Rick Simpson Oil (RSO) – Solvent Choice, Purity Grades and Recovery
Rick Simpson Oil, usually shortened to RSO, is a highly concentrated full spectrum cannabis extract taken orally from a syringe. It is made with a volatile solvent that then has to come back out. Since that never fully succeeds, the choice of solvent is not a matter of preference: whatever is in it ends up proportionally in the product and is swallowed.
What RSO is and where the name comes from
RSO names an origin, not a recipe. The term goes back to the Canadian engineer Rick Simpson, who publicised the method from 2003 onwards in connection with his own skin cancer diagnosis. Neither the starting material nor the solvent nor the soaking time is fixed, so two preparations under the same name can differ widely. The almost black colour comes from chlorophyll, pigments and waxes picked up during long soaking at room temperature; it says something about how the extract was made and nothing about its quality.
How the oil is made
Four steps: wash, separate the plant material, filter, recover the solvent. The heat of that last step also converts the acidic forms, which is why RSO is orally active without further heating; the conditions of that conversion are covered in the article on decarboxylation. Cold ethanol and short washes give a lighter, cannabinoid-rich extract. Warm or prolonged soaking pulls in more chlorophyll, waxes and water-soluble compounds and then calls for winterisation. The green tint can be removed afterwards: the liquid extract runs through an activated carbon column, known in processing as color remediation, and carbon binds chlorophyll reliably. How much cannabinoid content goes with it is disputed. A 2021 review cites up to half, figures from processing practice sit closer to 15 %, and with a carbon selected for the purpose and dosed sparingly the cannabinoid fractions can be kept close to unchanged. Only the direction is certain: carbon does not bind colour alone. Washing cold and briefly in the first place is cheaper than any decolourisation.
The solvent question is first a toxicity question
The only published chemical study of the common preparation routes is by Romano and Hazekamp from 2013. It compared ethanol, naphtha, petroleum ether and olive oil on the same starting material and reached a clear result: ethanol and olive oil capture the full range of cannabinoids and terpenes and are safe to consume. The naphtha and petroleum ether extracts, by contrast, retained substantial amounts of petroleum hydrocarbons. Neither is a defined substance; both are mixtures of varying composition whose components are partly neurotoxic and are listed by their own manufacturers as a possible cancer hazard. That expressly includes pharmacy-grade petroleum benzine sold for wound cleaning: the name sounds medical and the counter respectable, but the contents are a petroleum fraction like any other. That settles the question: undenatured ethyl alcohol at food or pharmacopoeia grade, sold as neutral alcohol, rectified spirit or pharmacy-grade ethanol.
The important warning concerns what people actually reach for. Methylated spirit is not a substitute. It is 94 to 96 % ethanol, but denatured, and since July 2013 the EU has prescribed a uniform mixture for that purpose: per 100 l of absolute alcohol, 3 l of isopropanol, 3 l of methyl ethyl ketone and 1 g of denatonium benzoate, better known as Bitrex. Isopropanol and methyl ethyl ketone are volatile and largely leave during evaporation. Denatonium benzoate is not. It stays behind and concentrates in exact proportion to how far the extract is reduced. The result is an oil carrying the most bitter substance known, and unlike a camping stove, this one gets eaten. Some countries add methanol as a further denaturant, which is a separate reason to avoid imported product.
Isopropanol sits in between. Under the European guideline on residual solvents it falls, like ethanol, into the low-toxicity class for which 50 mg per day or 5,000 ppm is treated as acceptable. For hexane the same figure is 2.9 mg, roughly one seventeenth. Isopropanol is therefore not the problem naphtha is, but it is not drinkable either and has to come out completely, whereas an ethanol residue is simply alcohol.
Purity grades and alternatives
Purity is an assurance backed by a certificate, not a label claim. The gases show this most clearly. Lighter refill is a blend of propane, n-butane and isobutane put together for burn behaviour rather than purity, and it carries longer-chain companions known in the trade as mystery oils. Against that stands instrument-grade n-butane or isobutane at a guaranteed 99.5 %, free of sulphur-based mercaptan odorants and supplied with an analysis showing the BTEX compounds benzene, toluene, ethylbenzene and xylene below a detection limit of 0.5 ppm. Even then suppliers recommend distilling the solvent before use. The point is not that gas becomes a home method, since it requires a pressure-rated closed system and is prohibited in Germany anyway. The point is the logic: without a certificate of analysis nobody knows what is in a solvent, and a hardware-store container comes with none.
The same logic applies to the alternatives, with a better outcome. Olive oil performed as well as ethanol in that same study, with no residue problem at all; the trade-off is that an oil extract cannot be concentrated by evaporation, so more of it is needed. Mechanical separation goes further still: cannabinoids sit in the glandular heads on the surface of the plant and can be separated without ever being dissolved. A controlled study on four chemovars reached a total cannabinoid content of 60.7 % using dry ice treatment and a 150 µm sieve, against 24.8 % in the fresh flower. No auxiliary substance, no residue, nothing to measure.
Recovery rather than evaporation
Where alcohol is used, the final step decides both residue and terpene retention. Every workable setup rests on one principle: the vapour is not released into the room but condensed back to liquid on a cold surface inside a closed system and collected. The solvent is then reusable, and because nothing reaches the room air, the hazard disappears at its source.
The setup is simpler than it sounds. It takes a round-bottom flask, a still head, a Liebig condenser and a receiving flask, all borosilicate glass and sold as a set by laboratory suppliers. The heat comes from a pot of water on a hotplate at 80 to 85 °C, the cooling water from a bucket via an aquarium pump rather than the tap. Ethanol boils at 78 °C, so no vacuum is needed. For anyone who would rather not handle glassware, a stainless steel pot still of the kind sold for distilling water and essential oils is the same principle ready-made. What matters is not the apparatus but that the vapour has a path into the receiver and none into the room.
A rotary evaporator is an improvement on that, not a requirement. It can pull a vacuum, which lowers the boiling point: instead of 78 °C the alcohol evaporates around 30 to 40 °C, which saves terpenes that simple distillation carries off. The operating rule is called 20/40/60, meaning coolant at 20 °C, vapour at 40 and water bath at 60, 20 K apart at each stage. Which pressure corresponds to those 40 °C is read from the solvent table of the particular device rather than guessed; in practice the most favourable working pressure sits somewhat below the tabulated value. A diaphragm pump realistically reaches about 10 mbar, and ethanol needs no more than that. Working without such a device costs terpenes and time, but nothing in terms of safety.
Open evaporation is the one variant that recovers nothing and releases the entire volume into the working space. Where it is used anyway it belongs outdoors, never in a cellar or garage, with heat applied through a water bath rather than directly and with no ignition source within reach.
What is left in the oil
No setup removes the solvent completely. The more concentrated the extract becomes, the more viscous it is and the more tightly it holds the remainder, and at that point additional heat drives off cannabinoids and terpenes rather than alcohol. A second low-temperature vacuum step therefore follows, targeting below 500 ppm in industry, while the pharmaceutical ceiling for ethanol sits at 5,000 ppm. The only reliable measurement is headspace sampling with gas chromatography. The nose sits orders of magnitude above that, with one exception: bitterness in the finished oil points to denatured alcohol, and that batch cannot be saved.
What the evidence says
The best-known claim made for RSO concerns cancer treatment. The National Cancer Institute states that highly concentrated THC and CBD oil extracts are illegally promoted as cancer cures and have not been evaluated in any clinical trial for antitumour activity or safety. Systematic reviews reach the same conclusion: effects described in preclinical work have not been followed by clinical trials in cancer patients, and the heterogeneity of the extracts prevents the literature from converging on a statement about efficacy. What is supported is the palliative use of cannabinoids, for which an oncology guideline exists. Clinically relevant as well: cannabidiol inhibits certain cytochrome P450 enzymes, the same route by which cytostatics and antiepileptics are metabolised. Then there is dose. A concentrate in the high double-digit percent range takes one to three hours to act orally, and that delay is where redosing comes from.
Legal status in Germany
Since 1 April 2024 cannabis has no longer fallen under the German narcotics act. Recreational use is governed by the Consumer Cannabis Act, medical cannabis by a separate Medical Cannabis Act. The statute contains two distinct prohibitions: manufacturing cannabis, and separately extracting cannabinoids from the plant, with only CBD extraction and analytical content determination exempt. Both carry up to three years imprisonment or a fine, and where the offence involves a so-called not insignificant quantity, defined as 7.5 g of pure THC, the law presumes an aggravated case punishable by three months to five years. With a concentrate that threshold is crossed after a few grams of product. Manufacturing follows the wording of the narcotics act and covers obtaining, producing, preparing, treating or processing, purifying and converting, with no exempt quantity.
That matters for the methods: mechanical sieving or pressing falls under the manufacturing prohibition, a solvent extraction falls under the extraction prohibition as well. For manufacturing there is at least an argument that a small private harvest ought to be exempt, since cultivation and possession are permitted and the permission would otherwise be worthless. That argument has not been settled. For the extraction prohibition it does not arise at all, because there is no permitted conduct alongside it for the ban to contradict. Standardised full spectrum extracts are available on prescription through pharmacies.
Sources
- Romano and Hazekamp 2013, Cannabinoids – Chemical evaluation of cannabis oils including Rick Simpson oil
- Hazekamp, International Association for Cannabis as Medicine – Questions and answers on cannabis oil, naphtha and residual solvents
- Composition of the euro denaturant and the volatility of denaturants, overview of methylated spirit
- Häberle Labortechnik – Product data for ethanol 94 % denatured with MEK, IPA and Bitrex
- Luna Technologies – Purity grades for butane and propane, odorants and contaminants
- Extractor Solutions – Analysis data for instrument-grade n-butane, BTEX below 0.5 ppm
- ICH Q3C R9, European Medicines Agency – Guideline on residual solvents, classes and daily exposure limits
- Büchi 20/40/60 rule for rotary evaporators, University of Wollongong – bath, vapour and coolant temperatures
- Julabo 2021 – Rotary evaporation in cannabis processing, water bath, condenser temperature and vacuum control
- BVV – Solvent recovery and vacuum purging, from simple distillation to falling-film evaporators
- Ramirez et al. 2021, Journal of Cannabis Research – Processing and extraction methods of medicinal cannabis
- Lab Alley – Activated carbon in hemp processing, decolourisation and the disputed data on cannabinoid loss
- Extraction Magazine 2019 – Polishing with silica and activated charcoal, practical figures on yield loss
- Molecules 2021 – Solventless separation with dry ice and sieving, measured cannabinoid content
- National Cancer Institute – Cannabis and Cannabinoids PDQ, health professional version
- Braun et al. 2024, Journal of Clinical Oncology – ASCO guideline on cannabis and cannabinoids in adults with cancer
- Vecerova et al. 2022, Medicina – Systematic review on the link between cannabis and anticancer treatment
- Section 2 KCanG, statutory text – prohibitions, the extraction ban and its exceptions
- Section 34 KCanG, statutory text – penalties, aggravated cases, attempt and negligence
- Federal Court of Justice, order of 10 July 2024, 3 StR 98/24 – cultivation against manufacturing
- Federal Court of Justice, order of 31 July 2024, 2 StR 204/24 – the same question expressly left open