Chemical Physics 290 (2003) 251–256
www.elsevier.com/locate/chemphys
Absolute absorption cross section measurements of
CO2 in the wavelength region 163–200 nm and the
temperature dependence
W.H. Parkinson, J. Rufus 1, K. Yoshino *
Harvard-Smithsonian Center for Astrophysics Cambridge, 60 Garden street, MS-50, Cambridge, MA 02138, USA
Received 18 November 2002
Abstract
Laboratory measurements of the absorption cross section of CO2 at the temperatures 195 and 295 K have been made
throughout the wavelength region 163–200 nm by using a high resolution grating spectrometer. Cross sections at 195 K
are smaller than those at 295 K, and the band structures are more emphasized as expected. In combining with our
previous measurements [J. Quant. Spectrosc. Radiat. Transfer, 55 (1996) 53], the absorption cross sections of CO2 are
available in the wavelength region 117.8–200.0 nm at 295 K and 117.8–192.5 nm at 195 K.
Ó 2003 Elsevier Science B.V. All rights reserved.
Keywords: Carbon dioxide; Cross sections
1. Introduction
Since absorption of radiation from the sun is
the predominant energy source in the atmospheres
of planets, the strength and spectral details of
photoabsorption by their major constituents
(CO2 ; H2 O; H2 ; CO; CH2 ; CH4 and N2 ) are paramount in the construction of atmospheric models.
In photochemical computations for the predominately CO2 Martian atmosphere, the problem of
*
Corresponding author.
E-mail address: [email protected] (K. Yoshino).
1
Present address: Space and Atmospheric Physics, London,
SW7-2AZ, UK.
the stability of the CO2 level and the recycling of
CO2 to maintain the O2 and CO levels remains
unsettled and has led recently to a number of new
models of the atmosphere [1,2]. All of these models
are very sensitive to the amounts of CO2 and to the
values of the absorption cross sections of CO2 and
H2 O around 200 nm.
The first measurements of photoabsorption
coefficients of CO2 were done by Wilkinson and
Johnston [3] in the wavelength region 144–167 nm
by the photographic technique. The first photoelectric measurements were reported by Inn et al.
[4] in the 106–180 nm region with a resolution of
. Nakata et al. [5] measured absorption
0.85 A
coefficients of CO2 in the wavelength region 58–
. All of those
167 nm with a resolution of 0.2 A
0301-0104/03/$ - see front matter Ó 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0301-0104(03)00146-0
252
W.H. Parkinson et al. / Chemical Physics 290 (2003) 251–256
measurements of photoabsorption coefficients depended on the basic parameters such as pressure
and temperature discussed by Okabe [6]. Heimerl
[7] presented photoabsorption cross sections of
CO2 (although he used absorption coefficients in
, and tabulated
his title) with a resolution of 3.4 A
in the wavelength region
the values for every 5 A
166–182 nm. Ogawa [8] presented cross sections of
CO2 in the wavelength region 172–216 nm with
, and tabulated the values
a resolution of 0.057 A
at the minima in the spectrum. Shemansky [9]
) of
made higher resolution measurements (0.038 A
extinction coefficients (cross sections) in the
wavelength region 210–250 nm and he also made
cross section measurements with low resolution (4
) at 170–210 nm. He concluded that the cross
A
section above 204 nm appeared to be due almost
entirely to Rayleigh scattering. DeMore and Patapoff [10] measured temperature and pressure dependence of extinction coefficients (cross sections)
in the wavelength region in 184–220 nm with a
by using a 5 cm pathlength and
resolution of 0.6 A
very high pressure of CO2 (up to 48 atmospheres).
As mentioned in their discussion, they did not
have enough resolution to see the temperature
dependence of band structures. Lewis and Carver
[11] obtained the cross sections of CO2 at temperatures of 200, 300, and 370 K in the range of
and with
175–197 nm with a resolution of 0.05 A
. Chan et al. [12] observed the CO2
steps of 0.5 A
electronic spectrum by using their dipole (e,e)
method of electron-impact excitation down to 6 eV
(206 nm). Our previous measurements [13] at 195
and 295 K covered the wavelength region 117.8–
. Julienne et al.
175.5 nm with a resolution 0.07 A
[14] calculated the temperature dependence of the
integrated cross sections of CO2 . Anbar et al. [15]
discussed the impact of high resolution and temperature dependence of cross section measurements of CO2 on photodissociation in the
atmosphere of Mars. A theoretical approach to the
photodissociation and assignments of CO2 are
presented by Sadeghi and Skodje [16].
We confirmed from the high resolution photographic spectrum of Cossart-Magos et al. [17]
taken with their 10 m spectrograph that resolution
available with the 6-m spectrometers at Center for
Astrophysics (CfA) and at the photon factory
(PF), KEK, Japan were good enough to use for
the cross section measurements. We report here
the results of absolute cross section measurements
of CO2 at the temperature 195 and 295 K in the
wavelength region 163–200 nm. In combination
with our previous results [13], cross section data
are available in the wavelength region 117.8–200.0
nm at 295 K and 117.8–192.5 nm at 195 K.
2. Experimental procedure and measurements
For the wavelength region below 173 nm, the
photoabsorption cross section measurements were
carried out at the 2.5 GeV storage ring at the PF,
KEK, Japan. A 6.65-m spectrometer, equipped
with a focal plane scanner and entrance and exit
slits of 20 lm 10 lm, was used to provide high
spectral resolution. A reciprocal dispersion of approximately 0.001 nm/mm was achieved in the
third order of a 1200 line l/mm grating blazed at
550 nm. A zero-dispersion predisperser was used
to reduce the bandpass of the synchrotron continuum radiation entering the spectrometer to
about 80 nm. The absorption cell, which is located
between the predisperser and the entrance slit assembly of the 6.65-m spectrometer, has two MgF2
windows (2 mm thickness) and provided an optical
path length of 24:36 0:01 cm. The cell can be
immersed in a dryice/alcohol bath to obtain a CO2
gas temperature of 195 K. We divided the spectral
region into 10 sections of about 1.0 nm extent. We
obtained data points every 0.001 nm at a resolution of 0.0008 nm with the counting period of 4 s.
For the wavelength region above 171 nm, the
6.65-m normal incidence, vacuum spectrometer at
the CfA was used in the first order of 2400 and
1200 l/mm gratings to obtain photoabsorption
cross section measurements of CO2 . Both gratings
, and provided an estimated
were blazed at 1500 A
instrumental width of 0.0021 and 0.0030 nm, respectively. To measure the weak photoabsorption
of the CO2 bands in the wavelength region 171–
200 nm, we required a high column density of the
gas. We obtained this by using a multi pass technique, a White cell [18]. The White cell was designed to have a distance of 1.50 m between two
main mirrors, and was set for 2–4 double passes
W.H. Parkinson et al. / Chemical Physics 290 (2003) 251–256
253
Fig. 1. Schematic diagram of the White cell setup. Only the
section between mirrors (1.5 m) can be cooled.
making a path length of 6–12 m. The set up of the
White cell is shown in a schematic diagram in Fig.
1. The cell was connected by one inch in diameter
Pyrex tubing to the entrance slit of the spectrometer, and to the light source assembly. Two fused
silica windows (Suprasil 1) were used, one at the
entrance slit and the other at the light source.
These connecting sections provided an additional
path length of 83.5 cm outside of the White cell
and these sections could not be cooled down to 195
K. The continuous background for the photoabsorption measurements was provided by a hydrogen discharge lamp operated at about 500 mA.
CO2 gas was frozen in a stainless cylinder immersed in liquid nitrogen, and the frozen product
(dry ice) was pumped by a Hg diffusion pump for
purification. The CO2 was warmed up slowly and
kept in the cylinder at high pressure. The CO2
pressure used in the White cell was varied from 1
to 1000 Torr depending on the wavelength region,
and was measured with a capacitance manometer
(MKS Baratron, 10 and 1000 Torr). The optical
paths between the mirrors in the White cell could
be cooled to 195 K, but the tubes connecting the
White cell to the spectrometer and a discharge
lamp were at room temperature (295 K), and
therefore a correction was required to the observed
absorption (optical depth). The absorption cross
section rðlÞ is obtained from the optical depth (the
natural log of the ratio of the incident intensity to
the intensity transmitted through a medium) divided by the column density.
The background continuum level was established by counting with the cell evacuated at the
beginning and end of each scan. We divided the
Fig. 2. Correction to the observed absorption at 195 K. The
estimated optical depths at 295 K (broken) is subtracted from
the observed optical depth (dotted) to obtain the optical depth
at 195 K (solid).
spectral region into 20 sections of about 1.5 nm
extent. The photoelectric scanning of each section
was continuous and took about 30 min. We obtained data points every 0.0018 nm at a resolution
of 0.0030 nm with the counting period of 4 s. At
each scan range, two other scans were obtained;
the back ground scan without any CO2 in the cell
and the scan of the emission spectrum of the
fourth positive bands of CO for wavelength calibration.
For the cross section measurements at 195 K, a
portion of the optical path (83.5 cm) could not be
cooled down, and the measured optical depth was
a combination of one at 295 K (outside of the
White cell) and the other at 195 K (inside of the
White cell). From the known previously measured
cross sections, and column density at 295 K, we
could deduce a contribution from the absorption
(optical depth) at 295 K (shown by broken curve
in Fig. 2). By subtracting optical depths at 295 K
from the observed optical depths (shown by dotted
curve), the optical depths at 195 K are available as
shown by solid curve.
3. Results and discussion
The cross section measurements at PF covered
the wavelength region 163.37–171.97 nm and were
254
W.H. Parkinson et al. / Chemical Physics 290 (2003) 251–256
Fig. 3. The absorption cross sections of CO2 at 295 and 195 K.
The cross sections at 295 K are given by the thin dotted line
(top), and those at 195 K by the dark solid line (bottom).
made without any wavelength calibration. The
cross section measurements at CfA covered the
wavelength region 170.80–197.63 nm. The wavelengths of the measurements at PF are calibrated
from the overlapped spectrum of the CfA measurements. The combined results of measurements
at PF and at CfA are presented in Fig. 3, where the
cross sections at 295 K are given by the thin dotted
line, and those at 195 K by the dark solid line. The
uncertainty of 5% in the measured cross sections
arises mostly from the statistical scattering in the
optical depth measurements. The temperature dependency of the cross sections becomes dominant
toward the longer wavelength. The cross sections
at 195 K are lower than those at 295 K, and the
band structures are more enhanced at 195 K as
Fig. 4. The absorption cross sections compared with previous measurements; for (a) 160–170 nm, (b) 170–180 nm, (c) 180–190 nm, and
(d) 190–200 nm. The darker curve represents our results at 295 K and the lighter curve at 195 K. The open circles are from Heimerl [7],
open diamond from Ogawa [8], small triangles from Shemansky [9], open squares from Lewis and Carver [11], and dotted curves from
our previous work [13].
W.H. Parkinson et al. / Chemical Physics 290 (2003) 251–256
expected with higher maxima and lower minima.
The cross section near 200 nm becomes flattened at
3 1024 cm2 , which might be the results of
Rayleigh scattering as mentioned by Shemansky
[9]. We can not extend the cross section measurements at 195 K to 200 nm, because of the limited
column density available.
As mentioned in Section 1, there are many
publications on the absorption coefficients (cross
sections). However, we compare our values with
only published tabulated values of Heimerl [7],
Ogawa [8], Shemansky [9], Lewis and Carver [11],
and Yoshino et al. [13]. Figs. 4a–d present the
cross sections at 195 and 295 K, each over a 10 nm
span compared with the other results. The cross
sections measured at 295 K (darker curve) appear
above those at 195 K (lighter curve). The values by
Heimerl [7] are presented by open circles with
in the wavelength region 166–182
resolution 3.4 A
nm. His values are larger than the present values
over all the wavelength region observed. The values of Ogawa [8] are presented by open diamonds
in the wavelength region 172–200 nm at a resolu. His values at absorption minima
tion of 0.057 A
agree well with ours down to 194 nm, and are
shifted away to higher values toward the longer
wavelength region. The higher resolution mea) by Shemansky [9] are presurements (0.038 A
sented by small open triangles. His values agree
very well with ours up to 194 nm, but the fine
structures are not present. Above 194 nm, his
values diverge strangely, do not present any
structures, and the values are shifted downward.
Lewis and Carver [11] obtained temperature
dependent cross sections of CO2 in the wavelength
, but
range of 175–197 nm with a resolution 0.05 A
with 0.5 A steps. Their tabulated values at 200 and
300 K, from Table 1 of Lewis and Carver [11], are
presented by open squares and compared with
ours at 195 and 295 K in Fig. 4. Their values at
both temperatures agree well with ours except for
steps,
the values above 188 nm. However, at 0.5 A
they failed to present the fine structures as can be
seen by comparing our Fig. 4 with their Fig. 7
through 9.
The measurements presented here are extension
toward the longer wavelength region from the our
previous measurements [13]. The overlapped area
255
Fig. 5. The absorption cross sections at 295 and 195 K in the
wavelength region 117–200 nm. The cross sections at 295 K are
given by the thin solid line (top), and those at 195 K by the dark
dotted line (bottom).
are presented in Fig. 4a, where the previous
measurements are presented by dotted curves.
The present values superpose over our previous
values in the overlapped region 164–175 nm, but
because of the higher resolution and accurate
wavelength calibration, we recommended the
present values. The combined cross sections in the
wavelength region 118–200 nm are presented in
Fig. 5 with two temperatures, 195 and 295 K. In
this scale, the temperature dependency is not seen
up to 150 nm, and the lower values of the cross
sections at 195 K are demonstrated clearly toward the longer wavelength. The cross sections
reported here are available at wavelength interval of 0.001 nm by request to kyoshino@cfa.
harvard.edu.
Acknowledgements
We thank F. Launay for providing their high
resolution spectrogram of CO2 and the referee for
some helpful suggestions. This work was supported by NASA grant NAG5-7859 to Harvard
College Observatory. A part of the measurements
were made with the approval of the Photon Factory Advisory Committee (proposal 91-149). We
thank Drs. K. Ito and T. Matsui for their help with
the Photon Factory measurements.
256
W.H. Parkinson et al. / Chemical Physics 290 (2003) 251–256
References
[1] V.A. Krasnopolsky, Icarus 101 (1993) 313.
[2] V.A. Krasnopolsky, J. Geophys. Res. 100 (1995) 3263.
[3] P.G. Wilkinson, H.L. Johnston, J. Chem. Phys. 18 (1950)
1440.
[4] E.C.Y. Inn, K. Watanabe, M. Zalikoff, J. Chem. Phys. 21
(1953) 1648.
[5] P.S. Nakata, K. Watanabe, F.M. Matsunaga, Sci. Light 14
(1965) 54.
[6] H. Okabe, Photochemistry of Small Molecules, WileyInterscience, New York, 1978.
[7] J. Heimerl, J. Geophys. Res. 75 (1970) 5574.
[8] M. Ogawa, J. Chem. Phys. 54 (1971) 2550.
[9] D.E. Shemansky, J. Chem. Phys. 56 (1972) 1582.
[10] W.B. DeMore, M. Patapoff, J. Geophys. Res. 77 (1972) 6291.
[11] B.R. Lewis, J.H. Carver, J. Quant. Spectrosc. Radiat.
Transfer 30 (1983) 297.
[12] W.F. Chan, G. Cooper, C.E. Brion, Chem. Phys. 178
(1993) 401.
[13] K. Yoshino, J.R. Esmond, Y. Sun, W.H. Parkinson, K.
Ito, T. Matsui, J. Quant. Spectrosc. Radiat. Transfer 55
(1996) 53.
[14] P.S. Julienne, D. Neumann, M. Krauss, J. Atmos. Sci. 28
(1971) 833.
[15] A.D. Anbar, M. Allen, H.A. Nair, J. Geophys. Res. 98
(1993) 10,925.
[16] R. Sadeghi, R.T. Skodje, J. Chem. Phys. 105 (1996)
7504.
[17] C. Cossart-Magos, F. Launay, J.E. Parkin, J. Molec. Phys.
75 (1992) 835.
[18] J.U. White, J. Opt. Soc. Am. 32 (1942) 285.