Demonstration and frequency noise characterization of a 17 $μ$m quantum cascade laser

TL;DR

This work demonstrates a continuous-wave, room-temperature CW DFB QCL at $17~\mu\mathrm{m}$ and investigates its spectroscopic and frequency-noise performance by performing broadband absorption spectroscopy of the $\nu_2$ mode in $\,\mathrm{N_2O}$ and by measuring the laser's frequency noise using a molecular line discriminator. The authors map the spectral coverage, calibrate absolute frequencies against the NIST dataset, and quantify the laser's noise through the frequency-noise power spectral density, revealing a low-frequency $1/f$ flicker regime and a white-noise plateau with $N_{\mathrm{w}} \approx 60\times10^{3}$ Hz$^2$/Hz leading to an intrinsic width $\Delta\nu_{\mathrm l} \approx 200$~kHz. Surprisingly, a detailed line-shape analysis shows an effective FWHM up to $\sim 350$~kHz at $T_{\mathrm int}=1$~s, far exceeding the theoretical intrinsic width $\Delta\nu_{\mathrm l,th} \approx 1$~kHz, suggesting new physics or unaccounted noise mechanisms at $17~\mu\mathrm{m}$. The results highlight the potential of long-wavelength, narrow-line MIR QCLs for atmospheric sensing, metrology, and spectroscopy of ultra-cold or large molecules, while also motivating development of improved theoretical models and stabilization strategies (e.g., locking to molecular lines, cavities, or frequency combs) for MIR frequency standards.

Abstract

We evaluate the spectral performance of a novel continuous-wave room-temperature distributed feedback quantum cascade laser operating at the long wavelength of 17 $μ$m. By demonstrating broadband laser absorption spectroscopy of the $ν$2 fundamental vibrational mode of N2O molecules, we have determined the spectral range and established the spectroscopic potential of this laser. We have characterized the frequency noise and measured the line width of this new device, uncovering a discrepancy with the current consensus on the theoretical modeling of quantum cascade lasers. Our results confirm the potential of such novel narrow-line-width sources for vibrational spectroscopy. Extending laser spectroscopy to longer wavelength is a fascinating prospect that paves the way for a wide range of opportunities from chemical detection, to frequency metrology as well as for exploring light-matter interaction with an extended variety of molecules, from ultra-cold diatomic species to increasingly complex molecular systems.

Figures (6)

• Figure 1: a) Optical output power and voltage versus drive current recorded for various QCL temperatures. b) Emission spectra, measured by a Fourier transform infrared (FTIR) spectrometer for various QCL temperatures and drive currents shown in the legend. c) Emission spectrum measured at 250 K and 550 mA at higher FTIR spectrometer resolution, demonstrating a Side Mode Suppression Ratio (SMSR) of at least 25 dB.
• Figure 2: Measured linear absorption spectrum of N2O (red curve) at a pressure of 500 Pa. The bar spectrum indicates the line frequencies and intensities reported in the NIST database makiOnline98 for five different N2O isotopologues, the main contributions coming from $^{14}$N$_2^{16}$O (99% abundance). The blue sticks are characterized by a better accuracy than the black sticks makiOnline98.
• Figure 3: QCL frequency noise measurement setup. A 24-cm long cell filled at a N2O pressure of 80 Pa is used for the QCL frequency noise analysis. The graph shows a measurement of the $P(9)$ line of the $v_2$ vibrational bending mode at 581.24 cm$^{-1}$ used as as a frequency-to-amplitude converter. The arrow shows the QCL frequency when recording the frequency noise power spectral density. The QCL is loosely locked to the side of the molecular line. Intensity fluctuations proportional to the laser frequency noise are recorded on the detector and processed by the Fast Fourier Transform (FFT) spectrum analyzer. Both the gas cell and the MCT detector are tilted to minimize optical feedback to the QCL.
• Figure 4: Frequency noise power spectral density (PSD) of the 17 µ m QCL (red line (i)). The contributions from the laser driver current noise (black line (ii)) and laser intensity noise (blue line (iii)) are also shown for comparison. The laser intensity contribution has been recorded with the same QCL parameters (temperature, current) as for the frequency noise measurement. The optical power hitting the detector is the same for both measurements. The green line is the $\beta$-separation line defined by $8 \ln(2) f / \pi^2$ and used to estimate the laser line width from the PSD measurement.
• Figure 5: a): FWHM as a function of integration time. Square points: FWHM inferred from a full line shape reconstruction based on the measured frequency noise PSD (red curve in Figure \ref{['fig:NoisePSD']}). Black line: FWHM determined using the $\beta$-separation line method. b): Corresponding QCL line shapes for three particular integration times: $1$ s (blue), $1$ ms (green) and 20 µ s (red). The same color code is used for the associated FWHM line widths in the a) panel.