🔬 optics

Photon statistics of resonantly driven spectrally diffusive quantum emitters

This paper theoretically demonstrates that analyzing photon statistics under resonant excitation allows for the discrimination between continuous and discrete spectral diffusion models in solid-state emitters, offering deeper insights into emission stability and clarifying the mechanisms behind recent experimental observations of B centers in hexagonal boron nitride.

Original authors: Aymeric Delteil, Stéphanie Buil, Jean-Pierre Hermier

Published 2026-02-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Aymeric Delteil, Stéphanie Buil, Jean-Pierre Hermier

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

Imagine you have a tiny, glowing light bulb embedded inside a solid piece of material (like a diamond or a crystal). This light bulb is special because it's supposed to emit single photons (particles of light) one by one, which is crucial for future quantum technologies.

However, there's a problem. The solid material surrounding the light bulb isn't perfectly still; it's like a crowded room where people are constantly bumping into each other. These bumps cause the light bulb's color (its frequency) to wobble and shift randomly over time. Scientists call this "spectral diffusion."

If the color shifts too much, the photons become "out of tune" with each other, making them useless for high-tech applications that require perfect synchronization.

The authors of this paper wanted to figure out how this color wobbling happens. They asked: Is the color drifting smoothly like a boat on a gentle wave, or is it jumping abruptly like a frog hopping from lily pad to lily pad?

To answer this, they didn't just watch the color shift directly (which is hard to do). Instead, they shined a laser on the emitter and looked at the pattern of the light flashes (photon statistics). They found that the way the light flickers tells a story about the underlying motion.

Here is the breakdown of their findings using simple analogies:

The Two Types of "Wobbling"

The paper compares two main theories for how the color shifts:

The Smooth Drift (Ornstein-Uhlenbeck Process):
- The Analogy: Imagine a drunk person walking home. They are swaying and drifting continuously. Their path is a messy, unbroken line. They don't teleport; they just move slowly and steadily in one direction before changing course.
- The Physics: The emitter's energy level drifts continuously. It's coupled to a huge crowd of tiny, independent fluctuators (like many people gently pushing the emitter).
The Discrete Jumps (Gaussian Random Jump Model):
- The Analogy: Imagine a frog sitting on a lily pad. It sits still for a while, then suddenly poof—it jumps to a new, random spot. It stays there, then jumps again. It never moves in between the jumps.
- The Physics: The emitter stays at one energy level for a while, then suddenly "hops" to a completely different energy level due to a charge moving nearby.

How They Tell the Difference

The researchers shined a laser on these emitters and measured how the light intensity fluctuated over time. They looked for two specific "fingerprints" to distinguish the smooth drift from the frog-like jumps.

1. The "Bunching" Test (How the light clumps together)

When the emitter is close to the laser's color, it glows brightly. When it drifts away, it goes dark.

The Smooth Drift: Because the color drifts slowly, once the emitter gets close to the laser's color, it tends to stay there for a while. The longer you shine the laser (increasing power), the longer it stays "in tune," and the light clumps together for a longer time.
- Result: The time the light stays "bunched" together changes depending on how strong the laser is.
The Discrete Jumps: The emitter sits still until it suddenly jumps away. It doesn't matter how strong the laser is; the time it sits still is determined by how often the frog decides to jump, not by the laser.
- Result: The time the light stays "bunched" together stays the same regardless of the laser power.

2. The "Histogram" Test (The shape of the brightness distribution)

If you take a long photo of the light's brightness over time and plot how often different brightness levels occur:

The Smooth Drift: The distribution of brightness looks like a standard, symmetrical bell curve (Poissonian). It's predictable.
The Discrete Jumps: The distribution becomes lopsided (skewed). You get a lot of average brightness, but you also get rare, very bright spikes. This happens because the emitter stays in a "bright" state for a random amount of time (exponentially distributed), creating a "Gamma distribution" shape.
- Result: If the brightness histogram is lopsided, it's a sign of jumps. If it's symmetrical, it's likely a smooth drift.

The Real-World Discovery

The authors applied this logic to a specific type of defect in hexagonal boron nitride (called a "B center"). Previous experiments showed these centers had spectral diffusion, but no one knew the mechanism.

By looking at the light statistics, they found that the "bunching time" did not change when they increased the laser power. This was the smoking gun. It proved that the B centers don't drift smoothly; they jump like frogs.

Summary

In short, the paper says: You don't need to see the color shift to know how it moves. By simply listening to the rhythm of the light flashes (photon statistics), you can tell if the emitter is drifting smoothly or hopping randomly. This helps scientists understand the "noise" in their quantum devices and figure out how to fix it.

They also noted that this method works for a specific type of emitter (B centers in boron nitride) and provides a new way to study other solid-state light sources without needing complex, high-speed equipment.

Technical Summary: Photon statistics of resonantly driven spectrally diffusive quantum emitters

Problem Statement
Solid-state single-photon emitters (e.g., quantum dots, color centers) are promising for quantum technologies due to their integration potential and photophysical quality. However, their performance is often limited by spectral diffusion (SD)—stochastic fluctuations in the emission wavelength induced by the host matrix. These fluctuations reduce the indistinguishability of emitted photons, a critical resource for quantum applications. While various experimental techniques exist to characterize SD (e.g., photon correlation Fourier spectroscopy, sub-linewidth filtering), they often face limitations in temporal or spectral resolution, particularly when SD occurs faster than the inverse count rate. Recent experimental work demonstrated that resonant laser excitation combined with photon correlations can reveal SD characteristics, but a detailed theoretical framework connecting the microscopic nature of SD processes to the resulting photon statistics was lacking.

Methodology
The authors theoretically investigate the photon statistics of a two-level quantum emitter driven by a resonant continuous-wave laser, subject to spectral diffusion. The study focuses on two distinct, stationary Markovian models that yield Gaussian inhomogeneous broadening:

Ornstein-Uhlenbeck (OU) Process: A continuous spectral diffusion model, representing emitters coupled to a large ensemble of independent fluctuators (e.g., charge or spin noise).
Gaussian Random Jump (GRJ) Model: A discrete spectral jump model, representing emitters coupled to a low number of fluctuators or trap states, where the frequency jumps at random times governed by a Poisson process.

The analysis assumes the adiabatic approximation ( $\tau_{SD} \gg T_1, 1/\Sigma$ ), where the emitter reaches a steady state faster than spectral fluctuations occur. The intensity response is modeled as $C(t) = C_{ss}(\omega(t))$ , converting spectral fluctuations into intensity fluctuations. The authors derive analytical expressions and perform numerical simulations (using Euler-Maruyama integration for OU and stochastic jump algorithms for GRJ) to analyze:

The intensity autocorrelation function $g^{(2)}(\tau)$ .
The short-time bunching $B(0) = g^{(2)}(0) - 1$ .
The bunching decay time $\tau_b$ .
Long-time intensity fluctuations, specifically the Mandel parameter $Q$ and the skewness $\gamma_1$ of the integrated intensity distribution.

Key Contributions and Results
The paper establishes that the photon statistics of resonantly driven emitters carry distinct signatures allowing discrimination between continuous (OU) and discrete (GRJ) spectral diffusion mechanisms.

Short-Time Bunching ( $B(0)$ ): The magnitude of bunching at zero delay is found to be independent of the specific SD mechanism (OU vs. GRJ) and the correlation time. It depends solely on the laser detuning, power, and the ratio of homogeneous to inhomogeneous linewidths. This allows for the extraction of the homogeneous linewidth (and thus coherence time) from static lineshape and bunching measurements without assumptions about the microscopic SD mechanism.
Bunching Decay and Power Dependence: The temporal decay of the normalized bunching $\tilde{B}(\tau)$ reveals the nature of the SD process:
- OU Model: The decay is non-exponential. Crucially, the characteristic bunching time $\tau_b$ increases with laser power (above saturation) due to power broadening, which prolongs the time the emitter stays near resonance. In the limit of high power, $\tau_b$ saturates at $\tau_{SD}/2$ .
- GRJ Model: The decay is strictly exponential. The bunching time $\tau_b$ is equal to the spectral diffusion correlation time $\tau_{SD}$ and is independent of laser power. The power only affects the probability of being on resonance (and thus $B(0)$ ), not the duration of the resonance.
Long-Time Intensity Fluctuations:
- Variance (Mandel Parameter): Both models exhibit increased intensity noise compared to the Poisson limit. However, the power dependence of this noise differs: for GRJ, noise reduction occurs near saturation power; for OU, it occurs at much higher powers where the homogeneous linewidth approaches the inhomogeneous width.
- Skewness: A critical distinction is found in the skewness of the integrated intensity distribution. The GRJ model yields a super-Poissonian skewness ( $\gamma_1 \approx 2\Delta I / \langle I \rangle$ ), characteristic of a gamma distribution arising from exponentially distributed stable periods between jumps. In contrast, the OU model yields a near-Poissonian skewness ( $\gamma_1 \approx \Delta I / \langle I \rangle$ ).

Significance and Claims
The authors claim that resonant excitation combined with photon statistics offers a rich, experimentally accessible method to characterize spectral diffusion processes beyond simple inhomogeneous line shape analysis. The primary significance lies in the ability to distinguish between continuous diffusion and discrete jumps using:

The power dependence of the bunching decay time (power-dependent for OU, power-independent for GRJ).
The skewness of the long-time intensity distribution (super-Poissonian for GRJ, Poissonian for OU).

The paper applies these findings to recent experimental data on B centers in hexagonal boron nitride (hBN). The observation of power-independent bunching times in these experiments leads the authors to attribute the SD mechanism of B centers to a discrete jump process (GRJ), providing new insight into their microscopic behavior. The framework is presented as generalizable to other SD models and experimental protocols, including non-resonant excitation with narrowband filtering, though the authors note that such extensions require careful interpretation of bunching times relative to the SD time. The work aims to aid the understanding and optimization of solid-state single-photon sources for quantum computing and networks.