Intensive drying: Anomaly and the chemical ... - ACS Publications

In Thomas Kuhn's celebrated study of scientific revolu- tions he pointed out that there is a need for detailed exami- nation of the way in which scien...
0 downloads 0 Views 4MB Size
Intensive Drying: Anomaly and the Chemical Community Harold Goldwhiie California State University, Los Angeles, CA 90032 In Thomas Kuhn's celebrated study of scientific revolutions he pointed out that there is a need for detailed examination of the way in which scientific communities grapple with and attempt to resolve discoveries that are anomalous in terms of the accepted views of the time (I). Within the domain of physical chemistry there is a relatively recent example of a set of anomalous observations, having to do with the effects of intensive drvine on fundamental nhvsicochemical properties, that illu&Ges both the responses of chemists to the unexpected and the way in which anomaly can polarize a group of chemists until an eventual resolution is achieved. This example centered on a distinguished British chemist, one of the Presidents of the Chemical Society. The anomaly discussed in this paper did not lead to even a minor scientific revolution; it illustrates, rather, the power of what Kuhn terms "normal science" in the resolution of anomaly. As early as the 18th century there were sporadic observations on the effects of small amounts of water on chemical properties, effects we would now term catalytic. For example, Bucquet, in 1773, found that quicklime (CaO) did not absorb dry fixed air (C02)though the moist gas was readily absorbed (2). Similar isolated observations were made during the early 19th century ( 3 ) .Perhaps the first systematic study of the effects of traces of water on gas reactions was carried out by H. B. Dixon (4). Harold Bailey Dixon (1852-1930) was originally a scholar of classical bent who was rescued for science during his undergraduate days a t Oxford by Vernon Harcourt, one of the founders of the study of chemical kinetics. After collaborating with Harcourt on various dynamic studies Dixon decided to pursue an area of chemical kinetics that, a t the time (1876), was very little in vogue in Britain, namely the study of gaseous explosions. There had been pioneering studies of this area by Davy, in connection with his investigations of mine explosions, and Bunsen had published in 1853 studies on the velocity of propagation of gas explosions. Early in his work. in 1877. Dixon made one of his most celebrated observations: a mixture of carbon monoxide and oxygen, carefully dried over ~ h o s ~ h o r pentoxide. us did not explode upon the passage of an electric spark. The addition of atrace 01water, or of other comoounds containina hvdroeen. such as diethvl . ether, led to a-vigorous explosi& when the mixture was sparked. Dixon's observations on the effect of moisture on this particular gas r c a ~ ~ i owere n repeatril and confirmed by other i htmists and led to a viwrous d e h a e in the rhemicnl literature of the period on the mechanism by which the "catalytic" water exerted its effect. Among others, the distinguished chemists Lothar Meyer, Mendeleev, and H. E. Armstrong contributed to the discussion (5).Dixon's studies of the effects of water on this reaction were later to lead to a wider and more controversial examination of the influence of traces of water on a ranee of nhvsico-chemical ~ r o ~ e r & ? s by one of his students, ~ e r i e r~re"reton t Baker (1862:1935). Baker, like Dixon before him, was a distinguished figure of the British scientific establishment (6). The son of a Lancashire clergyman, he was educated at Manchester Grammar School and won a scholarship to Balliol College, Oxford. After taking a first-class honors degree in natural science he became a demonstrator at Balliol, and an assistant to Dixon. In 1884 he left Oxford to become a schoolmaster at Dulwich College, and for over 20 years he combined school teaching

with fundamental research in physical chemistry, an unusual combination of talents and enterprises at any period. He was elected a Fellow of the Royal Society for his research accomplishments in this period. Baker returned to Oxford as Reader in Chemistry at Christ Church College and in 1912 succeeded Sir Edward Thorne as Director of the Chemistrv Department of the Imperial kollege of Science and 'Technorogy in London, a post he held until his retirement in 1932. Among other honors Baker received the Longstaff Medal of the Chemical Society in 1912 and the Davy Medal of the Royal Society in 1923. In 1926 he served as the President of the Chemical Society. I t is clear that despite his controversial views on the effects of intensive drying Baker was held in high esteem by the chemical community. Baker's work on intensive drying falls into two main categories. The first involved reactivity, following rather directly from Dixon's work; the second, and more novel, had to do with the physical properties of pure substances. The two categories are distinguished not only by their subject matter but also hy the challenges they posed to accepted theories of chemistrv. While the results on reactivitv could. i t .ohvsical . is clear in retrospect, be rationalized without major revision of the acceoted theories of the time, the results on ~hvsical properties kerious~ychallenged the foundations of ihirmodynamics. After Dixon's work on the carbon monoxideloxygen explosion, it was a natural extension for Baker to examine the effects of careful drying on other combustion reactions. In 1885 he published results (7) that showed that very dry sulfur, boron, or phosphorus do not burn in dry oxygen, whereas dry selenium, tellurium, arsenic, and antimony do. He then turned to combination and dissociation reactions (8). Dried NO and 0 2 showed no tendency to form NO,; dry ammonium chloride heated with drv CaO vielded no ammonia. More controversial was the observation that dry ammonia and HCI did not react to eive ammonium chloride. Baker's standard method gf intensive drying was to treat his reaaents with successive oortions of sublimed ~. h o s.o h o rus pcntuxiile. Other wrkers, whoco~~lrl not repeat his study ofthc1.1ck ofreaction hetw~cnNII1and HCl pointed i ~ ( 9t1 that each of these substances reacted with P205 SO that Baker could not have dried them as he had suggested. A more detailed description of his procedures followed from Baker, showing that they were plausible (10). Throughout his work Baker used very careful techniques for purification. He stressed, in a paper published near the end of his career ! I I ). lhar m a n i p ~ ~ l ; l t i v e ~ ~ ~wry r e o thigh ' a :mil unusual order u,its newisar) i n drying glassware and reagents ro achieer rearoducihle rewlfs in the srudv of intensi\.edr\,ine. Baker's " major contributions to chemistry were in the areas of analysis and precise atomic weight determinations, so his comment that "purity of the materials far beyond the standard required for atomic weight purposes is absolutely essential" is highly significant. Baker next examined dissociation reactions and reported that NHdCl, Hg&, and N ~ 0 3 ,all standard compounds showing gas-phase dissociation, did not dissociate when dried (12). In addition the boiling point of dry N203 was abnormally high. This led Baker to study the boiling points of dry nondissociatingliquids and produced his most controversial publications. He began his study of this topic around 1910 and had sealed a number of liquids over P205 just before the

-

Volume 64

Number 8 August 1987

657

outbreak of the First World War when he published his first results in this field (13). The war led Baker's work in other directions, and he did not resume his experiments on the dried liquids until the 1920's, when he had some striking results to puhlish (14). Bromine, dried for eight years, had its boiling point raised from 63 O C to 118 OC. Mercury, dried for nine years, had its hoiling point raised from 358 "C to 420425 "C. Ether, dried for nine years, had its boiling point raised from 35 "C to 83 O C : its vanor . -pressure at 20 "C. normally 442 mm, was now 374 mm. I t must be emphasized that these results were for liquids predistilled from the drving agents and tested for purity; in particular, it was cleariy shown that there were no volatile impurities containing phosphorus ptesent. However, ~aker's-methodfor determining "hoiling points" was criticized. He placed flasks containing the dried liquids into baths that were heated until ebullition occurred, and i t was the bath temperatures that were initiallv renorted as "boiline noints". Baker himself commented t h a t thermometers ig ihe condensing vapors "curiouslv" showed temoeratures onlv 1-2 OC above normal hoiling points. The work on intensively dried liquids was continued over the next decade, and i t is instructive to follow its progress in both original publications and the Annual Reports of the Chemical ~ u c i i tat~ that , period the most u,idely read review o i ~ . h r m i t r yin k:nglish ( 1 5 ) . The 1022 Ilvp(,rt (16) summarizes Dnkcr's work undrr the heading "Chemical Reaction" and describes it as "results which c k n o t be placed in any definite category (unless we invoke the hard-pressed word catalysis) and yet are clearly of fundamental significance". The next year (17) the same reporter, under the heading "Unclassified Phenomena", shows his bias as a Baker sup~ o r t e r .He cites a reference (18) that alleeedlv - . supports .. Diiker but that is in fact mther noncommittal, and he criticizes G. S . I.ewis, u,ho had published (19) a thermodynamic interpretation of Baker's results, for not realizing that his view was not original but had already been presented (20): "Even among chemists who accept them, these results are evidently less well-known than their importance merits." The theoretical importance of Baker's results on pure liquids was that they constituted a challenge to thermodynamics and views of liouid structure. Baker had his ownview ui !he causes uf his spectacular results and expounded it fnllv in his 192: I'rtsirlentinl ;\ddrrs.i to the Chemical Society'(21) entitled "Experiments on Molecular Association". He viewed pure liquids as being associated into dimeric or larger molecules, citing the well-known example of acetic acid. A trace of water was unusually effective in breaking down the dimers into monomers; thus, ordinarily dried liquids, containing traces of water, were unassociated. Intensively dried liquids were associated. "All liquids may be regarded as analogous to a dissociahle gas, such as nitrogen tetroxide." His mechanism of the action of water was based on a suggestion made in another context by J. J. Thomson (22). 1f there is an electrostatic attraction between two polar molecules and a polarizable Hz0 molecule is added to the svstem. the dioole induced in the HgO molecule lessens the initial attraction, and dissociation occurs. A similar explanation was ~ uforward t bv A. Smits. another contributor to this field, th;ugh with lesskmphasis bn the mechanism of action of the water. In Smits' view any phase of a pure substance may contain different species of different molecular complexities, and the equilibria between these species may be established at rates that may vary according to the amount 0 1 a.iit6.r. or o t h t ~ catalyst, present (A?). The di\.ision in the attitu(lcs of chemists to Baker's results isclearlvshown in twocontrasting reviews in the 1927 Annual Reports. In the first, under the title "Molecular Association", the reporters give a favorable review, and discuss the remarkable effects of water as a catalyst in molecular association (24). The second review, by a different reporter (25)in

1

858

Journal of Chemical Education

the section on "Kinetics", approaches the results more circumspectly and concludes "there is no longer any general belief in water as a 'universal catalvst' ". This renorter disringuisheb i)erwen herrrogenrow and homogeneous et'ferts of water and m~intsont thin whilt: surfnre effects of wawr in heterogeneous reactions are plausible, there are fe* homogeneous reactions (apart from ex~losions-a stranae omission on his part) in which an unch.kenged role for water is ohserved. For instance, althoneh there is a water effect in the NOiOz reaction, the reactionproceeds even in the absence of water. By 1928 even previously enthusiastic Baker supporters were chilled by the appearance of significant negative results. Work by Lenher and Daniels (26) had shown no effects after four years of drying benzene and carbon tetrachloride, and the reviewers stated (27): "The work . . . presents a conflict of evidence, and is therefore reported here without comment.'' The followine - .vear more neeative results aopeared. Timmermans found no effects on henzene, p-xvlene, or cvcluhexane after 3.5 months of drvine (281.I t was at this time that Baker, ably seconded by theaptly named Bone, made his pronouncement about the level of experimental care needed to obtain correct results in the area (11). Nevertheless, negative resnlts continued to accumulate. Haker had earlier r&orred experiments (21 I in which npplication oi an electric field to m(xlerntclr dried henzent had raised its boiling point by over 10 "C, which supported his polarization theory of the effect of water. Smits, originally a Baker advocate, could not repeat this work and suggested that the effect observed was simply superheating (29). Lenher pointed out that in dust-free liauids facile and considerable superheating can occur before ebullition. In Lenher's experiments henzene. of normal boiling point 80 'C. had to be heated in a bath to 106 'C before it boiled, but the temperature of the condensing vapor was normal (30). Smits, retracting earlier work, showed that reported effects on the vapor pressure of hexane were not re~roducihleand . . that there were problems in equilibrating liquid and vapor (31). He also showed that an increased vapor pressure of "dry" N204 was probably due to reaction between traces of nitric acid in the oxide and the drying agent. By now Baker was reaching the end of his career. He was 68 in 1930, within two years of retirement, and he seems to have lost his drive to puhlish his work in this field. In fact his last puhlications in 1929 and 1931 were on techniques rather than results (11,12), and many of his results were not pnblished directly but were summarized in his Chemical Society Presidential Addresses (21,321. Without further work from Baker's group, and with the growing body of negative results, work on intensive dryingvirtually ceased. After 1931 it was no longer reviewed in Annual Reports. The 1931 review by C. N. Hinshelwood is a dispassionate summary of the whole problem (33). The unchallenged observations that after intensive drying a liquid has to he heated to an unusually high temperature to give a stream of vapor can be interpreted in two ways. The dynamic interpretation is that of suoerheatine. Removal of dust and water dn)plets from the liquid reduces the rntr oievapurnti(m, and this can be interorered oualitarivelv hv the nucleation theory of bubble foimatio; The stat; tierrnodynamic interpretation of Baker and Smits is that there is a major displacement in the internal equilibria of the liquid when the last traces of water are removed. This is quite anomalous in terms oicurrent thermodynamic rheory because it implies a suhstimtial rhanrr in rhe frre enrrxy of a pure li(luid on the addition of a trace of water, whereas further smail amounts of water give either zero or very small changes in free energy. Normally free-energy changes are linear in concentration for the addition of small amounts of one substance to another. Not surprisingly, Hinshelwood opts for the normal scientific interpretation of the data. He opines that the dynamic

changes are real, and that they fall into the wide category of surface effects or catalysis. They are not yet well understood, hut at least they pose no challenge to the fundamentals of physical chemistry. The static changes he dismisses as probably spurious. But the static changes refused to disappear completely, even though the majority of chemists and reviewers ignored them. In 1934 Manley reported (34) that the refractive index of benzene underwent changes after intensive drying that could best be interpreted by Baker's theory. In 1937 Lacoss and Menzies published perhaps the best described and documented study on the effects of intensive drying on benzene (35), and concluded that the vapor pressure of benzene was lowered if it was dried a t room temperature and raised if it was dried a t 90 OC or 105 "C. They observed a maximum hoiling point elevation for benzene of 2 OC, well below that observed by Baker, but their work is much better descrihed, and apparently much more critical, than most earlier studies. Lacoss and Menzies' experiments attracted little attention and sank almost into ohlivion-almost, hut not quite. In J. R. Partington's massive survey of physical chemistry, there is a full review of intensive drying (36),and Partington shows clearly, by his footnoted comments, where his sympathies lie. He points out that "the quality of the P2O5 has an influence on the results". "Lack of attention to the purity of the PpOs has led to mistakes in the past." On cleaning the glassware to be used he notes "the complete unsuitability of chromic acid-sulphuric acid mixture . . . [which] should be banned in every laboratory." In apparent reference to those who doubt the validity of Baker's results in this difficult field he comments acidly: "Experimenters who do not care for its tediousness should enrich science in some other field." Nevertheless, Partington avoids the thorny question of the interpretation of Baker's results, he confines himself to reporting the data. On the whole question of the chemical significance of this topic one cannot improve on Hinshelwood's 1931 comment. "Experimentally the situation is not unlike that prevailing in psychical research where, we are told, most of the evidence can he ruled out hut a small obstinate residuum has to he contended with" (33). (I may remark, parenthetically, that to date no one has really contended with the obstinate residuum presented by the experiments of Lacoss and Menzies.) For the chemical historian the episode of intensive drying is an illustration of the power of normal science. The anomalies, at first presented as a group, are divided into those involving effects on dynamics and those involving static properties. The former can be classed with other homogeneous and heterogeneous catalytic problems. I t is true that this area is still full of puzzles, but they are the puzzles of

normal science, and there is expectation in the chemical community that they will eventually yield to existing theories. I t is only because of comparative complexity and the consequent difficulty of experimentation that they have not yielded so far. The effects involving static properties present a direct challenge to thermodynamic theory and so require a frontal counterattack. Experiments are shown to he irreproducible, and inevitably the two or three failures to produce anomalous results are weighted more heavily by the representatives (the reviewers) of the chemical community than are the orieinal anomalies. As counter examnles multinlv. ". the anomalies become buried and lost to view. The protagonists retire from the conflict. even thoueh " some issues remain unresolved, and a new generation has other, more pressing, problems to investigate. Eventually the whole episode is forgotten. Modern physical chemistry textbooks never mention intensive drying hut simply present the accepted theories of the thermodynamics of liquids, for textbooks of science, unlike those of history, record only the victors of scientific conflicts.

.

Literature Clted 1. Kuhn, T. S. The Structure oIScianlificReuoiutions; Univenity of Chicago: Chicago.

Bone. W.A.;B8ker.H.B. J. Chem. Soc. 1931.3349. 5. Bane, W . A.J. Cham. Sac. 1931,338. 6. Philip, J. C. J. Chem. Soc. 1935,1893. 7. Bsker, H. B. J Chem.Soc. 1885.47.349, 3. Bsker. H. B. J. Chzm.Soc. 1894.68.611. 9. Gutmann, S. Liebip's Annolen 1898.299,267. 10. Baker. H. B. J. Chrm.Sor. 1898,78,422. 11. Baker, H. B. J. Chem. Sor. 1929.1661;Bone, W.A. J . Cham.Soe. 1929.1664 12. Baker, H. B. J.Am. Chem.Sur.1931.53, 1810. 13. Baker, H. B.;Baker. M. J. Chem.Sac. Trans. 1912,101,233'3. 14. Baker, H. B. J.Chem.Soc.Trans. 1922,121,568. 15. Chrm. Reus. began publication in 1924. 16. Biiscoe, H.V. A. AnnuoiRepb. 1922.19.35-39, 17. Briseoe, H. V. A . Annual Repls. 1923,20, 32-34. 13. Roberts. H. M.:Bury,C.R.J. Chem.Sac. Trans. 1923,123,2037. 19, Lewis, G. N. J. Am. Chorn.Soe. 1923.45.2886. 20. Smits.A. ZPhysicol Cham. 1922, 100,477. 21. Baker, H. B. J. Cham.Soe. 1927,949. 22. Thornson, J. J. Phil. Mag. 1893.36.320. 23. Srnits. A. The Theory o/Allotropy:Longmans: London, 1922. 24. Brkcoe,H. V. A,: Rohinson,P.L. AnnuolRegts. 1927.24. 33. 25. Hinshehuood. C. N. Annuol Repta. 1327.24.321-322. 26. Lenhar, S.:Dsniels, F.Proc.N o t . Acod.Sci. 1928. 14,606. 27. Brisc0e.H. V.A.;Robinson,P.L. Annuol Repis. 1928.25.38-19. 28. Tirnmermana. J. Bull. Sor. Chim. Baig. 1929.88.160. 29. Srn1fs.A.J. Chem.Soe. 1928.2399. 30. Lenher,S.J.PhysChem. 1929,33.1579. L

31. Srnits,A.:Swarf,E.:Bruin,P. J. Cham. Soe. 1929,2712. 32. Baker. H.B.J. Chem Soc 1928,1051.

Volume 64

Number 8

August 1987

659