2005AJ....129....1L

We constrain the X-ray emission properties of Lyman break galaxies (LBGs) at z~3-6 using the ~2 Ms Chandra Deep Field North and ~1 Ms Chandra Deep Field South. Large samples of LBGs were discovered using the Hubble Space Telescope as part of the Great Observatories Origins Deep Survey (GOODS). Deep optical and X-ray imaging over the GOODS fields have allowed us to place the most significant constraints on the X-ray properties of LBGs to date. Mean X-ray properties of 449, 1734, 629, and 247 LBGs with z∼3, 4, 5, and 6, respectively, were determined using stacking techniques. When stacked, we detect X-ray emission from LBGs at z∼3 (∼7 σ) and from an optically bright subset (brightest 25%) of LBGs at z∼4 (∼3 σ); the latter is the highest redshift detection yet for ``normal'' galaxies in the X-ray band. The effective exposure times for these stacked observations are ~0.7 and 0.5 Gs, respectively. The derived average rest-frame 2.0-8.0 keV luminosities are 1.5x10⁴¹ and 1.4x10⁴¹ ergs/s, respectively. X-ray emission from these LBGs is likely due to high-mass X-ray binaries and Type II supernovae; the corresponding star formation rates are ~10-30 M_☉/yr. The X-ray-to-B-band mean luminosity ratio (L_X/L_B) at z∼3 is somewhat elevated with respect to that measured for starburst galaxies in the local universe (significance ∼3 σ). When stacking full samples of LBGs at z∼4, 5, and 6, we do not obtain significant detections (<3 σ) and derive rest-frame 2.0-8.0 keV luminosity upper limits (3 σ) of 0.9, 2.8, and 7.1x10⁴¹ ergs/s, respectively. These upper limits constrain any widespread active galactic nucleus (AGN) activity in these objects to be modest at best. Furthermore, we find that ∼0.5% of our LBGs from z~3 to z~6 are detected individually in the X-ray band. These LBGs have spectral shapes and luminosities characteristic of moderate-power AGNs (e.g., Seyfert galaxies and quasars).