A method to observe field-region oxide charge and inter-electrode isolation from $CV$-characteristics of $n$-on-$p$ devices

Abstract

$N$-on-$p$ silicon sensors will be utilized in the Compact Muon Solenoid (CMS) detector's tracker and High Granularity Calorimeter (HGCAL) in the High Luminosity upgrade of the Large Hadron Collider (HL-LHC). Among their several advantages in terms of radiation hardness over the traditional $p$-on-$n$ sensors in the extreme radiation environment of the HL-LHC are electron collection instead of holes and overlapping maxima of weighting and electric fields at the charge-collecting electrodes. The disadvantage of the multi-channel SiO$_2$-passivated $n$-on-$p$ sensors is the generation of an inversion layer under the Si/SiO$_2$-interface by a positive interface-oxide-charge ($N_\textrm{ox}$) that at high densities can compromise the position resolution by creating a conduction channel between the electrodes. This issue is typically addressed by including additional isolation implants ($p$-stop, $p$-spray) between $n^+$-electrodes. Focusing on the guard-ring regions of $n$-on-$p$ sensors where no isolation implants are applied between the electrodes, a capacitance-voltage ($CV$) characterization study of both 6-inch wafer test diodes and 8-inch HGCAL prototype and pre-series sensors showed a distinct threshold voltage ($V_\textrm{th,iso}$) in the $CV$-characteristics of a biased $n^+$-electrode when its enclosing guard-ring was left floating. When reproduced by simulations, the measured $V_\textrm{th,iso}$ was found to contain information on the field-region $N_\textrm{ox}$ and indicate the threshold where the two electrodes become electrically isolated by the influence of the reverse bias voltage. Together with previous studies on the inter-electrode isolation of irradiated $n$-on-$p$ sensors, the results indicate that position sensitive $n$-on-$p$ sensors without isolation implants may be feasible in future HEP experiments.