Within the SONYC – Substellar Objects in Nearby Young Clusters – survey , we investigate the frequency of free-floating planetary-mass objects ( planemos ) in the young cluster NGC 1333 .
Building upon our extensive previous work , we present spectra for 12 of the faintest candidates from our deep multi-band imaging , plus seven random objects in the same fields , using MOIRCS on Subaru .
We confirm seven new sources as young very low mass objects ( VLMOs ) , with T _ { \mathrm { eff } } of 2400–3100 K and mid-M to early-L spectral types .
These objects add to the growing census of VLMOs in NGC1333 , now totaling 58 .
Three confirmed objects ( one found in this study ) have masses below 15 M _ { \mathrm { Jup } } , according to evolutionary models , thus are likely planemos .
We estimate the total planemo population with 5-15 M _ { \mathrm { Jup } } in NGC1333 is \lesssim 8 .
The mass spectrum in this cluster is well approximated by dN / dM \propto M ^ { - \alpha } , with a single value of \alpha = 0.6 \pm 0.1 for M < 0.6 M _ { \odot } , consistent with other nearby star forming regions , and requires \alpha \lesssim 0.6 in the planemo domain .
Our results in NGC1333 , as well as findings in several other clusters by ourselves and others , confirm that the star formation process extends into the planetary-mass domain , at least down to 6 M _ { \mathrm { Jup } } .
However , given that planemos are 20-50 times less numerous than stars , their contribution to the object number and mass budget in young clusters is negligible .
Our findings disagree strongly with the recent claim from a microlensing study that free-floating planetary-mass objects are twice as common as stars – if the microlensing result is confirmed , those isolated Jupiter-mass objects must have a different origin from brown dwarfs and planemos observed in young clusters .