Solvent recrystallization experiments

Polymorph discovery by solvent recrystallization is based on the principle that the less stable forms recrystallize before the more stable ones. This is often referred to as Ostwald s law of stages [13]. As the system progresses from the liquid to the solid state, it will go to the state nearest in free energy to that of the liquid state. Frequently a metastable crystal form will appear and then convert to 

Crystal morphology Habit Prisms and plates elongated parallel Doubly terminated prisms 

Dispersion of optic axes V r (strong, crossed) V r (weak) 

Extinction Straight for sections perpendicular to Straight for prisms and 

The diuretic fmsemide illustrates how solvent recrystallization experiments can be used to discover new solid-state forms of older drugs. Fmsemide was marketed for many years with no reports of polymorphism. Doherty and York [14], however, noted some features of the differential scanning calorimetry (DSC) thermogram of fmsemide that suggested the possibility of polymorphs. They performed a number of solvent recrystallization experiments and discovered a new polymorph which they characterized analytically. At nearly the same time, Matsuda and Tatsumi [15] published the results of their experiments with fmsemide with the conclusion that there are five polymorphs and two solvates, illustrating the principle that the number of known forms is related to 

---

Usin mixed solvents for recrystallization - never assume that a mixed solvent system is a mixture of equal volumes of the two solvents.-The ratio of the two solvents is established practically during the recrystallization experiment. 

A few comments concerning the crystallization of carbanions are in order. These comments are based upon the personal experience developed in our own laboratory and also upon observations noted in the literature in the course of crystallizing enolate anions. Although alkali metal enolate anions are relatively unstable compounds, they have been prepared in the solid state, isolated, and characterized by IR and UV spectroscopy in the 1970s. Thus the ot-lithiated esters of a number of simple esters of isobutyric acid are prepared by metallation of the esters with lithium diisopropylamide in benzene or toluene solution. The soluble lithiated esters are quite stable at room temperature in aliphatic or aromatic hydrocarbon solvents and are crystallized out of solution at low temperature (e.g. -70 °C.). Alternatively the less soluble enolates tend to precipitate out of solution and are isolated by centrifugation and subsequent removal of the solvent. Recrystallization from a suitable solvent can then be attempted. The thermal stability of the lithiated ester enolates is dramatically decreased in the presence of a solvent with a donor atom such as tetrahydrofuran. 

Recrystallization experiments frequently yield crystals having different shapes and morphologies which are not necessarily different polymorphs. For example, Figs 8.2 and 8.3 show crystals of p-estradiol with distinctly different shapes but are, in fact, the same polymorph. The morphology differences are due to different crystallization solvents. It is important then to have some microscopical technique that allows one to distinguish between polymorphs. Optical crystallography, thermal microscopy and microspectroscopy have this ability. 

If a series of such experiments with pure solvents does not yield a satisfactory recrystallization solvent, the experiments above are repeated using solvent combinations. Solvents for this purpose must of course be... 

A setup similar to the preceding one is used in this experiment except that provision should be made for heating the reaction vessel (steam bath, oil bath, or mantle). Lithium aluminum hydride (10 g, 0.26 mole) is dissolved in 200 ml of dry -butyl ether and heated with stirring to 100°. A solution of 9.1 g (0.05 mole) of ra j-9-decalin-carboxylic acid (Chapter 16, Section I) in 100 ml of dry -butyl ether is added dropwise over about 30 minutes. The stirring and heating are continued for 4 days, after which the mixture is cooled and water is slowly added to decompose excess hydride. Dilute hydrochloric acid is added to dissolve the salts, and the ether layer is separated, washed with bicarbonate solution then water, and dried. The solvent is removed by distillation, and the residue is recrystallized from aqueous ethanol, mp 77-78°, yield 80-95 %. 

By the procedure described in the preceding experiment, 30 g (0.11 mole) of tri-phenylphosphine dissolved in 100 ml of acetonitrile is converted to triphenylphosphine dibromide. After the addition of the bromine has been completed, the cooling bath is removed, the flask is set up for vacuum distillation, and the solvent is removed. To the residue is added /7-chlorophenol (10.3 g, 0.08 mole), and the flask is heated at 200° (mantle, wax bath, or sand bath) until HBr ceases to be evolved (about 2 hours). The flask is cooled and the contents are steam distilled affording crude / -chlorobromo-benzene in about 90% yield. Recrystallization from benzene gives the pure product, mp 65-66°. 

The product is essentially pure. l,4-Disubstituted-5-amino-l,2,3-triazoles are readily isomerized 4 accordingly, care must be exercised in the recrystallization of such products from solvents. It has been found by experiment that the best practice, in order to avoid isomerization, is to heat the benzene to boiling before addition of the product for recrystallization. Repeated tests have shown that a single careless recrystallization of the product from benzene can increase the content of acidic isomer by as much as 4%. Polar solvents must be avoided. 

This procedure may be used for the preparation of a variety of a-amino ketones as is indicated in Table I, which summarizes most of the submitters experience with this reaction. Principal deviations from the procedure wifi be in the time required for a negative starch-iodide test and the nature and amount of extraction and recrystallization solvent. It is strongly recommended that any one using the reaction for the first time carry out the preparation on a-phenylethylamine before attempting to use it on other more valuable amines. 

One of the most common sources of contamination is the electrolyte since impurities in it would diffuse to the electrode and adhere to it during the course of the experiment. Impurities in the electrolyte can be reduced substantially by careful purification of solvent and solute. Distillation or ultrafiltration purifies water, the most common solvent. Usually solute materials can be bought in a very high purity, and whenever this is not the case, they can be cleaned by standard procedures such as recrystallization or calcination. Electrolysis of the electrolyte is also a common practice. Here, two sacrificial electrodes are immersed in the electrolyte and a potential is applied between them for about 36 hr in such a way that impurities are oxidized or reduced on their surfaces—the electrodes act as a garbage disposal thus the name of sacrificial electrodes. 

This allows one to prepare just double the amount of material in the same sized flask. The product obtained in this way, however, is slightly yellower than that obtained when more alcohol is used, but upon recrystallization it gives just as pure a product as that obtained by recrystallization of crude material made in the presence of more solvent. The results of many experiments lead to the conclusion that if large amounts of benzoin are to be prepared, the method described above is the better one. If, however, only a small amount is needed and a good grade of crude material is satisfactory, the larger amount of solvent is perhaps more desirable. 

The melting point of w-nitrocinnamic add is given in the literature as 195° and 196-197°. Beilstein describes it as yellow needles. Alcohol is mentioned as a good solvent for recrystallization but experience in this work showed that benzene is also satisfactory, a white w-nitrocinnamic acid being obtained from this solvent. 

Usual standard recrystallization conditions have been applied to the Preferential Enrichment experiment, except that highly supersaturated solutions are employed, because the supersolubility (a solubility obtained by dissolving the sample in a solvent on heating followed by being cooled) of the compounds showing Preferential Enrichment is considerably higher than that of the usual solubility at 25°C. 

It is essential that all the para isomer be precipitated, since it is not easily removed by recrystallization from solvents. The point at which the yellow />-nitrodimethylaniline ceases to precipitate and the orange w-nitrodimethylaniline begins to precipitate is difficult to judge without experience, since the solution itself has an orange color. If doubt exists, a small sample... 

---