Statistics of superfluorescent pulses in three-dimensional, relaxing samples: completing the quantitative theory and its experimental validation beyond one dimension and leading order in $1/N$

Superfluorescence — the cooperative, quantum-initiated spontaneous emission of a fully inverted ensemble of two-level atoms — produces delayed pulses whose shot-to-shot fluctuations are macroscopic. In 1979 Haake, King, Schröder, Haus, and Glauber computed complete probability distributions for the parameters of such pulses (delay times, first-peak durations, ringing structure, mean and variance of the radiated intensity) from a quasiprobability treatment of the one-dimensional Maxwell-Bloch equations with random initial polarizations, and stated in their conclusions what remained open: transverse (three-dimensional) effects, relaxation and dephasing ($T_1$, $T_2$, $T_2^{*}$), possible numerical suppression of phase disturbances (Hopf's critique), the lack of controlled experimental statistics, and unquantified higher-order $1/N$ corrections. Delay-time statistics were quantitatively closed (exact linear one-dimensional theory with $T_2^{*}$, cesium measurements, a three-dimensional large-Fresnel-number linear theory, and three-dimensional simulations with quantum initiation that reconcile the cesium single-shot data). The full pulse-shape statistics for three-dimensional, nonuniformly excited, relaxing samples — with controlled finite-$N$ accuracy and experimental validation of the full distributions — remain open and are actively re-encountered in x-ray, semiconductor, and perovskite superfluorescence.

Background

Superfluorescence (SF, also called superradiance in the extended-sample sense) is the cooperative spontaneous emission of a pencil-shaped ensemble of $N$ two-level atoms prepared, by pulsed pumping, essentially all in the excited state and then left to radiate freely. After a macroscopic delay time $\tau_D$ during which quantum fluctuations nucleate a macroscopic polarization, the ensemble emits a burst whose peak intensity scales as $N^{2}$ and whose duration scales as $1/N$ — as anticipated by Dicke's 1954 analysis of cooperative radiation — and whose emission is concentrated into narrow beams along the sample axis. It was first observed in optically pumped HF gas by Skribanowitz, Herman, MacGillivray, and Feld (1973), and in its cleanest single-shot form in a cesium atomic beam by Gibbs, Vrehen, and Hikspoors (1977). Because the initial state is an unstable equilibrium, the microscopic initiating noise is amplified to macroscopic size, so pulse parameters (delay, duration, shape) fluctuate strongly from shot to shot — a textbook example of macroscopic quantum fluctuations. By 1979 the theory had three branches. The mean-field single-mode theory of Bonifacio and Lugiato (1975) couples the atoms to one field mode with no spatial propagation; it admits analytic statistics but predicts shape-invariant pulses. The propagation theory of MacGillivray and Feld (1976) integrates the Maxwell-Bloch equations (the coupled propagation equation for the slowly varying field envelope and the two-level Bloch equations for polarization and inversion) along an extended sample, but must inject an ad hoc small initial 'tipping angle' $\theta_0$ — the angle by which the collective Bloch vector is displaced from the pole — to start the classical emission. The linear quantum initiation theory of Polder, Schuurmans, and Vrehen (1979) derives the same Maxwell-Bloch equations with a fluctuating quantum source and yields the relative delay-time standard deviation $\sigma(\tau_D)/\langle\tau_D\rangle = 2.3/\ln N$ for one-dimensional propagation; Ressayre and Tallet (1976) obtained the mean radiated intensity by decoupling expectation-value equations. Haake, King, Schröder, Haus, and Glauber (Phys. Rev. A 20, 2047 (1979)) unified these views: representing the density operator by a quasiprobability (a phase-space functional associated with the field and polarization envelope operators, obtained as the functional Fourier transform of a moment-generating functional), they showed that to leading order in $1/N$ the quantum dynamics reduces to the classical Maxwell-Bloch equations driven by random initial polarizations — Gaussian, with local width of order $1/\sqrt{N}$ — so that quantum expectation values become functional integrals over classical trajectories. Sampling that ensemble by a Fourier-series realization truncated at modest order, an 11-point spatial grid, and three-point Hermite quadrature (a Gaussian-type numerical integration over the initial-noise distribution), they computed probability distributions of the delay time, the first-peak duration, the ringing (the damped multi-peak tail following the first intensity maximum), and the mean and variance of the radiated intensity, in a strictly one-dimensional model: axial propagation only, radially uniform fields, valid for a cylinder with Fresnel number $F\approx 1$, where the Fresnel number $F = A/(\lambda L)$ ($A$ the transverse area, $\lambda$ the wavelength, $L$ the length) counts the transverse diffraction modes. All relaxation was omitted: the population decay time $T_1$, the homogeneous polarization dephasing time $T_2$, and the inhomogeneous (e.g. Doppler) dephasing time $T_2^{*}$. The paper's own conclusions listed what remained open. First, for $F\approx 1$ 'the radial dependence of the fields is not truly negligible', the pumped density decreases with radius, and only a ray-optics 'fiber' argument indicated 'the kind of correction to be anticipated from a three-dimensional treatment'. Second, relaxation was argued negligible only for the best cesium experiments; its quantitative effect on pulse statistics was not computed. Third, Hopf (Phys. Rev. A 20, 2064 (1979)), in a companion paper, argued that 'phase waves' — propagation of relative-phase disturbances between polarization and field — contribute roughly 30% of the statistics and conjectured that the low-order Fourier truncation and 11-point gridding suppress them; Haake et al. acknowledged they were 'investigating it further'. Fourth, the comparison with the four available cesium pulses showed discrepancies (experimentally observed first peaks up to a factor of two longer, with distinctly less ringing), and 'there are not yet sufficient experimental data to compare with the full content of our results'. Fifth, corrections to the classical dynamics were argued to be of order $1/N$ after the initial stage, but the neglected higher-order functional-derivative terms of the quasiprobability equation were never quantified at finite $N$. These five strands, one coherent objective — extend the quantitative statistical theory of superfluorescent pulses to realistic samples and test it — constitute the problem.

Problem Statement

Haake, King, Schröder, Haus, and Glauber (Phys. Rev. A 20, 2047 (1979)) computed, within an idealized one-dimensional model without relaxation and to leading order in $1/N$, the probability distributions of the parameters of superfluorescent pulses — delay times, first-peak durations, ringing structure (the damped secondary intensity maxima following the first peak), and the mean and variance of the radiated intensity. The source question, stated in the paper's own conclusions, is the completion of that statistical description under realistic conditions, as one coherent objective: a quantitatively reliable, experimentally validated statistical theory of superfluorescent pulses for realistic extended samples. Its four linked components are: (i) Three-dimensional geometry. Determine quantitatively how transverse (radial) field structure, the radially decreasing excitation density of realistic pumped cylinders, and the competition of axial with off-axial modes modify the predicted pulse statistics — the distributions of delay, duration, ringing, and their variances — for samples with Fresnel number (the ratio of the sample's transverse area to the product of optical wavelength and sample length, counting the transverse diffraction modes) of order one and above, where the one-dimensional treatment is only marginally applicable. (ii) Relaxation. Incorporate radiationless decay and dephasing — finite $T_1$ (population decay time), $T_2$ (homogeneous polarization dephasing time), and $T_2^{*}$ (inhomogeneous dephasing time), comparable to the pulse widths and delay times — and compute the resulting modifications of the ensemble moments of the pulse statistics, beyond the delay-time mean and variance. (iii) Accuracy of the statistical theory itself. Establish the accuracy of the leading-order (in $1/N$, where $N$ is the number of atoms) quasiprobability treatment (the phase-space functional representation of the density operator underlying the 1979 calculation): quantify the effect of the neglected higher-order functional-derivative terms on the fine statistics at finite $N$, and determine whether the spatial and Fourier discretizations used to sample the statistical ensemble suppress dynamical phase disturbances of the polarization-field relative phase, the effect Hopf conjectured (F. A. Hopf, Phys. Rev. A 20, 2064 (1979)). (iv) Experimental validation. Specify the controlled-ensemble measurement — fixed atom number, density, and active-volume shape across shots — that would deliver the shot-by-shot distributions of pulse parameters, and state the theoretical predictions those distributions are to be compared against, so that the full pulse-shape statistics can be tested directly, beyond the simulation-level reconciliation already achieved for the 1979 cesium single-shot data (experimentally observed first peaks up to a factor of two longer, with less ringing than predicted). As established in the literature record, delay-time statistics have been quantitatively closed in the regimes treated so far — one dimension with $T_2^{*}$, and the linearized three-dimensional theory for large Fresnel numbers — leaving the three-dimensional delay statistics at Fresnel numbers of order one within the scope of component (i); the cesium pulse-shape discrepancies were reconciled at the simulation level. The surviving open core is the full pulse-shape statistics — durations, ringing, and intensity variances — for three-dimensional, nonuniformly excited, relaxing samples, together with controlled finite-$N$ accuracy and experimental validation of the complete distributions. The objective comprises the four components above: (i) the modified distributions of delay, duration, and ringing and their variances for a nonuniformly excited three-dimensional sample at Fresnel numbers of order one; (ii) the same ensemble moments of the pulse statistics recomputed with finite $T_1$, $T_2$, and $T_2^{*}$, beyond the delay-time mean and variance; (iii) a quantified leading finite-$N$ correction to those fine statistics together with a determination of whether the sampling discretization suppresses the Hopf phase waves; and (iv) the specified measurement together with the theoretical distributions it would test. Components (i)-(iii) are independent sub-questions and component (iv) supplies the experimental-validation comparison; the four together complete the statistical description.

References

1. F. Haake, H. King, G. Schröder, J. Haus, and R. Glauber, 'Fluctuations in superfluorescence', Physical Review A 20, 2047 (1979), doi:10.1103/PhysRevA.20.2047, https://doi.org/10.1103/PhysRevA.20.2047 (source paper)
2. F. A. Hopf, 'Phase-wave fluctuations in superfluorescence', Physical Review A 20, 2064 (1979), doi:10.1103/PhysRevA.20.2064, https://doi.org/10.1103/PhysRevA.20.2064
3. F. Haake, J. Haus, H. King, G. Schröder, and R. Glauber, 'Delay-Time Statistics and Inhomogeneous Line Broadening in Superfluorescence', Physical Review Letters 45, 558 (1980), doi:10.1103/PhysRevLett.45.558, https://doi.org/10.1103/PhysRevLett.45.558
4. D. Polder, M. F. H. Schuurmans, and Q. H. F. Vrehen, 'Superfluorescence: Quantum-mechanical derivation of Maxwell-Bloch description with fluctuating field source', Physical Review A 19, 1192 (1979), doi:10.1103/PhysRevA.19.1192, https://doi.org/10.1103/PhysRevA.19.1192
5. Q. H. F. Vrehen and J. J. der Weduwe, 'Quantum fluctuations in superfluorescence delay times', Physical Review A 24, 2857 (1981), doi:10.1103/PhysRevA.24.2857, https://doi.org/10.1103/PhysRevA.24.2857
6. H. M. Gibbs, Q. H. F. Vrehen, and H. M. J. Hikspoors, 'Single-Pulse Superfluorescence in Cesium', Physical Review Letters 39, 547 (1977), doi:10.1103/PhysRevLett.39.547, https://doi.org/10.1103/PhysRevLett.39.547
7. F. P. Mattar, H. M. Gibbs, S. L. McCall, and M. S. Feld, 'Transverse Effects in Superfluorescence', Physical Review Letters 46, 1123 (1981), doi:10.1103/PhysRevLett.46.1123, https://doi.org/10.1103/PhysRevLett.46.1123
8. P. D. Drummond and J. H. Eberly, 'Transverse coherence and scaling in four-dimensional simulations of superfluorescence', Physical Review A 25, 3446 (1982), doi:10.1103/PhysRevA.25.3446, https://doi.org/10.1103/PhysRevA.25.3446
9. E. A. Watson, H. M. Gibbs, F. P. Mattar, M. Cormier, Y. Claude, S. L. McCall, and M. S. Feld, 'Quantum fluctuations and transverse effects in superfluorescence', Physical Review A 27, 1427 (1983), doi:10.1103/PhysRevA.27.1427, https://doi.org/10.1103/PhysRevA.27.1427
10. J. Mostowski and B. Sobolewska, 'Delay-time statistics in superfluorescence for large Fresnel numbers', Physical Review A 28, 2573 (1983), doi:10.1103/PhysRevA.28.2573, https://doi.org/10.1103/PhysRevA.28.2573
11. J. Mostowski and B. Sobolewska, 'Initiation of superfluorescence from a sphere', Physical Review A 28, 2943 (1983), doi:10.1103/PhysRevA.28.2943, https://doi.org/10.1103/PhysRevA.28.2943
12. F. P. Mattar and E. Friedman, 'Quantum fluctuations, phase wave, pump depletion, and diffraction dynamics in three-level superfluorescence', OSA meeting proceedings (1986), paper TUM9, doi:10.1364/oam.1986.tum9, https://doi.org/10.1364/oam.1986.tum9
13. C. Leonardi and A. Váglica, 'Fluctuations in superfluorescent beating', Optics Communications 53, 340 (1985), doi:10.1016/0030-4018(85)90274-3, https://doi.org/10.1016/0030-4018(85)90274-3
14. F. Haake, M. I. Kolobov, and H. Steudel, 'Dynamical models for forward-backward coupling in superfluorescence', Optics Communications 92, 385 (1992), doi:10.1016/0030-4018(92)90649-C, https://doi.org/10.1016/0030-4018(92)90649-C
15. L. Buchmann, G. Nikolopoulos, O. Zobay, and P. Lambropoulos, 'Passage-time statistics of superradiant light pulses from Bose–Einstein condensates', Journal of Physics B: Atomic, Molecular and Optical Physics 44, 025301 (2011), doi:10.1088/0953-4075/44/2/025301, https://doi.org/10.1088/0953-4075/44/2/025301
16. Peter P. Vasil'ev, Vojtech Olle, Richard V. Penty, and Ian H. White, 'Experimental study of delay-time statistics of superradiant pulses generated from semiconductor structures', Applied Physics Letters 103, 241108 (2013), doi:10.1063/1.4841395, https://doi.org/10.1063/1.4841395
17. A. Benediktovitch, V. P. Majety, and N. Rohringer, 'Quantum theory of superfluorescence based on two-point correlation functions', Physical Review A 99, 013839 (2019), preprint arXiv:1810.12280, doi:10.1103/PhysRevA.99.013839, https://arxiv.org/abs/1810.12280
18. Masaki Nasu, Kohei Kawamura, Tomoharu Yoshida, Jun Ishihara, and Kensuke Miyajima, 'Influences of quantum fluctuation on superfluorescent spectra observed by single-shot measurement for semiconductor quantum dots', Applied Physics Express 13, 062005 (2020), doi:10.35848/1882-0786/ab8b50, https://doi.org/10.35848/1882-0786/ab8b50
19. M. Biliroglu, G. Findik, J. Mendes, D. Seyitliyev, L. Lei, Q. Dong, Y. Mehta, V. V. Temnov, F. So, and K. Gundogdu, 'Room-temperature superfluorescence in hybrid perovskites and its origins', Nature Photonics 16, 324 (2022), doi:10.1038/s41566-022-00974-4, https://doi.org/10.1038/s41566-022-00974-4
20. M. Biliroglu, M. Türe, A. Ghita, M. Kotyrov, X. Qin, D. Seyitliyev, N. Phonthiptokun, M. Abdelsamei, J. Chai, R. Su, U. Herath, A. Swan, V. Temnov, V. Blum, F. So, and K. Gundogdu, 'Unconventional solitonic high-temperature superfluorescence from perovskites', Nature 642, 71 (2025), doi:10.1038/s41586-025-09030-x, https://doi.org/10.1038/s41586-025-09030-x
21. Lewis Ruks, 'Semiclassical Phase-Space Dynamics of Emitter Ensembles with Local Dissipation', arXiv:2602.14025 (2026), doi:10.48550/arXiv.2602.14025, https://arxiv.org/abs/2602.14025
22. P. D. Drummond and B. Opanchuk, 'Initial states for quantum field simulations in phase space', Physical Review Research 2, 033304 (2020), doi:10.1103/PhysRevResearch.2.033304, https://doi.org/10.1103/PhysRevResearch.2.033304
23. R. H. Dicke, 'Coherence in Spontaneous Radiation Processes', Physical Review 93, 99 (1954), doi:10.1103/PhysRev.93.99, https://doi.org/10.1103/PhysRev.93.99
24. N. Skribanowitz, I. P. Herman, J. C. MacGillivray, and M. S. Feld, 'Observation of Dicke Superradiance in Optically Pumped HF Gas', Physical Review Letters 30, 309 (1973), doi:10.1103/PhysRevLett.30.309, https://doi.org/10.1103/PhysRevLett.30.309
25. R. Bonifacio and L. A. Lugiato, 'Cooperative radiation processes in two-level systems: Superfluorescence', Physical Review A 11, 1507 (1975), doi:10.1103/PhysRevA.11.1507, https://doi.org/10.1103/PhysRevA.11.1507
26. J. C. MacGillivray and M. S. Feld, 'Theory of superradiance in an extended, optically thick medium', Physical Review A 14, 1169 (1976), doi:10.1103/PhysRevA.14.1169, https://doi.org/10.1103/PhysRevA.14.1169
27. E. Ressayre and A. Tallet, 'Basic Properties for Cooperative Emission of Radiation', Physical Review Letters 37, 424 (1976), doi:10.1103/PhysRevLett.37.424, https://doi.org/10.1103/PhysRevLett.37.424