All?gre Claude J. Isotope Geology

Подождите немного. Документ загружается.

CENOZOIC

MESOZOIC

PALEOZOIC

4.5

0.55

-30

⌠

-5

⌠

-5

⌠

NEOGENE

PLIOCENE

MIOCENE

OLIGOCENE

EOCENE

PALEOCENE

PALEOGENE

CRETACEOUS

JURASSIC

TRIASSIC

PERMIAN

CARBONIFEROUS

DEVONIAN

SILURIAN

ORDOVICIAN

CAMBRIAN

PRECAMBRIAN

CENOZOIC

MESOZOIC

PALEOZOIC

Formation

of the Earth

Hadean

Archean

Isua formation

Greenland

Barberton

Great Canadian

orogeny

Cambrian

Phanerozoic time

West-African

orogeny

Proterozoic

Grenville

orogeny

Pan-African

orogeny

Eoarchean

Paleoarchean

Mesoarchean

Neoarchan

Paleoproterozoic

Mesoproterozoic

Neoproterozoic

NEOGENE

PLIOCENE

MIOCENE

OLIGOCENE

EOCENE

PALEOCENE

PALE OGE NE

CRETACEOUS

JURASSIC

TRIASSIC

PERMIAN

CARBONIFEROUS

DEVONIAN

SILURIAN

ORDOVICIAN

CAMBRIAN

PRECAMBRIAN

570

500

435

395

345

280

235

195

141

Formation

of the Earth

First

humans

Precambrian

times

Phanerozoic

times

Phanerozoic times

4.55 Ga 1 Ga 100 Ma 10 Ma 1 Ma 100 ka 10 ka 1 ka 100 years

1 Ma 100 ka 10 ka 1 ka 100 years

231

230

234

8.9

226

500 Ma

107

Pd-

107

129

I-

129

182

Hf-

182

146

Sm-

142

244

Pu-Xef

CONDENSATION

FORMATION OF THE EARTH

METEORITES

Granites

Basalts

Kspaths

Muscovites

Biotites

Garnets

Granites, Basalts

Sm/Nd

or Lu/Hf

U/Pb

Rb/Sr

Re/Os

K/Ar

U/He

100 Ma10 Ma1 Ma100 ka10 ka

Basalts

Granites

Sphenes, Apatites

Zircons

Precambrian Paleoz

Mesozoic

Cenozoic Quaternary

Pleistocene Holocene

Cambrian

Ordovician Silurian

Devonian Carboniferous

Permian

Triassic

Jurassic

Cretaceous

Eocene

Oligocene

Pliocene

Miocene

Mn-

Fe-

Al-

Ca-

Plate 7 Ranges of the main geological clocks. Notice that the timescales are logarithmic.

and basalt arise. Granite is the es sential c onstituent of contin ental cru st.The continental

cru st is a permanent feature of the Earth’s surface and is very, very old, at least 4 Ga and

possibly older. When granites are systematically dated on a continent and the datings

mapped, geological age provinces are de¢ned. The classical example is that of Nor th

Amer ica, as we have seen, and the same is true of other continents. There are granites

aged 3.8 Ga (Greenland), 2.7 Ga (Canada, Scandinavia), and 2 Ga (Ivory Coast,

Sahara), but also granites aged 50 Ma (Himalayas) or even 5 Ma (Himalayas, Andes,

Alps) (see Plate 5).

Asgeologicalstudieshaveshown, theseprovinces correspondtowell-de¢nedepisodesin

geological history during which orogeny, that is, mountain building, occurred. Rock was

folded and metamorphosed and granite was formed by melting of the lower crust and was

injected intotheupper partofthe crust during these orogenic episodes. In accordance with

one ofthebasic principles ofplatetectonics, these piecesofcontinent£oaton the surface of

the mantle, they are broken up and drift about but are not swallowed up as suchwithin the

mantle by subduction.They therefore form a mosaic of blocks of di¡erent geological ages

which have accumulated, brokenup, and driftedtogether atthe Earth’s surface throughout

geologicaltime.

In the course of geologic al cycles, with the su ccession of erosion, sedimentation, meta-

morphism,andfoldingandformationofgranites,theisotopicgeologicalclockswehavestu-

died in the previous chapters are more or less reset and so date these episodes more or less

accurately. But a great part of the ancient continental material is preserved and recycled,

whether itis rejuvenated or not.

Basal t is the primary constituent of oceanic crust. Unlike c ontinental crust, oceanic

cru st is very young. Oceanic cru st arises from the mid-ocean ridges, spreads, and then

plunges at sub duction zones to be recycled in the mantl e. The oldest oceanic crust is

200 Ma,the averageagebeing80 Ma. Bycontrast, the rocks ofthe present-dayocean ridges

arevirtuallyofzero age. Oceanicbasaltsarethe p roductofmantle- meltingand as thatmelt-

ingoccurs athightemperatures isotopicequilibrium is achieved.This is whyo ceanicbasalt

re£ ects the isotopic compositions of the present-day mantle. Now, the mantle, being sub-

jected tovigorous convective motion, is probably well mixed. Itis onlytobe expected, then,

that the isotopic compositions of such a medium should be far less heterogeneous than

those of a medium that is segmente d like a mosaic of old pieces of various ages, which is

whatthe continental crustis (Figure6. 1).

Let us now try to explain the di¡erence in the strontium isotope structure of the conti-

nents and of the oceanic basalts in more quantitative terms. By the same token, we shall

bring together the reasoning already employed and which helped us i n developing

geochronology.

6.1.2 Isotope evolution diagrams and multi-episode

evolution models

One simple wayof representi ng changes in

Sr/

Sr ratios is to plot a diag ram of

Sr/

versus time. Ifwe write as our referencethe evolution ofan isotope ratio in a closed environ-

ment, by using the linear approximation, we obtain an equation we have already come

across:

221 Strontium isotope geochemistry



lt:

This is the equation ofa straightline inthe

; t



diagramwhose slope is l



and

so is proportional to th e Rb/Sr chemical ratio. The higher  ¼

Rb/

Sr is, the faster the

Sr/

Sr isotope ratio grows, an d the steeper the slope of the straight lin e of evolution

(Figure6.2).

Let us now consider a system comprising two episodes. We begin with a (

Sr/

Sr)

isotope ratio ¼

. During the ¢rst episode from T

to T

(the ages are counted from the

p resent day) th e system has a

Rb/

Sr ratio ¼

.Then, at T

, the chemical ratio changes

because th e initial system is separated into several sub-systems, A, B, and C, whose

Rb/

Sr ratios are noted 

, 

,and

, respectively. The three sub-systems evolve as

closedsystems untilth e presentday.The evolution equations for each medium arewritten:

Sub-system A ¼ 

present day

¼ 

þ l

ðT

 T

Þþl

¼ 

þ l

Sub-system B ¼ 

present day

¼ 

þ l

ðT

 T

Þþl

¼ 

þ l

Sub-system C ¼ 

present day

¼ 

þ l

ðT

 T

Þþl

¼ 

þ l

where 

isthe common isotope ratioat

¼ 

þ l

ðT

 T

Þ.

Graphically, in the (, T) plane, these three equations are represented ¢rst by a s ingle

straight line joining the origin tothe pointofconvergence (T

, 

) andthen by three straight

lineswhichdivergein proportiontotheirdi¡erentvalues (Figure 6.2).

Let us return now to the question of the di¡erence in dispersion between granites and

bas al ts,thatis,between continental crustand mantle materials.

The mantle has very low

Rb/

Sr ratios, ranging from 0.01 to 0.1. The evolution of its

Sr/

Sr on the isotope evolution di agram is almost a horizontal line if we take dispersion

Sr/

MORB

Granites

and

gneisses

0.700 0.750 0.800

Figure 6.1 Histogram of

Sr/

Sr ratios. Ratios are measured in the mid-ocean-ridge basalts (MORBs)

and in the granulites of the lower continental crust and the granitoids of the upper continental crust.

Notice the wide dispersion of values for granite compared with MORB values.

222 Radiogenic isotope geochemistry

measured on granite as our scale. The ratios representing the continental crust are much

higher: they rangefrom0.3 to 5. Letusgettheorderofmagnitude clear in our m inds.

Exercise

Suppose that granites from the continental crust derive from the mantle, which has a

Sr=



isotopic ratio 

¼0.701. Calculate the present-day

Sr/

Sr ratios of the granites assuming

that their ages are 3, 2, and 1 Ga, respectively, with

Rb/

Sr ratios of 1, 2, and 3, correspond-

ing to the time when they became granites, therefore when they were extracted from the

mantle. Plot the isotope evolution on an (,

) diagram.

Answer

Here is the matrix of the results for the

Sr/

Sr ratios of these granites (Table 6.1). Assuming

the mantle has a

Rb/

Sr ratio of 0.01, the evolution of the granites is shown in the (,

)

diagram in Figure 6.3. It shall be seen that the Rb/Sr ratios and age dispersions assumed in the

model are enough to explain the wide dispersion of continental granites observed.

Time

Differentiation

present

Rb/

Time

present

Rb/

Sr/

present

tg = λ

Figure 6.2 Diagrams of the evolution of isotope systems with time. Left: the

Rb/

Sr parent–daughter

ratio versus time. Right: the

Sr/

Sr ratios for the same systems under the same conditions. Top:

evolution of several closed systems where the

Rb/

Sr ratios are constant. Bottom: a two-stage

evolution. The ﬁrst episode has a constant Rb/Sr ratio and evolves in a closed system. At

a sudden

chemical differentiation event occurs giving rise to two systems (A) and (B) with two different Rb/Sr

ratios. Afterwards, the systems evolve separately as closed systems. Right: the chemical scenario as it

translates on the isotope evolution diagram.

223 Strontium isotope geochemistry

EXAMPLE

The history of the granites of western Greenland

A real case is represented in the strontium isotopic evolution diagram (Figure 6.4), that of the

granites of western Greenland studied by Steve Moorbath of the University of Oxford and his

various co-workers. Notice that this real diagram (almost) matches the diagram for the

theoretical model (see Moorbath and Taylor, 1981).

Letustiethisinwithwhatwelearnedin Chapter 3 abouttheisochron diagram.Inthethe -

oretical two-stage model of evolution with di ¡erentiation at T

, we represented the system

in an (,T) diagram.The equations for the threestraightlines were:



¼ 

þ l

Table 6.1 Present-day (

Sr/

Sr) ratios

Age (Ga)

 321

30.82880.78620.7436

2 0.7862 0.7578 0.7290

1 0.7436 0.7294 0.7152

3 2 1

0.800

0.750

0.700

Sr/

Time (Ga)

Figure 6.3 Theoretical (

Sr/

Sr,

) isotope evolution diagrams.

224 Radiogenic isotope geochemistry



¼ 

þ l



¼ 

þ l

In these equations the common isotope composition at timeT

is n oted 

.The three sys-

tems could havebeen represented in an (, ) diagram (Figure 6.5).This is the isochron dia-

gram, which is one of the foundations of geochronology.Why use ( , )ingeochronology

and (,T)in isotopegeochemistry? Theanswer is importantas itde¢nes a method.

In geoch ronology,what we measu re di rectly are

and

. It is th erefore normal to plot

the two ty pe s of measu rement that are to de¢ne a correlation diagram whose slope, that

is , th e age, we s hal l calc u late. In is otope ge och e m is t ry, we deter m i n e age s T and i n it ial

0.730

3 2 10

0.720

0.710

0.700

Sr/

Time (Ga)

Mantle

Greenland granites

Figure 6.4 Isotope evolution diagrams for rocks of western Greenland. A, Amitsoq gneiss; B, Nuk gneiss;

C, Ketelidian gneiss; D, Quorguq granites. After Moorbath and Taylor (1981).

λT

Sr/

Rb/

Figure 6.5 Equivalent to the diagram at the bottom of Figure 6.2, but plotted differently.

225 Strontium isotope geochemistry

ratios 

by is ochron constru ction diagrams on the variou s s ets of rock, of varied age s.

Ourdatawill therefore be 

and T. Itis normal, once again, toplotthe datafor one against

the other.Theunknown (provided by the slop e) is the

Rb/

Sr ratio, wh ich is therefore a

ch emical parameter.

6.1.3 Granites and granites

Inthetheoretical modelofevolutionofgranites intheexercise, weassumedthatthegranites

derived from di¡erentiation of the mantle.Their initial isotope ratios were the refore those

of the m antle. This idea of granite der iving from the mantle by chemical di¡erentiation

from basaltic m agmawas ¢rst proposed by Norman Bowen. Inthe 1920sand 1930she con-

ductedthe¢rstpetrologyexperimentsinthe courseofwhich he madeapieceofgranite arti-

¢ciallybycrystallizing a meltofbas altbystages (fractional crystallization) (Bowen,1928).

After thewar, a secondideaarose. Having¢rstbeen evokedbyBowen himselfitwasactu-

allydeveloped bythe German schooland particularlyHelmutWinkler. Hetoo m ade gran-

ite in the laboratory, but this time in a quite di¡erent way, by melting clastic se diments

(Winkler,1974). In petrology, thisprocess isknown as anatexis. Granite is thoughttobethe

endproductofextreme metamorphismwherethe increasein temperature causes melting.

Thesetwotheoriesontheoriginofgranitehaverepercuss ionsforisotopegeochemistry.If

a granite d erives from basalt, its initial

Sr/

Sr ratio is close to that of basalt and so very

low (close to 0.703). Onthe contrary, ifitderives from ancientsediment, which itselfderived

from the erosion of ancient g ranite, its

Sr/

Sr ratio will be much higher (more like 0.720

to 0.780). The corollary of this observation is that, by measuring the initial isotope ratios

(thatis, bycorrecting foragebytheiso chron construction), the origin ofgranites canbesys-

tematicallystudied.

EXAMPLE

Himalayan granites

If we measure the initial

Sr/

Sr isotope ratio of Lhasa granite (Tibet) dated 50 Ma, we ﬁnd

0.705 whereas the corresponding ratio for granite of the geographically very close High

Himalayas near Mount Everest, dated 30 Ma, is 0.780. Comparing these ratios with the

histogram in Figure 6.1 reveals that the two granites have very different sources, origins

and histories. The Lhasa granite Sr initial isotope ratio is similar to

Sr/

Sr ratios for the

mantle. It may originate directly from the mantle or indirectly from the intrusion of magma

from the mantle which has crystallized and remelted. By contrast, the High Himalayas granite

collected from the slopes of Mount Everest has a Sr isotope ratio situating it clearly among

values for the continental crust. This granite derives from remelting of older continental crust.

Himalayan granites are therefore of two separate origins (author’s unpublished results).

There, then, is a ¢rst example of an isotopic tracing technique. A straightforward mea-

surement of strontium isotope ratios has provide d a solution to a problem over which gen-

erations of petrologists have argued, that ofthe origin of granites, and to make a quick and

simple inventoryofthem.We shall return to this issue l ateron as it is central to the origin of

the continentalcrust.

226 Radiogenic isotope geochemistry

6.1.4 Basalt and oceanic basalt

Given the greatdi¡erence in

Sr/

Sr isotopic compositionbetween continental crust and

mantle, it is better, if we wish to determine the chemical and isotopic composition of the

mantle, to avoid taking samples from continental volcanoes (such as those of the Massif

Central in France, for example). In this way, we avoid contamination by continental mate-

rial, which is alwayspossible andwould skew results.We shal l therefore look ¢rst atoceanic

volcanism. Since the oceanic crust is basaltic, the

Sr/

Sr ratios of magmas and th e ocea-

nic crust aresimil ar, whichreducesthe riskofanydisastrous contamination.

Exercise

A basaltic magma with a

Sr/

Sr ratio of 0.7025 and a Sr concentration of 300 ppm passes

through granitic crust with a

Sr/

Sr ratio of 0.750 and a Sr concentration of 150 ppm. The

basalt assimilates granite in the proportions of 1%, 2%, and 5%. What are the isotope ratios of

the contaminated magmas? What precision is required on the

Sr/

Sr ratio measurement to

detect the phenomenon?

Answer

Notice that if we measure the isotope ratio to the nearest 1%, that is, 7 on the third ﬁgure,

we cannot detect these ﬁner points. At 1% we cannot detect anything for sure either. It can

only be seen with precision of at least 1%.

There are two types of oceani c basalt: m id-ocean- r idge basalt (MORB ),whichis

emplaced on the sea £oor at oceanic ridge crests and has very characteristic forms of £ow

known as pillow lavas, and ocean - island basalt (OIB) produced ver y largely by subaerial

volcanism and which forms island chains (Hawaii, Tahiti) or archipelagos (Azores,

Canaries, G ala

pagos).These two types of basalt are statistically di¡erent in terms of their

petrologyand chemistry.

The MORBs are said to be tholeiitic. They contain lower concentrations of alkali ele-

ments (sodium, potassium) and are slightly richer in silicon. Conversely, OIBs tend tohave

h igher alkal i co ntent s. (But again this is a statistical observation, because tholeiitic basalt

is found on ocean islands and some alkali basalts are found in the vicinity of mid-ocean

ridges.) Isotope analys is of both types of basalt (removing the altered parts for submarine

basalts) indicates that MORBshave

Sr/

Sr ratiosrangingfrom0.7022 to0.7034 whilethe

corresponding rangefor OIBsis 0.7030^0.7050.

The

Sr/

Sr isotope ratios of MORBs are statistically lower than the ratios for OIBs.

To understand this, we need to turn to geological observation, that is, the geological con-

text in which samples are collected. Samples of MORBs are collected from zones where

the ocean £oor is sp reading, as i f th e mantle were almost cropping out at the surface,

whereas OIB forms island chains ‘‘independently’’ of sea-£oor spreading. These are

plu mes of magma that cross the tectonic plates and so come from res ervoirs that lie

1% 2% 5%

0.702 72 0.702 96 0.703 65

227 Strontium isotope geochemistry

deeper than the ocean-ridge res ervoirs.This is the hot spot hypothesis proposed byJason

Morgan of Princeton Un iversity (Morgan, 1971). If we accept that S r isotope ratios are

p reserved during partial melti ng of the mantle and transport, this suggests that the man-

tle is isotopically strati¢ed, with an upper layer as the source of MORB, with a very low

Rb content (compared with Sr) and a lower layer as the source of OIB, which is a little

richer in Rb (compared with Sr) (Figure 6.6). Let us intro duce some quantitative con-

straints to make this clearer.

6.1.5 Comparative ‘‘prehistoric’’ evolution of MORB and OIB:

the concept of a model age

Suppose we have two oceanic volcanic rocks, one collected bydredging along the centerof

an oceanic ridge and the other collected from a Recent volcano of the Canary Islands.

Their

Sr/

Sr ratios are 0.7025 and 0.7045, respectively.What canwe say aboutthe prehis-

toryof the two rocks? Using just the Sr isotope ratios, we can assert that the rocks are from

di¡erent sources with di¡erent

Sr/

Sr signatures and that those di¡erences are ancient.

But how ancient? Suppose now that we measure the

Rb/

Sr ratios of these rocks and

¢nd 0.01and 0.1, respectively. Ofcourse, these ratios arenotthose ofthe sources i n the man-

tle.To make a volcanic rock from the solid mantle, a liquid magma has had to be created,

which i mplies partial melting of the mantle and associated chemical fractionation, and so

variations in the chemic al Rb/Sr ratio.T his is followed by the transferof the m agma to the

surface with possible contamination on its way up. But laboratory experiments, observa-

tions, an d measurementsonvolcanic series havetaughtus that rubidiu m enters the magma

much more readilythanstrontium. Consequently, we can infer that:

Upper mantle

source of MORB

Oceanic islands

Mid-ocean ridges

0.7035

0.7025

OIB source

mantle

Figure 6.6 The origins of mid-ocean-ridge basalt (MORB) and ocean-island basalt (OIB). The diagram

shows how the two types of oceanic basalt may have arisen on the basis of isotope analyses and

elementary geodynamics. The values are

Sr/

Sr ratios.

228 Radiogenic isotope geochemistry