A 100Mphoton/s time-resolved mini-silicon photomultiplier with on-chip fluorescence lifetime estimation in 0.13μm CMOS imaging technology

Tyndall, David and Rae, Bruce and Li, David and Richardson, Justin and Arlt, Jochen and Henderson, Robert; (2012) A 100Mphoton/s time-resolved mini-silicon photomultiplier with on-chip fluorescence lifetime estimation in 0.13μm CMOS imaging technology. In: 2012 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC). IEEE, GBR, pp. 122-123. ISBN 9781467303736 (https://doi.org/10.1109/ISSCC.2012.6176946)

Full text not available in this repository.Request a copy

Abstract

Time-correlated single photon counting (TCSPC) is a technique whereby low-light signals are recorded with picosecond timing resolution relative to a synchronized optical impulse excitation, in order to extract the characteristic fluorescence decay constant, or lifetime [1]. Typical TCSPC apparatus includes a pulsed optical source, a discrete detector such as an avalanche photodiode (APD) or photomultiplier tube (PMT), external time-to-digital conversion (TDC) hardware and a PC to compute the decay constant, resulting in a bulky, expensive and power-hungry acquisition system. A major limitation of this approach is the restrictively low photon count limit of 1-to-5% of the excitation rate, which is necessary in order to avoid distortion due to photon 'pile-up' caused by both long detector dead-time and the inability of the TDC hardware to process more than one event per excitation period. As such, promising applications of TCSPC including cell cytometry, confocal microscopy, high throughput screening (HTS), and functional near infrared spectroscopy (fNIRS) are severely limited by peak acquisition rates of 1MHz. Although 100MHz has been achieved [2], the approach used is restricted to fluorescent dyes with lifetimes less than 2ns. Recent advances in single-photon avalanche diodes (SPADs) and on-chip TDCs manufactured in standard CMOS processes have enabled TCSPC measurements to be performed by an imaging array [3]; however such devices produce data at over 25Gb/s, have low fill factors of ∼2% and pixel update rates are limited. Time-gated lifetime sensing significantly reduces the data bandwidth and processing time [4,5], but is photon inefficient and still limited by pile-up.