Broad-spectrum cannabis extracts — full cannabinoid and terpene profiles with THC removed — are one of the most commercially valuable product categories in the industry. They let you sell into markets where THC is a hard no: international exports, drug-tested consumer segments, pediatric formulations, and jurisdictions with zero-tolerance THC policies.
I helped pioneer the first commercial broad-spectrum products using chromatographic THC removal. This was before most of the industry even understood what broad-spectrum meant. Here's what I've learned about doing it right — and what most operators get wrong.
Why Broad-Spectrum Matters
Full-spectrum extracts contain everything the plant produces, including THC. For domestic markets with 0.3% THC limits on finished products, that's workable if you're formulating at low cannabinoid concentrations. But once you need to put meaningful amounts of CBD, CBG, or CBN into a product — or sell into a market with true zero-THC requirements — full-spectrum falls apart mathematically.
Isolate solves the THC problem by stripping everything out and giving you a single purified cannabinoid. But you lose the entourage effect — the synergistic interaction between cannabinoids, terpenes, and flavonoids that drives a lot of the therapeutic value. Isolate-based products are commodities. Margins are thin and getting thinner.
Broad-spectrum sits in the middle: you keep the complex cannabinoid and terpene profile, but you selectively remove THC. That's technically harder to produce, which is exactly why it commands better margins.
The Chromatography Approach
There are several ways to remove THC from a cannabis extract. The two most common at commercial scale are chromatography and selective crystallization/winterization. We went with chromatography because it gives you the most control over what stays and what goes.
The basic principle: you dissolve your crude or distillate extract in a solvent and pass it through a column packed with stationary phase media. Different cannabinoids have different affinities for the media, so they separate as they move through the column. You collect fractions — the THC fraction gets diverted, and the remaining fractions (CBD, CBG, CBN, CBC, terpenes) get recombined.
In practice, there are two main chromatographic methods for THC removal:
Flash Chromatography
Uses silica-based media and runs at moderate pressures. It's faster, handles larger volumes, and is more forgiving on operator skill. This is what most commercial operations should start with. We ran flash chromatography systems processing 5-10 kg of distillate per day with consistent THC removal to non-detect levels.
Preparative HPLC
Gives you finer separation control but runs slower, costs more per kg processed, and requires more technical expertise. It's worth considering if you're doing specialized isolations — pulling specific minor cannabinoids for pharmaceutical-grade products — but for standard broad-spectrum production, flash chromatography is the workhorse.
Where Operators Go Wrong
Starting with Bad Input Material
Chromatography is a purification step, not a miracle. If your crude extract is full of waxes, lipids, chlorophyll, and pesticide residue, your column will foul faster, your fractions will be dirtier, and your THC removal efficiency drops. You need clean distillate going in — that means proper winterization and degumming before chromatography.
Ignoring Terpene Preservation
Most terpenes are volatile. If your upstream distillation process runs too hot, you've already lost them before you even get to chromatography. We found that short-path distillation at controlled temperatures (140-160°C on the main body) followed by separate terpene fraction collection gave us the best starting material for broad-spectrum processing. You can also strip terpenes before distillation via steam distillation and reintroduce them post-chromatography.
Underestimating Solvent Management
Chromatography uses significant volumes of solvent — typically ethanol or heptane. Your solvent recovery and recycling system needs to be sized correctly or your consumables cost will destroy your margins. We ran closed-loop solvent recovery with 95%+ recovery rates. Anything below 90% and you need to re-evaluate your system.
Skipping Validation
Every batch needs third-party COA verification for THC levels. "Non-detect" means below the limit of quantification on the analytical method — typically <0.01% for HPLC-UV. If you're selling into international markets, some jurisdictions require even lower LOQ methods. Build testing costs and turnaround time into your production schedule.
Scale Considerations
At small scale (1-5 kg/day of distillate), a single flash chromatography unit with manual fraction collection works fine. One trained operator can manage the process.
At medium scale (5-20 kg/day), you need automated fraction collection, inline UV monitoring, and a dedicated solvent recovery system. This is where most commercial operations land. Expect a capital investment of $200-400K for the chromatography equipment alone.
At large scale (20+ kg/day), you're looking at multiple parallel columns or continuous chromatography systems (simulated moving bed). Capital costs go up significantly, but your per-kg processing cost drops. This only makes sense if you have consistent, high-volume demand for broad-spectrum material.
The Bottom Line
Broad-spectrum processing is one of the highest-value capabilities a cannabis manufacturer can build. It opens markets that are closed to full-spectrum products and differentiates you from the isolate commodity race. But it requires clean upstream processes, proper equipment, and operators who understand the chemistry.
We built this capability from scratch and used it to develop products for markets from Brazil to South Korea. If you're looking to add broad-spectrum processing to your operation, the investment is worth it — but get the fundamentals right first.