Two processes for extracting alumina from low-grade ores are described, one applicable to both clays and high-silica bauxites, and the other only to bauxites. In the former process, a line-containing sinter is prepared and annealed by cooling it slowly through the range of 1,300 0 to 1,200 0 C. The annealed sinter dusts to a powder which requires no grinding. This material is extracted with a solution containing about 200 g of Na2C03 and 150 g of NaCI/liter, and an extract is obtained in which the concentration of Al20 3 is 70 to 80 g/liter and that of Si02, 1 to 2 gfliter. By boiling the extract with a seed charge of synthetic sodalite, 3Na20·3AI20 3·6Si02·2NaCI, its silica content is reduced to 0.1 percent of the alumina content or less. Alumina, suitable for the manufacture of aluminum by electrolytic reduction, is precipitated by passing carbon dioxide into the desilicated solution. About 95 percent of the alumina in the clay is recovered by this method.The other process developed in this investigation involves the extraction of alumina from high-silica bauxites with sodium hydroxide-sodium chloride solution and recovery of soda and alumina from the residues by a modified sodalime-sinter method. Recoveries in excess of 90 percent of the alumina in high-silica bauxites are obtained.In cyclical operation of each process, the spent solution from the aluminaprecipitation step is used in the treatment of a fresh batch of material.When the bauxites are extracted with solutions containing mixtures of sodium hydroxide and sodium carbonate, sodium sulfate, sodium bromide, or sodium nitrate, the quantities of silica present in the extracts are lower, the higher the concentration of the salt. This is caused by the formation and increased stabilization of slightly soluble compounds r elated to sodalite. Comparisons of the X-ray diffraction patterns of various sodalite type compounds are presented. By the extraction of bauxite with sodium hydroxide solution, in the absence of added salts, a relatively soluble hydrated nephelite, Na20·AI203·2Si02·2H20, is formed, which has a crystal structure different from that of sodalite.

No abstract

A techni q ue fo r p reparin g, curing, a nd all t oclav in g 1-by 1-by 10-in ch cement-lime bars was developed , a nd t he expa ns ive characteri st ics of 80 comme rcial h ydrated limes we re determined . On t he bas is of chemi cal a na lys is a nd perce ntage of lInh y drated ox ide, the h ydrated lim es were cla ss ifi ed in to fo ur seri es : h igh-calcium , reg ul a rl y h ydrated dolom itic, high ly h ydrated d o lomi~i c , a nd m agnes ian . Data o n t h e expansions of cem ent-l ime ba rs pr epared in t he p roport ions of 2 parts ceme nt to 1 part lime, 1 par t cement to 1 par t lime, and 1 part ce ment to 2 parts lim e, b.v weig ht, a nd alltoclaved to 295 poun d per sq uare in ch gage p ress ure for 3 hours, showed t hat ba rs p repared with t he regulad y h ydrated do lomit ic lim es, whi ch had the high est per centages of u nh y drated ox ide, ha d t he hi ghest perce ntages of expa nsio n. Th e hig h-calcium lim es, characterized , in ge neral, by the 10IVest percen tages of nn hydrated ox ides, gave the lowest percen tages of expans io n. Most of t he hi ghl y hydra ted do lom it ic limes ha d percentages of unh yd rated ox ide a nd expansion t hat were comparable to t hose of the h igh-calcium limes. An increase in the p roportion of lime in t he ceme nt-lim c bars was atte nded by an inc rease in expans ion. The method for d eterminin g t he linear expa nsion of cem ent-lime bars a u~oclaved at a s team -gage pre s ure of 295 Ibl in. 2 (equi valent to a tem perat ure of 21 6 0 C .) was found to be rep rodu cible, by t hree indepe nde nt operato rs. T he effect of 17 differen t por tland cem ents on the expans ion of cement-lime bars showed t hat t he expansion valu es for a par ticul ar lime t ended to in crea se as t he expa ns ion value of t he co ns t ituent cemen t in creased. Arbi t ra ri ly s ub tractincr the expansion of thc ncat cement from t he t otal expansion gave thc m ost u niform res ult fo r the " net" expansion of t h e lime. \Vith an autocla ve specia ll y m odifi ed for ascer ta ini ng t he behavior of cem ent-lim e bars d ur ing t he course of a u toc lav ing, it was fouIld t hat on ly a sligh t am ou nt of t he total expansion occurred befor e a temperat ure of 150 0 C. was reached, b ut above HiO° C. a rapid expansion rate was noted , which in t urn taper ed off before a temperat ure of 2 16 0 C. was reach ed. R etarding t he rate o f heating resul tecl in a decrease in t he total expan sion . . F in a lly, from criteria set fo rt h fo r a p roced ure for deter m ini ng t he sou nd ness of hydrated limes, a test is p roposed wit h a suggested limi t of expansion o f l.0 percent.

An investigation was made of a particular kind of plaster failure that is characterized by the formation of blisters or bulges in the finish coat of plaster. Extensive surveys revealed that the failures are widespread; that several years elapse before the bulges appear and thereafter the number and extent of failures increase as the age of the buildings increases; that the failures occur more extensively and rapidly during warm humid weather; and that there is a marked similarity in the failures regardless of the kind of plaster base coats, type of construction base, varying job conditions, and type of workmanship. The usual white coat is prepared from a lime putty gaged with plaster of paris (gypsum gaging plaster). Chemical analysis of 88 samples of white-coat plaster that had failed showed that in every instance a dolomitic lime had been used in preparing the white coat. Regularly hydrated dolomitic limes contain about 32 percent of total MgO (by weight), of which only about 5 percent is hydrated and the remaining 27 percent is still present as unhydrated MgO. This highly incomplete hydration is due to the fact that the magnesia has been badly overburned in the production of dolomitic quicklimes (which contain CaO and MgO in nearly equal molecular proportions) and thus is inactive toward hydration in the usual hydrators. Furthermore, only about 20 percent of the hydration of the magnesia is completed at the end of the customary 1-day soaking period. The amounts of CaS04.2H20, CaCOs, Ca(0H)2, Mg(0H)2, and MgO in the hardened plasters were calculated from the chemical analysis and closely checked by the method of heat of solution. The presence of CaS04.2H20, Mg(0H)2, Ca(0H)2, and CaCOa was confirmed by thermal analysis. The average amount of hydration of the MgO for 88 samples of failed white coat was 59.0 percent-much in excess of the 20 percent attained after the customary 1 day of soaking the lime as a putty. Hydration of magnesia must, therefore, occur on the wall. The hydration of MgO is attended by marked expansion. Failures were also observed in base coats where dolomitic lime containing considerable unhydrated MgO was used in their preparation. It was shown that suggested causes other than hydration of MgO with its attendant expansion cannot account for the bulges that occur in plaster several years after a building has been erected. The precautions that should be exercised in specifying a proper lime in order to prevent future failures are discussed. Publication of National Bureau of Standards Research Paper RP1022, "Hydration of Magnesia in Dolimitic Hydrated Limes and Putties" [2], referred to in the letter to "Builduag and Building Management", soon led to consideration of the actual effect (if any) of unh^'drated magnesia (MgO) in partially hydrated dolomitic limes on the performance of mortars and plasters containing such limes, and the National Bm-eau of Standards soon became interested in this problem as it relates to plaster.1 to which it is to be applied. After it is applied, but before it hard...

No abstract