43/number in Great Britain
Carbon R Baptist
0042-207x/92$5.00+.00 @ 1992 Pergamon Press plc
and F Levy. Dkpartement
213 to 214/l
Optronique, Division Laboratoire d’Electronique, de Technologie Centre d’Etudes Nucl&aires de Grenoble 85X, F38041 Grenoble, Cedex, France
The interaction of carbon dioxide with a freshly broken glass surface is shown by considering the important decrease of the CO2 partial pressure and the large increase in the CH, one which occur in a vacuum chamber when a glass sheet is broken.
1. Introduction In this study, we present evidence, from a vacuum composition analysis, that upon glass breaking, CO, molecules interact strongly with the newly exposed glass surfaces. Concomitant with this, a strong CH, partial pressure increase is observed. With the aim of describing this observation, we wish first to recall the history of the experiments which have led us to observe this phenomenon and then present the results themselves. Our main goal was to analyse the vacuum composition inside a flat panel cathode ray tube in order to understand its influence on the stability of the electronic emission generated by field effect emitters’.‘. For this, a display was introduced into a vacuum chamber and then broken under specific conditions. The vacuum was analysed just before and just after breaking, so that information could be gained by analysing the differences between two successive spectra (for example, a strong argon partial pressure increase could be detected and its origin connected with the fabrication process of the microcathodes). In the following experiments we successively eliminated all the complex parts of the screen in order to proceed by dichotomy and reveal the reasons for all observed changes in the mass spectra. Finally, pure glass sheets were broken and the evolution of the principal m/e ratio analysed. In all cases, the CO, peak disappeared instantaneously from the mass spectrum when breaking the glass substrate. This fact led us to investigate more precisely this phenomenon. In the following paragraphs, we describe in more detail the experimental conditions and after this we present the results.
QUADRUPOLE IR-LAMP VACUUM GAUGE
Figure 1. Schematic of experimental apparatus.
record spectra from m/e = l&100 or to monitor 12 specific masses as a function of time. Both modes are operated in an autoranging mode, necessary for the observation of large differences between two successive spectra. Two different types of glass were used: sodalime or boro silicate. Results were similar for both types. Figure 2 presents two general overviews recorded just before and after breaking.
2. Experimental and results The experimental apparatus is presented schematically in Figure 1. The vacuum chamber is pumped down with a turbomolecular pump and can be isolated from the pump by closing a valve. This chamber is equipped with a LEDAMASS quadrupole analyser (LEDAMASS, Stoke-on-Trent, UK), an ir lamp for baking (2OO”C--12 h) and an ultra-high vacuum manipulator for breaking the glass plate. After baking, the vacuum is in the low 10m9 mbar range. During the experiments (done at room temperature), the valve is closed and the vacuum gauge turned off. The vacuum deteriorates slowly and when the total pressure reaches a value of -IO-’ mbar, the glass is broken; during this whole time spectra are continuously recorded. The quadrupole allows us to
Figure 2. Quadrupole mass spectra before (a) and after (b) breaking a glass plate. 213
and F Levy:
(5) We do not observe any increase for the peaks characteristic of water vapour (rather, a small decrease). As we suppose that the reaction
Figure 3. Evolution of the principal partial pressures before and after having broken the glass plate during the tenth spectrum.
As the valve was closed for about 3 min. the value of the total pressure was around IO-’ mbar. Figure 3 presents a temporal evolution of the main m/e ratio observed in the general spectrum. These values are not corrected for the relative gas sensitivity factors, which except for hydrogen are near 1. 3. Conclusion Our main conclusions
can be summarized
as follows :
(1) Breaking glass at room temperature under moderate stutic vacuum conditions (some IO- ‘-10m6 mbar) induces a strong decrease of the mass 44 peak (COJ intensity and a concomitant increase of the 16 and 15 peaks (CH,). (2) These changes are directly connected to the size of the freshly created glass surface. Large surfaces (10 cm’) induce a decrease by a factor of l/100-3 cm’ induces a decrease by a factor of l/40. The nature of the glass does not play a prominent part in the reaction of CO? with the surface. (3) The observed CO? decrease and CH, increase occur with the quadrupole filament either operating or in the switched-off position. (4) The CO peak does not suffer any sudden change when breaking glass; we can thus suppose that carbon monoxide is not directly involved in this process.
occurs at the surface, we conclude that a strong interaction ties the water molecules to the silicate molecules of the glass surface or to defects on the surface. (6) We do not have any clear explanation for the large pulse of H2 that appears and then returns immediately to its former level. (7) Similar results are obtained after the introduction of a c IO- ’ mbar CO? partial pressure into the closed chamber having a base pressure of some lOmy mbar. (8) Using the basic kinetic theory of gases we evaluated the number of collisions per cm’ and per s at IO ’ mbar. This number (- 5 x IO’ ‘) can explain that in one second all CO2 molecules included in the vacuum chamber (volume 15,000 cm ‘) strike the fresh surface and ‘disappear’ by interaction with the H 1 molecules also colliding with the glass surface. In conclusion, we think that we have observed the possible reaction of carbon dioxide with hydrogen on a fresh glass surface and their transformation into methane and water at room temperaturc; actually, we do not know whether the glass substrate acts as a catalyst or if energy is gained by breaking the silicate bonds. This observation is not obvious because CO? is a rather inert molecule and glass is not known to be very active chemically. It is not our purpose to study more deeply the reaction described above, because this phenomenon would be far from our technological preoccupations ; however, it seemed to us important to point out the preceding observations to specialists in the surface science community, because they could have consequences in the field of catalysis, in the field of radioactive waste storing and for environmental studies. References
’ R Meyer, Recent development
on ‘Microtips’ display at Leti, Fourrh In/ Vcrc, Microelectron Con/; Nagahama (Edited by S Namba, Y Nannichi and T Utsumi), p 6 (1991). ‘A Ghis. R Meyer, P Rambaud, F Levy and T Leroux. fEEE Trufis Ekc~/ron Dw. ED-38, 2320 (1991).