Viscosity, Crystallisation and Cannabitriol (CBT)

The problem: viscosity and crystallisation are two different failures industry practice, not published data

Cannabinoid distillate at room temperature is a resin, not a liquid — viscosity in the range of tens to hundreds of pascal-seconds depending on cannabinoid profile and temperature, which is to say it flows like cold honey or slower. That alone is a handling problem: it will not pass a filling needle at ambient temperature, it will not wick reliably in hardware designed for e-liquid, and it traps air. Crystallisation is a separate and worse failure. A cannabinoid in a near-pure state is a crystalline solid at room temperature, held in solution in the rest of the extract; when the solution is supersaturated, or is cooled, or is seeded by a crystal already present, the cannabinoid nucleates and grows crystals inside the product. In a cartridge that means a clogged wick, an unusable device, visible crystals that a customer reads as contamination, and a fill whose remaining liquid is no longer the composition on the label because part of the active has left solution. Crystallisation in the cartridge is the single most common reason a cartridge comes back.

• Two levers drive crystallisation: concentration relative to saturation, and temperature. Cold storage and shipping in winter are triggers.
• The purer the fraction, the worse the problem. A near-pure isolate has nothing in it to keep the isolate in solution.
• Seeding matters. A single crystal carried over from an upstream crystallisation step or a dirty vessel nucleates a whole batch.
• A supersaturated fill can sit clear for weeks and then crystallise in transit, which is why stability testing under thermal cycling is not optional.
• CBD is notoriously prone to this in high-CBD formulations; THCA and other acid forms crystallise readily; the neutral cannabinoids differ substantially from each other.

Sources: Hazekamp A 2007* · MELEK hemp-science shelf 2026* · ElSohly MA 2014*

Cannabitriol: structure and structural class in vitro

Cannabitriol is a genuine minor phytocannabinoid of Cannabis, not a synthetic additive invented by the vape trade, and it belongs to its own recognised structural class. The classifications used in the constituent literature — Turner and ElSohly in the 1980 review, ElSohly and Slade in the 2005 inventory, and Hanus in the 2016 unified critical inventory — enumerate roughly a dozen phytocannabinoid types, and cannabitriol (CBT) is one of them, sitting alongside the cannabigerol, cannabichromene, cannabidiol, tetrahydrocannabinol, cannabinol, cannabielsoin, cannabicyclol and cannabinodiol types. The parent structure is a tetrahydrocannabinol skeleton in which the Δ9 double bond is gone, a Δ6a(10a) double bond is present instead, and hydroxyl groups sit at C-9 and C-10. Counting the phenolic hydroxyl of the resorcinol ring, that is three hydroxyls on one molecule, which is where the triol name comes from. The pentyl parent is C21H30O4 with a nominal mass near 346, sixteen mass units above cannabinol and thirty-two above Δ9-THC — the difference of two oxygens, which is exactly what the biogenetic relationship implies. The hydroxylated, non-aromatic C-ring character is what distinguishes the class: cannabinol resolved the same oxidative pressure by aromatising its terpenoid ring, and cannabitriol resolved it by hydroxylating.

• Class membership is a structural statement, not a pharmacological one: the CBT type is defined by the 9,10-dihydroxy Δ6a(10a) arrangement.
• Stereochemistry is real and reported: trans and cis diastereomers exist, and the trans series has both enantiomers described in the literature.
• Do not confuse CBT the cannabinoid with the same three letters used elsewhere in the trade for unrelated blended additives. The chemistry above is what the literature means by cannabitriol.
• Do not confuse cannabitriol with cannabicitran, cannabiripsol or cannabielsoin; they are separate oxidised minor cannabinoids with separate skeletons, though cannabiripsol is polyhydroxylated and is discussed adjacent to the CBT group.

Sources: Chan WR 1976* · Turner CE 1980* · ElSohly MA 2005* · Hanuš LO 2016* · ElSohly MA 2014*

Known members of the cannabitriol group contested in vitro

The constituent-inventory literature describes a small set of CBT-type compounds isolated from Cannabis, and the list is worth stating precisely because most secondary writing about CBT treats it as a single substance. The core members are the pentyl parent in its stereoisomeric forms and the propyl homolog. Beyond those, the inventories record a set of C-10 ethers, and a positional isomer with the hydroxyls at C-8 and C-9. A methodological caveat belongs with the ethers: compounds bearing a 10-ethoxy group have been isolated from ethanol-processed material, and whether they are true plant constituents or artefacts formed when ethanol adds across the enol ether during extraction is not settled in the literature. That is exactly the kind of question a shelf like this should record as open rather than resolve by assertion.

Contested — caveat. The 10-ethoxy members may be artefacts of ethanolic extraction rather than true plant constituents; the literature records them without settling the question. The membership list is assembled from constituent-inventory reviews rather than from a single authoritative nomenclature standard, and naming across sources is inconsistent.

Sources: ElSohly MA 2005* · Turner CE 1980* · Hanuš LO 2016* · Boeren EG 1979* · Chan WR 1976* · Radwan MM 2008*

Occurrence, isolation history and why it is rarely on a COA in vitro

Cannabitriol is a natural trace constituent of Cannabis. It was first reported from hemp by Obata and Ishikawa in 1966, as an isolate whose structure was not then assigned; the triol structure and its stereochemistry were established by Chan, Magnus and Watson in 1976, and further CBT-type constituents were characterised through the late 1970s and 1980s by the Mississippi group whose serial Constituents of Cannabis sativa papers built the constituent inventory the field still uses. Abundance is low — CBT sits well down the list of minor cannabinoids, at a level where it is a footnote in a whole-plant analysis rather than a quantified peak. Two practical facts follow. First, its concentration in plant material is far too low for the plant to be the commercial source of the CBT used in formulation; material offered to processors is manufactured, and this shelf does not describe how. Second, it is absent from routine certificates of analysis for a purely analytical reason: you cannot quantify what you have no reference standard for, and commercial laboratories stock standards for the cannabinoids their customers ask about. That is the same bottleneck named on the research-frontier page, and it is why a formulator can add a material to a product and then not see it on the panel that product is sold against.

• Trace natural abundance. Present in the plant, not present at a level that supports isolation as a commercial route.
• The isolation history is 1966 report, 1976 structure, then progressive characterisation of the group in the constituent-inventory series.
• Not on a standard cannabinoid panel. If you want CBT quantified you must ask a laboratory whose scope includes it and which holds the standard.
• A formulation containing CBT and assayed on a standard panel reports a lower total cannabinoid mass than the fill actually contains, because part of the fill is invisible to the method.

Sources: Obata Y 1966* · Chan WR 1976* · Turner CE 1980* · ElSohly MA 2005* · ElSohly MA 2014*

Biogenetic relationships: CBT as an oxidation relative and proposed intermediate contested in vitro

The reason cannabitriol appears in a discussion of cannabinoid transformation is that its skeleton sits on the oxidative pathway that also produces cannabinol. Δ9-THC under oxidative pressure can lose the Δ9 alkene either by aromatisation of the terpenoid ring, which gives cannabinol, or by oxygen addition across the ring, which gives hydroxylated species including the Δ6a(10a) diol arrangement of cannabitriol. Turner and ElSohly set out a decomposition pathway from Δ9-THC to cannabinol in 1979, and the same body of work situates the hydroxylated intermediates that the CBT group represents. The literature therefore describes CBT-type compounds both as isolable constituents and as species on the degradation route between THC and its aromatised and hydroxylated end products, with the Δ6a(10a)-THC skeleton as the recurring junction. It should be stated plainly that the mechanistic detail here is inferred from isolated structures and model studies rather than from a fully mapped kinetic pathway in planta, so the intermediate role is a well-grounded proposal and not a closed question. No preparative route is given here; the synthetic literature on Δ6a(10a) and hydroxylated cannabinoids exists and is cited for its existence only.

• Mass arithmetic tells the same story: Δ9-THC at nominal 314, cannabinol at 310 after aromatisation and loss of hydrogen, cannabitriol at 346 after addition of two oxygens.
• Both cannabinol and the cannabitriol group are oxidative fates of the same alkene. That is why both accumulate in aged material.
• Consequence for a processor: CBT-type compounds, like CBN, are part of what a process-history readout on an aged or heat-abused extract looks like.

Contested — caveat. The intermediate role is a proposal supported by isolated structures, mass relationships and decomposition studies, not a kinetically mapped in-planta pathway. Treat the pathway diagram in secondary sources as a hypothesis with good structural support.

Sources: Turner CE 1979* · Turner CE 1980* · ElSohly MA 2005* · Hanuš LO 2016* · Pollastro F 2018*

CBT in formulation: what it actually does and why processors meet it as a working material contested industry practice, not published data

The reason a hemp processor encounters cannabitriol is entirely practical. CBT is sold to formulators as an anti-crystallisation and viscosity-modifying agent for cannabinoid vape formulations, and in that role it is not a novelty additive but one of the few materials that addresses the crystallisation problem without also thinning the fill and without adding a non-cannabinoid substance to an inhaled product. The working rationale the trade gives is structural: CBT is a cannabinoid, so it is fully miscible with cannabinoid resin and does not introduce a new chemical class into the aerosol, and its polyhydroxylated, non-planar structure disrupts the crystal packing of the cannabinoid it is dissolved in — it is a lattice interferent rather than a solvent. Reported working inclusion in the trade sits in the low single-digit percent range of the finished formulation, with higher loads used to hold high-CBD or isolate-heavy fills in solution. A formulator adding CBT is effectively choosing to solve a physical problem with a cannabinoid instead of with a diluent, which is a defensible choice for exactly the reasons the EVALI section below makes clear.

• Function: suppresses nucleation and crystal growth in supersaturated cannabinoid fills, and lowers apparent viscosity at a given temperature.
• Reported trade inclusion: commonly cited in the low single-digit percent range, higher for isolate-heavy or high-CBD formulations.
• Advantage over a non-cannabinoid diluent: it does not introduce a new substance class into an inhaled product, and it counts toward cannabinoid mass rather than diluting it.
• Disadvantage: it is not on standard analytical panels, its supply chain characterisation is variable, and its inhalation toxicology is not published.
• Material offered as CBT in the trade is not consistently characterised. Ask for a certificate of analysis with a named method and an identified analyte, and treat uncharacterised material as unknown.

Contested — caveat. The anti-crystallisation function and the inclusion range are industry practice. No peer-reviewed quantitative rheology or crystallisation-kinetics study of CBT in cannabinoid formulations was located in this pass, and the mechanistic explanation given here (lattice interference by a non-planar polyhydroxylated cannabinoid) is a structurally reasonable account rather than a measured result.

Sources: MELEK hemp-science shelf 2026* · ElSohly MA 2005* · Hanuš LO 2016*

The published record on CBT is thin, and that is a research gap this wiki is filling in vitro

It should be said directly rather than hedged into vagueness: compared with Δ9-THC and CBD, the published literature on cannabitriol is sparse. What exists is solid — the 1966 report, the 1976 structure determination, the constituent-inventory reviews that place it and enumerate its group, and the decomposition-pathway work that relates it to the oxidative fate of THC. What does not exist, as far as this pass could establish, is a receptor-pharmacology characterisation at CB1 and CB2 comparable to what is available for CBN, a quantitative rheology or crystallisation study documenting the formulation effect the trade relies on, any human pharmacokinetics, and any inhalation toxicology. That is a gap in the record, not a reason for scepticism about the compound: CBT is a structurally characterised natural product of known formula and known class with an established commercial use. The honest framing is that it is under-documented, that the specific questions are well defined, and that they are answerable with ordinary instrumentation. Naming them is more useful than repeating that it is understudied.

• Open question 1: CB1 and CB2 binding and functional activity of (−)-trans-cannabitriol, which would settle whether it contributes pharmacologically to a formulation or only physically.
• Open question 2: quantitative rheology and nucleation-suppression measurement across inclusion rates and cannabinoid profiles — the study the trade practice implies but does not have.
• Open question 3: thermal behaviour and degradation products at vaporiser coil temperatures, which is the toxicology question that matters for an inhaled additive.
• Open question 4: an accepted reference standard and a validated chromatographic method, without which CBT stays off certificates of analysis.
• Open question 5: whether the 10-ethoxy members are plant constituents or ethanolic-extraction artefacts.

Sources: Chan WR 1976* · Obata Y 1966* · ElSohly MA 2005* · Hanuš LO 2016* · Turner CE 1980* · MELEK hemp-science shelf 2026*

The rest of the viscosity toolkit industry practice, not published data

CBT is one lever among several and a competent formulation uses the cheap ones first. Temperature is the first lever and the free one: cannabinoid resin viscosity falls steeply with temperature, which is why filling is done warm, and why a cartridge that will not draw in a cold car will draw after a minute in a warm hand. It is a handling lever, not a product-stability lever, because it does nothing about the equilibrium the fill is sitting at. Formulation composition is the second and most underrated lever: crystallisation is a saturation phenomenon, so a broad cannabinoid profile with meaningful minor-cannabinoid content resists it far better than a near-pure isolate, and the practical implication is that chasing maximum purity and then fighting crystallisation is self-inflicted. Terpene content is the third lever, and it thins the fill genuinely and cheaply, at the cost of the hardware-compatibility and thermal-degradation limits set out on the terpene-blending page. Beyond those there is the class of added viscosity agents the trade uses, and it is here that the important safety line falls.

Sources: MELEK hemp-science shelf 2026* · Hazekamp A 2007* · Sleiman M 2016* · Meehan-Atrash J 2019*

The safety line: a viscosity agent chosen for rheology alone is how EVALI happened human data

This is the load-bearing sentence on the page. Choosing an additive for an inhaled product on the basis of its rheology, its cost and its appearance, without inhalation-toxicology data, is not a theoretical risk — it is the precise decision that produced the 2019 and 2020 outbreak of e-cigarette or vaping product use-associated lung injury. Vitamin E acetate was attractive because it is cheap, food-legal, oily, and close enough to cannabinoid distillate in colour and viscosity that a cut fill still looked and poured like an uncut one. Every property that recommended it was a property of the liquid in the jar, and none of them was a property of the aerosol in a lung. The outbreak record and the subsequent chemistry are covered in detail on the carriers and diluents page and on the adulterants page. The rule that follows for a formulator is narrow and absolute: for an inhaled product, the qualifying question about any additive is what is known about inhaling it, and food-grade status, GRAS status and cosmetic approval answer a different question. CBT is discussed above as a comparatively defensible choice partly because it is a cannabinoid already present in the plant and in the product class, but the honest version of that argument includes the fact that its inhalation toxicology is also unpublished.

• The correct question for any inhaled additive: what is the published inhalation data, and what does it do at coil temperature?
• Food-safe is not inhalation-safe. The two evaluations are not related and one does not imply the other.
• Appearance matching is a commercial property, not a safety property, and a diluent that is hard to detect by eye is a diluent that is hard to detect by eye.
• If a formulator cannot answer the inhalation question for an ingredient, the ingredient is an open risk and should be recorded as one in the formulation file.

Sources: Blount BC 2020* · Krishnasamy VP 2020 · U.S. Centers for Disease Control 2020* · Lanzarotta A 2020* · Wu D 2020* · U.S. Food 2020*