dr Buder is the principal investigator on the GALAH survey. 

Dr Buder provided the following explanation for expressing the results reported in Gao et al. (2020) as a ratio of lithium to hydrogen atoms. 

“The formulation of what A(X) is, is a bit strange and has a historical reason.

It is calculated as A(X) = log10(N_X / N_H) + 12.

If you turn that around, and simply want to figure out what the ratio of Li atoms to H atoms is, you would do the following — using BBN A(Li) as example:

A(Li) = 2.75 -> log10(N_Li / N_H) = 2.75 - 12 = -9.25 -> N_Li / N_H = 10^(-9.25), so less than 1/billion. You can also turn it around and state: how many hydrogen atoms are in a star compared to 1 Li atom -> 10^9.25, so more than 1 billion hydrogen atoms for each Li atom.

We have gone to these logarithmic descriptions, because H and He are so dominant in stellar atmospheres (only ~2% of elements in the Sun are not H or He). And the +12 was most likely added, because the inventors of this scale thought that no element would ever be reported if its number density is less than 10^(-12) compared to H.”