From:    Henry Shaw
To:      David Myers                              09-Jun-89 03:37pm
Subject: Re: IQ ETC.

 DM> I heard a creationist once claim that potassium-argon dating could
 DM> give misleadingly large dates because of argon penetrating rock.
 DM> An acquaintance of mine countered by saying that the argon
 DM> formed from potassium isn't the same isotope, and that leakage would
 DM> age UNDER-estimates.  true?

K-Ar dating is based on the decay of K-40 to Ar-40 by electron-capture.
K-40 can also decay to Ca-40 via beta decay.  About 12% of the time, a K-40
nucleus will decay to Ar-40 and 88% of the time, to Ca-40.  The K-Ca isotopic
system has only limited application for various technical reasons.

In traditional K-Ar dating, one crushes a mineral or rock sample and splits
the powder into two fractions.  One measures the K concentration in one
"split", usually by atomic emission or absorption spectrometry.  The other
split is placed into an evacuated tube and heated to temperatures exceeding
1400 deg C using an RF furnace.  At these temperatures, most rocks will melt,
and the Ar (and other gasses) are driven off from the sample.  It is important
to note that Ar is a noble gas, and does not readily form compounds.

One then adds a known quantity of Ar gas made up of a single isotope of Ar,
usually Ar-38, to the gas given off by the sample.  This is known as
"spiking",
and it allows one to calculate the absolute amount of Ar released by the
sample.  The resulting mixture is then purified to remove other gasses, and
the isotopic composition of the purified Ar is then measured using a mass
spectrometer.

Now even though Ar is not a proper constituent of minerals, is it possible
for some Ar to be trapped by the mineral when it forms, either in tiny
bubbles known as fluid inclusions, or in defects in the crystal structure, or
simply adsorbed on the surfaces of the sample powder.  In addition, the
various
vacuum systems used to extract the Ar are never perfect, and always contain
a little air (and therefore atmospheric Ar).  Atmospheric Ar contains other
stable isotopes of Ar besides Ar-40 (the only Ar isotope produced by the
decay of K-40).  It's isotopic composition is: Ar-36, 0.34%; Ar-38, 0.06%; and
Ar-40, 99.6%.  By measuring the amount of Ar-36 in the sample, and assuming
that the only Ar initially in the mineral, or that might have contaminated
the sample during later handling has an atmospheric composition, one can
subtract off the "atmospheric contribution" to the amount of Ar-40 measured,
and determine the amount of Ar-40 produced by radioactive decay.  It is then
a simple matter to calculate the age of the sample.

This method has four assumptions:  (1) that the system has been closed with
respect to K since it formed; (2) that the only Ar initially present in the
sample when it formed had atmospheric composition; (3) that no Ar has left
the sample since it formed; and (4) that only atmospheric Ar may have been
added to the sample at any time.  (One also takes for granted that the
laws of physics and chemistry have not changed in unknown ways over time.)

The consequences of each of these assumptions being violated are discussed
below.  Before doing this, it is important to understand that although
K is a common constituent of minerals, forming chemical bonds with the
surrounding atoms, when a K-40 atom decays to Ar-40, the crystal suddenly
finds itself with a Ar atom where a K atom should be.  This is an
energetically unfavorable situation and the crystal would "like" to
heal itself by getting rid of the Ar atom.  In order to do this, the atoms
in the crystal must reorganize themselves (diffuse).  At low temperatures,
this process is exceedingly slow, however at elevated temperatures, diffusion
can become more rapid.  Typically, the rate of diffusion increases
exponentially with 1/T (D = D0 * exp( -E/RT), where D0, E and R are
constants).
For typical minerals used in dating, the loss of Ar due to diffusion is
negligible below about 300 deg C, and can be nearly complete above that
temperature.

1) System closed with respect to K. -- Under some conditions, a K-containing
mineral can gain or lose potasium, either by exchanging some other alkali,
like Na, for the K or by recrystallizing.  If one could remove K without losing any
Ar, then this would result in the calculation of an incorrectly old age, if
one adds K without changing the Ar content, the the age would be too young.

However, since K is an integral part of mineral structures, while Ar is not,
it is usually impossible to gain or lose K without completely removing all the
Ar from the mineral.  When this happens, the isotopic clock is reset to the
time at which this disturbance occurred, and the age one calculates is not
that
for the original formation of the mineral, but the YOUNGER age corresponding
to the time at which the system was disturbed.

2 and 4) Only atmospheric Ar initially, or added later. --  This is a
relatively uncommon problem, however, consider the case in which a rock
(composed of a number of minerals) is buried deep in the earth's crust,
where it is heated up.  The minerals will start to undergo solid-state
reactions, converting into a new assemblage of minerals that are
stable at the higher temperature and pressure conditions present at
depth.  This process is known as metamorphism, and metamorphic
rocks form one of the three major classes of rocks.  As the minerals react,
they will release any trapped Ar produced by K-40 decay.  The Ar present in
this environment (i.e., as an interstitial fluid between the mineral grains)
is thus richer in Ar-40 than is atmospheric Ar, and may approach pure Ar-40 in
composition.  If a new mineral forms and entraps some of this Ar, then the new
minerals will have an "excess" of initial Ar-40.  If one trys to date this
sample, it will yield an anomalously OLD age.  Though rare, this phenomenon
has been documented in a number of cases.  Fortunately, there is a newer
K-Ar dating technique that can identify such cases and correctly account for
the isotopic composition of the initial Ar in a sample.  This technique is
called the Ar-40/Ar-39 method, the details of which would add another few
pages to this already overlong message.

3)  No Ar escape from the sample since it formed. -- This is by far the most
common problem with K-Ar dating.  As I mentioned before, at temperatures above
a few hundred degrees C, Ar diffuses out of minerals fairly rapidly.  Because
of the exponential dependence of diffusion rate on temperature, the
temperature range over which a given mineral goes from being "closed" to
"open" with respect to Ar diffusion is very small.  There is thus a "closure
temperature" for a given mineral that depends on the characteristics of that
mineral, below which, Ar is quantitatively retained and above which, Ar is
lost.  The extent of Ar-loss is dependent on the details of the
time-temperature history of the sample.  Loss of Ar without a corresponding
loss of K, will, of course, result in incorrectly YOUNG ages.  Again, the
Ar-40/Ar-39 dating method offers significant advantages in detecting, and
sometimes being able to "see through" the effect of the loss.

Because the closure temperatures for Ar in most minerals are relatively low
(on the order of 300 deg C), the K-Ar system is the most easily "reset" of
all the isotopic clocks used in dating rocks.  For that reason, it is not
usually used to date ancient rocks.  It is most useful for dating either
the time of metamorphism or dating rocks that have had simple geologic
histories and have never been deeply buried.
-!- TBBS v2.1/NM
 ! Origin: Diablo Valley PCUG-BBS, Walnut Creek, CA <415-943-6238> (1:161/55)