1. Chapter 13: Mid-Ocean Rifts
The Mid-Ocean Ridge System
Figure After Minster et al. (1974) Geophys. J. Roy. Astr. Soc., 36,

2. Ridge Segments and Spreading Rates
Slow-spreading ridges:
< 3 cm/a
Fast-spreading ridges:
> 4 cm/a are considered
Temporal variations are also known

3. Oceanic Crust and Upper Mantle Structure
4 layers distinguished via seismic velocities
Deep Sea Drilling Program
Dredging of fracture zone scarps
Ophiolites
4 layers of the oceanic crust and mantle distinguished on the basis of discontinuities in seismic velocities
Deep Sea Drilling Program rarely penetrates the volcanics, and then only to a maximum depth of 1500 m
Dredging of fracture zone scarps  samples from deeper sources, but no reliable stratigraphic control
Ophiolites = masses of oceanic crust and upper mantle thrust onto the edge of a continent or incorporated in mountain belts, now exposed by erosion

4. Oceanic Crust and Upper Mantle Structure
Typical Ophiolite
Figure Lithology and thickness of a typical ophiolite sequence, based on the Samial Ophiolite in Oman. After Boudier and Nicolas (1985) Earth Planet. Sci. Lett., 76,

5. Oceanic Crust and Upper Mantle Structure
Layer 1 A thin layer of pelagic sediment
Absent on newly generated crust at ridge axes, and thickens away from it
Figure Modified after Brown and Mussett (1993) The Inaccessible Earth: An Integrated View of Its Structure and Composition. Chapman & Hall. London.

6. Oceanic Crust and Upper Mantle Structure
Layer 2 is basaltic
Subdivided into two sub-layers
Layer 2A & B = pillow basalts
Layer 2C = vertical sheeted dikes
Layer 2 A  B with fracture in-filling by mineral deposition (some call both A, and then call 2B sheeted dikes)
Layer 2C = vertical sheeted dikes emplaced in the shallow brittle extensional environment at the ridge axis
Many dikes have only a single chill margin  later dikes split and intruded earlier ones
Figure Modified after Brown and Mussett (1993) The Inaccessible Earth: An Integrated View of Its Structure and Composition. Chapman & Hall. London.

7. Layer 3B is more layered, & may exhibit cumulate textures
Layer 3 more complex and controversial Believed to be mostly gabbros, crystallized from a shallow axial magma chamber (feeds the dikes and basalts)
Layer 3A = upper isotropic and lower, somewhat foliated (“transitional”) gabbros
Layer 3B is more layered, & may exhibit cumulate textures
The layering may be horizontal, but more commonly dips at angles locally up to 90o

8. Oceanic Crust and Upper Mantle Structure
Discontinuous diorite and tonalite (“plagiogranite”) bodies = late differentiated liquids
At the top of the gabbros in the Oman are small discontinuous diorite and tonalite (“plagiogranite”) bodies = late differentiated liquids filter pressed and mobilized along the gabbro-sheeted dike contact, and may extend up into the pillow layer
Figure Lithology and thickness of a typical ophiolite sequence, based on the Samial Ophiolite in Oman. After Boudier and Nicolas (1985) Earth Planet. Sci. Lett., 76,

9. Layer 4 = ultramafic rocks
Ophiolites: base of 3B grades into layered cumulate wehrlite & gabbro
Wehrlite intruded into layered gabbros
Below  cumulate dunite with harzburgite xenoliths
Below this is a tectonite harzburgite and dunite (unmelted residuum of the original mantle)
The boundary between layers 3 and 4 is, broadly speaking, the Moho
Upper portion of layer 4 is thought to be layered, and of cumulate origin (olivine and pyroxenes sink to the bottom of the axial magma chamber)
Below this is the original, unlayered, residual mantle material
Exactly what is the crust/mantle boundary?
1) Top of the original mantle
2) Mafic/ultramafic transition (top of the added ultramafic cumulates)
Is the mantle defined by petrogenesis or by composition?
A number of authors distinguish a seismic Moho from a petrological Moho

10. Petrography and Major Element Chemistry
A “typical” MORB is an olivine tholeiite with low K2O (< 0.2%) and low TiO2 (< 2.0%)
Only glass is certain to represent liquid compositions
MORBs are chemically distinct from basalts of other petrogenetic associations
Glass samples are very important chemically, because they represent liquid compositions, whereas the chemistry of phyric samples can be modified by crystal accumulation

11. The common crystallization sequence is: olivine ( Mg-Cr spinel), olivine + plagioclase ( Mg-Cr spinel), olivine + plagioclase + clinopyroxene
From textures and experiments on natural samples at low pressure the common crystallization sequence is:
Figure 7-2. After Bowen (1915), A. J. Sci., and Morse (1994), Basalts and Phase Diagrams. Krieger Publishers.

12. Fe-Ti oxides are restricted to the groundmass, and thus form late in the MORB sequence
(hence the early Fe-enrichment characteristic of the tholeiite trend on an ACF diagram)
Figure 8-2. AFM diagram for Crater Lake volcanics, Oregon Cascades. Data compiled by Rick Conrey (personal communication).

13. The major element chemistry of MORBs
Originally considered to be extremely uniform, interpreted as a simple petrogenesis
More extensive sampling has shown that they display a (restricted) range of compositions

14. The major element chemistry of MORBs
All analyses are of glasses, so that only liquid compositions are represented
Note the very low content of K2O and that all analyses are quartz-hypersthene normative (although olivine is common in the mode)

15. MgO and FeO Al2O3 and CaO SiO2 Na2O, K2O, TiO2, P2O5
Decrease in MgO and relative increase in FeO  early differentiation trend of tholeiites
Patterns are compatible with crystal fractionation of the observed phenocryst phases
Removal of olivine can raise the FeO/MgO ratio, and the separation of a calcic plagioclase can cause Al2O3 and CaO to decrease
SiO2 is a ~ poor fractionation index (as we’d suspected)
Na2O K2O TiO2 and P2O5 are all conserved and the concentration of each triples over FX range
This implies that the parental magma undergoes 67% fractionation in a magma chamber somewhere beneath the ridge to reduce the original mass by 1/3
Figure “Fenner-type” variation diagrams for basaltic glasses from the Afar region of the MAR. Note different ordinate scales. From Stakes et al. (1984) J. Geophys. Res., 89,

16. Conclusions about MORBs, and the processes beneath mid-ocean ridges
MORBs are not the completely uniform magmas that they were once considered to be
They show chemical trends consistent with fractional crystallization of olivine, plagioclase, and perhaps clinopyroxene
MORBs cannot be primary magmas, but are derivative magmas resulting from fractional crystallization (~ 60%)

17. Fast ridge segments (EPR) ® a broader range of compositions and a larger proportion of evolved liquids
(magmas erupted slightly off the axis of ridges are more evolved than those at the axis itself)
Magma chamber processes may be different at fast-spreading ridges than at slow ones
Fast ridge segments (EPR) display a broader range of compositions, and produce a larger proportion of evolved liquids than do slow segments
Also (not shown here) magmas erupted slightly off the axis of ridges are more evolved than those at the axis itself
Figure Histograms of over 1600 glass compositions from slow and fast mid-ocean ridges. After Sinton and Detrick (1992) J. Geophys. Res., 97,

18. For constant Mg# considerable variation is still apparent.
Figure Data from Schilling et al. (1983) Amer. J. Sci., 283,
Even when we compare for constant Mg# considerable variation is often still apparent
Fig shows the variation in K2O with Mg# for the MAR data set of Schilling et al. (1983)
MAR samples with high incompatibles come from the volcanic plateaus along the ridge, such as Jan Mayen, Iceland, and the Azores
K2O, TiO2, and P2O5 are incompatible elements: their concentration is unlikely to be affected greatly by % fractional crystallization (F = ) after separation from the mantle
The concentrations of these elements are more likely to reflect characteristics inherited from the mantle source

19. N-MORB (normal MORB) taps the depleted upper mantle source
Incompatible-rich and incompatible-poor mantle source regions for MORB magmas
N-MORB (normal MORB) taps the depleted upper mantle source
Mg# > 65: K2O < TiO2 < 1.0
E-MORB (enriched MORB, also called P-MORB for plume) taps the (deeper) fertile mantle
Mg# > 65: K2O > TiO2 > 1.0
There must be incompatible-rich and incompatible-poor source regions for MORB magmas in the mantle beneath the ridges (related to the lower and upper mantle reservoirs?)
N-MORB (normal MORB) taps the depleted (incompatible-poor) upper mantle source
Mg# > 65: K2O < TiO2 < 1.0
E-MORB (enriched MORB, also called P-MORB for plume) taps the deeper (incompatible-richer) mantle
Mg# > 65: K2O > TiO2 > 1.0
Major elements are not the best way to make these distinctions, which must be substantiated by trace element and isotopic differences

20. Trace Element and Isotope Chemistry
REE diagram for MORBs
Figure Data from Schilling et al. (1983) Amer. J. Sci., 283,
Also see two types
N-MORB has depleted trend
E-MORB has non-depleted
trend

21. Lack of distinct break suggests three MORB types
E-MORBs (squares) enriched over N-MORBs (red triangles): regardless of Mg#
Lack of distinct break suggests three MORB types
E-MORBs La/Sm > 1.8
N-MORBs La/Sm < 0.7
T-MORBs (transitional) intermediate values
Figure Data from Schilling et al. (1983) Amer. J. Sci., 283,
Ratios La/Sm vs. Mg#  slope for many REE analyses at once
Note that E-MORBs (squares) always have a higher La/Sm ratio than N-MORBs (open triangles): enriched regardless of Mg#
The lack of any distinct break between the enriched and depleted lavas suggests three MORB types
E-MORBs have La/Sm > 1.8
N-MORBs have La/Sm < 0.7
T-MORBs (for “transitional”) have intermediate values
Because T-MORBs form a continuous spectrum between N- and E-MORBs they may be the result of simple binary mixing of the two magma types
T-MORBs do not necessarily imply a third distinct source

22. N-MORBs: 87Sr/86Sr < 0. 7035 and 143Nd/144Nd > 0
N-MORBs: 87Sr/86Sr < and 143Nd/144Nd > , ® depleted mantle source
E-MORBs extend to more enriched values ® stronger support distinct mantle reservoirs for N- type and E-type MORBs
Figure Data from Ito et al. (1987) Chemical Geology, 62, ; and LeRoex et al. (1983) J. Petrol., 24,
Figure 13-11: 143Nd/144Nd vs. 87Sr/86Sr data for MORBs
N-MORBs plot as a relatively tight cluster with 87Sr/86Sr < and 143Nd/144Nd > , both of which indicate a depleted mantle source
E-MORBs extend the MORB array to more enriched values (higher 87Sr/86Sr and lower 143Nd/144Nd), providing even stronger support for the distinct mantle reservoirs for N-type and E-type MORBs
T-MORBs (not shown) exhibit intermediate mixed values

23. MORBs have > 1 source region
Conclusions:
MORBs have > 1 source region
The mantle beneath the ocean basins is not homogeneous
N-MORBs tap an upper, depleted mantle
E-MORBs tap a deeper enriched source
T-MORBs = mixing of N- and E- magmas during ascent and/or in shallow chambers
The time required for the isotopic systems to develop suggests that these reservoirs have been distinct for a very long time

24. Experimental data: parent was multiply saturated with olivine, cpx, and opx ® P range = GPa (25-35 km)
Figure Data from Schilling et al. (1983) Amer. J. Sci., 283,
Experimental data on the most likely parental MORBs and picrites indicates that the parent was multiply saturated with olivine, cpx, and opx in the P range of 0.8 to 1.2 Gpa (25-35 km)
Spinel lherzolite field, which is compatible with the REEs
lack of HREE depletion (expected if garnet were a residual phase) in the deeper garnet-lherzolite field
Lack of a europium anomaly (expected if plagioclase were residual) in the shallow plagiocalse- lherz. field

25. Implications of shallow P range from major element data:
MORB magmas = product of partial melting of mantle lherzolite in a rising solid diapir
Melting must take place over a range of pressures
The pressure of multiple saturation represents the point at which the melt was last in equilibrium with the solid mantle phases
Trace element and isotopic characteristics of the melt reflect the equilibrium distribution of those elements between the melt and the source reservoir (deeper for E-MORB)
The major element (and hence mineralogical) character is controlled by the equilibrium maintained between the melt and the residual mantle phases during its rise until the melt separates as a system with its own distinct character (shallow)
The depth of multiple saturation reflects the separation depth, which should be interpreted as the minimum depth of origin
The ultimate source can be much deeper than the km indicated by the experiments, perhaps as great as 80 km for N-MORB, and even deeper for plumes of E-MORB

26. MORB Petrogenesis Generation Separation of the plates
Upward motion of mantle material into extended zone
Decompression partial melting associated with near-adiabatic rise
N-MORB melting initiated ~ km depth in upper depleted mantle where it inherits depleted trace element and isotopic char.
Continue with: % partial melting increases to ~ 15-40% as diapirs of melting mantle rise toward the surface
% melted when reaches top depends on the source depth, temperature, and the rate of rise (and spreading)
Melting is terminated by conductive heat loss to the surface near the top of the column, perhaps aided by the consumption of clinopyroxene, which, when gone, will create a discontinuous temperature jump in melting
Figure After Zindler et al. (1984) Earth Planet. Sci. Lett., 70, and Wilson (1989) Igneous Petrogenesis, Kluwer.

27. Generation Region of melting Melt blobs separate at about 25-35 km
The region of melting is probably ~ 100 km wide, but is focused into the 3-8 km wide zone beneath the ridges
Melt blobs separate at about km where they are last in equilibrium with harzburgite residuum, and migrate to a depth of 1-2 km immediately beneath the ridge axis -> axial magma chamber
Figure After Zindler et al. (1984) Earth Planet. Sci. Lett., 70, and Wilson (1989) Igneous Petrogenesis, Kluwer.

28. Lower enriched mantle reservoir may also be drawn upward and an E-MORB plume initiated
The plume may be of independent (but geograph-ically coincidental) origin
The enriched plume undergoes decompression melting to form E- MORB
As with N-MORB, the melt will not segregate until shallower depths, where the major element and mineralogical character is determined
The E-MORB and N-MORB melt blobs may mix to varying degrees as funnel to the ridge (T-MORB)
Figure After Zindler et al. (1984) Earth Planet. Sci. Lett., 70, and Wilson (1989) Igneous Petrogenesis, Kluwer.

29. The Axial Magma Chamber
Original Model
Semi-permanent
Fractional crystallization ® derivative MORB magmas
Periodic reinjection of fresh, primitive MORB from below
Dikes upward through the extending and faulting roof
Relatively large (~ 5 km wide and 9 km deep)
Figure From Byran and Moore (1977) Geol. Soc. Amer. Bull., 88,

30. Dense olivine and pyroxene crystals  ultramafic cumulates (layer 4)
Crystallization near top and along the sides  successive layers of gabbro (layer 3)
Dense olivine and pyroxene crystals  ultramafic cumulates (layer 4)
Layering in lower gabbros (layer 3B) from density currents flowing down the sloping walls and floor?
Infinite onion model, since it resembled an infinite number of onion shells created from within and added to the walls
An elegant explanation for mid-ocean ridge magmatism and the creation of the oceanic crust
The open-system periodic replenishment of primitive magma and the continuous differentiation within the chamber also explained:
The narrow chemical range with a somewhat evolved character which reflected a near steady-state balance between differentiation and replenishment
Not perfectly steady-state  the chemical variation shown in the erupted volcanics
The more primitive nature of the volcanics toward the axis of the ridge and more evolved nature toward the flanks (observed by some investigators) could be explained by the fresh injections in the axial region and more advanced differentiation toward the cooler chamber walls
All-in-all an excellent and elegant model… Too bad it’s wrong
Figure From Byran and Moore (1977) Geol. Soc. Amer. Bull., 88,

31. Recent seismic work has failed to detect any chambers of this size at ridges, thus causing a fundamental shift away from this traditional view of axial magma chambers as large, steady-state, predominantly molten bodies of extended duration
Combines the magma chamber geometry proposed by Sinton and Detrick (1992) with the broad zone of volcanic activity noted by Perfit et al. (1994)
Completely liquid body is a thin (tens to hundreds of meters thick) and narrow (< 2 km wide) sill-like lens 1-2 km beneath the seafloor
Provides reflector noticed in detailed seismic profiles shot along and across sections of the EPR
Melt surrounded by a wider mush and transition zone of low seismic velocity (transmits shear waves, but may still have a minor amount of melt)
“Magma chamber” = melt + mush zone (the liquid portion is continuous through them)
As liquid  mush the boundary moves progressively toward the liquid lens as crystallization proceeds
Lens maintained by reinjection, much like the “infinite onion”
A modern concept of the axial magma chamber beneath a fast-spreading ridge
Figure After Perfit et al. (1994) Geology, 22,

32. The crystal mush zone contains perhaps 30% melt and constitutes an excellent boundary layer for the in situ crystallization process proposed by Langmuir
Seismic velocities are still low beyond the mush (transition zone where the partially molten material grades to cooler solid gabbro)
The small sill-like liquid chamber seems difficult to reconcile with the layered gabbros and cumulates, which appear to be more compatible with a large liquid chamber
In situ crystallization in the mush zone, however may be a viable alternative for gabbro formation
Much of the layering of ophiolite gabbros may be secondary, imposed during deformation of the spreading seafloor, and not by crystal settling
Figure From Winter (2001) An Introduction to Igneous and Metamorphic Petrology. Prentice Hall

33. Melt body ® continuous reflector up to several kilometers along the ridge crest, with gaps at fracture zones, devals and OSCs
Large-scale chemical variations indicate poor mixing along axis, and/or intermittent liquid magma lenses, each fed by a source conduit
Figure After Sinton and Detrick (1992) J. Geophys. Res., 97,

34. Model for magma chamber beneath a slow-spreading ridge, such as the Mid-Atlantic Ridge
Dike-like mush zone and a smaller transition zone beneath well-developed rift valley
Most of body well below the liquidus temperature, so convection and mixing is far less likely than at fast ridges
Distance (km) 10 5 2 4 6 8
Depth (km)
Moho
Transition
zone
Mush
Gabbro
Rift Valley
With a reduced heat and magma supply, a steady-state eruptable magma lens is relinquished in favor of a dike-like mush zone and a smaller transition zone beneath the well-developed rift valley. With the bulk of the body well below the liquidus temperature, convection and mixing is far less likely than at fast ridges
Figure After Sinton and Detrick (1992) J. Geophys. Res., 97,

35. Slow ridges are generally less differentiated than fast ridges
Nisbit and Fowler (1978) suggested that numerous, small, ephemeral magma bodies occur at slow ridges (“infinite leek”)
Slow ridges are generally less differentiated than fast ridges
No continuous liquid lenses, so magmas entering the axial area are more likely to erupt directly to the surface (hence more primitive), with some mixing of mush
Distance (km) 10 5 2 4 6 8
Depth (km)
Moho
Transition
zone
Mush
Gabbro
Rift Valley
Faster ridges with more persistent liquid chambers will, on average, undergo more advanced fractional crystallization in the liquid portion
Figure After Sinton and Detrick (1992) J. Geophys. Res., 97,

36. Figures I don’t use in class
Variation in CaO/Al2O3 with Mg# for basaltic glasses from the Afar region of the MAR. The shaded region represents the range of glass analyses. Vectors show the expected path for liquid evolution resulting from fractional crystallization of the labeled phase. From Stakes et al. (1984).
Figure From Stakes et al. (1984) J. Geophys. Res., 89,

37. Figures I don’t use in class
Variation diagrams for MORBs from the Mid-Atlantic Ridge. a. Plot of CaO/Al2O3 vs. Mg# /100 showing no discernible clinopyroxene trend. b. Pearce element ratio plot with 1:1 slope for (0.5(FeO+MgO)+2CaO+3Na2O) / K vs. Si/K, suggesting olivine + plagioclase fractionation. Data from Schilling et al. (1983).
Figure Data from Schilling et al. (1983) Amer. J. Sci., 283,