Cannabis microbial remediation through extraction recovers 60-80% of cannabinoid value from flower that failed regulatory testing for total yeast and mold (TYM), total aerobic count (TAC), or bile-tolerant gram-negative bacteria (BTGN). Ethanol extraction at -40C kills 99.9% of vegetative bacterial cells and mold hyphae on contact. Post-extraction winterization at -80C traps spores (2-10 microns) in the wax fraction, and sterile filtration at 0.2 microns removes everything cellular that survived. The process converts $200-2,000/lb of unsellable failed flower into compliant distillate or RSO at a processing cost of $50-200/lb. The catch: mycotoxins (aflatoxin B1, ochratoxin A) are small molecules (MW 312-404 Da) that survive extraction, survive filtration, and can concentrate 5-10x in the finished oil. Knowing which contaminants extraction eliminates and which it concentrates is the difference between remediation and contamination amplification.
Why Extraction Works for Microbial Remediation
Extraction kills or removes microorganisms through three mechanisms. First, ethanol at concentrations above 60% v/v is antimicrobial. It denatures cell membrane proteins and dissolves lipid bilayers. When you soak contaminated flower in 190-proof ethanol at -40C, the combination of solvent chemistry and cryogenic temperature destroys vegetative cells (bacteria, mold hyphae, yeast) within minutes. Second, the physical separation of extraction leaves cellular debris behind in the spent biomass. Intact spores, dead cell fragments, and biofilms stay in the plant material while cannabinoids dissolve into solution. Third, downstream processing (winterization, filtration, distillation) adds additional removal steps that are individually insufficient but collectively effective.
Hydrocarbon extraction (butane/propane) works differently. The solvent itself is not antimicrobial at processing temperatures. But the non-polar selectivity means it preferentially extracts cannabinoids and terpenes while leaving behind the water-soluble compounds that microorganisms depend on. The organisms are physically separated from the product, not chemically killed. This matters because if any step reintroduces moisture or compromises the cold chain, surviving organisms can recolonize.
CO2 extraction offers a middle ground. Supercritical CO2 (above 31.1C and 73.8 atm) has documented antimicrobial activity through cell membrane penetration and intracellular pH disruption. Several studies show 2-3 log reductions in microbial load during supercritical CO2 processing at 300+ bar. But the variable selectivity of CO2 means mycotoxin co-extraction depends heavily on pressure and temperature settings.
What Extraction Removes vs What Persists
This is the table that determines whether remediation is viable for your specific contamination profile. Not all microbial failures are equal, and extraction does not remove everything.
| Contaminant | Size | Ethanol Kills? | Winterization Removes? | 0.2um Filtration? | Concentration Risk | Net Result |
|---|---|---|---|---|---|---|
| E. coli (vegetative) | 1-3 um | Yes (>99.9%) | Partial | Yes | None | REMOVED |
| Salmonella | 0.7-1.5 um | Yes (>99.9%) | Partial | Yes | None | REMOVED |
| Aspergillus (hyphae) | 2-5 um | Yes | Yes (trapped in wax) | Yes | None | REMOVED |
| Aspergillus (spores) | 2-5 um | Partial (spores resist) | Yes (trapped in wax) | Yes | Low | REMOVED (filtration catches what ethanol misses) |
| Yeast (Candida, etc.) | 3-10 um | Yes | Yes | Yes | None | REMOVED |
| Aflatoxin B1 | 312 Da (molecule) | No (dissolves in ethanol) | No | No (passes through) | HIGH (5-10x) | CONCENTRATES |
| Ochratoxin A | 404 Da (molecule) | No (dissolves in ethanol) | No | No (passes through) | HIGH (5-10x) | CONCENTRATES |
| Endotoxins (LPS) | 10-20 kDa | No (heat-stable) | Partial | Partial (some pass) | Moderate | PARTIALLY REMOVED |
The critical takeaway: extraction eliminates everything cellular (bacteria, mold, yeast) but does not eliminate mycotoxins. If your flower failed for TYM or TAC but the mold species is non-mycotoxigenic (Penicillium chrysogenum, Cladosporium), extraction is a clean remediation path. If the contaminating species is Aspergillus flavus, A. parasiticus, or A. niger, you must test the remediated extract for mycotoxins before releasing it. The extraction solved the microbe problem but may have created a toxin concentration problem.
Extraction Method Selection for Remediation Batches
Not every extraction method handles contaminated material equally. The choice depends on the contamination type, the target product, and the downstream processing you have available.
| Method | Antimicrobial? | Mycotoxin Co-extraction | Recommended Temp | Best For | Avoid When |
|---|---|---|---|---|---|
| Cold Ethanol (QWET) | Yes (kills on contact) | Moderate (ethanol dissolves some mycotoxins) | -40 to -80C | TYM/TAC failures, distillate production | Confirmed mycotoxin-producing species |
| Hydrocarbon (BHO) | No (physical separation only) | Low (non-polar, poor mycotoxin affinity) | -20 to -40C | Mycotoxin-positive material (lower co-extraction risk) | Warm-temperature runs (microbial growth risk) |
| Supercritical CO2 | Moderate (2-3 log reduction at >300 bar) | Variable (pressure-dependent) | 35-60C at 300+ bar | High-value material where terpene preservation matters | Heavy TYM contamination (insufficient kill without downstream filtration) |
| Rosin Press | No | Low (mechanical, no solvent) | 180-220F | Not recommended for remediation | Any microbial failure (no kill step, no filtration path) |
The bottom line: cold ethanol is the default remediation method because it kills microbes on contact AND provides a clean downstream path (winterization, filtration, distillation). Hydrocarbon has a specific advantage when mycotoxins are the concern because its non-polar selectivity leaves most mycotoxins behind in the biomass. Rosin press has no remediation value because there is no kill step, no solvent contact, and no filtration path.
Post-Extraction Decontamination Protocol
Extraction alone is not enough. The extract needs three additional decontamination steps to reliably pass retesting.
Step 1: Winterization at -80C (24 hours minimum)
Standard winterization already removes waxes and lipids. For remediation batches, extend the hold time to 24 hours and use a 10:1 ethanol-to-crude ratio (by mass). The extended cold soak forces maximal precipitation of suspended particulates including spore fragments, cell wall debris, and endotoxin-laden lipopolysaccharide complexes. Filter through Whatman #1 qualitative paper first, then through a 0.45 micron membrane.
Step 2: Sterile Filtration at 0.2 Microns
This is the critical step for microbial remediation. A 0.2 micron membrane filter removes all bacteria (smallest pathogenic bacteria: Mycoplasma at 0.3 microns), all mold spores (2-10 microns), and all yeast cells (3-10 microns). Use PTFE or PVDF membrane filters compatible with ethanol. Flow rate drops significantly at 0.2 microns, so use positive pressure (10-15 PSI nitrogen) rather than vacuum to maintain throughput. Change filters when flow rate drops below 50% of initial, which indicates the filter is loading with captured particles.
For operations processing more than 5 liters of remediation crude per week, invest in a tangential flow filtration (TFF) system. TFF runs the liquid parallel to the membrane surface, continuously sweeping captured particles away. This maintains flow rate 5-10x longer than dead-end filtration and reduces filter consumption costs by 80%.
Step 3: Activated Carbon Treatment for Mycotoxins
If the contamination profile includes known mycotoxin-producing species (Aspergillus flavus, A. parasiticus, A. niger, A. ochraceus), add an activated carbon treatment step after sterile filtration. Mix 2-5% w/w activated carbon (coconut shell, 1000+ m2/g surface area) into the filtered crude at room temperature. Stir for 30 minutes. The high surface area and non-polar pore structure of activated carbon adsorbs aflatoxins with 60-90% efficiency depending on contact time and carbon quality. Filter the carbon slurry through a 0.45 micron membrane. This step adds $2-5/lb of material processed and removes 10-30% of cannabinoids along with the mycotoxins, which is the trade-off.
If you want the full step-by-step on winterization parameters, solvents, and temperature staging, we cover that entire process in our extraction training course.
The Mycotoxin Concentration Problem
This is the part that remediation equipment sellers do not talk about. When you extract 10 pounds of contaminated flower into 1 pound of crude oil, you have concentrated the cannabinoids by a factor of roughly 10x. If the flower contained Aspergillus-produced aflatoxins at 5 ppb, the crude oil may contain aflatoxins at 30-50 ppb after concentration, potentially exceeding the 20 ppb action limit used by most state testing programs.
The math is straightforward. If your starting material has a mycotoxin concentration of C ppb and your extraction yield is Y%, the theoretical mycotoxin concentration in the extract is:
C(extract) = C(flower) x (1/Y) x E
Where E is the extraction efficiency for the specific mycotoxin (how much of the mycotoxin ends up in the extract versus staying in the biomass). For ethanol extraction, E for aflatoxin B1 is approximately 0.4-0.7 (40-70% of the aflatoxin co-extracts). For hydrocarbon extraction, E drops to 0.1-0.3 (10-30% co-extraction) because aflatoxins are moderately polar and butane/propane are non-polar.
This means hydrocarbon extraction is actually SAFER for mycotoxin-contaminated material than ethanol extraction, despite ethanol being better at killing the organisms that produce the toxins. The irony: the method that kills the mold best also extracts the mold’s toxins most efficiently.
If you are processing material with confirmed mycotoxin-producing Aspergillus species:
- Use hydrocarbon extraction (lower mycotoxin co-extraction)
- Add activated carbon treatment post-extraction (adsorbs 60-90% of aflatoxins)
- Test the final extract for mycotoxins before release (not just microbial plate counts)
- If mycotoxin levels exceed state limits after remediation, the material cannot be saved. Destroy it.
State-by-State Remediation Rules
Not every state allows you to extract failed flower. Some mandate destruction. Some allow “reprocessing” but define it differently than “remediation.” The regulations are fragmented and changing. Check your state’s current rules before processing any failed material.
| State | Extraction Remediation Allowed? | Key Conditions | Retest Required? |
|---|---|---|---|
| Colorado | Yes | Failed flower can be “further processed” into concentrates. Must pass all testing on final product. | Yes (full panel) |
| Oregon | Yes | OAR 845-025-7580: failed harvest lots may be further processed. Remediation must be documented. | Yes (full panel) |
| Michigan | Yes | R 420.305: failed product may be remediated and retested. Processor must document remediation method. | Yes (full panel) |
| California | Limited | BCC regulations allow remediation for some failures but restrict methods. Irradiation and chemical treatment prohibited for inhalable products. | Yes (full panel) |
| Oklahoma | Yes | OMMA allows remediation. Must track and tag remediated batches through METRC. | Yes (full panel) |
| Washington | Yes | WAC 314-55-102: failed lots may be further processed. Must retest final product. | Yes (full panel) |
| Massachusetts | Yes (with restrictions) | CCC requires documented remediation plan. Some contaminants mandate destruction. | Yes (full panel) |
| Nevada | Limited | CCB allows remediation for microbial but not for heavy metals or pesticides. Method must be pre-approved. | Yes (full panel) |
Every state that allows remediation requires full-panel retesting on the finished product. Partial testing is not sufficient. The remediated extract must pass the same battery of tests that flower or concentrate would face: microbial, mycotoxin, heavy metals, residual solvents, potency, and terpenes. Budget $200-500 per retest per batch. If the remediated extract fails retesting, most states require destruction with no further remediation attempts.
For the full state-by-state licensing breakdown including extraction facility requirements, see our Cannabis Extraction License Requirements by State guide.
Cost-Benefit Analysis: When Remediation Pays and When It Does Not
| Variable | Small Batch (10 lbs) | Medium Batch (50 lbs) | Large Batch (200 lbs) |
|---|---|---|---|
| Failed flower value (destroyed) | $2,000-20,000 | $10,000-100,000 | $40,000-400,000 |
| Extraction processing cost | $500-2,000 | $2,500-10,000 | $10,000-40,000 |
| Additional filtration/carbon | $50-200 | $250-1,000 | $1,000-4,000 |
| Full-panel retest | $200-500 | $200-500 | $400-1,000 |
| Expected cannabinoid recovery | 60-80% | 65-80% | 70-85% |
| Total remediation cost | $750-2,700 | $2,950-11,500 | $11,400-45,000 |
| Recovered product value (distillate) | $1,200-12,000 | $6,000-60,000 | $24,000-240,000 |
The decision rule is simple: if recovered product value minus total remediation cost is positive, remediate. If the contamination includes confirmed mycotoxin-producing species and activated carbon treatment drops your cannabinoid recovery below 50%, the economics may not work for low-value biomass. High-value flower (indoor, high-potency) almost always justifies remediation. Low-value trim with mycotoxin contamination is often cheaper to destroy.
For the full decision framework covering all failure types (solvent, pesticide, microbial, potency), see our Failed Cannabis Batch? When to Remediate, Reprocess, or Destroy guide.
Common Failures and How to Diagnose Them
| Symptom | Root Cause | Diagnostic Test | Fix |
|---|---|---|---|
| Remediated extract fails retest for TYM | Filtration step skipped or filter pore size too large (>0.45 um). Spores passed through. | Check filter spec. If using 1-5 um, spores (2-5 um) will pass. | Refilter through 0.2 um PTFE membrane under 10-15 PSI nitrogen pressure. |
| Extract passes microbial but fails mycotoxin | Mycotoxins concentrated during extraction. Aflatoxin B1 co-extracted with cannabinoids. | Calculate concentration factor: (mass flower / mass extract) x mycotoxin extraction efficiency. | Add 2-5% w/w activated carbon treatment (30 min contact). If still failing, material is unsalvageable. |
| Cannabinoid recovery below 50% | Excessive activated carbon loading adsorbing cannabinoids along with mycotoxins. | Compare potency before and after carbon treatment. Loss >25% indicates overloading. | Reduce carbon to 1-2% w/w. Use coconut shell (higher selectivity for mycotoxins vs cannabinoids). Extend contact time to 45 min instead of increasing loading. |
| Remediated extract has off-flavors or dark color | Contaminated biomass had advanced decomposition. Mold metabolites (geosmin, 2-methylisoborneol) co-extracted. | Smell the crude before distillation. Musty or earthy notes indicate mold metabolite contamination. | Distillation will remove most volatile metabolites. For persistent off-flavors, add a CRC pass with 80/20 T5/B80 media stack. Accept slight terpene loss. |
| Recontamination after extraction (passes initially, fails at packaging) | Post-extraction handling introduced new microbial contamination. Non-sterile containers, exposed surfaces, or warm storage allowed regrowth. | Environmental monitoring: swab surfaces, test water for biofilm. Check storage temp (should be <5C). | Process in ISO 7+ clean environment. Use sterile containers. Store at 2-5C. Minimize time between filtration and packaging to <4 hours. |
Testing Protocol for Remediated Extracts
Do not skip testing. Remediated extracts must pass the same compliance panel as any concentrate product. But for remediation batches, add two tests that standard panels may not include:
- Mycotoxin panel: Test for aflatoxins (B1, B2, G1, G2) and ochratoxin A. Most state panels include this, but if your state only requires TYM/TAC/BTGN, add the mycotoxin panel voluntarily. A product that passes microbial but contains 50 ppb aflatoxin B1 is a liability regardless of regulatory requirements.
- Endotoxin testing (LAL assay): Bacterial endotoxins cause pyrogenic (fever-inducing) responses. Standard microbial plate counts do not detect dead-cell endotoxins. If the original contamination was gram-negative bacteria, the endotoxin load may persist in the extract even after all bacteria are killed. LAL testing costs $50-100 per sample and is especially important for inhalable products.
Sampling Protocol
Pull samples from three points in the process for chain-of-custody documentation:
- Pre-extraction: Sample the failed flower. Record the specific test results that triggered the failure (which organisms, at what CFU/g). Keep this sample in cold storage as reference.
- Post-filtration, pre-distillation: Sample the filtered crude. This confirms the filtration step worked. If this sample passes microbial, you know the kill and filtration steps are effective.
- Final product: Full compliance panel on the finished distillate, RSO, or crude. This is the sample that determines whether the remediated product can be sold.
Label every sample with the original failed batch number, remediation method used, date processed, and operator name. States that allow remediation require traceability from failed lot through remediation to final product. METRC or BioTrack entries must reflect the remediation step.
For the complete guide on residual solvent purging after extraction, including solvent-specific vacuum oven parameters, see our dedicated SOP.
Frequently Asked Questions
Can extraction remediate flower that failed for Aspergillus?
Extraction removes Aspergillus hyphae and spores effectively. Ethanol kills vegetative hyphae on contact, and 0.2 micron filtration catches spores (2-5 microns). The concern is mycotoxins: Aspergillus flavus and A. parasiticus produce aflatoxins that survive extraction and concentrate 5-10x in the extract. If the species is non-toxigenic (A. fumigatus, A. terreus in lower concentrations), extraction is a clean remediation path. If the species is A. flavus, add activated carbon treatment and test the final extract for mycotoxins before releasing it.
What is the best extraction method for microbially contaminated cannabis?
Cold ethanol extraction at -40 to -80C is the default choice. Ethanol at >60% v/v is antimicrobial, killing 99.9% of vegetative cells on contact. The cold temperature minimizes chlorophyll and wax co-extraction. For material with confirmed mycotoxin-producing mold, hydrocarbon (BHO) is actually safer because its non-polar selectivity leaves 70-90% of mycotoxins behind in the biomass, compared to ethanol which co-extracts 40-70%.
Does distillation kill remaining microbes in remediated extracts?
Short path distillation at 160-200C under vacuum (0.01-1 mmHg) kills any vegetative cells that survived extraction and filtration. However, distillation does not remove mycotoxins. Aflatoxin B1 has a decomposition temperature of 237-306C, well above standard distillation temperatures. Mycotoxins pass through the distillation apparatus with the cannabinoid fraction. Distillation is an effective final sterilization step for organisms but not for their metabolic toxins.
How much cannabinoid is lost during microbial remediation?
Standard extraction remediation (ethanol + winterization + filtration) recovers 60-80% of the original cannabinoid content. The 20-40% loss comes from: incomplete extraction of compromised trichomes (5-10% loss), winterization precipitate carrying some cannabinoids (5-10% loss), and filter cake retention (3-8% loss). Adding activated carbon treatment for mycotoxins costs an additional 10-30% cannabinoid loss depending on carbon loading. Total recovery with carbon treatment: 40-65%.
Can I remediate flower that failed for both microbial and pesticides?
Yes, but the processes must be sequenced correctly. Extract first (removes microbes), then apply adsorbent treatment for pesticides (see our pesticide remediation guide for adsorbent selection by pesticide class). The combined remediation will have lower cannabinoid recovery (40-55%) because both activated carbon (for mycotoxins) and specific adsorbents (for pesticides) remove some cannabinoids. Run a cost-benefit analysis before proceeding: if total recovery drops below 40%, the material may not be worth remediating.
What filter pore size is needed for microbial remediation?
0.2 microns is the standard for sterile filtration. This catches all bacteria (smallest pathogenic: 0.3 microns), all mold spores (2-10 microns), and all yeast (3-10 microns). Using 0.45 micron filters will catch most organisms but may allow some smaller bacteria through. Never use filters larger than 0.45 microns for remediation work. PTFE and PVDF membranes are compatible with ethanol and hydrocarbon solvents. Nylon membranes degrade in butane.
How do I know if my failed flower has mycotoxins?
Standard TYM testing identifies mold species but does not measure mycotoxin levels. If the lab report identifies Aspergillus flavus, A. parasiticus, A. niger, or A. ochraceus, request a separate mycotoxin panel (aflatoxins B1, B2, G1, G2 and ochratoxin A). Cost: $50-150 per sample. Most states set action limits at 20 ppb total aflatoxins. If the flower exceeds 5 ppb and you plan to extract (concentrating the toxins 5-10x), the finished oil will likely exceed 20 ppb without activated carbon treatment.
Is extraction remediation more cost-effective than irradiation?
For flower intended for sale as flower, irradiation (gamma or e-beam) is more cost-effective at $0.10-0.50/gram because it sterilizes without changing the product form. For flower intended for concentrate production, extraction remediation is the better path because you were going to extract anyway: the remediation step is built into the normal processing workflow, adding only the filtration and testing costs ($250-1,500 depending on batch size). The real comparison: irradiation keeps flower as flower (higher margin but limited by microbial failure stigma), while extraction converts failed flower to concentrate (lower per-gram margin but no market stigma attached to the finished oil).
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.
