DØ-note 4120

July 2003

Silicon Sensor Quality Assurance for the D0 Run IIb Silicon Detector: Procedures and Equipment

A. Bean1, T. Bolton2, M. Demarteau3, R. Demina4, D.Karmanov5, S. Korjenevski4, F. Lehner6, R. Lipton3, M. Mao3, M. Merkin5, R.L.McCarthy7, R. Sidwell2 and R.P Smith3

1 Kansas University, Lawrence, USA

2 Kansas State University, Manhattan, USA

3 Fermi National Accelerator Laboratory, USA

4 University of Rochester, USA

5 Moscow State University, Russia

6 University of Zurich, Switzerland

7 State University of New York at Stony Brook, USA

Abstract

This documents describes in detail the quality assurance (QA) program for the D0 Run IIb silicon sensors. The scope of the QA program, the responsibilities of the participating institutions, and the testing and measurement procedures are defined.

1. Introduction

Preparation for the Run IIb silicon tracking system for the D0 detector is underway. The design of the new system relies profoundly on the experience gained with the Run IIa silicon tracker and the quality assurance program established for the Run IIb silicon sensors themselves likewise builds on the Run IIa experience. The new tracking system incorporates ~ 2400 single-sided sensors in six barrel layers located at radii from 1.8 cm to 16 cm from the interaction point. The Run IIb detector is scheduled to be ready for installation by midyear 2006 and must maintain efficient tracking performance after experiencing the radiation damage expected (up to 1014 1 MeV equivalent neutrons/cm2) over its Run IIb lifetime (~10-15 fb-1 integrated Tevatron collider luminosity).

A thorough, reliable, and efficient silicon sensor probing and characterization effort, defined in a well-organized silicon quality assurance (QA) program is critical for the confirmation that only sensors of desired quality are procured and incorporated into the detector. This document describes the procedures, responsibilities and organization of the QA program established by the D0 collaboration for the silicon sensors for the RunIIb silicon tracking system. In particular the program includes:

  • Organization and responsibility for testing
  • Testing procedures and acceptance criteria
  • Testing equipment
  • Inventory control, documentation and database

In addition to incorporating prior experience gained during the fabrication of the Run IIa detector, the Run IIb QA program also incorporates testing procedures developed for the ATLAS SCT sensors and the CMS silicon sensors. Finally, it incorporates the consensus of many discussions within the D0 Run IIb silicon group while the design and performance parameters of the new system were developed.

In Section 2 of this document the general silicon sensor QA organization is described and its responsibilities are defined. Section 3 summarizes the silicon sensor specifications and terms of acceptance. Section 4 describes the silicon sensor QA program in detail. The inventory and testing database design and its requirements are described in section 6, and in Section 7 there is a list of measurement instrumentation and probing equipment which have been found useful by D0 and have been installed at the several silicon testing centers. The specifications for the sensors are given in Section 8

This document is also available on the D0 Run IIb sensor web page This web page contains also links to updated probing and testing information as well as results.

2. General D0 Run IIb Silicon Sensor QA Organization Issues

Quality assurance for the D0 Run IIb silicon sensors will be carried out at three primary D0 institutions and two secondary D0 institutions. The four primary testing centers and the corresponding coordinators are:

  • Fermilab, Batavia, USA. Local coordinator: Marcel Demarteau, Ron Lipton and Madame Mao
  • Kansas State University, Manhattan, USA. Local coordinator: Tim Bolton and Robin Sidwell
  • State University of New York, Stony Brook, USA. Local coordinator: Bob McCarthy
  • University of Rochester, USA. Local Coordinator: Regina Demina and Sergey Korjenevski

In the event of testing schedule backlogs additional QA activities are foreseen to occur at two secondary institutions:

  • Moscow State University, Moscow, Russia. Local coordinator: Michael Merkin
  • University of Zurich, Zurich, Switzerland. Local coordinator: Frank Lehner

The QA program defines a collaborative activity that provides for shared responsibility among the different institutions with defined testing procedures followed rigorously at all testing centers. Local coordinators at each institution are charged with the responsibility to determine that the local QA program is implemented consistently and that all tests are performed in accordance with this document. Prior to the performance of QA testing on production sensors at any institution, the capabilities and performance of that institution will be certified by the Run IIb silicon sensor coordinators.

Fermilab will serve as a central distribution and control center with dedicated testing and coordinating tasks. Its main tasks will be:

  • The initial registration of each sensor upon receipt from the vendor
  • The visual inspection of each sensor
  • The two key characterization measurements (I-V and C-V) on each sensor
  • The distribution of the sensors to the two other primary testing centers
  • The performance of additional sensor characterization measurements on sub-samples of the sensors as described in this document
  • The handling and shipping of sensors rejected by the QA program that are returned to the vendor
  • The overall monitoring of the QA program
  • The final acceptance and grading of the sensors

The last two items are coordinating and managing tasks and their completion will be the responsibility of the two RunIIb silicon sensor coordinators together with the local coordinator(s) at Fermilab.

Figure 1 shows a graphical representation of the QA program with the silicon sensor flow.

Figure 1: A graphical representation of the QA organization and silicon sensor flow.

3. Overview of Sensor Design, Specifications and Terms of Acceptance

3.1. General

The Run IIb silicon detectors are p+n type single-sided sensors, with AC coupling and biased through polysilicon resistors. The silicon sensors will have a single-guard ring with peripheral n-well feature as designed by Hamamatsu Photonics. There are three sensor types required for the RunIIb detector:

Layers / Active Length (mm) / Active Width (mm) / Strip pitch /readout pitch (m) / # readout channels / # of sensors + spares
0 / 77.36 / 12.8 / 25/50 / 256 / 144+50%=216
1 / 77.36 / 22.272 / 29/58 / 384 / 144+50%=216
2-5 / 98.33 / 38.34 / 30/60 / 639 / 1896+20%=2280

Layer 2-5 sensors will be manufactured on 6-inch wafers. Layer 0 and 1 sensors are compatible with a production on 4, 5 or 6-inch wafers.

The web page provides links to the detailed mechanical drawings of the sensors. These drawings are given wide circulation throughout D0 so that all testing personnel become familiar with the exact sensor or test structure geometry in precise detail. This detail comprises the location and character of fiducial marks, the location and character of bonding and testing pads, the on-sensor strip numbering keys, and the multiple-field scratchpad for unique sensor identification, vendor information and for QA pass/fail marks done by the testing centers.

3.2. Specifications

The detailed sensor specifications are reproduced in Section 8 of this note. The following table summarizes the mechanical and electrical specifications critical to the QA program:

Specifications: / Layer 0/1 / Layer 2-5
Wafer thickness / 32020m, wafer warp less than m / 32020m, wafer warp less than m
Depletion voltage / 40<V<300V / 40<V<300V
Leakage current / <100nA/cm2 at RT and FDV+20V, total current < 4A at 700V / <100nA/cm2 at RT and FDV+20V, total current < 16A at 350V
Junction breakdown / >700V / >350V
Implant width / 7m / 8m
Al width / 2-3 m overhanging metal / 2-3 m overhanging metal
Coupling capacitance / >10pF/cm / >12pF/cm
Coupling capacitor breakdown / >100V / >100V
Interstrip capacitance / <1.2pF/cm / <1.2pF/cm
Polysilicon bias resistor / 0.80.3 M / 0.80.3 M
Defective strips / <1% per sensor / <1% per sensor

3.3. Terms of acceptance

The silicon Distribution and Control Center (i.e. Fermilab) will receive all sensors from the vendor and record them as in preliminary compliance if the specified QA documentation received from the vendor indicates that all vendor-required tests meet the specification.

The Distribution and Control Center will expeditiously perform the sensor key tests described in detail below before distributing any sensors to the other testing sites. In addition, Fermilab will perform additional subset tests described below on a fraction of the sensors that pass the key tests. The subset tests focus on confirming the compliance of all sensors within a production batch.

Non-compliance of any sensor will be determined by Fermilab within 90 days of receipt of the sensor. The vendor will be notified of the non-compliance and if the vendor requests the opportunity for remeasurement, the noncompliant sensor(s) will be returned to the vendor. The Run IIb silicon sensor coordinators and the vendor can agree upon the acceptance of sensors that fail only marginally to meet specifications.

4. The QA program

The QA program for the Run IIb sensors consists of four main parts:

  1. Key Tests: The silicon sensor key tests are performed on every received sensor. The measurements/procedures belonging to the key tests are the most important ones because they confirm acceptable minimum performance of the sensor as an entity. The key testsare carried out at the Distribution and Control Center, i.e. at Fermilab.
  1. Subset Tests: The silicon sensor subset tests are conducted on a defined fraction of sensors which pass the key tests. The main goal of the subset tests is to verify compliance with the specifications in depth. The fraction of sensors which were subject to these tests was large during the prototyping stage. For production sensors, the fraction will depend on the overall quality of the subset test data and its corroboration of the vendor QA data. The subset tests are nearly all done in automatic probe stations under computer program control.
  1. Diagnostic Tests: The silicon sensor diagnostic tests are routinely performed on a fraction of sensors selected at random from those passing the key tests, continuously throughout the sensor delivery period, as well as on sensors with irregularities revealed in the key or subset sensor tests. The diagnostic tests measure in much more detail complex electrical parameters to provide a deeper insight into sensor quality and to monitor the vendor production process from batch to batch.
  1. Mechanical Measurements: The mechanical measurements are routinely performed throughout the production period on a fraction of sensors selected at random from those which satisfy the key tests. The measurements have been crucial in the prototyping phase to validate the mask design, and they can reveal wafer and production problems that might arise during production.

The current definition of the fractions of sensors which will be subject to the subset, diagnostic and mechanical tests is maintained in the D0 Run IIb Silicon Sensor QA Part Flow document. The fractions listed in this document will be updated on a continuing basis as the overall quality of the received sensors warrant. This document can be found on the web page

4.1. General Conditions/requirements of Testing/probing

4.1.1. Clean Room conditions and handling of sensors

All personnel who handle sensors shall undergo training and practice in the handling of sensors that exposes them to all phases of the manipulation of the sensors required by the QA program. Sensors not being tested shall be stored in clean containers in a dry storage area with restricted access.

All production sensors shall be handled at the testing centers in a clean room or a clean room housing with temperature (21 +/- 2oC) and humidity (40+/-10%) control. The clean room class is not specified, but personnel access to the area must be strictly limit and non-silicon activities (e.g. soldering, use of chemical reagents, use of power tools, etc.) prohibited. Note the actual humidity in the work environment should not fall below 30% so that electrostatic discharges don’t become a problem.

.

Protective clothes, masks, and gloves shall be worn by all personnel who handle sensors. While vacuum tweezers have customarily been the preferred method of lifting sensors, recent testing experience indicates that the tweezers may leave a temporary “footprint” on the detector that enhances the strip leakage currents where the vacuum cup of the tweezers contacts the sensor. Until this phenomenon is better understood, the use of vacuum tweezers should be limited. Because the sensors are packaged between two thin layers of cardboard, it is generally possible to gently and safely slide the sensors from the packaging to probe chucks, and back again, completely avoiding the use of tweezers.

In chapter 7 of this document are described the recommended test and measurement equipment that has been found to facilitate this QA program. A detailed guideline to cleanroom conditions and sensor handling can be found on the web (

4.1.2. Testing conditions

All electrical tests should be carried out in the specified clean room or clean room housing described in section 4.1.1. The electrical tests must be made in a light-tight dark box. To contact the sensors for probing, the probe tips must be placed only on special contact pads which are not used for production bonding during the assembly process (except bias line contact pads). These “test pads” or “probe pads” (either AC- or DC-pads) are indicated on the mechanical sensor drawings, which are available at

We suggest covering the chuck of the probe station with a special conductive rubber film prior to usage to prevent damage on the sensor backside during the positioning of the sensor on the chuck. If a conductive rubber sheet is not used, then great care has to be used in keeping the chuck clean, not to damage the sensors.

4.1.3. Sensor Database

A general database system for parts and components as well as for production and assembly of the entire Run IIb detector is being prepared at Fermilab. The database design and architecture is adopted from the ATLAS SCT group[1]. This database is a relational database in ORACLE and follows a client/server model. The database server is located at Fermilab. The web-interfaces for easy storage and retrieval of data are being developed. The database usage is part of the sensor QA program, and the database will be utilized to identify and track sensors and to store and retrieve sensor quality and grading information data for each silicon sensor.

Whenever testing data is logged to the database, the record must contain the following items:

  • Measurement data
  • Temperature/humidity
  • Date
  • Testing center
  • Comments

As noted in section 2, one of the tasks of the Run IIb silicon sensor coordinators will be to monitor the conformance of database entries to these requirements. The utility of the database is strongly a function of the clarity and comprehensivity of the data logged in it.

4.2. Key Sensor Tests: Performed on every Sensor

The key sensor test program represents the most important quality control test of the QA program. The key tests shall be performed on all sensors promptly at Fermilab upon their receipt from the vendor. The tests are not complex and they serve as a valuable first-inspection for sensor acceptance and performance grading. The key tests include an initial registration of the sensor in the database, a visual inspection, and I-V and C-V electrical measurements.

4.2.1 Initial Registration into the Database

This task is done at the central distribution center (Fermilab). Each shipment of sensors is checked for content and QA documentation from the vendor. For each sensor the test structure piece (“half moon”) for each wafer must be present. The vendor QA measurements are provided in both an electronic format (i.e. EXCEL spreadsheet) and printed copy. This material is verified for completeness and the data is inspected to ensure that each sensor it documents meets specification.

Specifically, the material is checked for:

  • serial ID-number of sensor
  • batch/lot number
  • wafer thickness
  • leakage current values at the range of specified voltages
  • depletion voltage
  • number of bad channels/strips
  • polysilicon resistor mean value and standard deviation or min/max
  • Aluminium resistance value on monitor structure
  • Implant resistance value on monitor structure
  • typical coupling capacitor mean value and breakdown value
  • eventually further measurement information based on teststructures

If any of the required information or the teststructure “half moon” for a sensor is missing, such omission shall be specifically flagged in the database and the sensor marked as noncompliant in the database. If any of the data indicates that the sensor does not meet specifications, the offending information is unambiguously flagged in the database and the sensor marked as noncompliant. The arrival date and present location (i.e. testing institute) of each sensor is entered.

4.2.2 Visual inspection

When the registration process is completed all sensors not marked as noncompliant are subject to visual inspection.

This key sensor test is also performed at the central distribution center, and its purpose is to identify sensors for which mechanical imperfection might impair electrical performance. The visual inspection identifies physical defects, damage, and edge chipping, and records the sensor region where the damage is located so that further electrical measurements can focus on that region. If edge damage is identified, the sensor must be sent for mechanical measurements when the visual inspection is complete.

The visual inspection is carried out on an x-y moving table equipped with a microscope having different objectives with magnification from 3x to 50x. A video camera connected to the microscope with a video monitor and video printing capability is required. A detailed guide to sensor visual inspections including some illustrative sensor defect pictures is on the web (

Procedure for visual inspection:

  1. Ensure that the x-y table is completely clean and clear of any debris.
  2. Remove sensor from its envelope/shipping container.
  3. Search for any signs of silicon debris in the sensor envelope or within the shipping box. If debris is present, be sure to remove it before eventually returning the sensor to the envelope, and identify the source of the debris during the visual scans of the sensor.
  4. Examine the back surface by eye. Take note of any blemishes or scratches.
  5. If there are indications of edge chipping, place the sensor on the probe station chuck (with the sensor still strip-side down on the shipping cardboard unless the chuck is covered with the recommended pad) and measure the width of the chipping. Take a picture if appropriate.
  6. Remove the sensor from the chuck and replace it on the chuck with the strip side facing upwards.
  7. Check that the serial number scratched on the identification pads matches the serial number on the sensor envelope.
  8. At high magnification, scan along all four edges, searching for edge chipping, scratching or other damage.
  9. Check the visibility and quality of the fiducial marks.
  10. With the same high magnification, scan along the bias resistors, searching for breaks, signs of processing defects or non-uniformity. Check for alignment of metalization with implant.
  11. Scan along the AC-bonding pads (the ones which will be used for the hybrid bonds) and verify that they are clean and not probed.
  12. At lower magnification, scan the full area of the sensor, making notes (and taking pictures where appropriate) of blemishes, scratches or other non-standard features.

13. Update the database, recording the completion of visual inspection and enter all comments and findings.