Persistent Single-Electron Emission Overview

• Persistent single-electron emission is the process where individual electrons are emitted as discrete charge quanta via mechanisms like Coulomb blockade, periodic gating, and field emission in xenon detectors.
• It encompasses engineered sources such as quantum dots and mesoscopic capacitors, as well as background emissions in liquid xenon devices, each characterized by unique control parameters like RC time constants and tunneling rates.
• This phenomenon is critical for advancing quantum electronics applications including on-demand electron sources, free-electron quantum optics, and improving low-energy thresholds in dark matter and neutrino detection.

Searching arXiv for recent and foundational papers on persistent single-electron emission and related single-electron sources. Persistent single-electron emission is the sustained release, recurrence, or clocked generation of individual electrons as discrete charge quanta rather than as an undifferentiated many-electron current. In contemporary literature, the term covers at least three experimentally distinct settings: Coulomb-blockade-controlled emission from a carbon nanowire vacuum point source that operates at room temperature and up to $I\sim1\,\mu$A (Kleshch et al., 2020); on-demand mesoscopic emitters based on quantum dots and mesoscopic capacitors, where one electron and one hole can be emitted per drive cycle [(Mahé et al., 2010); (Albert et al., 2011); (Fletcher et al., 2012); (Brange et al., 2020)]; and persistent or delayed single-electron backgrounds in two-phase xenon detectors, where Fowler–Nordheim tunneling, photoionization, delayed thermal emission, and Malter-type surface charging generate signals on timescales ranging from microseconds to hours [(0708.0768); (Santos et al., 2011); (Bodnia et al., 2021); (Va'vra, 1 Aug 2025)]. The common thread is discreteness of charge; the major distinctions are whether the phenomenon is engineered as a source or encountered as a background, and whether the governing kinetics are set by Coulomb blockade, periodic gate drive, interfacial extraction, or surface charging.

1. Conceptual scope and principal mechanisms

Persistent single-electron emission is not a single mechanism but a class of phenomena in which individual-electron transfer remains experimentally resolvable over extended operation. In nanoscale emitters, discreteness is imposed either by Coulomb blockade or by a periodic gate waveform that loads and ejects a single carrier per cycle. In two-phase xenon detectors, the same discreteness appears as isolated single-electron S2 signals, but there the origin is parasitic: photoionization of impurities or grids, delayed extraction of thermalized electrons, Fowler–Nordheim field emission from microscopic protrusions, or, on much longer timescales, Malter-type emission from resistive oxide patches [(Kleshch et al., 2020); (Bodnia et al., 2021); (Santos et al., 2011); (Va'vra, 1 Aug 2025)].

The relevant control parameters differ sharply across platforms. In Coulomb-blockaded emitters, the island capacitance $C$, tunnel resistance $R$, and gate coupling $C_g$ determine the charging energy $E_c=e^2/(2C)$ and the admissible emission rate. In periodically driven solid-state emitters, the key scales are the tunnel escape time $\tau$, the period $T$, the drive amplitude relative to the level spacing $\Delta$, and the tunnel transmission of the quantum point contact. In dual-phase xenon systems, extraction and drift fields, electrode surface quality, impurity content, and oxide resistivity set both the spontaneous rate and the persistence of delayed emission [(Mahé et al., 2010); (Albert et al., 2011); (Bodnia et al., 2021)].

A recurrent misconception is that all single-electron signals have the same physical origin. The literature does not support that view. Instead, it separates intentionally quantized emission in mesoscopic conductors and vacuum nanostructures from detector-specific backgrounds in liquid xenon, even when similar Fowler–Nordheim expressions appear in both contexts. This suggests that “persistent” is best understood operationally: it denotes indefinite repetition or long-lived recurrence of single-electron release, not a unique microscopic process.

2. Coulomb-blockade-controlled emission in vacuum point sources

A vacuum implementation of persistent single-electron emission was realized by integrating Coulomb-blockade physics with nanoscale field emission in a heterostructured tip composed of an ultra-sharp diamond needle coated by a thin amorphous-carbon layer from which a single carbon nanowire grows (Kleshch et al., 2020). Fabrication proceeds by field-emission-assisted Joule heating in ultra-high vacuum. Cyclic voltage ramps up to $1$–$10\,\mu$A gradually graphitize the diamond surface into amorphous carbon and, under the strong apex field from a biased mesh gate, induce surface diffusion and nucleation of an $C$0-bonded carbon nanowire with diameter $C$1–$C$2 nm and length tunable from $C$3 to $C$4 nm.

At the base of the nanowire, a Schottky-type barrier against the amorphous carbon forms a double-barrier emitter: a tunnel junction to the substrate with resistance $C$5 and capacitance $C$6, and a field-emission barrier at the nanowire/vacuum interface capacitively coupled to the gate through $C$7. The total island capacitance is $C$8, the charging energy is

$C$9

and the difference in electrostatic energy between $R$0 and $R$1 electrons on the nanowire is

$R$2

Steady-state occupations follow a master equation in terms of the rates $R$3 and $R$4 for tunneling into and out of the $R$5-electron state. Field emission at the nanowire apex is described by a curvature-corrected Fowler–Nordheim relation,

$R$6

and experimentally the current-voltage curve is well fitted by

$R$7

The experimental signatures are simultaneous in current and energy space. When the gate voltage is ramped, the emission current exhibits a clear staircase, while the normalized differential conductance $R$8 oscillates periodically with period $R$9. Energy-resolved spectroscopy shows “sawtooth” oscillations in the total-energy distribution $C_g$0: between conductance peaks only one emission line appears and shifts by $C_g$1, whereas at each conductance maximum a second line emerges, separated by the charging energy $C_g$2. From $C_g$3, one extracts $C_g$4; from the energy spacing $C_g$5, one obtains $C_g$6.

These single-electron features persist up to emission currents of order $C_g$7A and at room temperature, but they are suppressed in two distinct regimes. In the high-$C_g$8 case, with $C_g$9 ps and $E_c=e^2/(2C)$0 M$E_c=e^2/(2C)$1, Coulomb oscillations are washed out when $E_c=e^2/(2C)$2 becomes comparable to $E_c=e^2/(2C)$3, so that multiple charge states coexist. The condition for observing Coulomb blockade is

$E_c=e^2/(2C)$4

In the low-$E_c=e^2/(2C)$5 case, with $E_c=e^2/(2C)$6 fs and $E_c=e^2/(2C)$7 M$E_c=e^2/(2C)$8, Joule-heating-induced temperature rise broadens the energy lines and thermally smears the blockade once $E_c=e^2/(2C)$9 at $\tau$0 K. The critical current is $\tau$1A, corresponding to an effective tunneling rate $\tau$2 THz and an average inter-emission time $\tau$3 ps. The measured $\tau$4 values, spanning $\tau$5 fs to $\tau$6 ps, therefore define the operational window for continuous single-electron emission.

3. Clocked emission in mesoscopic conductors

In mesoscopic conductors, persistent single-electron emission is realized not as dc vacuum field emission but as indefinite, clock-synchronized repetition of a single-particle cycle. One implementation uses a quantum dot coupled to a conductor via a tunable quantum-point contact, driven by a fast square-wave gate voltage so that on the rising edge an electron tunnels out and on the falling edge a hole is emitted as the dot re-fills (Mahé et al., 2010). When $\tau$7, one electron and one hole are emitted in each half-cycle, yielding zero dc current but a quantized ac current of $\tau$8 per period. In the ideal regime, with unit emission probability and $\tau$9, the ensemble-averaged pulse is

$T$0

with a mirror pulse of opposite sign in the second half-cycle.

A closely related theoretical framework is the mesoscopic capacitor, modeled as a sub-micron cavity coupled through a QPC to a $T$1 edge state in the integer quantum Hall regime (Albert et al., 2011). In the semi-classical description, each period is divided into an absorption phase and an emission phase, with one-electron occupation $T$2. The characteristic correlation time is

$T$3

where $T$4. The high-frequency current-noise spectrum is

$T$5

In the phase-noise regime $T$6, the mean emitted-electron current approaches $T$7, and the Fano factor

$T$8

tends to zero, indicating nearly noiseless on-demand emission.

Time-domain control has also been developed in a dynamic single-electron transistor in the Coulomb-blockade regime, where a harmonic gate voltage

$T$9

drives repeated loading and unloading of a single-electron state (Brange et al., 2020). The time-dependent rates obey

$\Delta$0

with fitted parameters $\Delta$1, $\Delta$2, $\Delta$3 kHz, and $\Delta$4 kHz. Measured waiting-time distributions show a crossover from adiabatic to nonadiabatic dynamics as $\Delta$5 approaches and then exceeds the tunnel rates; in the high-frequency regime, the distribution develops peaks at $\Delta$6. The reported representative values are $\Delta$7 ms and $\Delta$8 ms at $\Delta$9 kHz, and $1$0 ms with $1$1 ms at $1$2 kHz.

A higher-energy variant uses a two-gate tunable-barrier quantum dot driven at $1$3 GHz to emit electrons with excess kinetic energy $1$4 meV into an empty edge channel (Fletcher et al., 2012). At $1$5 T, the phonon-emission probability $1$6 falls below $1$7, $1$8 exceeds $1$9, and from a $10\,\mu$0m source-detector separation the inferred inelastic scattering length is $10\,\mu$1m. Time-resolved spectroscopy yields a temporal wavepacket width $10\,\mu$2 ps and an energy spread of order a few meV; switching of two electrons into different paths is demonstrated with better than $10\,\mu$3 control. Taken together, these results establish the mesoscopic meaning of persistence: long-lived, cycle-by-cycle single-electron reproducibility.

4. Persistent and delayed single-electron emission in two-phase xenon detectors

In two-phase xenon time-projection chambers, persistent single-electron emission is primarily a background rather than a source function. The first measurements of the electroluminescence response to single-electron emission in such a detector were reported in ZEPLIN-II, where a mean single-electron S2 pulse of $10\,\mu$4 photoelectrons with $10\,\mu$5 photoelectrons was observed, consistent with an electroluminescence yield of $10\,\mu$6 VUV photons per extracted electron under the stated operating conditions (0708.0768). Throughout a 31-day background run, small secondary-like pulses appeared stochastically in the $10\,\mu$7–$10\,\mu$8s interval between S1 and main S2; roughly $10\,\mu$9–$C$00 of events contained one or more such pulses, corresponding on average to $C$01 single-electron signals per trigger, or $C$02 Hz. No significant time dependence was seen over the month, and the rate was interpreted as being far above expectations from thermionic emission or ordinary field emission at the stated mesh fields.

ZEPLIN-III extended this picture and separated several mechanisms: photoionization by VUV scintillation, field-induced extraction of “hot” electrons at the liquid surface, delayed thermal emission of electrons that fail immediate extraction, and Fowler–Nordheim field emission from cathode wires (Santos et al., 2011). Under typical conditions, the cross-phase extraction probability was $C$03 in the first science run and $C$04 in the second. The detector achieved $C$05 photoelectrons per extracted electron, with observed single-electron widths $C$06–$C$07 photoelectrons. The measured spontaneous or delayed single-electron rates were $C$08 s$C$09 in the dedicated single-electron run, $C$10 s$C$11 in the first science run, and $C$12 s$C$13 in the second science run; in a $C$14Cs-illuminated test, the raw rate was $C$15 s$C$16, reduced to $C$17–$C$18 s$C$19 by imposing $C$20–$C$21s vetoes. The time histogram of post-S1 single electrons also provided a direct in situ measure of the free-electron lifetime via

$C$22

with an example fit yielding $C$23s.

PIXeY quantified the field dependence of persistent single-electron backgrounds using $C$24Kr calibration events (Bodnia et al., 2021). By fitting the steady pre-S1 single-electron rate to a Fowler–Nordheim form,

$C$25

and using $C$26, PIXeY extracted a field-enhancement factor $C$27 and an effective protrusion area $C$28 cm$C$29. Over extraction fields $C$30–$C$31 kV/cm, the pre-S1 rate grew from $C$32 s$C$33 to $C$34 s$C$35, consistent with $C$36. PIXeY further found that the single-electron rate between S1 and S2, normalized to S1 area, declined by $C$37 as the drift field rose from $C$38 to $C$39 kV/cm, while the S2-tail single-electron rate tracked the extraction efficiency. Its $C$40s observation window showed quasi-steady single-electron rates punctuated by grid-photoelectric spikes rather than a clear multi-exponential decay.

A later interpretation argues that long-lived “hot spots” in LXe TPCs are consistent with a Malter-type surface-charging mechanism on native oxide films (Va'vra, 1 Aug 2025). In that framework, stainless-steel or Cu–Be wires develop $C$41–$C$42 nm oxide layers whose cryogenic resistivity reaches $C$43–$C$44, giving an oxide-film discharge time constant

$C$45

At $C$46, $C$47 s. The local field across a $C$48 nm film is estimated as

$C$49

yielding $C$50–$C$51 V/cm for $C$52. With an FN-style local current density $C$53, a $C$54m$C$55 hot spot at $C$56 A/cm$C$57 emits $C$58 e$C$59/s, sufficient for $C$60 Hz single-electron rates after geometric and detection losses. This framework was proposed specifically to explain persistence from minutes to hours and localization to recurring wire positions.

5. Temporal statistics, noise floors, and observables

The central diagnostic of persistent single-electron emission depends on platform. In clocked mesoscopic sources, average current alone is insufficient; short-time current correlations, waiting-time distributions, and high-frequency noise are used to distinguish true one-by-one emission from fluctuating multi-particle transport. For the on-demand quantum-dot source, the current correlator is defined as

$C$61

with noise spectral density

$C$62

In the perfect-emission limit, the irreducible contribution is the “quantum-jitter” noise floor,

$C$63

which arises from the exponential distribution of emission times even when every trigger emits exactly one electron (Mahé et al., 2010).

The mesoscopic capacitor reaches an analogous phase-noise regime when $C$64, where the only remaining fluctuations are timing fluctuations rather than missed cycles or extra particles (Albert et al., 2011). Its Lorentzian-type spectrum,

$C$65

vanishes at $C$66 and saturates at high frequency. The same model yields full counting statistics through the largest eigenvalue of the period propagator and gives $C$67 as $C$68, directly encoding near-ideal single-electron reproducibility.

In the dynamic SET, the waiting-time distribution provides a more direct temporal observable (Brange et al., 2020). For constant rates, it reduces to

$C$69

and for $C$70,

$C$71

Under periodic modulation, the distribution crosses over from a single broad adiabatic peak to a comb of peaks at $C$72, showing explicit locking of emission events to the drive period.

In the vacuum nanowire emitter, the relevant temporal observable is the competition among the average inter-emission time $C$73, the device $C$74 time, and the quantum uncertainty time $C$75 (Kleshch et al., 2020). There, persistence is not set by a clock but by continuous operation within the inequality $C$76. By contrast, in xenon TPCs the time-domain observables are delayed single-electron tails and spikes referenced to S1, S2, and the maximum drift time. PIXeY reported a prompt S1-induced burst at $C$77–$C$78s, a roughly constant background until S2, a constant single-electron-rate tail over $C$79–$C$80s after S2, and a sharp cathode spike at $C$81s (Bodnia et al., 2021). The contrast between these observables is important: in engineered emitters, fluctuations diagnose source fidelity; in TPCs, they diagnose background composition.

6. Applications, mitigation, and unresolved questions

Engineered persistent single-electron emission is pursued because it enables single-electron electronics, quantum information processing, and free-electron quantum optics. The carbon nanowire vacuum emitter was explicitly proposed as a platform that can be combined with femtosecond laser pulses for “laser-induced gating,” with the aim of synchronizing emission on sub-optical-cycle timescales and generating coherent ultrashort electron bunches for low-energy electron holography and ultrafast electron or X-ray imaging and spectroscopy (Kleshch et al., 2020). In solid-state circuits, clock-controlled quantum-dot pumps and mesoscopic capacitors provide the building blocks for coherent single-particle manipulation, with demonstrated wavepacket transport over several microns, high-energy emission, and path switching of individual electrons (Fletcher et al., 2012).

In liquid-xenon detectors, however, persistent single-electron emission sets the low-energy threshold. ZEPLIN-III concluded that a 3-electron threshold, corresponding to $C$82 photoelectrons, suppresses the $C$83 s$C$84 single-electron background to negligible coincidence rates, and it assessed coherent neutrino-nucleus scattering sensitivity in the few-electron regime (Santos et al., 2011). PIXeY likewise emphasized that persistent single-electron backgrounds are critical for sub-GeV WIMP and hidden-sector searches and proposed a background model in which Fowler–Nordheim and photoionization terms are combined with extraction efficiency (Bodnia et al., 2021).

Mitigation strategies are mechanism-specific. For field emission from protrusions, the literature recommends improving wire-surface quality, lowering $C$85, minimizing stressed area, applying insulating coatings or guard electrodes, and reducing grid voltage per unit length (Bodnia et al., 2021). For photoionization and delayed extraction, the recommended measures include ultra-high purification, operation at higher drift fields, software vetoes such as ignoring the first $C$86–$C$87s after S2, and gating grids to blank S2-tail fields (Bodnia et al., 2021). For Malter-type hot spots, the recommended design changes are more materials-focused: gold-plating of wires, alternative chemistries such as silver or palladium, avoidance of tight three-dimensional meshes that trap ions, and direct cryogenic measurements of oxide resistivity and dielectric constant (Va'vra, 1 Aug 2025).

A persistent ambiguity in the field is whether a given long-lived single-electron population in LXe is dominated by delayed extraction, photoionization, Fowler–Nordheim tunneling, or oxide-mediated feedback. The published record does not reduce the problem to one universal mechanism. PIXeY’s $C$88s observations support quasi-steady FN and photoionization components, whereas the Malter framework is explicitly aimed at seconds-to-hours hot spots. This suggests that persistence in xenon detectors is temporally stratified: microsecond behavior and multi-hour behavior need not share the same microscopic origin. By contrast, in engineered emitters the unresolved issues are more often optimization problems—reducing timing jitter without excessive level broadening, maintaining one-electron purity at high repetition rate, and extending coherence control from charge counting to full wavepacket interferometry.