WARNING AGAINST THE USE OF FLUORIFEROCS HYDROGEN PEROXIDE IN 
IN 
By W. F. HlLLEBRAND. 
Dunnington 1 has pointed out a source of error to be guarded against 
in estimating titanium in rocks and minerals by Weller's method, due, 
as he believes, to the partial reversion, iu certain cases, of ordinary 
titanic to meta-titanic acid, which does not afford a yellow color with 
hydrogen peroxide. It remains for me to indicate another source of 
error in the possible presence of fluorine in the hydrogen peroxide. 
For two years the colorimctric method has given reasonable satis- 
faction in this laboratory, but recently a new lot of hydrogen peroxide 
was purchased of a different brand from that hitherto used, and after 
a time it was noticed that the results obtained were in some instances 
far too high, and that no two determinations agreed. 
It is known that hydrogen peroxide does not produce a yellow color 
in titanium solutions carrying hydrofluoric acid or fluorides, and more- 
over the addition of even a drop of the dilute acid to an already 
peroxidized titanium solution weakens the color. For this reason it is 
necessary to take the greatest care to insure the complete expulsion of 
all fluorine when dissolving rocks or minerals by means of hydrofluoric 
and sulphuric acids prior to the colorimetric estimation. A drop of 
hydrofluosilicic acid acts similarly, but the latter reagent can not be 
made to completely discharge the color even if added in great excess. 
This, however, was not suspected as the cause of our trouble until, 
on referring to the circular of one of the leading makers of hydrogen 
peroxide in this country, whose product has always given satisfactory 
results in titanium work, it was found that among the various acids 
enumerated as usually to be found in the commercial article, hy droll uroic 
acid appears. Talbot and Moody, in the Technology Quarterly, v. 1U3, 
mention hydrofluosilicic acid as of frequent occurrence in the peroxide 
manufactured a few years ago. On examining the suspected peroxide 
by neutralizing with fixed alkali, evaporating to dryness, and heating 
with strong sulphuric acid, fluorine was detected by the odor of the 
acid evolved and by its action on glass. 
It is therefore imperative to use only hydrogen peroxide which is free 
from fluorine in estimating titanium, for its presence may utterly vitiate 
the results, even if only 2 or 3 cubic centimeters of the peroxide are 
employed. 
1 Jour. Am. Chem. Soc, Vol. XIII, p. 210. 
56 
