Silicon Wafer Rough

From OpenWetWare
Jump to navigationJump to search

NOTE TO ALL WHO EDIT Any help in writing this Manual is greatly appreciated. Grammar corrections, rearrangements, rewording and additions to the paper would be fantastic. However, I would ask that no one delete entire sentences without consulting me first. I believe that if we have "too much" in our rough draft it would be much easier to edit when we give it to Dr. Fields for the first time.


In order to ensure that expensive equipment and its components are working properly before they are attached together we must check for ourselves that the manufacturer produced the components correctly. Hence, when single sided AC coupled sensors are received the specs of the sensor must be checked within a certain tolerance. Also, due to the scale of the technology within the sensors improperly working strips are inevitable. Often times the manufacturer will have already tested for the improperly working strips. Using our own tools and methods we must confirm that these strips are indeed dysfunctional so that they can be accounted for when fitted into their final apparatus. Also we must ensure that the entire sensor and the individual strips of each sensor are functional within a certain tolerance.

Alexander T. J. Barron 14:30, 27 February 2009 (EST) Though not an especially new technology, silicon detectors remain unavailable as commercially mass-produced items. Researchers who want to use silicon detectors must place special orders with established silicon device fabricators. As no comprehensive system of quality control is in place, researchers must test all detectors individually to make sure of their quality and working tolerances. All silicon detectors make use of the P-N junction in various geometries according to taste in design. When a reverse bias voltage is applied over all the junctions, the entire detector becomes depleted past a certain voltage threshold, depending on the detector. One can vary layout of junctions to create multiple zones covering the entire detector where, once the detector has been fully depleted, a charged particle can trigger current spikes in individual strip channels. This localizes the position of the particle.

The prototype Forward Vertex Detector component silicon detectors are single-sided, meaning one side is sheathed in {METAL} while its opposite holds parallel strips of {TYPE} silicon, creating individual areas for high-resolution 2D path detection. Each strip is connected to a bias ring circumscribing and connecting to all strips via individual {polysilicon??} resistors for each strip. A second ground ring circumscribes the bias ring to provide additional measures for full detector depletion. A dielectric covering is applied over the surface with the exception of both directly- and capacitively-coupled testing pads. These pads allow researchers to probe each strip individually, or to test the detector as a whole for salient properties: leakage current, bulk capacitance, coupling capacitance, interstrip resistance, bulk breakdown voltage, and coupling capacitor breakdown voltage.

The prototypes in question, Hamamatsu S10938, arrive pre-tested for a variety of problems. In order to verify the quality and specifications of each detector, as well as the quality of testing by the manufacturer, we engage in a series of tests to determine all the salient qualities outlined above.

Single Sided AC Coupled Silicon Wafers

The Hammamatsu S10938 AC coupled prototypes were designed to detect elementary particles given off by heavy ion collisions as part of project PHENIX at the Relativistic Heavy Ion Collider (RHIC). The bottom side consists of conductive material where a high Voltage is applied. On top of the sensor 1664 conducting strips sit on each side of the sensor. Between the strips and the conducting base a semi-conducting material sits that has a conductivity proportional to the Voltage by _______________. Each Strip contains a DC pad, a Spy pad a bias resistor and two AC pads. These pads are the only points where the sensor can be contacted. The first set of AC pads are for testing the specs of the sensor and the second are for wire bonding during the actual experiment. It is very important to leave the AC pad closest to the Bias resistor unscathed. Any marks on AC pads make wire bonding nearly impossible.

First the sensor is brought to a proper high voltage so that the n type material acts as an inductor. With all of the charge carriers removed the bottom of the sensor and each strip act like a capacitor. The detection of particles occurs as it passes through the middle of the sensor causing a change in charge on the fully charged capacitor. This flux in the charge creates a pulse in the Voltage on the Capacitor. This pulse results from the charged particle being pulled toward the the side holding the opposite charge and destabilizing the balance of charge on the conductor. The conductor is connected to ground through the bias ring so that the circuit is complete.

Alexander T. J. Barron 14:29, 27 February 2009 (EST) Does the n-type Si act as an inductor?

Alexander T. J. Barron 14:23, 27 February 2009 (EST) I think we're going to need some detailed schematics and descriptions of the detector so all who read know what "spy pads" and "bias resistors" etc. are. The introduction would be a good place for all the physical descriptions - it would establish a baseline for those reading in order to understand the testing process delineated in the rest of the paper.


Probe Station

Probe Stations provide an easily maneuverable table in which the silicon wafer can be safely placed and probed. The Probe station must be compatible with a microscope and must be placed inside of a light tight box. Any light can cause voltage drops in the sensor and change test results. The probe station must be capable of holding a vacuum chuck where High Voltage can be applied and the sensor can be held. The probe station mainly allows us to easily move the sensor into the best place for the probing needles.

Vacuum Chuck and Air Pump

A Chuck of the sensor's proportions should be constructed in order to ensure that the sensor remains still during testing and setup. A vacuum pump should be used to ensure that the sensor does not slide or shift while the probing needles are down. Also the chuck should be made of a soft metal and smooth on top so that scratches are minimized during the placement of the sensors into the chuck. Severe scrathes and damages can occur during preparations if proper methods are not followed and a proper chuck is used.

Digital Multimeter & High Voltage Power Source

Connected to the vacuum chuck there must be a Power Source to supply the High voltage to the bottom side of the sensor. The high Voltage is required to deplete the sensor or test for the Depletion Voltage and Breakdown Voltage. The Multimeter must be capable of measuring the resistance up to the Mega Ohm scale, current to the pico amp scale, and Volts to the kV scale. Within the setup it would be most convenient to have logic that records the resulting measurement for varied inputs.

Probing Micro-needles with Micro-Positioners

The probing needles are used to create different circuits for testing different parts of the sensor. In order to test Capacitance, Bias resistance, and Current we must have needles that can act as the ______ that will eventually be attached. Two probing needles are needed to ensure that the bias ring is grounded and that the resulting measurements are read depending on our circuit we create. The micro-positoners allow for gentle place of the needle onto the sensor.


A microscope with a video camera attached allows for less tedious sensor placement and probing. With the view on a large screen one can slowly move the needle around until the shadow of the needle blocking the microscope appears. The needle tip must be on the scale of a micron so correct placement without a microscope is nearly impossible.


p-type semiconducting material

p type material is a conducting material which is placed in strips along the top of the sensor. These strips are always good conductors at good voltage. Each of the strips along the top of the sensor is made of this P type material where an abundance of negative charge holes are present. These negative holes create a material that acts like a strong conductor for .

n type

n type material is a semi-conducting material which acts as an inductor for a certain range of Voltages.

Bias Resistor

For each strip there is a Bias resistor that sits between the top side of the strip and the Bias ring. The Bias resistors is placed between the top side of the sensor and the bias ring so that when a particle is detected the pulse sent out will be greatly slowed. If the bias resistor was removed the pulsed cause by charged particles passing through the sensor would be so fast they would be undetectable.

In order to measure the Bias resistance place a probing needle on the DC pad and send High Voltage through

 The rate at which the sensor is slowed is given by...

DC Pads

DC Pads are Pads that are connected directly to the p type strips for the purpose of testing.

AC Pads

AC Pads are connected to the strips by a non conducting material ([math]\displaystyle{ SiO_2 }[/math]) so that a capacitor is created between the strip and the pad. The AC pads are where the additional logic and ___ chips will be attached during final assembly. Since the Bias resistor is so large the AC pad's capacitor prevents a large current from passing through the logic and overheating the amplifier and sensitive equipment that will be attached.

Spy Pads

Bias Ring

The Bias Ring is a conductive ring that surrounds the entire sensor and connects to every bias resistor. Bias ring connects to ground and completes the circuit

Guard Ring

IV Scan

Iv scans are tests of the resulting Current as Voltage is varied. The IV Scan can be performed on a single strip or the entire sensor depending on the spec's you wish to check. Figure A shows an example IV plot of an entire sensor that is working properly. Many errors can occur during the manufacturing of components of this size and the IV scan is the main method for checking the specs of our sensor.

Depletion Voltage

As Voltage is applied to the bottom of the wafer the potential difference begins to deplete the sensor of charge carriers between the p and n sides. When The depletion Voltage is applied to the sensor a resulting reverse bias will be present with a depletion range that extends to the entire length of the detector. The Depletion Voltage ([math]\displaystyle{ V_d }[/math])is the minimum Voltage required to obtain full depletion of the sensor. The depletion sets the lower limit of the sensor. Keeping the lowest full depletion voltage is extremely important due to the noise, heat and power consumption created from Voltage greater than the minimum depletion Voltage Refer to Figure A to see the point on an IV plot where the sensor become fully depleted.

Figure A: As the slope rises there is point where the IV curve turns into a straight line. The Depletion Voltage can be estimated by two linear fits. One linear fit found during the curve before it flattens out and another after the curve flattens out. The point where these two lines meet is the best estimate of the Breakdown Voltage.

Breakdown Voltage

Figure B: Here is an example of a sensor with a more limited range between the Full Depletion Voltage and the Breakdown Voltage. Sensor can vary greatly on their minimum Breakdown Voltage. The important thing to check is that the gap between the [math]\displaystyle{ V_b }[/math] and [math]\displaystyle{ V_d }[/math] is great enough to allow for easy depletion of the sensor. Also as sensors age the [math]\displaystyle{ V_b }[/math] can reduce. In order for a sensor to be useful it must have a tolerance significant enough for the age to not be an issue within a reasonable amount of time.

After obtaining full depletion the sensor will not maintain Capacitor like properties if the Voltages are brought up too high. At Some point the Voltage will "jump" the gap and begin to conduct again. The minimum voltage where the sensor begins to allow charges to pass is known as the breakdown Voltage ([math]\displaystyle{ V_b }[/math]). A good sensor will have plenty of leeway Voltage between the [math]\displaystyle{ V_d }[/math] and the [math]\displaystyle{ V_b }[/math]. Usually Operation Voltage is 20 Volts above Depletion voltage. The Breakdown Voltage sets the Upper limit of the Operating Voltage. A breakdown voltage less than 20 Volts larger than the Depletion Voltage renders a sensor or strip unusable.

The Breakdown voltage of the sensor is determined by [math]\displaystyle{ V_b = \epsilon_dd \ }[/math] Where,

[math]\displaystyle{ V_b \ }[/math] is the Breakdown Voltage
[math]\displaystyle{ \epsilon_d \ }[/math] is the dielectric strength
[math]\displaystyle{ d \ }[/math] is the plate separation or the thickness of the dielectric (d≈300microns)

Refer to Figure B to see the point where the Voltage begins to break down in the sensor.

Leaky Strips

Leaky Strips occur when the p-type material acts like a conductor more than the surrounding strips due to inconsistencies in the material. The sensor manufacturer will often send the test results along with the sensors and a good start begins by checking the leaky strips as measured by the manufacturer.

Resistance and Capacitance Measurements

Bias Resistor Measurements

To measure the Bias resistance of a single strip one must place a needle on the DC pad and the Bias ring. By applying a voltage to the DC pad and reading the out coming Voltage from the Bias ring one can measure the Voltage drop due to the Bias resistor in that strip.

Capacitor Short

When the report from the manufacturer reports a Capacitor short it is referring to the capacitor created on the AC pad. A short means that the AC pad is not fully insulated from the pad. To test for a capacitor short one must measure the current output from the AC pad during an IV scan. If the output current is significantly higher than the surrounding strips then there must be a short in the Capacitor. As long as the sensor flaw is isolated, understood and accounted for it will not be a problem when the sensor is being experimented with.