Cannabis extraction uses the same industrial solvents found in pharmaceutical manufacturing, metal degreasing, and circuit board cleaning. The extraction process is designed to remove them before the product reaches anyone. Residual solvent testing is the step that verifies the removal actually happened. Without it, you are taking the manufacturer’s word for it. In a market with inconsistent testing requirements, that word is not always reliable.
🧪 Lab Tested | 👩💼 Woman-Owned | 🏆 Est. 2017
How Cannabis Extraction Works
Cannabis concentrates, vape oils, and tinctures are not made by juicing a plant. They are made by dissolving cannabinoids and terpenes out of the plant material using a solvent, then separating that solvent from the cannabinoid-rich oil it now contains. The solvent’s job is to be an effective dissolver and then leave without a trace.
Different extraction methods use different solvents. Hydrocarbon extraction uses butane, propane, or a blend of the two. It is efficient and produces flavorful concentrates because hydrocarbons are good at capturing terpenes as well as cannabinoids. Ethanol extraction uses food-grade alcohol, which pulls a broader spectrum of compounds than hydrocarbons and scales more easily for large production runs. CO2 extraction uses pressurized carbon dioxide in a supercritical state, which requires no traditional solvents at all: when the pressure is released, the CO2 becomes a gas and leaves naturally, with no residue.
The extraction method matters for residual solvent risk. CO2 extraction has the lowest inherent solvent residue risk precisely because the “solvent” is a gas that dissipates completely on its own. Hydrocarbon extraction has the highest potential residue risk because butane and propane require active purging steps to remove them. Ethanol falls in between. A well-run operation using any of these methods can produce clean products. The testing is the verification, not the method itself.
Why Solvents Don’t Always Fully Evaporate
Every extraction method includes a purging step: a process of applying heat, vacuum, or both to drive the residual solvent out of the finished oil. When it works as intended, the solvent evaporates and only cannabinoids, terpenes, and other plant compounds remain. When it does not work as intended, trace amounts of solvent remain in the final product.
Several things can cause incomplete purging. Insufficient time at temperature is the most common: solvent molecules that have bonded with cannabinoid matrices release more slowly than free solvent, and cutting the purge short leaves them behind. Equipment quality matters too: vacuum ovens that do not reach consistent pressure across their entire chamber leave hot spots and cold spots that purge unevenly. Batch size affects purge efficiency because thicker oil purges more slowly than thin oil. And the specific solvent used changes the required conditions: butane and propane have different boiling points and vapor pressures than ethanol, so the purge parameters for one do not apply to the other.
None of this is exotic knowledge. Experienced extractors know the variables. The testing requirement exists because even experienced extractors working in well-equipped facilities produce batch-to-batch variation, and there is no way to confirm a batch’s solvent levels without measuring them directly.
One counterintuitive point: A product that smells like butane is not necessarily more dangerous than one that doesn’t. Butane has a noticeable odor at low concentrations, while some more toxic solvents like hexane have higher odor thresholds. The solvents you can smell easily may not be the ones posing the greatest risk. Testing by instrument, not by nose, is the only accurate method.
What Residual Solvent Testing Tests For
Residual solvent panels test for a defined list of compounds organized by the risk they pose when present above specific thresholds. The pharmaceutical industry, which has been managing residual solvent risk in drug manufacturing for decades, developed a classification system that most cannabis laboratories now adopt. Solvents are grouped into three classes based on their human health risk profile.
| Solvent | Class | Common Extraction Use | Health Concern | Typical Action Limit |
| Benzene | Class 1 (Avoid) | Not used in cannabis extraction; present as contamination | Known human carcinogen. No safe exposure level established. | 2 ppm (should ideally be ND) |
| Chloroform | Class 1 (Avoid) | Historical solvent, not used in modern operations | Hepatotoxic (liver damaging). Potential carcinogen. | 2 ppm |
| Butane | Class 2 (Limit) | Primary solvent in BHO (hydrocarbon) extraction | CNS depressant at high concentrations. Cardiac sensitization at extreme exposure. | 5,000 ppm (varies by state) |
| Propane | Class 2 (Limit) | Used with butane in hydrocarbon extraction blends | Similar to butane. Asphyxiation risk at extreme concentrations. | 5,000 ppm (varies by state) |
| Hexane | Class 2 (Limit) | Rarely used in licensed operations; possible contamination source | Neurotoxic at chronic exposure. Causes peripheral neuropathy. | 290 ppm |
| Toluene | Class 2 (Limit) | Not intentionally used; contamination from industrial processes | Neurotoxic at high doses. Developmental toxicant at fetal exposure. | 890 ppm |
| Ethanol | Class 3 (Acceptable) | Primary solvent in ethanol extraction | Low acute toxicity. Associated with ethanol toxicity only at high doses. | 5,000 ppm (most states) |
| Isopropanol | Class 3 (Acceptable) | Sometimes used as a cleaning solvent or secondary extraction aid | Low toxicity at residual concentrations. Slight CNS effects at high exposure. | 5,000 ppm |
| Acetone | Class 3 (Acceptable) | Secondary solvent or equipment cleaning | Very low toxicity at residual levels. Naturally present in the human body. | 5,000 ppm |
| Vitamin E acetate | Cutting agent (not classified) | Not used in legitimate extraction; added as adulterant | Safe to ingest. Produces ketene when inhaled, causing acute lung injury (EVALI). | Should be ND in all products |
Class 1 solvents are associated with unacceptable human health risk and should not be present in cannabis products above trace detection limits. Class 2 solvents are associated with known toxicity and carry defined limits. Class 3 solvents are considered acceptably safe at typical residual concentrations but are still tested because “acceptable at low levels” becomes “not acceptable at high levels” if purging was incomplete.
How the Test Works: GC-MS Explained
Residual solvent testing uses a technique called gas chromatography-mass spectrometry, abbreviated GC-MS. For identifying and quantifying trace volatile compounds, GC-MS is the standard across pharmaceutical manufacturing, forensic toxicology, and environmental analysis. Cannabis labs adopted it for the same reasons: it is precise, it is specific, and it cannot be fooled by compounds that look similar.
The GC half of the process separates compounds. A small sample of the cannabis product is placed in a sealed vial and heated. The volatile compounds (solvents, terpenes, anything that evaporates at the test temperature) enter the gas phase and are injected into a long, coiled column that acts as a separation highway. Different compounds travel through the column at different speeds based on their chemical properties. By the time they exit the column, they are separated in time: compound A arrives first, compound B arrives thirty seconds later, compound C arrives a minute after that.
The MS half of the process identifies them. Each compound that exits the column enters a mass spectrometer, which breaks it into charged fragments and measures their mass-to-charge ratios. Every compound produces a characteristic fragmentation pattern, its “mass spectrum,” which functions as a molecular fingerprint. The instrument compares each detected pattern against a reference library to identify what the compound is and uses the signal intensity to calculate how much of it was present.
GC-MS can detect compounds at the parts-per-million level and, in sensitive configurations, at the parts-per-billion level. The sensitivity is what makes it useful for safety testing: the instrument can confirm the absence of solvents down to concentrations far below those associated with any health concern. A result of “ND” from a GC-MS analysis is a meaningful negative, not a guess.
How to Read Residual Solvent Results on a COA
The residual solvent section of a COA lists each tested compound, the result (either a numerical value in parts per million or “ND” for not detected), and the action limit: the threshold above which the product fails. The format looks something like this:
| Compound | Result | Action Limit | Status |
| Butane | ND | 5,000 ppm | Pass |
| Propane | ND | 5,000 ppm | Pass |
| Hexane | 12.4 ppm | 290 ppm | Pass |
| Benzene | ND | 2 ppm | Pass |
| Ethanol | 842 ppm | 5,000 ppm | Pass |
| Vitamin E acetate | ND | ND required | Pass |
- All Class 1 solvents should show ND. Benzene, chloroform, and other Class 1 solvents have action limits in the low parts-per-million range, but any detectable result warrants scrutiny. If a Class 1 solvent shows a numerical result rather than ND, confirm it is well below the action limit and consider contacting the brand.
- “ND” is better than “0.00” for contaminants. ND means the instrument did not detect the compound above its measurement threshold. “0.00” is not a standard scientific notation for this result and may indicate data manipulation or a panel that was not actually run. See the red flags guide for more on this distinction.
- Check that Vitamin E acetate is on the panel. Many basic residual solvent panels test only the ICH/USP classified solvents and do not include cutting agents. Vitamin E acetate should appear explicitly on any panel for vape products. Its absence from the panel, not just from the results, is a gap.
- Action limits vary by state. Cannabis testing regulations differ between states. A product with a result of 3,000 ppm butane may pass in one state and fail in another depending on the applicable action limit. Confirm which limits apply to the jurisdiction where the product was tested.
The EVALI Connection
The 2019 EVALI crisis is the clearest illustration of what happens when residual solvent testing is absent from the supply chain. Vitamin E acetate is not technically a solvent (it is a diluent used to extend THC oil) but its detection falls within the residual solvent testing framework because it is a volatile organic compound that a GC-MS instrument can identify and quantify.
Every batch of vape products sold through a licensed operator with a complete residual solvent panel would have had its Vitamin E acetate content detected and flagged before reaching consumers. The 2,807 hospitalizations and 68 deaths that the CDC attributed to EVALI were products of the illicit market, where no such testing existed. The gap was not a failure of technology. The technology works. The gap was the absence of a testing requirement.
For the complete account of the 2019 crisis and what it means for product safety, see The 2019 Vaping Crisis: What Happened & What We Learned About Safety.
TribeTokes residual solvent testing: All TribeTokes vape products are tested for residual solvents including Vitamin E acetate as part of the full-panel COA. Batch-specific results are publicly available at tribetokes.com/certificates-of-analysis. No Vitamin E acetate or cutting agents are used in any TribeTokes products.
Frequently Asked Questions
Residual solvents are trace amounts of the chemical solvents used during cannabis extraction that remain in the finished product after the purging step. Common solvents used in cannabis extraction include butane, propane, ethanol, and isopropanol. When the purging process is complete and successful, these solvents leave the product entirely. When it is not, they remain at measurable concentrations. Residual solvent testing quantifies these concentrations and compares them against established safety limits.
At the concentrations present in properly purged products that pass testing, residual solvents do not pose a meaningful health risk. The danger arises when purging is incomplete or when testing is absent. Class 2 solvents like butane are associated with CNS effects at significant exposure. Class 1 solvents like benzene are associated with carcinogenicity at any meaningful exposure. Testing ensures that concentrations remain below the thresholds established for each compound’s risk profile.
The classification system originated in pharmaceutical manufacturing and describes the relative health risk of each solvent class. Class 1 solvents (benzene, chloroform, others) are associated with unacceptable risk and should be absent from any cannabis product. Class 2 solvents (butane, propane, hexane, toluene) have known toxicity and carry defined limits below which they are considered acceptable. Class 3 solvents (ethanol, isopropanol, acetone) have low inherent toxicity and are permitted at higher concentrations, though they are still tested to confirm purging was complete.
CO2 extraction has lower inherent residual solvent risk because the CO2 “solvent” is a gas that dissipates naturally when pressure is released. No liquid remains to purge. However, a well-run butane extraction with proper purging and testing produces a product with equivalent safety. A poorly run CO2 extraction can still produce contaminated products from other sources (pesticides, heavy metals, microbials). The extraction method is one variable; batch-specific testing is the verification that actually confirms what is in the product.
GC-MS stands for gas chromatography-mass spectrometry. Gas chromatography separates volatile compounds by their travel speed through a coiled column at elevated temperatures. Mass spectrometry identifies each separated compound by its fragmentation pattern and quantifies it by signal intensity. The combination provides both identification (what is it?) and quantification (how much is there?) at very low concentrations. GC-MS can detect solvents at parts-per-million levels. That sensitivity confirms solvent absence at concentrations well below any health-relevant threshold.
Not technically. It is a cutting agent, not a solvent used in extraction. But it is volatile enough to be detected and quantified by GC-MS, which is the same instrument used for residual solvent testing. A complete residual solvent panel that includes cutting agents will detect Vitamin E acetate. A basic panel that tests only ICH/USP-classified solvents will not. The distinction matters: if a COA’s residual solvent section does not explicitly list Vitamin E acetate, the product has not been tested for the compound most directly linked to the 2019 EVALI outbreak.
ND means “not detected”: the compound was not present above the method’s detection limit. GC-MS has defined detection limits for each compound, and results below those limits are reported as ND rather than a numerical value. For residual solvents and cutting agents, ND is the ideal result. It means the instrument specifically looked for the compound and did not find it at any concentration the method can reliably measure.
No. Residual solvent testing confirms the safety of the extraction process. It has no bearing on the cannabinoid content of the product or its effect on a drug test. Any TribeTokes product containing Delta-8 THC, Delta-9 THC, or THCa will produce a positive result on standard drug tests regardless of solvent test results. Drug tests screen for THC metabolites, which are produced by the liver from any THC-class cannabinoid irrespective of the extraction method used to make the product.
Full-Panel Residual Solvent Testing. Every Batch. COAs Publicly Available.
Including Vitamin E acetate. Woman-owned since 2017.