When Willard Libby and colleagues developed radiocarbon dating in the late 1940s, they faced an experimental obstacle: No radiation detection tool was sensitive enough to detect carbon-14 at its expected environmental concentrations—about 1 out of every 1012 carbon atoms.
To overcome that obstacle, Libby and coworkers built a sample chamber surrounded by multiple Geiger counters and significant shielding. It removed some background signal and allowed them to calibrate for still more, thereby enabling them to pick out the 14C signal.
As the field of radiometric dating has expanded to include isotopes with longer half-lives, lower abundances, and trickier contaminants, so has the range of experimental challenges. Calcium-41 was identified in the late 1970s as a potentially useful radioactive tracer for studying biochemical and geochemical cycles because of its prevalence in both biological organisms and Earth’s crust. And because 41Ca has a longer half-life than 14C—nearly 100 000 years rather than around 5700 years—it could date older specimens. But its low natural baseline abundance, about 1 out of every 1015 Ca atoms, kept 41Ca dating out of experimental reach.
Now Tong-Yan Xia, Wei-Wei Sun, and colleagues in Zheng-Tian Lu’s group at the University of Science and Technology of China in Hefei have built an atom-trap trace analysis (ATTA) setup1 (see figure 1) that can measure 41Ca at abundances of less than 10−16. They’ve demonstrated their system’s ability to measure 41Ca in natural samples of granite, seawater, and bone, and they are now collaborating with glaciologists and other experts to test the isotope as a tracer for radiometric dating.
Wei-Wei Sun, a graduate student in the Laser Laboratory for Trace Analysis and Precision Measurements at the University of Science and Technology of China, adjusts the blue laser that slows and traps calcium atoms in the group’s atom-trapping setup. (Courtesy of Wei-Wei Sun.)
Wei-Wei Sun, a graduate student in the Laser Laboratory for Trace Analysis and Precision Measurements at the University of Science and Technology of China, adjusts the blue laser that slows and traps calcium atoms in the group’s atom-trapping setup. (Courtesy of Wei-Wei Sun.)
Isolating isotopes
A major disadvantage of using Geiger counters to measure a radioactive isotope’s abundance is that the atoms are detectable only when they decay. So it revolutionized the field when researchers started using accelerator mass spectrometry (AMS) for carbon dating2 in the 1970s. The technique uses a magnetic field to sort ions in a beam by mass, so researchers could measure the abundance of a particular isotope directly instead of waiting to detect a decay product. As a result, measurements take less time and can use much smaller samples.
A challenge for mass-spectrometry techniques is separating rare isotopes from common ones of the same mass. Radioactive 14C may be easily distinguished from 12C, but it’s basically indistinguishable from nitrogen-14. To separate those isotopes, researchers have to give the species different charges so that they behave differently in the magnetic field. Carbon-14 atoms accept electrons to form a negatively charged beam, but 14N atoms don’t. So that technique solves the problem for carbon dating.
But for other elements, including Ca, the trick of making a negatively charged beam doesn’t work well. That’s where ATTA comes in.3 Introduced in 1999, the technique filters atoms based on their electronic structure rather than their mass.
An ATTA sample is first vaporized in an oven to create a neutral beam of atoms. Those atoms are then slowed by Doppler cooling and, eventually, loaded into a magneto-optical trap so that their fluorescence can be imaged (see figure 2a).
Individual calcium-41 atoms are picked out of a Ca beam by atom-trap trace analysis. (a) The fluorescence from a single trapped 41Ca atom is recorded as a CCD image. The pixels in the region of interest (ROI) are used to measure the signal. (b) When the laser frequency is tuned to be resonant with the electron transitions in 40Ca, 42Ca, or 43Ca, the beam traps many atoms at once, which produces a bright fluorescent signal (top panel). But when it’s tuned to 41Ca, it rarely catches more than a single atom at a time (bottom panel). (Adapted from ref. 1.)
Individual calcium-41 atoms are picked out of a Ca beam by atom-trap trace analysis. (a) The fluorescence from a single trapped 41Ca atom is recorded as a CCD image. The pixels in the region of interest (ROI) are used to measure the signal. (b) When the laser frequency is tuned to be resonant with the electron transitions in 40Ca, 42Ca, or 43Ca, the beam traps many atoms at once, which produces a bright fluorescent signal (top panel). But when it’s tuned to 41Ca, it rarely catches more than a single atom at a time (bottom panel). (Adapted from ref. 1.)
The main reason why AMS measurements of 41Ca can’t reach natural environmental levels is because of potassium-41 contamination; the species have the same atomic mass, and even after purification, the stable isotope is always present at the same low level as 41Ca. But the two have entirely different electronic structures. So when the ATTA laser is tuned to a 41Ca transition, it has no effect on 41K, and the contaminant is no longer an issue.
The problematic contaminants in ATTA are isotopes of the same element. Slightly different nuclear masses and dipole moments alter the isotopes’ energy levels and shift the electronic transition frequencies. But those shifts can be small, so isolating a particular transition can be tricky.
Picking out a peak
Figure 2b shows the ATTA spectra for the four Ca isotopes that Xia, Sun, and their colleagues measured. When they tuned their laser to the resonance of the more abundant isotopes (top panel), they caught so many atoms in the trap at once that the measurement yielded only an overall fluorescent signal. But at the 41Ca resonance, rarely was there more than one atom in the trap at a time, so the researchers could count individual atoms.
The fact that ATTA looks at only one isotope at a time differentiates it from typical spectroscopy measurements and is crucial for its success with rare isotopes. The 40Ca fluorescence peak is 16 orders of magnitude taller than the 41Ca peak, and the two are separated by only about 200 MHz. If they were measured together, the smaller peak would be completely swallowed by the tails of the larger one. Trapping the 41Ca atoms separately allows them to be measured in isolation and their tiny peak to emerge.
For dating studies, though, the quantity of interest isn’t just how many 41Ca atoms are present; it’s the ratio 41Ca/Ca. To figure out the Ca beam’s intensity, the researchers looked at the stable isotope 43Ca, which has the same hyperfine structure as 41Ca and is therefore trapped with a similar efficiency. Its isotopic abundance is 0.135%, which is too high for counting individual atoms. But the researchers extracted the trapping rate—how many atoms were caught per second—by turning the 43Ca trap on and off and analyzing the slope of the fluorescent signal.
Lu was part of the team that first developed ATTA and has been successfully applying the technique for years to dating glacier ice cores with argon and krypton isotopes. (See the Quick Study by Lu, Physics Today, March 2013, page 74.) Noble gases present the same challenge to AMS as Ca does—they don’t form negative ion beams, so it’s hard to separate the radioisotopes from more stable isotopes of the same nuclear mass.
So why did it take so long to apply ATTA to 41Ca? Researchers, including Lu, did attempt those experiments shortly after the technique’s development.4 And it was successful in applications for which the concentration could be raised above the natural background level of about 10−15. But it couldn’t measure lower levels.
Lu credits advances in lasers in part for their current setup’s improved performance: “Laser power is a lot higher, and laser frequency control is better—everything got better.” Whereas it was difficult to pinpoint an isotope’s transition frequency for cooling in earlier experiments, doing so is straightforward with modern lasers. Add to that an increase in power from about 60 mW to 3 W and an ATTA setup can reach a loading rate that is four orders of magnitude higher than it was in the technique’s early days.
Advances over the past 15 years in modern cold-atom experimental techniques were also necessary to improve the trap’s efficiency at capturing and holding atoms. One particularly important technique is so-called repumping: Each time an atom is excited in the Doppler-cooling process, it has a small chance of falling back to a state other than the ground state. When that happens, it can no longer be cooled by the laser and is lost. A repump laser directs the rogue atoms back to the ground state, thereby keeping more atoms around. It also helps the trap hold them long enough for imaging, which takes about 10 ms.
Going lower
The researchers measured the 41Ca abundances in a series of samples that they made using commercially available Ca with a known 41Ca abundance of 10−7 and an added dilutant. The measured abundances in those samples covered three orders of magnitude, from 4 × 10−13 to 5 × 10−16, and had an error of about 12%.
The lowest abundance the researchers measured was that of the dilutant itself—Ca collected from deep in a mine. They determined the dilutant’s 41Ca abundance by adding known amounts of it to the commercial Ca and measuring the resulting samples with ATTA. That calibration yielded an abundance of (8.4 ± 3.5) × 10−17, which agreed with the value the researchers found through ATTA measurements of the pure dilutant. And they expect to be able to reach still lower concentrations by strengthening the atom beam and improving the laser-cooling setup.
The researchers are already collaborating with other groups to explore geological applications for their device. One area of particular interest is exposure dating of glaciers. Calcium-41 is produced through the capture of cosmic-ray-induced neutrons by 40Ca, but the process happens only in rock and soil within the top few meters of Earth’s surface. So if rock gets buried under a glacier, the process stops, and the 41Ca starts to decay. Measuring a newly exposed rock’s 41Ca/Ca ratio should therefore reveal how long it’s been since the rock was last exposed.
A problem with that logic, though, is that baseline levels of 41Ca vary considerably, so it’s hard to establish the starting point from which the level declined. But that number is critical if researchers want to use the isotope for dating.
“We’ve been talking to experts,” says Lu, “because we have this new tool, but we don’t quite know what to do first.” One potential starting point they’ve identified is collecting and measuring the 41Ca concentration in water samples from various ocean bodies to see if it’s uniform. The thinking is that even if there is local variation in the concentration, the currents might even it out, and then the water could serve as a baseline in certain situations.
But that theory still needs to be tested, and future applications are still uncertain. “We’re just excited that now we can measure all kinds of samples and learn about 41Ca and its distribution on Earth,” says Lu. “We can now explore the world.”
