These investigations include several limitations and opportunities for refinement. Resources were limited during the course of this project, which led to limits on the number of replicates for the extraction treatments to two replicates, when it would have been preferable to perform at least three replicates for adequate precision and accuracy. The amount of biomass used in percolation treatments also had to be limited to approximately 50 g due to the size of the percolation cone that was available, to mimic home extractions. Future follow up studies should match marc mass in the percolation and maceration experiments.

Further limitations of these investigations include the potential heterogeneity in CBD concentration of the biomass, potentially resulting in variable CBD concentrations in the starting biomass. Consistent distribution of biomass CBD content and particle size was managed by mixing the hemp biomass repeatedly and sifting the biomass through a #5 sieve for an effective medium approximation of CBD. Our analysis shows an RSD of [6.19%] for CBD concentration after triplicate samples of the hemp biomass and triplicate analysis of each sample.

In these experiments we compared the solvent recovery, yield of CBD, additional input of pressing marc in two disparate techniques of extractions; maceration and percolation. We also mimicked home extractions that a person not versed in laboratory extractions of medicinal plants could perform in a typical home kitchen. All items used; screen, oven, glassware and press, are either easily afforded in Western countries or are readily available in a typical household. In addition, the particle size of the flower, which had been baked at 250°F for decarboxylation, was consistent throughout both macerations and percolations.

These data show that in the maceration method, pressing the marc significantly increases the content of CBD in the menstruum. Our results show that an additional 8.19% (∼458 mg) of CBD was added to the menstruum from the pressing procedure. The step of pressing the marc to express the retained menstruum is often problematic for people making tinctures at home. We showed with an inexpensive and readily avialable wine press, that was not designed for pressing the biomass of tinctures, that there is a significant value in pressing the marc for an increase yield in active compound(s).

As opposed to menstruum recovery in the 2-week maceration, the total menstruum recovery in the percolation was improved by 4.931% as compared to the fully pressed macerated tincture and 13.398% of the unpressed macerated tincture. However, what most production facilities of medicinal plant tinctures employing maceration technique use is a press supported by a hydraulics. The above figures on menstruum recovery come from an inexpensive wine press. Thus, it is conceivable that a hydraulic press, generally not available in the herbalist’s kitchen would yield a higher menstruum recovery.

Interestingly, we attempted to press the marc from the percolation cone and did not recover any viable amount of menstruum (Table 4). Menstruum recovery for percolation was more efficient over maceration, even when pressing the macerated marc. These investigations show that percolation as an extraction technique, has a significantly increased extraction efficiency of CBD and menstruum from Cannabis tinctures as compared to a maceration.

Although the efficiency of percolation vs. maceration has been long debated, percolation is another extraction technique, traced back to ancient Greece, that has long been used for botanical medicines (Raubenheimer, 1910). Percolations are preferred by some medicinal plant extraction facilities because they are more time and space efficient. Instead of macerating solutions sitting for extended time periods in large drums, as traditionally done in the production of medicinal plant liquid extracts, percolations are finished in hours, rather than the traditional maceration time of 2–4 weeks. After two independent extractions the percolations extraction efficiency of CBD was 80.10% on average while the 2 week maceration extraction efficiency was 63.52% on average. This investigation provides data to supplement a long-held debate on the efficiency of percolations over macerations.

Percolation extraction requires a percolator, a cone shaped vessel narrowing at one end and open at both ends (Majeodunmi, 2015). Pharmaceutical grade percolation cones are not widely available for the home enthusiast. However, certain wine shaped glass bottles or sparkling water bottles can be adapted. In the attempts to keep this technique replicable to a consumer in a home kitchen we used a Perrier bottle, which is commonly used by herbalists, with the bottom of the bottle carefully scored and then removed. The plant material is scantly pre-moistened with the liquor, sits for 24 h and is then packed in a percolation chamber with the proper packing pressure. The liquor is then run through the packed herb in the percolator which has a stop cock at the end to control rate of flow, which depending on the herb, is from one to six drops a second (Raaman, 2006). A screw on cap on a Perrier bottle is often substituted in home extractions instead of a stop cock.

Another advantage of percolation may be preservation of phytonutrients. Select phytochemicals degrade more rapidly in a liquid medium than in a solid medium (Fairbairn et al., 1976; Rogers et al., 1998; Livesey et al., 1999; Perry et al., 2000; Bilia et al., 2002; Yang, 2008; Peschel, 2016). Phytocompounds in a macerating hydroethanolic solution, or sitting in tincture form on a shelf waiting for a consumer purchase, degrade on varying timescales. For example, caffeic acid derivatives, important herb constituents due to immunological influences (Kilani-Jaziri et al., 2017), have shown a remarkable ability to degrade in just a few months in hydroethanolic solutions (Livesey et al., 1999). Degradation in hydroethanolic solutions has also been shown for cannabinoids (Fairbairn et al., 1976; Peschel, 2016) and the lipophilic CB₂ ligands known as the alkylamides (Livesey et al., 1999; Perry et al., 2000).

These results are particularly pertinent to other medicinal plant extractions in that the identical techniques used for Cannabis in these experiments are used for a wide variety of medicinal plants. Thus, these results may be pertinent to other lipophilic compounds that are extracted from medicinal plants. For example, in the case of Echinacea spp., one of the groups of active constituents are the alkylamides. Specific alkylamides bind with similar affinities to CB₂ as the endocannabinoids (50–60 nM) (Matovic et al., 2007; Raduner et al., 2006) and have also shown PPARγ activity (Spelman, 2009). Previous research demonstrates that there is no difference in alkylamide concentrations in Echinacea purpurea roots extracted for 24 h versus 2 weeks in 70% ethanol (Spelman et al., 2009), however these results did not compare a percolated extraction.

The log P, a measure of lipophilicity, is less than ideal for optimal pharmacokinetics for both CBD and the alkylamides previously mentioned. Thus, these examples may be relevant to other lipophilic compounds in medicinal plants. Moreover, the topological polar surface area (PSA) for both compounds is also less than ideal for optimal pharmacokinetics suggesting that at the least, these two lipophiles may act similarly in crude extractions.

Finally, it should be noted that these extractions were carried out with CBD, not CBDA. Given that the PSA differs considerably between CBD (PSA = 40.5 Å²) and CBDA (PSA = 77.8 Å²) further extraction efficiency of CBDA in ethanol is expected due to the higher PSA. Under aqueous extraction conditions, superior extraction of tetrahydrocannabinolic acid (THCA) over tetrahydrocannabinol (THC) has previously been shown (Hazekamp, 2007). Thus, extraction from dried hemp that is not decarboxylated will likely be advantageous for extraction in ethanol.

To summarize, using common tools available in a home kitchen, an extraction performed by percolation has a statistically significant higher extraction yield on average of 16.59% CBD than a 2-week extraction (p = 4.419 × 10⁻⁹). In addition, the percolation method of extraction exhibited a superior performance of precision between replicates and menstruum recovery by 4.93% as compared to the combined liquor from the drain and pressed marc.