What is a mass spectrometer?

From crime scenes to the bottom of the ocean, scientists are often analysing samples to find out what chemicals they contain, and this often involves a mass spectrometer. But what is a mass spectrometer, and how do they work?

Light can be split by a prism into a spectrum of different colours which take different paths. Mass spectrometers use different physics to produce a similar end result, but with molecules with different masses taking different paths, producing a 'mass' spectrum.

Mass spectrometers use an interesting mix of physics to charge, accelerate and then deflect the path of atoms and molecules. Once an electron is removed from an atom or molecule, they respond strongly to electric and magnetic fields, and are referred to as ions.

Making the ions fly through a strong magnetic field curves their path, but because the force is proportional to the charge, heavier molecules are deflected less, fanning the paths of the ions out similar to the spectrum produced by an prism. Charge sensitive detectors sit at the end of the flight path, and every time an ion is collected it produces a current, which is recorded and used to calculate the ratio of different masses.

If you use a sample with a huge mix of molecules with different masses, it can end up to be almost impossible to work out what a sample is made of. However, with careful preparation samples can reveal an enormous amount.

How NIWA uses mass spectrometers

NIWA uses the ratios of different types, or isotopes, of oxygen to perform a lot of its research. Not all oxygen atoms are the same. The centre, or nucleus of an atom is what makes up the majority of its mass, and consists of protons and neutrons. The number of protons (8) is always the same, but the number of neutrons can vary, and these different varieties are known as isotopes. Although over 99% have 8 neutrons (16O), oxygen also comes with 10 neutrons (18O around 0.2%), and 9 (17 O less than 0.04%). This last one is so rare we'll leave it out of the following example to make it simpler.

Because 18O is heavier than 16O, water molecules (H2O) which contain the 'heavier' oxygen are slightly less likely to evaporate than water comprised of the 'light' oxygen. This means that rain and snow has, on average, a slightly different ratio of 18O to 16O (δ18O) than the ocean. We can detect this small difference with a mass spectrometer. This effect is stronger for the warmer parts of the ocean, which leads to different parts of the ocean having different δ18O ratios, which we can see across the ocean today. 

Using δ18O to see into the past

During ice ages, water that evaporated from the ocean preferentially contained 16O and went on to be trapped as snow and ice at the poles, trapping the 16O, which means the overall the ration of 18O in the ocean went up. Because sea creatures use oxygen in their hard parts (shells), it can match the water they live in, and after they die their remains build up on the sea floor, it's possible to see a pattern of different ice ages going back hundreds of thousands of years in sediment cores from the seabed. These changes in δ18O are very small, but modern mass spectrometers are enormously sensitive, they've come a long way since they were invented in 1912.

We can use this information to work out what ocean conditions were like in the past when ancient snow was formed, or what ocean temperatures were like when sea creatures were alive. Both these techniques involve drilling cores deep into glaciers or the sea bed, allowing us to get data ranging back hundreds of thousands of years. There are limits to what can be determined from these kinds of temperature measurements, or proxies, as they are known, but when multiple different proxies agree with each other it greatly increases our confidence in the results.

For example, in younger ice cores you can see lines for each winter's heavier snowfall and count them like tree rings. It's also possible to calculate how much snow fell each year. You can take the same slice of snow and analyse its δ18O ratio, and see that it is opposite the signal in the ocean, i.e. as the 16O is removed from the ocean it ends up in icecaps. Rare events like volcanic eruptions, which can also be dated in a number of other ways, leave ash signals in ice cores and ocean sediments, which means we can increase our confidence in the conclusions we draw from these data. 

Carbon 13

Similar work is carried out with isotopes of carbon. Carbon 14 (14C) is famous for its use in dating because it undergoes radioactive decay, and the early parts of ice cores can be be dated in this way. However, there are two non-radioactive, or stable, isotopes of carbon 12C and 13C. These have different effects on chemical and biological reactions, similarly to δ18O, which can tell us about sources of carbon in the atmosphere of the day.

In practice, the samples we work with are usually carbon dioxide, or CO2, which gives us several likely combinations of stable isotopes of carbon and oxygen:

16O 16O 18O
16O 16O   16O 
 12C  13C  12C
 44  45  46

This means we can discover information about the temperature of ancient oceans and the sources of ancient carbon in carbon dioxide from the samples of CO2 from air trapped in different parts of an ice core. This allows us to see how ocean circulation patterns, carbon sinks and a whole host of other systems changed as previous ice-ages came and went over hundreds of thousands of years. This will help predict how adding greenhouse gases to the atmosphere today will go on to affect the earth in future. 

This is a more realistic model showing three different kinds of CO2 molecules in a mass spectrometer, below.

Sample gas is frozen back from the MAT253 after stable isotope analysis for later Carbon-14 analysis. [ Photo: NIWA ]