The Bonds of Conformity W. ~ . " N o ~ and e s the Initial Failure of the Lewis Theory in America Martin D. Saltzman Providence College, Providence, RI 02918 My own interest in the chemical side of the electron theory arose from an accident. Some twenty years or so a:o (18981,I had occasion to perform the old Hofmann experiment.. .The tube is filled with chlorine, and when a solution of ammonia is added, this chlorine should combine with an equal volume of hydrogen and liberate the nitrogen with which the hydrogen was combined. Thus, William Albert Noyes (1857-1941) described to an audience a t the Franklin Institute in 191.8 (1) how he first became interested in the role of the electron'in chemistry. W. A. Noves and several of his American contemooraries were among the first chemists to utilize the electron to explain organic structure and reactions. Their theoretical framework proved to be faulty, yet it did show that in the emereine American oreanic communitv there was an interest in c n h 6 n i n g theory with empiricall~dertvedexperimental evidenre. European organic chemists at this time were chieflv concerned with synthetic chemistry, and theory was not considered to he of any great importance, except in England. Ironically, the early interest of American chemists in the electronic interpretation of organic chemistry proved to be a major factor in why G. N. Lewis' electronic theory initially failed in his native country. Through the papers and letters of W. A. Noyes (2),we will attempt to show why this happened. William Albert Noves obtained his PhD in 1882 with Ira Remsen a t Johns ~ i p k i n after s undergraduate studies a t Grinnell College. Graduate education a t Johns Hopkins was organized in form and substance along the Germanic model that Remsen had been exoosed to in his studies with Fittie a t GAttingen (1869-72). In contrast to Remsen's lack of interest in theoretical organic chemistry, Noyes, early in his career, combined both fields along with an abiding interest in chemical education. Noyes' career involved academic and government posts including service a t Minnesota (1882-18831, Rose Polytechnic Institute (1886-19031, National Bureau of Standards (1903-1907), and finally the directorship of the chemical laboratories of the University of Illinois (1907-1926). In addition, among his accomplishments were the first editorships of Chemical Abstracts (1907-19101, Chemical Reviews (1924-1926), and The American Chemical Society Scientific Monoeranhs Noves was also editor of the .. . (1919-1944). . J~rurn,rlo / i h ~r l m ~ r ~ c o~n' h ~ m ~ ~ 0 1 Sfrom o c i1902-1927. ~I~ This was a crucial period for the Journal, as Noves helned in the merger of ems sen's American chemical ~ o & n a liito the Journal of the American Chemical Society. In 1901; when Noyes was a t Rose ~olytechnic,he assigned as a senior thesis project the problem of the course of the ammonia-chlorine reaction. The reaction was found to yield nitrogen trichloride and Noyes rationalized the results as follows: If we suppose what seems not inherently improbable, that all reactions involving the decomposition of molecules are preceded bv ionization of the oar& of these molecules. it would follow that eiementary mo~ecu~ks ionize into positive and negative parts (3). At this point, Noyes was applying the concept of ionization that Arrhenius had postulated in 1887. In 1904, in his
Silliman lectures a t Yale, J. J. Thomson discussed the arrangement of electrons in atoms and their role in chemical bonding. Thomson's model was a modern form of Berzelius' dualism in which the chemical bond was formed by "loss or eain of comuscles of electricitv which created chareed .. soecies" . ib,. ~ o ~ e s ' l u t ine r 1904 at the Inrernationnl Congress o i Arts and Sciences held in St. Luuis stated: If, however, we acrrpr rhe theory uf rl~ctnms.i t is evident that nu the rlrctrons musl be prrsent in r h r m d w ulr (31 an rle~~trolvrr rndtter i n wh~chmanner it ic tormrd. I t i> hut a strpfmhrr tutnr conclusion that the electrons are involved in every combination or separation of atoms and indeed, may he the chief factor in chemical combination ( 5 ) . T h e concept of the electron as the basis of bonding was also appreciated by other American organic chemists. Over the next two decades attempts to apply the electropolar hypothesis to organic reactions were made by Julius Stieglitz (1867-1937) a t Chicago, Harry S. Fry (1878-1949) a t Cincinnati, and K. George Falk (1880-1953) and James F. Nelson (1876-1965) a t Columbia. These chemists along with Noyes formed a distinct group who used what would become known as the electron theory of valence as the basis of their rationalization of organic structure and reactivity. Falk and Nelson (6) used the electron theory for the explanation of physical and chemical properties of organic compounds. Fry (6) wrote several papers as well as a monograph on aromatic substitution using the electron theory. Stieelitz ( 6 ) thoueht that with the electron theory he had discovered the ba& of why molecular rearrangements occur in certain organic compounds. Noyes himself did not contribute much in the way of applications to organic chemistry per se but considered the larger problem of proving the worth of the electropolar theory of chemical bonding. His eminence in American chemistry was viewed as an imprimatur by his collragt~rsin pressing their v i e w against their rrltirs. Thrir views quickly generared a wide vurirty of iritic ism. Typical of Noves' work in this era ! 1912) was his oauer on "Poisiblr ~ x ~ l a i a t i of o nSome phenomena of ~ o n i a i i o nby the Electron Theorv" " (7). .. Structural formulas for compounds such as ammonium hvdroxide and nitric acid in use durine this oeriod as shown helow gave no clue as to why these compounds behaved as hases or acids.
When we write the formulas in the light of the electron iheory, as shown in this diagram, we see that the negative oxygen of the hydroxyl may be held strongly by the positive nitrogen atom of while it is not held sostrongly by thenitrothe nitro group (NO*), gen atom of the ammonium group (NH4).The nitrogen atom of the latter group has lost four electrons and gained one, if we assume a localization of valence electrons (7).
Volume 61
Number 2 February 1984
119
Implicit in the electron theory was the belief that if a hond was formed by electron transfer from X to Y to produce the product x+YL, the reverse process could also occur to produce X-Y+. These two compounds were referred to as electronic isomers or eledromers. Noyes attempted to produce the pair N+3C13- and P3C13+ in order to prove the electron theory of valence. The compound N-CIS+ had already been prepared, and Noyes stated that if the other electrons could he prepared then: The discovery of such a compound would make it seem prohable that electrical charges are present in compounds and are not merely found during ionization of compounds in solution. This would give support to the hypothesis that atoms are held in combination by electrical forces (8). The attempts to prepare the electromer Nt3C13- or for that matter other electromeric pairs always failed, hut this did not dim the enthusiasm of Noyes and his circle for the hypothesis. The intrinsic problem with the theory can he viewed as follows. How do you treat bonding in ethane, for instance? Is one carbon positive and the other negative? As Stewart Bates remarked in 1914. It is difficult to conceive of two methyl groups in ethane as being so different from one another as would be the case if one of the molecules is positive and the other negative. In the vast number of reactions, two halves of such molecules react in an identical manner (9). The electron theory fails totally when applied to organic compounds, in which polarity does not exist. Yet Noyes, Stieglitz, Fry, Falk, and Nelson persisted against their detractors. To their credit these classically trained chemists were pioneers in the early development of physical organic chemistry in America. There was much interest in their ideas internationally as evidenced by an entire chapter devoted to the electron valence theom in Heinrich's "Theories of Organic Chemistry."The 1921 edition translated into End i s h bv Treat Johnson of Yale was used extensivelv . in .gradunte c~,urseshere a ~ abroad ~ d in the 1920's. There is no mentiun at all of thr Lewis theorv in the text and theonlv reference is in a footnote. Bv 1914 J. J. Thomson had modified his views concerning valence (10) to include nonpolar bonds. These could he formed by two similar-sized atomic spheres overlapping with the electrons shared equally. Thomson envisioned tubes of force emanating from the electron of one atom somehow anchoring itself to the nucleus of another atom and vice versa. This would lead to a nonpolar bond with the minimal number of electrons being two. In 1915 two significant papers concerning bonding came across Noves'desk as editor of the Journal of the American ~ h e m i c a k o c i e t y .One was by Albert parsois (1889-), and the other by Gilbert Newton Lewis (187E-1946). Both offered a comprehensive rationale for how all types of honds found in both oreanic and inorganic molecules could he understood on a single theoretical basis. Parsons' conception of valence involved the production of magnetic momenis by the circulation of elections in atoms. These magnetons, as he called them, were located a t the corners of a cube and could he shared in a variety of ways to produce the gradation in polarity found in F2 and HF, for kxample. ~ o y e was s forced to reject the paper because, as he recalled later @a), the Canadian physical chemist, William Lash-Miller, as one of the reviewers, had insisted that it was physics and not chemistry. Julius Stieglitz, a member of Noyes' circle, however, strongly recommended its publication (2a). Noyes saw merit in Parsons' idea, as did G. N. Lewis, in whose department Parsons worked, and he arranged to have it published in the Smithsonian Institution's Miscellaneous Collections (11). Noyes' respo&e to the magneton theory can he found in a lecture on valence delivered in 1917 and reprinted in the Journal of the Franklin Institute in 1918. 120
Journal of Chemical Education
The magnetan theory of the structure of the atom has also been developed and elaborated by Parsons. It cannot account for ionization, where, if we accept the electron theory at all, electrons must be transferred completely from the positive atom or group to the negative ( 1 2 ) . The Lewis paper which was accepted and appeared in April 1916 issue of the Journal of the American Chemical Society was entitled simply "The Atom and the Molecule" 113). In this naner Lewis described how bv treatine the chemical hondas'an electron pair one could rationalize d l the various tvoes .. of chemical honds as one continuum. Interest in the Lewis theory in America among organic chemists was minimal in the next decade. We shall now try to probe the reasons for this lack of interest by looking a t the way Noyes reacted to the Lewis theory. His problems mirrored the difficulties other American organic chemists had with Lewis' theory. One of the first recorded reactions of Noyes to the Lewis theory was his address to the AAAS meetine held in Baltimore on December 27,1917. His lecture was la& reprinted in Science in 1919. G. N. Lewis has proposed the hypothesis that carbon compounds are not held together by polar valences because they do not readily ionize. It seems possible that this is truein some cases, hut it is difficult to believe that there is any essential difference between the reaction of methyl iodide with silver nitrate and that of potassium iodide with the same agent (14). I n February 1917. Noyes himself published in the Journal of rhr ~ r n e r i c o n hem-icol ~ o c r r t ;dome new ideas ahuut bonding in a paper, "A Kinetic Hypothesis in Explaining the Function of Elertrons in the Chemiral Cmnhination uiAtoms" (15). The assumption was made that a single valence electron in rapid motion would bind two atoms hy the interaction of the negative electron with the two positive nuclei. For carbon atoms, Noyes drew the following picture:
- ",+-
,.A.
/
Why did other organic chemists have such difficulty with the electron-pair concept? One reason for this was G. N. Lewis himself. From all indications it appears that Lewis had little .. interest in organic chemistry. In his 1916 paper as well as his 1923 monoeraoh "Valence and the Structure of Atoms and ~ o l e c u l e s " f e ~if any substantial examples of application to oreanic molecules can he found. Lewis was aiding hv these omissions the chasm that was oerceived to e x i s between physical and organic chemistry'hy their respective practitioners. Further evidence comes from the wav Lewis built the department a t Berkeley which he took ovei in 1912. He recruited a group of highly innovative inorganic and physical chemists hut seems not to have done the same for organic chemistrv as Melvin Calvin (16)relates "From that time on. all the rest of the departmen; ufrhemistry were homemade." Gerald E.K. Rranrh was the unlv member ol'the deoartment publishing in the area of organicchemistry between1916 and 1926. His work for the most part was synthetic in nature, though he did do some kinetii studies. However, his papers do not use or mention the Lewis theory during this period. Branch in the 1930's would become one of the leaders of physical organic chemistry in America, however, primarily through his close contacts with C. K. Ingold which commenced in that period. Lewis' work also could not have aooeared a t a more inoo.. portune time in terms of its being accepted. European chemistry had come to a virtual standstill because of the war. With America's entry into the war Lewis himself was eliminated from further work in electron pair bonding because he joined the chemical warfare service. When he returned from France in 1919 he devoted his research efforts to thermodvnamics and not to the electron pair hond. In 1919, Irving Langmuir (1881-1957) a t the General Electric Company took up where Lewis had left off. His series ~
~
~~
of three papers appearing between 1919 and 1920 further amplified the electron pair concept and introduced the octet concept in detail (18). Langmuir was a tireless proselytizer for thr rlictron pair and hega
