PROBING POPULATION III STARS IN GALAXY IOK-1 AT z=6.96 THROUGH He II EMISSION

The He lambda 1640 emission line has been suggested as a direct probe of Population III (Pop III) stars at high redshift, since it can arise from highly energetic ionizing photons associated with hot, metal-free stars. We use the Hubble Space Telescope WFC3/F130N IR narrowband filter to probe He II lambda 1640 emission in galaxy IOK-1 at z = 6.96. The sensitivity of this measurement is greater than or similar to 5x deeper than for previous measurements. From this deep narrowband imaging, combined with broadband observations in the F125W and F160W filters, we find the He II flux to be (1.2 +/- 1.0) x 10 (18) erg s (1) cm (2), corresponding to a 1 sigma upper limit on the Pop III star formation rate (SFR) of similar to 0.5M(circle dot) yr(-1) for the case of a Salpeter initial mass function (IMF) with 50 less than or similar to M/M-circle dot less than or similar to 1000 and mass loss. Given that the broadband measurements can be fit with a UV-continuum spectral flux density of similar to 4.85x10(-10) x lambda(-2.46) erg s(-1) cm(-2) angstrom(-1), which corresponds to an overall SFR of similar to 16(-2.6)(+2.6) M-circle dot yr(-1), massive Pop III stars represent less than or similar to 6% of the total star formation. This measurement places the strongest limit yet on metal-free star formation at high redshift, although the exact conversion from He II luminosity to Pop III SFR is highly uncertain due to the unknown IMF, stellar evolution, and photoionization effects. Although we have not detected He II lambda 1640 at more than the 1.2 sigma level, our work suggests that a greater than or similar to 3 sigma level detection is possible with the James Webb Space Telescope.