The Theory of Types
Berzelius seized upon the notion that radicals could be the units of which organic molecules were built. He believed that organic molecules were built of radicals as inorganic molecules were built of individual atoms. Radicals, he came to think, were almost as indivisible and untouchable as the individual atoms themselves.
Berzelius maintained that the force holding atoms together in an inorganic molecule or in an organic radical was electrical in nature (which eventually turn out to be right). Every molecule, then, had to contain a positive portion and a negative portion, since only between oppositely charged elements was there attraction.
For simple inorganic substances, like sodium chloride, this notion of positive and negative was eventually shown to fit the facts. To make it fit organic substances, Berzelius had to insist that radicals consisted of carbon and hydrogen only, with carbon negative and hydrogen positive. He held that the benzoyl radical (C7H5O) did not and could not contain oxygen, which distorted the work done with that radical. Berzelius was also certain that it was impossible to substitute a negative element for a positive element without drastically changing the properties of a compound.
He was quickly shown to be wrong in that last contention. Dumas was an enthusiastic supporter of Berzelius, but one of Dumas's pupils, Auguste Laurent (1807-1853) managed, in 1836, to substitute chlorine atoms for several of the hydrogen atoms in the molecule of ethyl alcohol. This experiment delivered the fatal blow to Berzelius's views, for chlorine was considered negative and hydrogen positive, yet one could be substituted for the other without making a drastic change in the properties of a compound.
Furthermore, in this chlorinated compound, carbon must be attached directly to chlorine, and if both consisted of negative atoms, how could that be? Negative electric charges repelled each other. (For that matter, how could two chlorine atoms hold together to form a chlorine molecule? Such matters were not settled for another century).
Berzelius, grown testy and extremely conservative in his old age, refused to change his notions. Upon hearing of Laurent's report, he attached the new findings ferociously. Dumas had himself, in 1839, substituted chlorine for three of the hydrogen atoms in acetic acid. Nevertheless, in the face of Berzelius's displeasure, Dumas rather cravenly backed down and disowned Laurent.
Laurent held firm, however, and continued to accumulate evidence to the effect that radicals were not as indestructible and untouchable as Berzelius insisted, and that one must not overemphasize the matter of positive and negative. Berzelius's anger barred Laurent from the more famous laboratories, and while Berzelius lived, his version of the radical theory remained in being by the sheer force of his personality. With Berzelius's death in 1848, however, his theory died and Laurent's gained popularity.
Laurent abandoned all emphasis on electrical forces. He believed that an organic molecule had a nucleus (which might be a single atom) to which different radicals might be attached. Organic molecules might then be grouped into families or types (the theory of types). All the members of one type would have an identical nucleus to which any of a series of similar radicals could be attached; and within the radicals there would be considerable room for variation.
A particular molecular type might even extend into the realm of the inorganic. For instance, the water molecule (H2O) may be viewed as consisting of a central oxygen atom (the nucleus) to which two hydrogen atoms are attached. If, in place of one hydrogen atom, any of a series of radicals is substituted, a type of compound is built up that includes among its members water as well as various organic molecules.
If one substituted for the hydrogen atom a methyl group (CH3) or an ethyl group (C2H5), one would have CH3OH (methyl alcohol), and C2H5OH (ethyl alcohol), respectively. A vast number of other alcohols could be built up in the same way. And, indeed, alcohols not only have many similarities among themselves, but as a class, also show certain resemblances to water. The simpler alcohols, such as methyl alcohol and ethyl alcohol, will mix with water in all proportions. Sodium metal will react with alcohols as it will with water, though more slowly.
Between 1850 and 1852, the English chemist Alexander William Williamson (1824-1904) showed that the family of organic compounds called ethers could also be built up about the "water type". In their case both the hydrogens of water were substituted by organic radicals. The common ether, then beginning to be used as an anesthetic, has both hydrogens replaced by ethyl groups, so that it is C2H5OC2H5.
Earlier, in 1848, the French chemist Charles Adolphe Wurtz (1817-1884) had studied a group of compounds related to ammonia and called, amines. He showed they belonged to a type with a nitrogen nucleus. In ammonia a nitrogen atom was bound to three hydrogens. In amines organic radicals replaced one or more of these hydrogens.
The theory of types gained in popularity because it could be used to organize the rapidly proliferating numbers of organic compounds being studied. The Russo-German chemist Friedrich Konrad Beilstein (1838-1906) published a vast compendium of organic compounds in 1880, and utilized Laurent's theory of types to organize those compounds into a rational order.
Nevertheless, the theory, as it emerged from Laurent's work, remained incomplete. It still made use of radicals as units, and the question of molecular structure was evaded rather than answered. For a proper answer, one had to face up to the question: What is the actual atomic arrangement within the radicals themselves?
The theory of types impressed some chemists with the point that the oxygen atom consistently combined with two other atoms or radicals. It might combine with two hydrogen atoms to form water, with one hydrogen atom and an organic radical to form an alcohol, or with two organic radicals to form an ether. But always the oxygen atom attached itself to two entities.
In similar fashion, the nitrogen atom would always combine with three atoms or radicals. Men like Kolbe took to writing formulas for organic compounds in which such a constancy in the number of attachments to oxygen or nitrogen was taken for granted.
The point was made general by an English chemist, Edward Frankland (1825-1899). He was the first to become interested in organometallics compounds, those in which organic groupings were attached to atoms of metals such as zinc. (In the true organometallics compound, the atom of the metal is firmly attached to a carbon atom. Compounds such as zinc acetate (a type of substance known prior to Frankland's time) are salts of organic acids. In such salts the atom of the metal is attached to an oxygen atom, and these are not considered true organometallic compounds.) Here it was quite clear that each metallic atom would attach to itself only so many organic groupings, and that this number was different for different metals. Zinc atoms, for instance, would combine with two organic groupings, neither more nor less.
In 1852, Frankland advanced what later came to be known as the theory of valence (from a Latin word for "power"), which is the statement that each atom has a fixed combining power. For instance, a hydrogen atom, under ordinary conditions, will combine with only one other atom. This is also true of sodium, chlorine, silver, bromine, and potassium. All have a valence (oxidation number) of 1.
Oxygen atoms may combine with as many as two different atoms, as will calcium, sulfur, magnesium, and barium. All these elements have a valence (oxidation number) of 2. Nitrogen, phosphorus, aluminum, and gold have a valence (oxidation number) of 3. Iron could have a valence (oxidation number) of either 2 or 3, and so on. The matter of valence turned out, in the long run, to be nothing like as simple as it seemed at first. Nevertheless, even the simple form of the theory proved to be of inestimable worth.
For one thing, the concept of valence helped to clarify the difference between atomic weight and equivalent weight of an element. Even as late as the mid-nineteenth century, many chemists confused the two. (As they did valence and oxidation number).
It can be determined that 1 part of hydrogen will combine with 35.5 parts of chlorine, since 1 atom of hydroge will combine with 1 atom of chlorine to form hydrogen chloride, and the chlorine atom is 35.5 times as heavy as the hydrogen atom. That is, chlorine has an atomic weight of 35.5. But 1 part of hydrogen will not combine with all elements in proportion to their atomic weights. For instance, oxygen has an atomic weight of 16, but each oxygen atom combines with two hydrogen atoms, since oxygen has a valence (oxidation number) of 2. Therefore, 16 parts of oxygen combine with 2 parts of hydrogen. The equivalent weight of oxygen is the quantity of oxygen that combines with 1 part of hydrogen, and that is 16/2 or 8.
In the same way, the nitrogen atom, with an atomic weight of 14 and a valence (oxidation number) of 3, combines with 3 hydrogen atoms. The equivalent weight of nitrogen is therefore 14/3 or about 4.7.
In general, the equivalent weight of an atom is equal to its atomic weight divided by its valence (oxidation number).
Again, Faraday's second law of electrolysis states that the weight of different metals liberated by a given quantity of electric current is proportional to the equivalent weight of those metals. This means that a given amount of electric current will liberate only half as much by weight of a 2-valent (oxidation number) metal as it would of a 1-valent (oxidation number) metal of about equal atomic weight.
This situation can be explained by supposing that one "atom of electricity" is required to transport a single 1-valent (oxidation number) atom, while two are required for a single 2-valent (oxidation number) atom. This connection of valence (oxidation number) and "atoms of electricity: was not fully appreciated for another half-century.
(Before we go any further it should be noted that that the term "valence" and "oxidation number" are often interchanged inappropriately. "valence" as was later learned and defined refers to the number of electrons located in an outer orbital of an element. This was not known and distinguished until much later. "oxidation number" refers to the "charge" that an element or function group possesses. While there is a relationship between "valence" and "oxidation number" they are to be recognized as different entities. Remain aware of this whenever either of these terms are presented to you.)
The notion of valence was applied with particular force to the structure of organic molecules by Kekule. He began with the suggestion that carbon had a valence of 4 and proceeded, in 1858, to work out the structure of simpler organic molecules and radicals on that basis. The concept could be visualized after a Scottish chemist, Archibald Scott Couper (1831-1892), suggested that these combining forces between atoms (bonds, as they are usually called) be pictured in the form of small dashes. In this way, organic molecules could be built up like so many "Tinker-toy" structures.
Indeed, this representation made it possible to visualize quite clearly why organic molecules were so much larger and more complex, on the whole, than inorganic molecules. According to the Kekule concept, carbon atoms could attach themselves to each other by means of one or more of their four valence bonds to form long chains, either straight or branched. No other kind of atom seemed to have that ability in nearly as marked a fashion as carbon did.
Thus, the tree simplest hydrocarbons (molecules made up of carbon and hydrogen atoms only), which are methane (CH4), ethane (C2H6), and propane (C3H8), could be pictured with every carbon atom possessing four bonds and every hydrogen atom possessing one, as follows:
This series could be continued by stringing together carbon atoms for almost as long as one would care to. By adding oxygen with two bonds and nitrogen with three, one could represent the molecule of ethyl alcohol (C2H6O), and methylamine (CH5N) as follows:
Such structural formulas could be made more flexible if the existence of two bonds (a double bond) or three (a triple bond) between adjacent atoms were permitted. Thus, ethylene (C2H4), acetylene (C2H2), methyl cyanide (C3H3N), acetone (C3H6O), and acetic acid (C2H4O2) could be represented as follows:
Structural formulas showed such obvious usefulness that a number of organic chemists accepted them at once. They completely outmoded all attempts to depict organic molecules as structures built up of radicals. Nothing less than an atom-by-atom picture would do now.
In particular, a Russian chemist, Alexander Mikhalovich Butlerov (1828-1886), supported the new system. During the 1860's, he pointed out how the use of structural formulas could explain the existence of isomers. For instance, to use a very simple case, ethyl alcohol and dimethyl either, although possessing widely different properties, have the same empirical formula: C2H6O. The structural formulas of the two compounds are:
|Ethyl Alcohol||Dimethyl Ether|
It is no wonder that the change in arrangement of atoms leads to two sets of widely different properties. In the case of ethyl alcohol, one of the six hydrogen atoms is attached to an oxygen atom, while in dimethyl ether all six are attached to carbon atoms. The oxygen atom holds the hydrogen atom more weakly than the carbon atom does, so that sodium metal added to ethyl alcohol replaces just one-sixth of the hydrogen content. If sodium is added to dimethyl ether, it displaces no hydrogen at all. Thus, chemical reactions serve as guides to structural formulas, and the formulas in turn serve as guides to understanding reactions.
Butlerov dealt specifically with a type of isomerism called tautomerism, in which certain substances always appeared as mixtures of two compounds. If one of these compounds were isolated in pure form, it would promptly change over, in part, to the other. Butlerov showed that tautomerism consisted of a spontaneous shift of a hydrogen atom from a connection with an oxygen atom to a connection with a nearby carbon atom (and back again).
A major problem in the first few years of the structural formula involved benzene, a simple hydrocarbon with the empirical formula C6H6. No structural formula seemed to satisfy the valence requirements and at the same time to account for the great stability of the compound. That is, the structural formulas that were first suggested resembled those of other compounds which were very unstable.
Again it was Kekule to the rescue. One day in 1865 (according to Kekule himself), while in a semi-doze on a bus, it seemed to him that he saw atoms whirling in a dance. Suddenly, the tail-end of one chain attached itself to the head-end and formed a spinning ring. Until then structural formulas had been built up only of chains of carbon atoms, but now Kekule fastened on the notion of rings of carbon atoms as well. He suggested the following structural formula for benzene:
This explanation was accepted, and the concept of the structural formula was placed on a firmer bases than ever. (Nevertheless, the presence of three double bonds in benzene created a problem, for compounds with double bonds usually underwent certain types of reactions which benzene did not ordinarily undergo. It was nearly three-quarters of a century before the puzzle of the double bonds that didn't act like double bonds was explained).
Despite the usefulness of the structural formulas of Kekule, they did not entirely account for one particularly subtle type of isomerism. This involved light, which we must therefore briefly consider.
In 1801, Thomas Young (1773-1829), an extraordinary Englishman who was the first to understand the physiology of the eye, had conducted experiments which demonstrated that light behaved as though it consisted of tiny waves. Then, about 1814, Augustin Jean Fresnel, a French physicist (1788-1827), showed that the light waves belonged to the particular class known as transverse waves. These waves oscillate at right angles to the direction in which the wave as a whole is traveling. This situation is best visualized in connection with water waves, which are transverse in nature. Individual bits of water move up and down, but the wave itself moves outward.
Light waves are not confined to a surface and so need not merely move up and down. They can move right and left, or northeast and southwest, or northwest and southeast. In fact, there is an infinite number of directions in which a light wave can oscillate at right angles to its direction of travel. In a beam of ordinary light some waves are oscillating in one direction, some in another, some in still another. There is no one direction that is preferred.
If such a beam of light is sent through certain crystals, however, the orderly arrangement of atoms within the crystals forces the light beam to oscillate in some particular plane - a plane that will allow the light to slip past and between rows of atoms.
Light oscillating in one plane only is called polarized light. This name was given it in 1808 by a French physicist, Etienne Louis Malus (1775-1812). At that time, the wave theory had not yet been accepted, and Malus had a notion that light consisted of particles with north and south poles, and that in polarized light, all the poles were lined in the same direction. This theory quickly vanished, but the expression remained and is still used.
The properties and behavior of polarized light seemed to lie exclusively in the province of the physicist until 1815. In that year a French physicist, Jean Baptiste Biot (1774-1862), showed that if polarized light passed through certain crystals, the plane in which the waves undulated was rotated. Sometimes it was rotated in clockwise fashion (dextrorotation), sometimes in counterclockwise fashion (levorotation).
Among the crystals displaying this property of optical activity were those of organic compounds. Furthermore, some of these organic compounds, such as the various sugars, showed optical activity even when not in crystalline form, but in solution instead.
As it eventually turned out, there were substances that differed only in their optical properties. Otherwise identical, one substance would rotate the plane of polarized light clockwise, the other would rotate it counterclockwise. Sometimes still a third would not rotate the plane at all. The isomers racemic acid and tartaric acid, which Berzelius had discovered, differed in optical properties.
Such optical isomers were not readily explained by Kekule's structural formulas.
The first glimmer of understanding of optical activity appeared in 1848, when the French chemist Louis Pasteur (1822-1895) began work with crystals of sodium ammonium tartrate.
Pasteur noted that the crystals were asymmetric; that is, one side of a crystal had a small facet not present on the other. In some crystals the facet was present on the right side, in others on the left. Using a magnifying glass, he painstakingly separated the crystals with tweezers and dissolved each group separately. The properties of the solutions seemed identical but for the optical activity. One solution was dextrorotatory, the other levorotatory.
In seemed, then, that optical activity was the result of asymmetry. It seemed also that whether the plane of polarized light was twisted in one direction or another depended on whether otherwise identical crystals had a "right-handed" asymmetry or a "left-handed" one.
This theory was satisfactory when applied to crystals, but what about optical activity that persisted in solution? In solution substances existed not as crystals but as individual molecules floating about randomly. If optical activity had to imply asymmetry, then the asymmetry had to exist in the molecular structure itself.
The Kekule structural formulas did not show the necessary asymmetry, but this lack did not necessarily disprove the connection between asymmetry and optical activity. After all, the Kekule structural formulas were written two-dimensionally on the flat surface of a blackboard or a piece of paper. Surely, it was not to be expected that in reality organic molecules were two-dimensional.
It seemed certain that the atoms in a molecule must be distributed three-dimensionally. If they were, their arrangement then might well show the necessary asymmetry to account for optical activity. However, how as one to go about applying the necessary three-dimensionality to the molecule?
Atoms had never been seen and their very existence might simply be a convenient fiction used to explain chemical reactions. Was it safe to take their existence so literally that one should distribute them in three dimensions?
A young man was needed to take the next step, one who had not yet gained the wise caution that comes with years.
Molecules in Three Dimensions
Such a person was the young Dutch chemist Jacobus Hendricus Van't Hoff (1852-1911). In 1874 he had not yet completed work for his Ph.D., but he daringly suggested that the four bonds of the carbon atom were distributed in three-dimensional space toward the four apices of a tetrahedron.
To see this, imagine that three of the bonds of the carbon atom are arranged so as to resemble the legs of a tripod, while the fourth bond sticks directly upward. Each bond is then equidistant from the remaining three, and the angle between one bond and any of its neighbors is about 109o.
The four bonds of the carbon atom are thus arranged symmetrically about the atom, and asymmetry is introduced only when each of the four bonds is attached to a different kind of atom or group of atoms. Then the four attachments can be arranged in exactly two different fashions, one being the mirror image of the other. This pattern provides exactly the type of asymmetry Pasteur had found in his crystals.
Almost simultaneously, the French chemist Joseph Achille Le Bel (1847-1930) published a similar suggestion. The tetrahedral carbon atom is sometimes referred to as the Van't Hoff-Le Bel theory.
The tetrahedral atom explained so much so neatly that it was quickly accepted. Aiding in this was a book published in 1887 by the German chemist Johannes Adolf Wislicenus (1835-1902), which placed the authority of an older and particularly well-respected scientist behind the theory.
Most important of all, there was no blinking the facts. Compounds possessing asymmetric carbon atoms (those connected to four different types of groupings) possessed optical activity, while those that did not possess such atoms did not. Furthermore, the number of optical isomers was always equal to the number predicted by the Van't Hoff-Le Bel theory.
In the final decades of the nineteenth century the three-dimensional view of bonding was extended beyond the carbon atom.
The German chemist Viktor Meyer (1848-1897) showed that the bonds of nitrogen atoms, if viewed three-dimensionally, could also explain certain types of optical isomerism. The English chemist William Jackson Pope (1870-1939) showed this was applicable to other atoms, such as those of sulfur, selenium, and tin; the German-Swiss Alfred Werner (1866-1919) added cobalt, chromium, rhodium, and other metals.
(Beginning in 1891, Werner developed a coordination theory of molecular structure, the idea for which, according to him, came to him in his sleep, waking him at 2 A.M. with a start. Essentially, this theory held that the structural relationships between atoms did not need to be restricted to ordinary valence bonds. Instead, particularly in certain comparatively complex inorganic molecules, atom groups could be distributed about some central atom in accordance with certain geometric principles that did not seem to take ordinary valence into account. It was nearly half a century before notions of valence became subtle enough to include both the simple compounds fitting the notions of Frankland and Kekule and the coordination compounds of Werner as well.)
The idea of three-dimensional structure led quickly to further experiments. Viktor Meyer had shown that while atom groupings ordinarily could rotate freely about a single bond attaching them to the rest of the molecule, the bulk of nearby groups of atoms sometimes prevented this rotation. This situation, called steric hindrance, can be likened to a door that ordinarily moves freely in its hinges but may be blocked by some obstruction behind it. Pope went on to show that through steric hindrance it was possible for a molecule to be asymmetric as a whole. It would then show optical activity even though none of the constituent atoms were asymmetric in themselves.
The German chemist Johann Friedrich Wilhelm Adolf von Baeyer (1835-1917) used the three-dimensional view, in 1885, to picture carbon atoms arranged in planar rings. If the four bonds of the carbon atoms are pointed toward the four corners of a tetrahedron, the angle between any two of them is about 109.5o. Baeyer argued that in any organic compound there was a tendency to allow the carbon atoms to be so connected that the bonds remained at their natural angles. If the angle is force to change, the atom is placed under a strain.
If three carbon atoms were bound in a ring, they would form an equilateral triangle, with the angle between each pair of bonds equal to 60o. This separation is considerably different from the natural 109.5o angle, and for this reason 3-carbon rings are hard to form and, once formed, are easy to break.
A 4-carbon ring would form a square, with the angles 90o; a 5-carbon ring would form a pentagon with angles 108o; a 6-carbon ring would form a hexagon with angles 120o. It would seem then that a 5-carbon ring involves virtually no strain on the bonds of the carbon atom, and a 6-carbon ring involves only a small amount of strain. Baeyer's strain theory seemed to account for the preponderance of such rings in nature over rings of more than six or less than five atoms. (Baeyer's strain theory applies to rings with atoms in a single plane. It is not necessary for the atoms to be in a single plane, and all sorts of odd rings can be (and are) formed in which this restriction does not hold.)
Most dramatic of all was the work, in the 1880's, of the German chemist Emil Fischer (1852-1919) on the chemistry of the simple sugars. A number of well-known sugars share the identical empirical formula of C6H12O6. They also have many properties in common, but differ in others, notably in the extent of their optical activity.
Fischer showed that each sugar had four asymmetric carbon atoms, and that on the basis of the Van't Hoff-Le Bel theory, there should therefore be sixteen optical isomers. These isomers should be arranged in eight pairs; in each pair on isomer should rotate the plane of polarized light clockwise and to exactly the extent the other isomer rotates it counterclockwise.
Fischer proceeded to work out the exact arrangement of the atoms in each of the sixteen isomers. The fact that exactly sixteen isomers of the 6-carbon sugars have been found, divided into eight pairs, is strong evidence for the worth of the Van't Hoff-Le Bel theory. This same accuracy in prediction holds in the case of other types of sugars, of amino acids, and of any other types of compounds.
By 1900, the depiction of molecular structure in three dimensions, having well proved its value, was universally accepted.