Space-qualified, abuttable packaging for
LBNL p-channel CCDs, Part II
R.W. Besuner
1
, C. Baltay
2
, H.T. Diehl
3
, W.T. Emmet
2
, S.E. Harris
4
, P.N. Jelinsky
4
, J.C. Krider
3
,
D.L. Rabinowitz
2
, N.A. Roe
1
1
Lawrence Berkeley National Laboratory
2
Yale University
3
Fermi National Accelerator Laboratory
4
U.C. Berkeley Space Sciences Laboratory
ABSTRACT
Fully depleted, back-illuminated, p-channel CCDs developed at Lawrence Berkeley National Laboratory exhibit high
quantum efficiency in the near-infrared (700-1050nm), low fringing effects, low lateral charge diffusion (and hence
small, well-controlled point spread function), and high radiation tolerance. Building on previous efforts, we have
developed techniques and hardware that have produced space-qualified 4-side abuttable, high-precision detector
packages for 10.5ȝm pixel, 3.5k x 3.5k p-channel LBNL CCDs. These packages are built around a silicon carbide
mounting pedestal, providing excellent rigidity, thermal stability, and heat transfer. Precision fixturing produces
packages with detector surface flatness better than 10ȝm P-V. These packages with active areas of 36.8mm square may
be packed on a detector pitch as small as 44mm. LBNL-developed Front End Electronics (FEE) packages can mount
directly to the detector packages within the same footprint and detector pitch. This combination, along with identically
interfaced NIR detector/FEE packages offers excellent opportunities for high density, high pixel count focal planes for
space-based, ground-based, and airborne astronomy.
Keywords: JDEM, SNAP, SiC, Silicon Carbide, Radiation Tolerant
1. INTRODUCTION
Charge-coupled devices (CCDs) developed at Lawrence Berkeley National Laboratory (LBNL) have several
characteristics giving them advantages over other designs for astronomy, particularly space-based applications
1,2
. Back-
illumination through their relatively large, fully-depleted thickness (100-300μm) enables higher quantum efficiency
(QE) at near-infrared (NIR) wavelengths (up to one micron), with reduced fringing. Operation with high bias voltage
reduces lateral charge diffusion, and contributes to a small, well controlled point spread function (PSF), which is
essential for several types of science, for example precision morphology used in weak lensing observations
3,4
.
Employing p-channel architecture makes these devices intrinsically more radiation tolerant than conventional n-channel
technologies
5
.
Large-format versions of the LBNL CCDs (including 2k x 4k x 15μm, 2k x 2k x 15μm, and the presently-described 3.5k
x 3.5k x 10.5μm devices) have been packaged using various configurations and materials by teams at institutions
including University of California at Santa Cruz (Lick Observatory), LBNL, Fermi National Accelerator Laboratory
(Fermilab), and Yale University
6,7,8
. This paper reports on developments built upon those efforts, including recent
efforts described in part I of this paper
9
. The device described here and designed at LBNL is a 3.5k x 3.5k x 10.5μm-
pixel CCD, four of which are shown prior to dicing on the wafer in Figure 1. This device is manufactured on 150mm
wafers by DALSA Semiconductor, with thinning, backside processing, and metallization at the LBNL Microsystems
Laboratory
10
. It has an active area that is 36.8mm square, and overall size of 39.5mm x 38.9mm. Electrical
interconnection is through wirebond pads along two edges. For this application, the device is thinned to 200μm.
High Energy, Optical, and Infrared Detectors for Astronomy IV, edited by Andrew D. Holland, David A. Dorn
Proc. of SPIE Vol. 7742, 77420H · © 2010 SPIE · CCC code: 0277-786X/10/$18 · doi: 10.1117/12.856469
Proc. of SPIE Vol. 7742 77420H-1