The following response was sent to me by email on April 13, 2004. I am posting it because of the useful information it contains and because of its courteous tone. If, of course, the rate of decay were faster in the past, then fission track dates would still be far too old, despite these comments. Also, one might question how one can know for sure that the assumptions of this method are correct without knowing the real ages of the samples.
Recently I found your text on “The radiometric Dating Deception (on http://www.cs.unc.edu/~plaisted/ce/deception.html). Two things attracted me. The first, I am a Christian, the second, I am a fission tracker. Being a Christian has not kept me away from studying geology nor has it stopped me to get involved into radiometric age dating. My personal experience is in contrast to the content of the text on the cited web page that emphasizes the many errors that are said to occur when producing radiometric age data. I agree that in my career (covering about 400 age data) I have also had some ages that caused some headaches. However, from these, some turned out to contain very specific age information that forced us to reject a preliminary interpretation of the age data set. Some of the ages, less than 1%, however, turned out to remain strange (they are “wrong” by a factor of up to 2), and we do not know the reason for the difference from an expected value yet. As should be obvious from my explanation above, we hardly ever produce isolated values, but they commonly are in context to other age data, which provides a certain control on what is an “exceptional” age and what is “main stream”.
I would call myself an expert on fission track dating. I have used this method for more than 10 years and applied it successfully and in connection with other dating techniques on samples from Europe and Asia. Therefore, please allow to make some comments to the content of your web page. My comments will be restricted to the topic of the fission track method, to which you give more credit than to other methods (why?). I have also used other dating methods myself, but I think many of the comments to the fission track method could in part be used for those other methods, too.
“The next candidate dating method is fission track dating. Some minerals contain uranium 238 which decays by fission. It splits in two, and the pieces fly apart through the mineral, creating fission tracks. These tracks can be made visible by etching with an acid solution, and then counted. By knowing how much uranium 238 there is in a rock and by counting the number of fission tracks, one can measure the age of the rock.”
Yes, absolutely correct. However, this is only part of the story. It is important to realize that fission tracks in minerals have a specific length when they form, and this length is revealed by etching crystals on a grain-internal surface. When the etching fluid enters the crystal along tracks, cleavage planes or cracks, it happens from time to time that below the surface, but parallel to it, a track is revealed that we can observe in its full length. We call those tracks horizontal confined tracks. When we measure the length of those horizontal confined tracks in a sample (normally, 100 measurements per sample), we gain very important information about the thermal history of the sample after the onset of track accumulation. Be aware that the number of tracks will not provide an answer about how old a grain is or how old the rock is the grain was separated from. But: fission tracks are thermally unstable at high temperatures, and will only be accumulated when cooled below a certain temperature. If we heat the sample afterwards, tracks will be shortened and the age of the sample will become younger. The clue, however, is that we see the shortening of the tracks when measuring track lengths in that specific sample, which is routinely done. Thus, the take home messages: 1. Fission track ages are not formation ages, but are cooling ages. 2. The fission track method produces two data sets, which are closely connected to each other, ages and track length distributions.
"There are a number of problems with this method, and even geologists have had intense disagreements about its reliability. The ages often do not agree with what geologists expect. One problem is that certain constants involved in this method are not known or are hard to estimate, so they are calibrated based on the "known" ages of other rocks. If these other "known" ages are in error, then fission track dates are in error by the same amount."
Fact is that there is a growing community of fission trackers, which proves that the method has become very popular among geologists during the last few decades. If one realizes that the fission track age does not allow to determine the age of a rock (unless this rock cooled very rapidly, e.g. during volcanic eruptions or at impact events), but only the age of the rock cooling through a certain temperature range, there is a clear answer that we can give with this method. It is true that the fission track method relies on other dating techniques, but we are on our way to establish an absolute approach. Be also aware that the method started as an independent method in the 1960ies, and that already at that time people produced a lot of data which were much too young than expected from dating with other methods. In review, most of these discrepancies could be explained by the two points listed above, i.e. that the fission track method does not produce formation ages and that the reliability of an age can be controlled by looking at the track lengths. Routine track length measurements were only introduced in the 1980ies.
You are mentioning that “certain constants involved in this method are not known or are hard to estimate”. The age calculation formula contains several parameters, some of which are known with high precision such as e.g. the alpha-decay constant of uranium 238, some others may include larger errors. The most tricky parameter is the fission decay constant of uranium 238, which is the only relevant nuclide for the production of fission tracks. If one takes all physical estimates on this constant from the literature, they vary by +/- 20%. The fission track community has carefully evaluated among those existing values and among the applied methods and finally has agreed on a value that is intermediate to all estimates and was measured by a very precise method with a small potential of flaws. Be aware that the change of this constant may have a distinct influence on the age, however, only in the range of 20% and not of orders of magnitude. If one knows that fission track ages generally have a precision of +/- 5-10%, this is acceptable.
"Another problem is that fission tracks fade at high temperatures. So if there are too few tracks, the geologist can always say that most of them faded away. To get a fission track date, one has to know something about the temperature history of a rock."
I assume that this problem is answered by my explanations above. Track fading can be controlled by looking at the track lengths in a sample. There are, in addition, computer programs based on Monte Carlo algorithm that use age and track length information to generate a pattern of time-temperature paths in order to find out whether a certain data set allows one or more possible solutions (among which, hopefully only one fits with the geologic boundary conditions). You may argue that this sounds like a self-supporting system, however, tracker very often have their sample numbers encoded in order to avoid any misleading expectations while counting, and the modelling of the data is commonly done on the basis of a minimum of very simple assumptions in order to give the system the maximum freedom to find the right solution.
"Another problem is that uranium 238 can be removed from a rock by water. If a sample loses 99 percent of its uranium, then the fission track date will be 100 times too old. In fact, if a rock loses only about 1/350 of its uranium each year, then in 4000 years only one part in one hundred thousand of the uranium will remain, meaning that the date can approach a hundred thousand times too old. Now, 1/350 of the uranium each year is not much, especially when you consider that water occurs practically everywhere in the earth below a few hundred feet, and rocks shallower than this also become wet due to rainfall filtering down through the soil."
U loss, either by diffusion or dissolution is a process that does start from the surface of the grain we are looking at. Sometimes the grain surface is enlarged by cracks or cleavage planes that can be used as pathways for water to enter the crystal and by U to leave the crystal. However, on an internal surface of a grain, such a process should be visible by a gradient of U from rim to centre (zoning in U content). Zoning is a very common feature in some of the minerals. However, zoning by loss of U would lead to a U poor rim and a U rich core. The zoning that we observe is different from such a zoning and cannot be explained by U loss. Furthermore, the presence of complicate zoning patterns strongly suggests that U loss by U diffusion is not an important process, at least of no importance over the time range since grain formation. There are studies around that show hydrothermally altered zircon crystals, and these processes have an influence on the mineral composition of the grains, but it is also evident from such studies that the effects on the zircon crystal have a much stronger impact on major elements than on trace elements such as U. In general, crystalline matter still is the strongest way of binding elements.
"Another problem is knowing what is a fission track and what is just an imperfection in the rock. Geologists themselves suggest that imperfections are at times mistaken for fission tracks, and admit that fission tracks are not always easy to recognize. Textbooks have beautiful, clean pictures of fission tracks, but I doubt that these illustrations correspond to reality."
Yes, this is a good argument. I know some samples, where you are in difficulties in finding a single grain under the microscope with clear tracks and not full of other features that may be similar to tracks. There are, however, some good arguments to reject most of these features: First, tracks are straight, never curved. Second, they always show characteristic ends that show a strong younging of the etched track. Third, tracks are always randomly oriented. Never touch a grain with more than three tracks showing exactly in the same direction. Fourth, tracks have a characteristic range and a characteristic colour (these are just hints, some other inclusions are too big, some too small, others have the wrong light interference pattern). This helps a lot.
Nevertheless, the fission trackers have soon realized that different fission trackers count differently. Accordingly they introduced a calibration factor, which is established when counting age standards, which were dated by a range of other radiometric dating methods. In this context, it is important to realize that these age standards are over and over re-established. Go and check in the literature by looking for “fish Canyon tuff” and “Durango” to only mention two of the most important age standards for fission tracks. These two localities are still the subject of intense research, however, not discussing about order of magnitude, but discussion slight differences, in the range of 20% or less.
"Along this line, it is interesting to note that for every fission of uranium 238, there are over a million decays by a process called alpha decay, in which a helium nucleus is ejected from the nucleus of uranium. The alpha particle creates a long, thin trail of damage, and the former uranium nucleus recoils in the other direction, creating a short, wide track about one thousandth as long as a fission track. Not only this, but what's left of the uranium nucleus (having lost the helium nucleus) decays by thirteen more steps until it becomes lead, so there are over fourteen million other decays for every fission track. Over four million of these occur within a few days. All of these decays emit particles that damage the crystal structure. Some of these decays emit alpha particles, and some emit beta particles, which are energetic electrons. In addition, many millions of gamma rays are emitted, which are high-energy electromagnetic radiation like X rays, and also damage the crystal structure. Perhaps the damage created by all this radiation can be increased by chemical action and be etched by acid to appear like fission tracks. Or if two alpha particle trails are close enough together, perhaps they can damage the crystal enough so that their combined trail will be etched away by acid like a fission track."
There is abundant research on the impact of the different ways of radiation damage in the mineral zircon and, somewhat less, in the mineral apatite, the two minerals that are most frequently used for fission track dating. It is important to realize that the size of the different damage trails generally differ by several orders of magnitude. You are proposing that two alpha particle tracks may damage a crystal enough to produce a trail similar to a fission track if etched. This is similar to proposing that, seen from an airplane, two parallel little trails next to each other in a meadow look like a highway… Alpha particle trails are unlike fission track not trails of continued damage, but they are isolated damages along a trail that cannot be made visible by etching. You are listing big numbers of trails, however, think about the orders of magnitude of single atoms that effectively can be dislocated by radiation, and you will find out that there still is orders of magnitude more of such positions in a crystal lattice.
"Minerals are also subject to alteration by water, which may contain chemicals that react with the rock. Over long periods of time, all of these processes can damage the crystal structure, and it may be that when the mineral is etched with acid, track-like formations appear as a result."
You are right by stating that with time damage is accumulated. When a mineral grows, a large number of dislocations and crystal lattice defects are incorporated without any radiation or radioactive decay at all. My experience tells me that only a very small part of those defects can be mixed up with tracks (see above). I agree with you that over long periods of time, a lot can happen to those crystals, but all important processes leave their traces behind. I have dated grains from hard rocks with a lot of porosity and from unconsolidated sands and very strongly weathered rocks. It is surprising, however, how fresh the grains normally are, even though the rock they were separated from was strongly weathered. Be aware that the most frequently used minerals, apatite and zircon, are well known to be very stable and hard to destroy mechanically and by chemical treatment. It is, I confess, also known that the grain of apatites in old road cuts is zero, simply because the apatite grains are etched away by the organic acids in a tropical climate. But also in this case, grain remnants still show the correct number of tracks if compared with samples taken from nearby from fresh roadcuts. Thus, there is overall good evidence that alteration of apatite and zircon may only occasionally be important, but the traces of alteration are easy to see, and have no direct influence of the U in the crystal.
"Another problem is that fission tracks in some minerals, like zircons, can survive in lava, so the fission track date can be measuring an older event than the lava flow. Thus we cannot necessarily use this method to date the age of the fossils."
As explained above, fission tracks are subject to annealing when exposed to high temperatures. The annealing is a diffusion-controlled process, i.e. strongly time and temperature dependent. This means that e.g. for zircon an exposure to 700 °C (this would be a very cold lava), fission tracks would undergo only slight shortening if exposed only during five minutes. Exposed to a temperature of 400 °C for 100,000 years, however, would wipe away any fission tracks. Coming back to your example: If a zircon bearing sample falls into a rather cold granitic melt, and five minutes later is ejected of the melt, it may well happen that the fission track signal of the sample is only slightly disturbed. If the zircon sample, however, swims within the lava downhill during 3 hours, all fission tracks are gone and the clock reset to zero. We do not know what age will be correct unless we look at the track length distribution. In the five minute heating sample, many old tracks will be shortened, thus will be shorter than the tracks formed afterwards. In the 3 hour-sample, we will only find long tracks, in agreement with the most probably fast cooling of the lava at the earth’s surface.
"I think fission track dating has more potential than the other methods, but in view of all of these problems, I think we'll have to discard fission track dating as a reliable method."
Thanks for the compliments, but I disagree. And I have reasons to believe that my approach has been as critical as yours. Do not hesitate to ask for more details in case that my comments raised unanswered questions. Please accept also that I am not interested in a general discussion on “creation against evolution”. I would only be ready to discuss those topics I am well informed about, but will not discuss topics that are beyond my scientific horizon. In earth sciences, we love to make suggestions to other colleagues, but always respect their expert’s opinion.
Back to home page.