A large percentage of the detectors from the first production batches that were delivered to us from Micron Semiconductor exhibited unusual structure in the I/V plots which were measured on the 32-channel probe station. The detectors had nominal current, 1 nA or less, on the outside strips, but in the center region of the detector the measured current appeared to be orders of magnitude greater. The number of strips comprisng the high current region varied from detector to detector, from 10's of strips, all the way up to 100's of strips. The resulting plots led us to refer to the structure as a top hat effect. On examination of these data, we realized that the sum of the individual currents that we measured in the top hat region of the detector exceeded the total current we measured, being supplied to the detector through the bias. We concluded that the individual strip measurements were actually reading current from a bulk area that was greater than the width of one strip. We subsequently made measurements of the interstrip resistance. We found that the interstrip resistance in the top hat region of the detector was as low as 10's of k-Ohms, whereas in the region of the detector where the current was low, the interstrip resistance was several G-Ohms. We made several measurements to determine whether there was a metal short between strips, or from the strips to a guard ring. We did not find any evidence for this type of fault. Micron sent one of our detectors to the silicon group located at INFN, Pisa, Italy. They confirmed our measurements.

These detectors are coated with an environmental protection made of a thin polyimide layer. It is possible that the processing or application of this coating was flawed and that charge states were induced into the silicon/silicon dioxide interface, causing an effective short between strips. Micron has subsequently delivered several detectors coated with a silox passivation. None of these detectors has, thus far, exhibited unusual structure in the I/V distributions. The interstrip resistance is uniformly high. We have concluded that the original fault we observed was due to a processing flaw in the mixture or application of the polyimide coat. As the vendor does not have an explanation or solution to the problem, and since the silox detectors appear to behave nicely, we have decided to use only silox coated detectors for the MVD.