The Human Element I

Recent events surrounding potential adulteration of Cinnamon Bark oil from a large corporate supplier have inadvertently highlighted the problems associated with the overemphasis of instrumental analysis and especially GC/MS analysis, for proving the purity of essential oils.

Let us recollect. For quite some time now it has become almost imperative to provide GC/MS read-outs to concerned buyers of essential oils. This seems to be a consequence of the standpoints announced by different parties in this conversation.

There are the big corporate players who advertise the prowess of their testing labs, enabled by their ability to spend lavishly on equipment and personnel. The implication is that by having all the machinery money can buy, the oils associated with the brand must be pure. Another aspect of this conversation is that 3rd party testing has become quite en vogue. One implication may be that analysis performed by the owner of an oil may not be objective. Then there are those students of aromatherapy who request analysis so it helps them understand the make-up of the oil and understanding the therapeutic properties accordingly (i.e. if an oil is rich in alcohols it is different from one high in esters).

All these aspects carry quite some truth. More powerful equipment may well allow better insights. Sending a sample to two or three different labs is always highly instructive to say the least and the desire to understand the basic chemical make-up of an essential is all but tantamount to the concepts French style aromatherapy.

There is only one problem. We all have become so accustomed to the power of computation that we forget that there sometimes are sizable margins of error in the data that are generated. In our case the exclusive reliance on instrumental analysis and especially GC/MS has its own set of problems. The most overlooked issue is the way GC/MS are ran today.

Lets step back in time just a bit. As recently as 30 years ago Mass Sectra were interpreted on a case by case basis by chemists. A mass spectrum is a record of all the observable molecular fragments that arise when a specific molecule is shattered to pieces by bombardment with electrons. The range of fragments that arise is indicative of the molecule that was blown apart. It took, mostly rather specialized chemists who knew how to make sense of the many molecular fragments that show up in a Mass Spec to identify the original molecule. Spectra were interpreted one by one. (A classic introduction to the topic is “Identification of Essential Oil Components by Gas Chromatography/Mass Spectrometry,” by Robert P Adams)

However, when one analyzes an essential oil there are dozens of Mass Spectra, one for each component in the oil. Needless to say that for routine analysis this proved way to cumbersome to have a chemist sit down and interpret each spectrum separately. However, with the advent of computers these tedious tasks of interpreting the enormous amounts of data of not only one, but many Mass Spectra became manageable. Once a spectrum was clearly associated with a specific molecule it was stored in a computer library and software was written which compared the spectra generated in your GC/MS with those previously recorded. If your newly recorded spectrum matched the one in the library you had identified your component. But here comes the hitch. Almost never is there a 100% correlation or identification between your recorded spectrum and the one in the library. The software will return data that will say component x is identified, lets say, as linalyl acetate with a degree of 86% certainty. However with a degree of 75% certainty it could also be a closely related isomer. To chose the right component from the different choices offered the operator needs to have a good measure of experience .

I could go on and on about the idiosyncrasies and the potential for wrong conclusions in the process of interpreting GC/MS data. Identification of EO components becomes especially fuzzy when it comes to the complex field of sesquiterpene isomers. Most Mass Spec library setups are hopelessly over challenged when it comes to precise IDs for complex sesquiterpene mixtures. The take away message here is that even and especially computerized interpretation of the Mass Spectra of a GC/MS is not immune to delivering incorrect results. To quite some degree the reliability of such interpretations is based on the quality of the library. I.e an operator who has analyzed EO for ten years and added and added specific EO components to her or his specific library will have a better chance than someone else who starts out with a store bought library of common chemical components.

The above described issue is one inherent in todays ways of managing large amounts of information by computers. But there is yet another issue that is inherent to the process of analysis itself: purity and authenticity. Often mixed up in casual discussion these two terms really have different meaning. In the case of EO a point can be made that pure should mean that no impurities are present or have been added. Authentic is used to describe that an oil is the true representation of the components present in the specific plant that is referenced on the label.

Impurities or adulterations often are present in the form of easily discovered components such as perfume extender or also in the form of well known substances used to adulterate specific oils. Such gross adulteration is easily discovered with GC/MS. But the methods for adulterating or even reconstructing essential oils have reached an astounding level of refinement. For example Lavender EO. It contains, in an authentic sample, the terpene alcohol linalool. The proportion of linalool in different Lavender EO may well vary to some degree. While the addition 5% of natural linalool, derived from another EO, can ultimately be proven by strenuous analytical effort, routine GC/MS will not detect it. In other words neither GC/MS or any other instrumental method alone can prove authenticity of an essential oil.

That this is so can be verified by looking at some of the analyses posted on the internet. These analyses ostensibly demonstrate the purity of an oil, but anyone familiar with the composition of authentic Fine Lavender will notice quickly that quite some of the posted chromatograms reflect EO which are reconstructed to varying degrees.

The only way to truly grasp authenticity is to have access to verifiably authentic specimens. Such specimens can the be analyzed and serve as benchmark for future analysis.

One of my mentors once told me the real way to ascertain authenticity is to see the growing plants, the still and ideally be present during distillation and obtain a sample fresh from the still.

Clearly one cannot do this every time for every oil on the list. But we believe that the contact to the actual producer is an indispensable element of every attempt to procuring authentic oils. In our experience adulteration never happens at the level of the producer. The know how to reconstruct, standardize or adjust essential oils is found in the big brokerages who often can sell essential oils for less than they cost at the source.

I could discuss this endlessly, but suffice it to say, instrumental analysis is a very valuable tool in working with essential oils, but when it comes to the appreciation for truly authentic oils it will never replace the human element.