Petrology of the Mantle by 7023O8

VIEWS: 14 PAGES: 24

									                                   Petrology of the Mantle

Most evidence about the composition of the Earth’s mantle is indirect from composition of
basalts and seismic waves. The only direct evidence for the petrology and composition of the
very top of the mantle is from xenoliths and the bottom of ophiolite sequences.

xenoliths in basalts and kimberlites come from the lithosphere or asthenosphere: contain
garnet, plagioclase, or spinel peridotite. Garnet, plagioclase, and spinel are the dominant Al-
containing minerals in the mantle. Each is stable at different pressure regions (see attached
diagram).




Mantle Nodules in Volcanic Neck, Sierra Negro, N.M. + Diamond-bearing Kimberlite,
N.W.T., Canada
Ophiolites – ocean crust obducted onto Continents during Continent-Continent Collision (e.g.
Newfoundland)




Seismic data
lower crust:             Vp = 6.6-7.0 km/s g/cc

upper mantle lithosphere:
                       Vp = 7.8-8.4 km/s  g/cc

(the crust and the upper mantle together are referred to as the lithosphere)

asthenosphere (~60 to 150 km):
                       Vp = 7.5-7.8 km/s; 3-5% velocity decrease

                         >150 km velocity increases
The velocity data correspond to garnet peridotite with ~ 63% oliv + 35% opx + 2% cpx + 10%
garnet
The major oxides are SiO2, MgO, FeO. Minor oxides are CaO and Al2O3. Other elements are in
very minor to trace concentrations. The name given to the estimated original mantle composition
is pyrolite:
                                     SiO2     45.9
                                     TiO2     0.09
                                     Al2O3    1.57
                                     Cr2O3    0.32
                                     FeO      6.91
                                     MnO      0.11
                                     NiO      0.29
                                     MgO      43.5
                                     CaO      1.16
                                     Na2O     0.16
                                     K2O      0.12

400 km seismic discontinuity

The discontinuity is due to polymorphic transitions:

i) olivine spinel structure (tighter structure with shared faces of polyhedrons)
ii) pyroxene goes into a garnet structure: 2(Ca,Mg)2Si2O6 Ca,Mg)3(Mg,Si)2Si3O12 (majorite
garnet with Si in octahedral coordination)

600 km discontinuity
i) pyrope garnet ilmenite structure (all octahedral)
ii) garnet, opx perovskite structure with all Si on octahedral coordination. (True perovskite
   is CaTiO3.)
iii) spinel ilmenite structure + MgO below 1050 (all MgO, FeO and SiO2 are in octahedral
coordination.)
              MAJOR IGNEOUS ASSOCIATIONS OF THE OCEAN FLOOR

1) Mid-ocean basalts/gabbros (MORB’s)
2) Ocean island basalts (OIB’s)

These two magma series have differ primarily in the concentrations of minor elements, including
K, Ti, U, Th, La, and other large-ion-lithophile trace elements.


                                 MID-OCEAN RIDGE BASALTS
Characteristics:

 the major magma type is tholeiitic basalt

 slightly olivine to quartz normative (“basalt tetrahedron”)




 K2O < 0.3 wt.%, TiO2 < 2% (see table)




1, 2 – Olivine Basalt, 3,4,5,6 – Tholeiitic Basalt
 relatively low total alkalis (K2O + Na2O) for a given SiO2 or Al2O3 content
 include:
       a) “normal” N-MORB’s with <0.1% K2O, <1% TiO2, and very low concentrations of
                light rare earth elements (REE’s), such as La and Ce and other heavy elements,
                including U and Th and
       b) “enriched” E-MORB’s with >0.1% K2O, >1% TiO2 and are relatively enriched in light
                  REE’s




 N-MORB’s are dominant at all mid-ocean ridges. They are generated below mid-ocean ridges
   from the depleted upper mantle source.
 E-MORB’s are represented by plumes centered on mid-ocean ridges. Examples include Iceland
   and the Azores. They are thought to be derived from plumes generated near the 650 km
   discontinuity, from either
   a) a distinct lower mantle source, or
   b) boundary source enriched by subducted ocean crust trapped at the discontinuity
A major implication of the very low concentrations of alkali elements and most heavy, large trace
elements (large-ion lithophile) in N-MORB’s is that the upper mantle is depleted of these
elements. The extracted elements are now highly concentrated in the continental crust.
                                 Melting in the Upper Mantle


Adiabatic melting
      It is important to recognize that although the upper mantle convects, it is not molten.
       Likewise, deep mantle plumes are rising portions of hot solid mantle. Deep plumes do not
       contain melt; they ascend by “creep”.
      Adiabatic melting occurs when rising mantle does not cool along the geotherm, but
       instead it remains hot. This allows it to reach the mantle solidus that has a shallower slope
       on the P-T diagram (see diagram below).
      The melting is eutectic involving olivine, opx, and cpx. The melting continues as long as
       the eutectic condition remains. The extra heat that the rising mantle carries is used for the
       necessary “heat of fusion”. Once cpx melts-out, the melting stops (see below).




Phase equilibria

The appropriate phase diagram for explaining the composition of MORB’s involves the major
mantle phases olivine, orthopyroxene and clinopyroxene. The model diagram for this system is
the front side of the basalt tetrahedron:
The above diagram is for the Fe-free system. However, Fe is obviously very important in the
mantle. To get the Fe/Mg ratios of the minerals in the mantle, we can use appropriate binary
phase diagrams and the known Mg/(Mg+Fe) ratio, called the Mg#, of primary MORB’s which is
~0.70. For example, from the binary Mg2SiO4-Fe2SiO4 phase diagram we can figure out that
mantle olivine must be Fo90. This is also confirmed by the fact that during crystallization of
primary MORB’s, Fo90 is the composition of the first olivine. The diagram that illustrates phase
relations during melting of the real Fe-containing mantle is the following one:
                                            Ca(Mg,Fe)Si2O6

                                                      cpx




                                                      cpx+L

                                             e        1 atm
                                   o    l        +         L   p

                                            10 kbar            opx+L
                         ol                          opx                     qtz
                    (Mg,Fe)2SiO4                                             SiO 2
                                                                    ,
                                                       af ter Stolper 1983




Note that as opposed the simplified diagram above, this one is in terms of mole%, not wt.%.
The above diagrams shows two important things:

1) that the 1 atm. peritectic point becomes a eutectic with increase in pressure as the cpx-opx join
becomes a thermal high. This involves something called the Alkemade line and the Alkemade
theorem:

Alkemade line joins composition of two primary phases whose primary crystallization fields are
adjacent to a boundary curve separating the two phases.
In the diagram below on the left, the opx-cpx join is not an Alkemade line, because it crosses the
fo-di boundary, not the en-di boundary. In the diagram on the right, the join crosses the opx-cpx
boundary, thus the join is an Alkemade line.

                          Ca(Mg,Fe)Si2O6

                                 cpx           1 atm




                                     cpx+L


                   ol+L                    p

                                           opx+L
         ol                    opx                     qtz
    (Mg,Fe)2SiO4                                       SiO2


Alkemade theorem states that “Direction of falling temperature on the boundary curve is away
from the intersection of the boundary curve and its Alkemade line”. The implication is that what
was before a peritectic now becomes a eutectic point and the Alkemade line is a thermal ridge on
the diagram.

2) At high pressure, the composition of melts generated at the ternary eutectic has higher
        (Mg+Fe)/Si that for melts generated at shallow depths. Thus, melts generated at greater
        depth will tend to be olivine normative, whereas those generated at shallower depths will
        tend to be quartz normative.

The Ocean Crust

New ocean floor forms at the spreading ridge. The crust is generated from crystallization
products of new basaltic magmas/lavas.
The crust is 5 -10 km thick.

We know what the crust looks like from its pieces that were obducted onto the continents. Such
pieces of the crust are called ophiolites

The ocean crust consist of

   a)   cumulate products of crystallization within below-ridge mafic magma chambers
   b)   differentiated gabbroic melts
   c)   sheeted dikes
   d)   pillow basalts (extrusive)
   e)   sediments

Sea-water circulation through the hot rocks is a major geologic activity at mid-ocean ridges. It
results in:

   a) black smokers (massive sulfide deposits)




   b) ecosystems that rely of sulfur redox-reactions for their energy




   c) diagenetic alteration of the mafic minerals to serpentine and chlorite.
                                 OCEAN ISLAND BASALTS




There are two principle basalt types on ocean islands: tholeiites and alkali olivine basalts. Both
are generated by melting of a hot mantle plume that rises from the lower mantle.

A volcanic ocean island usually forms in several stages:

pre-shield stage: submarine eruption of alkali olivine basalts. This stage involves low degrees of
melting (few percent) of the plume at great depths (~90 km?)
shield-building stage: the most voluminous stage dominated by tholeiitic basalts. The basalts
probably form at a shallower depth as the rising plume continues to melt along the mantle solidus
as it rises.
post-shield stage: again alkali olivine basalts. Generation of these basalts probably represents the
last vestiges of melt as the plume ceases to actively generate melts for a brief period of time.




The difference in the compositions of tholeiites and alkali basalts is well illustrated on an alkali
vs. SiO2 diagram. The alkali basalts have substantially higher Na2O+K2O for a given
concentration of SiO2.
Fractionation Trends of OIB’s and Their Implications
The primary alkali olivine basalts and tholeiites on Hawaii fall in different parts of the Ol-Cpx-
Opx-Plag basalt tetrahedron. This is better shown by “projection” of the fractionation trends onto
the base of the tetrahedron. The two types of primary basalts have differentiation trends that go in
opposite directions, the alk. basalts toward nepheline (become more silica undersaturated) and
the tholeiites toward quartz (become silica saturated). What is the cause of this behavior?
Position of the Primary Melts
Their bulk composition of the primary melts reflects the movement of the ternary
peritectic/eutectic during melting of the cpx, opx, and olivine in the mantle with increase in
pressure:

                                                       Ca(Mg,Fe)Si2O6

                                                                 cpx




                                             primary tholeiites
                                         e                       1 atm
               primary alk. ol basalts             e      e
                                                                         p

                                     20 kbar
                                               15 kbar 10 kbar
                                ol                            opx                       qtz
                           (Mg,Fe)2SiO4                                                 SiO2
                                                                  after Stolper, 1983

Thus, the position of the primary alkali basalts implies that they are generated at greater depths
than the tholeiites.

Crystallization trends

The Ol-Cpx-SiO2 phase diagram at 1 atm has a thermal high on the olivine-cpx cotectic. The
thermal high is caused by the solid solution of cpx deep into the diagram. Because the primary
alkali olivine basalts and tholeiites fall on either side of the alkemade line when they get erupted
onto the surface, they must fractionally crystallize in opposite directions away from the thermal
high.
Origin of OIB’s

        The enrichment of various minor and large-ion-lithophile elements in OIB’s suggests that
their parent plumes originate in the lower mantle. There is good evidence that their source may
include portions of subducted slabs that have descended to the mantle-core boundary.

        Much of the heat necessary for generation of plumes may come from the outer core,
which is liquid and vigorously convecting. Decay of radioactive isotopes (40K, 238U, 235U, 232Th),
which occur in the source region at relatively high concentrations, also contributes to generation
of the plume.

        Melting of the rising plume probably involves adiabatic melting where the alkali olivine
basalts are generated by low degrees of partial melting as the plume material at ~60 km below the
ocean crust, whereas the tholeiites represent higher degrees of melting of the plume material as it
approaches the crust.
                           CONTINENTAL MAFIC PROVINCES

Mafic magmas in the continental crust are derived from the mantle. They may be analogues to
both OIB’s or MORB’s.

1) Plateau flood basalts (e.g., Columbia River basalts, 14-17 m.y.; Parana Basin, Mesozoic).

       These originate from plumes, in a manner similar to OIB’s
2) Cinder-cone/lava flow fields (e.g., Keewenawan Rift, Proterozoic; Rio Grande Rift, Tertiary
       and Quaternary; East African Rift, Tertiary).

       These provinces are also associated with plumes but occur in continental rifts. These
       magmas are often very alkalic and include carbonatites. The source regions of these
       magmas may get enriched in carbonate components by “mantle metasomatism” in the
       sub-continental mantle lithosphere.




3) Kimberlites (S. Africa; Siberia; Canada; Australia; minor in U.S.)

       They are generally found in non-orogenic parts of continental platforms. They are
       essentially serpetinized, porphyritic, phlogopite peridotites and occur as vertical <0.5
       Km2 pipes scattered over a given large area. They originate at >100 km in the mantle and
       often contain diamonds. They have affinities to carbonatites magmas and may have
       similar origin involving metasomatism of the sub continental lithosphere.




4) Gabbro intrusions (Skaergaard, Greenland; Stillwater, Montana; Muskox, NW Territories,
      Canada; Dufek, Antarctica; Kiglapait, Labrador).

       These intrusions have N-MORB-type compositions. Some may be related to initial rifting
       of continents during creation of the ocean floor.
5) Massive anorthosite provinces (Adirondacks, N.Y through Quebec to Labrador; Laramie
      Range, Wyoming)

      These are probably mega accumulations of plagioclase by floatation in large magma
      chambers near the Moho. Anorthosites are derived from tholeiitic magmas that have
      similarities to N-MORB’s.

								
To top