1. Clio: 3-5  m planet-finding AO camera Ari Heinze (Steward Observatory) Collaborators: P. Hinz (Steward), S. Sivanandam (Steward), M. Freed (Optical Sciences), A. Breuninger (Steward) First planet

2. 3-5  m Window High Background –Warm Telescope –Variable and High Sky Flux Atmospheric Throughput High Duty Cycle Readout Large Well-depth Detector Better contrast ratios M-band bump in planets only weakly dependent on age Observing catalogue includes nearby moderate age stars Better AO correction SPIE 2006 - Clio - Page 2 Near/Thermal IR Spectra of known extrasolar giant planets around G-type stars. Mass range from 1-10 M Jup. M-band bump present irrespective of age and mass, though near-IR flux decreases steeply with age. (Burrows et al. 2004)

3. Clio Design Diffraction-limited imaging from H through M-bands 3 imaging modes: f/35 (H and K-bands), f/20 (L and M-bands), and pupil imaging (for alignment of cold- stops) modes 320  256 large well-depth, high throughput InSb array optimized for 3-5  m imaging (Indigo Systems Inc.) Cooled optics (77K), baffling, and cold stops to minimize instrument thermal background Coronographic option built in (have ability to add field and pupil stops and PSF shaping wave plates) SPIE 2006 - Clio - Page 3

4. Clio Schematic SPIE 2006 - Clio - Page 4 Schematic of Clio in f/20 mode with ray trace.

5. Detector Performance Bias Voltage4.0V Temperature56 K Pixel Rate400 kHz Duty Cycle~ 90% Frame Rate  20 Hz QE 3-5  m 0.9 Well Depth 3.3  10 6 e - Dark Current 3.0  10 5 e - /s Read Noise800 e - Gain82-94 e - /DN Linearity1% Detector Characteristics Detector Noise Map SPIE 2006 - Clio - Page 5

6. Observing Strategy To remove high sky background: –Nod the telescope a few arcseconds after each exposure and close AO loop –Subtract nod pairs To improve duty cycle –Coadd batches for ~20 images for every exposure To improve sensitivity –Integrate in the L’ and M-band until sky flux fills ~1/2 of full well Typical integrations: 100 ms (M-band); 1500 ms (L’-band) –Sky background-limited and detector noise is negligible To mitigate long-lived speckles –Keep instrument derotator fixed allowing the sky to rotate –Rotate images by the parallactic angle to correct orientation during processing –Static speckles stay fixed on chip and are blurred out by rotation SPIE 2006 - Clio - Page 7

7. On-sky Performance Reach quoted 10  background-limited detection values at separations greater than 1.5” from the central star (after PSF subtraction and unsharp mask) L’ ~ detect 5 M jup 0.5 Gyr old planet at 10 pc M ~ detect 10 M jup 0.5 Gyr old planet at 10 pc Have better L’ sensitivity –Focus initial search in L’ –M-band follow up observations to obtain color and establish nature of source Very stable PSF –Good subtraction of PSF, better contrast FOV 15.5”  12.4” Plate Scale0.0496”/pix Throughput L’ 0.67 Throughput M 0.42 Emissivity*~10% 10  L’ (1 hr) 16.3 mag 10  M (1 hr) 13.5 mag PSFs of 5 different L’ exposures Clio Observational Parameters SPIE 2006 - Clio - Page 8 S T M

8. Vega in M-band Dust model and predicted planet position (Wilner et al. 2002) Clio 672 s M-band image (unsharp masked) First Light Observations ( ApJ, accepted ) SPIE 2006 - Clio - Page 9

9. Vega Sensitivity 1 arcsec Separation (arcsec) Companion mass limit (5 sigma) Macintosh et al. (ApJ 594, 538) using Keck at K-band Limit from 5 minute observation at M band Expected Planet? Metchev et al. (ApJ 582, 1102) using Palomar at H-band 10 M jup fake planet at 20 AU SPIE 2006 - Clio - Page 10 Vega Planet Sensitivity Comparison Vega Fake Planet

10. GJ450 in L’-band 2 arcsec Stellar properties: M dwarf 1 Gyr (X-ray) 9 pc Observation: 5355 s L’ exposure No detection Sensitivity Test: Monte Carlo simulation – 10  planets Blind test – Recovered all planets reliably GJ450 Exposure with fake 10  planets Nod artifact

11. GJ450 Monte-Carlo Planets Sep (arcsec)L’ magMass (MJ) 0.5112.5328.08 0.5613.3220.55 0.9515.359.85 1.1415.68.96 1.2715.967.66 1.5816.067.40 1.9016.516.05 2.5016.595.89 2.9116.386.44 2.9816.65.87 3.7116.516.05 3.9016.595.88 3.9316.625.83 5.0216.496.11 6.5216.436.29 6.5316.276.78 2.6916.575.91 Continued

12. GJ450 Sensitivity 6 M jup Background limited PSF subtracted Contrast Ratio in L’-band 10  Planet Mass Sensitivity SPIE 2006 - Clio - Page 12

13. PSF Suppression PSF sidelobes are over 7 magnitudes fainter at 3 λ /D away The pattern is stable and can be reliably subtracted off to reach the limit of the sky background PSF suppression is easier at M-band where Strehls are typically 90% SPIE 2006 - Clio - Page 13

14. Science Programs 6 pc M dwarf search [observed 4 stars] M through F star search (emphasis on solar- type stars) [observed 20 of 50 stars] A-type star search [this trimester] White dwarf search [this trimester] Vega search [observed twice] T-dwarf photometry [this trimester] SPIE 2006 - Clio - Page 14

15. Vega, M-band, 2850 sec (April)

16. Vega, M-band, ~1hr (June)

17. ξ Boo, L’, ~1 hr

18. ξ Boo, M, ~1 hr

19. BD+60 1417, L’, ~1hr

22. BD+60 1417, Ks, ~1hr

23. Stay tuned… The first direct image of a mature extrasolar planet orbiting a normal star may be only months away… And may well happen at the MMT.

24. Design Simulations MONTE CARLO SIMULATIONS Take spectra of 100 Myr - 5 Gyr old planets, and consider: Instrument Throughput Atmosphere Sky Background Telescope Emissivity (10%) Strehl Ratio Planet Photon Noise Detector Dark Current/Read Noise Best Filter Bandpass: M-band Best Bandpass S/N Plot. (Freed et al. 2004)

25. Outline Why 3-5  m? Clio Design Detector Performance Clio On-sky Performance Latest Scientific Results Planet-finding Sensitivities Science Program SPIE 2006 - Clio - Page 2