|Thesis abstract: |
The analysis of the optical signals plays a key role in many fields, from biology and medicine to communication and security. In this domain one of the most effective techniques is the Time Correlated Single Photon Counting technique, which is very advantageous especially when the light signals are extremely weak and fast.
Based on the detection of the single photons that compose the light and on the measurement of their arrival times, the TCSPC technique requires extremely high performance, especially in terms of time resolution, differential non linearity (DNL) and operating frequency.
In medicine and biology, for example, techniques like FRET(Förster Resonance Energy Transfer) or sFLIM (spectrally-resolved Fluorescence Lifetime Imaging) deeply rely on the time-resolved analysis of the optical signals with a resolution down to few tens of picoseconds and they play a key role in the studies that investigate the complex interaction between molecules involved in crucial issues such as the origin and growth mechanisms of tumors.
In recent years, then, a grown interest has arisen around advanced TCSPC techniques: concurrency, indeed, opens the way not only to a further speed up of the overall measurement time, but it also makes the record of different events occurring simultaneously possible, as required by some other important techniques like the Fluorescence Cross-Correlation Spectroscopy (FCCS).
However, TCSPC multichannel systems that are both available in commerce or presented in literature still suffer from a tradeoff between the number of channels and performance: the best commercial systems in terms of performance feature only a few parallel channels (typically four or eight channels), while the performance achieved with multichannel structures are quite far from the state of the art.
The aim of this project is thus the development of the integrated electronics necessary to build a TCSPC system featuring a very high number of channels while providing also very high performance on each single channel.
Up to now the best-in-literature results have been obtained with Single Photon Avalanche Diodes (SPAD) that have intrinsic advantages such as ruggedness, integrability, high photon detection efficiency, low power dissipation and insensitivity to magnetic fields.
However, this approach strictly requires a front-end electronics specifically designed to properly work with these sensors: for this reason the design of TCSPC acquisition systems is still an open challenge, all the more so when it comes to multichannel acquisition systems.
So, first of all, a front end electronics capable of both correctly operate the SPAD detector and extract the information about the arrival of the photon is of the utmost importance.
Above all, this pick up circuit should be suitable to be placed as close as possible to the sensor, since maintaining the integrity of the signal and minimizing the disturbances is essential.
Secondly, the circuit must be adapt to be employed in a multichannel system and this basically translates into two requirements: the crosstalk between the channels must be avoided and power consumption and area occupation have to be minimized.
The signal has then to be fed to a time measurement block, that is a circuit specifically designed to measure the arrival time of the photons with respect to a reference signal. Historically this block has been implemented following two different approaches: a purely digital solution based on a Time to Digital Converter or a Time to Amplitude Converter followed by an ADC.
In this project the time measurement block will be based on a TAC structure in order to achieve the best performance especially in terms of time resolution and DNL.
Again, to develop an integrated circuit suitable to be employed in a system with a very high number of channels, the crosstalk between the channels must be negligible ¿ also when many channels of the TAC array convert at the same time - as well as the power dissipation and the area of the chip must be minimized.
Last but not least, this project will include the design and development of a routing circuit. This circuit shall be capable of binding a SPAD with a time measurement circuit only when a photon is detected. In this way, it would be possible to optimize the number of converters, with clear advantages in terms of performance, miniaturization and scalability. This circuit must firstly provide a fair treatment of the detectors ¿ in order to avoid, for example, that an eventual hot pixel ends up with masking some other detectors ¿ and it should also guarantee that the maximum conversion frequency of the system is always achievable.
Since the employment of a smart router in conjunction with many acquisition channels in parallel could lead to some issues related to the elaboration of the signal coming from one single detector by different circuits (that can differ from each other, for example, for a static offset of the time measurement circuit) a calibration algorithm will be elaborated as well, in order to evaluate and compensate via software the differences between the paths that the output of a SPAD can go through.
In conclusion, the aim of this project is the design and development of the integrated circuits that can lead to the building of a TCSPC acquisition system capable of breaking the tradeoff between number of channels and performance that now limits the available systems.