What is a residual solvent in a cannabis extract?

A residual solvent is a volatile organic compound left over from the extraction or purification process. Solvents like ethanol, hexane or isopropanol are used to pull compounds from plant material, and trace amounts can remain in the finished extract if evaporation and purification steps aren't fully validated. Pharmacopoeial guidelines (ICH Q3C) classify these by toxicological risk and set permitted concentration limits.

How is residual solvent testing done in practice?

The standard method is headspace gas chromatography, either with mass spectrometry (GC-MS) or flame ionisation detection (GC-FID). A sample is heated in a sealed vial; volatile compounds move into the headspace above the sample and are injected into the chromatograph. GC-MS is preferred for complex matrices like cannabis extracts because the mass spectrum can distinguish residual solvents from terpenes that might otherwise co-elute.

Do all cannabis extracts sold in Australia need residual solvent testing?

Products manufactured to GMP standard and supplied through TGA-regulated pathways — including the Special Access Scheme and Authorised Prescriber arrangements — must comply with manufacturing quality requirements that include residual solvent limits consistent with ICH Q3C. A Certificate of Analysis (CoA) from a NATA-accredited laboratory should accompany regulated product.

Why do terpenes complicate residual solvent analysis?

Several cannabis terpenes — including linalool, alpha-pinene, and terpinolene — have boiling points and gas chromatography retention times that can overlap with common Class 3 solvents depending on the analytical column and temperature programme used. This creates potential for false positives unless mass spectrometry confirmation is used to verify the identity of each peak by its mass spectrum.

What should I look for on a Certificate of Analysis regarding solvents?

A well-prepared CoA lists each solvent relevant to the manufacturing process with a quantitative result (in ppm) alongside the ICH Q3C limit for that solvent class. It should name the analytical method (e.g. USP ), the detection technique, and the accreditation status of the testing laboratory. A simple 'Pass/Fail' with no numbers tells you much less about product quality and batch consistency than quantitative data does.

Residual Solvent Analysis

By Tom Whitfield · 1 July 2026 · 7 min read

The solvent that never quite leaves

Somewhere between the raw plant material and the finished extract, a solvent did the heavy lifting. Ethanol, hexane, supercritical CO₂, isopropanol , these are the workhorses of botanical extraction. Most evaporate cleanly under controlled conditions. But "most" and "cleanly" are not the same as "all" and "completely," and that gap is exactly what residual solvent analysis is designed to measure.

I came across a surprisingly thin shelf of Australian-specific literature on this when I was pulling together references for a colleague's piece on standardised extracts. The international pharmacopoeial literature is extensive; the local regulatory framing, less so. Worth pulling apart.

What residual solvents actually are

Residual solvents are volatile organic compounds (VOCs) that remain in a pharmaceutical or botanical extract after the manufacturing process. They're not there by design , they're artefacts of extraction and purification. Even small concentrations can carry risk depending on the compound, so pharmacopoeial bodies classify them into tiered categories based on toxicological profile.

The International Council for Harmonisation (ICH) guideline Q3C, which the TGA applies to medicines regulated in Australia, defines three classes:

Class 1 solvents , known or suspected human carcinogens, or environmental hazards. Benzene and carbon tetrachloride sit here. These should be avoided altogether in manufacturing; where unavoidable, limits are set extremely low (benzene at 2 ppm).
Class 2 solvents; non-genotoxic animal carcinogens, or possible causative agents of irreversible toxicity. Hexane (290 ppm), acetonitrile (410 ppm) and methanol (3,000 ppm) are in this bracket. Permitted daily exposure (PDE) limits apply.
Class 3 solvents, low toxic potential at current permitted levels. Ethanol, isopropanol, acetone and ethyl acetate fall here. The general limit is 5,000 ppm unless otherwise specified.

Supercritical CO₂, often cited as the "cleanest" extraction route, sits outside the solvent classification framework entirely, CO₂ is not a VOC solvent in the pharmacopoeial sense, though it doesn't eliminate the possibility of Class 2 co-solvents being used to improve polarity.

Why cannabis extracts get particular scrutiny

Cannabis extracts; whether they ultimately contain cannabidiol, tetrahydrocannabinol, or a broader range of cannabinoids including minor compounds like cannabigerol, are frequently produced using hydrocarbon or alcohol solvents at industrial scale. Butane hash oil (BHO) production, still common in some markets, uses Class 3-bordering solvents under conditions that aren't always controlled to pharmaceutical standards.

The distinction matters because Australia's TGA-regulated medicinal cannabis supply chain requires compliance with Good Manufacturing Practice (GMP). GMP-compliant facilities must validate their solvent removal processes and test finished product against established limits. But not every cannabinoid-containing product on a consumer's radar was manufactured under those conditions, and that's where the gap lives.

Full-spectrum and broad-spectrum extracts present a particular analytical challenge: they contain a wide array of compounds, including terpenes that are themselves volatile. Distinguishing a residual solvent peak from a naturally occurring terpene; say, myrcene or beta-pinene, in a gas chromatography run requires careful method validation. It's one of those situations where the chemistry is genuinely tricky, not just bureaucratically complicated.

The analytical method: headspace GC-MS

The gold standard for residual solvent analysis is headspace gas chromatography coupled with mass spectrometry (HS-GC-MS), or sometimes with flame ionisation detection (HS-GC-FID). Headspace sampling means the sample is heated in a sealed vial; volatile compounds partition into the gas phase above the sample matrix, and that vapour is injected into the chromatograph rather than the matrix itself. This protects the column from non-volatile matrix components and concentrates the analytes of interest.

The United States Pharmacopeia (USP) Chapter <467> and the European Pharmacopoeia (Ph. Eur.) 2.4.24 both describe validated headspace GC procedures for residual solvents. These are the methods Australian TGA-licenced manufacturers typically reference, often alongside ICH Q3C.

Method validation for cannabis matrices is not trivial. The lipid-rich, waxy character of many extracts can cause matrix effects, suppression or enhancement of analyte signals; that require matrix-matched calibration or standard addition methods to correct. I'd argue that laboratories testing cannabis products should be required to publish their validation data alongside CoA (Certificate of Analysis) results, not just the final numbers. The number without the method validation is not as informative as it looks. Plenty of labs don't make that easy for buyers to verify.

Regulatory context in Australia

The TGA's requirements for medicinal cannabis products flow through the Therapeutic Goods Act 1989 and the associated manufacturing principles, which incorporate GMP standards consistent with the Pharmaceutical Inspection Co-operation Scheme (PIC/S) PE 009. Products listed or registered on the Australian Register of Therapeutic Goods (ARTG) must demonstrate compliance with these standards, including acceptable residual solvent profiles.

Access pathways for unapproved medicinal cannabis, via the Special Access Scheme or through an Authorised Prescriber, don't remove the manufacturing quality requirement. The product itself must still have been manufactured to an appropriate standard; the access pathway changes how a prescription reaches the patient, not how the extract was made.

For isolate-based products, residual solvent risk is often lower in practice; purification steps involved in producing a high-purity isolate tend to drive solvent concentrations down, but that's a generalisation, not a guarantee. A poorly executed crystallisation or winterisation step can concentrate solvents unexpectedly.

Terpene interference and method specificity

Here's where things get genuinely interesting from an analytical standpoint. Several terpenes found in cannabis and other botanical extracts have boiling points and retention times that overlap with Class 3 solvents under certain GC conditions. Linalool, alpha-pinene, and terpinolene are all candidates for co-elution issues depending on column chemistry and temperature programming.

A false positive for a residual solvent is a real problem, it can hold up product release, trigger regulatory enquiries, and cost a facility significant time and money. A false negative is worse from a safety standpoint. Good method development sets the two concerns at roughly equal weight. The mass spectrometry component of GC-MS is the safeguard here: the mass spectrum of a terpene and the mass spectrum of, say, isopropanol are quite different, so MS confirmation can resolve co-elution artefacts that GC-FID alone would miss.

What a CoA should tell you

A legitimate Certificate of Analysis for a GMP-produced extract should report residual solvent results against the ICH Q3C limits for each solvent relevant to the manufacturing process. It should name the analytical method (typically USP <467> or equivalent), state the instrument type, and ideally reference the accreditation of the testing laboratory; in Australia, look for NATA accreditation (National Association of Testing Authorities).

Results reported simply as "Pass" with no quantitative values are, in my experience chasing references, less useful than they appear. A quantitative result, even one well below the limit, provides much more information about process consistency over time. Batch-to-batch trending of residual solvent concentrations is a legitimate process analytical technology (PAT) tool.

Between that and the terpene analysis on a full-spectrum extract, a thorough CoA for a complex cannabis product can run to several pages. That's not bureaucratic excess. That's what rigorous characterisation of a complex botanical matrix actually looks like.

Sources

, Tom Whitfield, Health librarian, references & fact-checking

Common questions

What is a residual solvent in a cannabis extract?: A residual solvent is a volatile organic compound left over from the extraction or purification process. Solvents like ethanol, hexane or isopropanol are used to pull compounds from plant material, and trace amounts can remain in the finished extract if evaporation and purification steps aren't fully validated. Pharmacopoeial guidelines (ICH Q3C) classify these by toxicological risk and set permitted concentration limits.
How is residual solvent testing done in practice?: The standard method is headspace gas chromatography, either with mass spectrometry (GC-MS) or flame ionisation detection (GC-FID). A sample is heated in a sealed vial; volatile compounds move into the headspace above the sample and are injected into the chromatograph. GC-MS is preferred for complex matrices like cannabis extracts because the mass spectrum can distinguish residual solvents from terpenes that might otherwise co-elute.
Do all cannabis extracts sold in Australia need residual solvent testing?: Products manufactured to GMP standard and supplied through TGA-regulated pathways — including the Special Access Scheme and Authorised Prescriber arrangements — must comply with manufacturing quality requirements that include residual solvent limits consistent with ICH Q3C. A Certificate of Analysis (CoA) from a NATA-accredited laboratory should accompany regulated product.
Why do terpenes complicate residual solvent analysis?: Several cannabis terpenes — including linalool, alpha-pinene, and terpinolene — have boiling points and gas chromatography retention times that can overlap with common Class 3 solvents depending on the analytical column and temperature programme used. This creates potential for false positives unless mass spectrometry confirmation is used to verify the identity of each peak by its mass spectrum.
What should I look for on a Certificate of Analysis regarding solvents?: A well-prepared CoA lists each solvent relevant to the manufacturing process with a quantitative result (in ppm) alongside the ICH Q3C limit for that solvent class. It should name the analytical method (e.g. USP <467>), the detection technique, and the accreditation status of the testing laboratory. A simple 'Pass/Fail' with no numbers tells you much less about product quality and batch consistency than quantitative data does.