CERN Accelerating science

RD50: 30th Workshop on radiation hard semiconductor devices for high luminosity colliders

by Michael Moll

The RD50 collaboration “Radiation hard semiconductor devices for high luminosity colliders” is working on the development of radiation-tolerant silicon sensors for the detectors of the HL-LHC (High Luminosity-LHC) experiments and started to evaluate challenges arising from radiation levels and technology demands anticipated for future detector systems at the FCC (Future Circular Collider). This article gives information about the last RD50 workshop, the general progress of the overall RD50 research program and highlights recent RD50 developments on precision timing detectors.

The 30th RD50 Workshop

The 30th RD50 workshop took place in Krakow, Poland, from 5th to 7th June 2017 [] and was organized by the AGH University of Science and Technology. With 70 participants, 40 oral presentations, 6 discussion sessions reviewing the status and the work plans of all RD50 research lines, plenty of new results as well as a session dedicated to FCC tracking detectors, the workshop was a scientific success.

The RD50 Collaboration and its R&D program

The RD50 collaboration was formed in 2001 and presently comprises, with the joining of the CNA Sevilla in June 2017, 56 institutions with 330 members. RD50 has brought forward many achievements among which the development and characterization of n-in-p type silicon-based tracking detector sensors is surely the most important. These devices have overcome the limitations in radiation hardness of the presently used p-in-n type silicon strip sensors in the inner tracking systems and assure a radiation tolerant detector performance for the HL-LHC operation phase. Further milestones were the development and industrialization of 3D sensors, the creation of reliable modelling and characterization tools for highly radiation damaged sensors and the profound understanding and identification of microscopic defect levels in the forbidden band gap of Silicon that constitute the main origin of all sensor performance degradation. Presently, the most intensive research and development fields are the development of radiation tolerant precision timing detectors reaching down to some 10s of picoseconds time resolution (see below for details) and the optimization of CMOS based Monolithic Active Pixel Detectors with respect to radiation tolerance. These two technologies shall enable on the one hand an effective pile-up mitigation by separating events in time that took place during the same bunch crossing and on the other hand deliver cost effective, low mass and fully integrated sensor concepts for future applications. Both approaches are coming along with the very challenging intention to be ready for installation in some specific detector areas already for the HL-LHC. For more details on the RD50 research program, the reader is referred to the RD50 website. In the following, a short summary on precision timing detectors will be given.

Precision timing detectors: How much to gain with gain?

4-Dimensional Particle Tracking

4-Dimensional Particle Tracking is adding the 4th dimension of high precision time information to the 3-dimensional spatial information delivered by tracking detectors. It thus enables to separate different particles not only by their track (i.e. path within the detectors of the LHC experiments) but also by the time they pass through the detectors. Every 25 nanoseconds the proton bunches of the LHC are crossing, but not all proton-proton collisions happen at exactly the same time. The collisions take place in a time window of several 10s of picoseconds (0.00000000001 seconds) over the interaction region in the heart of the experiments. Being able to separate the tracks in time with such a high resolution would thus allow for reduction of the number of overlapping tracks (i.e. the pile-up) and therefore increase the overall detector performance. But how to build detectors providing a spatial resolution down to 10 micrometres and a time resolution of 10 picoseconds for the harsh radiation environments of the LHC, HL-LHC or even FCC?

Solid State Detectors with intrinsic gain

Precision timing detectors require a good Signal to Noise ratio and a high signal slew rate (short risetime of the signal). A promising silicon-based option to achieve these properties is to implement an amplifying layer directly into the silicon sensor itself. RD50 is developing a technology that is very similar to that of an APD (Avalanche Photo Detector), where a highly doped gain layer provides an electric field strength being sufficiently high to create an avalanche of electrons (“charge multiplication”) from the charges generated by the absorbed photons. The so-called LGAD (Low Gain Avalanche Detector) concept is used which has compared to an APD only a very moderate gain (order of 10-100), hence the name. A schematic drawing and a simulated field strength profile for the region close to the top electrode (cathode) where the amplification is taking place are shown in Figure 1.

Performance and radiation damage issues of LGAD sensorsThe technology development of LGAD devices was initiated and driven in the framework of RD50 by IMB-CNM (Institute of Microelectronics of Barcelona) in collaboration with many RD50 member institutes. Within 5 years, 25 fabrication cycles with more than 200 test wafers were performed at CNM in order to optimize the LGAD design and production, and tailor the devices to various applications. Today, 4 sensor producers around the world (CNM, Spain - FBK, Italy - HPK, Japan and Micron, UK) have produced LGAD devices putting the commercial availability of the devices on solid ground. Furthermore, first prototype productions have been performed at BNL, USA. In parallel, promising tests on devices based on the Deep Diffused APD concept with a higher gain have been performed. The devices, produced by the company RMD in the USA, are awaiting first radiation tests.

In the course of the development it became quickly obvious that the sensor size and thickness plays a crucial role for optimizing the timing performance. It was for example demonstrated that with 45 micron thick and very small (1.7 mm2) LGAD devices a timing resolution of about 30 picoseconds can be reached and, if combining e.g. 3 layers of LGADs, the achieved resolution is less than 20 picoseconds. The performance, however, is compromised in extremely strong radiation fields as demonstrated in Figure 2. The signal decreases with increasing exposure to high energetic particles, which then leads to a degradation of the time resolution.

Silicon based precision timing sensors within the LHC Experiments

The reason for the degradation is found in a gradual destruction of the gain layer in a process that is called “acceptor removal” that originates from the de-activation of the Boron doping that is used to produce the high electric field strength. Mitigation techniques, like the replacement of the Boron by Gallium, or the enrichment of the gain layer with impurities like Carbon, which getter radiation induced defects and thus save the Boron from being de-activated, are being studied. Despite the degradation, the technology can be used after exposure to strong radiation fields as recent beam tests have demonstrated. After a neutron fluence of 1015 neq/cm2 (corresponding to the cumulated radiation fluence expected for detector layers at a radius of 20 cm in the HL-LHC after 3000 fb-1) still a time resolution of 60 picoseconds was reached with special biasing conditions (see Figure 3) giving hope that mitigation approaches will in the future enlarge the reach of this technology to this and even higher radiation levels.

The need for precision timing detectors for pile-up mitigation, the promising results obtained with the LGAD technology and its improved radiation hardness with respect to non-silicon timing detectors have already pushed this new technology into the LHC experiments. CMS-TOTEM CTTPS has installed the first layers of LGAD sensors and ATLAS Forward Proton (AFP) is considering this option. For the HL-LHC upgrade, LGAD sensors are the baseline technology for the ATLAS High Granularity Timing Detector (HGTD) and the CMS MIP Endcap timing layer.

Silicon-based precision Timing detectors are a very essential detector component for future experiments and at the same time a hot research and development field. RD50 is proud to have initialized this technology and pushed it within only 5 years from first design to first operation in an LHC Experiment! Stay tuned for the future and follow this challenging technology development.


More information:

RD50 website:

RD50 Co-spokespersons: Gianluigi Casse and Michael Moll

The Solid State Detector Lab at CERN (CERN-EP-DT): Michael Moll

30th RD50 Workshop, Krakow, June 2017:


Latest issues in pdf