Dual Tunable MZIs Stationary-Wave Integrated Fourier Transform Spectrum Detection

Xinyang Chen, Peijian Huang, Ning Wang, Yong Zhu, Jie Zhang2021 · DOI

TL;DR

This paper introduces a novel on-chip spectrometer design using dual tunable Mach-Zehnder interferometers (MZIs) and strategically placed gold nanowires. This approach overcomes spectral aliasing issues in traditional stationary-wave integrated Fourier transform (SWIFT) spectrometers, enabling high spectral resolution with a compact, integrated system suitable for applications like chemical analysis and Raman spectroscopy.

What Problem Does This Paper Address?

Traditional spectrometers, while powerful, are often bulky and expensive. Integrated on-chip spectrometers offer a more compact and cost-effective solution, but existing designs like arrayed waveguide gratings (AWGs) and planar concave gratings can require a large number of detectors, leading to complex structures and reduced signal-to-noise ratios. Fourier transform spectrometers (FTS) offer high resolution and signal-to-noise, but traditional implementations face challenges with on-chip integration. Specifically, stationary-wave integrated Fourier transform (SWIFT) spectrometers, while promising, often struggle with spectral aliasing due to under-sampling, which limits their achievable spectral resolution and bandwidth. This aliasing is a consequence of the Nyquist-Shannon criterion, which demands very small detector spacing that is difficult to achieve with conventional detectors. The paper aims to address these limitations by proposing an all-on-chip SWIFT system that can achieve high spectral resolution without the need for a large array of detectors and without the aliasing problem.

Methods

The core innovation is a "Dual Tunable MZIs Stationary-Wave Integrated Fourier Transform" (DTM-SWIFT) system. This system utilizes two tunable MZIs to generate interference fringes. Instead of relying on a large array of detectors to sample these fringes, the DTM-SWIFT employs an array of gold nanowires positioned asymmetrically along a straight waveguide (L5). These nanowires act as both scattering elements and sampling points for the evanescent wave of the interference fringes. By dynamically tuning the phase difference within the MZIs, different parts of the interference pattern are captured by the nanowires. This allows for the reconstruction of the full interferogram, which is then processed using Fourier Transform (FT) technology to obtain the spectrum. The use of gold nanowires is motivated by their ability to scatter light and enhance the intensity of the evanescent wave, making them suitable for sampling with standard CCD detectors. The paper also details the theoretical analysis of the optical distribution and the mathematical framework for signal reconstruction, including the influence of thermo-optic effects in Si3N4 waveguides for phase tuning.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis and simulation-based results for two types of spectra: a Gaussian spectrum and a Raman spectrum of an R6G molecule.

Key Insights and Ideas

  • Overcoming Spectral Aliasing with Nanowires: The most significant insight is the use of asymmetrically placed gold nanowires to sample interference fringes. This approach effectively circumvents the need for extremely dense detector arrays, a major hurdle for achieving high resolution in integrated SWIFT systems.
  • Dynamic Phase Tuning for Complete Interferogram: The dual tunable MZIs allow for precise control over the phase difference, enabling the system to capture different segments of the dynamic interferogram. This dynamic sampling is crucial for reconstructing the complete spectral information.
  • Si3N4 for Integrated Photonics: The choice of Si3N4 as the waveguide material is justified by its wide bandgap, transparency across a broad spectrum, and high refractive index contrast, which are advantageous for compact and efficient integrated photonic devices.
  • Nanowire Width as a Critical Parameter: The paper highlights that the width of the gold nanowires is a critical parameter. If the width is an integer multiple of the stationary wave period, it can lead to missing spectral information at specific wavelengths, underscoring the importance of careful design and fabrication.

Key Results

  • Gaussian Spectrum Reconstruction: The DTM-SWIFT system achieved a resolution of 6 nm for reconstructing a Gaussian spectrum using only six gold nanowires, with a temperature tuning range of less than 24.2 K and tuning times less than 100.
  • Raman Spectrum Reconstruction: For Raman spectrum reconstruction, the system achieved a resolution of 5 cm⁻¹ using 25 gold nanowires, with a temperature tuning range of less than 121 K and tuning times less than 500.
  • Peak Position Error: The reconstructed Raman spectrum showed a peak position error of less than 1 cm⁻¹ at the Raman peak of 1365 cm⁻¹.
  • Nanowire Width Impact: Simulations demonstrated that specific nanowire widths (e.g., 127.5 nm and 128.5 nm for certain wavelengths) could lead to missing spectral signals, while a width of 150 nm was found to be a cost-effective compromise for Raman spectroscopy.

Limitations and Future Work

  • Nanowire Width Optimization: While the paper identifies the importance of nanowire width, further optimization and experimental validation are needed to precisely control this parameter for optimal spectral reconstruction across a wide range of wavelengths.
  • Temperature Control: The system relies on thermal tuning for phase control. While the paper addresses temperature error, achieving precise and rapid temperature control in a compact on-chip system remains a practical challenge.
  • Integration with Raman Sensors: The authors suggest combining this DTM-SWIFT system with a waveguide Raman sensor to create a fully integrated on-chip Raman spectroscopy sensing and detection system, which represents a promising future direction.
  • Experimental Validation: The results presented are primarily based on theoretical analysis and simulations. Experimental verification of the proposed DTM-SWIFT system is a crucial next step.

Summary updated

BibTeX
@misc{chen2021dual,
  author = {Xinyang Chen and Peijian Huang and Ning Wang and Yong Zhu and Jie Zhang},
  title = {{Dual Tunable MZIs Stationary-Wave Integrated Fourier Transform Spectrum Detection}},
  year = {2021},
  doi = {10.3390/s21072352},
}
Figure