I realize that most of the hubbub about Curiosity, the newest Martian rover mission, was in August after the successful landing and subsequent initial instrument tests, but is it ever too late to talk about robots on other planets? No! Consider this my tiny endeavour to keep space science in the public’s eye, even when NASA’s not just landed a small car right side up on another planet with perfect precision.
Curiosity successfully landed on August 5th, and after a perfect landing, and all the instruments are in excellent working order. The instruments on board are all described on the NASA Jet Propulsion Lab’s site, with a blurb about how they work and what sort of things they’re looking for. For this post, I’m going to focus on the Canadian instrument on board.
The Alpha Particle X-ray Spectrometer, or APXS, is the only all-Canadian contribution to the mission. It’s designed by Ralf Gellert, a professor at the University of Guelph (previously known best for cows with windows in their stomachs and pigs with phosphate-reduced poop), built by Macdonald-Dettwiler Associates (who also built the Canadarm), funded by the Canadian Space Agency, and the research involves a group at Guelph as well as people at a few other Canadian universities. I’ve got some thoughts about what the APXS represents to Canadian science and science policy, but I’ll save that for a second post. I’ve heard press about various other instruments on board, including the laser spectrometer, the array of cameras, and the meteorological station, but not very much about how the APXS works and what it can do. So let’s talk about how it works!
The APXS is a pop-can sized instrument mounted on the robotic arm on Curiosity that uses two types of radiation to determine the elemental composition of rocks, soils, and dirt on Mars. Six radioactive Curium-244 sources covered with a thin titanium foil sit around an X-ray detector. Curium (element 96), like all elements heavier than uranium (element 92), radioactively decays by emitting an alpha particle, which is 2 protons and 2 neutrons bound together. When curium emits an alpha particle, it decays to plutonium-240, which then decays with a number of daughter atoms and x-rays. The x-rays from plutonium and alpha particle from curium are the radiation used to analyze a sample; the other by-products are typically of lower energy and are filtered out from the radiation beam by a foil over the sources. This radiation streams out of the sources in all directions, but to focus the beam, a metal ring surrounds the sources and the detectors to absorb radiation radiating at a wide angle.

Left: Photo of the APXS. The detector itself sits in the middle of of the ring behind the sources. Photo from JPL. Right: A schematic of how the APXS works. The instrument is shown as if it were pointing at a table-top. Picture from the University of Guelph.
Two kinds of radiation are used so that a wide range of elements can be detected. The alpha particles are used for Particle Induced X-ray Emission, or PIXE, which is used to detect elements with a low atmoic number. The X-ray radiation is used in X-ray Fluorescence, or XRF, and detects elements with a higher atomic number. Combining the the two technique means that elements from sodium (atomic number 11) to zirconium (atomic number 40) can all be clearly detected in the combined spectrum.
In PIXE analysis, particles (alpha particles here, but other particles like protons can also be used) are beamed at a sample. The particles are emitted with a known, fixed energy and collide with the atoms in the sample, where some of the particles will interact with the inner shells of the electron structure. Electrons in an atom sit in a series of increasingly energetic levels, each with a fixed energy. If a particle hits the atom and interacts with an electron, it imparts energy to the electron. Since the electron can only have a fixed amount of energy at any given level, if it gains energy it must jump up to a higher level, which leaves a gap in the original level it started at. An electron from a higher energy level (not necessarily the one that originally jumped up to a higher level) then falls down to fill the gap in the lower energy shell, and emits an X-ray with the difference in energy between the two levels to conserve energy. This is clearer with a picture!

Alpha particle comes in, excites electron and scatters off, then an electron (either the one that was bumped up in energy or another one) fall to a lower energy and the atom emits an X-ray carrying the difference in energy between the starting and finishing energy levels of the second electron (ie, the electron that drops down in energy).
The process is dependent on the inner (ie lower energy) shells of electrons interacting with the incoming particle. The particles are most likely to interact with the outermost shell or two of the electron levels of an atom, so if an atom has several shells, there’s too many electrons shielding the lower energy electrons for many interactions to occur. PIXE is then most effective for atoms with a low atomic number.
XRF does much the same thing, but instead of using a particle to induce the electron’s transition, it uses X-rays. X-rays are useful because they can interact more easily with a inner-shell electrons in heavier elements than the massive alpha-particles. XRF is then more useful for detecting heavier elements that PIXE cannot, but because the efficiency of the process is low at low atomic number, PIXE (which is a more efficient process at low atomic number) is needed to gather a robust spectrum.
But doing PIXE and XRF on Mars poses problems: even the very thin Martian atmosphere will absorb low energy X-rays, so the lightest element that the APXS can detect is sodium (atomic number 11). Hydrogen, carbon, nitrogen, and oxygen (atomic numbers 1, 6, 7, and 8 respectively) are all invisible to the APXS, and oxygen in particular is abundant in the samples. Most of the heavier elements are present as oxides, so the mineralogy becomes very important in the elemental analysis. Plus the atmosphere varies, like any atmosphere, and the temperature fluctuates, and generally the conditions are not as pristine as they are in a controlled lab setting. This means that there’s more calibration work needed on Mars than in the lab, and also that the invisible elements need to be carefully treated.
So when a sample is being analyzed, Curiosity’s robotic arm extends (slowly and carefully, to avoid any collisions) out towards the sample rock. A pressure sensor on the front of the instrument trips when the instrument touches the rock, and the protective doors that cover the aperture of the instrument when it’s not in use open. The electronics are switched on, and the detector starts to gather a spectrum of X-rays emitted at various energies from the sample. The instrument is left in place for a between a few minutes and a few hours to gather data, which is stored and then beamed back to Earth in the daily data dump. When it’s finished collecting data, the arm moves back from the rock, the doors close, and the next sample is chosen.
The spectra that come back look something like this:

A typical APXS spectrum. The horizontal axis is energy of the detected x-ray, and the vertical axis is the number of x-rays of a given energy detected. Image from JPL.
The horizontal axis is the energy of the detected X-ray, and the vertical axis is the number of X-rays detected at that energy. The spectra consist of peaks at specific energies, and the individual elements are identified by the energies of a series of characteristic electron transitions. The emitted x-ray of a given transition has a fixed energy, equal to the difference in energy between the two states of the electron. Since different elements have different atomic structures and different amounts of energy between the electron shells, the x-rays emitted by different elements undergoing either PIXE or XRF will have different energies. Working backwards, then, if a spectrum is collected and there is a peak at an energy corresponding to a transition of a given element, then that element is present in the sample. The precise composition of the sample is determined by calculating the relative area under each of the peaks.
The invisible elements complicate the analysis considerably, especially since there’s so much oxygen in the rocks. The mineralology of the rocks must be accounted for, and information from other instruments (including the cameras — visual identification is very helpful) is used to make initial estimates of the proportion of each element in its various oxidation states. For example, iron oxide may be ferrous (FeO) or ferric (Fe2O3), though to the spectrometer, it’s all just iron. So to do the analysis, knowledge about similar rocks on Earth is used to estimate the proportion of iron in each oxidation state. After all the expected oxygen from the bound minerals is accounted for, any “invisible element” left over may be bound water. Many minerals have water bound within the crystals, and Spirit and Opportunity found evidence of bound water in Martian rocks using this analysis.
Spirit and Opportunity were sent to Mars to find water in some form (and did!), while Curiosity was sent to examine whether or not at some point in its geological history, Mars could’ve sustained life. It’s important to note that it’s not looking for *life*, just evidence that conditions hospitable to life once occurred. Due to the raging success it’s had on Mars (the instrument on Opportunity is, to the best of my knowledge, still functioning, eight and a half years into a 90 day mission), other APXS’s are planned for missions to the Moon by the Indian and Russian space associations (the Chandrayaan-2 rover, landing hopefully in 2016) and to Comet 67P/Churyumov-Gerasimenko by the European Space agency (the Philae lander, landing in 2014).
More info: