Nails in the coffin of dark energy?

 kw: science, cosmology, dark energy, supernovae, supernovas, type ia supernova, metallicity

INTRODUCTION

The ΛCDM model of the Universe was proposed after two research groups (led by Adam G. Reiss and Saul Perlmutter) studied certain supernovae. "Λ" (Greek lambda) refers to the cosmological constant, first proposed by Einstein, that describes the expansion of spacetime. The research teams concluded that spacetime was not just expanding, but expanding at an increasing rate. This is called "cosmic acceleration." Their key observation was that distant Type Ia supernovae are fainter than expected. This soon led to the hypothesis that 75% of the energy content of the Universe is "dark energy", which is driving and accelerating the expansion.

When I first read about "dark energy" more than 25 years ago I thought, "How can they be sure that these supernovae are truly 'standard candles' over the full range of ages represented, more than ten billion years?" I soon considered, "Is the brightness of a Type Ia supernova affected by the metallicity of the exploding star?" and "Is it worth positing a huge increase in the energy of the Universe?" From that day until now I have considered dark energy to be the second-silliest hypothesis in cosmology (I may deal with the silliest one on another occasion).

On December 10, 2025, an article appeared that has me very excited: "99.9999999% Certainty: Astronomers Confirm a Discovery with Far-Reaching Consequences for the Universe’s Fate", written by Arezki Amiri. In the article, this figure demonstrates that I was on the right track. The caption reads, "Correlation between SN Ia Hubble residuals and host-galaxy population age using updated age measurements. Both the low-redshift R19 sample and the broader G11 sample show a consistent trend: older hosts produce brighter SNe Ia after standardization, confirming the universality of the age bias. Credit: Chung et al. 2025"

It reveals a correlation between the brightness of a Type Ia supernova and the age of its host galaxy. Galactic age is related to the average metallicity of the stars that make it up. Thus, more distant Type Ia supernovae can be expected to be fainter than closer ones, because more distant galaxies are seen when they were younger, and consequently had lower metallicity. This all requires a bit of explanation.

WHAT IS METALLICITY?

Eighty percent of the naturally-occurring chemical elements are metals. That means they conduct electricity. Astronomers, for convenience, call all elements other than hydrogen (H) and helium (He) "metals". The very early Universe consisted almost entirely of H and He, with a tiny bit of lithium (Li), element #3, the lightest metal. The first stars to form were not like any of the stars we see in our sky. They were composed of 3/4 hydrogen by weight, and 1/4 helium. The spectral emission lines of H and He are sparse and not strong. Thus, the primary way for such a star to shine is almost strictly thermal radiation from a "surface" that has low emissivity.

[Insert Fig2 and add a caption] By contrast, a star like the Sun, which contains 1.39% "metals", has many, many spectral lines emitted by these elements, even as the same elements in the outer photosphere absorb the same wavelengths. On balance, this increases the effective emissivity of the Sun's "surface" and allows it to radiate light more efficiently. The figure below shows the spectra of several stars. Note in particular the lower three spectra. These are metal-poor stars, and few elemental absorption lines are visible (The M4.5 star's spectrum shows mainly molecular absorption lines and bands). However, even such metal-poor stars, with less than 1/10th or 1/100th as much metals content as the Sun, are very metal-rich compared to the very first stars, which were metal-free.

Spectra of stars of different spectral types. The Sun is a G2 star, with a spectrum similar to the line labeled "G0".

One consequence of this is that a metal-poor star of the same size and temperature as the Sun isn't as bright. It produces less energy. Another consequence, for the first stars, is that they had to be very massive, more than 50-100 times as massive as the Sun, because it was difficult for smaller gas clouds to shed radiant heat and collapse into stars. Such primordial supergiant stars burned out fast and either exploded as supernovae of Type II or collapsed directly into black holes.

THE TWO MAIN TYPES OF SUPERNOVAE

1) Type I, little or no H in the spectrum

A star similar to the Sun cannot become a supernova. It fuses hydrogen into helium until about half of its hydrogen is gone. Then its core shrinks and heats up until helium begins to fuse to carbon. While doing so, it grows to be a red giant and gradually sheds the remaining hydrogen as "red giant stellar wind". When the helium runs out, the fusion engine shuts off and the star shrinks to a white dwarf composed mainly of carbon, a sphere about 1% of the star's original size, containing about half the original mass. For an isolated star like the Sun, that is that.

However, most stars have one or more co-orbital companion stars. For any pair of co-orbiting stars, at some point the heavier star becomes a red giant and then a white dwarf. If the orbit is close enough some of the material shed by the red giant will be added to the companion star, which will increase its mass and shorten its life. When it becomes a red giant in turn, its red giant stellar wind will add material to the white dwarf. The figure shows what this might look like.

White dwarfs are very dense, but are prevented from collapsing further by electron degeneracy pressure. This pressure is capable of resisting collapse for a white dwarf with less than 1.44 solar masses (1.44 Ms). That is almost three times as massive a the white dwarf that our Sun is expected to produce in about six billion more years. It takes a much larger star to produce a white dwarf with a mass greater than 1.4 Ms, one that began with about 8 Ms. Such a star can produce more elements before fusion ceases: C fuses to O (oxygen), O fuses to neon (Ne), and so on through Na (sodium) to Mg (magnesium). The white dwarf thus formed will be composed primarily of oxygen, with significant amounts of Ne and Mg. Such a stellar remnant is called an ONeMg white dwarf. Naturally it has more metals present than the original star did when it was formed, but less than a white dwarf formed from a higher-metallicity star.

Now consider a white dwarf with a mass a little greater than 1.4 Ms, with a companion star that is shedding mass, much of which spirals to the white dwarf, as the figure illustrates. When the white dwarf grows to 1.44 Ms, which is called the Chandrasekhar Limit, it will collapse as a powerful Type Ia supernova.

There are two other subtypes, Ib and Ic, that form by different mechanisms. While they are also no-H supernovae, there are differences in their spectra and light curve that distinguish them from Type Ia, so we don't need to consider them further.

2) Type II, strong H in the spectrum

Type II supernovae are important because they provide most of the metals in the Universe. They occur when a star greater than 10 Ms runs out of fusion fuel. It takes a star with 10 Ms to produce elements beyond Mg, from Si (silicon) to Fe (iron). Fe is the heaviest element that can be produced by fusion. These heavy stars experience direct core collapse to a neutron star, with most of the star rebounding from the core as a Type II supernova. During this blast, the extreme environment produces elements heavier than Fe also. (Stars that are much heavier can collapse directly to become a black hole.)

EVOLUTION OF UNIVERSAL METALLICITY

At the time the first stars formed, the Universe was metal-free. It took a few hundred million years for a few generations of supernovae to add newly-formed metals, such that the first galaxies were formed from very-low-metal stars and low metal stars. Even with very-low to low metallicity, smaller stars could form. Since that time, most stars have been Sun-size and smaller, though stars can still form with masses up to about 50 Ms.

Stars of these early generations smaller than about 0.75 Ms are still with us, having a "main sequence lifetime" exceeding 15 million years. I can't get into the topic of the main sequence here. We're going in a different direction.

Stars of the Sun's mass and heavier have progressively shorter lifetimes. Over time, the metallicity of the Universe has steadily increased. That means that the "young" galaxies discussed in the Daily Galaxy article (and the journal article it references) are more distant, were formed at earlier times in the Universe, and thus tend to have lower metallicity.

LOWER METALLICITY MEANS LOWER BRIGHTNESS

This leads directly to my conclusion. A Type Ia supernova erupts when a white dwarf, whatever its composition, exceeds the Chandrasekhar Limit of 1.44 Ms. This has made them attractive as "standard candles" for probing the distant Universe. However, they are not so "standard" as we have been led to believe.

Consider two white dwarfs that have the same mass, say 1.439 Ms, but different compositions. One is composed of C or C+O, with very low amounts of metallic elements. The other has a composition more like stars in the solar neighborhood, with 1% metals or more. As seen with stars, more metals lead to more brightness, for a star of a given mass. Similarly, when these two white dwarfs reach 1.44 Ms and explode, the one with more metals will be brighter than the other.

The final question to be answered: Is this effect sufficient to eliminate all of the faint-early-supernova trend that led to the hypothesis of dark energy in the first place? The headline to the article indicates that the answer is Yes. A resounding yes, with a probability of 99.9999999%. That's seven nines after the decimal. That corresponds to a 6.5-sigma result, where 5 sigma or larger is termed "near certainty".

The article notes that plans are in the works to use a much larger sample of 20,000 supernovae to test this result. I expect it to confirm it. The author also suggests that perhaps Λ is variable and decreasing. My conclusion is that dark energy does not exist at all. Gravity has free reign in the Universe, and is gradually slowing down the expansion that began with the Big Bang (or perhaps Inflation if that actually occurred).

That's my take. No Dark Energy. Not now, not ever.