$$ \def\MODULE#1{\left|\,#1\,\right|}% \def\PARENTHESES#1{\left(#1\right)}% \def\SQBRACKETS#1{\left[#1\right]}% \def\BRACES#1{\left\{ #1\right\} }% \def\LBRACE#1{\left\{ #1\right.}% \def\RBRACE#1{\left.#1\right\} }% \def\LSQBRACKET#1{\left[#1\right.}% \def\RSQBRACKET#1{\left.#1\right]}% \def\LPARENTHESIS#1{\left(#1\right.}% \def\RPARENTHESIS#1{\left.#1\right)}% \def\ANGLEBRACKETS#1{\left\langle #1\right\rangle }% \def\SPACELONG{\hspace{10mm}}% \def\SPACEMEDIUM{\hspace{5mm}}% \def\DEF{\overset{{\scriptscriptstyle \text{def}}}{=}}% \def\UPBRACE#1#2{\overset{{\scriptstyle #2}}{\overbrace{#1}}}% \def\UNDERBRACE#1#2{\underset{{\scriptstyle #2}}{\underbrace{#1}}}% \def\REALES{\mathbb{R}}% \def\IMAGINARIOS{\mathbb{I}}% \def\NATURALES{\mathbb{N}}% \def\ENTEROS{\mathbb{Z}}% \def\COMPLEJOS{\mathbb{C}}% \def\RACIONALES{\mathbb{Q}}% \def\DIFERENTIAL{\,\text{d}}% \def\PRIME{{\vphantom{A}}^{\prime}}% \def\ORDER#1{\mathcal{O}\PARENTHESES{#1}}% \def\DIRACDELTA#1{\delta_{D}\PARENTHESES{#1}}% \def\HEAVYSIDETHETA#1{\Theta_{H}\PARENTHESES{#1}}% \def\ATAN{\text{atan}}% \def\INDICATORFUNCTION#1{\mathbf{1}\BRACES{#1} }% \def\VECTOR#1{\boldsymbol{#1}}% \def\VERSOR#1{\hat{\VECTOR{#1}}}% \def\IDENTITY{\mathds{1}}% \def\CURL{\VECTOR{\nabla}\times}% \def\GRADIENT{\VECTOR{\nabla}}% \def\DIVERGENCE{\VECTOR{\nabla}\cdot}% \def\LAPLACIAN{\nabla^{2}}% \def\REALPART#1{\text{Re}\left(#1\right)}% \def\IMAGPART#1{\text{Im}\left(#1\right)}% \def\TENDSTO#1{\underset{{\scriptscriptstyle #1}}{\longrightarrow}}% \def\EVALUATEDAT#1#2#3{\left\lceil #1\right\rfloor _{#2}^{#3}}% \def\unit#1{\text{#1}} \def\TERA#1{\text{ T}\unit{#1}}% \def\GIGA#1{\text{ G}\unit{#1}}% \def\MEGA#1{\text{ M}\unit{#1}}% \def\KILO#1{\text{ k}\unit{#1}}% \def\UNIT#1{\,\unit{#1}}% \def\CENTI#1{\text{ c}\unit{#1}}% \def\MILI#1{\text{ m}\unit{#1}}% \def\MICRO#1{\text{ }\mu\unit{#1}}% \def\NANO#1{\text{ n}\unit{#1}}% \def\PICO#1{\text{ p}\unit{#1}}% \def\FEMTO#1{\text{ f}\unit{#1}}% \def\TIMESTENTOTHE#1{\times10^{#1}}% \def\PROB#1{\mathbb{P}\left(#1\right)}% \def\MEAN#1{\mathbb{E}\PARENTHESES{#1}}% \def\VARIANCE#1{\mathbb{V}\PARENTHESES{#1}}% \def\COLOR#1#2{{\color{#2}{\,#1\,}}}% \def\RED#1{\textcolor{red}{#1}}% \def\BLUE#1{\COLOR{#1}{blue!80!white}}% \def\GREEN#1{\textcolor{green!70!black}{#1}}% \def\GRAY#1{\COLOR{#1}{black!30}}% \def\GRAY#1{\COLOR{#1}{blue!35!white}}% \def\GUNDERBRACE#1#2{\GRAY{\UNDERBRACE{\COLOR{#1}{black}}{#2}}}% \def\GUPBRACE#1#2{\GRAY{\UPBRACE{\COLOR{#1}{black}}{#2}}}% \def\REDCANCEL#1{\RED{\cancel{{\normalcolor #1}}}}% \def\BLUECANCEL#1{{\color{blue}\cancel{{\normalcolor #1}}}}% \def\GREENCANCEL#1{\GREEN{\cancel{{\normalcolor #1}}}}% \def\BLUECANCELTO#1#2{\BLUE{\cancelto{#2}{{\normalcolor #1}}}}% $$
$\def\NEUTRONEQUIVALENT{\UNIT n_{\text{eq}}}$%Matías Senger
August 6, 2021
In this document I provide an update on the commissioning of setup for the long term studies of irradiated LGAD detectors at UZH. This setup is intended to periodically monitor the characteristics and performance of irradiated LGAD detectors while they are kept constantly biased for a long period of time.

Table of contents

Introduction

In this work a set of irradiated FBK detectors from the UFSD3.2 production   are going to be testedMore specifically, devices of "type" 4 and 10 from wafers 4, 10, 18 and 4A, irradiated to fluences of $4\TIMESTENTOTHE{14}\NEUTRONEQUIVALENT\CENTI m^{-2}$, $8\TIMESTENTOTHE{14}\NEUTRONEQUIVALENT\CENTI m^{-2}$, $15\TIMESTENTOTHE{14}\NEUTRONEQUIVALENT\CENTI m^{-2}$ and $25\TIMESTENTOTHE{14}\NEUTRONEQUIVALENT\CENTI m^{-2}$. . The tests consist in keeping them under working conditions (i.e. low temperature and biased) for an extended period of time, and regularly monitor a number parameters to look for deviations from the expected performance. For this, a dedicated setup is under implementation at UZH consisting of several hardware equipment and customized controlling and analysis software.
In this document the first report on its commissioning is presented. First the whole setup is described. Then some preliminary measurements are presented.

Setup description

In  a block diagram of the designed setup is shown. The setup was designed to handle up to eight devices simultaneously. The devices are individually controlled and monitored, independently from one another. Two CAEN high voltage power supplies provide the high voltage to bias the LGADs, while at the same time are used to monitor the bias current and regularly trace IV curves of each device. An oscilloscope and a PSI DRS4 evaluation board  are used to measure the signals coming out from each LGAD when exposed to beta radiation from an Sr-90 source. A robotic system moves the Sr-90 beta source and MCP-PMT time reference from one LGAD to the other. All the setup is automatically controlled by a computer which runs a customized software. A picture of the setup is shown in .
Block diagram of the designed setup.
Picture of the setup in its current state. The climate chamber hosts the readout boards with the LGADs. The two high voltage power supplies lie on the first shelve (red devices). The DC power supply is placed in the higher shelve. The oscilloscope and the PC are on the table.

Readout board

To provide flexibility, each device is mounted in an individual readout board which has a built in amplifier. The board, designed and produced in the context of this project, is based in the Santa Cruz board  . The circuit has some minor modifications and the board a completely new layout, to obtain a board with smaller dimensions. For details on this board and its performance please see reference  . Some pictures of this board can be seen in , and a microscopy picture of a device mounted in a Chubut board is shown in .
Pictures of the Chubut board  , developed in the context of this project.
Microscopy picture of a device mounted in the Chubut board.

High voltage power supplies

The bias voltage for the eight devices is provided by two CAEN modules, each with four outputs, which can be seen in the picture of (red apparatuses on the first shelve). One of the modules is model DT1419ET and the other is a DT1470ET. The modules are controlled through the network of the institute, optionally via USB. A simple and easy pure Python package was developed to control the instruments https://github.com/SengerM/CAENpy..

Acquisition system

To acquire the fast signals produced by MIP particles traversing the LGADs, a LeCroy WaveRunner 640Zi oscilloscope sampling at $40\GIGA s\UNIT s^{-1}$ is used. This oscilloscope has four input channels. To fully automatize the setup, a PSI DSR4 evaluation system   is planned to be used, though the software to control this device has not yet been implemented.

Timing reference

A Photonis PP2365AC MCP-PMT is going to be used as a reference for fast and precise triggering of the acquisition system. This device provides a time resolution $\lesssim10\PICO s$ Information provided to us by the manufacturer. Se also references      which describe the performance of a similar device for timing of minimum ionizing particles.. This device will be mounted in a robotic system, currently in design, to move together with a radioactive Sr-90 source along each of the eight test LGADs. The block diagram in  illustrates this.

Standby bias voltage and current monitoring

During the whole test the detectors will be kept under a standby voltage that can be configured individually for each device. Both the bias voltage and the bias current will be monitored constantly. In the very first measurements of standby bias current obtained for eight detectors simultaneously using this setup is presented. For this test the bias voltage for all the devices was chosen to be $66\UNIT V$, in the future appropriate voltages for each detector will be configured.
Very first measurements of the bias current for eight irradiated LGADs installed inside the climate chamber. The devices were all biased with $66\text{ V}$ for this test.

IV-curve measurements

The current-voltage (IV) characteristic of an LGAD detector provides valuable information. The setup measures the IV curve of each device periodically, using the values of bias voltage and current provided by the high voltage power supplies. In an IV-curve measurement example is shown (label "Measurement setup through CAEN"). The bias voltage and the current are the values that were feed into the Chubut board. The current corresponds to the sum of the current of the four pads (3 wire bonded to ground, one connected to the amplifier input, see ) plus the guard ring, which is wire bonded to ground. The amplifier input is kept at $\approx0\UNIT V$ DC, while the backside of the device is where the bias voltage is applied.
This measurement () is compared against an IV characterization of the same device made with a probe station before mounting the device in the readout board. In this case, the guard ring was grounded, one of the pads was kept fixed at $0\UNIT V$ and the bias voltage was applied to the backside through the chuck. The current shown under the label "Probe station" is the current flowing through the backside.
Example of an IV curve measured by the setup using one of the high voltage power supplies. The measurement is compared with another measurement of the same device made with a probe station before mounting the device in the readout board. Both measurements were taken at $-20\text{ °C}$.
As seen, the two measurements differ significantly; the current measured by the CAEN is considerably higher than the current measured by the probe station, and the current measured in the probe station has an extra kink at $\sim80-90\UNIT V$. To account for these differences let's consider the following expressions for the current in each case \[ \LBRACE{\begin{aligned} & I_{\text{CAEN}}\approx4I_{\text{pad}}+I_{\text{guard ring}}\\ & I_{\text{probe station}}\approx I_{\text{pad}}+I_{\text{guard ring}} \end{aligned} } \] after the connections described for each case. If we focus in the region where the gain layer depletes the change in the total current (i.e. the measured current) should be dominated by the variations in $I_{\text{pad}}$, i.e. \[ \EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{guard ring}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\ll\EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{pad}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\text{.} \] Following this idea we should observe that \[ \EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{CAEN}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\approx4\EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{probe station}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\text{.} \] The depletion of the gain layer is happening between $-35\UNIT V$ and $-40\UNIT V$ approximately. In the measurement using the probe station the increase between $-35\UNIT V$ and $-40\UNIT V$ is of \[ \EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{probe station}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\approx135\NANO A-94.3\NANO A=40.7\NANO A\text{.} \] For the measurement with the CAEN the current increase in the same voltage range is about \[ \EVALUATEDAT{\frac{\DIFERENTIAL I_{\text{CAEN}}}{\DIFERENTIAL V}}{\text{Gain layer depletes}}{}\approx350\NANO A-190\NANO A=160\NANO A\text{.} \] The ratio of these two currents is \[ \frac{160\NANO A}{40.7\NANO A}=3.9 \] which seem to indicate that the previous reasoning is correct.
With respect to the difference at $\approx-90\UNIT V$, the second kink in the measurement with the probe station, not present on the measurement with the CAEN, may be due to border effects in the pads left floating.
For reference, IV-curve measurements taken with the probe station of all the devices are shown in . To enable/disable traces simply click/double click in the legend.
IV curves of all the irradiated FBK devices measured in a probe station at UZH. The traces can be enabled/disabled by clicking in the legend to ease the visualization/compare among them. The color of the traces is linked with the fluence of irradiation. For all the measurements, the devices were connected in three points: 1) the back side where the bias voltage was applied and the $I_{\text{chuck}}$ current measured, 2) one of the four pads where a constant $0\text{ V}$ voltage was applied and the current $I_{\text{pad}}$ was measured, and 3) the guard ring was connected to ground. The guard ring current shown is $I_{\text{gr}}=I_{\text{chuck}}-I_{\text{pad}}$. The remaining 3 pads were left floating. All measurements were taken at $-20\text{ °C}$.

Charge collection measurements

An important figure of the LGAD detectors is the collected charge. To obtain good timing capability a charge of about $5\text{-}6\FEMTO C$ is required in the output of the detector  . The experimental setup is designed to periodically measure the collected charge of each device using the Sr-90 source of beta particles. The software for this is still under development, but as an example in  a signal produced by a beta particle in an HPK PIN diode is shown, together with a number of features extracted by the analysis software. This signal was recorded in a preliminary test of the current setup. The distribution of the collected charge for this test is shown in .
Example of a signal acquired with the setup using the Sr-90 source and the oscilloscope. This signal was recorded during a preliminary test using an HPK PIN diode biased with $55\text{ V}$. A number of features extracted by the analysis software are shown, in particular the collected charge (in arbitrary units of $\text{V s}$).
Example of collected charge distribution from a set of signals like the example one shown in .

Timing measurements

Perhaps the most relevant measurement is the time resolution of each detector; after all the detectors will be used for this. To monitor the time resolution of each detector a precise MCP-PMT reference detector will be used as trigger, as described previously. This detector was already ordered to the manufacturer, but it has not yet arrived to our lab. In the meantime a non-irradiated CNM LGAD detector with a time resolution of $27\PICO s$ will be used.

Conclusions

Work is ongoing on the commissioning of the setup for the long term tests of irradiated LGAD devices. The setup is already partially operational and monitoring the standby current of eight detectors. There is still work to do in terms both of hardware and software. It is expected to have it fully operational in about 2-4 months.

Footnotes

References

Ferrero, Marco, Roberta Arcidiacono, Marco Mandurrino, Valentina Sola, and Nicolò Cartiglia. An Introduction to Ultra-Fast Silicon Detectors: Design, Tests, and Performances. Boca Raton: CRC Press, 2021. https://doi.org/10.1201/9781003131946. The Chubut board, Matías Senger. https://msenger.web.cern.ch/the-chubut-board/. UCSC Single Channel. https://twiki.cern.ch/twiki/bin/view/Main/UcscSingleChannel. PSI DRS4 evaluation board, https://www.psi.ch/en/drs/evaluation-board. A. Ronzhin, S. Los, E. Ramberg, M. Spiropulu, A. Apresyan, S. Xie, H. Kim, A. Zatserklyaniy, Development of a new fast shower maximum detector based on microchannel plates photomultipliers (MCP-PMT) as an active element, 21 September 2014, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. https://doi.org/10.1016/j.nima.2014.05.039. A. Bornheim, C. Pena, M. Spiropulu, S. Xie, Z. Zhang, Precision timing detectors with cadmium-telluride sensor, 21 September 2017, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. https://doi.org/10.1016/j.nima.2017.04.024. Test Beam Studies Of Silicon Timing for Use in Calorimetry, A. Apresyan, G. Bolla, A. Bornheim, H. Kim, S. Los, C. Pena, E. Ramberg,A. Ronzhin, M. Spiropulu, and S. Xie. https://inspirehep.net/files/f6b1cd929d10bb6fe53679fd2f38d3c7. A. Ronzhin, M.G. Albrow, M. Demarteau, S. Los, S. Malik, A. Pronko, E. Ramberg, A. Zatserklyaniy, Development of a 10ps level time of flight system in the Fermilab Test Beam Facility, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. https://doi.org/10.1016/j.nima.2010.08.025.