A photocathode is a thin multilayer, usually deposited onto a transparent glass window. Its function is to absorb soft (visual) photons entering through the window, and to subsequently emit a detectable photoelectron into the adjacent vacuum. The probability that an incoming soft photon results in the emission of a photoelectron never exceeded 40%. With this high risk/high impact proposal we intend to raise this figure using new MEMS technology.

The photomultiplier, developed since 1934, has been the state-of-the-art soft photon detector until a decade ago. The photocathode is an essential component of this detector: it absorbs a soft photon, converts it into a photoelectron, and emits this electron into the vacuum. The Quantum Efficiency (QE) is defined as the probability to emit a photoelectron, per incoming soft photon. The QE never exceeded 0.4 for soft photons (400 nm < λ < 1000 nm).

At present, so-called Silicon Photomultipliers (Si-PMs) are rapidly penetrating the market of soft photon detectors because of their thin planar form, low mass, low cost, good efficiency and good time resolution. It can be foreseen that these detectors will approach full efficiency, but (shot) noise (dark current, dark counting rate) can’t be reduced below a certain level, and time resolution stays limited to 10 – 20 ps.

The other essential component of a photomultiplier is the electron multiplier: a photo-electron impinging on the surface of a dynode causes the emission of a multiple of secondary electrons at the impact point. By accelerating these electrons towards a next dynode, the number of electrons increases exponentially with the number of dynodes. An important property of amplification by electron multiplication in vacuum is the absence of shot noise.

Within the ERC-Advanced ‘MEMBrane’ project, the transmission dynode tynode has been developed: a very thin membrane, which emits, at the impact of an energetic electron on one side, a multiple of secondary electrons at the other side. We have realised tynodes with a transmission secondary electron yield of 5.5; a stack of 6 of these tynodes outputs a charge pulse of 20 k electrons, enough to drive digital circuitry without further amplification. The sub-ps spread in timing of these electrons enables the development of the fasted future detectors. The combination of a photocathode and a stack of tynodes form a detector with potentially sub-ps time resolution, an arbitrarily good 2D spatial resolution, while free of noise, outperforming Si-PMs except for its efficiency: we are therefore aiming for a higher QE.

When successful, we will develop a thin, planar, low cost detector for individual soft photons with outstanding properties, without disadvantages. This generic soft photon detector will enable timed tracking in particle physics experiments, and it will also open new markets such as Cherenkov-ToF PET scanners, machine viewing, prompt 3D imaging and self-driving of cars.