⚛️ phenomenology

The renormalization-group running of neutrino mixing is undetectable

This paper argues that the premise of directly observable renormalization-group running in neutrino mixing is incorrect because large logarithms in the self-energy cancel the scale dependence, meaning that neutrino mixing is a fixed property of the particle itself rather than a variable dependent on the momentum transfer of the production or detection process.

Original authors: Gabriela Barenboim, Mauricio Bustamante, Qinrui Liu

Published 2026-10-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Gabriela Barenboim, Mauricio Bustamante, Qinrui Liu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Neutrinos are the ghostly, nearly massless particles that stream through the universe by the trillions every second, passing through planets, stars, and human bodies without leaving a trace. They come in three distinct flavors: electron, muon, and tau. However, the neutrinos we detect are not fixed in these categories. As they travel across space, they oscillate, shifting from one flavor to another. This behavior is governed by a specific set of numbers called mixing parameters, which describe how the three flavors blend together. Scientists have spent decades measuring these numbers with increasing precision, hoping they hold the key to understanding why the universe is made of matter rather than antimatter and how particles acquired their mass. To make sense of these measurements, physicists often rely on a mathematical tool called renormalization-group running. This tool allows them to take a theory defined at extremely high energy scales, such as those present just after the Big Bang, and calculate how the mixing parameters would change as the energy scale drops to the levels we can test in laboratories today.

Recently, a growing number of researchers proposed a bold extension of this idea. They suggested that the mixing parameters do not just change over cosmic time or energy scales in a theory, but that they change in real-time depending on the energy of the specific experiment observing them. In this view, a neutrino produced in a low-energy pion decay would carry one set of mixing numbers, while the same neutrino detected in a high-energy collision would carry a different set. If true, this would mean that neutrinos could instantly change their flavor the moment they are created or detected, even before they have traveled any distance. It would also imply that experiments using different energy beams would measure different fundamental constants of nature. This idea gained traction in studies of high-energy astrophysical neutrinos and short-baseline experiments, promising new ways to discover physics beyond the standard model.

A new paper by Gabriela Barenboim, Mauricio Bustamante, and Qinrui Liu challenges this premise directly. The authors argue that the idea of energy-dependent mixing is fundamentally incorrect. They demonstrate that the mixing parameters are not properties of the experiment or the energy of the collision, but are intrinsic properties of the neutrino itself. To reach this conclusion, the team examined the quantum mechanical details of how neutrinos are produced and detected. They focused on the concept of the "mass shell," a state where a particle travels freely with a fixed relationship between its energy and momentum. Because neutrinos travel macroscopic distances between their source and a detector, they exist in this stable state. The authors show that the mixing matrix, which dictates how flavors blend, is fixed by the neutrino's self-energy—a quantum correction that depends only on the masses of the particles involved in the neutrino's internal loops, not on the energy of the collision that created or caught it.

The researchers explain that while the mathematical parameters in a theory do change with an arbitrary scale chosen by the physicist, the physical reality measured in an experiment does not. When a scientist chooses to calculate a process at a high energy scale, large mathematical terms appear in the equations that describe the neutrino's self-energy. These terms act as a counterbalance, rotating the changing parameters back to their original, fixed physical values. The premise that mixing changes with energy fails because it keeps the changing parameters but drops these necessary counter-balancing terms. Consequently, the mixing matrix remains constant regardless of whether the neutrino is produced in a low-energy pion decay or detected in a high-energy astrophysical event.

To prove this, the authors applied their reasoning to four specific theoretical frameworks where the energy-dependent mixing idea had been proposed: the Minimal Supersymmetric Standard Model, a seesaw model with a heavy neutral lepton, a seesaw model with a light scalar, and the Standard Model effective field theory. In every case, they found that the physical mixing matrix remained the same at production and detection. The only differences that could arise from the energy of the process were tiny corrections to the interaction points, which were far too small to be observed as a change in the mixing angles. The authors also pointed out that if the energy-dependent premise were true, short-baseline experiments would have already seen neutrinos changing flavor immediately upon leaving their source. Since these experiments have seen no such effect, the data already limits any potential change in mixing angles to less than 0.7 to 2.4 degrees between the energy of pion decay and a few GeV, ruling out the large shifts predicted by the premise.

The paper concludes that the renormalization-group running of neutrino mixing is a real and useful mathematical tool for connecting high-energy theories to low-energy observations, but it is not something that can be detected by comparing different experiments. The mixing parameters are fixed properties of the neutrino, determined by its mass and the particles it interacts with, not by the momentum transfer of the experiment. This finding clears up a significant misconception in the field, ensuring that future searches for new physics will focus on the correct signatures, such as new interactions during propagation or non-standard effects at the source, rather than looking for a change in the fundamental mixing constants themselves.

Technical Summary: The Renormalization-Group Running of Neutrino Mixing is Undetectable

Problem Statement
A growing body of literature posits that the renormalization-group (RG) running of neutrino mixing parameters is directly observable. This "premise" suggests that the mixing matrix entering the production or detection of a neutrino depends on the momentum transfer, $Q$ , of the interaction, specifically by evaluating the running mixing matrix $U(\mu)$ at the renormalization scale $\mu = Q$ . If true, this would imply that experiments operating at different energy scales (e.g., pion decay at $Q \sim m_\pi$ versus deep-inelastic scattering at $Q \sim \text{GeV}$ or $\text{TeV}$ ) would measure different mixing matrices. Proposed signatures include flavor change at zero distance from the source, shifts in the measured CP-violation phase, and anomalous flavor compositions of high-energy astrophysical neutrinos.

Methodology and Argument Structure
The authors refute this premise by distinguishing between the mixing matrix as a Lagrangian parameter, $U(\mu)$ , and the physical mixing matrix, $\hat{U}$ , which is an observable property of the neutrino itself. The analysis proceeds through the following logical steps:

Definition of Physical Mixing: The physical mixing matrix $\hat{U}$ is defined by the eigenvectors of the physical neutrino mass matrix, $m_{\text{eff}}$ . This matrix includes all quantum corrections (self-energies) evaluated for a neutrino on its mass shell ( $p^2 = m_i^2$ ). Since the neutrino propagates macroscopic distances between production and detection, it remains on its mass shell throughout.
Role of the Self-Energy: The self-energy $\Sigma(p^2)$ depends on the masses of particles in the loop (e.g., $W, Z, H$ , or new heavy states) and the renormalization scale $\mu$ , but crucially, it does not depend on the momentum transfer $Q$ of the external $W$ boson.
Cancellation Mechanism: The RG running of the parameter $U(\mu)$ is designed to cancel the $\mu$ -dependence of loop corrections. If one sets $\mu = Q$ (as the premise does) to minimize logarithms in vertex corrections, the self-energy on the external neutrino leg acquires large logarithms of the form $\ln(Q^2/M_X^2)$ , where $M_X$ are the masses of particles in the loop. These logarithms rotate the parameter $U(Q)$ back to the physical mixing $\hat{U}$ .
Vertex vs. Propagator Corrections: The authors demonstrate that large rotations of the mixing (the "degeneracy enhancement" responsible for significant RG running) arise only from the self-energy on the external neutrino line, where the propagator is near its pole. Vertex and box corrections, which do depend on $Q$ , lack this enhancement factor ( $1/(m_i^2 - m_j^2)$ ) and are suppressed by loop factors and couplings. Thus, they cannot mimic the large running predicted by the premise.
Explicit Models: The argument is verified in four specific contexts:
- The Minimal Supersymmetric Standard Model (MSSM) with large $\tan\beta$ .
- A seesaw model with a GeV-scale Heavy Neutral Lepton (HNL).
- A seesaw model with a light scalar singlet.
- The Standard Model Effective Field Theory (SMEFT).

Key Results

Universality of $\hat{U}$ : Every experiment, regardless of the momentum transfer $Q$ of the production or detection process, measures the same physical mixing matrix $\hat{U}$ . The premise that $U(\mu=Q_p) \neq U(\mu=Q_d)$ is incorrect because it retains the running parameter $U(Q)$ while discarding the counter-rotating logarithms in the self-energy that restore $\hat{U}$ .
Zero-Distance Flavor Change: If the premise were correct, neutrinos produced in pion decay and detected at a few GeV would exhibit flavor oscillations even at zero baseline ( $L=0$ ). The authors show that existing short-baseline data (NuTeV, NOMAD) already constrain any change in mixing angles between $Q \sim m_\pi$ and $Q \sim \text{few GeV}$ to be less than $0.7^\circ$ – $2.4^\circ$ .
Exclusion of Specific Benchmarks: In the HNL example, the premise predicts flavor changes at zero distance that are 7 to 67 times larger than current experimental limits, effectively ruling out the premise for these specific models.
Astrophysical Neutrinos: The predicted shifts in the flavor composition of high-energy astrophysical neutrinos (e.g., in IceCube) due to RG running do not occur. Any observed deviations from standard mixing expectations must be attributed to other new physics (e.g., propagation effects, non-standard interactions, or non-unitarity), not RG running.

Significance and Claims
The paper claims that the RG running of neutrino mixing is undetectable as a dependence on the momentum transfer of the process.

Clarification of RG Running: The authors emphasize that RG running is a valid and essential tool for relating high-scale model parameters (e.g., in Grand Unified Theories) to low-energy observables. However, this running describes how the Lagrangian parameters evolve with the renormalization scale $\mu$ to compensate for loop corrections; it does not imply that different physical experiments see different mixing matrices.
Correction of Literature: The work identifies and corrects a fundamental misconception in a number of recent studies (listed in Table I of the paper) that treated $U(\mu=Q)$ as the physical mixing.
Methodological Guidance: The paper advises that future phenomenological analyses should not parametrize potential new physics as a change in the mixing matrix between production and detection. Instead, any $Q$ -dependent flavor effects should be modeled as non-unitary corrections to the production/detection amplitudes (vertex and box diagrams), which are naturally small and distinct from the physical mixing.

In conclusion, the paper asserts that the mixing matrix is a property of the neutrino, fixed by its self-energy on the mass shell, and is independent of the process that produces or detects it. Consequently, the signatures proposed under the "running mixing" premise (zero-distance flavor change, energy-dependent mixing shifts) are not physical consequences of RG running.

Technical Summary: The Renormalization-Group Running of Neutrino Mixing is Undetectable

More like this