Pith. sign in

REVIEW 2 major objections 8 minor 24 references

Influence of a magnetic field on the frequency of a laser stabilized to molecular iodine

T0 review · 2 major / 8 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The frequency of an iodine-stabilized laser shifts linearly with applied magnetic field at (1062 ± 6) × 10^4 Hz/T, a measured Zeeman coefficient that forces magnetic shielding for 10^-15 stability.

desk verdict First weak-field Zeeman coefficient for the iodine R(34) 44-0 line, with a clean measurement but an unquantified cell tilt that leaves the stated uncertainty incomplete. read the letter →

arxiv 2506.00342 v1 pith:C3TAQBZM submitted 2025-05-31 physics.atom-ph physics.optics

Jonathan Gillot , Joannes Barbarat , Charles Philippe , Hector Alvarez-Martinez , Rodolphe Letargat , Ouali Acef This is my paper
classification physics.atom-phphysics.optics PACS 32.60.+i42.62.Eh
keywords Zeemaneffectmoleculariodinefrequencystabilizationlaserstabilitymagneticshieldinghyperfinetransitionmodulationtransferspectroscopyopticalreference
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports the first measured weak-field Zeeman coefficient for an iodine-stabilized laser transition. By applying a longitudinal magnetic field to the iodine cell, the authors find the locked laser frequency shifts linearly with field at a rate of $(1062 \pm 6) \times 10^4$ Hz/T at 514 nm for the $a_1$ hyperfine component of the R(34) 44-0 line. The practical consequence is that ambient magnetic-field fluctuations of a few tenths of a microtesla—such as those from nearby metro and train lines—push the frequency stability into the $10^{-14}$ range, and reaching the low $10^{-15}$ domain requires magnetic shielding. The paper also gives a preliminary transverse-field coefficient and shows that the shift grows with quantum numbers $J$ and $v$ for other iodine lines. If correct, this removes an often-neglected environmental term from the uncertainty budget of iodine frequency references, including space-mission backup lasers.

What carries the argument

The central object is the magnetically perturbed saturated-absorption line shape of molecular iodine, detected by modulation transfer spectroscopy on the first derivative of the $a_1$ hyperfine component. The magnetic field, generated by a calibrated solenoid wrapped around the cell and variable up to $70 \times 10^{-4}$ T, splits and shifts the hyperfine sub-levels; the zero crossing of the derivative line shape, which is the lock point of the laser, moves with the field. The measurement chain—laser frequency tripling, a mu-metal shield with attenuation 200, and a beat note against an ultra-stable cavity—converts that line-shape shift into a directly measurable frequency versus field slope.

What would settle it

Repeat the slope measurement with the cell axis precisely aligned with the solenoid axis, or with the tilt deliberately varied; if the slope moves outside the quoted $(354 \pm 2) \times 10^4$ Hz/T at 1542 nm, the reported coefficient is biased by the uncorrected angle.

Watch

Extended reading notes

Core claim

The central claim is that the frequency of a laser locked to the $a_1$ hyperfine component of the R(34) 44-0 transition of molecular iodine at 514.017 nm responds linearly to an applied magnetic field along the cell, with a measured slope of $(1062 \pm 6) \times 10^4$ Hz/T at 514 nm (a relative shift of $(1.82 \pm 0.01) \times 10^{-8}$ per tesla). This was established by locking a 1542 nm laser, frequency-tripled to 514 nm, to the first derivative of the line and recording the beat note against an ultra-stable cavity while stepping the solenoid current between 0 and 3 A. The beat frequency changed synchronously with the field, showed no drift artifacts after correcting for the cavity's small drift, and the fit over $\pm 3 \times 10^{-4}$ T had a correlation coefficient near 0.99. At stronger fields the line shape broadens and eventually splits, and other hyperfine lines ($a_1$:R(72) 46-0, $a_1$:P(90) 55-0, $a_1$:R(105) 50-0) show larger slopes, consistent with the Zeeman effect scaling with the quantum numbers. A preliminary measurement for a transverse field gives $(180 \pm 75) \times 10^4$ Hz/T at 514 nm. The authors conclude that weak magnetic fields are far from negligible for iodine-stabilized references and that shielding is mandatory to reach the lower $10^{-15}$ frequency-stability regime.

Load-bearing premise

The iodine cell is slightly tilted to avoid optical feedback, so the applied 'longitudinal' field is not exactly parallel to the laser beams, and the paper does not correct for this angle in reporting the Zeeman coefficient.

Editorial extensions

If this is right

  • Ambient horizontal field fluctuations of about $2 \times 10^{-7}$ T limit the frequency stability to roughly $3.6 \times 10^{-15}$, so a magnetic shield becomes a requirement for operation at the $10^{-15}$ level.
  • Vertical field fluctuations of about $3 \times 10^{-6}$ T, the largest local component, translate through the transverse-field coefficient to a stability limit near $9 \times 10^{-15}$.
  • For LISA-like requirements of 30 Hz/$\sqrt{\mathrm{Hz}}$, uncontrolled field fluctuations of order $10^{-5}$ T can push the frequency noise below requirements, motivating the twin shielding with attenuation $\geq 1000$ that the authors designed.
  • Because the Zeeman coefficient depends on the chosen transition, increasing with $J$ in the strong-field regime, transition selection offers a way to reduce magnetic sensitivity when picking an iodine reference line.
  • The measured coefficient gives a quantitative relation between residual magnetic field and frequency offset, so shield specifications can be set directly from a desired stability target.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable extension would be to measure the same slope with the cell axis accurately parallel to the solenoid field (no tilt): if the corrected slope differs by more than the stated $6 \times 10^4$ Hz/T uncertainty, the uncorrected tilt angle is biasing the reported longitudinal coefficient.
  • Because the transverse coefficient is smaller than the longitudinal one, the tilted-cell geometry would bias the measured slope slightly downward; determining the actual tilt angle would place an upper bound on this bias.
  • If the linear Zeeman behavior holds at even weaker fields, the same setup could serve as a sensitive magnetometer: a 1 mG ($10^{-7}$ T) field produces about 1 Hz shift at 514 nm, making the locked laser frequency a direct readout of ambient field.
  • The strong-field measurements of four transitions with different $J$ and $v$ offer a dataset for testing quantum models of the iodine Zeeman structure; a calculated coefficient matching these slopes would allow predicting magnetic sensitivity of other lines without measurement.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 8 minor

Summary. This brief report measures the Zeeman shift of the a1 hyperfine component of the R(34) 44-0 transition of molecular iodine at 514.017 nm, using a 1542-nm laser frequency-tripled to 514 nm and locked to the line via modulation-transfer spectroscopy. For a solenoid field applied nominally along the laser beams, the locked-laser beat-note frequency varies linearly with field: slope (354±2)×10^4 Hz/T at 1542 nm, corresponding to (1062±6)×10^4 Hz/T at 514 nm over ±3×10^-4 T (R²≈0.99). A preliminary transverse-field coefficient of (180±75)×10^4 Hz/T is also reported. Combining these slopes with measured ambient field fluctuations (≈0.2 μT horizontal, ≈3 μT vertical) yields estimated relative frequency instabilities of ≈3.6×10^-15 and ≈9.3×10^-15, respectively, which the authors use to explain the difference between shielded and unshielded Allan-deviation floors. The central result is an empirical coefficient whose error budget excludes the explicitly acknowledged cell-tilt angle and the solenoid calibration accuracy.

Significance. The paper supplies a genuinely missing data point: a weak-field (sub-mT) Zeeman coefficient for the a1 R(34) 44-0 line at 514 nm that is directly relevant to iodine-stabilized laser references, including the proposed LISA backup source. Several features of the experiment deserve explicit credit: the coefficient is obtained as a direct lock-point shift, not from a fitted model; the linear fit is documented with R²≈0.99 over ±3×10^-4 T; cavity drift is rejected by deliberate zero-field re-calibration and by an independent beat note against an ultra-stable cavity laser; and the shielded versus unshielded stability comparison at the few-10^-15 level supports the interpretation that ambient field fluctuations limit the system. The stability estimates follow by simple multiplication of the measured slopes by independently recorded local field fluctuations, with no free parameters. The result is metrological rather than fundamental, and its significance is moderate: it converts a device-level observation (worse long-term stability near metro lines) into a quantitative, transferable sensitivity coefficient.

major comments (2)
  1. [Section 3 (tilt statement before §3.1); §3.1 and Fig. 6] The unquantified cell tilt is load-bearing for the headline coefficient C_L=(1062±6)×10^4 Hz/T. Section 3 states that the room-temperature cell is 'very slightly tilted' so that 'a non-zero angle between the direction of BL and the beams appears, but this small angle is not considered in this study.' Under the paper's own operational definitions, the measured slope per unit solenoid field is C_L cosθ + C_T sinθ, with C_T=(180±75)×10^4 Hz/T from Section 3.2, so the tilt contributes at first order in θ. A 2° tilt changes the slope by about 6×10^4 Hz/T, equal to the full quoted uncertainty, and a 5° tilt by about 16×10^4 Hz/T; because C_T is nonzero at the 2.4σ level in the paper's own data, this is a genuine mixing effect rather than a hypothetical one. The authors should measure or bound the tilt angle, propagate the resulting systematic uncertainty, and state the solenoid calibration accuracy, which is absent from the current ±6×10^4 Hz/T budget.
  2. [Section 3.2] The transverse coefficient is under-specified and is used in ways that outrun its precision. The text reports only that 'the frequency beat note fbn exhibits a drift of (60±25)×10^4 Hz/T at 1542 nm' over a maximum transverse field of 1.11×10^-4 T, which corresponds to a total shift of only about 7 Hz at 1542 nm, i.e., of order the system's short-term Allan deviation; no statement is given of the number of field steps, the use of polarity reversal, or the fit procedure. This preliminary coefficient nevertheless feeds the vertical-field stability estimate of 9.26×10^-15 in Sections 3.2–3.3, and it is also the C_T needed to correct or bound the tilt contribution in the longitudinal measurement. The authors should describe the measurement in detail, justify or reduce the 42% uncertainty, and label the affected stability numbers and any tilt correction as preliminary.
minor comments (8)
  1. [Section 2 vs Section 3.2] Section 2 gives the transverse coil constant as BT=1×10^-5 T/A 'with values adjustable up to 3×10^-4 T,' while Section 3.2 states that the maximum transverse flux density with 10 A is 1.11×10^-4 T; these statements are mutually inconsistent, and the winding/turns description should be reconciled.
  2. [Table 2] The offsets listed for the three additional transitions (about −15 to +14 ×10^4 Hz/T, i.e., tens of kHz at 514 nm) are stated alongside a claim that the frequency zero is chosen at BL=0; the zeroing procedure for each dataset and the possible physical origin of these offsets should be clarified.
  3. [Abstract] The abstract's statement that uncontrolled fields of order 1×10^-4 T limit the stability to the upper 10^-14 domain conflates the static shift (1.8×10^-12 relative) with an instability, which is determined by field fluctuations; the assumed fluctuation amplitude should be stated.
  4. [Section 3.1, Fig. 6] The slope uncertainty (±2×10^4 Hz/T at 1542 nm, scaled to ±6×10^4 Hz/T) is reported without the number of fitted points, the fit weighting, or a characterization of the noise on the 5-min plateaus, and Fig. 6 does not show the fit residuals; adding the residuals and fit metadata would allow R²≈0.99 to be verified.
  5. [Sections 2 and 3.1] The measured coefficient is the shift of the locked zero-crossing for a particular modulation depth, pump/probe power balance, and beam geometry; the sensitivity of the slope to these line-shape parameters is not discussed, which limits the transferability of the quoted value to other iodine-stabilized systems.
  6. [Section 2] The solenoid calibration is described only by the sensor model (Bartington Mag-03); the sensor accuracy, the calibration position, and the assumed field homogeneity over the 40-cm cell should be stated, since a scale error in the 7.1×10^-4 T/A constant enters every reported slope.
  7. [Sections 3.1–3.3] The coefficient is measured on the uncooled room-temperature cell (iodine pressure <2 Pa) but is used to interpret stability data from the cooled cell at 1 Pa; the possible dependence of the effective zero-crossing slope on pressure is not addressed.
  8. [References] Several references are incomplete (e.g., [5], [8], and [22] lack pagination or article numbers, and [1] is a thesis without a title); the reference list should be completed for the archival record.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the Zeeman coefficient is a direct measurement, and the stability-impact estimate is an arithmetic propagation of measured quantities.

full rationale

The paper's central result, the longitudinal Zeeman coefficient (1062 ± 6) × 10^4 Hz/T at 514 nm, is obtained by a linear fit of measured beat-note frequency versus applied solenoid current, with the solenoid independently calibrated using a Bartington Mag-03 sensor. There is no derivation chain in which an output is defined in terms of an input or in which a fitted parameter is renamed as a prediction. The beat-note slope at 1542 nm (354 ± 2) × 10^4 Hz/T is converted to the 514 nm transition frequency by the stated factor of three, which is arithmetic, not circular. The subsequent frequency-stability estimates are obtained by multiplying the measured slope by measured local magnetic-field fluctuations, again without hidden fitted parameters. Self-citations are present (references [10], [15], [16], [17]) but they support experimental techniques and prior stability results, not the measured Zeeman coefficient; the Zeeman measurement is self-contained against an external reference beat note and a calibrated coil. The acknowledged unquantified cell tilt is a genuine experimental systematic uncertainty that could bias the slope, but that is a correctness risk, not circularity. No circular step could be identified under the paper's own equations or citations.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central claim is a measured quantity, so the ledger contains no free parameters used to construct the claim. The linear fit slope and offset are the result, not inputs. The assumptions are the standard experimental calibrations (uniform calibrated solenoid field, exact frequency tripling factor of 3, linearity of the Zeeman shift in the fitted range) plus unquantified choices: the cell tilt angle is ignored, and the coefficient measured on the room-temperature cell is applied to the cooled cell in the stability estimate.

assumptions (4)
  • domain assumption The frequency tripling factor between the 1542 nm laser and the 514 nm iodine transition is exactly 3.
    Used in Section 3.1 to convert the beat note slope at 1542 nm to the transition shift at 514 nm.
  • domain assumption The solenoid produces a uniform, calibrated longitudinal magnetic flux density BL = 7.1e-4 T/A along the beam path.
    Stated in Section 2; the calibration uses a Bartington Mag-03 sensor, but no calibration uncertainty is given.
  • domain assumption The Zeeman frequency shift is linear with BL in the ±3e-4 T range.
    Supported by R2 = 0.99 for the fit in Fig. 6, but only for the limited range studied.
  • ad hoc to paper The small tilt of the cell does not affect the measured longitudinal Zeeman coefficient.
    Section 3 states the cell is tilted to avoid optical feedback and that the angle is not considered, without estimating its impact.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Influence of a magnetic field on the frequency of a laser stabilized to molecular iodine." pith.science (2026). https://pith.science/paper/C3TAQBZM

@misc{pith2026250600342,
  author       = {Pith},
  title        = {Pith review of: Influence of a magnetic field on the frequency of a laser stabilized to molecular iodine},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C3TAQBZM}},
  note         = {Machine review of arXiv:2506.00342}
}
abstract

We report on the effect of a weak magnetic field applied on an iodine cell used to frequency stabilize a laser. A 1.5$~\mu$m laser is frequency tripled in order to excite the molecular transitions at 0.51$~\mu$m and frequency locked on a hyperfine line. With this frequency reference, we report short-term stability about $3\,\times\,10^{-14}~\tau^{-1/2}$, with a minimum value of $4\,\times\,10^{-15}$ at 200~s. The lower part of $10^{-15}$ frequency stability domain is reached, in our case, only by adding an efficient magnetic shield around the sealed quartz iodine cell. In order to quantify the Zeeman effect, we applied magnetic fields of several $\,\times\,10^{-4}$~T on the cell containing the iodine vapour. The Zeeman effect affects the lineshape transition in such a way that we observe a modification of the laser frequency. We have measured this linear Zeeman shift at $(1062\pm6)\,\times\,10^{4}$~Hz/T for the $a1$ hyperfine component of the R(34)~44-0 transition, near 514~nm by applying a magnetic field along the cell. Thus, in case of uncontrolled magnetic fields of an order of magnitude of 1$\,\times\,10^{-4}$~T, the frequency stability is limited in the upper of the $10^{-14}$ domain.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

  1. [1]

    Argence, vol

    B. Argence, vol. 7 Paris (2010)

    work page 2010

  2. [2]

    Barbarat, J

    J. Barbarat, J. Gillot, H. Alvarez-Martinez, M. Lours, D. Hol - leville, R.L. Targat, P.-E. Pottie, P. Wolf, P. Tuckey, O. Acef, F.-X. Esnault, T. Leveque, in ICSO 2018, vol 11180. SPIE (2019)

    work page 2019

  3. [3]

    Broyer, thesis, Université Pierre et Marie Curie - Paris VI (1977)

    M. Broyer, thesis, Université Pierre et Marie Curie - Paris VI (1977)

    work page 1977

  4. [4]

    Chen, thesis, University of Colorado (2005)

    L. Chen, thesis, University of Colorado (2005)

    work page 2005

  5. [5]

    Danzmann, the LISA study team, C lass

    K. Danzmann, the LISA study team, C lass. Quantum Gravity 13(11A) (1996)

    work page 1996

  6. [6]

    Gillot, S

    J. Gillot, S. Falzon Tetsing-Talla, S. Denis, G. Goavec-Merou, J. Millo, C. Lacroûte, Y. Kersalé, Opt. Express 30(20) (2022)

    work page 2022

  7. [7]

    Goncharov, S.V

    A.N. Goncharov, S.V. Gateva-Kosteva, M.N. Skvortsov, V.P. Che- botayev, Appl. Phys. B 52(4), 311–314 (1991)

    work page 1991

  8. [8]

    Goncharov, A.Y

    A.N. Goncharov, A.Y. Nevsky, M.N. Skvortsov. Appl. Phys. B 62(4) (1996)

    work page 1996

Show all 24 references
  1. [9]

    Gürlebeck, L

    N. Gürlebeck, L. Wörner, T. Schuldt, K. Döringshoff, K. Gaul, D. Gerardi, A. Grenzebach, N. Jha, E. Kovalchuk, A. Resch, T. Wendrich, R. Berger, S. Herrmann, U. Johann, M. Krutzik, A. Peters, E. M. Rasel, C. Braxmaier, Phys. Rev. D 97(12) (2018)

    2018

  2. [10]

    Hrabina, M

    J. Hrabina, M. Zucco, C. Philippe, T.M. Pham, M. Holá, O. Acef, J. Lazar, O. číp, Sensors (Basel) 17(1) (2017)

    2017

  3. [11]

    Jones, W.-Y

    R.J. Jones, W.-Y. Cheng, K.W. Holman, L. Chen, J.L. Hall, J. Ye, Appl. Phys. B 74(6), 597–601 (2002)

    2002

  4. [12]

    Kawamura, T

    S. Kawamura, T. Nakamura, M. Ando, N. Seto, T. Akutsu, I. Funaki, K. Ioka, N. Kanda, I. Kawano, M. Musha, K. Naka- zawa, S. Sato, T. Takashima, T. Tanaka, K. Tsubono, J. Yokoy - ama, K. Agatsuma, K.-s. Aoyanagi, K. Arai, A. Araya, N. Aritomi, H. Asada, Y. Aso, D. Chen, T. Chib...

    2018

  5. [13]

    Kokuyama, K

    W. Kokuyama, K. Numata, J. Camp. Appl. Opt. AO 49(32) (2010)

    2010

  6. [14]

    Mehlman, D

    A. Mehlman, D. Holleville, M. Lours, B. Pointard, R. L. Targat, P. Wolf, S. Bize, O. Acef, A. Boutin, K. Lepage, L. Fulop, H. Hal- loin, T. Zerguerras, J. Hrabina, A. Hola, L. Ouda, and R. Arguel, in ICSO 2022, vol 12777. SPIE (2023)

    2023

  7. [15]

    Philippe

    C. Philippe. thesis, Paris Sciences et Lettres (2017)

    2017

  8. [16]

    Philippe, E

    C. Philippe, E. Chea, Y. Nishida, F. du Burck, O. Acef. Appl. Phys. B 122(10) (2016)

    2016

  9. [17]

    Philippe, R

    C. Philippe, R. Le Targat, D. Holleville, M. Lours, T. Minh-Pham, J. Hrabina, F. Du Burck, P. Wolf, O. Acef, in 2016 EFTF. IEEE (2016)

    2016

  10. [18]

    Schkolnik, K

    V. Schkolnik, K. Döringshoff, F. Gutsch, C. Kürbis, M. Oswald, B. Probster, E. V. Kovalchuk, A. Bawamia, R. Smol, T. Schuldt, M. Lezius, R. Holzwarth, A. Wicht, C. Braxmaier, M. Krutzik, A. Peters, Phys. Rev. Appl. 11 (2019)

    2019

  11. [19]

    Schkolnik, K

    V. Schkolnik, K. Döringshoff, F. B. Gutsch, M. Oswald, T. Schuldt, C. Braxmaier, M. Lezius, R. Holzwarth, C. Kürbis, A. Bawamia, M. Krutzik, and A. Peters, EPJ Quant. Tech. 4 (1) (2017)

    2017

  12. [20]

    J.H. Shirley. Opt. Lett. OL 7(11), Nov. (1982)

    1982

  13. [21]

    Tricot, D.H

    F. Tricot, D.H. Phung, M. Lours, S. Guérandel, E. de Clercq, Rev. Sci. Instrum. 89(11) (2018)

    2018

  14. [22]

    Vigué, thesis, Université Pierre et Marie Curie, Paris VI (1978)

    J. Vigué, thesis, Université Pierre et Marie Curie, Paris VI (1978)

    1978

  15. [23]

    J. Ye, L. Robertsson, S. Picard, L.-S. Ma, J.L. Hall. IEEE 48(2) (1999)

    1999

  16. [24]

    Swierad, S

    D. Swierad, S. Häfner, S. Vogt, B. Venon, D. Holleville, S. Bize, A. Kulosa, S. Bode, Y. Singh, K. Bongs, E.M. Rasel, J. Lodewyck, R. Le Targat, C. Lisdat, U. Sterr. Sci .Rep. 6(1) (2016) Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in ...

    2016

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.