Simultaneous Co-magnetometry Explained

• Simultaneous co-magnetometry is a precision measurement method that interrogates multiple magnetic-sensitive channels concurrently to cancel common-mode noise.
• It employs differential encoding across various platforms—such as trapped ions, Bose-Einstein condensates, and solid-state sensors—to isolate weak observables from dominant background fields.
• Experimental implementations demonstrate enhanced precision and noise rejection, enabling rigorous tests of fundamental physics and novel quantum sensing applications.

Searching arXiv for recent and relevant papers on simultaneous co-magnetometry and related implementations. arXiv search query: simultaneous co-magnetometry magnetometry Ramsey quantum logic spectroscopy dual-species single-species NV Rb Xe BEC storage ring crystal Simultaneous co-magnetometry is a class of precision-measurement methods in which two or more magnetic-field-sensitive channels are interrogated in the same field environment at the same time, so that common-mode magnetic perturbations can be rejected while differential sensitivity to the target observable is retained. The channels can be different atomic species, different hyperfine or Zeeman manifolds of the same species, different transitions within identical molecules, mirror-symmetric sub-ensembles in a crystal, separate bunches in a storage ring, or hybrid solid-state and atomic sensors. Across these realizations, the defining feature is not merely the use of a reference sensor, but simultaneity: shared temporal sampling of the same magnetic environment within the same shot, scan, or experimental cycle (Rehmert et al., 21 Aug 2025, Gomez et al., 2019, Wu et al., 2018).

1. Core principle and formal structure

The canonical co-magnetometric observable is a ratio or linear combination constructed so that the dominant magnetic-field dependence cancels to first order. In a co-trapped ion implementation, the Zeeman splitting is measured simultaneously in $^{48}\mathrm{Ti}^+$ and in a $^{40}\mathrm{Ca}^+$ reference ion with a well-known $g$ factor, giving

$$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$

Because both ions experience the same magnetic field during the same experimental cycle, the field $B$ cancels to first order in the ratio (Rehmert et al., 21 Aug 2025).

A closely related structure appears in spin-precession devices. In a $^{87}\mathrm{Rb}$ spinor Bose-Einstein condensate, the $f=1$ and $f=2$ manifolds have nearly opposite gyromagnetic ratios,

$$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$

and the comagnetometer channel is the summed azimuth

$\theta^{(12)}\equiv \theta^{(1)}+\theta^{(2)}.$

The large $^{40}\mathrm{Ca}^+$0 terms cancel, leaving only the residual coupling proportional to $^{40}\mathrm{Ca}^+$1, so the summed angle is weakly coupled to external magnetic fields while remaining sensitive to perturbations that rotate both manifolds in the same direction (Gomez et al., 2019).

Single-species implementations use the same logic without requiring two distinct ensembles. In liquid acetonitrile-2-$^{40}\mathrm{Ca}^+$2C, two splittings,

$^{40}\mathrm{Ca}^+$3

are acquired simultaneously from one molecular sample, and the ratio

$^{40}\mathrm{Ca}^+$4

is largely insensitive to common-mode magnetic-field fluctuations (Wu et al., 2018). In dual-isotope xenon comagnetometry, the weighted combination

$^{40}\mathrm{Ca}^+$5

suppresses the common magnetic contribution while preserving sensitivity to additional spin-dependent interactions (Thrasher et al., 2019).

These examples illustrate a general point: simultaneous co-magnetometry is fundamentally a differential encoding strategy in which magnetic common mode is projected into a null or near-null channel, while the target perturbation is engineered to survive in the orthogonal or differential channel.

2. Experimental architectures

The experimental literature shows that simultaneity is compatible with a wide range of physical platforms. Some implementations use two different species in one apparatus; others use internal structure within one species or one material. What unifies them is co-location, common timing, and differential response.

Platform	Simultaneous channels	Demonstrated feature
$^{40}\mathrm{Ca}^+$6–$^{40}\mathrm{Ca}^+$7 two-ion crystal	Parallel Ramsey interferometers on Ca$^{40}\mathrm{Ca}^+$8 and Ti$^{40}\mathrm{Ca}^+$9	$g$0-level relative statistical uncertainty in Ti$g$1 $g$2 factors (Rehmert et al., 21 Aug 2025)
$g$3 spinor BEC	$g$4 and $g$5 hyperfine manifolds	$g$6 common-mode rejection (Gomez et al., 2019)
Liquid $g$7CH$g$8CN	Two nuclear-spin splittings from the same molecule	Gradient slope $g$9 (Wu et al., 2018)
Cs vapor	$$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$0 and $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$1 FID signals in one ensemble	Proton spin-gravity limit $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$2 (Yang et al., 2020)
$$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$3Xe–$$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$4Xe with embedded Rb	Continuous dual-isotope resonance tracking	White frequency-noise level $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$5 (Thrasher et al., 2019)
Eu$$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$6:YSO crystal	Four mirror-related sub-ensembles $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$7	Magnetic-field-induced shift cancellation better than $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$8 (Nima et al., 2024)
Pilot and signal deuteron bunches in a ring	Masked pilot bunch plus RF-driven signal bunch	Gating efficiency $$g_\mathrm{Ti}=g_\mathrm{Ca}\,\frac{\Delta E_\mathrm{Ti}}{\Delta E_\mathrm{Ca}}.$$9 (Slim et al., 2023)
Hybrid NV–Rb sensor	NV vector estimate plus Rb scalar estimate	Beyond $B$0 improvement in field-estimation error (Shalev et al., 21 Aug 2025)

Trapped-ion realizations emphasize exact temporal overlap and quantum-state control. The $B$1 logic ion provides a well-known reference $B$2 factor, sympathetic cooling, optical qubit readout, and logic readout of the $B$3 state in a linear segmented Paul trap (Rehmert et al., 21 Aug 2025). Cold-atom and vapor platforms instead exploit co-located ensembles and optical readout. The BEC device independently measures the $B$4 and $B$5 transverse magnetizations and azimuth angles by nondestructive Faraday rotation probing, whereas the Cs device extracts both hyperfine precession frequencies from the same optical FID record (Gomez et al., 2019, Yang et al., 2020).

Solid-state and beam-based platforms broaden the concept further. Mirror-symmetric Eu$B$6 ions in yttrium orthosilicate provide sub-ensembles with the same magnetic response but opposite sensitivity to possible T-violating new physics, and a storage-ring pilot bunch functions as a co-moving probe of the machine’s magnetic environment while the signal bunch undergoes RF-driven spin manipulation (Nima et al., 2024, Slim et al., 2023). Hybrid architectures such as NV centers paired with an Rb vapor cell use simultaneous vector and scalar sensing rather than a conventional cancellation ratio, but still implement co-magnetometric fusion of shared-field information (Shalev et al., 21 Aug 2025).

3. Interrogation protocols and observables

Simultaneous co-magnetometry is realized experimentally through diverse interrogation protocols, but each protocol creates paired observables from a shared evolution window. In the Ti$B$7 measurement, the sequence comprises ground-state cooling of the two-ion crystal, state preparation of $B$8 into a stretched Zeeman state in one of the four ground-state fine-structure manifolds $B$9, simultaneous rf Ramsey spectroscopy on both ions during a common dark time, fluorescence readout of Ca$^{87}\mathrm{Rb}$0, and quantum-logic readout of Ti$^{87}\mathrm{Rb}$1 by mapping Ti$^{87}\mathrm{Rb}$2 Zeeman information to shared motion and detecting that motion on Ca$^{87}\mathrm{Rb}$3 with a red-sideband rapid-adiabatic-passage pulse (Rehmert et al., 21 Aug 2025).

Other systems generate simultaneous channels spectroscopically rather than through parallel Ramsey interferometers. In continuous Xe comagnetometry, both isotopes’ NMR conditions are simultaneously satisfied by frequency modulation of the pulse repetition rate, the Xe precession is detected through the embedded Rb magnetometer, and the final comagnetometer variable is built from the two tracked resonance frequencies (Thrasher et al., 2019). In the pulsed dual-axis alkali-metal–noble-gas device, the measured optical-rotation transient is fit to

$^{87}\mathrm{Rb}$4

so that the fitted quadratures $^{87}\mathrm{Rb}$5 and $^{87}\mathrm{Rb}$6 encode orthogonal transverse responses simultaneously (Wang et al., 2024).

Simultaneous acquisition can also take the form of correlation readout. In covariance magnetometry with two NV centers, the two raw shot-by-shot outcomes are retained and the Pearson correlation

$^{87}\mathrm{Rb}$7

is computed, allowing common-mode magnetic fluctuations shared by both sensors to be distinguished from local, unshared noise (Rovny et al., 2022). In hybrid NV–Rb sensing, the outputs are fused geometrically: the NV subsystem provides a vector estimate, the Rb subsystem provides a scalar magnitude estimate, and a correction vector is chosen so that the corrected NV estimate has the same norm as the Rb measurement (Shalev et al., 21 Aug 2025).

The storage-ring pilot-bunch method uses yet another observable class. The principal measured quantity is the polarimeter asymmetry,

$^{87}\mathrm{Rb}$8

and the absence of oscillation at the spin-flip frequency in the masked pilot bunch verifies that it monitors the ring field without receiving the RF spin-rotator drive (Slim et al., 2023).

4. Noise rejection and systematic-error structure

The principal advantage of simultaneity is rejection of temporal magnetic noise that cannot be removed by interleaving alone. In the Ti$^{87}\mathrm{Rb}$9 experiment, earlier co-magnetometry methods are explicitly contrasted with the simultaneous scheme: interleaved measurements help with slow drifts but do not fully remove fast temporal fluctuations, whereas parallel Ramsey interferometers sample essentially the same instantaneous $f=1$0, so line-noise-driven magnetic fields at 50 Hz and harmonics largely cancel in the ratio. The result is that the $f=1$1-factor data recover the expected white-noise scaling more cleanly than the raw frequency data, and the dominant residual systematics become the static magnetic field gradient across the two-ion crystal and the ac-Zeeman shift caused by the trap drive (Rehmert et al., 21 Aug 2025).

Gradient systematics are a recurring theme in conventional dual-species comagnetometers, and several simultaneous single-species designs target them directly. In the liquid molecular device, the fitted slope of the single-species ratio with vertical gradient is

$f=1$2

whereas dual-species reference ratios show slopes $f=1$3 and $f=1$4, demonstrating suppression of first-order magnetic-field-gradient dependence by more than an order of magnitude (Wu et al., 2018). In the Cs FID comagnetometer, magnetic-field gradients and laser fields are highly suppressed because both signals come from the same atoms and the pump is blocked during readout, but asynchronous optical pumping and drift of residual magnetic field in the shield dominate the uncertainty budget (Yang et al., 2020).

Continuous and pulsed alkali–noble-gas systems reveal different systematic structures. The transversely polarized Xe comagnetometer emphasizes that the embedded Rb magnetometer phase shift

$f=1$5

must be calibrated and corrected; otherwise the apparent comagnetometer signal contains a derivative term from the magnetometer itself. With correction and improved stabilization of $f=1$6, the field suppression factor improves from about 75 to about 1800 (Thrasher et al., 2019). In the pulsed dual-axis $f=1$7Rb–$f=1$8Ne device, pump light shift is eliminated during measurement, probe beam pointing fluctuations in the pump-probe plane are fundamentally eliminated, and signal response to pump beam pointing fluctuations is suppressed by noble-gas compensation (Wang et al., 2024).

Other platforms demonstrate common-mode rejection by symmetry rather than by ratios alone. The $f=1$9Rb spinor BEC measures $f=2$0 common-mode rejection, in good agreement with theory (Gomez et al., 2019). The Eu$f=2$1:YSO crystal yields

$f=2$2

consistent with zero and corresponding to magnetic-field-induced shift cancellation better than $f=2$3 (Nima et al., 2024). In covariance magnetometry, by contrast, the limiting factor is often readout rather than field matching: the measured Pearson correlation is reduced by the product of two readout-noise factors, so correlation sensing is especially readout-limited (Rovny et al., 2022).

5. Multiparameter extensions and adjacent formulations

Simultaneous co-magnetometry intersects directly with multiparameter quantum metrology, although not every simultaneous magnetic measurement is a conventional co-magnetometer. A central example is simultaneous estimation of the amplitude $f=2$4 and frequency $f=2$5 of an AC magnetic field with

$f=2$6

In the uncontrolled problem, the generators for $f=2$7 and $f=2$8 are parallel, the quantum Fisher information matrix is singular, and joint estimation is unattainable. A quantum-control protocol resolves this by engineering the dynamics so the effective generators become orthogonal, yielding asymptotically diagonal QFIM,

$f=2$9

and restoring the optimal scalings $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$0 and $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$1 simultaneously (Isogawa et al., 19 Feb 2026).

A complementary line of work considers a thermal spin system used to estimate magnetic-field intensity and orientation jointly. With Hamiltonian

$$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$2

the ideal simultaneous strategy yields

$$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$3

but this advantage can be degraded or destroyed by coarsened measurement references for random rotations about the $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$4- or $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$5-axis. The $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$6-axis case is exceptional: the factor-of-two simultaneous-estimation advantage survives for all coarsening strengths considered (Xie et al., 2018).

Several works are explicitly described as co-magnetometry-like rather than standard co-magnetometry. Ancilla-assisted NV DC magnetometry uses a nuclear spin to upconvert a transverse DC field into an effective AC signal, allowing quantum lock-in detection up to the $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$7 limit rather than the Ramsey-limited $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$8 limit. This is in the co-magnetometry family in spirit, but the ancilla is not a second independent field sensor (Ajoy et al., 2016). Likewise, isotropic CPT magnetometry with simultaneous excitation of orientation and alignment resonances is an internal dual-resonance protocol rather than a two-species comagnetometer; its significance lies in eliminating dead zones and suppressing heading errors by driving complementary tensor channels in the same atomic ensemble (Ben-Kish et al., 2010).

These adjacent formulations show that simultaneity has two distinct roles in the broader field: common-mode rejection in true comagnetometers, and compatibility-enabling joint encoding in multiparameter sensing.

6. Scientific uses, limitations, and design implications

Simultaneous co-magnetometry is used where magnetic backgrounds or magnetic drifts would otherwise dominate a weak target signal. The applications identified in the literature include precision measurements of Landé $$\gamma^{(1)}=-\gamma_0-\gamma_s,\qquad \gamma^{(2)}=+\gamma_0-\gamma_s,$$9 factors and tests of CI+MBPT atomic theory in ions that cannot be directly laser cooled (Rehmert et al., 21 Aug 2025); searches for new physics on sub-millimeter length scales, precision studies of ultracold collision physics, and angle-resolved studies of quantum spin dynamics in spinor condensates (Gomez et al., 2019); measurements of hypothetical spin-dependent gravitational energy at the $\theta^{(12)}\equiv \theta^{(1)}+\theta^{(2)}.$0 level in single-species liquid-state NMR (Wu et al., 2018); constraints on proton spin-gravity coupling at the $\theta^{(12)}\equiv \theta^{(1)}+\theta^{(2)}.$1 scale in Cs vapor (Yang et al., 2020); tests of time-reversal violation in solid-state rare-earth systems (Nima et al., 2024); charged-particle EDM searches in storage rings (Slim et al., 2023); inertial rotation sensing and anomalous spin-coupling searches in alkali–noble-gas cells (Wang et al., 2024, Hedges et al., 2024); and spatial field mapping, gradiometric sensing, and multimodal quantum sensing in hybrid NV–Rb devices (Shalev et al., 21 Aug 2025). Covariance-based simultaneous sensing extends the application space to spatiotemporal structure factors and field-correlation dynamics (Rovny et al., 2022).

The limitations are platform-dependent. Trapped-ion implementations remain sensitive to static gradients and ac-Zeeman shifts (Rehmert et al., 21 Aug 2025). Continuous noble-gas systems require phase-shift calibration of the embedded magnetometer (Thrasher et al., 2019). Single-species vapor devices remain limited by asynchronous optical pumping and residual field drift (Yang et al., 2020). Hybrid vector–scalar fusion requires calibration of the background field $\theta^{(12)}\equiv \theta^{(1)}+\theta^{(2)}.$2 in unshielded environments (Shalev et al., 21 Aug 2025). Mirror-symmetric crystal devices are proof-of-principle demonstrations rather than full T-violation measurements (Nima et al., 2024). Pulsed dual-axis and perturbed alkali–noble-gas schemes are limited in current implementations by cell-fabrication, beam-collimation, and related imperfections rather than by the underlying concept (Wang et al., 2024, Hedges et al., 2024).

A common misconception is that co-magnetometry necessarily requires two different species. The contemporary literature does not support that restriction. Simultaneous co-magnetometers have been realized with different species, with two hyperfine manifolds of one ensemble, with multiple nuclear-spin transitions in identical molecules, with mirror-related sub-ensembles in one crystal, and with two spatially separated NV centers analyzed through covariance. This suggests that the decisive design criterion is not species multiplicity but whether the compared channels share the same field realization closely enough in space and time, while remaining differentially sensitive to the nuisance magnetic background and to the signal of interest.