Film Thickness Gauge Based on Interferometric Principle of Y-shaped Optical Fiber

Haoqi Wang; Jianguo Cao; Juntao He; Peilin Gao; Ruiyun Ma; Shaohua Wang; Yikai Dang; Yingke Li; Yingyuan Li; Yong Pan

REVIEW 4 major objections 5 minor 15 references

Film Thickness Gauge Based on Interferometric Principle of Y-shaped Optical Fiber

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A film-thickness gauge using a Y-shaped optical fiber and adjacent interference peaks reports 5% average error and 0.3% relative uncertainty on polymer films.

desk verdict A sound but overclaimed instrument paper; the accuracy numbers don't survive contact with its own table. read the letter →

arxiv 2412.07528 v1 pith:PYI4T3GC submitted 2024-12-10 physics.optics

Juntao He , Yikai Dang , Haoqi Wang , Shaohua Wang , Yingke Li , Ruiyun Ma , Yingyuan Li , Peilin Gao

show 2 more authors

Jianguo Cao Yong Pan

This is my paper · ORCID

classification physics.optics

keywords filmthicknessgaugeY-shapedopticalfiberthin-filminterferencewhite-lightinterferometrygratingspectrometerpolyimidePVCnon-contactmeasurement

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

The paper proposes a compact, non-contact film thickness gauge built from a Y-shaped optical fiber, a white LED, and a grating spectrometer, and claims it measures polymer film thickness from the wavelength separation of adjacent interference peaks. The central claim is that this simple arrangement achieves an average total error of about 5% and a relative uncertainty of 0.3% on polyimide and PVC films, with the best accuracy (roughly 1% error) on 0.1 mm films. A sympathetic reader should see this as an attempt to show that a low-cost, portable interferometric instrument can stand in for bulkier laboratory thickness measurement methods. The paper reports repeated measurements of nominally 0.025, 0.05, and 0.1 mm films and derives thicknesses using $d=\lambda_1\lambda_2/(2n(\lambda_1-\lambda_2))$.

What carries the argument

The load-bearing object is the Y-shaped fiber probe, a 6+1 fiber bundle in which six cores deliver light from the LED to the film surface and a single core returns the reflected light to the spectrometer; this geometry makes illumination and detection coaxial without touching the sample. The identity that carries the computation is $d=\lambda_1\lambda_2/(2n(\lambda_1-\lambda_2))$, which follows from writing the optical path difference as $2nd$ and requiring the two adjacent reflection maxima to satisfy $2nd/\lambda_1=j$ and $2nd/\lambda_2=j+1$. Its role is to convert a raw spectral measurement, the separation between neighboring peaks, into a thickness value using only the film refractive index $n$ as external input.

What would settle it

Measure the refractive index of a nominally 0.05 mm PI film at the two adjacent peak wavelengths with an independent method such as spectroscopic ellipsometry, insert that $n$ into $d=\lambda_1\lambda_2/(2n(\lambda_1-\lambda_2))$, and check whether the recomputed thickness still falls in the reported 54.3 to 58.9 $\mu$m range; if the true $n$ differs from the value the authors used by more than a few percent, the recomputed thickness leaves the claimed 5% error band.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a standard thin-film interference formula, applied to two adjacent peak wavelengths in the reflected spectrum collected through a Y-shaped fiber probe, yields thickness values that track nominal film thickness across repeated trials. The system illuminates the film through the six-core branch of a 6+1 Y-fiber, collects the reflected signal through the single-core branch, and a CCD grating spectrometer records the spectrum. Selecting two neighboring maxima gives $\lambda_1$ and $\lambda_2$, and with the film refractive index $n$ the thickness follows from the adjacent-order conditions $2nd/\lambda_1=j$ and $2nd/\lambda_2=j+1$. The reported measurements include mean values such as 99.6 $\mu$m for the 0.1 mm PI film and 27.7 $\mu$m for one 0.025 mm PI film, with a combined standard uncertainty of 0.3 $\mu$m; the authors state the average total error is 5% and the thickest film is measured within about 1%.

Load-bearing premise

The load-bearing premise is that the refractive index $n$ of each film is known accurately at the two peak wavelengths; the paper does not report the values used, and because the computed thickness scales linearly with $1/n$, any error in $n$ becomes a comparable percentage error in the reported thickness.

Editorial extensions

If this is right

If the claimed accuracy holds, transparent film thickness in the tens-to-hundreds of micrometers can be checked with a hand-portable, non-contact probe rather than a full laboratory ellipsometer or profilometer.
Because only two adjacent interference peaks are needed, a broadband LED and a modest-resolution CCD spectrometer suffice, keeping the system cost and size low.
The same measurement principle should transfer to other transparent films whose refractive index is known or measured separately.
The reported 0.3% relative uncertainty suggests the device is stable enough for repeated quality-control checks, even if its absolute accuracy is governed by the refractive-index input.

Reading between the lines

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

The refractive index $n$ used for the PI and PVC films is never reported; since the computed thickness is proportional to $1/n$, the claimed 5% accuracy can only be reproduced if the values used were within a few percent of the true film values at the peak wavelengths.
The 0.3% relative uncertainty is a repeatability-style uncertainty, not a total uncertainty; systematic effects such as peak-picking bias, fiber angle, and refractive-index error are separate from it and are reflected in the looser average-total-error claim.
A natural extension is to invert the same broadband spectrum for both thickness and refractive index simultaneously using several peak pairs, removing the need for a pre-supplied $n$.
The 6+1 Y-fiber geometry could be combined with a scanning stage to map thickness uniformity across a film rather than measuring a single spot.

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.

Desk Editor's Note

read the letter

Short version: the gauge works, but the numbers don't add up. The core idea is fine—a white-light interference thickness gauge built from a commercial Y-fiber reflection probe, a white LED, and a CCD spectrometer. The physics is textbook thin-film interference, and the device is cheap, portable, and non-contact. That's a genuinely useful practical combination for measuring 25–100 µm polymer films, even if the novelty ends there.

The problems are quantitative and concentrated in the error claims. The paper says the maximum error is no more than 15%, the absolute error is under 4 µm, and the average total error is 5%. Table 1 tells a different story. The first 0.05PI film reads 58.9 µm against a 50 µm nominal value (+17.8%, 8.9 µm error), and the first 0.025PI film reads 29.2 µm against 25 µm (+16.8%, 4.2 µm error). The mean absolute percentage error across the PI measurements is about 7%, not 5%. The relative uncertainty of 0.3% is repeatability only, not accuracy, and the standard deviations in Table 2 don't match the raw data—they look like standard errors of the mean.

On top of that, the refractive index n used in Eq. (1-11) is never reported. Thickness scales linearly with 1/n, so without n the measurement is unreproducible. The spectrometer resolution is also unspecified; adjacent peaks for 100 µm films are roughly 1 nm apart, so that matters. Peak selection is manual and done on "continuous and stable bands," which is an unquantified operator choice.

What the paper does well is describe the hardware clearly and show that measured thicknesses correlate with nominal values. The instrument is simple and plausibly useful for quality control. But as written, the central accuracy claims are contradicted by the paper's own data, and the missing n alone would justify rejection.

It's not an original contribution to physics, but it's a real engineering effort with checkable flaws. I would not cite it, and I wouldn't bring it to reading group as a model of analysis—maybe as a cautionary tale. That said, the errors are correctable, so a serious referee could turn this into a decent practical paper. Send it to review, but with the message that the accuracy claims and uncertainty budget need a major overhaul before it's publishable.

Referee Report

4 major / 5 minor

Summary. The manuscript reports a compact film-thickness gauge built on a Y-shaped optical-fiber reflection probe and a grating spectrometer. It derives the familiar adjacent-peak formula d = λ1λ2/[2n(λ1−λ2)] for transmission through a film of refractive index n and thickness d, describes the hardware design and assembly, and presents six repeated measurements on PI films (nominal 0.1, 0.05, 0.025 mm) and one PVC film. The central claims are an average total error of 5%, a maximum error not over 15%, an absolute error below 4 µm, and a relative uncertainty of 0.3%.

Significance. If fully validated, the instrument would offer a simple, portable, and low-cost method for measuring film thickness without contact, which could be useful in applied settings. The underlying interference relation is standard and correctly derived under the stated ideal conditions. However, the experimental validation as presented is incomplete: the refractive index used for the PI and PVC films is never stated, the spectrometer resolution is not quantified, and the accuracy claims are contradicted by the paper's own Table 1 in specific entries. These gaps prevent the reproducibility of the numeric results and undermine the headline accuracy statements as written.

major comments (4)

[Table 1, §4.2, §5] The stated error bounds are contradicted by the data in Table 1. The first 0.05 PI measurement is 58.9 µm against the 50 µm nominal value, an error of +17.8% and 8.9 µm; the first 0.025 PI entry is 29.2 µm against 25 µm, an error of +16.8% and 4.2 µm. Both exceed the claims in §4.2 and §5 of an overall maximum error not more than 15% and an error of less than 4 microns. The claim of an average total error of 5% also requires a defined nominal reference for the PVC sample, which is not supplied, and is not a robust description of the Table 1 data once these outliers are included.
[Eq. (1-11) and §4] The refractive index n of the PI and PVC films is never reported. Because d is directly proportional to 1/n, any error or arbitrary choice of n translates into the same percentage error in thickness. Without the n values used to compute the entries in Table 1, the entire calculation is unreproducible and the claimed 5% total error and 0.3% relative uncertainty cannot be independently checked. This missing input is load-bearing for the accuracy claim.
[§3, spectrometer description] The manuscript states that the spectrometer has "high resolution" but never specifies its spectral resolution or pixel spacing. Using the paper's own formula, adjacent interference peaks for a 50 µm film at λ≈550 nm and n≈1.5 are only about 2 nm apart, and for a 100 µm film they are about 1 nm apart. Unless the spectrometer can resolve such separations, the extracted peak wavelengths and hence the derived thicknesses can be systematically biased. The resolution must be reported for the accuracy claims to be meaningful.
[§4.2, Table 2] The uncertainty calculation is not transparent and conflates precision with accuracy. The 0.3% relative uncertainty is essentially a repeatability statement based on the standard error of six readings, yet Table 2 lists standard deviations that range from 0.30 to 0.79 µm across samples; the derivation of the single u_D = 0.3 µm value is not explained, especially the role of u_B. Moreover, this repeatability measure does not capture the dominant systematic errors, such as the unknown n and the observed bias of, for example, the first 0.05 PI sample mean of 55.4 µm versus the 50 µm nominal value.

minor comments (5)

[Eq. (1-2)] The expression written as "2√I1/I2 cosΔφ" should be 2√(I1 I2) cosΔφ; the product under the square root is missing parentheses.
[§4.2] The text refers to "taking into formula 1-10" when computing thickness; the correct reference is Eq. (1-11).
[Conclusion] The conclusion states that samples are "0.1 μm, 0.05 μm or 0.025 μm," but the experimental samples are 0.1 mm, 0.05 mm, and 0.025 mm; the units are inconsistent.
[§4.1, Table 1] The "scalar thickness" terminology should be defined as nominal thickness, and the nominal reference value for the PVC film is left ambiguous ("approximately 0.025–0.100 mm"), making the error analysis for PVC undefined.
[References] In the reference list, the entry by Guo et al. appears between [17] and [19] without its own number [18]; the citation [17-19] therefore does not match the formatted list.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the thickness formula is a first-principles interference derivation, measured values are compared independently against nominal scalars, and no fitted parameter or self-cited theorem forces the result.

full rationale

The paper's central estimate is obtained from Eq. (1-11), d = λ1λ2/(2n(λ1-λ2)), which follows from the adjacent-order interference conditions (1-9) and (1-10). This is a standard algebraic derivation from the physical model 2nd = jλ; the paper does not fit any parameter to the nominal thickness values, and the nominal values are not used in the calculation. The refractive index n is an assumed input rather than a fitted parameter, so an incorrect n would produce a reproducible but not circular error. The manual selection of 'continuous and stable bands' is an operator degree of freedom, but it does not make the output equal to an input by construction. There are also no load-bearing self-citations: the cited prior work is contextual background on thin-film interference and ellipsometry, not a uniqueness theorem or an ansatz that predetermines the result. The accuracy claims in §4.2 and §5 are contradicted by some entries in Table 1 (e.g., 58.9 µm for a 0.05PI film and 29.2 µm for a 0.025PI film), and the unreported refractive index values and spectrometer resolution hinder reproducibility. These are correctness and reporting problems, not circularity. Under the hard rule that circularity must be demonstrated by a specific reduction, no such reduction exists here; the honest finding is no significant circularity.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The ledger is dominated by unstated experimental conditions: the refractive indices of the test films and the spectral resolution are required by Eq. (1-11) and by the need to resolve adjacent peaks, but neither is provided in the manuscript. No new physical entities are introduced.

free parameters (1)

Refractive index n for PI and PVC films = not reported
Eq. (1-11) requires n to convert wavelength spacing to thickness; the paper never states the value used for polyimide or PVC, so all reported thickness values depend on an unstated input.

assumptions (5)

domain assumption The optical path difference in the film is ΔL = 2nd (Eq. 1-1), which assumes normally incident light.
Section 2, Figure 1B; the reflection probe geometry is not specified, so non-normal incidence would change the path length.
domain assumption Constructive interference occurs when ΔL/λ is an integer; interface phase shifts are ignored.
Equations (1-6) to (1-8); for a film on a substrate with different refractive indices, one reflection typically acquires a phase shift, shifting peak positions, though the adjacent-peak spacing formula is unchanged.
domain assumption The film thickness is uniform over the illuminated spot and the film is non-absorbing at visible wavelengths.
Section 2 states 'Because the thickness of the film to be measured is uniform'; the visible interference fringes imply transparency.
domain assumption The spectrometer can resolve adjacent interference peaks.
Section 3 describes a grating CCD spectrometer but gives no spectral resolution; for d=50 micrometers and n=1.5, adjacent peaks are roughly 1 to 2 nm apart, requiring sub-nanometer resolution.
domain assumption Nominal sample thickness values (0.025, 0.05, 0.1 mm) are accurate references.
Section 4.1 compares calculated values to 'scalar thickness'; no independent calibration of the films is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Film Thickness Gauge Based on Interferometric Principle of Y-shaped Optical Fiber." pith.science (2026). https://pith.science/paper/PYI4T3GC

@misc{pith2026241207528,
  author       = {Pith},
  title        = {Pith review of: Film Thickness Gauge Based on Interferometric Principle of Y-shaped Optical Fiber},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PYI4T3GC}},
  note         = {Machine review of arXiv:2412.07528}
}

read the original abstract

In this paper, a thin film thickness gauge based on the interferometric principle of Y-shaped optical fiber is proposed to achieve accurate measurement of film thickness. In this paper, the optical fiber, the interferometric principle and the film thickness calculation principle are introduced, and the interferometric thickness measurement system based on Y-shaped optical fiber is constructed. The system uses the special structure of Y-shaped optical fiber to transmit the optical signal generated by the light source to the surface of the thin film, and obtains coherent optical signals of different wavelengths through reflection and interference. The spectrometer is used to receive and interpret these interference signals, and the thickness of the film is calculated according to the wavelength difference of the peak positions of the adjacent stages, combined with the refractive index of the film. In the specific design, the paper elaborates on the design of each part of the instrument, including the selection and parameter setting of the light source, Y-fiber and spectrometer. Among them, the Y-shaped optical fiber, as the core component of the instrument, has the function of transmitting optical signals and detecting optical signals on the surface of thin films. At the same time, the paper also introduces the housing packaging and internal assembly process of the instrument to ensure the portability and stability of the instrument. The results show that the thickness gauge has high measurement accuracy and stability, which can meet the needs of practical applications.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

[3]

The main hardware is divided into three parts: a light sou rce, a Y -type optical fiber with a probe, and a spectrometer

Instrument design The construction of experimental instruments needs to meet: (1) The output of the light source optical signal and the input of the feedback optical signal and the external optical signal; (2) The light source is controllable; (3) Real-time collection and processing of optical signals. The main hardware is divided into three parts: a ligh...

work page
[4]

Nandi, E

S. Nandi, E. Olofsson, M. Bertolinom. Observation of Rabi dynamics with a short -wavelength free-electron laser [J] Nature, 2022, 608, 488-493

work page 2022
[5]

C. Wu, S. Kumar, Y . Kan, D. Komisar, Z. Wang, S. I. Bozhevolnyi, F. Ding. Room-temperature on-chip orbital angular momentum single-photon sources [J]. Science Advances, 2022, 8, 3075

work page 2022
[6]

R. M. Ma, R. F. Oulton. Applications of nanolasers [J]. Nature Nanotechnology, 2019, 14, 12-22

work page 2019
[7]

Z.Zhu, L. Wang, X. Huang, W. Huang. Room -temperature epitaxial welding of 3D and 2D perovskites [J]. Nature Materials, 2022, 21, 1042-1049

work page 2022
[8]

X. Y . Huang, H. B. Li, C. F. Zhang, S. J. Tan, Z. L. Chen, Z. D . Lu, X. Y . Wang, M. Xiao. Efficient plasmon -hot electron conversion in Ag –CsPbBr3 hybrid nanocrystals [J]. Nature Communications, 2019, 10, 1163

work page 2019
[9]

J. Guo, Y . Tang, L. Li, P. Jia, B. Song, X. Liu, X. Li, Z. Deng, Z. Lou, Y . Hu, F. Teng, L. Qin, Y . Hou. Aspect-ratio controllable growth of rectangular cesium lead bromide crystallites on PTAA modified substrates [J]. Journal of Materials Chemistry C, 2022, 10, 6473-6480

work page 2022
[10]

Markina, A

D.I. Markina, A. S. Sergey, M. A. Masharin, A. Peltek, P. Talianov, M. A. Baranov, A. Nikolaeva, M. V . Zyuzin, L. E. Zelenkov, A. P. Pushkarev, A. L. Rogach, S. V . Makarov. Excitonic versus Free -Carrier Contributions to the Nonlinearly Excited Photoluminescence in CsPbBr3 Perovskites [J]. ACS Nano, 2022, 9(1), 179-189

work page 2022

Show all 15 references

[11]

Z. L. Li, G. R. Mao, S. X. Dai, F. F. Xu, F. F. Chen. Third -order optical nonlinearity of CsPb(Br/I)3 metal halide perovskites nano -crystals embedded chalcogenide glass [J]. Optics Express, 2022, 30(16), 28647-28657

2022
[12]

Wang, Q.H

Wang, X.S. Wang, Q.H. Nie. Se -H-free As2Se3 fib er and its spectral applications in the mid-infrared [J]. Optics Express, 2022, 30(13), 24072-24083

2022
[13]

Measurement of the thickness and refractive index of very thin films and the optical properties of surfaces by ellipsometry[J]

Mccrackin F L, Passaglia E, Stromberg R R, et al. Measurement of the thickness and refractive index of very thin films and the optical properties of surfaces by ellipsometry[J]. Journal of Research of the National Institute of Standards and Technology, 2001, 106(3): 589-603

2001
[14]

Determination of the thickness of thin silane films on aluminium surfaces by means o f spectroscopic ellipsometry[J]

Franquet A, De Laet J, Schram T, et al. Determination of the thickness of thin silane films on aluminium surfaces by means o f spectroscopic ellipsometry[J]. Thin Solid Films, 2001, 384(1): Page 16 37-45

2001
[15]

Determination of film thickness and refractive index in one measurement of phase -modulated ellipsometry[J]

Pristinski D, Kozlovskaya V , Sukhishvili S A. Determination of film thickness and refractive index in one measurement of phase -modulated ellipsometry[J]. Journal of the Optical So ciety of America a-Optics Image Science and Vision, 2006, 23(10): 2639-2644. [16]Nestler P, Hel...

2006
[17]

Thin -film thickness profile and its refractive index measurements by dispersive white-light interferometry[J]

Ghim Y S, Kim S W. Thin -film thickness profile and its refractive index measurements by dispersive white-light interferometry[J]. Optics Express, 2006, 14(24): 11885-11891. Guo T, Chen Z, Li M H, et al. Film thickness measurement based on nonlinear phase analysis using a Linn...

2006
[19]

A Review of Thickness Measurements of Thick Transparent Layers Using Optical Interferometry[J]

Park J, Kim J A, Ahn H, et al. A Review of Thickness Measurements of Thick Transparent Layers Using Optical Interferometry[J]. International Journal of Precision Engineering and Manufacturing, 2019, 20(3): 463-477

2019

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [3]

The main hardware is divided into three parts: a light sou rce, a Y -type optical fiber with a probe, and a spectrometer

Instrument design The construction of experimental instruments needs to meet: (1) The output of the light source optical signal and the input of the feedback optical signal and the external optical signal; (2) The light source is controllable; (3) Real-time collection and processing of optical signals. The main hardware is divided into three parts: a ligh...

work page

[2] [4]

Nandi, E

S. Nandi, E. Olofsson, M. Bertolinom. Observation of Rabi dynamics with a short -wavelength free-electron laser [J] Nature, 2022, 608, 488-493

work page 2022

[3] [5]

C. Wu, S. Kumar, Y . Kan, D. Komisar, Z. Wang, S. I. Bozhevolnyi, F. Ding. Room-temperature on-chip orbital angular momentum single-photon sources [J]. Science Advances, 2022, 8, 3075

work page 2022

[4] [6]

R. M. Ma, R. F. Oulton. Applications of nanolasers [J]. Nature Nanotechnology, 2019, 14, 12-22

work page 2019

[5] [7]

Z.Zhu, L. Wang, X. Huang, W. Huang. Room -temperature epitaxial welding of 3D and 2D perovskites [J]. Nature Materials, 2022, 21, 1042-1049

work page 2022

[6] [8]

X. Y . Huang, H. B. Li, C. F. Zhang, S. J. Tan, Z. L. Chen, Z. D . Lu, X. Y . Wang, M. Xiao. Efficient plasmon -hot electron conversion in Ag –CsPbBr3 hybrid nanocrystals [J]. Nature Communications, 2019, 10, 1163

work page 2019

[7] [9]

J. Guo, Y . Tang, L. Li, P. Jia, B. Song, X. Liu, X. Li, Z. Deng, Z. Lou, Y . Hu, F. Teng, L. Qin, Y . Hou. Aspect-ratio controllable growth of rectangular cesium lead bromide crystallites on PTAA modified substrates [J]. Journal of Materials Chemistry C, 2022, 10, 6473-6480

work page 2022

[8] [10]

Markina, A

D.I. Markina, A. S. Sergey, M. A. Masharin, A. Peltek, P. Talianov, M. A. Baranov, A. Nikolaeva, M. V . Zyuzin, L. E. Zelenkov, A. P. Pushkarev, A. L. Rogach, S. V . Makarov. Excitonic versus Free -Carrier Contributions to the Nonlinearly Excited Photoluminescence in CsPbBr3 Perovskites [J]. ACS Nano, 2022, 9(1), 179-189

work page 2022