This page looks plain and unstyled because you're using a non-standard compliant browser. To see it in its best form, please upgrade to a browser that supports web standards. It's free and painless.
kilomentor | 24 January, 2007 18:26
Creating rugged scaleable processes has always been important for process chemistry. Exploratory and early development small scale reactions are most frequently purified using chromatography, but chromatography is not readily scaled up and it is expensive. Larger scale requires more rationally selected purification methods. A goal of this presentation is to show how using sulfuric acid esters can efficiently and economically improve preparative scale organic synthesis.
In 1998, Dennis P. Curran published Strategic-level Separations in Organic Synthesis: From Planning to Practice, Angew. Chem. Int. Ed., 1998, 37, 1174. In my opinion, for practical experimentalists, this paper is the most important paper ever written. This review organizes the theoretical principles of organic synthetic separations around the central concept of phase switching. According to Curran’s theoretical treatment of separation, if a substructure in the form of an intermediate, which is intended to be incorporated into a final synthetic target, is to be purified, the purification always uses transfers of the intermediate from one phase to another; for example, from solution to solid (crystallization or precipitation), distillation (liquid to gas), acid-base aqueous extraction (organic solution to aqueous solution), sublimation (solid to gas). Purification occurs because impurities, which have at least different physical properties from the target intermediate, are not able to completely make the switch and so a portion of the impurity is left behind in the original phase while most of the desired intermediate is switched from the first to the second phase. The power of the phase switch depends upon how much of the impurities are left behind. A well-chosen phase switch removes almost all the particular impurities of the crude mixture. A poorly chosen phase switch removes fewer impurities and often losses a significant portion of the desired material in the process.
Curran says that every intermediate has a natural phase to which it is most disposed. According to his definition, he is talking about a natural liquid phase in which it is particularly and characteristically soluble. Neutral organic molecules of 5 carbons or more are naturally soluble in organic solvents. This is the like dissolves like rule we are all familiar with. Carboxylic acids, even those containing 20 or 30 carbon atoms, are characteristically soluble in water most commonly as alkali metal salts. By this we mean not that they are insoluble in organic solvents but that when protonated, when showing their acidity, their most differentiating property, is that they can form salts which are extractable into water. Analogously, amines’ most characteristic behaviour is their solubility in an acid aqueous solution. Looked at in this way a separation like purifying a carboxylic acid by extracting the crude mixture dissolved in an organic solvent with aqueous base to form the water soluble salt and leaving neutral and basic organic impurities behind in the organic solvent, which we do as second nature, is viewed in terms of basic principles as a switch of the carboxylic acid to its natural phase, aqueous (basic) solution, which is effective purification because the neutral and basic impurities cannot make the switch and get left behind in the original organic solvent.
Sulfuric acid esters of hydroxyls have not been frequently used in organic chemistry even though alcohols and phenols, from which sulfuric acid esters are prepared, are among the most common functional groups.
Converting an alcohol or phenol into a sulfate ester is not a well-known derivatization. Yet converting an alcohol or phenol into a sulfuric acid ester can be very useful because a new very strongly acidic functionality is created in the intermediate that makes the substance significantly more water soluble in aqueous base, even in instances where the parent alcohol or phenol is practically insoluble in this medium. Simple hydrocarbon alcohols of more than five carbon atoms are characteristically insoluble in acidic, neutral or basic aqueous solution. Sulfuric acid derivatization changes these alcohol’s natural solution phase from organic solvent to water.
Phenols are slightly different. Phenols are slightly acidic and even a C20-C25 substance can be at least partially extracted into strongly basic aqueous solution using an excess of a strong alkali. Phenols natural phase would be strong aqueous base according to Curran’s classification. In practice however, the drastic conditions required to extract high molecular weight phenols into aqueous solution very often cause degradation in a complex substrate and so large, diversely functionalized phenols are more wisely described as naturally soluble in organic solvents. As a consequence, derivatization of higher molecular weight phenols to sulfuric acid esters changes their classification from organic soluble to water soluble.
Making a derivative of a desired functionality in order to make it amenable to phase shifting would not be so useful if one had to take this sulfate ester as starting material into the next step of the synthesis. This might cause more difficulties than the separation would be worth because the reactions of alcohols and sulfate esters are not interchangeable. However this derivatization is reversible; the alcohol or phenol can be regenerated by simple means in a single step without expensive reagents. This is in fact an essential requirement for the methodology to be useful for purification. Thus the combination of derivatization, phase switching and regeneration of the alcohol or phenol potentially can lead to a dramatic increase in purity as functionally different impurities, which cannot undergo the phase switch are removed.
What does this mean. If one has access to a method of separation, like sulfate ester formation, which provides a robust phase switch and which is selective even among isomers containing the same functional group (as this method is), at least in principle one could use a single step conversion which may only give a 20% yield of a desired intermediate and 80% closely related isomers or by-products and still fish-out the desired 20% of material from the reaction mixture using the power of the derivatizing phase switching method.
Thus for example if we take the example of the hydration of the olefin 4-methyl 2-pentene. The Markovnikov or anti-Markovnikov addition of water to this compound is not going to be highly selective and some substantial amount of the other compound can be anticipated. More to the point here, is that the separation of these isomeric alcohols does not appear like it will be simple. It is to be anticipated however that there will be a large difference in their rates of formation of sulfuric acid esters using say trimethylamine sulfur trioxide (which is commercially available). 4-methyl-2-pentanol would be expected to form a sulfuric acid ester much more rapidly than 4-methyl-3-pentanol so if these two substances are allowed to compete for only enough sulfur trioxide react to react with the former one should not be surprised to end up with a mixture in which only the former has reacted. After addition of potassium hydroxide to form the potassium salt it can be readily separated from the unreacted 4-methyl-3-pentanol and the 4-methyl-3-pentanol can be regenerated by mild acid hydrolysis in water. Presto! The two alcohols are cleanly separated.
The formation of alcohol sulfuric acid esters was better known early in the twentieth century. William B. Hardy and Mario Scalera published a paper, JACS 74 5212 (1952), showing that alcohols, phenols and even primary amines could be separated simply by selective reaction with the reagent triethylamine-sulfur trioxide. These industrial chemists were not able to explore the significance of these observations because their research was a distinctly targeted application. They urged others to follow up the lead. There are no future citations, which qualify as such follow up.
Hardy and Scalera’s intuition arose from seeing that although anthrahydroquinones reacted with triethylamine-sulfur trioxide essentially quantitatively to produce di-sulfuric acid esters when the substrate also contained α substituents anthranol sulfuric acid ester by-products were noticeably enhanced. They postulated that the α substituent hindered the formation of an ester at the phenol location close by the substituent. For example in the esterification of 1-chloroanthraquinone, the 1-chloro-9-anthranol-sulfuric acid ester is the principal reduction product.
They were able to separate 1- and 2- propanol from an equimolar mixture using the triethylamine- sulfur trioxide reagent with just sufficient reagent to react with one half of the alcohol groups present.
They were also able to separate 2-ethyl aniline from 4-ethyl aniline by selective competitive sulfonation of the amine of 4-ethyl aniline. It should be noted that both reactions proceeded well in the absence of the other isomer and in the presence of a sufficient quantity of chlorosulfonic acid. Although both reactions were fast the reaction of the 4-ethyl aniline must have been at least 100 times faster it allow it to become the quantitative product. 2-hydroxybiphenyl could be separated from 4-hydroxybiphenyl because in this case also the 4-hydroxybiphenyl was substantially more reactive.
Samples weighing 8.5 gm (0.05 mole) each of o-phenylphenol and p-phenyl phenol were dissolved at 70ºC in a solution consisting of 4.4 gm (0.11 mole) of sodium hydroxide in 58 cc of water. Nine and two tenths gm (0.11 moles) of sodium bicarbonate and 13.8 gm (0.075 moles) triethylamine sulfur trioxide were added. The mixture was stirred for 3.5 hours at 45-50ºC. After cooling, the mixture was extracted with ether. Two lower layers were diluted to 125 ml with water and 30 cc of 5 N sodium hydroxide was added. Addition of sodium chloride to the solution precipitated the sulfuric acid ester salt of p-phenylphenol in 81% yield, as shown by hydrolysis in acidic solution to pure p-phenyl phenol, mp 163-164ºC. The high purity of this product is demonstrated by the fact that a mixture comprised of 10% o-phenyl phenol and 90% p-phenyl phenol softened at 80ºC and melted at 112-157ºC.
A mixture of 30 gm (0.5 moles) each of 2-propanol and 1-propanol was treated in 125 ml of chloroform with 97 g (0.525 moles) of triethylamine-sulfur trioxide. The clear solution was allowed to stand at room temperature for two and one-half days. The solvent was then evaporated under vacuum in such a fashion that the temperature did not exceed 40ºC. Addition of 200 cc of hexane caused a layer separation and the lower layer weighing 177 gm was separated and cooled. Since crystallization could not be induced in this way a 3.5 gm (0.01 mole real) aliquot was added to 10 cc of water containing 2.2 gm (0.011 moles) of benzyl isothiouronium chloride. The vacuum dried solid obtained weighed 2.4 gm a 77.2% yield of the sulfuric ester of 1-propanol as the benzylisothiouronium salt; sintering point 113ºC; mp 115-116ºC.
Chlorosulfonic acid combined with either of dimethylaniline or pyridine can be used to make the sulfonate sates of phenols. The in situ product is the dimethylanilinium or pyridinium salt of the alcohol sulfonate and dimethyl anilinium/pyridinium hydrochloride. If an strongly alkaline solution of potassium hydroxide like 1:1 gm/gm is now added, the potassium salt of the alcohol-sulfonate is creates in the salt solution loaded with the common ion potassium. It is usually predominantly insoluble. Washing with a lipid dissolving solvent like diethyl ether removes the solvent pyridine or dimethylaniline and the small amount of water. The residue, which will contain potassium sulfate and potassium chloride can be usually extracted into hot 95% ethanol to take up the organic salts and leave behind the strictly inorganic materials. So long as the solution is filtered in a heated funnel, it can be clarified. Upon cooling solid crystals almost invariably separate and these can be recrystallized again this time from hot water to remove final traces of inorganic materials.
One can see in these potential treatments the advantage of the derivatization. The product is only poorly soluble in even polar organic solvents and usually only in hot water. It is also insoluble in lipid solvents. Combine this with the demonstrated propensity for the reaction to be highly sterically selective in competition reactions, makes the removal of even quite similar contaminants easy.
This methodology would not be particularly useful if it were not true that refluxing with water and a small amount of acid can hydrolyse the ester and return the starting alcohol or phenol in a purified condition.
This technology should always be considered alongside the alternatives. Phenols can be separated based on there relative acidities. They can also be separated based on the substitution pattern around the phenol group based on solid complexes with calcium chloride, calcium bromide, lithium bromide, magnesium chloride etc, US 4499312, US4429168 or US4447658 or US4424381(dihydric phenols). Phenols can be extracted from an aqueous solution where they have been neutralized exactly with caustic. Extraction with a solvent of medium polarity such as ethyl ether or ethyl acetate will cause sterically hindered phenols with large hydrophobic groups near the phenolic hydroxyl to hydrolyze to free phenol and alkali hydroxide and the free phenol will be extracted into the organic solvent. Impurities with an unhindered phenol will remain dissolved in the aqueous layer. This procedure is called dissociative extraction because the salt dissociates and reforms the free phenol which is extracted.
Alcohols form insoluble complexes with salts in hydrocarbon liquids. Calcium bromide and lithium bromide are the most common choices although calcium chloride anhydrous works in some cases.
Although the literature frequently makes light of the synthesis of these derivatives problems do arise. Hong, Scott and Rankin report difficulties preparing the O-sulfonate of the secondary alcohol N-(3,5-dichlorophenyl)- 2-hydroxysuccinimidewhich is an agricultural fungicide. This compound seemed to readily hydrolyze or degrade. This may be because the compound would be expected to be hydrolyzed in neutral or acidic solution while in base the succinimide itself would be cleaved or the sulfate could be expelled since it is a beta substituent to a carbonyl.
Chlorosulfonic acid at –78ºC in a 60:40 v/v mixture of anhydrous chloroform/ether was finally used to make the substance. 4 mmoles of substrate was treated at –78ºC with 8 mmoles of chlorosulfonic acid added slowly from a syringe during 30 minutes. Nitrogen was used to purge the free hydrogen chloride gas from the mixture. That this condition led to reaction indicates to me that the substance was certainly reactive enough! Work up was done by adjusting the pH to 7-8 using aqueous sodium bicarbonate (the calculated amount was used I would assume-the addition was done in a dry ice-acetone bath. The reaction contents were transferred to a regular ice bath and stirred. When the temperature of the contents reached the bath temperature, the white slurry was quickly filtered. The solid was washed with cold chloroform presumably to wash of starting material that might dry on the solid and then with 100 ml of ice-cold water ( to remove sodium sulfate and sodium chloride). The solid was resuspended in ice-cold water to give 0.95 gm (60%) of the sodium salt as the dihydrate. The material was apparently now reasonably stable since t was reported to melt with decomposition only at 163-172º C.
This use of chlorosulfonic acid alone it seems is following S.R. Sandler and W. Karo, in Organic Functional Group Preparations, Vol. 3, H.H. Wasserman, ed, Academic Press, San Diego, 1989, p 129. Another compendium which is claimed to cover this methodology is E.E. Gilbert, Sulfonation and Related Reactions, interscience Publishers, New York, New York, 1965.
Reagents include chlorosulfonic acid, triethylamine sulfur trioxide, pyridine sulfur trioxide, dimethylaniline sulfur trioxide, pyrophosphate, fuming sulfuric acid, and sulfamic acid.
The reaction of sulfuric acid with a variety of compounds in the presence of dicyclohexylcarbodiimide and a polar solvent deserves examination. The method is reported to give sulfate esters in good yield and is reported to be particularly useful for the formation of sulfate esters of compounds that are unstable to reagents such as chlorosulfonic acid or pyridine sulfur trioxide. [Mumma Lipids, 1, 221, (1966).; Hoiberg, Mumma, Biochim. Biphs. Acta 177, 149, (1969); Hoiberg and Mumma, J. Am. Chem. Soc. 91, 4273, (1969). This latter reference reports the selective sulfation of polyfunctional molecules.]
A mixture consisting of 12.1 gm (0.1 moles) each of 2-ethylaniline and 4-ethylaniline was placed in a 100 cc of chloroform in a flask fitted with a mechanical stirrer and 20.2 g (0.11 mole) of triethylamine sulfur trioxide was added. While stirring the temperature was allowed to rise gradually from 15 to 25ºC during a five-hour period. A white solid was filtered off at 7ºC and washed with hexane. It weighted 19.4 gm (64%); mp 101-103ºC. The identity of this material as the para isomer is shown by the fact that no depression in melting point occurred when it is mixed with an authentic sample. As further evidence, the ortho isomer yields an oil under these conditions.
General Procedure for o-phenol sulfonic acids and potassium salts
Acta Chemica Scandinavica B 32(1978) 489-498
One equivalent of phenol was treated with 1.1 equivalents of chlorosulfonic acid in dichloromethane at –20º C. the mixture was allowed to stand 2-3 hours until it had reached room temperature and was then poured into ice water. The amount of water was chosen so that the solution became 1.0-1,5 M with respect to the sulfonic acid.
A suitable amine (1.1-2) equivalents was added then added. The selection of the amine was made according to the principles presented, usually n,N-diethylaniline or 2-methylpyridine for more hydrophobic acids. If necessary, the pH was adjusted to less than 1 with sulfuric acid. The water layer was extracted with dichloromethane and, if the extraction was insufficient, a better solvent selected from Table 5 was used for the extraction. Some water was added to the combined organic layers containing the o-phenolsulfonic acid and the mixture was stirred and titrated with 1 M potassium hydroxide to pH 7-7.5.
The solvents were separated and the water layer washed twice with dichloromethane to remove the amine. The water was then evaporated and the potassium salt recrystallized from a suitable solvent.
Potassium 3,5-dimethylcyclohexylsulfate
One mole of of 3,5-dimethylcyclohexanol was treated with 1.1 moles of chlorosulfonic acid in ether at –10º C. The work-up procedure was analogous to that of o-phonolsulfonic acids using N,N-dimethylaniline and dichloromethane. Yield 71%.
O-phosphates are similar?
One would expect that the o-phosphate derivative would have similar abilities to transfer a material from the organic layer to become a hydrophilic substance. In fieser & fieser Vol. 5 the combination of phosphorous acid (H3PO3) and N-methylimidaszole in the presence of mercuric chloride is reported to give a zwitterionic species that reacts with alcohols to produce o-phosphates. The references are H. Takuku, Chem. Pharm. Bull. Japan 21, 1844 (1973) and 21, 2068 (1973).| « | September 2008 | » | ||||
|---|---|---|---|---|---|---|
| Su | Mo | Tu | We | Th | Fr | Sa |
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | 29 | 30 | ||||