Ask Ethan: How many generations of stars came before the Sun?

Here in our modern Universe, even in just our own Milky Way, we observe stars in all different stages of life:

• molecular gas clouds that are contracting and fragmenting,
• leading to protostars and young stellar objects,
• becoming full-fledged stars with protoplanetary disks around them,
• conventional stars burning through their fuel with their own fully-formed planetary systems,
• stars evolving into subgiants, giants, and even supergiants,
• stars dying in planetary nebulae, supernovae, and other life-ending events,
• and stellar remnants of now-extinct stars like white dwarfs, neutron stars, and black holes.

We can trace back the history of our Universe a full 13.8 billion years, to the earliest stages of the hot Big Bang, measuring the star-formation rate all throughout our cosmic history.

It was a full 9.2 billion years after the hot Big Bang first began, or roughly 4.6 billion years ago, that our Sun and Solar System began forming in our own Milky Way: some 27,000 light-years from the galactic center. And now, all this time later, human beings have risen to prominence here on planet Earth, having reconstructed our cosmic history more successfully than ever before. One of the questions we can ask, in the framework of how we got here, is how many generations of stars preceded the formation of our own? That’s what Mike Van Horn wants to know, asking:

“How many generations of stars had to precede the sun to account for the chemical composition of the sun and its planets? Star, supernova, star, supernova, and so on. How many cycles precede the sun?”

It’s a great question: one commonly asked by people curious about our Universe. But it’s not necessarily an easy question to answer, and a little investigation into the history of our Universe should help us understand why.

The visible light spectrum of the Sun, which helps us understand not only its temperature and ionization, but the abundances of the elements present. The long, thick lines are hydrogen and helium, but every other line is from a heavy element that must have been created in a previous-generation star, rather than the hot Big Bang.

What you see, above, is actually an image of our Sun. It’s not a “picture” of our Sun in the conventional sense: where we collect the light from its disk and no light from outside of its emissions. Instead, it’s an image of our Sun broken down spectroscopically: where the Sun’s light is all summed together, and then broken up into the individual wavelengths that span a portion of the electromagnetic spectrum. (In this case, it’s a visible light spectrum.)

The bright sections that you see represent parts of the spectrum where light is emitted the way sunlight normally gets emitted: as a series of blackbodies, all at different temperatures, from the outer portions of the Sun’s photosphere down to depths of tens of thousands of kilometers, where emitted sunlight can still escape.

The dark sections, however, represent elements that are present in the Sun in various concentrations: atoms that are neutral (or in a particular ionization state) and in the outer portions of the photosphere, where they can absorb that emitted sunlight and lead to dark bands. The darker and wider the band, the more abundant the atom. From this, as well as measurements of the other bodies in our Solar System, we can learn both:

• what fraction of our Sun is made up of heavy elements, as compared to the pristine hydrogen and helium that formed during the Big Bang,
• and what the ratios of those various heavy elements are to one another, going all the way up the periodic table.

The relative abundances of elements in the Solar System has been measured overall, with hydrogen and helium the most abundant elements, followed by oxygen, carbon, and numerous other elements. However, the compositions of the densest bodies, like the terrestrial planets, are skewed to be a vastly different subset of these elements. Overall, some ~90% of the atoms in the Universe, by number (but only ~70-72%, by mass), are still hydrogen, even after 13+ billion years of star-formation.

What we learn is that the most abundant elements, in order, are:

• hydrogen,
• helium,
• oxygen,
• carbon,
• neon,
• nitrogen,
• magnesium,
• silicon,
• iron,
• and sulfur,

and that hydrogen and helium make up more than 98% of the mass of the Sun, with all the other elements, combined, making up just a little over 1% of the Sun’s mass. In astronomy, we refer to the abundance of heavy elements (i.e., anything other than hydrogen or helium) as a star’s metallicity, and we use the Sun’s metallicity as our standard measuring stick: calling it solar metallicity.

When astronomers first began surveying stars and measuring their heavy element abundances — i.e., measuring their metallicities — they began dividing stars into two populations. There were stars that were of similar metallicity to the Sun, maybe a little more metal-rich or a little more metal-poor, and they called these “Population I” stars. But there were also significantly lower metallicity stars, particularly abundant in the globular clusters that were located far from the galactic center, which got the name “Population II” stars. Because the heavy elements found in our Universe are forged in stars, including in evolved stars and during stellar cataclysms (and cataclysms involving stellar remnants), a Population I star is generally assumed to have had more generations of stars preceding its formation than a Population II star.

The globular cluster Messier 69 is highly unusual for being both incredibly old, with indications that it formed at just 5% the Universe’s present age (around 13 billion years ago), but also having a very high metal content, at 22% the metallicity of our Sun. Its location may have something to do with its high metal content: it lies very close to the galactic center. The brighter stars are in the red giant phase, just now running out of their core fuel, while a few blue stars are the result of the mergers of initially lower-mass stars: blue stragglers.

But even Population II stars still have some quantity of heavy metals in them. And since we know — from the laws of physics and the conditions of the early Universe — that there was a period where nuclear reactions occurred shortly after the hot Big Bang (a period of Big Bang nucleosynthesis), we know that none of the stars we’ve ever spotted are made up purely of those pristine elements, alone. While hydrogen, helium, several of their isotopes, and tiny amounts of lithium (and beyond) were produced in those early stages, we’d have to find a star whose metallicity was less than one-ten millionth of solar metallicity for it to truly be pristine. Finding such a star would represent a third population of stars: Population III stars, made out of material forged in the Big Bang, alone.

As it stands, however, we’ve just set a new record (as of September 2025) for the most pristine star ever discovered in the Universe: a star discovered outside of our own galaxy, in the outskirts (specifically, the outer halo) of the Large Magellanic Cloud. This star is a red giant known as SDSS J0715-7334, and has just one-20,000^th the abundance of heavy elements that our Sun possesses. However, that’s around 1000 times richer in heavy elements than a true Population III star would possess. Even the most pristine star we’ve ever found isn’t truly pristine at all, but represents matter that’s been enhanced (or polluted, depending on your perspective) by enriched material arising from previous generations of stars.

This plot shows the iron abundance (x-axis) relative to the carbon abundance (y-axis) of a variety of low-metallicity stars, including the most metal-poor stars known. The newly discovered star from the Sloan Digital Sky Survey, J0715-7334, in the Large Magellanic Cloud, is the lowest-metallicity star known, with just 0.003% of the heavy element content of our Sun. However, it is still extremely far from being pristine.

With extremely low-metallicity stars such as this, we can actually make good comparisons to measurements we have of supernovae: where we can compare the relative abundance of the various heavy elements found within a star to the various ratios and abundances of elements produced by an individual supernova. This will show a dependence on the type of supernova (whether a core-collapse supernova, an exploding white dwarf, or a rarer, more exotic type) and the mass of the supernova’s progenitor (i.e., the mass of the white dwarf progenitor, the mass of the star before the core-collapse event, or the mass of whatever object produced the exotic supernova).

For SDSS J0715-7334, for example, 100% of its enrichment could be explained by the aftermath of just one prior supernova: a 30 solar mass core-collapse (type II) supernova. However, it’s worth noting three important caveats to that explanation.

1. The calculations about “enrichment from a single core-collapse supernova” are for a modern core-collapse supernova. We have never found a pristine, Population III star, and don’t know what types of elements their supernovae produce, and whether they’re similar to or different from modern ones.
2. The star SDSS J0715-7334 has fewer than a third of the overall heavy elements of the next-most-pristine known star, and about one-tenth of heavy elements found in the lowest-enrichment galaxies spotted by JWST.
3. And when it comes to the stars that we know of, the abundance of heavy elements found in stars isn’t uniform across a galaxy, but rather is more enriched near the galactic center and in the galactic plane, and is least enriched far from the galactic center and out of the galactic plane: in the outer galactic halo.

This color-coded map shows the heavy element abundances of more than 6 million stars within the Milky Way. Stars in red, orange, and yellow are all rich enough in heavy elements that they should have planets; green and cyan-coded stars should only rarely have planets, and stars coded blue or violet should have absolutely no planets at all around them. Note that the central plane of the galactic disk, extending all the way into the galactic core, has the potential for habitable, rocky planets, but stars facing away from the galactic center (far left and right) are much lower in heavy element abundance.

If we start asking, legitimately, “how many generations of stars needed to live and die before any specific star formed,” we find that’s actually an incredibly difficult question to address. The reason why it’s so challenging is straightforward: when we look at a star that we find in our modern Universe, today, it has been enriched by all of the contributions to the interstellar medium — and the interstellar gas cloud that led to this star’s formation — over the entire history that stars have existed and affected the abundance of this gas.

This brings up many questions we don’t yet know how to answer:

• what the history of gas and dust mixing in this particular galaxy is,
• how many star-forming regions, and how many times different star-forming episodes occurred within those regions, eventually affected this material,
• how many supernovae, planetary nebulae, kilonovae, tidal disruption events, or other stellar cataclysms contributed to the material that this star formed from,

and so on. When we look at the Universe, at any aspect of the Universe, all we typically get is a snapshot: of how things are at one particular instant in time. Everything else must be simulated and/or modeled and/or theoretically predicted, and that comes along with enormous uncertainties.

Put simply, there are many different ways to arrive at an “initial abundance” of heavy elements for a particular cloud of gas, involving small contributions from many generations of stars or large contributions from only a few generations of stars.

Artist’s illustration (left) of the interior of a massive star in the final stages, pre-supernova, of silicon-burning. (Silicon-burning is where iron, nickel, and cobalt form in the core.) A Chandra image (right) of the Cassiopeia A supernova remnant today shows elements like iron (blue), sulfur (green), and magnesium (red). Ejected stellar material can glow due to heat in the infrared for tens of thousands of years, and the ejecta from supernovae can be asymmetric and can have segregated elements within it, as shown here. In the right environment, this asymmetric material can be unevenly incorporated into future generations of stars.

Think about it. The first generation of stars — which we still haven’t ever found even one example of, mind you — are suspected to have been:

• extremely massive,
• with an average of around 10 solar masses (as compared to 0.4 solar masses for modern stars),
• and hence, very short-lived,
• and overwhelmingly likely to go supernova.

Those first stars, wherever you form them, are swiftly and surely going to enrich the entirety of the environment around them. They’re going to produce dust; they’re going to produce significant fractions of heavy elements; they’re going to help future generations of stars cool and contract more quickly, even when they’re just forming. Wherever you form these stars for the first time, you’re going to acquire some baseline amount of enrichment to your environment: somewhere likely around 0.001% to 0.01% of the present solar enrichment.

After that event, however, all of the stars formed from that now-enriched material will form with lower masses, on average, enabling them to have extremely long lives. If you form thousands of stars, then of those thousands of stars, perhaps only a few dozen of them will go supernova, with different supernova events separated by up to millions of years. The enrichment is not necessarily going to be symmetrical; near the center of the star-forming region, a single generation of massive stars could lead to heavy element enrichment that’s up to a few percent of solar metallicity. On the other hand, on the outskirts of those same regions, particularly where the gas doesn’t become well-mixed, an enrichment of just ~0.01% remains possible.

(This also implies, if you were curious, that we will require substantially deeper observations of the most metal-poor galaxies and distant gas clouds to find out whether they’re pristine, or just low-level enriched like SDSS J0715-7334 is.)

This is a blink comparison that plots the location of the red stars and blue stars that dominate the globular clusters in galaxies NGC 1277 and NGC 1278. It shows that NGC 1277 is dominated by ancient red globular clusters, but NGC 1278 contains many blue-colored ones. This is evidence that galaxy NGC 1277 stopped making new stars many billions of years ago, compared to NGC 1278, which has more young blue star clusters. The globular clusters located closer to the centers of these galaxies have preferentially higher metal contents, while the most distant ones have lower metal contents.

Then you can ask another question: based on the known exoplanets so far, which have just crossed the 6000 planet threshold, you can sort them into groups of whether the stars that have been monitored for planets either have known planets around them or don’t have any planets around them. If we confine ourselves to rocky planets — the kind that require sufficient quantities of heavy elements to form (i.e., planets like Earth) — we find something striking.

• Of all the planets ever found, 98.2% of them are around stars with at least 25% of the Sun’s metallicity: nearly all stars in this category have planets.
• Around stars with between 10-25% of the Sun’s metallicity, only 1.6% of known planets have been discovered; seldom do stars in this category have planets, and only very rarely do they have rocky planets.
• And of stars below 10% of the Sun’s metallicity, only 10 systems, ever, have been found to have an exoplanet. One system with about 1% has it, another one with about 2% has it, four with around 4% have one, and the others are all between 8-10%.

The second generation of stars, in other words, could possibly create a system that has planets around it, but we really need a much greater amount of enrichment to confidently have a good chance of creating Earth-like planets. It could possibly get there within just three generations, but it’s far more likely to require several generations, perhaps even rising into the double or triple digits of generations.

This diagram shows the discovery of the first 5000+ exoplanets we know of (a threshold crossed in 2022) and where they’re located in the sky. Circles show location and size of orbit, while their color indicates the detection method. No exoplanets have been found in globular clusters, and only one exoplanet has ever been found around stars with 1% or fewer of the Sun’s heavy element content. It’s as though a certain amount of enrichment is needed, prior to a star’s formation, for planets to be likely or even possible.

And why shouldn’t that be the case? After all, star-formation reached its peak in the Universe somewhere around 10-11 billion years ago, where stars formed at approximately 30-50 times the rate they’re forming in the Universe today. We know that in some star-forming regions in the Universe, star formation doesn’t just occur in one brief burst in a relatively confined region, like it does in the star-forming regions commonly found in the Milky Way. Instead, they can occur over an entire galaxy in a starburst event, or in a large star-forming region for hundreds of millions or even a billion years, continuously.

The old globular clusters found in the Milky Way, the ones that haven’t formed new stars in 10 billion years or more, have as little as 1% of the heavy elements found in the Sun, but the ones closest to the galactic center have up to 22% of the heavy elements found in the Sun. We can’t necessarily know “how many generations” of stars formed prior to the stars we find in those clusters, but we can be sure that the higher-metallicity ones have been enriched by greater numbers of prior generations of stars, overall, than the lower-metallicity ones.

What we can say is this. After billions of years of cosmic history,

• the closer you are to a galaxy’s center,
• the more massive the galaxy is that you’re in,
• and the closer to the center of a galactic plane you are (for a disk-containing galaxy),

the more enrichment you’re going to experience, signifying that more stars have lived, died, and enriched the gas and produced dust that will form future generations of stars.

The star-formation rate in the Universe is a function of redshift, which is itself a function of cosmic time. The overall rate, (left) is derived from both ultraviolet and infrared observations, and is remarkably consistent across time and space. Note that star formation, today, is only a few percent of what it was at its peak (between 3-5%), and that the majority of stars were formed in the first ~5 billion years of our cosmic history. Only about ~15% of all stars, at maximum, have formed over the past 4.6 billion years. Near “cosmic noon” when the greatest amounts of stars form, many galaxies have large and continuous, rather than sparse and bursty, regions of star-formation.

Overall, however, the lesson is this: it’s wrong to think about star-formation as the rare, isolated event that it appears to be today, in the outskirts of our late-time, thoroughly evolved Milky Way. Our galaxy, over the course of its history, likely experienced many events that led to sustained periods of star-formation: periods that lasted so long that stars formed, with many of them dying, then new stars forming, with many of them dying, over and over again, during the same star-forming episode.

How can you “count generations” in an event like that? The stars that form gather material from everything, cumulatively, that all of the different particles that form them experienced.

The thing we can be certain of is this:

• that many of the atoms that formed our Solar System had never been inside of even a single star previously,
• that other atoms were a part of one or more generations of stars,
• and that other atoms have been through several, possibly even dozens-to-hundreds, of star-forming episodes,

before joining together, some 27,000 light-years from our galactic center some 4.6 billion years ago, to form our Sun and Solar System. Sure, it might be more convenient to say “we’re at least a third-generation star,” but let’s try to dispel the myth that “generations” follow a similar math to stellar populations. Population III stars are the first generation of stars; Population II stars form as long as the overall enrichment remains low; Population I stars form once the enrichment is above a specific amount, whether it takes just a few or a great many generations to get there. Until we advance to the point where we can actually reconstruct the star-formation history of our own galaxy, rather than just infer a murky past from the evidence that remains today, this is likely the best we’ll be able to do.

Send in your Ask Ethan questions and suggestions to startswithabang at gmail dot com!