Distortion and attenuation free gain-assisted Superluminal Propagtion in a Rare-Earth Doped Crystal

Abstract: A light pulse that travels through a narrow spectral window, i.e. a frequency range of low absorption, within a strongly absorbing frequency range will be delayed due to strong dispersion. If however, the absorbing structure is inverted, such that it is amplifying instead, one will see that a pulse travelling through this window will have its pulse peak come out of the material before it is sent in. This phenomenon is called superluminal propagation, light with negative group velocity or fast light. In this thesis negative group velocity of light through anomalous dispersion of a Europium-doped YSO crystal was studied. A spectral window of low absorption was created in the inhomogeneous profile. Within this spectral window, an absorbing structure is made up of two identical, equidistant from the centre frequency peaks (ω = ω0±ωshift). The structure was composed of ions with strong oscillator strengths, collected through reshuffling ions to different hyperfine levels. This allowed for the collection of a total absorption that exceeds the natural level near the centre frequency. Between the two absorption peaks a region of strong dispersion exists, the magnitude of the dispersion here is far higher in between the peaks than elsewhere in the created spectral window. This structure is then inverted to get a gain structure instead of an absorbing structure. This changes the sign of the dispersion and hence created a region of strong anomalous dispersion and allowed for the propagation of a Gaussian pulse with negative group velocity. This method is unique in that it does not alter the pulse shape or cause significant attenuation. Furthermore, this method allows for the isolated generation of a fast light pulse without the simultaneous generation of a slow light pulse. Through Maxwell-Bloch simulations and using estimates of the experimental conditions an optimal recipe was determined to create the maximal pulse advancement with respect to its time-bandwidth product. The simulations for the maximal width of the pit gave a frequency width of 29 MHz and the optimal width of the two gain structures was determined to be 1 MHz with a hole of 0.3 MHz in between them. The optimal αL from simulation was 12. However, due to experimental factors, an absorption of αL = 4.8 had to be used. This allowed for a Gaussian propagation pulse with tFWHM = 5 μs. Experimentally a group refractive index of ng ≈-10135 was achieved and a pulse advancement of 10-14.2% of the tFHWM, probe pulses had little to no distortion and little to no attenuation/amplification when remaining within the linear regime. A non linear relationship for the optimal αL - tFWHM seems to exist.