VALENCE AND THE STRUCTURE OF MOLECULES
By Linus Pauling
Lecture to Teacher's Institute, Pasadena, Friday, Dec. 18, 1936.
4:30 PM, Culbertson Hall.
1. Introduction Considering how small molecules are, and how complex the arguments used by the great chemists of the last century in formulating the theories of valence and stereochemistry on the basis of the chemical properties of substances, I never cease to admire them and to marvel at their works. The theories of valence and stereochemistry may seem to be of importance to pure chemistry only; however, when we think of the great contributions of chemistry to applied science and industry, and of the significance which chemical theories have had in this contribution, we see that the practical value of these theories can hardly be overestimated. It is the theory of valence more than anything else which has converted chemistry from an art into a science.
I do not need to mention many of the structural formulas assigned to molecules by the methods of classical stereochemistry. For methane we write [the Lewis structure of] CH3, carbon being quadrivalent and hydrogen univalent, and the four carbon bonds being tetrahedrally directed. For ethylene we write [the Lewis structure of] C2H4, the quadrivalence of carbon being retained by the introduction of a double bond. For carbon dioxide we write [the Lewis structure of] CO2, for periodic acid [the Lewis structure of] HIO4, etc.
In 1916 a great step forward was made by G.N. Lewis at Berkeley, in correlating stereochemistry and valence with electronic structure. Lewis's principle contribution consists in the identification of the covalent bond with a pair of electrons held jointly by two atoms. In most cases the Lewis electronic structures are identical with the valence-bond formula except for this change: thus [see actual manuscript for hand drawings of Lewis structures of CH4 and C2H4]. In some cases, however, Lewis preferred to keep no more than four electron pairs around an atom and so changed the formula, writing, for example, [see actual manuscript for hand drawn Lewis structure of] HIO4. We now know that for first row elements lewis's idea was right, but not in general for others.
This makes a nice scheme, as far as it goes - and it goes a long way, too. there are, however, still some questions that can be asked, some points in doubt. For example, which of the two structures above for periodic acid is the correct one? Should CO be written [two Lewis structure hand drawings of CO double and triple bonded]? Is O=C=O really right for carbon dioxide? What is the structure of nickel carbonyl, Ni(CO)4? We can write [hand drawing of Ni(CO)4, but this seems strange in several ways.
And what is the structure of benzene?
Recent developments have permitted many of these questions to be answered. These developments have consisted in the discovery and application of several new experimental methods of studying the structure of molecules, largely physical in nature, and of several new theoretical ideas, mainly growing out of the new quantum mechanics. I shall discuss in the main one new experimental method and one new idea.
2. Electron Diffraction by Gas Molecules. In order to determine the arrangement of atoms in molecules use is made of the phenomenon of diffraction. This is not a new idea - about 130 years ago Thomas Young, who shares credit with Champollion for deciphering hieroglyphics, devised an apparatus for measuring the fineness of wool by looking at the diffraction pattern of a beam of light passing through the wool. Consider a wave front striking on a diatomic molecule. We see that there is interference in some directions and reinforcement in others. Knowing λ and angles of reinforcement we could calculate d. If the molecule had all orientations, as in a gas, the diffraction pattern would be somewhat smudged, but still a pattern. this was suggested by Debye in 1915, who later used x-rays with very long exposures (days); the idea is one of those which gave him the Nobel Prize in chemistry a few days ago. In 1930 Mark and Wiezl diffracted electrons. For ten years it has been known that λ=h/mv.
[5 slides, showing waves, Brockway's apparatus, CCl4 picture.
3. Results of Diffraction Studies. the most interesting result of this work is the striking verification of the ideas of the old stereochemists. CCl4 is formed to be tetrahedral. But also CHCl3 and CH2Cl2 are tetrahedral, with angles differing from 109º28' by not more than 2º or 3º. This substantiation is found almost always.
Another result of great interest refers to inter atomic distances. C-Cl = 1.76 Å is just the mean of C-C = 1.54 Å in diamond and Cl-Cl = 1.98 Å in Cl2. Hence we can assign radii to atoms. [See actual manuscript for the radii of atoms, carbon, nitrogen, oxygen, fluorine, chlorine, and hydrogen depending on a single, double, or triple bond.]
2 slides. CCl4 etc, metaldehyde - 1.43 Å, 1.54 Å. P4, S8. It is clear that these interatomic distances provide an experimental method of studying bond type.
4. Resonance Among Several Valence-bond Structures. Now let us discuss the new idea introduced by quantum mechanics. Suppose that we ask: is it necessary that a molecule such as CO have a definite valence-bond structure? The answer, which is part of the new idea, is no; instead the CO molecule may have (and does have) a structure which is neither C = O nor CO with a triple bond, but is somewhere between them, or which rather has some aspects of both. It is customary now to speak of the molecule as resonating between these two structures. this does not mean tautomerism - there is only one kind of Co molecules in carbon monoxide gas; rather the structure is a hybrid.
The second part of the new idea is a very important one - resonance stabilizes the molecule.
Now let us consider benzene. [Drawings of two benzene.] Benzene resonates and the molecule is stabilized; this causes the double bonds to loose their unsaturated character largely. Each of the six bonds has 1/2 double-bond character.
Graphite. [Drawing of the bond-line structure of graphite.] Each 1/3 double-bond character.
December
18, 1936
