Download Detection of Light: From the Ultraviolet to the Submillimeter PDF

TitleDetection of Light: From the Ultraviolet to the Submillimeter
File Size5.4 MB
Total Pages376
Document Text Contents
Page 1

Page 2

Detection of Light

Detection of Light provides a comprehensive overview of the important
approaches to photon detection from the ultraviolet to the submillimeter spectral
regions. This expanded and fully updated second edition discusses recently
introduced types of detector such as superconducting tunnel junctions, hot
electron bolometer mixers, and fully depleted CCDs, and also includes
historically important devices such as photographic plates. Subject matter from
many disciplines is combined into a comprehensive and unified treatment of the
detection of light, with emphasis on the underlying physical principles. Chapters
have been thoroughly reorganized to make the book easier to use, and each
includes problems with solutions as appropriate. This self-contained text
assumes only an undergraduate level of physics, and develops understanding as
it is needed. It is suitable for advanced undergraduate and graduate students, and
will provide a valuable reference for professionals in astronomy, engineering,
and physics.

george rieke is a Professor of Astronomy and Planetary Sciences at the
University of Arizona. After receiving his Ph.D. in gamma-ray astronomy from
Harvard University, he focused his work on the infrared and submillimeter
spectral ranges. He has been involved in instrumentation and detectors
throughout his career, applying them to the studies of planets, forming stars,
active galactic nuclei, and starburst galaxies. Rieke has also helped to establish
the foundations of infrared astronomy in areas such as calibration and
instrumental techniques, and is author or co-author of over 300 publications in
these areas.

Page 188

6.4 CMOS imaging arrays 175

dispersion in Si:As BIB detectors or with avalanche photodiodes). However, under
very low light level conditions where the CCD must be read out rapidly, the large
output signals overwhelm the CCD read noise and the signal to noise is improved. Performance

Currently available scientific CCDs have dimensions ranging roughly from 500 ×
500 to 4000 × 4000 or more pixels and achieve noise levels of two to five elec-
trons (rms). Peak quantum efficiencies (achievable from ∼0.3 to ∼0.9 µm) can be
80–100% if the device is anti-reflection coated. However, the quantum efficiency
falls fairly rapidly from ∼0.3 µm into the ultraviolet. These performance levels
require operation at an optimum temperature of ∼150 K. For comparison, front-
illuminated CCDs have peak quantum efficiencies of 30–40%, and front-illuminated
devices with anti-blooming gates may only reach 20%.

6.4 CMOS imaging arrays

In many ways, the back-illuminated CCD is virtually a perfect detector for the visible
spectral region, with: (1) nearly 100% quantum efficiency (if anti-reflection coated);
(2) read noise of a few electrons or even less, nearly equivalent to counting single
photons; (3) excellent imaging properties (∼100% fill factor, very small crosstalk,
good linearity); (4) virtually no dark current (when cooled to ∼150 K); and (5) avail-
ability in large formats (currently up to nearly 108 pixels). However, even virtual
perfection can have its flaws. CCDs operated to achieve the lowest noise are un-
avoidably slow to be read out. These detectors also impose rigid constraints on the
readout format, without the flexibility of random access to arbitrary subsets of their
pixels. The charge transfer, and hence many desirable performance attributes, can
be degraded by exposure to energetic ionizing radiation. Perhaps most importantly,
their production has become a relatively specialized sideshow in integrated circuit
fabrication, removed from the center stage of computer chip processing.

The unique advantage of a charge transfer architecture is that it can provide a
filled array on a single wafer without the presence of nonresponsive regions ad-
jacent to the pixels for individual readout amplifiers. However, as transistors have
been made smaller to allow increasing densities in computer electronics, their claims
on array real estate have shrunk sufficiently to challenge the necessity for this fea-
ture. High performance arrays can now be manufactured in standard complementary
metal–oxide–semiconductor (CMOS) facilities, with silicon photodiodes and their
readout circuitry produced on a single silicon wafer in a common series of processing

The simplest form of CMOS imager is designed virtually identically to the hybrid
infrared array circuit illustrated in Figure 6.2, noting that the photodiode sensors are

Page 189

176 6 Arrays

on the silicon chip itself rather than being grown in another material and hybridized
to a silicon readout chip. These “passive pixel sensors” achieve fill factors of 70 to
80% with read noises less than 50 electrons, in some cases significantly less.

Lower read noises can be achieved by providing gain in the amplifier for each
pixel, such as by using CTIAs. These “active pixel sensors” have been demonstrated
with noise of a few electrons at normal video frame rates (30 frames/s). Similar format
CCDs have read noises of 10–20 electrons at these frame rates. The price for this
low noise is a small fill factor, 20–30% currently, because of the space required for
the relatively complex readout amplifiers. The effective fill factor can be increased
to 50% or more by depositing a grid of tiny lenses over the array, one to concentrate
light onto each pixel. Cooling a detector array with such microlenses is not yet

The performance of CMOS arrays is limited currently by relatively low pixel
quantum efficiency as well as by the low fill factors discussed above. So far, most
of the development has emphasized competition with CCDs in consumer products,
not the demands of very low light level imaging. However, CMOS imagers have
important potential intrinsic performance advantages, such as greater resistance than
CCDs to the detrimental effects of energetic particles, as well as the potential to
simplify system design by integrating circuit elements with the sensor. Future efforts
can therefore be anticipated to improve their low light level performance and to
address other current performance limitations.

6.5 Direct hybrid PIN diode arrays

High performance arrays can also be manufactured in the same fashion as hybrid
infrared arrays, but using PIN silicon diodes. These devices have better red quantum
efficiency, greater resistance to blooming, and less susceptibility to cosmic-ray dam-
age than conventional CCDs. They can also have the flexibility of a nondestructive,
random access readout, which can be fast since charge transfer is not a concern. How-
ever, they have many of the manufacturing liabilities of the infrared arrays: complex
processing, indium bump hybridizing, reduced yields, and limitations in array size.

6.6 Array properties

6.6.1 Fixed-pattern noise

It is impossible to make the individual pixels of an array absolutely identical. The
differing pixel properties distort the electronic representation of the pattern of illumi-
nation on the array and therefore constitute a form of noise. Unlike the kinds of noise
discussed up to now, this noise is nonrandom and remains stable from one picture to
the next; it is therefore called fixed-pattern noise. A series of calibration steps can, in

Page 375

362 Index

responsivity (cont.)
of photoconductor 36
of photodiode 82
of QWIP 108
measurement of 135

Reticon 209
Richardson–Dushman equation 194, 309
rise time 16

Sν , Sλ, see flux density
S, see responsivity
sampling of integrating amplifier outputs

sampling up the ramp 133
saturation current 90
Schottky diode 99
Schottky diode mixers 307
Schrödinger equation 105
SEC vidicon, see secondary electron

conduction vidicon
secondary antenna 303
secondary electron conduction vidicon 213
Seebeck coefficient 266
semiconductor bolometer 240
semiconductor processing techniques 59
shot noise 71, 83
signal-induced background 208
signal processing in the element 40
silicon intensified target vidicon 213
silicon target vidicon 212
silver-halide grains 217
silver speck (in photography) 211, 221
single mode detector 289, 304
single sideband 278
SIS mixer, see superconductor–insulator–

superconductor mixer
SIT vidicon, see silicon intensified target

skipper readout 174
“solar blind” detector 192
solid angle 4
solid state photomultiplier 74
space charge 42
spatial frequency 12
specific heat (of bolometer materials) 255
specific intensity 6
spectral radiance 2
speed in photography 227
“spider web” bolometer 262
spiking on photoconductor output 67

spot sensitometer 233
SPRITE, see signal processing in the element
square-law detector 285
square-law mixer 277
SQUID, see superconducting quantum

interference device
SSPM, see solid state photomultiplier
ST vidicon, see silicon target vidicon
STJ, see superconducting tunnel junction
stressed detector 68
stripline tuning structures 315
superconducting bolometer 250
superconducting quantum interference device

superconducting tunnel junction 109
superconductivity 24
superconductor–insulator–superconductor mixer

superlattice 107
SUR, see sampling up the ramp

T , see transmittance
Tc, see critical temperature
TN, see noise temperature and thermal noise
T (λ), see transmittance
TDI, see time-delay integration
temperature coefficient of resistance 240
TES, see transition edge sensor
T-grains 228
thermal conductance (bolometer) 239
thermal detector 8
thermal diffusion charge transfer 161
thermal excitation 21

in diode mixer 309
in extrinsic material 62
in intrinsic material 48
in photodiode 92
in photoemissive detector 194
in photography 221
in STJ 111

thermal limit 293
thermal noise (bolometer) 247
thermistor bolometer 266
thermopile detector 266
Thévenin equivalent circuit 121
thinning of detectors 150, 154
TIA, see transimpedance amplifier
time-delay integration 40, 170
transient response 65
transimpedance amplifier 120

Page 376

Index 363

transition edge sensor 252
transmittance 5
transparent contacts 32
transverse contacts 32
trap 23

in CCD 154, 163, 164
role in photography 220

triple-correlated sampling 131
tuning circuits for SIS mixers 315
tunneling 81, 106
turbidity 234

ultraviolet CCDs 156
ultraviolet detector materials 84
ultraviolet flooding of CCDs 154
underexposure in photography 226

valence band 18
vidicon 211

waveform factor 136
waveguide 305
waveguide feed 305
well depth 133–34

in CCD 156, 166
white noise 46, 132
Wiedemann–Franz relation


Y-factor 297

“Z-plane technology” detector array 185
zone refining 60

Similer Documents