Geoscience Reference
In-Depth Information
100000
8
MORB source
80000
60000
40000
OIB source
25
Outgassing
20000
(
238
U/
3
He)
0
= 2200
0
5
4
3
2
1
0
Age (By)
Figure 11.13
Helium isotope evolution of the mid-ocean ridge and ocean island basalt mantle sources, showing
different
238
U/
3
He ratios. Outgassing of helium upon melting produces mantle residue with
higher
238
U/
3
He ratios and therefore potentially more radiogenic helium, i.e. higher
4
He/
3
He
ratios, at a later stage. The more conventional notation of He isotopic compositions using
atmospheric-He-normalized
3
He/
4
He ratios is shown on the right-hand side.
source have never been as intensely degassed as the MORB source, and so have never
passed close to the surface. The presence of neon, whose isotopic composition resembles
solar neon in Hawaiian basalts, also supports this argument. Rare-gas isotope compositions
had long been considered a solid proof of the primitive character of the mantle source of
hot spots until overwhelming evidence that the Nd, Hf, Pb, and O isotope compositions
are inconsistent with undifferentiated mantle made this hypothesis untenable. However, it
is probable that, deep in the mantle, there are still expanses that have remained intact since
the Earth's primordial differentiation. The lower mantle therefore is to some extent an inti-
mate mixture of primordial and recycled material. Such a statement is not as blunt as it
sounds and can be justified with the residence-time theory of
Chapter 6
: a residence time
of 1 Ga in the mantle would require that a proportion of e
−
4.5
/
1.0
1% of the initial
3
He
at the origin of the planet is still buried in the mantle. A residence time of 4.5 Ga would
bring this proportion to 37%. Certain refractory rocks (dunite, harzburgite representing old
lithospheric debris) may act as reservoirs for this “primitive” gas whose isotopic signature
is transferred by early diffusion of the primordial component into it.
≈
11.3 Mantle convection
Few problems have divided the Earth science community as much as the question of mantle
dynamics. While all concur that convection does really take place, the consensus about the
