CHAPTER TWO

The principles of radioactive dating

It can never be repeated enough that radioactive dating was the greatest revolution in the
geological sciences. Geology is an historical science which cannot readily be practised
without aprecisewayofmeasuring time. Itis safeto say thatno modern discoveryingeology
could have been made without radioactive dating: reversals of the magnetic ¢eld, plate tec-
tonics, the puzzle of the extinction of the dinosaurs, lunar exploration, the evolution of life,
human ancestry, notto mention the age ofthe Earth orofthe Universe!
The ages involved in the earth sciences are very varied.They are measured in years (yr),
thousands ofyears (ka), millions ofyears (Ma), and billions ofyears (Ga). Geological clocks
mustthereforebe varied too, with mean lives ranging from ayear to abillionyears.

2.1 Dating by parent isotopes

Imagine we have a radioactive isotope R and NR is the number of atoms of this isotope.
Suppose that geological circumstances (crystallization ofa rockor mineral, say) enclose an
initial quantity of R, i.e., the number of atoms of R at time zero, written NR ð0Þ, in a‘‘box.’’ If
the box has remained closed from when it ¢rst formed until today, the numberofatoms of R
remaining is NR ðtÞ ¼ NR ð0Þelt, where t is the time elapsed since the box was closed. If we
know the quantity NR ð0Þ and the decay constant l, by measuring NR ðtÞ we can calculate
the age t at which the box closed by using the radioactivity formula‘‘upside down’’:
  
1 NR ð0Þ
t ¼ ln :
l NR ðtÞ

Methods where the initial quantities of radioactive isotopes are well enough known are
above all those where the radioactive isotope is produced by irradiation by cosmic rays.
This is the case ofcarbon-14 (14C) and beryllium-10 (10Be).

Exercise
The half-life of 14C is 5730 years. The 14C content of the atmosphere is 13.2 disintegrations per
minute and per gram (dpm g1) of carbon (initial activity A0). We wish to date an Egyptian
artefact dating from approximately 2000 BC. What is the approximate activity (A) of this
artefact? If our method can measure 1 dpm, what mass of the (probably precious) sample will
have to be destroyed?
30 The principles of radioactive dating

Answer
If the half-life T ¼ 5730 years, the decay constant is l ¼ ln2=T ¼ 1:209  104 yr1 . 2000 BC
corresponds to a time 4000 years before the present, therefore since A0 ¼ 13.2 dpm g1 and
A ¼ A0 elt, A ¼ 7 dpm g1.
Making the measurement means using at least 1/7 g, or 142 mg of the sample.
As seen in the examples, the abundance of a radioactive isotope is estimated relative to a
reference. For 14C we use total carbon. The dating formula is therefore:
 
1 ð14 C=CÞ0
t ¼ ln 14
l ð C=CÞt

where (14C) and (C) represent the concentrations of 14C and total carbon, respectively.
In other cases, a neighboring stable isotope that is not subject to radioactive decay is taken
as the reference. So for 14C, we use stable 13C and we write:
" #
1 ð14 C=13 CÞ0
t ¼ ln :
l ð14 C=13 CÞt

This formulation has the advantage of bringing out isotopic ratios, that is, the ratios mea-
sured directly by mass spectrometry.

2.2 Dating by parent–daughter isotopes

2.2.1 Principle and general equations
The di⁄culty with datingby the parent isotope is ofcourse knowing NR ð0Þ, that is, knowing
exactlyhow many radioactive atoms weretrapped in thebox atthebeginning.This di⁄culty
is overcome by involving the stable daughter isotope produced by the disintegration noted
(D) (the asterisk being a reminder of the radioactive origin of the isotope R*).The parent ^
daughter relation is written:

ðRÞ ! ðDÞ
:
l

From the Curie ^ Rutherford ^ Soddylaw, we canwrite:

dNR ðtÞ
¼ l NR ðtÞ
dt

dND ðtÞ dN ðtÞ

¼  R ¼ l NR ðtÞ:
dt dt

Integrating the ¢rst equation yields the decay law, NR ðtÞ ¼ NR ð0Þ elt . The second is
therefore written:

dND ðtÞ
¼ l NR ð0Þelt ;
dt
31 Dating by parent–daughter isotopes

which integrates to:

ND ðtÞ ¼ NR ð0Þelt þ C:

The integration constant C is determined by writing in t ¼ 0, ND ¼ ND(0), hence:

C ¼ ND ð0Þ þ NR ð0Þ.This gives:

ND ðtÞ ¼ ND ð0Þ þ NR ð0Þð1  elt Þ:

But this expression leaves the troublesome unknown NR ð0Þ. This is advantageously
replaced by:

NR ð0Þ ¼ NR ðtÞ elt :

This gives:

ND ðtÞ ¼ ND ð0Þ þ NR ðtÞðelt  1Þ:

If the box remains closed for both the radioactive isotope and the radiogenic isotope, by
measuring the present values ND(t) and NR(t), we can calculate t, provided we know ND(0).
This canbe plotted as (ND(t), t).The slope ofthe curve at each point equals:
dND ðtÞ dN ðtÞ
¼  R ¼ lNR ðtÞ ¼ lNR ð0Þelt :
dt dt
It therefore equals lNR(t), at l times the parent isotope content. As this content decays con-
stantly, the slope ofthe tangent does likewise, and the curve is concave downwards.To calcu-
late an age, we write the dating equation:
  
1 ND ðtÞ  ND ð0Þ
t ¼ ln þ1 :
l NR ðtÞ

The values of ND(t) and NR ðtÞ can be measured, but t can only be calculated if ND(0) can be
estimated or ignored and, of course, if we know the decay constant l. Figure 2.1 illustrates
all these points.

EXAMPLE
Rubidium–strontium dating
Let us take rubidium–strontium dating by way of an example. As seen, 87Rb decays to
87
Sr with a decay constant l ¼ 1.42  1011 yr1. The parent–daughter dating equation is
written:
87 
1 Srðt Þ  87 Srð0Þ
t ¼ ln 87 Rbðt Þ
þ 1 ;
lRb

where 87Sr(0) is the quantity of 87Sr present at time t ¼ 0, and 87Rb(t) and 87Sr(t) the quantities
of 87Rb and 87Sr present at time t. (The term quantity must be understood here as the number
of atoms or atom–grams.) Notice that time can be reversed and the present time considered
as the starting point, which is more practical. The initial time is then in the past, age (t) such
that T ¼ t.
32 The principles of radioactive dating

ND = ND(0) + NR(0) (1–e–λt)

ND ND(0)
NR(0) = + (eλt –1)
Number of atoms of R or D N*R N*R

ND
N* R

N*R = NR(0) e–λt ND(0)

ND(0) NR(0)

0 2 4 6 0 2 4 6
Time graduated in periods Time graduated in periods

Figure 2.1 Left: the decrease in the radioactive isotope and the increase in the radiogenic isotope. Right:
the increase in the radiogenic/radioactive isotope ratio.
N*R number of atoms of the radioactive isotope (R)
ND number of atoms of the radiogenic isotope (D)
NR(0) number of atoms of R at time t ¼ 0
ND(0) number of atoms of D at time t ¼ 0
l radioactive decay constant

The equation is then written:

 
1 87 SrðpÞ  87 SrðT Þ
T ¼ ln 87 RbðpÞ
þ 1 ;
l

where 87Sr(p) and 87Rb(p) relate to the present-day values (p ¼ present), and 87Sr(T) relates to
the initial values at time (T). When dealing with minerals that are very rich in Rb such as
biotite and muscovite, the initial 87Sr is negligible relative to the 87Sr produced by decay of
87
Rb. For such systems, which are said to be radiogenically rich (or just rich for short), the
dating formula is written:
 87 
1 SrðpÞ
T ¼ ln 87 RbðpÞ
þ 1 :
l

This formula can be extended to rich systems in general:

 
1 ND ðpÞ
T ¼ ln þ1 :
l NR ðpÞ

As seen, if l is known, the age can be calculated directly from measurements of the present-
day abundances of ND(p) and NR(p). The only assumption made, but which is crucial, is
that the box to be dated, that is, the mineral or rock, has remained closed ever since the
time it formed and that closure was short compared with the age to be measured. This
is indeed the case when a mineral crystallizes or a magma solidifies as with volcanic or
plutonic rock.
33 Dating by parent–daughter isotopes

Remark
Where l is very small relative to the age, the exponential can be approximated by elt  1 þ l. This
is the case for the constant l of rubidium (l ¼ 1.42  1011 yr1) and for many others. The dating
formula is then:
 87 
1 SrðpÞ
T  87 RbðpÞ
:
l

Exercise
Suppose we have a specimen of biotite from Quérigut granite (Pyrénées Orientales, France)
whose age is to be determined from 87Rb ! 87Sr decay. We measure the total content of
Rb ¼ 500 ppm and of Sr ¼ 0.6 ppm. Knowing that Rb is composed of 87Rb and 85Rb in the
proportions 85Rb/87Rb ¼ 2.5, that the Sr is ‘‘pure’’ radiogenic 87Sr, and that the decay constant
is l ¼ 1.42  1011 yr1, calculate the age of the biotite in Quérigut granite.
Answer
The Rb content is written Rbtotal ¼ 85Rb þ 87Rb ¼ (2.5 þ 1) 87Rb, therefore:

87 Rbtotal
Rb ¼  142:8 ppm:
3:5

The 87Sr content is 0.6 ppm. (There is no need to come back to atom–grams since 87Rb and 87Sr
have virtually the same atomic mass.) We can therefore write directly:
 
1 0:6
T   298  106 yr ¼ 298 million years:
1:42  1011 142:8

If we had not adopted the linear approximation, we would have obtained:

 
1 0:6
T  ln þ 1  297:36  106 yr:
1:42  1011 142:8

As can be seen, the linear approximation is valid for the 87Rb/87Sr system when the age is not
too great.

2.2.2 Special case of multiple decay

Let us consider the case of the long-period 40K radioactive isotope which decays in two dif-
ferent ways, either by electron capture giving 40Ar or by   decay giving 40Ca, each with its
own decay constant le and l , respectively:
e–cap
40Ar

40K

40Ca
β–

What is the dating formula? Letus getbacktobasics:

d=dt½NK ðtÞ ¼ ðle þ l ÞNK ðtÞ
34 The principles of radioactive dating

where NK is the numberof 40K nuclides and NAr the numberof 40Ar nuclides.

d=dt½NAr ðtÞ ¼ le NK ð0Þeðle þl Þt :

Integrating using the usual notation with 40K and 40Ar0 gives:

40 le  ðle þl Þt
ArðtÞ ¼ 40 Ar0 þ 40 KðtÞ e 1
le þ l

with le ¼ 0.581  1010 yr1 and l ¼ 4.962  1010 yr1.

l ¼ le þ l ¼ 5:543  1010 yr1 :

The initial 40Ar is usually negligible.We are therefore generally dealing with rich systems
but not with very young systems where what is known as excess argon raises di⁄culties for
accurate age calculations.

Exercise
We analyze 1 g of biotite extracted from Quérigut granite by the 40K–40Ar method. The biotite
contains 4% K and 40K ¼ 1.16  10 4 of Ktotal.
The quantity of argon measured at standard temperature and pressure is 4.598  105 cm3
of 40Ar. What is the radiometric age of this biotite?

Answer
The dating formula to calculate the age is written:
40   
1 Ar le þ lb
T ¼ ln 40 K
 þ 1 :
le þ lb le

We must therefore calculate the 40Ar/40K ratio in atoms.

As there are 22 400 cm3 in a mole at standard temperature and pressure, the value of 40Ar
in number of moles is:
 
4:598  105
¼ 2:053  109 moles of 40
Ar:
22 400

The value of 40K is:

0:04  1:16  104

¼ 1:16  107 moles of 40
K:
40

Therefore:
  
109 2:053  109
T ¼ ln  0:1048 þ 1 ¼ 280 million years:
0:5543 1:16  107
87
Comparing this with the result obtained previously using the Rb–87Sr method, we find
about the same age but slightly younger.

There is another branched decay used in geology, that of 138La which decays into 138Ba
and 138Ce (Nakaı̈ etal.,1986).
35 Dating by parent–daughter isotopes

e–cap
138Ba λCap e– = 4.44 . 10–12 yr–1
138
La
138
Ce λβ = 2.29 . 10–12 yr–1
β–

There is also a case where even more intense multiple decay occurs, in the spontaneous
¢ssion of 238U, which yields a whole series of isotopes. Luckily, ¢ssion decay is negligible
compared with  decay. In the dating equation we can consider the constant l alone as the
decay constant, but allowance must be made, of course, for the ¢ssion products. For exam-
ple, for 136Xe we write the dating equation:
136
Xeradio lfission l t
238 U
¼Y ðe  1Þ
l
whereY is the yield of 136Xe produced during ¢ssion 0.0673, with l¢ssion ¼ 8.47  1017 yr1
and l ¼ 1.55  1010 yr1.
We saw when lookingatdatingby parentisotopes thatit was convenientto expressthe dat-
ing equation by introducing isotope ratios rather than moles of radioactive and radiogenic
isotopes.This is called normalization.Thus for 87Rb ^87Sr we use a stable strontium isotope,
86
Sr.The dating equation is then written:
" 87 87
! #
Sr Sr
1 86 Sr ðtÞ  86 Sr ð0Þ
t ¼ ln 87 Rb þ1 :
l 86 Sr

This is the form in which dating equations will be expressed from now on. A system will be
richwhen:
87 87
Sr Sr
86 Sr
ð0Þ 86 ðtÞ:
Sr

2.2.3 Main geochronometers based on simple

parent–daughter ratios

Rubidium ^Strontium 87Rb 87Sr; l ¼ 1.42  1011 yr1. The normalization isotope is
86
Sr. Developed in its modern formbyAldrich et al. (1953).

Potassium ^Argon 40K ^40Ar, with the constants already given.The reference isotope is
36
Ar. Developed in its modern formbyAldrich and Nier (1948a).

Rhenium ^Osmium 187Re  187Os with l ¼ 1.5  1011 yr1. The reference isotope is
186
Os and more recently 188Os. Developed by Luck, Birck, and Alle' gre in 1980 after a
¢rst attempt by Hirt etal. in1963.
These are the three simple clocks that are commonly found as rich systems in nature.We
shall see thatother forms ofdecay canbe used but under more di⁄cult circumstances.

Samarium ^Neodymium 147Sm  143Nd; l ¼ 6.54  1012 yr1. Normalization by 144Nd.

Developed by Lugmair and Marti in1977 after an attemptby Notsu etal. in1973.
36 The principles of radioactive dating

Lutetium ^Hafnium Lu  176Hf; l ¼ 2  1011 yr1. Normalization with

176 177
Hf.
Developed by Patchett and Tatsumoto (1980a,1980b).
These methods are supplemented by others related to radioactive chains which are to be
examined next, by the extinct radioactive methods covered in Section 2.4, and by the
induced radioactive methods examined in Chapter 4.

2.3 Radioactive chains

2.3.1 Principle and general equations
When 238U, 235U, and 232Th decay, theygive rise to three other radioactive isotopes which, in
turn, decay into new radioactive elements, and so on. The process stops when the last iso-
topes produced are the three lead isotopes 206Pb, 207Pb, and 208Pb, which are stable. It was
radioactive chains which allowed both Pierre and Marie Curie and Rutherford and
Soddy to discover the mechanisms of radioactivity. A radioactive chain can be represented
by writing:
ð1Þ ! ð2Þ ! ð3Þ ! ð4Þ ! . . . ! ðnÞ stable:

Decay involves  and  radioactivity. Alpha radioactivitygives o¡ helium nuclei.Their path,

in the (Z, N) plot, brings the end product into the valleyofstability.
Figure 2.2 shows the precise structure of the three chains. Mathematically, as studied by
Bateman (1910), radioactive chains can be described by the Curie ^ Rutherford ^ Soddy
laws written one after the other:
dN1 ðtÞ dN4 ðtÞ
¼ l1 N1 ðtÞ ¼ l3 N3 ðtÞ  l4 N4 ðtÞ
dt dt
dN2 ðtÞ ..
¼ l1 N1 ðtÞ  l2 N2 ðtÞ .
dt
dN3 ðtÞ dNn
¼ l2 N2 ðtÞ  l3 N3 ðtÞ ¼ lNn1 ðtÞ
dt dt
where Ni is the number of nuclides of elements i and Nn the number of nuclides of the ¢nal
stable isotope. Successive integration of these equations presents no real di⁄culty and is
even a good revision exercise for integrating ¢rst-order di¡erential equations with constant
coe⁄cients. Let us leave that for now and concentrate on a few simple and important limit-
ing cases.

2.3.2 Secular equilibrium: uranium–lead methods

We suppose that, in view of the length of geological time, the radioactive chain reaches a
stationary state where the content of all the intermediate radioactive isotopes remains con-
stant (this is known as secular equilibrium):
dN2 dN3 dNn1
¼ ¼  ¼ ¼ 0:
dt dt dt
37 Radioactive chains

234 238
U U
92 248 ka 4.5 Ba
234
Pa
230 1.2 m
90 Th 234
Th
Z 226
75 ka 24 d

88 Ra
1622 yr

222
86 Rn
38 d

210 214 218 235

84 Po Po Po U
2x10-4s 713 Ma 92
138 d 3.1 m 231
210 214 Pa
Bi Bi
50 d 20 m 34 ka
206 210 214 227
82 Pb Pb Pb Th 231 90
17 d Th
22 yr 27 m 227 26 h
Ac
22 yr
223
Ra 88
11 d

219
86 Rn
3.9 s
228 232
Th Th 90
215
1.9 yr 14 Ba
Po 228
84 Ac
2x10–3 s 6.1 h
211 224 228
Bi Ra Ra 88
2.2 m 3.6 d 6.7 yr
207 211
82 Pb Pb
° 207
36 m 220
Ti Rn 86
4.8 m 55 s

212
Po 216
Po 84
3x10–7 s 0.16 s
212
Bi
61 m

N
208 212
Pb Pb 82
208 11 h
TI
3.1 m
124 126 128 130 132 134 136 138 140 142 144 146

Figure 2.2 Radioactive chains represented on the proton–neutron graph. Each slot contains the symbol
of the isotope and its period T12. Identify the various types of radioactivity of the three chains for yourself.

In this case, we thereforehave:

l1 N1 ¼ l2 N2 ¼ l3 N3 ¼ . . . ¼ ln1 Nn1 ;

therefore:

dNn
¼ l1 N1 :
dt

It is as ifthere was just a single direct decay reaction (1) ! (n).

Considering the fact that in the course of geological time natural radioactive chains
rapidly reach equilibrium, the initial products give the end products directly. Dating equa-
tions can then be written:

206
Pb ¼ 238 Uðel8 t  1Þ þ 206 Pb0
207
Pb ¼ 235 Uðel5 t  1Þ þ 207 Pb0
208
Pb ¼ 232 Thðel2 t  1Þ þ 208 Pb0

where constants l8, l5, and l2 are those of 238U, 235U, and 232Th.
38 The principles of radioactive dating

238
U ! 206 Pb l238 ¼ l8 ¼ 1:551 25  1010 yr1
235
U ! 207 Pb l235 ¼ l5 ¼ 9:8485  1010 yr1
232
Th ! 208 Pb l232 ¼ l2 ¼ 4:9475  1011 yr1 :

This comes down to direct parent ^ daughter dating.Warning! Linear approximation can-
not be used with these chronometers. To be convinced of this, let us compare the values of
(elt  1) with (lt) for time intervals

of 1 and 2 billion years for 235U and 238U.
Let us consider the quantity ½ðe  1Þ  lt=½elt  1 100, which is the relative error
lt

that can be expressed as a percentage.

With 235U, for 1billion the error is 41.2% and for 2 billion 68.2%! With 238U, for 1billion
the error is 7.5% and for 2 billion 14.7%.These are highly non-linear systems, then, as can
be seen from the shape ofthe typical curves in Figure 2.3.

2.3.3 The special case of lead–lead methods

We have seen that on the geological timescale radioactive chains attain equilibrium and it
can be considered that 238U decays directly to 206Pb, 235U to 207Pb, and 232Th to 208Pb.
For uranium-rich minerals like zircon, uraninite, or even monazite, the initial amount of
lead can be considered negligible. Assuming the system has remained closed, we can
write:
206
Pb ¼ 238 Uðel8 t  1Þ
207
Pb ¼ 235 Uðel5 t  1Þ:

Taking the ratio, remembering that nowadays1 the ratio 238U/235U ¼ 137.8, we get:
207
 l5 t 
Pb 1 e 1
206 Pb
¼ :
137:8 el8 t  1

It can be seen that the 207Pb/206Pb isotope ratio of lead alone gives a direct measurement
of time. This function is implicit and calculating it requires prior numerical values
(Table 2.1).

Exercise
The 206Pb/207Pb ratio of a uranium ore deposit is found to be 13.50. What is the age of the ore
supposing it has remained closed since it crystallized and that common lead can be ignored?
Answer
We shall invert the ratio so Table 2.1 can be used. 207Pb/206Pb ¼ 0.074. The table shows that
the ratio measured lies between 1 and 1.2  109 years. The result can be improved either by

refining the table by calculating the ratio el5 t  1=el8 t  1 in the interval 1 and 1.2, or by

1
The two uranium isotopes decay each at their own rate from the time that they are first formed. Their
ratio is constant as there is no isotopic fractionation between them. However, this ratio has varied over
geological time.
39 Radioactive chains

1 0.6

206Pb*
0.8 0.4
238U

0.6 0.2
0.8

207Pb*/206Pb*
0.4 0 0.6
1 2 3 4 1 2 3 4

0.4

0.2
1
1

0
1 2 3 4

207Pb*
Time (Ga)
235U

0.5 0.5

0 0
1 2 3 4 1 2 3 4

Time (Ga)

1.0 87Rb-87Sr

T = 48.8
Age of the Earth
4.56 Ga

238U-206Pb
P/P0

0.5

T = 1.25
T = 0.70 T = 4.47
40K-40Ar

235U-207Pb

0.0
0 1 2 3 4 5
Age (Ga)

Figure 2.3 Non-linear decay of 235U and 238U and their end products 207Pb and 206Pb. (a) The radioactive
constants of these reactions are very different. The curve on the right shows the change in radiogenic
ratio 207Pb*/206Pb* versus time. (b) Comparison in the same figure of the principal radioactive clocks
used in geology, emphasizing the behavior of U–Pb systems compared to others.

using a linear approximation between 1 and 1.2  109 years. The value 207
Pb/206Pb for 1.2
billion is 0.080 12 and for 1 billion is 0.0725. The variation is therefore:

0:080 12  0:0725
¼ 3:8  105 per million years:
200

The difference is between 0.0740 and 0.0725, that is, 1.5  103 million years.
40 The principles of radioactive dating

Table 2.1 Numerical values of the radiogenic 207Pb/206Pb isotope ratio

207
Pb
Time (Ga) el1 t  1 el2 t  1 206 Pb
radiogenic

0 0.0000 0.0000
0.2 0.0315 0.2177 0.05012
0.4 0.0640 0.4828 0.05471
0.6 0.0975 0.8056 0.05992
0.8 0.1321 1.1987 0.06581
1.0 0.1678 1.6774 0.07250
1.2 0.2046 2.2603 0.08012
1.4 0.2426 2.9701 0.08879
1.6 0.2817 3.8344 0.09872
1.8 0.3221 4.8869 0.11000
2.0 0.3638 6.1685 0.12298
2.2 0.4067 7.7292 0.13783
2.4 0.4511 9.6296 0.15482
2.6 0.4968 11.9437 0.17436
2.8 0.5440 14.7617 0.19680
3.0 0.5926 18.1931 0.22266
3.2 0.6428 22.3716 0.25241
3.4 0.6946 27.4597 0.28672
3.6 0.7480 33.6556 0.32634
3.8 0.8030 41.2004 0.37212
4.0 0.8599 50.3878 0.42498
4.2 0.9185 61.5752 0.48623
4.4 0.9789 75.1984 0.55714
4.6 1.0413 91.7873 0.63930

The approximate age is therefore 1040 Ma. Direct calculation from the 207Pb/206Pb ratio for
1040 Ma gives 0.073 98. The age is therefore 1042 Ma, but such precision is illusory because
error is far greater (see Chapter 5) than the precision displayed.

2.3.4 The helium method

Natural chains feature many instances of  decay, that is, expulsion of 4He nuclei. Thus
238
U decay ultimately produces eight 4He, 235U decay produces seven 4He, and 232Th decay
produces six 4He.We can therefore write:

d 4 He
¼ 8l8 238 U þ 7l5 235 U þ 6l2 232 Th:
dt

Remark
This equation underlies the first helium dating method thought up by Rutherford (1906).

Integrating the equation, assuming 4He(0) ¼ 0, gives:

4
He ¼ 8 238 Uðel8 t  1Þ þ 7 235 Uðel5 t  1Þ þ 6 232 Thðel2 t  1Þ:
41 Radioactive chains

By admitting linear approximations, which is valid only for durations that are not excessive,
elt  1  lt and observing that 235U/238U ¼ 1/137.8 and that 232Th/238U  4 we obtain:
 
4 238 7
He ¼ U 8 l8 þ l5 þ6 4l2 t:
137:8
 7l5

By noting L ¼ 8l8 þ 137:8 þ 24l2 ¼ 14:889  1010 yr1 , we obtain
 
1 4 He
 :
L 238 U

This formula is valid for young ages.

Exercise
Take 1 kg of rock containing 2 ppm of uranium. How many cubic centimeters of 4He will it
have produced in 1 billion years?

Answer
At 2 ppm 1 kg of rock represents 103 g  2  106 ¼ 2  103 g of uranium, and 2  103 g of
2  103
uranium corresponds to ¼ 8:4  106 moles. Using the approximate formula:
238
4
He ¼ Lt U ¼ 14:89  1010  109  8:4  106
4
He  125  107 moles:

If 1 mole of 4He corresponds to 22.4 liters at standard temperature and pressure, the amount
of 4He produced in 1 billion years is 0.28 cm3.

A LITTLE HISTORY
The beginnings of radioactive dating
Rutherford performed the first radioactive age determination in 1906. He calculated the
amount of helium produced by uranium and radium per year and per gram of uranium and
found 5.2  108 cm3 yr1 g1 of uranium.
Ramsay and Soddy had measured the helium content (long confused with nitrogen
because it is inert like the noble gases) in a uranium ore known as fergusonite. The
fergusonite contained 7% uranium and 1.81 cm3 of helium per gram. Rutherford calculated
an age of 500 million years. The following year he found uranium ores more than 1 billion
years old! At that time Lord Kelvin was claiming the Earth was less than 100 million years old!
(See previous chapter.)

2.3.5 The fission track method

Fission reactions emit heavy atoms. The atoms ejected by ¢ssion create £aws in crystals.
Such defects ^ known as tracks ^ canbe shown up on mineral surfaces through acid etching
42 The principles of radioactive dating

Ionization Electrostatic Relaxation and

displacement elastic deformation

Figure 2.4 Creation of fission tracks in an insulating material.

which preferentially attacks damaged areas, those where atoms have been displaced (Price
and Walker,1962) (Figure 2.4).
The numberof ¢ssion tracks is written:
lfission 238
Fs ¼ Uðel t  1Þ
l

lfission ¼ 7  1017 yr1 t1=2 ¼ 9:9  1015 yr1 :

In practice, a thin section is cut and the number of surface tracks (and not in a volume) is
measured.The visible proportion s is calculated:
lfission 238
s ¼ q Uðel t  1Þ
l

where q is a geometric factor for switching from a surface to avolume.

To measure 238U, we take advantage of knowledge that 238U/235U ¼ 137.8 and measure
235
U by causing induced ¢ssion (by placing the ore under study in a reactor), which pro-
duces tracks that are revealed and counted.We then have the number of tracks before (s)
and after (i) irradiation.The geometric factor disappears:
s lfission 137:8 l t
¼ ðe  1Þ:
i l G

We calibratethe £ux ofneutrons inducing 235U¢ssion andthe e¡ective cross-section .2

By using a standard sample whose uranium content is known and which is irradiated at the
same time as the study sample and by counting the ¢ssion tracks produced, we can then
calculatet.

2
The effective cross-section is the probability of a reaction occurring, that is, here, the probability of
producing a fission with a given flux of neutrons. The characteristics of this will be seen in Chapter 4.
43 Radioactive chains

2.3.6 Isolation of a part of the chain and dating young

geological periods
The various radioactive isotopes of the radioactive chain belong to di¡erent chemical ele-
ments sothat, undercertaingeological conditions, one or twoisotopes inthe chain fraction-
ate chemically and become isolated, thereby breaking the secular equilibrium. Once
isolated they create a new partial chain in turn. Two straightforward speci¢c cases are of
practical importance.

The radioactive isotope becomes isolated on its own

First, when a radioactive isotope in the chain (but with a long enough period) becomes iso-
lated on its own, it gives rise to a partial chain, but being isolated from the parent it decays
according to the law:

Ni ðtÞ ¼ Ni ð0Þelt

where Ni(0) is the number of nuclides of the intermediate element at time t ¼ 0. If this num-
ber can be estimated, the decay scheme can be used as a chronometer. This is equivalent,
then, to dating by the parent isotope.

EXAMPLE
The ionium method and the rate of sedimentation
Thorium is virtually insoluble in sea water. Thus 230Th (still known as ionium from the
terminology of the pioneers of radioactivity), a decay product of 234U with T12 ¼ 75 ka,
precipitates on the sea floor, is incorporated in the sediment and so gradually buried. There,
now isolated, it decays.
At any depth x of sediment from the surface (Figure 2.5) we can write:
230
Thðx Þ ¼ 230 Thð0Þel230 t

where 230Th(0) is the surface content which is assumed constant over time. If the sedimenta-
tion rate is constant, time can be replaced by the ratio t ¼ x/Vs where Vs is the sedimentation
rate and x the length (depth):
230
Thðx Þ ¼ 230 Thð0Þ expð½x =V s lÞ

or in logarithms:
230  x
lnð230 ThðxÞÞ ¼ ln Thð0Þ  l:
Vs

The slope of the curve (ln 230Th, x) gives a direct measure of sedimentation rate and the
ordinate at the origin gives 230Th(0). (Note its order of magnitude of a millimeter per
thousand years.)
This method only works, of course, if it is assumed that 230Th(0), that is the thorium content
at the sediment surface, is constant and if the sediment has not been disturbed by chemical,
physical, or biological phenomena.
44 The principles of radioactive dating

100

230Thex(dpm g–1) 50

4.3 mm ka–1
10

1
0 20 40 60
Depth (cm)

Figure 2.5 Decreasing 230Th content in a core from the sea floor and determination of the
sedimentation rate. Thex is the excess thorium compared with the equilibrium value counted in
disintegrations per minute per gram (dpm g1).

Exercise
The lead isotope 210Pb (as a member of the radioactive chain) is radioactive with a decay
constant l ¼ 3.11  102 yr1. This natural radioactive lead is incorporated into ice deposited
in Greenland by forming successive layers of ice which can be studied like sedimentary strata.
The activity of 210Pb is measured at four levels in disintegrations per hour per kilogram of ice
(dph kg1). Table 2.2 shows the results.
Calculate the sedimentation rate of the ice. Assuming a constant rate and a compaction
factor of 5, how thick will the glacier be in 5000 years? Calculate the 210Pb content of fresh ice.
Answer
The dating equation is written noting activity by square brackets:
210 210 lt
Pb ¼ Pb e :

If the rate of deposition is V and height h, we have t ¼ h/V.

The equation becomes:
210 210
Pb ¼ Pb expðlh =V Þ

or
210  lh
ln Pb ¼ ln 210 Pb 0  :
V
If the 210Pb content has remained constant over time, the data points must be aligned in a (ln
[activity, h]) plot. The slope is therefore l=V . The data points are plotted on the graph and
the slope determined. This gives V ¼ 45 cm yr1.
45 Radioactive chains

Table 2.2 Activity of 210Pb with depth

Depth

0m 1m  20 cm 1.50 m  40 cm 2.50 m  50 cm

dph kg1 75 32  5 24  5 10  3

In 5000 years’ time, allowing for compaction, there will be 5000  45=5 ¼ 450 m of ice.
The 210Pb content is calculated: surface activity is 75 dph kg1 ice.
l210  N 210 ¼ 75 dph kg1
hence
75 75
N 210 ¼ ¼ ¼ 21:4  107 atoms of210 Pb nuclides
3:11102  ð8760Þ1 3:5  107
per kilogram of ice;

8760 being the number of hours in a year. As a mass that gives:

21:4  107  210
¼ 7:5  1014 kg of ice;
6:023  1023
where Avogadro’s number is in the denominator and the previous atomic mass of 210Pb in the
numerator.
The 210Pb content is 7.5  1017, which is very little! This shows the incredible sensitivity of
radioactive methods because 210Pb in Greenland’s glaciers really can be measured and used
for estimating the rate of sedimentation of ice.

The parent isotope is isolated and engenders its daughters

This is whathappenswithuranium; for example, when certain solidphaseslike calcium car-
bonate are precipitated uranium is entrained with calcium and then isolated. For the ¢rst
radioactive productofany notable half-life, we then have:
 
l23 N23 ðtÞ ¼ l1 UðtÞ el1 t  el23 t

where N2^3 is the number of nuclides in the third intermediate product in the chain.Why?
Because the chain includes very short-lived radioactive products such as thorium-234
(234Th) or protactinium-234 (234Pa) which reach equilibrium very quickly. Thus, 238U
decays to 234Th, an element whose lifespan is 24 days, then 234Pa, whose lifespan is 1.18
minutes: it can be considered, then, that 238U directly gives 234U whose half-life is 2.48  105
years for all types ofusual samples (corals, speleothems, travertines, etc.).
Such a method is applied, for example, to secondary mineralizations of uranium. The
soluble uranium migrates and is deposited further away, leaving the insoluble thorium
where it is. Itthen‘‘resumes’’ its decaygiving 230Th and we canwrite:
 
l230 230 Th ¼ l230 238 U el238 t  el230 t :
Therefore the age of migration canbe determined by measuring 230Th and 238U.
46 The principles of radioactive dating

Exercise
Uranium-238 (238U) decays to 234U which is itself radioactive with l ¼ 2.794  10 6 yr1.
Sea water is not in secular equilibrium relative to uranium isotopes as 234U weathers better
than 238U from rocks of the continental crust and is enriched in the rivers flowing into the
ocean. In activity, noted in square brackets [ ], ½234 U=238 Useawater ¼ 1:15. When limestone
forms from sea water, it is isotopically balanced with the sea water and so takes the value 1.15.
A fossil mollusk has been found in a Quaternary beach formation and its activity ratio measured
as ½234 U=238 U ¼ 1:05. Work out the dating equation. What is the age of the mollusk?
Answer
The dating equation is:
234  234 
U U  
238 U
¼ 238 elt þ 1  el234 t
U 0

because l234  l238 and l238 t  0 if t > 106 years.

This gives:
0h234 i 1
U
1 @ 238 U 0 1A
T ¼ ln 234 U and T ¼ 393 000 years.
l 238 U  1

2.3.7 General equation and equilibration time

We shall do an exercise to help understand equilibration time and by the same token give the
theoretical answer to the previous exercise.

Exercise
Establish the general equation for evolution of an isotope in a radioactive chain and where the
parent has a longer half-life and the daughter a markedly shorter half-life. We shall take the
example of 234U ! 230Th decay to get our ideas straight.
Answer
This is a review exercise for mathematics on integrating a linear differential equation with
constant coefficients:

d 234 U
¼ l238 238 U  l234 234 U
dt
d 230 Th
¼ l234 234 U  l230 230 Th
dt

with 238U being considered constant.

Integrating the two previous equations in succession for the example in question gives:
 
l234 234 U ¼ l234 234 U0 el234 t þ l238 238 U 1  el234 t
 
l230 230 Th ¼ l230 230 Th0 el230 t þ l234 234 U el234 t  el230 t :

These two equations can be written with the activity notation in square brackets: lN ¼ [N],
which simplifies notation and means the l constants can be dispensed with. This gives:
47 Radioactive chains

234 234 l t 238  

U ¼ U0 e 234 þ U 1  el234 t
230 230   
Th ¼ Th0 el230 t þ 234 U el234 t  el230 t :

This is the general equation for equilibration of a radioactive chain used for chronological
purposes.
From there, without resorting to long and tedions numerical simulation, let us try to answer
the question: at what speed does a greatly disturbed radioactive chain return to secular
equilibrium? Let us consider the preceding equation with l234 l230. The equation becomes:
 
l230 230 Th ¼ l230 Th0 el230 t þ l234 234 U 1  el230 t :

For the chain to achieve equilibrium, it is necessary and sufficient that:

230 234
l230 Th  l234 U:

For this el230 t must be virtually zero. So t must be less than to 6–10 periods of 230Th. The
chains therefore achieve equilibrium after more than six half-lives of the daughter product
(with the smaller radioactive decay constant).
If 234U/238U isotope fractionation occurs, which is the case in surface processes, it is 234U
which is the ‘‘limiting factor’’ for the 238U chain. The time required is therefore 1.5 million
years.
If there is no 234U/238U isotope fractionation, the limiting factor is 230Th and the equilibra-
tion time is 450 000 years.
For the 235U chain the limiting factor is 231Pa and therefore a time of 194 000 years. Both
chains of the two uraniums are equilibrated at about the same time in this case.
The isotopes of radioactive series used as geochronometers are those with decay constants of
more than one year.

Uranium-238 chain
234

U, half-life 248 ka, is used in sedimentary or alteration processes because 234U/238U

varies during alteration. The radioactive recoil of 234U extracts this uranium from its
crystallographic site and so it is easilyaltered.

230Th, half-life 75 ka.This element, named ionium, is certainly the most widely used for
surface andvolcanic processes.

226Ra, half-life 1.622 ka. It is used like 230Th but for shorter-lived surface or volcanic
processes.

210Pb, half-life 22 years. Used for studying glaciers, oceans or volcanics involving very
young phenomena.

Uranium-235 series
231

Pa, half-life 32.48 ka. This is the sister of 230Th but slightly more di⁄cult to master
analytically.
277

Ac, half-life 22 years. This element is not much used as yet because of di⁄culties in
making precise analyses.
48 The principles of radioactive dating

Thorium-232 chain
228

Ra, half-life 6.7 years.This is complementary to 226Rato constrain the timescale of Ra

fractionation.

228Th, half-life1.9 years. Averygood complement for 228Ra and 230Th.

These elements can be measured by alpha and gamma radioactive spectroscopyor by mass
spectrometry.The second technique is far more precise but less sensitive than counting. In
practice, 234U, 230Th, 226Ra, and 231Pa are measured by mass spectrometry and the others
by counting. Pb-210 canbe measured by mass spectrometrybutthis entailsgreat di⁄culties
and requires special precautions. For a general review see Ivanovich (1982).

A LITTLE HISTORY
The polemic that followed the discovery of radioactivity
Henri Becquerel discovered radioactivity almost by chance in 1898 while studying rays
from phosphorescent uranium salts excited by sunlight and trying to understand the
nature of x-rays discovered by Röntgen. But one day, although there was no sunlight, a
sample from the Joachimsthal mine in Bohemia spontaneously emitted radiation which
blackened a photographic plate, indicating as yet unknown properties of matter (see Barbo,
2003).
Some years later, when measuring the effect of these radioactive substances on an
electrometer (the particles ionized the air of the electrometer which then discharged)
Pierre and Marie Curie proposed calling the phenomenon radioactivity (activity created by
radiation). For them, it was the property certain substances, including uranium, had of
spontaneously emitting radiation.
They immediately came in for harsh criticism from British scientists relying on a crucial
observation: when the activity of 1 g of purified uranium was measured with an electro-
meter, the activity was less than that of 1 g of uranium contained in, say, 100 g of uranium
ore. In other words 1 g of uranium ‘‘diluted’’ in 100 g of inert rock was more ‘‘active’’ than 1 g
of pure uranium.
How could concentrated uranium be less active than diluted uranium? It smacked of
magic. It was Marie Curie who came up with the hypothesis of intermediate radioactive
products to explain this paradox. The discovery of radium must be set in this polemical
context, thus taking on its full significance. It was the second intermediate radioactive
product to be found after polonium.
The New Zealander Ernest Rutherford brought grist to the Curies’ mill and within a few
years the mechanisms of successive ‘‘cascade’’ decay was understood. Radioactive chains had
been discovered.

Shedding light on a paradox

Nowadays the paradox of diluted uranium being more active than pure uranium can be fully
explained. Suppose a radioactive chain is in equilibrium, say the 238U chain:

l1 N 1 ¼ l2 N 2 ¼ l3 N 3 ¼    ¼ ln N n

noting N1, N2, . . ., Nn the numbers of nuclides in the various isotopes of the chain. Thus in
secular equilibrium there is 14 times the l1N1 activity of 238U in the chain. (The input from
the 235U chain must be added to this although its contribution is small.)
49 Dating by extinct radioactivity

This is what happens in a rock several hundred million years old in which the chain has had
time to attain secular equilibrium.
Now, in purified uranium, the activity is merely l1N1.
So, if the ‘‘activity’’ of 1 g of uranium is measured, it will be 14 times less than the activity
of 100 g of 1% uranium ore. Indeed, the ore would be slightly more active because there must
also be at least 0.5 g of thorium whose chain also produces 11 times more activity than pure
thorium. However, as the thorium constant is 3.5 times less, the outcome would be an
increase in the uranium activity of about 1.5 times.
In all, the rock is about 15.5 times more active than purified uranium. This is the paradox
behind the polemic!

2.4 Dating by extinct radioactivity

2.4.1 The historical discovery
Chemical elements ^ thatis, the nucleithat make up mostoftheir mass ^ havebeen manufac-
tured in stars ever since the Universe came into being.This is nucleosynthesis. Of the nuclei
formed, some are stable and others radioactive. Among the radioactive nuclei some havevery
short half-lives: theydecayquicklygiving rise to their stable daughter isotopes. All ofthisgoes
on everywhere in interstellar space and is what provides the ordinary matter of the Universe.
These isotopes are incorporated into interstellar matterasgases ordust (see Chapter 4).
But suppose that nucleosynthesis of heavy elements (explosion of a supernova) occurred
in the vicinity of the place where the Solar System formed giving rise to certain short-lived
radioactive isotopes. Suppose too that the solid bodies of the Solar System form while these
radioactive isotopes are notyet extinct.These‘‘young’’radioactive isotopes will be incorpo-
rated with the other interstellar material in these solid bodies (planets or meteorites) and
there they will decay and give rise to daughter isotopes. The solid objects having received
such inputs will therefore have abnormal isotope abundances for the isotopes produced by
decayofthe radioactive isotope. Detecting such anomalies is therefore the ¢rst step in show-
ing the existence ofextinct radioactivity.
In1960 John Reynolds at Berkeley discovered a large excess of the isotope129 in the iso-
topic composition of xenon in the Richardton (H4) meteorite (see Reynolds, 1960)
(Figure 2.6). Now, this is not anyold isotopebutthe decay productof iodine-129 (radioactive
iodine we know how to make in nuclear reactors) whose half-life is 17 million years and
which, if it formed before the birth of the Solar System, has disappeared today. And indeed
naturallyoccurring iodine has onlya single isotope, 127I.
To prove that the excess 129Xe did come from the 129I, Peter Je¡rey and John Reynolds
came up with a most ingenious experiment combining neutron activation analysis, mass
spectrometry, and stepwise outgassing by temperature increments.They irradiated a sam-
ple ofthe Richardton meteorite using a£uxofneutrons.The 127Iwas transformedby nuclear
reaction (n, ) into radioactive 128I which transformed by   decay into 128Xe.This reaction
is written

127 
Iðn; Þ128 I ! 128 Xe:
50 The principles of radioactive dating

136 134 132 131 130129 128 126 124

x 10 x1
Xenon

Figure 2.6 Mass spectrum of xenon in the Richardton meteorite as measured by Reynolds. The bars
indicate the height of ordinary, say atmospheric, xenon.

They then heated the irradiated sample progressively, after placing it in the vacuum and puri-
¢cation line of a mass spectrometer. They analyzed the isotope composition of the xenon
extracted at each temperature increment and observed that the excess 129Xe was extracted
atthe same temperature as halfofthe 128Xe, whereas‘‘ordinary’’xenonwas extracted at a dif-
ferent temperature. Je¡rey and Reynolds (1961) concluded that 129Xe is situated at the same
(crystallographic) site as natural iodine and therefore is indeed the daughterof 129I.
This extraordinary discovery has two important consequences. First, it shows that,
before the Solar System formed (at a time in the past 5^10 times the half-life of 129I, that is,
85^170 million years), there was a synthesis of heavy chemical elements. In addition, this
radioactive decay provided an exceptional and unexpected dating tool for studying the per-
iod when meteorites (and also, as we shall see, the Earth) were formed (see Figure 2.7).We
shall concentrate on this aspect now.

2.4.2 Iodine–xenon dating

Let 129I*(0) bethe numberofnuclidesformedby nucleosynthesis attimet ¼ 0, de¢nedas the
end ofthe nucleosynthetic process.The radioactive iodine decaysby the law:
129 
I ðtÞ ¼ 129 I ð0Þelt :

Suppose now that, attimet1, some ofthe iodine is incorporated in a meteorite (A) andattime
t2 some other iodine in a meteorite (B).We can write:
129 
IA ðt1 Þ ¼ KA 129 I ð0Þelt1
129 
IB ðt2 Þ ¼ KB 129 I ð0Þelt2 :
51 Dating by extinct radioactivity

Condensation and
Nucleosynthesis of 129I planetary accretion Present
(around 4.55 Ga) day

ΔT

129Xe
accumulation

Time (arbitrary scale)

Figure 2.7 Evolution of 129I in the Solar System and its trapping in planetary bodies; T is the time
between the end of nucleosynthesis and the accretion of planetary objects.

Here KA and KB are the factors of incorporation of iodinebetweenthe interstellar cloud and
meteorites A and B, which is the ratio between iodine concentration in the interstellar
cloud and in the meteorite. For two meteorites of di¡erent chemical compositions, KA and
KB are di¡erent. In each meteorite, the 129I decays completely into radiogenic 129Xe, 129Xe*
(with an asterisk):
129
XeA ¼ 129
IA ðt1 Þ 129
XeB ¼ 129 IB ðt2 Þ:

How can thisbe used for dating as we do not know the values of KA, KB, and I(0)?
Iodine has a stable isotope 127I (the only one for that matter). We divide the express-
ions describing 129I decay by 127I. We can assume that the incorporation of iodine by the
meteorites obeys chemical laws, and so the same rules apply for the 129 isotope as for the
127 isotope. This means the Kvalues are the same for both isotopes. The K coe⁄cients can
therefore be removed from the equations describing the evolution of the isotopic ratios.This
gives:
129  129

I I
ðt Þ
127 I 1
¼ 127 ð0Þ elt1
A I
129  129 
I I
ðt Þ ¼ 127 ð0Þ elt2 :
127 I 2
B I

There remains one unknown in these equations, namely the ratio129 I =127 Ið0Þ, in other
words the (129I/127I) isotope ratioatthe endofnucleosynthesis. Itis areasonable assumption
that it was identical throughout the Solar System (and so for all meteorites).We can then
¢nd the ratiobetween the two isotope ratios ofour two meteorites:
129  
I
127 I
ðt1 Þ
129  A ¼ elðt2 t1 Þ :
I
127 I
ðt2 Þ
B
52 The principles of radioactive dating

By measuring the 129 I =127 IðtÞ ratios in both meteorites, we can in principle calculate the
time interval between the formation of the two meteorites (t2  t1). This method therefore
provides absolute ^ relative dating!
The problem comes down to measuring the (129I*/127I) isotope ratio at the time the
meteorite formed. Total 129Xe is the sum of initial 129Xe þ 129I*(t). Today 129I*(t) ¼ 129Xe*,
since 129I has decayed entirely.Therefore:
129
Xe ¼ 129 Xetotal 129 Xeinitial :

The interesting ratio is:

 
129
I 129
Xe 129 Xe Xetotal
127 I
ðtÞ ¼ 127 ¼  ¼ A elt
I Xetotal Itotal

with A ¼ 129 I=127 Ið0Þ atthe end of nucleosynthesis.

The problem comes down to measuring the fraction of radiogenic 129Xe* in total xenon,
and then measuring the chemical (Xe/I) ratio, since the iodine is all 127I.
We can calculate the age to the nearest coe⁄cient and therefore, by taking the ratio
between the values for the two meteorites, determine the relative age of the two meteorites.
The beautyofthis method lies in its capacity to measure verybrief intervals oftime between
the formation ofplanetaryobjectsbillions ofyears ago.

Exercise
The 129I/127I isotope composition measured on the Karoonda and Saint-Séverin meteorites is
1.3  104 and 0.8  104, respectively. Given that the half-life of iodine is 17 Ma, what is the
age difference between the two meteorites?
Answer
If T12 ¼ 17 Ma

ln 2
l¼ ¼ 4  108 yr1 :
T
The dating formula is applied:
129
I
127 I 1:3  104
129 I A ¼ ¼ el Dt ;
127 I B 0:8  104

hence Dt ¼ 1l lnð1:3=0:8Þ ¼ 12:1 million years. This age is actually the maximum interval
measured.

Exercise
Given that the half-life of 129I is 17 Ma, what is the shortest interval of time that can be
estimated, given the uncertainty in measuring the 129Xe*/127I ratio is 2%?
Answer
Suppose we always take the same reference, say, Karoonda. There will no longer be any error
relative to Karoonda but everything will be expressed in terms of this reference. Let us take
53 Dating by extinct radioactivity

the previous measurement and look at the limits of uncertainty. The value 0.8 has two limits,
at 0.816 and 0.784. Let us calculate the age t. We obtain 11.64 and 12.64, or 1 million years.
The age of Saint-Séverin (relative to Karoonda) is written 12.1  0.5 Ma.

2.4.3 Discoveries of other forms of extinct radioactivity

Since then many forms of extinct radioactivity have been discovered, which we shall use as
required. Each discoveryhas required the identi¢cation by experiment ofparent ^ daughter
relations in meteorites.We review themwith a fewbriefcomments.

Iodine ^Xenon 129I ^129Xe,  ¼ 25 Ma, discovered by Reynolds (1960).

Plutonium ^Xenon 244Pu ^ Xe¢ssion,  ¼ 84 Ma.This radioactivity, discovered by Kuroda

(1960), produces ¢ssion tracks and the ¢ssion isotopes of xenon 131Xe, 132Xe, 134Xe, and
136
Xe. It is an important supplement to the iodine ^ xenon method and was discovered
very shortlyafter it.

Samarium ^Neodymium 146Sm ^142Nd,  ¼ 146 Ma. This form of radioactivity, discov-
ered at San Diego by Lugmair et al. (1975), is interesting because it is has a long half-life
and allows us to connect long-duration phenomena that occurred around the time of 4.5
billionyears ago.

Aluminum ^Magnesium 26Al ^29Mg,  ¼ 1Ma. This form of radioactivity was ¢rst
detected in certain minerals from very ancient meteorites by the team of Gerald
Wasserburg at the California Institute of Technology (Caltech) (Lee et al., 1977). It was
of historical importance but it is probably more important still that aluminum is a deci-
sive constituent of meteorites (2^3%). It is probable that 26Al was instrumental in the
veryearly thermal historyofplanetesimals and its in£uence shouldbe addedtothe calcu-
lations alreadydone on this topic in Chapter1.

Palladium ^Silver 107Pd ^107Ag,  ¼ 9.4 Ma. This form of radioactivity was detected in
iron meteorites by the Caltech team (Kelly and Wasserburg, 1978). It has shown how
old these meteorites are. This means that metallic iron di¡erentiation is a very ancient
phenomenon in the processes offormation ofthe Solar System.

Manganese ^Chromium 53Mn ^53Cr,  ¼ 5.3 Ma.This form of radioactivity, discovered

in Paris by Birck and Alle' gre in 1985, is interesting because the manganese and chro-
mium fractionatebecause oftheir di¡erent volatilities.

Iron ^Nickel 60Fe ^60Ni,  ¼ 2.1Ma. This form of radioactivity has been found in just a
fewbasaltic meteoritesby Shukolyukovand Lugmair (1993) at San Diego. Itis important
because iron is averyabundant element.

Calcium ^Potassium 41Ca ^41K,  ¼ 0.143 Ma. This form of extinct radioactivity is
importantbecause of its short half-life. It was discovered by Srinivasan etal. (1994).

Hafnium ^Tungsten 182Hf ^182W,  ¼ 13 Ma. This form of extinct radioactivity, discov-
ered by Harper and Jacobsen (1994) at Harvard and then by Lee and Halliday (1995) at
the University of Michigan, is very important because Hf and W fractionate during
metal ^ silicate separation, allowing this separation to be dated, including in planets (dif-
ferentiation ofthe core).We shall use this scheme lateron.

Niobium ^Zirconium 92Nb^92Zr,  ¼ 36 Ma. This newcomer to the ‘‘club’’ of forms of

extinctradioactivity,discoveredinZurichby Schonbachleretal.(2002),isyettobe exploited.
54 The principles of radioactive dating

147Sm half-life by year of publication

2.0

T (1011yr)
1.6 T = 1.06 ± 0.01 x 1011yr

1.2

0.8
1930 1940 1950 1960 1970

Figure 2.8 Improvement in determination of decay constants over time: the example of 147Sm.

Chlorine ^Sulfur 36Cl ^98% 36Ar / 2% 36S,  ¼ 0.43 Ma was discovered recently by a
Chinese team, Lin etal. (2005). No anomalyon 36Ar was found, onlyon 36S.

Beryllium ^Boron 10Be ^10B, T ¼ 15 Ma was discovered by McKeegan, Chaussidon,

and Robert (2000).

Lead ^Thallium 205Pb ^205Tl, T ¼ 15.1 Ma was recently discovered by Neilsen,

Rehka« mper, and Halliday (2006).
Others have notbeen con¢rmed and are not listed here.

2.5 Determining geologically useful radioactive

decay constants
As just seen, what allows us to calculate age and is the very essence ofthe radioactive clock is
the radioactive constant l, namely the probability that a nucleus will decay. How can this be
determined? This is di⁄cult a priori, given that the constants are generally very small
because activity is low (see Figure 2.8). To simplify there are three methods more or less
derived from the dating method:
(1) direct measurementofradioactivityby lNactivity: ifwe know N, we can deduce l;
(2) measurement of accumulation of the daughter isotope (both these series of measure-
ments are done in the laboratory);
(3) geological‘‘comparison’’ofages obtained by various methods.
We shall examine these three techniques in the case of 87Rb.

2.5.1 Measurement of activity

We start with the fundamental equation describing decay
d 87Rb
dt ¼ l87Rb;

that is, the numberof  particles emitted by unittime is equal to l 87Rb.

Let us take 1kg of 87Rb, which corresponds to 103 =87  6:023  1023g (6.23  1023 is
Avogadro’s number), or 6.92  1024 atoms of 87Rb. If l ¼ 1.42  1011 yr1, the number of
 particles emitted in 1 year is 6.92  10241.41  1011, or 9.8264  1013 particles.
55 Useful radioactive decay constants

Remembering that 1 year  3  107 seconds, that corresponds to 3.275  10 6 disintegrations

per second, which is a measurable value: even10 g ofpure rubidium will su⁄ce.
The di⁄culty lies in measuring   particles. Some of these particles are absorbed by the
rubidium deposit itself. It is fundamental then to make layers of rubidium ofvariable thick-
nesses andto correct whatisknown as auto-absorption.These are tricky methods to master.

2.5.2 The radiogenic isotope produced

For 87Rb we try to measure the 87Sr accumulated. Take 1kg of pure 87Rb. How much 87Sr
does it produce in1year?
87
Sr ¼ 103 g1:42  1011 ¼ 1:42  108 g;
that is14.2 ng. As strange as it mayseem, such a quantity can easilybe measured with a mass
spectrometer by isotope dilution.
Naturally, in practice we try to useboth methods and to compare the results.The question
is, ofcourse, howdowe obtain1kgofpure oralmost pure 87Rb? Isotope separation is expen-
sive sowetry rather touse natural rubidium, inwhich there is onlya fraction of 87Rb.The dif-
¢culty is that there must be very little 87Sr in the rubidium being measured. The rubidium
must therefore be very carefully puri¢ed by chemical methods, which is di⁄cult for such a
large amountofrubidium.The measurement uncertaintystems from this.

2.5.3 The method of geological comparison

If we know the age of certain rocks from methods with decay constants that are easier to
determine (such as uranium) we can then calculate the constant lRb by measuring the
87
Sr/87Rb ratios on a series ofrocks or minerals whose (U/Pb) age is known.
This method too is di⁄cult to implement as we mustbe sure that the various systems have
remained closed, as we shall see in the next chapter. Even so, the method is essential for
ensuring the geological reliability of the di¡erent methods.To avoid geological di⁄culties,
much use is made of cross-calibrations with meteorites and moon rocks.Why so? Because
meteorites are rocks dating from the origin ofthe Solar System (and therefore old) and have
not been subjected to major ‘‘disruptive’’events.We proceed by trial and error combining
the di¡erent approaches. An international commission makes regular reviews and updates
the constants as need be.
Let us give three important geological comparisons that use the constants given in the
table.

Moon rocks
We chooseRock10072,which hascomeinforparticularlyclosescrutiny.AstheU/Pbratioislow
this dating method is notgoodbutitis agoodwayofcomparing Rb^Sr, K^Ar, and Sm^Nd.

Dating method

Rock10072 K ^ Ar Rb^ Sr Sm ^ Nd

Time (Ga) 3.520.4 3.570.05 3.570.03

56 The principles of radioactive dating

Eucrites (basaltic achondrites)

These are basalt meteorites, in other words ancient extraterrestrial lava £ows. Dating by
U ^ Pb is extremely precise because the U/Pb ratios are high. As said, the uranium decay
constants are the references. However, Rb ^ Sr is not very precise as Rb/Sr ratios are really
low. (We are anticipating a little on the next chapter.) The K ^ Ar, Lu ^ Hf, and Sm ^ Nd dat-
ings are also relatively precise.

Dating method

Eucrite U ^ Pb Rb ^ Sr K ^ Ar Sm ^ Nd Lu ^ Hf
meteorites

Time (Ga) 4.550.05 4.500.14 4.500.1 4.530.04 4.570.19

Ordinarychondrites
These are the most common meteorites characterized by the presence of chondrules.
Comparison of Pb ^ Pb and Rb ^ Sr datings is very precise. Similarly, K ^ Ar and Re ^ Os are
in suitable agreement.

Dating method

Ordinary chondrites U ^ Pb Rb ^ Sr K ^ Ar Re ^ Os

Time (Ga) 4.550.05 4.550.08 4.520.05 4.540.02

Table 2.3 Comparison of methods for determining radioactive decay constants

Isotope Laboratory counting Accumulation Geological comparison

238
U, 235U, 232Th, This is the reference Not used This is the reference for
and radioactive method and is (relatively) other methods
chains precise
87
Rb Di⁄cultbecause ofauto- Di⁄cultto show up Comparison between
absorption of  raysby because oftraces of initial methods and in
87
the powder being Sr, which is hard to particular between
counted remove meteorites is essential
187
Re Impossible: insu⁄cient This is thebest method; Comparison with other
 energy (high auto- we measure accumulated methods and particularly
187
absorption) Os meteorites remains
essential
176
Lu Poor determination Di⁄cultbecause ofthe Comparisonsbetween
importance of initial Hf methods are useful
147
Sm Poor determination Di⁄cultbecause of Comparisonsbetween
initial Nd, which is hard methods are useful
to remove
40
K Very precise counting Should be possible in a Comparison is di⁄cult
well-sealed £ask because 40Ar di¡uses
readily
57 Problems

2.5.4 Conclusions
Laboratory determination of uranium and thorium constants by counting is the most pre-
cise method.These decay constants are then taken as references forother measurements.
Table 2.3 shows the pros and cons of the di¡erent techniques for determining decay con-
stants for the di¡erent chronometers.

Problems
1 Given that the potassium content of the silicate Earth is 250 ppm,3 how much 40Ar is created in
4.5  109 years? 40K ¼ 1.16  104 Ktotal.
Given that 40Ar cannot escape from the Earth and the quantity of 40Ar in the atmosphere
is 66  1018 g, what conclusion do you draw?

2 Suppose a series of zircons gives the results in the table below. Calculate the 206Pb/238U,
207
Pb/235U, 207Pb/206Pb, and 208Pb/232Th ages. Which ages appear most reliable to you? If we
know that the samples must be of the same geological age, which age would you recommend?

206
U (ppm) Th (ppm) Radiogenic Pb (ppm) Pb/204Pb 207
Pb/204Pb 208
Pb/204Pb

415 86.4 30.5 1138 84.34 116.2

419 84.0 30.4 1984 130.7 162.5
482 85.0 32.6 2292 147.5 174.1
507 81.8 34.0 3301 206.0 229.4

3 The K–Ar ages of two volcanic rocks from the island of Santa Maria in the Azores are measured
as below.

Rock mass (g) K2O% Radiogenic 40Ar (1012 moles g1)

1.81 1.83 13.90

0.64 0.92 5.60

Calculate their ages. Given that the error is  5%, what can you say about these two rocks?

4 To apply the 230Th–238U method to carbonate rocks formed in the ocean, allowance must be
made for the fact that (234U/238U)0 ¼ 1.15 in activity and that 230Th derives entirely from 234U
decay.
(i) Draw up the complete 234U/238U dating equation.
(ii) Draw up the 230Th, 234U, 238U dating equation assuming that (230Th)0 ¼ 0.
5 The isotopes231Pa and 230Th are both insoluble and isolated while uranium isotopes remain in
solution.
(i) Draw up the dating equation based on the 231Pa/230Th ratio.
(ii) What is the time-span over which it can be applied?

3
Previously we used 210 ppm. Readers should be aware of variations in values used by different authors in
such determinations.