Silicon-based FTIR spectrometer

Silicon chip could make handheld FTIR spectrometers cheaper to manufacture

Silicon-based FTIR spectrometer

Silicon-based FTIR spectrometer

SHARE:





Feb. 23, 2018

Using silicon, researchers at the University of California San Diego and the University of Campinas in Brazil have developed the first silicon-based, miniaturized FTIR spectrometer. The advance could lead to the development of low-cost, portable chemical sensing devices for applications ranging from greenhouse gas monitoring to point-of- care diagnostics and quality control for drug development.

The findings were recently published in Nature Communications.

FTIR (Fourier Transform Infrared) spectrometers are one of the most widely used research instruments to identify and analyze chemical substances. They generally work by shining infrared light on a sample and then measuring how much light and what wavelengths get absorbed. The absorption pattern provides information on the sample’s chemical makeup.

Conventional FTIR spectrometers are bulky tabletop instruments, which makes them unsuitable for field use. Various miniaturized FTIR spectrometers are being developed so they can be integrated into drones, mobile phones and other portable devices for on-site spectral analysis. However, they are still expensive to manufacture.

To overcome these limitations, a team of researchers at UC San Diego and the University of Campinas built a miniature FTIR spectrometer device using silicon photonics—the same technology used to fabricate electronic chips in computers, smartphones and other consumer electronic devices. “Silicon photonics technology offers a platform for high-volume, low-cost manufacturing of miniaturized spectrometers,” said Mario Souza, the study’s first author. Souza, a visiting scholar from the University of Campinas, conducted the research while working in the lab of electrical engineering professor Shaya Fainman at UC San Diego.

However, integrating silicon photonics into an FTIR spectrometer involves several technical challenges, Souza explained. First, operating the device requires that the refractive index of the optical waveguide be thermally tuned, but this becomes problematic at high temperatures, which are required for high resolution. The refractive index change becomes non-linear and the waveguide undergoes significant thermal expansion. Second, silicon waveguides are highly dispersive—each wavelength effectively experiences a different change in refractive index.

The researchers addressed these challenges by inventing a laser calibration method that can quantify and correct the distortions caused by dispersion and non-linearity.

As a proof of concept, the team developed a small FTIR spectrometer chip (1 mm 2 ) using standard silicon photonics fabrication procedures. The chip was tested in a tabletop FTIR spectrometer. It produced a broadband test spectrum (7 THz wide around 193.4 THz) with a spectral resolution of 0.38 THz, which is comparable to the resolution of available commercial portable spectrometers operating at the same wavelength range. And because the distortions caused by dispersion and thermo-optic non-linearity were quantified and corrected, the device consumed 35 percent less power to achieve that resolution.

The researchers also note that the silicon-based FTIR spectrometer is not affected by fabrication imperfections. “Just like in the electronics industry, large-scale fabrication is the route to reducing costs in silicon photonics, but there are unavoidable chip-scale variations in the fabrication process. Most on-chip spectrometer proposals are very sensitive to such effects, whereas the FT design is transparent to it,” Souza said.

As a next step, the team plans to implement a fully functional packaged device consisting of an optimized design integrated with photodetectors, light source and input/output optical fibers.

Paper title:Fourier transform spectrometer on silicon with thermo-optic non-linearity and dispersion correction.

Authors of the study are Mario C. M. M. Souza and Newton C. Frateschi at the University of Campinas, Brazil; and Andrew Grieco and Yeshaiahu Fainman at UC San Diego.

Microsoft
Development Category (English)728x90