Essential Oils · Cymbopogon Oils
Citronella oil (Cymbopogon nardus): composition, yield
Cymbopogon nardus (L.) Rendle
- Family
- Poaceae
- Genus
- Cymbopogon
Species
| Binomial | Cymbopogon nardus |
|---|---|
| Accepted name | Cymbopogon nardus (L.) Rendle |
| WCVP status | accepted |
| Family | Poaceae |
| Genus | Cymbopogon |
| IPNI identifier | urn:lsid:ipni.org:names:396972-1 |
Source: Govaerts R. (ed.), World Checklist of Vascular Plants (WCVP), Royal Botanic Gardens, Kew, accessed on 21/09/2026.
Key facts
- The oil on this page comes from Cymbopogon nardus (L.) Rendle, a species of the family Poaceae.
- In one study of leaves collected in Brazil, oxygen-containing monoterpenes made up 90.61 % of the oil by relative peak area.
- The main compounds named by the authors include citronellal and geraniol; citronellal was also assayed by GC-FID, at 282.5 mg/mL.
- Fresh leaves were hydrodistilled in a Clevenger-type apparatus, and the yield is printed as “0, 7% (w/w)”, with no stated base.
- The ISO catalogue lists ISO 3849:2003 as “Published” and ISO 3849:1981 as “Withdrawn”.
De Toledo et al. (2016) describe C. nardus oil hydrodistilled from fresh leaves and analysed by GC-MS. Its composition is given as relative peak areas, and its citronellal content was assayed separately by GC-FID with external calibration.
Which plant does this oil come from?
In the World Checklist of Vascular Plants (WCVP, Kew), extract read on 9 November 2026, Cymbopogon nardus (L.) Rendle has the rank “Species” and the status “Accepted”, in the family Poaceae, under the identifier 406227.
The EFSA FEEDAP Panel (2024) describes C. nardus (L.) Rendle, which it calls “citronella grass, Ceylon citronella”, as a perennial grass species of the family Poaceae.
What is the oil assessed at European level?
In the EFSA FEEDAP opinion (2024), the additive under assessment is citronella oil from C. nardus (L.) Rendle. It is obtained from the fresh or partly dried leaves of the plant. The opinion describes it as a clear liquid, pale yellow to pale brownish yellow, with a “leafy and earthy” odour.
Which plant material was analysed?
The leaves studied by De Toledo et al. (2016) were collected in July 2013, in the morning. They came from the garden that the article names “Garden of Toxic and Medicinal Plants: Profa. Dra. Célia Cebrian de Araújo Reis”, at Universidade Estadual Paulista (Araraquara, São Paulo, Brazil). A voucher specimen, HRCB-60752, was deposited in the Rioclarense Herbarium of the Institute of Biosciences (Universidade Estadual Paulista, Rio Claro, São Paulo, Brazil).
Sawadogo et al. (2022) collected leaves of C. nardus (L.) Rendle and of “C. citratus (DC) Stapf”, as they write it, early in the morning. The leaves came from the botanical garden of IRSAT (latitude 12°25′ N, longitude 1°29′ W), in Ouagadougou, Burkina Faso. Voucher specimens were deposited at the herbarium of the Center of diversity (University Joseph KI-ZERBO, Ouagadougou). Their numbers are 17,950 for C. nardus and 17,940 for C. citratus.
What does the oil contain?
In De Toledo et al. (2016), the qualitative and quantitative composition shown in Table 1 comes from GC-MS. Relative amounts were calculated by peak-area normalisation, so the values are relative peak areas, not mass percentages. Oxygen-containing monoterpenes were the major constituents, at 90.61 %. The main compounds named in the text are citronellal (27.87 %), geraniol (22.77 %), geranial (14.54 %), “citronellol” (11.85 %) and neral (11.21 %).
The text writes “citronellol” where Table 1 writes “β-citronellol”, for the same value of 11.85.
Which compounds does the study’s table list?
The table below gives each compound as Table 1 of De Toledo et al. (2016) writes it, with its “Concentration (%)” value, a relative peak area for this one oil.
| Compound (as written in Table 1) | Concentration (%), relative peak area | Fact |
|---|---|---|
| β-myrcene | 0.09 | |
| n-octanal | 0.09 | |
| “D-limonene” | 2.47 | |
| cis-ocimene | 0.27 | |
| trans-ocimene | 0.17 | |
| “bergamal” | 0.37 | |
| linalool | 0.53 | |
| α-pinene oxide | 0.11 | |
| trans-rose oxide | 0.14 | |
| neo-isopulegol | 0.41 | |
| citronellal | 27.87 | |
| cis-4-decenal | 0.09 | |
| Decanal | 0.46 | |
| β-citronellol | 11.85 | |
| Neral | 11.21 | |
| geraniol | 22.77 | |
| geranial | 14.54 | |
| “citronellol acetate” | 0.31 | |
| geranyl acetate | 0.26 | |
| “β-cariophyllene” | 1.28 | |
| α-humulene | 0.12 | |
| γ-cadinene | 1.60 | |
| δ-cadinene | 0.36 | |
| citronellyl butyrate | 0.24 | |
| elemol | 0.11 | |
| “cariophyllene oxide” | 0.55 | |
| trans-cadinol | 0.16 | |
| α-muurolol | 0.30 |
How does the table group the compounds into classes?
| Class line (as written in Table 1) | Concentration (%), relative peak area | Fact |
|---|---|---|
| Monoterpene hydrocarbons | 3.00 | |
| Oxygen containing monoterpenes | 90.61 | |
| Sesquiterpene hydrocarbons | 3.36 | |
| Oxygen containing sesquiterpenes | 1.12 | |
| Other compounds | 0.64 | |
| Total identified | 98.73 |
How were the compounds identified?
De Toledo et al. (2016) ran GC-MS with electron-impact ionisation and a quadrupole mass analyser. The capillary column was an HP-5ms, 5 % diphenyl-polydimethylsiloxane, 30 m × 0.25 mm, with a 0.25 μm film. For this analysis, the oil was solubilised 1:100 (v/v) in chromatographic-grade hexane.
Helium was the carrier gas, at 1.00 mL/min (8.2 psi) and a linear velocity of 36.6 cm/s. The injector was at 250 °C, with 1 μL injected at a split ratio of 1:100. The oven program ran from 60 to 246 °C at 3 °C/min, over 62 min. The transfer line was at 280 °C, the detector at 150 °C, and the ionisation energy was 70 eV.
The authors identified the components by comparing their mass spectra with the NIST mass-spectral library, version 2.0, 2012 (243,893 compounds), and with literature data. They also calculated arithmetic retention indices by linear interpolation against the retention times of a C7–C30 n-alkane series. These indices were compared with published retention index values.
How was citronellal measured on its own?
De Toledo et al. (2016) also ran GC-FID, on an RTX-5MS capillary column of 5 % diphenyl-polydimethylsiloxane, 30 m × 0.25 mm, with a 0.25 μm film. Nitrogen was the carrier gas, at 1.00 mL/min (8.2 psi) and a linear velocity of 36.6 cm/s. The injector was at 250 °C, with 1 μL injected at a split ratio of 1:30. The oven rose from 70 °C to 180 °C at 4 °C/min, then from 180 °C to 250 °C at 10 °C/min.
In the same analysis, total run time was 34.5 min and the detector was at 280 °C. Citronellal was identified by its retention time and quantified by external calibration. The calibration solutions of a (+/−)-citronellal standard (≥ 95 % purity) in chromatographic-grade hexane were at 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0 and 5.0 mg/mL. The oil was solubilised 1:100 (v/v) in chromatographic-grade hexane, and all analyses were performed in triplicate.
By GC-FID, the retention time of citronellal was 11.2 min, according to De Toledo et al. (2016). The calibration curve had the equation y = 571,529.9016x − 102,555.3281, with an R² of 0.99976. The citronellal concentration in the oil, determined this way, was 282.5 mg/mL. This figure is a mass per volume of oil, a different quantity from the relative peak areas of Table 1.
What do the specifications of the assessed oil set?
In the EFSA FEEDAP opinion (2024), five components make up the specifications of the assessed oil. Citronellal was selected as the phytochemical marker. Table 1 of the opinion gives each specification, with the mean and range measured by GC–MS in four batches, in % GC area.
| Component (as written in Table 1) | Specification (% GC area) | Mean of four batches | Range | Fact |
|---|---|---|---|---|
| Citronellal | 3–6 | 4.61 | 3.92–5.12 | |
| Geraniol | 15–23 | 17.22 | 16.17–18.33 | |
| Citronellol | 3–8.5 | 4.37 | 3.95–4.57 | |
| Limonene | 7–11.5 | 6.42 | 5.75–7.18 | |
| (E)-Methyl isoeugenol | 7–15 | 12.13 | 9.89–13.89 | |
| Total | 44.733 | 43.27–46.33 |
For the Total line, the range gives the lowest and highest sums of the five components in the batches analysed. The four batches complied with these specifications by GC–MS, except for limonene, which was below the specification. The analysis of one batch by GC-FID, expressed as % GC area, showed compliance with the specification.
How fully is the composition of the assessed batches known?
The applicant provided a full GC–MS characterisation of the same four batches, according to the EFSA FEEDAP opinion (2024). Up to 104 constituents were detected, all identified, accounting on average for 100 % (99.9–100.1 %) of the GC area. Besides the five specified components, 28 others were each above 0.5 %. Together, these 33 compounds accounted on average for 92.9 % of the GC area (range 91.9–93.3 %).
The remaining 71 compounds, each between 0.02 and 0.49 %, accounted on average for 7.1 % of the total GC area. On the available characterisation data, the opinion considers this oil a “fully defined mixture”. Methyleugenol, which is not part of the specifications, was measured by GC–MS at a mean of 0.57 % GC area (range 0.41–1.00). The analysis confirmed its presence in all four batches.
Which compounds were found in an oil from Burkina Faso?
Sawadogo et al. (2022) analysed their oils on an Agilent gas chromatograph with a DB-5 capillary column (30 m × 0.25 mm, 0.25 μm film). It carried a flame ionisation detector (FID) interfaced with an Agilent model 5975 mass spectrometer. The oven rose from 60 °C to 165 °C at 8 °C/min, then to 280 °C at 20 °C/min, with a 1 min post-run at 280 °C. Helium flowed at 1.0 mL/min, and 1 μL (1/100 in acetone) was injected in split mode (1:150), injector at 250 °C, detector at 280 °C.
The mass spectrometer worked in electron-impact mode at 70 eV, with the electron multiplier at 1500 V and the ion source at 230 °C. Spectra were acquired in scan mode over m/z 33–450. Six compounds were identified in the C. nardus oil. Citronellal (41.7 %) and geraniol (20.8 %) were predominant.
The table below gives the “Cymbopogon nardus (%)” column of Table 2 of the article, each compound as written there.
| Compound (as written in Table 2) | Retention index | Cymbopogon nardus (%) | Fact |
|---|---|---|---|
| Citronellal | 1158 | 41.7 | |
| Citronellol | 1230 | 8.0 | |
| Geraniol | 1253 | 20.8 | |
| β-Elemene | 1372 | 11.0 | |
| α−Copaene | 1390 | 3.7 | |
| Hedycaryol | 1520 | 7.4 | |
| Hydrocarbon sesquiterpenes | 14.7 | ||
| Total | 92.6 |
In the same column, the Neral row (retention index 1242) and the Geranial row (1268) carry “-”. The EFSA opinion and Sawadogo et al. (2022) give different citronellal figures: a mean of 4.61 % GC area in the four assessed batches, against 41.7 in this column.
How was the oil obtained, and what yield was reported?
De Toledo et al. (2016) hydrodistilled 150 g of fresh C. nardus leaves in a Clevenger-type apparatus. The apparatus was attached to a 3 L round-bottom flask holding 1500 mL of deionised water. Residual water was separated from the oil by freezing.
The authors print the yield as “0, 7% (w/w)”. It is one value for this batch, and the text does not state the mass it is based on. The oil was stored under refrigeration until chemical analysis and biological tests.
The EFSA FEEDAP opinion (2024) states that the assessed oil is extracted by steam distillation from the fresh or partly dried leaves of C. nardus. The volatile constituents are condensed, then separated from the aqueous phase by decantation.
How much oil did fresh leaves yield in Burkina Faso?
Sawadogo et al. (2022) hydrodistilled “a mass of 400 g” of fresh leaves of C. nardus and C. citratus for 3 h in a Clevenger-type apparatus. The recovered oils were dried over anhydrous sodium sulfate and stored in darkness before use. The yield of each oil was calculated on a fresh-weight basis. For C. nardus, Table 1 of the article gives 1.37 ± 0.18, in “Yield % (w/w)”.
Of the oils in that study, C. nardus gave the highest yield and C. citratus the lowest.
What does the ISO catalogue list under ISO 3849?
The ISO catalogue entry for ISO 3849:2003 gives the following details.
| Item | ISO 3849:2003 | Fact |
|---|---|---|
| Publication date | 2003-03 | |
| Edition | 2 | |
| Number of pages | 6 | |
| Stage, on the archived copy of 8 November 2024 | “International Standard confirmed”, 90.93 | |
| Technical committee | ISO/TC 54 |
In the “Life cycle” section of that entry, ISO 3849:1981 appears under “Previously”, marked “Withdrawn”. ISO 3849:2003 appears under “Now”, marked “Published”. On the archived copy of 8 November 2024, the entry states that this publication was last reviewed and confirmed in 2020. The same copy states that “this version remains current”.
The ISO catalogue also lists ISO 3849:1981 as “[Withdrawn]”, at stage 95.99, under committee ISO/TC 54. The title of that entry is “Oil of Ceylon citronella”, and it names no botanical species.
What hazards are documented for this oil?
In the EFSA FEEDAP opinion (2024), the applicant provided no specific data on the safety of the additive for users. It did provide safety data sheets for citronella oil from C. nardus, in which concerns for users were identified. According to footnote 44, the sheet classifies the oil “Serious eye damage/irritation (H318; category 1)” and “skin corrosion/irritation (H315, category 2)”. It adds “skin sensitisation (H317, category 1)”; all three follow the criteria of Annex I of Regulation 1272/2008/EC (CLP).
For methyleugenol among the ingredients, the same sheet states “Suspected of causing genetic defects (H341)” and “Suspected of causing cancer (H351)”. The FEEDAP Panel concludes that the oil “should be considered as irritant to skin and eyes, and as a dermal sensitiser”. According to the opinion, unprotected users may be exposed to methyleugenol when the essential oil is handled.
See also: Palmarosa, Lemongrass, hydrodistillation, extraction yield calculator, drops ↔ ml converter, compare with Lemongrass (Cymbopogon citratus).
References
- Govaerts R. (ed.), World Checklist of Vascular Plants (WCVP), Royal Botanic Gardens, Kew — https://sftp.kew.org/pub/data-repositories/WCVP/wcvp.zip
- De Toledo L.G., Ramos M.A.D.S., Spósito L., Castilho E.M., Pavan F.R., Lopes É.D.O., Zocolo G.J., Silva F.A.N., Soares T.H., dos Santos A.G., Bauab T.M., De Almeida M.T.G., « Essential Oil of Cymbopogon nardus (L.) Rendle: A Strategy to Combat Fungal […] Caused by Candida Species », International Journal of Molecular Sciences 2016;17(8):1252, doi:10.3390/ijms17081252 — https://www.ebi.ac.uk/europepmc/webservices/rest/PMC5000650/fullTextXML
- International Organization for Standardization (ISO), online catalogue, ICS 71.100.60 « Essential oils », archived copy of 19/03/2026 (Internet Archive) — https://web.archive.org/web/20260319152944id_/https://www.iso.org/ics/71.100.60/x/
- International Organization for Standardization (ISO), catalogue record « ISO 3849:2003 », edition 2, 2003-03, archived copy of 08/11/2024 (Internet Archive) — https://web.archive.org/web/20241108223737id_/https://www.iso.org/standard/32034.html
- EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP), « Safety and efficacy of a feed additive consisting of an essential oil derived from the leaves of Cymbopogon nardus (L.) Rendle (citronella oil) for use in all animal species (FEFANA asbl) », EFSA Journal 2024;22(5):e8790, doi:10.2903/j.efsa.2024.8790 — https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11112456/fullTextXML
- Sawadogo I., Paré A., Kaboré D., Montet D., Durand N., Bouajila J., Zida E.P., Sawadogo-Lingani H., Nikiéma P.A., Nebié R.H.C., Bassolé I.H.N., « Antifungal and Antiaflatoxinogenic Effects of Cymbopogon citratus, Cymbopogon nardus, and Cymbopogon schoenanthus Essential Oils Alone and in Combination », Journal of Fungi 2022;8(2):117, doi:10.3390/jof8020117 — https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8878799/fullTextXML