To clarify the questions Pier raises regarding the H/Device,

The fundamental problem with the reactor design is not that the Cr gets
used up -- Cr is relatively cheap -- but that it cannot be removed because
the Cr2O3 is well bonded to itself.  And as Pier points out, anything that
readily dissolves the chromium oxide is likely to dissolve the silicon
dioxide as well.  One possibility is to use HCl, which reacts with the
Cr-oxide to produce chromium (III) cloride.  This is a water-soluble,
greenish-blue compound.  My thought is to set up a slow pump-through
system that will slowly dissolve and remove the oxide from the reactor
tube.  The resulting CrCl3 solution will have to be disposed of as haz.
waste (though any excess acid could be neutralized and the solution is
relatively safe (non carcinogen, non flammable; incompatibles are strong
oxidizing agents only)).

Eric Steig

On Mon, 7 Feb 2000, Pier de Groot wrote:

> Dear Francesca and others,
>
> You better do not use Br-pentafluoride (BrF5) for recovering your Cr
> metal. I understood you want to 'reactivate' (reduce) the Cr in the
> reactor without removing it first, which in theory might be a good idea.
> If you react with BrF5 the silica glass will be etched and eaten away -
> rather slowly at room temperature and quickly if you heat up a bit -
> besides the dangerous operation to carry out  and control the reaction.
> Also, I expect the Cr2O3 to turn into CrF3 or other more complex
> fluoride-compounds (also depending on any polution adsorbed on the
> Cr-Cr2O3). You rather must think about reducing reactors instead of a
> stronger oxidator than O2, such as H2 or CH4, if you want to remove the
> oxygen from the metal - although this is just a rough idea - there will
> be more (and probably better) reductors. However, I see still some other
> problems in doing so - polutions from samples adsorbed on the Cr-Cr2O3
> should be removed too, or else it will pile up and render the reactor
> useless after a few 'reactivation' turns. I am wondering if reworking of
> the Cr really pays off if considering the time and costs (salary of
> analyst doing the job included) to get it right. The silica (or quartz)
> glass reactor is another story, of course - if one can reload it with
> fresh Cr it certainly will be economic to do so.
>
> I hope 'my two pennies worth' is giving you a better idea about the
> problems you face if reactivating/reducing the Cr in the reactor.
>
> Pier.
> --
> **************************************************************
> Dr. P.A. de Groot
> University of the Witwatersrand
> Economic Geology Research Unit
> Department of Geology
> Private Bag 3
> 2050 Johannesburg
> South Africa
> Tel. +27 11 7162564
> Fax. +27 11 3391697
> E-mail <[log in to unmask]>
>
> Visit the combined HOME-PAGE of EGRU-Geology on the Internet:
>   http://www.wits.ac.za/science/geology/index.htm
> **************************************************************
>