Acid Catalyst Comparison for CBD Isomerization: BF3, pTSA, CSA, H2SO4 and More

The Short Answer

The catalyst you choose determines which cannabinoid you make, how fast you make it, and how much cleanup you need afterward. BF3 etherate pushes CBD toward delta-9-THC with 70-85% conversion in 1-3 hours at 60-80C. pTSA and CSA favor delta-8-THC through a longer ring closure pathway, hitting 80-95% conversion in 4-12 hours at 80-120C. Sulfuric acid is cheap and fast but generates 15-30% byproducts that complicate downstream purification. Amberlyst-15 runs solventless on a hot plate and converts 60-75% over 6-24 hours, but the heterogeneous contact limits throughput. Citric acid barely works. Every operator running isomerization at any scale needs to match the catalyst to the target cannabinoid, the available equipment, and the acceptable byproduct load. This guide is that match.

Why Catalyst Selection Is the Entire Ballgame

CBD isomerization is a ring-closure reaction. The open terpene ring in CBD cyclizes into the closed ring structure of THC. That sounds simple. It is not. The cyclization can close into delta-9-THC, delta-8-THC, delta-4(8)-iso-THC, or a dozen other positional isomers depending on three variables: the acid strength of the catalyst, the reaction temperature, and the time you give it.

Strong Lewis acids (BF3, ZnBr2) coordinate with the hydroxyl group on CBD and direct the ring closure toward the thermodynamically less stable but commercially more valuable delta-9 isomer. Brønsted acids (pTSA, CSA, H2SO4) donate a proton to the terpene double bond and favor the thermodynamically stable delta-8 isomer. This is not a preference. It is a mechanistic consequence of how each catalyst type initiates the cyclization.

The practical result: choosing the wrong catalyst for your target product means you either get the wrong cannabinoid, generate excessive byproducts, or both. A delta-9 operation running pTSA will produce mostly delta-8 and wonder why their HPLC looks wrong. A delta-8 operation running BF3 will overshoot into delta-9 and accumulate byproducts that no amount of distillation removes cleanly.

The Catalyst Comparison Table

This table compares every catalyst used in CBD isomerization at production and bench scale. No other resource puts all of these side by side with specific conditions, yields, and failure thresholds. Use it as a decision framework, not a recipe. Your starting material purity, solvent system, and target product determine which row applies to you.

Reading this table: “Conversion” means total CBD consumed, not target cannabinoid yield. A catalyst showing 80% conversion with 20% byproduct load delivers 60% usable product. The gap between conversion and usable yield is your purification burden. High-byproduct catalysts cost less up front but cost more in downstream purification steps.

Catalyst Deep Dives

BF3 Etherate: The Delta-9 Workhorse

Boron trifluoride diethyl etherate is the standard Lewis acid catalyst for delta-9-THC production. It coordinates with the phenolic hydroxyl on CBD, activating the terpene ring for electrophilic cyclization. The Lewis acid mechanism favors kinetic control, which directs ring closure to the delta-9 position rather than the thermodynamically stable delta-8.

Running BF3 correctly requires anhydrous conditions. Water decomposes BF3 etherate into boric acid and HF, killing catalytic activity and generating a corrosive byproduct that attacks glassware. The reaction solvent is typically dichloromethane (DCM) or toluene. DCM gives faster reaction rates but requires careful temperature control below its 40C boiling point. Toluene allows higher temperatures (60-80C) and is the preferred solvent for production-scale runs.

The critical window: 5-10 mol% BF3 at 65C in toluene with a 1:5 CBD-to-solvent ratio. Under these conditions, expect 75-85% conversion to delta-9-THC in 90-120 minutes. Pushing beyond 3 hours or exceeding 15 mol% catalyst loading drives the reaction past delta-9 and into unidentified byproducts that appear as mystery peaks between 5 and 8 minutes on a standard C18 HPLC column.

Quenching is non-negotiable. BF3 does not stop catalyzing when you stop heating. Add saturated sodium bicarbonate solution (1:1 volume ratio) to neutralize the acid and halt the reaction. Skip the quench and your product continues isomerizing during solvent recovery.

pTSA: The Delta-8 Standard

Para-toluenesulfonic acid is the most reliable Bronsted acid catalyst for delta-8-THC production. It donates a proton to the terpene double bond in CBD, initiating a carbocation intermediate that cyclizes preferentially to the thermodynamically stable delta-8 position.

pTSA is forgiving. The reaction window is wide: 5-10 mol% in toluene or heptane at 80-120C for 4-12 hours. Higher temperatures accelerate conversion but increase the delta-8 to delta-9 ratio (delta-9 becomes a significant byproduct above 110C). The optimal balance for maximum delta-8 with minimal delta-9 contamination is 100C for 6-8 hours at 7 mol% loading.

The monohydrate form of pTSA is easier to handle than the anhydrous form and works identically in isomerization. Use it. The water of crystallization does not affect the reaction at these concentrations.

Post-reaction cleanup: wash the organic layer with saturated sodium bicarbonate (2x), then distilled water (1x). pTSA is water-soluble and washes out cleanly, unlike Lewis acids that can form emulsions. This is the single biggest practical advantage of pTSA over BF3: the workup is simple.

CSA: The Precision Bronsted Acid

Camphorsulfonic acid behaves similarly to pTSA but offers finer control over selectivity. CSA is a chiral sulfonic acid, and while its chirality does not significantly influence the isomerization outcome (CBD-to-THC is not an enantioselective reaction), its steric bulk modulates reaction kinetics.

CSA runs slower than pTSA at equivalent loading and temperature, which is actually useful for operators who need predictable endpoints. At 10 mol% in toluene at 90C, expect 80% conversion to delta-8 in 6-8 hours with 3-5% byproducts. The byproduct profile is cleaner than pTSA at equivalent conversion because the slower kinetics reduce over-isomerization.

The tradeoff: CSA costs 2-3x more than pTSA with no significant yield advantage. It is the right choice for operators who prioritize clean HPLC profiles over cost efficiency. If your downstream distillation can handle 5-8% byproducts, save the money and use pTSA.

Sulfuric Acid: Cheap, Fast, Dirty

Sulfuric acid is the brute-force option. It works at room temperature, converts CBD rapidly (30-60 minutes at concentrated loading), and costs almost nothing. It also generates the highest byproduct load of any catalyst on this list.

The problem is specificity. Sulfuric acid is a strong, non-selective Bronsted acid. It protonates every available site on the CBD molecule, not just the terpene double bond. The result is a mixture of delta-8, delta-9, delta-4(8)-iso-THC, and degradation products. At 3-5 mol% with careful temperature control (room temp, 2-3 hours), you can limit byproducts to 15-20% of the total conversion. But most operators running H2SO4 at scale report 20-30% byproduct loads that require two distillation passes to reduce to acceptable levels.

Use sulfuric acid only if: (a) you do not care which THC isomer you produce, (b) you have distillation capacity to handle heavy cleanup, or (c) you are running bench-scale feasibility tests and want the cheapest possible proof of concept. For production, switch to pTSA or BF3.

HCl: The Underperformer

Hydrochloric acid works but underperforms every other catalyst on this list in both conversion and selectivity. The aqueous phase creates biphasic reaction conditions that limit contact between catalyst and substrate. Conversion rarely exceeds 75% even with extended reaction times, and the byproduct profile is unpredictable.

Some operators attempt to use dry HCl gas bubbled through a CBD-solvent mixture. This approach improves contact but introduces handling hazards (HCl gas is corrosive and toxic) that are not justified by the marginal improvement in conversion. For any application where HCl seems like the right choice, pTSA does the same job better, cheaper, and safer.

Zinc Halides (ZnBr2, ZnCl2): The Middle Ground

Zinc bromide and zinc chloride are Lewis acids that target delta-9-THC but with lower reactivity than BF3. They require higher loading (10-30 mol%) and longer reaction times but are easier to handle because they are stable crystalline solids at room temperature. No fuming, no decomposition from moisture at practical levels.

ZnBr2 outperforms ZnCl2 because bromide is a better leaving group, making ZnBr2 a stronger Lewis acid. At 20 mol% in DCM at 65C, ZnBr2 delivers 70-80% conversion to delta-9 in 3-4 hours. ZnCl2 under identical conditions converts 55-70% and generates more byproducts.

The zinc halide approach makes sense for operators who want delta-9 but cannot safely handle BF3 etherate (which is moisture-sensitive, corrosive, and releases toxic boron trifluoride gas). The yield penalty is 10-15% versus BF3, but the operational simplicity may justify the tradeoff in facilities without fume hoods rated for gas-phase Lewis acids.

Amberlyst-15: The Solventless Option

Amberlyst-15 is a macroreticular sulfonic acid ion-exchange resin. Unlike every other catalyst on this list, it is a solid. This means the reaction can run solventless: CBD distillate heated directly with Amberlyst-15 beads on a hot plate or in a flask with a stir bar. No solvent means no solvent recovery, no aqueous workup, and no biphasic separation.

The tradeoff is speed and conversion. Heterogeneous catalysis (solid catalyst + liquid substrate) has lower contact efficiency than homogeneous catalysis (dissolved catalyst). Expect 60-75% conversion over 6-24 hours at 100-140C. The product is predominantly delta-8, consistent with the Bronsted acid mechanism of its sulfonic acid active sites.

Amberlyst-15 is reusable. After the reaction, filter out the beads, wash with ethanol, dry, and reuse for 3-5 cycles before catalytic activity drops below useful levels. At $50-90/kg, the effective cost per batch is lower than any other catalyst after 3 reuses.

Limitation: Amberlyst-15 has a maximum operating temperature of 120C (standard grade) or 140C (dry grade). Exceeding this decomposes the resin and releases sulfonic acid into the product, contaminating your extract. Always use the dry-grade resin and monitor temperature closely.

Citric Acid: Why It Barely Works

Citric acid appears in DIY forums as a “safe” and “natural” isomerization catalyst. It is neither effective nor practical. Citric acid is a weak triprotic acid with a first pKa of 3.13. Compare that to pTSA (pKa -1.34) or sulfuric acid (pKa -3). The proton donation capacity of citric acid is orders of magnitude weaker than any effective isomerization catalyst.

At 20 mol% loading in a sealed reactor at 130C for 24 hours, citric acid converts 10-20% of CBD. The product is a mixture of delta-8 and delta-9 with no useful selectivity. The remaining 80-90% is unreacted CBD. You are paying for heat and time to produce a product that is mostly starting material.

If you see a recipe using citric acid for isomerization, the person posting it either confused it with a different acid, never ran HPLC on their product, or is measuring “success” by color change rather than actual cannabinoid conversion.

Decision Framework: Choosing the Right Catalyst

Common Failures and How to Diagnose Them

Failure 1: Low Conversion Despite Correct Conditions

Symptom: HPLC shows 30-50% unreacted CBD after the expected reaction time. Product is mostly starting material with trace cannabinoids.

Root cause: Catalyst was deactivated before or during the reaction. For BF3, this means moisture contamination: even 0.5% water in the solvent or CBD feedstock generates boric acid and kills catalytic activity. For pTSA/CSA, this means the starting material contained basic contaminants (residual sodium bicarbonate from a prior wash, alkaline adsorbents from CRC) that neutralized the acid.

Diagnostic test: Run a control reaction with known-pure CBD isolate (99%+) under identical conditions. If the control converts normally, your feedstock is contaminated. If the control also fails, your catalyst is degraded (check expiration, storage conditions, container seal).

Fix: For BF3: use molecular sieves (3A) in your solvent for 24 hours before the reaction. Verify CBD feedstock water content is below 0.1% (Karl Fischer titration). For Bronsted acids: wash your CBD distillate with dilute acid (0.1N HCl) before isomerization to strip basic contaminants, then dry over anhydrous sodium sulfate.

Failure 2: Wrong Isomer Ratio

Symptom: Running BF3 but getting mostly delta-8. Or running pTSA but getting significant delta-9. The HPLC profile does not match the expected product for your catalyst.

Root cause: Temperature is the most common variable. BF3 above 85C in toluene shifts selectivity toward delta-8 because the thermodynamic product becomes favored at higher energy input. pTSA below 80C runs so slowly that kinetic delta-9 formation competes with the intended delta-8 pathway.

Diagnostic test: Pull 0.5 mL aliquots every 30 minutes during the reaction and run HPLC. Plot delta-8:delta-9 ratio vs. time. If the ratio shifts during the reaction, your temperature control is drifting. If the ratio is wrong from the first aliquot, your catalyst loading or solvent system is the issue.

Fix: For BF3: maintain 60-70C in toluene, never exceed 80C. Use an oil bath with a PID controller, not a hotplate with a dial. For pTSA: ensure reaction temperature reaches 90-100C within the first 30 minutes and stays there. Slow heating gives CBD time to form the kinetic delta-9 product before the reaction reaches Bronsted-acid-optimal temperature.

Failure 3: Excessive Byproducts (>15% Unknown Peaks)

Symptom: HPLC shows 15-30% of total area as unidentified peaks between CBD and THC retention times, or late-eluting peaks after THC.

Root cause: Over-catalysis. Too much catalyst, too high temperature, or too long reaction time. Once all CBD is consumed, the catalyst does not stop. It continues isomerizing THC into positional isomers and degradation products. The reaction went past the optimal endpoint. See the byproducts identification guide for specific peak assignments.

Diagnostic test: Time-course HPLC (aliquots every 30-60 minutes). Plot total byproducts vs. time. The curve should plateau and then rise. The optimal endpoint is where CBD is below 5% and total byproducts are below 10%. Every minute past that point generates diminishing returns on conversion and increasing byproduct accumulation.

Fix: Reduce catalyst loading by 30% and extend reaction time to compensate. A slower reaction gives you a wider window to hit the optimal endpoint before byproducts climb. For BF3: drop from 10% to 7% and extend from 2 to 3 hours. For pTSA: drop from 10% to 5% and extend from 6 to 10 hours. Always quench immediately when target conversion is reached.

Failure 4: Emulsion During Workup

Symptom: Adding water or bicarbonate solution to quench creates a stable emulsion that will not separate, even after 24 hours.

Root cause: Lewis acid catalysts (BF3, ZnBr2, ZnCl2) form metal-organic complexes that act as surfactants. These complexes stabilize the interface between the organic and aqueous phases. The more catalyst you used and the longer the reaction ran, the worse the emulsion.

Diagnostic test: Add 5% NaCl to the aqueous phase (salting out). If the emulsion breaks within 30 minutes, the complexes are the cause. If it persists, you have a different problem (chlorophyll or wax contamination from crude feedstock).

Fix: Prevention is better than cure. Winterize your CBD feedstock before isomerization to remove waxes and lipids. Use the minimum effective catalyst loading. For persistent emulsions: centrifuge at 3,000 RPM for 10 minutes, or add celite and vacuum filter through a Buchner funnel.

Failure 5: Catalyst Not Dissolving

Symptom: Solid catalyst (pTSA, CSA, ZnBr2, ZnCl2) sits undissolved at the bottom of the flask. Reaction proceeds slowly or not at all.

Root cause: Solvent is too nonpolar. pTSA dissolves well in toluene and DCM but poorly in hexane and heptane. ZnBr2 dissolves well in DCM and ethanol but poorly in hydrocarbons. The catalyst must be in solution (or, for Amberlyst-15, in physical contact) to function.

Diagnostic test: Add 10% ethanol or methanol to the reaction mixture. If the solid dissolves and conversion rate increases, your solvent system was the problem.

Fix: Match the catalyst to the solvent. pTSA + toluene is the default pairing. BF3 + DCM or toluene. ZnBr2 + DCM. If you must use a hydrocarbon solvent (heptane, pentane), pre-dissolve the catalyst in minimal DCM and add that solution to the reaction. Do not co-solvent with alcohol above 5% volume fraction or it will quench certain Lewis acid catalysts.

Starting Material Purity Requirements

Every catalyst on this list performs worse with crude feedstock. The minimum starting material purity for reliable isomerization is 85% CBD by weight. Below that threshold, the contaminants consume catalyst, generate unpredictable byproducts, and make HPLC interpretation unreliable.

For production-scale operations targeting consistent HPLC profiles batch to batch, start with CBD distillate at 90%+ purity or CBD isolate at 98%+. The cost of higher-purity feedstock is recovered in reduced catalyst consumption, lower byproduct cleanup, and fewer failed batches.

Critical contaminants that must be removed before isomerization:

• Water: Deactivates Lewis acids (BF3 especially). Target: below 0.1%.
• Waxes and lipids: Create emulsions during workup and foul distillation columns. Winterize before isomerization.
• Chlorophyll: Co-converts into degradation products that contaminate the target cannabinoid. Run CRC or activated carbon treatment before isomerization.
• Residual solvents: Butane, ethanol, and isopropanol below 500 ppm. Higher levels dilute the reaction and shift kinetics.
• Other cannabinoids: CBN, CBC, and CBG do not interfere significantly. THC in the starting material skews your conversion calculation but does not affect the reaction.

Scale-Up Considerations

Moving from bench (1-10g CBD) to production (100g-1kg CBD) changes three things: heat transfer, mixing, and quench timing.

Heat transfer: A 1L flask on a hot plate reaches 80C in 5 minutes. A 20L reactor takes 30-45 minutes. During that ramp, the catalyst is active but the reaction is running at sub-optimal temperature. This means more time at low temperature = more kinetic (delta-9) product in Bronsted acid reactions. For pTSA targeting delta-8: preheat the solvent to reaction temperature BEFORE adding the catalyst-CBD mixture. Do not add everything cold and heat together.

Mixing: Insufficient agitation at scale creates concentration gradients. The CBD near the impeller converts faster than the CBD near the vessel walls. The result is a broad product distribution rather than a sharp peak. Overhead mechanical stirring at 200-400 RPM with a PTFE paddle is the minimum for reactors above 5L.

Quench timing: At bench scale, quenching is instant: pour the reaction mixture into bicarbonate solution. At production scale, adding quench solution to a 20L reactor takes 5-10 minutes. During that time, the un-quenched portion continues reacting. Solution: add quench solution at the same rate the reaction would consume it (approximately 0.5 L/min per liter of reaction volume), or cool the reactor to below 40C before quenching to slow the reaction enough that the quench addition time does not matter.

Frequently Asked Questions

Which catalyst produces the highest purity delta-9-THC?

BF3 etherate at 7-10 mol% in toluene at 65C for 90-120 minutes produces the cleanest delta-9 profile: 75-85% delta-9-THC with 5-10% byproducts. The key is quenching with sodium bicarbonate the moment HPLC shows CBD below 5%. Every additional minute past that point generates 1-2% more byproducts without meaningful increase in delta-9 yield.

Can I reuse any of these catalysts?

Only Amberlyst-15. It is a solid resin that is physically filtered from the product, washed with ethanol, dried at 60C for 4 hours, and reused. Expect 3-5 cycles before conversion drops below 50% of fresh activity. All liquid-phase catalysts (BF3, pTSA, CSA, H2SO4, HCl, zinc halides) are consumed or washed out during workup and cannot be recovered economically.

What is the cheapest way to convert CBD to delta-8?

pTSA monohydrate. At $20-40/kg with 5-7 mol% loading, the catalyst cost per gram of converted CBD is $0.02-0.05. The reaction runs in cheap solvents (toluene or heptane), workup requires only sodium bicarbonate and water, and byproducts stay below 8% with proper temperature control. No other catalyst matches this cost-to-yield ratio for delta-8 production.

Why does sulfuric acid produce so many byproducts?

Sulfuric acid is a non-selective strong acid. It protonates every nucleophilic site on the CBD molecule, not just the C1 terpene double bond that initiates isomerization. The result is multiple competing cyclization pathways producing delta-8, delta-9, delta-4(8)-iso-THC, abnormal cannabinoids, and ring-opened degradation products simultaneously. Selective catalysts like pTSA and BF3 have steric or electronic properties that favor one pathway over others. Sulfuric acid has no such selectivity.

Is citric acid a viable isomerization catalyst?

No. Citric acid has a pKa of 3.13, which is 4 orders of magnitude weaker than pTSA (pKa -1.34). At 130C for 24 hours, citric acid converts 10-20% of CBD. The remaining 80-90% is unreacted starting material. At the energy cost of running a sealed reactor at 130C for a full day, you would spend less money buying pTSA and finishing the reaction in 6 hours at 100C with 90% conversion.

Do I need a fume hood for these reactions?

BF3 etherate requires a fume hood. It releases boron trifluoride gas (toxic, corrosive) when exposed to moisture or heat above its decomposition temperature. HCl gas reactions require a fume hood. All other catalysts on this list can be handled safely on a bench with standard PPE (nitrile gloves, safety glasses, lab coat) and adequate ventilation. Toluene and DCM solvents always require a fume hood regardless of catalyst choice.

What happens if I use too much catalyst?

Over-catalysis drives the reaction past the optimal endpoint. CBD converts completely, then delta-8 or delta-9 continues isomerizing into positional isomers and degradation products. The HPLC shows a forest of small peaks instead of one or two clean product peaks. At 2x the recommended loading, expect byproducts to double from 8% to 15-20%. At 3x loading, the reaction can become exothermic and difficult to control, especially with BF3 and H2SO4.

How do I choose between BF3 and ZnBr2 for delta-9?

BF3 gives higher conversion (75-85% vs 60-80%) and faster reaction (1-3h vs 2-6h) but requires anhydrous technique, a fume hood rated for corrosive gases, and careful quenching. ZnBr2 is a stable crystalline solid that tolerates ambient moisture, dissolves in common solvents without fuming, and quenches cleanly. If your lab has the infrastructure for BF3, use it. If not, ZnBr2 delivers delta-9 with a 10-15% yield penalty but dramatically lower operational complexity.

The Bottom Line

The catalyst is the single decision that determines your product, your yield, and your cleanup burden. Choosing a catalyst is not a chemistry preference. It is an engineering decision with direct cost and quality consequences. Match the catalyst to your target cannabinoid, your equipment capability, and your tolerance for downstream purification. The comparison table above gives you the data to make that call. The full isomerization SOP covers the step-by-step process once you have selected your catalyst.

Ready to level up your extraction game? Contact WKU Consulting for personalized guidance on building your extraction lab.

For more deep dives into cannabis chemistry, extraction SOPs, and lab design, subscribe to the WKU Consulting YouTube channel.