1911 Encyclopædia Britannica/Stereo-isomerism

by Jacobus Henricus van 't Hoff

STEREO-ISOMERISM, or Stereomerism, a term introduced by Victor Meyer (by way of his denomination stereo-chemistry for “chemistry in space”) to denote those cases of isomerism, i.e. the difference of properties accompanying identity of molecular formulae, where we are forced to admit the same atomic linking and can only ascribe the existing difference to the different relative position of atoms in the molecule.

Historical.—Considerations concerning the relative position of atoms have been traced back as far as Swedenborg (1721); in more recent times the first proposal in this direction seems due to E. Paterno (1869), followed by Auguste Rosenstiehl and by Alexis Gaudin (1873). The step made by J. A. Le Bel and J. H. van't Hoff (1874) brought considerations of this kind in the reach of experimental test, and so led to " stereo- chemistry." The work of Louis Pasteur on molecular asymmetry in tartaric acid (i860) touched stereo-chemistry so nearly that, had structural chemistry been sufficiently developed then, stereo-chemistry might have originated fourteen years earlier; it happened, however, that Wislicenus's investigation of lactic acids (1869) immediately stimulated Van't Hoff's views. The fundamental conceptions, of Le Bel and Van't Hoff differ in that the former are based on Pasteur's notions of molecular asymmetry, the latter on structural chemistry, especially as developed by August Kekule for quadrivalent carbon. Both seem to lead to the same conclusions as to stereo-isomerism, but the latter has the advantage of allowing a more detailed insight, whereas the former, which is free from hypothetical conceptions, is of absolute reliability.

As our knowledge of stereo-isomerism originated in the chemistry of carbon compounds and found the largest development there, this part will be treated first.

Stereo-isomerism in Carbon Compounds.

1. The Asymmetric Carbon Atom.—Though stereo-chemistry is based on the notion of atoms, there is not the least danger that it may break down when newer notions about those atoms are intro- duced. Even admitting that they are of a compound nature, i.e. built up from smaller electrical particles or anything else and able to split up under given conditions, their average lapse of existencs is long enough to consider them as reliable building-stones of the molecule, though these building-stones may give way now and then, as our best ordinary ones by the action of an earthquake. Another thing which stereo-chemistry abstracts beforehand is the movement of atoms, which is generally accepted to exist, but becoming less as the temperature sinks and disappearing at absolute zero. And so the following symbols, representing atoms in a fixed position, may correspond to these last circumstances, whereas at ordinary temperatures atoms may vibrate, for instance, with these fixed positions as centres.

The first development from structural to stereo-chemistry was to consider the relative position of atoms in methane, CH₄. Structural chemistry had proved that the four atoms of hydrogen were linked to carbon and not to each other, thus

C	—H
	—H
	—H
	—H

and not for example H—H·C<H
H, but how the four were grouped remained to decide.

The decision is derived as follows:—

If the four hydrogen atoms are supposed to be in a plane on one side of the carbon atom as above, two methylchlorides CH₃Cl should be possible, viz. :—

C	—Cl	and	C	—H
	—H			—Cl
	—H			—H
	—H			—H

Such isomeric compounds have never been found, but they appear as soon asthe four atoms (or groups of atoms) to which carbon is combined are different, for example in CHFClBr, fiuorchlorbrom- methane. Then and only then two isomeric compounds have been regularly observed, and the sole notion about relative position of atoms in methane which explains this fact is that the four groups combined with carbon are placed at the summits of a tetrahedron whose centre is formed by carbon. The two possibilities are then represented by:—

Fig. 1.	Fig. 2.
Fig. 1.	Fig. 2.

These groupings have the character of enantiomorphism, i.e. they are non-identical mirror images. If any of the two differences in the summits is given up, for example, F substituted by Cl with the formation of CHCUBr, the enantiomorphism disappears.

The isomerism corresponding to this difference in relative position is the simplest case of stereo-isomerism. The carbon atom in the special condition described, linked to four different atoms or groups, is denominated “asymmetric carbon”, and will be denoted in the following formulae as C. Stereo-isomerism exists in tartaric acid, HO₂C·CH(OH)·CH(OH)·CO₂H (studied by Pasteur), in the lactic acid, CH₃·CH·OH·CO₂H (studied by Wislicenus), while the simplest case at present known is the chlorobromofluoracetic acid, C·Cl·Br·F·CO₂H, obtained by Schwartz. This stereo-isomerism, due to the presence of asymmetric carbon, is of a characteristic kind, which is in perfect accordance with the theory of its origin, being the most complete identity combined with the difference that exists between the left and right hand. All the properties which cannot differ in this last sense are identical, viz.: melting and boiling point, specific gravity, &c. But the crystalline form, which may show enantiomorphism, indeed shows this difference in the isomers in question; and especially the behaviour (in the amorphous state) towards polarized light differs in the sense that the plane of polarization is turned to the left by the one isomer, and exactly as much to the right by the other, so that they may be termed " optical antipodes." All these differences disappear with the asymmetric carbon, and the succinic acid, HO₂C·CH₂·CH₂·CO₂H, from tartaric acid is optically inactive and shows no stereo-isomerism.

2. Compounds with more than one Asymmetric Carbon Atom.—Stereo-isomerism and the space relation of atoms in compounds with higher asymmetry can best be developed by aid of graphic representations, founded on the notion of space relations in ethane, H₃C·CH₃. A consequence of the tetrahedral grouping in methane is the configuration given in fig. 3, where the six hydrogen atoms are substituted by six atoms or groups R₁, . . . R₆. The second (above) carbon atom is supposed to be at the top of the lower tetrahedron, and vice versa. Each other position, obtained by turning R₁R₂R₃ around the ·C—C· axis, is also possible, but since no isomerism due to this difference of relative position, which might already show itself in ethane, has been observed, we may admit that one of the positions obtained by the above rotation is the stable one, and fig. 3 may represent it. For simplicity’s sake this figure may be projected on a plane by moving R₃ and R₆ respectively upward and downward, with R₁R₂ and R₄R₅ as axes, which leads to the first of the four configurations representing the stereo-isomers possible in the above case. They differ in the two possible spatial arrangements of R₁R₂R₃ and R₄R₅R₆:—

Fig. 3.

As one asymmetric carbon introduces two stereo-isomers and two introduce four, n asymmetric carbon atoms will lead to 2ⁿ isomers. They are grouped in pairs presenting enantiomorphic figures in space, as do the first and the last of the above symbols, which correspond to the character of optical antipodes, whereas the first and second correspond to greater differences in melting points, &c. A well- studied example is offered by the dibromides of cinnamic acid, C₆H₅⋅CHBr⋅-CHBr⋅CO₂H. They have been obtained by Liebermann in two antipodes melting at 92°, and two other antipodes, differing in optical rotation from the first, and melting at 195°.

A simplification is introduced when the structural formula shows symmetry, as is the case in R₁R₂R₃C⋅CR₃R₃₂R₁. The four above- mentioned symbols then are reduced to three:—

of which the first and last show the enantiomorphism corresponding to the character of optical antipodes, while the second shows sym- metrv and corresponds to an inactive type. A well-studied example is offered here by tartaric acid: the two antipodes, often denoted as d and l, have been found, viz. in the ordinary dextrogyre form and the laevogyre form, prepared by Pasteur from racemic acid, while the third corresponds to mesotartaric acid; such internally compensated compounds are generally termed “meso.”

3. Cyclic Compounds.—Three or more carbon atoms may link together so as to produce ring systems such as

It is in these cases that the principle of the asymmetric carbon, which in the above case leads to 2³ = 8 stereo-isomers, is easily applied by means of graphical representations in a plane, derived from the space relation shown in fig. 4. The six groups, R₁ . . . R₆, are either under or above the plane in which the carbon ring is supposed to be situated, and this may be indicated by the following symbol:—

where the carbon atoms are supposed half-way between R₁ and R₂, R₃ and R₄, R₅ and R₆.

One of the most simple examples is offered by the trimethylene- dicarboxylic acids

for which three formulae can be deduced:—

the first, where the carboxyl groups ⋅CO₂H lie on the same side of the carbon ring, called, as Von Baeyer proposed, the cis-form, the others trans-forms. The trans-forms show enantiomorphism and correspond to optical antipodes, whereas the first symbol may be considered as corresponding to mesotartaric acid, symmetrical in configuration and inactive; this third stereo-isomer has also been met with.

Special attention has been given to those ring systems of the general form:—

This trans-form corresponds to a cis-form, where both R₂ and R₁ are on the same side of the plane containing the ring, These latter are enantiomorphic in the ordinary sense of the word, but the particular feature is that the trans-form, though offering nc plane of symmetry, is yet identical with its mirror image, and thus not enantiomorphic and not corresponding to optical antipodes but to the meso-form.

There correspondences have been realized by Emil Fischer in derivatives of alanine, H₃C⋅CH(NH₂)⋅CO₂H, which exists in two antipodes d and l. Two of these molecules can be combined to alanyl-alanine:H₃C⋅CH⋅NH(COC⋅H⋅NH₂⋅CH₃)⋅CO₂H, which, as containing two asymmetric carbons, may be had in four stereo- isomers dd, ll, dl and ld. In their anhydrides

we meet the above type, and find that dd and 11 formed the predicted antipodes, while the anhydride of dl and Id is one and the same substance, without any optical activity. Such cases are often termed "pseudo-asymmetric."

4. Isolation of Optical Antipodes.—The optical antipodes are often found as natural products, as is the case with the Ordinary or d-tartaric acid ; generally only one of the two forms appears, the second form (and, more generally both forms) being obtained synthetically. This is a problem of particular difficulty, since the artificial production of a compound with asymmetric carbon, from another which has no asymmetric carbon, always produces the two antipodes in equal quantity, and these antipodes, by their identity in most properties, e.g. melting and boiling point, solubility, and also on account of their analogous chemical behaviour, cannot be separated by customary methods, the application of which is rendered still more difficult by the formation of a so-called racemic compound.

The method called “spontaneous separation” was first observed by Pasteur with racemic acid, which in its double sodium and am- monium salt crystallized from its aqueous solution in two enantio- morphic forms, which could be separated on examination. One of the two proved to be the ordinary sodium-ammonium-tartrate, the other its laevogyre antipode; thus /-tartaric acid wasdiscovered, and racemic acid proved to be a combination of d- and /-tartaric acid. The further examination of this particular transformation showed that it had a definite temperature limit. Only below 27° is Pasteur's observation corroborated, while above 27° a racemate appears; these changes are due to a chemical action taking place at the given temperature between the solid salts:—

2C₄O₆H₄NaNH₄·4H₂O ⇄ (C₄O₆H₄NaNH₄)₂·2H₂O+6H₂O,

one molecule of the d- and one of the l-tartrate forming above 27°, the racemate with loss of water, while under 27° the opposite change occurs. This temperature limit, generally called transition-point, was discovered by Van’t Hoff and Van Deventer. It is the limit where the possibility of spontaneous separation begins, and is relatively rare, so that this way of separation is an exceptional one, most antipodes forming a racemic compound stable at all temperatures that come into question.

The use of optically active compounds in separating antipodes is of the greatest value. The general principle is that the compounds which the d- and l-form give with a different active compound, for instance d producing dd and ld, are by no means antipodes and so exhibit the ordinary differences, e.g. in solubility, which allow separation. It was in this way that Pasteur split up racemic acid by cinchonine. This method has since been applied to the most various acids; bases may be split in an analogous way; artificial conine was separated by Ladenburg by means of (d-tartaric acid, and one of these antipodes proved to be identical with natural conine. Aldehydes and ketones on the other hand may be split up by their combinations with an active hydrazine, &c, and so this method is by far the most fruitful.

The formation of a racemic compound built up from dd and ld has also been observed in the so-called partial racemate. An example is the racemate of strychnine. It is in this case also that the transition-point forms the limit of possible separation, determined by Ladenburg and G. Doctor to be 30°. Such partial racemic combination however occurs only in exceptional cases, else it would have invalidated this method, as it did spontaneous separation.

A different way of using active compounds in producing antipodes consists in the so-called asymmetric synthesis. The method consists in the introduction of an active complex before that of the asymmetric carbon ; both stereo-isomers need not then form in the same quantity. W. Marckwald and A. McKenzie, who chiefly worked out this method, found, for example, that the salt of methylethylmalonic acid, C(CH₃) (C₂H₅) (CO₂H)₂ , with the active brucine forms on heating the corresponding salt of d- and /-methylethylacetic acid C(CH 3 ) (C 2 H 5 )H(C0 2 H), with the /-antipode in slight excess.

5. Configuration of Stereo-isomers.—The conception of asymmetric carbon not only opens the possibility of determining when and how many stereo-isomers are to be expected, but also allows a deeper insight into the relative position of atoms in each of them. The chief indication here lies in the configuration of the meso-type, already given for mesotartaric acid ; the corresponding alcohol, the natural sugar erythrite, which produces this acid by oxidation, consequently corresponds to CH 2 OH

H—C—OH

H—C—OH

CH 2 OH. In the glutaric acids, H0 2 C- (CH-OH) 3 -CO.H, the structural symmetry again leads to meso-forms

OHOHOH OH H OH

HO,C-

-U-<

C—C0 2 H and H0 2 C

H H H

-cUU-

H OH H

-C0 2 H.

They are respectively obtained by the oxidation of ribose and natural xylose, stereo-isomers of the formula COH(CHOH) 3 CH 2 OH ; the latter produces active tartaric acid and so decides that the second formula is that of the corresponding trioxyglutaric acid, the first remaining for that obtained from ribose.

In such and analogous ways the configuration of meso-types may be fixed with absolute certainty. The decision is more difficult in the case of antipodes. For tartaric acid it is certain that the d- and l-forms correspond to

CO2H HCOH HOCH C0 2 H

C0 2 H

and

HOCH HCOH

A

2 H,

but which of the two represents the ordinary <?-acid is unknown. Emil Fischer proposed to decide provisionally in an arbitrary way and admit for the d- the first formula. Then we may conclude that the natural malic acid, which may be obtained by the reduction of /-tartaric acid, is C0 2 H

HO

h 2

OH

o 2 h,

while the natural xylose, which produces /-tartaric acid by the substitution of C0 2 H for CHO-CHOH, corresponds to

CHO

HCOH

HOCH

HCOH

The results obtained in these and analogous ways have proved to be of value in the study of enzymes, e.g. such complex organic substances as zymase in yeast, which is able to produce in small quantity an unproportioned large amount of chemical change, in this case the transformation of the sugar glucose, C6Hi 2 06, into alcohol and carbonic acid

C6H 12 06 = 2C 2 H 6 0-r-2C0 2 .

These enzymes have an extremely specific action, producing, for instance, the change in ordinary natural glucose, but not at all in its artificial antipode, and so they are often valuable means of isolating an antipode from the inactive mixtures or racemic compounds; this method has indeed been used for the isolation of the glucose- antipode from the artificial racemic form. The fundamental fact here is due once more to Pasteur, but Emil Fischer added that sugars are acted upon by zymase in an analogous way if their configuration shows a certain amount of identity. For example yeast acts on

d-Glucose HCO

HCOH

HOCH

HCOH

HCOH

H 2 COH

^Mannose HCO

HOCH

HOCH HCOH HCOH H 2 COH

(/-Fructose H 2 COH

Ao

HOCH HCOH HCOH H 2 COH,

t

H 2 OH.

and we observe that the three formulae agree indeed in the lower four-carbon chain. This particular behaviour led Fischer to the expression that the enzyme-action on given substances needs a corresponding feature as " lock and key." There are indications that in the synthesis by enzymes, of which examples have been realized in fats, sugars, glucosides and albuminoids, an analogous behaviour prevails.

6. Mutual Transformation of Antipodes.—Thus far we have supposed the molecule to be stable with atoms in fixed places, as may be the case at absolute zero ; in reality, at ordinary temperatures, atoms probably are endowed with movement, and this may be supposed to take place along the fixed places just mentioned as centres, which movement can go so far as to lead to total transformation, the one stereo-isomer changing over into the other. These cases may be considered now.

As a general rule the liquid, gaseous or dissolved antipode is it itself unstable, tending to be transformed into inactive complexes. Temperature may accelerate this, and, as a rule, sufficient heat will produce the loss of optical activity, half of the original compound having changed over into its optical antipode. This transformation has been often used for preparing the latter, as was first done by Le Bel with the optically active amyl alcohol, HC(CH₃)(C₂H₆)(CH₂OH), rendering it inactive by sufficient heating, and separating from the obtained complex the stereo-isomer. Walden found that in some cases analogous transformations take place at ordinary temperature, as for instance with d-phenylbromacetic acid, which within three years totally lost its considerable rotative power; this transformation has been termed "autoracemization." It explains that till now the most simple compounds with asymmetric carbon have not yet been obtained in antipodes; active CHClBrF might be obtained by treating chlorobromonuoracetic acid with potash, but autoracemization, which especially shows itself when halogens are linked to the asymmetric carbon, might, without special precautions, lead to an inactive mixture of antipodes.

When two asymmetric carbons are present, four stereo-isomers are possible, which may be represented by:—

(1) A+B, (2) −(A+B), (3) A−B, (4) −(A-B),

(1) and (2), as well as (3) and (4), being antipodes. The stable form will be in this case also the inactive mixture, corresponding in the solid state either to (1), (2) or (3), (4). In the last case, suppose the primitive compound is (1), the first step towards stability may be the production of (3), so that practically one stereo-isomer changes over into another of a different type. Such has, for instance, been proved by Bechmann for l--menthol,

H—C—C₃H₇

H₂C CH₂

H₂C CO

V

H—C—CH₃,

which on heating produces a form rotating in opposite sense, though not the antipode. Probably H and CH₃ in the lower asymmetric carbon have changed places. A further treatment at high temperature might prob- ably produce the inactive mixture of this menthol and its antipode. Fig. 5. 7. Doubly-Linked Carbon Atoms.—When carbon atoms are doubly linked, as in derivatives of ethylene, H₂C:CH₂, the two tetrahedra representing the four groups around each carbon may be supposed to have two summits combined, as was supposed with one in simple linking. Fig. 5 represents this supposition, from which follows that the six atoms in question are situated in a plane and may be represented by a plane figure:

R₁·C·R₂

| |

R₃·C·R₄.

The chief consequence is that as soon as the two atoms or groups attached to each carbon are different, two stereo-isomers may be looked for:—

Rj-C-R, R r CR 2

I! II

RrC-Rj R2CR1.

Such has been found to be the case, fumaric and maleic acids, H-C-COjH H-C-COsH

HOjC-C-H H-C-C0 2 H,

forming the oldest and one of the most simple examples ; the simplest is α-chlorpropylene (H₃C)HC:CClH.

The nature of this stereo-isomerism is quite different from that in antipodes. There is no enantiomorphism in the supposed configurations, and so no. rotatory power, &c, in the corresponding compounds, which, on the other hand, show differences far deeper than antipodes do, having different melting points, solubility, heat of formation, chemical properties, &c, behaving in these as ordinary isomers. These isomers, having some relation to those in cyclic compounds, may be also denoted as cw-(maleic) and trans-(fumaric) forms, a close analogy existing indeed in those ring systems of which the simplest type is:—

this has been realized in the 1, 3-tetramethylene dicarboxylic acids, which exist in a trans- and cis-form:—

When two double carbon Unkings are present, as in H 2 C:C:CH 2 , the four hydrogen atoms form the summits of a tetrahedron according to the development in fig. 4 ; and consequently the introduction of different groups may bring enantiomorphism and optical antipodes. This has been realized in the compound 1-methyl-cyclo-hexylidene-4-acetic acid (formula I.), first prepared by W. H. Perkin and W. J. Pope in 1908, and resolved into its components by fractional crystallization of its brucine salt by Perkin, Pope and Wallach. The substance resolved by W. Marckwald and R. Meth in 1906, which was regarded as this acid, was really the isomeric i-methyl-∆3-cyclo-hexene-4-acetic acid (formula II.), which contains asymmetric carbon atoms (see Journ. Chem. Soc., 1909, 95, p. 1791; cf. ibid., 1910, 97, p. 486).

H₃CHrCftv /H H 3 C

>C:C<

H 2 -ch/ \co 2 h, h

8. Numerical Value of Optical Rotation.—To express the value of optical rotation either specific or molecular rotation may be chosen, the first being the deviation caused by a layer of 1 decimetre in length when the substance in question is supposed to be present with specific gravity 1, the latter is this value multiplied by one-hundredth of the molecular weight. Specific rotation is indicated by [a] ! D , where the suffix indicates the wave-length of the light in question, D being that of the sodium line, and t the temperature; [M]j-, is the corresponding value of molecular rotation. Both values vary with the solvent used, and probably are most adapted to solve problems touching relations of rotatory power and configuration, when they apply to extreme dilution in the same liquid.

One of the most general rules, relating to rotatory power, is that for electrolytes, i.e. salts in aqueous solution, viz. the limiting rotation in dilute solution only depends on the active radicle. Oudemans found that for such active bases as quinine in its salts with hydrochloric, nitric, chloric, acetic, formic, sulphuric, oxalic, phosphoric, perchloric acids the specific rotation (calculated for the base) only varies from −272° to −288°; H. H. Landolt found the same thing for active acids, the mono lithium, sodium, potassium and ammonium tartrates varying only between 27·5° and 28·5° (calculated for the acid). A corresponding rule may be expected where both base and acid have rotatory power; the molecular rotation will be the sum of those for base and acid in salts with inactive radicles. Each of these rules finds sufficient explanation in Arrhenius's view of electrolytic dissociation, which admits that diluted electrolytes are split up in their ions, and so the salts of quinine (Q) owe their rotatory power to the ion QH, those of acid tartrates to the ion C₄H₅O₆, and quinine-tartrate to both.

With non-electrolytes relations are less evident. One general observation is that non-saturation, especially cyclic structure, augments rotatory power. The saturated compounds, hydrocarbons, alcohols, ethers, amines and acids rarely show specific rotations higher than 10°, and some of them, as mannite, CH₂OH(CHOH)₄CH₂OH, for instance, show such small values that only a more thorough investigation, due to the theoretical probability of rotatory powers in asymmetric natural products, has detected the optical activity.

Unsaturated compounds generally show larger rotative powers; amyl alcohol with −5° produces an aldehyde with 15 ; succinic (diamyl) ether with −9° produces fumaric ether with 15°, &c. Cyclic configuration especially leads to the highest values known: the lactic acid with 3 leads to a lactone with −86°,

H₃C-CH-COO

I I

OOC-HC-CH 3 ;

mannosaccharic acid, H0 2 C(CHOH)4CO 2 H, to a dilactone (with two rings, formed by the loss of two molecules of water) with 202°, whereas the original acid only shows a small rotation.

A second Conception, which connects rotation with configuration in non-electrolytes, is due to Alexander Crum Brown and P. A. Guye. It starts from the simple assumption that, as rotatory power is due to the difference of the four groups around the asymmetric carbon, so its amount may correspond to the amount in this. So, generally speaking, take some property, denoted by Ki, . . . K t respectively, a function :—

(K₁-KO (Ki-K.) (Id-KO (K.-K.) (K 2 -K 4 ) (K.-KO

would express what is wanted. It becomes zero when two groups are equal ; it changes its sign, retaining its value, when K₁ is interchanged with K 2 , &c. The chief difficulty in application is to point out that property which is here dominating. It has been supposed to be weight, and then the above expression divided by (K₁+K₂+K₃+K₄)⁶ might be proportional to specific rotation. This explains, for instance, that in the homologous series of glyceric ethers HO
 HCCH₂OH
CO₂R , augmenting the heaviest group, ·CO₂R, first augments the specific rotation, which then passes through a minimum (the theoretical limit being zero):—

Ether of methyl, ethyl, propyl, butyl, hexyl, octyl,

[α]_D＝−4·8°, −9·2°, −12·9°, −13·2°, −11·3°, −10·2.

But the serious objection is met that groups of equal weight and different structure often allow considerable rotatory power as in methyl acetylamygdalate, with −146°, though in the formula C₆H₅HC(OC₂H₃O)(CO₂CH₃) the third and fourth groups are of equal weight. It is in this way especially that other properties might be tested, such as volume or density, and perhaps qualities related to light, such as refractive power and the dielectric constant. Attempts to connect the rotatory power of a compound with more asymmetric carbons to the action of each of these separately, i.e. by the so-called optical superposition have not been very successful. In the four stereo-isorceric acids CO₂H(CHOH)₃CH₂OH of the following configurations

CO₂ H

HCOH HOCH HOCH

H 2 COH

i-arabonic acid,

C0 2 H

I HCOH

C0 2 H HOCH HCOH

CO2H HOCH HOCH

HCOH

H 2 COH

d-lyxonic

acid,

HOCH

H 2 COH

(i-xylonic

acid.

HCOH

HCOH

I H₂COH

d-ribonic acid,

We might suppose the upper asymmetric carbon to produce a rotation +A or −A, the other ± B and ± C. The rotations then were A−B−C, A + B+C, -A-B + C and −A + B−C or zero in total. This supposition is in so far related to that of Crum Brown and Guye that it admits that the smallest conceivable change, i.e. stereo-isomeric change, in one group does not influence the rotation caused by the asymmetric carbon attached to it. It has not been tested in this case, but substances as propyl- and isopropyl-glycerate only differ in specific rotation from −12·9° to −11·8°, and might prove identical in the same solvent ; the sharpest test might be afforded by propylisopropylacetic acid.

9. Steric Hindrance.—The difference in the relative positions of atoms not only explains the different behaviour of optical antipodes, as has been indicated, but also gives some indication where no optical activity is concerned.

In the stereo-isomerism of ethylene compounds, taking maleic and fumaric acid as examples, space relations chiefly indicate that in one of the two the carboxyl groups CO2H are nearer. Such seems indeed to characterize maleic acid. It easily gives an anhydride of the cyclic formula

HC—CO || \0 and, inversely, when cyclic

HC-CO

compounds such as benzene are broken down by oxidizing agents, it is maleic and not fumaric acid that appears. On the other hand the presence of the two negative carboxyls makes maleic acid the stronger acid but less stable, with a pronounced tendency to change over into fumaric acid; this goes hand in hand, according to a general rule, with smaller heat of formation, "lower melting point and increased solubility.

In the cyclic compounds analogous phenomena occur. The formation of lactones, i.e. cyclic anhydrides derived from oxy-acids by interaction of hydroxyl and carboxyl, presents one of them. In the oxy-acids of the fatty series a particular feature is that from the isomers, denoted as o, and y, &c. H0 2 C-CHOH(CH 2 )„CH 3 , H0 2 C-CH 2 -CHOH-fCH 2 )„. 1 CH 3 ,H0 2 0(CH«) 2 -CHOH(CH 2 ) n - 2 CH s ,&c. the 7-compounds most easily form a lactone, though in the o-series

carboxyl and hydroxyl run nearer. The tetrahedral arrangement, however, as shown in fig. 6, explains that A, one of the groups attached to the carbon atom Ci, is fairly near C&, one of the groups attached to the carbon atom C 4 (the angle A being 35 ) ; A would correspond to the hydroxyl forming part of carboxyl around Ci; Cs to the hydroxyl linked with the carbon atom in the -y-position.

A third consideration on analogous ground is that of " steric hindrance." It was introduced by Victor Meyer’s discovery that derivatives of benzoic acid, having two substituents (X and Y) in the immediate neighbourhood of carboxyl :—

COjH

are unable to form ethers in the ordinary way, by treating with methyl alcohol and hydrochloric acid, whereas the isomers having only one of the substituents Y in 4 (X in 6) readily do ; it was suggested that the presence of X and Y near C0 2 H prevented the access to the latter. This argument has not been completely established, but a large amount of quantitative corroboration has been brought together by N. A. Menschutkin, who has found that in alcohols the more the hydroxyl group is surrounded by substituents (for instance CH 8 ) the slower esterification (with acetic anhydride in acetone at 100 ) takes place, the ratio of rates being

Methyl alcohol H 3 COH ... 100

Ethyl alcohol H 3 C-CH 2 -OH . . 48

Dimethyl carbinol (H 3 C) 2 CH-OH . 14

Trimethyl carbinol (H₃C) 3 OOH . . 0·8

Stereo-isomerism in Other Elements.

Phenomena analogous to those observed in carbon compounds might also exist in derivatives of other quadrivalent elements ; and only the relative stability of carbon-compounds makes every form of isomer, which often is unstable, more easily obtainable in organic chemistry. Nevertheless it has been possible to obtain stereo-isomers with different elements, but, as expected from the above, especially in derivatives containing carbon. Some of them have the character of optical antipodes and are more easily considered from a theoretical point of view ; others have not.

1. Optically Active Stereo-isomers.—Most closely related to the phenomena with carbon are those with sulphur, selenium, tin and silicon, when these elements behave as quadrivalent. S. Smiles {Journ. Chem. Soc, 1900, 77, pp. 1072, 1174; 1905, 87, p. 450) split up such derivatives of methylethyl-thetine as C 2 H 6V XHrCO-CeH,,

X,

CH

^Br

obtained by condensing methylethyl sulphide with oi-bromacetophenone, by means of the salt with d-bromocamphosulphonic acid, into optical antipodes.

W. J. Pope and A. Neville (Journ. Chem. Soc, 1902, 18, p. 198) succeeded in the same way with a selenium compound CaHsv /CH 2 C0 2 H >Se< ;

H 3 (X ^Br W. J. Pope and S. J. Peachey (Journ. Chem. Soc, 1900, 16, pp. 42, 116) with a compound of tin (tin methylethylpropyl iodide) C 2 H S \ /C3H7

>Sn< H 3 (X M ; Kipping (Journ. Chem. Soc, 1904, 20, p. 15; 1907, 23, p. 9) with one of silicon (benzylethylpropyl silicol)

(~ 2 Hfiv /C3H7

CeHj-CH/ X>H.

These facts may be explained in the same way as with carbon, by admitting tetrahedral grouping. A special feature, however, wanting with carbon, is that compounds with one atom only of the element in question have been obtained as antipodes. A second observation of some interest is that the compounds in question are electrolytes and that, as in solutions, where they are split up into ions, activity must be due to the last, the ionic complex, for instance, RiR 2 R s S, must cause optical rotation.

Optical antipodes have also been obtained with quinquevalent nitrogen in compounds of the type: R;R 2 R 3 RiNR 6 . Le Bel observed these in methylethylpropyl-isobutylammonium chloride; since then Pope and Peachey and Wedekind studied the same question more thoroughly, and as a general result it is now stated that ammonium compounds with four different radicals behave as asymmetric carbon compounds. The explanation may be that the four radicals arrange themselves in the two possible tetrahedral configurations, and that the fifth element or group, e.g. chlorine or hydroxyl, more loosely linked, finds its fittest place, as shown in figs. 7 and 8.

Fig. 7.

Fig. 8.

2. Stereo-isomers^ Without Optical Activity.—The chief cases here belong to the derivatives of nitrogen with double linking and the metallic compounds which have been chiefly studied by Werner.

The nitrogen compounds showing stereo-isomerism belong to two classes, according to the structural formulae, containing C :N or N :N ; in their general behaviour they seem related to the ethylene derivatives.

The first group was detected by Victor Meyer and Goldschmidt in C 6 H 6 'C:NOH benzildioxime : I

C 6 H 6 -C:NOH. Later investigations, especially by Hantzsch, showed that a grouping Ri-C-R 2 R,CR 2

gives rise to stereo-isomerism, the supposed difference being that X is either more close to Ri or to R 2 . This peculiarity is observed in the aldoximes and ketoximes, derived from aldehydes and ketones on treatment with hydroxylamine, and the two simplest examples are ethyl-aldoxime H 3 C-CH:NOH, and phenyl-benzyl-ketoxime, (C 6 H 6 ) (C6H 4 CH 2 ) C:NOH. As the behaviour of these stereo-isomers much resembles that of ethylene-compounds, they are often indicated as cis- and trans-forms.

The second stereo-isomerism in nitrogen-compounds was detected by Schraube in potassium benzenediazotate, and may perhaps be reproduced by the following symbols:—

C₆H₅·N and C₆H₅N

KON NOK.

The last group of stereo-isomers, in which insight is most difficult yet, is that of Werner's complex metallic compounds, observed with cobalt, platinum and chromium. No enantiomorphous character throws light here, and there is no relation to ethylene derivatives.

With cobalt the fact is that in the hexammonic cobalt salts, e.g. Co(NH 3 ) 6 Cl 3 , when NH 3 C1 is substituted by NO₂ isomerism appears as soon as the number of substituents is two ; Jorgensen's flavo-salts Co(NHs)4(NO₂ ) 2 Cl, and Gibbs's isomeric croceo-salts offer examples. Werner puts forward that a grouping of (NH S ) 6 at the summits of a regular octahedron around Co may explain this. Platinum compounds such as (H 3 N) 2 PtCU have been obtained in two forms, Werner admitting here the following plane configura- tions : —

H.N' \NH 3 H 3 lN T/ X C1

Chromium shows a behaviour analogous to that of cobalt, and analogous space-formulae may be used here. But, in a general way, at present it is extremely difficult to decide upon their value.

Bibliography. — The standard authority is C. A. Bischoff and P. Walden, Handbuch der Stereochemie (1894), with the two supple- mentary volumes (by Bischoff) entitled Materialien der Stereochemie (1904), containing abstracts of papers up to 1902. A. W. Stewart, Stereochemistry (1908), is a comprehensive survey. The views of A. Hantzsch and A. Werner are developed in Hantzsch, Grundriss der Stereochemie (1904) , and Werner, Lehrbuch der Stereochemie (1904). Other works are: H. Landolt, Optische Drehungsvermogen (2nd ed., 1898) ; Eng. trans. Optical Activity and Chemical Composition (1900) ; J. H. van't Hoff, with a preface by J. Wislicenus, Die Lagerung der Atome im Raume (3rd ed., 1908), Eng. trans, (with an appendix by Werner dealing with stereo-isomerism among other elements) by A. Eiloart, entitled The Arrangement of Atoms in Space (1898); J. B. Cohen, Organic Chemistry (1907), pp. 56-171. Pamphlets, &c, dealing with special subjects, are W. van Ryn, Die Stereochemie des Stickstoffs (1897) ; E. Wedekind and Frohlich, Zur Stereochemie des funfwertigen Stickstoffs (1907); Jones, "The Stereochemistry of Nitrogen," Brit. Assoc. Rep. (1903); M. E. Scholtz, " Die optisch- aktiven Verbindungen des Schwefels, Selens, Zinns, Siliziums und Stickstoffs," Ahrens' Sammlung (1898, 1907); A. Ladenburg, " Uber Racemie," Ahrens' Sammlung (1903) ; Meyerhoffer, Gleichgewichte der Stereomeren (1907) ; lectures delivered by Walden and Werner in the German Chemical Society {Ber., 1905, 38, p. 345; 1907, 40, p. 15). For the stereo-isomerism exhibited by ammino compounds see Cobalt and Platinum; also Werner, Ann., 1910, 375, p. 1. Recent progress is reported in The Annual Reports of the Chemical Society (annual since 1904). (J. H. van't H.)