Transcript Lecture 12: Surface Processes I: chemical and

Lecture 19: Noble Gas Geochemistry
• Questions
– What are the isotopic and elemental distributions of the
noble gases He, Ne, Ar, Kr, Xe, and Rn in the Earth?
– What do the noble gases tell us about the present
geodynamic structure of the Earth:
• Is the mantle layered?
• Is there a primordial component left in the mantle?
– What do the noble gases tell us about the history of the
Earth system:
• Where did the atmosphere come from?
• When did the earth form?
• When did the atmosphere form?
• Reading
– A few things in Chapter 8 of White but there is also a
good book by Ozima and Podosek, Noble Gas
Geochemistry.
1
Noble Gases
• He: two stable isotopes, 3He and 4He
• Ne: three stable isotopes, 20Ne, 21Ne, 22Ne
• Ar: three stable isotopes, 36Ar, 38Ar, 40Ar
• Kr: six stable isotopes, 78, 80, 82, 83, 84, 86,
(but no interesting anomalies)
• Xe: nine stable isotopes: 124, 126, 128, 129,
130, 131, 132, 134, 136 (and numerous
interesting anomalies)
• Rn: no stable isotopes (longest lived is 222Rn
with half-life 3.8 days)
2
Noble Gases
• Atmospheric composition:
• Elemental
–
–
–
–
–
He: 5.24 ppm = 2 x 10-5x solar
Ne: 18.18 ppm = 0.01x solar
Ar: 0.934% = 93400 ppm = 1000x solar
Kr: 1.14 ppm = 12x solar
Xe: 0.087 ppm = 6x solar
• Isotopic
– 4He/3He = 750000; 3He/4He = 1.3 x 10–6 = 1 RA
– 20Ne/22Ne = 9.8, 21Ne/22Ne = 0.03
– 40Ar/36Ar = 295.5
3
Geochemistry of He
• 3He is not manufactured by any endogenous terrestrial
process.
– Essentially all 3He in the Earth is either primordial or
brought in by cosmic dust
• 4He is radiogenic, produced by all alpha-decays
– 8 atoms of 4He per 238U, 7 for every 235U, 6 for every 232Th
• He is the only noble gas that is able to escape from the
top of the present atmosphere.
– Residence time of He in atmosphere:
• Mass of 3He = 1.3 x 10–6 * 5.24 ppm * 5.2 x 1021 g
• Flux of 3He = 1000 moles/yr = 3000 g/yr
• Residence time = 10 Ma
– This makes He geochemistry pretty easy, because
atmospheric contamination is usually negligible.
4
Geochemistry of He
• Just how bad is atmospheric contamination for other
noble gases?
5
Geochemistry of He
• Even for He, OIB samples are often contaminated, but MORB
are usually deep enough to give magmatic He
Note
systematically
low R/Ra in
HIMU OIB
localities
6
Geochemistry of He
• MORB samples are
relatively homogenous in
3He/4He, at 8±1 R , except
A
where hotspot contaminated
(a circular argument?)
• OIB samples have a wide
range, from 1 - 50 RA.
• Continental samples,
sediments, etc. have very
low 3He/4He (~0.01 RA),
since they are totally
degassed and contain only
radiogenic 4He.
7
Geochemistry of He
• What do these He isotope ratios mean? They are a
measure of time-integrated (U+Th)/3He ratio:
(
)
(
)
(
)
238
235
232
He æ 4 He ö
U l238t
U l 235t
Th l 232t
= ç 3 ÷ +8 3
e
-1 + 7 3
e
-1 + 6 3
e
-1
3
He è He ø o
He
He
He
4
• How can we change (U+Th)/3He ratio, so that with time
we will develop reservoirs with different He isotope
ratios? Two ways:
– Add or subtract U and Th, add or subtract 3He.
• Depleted upper mantle has ~2 ppb U, Th/U = 2.5, and
3He/4He = 8 R . Presumably it has been degassed by
A
cycling through mid-ocean ridges.
– Nonvolatile elements are cycled between upper mantle and
oceanic crust, but only leave the system if accreted to
continents; whereas He that reaches ridges is lost to space.
• All recycled materials are very low in 3He.
8
Geochemistry of He
• So where do OIB with 3He/4He = 30 RA come from?
• If primitive mantle is involved, it has 21 ppb U and Th/U
= 3.8; 3He/4He would be lower than depleted mantle, if
3He content were the same.
• It follows that there must be an undegassed reservoir that
retains much primordial 3He…this is the most
straightforward way to explain the low (U+Th)/3He.
– Don Anderson likes to suggest alternatives:
• Subduction of 3He from cosmic dust…not practical given diffusion of
3He in subduction zones.
• CO2 fluid inclusions in upper mantle with very low U content?
• To remain undegassed, the reservoir must never be
circulated past a mid-ocean ridge. Geochemists
therefore presume it resides in the lower mantle and that
convection is either layered or sluggish.
– Increasingly popular heretical view: He more compatible than
U during melting??
9
Heat/Helium imbalance
• A further geochemical argument for some layering in
mantle convection considers the flow of heat and of 4He
from the Earth
• 4He is produced by radioactive decay and we know quite
precisely how much heat is liberated per alpha particle for
terrestrial U and Th abundances (10-12 J/4He)
• Heat alone suggests layering: BSE U, Th and K produce
19.2 TW; crustal heat production is 5-10 TW, and a
uniformly depleted mantle produces 7 TW. Another 2-7
TW is elsewhere.
• The 4He flux from the mantle corresponds to only 2.4 TW
of heat production
• Terrestrial heat flux of 44 TW includes 5-10 from crust and
3-7 from core; mantle heat flux is 27–36 TW of which only
18–22 is primordial heat; radiogenic heat, then should be 914 TW
• This implies a boundary layer in the mantle that passes heat
(by conduction) but mostly retains 4He
10
Geochemistry of He
• So where does the 3He in the upper mantle come from?
– Cycling time of upper mantle through ridge system is relatively
short, about 500 Ma, so it should be thoroughly degassed.
– 3He is not subducted or recycled, as far as we know
– Hence 3He must be added to upper mantle from the undegassed
part of the mantle.
• Two basic mechanisms: diffusion across 670 km discontinuity, or
transport by plumes (any part of plume that does not melt or reach
surface adds its He to the upper mantle system).
• So which of the mantle zoo species is the carrier of high
3He/4He ratios?
– None of the end members DMM, HIMU, EMI, or EMII!
– DMM is degassed and the others are probably all recycled
crust of one kind or another.
– But trace element ratios show no sign of a primordial
component in OIB…is He decoupled from lithophiles?
• Probably the abundance of He in the undegassed reservoir is so high
that it can dominate He budget without showing up in other tracers
11
Geochemistry of He
•
3He/4He
is one of the reasons
people make up “internal
components” like FOZO, C, and
PREMA.
• Do mixing hyperbolas curve the
wrong way? Undegassed
reservoir should be high in [He].
12
Geochemistry of Ne
• Neon can tell a very similar story to He, about
degassing and radiogenic ingrowth
– Advantage is that Ne has three isotopes
– Disadvantage is atmospheric contamination
•
•
is primordial and abundant, an a nuclide
21Ne is manufactured in abundance by nucleogenic
reactions in the mantle:
20Ne
– 18O(a, n)21Ne and 24Mg(n, a)21Ne
– U drives both reactions: a-decay as well as neutrons
from spontaneous fission of 238U (and 244Pu in the early
days); O and Mg are the most abundant target atoms in
the mantle. Production ratio of 21Ne/4He is ~10–7.
• There may be a small nucleogenic production of
22Ne, but it is probably negligible.
13
The Ne 3-isotope diagram
• Nucleogenic production moves degassed mantle to the right;
undegassed mantle moves less because of low U/22Ne.
14
The Ne 3-isotope diagram
• Note that atmospheric composition and
solar composition lie (approximately) on a
mass fractionation line
– Any physical or chemical process,
thermodynamic or kinetic, fractionates isotopes
by mass and will change 20Ne/22Ne twice as
much as 21Ne/22Ne.
– (atm contains a little nucleogenic 21Ne)
• But if Earth started out with Solar Ne
isotope composition, why is atmosphere
heavier in isotopic composition?
– Preferential degassing of light isotopes would
make atmosphere lighter than mantle.
– Best story is hydrodynamic escape: earliest
earth atmosphere was dominated by H2, whose
escape flux under influence of early solar wind
was so big it could carry away other atoms
along with it. Light Ne was preferentially lost
from the earth, leaving an isotopically heavy
atmosphere.
15
The Ne
3-isotope
diagram
• Best MORB
data are on
the famous
gas rich
“popping
rock” sample,
2PD43, from
the Atlantic.
• Highest quality MORB data lie on a mixing line through
atmosphere, even though erupted under several km of water! The
least contaminated samples so far measured do not quite extend up
to solar 20Ne/22Ne, perhaps due to 22Ne production, perhaps due to
contamination of even the best sample.
16
The Ne
3-isotope
diagram
Quality OIB data
are extremely hard
to get for heavy
noble gases, since
subaerial eruptions
are totally degassed
and contaminated
by air.
Loihi is an
exception, because
it is submarine, and
so less degassed on
eruption.
• Highest quality OIB data also lie on a mixing line through
atmosphere, but with a clearly steeper slope than MORB. This
mantle component apparently has solar 20Ne/22Ne, but less
radiogenic 21Ne/22Ne, consistent with being less degassed.
17
Coupled He-Ne systematics
• Extrapolating each Ne
measurement along a line through
atmosphere to solar 20Ne/22Ne gives
21Ne/22Ne
extrap, which varies
coherently with He isotopes and
indicates two component mixing
along southern mid-Atlantic ridge
between MORB source and
undegassed OIB reservoir.
18
•
•
•
•
Argon Geochemistry
There are several aspects to Ar geochemistry,
because 40Ar radiogenic production is so large.
Geochronology…we have already covered the KAr and 39Ar-40Ar methods.
Consider how difficult it is to get good mantle Ar
signals, when the air is 1% Ar.
But it is not hopeless, since 40Ar/36Ar contrast
between reservoirs is so big.
• Correlation with Ne in
popping rock data, where
three-isotope systematics let
us fix upper mantle value,
shows that for upper mantle
25000 ≤ 40Ar/36Ar ≤ 45000
(atmosphere is 295.5!).
19
Argon Geochemistry
• We can also use the Ar 3isotope diagram (in the few
cases where good 38Ar data
are available) or correlation
with neon to get 40Ar/36Ar of
plume component from
Loihi data. Same story:
undegassed, less radiogenic
than MORB source.
Possibly, air has higher
36Ar/22Ne ratio than solar,
consistent with atmospheric
mass fractionation by
hydrodynamic escape.
20
Argon geochemistry
• Because we also have some idea of terrestrial 40K budget,
we can confirm the existence of an undegassed lower
mantle not from 40Ar/36Ar ratios but from total 40Ar
abundance. Remember this box model?
21
Terrestrial Xenology
• Xe has many isotopes, but they fall into three groups:
– The non-radiogenic isotopes 124Xe, 126Xe, 128Xe, 130Xe
– 129Xe, the daughter of extinct 129I (half life 17 Ma)
– The fission products 131Xe, 132Xe, 134Xe, 136Xe, produced in
slightly different relative abundances by spontaneous fission of
long-lived 238U and extinct 244Pu (half-life 82 Ma).
• Pepin’s curse: “Let every element have isotope anomalies at every mass
number…like Xenon.”
22
Terrestrial Xenology
• Fission yields:
• Why no yield of 124, 126, 128, and 130?
1024 a
1020 a
23
Terrestrial Xenology
• Fundamental observation: MORB data differ from
atmosphere, and show mixing between air and a
component with excesses of both 129Xe (from 129I decay)
and fissiogenic Xe isotopes (from Pu and/or U decay)
Continental samples
(granites, etc.) show
only fissiogenic Xe
from U decay.
CO2 well gases from
continents show mixing
between continental and
mantle components.
No OIB has ever shown
Xe isotopes different
from air -- either totally
contaminated or lower
mantle has atmospheric
composition, we do not
know which!
24
Terrestrial Xenology
• Facts proven by the MORB Xe data:
– Earth formed while 129I, half-life 17 Ma, was still alive.
– Present atmosphere formed by degassing from interior of Earth while 129I
was still alive, so that ingrowth of 129Xe in degassed residual mantle could
generate high 129Xe/130Xe mantle.
Having three clocks (I, Pu, and U), it is
possible to constrain several times:
– Atmosphere began to be retained (i.e.
accretion was complete) just after moonforming impact, 50 to 70 Ma after solar
system formation at 4.556 Ga.
– Degassing was then initially very rapid,
with 80% of the remaining Xe transferred
from upper mantle to atmosphere within the
next 20-30 Ma (but not instantaneous, or
mantle 129Xe would be much bigger).
– Degassing since has been slow, but 99%
of upper mantle Xe is now in the
atmosphere. Possibly this was onset of
layered convection and lower mantle
retains 20% of original Xe.
25
Terrestrial
Xenology
• Here is a simple
degassing and gas-loss
model: total loss of
anything degassed until
closure, total retention
of everything since.
• With suitable choice of
initial Xe composition,
129I and 244Pu
abundances, you can
make both I and Pu
clocks give same age,
90 Ma after origin of
solar system
26
Terrestrial Xenology
• It is hard to tell how much of
the fissiogenic Xe in the
MORB data comes from early
244Pu decay and how much
from continuing 238U decay
over the whole age of the
Earth.
• However, very high precision
data in 1998 showed that
~30% of the fissiogenic Xe in
MORB is from 244Pu decay
based on slightly different
production ratios of 131Xe,
132Xe, 134Xe, and 136Xe.
• Note air has only 244Pugenerated fission Xe.
27
Terrestrial Xenology
• Even better data from Hadean zircons, published 2004: clear
evidence of 244Pu in the fission Xe.
28
Terrestrial Xenology
• Unlike Ne and Ar, the mantle composition for the
nonradiogenic Xe isotopes is similar to atmospheric, not
to solar. Hence either
– (1) whatever fractionated the atmosphere in Ne and Ar was
unable to separate Xe isotopes, presumably because Xe is too
heavy to escape, or
– (2) the upper mantle Xe is mostly recycled atmosphere because
Xe is retentive enough to be subducted.
29
Xe and the Open-system
upper mantle
• Recall the argument based on Pb and Th
isotopes that there must be a leak from
lower mantle to upper mantle of Pb.
We made the same case for He.
Likewise, you can construct a model
that makes Xe isotopes work if upper
mantle is an open system.
• The open-system model is too hard to
solve without the steady-state
assumption, but even with this limit it is
much more powerful than the residual
(or He-leak only) model
Sorry: transposed
30
Attempts at Synthesis
• So we have a problem:
– We know there exists an undegassed reservoir in the Earth (from 40Ar), and
we see noble gases derived from that reservoir in OIBs (based on 3He/4He,
21Ne/22Ne
40
36
extrap, Ar/ Ar).
– But we do not see obvious evidence for a primordial reservoir in radiogenic
lithophile isotope ratios (Sr, Nd, Pb), and the trace element ratios (Nb/U,
Ce/Pb, etc.) in OIB sources are clearly not primordial.
– Yet we know of no way to differentiate a reservoir without degassing it.
• Perhaps OIBs sample only recycled material in lithophile elements, but
noble gases somehow diffuse into their sources either as they traverse the
lower mantle or across the boundary layer at 670 km.
• Perhaps it is all in the mixing ratios…recycled crust is high in
incompatible lithophile elements but very low in noble gases, so a small
lower mantle component might only be seen in effect on noble gas
isotopes.
• Perhaps it can be done with different residence times for each system.
• Perhaps noble gases are compatible at high pressure, so that early
differentiation in a deep magma ocean could alter trace element ratios
without removing noble gases.
• Or something else: core pumping, non-chondritic Earth, various
31
Andersonian ideas, etc.