A lot of people think that doing an Earth Science PhD involves looking at rocks. Most of the time they’d be wrong (experimental petrology = making pretend rocks; geochemistry = water; geophysics = computers; palaeontology = colouring in), but just occasionally, I do get to play with the real thing. From sample pick-up to analysis, it’s quite a journey – allow me to explain…
The first step is to find an exotic fieldwork location, and then justify why it’s the best place in the world to sample the type of rock you’re after. In my case, this is the Lesser Antilles, an island arc chain of volcanoes which are especially proficient in erupting cumulate xenoliths (coarse grained fragments of magma chamber wall).
Next, select your samples. It is important to record the geological context in which you found the rocks, the geographical location, and also to have a customs-cleared method of transportation back to the UK! In this case, a cargo ship full of bananas helped us get 200 kg of cumulates back to Bristol.
Hand specimens can only provide limited information. To investigate the processes and timescales involved in the formation of the sample, we need to retain mineral textures. The best way of doing this is to make thin sections. The rock is cut up into matchbox-sized pieces and ground down to a thickness of ~30µm (0.03 mm, or 1/3 of the thickness of a human hair). Glued to a glass slide, the wafer of rock is sufficiently thin for light to pass through and can be viewed using a petrological microscope.
The skilled use of a petrological microscope allows identification of mineral phases and reveals basic compositional information; however, it gets tricky to see anything smaller than 200µm (0.2mm) in diameter. If we want to start investigating on such a scale, we need a microscope with a little more complexity.
The Scanning Electron Microscope (SEM) uses a beam of electrons, rather than visible light, to generate images and can easily magnify up to 4000 times, revealing objects of a few microns. The most useful mode for petrologists like myself is Back Scattered Electron imaging where the brightness of the image is proportional to the atomic mass of the compounds in the field of view.
If you want to find out what a particular mineral is actually made of, you need EPMA (electron probe microanalysis). This technique is used to gather quantitative compositional information about elements that are present in concentrations of more than 100 parts per million (ppm). Similarly to the SEM, a small area (1-2µm diameter) of the sample is bombarded with a beam of electrons. These incident electrons collide with, and cause ejection of, electrons from the inner shell of atoms. An electron from a higher energy level will fall down to fill the vacancy; to conserve energy it will emit x-rays. The wavelength of this radiation is is characteristic to each element. The x-rays are either collected for energy dispersive spectrometry (EDS) to produce immediate semi-quantitative spectra (see image) or wavelength dispersive spectroscopy (WDS), which takes longer but is more sensitive and gives better detection limits (and ultimately the numbers that you see in papers).
In igneous petrology, we are also interested in trace elements (atoms that constitute less than 0.01 wt % of any sample). One way of investigating trace elements for individual minerals is to use a technique with an impressively long name: laser ablation inductively coupled plasma mass spectrometry (or LA-ICP-MS). I shall translate!
A pulsing laser beam is created by oscillating electric currents (ie. it is inductively coupled) and is fired at a sample spot. The laser erodes material from a small area of the surface (this is the laser ablation). The ejected particles are then ionised (typically a given a positive charge) by an argon plasma at temperatures of about 8,000ºC before being transported into a mass spectrometer which separates the ions based on their mass:charge ratio. It is a destructive technique (see image) but is extremely sensitive; elements can be detected down to ppb (one part per billion) limit.
This is by no means a prescriptive or exhaustive list as to what you can do with a rock. There are myriad other possibilities, including secondary ion mass spectroscopy (SIMS), which looks at isotopes and atomically light elements, and Fourier transform infrared spectroscopy (FTIR). Or you can ditch the thin section, and just crush up the sample to do bulk analysis (major, trace and isotopic composition). Once you have the rock, the world really is your oyster.