The Science of Gas Chromatography

Gas Chromatography/Mass Spectrometry (GC-MS, GCMS or GC/MS) is an analytical method that is used to determine the composition of substances within a test sample. As the name describes, the system is composed of a gas chromatograph to separate the substances (analytes) and a mass spectrometer to identify these analytes. This is the gold standard for analysis in many fields such as forensics or petrochemical studies as GC-MS allows for the unequivocal identification of substances.

GC-MS is usually associated with any sample that can go through a GC, which means that the substances of interest must be volatile and generally not damaged by heat (450C). As such, GC-MS is usually used for: pesticide analysis, food safety, food quality, environmental analysis (air, water, soil), petrochemical studies, forensics applications such as drug detection, arson/fire investigation, cause of death or toxicology investigations. More recently, GC-MS has been extensively used in biochemistry for the identification of metabolites (primary or secondary), metabolomic studies, plant research and even archaeology and geologic studies! In an ironic twist, GC-MS has even been used to study samples from Mars during probe missions since the 1970s! How cool is that?

GC-MS systems touch nearly every aspect of our lives and the data from these systems directly affects decisions regarding our quality of life.

Are all GC-MS systems the same?

No. There are two broad approaches to GC-MS technology, either routine applications or non-routine. For routine applications, you can define fairly closely what you are looking for. This is what we call known-knowns and known-unknown analysis. For non-routine applications, you tend to look for substances that have never been seen before, are not expected and/or are not in a routine library. For example, one might be looking for designer street drugs or new doping drugs in athletes. This is unknown-unknown analysis and is the most complex analysis possible.

“Known knowns”

Essentially, you have either a very defined target list and you are only interested in these (known-knowns). You will ignore any other substance that might be found. For example you are manufacturing synthetic perfumes and you want to know that the manufacturing plant has produced the perfume according to your recipe. As you know the exact recipe and the exact output, you have a very defined target list (number of substances, exact identity, percentage ratio). Typical systems used here are entry level TOF-MS systems, Quadrupole-MS systems or MS/MS systems such as triple-quadrupoles (though rare due to the cost).

“Known unknowns”

In this case you are working slightly beyond the known-knowns. Essentially, the substances you are interested in belong to a much longer list and might or might not be present. In such a case, you do not know exactly which substance you are looking for but you know that it belongs to a certain category. For example in the synthetic perfume, you may be interested in finding out exactly which skin irritants might be present. You have a list from the US or the EU as to which substances are banned and you want to make sure that these substances are not present. Typical systems used here are entry level TOF-MS systems,  MS/MS systems such as triple-quadrupoles or Quadrupole-MS systems though these are getting replaced by TOF-MS systems due to their performance in this field.

“Unknown unknowns”

In this final case, you are looking for a needle in a haystack while being blindfolded and not knowing what a needle looks like. This is what GC-MS systems were originally designed for until technical difficulties forced manufacturers to stop dreaming so big. In unknown-unknown analysis, you have no idea of the nature of the analytes, which category they belong to or even if they are present. This means that there are no libraries to refer to and the best you can hope is to have perfect separation from the GC, followed by a very accurate MS measurement to give you precise empirical information about the molecule. This is where you will use high-resolution mass spectrometers such as a HRTOF-MS system.


Why is GC-MS so useful?

GC-MS has rapidly grown in popularity over the last few decades as it provides a fairly accurate identification of the substances at ultra-trace levels. Traditional GC uses dumb detectors which enable the user to determine that at a specific moment in the analysis X amount of a substance is present. Based on the experimental conditions, a fairly good guess as to the identity of this substance is possible by using the retention time of that substance. However, in many cases several substances could have the same retention time, so the identification is not guaranteed.

Using a Mass Spectrometer, it is possible for the analyst to reference the output of the mass spectrometer with a library and thereby use not only the retention time, but also the mass spectrum at that time to identify the substance. The more fragments present in a mass spectrum, the better the identification will be.

This is similar to identifying a suspect using a fingerprint. The more unique features in a fingerprint, the more certain we are that we can match the fingerprint to our reference point.


What about LC-MS?

Liquid Chromatography-Mass Spectrometry is a related technique to GC-MS. Due to its current popularity, many people think that it is replacing GC-MS. However, this is only because scientists tend to become famous and/or make money when they publish or present something new. As LC-MS is a newer technique in general, more remains to be discovered and therefore more noise is made about LC-MS.

In any good laboratory you will have both a GC-MS and an LC-MS. These are complementary systems and have a little bit of overlap. For example, anything that is volatile (e.g. you can smell) will tend to be done by GC-MS while anything that has a lot of sugar will tend to be done by LC-MS. This is because sugars tend to burn at high temperature (make a braai or barbecue with sugar-based marinades and see what happens to the grill).

Some substances such as amino acids can be done by both techniques but are found at lower levels by GC-MS. The same can be said for Fatty Acids and other organic acids. These seem easier by LC-MS because you can pretty much inject the samples directly while by GC-MS you need to derivatize the samples to allow the substances to go into the GC. But the trade off is that the LC-MS tends to be 100 times less sensitive than the GC-MS for these.


Is GC-MS the “Gold Standard” for analysis?

While GC-MS is regarded as a gold standard for substance identification, the reality is slightly more complex. While a GC-MS can be used to perform a 100% specific test to positively identify the presence of a particular substance, certain criteria need to be kept in mind. For example, if the GC-MS is operated in Selected Ion Mode (SIM), the test is less reliable. This is at best a nonspecific test which could statistically suggest the identity of the substance, though this could lead to false positive identification. The correct way to use the GC-MS for substance identification, especially in the forensic field is to perform a full scan analysis. With most GC-MS systems, this  means sacrificing sensitivity, requiring more substance to be present. Only the TOFMS systems allow for full scan analyses without sacrificing sensitivity.