Sciencemadness Discussion Board

TLC solvents

chemrox - 19-2-2009 at 23:25

I need to adjust my TLC tank solvent system. I have Me2Cl2, EtOAc, MeCN, trace of Et3N -- 2:1:0.2:tr. Its not separating the amides from the amines enough. The spots overlap with an rf difference of 0.5-1.0. I need to increase the separation. This system did work well separating the primary and secondary amines and the amines from other cmpds. Thanks for your suggestions.
CRX

Sandmeyer - 20-2-2009 at 17:41

What's the structure of the compounds you want to separate? In my experience in order for amines to migrate Et3N (or highly polar eluent) is needed. Try heptane:EtOAc 1:1 to begin with, silica is slightly acidic and without base it should be very easy to separate amines from amides since amine should remain on the baseline. "Me2Cl2" = DCM?

Arrhenius - 20-2-2009 at 22:58

Can you try to explain the plate (draw it if necessary). Name all spots in the mixture by Rf. If they're running together, your solvent system may be no good, or you're spotting too much.

What you need is a non-polar solvent (hexane), all of what you listed would be considered polar. Remove the Et3N and amines will have Rf of about zero from pure EtOAc or EtOAc:hexane provided your compound isn't huge otherwise. Amides should be easily separable from amines.

If you need to separate a mixture, I would suggest acid-base extraction to remove the amines from the amides (though amides can be very water soluble).

smuv - 20-2-2009 at 23:57

Get a small sample of the Amide/amine mixture, then take a pipette, stick it into a bottle of HCl pull just the semi-anhydrous HCl vapour sitting over the HCl solution and then bubble this through the small sample of the amide/amine mixture. Then spot/elute this small sample. After this treatment, the amine will definitely be at baseline with all but the most polar solvents.

Edit:

@Sandmeyer thanks for introducing me to Royksopp
@ Chemrox, upon second read, I realize that I don't fully understand what you are aksing...Forgive me if i am completely off base.

[Edited on 2-21-2009 by smuv]

chemrox - 21-2-2009 at 12:14

I'm monitoring the progress of a conversion: amine --> amide by TLC. I'm not getting big enough RF difference to separate the spots but I just saw something in these responses I should have snapped to immediately. My spots are too damned big. If I spotted less I'd have adequate separation to make sure the amine reactant was consumed. BTW- I visualize with I2. It seems to work will and is simplicity itself. I have a tank with I2 in it that the plates go in when dried. I also have an H2SO4 sprayer but haven't found it necessary yet. @ Sandemeyer, yes my non-standard abbreviation for DCM was Me2Cl2 (I have no idea why I used it ..) so while Smuv has a great idea and I will make a note on the technique involved, I simply need to pull another sample and dilute before spotting.

stoichiometric_steve - 21-2-2009 at 13:01

Quote:
Originally posted by chemrox

my non-standard abbreviation for DCM was Me2Cl2


not only non-standard but plain wrong

Globey - 21-2-2009 at 19:34

Quote:
Originally posted by stoichiometric_steve
Quote:
Originally posted by chemrox

my non-standard abbreviation for DCM was Me2Cl2


not only non-standard but plain wrong


Nope: chemrox's abbreviation makes perfect sense. In place of CH he has Me. Unless your just being picky because of the two free -'s. But electrons aren't generally shown in shorthand.

(For everyone's info, stoichiometric_steve just sent me an unsolicited PM, the contents of which: "You are an idiot"

Any you, SS, are obviously a nice guy, and not some antisocial miserable fuk.:) Chow;)

497 - 21-2-2009 at 19:49

Uhm... remind me again how there are there two carbons in DCM?

Globey - 21-2-2009 at 20:13

There aren't, there's 2 hydrogens one carbon and a two chlorines. It's just short hand, and there's no mistaking it because methylene chloride is the only real compound this could be.

HydroCarbon - 21-2-2009 at 20:19

Generally the more polar the compound being tested the more polar your eluent should be. In TLC the plate and the eluent compete for the substance being analyzed. Therefore if you want it to move further, use a solvent that it is more soluble in. And yes, over spotting will lead to spots overlapping and tailing. I usually use a double open capillary tube to spot with.

On the note of abbreviation, if Me stands for methylene then methylene chloride should be abbreviated MeCl2, not Me2Cl2. Although, DCM is much easier, and quite standard.


[Edited on 21-2-2009 by HydroCarbon]

chemrox - 21-2-2009 at 22:57

Yeah but its lousy shorthand and I won'y use it again. Anyway, I was reading some papers from the '70's that suggested that saturation of the chamber cn affect Rf values greatly and that, and this was the kicker, that saturated chambers give poorer separations than unsaturated ones (Sherma, ed., 1973). I have requestd the original papers because I have been using sheets of filter paper in my tank to insure saturation of the vapor phase. I may want to change everything to storing the solvent system when not in use. Sorry about the CHCl2 thing ... mea maxima culpa OK?

DJF90 - 22-2-2009 at 01:01

Shouldn't that be CH2Cl2...? ;)

HydroCarbon - 22-2-2009 at 10:34

Quote:
Originally posted by chemrox
Yeah but its lousy shorthand and I won'y use it again. Anyway, I was reading some papers from the '70's that suggested that saturation of the chamber cn affect Rf values greatly and that, and this was the kicker, that saturated chambers give poorer separations than unsaturated ones (Sherma, ed., 1973). I have requestd the original papers because I have been using sheets of filter paper in my tank to insure saturation of the vapor phase. I may want to change everything to storing the solvent system when not in use. Sorry about the CHCl2 thing ... mea maxima culpa OK?


Interesting, I've always done the filter paper thing too with the purpose of halting evaporation from the plate during the process. It would seem that evaporation from the plate due to the chamber not being as saturated would slow the rise of the solvent, possibly allowing for better separation.

[Edited on 22-2-2009 by HydroCarbon]

Arrhenius - 22-2-2009 at 12:33

Iodine is good, but kindof slow. I would highly recommend using a basic solution of KMnO4. Just dip the plate, and heat it lightly with a heat gun. This works well with any oxidizable analyte. Yes, TLC spotting solutions should be VERY dilute. Get some hexane, I promise it will work wonders for you! Hexane:Ethyl Acetate is so widely used as a solvent system because it generally gives tight spots. DCM is not so great as a TLC solvent. If you can't get spots to move in 100% EtoAC, try adding a very small amount of MeOH (too much and you'll have problems with the silica dissolving). Good luck.

Ebao-lu - 22-2-2009 at 13:44

Quote:
My spots are too damned big. If I spotted less I'd have adequate separation to make sure the amine reactant was consumed.
You can try to dilute your RM before spotting, maybe this could help? And what is the Rf of the higher spot, and is it amide or amine? You can use a small amount of NH3 as well, but i need to enquire to be sure.. Still, it is strange that the method was previously used for amines separation, and now it does not work.
ps: stop trolling chemrox for DCM! he already understood

[Edited on 22-2-2009 by Ebao-lu]

chemrox - 22-2-2009 at 21:23

Thanks for the solvent suggestions. I think I will move to hexane:EtOAc. I changed to EtOAc:MeOH (tr Et3N). The main improvements came from dilution and prventing the vapor from saturating the tank. The latter resulted in higher Rfs and better separation. Complete separation was achieved and a clear distinction of the spots was made. I'm happy to say that the reax went to completion with no reactant amine left. I've changed my TLC procedure to storing the solvent in a bottle and placing in the tank just before the plate. I've been using glass plates. I cut them into strips with a glass cutter for the one dimension runs. I also have a hook for hanging filter paper strips but large test tubes work OK for this too. Basic KMnO4? I'll try it. NH4OH/KMnO4? NaOH/KMnO4? Either one? I can char then with con H2SO4 too. I2 takes about 10 min.

Arrhenius - 22-2-2009 at 22:37

Maybe I've misunderstood, but you definitely want the vapor to saturate the tank. This is essential for repeatable TLC results & tight spots. If you've had to add Et3N to get the amine spot to move up from the origin, adding hexane won't help you (decrease polarity = decrease Rf, you know this). I use 9:1 EtOAc:MeOH (no triethylamine) with the amines I'm currently working with. I have also used ammonia saturated MeOH. It just takes some time to find out what works.

I'll try to grab the recipe from work... it's something like 1M NaOH plus KMnO4 in water. The spots will be yellow/brown and your plate pink. Works well with glass plates, and solution lasts a long time before going bad. Takes about 30seconds to visualize the plate. I've dipped & sprayed H2SO4, but I would normally avoid it. It's a mess, and the spots are faint (though different colors, which is neat), and you generally have to load the plate quite heavily. Of course UV is the easiest visualization if you've got UV dense analytes.

[Edited on 22-2-2009 by Arrhenius]

Ebao-lu - 22-2-2009 at 23:13

Arrhenius: thatks for nice method of spots viewing!

chemrox - 23-2-2009 at 18:39

@Arhenius: Yes- to the repeatability consideration. I pondered that and concluded there's no other way. Unless- everyone follows the same process for setting up an unsaturated tank or at least publishes the proces along with the data. Even then...The damned thing is this; you do get higher Rf's and better separation with unsaturated tanks. Oh well, I don't see it as a feasible option. Do you use a wick inside your tank?

Arrhenius - 23-2-2009 at 23:32

I mean... realistically... Rf values will vary between your TLC plate and mine simply from the sorbent batch or whatever, but for your own purposes having repeatable plates is good. If the Rf is too high, the proper thing to do is to decrease the polarity of your solvent by lowering the amount of methanol, or if it's too high in 100% ethyl acetate, by adding hexanes. This is of course just simple TLC for checking reactions. There are lots of fancy mixes with toluene, ether, acids, chloroform, etc. but these are really only appropriate for high performance separations for quantitative work, or identification in complex mixtures. I would not generalize your statement that a non-saturated tank lowers Rf and improves saturation, as I'm sure this is not always the case.

My plates are generally ~2cmx5cm for checking reaction progress or workup purity. I do generally fold a filter paper up and stick it on the side of the chamber, and I keep a lid on the chamber and typically run up to a dozen plates in a sitting. So yes, a wick is a good habit, albeit not essential for this small of a plate.

If you were running a 20cmx20cm plate, or 2D, you would find that the solvent front is uneven (sagged in the middle) if you don't saturate the tank.

[Edited on 23-2-2009 by Arrhenius]

chemrox - 24-2-2009 at 22:12

Quote:
Originally posted by Arrhenius
I would not generalize your statement that a non-saturated tank lowers Rf and improves saturation, as I'm sure this is not always the case.
[Edited on 23-2-2009 by Arrhenius]


Maybe- I experimented with it after reading an excerpt from a paper by Sherma in the Journal of Chromatography. I don't mean to wax authoritarian. I haven't gotten the original paper and found out how many experiments the statement was based on or what compds were involved.

As to the solvent front. The first thing I noticed wth the unsaturated tanks was an uneven solvent front. I'm using 1-2 cm X 12 cm plates. I'm cutting them from 12 X 12 plates. It's awfully hard glass and I have a lot of bad cuts. Sometimes I have the window shop cut up a couple of plates for me.

Arrhenius - 24-2-2009 at 22:24

In terms of synthetic work, a 12cm long plate is pretty huge. I don't know if you're having to buy them, but if so you can probably make your money go a little further by using a small plate. Seriously, it probably sounds silly that I use a 2cmx5cm plate, but it works well 99% of the time. The only time I've used larger plates is for a better assesment of purity, or for 2D analysis of raw extracts & radiotracers. Try that KMnO4 stain out, it's quite a bit more sensitive than I2.

Another trick to get basic compounds to move on your plates is to soak the plate in EtOAc with a little bit of triethylamine, rather than putting the TEA in your mobile phase. Just a thought. Sounds like you've got things figured out though.

TLC ELUENTS BY COMPOUND TYPE

Daedalus - 19-7-2013 at 17:20

Is anyone aware of a reference work on thin layer chromatography that contains a chart of appropriate eluent mixtures arranged by the class or sub-class of compound to be eluted (i.e., what solvent mixture is best for amines, ketones, aldehydes, etc.)? I know that, with silica at least, the more polar the compound, the more polar the eluent required to move it up the plate. Often times, however, it seems like much guess-work is involved in selecting the ideal mixture to effect good spot separation between compounds of similar polarities. In my searches so far I have found only general guidelines but nothing remotely comprehensive. Is anyone familiar with such a work, or must we go by "feel" in each specific situation?

byko3y - 20-7-2016 at 19:30

I have similar question. What confuses me the most is that mixtures of solvents are usually used instead of a single solvent, e.g. acetonitrile-ethanol, chloroform-methanol-water. I wanto to empasize the last example, because it could be substituted with a pure methanol or ethanol, but I'm sure there's the reason why a mixture works better than a single solvent. I've already spent some time searching the internet trying to find some explanation, yet most of the authors have no idea about solvent selection, except general rules "more polar solvent for more polar compound" and "guess and try". I'm pretty sure in a few days I will be able to find some answers, but maybe some of you already have one.

[Edited on 21-7-2016 by byko3y]

Metacelsus - 21-7-2016 at 05:13

You can easily adjust the ratio of solvents in a mixture if you want to tweak the retention factors. You can't do this when using a pure solvent.

In my experience, there aren't general rules beyond those you've mentioned. "Guess and try" is very much a part of selecting TLC solvents. Based on the results of previous guesses, successive guesses can be improved.

byko3y - 21-7-2016 at 12:31

You can easily adjust lipophilicity of mobile phase by going through the row of methanol-ethanol-propanol-isopropanol-butanol-pentanol...
And the retention is usually can be approximately predicted, for example, here is a graph for reverse-pase chromatography of homologs, which show us the fact that same type of solutes behave similary for the fixed type of stationary and mobile phase, while later can have different composition.
Rf for aromatics.png - 49kB

Graph of dependance of logP to composition of solvent is usually either straight line (1 and 3), or downward curve looking like a parabola branch (2), or falling and then rising curve (4) caused by interaction of silica acidic groups with basic amines. I'd say the reason for the drop in retention time for mixed solvent is caused by three components interaction (mobile, stationary phases, solute), when the solvent is capable of occupying the stationary phase surface just as well, as dissolving the solute. In fact, most of the curves for non-basic compounds are of type 2.
Retention curve type.png - 71kB

Recently I managed to guess the solvent for chromatography in the very first try, the one that caused an almost complete separation of closely related amine hydrochlorides on TLC (Rf=0.05 first, Rf=0.3 second, Rf=1.0 third) just by knowing their solubility in the used solvent (chloroform: 96%alcohol = 1:1), while usual solvents from literature would give me some really shitty separation, like 0.37 vs 0.42 for chloroform-acetic acid-water.
The guess-and-try method is even more unrelaible because of the need for mixture of solvents. Chloroform is needed to separate the sample constituents, but chloroform alone can't efficiently elute the sample, thus leaving everything at start. Pure alcohol would also work for my case, but give slightly worse separation.
Usage of a sinlge solvent might be advantegous e.g. when you separate basic amine from nonbasic amine: basic amine has higher affinity for stationary phase, thus higher retention time, but if you try to use a mixture of organic-protic solvent, the retention time of basic amine will be lowered so much it become the same as retention time of nonbasic amine, because the water negated the sample-silica interaction, making both amines equal.
isocraticrelationships.gif - 10kB
The example on the picture shows the convenience of separation of the three compounds using a pure acetonitrile, which causes the sample compounds to completely interact with stationary phase active sites. Amitriptyline has extremely low water solubility, thus high retantion time for pure water.
It could be analyticaly deduced even for unknown compounds, the only thing I need to know is that one compounds is completely neutral, the other is slightly basic, and the third is notably basic.
If I know the compounds have different octanol-water partition coefficients, I could blindly use alcohol (or water for low logP compounds), which would give a good separation most of the times.
I believe that just by knowing interactions of eluent-stationary phase, eluent-elute, and elute-stationary phases I can approximately pretict the separation with a good rate of success. But I cannot find much guides in my objective.