DIY Thin Layer Chromatography (TLC) of cannabinoids at home - tutorial

Discussion in 'Advanced Marijuana Cultivation' started by PhenoMenal, Nov 16, 2017.


    LostInEthereal Well-Known Member

    Unfortunately these days I don't dabble too much, but just for lack of trying. I used to source everything through the darknet but situation is pretty sketch and has been for some time. I just scoped out the website I purchased my kit from and it doesn't have any labeling or MSDS information (a bit odd actually) for whatever solvents and other chemicals used. I know the solvent was quite strong and I actually avoided using it because of my roommates cats lol, I didn't want to expose them to any harsh chemicals. I do know the kit was non specific towards MDMA, just that was likely to be the most common use but I wonder if it could differentiate between isomers of ketamine for example, would be interesting.

    #hugs'n'glowsticks brother, haha

    EDIT - weird formatting and added hashtag
    PhenoMenal likes this.

    Sureshot2 Well-Known Member

    This is fantastic work, greatly appreciated and definitely one of the better advanced section threads on this site. I have most of the materials already, just gotta pick up some of the fast blue, lye and TLC plates and then I'll give this a go.


    PhenoMenal Well-Known Member

    I guess all labs have their good days and bad days ... well, (sadly), amateurs have more than their fare share of bad days :( :)

    I've just done another TLC today, this time on a CBD oil, as well as C99 by FemaleSeeds Co (perhaps the grapefruit pheno or just another pheno thats popped out from whatever theyve crossed C99 with... def not the pineapple/tropical/classic C99 tho, but very much still in the neighborhood).

    It's not my cleanest TLC though! As I was only doing 2 samples I decided to cut the silica plate in half so I could do two samples per halfplate, each in a different eluent for the mobile phase.

    But, unfortunately there was some minor interaction between the lanes. From now on, I think I will only cut a plate in half when doing 1 sample (1 lane per halfplate) in 2 eluents. The aluminium silica plates are only about $1-2 each anyway, and TLC isn't something I do on a regular basis anyway, so I should stop being a tightass with those!

    A little ant also ran across the silica plate, probably leaving a pheromone trail, the little ****! I highly doubt it would affect the TLC in any noticeable way though, but I wouldn't want to bet on it, especially as this was one of my worse TLC runs, although that could just be a coincidence.

    Anyway here is the first halfplate (the two lanes on the RIGHT), which I developed with Chloroform as the mobile phase eluent. On the LEFT, the smaller one is just a copy from an image ive already posted here, which was the only other time I TLC'd a CBD oil sample (pretty much all my other testing has been with buds). I'm not sure why the Grapefruit lane isn't straight, perhaps there was a slight bend in the plate being only a half-plate, but I haven't seen that problem before with other halfplates, and even if it was bent you'd think it would still go upright, so perhaps it was just slightly oriented off-center due to being in a cylindrical jar with a raised bottom which might've offset it ever so slightly)


    I've also TLC'd an official Grimm Brothers C99, so for comparison you can see that in the last image in [this post] a bit earlier in this thread.

    ps. you can see two dark dots to the right of the Grapefruit lane's CBG dot ...they're clumps of Fast Blue BB (i wasn't patient enough with the coffee filter paper and pushed a bit too hard, so some FBBB ended up in the mix, whoops... be very gentle when squeezing the FBBB through the coffee filter paper)

    While I can't remember what the Unknown CBD oil i tested is called, and although I only used Hexane : Diethyl Ether as the eluent, you can see that its profile is very similar to the CBD oil I tested today (Hempworx 750). Note: This is not an endorsement, and I'm not saying you should or shouldn't get it. There's a lot of good/legit CBD oils out there, but sadly a lot of scamming going on, and that's ****ing with peoples lives, and in the case of cancer patients they simply don't have time, OR money, OR health & wellbeing to lose to scammers. Always do a lot of googling about a CBD oil before you buy it - do some basic background checks on the company, and search the web to see if other people have found it helpful.

    This particular CBD oil, Hempworx 750, purports to contain 750mg of CBD (in a 30ml/1oz bottle, so that would be 2.5% CBD, W/V%), and comes infused in hemp seed oil, and peppermint oil for taste. The word from my friend going through cancer who's just tried it said it tasted good, which might be good for people who don't like that planty taste some oils have. Or simply add a drop or two of your own peppermint oil to whatever CBD oil you have, no rocket science there!

    I've never seen that "W" shape/effect that can be seen in the CBD oil's CBD spot on the TLC plate ... I'm wondering if that's a component of the peppermint oil just ahead of it that might've caused that (clearly it has SOME other molecular components ahead of it near the top, as you can see how much lower on the plate the CBD dot is... CBD appears above THC, so you can tell where the Grapefruit's CBD dot would exist if it had one). It'd be great to have another try on a full plate etc, but the stuff is expensive and comes in tiny bottles. What's the dark-orange line on the top of the CBD spot? I'm not sure - perhaps its due to the minor interaction with the red THC dot in the lane beside it, because i've never seen such a 'line' before... always DOTS. However, perhaps the molecule dot above it is 'restrictive' in a way, something like a wall, which might lead to a buildup along the edges and therefore resulting in a darker color from the higher density, but this is pure speculation and guessing at possibilities.

    So, just judging from those two hemp-derived CBD oils (granted, not a big sample size), my takeaway is this:
    • CBD oils are a great way to LOAD UP on CBD, without having to worry about THC taking over. As CBD isn't psychoactive obviously we can handle a LOT of CBD, but we can only handle so much THC before we fall asleep, and that's a significant limiting factor. I now I'm not saying anything new there, but is an important aspect for patients. Some medical trials with CBD have used very high levels of CBD, and you simply can't obtain those levels with for example "1:1" CBD:THC strains.
    • However, a lot of trials also show the SYNERGY that CBD and THC have - THC often seems to positively increase CBD's effects (something the cannabis community has known for a long time, and is now finally getting scientifically affirmed). So, after your medicinal CBD oil drops, perhaps have "a quarter of a cookie" or something to introduce a small but manageable amount of THC. (Or the whole cookie, or four, whatever) :)
    Anyway, so that was the plate developed with just pure Chloroform as the mobile phase eluent ...
    but I also use Hexane : Diethyl Ether (4:1) ... I'm still not sure which eluent I prefer (they both seem to have pros and cons, but also complement each other nicely), and I've had nothing but success with it so far ...
    ... but NOT TODAY! Just one of those days unfortunately! ...


    WHOA! Massive fail ..... the extract didn't even get out of the starting blocks!!! (even though the eluent crossed the finishing line!)

    The only saving grace I guess, is that yes we can see the CBD Oil lane is clearly showing a fat orange CBD dot, while the C99 lane is clearly showing a fat red THC dot... it's not much information, but it's better than nothing and still definitely provides a slightly clearer indication than Beam's Test, as Fast Blue BB is far more specific in indication of cannabinoids. But, for whatever reason, it failed to SEPARATE the substance into its lesser molecules, so we've basically put a bread in the oven and it's failed to rise due to some problem probably with the yeast.

    In case you're wondering, yes the eluent did capillary-action all the way to the top (and you can see some minor staining at the top), so that's not the problem.

    • The extract failed to be moved up the plate via capillary action.
    So, what's responsible for driving it up? there are two fundamental parts...
    1. first we use Hexane to create the EXTRACT... we need to extract the cannabinoids from the target substance (Hexane is something of a molecular magnet)
    2. then we use the eluent (eg. Chloroform, or Hexane : Diethyl Ether) for the MOBILE PHASE, where capillary action drives the eluent up to the top of the plate, separating molecules into groups in the process
    I believe I've made a mistake in either of these two.

    POSSIBLE CAUSES: (the only ones i can think of - it's a fairly straightforward procedure afterall)
    • MY BET: perhaps my hexane : diethyl ether measurement was slightly off (i don't know how sensitive the ratio is). In one of my first posts I showed experiments with different eluents that help show how important it is to use "the right eluent for the job (and there aren't many for this)", but I'm confident I was within 1ml of 8ml of hexane, and pretty sure I was close to the 2ml for diethyl ether (should be 4:1 ratio). I've never had a problem with it before, but measuring this time was trickier as my syringes now no longer reach into the ever-depleting level of eluent in the bottle, so I was happy at the time to be less precise. However, there is obviously a reason why the two are used together, and there's obviously a reason for the 4:1 ratio ... so, perhaps there is no elbow room there. Perhaps we need to be very precise with this mix. But in my previous experiments with different eluents, there was never this "gridlock" problem! Even water had no problem driving forward, albeit without separation. So, if this is the problem, the solution is of course simple: be precise ... it's not difficult, especially now in retrospect haha
    • UNLIKELY 2ND BET: perhaps I didn't use enough hexane in the extraction phase (possibly resulting in extract which is too 'sticky'), but it didn't affect the Chloroform plate.
    • UNLIKELY: I actually used 4 drops per lane, usually I just use 3, but surely that 1 extra drop couldnt have this much effect, especially as it didn't affect the Chloroform plate.
    • UNLIKELY: surely that ant that walked across the plate didn't leave a pheromone trail that could have this much effect, especially as it didn't affect the Chloroform plate.
    I might be able to revisit this in the future to try to set up a couple of experiments which can test some of these possibilities to try to rule them out.

    Due to lack of publicly available knowledge about this, I basically have no other option but to keep experimenting. Hopefully some of you will too :)
    Last edited: Dec 13, 2017

    BabyLobsterito Active Member

    I greatly enjoyed reading this. Got me wanting to bust out the lab equipment...
    Glad I decided to check in after several months!

    raggyb Active Member

    I want to start with Beams test. See if I can handle it LOL:clap:

    PhenoMenal Well-Known Member

    the beauty of Beam's test is how simple it is, and just two easy-to-source ingredients which are quickly and easily prepared ... if you can't handle Beam's test you'll probably struggle putting butter on toast too :clap:
    Stink Bug and raggyb like this.

    raggyb Active Member

    Ia canna butter mya toasta

    raggyb Active Member

    Just Jim Beam,, I mean KOH and Ethanol

    raggyb Active Member

    But seriously man, this is really a cool post. Thanks for doing all that work and sharing!
    Stink Bug

    Stink Bug Well-Known Member

    First class post. Well done.

    PhenoMenal Well-Known Member

    mate I'm not sure if Jim Beam's Test would work as that is < 50% ethanol, plus it's already coloured not clear so maybe it might work but would be tricky to read the result, but what I am sure of is that it would be a waste of bourbon. I know you were only joking, but I still hope you felt guilty as soon as you pressed Enter!

    You can buy 95% ethanol like "Everclear" but its very expensive ... also you can distill your own but very time-consuming and i think 95.6% is the max you can get it to.

    I just use methylated spirits, which is dirt cheap (~$3 / Litre) and can be sourced from local grocery store - dont even need to go to a hardware store usually, you can also order it online, and it's still ~95% ethanol, and visually clear.... basically just Everclear with another substance to prevent people drinking it... and BANG the price goes down! Gets the job done nicely, cheaply, and very obvious color differences when CBD is present, so I say just use methylated spirits - don't bother with pure ethanol... or whisky. :)
    Last edited: Jan 18, 2018

    p0opstlnksal0t Well-Known Member

    Have any of you found a chart/overlay for the plates like what alphacat's kit uses to get a "not too accurate" quantitative reading of the terpenes?




    raggyb likes this.

    raggyb Active Member

    I don't know what that is but it looks pretty cool, stinksalot. I'm still messing around with the Beams test idea. Found a recipe for making your own koh. I'm going to try.
    p0opstlnksal0t likes this.

    p0opstlnksal0t Well-Known Member

    keep us updated bro. im just now looking at starting up TLC for myself and also charging a small fee for local Care Givers to get cheap tests done. The ability to actually quantify (Alpha-Cat says their tests are quantifiable to 1.0% accuracy) each terpene makeup would be an astronimical benefit as opposed to just seeing each terpenes ratio to the others. I know the only way this could be possible is if each input material is minutely measured to exacting precision. the solvent ratios, dye, mixed media, flower, etc... all have to be perfectly proportioned and measured so that the end resulting dot sizes are quantifiable in the end. I would either need to utilize Alpha-Cats charts for measurement or use some type of software to scan the plates and measure the resulting profiles and create a database of ranges. I would then need to pay to send in a few flowers that ive tested at home to a lab that is accurately quantifying these terpenes so that i could build this online database to use the data points to extrapolate my at home TLC data points.
    raggyb likes this.

    raggyb Active Member

    cool. I'm going to start trying to make KoH this weekend.

    PhenoMenal Well-Known Member

    raggyb, "make KOH"??? !?!?!?? no! just buy it from ebay or other, it's very cheap, and you don't need much at all to do a Beam's test, so a little goes a long way :) KOH is caustic too so it's quite different and more dangerous to say brewing your own ethanol! For TLC, making your own KOH is just unneeded extra time and expense.

    PhenoMenal Well-Known Member

    my advice is don't bother trying to get % from home TLC, there are too many variables that all need to be exactly 100% just to get semi-accurate readings. TLC is a qualitative test, not quantitative. You really need to send it to a lab if you want %'s you can rely on.

    It's still a beautiful tool for revealing both the EXISTENCE and RATIO and APPROXIMATE QUANTITY and overall FINGERPRINT of cannabinoids though!!!♥ just not accurate %.

    You mentioned selling it as a service, but I feel you would be doing a disservice to your clients by using TLC to provide %'s because you could be off by as much as 5-10%, which is a huge error margin considering strains top out at ~30% THC. If people are willing to pay for such a test they generally want accurate results, and TLC's margin of error simply cannot accommodate that.

    For example, in the AlphaCat image you posted for example, look at the 12% and 17% blobs:
    ... almost identical to the naked eye yet 5% difference. I would encourage you to provide a scan IMAGE of the plates to clients, rather than %'s, because it really is afterall a VISUAL test (hence the need for the special Blue B/BB dye!), a non-digital test, and one that doesn't provide concrete numbers. :)
    Last edited: Feb 3, 2018

    ANC Well-Known Member

    Here is my favorite method

    PhenoMenal Well-Known Member

    ANC, smoking is an easy detection method ... but only for THC! not for CBD (or CBG, or THCV, or CBC, or CBN etc)

    Let me flip this to you -- "Here's a bud, smoke it and tell me how much (IF ANY) CBD it has" ... :) You simply cannot answer this question by smoking. (I was drawn to TLC because a family member was diagnosed with terminal cancer and was having a lot of good results with CBD, so we needed a way to detect with confidence)

    Even if you give it to somebody in pain for example, they might smoke it and still give it a negative rating even if it has high CBD ... so clearly humans smoking is not a good indicator as to whether or not a strain has CBD etc.
    This is where TLC (and to some extent Beam's test) really show their unique value. :)

    ps. having said all that, mate I have to agree with you though smoking is also my favorite THC detection method haha :)
    Last edited: Feb 3, 2018

    raggyb Active Member

    Ok, stupid question, but how dangerous can it be if it's used for making soap. I found the instructions from soap makers. Soap doesn't exactly make my boots tremble.

Share This Page