Water-Based Cannabis Extraction: Four Methods Compared

Can hot water recover useful cannabis compounds as effectively as equipment that uses ultrasound or microwaves? The answer depends on what researchers want to collect. Producing more extract, concentrating a particular compound, and obtaining a consistent product are related goals, but they require different measurements.

A study published in Industrial Crops and Products compares four water-based extraction protocols applied to the same cannabis material.¹ Simple infusion produced the highest total yield, while ultrasound led on certain individual compounds. The findings offer a promising starting point for designing extraction around a desired chemical profile.

Why Water-Based Cannabis Extraction Matters

Extraction transfers substances from plant material into a liquid. Water is especially useful for recovering many polar compounds, whose chemical structures interact readily with it. Cannabis also contains constituents with different extraction requirements, so changing the liquid, temperature, or treatment can change the resulting mixture.

Think of brewing tea. Weighing everything left after evaporating the water would reveal how much material entered the cup. It would not tell you the amount of one particular compound. Cannabis extraction presents the same distinction between total material and chemical composition.

Using water as the extraction solvent offers an appealing research direction. Its usefulness still depends on which compounds are recovered, what further processing is needed, and how consistently the procedure performs.

How Researchers Compared Four Extraction Methods

The team studied dried leaves, stems, and flowers from female plants of the Moroccan Khardala variety. All procedures used portions of one homogenized batch, helping reduce differences in the starting material.

Infusion involved pouring boiling water over the plant powder and allowing it to stand. Soxhlet extraction repeatedly circulated hot water through the material under reflux. Ultrasound applied sound waves through a water bath, while microwave extraction supplied controlled microwave heating.

Each protocol was performed on three separate 20-gram portions. However, the operating conditions differed. Infusion and microwave extraction involved three cycles totaling 90 minutes; ultrasound totaled 180 minutes. Those procedures each used 600 milliliters of water. Soxhlet ran continuously for four hours with 200 milliliters.

These differences matter. The experiment compared complete protocols rather than isolating the effect of each technology under equal conditions. Its results cannot establish that one machine or method is universally superior.

What the Cannabis Extracts Contained

Infusion recovered 30.05% of the starting plant mass as dry extract, the highest yield. It also produced the highest readings in broad color-based tests associated with phenolics and flavonoids. Those tests provide comparative chemical indicators, rather than precise inventories of every compound present.

More targeted measurements revealed another pattern. Ultrasound had the highest measured cannflavin B concentration, at 67.12 milligrams per kilogram of crude extract, compared with 0.68 for infusion. Cannflavin B belongs to the flavonoid family, a different chemical group from cannabinoids such as CBD and THC.

Infusion had the highest measured total CBD, while ultrasound had the highest measured total THC. These are preliminary comparisons: the individual compounds were measured only once per extract, without repeated determinations to establish uncertainty.

Yield values are means from three independent extraction procedures. Compound-specific findings are descriptive single measurements. The paper’s total CBD and total THC values include adjusted contributions from their acidic forms.

The cannabinoid concentrations describe crude extracts after water removal. Dilution and formulation would change the amounts in a finished product. The experiment also did not measure absorption in the body, so recovering a compound with water does not establish that a product will deliver it more efficiently.

Why Total Yield Can Mislead

A hypothetical extract twice as heavy could contain the same amount of a target compound if its concentration were halved. Yield and concentration therefore need to be considered together.

The Soxhlet result illustrates the issue: it delivered the second-highest mass yield, yet none of the five targeted cannabinoids was detected. That does not prove heating destroyed them. The researchers also identify incomplete recovery, conversion, adsorption, and analytical losses as possible explanations.

A useful next experiment would track selected compounds through the starting plant, recovered extract, and remaining material. That accounting could distinguish successful recovery from concentration changes or losses.

Matching Cannabis Extraction to the Intended Product

The broader opportunity is to begin with a defined target. Researchers seeking a particular flavonoid may need different conditions from those developing a standardized cannabinoid extract. A larger yield can be useful, but additional extracted material may also require separation later.

This connects with MyCannabis’s coverage of CBD purification after extraction. A practical process must consider the steps that turn an initial extract into the intended ingredient. Improving extraction in isolation may leave the hardest purification problem unresolved.

A separate 2026 study of solvent choice and cannabis plant parts found different optimal solvent mixtures for leaves and flowers across its selected laboratory measures. It used water mixed with ethanol and acetone, so its results are not directly interchangeable with this water-only comparison.

The implication is useful: separating plant parts could be another variable worth investigating. Because the current study combined leaves, stems, and flowers from one batch, it cannot show whether the same protocol would suit each component or another variety.

Standardization would mean reproducing a chosen composition within defined limits. That could also make later research easier to interpret: when the preparation stays consistent, investigators have a clearer basis for examining its effects. These protocols provide starting points for that work.

What Antioxidant and Safety Tests Reveal

The extracts underwent two chemical antioxidant screening tests, called DPPH and ABTS. Both measure responses to laboratory radicals, but their chemistry differs. Ultrasound required the lowest concentration to reach 50% inhibition in DPPH; infusion led on the corresponding ABTS measure.

These different rankings reinforce the importance of defining the question. They do not establish antioxidant benefits in people. Repeating reactions from the same extract also cannot demonstrate consistency across different harvests.

The acute mouse experiment provided another layer of information. No deaths or overt clinical signs occurred after a single oral dose of 2,000 milligrams of extract per kilogram of body weight, followed by 14 days of observation.

However, blood measurements showed extract-dependent changes, including increased liver-associated enzymes with infusion and increased creatinine with Soxhlet and microwave extracts. Blood analyses involved only three mice per group. These observations warrant follow-up without establishing organ injury, human safety, or harm from ordinary consumer use.

How to Evaluate Greener Cannabis Extraction

Water-based processing creates opportunities to reconsider solvent use, but environmental performance depends on the complete workflow. Heating, repeated extraction, and removing water from the final extract all require attention. This study did not measure overall energy consumption, commercial costs, or lifecycle impacts.

Research on balancing extraction yield with resource use offers a useful model. A 2025 hemp-seed study evaluated separate strategies for maximizing protein and phenolic recovery and reducing water and energy requirements. Its seed-based system differs substantially, but the experimental question is transferable.

For future cannabis work, success could mean reaching a specified composition with fewer resources while maintaining repeatability. That is a more informative goal than maximizing recovered mass alone.

Turning Promising Extracts Into Consistent Products

The findings suggest three practical research priorities:

• Repeat compound measurements across independent batches and varieties.
• Compare protocols using matched conditions and measured resource consumption.
• Evaluate stability and repeated-dose safety for clearly characterized extracts.

Reliable measurement supports each step. MyCannabis’s coverage of cannabis testing and product consistency explores that connection. For consumers, the corresponding habit is checking the finished product’s batch information. The CBD oils and tinctures buying guide provides additional context for comparing liquid formulations.

The next decisive experiment would combine controlled extraction conditions with repeated chemical measurements across different harvests. That could show whether these promising differences can become dependable ways to produce a chosen cannabis composition.

References:

¹ Chahbouni, M., Majdouli, K., Dardouri, A., Elhichou Dbilen, H., Herradi, W., Samine, N., Jamal Eddine, O., Meftah Elkhair, M., Driouch, L., El-Mernissi, R., Gmouh, S., & Amine, A. (2026). Influence of green aqueous extraction techniques on selected phytochemical markers, biological screening responses, and acute oral toxicity of Cannabis sativa L. extracts. Industrial Crops and Products, 254, Article 124760. https://doi.org/10.1016/j.indcrop.2026.124760