GC-MS Testing for Essential Oils: Reading a Sage Report

The certificate of analysis we publish for sage oil is not a GC-MS report. It lists three physical constants for one batch and no compound percentages, so it cannot tell you the thujone level, which is the figure that matters most for sage. A buyer with a safety reason should ask for a gas chromatography report in writing before ordering. This article explains how to read one when it arrives, using published sage data as the worked example.

Sage Essential Oil bottle

Sage Essential Oil

Our own Sage Essential Oil, in sizes from 0.5 oz to bulk.

From $13.99

Shop Now →

What GC-MS reports, and what it does not

Gas chromatography separates an oil into its volatile compounds as they travel through a long column at different speeds. A flame ionisation detector (FID) records how much of each compound comes off; a mass spectrometer (MS) helps identify what each one is. Many reports pair the two. The result is a table of compounds with a percentage beside each.

The percentage is almost always the share of total peak area, not the share of weight. ISO 9909, the sage oil standard, says the proportions are those "indicated by the integrator". An EFSA opinion on a sage oil additive (2024) states that its results are expressed as percentage of GC peak area, assuming the sum of all detected peak areas is 100%. A Tunisian sage study (Khedher and colleagues, 2017) likewise computed amounts from FID peak area percentages. Different compounds can give the detector different signals per gram, so area percent and weight percent are close cousins, not twins. Treat "25% alpha-thujone" as "about a quarter of the peak area".

The reading order

  1. Identity of the sample. Batch or lot number, species name and date. A report with no lot cannot be tied to the bottle.
  2. Method and instrument. FID, MS or both, column, temperature programme. Khedher and colleagues used an FID run for amounts and a separate MS run for identification, matching spectra to libraries and, for some compounds, to authentic standards.
  3. Total accounted for. Khedher reports 49 compounds identified, 97.97% of the oil. If a table sums to 60% or lists four compounds only, ask what the rest is.
  4. Marker compounds against a window. For sage: alpha-thujone, beta-thujone, camphor, 1,8-cineole.
  5. The thujone line. Combine both isomers and see the arithmetic below.
  6. Everything else (limonene, linalool, borneol, viridiflorol, humulene) is context; ISO 9909 gives limits on some and none on others.

A worked comparison

Here are three published views of the main markers, each as percent of peak area, with ISO 9909's window from its Table 1:

CompoundISO 9909 windowEFSA, 5 batches, GC-FID (mean, range)EFSA, GC-MS (mean)Tunisian leaf oil (Khedher 2017)
Alpha-thujone18 to 4324.96 (23.54 to 26.00)20.4818.83
Beta-thujone3 to 8.55.60 (4.98 to 6.16)5.614.46
Camphor4.5 to 24.518.74 (18.50 to 18.87)17.5825.14
1,8-Cineole5.5 to 1310.70 (10.33 to 10.99)10.6314.14

Three lessons sit in that table. First, the same oil reads 24.96 for alpha-thujone by FID and 20.48 by MS in the EFSA opinion, a gap of 4.5 points from two detectors (the opinion's text says the MS work covered the same five batches, while one table caption says 10, so the sample behind the MS column is not clear). A report that does not say which detector gave the numbers is hard to compare with another. Second, the Tunisian oil has camphor at 25.14 and 1,8-cineole at 14.14, both above ISO's maxima of 24.5 and 13 by our reading, though its authors say it fits the ISO profile. Third, the EMA's assessment report, citing Bradley (2006), gives wider literature ranges (alpha-thujone 10% to 60%, camphor 5% to 20%) and calls the composition very variable. A single batch report describes that batch.

The thujone line, with arithmetic

Add alpha and beta. ISO's windows allow a combined 21% to 51.5%. The EMA's public statement (2016) says sage oil is characterised by high levels of thujone, that thujone is toxic and may cause seizures at high doses, and restricts sage leaf preparations to 6.0 mg of thujone per person per day for at most two weeks. EFSA applied a benchmark dose lower limit of 8 mg per kg of body weight per day for alpha-thujone, derived from rat seizure data.

To turn a report into milligrams, take the share of oil that is thujone and multiply by the weight of oil. One drop is about 0.05 ml, about 46 mg at a density of 0.915 (the density on our published certificate, one batch; weigh rather than count drops). At 21% thujone that is about 9.6 mg per drop, at 51.5% about 23.6 mg, and because the report is area percent, treat these as orientation, not exact. The EMA's 6.0 mg is an oral ceiling for a different material, so it shows scale and is not a skin or inhalation limit. What it shows is why a report that merely says "thujone present" is not enough. You need the number, the unit and the method.

What our certificate does and does not establish

The certificate we publish covers batch SAGHMCSOESC001A: relative density 0.915, refractive index 1.469 and optical rotation -10 degrees at 20 C, with the batch's refractive index and rotation sitting at the edges of the stated acceptance ranges. Against ISO 9909's physical limits, density and refractive index are inside and the rotation of -10 is outside ISO's +2 to +30. We cannot say why from a certificate of constants; a GC-MS would be the way to find out. The certificate also carries no species-authentication result beyond the name, no origin and no laboratory name. For a deeper look at how that sits with grade words, see the explainer on who grades essential oils.

What GC-MS cannot do

It does not, by itself, prove an oil is unadulterated. A 2024 review of authenticity methods in Food Chemistry (Yang and colleagues) notes that essential oils are volatile and prone to oxidation, that quality varies with harvest season and extraction method, and that widespread adulteration makes assessment difficult, so quick and effective identification techniques are needed. We have not summarised the techniques behind that paywalled article. The practical point is smaller: a report with all four markers inside a window is consistent with sage oil, and a report with a marker far outside is a prompt for questions. It is not a verdict either way.

Five questions to put in writing

  1. Which lot is this report for, and is that lot on the label of what I will receive?
  2. Which method and detector produced the percentages (FID, MS, both)?
  3. What are alpha-thujone and beta-thujone, as percent of peak area, and what does the table sum to?
  4. Who ran the analysis, and can I see the report itself and not a summary?
  5. Does the seller state a specification (for example ISO 9909) and where this lot sits against it?

Take the answers to whoever advises you on safety: a physician for pregnancy or a seizure disorder, a product-safety professional for a workplace limit. For how sage oil is made and what the process involves, see the steam distillation overview, and for the thujone safety background see what the number means and who should not use sage oil. Our buying guide is here.

Sage Essential Oil bottle

Sage Essential Oil

Sage Essential Oil from our own shop, in sizes from 0.5 oz to bulk.

From $13.99

Shop Now →

Sources & Further Reading