PHOBOS Experiment

John J. Ryan, Ph.D.

PHOBOS

PHOBOS is one of the experiments being built at the Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory (BNL), on Long Island, NY. The collaboration includes Argonne, Brookhaven, Krakow, MIT, UIC, Maryland and Yale. PHOBOS is now in the construction phase, and initial operation is expected in late 1999, when the first beams from RHIC become available.

The experiment will study nucleus-nucleus collisions at total center-of-mass energies up to 200GeV per nucleon for collision systems as heavy as Au+Au. The aim of the PHOBOS experiment is to search for manifestations of new phases of hadronic matter, which are theorized to be produced under the conditions of extreme energy densities which are expected to occur in these collisions.

The experiment will measure the production of particles emitted over essentially the entire solid angle, using a Multiplicity Array composed of various types of silicon detectors and incorporating a Vertex Detector for the determination of the interaction point. Approximately 1% of the particles emerging from a collision will enter one of the two Spectrometers, which consist of planes of silicon detectors arranged at the entrance of and inside of a dipole magnet.

PHOBOS Design

I worked with Wit Busza from M.I.T. and Steve Manly from Yale to layout the initial design of the PHOBOS detector. We listed our physics goals on one hand and our financial constraints on the other. Working down to the minimum detector specifications and the maximum number of elements and electronic channels, we found the problem to be fairly tightly constrained, and we came away with a design which has held up to subsequent Monte Carlo GEANT simulations.

Silicon Detectors

I spent the majority of my time on PHOBOS working with the silicon detectors. I was the Manager of the Silicon Detectors for the experiment. 85% of the experiment's detectors are being constructed from silicon. I was responsible for the design of the devices, and I had done a lot of prototype work in choosing the design. This work included testing the technically challenging aspects of the devices by building prototypes and by developing testing procedures.

The PHOBOS project needed pad (pixel) detectors with low mass but high rate capability. I worked closely with Dr. Hobie Kraner at Brookhaven National Laboratory's Instrumentation Division in this research and development work.

We experimented with several types of insulating material applied over a silicon pad device, over the top of which we etched metal traces to carry the signals from the pads to the edges of the device. With such thin layers of dielectric, we had to investigate the capacitive coupling of the traces to the other pads over which they ran. We performed detailed measurements of capacitance and cross talk.

I designed a prototype device to measure the parameters necessary to operate an AC-coupled silicon device, biased by a FOXFET structure. Hobie Kraner produced this device at BNL, and I tested it at MIT.

The National Central University in Taiwan produced a silicon pad detector for use in the CERN experiment WA98. As MIT was involved in this experiment, we tested this device at MIT. This device was AC-coupled, biased with polysilicon resistors. We performed initial tests on an Aelisi probe station and more extensive tests with the devices bonded to hybrid electronics.

Software

With Clive Halliwell at the University of Illinois at Chicago, I laid out the initial PHOBOS program environment, carrying over our experience from Fermilab Experiment-665. With help from several others, we wrote up a description of the PHOBOS Object-Oriented Data structure; I provided some of the low level methods for creating and accessing these sub-objects.

I worked with Clive to develop a pattern recognition program for the data from our silicon spectrometer arms. This method first picked up hits and fit them to straightlines, from the vertex to the edge of the dipole magnet. Then, it projected these trajectories into the magnet to pick up hits in those silicon planes. Finally, the program performed a full fit, coupling the straightline and the circle at the field edge, using a hard-edged field approximation. The pattern recognition was complicated by fringe fields, but the major problem was multiple coulomb scattering from plane to plane.

Last modified: Mon Sep 20, 1999