Fig. 1. Normalized output energy of the regenerative amplifier as a function of time over a 3.5-month period.

Fig. 2. Morphology of a damage site on a QWP surface in the Regen cavity. (a) Optical microscopy of the QWP damage site with $1000\times$ magnification; (b) contour map of the QWP damage site.

Fig. 3. Morphology of a damage site on a WG surface of the PC. (a) Optical microscopy of the WG damage site with $350\times$ magnification; (b) contour map of the WG damage site.

Fig. 4. Morphology of the whole damage sites of WG surface of PC in the Regen.

Fig. 5. Morphology of a damage site on an Nd:glass rod surface caused by LID in the LID test experiment when laser energy fluence exceeds the damage threshold of the Nd:glass rod surface.

Fig. 6. (a) Raman spectra of deposits on the damaged QWP surface in comparison to that of carbon; (b) Raman spectra of deposits on the surface of damaged WG of PC in comparison to that of carbon.

Fig. 7. The TIC for the AMC sampling made in the Regen cavity.

Fig. 8. The regenerative amplifier with a CDN purge system, which is installed in the Regen cavity.

Fig. 9. The output energy of the Regen with a CDN purge system working versus time.

Fig. 10. (a) The diagram which describes that the ghost beam irradiates on the gap of the rotation stage of QWP2. The solid red line represents propagation of the main laser beam, and the dashed blue line represents propagation of the ghost beam generated by reflection of the main beam from a surface of QWP2. (b) A photo of rotation stage.

Fig. 11. A photo of the PC used in the Regen. The external aperture stop of the stainless steel window holders was marked with a red circle.

Fig. 12. Normalized output energy of the regenerative amplifier as a function of time before and after improvement.