The Cs2 Atlas
in the 
9700-11000 cm-1
spectral range


 
back to the welcome page
back to LAC. homepage
Direct access to the plates and tables

I. Experimental conditions

The absorption spectrum of the Cs2 molecule in the 9700-11000 cm-1 spectral range is presented in this Atlas. It is due to the transition between the (A) 1S+u and (X)1S+g electronic states. This spectral region is included in two previously published atlases [1,2].
The present work was performed for two reasons:
- the experimental conditions proposed to the potential users are easiest to implement than in the two previously cited references ;
- the hyperfine structure affecting the lines are of much fainter intensity than those in the I2 case, leading to a better wavenumber determination.

The used absorption cell is represented in Fig. 1. It is a stainless steel pipe, 480 mm long and with 63 mm external diameter. The active part is approximately the length of the heating oven i.e. 250 mm. The grid (250 mesh), made of stainless steel (type 304), is realized with a 40 mm diameter (D) thread, the distance O between two consecutive threads being 61 mm (Fig. 2). The grid is composed of three layers of this wire gauze and extends after the cooling zone. The metal, evaporating in the central part, moves through the mesh due to capillarity. The pressure of the metallic vapour is balanced by an inert gas, argon, preventing from potassium condensation on the windows [3]. The heat pipe is inserted in a White type cell, ruled for four passes (Fig. 1), and giving a one meter absorption length [4].

The Cs2 absorption spectrum was recorded under the following source conditions :

argon pressure 6 mbarr (610 P) ; temperature 550 K. The continuous background spectrum is provided by a 250 W halogen type lamp (tungsten wire), the electrical power being limited to 90 W during the experiment. The Fourier transform spectrometer used for the spectrum recording is described in [5].

II. Presentation of Plates and Table

Each page of the atlas, noted Page 001, 002,...reproduce equally spaced ranges of the spectrum. A small overlapping spectral range exists between successive plates. The vertical scale, indicating the absorption of each line profile, is given in arbitrary units but the continuous background is normalized to unity for each plate. The estimated local noise [6] is visualized at right of the vertical axis by a line portion limited by two opposed triangles. The wavenumber scale is equal to 0.2 cm-1/cm when the distance between abcissae ticks is 5 cm. Each reference line is marked by a vertical tick, located in the upper part of the spectrum. These ticks are numbered each 5 value, the corresponding ticks being longer. A spectral line is retained as a reference one if the two criteria are fullfiled :
The evolution of the background, not really continuous due to various effects, is depicted in Fig. 3a.
The Tables of wavenumbers, noted Sigma P001,... Sigma Pn, correspond to each plate Page001,...Page n. They are organized as follows :
First column : numeration number near the vertical ticks ;
Second column : vacuum wavenumber in cm-1 ;

Third column : relative absorption intensity (multiplied by 100) ;

Fourth column : width of the line at half height in 10-3 cm-1.

Additional informations about the characterization of the lines can be found in [6].

III. Calibration of the wavenumbers and accuracy
 

A complementary experiment was performed on the atomic resonance transition

6p 2P1/2---6s 2S1/2 of cesium, appearing in the considered spectral region. An accurate frequency measurement of this transition gave [7] n = 335 116 48 807 (41) kHz or s = 11178.2681607 (14) cm-1. The Fourier measurement, 11178.2710 cm-1, shows a difference lower than 0.003 cm-1. However the absorption and emission spectra were successively recorded. As a consequence systematic errors due to the interferometer, to the electronic path difference servo-control, to the source,... can perturb the measurements. These different errors were analysed and evaluated [8]. The major uncertainty is due to the wavenumber precision of the Xe radiation used to monitor the path difference in the interferometer. All these errors lead to a 0.0013 cm-1 uncertainty.

In conclusion, for the stronger lines, the uncertainty of the reported wavenumbers should not exceed 0.005 cm-1.
 
 

References

 
    [1] S. Gersternkorn, J. Vergès, J. Chevillard, Atlas du spectre d'absorption de la molécule d'iode (11000-14000 cm-1). Laboratoire Aimé Cotton, C.N.R.S. II, Bâtiment 505, Campus d'Orsay, 91405 Orsay Cedex, France.
    [2] S. Gersternkorn, P. Luc, J. Vergès, Atlas du Spectre d'absorption de la molécule d'iode (7220-11200 cm-1). Laboratoire Aimé Cotton, C.N.R.S. II, Bâtiment 505, Campus d'Orsay, 91405 Orsay Cedex, France.
    [3] C.R. Vidal , J. Cooper, J. Appl. Phys. 40, 3370-3374 (1969).
    [4] J.U.White, J. Opt. Soc. Am. 32, 285-288 (1942).

    [5] J. Connes, H. Delouis, P. Connes, G. Guelachvili, J.P. Maillard, G. Michel, Nouv. Rev. Opt. Appl. 1, 3-22 (1970).

    [6] H. Delouis, Thèse d'Etat, Université Paris XI Orsay (1973).

    [7] Th. Udem, J. Reichert, R. Holwarth, T.W. Hansch, Phys. Rev. Lett. 82, 3568-3571 (1999).

    [8] G. Guelachvili, Thèse d'Etat, Université Paris XI Orsay (1973).

 
back to the welcome page
back to LAC. homepage
Direct access to the plates and tables