Home >
Community >
Choice of solvent during thin layer chromatography
Upvote
32
Downvote
+ Biochemistry
+ Chemistry
Posted by
Marry Harry
Choice of solvent during thin layer chromatography
Chromatography is really an empirical science. The fluid mechanical side and the adsorption phenomenon is somewhat well understood but choosing mobile phases is essentially based on trials, sadly to date. If someone has given you a ratio, say 8:2, it means someone must have tried it and they found that this mobile phase works the best. The main criterion for choosing the mobile phase is that all the mixture components must be soluble in it. Thus if glycine did not move, this means that the mobile phase has no affinity for this solute (it is perhaps insoluble in it) and it is a poor choice. One then starts to tune the polarity of the solvent mixture. Think about a petroleum sample, which literally has thousands and thousands of different molecules. Now think yourself: will they all dissolve in 100% hexane? You do need an additional solvent.
There is no theoretical restriction in chromatography to use multiple solvents or a single solvent. For example in gas chromatography, you always use a pure gas. The reason for using a mixture of solvents in liquid chromatography originates from the concept of "general elution problem of chromatography" which basically says that there is no single adsorbent or a single mobile phase which can separate everything. Not only separation scientists use a single mixture composition of mobile phase (isocratic), they can also continuously change the mobile phase composition as a function of time. This is called gradient elution. You cannot do that with thin layer chromatography without using fancy stuff.
People are trying to implement machine learning, so that one can enter the molecular structures contained in the mixture, available stationary phases, and with a lot of statistics and tons of experimental data, the computer can tell you to try a certain mobile phase composition.
Chromatography is really an empirical science. The fluid mechanical side and the adsorption phenomenon is somewhat well understood but choosing mobile phases is essentially based on trials, sadly to date. If someone has given you a ratio, say 8:2, it means someone must have tried it and they found that this mobile phase works the best. The main criterion for choosing the mobile phase is that all the mixture components must be soluble in it. Thus if glycine did not move, this means that the mobile phase has no affinity for this solute (it is perhaps insoluble in it) and it is a poor choice. One then starts to tune the polarity of the solvent mixture. Think about a petroleum sample, which literally has thousands and thousands of different molecules. Now think yourself: will they all dissolve in 100% hexane? You do need an additional solvent.
There is no theoretical restriction in chromatography to use multiple solvents or a single solvent. For example in gas chromatography, you always use a pure gas. The reason for using a mixture of solvents in liquid chromatography originates from the concept of "general elution problem of chromatography" which basically says that there is no single adsorbent or a single mobile phase which can separate everything. Not only separation scientists use a single mixture composition of mobile phase (isocratic), they can also continuously change the mobile phase composition as a function of time. This is called gradient elution. You cannot do that with thin layer chromatography without using fancy stuff.
People are trying to implement machine learning, so that one can enter the molecular structures contained in the mixture, available stationary phases, and with a lot of statistics and tons of experimental data, the computer can tell you to try a certain mobile phase composition.
I believe you are missing the punchline of $elementary$ chromatography principles- good for the beginning student. The partition coefficient of a given solute cannot be zero! It can only be non-zero, if the compound is soluble in it. Polarity is really a poor term, IUPAC discourages it. I wish there were a better term for it. Organic chemists use it a lot, but polarity doesnt mean anything.More
This is not a nice answer but unfortunately it has been accepted. Solubility is absolutely required just because molecules cannot walk by themselves along the column. However has little to do with the mobility of compounds and thus, with their separation. This misleading idea is reiterated by " Thus if glycine did not move, this means that the mobile phase has no affinity for this solute (it is perhaps insoluble in it) an...".Technically, chromatography uses eluants, it is just contingent they are also solvents and we call them like that. It is true that eluant it is choose by feeling and...More
the adsorbing phase and the liquid one, and solubility might, in principle, play no role at all still being a prerequisite. Mixing solvents allows for an eluant with the right polarity* for the best separation. In the given example using just dichloromethane would have push everything faster, resulting in no separation. A 7:3 mixture would probably have worked anyway, at least with an appropriate colum. *polarity to be taken broadly together with affinity and specific shape and chem interactionMore
Upvote
VOTE
Downvote
more replies
The answer given by @M. Farooq may be interpreted as «the optimization of TLC depends on the choice of the eluents used, which till now (by broader advent of machine learning) was empirical».
In my opinion, this should not the impression to retain. Of course the optimization of TLC depends 1) on the eluent / eluent mixture, but should be seen in context of 2) the stationary phase available (e.g., silica, basic / neuter / acidic aluminum oxide, reverse-phase) and 3) the nature of the sample itself (e.g., molecules which may donate protons to / accept protons from the stationary phase / mobile phase, or not) as most common factors, too.
It is true instruction classes and research groups often stick to «typical» mixtures (e.g., petrol ether / ethyl acetate) to perform TLC for analysis and testing, often as preparation for eventual column chromatography on preparative scale, for «typical» samples. Here and there, there are special recipies to address peculiar separation problems, too.
On the other hand, the optimization a TLC already is formalized algorithmically, for example by Glajch et al. 1982 (up to quaternery eluent systems, here), or with the PRISMA method by Nyiredy et al., back in 1985 (here, English abstract). For the later approach, after an initial thought about the separation problem ahead
multiple solvents, grouped for their properties, are tested as such (purely) as eluent. The results of this initial screening then are further optimized in mixing solvents as eluents
This formalized approach include computations which may be performed independent of a computer. A not-too old review in 2000 (here) explicitly reflects again how to optimize TLC, including the PRISMA method.
The answer given by @M. Farooq may be interpreted as «the optimization of TLC depends on the choice of the eluents used, which till now (by broader advent of machine learning) was empirical».
In my opinion, this should not the impression to retain. Of course the optimization of TLC depends 1) on the eluent / eluent mixture, but should be seen in context of 2) the stationary phase available (e.g., silica, basic / neuter / acidic aluminum oxide, reverse-phase) and 3) the nature of the sample itself (e.g., molecules which may donate protons to / accept protons from the stationary phase / mobile phase, or not) as most common factors, too.
It is true instruction classes and research groups often stick to «typical» mixtures (e.g., petrol ether / ethyl acetate) to perform TLC for analysis and testing, often as preparation for eventual column chromatography on preparative scale, for «typical» samples. Here and there, there are special recipies to address peculiar separation problems, too.
On the other hand, the optimization a TLC already is formalized algorithmically, for example by Glajch et al. 1982 (up to quaternery eluent systems, here), or with the PRISMA method by Nyiredy et al., back in 1985 (here, English abstract). For the later approach, after an initial thought about the separation problem ahead
multiple solvents, grouped for their properties, are tested as such (purely) as eluent. The results of this initial screening then are further optimized in mixing solvents as eluents
This formalized approach include computations which may be performed independent of a computer. A not-too old review in 2000 (here) explicitly reflects again how to optimize TLC, including the PRISMA method.
Very well explained, did you notice the interesting footnote of the PRISMA paper "Die Kunst des Chromatographierens besteht nach wie vor darin, die stationäre Phase auf die mobile Phase und beide auf die Trenntechnik abzustimmen.” which DeepL nicely translated as "The art of chromatography is still to tune the stationary phase to the mobile phase and both to the separation technique". The emphasis on art is mine. –More
Chromatography is really an empirical science. The fluid mechanical side and the adsorption phenomenon is somewhat well understood but choosing mobile phases is essentially based on trials, sadly to date. If someone has given you a ratio, say 8:2, it means someone must have tried it and they found that this mobile phase works the best. The main criterion for choosing the mobile phase is that all the mixture components must be soluble in it. Thus if glycine did not move, this means that the mobile phase has no affinity for this solute (it is perhaps insoluble in it) and it is a poor choice. One then starts to tune the polarity of the solvent mixture. Think about a petroleum sample, which literally has thousands and thousands of different molecules. Now think yourself: will they all dissolve in 100% hexane? You do need an additional solvent.
There is no theoretical restriction in chromatography to use multiple solvents or a single solvent. For example in gas chromatography, you always use a pure gas. The reason for using a mixture of solvents in liquid chromatography originates from the concept of "general elution problem of chromatography" which basically says that there is no single adsorbent or a single mobile phase which can separate everything. Not only separation scientists use a single mixture composition of mobile phase (isocratic), they can also continuously change the mobile phase composition as a function of time. This is called gradient elution. You cannot do that with thin layer chromatography without using fancy stuff.
People are trying to implement machine learning, so that one can enter the molecular structures contained in the mixture, available stationary phases, and with a lot of statistics and tons of experimental data, the computer can tell you to try a certain mobile phase composition.
Chromatography is really an empirical science. The fluid mechanical side and the adsorption phenomenon is somewhat well understood but choosing mobile phases is essentially based on trials, sadly to date. If someone has given you a ratio, say 8:2, it means someone must have tried it and they found that this mobile phase works the best. The main criterion for choosing the mobile phase is that all the mixture components must be soluble in it. Thus if glycine did not move, this means that the mobile phase has no affinity for this solute (it is perhaps insoluble in it) and it is a poor choice. One then starts to tune the polarity of the solvent mixture. Think about a petroleum sample, which literally has thousands and thousands of different molecules. Now think yourself: will they all dissolve in 100% hexane? You do need an additional solvent.
There is no theoretical restriction in chromatography to use multiple solvents or a single solvent. For example in gas chromatography, you always use a pure gas. The reason for using a mixture of solvents in liquid chromatography originates from the concept of "general elution problem of chromatography" which basically says that there is no single adsorbent or a single mobile phase which can separate everything. Not only separation scientists use a single mixture composition of mobile phase (isocratic), they can also continuously change the mobile phase composition as a function of time. This is called gradient elution. You cannot do that with thin layer chromatography without using fancy stuff.
People are trying to implement machine learning, so that one can enter the molecular structures contained in the mixture, available stationary phases, and with a lot of statistics and tons of experimental data, the computer can tell you to try a certain mobile phase composition.
More
VOTE
VOTE
VOTE
VOTE
VOTE
VOTE
The answer given by @M. Farooq may be interpreted as «the optimization of TLC depends on the choice of the eluents used, which till now (by broader advent of machine learning) was empirical».
In my opinion, this should not the impression to retain. Of course the optimization of TLC depends 1) on the eluent / eluent mixture, but should be seen in context of 2) the stationary phase available (e.g., silica, basic / neuter / acidic aluminum oxide, reverse-phase) and 3) the nature of the sample itself (e.g., molecules which may donate protons to / accept protons from the stationary phase / mobile phase, or not) as most common factors, too.
It is true instruction classes and research groups often stick to «typical» mixtures (e.g., petrol ether / ethyl acetate) to perform TLC for analysis and testing, often as preparation for eventual column chromatography on preparative scale, for «typical» samples. Here and there, there are special recipies to address peculiar separation problems, too.
On the other hand, the optimization a TLC already is formalized algorithmically, for example by Glajch et al. 1982 (up to quaternery eluent systems, here), or with the PRISMA method by Nyiredy et al., back in 1985 (here, English abstract). For the later approach, after an initial thought about the separation problem ahead
(source, open access)
multiple solvents, grouped for their properties, are tested as such (purely) as eluent. The results of this initial screening then are further optimized in mixing solvents as eluents
(source, open access)
This formalized approach include computations which may be performed independent of a computer. A not-too old review in 2000 (here) explicitly reflects again how to optimize TLC, including the PRISMA method.
The answer given by @M. Farooq may be interpreted as «the optimization of TLC depends on the choice of the eluents used, which till now (by broader advent of machine learning) was empirical».
In my opinion, this should not the impression to retain. Of course the optimization of TLC depends 1) on the eluent / eluent mixture, but should be seen in context of 2) the stationary phase available (e.g., silica, basic / neuter / acidic aluminum oxide, reverse-phase) and 3) the nature of the sample itself (e.g., molecules which may donate protons to / accept protons from the stationary phase / mobile phase, or not) as most common factors, too.
It is true instruction classes and research groups often stick to «typical» mixtures (e.g., petrol ether / ethyl acetate) to perform TLC for analysis and testing, often as preparation for eventual column chromatography on preparative scale, for «typical» samples. Here and there, there are special recipies to address peculiar separation problems, too.
On the other hand, the optimization a TLC already is formalized algorithmically, for example by Glajch et al. 1982 (up to quaternery eluent systems, here), or with the PRISMA method by Nyiredy et al., back in 1985 (here, English abstract). For the later approach, after an initial thought about the separation problem ahead
(source, open access)
multiple solvents, grouped for their properties, are tested as such (purely) as eluent. The results of this initial screening then are further optimized in mixing solvents as eluents
(source, open access)
This formalized approach include computations which may be performed independent of a computer. A not-too old review in 2000 (here) explicitly reflects again how to optimize TLC, including the PRISMA method.
More
VOTE
VOTE
VOTE