Irising time of microchannel plate intensifiers with fiberopticcathode windows has been reduced to less than 100 ps. This
is achieved by metalizing the fiberoptic's cladding glass to reduce cathode substrata resistance without reducing transmission
of the cathode window.
During irising, a bright ring is seen, first at the edge as it propagates toward the center. This time lag is caused by the
distributed time constant of the substrata resistance and the cathode-to-MCP capacitance. Since the capacitance is fixed by
restraints of tube geometry, our goal was to reduce the distributed resistance sufficiently to achieve sub-nano-second
The process involves a number of steps. First, an approximately 34-nm-thick layer of nickel is evaporated over the cathode
side of the fiberoptic window. Next, the nickel is coated with photoresist, which is exposed through the faceplate and the
nickel, using the fiberoptics as a self-aligning mask. The photoresist is developed and washed off, exposing the nickel
covering the active fiber area or core glass. At this point, photoresist still covers the nickel over the cladding glass or
webbing, protecting it when the exposed nickel covering the core glass is etched away. Finally, the remaining photoresist is
removed, leaving only the nickel covering the fiberoptic webbing.
We processed several 18-mm faceplates from ITT, Fort Wayne, Indiana, and ITT used the plates to construct intensifiers
which we tested for irising time. Photographs, as well as computer processed 3-D images of the output showed no irising.
This means that any irising that occurs is clearly faster than the time resolution of our measurement system (which is
estimated to be less than 50 ps).