Stark Comb: Quantum Control & Spectroscopy

• Stark comb refers to systems combining discretized frequency combs with Stark-induced phase or frequency control, enabling applications in quantum memory and spectroscopy.
• These architectures leverage electric-field tuning to achieve on-demand recall in atomic frequency comb memories and spatially map microwave comb lines in Rydberg vapor-cell arrays.
• Key performance factors include comb tooth width, available microwave power, and precise AC Stark shift calibration, which together govern operational limitations and measurement accuracy.

“Stark comb” denotes a family of comb-based spectroscopic and quantum-control architectures in which the Stark effect is integral to frequency selection, phase control, or parameter extraction. In the literature considered here, the term is used in more than one technically specific sense. In one usage, it refers to an atomic frequency comb memory whose rephasing is stopped and restarted by electric-field pulses through the linear Stark effect (Horvath et al., 2020). In another, it denotes a hybrid structure formed from a microwave frequency comb and a position-dependent Stark field in a Rydberg vapor-cell array (Jiao et al., 30 Sep 2025). Closely related work also treats the Stark effect as the dominant metrological limitation in direct frequency-comb spectroscopy of cesium (Kim et al., 2017), or uses dual frequency combs as the spectrometer for vibrational Stark spectroscopy in the mid-infrared (Szczepaniak et al., 2019). This suggests that “Stark comb” is best understood as a contextual term rather than a single standardized device class.

1. Terminology and scope

The principal usages represented in the cited literature differ in both physical platform and operational objective. One meaning is explicitly tied to quantum memory: a standard atomic frequency comb (AFC) is modified so that collective emission is suppressed and later revived by Stark-induced phase control of two ion subensembles in $\mathrm{Pr^{3+}:Y_2SiO_5}$ (Horvath et al., 2020). A second meaning is explicitly architectural: the “Stark comb” comprises a microwave frequency comb (MFC) and a position-dependent Stark field, which together map discrete comb lines to different positions in a scalable Rydberg vapor-cell array (Jiao et al., 30 Sep 2025).

A broader comb–Stark relationship appears in two additional contexts. In direct frequency-comb spectroscopy of the cesium $6S_{1/2}\to 8S_{1/2}$ two-photon transition, the decisive systematic uncertainty is the laser-induced AC Stark shift, which is measured as a function of comb power and removed by extrapolation to zero power (Kim et al., 2017). In vibrational Stark spectroscopy, a dual frequency-comb QCL spectrometer is used to recover Stark spectra and extract the Stark tuning rate of fluorobenzene in frozen 2-MeTHF, with the same derivative-based analysis used in conventional FTIR VSS (Szczepaniak et al., 2019).

Usage	Physical system	Operational role of the Stark effect
Stark-controlled AFC memory	$\mathrm{Pr^{3+}:Y_2SiO_5}$, projected $\mathrm{Eu^{3+}:Y_2SiO_5}$	Stops and restarts AFC rephasing by opposite linear Stark phases of two ion classes
Rydberg-array Stark comb	Scalable Rydberg vapor-cell array	Maps cell position to microwave resonance so each MFC line acts as one cell’s LO
Comb spectroscopy with Stark-limited accuracy	Atomic cesium hot vapor DFCS	AC Stark shift is the dominant systematic uncertainty
Dual-comb vibrational Stark spectroscopy	Fluorobenzene in frozen 2-MeTHF	Stark-induced vibrational difference spectra measured with QCL dual combs

A common misconception is that “Stark comb” necessarily means a comb spectrum created by Stark splitting alone. The cited AFC and Rydberg papers state otherwise. In the AFC case, the comb remains an AFC whose collective phase evolution is electrically gated (Horvath et al., 2020). In the Rydberg case, the comb is hybrid: discrete MFC lines in frequency space combined with a spatial Stark map of the atomic resonance (Jiao et al., 30 Sep 2025).

2. Stark-controlled atomic frequency comb memory

In the AFC-memory context, the problem is the fixed recall time of standard AFC storage. For an AFC with tooth spacing $\Delta$, rephasing occurs automatically at

$t=\frac{1}{\Delta}.$

True on-demand recall in conventional AFC is commonly achieved by spin-wave transfer using bright optical control pulses, but that route introduces noise from intense optical pulses and substantial experimental complexity (Horvath et al., 2020).

The Stark-controlled extension replaces optical control pulses with electric-field control. In $\mathrm{Y_2SiO_5}$, when an electric field is applied along the crystallographic $b$ axis, the rare-earth ions split into two electrically inequivalent classes with equal-magnitude and opposite-sign linear Stark shifts. If the Stark shift magnitude is $\Omega$, then during a pulse of duration $T$ the two classes acquire phases $6S_{1/2}\to 8S_{1/2}$0 and $6S_{1/2}\to 8S_{1/2}$1 (Horvath et al., 2020). The relevant shift is

$6S_{1/2}\to 8S_{1/2}$2

The collective state after absorption is written as

$6S_{1/2}\to 8S_{1/2}$3

If the first Stark pulse satisfies

$6S_{1/2}\to 8S_{1/2}$4

the two classes become opposite in phase in the emitted field contribution, and the AFC echo is suppressed. A second identical pulse later removes the differential phase, so recall occurs at a later revival time

$6S_{1/2}\to 8S_{1/2}$5

The retrieval time is therefore selectable in discrete steps of $6S_{1/2}\to 8S_{1/2}$6 rather than being fixed at the first AFC echo (Horvath et al., 2020).

The experimental realization used $6S_{1/2}\to 8S_{1/2}$7 at $6S_{1/2}\to 8S_{1/2}$8 K, 500 ppm doping, and the $6S_{1/2}\to 8S_{1/2}$9 transition at 606 nm. The prepared AFC had $\mathrm{Pr^{3+}:Y_2SiO_5}$0, giving a basic revival period of about $\mathrm{Pr^{3+}:Y_2SiO_5}$1. The input optical pulse was Gaussian with FWHM $\mathrm{Pr^{3+}:Y_2SiO_5}$2. The Stark pulse had amplitude $\mathrm{Pr^{3+}:Y_2SiO_5}$3 V and FWHM $\mathrm{Pr^{3+}:Y_2SiO_5}$4 ns, corresponding to the $\mathrm{Pr^{3+}:Y_2SiO_5}$5 phase condition for the two ion classes (Horvath et al., 2020).

The reported recall efficiency was $\mathrm{Pr^{3+}:Y_2SiO_5}$6 for a storage time of $\mathrm{Pr^{3+}:Y_2SiO_5}$7. With weak coherent states of mean photon number $\mathrm{Pr^{3+}:Y_2SiO_5}$8, the measured signal-to-noise ratio was $\mathrm{Pr^{3+}:Y_2SiO_5}$9 for $\mathrm{Eu^{3+}:Y_2SiO_5}$0 storage, and the unconditional noise floor was effectively detector dark counts because no optical control pulse was used (Horvath et al., 2020). The paper further reports recall up to the tenth echo with a ratio $\mathrm{Eu^{3+}:Y_2SiO_5}$1 compared to the no-revival case.

The main storage-time limitation in $\mathrm{Eu^{3+}:Y_2SiO_5}$2 is not the nominal AFC echo time $\mathrm{Eu^{3+}:Y_2SiO_5}$3, but dephasing from finite comb tooth width. The forward-efficiency model used for the data is

$\mathrm{Eu^{3+}:Y_2SiO_5}$4

with

$\mathrm{Eu^{3+}:Y_2SiO_5}$5

Experimentally, the narrowest comb tooth width analyzed was $\mathrm{Eu^{3+}:Y_2SiO_5}$6, linked to an optical coherence time of about $\mathrm{Eu^{3+}:Y_2SiO_5}$7 (Horvath et al., 2020).

Using an impedance-matched cavity model for $\mathrm{Eu^{3+}:Y_2SiO_5}$8, with assumed linewidth $\mathrm{Eu^{3+}:Y_2SiO_5}$9, optical depth $\Delta$0, and mirror reflectivities $\Delta$1 and $\Delta$2, the paper estimates $\Delta$3 for storage time $\Delta$4 (Horvath et al., 2020). A plausible implication is that the protocol’s usefulness depends less on the Stark-control principle itself than on whether the host material supports sufficiently narrow comb peaks.

3. Hybrid spatial–frequency Stark combs in Rydberg microwave reception

In the Rydberg-microwave context, the Stark comb is defined explicitly as

$\Delta$5

(Jiao et al., 30 Sep 2025). The purpose is to overcome the narrow instantaneous bandwidth of a single Rydberg vapor-cell receiver by converting a set of cells into parallel atomic heterodyne channels.

The receiver uses a Rydberg EIT ladder with

$\Delta$6

Without Stark field, the $\Delta$7 microwave transition frequency is about $\Delta$8. Under a position-dependent Stark field, it becomes

$\Delta$9

or, in the one-dimensional array geometry,

$t=\frac{1}{\Delta}.$0

(Jiao et al., 30 Sep 2025).

The MFC in the proof-of-principle experiment had 21 lines, 10 MHz spacing, and center frequency around 8.13 GHz. The Stark field was engineered so that neighboring cell positions differed in resonance by about 10 MHz, matching the comb spacing. The operating condition is therefore that one cell position $t=\frac{1}{\Delta}.$1 is tuned to one comb line $t=\frac{1}{\Delta}.$2, so that the comb line acts as the local oscillator for that cell, while other comb lines remain off-resonant and produce weak, negligible responses (Jiao et al., 30 Sep 2025).

Heterodyne detection then proceeds by mixing the input signal with the nearest comb line in the atomic medium, generating a beat note with

$t=\frac{1}{\Delta}.$3

In the reported measurements, beat-note peaks were taken at $t=\frac{1}{\Delta}.$4. A single cell has instantaneous bandwidth about $t=\frac{1}{\Delta}.$5, so the comb spacing was chosen as

$t=\frac{1}{\Delta}.$6

for contiguous coverage (Jiao et al., 30 Sep 2025).

The total-bandwidth scaling law is

$t=\frac{1}{\Delta}.$7

With $t=\frac{1}{\Delta}.$8 effective channels, the demonstrated total instantaneous bandwidth was

$t=\frac{1}{\Delta}.$9

covering 8.025 GHz to 8.235 GHz (Jiao et al., 30 Sep 2025). The overall sensitivity was reported as $\mathrm{Y_2SiO_5}$0, while at the 8.13 GHz channel the minimum detectable field was $\mathrm{Y_2SiO_5}$1 in $\mathrm{Y_2SiO_5}$2, corresponding to $\mathrm{Y_2SiO_5}$3 (Jiao et al., 30 Sep 2025).

The physical array was emulated by moving one vapor cell to 21 positions. The cell size was $\mathrm{Y_2SiO_5}$4, and the minimum spacing between two cells was reported as 0.23 cm. The position-dependent Stark field was generated by a 120 MHz RF field applied to aluminum plates of size $\mathrm{Y_2SiO_5}$5, with 12 cm edge separation (Jiao et al., 30 Sep 2025). Probe and coupling lasers at 852 nm and 509 nm counterpropagated through the cell, with $\mathrm{Y_2SiO_5}$6, $\mathrm{Y_2SiO_5}$7, and beam waists $\mathrm{Y_2SiO_5}$8 and $\mathrm{Y_2SiO_5}$9, respectively.

The paper’s “arbitrary instantaneous bandwidth” claim is explicitly conditional: it holds provided one can design the position-dependent Stark fields properly and provide enough MFC lines with sufficient power (Jiao et al., 30 Sep 2025). The same paper also states explicit constraints: available MFC power limits line count, sensitivity degrades at larger Stark tuning because the microwave dipole moment decreases, and excessive Stark field may cause state mixing and shift nearby cells out of usable resonance conditions. This indicates that “arbitrary” is a scalability statement in principle rather than the absence of engineering bounds.

4. AC Stark effects in direct frequency-comb spectroscopy

A distinct comb–Stark relation appears in direct frequency-comb spectroscopy of the cesium $b$0 two-photon transitions in a hot vapor cell (Kim et al., 2017). Here the comb is the spectroscopic source itself, and the dominant Stark-related issue is not comb formation or phase gating, but laser-induced AC Stark shift as the principal uncertainty in an absolute-frequency measurement.

The optical frequency comb was generated by a mode-locked Ti:sapphire oscillator centered near $b$1 ($b$2) with bandwidth $b$3 ($b$4 FWHM). Its tooth frequencies satisfy

$b$5

and for the two-photon resonance,

$b$6

The scan conversion is

$b$7

with $b$8, so that $b$9 in $\Omega$0 corresponds to about $\Omega$1 in optical frequency (Kim et al., 2017).

The Doppler-free geometry is produced by coherent control rather than by any special Stark engineering. The pulse spectrum is divided at the exact half-frequency of the two-photon resonance into red and blue parts, temporally separated in a 4-$\Omega$2 pulse shaper, with the red pulse delayed by $\Omega$3 relative to the blue pulse. A counter-propagating replica is sent in the opposite direction, so Doppler-free two-photon absorption occurs where opposite-direction red and blue sub-pulses arrive simultaneously (Kim et al., 2017). The paper explicitly states that this coherent-control method addresses the Doppler problem, whereas the AC Stark shift remains the dominant systematic.

Detection uses $\Omega$4 fluorescence from the cascade

$\Omega$5

The observed spectrum contains the resolved lines $\Omega$6 and $\Omega$7. Under representative conditions of 35 mW average comb power, 10 nm spectral width, and averaging 30 scans, the linewidth is about $\Omega$8 FWHM, larger than the $\Omega$9 natural linewidth because of residual Doppler effects, laser linewidth, and transit-time broadening (Kim et al., 2017).

The reported absolute frequencies are

$T$0

for $T$1, and

$T$2

for $T$3, with fractional uncertainty below about $T$4 (Kim et al., 2017).

Among the systematic effects considered—pressure shift, transit-time broadening, Zeeman shift, and AC Stark shift—only the AC Stark shift was significant. The authors state that in this weak-field comb regime the AC Stark shift is linearly proportional to the average laser intensity, explicitly not the peak intensity. Using a reference coefficient of about

$T$5

they estimate an expected shift of about $T$6 at 50 mW. They then measure the line center as a function of laser power and extract slopes of

$T$7

for $T$8 and

$T$9

for $6S_{1/2}\to 8S_{1/2}$00 (Kim et al., 2017).

At the typical operating power of 34.8 mW, the AC Stark correction uncertainty contribution was taken as 289 kHz for $6S_{1/2}\to 8S_{1/2}$01 and 358 kHz for $6S_{1/2}\to 8S_{1/2}$02, dominating the final 0.33–0.40 MHz error bars (Kim et al., 2017). The practical procedure is explicit: measure frequency versus comb power, fit a linear trend, and extrapolate to zero power. This suggests that in comb-based metrology the Stark effect may be central even when the phrase “Stark comb” does not denote the apparatus itself.

5. Dual frequency combs in vibrational Stark spectroscopy

In vibrational Stark spectroscopy, the comb device is the spectrometer rather than the sample-side control mechanism. The cited work demonstrates a mid-infrared QCL dual-frequency-comb spectrometer for vibrational Stark spectroscopy on fluorobenzene in frozen 2-MeTHF, benchmarking it against FTIR (Szczepaniak et al., 2019). The experiment shows that dual-comb spectroscopy can recover the same Stark spectral information and the same Stark tuning rate as FTIR, but in much shorter acquisition time.

The spectrometer was an IRsweep IRis-F1 based on two free-running QCL frequency combs. Each comb spanned about $6S_{1/2}\to 8S_{1/2}$03 and was centered near $6S_{1/2}\to 8S_{1/2}$04. The overlap region was 1173 to $6S_{1/2}\to 8S_{1/2}$05, giving 57 $6S_{1/2}\to 8S_{1/2}$06 of covered bandwidth and spectral point spacing $6S_{1/2}\to 8S_{1/2}$07. Each comb tooth had average power $6S_{1/2}\to 8S_{1/2}$08, and the total CW output power was about 700 mW (Szczepaniak et al., 2019).

The dual-comb mapping uses multi-heterodyne detection. If the repetition frequencies are $6S_{1/2}\to 8S_{1/2}$09 and $6S_{1/2}\to 8S_{1/2}$10, the beat-note spacing is

$6S_{1/2}\to 8S_{1/2}$11

and the time resolution is

$6S_{1/2}\to 8S_{1/2}$12

Because the QCL cavities are short and have high repetition rates, the quoted time resolution is $6S_{1/2}\to 8S_{1/2}$13 (Szczepaniak et al., 2019).

The sample was 100 mM fluorobenzene in 2-MeTHF, frozen as an isotropic glass. The Stark cell used two $6S_{1/2}\to 8S_{1/2}$14-thick, 12.7 mm diameter CaF$6S_{1/2}\to 8S_{1/2}$15 windows with 4.5 nm Ni coatings as electrodes. The sample thickness, determined by interferometry at room temperature, was $6S_{1/2}\to 8S_{1/2}$16 for DCS and $6S_{1/2}\to 8S_{1/2}$17 for FTIR (Szczepaniak et al., 2019). The applied electric field was estimated by

$6S_{1/2}\to 8S_{1/2}$18

For DCS, 2.0 kV and 3.0 kV across $6S_{1/2}\to 8S_{1/2}$19 corresponded to $6S_{1/2}\to 8S_{1/2}$20 and $6S_{1/2}\to 8S_{1/2}$21, respectively.

The Stark observable is the field-on minus field-off absorbance spectrum, analyzed through the derivative expansion

$6S_{1/2}\to 8S_{1/2}$22

The absorbance band $6S_{1/2}\to 8S_{1/2}$23 was modeled with a pseudo-Voigt profile. The second-derivative coefficient was converted to the Stark tuning rate using

$6S_{1/2}\to 8S_{1/2}$24

in the paper’s analysis convention (Szczepaniak et al., 2019).

The extracted Stark tuning rates for the fluorobenzene C–F stretch were

$6S_{1/2}\to 8S_{1/2}$25

for DCS and

$6S_{1/2}\to 8S_{1/2}$26

for FTIR, both consistent with the previously reported value $6S_{1/2}\to 8S_{1/2}$27 (Szczepaniak et al., 2019). The DCS benchmark used total acquisition time $6S_{1/2}\to 8S_{1/2}$28, whereas the FTIR benchmark used $6S_{1/2}\to 8S_{1/2}$29, giving a factor of 250 in acquisition time. The paper also states that the Stark response was already visible with 128 ms acquisition time in DCS (Szczepaniak et al., 2019).

The raw fit-based SNR values at matched $6S_{1/2}\to 8S_{1/2}$30 resolution were 9.5 for DCS and 20.2 for FTIR, but after correcting for acquisition time and field strength the comparison gave an approximately $6S_{1/2}\to 8S_{1/2}$31 DCS SNR advantage (Szczepaniak et al., 2019). Since the DCS noise was wavelength dependent, the data were fitted with pointwise inverse-variance weighting, an analysis feature not used for the FTIR comparison.

This branch of the literature does not use “Stark comb” to name the sample-side physics. Instead, the Stark effect defines the spectroscopic observable, while the dual comb supplies fast, bright, multiplexed mid-IR acquisition. A plausible implication is that the phrase can designate either a Stark-engineered comb device or a comb-enabled Stark measurement, depending on disciplinary context.

6. Conceptual relations, distinctions, and recurring design principles

Despite their different implementations, the cited systems share a common structural pattern: the comb provides a discretized frequency or time grid, while the Stark effect provides tunability, phase control, or the perturbation to be measured. In the AFC memory, the discretization is the tooth spacing $6S_{1/2}\to 8S_{1/2}$32, and the Stark effect changes the collective phase relation between two ion classes so that echoes can be suppressed and revived at chosen revival slots (Horvath et al., 2020). In the Rydberg receiver, the discretization is the set of MFC lines, and the Stark field maps those lines onto distinct cell positions (Jiao et al., 30 Sep 2025). In cesium DFCS, the comb defines the absolute optical-frequency relation, while AC Stark shift sets the dominant uncertainty budget (Kim et al., 2017). In vibrational Stark spectroscopy, the dual comb multiplexes the spectral acquisition of a field-induced derivative signal whose parameter of interest is the Stark tuning rate (Szczepaniak et al., 2019).

There are equally important distinctions. The AFC and Rydberg papers describe explicit Stark-engineered device functionality: echo gating in one case, bandwidth scaling through spatial-frequency multiplexing in the other (Horvath et al., 2020, Jiao et al., 30 Sep 2025). By contrast, the cesium and fluorobenzene papers show comb-based spectroscopy in which the Stark effect is, respectively, a systematic shift to be removed or the physical response to be quantified (Kim et al., 2017, Szczepaniak et al., 2019). The term therefore spans at least three operational categories: Stark control of collective phase, Stark tuning of resonance assignment, and Stark-sensitive comb spectroscopy.

Several limitations recur across the literature. In Stark-controlled AFC, storage time is ultimately limited by comb tooth width and optical coherence rather than by the first AFC revival time (Horvath et al., 2020). In the Rydberg-array architecture, scalability is limited by available MFC power, Stark-field engineering, and the risk of state mixing at large tuning (Jiao et al., 30 Sep 2025). In cesium DFCS, the metrological bottleneck is the AC Stark shift, which must be removed empirically by power extrapolation (Kim et al., 2017). In dual-comb VSS, practical limitations include limited spectral bandwidth relative to FTIR, uneven comb power distribution, and duty-cycle inefficiency in the implementation used (Szczepaniak et al., 2019).

Taken together, these works indicate that “Stark comb” is not a single canonical object but a cross-cutting designation for comb-enabled systems in which Stark physics is operationally central. The most literal current definitions are the Stark-controlled AFC memory (Horvath et al., 2020) and the Rydberg hybrid spatial–frequency comb (Jiao et al., 30 Sep 2025). The broader literature shows that even where the phrase is not the formal device name, Stark effects can set either the dominant control primitive or the dominant uncertainty mechanism in frequency-comb-based spectroscopy (Kim et al., 2017, Szczepaniak et al., 2019).