Elementary, said he

80s-tastic plasma lamp (Wikipedia)

Today I was mostly preparing lake sediment core samples for ICP. ICP stands for Inductively Coupled Plasma, and comes either in the MS (mass spectrometry) or OES (optical emission spectrometry) variety. Both are methods for measuring the different elements present in a sample. The ICP bit is a 'torch' containing a gas, typically argon, that is ionised by heating it via electromagnetic induction (remember those weird glowing globe plasma lamps?). Still confused? This means there is a coil wrapped around the torch, which produces a very strong electromagnetic field when turned on. The argon gas is ''lit' by an electrostatic spark, and the gas becomes ionised.
The sample (dissolved in liquid) is sprayed into the argon flame, and also becomes ionised. As different components in the sample become ionised they gave off a characteristic energy. The MS and OES parts is the bit that detects the elements present. MS does this by measuring their atomic mass, whilst OES measures the wavelengths of energy emitted by the elements. Each element produces a characteristic wavelength, so we can identify them on this basis.

Preparing samples from Lake Nineris, Latvia

And why do we want to measure elements you may ask? ICP is another one of those lab methods that has been developed for forensic and environmental applications then adopted by archaeologists. It is used routinely to look for metal toxicity in human blood and tissue samples, and to identify levels of pollutants in water. In archaeology, the idea is the same, but we are looking at the past history of such 'contamination' events. In the lake cores we have from the Baltic for the Ecology of Crusading project, we are measuring a range of elements in the sediments which can be linked to human activity. The lakes are located close to crusader castle sites, and pilot studies have linked signals such as increased lead to the castle construction. Other signals can be linked to activities such as landscape clearance.

Dissolving sediments in HNO₃

Before we can analyse the samples, they have to be prepared for the ICP machine. This involves grinding/sieving them to less than 20mm and dissolving them in concentrated nitric acid. Grinding up and weighing packets of soil takes longer than you'd think. They'll sit in the acid overnight, then will be heated for 9 hours tomorrow (checking at intervals for explosions and whatnot). The resulting solution is then filtered, diluted and injected into the ICP.

So far I estimate there are about 60 samples per lake core, and maybe 15 lake cores. Eep.