368 



UNITED STATES MINERAL RESOURCES 



it the favored source, and at times the dominant 

 source, of lithium in early years of the industry, 

 but in recent decades its share of the market has 

 been small. 



Other lithium-bearing minerals in pegmatites are 

 zinnwaldite and allied micas, triphylite-lithiophilite 

 and similar phosphates, and bikitaite. None of these 

 have been commercially important sources of lith- 

 ium, nor are any likely to become important in the 

 future. 



PETROGENESIS 

 CRYSTALLIZATION OF LITHIUM-RICH PEGMATITE 



How lithium pegmatites crystallize has been 

 greatly elucidated by the experimental work of 

 Stewart (1960, 1963, and written communs.). The 

 geochemical development of lithium pegmatites has 

 the remarkable result that in deposits of spodumene, 

 petalite, or lepidolite the grade is rarely much lower 

 than 1.0 percent or much greater than 2.0 percent 

 LizO. Stewart has found that the system albite- 

 quartz-spodumene-water has a thermal minimum 

 (analogous to the well-known granite minimum) at 

 a composition characteristic of homogeneous spodu- 

 mene pegmatites and of feldspar-quartz-spodumene 

 zones of zoned pegmatites. The magma may at the 

 outset precipitate only quartz and feldspar, forming 

 the spodumene-free units observed in the outermost 

 parts of many spodumene pegmatites, but when the 

 lithium concentration of the melt reaches that of 

 the thermal minimum, a lithium mineral will begin 

 to be precipitated. Ordinarily the lithium mineral 

 will be spodumene, which crystallizes from the melt 

 at low temperatures and high pressures. Under lower 

 pressure (Pb^o less than about 2,000 bars) the 

 liquidus does not reach the low temperature of 

 spodumene crystallization, and the higher tempera- 

 ture mineral, petalite, must crystallize from the 

 melt. Not uncommonly, as the solidified pegmatite 

 becomes cooler, the petalite changes to spodumene 

 plus quartz. 



As crystallization from the melt proceeds, the 

 composition must remain feldspathic and have an 

 LisO content of no more than about 2 percent. Yet 

 inner zones of some pegmatites are exceedingly 

 quartzose, and a few inner units contain more than 

 2 percent LijO. Stewart discovered that the solids 

 dissolved in the gas coexisting with the melt are 

 highly siliceous. He thus attributes quartzose inner 

 zones and also units of abnormally high lithium con- 

 tent to precipitation from the gas. 



Hydrothermal investigations by Munoz (1971) 

 show that lepidolite is stable only at low tempera- 



tures in pegmatitic mineral assemblages and that it 

 can form by subsolidus reaction of spodumene and 

 potassic feldspar in the presence of a fluorine-bearing 

 aqueous gas. Textural and structural relations of 

 some lepidolite to spodumene and microcline at many 

 localities indicate that it does indeed form by re- 

 placement during cooling of pegmatite. Yet most 

 lepidolite is in the innermost zone of the pegmatite, 

 quite separate from any zone containing other lith- 

 ium minerals. Though such lepidolite zones have a 

 very disordered texture, with abundant veining and 

 corrosion features, they lack evidence of preexist- 

 ing lithium minerals. The core of the Bikita pegma- 

 tite has quartz as the principal associate of lepido- 

 lite and may have crystallized directly from the 

 siliceous gas described by Stewart. In other pegma^ 

 tites, however, feldspar is abundant in the lepidolite 

 core, and its origin thus becomes implausible as a 

 product either of the gas or of the silicate melt. 

 Nevertheless, if lepidolite can form at low tempera- 

 tures by subsolidus reactions as a replacement min- 

 eral in outer zones, then it should also be stable as 

 a direct precipitate of whatever residual fluid re- 

 mains at these temperatures trapped in the center 

 of the pegmatite. 



ORIGIN OF PEGMATITIC MAGMA 



How a lithium-rich magma is generated can be 

 stated in generalities but with few details. Locali- 

 ties containing pegmatites ordinarily also contain 

 granite of approximately the same age that is uni- 

 versally regarded as the source of the pegmatitic 

 fluid or in some other way closely related to it. The 

 granites that accompany pegmatites, being of low 

 calcium content and in some instances having vir- 

 tually no calcium at all, are probably formed by 

 melting of calcium-poor sediments rather than by 

 fractionation of basic magmas. If so, then only 

 where lithium is available in the zone of the melt- 

 ing — presumably from clays that have been carried 

 to great depths — can lithium pegmatites result, 

 which would account for the absence of lithium 

 minerals in many pegmatite regions. 



How granites, which rarely contain much more 

 than 100 ppm (parts per million) lithium, can gen- 

 erate pegmatitic magma of generally granitic com- 

 position but with several thousand parts per million 

 lithium is a formidable question. In the Georgia 

 Lake area, Ontario (Pye, 1965, p. 52-54), and in 

 the Black Hills of South Dakota, the granite near 

 lithium pegmatites tends to be of higher lithium 

 content than elsewhere. This might be viewed as 

 implying that the granitic magma fractionated to 

 produce a lithium-rich magma were it not that 



