All?gre Claude J. Isotope Geology

Подождите немного. Документ загружается.



volcanic rock





mantle source

The chemical Rb/Sr ratios measured on the basalts are therefore the upper bounds of the

source ratios ofthesebasalts. Letus makeafurther assumption: as these twovolcanic rocks

come from the mantle (we shall suppose that at one time they b oth belonged to the Earth’s

primitivemantlewhichwashomogeneous,identical throughout) andthen thattheir source

domains separated(orbecameindividual ized)throughgeodynamic andgeochemicalpro-

cessesofvarying complexity.

Can this sc enario be put into quantitative terms? L et us put it graphically ¢rst.We plot a

(

Sr/

Sr, T) graph and construct vectors of evolution for the sourc es of the two rocks

(Figure 6. 7).They pass through the present-day measurement points, the two isotope com-

positions.Their maximum slopes are l

Rb=



.The two vectors intersect at about1.4

billionyears. (Plotthegraphto checkthis as an exercise.)

Ifwe address th e questionusing algebra, we canwrite the evolution equations for the two

sources by noting the

Sr/

Sr isotope ratios  and the

Rb/

Sr ratios . Using a linear

approximation,weget:



ridge

¼ 

þ l

ridge



Canaries

¼ 

þ l

Canaries

Ifthe twovolcanic rocks arefro m identical mediawhich separated attimeT

, we can calcu-

lateT

andtheircompositionbypositing 

¼

and T

¼T

This thengives:



ridge

 

Canaries



ridge

 

Canaries



With l ¼1.42 10

11

1

,wegetT

¼1.40 billionyears.

But, one might object, we have used the

Rb/

Sr ratios of the volcani c rocks and not

those of the mantle. Now, as we said, these are maximum values and not real values. This

means that the1.4 billion is a minimum value.The two real vectors may well intersect at an

Sr/

1max

Canary

Islands

Mid-oceanic

ridge

Time

Figure 6.7 Calculation of a model age of differentiation of two basalts, one from the Canary Islands and

one from a mid-ocean ridge. The solid lines show the vectors of isotope evolution for (Rb/Sr) measured

on the rocks and the dashed lines are the more plausible vectors.

229 Strontium isotope geochemistry

earlier time. So the true age of di¡erentiationT

is greater than (or equal to) the age calcu-

latedwithstrontium ¼T

T

To interpret the

Sr/

S r ratios of the two basalts, one of MORB and one of OIB, we

presume the two domains of the mantle from which the basalts were extracted were

separated more th an 1.4 billion years ago and evolved independently thereafter.This is a

strong constraint as to the present-day structure of the mantle and as to the way it has

evolved historically. However, one point has to be clari¢ed. This reasoning assumes, as

said,th atwhenthe magmaformsatdepth, itis isotopicallyre-equilibratedwiththesource

domain, that is, it has the same isotopic composition as the source domain (this is the

isotopic equilibration assumption) and that subsequently, during its time at the surface,

the b asalt’s isotopic composition is not altered either by mixing with the crust (which is

why we chose two oceanicb asalts), or by interac tion withsurface £uids (the closed system

assumption), or by weathering.With a few precautions, these assumptions are gen erally

borne out.

As canbe seen, the pr inciplesbehind this model areverysi mple and yettheyleadtovery

i mp or tant results.

Remark

Until now, with dating, we were concerned about the history of rocks. In studying initial isotope

ratios, we are interested in their pre-history and in that of the chemical elements of which they are

composed.

Exercise

The

143

Nd/

144

Nd isotope ratios were measured on the same basalts as before from the mid-

oceanic ridge (0.513 50) and the Canaries (0.512 95). The

147

Sm/

144

Nd ratios are 0.25 for the

MORB and 0.20 for the OIB. Supposing the two reservoirs of these rocks ﬁrst had a common

history in the same mantle reservoir before separating and differentiating at time

, write

the evolution equations for the

147

Sm/

143

Nd system. What is the approximate age

which the two media of origin of the two basalts separated?

Given that during genesis of basalts from the mantle, the magma is more enriched in

neodymium than in samarium, is the age calculated a minimum or a maximum? By compar-

ing it with the model age for strontium, can you improve your knowledge of

? What do you

conclude from this?

Answer

¼1.69 Ga and



, therefore

¼1.55 0.13 Ga.

6.1.6 The problem of mixtures

In the phenomena we have considered so far, the process by which magma is generated is

straightforward enough: partial melting with isotope equilibration followed by transfer

towards the surface. In fact, actual geological phenomena are more complexbecause mix-

ing phenomenaoc cur frequently in nature.Wehavealreadycome acrossthesewhen dealing

withproblems ofgeochronologyandthen in the exercisein Section6. 1.4.

230 Radiogenic isotope geochemistry

Togetour ideas straight,letusbeginwithan example. Letus considerthe caseoftwo con-

tinental basalts locate d in, say, La Cha ˆıne des Puys in the Massif Central ( France). One of

thebasalts contains xenoliths

ofmantle rock(peridotite), showing thatthe magmadid not

remainin a magma chambe rasitwastransferredtothesurface,otherwisetheheavierxeno-

liths would have settled out. Its

Sr/

Sr ratio is similar to that of peridotitic xenoliths at

0.7035.The second basalt contains slightly more silica and does not c ontain any xenoliths

ofdeep origin, but, on the contrary, xenoliths ofgranite which were probably scoured from

the conti nental crust.Th e second basalt has a

Sr/

Sr ratio of 0.706.The granitexenoliths

have

Sr/

Sr ratiosof 0.740.

Weinterpretsuchanoccurrencebyconsideringthatthe secondbasalthasbeen contam i-

nated by continental cru st.What is the extent of this contamination? From the mixing for-

mula alreadystudied atthe endof Chapter 3:



mixture



mantle magma

x þ



continental crust

ð1  xÞ

where x ¼ðm

Þ=ðm

þ m

Þ is the m ass of un contaminated magma, m

the mass of conti nental cr ust dissolved in the mag ma, C

the strontium concentration

ofthe un contaminated magma ¼30 0 ppm, and C

the strontium concentration of conti-

nental c ontaminants ¼100 ppm.

With (

Sr/

Sr)

mixture

¼0. 705, (

Sr/

Sr)

magma

¼0. 7035 , and (

Sr/

Sr)

¼0. 740, we get:

x ¼

0:740  0:706

0:740  0:7035

¼ 0:931

for the proportion ofstrontium ofdeep origin (mantle).

Let us now determine the contaminated quantity. If we posit u ¼ m

=ðm

þ m

Þ,

which is the proportion of the continental crust assimilated, we ¢nd u ¼0.18 or 18 %. The

magmais contaminated byabout 20%in mass.

Wherewehaveseveralsamples,masscontaminationphenomenamaybeshownupbythe

linear relation

Sr=

Sr; 1=C



where C

is the strontium c oncentration. (Lookback at

Chapter 3.)

Exercise

Analysis of a series of rocks related to a carbonatite

volcano yields the results shown in

Table 6.2.

Given that the granite country rock contains 100 ppm of strontium and has a

Sr/

Sr ratio

of 0.715, can this suite of rocks be interpreted as contamination of a deep magma in a granitic

environment? How can this be demonstrated? Can the isotope composition of the deep

magma be ascertained and what is its value?

Answer

Yes, it is a mixture. This can be shown by applying the (1/

) test of Chapter 3, Section 3.4.2.

The composition of the uncontaminated magma is 0.7030.

Xenoliths are pieces of rock that are ‘‘foreign’’ to the magma, having been picked up during its transfer

towards the surface.

Look up what this and the words in the table mean in a petrology textbook.

231 Strontium isotope geochemistry

Table 6.2 Analysis of volcanic rocks

Rock

Sr/

Sr C

(ppm)

Carbonatite 0.7040 1000

Ijolite 0.70 62 350

Syenite 0 .7070 280

Nephelinite 0.7107 150

μ =

Rb/

Time

Second

stage

First

stage

mixture

α =

Sr/

Time

mixture

α =

Sr/

Time

μ =

Rb/

Time

α =

Sr/

Time

d e

Figure 6.8 Evolution over time of two systems involved in mixing at

. (a) The

Rb/

Sr ratios do not

change until

, when mixing occurs. (b) The corresponding evolution of the

Sr/

Sr ratio.

Sr/

Sr ratio at

. The shift could also be negative!

This is the only case where the radiogenic

Sr/

Sr ratio can decrease. Otherwise, it invariably increases

as a result of radioactivity. (d, e) Evolution of two systems which underwent isotopic exchange at

without changing their chemistry.

232 Radiogenic isotope geochemistry

These mixing phenomen a may occur at depth or n ear the surface during the system’s

isotopic evolution, for example, during tectonic phenomena where rocks are mixed by

th inni ng and folding or by breaking and mixing, or during mantle convection processes

associated with stirring.Theyalso oc cur, of course, in the processes oferosion, sedimenta-

tion,transport,and theformation ofoceansand lakes.Theyhavesubstantialrepercussions,

asweshall seeon manyoccasions inwhat follows.Each t im e,th ey int ro du c e a hiatu s i n th e

(

Sr/

Sr,t) isotope evolution diagram.This isotope ratio may rise or fall suddenly. T he

chemical

Sr/

Sr ratio mayalso change abruptly.Thus, theslopeofevolution after mixi ng

changes suddenly(Figure 6.8).

6.1.7 Isotopic exchange

We sp oke of isotope exchange at the beginning of this book when discussing the isotopic

dilution method. But the phenomenon extends well beyond the con¢nes of the laboratory.

In nature, when two systems (rocks, minerals, water/rock mixtures) are i n chemi cal equili-

brium but have di¡erent isotope compositions, both systems exchange their atoms to tend

towards isotopic homogeneity (while being, it is worth repeating, chemi cally heteroge-

neous).The¢nalisotopic compositionforheav yelementsisequal,ofc ourse,totheweighted

mean ofthe systems involved.

As we have said, it is the di¡usion process es that control the rate at which isotopic

exchanges occur.They are faster in liquid phases than in solid phases and are activate d by

temperature.

Wehave madeuseofthisphenomenon i n the case oftheisochron diagram. Here, wewish

to remind readers that it plays a role in all chemical processes, but naturally this isotopic

exchange maybe partialor total dependingon circumstances.

6.1.8 The Schilling effect

Jean-Guy Schilling of Rhod e Island University discovered an intermediate structure

between MORB and OIB, which appears to complicate matters but in fact con¢rms the

dual origins of the basalts. Hot-spot injections occur at some points on the mid-oceanic

ridges, as re£ected by the ridges being subjected to topographic bombardment (see review

inSchil ling,1992) (Figure6.9).

The ¢rst reported an d most spectacular case of such hot-spot injection beneath mid-

oceanicridgesisthatofIceland.Animmensevol canicstructureofdeeperoriginliesastrid e

the mid-Atlantic ridge creating a huge oceanic island. A less spectacular case is found

on the Atlantic ridge n ear the Azores. There the bombment remains submarine, because

the emerged islands areshifte d tothe east.

Now, analysis of the isotope comp ositions of these mixed structures reveals

Sr/

values that are intermediatebetween MORB and OIB values.The distribution of

Sr/

isotope ratios along the mid-ocean ridge follows th e pro¢le of the topography (Figure 6.9).

It is only at some considerable distance from the hot spot that isotope values typical of

MORBrecur.

233 Strontium isotope geochemistry

Whentheareasoftheworldwh eretheSchillinge¡ectoccurs are removedfrom thestatis-

tics, the dispersion of MORB values declines. The range of variation is then just

0.7022^0.7027, and so is far nar rower than the var iation observed for OIB, which is

0.7032^0.7050.This observation, which has been con¢rmed for m anyocean ridges, shows

there is a clear statistical distinction between MORB sources with a narrow range of very

low

Sr/

Sr ratios and the higher and more scattered

Sr/

Sr ratios ofthe hot spots.This

shows that samples must be collected from ocean ridges using knowledge ofthe geolog ical

and geodynamic context before interpreting the isotope m easurements and before distin-

guishing the MORB i nteracting with hot spots from the others (Hart et al., 1973)

(Figure6. 1 0).

6.2 Strontium–neodymium isotopic coupling

The development of neodymium isotopic geochemistry was a key moment in

isotope geology. It was brought about by three teams: th e p re s ent aut ho r ’s team at the

Institutde Physique du Globe in Paris (Richard et al., 1976), Jerry Was serburg’s team atthe

California Institute of Technology (DePaolo and Wasserburg, 1976a, b), and the team of

Keith O’Nions then at Columbia University, New York (O’Nions et al., 1977). This was a

decisive stage in the use of isotope trace rs for understan ding geodynamic processes and

the historical evolution of terrestrial reser voirs because, for the ¢rst time, two chemically

di¡erentisotopic tracersyieldedcoherent, co mplem entaryinformation.

Sr/

sea level

63°65° 61°

Latitude N

Depth

58° 52°

63°65° 61° 58° 52°

500 m

0.7026

0.7030

Reykjanes

ridge

Iceland

Figure 6.9 Study of samples collected south of Iceland. Variation in the

Sr/

Sr ratio along the mid-

oceanic ridge southwards from Iceland along with variation in submarine topography. The parallel is

astonishing. After Hart

.(1973).

234 Radiogenic isotope geochemistry

As said in the previous chapters,

147

Sm de cays by  radioactivity to

143

Nd with a half-

life of1.5576 10

years and a decay const ant l of 6.42 10

12

1

. It is possible, then, in

principle to study variations in

143

Nd/

144

Nd isotope ratios in basalts and granites as

with

Sr/

Sr ratios.The di⁄culty is th at these Nd isotope variations are ve ry small, of

the order of 1 per 10 000. It was therefore not until very precise mass spectro-

metry techniqu es were developed for the Apollo missions that this program could be

carried out.

6.2.1 Neodymium isotope variations

It was therefore not until1975 in Paris, and then 6 months later in Pasadena, that variations

in the isotope composition of neodymium ofterrestrial originwere shown to existbydraw-

ingon thegiantleap in mass spectrometry:



MORBshave

143

Nd/

144

Nd isotope ratiosofab out 0.51320;



OIBshave

143

Nd/

144

Nd isotope ratiosofabout 0.5128;



the isotope ratios ofgranitic rocks range w idely from 0.508 to 0.511d epending on the age

ofthe rock.

Qualitatively, the phenomenon is the sam e as for

Sr/

Sr ratios, exceptthatthevariations

are in the opposite direction.The most radiogenic strontium ratios are for granites, while

the most radiogenic neodymium ratios are found in MORBs.This inverse variation called

for closer study and exami nation of the isotope correlation between Sr and Nd measured

onthe samesamples.

6.2.2 The "

notation

To make the variations easier to read, we shall express the Nd variations as variations rela-

tive to a reference sample. This reference sample is chosen from among meteorites whos e

E MORB

MORB

Sr /

0.702 0.703 0.704 0.705

OIB

Figure 6.10 Histogram of isotope ratios measured on various types of oceanic basalts. Hoffmann has

called the MORB subjected to the Schilling effect E-MORB. Normal N-MORB is located outside of these

peculiar areas.

235 Strontium–neodymium isotopic coupling

p resent-day isotope ratio is constant (whichever meteorite it maybe) at 0.512 638.

The "

unitisde¢ned(seeDePaol o andWa s s e r b u r g ,1976a)by:

143

144



sample



143

144



chondrite

143

144



chondrite

10

Inshort,"

isarelativevariationexpressedper10m il,whichis more convenientforexpres-

sing variation s and can be used to m anipu late numbers like 2,10, or 20 whether positive or

n egative.

Exercise

Calculate the "

ratios for two samples: one is a MORB whose

143

Nd/

144

Nd ratio is 0.513 10

and the other a billion year-old granite whose

143

Nd/

144

Nd ratio is 0.509 00.

Answer

By using the isotope ratio of chondrites as a reference, we ﬁnd " ¼þ9.36 for the MORB and

" ¼–70.6 for the granite.

Exercise

The isotopic composition of a mixture (noted

) of two components A and B of isotopic

composition

and

ð1 

Þ;

being the mass fraction of component A in the mixture. Show that if we use the " notation,

the mixture "

between two components (A) and (B) with "

and "

can be written:

¼ "

þ "

ð1 

Þ:

Answer

Begin with

(1 –

). Subtracting

(standard) from each side gives:

–

¼(

–R

)

þ(

–R

)(1 –

Dividing both sides by

and multiplying by 10

gives:

¼ "

þ "

ð1 

6.2.3 The Sr–Nd isotope correlation of basalts

As mightbe expected after the initial measurements, when the Srand Nd isotope composi-

tions are measured on a series ofoceanic basalts, they are found to be anti-correlated.The

Sometimes another value is given in the literature because of the way stable Nd isotopes are normalized.

This diﬀerence is removed with the " notation.

236 Radiogenic isotope geochemistry

correlation observed is good (Figure 6.11) and is an inverse correlation (DePaolo and

Wa s s e r b u r g ,1976b; Ric hard etal., 1976; O’Nions etal.,1977).

Oceanic basalts de¢ ne a straightline ofcorrelationwith lowdispersion.Th is correlation

is very important as itshows coherence between Nd and Sr isotopevariations.These vari a-

tions, then, are not caused bysome minorchemical processes speci¢c to Sror Nd elements

but bygeneral grand-scale geological processes. It can be consi dered, at least as a work ing

hypothesis,thattheSrand Nd isotope ratios togetheract as tracersforthe major g eological

phenomena, becausetogether theykeep a recordofthesephenomena.

Firstofall,twoessentialreferenceparameterscanbeextractedfromthiscorrelation:th eNd

and Sr isotope values for the Bulk Earth and the Sm/Nd and Rb/Sr ratios for the Bulk

Eart h.TheBulkEarth is a complexsystem comprisingvarious entities: the continental crust,

oceanic crust, mantle, and core. It is noteasy to obtain a single meanvaluefor thewhole. But,

conversely, whenwe doobtain suchavalue,itacts asa‘‘universal’’referenceforallthe others.

Let us come backbrie£y to the meteorites we took as the reference standard for calculat-

ing the  parameters of neodymium (as already speci¢ ed in C hapters 4 and 5). Meteorites

have extremely varied chemical compositions. Some have chemic al compositions quite

close tothatofthe Sun.These‘‘primitive’’meteorites are carbonaceous chondrites. Others,

byc ontrast, have chemical compos itions analogous to thatof basalt.Theyare the products

ofvolcanism and are known as basalticachondrites.Yetothers are made up of iron^nickel

allo ys: these are the iron meteorites. Other better known because more abundantones are

chemically intermedia te: these arethe ordinarychon dr ite s as seen intheprevious chapters

(see Jacobsen and Wasserburg,1980). All ofthese meteorites are ofthesame age towithin

0.703 0.704 0.705 0.706

–5

–10

0.511

976

0.512 638

0.513 300

143

Nd/

144

Sr/

Bulk Earth

for

143

Nd /

144

Bulk Earth

for

Sr/

OIB

MORB

Figure 6.11 Strontium–neodymium isotope correlation established by the Paris, Pasadena, and

Columbia teams (1976–7) on oceanic basalts. Thousands of measurements have been added since,

and we shall see later how the initial correlation has been altered. The correlation is used for

determining the

Sr/

Sr ratio of the Earth from the

143

Nd/

144

Nd value of the Earth as determined

from meteorites (and then taken as a reference for calculating ").

237 Strontium–neodymium isotopic coupling

a few million years: around 4.55 0.05 Ga. Now, all of them have the same present-day

143

Nd/

144

Nd ratio of 0.512 638, but not the same

Sr/

Sr ratio !

It is the refore reasonable

to suppose that if these meteorites, which represent the primordial planetary material,

havethesame Ndisotope comp ositionandanear-constant

147

Sm/

144

Nd ratio,theprimiti v e

material of the E arth must also have the same isotopic and chemical composition.These

ratios arethereforeprobably those ofthe Bulk E arth.

On the strength of this and of the (Sr^Nd) isotope correlation, we can determine the

strontium isotope composition of the E arth (Figure 6. 11 )and,assuminganageof4.55Ga

for the Earth, we can determine its

Rb/

Sr ratio. Note that we know the i nitial

(

Sr/

Sr)

ratio whi ch is the initial ratio of meteorites of 0.698 98 and is assumed to be the

samefor the early materialofthe Earth.

Exercise

Calculate the Rb/Sr ratio of the Earth given that the present-day

Sr/

Sr ratio is

0.7050.

Answer

We write the equation for evolution in a closed system, which is legitimate for the Bulk Earth

since neither Rb nor Sr escapes from it. If the age of the Earth is

, we have:



ðe

 1Þ:

Hence:







ðe

 1Þ

0:705  0:6989

0:006 674 5



Earth

¼ 0:091 0:004:

The second stage is more difﬁcult than it looks. We must shift from an atomic isotope ratio to

a gravimetric chemical ratio:

total

¼ð

Rb þ

RbÞatomic mass of Rb

total

Rb 1 þ



 atomic mass of Rb

total

¼ð

Sr þ

SrÞatomic mass of Sr

The explanation for this is that Sm and Nd are ‘‘refractory’’ elements, that is, they do not vaporize easily.

Strontium is refractory too, but Rb is volatile. When solid bodies were created in the primordial solar

nebula, fractionation between refractory and volatile elements was an essential phenomenon. However,

refractory elements behave quite coherently. Thus Sm/Nd ratios of meteorites are almost constant

whereas Rb/Sr ratios vary greatly.

238 Radiogenic isotope geochemistry