HIGH-RESOLUTION SPECTRA AND PHOTOABSORPTION COEFFICIENTS FOR CARBON MONOXIDE ABSORPTION BANDS BETWEEN 94.0 nm AND 100.4 nm

HAL Id: jpa-00227424

https://hal.archives-ouvertes.fr/jpa-00227424

Submitted on 1 Jan 1988

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

HIGH-RESOLUTION SPECTRA AND PHOTOABSORPTION COEFFICIENTS FOR CARBON MONOXIDE ABSORPTION BANDS

BETWEEN 94.0 nm AND 100.4 nm

K. Yoshino, G. Stark, P. Smith, W. Parkinson, K. Ito

To cite this version:

K. Yoshino, G. Stark, P. Smith, W. Parkinson, K. Ito. HIGH-RESOLUTION SPECTRA AND PHOTOABSORPTION COEFFICIENTS FOR CARBON MONOXIDE ABSORPTION BANDS BE- TWEEN 94.0 nm AND 100.4 nm. Journal de Physique Colloques, 1988, 49 (C1), pp.C1-37-C1-40.

�10.1051/jphyscol:1988104�. �jpa-00227424�

JOURNAL DE PHYSIQUE

Colloque Cl, Supplement au n03, Tome 49, Mars 1988

HIGH-RESOLUTION SPECTRA AND PHOTOABSORPTION COEFFICIENTS FOR CARBON MONOXIDE ABSORPTION BANDS BETWEEN 94.0 nm AND 100.4 nm

K. YOSHINO, G. STARK, P.L. SMITH, W.H. PARKINSON and K. ITO*

Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, M A 02138, U.S.A.

'photon Factory, National Laboratory for High Energy Physics, Oho-Machi, ~sukuba-gun, Ibaraki 305, Japan

Abstract

Theoretical descriptions of the abundance and excitation of CO in interstellar clouds require accurate and complete information on many molecular properties. In particular, data such as line positions, line widths, and photoabsorption coefficients, which will allow the photodissociation rate for CO to be determined, are needed.

We have measured some photoabsorption coefficients for CO at a resolving power more than 20 times greater that that used in previous work. Most of our results for bands between 94.0 and 100.4 nm are larger than the values of Letzelter et al.

1. Introduction

Observations and studies of interstellar carbon monoxide (CO) provide important insights into the formation and properties of interstellar clouds [ l ] . CO is believed to be the most abundant molecule after hydrogen, and, because it is much more readily observed than H2, is used as a tracer of the large-scale distribution of molecular gas in galaxies. This application of CO observations requires that the abundance and excitation, and, therefore, the physical and chemical properties, of CO be known. An indication of the uncertainties in the spectral data for CO until now is provided by Table 2 of [l], where oscillator strengths of a number of bands are presented; they differ, in some cases, by factors of 5 .

An important process for which more data are needed is photodissociation [l].

Destruction of CO by this means occurs through line absorptions, at wavelengths between 91.2 nm and 111.8 nm, into predissociating states. In diffuse clouds, the rate of photodissocia- tion of CO is reduced because of shielding by atomic and molecular hydrogen lines that are accidentally coincident with CO absorption features. On the other hand, in dense clouds, dissociation of CO is enhanced by emission from Hp that has been excited by cosmic-ray- generated secondary electrons [2). Quantitative understanding of these effects requires high resolution spectral data so that the molecular states that contribute can be identified and their line positions, line widths, and absorption coefficients, or band oscillator strengths, can be determined.

2. Experimental Method

2.1 Measurements Using Synchrotron Radiation Continuum

The main set of measurements was made with the 6.65-m vertical-dispersion spectro- graph/spectrometer at the Photon Factory (an electron storage ring) of the Japanese National Laboratory for High Energy Physics (KEK) 131. The fifth order of a 1200 line mm-' grating and 10-pm entrance and exit slits were used. The theoretical resolution was less than 3x10-* nm (3 mA), but we believe that the actual value was about 7 ~ 1 0 - ~ nm (7 mA).

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1988104

C 1-38 JOURNAL DE PHYSIQUE

The spectrometer tank contained the CO and was used as the absorption cell. The nominal absorption path length was 12.5 m, with exact values depending upon the grating settings used for the measurements. Typical tank pressures were 5 ~ 1 0 - ~ to 5 ~ 1 0 - ~ Torr (0.007 to 0.07 Pa) so that the nominal column densities were 2x1016 to 2x1017 cm2. Absorption spectra at two values of column density were obtained for most spectral regions. Laboratory temperatures were 23+1 C. The low tank pressures were difficult to measure directly and accurately. As a consequence, the column density of the absorbing gas was selected by filling a small volume, which was a measured fraction of the spectrometer tank volume, to a pressure that could be measured with a capacitance manometer and, then, expanding the test sample into the main tank. We estimate that the the initial column densities could be determined with an uncertainty, at the 20 level of confidence, of f 10 percent. However, the gas sample in the tank changed continuously in density and composition because of two effects: leakage from the tank through the spectrometer slit and outgassing from the tank walls (and/or leaks) into the spectrometer. For most measurements, the latter rate was much smaller. Therefore, we added CO through a needle valve to keep the tank pressure, and, thus, the CO column density, constant. There was no gauge that could directly measure the tank pressure during the spectrum scans, but an ionization gauge in the volume outside of the slit was used to monitor the relative value of the pressure in the tank. The readings on this gauge were shown to be proportional to the tank pressure. We estimate that, during the spectrum scans, the column densities could have deviated by as much as f 15 percent from their initial values, and have increased the uncertainties in our results accordingly.

Most spectral regions were scanned in 0.15 nm portions. The scans were begun with no gas in the main tank. After measuring the background continuum level, I,, over 0.01 nm, the gas was injected. After 0.13 nm of the spectrum, I, had been scanned, the tank was evacuated and a final short portion of the continuum was studied. The purpose of the short portions of background continuum is discussed in 93. The step size was 0.00025-nm (2.5 mA). The integration time at each step was 3 S; the dead time between steps was about 2 s. Signal count rates were about 300 to 600 S-l; the dark count rates were less that 2 ^S-'. Two fixed photodiodes, located at either end of the 25-cm long exit plane, monitored the time dependence of the background continuum level. In most cases, it decreased bjr less than 5 percent during the measurements. Because of fluctuations in the position of the orbit of the electrons in the storage ring, the 'scatter' in the signal and monitor counts was about 1.5 times larger than expected from Poisson statistics.

2.2 Additional Measurements

In an effort to detect systematic errors that might have affected the data obtained at KEK, additional measurements were made using the 6.65-m spectrometer at the Center for Astrophysics (CfA) [4]. The second order of a 2400 line mm-' grating and 10-pm slits were used. Our experience has shown that a resolution of 7 ~ 1 0 - ~ nm (7 mA) can be obtained with this apparatus. A condensed discharge in helium was used to produce a background continuum.

The signal-to-noise ratio was poor in the region of all the bands studied except S(0). At the peak of this band, the measurements at CfA gave a / n = (9.3f 1.5)xlO-l6 cm2, where a is the absorption coefficient and n is the absorber density in cm-3. The measurements at KEK gave a / n = (9.4&1.8)~10-'~ cm2.

3. Analysis of the Data

The analysis followed that of Yoshino et al. [S]. We considered the signal, M, from the monitor photodiodes and the absorption spectrum signal, L The decrease in M during the scans was linear for all runs analyzed and was consistent with the difference between the values of the short scanned portions of background continuum. Therefore, a linear interpolation was used to approximate I,. I, and I were comprised, in part, of scattered light, so that their relationships to the actual values of background, i,, and signal, I, were:

i,, + si, = I, and i + si,, = I,

1 I + where is where si, is the scattered signal. Thus a: is given by: a:/n = -

e

the length of the absorbing column. s was determined by considering absorption data obtained at two values of absorber column density for the J(O), W(O), L(O), and L f ( 0 ) bands: s was varied until, for both measurements on a band, the integrated absorption coefficient was the same.

The four measurements of s gave values that clustered around s = 0.22. 0.30 was a definite upper limit to S; 0.15 was estimated to be a lower limit at the 20 level of confidence. The uncertainty in the absorption coefficient data that resulted from the uncertainty in s was 1 8 percent. When combined in quadrature with the 10 and 15 percent uncertainties in the column density determination, we obtain an uncertainty of 1 2 0 percent, at the 20 level of confidence, for most of our results.

A small portion of the data obtained are presented in Figure 1. Because a / n is a temperature-dependent quantity, we have chosen not to call it a 'cross section'. Our high resolving power is evident when one compares Figure 1 to the first figure of Letzelter e t al. 161, whose stated resolving power is 20 times less than ours; the bands in our figure correspond to those numbered 25 and 26 in theirs. The apparent band 'head' at 96.87 nm is actually a superposition of the P-branch of the L(0) band and the R-branch of the L f (0) band. Our high resolution, sufficient to resolve the rotational lines, is required if separate oscillator strengths for each band are to be determined from the absorption coefficients.

wavelength (nm)

Figure l . Absorption Coefficient for the L(0) and LI(0) Bands of CO 4. Results

Our absor~tion coefficient data are s resented in Table 1. In order to facilitate comparisons, we have integrated our data over the same wavelength ranges as used in [6]. Our integrated values of a / n are greater than the values of Letzelter e t al. [6] for all bands except the blend of L(O), LI(O), and K(0). The discrepancy may be attributable to the difference between the resolving powers of the spectrometers used: Because of saturation effects, measurements made at low resolution can underestimate values of the absorption coefficient.

We intend to make additional high-resolution measurements in order to obtain additional data for comparison with [6] and to reduce the uncertainties in the results presented here.

Cl-40 JOURNAL DE PHYSIQUE

Table 1. Integrated Values of g (in 10-l6 cm2) n

- - P-

wavelength range column density(a) assignment this work Letzelter et al. ratio(b)

(nm) (10'6 cm-2) (10-l6 cm2) (10-16 cm2)

0.99

+

0.10

+

2.3 1 0.46 1.2

+

+

(c) 1.3

+

2.9

+

(a) If a column density is given in parentheses, it was used only for determining the scattering.

(b) (this work)/(Letzelter et al.) (c) bands are partially overlapped (d) not resolved

Acknowledgments

This research was supported in part by NASA Grant NSG-7304 to Harvard University and by a joint JSPS-NSF Cooperative Research Program.

References

[l] van Dishoeck, E. F., and Black, J. H, The Abundance of Interstellar CO, in PHYSICAL PROCESSES IN INTERSTELLAR CLOUDS, ed. M. Scholer, NATO Advanced Study Institute 1986.

i2] Gredel, R., Lepp, S., and Dalgarno, A., The C / C O Ratio in Dense Interstellar Clouds, Astrophys. J., submitted.

[3] Ito, K., Namioka, T., Morioka, Y., Sasaki, T., Noda, H., Goto, K., Katayama, T., and Koike, M., High-Resolution VUV Spectroscopic Facility at the Photon Factory, Appl. Opt.

25, 837-847 (1986).

[4] Yoshino, K., Freeman, D. E., and Parkinson, W. H., Photoelectric Scanning (6.65-m) Spectrometer for VUV Cross-Section Measurements, Appl. Opt. 19 66-71 (1980).

[5] Yoshino, K., Freeman, D. E., Esmond, J. R., and Parkinson, W. H., High Resolutton Absorption Cross Section Measurements and Band Oscillator Strengths of the (1,O)-(12,O) Schumann-Runge Bands of 0 2 , Planet. Space Sci. 31 339-353 (1983).

[6] Letzelter, C., Eidelsberg, M., Rostas, F., Breton, J., and Thieblemont, B., Photoabsorption and Photodissociation Cross Sections of CO between 88.5 and 115 nm, Chem. Phys. 114, 273-288 (1987).