Giant planets are usually thought to form within a few tens of AU of their host stars , and hence it came as a surprise when we found what appeared to be a planetary mass ( \sim 0.008 M _ { \odot } ) companion around the 5 Myr-old solar mass star 1RXS J160929.1-210524 in the Upper Scorpius association .
At the time , we took the object ’ s membership in Upper Scorpius —established from near-infrared , H - and K -band spectroscopy— and its proximity ( 2.2″ , or 330 AU ) to the primary as strong evidence for companionship , but could not verify their common proper motion .
Here , we present follow-up astrometric measurements that confirm that the companion is indeed co-moving with the primary star , which we interpret as evidence that it is a truly bound planetary mass companion .
We also present new J -band spectroscopy and 3.0-3.8 \mu m photometry of the companion .
Based on a comparison with model spectra , these new measurements are consistent with the previous estimate of the companion effective temperature of 1800 \pm 200 K. We present a new estimate of the companion mass based on evolution models and the calculated bolometric luminosity of the companion ; we obtain a value of 0.008 _ { -0.002 } ^ { +0.003 } M _ { \odot } , again consistent with our previous result .
Finally , we present angular differential imaging observations of the system allowing us to rule out additional planets in the system more massive than 1 M _ { Jup } , 2 M _ { Jup } and 8 M _ { Jup } at projected separations larger than 3″ ( \sim 440 AU ) , 0.7″ ( \sim 100 AU ) and 0.35″ ( \sim 50 AU ) , respectively .
This companion is the least massive known to date at such a large orbital distance ; it shows that objects in the planetary mass range exist at orbital separations of several hundred AU , posing a serious challenge for current formation models .