Separation of a Mixture of Ketones of Similar Structure Using Dissociative Extractio

 

KiloMentor blog articles normally are restricted to assessing and highlighting what the author has found useful and important during his career— applying organic chemistry, most significantly in the scale-up of reaction schemes for fine chemicals and pharmaceuticals.

This article is more speculative and requires laboratory experimentation for validation. It proposes a strategy for work-up, isolation, and purification which as far as the author is aware is not backed-up by experimental science. He suggests ways that the technology can build towards taking more into account making these ‘work-ups’ more dependably rugged by incorporating functionality into intermediates that enable reversible derivatizations that make useful partitioning of reaction mixture components between an aqueous and an immiscible organic phase possible.

Although this proposal is speculative in that it is not experimentally verified, it is realistic to expect success based on what chemistry already teaches us. The basic idea is to modify a ‘neutral’ non-ionizable intermediate’s structure so that in a modified form it can be usefully partitioned between an aqueous and an organic phase.

A reader will more easily understand the general principle after examining the specific application to the ketone functionalized as exemplified below.

__________________________________________________

Suppose you need to separate a synthetic reaction mixture of 7-phenyl-3-heptanone (A) and an iodinated product 7-(4’-iodophenyl)-3-heptanone (B). What methods might be applicable to retrieve the two compounds both separated and pure?  Distillation is the most immediate thought, but the boiling points will both be quite high and iodo compounds can be unstable at higher temperatures. Well— vacuum distillation or steam distillation… At an industrial scale vacuum distillation is limited in the actual reduction in pressure that can be dependably obtained. Steam distillation, even vacuum steam distillation, produces large volumes of waste water. It probably requires special equipment and the large maximum volume limits the throughput for the step.  Perhaps the uniodinated aromatic compound could be selectively sulfonated, then selectively extracted into dilute aqueous alkali, acidified, taken back into a water-immiscible organic solvent, concentrated, and desulfonated to give the original 7-phenyl-3-heptanone.

Thinking about this problem caused me to ask myself, “Why is this such a difficult problem?”  The reason I think is that both substances are neutral. What if the problem were to separate the two similarly related carboxylic acids? Although the two would have very close to the same pKas, the lipophilicity difference should make dissociative extraction between hydrocarbon and water phases using just enough base to neutralize the 7-phenyl-3-heptanone work well.

But to apply such a strategy an ionizable functionality is required in the molecules. The carboxylic acid provides this; the ketone does not!

Suppose then I try to provide one. Suppose the mixture is treated with enough R, S-tartaric acid to completely form ketals of both compounds. Each substance gives rise to a pair of optical isomers (two R/S pairs) but each racemic pair has a simple NMR because the ketal carbon is not a chiral center.

From the spectrum of the mixture of tartrate ketals, the proportions of each can be closely estimated. Now, just as in the simple carboxylic acid model, the pKas will not substantially differ but the lipophilicity difference between them will be substantial. Dissociative extraction between hydrocarbon and water phases using just enough base to neutralize the ketal tartrate derived from 7-phenyl-3-heptanone should work well. The uniodinated compound should end up substantially in the aqueous phase as the monosodium salt. The brominated compound should be retained in the hydrocarbon layer.

Note that neither the R- tartaric acid nor the S-tartaric acid will work in this method because then each of the compounds will give two ketals creating two sets of diastereomers in the mixture. 

The Quinoline Tag to enable Phase Switching Isolations in Chemical Synthesis Process Development

 

The KiloMentor blog is dedicated to transmitting useful ‘tricks of the trade’ to the synthetic organic chemists who do the actual laboratory experimentation that creates chemical processes that are intended to be performed at scale.

The KiloMentor blog therefore particularly highlights the importance of the workup of chemical reactions with their separations, isolations, and purification operations since it is these that consume the most time in the plant setting. It is also these methods which are most difficult to uncover by electronic database searching.

In 1999, Hélène Perrier and Marc Labelle published a paper titled Liquid-Phase Synthesis with Solid-Phase Workup: Application to Multistep and Combinatorial Syntheses, J. Org. Chem. 1999, 64, 2110-2113 which was designed to make the workup of every step of a reaction sequence a simple filtration until the tagged target molecule was completely assembled whereupon the facilitating functional ‘tag’ could be cleaved off to leave the target molecule itself in practically pure form.

Another way of looking at what they were enabling is an alcohol-protecting group that so long as it remains in process intermediates makes them simple to isolate and purify from their reactions. In the terminology of Dennis P. Curran, it makes phase switching of these intermediates possible. In the terminology of Jun-ichi Yoshida, it introduces tags on these intermediates.

These chemists discovered that if an alcohol functionality in a starting material was made into the ester of 3-quinoline carboxylic acid then other functionalities in that starting material could be used to elaborate some larger assemblage and all the intermediates in that sequence could be easily isolated and purified by first precipitating them from the reaction solvents as insoluble sulphuric acid salts and then reconstituting the free quinoline by neutralizing the salts and extracting the intermediate back into the solvent for the next reaction in the desired synthetic route.

The salts with sulphuric acid were found to be the choice that most dependably crystallized from a wide range of solvents. The original paper should be consulted for the experimental details.

Finally, in their examples, the 3-quinoline carboxylic acid could be recovered at the end of their test sequences and was sufficiently pure to be reused!

Although these authors don’t go that far, KiloMentor concludes that protection groups generally, besides protecting functional groups, should be selected to promote the isolation and purification of the intermediates into which they are introduced.

Methyl Ethyl Ketone (2-Butanone) as a Uniquely Useful Solvent

Anhydrous Hydrochloride and Hydrobromide Salts

The reason that I write the blog KiloMentor is to pass along whatever ‘tricks of the trade’ I have picked up in my career doing organic synthesis. While I was Research Director and President at QuébéPharma Recherche Inc. in Montreal Canada, a consultant, Lars Svendsen, said in an aside to our discussions that methylethylketone (MEK, 2-butanone) was a particularly good solvent for making a hydrochloride or hydrobromide salt because the water could be removed by heating to distill the water azeotrope. 

The azeotrope between MEK and water boils at 73.4℃ and has 12% water in the distillate. Since MEK itself boils at 79.6℃ the temperature gap between it and the azeotrope will be sufficient that a distillation column will not be required to take off fairly uncontaminated azeotrope.

It is not clear whether this process of making the solution anhydrous is performed before or after adding the basic group that is to be formed into a salt. It would logically be more likely that the heating to distill the azeotropic composition would be done after the neutralization. Distilling first would likely drive off some of the volatile hydrogen chloride gas and could cause some degradation of the ketone solvent.

MEK/Water Liquid-Liquid Extractions

KiloMentor is always on the lookout for mixtures of two liquids that are not completely miscible but could partition similar solutes usefully between the two layers. This would make liquid-liquid partitioning a practicable separation method.  A mixture of methyl ethyl ketone and water might do that. I calculate that at 25℃, water will dissolve 18.2% of its weight of MEK while a MEK phase will dissolve 12.5% of its weight of water.

Other Useful Methyl Ethyl Ketone Azeotropes

Besides its behavior with water, MEK  forms lower boiling azeotropes with ethyl acetate (77℃/18% in distillate) and cyclohexane (71.8℃/40.0% in distillate).  

Managing Competing Reactions

Varying Conditions

The yield of a desired reaction transformation in an organic synthesis can often be reduced by the loss of a portion of the starting material in a competing reaction or reactions. The standard strategy then is to vary the reaction conditions including most prominently the reaction time, the reaction temperature, and the solvent. The logic is that each different reaction course is likely to respond differently to changes, so there is a good chance that conditions can be found where the desired reaction is increasingly favoured.

 Common Trail Changes

A change in the solvent aims to change the difference in the free energies of activation between the different reactions. A change in reaction temperature aims to change the ratio of reaction rates as a function of temperature without changing the actual difference in free energies of activation. A change in reaction time targets the difference in rates directly. A change in the stoichiometry of the starting materials is aimed at discovering a difference in the molecularity of the competing reactions.

Catalysis

There are other less obvious but potentially effective targets. If catalyzed versions of the desired reaction are known or can be conceived, addition of a catalyst can accelerated the desired transformation relative to alternative outcomes. Addition of a catalyst is the most dramatic way to change one free energy of activation relative to another. Furthermore, because catalytic mechanisms are specific in many cases, it is much less likely that both reactions would be affected.

A much, much, less frequent situation arises where the less favoured reaction can be inhibited. Such a rarer situation has been described in the paper, Effect of Amines on O-Benzyl Group Hydrogenolysis,[ Bronislaw P. Czech and Richard A. Bartsch, J. Org. Chem. 1984, 49, 4076-4078 ].As the example, a substance containing both a monosubstituted olefin and an alkyl benzyl ether was treated with hydrogen gas and Pd/C in alcohol with the addition of a non-stoichiometric amount of a non-aromatic amine. In the presence of the added amine only olefin reduction occurred. Benzyl n-nonyl ether was investigated as a test substrate. Using the same conditions but in one case in the presence of 5 mol % n-butylamine and in the other in its absence, the first case gave complete debenzylation and the latter case gave none. This may be a fairly common case in the situation of competing reactions on heterogeneous catalysts. If different sites are responsible for the catalysis of the two reactions one group of sites can be selectively poisoned. Thus we can understand the differing activities of a catalyst prepared with different protocols.

Different Fates of a Common Intermediate

How would one create larger differences in the product ratios arising from differing fates of a common intermediate?

This is the most troublesome situation because there is a common free energy of activation going to the common intermediate so changing that activation energy will not change a product mixture. Also, any subsequent free energies of activation passing from the high energy common intermediate to the transition states will be small and the difference between these two even smaller. It is the rate of quenching of the activated common intermediate that may be crucial here. Take the example of the 5-hexenyl radical. After only a brief existence the material trapped will arise from the cyclopentylmethyl radical.

Can Chemists Make Educated Predictions of Catalysts for Chemical Reactions?

 

When two different reactions are competing with each other to consume your starting material, it might be possible to find some catalyst that only promotes the reaction course you desire. When this happens the reaction that is not catalyzed is suppressed in relative terms. It no longer competes as effectively.  It no longer misdirects as much of your starting material(s) as before.

A catalyst can be very specific for a particular substrate undergoing a particular reaction. Enzymes are the most common examples of this situation. In fact, they dominate this class of performance. Chemists so far cannot regularly predict or prepare enzyme-like catalysts. They are just too demanding in their structures. Sometimes an enzyme can be found that catalyze the transformation you want. Such enzyme use is not what I want to talk about here.

Let us just look at the first kind- catalysts; those that are not enzymes. The question in the title is a practical one.  A chemist might want to use such a catalyst to decrease the severity of a reaction’s conditions so that it would proceed under milder conditions. At the same time or alternatively, (s)he might want to suppress a competing reaction where only the desired transformation would be accelerated by such catalysis.

Many catalysts function by reducing the height of the activation energy-hill for the reaction in question. Solvent choice is the most common example of this type of reaction acceleration. It is so common that it is not even recognized as catalysis. But solvent molecules do meet the criteria for being catalysts. A well chosen solvent lowers the activation energy for its reaction, doesn’t itself get consumed in the reaction, and is not part of the reaction equation linking starting materials and products and is conserved during the reaction. Now most often catalysts are used in sub- molar amounts compared to the reaction components and this is not true of solvents which are use in massive excess. Nevertheless, the quantities of catalyst required are not an element of the definition of cataalyst. Since we don’t want to consider large changes that massively modify the chemical environment of the reacting species I need to specify in this discussion that the catalyst must be a chemical species added usually in sub-molar amounts, even though once a catalytic effect can be demonstrated at low levels the concentration of the catalytic species can be increased within reason while still retaining the essential same conditions of the uncatalyzed reaction.

To predict catalysis chemists need to know or hypothesize the mechanism of the uncatalyzed reaction they are seeking to enhance. Furthermore, the chemist must learn or hypothesize what the pertinent transition state with the highest activation energy looks like. Then, (s)he must imagine a path that avoids that transition state but provides a new scheme for bond-making and bond-breaking that leads to the same chemical result.

This leads to the new question: are there predictable classes of solutions to this problem of imagining a new pathway?

Covalently-Bound Catalyst Intermediates

We can imagine a situation where the starting material(s) first covalently bind not to each other but instead one of the starting species reacts first with the catalyst and only then does another starting material species react further with the catalyst-starting material complex giving rise eventually to products and expelling the catalyst molecule unchanged.

Now the covalently-bound catalyst intermediate can be of the same structural type as the complex formed with the other starting material or it can be different. In the case of the catalysis of imine formation by the catalysis of beta glycine, the covalently-bound catalyst intermediate is an imine like the desired final product. In the case of the catalysis of hydrogenation by formation of a complex between platinum metal and hydrogen, the covalently-bound catalyst intermediate is quite different from the final product type.

It is not sufficient that the chemist can imagine bond reorganization through some catalyst-involved scheme, for that scheme be used in nature the overall activation energies must all be less than the controlling activation step in the original uncatalyzed scheme. This can only be shown by experiment! 

This new scheme to reorganize the bonding by somewhat simultaneously making new bonds and breaking old ones must meet the criterion of satisfying the least motion principle. This is accomplished in the practice of organic chemistry by drawing a scheme using hooked arrows. If such a scheme can be drawn there is a possibility that a realistic mechanism exists.

Increased Encounter Type Catalysis

A catalytic effect is possible without there ever being catalyst covalently bound to any of the reaction moieties. Reaction rate is determined by two factors. One represents the likelihood of the reactants transforming to products when they accidentally collide. A second increases the likelihood of such a collision. It is this second type which I have called Increased Encounter Type Catalysis.

Increased encounter catalysts elevate the effective concentrations of the reacting starting materials in some common phase which they both simultaneously occupy. Phase-transfer catalysis is of this type. 

Transition metal catalysis is often of this type. Imagine that two reactant molecules each form transient complexes with the same transition metal; that is they both are ligands to the same metal atom at the same instant. As a consequence they are being brought close to each other which is to say both of their concentrations in the same micro-environment is very high. These species are not covalently bound to the metal. The bonding is Lewis acid-Lewis base type. It is possible to imagine that these ligands ultimately react together at a higher than normal rate because they have been more effectively brought close together.

Most Preferred Conformation Catalysts

Outside of enzymatic catalysis, this type of catalysis is most frequently observed or imagined for intramolecular reactions. In order for a reaction (the reorganization of bonding) to occur it is insufficient that two loci of potential reaction bounce into one another. They must encounter each other in an orientation that facilitates least-motion bond reorganization by a somewhat simultaneous bond-breaking and bond-forming.

The catalysis of the acyloin condensation of long-chain dicarboxylic esters by sodium metal is an example of this. Both ester groups are attracted to the surface of the sodium metal (they form complexes with the metal). When they are both attracted to the same surface they are brought closer together as the long chain assumes a conformation that facilitates this. 

Acid Catalysis

When bonds are made or broken, charged intermediates are often formed which are higher in energy than the reactants. Since the intermediate is higher in energy than the reactants, the transition state would be even higher in energy, and hence more closely resemble the charged intermediate. Anything that can stabilize the charges on the intermediate and hence the developing charges in the transition states will lower the energy of the transition state and catalyze the reaction. Charge development in the transition state can be decreased by either donation of a proton from general acids (like acetic acid or a protonated indole ring) to an atom such as a carbonyl O which develops a partial negative charge in the transition state when it is attached by a nucleophile. Proton donation decreases the developing negative in the transition state.

Base Catalysis

Alternatively, a nucleophile such as water which develops a partial positive charge in the transition state as it begins to form a bond to an electrophilic C in a carbonyl, for example, can be stabilized by the presence of a general base (such as acetate or the deprotonated indole ring). Proton abstraction decreases the developing positive charge.

As far as predicting acid or base catalysis is concerned; if you know or hypothesize that the reaction you are trying to catalyze requires that hydrogen atoms be transferred to get from starting materials to products then acid, base, or acid/base catalysis may work.

  

Tautomeric Shift Catalysis

In many chemical reactions,  the bond reorganization requires a simultaneous shifting of a hydrogen from one atom to another. This transfer can often be made less energy demanding by interposing another molecule which can act catalytically by accepting the hydrogen from one location in the transition state and delivering a hydrogen atom to another location in the product by accepting the hydrogen on one atom of the catalyst and delivering a hydrogen atom from a different atom of the catalyst all the while satisfying the principle of least motion. Catalytic species that can do this typically exist as tautomers. Tautomers are molecules for which usually two approximately equal energy structures exist that differ by the point of attachment of a hydrogen atom. The carboxylic acid functional group can have two such forms. The hydrogen of the carboxyl can be attached to either of two equivalent oxygens. A beta diketone has the diketone form and the veto enol form.

Free Radical Chain Initiators are not Catalysts

Some substances only resemble catalysts. Initiators of free-radical reactions share with true catalysts the property that they both are often only required in sub-stoichiometric amounts. These initiators however are consumed during the free-radical process. The initiator is destroyed when it does its function and so does not meet a criterion of a catalyst.

Filter Clogging

 

Sometimes during a synthetic organic reaction workup, small particulate matter clogs an attempted filtration. This situation is called filter ‘blinding’. The cause of this blinding may be the major component itself or some small particulate solid or sticky impurity. Filter papers clog more readily than filter cloths and filter cloths differ from each other depending upon the material of their construction. When filter paper clogs in developing the small-scale laboratory procedure, filter pads cut from the filter cloth material used in your pilot plant or plant should be tested. Perhaps what you are seeing using filter paper in the lab will be more manageable at scale but you need to confirm this.

If the blinding of the filter is attributable to impurities and not the major compound of interest, a first filtration of the solution before crystallization is induced can remove a useful amount of the material that is blocking your main filtration. This ‘sparkling’ filtration can use a thin layer of a filter aid even include charcoal.

To overcome the most difficult problems filter aids (Celite^®, Hyflo^®, kieselgur) are added to the hot mixture before the main product has precipitated, the solution is stirred vigorously, and then slowly cooled to bring out the main product onto the inert solid. The filter aid and product are collected together and the majority of the dissolved byproducts, co-products, and processing chemicals may be carried away with the filtrate. The product can be dissolved off the filter-aid and the workup of the product, now in solution, continued.

An Advance in Making Inclusion Complexes: US 2,620,716

 

This patent pertains to a treatment of crude petroleum which separates some fractions containing particular classes of chemical structures from other fractions containing other classes. It makes reference to making inclusion complexes that trap and make insoluble and hence filterable certain kinds of chemical structures while leaving undisturbed in the bulk oil other structural types.  Its examples use the chemical agents urea and thiourea but their use is not new and not the claimed invention. KiloMentor has written blogs about such inclusion complexes. They are in fact old.

What this inventor claims is new is the claimed discovery that certain ‘contacting agents’ accelerate and make more dependable the formation of these valuable complexes and by so doing makes their application practical.

So much for the generalities!! That is how patent lawyers talk!

The teaching that excites me is that straight chain hydrocarbon-like molecules quickly and dependably form insoluble, retrievable, filterable complexes with urea when stirred in neat acetic anhydride, neat sulfur dioxide, and ‘perhaps’ in neat furfural. 

Well, KiloMentor is interested in techniques that are promising for separating very similar chemical structures by rugged cheap methods that work at scale. Well, urea, sulfur dioxide, furfural, and acetic anhydride are inexpensive and urea inclusion complexes have the ability to separate straight chains from branched chains.