Several specific questions about RadioCarbon Dating using AMS

Hi PhiloNibbler,

I'll start with simple answers and more details can be filled in as necessary.

First, about the age of your references, they date from the 1980's and 1990's, more than 20 years ago when use of AMS (Accelerator Mass Spectrometry) for radiocarbon dating was still being researched and refined. More recent research about background measurement levels does seem to exist, but not too much as this issue has long been considered settled. Research would only resume if substantive questions were raised.

Your discussion and excerpts mention counting ¹⁴C atoms but don't mention that ¹²C and ¹³C atoms are also counted, so since I can't be sure you understand that radiocarbon dating actually measures a ratio, let me briefly describe it. Just counting ¹⁴C atoms won't provide a date. Counting ¹²C is also necessary, and the ratio of ¹⁴C atoms to ¹²C atoms yields the date. The ratio of ¹⁴C atoms to ¹³C can also be used (the levels of both ¹²C atoms and ¹³C in the atmosphere are relatively constant).

PhiloNibbler writes: Now I read this as saying that they ran the test without a sample and measured an age of 61000 yrs age. And as I understand the process the tantalum wire is normally used to hold the graphite in the sample holder but in this case it's bare or the sample chamber is empty. Right? And the machine background measured is just stray carbon molecules that are stuck in the AMS?

This background measurement is very unlikely to come from "stray carbon molecules that are stuck in the AMS." They aren't specific about the cause, but it likely was just how their machine behaved. Their counts of ¹⁴C atoms are evidently in some way dependent upon the beam current, and what they were saying is that the beam current never goes to zero, even when no carbon sample is present. Keep in mind that they were working more than 30 years ago with the technology available in 1984.

The reason this background measurement is unlikely to stem from "stray carbon molecules" is because those carbon molecules would most likely be carbon dioxide from the air. The carbon in the air is constantly replenished with ¹⁴C and so has high levels of ¹⁴C compared to old carbon in ancient biological material. Hence carbon dioxide would produce a relatively high ¹⁴C count relative to the ¹²C count and date to zero years old, not 61,000 years old. The more ¹⁴C relative to ¹²C the younger the sample.

My only question here is what is meant by "with the ion source closed off from the remainder of the beam line, ≥104 ka (0 counts in 20.5 min of counting). With the ion source containing an empty aluminum target holder open to the beam transport system, a 14C count rate equivalent to .009 pMC (74 ka) was measured" in laymens terms.

The "ion source" is the part of the AMS machine that turns the carbon molecules in the sample into ions (charged molecules). Closing off the "ion source" must mean that the connection of the "ion source" to the rest of the machine is in some way blocked. It's slightly different from having the "ion source" unblocked but with no sample in place.

If we give the label "new carbon" to carbon samples with recent interactive contact with the air, then new carbon contamination would cause samples to date newer, not older.

If we give the label "old carbon" to ancient carbon sources with no recent interactive contact with the air (e.g., long-buried carbon such as in coal that has no exposure to radioactivity), then old carbon contamination could cause samples to date older, but it would have to be a great deal of contamination. For example, a 5,000 year-old sample would require contamination larger than the sample itself to make it date to 50,000 years old. There's also the question of where so much stray old carbon would come from and yet not be detected during cleaning and calibration.

--Percy

PhiloNibbler writes: indicates that they were testing the machine for background readings using a known powder with low c14 concentration and received a total background of .03 ± .015 Pmc (percent modern Carbon) total due to electronic noise, stray particles, and possible introduction of (modern?) carbon14 during preparation of the sample. So the background here measured is mostly possible contamination and machine background from scattered particles?

We need someone more knowledgeable to chime in, but here's what I think is going on. The negative ions from the sample are accelerated to a moderate kinetic energy¹, and ions of different mass split into their own ion beams based upon atomic weight. This means the ¹²C, ¹³C and ¹⁴C ions are split into three separate beams. The recombinator uses a carefully designed magnetic field to recombine the ion beams into a single beam, which then enters the mass spectrometer portion of the machine.

Before the three beams are recombined the individual beams can be individually blocked, allowing researchers to determine the base level ion counts in the mass spectrometer for each carbon isotope. For example, by blocking the ¹²C ion beam they can determine how many ¹²C ions the mass spectrometer counts when the ion source is providing no ¹²C ions.

It bears repeating that by far the most abundant sources of environmental carbon are comparatively rich in ¹⁴C, and the more ¹⁴C reaching the mass spectrometer the younger a sample will date. This makes contamination an unlikely cause of very old carbon dates. They are most likely caused by very old samples.

--Percy

1: Mass recombinator for accelerator mass spectrometry

PhiloNibbler writes: Thin, by GRI do you mean Global Research Inititive? and SDA Seventh Day Adventists?

He's referring to the link to the GRI SDA (Geoscience Research Institute of the Seventh Day Adventists) that you provided in your Message 1:

PhiloNibbler in Message 1 writes: Specifically, I was reading paul giem's page http://www.grisda.org/origins/51006.htm

--Percy

I found I couldn't be certain what Coyote actually meant and would like to get more clarity. He seemed to be saying that 5900 BP and 9400 BP shells can exist at the same level and location at some archeological sites, but that doesn't seem possible without some inter-level contamination, so I think he must have meant something else. Maybe that you can't know a level's date range unless you are thorough in submitting samples for dating? Still, a date range for a level that spans 3500 years seems very extreme and unlikely. So I'm left uncertain what Coyote meant.

--Percy

Thanks for the reference to Bertsche's RATE’s Radiocarbon: Intrinsic or Contamination?. It fills in a few very useful details, which I attempt to summarize here. I occasionally add some additional information and my own comments.

Typical sample sizes are 1 mg of carbon. A poppy seed found in an ancient jar might weigh around 0.3 mg, and I'm guessing the carbon content would below 0.1 mg. Bertsche says that small an amount is manageable, but precision begins to "suffer below about 0.1 mg of carbon due to counting statistics."

Beta Analytic Radiocarbon Dating requests sample sizes of 20 mg because the pretreatment reduces the mass significantly.

Contamination in situ, during collection and during storage can be significant. This makes a lot of sense. The poppy seed in an ancient jar might have been isolated from modern carbon for millennia, but once opened modern carbon floods in. Carbon is everywhere, in the air, in people, in animals, in bacteria, and even in the sterile gloves an archaeologist might wear while collecting samples.

Contamination is measured in units of pMC (percent modern carbon). Apparently contamination is always assumed to be modern carbon (high ¹⁴C content). Samples will inevitably have widely varying pMC's. A sloppily gathered sample could easily contain more modern carbon than sample carbon (Coyote's comments about this possibility might be valuable).

Laboratory contamination contributions vary, but values around 0.1 pMC are typical

AMS instrument contamination contributions also vary, but values around 0.03 pMC are typical.

Bertsche's purpose is to critique John Baumgardner's RATE claims. Baumgardner reanalyzed "90 previously purblished radiocarbon AMS dates of old samples," and RATE gathered new samples and had them AMS dated. Baumgardner claims that all material, even non-biological, contains intrinsic radiocarbon. Bertsche's reference 1 and 2 are Baumgardner's papers, and I provide links here:

1. ¹⁴C Evidence for a Recent Global Flood and a Young Earth
   
2. MEASURABLE ¹⁴C in Fossilized Organic Materials: Confirming the Young Earth Creation-Flood Model

The abstract of reference 1 briefly describes Baumgarders results and conclusions, which I paraphrase here:

All organic samples from the Phanerozoic (500 million years ago until the present) contain "significant and reproducible amounts of ¹⁴C." The technical literature describes many instances of ¹⁴C/C ratios between 0.1 and 0.5 percent of the modern ratio. Because these levels are a factor of 10 above the threshold levels of most AMS labs they represent real ¹⁴C levels when none should be present according to uniformitarian principles.

Another problem is that the ratios are uncorrelated with their position in the geological record. The older the strata in which the sample is found, the less ¹⁴C should be present, yet there is no correlation of ¹⁴C levels with age of strata. This calls into question the uniformitarian assumption that samples from different geological levels were deposited in different eras.

Because these samples from ancient strata date to 44,000 to 57,000 years ago, they must all have been buried relatively contemporaneously only a few thousand years ago during a "global cataclysm."

In order for ¹⁴C levels to be so low in life only a few thousand years old, the ¹⁴C/C ratios must have been much lower then.

The most generous thing I can say about this is that Baumgardner is letting his flood assumptions affect his judgment, because he makes very simple and fundamental errors. If I weren't being generous then I would say that Baumgardner is purposefully misrepresenting the data knowing that fellow creationists will either ignore the misrepresentations or not even notice them.

Baumgardner says that 0.1 to 0.5 percent of the modern ¹⁴C/C ratio is a factor of 10 above AMS threshold levels. By "AMS threshold levels" he means of the AMS instrument itself and doesn't include sample contamination, so his claim that this says something meaningful about AMS reliability is wrong. The instrument threshold is an order of magnitude smaller than sample contamination and so is overwhelmed by it.

Betsche notes that Baumgardner improperly lumps geologic samples together with biologic samples when he draws conclusions about the lack of correlation with geologic level. Baumgardner also ignores the fact that both coal and carbonates easily absorb atmospheric CO₂, providing extra ¹⁴C that skew the AMS results. He needs to treat such samples differently, but doesn't.

I've left out a lot from Bertsche's conclusions, but here's the bottom line: There is no evidence that “all carbon in the earth contains a detectable and reproducible ... level of ¹⁴C.”

--Percy

Edited by Percy, : Fix typos.

Okay, thanks, I think I get it. If you take a multi-level site and attempt to date all the different levels using a poorly selected sample array, say only abalone, then you'll get an incomplete and potentially erroneous impression.

--Percy

PhiloNibbler writes: Aside from Baumgarder's motives...

I should be a bit more detailed about how I really feel about Baumgardner's stance. No one invests a lifetime in a cause they don't sincerely believe in, and so Baumgardner must believe the Earth is only a few thousand years old with all his heart. It's difficult to imagine the machinations his mind must go through to write things like this from ¹⁴C Evidence for a Recent Global Flood and a Young Earth:

quote:

---

After one million years (175 half-lives), the amount of ¹⁴C remaining is only 3 × 10^-53 of the initial 14C concentration—so vanishingly small as to exclude even a single ^14C atom in a beginning mass of ¹⁴C equal to the mass of the earth itself. However, in samples with uniformitarian ages between one and 500 million years, the peer-reviewed radiocarbon literature documents scores of examples of ¹⁴C/C ratios in the range 0.1–0.5 percent of the modern ¹⁴C/C ratio. The lower limit of this range is a factor of ten above the detection threshold of most AMS laboratories in the world.

---

The obvious and blatant error is that he's ignoring sample contamination. He's making an argument that assumes that every bit of ¹⁴C above the instrument threshold level was actually from the sample and not contamination.

Baumgardner is a brilliant scientist. Even brilliant scientists can be wrong, even profoundly wrong, but not about things as basic and simple as contamination. He's declaring, "Contamination outside the instrument doesn't exist," when the reality is that sample contamination is at least an order of magnitude larger than instrument contamination.

How does Baumgardner reach such a dunderheaded conclusion? It appears he trusts someone else's dunderheaded conclusion, though on some level Baumgardner must know it's wrong. He relies upon Paul Giem's paper Carbon-14 Content of Fossil Carbon. About Giem's paper Baumgardner says on page 590:

quote:

---

He further showed contamination of the ¹⁴C-bearing fossil material in situ was unlikely, but theoretically possible, and was a testable hypothesis." While contamination during sample preparation was a genuine problem, the literature showed it could be reduced to low levels by proper laboratory procedures. He concluded the ¹⁴C detected in these samples most likely originates from the organisms themselves from which the samples are derived. Moreover, because most fossil carbon seems to have roughly the same ¹⁴C/C ratio, he deemed it a clear logical possibility that all these fossil organisms had lived together on earth at the same time.

---

A reading of Giem's paper reveals that he understands the possible sources of contamination, but after many words he unjustifiably declares that the contamination problem is solvable (i.e., reduce contamination to indetectable levels) and that therefore contamination can be ignored. But he did no dating himself. He was reviewing papers about dating done by others where they described the contamination problem and definitely did not think it had been solved, as Bertsche's discussion makes clear.

So is that why Baumgardner makes the errors he does, misplaced trust? That's a lot of trust. If you were a chemist, would you believe a paper by your most beloved collaborator declaring that dust in beakers doesn't affect experiments? Of course not. Baumgardner has no excuse for accepting such specious declarations about sample contamination. It's inexplicable.

--Percy