The PPS tracking system - Performance in LHC-Run2 Prospects for LHC-Run3 M.M. Obertino

The PPS tracking system
Performance in LHC-Run2
 Prospects for LHC-Run3

 M.M. Obertino
 Università di Torino, INFN Torino
 on behalf of the CMS and TOTEM collaborations

The CT-PPS - now PPS - project
 Approved in Dec. 2014 by LHCC and CERN Research Board as common CMS-TOTEM project
 (CT-PPS à CMS-TOTEM Precision Proton Spectrometer )

 Since April 2018, CT-PPS is a standard component of CMS, with name PPS

 CERN-LHCC-2014-021
M.M.Obertino – Università e INFN Torino – IPRD19

 The main target of the PPS physics program is the study of Central Exclusive Production (CEP) processes

 pp à p X p X = high-ET jets, WW, ZZ, gg …
 where both protons remain intact

 PROCESSES STUDIED IN DETAIL FOR THE CT-PPS TDR FIRST PHYSICS RESULT, JHEP07 (2018) 153
 [CERN-LHCC-2014-021
 2

The Precision Proton Spectrometer
 Introduction to PPS
 Events of interest are characterized by:
 § two leading protons on opposite sides of the CMS interaction point (IP)
 • Near beam magnetic spectrometer at IP5 of the LHC
 § possibility to constrain the event kinematics by matching the momenta of the central system (X) and
 • Joint CMS+TOTEM project to include horizontal Roman Pots in the standard, high luminosity,
 the leading protons CMS data taking, TDR in 2014 [CERN-LHC-2014-021]
 • Started in 2016 (one year early) thanks to the availability of the legacy TOTEM silicon strips
 PPS has been designed for measuring the scattered protons on both sides of CMS
 detectors
 TRACKING
 • Collected
 in standard LHC running conditions
 data throughout the whole LHC Run 2
 SYSTEM
M.M.Obertino – Università e INFN Torino – IPRD19

 ~200 m

 IP5
 LHC Sector 45
 LHC Sector 56 DETECTOR STATIONS

 • Two complementary measurements
 ü LHC magnets used •toTracking
 bend protons
 detectors measure the proton displacement with respect to the beam, which is
 translated into energy-momentum loss thanks to the knowledge of the beam optics
 ü Tracking detectors used to measure the proton displacement with respect to the beam, which is
 • Timing detectors measure the proton arrival time in both arms w.r.t. the reference clock
 translated into momentum lossthe
 that keeps (x)two
 thanks to the
 stations knowledge
 synchronized, of the
 allowing thebeam optics
 calculation of the longitudinal position
 of the interaction
 ü Timing detectors to disentangle pile-up
 E. Bossini’s talk on Thursday
 A. Bellora Low-x 2019 – Recent results from PPS and PPS status and prospects – Nicosia, Cyprus
 3

Tracking detectors:
Experimental challenges
 experimental challenges
 § To maximize the acceptance for low x protons detectors need :
 Roman Pots need to operate at few mm from the beam to maximize acceptance
 ü to operate as close as possible to the beam without affecting LHC stable operation
 • RF shielding installed to limit the impedance caused by the RP insertion
 Roman Pots (RP) PPS physics
 case
 Tracking Proton beam
 Experimental
 Detectors In 2017-18 CT-PPS RPs inserted
 apparatus at
 RF shield 12 sbeam+0.3 Operation
 mm (~1.5 in mm)
M.M.Obertino – Università e INFN Torino – IPRD19

 distance from the LHC beam
 LHC Run 2
 Tracking
 efficiency
 Detector
 ü to be fully efficient as close as possible to their alignment and
 proton
 mechanical
 Detectors must tolerate edge
 high levels of non-uniform irradiation reconstruction
 RADIATION LEVELS
 • For ~100 fb-1 (Run 2): ~5 ∙ 10 15
 protons/cm2 (Semi)Exclusive
 IN THE DETECTOR
 → ℓ+ ℓESTIMATED
 −
 USING
 in tracking detectors 2016 results
 TOTEM DATA AND
 • §Spatial resolution
 Detectors must required:
 tolerate~10-30
 highμm
 levels of non-uniform irradiation SIMULATION (LINT ~
 Prospects for -1
 • Timing resolution required: ∼ 20 ps, to reject high pileup LHC Run 100
 3 fb )
 à Proton flux up to ~5 · 1015 protons/cm2 for ~100 fb-1
 Dose: 1.61 Mrad/ fb-1

Bellora Low-x 2019 – Recent results from PPS and PPS status and prospects – Nicosia, Cyprus 5
 4

PPS detector configurations in LHC-Run2
 Data taking in 2016, 2017 and 2018 with different detector configurations
 Data taking with CMS
M.M.Obertino – Università e INFN Torino – IPRD19

 Exploratory Strip + Pixel + Pixel + Single and
 phase Strip (+Diamond) Diamonds+UFSD double diamonds
 2015 2016 2017 2018

 LINT~15 fb-1 LINT~40 fb-1 LINT~60 fb-1
 Data recorded with 88% OF DATA
 39% OF DATA 93% OF DATA
 tracking RPs inserted
 RECORDED BY CMS RECORDED BY CMS RECORDED BY CMS

 Exploratory phase in 2015 and 2016; very high stability in both 2017 and 2018

 PPS integrated luminosity in LHC-Run 2: ~115 fb-1
 5

Tracking detector – 3D pixels
 Six detector planes per station tilted by 18.4° to improve resolution
 § 3D silicon pixel sensors read out with 4 or 6 PSI46dig ROCs (same ROC
 usedin layers 2-3-4 of the CMS pixel detector) depending on sensor size

 § Pixel area: 100x150 µm2
 Rpix § Sensor thickness: 230 µm
 portcard § Column depth: 200 µm
M.M.Obertino – Università e INFN Torino – IPRD19

 § Column diameter: 10 µm
 § 200 µm slim edge

 § Modules wire-bonded to a flex circuit connected
 to the RPix portcard (interface between modules
 16 mm
 and front-end electronics)
 § Same front-end boards for data (FED) and
 24 mm

 control (FEC) as for the CMS Phase I pixel tracker
 upgrade
 2X3 (2x2) SENSOR – 6 (4) ROCs

 § Mechanics and cooling adapted from TOTEM tracking system.
 § Operation at -20 °C and in vacuum ( p < 20 mbar) 6

Tracking detector - Silicon strips
for a heater and a PT100 resistance thermometer for temperature control.
the motherboard, together with clock and trigger input signals, HV and LV power, and c
connector each VFAT will send the trigger outputs and the serialised tracking data from m
 m
 Same detectors used in the tracking RPs of the TOTEM experiment [2]
used to monitor detector leakage currents, power supply voltages and temperature. Via 35
Each VFAT provides tracking and trigger generation signals from 128 strips. The D
 10 planes of micro-strip silicon detectors per RP
input channels of 4 readout chips “VFAT” (section 7.1), and a Detector Control Unit (
The silicon detector hybrid (figure 4.13) carries the detector with 512 strips wire-bon
that are relevant to the mechanical construction and the detector monitoring are discu § 512 strips per plane, ±45o orientation
The generalised electronics are described in more detail in chapter 7. Only the electro
 § Thickness: 300 µm - Pitch: 66 µm
 § Edgeless technology: ~50 µm inactive edge
 4.4 On-detector electronics

four VFAT readout chips.
 (CTS on the edge facing the beam)
M.M.Obertino – Università e INFN Torino – IPRD19

Figure 4.13: Two RP hybrids mounted back-to-back. Each hybrid carries the silicon d § Resolution ~20 µm
 § Designed for low-luminosity running:
 ⚠ Lifetime: up to 5x1014 p/cm2 integrated flux
 ⚠ No multi-track capability
 Figure 4.10: CURRENT TERMINATING
 Digital readout provided by Picture
 4 VFAT2of a Planar Edgeless Detector with CTS (top).
 STRUCTURE (CTS)
 of the chip cut region (bottom)!'**#%/)/#*1+%$/+%,)*+%,
 shows the details of the CTS.
 chips located on a flexible circuit !"#$%&'()*+%, 6+$?+%,)#"#!/*@0#

 connected to a motherboard 8 8 -" 2+3
 9:; 9; 8 AB

 4.3.2 The silicon detector for the TOTEM Roman Pots
 4

 Front-end boards forThe
 data and control, ( . ?/*+(
 advantages offered by the
 & CTS have led TOTEM to choose th
 ( ?/*+( .

 VFAT2 mechanics and cooling from TOTEM
 tors. Geometry
 ! 1
 and granularity have been adapted to the specific re
 %&/5(#)6'"7

 tracking system.
 and spatial resolution.
 !'/)#0,#)))))
 -" % .

 The detectors have been developed and produced in a joint ?#%?+/+C#)C@"'1#

 [2] The Totem Collaboraytion et al., The TOTEM Experiment at theCERN
 CERNand Megaimpulse, a spin-off company
 Figure 4.8: Cross-section of a silicon detector with from
 CTS in the theplane
 Ioffe
 parallelPh
 to
 Large Hadronscheme. Collider, 2008 JINST 3 S08007
 biasing
 + 7

DETECTOR PERFORMANCE

2D track impact point distributions
 LHC SECTOR 45 IP5 LHC SECTOR 56
 BEAM 2017 DETECTOR CONFIGURATION
M.M.Obertino – Università e INFN Torino – IPRD19

 2018 DETECTOR CONFIGURATION

 The RP located at 210 m tilted by 8° around the z axis.
 9

Si-strip detector efficiency
 Two major sources of inefficiency (radiation damange and no multi-track capability) studied separately
 0.9 7
 Only single-track events can be

 Efficiency
 CMS Preliminary 2017 postTS2
 reconstructed with the Si-strip detector 0.8 6
 à efficiency decreases with number of 5
 interactions per bunch crossing (pile-up) 0.7
 4
 ~50% efficiency at pile-up 40 0.6
M.M.Obertino – Università e INFN Torino – IPRD19

 3
 0.5
 NUMBER OF TRACKS RECONSTRUCTED BY THE PIXEL DETECTOR AS A 2
 FUNCTION OF PILE-UP 0.4
 1
 0.3 0
 10 20 30 40 50 60
 Pile-up

 More than 40% of events with multiple tracks in standard
 data taking conditions (PU~40)
 The increasing number of multi-track events with pile-up
 shows the advantage of a pixel detector with respect to a
 strip one
 Pile-up 10

Effect of radiation on strip detector efficiency
M.M.Obertino – Università e INFN Torino – IPRD19

 Efficiency maps produced:
 § for the region relevant for physics (covered by the pixel detector acceptance, and below the
 collimator aperture limits)
 § with the first data collected (LLINT ~ 2.4 fb-1)

 Radiation damaged area clearly visible, especially in sector 45

 à Detector packages replaced during 2017 LHC technical stop
 11

Pixel performance: efficiency
 Efficiency estimation based on tracks reconstructed within the same tracking station

 RP EFFICIENCY
 AT BEGINNING
 OF DATA TAKING

 SIMILAR RESULTS
M.M.Obertino – Università e INFN Torino – IPRD19

 IN 2018

 Overall very good performance - average efficiency above 99%
 LINT = 8 fb-1 LINT = 11.7 fb-1
 The lower efficiency region visible in the right part of the two maps is due to the fact that both RPs
 have one plane 2x2 not covering that region, and to the presence of malfunctioning ROCs on
 other planes.

 Less than 0.05% bad/noisy pixels
 12

Radiation effect on the ROC
 Pixel ROC (PSI46dig) not optimised for non-uniform irradiation.

 Non-uniform irradiation causes a difference between the analog current supplied to the most and the
 least irradiated pixels.
 Voltage

 NON-IRRADIATED PIXEL
M.M.Obertino – Università e INFN Torino – IPRD19

 HIGHLY IRRADIATED PIXEL

 PIXELS NOT RESPONDING IN THE SAME 25 ns
 CLOCK WINDOW (BX)

 Time

 ü Irradiation studies performed before installation at LHC showed that after a dose corresponding to
 LINT (LHC) ~8 fb-1, the drift of the useful time window for signals in most irradiated pixels exceeds 25 ns
 ü To mitigate the impact on the data quality, the tracking stations were lifted up during LHC technical
 stops (TS) to shift the occupancy maximum away from the damaged region.
 13

Effect of radiation on RP efficiency (2018)
 EVOLUTION OF THE RP EFFICIENCY MAP IN THE DETECTOR REGION CLOSEST TO THE BEAM FOR RP 220 FAR (worst case)

 BEGINNING OF DATA TAKING TS 21.0 fb-1 TS 50.3 fb-1 57.8 fb-1

 SECTOR
 45
M.M.Obertino – Università e INFN Torino – IPRD19

 SECTOR
 56

 Deterioration of the RP
 BEAM RP SHIFTED BY RP SHIFTED BY efficiency caused by
 [ always in x=0, y=0 ] 0.5 mm DOWNWARDS 0.5 mm DOWNWARDS the radiation damage 14

Effect of radiation on RP efficiency (2018)
 IRRADIATION PEAK AREA

 IRRADIATION PEAK AREA
M.M.Obertino – Università e INFN Torino – IPRD19

 Average efficiency calculated every ∼ 1 fb-1 in the critical region around the irradiation peak
 Drop in the efficiency due to irradiation clearly visible in the critical region; recovery after each
 LHC technical stop (TS) due to vertical movement of the RPs.
 à This plot will be used as monitoring tool in LHC-Run3
 15

in LHC Run 3 (2021-2023)
 PPS prospects for LHC-Run3 DESIGN AND FIRST PROTITYPE OF
 THE NEW DETECTOR PACKAGE

 PPS will operate as a full CMS subsystem in LHC-Run3
 • Experimental (2021
 Sensor
 apparatus: - 2023) for Run3
 production
 TRACKER SYSTEM in LHC-Run3 • Production
 Tracking: on 6′′ wafers; 3D sensors with one-side columns, ’2E’ geometry NEW FLEX
 • 220 •
 2 horizontal “2×2” and “1×1” sensor size
 stations with6silicon 3D pixels sensors Wafer layout CIRCUIT
 § 2 Roman Pots per side, at 210 m Contract
 and mCERN
 through from CMS-IP;
 tender: detector planes
 (similar to the 3D pixels designed for the
 per RP (same as 2018) • minimal requirement
 CMS Phase2of 100 good 2×2
 Tracker sensors;
 R&D)
 • minimal requirement of 4 good (≥50% yield) wafers;
 • PROC600 readout chip
 Contract adjudged to FBK Trento
 § New 3D silicon pixel sensors will be produced
 (same as CMS bypixelFBK detector layer 1)
 • delivery in February-March 2020
M.M.Obertino – Università e INFN Torino – IPRD19

 • New internallyCell detail motorized detector package,
 ü single side technology to distribute the radiation damage
 ü 2x2 sensor geometry and reduce its impact
 ü 150 µm thick • Timing:
 ü 2E electrode configuration • 2 horizontal stations New support structure
 equipped with double-layer diamond sensors
 23/09/2019 PPS General Meeting, CERN !2

 First prototype with new design produced in INFN Ge workshop
 § ROC: PROC600 (same as layer 1 of the CMS pixel
 • Optimized readout detector)
 electronics PIEZOELECTRIC
 Features: MOTOR
 • simplified, easier-to-assemble structure;
 § New flex circuit design (different ‘look’ but very• similar design)
 reduced friction, with rolling balls;
 • ad-hoc mechanical coupling with piezo-actuator
 for bi-directional active movements;
 A. Bellora Low-x 2019 – Recent results from PPS and PPS status and prospects – Nicosia, Cyprus
 § New detector package with internal movement• simple
 system, to position
 (⟹ rad-hard) better distribute
 sensor included
 Tested in Genova starting end of July
 the radiation damage • successful movement tests in both directions,
 à 12 positions spaced by 500 µm, possibility under
 of handling
 realistic work conditions;
 • assembly procedure critical: steps to be followed
 more than 50 fb with minimal efficiency with
 -1 losscare;
 • few improvements implemented; others planned 16

Conclusion
 PPS has proven the feasibility of continuously operating a near-beam proton spectrometer at a high-
 luminosity hadron collider and has collected ~!!" fb-1 of data during LHC-Run 2

 Tracking system has been successfully operated since the beginning
 of LHC-Run2 with overall very good performance:
 § smooth data taking
 § high efficiency and tracking capability
M.M.Obertino – Università e INFN Torino – IPRD19

 despite the high and non uniform irradiation conditions

 The preparation for LHC-Run 3 is ongoing
 • New detectors are getting ready to be installed for the future
 data taking
 • A rich physics programme lies ahead, with many final states to be
 studied

 17

BACKUP SLIDES

Data taking
 CT-PPS collected:
 § ~ 88% of the full statistics recorded by CMS in 2017
 à ~ 40 fb-1 with RP data
 § ~ 93 % of the full statistics recorded by CMS in 2018
 à ~ 60 fb-1 with RP data
M.M.Obertino – Università e INFN Torino – IPRD19

 MD+
 TS2
 CMS Online Luminosity CMS Online Luminosity

 MD
 MD+TS1+
 CMS Online Luminosity high b run

 19

Tracking Detector - Silicon 3D Pixels
 3D sensor technology chosen for its intrinsic high radiation hardness and the possibility to implement
 slim edges.
 3D sensors produced by CNM: double side process with 2E and 1E electrode configuration
 no passing-through columns
 § Pixel area: 100x150 µm2
 § Sensor thickness: 230 µm
M.M.Obertino – Università e INFN Torino – IPRD19

 § Column depth: 200 µm
 § Column diameter: 10 µm

 VBIAS=40V

 3D sensors bump-bonded to the PSI46dig ROC extensively tested
 in laboratory and with beam at FNAL [1]
 ü 200 µm slim edge made of a triple p-type column fence;
 reduced at ~50 µm by increasing the bias voltage (for 2E
 electrode configuration)

 100 µm
 ü Spatial resolution for 2E (1E) sensors tilted by 20° : 22 µm (25 µm)
 200 µm

 [1] F. Ravera, The CT-PPS tracking system with 3D pixel detectors, Pixel2016 Workshop
 20

Tracker performance: hit residuals
 Hit residuals for single planes are evaluated with respect to the local track reconstructed in the
 Pixel Roma Pot

 Hit residuals
 CMS-TOTEM Preliminary for 1 pixel
 plane
 2017 DATA
M.M.Obertino – Università e INFN Torino – IPRD19

 § Residuals are consistent
 sx = 17 µm with those obtained at
 the beam tests

 § Similar results in 2018

 ü The pixel tracker works as expected
 ü Track resolution under final evaluation (~ 20 µm)
 21

Effect of radiation on RP efficiency (2017)
 EVOLUTION OF THE RP EFFICIENCY MAP IN THE DETECTOR REGION CLOSEST TO THE BEAM FOR LHC SECTOR 45

 8.8 fb-1 TS 18.9 fb-1 29.5 fb-1
M.M.Obertino – Università e INFN Torino – IPRD19

 RP SHIFTED BY
 BEAM The deterioration of the As the integrated luminosity
 1 mm UPWARDS
 [ always in RP efficiency caused by increases, a second damage
 x=0, y=0 ]
 the radiation damage region appears at the new
 occupancy maximum

 Detectors in sector 56 suffered smaller radiation damage (barely visible in the maps) because of
 the different irradiation profile.
 22

Effect of radiation on RP efficiency (2018)
 IRRADIATION PEAK AREA COMPLEMENTARY AREA
M.M.Obertino – Università e INFN Torino – IPRD19

 Average efficiency calculated every ∼ 1 fb-1 : IRRADIATION PEAK AREA
 § in the critical region around the irradiation peak (left plot)
 § in the complementary region (right plot)
 Drop in the efficiency due to irradiation clearly visible in the COMPLEMENTARY AREA

 critical region; recovery after each LHC technical stop (TS)
 due to vertical movement of the RPs.
 Average efficiency in the complementary area high and
 almost constant during the whole data taking.
 à These plots will be used as monitoring tool in LHC-Run3
 2323

RP efficiency versus proton fractional momentum loss (x)
M.M.Obertino – Università e INFN Torino – IPRD19

 Track efficiency versus the proton fractional momentum loss x at the beginning and at the end
 of 2018 data taking period for sector 45 (left) and sector 56 (right).
 § Efficency shown in the x range corresponding to the detector acceptance
 § Efficiency in each x bin evaluated using the values from the RP efficiency maps

 These plots show the impact of the sensor radiation damage on physics studies
 24

Si strip detectors – “multi-track” inefficiency
 ◦ RPs consist of 5+5 planes with mutually perpendicular strip orientations (denoted U and V)

 ◦ If two tracks pass through the detector (black dots), they activate 2 U strips (blue) and 2 V
 strips (red), yielding 4 crossings out of which only 2 are real
M.M.Obertino – Università e INFN Torino – IPRD19

 Only single-track events can be reconstructed

 Efficiency evaluated, using zero-bias data, as:

 Sufficient signal = one or more track patterns recognised OR hit multiplicity greater
 than threshold for pattern-recognition to give up

 25

Effect of radiation on strip detector efficiency
M.M.Obertino – Università e INFN Torino – IPRD19

 Efficiencies vs ξ measured in the most irradiated region of y (plotted for ξ values below aperture
 limitations)

 Radiation damaged area clearly visible, especially in sector 45

 26

Roman Pots
M.M.Obertino – Università e INFN Torino – IPRD19

 Each station includes 3 RPs Cylindrical RP specifically designed
 for PPS to reduce the impedance and
 host larger detectors.

 Equipped with a 300 µm thick window
 towards the beam (thickness required
 to compensate the pressure gradient
 on the larger window).

 Tracking RPs equipped with a thin No vertical stations, alignment done
 window 150 µm thick toward the by propagating tracks from the
 beam tracking stations.
 27