A revolution has come in the first of the Big Bang’s 4 cornerstones

There’s a huge problem with our attempts to uncover the origins of the Universe: we only see our Universe as it is today, 13.8 billion years after the conditions marking the hot Big Bang first came into existence. Sure, we look at objects or events that occurred very far away, and by doing so, we can see the Universe as it was long ago, with the arriving light encoding the conditions that were present when that light was emitted. But because it’s so far away, even with our biggest, most powerful telescopes, we can only see the brightest objects and events that existed back then, and only at low resolution.

Due to these observational limitations, if we want to see faint features in great detail and to extraordinary precisions, we are confined to looking close by: to things that exist in the here-and-now, in the late-time Universe. For three of the Big Bang’s cornerstones,

• the expanding Universe,
• the growth and evolution of galaxies, galaxy clusters, and the cosmic web,
• and the leftover glow of radiation, the cosmic microwave background (CMB),

seeing the Universe out to great distances at the highest precision possible is how we thought up, confirmed, and validated the Big Bang picture up to the present day.

But the fourth cornerstone, Big Bang nucleosynthesis, has long been the lowest-precision and most poorly-measured (and most often disputed) line of evidence that the Big Bang provides. But here in the latter half of 2026, that is beginning to change, as the LBT Yp Project has just released a series of papers with the most accurate, precise measurements of the primordial helium abundance ever. Perhaps surprisingly, the evidence comes not from extremely early times, but from the nearby Universe. Here’s what these newest results are all about.

Historically, what we now know as the Big Bang was anything but a foregone conclusion. When Einstein first put forth general relativity, the prevailing view of the Universe wasn’t just that it was eternal, static, and stable, but that the Milky Way itself represented the full extent of the cosmos. With better observations and the theoretical development of the consequences of general relativity, however, things began to change. In the 1920s, observations of the spiral nebulae proved that they were indeed extragalactic objects: galaxies — or, as they were called then, “island universes” — unto themselves.

With measurements of galactic distances and redshifts together, we saw that the farther away a galaxy was, the faster it appeared to recede from us. By folding in equations derived from Einstein’s general relativity, we uncovered the concept of the expanding Universe: the notion that space itself is expanding, and that the distances between two unbound objects increases over time.

If you were to extrapolate today’s expansion back into the past, it implies that, at earlier times, the Universe was smaller, denser, hotter, and less evolved. The farther back you go, the smaller, denser, hotter, and more uniform (and less “clumpy”) it was. As first worked out by George Gamow and his collaborators in the 1940s, this would lead to three major consequences in addition to cosmic expansion:

• a cosmic web of galaxies, galaxy clusters, and even grander structures that grows and evolves over time,
• a bath of now-cold radiation, left over and subsequently redshifted from the transition between a hot, ionized plasma state (where electrons and photons bounce off each other) to a state of neutral atoms (where photons just stream, freely, through space),
• and an early period where it was so hot and dense that protons (and neutrons) could fuse together, creating composite atomic nuclei even before any stars had formed.

The “smoking gun” evidence for the Big Bang would arrive in the mid-1960s, when the bath of leftover radiation — originally known by the poetic name of “cosmic fireball” but now simply known as the CMB — was observationally discovered. But for the 20 years in between Gamow’s proposal and the CMB’s discovery, the origin of the elements heavier than hydrogen was the main battleground for the theory of our cosmic origins. Those elements clearly exist, but the big question was how they came to be? Gamow asserted that they were forged in the crucible of the Big Bang, while his main rival and proponent of the Steady-State alternative to the Big Bang, Fred Hoyle, asserted that they were created in stars.

The 1950s were the key decade for working out precisely how nuclear fusion occurred in stars, and Hoyle, along with Geoffrey and Margaret Burbidge, plus experimental physicist Willie Fowler, were among the most important figures in determining the nuclear processes that created those elements. Gamow’s scenario — of the hot Big Bang — would turn out to be important for the nucleosynthesis of the lightest elements and their isotopes alone:

• hydrogen (with a single proton),
• deuterium (with a proton and a neutron),
• tritium (with a proton and two neutrons, which decays into helium-3),
• helium-3 (with two protons and one neutron, which is stable),
• helium-4 (with two protons and two neutrons, which is the second most abundant isotope in the Universe behind hydrogen),
• lithium-7 (with three protons and four neutrons, which only occurs in about a billionth of the abundance of hydrogen),
• beryllium-7 (with four protons and three neutrons, which decays into lithium-7),

and negligible amounts of everything else. For making the majority of other elements, including carbon and oxygen, stars are required.

Over the remainder of the 20th century and the first quarter of the 21st century, huge advances occurred along the fronts of all four cornerstones. We measured the expanding Universe out to more than 10 billion light-years away, discovering how the expansion rate evolves and revealing the dominant component of the cosmic energy budget: dark energy. Today, the expansion rate is known to a precision of about 1% through the cosmic distance ladder.

Galaxies and galaxy clusters are seen to evolve in size, scale, mass, and stellar populations, and large-scale structure measurements reveal a Universe that grows clumpier and more clustered with cosmic time. Insights include a special feature known as the acoustic scale that leaves an imprint based on the amount of normal matter and dark matter found within the Universe, with recent evidence for an individual acoustic oscillation found in our own cosmic backyard.

And the crown jewel of modern cosmology is the CMB, where we’ve measured not just the overall temperature and spectral properties of that primordial radiation from just 380,000 years after the Big Bang, but have exquisite information about the imperfections — or departures from perfect uniformity — of those temperature fluctuations, as well as polarization measurements of the radiation. The CMB, more than any other data set, informs our modern picture of the Universe.

But the cornerstone of Big Bang nucleosynthesis, which once looked poised to be the decisive test of the Big Bang versus the Steady-State cosmological picture, has proven to be the least precise of the four main cornerstones of cosmology. Theoretically, the nuclear physics calculations for nucleosynthesis in the early Universe are quite complicated, but the physics has been known for many decades, and calculations are performed in a straightforward fashion. As the Universe, once filled with protons-and-neutrons in equal abundance, expands and cools, several things happen in sequence.

• It becomes easier for neutrons and neutrinos (or neutrons and positrons) to produce a proton and electron (or a proton and an antineutrino) than the other way around, as the kinetic energy of particles drops, the primeval balance tilts in favor of protons.
• Electrons and positrons annihilate away, removing all of the positrons, and the weak interactions freeze-out.
• Protons and neutrons fuse together to create deuterons, but the hot radiation of the Universe blasts them apart, preventing the build-up of a significant abundance of any complex nuclei.
• Free neutrons, with a half-life of a little over 10 minutes, begin to decay, further tilting the balance of the Universe in favor of protons and away from neutrons.
• And then, finally, the temperature of the Universe cools enough so that nuclear reactions can proceed unabated, creating the light elements and their isotopes according to the known equations of nuclear physics.

From 1 second after the Big Bang until about 20 minutes after the Big Bang, this nucleosynthetic period is of incredible cosmic importance.

Once those elements and their isotopes are created, the unstable ones decay, leaving a “relic abundance” of these initial conditions that remains unchanged until stars begin to form. Because the equations are so straightforward, it turns out that there are only two things that impact the results that we get for the abundance of these elements:

• the baryon-to-photon ratio, or how many CMB photons exist for every baryon (proton or neutron) that exists,
• and the total amount of radiation, which is dependent on the number of photons and the number of neutrinos.

If we could simply measure the initial abundance of any of these light elements and isotopes precisely and without ambiguity, whether helium-4, deuterium, helium-3, or lithium-7, we could gain an incredibly precise window into our cosmic history, independently of any of the other three cosmic cornerstones.

But measuring the initial abundance of the light elements is challenging, for one big reason: the Universe, over the course of its history, is very good at forming stars. In fact, we have never yet found a population of normal matter that was robustly demonstrated to have never formed stars in its past. One promising line of inquiry is to use quasar absorption features: where bright quasar light passes through very distant clouds of gas that have scant evidence for forming stars in their past, and to measure the strength of their deuterium lines versus the strength of their hydrogen lines. This yields a solid deuterium abundance, but drawing cosmological conclusions isn’t currently being improved by improving the observations. The reason is simple but annoying: there are theoretical uncertainties around the deuterium cross-section for the production of tritium and helium-3, and better observations won’t fix that.

Since helium-4 is the second most abundant element in the Universe, its abundance is much easier to measure, but the theory of Big Bang nucleosynthesis poses a problem here: the helium-4 abundance is rather insensitive to variations in the baryon-to-photon ratio. If the baryon-to-photon ratio were to range between 1-to-ten-billion versus 1-to-one-billion, the:

• lithium-7 abundance would vary by about a factor of 10,
• helium-3 abundance would vary by about a factor of 5,
• deuterium abundance would vary by about a factor of 40,

while the helium-4 abundance varies only by about 13%: from 22.3% of the mass of the baryons to 25.3% of the mass of the baryons.

In other words, if you want to tightly constrain the parameters of the Universe from Big Bang nucleosynthesis, and you use helium-4 to do it, you can’t just measure the helium-4 abundance to a few percent and learn something meaningful about the Universe. You have to go all the way to the third or even the fourth significant figure to place good constraints on the cosmological parameters that govern reality. But this is not an easy task; as of 2015, when the most robust study (for its time) was conducted, a precision of 1.6% was achieved: an uncertainty of around ±0.004 on the helium fraction, which is still too large to discriminate between a baryon-to-photon ratio of 1-to-four-billion versus 1-to-eight-billion.

However, just last year, in 2025, a new research programme to advance the field of BBN was proposed. The basic idea is to:

• use nearby low-mass galaxies that are easy to measure well,
• select the most metal-poor (low metallicity, or the least processed/most pristine) ones among those that qualify,
• make homogeneous observations of ionized nebular regions within them all, using the same high-powered instrument with a consistent calibration scheme,
• fold in observations of the third most-abundant element in the Universe (oxygen) to further eliminate the least pristine objects from your sample,
• measure the density and temperature of the electrons within the nebular regions inside these galaxies to better understand how the strength of the helium-to-hydrogen lines correspond to both the actual underlying abundance of the helium-to-hydrogen ratio as well as the ionization of those elements,
• and then screen the surviving ionization regions within those galaxies for potential errors, sources of bias, and low-quality and reliability, to “cut” your sample down to only the most reliable low-metallicity data points,

to give you the best measurements of the primordial helium abundance ever assembled, and then to extract the cosmological parameters associated from that abundance as a further product of those results.

The instrument the team used was the Large Binocular Telescope, which is basically two giant 8.4-meter diameter mirrors attached to a single mount (a prototype of the Giant Magellan Telescope), they measured a total of 62 galaxies over a four year time period, with a data set whose end result had 41 targets, including 15 high signal-to-noise targets with extremely low metallicities: an oxygen-to-hydrogen ratio of 0.004% or less. (For comparison, the Sun’s oxygen to hydrogen ratio is 0.05%, or about a factor of 12 larger than those limits.)

Then, from this data set, they plotted the inferred number density of helium atoms to hydrogen atoms as a function of metallicity, and extrapolated back to “what would this imply when we go all the way back to zero metallicity, or the conditions present before any stars had formed?” The answer they get corresponds to a primordial helium abundance where 8.146% of the number of atomic nuclei are helium nuclei, as opposed to 91.83% hydrogen nuclei. (Deuterium and helium-3 make up the rest.) When they convert this numerical ratio to a mass fraction, they get the most precise value ever obtained for the primordial helium abundance: 24.58%, with an uncertainty of just ±0.13%, which corresponds to an error of just half-a-percent of the overall value.

For comparison, the best “independent” determination of the primordial helium abundance comes from the CMB, and in particular from the ESA’s Planck satellite, which gave a value of 24.67% ± 0.02%. That value may be marginally improved by the Simons Observatory, but not beyond that, since the CMB S4 program was cancelled in 2025 by the USA’s National Science Foundation and Department of Energy. But as newer, more powerful optical telescopes come online — the ELT and the GMT in particular — the BBN measurements for the primordial helium abundance, as well as for the baryon-to-photon ratio and constraints on additional “exotic” species of neutrinos, should continue to improve in the decades to come.

With these new results, BBN from the observed helium abundance has now passed several milestones that, even a few years ago, seemed unlikely.

• The primordial helium abundance has been measured to a value with an uncertainty of just 0.5%: smaller uncertainty than on the expansion rate directly from the cosmic distance ladder.
• The baryon-to-photon ratio is consistent across all data sets: about 1-to-6.1 billion, with the helium abundance alone providing constraints that it’s between 1-to-4.9 billion and 1-to-6.4 billion.
• The determination of the baryon-to-photon ratio from the largely-insensitive helium abundance has now been made to greater precision than from the incredibly-sensitive deuterium abundance, despite a difference in sensitivity by a factor of 300.
• And, perhaps most importantly, we now have a validated picture of the hot Big Bang to extremely high precision across all four cosmic cornerstones to the 1% level or better: from when the Universe was just 1 second old through the CMB at 380,000 years old to modern observations of stars and galaxies and cosmic structure, from hundreds of millions of years of age until the present, 13.8 billion years after the Big Bang.

Big Bang nucleosynthesis, of the four cosmic cornerstones, has been the only one with its share of doubters even among mainstream cosmologists. No longer is there any justification for that to be the case! Thanks to the LBT Yp project, BBN is stronger than ever as a cornerstone of the Big Bang, and is well-positioned to improve even further in the years and decades to come!

This article is featured on Big Think.