DESIGN AND USE OF A LARGE-FORMAT CCD INSTRUMENT FOR THE IDENTIFICATION AND STUDY OF DISTANT GALAXY CLUSTERS by GERARD ANTHONY LUPPINO B.S. Astronomy and Physics The Pennsylvania State University (1983) SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN PHYSICS at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY August, 1989 @ Massachusetts Institute of Technology 1989 Signature of Author Department of Physics August, 1989 Certified by George R. Ricker Thesis Supervisor Accepted by George Koster Chairman, Department Committee SEP 1 2 1989 ARCHjvre DESIGN AND USE OF A LARGE-FORMAT CCD INSTRUMENT FOR THE IDENTIFICATION AND STUDY OF DISTANT GALAXY CLUSTERS by GERARD A. LUPPINO SUBMT17ED TO THE DEPARTMENT OF PHYSICS OF THE MASSACHUSEITS INSTITUTE OF TECHNOLOGY ON AUGUST 1989, INPARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY INPHYSICS ABSTRACT This dissertation describes a search for distant clusters of galaxies (z>0.3). Cluster candidates were selected from a sample of radio sources with steep, low-frequency spectra since such radio sources are often found in rich nearby galaxy clusters. The steep-spectrum radio source (SSRS) is thought to be produced by the confinement of the radio plasma by the thermal pressure of a hot intracluster gas; therefore, we expect clusters discovered by this selection criterion to be bright x-ray sources and excellent targets for future x-ray satellites such as ROSAT, ASTRO-D and AXAF. These objects are also more likely to be true physical associations rather than the chance superpositions of galaxies that may plague some optically selected samples, particularly those at faint magnitudes. In order to identify the optical counterpart of the SSRS and see if the radio source was part of a distant cluster of galaxies, optical images to faint limiting magnitudes were required. Therefore, an imaging instrument incorporating mosaics of CCDs was designed and built . With this instrument, it is possible to observe faint, distant clusters of galaxies with spatial resolution limited by atmospheric seeing, while at the same time covering a relatively large field of view. The major task in the creation of this instrument was the design and construction of two separate CCD camera systems; each based on different CCD mosaics. The first camera system incorporates four prototype, TI 850 x 750 virtual-phase CCDs (22.3 gim square pixels) arranged in a square pattern with the device packages abutted. The second camera system is designed to operate a four-chip CCD imager made by MIT Lincoln Laboratory. The imaging area of this CCD mosaic is 840 x 840 pixels (27 pm square pixels) formed by abutting four 840 x 420 framestore devices with seam losses less than 6 pixels. Details of the CCD camera electronics design as well as details of the mechanical design of the CCD cameras and LN 2 dewars are described. A total of 30 SSRS fields were observed with this CCD instrument mounted on the 1.3m and 2.4m telescopes of the Michigan-Dartmouth-MIT (MDM) Observatory on Kitt Peak. Additional CCD images of 5 of these radio sources were obtained with the Kitt Peak 4m telescope equipped with the KPNO prime focus CCD. Of these 30 SSRSs, 26 have been optically identified. In all, 15 of the observed fields contain visible clusters of galaxies, some of which are quite rich. Of the remaining 15 SSRSs, 6 are identified with faint galaxies , 2 are identified with poor groups, 3 are probably quasars, and 4 of the fields have uncertain or no optical counterpart. The median value for the estimated redshift of the 15 clusters is z=0.43 and the median richness is Abell Class 2. Nearly all of the radio sources have classical double-lobed morphology, and in half of the sources associated with clusters, the radio source is identified with a fainter cluster member rather than with the brightest cluster galaxy. In all cases where the cluster is BM Type I, however, the radio source is identified with the optically dominant galaxy. Observations are also presented of three distant clusters that were not part of the radio selected sample: 0414+009, an x-ray bright BL Lac object coincident with the optically dominant galaxy in a Richness Class 1 cluster; 1217+1006, a compact group containing an extremely blue (B-R) ~ 0.0 object that may be either a QSO or a BL Lac; and 1358+6245, a new, x-ray luminous (Lx (0.5-4.5 keV) = 8.4x10 44 erg s-l), extremely rich (Richness Class 3 or 4), distant cluster of galaxies (z=0.323) which exhibits the Butcher-Oemler effect with a blue galaxy fractionfb = 0.18 orfb = 0.10 depending on the background galaxy correction. ACKNOWLEDGMENTS This instrument and scientific program would not have been possible without the help of a large number of people. First of all, I would like to thank my thesis advisor, George Ricker, who gave me access to a large number of CCDs and allowed me to use some of the best of them in this instrument. He also provided guidance and encouragement, and allowed me the freedom to be creative in the instrument design. I am also indebted to Ian McHardy and Brin Cooke, who developed this method for searching for distant galaxy clusters, and graciously accepted me as a collaborator. I appreciate the constructive comments and advice I received from my thesis committee members, Claude Canizares, and John Tonry. I am especially grateful to John, who taught me how to think as a scientist, and changed my perspective on instrumentation, urging me to "put the science first and let the science requirements drive instrument designs." John was always there to answer questions and help me wring the most out of my data. In addition, as the reader will later discover, nearly all of the software tools I use for data reduction in this thesis were developed and kindly provided by John. Also, thanks to Paul Schechter for helpful advice and criticism during the final stages of this project. During my first three years as a graduate student, I worked closely with John Vallerga, who taught me almost everything I know about CCDs. The time I spent working with John was the most "productive fun" I ever had in the laboratory. And, I can never thank Roland Vanderspek enough. During my entire time at MIT Roland has been my best friend, companion, and advisor. He has helped me with virtually everything and I will be eternally grateful. I also appreciated the "baked goods" that often appeared when I was working on some particularly tedious or difficult problem. The designs for much of the CCD electronics used in the instrument described later are based on original designs by John Doty. I first met John when I visited MIT 6 1/2 years ago, and he greatly influenced my decision to come to MIT to work in the CCD Lab (then the Balloon Lab). I have never regretted that decision. The Center for Space Research (CSR) CCD Lab has been a great place to work: partly because of the facilities, but mostly because of the people. I thank Mark Bautz and Pat Mock for relieving me of my responsibilities for the laboratory x-ray CCD system so that I could spend my time working on optical instrumentation and the scientific program for this thesis. Mark has also been an endless source of scientific advice and information about clusters of galaxies, and I appreciate the time he spent answering many of my questions. Pat not only helped me finish some unfinished electronic designs, he designed and built the hardware to interface the laboratory CCD cameras to our SUN and MicroVAX computers. Pat also went observing for me when I was too busy. Lawrence Shing and Ed Boughan deserve the credit for debugging the electronic design for the Lincoln Laboratory CCD system and actually constructing the circuits and making them work while I was busy writing this thesis. I have also enjoyed working with (as well as sharing a house with) George Mitsuoka (thanks for all the stargate) and Steve Rosenthal. A few years ago, I got my feet wet working at MDM Observatory when George and I upgraded the MASCOT with new CCDs and electronics. George also wrote MIDAS to interface the MASCOT to one of our old generic computers. Recently, Steve wrote the CCD tool program and ultimately interfaced the MASCOT and (later) the BRICC to the MDM Observatory SUN computers. I must also thank Mike Decker, who helped me sift through many TI-4849 CCDs while I was searching for the one to use in the BRICC, and Suzan DeFreitas who helped with much of the galaxy cluster data reduction and with the construction of some of the instrument hardware. I thank Rosemary Hanlon for her help with some of the figures (and for dealing with Graphic Arts), and for her useful tips concerning WORD and the Macintosh. And thanks, Shep, for always making me laugh. CSR machinists David Breslau and Leo Rogers were responsible for actually bringing the instrument design to life. During the course of construction, their excellent suggestions often improved the instrument design and their craftsmanship and attention to detail is evident in the final product. In addition, I am especially grateful to our CSR purchaser, Dan Calileo, who was always able to obtain the parts and services I required. I would like to thank Barry Burke and his co-workers in Group 87 at MIT Lincoln Laboratory for their marvelous CCDs. I have enjoyed working with Barry and I am grateful for the education he has given me concerning some of the details of CCD design and manufacture.