Solvent residues
A solvent residue is extraction liquid left behind in an extract. To take an extract from a plant, the plant is washed with a liquid, the wanted substances pass into that liquid, and the liquid is then driven off. That liquid is called the solvent. With a liquid solvent the driving off is never complete, and some of it can stay in the product. The European Union defines this as a residue that is "unintentional, but technically unavoidable" [1]. What stays behind depends on the solvent used.
How is an extract made, and why does solvent stay behind?
The solvent takes the product, then it is taken back. Ground plant material meets the solvent in a vessel. The solvent pulls the oil, the colour or the aroma substance out of the plant. The mixture is then heated, the solvent evaporates and is collected for reuse. Most vegetable oils are made this way, with a solvent called hexane [2]. The solvent can take out up to 99% of the oil in the seed [3].
This is where the residue comes from. Recovery is never complete. Fresh hexane is added to the plant all the time. One third of what is added stays as traces in the oil and in the meal (the plant material left after the oil is removed) [4]. That is why the residue question exists.
A gas solvent escapes when the pressure drops. Liquefied gases such as propane or butane behave like a liquid under pressure. When the pressure drops they turn back into gas and escape. The extract stays behind, the solvent itself does not [2]. Supercritical carbon dioxide (CO₂) extraction works on the same principle. Supercritical is the state in which a gas, under pressure and heat, acts as a solvent the way a liquid does. No organic solvent is used, and the product comes out solvent-free [3].
The solvent also changes the extract itself. Hexane pulls more material out of the plant. From spearmint leaf, hexane gave 1.7% extract and CO₂ gave 0.5%. The two extracts differed in their constituents and in their smell [5]. Oils were also taken from six seeds with hexane, with CO₂ and by cold pressing. The fatty acid composition did not differ significantly between methods, while the CO₂ oils contained more plant sterols (cholesterol-like substances found in vegetable oils) [3].
Which method can leave what?
An oleoresin is the oily extract of a spice taken with a solvent. A tincture is the liquid obtained by steeping a plant in alcohol. A co-solvent is a small amount of ethanol added to the CO₂.
| Method | Solvent | What can remain | Where on the certificate |
|---|---|---|---|
| Hexane extraction (oil, oleoresin) | hexane | traces of hexane | the "residual solvents" or "residual hexane" line |
| Alcohol extraction (tincture, liquid supplement) | ethanol | ethanol, left in on purpose in most products as the carrier | the ethanol content line |
| Acetone or ethyl acetate extraction | acetone, ethyl acetate | traces | the "residual solvents" line |
| Supercritical CO₂ extraction | CO₂ | no CO₂ remains. Traces of ethanol if a co-solvent was used | the "residual ethanol" line, if a co-solvent was used |
| Steam distillation (essential oil) | steam | none | no line |
| Cold pressing (fixed oil) | none | none | no line |
Why can a CO₂ product have an ethanol line? CO₂ dissolves oil-like substances well and water-loving substances poorly. For this reason some extractions add a small amount of ethanol to the CO₂. To take plant sterols and vitamin E from a rapeseed oil by-product, 5% ethanol was added, and both yield and purity rose [6]. The ethanol is driven off afterwards, but a trace can remain. Where a co-solvent was used for higher yield, the name "CO₂ extract" does not close the residue question completely; it makes it smaller.
Is there a limit?
In food there are three lists. The European Union keeps food extraction solvents in three lists. For those on the first list there is no numerical limit; good manufacturing practice (a way of working that keeps residues as low as technically possible) is enough. This list holds propane, butane, ethyl acetate, ethanol, CO₂, acetone and nitrous oxide. For those on the second list, the permitted use and an upper limit are written down. For hexane the limit is 1 mg/kg in oil (one milligram in one kilogram of product) and 10 mg/kg in defatted protein products. Dichloromethane is permitted on this list only for removing caffeine and bitter or irritant substances from coffee and tea, with a limit of 2 mg/kg in roasted coffee and 5 mg/kg in tea. The third list is for solvents used in making flavourings. There, hexane is limited to 1 mg/kg and dichloromethane to 0.02 mg/kg [1]. 1 mg/kg means one gram in a tonne of oil.
The European Union limit is among the strictest. Countries set different limits for hexane in oil. The European Union's 1 ppm (one part per million, the same as 1 mg/kg) is the lowest among the countries compared [4].
In medicines and supplements there are three classes. The international pharmaceutical guideline divides solvents into three classes. Class 1 solvents are to be avoided, benzene among them. For benzene the limit is below 10 ppm. Class 2 solvents are to be limited, and each has its own limit. For methanol it is 3,000 ppm [7]. Hexane is in this class, with a limit of 290 ppm [4]. Class 3 solvents have low toxic potential. Ethanol, acetone and ethyl acetate are here, and the limit is usually 5,000 ppm, which is 0.5% [7]. These classes are written for medicines. Surveys of supplements use them as the yardstick as well [8], [9].
How much actually remains?
Residues measured in supplements are usually below the limit. In Serbia, 103 liquid herbal supplements bought from pharmacies were measured. Methanol was found in 38 products, and only one exceeded the European Pharmacopoeia limit of 0.05% (500 ppm) set for alcoholic preparations. Other solvents were found rarely and at low levels [10]. In Japan, among 29 supplements, hexane was found in one product and methanol in four, and none exceeded the limit [8].
The residue does not always come from the extraction. In the same Japanese study, benzene was found in eight black-coloured products. The raw material these products had in common was activated carbon. The study looked for benzene in the activated carbon itself and found it there [8]. The residue came not from the extract but from another ingredient.
In an unregulated market the rate came out high. In South Africa, 279 cannabis-based samples were measured between 2019 and 2021, from crude extract to finished product. 111 of them (37%) exceeded the pharmaceutical limit for at least one solvent. In sticky, thick extracts taken with a solvent, the solvent is harder to remove [9].
How is it read on the certificate?
The line is called "residual solvents" or "residual hexane". The certificate of analysis (CoA) is the measurement report that comes with each batch. The method is most often gas chromatography, usually with the headspace technique [7]. The sample is heated in a small sealed vial. The vapour that collects above it is drawn off and fed into the instrument. Each solvent comes out at its own time and its amount is read. One method measured 19 solvents at once [7]. On the certificate, the method name and the result are written side by side.
"Not detected" on its own is not enough. It means below the lowest level the instrument can see. That level is called the limit of detection, and its number must be stated. For example, "not detected, limit of detection 1 mg/kg" and "not detected, limit of detection 50 mg/kg" are not the same thing.
Whether the line is there depends on the method. An essential oil taken by steam distillation (oil carried out of the plant by steam and condensed) contains no solvent. Neither does a cold-pressed oil. In both, no line is expected. In an oleoresin taken with hexane or acetone, a line is expected. In a CO₂ extract, the absence of the line says one of two things: no co-solvent was used, or one was used but not measured. The method written on the technical data sheet (TDS, the permanent description of the product) tells the two apart.
The phrase "free of solvent residues" means something together with a measurement. The phrase is confirmed by a measurement on the certificate written with its limit of detection. Without a method and a measurement it cannot be confirmed.
References
- European Parliament and Council of the European Union (2009). Directive 2009/32/EC on the approximation of the laws of the Member States on extraction solvents used in the production of foodstuffs and food ingredients (recast), Article 1(2)(b) (definition) and Annex I, Parts I to III (permitted extraction solvents and maximum residue limits). Official Journal of the European Union. https://eur-lex.europa.eu/eli/dir/2009/32/oj
- Cravotto C., Claux O. et al. (2023). Leading Edge Technologies and Perspectives in Industrial Oilseed Extraction. Molecules. https://doi.org/10.3390/molecules28165973
- Schoss K., Glavač N. (2024). Supercritical CO2 Extraction vs. Hexane Extraction and Cold Pressing: Comparative Analysis of Seed Oils from Six Plant Species. Plants. https://doi.org/10.3390/plants13233409
- Rapinel V., Claux O. et al. (2020). 2-Methyloxolane (2-MeOx) as Sustainable Lipophilic Solvent to Substitute Hexane for Green Extraction of Natural Products. Properties, Applications, and Perspectives. Molecules. https://doi.org/10.3390/molecules25153417
- Pino J., Garcia J. et al. (1999). Comparison of Solvent Extract and Supercritical Carbon Dioxide Extract of Spearmint Leaf. Journal of Essential Oil Research. https://doi.org/10.1080/10412905.1999.9701107
- Jafarian Asl P., Niazmand R. et al. (2020). Extraction of phytosterols and tocopherols from rapeseed oil waste by supercritical CO2 plus co-solvent: A comparison with conventional solvent extraction. Heliyon. https://doi.org/10.1016/j.heliyon.2020.e03592
- Kattel K., Crist R. et al. (2026). Gas Chromatography Method for Quantitation of Residual Solvent Impurities in Nanoformulations. Methods and Protocols. https://doi.org/10.3390/mps9040110
- Takahashi M., Sugaya N. et al. (2021). Detection of Residual Solvents in Commercial Supplements Using Headspace (HS)-GC-MS. Yakugaku Zasshi. https://doi.org/10.1248/yakushi.21-00123
- Viviers H., Petzer A. et al. (2022). An assessment of solvent residue contaminants related to cannabis-based products in the South African market. Journal of Cannabis Research. https://doi.org/10.1186/s42238-022-00130-3
- Vojvodić S., Čonić B. et al. (2023). Safety assessment of herbal food supplements: Ethanol and residual solvents associated risk. Journal of Food Composition and Analysis. https://doi.org/10.1016/j.jfca.2023.105483