Fe-Ti Anorthosites

Nearly all the titanium mined in the world today comes either directly from magnetite-ilmenite ore bodies associated with anorthosite complexes or placer deposits derived from the weathering of these complexes.


Anorthosite complexes occur in a rather narrow band stretching from the San Gabriel Mountains eastward along the Great Lakes into New England. Concentration of anorthosite complexes is so great in the Adirondacks this deposit type is often termed "Adirondack Fe-Ti deposits". Also occur in a second narrow band across northern Europe. In a Pangaea reconstruction the complexes all lie in a single belt. More peculiar is the fact that all anorthosite complexes are approximately 1.3 BY. That such massive bodies of anorthosite should be emplaced during a relatively short span of the earth’s history is quite puzzling.


Ore deposits are essentially discordant within the host while varying in both shape and size. Some are tabular while others are lensoidal or dike-like. Vary in size from a few million tons to well over 100 million tons. Largest is the Lac Tio (Allard Lake) deposit 1200 meters wide and long by 100 meters thick. Typical ore grade in excess of 30% TiO2.


Anorthosite complexes typically consist of an older massive gabbro overlain by younger anorthosite. The ores lie within the anorthosite, generally at the contact with the older gabbro. Layering within both the host and ore is absent. Ores commonly contain inclusions of anorthosite indicating they are younger. Cataclastic fabrics present in the anorthosites and gabbros, but lacking in the ores are thought to indicate the ore mineralization has migrated to its present location by some tectonic process.

The host rock is generally a medium grained, anorthosite with the plagioclase andesine to labradorite in composition. With addition of pyroxene local gradations to gabbro are present. Ore minerals consist of magnetite, ilmenite, hematite, ulvospinel and rutile. Apatite is a common accessory as are nickel and copper sulfides. Ore textures vary from euhedral crystals of magnetite and ilmenite to very common exsolution intergrowths of ilmenite-hematite and magnetite-ulvospinel. Occasional magnetite-ilmenite exsolutions are present where oxidation of early-formed spinel has occurred. Some anorthosite complexes are surrounded by zones of charnokite, a peculiar pyroxene-rich anhydrous metamorphic rock. If is thought to form by dehydration of metasediments with accompanying iron addition from the magmatic fluids.

The presence of coexisting oxide phases has permitted the use of phase diagrams to fix temperatures of crystallization. Temperatures have been found to lie in the range from 600-800 ºC. These relatively low temperatures compared to those for crystallization of anorthosites (800-l000ºC) supports the paragenetic relationship discussed above. This presents a problem for theories of crystallization for these ores which favor magmatic segregation with settling of the denser oxides to the base of the anorthosite complex. For this mechanism to be viable the ores and anorthosites must be co-genetic (same temperature of formation) which they clearly are not. More recently it has been suggested the ores are the products of immiscible liquid segregation when the oxygen fugacity reaches a critical point. The oxides crystallize within the interstices of the older silicate phases. Eventually their density is sufficient to displace the silicate minerals upward causing the cataclastic fabric in the anorthosites.


The "billiard ball model" proposed above is generally accepted by most geologists. Controversy remains about the narrow age constraints for anorthosite complexes. One model suggests that the linear trend of the anorthosite belt is indicative of a rift system. However, rocks in active rifts today are rarely anorthosite so there must be additional complications to the model. The Earth's history prior to 1.5BY appears to be one of very stable cratonic land masses and little plate tectonics as we know it today. In fact, active rifting does not seem to have occurred until about 1 BY ago. Thus, the period from 1.5 to 1 BY may be one of transition. It has been suggested that perhaps the linear trend does not represent an active rift, but rather an aborted rift in the lower continental crust/upper mantle.

The most applicable model hypothesizes the generation of magmas in the upper mantle which rise to the crust/mantle boundary. There, the upward movement is arrested and differentiation occurs removing mafics and producing an Al-rich magma. The differentated magma then rises upward into the lower crust along the aforementioned aborted rift to crystallize as an anorthosite complex.

Tellnes, Norway Ilmenite Deposit


Lies beneath Lake Tellnesvann about 7km from the coast and between the cities of Egersrund and Tellnes.


Mining of ilmenite dates to the 18th century when the area was first worked for iron. The first titanium mining occurred in 1918. The Tellnes deposit, half hidden by the lake was discovered by aeromagnetic survey in 1954 and began production in 1957.


The Egersrund complex, a major area of mafic rocks dominated by anorthosite, outcrops over approximately 1000 sq. km. The complex lies at the southwest tip of the Baltic Shield and is one of the youngest units in the shield, dated at about 1.1 BY. It is surrounded by intensely deformed granite gneiss. The basal unit of the complex is anorthosite, overlain by gabbro and monzonite. A charnockite alteration zone rims the complex.

The Tellnes deposit consists of one large lensoidal body surrounded by porphyritic anorthosite (Figure). The contacts between the ore and host rock are generally sharp, however, at the flanks of the ore zone xenoliths of anorthosite are common. Banding is either weak or absent. The main

ore mineral is ilmenite containing 12% hematite in exsolution. Magnetite occurs as euhedral crystals. Apatite is also common with trace quantities of chromium spinel.


Generally thought to be the product of immiscible liquid segregation, but still remains controversial among European geologists.

Charcteristics of Fe-Ti Anorthosite Complexes

  1. Restricted to Late Proterozoic, particularly 1.2-1.7 BP.
  2. Occur in a geographically continuous belt.
  3. Consist predominantly of anorthosite (An>35) and noritic gabbro.
  4. Charnockite, an anhydrous felsic rock, is common surrounding the anorthosite complexes.
  5. Fe-Ti mineralization in the form of ilmenite, hematite, magnetite and rutile occurs at the gabbro/anorthosite contact.
  6. Genesis not well understood, but thought to be related to buoyant blockage of magma at the crust-mantle boundary and subsequent partial melting and fractionation.