— 4 —
itself may have changed. You can check the current page or check for previous versions at the Internet Archive.
Yahoo! is not affiliated with the authors of this page or responsible for its content.
4
4
Source Standards for the Radiometric
Calibration of Astronomical Instruments in the
VUV Spectral Range Traceable to the Primary
Standard BESSY
J
ORG
H
OLLANDT
Physikalisch-Technische Bundesanstalt
Berlin, Germany
M
ICHAEL
K
UHNE
Physikalisch-Technische Bundesanstalt
Berlin, Germany
M
ARTIN
C.E. H
UBER
International Space Science Institute
Bern, Switzerland
B
URKHARD
W
ENDE
Physikalisch-Technische Bundesanstalt
Berlin, Germany
On the basis of a high-current hollow-cathode discharge we have developed two trans-
fer source standards suitable for the radiometric calibration of vacuum-ultraviolet (VUV)
telescopes. The source standards are transportable and (in their current design) produce
collimated beams of 5 mm (grazing-incidence region) and 2.5 mm, 5 mm, 10 mm and
15 mm (normal-incidence region) diameter. By irradiating the entrance aperture of the
telescope with this beam, the overall spectral response of the instrument can be determined
and spectral-responsivity variations over the entrance aperture can be directly evaluated.
The transfer standards described in this paper have been calibrated in the radiometry labo-
ratory of the Physikalisch-Technische Bundesanstalt (PTB) by use of the calculable spec-
tral photon ux of the Berlin electron storage ring for synchrotron radiation BESSY I: a
primary radiometric VUV source standard. The output of the source standards has been
determined at 57 emission lines covering the wavelength range 15 nm to 150 nm. The
photon ux in these emission lines ranges from 10
4
s
1
to 10
9
s
1
and the overall relative
standard uncertainty of the photon ux in any given line is found to be not more than 8 %.
Note: This paper is an updated version of J. Hollandt et al., 1996, A&AS 115, 561
4.1
Introduction
The radiometric calibration of instruments is based on either source or detector stan-
dards. Since in the vacuum-ultraviolet (VUV) spectral domain reliable celestial standards
are still being established, the radiometric calibration of space-borne VUV instruments
must be based on laboratory standards.
In the past, calibrations in the vacuum-ultraviolet have been made almost exclusively
by use of detector standards rather than source standards. The detector standard was used
to determine the photon ux entering the telescope when the complete instrument was illu-
minated with a monochromatic photon beam. The ratio of the instrument detector response
1
2
4. S
OURCE
S
TANDARDS
T
RACEABLE TO THE
P
RIMARY
S
TANDARD
BESSY
and the measured monochromatic ux entering the telescope then represented the spectral
responsivity of the complete instrument at this particular wavelength. This procedure was
repeated at different wavelengths in order to establish the wavelength-dependent spectral
responsivity of the instrument. For practical reasons, in many cases the overall calibration
was made in two (or more) steps: the transmission of the telescope and the responsivity of
the spectrometer-detector system had to be measured separately [Reeves et al., 1977].
A wide variety of VUV detector standards is available in the form of photoemission
diodes [Caneld and Swanson, 1987], sodium-salicilate-based detectors [Philips, 1984],
proportional counters [Henke and Tester, 1975; Kroth et al., 1990] or photoelectric ion-
ization chambers [Samson, 1964; Samson and Haddad, 1974]. The relative uncertainties
of these detector standards are of the order of 10 % (1
), and, particularly for photoemis-
sion diodes and sodium-salicilate-based detectors, long-term stability is a problem. The
possibilities offered by the signicantly lower uncertainties that have been achieved using
electrical substitution radiometers operating at liquid-helium temperature as the primary
detector standard [Rabus et al., 1997; Shaw et al., 1999] will be briey addressed in Sec-
tion 4.5 of this paper.
In contrast to the rather complex procedures that are required to accommodate the use
of detector standards, a radiometric instrument calibration based on source standards is
straightforward. If the radiation of the source is well-matched to the demands of the cali-
bration and a level of uncertainty of a few percent is sufcient, the instrument to be cali-
brated is illuminated with the source of known spectral emission and the spectral response
of the detector is measured. This classical approach to radiometric calibration could not,
however, be used in the VUV until about two decades ago, since such sources were not
available.
During the last two decades, VUV source standards in particular, dedicated electron
storage rings (in fact, the primary radiometric VUV standard) have become available and
several instruments have been calibrated with the SURF-II storage ring at the U.S. National
Institute of Standards and Technology (NIST) [Furst et al., 1995]. Similar activities took
place at the positron storage ring SUPER-ACO at the Institut dAstronomie Spatiale (IAS)
at Orsay [Besson et al., 1989].
The spectral distribution of radiation at most synchrotron sources (see Figure 4.1 for
BESSY I) is usually not well matched to the VUV emission of the Sun or other stars.
The increase of photon ux with decreasing wavelength of the typical synchrotron source
can cause extremely high contributions from high orders of the spectrometer grating, es-
pecially for grazing-incidence optics. In order to overcome some of the major difculties
that occur when space instruments are calibrated by direct use of synchrotron radiation,
we have developed transfer source standards. Being based on a conventional emission-
line source (a hollow-cathode discharge source) with a collimated output beam, our trans-
fer standards avoid the problems associated with the uncollimated, spectrally-continuous,
synchrotron radiation, whose polarization, moreover, varies with the viewing angle. The
hollow-cathode discharge source we describe here emits well-separated emission lines and
thus avoids the often severe problem of high-order contributions that occur in grating in-
struments. It also provides a wavelength calibration as a matter of course. The transfer
standards in question are transportable and can, accordingly, be mounted on the test tank
in the laboratory, where the instrument to be calibrated has been developed. Thus there is
no need for a special calibration tank nor is there a need to transport the instrument for cal-
ibration. This also removes schedule incompatibilities as they are bound to occur with the
4.2. Transfer Source Standards
3
Figure 4.1: Comparison of a black-body radiator at a temperature of 3200 K and the
synchrotron-radiation spectrum of the electron storage ring BESSY I at an electron energy
of 340 MeV and 850 MeV.
inexible beam-scheduling that is mandatory for multi-purpose storage ring facilities. A
given transfer standard can, in principle, also be brought to different instruments so that an
intercalibration by use of the same standard can be achieved. A further interesting property
of our transfer source standards is that they produce detector count-rates in the telescope
systems under calibration which are comparable with those expected for the instruments
when in ight.
4.2
Transfer Source Standards
4.2.1
Hollow-cathode Source
The source standards for the measurement of the spectral responsivity of astronomi-
cal instruments are based on a high-current, hollow-cathode, glow-discharge source (Fig-
ure 4.2). This source emits intense, unpolarized, line-radiation from the buffer gas and the
cathode material (99.5 % aluminium) at wavelengths above 15 nm (Figure 4.3). Equipped
with a compact, two-stage, differential pumping system, the source allows windowless
observation of its VUV radiation under ultrahigh-vacuum conditions. The ux-limiting
aperture stop with a diameter of 0.6 mm is part of the differential pumping system.
As a radiometric transfer standard the source is operated with a xed current (e.g.,
1 A) with a current stability better than 1 mA. The voltage drop over the hollow-cathode
discharge (e.g., 500 V) is sensitive to buffer-gas pressure (typically 1 mbar). It can be sta-
4
4. S
OURCE
S
TANDARDS
T
RACEABLE TO THE
P
RIMARY
S
TANDARD
BESSY
bilized (to better than
1 V) by regulating the buffer-gas ow with an automatic needle-
valve (Balzers RME 010) and a self-regulating voltage control unit. In this mode of oper-
ation, the radiant intensity of the VUV emission lines is reproducible within
5 % over a
period of 40 operating hours.
"!$#&%
')(1032
465&78469A@8B&CED
F&G
H6IPIQI8R8H6SPT&U
V&W8X`Y6a
b6ced)f8gihPp`q
r
s8t6g&c
uwvyx
8P6)
P8`6
66ed
f6g
hi
jklem)n&oPp6p&q
r
s8t
uv
wPxyv
z
{yz
|}z
~8
)8&
|w
e
Figure 4.2: Longitudinal section of the hollow-cathode source with an integrated two-
stage, differential pumping system.
In order to provide a large number of calibration lines within the spectral range from
15 nm to 150 nm, the hollow-cathode source is alternately operated wi