Polyfunctionality
Chapter 8 detailed Kolbe's route to his carbonic acid theory. There were several essential novelties in this theory, as compared to his earlier beliefs; he now fully accepted carbon tetravalence for most, but
not all, organic compounds. He had also adopted a substitutionist viewpoint and given up all traces of the copula theory. The research of Wurtz and Debus in 1857 and 1858 had convinced him of the dibasicity of oxalic acid, hence the need to double his formula for it. Furthermore, by extension, he now accepted the generic category of polybasic acids.
Polybasic acids are examples of polyfunctional organic compounds. For years, Kolbe had resisted a direct theoretical confrontation with this kind of compound, and for good reason. His chemical instinct, developed to maturity in the electrochemical tradition, was to identify a single, central carbonaceous focus for a compound and to use that atom or radical as the theoretical centerpiece of the formula. Hetero-atoms, especially oxygen, were grouped together as much as possible. Polyfunctionality was possible to formulate in this style, and Kolbe often did this before 1858, but only when all but one functional group remained in the background and hence were formulable as substituents within the radicals attached to the carbon group at the focal point. To relinquish this viewpoint would be to accept the structuralists' (Gerhardtian) thesis of the chemical equality of all carbon atoms in the molecule, thus to relinquish the last vestige of dualism. To be sure, by the time of Gerhardt's death Kolbe saw much of value in the Frenchman's system. Where the type theorists had gone seriously wrong, he thought, was in their subsequent development of multiple and mixed types. As it happens, multiple and mixed types were a result of the typists' struggle with polyfunctionality.
Wurtz in particular was generating phenomenal numbers of novel polyfunctional organic amines, alcohols, acids, and aldehydes in the years from 1855 to 1861. Chief among these were the two- and three-carbon organic acids and alcohols. Wurtz succeeded in oxidizing glycol to glycolic acid and then to oxalic acid. Since glycol had been prepared from ethylene, Wurtz argued that oxalic acid must have as much carbon as ethylene; moreover, glycol was seen to be as much the alcohol of glycolic acid as of oxalic acid, and glycolic acid was formulated as "diatomic" and dibasic (which is to say that both hydroxyl and carboxyl hydrogen atoms were acid in character). In modern terms,[10]
|
By analogous reactions, propylene was converted to propylene glycol and then to lactic acid, and analogous claims could be made for these as well:[11]
|
Wurtz then found that phosphorus pentachloride could be used to replace both hydroxyl groups of lactic acid with chlorine atoms, to make "[chloro]lactyl chloride," and then by reaction with alcohol, "chloro-lactic [ethyl] ester."[12] The existence of these compounds underlined for Wurtz the dibasic character of lactic acid and also its close relationship to propylene glycol, which underwent an analogous reaction with phosphorus pentachloride.
In the meantime, Heinrich Debus, a Bunsen protégé then at Queen-wood College, had developed a means of oxidizing alcohol to produce many of these same polyfunctional two-carbon compounds, but also including glyoxal (the dialdehyde) and glyoxylic acid (the aldehyde-acid). Kekulé and his student R. Hoffmann demonstrated how to make monochloroacetic acid, from which could be derived glycolic acid and glycocoll (glycine or aminoacetic acid). Strecker showed how to oxidize glycocoll to glycolic acid, and alanine (aminopropionic acid) to lactic acid. There seemed to be clear genetic relationships among all these compounds, and all could be formulated using multiple and mixed types.[13]
Kolbe needed to respond to all of these novel reactions by providing interpretations consistent with his own ideas. Chapter 8 showed why Kolbe formulated glycol as a hydrated oxide,
rejecting Wurtz' claim that it is an alcohol. For Kolbe, the lone hydrogen in the brace meant that glycol cannot be oxidized to an acid since such oxidation requires a minimum of two hydrogen atoms attached to the carbonyl group. It could, however, be oxidized to a two-carbon homolog of glycerin, that is, replacing the lone hydrogen by a third oxygen function. In Kolbe's formulation, this would be 3HO.C2 H3 .C2 O3 , a triple oxide hydrate. That this reaction had not yet been accomplished was not a problem for Kolbe.[14]
Wurtz had actually done what Kolbe considered impossible, in oxidizing ethylene glycol to glycolic acid, and propylene glycol to lactic acid. Kolbe responded that rearrangements must take place: in oxidizing glycol, one hydrogen of the methyl was replaced by a "hydrogen peroxide" radical (O2 H, i.e., hydroxyl) so that the other carbon now had room for full oxidation to the acid. Consequently, there must be a
yet undiscovered alcohol whose oxidation yields glycolic acid without rearrangement, the "true" glycol or double alcohol, isomeric with Wurtz' compound and unrelated to ethylene. Milder oxidation of this hypothetical substance should also yield a new aldehyde isomeric with acetic acid. Finally, if Wurtz' compound, ethylene oxide hydrate, could be dehydrated, it should yield ordinary aldehyde. In short, Kolbe affirmed, ethylene glycol and propylene glycol are not alcohols at all and have no substantive genetic relationship to glycolic or lactic acids. The latter are derivatives of acetic and propionic acids, as Kekulé, Hoffmann, and Strecker had shown. To be sure, glycolic and lactic acids each possess a radical (O2 H) substituted for hydrogen, but the acids are monobasic, not dibasic as Wurtz had asserted.
Kolbe made all these claims in fascicle 8/9 of his textbook, written in the summer or fall of 1858 and published early in 1859. The driving force of his resurgent research program begun late in 1858 was the examination and substantiation of these ideas, which he regarded as a direct outgrowth of his carbonic acid theory. That this theory was generating interpretations and predictions different from those of the type theorists validated Kolbe's sense that he had developed a powerful theory different from and superior to the school he so heartily despised—to what would soon be known as structure theory.
As we have seen, there were indeed some distinctive aspects of Kolbe's approach. Kolbe took formulation far more seriously and literally than did the type theorists, believing that discerning the constitutions of molecules was a straightforward process of applying rigorous deduction to skillfully gathered hard evidence, and he was convinced that one could arrive at ultimate formulas in this fashion. He accepted carbon tetravalence ("tetratomicity") for most organic compounds, but he could not countenance carbon chain formation. Kolbe's proximate radicals (such as methyl and carboxyl in acetic acid, or methyl, hydroxymethylene, carbon dioxide, oxygen, and water in lactic acid) were combined with one another as discrete molecular units and presumably coulombically. In formulating any compound, the goal was to identify a single governing radical, which he named the fundamental radical that held as many other proximate radicals together as equaled its combining capacity.
His approach made it difficult for him to deal with polyfunctionality in organic compounds, especially when the functional groups in a single molecule were not all the same. Once he was compelled to confront polyfunctionality by the work of Wurtz, Debus, Kekulé, and others, his theory became much more powerful. But even thereafter, his discomfort with the phenomenon is revealed in many subtle ways. This discomfort was the ultimate source of his strong disagreement with
Wurtz over the nature of glycolic and lactic acids; rather than alcohol-acids, Kolbe was convinced that they were ordinary monofunctional acetic and propionic acids, merely with a substituent O2 H group. Kolbe even admitted publicly that it had been difficult for him to reach the conclusion that O2 H could replace H in organic compounds.[15]
Despite these differences, Wurtz' and Kolbe's chemical ideas were substantially similar. All of Kolbe's reasoning and all of his formulas could easily be translated into type notation, and vice versa. Their disagreements were usually unrelated to the distinctions that did exist between their theories. Wurtz could point to the genetic relationship between glycol and glycolic and oxalic acids to argue that glycol was a dialcohol corresponding to a reduced form of the two acids, but he then ran into trouble explaining—or rather, did not even try to re-solve—the precise genetic relationship between ethylene and glycol, a point seized upon by Kolbe. Kolbe could easily account for the latter genetic relationship, but had to suppose a rearrangement and to posit undiscovered isomeric alcohols to explain the former. This set of problems was eventually solved, at least in principle, by Alexander Crum Brown's argument in 1864 that ethylene was not in fact CH3 CH, as virtually everyone had assumed by analogy to carbonic oxide, but rather CH2 CH2 .[16]