Professor P. Federlin is thanked for helpful comments and a copy of Ref. [7 b]. The assistance of Annette Allen, Professor Jean-Marie Lehn, and Dr. Valerij Nikolaev is gratefully acknowledged, as is financial support by the Natural Sciences and Engineering Research Council of Canada, the Petroleum Research Fund administered by the American Chemical Society, and a Killam Fellowship from the Canada Council for the Arts.
An explosive history: Diphenylketene was isolated and identified by Staudinger (see photo) in 1905, as the first example of this class of highly reactive compounds. Others had previously generated these species, but missed the opportunity to claim their discovery. This Essay provides an enlightening view of the history behind these useful reagents.
Organic reactive intermediates1 are the species involved in multistep organic reactions, and the beginning of the 20th century saw convincing experimental evidence for these entities become available. The discovery by Gomberg in 1900 that triphenylmethyl chloride and zinc formed the stable triphenylmethyl radical (1) [Eq. (1)(1)],2 was widely noted, and Gomberg narrowly missed the award of the Nobel prize for his discovery.2c This was quickly followed in 1901 by the demonstration that triphenylmethyl chloride in an ionizing solvent also formed the triphenylmethyl cation 2 [Eq. (1)].3
A carbanion intermediate was suggested by Clarke and Lapworth in 1907 to occur in the benzoin condensation,4a and the conversion of 1 or triphenylmethyl chloride with sodium into triphenylmethylsodium was demonstrated in 1914 by Schlenk and Marcus, who studied the ion pairing of the intermediate by measuring the conductivity.4b The name carbanion was coined by Wallis and Adams in 1933.4c
The intervention of carbenes was inferred to explain the Buchner reaction [Eq. (2)(2)]5a–d and analogous reactions of diazomethane.5e Many examples of stable carbocations, carbanions, and carbon-centered free radicals have been isolated, as have carbenes,5f,g which formerly seemed the most unruly of the reactive intermediates.
These findings combined with an emerging understanding of the theory of organic chemistry incorporating reaction kinetics and stereochemical and structural studies created what came to be known as physical organic chemistry. After decades of work, argument, and controversy the pivotal role of reactive intermediates was well recognized by 1940, marked by the text “Physical Organic Chemistry” by Hammett.6 As pointed out by Leffler in 19561a reactive intermediates in reactions were assumed “which have not been isolable for direct experimental study” but whose existence could be inferred by comparison to observable species “of the same type as the hypothetical intermediate” and which “have a great variety of degrees of stability, depending upon their structures. The extrapolation to the properties of the hypothetical intermediate is therefore a continuous one.” Now with vastly improved methodology many formerly hypothetical intermediates have been observed directly, and even isolated. There are still hold-outs which are too short-lived for observation, or whose preparation has eluded the synthetic skills of the investigators, and these provide a continuing challenge.
In 1903 a young instructor (Unterrichts-Assistent) Hermann Staudinger joined the laboratory of Thiele at the Kaiser-Wilhelms-Universität of Strassburg (Strasbourg in French) in the Alsace.7a Strasbourg occupies a strategic position on the Rhine River on the border between Germany and France and over the centuries has been part of one or the other, depending on the political and military situation. After the Franco-Prussian war of 1870–1871 it had become part of the new Germany, and the University was endowed with an imposing chemical laboratory (Figure 1), which still stands.7b The University is now a prominent center of chemistry in France, the Université Louis Pasteur.
In 1905 Staudinger reported the discovery of the first example of an unexpected new family of reactive intermediates, diphenylketene (6).8a As described later by Staudinger in his scientific autobiography8b he was inspired by the work of Gomberg to attempt the preparation of new free radicals. He examined the reaction of α-chlorodiphenylacetyl chloride (5) with zinc and was rewarded with the isolation of 6, the first demonstrated example of these highly reactive organic compounds [Eq. (3)(3)]. Diphenylketene (6) was obtained as a low-melting-point solid, and so the identity of this species was secure, although ketenes are in general highly reactive, and the existence of some examples is still in doubt.
Staudinger seized the opportunity and pursued the study of ketenes for the next 20 years (Figure 2), after which he devoted himself fulltime to the exploration of polymer chemistry, for which he won the Nobel Prize in 1953. Within two years of the initial discovery dimethylketene (7) had also been prepared and found to undergo [2+2] dimerization to give the symmetrical cyclobutanedione 8 [Eq. (4)(4)],9a and 6 was found to participate in cycloadditions with the imine 9 to form β-lactam 10 (Staudinger reaction) [Eq. (5)(5)] and with carbonyl compounds to form reactive β-lactones.9b The reaction of 6 with cyclopentadiene gave a product9b that was later proven to be the [2+2] cycloaddition adduct 11 [Eq. (6)(6)]. These discoveries were made two decades before the development of the Diels–Alder reaction, and are still mainstays of synthetic organic chemistry. Cycloadditions have remained the most distinctive, useful, and intellectually stimulating aspect of ketene chemistry.
Staudinger, Wedekind, or Wolff
The discovery by Staudinger had, however, been foreshadowed by the investigations of Wedekind, who in 1901 at the University of Tübingen reacted diphenylacetyl chloride (12) with nPr3N, and noted the formation of nPr3NHCl. He proposed the formation of the intermediate 13 (this was represented as 13 a) [Eq. (7)(7)], which appears equivalent to ketene 6 but was not isolated or characterized.10a Wedekind wrote “Herewith is made the hypothesis, that the atomic grouping (R1.R2).C2O. is temporarily capable of existence in solution.”10a This was 15 years before the proposal of the electron pair bond by Lewis, but the idea of the tetravalence of carbon was well established. However, in the next year Wedekind was more equivocal about the existence of the ketene.10b This work was known to Staudinger and frequently acknowledged by him,8a but Wedekind missed his chance to be recognized as the discoverer of ketenes. He worked in this area for a decade, but there is no mention of this research in his obituary.10c,d
Ketenes had also been generated by a completely different route by Ludwig Wolff (Figure 3) at the University of Jena in 1902, by a reaction that became known as the Wolff rearrangement.11 This involved heating the diazo ketone 14 (formulated at that time as having a cyclic structure) at refux in water to form the ketene 15. The latter led to the unobserved acid 16, which underwent decarboxylation to give 17 [Eq. (8)(8)].11a Wolff recognized that rearrangement had occurred and that the acid 16 was a key intermediate. But rather than assuming ketene formation, he proposed that the addition of water occurred to give 18, which underwent a double migration to form 16, and cited the analogy of the pinacol and benzilic acid rearrangements [Eq (9)(9)].11a The proposal of a ketene intermediate in the reaction was made by Schröter in 1909, after ketenes had become well-recognized intermediates.11c This reaction was adapted as the Arndt–Eistert reaction for chain elongation, with a proposal for a carbene intermediate,11d and was developed for photochemical ketene generation by Horner.11e,f The photochemical version of the reaction is particularly valuable for the study of reactive intermediates, as highly reactive ketenes can be generated in various media from the reasonably stable diazo ketones. Addition of water to the ketenes gives observable acid enols as further observable reactive intermediates, which then form the carboxylic acids.
Why is Staudinger and not Wedekind or Wolff credited with the discovery of ketenes? Recognition of priority for discoveries is of intense interest to scientists, but there is no apparent record of dispute between Staudinger, Wedekind, and Wolff. Staudinger was a highly prolific investigator, and with more than 50 papers and a book on ketenes12a following his 1905 work in rapid order, his position as the founder of ketene studies has not been challenged. The nature of chemical discovery has been discussed by Berson,12b and Staudinger's clear recognition of what he had discovered and extensive further study established him as the founder of the field. Wedekind had correctly identified the structure of the ketene as an unisolated reactive intermediate,10a but was not firm in his conviction and missed his opportunity.
Staudinger's further career in the tumultuous development of polymer chemistry, as well as his complex personal life and sometimes outspoken political views, have generated much comment.12c–e Wolff has the honor of two well-known named reactions, the Wolff-rearrangement and the Wolff–Kishner reaction, and can be satisfied that every student of organic chemistry still sees his name. Ironically Berson's comments on the need to recognize a discovery were directed at Fittig, Thiele's predecessor and Wolff's supervisor at Strassburg, who had discovered the rearrangement of pinacol with acid, but did not recognize what had occurred, and so the reaction is known as the pinacol rearrangement, an example of the broader class of Wagner–Meerwein rearrangement. As noted above Schröter was the first to formulate the ketene intermediate in the Wolff rearrangement,11c and by Berson's criterion Wolff does not deserve credit for the ketene discovery.12b
Ketene and Its Dimer: Wilsmore's Contribution
Staudinger was not alone for long in the study of ketenes, for in 1907 from a completely different perspective Wilsmore (Figure 4) prepared CH2CO (19) from the thermolysis of acetic anhydride or acetone with a hot platinum wire [Eq. (10)(10)].13a The parent ketene is a noxious and poisonous gas, and its preparation at University College (London) inspired “The Ketene Song”, perhaps sung at the occasion of a departmental party.13b
The Ketene Song
There is a new substance discovered
In a room that's just over the way.
—The inventor's assistant recovered,
Though it was a near squeak, doctors say—.
But though it is ages & ages
E'er the world saw as much as a grain,
Later on you will hear, say the sages,
Of keten again.
It has quite a good constitution
—C twice and H twice and an O—:
And even in weakest dilution
Of its presence you will very soon know.
For the smell of that simple creation
Will grasp at your nose & remain
And hours after you'll sneeze in iration
At ketene again.
It's a simple enough preparation.
You stick in acetic a wire
Which gives the required dehydration
When raised to a red heat or high'r.
Then out comes a torrent of gases
Bearing liquid along in their train:
When this you have trapped ere it passes,
You've keten again
When the O.L. in slumber reposes
And room No 7 is asleep
Dr. Wilsmore, as usual, proposes
To let loose the winds on the deep.
And the sleepers, awakened, grab vainly
After beakers that scatter like rain,
While A and B rage quite insanely
At ketene again.
(A & B were two small research labs, and O. L. refers to the Organic Lab.)
Staudinger and Klever shortly thereafter reported the preparation of ketene by the zinc debromination of bromoacetyl bromide.13c There was a brisk dispute between Staudinger and Wilsmore as to the priority for this discovery as well as to the purity of Wilsmore's sample of ketene and whether it had the formula CH2CO or HCCOH, another possibility noted by Wilsmore. This was resolved in favor Wilsmore and the ketene structure.13d,e
The development of ketene chemistry has been bedeviled with controversy, most prominently with the structure of ketene dimer 20 [Eq. (10)], originally suggested by Wilsmore in 190814a to have the acetylketene structure CH3COCHCO. This debate lasted for more than 40 years, and elicited the comment from one worker in the field “The extraordinary chemical behavior of the ketene dimers has lent exceptional interest to that class of substances, and the controversy which has raged for decades over the structures of the compounds is without parallel in the study of small molecules.”14b This was finally settled definitively by chemical investigations,14c electron diffraction,14d and X-ray diffraction.14e The continual dispute regarding the chemistry of ketenes may be attributed to their extraordinary reactivity and unique structures.
Hurd (Figure 5) was active in the discussion of the structure of diketene, and carried out many studies in ketene chemistry.15 He modified Wilsmore's procedure for ketene preparation by pyrolysis of acetone with the widely used apparatus known as the “Hurd lamp” (Figure 6).15e This method was also used for the industrial preparation of ketene for conversion into acetic anhydride.
Additions of nucleophiles to unsymmetrical ketenes generate new stereogenic centers, and this led to a long-lasting interest in steroselective reactions of ketenes. In an early study of asymmetric synthesis Richard Weiss at the University of Vienna treated phenyl(4-tolyl)ketene with the optically active alcohol menthol, and after saponification claimed to have obtained an acid with optical activity.16a The chirality of the acid arises from the different aryl groups in Ph(4-Tol)CHCO2H, and the claim that this compound exhibited significant optical activity was improbable and was eventually disproven.16b,c Even though this initial effort was not successful, the idea was to prove fruitful, and was later developed by Pracejus (Figure 7) at the University of Rostock into a practical reaction for aryl(alkyl)ketenes. Pracejus observed significant stereoselectivity in the addition of methanol to phenyl(methyl)ketene (21) catalyzed by the chiral base brucine to form 22 [Eq. (11)(11)]. Ester 22 is a precursor of α-arylalkanoic acids, which are of major pharmaceutical importance.16d,e The Gesellschaft Deutscher Chemiker (German Chemical Society) established the Horst Pracejus Prize for advancement of research in chirality in his honor, first awarded in 1999.
Stereoselective addition of chiral amines to aryl(alkyl)ketenes was also studied by Pracejus,16f and reaction of phenyl(methyl)ketene (21) with (S)-1-phenylethylamine in toluene at −100 °C gave the chiral amide (S,S)-23 [Eq. (12)(12)].16f
Many types of ketenes have been made, including bisketenes,17a that bear two ketene groupings in the same molecule. The first example, carbon suboxide (24) was prepared in 1906 by Diels and Wolf by the dehydration of malonic acid [Eq. (13)(13)],17b and then in 1908 by Staudinger and Bereza by dehalogenation [Eq. (13)].17c However, the isolation of further bisketenes proved difficult, and the simplest example, the carbon monoxide dimer 25, has been the object of continuing theoretical17d,e and experimental17f examination. Although an energy minimum structure for this molecule has been calculated,17f it is still unknown experimentally. It was concluded that this species is an intrinsically short-lived molecule that would rapidly dissociate to CO within a few nanoseconds.17f
Staudinger made a sustained effort to prepare other bisketenes, but was unsuccessful in efforts to obtain the linear α,ω-bisketenes, including the 1,2-bisketene 26.9b, 12a, 18a This was eventually observed in 1982 when it was formed by a double Wolff rearrangement of bis(diazo ketone) in an argon matrix at 10 K. The species was identified by the IR absorption at 2125 cm−1 and its conversion into the diester with methanol [Eq. (14)(14)].18b Surprisingly, such 1,2-bisketenes were shown by computational studies and confirmed by X-ray crystal-structure determination to have twisted, almost perpendicular conformations instead of the typical coplanar structures of 1,3-butadienes.18c,d
Among the distinctive characteristics of ketenes are their propensity for dimerization and their sensitivity towards hydration, but a notable exception is di-tert-butylketene (27
). This ketene, which was prepared in the laboratory of Newman (Figure 8) at Ohio State University in 1960,19a is protected by its bulky substituents. Another exception is trimethylsilylketene (28), discovered by Shchukovskaya, Pal'chik, and Lazarev in 1965 in Leniningrad (now St. Petersburg; Russia) (Figures 9 and 10).19b,c Neither of these ketenes is known to dimerize,19a,b and both are much less reactive towards water than CH2CO. The stabilization by the trimethylsilyl group in 28 is due to the β-silicon effect and the general tendency of ketenes to be stabilized by electropositive groups through electron donation from ketenyl CM σ bonds to the in-plane carbonyl π orbital.19d The first acylketene was EtO2CCEtCO, also discovered by Staudinger in 1909,19e which was only stable at −80 °C, while the crowded example 29, first prepared in 1978 in Leningrad, appeared to be stable indefinitely as a neat liquid.19f
Just as ketenes are stabilized by electropositive groups they are destabilized by electronegative groups. Thus halo ketenes are highly reactive, and despite early efforts directed toward their preparation,20a–c were unknown until 1965–1966, when their formation as unobserved intermediates was reported from three different laboratories.20d–f Dichloroketene (31) was prepared by the dehydrochlorination of dichloroacetyl chloride (30), and trapped with cyclopentadiene by [2+2] cycloaddition to form bicyclo[3.2.0]pent-2-ene-6-one 32 [Eq. (15)(15)].20d Hydrolysis of 32 is a simple and efficient method for the preparation of tropolone (33) [Eq. (15)].20d,g Brady of the University of North Texas (Figure 11)20e was a leading pioneer in the development of haloketene chemistry, including fluoro, chloro, and bromo derivatives.
Novel routes to ketenes were found by the reversal of cycloadditions, and these reactions have been extensively exploited in mechanistic studies and in the synthesis of complex molecules. Thus it was discovered by Smith (Figure 12) at the University of Minnesota that cyclobutenone 34, obtained from [2+2] cycloaddition of diphenylketene with phenylacetylene, undergoes thermal ring opening to the alkenylketene 35, which gives intramolecular [4+2] cycloaddition with the pendant phenyl group to form 36 (Smith–Hoehn reaction) [Eq. (16)(16)].21a Barton (Figure 13) and Quinkert discovered the facile photochemical ring opening of cyclohexadienones to give dienylketenes (Barton–Quinkert reaction).21b This process has even been observed with the parent 37, a tautomer of phenol generated at 77 K, to form 38 [Eq. (17)(17)].21c
The exquisite stereochemistry observed in the [2+2] cycloaddition of ketenes was one of the cornerstones of the development of the rules of orbital symmetry by Woodward and Hoffmann,22a and the investigation of these reactions was pursued in many mechanistic studies, particularly by Huisgen (Figure 14) in München (Germany).22b,c One of the first examples of the [2+2] cycloaddition of a ketene with a carbon–carbon double bond was the reaction of diphenylketene (6) with cyclopentadiene [Eq. (6)].9b Surprisingly it was found by Machiguchi (Figure 15) at Saitama University that the net [2+2] cycloaddition observed in this canonical process was not the initial reaction in the sequence, but that instead (as observed at low temperature) a [4+2] cycloaddition to carbonyl oxygen formed 39, which upon warming underwent Claisen rearrangement to 11 [Eq. (18)(18)].22d
Ketene chemistry in its first 100 years has been a microcosm of organic chemistry as a whole, with useful contributions in synthesis, theory, mechanisms, and practical applications. The major types of ketenes and their typical reactions as described above are now being increasingly applied in new and ingenious ways. Ketenes have always attracted many of the most talented individuals in chemistry, including many Nobel prize winners.23 Ketenes have found many industrial applications, ranging from the seemingly mundane manufacture of acetic acid and acetic anhydride, to the widespread use of ketene dimers from fatty acids as coatings for paper,24a the sophisticated application of the Süs photochemical Wolff rearrangement24b for the manufacture of the integrated circuits used in computer chips,24c,d and the use of ketenes in the preparation of drug candidates. The unique structural features of ketenes and their high reactivity ensure the vitality of these species in their second century.