The transfer of high-quality time-frequency signals between remote locations underpins a broad range of applications including precision navigation and timing, the new field of clock-based geodesy, long-baseline interferometry, coherent radar arrays, tests of general relativity and fundamental constants, and the future redefinition of the second [1-7]. However, present microwave-based time-frequency transfer [8-10] is inadequate for state-of-the-art optical clocks and oscillators [1,11-15] that have femtosecond-level timing jitter and accuracies below 1E-17; as such, commensurate optically-based transfer methods are needed. While fiber-based optical links have proven suitable [16,17], they are limited to comparisons between fixed sites connected by a specialized bidirectional fiber link. With the exception of tests of the fundamental constants, most applications instead require more flexible connections between remote and possibly portable optical clocks and oscillators. Here we demonstrate optical time-frequency transfer over free-space via a two-way exchange between coherent frequency combs, each phase-locked to the local optical clock or oscillator. We achieve femtosecond-scale timing deviation, a residual instability below 1E-18 at 1000 s and systematic offsets below 4E-19, despite frequent signal fading due to atmospheric turbulence or obstructions across the 2-km link. This free-space transfer would already enable terrestrial links to support clock-based geodesy. If combined with satellite-based free-space optical communications, it provides a path toward global-scale geodesy, high-accuracy time-frequency distribution, satellite-based relativity experiments, and "optical GPS" for precision navigation

We demonstrate coherent dual frequency-comb spectroscopy for detecting variations in greenhouse gases. High signal-to-noise spectra are acquired spanning 5990 to 6260 cm -1 (1600 to 1670 nm) covering ~700 absorption features from CO 2 , CH 4 , H 2 O, HDO, and 13 CO 2 , across a 2-km open-air path. The transmission of each frequency comb tooth is resolved, leading to spectra with <1 kHz frequency accuracy, no instrument lineshape, and a 0.0033-cm -1 point spacing. The fitted path-averaged concentrations and temperature yield dry-air mole fractions. These are compared with a point sensor under well-mixed conditions to evaluate current absorption models for real atmospheres. In heterogeneous conditions, timeresolved data demonstrate tracking of strong variations in mole fractions. A precision of <1 ppm for CO 2 and <3 ppb for CH 4 is achieved in 5 minutes in this initial demonstration. Future portable systems could support regional emissions monitoring and validation of the spectral databases critical to global satellitebased trace gas monitoring.

We demonstrate a high-accuracy dual-comb spectrometer centered at 3.4 μm. The amplitude and phase spectra of the P, Q, and partial R branches of the methane ν 3 band are measured at 25 to 100 MHz point spacing with resolution under 10 kHz and a signal-to-noise ratio of up to 3500. A fit of the absorbance and phase spectra yields the center frequency of 132 rovibrational lines. The systematic uncertainty is estimated to be 300 kHz, which is 10 −3 of the Doppler width and a 10-fold improvement over Fourier transform spectroscopy. These data quantify the accuracy and resolution achievable with direct comb spectroscopy in the midinfrared.

We have studied the effect of growth and design parameters on the performance of Si-doped GaN/AlN multiquantum-well ͑MQW͒ structures for intersubband optoelectronics in the near infrared. The samples under study display infrared absorption in the 1.3-1.9 m wavelength range, originating from the photoexcitation of electrons from the first to the second electronic level in the QWs. A commonly observed feature is the presence of multiple peaks in both intersubband absorption and interband emission spectra, which are attributed to monolayer thickness fluctuations in the quantum wells. These thickness fluctuations are induced by dislocations and eventually by cracks or metal accumulation during growth. The best optical performance is attained in samples synthesized with a moderate Ga excess during the growth of both the GaN QWs and the AlN barriers without growth interruptions. The optical properties are degraded at high growth temperatures ͑Ͼ720°C͒ due to the thermal activation of the AlN etching of GaN. From the point of view of strain, GaN/AlN MQWs evolve rapidly to an equilibrium average lattice parameter, which is independent of the substrate. As a result, we do not observe any significant effect of the underlayers on the optical performance of the MQW structure. The average lattice parameter is different from the expected value from elastic energy minimization, which points out the presence of periodic misfit dislocations in the structure. The structural quality of the samples is independent of Si doping up to 10 20 cm −3 . By contrast, the intersubband absorption spectrum broadens and blueshifts with doping as a result of electron-electron interactions. This behavior is independent of the Si doping location in the structure, either in the QWs or in the barriers. It is found that the magnitude of the intersubband absorption is not directly determined by the Si concentration in the wells. Instead, depending on the Al mole fraction of the cap layer, the internal electric field due to piezoelectric and spontaneous polarization can deplete or induce charge accumulation in the QWs. In fact, this polarization-induced doping can result in a significant and even dominant contribution to the infrared absorption in GaN/AlN MQW structures.

Mid-infrared dual-comb spectroscopy has the potential to supplant conventional high-resolution Fourier transform spectroscopy in applications that require high resolution, accuracy, signal-to-noise ratio, and speed. Until now, dual-comb spectroscopy in the mid-infrared has been limited to narrow optical bandwidths or to low signal-to-noise ratios. Using a combination of digital signal processing and broadband frequency conversion in waveguides, we demonstrate a midinfrared dual-comb spectrometer that can measure comb-tooth resolved spectra across an octave of bandwidth in the mid-infrared from 2.6-5.2 µm with sub-MHz frequency precision and accuracy and with a spectral signal-to-noise ratio as high as 6500. As a demonstration, we measure the highly structured, broadband cross-section of propane (C3H8) in the 2860-3020 cm -1 region, the complex phase/amplitude spectrum of carbonyl sulfide (COS) in the 2000 to 2100 cm -1 region, and the complex spectra of methane, acetylene, and ethane in the 2860-3400 cm -1 region.Mid-infrared spectroscopy is a powerful technique for the multispecies detection of trace gases with applications ranging from the detection of hazardous materials, to environmental monitoring and industrial monitoring. Compared to the near-infrared, where laser sources are more plentiful, the techniques for measuring mid-infrared spectra are more limited. Mid-infrared spectra are most commonly acquired by Fourier transform spectroscopy (FTS), which provides accurate and high resolution spectra but requires a scanning delay arm and blackbody source leading to large instruments and long acquisition times. Dual-comb spectroscopy (DCS) is a high-performance alternative to conventional FTS providing high resolution, absolute frequency accuracy, fast acquisition times, long interaction lengths, broad bandwidth coverage, and high signal-to-noise ratio [1,2]. The advantages of speed and long path length are of particular relevance to non-laboratory applications, for example in open-path atmospheric monitoring or industrial process monitoring [3][4][5][6]. However, up until now, DCS has only been demonstrated with its full panoply of advantages in the near-infrared, from ~ 1 to 2 µm [7-10]. The near-infrared has much more limited applications compared to the mid-infrared since molecular crosssections are typically 1000 times weaker, if they exist at all. DCS in the mid-infrared has indeed been actively pursued [11][12][13][14][15][16][17][18][19][20][21][22][23][24], yet it is not competitive with high-resolution conventional FTS, limited by the coherence and/or the bandwidth of mid-infrared comb sources.Quantitative broadband mid-infrared DCS requires addressing strong overlapping requirements on the underlying mid-infrared frequency combs: they must produce broad and relatively flat optical spectra while maintaining mutual coherence over the measurement time. Without coherence, adjacent comb teeth blend together, sacrificing orders of magnitude in spectral resolution and obscuring both the frequency and amplit...