Just for clarity, you are extracting THCA (i.e., not decarbed) and getting those numbers. A lot of people might think you are extracting THC. OR do those numbers refer to the extraction of CBDA? When using protic and polar non protic solvents, I do not think a lot of extractors or readers on this forum appreciate the vast difference of solubilities between THC and CBDA…your numbers seem to reflect perhaps ethanol extractions of CBDA.
and I would assume you are using 95/5 etoh/h20 plus the 8% H2O residual in dried Hemp biomass. You can certainly add a lot more water and get similar numbers.
We’ve done testing with THCA, CBDA, and partially decarbed CBD&CBDA biomass for alcohol extractions, and had similar results.
Our ethanol testing was primarily done with standard 190 proof, though we did a bunch of testing down as low as I believe 50/50 and observed some very interesting results at much higher water content than most extractors use.
Here is an example of two aliquots of water hemp
extract removed at different steps in a multi step water process (three buckets and a Buchner)…in this case “tuned” to separate CBD and CBDA both in high purity. TLC/silica of both aliquots. Left columns CDB; right the CBDA aliquots. Samples are in duplicates at 3:1 L/R. Water tuned
Lab test results have come in for Red#2 and Blu#2. I’m adding the original RECE and RICE lab tests for comparison. The originals were a quick wash that yielded about 2g of oil. The Red#2 and Blu#2 were 10 minute soaks resulting in over 5g each as seen in the table below.
For the record, all these tests were run using the same hemp purchased from BuyOregonHemp.com. The COA is presented above. All these test runs were done as a single wash by filling the jar to cover the plant material, then agitated. No second wash or additional clean alcohol added to enhance the yields.
I find it interesting the TC and CBD totals only dip a little with a 10 minute soak vs a quick wash under a minute. The surprising thing to me is the volume of oil went from 2g to over 5g and barely lowered the potency.
Red#2 was decarbed in the oven for 40 min at 240f before extraction. I wanted to see how that would change the yield after distilling in water. You can see in the table both resulted in similar yields, but the lab tests tell a different story. The CBD is significantly higher saying I suck at post decarbing! Besides that the CBN is about the same between the two. All these yield tests were reduced down to oil and ended when there was no more C02 bubbling so I thought these numbers would be higher.
Observing the final yield #s, a 10 minute soak shows it has yielded more oil with marginally more plant compounds. To your your point @cyclopath, there’s still a discrepancy between the COA #s and the final Red and Blu yields, though much better than the initial yield of 2g per 2oz of hemp. This probably means I have to do more work on optimizing the wash as @Lincoln20XX pointed out above.
If I wanted to try to optimize for complete extraction at the home scale, I’d probably start with something like this:
Size-reduce input biomass via your preferred method
Put biomass into a mason jar
Add 1/4 to 1/3 of total solvent into jar
Agitate vigorously (shake it like it owes you money AND said mean things about your mother) for about a minute
Drain off solvent
Add additional 1/4 to 1/3 solvent. Repeat agitate/drain/fill until solvent is all used up.
Doing a total of three or four rounds of this should get you about as much of your desireables as you’re going to be able to get in a home setting with minimal effort and equipment.
I’d be willing to bet using a 90% alcohol concentration would yield better than 99+%
Also, if you’re going to the effort of paying for lab tests in service of making this better, getting yourself a higher precision scale would be a good thing.
Hemp is so cheap…is it worth good solvent and recovery time to go after residuals? I’m asking…does anyone have the numbers as a break even point. Once you add alcohol to hemp biomass…you want in and out before the greenies…so to speak. It is more getting it to drain or press-drain with out the undesirables. You have hundreds of thousands of trichomes with diffusion cross sections vs the leaf structures…so there is time and cannabinoid yield and then there is time and undesirable yields. It seems to me to take longer pressing the solvent volume out than it does to do the extraction (if not using centrifugal or press plate forces)…i.e., the home cook. A quick French press (like for coffee) might be an easy way to address this issue in a kitchen setting. Also if you are doing your pre-extraction as you have mentioned earlier…you could press your excess water out…prior to the EtOH extraction step. Otherwise you have a rather un-metered water/Ethanol ratio and unknowns galore.
Of course if you WANT the green and other constituents of the long soak…Then leave it overnight. Certainly there is a history and tradition of “total extract” when it comes to RSO. Above I am only addressing cannabinoid extraction and its subtleties with respect to the OP’s concept of RxCE.
I’ve been doing a single 30 second to 1 minute washes using 99% ISO for my top shelf oil. It’s cherry picking to avoid plant matter. I found that using 99% leaves the plant crunchy like dry corn flakes. 91% ISO and 95% EtOH leaves the plant wet like soggy corn flakes. Using 99% really does avoid penetrating the plant. @Lincoln20XX suggestion above makes sense going forward. I found the acidic state and long soaks are harsher on the throat, not so much chlorophyll so dark oil can be very smooth when fully decarbed.
Removing the water is part of the process. When rehydrating with EtOH for the final reduction, water is avoided. Its easy to remove the waste water, but can be messy resulting in transfer losses if not handled right. Dumping out the waste water is a brand new experience for making RSO.
After I ran those tests doing 10 minute soaks, I saved the biomass and let that soak over 24 hours. Ran 7 more tests on that long wash to work out how to break emulsions. The resulting oil was sooooo nasty I threw it out. I used 9ml in a batch of gummies. It was very weak and tasted horrible so all those got tossed in the trash. Long soaks should be avoided. Turns out breaking emulsions is easy using up to 0.5% salt to water ratio. The salt disrupts the hydrogen bonding and immediately brings the yield up to the control case yields. It’s easy enough to remove when dumping the waste water by doing a quick rinse of the captured oils with cold tap water.
I’m all for better looking products, but I do have some concerns with efficacy. Everyone seems concerned with cannabinoids (as they should). What about the various molecules that may act as cofactors to cannabinoids to increase bioavailability? Have you noticed any difference in how well the oil performs in comparison with the older “La Brea”?
On molecules, here’s what Gray Wolf said in another forum.
On performace, one individual on another forum, is having success with it’s higher potency.
Hey @Lincoln20XX, I just ran a yield test with 15% CBD and your approach yielded 93% of the volume of the control case. I’ll published those results soon. Thank you for that. I’m going to add this to the process and RxCE Shop Manual with credits.
Thanks for the info @MimiEmu. I currently work in Florida and any use of ethanol must be in a closed loop system. Because of this law, i’m trying to stay away from ethanol outside of the X40. In this formulation it states the need to reintroduce ethanol into the solution. Would it be possible to just use more isopropyl or is the ethanol a necessity for the final reduction?
The use of Ethanol in the second half provides two features. It has a dielectric constant at 24 vs ISO 17 so it’s more polar which drives the Rapid Winterization. Maybe adding bit of water to ISO to take it to 87% would work. (That 87% was mentioned by @beaker or @cyclopath in a different thread as being similar to Ethanol’s polarity). That needs to be tested to see if it works as well as Ethanol for Rapid Winterization. Fall back is going back to the freezer, but then, what is the benefit of this process?
Secondly, it being GRAS, makes it an excellent cleanup solvent to avoid any Residual Solvent violations. Using it in the second stage reduction and decarbing makes for a very robust residual solvent remediation.
So, I believe you could use ISO for the second stage, but dewaxing might not work as well. You must decarb at the end as this is a literal firewall to prevent residual solvents in the oil.
I’m sure our reactor and vac pumps could remove any residuals in the material. Thanks again for all your assistance. I’m not sure when i will get the chance to test this in our lab, but when i do i will be sure to update you with the results.
Here’s the results of a quick test for maximizing yield on solvent extraction. I ran four 2oz extractions. The first, ‘Grn’ was the control case that was straight away boiled down to oil (RSO) to establish a maximum yield. The second, ‘Blu’, was your approach. The third and fourth I tried something new. The 99% ISO was heated to 150f then used for a 30 second and 1 minute soak. So here are the results:
Edited for clarity…
Label Alc Time Washes Salt Yield Waxes
Grn 99% 10m 1 N 6.5g N/A
Blu 91% 5min 3 Y 6.0g 1.3g
Red 99% 30s 1@150f Y 4.6g 0.8g
YlW 99% 1min 1@150f Y 4.8g* 0.9g
*noting a small transfer loss, maybe 0.1g spilled…
Couple points. I used 91% at room temp with ‘Blu’ knowing this would penetrate the plant just a bit. Reason being, I expect 91% to be used on a regular basis by home users. 91% is what you recover out of a still/water distiller. Same rules apply to 200 proof vs 190 proof ethanol.
99% will have to be an ongoing new purchase for every extraction. You don’t want to dry out ISO thru refluxing as you risk generating peroxides. Over time, the alcohol evaporates and can leave peroxide crystals. These crystals are unstable, IE explosive. There are instances where 99% refluxed ISO stored long term has formed crystals and exploded when disturbed. Water disrupts the whole peroxide issue, so it is safe to recover 91% ISO. It’s removing the 9% water to make 99% ISO that it becomes vulnerable to peroxides.
Visually comparing the ‘Blu’ wash to the others showed it to be marginally greener, so using 91% quickly works well. There was very little ‘greenies’ pulled using 91% this way. The ‘Red’ and ‘Ylw’ tests cases were interesting, but yielded less than I had hoped for. The ISO was initially heated to 150f then poured over the plant for a single wash. After winterization, I collected the waxes by rehydrating them in ISO. This rehydrated wax solution produced an amazing wax bloom before being boiled it down for measurement. Got pics of everything if you want to see any of it.
So, there it is in numbers. The ‘Blu’ test weight + the measured waxes shows a higher total weight than the control case. The control case was a single wash boiled down so it contains it’s share of waxes.
You can ignore the the ‘Salt’ column. It indicates the use of salt to break the emulsion during distillation. It prevents oil losses with water distillation.
I think that you should edit your post so the text under the table isn’t so hard to read.
I tend not to recommend working with solvents that one has applied heat to in a home context, or even in an industrial one, for that matter.
With that said, if you’re willing to heat up your solvent, doing the same procedure that I’d recommended initially, but with hot solvent, would possibly work even better.
Something between 35°C and 50°C is probably sufficient.
It’s also entirely possible that you’ll not get more cannabinoids, but just more other shite.
It’s also entirely possible that you’ll not get more cannabinoids, but just more other shite.
Thats why I used 99% ISO, not 91% as it penetrates the plant to easily. Using heat was just an experiment to see what happens. I’ll do a 5th test at 37c/100f with 3 quick warm washes.
Now that I can read all that easily, yes showing pictures to the class would be good.
Depending on how far you want to go down the extraction optimization rabbit hole, you might find that having water fractions substantially greater than 9% might prove useful.
When I get to my laptop I’ll DM you some research we had commissioned years ago on the topic that you might find interesting.
Doesn’t look that much different, but the Blu jar is marginally less magenta, little more green due to using 91% ISO where it grabbed some color. All these were then filtered thru an KN95 mask. N95 and KN95 masks are 0.3 micron fast flow filters.
Besides that the CBN is about the same between the two. All these yield tests were reduced down to oil and ended when there was no more C02 bubbling so I thought these numbers would be higher.