Dear Dr. Horstmann,

There are four possibilities to dissipate energy by the filament. These 
are : radiation, ends effect, electron emission and heat conductivity. 
More than 95 % of energy runs out by radiation, some about 4% by 
filament ends, less than 1% by electron emission and much below 0.1% 
by heat conductivity. Therefore it is not heat conductivity of gas in the 
ion source that increases the filament current. 

Electron emission current is descibed by Duschmann's formula as a 
nonlinear function of both filament temperature and work function of 
the filament material. Exponent in this formula makes the work 
function the factor having great influence on electron emission 
current. Change of 1 eV in work function produces a 4 orders of 
magnitude change in electron emission. The larger the work function 
is, the smaller the electron emission one gets. The effect of slight 
increase of work function due to surface oxidation is well known and 
may exist when gas is introduced to the source. The non-homogenic 
monoatomic layer  of oxygen disappeares when the source is empty. 

The electron current stabilizer you operate is a proportional one. It will 
increase the filament current (temperature) in order to stabilize the 
electron emission current if the work function of filament material 
increases. If only the stabilizer is OK, the whole phenomenon does not 
disturb the ion current stability, and therefore I hope that your results 
should be OK.

In case you want to see the energy dissipation problem a little closer, I 
may send you the simple (but not user-friendly) Pascal 6.0 program 
that simulates the filament operation.

With best regards,

*************************************************
Tomasz Durakiewicz
Mass Spectrometry Lab, Institute of Physics
Maria Curie-Sklodowska University
pl. M.C. Sklodowskiej 1
20-031 Lublin, POLAND
tel. (0-81) 37 61 78, fax (0-81) 33 669
*************************************************