Cambridge Physicists Directly Image Quantum Vacuum Fluctuations Using Ultracold Atomic Gas

Cool atoms down to near absolute zero and confine them in a uniform two-dimensional box. What remains looks like a perfectly still condensate—but according to the principles of quantum mechanics, even the lowest-energy ground state cannot leave physical quantities perfectly at rest.

A research team led by Yansheng Zhang at the Cavendish Laboratory, University of Cambridge, has published a preprint describing an experiment that simulated a relativistic sine-Gordon field using a two-component Bose-Einstein condensate of ultracold potassium-39 atoms, and directly imaged the spatial pattern of quantum fluctuations present in its ground state. The paper appeared on the preprint server arXiv on August 20, 2026 (arXiv:2608.20311).

Some science media coverage has used phrases like "directly photographing the vacuum," but this achievement did not capture an image of empty space itself or the vacuum of the electromagnetic field. Rather, the researchers engineered the spin degrees of freedom of a precisely controlled ultracold atomic gas in the laboratory into an "analog quantum field," and captured the vacuum fluctuations inherent to that field through state-selective spatial imaging. With this distinction in mind, this article examines why the observation matters for the fundamental physics of quantum fields.

Note that the paper is a seven-page preprint with three figures that has not yet completed the peer-review process at an academic journal.

Beyond half a century of indirect evidence: why the fluctuations themselves remained unseen

In modern physics, the vacuum is not complete nothingness. Heisenberg's uncertainty principle, a cornerstone of quantum mechanics, dictates that conjugate physical quantities cannot in principle be simultaneously fixed. This property applies directly to quantum fields with their own degrees of freedom.

Each mode of a quantum field is described as an independent quantum harmonic oscillator. Just as the position and momentum of a harmonic oscillator do not commute, the value of a quantum field and its rate of change over time (the conjugate momentum) cannot be simultaneously determined. As a result, even in the ground state of minimum energy at absolute zero—the "vacuum"—each point in the field retains unavoidable random fluctuations arising from zero-point energy. These are vacuum fluctuations.

Vacuum fluctuations have long been recognized as the fundamental driving force behind numerous physical phenomena. Spontaneous emission, in which an atom in an excited state emits a photon and naturally transitions to the ground state; the Casimir effect, in which two metal plates facing each other in an otherwise empty vacuum are pulled together; and Hawking radiation, in which particle pairs are created near a black hole's event horizon—none of these phenomena can be explained without the existence of vacuum fluctuations.

However, what previous experiments captured were merely the "results" of vacuum fluctuations acting on macroscopic matter or electromagnetic fields. In the Casimir effect, this meant an averaged mechanical quantity—the force acting on the plates; in spontaneous emission, it meant detection of the emitted photon. Directly capturing an image of how the field itself fluctuates spatially, and with what correlations across different length scales, has proven extremely difficult.

Comparison axis	This experiment (BEC analog system)	Electromagnetic field vacuum (free space)	Previous indirect evidence (Casimir force, etc.)
Field under study	Ultracold atoms' spin degrees of freedom (relative phase and particle number difference)	Electromagnetic field in free space (photon field)	Electromagnetic field or atomic electron orbitals
Observation method	Direct snapshot of spatial field fluctuations via absorption imaging	Direct spatial imaging extremely difficult	Detection of macroscopic forces or emitted photons
Theoretical field model	Massive relativistic sine-Gordon field	Quantum electrodynamics (QED, massless gauge field)	Interaction effects of quantum electrodynamics
Temperature and purity	$T=(6\pm 6)\text{ nK}$ (thermal occupation number $N_k\lesssim 0.1$)	Cosmic microwave background (~2.7 K) or ultracold shielded environment	Laboratory temperature to ultracold
Parameter controllability	Effective mass and coupling constant dynamically tunable via Rabi frequency	Physical constants fixed, cannot be externally altered	Only geometric configuration (plate spacing, etc.) adjustable

How potassium atoms weave a "sine-Gordon field"

To overcome this challenge, the Cambridge research team constructed an analog quantum simulation system using an ultracold Bose-Einstein condensate (BEC).

The stage for the experiment was a potassium-39 (${}^{39}\text{K}$) atomic gas confined in a two-dimensional flat box trap made of light (an optical box trap). The atoms were cooled to the extreme limit, forming a highly pure two-dimensional condensate in which thermal motion had almost completely vanished. In this system, the researchers coherently coupled two different hyperfine spin states of the potassium atoms, realizing a uniformly mixed gas.

In this two-component system, it is the spin degree of freedom that behaves as the quantum field. Specifically, the relative phase $\varphi (x,y)$ between the two spin states and the local particle-number imbalance $Z(x,y)$ at each point serve as the field variables. These two physical quantities satisfy canonical commutation relations analogous to the position-momentum relation $[x,p]=i\hslash$ of a harmonic oscillator.

The research team precisely controlled the Rabi coupling frequency ${\Omega }^{\prime }$, which drives the two spin states with electromagnetic waves (radio frequency), and the atomic interaction energy ${\mu }_s^{\prime }$. In the regime where interactions overwhelm the externally applied coherent coupling ($\hslash {\Omega }^{\prime }/{\mu }_s^{\prime }\ll 1$, known as the Josephson regime), this spin field theoretically behaves as a sine-Gordon field.

The sine-Gordon model is a type of massive relativistic quantum field whose potential is expressed as $U(\varphi )\propto \cos \varphi$. In this system, the field's effective mass is given by $m^{\text{*}}=m\sqrt{2\hslash \Omega /{\mu }_s}$, using the atom's bare mass $m$, the Rabi coupling frequency $\Omega$, and the interaction ${\mu }_s$—meaning the field's mass can be freely manipulated simply by adjusting the externally applied Rabi frequency.

In earlier work (arXiv:2603.08840, published March 2026), the same team had already demonstrated that this platform follows a relativistic dispersion relation with a tunable mass gap, and that it forms topological domain walls—phase defects in 2+1 dimensions. Building on that precisely controlled system, the present study set out to visualize the ground-state fluctuations themselves after eliminating thermal excitations to the greatest extent possible.

Two techniques for revealing the fluctuations: quench amplification and adiabatic direct imaging

Fluctuations in the ground state of a quantum field are exceedingly small and easily buried in detector noise or measurement error. To address this challenge, the research team employed two independent approaches.

The first technique is parametric amplification through a rapid quench. The spin field is first prepared in its ground state within a regime of strong Rabi coupling. From there, the Rabi frequency ${\Omega }^{\prime }$ is abruptly reduced to a small value over an extremely short time. This parametrically excites the tiny vacuum fluctuations already present in the field, amplifying them into measurable oscillations in amplitude as the system evolves in time.

If the system were a classical harmonic oscillator with zero energy in its initial state, no oscillation would be excited even after an abrupt change in parameters. The fact that macroscopic oscillations appeared after the quench is evidence that fluctuations originating from zero-point energy were present in the initial state.

The second technique is adiabatic direct observation, which captures the fluctuations as they are without amplification. The research team slowly and linearly reduced the Rabi frequency ${\Omega }^{\prime }$ over 99 milliseconds, guiding it down to the target final value. This scan rate was designed to satisfy the adiabatic condition $d{\omega }_k/dt<{\omega }_k^2$ relative to the natural frequency ${\omega }_k$ of each mode of the field. This allows each mode to continuously track the ground state of the changing Hamiltonian without being excited.

At high wavenumbers, the natural frequency ${\omega }_k$ is large, so the adiabatic condition $d{\omega }_k/dt<{\omega }_k^2$ holds and the ground state remains stably preserved. At low wavenumbers, however, the natural frequency ${\omega }_k$ becomes small, creating a region where the adiabatic condition breaks down relative to the scan rate. The research team identified this boundary and applied a strict distinction when interpreting the data.

State measurement was performed using spatially resolved, in-situ state-selective absorption imaging. Measurements of atomic density directly yielded the particle-number imbalance $\delta Z(x,y)$. To measure fluctuations in the relative phase $\delta \varphi (x,y)$, meanwhile, a radio-frequency pulse was applied immediately before imaging to rotate the spin vector on the Bloch sphere, converting phase fluctuations into particle-number differences.

The observed fluctuations in $\delta Z$ and $\delta \varphi$ were confirmed to oscillate in time in quadrature—analogous to the momentum and position of a classical oscillator.

The decisive test that ruled out thermal noise: the 6-nanokelvin threshold

The resulting snapshots recorded random patterns of light and dark at various spatial scales (wavenumbers $k$) simultaneously. The most rigorous test that had to be applied here was determining whether these patterns arose from "quantum vacuum fluctuations" or from residual "thermal noise" or measurement noise in the experimental system.

The research team performed a detailed analysis of the power spectrum of the observed fluctuations. If the fluctuations arise from a quantum mechanical ground state, the fluctuation amplitude $A_Y$ should theoretically follow a ${\omega }_k^{-1}$ scaling law relative to the mode's natural frequency ${\omega }_k$. If thermal excitation dominates instead, the spectrum's frequency dependence takes a different form.

The amplitude $A_Y$ obtained from the direct-observation data matched the ground-state theoretical prediction of ${\omega }_k^{-1}$ scaling across a wide range of wavenumbers $k$ and final Rabi frequencies ${\Omega }_f^{\prime }$.

To further quantify the influence of thermal excitation, the team performed a fit using finite temperature $T$ as a variable. This analysis yielded an effective spin temperature of:

Expressed relative to the interaction energy ${\mu }_s^{\prime }$, this corresponds to $(0.2\pm 0.2){\mu }_s^{\prime }$—consistent with absolute zero ($T = 0$) within experimental error. Across the entire measurement region where the adiabatic condition held, the thermal occupation number $N_k$ of each mode corresponding to a temperature of $k_{B}T=0.2{\mu }_s^{\prime }$ remained at or below $0.1$. In other words, thermally excited particles (quasiparticles) were essentially absent, confirming that the dominant component of the observed spatial pattern was purely quantum fluctuation.

The absolute value of the amplitude $A_Y$ also matched the $T=0$ theoretical prediction, using a detection efficiency $\eta$ calibrated independently from the quench-amplification experiment. Values derived from multiple independent lines of analysis all point to the same quantum-theoretical conclusion.

Toward cosmological puzzles: the horizon opened by analog quantum fields, and remaining challenges

The significance of this study extends beyond capturing a spatial image of vacuum fluctuations—it also demonstrates the possibility of directly simulating nonlinear relativistic quantum fields in the laboratory, systems that are notoriously difficult to compute theoretically.

In the paper, the research team suggests that by extending this platform, laboratory-scale approaches could become possible for the following unresolved physical phenomena:

• False-vacuum decay: a phase transition phenomenon in which a metastable vacuum state tunnels into the true vacuum
• Particle creation via parametric instability: the mechanism of matter creation through an expanding universe or a strong external field
• The quantum Kibble-Zurek mechanism: the universal law governing the density of phase defects generated when crossing a quantum phase transition
• Cosmic reheating/preheating: the process by which vacuum energy converts into heat or particles after inflation ends
• Dynamics of topological domain wall networks: the evolution of phase defects believed to have formed in the early universe

In particular, combining this approach with periodic driving (Floquet engineering) is expected to enable direct simulation of relativistic bubble nucleation processes in the early universe.

However, several premises and limitations must be properly understood when interpreting this achievement.

First, as noted above, this system is an "analog simulator" that maps collective excitations (spin waves) of a two-dimensional BEC onto a relativistic field—it did not directly observe the vacuum of the electromagnetic field itself. It exploits a mathematical isomorphism in which the effective action describing the system coincides with the sine-Gordon model within a specific regime.

Second, part of the experimental data shows a breakdown of the adiabatic condition. As shown in Figure 3 of the paper, at long wavelengths (low wavenumber $k$), the natural period becomes long enough that even a 99-millisecond scan time fails to satisfy the adiabatic condition, creating a region (shaded in the paper's parameter space) that requires careful treatment when interpreting it as ground-state fluctuation.

Third, the paper remains at the preprint stage; independent peer review and replication by other research groups remain future tasks.

A swarm of atoms confined near absolute zero has captured an image of the quantum "presence" that continuously wells up throughout space. How far this precise control technology can push back the boundaries of computing the nonequilibrium dynamics of relativistic fields—dynamics that have long confounded theoretical calculation—remains to be seen through future verification.