1. INTRODUCTION

The ever-increasing demand for higher angular, spectral, and temporal resolution over a broad wavelength range is what motivates continuing developments across many fields of instrumentation for astronomy. Since the 1980s, the demands of observing faint celestial sources have led to huge leaps in adaptive optics (AO), robotic positioning, and detector technologies. The coming generation of extremely large, 30-m class telescopes has ushered in a new era of technological developments. With these future systems, astronomers hope to better understand the formation of planets, stars, and galaxies to characterize the atmospheres of exoplanets in the habitable zone, image the distant corners of our universe, capture the dynamics of active galactic nuclei, and come closer to unraveling the mystery of black hole physics.

During the past two decades, astrophotonics [1–4] has emerged as a technology for astronomical instruments, providing photonic components that guide, shape, and manipulate the light coming from telescopes. These fiber- or chip-based components have an advantage in that they can reduce the footprint and weight compared to their free-space, bulk optics counterparts, thus lowering size, weight, and often cost of an instrument. Photonics-based solutions can also reduce the effects of environmental changes on the different beam paths, such as temperature variations or vibrations, hence significantly improving instrument stability. With cutting-edge manufacturing methods and the devices’ small size, complex waveguide structures and small-scale optical interactions are possible, leading to new capabilities that are nearly impossible to realize otherwise.

Two main developments are currently ongoing to push the limits of astronomical observations: first, the advent of the new generation of 30-m class telescopes, such as the Extremely Large Telescope (ELT) currently under construction at Cerro Armazones in Chile. These new facilities will enable the detection of smaller details and fainter objects, due to better angular resolution and larger photon collection areas, respectively. Second, progress in AO can significantly reduce the wavefront distortion caused by air turbulence in the atmosphere—so-called extreme AO is capable of detailed wavefront sensing and fast control, using a large number of actuators in the control mirror. This means not only higher angular resolution but also better coupling into single-mode fibers (SMFs) and photonic components for light processing, which is crucial for applications with low light levels. To utilize the advantages of these investments, astronomical instruments have to progress in parallel to improve their functionality, stability, and efficiency, with the help of astrophotonics.

2. HIGHLIGHTS OF THIS FEATURE ISSUE

In the following, we provide a summary of astrophotonic topics to provide a context for the papers that appear in this feature issue.

A. Fibers and Photonic Lanterns

To collect and distribute light from the telescope, multi-mode fibers (MMFs) or SMFs can be used instead of or in addition to free-space optics. This enables flexible designs with mechanically decoupled optical systems and instruments. While more light can be coupled into MMF, which is particularly important in the case of imperfect AO correction, some applications require or even benefit from a single mode of light (e.g., mode-dependent filtering wavelength in fiber Bragg gratings (FBGs), phase preservation in interferometry). However, if the fiber mode is not perfectly matched, a large fraction of light will be lost, degrading significantly the instrumental throughput. A review by Ellis et al. [5] discusses how light can be efficiently coupled into astronomical instruments using fibers. In particular for applications using MMF, complex effects inside the fiber (e.g., modal noise) make it hard to predict the resulting throughput. This issue includes a paper by Hernandez et al. [6], presenting a novel numerical model for field propagation in MMF for astronomical applications to guide the design of spectrographs by using precise end-to-end simulations.

Photonic lanterns (PLs) [7,8] are an astrophotonic solution where a multi-mode input is converted into many single-mode (or differently configured multi-mode) outputs. For the last 20 years, the design and manufacturing of PLs has been an active research field due to their unique capability to adiabatically couple one input fiber into multiple output fibers (and vice versa). This feature issue includes four papers on different aspects of PLs. Lin et al. [9] investigate PL designs in the context of diffraction-limited spectrometry. Moraitis et al. [10] report the first laboratory tests of a PL prototype developed for PolyOculus, and demonstrate coupling efficiencies of 91%, which is higher than that of concepts using free-space optical linkages (with 80% efficiency). Diab et al. [11] propose designs of mode-selective PLs (MSPL) and present simulation results demonstrating the potential of such MSPLs to reduce the number of required outputs. Davenport et al. [12] explore fiber packing topology to improve the manufacturing of PL devices. The authors fabricate PLs with 19 and 37 SMFs and find a good match when comparing the resulting SMF positions to theoretical values, demonstrating that topological circle packing data are a good predictor for optimal PL parameters.

B. Fiber Bragg Gratings

Ground-based near-infrared (NIR) observations are adversely affected by the presence of an extremely bright and variable background produced by the rotational and vibrational de-excitation of hydroxyl (OH) molecules at an altitude of 90 km. Nearly two decades ago, complex FBGs were introduced [13] as a viable solution to suppress these OH emission lines, which led to the development of the first FBG-based OH suppression demonstrator called GNOSIS [14]. While GNOSIS, commissioned at the 3.9-m Anglo-Australian Telescope, used the existing IRIS2 spectrograph, a dedicated NIR spectrograph optimized for fiber feed was used in its successor instrument PRAXIS. PRAXIS incorporated the OH suppression module inherited from GNOSIS, and the on-sky demonstration [15] showed promising results validating the notch filter characteristics. These FBG filters were inscribed in SMFs and spliced with PLs for coupling light from MMFs to SMFs and then back to the MMFs after filtering. This design made the system bulky and resulted in ${3} \times 133$ fusion splices incurring losses at each splice junction. However, to date, this is the only set of FBG-based OH suppression filters existing that was tested on-sky. Research efforts on novel approaches to design and manufacture photonic filters for OH suppression are ongoing. One such effort is fabricating FBG notch filters in multi-core fiber (MCF), which can significantly reduce splice losses and design complexities. Goebel et al. [16] report the fabrication of FBG notch filters in different types of MCFs using ultrashort laser pulses and the phase mask technique. A significantly large core-to-core deviation of the resonance wavelength of up to 0.45 nm was observed in an FBG inscribed in a commercially available seven-core fiber. This will render the filters unsuitable for use in astronomy. Two options are presented in this paper to overcome the large core-to-core wavelength variation of the FBGs inscribed in an MCF, thus fulfilling the requirements for OH suppression in astronomy. By suppressing the OH emissions, exciting observations come into reach, such as the search for positronium, an exotic and short-lived atom that consists of an electron and its antiparticle, the positron. In this issue, Robertson et al. [17] discuss the search for positronium with an OH-suppressed diffraction-limited spectrograph, aiming for the infrared (IR) light that is expected to be emitted by a radiative process before annihilation in the sky.

C. Spectrographs and Frequency References

Key components of the spectral analysis of astronomical objects are spectrographs and spectrometers, where light is split into its different frequency components and individual intensities are recorded, providing, e.g., information about the chemical composition of the source or its motion via Doppler shift. Frequency resolution, stability, and dynamic range are important parameters characterizing such devices. For compact spectrographs, SMFs are ideally used. This issue includes several papers on fiber-fed spectrographs, with Haffert [18] comparing the telescope mode with the fiber mode and discussing the fundamental coupling limit of single-mode integral-field spectroscopy. Anagnos et al. [19] investigate how fiber-fed spectrographs can utilize MCFs with 3D printed micro-lenses for a space-saving solution obtaining high spatial resolution spectroscopy, including a first on-sky test at the Subaru telescope.

While large diffraction gratings or prisms can be used for the dispersion of light, chip-based or integrated spectrographs provide a photonic alternative with a much lower footprint and comparable resolution. Brand et al. [20] present a compact serpentine integrated grating spectrograph with high resolution (resolving power ${R} = 125.000$), based on a 2D dispersive serpentine optical phased array, with which they measured spectral lines separated by 12 pm in the NIR. For the manufacturing process of astrophotonic components, precision and quality as well as accessibility and cost are important factors. Gatkine et al. [21] discuss these for arrayed waveguide gratings on silicon nitride, considering commercially available multi-project wafer options, and evaluate associated waveguide geometries and tolerances.

For calibration of any spectrograph, a reference emission line spectrum is required, which should ideally have high frequency stability, a fairly even intensity distribution, and a spacing appropriate for the spectrograph. As an alternative to emission spectra of gas lamps, stabilized frequency combs provide an ideal ruler with known intervals between the lines and an absolute frequency reference to an atomic transition. Frequency combs can be realized using different methods. With their properties matching the requirements for astronomy spectrograph calibration, photonic frequency combs—or astro-combs in the context of astronomy—are suitable candidates. Chae et al. [22] present a green astro-comb based on Ti:sapphire and a mode-selective cavity, which has been tested at the high dispersion echelle spectrograph for the Okayama Telescope of the National Astronomical Observatory of Japan. Cheng et al. [23] discuss a frequency comb in the IR based on a Ti:sapphire master comb and an optical parametric oscillator (OPO), for applications such as the high-resolution spectrograph (HIRES) at the ELT. As an alternative to astro-combs, Fabry–Perot etalons can be used to filter broadband light. Tang et al. [24] demonstrate their design, characterization, and thermal performance of a fiber Fabry–Perot etalon-based wavelength calibrator system for high-precision spectroscopy. This issue also includes a different approach from Bonduelle et al. [25], who present the proof of concept of a multiplexed integrated optics Fourier transform spectrometer based on lithium niobate waveguides, where interference fringes are sampled using spatially displaced nanogrooves.

D. Coronagraphy

In some cases, astronomers need to observe faint structures or objects that are many orders of magnitude brighter. The historical origins of coronography began with solar observations, where fascinating structures and events in the corona (such as coronal loops, solar flares, or coronal mass ejections) are usually hidden in the solar glare. A remarkable but imperfect coronograph offered by nature is the Moon during solar eclipses. (Since the Moon is slowly spiraling away from Earth, it is fortuitous that we live at a time when coronography with the Moon’s disk is even possible.) To observe the faint outer regions of the Sun at any time, an opaque coronagraphic mask can be placed on top of the image of the Sun to simply hide the light coming from the bright solar disk [26]. With the advent of exoplanetary science, an additional exciting application emerged for coronagraphy: revealing planets that are more than five orders of magnitude fainter than their host stars. Due to advanced coronagraph designs, coupled with decent AO correction, young giant gaseous planets outside of our solar system have been directly imaged through their emitted light. There are different techniques to build coronagraphs dedicated to high-contrast imaging of exoplanets: using phase or amplitude masks, located in either the focal or pupil plane [27]. For instance, phase masks can be used to cause destructive interference for the central starlight while the light from potential nearby planets would be preserved. The vector-apodizing phase plate (vAPP) [28,29] is such a coronagraph consisting of a phase mask placed in a pupil plane (thus showing high stability to tip–tilt variations). A vAPP can be manufactured with high precision by using liquid-crystal-based technology. This issue includes a review by Doelman et al. [30] about the design process, development, commissioning, on-sky performance, and scientific results of a vAPP coronagraph manufactured with liquid crystals. The design of phase masks is very complex. Wong et al. [31] use optimized automatic differentiation tools to apply gradient descent methods to phase retrieval and phase mask design. Among other examples, the authors present simulations in which their code is used to optimize coronagraphs for direct imaging of exoplanets, achieving results that are comparable to existing state-of-the-art solutions. The problem of phase retrieval is universal in nature, and the code, which is made available to the community, can be used for a much larger range of imaging applications and the design of astrophotonic components.

E. Interferometry and High Angular Resolution Imaging

Interferometry at optical and IR wavelengths enables high-resolution imaging of astronomical objects by combining the beams of telescopes physically separated by tens or hundreds of meters, e.g., in an array such as CHARA or VLTI. At a given wavelength, the angular resolution provided by an interferometer depends on the physical distance between the telescopes, whereas it depends on the size of the primary mirror for a single dish telescope. In the visible and IR, the angular resolution of current interferometers surpasses those of single dish telescopes by a factor of 10 to 40. Interferometry has delivered historical scientific data from the onset, when the diameter of a star was directly measured for the first time [32], to images of convective cells on the stellar surface of red giants [33] or to recent µas astrometric measurements of stellar orbits around the supermassive black hole in the galactic center with GRAVITY at VLTI [34], which led to the 2020 Nobel Prize in Physics. In contrast to radio very long baseline interferometry (VLBI), the high frequency of optical and IR light prevents direct detection of the phase. Therefore, interferometric measurements require the overlap of light—by either directly combining light from two or more telescopes or overlapping telescope light with a local oscillator in heterodyne interferometry [35], imprinting the phase on the resulting radio frequency, which is then correlated. While free-space beam combining optics quickly scale up in volume with each interferometric baseline, e.g., when more telescopes are included in the interferometer, photonic beam combiners can overlap light in a compact device. Research is ongoing to extend, e.g., available wavelength range, number of baselines, throughput, and stability of photonic beam combiners. Benoit et al. [36] have designed and experimentally tested an interferometric beam combiner laser-written in standard, commercial infrasil glass to operate as a two-beam interferometer in the astronomical K-band (2.2 µm) at the CHARA array (USA), using asymmetrical cores for an achromatic near 50/50 coupling ratio. This further demonstrates the versatility of the ultrafast laser inscription [37] platform to process a large variety of glass substrates. With the beam combiner being at the heart of an astronomical interferometer, photonic technologies allow exploration of different architectures: taking advantage of the 3D processing capabilities of laser writing, Nayak et al. [38] tested for the first time a prototype discrete beam combiner (DBC) at the William Herschel Telescope on the bright stars Vega and Altair. For future high-resolution imaging with large telescopes at the diffraction limit by means of aperture masking, Cvetojevic et al. [39] describe a method of building hybridized 28-baseline pupil-remapping photonic interferometers. This photonic device is capable of performing aperture masking with eight subapertures of a large monolithic or segmented telescope, implementing ABCD pairwise beam combination.

Similar to a coronagraph, the method of nulling interferometry makes use of destructive interference between paths to remove the signal from a bright object and uncover a fainter signal next to the bright source. With high enough dynamic range, this method is a promising candidate for direct imaging of Earth-like exoplanets, which are around 10 orders of magnitude fainter than their central star. With design and simulations for a tricoupler for GLINT, utilizing the third dimension available with laser writing for equal distances between the waveguides, Martinod et al. [40] present a beam combiner concept that has fringe tracking capabilities included on the chip. With such combined functionality, much fainter signals from orbiting planets could be detected.

Space-based interferometers are a logical next step for extremely long baselines without wavefront distortions from atmospheric turbulence, but several technological challenges have yet to be overcome. Worldwide, research groups are working on precise and reliable technology that will finally enable such an endeavor. For all space-based missions, technologies to reduce the size, weight, and power requirements of the instrument are required. Due to their small footprint and monolithic design, photonic devices are predestined for such applications, but they cannot replace all optical components. For lenses, alternatives such as diffractive lenses can provide solutions, and Wang et al. [41] design a high-harmonic diffractive lens color compensation system for space applications.

Quantum physics is an integral part of astronomy, first on the fundamental level in terms of photon statistics and telescope resolution and, second, in terms of the potential of using quantum technology and quantum photonics for astronomy applications. In [42], Gumpel et al. continue their investigation of quantum amplification methods for super-resolution (i.e., below the diffraction limit) imaging. By amplifying, the momentum and thus the position can be extracted from multiple instead of a single photon. Spontaneous emissions add significant noise, which have to be measured and subtracted from the amplified signal. In addition, only amplification above a certain threshold is used, which limits the relative noise contribution at the cost of losing sensitivity—a trade-off required to obey the Heisenberg uncertainty limit. Here, the authors use centimeter-scale components made from dye in solid polymer to amplify broadband light by means of stimulated emission. While super-resolution has not yet been achieved in this non-diffraction-limited setup, an improvement of spatial resolution is experimentally shown and holds potential for further optimization and application in high-resolution imaging.

For interferometry in particular, quantum technology could play an important role. Bland-Hawthorn et al. [43] present an exciting solution for transporting light from telescopes to a central processing center for beam combination as part of a very long baseline interferometer. Here, the possibility of quantum-memory-based, quantum hard drives to record and transport the phase and amplitude information of the incoming light is assessed. Analogous to Young’s double-slit interferometer, photons have to be stored without knowledge of their properties so as not to destroy the interference pattern—this is possible only with quantum devices. This paper is one of several examples demonstrating how progress in astrophotonics and in quantum technology can be mutually beneficial, and we foresee a closer cooperation between these fields in future.

3. SUMMARY

With this joint feature issue of JOSA B and Applied Optics, we include papers that present a range of technological advances in astrophotonics, from simulation during an early design stage to experimental results and on-sky tests. Astrophotonics requires interdisciplinary solutions, including input from manufacturing methods and material considerations, fiber optics and detectors, quantum physics, space technology, and of course astronomy. The presented work will surely find its way into existing telescope systems to upgrade instruments and improve their science output, as well as into future instruments for next-generation telescopes on ground and in space.