Skip to main content

What Is Stellar Nucleosynthesis?

Leonard Kelley holds a bachelor's in physics with a minor in mathematics. He loves the academic world and strives to constantly explore it.

Our star produces light in many different portions of the spectrum, with the biggest benefit to us coming in the visible portion. This light is a result of nuclear processes occurring inside the star, and this is occurring throughout the Universe. In fact, all the elements besides hydrogen and helium were for the most part a result of stellar processes involving nuclear reactions.

How these reactions occur and the way they make elements is a complex but fascinating topic. Through stellar nucleosynthesis, we have gained all the appreciable material needed to make our lives possible. Here is how its done.

Standard Nuclear Fusion

Stars make their energy by exploiting the strong nuclear force, which governs the bonding of atomic components. By merging and breaking these bonds in a selective manner, stars can release tons of energy that help sustain them. As you combine atoms, the distance between them in the new nuclei decreases, releasing the binding energy. Fission can also work this way, with nuclei decaying into a more tightly compacted nucleus. Iron is the most tightly bound element, meaning it’s the dead end for fusion or fission energy release methods (Seeds 126).

The fusion process starts off with making 4 protons into a helium (2 protons, 2 neutrons, and 4 electrons), with binding energy being released in the process. We know this energy is expelled because a helium nucleus is 0.7% lighter than the 4 hydrogen atoms, and since mass and energy are interchangeable we know the missing amount was converted to energy. Now, this reaction releases very little energy, and for a star like the sun it means that 1038 reactions a second is required, converting 5 million tons of mass to energy during that time frame. This seems completely unreasonable until you realize that only 0.07% of the star’s mass is lost as energy through this, leaving plenty behind (Ibid)

The process of converting the protons into helium is known as the proton-proton chain, and the actual step-by-step process goes like this. First, two protons collide and form a deuterium atom (which has a proton and a neutron in the nucleus) along with a positron and a neutrino. Then a deuterium atom and a proton collide to make 3He (called light helium) along with a gamma ray, a traditional signal of nuclear fusion scientists’ use, along with neutrinos, to gather clues about this process (Seeds 127, Arnett 1360).

Finally, a light helium collides with a light helium to make 4He (normal helium) along with two free protons that can go back into the process to start a new chain. Note that we needed two light helium, meaning that the first two steps happen twice on our way to getting to 4He. We need 6 protons to build up to a helium but because 2 get released back it’s a net process of 4 (Ibid).

It is worth noting here that there are other proton-proton chains besides the one mentioned above. The first two steps are the same, with the light helium being used in other ways. One alternate route, the second proton-proton chain, takes light helium and 4He and makes 7Be and a gamma ray, then the 7Be and an electron combine to make 7Li and a neutrino. The 7Li and a proton then combine to make two 4He. The third proton-proton chain involves light helium and 4He colliding to make 7Be and a gamma ray, then the 7Be and a proton collide to make 8B and a gamma ray. The 8B decays into 8Be with a position and a neutrino being emitted, and finally the 8Be decays into two 4He (Wallerstein 626).

A cross-section of an oxygen burning zone inside a 20 solar mass star. Color is used to indicate the amount of 20-Ne present (violet being 1% to red being 0%).

A cross-section of an oxygen burning zone inside a 20 solar mass star. Color is used to indicate the amount of 20-Ne present (violet being 1% to red being 0%).

The spare energy we noted before can now be accounted for – it’s those spare neutrinos, positrons, and gamma rays. The positrons combine with free electrons to make more gamma rays, the gamma rays can be absorbed by the surrounding gas with a small fraction actually escaping from the star, and the neutrinos, due to their weak interaction with normal matter, essentially pass through into space with no problems at all (Seeds 127, Wallerstein 626, Arnett 1360).

All of this in of itself drives heat up, further explaining the energy conversions we needed before. These energy signatures match what we see in laboratory results, and that along with spectral clues of stars showing us these elements gives us confidence that our models are right (Ibid)

For fusion to occur efficiently, we need our atoms to be in very close proximity to each other. This is challenging, because in a star atomic nuclei are protons and neutrons, therefore having a net positive charge. Like charges repel each other, so when we fuse nuclei they are actively fighting it off due to their Coulomb barrier, or the energy needed to overcome the electrical repulsion. We need the nuclei to be moving very fast then so they do not have a chance of repelling each other. This means the hotter the conditions, the faster the particles are moving, and the closer we go to the core the greater the pressure and density is. Therefore, traveling into a star reveals layers of nuclear fusion, with more complex chains occurring the deeper we go (Seeds 127).

One of these more complex chains is the carbon-nitrogen-oxygen cycle, otherwise known at the CNO cycle. It’s how more massive stars can make further helium from hydrogen but requires more energy because of the higher elements used in it. With carbon having 6 times the charge of a hydrogen, we need more energy to overcome the Coulomb barrier. The proton-proton chain we discussed earlier works best at 10 to 14 million Kelvin but for CNO we need temperatures greater than 16 million Kelvin. As we shall see, the CNO cycle is actually more efficient than the proton-proton chain (182).

First, we get a 12C to collide with a proton, making 13N and a gamma ray. Then, our 13N decays into 13C with a positron and a neutrino resulting as well. Note that we lose a proton here, it becoming a neutron instead with the positron and neutrino as by-products. This also happened in the proton-proton chain. In our CNO cycle, we will now have a successive series of steps with proton s and gamma rays involved. Our 13C and a proton collide to form 14N and a gamma ray, then the 14N and a proton make 15O and a gamma ray. The 15O decays into 15N with a positron and a neutrino resulting, so another proton lost here. Finally, the 15N and a proton become 4He and 12C, with that final carbon being used to help start the process all over again (Seeds 182, Wallerstein 626).

The amount of different types of carbon being released back into the interstellar medium, according to the stellar class.

The amount of different types of carbon being released back into the interstellar medium, according to the stellar class.

The CNO cycle does have an extra little surprise to it. About once every 2500 reactions, the 15N and the proton become 16O and a gamma ray instead. The 16O and a proton combine to make 17F, which decays into 17O along with a position and a neutrino. Finally, the 17O and a proton collide to make 4He and 14N, which can then go back into the CNO cycle and be used again. It’s a very efficient process, taking higher synthesized elements and making use of them over and over again (Wallerstein 626).

As we build up in our chain to heavier and heavier elements, we need hotter and hotter temperatures to overcome those Coulomb barriers. To get helium to fuse requires temperatures exceeding 100 million Kelvin. One of these processes is known as the triple-alpha process, because it needs 3 alpha particles, aka 4He, to make 12C. First, we get 4He and 4He to combine and make 8Be plus a gamma ray. Then, the 8Be and another 4He combine to make 12C and another gamma ray. Note here that 8Be is very unstable, and wants to decay back to two alpha particles. Therefore, we need to supply as much energy as possible so that it can be fused further before this decay occurs. And as a further note there are many more possible fusion routes for higher elements, the triple alpha process was just one example (Seeds 182, 200-1).

As a star uses up hydrogen, the helium present cannot fuse right away because the necessary 100 million Kelvin hasn’t been reached yet. So we develop a helium core of a sort, the material gathering on the inside of the star. After a certain portion of hydrogen has been converted, the outward pressure the reactions were providing diminished, allowing gravity to further condense the material and raise temperatures, eventually getting to a larger and larger helium core (Seeds 198-9, Wallerstein 626).

All of this is occurring at the core of our giant star, with our helium nuclei flying around at higher and higher speeds while the electrons remain trapped. The overall charge of 2+ that a helium nuclei processes means more energy is needed to overcome the Coulomb barrier and force the nuclei to fuse. Eventually, sufficient temperatures are met but how the helium fuses depends on the mass of our star (Seeds 200-1).

For stars more than 3 solar masses, they contract fast, heating up and allowing helium fusion to occur at a nice rate. For less massive stars though, things are rough. Because they condense slower with less gravitational forces, we gain more degenerate matter which then breaks our pressure temperature thermostat, no longer correlating temperature as strongly with pressure. Energy production cannot be regulated this way, but we eventually do get to the temperature needed to fuse helium via accumulating degenerate matter but at a pressure less than in a massive star. Therefore, we can get a helium flash since we lack sufficiency force to suppress the burning of helium, generating tremendous thermal effects in a runaway explosion (201).

Fascinatingly, this helium flash is barely detectable from the surface! The star retains its structure during this because the outer layers actually absorb the energy released during the helium flash. In just a few minutes to a few hours (which over the span of a star is barely perceptible), the helium flash is over, the degeneracy is removed, the pressure temperature thermostat is restored, and normal helium fusion occurs with a fusing hydrogen boundary surrounding it. The outer layers now contract back towards the core since the energy production is again regulated with gravity, increasing the temperature of the surface (since its area has decreased), and our star again moves towards the left side of the H-R diagram. This looping back and forth occurs each step we take in our fusion journey. Similar circumstances arise when we build up our element chain, with temperatures getting hotter and hotter. At 1 billion Kelvin, oxygen and carbon can be used to make many isotopes of elements from neon to sulfur, and so on until we can start to fuse iron at around 3-5 billion Kelvin (Seeds 198-9, 200-1; Wallerstein 626).

It is here that things change. A star cannot synthesize this way forever, and the end of the road is iron. We cannot get enough heat to overcome the Coulomb barrier, outward pressure from nuclear reactions goes down and gravity starts to condense things further. The shells that can still fuse feel this and increase their rate of production but are unable to effectively fight off the crushing conditions (Lally 47, Seeds 214).

Iron takes away further energy due to becoming degenerate, or where electrons are forced into the lowest possible orbitals available. Spare gamma rays get absorbed by spare matter too, further taking energy away. It gets to the point where the core cannot take it anymore and within 0.1 seconds it collapses. It’s hard to tell for sure what happens here because of the extreme conditions, but a shockwave ensues, tearing the star apart and leaving a neutron star of a black hole) (Ibid).


R-Process Production

The supernova is one example of a rapid neutron capture process (r-process) production of heavy elements. For years, it was thought that the stream of neutrons hitting elements as the shock wave progressed in this violent event lead to massive isotopes that decayed down into lighter but still heavy elements via beta decay. That is when a neutron breaks down into a proton and electron (so total charge is still zero). Because these events happen with massive stars, it could take as little as 10 million years to get some output of elements. And with 99% of the energy from the supernova being released via the neutrons, it should be energetic enough for element synthesis (Lally 47-8, Wallerstein 626, Truran 1293, Seeds 214).

This seems straightforward enough, but simulations along with spectral analysis shows that the expected amount of heavy elements coming from these isn’t enough to account for the heavy elements seen in the Universe. This led David Schramm and James Lattimer to develop another r-process route via kilonovae, where the orbiting neutron stars collide with each other. This would be a great place for r-processes to occur, because of all those neutrons at our disposal, suddenly being thrust into a violent scenario necessary for heavy isotope production (Lally 48).

Brian Metzger was able to figure out the expected radioactive signature such an event should generate, with an initial color of blue (from the lighter elements being made) then a shift to red (as the heavy elements are made). The trouble is being able to resolve our spectral data to see this, because most of these kilonovae happen very far away from us. You have to essentially luck your way into a kilonova observation (Ibid).

Fortunately, LIGO entered the picture and suddenly we gained an early warning system via gravitational waves. These come to us before the expected EM signal, so if you get a hit with LIGO then you can rush to find the location optically. And so far kilonovae seen do have the expected signature, confirming that these too are the site of r-process synthesis (48-9).

But even with this, we still are missing the necessary production seen in the Universe. The biggest clue that his is so comes from old halo stars around our galaxy. Lots of these are old stars and yet they contain lots of heavy elements known to be made via the r-process. This is strange, because these stars would have been born at a time when very few, if any, supernovas or kilonovas had happened. How then did these old stars get their contents? (Lally 49, Wallerstein 627)

A possible solution could be collapsars, developed by Stan Woosley. He postulates that in the early Universe, stars got very large compared to what we are used to, and at the end of their lives they were simply too large to go supernova. Instead, the collapsed under their immense gravity, releasing lots of radiation in the process. The pressure of this radiation flowing outward could have given rise to r-process elements that then seeded the next generation of stars that today exist in the halo of our galaxy (Lally 49).

It is also quite possible that there are still some unknowns yet to be resolved with the r-process, such as the neutron capturing mechanism, the beta decay itself, how elements undergo fission if unstable, and even the supernova conditions themselves (Truran 1295).

S-Process Production

The r-process isn’t the only potential route for heavy element synthesis, it turns out. There is also a slow neutron capture process (s-process) at play, which also makes use of beta decay. In this instance however the neutrons are captured passively, then decay down, so it’s a slow buildup of our elements.

How slow? Well, on the average we expect an element to capture a free neutron every month, which is a low occurrence. S-processes are likely to occur in the helium-burning regions of stars, so possible at the core or at a certain layer of the star, depending on its age. It is also more likely to happen in less massive stars, about 1-3 solar masses, and can take 100+ million years (Lally 47, Wallerstein 626, Truran 1293).

Signs of s-process have been spotted in the spectrums of red giant stars, such as R Andromeda. That star was one of the first signs of nucleosynthesis occurring inside stars when P.W. Merrill spotted technetium in the spectrum. This element is unstable, with the longest living isotope of it having a half-life of 2.6 million years. Seeing as how these stars have a lifespan nearly 1000 times that, it was a clear sign of active nucleosynthesis, since the element had to be created recently (Lally 47, Wallerstein 625).

But what is supplying the neutrons? Several possible reactions inside stars could give way to free neutrons conducive to the s-process. 13C and 4He can combine to make 16O and a neutron, 18O and 4He can collide to make 21Ne and a neutron and 21Ne plus 4He can make 24Mg and a neutron. But this requires plenty of those isotopes of carbon and neon, and that means they too have to be made at a reasonable rate (Wallerstein 626, Truran 1295-6).

Finding the Right Path

With two options for heavy element synthesis, it can be tricky trying to find out which process is occurring for a given situation. The biggest difference between the processes is in the time for the neutron capturing to occur, leading to the beta decay. But if we can map out the expected contributions of each process and compare with what is present in a given stellar environment, we can potentially get a trace history at hand. It also helps that certain elements are likelier to be made under certain processes, like barium being mainly an s-process element and europium being mainly an r-process (Truran 1294-5).

Finding ways to refine our simulations of the interior lives of stars is also yielding great insights. All those thermonuclear reactions inside a star happen so fast that the plasma containing it is not able to disperse the elements fast enough, so the layers remain rather fixed. It takes some thermal imbalance then to cause mixing between the layers, meaning we can have great regions with no convection at all, but certainly so at the boundary layers (Arnett 1360).

This is where the simulations come into play, and oftentimes we use approximations to simply, then build up in complexity from there. We assume that the regions of convection are about spherical, with all the random effects essentially creating fine, discrete layering. But supernova data shows that these approximations will fail at these extreme conditions (Ibid).

When we look at the spectral lines for these events, they don’t match our spherical approximations very well. In fact, it often shows many radioactive materials seemingly piercing their way through, with gamma rays being seen weeks before expected if in a truly discrete layering scenario. Modeling based on these results now points to some interesting behavior at the oxygen burning shell (1360-1).

In the hour leading up to the supernova, unburned fuel tends to be convected downward in unmixed packages. Much of this is made of 20Ne and 12C, which get burned quickly in the hotter, denser conditions and cause energy to be released and overcome that oxygen burning layer as hot spots form. These can fluctuate in duration, location, and intensity, leading to extended mixing and non-spherical results (Ibid).

Another insight has been gained when looking at binary systems, of which most stars are a part of in the Universe. Stars near the asymptotically giant branch of the HR diagram are amongst the largest stars in the Universe, are often in binary pairs, and contain lots of carbon. Binary pairs often undergo tidal forces as they grow, with the outer layers being syphoned off by gravity from one star to the other. Depending on the current moment in the lifespan of the stars, the layers taken away can reveal interior, more complex, elements while altering the chemical path of the other star. This too can alter nucleosynthetic models (Jeffery 345-6).

So next time you see a piece of jewelry or you operate some electronics, appreciate how that material was made. Not just by the hands of man, but in the complex interiors of stars.

Works Cited

Arnett, David and Grant Bazan. “Nucleosynthesis in Stars: Recent Developments.” Science , May 30, 1997, New Series, Vol. 276, No. 5317 (May 30, 1997), pp. 1360- 1361.

Jeffery, C. Simon and Christopher A. Tout, John C. Lattanzie. “Nucleosynthesis in Binary Stars.” Science Vol. 311, 20 Jan. 2006. Print. 345-6

Lally, Sapphire. “Where does gold come from?” New Scientist. New Scientist, 24 Jul. 2021. Print. 47-9.

Seeds, Michael A. Horizons. Tenth Edition, Thompson Brooks/Cole, Belmont, CA. 2008. Print. 126-7, 182, 198-9, 200-1, 214.

Truran, James and John J. Cowan, Catherine A. Pilachowski, Christopher Sneden. “Probing the Neutron-Capture Nucleosynthesis History of Galactic Matter.” Publications of the Astronomical Society of the Pacific, 114:1293–1296, 2002 December

Wallerstein, George. “Astronomical Evidence for Nucleosynthesis in Stars.” Science. 8 Nov. 1968, Volume 162, Number 3854. 625-7.

This content is accurate and true to the best of the author’s knowledge and is not meant to substitute for formal and individualized advice from a qualified professional.

© 2022 Leonard Kelley