of the precipitate. Also shoxm are the pH values obtained simultaneously 

 for the same ingredients not aerated and surface oxidation minimized by 

 a stream of nitrogen in a covered vessel which v;as opened only when pH 

 readings were taken. The data suggest that the particles in suspension 

 when first formed are hydroxides of some type, such as ferrous hydroxide, 

 and give rise to a pH value higher than that of the surrounding true 

 solution, as mentioned above, and slowly change to the oxide by the spon- 

 taneous loss of water to form the theoretical anhydride as represented 

 empirically in the equation: 



Fe(0H)2 -? FeO + HgO 



After the pH of the unaerated sample dropped from 6.1 to U.7 in the first 

 hour and i;5 minutes, the readings taken in the next 18 hours remained the 

 same. This vfould indicate that the above possible reaction is reversible 

 and that it gradually establishes equilibrium on standing, at pH of U.7 

 in this particular case. It has been observed that the precipitate is 

 much more readil.y oxidized when first formed. Thus it is postulated that 

 the hydroxide is oxidized rather than the corresponding oxide. There are 

 additional reasons for believins' this to be the case. It is known that 

 magnetite is moi-e easily oxidized in strongly alkaline solutions (11), 

 and the slow drop in pH below U,7 (20 to 30 hrs.) of the aerated sample 

 before reaching an equilibrium value also indicates this, 



'..'hen the pH of the aerated sample becam.e lovrer than li.6, the instrument 

 drift reversed, tending to give lower readings as particles coalesced on 

 the electrodes. This alue coincides v/ith the equilibrium value reached 

 by the samrle not e:;posed to oxidizing conditions. Possibly this is 

 related to selective adsorption involving "ferrous acid" (II Fe02 ^'^ FeOOH) 

 with very high total surface area in the early stages of precipitation. 

 However, more experiinnntal v/ork will be necessary before adequate explana- 

 tion can be made as to the unusual reversal, 



./hen ferrous sulfate :jas treated with an excess of ammonium hydroxide 

 and aerated for a period of 72 hours, magnetite was the predominant pro- 

 duct, as shown by its strong ferromagnetic property, black color, and X-ray 

 diffraction pattern, A second sa;iiplG was prepc^red s:i.milariy, but was per- 

 mitted to stand 2Lt hours, filtered to remove the sclr.ble portions and then 

 allovred to aerate for a period of 2[i hours vrhir.e suspended in distilled 

 water. Ihe oxide in this case was found to be mainly the '.'-monohydrate 

 as shown by X-ray diffraction patterns and ferromagnetic behavior of a 

 portion dehydrated at a temperature of 250°C. i: third sample was prepared 

 by adding small portions of amraonium hydroxide to ferrous sulfate and 

 oxidizing the r:recipit,ate wliile xiiider acid conditions, Tliis sample vfas 

 found to contain the > -form^ but tests shovred that it was less pure. 

 From these resu].ts it appeared that the .■ -form was obtained at a pH of 

 about 6 to 7 and that a contaminating substance was formed at a lower pH 

 value. As already described, exact control of pH for this heterogeneous 

 system is difficult but it apT-eared desirajle to m.ake further tests 

 regulating the pH, i,€,, the average value, as ^'vell as possible. This 

 was done by titrating Ot$ll, FeSOj_^ v;ith IM. NaOH at such rate as would 



K - 5 



