Photonic ring resonators used as wavelength notch filters are a promising novel solution to enable astronomical instruments to remove the signal from atmospheric OH emission in the near-infrared wavelength range. We derive design requirements from theory and finite difference time domain simulations. We find rings with radii less than 10 microns provide an adequate free spectral range for silicon nitride abd less than 3 microns for silicon. One challenge for this application is the requirement for many rings in series to suppress particular wavelengths within 0.2nm. We report progress in fabricating both silicon and silicon nitride rings for OH suppression.
Integrated optics has the potential to play a transformative role in astronomical instrumentation. It has already made a significant impact in the field of optical interferometry, through the use of planar waveguide arrays for beam combination and phase-shifting. Additionally, the potential benefits of micro-spectrographs based on array waveguide gratings have also been demonstrated.
Here we examine a new application of integrated optics, using ring resonators as notch filters to remove the signal from atmospheric OH emission lines from astronomical spectra. We also briefly discuss their use as frequency combs for wavelength calibration and as drop filters for Doppler planet searches. We discuss the theoretical requirements for ring resonators for OH suppression. We find that small radius (< 10 μm), high index contrast (Si or Si3N4) rings are necessary to provide an adequate free spectral range. The suppression depth, resolving power, and throughput for efficient OH suppression can be realised with critically coupled rings with high self-coupling coefficients.
We report on preliminary laboratory tests of our Si and Si3N4 rings and give details of their fabrication. We demonstrate high self-coupling coefficients (> 0:9) and good control over the free spectral range and wavelength separation of multi-ring devices. Current devices have Q ≈ 4000 and ≈ 10 dB suppression, which should be improved through further optimisation of the coupling coefficients. The overall prospects for the use of ring resonators in astronomical instruments is promising, provided efficient fibre-chip coupling can be achieved.
ULTIMATE is an instrument concept under development at the AAO, for the Subaru Telescope, which will have the unique combination of ground layer adaptive optics feeding multiple deployable integral field units. This will allow ULTIMATE to probe unexplored parameter space, enabling science cases such as the evolution of galaxies at z ~ 0:5 to 1.5, and the dark matter content of the inner part of our Galaxy. ULTIMATE will use Starbugs to position between 7 and 13 IFUs over a 14 × 8 arcmin field-of-view, pro- vided by a new wide-field corrector. All Starbugs can be positioned simultaneously, to an accuracy of better than 5 milli-arcsec within the typical slew-time of the telescope, allowing for very efficient re-configuration between observations. The IFUs will feed either the near-infrared nuMOIRCS or the visible/ near-infrared PFS spectrographs, or both. Future possible upgrades include the possibility of purpose built spectrographs and incorporating OH suppression using fibre Bragg gratings. We describe the science case and resulting design requirements, the baseline instrument concept, and the expected performance of the instrument.
Atmospheric emission from OH molecules is a long standing problem for near-infrared astronomy. PRAXIS is a unique spectrograph, currently in the build-phase, which is fed by a fibre array that removes the OH background. The OH suppression is achieved with fibre Bragg gratings, which were tested successfully on the GNOSIS instrument. PRAXIS will use the same fibre Bragg gratings as GNOSIS in the first implementation, and new, less expensive and more efficient, multicore fibre Bragg gratings in the second implementation. The OH lines are suppressed by a factor of ~1000, and the expected increase in the signal-to-noise in the interline regions compared to GNOSIS is a factor of ~ 9 with the GNOSIS gratings and a factor of ~ 17 with the new gratings. PRAXIS will enable the full exploitation of OH suppression for the first time, which was not achieved by GNOSIS due to high thermal emission, low spectrograph transmission, and detector noise. PRAXIS will have extremely low thermal emission, through the cooling of all significantly emitting parts, including the fore-optics, the fibre Bragg gratings, a long length of fibre, and a fibre slit, and an optical design that minimises leaks of thermal emission from outside the spectrograph. PRAXIS will achieve low detector noise through the use of a Hawaii-2RG detector, and a high throughput through an efficient VPH based spectrograph. The scientific aims of the instrument are to determine the absolute level of the interline continuum and to enable observations of individual objects via an IFU. PRAXIS will first be installed on the AAT, then later on an 8m class telescope.
For the past forty years, optical fibres have found widespread use in ground-based and space-based instruments. In most applications, these fibres are used in conjunction with conventional optics to transport light. But photonics offers a huge range of optical manipulations beyond light transport that were rarely exploited before 2001. The fundamental obstacle to the broader use of photonics is the difficulty of achieving photonic action in a multimode fibre. The first step towards a general solution was the invention of the photonic lantern1 in 2004 and the delivery of high-efficiency devices (< 1 dB loss) five years on2. Multicore fibres (MCF), used in conjunction with lanterns, are now enabling an even bigger leap towards multimode photonics. Until recently, the single-moded cores in MCFs were not sufficiently uniform to achieve telecom (SMF-28) performance. Now that high-quality MCFs have been realized, we turn our attention to printing complex functions (e.g. Bragg gratings for OH suppression) into their N cores. Our first work in this direction used a Mach-Zehnder interferometer (near-field phase mask) but this approach was only adequate for N=7 MCFs as measured by the grating uniformity3. We have now built a Sagnac interferometer that gives a three-fold increase in the depth of field sufficient to print across N ≥ 127 cores. We achieved first light this year with our 500mW Sabre FRED laser. These are sophisticated and complex interferometers. We report on our progress to date and summarize our first-year goals which include multimode OH suppression fibres for the Anglo-Australian Telescope/PRAXIS instrument and the Discovery Channel Telescope/MOHSIS instrument under development at the University of Maryland.
Astrophotonics is a rapidly developing area of research which applies photonic technology to astronomical instrumentation. Such technology has the capability of significantly improving the sensitivity, calibration and stability of astronomical instruments, or indeed providing novel capabilities which are not possible using classical optics. We review the development and application of speciality fibres for astronomy, including multi-mode to single-mode converters, notch filters and frequency combs.In particular we focus on our development of instruments designed to filter atmospheric emission lines to enable much deeper spectroscopic observations in the near-infrared. These instruments employ two novel photonic technologies. First, we have developed complex aperiodic fibre Bragg gratings which filter over 100 irregularly spaced wavelengths in a single device, covering a bandwidth of over 200 nm. However, astronomical instruments require highly multi-mode fibres to enable sufficient coupling into the fibre, since atmospheric turbulence heavily distorts the wavefront. But photonic technologies such as fibre Bragg gratings, require single mode fibres. This problem is solved by the photonic lantern, which enables efficient coupling from a multi-mode fibre to an array of single-mode fibres and vice versa. We present the results of laboratory tests of these technologies and of on-sky experiments made using the first instruments to deploy these technologies on a telescope. These tests show that the fibre Bragg gratings suppress the night sky background by a factor of 9. Current instruments are limited by thermal and detector emission. Planned instruments should improve the background suppression even further, by optimising the design of the spectrograph for the properties of the photonic components. Finally we review ongoing research in astrophotonics, including multi-moded multicore fibre Bragg gratings, which enable multiple gratings to be written into the same device simultaneously, femtosecond direct-write photonic lanterns and Bragg gratings, and complex notch filters and frequency combs using microring resonators, and plans for future astrophotonic instrumentation.
All spectrographs unavoidably scatter light. Scattering in the spectral direction is problematic for sky subtraction, since atmospheric spectral lines are blurred. Scattering in the spatial direction is problematic for fibre fed spectrographs, since it limits how closely fibres can be packed together. We investigate the nature of this scattering and show that the scattering wings have both a Lorentzian component, and a shallower (1/r) component. We investigate the causes of this from a theoretical perspective, and argue that for the spectral PSF the Lorentzian wings are in part due to the profile of the illumination of the pupil of the spectrograph onto the diffraction grating, whereas the shallower component is from bulk scattering. We then investigate ways to mitigate the diffractive scattering by apodising the pupil. In the ideal case of a Gaussian apodised pupil, the scattering can be significantly improved. Finally we look at realistic models of the spectrograph pupils of fibre fed spectrographs with a centrally obstructed telescope, and show that it is possible to apodise the pupil through non-telecentric injection into the fibre.
PRAXIS is a second generation instrument that follows on from GNOSIS, which was the first instrument using fibre
Bragg gratings for OH suppression to be deployed on a telescope. The Bragg gratings reflect the NIR OH lines while
being transparent to the light between the lines. This gives in principle a much higher signal-noise ratio at low resolution
spectroscopy but also at higher resolutions by removing the scattered wings of the OH lines. The specifications call for
high throughput and very low thermal and detector noise so that PRAXIS will remain sky noise limited even with the
low sky background levels remaining after OH suppression. The optical and mechanical designs are presented. The
optical train starts with fore-optics that image the telescope focal plane on an IFU which has 19 hexagonal microlenses
each feeding a multi-mode fibre. Seven of these fibres are attached to a fibre Bragg grating OH suppression system while
the others are reference/acquisition fibres. The light from each of the seven OH suppression fibres is then split by a
photonic lantern into many single mode fibres where the Bragg gratings are imprinted. Another lantern recombines the
light from the single mode fibres into a multi-mode fibre. A trade-off was made in the design of the IFU between field of
view and transmission to maximize the signal-noise ratio for observations of faint, compact objects under typical seeing.
GNOSIS used the pre-existing IRIS2 spectrograph while PRAXIS will use a new spectrograph specifically designed for
the fibre Bragg grating OH suppression and optimised for 1.47 μm to 1.7 μm (it can also be used in the 1.09 μm to 1.26
μm band by changing the grating and refocussing). This results in a significantly higher transmission due to high
efficiency coatings, a VPH grating at low incident angle and optimized for our small bandwidth, and low absorption
glasses. The detector noise will also be lower thanks to the use of a current generation HAWAII-2RG detector.
Throughout the PRAXIS design, from the fore-optics to the detector enclosure, special care was taken at every step along
the optical path to reduce thermal emission or stop it leaking into the system. The spectrograph design itself was
particularly challenging in this aspect because practical constraints required that the detector and the spectrograph
enclosures be physically separate with air at ambient temperature between them. At present, the instrument uses the
GNOSIS fibre Bragg grating OH suppression unit. We intend to soon use a new OH suppression unit based on multicore
fibre Bragg gratings which will allow an increased field of view per fibre. Theoretical calculations show that the gain in
interline sky background signal-noise ratio over GNOSIS may very well be as high as 9 with the GNOSIS OH
suppression unit and 17 with the multicore fibre OH suppression unit.
Photonic lanterns are an important enabling technology for astrophotonics with a wide range of potential applications including fibre Bragg grating OH suppression, integrated photonic spectrographs and fibre scramblers for high resolution spectroscopy. The behaviour of photonic lanterns differs in several important respects from the conventional fibre systems more frequently used in astronomical instruments and a detailed understanding of this behaviour is required in order to make the most effective use of this promising technology. To this end we have undertaken a laboratory study of photonic lanterns with the aim of developing an empirical model for the mapping from input to output illumination distributions. We have measured overall transmission and near field output light distributions as a function of input angle of incidence for photonic lanterns with between 19 and 61 cores. We present the results of this work, highlight the key differences between photonic lanterns and conventional fibres, and illustrate the implications for instrument design via a case study, the design of the PRAXIS spectrograph. The empirical photonic lantern model was incorporated into an end-to-end PRAXIS performance model which was used to optimise the design parameters of the instrument. We describe the methods used and the resulting conclusions. The details of photonic lantern behaviour proved particularly important in selecting the optimum on sky field of view per fibre and in modelling of the instrument thermal background.
The KOALA optical fibre feed for the AAOmega spectrograph has been commissioned at the Anglo-Australian
Telescope. The instrument samples the reimaged telescope focal plane at two scales: 1.23 arcsec and 0.70 arcsec per
image slicing hexagonal lenslet over a 49x27 and 28x15 arcsec field of view respectively. The integral field unit consists
of 2D hexagonal and circular lenslet arrays coupling light into 1000 fibres with 100 micron core diameter. The fibre run
is over 35m long connecting the telescope Cassegrain focus with the bench mounted spectrograph room where all fibres
are reformatted into a one-dimensional slit. Design and assembly of the KOALA components, engineering challenges
encountered, and commissioning results are discussed.
KOALA, the Kilofibre Optimised Astronomical Lenslet Array, is a wide-field, high efficiency integral field unit
being designed for use with the bench mounted AAOmega spectrograph on the AAT. KOALA will have 1000
fibres in a rectangular array with a selectable field of view of either 1390 or 430 sq. arcseconds with a spatial
sampling of 1.25" or 0.7" respectively. To achieve this KOALA will use a telecentric double lenslet array with
interchangeable fore-optics. The IFU will feed AAOmega via a 31m fibre run. The efficiency of KOALA is
expected to be ≈ 52% at 3700A and ≈ 66% at 6563°Å with a throughput of > 52% over the entire wavelength
GNOSIS has provided the first on-telescope demonstration of a concept to utilize complex aperioidc fiber Bragg
gratings to suppress the 103 brightest atmospheric hydroxyl emission doublets between 1.47-1.7 μm. The unit is
designed to be used at the 3.9-meter Anglo-Australian Telescope (AAT) feeding the IRIS2 spectrograph. Unlike
previous atmospheric suppression techniques GNOSIS suppresses the lines before dispersion. We present the
results of laboratory and on-sky tests from instrument commissioning. These tests reveal excellent suppression
performance by the gratings and high inter-notch throughput, which combine to produce high fidelity OH-free
Ground based near-infrared observations have long been plagued by poor sensitivity when compared to visible
observations as a result of the bright narrow line emission from atmospheric OH molecules. The GNOSIS instrument
recently commissioned at the Australian Astronomical Observatory uses Photonic Lanterns in combination with
individually printed single mode fibre Bragg gratings to filter out the brightest OH-emission lines between 1.47 and
1.70μm. GNOSIS, reported in a separate paper in this conference, demonstrates excellent OH-suppression, providing
very “clean” filtering of the lines. It represents a major step forward in the goal to improve the sensitivity of ground
based near-infrared observation to that possible at visible wavelengths, however, the filter units are relatively bulky and
costly to produce.
The 2nd generation fibre OH-Suppression filters based on multicore fibres are currently under development. The
development aims to produce high quality, cost effective, compact and robust OH-Suppression units in a single optical
fibre with numerous isolated single mode cores that replicate the function and performance of the current generation of
“conventional” photonic lantern based devices. In this paper we present the early results from the multicore fibre
development and multicore fibre Bragg grating imprinting process.
Ring resonators are a looped waveguide coupled to an input and an output waveguide. They can be used
to filter, and drop, a series of wavelengths at the resonant frequencies of the ring. Both these properties are
useful for astronomical applications. The dropped signal provides a frequency comb that can be used to provide
accurate wavelength calibration. The free spectral range of such a device is larger than that from a laser comb,
removing the requirement to perform subsequent filtering. The filtered signal could be used to suppress specific
wavelengths, e.g. corresponding to atmospheric emission lines. We present the expected performance of devices
designed for both applications and discuss their advantages and limitations.
Fibre Bragg grating (FBG) OH suppression is capable of greatly reducing the bright sky background seen by near infrared
spectrographs. By filtering out the airglow emission lines at high resolution before the light enters the spectrograph this
technique prevents scattering from the emission lines into interline regions, thereby reducing the background at all wavelengths.
In order to take full advantage of this sky background reduction the spectrograph must have very low instrumental
backgrounds so that it remains sky noise limited. Both simulations and real world experience with the prototype GNOSIS
system show that existing spectrographs, designed for higher sky background levels, will be unable to fully exploit the sky
background reduction. We therefore propose PRAXIS, a spectrograph optimised specifically for this purpose.
The PRAXIS concept is a fibre fed, fully cryogenic, fixed format spectrograph for the J and H-bands. Dark current
will be minimised by using the best of the latest generation of NIR detectors while thermal backgrounds will be reduced
by the use of a cryogenic fibre slit. Optimised spectral formats and the use of high throughput volume phase holographic
gratings will further enhance sensitivity. Our proposal is for a modular system, incorporating exchangeable fore-optics
units, integral field units and OH suppression units, to allow PRAXIS to operate as a visitor instrument on any large
telescope and enable new developments in FBG OH suppression to be incorporated as they become available. As a high
performance fibre fed spectrograph PRAXIS could also serve as a testbed for other astrophotonic technologies.
GNOSIS is an OH suppression unit to be used in conjunction with existing spectrographs. The OH suppression
is achieved using fibre Bragg gratings (FBGs), and will deliver the darkest near-infrared background of any
ground-based instrument. Laboratory and on-sky tests demonstrate that FBGs can suppress OH lines by 30dB
whilst maintaing > 90% throughput between the lines, resulting in a 4 mag decrease in the background.
In the first implementation GNOSIS will feed IRIS2 on the AAT. It will consist of a seven element lenslet
array, covering 1.4" on the sky, and will suppress the 103 brightest OH lines between 1.47 and 1.70 μm. Future
upgrades will include J-band suppression and implementation on an 8m telescope.
ERASMUS-F is a pathfinder study for a possible E-ELT 3D-instrumentation, funded by the German Ministry for
Education and Research (BMBF). The study investigates the feasibility to combine a broadband optical spectrograph
with a new generation of multi-object deployable fibre bundles. The baseline approach is to modify the spectrograph of
the Multi-Unit Spectroscopic Explorer (MUSE), which is a VLT integral-field instrument using slicers, with a fibre-fed
input. Taking advantage of recent developments in astrophotonics, it is planed to equip such an instrument with fused
fibre bundles (hexabundles) that offer larger filling factors than dense-packed classical fibres.
The overall project involves an optical and mechanical design study, the specifications of a software package for 3Dspectrophotometry,
based upon the experiences with the P3d Data Reduction Software and an investigation of the
science case for such an instrument. As a proof-of-concept, the study also involves a pathfinder instrument for the VLT,
called the FIREBALL project.
We demonstrate for the first time an imaging fibre bundle ("hexabundle") that is suitable for low-light applications in
astronomy. The most successful survey instruments at optical-infrared wavelengths today have obtained data on up to a
million celestial sources using hundreds of multimode fibres at a time fed to multiple spectrographs. But a large fraction
of these sources are spatially extended on the celestial sphere such that a hexabundle would be able to provide
spectroscopic information at many distinct locations across the source. Our goal is to upgrade single-fibre survey
instruments with multimode hexabundles in place of the multimode fibres. We discuss two varieties of hexabundles: (i)
closely packed circular cores allowing the covering fraction to approach the theoretical maximum of 91%; (ii) fused noncircular
cores where the interstitial holes have been removed and the covering fraction approaches 100%. In both cases,
we find that the cladding can be reduced to ~2μm over the short fuse length, well below the conventional ~10λ thickness
employed more generally. We discuss the relative merits of fused/unfused hexabundles in terms of manufacture and
deployment, and present our first on-sky observations.
We present FLEX, an instrument to demonstrate the power of fibre Bragg grating OH suppression. This ground
breaking technology promises great gains in sensitivity for near infrared instrumentation and the time is now
right for a pioneer instrument to prove the effectiveness of the technique. Our proposal is for an adaptive optics
fed integral field unit for an 8 metre class telescope. We envisage a 61 element IFU with 0.22" sampling and a
2.2" field of view. J and H-band OH suppression units would cleanly suppress the atmospheric emission lines,
effectively lowering the sky background by 3 or 4 magnitudes respectively. The capabilities of FLEX will make
it ideal for deep Epoch of Reionisation studies, as well as studies of star formation at z~1-4. To enable rapid
and economical deployment FLEX would use an existing near infrared spectrograph with R ≈ 1000 and employ
facility adaptive optics.
Mapping out stellar families to trace the evolutionary star formation history of the Milky Way requires a spectroscopic facility able to deliver high spectral resolution (R≥30k) with both good wavelength coverage (~400 Ang) and target multiplex advantage (~400 per 2 degree field). Such a facility can survey 1,200,000 bright stars over 10,000 square degrees in about 400 nights with a 4-meter aperture telescope. Presented are the results of a conceptual design study for such a spectrograph, which is under development as the next major instrument for the Anglo-Australian Observatory. The current design (that builds upon the AAOmega system) makes use of a White Pupil collimator and an R3 echelle that is matched to the existing AAOmega cameras. The fibre slit can be reconfigured to illuminate the Pupil relay side of the collimator mirror bypassing the echelle, thus preserving the lower dispersion modes of the AAOmega spectrograph. Other spectrograph options initially considered include use of an anamorphic collimator that reduces the required dispersion to that achievable with VPH grating technology or possible use of a double-pass VPH grating.
AAOmicron is a wide-field, fiber-fed, multi-object, near-infrared spectrograph concept for the Anglo Australian
Telescope (AAT). It is one of a number of instruments concepts (primarily for bright time use) recently considered to
complement the existing instrumentation and in particular the highly popular AAOmega system (primarily dark and grey
time usage). AAOmicron has a two-degree field of view, 240 robotically configured fibers and operates between 0.98
and 1.75μm at a resolution of R~3500. AAOmicron offers a broad suite of applications from the study of low-mass stars,
to determining the structure of the high-redshift Universe. We present an overview of the instrument concept, which is
based heavily on the highly successful AAOmega system, and describe how the AAOmega spectrograph design could be
adapted for near-infrared observations to provide a highly cost effective and scientifically compelling instrument.
FLEX is a concept for a fully OH suppressed near infrared integral field spectrograph, being developed at the AAO.
FLEX will be the first instrument to employ fibre Bragg gratings for OH suppression, a radical new technology which
cleanly suppresses the atmospheric OH emission lines at 30dB whilst maintaining a high overall throughout of ~90%. In
this paper we simulate the expected performance of FLEX, and discuss its impact on the science case. FLEX will
effectively make the near-infrared sky 4 mags fainter in the H band and 3 mags fainter in the J band, offering
unprecedentedly deep views of the near-infrared Universe. The FLEX concept is optimised for the identification of the
sources of first light in the Universe - high redshift galaxies or quasars identified through Lyman-alpha emission or a
Lyman break in the continuum spectrum. As such it will consist of a 2x2" integral field unit, composed of a 61 lenslet
hexagonal array, feeding an existing moderate spectral resolution spectrograph, via an OH-suppression unit. We have
simulated the performance of FLEX and show that it can provide robust identification of galaxies at the epoch of
reionisation. A FLEX-like instrument on an ELT could measure the ionisation and enrichment of the inter-galactic
medium beyond a redshift of 7 via metal absorption lines.
IRIS2 is a near-infrared imager and spectrograph based on a HAWAII1 HgCdTe detector. It provides wide-field (7.7’×7.7’) imaging capabilities at 0.4486”/pixel sampling, long-slit spectroscopy at λ/Δλ≈2400 in each of the J, H and K passbands, and the ability to do multi-object spectroscopy in up to three masks. These multi-slit masks are laser cut, and have been manufactured for both traditional multiple slit work (≈20-40 objects in a 3’×7.4’ field-of-view), multiple slit work in narrow-band filters (≈100 objects in a 5’×7.4’ field-of-view), and micro-hole spectroscopy in narrow-band filters allowing the observation of ≈200 objects in a 5’×7.4’ field.