研究CMS高粒度量能器之原型於CERN SPS測試粒子束下的表現

國立台灣大學理學院物理學研究所 碩士論文

Graduate Institute of Physics College of Science

National Taiwan University Master Thesis

研究 CMS 高粒度量能器之原型於

CERN SPS 測試粒子束下的表現

Performance of a Novel CMS High Granularity Calorimeter(HGCAL) Prototype in Beam Tests at the

CERN SPS 簡嘉泓 Chia-Hung Chien

指導教授： 裴思達 教授 Advisor: Prof. Stathes Paganis

中華民國一百零八年二月

February 2019

Performance of a Novel CMS High

Granularity Calorimeter(HGCAL) Prototype in Beam Tests at the CERN SPS

Master Thesis

Chia-Hung Chien

Physics Department, National Taiwan University, No. 1, Sec. 4, Roosevelt Rd., Taipei 10617, Taiwan (R.O.C.)

Advisor Stathes Paganis

Physics Department, National Taiwan University, No. 1, Sec. 4, Roosevelt Rd., Taipei 10617, Taiwan (R.O.C.)

email:[email protected]

Feburary 2019

Acknowledge

Firstly, I would like to thank my advisor, professor Stathes Paganis, for his kindness and taking good care of me. He provided valuable advises and analysis strategy during the research and always willing to discuss with me about technical or physics issues. The period of doing research wouldn’t be so nice without his accompany. He also introduce me to the world of particle physics and detectors. I also want to thanks my roommate 劉子禎 for teaching me the coding skill and supervising me during the preparation for application of master degree.

I also appreciate having opportunities to attend the beam tests at CERN. There, I learn a lot from the experts, especially from Steen Arnaud and professor 呂榮祥. They taught me the correct attitude to analyze each problem and how to systematically debug, which is the most crucial ability in practical aspect. Moreover, working with all the people from beam tests (include HGCAL and AHCAL people) is truly a great honor to me. I will not forget the first beam test in July, 2017. It was the first time I realize how complicate the data taking of a prototype could be and the first time I feel the ecstasy of doing the physics. I also wants to thank Andreas Psallidas, 陳柏勳, 陳以牧, 李侑穎, 余佩容, 李家 豪, 吳新業, 林子路 and professor 陳凱風, they enriched my life in Geneva.

All members from high energy physics in Taiwan also taught me a lot: professor 侯 維恕, 熊怡, 張寶棣, 王名儒, postdoc 趙元, and all the colleagues in NTU CMS: 蔡連勝, 林俊廷, 楊宜知, 陳溥生, 徐啟峻, 黃文亮, 陳孟呈, 高裕維, 蘇俊瑋, 馬聖凱. Discussing physics, coding skills and the technical issues with them benefits a lot.

In the end I want to thank 黃惇暉 for providing a place for me to stay, I couldn’t accomplish this thesis without his help (also KFC is good). And of course, my family, for their support in both economically and spiritually to pursue high energy physics. There are still lots of people I didn’t mention, I feel sorry about not be able to list each of your names here. I really appreciate all the help and concern from you!

摘要

世界上最大且能量最高的粒子加速器–大型強子對撞機(LHC)，正邁向「高亮度」

(high luminosity)的運行階段，在下個階段，LHC 會產生現階段 10 倍的累積亮度 (integrated luminosity)，這會使偵測器面臨兩個重大的挑戰：輻射損傷(尤其是偵 測器前緣)以及源於碰撞點產生的事件數上升使得單一事件的分析變得困難。

高粒度量能器(HGCAL)是緊湊秒子線圈實驗(CMS)的其中一項升級計畫，

HGCAL 將會取代現有的量能器兩端，包含電磁量能器與強子量能器的部分。其 中電磁量能器與一大部分強子量能器會採用 0.5 到 1 平方公分大小的矽感應器，

而其他部分則會以小的閃爍體探測器並運用矽光電倍增器做讀出，其中矽感應器 精確的時間量測能幫助分辨短時間內產生的大量事件。

在 2016 年，以現存 CALICE 實驗開發的 Skiroc2 前端積體電路製程的第一個六 角形矽偵測模組(module)已投入測試，而新的前端積體電路–Skiroc2cms 也在 2017 及 2018 年投入測試。

本篇論文會以 2016 的測試粒子資料為主體進行 2 項研究，第一項是電子能量的 回推(重建)，有兩種方法會被運用並比較；第二項研究則是運用射叢粒子(particle shower)在偵測器內的型態所定義的變數分辨電子與π介子。

本篇將詳細說明 2016 年所進行的粒子測試以及其實驗裝置、訊號重建與蒙地卡 羅(Monte carlo)模擬的方法，而 2017/2018 年所進行的粒子測試之資料重建方法 也將被大略提及。

關鍵字：粒子物理、高能物理、量能器、偵測器、量能方法、能量重建、粒子辨 識

Abstract

The world’s largest and most powerful particle accelerator, Large Hadron Col- lider(LHC), is proceeding to the High Luminosity phase. LHC will deliver 10 times more integrated luminosity than now. It will lead to significant challenges for ra- diation damage and event pileup on detectors, especially in the endcap part of the detector. High-Granularity Calorimeter(HGCAL) is the chosen technology by the Compact Muon Solenoid(CMS) experiment as part of the phase 2 upgrading program. Consisting of the electron-magnetic and hadronic sections, HGCAL will replace the existing endcap calorimeters.

The electromagnetic section and a large fraction of the hadronic section will be based on hexagonal silicon sensors of 0.5–1 cm² cell size, while the rest of the hadronic section will use small scintillator with silicon photomultiplier(SiPM) read- out. The high-precision in timing capabilities of silicon will be helpful to pileup rejection.

First hexagonal silicon modules using the existing Skiroc2 front-end ASIC devel- oped for CALICE has been tested in 2016. New front-end ASIC named Skiroc2cms is tested in 2017 and 2018.

This thesis will provide 2 studies based on one of the beam test data in 2016.

The first study is the energy calibration of the electron. Two different methods will be explained and compared. The second study is the electron pion separation by the shower-shape variable. The test beam setup, data reconstruction and Monte Carlo generation of the 2016 test beam will be mentioned.

Furthermore, the studies of the data reconstruction in 2017/2018 beam tests will be briefly explained.

keywords: Particle physics, High energy physics, Calorimeter methods, Calorimeter, Detector, Energy reconstruction, Particle identification

Contents

1 Introduction 1

1.1 Energy loss in the material . . . 1

1.1.1 Minimum ionizing particle (MIP) . . . 3

1.1.2 Landau fluctuation . . . 3

1.2 Calorimetry . . . 3

1.2.1 Particle shower . . . 4

1.2.2 Electromagnetic shower . . . 4

1.2.3 Hadronic shower . . . 5

1.3 EM calorimeter performance . . . 6

2 The large hadron collider (LHC) and the compact muon solenoid (CMS) 8 2.1 The large hadron collider . . . 8

2.2 The compact muon solenoid detector . . . 9

2.2.1 Coordinate system in the CMS detector . . . 9

2.2.2 Magnet configuration . . . 11

2.2.3 Tracking system . . . 11

2.2.4 Electromagnetic calorimeter(ECAL) . . . 12

2.2.5 Hadronic calorimeter(HCAL) . . . 14

2.2.6 The muon detectors . . . 14

2.2.7 Trigger system . . . 15

3 The CMS phase 2 upgrading and the high granularity calorimeter (HG- CAL) 17 3.1 Silicon as active material in sampling calorimeter . . . 18

3.2 Overview to HGCAL . . . 19

3.3 Motivation for beam test . . . 21

4 Test beams and the H2 beam line 22 4.1 SPS and the target 2 (T2) . . . 22

4.2 H2 beam line . . . 23

4.3 Beam generation . . . 23

5 Beam test prototypes 24 5.1 Module . . . 24

5.1.1 Module in 2016 . . . 26

5.1.2 Module in 2017/2018 . . . 27

5.2 Trigger system of beam test . . . 27

5.3 Delayed wire chamber (DWC) . . . 28

5.4 Setups . . . 29

5.4.1 2016 CERN beam test - configuration 2 . . . 29

5.4.2 2017 July beam test . . . 30

5.4.3 2018 June beam test . . . 32

6 Readout chips 33 6.1 Skiroc2 . . . 33

6.2 Skiroc2cms . . . 33

6.2.1 Skiroc2cms single moudule data taking . . . 34

6.2.2 Skiroc2cms CHIP configuration . . . 35

6.2.3 Skiroc2cms readout chain and DAQ . . . 41

7 Data reconstruction 42 7.1 Data reconstruction in 2016 . . . 42

7.1.1 Pedestal and noise study . . . 43

7.1.2 MIP calibration . . . 45

7.1.3 Gain calibration . . . 45

7.2 Data reconstruction in 2017/2018 . . . 47

8 Monte Carlo (MC) sample generation 50 8.1 Data/MC comparison in 2016 . . . 51

9 Performance of the prototypes 58 9.1 Electron energy calibration by 2016 CERN 8-module data . . . 58

9.1.1 Data set . . . 58

9.1.2 dEdx method . . . 59

9.1.3 Sampling fraction (SF) method . . . 59

9.1.4 Comparison . . . 64

9.2 Pi/e separation from shower-shape variable by 2016 CERN 8-module data 66 9.2.1 Data set . . . 67

9.2.2 2D re-weighting of MC beam profile . . . 67

9.2.3 Selection . . . 68

9.2.4 Electron contamination in pion data . . . 70

9.2.5 Lateral shower-shape variables . . . 71

9.3 Brief look to 2018 CE-E prototype performance . . . 75

10 Conclusion and outlook 76

List of Figures

1.1 -^dE_dx of muon, pion and proton . . . 2

2.1 CMS detector . . . 10

2.2 CMS tracker . . . 12

2.3 CMS ECAL . . . 13

2.4 CMS HCAL . . . 14

2.5 CMS muon system . . . 15

2.6 CMS L1 trigger concept . . . 16

3.1 LHC upgrading schedule . . . 17

3.2 Material design for HGCAL . . . 19

3.3 Simulated radiation dose in HL-LHC . . . 20

4.1 SPS beam status . . . 22

5.1 Module design . . . 24

5.2 Sensor design . . . 25

5.3 Complete module . . . 26

5.4 DWC setup . . . 28

5.5 2016 setup . . . 29

5.6 2017 July setup . . . 30

5.7 2017 July CE-E . . . 31

5.8 2017 July CE-H . . . 31

6.1 Single module test system . . . 34

6.2 Skiroc2cms memory map . . . 36

6.3 Skiroc2cms DAQ . . . 41

7.1 ADC distribution of all types of cells . . . 44

7.2 MIP calibration . . . 45

7.3 HG to LG behavior . . . 46

7.4 Pulse,TS and SCA . . . 47

8.1 Simulated upstream materials . . . 50

8.2 Evisible Data/MC comparison . . . 52

8.3 E1/Evis of 6 energies . . . 53

8.4 E8 divide Evis . . . 54

8.5 SHD calculate by Elayer . . . 56

8.6 SHD calculate by Emax . . . 57

9.1 dEdx EReco . . . 60

9.2 Energy linearity obtained from E_visible . . . 61

9.3 SF⁻¹ . . . 62

9.4 Shower depth . . . 62

9.5 SF⁻¹ slope . . . 63

9.6 SF E_Reco . . . 64

9.7 comp correctness . . . 65

9.8 Final resolution . . . 66

9.9 125GeV pion 2D re-weight . . . 67

9.10 100GeV electron 2D re-weight . . . 68

9.11 Observables before cut . . . 69

9.12 electron contamination in pion data . . . 70

9.13 single layer containment variable . . . 72

9.14 containment sum MC comparison . . . 72

9.15 containment sum for electron and pion . . . 73 9.16 delta-ray candidates Data/MC . . . 73 9.17 E1/E_vis and SHD between 20 GeV e⁻ and delta-ray candidates . . . 74

List of Tables

6.1 Skiroc2cms configuration part 1 . . . 37

6.2 Skiroc2cms configuration part 2 . . . 38

6.3 Skiroc2cms configuration part 3 . . . 39

6.4 Skiroc2cms configuration part 4 . . . 40

9.1 Data statistics - energy calibration . . . 59

9.2 2018 October beam status . . . 65

9.3 Table of statistics - πe separation . . . 69

1 Introduction

The standard model is one of the most successful theoretical model to describe three of the four fundamental interactions to date. It has predict the existence of several fundamental particles, include the Higgs boson, which is discovered in 2012.

Despite the success of the standard model, there are still phenomena which can not be explained: the CP violation, the existence of the dark matter and the dark energy, failing to incorporate the gravitation, etc. It is clearly that the standard model is not the end of the story for particle physics. Detailed measurements are still required to understand the property of the Higgs boson, also the process of probing physics beyond standard model have never stopped.

The technology of the detectors plays an essential role in collecting and selecting the data for potentially interesting physics events. To reconstruct a physics event, most useful information is the position, momentum and energy of the final states of the short-lived particles. Due to the detectors usually have specific function, obtaining these observables needs the cooperation between various kinds of detectors. The calorimeter, which is responsible for measuring the energy, is one of the most indispensable component in a particle physics experiment.

This chapter will go through the vital physical process creates in the calorimeter as well as the property of the machine. Moreover, the principle of the energy detection and the category of the calorimeter will also be discussed.

1.1 Energy loss in the material

Before a particle can be detected, it must undergo some sort of the interaction. By observing the aftermath of the interaction, the information of the original particle can be retrieved. During the penetration of the particle through matter, energy of the particle is deposited in the material. Hence, measuring the energy loss of particle in certain material is an imperative task in particle detection.

When a heavy charged particle(heavier than electron) pass through a material, it deposits energy through Coulomb interaction. The interaction will cause the excitation or ionization of atomic electrons in the material. In particle physics, the factor -dE/dx is chosen to describe such an effect, which is the energy loss per unit length.

In the early 1930s, Bethe and Bloch treat the energy loss problem in the framework of quantum mechanics. Improved from Bohr’s calculation[22], they have deduced an equation that is valid for heavy charged particles (M_incident >>M_electron). This equation is now known as Bethe-Bloch equation:

< −dE

dx >= 4π

m_ec² · n_eZ_i² β² · ( e²

4π₀)²·

"

1

2ln(2m_ec²β²γ²T_max

I² ) − β²− δ(βγ) 2

#

(1.1)

This equation is valid for a particle with spin 0, speed v and charge Z_i electron charge incident into a material with a mean ionization potential I. Here m_e stands for electron mass, c is the speed of light, n_e is the electron density of the material and β = v/c, γ = 1/^q(1 − β²) as the common notation in relativity. T_max is the maximum kinetic energy which can be imparted to a free electron in a single collision, which is approximately 2mec²β²γ² in this case. δ term in the very last is the density effect of the absorber. The behavior of the equation is shown in Figure 1.1.

Figure 1.1: -^dE_dx of muon, pion and proton

At first, the curve behaves like β⁻². The curve then stop dropping at a point where βγ ∼ 3. After the minimum point, the curve starts to rise with a very small slope ∼ ln(βγ)². Picture from [19].

As shown in the plot, the minimum point of the formula is independent of the particle type. Particles around this region is called minimum ionizing particle (MIP).

1.1.1 Minimum ionizing particle (MIP)

As mentioned previously, the MIP particle is defined by the Bethe-Bloch equation. Due to its special feature, the MIP particle can be viewed as a standard candle in studying the detector response. The response of MIP is used as a bridge to calibrate the detector signal (usually in ADC, which indicates the height of a analog pulse) with real physics energy (MeV).

In the practical point of view, muons are usually treated as the MIP particle because muons interact by pure ionization and do not participate in any showering process^†.

1.1.2 Landau fluctuation

While the average behavior for a charged particle pass through a fixed thickness material was mentioned previously, the energy deposition will vary due to the statistical nature of the interactions with individual atoms in the material. In the process of penetration, small energy transfer happens more frequently than the large energy transfer. The events of large energy transfer is mainly due to a head-on collision with the atomic electron or from nuclear interaction. As a result, the energy loss distribution become asymmetric.

The distribution is biased to the low energy part, which can be well described by Landau- Vavilov (or Landau) distribution.

The most probable value of the energy loss in a thickness x material is given by the equation:

W_mp= ξ(lnξ

⁰ + 0.198 − δ), with ξ = 2πn_eZ_i²e⁴

mv² x (1.2)

Where δ is the density effect of the material and ⁰ is the cutoff on low energy transfer for a single interaction. The parameter ⁰ was chosen by Landau so that the mean energy loss will agree with Bethe-Bloch equation.[22]

1.2 Calorimetry

The main subject of the calorimeter, is to measure the total energy of the incident particle via total absorption. During the absorption the particle energy is transformed into a measurable quantity, such as charge and light.

Most of the calorimeters are position sensitive, as a result they are usually segmented with independent signal readout. From the lateral geometry, calorimeters can be catego- rized again into two kinds: the homogeneous calorimeter and the sampling calorimeter.

For a ”homogeneous” calorimeter, it is segmented only in the lateral design, where one material serves to absorb the energy and in the mean time to provide the measureable

†The shower process will be discussed in Chapter1.2.1

signal. For those calorimeters which have both lateral and longitudial segments, they are

”sampling” calorimeter. The sampling calorimeter has two kinds of layers, the absorber layers which absorb most of the energy and expand the shower (usually dense materials with high atomic number, passive material); the active layers which measure the energy signal. While almost all energy are converted into signal for the homogeneous calorimeter, only a fraction of the energy contributes to the measured signal in sampling calorimeter.

The energy measurement of the photon, electron, taus and jets as well as the iden- tification of the muon track can be done by the calorimeter. Moreover, for a hermetic calorimeter, finding the missing transverse energy in the experiment is also possible.

1.2.1 Particle shower

While the Coulomb interaction of the heavy charged particle is explained by equation1.1, the dominant process for energy loss of higher energy particle will no longer dominant by the ionization and excitation. (For electron, the limit is ∼ 100M eV .) In such energy scale, the incoming particle interacts with the matter and create new particles with lower energy. The newly produced particles interact in the same way and hence create a cascade of secondary particles, the whole process is called ”particle shower”. The shower process will continue until the energy of each member is lower than a threshold and the shower is then slowly dies out as particles are absorbed by the material. The shower process could produce millions or even billions new particles (depends on the energy of the incident particle) before it stops.

Based on the type of the incident particle, the shower can be separated into the electromagnetic shower and the hadronic shower. The former is produced by particles (usually electron and photon) interact primarily via the electromagnetic force. The latter is produced by hadron, which mostly interact via strong interaction. Due to the different feature of the shower, the calorimeter is usually designed as two parts: the electromagnetic section and the hadronic section.

Various of parameters are invented to describe the showering process, include the radiation length (X₀) and interaction length (λ_I), critical energy, Moli´ere radius, etc.

1.2.2 Electromagnetic shower

The electromagnetic (EM) interaction has been studied for hundreds of years, within several epochal theories tried to explain this fundamental interaction, the quantum elec- trodynamics (QED) is the most successful one.

As the electron and photon are the dominant component in the EM shower, the process of these two particles are listed in the following paragraph. The phenomena of the EM process for the electron in the matter includes the loss of energy due to excitation or ionization of atomic electrons, the energy loss due to photon emission (bremsstrahlung) and the Coulumb scattering of charged particles with atomic nucleus. The interaction of photon, include the photoelectric effect, Compton scattering and pair production are also

describe completely by the QED.

In spite of the complex of shower development, the main shower features can be describe by simple empirical functions. Here 3 parameters are introduced: the radiation length, the cutoff energy and the Moli´ere radius. The radiation length (X₀) is defined to provide an estimation for the EM shower. The radiation length of a material is the mean length to reduce the energy of an electron by the factor of 1/e, where e is the base of the natural logarithm. And based on the same X₀ definition, a photon beam reduce its energy by 1/e after 9/7X₀.

The X₀ can be approximate by the following formula:

X0(cm) = 716.4A Z(Z + 1) · ln^√287

Z

· 1

ρ (1.3)

Where A and Z is the atomic number and atomic weight. And ρ is the density of the material. For a compound material with

multiple ingredients can also be calculated (refer to [18]).

The energy loss of a high energy electron is dominant by the bremsstrahlung process, while for high energy photon the energy loss is dominant by pair production. As the energy drops, the energy loss of electron will gradually taken over by ionization. In the end the shower dies out. The critical energy (E_c) is defined as the energy where ionization takes over, which can be estimate by:

K

Z + K_c (1.4)

Where K and K_c are constant related to the material.

The lateral information of a shower can be describe by the Moli´ere radius. It contains 90 % of the shower, which can also estimate mathematically by:

RM = 21.2(M eV )X0

E_c (1.5)

All these parameters are very useful when designing the electromagnetic part of a calorimeter, which helps decide the material and the size of the segments.

1.2.3 Hadronic shower

The hadronic shower is more complicated, as it relies on the incoming particle interact with nucleus via the strong interaction, which describe by the quantum chromodynamics (QCD). The secondary particles produced by strong interaction can be again separated into the EM part (generated from π⁰ component) and haronic part (p,n,π⁺).

Due to the small cross-section for the strong interaction, the hadronic shower always starts deeper in the calorimeter. The hadronic shower is also characterized by event-by- event fluctuations and with much broader spread compare to the EM shower.

The nuclear interaction length (λ_I) is used to estimate the behavior of hadronic shower, which is related to the mean free path of a particle traveling in the sea of nucleus. The definition of λ_I is:

λ_I = A NAσtotal

(1.6)

The N_A is the avogadro constant, the A is the atomic weight and σ_total is the total cross section for the strong interaction.

Due to the characteristic of the hadronic shower, the hadronic part of the calorimeter usually composed with thicker absorbers and in general, placed outside the EM part of the calorimeter.

1.3 EM calorimeter performance

The criteria of judging the performance of a calorimeter has covered lots of aspects: the precision of the energy measurement, the degradation of the active material response after radiation damage, the radiation hardness of the electronics, the power consumption of the detector, etc. Among all of these aspects, only the precision of the physics measurement will be mentioned below. While some of the discussion can also be applied in the hadronic calorimeter, here the discussion will focus on the EM calorimeter.

The detector linearity, which means the detector response should be linear in all the interested energy region. For a homogeneous calorimeter, the amplitude of the collected signal is propotional to the incident particle energy, the calibration is straight-forward.

However, for the case of a sampling calorimeter, different methods can be developed based on the usage of the information obtained from the slice of the shower. Different active layers can be treat with corresponding weight, and the estimation for the number of MIPs in each active layer can be utilize as a tool to trace back the information of the absorber layer, particle identification from the shower shape can also help the procedure for the energy reconstruction.

Energy resolution of a calorimeter is another important parameter in calorimetry.

The energy resolution is usually defined as σ_E/E in a given beam energy (E). Hence the resolution is usually shown as a function of the energy. The dominant term in the energy resolution is usually due to sampling fluctuations, which is Poisson distribution in nature.

For an ideal calorimeter, the intrinsic energy resolution can be expressed by:

σ_E

E = a

√E (1.7)

However, in the real calorimeter the detector response should also be considered. For example the noise of the electronics, the imperfection in the mechanical structure, the

temperature gradient, etc. Due to such effects, the energy resolution is then become the form:

σ_E

E = S

√E ⊕ N

E ⊕ C (1.8)

Where the S,N and C are all constants, and the ^Lis the quadratic sum.

The S is the stochastic term, which relates to the fluctuation in the physical development of the shower. The N is the noise term, which mostly related to the electronics noise in the readout chain. The last C is the constant term, which is mainly judged by the quality of the calorimeter. The constant term is probably the most important term in the detector design, since it is mainly effected by the radiation damage, the detector geometry or the non-homogeneous temperature distribution of the detector.

2 The large hadron collider (LHC) and the compact muon solenoid (CMS)

The Large Hadron Collider (LHC) is currently the most powerful and largest particle ac- celerator in the world. Using the tunnel which previously served Large Electron Positron (LEP), the LHC is now lying 175 m underground at the France-Switzerland border with the circumference 27 km. Regarding its scale, the LHC is a marvelous machine accom- plished by European Organization for Nuclear Research (CERN) with the efforts of 10,000 scientists and engineers from over 100 countries.

Designed to maintain proton (or heavy-ion) beams circulating in the opposite way, the LHC has created four main interaction points where the protons collide. Situated in one of the interaction point, the compact muon solenoid (CMS) works as a main experiment in the LHC to measure the physical quantities originated from the proton-proton collision (pp collision) or lead-lead heavy-ion collision.

A brief overview of the LHC and the CMS will be introduced later in this chapter.

For the LHC part, only the process of producing the proton beam will be mentioned.

Moreover, the main focus will lie on the idea of design and operation setting of these great machines.

2.1 The large hadron collider

As mentioned previously, the whole LHC is placed 175m underground. This decision allows the LHC to effectively decrease the bias from cosmic background, though, posting tough engineering condition during the construction.

The main target of the LHC is to produce two opposite proton beams circulate in the beam pipe and make the bunches collide at interaction points. Various magnets are used to direct the beams around the accelerator. These magnets have 3 main purposes:

bend the beam, focus the beam and squeeze the beam just before the interaction points.

Thanks to the magnets, in the early stage of the LHC operation, 3.5 tera-electronvolts (TeV) proton beam was generated. In 2015 - 2018 the beam energy will be further raised to 6.5 TeV and colliding with center-of-mass energy of 13 TeV.

Before the proton beams reach the desirable energy, these groups of protons (also called bunches) have to undergo multiple stages of acceleration.

In the first stage which take place in Linear accelerator 2 (Linac 2), protons are obtained by sending hydrogen (H₂) gas through an electric field to strip off its electrons.

Then these protons are injected to the Linac 2 to accelerate to an energy scale of 50 MeV.

In the second stage, the beam obtained from the Linac 2 has been sent to the Proton Synchrotron Booster (PSB), which is the first circular accelerator in the stage. The PSB accelerates the beam from 50 mega-electron volt (MeV) to 1.4 giga-electron volt (GeV) for injection into the Proton Synchrotron (PS). PS further push the beam energy to reach the scale up to 25 GeV and passing the beam to the Super Proton Synctron (SPS). The

SPS is the final stage before filling the protons into the LHC, here, the proton beam can be accelerate to up to 450GeV.^† At the stage of the LHC, the beams are separated into 2 opposite directions to prepare for the collision. Finally, the LHC main ring raise the speed of the proton beams to 0.999999 speed of light, correspond to 6.5 TeV designated energy.

In the LHC main ring, the protons are separated into maximum 2556 bunches.[1]

Each of the bunch contains 1.15 × 10¹¹ protons, with a time interval of 25 ns between each bunch crossing at the four interaction points (CMS, ATLAS, ALICE and LHCb).

Due to frequent collision-rate and density of the beam (1.15×10¹¹ protons per bunch), the average luminosity at the interaction points achieves an order of 10³⁴cm⁻²s⁻¹. The collisions happens at each bunch crossing is defined in the term ”pileup”. During the period from early 2015 to nearly the end of 2018, the pileup is estimated to be 25 ∼ 60.

2.2 The compact muon solenoid detector

The CMS detector is a multi-purpose apparatus for particle detection. The prime purpose of this detector includes studying the Standard Model (search of the Higgs boson), prob- ing beyond standard model physics such as supersymmetry or extra dimensions, even revealing the mystery of dark-matter. While the search of the Higgs boson has been accomplished in 2012, more detailed measurement are still ongoing.

In a shape of a hollowed cylinder, the detector can be separated into the barrel part and the end-cap part. The dimension of the detector is 15 m in diameter, 21 m in length.

Sub-systems are contained in the CMS detector. The superconducting solenoid, which generates an uniform magnetic field parallel to the beam and contains most part of the detector. For the sub-systems measuring particle momentum and energy, start from the inner most region lies the tracking sub-system. Calorimeters are placed outside the tracking sub-system to measure the energy of the particle. Finally, the muon system, locates outside the solenoid, which provide the information of muons. The CMS detector layout is shown in Figure 2.1.

2.2.1 Coordinate system in the CMS detector

The CMS is a cylindrical shape apparatus, which is symmetric in the z axis. Consider the physics events generated from the interaction point, it is reasonable to use the (r,θ,φ) spherical coordinate system to describe both the physics event and the detector system.

However, in the hard process occurring at the TeV scale, the center-of-mass of the event is likely to be boosted in the z axis. This effect will make the ∆θ measurement

†SPS is not only the beam source to the LHC, but also some important experiments. Moreover, the SPS also pass the proton beam to the north area where generates the beam source for the HGCAL beam tests.

Figure 2.1: CMS detector plot from [2].

between two fragments originate from the interaction point no longer Lorentz-invariant.

Thus, a Lorentz-invariant measure, the rapidity (y), is introduced to deal with the boosts in the z axis.

y ≡ 1

2ln(E + cp_z

E − cp_z) (2.1)

The E is the energy of the final state, p_z is the momentum alone the z axis and c is the speed of light.

Moreover, the energy of the final state is high compare to its rest mass in a hard process of TeV scale, namely E = mc²γ >> mc². The approximation E → c|p| can be applied, which gives a new measure called the pseudorapidity (η).

y ≡ −ln(tan(θ

2))) (2.2)

The θ is the polar angle measured from the beam axis.

Usually, for precise physics description of angular distribution for two tracks, the rapidity is used (denoted as ∆_ij). While for describing the detector geometry or fast computation of physical quantities, the pseudorapidity is used. In this case the angular distribution for two tracks is expressed in ∆R.

2.2.2 Magnet configuration

The super conducting coil is designed to be 4 tesla, but operates with 3.8 tesla in currently stage (2018). The strength of the magnetic field can help to bend the trajectory of the charged particle. Since the field is designed parallel to the beam direction, the transverse momentum of a charged particle can affect the behavior of the particle path. For the particles with smaller transverse momentum, the trajectory is bent more. As a result, reconstructing the particle path can provide a measurement of the momentum.

2.2.3 Tracking system

The tracking system of the CMS detector aims to measure the trajectory of charged particles and provides the reconstructed information about the secondary vertices. This silicon-based sub-system, is composed of the pixel part and the silicon strip part. The decision of the cell size or type is mainly driven by the radiation flux and the desirable position resolution.

Silicon pixel detector situated in the inner-most region, is designed in a size of shoebox.

With a 2D sensor component of 100 × 150 µm cells by the original design. With a coverage

up to |η| ∼ 2.5 and such fine 2D cells, the detector provide high spatial resolution (∼ 15 - 20 µm) around the interaction point.^†

The strip detectors (TIB,TID,TOB, and TEC in Figure2.2), placed right outside the pixel detectors, are composed of three different sub-systems. The Tracker Inner Barrel and Disks (TIB/TID), the Tracker Outer Barrel (TOB) and the Tracker End-Caps(TEC).

At intermediate radii (20cm < r < 55cm), strips with 10cm×80 µm are used, while in outer region (55cm < r < 110cm) the strip pitch can be further loosen. Different thickness of silicon is also considered based on the radiation flux of the position.(500 µm thickness at outer tracker region and 320 µm in the inner tracker) The ultimate acceptance of the tracker ends at |η| ≈ 2.5 as the pixel detectors coverage.

Figure 2.2: CMS tracker plot from [2].

2.2.4 Electromagnetic calorimeter(ECAL)

The electromagnetic calorimeter (ECAL) is a sub-system to measure the energy of parti- cles through the electromagnetic showering process (mainly photon and electron). Both barrel part of ECAL(EB) and end-cap part of ECAL take lead tungstate (PbWO4) crys- tals as the material for detection. In addition, a preshower detector is placed in front of the end-cap crystals. Shown in Figure 2.3.

The EB covers the range |η| < 1.479, with the crystal cross-section ≈ 0.0174 × 0.0174 in η - φ. The length of the crystal is 230 mm, which corresponding to 25.8 X0, for the crystals in the ECAL end-cap (EE), the length decrease to 220 mm (24.7 X0). The EE covers the region from 1.653 < |η| < 3.0.

†The pixel modules are actually upgraded in 2017, so the current cell size could be smaller than this number. Smaller sensor size is more radiation-tolerant and provide more precise measurement. Also the layer number in the barrel part is now 4 instead of 3, the end-cap detector has also been upgraded. See [3] for the upgrading information.

The CMS Preshower detector, which is placed in the region 1.653 < |η| < 2.6, is made of two planes of lead followed by silicon sensors. The aim for this detector is to identify neutral pions in the end-cap. Moreover, it can help the identication of electrons against MIPs while improveing the position determination at the same time^†.

Figure 2.3: CMS ECAL plot from [4].

†The preshower strips are 2 mm wide, which is finer than the 3 cm-wide ECAL crystals.

2.2.5 Hadronic calorimeter(HCAL)

In the ECAL region, only the photon and electron will be fully captured while hadronic particles and muons can easily pass through ECAL with almost no interaction. Thus, the hadronic calorimeter (HCAL) is designed to capture the rest of the particles.

The schematic plot for the CMS HCAL is shown in Figure2.4. The CMS HCAL can be separated into three parts. The HCAL barrel (HB) and outer barrel (HO) cover |η| <

1.6, while the HCAL end-cap (HE) covers 1.5 < |η| < 3.0 and forward calorimeter (HF) covers the region 3.0 < |η| < 5.2.^†

The HCAL is deigned as a sampling calorimeter, which use steel and brass as absorbers and scintillator layers as readout layer to form the sandwich structure. The steel is used at both the first layer and last layer to form the supporting structure, and brass are placed in the middle of the steel planes. The HB effective thickness increases with polar angle (θ) as 1/sinθ, resulting in 10.6 λ_I at |η| = 1.3. In the end-cap case it’s around 10 λ_I combine with the ECAL end-cap.

Figure 2.4: CMS HCAL, plot from [2].

2.2.6 The muon detectors

Muon has a characteristic that can penetrate through several meters of absorber (ECAL and HCAL) with only few interactions. Thus, the muon detector is placed in the outermost part of the CMS. Due to a large area needs to be covered by the muon sub-system, gaseous detectors are widely considered as the design. The concept of muon detection is to probe

†Note that the HB, HO, HE do share some coverage. For |η| < 1.3 region, HB has full coverage.

the signal produced from ionized gaseous molecule, which is caused by muon passing through.

The CMS muon system has 3 functions: muon identication, momentum measurement and triggering. Thanks to the solenoid and the flux-return yoke, good muon momentum resolution and trigger capability can be achieved.

There are 3 main detectors in the muon sub-system. The barrel drift tube (DT) chambers measures the muon particle in the barrel region with |η| < 1.2. For the end-cap region, the Cathode strip chambers (CSC) are designed as a 2D array with anode wires and cathode strips which are perpendicular to each other for fast measurement. The CSC covers 0.9 < |η| < 2.4. The last component, the resistive plate chambers (RPC), covers |η| < 1.6, provides an independent measurement from DT and CSC. The CSC is a dedicated trigger system which operated in avalanche mode to ensure good operation at high rates. In Figure 2.5, the |η| coverage of the muon system and the schematic plot of the 3 detectors are shown.

Figure 2.5: CMS muon system plot from [5].

2.2.7 Trigger system

As mentioned previously, the LHC beam crossing interval is 25 ns, correspond to 40 MHz.

It is impossible to readout all the data produced from each hardware unit. As a result, a trigger system is designed in the CMS to reduce the event rate needs to be stored.

Undergo two steps of event selection, Level-1 (L1) Trigger and High-Level Trigger (HLT), respectively. The potentially interesting events are selected and the data size is reduced dramatically by the trigger system. The L1 Trigger reduced the event rate to 30kHz and the HLT further reduced the rate down to minimum 100 Hz.

The L1 trigger consists of the programmable electronics. It has local, regional and global components. See Figure 2.6. The local component is based on energy deposits in calorimeters and track segments as well as the hit patterns. The regional component combine these information and use pattern logic to rank and sort these objects. The global trigger has the right to reject or accept the event and it will pass the information to HLT for further calculation.

The HLT trigger is a software system composed of a 1000 computers farm. It performs complex calculation to the complete readout data. There are many algorithms can be selected for reconstruction physics objects, thus the HLT has a rate ranges 100 ∼ 20,000 Hz.

Figure 2.6: CMS L1 trigger concept plot from [2].

3 The CMS phase 2 upgrading and the high granu- larity calorimeter (HGCAL)

As mentioned in Chapter2, the LHC is currently producing the proton beam and colliding with a center-of-mass of 13 TeV with averaged luminosity of 1.7 × 10³⁴cm⁻²s⁻¹ (2018).

However, this will no longer be the case in the future. Start from the early stage of the LHC, various of upgrades has been done to improve the collision center-of-mass energy and to raise the luminosity. The history of the LHC operation as well as the target for future can be found in Figure 3.1.

Figure 3.1: LHC upgrading schedule

The upper time line describe the energy scale as well as the beam status of the LHC. The bottom time line labels the improtant milestone for HL-LHC design. Picture from [14].

By the end of 2018, the LHC has the second shut down (LS2) to further raise the collision energy to 14 TeV. At the mean time, the experiments include the CMS will also upgrade the detectors. In the LS2, the so called phase-1 upgrading will be performed to make the apparatus cope with the condition in the Run3. The overview of phase-1 upgrading for the CMS experiment can be found in [15].

The next upgrade for LHC is planned in the third long shutdown (LS3), which will take place from 2024 ∼ 2026. In the LS3, the LHC will further upgrade its luminosity to 5 × 10³⁴cm⁻²s⁻¹. This upgrade will open a new era of operation in the LHC, called high- luminosity LHC (HL-LHC), or the phase 2 operation of the LHC. The phase 2 upgrade program of the CMS can be found in [16].

The upgrade of the luminosity will definitely pose challenges to the whole experiment,

include the radiation damage to the detector (includes the on-detector electronics) and the high pileup condition. Moreover, the integrated luminosity in HL-LHC era is ten times more than the current LHC, which will degrade the performance of the current ECAL and HCAL to an unacceptable level. Thus, a replacement of the CMS calorimeters is planned and will take place in the LS3. The high granularity calorimeter (HGCAL) is the chosen technology for the replacement of the end-cap part of the calorimeter.

HGCAL is designed as a sampling calorimeter in both electromagnetic and hadronic section. A considerable percentage of the HGCAL will utilize silicon as the active ma- terial, and the scintillator with direct SiPM readout will be used in the rest part of the calorimeter. This chapter will first visit the feature of the silicon and briefly explain why it is chosen as a component of the HGCAL. Then a overview to HGCAL will be given, as well as the motivation of the HGCAL beam tests.

3.1 Silicon as active material in sampling calorimeter

The semiconductor detectors, or the solid-state detectors have long been used in the particle physics since late 1950’s[23] . Among all the semiconductor material, the sili- con is particularly useful when designing a particle tracker, which requires high position resolution.

The principal of particle detection in silicon detector is quite different from other crystalline material. The doped silicon needs a bias voltage to create a depletion region by removing all free charge carriers in the area. The thickness of the depletion region has a strong correlation with the charge collection efficiency in silicon. As the bias voltage raises, the charge collection efficiency of silicon increase until a limit where the full depletion is achieved. In practical, a voltage higher than the full-depleted voltage will be supplied to the silicon and generates a stable electric field for charge collection. The passage of ionizing radiation in silicon creates electron-hole pairs, which are then collected by the electric field.

The characteristics of the silicon as a active material are listed:

• High energy resolution: The average energy required to produce a electron-hole pair is small^†, which points to good energy resolution.

• Fast response time: After the electron-hole pair is created, finishing the charge collection cost only an order of few nanoseconds.

• High radiation tolerance: Due to the silicon has been widely used in the tracking system of plenty of the experiments. The efficiency drop of the silicon after radiation has been well-understood and proven to be endurable even in the HL-LHC radiation environment.

These feature allow the silicon sensor to take place in the calorimeter design as the main challenge of the HL-LHC era is the pileup and radiation.

†An order of magnitude smaller than the gas ionization in gaseous detector.

3.2 Overview to HGCAL

As the new end-cap calorimeter for the CMS, the HGCAL will cover the region from 1.5 < |η| < 3.0 region. The calorimeter is again separated into the electromagnetic and hadronic sections, namely the calorimeter end-cap electromagnetic section (CE-E) and calorimeter end-cap hadronic section (CE-H). The material in active layers and absorber layers are different from the CE-E and CE-H.

The whole system will be cooled to a temperature of -30 ^oC to ensure the level of the electronics noise resulted from the increasing leakage current as well as to guarantee the lose of charge collection efficiency stay acceptable after irradiation.

A detailed schematic plot as well as the basic property for the HGCAL is shown in Figure 3.2.

Figure 3.2: Material design for HGCAL Picture from [6].

In order to withstand the radiation level in HL-LHC, the HGCAL is designed as a silicon and scintillator based calorimeter. In the high radiation region, silicon layers are used due to its ability to withstand the high radiation. For the low radiation region, the scintillator with direct SiPM readout is used. The boundary of the two material is decided using the FLUKA simulation[17] shown in Figure3.3. The highest fluence is around 10¹⁶ neq/cm^{2 †} and the highest dose is around 2 MGy.

Furthermore, the HGCAL is designed as a unprecedented sampling calorimeter with 28 layers in CE-E and 24 layers in CE-H. Aims to provide good two-shower separation, the lateral cell size (granularity) has to be smaller than the Moli´ere radius, which is 2.8 cm in CE-E. The cell size of all active layers are decided to be 0.5 ∼ 1 cm.

Figure 3.3: Simulated radiation dose in HL-LHC Picture from [7].

In addition to the challenges, the interested physics will be the detailed study for Higgs boson and SM processes, the searches for physics beyond the SM include reactions initiated by vector boson fusion (VBF) and highly boosted objects. In the HL-LHC period, one of the goal is to trigger on the narrow VBF jets, as well as merged jets.

Hence, good jet identification/measurement will also be a crucial task for the detector.

†Using 1 MeV equivalent neutrons per cm²

3.3 Motivation for beam test

To validate the idea of the HGCAL design and to test the electronics circuit that will be used in the phase 2. Various research and design are undergoing, such as the development of the sensor, front-end electronics, back-end electronics and the HGCAL beam tests.

Several beam tests were done since 2016. The prototypes with the official absorber material were built. The active silicon sensors are composed into real modules which have their own data acquisition and trigger system. The whole setup is then tested by the known energy particle beams to study the performance of the official design.

The main goal of the beam test is to systematically understand the design of the module, and to get good agreement between data and Monte Carlo simulation in both well-known electromagnetic shower and the complicate hadronic shower. In the end, the beam test is expecting to proof the power of the HGCAL.

4 Test beams and the H2 beam line

Test beams are the particle beams generated in certain purpose with usually a fixed energy or the high purity of specific kind of long-lived particle. These beams are made to provide a reference to study the performance of new apparatus.

The HGCAL prototype has been tested in different beam facility started from 2016, includes FNAL, DESY and SPS H2 and H6 beam line.^†

This chapter will provide a basic introduction for the main beam facility used by HGCAL prototype, the H2 beam line. In addition, a more detailed description of the SPS will be given as it is the upstream source of the whole test beam.

4.1 SPS and the target 2 (T2)

The Super Proton Synchrotron (SPS) is the second-largest machine in CERN’s accelerator complex mentioned in Chapter 2. With 7 km circumference, the particles came from the Proton Synchrotron(PS) are accelerated in the SPS and used as the source of both LHC and various famous experiments, such as NA61/SHINE and COMPASS.

Besides the proton, particles like sulphur and oxygen nuclei, electrons and positrons can also be provided depends on the demand of the experiments down stream.

The general beam status in SPS can be explained by a 45 ∼ 60 second ”cycle”. Within a cycle there are usually 2 spills with a duration for approximately 5 seconds, as shown in the Figure 4.1. This status is important reference during the data taking. Test beam users can judge the information from the monitor to see if the beam is received in the upstream.

Figure 4.1: SPS beam status

The yellow curve is the relative particle luminosity, while the white curve is the magnetic field upstream. The perpendicular line is the updating bar, which goes from left to right and complete a cycle. Plot from [10].

†The H6 is once used in the end of 2017, rest of the beam tests at SPS took place in H2 beam line.

The extracted 400 GeV proton beam from the SPS is transported over 1 km in the help of magnets and then split into three parts, each one is directed towards specific primary targets where secondary particles are created. There are 4 fixed targets: T2 and T4, T6, T10, while the first two produce beams for the EHN1 (Experimental Hall North 1) and received by H2,H4,H6 and H8 beam line. The EHN1 is where the HGCAL beam test took place and H2 beam line is located down stream of the target T2. The T2 target station has several different lengths beryllium (Be) plates. The target plate is chosen based on the requested secondary particle momentum and type[24].

4.2 H2 beam line

After the secondary beam is generated in the T2, the H2 beam line transport the mo- mentum selected secondary particles to the EHN1. Users at H2 can further optimize the beam status by modifying the magnets and the collimators in the beam line.

The H2 beam line features by its high-energy and high-resolution beam. There are various particles with different momentum can be selected: protons, charged pion, muon and electron. The electron purity can vary from 10 ∼ 99.5 percents based on the request momentum and the the optimization of the beam line.^† The maximum ∆p/p acceptance of the line is 2 percent.

4.3 Beam generation

Electron/positron, muon and charged pion are selected to test the performance of the HGCAL prototype. These long lived particles are generated from different process.

• Electron: A secondary e⁺ or e⁻ beam can be produced by producing π₀ in a thin target. The decay chain is π₀ → 2γ → 2e⁺+ 2e⁻.

• Charged pion: Generated from incident proton into the fixed target.

• Muon: Produced by the charged pion decay. In 2016 test beam the pion beams are pointed to a thick concrete, thus an unknown energy muon beam is generated.^‡

†Usually higher momentum has lower purity in both electron and pion case.

‡Different methods are used afterward as in 2017 and 2018 muon beam with fixed energy are also available.

5 Beam test prototypes

5.1 Module

A single module in the official HGCAL design is comprised of 4 parts, which are shown in Figure 5.1. The basic function of the 4 components are listed below:

• Base plate: The base plate has a 3 in 1 function in the test beam. First, it provides the mechanical rigidity to the module and prevent the fragile silicon sensor from being distorted or bent. Second, it is the thermal path between the PCB and the cooling plate, which helps to send out the heat produced from the units on the PCB (Chips, FPGAs...). The last function is that it also serves as an absorber in the beam if it is made by the metallic material.^†

• Kapton sheet: A gold-surfaced sheet to allow biasing of the back side of the silicon sensor while also served as a insulation between sensor and the base plate.

• Silicon sensor: The active material with ∼ 1 cm in diameter hexagonal cells. There are 133 or 134 cells on a full 6-inch sensor. Some of the cells have difficulty to wire bond, so these cells are ”merged” to the nearby cell which share the same readout.

Non-complete hexagonal cells around the edge are also specially treated, to match the final 127 channels in the readout chain. There are 4 cell types categorized based on the shape of the cells, which are inner calibration cell/outer calibration cell, mouse bite, the full hexagonal cell and half cell. See Figure 5.2.

• Printed circuit board(PCB): This component is where the readout chip and the FPGA mounted. With aluminium wire bonds through holes in the PCB, analog signal collected from the silicon sensor are routed to the readout chips. The digiti- zation is done in the readout chip.

Figure 5.1: Module design, plot from [7].

†In 2018 some of the module used PCB as base plate.

The composition is the same in all HGCAL test beams while the difference between 2016 and 2017/2018 module will be visited later.

Figure 5.2: Sensor design, plot from [6].

This picture shows the dimension of the silicon sensor and different cell types are labeled. Although the inner calibration cell is not always existing, the sensor layout stays the same in all the beam tests. The thickness of the active region, however differs from 2016 to 2017/2018.

The complete module of the 2016 and 2017/2018 is shown in Figure 5.3.

In general, the silicon sensors and the hexaboard are hexagonal with small cutouts at each of the six corners. The cutouts provide access to the positioning and mounting holes in the base plate. They also provide access to a portion of the Kapton-Au layer for wire bond connections to the hexaboard, for the biasing of the sensor back-plane.

Figure 5.3: Complete module

The 2016 module is shown in the left, plot from [6]. The 2017/2018 module is shown in the right.

The inner/outer calibration cell is a special design to cope with the radiation damage in the HL-LHC era. On the cell it is separated into 2 parts: the inner calibration cell which is designed 1/9 th smaller in the area in order to see MIP signal in the full run.

The outer calibration cell which take the rest of the area. The inner calibration cell can serve as a reference after the big cells suffer from radiation damage and begin to lose the efficiency. The result obtain from inner/outer calibration cell will then be applied to other cells on the sensor and guarantee the performance of the module.

Although the over all introduction to a module is described here. There are still period based design in 2016 and 2017/2018, which will be surveyed afterwards.

5.1.1 Module in 2016

For the 2016 module, most information are retrieved from [6]. In this period, all the base plates are made by 25 % copper and 75 % tungsten(CuW). These base plate are made in 3 thicknesses: 0.8 mm (∼ 0.15 X₀), 1.2 mm (∼ 0.25 X₀) and 1.6 mm (∼ 0.5 X₀).

For the sensor part, the mouse bite cells are two half cells joint together and share the readout. The 2016 sensor is made 320 µm physically but with a depleted thickness of 200 µm.

The PCB design in 2016 is double layer PCB, which locates each side of the sensor. The first PCB connect the silicon sensor from wire bonds and collect the signal. The second PCB serves as readout PCB, which connected to the first, has the front-end electronics mounted. This readout PCB also transmit the digitized data to the outside world. There

are 2 ASICs (chips) used on the PCB, each of them response for 64 sensor cell.

5.1.2 Module in 2017/2018

In 2017/2018, different materials are used in the base plate include pure copper base plate, PCB base plate, copper with kapton and the rest are again copper-tungsten.^† With different material in the base plate, multiple tests related to the total noise and the common mode noise are performed. Most of the base plate are 1.2 mm, while those with special design are 2.4 mm.

The mouse bite cell which made by joining two half cell are no longer fully readout, which means only one of the half cell is readout. The overall channels in the readout chain remains 127.

The PCB design is single layer PCB, which can collect the signal from the wire bond and directly transmit the analog signal to the front-end electronics on the same PCB.

There are 4 ASICs (chips) used on the PCB, each of them response for 32 sensor cells.

5.2 Trigger system of beam test

The trigger of the HGCAL beam tests usually consists of two scintillators photo-multiplier (PMT) aligned with the module center. The biggest scintillator is 10 × 10 cm, which is smaller than one full module, this is good for selecting the events in the center while also leads to a result that the peripheral cells have too few statistics in the muon run to perform MIP calibration.

In the setup with only CE-E stack or the pion/muon driven beam tests, the scintillators are placed in the distance of around 40 cm. This distance is usually enough to select good events with straight incident. The signals from all PMTs are fed into a coincidence unit.

If the coincidence condition is achieved, the output signal from the trigger system is then fed into the data acquisition (DAQ) system. The DAQ system executes the command of the readout and storage process.^‡

†Usually the grounding path differ from material to material.

‡Note that in Skiroc2cms chip, there is no buffer design. The latency of the trigger path should be short in order to finish the trigger process in 275 ns.

5.3 Delayed wire chamber (DWC)

The delayed wire is the most commonly used supporting detector in the HGCAL beam tests at H2 beam line. The DWC is built in an aluminium case of 220 × 220 × 56 (mm), with two kapton windows of 110 × 110 (mm) and a thickness of 25 µm. The picture of DWCs is shown in Figure 5.4.

Figure 5.4: DWC setup

The supporting frame and the DWC is shown. There are 4 signal readout and a high voltage cable come from the back of the DWC. The functioning part is labeled by the red frame.

The DWC is a gaseous detector which works like the Multi Wire Proportional Chamber (MWPC). When a particle path through, the ionized gas will be accelerated by the high voltage between the thin cathode and anodes. At the end, the avalanche multiplication happens around the anode and the signal is readout. The DWC has 4 direction TDC(time to digital converter) readouts (top, bottom, left and right). By calculating the delay between x or y, the point where the particle penetrates is retrieved.

By using multiple DWCs in the beam line, the track of the particle can be recon- structed. The extrapolation of this track to the HGCAL prototype can serve as a refer- ence for studying position resolution as well as an event selection to reject the particle deviated from the normal trajectory. The resolution of the DWC in the H2 beam line is 0.2 mm, which is enough for the cell size of the sensor (∼1cm).

More information of the DWC can be found in [11].

5.4 Setups

5.4.1 2016 CERN beam test - configuration 2

In 2016, different beam tests have taken place at FNAL and at CERN. Here only the configuration 2 setup at CERN will be mentioned. The 2016 beam tests are based on CE-E prototype, the Figure 5.5shows the setup for configuration 2. The detector consists

Figure 5.5: 2016 setup (CERN config 2), plot from [6]

The X₀ of all the layers are calculated and labeled in the schematic plot.

Different materials of the absorbers are expressed with the in different color.

of 8 single module layers and the absorber layers. In addition, a lead block of 4.5X₀ is placed before the first module. From then on, a sandwich structure is repeated 8 times by the order of module-absorber. The absorber layers are all consist of 3 or 4 tungsten plates and/or a lead plate. In most of the detector, the modules are placed right side of the Cu cooling plate, while in layers 5 and 7 the module is put on the left side of the Cu cooling plate.

5.4.2 2017 July beam test

In 2017, 3 beam tests were taken at CERN: May, July and September/October. The July test beam is a milestone for the first system test to achieve the original design of HGCal. It combines CE-E and CE-H part of the HGCal detector and include the CALICE analogue hadron calorimeter (AHCAL) to collect the hadronic shower after the CE-H.

CALICE analogue hadron calorimeter (AHCAL) is a scintillator-based sampling calorime- ter with silicon PMT (SiPM) readout. The ASIC used is called the SPIROC, and the detector also has its own independent DAQ system. [13] The absorber between different layers are steel. The AHCAL is selected to act as the scintillator part of the official HG- CAL design. Although in the latter CE-H section the official design is a mixture of silicon and scintillator, it is still a nice trial to validate the performance of the material.

The full setup of this beam test can be found in Figure 5.6, while Figure 5.7 and Figure 5.8 gives a more detailed look to the silicon based prototype. In the CE-E, lead, copper and iron are placed as absorber (and CuW base plate). Iron absorber is placed in both CE-E and CE-H part of the prototype instead of tungsten. 3 silicon modules are placed in the CE-E as 3 single module layers, and 8 of the module are placed in CE-H with 2 single-module layer and 2 triple-module layer. The 3 module structure is made in order to contain the expected hadronic shower width.

Figure 5.6: 2017 July setup The full set up for 2017 July setup.

Figure 5.7: 2017 July CE-E, plot from Shilpi Jain.

The label EE is the old term used in the CMS phase 2 technical proposal.

Figure 5.8: 2017 July CE-H, plot from Shilpi Jain.

The label FH is the old term used in the CMS phase 2 technical proposal.

5.4.3 2018 June beam test

In early 2018, 3 modules are brought to DESY for beam test in order to obtain a better understanding for both the sensor and the electronics for Skiroc2cms 6” module. With the knowledge learned from DESY, 2 beam tests are taken at CERN in June and October.

June test beam is the first test to realize the full design of the CE-E in the HGCal technical design report (TDR) [7], with 28 layers single module prototype.

The absorbers are chosen to be 1X₀ in each layer, and ∼ 28 X₀ in total. The cassette structure is developed to contain 2 module back to back with good attach to the cooling plate. The cassette are constructed to keep roughly the same radiation length between all module layers, and the structure also make the transform safer. With such an effort, the prototype is finally ”compressed” into a 75 cm detector, with 28 layers of single module.

The aim for this setup is to test the behavior of various energy for electron and compare the agreement between data and Monte Carlo. The plan for October beam test will be having an order of 100 module prototype, with both EE, FH and the new AHCAL prototype ^†.

†Other than the AHCAL used in 2017 test beam.

6 Readout chips

The analysis work is highly related to the original data format sent from the readout chips. Based on the memory mapping of the chip, different decoding method are applied.

Two kinds of readout chip are chosen to carry out the test beams, namely, Skiroc2 (Silicon Kalorimeter Integrated ReadOut Chip) in 2016, and Skiroc2cms in both 2017 and 2018 periods. This chapter will states on the decoding methods of both chip.

Both Skiroc2 and Skiroc2cms ASIC are designed by Omega group and both the ASICs are based on Skiroc, an ASIC applied in CALICE experiment.

6.1 Skiroc2

Skiroc2 has 64 channels, while in each channel exists preamplifier and two separate slow shapers, a fast shaper, the self-trigger and a fifteen-cell pipeline, as well as a 12-bit ADC.

Only some of the functionalities of Skiroc2 are used, two slow shapers with different gains running with external trigger mode (provide by the test beam trigger system). The two shapers have a fixed ratio of 10:1, namely high gain(HG, 10) and low gain(LG, 1) circuits.

When the trigger signal was received, both HG and LG in all 64 channel are readout and stored.[2]Note that one channel remains unconnected to the sensor since only 127 wire bonds are attached to the sensor. The readout chain for data acquisition (DAQ) in 2016 used a costom DAQ boards based on off-the-shelf components. More information and photograph of the full DAQ chain in 2016 can be found in [6].

6.2 Skiroc2cms

Skiroc2cms has 64 channels. Half of these channels are connected to the silicon sensor, while their twin channels which lie in the same position are left unconnected. Both fast and slower shaper circuits are readout in 2017/2018 period. However, only the slow shaper circuits are used in the charge measurement. The functionality of the slow shaper remains the same(10:1 HG and LG circuits) while there is a new circuit called Time Over Threshod(TOT). Another circuit, the Time Of Arrival (TOA), is also included for providing a precise (∼ tens of pico-second resolution) timing measurement. 10-bit DAC threshold can be set in both TOT and TOA. TOT is designed to cover the signal range even after the LG circuit is saturated so that the dynamic range set in HGCal TDR (1 fC

∼ 10 pC) can be fully covered. With HG, LG and TOT circuits for signal measurement and TOA circuits for timing measurement, the HGCal prototype has a chance to realize the original purpose for a 5D detector (energy, time, x, y, z).

Both connected and non-connected channels(HG,LG,TOT and TOA) are readout and stored for analysis purpose, mainly for noise study.