| We study the optical nonlinearities generated in gallium phosphide as a result of optically pumping 60 meV above the indirect band edge using picosecond pulses at 532 nm. Using various pulsewidths and carrier densities, the physical mechanisms responsible for the nonlinearities are identified.; Fits to the measured data allow us to extract values for the absorption cross section per free-carrier pair ({dollar}1times10sp{lcub}-18{rcub}{dollar} cm{dollar}sp2{dollar}), the index change per free-carrier pair ({dollar}-3.1times10sp{lcub}-22{rcub}{dollar} cm{dollar}sp3{dollar}), and the two-photon absorption coefficient (20 cm/GW). The experiments also reveal a constant indirect absorption coefficient at 532 nm, even for carrier densities in excess of 10{dollar}sp{lcub}19{rcub}{dollar} cm{dollar}sp{lcub}-3{rcub}{dollar}. Consequently, the optically excited carrier densities are accurately known, allowing us to determine the per carrier changes given here. Specifically, the measured per carrier index change is independent of carrier density as predicted by a simple Drude/Lorentz model which neglects screening and renormalization effects. Furthermore, the measured sign and magnitude are consistent with such a model. This is in contrast to theoretical predictions based on the plasma theory of Banyai and Koch. That theory predicts a density dependent per carrier index change dominated by screening and renormalization effects and having a much larger magnitude than is measured here.; The measured two-photon absorption coefficient is an order of magnitude larger than expected for this wavelength, based on a well known two-band model. Explanations for this discrepancy are presented, based on higher lying critical points and Coulomb enhancement effects. Further experiments are suggested to firmly establish the role played by each of these.; One motivation for this work was to identify pulsewidth independent nonlinearities for use in visible wavelength optical limiting for eye protection. This type of behavior is confirmed in an experiment using two different laser pulsewidths. One apparent drawback of this technique is the low damage threshold of this material at such wavelengths. However, we show this damage can be used to complement the limiting response and it provides only negligible image degradation in low f-number optical systems. This identifies a promising technique for eye protection over dynamic ranges in excess of 45 dB. |