The Good, The Bad And The Ugly

The Organic, the Aqueous, and the Fluorous phasesin their Mexican Stand-Off. Taken from [1].

‘Every gun makes its own tune.’ 

– The Man with No Name (Clint Eastwood), in ‘The Good, The Bad And The Ugly’.

 

Chemistry, like life, is dictated by the golden rule 'like dissolves like'. Just like a Western film - the showdown is always between a good guy and a bad guy, right? But then there is the great Sergio Leone, who shows that human nature is not that simple (and Eastwood makes his day in this way). There can be 'The Good, The Bad, and The Ugly', and in chemistry there can be someone else who doesn't want to mend any fences with neither the greasy organic nor the watery aqueous. That, has something to do with a guy called 'Fluorine'.

A CH4 Molecule. Taken from [2].

A CF4 molecule. Taken from [3].

Fluorine is always special. It ranks the first in electronegativity (a really electron density bully here), and its size is huge. 4 small hydrogen atoms surround a big carbon atom to form methane (CH₄), but when the hydrogen is swapped as fluorine, the resulting CF₄ becomes 4 huge fluorine atoms burying 1 skinny

poor carbon atom. If you imagine you have a long carbon chains with all fluorines substituted on it - the result can be said as a 'fluorine blanket'.

That leads us to a new concept known as 'Fluorous'. 'Fluorous', to start with, is related to fluorine.  Yet to understand how we can get something from this, we have to be careful about the meaning. A compound with fluorine atoms on it does not necessarily mean it is 'fluorous', as 'fluorous' takes a rather narrow definition. First, you have to have a lot of fluorine atoms in a compound to be fluorous (the current understanding is that the fluorine content should contribute to at least 60% of the total molecular weight¹), and second, the fluorine atoms have to be closely packed on the organic molecule, not dispersed throughout a fatty organic compound with a large molecular weight. If these 2 criteria can be fulfilled, when you dissolve a fluorous compound or solvent into an organic phase, even though it is non-polar (means it hates water), it can still be separated into 2 phases, and so a fluorous-organic bi-phase can be established. This is great because you can start to engineer novel concepts in reaction design and catalysis, and this should simplify separation and purification procedures.

This time, I will share with you 2 pieces of great work that is related to fluorous chemistry.

by Ed Law
27/03/2015

Reference:

1. http://www.fluorous.com/journal/?paged=65

2. http://commons.wikimedia.org/wiki/File:Methane-3D-space-filling.png

3. http://commons.wikimedia.org/wiki/File:Tetrafluoromethane-3D-vdW.png

4. I. T. Horváth, J. Rábai , Science, 1994, 266, 5182, 72-75. 

5. For those organic compounds which have a fluorine content of more than 60%, they are specifically known as ‘heavy’ fluorous compound. There are also ‘light’ fluorous compounds (which has <40% Fluorine by weight, see Handbook of Fluorous Chemistry, Chapter 8 and references cited therein). In my following 2 articles, the term ‘fluorous’ means ‘heavy’ fluorous compounds.

Shake Hands In The Fluorous Phase

Match (left) and Mismatch (right). When a (S)-Fluorous BINOL interacts with a (R)-amino alcohol (left), it results in strong fluorescence. This is not the case for the interaction between a (S)-Fluorous BINOL and a (S)-amino alcohol (right), in which only a little fluorescence can be detected. Taken from [1].

The first paper regards the development of a chemical sensor molecule that can sense enantiomeric phenomenons in a fluorous phase. The compound is a chiral perfluoroalkyl-BINOL based diketone, and the researchers have shown that this new compound can serve as an enantioselective fluorescent sensor when it is dissolved into a fluorous phase. That kind of suggests that, when an enantiomer of this compound encounters a substrate with a particular enantiomeric configuration, they will interact and result in a large enhancement in fluorescent signals. The researchers have hypothesized, and also have confirmed that the rationale of this is due to formation of large-sized aggregates which can then be observed by dynamic light-scattering techniques.

The first generation (S)-BINOL. This candidate is not 'fluorous' enough to serve as a fluorous sensor. Modified from [1].

Now the big question, how can they make a greasy BINOL to partition into a fluorous phase? Their first generation compound will not work, and that illustrates the point of the 2 criteria I have mentioned above. True, it does contain 2 attention-attracting CF₃ groups (which medicinal chemists love because they often positively impact the activity of a drug molecule), but the compound is simply not fluorous enough. The fluorine content is way lower than 60%, and they are far away from each other. So, as they have confirmed, this compound simply dissolves in the organic solvent and does not crash out as a second phase. So how can they solve the problem? You get it – increase the fluorine content of the BINOL compound. So, the second generation BINOL they use in this work, consists of 2 C₇F₁₅ groups (a total of 2 sets of 15 closely-packed fluorine atoms here!) and this is more than enough for the compound to partition into a fluorous phase. 

The novel fluorous (S)-BINOL. This compound is fluorous enough to preferentially partitioned into the fluorous phase in a bi-phasic system. Modified from [1]. 

To explain the reaction they study, I will use a notation and a diagram here. The reaction involves the BINOL and also an amino alcohol, and both of these compounds are enantiomeric. The 2 enantiomers of the BINOL are denoted as ‘A’ and ‘a’ (R and S configuration respectively), and the 2 enantiomers of the amino alcohol are denoted as ‘B’ and ‘b’ (ditto).

The 'Match' cases that will lead to intense fluorescence signals. Modified from [1].

What they have discovered is that the R enantiomer of the BINOL can interact only with the S enantiomer of the amino alcohol to give a significant increase in fluorescence signal, and vice versa. That means ‘A’ interacts with ‘b’, and ‘B’ interacts with ‘a’. They have also analyzed the reaction mixture and are able to propose that a nucleophilic addition has occurred between the BINOL and the amino alcohol to form an oxazolidine structure. If A interacts with B (or a interacts with b), the fluorescence intensity is much smaller – which means a ‘mismatch’ in pair takes place. So, this is seen as a sort of chiral recognition. The scope of the amino alcohol can be expanded and even diamines have been tried.

This is certainty a great reaction done in the fluorous phase, as it involves a highly perfluorinated BINOL compound. Coupled with fluorescence techniques, that makes the chiral recognition technique easily observable and quantifiable at the same time.

by Ed Law

27/03/2015 

Reference:

1. Enantioselective Fluorescent Recognition in the Fluorous Phase: Enhanced Reactivity and Expanded Chiral Recognition. C. Wang, E. Wu,  X. Wu, X. Xu, G. Zhang, L. Pu.

DOI: 10.1021/ja512569m

Fluoro-Soap

Structure of a micelle. Taken from http://en.wikipedia.org/wiki/File:Micelle.png

This great paper is on the Chemistry-Biology interface. The group has developed a novel fluorinated (rightfully a ‘fluorous’) detergent that can find potential applications in membrane biology. Membrane proteins are biologically important, they serve as many drug targets, and infamously difficult to deal with. Detergents are often used to solubilize the membrane proteins so that they could be manipulated for further studies. Detergents are amphiphiles, which means they have both a polar and a non-polar part, so that they can interact with the phospholipid and thereby disrupt the membrane bilayer structure.

As I have said, a fluorous compound does not necessarily like the 'fatty' organic layer, so a 2-phase system can be generated potentially. However, traditional wisdom suggests that although fluorinated compounds don't like aqueous phase too, that does not make them good detergents. The reason is two fold - first, the fact that fluorous and organic are not miscible means that it can be hard for fluorous compounds to interact with the organic membranes. And fluorine atoms are large compared to hydrogen atoms, so sterically it can be tough for them to mingle with the membranes. So, fluorous compounds are considered detergent-resistant, and they are less likely as candidates of great detergents. That is not the end of the world - because if there exists a ‘fluorinated detergent’, then its fluorinated tail will be unlikely to interact with the hydrocarbon part of the membrane, and then the protein-membrane properties will not be affected and the integrity of the membrane protein will be restored. This paper just shows one of these cases.

The structure of F6OM, the detergent the researchers developed, and F6OPC, another detergent to compare with. Taken from [1].

The researchers have synthesized a novel fluorinated detergent known as F6OM, which consists of a carbohydrate end and a fluorous end (C₆F₁₃). They have compared F6OM with another fluorinated detergent known as F6OPC, to show how different their properties can be. While F6OPC also consists of a C₆F₁₃ terminal, it also consists of a polar, and indeed zwitterionic, end - a cationic ammonium and an anionic phosphate here. The respective self-assemblies of the 2 contenders are very different - F6OM appears as long rods but F6OPC looks like small spheres. 

By using light scattering experiment, the group discovers that micellar F6OM can solubilize with  a derivative of phosphocholine (POPC), and this solubilization can be enhanced by increasing the F6OM to a certain concentration, or increasing the temperature. This is  not the case for the counterpart F6OPC. The observations agree to conventional understanding – a higher temperature should encourage membrane destabilization and a faster detergent translocation. The researchers carry out further isothermal titration calorimetry (ITC) to develop a quantitative understanding of the phenomenon. Furthermore, they can establish a phase equilibrium of F6OM, as present in a bilayer or a micelle. At a medium concentration, both types are seen to co-exist. One thing I would like to point you to is the use of  31P and 19F NMR in this work. The dynamic upfield shift of CF₃ signal in 19F NMR signifies solubilization.   

The SDS-Page analysis for refolding experiment. When the concentration of F6OM reaches a certain level, folded protein can be observed. This is not so in the case for F6OPC, even when its concentration is raised to the same level as F6OM. Taken from [1].

An interesting aspect of membrane biology is the study of unfolding / refolding of proteins and membranes, in a sense you are building (or reconstituting) the membrane architecture from scratch. Our F6OM turns out to be a great candidate as a chaperone for this. By adding a CH₂ spacer between the fluorous group and the carbohydrate section, the analogue shows promise. First, this analogue has a self-assembly to make it appear like a membrane bilayer, and then a phospholipase (OmpLA) can be refolded, and the whole protein-membrane complex becomes a functional proteolipase. The SDS-Page analysis shows that when the concentration of  F6OM reaches a certain level, the phospholipase can be refolded into an active state, and this seems to out-rival the other contender, F6OPC.

I think this is a great paper that illustrates the use of a chemical compound for biological applications. There are a lot of nice techniques inside – electron microscopy, light scattering, ITC, Fluorescence Spectroscopy, NMR and various assays. I am aware of some of them but I do not have practical experiences on these for my work, so I have learnt a lot from this paper. I encourage you to read more about those techniques, especially if you are working in biochemistry / chemical biology!

by Ed Law

27/03/2015

Reference:

1. A Fluorinated Detergent for Membrane-Protein Applications E. Frotscher, B. Danielczak, C. Vargas, A. Meister, G. Durand, S. Keller.

DOI: 10.1002/anie.201412359