Dear Attila, and others interested,

>I think I am not alone who would be interested in your experiences
>with the induction generator. Have you observed correlation
>between the appropriate frequency and the mineral studied? What
>is the frequency that you use for biotite and amphibole?
>
As I said in my earlier postings, it was very difficult for me to get clear
answers to the problem of plasma formation, and much of the comments I can
make would require proper confirmation.

Having said that, however, there it goes my suspicions:

Plasma formation requires a low pressure gas and some sort of energy to get
it "ignited" (excited), so the problem has nothing to do with biotites or
amphiboles or any other rock or mineral, but with the "gases" (be it H2,
CO2, H2O, ... or any other) evolved when heating them. Each of these gases
has a different minimum activation threshold, ... etc. I will not go
through this avenue any further, because I must confess my absolute
ignorance ....

However, if you provide the adequate energy, you will ignite your gas, and
isotope fractionation does occur.

People with better knowledge can illustrate us all.  Simon ??? ((I'm trying
to get the paper you mentioned, but that year is not in our library. It
will take me a few days. I'm intrigued to see the solution you found))

What I observed was that low-Fe minerals and whole rocks posed no problems
in our old 2 MHz induction heater. I assume that these minerals release
water; this water is collected, and analysis proceeds as usual without
problem.

However, Biotites and certain chlorites, as well as some amphiboles would
now and again, and again and again produce a plasma. Why? I tryed to
include a piece of stainless steel inside the reaction vessel, and that
eased the problem a little bit. The larger the steel mass, the more it
improved, although plasma always persisted. On the other hand it took
longer to completely dehydrate the sample. Why?

Well, a possible expalanation is that these minerals have Fe, and at high
temperature, Fe is a powerful reductant. My hypothesis is that Fe reduced
the water released immediately in the reaction vessel, producing H2 still
inside the RF field created inside the work coil. It was this H2 that had a
resonance frequency close to the frequency of the induction heater (I
insist, in the range of around 2 MHz), and it was hydrogen that produced
the plasma. Since hydrogen does not freeze at LN temperature, adding
further cool traps, as I did, was useless. The Toepler pump, however
efficient it might be, was by all means slow, so the hydrogen stayed within
the RF field for a fair amount of time, and got fractionated.

The effect of adding the steel piece was just to change the inductance,
such that less energy would be delivered to the crucible. Less energy meant
that water was being released more slowly, and therefore, reduction to H2
was also proceeding more slowly.

I asked the physicochemists in the University about this, and they were
unable to direct me to anywhere with the resonance frequency of Hydrogen
tabulated, so I cannot confirm the above. However, it seems to me a
reasonable explanation. After all, the very same principle works in the
fluorescent tubes lighting my office right now; only with a different low
pressure gas.

On top of that, plasma formation vanished after we changed to a lower
frequency (variable, 50 to 200 kHz) induction heater. Plasma formation has
not been observed even once in two years since we changed the heater.
However, to impress my students, I sometimes use our old heater -with the
very same samples we are now routinely analysing successfully- to ignite a
plasma ... and it has never let me down as yet!

If somebody out there can offer me a fully scientific explanation, I would
very much appreciate it. Although in practical terms I have solved my
problem, I feel uneasy about not being able to get a fully satisfactory
explanation.

Cheers,

Clemente


-----------------------------------------------
Dr. Clemente Recio
Stable Isotope Laboratory
Fac. de Ciencias
Univ. de Salamanca
E 37008 - Salamanca
      Spain

Phone: (+34) 23 29 44 00 Ext. 1540
Fax:   (+34) 23 29 45 14
E-mail: [log in to unmask]
-----------------------------------------------