SiPM Photon Detection Efficiency (PDE) Calculator

Formula

How this is calculated

A Silicon Photomultiplier (SiPM) consists of thousands of microscopic Geiger-mode avalanche photodiode (GAPD) microcells wired in parallel. Three successive processes determine whether an incident photon produces an output pulse.

First, the photon must be absorbed by the silicon lattice and generate an electron–hole pair. The probability of this is the Quantum Efficiency (QE), which depends on the anti-reflection coating, the wavelength of the light, and the doping profile of the silicon. Modern SiPMs achieve QE of 20–50% at their peak wavelength (typically 420–550 nm). Second, the generated carrier must be in an electrically active region; if it lands in the dead-area border (quench resistor, metal interconnect, trench) it is lost. The Fill Factor (FF) is the ratio of active area to total pixel area, ranging from about 30% in fine-pitch devices to over 80% in coarse-pitch ones. Third, the carrier must trigger a self-sustaining Geiger avalanche with probability P_G, which rises with the overvoltage (Vbias − Vbreakdown) and can exceed 90% at high overvoltage. PDE is the product of all three: PDE = QE × FF × P_G.

The FractionBar shows the four fates of 100 incident photons: those lost because they are not absorbed (QE loss), those absorbed but landing in inactive area (FF loss), those that reach an active cell but fail to trigger an avalanche (P_G loss), and those that produce a count (PDE). Increasing any one factor proportionally increases PDE, but practical devices trade off FF against dynamic range (smaller cells → lower FF but more cells → higher saturation level), and higher overvoltage improves P_G at the cost of more dark counts.