Interferometric modal splitting enables a broadband, dual-polarization on-chip spectrometer
Karl Johnson, Vladimir Fedorov, Dmitrii Belogolovskii, Andrew Grieco, Noah A. Rubin, Yeshaiahu Fainman · 2025
Read Interferometric modal splitting enables a broadband, dual-polarization on-chip spectrometer by Karl Johnson, Vladimir Fedorov, Dmitrii Belogolovskii, Andrew Grieco, Noah A. Rubin, Yeshaiahu Fainman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-07300.txt.
Introduction
Karl Johnson and colleagues demonstrated a tiny light analyzer built on a chip in 2025. Spectrometers split light by color to identify materials, and chip versions suit portable sensors. A stubborn limit is that chip guides usually carry only one polarization of light, the direction its waves vibrate, over a useful color range.
Their heat-tuned silicon device turned that flaw into a feature: the two polarizations separate on their own in the output pattern, even with a single plain detector. Tests recovered both polarization spectra across a wide near-infrared band with strong separation between them. Added calibration tricks also improved single-polarization use. Gas sensing and health monitors could gain cheaper, smaller analyzers.
Catalog record
| CREATOR | Karl Johnson, Vladimir Fedorov, Dmitrii Belogolovskii, Andrew Grieco, Noah A. Rubin, Yeshaiahu Fainman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-07300
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Interferometric modal splitting enables a broadband, dual-polarization on-chip spectrometer Karl Johnson, Vladimir Fedorov, Dmitrii Belogolovskii, Andrew Grieco, Noah A. Rubin, Yeshaiahu Fainman Abstract 1 Abstract The modal dispersion of waveguides often limits integrated photonic devices to operation with a single polarization state. This presents a challenge for sensing and spectroscopy applications, which often require polarization diversity over wide bandwidths with high throughput. Here, we show that an unmodified thermally-driven silicon photonic Fourier transform spectrometer exhibits a polarization-separating effect in the frequency domain, even though only one polarization-insensitive detector is used. Using this effect, we experimentally demonstrate a simple on-chip spectrometer capable of extracting two-polarization spectra over a wide 1480-1630 nm bandwidth with a greater than 20 dB polarization extinction ratio. These specifications would be highly challenging to achieve using existing, conventional on-chip polarization-splitting techniques. We additionally demonstrate several improvements in calibration and testing that improve the performance of on-chip Fourier transform spectrometers even in the single-polarization case. The “interferometric mode splitting” principle which this spectrometer exemplifies is general to various on-chip spectrometer architectures, other spatial modes, and technologies other than thermally-driven Fourier transform spectrometers. Interferometric mode splitting shows promise as a general approach for robust and fundamentally broadband detection of orthogonal modes in guided-wave sensing. 2 Introduction Developments in integrated photonics have revolutionized optical communications, and have brought new capabilities for basic research [ 1 , 2 , 3 ] . The unique physics and size advantages of on-chip optics have also stimulated substantial interest in optical sensing applications, both in miniaturizing existing technologies and in realizing new ones [ 4 , 5 , 6 , 7 ] . In recent years, miniaturized on-chip optical spectrometers have seen notable advancements [ 8 ] . The resolution, bandwidth, and throughput of on-chip spectrometers have dramatically increased, while the footprint of such devices has continued to shrink [ 9 , 10 , 11 ] . However, transitioning optical systems to a chip often introduces unique challenges. Most integrated photonic devices are designed to operate using only one waveguide mode, corresponding to a single spatial mode with one polarization. In optical communications, mature solutions exist for supporting two polarization modes, through a combination of polarization-splitting components [ 12 , 13 ] and signal processing to account for polarization fluctuations in incoming light and in the fiber-optic channel [ 14 , 15 ] . Meanwhile, for optical sensing applications - such as optical spectroscopy - in which light may move on and off the chip and interact with the environment, this single-mode limitation is a more serious issue. Practical applications of optical spectroscopy demand the highest bandwidth and throughput possible. Supporting only one waveguide mode yields a substantial throughput limit for coupling extended light sources on-chip [ 16 ] . Additionally, single-polarization operation may lead to ambiguities or artifacts in the measurement if incoming light has a spectrally-varying polarization state. One method to achieve multi-mode operation is to couple off-chip light into several waveguide modes and use an on-chip mode or polarization splitter to route the light to multiple copies of the on-chip device [ 12 ] . However, this increases system complexity, and designing such on-chip mode splitters to have a wide bandwidth is highly challenging. For example, on-chip broadband polarization splitters typically exhibit poor polarization extinction ratios (PER’s) of just 12-16 dB over the wide bandwidths of interest for spectroscopy applications [ 17 , 18 , 19 ] . Guided wave technologies such as integrated photonics struggle to simultaneously operate with more than one mode for a simple reason: every mode in a waveguide has different properties, all of which are highly wavelength-dependent. Substantial design effort can yield devices that behave similarly for multiple modes at specific wavelengths, but such efforts to “correct” the differences between modes are inherently narrowband [ 14 ] . On the other hand, in the field of free-space polarization optics, highly broadband polarization splitters have been constructed for hundreds of years [ 20 ] . This bandwidth is derived by harnessing the differences in absorption or refractive index between polarizations that remain valid over broad bandwidths. This suggests that harnessing the differences between modes rather than laboring to “correct” them is a more robust approach for broadband management of several modes in integrated photonics. Indeed, existing integrated mode splitters leverage the difference in effective index between modes for their operation; however, mode splitters are coupled-mode devices relying on phase-matching, which is also a fundamentally bandwidth-limited technique [ 21 ] . Figure 1: a) Typical methods of polarization or mode diversity in optical systems require beam splitters, which suffer from poor bandwidth and PER on-chip. b) In the method proposed here, modes share a common physical channel in which they experience different optical responses to a common perturbation (e.g., electric field, stress, or heat). These different responses enable mode separation in post-processing, avoiding the challenges inherent to physical separation over broad bandwidths. In contrast to the narrowband nature of coupled-mode devices, interferometric structures can convert phase variations between modes into intensity variations (fringes) across very broad bandwidths. In this work, we propose that interferometric photonic sensing architectures can utilize this inherent advantage to naturally separate modes over broad bandwidths (Fig. 1). To demonstrate this general principle of “interferometric modal splitting” (IMS), we show that temperature-derivative modal dispersion (TDMD) causes frequency-domain polarization splitting in a thermally actuated Fourier transform spectrometer. This phenomenon manifests in the waveguides themselves, and simple signal processing is used to extract the two spectra from data measured using a polarization-insensitive detector. There are no changes to the spectrometer design or architecture from previous works [ 22 , 23 ] . The demonstrated spectrometer is capable of simultaneously measuring spectra of two orthogonal polarizations with a wide bandwidth (1480-1630 nm), high extinction ratio (approximately 20 dB), and moderate resolution ( R=\frac{\lambda}{\Delta\lambda}\approx 600 for the TE mode and R\approx 1500 for the TM mode, or \Delta\lambda= 2.5 and 1 nm, respectively). The measured bandwidth and extinction ratio reach the limits of our test equipment. Beyond the demonstration here, IMS easily generalizes to orthogonal modes other than polarization, more advanced spectrometer architectures, and modal differences other than TDMD. 3 Results 3.1 Temperature-derivative modal dispersion and frequency-domain mode splitting In an interferometer with a monochromatic input of frequency \nu_{0} , the fluctuations in output intensity (interferogram) are sinusoidal with the optical path length (OPL) difference between the two sides (Fig. 2a-b), I_{\text{out}}=I_{0}\cos(\phi_{2}-\phi_{1})=I_{0}\cos(k_{0}(\text{OPL}_{2}-\text{OPL}_{1}))=I_{0}\cos(k_{0}\Delta\text{OPL}), (1) where I_{0} is the intensity of the monochromatic input, k_{0} is the free space wavenumber \frac{2\pi\nu_{0}}{c_{0}} , and \text{OPL}_{1,2} are defined in Fig. 2a. Note that here we drop the constant intensity offset of this interferogram (which is required for the intensity to remain positive) for brevity. In this work, we consider interferometers in which the path length \Delta\text{OPL} is varied by use of heat and the thermo-optic effect. In the case of a thermally-driven, balanced interferometer with L_{1}=L_{2} and a linear thermo-optic response, the measured interferogram is dependent only on \Delta T=T_{2}-T_{1} , with \Delta\text{OPL}=L\frac{dn_{\text{eff}}}{dT}\Delta T such that I_{\text{out}}(\Delta T)=I_{0}\cos(\frac{2\pi\nu_{0}}{c_{0}}L\frac{dn_{\text{eff}}}{dT}\Delta T). (2) This has a frequency with respect to \Delta T of f_{T}=\frac{L}{c_{0}}\nu_{0}\frac{dn_{\text{eff}}}{dT}\Big|_{\nu=\nu_{0}}. (3) The one-to-one relationship between frequency in the interferogram and optical frequency is the classic operating principle of a Fourier transform spectrometer.