This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Geol. 655 Isotope Geochemistry
Lecture 9 Spring 2003 GEOCHRONOLOGY V
THE U-TH-PB SYSTEM: ZIRCON DATING
Zircon (ZrSiO4) is a mineral with a number of properties that make it extremely useful for geochronologists. First of all, it is very hard (hardness 71/ 2), which means it extremely resistant to mechanical weathering. Second, it is extremely resistant to chemical weathering and metamorphism.
For geochronological purposes, these properties mean it is likely to remain a closed system. Third, i t
concentrates U (and Th to a lesser extent) and excludes Pb, resulting in typically very high 238U/204Pb
ratios. It is quite possibly nature's best clock. Finally, it i s
reasonably common as an accessory phase in a variety of
igneous a nd metamorphic
The very high 238U/204Pb r atios i n z ircon ( and s imilar
high µ m inerals s uch a s
sphere and a patite) provide
some special geochronological
opportunities and a s pecial
diagram, t he c oncordia d i a gram , has been developed to
take advantage of them. T he
discussion that follows can be
applied to any other system
U/204Pb ratios, but in p ractice, z ircons c onstitute t h e
principle t arget for Pb geochronologists.
A concordia diagram is simply a p lot o f 206Pb*/238U v s.
Pb*/235U. You should s atisfy
yourself that both of these r atios are proportional to t ime.
In essence, the concordia d iagram is a plot of the 238U–206Pb
age against the 235U–207Pb a ge.
The ‘concordia’ curve on such a
diagram that is the locus of
points where the 238U–206Pb a ge
equals the 235U–207Pb age. Such
ages are said to be c oncordant.
Figure 9.2 is an example of a
The best way to think about
evolution of Pb/U ratios is to
imagine that the diagram i t - Figure 1. Upper. Separated Zircon crystals. Notice the zoning.
self evolves with time, along Lower. Strongly zoned zircon showing differing ages of spots anawith its axes, w hile the a c- lyzed by ion probe.
55 February 6, 2003 Geol. 655 Isotope Geochemistry
Lecture 9 Spring 2003
1 4.5 Ga 0.8
238 U 4.0 Ga
3.5 Ga 0.6 3.0 Ga
20 40 60 80 100 207 Pb/235 U Figure 9.2. The concordia diagram.
tual data point stays fixed. Let’s take a 4.0 Ga old zircon as an example. When it first formed, or
“closed”, it would have plotted at the origin, because had anyone been around to analyze it, t hey
would have found the 207Pb*/235U and 206Pb*/238U ratios to be 0. I nitially, 207Pb*/235U would have i ncreased rapidly, while the 206Pb*/238U would have been increasing only slowly. This is because 4.0 G a
ago there was a lot of 235U around (recall that 235U has a short half-life). As time passed, the i ncrease in 207Pb*/235U would have slowed as the 235U was ‘used up’. So imagine that the diagram i nitially 'grows' or 'expands' to the left, expanding downward only slowly. Had someone been around
3.0 Ga ago to determine 'zircon' ages, he would have drawn it as it appears in Figure 9.3 (of course, h e
would have labeled the 3.0 Ga point as 0, the 4.0 Ga point as 1.0, etc.).
Any zircon that has remained as a completely closed system since its crystallization must plot on
the concordia line. What happens when a zircon gains or looses U or Pb? Let’s take the case Pb loss,
since that is the most common type of open-system behavior in zircons. The zircon must lose 207Pb and
Pb in exactly the proportions they exist in the zircon because the two are chemically identical. In
other words, a zircon w ill
not lose 206Pb in preference
to 207Pb or visa versa.
L et’s t ake t he s pecific
case of a 4.0 Ga zircon t h a t
experienced some Pb loss
during a metamorphic
event at 3.0 Ga. If the loss
was complete, the zircon
would have been reset and
would have plotted at t h e
origin in Figure 9.3. W e
cannot, of course, d istin0.1
guish a zircon completely
reset at 3.0 Ga from one
that c rystallized a t 3 .0
Ga, but suppose it lost only
half its Pb at t hat t ime.
207 Pb/235 U
During the Pb loss, t h e
Figure 9.3. A concordia diagram as it would have been drawn at 3.0 Ga.
Pb/238U a nd 207Pb/235U
56 February 6, 2003 Geol. 655 Isotope Geochemistry
00 Spring 2003
4.55 Ga a
4.0 Ga 1.2
0 206Pb/238U 206Pb/238U Lecture 9 3.0 Ga
10 20 30 40 50
207Pb/235U 60 70 b 4.0 Ga 4.55 Ga 3.0 Ga Crystallization
0 10 20 30 40 50 60 70 80 90
207Pb/235U Figure 9.4. (a) Concordia diagram as it would have appeared at 3.0 Ga. Three zircons that e xperience variable amounts of Pb loss move from the 4.0 Ga point on the concordia curve ( their
crystallization age) toward the origin. (b) The same three zircons as they would plot at present. The three define a cord between 3.0 Ga and 4.0 Ga. A possible interpretation of this result
would be that 4.0 Ga is the crystallization age and 3.0 Ga is the metamorphic age.
would have both decreased by half. Consequently, the point would have migrated half way to t h e
origin. At 3.0 Ga, therefore, it would have plotted on a ‘cord’, i.e., a straight line, between its i nitial
position on the concordia curve, the 4.0 Ga point, and the origin (Figure 9.4a) at 3.0 Ga. Had it l ost
some other amount of Pb, say 30% or 80%, it would have plotted on the same cord, but further or nearer
the origin. The line is straight because the loss of 207Pb is always directly proportional to the loss of
Pb. The origin in Figure 9.3a corresponds to the 3.0 Ga point on the concordia in Figure 9.4b. So, in
Figure 9.4b, the zircon would lie on a cord between the 4.0 Ga and the 3.0 Ga point. We would say t his
is a 'discordant' zircon.
The intercepts of this cord with the concordia give the ages of initial crystallization (4.0 Ga) and
metamorphism (3.0 Ga). So if we can determine the cord on which this discordant zircon lies, we can
determine the ages of both events. Unfortunately, if our only data point is this single zircon, we can
draw an infinite number of cords passing through this point, so the ages of crystallization and m etamorphism are indeterminate. However, we can draw only 1 line through 2 points. So by measuring
two zircons (or populations of zircons) that have the same crystallization ages and metamorphism
ages, but have lost different amounts of Pb, and hence plot on different points on the same cord, t h e
cord can be determined. The closure age and partial resetting ages can then be determined from the i nterecepts. (as usual in geochronology, however, we are reluctant to draw a line through only two
points since any two points define some line; so at
least three measurements are generally made). In
practice, different zircon populations are selected
based on size, appearance, magnetic properties,
color, etc. While zircon is generally a trace mineral, only very small quantities, a few m illigrams,
are needed for a measurement. Indeed, it is possible to analyze single zircons and even parts of z ircons.
U gain would affect the position of zircons on t h e
concordia diagram in the same manner as Pb loss;
the two processes are essentially indistinguishable on the concordia diagram. U loss, on the
other hand, moves the points away from the o ri- Figure 9.5. A concordia plot showing h ypothetigin at the time of the loss (Figure 9.5). In t his cal zircons that crystallized at 4.0 Ga and lost U
case, the zircons lie on an extension of a cord above during metamorphism at 3.0 Ga.
57 February 6, 2003 Geol. 655 Isotope Geochemistry
Lecture 9 Spring 2003 0.90 the concordia. As is the case
for Pb loss, the upper i ntercept of the cord gives the i niBanded Gneiss BGXM
tial age and the lower i nter3800
cept gives the age of U loss.
U loss in less common than Pb
loss. This is true for two r easons. First, U is happy in t h e
zircon, Pb is not. Second, Pb
will occupy a site damaged
by the alpha decay, making
diffusion out of this site e as0.60
ier. Radiation damage is a
significant problem in zircon
geochronology, and one of
the main reasons ages can be
50.0 imprecise. U-rich zircons are
particularly subject to r adia0.90
tion damage. Heavily damaged crystals are easily recognized under the microscope
Porphyritic Gneiss SP-405
4000 and are termed metamict.
Pb gain in zircons is not
predictable because the i so3600
topic composition of the Pb
gained need not be the same
as the composition of the Pb
in the zircon. Thus Pb gain
would destroy any age r elaEquant
tionships. However, Pb gain
is m uch l ess l ikely t han
other open system behaviors.
Zircons that have suffered
multiple e pisodes o f o pen
system behavior will h ave
U-Pb s ystematics t hat a re
difficult t o i nterpret a nd
Figure 9.6. Concordia diagrams showing ion probe Pb-U analyses of
could b e i ncorrectly i nterAcasta gneiss zircons. Size of point is proportional to 1 s a nalytical
preted. For example, zircons
uncertainty. Triangles are zircon analyses done by conventional mass
lying on a cord between 4.0
spectrometry. From Bowring, et al, 1989.
and 3.0 Ga that subsequently
lose Pb and move on a second cord toward the 2.0 Ga could be interpreted as having a metamorphic a ge
of 2.0 Ga and a crystallization age of between 4.0 and 3.0 Ga.
Continuous Pb loss from zircons can also complicate the task of interpretation. The reason is that in
continuous Pb loss, zircons do not define a straight line cord, but rather a slightly curved one. Again
imagining that the concordia diagram grows with time, a zircon loosing Pb will always move toward
the origin. However, the position of the origin relative to the position of the zircon moves with t ime
in a non-linear fashion. The result is a non-linear evolution of the isotopic composition of the zircon.
Given the mechanical and chemical stability of zircon, it should not be surprising that the oldest
terrestrial material yet identified is zircon. Until a decade ago, the oldest dated terrestrial rocks
were the Isua gneisses in Greenland. These are roughly 3850 Ma old. Work published in 1989, r evealed that the Acasta gneisses of the Slave Province (Northwest Territories, Canada) are 3.96 G a 206Pb/238U 4000 206Pb/238U . 58 February 6, 2003 Geol. 655 Isotope Geochemistry
Lecture 9 Spring 2003 old. These ages were determined using an ion probe to date the cores of zircon crystals extracted from
these gneisses. Concordia diagrams for these gneisses are shown in Figure 9.6.
Zircons having ages in the range of 4100-4260 Ma have been identified in quartzites at Mt. N arryer
and the Jack Hills in western Australia (e.g., Compston and Pidgeon, 1986). The quartzites t hemselves are metamorphosed sandstones that were probably deposited about 3100-3300 Ma. They contain zircons derived from a number of sources. A small fraction of these zircons has cores that are in
the range of 4100-4200 Ma. The zircons were analyzed by a specially built high resolution ion probe
at the Australian National University nicknamed ‘SHRIMP’. The great advantage of this instrument over conventional analysis of zircons is not only that individual zircons can be analyzed, but i ndividual parts of the zircons can be analyzed. The Mt. Narryer zircons have had complex h istories
suffering multiple metamorphic events between 4260 and 2600 Ma. The principle effect was t h e
growth of rims of new material on the older cores around 3500 Ma. Conventional analysis of these z ircons would not have recognized the older ages. The cores of these zircons, however, proved to be
nearly concordant at the older ages. These ages determined by ion probe were i nitially highly controversial. By and large, however, the community has come to accept them as reliable, when p erformed carefully.
Subsequently even older zircons (would be more correct to say parts of zircons), were discovered in
the Jack Hills of Australia. An ion probe date on one part of one of these zircons (Figure 9.6) is shown
4.404 Ga ±8 Ma. Thus the oldest known terrestrial materials are approaching the oldest ages from
other planetary bodies, including the Moon, Mars, and asteroids (as represented by meteorites). T hey
remain, however, significantly younger than the age of the Solar System, which is 4.556 Ga. N evertheless, these very old ages seem to demonstrate that it is zircons, not diamonds, that “are forever”. REFERENCES AND SUGGESTIONS FOR FURTHER READING
Bowring, S. A., I. S. Williams, and W. Compston, 3.96 Ga gneisses from the Slave province, Northwest Territories, Canada, Geology, 17: 971-975, 1989.
Compston, W. and R. T. Pidgeon, Jack Hills, evidence of more very old detrital zircons in Western
Australia, Nature, 321:766-769, 1986.
Wilde, S. A., J. W. Valley, W. H. Peck and C. M. Graham, Evidence from detrital zircons for the e xistence of continental crust and oceans on the Earth 4.4 Gyr ago, Nature, 409:175-178, 2001. 59 February 6, 2003 ...
View Full Document
This note was uploaded on 03/03/2012 for the course EAS 656 taught by Professor White during the Fall '10 term at Cornell University (Engineering School).
- Fall '10