What Heat Actually Does to Cannabis: The Chemistry of Turning THCA Into THC

Cannabis synthesizes the carboxylic acid forms of its major cannabinoids: tetrahydrocannabinolic acid A (THCA-A), cannabidiolic acid (CBDA), and cannabigerolic acid (CBGA). Heat drives off the carboxyl group as carbon dioxide, leaving THC, CBD, and CBG.

This page is chemistry, not clinical evidence. The measurements below come from chromatography and spectroscopy laboratories working with weighed samples in controlled ovens. None of it is a trial, and none of it establishes a therapeutic outcome.

Patients who prepare cannabis at home are running a chemical reaction with no instrumentation and no way to verify the result. Understanding what governs the reaction is the only substitute for measurement they have.

The same chemistry explains why a laboratory result for total THC and the amount of active THC in a finished preparation are different numbers, and why acid precursors keep turning up in the blood of people who only smoked.

The best characterized dataset comes from Wang and colleagues at the National Center for Natural Products Research at the University of Mississippi, published in Cannabis and Cannabinoid Research in 2016. They heated cannabis extract in a vacuum oven at 80, 95, 110, 130, and 145 degrees Celsius for intervals up to 60 minutes, then quantified nine cannabinoids by ultra-high-performance supercritical fluid chromatography with photodiode array and mass spectrometric detection.

The reaction followed first-order kinetics, meaning the rate depends on how much acid remains. From the temperature dependence of the rate constants the authors calculated an activation energy of 88 kJ/mol for THCA-A, consistent with an earlier published value of 84.8 kJ/mol. The corresponding figures for CBDA and CBGA were higher, at 112 and 109 kJ/mol.

That difference is not academic. Across the temperatures tested, the rate constants for THCA-A were approximately twice those of CBDA and CBGA. THCA decarboxylates faster than CBDA under the same conditions, an ordering independently reported by Ryu and colleagues in Molecules in 2021 using a separate high performance liquid chromatography method. Any protocol tuned to fully convert THCA will leave CBDA partially unconverted.

In the Mississippi work, THCA-A concentration approached zero in 30 minutes at 110 degrees Celsius, 9 minutes at 130 degrees, and 6 minutes at 145 degrees. Below 100 degrees the reaction did not reach completion within the 60-minute observation window at all. At 145 degrees the rate was fast enough that the reaction order became difficult to determine.

Other laboratories using different apparatus land in compatible territory. Olejar and Kinney, in the Journal of Cannabis Research in 2021, used pressurized liquid extraction with water as the solvent and found an optimum conversion temperature of 120 degrees Celsius for six minutes for THC, with 140 degrees optimal for CBD, CBG, CBC, and CBDV. A 2024 follow-up from the same group modeled the reaction across 80 to 160 degrees and one to 90 minutes and again found it to be pseudo-first order.

The practical reading is that temperature and time are interchangeable within limits, and that the limits are set by what else the heat is doing. Lower and longer favors completeness. Higher and shorter risks losing what you just made.

The Mississippi group tracked the total molar concentration of acid plus neutral form, which is the honest way to see whether material is disappearing. At 110 degrees Celsius, converting THCA-A to THC produced a 7.94 percent loss in total molar concentration. The same procedure applied to CBDA produced an 18.05 percent loss at 110 degrees and 25.2 percent at 130 degrees. CBGA was worse still, at 52.67 percent.

THCA-A was the cleanest of the three. The authors described its decarboxylation as essentially stoichiometric with no side reactions, and importantly, heating in a vacuum oven in the dark produced no detectable cannabinol, the usual oxidation product of THC. That absence is a result about vacuum ovens, not about kitchens. Remove the vacuum and add light and oxygen and the oxidation pathway reopens.

Higher-temperature apparatus does worse. Dussy and colleagues, publishing in Forensic Science International in 2005, reported that even optimized analytical equipment recovered a maximum of about 70 percent of THCA-A as THC, and that simulating the smoking process recovered only about 30 percent. A 2025 paper in Pharmaceutics examining cannabis oils found that gas chromatography inlet decarboxylation without derivatization achieved only a 50 to 60 percent conversion rate, which meant total cannabinoid content was being underestimated.

The assumption that combustion completely decarboxylates THCA turns out to be measurably wrong. Jung and colleagues, writing in the Journal of Mass Spectrometry in 2007, applied liquid chromatography tandem mass spectrometry to urine and blood serum collected during police stops of drivers suspected of drug impairment. They detected THCA-A at concentrations up to 10.8 ng/mL in urine and 14.8 ng/mL in serum.

The molar ratio of THCA-A to THC in most serum samples ran between roughly 5 and 18.6 percent. Intact acid precursor was reaching the bloodstream of people who had smoked, which is only possible if decarboxylation during smoking was incomplete.

Vaporization behaves similarly. Busardo and colleagues, in Pharmaceuticals in 2021, gave vaporized medical cannabis to 14 healthy volunteers and measured acid precursors alongside neutral cannabinoids in serum, oral fluid, and urine. THCA and CBDA appeared with time courses parallel to their neutral counterparts but at lower concentrations, because vaporization decarboxylated them only partly.

Heat also creates compounds that were not in the plant. A 2025 study in the Journal of Chromatography A ran cannabinoids through an electronic cigarette at coil powers from 45 to 105 watts and identified thermal conversion products including hexahydrocannabinol derivatives, cannabinol, delta-8-THC isomers, and a cannabidiol quinone. Most increased with coil power. Decarboxylation is one reaction among several that heat can start.

Interest in THCA as a therapeutic agent in its own right runs into a chemistry problem. McPartland and colleagues, in Cannabis and Cannabinoid Research in 2017, measured the binding affinity of THCA-A at human CB1 and CB2 receptors in transfected cells and found it weak, with approximate Ki values near 3.1 micromolar at CB1 and 12.5 micromolar at CB2. THC showed 62-fold greater affinity at CB1 and 125-fold greater at CB2.

The authors then raised the complication that matters most. Their certified THCA-A reference standard contained 2 percent THC. They concluded that, given known decarboxylation kinetics, contamination of THCA preparations with THC is close to unavoidable, and suspected that part of their own binding curve was an artifact of that contamination, meaning the true affinity of THCA-A is likely weaker still.

The implication is not that THCA does nothing. It is that any study attributing an effect to THCA has to demonstrate that the effect was not produced by the THC that formed while the material sat on a shelf or in a solvent.

Every number above was generated in a laboratory with weighed samples, controlled atmospheres, and analytical verification of the product. A home oven has none of those. Heat distribution is uneven, moisture content varies, the mass and geometry of the material change the effective temperature at the center of the sample, and there is no way to confirm the result.

Published temperature and time charts for home decarboxylation generally do not cite the apparatus their numbers came from. The values in this article come from a vacuum oven, a pressurized liquid extraction system, a gas chromatography inlet, and an electronic cigarette coil, and those four give different answers for good physical reasons.

There is also no clinical literature here at all. No trial has compared decarboxylation protocols against a patient outcome. The reasonable claims are chemical: the reaction is first order, THCA converts faster than CBDA, completeness improves at lower temperatures held longer, and material is lost at every temperature tested.

As analytical chemistry, this is solid work. Kinetic parameters derived from concentration versus time curves across five temperatures, with mass balance tracked and validated chromatographic methods, are reproducible measurements rather than estimates. Independent laboratories using supercritical fluid chromatography, high performance liquid chromatography, pressurized liquid extraction, and infrared spectroscopy have converged on the same reaction order and the same relative ordering of cannabinoid acids.

As guidance for anything a patient does, it is an extrapolation. The gap between a weighed sample in a vacuum oven and a tray of plant material in a domestic oven is large and has not been quantified in any published study.

Every laboratory used a different starting matrix. The Mississippi group heated a dried methylene chloride extract. Others heated intact flower, hemp biomass in water under pressure, or oil. Matrix affects heat transfer, and the same group showed it also affects yield: pure CBDA standard lost 13.75 percent of total molar concentration at 110 degrees while CBDA in extract lost 18.05 percent.

Vacuum ovens are not representative of ordinary heating. Excluding oxygen and light is precisely what prevents THC from oxidizing to cannabinol, so the clean stoichiometry reported for THCA-A is a best-case result, not a typical one.

The commonly quoted home decarboxylation charts, including the one in the original version of this article, generally have no attributed source. Temperature and time pairs stated to the minute should be treated as unverified unless the apparatus is specified.

This literature does not show that any particular decarboxylation protocol produces a better clinical outcome. There is no trial. The endpoints are concentrations, not symptoms.

It does not establish safe or effective home procedures. None of these studies was conducted in a kitchen, and none measured what a home preparation actually contains.

It does not resolve what happens to terpenes. Volatile compounds are lost during heating, and while that is chemically expected, the studies cited here quantified cannabinoids rather than the aromatic fraction.

It does not tell you the potency of the result. Without analytical verification, a decarboxylated preparation has an unknown THC content, which is the single most important safety limitation of preparing cannabis at home.

This reaction sits underneath several practical questions that are usually discussed separately. It explains why laboratory certificates report total THC as a calculated sum of THC plus a conversion factor applied to THCA. It explains why acid precursors show up in toxicology results from people who only smoked. It explains why a raw cannabis preparation and a heated one are chemically different products.

It also intersects with the growing market in cannabinoids that are made rather than grown. The same thermal chemistry that removes a carboxyl group can, at other temperatures and in other apparatus, produce hexahydrocannabinol, delta-8-THC isomers, and oxidation products that were never present in the plant. Heat is not a purification step.

I get asked about decarboxylation temperatures more than almost any other preparation question, and I have learned to answer it differently than people expect. The chemistry is real and the published numbers are trustworthy for the apparatus they were measured in. What I cannot give anyone is a home procedure with a predictable outcome, because nobody has measured that.

What I tell patients who insist on preparing cannabis themselves is to accept that they are producing a preparation of unknown strength, and to dose accordingly: start at a fraction of what they think is reasonable, wait, and write down what happened. The chemistry is a reason to be humble about the product, not a reason to be confident about it.

The other thing worth saying plainly is that lower and slower is the better failure mode. Underheating leaves some THCA unconverted, which mostly wastes material. Overheating destroys what you made and can generate compounds nobody asked for.

Decarboxylation of THCA to THC is a first-order thermal reaction with an activation energy near 88 kJ/mol, roughly twice as fast as the equivalent reaction for CBDA. In a controlled vacuum oven the conversion completed in about 30 minutes at 110 degrees Celsius and about 6 minutes at 145, with a measurable loss of material at every temperature. Those figures describe laboratory apparatus. They do not translate into a verified home protocol, no clinical study has compared decarboxylation methods against patient outcomes, and any preparation made without analytical testing has an unknown potency.

Take the reaction order, the relative rates, and the direction of the temperature and time tradeoff. Those are robust across laboratories. Leave behind any specific oven setting presented as optimal, because optimal was defined inside a vacuum oven or a pressurized extraction vessel and the number does not carry over. Also leave behind the idea that smoking or vaporizing fully converts the acids, since intact THCA reaches the bloodstream after both.