2012 SALT Meeting SALT 2 GI SALT Second

2012 SALT Meeting: SALT 2 GI SALT Second Generation Instruments (2 GI): The Half-Offner Spectrograph As a Spectroscopic Building Block Darragh O’Donoghue Southern African Large Telescope in collaboration with Chris Clemens Univ. of North Carolina pg 1

2012 SALT Meeting: SALT 2 GIs Can be one or more items from: • • Additional spectroscopic capability (wider λ, IFU(s) …) Adaptive optics and instruments to match Telescope wider field and instruments to match Specialised instruments (coronography, Fourier Transform Spectrometers. . . ) This is by no means an exhaustive list. Today I am going to talk about a new kind of optical spectrograph which can be used in a wide variety of ways, and present one specific example of how it might be used on SALT. However, the talk is NOT a specific instrument proposal. It is rather to show a new kind of instrument which may be of interest to those with specific science cases which could be met by this instrument along with an appropriate front end. I will certainly be paying close attention to the radio projects to be talked about for the rest of the day. pg 2

2012 SALT Meeting: SALT 2 GI The Horror Of The Modern Optical / NIR Spectrograph On today’s 10 -m class spectrographs, you will find instruments which have / are: • • • Physically large: typically ~2 metre in linear dimension so the mechanics are difficult (need to keep optics in place to 25 microns) Optics are large and heavy: 250 - 300 mm in diameter Numerous elements: 12 – 20 lens elements is commonplace. Many … … lenses are made of Ca. F 2, the crack cocaine of the optical designer: v Expensive v Fragile v Vulnerable to thermal shock or thermal gradient shock The most desperate, UV light-craving addicts will use Ba. F 2 or even Na. Cl as lens substrate material. Antireflection coatings cost an arm and a leg I have a right to criticise: my design of SALTICAM includes many of the above (but has not had optical problems – my explanation for this is “SALTICAM did not 3 pushpgthe limits of the technology and I was lucky”).

2012 SALT Meeting: SALT 2 GI GMOS on Gemini is an example: pg 4 Gemini Multi-Object Spectrograph (GMOS)

2012 SALT Meeting: SALT 2 GI GEMINI Multi-Object Spectrographs (GMOS) Optical train of Gemini Multi-Object Spectrograph pg 5

2012 SALT Meeting: SALT 2 GI GEMINI Multi-Object Spectrographs (GMOS) 1 metre Optical diagram of Gemini Multi-Object Spectrograph pg 6

2012 SALT Meeting: SALT 2 GI IMACS on Magellan 600 mm pg 7

2012 SALT Meeting: SALT 2 GI Physically large with lots of large and delicate glass ~ half the lenses are either Ca. F 2 or Na. Cl pg 8

2012 SALT Meeting: SALT 2 GI … Worse Horror Is To Come … MOBIE: Low/med resolution spectrograph for the 30 Metre Telescope (TMT). pg 9

2012 SALT Meeting: SALT 2 GI I have been thinking for a number of years: “Can it be simpler? ” Offner imaging spectrometer designs seemed like a promising avenue to explore for a solution. pg 10

2012 SALT Meeting: SALT 2 GI 1973 Patent for Abe Offner’s 1 -to-1 Relay: pg 11

2012 SALT Meeting: SALT 2 GI 1973 Patent for Abe Offner’s 1 -to-1 Relay: Folding flat mirror It has EXCELLENT image quality as it Is corrected for all 5 primary optical aberrations Identical concave tertiary mirror Concave spherical primary mirror Convex secondary mirror pg 12

2012 SALT Meeting: SALT 2 GI 1973 Patent for Abe Offner’s 1 -to-1 Relay: Application: lithographic production of printed circuits: Light passing through a scanning mask … pg 13 … is imaged on to a silicon wafer

2012 SALT Meeting: SALT 2 GI Chris Clemens is one of only 3 commercial suppliers of astronomical VPH gratings in the world, having spent 10 years building up a fabrication facility and learning how to make them. Chris and I discussed Offner-type designs for some years. My ongoing question to Chris was: “Can you make a convex VPH grating? ” In 2012 February, his answer was “Yes”. But what are VPH gratings and why are they relevant to Offner spectrometers? pg 14

2012 SALT Meeting: SALT 2 GI Volume Phase Holographic gratings: astronomical dispersers of choice Transmission gratings using DCG (dichromated gelatin) processed after exposure to a laser hologram fringes. This imprints a modulation in the refractive index, as planes of constant n. pg 15

2012 SALT Meeting: SALT 2 GI Tuning VPH gratings for efficiency: the ‘super blaze’ Rotate the grating and the peak efficiency “travels along” the superblaze curve Su pe rb VPH at specific incident angle pg 16 laz ec urv e Tra sur dition face al relie f

2012 SALT Meeting: SALT 2 GI The first curved VPH grating design pg 17

2012 SALT Meeting: SALT 2 GI The first “Half Offner”-type spectrograph design pg 18

2012 SALT Meeting: SALT 2 GI The first “Half Offner”-type spectrograph design pg 19

2012 SALT Meeting: SALT 2 GI Proof Of Concept Prototype: 2012 May-June Design, procurement, commission and try-out in 6 wks • • • 1 x off-the-shelf spherical mirror, 50 mm in diameter, from Optosigma 2 x off-the-shelf meniscus lenses from JML Optical … … turned into a curved VPH grating in Chris’ lab Some optical bench fixturing Detector: white paper (max sensitivity with the lab window blinds down) Later we got hold of a decent CCD camera but we only got eyeball estimates of performance (due to time limits) 20 mm Layout of the mini prototype pg 20 Resulting spectrum

2012 SALT Meeting: SALT 2 GI Developing The Half Offner Into A Real Instrument • Design a full-sized, useful, astronomical instrument of interest for SALT. For the real SALT, the constraints on better performance include: v SALT has moderately poor seeing: 1 arcsec rare; 1. 4 -2. 0” typical. Median seeing was believed to be ~1 arcsec. Image slicing would help enormously. v SALT is slow to set up, and has limited track time so shutter-open time is squeezed. Reducing spectroscopic setup time will help substantially. v Multi-object spectroscopy on SALT is difficult and exacerbates the previous point. v Many SALT proposals are for spectroscopy of single objects. v Many SALT proposals would like to do ultraquick “point and shoot” type spectroscopy (To. Os, transient follow-up or short exposure-many object programs) but are frustrated by poor shutter-open efficiency due to long setup time. pg 21

2012 SALT Meeting: SALT 2 GI Developing The Half Offner Into A Real Instrument • In response to the above, we (Chris Clemens, Gerald Cecil and I) have come up with the following concept for SOAR and SALT for transient follow-up: v IFU-fed for image slicing to beat seeing and minimize setup time. Single IFU to begin with for single object programs, expansion to many for multi-object spectroscopy. Multiple IFUs to be deployable. v Resolution 3000 covering the wavelength range 380 -900 nm v Individual spectra will span 140 - 200 nm. 3 channels covers it all. This is NOT a SALT 2 nd Generation instrument proposal per se. It is a rather a prototyping and characterization design. It may prove to be a competitive SALT 2 GI. Or another variation of the spectrometer as the basic unit, in a design aiming at other science goals. More prototyping work is needed first. • So … the immediate aim is to build a full-scale prototype of the basic spectrometer for laboratory characterization: • To measure image quality and resolution • To measure efficiency • Assess overall instrument performance (ghosting, stray light etc. ) pg 22

2012 SALT Meeting: SALT 2 GI IFU-fed R=3000 Spectrograph • • • This is an R = 3000 spectrograph covering λλ 380 -900 nm Detector is 25 mm in the spectral direction and 10 mm field in the spatial Single spectrum covers ~170 nm. A 3 -channel system does it all in 1 shot pg 23

2012 SALT Meeting: SALT 2 GI R=3000 Spectrograph “Spot diagrams” show instrument has excellent image quality Wavelength (nm): -5 mm Spatial field 0 mm 5 mm pg 24 520 580 630 690 30 microns = 2 pixels on typical CCD

2012 SALT Meeting: SALT 2 GI R=3000 Spectrograph But you promised us a 2 element spectrograph … Two field lenses were added by Offner to his relay in a later patent to widen the field of view. Ours arise for the same reason and yield an even more compact instrument. They are small and made of any glass which transmits the light (BK 7 or, for UV, fused silica – the two most common glasses. No Ca. F 2, Na. Cl or anything fragile and expensive). One version of this design uses 2 off-the-shelf field lenses costing $300. Field lenses pg 25

2012 SALT Meeting: SALT 2 GI R=3000 Spectrograph But … what about the focal reducer? This compresses the scale and speeds up the beam by a factor of 2 to reduce the size of the detector (and the optics). It uses 8 lenses, the largest being 12 mm in size and all the rest under 7 mm. BUT … half the lenses are Ca. F 2. Another focal reducer I have designed is much better: it is half as long and uses 1 glass type (e. g. BK 7 or silica). No Ca. F 2. It may be worth patenting so I don’t want to show it ye. pg 26

2012 SALT Meeting: SALT 2 GI The IFU Gerald Cecil (UNC) has been developing, in collaboration with Josh Bland. Hawthorn (U of Sydney), IFUs positioned by efficient robotic positioners: v “Heavily fused” and “Lightly fused” Hexabundles comprising 61 x 100 micron fibers in a hexagonal packing arrangement. v They currently prefer the lightly fused devices for astronomy: they report 84 per cent fill-factor if etching the ends of the fibers is done. v One can purchase a device now for $12 500 for a device now This technology may need to become more mature. We are currently thinking of, as an alternative (or preferably), a scheme similar to the GMOS IFU by Allington-Smith et al. (2002) which uses lenslets to feed fibers. These will have 100 per cent fill factor. pg 27

2012 SALT Meeting: SALT 2 GI Heavily-fused Hexabundles: FRD problems pg 28

2012 SALT Meeting: SALT 2 GI Lightly-fused Hexabundles: 84 % fill factor pg 29

2012 SALT Meeting: SALT 2 GI Rough concept for lenslet array: 100 % fill factor 4” on sky 100 mu = 0. 45” on sky . . . 9 x 9 lenslet + fiber array fibers lenslets pg 30 81 x 75 micron fibers

2012 SALT Meeting: SALT 2 GI Cost Of The R=3000 Spectrograph Hardware • Optics: v 140 mm spherical mirror v 50 mm curved VPH grating v 2 x 70 mm field lenses v Focal reducer v Optics coatings v Integral field unit $ 3000 $ 7000 $ 5000 $ 10000 • Mechanics: $ 3000 • Electronics & software (grating changer): $ 5000 • Detector & software: v 1 x 25 mm Andor EMCCD: $ 37000 $ 80000 pg 31

2012 SALT Meeting: SALT 2 GI Conclusion The Half Offner is a compelling new kind of spectrograph which is: v Very simple v v Very compact Very inexpensive and might turn into the spectrograph design of choice in the future if its initial promise pans out. One way or another, a full-sized lab prototype of the basic spectrometer (no IFU and no focal reducer) will be built here at SAAO and more quantitative information about its performance should be in hand by the second half of next year. Specific instrument concept proposals should be available by then as well. pg 32

2012 SALT Meeting: SALT 2 GI pg 33

2012 SALT Meeting: SALT 2 GI pg 34