We report on recent progress in the development of our focal plane imaging system for the detection and characterization of small fabrication errors in diffraction gratings. The instrument uses a purpose-designed high dynamic range imaging method in conjunction with a low-cost digital camera to acquire images with a dynamic range that can now exceed eight orders of magnitude. The sensitivity and utility of the instrument is demonstrated with measurements of three different diffraction gratings. Avenues for further possible improvements of the instrument are discussed.
Silicon immersion gratings will allow the Giant Magellan Telescope Near-IR Spectrograph (GMTNIRS) to achieve continuous coverage over the entire J, H, K, L and M photometric bands with resolution R~65,000 at J, H and K and R~80,000 at L and M. Gratings for J, H and K will be blazed at R3, while the L and M gratings will be blazed at R4 to achieve the desired resolution. The higher blaze angle of the L and M gratings requires that we use 150mm diameter substrates rather than the 100mm substrates that our standard process was built for. In order to accommodate the larger substrates we have constructed a custom UV exposure system for contact printing of grating lines, and constructed fixtures for coating and etching of the larger substrates. These updates to our process have resulted in the successful production of a grating for the GMTNIRS M-band.
Silicon immersion gratings take advantage of the high index of refraction of silicon (3.4) to significantly improve the performance and reduce the volume of near-infrared spectrographs. The immersion gratings we discuss here are produced by contact photolithography. Lithography is followed by plasma etching of a silicon nitride hard mask, which defines the pattern for wet etching of the silicon v-grooves in potassium hydroxide that form the blazed grating. We have shown that interference fringes between the photomask and the polished silicon nitride on silicon substrate produce a phase error in the completed grating. With our standard process, the lines in photoresist formed during the lithography step have a slope with an additional “foot” at the base of the line. The thickness of this foot can vary and may be partially etched away causing a shift in the position of the line during etching. To reduce the effect of the foot, we have added a plasma etch step designed to remove the foot prior to completing the silicon nitride etch. We have also found that thinning the photoresist to better control the profile formed during contact printing and subsequent etching results in very uniform gratings over a 125 mm grating length. We will also describe a method to predict the phase uniformity at the patterning stage, which allows us to pattern and evaluate the potential grating before etching, saving both time and material costs.
The Immersion GRating INfrared Spectrometer (IGRINS) was designed for high-throughput with the expectation of being a visitor instrument at progressively larger observing facilities. IGRINS achieves R∼45000 and > 20,000 resolution elements spanning the H and K bands (1.45-2.5μm) by employing a silicon immersion grating as the primary disperser and volume-phase holographic gratings as cross-dispersers. After commissioning on the 2.7 meter Harlan J. Smith Telescope at McDonald Observatory, the instrument had more than 350 scheduled nights in the first two years. With a fixed format echellogram and no cryogenic mechanisms, spectra produced by IGRINS at different facilities have nearly identical formats. The first host facility for IGRINS was Lowell Observatory’s 4.3-meter Discovery Channel Telescope (DCT). For the DCT a three-element fore-optic assembly was designed to be mounted in front of the cryostat window and convert the f/6.1 telescope beam to the f/8.8 beam required by the default IGRINS input optics. The larger collecting area and more reliable pointing and tracking of the DCT improved the faint limit of IGRINS, relative to the McDonald 2.7-meter, by ∼1 magnitude. The Gemini South 8.1-meter telescope was the second facility for IGRINS to visit. The focal ratio for Gemini is f/16, which required a swap of the four-element input optics assembly inside the IGRINS cryostat. At Gemini, observers have access to many southern-sky targets and an additional gain of ∼1.5 magnitudes compared to IGRINS at the DCT. Additional adjustments to IGRINS include instrument mounts for each facility, a glycol cooled electronics rack, and software modifications. Here we present instrument modifications, report on the success and challenges of being a visitor instrument, and highlight the science output of the instrument after four years and 699 nights on sky. The successful design and adaptation of IGRINS for various facilities make it a reliable forerunner for GMTNIRS, which we now anticipate commissioning on one of the 6.5 meter Magellan telescopes prior to the completion of the Giant Magellan Telescope.
The Immersion Grating Infrared Spectrometer (IGRINS) is a revolutionary instrument that exploits broad spectral coverage at high-resolution in the near-infrared. IGRINS employs a silicon immersion grating as the primary disperser, and volume-phase holographic gratings cross-disperse the H and K bands onto Teledyne Hawaii-2RG arrays. The use of an immersion grating facilitates a compact cryostat while providing simultaneous wavelength coverage from 1.45 - 2.5 μm. There are no cryogenic mechanisms in IGRINS and its high-throughput design maximizes sensitivity. IGRINS on the 2.7 meter Harlan J. Smith Telescope at McDonald Observatory is nearly as sensitive as CRIRES at the 8 meter Very Large Telescope. However, IGRINS at R≈45,000 has more than 30 times the spectral grasp of CRIRES* in a single exposure. Here we summarize the performance of IGRINS from the first 300 nights of science since commissioning in summer 2014. IGRINS observers have targeted solar system objects like Pluto and Ceres, comets, nearby young stars, star forming regions like Taurus and Ophiuchus, the interstellar medium, photo dissociation regions, the Galactic Center, planetary nebulae, galaxy cores and super novae. The rich near-infrared spectra of these objects motivate unique science cases, and provide information on instrument performance. There are more than ten submitted IGRINS papers and dozens more in preparation. With IGRINS on a 2.7m telescope we realize signal-to-noise ratios greater than 100 for K=10.3 magnitude sources in one hour of exposure time. Although IGRINS is Cassegrain mounted, instrument flexure is sub-pixel thanks to the compact design. Detector characteristics and stability have been tested regularly, allowing us to adjust the instrument operation and improve science quality. A wide variety of science programs motivate new tools for analyzing high-resolution spectra including multiplexed spectral extraction, atmospheric model fitting, rotation and radial velocity, unique line identification, and circumstellar disk modeling. Here we discuss details of instrument performance, summarize early science results, and show the characteristics of IGRINS as a versatile near-infrared spectrograph and forerunner of future silicon immersion grating spectrographs like iSHELL<sup>2</sup> and GMTNIRS.<sup>3</sup>
Silicon immersion gratings make near-IR spectrographs compact and allow them to have continuous wavelength coverage over a large bandwidth. We have produced an exceptional silicon immersion grating that approaches optical perfection in terms of surface error. This grating has a peak-to-valley error of 79 nm over a 25 mm beam, which exceeds the 85 nm requirement to have λ/4 peak-to-valley error at the shortest wavelength where silicon immersion gratings can be used. In order to reduce the level of large-scale errors we have honed our contact printing method by optimizing our UV exposure system, introducing additional process checks and inspections and carefully evaluating large-scale errors in the gratings produced.
We have explored a number of lithographic techniques and improvements to produce the resist lines that then define the grating groove edges of silicon immersion gratings. In addition to our lithographic process using contact printing with photomasks, which is our primary technique for the production of immersion gratings, we explored two alternative fabrication methods, direct-write electron beam and photo-lithography. We have investigated the application of antireflection (AR) coatings during our contact printing lithography method to reduce the effect of Fizeau fringes produced by the contact of the photomask on the photoresist surface. This AR coating reduces the amplitude of the periodic errors by a factor of 1.5. Electron beam (e-beam) patterning allows us to manufacture gratings that can be used in first order, with groove spacing down to 0.5 micrometer or smaller (2,000 grooves/mm), but could require significant e-beam write times of up to one week to pattern a full-sized grating. The University of Texas at Austin silicon diffractive optics group is working with Jet Propulsion Laboratory to develop an alternate e-beam method that employs chromium liftoff to reduce the write time by a factor of 10. We are working with the National Institute of Standards and Technology using laser writing to explore the possibility of creating very high quality gratings without the errors introduced during the contact-printing step. Both e-beam and laser patterning bypass the contact photolithography step and directly write the lines in photoresist on our silicon substrates, but require increased cost, time, and process complexity.