Simulations push frontier of pixel detectors forward
Jennifer Toes (CERN), 27/05/2016

 


Figure 1: Schematic showing a TCAD simulation of a sensor with active-edge technology
and a guard ring in the edge. The electric field distribution is illustrated in this figure

New simulation results from a CERN doctoral researcher involved in Work Package 7 of the AIDA-2020 project will help to develop new active-edge sensors for pixel detectors.

WP7 focuses on the development of advanced hybrid pixel detectors, including the definition of optimal configurations for silicon pixel sensors. Task 7.2 employs the use of TCAD (Technology Computer Aided Design) simulations to do just that.

TCAD simulations allow researchers to model and study the behaviour of a semiconductor detector and provide information on its performance. This helps to locate areas of underperformance so researchers can take this data into consideration to enhance future detectors by selecting the most efficient design.

Equally, the use of TCAD simulations has a cost-effective purpose, as designing, manufacturing and testing multiple prototypes can be costly – the simulations allow designers to identify potential weak spots before they are put into production.

“Since the cost of production is high, it is necessary to optimise the sensor design in simulation to make sure the device will behave as expected and avoid defects or breakdowns” explains Niloufar Alipour Tehrani, a doctoral researcher based at CERN. Tehrani works on the development of active-edge sensors within the Linear Collider Detector group (LCD) at CERN as part of the Compact Linear Collider (CLIC) experiment.

Active-edge technology is used for pixelated planar silicon sensors to extend the backside voltage to the trench area. This allows for a fully depleted sensor all the way up to its physical edge and thus increases its efficient area. If all goes to plan, the idea is for these sensors to seamlessly tile the innermost detector (vertex detector) of CLIC. Figure 1 shows a simulation of the cross-section of an active-edge sensor, demonstrating how the backside implantation and bias voltage has been extended right up to the edge.

However, the extended edge implant makes the sensor vulnerable to potential breakdowns due to the high potential gradient between the active edge and the last pixel. “Using TCAD simulations, we can study the breakdown behaviour for different configurations of the sensor,” explains Tehrani. Using the Sentaurus Device simulator from Synopsys, Tehrani input different sensor configurations which were produced by the semiconductor sensor production company Advacam

The proposed new design of the sensor is only 50-150 micrometers thick, and has better breakdown behaviour and higher detection efficiency. The design also includes a guard ring, which consists of an n-implant with a metallic contact on top, surrounding the pixel matrix close to the edge. The guard ring thereby smoothes the potential transition between the edge and the neighbouring pixels, which is vulnerable to breakdown due to the high gradient of potential between these areas.

The sensors were produced over the course of a year, and then trialled during beam tests by Tehrani. Their performance was then analysed and compared to the original TCAD simulations to verify the results and identify possible inconsistencies.

“[The] next step would be to study the active edge designs with 3D simulations” says Tehrani, which would allow researchers to understand the efficiency in between the pixels close to the edge.

With ever changing requirements in resolution, quality and cost effectiveness for these sensors, identifying the most efficient technological configuration before production even begins allows researchers to push the frontiers of detector technology forward to meet their needs.

You are here