November_EDFA_Digital

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 21 NO. 4 24 This layer serves bothas aprotective layer against possible unwantedmilling fromthe FIBandas abaseline resistor to monitor the resistance during FIB-induced deposition. It is worth noting that 10 minutes were required to deposit this first layer of approximately 60 kOhm resistor using the ebeam, which justifies use of the FIB to deposit the 1 kOhm, as its deposition rate is much higher. Once this baseline resistorwas deposited, the stagewith theprobing systemwas tilted to FIB position at 54°, alongside the GIS nozzle, as shown in Fig. 4. At this tilt angle, horizontal alignment and landing of the probes in contact with the terminal pads was rapidly achievedwith the nanometricmotion resolution probers. Prior to deposition, the baseline resistor was measured again, to check the electrical connection between the probes. Platinum deposition was then initiated at the same time as a 1V bias was appliedwith the source-meter unit between the probes. The current flowing between themwas continuously monitored during the deposition as depicted in Fig. 5. The FIB and gas insertion were stopped once the desired value of 1 mA was reached. The obtained resis- tance was then measured at exactly 1.00 kOhm, which confirms the advantage of being able to measure the resistance while it is being deposited. As the deposition did not take place on an entirely flat substrate (Fig. 6), the sample topography impacts the speed at which the resistance changes over time. In fact, because it takes more time to deposit platinum on the vertical sides of the metal tracks, these areas act like bottlenecks for the current, increasing overall resistance. Fig. 3 Platinumlayer depositedbetween the twopads using the electronbeam. It serves bothas aprotective layer fromundesired FIBmilling and as a baseline resistor for the in situ resistance monitoring. Fig. 4 Chamber scope view of the platform in operating conditions at FIB tilt position with the GIS nozzle inserted. The sample is at the coincidence point of the two beams. Fig. 5 Live monitoring of the current between the two probeswhile the resistor is beingdeposited. Avoltage difference of 1V is applied to the probes. Fig. 6 SEMimageunder 54° stage tilt of the resistor obtained at the end of the FIB induced deposition. The probes are landed on the resistor terminals to measure its value. (continued on page 26)