Schematic representation of processes associated with the formation - PowerPoint PPT Presentation

Schematic representation of processes associated with the formation of layered mafic complexes Venting of fractionated Assimilation of magma to surface volcano country rock Country rock Heat loss Density plume “Fountain” Convection Density currents Compaction and crystallization of Layered sequence accumulated along floor intercumulus melt Chilled margin Multiple influxes of primary magma Country rock with from deeper source (recharge) thermal aureole

The Skaergård Intrusion in E. Greenland This intensively studied body of mafic igneous rock is a 34 95 classic example of extreme closed- system fractional crystallization, e.g., 68 olivine ranges from 67 Fo 70 at bottom of LZ to Fo 0 at the top of 18 UZ 53 27 The numbers on map refer to the location of 4 102 75 samples, thin sections 83 of which we are 89 studying in the lab. MBS: Marginal Border Series After Stewart and DePaolo (1990) Cont. Min. Pet., 104, 125

Cross section through the Skaergaard Intrusion looking down dip (after Hoover (1978) Carn. Inst. Wash Yb., 77, 732-739

Stratigraphy, major subdivisions, phase layering and cryptic variations in Skaergaard Intrusion Note: (1) the gap in appearance of olivine as a cumulus mineral (no olivine in MZ), (2) restricted range of pigeonite, (3) extreme Fe-enrichment in olivine and cpx, (4) the widespread occurrence of Fe-Ti oxides, (5) the occurrence of quartz and K- feldspar in granophyres in UZ, (6) the α , β , and γ zones in the UBS correspond (in a rough geochemical sense) to LZ, MZ and UZ, resp., (7) the Sandwich Horizon is actually a thin layer that contains the “dregs” of the fractional crystallization process. After Wager and Brown (1968) Layered Igneous Rocks, and Naslund (1983) J. Petrol., 25, 185-212.

Phase equilibria in MgO-FeO-SiO 2 system at P = 1 atm (Bowen and Schairer, 1936) (The experiments were done under strongly reducing conditions in eq m with metallic Fe) This phase diagram shows the change in Fe/(Fe+Mg) in olivine and Olivine compositions pyroxene with progressive fractional Pyroxene compositions crystallization as liquid moves from a → b → c → d → e → 1178. Ol 1 in eq m with liq a Ol 2 in eq m with liq b (and px 6) Ol 3 in eq m with liq d (and px 8 + sil) Ol 4 in eq m with liq e (and sil) Ol 5 in eq m with liq 1178 (and sil) Px 6 in eq m with liq b (and ol 2) Px 7 in eq m with liq c (and sil) Px 8 in eq m with liq d (and ol 3+ sil) 7 8 6 e Although this is a simple ternary system, the phase relations provide a realistic view of changing Fe# during 5 Olivine gap 3 4 fractional crystallization 2 1 d : isobaric invariant point. Reaction at d : En + L → Fo + Silica

The differentiation of the Skaergaard Intrusion 1939: Publication of Wager and Deer’s “Skaergaard memoir” that apparently presented incontrovertible evidence for extreme Fe enrichment as a result of crystal fractionation in a closed system. This fractionation trend was predicted by Fenner but disputed by Bowen. 1968: Publication of book by Wager and Brown “Layered Igneous Rocks” in which they calculated the compositions of successive liquids using the composition of the chilled margin (EG 4507) as the initial magma and subtracting the computed average bulk compositions of each of the zones in the layered series. They also assumed that the Sandwich Horizon represents a liquid composition and that the granophyres were the last liquid products of fractionation separated by filter pressing. They computed that the Hidden zone formed ~70% of the intrusion. The maximum FeO+Fe 2 O 3 content of ~22% was reached by the liquid at top of UZb. SiO 2 content of successive liquids remained below 50% through ~95% of the crystallization. Liquid comps Crystal extracts A=Na 2 O+K 2 O F=FeO+0.9Fe 2 O 3 M=MgO SH Sandwich Horizon MG Melanogranophyre SG Sydtoppen granophyre TS Tinden Sill Data from Wager and Brown (1968) Data from Wager and Brown (1968) SiO 2 A

The differentiation of the Skaergaard Intrusion (cont.) 1987: Hunter and Sparks (CMP, 95, 451) questioned Wager’s calculations (HERESY!!) They noted that such extreme Fe enrichment is not observed in volcanic basalt systems so they set about redoing the calculations. First let’s look at the volcanic data, e.g., Iceland FeO vs SiO 2 data for Icelandic tholeiitic systems. Composition of the Iceland flows do not exceed ~15% FeO. The calculated cumulate extracts are those required to model fractionation intervals (ferrobasalt to icelandite to dacite to rhyolite). AFM diagram for Icelandic basalt flows Compositional fields of FeO-rich basalts from C which show less extreme Fe enrichment (flood basalts) and O (oceanic basalts). and substantial amounts of rhyolite.

The differentiation of the Skaergaard Intrusion (cont.) Hunter and Sparks (1987) re-calculated Skaergaard magma compositions FeO vs SiO 2 :Hunter and Sparks’s recomputed liquid compositions were much less Fe-rich (16 wt. % to 17.5 wt. % FeO). Hunter and Sparks used a new chilled margin composition (KT39) and recalculated AFM plot . The numbers 70-95 indicate the the bulk compositions of the major zones. They % crystallization. Hunter and Sparks model claimed that including the Fe-Ti oxides was critical. produced significant amount of dacite and Their fractionated liquid trend was similar to that rhyolite liquids—not obvious in field. observed in volcanic sequences, e.g., Loch Ba

Summary of AFM digrams Data from Wager and Brown (1968) In the middle panel, the calculated liquid-line-of-descent curves for Skaergaard liquids are based on two models. The upper curve is for 50% SiO 2 and the lower curve for 52% SiO 2 ferrobasalt parent magma compositions.

Summary of FeO versus SiO 2 diagrams Figures reproduced from: Hunter, RH and Sparks RSJ (1987) The differentiation of the Skaergaard Intrusion. Contrib. Mineral. Petrol., 95, 451-461 In the lower panel, the calculated liquid-line-of- descent curves for Skaergaard liquids are based on two models. The upper curve is for 50% SiO 2 and the lower curve for 52% SiO 2 ferrobasalt parent magma compositions. Liquid compositions are those at the top of the zones indicated. LZi is initial LZ liquid calculated by adding the LZ cumulate average to the parental ferrobasalts

Neil Irvine’s Skaergaard Model Irvine et al. (1998) Geol. Soc.Am. Bull., 110, 1398-1447 .