Carbon’s only naturally occurring radioactive isotope, 14C, is exceedingly rare. Produced when neutrons from cosmic rays interact with nitrogen, the radioisotope makes up just one part per trillion of the carbon in Earth’s atmosphere. Yet because of its continual production, its diffusion through the planet’s carbon cycle, and its long half-life of 5700 years, 14C is routinely used to date organic matter as old as 50 000 years. Archaeologists, forensic scientists, and environmental researchers, among others, essentially measure the concentration of radiocarbon in a sample to determine its age.
Since the late 1970s, accelerator mass spectrometry (AMS) has served as the benchmark method for the job. In that approach, samples are burned, chemically converted to graphite, and bombarded with cesium ions. The negative carbon ions ejected from the solid samples are then accelerated to a few percent of the speed of light and their mass-to-charge ratios deduced from their trajectories through electric and magnetic fields. Fortuitously, the most common isotope of nitrogen in the atmosphere, 14N, forms no stable negative ion; and its absence eliminates its otherwise large interference with the 14C signal. Likewise, 12CH2 and 13CH molecules are broken apart during a later, electron-stripping stage and don’t survive to interfere with the signal.
Both effects help free 14C signals from background noise. But although the technique is powerful—and applicable to other trace elements—the spectrometers can cost millions of dollars and often require a dedicated facility to maintain their electrodes at hundreds of thousands to millions of volts in a vacuum.
A technically simpler approach also begins with burning a sample, but only to transform its carbon atoms into carbon dioxide molecules. With their strong vibrational absorptions in the mid-IR, the many isotopic combinations of CO2 can be distinguished optically. The challenge is to measure the intensities of their spectral lines to determine the concentration ratios. The task is not easy if the goal is to count trace isotopes in a sea of abundant ones. The CO2 molecule has hundreds of vibrational and rotational lines, many of them closely spaced in frequency. And even the most stable lasers suffer from intensity fluctuations.
Five years ago Iacopo Galli and his colleagues at Italy’s National Institute of Optics and the European Laboratory for Non-Linear Spectroscopy adapted an optical method that is immune to such fluctuations to measure trace amounts of radiocarbon dioxide.1 Although its sensitivity fell short of that of AMS by more than an order of magnitude, the method, called saturated-absorption cavity ring-down (SCAR) spectroscopy, could be done on a tabletop and was less expensive than AMS. In the years since then, the researchers embarked on a program to reengineer their proof-of-principle demonstration. Their latest implementation2 reaches a sensitivity of just five parts in a quadrillion (1015). That’s within a factor of two of the AMS state of the art.
SCAR
The group’s method entails filling an optical cavity with CO2 and illuminating the cavity with an IR laser beam tuned to a molecular transition in 14C16O2. Thanks to the thousands of round-trips the beam makes between cavity mirrors, the optical path length is on the scale of kilometers. When the beam is turned off, the intensity of the light remaining in the cavity “rings down,” decreasing over the roughly 100 µs it takes the stored light to leak out. The rate at which that transmitted signal decreases depends on two contributions—how quickly light leaks from the cavity because of the mirrors’ imperfect reflectivity and how much of the light is absorbed by 14C16O2.
Since the mid 1990s, when they first applied cavity ring-down spectroscopy to more abundant molecules, scientists would routinely measure the ring-down intensity twice to disentangle the two contributions: once with the laser frequency off the absorption line to effectively measure an empty cavity in order to account for purely mirror losses and a second time with the laser frequency moved back on resonance to account for the molecular absorption.3
In 2010 Galli and his team realized that they could disentangle the effects in a single measurement, and SCAR was born.4 The trick is to shine enough photons into the gas-filled cavity to saturate the molecules’ vibrational transitions: With half the molecules in the excited state and half in the ground state, the gas becomes transparent to light circulating through the cavity. And even after the photon source is turned off, the transparency persists for several microseconds, until the intensity decays enough for excited-state molecules to relax and reabsorb some of the remaining photons.
The advance made it possible for Galli and colleagues to subtract the mirror losses—the only contributors to the first part of the decay curve—from the absorption contribution encoded in the curve’s tail. It also bought the researchers a factor of 20 boost in sensitivity relative to conventional cavity ring-down spectroscopy, and it enhanced the frequency resolution by three orders of magnitude. In essence, the absorption coefficient of 14C16O2 in the cavity is proportional to the difference between the unsaturated and saturated decay rates; from that relationship and a knowledge of the gas sample’s pressure and temperature, they could determine the molecules’ concentration to tens of parts per quadrillion.
Back to the drawing board
To achieve the additional order-of-magnitude improvement in sensitivity in the new setup, the researchers redesigned their laser system and optical cavity from scratch; the figure shows a schematic. Whereas a bulky titanium:sapphire laser linked to an optical frequency comb was used in the earlier, more complicated version of the experiment, the new setup uses two quantum cascade lasers—one of which serves as a frequency stabilizer for the other, used to probe. The 1-m-long cavity, whose volume was reduced by an order of magnitude to lower the amount of gas needed for a measurement, now boasts higher-reflectivity mirrors, which increase the optical path to about 5 km, a 40% gain. Among other modifications, the team swapped out their dry-ice bath for a cryocooler that lowers the cavity’s temperature to 170 K. The modest 25 K temperature reduction suppresses interference from the wings of isotopic absorption lines close to the one chosen for 14C16O2.
Optical detection of radiocarbon dioxide. An optical cavity filled with carbon dioxide is illuminated by a continuous-wave quantum cascade laser (QCL1) tuned to excite a specific molecular transition in 14C16O2. When the light from QCL1 is turned off—blocked from the cavity by an acousto-optic modulator (AOM)—light intensity in the cavity “rings down,” or decays over time, as photons leak through the mirrors or are absorbed by the gas. In their experimental setup,2 Iacopo Galli and colleagues detect the intensity signal through the far mirror and extract the contribution of molecular absorption to its decay rate. Because the laser linewidth is far narrower than the molecular line, they scan the QCL1 frequency across 600 MHz—measuring the decay time at 10 MHz intervals—and generate an absorption curve whose spectral area determines the concentration of 14C16O2. Three feedback loops (orange, green, and blue) stabilize the frequency of QCL1 and keep it locked to a cavity resonance while it is being tuned across the target transition. In particular, a second quantum cascade laser (QCL2), which is frequency locked to a transition of nitrous oxide (N2O), acts as a strong reference frequency. The “beat note” from the interference of the lasers’ two slightly different wavelengths is fed into a piezoelectric plate (PZT) that adjusts the separation of the optical cavity’s mirrors.
Optical detection of radiocarbon dioxide. An optical cavity filled with carbon dioxide is illuminated by a continuous-wave quantum cascade laser (QCL1) tuned to excite a specific molecular transition in 14C16O2. When the light from QCL1 is turned off—blocked from the cavity by an acousto-optic modulator (AOM)—light intensity in the cavity “rings down,” or decays over time, as photons leak through the mirrors or are absorbed by the gas. In their experimental setup,2 Iacopo Galli and colleagues detect the intensity signal through the far mirror and extract the contribution of molecular absorption to its decay rate. Because the laser linewidth is far narrower than the molecular line, they scan the QCL1 frequency across 600 MHz—measuring the decay time at 10 MHz intervals—and generate an absorption curve whose spectral area determines the concentration of 14C16O2. Three feedback loops (orange, green, and blue) stabilize the frequency of QCL1 and keep it locked to a cavity resonance while it is being tuned across the target transition. In particular, a second quantum cascade laser (QCL2), which is frequency locked to a transition of nitrous oxide (N2O), acts as a strong reference frequency. The “beat note” from the interference of the lasers’ two slightly different wavelengths is fed into a piezoelectric plate (PZT) that adjusts the separation of the optical cavity’s mirrors.
With a footprint less than 2 square meters, the team’s tabletop system is one-quarter the size of the most compact 14C-dedicated AMS systems available, and one-hundredth that of typical AMS facilities in national laboratories. Although its sensitivity and precision still don’t match those of AMS, SCAR can measure concentrations spanning six orders of magnitude. It also doesn’t entail destroying gas samples in the process of analyzing them. Galli and company envision that with its compact size, acquisition times of about an hour, and a need for just 6 mg of carbon in a sample, the technique should be practical for use in the field.
Indeed, radiocarbon applications go well beyond dating samples. Pharmaceutical companies can now lace their drugs under development with a small concentration of 14C, a procedure known as microdosing. A decade ago the US Food and Drug Administration and European Medicines Agency approved such 14C tracers for in vivo human drug studies. The tracer concentrations—well below therapeutic levels—allow physicians to monitor a drug’s path through the body to where it eventually settles. That application doesn’t require state-of-the-art sensitivity, and pharmaceutical firms are likely to want their own detectors.
